Download Electric field, Magnetic field and Magnetization: THz time

Electric field, Magnetic field and Magnetization:
THz time-domain spectroscopy studies
Proefschrift
ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van de Rector Magnificus prof. ir. K. C. A. M. Luyben,
voorzitter van het College voor Promoties,
in het openbaar te verdedigen op maansdag 6 July 2015 om 12:30 uur
door
Nishant Kumar
Master of Science in Photonics
Cochin University of Science and Technology, Cochin, India
geboren te Patna, India.
Dit proefschrift is goedgekeurd door de promotor:
Prof. dr. P. C. M. Planken
Samenstelling promotiecommissie:
Rector Magnificus,
Prof. dr. P. C. M. Planken,
Dr. A. J. L. Adam,
Prof. dr. H. P. Urbach,
Prof. dr. H. J. Bakker,
Prof. dr. L. D. A. Siebbeles,
Prof. dr. Ir. L. J. van Vliet,
Dr. W. A. Smith,
voorzitter
Technische Universiteit Delft/ARCNL, promotor
Technische Universiteit Delft, copromotor
Technische Universiteit Delft
FOM-Instituut voor Atoom- en Molecuulfysica
Technische Universiteit Delft
Technische Universiteit Delft
Technische Universiteit Delft
This work was funded by the Nederlandse Organisatie voor Wetenschappelijk
Onderzoek (NWO) and the Stichting voor Technische Wetenschappen (STW).
c 2015 by N. Kumar
Copyright All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form or by any means: electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission of the author.
isbn:
Printed in the Netherlands by Ipskamp Drukkers, Enschede.
A free electronic version of this thesis can be downloaded from:
http://www.library.tudelft.nl/dissertations
Author email: [email protected]
To my parents
Contents
1 Introduction
1.1 Terahertz radiation . . . . . . . . . . .
1.2 Applications of THz radiation . . . . .
1.3 THz time domain spectroscopy . . . .
1.4 Generation of THz radiation . . . . . .
1.4.1 Biased semiconductor emitters
1.4.2 Photo-Dember effect . . . . . .
1.4.3 Optical Rectification . . . . . .
1.5 Terahertz detection mechanisms . . . .
1.5.1 Electro-optic detection . . . . .
1.5.2 Magneto-optic detection . . . .
1.6 Measuring the THz electric near-field .
1.7 Scope and organization of the thesis .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
2
2
4
4
6
6
8
8
9
9
11
2 THz near-field Faraday imaging
2.1 Metamaterials . . . . . . . . . . . . . . . . . . . .
2.1.1 Split-ring resonator . . . . . . . . . . . . .
2.2 Imaging the terahertz magnetic field . . . . . . .
2.2.1 Measuring the terahertz magnetic far-field
2.2.2 Imaging terahertz magnetic near-field . .
2.3 Experimental . . . . . . . . . . . . . . . . . . . .
2.3.1 Sample fabrication . . . . . . . . . . . . .
2.3.2 Results and Discussions . . . . . . . . . .
2.3.3 Single point measurement . . . . . . . . .
2.3.4 Two dimensional distribution . . . . . . .
2.4 Double Split Ring Resonator . . . . . . . . . . .
2.4.1 Single point measurements . . . . . . . .
2.4.2 Two dimensional distribution . . . . . . .
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . .
2.6 Complementary split ring resonators . . . . . . .
2.6.1 Sample fabrication . . . . . . . . . . . . .
2.6.2 Single point measurement . . . . . . . . .
2.6.3 Two dimensional distribution . . . . . . .
2.6.4 Conclusion . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
13
15
17
17
18
19
19
20
21
23
25
26
26
27
29
30
31
31
33
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
iv
3 THz emission from ferromagnetic metal thin films
3.1 Ferromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Laser-induced ultrafast demagnetization . . . . . . . . . . . . . .
3.2.1 Historical review . . . . . . . . . . . . . . . . . . . . . . .
3.3 THz emission from non-magnetic metal thin films . . . . . . . . .
3.4 THz emission from ferromagnetic metal thin films . . . . . . . .
3.5 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.5.1 Sample fabrication . . . . . . . . . . . . . . . . . . . . . .
3.5.2 THz generation and detection setup . . . . . . . . . . . .
3.5.3 Magnetic force microscopy . . . . . . . . . . . . . . . . . .
3.6 Result and discussions . . . . . . . . . . . . . . . . . . . . . . . .
3.6.1 THz emission from cobalt thin film . . . . . . . . . . . . .
3.6.2 Azimuthal angle dependence . . . . . . . . . . . . . . . .
3.6.3 THz emission in back reflection . . . . . . . . . . . . . . .
3.6.4 Thickness dependent THz emission . . . . . . . . . . . . .
3.6.5 MFM measurements . . . . . . . . . . . . . . . . . . . . .
3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.8 Effect of capping layer on the Terahertz emission . . . . . . . . .
3.8.1 THz emission from Pt/Co thin films . . . . . . . . . . . .
3.8.2 Relation between the magnetic order and THz emission .
3.8.3 Azimuthal angle dependence . . . . . . . . . . . . . . . .
3.8.4 Thickness dependent THz emission . . . . . . . . . . . . .
3.8.5 Effect of changing the order of the films on THz emission
3.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
35
35
37
37
41
41
42
42
43
44
45
45
46
48
49
50
52
54
55
56
57
58
59
60
4 THz emission from BiVO4 /Au thin films
4.1 Motivation . . . . . . . . . . . . . . . . . .
4.2 THz generation from semiconductors . . . .
4.2.1 Surface field effect . . . . . . . . . .
4.2.2 Photo-Dember effect . . . . . . . . .
4.3 Bismuth Vanadate . . . . . . . . . . . . . .
4.3.1 BiVO4 structure . . . . . . . . . . .
4.3.2 Preparation of BiVO4 thin film . . .
4.4 Experimental . . . . . . . . . . . . . . . . .
4.4.1 Sample fabrication . . . . . . . . . .
4.4.2 THz generation and detection setup
4.5 Results and discussion . . . . . . . . . . . .
4.5.1 Thickness dependent THz emission .
4.6 Conclusion . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
61
61
61
62
64
65
66
67
68
68
68
70
73
74
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5 Conclusions
77
Appendix
79
Bibliography
80
Summary
95
v
Samenvatting
97
Acknowledgements
101
Biography
105
Chapter
1
1.1
Introduction
Terahertz radiation
Terahertz (THz) radiation is electromagnetic radiation which spans the frequency
range from 0.1 THz to 10 THz. In terms of wavelengths, it ranges from 30 µm to
3 mm. The THz region of the electromagnetic spectrum lies between the microwave
and the infrared regions, as shown in Fig. 1.1. Until recently, this region was also
known as the “THz gap” since efficient sources and detectors for THz radiation
were not available. The THz region is located where both electronic means and
optical means to generate light exist [1]. For example, low frequency THz radiation
can be made by electronically multiplying a lower frequency source [2]. A methanol
gas laser, on the other hand, is an example of an optical source of THz radiation [3].
THz radiation has many interesting applications. For example, it can pass through
many dielectric materials which are opaque to visible radiation and can thus be
used for imaging purposes. Unlike X-rays, THz radiation is non-ionizing and has
little effect on biological samples [4]. Many crystalline organic materials have
unique absorption spectra in the THz range which can be used as an optical
fingerprint and therefore can be used to identify the chemical structure of the
materials [5]. For example, Walther et al. showed that, by using THz radiation,
polycrystalline sucrose can be easily differentiated from other sugars [6].
Electronics
Radio
waves
Micro
waves
106
109
Photonics
THz
gap
1012
Infrared Visible X-rays
1014
1015
1018
Υ-rays
1021
Frequency (Hz)
Figure 1.1: Terahertz frequency region in the electromagnetic spectrum.
Chapter 1. Introduction
1.2
2
Applications of THz radiation
THz radiation is very effective for the analysis of solids, liquids or gases. Several
materials have characteristic strong absorption lines in the THz frequency range.
For many gas molecules, the energy required for the transitions between the rotational energy levels lies in the THz region. For example, dichloromethane has
rotational lines with transitions up to 2.5 THz [7].
In the solid crystalline phase, the atoms or molecules are held close to their equilibrium positions which leads to collective lattice vibrations at certain frequencies.
These low frequency vibrations of molecules in the solid state can often absorb
THz radiation [7]. Furthermore, in the case of semiconductors, absorption of THz
radiation is due to free electrons. THz radiation is non-ionizing radiation and can
be used for non-invasive and non-destructive imaging and spectroscopy.
THz radiation can pass through many packaging materials such as cardboard,
cloth, paper, ceramics and plastics. Hence, THz radiation can be used for package
inspection, security screening and quality control. For example, in 1995, Hu et
al. showed that THz radiation can be used to find defects in computer chips [8].
THz radiation is safe for humans and could be used for biomedical imaging [9].
It might be used for cancer detection, endoscopy and detection of tooth decay
[10]. THz radiation is often used for the characterization of old paintings, the
detection of explosives, drug screening, food inspection and to identify the chemical
composition of materials [11, 12]. THz radiation can not only image the hidden
underlying paint layers but can also give spectroscopic information on different
paint layers [13, 14]. THz radiation also has potential applications in the field of
near-field imaging. Using a THz near-field microscope, it is possible to achieve a
subwavelength spatial resolution [15, 16].
THz radiation has a lot of potential to be used for communication purposes.
Compared to microwaves, the THz band provides a larger bandwidth and, consequently, a potentially higher transmission rate. The low power and low efficiency
of THz sources and the strong absorption by water vapor present in the atmosphere are some disadvantages of using THz radiation. However, it is possible to
use THz radiation for communication over shorter distances and for satellite to
satellite communication, where atmospheric absorption is not a problem [17].
1.3
THz time domain spectroscopy
Terahertz time domain spectroscopy (THz-TDS) is a spectroscopic technique that
uses THz radiation to probe the properties of materials. The basic idea of THzTDS is that we measure the electric field of the THz pulse as a function of time. If
the THz pulse passes through a material, its time profile gets changed compared
to the reference pulse. The reference pulse can be a pulse propagating in vacuum
or in a medium with known properties. By comparing the THz pulse transmitted through the medium with the reference THz pulse we can find the changes
introduced by the material [18].
In Fig. 1.2 we show a typical experimental setup for THz generation and detection. A femtosecond laser pulse is split into two parts using a 80:20 beam splitter.
Chapter 1. Introduction
3
The part of the beam with higher energy is called the pump beam and the lower
energy one is called the probe beam. The pump pulse is incident on the emitter at a 45◦ angle of incidence and the THz radiation is collected in the reflection
direction. The experimental setup is described in detail in chapter 3 and chapter 4.
Beam
Splitter
Femtosecond
Laser
Wollaston
prism
Probe
Beam
Detection Quarter
Crystal Waveplate
Pump
Beam
THz
Beam
Differential
Photodetectors
Electro-optic detection
Sample
Parabolic
Mirrors
THz reflection setup
Figure 1.2: Terahertz far-field experimental setup.
A typical measurement of the THz electric field as a function of time is shown in
Fig. 1.3(a). The corresponding THz frequency spectrum is obtained by taking the
Fourier transform of the emitted THz electric field, which is shown in Fig. 1.3(b).
THz-TDS measures the electric field instead of the intensity of a light pulse and
hence gives information about both the amplitude and the phase of the light.
Using THz-TDS, the absorption coefficient and the refractive index of a sample
at different frequencies can be calculated simultaneously without the need for a
physical model of the absorption [19].
(a)
(b)
Figure 1.3: (a) Measured temporal waveform of a THz pulse generated in a gallium
phosphide (GaP) (110) crystal by optical rectification and (b) the corresponding
calculated frequency spectrum. As a detection crystal we have selected GaP, which
is 300 µm thick.
Chapter 1. Introduction
1.4
4
Generation of THz radiation
There are different ways of generating THz radiation and in this thesis we will
focus only on the optical methods of THz generation. This includes generation
of THz radiation from biased semiconductor emitters, THz emission due to the
photo-Dember effect and THz emission due to optical rectification.
1.4.1
Biased semiconductor emitters
When a femtosecond laser pulse is incident on a semiconductor, if the energy of
the photons is greater than the bandgap energy of the semiconductor, electronhole pairs are generated. Due to the applied bias these carriers accelerate and a
transient photocurrent is formed on a subpicosecond time scale. The bias can be
either an externally applied voltage, as in the case of a photoconductive antenna,
or it can be an intrinsic electric field, as in the case of depletion field emitters.
Photoconductive antenna - The photoconductive antenna (PCA) is one of
the most common and efficient sources of THz radiation. A schematic diagram of a
PCA is shown in Fig. 1.4. In a PCA, we have a semiconductor with two electrodes
attached to the surface [20]. These electrodes are biased with an externally applied
voltage. When charge carriers are excited using a fs laser pulse, this external bias
accelerate the carriers to form a photocurrent J(t).
+V
SI-GaAs
substrate
Current
+
THz
pulse
fs laser
pulse
-
+
-
+
Silicon
lens
Electrodes
-V
Figure 1.4: Schematic diagram of generation of THz radiation from a photoconductive antenna. When a femtosecond laser pulse is incident on the semiconductor,
electron-hole pairs are generated. These carriers are accelerated due to the applied
bias and form a current on a sub picosecond time scale, which generates electromagnetic radiation in the THz range.
Since, the mobility of electrons is usually much higher than the mobility of holes,
the contribution of holes in the transient current can typically be neglected. The
Chapter 1. Introduction
5
magnitude of the transient photocurrent is described as [21] :
J(t) = N (t) |e| v(t) = N (t) |e| µE b ,
(1.1)
where N(t) is the density of photoexcited electrons, e = 1.6 × 10−19 C is the
charge per electron, v(t) is the velocity of the electrons, µ is the mobility of the
electrons and E b is the bias electric field.
The photocarrier density is a function of time and depends both on the carrier
life time τ c and the shape of the laser pulse. The charge carriers recombine and
the photocurrent decays. This transient photocurrent generates electromagnetic,
pulsed radiation in the THz frequency range. The THz radiation is emitted in
the direction of the propagation of the laser pulse and also in the direction of the
reflected laser pulse. The polarization of the generated THz radiation is parallel to
the applied bias field. The far electric field of the radiated THz radiation, ET Hz ,
is proportional to the first time derivative of the photocurrent J(t),
ET Hz ∝ ∂J(t)/∂t = |e| [N (t)∂v(t)/∂t + v(t)∂N (t)/∂t].
(1.2)
Equation 1.2 demonstrates that the time varying photocurrent and, hence, the
THz emission depends on two phenomena: 1) acceleration of carriers and 2) ultrafast variation in carrier density. The number of free carriers depends on the
optical power provided by the laser pulse and on the material itself whereas the
acceleration of these free carriers depends on the applied bias voltage. However,
the laser power and the bias voltage can be increased only till a threshold limit
because of the device saturation and/or the device breakdown [22].
In our setup, we have used semi insulating-GaAs as the semiconductor substrate
for fabricating the photoconductive antenna. On the back of the GaAs substrate,
we have used a hyperhemispherical silicon lens for coupling out and collimating
the THz radiation efficiently. Also, instead of applying a DC bias voltage, we have
used a 50 kHz, 400 V square wave ac bias voltage. When the voltage changes
from -400 V to +400 V, the sign of the THz signal also changes. Hence, the
signal detected by the lock-in is double the signal obtained with a DC bias voltage
that changes from 0-400 V. As a result, the measured signal is increased without
actually increasing the THz amplitude [23].
Depletion field emitter - THz emission from a depletion field is similar to
THz emission from a photoconductive antenna. Charge carriers are generated by
the photoexcitation of a semiconductor surface using a femtosecond laser. However, instead of applying an external bias voltage as in PCA, in this case the charge
carriers are accelerated due to an intrinsic electric field, present in the semiconductor. This intrinsic electric field can be formed when a metal comes in close contact
with an n-type semiconductor material. On contact, the electrons move from the
semiconductor to the metal and an electric field is formed near the surface. This
electric field is called the depletion field [24]. The direction of this field is from the
semiconductor to the metal. The depletion field drives the two kinds of carriers
in opposite directions and produces a photocurrent which leads to the formation
of a dipole-like layer in the direction of the surface normal. This transient dipole
emits a THz pulse [25]. The direction of the surface depletion field depends on
Chapter 1. Introduction
6
the type of doping. For n-type and p-type doping, the direction of the dipole is
opposite and hence the polarities of the emitted THz pulses are opposite too. The
formation of the depletion field and THz generation due to the depletion field are
discussed in more detail in section 4.2.1 of this thesis.
1.4.2
Photo-Dember effect
In the case of narrow-bandgap semiconductors, when the depletion field is weak,
THz generation is mostly due to the photo-Dember effect. In the photo-Dember
effect, the laser light is strongly absorbed by the semiconductor material, so that
the photoinduced electron-hole pairs form a concentration gradient close to the
surface of the semiconductor. Since electrons often have a higher mobility than
holes, they are able to diffuse faster. The combination of the concentration gradient and the difference in mobilities generate a transient dipole which emits THz
radiation [26]. The electric field of the emitted THz radiation, ET Hz , is directly
proportional to the derivative of the diffusion current Jd ,
ET Hz ∝
∂Jd
.
∂t
(1.3)
InAs and InSb are two examples of photo-Dember based THz emitters. Both
materials have a very high ratio of electron to hole mobilities and a very narrow
bandgap [27]. The photo-Dember effect is different from the depletion field effect.
The polarity of THz radiation emitted by the photo-Dember effect does not depend
on the doping type but it shows a strong dependence on temperature [28]. These
differences make it possible to separate the contributions to the THz emission
by the photo-Dember effect and from the current surge in the surface depletion
region. THz generation due to the photo-Dember effect is discussed in more detail
in section 4.2.2 of this thesis.
1.4.3
Optical Rectification
Another commonly used technique for the emission of THz radiation is optical rectification. The schematic representation of THz generation by optical rectification
is shown in Fig. 1.5. Optical rectification is a non-resonant method of generating
THz radiation meaning that no absorption is needed to create THz light. It is basically difference-frequency generation with the difference frequency close to zero.
It is a second-order non-linear optical effect [21]. The dielectric polarization of the
material is directly proportional to the applied electric field.
P = ε0 χ(E)E
(1.4)
Here, ε0 is the permittivity of free space and χ(E) is the electric susceptibility.
The nonlinear optical properties of the material can be described by expanding
the susceptibility χ(E) into a power series of the electric field E [18].
P = ε0 (χ1 + χ2 E + χ3 E 2 + χ4 E 3 + . . .)E
(1.5)
Chapter 1. Introduction
7
Here, optical rectification comes from the second term of the equation after the
equality sign. If we consider an optical electric field E described by E = E 0 cos ωt,
where E 0 = E 0 (t) for a laser pulse, then the second-order nonlinear polarization
P (2) is [18]:
P (2) = ε0 χ2 E 2 = ε0 χ2 (E 0 cos ωt)2 = ε0 χ2
E02
(2)
(2)
(1 + cos 2ωt) = POR + PSHG (1.6)
2
The second-order nonlinear polarization is a sum of two terms. The first term is
a quasi DC polarization (quasi, because E0 is not a constant but time-dependent
since we are dealing with a laser pulse), which results from the rectification of the
incident optical electric field by the second-order nonlinear electric susceptibility of
the material. The second term shows a cos 2ωt dependence and describes second
harmonic generation. Here, only the first term is relevant to the generation of
THz radiation [18]. Optical rectification occurs only in those crystals that are
not centrosymmetric – that is, crystals that do not display inversion symmetry.
In such non-centrosymmetric crystals, the second-order non-linear susceptibility
χ2 6= 0. When a femtosecond pulse is incident on such a crystal, due to the
optical rectification of the femtosecond laser pulses, subpicosecond THz pulses are
generated. In the far field, the radiated electric field E(t) is proportional to the
(2)
second time derivative of POR [21]:
(2)
(2)
ET Hz ∝ ∂ 2 POR /∂t2 ∼ ∂ 2 E0 /∂t2
(1.7)
Where, again, E 0 = E 0 (t)
Laser pulse
(fs)
THz pulse
(ps)
Non-linear
optical crystal
Figure 1.5: Schematic representation of THz generation by optical rectification. A
femtosecond laser pulse is incident on a nonlinear optical crystal and due to optical
rectification of the femtosecond laser pulses, THz pulses are generated.
One of the most important factors to take into account during THz generation from non-linear non-resonant optical processes is phase matching. The phase
matching condition is satisfied when the group velocity of the optical beam is equal
to the phase velocity of the THz beam [21, 29]. The advantage of using optical
Chapter 1. Introduction
8
rectification in non-absorbing materials, is that the response of the crystal is essentially instantaneous, in prinicple allowing for very short THz pulse durations. The
bandwidth depends only on the width of the laser pulse and the phase matching
conditions in the generation and detection crystals [21].
1.5
Terahertz detection mechanisms
The most common methods for detecting pulsed THz radiation are photoconductive detection and electro-optic/magneto-optic detection. Here, we describe only
the free-space electro-optic and magneto-optic detection methods in detail, because only these detection methods are used in the experiments described in this
thesis.
