Download Ultrafast Excited State Twisting Dynamics of Molecular Systems in

Document related concepts

Old quantum theory wikipedia , lookup

T-symmetry wikipedia , lookup

Circular dichroism wikipedia , lookup

Nuclear physics wikipedia , lookup

Transcript
Ultrafast Excited State Twisting Dynamics of
Molecular Systems in Condensed Phase
A Thesis Submitted
in Partial Fulfillment of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
by
SHAHNAWAZ RAFIQ RATHER
to the
DEPARTMENT OF CHEMISTRY
INDIAN INSTITUTE OF TECHNOLOGY KANPUR
KANPUR, INDIA
September, 2013
ii
iii
STATEMENT
I hereby declare that the work manifested in the thesis entitled “Ultrafast
Excited State Twisting Dynamics of Molecular Systems in Condensed Phase”
is the result of research carried out by me in the Department of Chemistry, Indian
Institute of Technology Kanpur, India under the supervision of Dr. Pratik Sen.
In keeping with general practice of reporting scientific observations, due
acknowledgements have been made whenever the work described is based on the
findings of other investigators.
September, 2013
IIT Kanpur
(Shahnawaz Rafiq Rather)
iv
vi
vii
DEPARTMENT OF CHEMISTRY
INDIAN INSTITUTE OF TECHNOLOGY KANPUR
CERTIFICATE OF COURSE WORK
This is to certify that Mr. Shahnawaz Rafiq Rather has satisfactorily completed
all the courses required for the Ph.D degree. The courses include:
CHM621
Chemical Binding
CHM629
Principles of Physical Chemistry
CHM664
Modern Physical Methods in Chemistry
CHM693
Synthesis of Advanced Materials
CHM636
Physical Photochemistry
CHM684
Computational Programming for Chemistry
CHM 799
Research
CHM800
General Seminar
CHM801
Graduate Seminar
Mr. Shahnawaz Rafiq Rather was admitted to the candidacy of the Ph.D degree
in November, 2010 after he successfully completed the written and oral qualifying
examinations.
Head
Department of Chemistry
Indian Institute of Technology Kanpur
Kanpur – 208016, India
Convenor, DPGC
Department of Chemistry
Indian Institute of Technology Kanpur
Kanpur – 208016, India
viii
ix
Dedicated to my Mom and Dad
x
xi
Acknowledgements
First and foremost, with due humble submissions I would like to thank my
Almighty, most Gracious and most Compassionate for blessing me with the attire
of a researcher and being always benevolent in guiding my life.
It is my sincere pleasure to express my heartfelt and overwhelming gratitude to
my thesis supervisor Dr. Pratik Sen for being an excellent mentor, a source of
unstinting encouragement, inspiration and motivation. Since the day I joined at IIT
Kanpur, he has not only been my research guide, but also a guardian with whom I
have never felt any reluctance to share things of concern from the perspectives of
research and many more. The field of ultrafast spectroscopy was a farfetched thing
at the time when I joined this laboratory, but in due course of time it all looked
clear and comprehensible, and this all happened because of the zeal and ardent
devotion of my supervisor towards all of we labmates. I am highly indebted to him,
as he shaped the most important part of my academic career by not only guiding
me but also giving me a complete freedom to express my thoughts and explore
various other domains. Whenever I was having any kind of difficulty especially in
analyzing the ultrafast data, he would come and sit by my side on a chair and then
we would sort out the problem together. To me, this journey of completion of
dissertation with my mentor has been highly invigorating, filled with lots of good
memories and will certainly help me in achieving success in future. My special
thanks are also due for Madam Manila Sen, who has occasionally prepared very
good dishes for us and made us feel at home. Lots of love is for their daughter
Suchismita, whom I wish a very lovely future ahead.
I offer my sincere thanks to Prof. S. Manogaran, Prof. S. Manoharan, Prof. D.
Goswami, Dr. M. Ranganathan for the highly informative and interesting courses
they taught me as a part of my course work. From their teachings I have learned a
lot, especially in the field of Quantum Chemisty. Special mention goes to Dr. N. N.
Nair for helping me in learning Gaussian software and various other fruitful
discussions. My thanks are also due for Prof. A. Chandra and Prof. K. Srihari
whose occasional lectures have been very inspiring.
I express my thanks to Prof. Kankan Bhattacharyya (IACS, Kolkata), Dr.
Sobhan Sen (JNU, New Delhi) for letting me use their lab facilities. I also thank
Prof. R. Gurunath for the collaborative work that is manifested in this thesis.
I extend my warm gratitude to the Department of Chemistry, IIT Kanpur for
making me a part of it, and its student community which has always made me felt
proud. Thanks are due to the former (Prof. V. Chandrashekhar and Prof. R. N.
Mukherjee) and current (Prof. P. K. Bharadwaj) Heads of the Department. Our
office staff especially Sudha madam and Geeta madam deserves a whelming
xii
acknowledgement for being very cooperative and helpful whenever needed. I
would also like to acknowledge the Departmental Chemical Society and
Departmental Post-Graduate Committee for being their member.
A sincere gratitude is due to the Department of Chemistry, University of
Kashmir from where I have pursued my Master’s program. I would like to thank
Prof. G. M. Rather, Prof. K. Z. Khan, Prof. G. M. Peerzada, Dr. Aijaz A. Dar, Dr.
Mohsin A. Bhat, Mr. Masood A. Rizvi and Dr. Altaf H. Pandit for being excellent
teachers and aspiring faculties who have wished always best for me and inspired
me to pursue research as my career.
I would like to acknowledge “Sheikh-ul-Alam Model School Chakura” from
where I have learned the humane morality and the value of education as a child. I
would also thank all the teachers from that place who have enriched me with
knowledge besides imbibing me with a good moral character.
My life inside the lab has always been an unforgettable journey alongside my
colleagues, who have been my best friends. Rajeev, Shradhey and me have been
together since the day first, we worked together, laughed together on funny
moments and enjoyed every bit of each other’s company especially during the lab
build-up phase. My other lab-mates Bhaswati, Gyanesh, Puspal, Vaisakh, Vipin
and Faizi have been very caring and loving and have made my stay a memorable
one. Due acknowledgments are to all the project students Shyamasish, Mainak,
Soumen, Arghav, Nirmal, Snigdha, Sharmishta, Barun, Shubrangshu, Sayani,
Ashish, Ankur, and Sunil who have worked in our lab since its inception.
Since the day I joined IIT Kanpur, my life has been blessed with the presence of
many caring and affectionate friends. I thank Ravi, Jhasaketan, Musheer, Chetan,
Rajeev, Shradhey, Akhilesh, Amit, Archana, Ruchi, Naziya, Chandrajeet, Keshav,
Chandan, Basanta, Prosenjit, Joydeb, Ankit, Dhiman, Arwind, Asif, Ashish, Akram,
Sameer, Rohan, Alok and Biswajit for making my stay very pleasant. My friends
from Kashmir Irshad, Tajamul, Jan, Hilal, Muntazir, Younis, Imtiyaz and Pervez
have always been a source of joy for me at IIT Kanpur. Friends from University of
Kashmir deserve a very warm acknowledgment for sharing two important years
with me and being with me afterwards. I thank Tariq, Umer, Mukhtar, Waseem,
Shahid, Arifa, Saima, Sabreena, Bilal, Haroon, Irfan, Ishtiyaq, Mushtaq, Zahid,
Mushtaq Ahmad Teli, Hamid, Afaq and Ajaz for being there whenever I needed
them and encouraging me for doing better and better. A special mention to my
dear friends at home, Jahangir, Younis, Nisar, Inam, Altaf, Pervez, Jawaid, Feroz,
and many more, who all have made me feel special.
I am very grateful to Council of Scientific and Industrial Research (CSIR),
India for providing fellowship and also to Research and Development section IIT
Kanpur for their best cooperation.
xiii
Devout thanks to my Mom and Dad, my siblings– Rafia, Shahbaz, Haqnawaz,
and sweet Saima, my dada and dadi, my uncle and aunt, my two lovely cousins–
Muneeb and Mueed and all my relatives for their never ending support, love,
encouragement, for entrusting me and wishing me all the best things in the world
and hence making this dissertation a success. I would like to specially thank my
dad, who has always been a source of tireless motivation and inspiration for me
since my childhood. He has always been a torchbearer in supporting me for my
cause and emphasized me on being a good human besides a good researcher.
From very early I have tried to emulate my dad’s zeal, determination, hard work
and humility. But, I feel still far from ideal. Finally, a very special thanks to Faiqul
Nisa for being a part of my life.
Shahnawaz Rafiq Rather
xiv
xv
SYNOPSIS
Name of the Student:
Shahnawaz Rafiq Rather
Roll Number:
Y9107081
Degree for which the Thesis is Submitted:
Ph.D
Department:
Chemistry
Thesis Supervisor:
Dr. Pratik Sen
Thesis Title:
Ultrafast Excited State Twisting
Dynamics of Molecular Systems in
Condensed Phase
September, 2013
Month and Year of Submission:
This thesis reports the nature of potential energy surface (PES) along the
geometrical coordinate(s) responsible for the excited state relaxation of molecular
systems in condensed phase, using femtosecond laser spectroscopy and theoretical
calculations. The excited state relaxation of several molecular systems involves
torsional motion of molecular fragments, and such twisting dynamics acts as the
major non-radiative decay channel. These torsional motions cause depletion of the
excited state population following a certain specific path defined by the PES. The
main objective was to understand the mechanistic pathways of electronic relaxation
of the photo-excited molecules, and to reveal the non-radiative channels rendering
the molecules highly non-fluorescent. An emphasis was given to understand the
influence of molecular structure in promoting a particular coordinate as a nonradiative relaxation channel. In this regard, the effect of twisting dynamics of
various molecular fragments was exclusively established for (S)-(−)-1-(4Nitrophenyl)-2-pyrrolidinemethanol and trans-4-dimethylamino-4-nitrostilbene.
Ultrafast twisting dynamics of auramine-O in compliance with temperature
dependent measurements was studied to reveal the presence of an activation barrier
in an otherwise thoroughly accepted barrierless excited state relaxation model of
auramine-O. Further, the knowledge of excited state relaxation mechanism of
model chromophore analogs was implemented to explain the highly fluorescent
xvi
nature of naturally occurring Green Fluorescent Protein. The ultrafast excited state
relaxation behaviour of a triphenylmethane dyes was contemplated and
subsequently used to estimate the viscosity of water in a nano-confined region.
Summary of the Work Done
(a) Excited State Relaxation Dynamics of 4-Nitrophenyl Pyrrolidinemethanol
The ultrafast excited state relaxation dynamics of a non-linear optically active
push-pull dye, (S)-(−)-1-(4-Nitrophenyl)-2-pyrrolidinemethanol (NPP), was carried
out under the regime of femtosecond fluorescence up-conversion measurements in
augmentation with quantum chemical calculations. Non-linear optical activity of
such type of systems has been ascribed to the extent of charge separation under the
optical effect. Due to the presence of a lone pair on the heteroatom of this
molecule, such systems can as well offer some degree of spin orbit coupling and
hence may change its photophysics drastically. The main motivation was thus to
trace the relaxation pathways which guide the depletion of first singlet excited
state, in such a way that it is virtually non-fluorescent. The femtosecond
fluorescence transients of NPP are best fitted by a bi-exponential function. The
first time component of few hundred femtoseconds was ascribed to the ultrafast
twisted intramolecular charge transfer (TICT). The occurrence of charge transfer is
substantiated by the large dipole moment change during excitation and also the
construction of intensity and area normalized time resolved emission spectra
(TRES and TRANES) of NPP in acetonitrile exhibited a two state emission on
behalf of decay of LE state and rise of CT state with a Stokes shift of 2000 cm-1
over a time scale of 1 ps. The second time component of few picoseconds is
attributed either to internal conversion (IC) or to intersystem crossing (ISC)
depending upon the polarity of the medium, with intersystem crossing dominating
over internal conversion in non-polar solvents and vice versa in polar solvents.
Ground and excited states (singlets and triplets) potential energy surfaces were
calculated as a function of –NO2 torsional coordinate, which revealed the
perpendicular orientation of –NO2 in the excited state minimum relative to the
xvii
planar ground state conformation and hence indicating the involvement of nitro
torsion as the main relaxation coordinate. The viscosity dependence of
fluorescence transients augments the proposition of considering –NO2 group
torsional motion as the main excited state relaxation coordinate.
(b) Dielectric Mediated Relaxation Dynamics of trans-4-Dimethylamino-4Nitrostilbene
Femtosecond fluorescence up-conversion technique was employed to
reinvestigate the intriguing dependence of fluorescence quantum yield of trans-4dimethylamino-4-nitrostilbene on dielectric properties of the media. In polar
solvents like methanol and acetonitrile, the bi-exponential fluorescence transients
were assigned to intramolecular charge transfer (ICT) dynamics and secondly to
the depletion of the ICT state to the ground state via internal conversion along the
torsional coordinate of nitro moiety. The viscosity independent decay dynamics
revealed the absence of any torsional coordinate involved during the charge
transfer process. In polarizable solvent the fluorescence transient are best fitted by
a three exponential function. The first time component was assigned to ICT
dynamics on a 2 picosecond time scale. Excited state intramolecular charge
transfer dynamics was established by constructing time resolved intensity and area
normalized emission spectra. Second time component was assigned to the
depletion of the ICT population along the torsional coordinate of double bond,
which decays directly to the ground state via a conical intersection or avoided
crossing at perpendicular configuration leading to the ground state trans- and cisisomers. This state which is having a dominant π-π* character may also decay by
intersystem crossing to the n-π* triplet manifold. Another pathway contributing to
the decay of ICT state involves the torsional coordinates of dimethylanilino and/or
nitrophenyl moieties. The motion about these torsional coordinates lead to the
formation of a conformationally relaxed state, which depletes back to the ground
state radiatively, and is responsible for the high quantum yield of DNS in
moderately polar solvents.
xviii
(c) Establishing the Presence of an Activation Barrier in an Otherwise
Barrierless PES of Auramine-O
A widely acclaimed model for the excited state relaxation dynamics of
auramine-O involves orientational relaxation of dimethylanilino moieties along the
barrierless excited state potential energy surface (PES). Such a model would
necessitate similar excited state dynamics in media offering similar viscous drag.
However, we have observed an interesting experimental observation showing
auramine-O to have ca. 8 times larger fluorescence quantum yield in chloroform
than in methanol, though both solvents have the same viscosity. The femtosecond
fluorescence transients of auramine-O in chloroform surprisingly depict a rise time
of ca. 200 fs which has not been observed before. This, along with the
simultaneous observation of unexpectedly large fluorescence lifetime and multiexponential transients (in chloroform) questions the thoroughly accepted
barrierless model of auramine-O relaxation dynamics, as the barrierless model
would demand a short lifetime and single-exponential decay. Temperature
dependent quantum yield measurements along with solvent dependent excited state
multi-coordinate calculations further unveil the exact nature of PES. All these
results concomitantly conclude that, at-least in chloroform, upon photo-excitation
auramine-O must pass over an activation barrier before dumping the excited state
population into ground state via a sink function through adiabatic coupling of the
electronic states.
(d) Ultrafast trans-cis Isomerization of GFP Chromophore Analogs and the
Effect of Protein Scaffold
Torsional motion mediated multi-coordinate relaxation pathway was
introduced to explain the excited state relaxation behavior of two analogs of Green
Fluorescent
Protein
(GFP)
chromophore,
(4Z)-4-(4-N,N-
Dimethylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPI)
and
(4Z)-4-(4-N,N-Diphenylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-
imidazolin-5-one (DPPI), using femtosecond fluorescence up-conversion and
xix
transient absorption spectroscopic techniques. The fluorescence transients are
found to be bi-exponential in nature. The first component with few picosecond
timescale has been assigned to the formation of the twisted intramolecular charge
transfer (TICT) from the directly excited Franck-Condon state along the torsional
coordinate of the donor N,N-disubstituted amine moiety. The existence of the
TICT state is authenticated from both the solvent polarity and viscosity
dependence dynamics. Completely non-fluorescent nature and absence of triplet
yield emphasizes that internal conversion must be the only possible channel for the
excited state relaxation back to the ground state. Viscosity dependence of the
second time component rules out the volume conserving simultaneous rotation of
the bridging bonds, referred to as “Hula twist”. The other probable non-radiative
decay channel is either a rotation about a single or a double bond. Time dependent
density functional theory calculations qualitatively witnessed an activated channel
along the twist coordinate of exocyclic double bond channelizing the charge
transfer state to the point of avoided conical intersection between S 1 and S0
electronic surfaces. Phenomenological kinetic relaxation scheme was formulated
based on the global analysis of transient absorption data of DPI and DPPI in
methanol, which confirms the existence of three transitory states; which has been
ascribed to locally excited state, charge transfer state, and conical intersection
existing between the excited and the ground state surfaces. This detailed study has
been extended to explain the influence of protein scaffold in supressing the nonradiative pathways and hence enabling GFP highly fluorescent.
(e) Viscosity of Water in a Nano-confined Environment Through the
Ultrafast Excited State Relaxation Dynamics of Malachite Green
Ultrafast dynamics study of malachite green (MG) was carried out to confirm
the excited state relaxation mechanism and subsequently to probe the microviscosity of water trapped in a nano-confined environment using AOT reversemicelle as a model system. Experimental results reveal a strong dependence of S 1
state relaxation dynamics of MG on solvent viscosity while a very weak
xx
dependence has been observed for the S2 state relaxation. The time dependent
density functional theory (TDDFT) calculations have been used to construct
potential energy surfaces of malachite green by pursuing an intramolecular rotation
along torsional coordinate of the phenyl rings. In augmentation with the
experimental observations, the computational results comprehend the existence of
a conical intersection along the S1 and S0 PESs, which leads to mixed vibrational
levels of S1 and S0 characteristics. The results suggest that the conical intersection
is along the torsional coordinate of N,N-dimethylamino substituted phenyl ring.
Correlating the observed dynamics of MG in confined system with the relaxation
time of MG in different glycerol-water mixtures, we assert the determination of
micro-viscosity of water inside the AOT reversed micellar system. The data
confers that the micro-viscosity of water in AOT water pool of w0 = 2 (9 cP) is
almost 9 times higher than the bulk water. As we increase the w0 from 2 to 40, the
micro-viscosity decreases monotonically to 5.68 cP and the decrease is observed to
be exponential.
xxi
TABLE OF CONTENTS
Chapter 1
1.1
1.2
1.3
1.4
Chapter 2
2.1
2.2
Motivation
Thesis Outline
Photophysical Processes: An Overview
Introduction
Absorption of Light and Subsequent Formation of the
Excited State
Fate of the Excited State and various Other Perspective
1.3.1 An Energy Surface Description of Photophysical
Processes
1.3.2 The Born-Oppenheimer ApproximationQuantum
Mechanical Perspectives
1.3.3 Perturbation Theory and Photophysical Processes
1.3.4 Vibronic Coupling
1.3.5 Radiative TransitionsOrbital Configuration Mixing
and Multiplicity Mixing
1.3.6 Spinorbit Mixing: A Mechanism for Inducing
Spin Changes
1.3.7 SpinOrbit Coupling and Radiative Transitions
1.3.8 Twisted Intramolecular Charge Transfer in
the Excited State
1.3.9 Photophysical Radiationless Processes
1.3.10 Conical Intersections
1.3.11 Rate of Vibrationless Relaxation
1.3.12 Internal Conversion and Relationship to the Excited
State Structure
1.3.13 The Energy Gap Law
1.3.14 The “Loose Bolt” and “Free Rotor” Effects
1.3.15 Intersystem Crossing and Selection Rules
Twisting Dynamics and Ultrafast SpectroscopyPrologue
References
Experimental Methods
Steady State Measurements
Femtosecond Fluorescence Up-conversion Measurements
2.2.1 Excitation of the Sample with SHG Pulse
2.2.2 Collection of the Fluorescence and Delay of the
xxv
xxix
142
3
3
6
8
10
12
15
16
18
21
22
23
26
28
29
30
31
33
35
39
43─78
45
45
47
xxii
2.3
2.4
2.5
2.6
Chapter 3
3.1
3.2
3.3
3.4
Chapter 4
4.1
4.2
Gate Pulse
2.2.3 Signal Acquisition and Data Analysis
2.2.4 Deconvolution Procedure
Time Resolved Emission Spectra
Femtosecond Transient Absorption Measurements
2.4.1 Pulse Duration, Time Resolution and
Spectral Sensitivity
2.4.2 Transient Absorption Setup
Quantum Mechanical Calculations
2.5.1 ExchangeCorrelation Functionals
2.5.2 Basis Set
68
2.5.3 Solvent Effects
Synthesis of Model GFP Chromophore Analogs
2.6.1 General Aspects
2.6.2 General Procedure for Preparation of Oxazolones
2.6.2.1
Results
2.6.3 General Procedure for Preparation of DPI and DPPI
2.6.3.1
Results
References
Excited State Relaxation Dynamics of 4-Nitrophenyl
Pyrrolidinemethanol
Introduction
Results
3.2.1 Steady State Absorption and Emission
Measurements
3.2.2 Femtosecond Fluorescence Up-conversion Study
3.2.3 Quantum Chemical Studies of Ground and
Excited States
Discussions
Conclusion
References
Dielectric Mediated Relaxation Dynamics of trans-4Dimethylamino-4-Nitrostilbene
Introduction
Results
4.2.1 Steady State Fluorescence Measurements
49
51
52
54
57
60
61
65
67
69
69
69
70
71
71
72
74
79─100
81
83
83
86
87
88
96
97
101─128
103
105
105
xxiii
4.3
4.4
Chapter 5
5.1
5.2
5.3
Chapter 6
6.1
6.2
6.3
4.2.2 Viscosity Dependent Femtosecond
Fluorescence Transients
108
4.2.3 Fluorescence Transients in Highly Polar Solvents
109
4.2.4 Fluorescence Transients in Slightly Polar Solvents
110
4.2.5 Polarity Dependent Behaviour of Fluorescence Transients in
AcetonitrileCarbon tetrachloride Mixture
112
4.2.6 Quantum Chemical Calculations
113
Discussion
113
Conclusion
123
References
124
Establishing the Presence of an Activation Barrier in an
Otherwise Barrierless PES of Auramine-O
129─150
Introduction
Results and Discussion
5.2.1 Steady State and Time Resolved Fluorescence
Measurements
5.2.2 Temperature Dependent Quantum Yield
Measurements
5.2.3 Quantum Mechanical Calculations
Conclusion
References
131
132
132
139
142
148
149
Ultrafast trans-cis Isomerization of GFP Chromophore Analogs
and the Effect of Protein Scaffold
151─188
Introduction
Results
6.2.1 Steady State Absorption and Emission Measurements
6.2.2 Femtosecond Fluorescence Up-conversion Study
6.2.2.1 Viscosity Dependent Fluorescence
Kinetics of DPI and DPPI
6.2.2.2 Polarity Dependent Fluorescence Kinetics of
DPI and DPPI
6.2.3 Quantum Mechanical Calculations
6.2.4 Transient Absorption Measurements
6.2.4.1
DPI in Methanol
6.2.4.2
DPPI in Methanol
Discussion
153
154
154
157
157
160
161
163
163
164
166
xxiv
6.5
Chapter 7
7.1
7.2
7.3
7.4
6.3.1 Fluorescence Up-conversion and Quantum
Mechanical Study
6.3.2 Transient Absorption Study
6.3.3 Global Analysis of the Transient Absorption Data
Conclusion
References
166
172
179
183
184
Viscosity of Water in a Nano-confined Environment through the
Ultrafast Excited State Relaxation Dynamics
of Malachite Green
189-212
Introduction
191
Computational Method
195
Results and Discussion
195
7.3.1 Steady State Absorption and Emission
Measurements of Glycerol-Water Mixtures
195
7.3.2 Femtosecond Fluorescence Up-conversion Study
197
7.3.3 Quantum Chemical Calculations
200
7.3.4 Microviscosity of Water in the Nano-pool of
AOT Reverse Micelle
202
Conclusion
207
References
208
Concluding Remarks and Future Outlook
213─214
List of Publications
215─216
xxv
MOTIVATION
In a somewhat philosophical vein, it is sometimes stated that the determination
of the potential energy surface (PES) is the goal of all experimental work. In this
sense, the parameters; equilibrium and transition state geometries, vibrational
frequencies, and barrier heights are mere aspects of the PES and could be estimated
easily. The ground and excited electronic state properties of a molecular system
ultimately depends on the shape of the PES. Question is, once a molecule is
promoted to an excited state by absorption of a photon, what decides the fate of the
excited state in space and in time? When does an excited state decay radiatively or
how an excited state is rendered non-radiative? How the excite state processes like
conformational changes, charge transfer, isomerisation, etc. control the dynamics
of the excited state?
In last two decades, femtosecond laser spectroscopy has revealed an
appreciable amount of molecular insights into the primary steps of ultrafast
processes occurring in the excited states.1-5 Ultrafast laser spectroscopy has been
profound in exploring the dynamical processes occurring during the excited state
relaxation of molecular systems through the interaction of the system with the
environment, which may involve exchange of energy and momentum. Such a
response is not instantaneous, rather occurs in some finite time, which is very
crucial in ascertaining the complex excited state relaxation dynamics of molecular
systems in condensed phase.2 Alongside the advancement in experiments,
theoretical calculations have proven very facilitative in revealing the approximate
nature of electronic states of interest.6-9 Once complete relaxation mechanism of
the molecular system is established, using chemical judgement one can combine
the information to produce a satisfactory and accommodative geometrical
coordinate and hence the PES. Relaxation studies of photo-excited systems might
contribute to a better insight in a wide range of phenomena of fundamental
importance. These phenomena include electronic charge redistribution, 10-12
electron and proton transfer, large amplitude intramolecular torsional motions,
xxvi
trans-cis isomerization, vibrational relaxation, internal conversion, intersystem
crossing, etc.10-23
References
1. Duxbury, D. F. Chem. Rev. 1993, 93, 381.
2. Glasbeek, M.; Zhang, H. Chem. Rev. 2004, 104, 1929.
3. Polívka, T.; Sundström, V. Chem. Rev. 2004, 104, 2021.
4. Zhao, G. -J.; Han, K. -L. Acc. Chem. Res. 2012, 45, 404.
5. Amdursky, N.; Erez, Y.; Huppert, D. Acc. Chem. Res. 2012, 45, 1548.
6. Dreuw, A.; Head-Gordon, M. Chem. Rev. 2005, 105, 4009.
7. Jödicke, C. J.; Luthi, H. P. J. Am. Chem. Soc. 2003, 125, 252.
8. Perun, S.; Sobolewski, A. L.; Domcke, W. J. Am. Chem. Soc. 2005, 127,
6257.
9. Liu, F.; Morokuma, K. J. Am. Chem. Soc. 2012, 134, 4864.
10. Barabara, P. F.; Jarzeba, W. Adv. Photochem. 1990, 15, 1.
11. Yeh, A. T.; Shank, C. V.; McCusker, J. K. Science 2000, 289, 935.
12. Changenet-Barret, P.; Espagne, A.; Katsonis, N.; Charier, S.; Baudin, J. –B.;
Jullien, L.; Plaza, P. Martin, M. M. Chem. Phys. Lett. 2002, 365, 285.
13. Tolbert, I. M.; Solnstev, K. M. Acc. Chem. Res. 2002, 35, 19.
14. Hewitt, J. T.; Vallett, P. J.; Damrauer, N. H. J. Phys. Chem. A 2012, 116,
11536.
15. Kang, Y. K.; Duncan, T. V.; Therien, M. J. J. Phys. Chem. B 2007, 111,
6829.
16. Piechowska, J.; Huttunen, K.; Wrobel, Z.; Lemmetyinen, H.; Tkachenko, N.
V.; Gryko, D. T. J. Phys. Chem. A 2012, 116, 9614.
17. Rettig, W.; Maus, M. Conformational changes accompanying intramolecular
excited-state electron transfer. In Conformational Analysis of Molecules in
Excited States; Waluk, J., Ed.; Wiley-VCH: New York, 2000.
18. Rini, M.; Kummrow, A.; Dreyer, J.; Nibbering, E. T. J.; Elsaessser, T.
Faraday Discuss. 2003, 122, 27.
xxvii
19. Liu, R. S. H. Photochem. Photobiol. 2002, 76, 580.
20. Baskin, J. S.; Yu, H. Z.; Zewail, A. H. J. Phys. Chem. A 2002, 106, 9837.
21. Yu, H. Z.; Baskin, J. S.; Zewail, A. H. J. Phys. Chem. A 2002, 106, 9845.
22. Rini, M.; Holm, H. K.; Nibbering, E. T. J.; Fidder, H. J. Am. Chem. Soc.
2003, 125, 3028.
23. Fainberg, B. D.; Huppert, D. Adv. Chem. Phys. 1999, 107, 191.
xxviii
xxix
THESIS OUTLINE
The thesis is divided into following chapters
Chapter 1.
This chapter describes in detail the basic photophysical processes a
molecule can undergo, upon interaction with light. At the end a
prologue
is
mentioned
in
brief,
the
relevance
of
twisting/configurational changes and the ultrafast time resolved
spectroscopy.
Chapter 2.
This chapter throws light on the various experimental techniques
which have been used to understand the excited state dynamics of the
molecular systems in condensed phase.
Chapter 3.
This chapter describes our work on the role of twisting motion in the
excited state relaxation behaviour of a hetero-atom containing nonlinear optically active push-pull system.
Chapter 4.
In this chapter, we have described the solvent polarity dependent
excited state relaxation dynamics of a stilbene based push-pull
system. The role of twisting motion of various molecular fragments,
in deciding the excited state relaxation behaviour, has been discussed
in detail.
Chapter 5.
This chapter reports the spectroscopic evidence of the presence of an
activation barrier in the otherwise barrierless excited state potential
energy surface of a highly non-fluorescent dye-Auramine-O.
Chapter 6.
This chapter enunciates the role of protein scaffold in restraining the
torsional degrees of freedom of the green fluorescent protein
chromophore, by studying the excited state relaxation behaviour of
two synthetic analogs of the GFP chromophore.
Chapter 7.
This chapter deals with the determination of viscosity of water in a
nano-confined environment through the ultrafast excited State
Relaxation Dynamics of Malachite Green.
xxx
Chapter 1
Photophysical Processes: An Overview
2
Chapter 1
Chapter 1
3
1.1. Introduction
“Molecular photochemistry is a science concerned with description of physical
and chemical processes induced by the absorption of photons, in terms of a
concrete mechanistic model based on molecular structures and their implied
properties”, as quoted by Nicholas J. Turro in his book “Modern Molecular
Photochemistry”.1 In all photophysical and photochemical processes, activation of
the molecular system is mediated by absorption of a photon of suitable energy.
This process of absorption can promote the molecule into an electronically excited
state, when the available energy is in the UV-vis region and is in resonance with
the energy difference between the ground and the excited state. A molecule in a
specific electronic state may exist with various nuclear configurations, and each
configuration in space corresponds to a particular potential energy of the system.
Such a map of potential energy against nuclear configurations represents potential
energy surface for a given electronic state. The notion of potential energy surfaces
can be used to merge the realm of molecular structure, energetics and dynamics.
1.2. Absorption of Light and Subsequent Formation of the Excited State
Quantum mechanics replaces the classical precise position of nuclei in space
and the associated motion by the concept of nuclear or vibrational wave function,
. This function is not as restrictive in confining the nuclear configuration to the
regions of space as defined by the classical potential energy curves. Classically, for
a transition to occur, molecule should have similar nuclear configuration and
similar momentum in the two electronic states at the instant of transition. However,
quantum mechanics relies on the net positive overlap between the vibrational wave
function of the two electronic states undergoing transition, which is given by the
Franck-Condon overlap integral
χ i χ f . The probability of the electronic
transition between two electronic states is given by the square of vibrational
overlap integral, called as Franck-Condon factor,
i  f
2
. The larger the
difference between the two quantum numbers among which the transition is
4
Chapter 1
occurring, the more likely it is that the nuclear configuration and the momentum of
the two states will be different, and less probable will be the transition, which is
exactly the result anticipated by the Franck-Condon (FC) principle. An illustration
(Figure 1.1) shows the quantum mechanical basis of FC principle for absorption
process between the two electronic states. The most probable transition will be the
Figure 1.1. Quantum mechanical interpretation of the Franck-Condon principle.
one, which furnishes maximum overlap between the two vibrational states, which
in this case is from v = 0 of  0 to v = 3 of  * . Other transitions will occur, but
with less probability. There are various implications of FC principle depending on
the relative position of equilibrium nuclear configuration in the excited electronic
state. It must be borne in mind that, due to the different electronic distribution in
the excited state compared to that of the ground state, a different Morse potential is
required to represent the excited state. So the equilibrium position of the nuclei in
excited state may be different than the ground state nuclear configuration. In figure
1.2(a), a situation is shown in which the two potential energy curves are similar for
the two electronic states and are not displaced horizontally, which implies the
equilibrium nuclear separation is same for  0 and  * .Under such circumstances,
according to the Franck-Condon principle, relatively strong v = 0  v = 0
transition is observed for both absorption and emission and the 0  0 band of the
absorption and emission spectra overlap significantly. Consequently, the transition
Chapter 1
5
Figure 1.2. Potential energy curves of ground ( 0 ) and excited electronic ( * ) state
showing the Franck-Condon allowed transitions for (a) same equilibrium nuclear
configuration of two electronic states, (b) excited state equilibrium nuclear configuration
shifted to the right side of ground state nuclear configuration. (c) Absorption to a point above
the asymptote on *.
to higher vibrational quantum number is very weak. Figure 1.2(b) represents a
situation in which the equilibrium nuclear configuration of excited state potential
energy curve is significantly displaced relative to the minimum in ground state
curve. Such configuration of electronic states enables transitions v = 0  v = 2 or 3
to be more intense than rest of the transitions on either side of the maximum.
Anthracene is an example wherein the excited-state PES is slightly displaced from
the ground state because of slight bending in excited state to make a “v” shape
symmetry about 9,10 positions of the molecule.1,2
In figure 1.2(c), absorption of the photon takes the molecule to a higher
vibrational quantum number above the asymptote, so that the molecule no longer
feels the restoring force and proceeds towards dissociation, in case it is diatomic,
or in polyatomic molecule it can cause depletion of the excited state from some
higher vibrational level, wherein the nuclear configuration is far from the
equilibrium configuration of the excited state. The absorption spectra of such
molecules will be devoid of any vibrational structure, instead will exhibit
continuity.
Chapter 1
6
In condensed phase, since the rate of vibrational relaxation is rapid, emission
generally occurs from the v = 0 vibrational level of the lowest electronic excited
state to some higher vibrational quantum number of the ground state. Analogous to
absorption, under the regime of Franck-Condon principle, the most intense will be
the one that occurs vertically, maintaining the same nuclear configuration of the
excited and ground state and having maximum Franck-Condon factor. In general,
the equilibrium nuclear separation of the ground state potential energy curve is
smaller than that of excited state curve (since excited state possesses electron in
anti-bonding orbital), thus the vertical transition from excited state produces an
elongated ground state immediately after transition, while as absorption produces a
compressed excited state.
The absorption of UV-vis light by a molecule promotes an electron from an
initially occupied, low energy orbital (ground state) to a high energy, previously
unoccupied orbital (excited state). The excited electronic states can have two spin
configurations. In one state, the electron spins are paired (opposite spins) and the
state carries zero resultant spin magnetic moment and in other state the electron
spins are unpaired (parallel) with a finite net spin magnetic moment. In presence of
a magnetic field the state with paired spins remains a single state and is termed a
singlet state, while as the state with unpaired spin splits into three quantized states,
and is hence termed as triplet state.
1.3. Fate of the Excited State and Various Other Perspectives
The electron once promoted, won’t stay in the excited electronic state (A*),
rather it will return back to the ground state, which can be accomplished by many
means as shown by the schematic representation in figure 1.3. The energized
excited state can relax back to ground state by radiative decay known as emission
or by a radiationless process. It may also undergo a chemical reaction to form some
product state. The first two phenomena are photophysical process and can be
envisaged in a more illuminating way by means of Jablonski diagram (Figure 1.4),
Chapter 1
7
Figure 1.3. A global paradigm for reactions, a molecule is feasible to undergo, once
promoted to excited state by absorption of a photon of light.
while as the later come under the paradigm of photochemical processes. In this
thesis, we will be only concerned with the photophysics of the molecules.
Photophysical processes may be defined as transition which interconvert
excited states with each other or excited states with the ground state and are
classified into two groups, radiative and radiationless processes.1 The possible
Figure 1.4. Jablonski diagram. SSinglet states; TTriplet states; AAbsorption;
FFluorescence; PPhosphorescence; ICInternal Conversion; ISCIntersystem
Crossing; VRVibrational Relaxation.
radiative transitions between the various electronic states include;
1. The spin-allowed singletsinglet absorption of photons (S0 + h  S1).
2. The spin-forbidden singlettriplet absorption of photons (S0 + h  T1).
Chapter 1
8
3. The spin-allowed singletsinglet emission of photons, (S1  S0 + h), called
fluorescence.
4. The spin-forbidden triplet-singlet emission of photons, (T1  S0 + h), called
phosphorescence.
While as the radiationless processes between electronic states can be of following
types;
5. The spin-allowed transition between states of same multiplicity (S1  S0 +
heat), called internal conversion.
6. The spin-forbidden transition between states of different multiplicity (S 1  T1
+ heat) called intersystem crossing.
7. The spin-forbidden radiationless transition between the triplet state and the
ground state, (T1  S0 + heat), also called as intersystem crossing.
To understand which of these possible processes are most probable, we desire
information on the relative rates of all the possible photophysical processes which
compete for deactivation of the concerned state. One can estimate the rate of these
reactions from experimentation or through computation and these relative rates can
determine the probability of these possible processes that can occur from the state
of interest.
1.3.1. An Energy Surface Description of Photophysical Processes
For a complete analysis of these competing photophysical processes, it is
important to keep track of a number of structures, energies and dynamics, which
can be nicely handled by the paradigms associated with potential energy surfaces
(PES). PES displays the potential energy of the molecular system against the
varying molecular structure of the system. The motion of a molecule over a
particular potential energy surface can be visualised in terms of motion of a
representative point corresponding to each nuclear configuration along the reaction
coordinate. In this way, one can envision photophysical processes in terms of
motion of a representative point on a PES. The important topological features of a
Chapter 1
9
hypothetical ground and excited state PES are the maxima and minima on each
surface, the nuclear geometries for which the surfaces are far apart in energy, the
relative disposition of the maxima and minima to each other, and the geometries
for which two surfaces come close to one another in energy. The important features
of maxima and minima of the ground and excited state surfaces can be summarized
below.
Spectroscopic (FranckCondon) Minima. The interaction of a molecule in
ground state minimum with a photon may promote it to the excited state surface
minimum or the relaxation of the molecule may occur from excited state minimum
to the ground state minimum, such minima are called as “spectroscopic” or
“Franck-Condon” minima. It has been observed that radiative jumps occur with the
highest probability between surfaces for which there is a minimum and also similar
nuclear configuration in both the excited and ground state surfaces. Intervention of
small barrier on the surface may be overcome on behalf of thermal energy from the
collisions. If the barrier height is high, then the representative point will be trapped
in the excited state minimum until the representative point returns to ground state
by either emission of a photon or radiationless deactivation.
SurfaceCrossing Minima. The minimum in the excited state PES may undergo
crossing with the ground state PES and such events are called surface crossings.
Depending on electronic interactions in the ground and excited state, the crossing
may be weakly avoided or strongly avoided. If the representative point approaches
a weakly avoided crossing, a very fast transition to the ground state will occur and
a funnel is said to exist on the excited state surface. Barriers that exist in the
ground state surface can also be visualised as having an approximate surface
crossing origin apriori to the weak electronic interactions. Such barriers will also
lead to the surface crossing of excited state with the ground state. For the weak
avoided crossing, the barrier needs to be large and strongly avoided crossing will
be possible in case of small barrier in the ground state surface.
Chapter 1
10
1.3.2. The
Born-Oppenheimer ApproximationQuantum
Mechanical
Perspectives
The Born-Oppenheimer (BO) approximation proposes that the low mass,
rapidly moving, negatively charged electrons can immediately adjust their
distribution relative to the positively charged, heavily massive, and slowly moving
nuclei. As per this approximation, the motion of electrons in orbitals is much more
rapid than nuclear vibrational motions.3 This allows the motion of electrons and
nuclei to be treated separately and independently. This salient feature simplifies the
solution of Schrödinger equation ( Hˆ   E ) as it permits the calculations of
electronic wave function to be computed for any nuclear configuration. The energy
obtained from such a calculation is “Potential Energy” in the sense that it
corresponds to total electronic energy without any contribution of kinetic energy of
the nuclei. In principle, the nuclei can be moved to all possible geometries and
henceforth the electronic energy can be computed for each such configuration of
the nuclei. This procedure will help in finding the equilibrium position of nuclear
coordinates and thus providing the most stable configuration of the nuclei. It is
possible within the Born-Oppenheimer approximation to compute the energies of
ground or excited electronic states against a particular nuclear coordinate of
interest.
For ground state, the net electron spin being zero, spin is not involved in the
solution of Schrödinger equation, which is certainly not the case with excited
states, as spin is critical in determining the overall photophysical pathways an
excited state is going to adopt. The BO approximation allows an approximate wave
function of the molecular system () to be computed in terms of three independent
wave functions;
 ~0  S
(1.1)
The wave function  0 represents an approximate electronic wave function for the
electron position and electron orbital motion in space about the nuclei. The wave
function , represents an approximate vibrational wave function and S represents
an approximate spin wave function. Thus in a “zero order” approximation, which
Chapter 1
11
assigns purity to the electronic states, the molecular wave function is a product of
three independent approximate wave functions. A significant interaction between
the electrons and the vibrations (called vibronic coupling) or between the spins and
the orbiting electrons (spinorbit coupling) cause break down of the BO
approximation.
One of the postulates of quantum mechanics is that the expectation value of any
observable molecular property (P) can be obtained in terms of a matrix element
given below
Pavg

 Pˆ 
(1.2)
where P̂ is the quantum mechanical operator of the observable property, P. This
property can be the energy of an electronic state, the dipole moment of the state,
the transition probability between the states, etc. Here we attempt to attribute a
qualitative meaning to this matrix element from the perspectives of photophysics
of the molecule. The magnitude of the matrix element provides basis for estimating
energy of the electronic states and for proposing selection rules for the probability
of transition between the desired electronic states.4,5 The approximate BO wave
functions are employed to compute matrix elements. Thus if the wave function of
the system is given as per equation 1.2 then the matrix element would be;
Pavg
~
 0  S Pˆ  0  S
(1.3)
Since the electronic wave function,  0 is usually approximated as a product of one
electron molecular orbitals, i
0
~ 12 n
(1.4)
The approximate value of Pavg is given by the matrix element;
Pavg
~
(12 n )  S Pˆ (12 n )  S
(1.5)
This level of approximation is essentially called as “zeroorder approximation”, as
we are still within the limits defined by Born-Oppenheimer approximation, i.e., the
different parts of the overall wave function are treated separately and independent
of each other. However zero-order approximation does not hold good for the
Chapter 1
12
situations, wherein mixing of the oneelectron wave functions takes place and
electronelectron interactions are introduced. The mixing of wave functions is
essentially a mathematical process that proposes a better approximation of the
wave functions in firstorder than there was in zeroorder and the resultant wave
function may represent the system to a better extent.
1.3.3. Perturbation Theory and Photophysical Processes
The photophysical processes that a molecule in general can undergo are the
radiative and radiationless transitions between electronic states. These processes
may involve transitions between states of same or different symmetry (n*,
*) and may also involve transitions between states of same or different
multiplicity (singlet, triplet). By knowing the proper operator for the rate of
transition between the electronic states and the wave functions of the electronic
states, one can calculate the rate of transition and hence the probability of the
transition between the desired states using quantum mechanics as shown below;
k for  1 to  2 transition
P12
~
 1 Pˆ12  2
2
(1.6)
For example, the matrix element corresponding to the rate of transition between
ground and excited state involves the wave function of the ground and excited state
and P represents the appropriate operator corresponding to the interaction of
ground state with the electromagnetic field. More generally, the interaction of
operator with the ground state “distorts” the ground state wave function  1 and if
this makes  1 look like  2 , a transition between the two states can be triggered.
Wave mechanically, one can consider the influence of operator in terms of mixing
of the wave function and effective mixing of the two states occur under condition
of resonance between the waves  1 and  2 . This notion of mixing provides an
efficient quantum mechanical intuition for visualisation of all possible
photophysical transitions. The qualitative evaluation of the matrix elements
provide useful selection rules for such transitions, which serve as a guide for the
probability of a given transition among various possibilities.
Chapter 1
13
Perturbation theory is a mathematical method for using weak perturbations to
carry out mixing of wave functions of the system in a manner so as to achieve
better approximation of the true system. Considering a zero order approximate
wave function reasonably close to true wave function of the system, perturbation
theory can be employed to distort the approximate wave function to look more like
the true one. Weak perturbations are often considered to be responsible for
triggering the transitions between the electronic states. In essence, the perturbation
operator P1→2 interacts with the initial state ( 1 ) and enables mixing with the final
state ( 2 ) creating resonance between the two states, which can be expressed by
the following equation;
 1  Pˆ12   1   2   2
(1.7)
where  is the measure of the amount of  2 that is mixed into  1 as a result of
perturbation, and is called as mixing coefficient. The more the magnitude of
mixing coefficient, more probable will be the transition between states. Its
magnitude is directly related to the strength of perturbation and inversely related to
the separation between the states, and is given by,
   1 Pˆ12  2 / E12
(1.8)
If the transition between states is fully allowed, then the rate of such a transition is
limited by zero point electronic motion under conditions of constant nuclear and
spin configurations. However, in case the nuclear or spin configurations change
during the transition, the rate will be limited by such changes and not by the time it
takes to make a zero point motion. Thus when a rate is slower than the maximal
rate, the structure and dynamics of electronic, vibrational and spin wave functions
may serve as bottlenecks in determining the rate of transitions between the desired
states. So the actual observed rate of a transition can be written in terms maximum
0
possible rate constant ( k max
) and the product of prohibition factors for the
electronic (fe), vibrational (fv) and spin (fs) aspects of the transitions.
0
k obs  k max
 fe  fv  fs
(1.9)
Chapter 1
14
0
In most of the circumstances, kobs is much smaller than kmax
. The rate of the
transition between zero order states whose purity is eliminated by perturbation is
given by Fermi’s golden rule;6
k obs
~