1.5.1
Electro-optic detection
The electro-optic detection method relies on a change of the polarization of the
probe pulse induced by the instantaneous electric field of the THz pulse [30].
The schematic diagram for electro-optic detection of THz radiation is shown in
Fig. 1.6. Initially, when no THz radiation is incident on the electro-optic crystal,
the linearly polarized probe beam remains linear even after passing through the
crystal. Then, the linearly polarized probe beam passes through a quarter wave
plate and becomes circularly polarized. A Wollaston prism splits this circularly
polarized light into two beams with equal intensities and orthogonal polarizations.
These two beams are focused on two photodiodes of a differential photodetector
and since both components have the same intensity, the detector signal is zero.
On the other hand, when THz radiation is incident on the electro-optic crystal,
the electric field of the THz pulse induces a birefringence in the crystal which is
proportional to the instantaneous THz field [31]. As a result of the birefringence,
when the linearly polarized probe beam passes through the crystal, it becomes
slightly elliptical. The quarter wave plate turns this into an elliptically polarized
beam that deviates slightly from a circularly polarized beam. The Wollaston
prism splits the elliptically polarized probe beam into two beams with orthogonal
polarizations but, now, with unequal intensities. The output of the differential
photodetector is proportional to the difference in the intensities of the two beams
which is directly proportional to the instantaneous THz electric field. Hence, by
measuring the change in the polarization of the probe beam as a function of delay
between the probe pulse and the THz pulse, the THz electric field as a function
of time can be measured, as shown in Fig. 1.6 [32].
Zinc telluride (ZnTe) and Gallium phosphide (GaP) are the most commonly
used electro-optic crystals for THz detection. By increasing the thickness of the
detection crystal we can increase the sensitivity of THz detection. However, increasing the thickness of the detection crystal also increases phase mismatching
effects [18]. This leads to a smearing out of the detected THz field as a function of time, which can have a negative effect on the shape of the detected THz
spectrum [33].
Chapter 1. Introduction
9
Polarization
states
Probe
beam
THz
beam
Detection
Crystal
THz pulse
Quarter
Waveplate
Wollaston
prism
Differential
Photodetector
EO crystal
ETHz(t)
Probe
Iprobe(t)
Figure 1.6: Schematic diagram of THz detection using free space electro-optic
sampling. The synchronised probe pulse samples the complete electric field of the
THz pulse by varying the delay between the THz pulse and the probe pulse.
1.5.2
Magneto-optic detection
Similar to electro-optic detection we also have a magneto-optic detection scheme
for THz radiation. In magneto-optic detection, instead of measuring the electric
field of THz radiation, we measure the THz magnetic field using the Faraday effect.
The linear polarization of the probe beam is rotated due to the THz magnetic field
which is measured using a magneto-optic crystal [34,35]. The free space magnetooptic detection of THz radiation is discussed in more detail in section 2.2.1.
1.6
Measuring the THz electric near-field
THz radiation has been widely used for imaging applications. However, the spatial
resolution that can be achieved when imaging with electromagnetic waves is limited
by diffraction. The size of the smallest objects that can be spatially distinguished is
theoretically about half of the wavelength of light. For a wavelength corresponding
to a frequency of 1 THz, the far field spatial resolution of an image is limited to
approximately 150 µm in vacuum. This constitutes a major issue in THz imaging
of subwavelength-sized objects [36]. There are different techniques to break the
diffraction limit and to achieve a better spatial resolution. One way to overcome
this limit is to work in the near-field region of the sample [37]. When the size
of the object is bigger or comparable to the wavelength of light, it has clearly a
visible effect in the far-field, but when the object is of sub-wavelength dimensions,
Chapter 1. Introduction
10
it affects the field only in a volume around the object which is comparable to the
size of the object itself. Hence, for the detection of subwavelength sized objects we
need a measurement method which can measure the field in the immediate vicinity
of the object. So, the idea is to capture the THz wave in the near field, very close
to the sample surface [37]. The advantage of THz near-field imaging over optical
near field imaging techniques is that it measures the electric field, rather than the
intensity of the light. Many schemes have been proposed for imaging the THz
electric near field, most of which involve an aperture, a tip or an electro-optic
detection crystal [38]. In experiments related to the one described in chapter 2 of
this thesis, electro-optic detection is used for imaging the THz near-field. For this
reason, a brief description of this technique follows:
Electro-optic detection of the THz electric near-field : In near-field
electro-optic detection, a tightly focused probe beam is used to detect the THz
electric near-field in a small volume [39]. The schematic diagram for this is shown
in Fig. 1.7. Gallium phosphide (GaP) is taken as a detection crystal [39]. On
top of the GaP crystal a reflective coating for the 800 nm beam, which consists
of 130 nm of SiO2 and 200 nm of Ge, is deposited. The sample in Fig. 1.7, illustrated as a hole in a gold film, is illuminated with a THz pulse from the top
and a probe pulse is incident on the sample from below. The probe pulse samples
the THz electric near-field of the structure and gets reflected due to the reflective
coating. Then, the probe beam passes through a λ/4 waveplate and a Wollaston
prism and is finally incident onto a differential detector. The coating also prevents
any probe light from reaching the sample, getting scattered and measured by the
detector [39]. The spot of the probe beam is used as a synthetic aperture and
only that part of THz radiation is detected which is present in the path of the
probe beam. The radius of the focal spot of the probe beam (800 nm) is related
to its own wavelength which is much smaller than the THz wavelength. Thus, this
method circumvents the THz diffraction limit [37,40]. By selecting the orientation
of the detection crystal, it is possible to select for which component of the THz
electric near-field vector an electro-optic detection setup is sensitive. For example, (100) oriented GaP or ZnTe crystals measure the component of the electric
near-field which is perpendicular to the sample surface, being blind to the in-plane
(x and y) components of the electric near-field [39]. Similarly, we can measure
the in-plane (x and y) components using a (110) crystal orientation [33, 37]. The
electro-optic crystal is mechanically raster scanned in all three directions to measure the electric near field and pixel by pixel an image is obtained. To obtain a
high spatial resolution, the electro-optic crystal should be thin and the interaction
region should be small. Hence, the probe beam should be well focused onto the
detection crystal [38].
Adam et al. measured the THz electric near-field in the vicinity of subwavelength
sized metallic spheres [40]. The same technique has been used for measuring the
electric near-field of many other structures and to perform microspectroscopy in
the THz frequency range [41–47].
In 2007, Bitzer et al. measured the in-plane electric near field distributions of
split-ring resonators [48]. A split-ring resonator (SRR) is a single ring, or concentric rings, of metal containing a gap. Later, the same group showed THz electric
Chapter 1. Introduction
11
THz
beam
Sample
scan
Gold
Ge
SiO2
GaP
Detector
Probe
beam
Figure 1.7: Schematic diagram of THz detection using free space electro-optic
sampling
near-field measurements of a complementary split-ring resonator (CSRR). They
showed that the magnetic near-field of a SRR and a CSRR can be calculated
from the electric near-field measurements [49]. Direct measurement of the magnetic near-field at THz frequencies is very challenging. Until recently, only far
field measurements of the THz magnetic field, using the Faraday effect, have been
shown by Riordan et al. [35].
1.7
Scope and organization of the thesis
Till now, we have discussed the generation and detection of, mostly, the THz electric field and the use of static electric fields for the generation of THz light. In
general, very little work has been done in the THz domain involving magnetic
fields. In this thesis we study magnetic field aspects of THz generation and detection. In Chapter 2, we report on the first direct measurement of the THz magnetic
near-field of split ring resonators using a magnetic field sensitive material. The
THz electric near-field of such a split-ring resonator has been measured before
but the magnetic near-field of a split-ring resonator is relatively weak and has
never been measured before. In chapter 3 we discuss emission of THz radiation
due to ultrafast demagnetization of ferromagnetic thin films. The THz electric
Chapter 1. Introduction
12
field emitted from thin cobalt films changes sign when the sample is rotated by
180◦ . However, for thicker cobalt layers, we observe the development of an azimuthal angle-independent contribution to the THz emission. Hence, for thick
cobalt films, the polarity remains unchanged with 180◦ sample rotation. We correlate these findings with a change in the magnetization of these films from in-plane,
to out-of-plane for increasing Co layer thickness, as measured using magnetic force
microscopy.
In contrast to the previous chapters, in chapter 4 the magnetic field does not
play a role. In this chapter, we show generation of THz light from BiVO4 /Au thin
films. The motivation behind this study is that BiVO4 is a wide-bandgap semiconductor, similar to Cu2 O. Recently it has been shown that when femtosecond
laser pulses with 800 nm wavelength are incident on Cu2 O/metal interfaces, strong
THz emission is observed. This is surprising, because the energy of the corresponding photons is much smaller than the bandgap of cuprous oxide. Therefore, it is
interesting to try other large bandgap materials as well, especially BiVO4 , which
is technologically relevant. BiVO4 is widely used in the pigment industry and
has potential applications for photoelectrochemical water splitting. We find that
BiVO4 /Au interfaces emit small amplitude THz pulses, when illuminated with
below-gap femtosecond laser pulses. By studying the THz radiation emitted from
these interfaces we propose that the most likely cause of the THz emission is the
photo-Dember effect. Finally, chapter 5 summarizes the experiments described in
this thesis.
Chapter
2
2.1
THz near-field Faraday
imaging
Metamaterials
Recently, there has been a lot of interest in metamaterials because of their exotic
properties and potential applications. These materials can manipulate light in
remarkable ways and have optical properties which are not available in nature.
Metamaterials provide a design-based approach to create novel electromagnetic
functionality. This functionality spans the spectrum from the microwave domain
to the visible domain with examples including cloaking and superlensing [50]. One
can think of metamaterials as pseudo materials in which periodically or randomly
distributed structures constitute the “atoms”. The size of the structures and the
spacing between them is much smaller than the wavelength of the electromagnetic
radiation. When electromagnetic radiation is incident on these materials, we can
pretend that the material is electromagnetically homogeneous. The properties of a
metamaterial are determined by the properties of the material from which each of
the individual elements is formed, the shape or the structure of the individual elements and the interaction between them. Metamaterials are characterized by two
fundamental macroscopic parameters: relative permittivity (εr ) and permeability
(µr ). The electric permittivity and magnetic permeability define the response of
the material when an oscillating electric field or magnetic field is applied. In the
case of metamaterials, the effective permittivity or permeability is an average or
collective response of all the elements of which the metamaterial is made. By
properly designing the structures we can control the relative permittivity and permeability. This has been used recently for various applications like imaging with
subwavelength spatial resolution, artificial magnetism, and to obtain a negative
refractive index [51].
In Fig. 2.1 we show the optical properties of the materials for different signs
of the permittivity (εr ) and permeability (µr ). In the case of ordinary optical
materials, the permittivity and permeability are usually positive (εr > 0, µr > 0),
as shown in the first (top-right) quadrant. The refractive index of such materials
√
is given by n = εr µr . In this case, the phase velocity is in the direction of the
~ the magnetic field H,
~ and the wave vector
Poynting vector. The electric field E,
~k form a right handed set of vectors. Hence, these materials are also called right
handed materials, which support forward propagating waves.
Chapter 2. THz near-field Faraday imaging
14
μr
Evanescent decaying wave
Forward propagating wave
E
S
S
H
k
εr > 0, μr > 0
εr < 0, μ r > 0
n = (εr μr )¹/² > 0 (real)
n = (εr μr )¹/² (imaginary)
Metals, doped semiconductors
(below plasma frequency)
Ordinary optical materials
Backward propagating wave
Evanescent decaying wave
εr
E
S
k
S
H
ε r< 0, μ r< 0
εr > 0, μr < 0
n = (εr μr )¹/² < 0 (real)
n = (εr μr )¹/² (imaginary)
Artificial Metamaterials
Some ferromagnetic metals
(up to GHz)
Figure 2.1: Classification of materials based on their permittivity, εr , and permeability, µr . Here, materials are divided into four different quadrants. (Quadrant I):
Both, εr and µr are positive. Most ordinary optical materials fall in this quadrant.
(Quadrant II): εr is negative and µr is positive. Metals and heavily doped semiconductors below plasma frequency fall in this quadrant. (Quadrant III): Both, εr and
µr are negative. No natural materials, only metamaterials show such characteristics
(Quadrant IV): εr is positive and µr is negative. Some ferromagnetic materials near
~ is the electric component
resonance frequency belong to this quadrant. Here, E
~
of the plane wave, H is the magnetic component of the plane wave, ~k is the wave
~ is the Poynting vector, where the Poynting vector is defined as S
~ =
vector and S
~ H.
~
E×
However, metals form an exception having a negative value for the real part
of the permittivity in a large range of the electromagnetic spectrum. Metals and
doped semiconductors can display a negative permittivity and a positive perme-
Chapter 2. THz near-field Faraday imaging
15
ability (εr < 0, µr > 0), below the plasma frequency and fall in the second
(top-left) quadrant. On the other hand, the permeability of ferromagnetic materials is negative whereas their permittivity is positive (εr > 0, µr < 0), near the
ferromagnetic resonance frequency [50] and they thus fall in the fourth (bottomright) quadrant. For materials with either permittivity or permeability less than
zero, the refractive index is imaginary, which supports the existence of evanescent
waves.
In 1968, Veselago proposed the concept of materials having simultaneously
a negative permittivity and a negative permeability (εr < 0, µr < 0, third
quadrant/bottom-left quadrant), at a specific frequency [52]. In 1996, Pendry
argued that a negative permittivity can be achieved by a periodic array of thin
metallic wires and confirmed it experimentally in 1998 [53, 54]. Subsequently, in
1999 Pendry showed that an effective negative permeability can be achieved by
using split ring resonators [55]. In 2000, Smith et al. demonstrated that on combining an array of split-ring resonators (negative permeability), with an array of
metallic thin wires (negative permittivity), a negative refractive index is achieved
√
in the microwave regime, n = − εr µr [56, 57]. Hence, such materials are also
called “double negative materials”. In this case, the phase velocity is opposite
to the flow of energy or poynting vector, and instead of forming a right handed
~ H,
~ ~k forms a left handed set of vectors. Hence, these mateset of vectors, E,
rials are also called left handed materials, which support backward propagating
waves [58]. In such materials, Snell’s law, the Doppler effect, Cherenkov radiation
etc. are completely reversed with respect to materials with a positive refractive
index. Lately, several metamaterial structures have been realized with various
types of constituents such as thin wires, swiss rolls, split-ring resonators (SRRs),
pairs of rods, pairs of crosses, fishnets etc [59].
2.1.1
Split-ring resonator
The split-ring resonator (SRR) is one of the most common and most widely used
metamaterial elements. Typically, SRRs are fabricated using highly conducting
metals and they are used to obtain a negative magnetic permeability. In Fig. 2.2(a)
we show the design of a single split ring resonator (sSRR) where d is the length
of the arm, t is the width of the arm, g is the gap width and h is the thickness of
the metal of the split-ring resonator. A SRR can be represented by an LC circuit.
The equivalent circuit for a sSRR is shown in Fig. 2.2(b). The capacitance C is
associated with the charge accumulation at the gaps and the inductance L with
the current circulating in the resonator [60]. The capacitance (C) of a SRR is
generally calculated as [61],
C = ε0 εg
ht
g
(2.1)
where ε0 is the vacuum permittivity, εg is the effective relative permittivity of
the material in the gap. εg is influenced not only by the medium inside the gap
but also by the dielectric constant of the substrate.
Chapter 2. THz near-field Faraday imaging
a)
16
Inductor Capacitor
b)
d
L
C
h
t
g
Equivalent LC circuit
sSRR
c)
d
g
t
g
dSRR
Figure 2.2: (a) Design of the single split ring resonator (sSRR) (b) Equivalent
LC circuit (c) Design of the double split ring resonator (dSRR). The incident THz
electric-field (blue) is polarized parallel to the gap. It generates current flowing in
the arms of the resonator, leading to a single magnetic “dipole” for the sSRR and
two opposite magnetic “dipoles” for the dSRR.
The inductance (L) of a SRR depends on the geometric shape of the ring. For
a planar square split-ring resonator, it is given by [62],
L = Nµ
d2
d2
= µ0
h
h
(2.2)
Here, N (number of turns in coil)=1 and µ = µr µ0 (where µr =1 for air).
Where, µ is absolute permeability, µr is relative permeability and µ0 is the permeability of free space.
The resonance frequency is given by [62],
fLC =
1
√
2π LC
(2.3)
Chapter 2. THz near-field Faraday imaging
fLC
1
=
2πd
r
17
g
1
1
=
√
t ε0 µ0 εg
2πd
r
g c0
√
t εg
(2.4)
where c0 is the velocity of light in vacuum.
The resonance wavelength is given by,
λLC
c0
√
= 2πd εg
=
fLC
r
t
.
g
(2.5)
The resonance frequency and wavelength of the SRR depend on the dimensions
of the SRR and hence can be tuned by scaling the geometrical parameters of the
SRR. When the incident electric field is parallel to the arm containing the gap
of the SRR, the electric field capacitatively couples to the SRR and generates a
current in the loop. This circulating current generates a magnetic field in the SRR.
The magnetic field, induced by the current, is strongest at the resonance frequency
of the SRR, which is determined by the loop inductance and gap capacitance as
depicted in Fig. 2.2(a) and 2.2(b).
The single SRR (sSRR) in Fig. 2.2(a) is but one of a myriad of design possibilities
for subwavelength magnetically active resonators. For example, Fig. 2.2(c) depicts
a double SRR (dSRR) which is simply two sSRRs placed back-to-back. Electric
field excitation drives counter circulating currents in this structure resulting in two
oppositely directed “magnetic dipoles”. Thus, the magnetic dipoles cancel and the
bulk effective response of an array of dSRRs is described by an effective electric
permittivity. In short, the magnetic fields associated with these magnetic “dipoles”
are of opposite sign, originate from a subwavelength area, and thus largely cancel
in the far field.
2.2
Imaging the terahertz magnetic field
When electromagnetic radiation interacts with matter, usually the magnetic field
component of light couples very weakly to the atoms compared to the electric field
component. Thus, it is very difficult to detect the effect of a magnetic field component on matter. Enhanced magnetic light-matter interaction can be achieved
by using artificial magnetic “atoms”. In metamaterials, by tailoring the geometry
of the constituent resonating structures, an enhanced response to the magnetic
field can, in principle, be achieved. Usually, when we study metamaterials, we
study their response in the far-field which gives information about the macroscopic
parameters, like effective relative permittivity and permeability. However, their
unique optical properties are derived from the near-field interactions including
magnetic near-field interactions. So, to understand the properties of metamaterials, near-field information is important.
2.2.1
Measuring the terahertz magnetic far-field
When linearly polarized light passes through a transparent magneto-optic material placed in a uniform magnetic field, the transmitted light has its polarization
Chapter 2. THz near-field Faraday imaging
18
rotated. This effect is known as the Faraday effect (see Fig. 2.3(a)). The rotation
angle is proportional to the component of the magnetic field in the propagation
direction of the probe beam. The rotation angle of the polarization of the probe
beam is θ(t) = V B(t)Lcosγ, where V is the Verdet constant of the material, B(t)
is the magnitude of the magnetic field, L is the interaction length of the THz and
optical beams inside the crystal and γ is the angle between the direction of the
magnetic field and the propagation direction of the probe beam. For maximum
rotation, γ = 0, i.e. the propagation direction of the probe beam and the direction
of the magnetic field are parallel to each other. In 1997, Riordan et al. measured
the transient magnetic field component of THz radiation using free space magnetooptic sampling [35]. The authors demonstrated that the magnetic component of
THz radiation induces a circular birefringence in the magneto-optical sensor via
the Faraday effect and when a probe beam passes through the crystal, the linear
polarization of the probe beam is rotated (see Fig. 2.3(b)).
(a)
(b)
B
β
Magneto-op"c
sensor
B THz
E
THz
beam
Plane of
polariza"on
d
Plane of polariza"on
rotated by angle β
Faraday’s rota"on
Linearly polarized
probe beam
Free space magneto-op"c sampling
Figure 2.3: (a) Faraday rotation (b) Free space magneto-optic sampling.
2.2.2
Imaging terahertz magnetic near-field
It is difficult to measure the electric and magnetic near-fields experimentally because they are strongly localized and cannot be observed using conventional, far
field, imaging techniques which cannot “see” objects smaller than half-of-a wavelength. Although, it is possible to calculate the magnetic near-field from measurements of the electric near-field [41, 48, 63], such a method amplifies noise and
inaccuracies. In addition, a calculation of the magnetic near-field requires two
measurements, namely that of Ex and Ey with equal sensitivity. At optical wavelengths, only indirect measurements of the magnetic near-field [64], the amplitude
of the magnetic field [65], or its polarization have been reported [66]. At microwave
frequencies, only a simple microstrip line has been investigated [67] while no direct
measurements have been reported at THz frequencies. In fact, for both sSRRs and
dSRRs, the near-field magnetic distribution is expected to be non-trivial. Nevertheless it is important to have magnetic near-field information because deep
subwavelength measurements of the magnetic near-field with high spatial resolution can shed light on the strength and distribution of the local magnetic field and
the near-field magnetic interaction between neighboring resonators.