  1 P̂12  2

2
(1.10)
The term,  refers to the density of states that are capable of effectively mixing the
two states  1 and  2 . This rule provides a basis of selection rules in transitions
that are triggered by weak interactions, e.g. if the value of the matrix element in the
above equation is zero, then the transition is rigorously forbidden. For majority of
the cases, the most important perturbation for mixing electronic wave functions is
vibrational nuclear motion coupled to orbital motion of the electrons, called as
vibronic coupling. Under such case, the matrix element that vibrationally mixes
two states  1 and  2 is given by  1 P̂vib  2 . If only the vibrational motion is
responsible for the mixing of two states, then the observed rate of transition will be
proportional to the square of vibrational overlap integral 1  2
2
called as Franck-
Condon factor.
k obs
0
 k max
 1 P̂vib  2
 
E122

2

 


1
2
2

(1.11)
If the transition between states does not involve any change in spin of the electrons
then there will be no contribution of prohibition by the spin configuration i.e., fs =
1. However, under cases wherein the transition involves change of spin, the
perturbation operator has to couple the states of opposite spins. The important
perturbation available to make a pair of parallel spins look like a pair of antiparallel spins is the coupling of the electronic spin motion with the electron orbital
motion termed as spin-orbit coupling. The matrix element to take care of such
coupling is  1 P̂so  2 with P̂so as spin-orbit coupling operator. The modified rate
of transition is given below;
k obs
0
 k max
 1 P̂vib  2

E122

2
   1 P̂so  2

E122
 
2


  1  2

2

(1.12)
Chapter 1
15
1.3.4. Vibronic Coupling
Zero-order description of electronic states arises from the Born-Oppenheimer
approximation of assuming frozen and non-vibrating nuclear geometry. Replacing
the concept of pure vibrationless molecules with a vibrating molecule will certainly
modify the zero-order description of electronic states. The states of a vibrating
molecule are called as vibronic states rather than pure electronic states. The
vibrations of a molecule serve as a source of weak perturbation on the approximate
wave function and cause the mixing of the states, which eventually will promote a
transition between vibronic states of interest. The energy of the perturbation and
the rate of transition from  1 and  2 under such conditions are given below;
Ev
k obs

~
 1 P̂vib  2
E12
2
  1 P̂vib  2
(1.13)
2
(1.14)
From perturbation theory, if E12 is large, the mixing of the states is small and vice
versa. If vibronic mixing energy (Ev) is small compared to E12, the mixing of the
states is expected to be small. The value of vibronic mixing energy is in general of
the order of 1 ‒ 10 kcal mol-1. Vibronic interactions in case of energy separation,
E12 > 50 kcal mol-1, do not significantly mix electronic states. Consequently, the
vibrations of the ground state do not mix with the vibrations of the excited states
during absorption process. On the contrary, vibrational interactions are very
effective in mixing electronic excited states on behalf of less energy separation in
the order of few kilocalories per mole between them. Thus vibrational motion of
the appropriate type can be effective in triggering mixing of the zero-order states
and such effects are of importance in determining the rate of transitions from
excited states. One must remember that, every vibration is not capable of causing
mixing of the states. For example, in case of methyl group as a radical, anion or
carbonium ion, during an in-plane vibration, the perpendicular p- orbital retains its
pure p- character and also since the system remains planar during the in-plane
bending vibration, the energy of the orbital is not expected to change and
16
Chapter 1
conclusively the vibronic coupling will be very weak. On the contrary, an out-ofplane vibration (umbrella-flipping) breaks the symmetry of the molecule and
causes a change in the hybridization of the central carbon atom. The pure p- orbital
changes shape in response to more electron density is one side of the plane relative
to the other side. Due to this momentary re-hybridization, the pure p- orbital begins
to take on some s- character. This sp- mixing decreases the energy of the resulting
orbital compared to the pure p- orbital. The matrix element for vibronic mixing
will have a different magnitude for in-plane and out-of-plane vibration. For inplane vibration,  1 P̂ip  2 equals 0, and for out-of-plane vibration  1 P̂op  2 has
a finite value. The vibronic mixing matrix thus provides a first-order mechanism
for the transition from one vibronic state to another, even though the electronic
transition is strictly forbidden in the zero-order approximation.
In quantum mechanical terms, Franck-Condon principle states that the most
probable transition between electronic states occur when the wave function of the
initial vibrational state (  1 ) most closely resembles the wave function of the final
state (  2 ). Analogous to orbital overlap integral  1  2 , one can define the extent
of overlap between vibrational wave functions 1  2 to define the degree of the
overlap integral of the two vibrational wave functions  1 and  2 . The larger the
value of overlap integral, the more probable is the vibronic transition.
1.3.5. Radiative Transitions–Orbital Configuration Mixing and Multiplicity
Mixing
For radiative transitions,2,3,7,8 first order mixing is generally a significant
mechanism for allowing processes that are rigorously forbidden in zero-order
approximation. A state described in terms of a single electron orbital configuration
(or spin multiplicity) may acquire characteristics of second electronic orbital
configuration (or spin multiplicity) and hence undergo mixing to promote
transitions which are otherwise forbidden in zero-order approximation. An
absorptive transition from n to * (or emissive transition from * to n) with the
Chapter 1
17
assumption that there are no electron-electron or electron-vibrational coupling, will
render them zero-order or pure states. The oscillator strength f, which is related to
the probability of the transition between the states, will be zero for such states
under zero-order approximation, as the two orbitals are strictly orthogonal and
hence no orbital overlap is present. Thus we can write n Pˆ  *  0 , where P̂ is the
operator for rate of transition. Under first order approximation, the vibrations or
electron-electron interactions may cause mixing of the n-* and -*
configurations and consequently n and * orbitals are no longer rigorously
orthogonal, which will render the transition rate matrix element non-zero,
n Pˆ  *  0 . Thus, the transition which is completely forbidden in zero-order
approximation is now partially allowed under the first order approximation
triggered by a vibronic coupling. The description of the initial excited electronic
state S1 (n-*) is now given by;
S1 (mixed )  n   *  (   *)
(1.15)
where  is the measure of the extent of mixing. Since the S0 → ‒* transition is
allowed, so the contribution of ‒* (S2) into the rigorously forbidden transition
n‒* (S1) is rendered partially allowed, although the oscillator strength will be low
compared to a fully allowed transition. The oscillator strength for such a case is
given by;
f (S 0  S1 )  2 f (S 0  S 2 )
(1.16)
In other words, the first order S0 → S1 (mixed) transitions are allowed only to the
extent that S2 (‒*) is mixed into S1 (n‒*) and the amount of mixing is given by
, which can be predicted from perturbation theory as;

 a P̂  b
(1.17)
E a  Eb
Replacing  in eq. 1.16 yields;
f ( S 0  S1 ) 
n   * P̂    *
E  *  E n  *
2
f ( S 0     *)
(1.18)
Chapter 1
18
The theory predicts that the transition from S0 → S1 will posses all the
characteristics properties of S0 → S2 transition, except that the probability of
transition will be limited by a factor 2.
The same idea of mixing can be applied to the qualitative evaluation of a spinforbidden transitions i.e., singlet → triplet. The triplet n‒* and singlet ‒*
mixing is dominant for a transition involving excitation from S0 (n2) → T1 (n‒*)
and it is the spin‒orbit coupling that mixes singlet and triplet states. The oscillator
strength of such a transition is given by;
3
f ( S 0  T1 ) 
( n   *) P̂SO 1 (    *)
E  *  E n  *
2
f ( S 0     *)
(1.19)
The illustration of vibronic mixing and spin orbit coupling to render rigorously
forbidden transitions partially allowed is shown in figure 1.5.
1.3.6. Spin–Orbit Coupling: A Mechanism for Inducing Spin Changes
As per perturbation theory,1-3,9,10 an interaction that either rephrases or flips
electron spin will cause the pure initial excited electronic state to evolve in time to
an oscillating mixture of singlet and triplet states. For the interaction to be effective
in causing resonance, there must be an interaction that causes a finite mixing of the
electronic wave functions, and also the energy of separation of the two states must
be on the order of interaction energy. The interaction is facilitated by means of
coupling of angular momenta associated with electron spin and with orbital
motion.
In zero-order approximation, the matrix element for the interaction energy via
spin-orbit coupling has the form,  1 Hˆ SO  2 , where HSO is the operator for spinorbit coupling,  1 and  2 are the initial and final orbitals involved. The spin-orbit
coupling can be conveniently visualised in terms of interaction between the
magnetic moment (S) due to electron’s spin angular momentum (S) and magnetic
moment (L) due to electron’s orbital angular momentum (L). The strength of
Chapter 1
19
Figure 1.5. The transition from zero-order to first-order involves turning on a mechanism for
mixing of states. Top: Vibronic coupling renders S0 → n‒* weakly permissible, which is
otherwise strictly forbidden in zero-order approximation. Bottom: spin-orbit coupling
renders singlet t triplet transition partially allowed in fist-order approximation.
coupling depends on the relative orientation of the magnetic moments, their
magnitudes and separation in space. The operator HSO has the form;
H SO   SOSL ~  SO  S  L
(1.20)
where SO is the spin orbit coupling constant and is related to the nuclear charge
that an electron feels as it revolves around the nucleus involved. The spin orbit
coupling energy is given by the following matrix element;
ESO   1 H SO  2   1  SOSL  2 ~  1  SO  S  L  2
(1.21)
In molecules, spin-orbit coupling is a local effect and is effective when the electron
is located near the heavy atom. Intersystem crossing requires spin change of a
system from +1/2 → -1/2 or -1/2 → +1/2, which would cause a change in total
angular momentum of the system by one unit (ħ) and hence overriding the
20
Chapter 1
conservation of angular momentum principle. So in order to conserve the angular
momentum, there should be a simultaneous change in angular momentum
somewhere else. The p- orbital coupled to the electron spin can produce exactly
one unit of orbital angular momentum (ħ) by undergoing a rotation of 90 about an
arbitrary axis (z- axis), and overlapping with an adjacent p- orbital, i.e., interconversion between px and py. The requirement of changing orbital angular
momentum is reflected in the form of HSO operator. For effective change of the
orbital angular momentum, the energy of the two inter-converting px and py orbitals
should be almost same. The mixing of spin states generates orbital angular
momentum due to a px → py jump, which is required to couple with the spin
angular momentum change with the orbital angular momentum change. The
rotation about an axis corresponds physically to the jump of an electron from a px
orbital to an empty py orbital and this rotation generates one unit of angular
momentum about the z-axis. The orbital angular momentum possesses an
associated magnetic moment, which can couple with the spin magnetic moment
and induces spin reorientation or spin rephasing. This coupling of the magnetic
moments means that S serves as a source of torque that tends to twist the porbital and makes it to rotate by 90 about the z-axis. Reciprocally, we can put it in
a way that L serves as a torque that tends to change the spin vector from  →  or
 →  orientation. This generation of torque by coupling the angular momenta is
effective in triggering spin changes in transitions when there is significant matrix
element capable of coupling the initial and final electronic states.
In reality, the electrons in molecules move around in space about a framework
of positively charged nucleus. In case of one electron hydrogen like atom, the
strength of the spin-orbit coupling parameter SO is proportional to the fourth
power of atomic number (Z4). The electron in an orbital is not placed at a fixed
distance from the nucleus, rather for some of the time the electron finds itself more
close to the nucleus. During that time, under the influence of stronger nuclear
attraction the electron accelerates to very high speeds. The magnetic field
Chapter 1
21
generated by a negatively charged particle is proportional to its velocity and if the
particle is accelerated almost to the limit of relativistic velocity, the associated
orbital angular momentum will be very large and hence the spin-orbit coupling
parameter will be large. As a result there is a heavy atom effect on the rate of spinorbit induced transitions. The illustration of consequences of spin-orbit coupling is
shown below in figure 1.6.
Figure 1.6. Oversimplified model of an electron in a p-orbital undergoing spin flip. The top
panel shows how the spin magnetic moment changes due to spin-orbit coupling and the
bottom panel shows the vector model of the change in precession of the electron upon
coupling of sin and orbital angular momentum.
1.3.7. Spin–Orbit Coupling and Radiative Transitions
The oscillator strength or emission rate of spin-forbidden transitions is
significantly smaller than the spin allowed transitions, which is due to the very low
probability of the occurrence of spin-orbit coupling between the desired states.
Experimentally, it has also been found that the oscillator strength for spinforbidden radiative transition S0(n2)  T(n–*) is greater than S0(2)  T(–*)
transition, which is the opposite situation of S0  Sn transitions wherein f (–*)
>f (n–*). Before delving into the reason, we must know that oscillator strength (f)
has three contributions, electronic factor (fe), vibrational factor (fv) and spin factor
(fS). For singlet–singlet transitions, where spin factor has no contribution, we know
Chapter 1
22
that fefv (–*) > fefv (n–*) because  (–*) >  (n–*), where  is the molar
extinction coefficient. Thus for spin-forbidden radiative transitions, fS (n–*) >> fS
(–*).
This situation can easily be understood by considering the example of
formaldehyde for S0(n2)  T(n–*) and ethylene for S0(2)  T(–*) spinforbidden radiative transitions. The spin change in formaldehyde is due to an n–*
transition, which may be visualised as a jump of an electron from a p- orbital
localized on oxygen (say, px) in the plane of the molecule to a p- orbital (say, py)
perpendicular to the plane of the molecule. The simultaneous px  py orbital jump
is thus a one-center jump involving an orbital angular momentum change. This
type of situation is thus precisely what is required for generating orbital angular
momentum and favors strong spin-orbit coupling. For ethylene, one can easily
observe that for planar ground state, there is no orbital of low energy in the
molecular plane into which  electron can jump. Consequently, there are no onecenter spin orbit interactions to help spin flip and hence no spin-orbit coupling
exists in ethylene. This implies that the magnitude of the matrix element of spinorbit coupling for n2  n–* is greater than 2  –* transition. Thus we can
write;
f [S0 (n 2 )  T(n   *)]  f [S0 ( 2 )  T(   *)]
(1.22)
This represents a generalized situation for occurrence of spin-orbit coupling and is
called as El-Sayed’s rule.11,12
1.3.8. Twisted Intramolecular Charge Transfer in the Excited State
Transfer of charge from a donor moiety to an acceptor moiety coupled via a conjugated spacer has always been important from the photophysical and
photochemical perspectives. They offer very different excited state properties and
have found their implications in many fields like photosynthesis, solar energy
conversion, photonics, optoelectronics, etc.13-17 Due to the conceived applications
of such type of molecular structures, knowledge of electronic structure is cardinal
Chapter 1
23
to the understanding of their properties and also the ability to tune the electronic
structure of such systems through custom design. The most important prototype to
such class of molecular systems is N, N–dimethylaminobenzonitrile (DMABN).
According to widely accepted model by Grabowski and coworkers17, in polar
solvents, upon photo-excitation DMABN undergoes a twisted intramolecular
charge transfer (TICT) from the donor dimethylamine group to the acceptor cyano
group leading to the observation of dual emission bands in the fluorescence
spectra. It was proposed that the excited state may be stabilised by the twisting of
the dimethylamino group which facilitates transfer of charge from donor to
acceptor group. The two bands were ascribed to the emission from the locally
excited state and the TICT state. However, one must remember, it is not necessary
that every twisted charge transfer event leads to the observation of dual
fluorescence emission. We can observe dual emission in cases when both the
locally excited state and the TICT state decay radiatively. On the contrary, if either
of the two states is non-fluorescent, then even though there is a charge transfer in
the excited state, we will still only get a single emission band corresponding to
whichever state is fluorescent. Most of the cases, it has been observed that, if the
charge transfer process involves a torsional coordinate, then the TICT state formed,
will be non-fluorescent, because the torsional coordinate will provide an important
channel for the non-radiative decay of the molecule, which has been established
profoundly in many classes of molecules.13-17
1.3.9. Photophysical Radiationless Processes
Radiationless transitions between electronic states are a form a electronic
relaxation in which electronic energy is converted into the kinetic energy
associated with nuclear vibrational motion.3 Such transitions include two types; (1)
radiationless transitions between electronically excited states and (2) radiationless
transitions between lowest excited state of a given multiplicity and the ground
state, as can be comprehended from figure 1.7.
24
Chapter 1
Figure 1.7. Radiationless photo-physical processes of interest.
Quantum mechanics like classical mechanics treats the problem of radiationless
processes in terms of motion of a representative point on the energy surface. This
representative point is the instantaneous nuclear configuration of a molecule in the
ground and the excited electronic states and will make a radiationless jump from
one electronic surface to another at certain critical nuclear configurations (rc) The
Born-Oppenheimer approximation allows the solution of electronic wave equation
to be formulated in terms of motion of nuclei only, because for each nuclear
geometry it is assumed that there is only one electronic distribution for the ground
and the excited state. This process of computing electronic distribution is called as
adiabatic approximation and the surfaces thus generated are known as adiabatic
surfaces. The nature of interactions between the various electronic states of a
molecule apart from the selection rules depends exclusively upon the shape of the
potential energy surfaces of these electronic states.
Figure 1.8 provides a convenient framework for understanding potential energy
surface relationships and radiationless transitions between them. A perfect zeroorder surfaces for which the two wave functions cross at critical nuclear
configuration but do not mix at all. Even if the energy of the two states at rc is
identical, the representative point which starts on excited state surface ( 2 )
maintains its electronic characteristics as it crosses  1 through the point rc; that is
the two states do not mix. Under such circumstances the probability of
radiationless transition from one surface to another is one (Figure 1.8(a)). Figure
Chapter 1
25
Figure 1.8. Examples of surface topologies for a two-surface system. (a) “perfect” crossing
for which the transition is strictly forbidden, (b) and (c) “Weakly” and “Strongly” avoided
crossings for which transitions are possible near critical nuclear configuration, rc. (d)
“Matching” surfaces for which transitions are very improbable. (e) “Touching” surfaces for
which the surfaces have similar energies once rc has reached.
1.8(b) represents a situation where in the two surfaces constitute a weakly avoided
situation and hence the surface crossing is highly probable, thus rendering the
probability of radiationless transitions close to one. The importance of such a
crossing is that in some trajectories, the wave function will continue to look like
 2 after passing through rc and in others the wave function will change to  1 .
Strongly avoided crossing at rc is represented by figure 1.8(c). Herein the
representative point is expected to reach a certain equilibrium configuration and to
undergo oscillatory motion like a harmonic oscillator over the potential energy
surface. A jump from the upper excited state may occur near this minimum,
Chapter 1
26
however, the jump is expected to occur on a slower rate because of the large
energy gap between the surfaces. The matching surfaces (Figure 1.8(d)) differs
from the situation in figure 1.8(c) in terms of zero order linkage between the upper
and lower surfaces which is absent in former case. This means, there is no dynamic
link between  2 and  1 which can couple them at the purely electronic level. The
final situation which can exist between two states is shown in figure 1.8(e) and is
termed as “Touching” surfaces, wherein the surfaces approach each other
asymptotically and at critical nuclear configuration they become close in energy
and essentially touch each other. Such a situation is typical of processes like
breaking of a sigma bond or rotation about a - bond.
Essentially the zero order states  2 and  1 are capable of undergoing resonance
near critical nuclear configuration. Once resonance occurs between states, the
electronic motion is no longer adequately described alone by the zero order wave
functions, which implies the adiabatic Born-Oppenheimer approximation is no
longer accurate, as the nuclear motion and the representative point are no longer
unanimously controlled by a single surface  2 and  1 , rather are now defined by a
mixed state [1  2]
Intial state  1
  1  2    1 or  2
Final state
(1.23)
1.3.10. Conical Intersections
The energy surfaces corresponding to two electronic states of exactly same
symmetry do not cross but avoid each other at the critical nuclear geometry. Such a
formulation is called as “noncrossing rule”. Essentially, for an adiabatic surface,
two electronic states having same symmetry, energy and geometry may undergo
some quantum mechanical mixing of the wave functions of the excited and ground
state to produce two adiabatic surfaces. The result will be the formation of two
states with one state having higher energy and the other state of lower energy
causing surface avoiding of the states. This rule applies strictly to molecules
possessing high symmetry along the reaction coordinate at critical nuclear
Chapter 1
27
configuration and such a high symmetry is possible only in diatomic molecules.
Noncrossing rule is not a strict selection rule for polyatomic molecules possessing
a little or no symmetry with respect to the reaction coordinate at the critical
geometry. As a result of which the two electronic states of same formal symmetry
may actually cross each other. This feature of electronic states was postulated long
ago,18,19 but there was very little experimental or computational evidence as to
whether surfaces actually cross or not. It was profoundly established especially by
means of computational analysis of energy surfaces, which convincingly showed
that intersections of two potential energy surfaces are quite common. 20-25 Based on
the computational analysis of surface crossings, it was emphasized, that in three
dimensions, the energy surfaces have the form of a double cone in the immediate
vicinity of the touching point. At the touching point of these two cones, the wave
functions for the two surfaces are degenerate and this double cone (Figure 1.9) is
termed as a conical intersection (CI).
An important feature of conical intersections is that, it serves as a very efficient
funnel and takes the representative point on the excited state surface rapidly
towards the ground state surface. The driving coordinate of the motion of
representative point over the excited state surface is essentially a vibrational
Figure 1.9. “Double Cone” representation of conical intersection.
Chapter 1
28
relaxation of the system. One should remember that the conical intersection, in
general, may not be the rate limiting step in a radiationless pathway between states
of same multiplicity. Once the representative point enters the region of
intersection, the passage towards the ground state may occur with unit probability
and hence the rate of the relaxation through a conical intersection is determined
mainly by the slope at which the representative point passes from Franck-Condon
point to the CI. If barriers intervenes the excited state potential energy surface, then
the rate of the radiationless decay will be controlled by the height of the barriers
itself.
At the crossing point of two surfaces i.e., conical intersection, the energy gap
between them is essentially zero and hence the classical probability for the
transition between the two states through conical intersection is unity. Under the
circumstances of no restriction from spin prohibition on the motion of
representative point, the rate of transition from the excited state surface to the
ground state surface will be controlled by the time scale of intramolecular
vibrational relaxation (IVR) which ranges approximately from 100 fs – 10 ps for
most of the organic molecules. It is hereby important to make a distinction between
the conical intersection (CI) and the avoided surface crossing (ASC) between two
electronic states. In ASC the excited state minimum corresponds to a thermally
equilibrated critical nuclear geometry, and a representative point undergoes
oscillations with a characteristic zero point motion. On the contrary, the motion of
a representative point through a conical intersection to the ground state occurs very
rapidly so that there is not enough time for thermal equilibration of the system.
1.3.11. Rate of Vibrational Relaxation
The molecule upon encounter with a photon is promoted to some specific
vibrational level of the excited electronic state in accordance with the FranckCondon principle. The immediate response of the molecule in upper electronic
state is to render the excess vibrational energy to exchange with the surroundings
and this process of energy dissipation is called as vibrational relaxation. For most
Chapter 1
29
of the organic molecules this process occurs with rate constants typically of 1012 –
1014 s-1. In essence, the vibrations within a molecule and the vibrations of the
molecules in the environment have the ability to rapidly accept the energy of
nuclear motion of the molecule and convert it into many degrees of vibrational
motion. Since the number of vibrational, rotational and translational energy levels
of the solvent environment of the excited molecule are essentially continuous, such
energy levels may serve as a classical heat bath, and hence can take up any amount
of vibrational energy that the excited molecule needs to dispose off.
26-29
The
general qualitative observations of vibrational relaxation infer that intramolecular
vibrational relaxation is comparatively faster than that of intermolecular vibrations.
Also, the vibrational relaxation generally is not a rate limiting step in the electronic
relaxation of the excited state. It is also generally observed that the rate of
vibrational relaxation is slightly faster in excited state than in ground state.
1.3.12. Internal Conversion and Relationship to the Excited-State Structure
Figure 1.7 represents a few processes which constitute the relaxation of excited
state into ground state via internal conversion, which can be defined as the
radiationless electronic relaxation between states of same spin multiplicity, and
hence occurs between states among singlets and among triplets. Under conditions
of very low temperatures like 77 K or in a very rigid matrix, processes like
photoreactions, bimolecular quenching, etc. may be completely avoided, while as
the fluorescence and phosphorescence processes are readily observed with
significant quantum yields. In addition, the confinement provided by the rigid
matrix inhibits radiationless transitions that require loose bolts and free rotors, and
low temperature restricts radiationless processes that require passage over energy
barriers. Considering such factors, following simple deductions have been
generalized for rigid aromatic hydrocarbons.30 Fluorescence generally occurs from
only from S1 to S0; Phosphorescence occurs mainly from T1 to S0; Sn and Tn
emissions are extremely rare. The quantum yield of fluorescence and
phosphorescence are independent of excitation wavelength. The sum of the
Chapter 1
30
quantum yields for fluorescence and phosphorescence is approximately equal to
unity. These all observations are essentially a deduction of very fast internal
conversion process from upper excited states to S1 and or T1 and also due to very
fast vibrational relaxation within an electronically excited state. On the contrary,
internal conversion from S1 to S0 is much slower, because of energy gap law and
poor Franck-Condon factors, and molecules that are relatively rigid due to
structures of environment cannot compete with fluorescence and intersystem
crossing. As an example we can choose anthracene for the estimation of rate of
internal conversion from upper excited singlet states. Even though the fluorescence
from upper singlet states is not observed experimentally we can still calculate the
rate of relaxation from such states. For example, the S0 to S3 absorption of
anthracene has maximum oscillator strength at 252 nm with an extinction
coefficient of ca. 2 x 105 M-1 cm-1. Using the equations;1 the rate of fluorescence is
S
k F0 ~ 103  max ~ 2  109 s 1 and using  F2  k F0 / k D  10 4 one can calculate the value of
deactivation rate constant to be equal to k D  10 4 k F ~ 2 1013 s 1 . This implies, that
S3 deactivates to S1 with a rate constant on the order of intramolecular vibrational
relaxation. Even though one excites anthracene to S3 state, the emission from S1
state is observed exclusively, which concludes that the internal conversion rate
constant of Sn to S1 is of the order of 10-13 s-1. Evidently, electronic relaxation by
internal conversion from upper electronic states is rate limited only by nuclear
vibrational motion and not by electronic relaxation. This in turn suggests that zeroorder crossings are common in Sn states (n>1) and that the critical nuclear
geometries may be readily achieved during vibrational motion of S n in its zero
vibrational quantum number.
1.3.13. The Energy Gap Law
Internal conversion between S1 and S0 electronic states, not undergoing zero
order surface crossings must occur via “Franck-Condon forbidden” mechanism,
which means the nuclei in one of the electronic state must undergo a rather drastic
Chapter 1
31
change in both position and momentum as a result of the transition, since the net
overlap of the vibrational wave functions in the two states is small. Such situations
are characterized with internal conversion being limited by the Franck-Condon
overlap factor, fv. Considering 1013 s-1 as the approximate order of maximum rate
of internal conversion, then one can write;
k IC ~ 1013 f v
(1.24)
One can estimate fv from spectral data.31 It is both evidenced by experiments and
theoretical calculations that the overlap factor is a very sensitive function of the
energy gap, E, which is the difference between the zero point vibrational energy
levels of the states undergoing internal conversion.32,33 An expression for rate of
internal conversion is deduced as;
k IC ~ 1013 exp( E )
(1.25)
The energy gap law can thus be attributed to the change in the Franck-Condon
factors, which become increasingly unfavourable with increasing energy separation
between the states as can be predicted from above equation.
Experimentally it has been observed that for rigid molecular systems, internal
conversion process is usually negligible in comparison to the fluorescence or
intersystem crossing if the energy gap (E) is larger than 50 kcal mol-1 unless there
is some vibrational motion that is favourable for inducing the radiationless
transition (i.e., a free rotor or loose bolt in the molecule).
1.3.14. The “Loose Bolt” and “Free Rotor” Effects
Certain vibrations in a particular electronic state have a possibility to act as a
promoter of radiationless transitions, if they carry both the representative nuclear
configuration on a surface, near critical nuclear configuration, and also help to
trigger the radiationless transition by facilitating an appropriate vibronic
interaction. In order to conserve the energy changes during the electronic
relaxation, some vibrations must act as an acceptor of the energy difference
between electronic states. If a vibration can act as both the promoter of the
32
Chapter 1
radiationless transition and also the acceptor the excess energy corresponding to
the difference of the two electronic states, such a vibration will be very effective in
causing radiationless relaxation of the excited electronic state.34,35 Stretching of a C
─ C sigma bond and the twisting of a C ═ C  bond represent two exemplar
promoter and acceptor vibrations in organic molecules. Both types of vibrations at
the extreme nuclear configurations cause surface touching as shown in figure 1.10
below. These two mentioned vibrations may escort a representative point on an
Figure 1.10. Schematic representation of (a) the stretching of a  bond and (b) the twisting of
a  bond.
excited electronic surface to a region near critical nuclear configuration and if this
vibration also causes vibronic mixing of the excited and the ground state, then the
radiationless transition between the states of interest is highly probable. The
stretching vibration is analogous to a “loose bolt”, and is set in motion by other
moving part of the molecule and there by taking the excess kinetic energy
produced during the motion.34 This excess energy may lead to complete
dissociation of the bond, implies that the promoter stretching vibration facilitates
an irreversible separation of the atoms constituting the vibration.
The twisting of a double bond is analogous to the “free rotor” which accepts the
excess energy and begins to rotate. This situation is responsible for trans-cis
isomerisation. These types of motions are the primary source of radiationless
transitions in the molecules and enable the molecule to undergo internal
Chapter 1
33
conversions. Restricting such type of motions will prevent the occurrence of
radiationless processes and increase the probability of radiative transition. For
example, at room temperature trans-stilbene is weakly fluorescent (F = 0.05) and
cis-stilbene is non-fluorescent (F = 0.00). On the contrary, under the similar
experimental conditions the structurally constrained derivatives of above two
molecules are highly fluorescent. (F = 1.0).36,37
1.3.15. Intersystem Crossing and Selection Rules
As discussed earlier in section 1.3.6, intersystem crossing (ISC) is facilitated by
either “spin flip” or “spin rephrasing”. In zero-order approximation, based on the
presumed separation of electronic and spin motions, there will be no intersystem
crossing. In other words, if a molecule is in an initial singlet or in a triplet state, it
will stay there forever; irrespective of, even there is a crossing between the singlet
and the triplet state. The presence of spin-orbit coupling in the system will enable
mixing of the states near critical nuclear configuration at the crossing point and
consequently intersystem crossing will occur. The important prerequisite is that a
significant magnetic field capable of changing the spin multiplicity must exist,
besides, the interaction between the spin and orbital magnetic moments must
operate effectively in the timescale when the representative point is in the region
near critical nuclear geometry. Also, near critical nuclear configuration, the orbital
transition involved must possess the character of a px  py orbital jump to generate
orbital angular momentum. Further, the orbital transition should be predominantly
localized on a single atom.
Carbonyl group provides a qualitative example for understanding spin-orbit
coupling, wherein the singlet state may be derived from an n–* or a –*
configuration. The system will provide four possible intersystem crossings from
S1, depending on the electronic configuration of starting and final states as shown
below.
Chapter 1
34
1
n–* 
3
–*
(1.26)
1
n–* 
3
n–*
(1.27)
1
–* 
3
n–*
(1.28)
1
–* 
3
–*
(1.29)
The atomic orbital representation of the first type of intersystem crossing (1.26) is
shown in figure 1.11(a), wherein S1 state has unpaired electrons in n and *
orbitals located on the same oxygen atom and a pair of electrons in low lying 
orbital. During ISC an electron jumps from  orbital to the n orbital and hence
accounting for the conservation of angular momentum during the spin-flip. Hence
for the transition of the type given by equation 1.26, the ISC is allowed because of
effective spin orbit coupling. The ISC involving orbitals shown in equation 1.27
does not involve an orbital change (Figure 1.11(b)). Thus neither the spinflip in
Figure 1.11. Qualitative orbital description of the basis for the allowed and forbidden
intersystem crossings. (a) and (c) and allowed, while as (b) and (d) are forbidden.
Chapter 1
35
n orbital nor the spinflip in * orbital will induce any accompanying change in
orbital angular momentum and hence the spinorbit coupling is rigorously
forbidden. Same theory can be applied to elucidate the occurrence and prohibition
of intersystem crossing in system with orbital configurations of the type shown in
equation 1.28 and 1.29 respectively. The orbital representation of these two
conditions is demonstrated in figure 1.11(c) and 1.11(d). The following selection
rules (i.e., El-Sayed’s rules) can thus be deduced for S1 → T1 intersystem crossing
S1(n–*)

T1(–*)
Allowed
S1(n–*)

T1(n–*)
Forbidden
S1(–*)

T1(n–*)
Allowed
S1(–*)

T1(–*)
Forbidden
in zero order. Extending the same level of selection rules to the spin-orbit coupling
in S0 → T1 transitions, one conforms to the following deductions;
T1(n–*)

S0(n2)
Allowed
T1(–*)

S0(2)
Forbidden
Although spin-orbit coupling in zeroorder occurs strictly according to the
selection rules mentioned above. However, some perturbations may induce finite
amount of spinorbit coupling to render the transition partially allowed. A
perturbation can be in the form of outofplane vibration which may cause some
weak mixing of the singlet and triplet states and will subsequently trigger
spinorbit coupling.
1.4. Twisting Dynamics and Ultrafast Spectroscopy ─ Prologue
Mostly, optical properties of a molecule are strongly influenced by the medium.
In addition to prevalence of static interactions between the molecule and the
environment, the interactions may also lead to dynamic effects via exchange of
energy and momentum. Of the important dynamical effects is the response of
solvent molecules to the photo-induced changes in the charge distribution of
36
Chapter 1
chromophore molecules. The finite response time of the environmental response
helps in determining the excited state relaxation dynamics of the molecule itself.
Intramolecular torsional motions is one of the important relaxation mechanism of
certain chromophores in addition to many other dynamical processes occurring in
them.38-54 The typical timescale of these dynamical relaxation processes is on the
order of tens of femtoseconds to hundreds of picoseconds. One can essentially
conclude that to a large extent the knowledge of these dynamical processes have
only been possible because of the advent of ultrafast laser pulses.38 Ultrafast time
resolved laser spectroscopy is one of the most powerful tools for the direct
elucidation of the relaxation processes.39-41 The pulsed excitation of the molecular
system determines the start of the photophysical processes and the subsequent
monitoring is pursued by various techniques like femtosecond fluorescence upconversion spectroscopy, femtosecond transient absorption spectroscopy, etc.
In certain molecular systems intramolecular configurational (twisting) modes
occasionally in combination with solvation modes, control the excited state
relaxation dynamics. The solvent has a profound influence on the rate of
intramolecular configurational changes especially in terms of collisional friction,
which largely affects the isomerisation or twisting dynamics. More importantly,
the functional group twisting is most of the time related to intramolecular charge
transfer and hence the concept twisted intramolecular charge transfer (TICT)
dynamics within the molecule. Such type of conformational changes are of
importance to numerous processes in chemistry and biology. 42,43 For such systems,
it has been proposed that very high relaxation rates can be obtained in case the
potential energy surfaces along a reaction coordinate lacks an activation barrier.4246
Solvent viscosity will then influence the excited state dynamics to a great extent.
This has been extensively investigated for photoexcited di- and tri- phenylmethane
dye molecules showing twisting dynamics of phenyl rings.44-46 Auramine-O is
another dye which is weakly fluorescent in less viscous solvents and highly
fluorescent in viscous solvents, DNA, and polymeric acids and was concluded that
the diffusion controlled twisting motion of the phenyl rings are the main
Chapter 1
37
coordinates of excited state relaxation.47-50 The non-fluorescent nature of the
molecule in low viscous solvents was ascribed to the fast diffusion of the phenyl
rings, which will be restricted in more viscous solvents. Time resolved upconversion and transient absorption spectroscopic studies have further confirmed
the mechanism of excited state relaxation.47-50
Time resolve fluorescence experiments have also been reported for another dye,
Michler’s Ketone, an analogue of auramine-O. It was proposed that twisting of the
phenyl rings is of dominant influence on both the dynamics of the Stokes shift and
on the radiative decay.51 It was also proposed for certain TICT molecules that an
enhancement in the dynamics of the Stokes shift may occur by considering a two
dimensional approach with twisting motion and solvation, as the main relaxation
processes.52-54 There can be several limiting cases, wherein solvent relaxation is
either slow or fast than the intramolecular torsional motion. In the slow solvent
relaxation, proposed to be operational for Michler’s Ketone the reaction is initiated
by solvation, and soon taken over by twisting which determines the subsequent
dynamics of the excited state.52-54
Chromophores have been found useful in initiating signal transduction in a
photoreceptor via absorption of a photon and undergoing ultrafast intramolecular
conformational changes, like cis-trans isomerisation, which may alter the changes
in the environment. At a later stage, this leads to the appearance of the signalling
state and eventually the biological response. Such phenomena have been studied
exclusively for rhodopsins,55-57 bacteriorhodopsins,58 and phytochromes,59 etc.
Coumaric acid chromophore in proteins (styryl dye chromophore) is a
photoreceptor in wild type photoactive yellow protein (PYP) and many other
protein systems,60 for which fast transient fluorescence behaviour has been
ascribed to the intramolecular conformational changes and not to the solvation. 60-62
The molecular fragments follow a motion along the barrierless excited state
potential energy surface leading to trans-cis isomerisation. The relaxation model of
PYP chromophore was proposed to be similar to that of retinal in
bacteriorhodopsin.60-62 The model proposes, after pulsed excitation of the
38
Chapter 1
chromophore, the Franck-Condon state decays very rapidly (<50 fs) into an almost
flat excited state potential. The molecule in the excited state is now in a reactive
region typical of a twisted chromophore, where its potential energy surface shows
an avoided crossing with that of S0, following which the molecule may twist
further and decay non-radiatively back into the trans- state or twist into the cisconformation. Due to the in-homogeneity in the surroundings of the chromophore
in the protein pocket, a spread in the twist angle of the excited state at the crossing
region exists, which will consequently induce in-homogeneity in the excited state
lifetime of the emissive state. This inhomogeneous distribution of the lifetimes
thus explains the multi-exponential fluorescence decays of the native PYP.60-62
Thus twisting dynamics is an important coordinate providing a trajectory for the
excited state relaxation of molecular systems in condensed phase. The following
chapters will first throw light on the ultrafast spectroscopic techniques used for
studying such conformational dynamics of molecular systems. Exclusive role of
twisting dynamics in deciding the fate of excited states and the subsequent
photophysics will be discussed thoroughly throughout this thesis.
Chapter 1
39
References
1. Turro, N. J. Modern Molecular Spectroscopy, The Benjamin/Cummings
Pub. Company: US, 1978.
2. McGlynn, S. P.; Azumi, T.; Kinoshita, M. Molecular Spectroscopy of the
Triplet State, Prentice Hall, Englewood Cliffs, NJ, 1969.
3. Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular
Photochemistry of Organic Molecules. University Science Books, Sausalito,
California, 2010.
4. Atkins, P. W.; Friedman, R. Molecular Quantum Mechanics, 5th Edition,
Oxford University Press, Oxford, UK, 2005.
5. Klessinger, M.; Michl, J. Excited States and Photochemistry of Organic
Molecules, VCH Publishers, New York, 1995.
6. Atkins, P. W. Molecular Quantum Mechanics, Oxford University Press,
Oxford, UK, 1983, Chapter 8.
7. Deward, M. J. S.; Dougherty, R. C. The PMO Theory of Organic Chemistry,
Plenum, New York, 1974.
8. Hochstrasser, R. M.; Marzzallo, A. Molecular Luminescence, E. Lim, ed.;
Benjamin, W. A. New York, 1969.
9. Salem, L.; Rowland, C. Angew. Chem. Intl. Ed. 1972, 11, 92.
10. Doubleday, C. E.; Turro, J. N. Jr.; Wang, J. F. Acc. Chem. Res. 1989, 22,
199.
11. El-Sayed, M. A. J. Chem. Phys. 1963, 38, 2834.
12. El-Sayed, M. A. J. Chem. Phys. 1962, 36, 573.
13. Blankenship, R. E. Molecular Mechanisms of Photosynthesis, Blackwell
Science, Oxford, 2002.
14. Sansar, A. Chem. Rev. 2003, 103, 2203.
15. Gratzel, M. Acc. Chem. Res. 2009, 42, 1788.
16. Iwamura, M.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2007, 129, 5248.
17. Grabowski, Z. R.; Rotkiewies, K. Rettig, W. Chem. Rev. 2003, 103, 3899.
18. Teller, E. J. Phys. Chem. 1937, 41, 109.
Chapter 1
40
19. Teller, E. Israel J. Chem. 1969, 7, 227.
20. Bernardi, F.; De, S.; Olivucci, M.; Robb, M. A. J. Am. Chem. Soc. 1990,
112, 1737.
21. Bernardi, F.; Olivucci, M.; Robb, M. A. Acc. Chem. Res. 1990, 23, 405.
22. Bernardi, F.; Olivucci, M.; Robb, M. A. Chem. Soc. Rev. 1996, 20, 321.
23. Olivucci, M. (Editor) Computational Photochemistry, Theoretical and
Computational Chemistry; Vol. 16 Elsevier B. V. Amsterdam, The
Netherlands, 2005.
24. Domcke, W.; Yarkony, D. R.; Köppel, H. Conical Intersections: Electronic
Structure, Dynamics and Spectroscopy, World Scientific Publishing Co.,
Singapore 2004.
25. Domcke, W.; Yarkony, D. R. Annu. Rev. Phys. Chem. 2012, 63, 325.
26. Iwaki, L. K.; Dlott, D. D. In Encyclopedia of Chemical Physics and
Physical Chemistry, Vol. III, Moore, J. H.; Spenser, N. D. (eds.) Institute of
Physics Publishing, Philadelphia, 1999.
27. Iwaki, L. K.; Deak, J. C.; Rhea, S. T.; Dlott, D. D. In Ultrafast Infrared and
Raman Spectroscopy, Fayer, M. D., ed., Dekker, New York, 2001.
28. Jiang, Y.; Blanchard, G. J. J. Phys. Chem. 1994, 98, 9417.
29. Macpherson, A. N.; Gillbro, T. J. Phys. Chem. 1998, 102, 5049.
30. Birks, J. B. Photophysics of Aromatic Molecules, John Wiley  Sons, Inc.
New York, p. 143, 1970.
31. Byrne, J. P.; McCoy, E. F.; Ross, I. J. Aust. J. Chem. 1965, 18, 1589.
32. Siebrand, W. J. Chem. Phys. 1967, 46, 440.
33. Siebrand, W. J. Chem. Phys. 1967, 47, 2411.
34. Lewis, G. N.; Calvin, M. Chem. Rev. 1939, 25, 272.
35. Lin, S. H. J. Chem. Phys. 1969, 44, 3759.
36. Gegion, D.; Muskat, K. A.; Fischer, R. J. Am. Chem. Soc. 1968, 90, 2097.
37. DeBoer, C. D.; Schlessinger, R. H. J. Am. Chem. Soc. 1968, 90, 803.
38. Mukamel, S. Principles of nonlinear Spectroscopy, Oxford University Press:
New York, 1995.
Chapter 1
41
39. Grabowski, Z. R. Pure Appl. Chem. 1992, 64, 1249.
40. Grabowski, Z. R. Pure Appl. Chem. 1993, 65, 1751.
41. Rettig, W.; Maus, M. Conformational changes accompanying intramolecular
excited state electron transfer. In Conformational Analysis of Molecules in
Excited States; Waluk, J. Ed.; Wiley-VCH: New York, 2000.
42. Zewail, A. H. J. Phys. Chem. A 2000, 104, 5660.
43. Sundström, V. Prog. Quantum Electron. 2000, 24, 187.
44. Sundström, V.; Gillbro, T. J. Chem. Phys. 1984, 81, 3463.
45. Ben-Amotz, D.; Harris, C. B. J. Chem. Phys. 1987, 86, 6119.
46. Jurczok, M.; Plaza, P.; Martin, M. M.; Rettig, W. J. Phys. Chem. A 1999,
103, 3372.
47. Oster, G.; Nishijima, Y. J. Am. Chem. Soc. 1956, 78, 1581.
48. Changenet, P.; Zhang, H.; van der Meer, M. J.; Glasbeek, M.; Plaza, P.;
Martin, M. M. J. Phys. Chem. A 1998, 102, 6716.
49. van der Meer, M. J.; Zhang, H.; Glasbeek, M. J. Chem. Phys. 2000, 112,
2878.
50. Bagchi, B.; Fleming, G. R.; Oxtoby, D. W. J. Chem. Phys. 1983, 78, 7375.
51. van Veldhoven, E.; Zhang, H.; Rettig, W.; Brown, R. G.; Hepworth, J. D.;
Glasbeek, M. Chem. Phys. Lett. 2002, 363, 189.
52. Fonseca, T.; Kim, H. J.; Hynes, J. T. J. Mol. Liq. 1994, 60, 161.
53. Fonseca, T.; Kim, H. J.; Hynes, J. T. J. Photochem. Photobiol. A 1994, 82,
67.
54. Kim, H. J.; Hynes, J. T. J. Photochem. Photobiol. A 1997, 105, 337.
55. Unger, V. M.; Hargrave, P. A.; Baldwin, J. M.; Schertler, G. F. X. Nature
1997, 389, 203.
56. Stenkamp, R. E.; Teller, D. C.; Palczewski, K. ChemBioChem 2002, 3, 963.
57. Ernst, O. P.; Bartl, F. J. ChemBioChem 2002, 3, 968.
58. Haupts, U.; Tittor, J.; Oesterhelt, D. Annu. Rev. Biophys. Biomol. Struct.
1999, 28, 367.
Chapter 1
42
59. Quail, P.; Boylan, M. T.; Parks, B. M.; Short, T. W.; Xu, Y.; Wagner, D.
Science 1995, 268, 675.
60. Changenet, P.; Zhang, H.; van der Meer, M. J.; Hellingwerf, K. J.; Glasbeek,
M. Chem. Phys. Lett. 1998, 282, 276.
61. Haran, G.; Wynne, K.; Xie, A.; He, Q.; Chance, M.; Hochstrasser, R. M.
Chem. Phys. Lett. 1996, 261, 389.
62. Hasson, K. C.; Gai, F.; Anfinrud, P. A. Proc. Natl. Acad. Sci. USA 1996, 93,
15124.
Chapter 2
Experimental Methods
44
Chapter 2
This chapter enlists the experimental methods which have been used in carrying
out the research work presented in this thesis. I will mainly describe the
fundamental theory, basic principles and the instrumental details of time resolved
spectroscopic technique namely femtosecond fluorescence up-conversion and
transient absorption. Numerical methods of analysis of the experimental data and
the method of preparation of the materials used during the work will be described.
The usage of time-dependent density functional theory is also discussed in detail.
Chapter 2
45
2.1. Steady State Measurements
The steady state absorption spectra of the samples in condensed phase were
measured in Shimadzu UV-2450 and JASCO V-670 spectrophotometer. Steady
state emission spectra were measured in Fluoromax-4, Horiba Jobin-Yvon and
Fluorolog 3-21, Horiba Jobin-Yvon spectrofluorometer.
2.2. Femtosecond Fluorescence Up-conversion Measurements
Time resolved fluorescence up-conversion was first reported by Mahr and
Hirsch in 1975.1 A chromophore in condensed phase is excited by an ultrashort
optical pulse, and its emission at a specific wavelength is mixed with another
ultrashort pulse called as gate pulse in a nonlinear optical (NLO) crystal. Sumfrequency generation occurs at a well-defined crystal orientation when phasematching conditions are met and the up-converted signal is hence generated and
subsequently detected. The time evolution of fluorescence is obtained by delaying
the gate pulse with respect to pump pulse using a mechanical delay stage. By
plotting the intensity as function of time delay between pump and the gate pulses, a
“kinetic trace” of the fluorescence at the chosen wavelength is obtained.2
Fluorescence up-conversion has an advantage over time correlated single
photon counting techniques in terms of its time resolution, which is limited by
pulse duration itself. It can achieve a time resolution of about 100 fs i.e., any
excited state event or a process e.g. charge transfer, proton transfer, trans-cis
isomerization, dynamic solvation or any other molecular rearrangement occurring
in such ultrafast timescales can be traced using such a technique. In fluorescence
up-conversion measurements the excitation of the sample is carried out by an
ultrashort pulse (~ 100 fs) of light to achieve high resolution in time domain. The
excitation by ultrashort pulses over a wide wavelength range has been made
possible by means of the tunable lasers, and also the occurrence of nonlinear
optical phenomena, like second and third harmonic generation, have increased the
diversity of wavelengths needed for excitation of different chromophores. In the
fluorescence up-conversion setup (FOG 100, CDP Systems, Moscow, Russia),
46
Chapter 2
samples were excited using second harmonic generated optical pulses from a one
box mode-locked Ti:Sapphire oscillator (Mai-Tai-HP, Spectra Physics, USA)
operating at a working frequency of 80 MHz. The oscillator has an average power
of ca. 2.5 W and a peak power of ca. 300 kW at 750 nm. The diagram of the
experimental setup of femtosecond fluorescence up-conversion is schematically
shown in scheme 2.1. The ultrashort femtosecond pulses of fundamental frequency
Scheme 2.1. Schematic diagram of femtosecond fluorescence up-conversion setup
1 from the Ti-Sapphire oscillator are directed onto the -barium borate (BBO)
(0.5 mm,  = 38,  = 90) nonlinear crystal (NC1) by means of mirrors M1, M2
and lens L1 resulting in the generation of a second harmonic beam of frequency
21. Lens L2 collimates the fundamental and the second harmonic beams onto the
beam splitter BS1. BS1 allows 1 to pass through, which proceeds towards the
optical delay stage directed by two mirrors M3 and M4, while as 2 1 is reflected
by beam splitter towards the Berek waveplate (B). The residual fundamental
frequency 1 is cut off by means of a blue filter F1, and also a set of neutral
density filters to regulate adequate intensity of 2 1 are placed before ‘B’. Berek
waveplate is placed there in order to change the polarization of 21 to magic angle
(54.7) relative to 1. 21 is reflected back from another beam splitter BS2 and is
Chapter 2
47
focussed onto the sample (S) by means of focussing lens L3. To avoid photobleaching of the sample by exposure at a single spot, the sample is kept inside a
rotating cell. Once the pump pulse ‘2 1’ hits the sample, absorption of light takes
place and the molecules are promoted into the excited state. The excited state emits
fluorescence (2), which is collected by another lens L4. Filter F2 is placed just
after the lens to filter out the residual amount of 2 1. Fluorescence (2) from the
sample and the gate pulse ( 1) from the optical delay line are collimated by means
of lens L5 onto another nonlinear 0.5 mm thick BBO crystal ‘NC2’ (0.5 mm,  =
38,  = 90). This is known as the sum-frequency generation crystal as it causes
mixing of 1 and 2 and produces a sum frequency signal ( 3 = 1 + 2). This sum
frequency signal proceeds towards the monochromator and is detected by a
photomultiplier tube (PMT). The output of PMT is connected to computer and the
data collection is done by the software “Lumex”.
There are many stages involved in the data acquisition by the fluorescence upconversion setup briefly described here. The details are given below;
2.2.1. Excitation of the Sample with SHG Pulse
Earlier it was believed that induced polarization in a dielectric material is
linearly proportional to the applied electric field. However since 1960, when the
high intensity light sources became available, it was realized that linearity is a mere
approximation. Instead, the polarization in a dielectric material can be expanded in
terms of applied electric field as;
( 3)
Pk (t )   0 ( ik(1) Ei (t )  ijk( 2) Ei (t ) E j (t )  ijkl
Ei (t ) E j (t ) Ek (t )  )
(2.1)
(1)
0 is the permittivity of free space.  ik is the first order or linear susceptibility,
( 2)
( 3)
while as  ijk and  ijkl are the nonlinear susceptibilities and are called respectively
as second and third order susceptibilities. The second and third terms in the above
equation represent the second and third order optical response of the dielectric
material towards the applied electric field. Second harmonic generation (SHG) is a
Chapter 2
48
coherent optical process of radiation of dipoles in the material, dependent on the
second term of the nonlinear polarization, which can be illustrated as follows for a
laser beam of electric field strength of
E (t )  Ee it  c.c
(2.2)
Where, c.c is the complex conjugate term. The nonlinear polarization that is
generated in the crystal is given by;
P( 2) (t )   0  ( 2) E 2 (t )
P( 2) (t )  2 0  ( 2) EE*  ( 0  ( 2) E 2e i 2t  c.c)
(2.3)
First term in the above equation does not contain any frequency term, and hence
does not lead to generation of electromagnetic radiation; rather it is responsible for
another process known as optical rectification. In this process a static electric field
is created inside the nonlinear crystal. The term in brackets of above equation is
responsible for generation of second harmonic signal. One can essentially consider
the molecules of the NLO material as dipoles and under the influence of the
electric field these dipoles oscillate with the applied electric field of frequency 1,
and radiates electric field of frequency 2 1 as well as 1. So the near infrared input
light comes out as near UV light. In centro-symmetric materials, SHG cannot be
observed, because of the inversion symmetries in polarization and electric field.
The odd terms will only survive, thus the second order harmonics are not observed.
An important criteria critical for SHG to occur is the phase matching of the
radiated electric field of 1 and 21. They have to be in phase, so as not to
interfere destructively. The applied electric field is strong enough to generate the
second order radiation. So we could think of SHG signal as a perturbation and thus
those lights of two frequencies should be added, and not cancel out each other. The
total output is sum or integration of all the dipoles at every position in the dielectric
material. So the maximum is achieved when the phases are same.
In the up-conversion setup we used, The SHG wave is generated using a
nonlinear optical BBO crystal, and the SHG beam is used to excite the sample,
while as the fundamental beam is allowed to proceed towards the optical delay
Chapter 2
49
line. SHG signal is passed through a Berek waveplate before it irradiates the
sample. Berek waveplate maintains the polarization of SHG signal 54.7 with
respect to the fundamental beam in order to avoid effects due to rotational
diffusion of the molecules. Berek comprises of a set of prisms made of calcite
crystals, in which tilt and rotation of the prisms change the polarization of the laser
beam. The laser beam is focussed on the sample by means of a lens. The sample in
condensed phase are kept in a rotating cell driven by a motor, in order to avoid the
continuous exposure of the sample at a single spot to the laser beam, which may
cause photo-bleaching of the sample.
2.2.2. Collection of the Fluorescence and Delay of the Gate Pulse
Once the SHG laser beam irradiates the sample in condensed phase, absorption
of the photons take place and such an act of absorption takes the molecules to the
excited electronic states. During relaxation, the molecule emits photons in the form
of fluorescence, which are collected and collimated by a lens. The small amount of
the excitation light may also leak out of the sample because of high photon density
and may get involved in the frequency mixing process and will hence strongly
affect the up-converted signal. To avoid such a problem, filters are used to block
the excitation beam completely, so that only pure fluorescence proceeds towards
the sum frequency generation crystal, which is actually another nonlinear optical
BBO crystal.
As mentioned above, fluorescence up-conversion works on the principle of
frequency mixing of the fluorescence from sample and the gate pulse inside the
nonlinear BBO crystal. The only difference between second harmonic generation
and the sum frequency generation process is that, in the later the frequencies
undergoing mixing are different, while as in former they are same. In a nonlinear
crystal, different frequencies of the incident field can be written as sum of two
frequency components from equation 2.2 as;
E(t )  E1e i1t  E2e i2t  c.c
(2.4)
Chapter 2
50
For incident field of the above form, the polarization created inside the nonlinear
crystal is of the form;
P
(2)
 2 2i1t  E 2 e 2i 2 t  2E E e i (1  2 )t 
1 2
(2)  E1 e
21