Chapter 2. THz near-field Faraday imaging
2.3
19
Experimental
We directly measure the magnetic near field of SRRs that are resonant at terahertz frequencies. This is accomplished using near-field terahertz time domain
spectroscopy (THz-TDS). Our structures are deposited on terbium gallium garnet
(TGG) substrates. TGG is a magneto-optic crystal providing a linear Faraday
rotation, that is, a rotation of the plane of polarization of an optical beam is linearly proportional to the strength of an external magnetic-field pointing in the
optical beam propagation direction. TGG doesn’t show any second order electrooptic effects and it is generally used as a polarization rotator or isolator. In
combination with THz-TDS, TGG (Verdet constant of 60 radT −1 m−1 at 800 nm)
has previously been used to measure the free-space time-dependent magnetic field
component parallel to a probe beam [35]. Adapting this technique allows us to
measure the two-dimensional spatial distribution of the magnetic near-field which
strongly varies in a small region of space of only several tens of microns, about
two orders of magnitude smaller than the wavelength of the THz light.
2.3.1
Sample fabrication
The sSRR and dSRR resonators (as shown in Fig. 2.4(a) and Fig. 2.4(b)) have
been fabricated on TGG. Table 2.1 details the dimensions and simulated resonance frequencies of the SRRs. The resonance frequency given in the table is the
resonance frequency of the resonator on TGG, not the resonance frequency of a
resonator in free space. The presence of TGG lowers the resonance frequency. The
numerical simulations were performed by using the commercial software package
CST Microwave Studio. The measured and calculated resonance frequencies for
different SRRs and a dSRR are also listed in the table.
For calculating the resonance frequency of a split-ring resonator analytically,
we use the formula given in equation 2.4. As we can see from the equation, the
calculated resonance frequency depends on the geometrical parameters of a splitring resonator and also on the effective relative permittivity (εg ) of the material in
the gap, which is in air in this case. However, εg is also affected by the dielectric
constant of the substrate, which is TGG. Hence, the actual εg should be some
weighted average of the dielectric constants of the TGG substrate and air. We
have calculated the resonance frequencies of SRRs for εg = εT GG = 12.4 and for
εg = εair =1 and then we get a range in which the actual resonance frequency will
be present, which is shown in table 2.1.
Before fabrication, we first deposit a reflective coating for the 800 nm beam
consisting of 130 nm of SiO2 and 300 nm of Ge on top of a 1 mm thick (111)
TGG crystal. Standard electron-beam lithography was used for patterning the
resonators, which consists of 200 nm thick gold with a 10 nm layer of titanium
for adhesion to the Ge layer. The design of the sample is shown schematically in
Fig. 2.5(a).
Chapter 2. THz near-field Faraday imaging
20
Table 2.1: Summary of the parameters of different single split resonators and the
double split resonator simulated and measured in the near-field with the parameters
indicated in the drawings below.
Sample
Name
sSRR-1
sSRR-2
sSRR-3
dSRR-1
Arm
length
d
(µm)
90
90
70
90
Arm
width
t
(µm)
10
15
10
10
Gap
width
g
(µm)
5
15
10
5
Resonance
Frequency
Simulated
(THz)
0.155
0.185
0.224
0.170
Resonance
Frequency
Measured
(THz)
0.166
0.181
0.250
0.174
Resonance
Frequency (range)
Calculated
(THz)
0.106 – 0.375
0.150 – 0.530
0.194 – 0.682
–
d
d
d
d
Figure 2.4: (a) Schematic drawing of a sRR and (b) a dSRR
2.3.2
Results and Discussions
The schematic of the experimental setup is shown in Fig. 2.5(b). In our experiment,
a single-cycle, broadband THz pulse propagating in the ẑ-direction with an electric
field polarization in the ŷ-direction, is incident on a single resonator. The singlecycle, broadband (0 - 3 THz) THz pulse is generated using a Ti:sapphire laser
producing 15 fs pulses, which are focused on the surface of a semi-insulating GaAs
crystal biased with a 50 kHz, ±400 V square wave. A silicon hyper-hemispherical
lens is glued on the back of the crystal to collimate the emitted THz radiation.
The THz beam is then further collimated and refocused using gold-plated parabolic
mirrors [23]. The THz beam at focus covers a larger area than that of a single
resonator. At the same time, a synchronized, femtosecond probe laser pulse propagating in the (−ẑ)-direction is focused in the crystal to an approximately 5 µm
diameter spot immediately below the structure, using a reflective objective. The
Ge/SiO2 reflection coating on the crystal reflects the probe beam. Due to the
induced magnetic field of SRRs, the polarization of the probe beam experiences a
Faraday rotation. The (111) orientation of the TGG crystal ensures that the probe
pulse will only experience a Faraday rotation by a magnetic field component Hz
aligned with the propagation direction of the probe beam. This means that the
Chapter 2. THz near-field Faraday imaging
a)
21
b)
Figure 2.5: (a) Design of the sample (b) Rotation of the probe polarization due to
the magnetic near-field present inside the TGG crystal.
setup is blind to both the incident magnetic field, polarized in the x̂-direction, and
any other magnetic field in the ŷ-direction. In practice, no change in probe polarization was detected in the absence of the metallic split-ring resonators. Therefore,
we can safely assume that the probe beam polarization will be linearly rotated only
in the presence of a longitudinal magnetic field Hz inside the crystal. A differential
detector, combined with a λ/2 wave plate and a Wollaston prism measures this
rotation [68]. The instantaneous THz magnetic field is linearly proportional to the
differential detector signal.
The THz magnetic near-field as a function of time is obtained by optically &
rapidly delaying the probe pulse via the optical delay stage, with respect to the
THz pulse while measuring the Faraday rotation. This technique measures the field
and thus both the amplitude and the phase of the magnetic near-field are obtained.
Because the signal is weak, the time-dependent signal at a single position is an
average over 200000 temporal scans and was obtained in less than an hour time.
To measure the two-dimensional spatial distribution of the magnetic near-field,
the sample is raster scanned in the xy-plane. The temporal average scan number
is reduced to 10000 per pixel for the 2D-scan. A typical, 25 ps long scan of the
THz magnetic near-field was obtained by stitching two 15 ps long scans together.
2.3.3
Single point measurement
Fig. 2.4(a) shows a drawing of the sSRR patterned on the TGG crystal. The
incident electric field is polarized along the ŷ-axis, parallel to the arm containing
the gap of the resonator. The sSSR covers an area of 90 µm by 90 µm, the width
of each arm is 10 µm and the gap is 5 µm wide as shown in Table 2.1.
In Fig. 2.6(a) we plot the measured magnetic near-field Hz (t) induced by the
incident electric field at a single fixed position inside the sample sSRR-1, indicated by a cross in the insert of Fig. 2.6(a). To confirm that we measure the
Chapter 2. THz near-field Faraday imaging
a)
22
b)
1.2
0
|Hz| (arb. units)
Hz (arb. units)
E
sSRR-1
sSRR-2
sSRR-3
1
0.8
0.6
0.4
E
0.2
5
10
15
Time (ps)
c)
20
25
0
0.1
d)
0.2
0.3
0.4
Frequency (THz)
0.5
e)
max
0º
180º
0
Current Distribution
at 166 GHz
Amplitude of the Hz field
at 166 GHz
Phase of the Hz field
at 166 GHz
Figure 2.6: (a) Measurement of the time dependent out-of-plane magnetic nearfield Hz (t), induced by the electric field incident for the two different orientations
of the sSRR shown in the insets. Measurements are taken at the positions indicated
by the crosses. (b) Amplitude spectra calculated from the time-dependent magnetic field Hz (t) for the three different sSRR with dimensions given in Table 2.1.
(c) Calculated surface current density at the resonance and two dimensional spatial
distribution of the calculated d) amplitude and e) phase of Hz at the crystal surface
at the resonance frequency of 166 GHz. One can see the 180 degree phase difference
between the fields on the inside and outside of the structure.
magnetic field induced by the structure, the structure is rotated by 180 degree
around the z-axis. The incident electric field being unchanged, this should reverse
the direction of the current and thus reverse the direction of the magnetic nearfield vector. Indeed, the figure shows that the measured Hz (t) is opposite in sign
compared to the previous measurement confirming that we indeed measure the
magnetic near-field. The oscillations found in the two time traces indicate that
the structure behaves like a resonator. Time traces of the magnetic near-field of
two other sSRRs with dimensions shown in Table 2.1 have also been measured.
The spectral content of the three measured magnetic field time traces, obtained by
fast-Fourier transforming these traces, is shown in Fig. 2.6(b). Each sSRR shows a
single large peak in its frequency spectrum, which corresponds to the strong oscillations observed in the time trace of the magnetic near-field. The peak frequency
for the three different sSRRs are 0.155, 0.185, and 0.224 THz, respectively. These
Chapter 2. THz near-field Faraday imaging
23
resonances correspond to the ones found in far-field transmission measurements
of the LC response of arrays made of similar SRRs [69]. The peak frequency is a
clear function of the sSRR dimensions: the smallest resonator (sSRR-3) exhibits
the highest resonance frequency at 0.224 THz. To confirm that the measured
peak frequencies correspond to the resonance frequencies of the SRRs, we have
performed finite integration technique (FIT) simulations on these structures usR a commercial software package. The gold layer was
ing CST Microwave Studio,
taken as a perfect conductor, which is a reasonable assumption at THz frequencies.
The reflective layers were neglected. The index of refraction of the TGG crystal
at THz frequency was taken to be 3.75, equal to the value that we have measured.
The peak positions calculated by the FIT simulations are at frequencies of 0.166,
0.181, and 0.250 THz respectively, which agrees well with the experimental values. Fig. 2.6(c) shows the calculated surface current densities at the resonance
frequency (166 GHz) of the sSRR-1 sample. When the single split-ring resonator
is excited with an incident electromagnetic wave, a spatially circulating and temporally oscillating electric current is induced in the metallic ring. One can see
that the strongest current is inside the long arm of the structure, and that it is
particularly strong near the corner. This can be understood intuitively: the electrons flowing through the arm would prefer to take the shortest path, i.e. hugging
the bend, resulting in a stronger current at the inside of the corner. This current
creates a time-dependent magnetic field, which is oriented normal to that plane,
i.e. along the z-axis. It corresponds to the field component that we measure in
our magneto-optical detection setup.
2.3.4
Two dimensional distribution
We have also measured the two-dimensional spatial distribution of the magnetic
near-field Hz at the resonance frequency. These 2D measurements give information
about the distribution of the field inside and outside the ring. As we can see in
the measurement in Fig. 2.7, the field is only measurable inside the resonator and
within our measurement accuracy no field was measured at positions outside the
sSRR. The strongest field is measured in a region opposite the gap, close to the
long arm. Both observations can be understood by the fact that the current is
stronger in the long arm than in the arm containing the gap. This leads to a
stronger magnetic near-field near the long arm, on the inside of the ring. This
is supported by the calculations shown in Fig. 2.6(d) and 2.6(e) where we plot
the calculated amplitude and phase of Hz in the plane below the structure at the
resonance frequency. Interestingly, these calculations predict a 200 times stronger
magnetic field near the corner, close to the long arm, compared to the incident
magnetic field strength. The calculated 2D magnetic field distribution plotted in
Fig. 2.6(d), however, differs from the measured 2D distribution. In the calculation,
the field is strongly localized near the long arm, whereas the magnetic field has
expanded to fill the resonator in the measurement. To better understand the
discrepancy between experiment and simulation, we have calculated 2D spatial
distributions of the magnetic near-field inside the TGG crystal at four different
distances from the surface at z = 0, -10, -20 and -30 µm. These results are shown
Chapter 2. THz near-field Faraday imaging
24
194
0
13
0
µm
0
-194
28
0
13
µm
0
µm
-10
13
0
-28
16
0
0
µm
-20
13
130 µm
-16
11
0
13
0
µm
-30
z (µm)
130 µm
-11
Figure 2.7: Measured (left) and calculated (right) two-dimensional spatial distributions of the magnetic near-field Hz at the resonance frequency of sample sSRR-1:
for z = 0, -10, -20 and -30 µm. The 2D measurement agrees mostly with the
calculated spatial distributions between 10 and 20 µm below the surface.
in Fig. 2.7 along with the measurement. In both cases, the total area covered is
130 µm by 130 µm.
Although the field is mainly concentrated near the edge of the metal at the
plane z = 0, it gradually changes into an uniform distribution when the distance
z to the structure increases. One can see that the measurement resembles the
calculation for a distance between 10 to 20 µm from the surface. This shows that
we measure directly in the near-field, at a distance much smaller than the size
of the object. We reach a spatial resolution of about 10 µm, much smaller than
the 1.88 mm vacuum wavelength that corresponds to the resonance frequency of
0.16 THz. This corresponds to a value of about λ/200, more than two orders of
magnitude below the diffraction limit.
As the simulations also show, the longitudinal component of the magnetic field
amplitude decreases rapidly with distance from the surface but at 30 µm below,
the calculated magnetic field strength is still 5 times larger than the incident field
strength. At the average depth where we measure the magnetic field, the calculated
enhancement explains why we are able to measure the magnetic near-field at all,
despite the fact that we sample the field in a very small volume only.
In principle, our measurement method doesn’t sample the field at a single depth
only but, it integrates the field over the entire length of the crystal. To understand
Chapter 2. THz near-field Faraday imaging
25
0.25
a)
b)
100
0.15
130 µm
|Hz| (Arb. units)
0.2
0.1
0
0.05
0
-100
0
50
100
150
200
250
Depth (µm)
300
130 µm
Figure 2.8: (a) Magnitude of the magnetic near-field, Hz , inside the TGG crystal,
vs. distance to the structure at the surface. The line has been taken below the cross
indicated in the drawing. (b) Two dimensional spatial distribution of magnetic nearfield after integration from z = 0 to z = -300 µm inside the crystal. The magnetic
field distribution matches exactly to the distribution calculated at z = -20 µm
distance, shown in fig. 2.7
.
why we observe a magnetic field distribution at an effective depth of 10-20 microns,
we plot in Fig. 2.8(a), the calculated magnetic field component Hz as a function
of depth z inside the crystal at the location indicated by the cross in the figure.
The figure shows that as we move away from the surface the magnetic field decays
rapidly over a distance of about 30 microns and becomes negligibly small at larger
distances. The largest contribution to the signal, therefore, comes from a region of
space less than about 30 microns away from the surface. In Fig. 2.8(b), we plot the
two-dimensional distribution of the field calculated by integrating the field along
the length of the crystal at each point. Clearly, this calculation strongly resembles
both the measured distribution and the calculated one for a depth of 20 microns.
2.4
Double Split Ring Resonator
Additionally, we have performed measurements on a double split ring resonator
(dSRR). This sample, dSRR-1, is composed of two sSRRs sharing a middle arm;
the dimensions are shown in Table. 2.1 and the drawing is shown in Fig. 2.4(b).
When the THz electric field is polarized parallel to the gaps along the ŷ-axis it
generates at one moment in time, a clockwise running current in the left ring and a
counterclockwise running current in the right ring. Some time later, the situation
reverses since the currents oscillate in time. Due to the opposite directions of
the currents, we have at a moment of time a magnetic-field component pointing
down into the plane in the left ring, while in the right ring it is pointing up. This
means that we have time-dependent magnetic fields of opposite direction in a deep
Chapter 2. THz near-field Faraday imaging
26
subwavelength sized region of space, which thus more or less cancel in the far-field.
We note that only near-field measurements are capable of discerning these fields.
2.4.1
Single point measurements
Fig. 2.9(a) shows a measurement of the time-dependent magnetic near-field Hz (t)
of the dSRR structure at two different locations indicated by two crosses in the
insert of Fig. 2.9(a): one inside the left ring and the other inside the right ring.
In both measurements, the presence of long-lasting temporal oscillations indicates
that the structure has a well defined resonance. The two time traces of the magnetic near-fields are opposite in sign for the two locations. This means that the
component of the magnetic near-field Hz (t) points into the plane for the left ring,
and out of plane for the right ring. The spectrum of |Hz |, calculated from the
time-domain measurement is plotted in Fig. 2.9(b) and shows a strong peak at
0.17 THz. The resonance frequency was calculated again via FIT simulations, and
was found to be 0.174 THz, in good agreement with the experiment.
The current in the dSRR in the central arm is larger than the current in a sSRR
and distributed uniformly across its width, because the current is fed by two
identical resonators rather than just one. The presence of the magnetic field is
mainly due to this high current flowing in the middle arm, along the ŷ-direction as
shown in Fig. 2.9(c). This creates a magnetic near-field with field lines describing
roughly circles around the arm as schematically shown in the insert of Fig. 2.9(b).
2.4.2
Two dimensional distribution
The two-dimensional spatial distribution of the magnetic near-field component Hz ,
below the dSRR at the resonance frequency of 0.17 THz is shown in Fig. 2.10. The
total area covered is 140 µm by 140 µm. The measurement shows that there is
little or no field Hz outside the structure and at the location of the middle arm.
In contrast, a field Hz is present in the left and right ring and is strongest in the
area of the dSRR near the middle arm. Due to the structure of the resonator,
clockwise and anti-clockwise oscillating currents exist in the left and the right ring
and therefore the magnetic fields Hz in both resonators are opposite in direction.
The change of sign of the magnetic near-field component, from positive (red) to
negative (blue) occurs within 10-15 µm, a distance two orders of magnitude smaller
than the vacuum wavelength of 1.7 mm. This is also shown in Fig. 2.9(d) and 2.9(e)
where we plot the calculated amplitude and phase of Hz in the plane below the
dSRR structure at the resonance frequency.
As in the case of the sSRR, for DSRR also, measured 2D distribution of the
magnetic near-field doesn’t match the 2D distribution calculated at z = 0. To
understand this difference, Fig. 2.10 shows the calculated spatial distribution of
the magnetic near-field inside the crystal at various distances from the crystal
surface at z = 0 (at surface), -10, -20 and -30 µm. The calculated field distribution
at z = 0, shows inside each ring a distribution resembling the one of the sSSR,
and the field is mainly concentrated along the middle arm of the structure. As
z increases, it gradually expands and fills up the resonator. The measurement
resembles the calculation for a distance between 10 to 20 µm, confirming again
Chapter 2. THz near-field Faraday imaging
27
b)
a)
0
x
1
y
|Hz| (arb. units)
|Hz| (arb. units)
Right Side
Left Side
0.8
z
0.6
0.4
X
5
X
10
15
Time (ps)
c)
0.2
20
25
0
0.1
d)
0.2
0.3
0.4
Frequency (THz)
0.5
e)
max
0º
180º
0
Current Distribution
at 174 GHz
Amplitude of the Hz field
at 174 GHz
Phase of the Hz field
at 174 GHz
Figure 2.9: (a) Measurement of the time dependent magnetic near-field Hz at
positions indicated by the crosses in the Fig. 2.9(a), induced by the electric field
incident on the structure. The field is reversed in sign for the left and right part of
the structure as shown in the inset. (b) Associated spectrum of the out of plane
magnetic near-field (c) Calculated surface current density and (d) two dimensional
spatial distribution of the calculated amplitude and (e) phase of Hz at the crystal
surface z = 0, immediately below the structure for the resonance frequency of 174
GHz.
that we are probing the magnetic-field at an average distance of 10-20 µm from the
structure, once again confirming that we are measuring the magnetic field in the
near-field region. Moreover, as we move away from crystal surface, the magnitude
of the magnetic near-field decreases. At z = 0, the magnetic near-field Hz is 155
times stronger than the incident magnetic field, while at 30 µm distance from the
crystal surface, the strength of the calculated Hz is still 8 times stronger than the
incident magnetic field.
2.5
Conclusion
While our MM/TGG magneto-active devices have enabled direct imaging of the
magnetic field with a resolution of λ/200, numerous other possibilities are worthy
of detailed exploration. This includes further optimization of the response to
create compact devices such as dynamic Faraday isolators. In addition, SRRs
Chapter 2. THz near-field Faraday imaging
28
155
0
14
0
µm
0
-155
20
0
0
µm
14
0
µm
-10
14
-20
13
µm
0
14
140 µm
-20
0
-13
0
14
z (µm)
8
µm
-30
0
140 µm
-8
Figure 2.10: Measured (left) and calculated (right) two-dimensional spatial distributions of the magnetic near-field Hz at the resonance frequency of sample dSRR-1:
for z = 0, -10, -20 and -30 µm. The 2D measurement agrees mostly with the calculated spatial distributions between 10 and 20 µm below the surface.
provide a unique pathway to locally excite magnetic materials with well-defined
high frequency fields to interrogate, for example, magnetic field induced switching
or control of ferromagnets initiated by an applied picosecond electric field - that
is, creating dynamic magneto-electric materials. This is essentially what we have
accomplished at a basic level with our MM/TGG. It is the incident electric field
which induces the SRR magnetic dipole that, in turn, induces the TGG Faraday
rotation at near-infrared frequencies.
Finally, recent advances in generating high-field THz pulses will be of interest
for magnetic structures similar to what we have presented [70]. For example, an
incident THz pulse with a peak electric field of 1 MV/cm has a corresponding peak
magnetic field of 0.3 Tesla. A field enhancement of 200 suggested by our numerical
calculation would correspond to a local magnetic field of 60 Tesla of picosecond
duration in the plane of the SRRs, sufficient to interrogate the dynamic magnetic
properties of numerous materials.
Chapter 2. THz near-field Faraday imaging
Complementary split ring resonators
A complementary split-ring resonator (CSRR) is the “negative” of the split-ring
resonator (SRR). The CSRR can be realized by replacing the metal area of the
SRR with nothing and filling the empty area with the metal. The schematic
drawing of a CSRR is shown in the Fig. 2.11(a). Here, d and t represent the
length and width of the arm of the CSRR and g represents the width of the metal
region connecting the voids. The concept of CSRRs was introduced by Falcone
in 2004 [71]. The origin of the word “complementary” of the CSRR derives from
the fact that the electromagnetic behavior of a SRR and a CSRR are almost dual
or complementary to each other. For example, if we measure the THz electric
field transmission in the time domain, for a SRR, the transmission decreases and
shows a dip at the resonant frequency. However, for a CSRR, the transmission is
enhanced and shows a peak at the same resonance frequency [72]. Also, a SRR
shows a negative permeability whereas a CSRR shows a negative permittivity [73].
a)
d
d
t
g
Gold
TGG
b)
H
c)
E
Gold
H
Gap
THz E-field
polarisation
Gold
Bridge
Void
E
d
2.6
29
THz E-field
polarisation
Figure 2.11: (a) The schematic drawing of the CSRR describing the parameters; d
= length of the arm of the CSRR, t = width of the arm of the CSRR, g = width of
the metal region connecting the voids. (b) The schematic drawing of a SRR. The
incident THz electric field is parallel to the arm containing the gap and (c) In case
of a CSRR, the incident THz electric field is perpendicular to the empty (no gold)
arm containing the “bridge”.
One way to properly excite a SRR under perpendicular incidence is to have
Chapter 2. THz near-field Faraday imaging
30
the incident THz electric field parallel to the arm containing the gap. Since the
electromagnetic behavior of a SRR and a CSRR are complementary to each other,
the role of the electric and magnetic field is interchanged [74]. Hence, to excite
the CSRR the electric field should be perpendicular to the arm containing the
“bridge”. In Fig. 2.11(b) and 2.11(c), we show the typical directions of the incident
THz electric field for a SRR and a CSRR respectively used in our experiments.
As mentioned before, in a SRR, the incident THz electric field drives a circulating
current in the loop and we have a charge accumulation near the gap. As a result,
the electric field is enhanced near the vicinity of the gaps. In a CSRR, the incident
THz electric field drives current in the surrounding metal, and the current is the
strongest through the bridge. However, the resonant response shown by both,
the SRR and the CSRR, is purely electrical in nature. According to Babinet’s
principle, when the direction of the applied THz electric field should be rotated by
90◦ with respect to the original metamaterial or, equivalently, when the sample is
rotated by 90◦ , the complementary metamaterials should exhibit complementary
transmission properties. That means, in place of a transmision dip in original
metamaterials, we observe a transmission peak for complementary metamaterials
[75].
2.6.1
Sample fabrication
CSRRs with different dimensions are fabricated on a 1 mm thick (111) TGG
substrate. Similar to an SRR, before fabrication, a 130 nm thick SiO2 layer and
a 300 nm thick Ge layer are deposited on top of the TGG crystal as a reflective
coating for the 800 nm beam. The CSRR pattern is written onto a negative
e-beam resist using standard electron beam lithography. Subsequently, a 10 nm
thick adhesion layer of titanium and a 200 nm thick layer of gold are deposited and
then the final structure is obtained using the lift-off process. A scanning electron
microscope picture of a CSRR is shown in Fig. 2.12. The CSRR shown in the
Metal
90 μm
5 μm
10 μm
Empty
Empty
Metal
90 μm
50 μm
Figure 2.12: The scanning electron microscope image of a CSRR.
Chapter 2. THz near-field Faraday imaging
31
image covers an area of 90 µm by 90 µm. The width of each arm is 10 µm and
the width of the metal bridge is 5 µm.
2.6.2
Single point measurement
A
The electric field of the incident THz radiation drives a current in the surrounding
metal, which generates a magnetic field. In Fig. 2.13 we plot the measured magnetic near-field at two different positions, indicated by the arrows in the insert of
Fig. 2.13. The time-dependent signal at a single position obtained from a CSRR is
extremely weak, much weaker than the signal obtained from a SRR. The signal is
averaged over 200000 temporal scans and was obtained in less than an hour time.
The oscillations found in the two time traces indicate that the structure behaves
like a resonator. We see that the measured Hz (t) at both positions are opposite
in sign. This is because the direction of the magnetic near-field, generated by the
current flowing through the metal bridge region, is opposite on the two sides of
the bridge. However, the signals were too weak to measure the two-dimensional
spatial distribution of the magnetic near-field of a CSRR.
Figure 2.13: Measured magnetic near-field Hz (t), induced by the incident electric
field at two different positions of the CSRR, indicated by the arrows.
2.6.3
Two dimensional distribution
In Fig. 2.14(a) we show the calculated two-dimensional spatial distribution of the
magnetic near-field Hz of the CSRR at the resonance frequency. These calculations
give information about the distribution of the field inside and outside the “antiring”. We see that the magnetic field is strongest near the “bridge” region which
Chapter 2. THz near-field Faraday imaging
32
connects the two voids. This is because the strongest current is flowing through
that area. In Fig. 2.14(b) we show the spatial distribution of the magnetic nearfield for a SRR. In this case the applied THz field is polarized parallel to the arm
containing the gap. The magnetic near-field is maximum near the the long arm,
especially around the corners. This is because at the corner there is contribution
from “two” currents. One flowing to the corner, the other flowing away from
it at a right angle. We see that the spatial distribution of the magnetic nearfield for a SRR and for a CSRR are different but the strength of the magnetic
field distribution is similar in both cases. However, we were not able to measure
the two-dimensional spatial distribution of the magnetic near-field of the CSRR
experimentally because the signals were extremely weak.
100
Amplitude of z-component
100
Amplitude of z-component
0.2
0.2
50
0
0.1
-50
-100
-100
Y (μm)
Y (μm)
50
0
0.1
-50
-50
0
X (μm)
(a)
50
0
100
-100
-100
-50
50
0
X (μm)
0
100
(b)
Figure 2.14: The calculated 2D spatial distributions of the magnetic near-field for
a) CSRR b) SRR. For CSRR the magnetic field is strongest in the “bridge” region
and for SRR, the magnetic field is strongest in the vicinity of the long arm, in the
corner. The dimensions of SRR and CSRR are taken to be the same. The outer
dimension is 90 µm by 90 µm, width of each arm is 10 µm and gap is 5 µm.
We tried various things to understand why the signal is so weak for a CSRR
compared to a SRR, when according to the simulation results we expect signals of
similar strength. First, we fabricated the complementary structures on different
TGG crystals and measured the signal but found that there was no effect on the
strength of the signal. Then, we removed the SRR structures from the “original”
TGG crystal and made complementary structures on the same crystal but still the
signal remained small. This was done to determine that TGG crystal on which
the original SRR was made was not somehow “special”. One possible reason can
be that the current flowing through the bridge region in CSRR generates opposite
magnetic fields on two sides of the bridge. Since we investigate the field over
a certain propagation length of the probe beam, along a cone perpendicular to
the surface, perhaps the probe sees both positive and negative z-components of
the field. As a result, the magnetic fields cancel out each other partially and we
Chapter 2. THz near-field Faraday imaging
33
measure a weaker signal from a CSRR compared to a SRR. However, the exact
reason why the signal from the CSRR is much weaker, still remains a bit of a
mystry.
2.6.4
Conclusion
CSRRs are resonant structures which show complementary electromagnetic behaviour with respect to SRRs. The THz magnetic near-field of these CSRRs is
very weak and extremely difficult to measure. After averaging for a long time,
we could measure the magnetic near-field signal at a single position of the CSRR
but we were not able to measure the two-dimensional spatial distribution of the
magnetic near-field of the CSRR.
Chapter
3
3.1
THz emission from
ferromagnetic metal thin
films
Ferromagnetism
Magnetic materials play a very important role in modern day life. From fridge
magnets to data storage technology, they are used for various applications. Apart
from the applications, magnetic materials are also fascinating on a fundamental
level. These materials can be divided into five different classes based on their
magnetic behavior: diamagnetic, paramagnetic, ferromagnetic, ferrimagnetic and
antiferromagnetic. In Fig. 3.1(a) we show the magnetic behavior of different types
~ is applied to a material, which
of materials. When an external magnetic field, H,
~
induces an opposite magnetization M , the material is called diamagnetic. In gen~ and the response of
eral, the relationship between the applied magnetic field H
~
~
~
~
the material B is written as B = µ0 (H + M ) where µ0 is the magnetic perme~ is the magnetization of the medium which is defined
ability of free space and M
as the magnetic moment per unit volume [76]. Diamagnetism is a property of all
materials but other types of magnetism, like paramagnetism and ferromagnetism,
are present only in the materials with partially filled electron shells. When electron shells are completely filled, the total magnetic dipole moment of electrons is
zero but in the presence of unpaired electrons we can have a net magnetic moment. These magnetic moments are oriented in random directions but when an
external magnetic field is applied they tend to align parallel to the applied field.
This effect is called paramagnetism and such materials are called paramagnets,
e.g. aluminium, platinum, uranium etc. Unlike paramagnets, ferromagnets (like
cobalt, iron, nickel etc.) retain a component of the magnetization even after the
external magnetic field is removed. In ferromagnets the interaction between the
magnetic moments is stronger and a magnetic order is formed. As a consequence,
even in the absence of an external magnetic field, the magnetic dipoles tend to
spontaneously align parallel to each other and give rise to a non-zero magnetization. In the case of antiferromagnetic materials, the adjacent magnetic moments
are equal in magnitude but opposite to each other so there is no net magnetization
and they behave like paramagnets. For ferrimagnets, the magnetic moments of
the adjacent magnetic dipoles are opposite but not equal which results in a net
magnetic moment [77]. In this work we focus only on ferromagnetic materials, in
particular, cobalt.
Chapter 3. THz emission from ferromagnetic metal thin films
Diamagnetic
No permanent dipoles.
External magnetic field induces
an opposite magnetic field.
No external
magnetic field applied
36
With external
magnetic field applied
Paramagnetic
Ferromagnetic
B
Ferrimagnetic
Antiferromagnetic
Figure 3.1: (a) Different types of magnetic behaviors shown with the help of magnetic moments represented by arrows. (b) Magnetic domains without any external
magnetic field applied and with an external magnetic field applied
Origin of ferromagnetism: In ferromagnetic materials the magnetic moments
are strongly coupled to each other. This strong coupling between the magnetic
moments is a consequence of a particular quantum mechanical effect known as
exchange interaction, a description of which is beyond the scope of this thesis.
The exchange interaction between neighboring ions forces the individual moments
into parallel alignment whereas the magnetic dipole-dipole interaction tends to
orient the magnetic dipoles to be opposite to each other. In case of magnetic
materials, the exchange interaction is much stronger than the magnetic dipoledipole interaction, and as a result the dipoles tend to align in the same direction
[78].
Curie temperature: All the ferromagnetic materials have a characteristic temperature above which they lose their ferromagnetic property. This temperature is
called the Curie temperature Tc . Above this temperature, the thermal agitation
is sufficient to overcome the magnetic order present and the ferromagnet starts
behaving like a paramagnet. At room temperature, the exchange energy of ferromagnetic materials is much greater than the thermal energy but when the thermal
energy exceeds the exchange energy, the coupling breaks down and the ferromagnetic property disappears. The Curie temperature depends on the material, and
for cobalt, iron and nickel, Tc is 1400, 1040 and 630K respectively [79].
Magnetic domains: Domains are small regions in the ferromagnetic materials,
within which the magnetic dipoles are aligned parallel to each other. In ferromagnetic materials, the exchange interaction is large so the magnetic dipole moments
tend to align in the same direction. Ideally, therefore, ferromagnetic materials
should have a single domain and the magnetization should be in the same direction throughout the sample. However, a ferromagnetic sample having a single
domain has a macroscopic magnetization. This magnetization generates a magnetic field around the sample which is opposite to its own magnetization. This
field tends to demagnetize the material and hence it is called a demagnetizing field.
Since it is apparently energetically advantageous to reduce the magnetic energy
Chapter 3. THz emission from ferromagnetic metal thin films
37
of the system, the magnetization is split into several domains. In the absence of
any magnetic field, the magnetization vectors in different domains are oriented in
different directions, so the net magnetic field is small. Ferromagnetic domains are
separated by thin walls, in which the direction of magnetization rotates from one
domain’s direction to the other domain’s direction. These boundaries are called
domain walls or Bloch Walls [80]. In Fig. 3.1(b) we show a schematic of magnetic domains and domain walls inside a ferromagnet. When no external magnetic
field is present, magnetic dipoles are oriented in random directions but when a
magnetic field is applied the magnetic dipoles are aligned in the direction of the
applied magnetic field.
The state of aligned domains is extremely stable but the magnetization can be
destroyed if the magnets are subjected to a magnetic field, stress or heat. The
simplest way to affect a magnet is to apply an external magnetic field. If the
externally applied magnetic field is sufficiently strong and opposite in polarity
then the magnet gets demagnetized. The required field is known as a coercive
field and it depends on the material. Another way is to provide stress to the
magnets, since the vibration induced in the magnet can randomize the magnetic
dipoles and destroy the magnetic ordering. One other possible situation is when
a magnet is heated beyond its Curie temperature, then the domains are oriented
in random directions and the material comes into a demagnetized state.
3.2
Laser-induced ultrafast demagnetization
When a femtosecond laser pulse is incident on the ferromagnetic sample, a demagnetization may take place on the femtosecond timescale. To study this ultrafast
demagnetization, typically a pump-probe technique is used, in which a fs laser
pulse is split into two parts: an intense pump pulse and a weaker probe pulse
with a delay. The pump pulse excites the sample and perturbs the magnetization
whereas the probe pulse detects the change in the magnetization. Most commonly, a change in the intensity of the reflected probe pulse or a change in the
polarization of the probe pulse is measured to probe the magnetization of the sample [81]. The ultrafast demagnetization has attracted a lot of attention because
it has potential applications in magnetic data storage technology and, recently,
because of evidence that changes in the magnetization give rise to the emission of
a terahertz electromagnetic pulse [82, 83]. From a fundamental perspective it is
very interesting, even though the phenomenon of ultrafast demagnetization is not
understood completely. Below, we describe briefly the work related to ultrafast
demagnetization, as reported previously in the literature.
3.2.1
Historical review
In the past few decades a lot of research has been done in the field of laser-induced
ultrafast demagnetization. The first experiment was performed by Agranat et al.
in 1984, when he demonstrated that the interaction of picosecond laser pulses
with nickel thin films does not demagnetize the film, even if the temperature
Chapter 3. THz emission from ferromagnetic metal thin films
38
reaches double the Curie temperature [84]. Similar results were initially shown by
Vaterlaus et al. for iron thin films but a few years later he performed a pioneering experiment with Gadolinium thin films and concluded that demagnetization
takes place on the time scale of tens of picoseconds or slower [85, 86]. In 1996,
Beaurepaire et al. measured the magnetization as a function of time using the
magneto-optical Kerr effect (MOKE). They observed that when 60 fs laser pulses
are incident on a 22 nm thick nickel film, the magnetization drops by 50 percent
within the first picosecond [87]. It was a real breakthrough, and the demagnetization on a sub-picosecond timescale strongly increased the interest in the field of
ultrafast demagnetization. These results were soon confirmed by other research
groups by different experimental methods. Hohlfeld et al. used second harmonic
measurements and Scholl et al. used two photon photoemission (2PPE) experiments for obtaining the magnetization as a function of time [88, 89]. In Hohlfeld’s
experiment, a femtosecond laser pulse excites the sample. By measuring the sum
and difference of the reflected SHG signals for opposite magnetization directions
the change in magnetization is obtained. In 2PPE experiment, first the pump pulse
heats the sample and then after a delay, electrons are photoemitted by the probe
pulse. The spin-polarization of these photo-excited electrons is measured using a
Mott detector. The polarization of the photoelectrons depends on the temperature
and, hence, upon heating, there is a change in the spin polarization which provides
information about the change in the magnetization. In 2004, it was shown that
in nickel the loss of magnetization can result in emission of terahertz (THz) radiation [83]. In 2007, Stamm et al. pumped nickel thin films using a fs laser pulse
and then studied the demagnetization using a 100 fs X-ray pulse as a probe. Using
this technique it was possible to separate the contributions from the spin and the
orbital angular momentum [90]. However, till now, the mechanism behind ultrafast laser induced demagnetization has not been understood completely. Below,
we discuss some of the phenomenological and microscopic models that have been
used by other research groups to describe the ultrafast demagnetization process.
Phenomenological model: To understand the process of ultrafast demagnetization, Beaurepaire et al. proposed a phenomenological model which is based on
the experimental observations without assuming any particular microscopic mechanism [87]. The authors extended the two temperature model (2TM) given by
Ali et al. for characterizing electron dynamics in normal metals [91]. In the two
temperature model, electrons and phonons are considered as two different heat
baths which are coupled so they can exchange energy. The pump laser pulse is
absorbed only by the electron bath and hence its temperature is increased. Then,
these electrons thermalize through electron-electron interaction and then transfer
the energy to the lattice by electron-phonon interaction. To explain the results
obtained from laser-induced ultrafast demagnetization in ferromagnetic metals,
Beaurepaire et al. added a spin bath and proposed a three temperature model
(3TM) [83]. According to this model, there are three different heat baths, as
shown in Fig. 3.2: The electron bath at temperature Te , the lattice bath at temperature Tl and the spin bath at the temperature Ts . In this case also, the initial
heat provided by the laser is absorbed only by the electron bath. The excited
Chapter 3. THz emission from ferromagnetic metal thin films
39
electrons interact with the lattice and deposit heat into the lattice system. Subsequently, this heat is transferred to the spin system bringing spins to an equilibrium
temperature with the electrons and the lattice. The increase in temperature of
the spin bath results in a reduction of the magnetization. It was observed that the
electron-lattice and electron-spin interactions occur on the picosecond timescale
while electron-electron interactions take place on the femtosecond timescale. The
redistribution of the laser power or heat absorbed by the electrons among the
three systems (electron, spin and lattice) is depicted by three different differential
equations:
Ce (Te ) (dTe /dt) = - Gel (Te -Tl ) - Ges (Te -Ts ) + P(t),
Cs (Ts ) (dTs /dt) = - Ges (Ts -Te ) - Gsl (Ts -Tl ),
Cl (Tl ) (dTl /dt) = - Gel (Tl -Te ) - Gsl (Tl -Ts ),
Te
TL
eelectrons
Lattice
Ts
Figure 3.2: Three temperature model
Here Ce is the specific heat of the electronic bath, Cl is the specific heat of
the lattice bath and Cs is the specific heat of the spin bath. P(t) is the initial
excitation provided by the laser source. Since the laser is initially absorbed only
by the electron bath, it is added only to the first equation. Gel , Ges and Gsl
are the electron-lattice, electron-spin and spin-lattice interaction constants, which
describe the rate of energy exchange between the respective baths. The three
temperature model gives a good intuitive physical understanding but does not take
care of angular momentum conservation. However, many microscopic mechanisms
have been proposed which take spin angular momentum conservation into account,
which will briefly be discussed below.
Chapter 3. THz emission from ferromagnetic metal thin films
40
Microscopic model: During the process of demagnetization, the spin moments are misaligned and the total spin angular momentum of the system is decreased. The angular momentum cannot be destroyed so it has to be transferred
from the spin system. The biggest challenge is to discover through which channel
this spin angular momentum is getting transferred and how fast this transfer is. In
2000, Zhang et al. investigated the ultrafast demagnetization and proposed that
the demagnetization happens because of the combined action of spin-orbit coupling
and the external laser field [92]. The required angular momentum is transferred
directly from the laser photons to the spins. In that case, a linearly or circularly
polarized pump pulse should have a different effect on the demagnetization but
Dalla Longa et al. observed that the polarization of light has a negligible effect on
the timescale of demagnetization [93]. Also, according to this model the demagnetization time is limited by the pulse duration of the laser but later it was shown that
this is not the case. In 2005, Koopman proposed a microscopic model for ultrafast
demagnetization [94]. He proposed that the spin angular momentum is transferred
to the lattice and that the demagnetization is mediated by the phonon scattering.
Koopman’s model relates the demagnetization time to the Gilbert damping, but
Radu et al. changed the Gilbert damping parameter by doping and showed that it
does not modify the demagnetization constant [95]. Recently Battiato et al. presented a microscopic model for femtosecond laser-induced demagnetization [96].