(t )   0 


i (1  2 )t
 cc  2( E1E1*  E2 E2* ) 
 2E1E2e
(2.5)
This expression contains various quantities describing the different physical
processes occurring inside the nonlinear crystal under the influence of applied
field. The term with ( 1+2) describes sum frequency generation, and the term
with (1-2) is the difference frequency generation. However, typically only one
of the frequency components will be present with significant intensity generated by
optical rectification. For such a process to occur, phase matching condition
between the two pulses should be fulfilled and this process cannot be achieved for
more than one frequency component of the above equation.
The frequency mixing of fluorescence ( 2) from the sample and the
fundamental gate pulse ( 1) inside the BBO crystal produces light with sumfrequency (1+2). The intensity of the sum-frequency signal (Isf) at a given delay
time () of the gate beam is proportional to the convolution function of the
fluorescence intensity (If) and the intensity of the gate beam (Ig) as,

I sf ( )   I f (t ) I g (t   )dt

(2.6)
Optical delay line maintains the time interval between the pump pulse (2) and the
gate pulse () by changing the path length of the gate pulse through the delay
stage. By controlling the arrival of the photon pulses of gate beam at the sum
frequency generation crystal, we can exactly measure the variation of fluorescence
intensity with the time (). The schematic illustration shown below will help in
understanding, how the fluorescence from the excited state can be measured as a
function of time as shown in scheme 2.2.
Chapter 2
51
2.2.3. Signal Acquisition and Data Analysis
The presence of sum frequency generation crystal causes mixing of the
fluorescence (2) from the sample and the fundamental gate pulse ( 1) resulting in
the formation of an extra frequency component ( 3) whose frequency is the sum of
the former two frequencies. The frequencies 1 and 2 are not allowed to proceed
forwards and are halted, while as the sum frequency generated light is collected
and collimated by a lens and allowed to proceed into the double monochromator to
select the exact sum-frequency wavelength. The monochromator passes the sumfrequency signal to photomultiplier tube (PMT), which shows very high sensitivity
to the wavelengths in UV range. The data acquisition is in the form of intensity of
Scheme 2.2. Measurement of fluorescence intensity as a function of time delay in the gate
pulse.
sum-frequency signal and the time delay between pump and the gate pulse. These
time resolved fluorescence data are called as “kinetic traces”. The fluorescence
lifetime of a molecule is defined as the rate of depopulation of the excited state
Chapter 2
52
following excitation by a pulse of zero width in the time domain i.e., -function.
However, in practice no pulse has zero width in the time domain, rather the
excitation pulse have a definite width, which in our case is in femtoseconds. The
assumption of -function optical excitation holds well only, if the lifetime of
excited state is much larger than the excitation pulse. The measured fluorescence
departs from the actual fluorescence response function due to finite width of the
excitation pulse. These all factors cause broadening of the optical excitation pulses
resulting in the larger instrument response function (IRF) than the actual width of
excitation pulse.
2.2.4. Convolution Procedure
For molecules decaying on a much faster time scale, the instrument response
function may prove decisive in deciding the lifetime of the excited state and also to
reveal the nature of various excited state process occurring in it. In essence, the
measured fluorescence decay of the molecule is a convolution of the fluorescence
from the molecule and the instrument response function. To extract the information
related to fluorescence exclusively, it is necessary to deconvolute the experimental
kinetic traces obtained during the measurement. The measured fluorescence decay
is mathematically represented by the following convolution integral,
t

F (t )  I (t ) P(t  t )dt 
(2.7)
0
where, t defines variable time delays of infinitesimally small time widths, dt. F(t)
is the fluorescence intensity at any time t, and I(t) is the intensity of the exciting
light at time t. P(t-t) is the response function of the experimental setup. For a
single exponential decay of the excited state population, P(t-t) can be written as
P(t-t)= exp[-(t-t)/]. The above equation takes the form
t

F (t )  exp(t /  ) I (t ) exp(t  /  )dt 
0
(2.8)
Chapter 2
53
while as F(t) is the observed fluorescence decay and I(t) is the instrument response
function, which are obtained experimentally by measuring fluorescence traces of a
sample solution and the sum-frequency signal of water Raman and the gate pulse
respectively. The deconvolution is based on an iterative least square regenerative
convolution method.3-7 An excitation pulse profile is first recorded, which in the
present setup is obtained by measuring sum-frequency signal of water Raman and
the gate pulse. Cross-correlation of the fundamental probe pulse and water Raman
signal can essentially enable us to calculate the IRF of the system, as Raman signal
is instantaneous and the cross-correlated signal only shows dependence on the
pulse-width and the group velocity dispersion introduced in the pump and gate
pulses throughout the path length of the laser beams. The IRF can be perfectly
modelled with a Gaussian function, and the time resolution of the up-conversion
setup is directly obtained from the full-width-half-maximum (FWHM) of this
Gaussian shaped instrument response function fetching a value of ca. 250 fs. The
fluorescence transients from the up-conversion setup were fitted using convolution
procedure in IGOR Pro software using the following equation,
1
F (t ) 
2
  2 (t  t ) 
  2   (t  t ) 
G
0
0 


ai exp

erf  G i
 2 i


 i 
2 i G



i

(2.9)
where, G is the width of Gaussian shaped instrument response function and t0 is
the time offset between the IRF and the fluorescence transient. ai and i are the
respective amplitude and the time constant obtained during the fitting of the
experimental data. Error function, also called as Gaussian error function is a
special function of sigmoid shape and is defined as;
erf ( x) 
x
exp(t
 
2
2
)dt
0
The decay transients have been fitted either with single exponential or with multiexponential functions, as we will see it in all the succeeding chapters of the thesis.
Chapter 2
54
2.3. Time Resolved Emission Spectra
Excited state processes in a molecule result in complex time-dependent decays
and the fluorescence traces depend on the observation wavelength because of the
time needed for the relaxation of directly excited Franck-Condon (FC) state to
evolve into relaxed state or some intermediate state. Suppose, if emission could be
recorded at any instant following the excitation pulse, then we can continuously
monitor the emission spectra starting as the molecule relaxes from FC to relaxed
state. These emission spectra, representing discrete times following pulsed
excitation, are called the time resolved emission spectra (TRES). The procedure
starts with measurement of time resolved decay transients at a number of
wavelengths across the emission spectra, I(,t).8-10 Since the intensity depends on
both wavelength and time, the total intensity can be written as a product of two
functions individually dependent on wavelength and time as;
I (, t )  A( ) B(t )
(2.10)
Time dependent part B(t) can be obtained from the best fits of the fluorescence
transients. Wavelength dependent part A(), on the other hand can be obtained
from steady state emission spectrum Iss(), (intensity of emission corresponding to
the wavelength at which decay is measured). Thus one can write;
I ss ( ) 


0
0
 I (, t)dt  A() B(t)dt
(2.11)
where,
n
B(t ) 
 ai exp(t /  i )
i 1
(2.12)
Thus,
A( ) 
I ss ( )


0
B(t )dt

I ss ( )
 ai i
i
(2.13)
Chapter 2
55
We can write the wavelength and time dependent fluorescence intensity in the
modified form as;
I ( , t )  A( ) B(t ) 
I ss ( )
n
ai exp(t /  i )


ai i i 1
(2.14)
i
I(,t) represents the fluorescence intensity corresponding to a particular
wavelength at a particular instant of time. One can construct TRES at different
times during the relaxation by substituting the steady state fluorescence intensity
and the fitting parameters of the decay transients at various wavelengths. Because
TRES represents essentially the intensity values at different times as a function of
wavelength and hence contains only few points, it is necessary to fit the data with
some smooth function capable of modelling its line-shape. Lognormal line-shape
function was introduced to fit the TRES data.8
2

 ln(1  2b(  p ) / )  

I ( )  I 0 exp ln( 2)

 

b

 

(2.15)
for, 2b(   p ) /     1
and, I ()  0, for   1
It is a four parameter function and describes an asymmetric line-shape, which
reduces to a Gaussian function in the limit b = 0. In this equation; I0 is the peak
height, p is the frequency corresponding to the peak of the spectra in cm -1, b is
asymmetric factor (usually negative e.g. -0.3),  is the width parameter usually of
the order of few thousand cm-1.
Time dependence of the Stokes shift can be obtained by plotting the variation of
peak or mean frequency with spectral time as obtained from fitted TRES data. In
the work presented here, mostly the variation of peak frequency with time has been
used. In Chapter 6, we used the variation of mean frequency or first moment with
time to understand the evolution of Stokes shift with time. The use of first moment
is mainly recommended for cases where in there is prominent broadening of the
emission spectra with time, which cannot be incorporated by choosing the
Chapter 2
56
variation of peak frequency. The first moment of the TRES is calculated by using
following equation,

 .I ( )d
  0


I ( )d
(2.16)
o
Time resolved area normalized emission spectra (TRANES) is a one-step
extension of the frequently used TRES method, calculated by equating the area of
each spectrum in TRES to a constant value; e.g. the area at time, t = 0. The
interpretation of TRANES spectra is relatively simple. For example, observation of
identical TRANES spectra confirms emission from a single species and an
isoemissive point in the TRANES spectra confirms emission from two species. An
isoemissive point in the TRANES spectra is the necessary and sufficient condition
to confirm the presence of two emissive species irrespective of their origin.
Isoemissive point occurs at a wavelength at which the ratio of the radiative rates at
that wavelength is equal to the ratio of total radiative rates of the two emissive
species. The isoemissive point is the equivalent of the isosbestic point in
absorption spectroscopy.
Chapter 2
57
2.4. Femtosecond Transient Absorption Measurements
Femtosecond transient absorption (TA) spectroscopy represents the natural
evolution of flash photolysis since the introduction of ultrashort laser sources. 11-15
The principles of the technique are essentially the same though much more care is
necessary in order to carry experiments with higher time resolution. The
fundamental idea of the method is to use an ultrashort pulse to excite the system
under study and to follow the course of photo-reaction by monitoring the change in
absorption properties of the system. The excitation pulse instantly increases the
energy of the system and triggers a chain of spontaneous and stimulated reactions.
With pulse duration of about 100 fs it is possible to access most of the dynamics
occurring in the excited electronic states: internal conversion, intramolecular
vibrational relaxation, intersystem crossing, excited state reactions like charge
transfer, proton transfer, isomerisation, and many other processes leading to
reactive pathways. This technique is somehow complementary to time-resolved
fluorescence measurements, though in principle it is capable of providing larger
amount of information about the system of interest. While fluorescence upconversion measurements are attainable only with reasonable fluorescent samples
(with a quantum yield of ~ 1%), TA can be utilized in any system and can also
probe the dark states. The only requirement is that a transition from an excited state
to higher states is allowed and possesses a dipole moment strong enough to make it
observable. This is highly probable if the observation is extended in a wide spectral
range (from mid IR up to UV).
The TA measurement involves two femtosecond laser pulses, a
monochromatic energetic pump pulse (Ipump) which triggers the photoreaction and
a weak (broad or monochromatic) probe pulse (Iprobe) (Scheme 2.3). The pump
pulse, which passes through a certain volume of the sample solution, is resonant
with an electronic transition of the photo-system of interest. Thus the pump pulse
induces transition in certain amount of molecules to their excited state by a vertical
Franck-Condon transition (usually only a few percent of molecules are excited
within the excited volume, depending on the pump power and absorption cross
Chapter 2
58
section of the molecules). The probe pulse passes through the same volume after a
certain delay time (t) with respect to pump pulse. For each time delay
Scheme 2.3. Schematic depiction of the transient absorption spectroscopy principle
between the pump and the probe pulse, both the intensities of the probe pulse with
pump
0
and I probe
and without pump pulse ( I probe
) are measured. In this way, difference
absorption spectra, A (,t) is calculated to measure the signal related to the
excited states and the formation of product states.
0
 I probe

A  log
pump 
I probe 

For broad probe pulses, A spectra are obtained straightforward, by dispersing the
white light continuum onto a Coupled Cluster Detector (CCD). There are various
types of processes which can occur when the molecules are excited by a strong
pump pulse depending upon the nature of the molecule and the process it is
susceptible to undergo. The various contributions to the transient absorption
spectra (Scheme 2.4) are given below.
1. Ground state bleaching or depopulation signal: A fraction of the
molecules within the excitation volume have been promoted to the excited state by
means of a pump pulse and consequently the population in the ground state is
decreased. Consequently only small number of molecules will absorb the probe
Chapter 2
59
light and more will be transmitted through. In this way the amount of probe light
reaching the CCD will be more in presence of pump compared to in absence of
pump and hence as per above equation, absorption difference will be negative in
the wavelength region of ground state absorption.
2. Stimulated emission: For a two-level system, the Einstein coefficients for
absorption from ground state to the excited state (A12) and stimulated emission
from the excited to the ground state (A21) are identical. Stimulated emission to the
Scheme 2.4. Contribution of various processes to the difference absorption spectrum (OD):
ground state bleaching (GSB), stimulated emission (SE), excited state absorption (ESA), and
product absorption (PA).
ground state is possible occur once probe light of suitable wavelength passes
through the sample during the lifetime of the excited state. The stimulated emission
will occur only for optically allowed transitions and will have a spectral profile
similar to the steady state fluorescence spectrum of the chromophore and it will be
Stokes shifted from the ground state bleaching. In stimulated emission, a photon
from the probe pulse induces emission of another photon from the excited
molecule and pushes the molecules to ground state. The photon produced by
stimulated emission is emitted in the same direction as that of probe photon and
hence both will be detected. Stimulated emission results in an increase of light
intensity on the detector, corresponding to a negative A signal. Under certain
60
Chapter 2
circumstances the Stokes shift may be so small that the stimulated emission band
spectrally overlaps with the ground state bleaching and emerge as a single band.
3. Excited state absorption: Upon absorption of the photons of pump pulse,
the molecules are prompted to excited states. Optically allowed transitions from
the excited states of the chromophore to higher excited states may exist in certain
wavelength regions, and absorption of the probe pulse at these wavelengths will
occur. Consequently the intensity of the probe pulse in these wavelength regions
will be higher in absence of pump pulse than in presence of pump pulse and hence
a positive signal in the difference absorption spectrum will be observed in the
wavelength region of the excited state absorption.
4. Product absorption: After excitation of the sample by means of a pump
pulse, reactions may occur that results in the formation of a transient or some long
lived intermediate state, such as charge transfer states, proton transfer states,
isomerisation, triplet states etc. The absorption of such species will appear as a
positive signal in the difference absorption spectrum. Ground state bleach will
appear in the wavelength region where the chromophore on which the product state
resides has ground-state absorption.16,17
2.4.1. Pulse Duration, Time Resolution and Spectral Selectivity
Optical pulses as short as 5 fs are now employed for transient absorption
measurements.18,19 For Gaussian-shaped pulses, the time-bandwidth product t
= 0.44 necessitates that for a short pulse duration a large spectral bandwidth is
expected. For example, a 10 fs pulse with centre at 800 nm wavelength has a
spectral bandwidth of 4.4 x 1013 Hz at full-width half maximum (FWHM), which
corresponds to about 100 nm in wavelength domain. Thus one has to make a
proper selection of the time resolution and the spectral sensitivity. With a 10 fs
optical pulse at 800 nm, one would simultaneously excite all the cofactors with a
100 nm bandwidth pulse. To selectively excite any of the cofactors, one has to use
a pulse of ca. 30 nm bandwidth, which implies a pulse of time duration not less
than ca. 30 fs. For some systems where the difference in absorption between
Chapter 2
61
various chromophores is significantly smaller, optical pulses of 100 fs time
duration are needed. 20-23
On very fast timescales, transition absorption signals have contributions from
process in addition to the ones mentioned above. There are mainly three types of
potential artifacts in TA spectroscopy; known as ‘two-photon absorption’,
‘stimulated Raman amplification’ and ‘cross-phase modulation’.24 As transient
absorption signals result from light matter interaction through the third order
nonlinear susceptibility (3), non-sequential light interactions that do not represent
population dynamics of electronic states will contribute to the signals.
Femtosecond coherent artifacts are produced by the simultaneous action of one
photon from the pump pulse and another from probe pulse. Such undesired signals
can be ignored by excluding the initial phases of the femtosecond dynamics from
the data interpretation and analysis, or they may be included in the analysis by
considering their physical origin. Under such circumstances assumptions need to
be taken about the line shapes and dephasing times of the chromophore in
question.25 Cross phase modulation effects are due to a change in the index of
refraction of solvent and sample cell induced by the pump beam and give rise to
oscillatory patterns around zero time. Such artifacts can in principle be removed
from the data set by subtracting the signal obtained from sample cell with solvent
in it.26
2.4.2. Transient Absorption Setup
The laser system used for transient absorption experiments consisted of a mode
locked oscillator (Spectra Physics, Mai Tai SP) and a Ti:sapphire regenerative
amplifier (Spectra Physics, Spitfire Pro XP) pumped by a 20W Q-switched
Nd:YLF laser (Spectra Physics, Empower). The regenerative amplifier generated
50 fs pulses, centered at 800 nm, at a 1 kHz repetition rate with energy of 4 mJ per
pulse. The schematic illustration of our transient absorption setup is shown below
in scheme 2.5. The fundamental beam was directed to the setup by mirrors ‘MA’
and ‘MB’ and was split into two beams by means of a beam splitter ‘MC’ with
62
Chapter 2
major proportion being used to generate the pump pulses. Most of the
photophysically and photochemically important molecular systems, since absorb
light in the UV-vis region, so as a matter of fact frequency doubling or tripling of
the fundamental beam is required. The frequency doubling was performed in a 0.2
mm BBO crystal. The fundamental (800 nm) and the second harmonic beam (400
nm) were aligned collinearly in another 0.2 mm BBO crystal to generate third
harmonic light (266 nm). To cause a temporal overlap between the 800 nm and 400
nm light, 800 nm light was delayed in time by means of CaF2 crystal. The second
and third harmonic light is generated inside a JANOS tripler, and the respective
beams are directed to mirrors ‘MG’ and ‘MH’. A small portion of the fundamental
light was allowed to pass through a window ‘W1’ in the FemtoFrame - II and then
directed to fall onto the retro-reflector ‘RR’ by mean of mirrors ‘M1’ and ‘M2’ and
iris ‘D1’. The retro-reflector is mounted onto a computer controlled motorized
delay stage, which controls the time of arrival of probe pulses relative to pump
pulses onto the sample cell. The returned beam from the retro-reflector is directed
onto the beam splitter ‘BS’ by means of mirrors ‘M3’ and ‘M4’. Most of the light
passes through the beams splitter and traverses through a variable linear neutral
density filter ‘F1’, an iris ‘D2’, a focusing lens ‘L1’ to fall onto the 0.3 mm
sapphire plate. The neutral density filter ‘F1’ and iris ‘D2’ control the amount of
light and lens controls the focusing of the light onto the sapphire plate for
optimized generation of the white light continuum. The other part of the beam
from BS is allowed to reach a photodiode placed just behind the BS (vide infra).
The white light continuum with spectral bandwidth of 450 – 770 nm generated
from the sapphire plate is deflected by a series of plane mirrors from ‘M5’ – ‘M9’
onto a concave mirror ‘M10’ interceded by two dielectric notch filters ‘Df1’ and
‘Df2’ and an iris ‘D3’. Dielectric notch filters are used in order to cut off the
fundamental 800 nm light from the white light continuum. The concave mirror
‘M10’ inclines the probe white light continuum into the optical fiber cable by
means of a plane mirror ‘M11’ and a small focal length lens. The path is also
interceded by a variable circular neutral density filter to control the amount of
Chapter 2
63
Scheme 2.5. Schematic representation of a femtosecond transient absorption setup in our
laboratory.
probe light reaching the 200 m optical fiber cable. The optical fiber cable takes
the probe white light continuum to the spectrograph wherein the light is dispersed
64
Chapter 2
by a diffraction grating onto a multichannel coupled cluster detector ‘CCD’ meant
for measuring the intensity of the incoming light at large number of wavelengths
simultaneously. The second or third harmonic pump beams from the JANOS
tripler are directed inside the FemtoFrame – II by means of mirrors ‘MG’ and
‘MH’ thorough a window ‘W2’. The pump beam then passes through
electronically controlled chopper ‘CH’, a 50 cm focal length lens, a half waveplate
‘HWP’, a variable circular neutral density filter ‘F2’ and an iris ‘D4’ onto the
mirror ‘M12’. ‘M12’ directs the pump pulses into the sample cell maintaining a
small angle of divergence with respect to the probe beam, so that the transmitted
pump pulses should not pass onto the optical fiber cable. The alignment of the
pump and the probe beams is maintained in a way to furnish a spatial and a
temporal overlaps between the two pulses within the observation volume. The
chopper placed in the pump path is controlled electronically and its rotation is
synchronized with the photodiode placed behind the BS in the probe path. The data
recording of the CCD is also synchronized with the photodiode. Halfwave plate
‘HWP’ is meant to control the polarization of the pump pulses relative to the probe
pulse and for current measurements is set at 54.7 (magic angle) to avoid any
contributions from the rotational diffusion of the molecules. The beam waist of the
pump light is usually set approximately double than that of the probe light and also
the measurements are performed with pump energy of ca. 1 J. The data is
recorded in a FemtoFrame 2.4.6 software based on LabView. Data analysis is
performed by the Femtosuite software. For global analysis, GLOTARAN software
was used, which is freely available on internet.27 The modeling of the procedure
for global analysis and the assessment of the obtained information is mentioned in
the literature.28-31 The fitting of the kinetic traces was performed in IGOR,
wavemetrics USA, while as the TA data was loaded into the IGOR by means of
procedure written by Matt Sfeir, Brookhaven National Laboratory which is
available free on internet.
Chapter 2
65
2.5. Quantum Mechanical Calculations
The treatment of electronic states of molecular systems is by far the most
common application of theoretical calculations in chemistry. If the potential energy
surfaces of the electronic states and their couplings are to be evaluated as needed
while trying to devise a model for the excited state relaxation of a molecular
system, it becomes highly imperative that the method for generating these PESs be
computationally amenable. More importantly, one must be carefully not to
sacrifice accuracy, at least qualitative, in the process of finding an efficient model
for generating the PESs. The best choice for carrying out such dynamical studies
from the stand point of their accuracy is ab-initio quantum chemistry. The central
equation representing the time evolution of the correlated motion of nuclei and
electrons in a molecule is the time dependent Schrödinger equation.
i

 H
t
(2.17)
where  is the wave function describing the state of system.
Hohenberg-Kohn density functional theory (DFT) is one of the most popular
method and is strictly limited to ground states only, which excludes its usage in
photochemistry.32 Several routes have been followed to extend the conventional
DFT to excited states, among which time-dependent density functional theory is
the most popular method to treat excited states.33-35 Excited states are solutions of
the time-independent stationary Schrödinger equation and time dependent response
theory is used as a trick to reduce electronic excitations to ground state properties.
If a molecule in the ground state is subjected to a periodic perturbation by a
uniform electric field E oscillating at certain frequency, . The current density and
the electronic charge of the molecule are given by one-particle density matrix (t).
This density matrix will perform driven oscillations about its ground state value
(0). Fourier transform of (t) will provide the amplitude of these oscillations
denoted by (). Based on elementary perturbation theory, the amplitude of the
oscillations can be expanded in powers of electric field E.
Chapter 2
66
 ( )   (0) 
 
 
 0n 0n   0n  0n
    0n    0n
n 


 E  O( E 2 )


(2.18)
Once the frequency  of the applied field approaches the excitation energy 0n,
there occurs a resonance catastrophe and the amplitude of the oscillation diverges.
Analogous to a system of harmonic oscillators,
36,37
the excitation energies 0n are
the eigenfrequencies of the electrons in the molecule, and the transition density
matrices () are the corresponding collective modes. On inversion of the relation
between amplitude of oscillations and the electric field, the excitation energies are
obtained as eigenvalues of an electronic Hessian, which is imagined as a matrix of
second derivative of the electron energy with respect to the electronic degrees of
freedom. In this way, any ground state formulation can be extended to excited
states, provided time-dependent response is well-defined. In time-dependent KohnSham (TDKS) framework,38 a system of N non-interacting particles is considered
whose density is constrained to the physical density (t,x). This leads to the timedependent Kohn-Sham equation
i

 j (t , x)  H [  ](t , x) j (t , x)
t
(2.19)
The effective one-particle Hamiltonian H[](t,x) = 2(t,x)/2 + s[](t,x) is
comprised of kinetic energy part and a local time-dependent external potential
s[], which is a unique functional of (t,x) for a given initial state, as postulated
by Runge and Gross;38
s [  ](t, x)  ext (t, x)  C [  ](t , x)   xc [  ](t, x)
(2.20)
This potential is composed of one-particle external potential ext, the timedependent Coulomb potential C[](t,x), and the time-dependent exchangecorrelation potential xc[](t,x). The exchange-correlation potential comprises all
the non-trivial many-body effects. In ordinary DFT, xc is normally written as a
functional derivative of exchange-correlation energy. This follows from a
variational derivation of the Kohn-Sham equations starting from the total energy. It
is not however, straight forward to extend this formalism to the time-dependent
Chapter 2
67
case due to problems related to causality.39 The problem was solved by van
Leeuwan in 1998, by using the Keldysh formalism to define exchange-correlation
potential in terms of a new action potential.40
2.5.1. Exchange-Correlation Functionals
The exchange and correlation energy are the difference between the exact total
energy of a system and the classical Hartree energy. It is highly improbable to treat
exchange and correlation exactly and approximations are needed. The quality of a
TDDFT calculation is determined by how close the approximate exchange and the
correlation approaches to the exact value.
The most common and universally used approximation in TDDFT in the
Adiabatic Approximation (AA).41 It replaces the time-dependent exchangecorrelation potential by its static counterpart, evaluated at the time-dependent
density level. The resulting potential is instantaneous or local in time and hence
can work in cases where temporal dependence is small i.e., when the time
dependent system is locally close to equilibrium. Certainly this is far from the
reality under conditions when one is studying the interactions of strong laser light
with matter. The introduction of Local Density Approximation (LDA) into AA
resulted in another exchange-correlation functional called Adiabatic Local Density
Functional (ALDA), although it retained the problems of LDA, but still improved
the results.
Hybrid functionals are a class of approximations that interpolates between
Hartree-Fock (HF) theory and semi-local functional, and the fraction of HF
exchange is controlled by the exchange parameter. The exact exchange energy
functional is expressed as Kohn-Sham orbitals rather than the density, so is also
called as implicit density functional. This approach was initially introduced by
Axel Becke in 1993.42,43 One of the most commonly used versions of hybrid
functionals is B3LYP, which stands for Becke, 3-parameter, Lee-Yang-Parr. The
exchange-correlation functional is:
Chapter 2
68
B3LYP
LDA
E xc
 E xc
 a0 ( E xHF  E xLDA )  a x ( E xGGA  E xLDA )  ac ( EcGGA  EcLDA )
(2.21)
GGA
GGA
Where a0 = 0.2, ax = 0.72, and ac = 0.81. Ex and Ec are Generalized Gradient
Approximations (GGA): the Becke 88 exchange functional44 and the correlation
LDA
functional45 of Lee, Yang and Parr and Ec is the Vosko, Wilk, and Nusair
(VWN)46 local density approximation to the correlation functional. Most of the
Theoretical calculations mentioned in the thesis are performed using TDDFT with
B3LYP functional. Sometimes, the excited state energies incorporating solvent
were calculated using PBE0 functional.
One major limitation of approximate exchange-correlation functionals used in
TDDFT is that the exchange-correlation potentials do not exhibit the correct 1/r
asymptotic behaviour (r is electron-nucleus distance), rather falls of too rapidly.47
Also, for many exchange-correlation functionals, the corresponding exchange
potential is too high in the inner regions, resulting in lesser bound virtual orbitals.
This results in significant errors under conditions in which electron drifts away
from the nucleus. Even though this limitation was improvised by using
asymptotically correct potentials, but for systems with extended -conjugation48
and for molecules undergoing charge-transfer,49-51 the applicability of TDDFT is
limited. Especially, excitation energies of long-range charge transfer states in
weakly interacting molecular systems are drastically underestimated. Because of
such limitation, the results of theoretical calculations have mainly been used to
help in elucidating the experimental observations at the qualitative level only.
2.5.2. Basis Set
A basis set in quantum chemistry is a set of functions which are combined in
linear combinations to create molecular orbitals. These functions are typically
atomic orbitals centred on atoms and the calculations are usually performed using a
finite set of basis functions. For most of the calculations presented in this thesis,
we preferred to use basis set with significant number of basis functions like 631+G(d,p), 6-311++G(d,p). The authenticity of the calculations was usually
Chapter 2
69
performed by comparing the excitation wavelength, transition moments, and
oscillator strength with that of experimental values.
2.5.3. Solvent Effects
Electronic absorption and emission spectra usually exhibit a marked solvent
dependence. A common approach to incorporate these solvent effects in quantum
chemical calculations is the use of self-consistent reaction-field (SCRF) continuum
models where the explicit solvent structure is not considered. However its main
advantage is that the solute’s electronic structure can be treated quantitatively and
polarization effects are evaluated. All of the SCRF methods provide special
attention to the determination of solvation Gibb’s energy. Among all the SCRF
methods, the Polarizable Continuum Model (PCM)52 has proven to be a reliable
tool for the description of electrostatic solute-solvent interactions. In this model,
solvent effect is approximated as a polarisable continuum, with solute molecule
placed in a cavity formed by interlocking spheres, whose dielectric constant is set
to one. The presence of solvent molecules in the immediate surroundings leads to
an additional external potential which depends on the charge density of the
electrons. The use of DFT within the continuum models for solvation is an
efficient way to model solvent effects.53,54
2.6. Synthesis of Model GFP chromophore Analogs
2.6.1. General Aspects
The melting points were determined in a JSGW melting point apparatus.
Column chromatography was performed with silica gel 100-200 mesh particle size.
TLC analyses were performed with MERK Kieselgel 60 F254 plates. 1H NMR
spectra were recorded in JEOL 400 (400 MHz) and JEOL 500 (500 MHz) NMR
spectrometers in solution of CDCl3 with the solvent as the internal standard.
13
C
NMR spectra were recorded in JEOL 400 (100 MHz) and JEOL 500 (125 MHz)
NMR spectrometers with complete proton decoupling. IR spectra were recorded on
Chapter 2
70
a Bruker Vector 22 FT-IR spectrometer. Mass spectral analyses were carried out
on a Waters ESI-QTOF instrument.
All solvents were purchased from local company SD Fine Chemicals (India)
and were purified by established procedures.55 4-N,N-dimethylaminobenzaldehyde,
N,N-dimethylformamide, N-benzoylglycine, acetic anhydride, aniline, anhydrous
zinc chloride and anhydrous sodium acetate were purchased from SD Fine
chemicals (India). Zinc chloride was dried by strong heating under vacuum and
sodium acetate was dried under strong heating for 4-5 hours before use.
Triphenylamine was purchased from Sigma-Aldrich (USA). It was used in
preparation of tripehylamine aldehyde by Vilsmeier-Haack reaction. 56 Phosphorus
oxychloride was purchased from Spectrochem. N,N-dimethyl formamide was
distilled under vacuum before use. Acetic anhydride and aniline were distilled
under vacuum prior to use. All other chemicals used in synthesis were used as such
unless otherwise mentioned. For spectroscopic studies all the solvents were
purchased from Sigma-Aldrich and were used as such. Synthesis of the GFP
chromophore analogs DPI and DPPI is shown in scheme 2.6.
Scheme 2.6. Synthesis of GFP chromophore analogs DPI and DPI
2.6.2. General procedure for preparation of (4Z)-4-arylmethylene-2-phenyl5(4H)-oxazolones (2a, 2b)57,58
Chapter 2
71
A mixture of N-benzoylglycine, I (1.0 g, 5.6 mmol, 1 equiv.), p-disubstituted
aromatic aldehyde, IIa-b (IIa for DPI and IIb for DPPI; 5.6 mmol, 1 equiv.),
anhydrous sodium acetate (0.46 g, 5.6 mmol, 1 equiv.) and acetic anhydride (2.6
mL, 28.0 mmol, 5 equiv.) was refluxed at 160 ºC for 2 h. The reaction mixture was
cooled to room temperature and 10 mL ethanol was added and kept at 0 °C for 20
minutes. The deposited solid was filtered under suction and washed with ice-cold
ethanol (10 mL 2) and warm water (50 mL). The crude product obtained was
dried in vacuum desiccator. The crude product was purified by a short pad of silica
gel chromatography (60-120 mesh).
2.6.2.1. Results
(4Z)-4-(4-N,N-Dimethylaminobenzylidene)-2-phenyl-5(4H)-oxazolone (IIIa).
Orange powder; Yield 54%; m.p. 182-183 °C; 1H NMR (CDCl3, 400 MHz): 3.12
(s, 6H, NMe2), 6.75 (d, J = 8.7 Hz, 2H, ArH), 7.21 (s, 1H, =CHAr), 7.48-7.57 (m,
3H, ArH), 8.13-8.16 (m, 4H, ArH).
(4Z)-4-(4-N,N-Diphenylaminobenzylidene)-2-phenyl-5(4H)-oxazolone (IIIb).
Red powder; Yield: 3.68 g (69%); m.p. 168-170 °C; Rf 0.5 (20% Ethyl acetatePetroleum ether). IR (KBr) max/cm-1: 1706, 1635, 1584, 1491, 1426, 1285, 1152.
1
H NMR (CDCl3, 400 MHz):  7.05 (d, 2H, J = 9.8 Hz, ArH), 7.13-7.20 (m, 7H,
=CHAr + ArH), 7.32-7.35 (m, 4H, ArH), 7.47-7.51 (m, 2H, ArH), 7.54-7.58 (m,
1H, ArH), 8.07 (d, 2H, J = 8.7 Hz, ArH), 8.13 (d, 2H, J = 8.8 Hz, ArH). 13C NMR
(CDCl3, 100 MHz): δ 120.5, 124.9, 126.1, 126.2, 126.4, 128.2, 128.9, 129.7, 130.6,
132.0, 132.9, 134.1, 146.5, 150.8, 162.0, 168.2.
ESI-MS+ m/z Calcd for
C28H21N2O2: 417.1603 [M+H], found 417.1607.
2.6.3. General procedure for preparation of (4Z)-4-arylmethylene-1,2diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPI and DPPI)59-61.
Chapter 2
72
A mixture of oxazolone (0.1 g, 0.34 mmol), freshly distilled aniline (0.038 mL,
0.41mmol) and anhydrous zinc chloride (0.005g, 0.034 mmol, 10 mol%) was fused
at 170 °C under nitrogen atmosphere for 2 h. The reaction mixture was cooled to
room temperature and quenched with water. The organic matter was extracted with
ethyl acetate (20 mL 3) and washed with saturated sodium bicarbonate solution.
The extract was dried over anhydrous sodium sulphate and concentrated to get
brown colored crude product. It was further purified by silica gel (100-200 mesh)
column chromatography.
2.6.3.1. Results
(4Z)-4-(4-N,N-Dimethylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5Himidazolin-5-one (DPI).
Isolated yield: 0.084 g (67%); m.p. 198-200 °C; Rf 0.5 (30% Ethyl acetatePetroleum ether). IR (KBr) max/cm-1: 1702, 1631, 1607, 1531, 1492, 1323, 1292,
1197. 1H NMR (CDCl3, 500 MHz): δ 3.09 (s, 6H, N(CH3)2), 6.75 (d, J = 8.8 Hz,
2H, ArH), 7.19 (d, J = 7.7 Hz, 2H, ArH), 7.26 (s, =CHAr + CHCl3), 7.28 -7.31 (m,
3H, ArH), 7.36 -7.42 (m, 3H, ArH), 7.55 (d, J = 8.4 Hz, 2H, ArH), 8.23 (d, J = 8.5
Hz, 2H, ArH). 13C NMR (CDCl3, 125 MHz): δ 40.2, 111.9, 122.7, 127.5, 128.1,
128.3, 129.1, 129.4, 129.6, 130.7, 131.1, 134.7, 135.0, 135.3, 151.9, 157.2, 170.6.
ESI-MS+ m/z Calcd. for C24H22N3O: 368.1762 [M+H], found 368.1760.
(4Z)-4-(4-N,N-Diphenylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5Himidazolin-5-one (DPPI)
Isolated yield: 0.073 g (60%); m.p. 216-218 °C; Rf 0.45 (20% Ethyl acetatePetroleum ether). IR (KBr) max/cm-1: 1712, 1634,1585, 1491, 1446, 1285, 1163.
1
H NMR (CDCl3, 500 MHz): δ 7.06 (d, J = 8.8 Hz, 2H, ArH), 7.12 (t, J = 7.5 Hz,
2H, ArH), 7.17 -7.19 (m, 6H, =CHAr + 5ArH), 7.28 -7.33 (m, 7H, ArH), 7.36 7.44 (m, 4H, ArH), 7.53 (d, J = 7.5 Hz, 2H, ArH), 8.16 (d, J=8.6 Hz, 2H, ArH). 13C
NMR (CDCl3, 125 MHz): δ 129*2, 124.5, 125.9, 127.5, 128.3, 128.4, 129.2, 129.5,
129.6*2, 131.0*2, 134.1, 135.1, 136.5, 146.7, 150.1158.7, 170.6. CHN Analysis:
Chapter 2
73
%C 83.36 (83.07), %H 5.33 (5.13), %N 8.60 (8.55), %R 2.71 (3.25), where the
values in parenthesis are calculated. ESI-MS+ m/z Calcd. for C34H26N3O:
492.2075 [M+H], and experimentally obatined value is 492.2076.
Chapter 2
74
References
1. Mahr, H.; Hirsh, M. D. Opt. Comm. 1975, 13, 96.
2. Kahlow, M. A.; Jarzeba, W.; DuBruil, T. P.; Barbara, P. F. Rev. Sci.
Instrum. 1988, 59, 1098.
3. Ó Connor, D. V.; Phillips, D. Time Correlated Single Photon Counting,
Academic Press: New York, 1984.
4. Fleming, G. R. Chemical Applications of Ultrafast Spectroscopy; Oxford
University Press: New York, 1986.
5. Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd Edition,
Springer, 2006.
6. Tkachenko, N. V. Optical Spectroscopy Methods and Instrumentation,
Elsevier, 2006.
7. Bevington, P. R. Data Reduction and Error Analyses for the Physical
Sciences, McGraw-Hill, 1969.
8. Maroncelli, M.; Fleming, G. R. J. Chem. Phys. 1987, 86, 6221.
9. Eastern, J. H.; DeTomma, R, P.; Brand, L. Biophys, J. 1976, 16, 571.
10. Badea, M. G.; Brand, L. Methods Enzymol. 1979, 61, 378.
11. Femtochemistry and Femtobiology, Sundström, V. (ed.), World Scientific,
Singapore, 1997.
12. Femtochemistry, Chergui, M. (ed.), World Scientific, Singapore, 1996.
13. Schreiber, E. Femtosecond Real-Time Spectroscopy of Small Molecules and
Clusters, Springer, New York, 1998.
14. Zewail, A. H. J. Phys. Chem. 2000, 104, 5693; and references therein.
15. Porter, G. Femtosecond Chemistry, Manz, J., and Wöste, L. (eds.) Vol. 1,
VCH Weinheim, 1995, p. 3.
16. Arlt, T.; Schmidt, S.; Kaiser, W.; Lauterwasser, C.; Meyer, M.; Scheer, H.;
Zinth, W. Proc. Natl. Acad. Sci. USA 1993, 90, 794.
17. Kennis, J. T. M.; Shkuropatov, A. Y.; Van Stokkum, I. H. M.; Gast, P.;
Hoff, A. J.; Shuvalov, V. A.; Aartsma, T. J. Biochemistry 1997, 36, 16231.
Chapter 2
75
18. Cerullo, G.; Polli, D.; Lanzani, G.; De Silvestri, S.; Hashimoto, H.; Cogdell,
R. J. Science 2002, 298, 2395.
19. Nishimura, K.; Rondonuwu, F. S.; Fujii, R.; Akahane, J.; Koyama, Y.;
Kobayashi, T. Chem. Phys. Lett. 2004, 392, 68.
20. Streltsov, A. M.; Vulto, S. I. E.; Shkuropatov, A. Y.; Hoff, A. J.; Aartsma,
T. J.; Shuvalov, V. A. J. Phys. Chem. B 1998, 102, 7293.
21. Vos, M. H.; Breton. J.; Martin, J. L. J. Phys. Chem. B 1997, 101, 9820.
22. Durrant, J. R.; Hasting. G.; Joseph, D. M.; Barber, J.; Porter, G.; Klug, D. R.
Proc. Natl. Acad. Sci. 1992, 89, 11632.
23. Groot, M. L.; Van Maurik, F.; Eijckelhoff, C.; Van Stokkum, I. H. M.;
Dekker, J. P.; Van Grondelle, R. Proc. Natl. Acad. Sci. 1997, 94, 4389.
24. Mukamel, S. Principles of Nonlinear Optical Spectroscopy, Oxford
Univeristy Press, New York, 1995.
25. Novoderezhkin, V. I.; Palacios, M. A.; Van Amarongen, H.; Van Grondelle,
R. J. Phys. Chem. B 2004, 108, 10363.
26. Kovalenko , S. A.; Dobryakov, A. L.; Ruthmann, J.; Ernsting, N. P. Phys.
Rev. A 1999, 59, 2369.
27. Snellenburg, J. J.; Laptenok, S. P.; Seger, R.; Mullen, K. M.; van Stokkum,
I. H. M. J. Stat. Soft. 2012, 49, 1.
28. Van Stokkum, I. H. M.; Larsen, D. S.; van Grondelle, R. Biochim. Biophys.
Acta 2004, 1657, 82.
29. Berera, R.; van Grondelle, R.; Kennis, J. T. M. Photosynth. Res. 2009, 101,
105.
30. Ruckebusch, C.; Sliwa, M.; Pernot, P.; de Juan, A.; Tauler, R. J. Photochem.
Photobiol. C: Photochem. Rev. 2012, 13, 1.
31. Niedzwiedzki, D. M.; Sullivan, J. O.; Polivka, T.; Birge, R. R.; Frank, H. A.
J. Phys. Chem. B 2006, 110, 22872.
32. Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, 864.
33. Gross, E. K. U.; Dobson, J. F.; Petersilka, M. Top. Curr. Chem. 1996, 181,
81.
Chapter 2
76
34. Van Leeuwen, R. J. Mod. Phys. B 2001, 83, 4361.
35. Onida, G.; Reining, L.; Rubio, A. Rev. Mod. Phys. 2002, 74, 601.
36. Thouless, D. J. The Quantum Mechanics of Many-Body Systems, vol. 11 of
Pure and Applied Physics, Academic Press, New York, 2nd Edn., 1972.
37. Tretiak, S.; Chernyak, V. J. Chem. Phys. 2003, 119, 8809.
38. Runge, E.; Gross, E. K. U. Phys. Rev. Lett. 1984, 52, 997.
39. Gross, E. K. U.; Ullrich, C. A.; Gossman, U. J. Density Functional Theory,
vol. 337 of NATO ASI, Ser. B, Plenum Press, New York, 1995.
40. Van Leeuwan, R. Phys. Rev. Lett. 1998, 80, 1280.
41. Gross, E. K. U.; Kohn, W. Adv. Quantum Chemistry 1990, 21, 255.
42. Becke, A. J. Chem, Phys. 1993, 98, 1372.
43. Becke, A. J. Chem, Phys. 1993, 98, 5648.
44. Becke, A. D. Phys. Rev. A 1988, 38, 3098.
45. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
46. Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
47. Tozer, D. J.; Handy, N. C. J. Chem. Phys. 1998, 109, 10180.
48. Cai, Z. –L.; Sendt, K.; Reimers, J. R. J. Chem. Phys. 2002, 117, 5543.
49. Drew, A.; Dunietz, B. D.; Head-Gordon, M. J. Am. Chem. Soc. 2002, 124,
12070.
50. Tozer, D. J.; Arnos, R. D.; Handy, N. C.; Roos, B. J.; Serrano-Andres, L.
Mol. Phys. 1999, 97, 859.
51. Sobolewski, A. L.; Domcke, W. Chem. Phys. 2003, 294, 73.
52. Amovilli, C.; Barone, V.; Cammi, R.; Cancès, M.; Cossi, M.; Mennucci, B.;
Pomelli, C. S.; Tomasi, J. Adv. Quantum Chem. 1998, 32, 264.
53. Hall, R. J.; Davidson, M. M.; Burton, N. A.; Hillier, I. H. J. Phys. Chem.
1995, 99, 921.
54. Namazian, M. J. Mol. Struct. (Theochem.) 2003, 664, 273.
55. Perrin D. D.; Armarego W. L. F. Purification of Laboratory Chemicals, 4th
ed.; Butterworth- Heinemann, 1997.
56. Lai G.; Bu X. R.; Santos J.; Mintz E. A. Synlett 1997, 1275.
Chapter 2
77
57. Hoshina H.; Tsuru H.; Kubo K.; Igarashi T.; Sakurai T. Heterocycles 2000,
53, 2261.
58. Rafiq S.; Rajbongshi B. K.; Nair N. N.; Sen P.; Ramanathan G. J. Phys.
Chem. A 2011, 115, 13733.
59. Badr M. Z.; El-Sherief H. A. H.; Tadros M. E. Bull. Chem. Soc. Jpn. 1982,
55, 2267.
60. Rajbongshi B. K.; Ramanathan G. J. Chem. Sci. 2009, 121, 973.
61. Mukerjee A. K.; Kumar P. Heterocycles 1981, 16, 1995.
78
Chapter 2
Chapter 3
Excited State Relaxation Dynamics of
4-Nitrophenyl Pyrrolidinemethanol
Shahnawaz Rafiq et al. J. Phys. Chem. A 2011, 115, 8335.
80
Chapter 3
This chapter reports the ultrafast excited state relaxation dynamics of an NLO
dye, (S)-(−)-1-(4-Nitrophenyl)-2-pyrrolidinemethanol, carried out under the
regime of femtosecond fluorescence up-conversion measurements in augmentation
with quantum chemical calculations. The primary concern was to trace the
relaxation pathways which guide the depletion of first singlet excited state, in such
a way that it is virtually non-fluorescent. Ground and excited states (singlets and
triplets) potential energy surfaces were calculated as a function of –NO2 torsional
coordinate, which revealed the perpendicular orientation of –NO2 in the excited
state minimum relative to the planar ground state conformation. The fluorescence
transients in the femtosecond regime show bi-exponential decay behavior. The first
time component of few hundred femtoseconds was ascribed to the ultrafast twisted
intramolecular charge transfer (TICT). The occurrence of charge transfer is
substantiated by the large dipole moment change during excitation. The
construction of intensity and area normalized time resolved emission spectra
(TRES and TRANES) of NPP in acetonitrile exhibited a two state emission on
behalf of decay of LE state and rise of CT state with a Stokes shift of 2000 cm-1
over a time scale of 1 ps. The second time component of few picoseconds is
attributed either to internal conversion (IC) or to intersystem crossing (ISC)
depending upon the polarity of the medium, with intersystem crossing dominating
over internal conversion in non-polar solvents and vice versa in polar solvents.
Chapter 3
81
3.1. Introduction
Molecules exhibiting nonlinear optical (NLO) properties have received
significant attention because of their possible applications as nonlinear optical
devices in the fields of spectroscopy, telecommunications, optical data storage and
processing etc.1-3 In general, second-order NLO active chromophores consist of
strong electron donating (D) and electron withdrawing (A) moieties coupled via a
-conjugated spacer. Such kind of D –  bridge – A combinations are commonly
referred to as “push-pull” systems.1-6 It is believed that the capability of a molecule
to exhibit NLO phenomena is related to the degree of charge separation and this
charge redistribution takes place on an ultrafast timescale.7-9 During the past few
decades, the excited state intramolecular charge transfer (ICT) dynamics of such
type of “push-pull” systems have been the subject of many experimental and
computational studies.7-9 The important thing about twisted intramolecular charge
transfer (TICT) processes is that the pi-orbitals of donor and acceptor moieties are
oriented at some angle with respect to each other. 7-9 Photo-excitation causes the
formation of locally excited (LE) singlet state, which relaxes to the more polar
TICT singlet state and emission is thus an outcome of both the states.7-11 The
widely studied compound in this regime is dimethylaminobenzonitrile, which
shows a very prominent TICT emission.7-9 There are many other molecules which
have been studied till date to investigate the charge transfer dynamics, but very
little concern has been paid towards studying the overall excited state dynamics of
compounds containing nitro group as an acceptor moiety separated from the donor
group through a -conjugated spacer.
The occurrence of nonbonding electrons in many substituted aromatic
hydrocarbon systems govern the presence of upper excited states, which in some
cases significantly increase the spin-orbit coupling and hence drastically change
the photophysics and photochemistry of the molecule.12-17 This feature essentially
hastens the intersystem crossing (ISC) channels rendering such aromatic molecules
highly nonfluorescent with fluorescence quantum yields of ca. 10 -3. Thus the
excited state process will mainly occur from triplet state instead of singlet. 17,18 In
Chapter 3
82
1980, Hamanoue and coworkers first suggested the involvement of upper n-*
triplet states as a receiver of electrons from singlet state S1 during ISC in 9nitroanthracene.12,13 The first identification of upper triplet states with adequate
electronic configuration to couple to the S1 state in nitroaromatics was shown by
Peon and coworkers.19 The spin-orbit coupling due to the presence of –NO2 group
was assigned to the observed lack of fluorescence and high triplet yield of this
molecule.12,13 Also, the electron withdrawing nature of –NO2 group permits the
extension of ring aromaticity and hence helps in stabilizing the electronically
excited states, and as well imparting a charge transfer character.17,19-21 Transient
absorption measurements on several nitroaromatic systems have inferred few
picosecond build up time for triplet population.22,23 Ernsting and co-workers
probed the deactivation kinetics of p-nitroaniline through transient absorption
technique and semi-empirical calculations. They revealed that internal conversion
process was controlled by the twisting motion of –NO2 group to the perpendicular
conformation towards the local minimum of the potential energy surface along the
torsional coordinate.24,25 In 2007, Peon et al. measured the transient fluorescence
from several nitrated polycyclic aromatic hydrocarbons and proposed ISC as the
main decay channel of S1 state with the time duration of few tens of
femtoseconds.26 Recently Crespo-Hernández et al. in 2008, based on transient
absorption measurements and quantum chemical calculations, reported an ISC time
constant of ~7 ps corresponding to the decay from * S1 state to n* T3 state for
1-nitropyrene.21
The orientation of nitro group has been suggested to be
responsible for controlling the rate of ISC from singlet to triplet states. The
torsional angle of nitro group upon excitation changes its orientation relative to the
ground state which increases the efficiency of spin orbit coupling between the
singlet and triplet states and hence instead of showing fluorescence, the molecules
decays through intersystem crossing channel.19
Apart from experimental findings, many computational studies have been as
well devoted to understand the photophysics of nitrated aromatic systems. Time
dependent-density functional theory (TD-DFT) was employed to estimate the
Chapter 3
83
relative energies of singlet and triplet manifolds which assisted in correlating the
rate of ISC with the energy alignment of singlet and triplet states. 16,17,21 From these
calculations, Crespo-Hernández et al. observed that in 1-nitropyrene, there is a
local minima at nitro group torsional angle of 27.46 in the fully optimized ground
state, however in the first singlet excited state the nitro group is almost coplanar
with the aromatic system.21
In this chapter, we present a detailed femtosecond fluorescence up-conversion
measurements and quantum chemical calculations on (S)-(−)-1-(4-Nitrophenyl)-2pyrrolidinemethanol (NPP) (Figure 3.1). NPP is a nitrated push-pull molecule
exhibiting nonlinear optical properties. The emphasis is being given to understand
its excited state relaxation mechanism.
3.2. Results
3.2.1. Steady State Absorption and Emission Measurements
NPP
exhibits
a
single
broad
structureless absorption band in the
solution phase. Figure 3.2(a) shows the
steady state absorption spectra of NPP in
four solvents of different polarity. In
nonpolar tetrahydrofuran, the absorption
Figure 3.1. Structure of NPP.
maximum was centred at 392 nm.
Increasing the polarity of the medium shifts the absorption maximum
monotonically towards red side and in case of water it was found at 422 nm,
reflecting a strong positive solvatochromism. The emission spectra (shown in
figure 3.2(b)) also manifest strong solvatochromism as is dictated by the
appreciable bathochromic shift of emission maxima with increasing solvent
polarity. In tetrahydrofuran, the emission maximum is centred at ca. 453 nm, while
as in water it is ca. 520 nm. In acetonitrile, NPP shows the molar extinction
coefficient of 22800 M-1cm-1 and the oscillator strength was calculated to be 0.33.
The fluorescence quantum yield of NPP was found to be ~ 51  10-4 in
84
Chapter 3
acetonitrile using Coumarin 153 as reference, which indicates its highly nonfluorescent nature. The observed strong positive solvatochromism is an indicative
of larger dipole moment change during photo-excitation.
The ground and excited state dipole moments were calculated using the
solvatochromic theory based on Onsager solvent reaction field, established by
Kawski et al.,27a-d whose applicability has been authenticated on a large number of
Figure 3.2. Steady state absorption spectra (a) and fluorescence spectra (b) of NPP in four
different solvents; tetrahydrofuran (black, ― - - ―), ethanol (blue, ― ― ―), acetonitrile
(green, ············) and water (red, ———). The respective absorption maxima are 391, 392,
399 and 422 nm and the respective emission maxima are 453, 458, 486, and 523 nm. The
broken region in each of the emission spectrum is the solvent Raman peak which was
manually deleted in order to make spectra look clear.
molecules.28a-e The theory differs from the Lippert-Mattaga29 propositions, which
considers the polarizability of both ground and excited state to be zero ( = 0).
Kaswki et al. did not neglected the polarizability, rather the isotropic polarizability
() of the solute was introduced as 2/a3 = 1 (where, a is Onsager cavity radius)
and following solvent reaction field factors were deduced;
(3.1)
(3.2)
These factors depend upon dielectric constant,
and refractive index, n of the
solvent. The steady state absorption and emission characteristics of NPP were
Chapter 3
plotted as a function of these solvent polarity parameters
85
and
as shown in figure 3.3(a) and 3.3(b) respectively, following below
mentioned equations
 a  f  m1 f ( , n)  c
(3.3)
 a  f  m2[ f ( , n)  2 g (n)]  c
(3.4)
Figure 3.3. Plots of (a) Stokes shift