They argue that in ferromagnetic metals, where the number of spin-up and spindown electrons is different, majority spin electrons (conventionally the majority
spin electrons are referred to as ”spin up electrons” and the minority spin electrons as ”spin down electrons”) have a longer mean free path. This may lead to a
depletion of majority carriers in the magnetic film and a transfer of magnetization
away from the surface. As a result, a spin current is established leading to the
transfer of magnetization away from the surface. Subsequently, Eschenlohr et al.
also supported the idea of superdiffusive spin transport as the mechanism responsible for the ultrafast demagnetization [97]. In this work the authors deposited a
thin layer of gold on top of a ferromagnetic thin film in such a way that only a
small part of the incident laser energy is able to reach the ferromagnetic film. Surprisingly, they observe that the efficiency of the ultrafast demagnetization remains
the same. This indicates that for ultrafast demagnetization it is not necessary to
directly illuminate the ferromagnetic metal but that it can also be achieved by the
transport of electrons excited by the laser elsewhere. Yet again Schellekens et al.
showed that for ultrafast demagnetization, superdiffusive spin transport may have
an effect but it is not the dominating one [98].
Clearly, the microscopic mechanism behind ultrafast demagnetization and behind the generation of THz light in these ferromagnetic films is not yet completely
understood. Determining the microscopic mechanism behind ultrafast demagnetization is an interesting topic by itself, but it is not the topic of this chapter,
which focuses on the generation of THz light by ultrafast laser excitation of ferromagnetic thin films. However, it may be possible that THz emission may, in
the near future, turn out to be a useful tool to study demagnetization, provided
an unambiguous connection between THz emission and ultrafast demagnetization
can be made. Confirming this connection is the main goal of the work described
Chapter 3. THz emission from ferromagnetic metal thin films
41
in this chapter.
3.3
THz emission from non-magnetic metal thin films
The most common way of generating terahertz (THz) radiation is by illuminating
non-linear optical materials and semiconductors with ultrashort laser pulses. The
emission of THz radiation from non-linear crystals and semiconductors is due to
laser-induced changes in current and/or a polarization in the sample [99]. THz
radiation can be also generated by irradiating thin metal films after illumination
with femtosecond laser pulses [100]. In 2004 Kadlec et al. showed emission of THz
pulses from gold thin films after illumination with femtosecond laser pulses [101].
But they could not detect THz emission from metal films thinner than 100 nm.
In 2012, Ramakrishnan et al. showed that it is possible to generate THz radiation
from continuous gold films thinner than 100 nm in the Kretschmann geometry by
exciting surface plasmons [102]. THz emission from flat metal thin films was also
shown by Suvorov et al., who used a bolometer for detecting the emitted THz radiation [103]. Apart from flat metal thin films, many other research groups showed
THz generation from structured metal surfaces. In 2007, Welsh et al. showed THz
emission from gold-coated glass nanogratings by exciting surface plasmons [104].
In 2011 Ramakrishnan and Planken showed emission of THz pulses from percolated gold thin films [105]. Subsequently, Polyushkin et al. showed THz emission
from arrays of silver and gold nanoparticles, and from percolating silver films [106].
In 2012 Kajikawa et al. showed that gold nanospheres on top of a gold surface
emit THz pulses when they are illuminated with femtosecond laser pulses [107]. In
all these cases, second and higher-order nonlinear optical processes were responsible for the THz emission from non-magnetic metal thin films. At low intensity,
second-order optical rectification is the dominating generation mechanism whereas
at higher intensities the acceleration of photoexcited electrons by the pondermotive
forces is responsible for the THz emission [100].
3.4
THz emission from ferromagnetic metal thin films
More recently, there has been a lot of interest in THz generation from ferromagnetic
metal thin films. The interest in ferromagnetic thin films is motivated by the
fact that such films have potential applications in magnetic data storage. As
mentioned before, Beaurepaire et al. were the first to show that laser-induced
ultrafast demagnetization of ferromagnetic nickel films results in the emission of
THz electromagnetic pulses [83]. Far away from the sample, the radiated electric
field E(t) was assumed to be proportional to the second time derivative of the
magnetization (d2 M/dt2 ). In their research, the authors suggested that thermal
effects are responsible for the ultrafast demagnetization. When laser pulses are
incident on the sample, the temperature in the pump spot increases and as soon
as it reaches the Curie temperature, demagnetization starts immediately. In 2004,
Hilton et al. showed terahertz emission from iron thin films due to a second-order
magnetic nonlinearity. Without external magnetic field present, the ferromagnetic
sample is multi-domain. However, when averaged over these multiple magnetic
Chapter 3. THz emission from ferromagnetic metal thin films
42
domains, it still has a non-negligible net magnetization. In 2012, Shen et al.
also reported on terahertz emission from Ni-Fe alloy thin films through ultrafast
demagnetization [108]. A complicating and, perhaps, underestimated factor is that
non-magnetic metal films are known to emit THz radiation upon illumination
with a femtosecond laser too (as mentioned above), and that ”non-magnetic”
contributions to the THz emission from ferromagnetic films cannot a priori be
excluded. More experiments are thus essential to provide further information on
the origin of THz emission from these materials.
Here, we show measurements of THz emission from cobalt (Co) thin films, illuminated with femtosecond laser pulses from a Ti:sapphire oscillator. We find that
for Co thicknesses smaller than about 40 nm, the THz electric field polarization
rotates when the sample is rotated around the surface normal. As a result, when
only the p- or s-polarized field component is detected, the THz field changes sign
every 180◦ when the sample is rotated. Such behavior is typically absent in experiments on femtosecond laser induced THz emission from non-magnetic metals, and
suggests ultrafast changes in an in-plane magnetization as the source of the THz
emission for these thicknesses. For increasing thicknesses, however, an additional,
azimuthal angle-independent, contribution to the signal is observed which grows
in size with respect to the angle-dependent contribution. This angle-independent
contribution is attributed to ultrafast changes in an out-of-plane magnetization
which emits only p-polarized THz light and which dominates the emission for
thicknesses larger than about 175 nm. This is consistent with magnetic force
microscopy (MFM) measurements on these samples that show that for low thicknesses, the magnetization is predominantly in-plane, whereas for larger thicknesses
the magnetization acquires a strong out-of-plane character. Our results show that
for cobalt thin films, the emission of THz light is strongly correlated with the
magnetization dynamics.
3.5
3.5.1
Experimental
Sample fabrication
Cobalt thin films of different thicknesses (2 nm - 140 nm) were deposited on glass
substrates by electron beam evaporation at a rate of 1 Å/s. The thickness of the
evaporated thin films was measured using a quartz crystal resonator positioned
inside the evaporation chamber. The crystallinity of the deposited cobalt thin
films was investigated by standard X-ray diffraction (XRD) measurements. Fig. 3.3
shows the experimental XRD data for a 100 nm thick cobalt film deposited on the
glass substrate. The large bump in the XRD data around 2θ = 30◦ is due to the
amorphous nature of the glass substrate while the presence of several sharp peaks
shows that the cobalt film is polycrystalline. The origin of the small peak near 2θ
= 49◦ is ambiguous. It is either due to the neighboring hexagonal (100) plane of
Co or due to the presence of cobalt oxide (CoO).
Chapter 3. THz emission from ferromagnetic metal thin films
43
Due to amorphous
glass substrate
?
Figure 3.3: XRD measurement of a 100 nm thick cobalt film deposited on the glass
substrate.
3.5.2
THz generation and detection setup
The experimental setup used for our measurements is schematically shown in
Fig. 3.4(a). We have used a Ti:sapphire oscillator which generates p-polarized
light pulses of 50 fs duration. These pulses are centered at a wavelength of 800 nm
with an average power of 800 mW. The laser beam is split into a pump beam and
a probe beam by a 80:20 beam splitter. When the pump beam is incident on the
sample at a 45◦ angle of incidence, THz radiation is generated. The generated
THz pulses are collected and focused using parabolic mirrors onto a 0.5 mm thick
zinc telluride (ZnTe) (110) detection crystal. The probe beam is also focused on
the detection crystal. The instantaneous electric field of the THz radiation induces
THz emission at 45° angle of incidence
Beam
splitter
Pump
Beam
THz emission at 0° angle of incidence
(Back-reflection setup)
Beam
Splitter
Probe
Beam
Detection
Crystal
THz
Beam
Pump
Beam
Probe
Beam
Detection
Crystal
THz
Beam
45°
Mirror
Cobalt thin film
deposited on glass
Parabolic
Mirrors
(a)
Parabolic
Mirrors
Cobalt
thin film
(b)
Figure 3.4: Experimental setup for the generation and detection of THz pulses.
The pump beam is incident on the sample at (a) a 45◦ degree angle of incidence
and (b) a 0◦ angle of incidence.
Chapter 3. THz emission from ferromagnetic metal thin films
44
birefringence in the detection crystal. When the probe beam passes through the
detection crystal, it is modified from a linearly polarized to an elliptically polarized
beam. The amount of ellipticity is proportional to the instantaneous THz electric
field. The probe beam then passes through a quarter waveplate and a Wollaston
prism and is finally focused on the photo-diodes of a differential detector. By
varying the delay between the pump and the probe pulse we obtain a full 20 ps
long THz electric field time-trace. In Fig. 3.4(b) we show the setup used for THz
emission and detection at a 0◦ angle of incidence. In this case the pump beam
passes through a hole in the parabolic mirror and is incident on the sample at zero
degree angle of incidence. The THz radiation emitted from the sample is collected
in the back-propagating direction.
3.5.3
Magnetic force microscopy
Magnetic force microscopy (MFM) is used to determine the size of magnetic domains having an out-of-plane magnetization. In MFM there is a sharp tip, which
is coated with a ferromagnetic thin film and magnetized along the axis of the tip.
During MFM measurements, this tip first scans the surface and gathers the topographical information over a sample area. During a second scan, in which the tip
is slightly raised, this information is used to maintain a locally constant separation between the tip and the sample throughout the image. This second scan then
only measures the long-range magnetic interactions [109]. In Fig. 3.5 we show a
schematic diagram of the principal of operation of a magnetic force microscope.
MFM is not sensitive to the in-plane magnetization but only to a magnetization
perpendicular to the plane. MFM can be used for investigating the domains and
domain walls of magnetic thin films, nanoparticles and nanostructures [110].
First scan
Lift mode scan
Cantilever
Magnetic
tip
Lift Height
Surface
Surface
Figure 3.5: Schematic diagram of working of a magnetic force microscope. During
the first scan the tip scans the surface of the sample and then lifts up and scans
again, this time following the topography of the sample to gather the magnetic force
images.
Chapter 3. THz emission from ferromagnetic metal thin films
3.6
3.6.1
45
Result and discussions
THz emission from cobalt thin film
In Fig. 3.6(a) we plot the THz electric field as a function of time, emitted from a
100 nm thick cobalt film deposited on a glass substrate, illuminated with femtosecond laser pulses at a 45◦ angle of incidence (see Fig. 3.4(a)). The amplitude of
the emitted THz radiation is fairly weak and is roughly 0.4% of the THz emission
from a conventional semi-insulating GaAs (100) surface depletion field emitter.
The emitted THz amplitude increases linearly with the laser power incident on
the sample as shown in Fig. 3.6(b). This suggests that a second-order non-linear
process is responsible for the THz emission. The red solid line is a guide to the
eye. In Fig. 3.6(c) we plot the emitted THz amplitude and the optical absorption
as a function of pump-beam polarization angle for a 40 nm thick cobalt film. A
0◦ angle corresponds to a p-polarized pump beam, while a 90◦ angle corresponds
to an s-polarized beam. We attribute the dependence on the polarization angle to
changes in the efficiency with which the pump light is coupled into the film.
(a)
(b)
(c)
Figure 3.6: (a) Measured THz electric field vs. time, emitted from a 100 nm thick
cobalt film deposited on the glass substrate (b) Pump power dependence of THz
emission from a 100 nm thick cobalt film. (c) The measured percentage of absorbed
pump power (blue) and the electric field amplitude of the THz pulses emitted (red)
from a 100 nm thick cobalt film, as a function of pump beam polarization.
Chapter 3. THz emission from ferromagnetic metal thin films
46
To investigate the relation between the THz emission and the magnetic order of
the thin film, we applied an external magnetic field using permanent magnets, as
shown in Fig. 3.7(a). We observe that upon reversal of the applied magnetic field
direction, the THz electric field is also reversed, as shown in Fig. 3.7(b), where we
plot the THz electric field amplitude as a function of time for a film thickness of
40 nm. Since the sample magnetization is reversed when we flip the direction of
the applied magnetic field. This suggests that there is a strong connection between
the magnetic order and the polarity of the THz pulse.
N
N
fs laser
pulse
N
S
S
N
S
S
cobalt
thin film
magnets
(a)
(b)
Figure 3.7: (a) Schematic detail of the setup used to apply an external magnetic
field to cobalt thin films. (b) Measured THz emission from a 40 nm thick Co film
on glass, as a function of time. Black and red traces indicate THz emission with
magnetic fields applied in opposite directions.
3.6.2
Azimuthal angle dependence
In Fig. 3.8(a) we plot the p-polarized THz amplitude emitted by a 40 nm thick
Co film, as a function of sample azimuthal angle. The pump beam incident angle
is 45◦ and there is no externally applied magnetic field. The figure shows that the
amplitude of the THz signal shows a sinusoidal-like dependence on azimuthal angle.
The generated THz electric field flips sign when the sample is rotated by 180◦ . This
kind of dependence is typically not observed in experiments on THz emission from
non-ferromagnetic metals. This suggests that an in-plane magnetization change is
responsible for the THz emission. We performed the same experiment with cobalt
films of different thicknesses, in the range of 10 nm - 250 nm. For cobalt films
with thicknesses less than 90 nm, the azimuthal angle dependence is similar to that
measured for the 40 nm thick sample, meaning that the polarity of the THz signal
changes when the sample is rotated by 180◦ , as shown in Figs. 3.8(b) and 3.8(c).
For 90-150 nm thick cobalt films, upon rotation, the THz signal changes polarity
but an apparent positive offset reduces the angular range in which the signal
changes sign, as shown in Figs. 3.8(d) and 3.8(e). When the thickness of the cobalt
thin film is greater than 175 nm, an azimuthal angle dependence is still observed
but there is no sign change anymore. An example of this is shown in Fig. 3.8(f)
where we plot the change in THz amplitude as a function of azimuthal angle, for
Chapter 3. THz emission from ferromagnetic metal thin films
47
a 250 nm thick cobalt film. As we will argue below, this suggests that for thick
cobalt films, along with an in-plane magnetization, an additional perpendicular
magnetization component is also present. Since the in-plane magnetization is no
longer the dominant component, when we rotate the sample, the transient does
not change sign and remains positive.
(a)
(b)
(c)
(d)
(e)
(f )
Figure 3.8: The azimuthal angle dependence of the measured THz electric-field
amplitude emitted by a (a) 40 nm (b) 60 nm (c) 80 nm (d) 125 nm (e) 150 nm
and (f) 250 nm thick cobalt film.
Note that for the thinnest sample, 40 nm, it seems that the range of azimuthal
angles where the amplitude is negative, is smaller than for the next, 60 nm thick,
sample. We are not really sure about the reason for this, but perhaps this may be
explained if we can assume that, in the 40 nm case, the location where the laser
beam hits the sample has a small offset with respect to the rotation axis. If this is
true, then upon rotation of the sample, the laser would not hit exactly the same
spot all the time. Instead it would, to a certain extent, also hit different parts of
the sample which may give rise to somewhat different THz emission amplitudes
and, thus, somewhat different looking azimuthal angle dependence.
Relation between the components of the THz electric field: If changes
in the in-plane magnetization are responsible for the THz emission, then rotating
the sample around the surface normal should also rotate the THz field polarization.
In general, when the sample magnetization is in-plane, both, s- and p-polarized
THz light should be present. In Fig. 3.9 we plot the measured azimuthal angle
dependence for both p-polarized and s-polarized components in the generated THz
pulse for a 30 nm thick Co film. The pump beam is p-polarized and is incident at
a 45◦ angle. The figure shows that when we rotate the sample, both components
change in such a way that when the amplitude of the p-polarized THz emission
is maximum, the amplitude of the s-polarized THz emission is zero and vice-
Chapter 3. THz emission from ferromagnetic metal thin films
48
versa. This supports the assumption that changes in the in-plane magnetization
are responsible for the THz emission from this sample.
Figure 3.9: Measured p-polarized and s-polarized component of the emitted THz
electric field amplitude for a 30 nm thick cobalt film.
At the same time, for the samples with the magnetization in the perpendicular
direction only s-polarized THz emission should be observed. However, for the
250 nm thick Co film, the results are different, as already shown in Fig. 3.8(f).
Here only p-polarized THz light is observed and this shows only a weak azimuthal
angle dependence. No measurable s-polarized THz emission was observed from the
250 nm thick cobalt film. For the 250 nm thick cobalt sample, the weak azimuthalangle dependence observed in the p-polarized THz emission shown in Fig. 3.8(f)
indicates that a contribution from a weak in-plane magnetization is present. The
apparent azimuthal-angle independent contribution to this signal (the offset) is
most probably caused by an electric-dipole contribution, similar to what is seen
in non-magnetic metals. This highlights that the emission of terahertz radiation
from such ferromagnetic thin films is not necessarily only due to the changes in
the magnetization but can also have an electric contribution. Probably, a weak
s-polarized component is present but presumably too weak to measure in view of
the small signal to noise ratio observed for this film thickness.
3.6.3
THz emission in back reflection
In Fig. 3.10 we plot the THz electric field measured in the backreflected direction
as a function of time from cobalt thin films, detected at a 0◦ angle of incidence
for two different film thicknesses, 40 nm and 250 nm. We observe that for the
40 nm thick cobalt film, THz emission is detected in the back-reflected direction,
but when the 250 nm thick cobalt film is illuminated, we do not detect any THz
emission. For the 40 nm thick cobalt film, the magnetization is assumed to be
in plane and so THz emission in the back-reflected direction would be allowed.
However, for the 250 nm thick cobalt film, the magnetization of the film is in
the perpendicular direction and since an oscillating magnetic dipole oriented in
the normal direction cannot emit an electric field in the same direction, we don’t
Chapter 3. THz emission from ferromagnetic metal thin films
49
Figure 3.10: Measured THz electric field vs. time from 40 nm (black) and 250 nm
(red) thick cobalt films at a 0◦ angle of incidence.
detect any THz emission in the back-reflected direction. However, we note that
in the zero degree angle configuration a part of the pump beam is blocked while
passing through the hole in the parabola, which results in smaller THz emission
signals. This makes it more difficult to detect weak backreflected THz emission
from a small in-plane magnetization, if present.
3.6.4
Thickness dependent THz emission
In Fig. 4.12(a) we plot the reflected, transmitted and absorbed pump power as
a function of the thickness of the cobalt thin films deposited on glass substrates.
The absorbed pump power is obtained by measuring the incident, reflected and
transmitted pump power from the samples. We see that the absorption of the
cobalt thin films increases as we increase the thickness but becomes eventually
constant for thicknesses larger than about 20 nm. In Fig 4.12(b), we show the
amplitude of the THz radiation emitted at a 45◦ angle of incidence as a function
of the thickness of the cobalt thin films. For each film, the sample is rotated to
find the maximum THz amplitude emitted by the film. We observe that there
is no direct correlation between the absorbed pump power and the emitted THz
radiation. When we increase the thickness of the cobalt thin film, THz emission initially increases with increasing thickness, peaks around 40 nm and then
decreases. We propose that the in-plane magnetization component increases as
we increase the thickness of the cobalt film, but around 40 nm a perpendicular
magnetization component develops. As we further increase the thickness of the
cobalt film, this perpendicular component of the magnetization grows bigger and
more dominant. If we can assume that THz radiation emitted due to changes in
the perpendicular magnetization component and that emitted by changes in the
in-plane magnetization component are opposite in phase, then these contributions
partially cancel each other. Consequently, with increasing thickness, as the outof-plane magnetization component becomes stronger, the emitted THz amplitude
becomes smaller.
Chapter 3. THz emission from ferromagnetic metal thin films
50
Figure 3.11: (a) Percentage reflection, transmission and absorption of the pump
laser pulses by different thicknesses of cobalt thin films. (b) p-polarized THz emission as a function of thickness of cobalt film deposited on the glass substrate.
3.6.5
MFM measurements
In order to confirm our assumption that changes in the orientation of the magnetic domains are responsible for the observed changes in THz emission as we
increase the sample thickness, cobalt thin films were studied using magnetic force
microscopy (MFM). Polycrystalline cobalt thin films typically possess an in-plane
magnetization, but above a critical thickness a perpendicular magnetization component is expected [111]. This cross-over thickness for cobalt thin films is around
40 nm [112]. For thicker samples, the perpendicular magnetization component
gives rise to a stripe domain pattern with alternating dark and bright contrast
indicating domains pointing up and down. Saito et al. have demonstrated such
domains for thin Ni-Fe films [113].
In Fig. 3.12 we show the MFM measurements of cobalt thin films with different
thicknesses deposited on the glass substrate. We see that for thin cobalt films, with
thicknesses less than 40 nm, there are no stripe domains present but the stripe
pattern appears for thicker cobalt films, showing the presence of a perpendicular
magnetization component. We also observe that the width of the stripes is a
function of film thickness and for much thicker films (>200 nm) they grow wider.