a
 vs. solvent polarity parameter f  ,n (b) sum of
 vs. solvent polarity parameter f  ,n  2gn. The
 f
absorption and emission maxima  a  f
labels shown in the graph are solvents (1) dimethylsulfoxide, (2) acetonitrile, (3) methanol,
 (8) octanol and (9) water
(4) ethanol, (5) iso-propanol, (6) butanol, (7) dimethylformamide,
used in this study. The respective slopes are m1 = 5360 and m2 =15850.
The respective slopes are m1 = 5360 cm-1 and m2 = 15850 cm-1. Based on above
solvent polarity parameters, the ground state dipole moment is given by;
(3.5)
Assuming the symmetry of the investigated solute molecule remains unchanged
upon electronic transition, and also the ground and excited state dipole moments
are parallel (also predicted by TD-DFT calculations) which renders the
polarizability of both states similar, the excited state dipole moment can be
calculated as;
,
(m2 > m1)
(3.6)
Chapter 3
86
It was also observed by Kawski et al.31d that the shape of molecule has not a
significant effect on the determined values of dipole moments, however the dipole
moments does depend upon the size of the Onsager cavity. In the present case, the
Onsager cavity radius was taken as 50 % of the long axis of the optimized ground
state structure of NPP, which yielded a = 4.76 Å. By substituting the value of all
the parameters in equations (3) and (4), the dipole moment values obtained are; g
= 7.4 D and e = 14.9 D with a change in dipole moment of 7.5 D. The large
transition dipole moment of NPP suggests the involvement of charge separation in
its electronic excited state.
3.2.2. Femtosecond Fluorescence Up-Conversion Study
Femtosecond time resolved fluorescence transients30 were recorded by exciting
NPP at 405 nm. The fluorescence signal decays on timescales ranging from
hundred femtoseconds to a few picoseconds depending upon the nature of the
solvent. In case of water, the fluorescence transient at 500 nm was fitted to a biexponential function with 1 = 250 fs (96%) and 2 = 690 fs (4%) (Table 3.1). In
acetonitrile as well, NPP decays bi-exponentially with 1 = 350 fs (80%) and 2 =
900 fs (20%) at 490 nm. On decreasing the polarity, the decay time constants
increased and for tetrahydrofuran (at 500 nm), the respective time constants are 1
= 650 fs (84%) and 2 = 2.22 ps (16%). The femtosecond transients are shown in
figure 3.4(a). We also testified the compliance with few longer chain alcohols as
Table 3.1. Fitting parameters for fluorescence transients of NPP in different solventsa
Solvent
Water
Acetonitrile
Ethanol
Butanol
Octanol
Tetrahydrofuran
a
81.00
35.50
24.55
17.43
09.87
07.40
a1
1 (ps)b
a2
2 (ps)b
0.96
0.96
0.86
0.81
0.68
0.84
0.25
0.35
0.56
0.62
0.65
0.65
0.04
0.04
0.14
0.19
0.32
0.16
0.69
0.90
1.67
2.00
2.72
2.22
Data are described by double exponential functions, convoluted with the instrument
response function. The sum of amplitudes a1 and a2 is normalized to one. b 0.10 ps
Chapter 3
87
solvent in our measurements as mentioned in table 3.1 and transients are shown in
figure 3.4(b). In all alcohols studied, the data was fitted bi-exponentially. It is also
observed that with increase in length of carbon chain, the magnitude of time
constants increases. In order to have a vivid knowledge about the ultrafast time
dependent dynamics following excitation to Franck-Condon state, the decay
transients of NPP in acetonitrile were measured at wavelengths ranging over whole
of the emission spectra from 440 to 590 nm. From the fitting parameters, time
resolved intensity and area normalized emission spectrum (TRES and TRANES)
were constructed.
Figure 3.4. Femtosecond time resolved fluorescence transients of NPP in different solvents;
(a) water (red), acetonitrile (green), ethanol (blue) and tetrahydrofuran (black) and (b) ethanol
(blue), butanol (pink) and octanol (orange). All the traces are fitted by a sum of two
exponentials. Solid lines indicate fitted decays convoluted with the instrument response
function.
3.2.3. Quantum Chemical Studies of Ground and Excited States
The ground state of NPP was optimized in Gaussian 03 by using B3LYP/6311++G(d,p) functional.31 The optimized structure of NPP is characterized by the –
NO2 dihedral angle of –0.14. The single point energy calculations of ground state
were performed for the torsional motion of –NO2 group from its equilibrium planar
position with an interval of 10 away from the equilibrium. The corresponding
excited state energies were computed with the TD-DFT framework under the
regime of vertical excitations. All the potential energy values were plotted as a
Chapter 3
88
function of torsional angle of –NO2 group considering the ground state optimized
energy as the zero energy. The S0, S1, T1, T2 and T3 potential energy surfaces are
shown in figure 3.6(a) (vide infra). The shape of the frontier molecular orbitals of
NPP corresponding to the ground state, Frank-Condon state and the relaxed state
provide very substantial information regarding the distribution of charge in each
state. Figure 3.7 (vide infra) shows the HOMO of B3LYP/6-311++G(d,p)
optimized ground state, LUMO of vertically excited Frank-Condon state using TDDFT/ B3LYP/6-311++G(d,p) basis and LUMO of the excited state configuration
which represents minimum in the S1 potential energy surface. Vertical excitation
energies of fully optimized NPP structure were also computed in different solvents
using TD-PBE0/NE-IEFPCM/6-311++G(d,p)B3LYP/IEFPCM/6-311++G(d,p)
level of theory.32,33 The relative potential energy values of S0, S1, and T1, T2 and T3
in vacuum and four different solvents (THF, ethanol, ACN and water) are
mentioned in table 3.2 and also shown in figure 3.6(b). These single point potential
energy values represent the potential energy of respective states at the FranckCondon non-equilibrium level and not the solvent relaxed energy values.33
Qualitatively, it was observed that as the polarity of solvent increases from vacuum
to water, the energy states are stabilized to different extents in different solvents in
such a way that as we reach to the high dielectric medium water, S 1 and T2 energy
states have almost became degenerate.
3.3. Discussions
Steady state measurements infer that there is an appreciable change in the
dipole moment of NPP on photo-excitation because of charge transfer. The
fluorescence quantum yield is found to be very small (51  10-4) in acetonitrile
suggesting the involvement of non-radiative relaxation pathways. The femtosecond
time resolved fluorescence study reveals very fast decay kinetics in all the solvents
studied with two distinct ultrafast time components. Following the previous
interpretation on various nitrated polyaromatic hydrocarbons and paranitroaniline,10,11,17,26 we propose the bi-exponential decay of S1 state in NPP is
Chapter 3
89
because of the relaxation of Frank-Condon state through many stages. This
relaxation is a consequence of some molecular rearrangements, which direct the
excited state to decay extremely fast just after the photo-excitation, subsequently
followed by slow decay kinetics. Sub-picosecond components in the fluorescent
transients are often assigned to processes like internal conversion, charge transfer,
solvation, vibrational redistribution and cooling.25,34,35 The fastest process is the
intramolecular vibrational redistribution of high frequency internal modes
developing on a 10 fs time duration which we cannot observe under the prevailing
experimental constraints.25,36,37 The processes between 100 fs to few picoseconds
in NPP include intramolecular charge transfer and intersystem crossing as a
consequence of torsional motion of –NO2 group during its equilibration to
intramolecular charge transfer state.25
Femtosecond fluorescence up-conversion measurements of NPP in acetonitrile
have been performed at eleven different wavelengths throughout the emission
spectrum to authenticate the proposition about excited state intramolecular charge
transfer, which exhibit strong wavelength dependent fluorescence transients. On
the blue side of the emission spectra, the decay is bi-exponential in nature. The fast
time constant constitutes the major part of the relaxation process. For example, at
440 nm, the observed components are 1 = 60 fs (0.96) and 2 = 470 fs (0.04%).
While at the red end, 590 nm, the decay of time constant 520 fs is preceded by a
distinct rise constant of 80 fs. A few representative decay profiles of NPP in
acetonitrile solution are shown in figure 3.5(a).
The decay component at the short wavelength region accompanied by the fast
rise component at the longer wavelength region gives clear evidence of the
transition from the LE to CT state. The time resolved intensity and area normalized
emission spectra (TRES and TRANES) were constructed using the parameters of
best fit to the fluorescence decays and the steady state emission spectra emission
(Figure 3.5(b) and 3.5(c)).38a,b The TRES exhibits a dynamic Stokes shift of 2000
cm-1 within 1 ps indicating that the charge transfer dynamics is a very fast in NPP.
The emission in the very early time reflects the existence of LE state
Chapter 3
90
Figure 3.5. (a) Few representative femtosecond up-conversion decay transients of NPP in
acetonitrile at wavelengths; 450 nm (red), 460 (blue) 470 nm (black), 490 (green) and 590 nm
(pink). The decay at 570 nm shows an initial ~ 100 fs rise time followed by a 800 fs decay
time. (b) Time resolved emission spectra (TRES) and (c) Time resolved area normalized
emission spectra (TRANES) of NPP in acetonitrile constructed from the best fit parameters
of fluorescence decays and emission intensity at different times following excitation.
and in the latter time a distinct charge transfer emission is observed. TRANES
clearly shows the presence of iso-emissive point indicating the LE to CT state
transformation.
Based on previous literature and from our results, we propose, –NO2 torsional
motion constitutes the reaction coordinate for the excited state relaxation dynamics
in NPP.21,25,39 The nature of the potential energy surfaces of NPP were calculated
as a function of torsional angle of –NO2 group (see Figure 3.6(a)). It suggests that
the vertical excitation of the ground state optimized geometry leads to the excited
singlet state having very high energy, corresponding to the directly excited
nonrelaxed nuclear configuration. The ground state energy of NPP keeps on
Chapter 3
91
Figure 3.6. (a) Potential energy surfaces of ground state (Black, ), first singlet excited state,
S1 (Red, •), triplet excited states; T1 (blue, ▲), T2 (green, ♦) and T3 (orange, ▼) plotted as a
function of –NO2 group torsional angle in NPP. The ground state potential energy values are
calculated with the basis B3LYP/6-311++G(d,p) for every torsional angle of –NO2 group.
The excited state potential energy value corresponding to every torsional angle are obtained
considering vertical excitation using TD-DFT/B3LYP/6-311++G(d,p). (b) Vertical transition
energies of fully optimized NPP structure in vacuum and in presence of various solvent
media of different.
increasing as a –NO2 group dihedral angle change from its equilibrium geometry.
However, the energy of singlet excited state decreases until the minimum in first
singlet excited state potential energy surface is reached. This minimum
corresponds to the relaxed nuclear configuration and is represented by the –NO2
group perpendicular configuration of the torsional angle. Thus, the S 1 state is
characterized by the presence of a potential well at the torsional angle of – NO2
group, which renders its orientation perpendicular to the phenyl ring relative to its
stable planar configuration in the ground state. As shown in figure 3.6(a), the
rotation of –NO2 group along the twist coordinate results in the formation of a
prominent conical intersection (CI) between the S1 and T3 states. This CI occurs at
the torsional angle of ca. 32. These calculations also govern the presence of
another CI between the S1 and T2 states at ca. 70. Based on these observations, we
propose that NPP upon photo-excitation to the Franck-Condon singlet state loses
some energy and moves down the barrierless potential energy surface towards the
less energetic relaxed state through the twisting motion of –NO2 group. Moving
92
Chapter 3
down the barrierless surface leads to the occurrence of CI between singlet state S1
and the triplet state T3 and T2. The CI between S1 and T3 or T2 states may furnish
an essential nonradiative pathway for the excited state relaxation of S1 to triplet
manifold involving intersystem crossing (ISC). Also, the presence of a minimum in
the singlet state at perpendicular configuration may open channel for internal
conversion with the ground state. The fate of the excited state will thus
predominantly be decided by the relative stability of the electronic state at various
nuclear configurations, which will be controlled by the nature of the electronic
state. Thus according to the present TD-DFT calculations, excited state
intramolecular charge transfer and non-radiative channels occur along the
barrierless torsional motion of –NO2 group. However the location of conical
intersection calculated using TD-DFT is not certain. The conical intersection is
being predicted between adiabatic potential energy surfaces, while as the main
feature of conical intersection is its non-adiabatic nature and hence this may
eventually predict CI with less certainty.40 It is also important to know that the
local functionals used within TD-DFT do not allow for sufficient accuracy to
describe long range charge transfer transitions as well.41
The frontier molecular orbitals of NPP in the ground and excited states reveal
significant information about the distribution of charge density. As shown in figure
3.7, the distribution of charge density is altogether different relative to each other.
HOMO of ground state is characterized by the uniform spread of charge density.
However in the LUMO of relaxed state the electron density is shifted more
towards –NO2 group relative to the LUMO of nonrelaxed excited state. Such kind
of gradient of electron density distribution in the excited state suggests a
significant charge transfer from pyrrolidine moiety to the –NO2 group. This charge
transfer lies along the twisted coordinate of –NO2 group and is thus known as
excited state twisted intramolecular charge transfer (TICT) process.
In order to understand the fate of the singlet excited state, once it has reached
to the minimum of the potential well, opted to visualize the influence of dielectric
properties of solvents on the relative stabilities of the electronic states. The
Chapter 3
93
Figure 3.7. Frontier molecular orbitals of NPP: (a) HOMO of B3LYP/6-311++G(d,p)
optimized ground state. (b) LUMO of vertically excited Frank-Condon state. (c) LUMO of
the excited state NPP configuration representing minimum in the S1 potential energy surface.
Table 3.2. Vertical excitation energy values of S1, T1, T2 and T3 of a fully optimized NPP
ground state structure in solvents of different dielectric properties.
Solvent
S1 (eV)
T1 (eV)
T2 (eV)
T3 (eV)
S1-T2 (eV)
Tetrahydrofuran
3.30
2.27
3.16
3.50
0.14
Ethanol
3.26
2.20
3.21
3.54
0.05
Acetonitrile
3.25
2.18
3.22
3.55
0.03
Water
3.24
2.17
3.23
3.56
0.01
solvents are tabulated in table 3.2 and also shown in figure 3.6(b). Qualitatively, it
was observed that as the polarity of medium increases from vacuum to water, S 0,
S1 and T1 gain stability, while as the energy of T2 and T3 increases. Such a
variation in the energy of these electronic states with increasing solvent polarity
throws light on the symmetry of these states. From the fundamental knowledge, the
energy of the  - * state decreases with increasing solvent polarity, while as the
energy of n - * state increases with solvent polarity. Such a dependence helps us
to assigns  - * symmetry to S0, S1 and T1 and n - * symmetry to T2 and T3
electronic states. As per El-Sayed rules,42 the intersystem crossing between the two
electronic states is possible only if they are having opposite symmetry, which in
this case is possible between the minimum in S1 state and the T2 state. It can be
visualized as, once the molecule reaches to the minimum in the S1 potential energy
surface, due to the intersection of the oppositely symmetric S 1 and T2 states as
Chapter 3
94
shown in figure 3.6(a), intersystem crossing will be the essential non-radiative
pathway. With increase in solvent polarity, since the energy of S1 state will
decrease and that of T2 state will increase creating strong avoiding and hence will
restrict any kind of intersystem crossing between the two states, rather will bring
the minimum of singlet state more closer to the ground state perpendicular
configuration and will open internal conversion as the main non-radiative channel
in the polar solvents.
Based on above discussion, we assign the observed first subpicosecond time
component in the fluorescence transients to the twisted intramolecular charge
transfer from pyrrolidine moiety to the –NO2 group. In non-polar tetrahydrofuran,
the fast component (1) is 650 fs. As we increase the polarity from THF to water, 1
decreases to 250 fs as shown in table 3.1. This decrease in the magnitude of time
constant with increase in solvent polarity is ascribed to the stability of charge
transfer state in the highly polar solvents compared to that in non polar solvents
and hence substantiates our proposition of assigning the first time component to
the charge transfer process in the excited state. It is also observed that, as we
increase the viscosity of the medium from ethanol (1.087 cP) to n-octanol (7.21
cP),29,43 although the polarities are not much different, 1 increases from 560 to 650
fs respectively. The change in magnitude of time constant with change in viscosity
of medium is accredited to the large amplitude motion of –NO2 group, which
shows viscosity dependence and has been proposed earlier for a number of
molecules.44-46 While the first component is the twisted intramolecular charge
transfer of few hundred femtoseconds, the second time component is assigned
either to the ISC from S1 state to triplet manifold or to the internal conversion
process between S1 and S0. (Scheme 3.1) In non-polar solvents intersystem
crossing is the main non-radiative channel, while as in polar solvents, internal
conversion will be the governing non-radiative channel. In THF, being the nonpolar solvent, the observed time constant (2) is 2.2 ps and as the medium is
changed to more polar water, the time constant decreases to 0.69 ps showing a
strong dependence on polarity of the solvent. As we know, the timescale of
Chapter 3
95
Scheme 3.1. Representation of the excited state relaxation mechanism of NPP. Following
charge transfer, In non-polar solvents intersystem crossing is the main non-radiative pathway,
while as in polar solvents, dominant relaxation channel is internal conversion.
intersystem crossing is slower than that of internal conversion, which can be
manifested from the femtosecond results with larger second time component in
non-polar solvents and a faster decay in polar solvents. This time component has
also shown its dependence on viscosity of the solvent with 2 increasing from 1.67
ps in ethanol to 2.72 ps in n-octanol. In general, increase in viscosity hinders the
torsional motion of –NO2 group down the barrierless S1 potential surface and delay
the occurrence of intersystem crossing (ISC). Thus solvent’s role in determining
the fluorescence decay behaviour may be ascribed to the effect of polarity on the
relative energies of singlet and the triplet manifolds, which changes the degree of
spin orbit coupling, as well as the viscosity of the solvents.
Chapter 3
96
3.4. Conclusion
In this work, we have directly shown the involvement of the –NO2 group
rotation as the main coordinate in the excited state relaxation dynamics of an NLO
dye
(S)-(−)-1-(4-Nitrophenyl)-2-pyrrolidinemethanol.
The
steady
state
measurements report an appreciable change in dipole moment of 7.55 D in the
molecule. The quantum yield calculations comprehend the non-fluorescent nature
of NPP with NPP ~ 51 x 10-4 in acetonitrile. Quantum chemical calculations in
vacuum predict the presence of few triplet states in between the ground and first
singlet excited state with the occurrence of conical intersection between S 1 and
upper triplet maifold. Femtosecond fluorescence transients of NPP in different
solvents shows biexponential decay behavior. The faster time component of few
hundred femtoseconds was assigned to twisted intramolecular charge transfer
(TICT) from pyrrolidine moiety to –NO2 and the second time constant of few
picoseconds was attributed to either to intersystem crossing (ISC) from S 1 state to
some upper triplet manifold.or to the internal conversion between the two singlet
states. The excited state charge transfer dynamics was authenticated by the
construction of intensity and area normalized time resolved emission spectra
(TRES and TRANES), which inferred a decay of LE state and fast rise of CT state
within few hundred femtoseconds. Solvent polarity is found to have a dominant
effect in deciding the fate of the TICT state. In non-polar solvents, due to the close
intimacy of the oppositely symmetric S1 and T2 states, intersystem crossing
dominates over internal conversion, while as in polar solvents, energetic separation
of the above states disables intersystem crossing and instead facilitates internal
conversion between S1 and S0. The effect of using long chain alcohols elucidates
the consideration of choosing torsional motion of –NO2 group as the main
relaxation coordinate.
Chapter 3
97
References
1. Rizzo, F.; Cavazzini, M.; Righetto, S.; Angelis, F. D.; Fantacci, S.; Quici, S.
Eur. J. Org. Chem. 2010, 4004.
2. Zyss, J.; Kelley, P.; Liao, P. F. Molecular nonlinear optics: Materials,
Physics and Devices, Academic Press, 1993.
3. Schneider, A.; Guenter, P. Ferroelectrics 2005, 318, 83.
4. Marder, S. R.; Kippelen, B.; Jen, A. K. -Y.; Peyghambarian, N. Nature
1997, 388, 845.
5. Facchetti, A.; Beverina, L.; Van der Boom, M. E.; Dutta, P.; Evmenenko,
G.; Shukla, A. D.; Stern, C. E.; Pagani, G. A.; Marks, T. J. J. Am. Chem.
Soc. 2006, 128, 2142.
6. Davies, J. A.; Elangovan, A.; Sullivan, P. A.; Olbricht, B. C.; Bale, D. H.;
Ewy. T. R.; Isborn, C. M.; Eichinger, B. E.; Robinson, B. H.; Reid, P.J.; Li,
X.; Dalton, L. R. J. Am. Chem. Soc. 2008, 130, 10565.
7. Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Chem. Rev. 2003, 103, 3899.
8. Glasbeek, G.; Zhamg, H. Chem. Rev. 2004, 104, 1929.
9. Mondal, J. A.; Sarkar, M.; Samanta, A.; Ghosh, H. N.; Palit, D. K. J. Phys.
Chem. A 2007, 111, 6122.
10. Huppert, D.; Rand, S. D.; Rentzepis, P. M.; Barbara, P. F.; Struve, W. S.;
Grabowski, Z. R. J. Chem. Phys. 1981, 75, 5714.
11. Lippert, E.; Rettig, W.; Bonancic-Koutecky, V.; Heisel, F.; Mieche, J. A.
Adv. Chem. Phys. 1987, 68, 1.
12. Hamanoue, K.; Hirayama, S.; Nakayama, T.; Teranishi, H. J. Phys. Chem.
1980, 84, 2074.
13. Hamanoue, K.; Hirayama, S.; Nakayama, T.; Teranishi, H. Chem. Lett. 1980,
407.
14. Anderson, R. W., Jr.; Hochstrasser, R. M.; Lutz, H.; Scott, G. W. Chem. Lett.
1980, 407.
15. Heinz, B.; Schmierer, T.; Laimgruber, S.; Gilch, P. J. Photochem. Photobiol.
A 2008, 199, 274.
Chapter 3
98
16. Mohammad, O. F.; Vauthey, E. J. Phys. Chem. A 2008, 112, 3823.
17. Collado-Fregoso, E.; Zugazagoitia, J. S.; Plaza-Medina, E. F.; Peon, J. J.
Phys. Chem. A 2009, 113, 13498.
18. Zugazagoitia, J. S.; Collado-Fregoso, E.; Plaza-Medina, E. F.; Peon, J. J.
Phys. Chem. A 2009, 113, 805.
19. Zugazagoitia, J. S.; Almora-Diaz, C. X.; Peon, J. J. Phys. Chem. A 2008,
112, 358.
20. Hamanoue, K.; Nakayama, T.; Kajiwara, K.; Yamanaka, S.; Ushida, K. J.
Chem. Soc., Faraday Trans. 1992, 88, 3145.
21. Crespo-Hernández, C. E.; Burdzinski, G.; Arce, R. J. J. Phys. Chem. A 2008,
112, 6313.
22. Thomsen, C. L.; Thogersen, J.; Keiding, S. R. J. Phys. Chem. A 1998, 102,
1062.
23. Gurzadyan, G.; Goerner, H. Chem. Phys. Lett. 2000, 319, 164.
24. Farztdinov, V. M.; Schanz, R.; Kovalenko, S. A.; Ernsting, N. P. J. Phys.
Chem. 2000, 104, 11486.
25. Kovalenko, S. A.; Schanz , R.; Farztsinov, V. M.; Hening, H.; Ernsting, N. P.
Chem. Phys. Lett. 2000, 323, 312.
26. Morales-Cueto, R.; Esquivelzeta-Rabell, M.; Saucedo-Zugazagoitia, J.; Peon,
J. J. Phys. Chem. A 2007, 111, 552.
27. (a) Bilot, L.; Kawski, A. Z. Naturforsch, 1962, 17a, 621. (b) Bilot, L.;
Kawski, A. Z Naturforsch. 1963, 18a, 256. (c) Kawski, A. Acta. Phys. Polon.
1966, 29, 507. (d) Kawski, A. Z. Naturforsch. 2002, 57a, 255.
28. (a) Kawski, A.; Kuklinski, B.; Bojarski, P. Z. Natureforsch, 2001, 56a, 407.
(b) Kawski, A.; Kuklinski, B.; Bojarski, P. Chem. Phys. Lett. 2007, 448, 208.
(c) Kawski, A.; Bojarski, P.; Kuklinski, B. Chem. Phys. Lett. 2008, 463, 410.
(d) Inamdar, S. R.; Nadaf, Y. F.; Mulimani, B. G. J. Mol. Struct. 2003, 624,
47. (e) Rajbongshi, B. K.; Sen, P.; Ramanathan, G. Chem. Phys. Lett. 2010,
494, 295.
Chapter 3
99
29. (a) Lippert, V. E. Z. Electrochem. 1957, 61, 962. (b) Mataga, N.; Kaifu, Y.;
Koizumi, M. Bull. Chem. Soc, Jpn. 1956, 29, 465.
30. (a) Sen, P.; Roy, D.; Mondal, S. K.; Sahu, K.; Ghosh, S.; Bhattacharyya, K.
J. Phys. Chem. A 2005, 109, 9716. (b) Sen, P.; Mukherjee, S.; Patra, A.;
Bhattacharyya, K. J. Phys. Chem. B 2005, 109, 3319.
31. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R.
E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, K. N.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.;
Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson,
G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A.
G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 03, revision D.01; Gaussian, Inc.:
Wallingford, CT, 2004.
32. Reichardt, C.; Vogt, R. A.; Crespo-Hernandez, C. E. J. Chem. Phys. 2009,
131, 224518.
33. Scalmani, G.; Frisch, M. J.; Mennucci, B.; Tomasi. J.; Cammi, R.; Vincenzo,
B. J. Chem. Phys. 2006, 124, 094107.
34. Kasajima, T.; Akimoto, S.; Sato, S.; Yamazaki, I. J. Phys. Chem. A 2004,
108, 3268.
35. Peon, J.; Zewail, A. H. Chem. Phys. Lett. 2001, 248, 255.
36. Horng, M. L.; Gardecki, J. A.; Papazyan, A.; Maroncelli, M. J. Phys. Chem.
2005, 99, 17311.
37. Kovalenko, S. A.; Ruthmann, J.; Ernsting, N. P. J. Chem. Phys. 1998, 109,
1894.
Chapter 3
100
38. (a) Maroncelli, M.; Fleming, G. R. J. Chem. Phys. 1987, 86, 6221. (b) Koti,
A. S. R.; Krishna, M. M. G.; Periasamy, N. J. Phys. Chem. A 2001, 105,
1767.
39. Takezaki, M.; Hirota, N.; Terazima, M.; Sato, H.; Nakajima, T.; Kato, S. J.
Phys. Chem. A 1997, 101, 5190.
40. (a) Kaduk, D.; Voorhis, T. V. J. Chem. Phys. 2010, 133, 061102. (b) Levine,
B. G.; Ko. C.; Quenneville, J.; Martinez, T. J. Mol. Phys. 2006, 104, 1039.
41. (a) Stein, T.; Kronik, L.; Baer, R. J. Am. Chem. Soc. 2009, 131, 2818. (b)
Dreuw, A.; Weisman, J. L.; Head-Gordon, M. J. Chem. Phys. 2003, 119,
2943. (c) Liao, M. S.; Lu, Y.; Parker, V. D.; Scheiner, S. J. Phys. Chem. A
2003, 107, 8939.
42. El-Sayed, M. A. Acc. Chem. Res. 1968, 1, 8.
43. Viswanath, D. S.; Ghosh,T. K. Viscosity of Liquids; Theory, Estimation,
Experiment, and Data, Springer, 2007.
44. Rafiq, S.; Yadav, R.; Sen, P. J. Phys. Chem. B 2010, 114, 13988.
45. Duxbury, D. F. Chem. Rev. 1993, 93, 381.
46. Yoshizawa, M.; Suzuki, K.; Kubo, A.; Saikan, S. Chem. Phys. Lett. 1998,
290, 43.
Chapter 4
Dielectric Mediated Relaxation Dynamics
of trans-4-Dimethylamino-4-Nitrostilbene
Shahnawaz Rafiq et al., J. Chem. Phys. 2013, 138, 84308.
102
Chapter 4
Femtosecond fluorescence up-conversion technique was employed to
reinvestigate the intriguing dependence of fluorescence quantum yield of trans-4dimethylamino-4-nitrostilbene (DNS) on dielectric properties of the media. In
polar solvents like methanol and acetonitrile, the two time components of the
fluorescence transients were assigned to intramolecular charge transfer (ICT)
dynamics and to the depletion of the ICT state to the ground state via internal
conversion along the torsional coordinate of nitro moiety. The viscosity
independence of the first time component indicates the absence of any torsional
coordinate in the charge transfer process. In polarizable solvent (carbon
tetrachloride) the fluorescence transients show a triple exponential behaviour. The
first time component was assigned to the formation of the ICT state on a 2
picosecond time scale. Second time component was assigned to the relaxation of
the ICT state via two torsion controlled channels. First channel involves the
torsional motion about the central double bond leading to the trans-cis
isomerization via a conical intersection or avoided crossing. The other channel
contributing to the depopulation of ICT state involves the torsional coordinates of
dimethylanilino and/or nitrophenyl moieties and leads to the formation of a
conformationally relaxed state, which subsequently relaxes back to the ground
state radiatively, and is responsible for the high fluorescence quantum yield of
DNS in polarizable solvents like carbon tetrachloride, toluene, etc. The excited
singlet state which is having a dominant π-π* character may also decay via
intersystem crossing to the n-π* triplet manifold and thus accounts for the
observed triplet yield of the molecule in polarizable solvents.
Chapter 4
103
4.1. Introduction
Photo-induced excited state intramolecular charge transfer (ICT) dynamics in
-conjugated donor and acceptor systems have been studied extensively for
decades.1-14 Such dynamics define the competence of certain molecular systems in
photosynthesis and solar energy conversion.1-3 Due to the existing and conceived
applications of such systems in the field of photonics and optoelectronics,
knowledge of electronic structure is cardinal to the understanding of their
properties, and also the ability to tune the electronic structure holds the promise of
custom design. Excited state charge transfer dynamics unravelled at an ultrafast
timescale has been very important to understand the mechanistic aspects of the
molecular events.4-6 4-(N,N-dimethylamino)benzonitrile (DMABN)7-10 is one of
the extensively studied prototype of such systems. According to a widely accepted
model by Grabowski and co-workers, in polar solvents DMABN undergoes a
twisted intramolecular charge transfer from the donor dimethylamine group to the
acceptor cyano moiety on photo excitation.7a,7b The twisting of the dimethylamine
group about the plane of the molecule induces decoupling of the donor and
acceptor groups resulting in quenching of fluorescence. 7-10 Similar to DMABN,
Trans-4-(N,N-dimethylamino)-4'-cyanostilbene (DCS) has also been studied
thoroughly by time-resolved fluorescence, transient absorption, kerr ellipsometry
spectroscopy, etc.11-13 On vertical excitation, the Franck-Condon (FC) state
undergoes electronic reorganization to an ICT state with high polar character
having nearly planar conformation.11 Subsequently, the ICT state depopulates itself
to a phantom state, P* (less polar and non-fluorescent) along the twist coordinate
of double bond, which leads to the trans-cis isomerization. However in polar
solvents, the energy barrier between the ICT and P* state increases resulting in
reduction of isomerization. Instead, the ICT state decays to a highly polar and
fluorescent conformationaly relaxed state (CRICT) along the dimethylanilino
group twist coordinate.11-13
Another candidate belonging to the class of push-pull stilbenes is trans-4-(N,Ndimethylamino)-4-nitrostilbene, (DNS) which has been studied by many groups
104
Chapter 4
on account of its interesting and surprising fluorescence properties. 14-26 The
molecule has been exploited as a potential nonlinear material and also as a color
tuner in OLEDs.14,15 Compared to DCS, DNS shows a peculiar fluorescence
behavior as a function of solvent polarity. The fluorescence quantum yield
increases from weakly polar to polarizable solvents and then almost goes to zero in
highly polar solvents with a very large red shift in the emission maximum. 18 The
role of solvent polarity on the deactivation mechanism of excited state of DNS was
pioneered by Lippert,16a Schulte-Frohilinde,16b Fischer,16c and co-workers. Since
then a lot of work was carried out both experimentally and theoretically on this
molecule especially from the mechanistic perspectives.17-26 In non-polar solvents,
the reason of being non-fluorescent has been ascribed to trans-cis isomerization via
triplet mechanism involving two triplet manifolds.17-19 It is proposed that, the
excited trans-singlet state (1t*) undergoes intersystem crossing to the excited transtriplet state (3t*) and is followed by twisting of the C═C double bond, which
populates the perpendicular triplet configuration ( 3P*). Second intersystem
crossing from the 3P* state to the perpendicular configuration in the ground state
(perp) leads to trans-cis isomerization, as shown below.17-19
1
t*3 t*3P*  perp  cis  trans
In polarizable solvents, DNS is observed to have a high fluorescence quantum
yield in the order of ca. 0.5.18b,c This has been illustrated on behalf of the
depopulation of 1t* to a transoid excited state configuration A* (different than
solvent relaxed 1t* state). The major depletion has been mentioned to be a doubly
activated pathway given by18b
1
t*  A*1P*1P  cis  trans
The intersystem crossing and internal conversion may also contribute, but to a
lesser extent. Insignificant intersystem crossing in polarizable solvents enunciates
the occurrence of an energy barrier between A* and perpendicular singlet
configuration (1P*), which will eventually reduce the yield of isomerization. 18b In
polar solvents, the trans-cis isomerization is highly improbable, because of the
lowering of energy of highly polar A* state due to solvation. The A* state in polar
Chapter 4
105
solvents has a maximum contribution from the torsional motion of the nitro group,
and the twisted state being highly non-radiative, resulting in almost zero quantum
yield of DNS.18,19
Semiempirical calculations (SAM1) were employed by Farztdinov et al.23 to
establish the polarity dependence of fluorescence quantum yield, intersystem
crossing and trans-cis isomerisation of DNS. According to them, in non-polar
solvents, the FC state undergoes ISC via a planar S1 conformation. The second ISC
step leads to trans-cis isomerization via the formation of perpendicular triplet
conformation. In polarizable solvents, the FC state decays to planar conformation
which is followed by depletion to the singlet perpendicular geometry. In highly
polar solvents, the relaxation is predominantly controlled by the torsional motion
of nitrophenyl group, which renders the twisted state to be nonradiative.23
Theoretical calculations on this molecule were also performed by Vijayakumar, 24
Petsalakis25 and Murugan26 to explore various other aspects of the molecule.
Although a significant attention has been paid to understand the photophysics
of such an interesting push-pull system, the most important process in this
molecule i.e., the intramolecular charge transfer has not been studied and
established explicitly. Also the possible route of trans-cis isomerization is highly
ambiguous. The contribution of the different groups within the molecule towards
the relaxation dynamics of the excited state also deserves a keen attention. The
present work showcases in detail, the ultrafast events occurring in the excited state
immediately after photo-excitation by employing femtosecond fluorescence upconversion measurements as a function of solvent parameters.
4.2. Results
4.2.1. Steady State Fluorescence Measurements
To have a vivid cognition about the effect of viscosity on the fluorescence
quantum yield of DNS (Scheme 4.1), we choose a mixture of glycerol and
methanol with very different viscosities (glycerol = 945 cP, methanol = 0.54 cP),
Chapter 4
106
but almost similar dielectric constant
(glycerol:
ETN  0.762 ).
ETN  0.812 ,
methanol:
The mixing of these two
solvents will vary the viscosity of the Scheme 4.1. Molecular structure of trans-4medium markedly without affecting
dimethylamino-4-nitrostilbene (DNS).
the polarity to a significant extent. Also, the fluorescence intensity of DNS is
insensitive to the change in polarity of the medium in the higher polarity regime. 18
Under these criteria, we can evidently visualize the effect of viscosity on the
fluorescence properties of the molecule. The absorption maximum of DNS in pure
methanol is 426.0 nm, and as the proportion of glycerol increases to 70% (v/v), the
absorption maximum monotonously red shifts to 438.6 nm. The extinction
coefficient of DNS in these solvent mixtures is found to be similar (Figure 4.1(a)).
Emission spectrum with maximum of ca. 690 nm in pure methanol was obtained
by exciting the sample using 410 nm light which exhibits a continuous steady red
shift on increasing the proportion of glycerol in methanol and is found to be ca.
710 nm in 70% (v/v) glycerol-methanol mixture (Figure 4.1(b)). Fluorescence
intensity shows a regular increase as the proportion of glycerol is raised from 10%
to 70% v/v in methanol (Figure 4.1(c)). The fluorescence intensity at the emission
maxima almost doubles as the coefficient of viscosity raises from 1.15 cP (10%
v/v) to 100.75 cP (70% v/v) and follows a power law dependence giving an 
value of 0.63. The dependence of fluorescence intensity on the viscosity of the
solvent rationalizes the involvement of torsional motion of molecular fragments in
the excited state.
As observed before,18 the dielectric property of the solvent executes a very
significant and surprising effect on the fluorescent properties of DNS. In polar
solvents like acetonitrile ( ETN  0.460) and methanol ( ETN  0.762) the absorption
maxima are 430 nm and 426 nm respectively and the emission maxima are ~830
nm ~690 nm respectively (Figure 4.2). A very large Stokes shift and a small
quantum yield (f < 0.002)18b characterizes the excited state of DNS in polar
Chapter 4
107
Figure 4.1. (a) Absorption spectra of DNS in different glycerol-methanol mixtures on an
absolute extinction coefficient scale. (b) Steady state fluorescence spectra of DNS in different
glycerol-methanol binary solvent mixtures obtained by using an excitation wavelength of 410
nm. (c) Plot of fluorescence intensity against the viscosity of the medium.
Figure 4.2. (a) Steady state absorption spectra of DNS in acetonitrile-CCl4 binary solvent
mixture are shown in absolute extinction coefficient scale. (b) Steady state fluorescence
spectra of DNS in different acetonitrile-CCl4 binary solvent mixtures obtained by using an
excitation source of 410 nm.
108
Chapter 4
solvents. While as, in completely non-polar solvents, like cyclohexane ( ETN  0.006) ,
absorption maximum is at 417 nm and emission maximum is at 470 nm (Stokes
shift = 2700 cm-1) and the fluorescence quantum yield is 0.33.18b In polarizable
solvents like carbon tetrachloride, toluene ( ETN  0.099) , etc DNS is characterized
by moderate Stokes shift of 5392 cm-1 and 6211 cm-1 associated with a high
quantum yield of 0.40 and 0.53 respectively.18b,20a
4.2.2. Viscosity Dependent Femtosecond Fluorescence Transients
The fluorescence transients of DNS were recorded at 605 nm in seven different
v/v proportions of glycerol in methanol binary solvent mixture by exciting at 410
nm, as shown in figure 4.3(a). The decays are very fast and are best fitted by a biexponential function (Table 4.1). In 10% (v/v) glycerol - methanol mixture ( =
1.15 cP), the two time constants are found to be 480 fs and 2.15 ps. As the
viscosity of the medium increases to 100.75 cP (for 70% v/v glycerol - methanol
mixture) the two time constants changed to 410 fs and 3.15 ps respectively. The
first time component does not show any dependence on the viscosity of the solvent
as observed in the experiments (Table 4.1). While as, the second time component
exhibits a monotonous increase, as the viscosity of medium increases and is found
Figure 4.3. (a) Fluorescence transients of DNS recorded at 605 nm in various proportions
(v/v) of glycerol methanol mixtures. (b) A few representative fluorescence transients of DNS
in pure methanol monitored at different wavelengths. DNS has been excited by a 410 nm
femtosecond pulse.
Chapter 4
109
Table 4.1. Bi-exponential fitting parameters of fluorescence transients of DNS in different
proportions (% v/v) of glycerol-methanol mixture.
Volume % of
Glycerol in
Viscosity (cP)
a1
a2
1a (ps)
2a (ps)
Methanol
10
1.15
0.76
0.48
0.24
2.15
20
2.42
0.73
0.54
0.27
2.20
30
5.09
0.63
0.48
0.37
2.30
40
10.74
0.74
0.50
0.26
2.45
50
22.65
0.65
0.46
0.35
2.50
60
47.77
0.67
0.45
0.33
2.70
70
100.75
0.67
0.41
0.33
3.15
100 fs
a
to follow a power law dependence ( = 0 + c), which yields an  value of 0.53,
0 is 2.05 ps, and c is 0.09 ps cP-. The observed value of the exponent () is very
similar to that obtained by fitting the steady state fluorescence intensity vs.
viscosity (Figure 4.1(c)) which is 0.63. This value of  indicates the involvement
of diffusive like motion in the excited state. This dependence of the longer time
component on viscosity infers the involvement of torsional coordinate in the
excited state of DNS.
4.2.3. Fluorescence Transients in Highly Polar Solvents
Although we have already recorded the fluorescence transients of DNS in
highly polar glycerol-methanol mixtures, to have a vivid understanding of the
excited state behaviour of DNS, decay transients were also recorded in pure
methanol over a range of wavelengths. It must be borne in mind, that the steady
state fluorescence spectrum of DNS in methanol shows a large Stokes shift with
emission maximum at ca. 690 nm. It was technically difficult with the present
fluorescence up-conversion setup to measure the decays in the full range of the
emission spectrum and thus the transients were measured at wavelengths stretching
only on the blue side of the emission maximum. Understandably, the complete
picture of the excited state behaviour of DNS in methanol will not be possible to
portray, but still one can have some information which may help in establishing
110
Chapter 4
Table 4.2. Fitting parameters of fluorescence transients of DNS in pure methanol. The
transients were recorded at various wavelengths on exciting at 410 nm.
a1
a2
em (nm)
1a (ps)
2a (ps)
500
0.94
0.05
0.06
0.40
520
0.92
0.15
0.08
0.69
530
0.96
0.21
0.04
1.20
545
0.91
0.23
0.09
1.04
560
0.88
0.27
0.12
1.31
575
0.85
0.35
0.15
1.66
590
0.82
0.42
0.18
2.00
605
0.73
0.46
0.27
2.05
620
0.70
0.65
0.30
2.78
630
0.59
0.61
0.41
2.70
100 fs
a
pathway for excited state relaxation dynamics. The decay transients were obtained
at 10 wavelengths ranging from 500 nm to 630 nm. The molecule decays very fast
over a total time stretch of 5 ps for all the wavelengths mentioned above. All the
decays are bi-exponential in nature with the first time component of few hundred
femtoseconds and the second time component of few picoseconds, which are
shown in figure 4.3(b) and the fitting parameters are shown in table 4.2. Both the
time constants show a slight but regular increase, as the decays are monitored
towards the longer wavelengths. At 500 nm, the two time constants are 50 fs (0.94)
and 0.40 ps (0.06) which increase to 610 fs (0.59) and 2.70 ps (0.41) when
monitored at 630 nm.
4.2.4. Fluorescence Transients in Polarizable Solvent
DNS shows comparatively large fluorescence quantum yield in polarisable
solvents (more polar than the hydrocarbons) like chloroform, benzene, toluene,
carbon tetrachloride etc.18b,20a In carbon tetrachloride the emission maximum of
DNS is found to be at 540 nm. The decay transients of DNS in carbon tetrachloride
have been recorded at 14 different wavelengths covering the complete emission
spectra ranging from 470 nm to 615 nm and a few representative fluorescence
transients are shown in figure 4.4(a). The decays were best fitted by a three
exponential function and the fitting parameters are tabulated in table 4.3. In all the
Chapter 4
111
Figure 4.4. (a) Fitted data of a few representative fluorescence transients of DNS in carbon
tetrachloride by using an excitation wavelength of 410 nm. (b) Fluorescence transients of
DNS monitored in various proportions (% v/v) of carbon tetrachloride in acetonitrile
monitored at 680 nm.
Table 4.3. Fitting parameters of wavelength dependent fluorescence transients of DNS in
carbon tetrachloride. The third time component was fixed to 3 ns, however its amplitude was
not fixed.
a1
a2
a3
em (nm)
1a (ps)
2a (ps)
3 (ps)
470
0.63
1.29
0.14
18.7
0.23
3000
480
0.62
1.44
0.15
22.5
0.23
3000
490
0.55
1.53
0.15
29.8
0.30
3000
500
0.45
1.93
0.18
53.2
0.37
3000
510
0.39
1.96
0.20
61.0
0.41
3000
520
0.30
2.50
0.25
83.9
0.45
3000
530
0.19
3.20
0.30
80.0
0.51
3000
540
0.11
3.95
0.32
68.0
0.57
3000
550
-1.42
0.12
0.87
65.0
1.55
3000
560
-0.13
0.36
0.41
97.9
0.72
3000
570
-0.32
0.74
0.55
146.0
0.77
3000
585
-0.43
0.82
0.53
140.0
0.90
3000
600
-0.37
2.13
0.61
157.0
0.76
3000
615
-0.35
2.38
0.57
159.0
0.78
3000
100 fs
a
transients, the third time constant was fixed to 3 ns, based on the previously
reported value of lifetime of DNS in benzene20b (DNS shows similar behaviour in
Carbon tetrachloride and benzene). At 470 nm, the other two time constants are
found to be 1.29 ps and 18.7 ps. As the transients are measured towards the
wavelength representing the maximum in the emission spectra (540 nm), the first
time constant increases monotonously to 3.95 ps while as the second time
112
Chapter 4
component shows an irregular behaviour with an overall increases to 68 ps. At 550
nm and onwards, the first time component of the fluorescence transient shows a
negative amplitude inferring the formation of a new state, whose rise time keeps on
increasing as the decays are measured towards the red side of the emission spectra.
The second time component in this wavelength regime keeps on increasing. At 615
nm, the three time constants are 2.38 ps (-0.35), 159.0 ps (0.57), and 3.0 ns (0.78).
4.2.5. Polarity Dependent Behaviour of Fluorescence Transients in
Acetonitrile-Carbon tetrachloride Mixture
This mixture provides a perfect environment to continuously monitor the nature
of the excited state as one descends from highly polar to polarizable medium. In
pure acetonitrile, the emission maximum is found to be at ~830 nm and as the
proportion of carbon tetrachloride is raised, the emission spectra shows a
monotonous blue shift with its maximum at 540 nm in pure carbon tetrachloride.
The time dependence of the fluorescence of DNS at 680 nm in pure acetonitrile
decays very fast and was best fitted by a sum of two exponentials with time
constants of 250 fs (0.94) and 4.4 ps (0.06). For 10% (v/v) carbon tetrachloride in
acetonitrile, the decay becomes slower and the fitting parameters are 320 fs (0.94)
and 16.3 ps (0.06). Similar bi-exponential decay patterns were observed up to 30%
carbon tetrachloride in acetonitrile with the time constants raising up to 640 fs
(0.85) and 26.0 ps (0.15) respectively. All the fitting parameters along with the
monitoring wavelengths are mentioned in table 4.4 and a few representative decay
transients are shown in figure 4.4(b). Surprisingly, when the proportion of carbon
tetrachloride in acetonitrile is raised to 40 % (v/v) and above, another time
component appears in the fluorescence transients. At 40% (v/v) the three time
constants along with contributions are 50 fs (-1.74), 1.15 ps (0.64) and 64.9 ps
(0.10) and as the proportion of carbon tetrachloride in acetonitrile is raised beyond
40% (v/v), the growth component becomes much more evident with the time
constants of 660 fs (-0.29), 17.52 ps (0.56) and 175.5 ps (0.73) for 80% (v/v)
carbon tetrachloride in acetonitrile.
Chapter 4
113
Table 4.4. Fitting parameters of fluorescence transients of DNS recorded in a binary mixture
of carbon tetrachloride and acetonitrile. All the transients were recorded at 680 nm upon
exciting at 410 nm.
% (v/v) of CCl4
a1
a2
a3
1a (fs)
2a (ps)
3a (ps)
in aetonitrile
0
0.94
250
0.06
4.4
10
0.94
320
0.06
16.3
20
0.87
500
0.13
18.5
30
0.85
640
0.15
26.0
40
-1.74
50
0.64
1.15
0.1
64.9
50
-2.68
50
1.26
1.72
0.42
40.4
60
-4.57
100
3.52
3.17
2.05
66.3
70
-0.42
490
0.88
3.64
0.54
63.8
80
-0.29
660
0.56
17.52
0.73
175.5
100 fs
a
4.2.6. Quantum Chemical Calculations
The optimized ground state structure of DNS is characterized by planar
conformation of the nitrophenyl and N,N-dimethylanilino moieties about the
central double bond. The solvent dependent ground and excited state potential
energy surfaces of the singlet and triplet manifolds were constructed along the
torsional coordinates of N,N-dimethyaniline, nitrophenyl and double bond torsional
coordinates. The Franck-Condon state is having the same planar geometry as that
of the optimized ground state. The first excited singlet and triplet states are having
a dominant -* character, while as the second excited triplet state is placed at
energy higher than first singlet excited state and is having a dominant n-*
character. The energy of the second triplet manifold is comparatively higher in
polar solvents than in less polar solvents. Because of the inherent artifacts in the
TDDFT to entertain long range charge transfer, the calculations have been used
mainly for the qualitative understanding.27
4.3. Discussion
The viscosity dependent steady state and time resolved measurements of DNS
provide a stringent confirmation of the involvement of torsional motion within the
molecule during the depletion of the excited state. The viscosity dependence of the
114
Chapter 4
fluorescence intensity in different proportions of glycerol-methanol mixtures (as
shown in figure 4.1) reveals small change in intensity during the initial increase of
the proportion of glycerol in methanol and becomes significant once the proportion
of glycerol in methanol approaches to 30% (v/v). Although there is an increase in
fluorescence intensity with increase in the viscosity of the medium, even at 70%
(v/v) glycerol-in-methanol the value of fluorescence quantum yield is found to be
very small. The non-fluorescent nature of DNS in more polar, even highly viscous,
solvents is also evident from the very fast fluorescence decay of the molecule on a
few picosecond timescale. The decay transients (as shown in figure 4.3 and
mentioned in section 4.2.2) suggest a major contribution from the fast few hundred
femtosecond time component, which practically shows no dependence on the
viscosity of the medium. While as, the second time component shows a perceptible
dependence on the viscosity of the medium, which infers the involvement of
torsional motion in the excited state relaxation pathway.
As mentioned in section 4.2.3, the fluorescence transients of DNS in methanol
show emission wavelength dependence and are biexponential in nature. The
existence of two time components inherently designates the decay of the FC state
into a transitory state in a few hundred femtoseconds timescale, which then
subsequently decays further in a few picosecond timescale. Both the time constants
keep on increasing as the transients are monitored at wavelengths towards the red
side of the emission spectrum starting from 500 nm. The contribution of first time
component decreases and that of second time component increases with increase in
monitoring wavelength. The decrease of the contribution of the first time
component suggests the depletion of FC state into another transient state, which
may show growth, if the decays are recorded on the red side of the fluorescence
maximum. However as mentioned earlier, due to the technical problems in
obtaining the data at higher wavelength region, we cannot ascribe much to the
nature of the first time component observed in the excited state dynamics of DNS
in strongly polar solvents. Instead we choose carbon tetrachloride wherein we can
monitor the decay transients at wavelengths stretching over the full emission range.
Chapter 4
115
In carbon tetrachloride, the decay transients exhibit a very significant
dependence on the emission wavelength as can be seen in figure 4.4(a), with the
time constants tabulated in table 4.3, giving a quantifiable extent of dependence.
To have a more vivid understanding of the events occurring in the earlier time
scales, time resolved emission spectra were constructed.28 The constructed time
resolved intensity normalized emission spectra (TRES) is shown in figure 4.5(a).
At zero time the emission spectra shows a maximum at 515 nm and is gradually
red shifted as time increases. At 20 ps, the emission maximum is found to be at
538 nm with an overall Stokes shift of 800 cm-1 and after that no significant shift is
Figure 4.5. (a) Time resolved emission spectra of DNS in CCl4 constructed from the fitting
parameters of the fluorescence transients recorded at different wavelengths of the emission
spectrum. (b) Time resolved area normalized emission spectra characterized by an
isoemissive point at 525 nm. (c) Plot showing the variation of wavelength corresponding to
emission maximum in the TRES as a function of time. The data is best fitted by a biexponential function.
116
Chapter 4
observed. To know the actual nature of the excited state process, the time resolved
area normalized emission spectrum (TRANES)29 has been constructed as shown in
figure 4.5(b). TRANES clearly indicates a two state process with a well defined
iso-emissive point. The initial emission with a maximum at 515 nm keeps on
decreasing and a new band at 538 nm is formed as time progresses from 0 to 20 ps.
After 20 ps, no further increase in the intensity of the 538 nm band or decrease of
515 nm band is observed. The time dependent change of the emission maximum is
shown in figure 4.5(c). The emission maxima shows convergence at 80 ps and the
data could be best fitted with a bi-exponential function. The observed two time
constants are 2.2 ps (0.88) and 24.8 ps (0.12). This suggests that the total Stokes
shift of 800 cm-1 has a major contribution of 2.2 ps time constant. Based on the
experimentally observed growth in the fluorescence transients and on the existence
of a well defined iso-emissive point in the TRANES, the first time component is
ascribed to the formation of an intramolecular charge transfer (ICT) state. Briefly,
on photo-excitation, the initial FC state undergoes a rapid charge re-distribution,
within donor dimethylamine group and the acceptor nitro group, to form a charge
transfer state resulting in the observed rise time (growth) in the fluorescence
transient at the red edge of the emission spectrum. Assignment of this time
component to the intramolecular charge transfer dynamics can also be established
from the dependence of this time component on the polarity of the medium. In high
polar solvents like methanol and acetonitrile (see sections 4.2.3 and 4.2.5), the first
time constant is smaller than observed in polarizable solvents like carbon
tetrachloride. The depletion of FC state to ICT state in polar solvents will be very
fast compared to that in polarizable solvents on behalf of the extra stability gained
by the ICT state in polar solvents, which will consequently increase the forward
transformation rate of FC to ICT state and hence decreases the time constant.28 One
important point to mention here is that the first time component does not show any
dependence on solvent viscosity and hence clearly indicates the absence of the
involvement of any torsional motion in the charge transfer process. Here we have
experimentally established that the intramolecular charge transfer dynamics in
Chapter 4
117
DNS occurs without any involvement of torsional motion of any moiety within the
molecule.
The decay of the FC state to the ICT state thus explains the first time
component observed in the decay of DNS in both polar as well as polarizable
solvents. The second time component in polar solvents shows a monotonous
dependence on the viscosity of the medium. The small incremental dependence of
this time component on the viscosity of the medium (as mentioned in section
4.2.2.) points towards the involvement of torsional motion of groups like
dimethylamine or nitro moiety in the excited state. Based on picosecond timeresolved coherent anti-Stoke Raman spectroscopy, Oberlé and coworkers19a
postulated that the charge transfer dynamics involves rotation of nitro group. We
agree, in terms of that, there is a nitro rotation, but at the same time we concluded
that nitro rotation does not occur during the charge transfer process, as we have
already established the absence of involvement of any torsional coordinate during
this process. If there would have been any nitro rotation during this process, then
its time scale in polar solvent should have been much smaller (on a few hundred
femtosecond timescale) than what they have observed. Their observation of the
shift of nitro stretching frequency on the picosecond timescale in acetonitrile
solvent may occur during the subsequent relaxation of the charge transfer state.
This fact can as well be established by mentioning that they did not observe any
shift in nitro stretching frequency in toluene. Neverthless, if the charge transfer
process invloves the rotation of nitro group, they must have observed a shift in the
nitro stretching frequency of DNS in toluene.19a It has previously been reported
that in polar solvents, the fluorescent quantum yield and the rate of intersystem
crossing is negligible,18 so the only pathway for the nonradiative decay of the
charge transfer state is internal conversion through conical intersection or avoided
crossing. The viscosity dependent second time component in the observed
fluorescence transient of DNS is thus assigned to the depletion of the CT state to
the ground state via a possible conical intersection along the torsional coordinate of
the nitro group (Scheme 4.2). The occurrence of internal conversion instead of
118
Chapter 4
Scheme 4.2. Proposed excited state relaxation channels of DNS in highly polar solvents
(acetonitrile).
intersystem crossing in polar solvents is also enabled by the El-Sayed rules,30
through the existence of high energy n-* triplet state relative to -* singlet state
(as determined by TDDFT calculations and mentioned in section 4.2.6), which
excludes any chance of spin-orbit coupling between the different multiplicity states
in highly polar solvents.
The decay transients of DNS recorded in acetonitrile (polar) and carbon
tetrachloride (polarizable) mixture exhibit very intriguing nature of time
components as can be seen in figure 4.4(b) and table 4.4 (see section 4.2.5). In pure
acetonitrile, the excited state decays bi-exponentially with two time constants of
250 fs (0.94) and 4.4 ps (0.06). As the proportion of carbon tetrachloride is raised
to 30% v/v, the transient can still be best fitted by a bi-exponential function with a
regular increase in the value of time constants. In 30% v/v of carbon tetrachloride
in acetonitrile, the measured time constants are 640 fs (0.85) and 26.0 ps (0.15).
Once the proportion of carbon tetrachloride is raised to 40% v/v in acetonitrile, the
decay becomes triple exponential with time constants of 50 fs (-1.74), 1.15 ps
(0.64), and 64.9 ps (0.10). On further increase of carbon tetrachloride, the growth
component keeps on increasing and a similar increase is also observed for the other
Chapter 4
119
two time components. For 80% v/v carbon tetrachloride in acetonitrile, the three
components are 660 fs (-0.29), 17.5 ps (0.56), and 175 ps (0.73). All the transients
were recorded at wavelength 680 nm, which represents the blue side of the steady
state emission spectrum of DNS in acetonitrile and red side of the emission
spectrum in carbon tetrachloride. Initially at lower proportions of carbon
tetrachloride in acetonitrile, the two time components can be ascribed to the
intramolecular charge transfer event and the subsequent relaxation of ICT state to
the point of conical intersection or avoided crossing via the rotation of nitro group
as already established above. Till 30% v/v of carbon tetrachloride in acetonitrile,
the first time component increases from 250 fs to 640 fs beacause of the relatively
slower rate of charge transfer process in carbon tetrachloride - acetonitrile solvent
mixture than pure acetonitrile ascribed to the decreasing polarity of the medium.
An increase is also been observed for longer time component which is due to the
suppression of the nonradiative pathway, which may be originating from the
inclusion of carbon tetrachloride in the solvation sphere. This indicates that
whichever nonradiative channel is operational in acetonitrile, is either missing or
contributing ineffectively to the decay of excited state of DNS in carbon
tetrachloride. Considering the fluorescent nature of the ICT state in carbon
tetrachloride, we propose the involvement of a pathway different than the one
operational in polar solvents in the depletion of the ICT state. The dominant effect
of carbon tetrachloride on the excited state relaxation dynamics can be retrieved by
the observation of the rise instead of decay in the first time component for 40% and
greater v/v proportions of carbon tetrachloride in acetonitrile. The appearance of
the growth component may be due to the effect of carbon tetrachloride in the
solvation sphere. On this basis, we established that the first time component
accounts for the formation of the charge transfer state. The second and third time
component of the fluorescence transients of DNS in carbon tetrachloride acetonitrile mixture increases with increase in the proportion of carbon
tetrachloride. The second time constant increases from 1.15 ps (40% v/v) to 17.5
ps (80% v/v) and the third time constant increases from 4.4 ps (0% v/v) to 175.0 ps
120
Chapter 4
(80% v/v) (Table 4.4). In pure carbon tetrachloride, at the emission maximum,
these two time constants are 68 ps and 3 ns. Previous studies indicate that in
carbon tetrachloride, the fluorescence quantum yield is 0.4 and also the occurrence
of trans-cis isomerization was established and proposed to occur via a triplet
route.18
The large value of the fluorescence quantum yield and trans-cis isomerization
yield inferences towards the involvement of radiative and isomerization channels
during the excited state relaxation of DNS in carbon tetrachloride. Solvent
dependent quantum chemical calculations were carried out to construct the ground
and excited state potential energy surfaces along the torsional coordinates of
dimethylanilino and nitrophenyl moieties as shown in figure 4.6(a) and 4.6(b) (see
section 4.2.6). The nonintersecting nature of the excited singlet and ground state
in carbon tetrachloride along the dimethylanilino and nitrophenyl groups do not
lead to any conical intersection. This inherently predicts the existence of radiative
channel along the torsional coordinate of these two fragments and hence is
proposed to be responsible for the large fluorescence quantum yield of DNS in
carbon tetrachloride. Ground and excited state potential energy surfaces were also
constructed along the torsional coordinate of central double bond as shown in
figure 4.6(c). The singlet excited state surface intersects with the ground state
surface on its way towards the perpendicular configuration of the central double
bond forming an avoided conical intersection. This channel may be responsible for
the trans-cis isomerization of DNS in carbon tetrachloride. On its progression
towards the perpendicular configuration, the singlet excited state with -*
character may also undergo intersystem crossing with the nearby second triplet
manifold with n-* character on behalf of the symmetry breakdown by spin-orbit
coupling (Scheme 4.3). This triplet state decays to first triplet state, which then
relaxes to the ground state trans- isomer and is responsible for the previously
observed triplet yield of the molecule in completely non-polar and polarizable
media.18b It is important to mention that there is an activation barrier present in the
singlet excited state potential energy surface between the ICT state and the
Chapter 4
121
Figure 4.6. Potential energy surfaces of ground and excited states of DNS in carbon
tetrachloride solvent along the torsional coordinate of (a) dimethylanilino moiety, (b)
nitrophenyl moieties, and (c) central double bond. The designation of the states are; ground
state (Red, ), First triplet state (Black, ▲), First singlet state (Blue, ♦) and second triplet
state (Green, ▼).
Scheme 4.3. The relative stability of the different electronic states in polarizable and highly
polar solvents, predicting the fate of ISC.
122
Chapter 4
perpendicular conformation along the double bond rotation. As the polarity of the
medium increases, the ICT state will become more stable which will eventually
increase the height of the activation barrier and will decrease the yield of trans-cis
isomerization and almost zero in highly polar solvents. The observation of zero
triplet yield in the polar solvents is because of the lowering of the energy of -*
singlet state and increase in the energy of n-* triplet state (as evident from our
TDDFT calculations), which will exclude any chances of spin-orbit coupling
between the two states (Scheme 4.3) by increasing the energy separation, and
hence no triplet yield was observed in polar solvents.
On the basis of above discussion, the experimentally observed second time
component of DNS in carbon tetrachloride (24.8 ps time component in figure
4.5(c)) is ascribed to the lifetime of ICT state, which relaxes via two routes. One is
the non-radiative pathway via double bond torsional coordinate leading to trans-cis
isomerization. The other channel involves the torsional motion of N,Ndimethylaniline and nitrophenyl leading to the formation of a conformationally
relaxed radiative state, as evident from TDDFT calculations (Scheme 4.4). The
trans-cis isomerization does not occur via the triplet mechanism, rather via conical
Scheme 4.4. Proposed representation of the excited state relaxation pathways of DNS in
polarizable solvents (carbon tetrachloride).
Chapter 4
123
intersection between the singlet excited state and the ground state. The third time
component, largest among all, is ascribed to the radiative decay of the
conformationally relaxed singlet state and is responsible for the high fluorescence
quantum yield of DNS in polarizable solvents like carbon tetrachloride, toluene
etc. In conclusion, the observed three time components in carbon tetrachloride are
assigned sequentially to the charge transfer dynamics, depletion of the ICT state
and radiative decay from the conformationally relaxed singlet state.
4.4. Conclusion
The present study revealed important mechanistic details pertaining to the
relaxation behavior of trans-4-dimethylamino-4-nitrostilbene (DNS) in highly
polar and polarizable media. Upon photo-excitation, DNS undergoes an ultrafast
intramolecular charge transfer without any change in the molecular geometry and
the rate of charge transfer strongly depends on the dielectric properties of the
medium. In polar solvents, the ICT state decays along the torsional coordinate of
acceptor nitro moiety and depletes back to trans- ground state via internal
conversion through a conical intersection or avoided crossing. In polarizable
solvents like carbon tetrachloride, the ICT state relaxes via two channels. One
pathway accounts for the nonradiative decay of the ICT state and involves
torsional motion about the central double bond leading to the trans-cis
isomerization. This route is intervened by the presence of an activation barrier and
thus the isomerization yield strongly depends on the polarity of the medium. The
torsion about the double bond may also lead to the intersystem crossing between
the π-π* singlet state and the n-π* triplet manifold. The other pathway is the
depletion of ICT state along the torsional motion coordinate of dimethylanilino
and/or nitrophenyl moiety to a conformationally relaxed singlet excited state. This
newly formed state depletes radiatively to the ground state and hence is responsible
for the high fluorescence quantum yield of DNS in polarizable solvents like carbon
tetrachloride.
Chapter 4
124
References
1. (a) Blankenship, R, E.; Molecular Mechanisms of Photosynthesis, Blackwell
Science, Oxford. 2002. (b) Gehlen, J. N.; Marchi, M.; Chandler, D. Science
1994, 263, 499. (c) Meech, S. R.; Hoff, A. J.; Wiersma, D. A. Proc. Natl.
Acad. Sci. USA 1986, 83, 9463.
2. (a) Sancar, A. Chem. Rev. 2003, 103, 2203. (b) Yu, X.; Eymur, S.; Singh,
V.; Yang, B.; Tonga, M.; Bheemaraju, A.; Cooke, G.; Subramani, C.;
Venkataraman, D.; Stanley, R. J.; Rotello, V. M. Phys. Chem. Chem. Phys.
2012, 14, 6749.
3. Grätzel, M. Acc. Chem. Res. 2009, 42, 1788.
4. (a) Iwamura, M.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2007, 129,
5248. (b) Klosterman, J. K.; Iwamura, M.; Tahara, T.; Fujita, M. J. Am.
Chem. Soc. 2009, 131. 9478.
5. Mondal, J. A.; Sarkar, M.; Samanta, A.; Ghosh, H. N.; Palit, D. K. J. Phys.
Chem. A 2007, 111, 6122.
6. (a) Nicolet, O.; Banerji, N.; Pages, S.; Vauthey, E. J. Phys. Chem. A 2005,
109, 8236. (b) Mohammed, O. F.; Vauthey, E. J. Phys. Chem. A 2008, 112,
5804.
7. (a) Lippert, E. Angew. Chem. 1961, 73, 695. (b) Grabowski, Z. R.;
Rotkiewies, K.; Rettig, W. Chem. Rev. 2003, 103, 3899.
8. (a) Kwok, W. M.; Ma, C.; Matousek, P.; Parker, A. W.; Phillips, D.; Toner,
W. T.; Towrie, M.; Umapathy, S. J. Phys. Chem. A 2001, 105, 984. (b)
Kwok, W. M.; Ma, C.; George, M. W.; Grills, D. C.; Matousek, P.; Parker,
A. W.; Philips, D.; Toner, W. T.; Towrie, M.; Photochem. Photobiol. Sci.
2007, 6, 987.
9. (a) Kohn, A.; Hattig, C. J. Am. Chem. Soc. 2004, 126, 7399. (b) Gomez, I.;
Reguero, M.; Boggio-Pasqua, M.; Robb, M. A. J. Am. Chem. Soc. 2005,
127, 7119. (c) Cogan, S.; Zilberg, S.; Hass, Y. J. Am. Chem. Soc. 2006, 128,
3335.
Chapter 4
125
10. Rhinehart, J. M.; Mehlenbacher, R. D.; McCamant, D. J. Phys. Chem. B,
2010, 114, 14646.
11. (a) Gilabert, E.; Lapouyade, R.; Rullière, C. Chem. Phys. Lett. 1988, 145,
262.(b) Gilabert, E.; Lapouyade, R.; Rullière, C.; Chem. Phys. Lett. 1991,
185, 82. (c) Lapouyade, R.; Czechka, K.; Majenz, W.; Rettig, W.; Gilabert,
E.; Rullière, C. J. Phys. Chem. 1992, 96, 9643. (d) Eilers-Konig, N.; Kuhne,
T.; Schwarzer, D.; vohringer, P.; Schroeder, J. Chem. Phys. Lett. 1996, 253,
69.
12. (a) II’chev, Y. V.; Zachariasse, K. A. Ber. Bunsen-Ges Phys. Chem. 1997,
101, 625. (b) Abraham, E.; Oberlé, J.; Jonusauskas, G.; Lapouyade, R.;
Rullière, C. Chem. Phys. 1997, 214, 409.
13. (a) Kovalenko, S. A.; Schanz, R.; Senyushkina, T. A.; Ernsting, N. P. Phys.
Chem. Chem. Phys. 2002, 4, 703. (b) Amatatsu, Y. Chem. Phys. Lett. 2003,
369, 673. (c) Druzhinin, S. I.; galievsky, V. A.; yoshihara, T.; Zachariasse,
K. A. J. Phys. Chem. A 2006, 110, 12760. (d) Arzhantsev, S.; Zachariasse,
K. A.; Maroncelli, M. J. Phys. Chem. A 2006, 110, 3454. (e) Kubicki, A. A.
Chem. Phys. Lett. 2007, 439, 243.
14. (a) Rijkerberg, R. A.; Bebeelar, D.; Buma, W. J.; Hofstraat, J. W. J. Phys.
Chem. A 2002, 106, 2446. (b) Makowska-Janusik, M.; Reis, H.;
Papadopoulos, M. G.; Economou, I. G.; Zacharopoulos, N. J. Phys. Chem. B
2004, 108, 588. (c) Yan, J.; Liu, L.; Ji, L.; Ye, M.; Xu, L.; Wang, W. J.
Phys. D: Appl. Phys. 2004, 37, 1597.
15. (a) Nakabayashi, T.; Wahadoszamen, M.; Ohta, N. J. Am. Chem. Soc. 2006,
127, 7041. (b) Benkova, Z.; Cernusak, I.; Zahradnik, P. Struct. Chem. 2006,
17, 287. (c) Georgiadou, D. G.; Vasilopoulou, M.; Pistolis, G.; Palilis, L.;
Dimotikali, D.; Argitis, P. Phys. Stat. Sol. 2008, 205, 2526.
16. (a) Lippert, E. Z. Naturforsch. 1955, 10a, 541. (b) Schulte-Frohlinde, D.;
Blume, H.; Güsten, H. J. Phys. Chem. 1962, 66, 2486. (c) Gegiou, D.;
Muszkat, K. A.; Fishcer, E. J. Phys. Chem. 1968, 90, 3907.
Chapter 4
126
17. (a) Bent, D. V.; Schulte-Frohlinde, D. J. Phys. Chem. 1974, 78, 451. (b)
Schulte-Frohlinde, D.; Görner, H. Pure Appl. Chem. 1979, 51, 279. (c)
Görner, H. Ber. Bunsenges Phys. Chem. 1984, 88, 1199.
18. (a) Görner, H. J. Photochem. Photobiol. A 1987, 40, 325. (b) Gruen, H.;
Görner, H. J. Phys. Chem. 1989, 93, 7144. (c) Görner, H. Ber. Bunsenges
Phys. Chem. 1998, 102, 726. (d) Gurzadyan, G.; Görner, H. Chem. Phys.
Lett. 2000, 319, 164.
19. (a) Oberlé, J.; Abraham, E.; Jonusauskas, G.; Rullière, C. J. Raman
Spectrosc. 2000, 31, 311. (b) Oberlé, J.; Jonusauskas, G.; Abraham, E.;
Lapouyade, R.; Rullière, C. Bull. Chem. Soc. Jpn. 2002, 75, 1041.
20. Lapouyade, R.; Kuhn, A.; Letard, J.; Rettig, W. Chem. Phys. Lett. 1993,
208, 48.
21. Moran, A. M.; Bartholomew, G. P.; Bazan, G. C.; Kelley, A. M. J. Phys.
Chem. A 2002, 106, 4928.
22. Yang, J. -S.; Liau, K. -L.; Hwang, C. -Y.; Wang, C. -M. J. Phys. Chem. A
2006, 110, 8003.
23. Farztdinov, V. M.; Ernsting, N. P. Chem. Phys. 2002, 277, 257.
24. Vijayakumar, T.; Joe, I. H.; Nair, C. P. R.; Jayakumar, V. S. Chem. Phys.
2008, 343, 83.
25. Petsalakis, I. D.; Georgiadou, D. G.; Vasilopoulou, M.; Pistolis, G.;
Dimotikalo, D.; Argitis, P.; Theodorakopoulos, G. J. Phys. Chem. A 2010,
114, 5580.
26. Murugan, N. A.; Kongsted, J.; Rinkevicius, Z.; Aidas, K.; Mikkelsen, K. V.;
Agren, H. Phys. Chem. Chem. Phys. 2011, 13, 12506.
27. (a) Autschbach, J. ChemPhysChem. 2009, 10, 1757. (b) Niehaus, T. A.;
March, N. H. Theor. Chem. Acc. 2009, 125, 427. (c) List, N. H.; Olsen J.
M.; Rocha-Rinza, T.; Christiansen, O.; Kongsted, J. Int. J. Quantum Chem.
2012, 112, 789.
Chapter 4
127
28. (a) Rafiq, S.; Yadav, R.; Sen, P. J. Phys. Chem. A 2011, 115, 8335. (b)
Rafiq, S.; Rajbongshi, B. K.; Nair, N. N.; Sen, P.; Ramanathan, G. J. Phys.
Chem. A 2011, 115, 13733.
29. Koti, A. S. R.; Krishna, M. M. G.; Periasamy, N. J. Phys. Chem. A 2001,
105, 1767.
30. El-Sayed, M. A. J. Chem. Phys. 1963, 38, 2834.
128
Chapter 4
Chapter 5
Establishing the Presence
of an Activation Barrier in an
Otherwise Barrierless PES of Auramine-O
Shahnawaz Rafiq et al., J. Chem. Phys. 2013, 139, 124302.
130
Chapter 5
This chapter questions the widely acclaimed model for the excited state
relaxation dynamics of auramine-O involving orientational relaxation of
dimethylanilino moieties along the barrierless excited state potential energy
surface (PES). Such a model would necessitate similar excited state dynamics in
media offering similar viscous drag. However, an interesting experimental
observation shows auramine-O to have ca. 8 times large fluorescence quantum
yield in chloroform than in methanol, though both solvents are having same
viscosity. The femtosecond fluorescence transients of auramine-O in chloroform
surprisingly depict a rise time of ca. 200 fs which has not been observed before.
This, along with the simultaneous observation of unexpectedly large fluorescence
lifetime and multi-exponential transients in chloroform questions the thoroughly
accepted barrierless model of auramine-O relaxation dynamics, as the barrierless
model would demand a short lifetime and mono-exponential decays. Temperature
dependent quantum yield measurements along with solvent dependent excited state
multi-coordinate calculations further unveil the exact nature of PES. These all
results concomitantly conclude that in chloroform, upon photo-excitation
auramine-O must pass over an activation barrier in space and time before
damping the excited state population into ground state via a sink function through
adiabatic coupling of the electronic states.
Chapter 5
131
5.1. Introduction
Ultrafast spectroscopy of intramolecular re-orientational relaxation dynamics
has been studied to a great extent to explore their utility as molecular rotors.1-9 A
representative class of such molecules is triphenylmethane dyes which include
malachite green, crystal violet, brilliant green, etc.9-12 The use of malachite green to
act as a reporter of the microviscosity of nano-confined water has been
experimentally illustrated by us exploring its property of being a barrierless
molecular rotor.6 Auramine-O (AO) also shows similar type of relaxation
behaviour as has been reported for triphenylmethane dyes.10 Hasegawa and coworkers exploited the barrierless torsional motion of AO to estimate the
microviscosity of AOT reversed micellar system.13 AO incurs its applicability as a
fluorescent probe for investigating the structure and function of proteins,14,15
detection of amyloid fibrils,16 reporter of polymerization,17,18 etc., primarily
accounted to its barrierless excited state potential energy surface (PES) along the
torsional coordinates of dimethylanilino groups. In 1956, Oster and Nishijima for
the first time postulated the role of internal rotation in AO by studying the effect of
viscosity on its quantum yield.19 Forster and Hoffmann20 (FH) proposed that
torsional motion of the phenyl rings occur along the barrierless PES leading to
radiationless decay to the ground state, and reported an 2/3 dependence of
fluorescence quantum yield. Bagchi, Fleming and Oxtoby21 (BFO) modelled such
barrierless process considering the participation of a position dependent or
nonlocal sink function on the parabolically shaped PES. This sink function induces
an enhanced non-radiative decay rate and once the system has relaxed to the
minimum of the PES, the population density drops drastically and henceforth
subsides the fluorescence intensity.21 Such a model would necessitate the single
exponential nature of excited state relaxation. Martin and Glasbeek22-25 however
opposed the FH and BFO models on the basis of non-exponential behaviour of
excited state of AO. Hirose et al. proposed that the relaxation dynamics of AO is
effected not only by solvent viscosity but also by ultrafast solvation dynamics.26
Meech and coworkers27 proposed that in bulk water, since the solvation is very
Chapter 5
132
fast, it promotes a facile barrierless formation of charge transfer state and hence
relaxation of AO is decided primarily by aqueous solvation and not by the solvent
viscosity. Palit and co-workers28 suggested the involvement of two transition states
during the excited state relaxation of AO in aprotic and alcoholic solvents. The
photophysics of AO thus poses a question mark to the widely accepted barrierless
nature of PES.
The work presented here, is based on an interesting observation about the
fluorescence quantum yield of AO in two solvents of same viscosity, namely
methanol and chloroform. As shown in figure 5.1, AO was found to have large
quantum yield in chlorinated solvents, especially chloroform than alcohols and
other polar solvents. Considering the barrierless nature of PES, the fluorescence
quantum yield is expected to be identical in solvents of similar viscosity. However,
the observation is otherwise. In this contribution, we try to explain such interesting
observation
using
time
resolved
femtosecond fluorescence measurements,
temperature
dependent
fluorescence
properties along with time dependent density
functional theory calculations.
Scheme 5.1. Structure of auramine O
5.2. Results and Discussion
5.2.1. Steady State and Time Resolved Fluorescence Measurements
The fluorescence quantum yield (f) of AO in methanol is 2.4 x 10-4 ( 0.1 x
10-4) (taking Coumarin 152 in 50% ethanol-in-water as standard). In chloroform, f
is 2.0 x 10-3 ( 0.1 x 10-3), which is 8 times larger than in methanol, though having
same viscosity (Figure 5.1). To understand this anomaly, we have carried out
ultrafast femtosecond dynamics of AO in these media. In methanol, fluorescence
transients of AO were measured at wavelengths ranging from 430 to 600 nm by
exciting at 405 nm. All the transients can be fitted by a sum of two exponentials,
with first time component (1) of < 1 ps and second time component (2) stretching
Chapter 5
133
till 5 ps (Figure 5.2(a)). For 430 nm, 1 = 70 fs and 2 = 420 fs, at the peak
wavelength (480 nm), 1 = 300 fs and 2 = 825 fs, and on the red side of the
emission spectrum (600 nm), 1 = 1.1 ps and 2 = 4.0 ps (Table 5.1(a)). The
average lifetime increases from 100 fs at 430 nm to 1.50 ps at 600 nm. The
Figure 5.1. Steady state fluorescence spectra of auramine-O in various solvents of different
viscosities and different dielectric constants
fluorescence transients at various wavelengths were combined with steady state
fluorescence spectrum to reconstruct time resolved emission spectra (TRES). 29 The
reconstructed experimental points and the lognormal fits of the TRES in methanol
are shown in figure 5.2(b) and clearly infer the occurrence of dynamic Stokes shift
and broadening of the emission spectra with time. An overall Stokes shift of 1100
cm-1 could be observed till 2.5 ps, at which the population is just 2.8% of the
population at zero time. Similar to the previous time resolved studies, 22-26 we also
have not observed any rise time in the transients of AO in methanol which is in
agreement with the barrierless nature of excited state PES on behalf of the rapid
loss of fluorescence intensity on a timescale much faster than the Stokes shift. In
order to delve into the above mentioned steady state observation, femtosecond
fluorescence transients were measured for AO in chloroform (Figure 5.3(a)). The
transient at 430 nm is best fitted by three exponentials with time constants 150 fs,
134
Chapter 5
Figure 5.2. (a) Femtosecond fluorescence transients of auramine-O in Methanol (shown till 5
ps) at different wavelengths covering the full range of emission spectrum, and (b)
Reconstructed time resolved emission spectra (TRES) of auramine-O in methanol. Crosses
() represent the reconstructed experimental intensity and solid lines are the lognormal fits to
the experimental data.
Figure 5.3. (a) Femtosecond fluorescence transients of auramine-O in chloroform recorded at
different wavelengths (shown till 10 ps), and (b) Reconstructed time resolved emission
spectra (TRES) of auramine O in chloroform. Crosses () represent the reconstructed
experimental intensity and solid lines are the lognormal fits.
840 fs, and 3.72 ps (Table 5.1(b)). Till 530 nm, the transients can still be fitted with
three exponentials with all the three time components having positive amplitude.
The average lifetime increases from 1.24 ps to 23.02 ps in this wavelength region.
The transient at 540 nm, however, shows first time constant with negative
Chapter 5
135
amplitude, and the time constants are 130 fs (-0.50), 5.09 ps (1.21), and 27.01 ps
(0.29). Such rise time is consistently observed for all the wavelengths beyond 540
Table 5.1. Fitting parameters of femtosecond fluorescence transients of auramine O in (a)
Methanol, (b) Chloroform, and (c) Dimethylsulfoxide.
em, nm
1a, ps (a1)
2 a, ps (a2)
3 a, ps (a3)
(a) In Methanol ( = 0.54 cP)
430
440
450
460
470
480
490
500
510
520
530
540
550
560
580
600
430
440
450
460
470
480
490
500
510
520
530
540
550
565
580
595
610
440
490
550
100 fs
a
0.07 (0.91)
0.42 (0.09)
0.19 (0.84)
0.64 (0.16)
0.26 (0.75)
0.74 (0.25)
0.24 (0.61)
0.72 (0.39)
0.29 (0.58)
0.80 (0.42)
0.30 (0.51)
0.82 (0.49)
0.30 (0.42)
0.84 (0.58)
0.22 (0.30)
0.83 (0.70)
0.29 (0.28)
0.89 (0.72)
0.69 (0.74)
1.31 (0.26)
0.82 (0.82)
1.31 (0.18)
0.89 (0.37)
1.01 (0.63)
0.66 (0.66)
1.63 (0.34)
0.83 (0.75)
1.95 (0.25)
0.99 (0.85)
2.74 (0.15)
1.09 (0.87)
4.02 (0.13)
(b) In Chloroform ( = 0.54 cP)
0.15 (0.63)
0.84 (0.25)
0.22 (0.57)
1.24 (0.33)
0.35 (0.49)
1.78 (0.39)
0.50 (0.42)
2.74 (0.49)
0.57 (0.37)
3.11 (0.53)
0.68 (0.35)
3.61 (0.56)
1.03 (0.34)
4.32 (0.57)
1.20 (0.31)
4.72 (0.60)
1.63 (0.33)
5.48 (0.57)
1.81 (0.32)
5.72 (0.57)
2.63 (0.43)
7.59 (0.46)
0.13 (-0.50)
5.08 (1.21)
0.13 (-0.65)
4.00 (1.24)
0.19 (-0.63)
4.57 (1.12)
0.18 (-0.45)
5.12 (0.97)
0.16 (-0.80)
5.01 (1.03)
0.19 (-0.72)
5.29 (0.92)
(c) In Dimethylsulfoxide ( = 1.99 cP)
0.16 (0.82)
1.06 (0.18)
0.55 (0.75)
2.74 (0.25)
0.61 (0.43)
3.10 (0.57)
3.72 (0.12)
5.47 (0.10)
6.97 (0.12)
10.98 (0.09)
13.86 (0.10)
16.81 (0.09)
21.93 (0.09)
25.72 (0.09)
30.77 (0.10)
32.52 (0.11)
39.98 (0.11)
27.01 (0.29)
23.01 (0.40)
29.95 (0.51)
37.69 (0.48)
35.38 (0.77)
37.52 (0.90)
136
Chapter 5
nm as shown in table 5.1(b). This is for the first time, that someone has observed a
rise in the fluorescence transients of AO. Till now, all the work related to excited
state dynamics of AO in several solvents (water, ethanol, decanol, etc.), does not
report any rise time in fluorescence transients.23-25,27 The proposed models of
barrierless excited state relaxation dynamics of AO in the last three decades20-27
also does not explain the observed large average fluorescence lifetime of AO in
chloroform, although with viscosity same as that of methanol. The time scale of the
decay of AO in DMSO, though being 4 times more viscous, is even faster than in
chloroform (Table 5.1(c) and Figure 5.4). The two parameters, which contradict
the previous notions of AO relaxation dynamics in chloroform, are the initial rise
time, which had never been observed before, and the unexpectedly large
fluorescence lifetime (Table 5.1(b)). BFO theory also predicts the multiexponential decay in highly viscous solvents credited to inherent properties of the
relaxation in absence of a barrier,21 whereas the decay in chloroform is three
Figure 5.4. Comparison of fluorescence transients of AO in methanol, chloroform and
dimethylsulfoxide at three emission wavelengths. One can observe that the decay in
chloroform is much slower than DMSO and methanol.
Chapter 5
137
exponential, although is less viscous. In chloroform the measured dynamic Stokes
shift of 1250 cm-1 (Figure 5.3(b)) is stretched over a longer time period with actual
population decaying to 3.0% till 20 ps.
For quantitative understanding of the time evolution of the emission spectra of
AO in methanol and chloroform, first moment of the emission bands are plotted in
figure 5.5. In methanol, the variation of first moment with time can be best fitted
by single time constant of 1.50 ps, suggesting the involvement of two states in the
excited state manifold of AO. In chloroform, the time-dependent variation of the
first moment of the emission spectra is best fitted by a bi-exponential function with
two time constants of 550 fs and 14.20 ps. This suggests the involvement of three
states in the excited state manifold of AO in chloroform. In methanol, as
mentioned above, the fluorescence kinetic traces are bi-exponential and no growth
has been observed at any wavelength of the emission, while as in chloroform, the
fluorescence kinetic are found to be tri-exponential with a rise time of 100 – 200
fs. It apparently seems plausible that the process which occurs in 100 – 200 fs
timescale in AO in chloroform is much faster in methanol, and was not observed
Figure 5.5. Evolution of the reconstructed fluorescence spectra of auramine in Methanol
(Red,▲) and Chloroform (Black,●) in terms of first moment. The solid black lines are the
fitted lines.
138
Chapter 5
within the current instrument response limit. Since the solvation time of methanol
(~5 ps)30 is more than the solvation time of chloroform (~ 2.8 ps), 30 it would
necessitate that if this rise time is because of solvation, then it certainly have been
observed in methanol as well. The solvation process can be also eliminated based
on the fact that the solvation time of chloroform cannot be 100 - 200 fs (observed
rise time). Meech and co-workers also proposed that in slow solvent response
media, the AO excited state relaxation dynamics is not controlled by the solvation
process, rather microscopic viscosity is the controlling factor.27 Considering this
observation, and also knowing the structure of the highest occupied molecular
orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) from
theoretical calculations (vide infra), the first process occurring in the excited state
relaxation of AO was assigned to the charge transfer process from locally excited
(LE) state to the charge transfer (CT) state. This is to emphasize that the charge
transfer process was authenticated only in chloroform and not in methanol. This
may be due to the fact that the CT state is more stabilized in methanol than in
chloroform, which will cause the depletion of the LE state to the CT state to occur
at much faster rate in methanol than in chloroform, and hence is not observable
within the present time resolution limit. Based on these facts, the first spectral shift
of 550 fs obtained in the time-dependent variation of first moment of AO in
chloroform was assigned to the charge transfer process.
Once the CT state has formed from the LE state, the subsequent relaxation of
the excited state continues on behalf of the torsional motion of N,Ndimethylanilino groups resulting in the formation of twisted intramolecular charge
transfer (TICT) state, which depletes its population through conical intersection via
internal conversion to the vibrationally hot ground state. On this basis, the only
time constant in methanol, and second time constant in chloroform of timedependent variation of first moment is ascribed to the torsional motion of the N,Ndimethylanilino rings, which occurs in 1.5 ps in methanol and 14.2 ps in
chloroform. Thus we can say that the observed dynamic Stokes shift in methanol is
primarily due to the torsional motion of the rings which occur at much faster rate
Chapter 5
139
than the solvation. Hence, in this slow solvent relaxation limit, the twisting motion
determines the dynamics of the excited state relaxation. As was proposed for
Michler’s Ketone,31,32 for Michler’s Ketone,33,34 we suggest that for AO in
methanol an inhomogeneous distribution of stabilized twisting configurations may
exist, because of which the influence of short time dynamics of the solvent may be
drastically changed.10,34 The proposition of inhomogeneous distribution of twisting
configurations can be authenticated from the observation of the broadening of the
time-resolved emission spectra in figures 5.2(b) and 5.3(b).
On a similar analogy, the second time constant of 14.2 ps in the time dependent
variation of fist moment of AO in chloroform is also ascribed to the twisting
dynamics. In this case, as the solvent relaxation time of chloroform is 2.8 ps,30 the
solvation and twisting motion may occur together till the system is fully solvated,
and afterwards the twisting dynamics primarily controls the excited state
relaxation. It apparently seems that the large value of the dynamic Stokes shift due
to the twisting motion in chloroform (14.2 ps) than in methanol (1.5 ps) is due to
the larger in-homogeneity in the distribution of twisting configurations of AO in
chloroform than in methanol. This prominent in-homogeneity in chloroform may
result in the formation of a prominent activation barrier between the CT state and
the final TICT state. The presence of an activation barrier between the CT and
TICT state will not only explain the large fluorescence quantum yield in
chloroform, but will also throw light on the observed rise time and the multiexponential fluorescence transients.
5.2.2. Temperature Dependent Quantum Yield Measurements
The existence of the proposed activation barrier was further confirmed by
temperature dependent steady state fluorescence measurements of AO in methanol
and chloroform. These measurements enabled us to calculate the activation energy
for the orientational relaxation of dimethylanilino groups in AO in either solvent.
Upon decrease in temperature from 323 K to 263 K, we observe an increase in
fluorescence quantum yield of AO. As proposed, the increase of fluorescence
Chapter 5
140
intensity with decreasing temperature may be because of the presence of an
activation barrier in the excited state PES. However, one may argue that the
increasing solvent viscosity on decreasing the temperature may also lead to
increasing in fluorescence quantum yield of AO. Both the factors will increase the
fluorescence quantum yield of the molecule on behalf of decreasing the rate of
orientational relaxation, which is given by Debye-Stoke-Einstein equation;
k or  T / 
where

(5.1)
is shear viscosity of the medium and T is the absolute temperature. As the
orientational relaxation is the operative non-radiative channel in AO relaxation, we
can relate it to its fluorescence quantum yield by   k r /( k r  k or ) wherein k nr  k or
and one can obtain the following relation;7,35,36
 /(1  )  k r / k or   / T
(5.2)
This equation is applicable only if the variation of quantum yield is linear with  / T
. Otherwise the exact degree of dependence () can be estimated by following
equation;7,35,36
 /(1   )  k r / k or  C ( / T ) 
(5.3)
where C is the proportionality constant. The magnitude of  will enable us to sort
out the degree of dependence of fluorescence quantum yield on temperature of the
medium. This equation has been used to fit the experimental data as shown in
figure 5.6(a) yielding an  value of 0.42 for AO in methanol and 0.84 in
chloroform, which is a direct indicative of the fact that the degree of dependence of
quantum yield on
 /T
is significantly higher in chloroform than in methanol. Since
the change in fluorescence quantum yield is a combined effect of two factors as
mentioned above, in order to envisage the effect of viscosity alone on the rate of
orientational relaxation, one can calculate the activation energy for viscous flow
using Arrhenius-type equation;
   0 exp(E( ) / kT )
(5.4)
Chapter 5
Figure 5.6. Plots showing variation of (a)
 /(1   )
141
vs /T (K) and (b)
ln( ) vs 1 / T ( K ) ,
and (c)
1/T (K) for auramine O in methanol (Red,▲) and chloroform (Black,●). The
unit of E() and E(or) is kcal mol-1. The solid black lines are the fitted lines.
ln(1 / T (1 /   1) vs
where
E ( )
is the activation energy of the viscous flow of the solvent. Fitting the
curves with rearranged form of above equation ( ln()  ln( )  E( ) / RT) ) gives
0
E ( )
value of 2.49 kcal mol1 for methanol and 1.74 kcal mol1 for chloroform (Figure
5.6(b)).
E ( )
being higher for methanol, suggests that methanol should offer more
resistance on decreasing the temperature to the motion of dimethylanilino groups
than chloroform considering the barrierless nature of PES, and should inherently
lead to increase in quantum yield of AO in methanol than in chloroform. However
the observation is otherwise, which suggests the existence of an activation barrier
in AO excited sate PES. Waldeck and Fleming35 combined equations 5.2 and 5.4 to
calculate the actual activation energy for the rate of orientational relaxation of
molecular fragments as;
Chapter 5
142
Table 5.2. Degree of dependence and activation energies for solvent viscosity (E()) and
internal rotation rate (E(or)).
Solvent
E() kcal mol-1
E(or) kcal mol-1

Methanol
0.42
2.49
0.71
Chloroform
0.83
1.74
1.36
kor / k r  ln(1/ T (1 /   1))  ln(c / 0 )  Eor  / RT
where
E (or)
(5.5)
represents the activation energy of rate of orientational relaxation. The
above equation gives activation energy for the rate of orientational relaxation as
0.71 kcal mol1 in methanol and 1.36 kcal mol1 in chloroform as shown in figure
5.6(c) and also tabulated in table 5.2. It implies that the rotation of dimethylanilino
moieties during the excited state relaxation along PES is subjected to an inherent
activation barrier of 0.71 kcal mol-1 in methanol which is almost equal to the
energy available at room temperature (kT = 0.59 kcal mol-1) and hence we can
quantitatively ascribe barrierless nature to the excited state PES of AO in
methanol. The occurrence of a rise in fluorescence transient of AO in methanol is
thus expected, if transients are recorded at lower temperatures. In chloroform,
however the activation barrier height is almost double to that in methanol, which
may be responsible for the observation of rise time. The existence of the significant
activation barrier between the CT and the TICT states will increase the stability of
the CT state and hence justifies the high fluorescence quantum yield and observed
rise in the fluorescent transients of AO in chloroform. Accordingly the two time
constants in the time dependent variation of first moment of AO emission in
chloroform have been assigned to the charge transfer process, and the subsequent
passage of CT state to the TICT state over the activation barrier.
5.2.3. Quantum Mechanical Calculations
The existence of an excited state activation barrier and consequently the
intermediate state were further authenticated by time dependent density functional
theory calculation using Gaussian 09.37 The ground state of AO is pushed out of
planarity by a dihedral angle of 31.7 between the two dimethylanilino groups
Chapter 5
143
about the central sp2 carbon. The distribution of electronic density in HOMO and
LUMO suggest transfer of charge from dimethylamine groups towards the central
region, which present itself as a theoretical evidence of the redistribution of
electronic density once the molecule is promoted to higher electronic state (Figure
5.7). Separate optimizations were carried out in presence of chloroform and
Figure 5.7. Structure of HOMO and LUMO of auramine-O obtained at B3LYP/6-311+(g,d)
level of theory. The distribution of electronic density in the two molecular orbitals suggests
the involvement of intramolecular charge transfer in the excited state.
methanol using polarisable continuum model (PCM). TDDFT was used to
calculate the energies of 20 electronically excited states corresponding to each 361
geometries of AO by varying the dihedral angles of both the dimethylanilino
groups with 10 interval using PBE/6-311+g(d,p)/dga1 functional under PCM
approximation. The energy of the ground state and the first excited electronic state
was plotted against the theta () and phi () dihedral angles to generate three
dimensional potential energy surfaces as shown in figure 5.8. Upon photoexcitation, AO is promoted from ground state ‘O’ to the LE state ‘P’ having same
geometry as of former. The fate of the excited state is decided by the rotational
diffusion of dimethylanilino groups along the  and  dihedrals angles. As can be
observed from the excited state PES, there are two local minima ‘Q1’ and ‘Q2’
corresponding to the individual rotation of  and  dihedrals with exact coordinates
of  = 60.8,  = -31.6 (EQ1 = 43.55 kcal mol1) and  = -31.6,  = -60.8 (EQ2 =
43.58 kcal mol1) in both chloroform and methanol. In addition, there is another
minimum ‘R’ whose coordinates are  = 60.8,  = 60.8 (ER = 48.03 kcal mol-1).
144
Chapter 5
Figure 5.8. (A) Three dimensional ground and excited state potential energy surfaces of
auramine O procreated along the torsional coordinates of theta () and phi () dihedral
angles. (B) and (C) are the contour representations of the ground and excited state PES.
These surfaces are drawn while assigning zero energy to the optimized geometry ‘O’. ‘P’ is
the Franck-Condon geometry, ‘Q1’, ‘Q2’ and ‘R’ are the excited state minima, and ‘S’ is the
ground state maximum located directly below ‘R’.
The energy in brackets is the energy of these minima in chloroform relative to zero
energy of the ground state. In methanol the respective energies are E Q1 = 45.61,
EQ2 = 45.62, and ER = 49.72 kcal mol-1. Although the energy of first two minima is
less than the energy of third minimum, we still propose that the excited state sink
function is located at the bottom of third minimum (R) based on the presence of a
global maximum ‘S’ ( = 60.8,  = 60.8) in the ground state exactly below ‘R’.
This minimum in AO will act as a funnel for the radiationless decay of excited
state population into the ground state via adiabatic coupling of the electronic states.
The energy of the ground state maximum (ES) is 28.78 kcal mol-1 in chloroform
Chapter 5
145
Figure 5.9. The contour plots of the ground (bottom) and excited (top) electronic states of
auramine-O in methanol (left) and chloroform (right) along the torsional coordinates of theta
and phi dihedral angles. As can be interpreted from the excited state contours, chloroform
presents a higher barrier to the motion of the dimethylanilino moieties along the paths ‘Q1’ to
‘R’ and ‘Q2’ to ‘R’. In both solvents, excited state follows same relaxation pathways, the
only difference is the height of the activation barrier which renders auramine-O to fluoresce
more in chloroform than methanol.
and 27.27 kcal mol-1 in methanol. These all energy values though do not signify
much at the quantitative level, but they qualitatively provide necessary information
to support the experimental findings. As per energy statistics corresponding to the
minima in the excited state surface, the minimum energy path will be from ‘P’ to
‘Q1’ or ‘Q2’ due to being individually barrierless in nature. However, their
individual local paths actually lead to a global path from point ‘P’ to ‘R’ via ‘Q1’
or ‘Q2’, with an average activation barrier of 8.41 kcal mol -1 in methanol and 8.67
kcal mol-1 in chloroform (Figure 5.9). The barrier height is found to be higher in
chloroform than methanol by 0.26 kcal mol1.Although this value is insignificant
146
Chapter 5
quantitatively, however, qualitatively it presents a larger barrier to the motion of
dimethylanilino moieties in chloroform than in methanol. In TDDFT calculations,
the starting point in the excited state potential energy surface is the charge transfer
state, as identified from the femtosecond time resolved measurements. Because of
the inherent limitations of TDDFT method, it was unable to identify the charge
transfer process in the excited state.
In summary, the initial relaxation dynamics of AO from the Franck-Condon
point (LE state) to the CT state is described by a barrierless PES. After the CT
state, the PES in no longer barrierless, rather, the representative nuclear
configuration has to pass over an activation barrier to reach to the conical
intersection. It is proposed that the LE state and the immediately formed CT state
are the emissive states, while as TICT state is completely non-emissive in nature.
In methanol the charge transfer process was not observed, may be because of the
limited resolution of the measurement. Also in this case, the activation barrier
being of the order of room temperature thermal energy (kT), render the PES
barrierless and the passage from CT to TICT state occurs in 1.5 ps. While in
chloroform, the charge transfer process from the LE state occurs in 550 fs time
scale, and the time scale of the CT to TICT conversion is 14.2 ps with an activation
barrier higher than that in methanol (Scheme 5.2). The rate limiting process in the
excited state relaxation of AO is the motion of N,N-dimethylanilino rings over the
activated PES between the CT and the TICT state. The two solvents (methanol and
chloroform) since differ from each other mainly in the polarity; ETN (MeOH) =
0.762, ETN (CHCl3) = 0.259, the different excited state relaxation mechanism can
be ascribed to the difference in this property of the medium. Especially, due to
more polar nature of methanol over chloroform, the stability of the TICT states
will be more and the height of the barrier may be reduced, as was proposed by
Eisenthal and co-workers for p-dimethylaminobenzonitrile.38 This, in compliance
with the proposition of less in-homogeneity of twisting configurations in methanol
than in chloroform, indicates the existence of a barrierless PES in methanol. The
unexpectedly high fluorescence quantum yield, large excited state lifetime along
Chapter 5
147
Scheme 5.2. Excited state relaxation scheme of auramine-O in chloroform showing the
charge transfer process from LE to CT state and the subsequent twisting motion from the CT
to the TICT state over an activation barrier.
with the observation of rise time in fluorescence transients of AO in chloroform
can hence be explained by the presence of a prominent activation energy barrier
between the CT state and the TICT state.
148
Chapter 5
5.3. Conclusion
The comparative study of excited state relaxation dynamics of auramine-O in
methanol and chloroform pose questions to the widely accepted barrierless
relaxation model of excited state population decay. Observation of multiexponential decay, rise in fluorescence transients, much slower fluorescence decay,
and unusual temperature dependence of AO in chloroform, point towards the
existence of an inherent activation energy barrier in the excited state potential
energy surface. The existence of larger activation barrier of AO in chloroform than
in methanol was indicated by TDDFT calculation. Temperature dependent
fluorescence quantum yield measurements predict the height of activation barrier
to be double in chloroform than in methanol, where the PES is essentially
barrierless at room temperature.
Chapter 5
149
References
1. Takeuchi, S.; Ruhman, S.; Tsuneda, T.; Chiba, M.; Taketsugu, T.; Tahara, T.
Science 2008, 322, 1073.
2. Rafiq, S.; Sen, P. J. Chem. Phys. 2013, 138, 84308.
3. Sahoo, D.; Chakravorti, S. Phys. Chem. Chem. Phys. 2008, 10, 5890.
4. Duxbury, D. F. Chem. Rev. 1993, 93, 381.
5. Bulic, B.; Pickhardt, M.; Schmidt, B.; Mandelkow, E. M.; Waldmann, H.;
Mandelkow, E. Angew. Chem. Int. Ed. 2009, 48, 1741.
6. Rafiq, S.; Yadav, Y.; Sen, P. J. Phys. Chem. B 2010, 114, 13988.
7. Stsiapura, V. I.; Maskevich, A. A.; Kuzmitsky, V. A.; Uversky, V. N.;
Kuznetsova, I. M.; Turoverov, K. K. J. Phys. Chem. B 2008, 112, 15893.
8. Amdursky, N.; Erez, Y.; Huppert, D. Acc. Chem. Res. 2012, 45, 1548.
9. Sulatskaya, A. I.; Kuznetsova, I. M.; Turoverov, K. K. J. Phys. Chem. B 2012,
116, 2538.
10. Glasbeek, M.; Zhang, H. Chem. Rev. 2004, 104, 1929.
11. Ben-Amotz, D.; Harris, C. B. J. Chem. Phys. 1987, 86, 4856.
12. Nagasawa, Y.; Ando, Y.; Okada, T. Chem. Phys. Lett. 1999, 312, 161.
13. Hasegawa, M.; Sugimura, T.; Suzaki, Y.; Shindo, Y. J. Phys. Chem. 1994, 98,
2120.
14. Steiner. R. F.; Albaugh, S.; Nenortas, E.; Norris, L. BioPolymers 1992, 32, 73.
15. Weers, J. G.; Maki, A. H. Biochemistry 1986, 25, 2897.
16. Amdursky, N.; Huppert, D J. Phys. Chem. B 2012, 116, 13389.
17. Wang, Y.; Morawetz, H. Macromolecules 1986, 19, 1925.
18. Pereira, R. V.; Gehlen, M. H. J. Phys. Chem. B 2006, 110, 6537.
19. Oster, G.; Nishijima, Y. J. Amer. Chem Soc. 1956, 78, 1581.
20. Forster, T.; Hoffmann, G. Z. Phys. Chem. N. F 1971, 75, 63.
21. Bagchi, B.; Fleming, G. R.; Oxtoby, D. W. J. Chem. Phys. 1983, 78, 7375.
22. Martin, M. M.; Plaza, P.; Changenet, P.; Meyer, Y. H. J. Photochem.
Photobiol. A: Chem. 1997, 105, 197.
Chapter 5
150
23. Changenet, P.; Zhang, H.; van der Meer, M. J.; Glasbeek, M.; Plaza, P.;
Martin, M. M. J. Phys. Chem. A 1998, 102, 6716.
24. van der Meer, M. J.; Zhang, H.; Glasbeek, M. J. Chem. Phys. 2000, 112, 2878.
25. Glasbeek, M.; Zhang, H.; van der Meer, M. J. J. Mol. Liq. 2000, 86, 123.
26. Hirose Y.; Yui, H.; Sawada, T. J. Phys. Chem. B 2004, 108, 9070.
27. Kondo, M.; Heisler, I. A.; Meech, S. R. J. Phys. Chem. B 2010, 114, 12859.
28. Singh, C.; Modak, B.; Mondal, J. A. K. J. Phys. Chem. A 2011, 115, 8183.
29. Rafiq, S.; Rajbongshi, B. K.; Nair, N. N.; Sen. P.; Ramanathan, G. J. Phys.
Chem. A 2011, 115, 13733.
30. Horng, M. L.; Gardecki, J. A.; Papazyan, A.; Maroncelli, M. J. Phys. Chem.
1995, 99, 17311.
31. Fonseca, T.; Ladanyl, B. M. J. Phys. Chem. 1991, 95, 2116.
32. Singh, A. K.; Palit, D. K.; Mittal, J. P. Res. Chem. Intermed. 2001, 27, 125.
33. Su, S. –G.; Simon, J. D. J. Phys. Chem. 1989, 93, 753.
34. Kim, H. J.; Hynes, J. T. J. Photochem. Photobiol. A 1997, 105, 337.
35. Waldeck, D. H.; Fleming, G. R. J. Phys. Chem. 1981, 85, 2614.
36. Loutfy, R. O.; Arnold, B. A. J. Phys. Chem. 1982, 86, 4205.
37. Frisch, M. J. et al. Gaussian 09, Revision A.1, Gaussian, Inc., Wallingford CT,
2009.
38. Hicks, J.; Vandersall, M.; babarogic, Z.; Eisenthal, K. B. Chem. Phys. Lett.
1985, 116, 18.
Chapter 6
Ultrafast trans-cis Isomerization of GFP
Chromophore Analogs and the
Effect of Protein Scaffold
Shahnawaz Rafiq et al., Manuscript under preparation
Chapter 6
152
Torsional motion mediated multi-coordinate relaxation pathway was
introduced to explain the excited state relaxation behaviour of two analogs of
green
fluorescent
protein
(GFP)
chromophore,
namely
(4Z)-4-(4-N,N-
Dimethylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPI)
and
(4Z)-4-(4-N,N-Diphenylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-
imidazolin-5-one
(DPPI),
using
femtosecond
fluorescence
up-conversion,
femtosecond transient absorption and quantum mechanical calculations. The
fluorescence transients are found to be bi-exponential in nature. The first
component with few picosecond timescale has been assigned to the formation of
the twisted intramolecular charge transfer (CT) from the directly excited FranckCondon state along the torsional coordinate of the donor N,N-di-substituted amine
moiety. The existence of the CT state is authenticated from both the solvent
polarity and viscosity dependent dynamics. Completely non-fluorescent nature and
absence of triplet yield emphasizes that internal conversion must be the only
possible channel for the excited state relaxation back to the ground state. Viscosity
dependence of the second time component rules out the volume conserving
simultaneous rotation of the bridging bonds, referred to as “Hula twist”. The other
probable non-radiative decay channel is either a rotation about a single or a
double bond. Time dependent density functional theory calculations qualitatively
witnessed an activated channel along the twist coordinate of exocyclic double bond
channelizing the charge transfer state to the point of conical intersection between
S1 and S0 electronic surfaces. Phenomenological kinetic relaxation scheme was
formulated based on the global analysis of transient absorption data of DPI and
DPPI in methanol, which confirms the existence of three states in S 1 potential
energy surface (PES); which has been ascribed to locally excited state, charge
transfer state, and conical intersection existing between the excited and the ground
state surfaces. This detailed study reveals the role of protein scaffold in
suppressing the non-radiative pathways and hence enabling native GFP highly
fluorescent.
Chapter 6
153
6.1. Introduction
Crystal structure of green fluorescent protein (GFP)1-2 from Aequorea victoria
revealed that the chromophore, para-hydroxybenzylideneimidazolin-5-one (HBDI)
is rigidly held by a network of hydrogen bonds at the centre of an 11-stranded βbarrel formed by the protein coat. The chromophore of GFP is formed via a posttranslational internal cyclization of the Ser65-Tyr66-Gly67 tripeptide followed by
dehydrogenation of the Tyr66 residue without the aid of any cofactor or substrate
other than oxygen.3-5 Because of the autocatalytic mechanism of chromophore
formation, genetic encoding of GFP expresses the gene with in situ generation of
an intense green fluorescence. Fusion of GFP to a protein does not seem to alter
the function of the host protein. GFP is resistant to photo-bleaching, heat, alkaline
pH, salts, detergents and many proteases.6 All these properties of GFP together
with its intense fluorescence properties enable its vast range of applications from
gene expression studies7-9 to fluorescence resonance energy transfer (FRET)
studies10-12 which revolutionized the field of cell biology. One of the most
intriguing observations with GFP is that its loses fluorescence when denatured with
proteases.13-14 Peptide fragments containing the GFP chromophore produced on
denaturation of GFP behaved essentially as non-fluorescent (fluorescence quantum
yield
~
2×10-4).15
Interestingly,
synthetic
model
compounds
of
GFP
chromophore15-17 and some blue and red shifted mutants of GFP18-20 are also
similarly, weakly fluorescent. The loss of fluorescence of GFP up on denaturation
is the result of collapse of the rigid hydrogen bonded network structure and gain of
torsional degrees of freedom by the chromophore. At low temperatures, or in
solvent glasses which freeze internal rotational degrees of freedom of the solute
molecules, the model chromophores of GFP become as fluorescent as GFP
itself.15,21,22 Over one decade there has been extensive research to understand the
underlying mechanism of the loss of fluorescence in weakly fluorescent GFP
mutants and synthetic model compounds of GFP chromophore.15-20 Dihedral angle
freedom promoted internal conversion (IC) is being suggested as the cause of weak
fluorescent properties of these model chromophores. Weak fluorescence property
Chapter 6
154
is
essentially
a
characteristic
feature
of
flexible
molecules
such
as
triphenylmethane dyes where rotation of the phenyl rings is known to be the cause
of dominant radiationless relaxation.23-25 The fact that GFP chromophore model
compounds undergo IC by torsional relaxation is evidenced by temperature
dependent fluorescence studies21,22 and highly fluorescent GFP chromophore
model compounds where the rotation along the bridging single and double bonds
are locked.26 In the nearest analogs of the GFP chromophore – phydroxybenzylideneimidazolin-5-one (p-HBDI) rotation along the bridging single
bond,27-29 exo-methylene double bond16,30-32 as well as a concerted rotation along
both bonds known as hula-twist21,22, 33-36 has been controversially suggested over
the last decade to be responsible for internal conversion. In contrast to the
mechanism of IC in p-HBDI, internal conversion in model GFP chromophore
analogs containing strong electron donors such as dimethylamino group, has been
rigorously shown by us to occur via rotation along the bridging exo-methyelene
double bond.37 A multi-coordinate relaxation involving intramolecular charge
transfer (ICT) along the twist coordinate of the donor group and torsional motion
along the bridging exo-methylene double bond operates in the radiationless
deactivation of these chromophores. Both ICT and IC were observed to be highly
dependent on bulkiness of the donor, and also on viscosity. In this chapter we
throw light on the excited state relaxation mechanism of two other structural
analogs of GFP differing in the nature of substituent, and try to find out how these
substituents will affect the excited state dynamics of these analogs.
6.2. Results
6.2.1. Steady State Absorption and Emission Measurements
The UV-vis absorption spectra of two GFP synthetic chormophore analogs
(Scheme
6.1),
(4Z)-4-(4-N,N-Dimethylaminobenzylidene)-1,2-diphenyl-1,4-
dihydro-5H-imidazolin-5-one
(DPI),
and
(4Z)-4-(4-N,N-
Diphenylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPPI)
were measured in a number of solvents with different viscosity and polarity. Both
Chapter 6
155
Scheme 6.1. Structure of the two GFP synthetic chromophore analogs; (4Z)-4-(4-N,NDimethylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPI) and (4Z)4-(4-N,N-Diphenylaminobenzylidene)-1,2-diphenyl-1,4-dihydro-5H-imidazolin-5-one (DPPI).
the analogs show a single broad absorption band located between 400 nm and 480
nm in the range of eight solvents. All the spectra have similar shape except that in
cyclohexane, where a vibrational structure can be seen. The position of the
absorption maximum depends on the nature of the substituent in the benzylidene
ring, and on polarity of the solvent. In contrast to p-HBDI which shows a mild
solvatochromism,38 the two analogs show substantial bathochromic shift with
increase in solvent polarity. In cyclohexane DPI absorbs at 440 nm, while in
methanol and 50% v/v glycerol in methanol, the absorption is red shifted to 468
nm and 476 nm respectively, because of the increasing solvent polarity. The
absorption of DPPI is similarly red shifted from 452 nm in cyclohexane to 466 nm
in methanol and 473 nm in 50% v/v glycerol in methanol. In DPI, the large
bathochromic shift arises from strong coupling of the electron donating effect of
the dimethylamino group with the electron withdrawing effect of the carbonyl
group of the imidazolin-5-one ring.39 In DPPI, such coupling is reduced because of
delocalization of the lone electron pair present on the nitrogen atom as the two
benzene rings are connected to the nitrogen atom, which enables less
solvatochromism in DPPI compared to DPI. The fluorescence spectra of both the
analogs are shown in figure 6.1. In non-polar solvents, prominent vibrational
structure is observed. In contrast to p-HBDI, where fluorescence spectra show
small solvatochromism and weak dependence on solvent viscosity,38,40 the
fluorescence spectra of DPI and DPPI show significantly large bathochromic shift
156
Chapter 6
Figure 6.1. Steady state fluorescence spectra of (a) DPI and (b) DPPI in a mixture of varying
proportions of glycerol in methanol. This clearly indicates the increase in fluorescence
intensity with increasing solvent viscosity. (c) and (d) represent the fluorescence spectra of
DPI and DPPI respectively in solvent of different polarities indicating the occurrence of
bathochromic shift with increasing solvent polarity.
with increasing solvent polarity and the fluorescence intensity strongly depends on
solvent viscosity. In cyclohexane, DPI emits at 483 nm, while in methanol and
50% v/v glycerol in methanol, the emission is red shifted to 538 nm and 547 nm
respectively. The increasing bathochromic shift with solvent polarity suggests
increase in dipole moment in the excited state and therefore intramolecular charge
transfer (ICT) may be anticipated as was suggested in similar analogs.28,37,40-42 In
cyclohexane, DPPI emits at 494 nm, while in methanol and 50% glycerol in
methanol, the emission is red shifted to 629 nm and 630 nm respectively. The
presence of phenyl rings in place of methyl groups on nitrogen in DPPI enhances
the fluorescence, and is probably due to the enhanced conjugation length in DPPI
and the effect is known as amino conjugation effect.40-42 While the emission is red
Chapter 6
157
shifted from methanol to 50% v/v glycerol in methanol in case of DPI, emission of
DPPI is not red shifted, probably due to weak coupling between electron donating
N,N-diphenylamino group and the electron withdrawing effect of the imidazolin-5one carbonyl group.
6.2.2. Femtosecond Fluorescence Up-conversion Study
Femtosecond fluorescence up-conversion measurements43 for the two analogs was
undertaken in order to understand their excited state relaxation dynamics. This
study is a follow up of the work which has already been published, wherein the
ultrafast dynamics of two chromophore analogs was reported and discussed in the
context of non-fluorescent nature of synthetic HBDI analogs and the corresponding
correlation with the wild type green fluorescent protein.37 The chromophores
included in this study differ in the nature and position of substitution, which will
have an impact on the steepness of the decay transients and hence will report the
influence of substitution on the intriguing photophysics of GFP chromophore
analogs.
6.2.2.1. Viscosity Dependent Fluorescence Kinetics of DPI and DPPI
Fluorescence transients of the two analogs reported here were obtained by
exciting the samples at 435 nm and recording the transients at wavelength
corresponding to the emission maximum. For viscosity dependent studies, a
solvent mixture ranging from pure methanol to 70% v/v glycerol in methanol was
chosen. The non-exponential fluorescence transients were recorded in a total of
eight solvent with viscosity ranging from 0.54 cP in methanol to 100.75 cP in 70%
v/v glycerol-methanol mixture. The measured fluorescence transients of DPI and
DPPI along with the fitted lines are shown in the figure 6.2(a) and 6.2(b) and the
fitting parameters are tabulated in table 6.1. In pure methanol, the two time
constants of the bi-exponential fit are 680 fs and 6.2 ps for DPI, and 8.0 ps and
38.7 ps for DPPI. As the viscosity of the medium increases from methanol to 70%
v/v glycerol-methanol mixture, the first time component increases monotonously
Chapter 6
158
Figure 6.2. Fluorescence transients of (a) DPI and (b) DPPI in a glycerol-methanol mixture,
which offers varying viscosity to the molecular relaxation. The data is best fitted by biexponential function.
Table 6.1. Bi-exponential fitted parameters 1 and
mixtures of varying viscosity.
Viscosity (cP)
DPI
Glycerola
a
1 (fs)
2 (ps)
MeOH
0.54
680
6.20
a
2
of DPI and DPPI in glycerol-MeOH
DPPI
1
a
(ps)
2
a
(ps)
8.00
38.70
1.14
830
7.38
9.40
46.00
2.41
1250
9.70
10.90
54.60
5.09
1640
13.00
12.90
63.90
10.74
1950
17.50
15.40
68.40
22.65
2800
28.00
18.00
82.20
47.77
3700
42.70
26.40
117.00
100.75
5400
74.30
41.50
172.00
100 fs
from 680 fs to 5.4 ps in DPI, and from 8.0 ps to 41.5 ps in DPPI, while as the
second time constant increases from 6.2 ps to 74.3 ps in DPI, and 38.7 ps to 172.0
ps in DPPI respectively. The extent of dependence of time constants on the
viscosity of the medium is usually represented in terms of power law empirical
relationship between the time constants and the coefficient of viscosity of the
medium as.25
Chapter 6
159
Figure 6.3. Plots showing the variation of time components with coefficient of viscosity of
glycerol-methanol mixtures. (a) first time component of DPI, (b) second time component of
DPI, (c) first time component of DPPI, and (d) second time component of DPPI. The data is
fitted by power law equation shown above.
The power factor “alpha” ( ) gives us the degree of dependence of lifetime on
viscosity of the medium, and it ranges from 0 and 1. Time constants of both the
analogs in different solvent mixtures were plotted against the coefficient of
viscosity and the plots are shown in figure 6.3. The degree of dependence of time
components
1
and
2
on the coefficient of viscosity are 0.41 and 0.61 for DPI and
in DPPI, the alpha value are found to be 0.38 and 0.32 for
1
and
2
respectively.
The significant dependence of time components on viscosity of the medium infers
the involvement of large amplitude torsional motion in the excited state relaxation
of the molecules.
Chapter 6
160
6.2.2.2. Polarity Dependent Fluorescence Kinetics of DPI and DPPI
In order to understand the effect of polarity on the decay kinetics of DPI and
DPPI, fluorescence transients were recorded in aprotic solvents of almost similar
viscosity but different polarity; namely cyclohexane, acetone, and acetonitrile. The
decay transients of DPI and DPPI in these solvents are shown in figure 6.4(a),
6.4(b) and the time constants are tabulated in table 6.2. From the fitted parameters
Figure 6.4. Fluorescence transients of (a) DPI and (b) DPPI measured in three iso-viscous
solvents but with different polarities; Cyclohexane (black), Acetone (blue) and acetonitrile
(red).
Table 6.2. Fitting parameters of bi-exponential decay profiles of DPI and DPPI in various
solvents
Solvent Details
DPI
DPPI
solvent