Based on these measurements we can confirm that the contrasting azimuthal angle
dependence behavior shown by films with different thicknesses arises due to a
change from a predominantly in-plane magnetization to a predominantly out-ofplane magnetization for increasing thickness.
It is actually surprising that a net magnetization is present in our samples. If the
orientation of many domains within the spotsize of the laser were truly random,
no net magnetization would be present. In principle, a net magnetization can be
induced by the atomic order of the underlying substrate but that is less likely in
our case for which all substrates are made of glass. However, a net magnetization
can also result from a nonzero net angle of incidence of cobalt atoms impinging on
the substrate during deposition [114]. A net magnetization can also result from
Chapter 3. THz emission from ferromagnetic metal thin films
51
Figure 3.12: Magnetic force microscope images of cobalt thin films on glass with
different thicknesses. No domains are observed for thin cobalt films; domains start
appearing when the thickness of the film crosses the critical thickness (40 nm). For
thicker cobalt films, the width of the domains increases as we increase the thickness
of the film.
stray magnetic fields, for example, when the cobalt film is deposited on the substrate using electron beam deposition in which a magnetic field is used to direct
an electron beam to the cobalt target to melt it. We note that the MFM measurements for cobalt films with thicknesses greater than 40 nm give the impression
of some order in the way the out-of-plane magnetic domains are organized. It
is unclear whether this also implies that there is a net magnetization. Although
the exact origin of the net magnetization is currently not well understood, the
purpose of the current work is to discover whether changes in this magnetization
are responsible for the THz emission. Similar conclusions were drawn by Hilton
et al. who studied iron thin films [114].
Our results can be summarized by the four figures in Fig. 3.13. We show that
when the femtosecond laser pulses are incident on the thin cobalt films with thicknesses less than 40 nm, i.e. when the magnetization is in-plane, THz emission is
observed at both a 45◦ angle of incidence and a 0◦ angle of incidence, as shown in
Fig. 3.13(a) and Fig. 3.13(b) respectively. For thicknesses in the range of 40 nm
- 175 nm, in addition to an in-plane magnetization component, a perpendicular
magnetization component is also present. The perpendicular magnetization component for these samples is relatively weak and the net magnetization mostly lies
in the plane of the sample. Hence, the THz emission from these cobalt films is observed at both angles of incidence. When femtosecond laser pulses are incident at
a 45◦ angle of incidence on the cobalt films thicker than 175 nm, with the sample
magnetization predominantly perpendicular to the plane, a relatively weak THz
Chapter 3. THz emission from ferromagnetic metal thin films
52
THz pulse
THz pulse
fs laser
pulse
fs laser
pulse
45°
0° angle of
incidence
thickness < 40 nm
thickness < 40 nm
(a)
(b)
No THz emission
fs laser
pulse
THz pulse
45°
fs laser
pulse
0° angle of
incidence
thickness >175 nm
thickness >175 nm
(c)
(d)
Figure 3.13: Schematic overview of our results. (a) THz generation from thin
cobalt films at a 45◦ angle of incidence (b) THz generation from thin cobalt films
at a 0◦ angle of incidence (c) THz generation from thick cobalt films at a 45◦ angle
of incidence (d) Absence of THz emission from thick cobalt films at a 0◦ angle of
incidence.
emission is observed compared to the emission from a 40 nm thick cobalt film,
as shown in Fig. 3.13(c). However, when the laser pulses are incident on these
films at 0◦ angle of incidence, no THz emission is detected in the back-reflected
direction (Fig. 3.13(d)).
3.7
Conclusion
We have demonstrated that the THz radiation emitted by cobalt thin films upon
illumination with ultrashort laser pulses shows a different azimuthal angle dependent behavior depending on the sample thickness. For cobalt thin films with
thicknesses less than 175 nm, the THz signal changes sign when the sample is
rotated around the surface normal whereas no sign change is observed for thicker
cobalt films. This behavior is attributed to the change in the direction of magnetization from in-plane to out of plane as the film thickness increases. THz emission
at zero degree angle of incidence is consistent with this change in orientation of
the magnetization. The maximum amplitude of the THz radiation emitted from
cobalt thin films at a 45◦ angle of incidence depends on film thickness. The THz
amplitude increases with increasing thickness, peaks around 40 nm and then decreases. The measurements are consistent with MFM measurements of cobalt films
showing the increasing presence of out-of-plane magnetization for increasing film
thicknesses. Our results provide strong evidence that THz emission from cobalt
Chapter 3. THz emission from ferromagnetic metal thin films
53
thin films after illumination with femtosecond laser pulses, is the result of rapid
changes in the magnetization.
Chapter 3. THz emission from ferromagnetic metal thin films
3.8
54
Effect of capping layer on the Terahertz emission
From the above discussion, we can state that the THz radiation emitted from ferromagnetic thin films is most likely due to an ultrafast change in the magnetization.
The process of ultrafast demagnetization of ferromagnetic metals is governed by
the demagnetization time.
~ is placed in an effective field H
~
When a ferromagnet with magnetization M
(which is a combination of the external magnetic field and all the internal fields
acting on the magnetization such as, demagnetizing field, anisotropy field, exchange field etc.) the magnetization ultimately relaxes towards the field axis due
to a damping term. This damping is known as Gilbert damping and it occurs on
the time scale of hundreds of picoseonds [115,116]. In 2005, Koopmans established
a relation between the Gilbert damping and the demagnetization taking place on
a sub-picosecond time scale. He showed that the relatively slow process of Gilbert
damping is correlated with the damping of single-electron precession of an electron
spin initially not aligned with rest of the electron spins [117]. According to the
model, the demagnetization time (τm ) is inversely proportional to the so-called
Gilbert damping constant (α) [117].
τm ∝
1
α
(3.1)
The value of the damping factor is a material property. In the literature, the
value of α ranges from 0.005 to 0.1. For example, α = 0.031 for a thin bi-layer
of CoFe/NiFe [118]. For magnetic layered systems, the Gilbert damping constant
is enhanced [119]. In 2008, Malinowski et al. showed that the damping constant
of a ferromagnetic thin film can be tuned by using non-magnetic metallic capping
layers [120]. In 2011, Barman et al. showed that with an increase in the number
of layers in a [Co/Pt]N multilayer system, the Gilbert damping constant increases
[121]. Thus, by varying the capping layer, the demagnetization time and the
magnitude of the demagnetization process in a ferromagnetic film can be changed.
In 2012, Shen et al. showed experimentally that the amplitude of the THz
radiation emitted from a ferromagnetic thin film is strongly correlated with the
Gilbert damping constant [108, 122]. The larger the Gilbert damping constant
of the ferromagnetic film, the faster the demagnetization process and the larger
the emitted THz signal. Therefore, by changing the capping layer, the damping
of the ferromagnetic thin film and the amplitude of the emitted THz radiation
from the ferromagnetic/metal interface can be tuned. However, recently, Radu et
al. showed that when the Gilbert damping parameter of permalloy (Ni80 Fe20 ) is
changed by doping with rare earth elements, the demagnetization constant is not
inversely proportional to the damping constant [95].
An alternative explanation for why the THz emission changes with a capping
layer was given based on the spin transport mechanism [96]. As mentioned before,
Battiato et al. proposed that when a femtosecond laser pulse is incident on a
ferromagnet-metal interface, a spin current is established from the ferromagnet to
the nonmagnetic capping layer. Thus, the authors argue that the laser-induced
ultrafast demagnetization of ferromagnetic metals is a transport effect. Subse-
Chapter 3. THz emission from ferromagnetic metal thin films
55
quently, Kampfrath et al. demonstrated different transport dynamics for ruthenium capped iron thin films and for gold capped iron thin films, where gold has a
higher electron mobility than ruthenium [123]. In this paper they explain that, microscopically, the electrons injected from an iron film to the gold layer will occupy
sp states and have a higher velocity. However, in the case of ruthenium, electrons
will fill d-band states and have a lower velocity. Hence, transport of the electrons
is much slower in the ruthenium layer compared to the gold layer. As a result,
the demagnetization in Fe/Ru is slower than the demagnetization in Fe/Au [123].
Therefore, the presence of a capping layer can increase or decrease the demagnetization time and the amount of demagnetization, which in turn determines the
emission of THz radiation.
3.8.1
THz emission from Pt/Co thin films
The amplitude of the emitted THz radiation changes when a non-magnetic metal
layer is deposited on top of a ferromagnetic metal thin film. In 2012, Shen et al.
selected FeNi as a ferromagnetic layer and aluminum, copper and tantalum, one at
a time, as a non-magnetic metal layer [108]. The authors show that the amplitude
of THz radiation emitted from NM/FeNi/NM layers depends on the non-magnetic
metal layer. In 2013, Kampfrath et al. showed that when a thin layer of ruthenium
or gold is deposited on top of an iron thin film, strikingly different THz signals
from the ruthenium and gold covered iron films are observed [123].
In our work, we choose platinum (Pt) as the non-magnetic layer. A 5 nm thick
platinum film is deposited as a capping layer on cobalt films of different thicknesses
(5 nm - 100 nm), by electron beam evaporation at a rate of 1 Å/s. The laser pulse
is incident on the platinum side and THz emission is measured in the reflection
direction, as shown in Fig. 3.14(a). In Fig. 3.14(b), we show the THz electric field
fs laser
pulse
5
40
nm
nm
Pt
THz
pulse
Co
Figure 3.14: (a) Schematic diagram of the experimental setup. A femtosecond
laser pulse is incident on a 5 nm Pt/40 nm Co sample from the platinum side at a
45◦ angle of incidence and the emitted THz radiation is measured in the reflected
direction. (b) Comparison of the THz electric field emitted from a 5 nm Pt/40
nm Co sample, with the THz electric field emitted from a 40 nm thick cobalt film
deposited on the glass substrate.
Chapter 3. THz emission from ferromagnetic metal thin films
56
as a function of time, emitted from a 5 nm Pt/40 nm Co sample and from a 40 nm
thick bare cobalt film. We observe that the emission from the 5 nm Pt/40 nm Co
sample is much stronger compared to the emission from the 40 nm thick cobalt film.
It is important to mention that we did not observe any THz emission from 5 nm of
platinum deposited on glass, when excited with femtosecond laser pulses. Hence,
it is really remarkable that when such a thin film of platinum is deposited on top
of a cobalt thin film, it can increase the amplitude of the emitted THz radiation
considerably. The increase in the amplitude of THz emission is attributed to the
considerably larger Gilbert damping constant of Pt/Co compared to that of a
cobalt thin film [124–126].
In Fig. 3.15 (a), we plot the emitted THz amplitude as a function of the laser
power incident on the Pt/Co sample. Similar to the results obtained with the
cobalt thin film, THz emission from Pt/Co also shows a linear dependence on the
incident pump power. It indicates that a second-order nonlinear optical process is
responsible for the emission of THz radiation. In Fig. 3.15(b) we plot the emitted
THz amplitude and the optical absorption as a function of the incident pump beam
polarization for the 5 nm Pt/40 nm Co sample. The p-polarized pump beam is
incident on the sample at a 45◦ angle. We observe that the THz amplitude and
the optical absorption change with changes in the polarization of the pump beam
but this variation can most likely be attributed to the coupling efficiency of the
pump beam into the film.
(a)
(b)
Figure 3.15: (a) Pump power dependence of the THz electric field amplitude
emitted from a 5 nm Pt/10nm Co sample. (b) The measured percentage of the
absorbed pump power (blue) and the amplitude of the THz electric field (red)
emitted from the Pt/Co sample, as a function of pump beam polarization.
3.8.2
Relation between the magnetic order and THz emission
To investigate the relation between the magnetic order of the film and the polarity
of the emitted THz electric field, we applied an external magnetic field to a 5 nm
Pt/10 nm Co sample using permanent magnets. In Fig. 3.16(a) we show the
Chapter 3. THz emission from ferromagnetic metal thin films
57
schematic diagram of the setup used to apply the external magnetic field on the
Pt/Co samples. The direction of the external magnetic field is in the plane of the
sample. When femtosecond laser pulses are incident on the sample at a 45◦ angle of
incidence, THz emission is observed. We measure the emitted THz electric field as
a function of time in the presence of the magnetic field. We observe that when the
direction of the applied magnetic field is reversed, the polarity of the THz signal
is also reversed, as shown in Fig. 3.16(b). This suggests that the magnetization of
the 5 nm Pt/10 nm Co sample is in-plane.
N
fs laser
pulse
S
S
S
N
A
N
N
S
Pt/Co
layers
magnets
(a)
(b)
Figure 3.16: (a) Schematic diagram of the setup used to apply an external magnetic
field to the Pt/Co samples. (b) Measured THz electric field emitted by a 5 nm
Pt/10 nm Co sample, as a function of time. Black and red traces indicate THz
emission with magnetic fields applied in opposite directions.
3.8.3
Azimuthal angle dependence
In Fig. 3.17, we plot the amplitude of the p-polarized THz emission from different
Pt/Co samples as a function of sample azimuthal angle. For each sample, the
thickness of the platinum layer is 5 nm while the thickness of the cobalt film is
varied. The p-polarized pump beam is incident on the samples at a 45◦ angle of
incidence and there is no externally applied magnetic field. We observe that the
emitted THz signal shows a sinusoidal dependence on the azimuthal angle. When
the sample is rotated by 180◦ , the polarity of the THz signal is reversed. This
suggests that the magnetization of the sample is pre-dominantly in-plane for all
the samples.
Chapter 3. THz emission from ferromagnetic metal thin films
(a)
(d)
(b)
(e)
58
(c)
(f )
Figure 3.17: The azimuthal angle dependence of the measured THz electric field
amplitude emitted by different samples with (a) 10 nm (b) 30 nm (c) 50 nm (d)
60 nm (e) 70 nm (f) 80 nm thick cobalt film. The thickness of the platinum layer
is 5 nm for each sample.
3.8.4
Thickness dependent THz emission
We measure the reflected, transmitted and absorbed pump power as a function
of the thickness of the cobalt films for all Pt/Co samples. The thickness of the
platinum layer is constant (5 nm) for all samples. The absorbed power increases
initially but quickly becomes constant as we increase the thickness of the cobalt
layer, as shown in Fig. 3.18(a). In Fig. 3.18(b), we plot the amplitude of the
THz radiation emitted at a 45◦ angle of incidence as a function of the thickness
of the cobalt thin films, with a 5 nm platinum layer on top. For each thickness,
the sample is rotated until the maximum THz amplitude is recorded. We observe
that the emitted THz amplitude shows a thickness dependent behavior. When we
increase the thickness of the cobalt thin film, the THz emission initially increases
with increasing thickness, peaks when the thickness of the cobalt film is around
7 nm and then decreases. The thickness dependence behaviour demonstrated by
Pt/Co thin films is completely different than the thickness dependence behaviour
demonstrated by cobalt-only thin films, as shown in Fig. 3.12(a). Whereas, the
increase in the emitted THz electric-field amplitude is expected when cobalt is
capped with a thin layer of Pt, it is surprising that the maximum THz emission is
observed for a Co layer thickness of 7 nm only. The thickness-dependent THz emission from Pt/Co samples has not been thoroughly studied and more experiments
are required to explain this behavior.
Chapter 3. THz emission from ferromagnetic metal thin films
Thickness of cobalt film (nm)
(a)
59
(b)
Figure 3.18: (a) Percentage reflection, transmission and absorption of the pump
laser pulses by different Pt/Co samples. (b) p-polarized THz emission as a function
of thickness of cobalt film. The thickness of platinum film is 5 nm for each sample
while the thickness of the cobalt film is varied from 5 nm to 100 nm.
3.8.5
Effect of changing the order of the films on THz emission
In 2013, Eschenlohr et al., observed that the efficiency of the ultrafast demagnetization remains the same even when only 7% of the incident pump pulse energy
reaches the ferromagnetic nickel layer [97]. This supports the idea that for ultrafast demagnetization and emission of THz pulses, direct illumination of laser
pulses is not required and that demagnetization can be achieved by the transport
of electrons. To investigate whether the same idea holds true for Pt/Co thin films,
we prepared two samples, 10 nm Pt/10 nm Co and 10 nm Co/10 nm Pt. For the
first sample, a 10 nm thick platinum layer is deposited on top of a 10 nm thick
Figure 3.19: Comparison of THz amplitude emitted from a 10 nm Pt/10 nm Co
(red trace) with 10 nm Co/10 nm Pt (black trace) deposited on a glass substrate.
The emitted THz amplitudes are comparable even when the order of the films is
changed.
Chapter 3. THz emission from ferromagnetic metal thin films
60
cobalt film. Whereas, for the second sample, the order of the thin films is swapped
and the cobalt layer is deposited on top of the platinum film. Both samples are
illuminated with femtosecond laser pulses at a 45◦ angle of incidence and THz
emission is measured. In Fig. 3.19, we show the measured THz emission from the
10 nm Pt/10 nm Co (red) and from the 10 nm Co/10 nm Pt (black) samples. We
observe that the emitted THz amplitude is comparable for both samples. Despite
the fact that when the platinum layer is deposited on top, much less incident laser
energy reaches the ferromagnetic metal, there is not much effect on the amplitude
of the THz emission. This supports the idea that the THz generation from Co/Pt
samples is mainly due to transport of electrons and not due to heating effects in
the Co layer.
3.9
Conclusions
In conclusion, we have shown THz emission from Pt/Co thin films after illumination with femtosecond laser pulses. We show that the presence of a platinum
capping layer can increase the emission from cobalt thin films. The emitted THz
radiation changes sign when the sample is rotated by 180◦ around the surface
normal which indicates that the magnetization is in-plane. We also show that the
maximum amplitude of the THz radiation emitted from Pt/Co films at a 45◦ angle
of incidence shows a thickness dependent behavior. However, more experiments
and studies are required to understand the azimuthal angle dependence and the
thickness dependence behavior of the THz electric field amplitude emitted from
Pt/Co samples.
Chapter
4
4.1
THz emission from
BiVO4/Au thin films
Motivation
Recently, THz emission from Cu2 O/Au interfaces was reported upon illumination
with femtosecond laser pulses centered around 800 nm [127]. These experiments
are surprising because the bandgap of Cu2 O is much larger than the energy of
photons corresponding to a 800 nm wavelength. Therefore, it is interesting to try
other large bandgap materials as well, especially BiVO4 which is technologically
relevant. BiVO4 is an exciting material, which is widely used in the pigment industry and has potential applications for photoelectrochemical water splitting. The
study of THz pulses emitted from such semiconductors and semiconductor/metal
interfaces can provide us with information about the carrier dynamics and the
carrier transport properties in the semiconductors. In cases where the THz generation mechanism is initially not known, careful examination of the properties of
the emitted waveforms can help us discover the source of the emission.
4.2
THz generation from semiconductors
Terahertz (THz) pulses can be generated by exciting non-linear optical crystals,
metals and semiconductors using ultrashort laser pulses that excite currents and
polarizations [99, 128]. For example, THz radiation can be generated by illuminating thin semiconductor layers deposited on metal surfaces [92, 129]. When a
femtosecond laser pulse is incident on such a semiconductor-metal junction, a transient current is formed in the Schottky depletion layer of the metal/semiconductor
interface, which gives rise to the emission of an electromagnetic transient in the
THz range. In the past, THz emission from mostly conventional semiconductors
like gallium arsenide, silicon, germanium and some unconventional ones, such as
cuprous oxide, has been studied [130]. Another process taking place near the
semiconductor surface which can give rise to the THz emission of THz radiation is
the photo-Dember effect. When femtosecond laser pulses are incident on a semiconductor, electron-hole pairs can be generated near the surface, which diffuse
away into the bulk. Due to the difference in mobilities between electrons and
holes, one type of charge carrier moves faster than the other. This results in the
formation of a transient dipole in the vicinity of a semiconductor surface, which is
Chapter 4. THz emission from BiVO4 /Au thin films
62
known as the photo-Dember effect. This transient dipole gives rise to THz generation [131]. Hence, the generation of a transient current or a transient dipole can
be either due to the presence of a built-in field present inside the semiconductor or
it can be due to the photo-Dember effect. Below we discuss these two important
mechanisms, carrier acceleration in the surface field, and the photo-Dember effect,
which play an important role in the case of THz generation from surfaces and
interfaces.
4.2.1
Surface field effect
When a metal comes into contact with a semiconductor, the metal-semiconductor
junction forms either a Schottky or an ohmic contact, depending on the work
functions of the metal and the semiconductor. The work function is defined as the
energy required to bring an electron from the Fermi level to the vacuum level [24].
When the work function of the metal (φm ) is greater than the work function of
the semiconductor (φsemi ), i.e. when φm > φsemi , a metal-semiconductor junctions form a Schottky junction. On the other hand, when φm < φsemi metalsemiconductor junction forms an ohmic junction. In Fig. 4.1(a) we show the
schematic of the band diagram of a metal. At the top we show the vacuum
level O, EF m is the Fermi energy and φm is the work function of the metal. In
Fig. 4.1(b) we show the band diagram of an n-type semiconductor. Since it is an
n-type semiconductor, the Fermi level is closer to the conduction band. Ev is the
top of the valence band, Ec is the bottom of the conduction band, EF n−semi is the
Fermi energy of the semiconductor and Eg is the band gap. The work function of
the semiconductor is φsemi and the electron affinity is χ. The electron affinity is
Vacuum
level O
Conduction band
O
χ
фsemi
фm
E Fm
Eg
Ec
E Fn-semi
Ev
Valence band
фm : Metal work function
(a)
χ : Electron affinity
фsemi : Work function of semiconductor
(b)
Figure 4.1: a) The schematic of the energy band diagram of a metal. EF m is
the Fermi energy of the metal and eφm is the work function of the metal b) The
schematic of the energy band diagram of an n-type semiconductor. Ev is the top of
the valance band. Ec is the bottom of the conduction band, EF semi is the Fermi
energy of the semiconductor, Eg is the band gap, φsemi is the work function of the
semiconductor and χ is the electron affinity.
Chapter 4. THz emission from BiVO4 /Au thin films
63
defined as the energy required to bring an electron from the conduction band to
the vacuum level [132].
When a metal comes into contact with a semiconductor, the Fermi level of the
two materials must match at thermal equilibrium [24]. In Fig. 4.2(a), we show the
junction of a metal with an n-type semiconductor. For an n-type semiconductor,
when a metal and a semiconductor come into contact with each other, electrons
move from the conduction band of the semiconductor to the metal. As a result,
we have a net positive charge on the semiconductor and a net negative charge on
the metal side. Hence, an electric field is formed in the direction from the semiconductor to the metal. Far away from the junction, the semiconductor behaves
like a typical n-type semiconductor but field-induced band bending occurs near
the metal-semiconductor junction. For an n-type semiconductor, when electrons
move from the semiconductor to the metal they not only move from the surface
of the semiconductor but also from a region near the interface, within the semiconductor, which is thus depleted of carriers. This region is called the depletion
region or the space charge region. The electric field across the depletion layer is
called the depletion field and the corresponding potential difference between the
metal and the semiconductor is called the contact potential [133]. The contact potential depends on the work functions of metal and semiconductor. The strength
of the depletion field depends on the extent of the bending of the valence and
conduction bands of the semiconductor near the surface. The depletion layer acts
like a potential barrier. Electrons in the semiconductor must pass over a potential
Depletion
region
Vacuum
O
level
фm
EF
Metal
Conduction
band
фb
O Vacuum level
χ n-semi
Ec
фn-semi
EFn-semi
+
+
+
+
Valence
band
Ev
E (Depletion field)
V0
n-type semiconductor
Contact potenial eV0 = фm - фn-semi
Schottky barrier ф = ф - χ n-semi
b
m
Figure 4.2: The schematic of the energy band diagram of a metal-semiconductor
junction. Here, eV0 = φm - φn−semi is the potential barrier and φb = φm - χn−semi
is the Schottky barrier.
Chapter 4. THz emission from BiVO4 /Au thin films
64
barrier of height eV0 in order to get into the metal, where eV0 = φm - φn−semi .
Similarly, electrons in the metal must pass a potential barrier of height φb in order
to get into the semiconductor, where φb = φm - χn−semi , where φb is known as
the Schottky barrier [134].
Due to the depletion field present at the Schottky barrier, the carriers photoexcited in this region accelerate and form a transient photocurrent normal to the
surface, giving rise to THz emission [135–138]. The stronger the depletion field,
the stronger the acceleration of the charge carriers, and thus the stronger the THz
emission. THz emission caused by the surface field effect has been demonstrated
for several semiconductors like GaAs, InP, and InN [139–141].
4.2.2
Photo-Dember effect
In the case of narrow-bandgap semiconductors, where the surface depletion field is
weaker, we can have THz emission through the photo-Dember effect. In the photoDember effect, the incident light is absorbed near the surface of the semiconductor.
When the absorption is strong, more charge carriers are generated near the surface
of the semiconductor compared to deeper into the material. As a result, a nonuniform carrier distribution is built up and a carrier gradient is formed. When, in
addition, the mobilities of the electrons and holes are also different, they diffuse
with different velocities. As a result, a transient dipole is rapidly formed near
the surface. This time-dependent dipole, which is parallel to the concentration
gradient and perpendicular to the excited surface, gives rise to THz radiation
[131, 142].
The diffusive currents due to electrons (Je ) and holes (Jp ) are given by [27],
Jn ∝ eDe
d∆p
d∆n
, Jp ∝ −eDh
dx
dx
(4.1)
where e is the electron-charge, ∆n (∆p) is the density of the photo-generated
electrons (holes) and De (Dh ) is the diffusion coefficient of the electrons (holes).
The diffusion coefficient, D, is defined as, D = kBeT µ . where µ is the mobility of
the charge carrier, kB is the Boltzmann constant and T is the temperature of the
carrier distribution.
The photo-Dember current is given by J = Jn + Jp . The mobility and kinetic
energy of electrons are typically much larger than those of the holes so the contribution from holes can often be neglected. Hence, THz radiation due to the
photo-Dember effect is approximately proportional to the electron mobility. The
emitted THz electric field can be expressed as [27],
∂Jn
∝ µe (T )
(4.2)
∂t
In narrow bandgap semiconductors like, InN, InAs, InSb, the difference between
mobilities of electrons and holes is large. THz emission from the photo-Dember
effect has been reported for many such narrow band-gap semiconductor materials
[140, 143, 144].
In 2010, Klatt et al. showed THz emission from the lateral photo-Dember
currents by partially masking the semiconductor surface with a metal layer [145].
ET Hz ∝
Chapter 4. THz emission from BiVO4 /Au thin films
65
In this case, the diffusion of charge carriers is along (parallel to) the surface.
This can be achieved by partially covering a semiconductor surface with a metal
layer. When femtosecond laser pulses, with energy higher than the bandgap, are
incident on the metal-semiconductor interface, an asymmetrical distribution of
photo-generated carriers is formed. These carriers diffuse freely and the diffusion
current going towards and under the metal mask is greater than that going away
from it. Hence, a net diffusion current is formed which gives rise to THz emission
[145].
800nm fs
laser pulse
Emitted
THz pulse
Holes
Electrons
Semiconductor
Figure 4.3: Schematic representation of longitudinal Photo-Dember effect, which
arises due to the concentration gradient and different mobility of charge carriers.
When the surface field effect is the principal mechanism responsible for the
generation of THz radiation, the polarity of THz radiation is opposite for n-type
and p-type semiconductors. This is because the direction of the depletion field
flips according to the doping type. However, if THz emission is mainly due to
the photo-Dember effect, the polarity of the THz radiation remains the same for
n-type and p-type semiconductors. Irrespective of the type of the doping of the
semiconductor, the direction of diffusion of the carriers will not change.
In the following section we describe some of the properties of the semiconductor Bismuth Vanadate, the material that we have used in the THz generation
experiment described later in this chapter.
4.3
Bismuth Vanadate
Bismuth Vanadate (BiVO4 ) is a semiconductor material with broad applications.
It is non-toxic in nature and has the ability to replace toxic pigments like lead,
chromate etc. in the pigment industry [146]. It is greenish-yellow in color and
strongly absorbs visible radiation. Due to its high visible light absorption and high
chemical stability, BiVO4 acts as an efficient catalyst and splits water into hydrogen and oxygen upon illumination. It also displays other interesting properties
Chapter 4. THz emission from BiVO4 /Au thin films
66
such as, ferroelasticity, ionic conductivity and photochromism [147, 148]. BiVO4
exists mainly in three types of crystalline forms: orthorhombic pucherite, tetragonal dreyerite and monoclinic clinobisvanite. These polymorphs have different
physical properties depending on their crystal structure and electronic structure.
Among the three available crystal phases of BiVO4 , monoclinic BiVO4 (m-BiVO4 )
is an important material with many applications. It is a wide band semiconductor
with a bandgap of 2.4 eV and it exhibits much higher photocatalytic activity with
respect to the other polymorphs [149–151].
4.3.1
BiVO4 structure
Crystal structure: In the monoclinic clinobisvanite structure, each unit cell consists of four bismuth (Bi) atoms, four vanadium (V) atoms, and sixteen oxygen
(O) atoms [147]. Each V atom is connected to four O atoms and forms a VO4
tetrahedron and each Bi ion is connected to eight O atoms and forms a BiVO8
dodecahedron. These two are the basic structural units of the monoclinic clinobisvanite structure. Now each BiVO8 dodecahedral unit is connected to 8 such
VO4 units. Also, each O atom is connected to 2 Bi centers and one V center.
In the monoclinic clinobisvanite BiVO4 crystal structure, Bi atoms and V atoms
are arranged alternate to each other along the crystal axis which gives BiVO4
the properties of a layered structure [147]. The crystal structure of monoclinic
O-atom
V-atom
Bi-atom
VO4
BiO8
Figure 4.4: Crystal structure of BiVO4
clinobisvanite is shown in Fig. 4.4. In the case of monoclinic clinobisvanite, the
Chapter 4. THz emission from BiVO4 /Au thin films
67
length of four Bi-O bonds in BiVO4 and two V-O bonds in VO4 are different so
both VO4 and BiVO8 are slightly distorted. Due to this distortion, the centers of
positive and negative charges are shifted and we have an internal electric field and
an enhanced static polarization. This effect improves the separation efficiency of
electron-hole pairs and as a result we have a better photocatalytic activity.
Electronic structure: Like the crystal structure, the electronic structure is
also an important factor for deciding the photocatalytic activity of the material.
The bandgap of monoclinic clinobisvanite BiVO4 was measured to be around 2.4
- 2.5 eV, which is smaller than the bandgap of other BiVO4 polymorphs. Due to
the smaller bandgap, monoclinic BiVO4 shows enhanced absorption and improved
photocatalytic activity. In monoclinic clinobisvanite BiVO4 , the effective mass of
electrons is 0.9 m0 and effective mass of holes is 0.7 m0 . The effective mass of
holes in monoclinic BiVO4 is smaller than in the other BiVO4 polymorphs and in
other oxide semiconductors which are used for photocatalysis. When the carriers
are lighter, they have a higher probability of reaching the reaction sites within
their life-time, which leads to better photocatalytic properties [146, 147, 152].
4.3.2
Preparation of BiVO4 thin film
For our THz generation experiments, thin films of BiVO4 were prepared using a
spray pyrolysis method [153,154]. The schematic of the setup is shown in Fig. 4.5.
For preparing the precursor solution, Bi(NO3 )3 .5H2 O is dissolved in acetic acid
and VO(AcAc)2 in absolute ethanol. These two solutions are mixed and excess
ethanol is added to make the solution 4 millimolar (1 millimolar concentration is
equal to 10−3 moles of solute per litre of solution). This solution is sprayed onto
N2 flow
N2
flow
Nitrogen
( N2 )
N2
flow
Spray
controller
Nozzle
N2 flow
BiVO4
precursor solution
Samp
Sample
ple
Heating plate
Figure 4.5: Schematic of spray pyrolysis setup for deposition of BiVO4 films
a substrate with the help of a nozzle which is driven by nitrogen gas. Each spray
Chapter 4. THz emission from BiVO4 /Au thin films
68
cycle lasts 5 s and then there is a delay of 55 s which allows the solvent to evaporate.
The number of spray cycles determines the final thickness of the film deposited on
the substrate. The substrate is kept on a hot plate which is 20 cm away from the
nozzle, and whose temperature is maintained at 450◦ C. After spraying BiVO4 , the
substrate is left on the hot plate for around 2 hours [153, 154]. The preparation
method has a strong effect on the properties of BiVO4 . For example, BiVO4
prepared by the aqueous process has a much higher photocatalytic activity than
a conventional solid-state method while both result in the same crystal structure.
BiVO4 prepared using the spray-pyrolysis method is reported to be monoclinic in
nature and an n-type semiconductor [155, 156].
In the following section we discuss the results of THz emission experiments
performed on these BiVO4 films.
4.4
4.4.1
Experimental
Sample fabrication
For the sample, a 100 nm thick gold film was first deposited on the glass substrate
using electron beam evaporation. For adhesion purposes, a 10 nm thin chromium
layer was deposited first. Then, different thicknesses (20 nm - 300 nm) of BiVO4
were deposited on top of the gold film using the spray pyrolysis technique. During
the deposition of BiVO4 , the sample (glass substrate coated with gold thin film) is
heated at 450◦ C for around 2 hours. As a result of the heating, the surface of the
bare gold becomes rough and discontinuous and there is a formation of big islands.
In Fig. 4.6(a) we show a scanning electron microscope (SEM) image of a 100 nm
thick gold film after heating at 450◦ C for around 2 hours. Heating of the sample
while spraying leads to drying and crystallization of BiVO4 . In Fig. 4.6(b) we show
the topography of a 20 µm x 20 µm area of a 200 nm thick BiVO4 film, deposited
directly on a bare glass substrate, as measured with an atomic force microscope.
The surfaces of the prepared BiVO4 thin films are quite rough. For a 200 nm thick
film, the root mean square (RMS) roughness was estimated to be around 20 nm.
In Fig. 4.6(c), we show the measured reflection spectrum of a 100 nm thick BiVO4
film deposited on top of a gold film with a 100 nm average thickness. The reflection
from BiVO4 /Au is normalized with respect to the reflection from an aluminum thin
film. The figure shows that for increasing wavelengths the normalized reflection
drops to approximately 25 percent for wavelengths longer than about 550 nm.
The optical bandgap of BiVO4 is 2.4 eV (≈ 520 nm) and hence it has a strong
absorption in the ultraviolet and visible region. An aluminum mirror is used for
measuring the spectrum of the source in the range of 400 nm - 900 nm as it has
a very high and a nearly spectrally flat reflection in this wavelength region of the
spectrum.
4.4.2
THz generation and detection setup
The experimental setup is schematically shown in Fig. 4.7. We use a Ti:Sapphire
oscillator which has a center wavelength of 800 nm, a pulse duration of 50 fs, an
Chapter 4. THz emission from BiVO4 /Au thin films
(a)
69
(b)
10 μm
µm
20
20
µm
A
Annealed Au/Glass
(c)
Figure 4.6: (a) Scanning electron microscope (SEM) image of a 100 nm thick
gold film after annealing at 450◦ C for around 2 hours. (b) AFM scan showing the
topography of a 20 µm x 20 µm area of a 200 nm thick BiVO4 film deposited on
the glass substrate. (c) UV-vis reflection spectra of a 100 nm thick BiVO4 film
deposited on a 100 nm thick gold film.
average power of 800 mW and a repetition rate of 11 MHz. When ultrashort pulses
from the laser oscillator are incident on the sample at a 45◦ angle of incidence,
THz emission is observed. A pair of gold coated parabolic mirrors is used for
collecting, collimating and finally, focusing the THz radiation onto a 500 µm thick
zinc telluride (ZnTe) (110) electro-optic detection crystal. The electric field of the
THz radiation induces a small birefringence in the electro-optic crystal. At the
same time, a part of the same ultrashort laser pulse that was used to generate
the THz pulse, is incident on the detection crystal. When the linearly polarized
probe beam propagates through the electro-optic crystal, due to the THz induced
birefringence, it acquires a small elliptical polarization. Then, the probe beam
passes through a Wollaston prism which separates the beam into two orthogonal
components. A differential detector consisting of two photodiodes measures the
difference in the intensities which is proportional to the instantaneous THz electric
field. By varying the time-delay between the pump pulse and the probe pulse, the
THz electric field is measured “stroboscopically” as a function of time.
Chapter 4. THz emission from BiVO4 /Au thin films
Beam
Splitter
Femtosecond
Laser
Probe
Beam
Pump
Beam
THz
Beam
70
Wollaston
prism
Detection Quarter
Crystal Waveplate
Differential
detector
Electro-optic detection
BiVO4
Sample
Parabolic
Mirrors
Figure 4.7: Experimental setup for the generation of THz light from BiVO4 thin
layer illuminated with femtosecond laser pulses.
4.5
Results and discussion
In Fig. 4.8(a) we show the measured THz electric field as a function of time,
emitted from a 100 nm thick BiVO4 film deposited on a 100 nm thick gold film.
The amplitude of the emitted THz electric field is roughly around 0.2% of the
THz emission from a conventional semi-insulating GaAs (100) surface depletion
field emitter and is comparable to the emission from percolated gold [105]. Tight
focusing of the pump beam is avoided to prevent any damage to the sample. The
emitted THz amplitude increases linearly with the laser power incident on the
(a)
(b)
Figure 4.8: (a) THz electric field emitted from a 100 nm thick BiVO4 film deposited
on a 100 nm thick gold film, plotted vs. time. (b) Emitted THz amplitude plotted
as a function of the incident laser power.
sample as shown in Fig. 4.8(b). This suggests that a second order non-linear
optical process is responsible for the THz emission.
Chapter 4. THz emission from BiVO4 /Au thin films
71
In the literature, BiVO4 thin films deposited on a gold surface are reported
to show a diode-like behavior, suggesting that the BiVO4 /Au junction forms a
Schottky interface. In Fig. 4.9(a), we schematically show the energy band bending
between gold and BiVO4 , which is an n-type semiconductor. In order to determine
if the depletion field is responsible for the THz generation in our case, we included
SiO2 dielectric layers of varying thickness between the BiVO4 layer and the gold
layer. Due to the SiO2 layer, the carrier transport between gold and BiVO4 is
strongly reduced, which should hinder the formation of a depletion field. However,
during the deposition of BiVO4 , due to heating, there is a risk of destroying the
ultrathin SiO2 layer. In that case, the BiVO4 film can again come into direct
Depletion
Layer
BiVO4
Gold
Barrier
Fermi Height
Level
Conduction band
EF
Valence Band
(a)
(b)
Annealed SiO2/Au/Glass
(c)
(d)
Figure 4.9: (a) The energy band bending between BiVO4 and gold thin films
(b) I-V characterization of a bare gold thin film and a gold thin film with 20 nm
SiO2 deposited on top. No current is measured for the SiO2 /Au sample, which
confirms that even after annealing, the silica layer is not destroyed. (c) SEM image
of SiO2 /Au sample after annealing at 450◦ C for around 2 hours. (d) Comparison
of THz emission from BiVO4 /Au and BiVO4 /5 nm SiO2 /Au. The emitted THz
amplitudes are comparable even after the inclusion of SiO2 layer. The thickness of
the BiVO4 film is 100 nm.
contact with the gold layer and form a Schottky junction. To check if the SiO2
layer is still intact after heating for 2 hours, we inspect the samples using current-
Chapter 4. THz emission from BiVO4 /Au thin films
72
voltage (I-V) measurements. In Fig. 4.9(b) we show the I-V curves for the bare
gold thin film and the gold thin film with 20 nm SiO2 layer on top. We measure
a significant current for the bare gold thin film but very little current is measured
for the SiO2 /Au layers which confirms that even after heating, the SiO2 layer is
not getting destroyed and remains intact.
Moreover, the sample is also characterized using the scanning electron microscopy (SEM) technique. In Fig. 4.9(c) we show the SEM image of the gold
thin film coated with a 20 nm SiO2 layer, after heating at 450◦ C for 2 hours. Interestingly, it is observed that when a 20 nm thin SiO2 layer is present on the top
of gold, the surface of the annealed BiVO4 /Au sample is much smoother compared
to the surface of the annealed bare gold film. This gives a clear indication that
the SiO2 layer, sandwiched between the gold thin film and the BiVO4 thin film,
remains intact even after heating. In Fig. 4.9(d) we compare the amplitude of the
THz radiation emitted from BiVO4 /Au and from BiVO4 /SiO2 /Au samples with a
silica layer thickness of 5 nm. We observe that the THz amplitude remains largely
unaffected after the inclusion of the thin SiO2 layer. Similar results are observed
with thicker layers of silica sandwiched between gold and BiVO4 . Even for 10 nm,
20 nm and 40 nm thick silica layers, THz emission is only a little weaker (not
shown here).
The above results make it less likely that the THz generation is due to carrier
acceleration in the depletion field associated with the BiVO4 /Au Schottky interface. The excitation of BiVO4 thin films deposited on a glass substrate using
femtosecond laser pulses does not produce any measurable THz emission. This
excludes the possibility of THz emission due to a surface depletion field or the
surface photo-Dember effect [26, 27, 157, 158]. We therefore propose a new generation mechanism based on the longitudinal photo-Dember effect. In a typical
longitudinal photo-Dember effect, schematically shown in Fig. 4.3, electron-hole
pairs are generated in the vicinity of a semiconductor surface by photo-excitation.
Due to a difference in mobilities, electrons and holes move with different velocities.
800 nm fs
laser pulse
Air
THz radiation
High
intensity
BiVO4
Au
Low
intensity
Figure 4.10: Intensity distribution in the vertical direction due to the formation of
standing wave pattern. The maximum intensity is at air/ BiVO4 interface whereas
the minimum is at BiVO4 /Au interface.