Cyclohexane
a
a
a
a
2.00
(cP)
0.89
(ps)
1.35 (0.49)
(ps)
13.50 (0.51)
(ps)
5.12 (0.50)
(ps)
132 (0.50)
Acetone
20.56
0.30
0.96 (0.47)
12.90 (0.53)
3.13 (0.70)
142 (0.30)
Acetonitrile
35.96
0.34
0.63 (0.50)
9.90 (0.50)
2.39 (0.67)
91 (0.33)
Methanol
32.63
0.54
0.68 (0.48)
6.20 (0.52)
8.00 (0.67)
38 (0.33)
Ethanol
24.35
1.08
1.19 (0.53)
8.60 (0.47)
9.20 (0.55)
56 (0.45)
Butanol
17.43
2.61
1.55 (0.59)
12.60 (0.41)
11.80 (0.46)
86 (0.54)
1
2
1
2
100 fs
a
we can infer that, as the polarity of the medium increases from cyclohexane to
acetonitrile, there is a gradual decrease in the value of first time component. For
DPI, time constant
1
changes from 1350 fs in cyclohexane to 630 fs in acetonitrile.
Chapter 6
161
There is also a similar trend in the first time component of DPPI, where in
1
decreases from 5.12 ps in cyclohexane to 2.39 ps in acetonitrile. Although the
variation of first time component is monotonous with increasing polarity of the
solvents, we did not observe any such sequential variation in the second time
component with the change in solvent polarity. Other than measuring transients in
solvents of different polarity and similar viscosity, fluorescence traces of DPI and
DPPI were also measured in three alcoholic solvents; methanol, ethanol, and nbutanol. The bi-exponential fitted parameters are tabulated in table 6.2 and the
transients are shown in figure 6.5(a) and 6.5(b). The first and second time
Figure 6.5. Fluorescence transients of (a) DPI and (b) DPPI measured in three alcoholic
solvents of similar polarity but different viscosities; Methanol (black), Ethanol (blue) and nButanol (red).
component in DPI increases from 680 fs to 1550 fs and 6.20 ps to 12.60 ps, while
as in DPPI there is an increase from 8.00 ps to 11.80 ps in case of first time
component and for second time component from 38 ps to 86 ps, as we change the
solvent from methanol to n-butanol. Conclusively, we can put it in words, as
hydrogen bond donating efficiency decreases from methanol to n-butanol, the
overall decay becomes slower as seen in figure 6.5(a) and 6.5(b).
6.2.3. Quantum Mechanical Calculations
Ground and excited state potential energy surfaces of one among the two
analogs (DPPI) were constructed based on quantum mechanical time dependent
162
Chapter 6
density functional theory calculations in Gaussian using B3LYP/6-311+g(d,p).44-47
The three dimensional PES’s, shown in figure 6.6(a) were obtained against the
twist coordinates of bridging single bond and exocyclic double bond, designated as
gamma () and beta () rotations respectively. The ground state is characterized by
dihedral angles of  = -2 and  = 0. Individual plots are drawn in figure 6.6(b)
and 6.6(c) showing the variation of ground and excited state potential energy of
nuclear configurations against the gamma and beta torsional coordinates, while as
restraining the other coordinates to their equilibrium value.
Figure 6.6. (A) Potential Energy Surfaces of ground electronic state (Green, S 0) and first
singlet state (Brown, S1) of DPPI using TDDFT. The surfaces are plotted with respect to
ground state optimized geometry considered as zero of energy. (B) Potential energy of S0
and S1 electronic states along  torsional coordinate keeping  = 0 and (C) Potential energy
of S0 and S1 electronic states along  torsional coordinate keeping  = -2.
Chapter 6
163
6.2.4. Transient Absorption Measurements
6.2.4.1. DPI in Methanol
The time resolved absorption studies of DPI in methanol were recorded in 1 mm
sample cell in the spectral range of 450 – 775 nm with probe delay time extended
till 40 ps as shown in figure 6.7. The OD of the solution in 1 mm path length
Figure 6.7. Time resolved absorption spectra of DPI in methanol with the annotation shown
on the right side of the graph. (A) -0.06 – 0.10 ps, (B) 0.12 – 2.50 ps, and (C) 3.0 – 40.0 ps.
In the botton panel the negative spectra represented by Red dashed line (----) is the negative
of steady state absorption spectrum of DPI in methanol and the blue (----) curve is the
negative of steady state emission spectrum of DPI in methanol. In each panel, the spectra at 1.00 ps is drawn as a reference.The black dashed arrows show the evolution of absorption
signals with time.
164
Chapter 6
sample cell was 0.3 at excitation wavelength of 400 nm. The transient absorption
(TA) spectra of DPI in methanol undergo very rapid evolution of absorption
features. The time resolved absorption spectra (Figure 6.7) show a negative signal
in the wavelength range 450 – 500 nm at very early time, which rises till 100 fs
and this signal also lies within the spectral range of steady state absorption
spectrum and hence is ascribed to the ground state bleaching (GSB) of the
molecule. The spectrum at early time is also characterized by the presence of broad
positive band ranging from 550 – 775 nm and has been ascribed to the excited state
absorption (ESA). The presence of ESA in the same spectral region as that of
fluorescence spectrum seems to mask the prominent appearance of stimulated
emission (SE) band. Keen observation of the time evolution of ESA band in the
spectral window of 550 – 775 nm from -60 fs to 100 fs depicts the shift of
absorption maximum from 660 nm to 643 nm, and this blue shift continues to
appear till 2.5 ps, wherein the ESA maximum is 620 nm. The blue shift of this
ESA band is also accompanied by continuous decrease in intensity starting from
120 fs onwards. The time evolution of the ESA band also shows an initial
appearance of a shoulder, which gradually transforms into a very prominent ESA
peak centred at 510 nm with a somewhat isosbestic point between the two bands at
ca. 540 nm (Figure 6.7(b)). The ESA band at 510 nm continues to grow till it
reached maximum at 2.5 ps, and afterwards the population starts to decrease till it
becomes zero at 40 ps. The ground state bleaching signal in the spectral range of
450 - 500 nm starts slowly decreasing in intensity from 100 fs and continues to
decrease till 40 ps.
6.2.4.2. DPPI in Methanol
The time resolved absorption spectra of DPPI in methanol is obtained by
exciting at 400 nm and the time evolution of the spectra were mapped in the
wavelength range of 450 – 775 nm. Based on the occurrence of various transient
signals, the TA spectra have been segmented in four time windows as shown in
figure 6.8. The transient spectra in the time range from -0.06 ps to about 0.3 ps is
Chapter 6
165
characterized by increase in excited state absorption signal in the spectral range of
550-775 nm. Above 0.3 ps the intensity of this ESA band decreases and starts to
evolve into another ESA band on the blue side with a maximum at ca. 525 nm. In
the figure 6.8(c) we can also see the evolution of another ESA signal which gains
intensity after 4.3 ps and reaches maximum at ca. 20.0 ps. A clear isosbestic point
Figure 6.8. Time resolved absorption spectra of DPPI in methanol with the annotation shown
on the right side of the graph. (A) -0.6 – 0.24 ps, (B) 0.30 – 2.80ps, (C) 4.3 – 18.7 ps, and (D)
= 19.7 – 174 ps. In the botton panel the negative spectra represented by Red (----) is the
negative of steady state absorption spectrum of DPPI in methanol and the blue (----) curve is
the negative of steady state emission spectrum of DPPI in methanol. In each panel, the
spectra at -1.00 ps is drawn as a reference.The black dashed solid arrows show the change in
difference absorption of the spectral signals with time.
166
Chapter 6
at 620 nm is observed during this time period. In the later time, the two apparently
merged ESA signals continues to decay and reaches almost to zero as can be
visualized in figure 6.8(d). It appears that the observation of transient absorption
signal in the wavelength region around 580 nm is obscured because of the presence
of both ESA and stimulated emission in the same spectral region, such that neither
the SE signal nor the ESA signal is prominent. Another important spectral region
in the acquired transient absorption data of DPPI is the presence of negative signal
in 450 – 500 nm spectral region which is ascribed to the ground state bleaching
based on the location of steady state absorption spectrum in the same region. This
negative signal initially rises till ca. 0.24 ps and then prevails there for a long time
with a continuous decrease in intensity and almost goes to zero when all the excite
state absorption signals have depleted. The initial rise in this negative signal is
because of the GSB within the duration of the pump pulse, and afterwards it is
actually the recovery of the ground state (GSR) which will continue till the excited
states have mostly been depopulated. On the red side of the transient absorption
data, at later time, there is another negative signal which persists till the end and is
in the same region as of the tail of the steady state emission spectrum and hence is
ascribed to the stimulated emission.
6.3. Discussion
6.3.1. Fluorescence Up-conversion and Quantum Mechanical Study
Several possible excited state relaxation pathways have been suggested to
explain the highly non-fluorescent nature of denatured green fluorescent protein
(GFP), isolated GFP chromophore, and all other synthetic analogs of paraHBDI.21,22,26,37 Here, we tried to look into the possible non-radiative relaxation
pathways operational in the excited state of two GFP chromophore analogs, which
render them highly non-fluorescent. The effect of substitution will be given
emphasis in order to understand the non-radiative channels more precisely.
The fluorescence up-conversion transients of DPI and DPPI show a dependence
on the solvent viscosity, which is reflected in terms of the significant monotonous
Chapter 6
167
increase of the two time components on increasing the viscosity of the media. Such
viscosity dependence infers the involvement of large amplitude torsional motion
within the molecule during the excited state relaxation from the Frank-Condon
state. The probable non-radiative relaxation pathways as reported earlier,21,22,26,37
may include torsional motion about the single bond, or about the exocyclic double
bond or the simultaneous rotation of both the bridging bonds famously termed as
“Hula twist”. The simultaneous rotation of the bridging bond can easily be ruled
out, because it is essentially a volume conserving process and is expected to show
no viscosity dependence. While as, the chromophores under investigation show
significant viscosity dependence as evidenced by a significant value of
from the
power law fitting of the time constants as shown in figure 6.2.
From the polarity dependent fluorescence transients it was observed that the
first time component decreases monotonously as the polarity of media increases
from cyclohexane to acetonitrile, while as no such order is observed for second
time component. This monotonous dependence of first time component on polarity
of the medium infers the existence of an intermediate state lying below the directly
excited Franck-Condon (FC) state. The rate of conversion of FC state into this
intermediate state increases with increases in polarity, which inherently points
towards the extra stability gained by the intermediate state in more polar solvents.
This can be possible if the intermediate state is having more dipole moment than
the FC state, which can occur on behalf of substantial amount of charge separation
in the intermediate state. The steady state fluorescence measurements as reported
in section 6.2.1 also indicate the existence of large dipole moment in the excited
state. In DPI and DPPI, upon photo-excitation, the electron donating amino group
can shift electron density towards the electron withdrawing carbonyl moiety in the
imidazolin-5-one group via the extensive conjugation between the donor and
acceptor moiety. The intermediate state thus will be a charge transfer (CT) state
and hence as polarity increases, the stability of CT state increases and thus the rate
of conversion from FC to CT state increases leading to fast decay of LE state. Thus
the first time constant decreases with increase in the solvent polarity. Based on
168
Chapter 6
these details, we ascribe the first time component to the intramolecular charge
transfer process from N,N-substituted amine to the carbonyl group. The existence
of CT state has been well established by the time resolved intensity normalized and
area normalized emission spectra (TRINES and TRANES) in a similar type of
molecules with same electron donor and electron acceptor moieties, in our previous
work.37
As discussed earlier, the first time component of the fluorescence transients
also show a strong dependence on the solvent viscosity, which can be explained
considering the involvement of internal rotational diffusion within the molecule
during the charge transfer process. Thus the charge transfer state is actually a
twisted intramolecular charge transfer (TICT) state. The immediate outcome of the
twisted nature of charge transfer can be rationalized in the bulkiness of the
molecules. In DPPI, the two methyl groups on amino nitrogen have been
substituted by two phenyl groups, which results in an increase in the first time
constant from 680 fs to 8.0 ps in pure methanol. This increase in the value of first
time constant from DPI to DPPI is ascribed to the increased bulkiness of the donor
group. The rotation of the bulkier group senses more restrictions compared to the
less bulky group and hence leads to an increased time constant in DPPI.
The second time component refers to the subsequent depletion of charge
transfer state. The quantum yield of such class of analogs has been found to be
very small ~10-3, which renders the molecules non-fluorescent. There is as well no
report of any triplet yield observed in these molecules, and hence the only nonradiative relaxation pathway which has been established for p-HBDI and the
analogs under consideration, is internal conversion (IC). As predicted from the
viscosity dependent relaxation dynamics, the second time component exhibits a
strong dependence on viscosity of the medium. For DPI, the second time constant
increases from 6.2 ps to 74.3 ps and in DPPI it increases from 38.7 ps to 172.0 ps
as the solvent viscosity increases from 0.54 cP (pure methanol) to 100.75 (70% v/v
glycerol in methanol) . This suggests the involvement of torsional motion
coordinate during the depletion of charge transfer state towards the point of conical
Chapter 6
169
intersection between the S1 and S0 electronic states. The important point is to
locate the torsional coordinate responsible for the non-radiative decay of the charge
transfer state.
Among the three possible excited state non-radiative pathways, the “Hula twist”
has already been ruled out based on the viscosity dependent relaxation dynamics.
The other two pathways which may be operational are the single bond rotation or
the rotation of exocyclic double bond. The three dimensional PESs along these two
torsional coordinates are shown in figure 6.6(a). The energy of S 1 electronic state
along the gamma () coordinate keeps on rising as it moves away from the directly
excited Frank-Condon state. The FC state represents the local minimum on this
surface along the single bond torsion (Figure 6.6(b)). Hence these calculations do
not suggest occurrence of any conical intersection or an avoided crossing between
the ground and the excited state along the single bond torsional coordinate.
Restraining gamma () torsional angle to -2 and twisting the beta () dihedral
angle away from its equilibrium geometry, is an equivalence of finding the effect
of torsion about exocyclic double bond (Figure 6.6(c)). As dihedral angle  rotates
away from the its equilibrium 0 position, the energy of first excited singlet state
starts rising, reaches to the maximum of the surface at around 70 of  and then
steeps down reaching to minimum at 90. At 90 the S1 energy equals the ground
state energy of that particular configuration and hence paves path for the two
electronic states to the point of conical intersection or avoided crossing. From these
calculations, the rotation about the exocyclic double bond is intervened by an
activation barrier of around ~20 kcal mol-1. Since the excited state of the molecule
involves a long range charge transfer, the potential energy barrier computed using
TDDFT may have quantitative errors.48 Thus the results here are used only to have
qualitative information about the involvement of a specific dihedral angle in the
excited state relaxation mechanism towards the ground state.
The whole excited state relaxation process starting from Frank-Condon state
through the charge transfer state towards the point of conical intersection can be
170
Chapter 6
explained in terms of an involvement of multi-coordinate torsional motion. Upon
photo-excitation, the ground state equilibrium geometry is promoted to the FrankCondon state with its nuclear coordinates intact following the Franck-Condon
principle. Within a timescale of few hundred femtoseconds for DPI and few
picoseconds for DPPI, electronic charge redistribution occurs and shuffles the
electron density from donor amino group towards the acceptor carbonyl moiety
leading to the formation of an intramolecular charge transfer state. This charge
transfer process from donor to acceptor group occurs along the twist coordinate of
the donor moiety, which in this case is N,N-di-substituted amine. Transfer of
electron density from donor to acceptor group leads into an intermediate state
having increased electron density between nitrogen and benzylidene group (Figure
6.9). This also increases electron density of the single bond and essentially reduces
double bond character of the exocyclic double bond. Such kind of charge
Figure 6.9. Electron density maps of the highest occupied molecular orbital (HOMO) and the
lowest unoccupied molecular orbital (LUMO) showing less electron density at the exocyclic
double bond in the LUMO.
redistribution induces rigidness to the N,N-di-substituted benzylidene moiety. The
ICT state, which is now characterized by having a reduced double bond character
undergoes torsional motion of substituted benzylidene group about the double
bond, and depletes via an activated pathway to the point of conical intersection.
Since the point of conical intersection or avoided crossing between the two states is
characterized by a  torsional angle of 90, it may return to the ground state either
with the same starting geometry or to the other geometrical isomer. Implies, if the
molecule being irradiated is a Z-isomer, the excited state rotation about the double
Chapter 6
171
bond, having a local minimum at 90 may return back to ground state either to the
same Z-isomer or to the E-isomer as can be visualized from scheme 6.2. The
Scheme 6.2. Complete relaxation scheme of DPI and DPPI starting from the Franck-Condon
geometry (FC) till the formation of ground state trans- and cis- isomers through conical
intersection at perpendicular geometry of the exocylic double bond.
involvement of  coordinate torsional motion has been authenticated
experimentally by 1H-NMR and HPLC measurements on the similar type of
molecules with the same donor and acceptor moiety, by confirming the presence of
both Z- and E-isomers after irradiating the Z-isomer only.37
The effect of hydrogen bonding can be visualized by monitoring the variation
of second time component of DPI and DPPI in different alcohols. In DPI,
2
increases from 6.20 to 12.60 ps, and in DPPI it increases from 38.0 to 86.0 ps as
the medium changes from methanol to n-butanol (Table 6.2). Although we cannot
rule out the effect of viscosity of the solvent media, the increase in time constant,
however, is more than as expected from the estimated effect of viscosity from the
known values of
. The additional increase in the time constant in going from
methanol to n-butanol is ascribed to the decreasing extent of hydrogen bonding
172
Chapter 6
between solvent molecules and the chromophore in excited state. Similar kind of
dependence has been seen by Vauthey and co-workers.49 In essence, the charge
transfer process results in increased negative charge on the carbonyl group, which
in turn paves path for enhancement of hydrogen bonding ability with the solvent
molecules in the excited state. The stretching vibrational modes of hydrogen bonds
act as a sink for the non-radiative decay of charge transfer state by dissipating
electronic energy into the vibrational modes of hydrogen bonds between the
carbonyl oxygen and the hydrogen of solvent molecules. The decrease of hydrogen
bond donating ability of long chain alcohols compared to the short chain alcohols
leads to inefficient transfer of energy to the vibrational modes and hence
decelerating the decay of S1 state, which consequently leads to larger time constant
of molecules in longer chain alcohols. Acetonitrile and methanol can make a better
complimentary pair to understand the effect of hydrogen bonding in a rigorous
manner. In acetonitrile, DPI and DPPI show a second time constant of 9.9 ps and
91.0 ps, while as in methanol, the same time component now changes to 6.2 ps and
38.0 ps. The two solvents under consideration have almost same coefficient of
viscosity and same dielectric constant, so the parameter which decides the
difference between the two solvents is the Kamlet-Taft parameter “ ” (index of
hydrogen bond donating tendency), which in case of acetonitrile is 0.19, and in
methanol, the value of
is 0.93. Methanol being an efficient hydrogen bond donor
furnishes a platform for efficient transfer of excited state electronic energy of DPI
and DPPI into the vibrational modes of hydrogen bonds and hence dampens the
excited electronic state more quickly than in acetonitrile.
6.3.2. Transient Absorption Study
Based on the temporal evolution of the transient absorption signals as reported
in section 6.2.4, we propose the existence of at least three distinct states in the
excited manifold of DPI in methanol as discussed below. The locally excited (LE)
state has a broad absorption spectrum with a maximum at around 660 nm. The
kinetic traces from 620 – 730 nm are best fitted by a bi-exponential function with
Chapter 6
173
the first time constants in the order of few hundred femtoseconds and the second
time constant of around 6 ps. At 660 nm, the two time constants are 770 fs and 6.0
ps. Following the up-conversion study, the first ~ 700 fs time constant is ascribed
to the decay of LE state. In the time resolved spectra, as mentioned in section
6.2.4.1, there is a shift of this ESA band to the blue side, and hence inferring the
mixture of ESA bands in this region with the first ESA having maximum at 660
nm and second ESA with a maximum at 620 nm. The temporal evolution of one
ESA band into another suggests the depletion of LE state into some intermediate
state, which decays slowly with a time constant corresponding to the second time
component of the kinetic traces measured between 620 – 730 nm (Table 6.3). The
growth of the intermediate state from the LE state is observed from the kinetic
traces in the region from 550 nm to 590 nm (Table 6.3). As mentioned in the
Table 6.3. Lifetime of four components obtained
transient absorption data of DPI in methanol.
Wavelength (nm)
1 (ps)
2 (ps)
460
0.08 (r)
470
0.09 (r)
480
0.11 (r)
485
0.19 (r)
490
0.35 (r)
500
0.47 (r)
510
0.06 (d)
0.09 (r)
520
0.08 (d)
0.18 (r)
525
0.10 (d)
0.13 (r)
530
0.07 (d)
0.16 (r)
550
0.09 (d)
0.14 (r)
560
0.08 (d)
0.11 (r)
570
0.07 (d)
0.08 (r)
580
0.07 (d)
0.07 (r)
590
0.06 (d)
0.07 (r)
620
1.06 (d)
5.97 (d)
640
0.99 (d)
6.61 (d)
660
0.77 (d)
6.00 (d)
680
0.62 (d)
5.98 (d)
700
0.57 (d)
6.56 (d)
730
0.34 (d)
6.75 (d)
by fitting the temporal profiles of the
3 (ps)
1.43 (r)
2.25 (r)
1.39 (r)
1.16 (r)
1.00 (r)
0.47 (r)
0.67 (r)
1.01 (r)
1.20 (r)
1.74 (r)
0.74 (d)
1.35 (d)
2.49 (d)
2.92 (d)
3.04 (d)
4 (ps)
6.20 (r)
10.69 (r)
12.31 (r)
17.42 (r)
2.53 (d)
7.27 (d)
9.08 (d)
11.51 (d)
13.60 (d)
16.64 (d)
Letters ‘d’ and ‘r’ given inside the bracket indicates that the time component is associated
respectively with the decay or rise of transient absorption signal.
174
Chapter 6
results section, this absorption band is essentially a mixture of at least two states.
The transients (Figure 6.10(c)) in this region are best fitted by three exponentials,
with first time component of sub-hundred femtosecond duration, second time
component is a rise suggested to be the growth of the intermediate state, and the
third time component is the decay of the intermediate state (or it can also be the
mixture of the decay of ‘LE’ and the intermediate state). The growth component
keeps on decreasing and eventually diminishes on the red side of the transient
absorption spectrum, wherein the LE state dominates. The lifetime components are
not consistent throughout the spectral region due to not being the signals from the
pure states, rather representing a mixture of states. At later time, a prominent ESA
band with maximum at 510 nm is observed. The kinetic traces (Figure 6.10(b))
between 510 – 540 nm are best fitted by four exponentials. The first time
Figure 6.10. Kinetic traces of DPI in methanol at selective wavelengths in the important
spectral regions. The best fits of the measured transients are shown as solid lines with the
fitting parameters shown in table 6.3.
Chapter 6
175
component of sub 100 fs is a fast decay ascribed to the decay of LE or intermediate
state or a mixture of both. Following this, there are two rise time components, the
first one of around 100 – 200 fs, and second one of around 1 – 2 ps. The fourth
time component is a decay, which increases from 9 ps to 16.6 ps from 510 nm to
530 nm. This decay being the largest decay time in the system is ascribed to the
lifetime of the final state in the excited PES. This state grows from the intermediate
state with the rise time given by the second rise component of these kinetic traces.
The authentication of the existence of three states in the excited manifold of DPI in
methanol is also suggested by the three exponential kinetic traces in the 450 – 485
nm spectral region, which represents the ground state recovery (GSR) of the
molecule. The three time constants are obtained at all the wavelengths of the
ground state bleaching spectral region (Table 6.3). On solving the kinetic model
for the rate of change of population during GSR dynamics, based on the proposed
model, (Scheme 6.3) three exponential kinetic traces are expected on behalf of
recovery from the LE, intermediate and the final state given by above equation.
d (GS )
 a1 exp( (k1  k 2 ) / t )  a2 exp( (k 3  k 4 ) / t )  a3 exp( k 5 / t )
dt
The three time constants obtained at these wavelengths are given in table 6.3 with
first time constant of few hundred femtoseconds corresponding to the recovery on
behalf of LE state, second time constant of ca. 1-2 ps to the recovery on behalf of
Scheme 6.3. Kinetic scheme for the excited state relaxation of DPI and DPPI in methanol
based on transient absorption measurements.
176
Chapter 6
intermediate state, and final recovery time constant of around 16 ps is credited to
the recovery from the final state.
Similarly, for DPPI in methanol, the transient absorption data also suggest the
existence of three states in the excited manifold. In this case, the LE state
absorption maximum is located at ca. 650 nm (Figure 6.8). This ESA signal does
not persist for long and decays very fast. The kinetic trace (Figure 6.11(d)) at 650
nm is best fitted by a three exponential function with the time constants of 1.12 ps
(d), 5.92 ps (d), and 56.0 ps (d) as tabulated in the table 6.4. In 630 – 660 nm
Figure.6.11. Kinetic traces of DPI in methanol at selective wavelengths in the important spectral
regions. The best fits of the measured transients are shown as solid lines with the fitting parameters
shown in table 6.4.
Chapter 6
177
region, we can still fit the transients with three decay time components. Based on
the time evolution of the ESA band in this region, the first time component is
ascribed to the decay of ESA of the LE state, which can be observed at all the
wavelength from 570 to 730 nm (Table 6.4 and Figure 6.11(c-e)). On the blue side
of 570 nm, in the wavelength region of 495 – 550 nm, two distinct growth
components and a large decay time component was observed. The two growth
components seem to merge together at wavelengths from 570 nm and appear as a
single growth component (second time constant at 570 – 620 nm) which vanishes
at 620 nm (Table 6.4 and Figure 6.11). Among the two growth components, the
first component is suggested to be the formation time of the intermediate state,
which decays in a timescale given by the second time component of the transients
at wavelengths from 630 – 730 nm (Table 6.4). However, one must remember that
we cannot ascribe the time scale of formation and decay of the intermediate state
exactly, because of the overlap of many signals in this region. The second growth
component in the mentioned spectral region of 495 – 550 nm is assigned to the
formation of the final state, whose ESA signal appears to have a maximum at
around 530 nm. Although, the ESA band in this spectral region appears to be
narrow initially (Figure 6.8(b)), at later times the signal broadens and appears to
encompass the ESA of the intermediate state as well. In this wavelength region of
495 – 530 nm, we would expect an initial GSB, followed by the growth, and
subsequently the decay of the ESA signal. Considering such an observation, we
found that the transients at 495 – 530 nm spectral region are best fitted by three
time components. Out of which the second component has negative amplitude
indicating the growth of a state with an average formation time of ca. 20.0 ps. This
state decays with an average time constant of ~ 60 ps. Since, in the whole spectral
region, this component is the largest decay time component, we ascribe it to the
lifetime of the final state in the excited potential energy surface.
On the red side of the transient absorption spectra from 670 – 750 nm, we
would expect the observation of decay of ESA signals and then the stimulated
emission (Figure 6.11(e)). The transients in this wavelength region are best fitted
Chapter 6
178
by three time components, out of which the first one corresponds to the lifetime of
the LE state. The second time component is ascribed to lifetime of the intermediate
state whose ESA maximum is located somewhere in the middle of the extreme red
and extreme blue ESA bands. The third time component which initially
corresponds to the decay of the ESA signal of the final state from 490 – 660 nm is
now no longer observed. Instead the presence of a negative signal infers the
occurrence of a stimulated emission, whose lifetime is ~ 50 ps. (Table 6.4) The
Table 6.4. Lifetime of three components obtained by fitting the temporal profiles of the
transient absorption data of DPPI in methanol.
Wavelength (nm)
465
470
475
480
485
490
495
500
505
510
515
520
530
540
550
570
580
590
600
610
615
620
630
640
645
650
655
660
670
675
680
690
700
710
720
730
1
(ps)
0.13 (r)
0.13 (r)
0.14 (r)
0.06 (r)
0.06 (r)
0.05 (r)
0.33 (r)
0.39 (r)
0.43 (r)
0.46 (r)
0.55 (r)
0.59 (r)
0.76 (r)
1.01 (r)
1.57 (r)
0.07 (d)
0.27 (d)
0.41 (d)
0.53 (d)
0.72 (d)
0.77 (d)
0.96 (d)
0.15 (d)
0.89 (d)
0.96 (d)
1.12 (d)
1.37 (d)
1.57 (d)
1.89 (d)
2.04 (d)
1.75 (d)
1.87 (d)
1.85 (d)
1.64 (d)
1.37 (d)
1.14 (d)
2
(ps)
2.41 (r)
4.73 (r)
5.46 (r)
5.39 (r)
3.50 (r)
0.64 (r)
9.20 (r)
9.61 (r)
7.65 (r)
6.73 (r)
7.64 (r)
6.06 (r)
8.72 (r)
11.80 (r)
15.70 (r)
8.02 (r)
9.56 (r)
10.70 (r)
12.10 (r)
14.20 (r)
21.90 (r)
25.00 (r)
1.50 (d)
4.34 (d)
5.33 (d)
5.92 (d)
6.16 (d)
6.84 (d)
7.60 (d)
7.83 (d)
6.65 (d)
6.10 (d)
8.00 (d)
7.33 (d)
7.03 (d)
7.37 (d)
3
(ps)
56 (r)
57 (r)
58 (r)
52 (r)
37 (r)
9 (r)
74 (d)
64 (d)
69 (d)
66 (d)
64 (d)
66 (d)
61 (d)
59 (d)
54 (d)
65 (d)
62 (d)
60 (d)
58 (d)
59 (d)
49 (d)
46 (d)
64 (d)
57 (d)
56 (d)
56 (d)
51 (d)
42 (d)
22 (r)
49 (r)
54 (r)
66 (r)
48 (r)
45 (r)
64 (r)
58 (r)
Letters ‘d’ and ‘r’ given inside the bracket indicates that the time component is associated
with the decay or rise of transient absorption signal respectively.
Chapter 6
179
negative signal in this spectral region is essentially the tail of the steady state
emission spectra and hence the lifetime obtained here will be comparatively higher
than the lifetime obtained at emission maximum in the fluorescence up-conversion
measurements. The transients in the GSB region of 470 – 480 nm (Figure 6.11(a))
are best fitted by three time components (Table 6.4). The first time constant of
about 100 fs is because of the ground state recovery from the locally excited state
and occurs very fast. The second and third time components correspond to the
ground state recovery (GSR) of the population from intermediate state and the final
state respectively. The GSR from the final state occurs with a time constant of
approx. 58 ps which is almost same as that of the decay of final state.
In general, we can ascribe the observed transient absorption data of DPPI in
methanol to the existence of three states in the excited manifold, similar to the DPI.
In this case the time constants are different because of the presence of bulkier
substituents.
6.3.3. Global Analysis of the Transient Absorption Data
Global analysis of the time resolved data was pursued using sequential kinetic
scheme, wherein the back reactions are ignored based on the assumption that the
energy loses are large enough so that reverse reaction rates are negligible.50 Such a
global analysis produces the evolution associated difference spectra (EADS) as
shown in figure 6.12(a) and (c) from which decay associated difference spectra
(DADS) (Figure 6.12(b) and (d)) are evaluated. While as EADS represent the
spectral evolution with time, e.g. the second EADS rises with the first lifetime and
decays with the second lifetime, DADS are interpreted as loss or gain of absorption
with a certain lifetime component. If the sequential kinetic scheme represents the
true physicochemical picture, the EADS correspond to true species associated
difference spectra (SADS) characterizing the intermediate states. In most of the
cases, the EADS and DADS represent a mixture of states with one state
dominating the other.50-55 For both the molecules (DPI and DPPI) the minimum
number of components necessary to fit the data for global analysis is four. Out of
180
Chapter 6
Figure 6.12. The top panel (A and C) of the figure represents the Evolution Associated
Difference Spectra (EADS) of DPI and DPPI and the bottom panel (B and D) shows the
Decay Associated Difference Spectra (DADS) of DPI and DPPI respectively. Corresponding
time of evolution or decay of the spectra are shown in the form of annotation with each
graph. In practice the global fitting produced four EAS and DAS, out of which the first one
was found to be due to coherent artifact and hence for the sake of brevity is not shown here.
which the first EADS and DADS was found to resemble the coherent artifact
spectra and hence for the sake of brevity is not shown in the figure. Figure 6.12(a)
and (c) shows the three EADS and DADS of DPI. The first DADS (Red) with a
broad excited state absorption depict a loss of intensity of this band on the red side,
and loss of stimulated emission (maximum at ca. 500 nm). This DADS decays in
0.5 ps timescale. The second DADS (Blue) differs much from the first DADS and
has an ESA band at ca. 620 nm, the SE band is now at ca. 540 nm with less
intensity and a second ESA band also appears on the blue side of the spectrum.
The DADS is also having a prominent GSB in 450 – 500 nm region. This DADS
grows with time constant of 0.5 ps and decays in 3.50 ps. The third DADS (Green)
has a prominent ESA band at ca. 510 nm and the presence of GSB band is also
observed. The SE or ESA around 600 nm is not observed in this DADS. The third
Chapter 6
181
DADS decays with 15.6 ps timescale. The first EADS is characterized by a large
GSB in 450-500 nm range. While as the subsequent EADS show a gradual
decrease of this band. EADS also shows a gradual decrease in the ESA around 650
nm, however the ESA do not belong to the same state, rather due to the continuous
shifting of this band to the blue side enables us to confirm the existence of
intermediate state. The second EADS evolves into the third EADS, which is
characterized by the ESA of the final state. The second DADS or second EADS
need not to represent a pure intermediate state, it may be rather a mixture state of
the first and third EADS and DADS. The important processes and spectral regions
which can thus be envisaged from the global analysis are GSB (450-500 nm), SE
(500-550), ESA band 1 (550-775 nm), ESA band 2 (550-775 nm) and ESA band 3
(480-550 nm).
The EADS and DADS of DPPI are shown in figure 6.12(c) and (d)
respectively, wherein these spectra show similar time evolution and spectral
characteristics as that of DPI. The first DADS (Red) decays with a time constant of
1.1 ps having a broad ESA band on the red side and a SE at ca. 575 nm. The
second DADS (Blue) has an ESA band at around 670 nm and a strong SE band at
575 nm. The ESA band which appears broad in first DADS is no longer broad in
the second DADS probably because of the presence of SE band. This DADS
decays in 6.9 ps. The third DADS (Green) has a broad ESA band, and also a SE
band, which decays with a lifetime of 67.0 ps. The broadness of the ESA band is
probably because it represents a mixture of the intermediate and the final state. The
EADS (Figure 6.12(c)) shows a continuous decrease of ground state bleaching
signal. In the second EADS, one could observe two sharp ESA bands at 660 nm
and 525 nm with a dip in the spectral region between them. However, in the third
EADS, due to apparent decrease in the intensity of the stimulated emission, the
ESA band now appears broad and represents a mixture of the intermediate and the
final states.
From the time dependent absorption spectra, wavelength dependent kinetic
traces, and global analysis of the transient absorption measurements of DPI and
182
Chapter 6
DPPI in methanol, we propose the existence of three states in the excited manifold
of these molecules. The three states are; locally excited state (LE), intermediate
state and the final state.
In summary, the proposition of the presence of a CT state in the excited state
PES of DPI and DPPI through the fluorescence up-conversion measurements is
confirmed by the existence of the intermediate state through transient absorption
measurements. Thus we assign the intermediate state to the charge transfer state for
these molecules. The depletion of the CT state to a point of conical intersection,
leading to trans-cis isomerization is also confirmed by the finite lifetime of final
state from transient absorption measurements. This final state is thus assigned to
the global minimum in the excited state PES, where the internal conversion occurs
with the ground state through conical intersection. This state being dark in nature,
fluorescence up-conversion study was unable to report its features. Transient
absorption measurements however, could identify the formation and decay of this
state. The ground state recovery dynamics is also in accordance with the
proposition of the existence of three states in the excited manifold of the two
analogs under investigation. The involvement of twisting motion of the exocyclic
double bond (facilitated by charge transfer) in the excited state is credited to be
responsible for the non-fluorescent nature of the bare chromophores. In analogy, if
such twisting motion could be restricted, the fluorescence quantum yield must
increase. In case of native GFP, the chromophore being present in a protein
scaffold, its torsional degrees of freedom are restricted and hence making it highly
fluorescent by abandoning the non-radiative pathways.
Chapter 6
183
6.5. Conclusion
The current study of the ultrafast excited state relaxation dynamics of two
HBDI analogs, DPI and DPPI is devoted to understand the restraining role of
immediate amino acid environment around the p-HBDI chromophore in GFP,
which renders it highly fluorescent by abandoning the non-radiative pathways. The
involvement of torsional coordinate about the donor group attributes twisted nature
to the charge transfer state, which evolves from Franck-Condon state on a time
scale of few picoseconds. The charge transfer state depletes to the region of conical
intersection between S1 and S0 state along the torsional motion about exocyclic
double bond. Femtosecond fluorescence up-conversion results and assistance from
time dependent density functional theory calculations helped us in ruling out the
volume conserving simultaneous rotation of bridging bonds “Hula twist”, and
rotation about single bond. While as, an activated rotation about the exocyclic
double bond confirmed by quantum chemical calculations, leads to relaxation of
excited state into ground state with two geometrical isomers. The spectral and
temporal evolution of the transient absorption signals along with the global
analysis confirms the existence of three states in the S1 potential energy surface
sequentially as locally excited, charge transfer, and the conical intersection
position.
Chapter 6
184
References
1. Ormö, M.; Cubitt, A. B.; Kallio, K.; Gross, L. A.; Tsien, R. Y.; Remington, S.
J. Science 1996, 273, 1392.
2. Yang, F.; Moss, L.; Phillips, G. Nat. Biotech. 1996, 14, 1246.
3. Heim, R.; Prasher, D. C.; Tsien, R. Y. Proc. Natl. Acad. Sci. USA 1994, 91,
12501.
4. Cubitt, B.; Heim, R.; Adams, S. R.; Boyd, A. E.; Gross, L. A.; Tsien, R. Y.
Trends Biochem. Sci. 1995, 20, 448.
5. Reid, B. G.; Flynn, G. C. Biochemistry 1997, 36, 6786.
6. Ehrmann, M. A.; Scheyhing, C. H.; Vogel, R. F. Lett. Appl. Microbiol. 2001,
32, 230.
7. Chalfie, M.; Tu, Y.; Euskirchen, G.; Ward, W. W.; Prasher, D. C. Science
1994, 263, 802.
8. Zimmer, M. Chem. Rev. 2002, 102, 759.
9. Bottin, A.; Larche, L.; Villalba, F.; Gaulin, E.; Esquerré-Tugayé, M. T.;
Rickauer, M. FEMS Microbiol. Letts 1999, 176, 51.
10. (a) Heim, R.; Tsien, R. Y. Curr. Biol. 1996, 6, 178. (b) Schüttrigkeit, T. A.;
Zachariae, U.; Feilitzsch, T. V.; Wiehler, J.; Hummel, J. V.; Steipe, B.; MichelBeyerle, M. E. ChemPhysChem 2001, 2, 325.
11. Pollok, B. A.; Heim. R. Trends Cell Biol. 1999, 9, 57.
12. Pozzan, T. Nature 1997, 388, 834.
13. Ward, W. W.; Cody, C. W.; Hart, R. C.; Cormier, M. J. Photochem. Photobiol.
1980, 31, 611.
14. Bokman, S. H.; Ward, W. W. Biochem. Biophys. Res. Comm. 1981, 101, 1372.
15. Niwa, H.; Inouye, S.; Hirano, T.; Matsuno, T.; Kojima, S.; Kubota, M.;
Ohashi, M.; Tsuji, F. I. Proc. Natl. Acad. Sci. USA 1996, 93, 13617.
16. Follenius-Wund, A.; Bourotte, M.; Schmitt, M.; Lyice, F.; Lami, H.;
Bourguignon, J. J.; Haiech, J.; Pigault, C. Biophys. J. 2003, 85, 1839.
Chapter 6
185
17. Rajbongshi, B. K. Photophysical, Crystallographic and Photovoltaic Studies
on Imidazolin-5-ones. Ph. D thesis, Indian Institute of Technology Kanpur,
2011.
18. Kummer, A. D.; Kompa, C.; Lossau, H.; Pöllinger-Dammer, F.; MichelBeyerle, M. E.; Silva, C. M.; Bylina, E.; Coleman, W. J.; Yang, M. M.;
Youvan, D. C. Chem. Phys. 1998, 237, 183.
19. Kummer, A. D.; Wiehler, J.; Rehaber, H.; Kompa, C.; Steipe, B.; MichelBeyerle, M. E. J. Phys. Chem. B 2000, 104, 4791.
20. Kummer, A. D.; Wiehler, J.; Shüttrigkeit, T. A.; Berger, B. W.; Steipe, B.;
Michel-Beyerle, M. E. Chembiochem 2002, 3, 659.
21. Webber, N. M.; Litvinenko, K. L.; Meech, S. R. J. Phys. Chem. B 2001, 105,
8036.
22. Litvinenko, K. L.; Webber, N. M.; Meech, S. R. J. Phys. Chem. A 2003, 107,
2616.
23. Duxbury, D. F. Chem. Rev. 1993, 93, 381.
24. Ben-Amotz, D.; Harris. C. B. J. Chem. Phys. 1987, 86, 4856.
25. Rafiq, S.; Yadav, R.; Sen, P. J. Phys. Chem. B, 2010, 114, 13988.
26. Wu, L.; Burgess, K. J. Am. Chem. Soc. 2008, 130, 4089.
27. Martin, M. E.; Negri, F.; Olivucci, M. J. Am. Chem. Soc. 2004, 126, 5452.
28. Rajbongshi, B. K.; Sen, P.; Ramanathan, G. Chem. Phys. Lett. 2010, 494, 295.
29. Stavrov, S. S.; Solntsev, K. M.; Tolbert, L. M.; Huppert, D. J. Am. Chem. Soc.
2006, 128, 1540.
30. Gepshtein, R.; Huppert, D.; Agmon, N. J. Phys. Chem. B 2006, 110, 4434.
31. Voityuk, A. A.; Michel-Beyerle, M. E.; Rösch, N. Chem. Phys. Lett. 1998,
296, 269.
32. Solntsev, K. M.; Poizat, O.; Dong, J.; Rehault, J.; Lou, Y.; Burda, C.; Tolbert,
L. M. J. Phys. Chem. B 2008, 112, 2700.
33. Litvinenko, K. L.; Webber, N. M.; Meech, S. R. Chem. Phys. Lett. 2001, 346,
47.
Chapter 6
186
34. Megley, C. M.; Dickson, L. A.; Maddalo, S. L.; Chandler, G. J.; Zimmer, M. J.
Phys. Chem. B 2009, 113, 302.
35. Mandal, D.; Tahara, T.; Webber, N. M.; Meech, S. R. Chem. Phys. Lett. 2002,
358, 495.
36. Mandal, D.; Tahara, T.; Meech, S. R. J. Phys. Chem. B 2004, 108, 1102.
37. Rafiq, S.; Rajbongshi, B.K.; Nair, N. N.; Sen, P.; Ramanathan, G. J. Phys.
Chem. A 2011, 115, 13733.
38. Dong, J.; Solntsev, K. M.; Poizat, O.; Tolbert, L. M. J. Am. Chem. Soc. 2007,
129, 10084.
39. Petkova, I.; Dobrikov, G.; Banerji, N.; Duvanel, G.; Perez, R.; Dimitrov, N.;
Nikolov, P.; Vauthey, E. J. Phys. Chem. A 2010, 114, 10.
40. Bhattacharjya, G. From Molecules to Materials: Design, Synthesis,
Characterization and Possible Applications of Imidazolin-5-ones. PhD thesis,
2006, Indian Institute of Technology Kanpur.
41. Rajbongshi, B. K.; Ramanathan, G. J. Chem. Sci. 2009, 121, 973.
42. Bhattacharjya, G.; Agasti, S. S.; Ramanathan, G. ARKIVOC 2006, 10, 152.
43. Sen, P.; Roy, D.; Mondal, S. K.; Sahu, K.; Ghosh, S.; Bhattacharyya, K. J.
Phys. Chem. A 2005, 109, 9716.
44. Frisch, M. J. et al. Gaussian 03, rev. C.02; Gaussian, Inc.: Wallingford, CT,
2004.
45. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
46. Jamorski, C.; Casida, M. E.; Salahub, D. R. J. Chem. Phys. 1996, 104, 5134.
47. Bauernschmitt, R.; Ahlrichs, R.; Hennrich, F. H.; Kappes, M. M. J. Am. Chem.
Soc. 1998, 120, 5052.
48. (a) Autschbach, J. ChemPhysChem 2009, 10, 1757. (b) Niehaus, T. A.; March,
N. H. Theor. Chem. Acc. 2010, 125, 427. (c) List, N. H.; Olsen, J. M.; RochaRinza, T.; Christiansen, O.; Kongsted, J. Int. J. Quantum Chem. 2011, 112,
789.
Chapter 6
187
49. (a) Sherin, P. S.; Grilj, J.; Tsentalovich, Y. P.; Vauthey, E. J. Phys. Chem. B
2009, 113, 4953. (b) Furstenberg, A.; Vauthey, E. Photochem. Photobiol. Sci.
2005, 4, 260.
50. Snellenburg, J. J.; Laptenok, S. P.; Seger, R.; Mullen, K. M.; van Stokkum, I.
H. M. J. Stat. Soft. 2012, 49, 1.
51. Kovalenko , S. A.; Dobryakov, A. L.; Ruthmann, J.; Ernsting, N. P. Phys. Rev.
A 1999, 59, 2369.
52. Novoderezhkin, V. I.; Palacios, M. A.; Van Amarongen, H.; Van Grondelle, R.
J. Phys. Chem. B 2004, 108, 10363.
53. Van Stokkum, I. H. M.; Larsen, D. S.; van Grondelle, R. Biochim. Biophys.
Acta 2004, 1657, 82.
54. Berera, R.; van Grondelle, R.; Kennis, J. T. M. Photosynth. Res. 2009, 101,
105.
55. Ruckebusch, C.; Sliwa, M.; Pernot, P.; de Juan, A.; Tauler, R. J. Photochem.
Photobiol. C: Photochem. Rev. 2012, 13, 1.
188
Chapter 6
Chapter 7
Viscosity of Water in a Nano-confined
Environment through the Ultrafast
Excited State Relaxation
Dynamics of Malachite Green
Shahnawaz Rafiq et al., J. Phys. Chem. B 2010, 114, 13988.
Chapter 7
190
This
chapter
reports
the
femtosecond
fluorescence
up-conversion
measurements of malachite green (MG), carried out to confirm the excited state
relaxation mechanism and subsequently to probe the microviscosity of water
trapped in a nano-confined environment using AOT reverse-micelle as a model
system. Experimental results reveal a strong dependence of S1 state relaxation
dynamics of MG on solvent viscosity while a very weak dependence has been
observed for the S2 state relaxation. The time dependent density functional theory
(TDDFT) calculations have been used to construct potential energy surfaces (PES)
of malachite green by pursuing an intramolecular rotation along torsional
coordinate of the phenyl rings. In augmentation with the experimental
observations, the computational results comprehend the existence of a conical
intersection along the S1 and S0 potential surfaces, which leads to mixed
vibrational levels of S1 and S0 characteristics. The results suggest that the conical
intersection is along the torsional coordinate of N,N-dimethylamino substituted
phenyl ring. Correlating the observed dynamics of MG in confined system with the
relaxation time of MG in different glycerol-water mixtures, we assert the
determination of microviscosity of water inside the AOT reversed micellar system.
The data confers that the microviscosity of water in AOT water pool of w 0 = 2 (9
cP) is almost 9 times higher than the bulk water. As we increase the w 0 from 2 to
40, the microviscosity decreases monotonically to 5.68 cP and the decrease is
observed to be exponential.
Chapter 7
191
7.1. Introduction
The ground and excited state dynamics of triphenylmethane (TPM) dyes in
solution is a sensitive probe of the influence of environment on intramolecular
motion along a reactive potential energy surface.1-5 Number of studies have been
carried out to understand the molecular structure, electronic states and excited state
relaxation dynamics of TPM dyes.1-5 The dynamics of first singlet excited state (S1)
of TPM dyes is observed to show a strong viscosity dependence ascribed to
strong coupling between electronic states and the torsional degrees of freedom. 1-5
The nature of molecular electronic states of TPM dyes bears a strong correlation
with the angle between the central sp2 hybridized carbon and the plane of phenyl
rings. The fluorescence lifetime, quantum yield and the ground state recovery
(GSR) time of TPM dyes have undergone a great deal of study. In general, these
quantities have been found to be proportional to the viscosity of medium ( ) raised
to a power ranging roughly between 0.33 and 0.66.2,4,5 The model for the
radiationless decay of the excited singlet state to the ground state is believed to
involve rotation of the phenyl rings towards an equilibrium geometry displaced
from that in the ground state. It is proposed that the internal conversion rate is
much higher near the equilibrium (through a conical intersection) geometry of the
excited state and the viscosity dependence of the GSR arises simply from the
solvent frictional effect in resisting the phenyl ring rotation. The initial ground
state is subsequently repopulated by reverse torsional motion from high up on the
ground state surface.1-5 Because of the friction dependence of the excited state
lifetime, TPM dyes can be used as a local viscosity probe. The effect of solvent
polarity is not found to be significant on the excited state dynamics in solvents of
similar viscosities but with different polarity.6 The weak solvent polarity
dependence suggests that the potential energy surfaces are to be similar in the
different environments. One of such TPM dye is malachite green (MG) and has
been as well studied to a great extent. Contrary to the Kasha’s rule, MG has been
found to show fluorescence emission from S2 state as well, imparting its
significance in excited state processes.4a,5b Various efforts have been made to
192
Chapter 7
characterize the relaxation dynamics of S1 and S2 states of MG. Using femtosecond
fluorescence spectroscopy, Mokhtari et al. suggested the presence of a hot ground
state in the relaxation pathway of the S1 state of MG.5a Several other transient
absorption studies also reveal the same conclusion for the excited state relaxation
pathway of MG.1-4 All studies emphasized the presence of a conical intersection
between the S1 and S0 states of MG. Yoshizawa et al. studied the S2 fluorescence
of MG by fluorescence up-conversion spectroscopy in different solvents.5b
According to them, the S2 state relaxation follows a single exponential decay
kinetics in low viscous solvents. The origin of the decay component has been
explained by both the torsional configuration change and the internal conversion
from S2 state to the S1 state.5b In another work, Bhasikuttan et al.5c on the basis of
kinetic data along with the time resolved fluorescence anisotropy measurements
proposed a relaxation pathway along an interaction of S2 and S1 potential energy
surfaces forming a conical intersection. It has also been proposed that the conical
intersection is along the torsional coordinate of unsubstituted phenyl ring of the
MG dye.5c
The GSR dynamics of MG have already been measured in several systems to
understand the microviscosity of the medium. Canva et al. studied the viscosity of
xerogel matrix by probing the GSR dynamics of MG and found that the viscosity is
about 25 cP in the matrix.7a Nakatsuka et al. measured the fluorescent decay time
of MG doped in an onion cell and try to correlate with the microscopic dynamics in
different part of the cell.7b Recently, Nagasawa et al. utilize MG to probe
microscopic viscosity of water-alcohol binary mixture.7c There has also been a lot
of work regarding the relaxation dynamics of MG at the different interfaces.8
Till date a lot of work has been devoted to understand the dynamics of MG in
bulk solvents and at the bulk interface. However, only a little concern has been
paid to explore its dynamics in biological environment. For a first choice, one can
study the dynamics of MG in reverse-micelle which can be an elegant model to
mimic the biological systems.9 Among all the reverse-micellar systems, AOT
(sodium dioctylsulfosuccinate, Aerosol-OT) has been studied extensively using
Chapter 7
193
many techniques and it is reported that water trapped in the nano-cavity behaves
quite differently than the bulk water.10 One of the important question is, what is the
nature of water in the water pool inside the reverse micelle? In the simplest model,
many authors have postulated the presence of two components of water in the
reverse micelles, one is the interfacial (shell) water, which exhibits distinct
properties than bulk water, and the other is interior (core) water, which behaves
similarly to bulk water.11-16 Several groups have employed this concept to elucidate
the structure and dynamics of solvents inside the reverse micelles. Levinger and
Fayer have used a combination of steady state, vibrational lifetime, orientational
relaxation and vibrational echo spectroscopies to study the OD stretch of diluted
HOD in H2O in AOT reverse-micelles based on the core-shell model.15,17,18 They
concluded that confinement by an interface to form an nanoscopic water pool is a
primary factor governing the slow dynamics of nanoscopic water rather than the
presence of charged groups at the interface. Correa’s group has studied the new
insights into the photophysics of fluorescent molecules like Coumarin 343 and
prodan.19-21 Recently they have used a cationic reverse micelle to create water with
super Hydrogen-Bond-Donor capacity for enzymatic catalysis.22 Amararene et al.
reported that, even in a large water pool (w0 = 27), the compressibility is 2 times
larger than that in bulk water.24a A 25% shift in liberation frequency at 670 cm-1 is
reported in AOT reverse-micelle compared to that in bulk water.16 The terahertz
(THz) absorption spectrum of confined water molecules is in the region of 0.1−1.3
THz which is very different from that of bulk water, like in a water pool size of 1.5
nm the terahertz frequency shifts from 25 cm-1 to about less than 10 cm-1 in 2.9 nm
water pool.24b
The size of the water pool in reverse-micelles is expressed in terms of a ratio of
number of water molecules to the number of surfactant molecules and is denoted
by w0. With an increase in the magnitude of w0, the radius of the water pool (rw)
increases. On the basis of NMR data Maitra10a formulated an expression for the
radius of water core (rw) inside the reverse micelle as
Chapter 7
194
(7.1)
where, n is the mean average aggregation number of surfactant per droplet.
Kinugasa et al. estimated the size of AOT reverse micelle on the basis of viscosity
measurements based on the following expression.23
(7.2)
where dwp, Vw, Cw, Cs, nag represent diameter of water pool, volume of water,
concentration of water, concentration of surfactant and aggregation number
respectively. For w0 between 2 and 20, one can derive the following simplified
equation for the diameter of water pool of AOT reverse-micelle
(7.3)
In the present study, we suggest the nature of potential energy profile of S0, S1
and S2 electronic states of MG and also the probable involvement of the three
different phenyl groups on the relaxation dynamics by using the femtosecond
fluorescence up-conversion measurements in different glycerol-water mixed
solvents and quantum mechanical calculations in vacuum. As discussed by
different authors,1-5 the rotation of phenyl rings in the excited state is the major
deactivation process. However, there was no such attempt to investigate which of
the phenyl ring(s) is responsible for such rapid deactivation of the excited state.
From the quantum mechanical calculations we also estimate the contribution of
different phenyl rings in the
deactivation
observed
process.
viscosity
The
dependent
relaxation dynamics has been
further exploited to determine the
microviscosity of water trapped in
a nano-confined environment with
different dimensions using AOT
reverse-micelle as a model system.
Scheme 7.1. Structure of Malachite Green.
Chapter 7
195
7.2. Computational Method
Ground state optimization of malachite green has been performed by using
density functional theory (DFT) calculations using B3LYP hybrid functional with
6-31+g(d,p) basis set. A rigorous formalism (time dependent-DFT) has been used
to calculate the transition energies within the DFT framework using the same
hybrid function and the same basis set. All calculations were performed using
Gaussian 03 software. We have considered the optimized structure obtained in
vacuum to hold good for excited state calculations assuming not much change in
the corresponding geometries. Transition energies obtained from TD-DFT
calculations are the vertical excitation energies obtained without zero point
corrections.
Malachite green contains two N,N-dimethyl substituted phenyl rings and
another un-substituted phenyl ring (Scheme 7.1). To determine the potential energy
profile of ground, first and second excited states of MG, we determine the energies
of different geometries by rotating the dihedral angles of different phenyl rings.
For the estimation of the effect of different phenyl rings on the relaxation process,
we have selectively rotated the un-substituted phenyl ring (C24-C23-C1-C13
dihedral angle) and substituted phenyl ring (C15-C13-C1-C23 dihedral angle)
linked to central sp2 hybridized carbon atom and compared it with the rotation of
all the phenyl rings at different extent from its equilibrium position.
7.3. Results and Discussion
7.3.1. Steady State Absorption and Emission Measurements in GlycerolWater Mixtures
The absorption spectra of MG in water show two peaks, one at 425 nm
corresponding to S2  S0 absorption and other at 617 nm corresponding to S 1 
S0 absorption. On increasing the proportion of glycerol from 0% to 70% glycerol
in water, there is a little red shift in the absorption maxima from 425 nm to 430 nm
for S2  S0 absorption band and from 617 nm to 626 nm for S1  S0 absorption
band as shown in figure 7.1(a). Up on excitation at 410 nm, the emission spectrum
196
Chapter 7
Figure 7.1. (a) Steady state absorption and (b) emission spectra of malachite green in bulk
water (red), 30% glycerol (green) and 70% glycerol (blue) in water.
of MG shows two fluorescence bands. The band in the region of 430 to 600 nm is
attributed to the S2 fluorescence. In this case, the Raman scattering from the
solvent is predominant over the fluorescence as the fluorescence quantum yield is
very low. The fluorescence in the 620 to 750 nm region clearly shows a maximum
at 667 nm in bulk water and has been assigned to the S 1 fluorescence. On
increasing the proportion from 0% to 70% glycerol in water, the S 1 emission
maximum is found to be at 665 nm which is 2 nm blue shifted from the MG in
water. The intensity of emission for both the S1 and S2 fluorescence show strong
viscosity dependence (Figure 7.1(b)). The increase in fluorescence intensity is not
similar in S1 and S2, rather it is observed that S1 fluorescence intensity increases
more rapidly compared to S2 fluorescence intensity. At 520 nm (S2 fluorescence
region) the fluorescence intensity increases 3.1 times in 70% glycerol-water
mixture compared to bulk water whereas, at 665 nm (S1 fluorescence region) there
is a 5.1 times increase in the fluorescence intensity. The origin of the viscosity
dependent emission intensity can be interpreted as a consequence of change of
non-radiative decay rate constant in the molecule as already suggested by several
authors.1-5 From the emission spectra it is also observed that the spectral widths
and peak energies do not change much with increase in viscosity which interprets
that the electronic states in MG are meagerly affected by the solvent viscosity.
Chapter 7
197
7.3.2. Femtosecond Fluorescence Up-conversion Study
Femtosecond resolved fluorescence lifetime measurements25 have been carried
out in different glycerol-water mixtures of varying viscosity in order to understand
the relaxation pathways of S1 and S2 electronic states of MG. The fluorescence
lifetime of S1 and S2 states were obtained at a constant laser excitation source of
410 nm. The viscosity dependent fluorescence transients of S 1 and S2 states are
shown in figure 7.2. In water the fluorescence transient of S 1 state shows biexponential behavior and can be fitted with two time constants of S1, rise = 100  50
fs and S1, decay = 660  50 fs (Figure 7.2(a).). Transient absorption measurements
of MG in water reveals the presence of three time constants 270 fs,
Figure 7.2. Fluorescence transients of malachite green (a) measured at 670 nm (S1 state) in
bulk water (red), 30% glycerol (green) and 70% glycerol (blue) in water (b) measured at 500
nm (S2 state) in bulk water (red), 30% glycerol (green) and 70% glycerol (blue) in water. The
sample was excited at 410 nm. The solid line represents the best fit lines.
630 fs and 3000 fs which have been assigned to equilibration of high-frequency
internal modes in the S1 state, the relaxation to the intermediate state Sx, and the
relaxation from Sx state to the ground state, respectively.3c Similar to the present
work, Yoshizawa et al. got a bi-exponential fluorescence transient for MG in water
with a rise component of 430 fs followed by a decay component of 540 fs. 5b
Mokhtari et al. also got a bi-exponential fluorescence transient for MG in water
(ex = 625 nm) with two time constants of 150 fs and 600 fs respectively. 5a The
observation of only two time constants in the fluorescence transient data for S 1
198
Chapter 7
state relaxation is because the intermediate state Sx is non-fluorescent. The longer
component has been interpreted in terms of the intramolecular diffusion towards
the optimum configuration,5a which essentially corresponds to the relaxation from
S1 state to intermediate state Sx. With increase in viscosity of the solvent to 70%
glycerol-in-water, the rise component of the S1 fluorescence changes from 100 fs
to 190 fs and is thus almost independent of viscosity, while as the decay
component shows a strong dependence on viscosity. In 70% glycerol-water
mixture the S1, decay increases from 660  50 fs (in bulk water) to 7300  100 fs.
The respective relaxation kinetic parameters of MG in different glycerol-water
mixtures are tabulated in table 7.1. The increase in decay time of S 1 state (S1, decay)
is attributed to the larger dependence of torsional configuration change on viscosity
i.e., the restricted rotation of phenyl rings of MG. 4,5 The fluorescence
Table 7.1. Kinetic parameters of relaxation dynamics of malachite green in different
glycerol-water mixtures.
% of
 (cP)
S1,rise (fs)a
S1,decay (fs)b
S2 (fs)b
Glycerol
0
1.005
100
660§
260
§
10
1.31
110
840
280
30
2.50
120
1420
350
50
6.00
140
2900
410
60
10.80
150
4300
480
67
17.70
160
6060
500
70
22.50
190
7620
520
50fs. b100fs
a
transients of the S2 state (monitored at 500 nm) were found to be single
exponential (S2) in nature (Figure 7.2(b)). In pure water the observed decay time is
260  50 fs. On increasing the solvent viscosity to 70% glycerol in water, there has
been a little increase in the lifetime of S2 state from 260  50 to 510  50 fs.
Yoshizawa et al. also observed similar decay behavior for MG in water (270 fs).5b
This may be assigned to both the torsional configuration change and the internal
conversion from S2 to S1 state. Figure 7.3 shows the dependence of S1 rise, S1
decay and S2 decay time of MG on solvent viscosity deduced from the fitting
Chapter 7
199
Figure 7.3. Dependence of the time constants S1, decay (●), S2 (▲) and S1, rise (♦) of MG on
viscosity () of different glycerol-water mixtures.
parameters. It can be readily seen that the rise time of the S 1 state (S1,rise) and
decay time of the S2 state (S2) have very meager dependence on the solvent
viscosity. S1,
rise
and S2 components both show a very weak power law
dependence on viscosity as 0.19 and 0.21 respectively. While as S1 decay (S1, decay)
shows a much larger dependence on viscosity with the decay time increasing with
increase in viscosity and follows an 0.75 dependence. From these lifetime
measurements it infers that only the S1 excited state decay time is being affected by
the viscosity of solvent while as S2 decay time constant seems to be inert with
respect to changing viscosity.
For another triphenylmethane dye with similarly substituted phenyl rings
(crystal violet) Ben-Amotz et al. found the barrierless torsional motion of the
phenyl rings are responsible for the rapid relaxation of excited state. They reported
the effect of temperature on the relaxation dynamics of crystal violet and found
that for iso-viscous solvents, at different temperatures there is no change in the
relaxation time which rejects the presence of any activation energy barrier in the S 1
potential energy surface.2b As MG is as well a similar type of dye, one can assume
that the excited state relaxation of MG is also barrierless in nature. Since MG has
200
Chapter 7
two different kinds of phenyl rings, it is important to have a deeper understanding
of the effect of internal rotation of different types of phenyl rings on the relaxation
process of MG. Bhasikuttan et al. proposed a relaxation pathway along an
interaction of S2 and S1 potential energy surfaces forming a conical intersection
and also proposed that the conical intersection is along the torsional coordinate of
un-substituted phenyl ring of MG.5c Here, we tried to investigate the role of
different phenyl rings on relaxation process and the nature of potential surfaces of
different energy states by quantum mechanical calculations.
7.3.3. Quantum Chemical Calculations
We tried to explore the relaxation pathways and the role of different phenyl
rings of MG through time dependent density functional theory (TDDFT)
calculations to reveal the mechanistic details of the viscosity dependence of
relaxation time. The ground state optimization of malachite green in vacuum has
been performed through density functional theory (DFT) calculations using
B3LYP hybrid function with 6-31+g(d,p) basis set. In the ground state optimized
structure of MG, the dihedral angle between the un-substituted phenyl ring and the
central carbon atom (C24-C23-C1-C13) is ~43º (Scheme 7.1) and the angle
between two N,N-dimethylamino substituted phenyl rings with the central carbon
atom (i.e., C15-C13-C1-C23 and C3-C2-C1-C23) is ~28.5º. The ground state
dipole moment of MG in vacuum is found to be 3.28 D. A rigorous formalism
(time dependent-DFT) is employed to calculate the transition energies within the
DFT framework using the same hybrid functional and the same basis set. The
calculated transition energies of the S1  S0 and S2  S0 transition are found to be
~489 nm and ~403 nm with very high oscillator strength of 0.86 and 0.34
respectively. These calculated values of transition energies are over estimated as
compared to the experimental results. However, the calculated ratio of S1  S0 and
S2  S0 oscillator strength is 2.53 which are in quite good agreement with the
experimental value of ~2.7. To monitor the shape of the potential energy surfaces
of MG we first rotate all the three phenyl rings in clockwise and anticlockwise
Chapter 7
201
directions from the geometrically optimized position. The data obtained from these
calculations when plotted in the form of potential energy versus torsional
coordinate (dihedral angle) depict the similar nature of potential for both the S1 and
S2 states with a little potential barrier around 45º as shown in figure 7.4(a), which
narrates the similar dependence of excited state decay behavior of both S 1 and S2
states on the viscosity of the solvent. However the present experimental
observations do not favor this type of relaxation dependence on viscosity. Similar
kind of potential energy plots for S0, S1 and S2 states are obtained on rotating only
the un-substituted phenyl ring as shown in figure 7.4(b). Upon rotation of one
among the two N,N-dimethylamino substituted phenyl rings, the potential energy
Figure 7.4. Potential energy surfaces (PES) of malachite green in vacuum for the ground and
first two excited states along the torsional coordinate of (a) all the three phenyl rings (b) unsubstituted phenyl ring and (c) N,N-dimethylamine substituted phenyl ring.
Chapter 7
202
curves we obtained are shown in figure 7.4(c). The plot clearly show the barrierless
potential energy variation of S1 with the torsional motion of the N,Ndimethylamino substituted phenyl ring supporting the previous understanding and
our experimental data. This barrierless potential energy curve clearly shows the
dependence of S1 relaxation dynamics on viscosity. The S2 curve has a potential
well i.e., the S2 relaxation dynamics do not change much with the rotation of
substituted phenyl ring, which is also observed in the present femtosecond
fluorescence spectroscopic measurements. This leads to the meager dependence of
S2 relaxation dynamics on solvent viscosity and thus also supports our
experimental observations. The theoretical calculations are also an indicative of the
presence of a conical intersection or an avoided conical intersection between the S 0
and S1 electronic states. This may be regarded as an evidence of the presence of
some intermediate state (Sx) in the deactivation pathway of S1 state. The present
calculations thus suggest that the rotation of the substituted phenyl ring in MG may
be responsible for the excited state relaxation dynamics of both S1 and S2
electronic states.
7.3.4. Microviscosity of Water in the Nano-pool of AOT Reverse Micelle
The observed viscosity dependence of S1 lifetime of MG is exploited to
determine the microviscosity of a nano-confined environment. Previously there
was an attempt to measure the viscosity of micellar interface by Bhattacharyya and
co-workers.26
They
measured
the
photo-isomerization
of
3,3'-
diethyloxadicarbocyanine iodide in different micelles and found the viscosity of
micellar interface to be very large compared to bulk water. Hasegawa and coworkers used Auromine-O (AuO), whose fluorescence quantum yield increases
with increasing solvent viscosity, as a probe to determine the microviscosity of
AOT reverse-micelle. They have observed that for AOT reverse-micelle of w0 = 4
the microviscosity is about 80 times higher compared to bulk water. 27 Levinger’s
group and Fayer’s group have contributed a lot to understand the rigidity of water
inside the reverse micelles.28-30 In present study, we choose the AOT reverse-
Chapter 7
203
micelle in n-heptane with varying water pool sizes, w0 = 2, 5, 10, 20, 30, 40 as a
model nano-confined environment. Being an ionic species, MG is completely
insoluble in n-heptane and also in the reverse micellar side chain, which enables it
to be found only inside the water pool. We are relying on the core-shell model of
water inside the AOT reverse-micelle.31 Since the dye molecules are carrying a
positive charge and the AOT surface is negative, the dye molecules will prefer to
stay in the interfacial region rather than in bulk. The work of Crans and Levinger
especially on the highly charged inorganic molecule, decavanadate has elucidated
the structure of water in reverse micelle to a large extent and concluded that the
decavanadate ion prefers to reside in the interfacial region of reverse micelle using
many techniques.29-32 Thus the viscosity which we are reporting is actually the
micro-viscosity at the interface.
The absorption maxima of MG in water pool of the AOT reverse-micelle are
found to be at 425 nm (S2  S0) and 630 nm (S1  S0) while as emission maxima
of the S2 and S1 fluorescence are observed at ca. 470 nm and 690 nm respectively
as is shown in figure 7.5(a) for water pool size w0 = 20. To determine the viscosity
of water inside the reverse-micellar system, we measured the S1 state relaxation
Figure 7.5. (a) Steady state absorption spectra (red) and Fluorescence spectra (black) of
malachite green in AOT reverse micelle of water pool size w0 = 20. (b) Fluorescence
transients of malachite green in AOT reverse micelle of different water pool sizes; w0 = 2
(red), w0 = 10 (green) and w0 = 40 (black). The sample was excited at 410 nm. The solid line
represents the best fits of the experimental data.
204
Chapter 7
time of MG in the water pool. The MG was excited with an excitation source of
410 nm and the fluorescence transients were collected at magic angle polarization
at 670 nm. Figure 7.5(b) shows the fluorescence transients of MG in AOT reversemicelle with different amount of water loading (w0 = 2, 10, 40). It is observed that
the S1 decay of MG inside the water pool exhibits a bi-exponential behavior with a
fast rise component and slower decay component similar to that in bulk water and
in different glycerol-water mixtures (Table 7.2). In all the samples (w0 = 2, 5, 10,
20, 30, 40), we observed that the rise time (S1, rise) is almost constant with a value
of 250  50 fs. The decay time of the S1 state (S1,
decay)
however keeps on
decreasing from 3750 fs in w0 = 2 to 2650 fs in w0 = 40. Thus with increase in size
of water-pool the decay time of S1 state keeps on decreasing and it is expected to
decrease until the water trapped inside reverse-micelle exhibits the characteristics
of bulk water. On correlating the decay time of S1 state of MG in AOT reversemicelle interface with the viscosity dependent S1 decay time of MG in glycerolwater mixtures, we estimate the microviscosity of water trapped inside the
nanocavity as shown in figure 7.6(a). The measured S1 decay time readily reveals
that the water trapped inside the reverse-micelle is more viscous than bulk water.
In w0 = 2 the measured microviscosity is 9.0 cP which is 9 times higher compared
to the bulk water. With increase in the size of water-pool from w0 = 2 to w0 = 40
the microviscosity keeps on decreasing from 9.0 cP to 5.68 cP as shown in figure
7.6(b). Without a proper mathematical formulation we propose the decrease of
Table 7.2. Observed value of S1,decay of malachite green and estimated viscosity of water
inside the water pool of different AOT reverse-micelle with varying water content.
w0
S1,decay (fs)a
 (cP)b
2
3750
9.00
5
3500
8.21
10
3200
7.30
20
2800
6.12
30
2750
5.97
40
2650
5.68
100fs. b0.3 cP
a
Chapter 7
205
Figure 7.6. (a) Viscosity of water inside the AOT reverse micelle of water pool sizes w0 = 2,
5, 10, 20, 30, 40 directly obtained by comparing the S1, decay of MG with the viscosity
dependent dynamics of MG shown in figure 3. (b) Exponential decrease of viscosity of water
inside the AOT reverse-micelle as a function of water pool size (w0).
microviscosity with increasing size of nanocavity to follow an exponential
behavior. These experimental results thus confirm that the behavior of a solvent is
quite different when it is trapped in the interfacial region of nano-dimensional
cavity compared to that in bulk as already discussed by many authors.10-15,23,37
As observed from the experimental results, microviscosity increases with
decrease in the water-pool size. It is well documented that in smaller water-pool
sizes there is higher water structure.11,13,17,29,30,32 Bakker and co-workers proposed
that there are almost six to seven hydrogen bonds per surfactant molecule, closely
matching the three sulphonate oxygen atoms per AOT molecule, each of which can
form two hydrogen bonds.33 Levinger and others also observed a slow solvation
dynamics in the interior of reverse micelle compared to that in bulk water. 34-36 All
these were interpreted as a consequence of stronger hydrogen bonding of water
molecules in the interfacial region than in the bulk of reverse micelle. 15,37 As
malachite green is a positively charged molecule it always prefers to stay in the
negatively charged interface of AOT reverse micelle. We strongly believe that the
observed high microviscosity of the interface in small water pool compared to
larger one is due to the stronger hydrogen bonding in the interfacial region.
Previous results also show that the dynamics is getting faster as the size of the
206
Chapter 7
water pool increases and were explained as an outcome of weakening of the water
structure. Consequently, the friction posed by the water molecules will also reduce
with increase in the size of the water pool. We believe this as the origin of the
water pool size dependent microviscosity of the AOT reverse micelle.
Chapter 7
207
7.4. Conclusion
The ultrafast relaxation mechanism of malachite green (MG) has been revisited
by using femtosecond fluorescence up-conversion spectroscopy combined with
quantum mechanical calculations. The femtosecond measurements and theoretical
calculations completely oppose the view proposed by Bhasikuttan et al.5c In this
Chapter we propose that conical intersection does not exist between S 2 and S1
states, rather it is between S1 and S0 electronic states as suggested by several other
authors.1-4,5a-b This work also proposes that the conical intersection is not along the
torsional coordinate of un-substituted ring, rather it is along the torsional
coordinate of N,N-dimethylamino substituted phenyl ring. The femtosecond
fluorescence up-conversion measurements and TD-DFT calculations support the
viscosity dependent barrierless relaxation of the S1 state and it also justifies the
viscosity independent behavior of the S2 state relaxation due to the presence of a
potential well and an activation energy barrier.
Finally, the microviscosity of interfacial water inside the nano-water pool of
AOT reverse-micelle has been measured by correlating the relaxation time of MG
in reverse-micelle with the relaxation time of MG in different glycerol-water
mixtures. The data reveals that the decay time of S1 excited state of MG in AOT
water pool with w0 = 2 is almost 5.6 times greater than in bulk water, which leads
us to conclude that the microviscosity of water in the nano water pool of AOT
reverse-micelle (w0 = 2) is almost 9 times higher than the viscosity of bulk water. It
is also observed that as the size of the water pool increases from w0 = 2 to w0 = 40,
the microviscosity of water goes on decreasing and the decrease is assumed to
follow an exponential nature.
Chapter 7
208
References
1. Duxbury, D. F. Chem. Rev. 1993, 93, 381.
2. (a) Ippen, E. P.; Shank, C. V.; Bergman, A. Chem. Phys. Lett. 1976, 38, 611.
(b) Ben-Amotz, D.; Harris, C. B. J. Chem. Phys. 1987, 86, 4856. (c)
Mokhtari, A.; Fini, L.; Chenoy, J. J. Chem. Phys. 1987, 86, 3429.
3. (a) Sundstorm, V.; Gillbro, T.; Bergstorm, H. Chem. Phys. 1982, 73, 439.
(b) Sundstorm, V.; Gillbro, T. J. Chem. Phys. 1984, 81, 3463. (b) Saikan, S.;
Sei, J. J. Chem. Phys. 1983, 79, 4154. (c) Mokhtari, A.; Fini, L.; Chesnoy, J.
J. Chem. Phys. 1987, 87, 3429.
4. (a) Janowski, A.; Rzeszotarska, J. J. Lumn. 1980, 21, 409. (b) Nagasawa, Y.;
Ando, Y.; Kataoka, D.; Matsuda, H.; Miyasaka, H. J. Phys. Chem. A 2002,
106, 2024. (c) Nagasawa, Y.; Ando, Y.; Okada, T. Chem. Phys. Lett. 1999,
312, 161.
5. (a) Mokhtari, A.; Chebira, A.;Chesnoy, J. J. Opt. Soc. Am. B 1990, 8, 1551.
(b) Yoshizawa, M.; Suzuki, K. Kubo, A.; Saikan, S. Chem. Phys. Lett. 1998,
290, 43. (c) Bhasikuttan, A. C.; Sapre, A. V.; Okada, T. J. Phys. Chem. A
2003, 107, 3030.
6. (a) Cremers, D. A.; Windsor, M. W. Chem. Phys. Lett. 1980, 71, 27. (b)
Martin, M. M.; Breheret, E.; Nesa, F.; Meyer, Y. H. Chem. Phys. 1989, 130,
279.
7. (a) Canva, M.; Saux, G. L.; Georges, P.; Brun, A.; Chaput, F.; Boilot, J. P.
Chem. Phys. Lett. 1991, 176, 495. (b) Nakatsuka, H.; Hirai, M.; Sekine, S.;
Suzuki, Y.; Hattori, T. Jpn. J. Appl. Phys. 1999, 38. L324. (c) Nagasawa,
Y.; Nakagawa, Y.; Nagafuji, A.; Okada, T.; Miyasaka, H. J. Mol. Str. 2005,
735, 217.
8. (a) Morgenthaler, M. J. E.; Meech, S. R. Chem. Phys. Lett. 1993, 202, 57.
(b) Shi, X.; Borguet, E.; Tarnovsky, A. N.; Eisenthal, K. B. Chem. Phys.
1996, 205, 167. (c) Punzi, A.; Martin-Gassin, G.; Grilj, J.; Vauthey, E. J.
Phys. Chem. B 2009, 113, 11822. (d) Sen, P.; Yamaguchi, S.; Tahara, T.
Faraday Diss. 2010, 145, 411.
Chapter 7
209
9. (a) Bhattacharyya, K. Acc. Chem. Res. 2003, 36, 95. (b)Bhattacharyya, K.
Chem. Comm. 2008, 25, 2848. (c) Fenn, E. E.; Wong, D. B.; Fayer, M. D.
Proc. Natl. Acad. Sci. 2009, 106, 15243. (d) Setua, P.; Seth, D.; Sarkar, N.
Phys. Chem. Chem. Phys. 2009, 11, 8913.
10. (a) Maitra, A. J. Phys. Chem. 1984, 88, 5122. (b) Moulik, S. P.; De, G. C.;
Bhowmik, B. B.; Panda, A. K. J. Phys. Chem. B 1999, 103, 7122. (c) Boyd,
J. E.; Briskman, A.; Sayes, C. M.; Mittleman, D. M. Colvin, V. L. J. Phys.
Chem. B 2002, 106, 6346.
11. Levinger, N. E.; Swafford, L. A. Annu. Rev. Phys. Chem. 2009, 60, 385.
12. Cringus, D.; Bakulin, A.; Lindner, J.; Vohringer, P.; Pshenichnikov, M. S.;
Wiersma, D. A. J. Phys. Chem. B 2007, 111, 14193.
13. Cringus, D.; Lindner, J.; Milder, M. T. W.; Pshenichnikov, M. S.;
Vohringer, P.; Wiersma, D. A. Chem. Phys. Lett. 2005, 408, 162.
14. Dokter, A. M.; Petersen, C.; Woutersen, S.; Bakker, H. J. J. Chem. Phys.
2008, 128, 44509.
15. Piletic, I. R.; Moilanen, D. E.; Spry, D. B.; Levinger, N. E.; Fayer, M. D. J.
Phys. Chem. A. 2006, 110. 4985.
16. Venables, D. S.; Huang, K.; Schmuttenmaer, C. A. J. Phys. Chem. B 2001,
105, 9132.
17. Moilanen, D. E.; Levinger, N. E.; Spry, D. B.; Fayer, M. D. J. Am. Chem.
Soc. 2007, 129, 14311.
18. Tan, H. S.; Piletic, I. R.; Riter, R. E.; Levinger, N. E.; Fayer, M. D. Phys.
Rev. Lett. 2005, 94, 57405.
19. Correa, N. M.; Levinger, N. E. J. Phys. Chem. B 2006, 110, 13050.
20. Novaira, M.; Biasutti, M. A.; Silber, J. J.; Correa, N. M. J. Phys. Chem. B
2007, 111, 748.
21. Novaira, M.; Moyano, F.; Biasutti, M. A.; Silber, J. J.; Correa, N. M.
Langmuir 2008, 24, 4637.
22. Moyano, F.; Falcone, R. D.; Mejuto, J. C.; Silber, J. J.; Correa, N. M. Chem.
Eur. J. 2010, 16, 8887.
Chapter 7
210
23. Kinugasa, T.; Kondo, A.; Nishimura, S.; Miyauchi, Y.; nishii, Y.; Watanabe,
K.; Takeuchi, H. Colloids and Surfaces A: Physiochem. Eng. Aspects 2002,
204, 193.
24. (a) Amararene, A.; Gindre, M.; Le Huerou, J.-Y.; Nicot, C.; Urbach, W.;
Waks, M. J. Phys. Chem. B 1997, 101, 10751. (b) Boyd, J. E.; Briskman, A.;
Sayes, C. M.; Mittleman, D. M. Colvin, V. L. J. Phys. Chem. B 2002, 106,
6346.
25. Sen, P.; Roy, D.; Mondal, S. K.; Sahu, K.; Ghosh, S.; Bhattacharyya, K. J.
Phys. Chem. A 2005, 109, 9716.
26. Pal, S. K.; Datta, A.; Mandal, D.; Bhattacharyya, K. Chem. Phys. Lett. 1998,
288, 793.
27. Hasegawa, M.; Sugimura, T.; Suzaki, Y.; Shindo, Y. J. Phys. Chem. 1994,
98, 2120.
28. Corbeil, E. M.; Riter, R. E.; Levinger, N. E. J. Phys. Chem. B 2004, 108,
10777.
29. Baruah, B.; Roden, J. M.; Sedgwick, M.; Correa, N. M.; Crans, D. C.;
levinger, N. E. J. Am. Chem. Soc. 2006, 128, 12758.
30. Baruah, B.; Swafford, L. A.; Crans, D. C.; Levinger, N. E. J. Phys. Chem. B
2008, 112, 10158.
31. Crans, D. C.; Rithner, C. D.; Baruah, B.; Gourley, B. L.; Levinger, N. E. J.
Am. Chem. Soc. 2006, 128, 4437.
32. Sedgwick, M. A.; Crans D. C.; Levinger, N. E. Langmuir 2009, 25, 5496.
33. Dokter, A. M.; Woutersen, S.; Bakker, H. J. Proc. Natl. Acad. Sci. USA
2006, 103, 15355.
34. Riter, R. E.; Willard, D. M.; Levinger, N. E.; J. Phys. Chem. B 1998, 102,
2705.
35. Willard, D. M.; Riter, R. E.; Levinger, N. E. J. Am. Chem. Soc. 1998, 120,
4151.
36. Nandi, N.; Bhattacharyya, K.; Bagchi, B. Chem. Rev. 2000, 100, 2013.
Chapter 7
211
37. Piletic, I. R.; Moilanen, D. E.; Levinger, N. E.; Fayer, M. D. J. Am. Chem.
Soc. 2006, 128, 10366.
212
Chapter 7
Chapter 7
213
Concluding Remarks and Future Outlook
In this thesis, we have studied many fundamental aspects of the role of twisting
dynamics in determining the fate of excited state relaxation behavior of molecular
systems in condensed phase, using femtosecond fluorescence up-conversion
spectroscopy along with a support from theoretical calculations. The use of
femtosecond transient absorption spectroscopy in defining the dynamics of the
molecules can also be seen in one of the chapters. Further, once the excited state
relaxation mechanism of a molecular system has been established, it has been
exploited to understand the existence of some natural phenomena like the highly
fluorescent nature of wild type GFP. It can also be manipulated to define the
variation of the shape of potential energy surfaces as a function of solvent
properties, and once the shape of potential energy surface is established, it can be
used to know the properties of the media, like microviscosity of a nano-confined
water.
Although the above mentioned techniques have revealed very useful
information about the excited state properties of molecular systems, there are
various other aspects, which once explored, can give further information about the
complete excited state depletion mechanism. The femtosecond transient absorption
spectroscopy1-3 can be used exclusively on all of these molecular systems to reveal
especially the nature of the dark states, which are otherwise unable to probe
through time resolved fluorescence spectroscopy. Such facility has recently been
available in our laboratory and the work is being pursued to study excited state
dynamics in much more detail. Femtosecond pump-dump-probe spectroscopy3-7
can prove to be very essential tool to have a vivid knowledge about the nature of
potential energy surfaces of molecular systems at the experimental level. Time
resolved Raman spectroscopy8-11 is another technique, which can be utilized in
analyzing the structural changes of a particular functional group upon perturbation
by the pump pulse. Irrespective of predicting the involvement of particular
molecular fragments indirectly, one can actually locate the exact molecular
Chapter 7
214
fragment undergoing structural changes in the excited state using time resolved
Raman spectroscopy. Thus, the pump-dump-probe and time resolved Raman
spectroscopy, in conjunction with time resolved fluorescence and absorption
spectroscopy can reveal a very authentic and complete excited state relaxation
behavior of the molecular systems of interest.
References
1. Chen, J.; Thazhathveetil, A. K.; Lewis, F. D.; Kohler, B. J. Am. Chem. Soc.
2013, 135, 19290.
2. Wang, L.; Puodziukynaite, E.; Grumstrup, E. M.; Brown, A. C.; Keinan, S.;
Schanze, K. S.; Reynolds, J. R.; Papanikolas, J. M. J. Phys. Chem. Lett.
2013, 4, 2269.
3. Changenet-Barret, P.; Loukou, C.; Ley, C.; Lacombat, F.; Plaza, P.; Mallet,
J. –M.; Martin, M. M. Phys. Chem. Chem. Phys.2010, 12, 12715.
4. Kim, P. W.; Rockwell, N. C.; Freer, L. H.; Chang, C. –W.; Martin, S. S.;
Lagarias, J. C.; Larsen, D. S. J. Phys. Chem. Lett. 2013, 4, 2605.
5. Kim, P. W.; Freer, L. H.; Rockwell, N. C.; Martin, S. S.; Lagarias, J. C.;
Larsen, D. S. J. Am. Chem. Soc. 2012, 134, 130.
6. Wei, Z.; Nakamura, T.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2011,
133, 8205.
7. Bismuth, O.; Komm, P.; Friedman, N.; Eliash, T.; Sheves, M.; Ruhman, S.
J. Phys. Chem. B 2010, 114, 3046.
8. Su, T.; Ma, J.; Li, M. –D.; Guan, X.; Yu, L.; Phillips, D. L. J. Am. Chem.
Soc. 2012, 134, 14858.
9. Kruglik, S. G.; Yoo, B. –K.; Franzen, S.; Vos, M. H.; Martin, J –L.;
Negrerie, M. Proc. Natl. Acad. Sci. USA 2010, 107, 13678.
10. Shim, S.; Dasgupta, J.; Mathies, R. A. J. Am. Chem. Soc. 2009, 131, 7592.
11. Kukura, P.; McCamant, D. W.; Yoon, S.; Wandschneider, D. B.; Mathies,
R. A. Science, 2005, 310, 1006.
Chapter 7
215
List of Publications
1. Microviscosity Inside a Nanocavity: A Femtosecond Fluorescence Upconversion study of Malachite Green
Shahnawaz Rafiq, Rajeev Yadav, Pratik Sen, J. Phys. Chem. B 2010, 114,
13988.
2. Femtosecond Excited-State Dynamics of NPP: Evidence of Twisted
Intramolecular Charge Transfer and Intersystem Crossing Involving the
Nitro Group
Shahnawaz Rafiq, Rajeev Yadav, Pratik Sen, J. Phys. Chem. A 2011, 115,
8335.
3. Excited State Relaxation Dynamics of Model Green Fluorescent Protein
Chromophore Analogs: Evidence for Cis - Trans Isomerism
Shahnawaz Rafiq1, Basanta K. Rajbongshi1, Nishant Nair, Pratik Sen,
Gurunath Ramanathan, J. Phys. Chem. A 2011, 115, 13733.
4. Trinuclear Bright Red Luminophore Containing Cyclometallated Ir(III)
motifs
V. Chandrashekhar, S. M. Wahidur Rahaman, T. Hajra, D. Das, T. Ghatak,
Shahnawaz Rafiq, Pratik Sen, Jiten K. Bera, Chem. Commun. 2011, 47,
10836.
5. Origin of Strong Synergism in Weakly Perturbed Binary Solvent Solvent: A
Case Study of Primary Alcohols and Chlorinated Methanes
Shradhey Gupta, Shahnawaz Rafiq, Mainak Kundu, Pratik Sen, J. Phys.
Chem. B 2012, 116, 1346.
6. Dielectric Controlled Excited State Relaxation Pathways of a Representative
Push-Pull Stilbene: A mechanistic Study using Femtosecond Fluorescence
Upconversion Technique
Shahnawaz Rafiq, Pratik Sen J. Chem. Phys. 2013, 138, 84308.
7. Spectroscopic Evidence of the Presence of an Activation Barrier in the
otherwise Barrierless Excited State Potential Energy Surface of AuramineO: A Femtosecond Fluorescence Up-conversion Study
Shahnawaz Rafiq, Pratik Sen, J. Chem. Phys. 2013, 139, 124302.
216
Chapter 7
8. Stereoselective Synthesis of a Z-acrylonitrile Derivatives: Catalytic and
Acetylcholinesterase Inhibition Studies
Mehtab Parveen, Ali Mohammad Malla, Mahboob Alam, Musheer Ahmad,
Shahnawaz Rafiq, New J. Chem. 2014, accepted.
9. Coupled Proton–Electron Transfer Process in a New Heterocyclic Based Hydrogen
Bond Relay System.
Md. Serajul Haq Faizi,1 Shahnawaz Rafiq,1 Gyanesh Kumar Saxena, Pratik Sen,
Manuscript Submitted.
10. A Combined Femtosecond Fluorescence Up-conversion and Femtosecond
Transient Absorption Spectroscopic Study of two GFP Chromophore
Analogs: Understanding the Weakly Fluorescent Nature of Unfolded GFP
Shahnawaz Rafiq1, Basanta K. Rajbongshi1, Gurunath Ramanathan, Pratik
Sen, Manuscript under Preparation
11. Discovery of β-Carboline Analogue as a Probe for Optical pH Sensor
Shashi Dighe, Shahnawaz Rafiq, Visakh K. Mohan, Pratik Sen, Sanjay
Batra, Manuscript under preparation
12. Appearance of Slow Solvation Dynamics and Diffusion Coefficient in
Methanol-Chloroform Binary Solvent Mixture: A Case of Synergistic
Solvation in Mixed Solvents
Shradhey Gupta, Shahnawaz Rafiq, Bhaswati Sengupta, Rajeev Yadav,
Pratik Sen, Manuscript under preparation