Chapter 4. THz emission from BiVO4 /Au thin films
73
As a result, a dipole perpendicular to the surface is formed which emits THz radiation. Interestingly, the longitudinal Photo-Dember effect may also be realized
in a slightly different way. When laser light is incident on the BiVO4 thin film
deposited on gold, due to the interference of light reflected from the air/BiVO4
and BiVO4 /Au interfaces, a standing wave pattern is formed. As a result, we
may have a higher intensity (anti-node) at the Air-BiVO4 interface and a lower
intensity (node) at the BiVO4 /Au interface, as shown in Fig. 4.10(b). Because of
this, we have a higher absorption near the Air/BiVO4 interface and, as a result,
more carriers than near the BiVO4 /Au interface. In this way, we get a concentration gradient which, combined with a difference in the mobility of electrons
and holes, gives rise to the THz emitting transient dipoles. In BiVO4 , the hole
mobility (µh = 2.25 × 10−3 cm2 /Vs) is higher than the electron mobility (µe =
1.75 × 10−3 cm2 /Vs). To further test whether this explanation is plausible we
have also studied samples in which the gold layer has been replaced by indium tin
oxide (ITO). ITO is a conducting oxide which reflects THz light and transmits
light with a wavelength of 800 nm quite well. In the absence of the gold layer,
multiple reflections are strongly reduced and no significant charge carrier gradient
is formed. As a consequence, no THz emission is observed when the gold layer is
replaced by ITO layer. This provides supporting evidence to our interpretation of
the generation mechanism of THz radiation from the BiVO4 /Au interfaces.
4.5.1
Thickness dependent THz emission
BiVO4 thin films of different thicknesses (20 nm - 300 nm) were deposited directly
onto the glass substrates. In Fig. 4.11(a), we show the reflected, transmitted and
Figure 4.11: (a) Percentage reflection, transmission and absorption of the pump
laser pulses as a function of the thickness of BiVO4 . Different thicknesses of BiVO4
are directly deposited on top of glass substrates. (b) Percentage reflection and
absorption of the pump laser pulses as a function of thickness of BiVO4 . In this
case, different thicknesses of BiVO4 are deposited on top of a 100 nm thick gold
layer. The thickness of the BiVO4 film is varied from 20 nm to 300 nm.
absorbed pump power as a function of the thickness of the BiVO4 films at a 45◦
angle of incidence. The measurements show that there is a very small reflection
Chapter 4. THz emission from BiVO4 /Au thin films
74
from BiVO4 /glass samples. As the thickness of the BiVO4 film increases, the
absorption also increases. For a 300 nm thick film of BiVO4 deposited on glass,
around 75 percent of the incident light is absorbed.
As another set of samples, we deposited different thicknesses of BiVO4 on a
100 nm thick gold film. Since the 100 nm thick gold film blocks the pump beam,
there is no transmission in this case. The reflected and absorbed pump power as
a function of the thickness of the BiVO4 film deposited on gold film are shown in
Fig. 4.11(b). The absorption again increases with the increase in the thickness of
BiVO4 film and it is high compared to the BiVO4 /glass samples.
In Fig. 4.12, we show the amplitude of the THz radiation emitted from the
BiVO4 /Au interface as a function of the thickness of the BiVO4 layer. When we
increase the thickness of the BiVO4 thin film, the THz emission initially increases
with increasing thickness, peaks around 100 nm and then decreases. The results
suggest that as we increase the thickness of the BiVO4 thin film, the absorption
of pump light increases which is expected as the interaction length of the pump
light with BiVO4 increases (Lambert-Beer law). More absorption leads to the
generation of more charge carriers, and, as a result, the THz amplitude increases.
As we further increase the thickness of the BiVO4 layer, the absorption becomes
high and less light reaches the gold surface. As a result, the standing wave is much
less pronounced and so the THz amplitude decreases again.
Figure 4.12: THz emission amplitude from BiVO4 /Au interface as a function of
the thickness of the BiVO4 layer.
4.6
Conclusion
In conclusion, we demonstrate THz emission from BiVO4 /Au thin film interfaces
after illumination with femtosecond laser pulses and investigate the possible generation mechanisms. Based on the experimental results and observations, we pro-
Chapter 4. THz emission from BiVO4 /Au thin films
75
pose that the longitudinal photo-Dember effect is the mechanism responsible for
the THz generation.
Chapter
5
Conclusions
Terahertz time-domain (THz-TDS) spectroscopy has proven itself as a promising spectroscopy tool with great potential for a wide range of applications. This
technique is based on the generation and detection of the electric field of subpicosecond THz pulses using femtosecond near-infrared laser pulses. THz radiation
has lots of applications in the field of imaging and spectroscopy. Recently, there
has a been a lot of effort to find the sources which can generate high power THz
radiation. THz emission from different materials can also provide information on
the properties of the emitting material. THz radiation can be used for imaging the
near-field of metamaterial elements with a subwavelength spatial resolution. The
near-field information of metamaterial elements is important because it provides a
clear description of the microscopic fields that give metamaterials their remarkable
properties.
In chapter 2, for the first time, we present the measurement of the two dimensional spatial distribution of the THz magnetic near-field of split ring resonators
(SRR). SRRs are one of the most common metamaterial elements. The SRRs are
fabricated by deposition of gold on TGG crystal substrates. In the past, the THz
electric field of such SRRs was measured but, until now, the magnetic field was
never directly measured. This is because the THz magnetic near-field of SRRs is
very weak and difficult to measure experimentally. To measure the THz magnetic
near field of a SRR, THz radiation is incident on the SRR in such a way that
the THz electric field is parallel to the gap of the SRR. As a result, an oscillating
current is created in the ring which generates a magnetic field perpendicular to
the SRR surface which is measured using free space magneto-optic sampling. The
experimental results are in good agreement with the simulations. A spatial resolution of about 10 microns is reached, which is more than two orders of magnitude
smaller than the resonant wavelength of these structures. We have also measured
the THz magnetic near-field of a complementary split-ring resonator (CSRR).
However, the THz magnetic near-field signal for the complementary structure is
very weak compared to the signal for the regular split-ring resonator. This work
opens up new possibilities for studying the THz near-field of metal structures and
metamaterial elements.
Chapter 5. Conclusions
78
In Chapter 3 we discuss THz emission from ferromagnetic thin films. It is often forgotten that nonmagnetic metals are also capable of emitting THz light and
that such a contribution to the emitted THz field cannot a priori be excluded for
magnetic metals. Our work clearly establishes a strong correlation between the
magnetization and the emission of THz light following the excitation of cobalt with
a femtosecond laser pulse. It does this by highlighting the role of the orientation
of the magnetization in the terahertz emission from ferromagnetic thin films. We
find that, as we increase the cobalt film thickness, the polarization direction of the
emitted THz pulse changes, correlating with the transition from a predominantly
in-plane to a predominantly out-of-plane magnetization, as measured with magnetic force microscopy. It was also seen from our experiments that THz emission
from cobalt thin films shows a thickness dependent behavior. Furthermore, we
have investigated the effect of capping layers on the amplitude of the emitted THz
radiation. Relatively strong THz emission was observed from Pt/Co thin films.
More research is needed to lead us to an answer on the role of the capping layer
on the THz emission from ferromagnetic thin films.
THz radiation can be generated by exciting nonlinear crystals, semiconductors
and metals using ultrafast laser pulses. Usually semiconductor thin films do not
emit strong THz radiation when infrared femtosecond laser pulses are incident
on them. THz emission from thin films of semiconductors can be enhanced when
they are deposited on gold surfaces. For example, Cu2 O/Au interfaces emit strong
THz radiation when illuminated with femtosecond laser pulses. Similar to Cu2 O,
BiVO4 is a wide bandgap semiconductor material which has exciting technological
applications. Chapter 4 discusses THz emission from BiVO4 /Au interfaces when
illuminated with femtosecond laser pulses. From the analysis of the emitted THz
pulses we show that the longitudinal photo-Dember effect is the most likely mechanism behind the THz emission. The photo-dember effect is the formation of a
transient dipole in the vicinity of a semiconductor surface which takes place due to
photoexcitation of semiconductor with femtosecond laser pulses. When light is incident on a semiconductor material, electron-hole pairs are generated. Due to the
difference in mobility, electrons and holes move with different velocities and form
a carrier gradient. As a result, a time-dependent dipole is formed perpendicular
to the surface, which gives rise to the emission of a THz pulse.
Finally, one may look at each of the three main chapters in this thesis by zooming
out a little bit to realize that each of these chapters emphasizes a different aspect
of the use of THz radiation. Chapter 2 shows how THz radiation can be used to
study magnetic near fields of designed structures. Chapter 3 describes a new kind
of source of THz light, where as chapter 4 shows how THz emission may help us
understand some of the optical properties of BiVO4 . THz science and technology
is an exciting field to work in.
Photoelectrochemical water splitting
Recently, BiVO4 has been discovered as a promising material for water-splitting,
as it can be used as a catalyst for photoelectrochemical (PEC) water splitting
reactions. BiVO4 is an n-type semiconductor and it has a strong absorption in the
visible region. Moreover, the band edges of BiVO4 are placed in a suitable way,
i.e. Ec > Ered and Ev < Eox for the oxidation and reduction of water, illustrating
why it is considered to be a good candidate for PEC water splitting.
Photoelectrochemical (PEC) water splitting has drawn a lot of interest due to
the potential for the production of clean and renewable energy. In PEC water
splitting, hydrogen and oxygen are produced from water using sunlight and some
specific semiconductor materials. The semiconductor material acts as an anode
and a metal acts as a cathode and both, anode and cathode, are immersed in an
aqueous electrolyte, as shown in the Figure 5.1. The photogenerated holes move
towards the semiconductor-electrolyte interface and where they undergo a wateroxidation reaction. At the same time, the electrons are transferred to the metal
cathode through an external circuit and then undergo the water reduction reaction.
The oxidation and reduction reaction for water splitting are shown below.
eˉ
eˉ
h
eˉ
H2
Semiconductor
O2
Metal
Figure 5.1: Schematic of photoelectrochemical water splitting
When the electrolyte is acidic:
2H2 O + 4h+ = 4H + + O2 (Oxidation)
Chapter 5. Conclusions
4H + + 4e− = 2H2 (Reduction)
————————————————2H2 O = 2H2 + O2
When the electrolyte is basic:
4H2 O + 4e− = 2H2 + 4OH − (Oxidation)
4OH − + 4h+ = 2H2 O + O2 (Reduction)
—————————————————–
2H2 O = 2H2 + O2
80
Bibliography
[1] Y.-S. Lee, Principles of Terahertz Science and Technology. Springer, 2009.
[2] R. Miles, P. Harrison, and D. E. Lippens, Terahertz sources and systems.
Springer, 2001.
[3] T. Y. Chang, T. J. Bridges, and E. G. Burkhardt, “cw submillimeter laser
action in optically pumped methyl fluoride, methyl alcohol, and vinyl chloride gases,” Applied Physics Letters, vol. 17, no. 6, pp. 249–251, 1970.
DOI:10.1063/1.1653386
[4] G. J. Wilmink and J. E. Grundt, “Invited review article: Current state of
research on biological effects of terahertz radiation,” Journal of Infrared,
Millimeter, and Terahertz Waves, vol. 32, no. 10, pp. 1074–1122, 2011.
DOI:10.1007/s10762-011-9794-5
[5] J. F. Federici, B. Schulkin, F. Huang, D. Gary, R. Barat, F. Oliveira, and
D. Zimdars, “THz imaging and sensing for security applications – explosives,
weapons and drugs,” Semiconductor Science and Technology, vol. 20, no. 7,
pp. S266–S280, 2005. DOI:10.1088/0268-1242/20/7/018
[6] M. Walther, B. M. Fischer, and P. U. Jepsen, “Noncovalent intermolecular
forces in polycrystalline and amorphous saccharides in the far infrared,”
Chemical Physics, vol. 288, pp. 261–268, 2003.
DOI:10.1016/S03010104(03)00031-4
[7] R. Chakittakandy, “Quasi-near field terahertz spectroscopy,” Ph.D. dissertation, Delft University of Technology, The Netherlands, 2010.
[8] B. B. Hu and M. C. Nuss, “Imaging with terahertz waves,” Opt. Lett., vol. 20,
no. 16, pp. 1716–1718, 1995.
DOI:10.1364/OL.20.001716
[9] E. Pickwell, B. E. Cole, A. J. Fitzgerald, M. Pepper, and V. P. Wallace, “In vivo study of human skin using pulsed terahertz radiation,”
Physics in Medicine and Biology, vol. 49, no. 9, pp. 1595–1607, 2004.
DOI:10.1088/0031-9155/49/9/001
[10] C. Yu, S. Fan, Y. Sun, and E. Pickwell-MacPherson, “The potential of terahertz imaging for cancer diagnosis: A review of investigations to date,”
Bibliography
82
Quantitative Imaging in Medicine and Surgery, vol. 2, no. 1, 2012.
DOI:10.3978/j.issn.2223-4292.2012.01.04
[11] C. Zhang, K. Mu, X. Jiang, Y. Jiao, L. Zhang, Q. Zhou, Y. Zhang, J. Shen,
G. Zhao, and X.-C. Zhang, “Identification of explosives and drugs and inspection of material defects with THz radiation,” Proc. SPIE, vol. 6840,
2008. DOI:10.1117/12.760133
[12] W. R. Tribe, D. A. Newnham, P. F. Taday, and M. C. Kemp, “Hidden
object detection: security applications of terahertz technology,” Proc. SPIE,
vol. 5354, pp. 168–176, 2004.
DOI:10.1117/12.543049
[13] A. J. L. Adam, P. C. M. Planken, S. Meloni, and J. Dik, “Terahertz imaging
of hidden paint layers on canvas,” Opt. Express, vol. 17, no. 5, pp. 3407–3416,
2009. DOI:10.1364/OE.17.003407
[14] E. Abraham, A. Younus, J. Delagnes, and P. Mounaix, “Non-invasive investigation of art paintings by terahertz imaging,” Applied Physics A, vol. 100,
no. 3, pp. 585–590, 2010.
DOI:10.1007/s00339-010-5642-z
[15] S. Hunsche, M. Koch, I. Brener, and M. Nuss, “THz near-field imaging,” Optics Communications, vol. 150, no. 1-6, pp. 22–26, 1998.
DOI:10.1016/S0030-4018(98)00044-3
[16] Q. Chen, Z. Jiang, G. X. Xu, and X.-C. Zhang, “Near-field terahertz imaging
with a dynamic aperture,” Opt. Lett., vol. 25, no. 15, pp. 1122–1124, 2000.
DOI:10.1364/OL.25.001122
[17] M. J. Fitch and R. Osiander, “Terahertz waves for communications and
sensing,” John Hopkins APL Technical Dig., vol. 25, no. 4, pp. 348–355,
2004.
[18] I. Wilke and S. Sengupta, Nonlinear Optical Techniques for Terahertz Pulse
Generation and Detection–Optical Rectification and Electrooptic Sampling.
CRC Press, 2007.
DOI:10.1201/9781420007701.ch2
[19] D. Saeedkia, Handbook of Terahertz Technology for Imaging, Sensing, and
Communications. Elsevier Science & Technology, 2013.
[20] D. H. Auston, K. P. Cheung, and P. R. Smith, “Picosecond photoconducting
hertzian dipoles,” Applied Physics Letters, vol. 45, pp. 284–286, 1984.
DOI:10.1063/1.95174
[21] X.-C. Zhang and J. Xu, Introduction to THz wave photonics.
2010.
Springer,
[22] N. Khiabani, “Modelling, design and characterisation of terahertz photoconductive antennas,” Ph.D. dissertation, University of Liverpool, United
Kingdom, 2013.
Bibliography
83
[23] G. Zhao, R. N. Schouten, N. van der Valk, W. T. Wenckebach, and P. C. M.
Planken, “Design and performance of a THz emission and detection setup
based on a semi-insulating GaAs emitter,” Review of Scientific Instruments,
vol. 73, no. 4, pp. 1715–1719, 2002. DOI:10.1063/1.1459095
[24] S. Sze and K. Ng, Physics of semiconductor device.
publication, 2007.
DOI:10.1002/0470068329
Wiley-Interscience
[25] K. Radhanpura, “All-optical terahertz generation from semiconductors: Materials and mechanism,” Ph.D. dissertation, Institute for Superconduting
and Electronic Materials, University of Wollongong, Australlia, 2012.
[26] V. Apostolopoulos and M. E. Barnes, “THz emitters based on the photodember effect,” Journal of Physics D: Applied Physics, vol. 47, p. 374002,
2014. DOI:10.1088/0022-3727/47/37/374002
[27] S. Kono, P. Gu, M. Tani, and K. Sakai, “Temperature dependence of terahertz radiation from n-type InSb and n-type InAs surfaces,” Applied Physics
B, vol. 71, pp. 901–904, 2000. DOI:10.1007/s003400000455
[28] P. K. Lockhart, “Effects of band structure and transport properties on terahertz emission from iii-v semiconductors,” Ph.D. dissertation, Rensselaer
Polytechnic Institute, USA, 2007.
[29] A. Nahata, A. S. Weling, and T. F. Heinz, “A wideband coherent terahertz spectroscopy system using optical rectification and electro-optic sampling,” Applied Physics Letters, vol. 69, no. 16, pp. 2321–2323, 1996.
DOI:10.1063/1.117511
[30] Q. Wu and X. C. Zhang, “Free space electro-optic sampling of terahertz
beams,” Applied Physics Letters, vol. 67, no. 24, pp. 3523–3525, 1995.
DOI:10.1063/1.114909
[31] N. C. J. van der Valk, “Towards terahertz microscopy,” Ph.D. dissertation,
Delft University of Technology, The Netherlands, 2005.
[32] G. Gallot and D. Grischkowsky, “Electro-optic detection of terahertz radiation,” J. Opt. Soc. Am. B, vol. 16, no. 8, pp. 1204–1212, 1999.
DOI:10.1364/JOSAB.16.001204
[33] N. C. J. van der Valk, T. Wenckebach, and P. C. M. Planken,
“Full mathematical description of electro-optic detection in optically
isotropic crystals,” J. Opt. Soc. Am. B, vol. 21, pp. 622–631, 2004.
DOI:10.1364/JOSAB.21.000622
[34] D. L. Woolard, W. R. Loerop, and M. S. Shur, Electronic Devices and Advanced Systems Technology. World Scientific Pub Co Inc, 2003.
[35] J. A. Riordan, F. G. Sun, Z. G. Lu, and X. C. Zhang, “Free-space transient
magneto-optic sampling,” Appl. Phys. Lett., vol. 71, pp. 1452–1454, 1997.
Bibliography
84
[36] X.-C. Zhang, “Terahertz wave imaging: horizons and hurdles,” Physics in
Medicine and Biology, vol. 47, no. 21, p. 3667, 2002.
[37] P. C. M. Planken, A. J. L. Adam, and D. Kim, “Terahertz near-field imaging,” in Terahertz Spectroscopy and Imaging, ser. Springer Series in Optical Sciences, K.-E. Peiponen, A. Zeitler, and M. Kuwata-Gonokami, Eds.
Springer Berlin Heidelberg, 2013, vol. 171, pp. 389–413. DOI:10.1007/9783-642-29564-5
[38] A. Adam, “Review of near-field terahertz measurement methods and their
applications,” Journal of Infrared, Millimeter, and Terahertz Waves, vol. 32,
no. 8-9, pp. 976–1019, 2011.
DOI:10.1007/s10762-011-9809-2
[39] N. C. J. van der Valk and P. C. M. Planken, “Electro-optic detection of subwavelength terahertz spot sizes in the near field of a metal
tip,” Applied Physics Letters, vol. 81, no. 9, pp. 1558–1560, 2002.
DOI:10.1063/1.1503404
[40] A. J. L. Adam, J. M. Brok, M. A. Seo, K. J. Ahn, D. S. Kim, J. H. Kang,
Q. H. Park, M. Nagel, and P. C. M. Planken, “Advanced terahertz electric near-field measurements at sub-wavelength diameter metallic apertures,”
Opt. Express, vol. 16, pp. 7407–7417, 2008. DOI:10.1364/OE.16.007407
[41] M. A. Seo, A. J. L. Adam, J. H. Kang, J. W. Lee, S. C. Jeoung, Q. H. Park,
P. C. M. Planken, and D. S. Kim, “Fourier-transform terahertz near-field
imaging of one-dimensional slit arrays: mapping of electric-field-, magneticfield-, and poynting vectors,” Opt. Express, vol. 15, no. 19, pp. 11 781–11 789,
2007. DOI:10.1364/OE.15.011781
[42] M. A. Seo, A. J. L. Adam, J. H. Kang, J. W. Lee, K. J. Ahn, Q. H. Park,
P. C. M. Planken, and D. S. Kim, “Near field imaging of terahertz focusing
onto rectangular apertures,” Opt. Express, vol. 16, no. 25, pp. 20 484–20 489,
2008. DOI:10.1364/OE.16.020484
[43] L. Guestin, A. J. L. Adam, J. R. Knab, M. Nagel, and P. C. M. Planken,
“Influence of the dielectric substrate on the terahertz electric near-field
of a hole in a metal,” Opt. Express, vol. 17, pp. 17 412–17 425, 2009.
DOI:10.1364/OE.17.017412
[44] L. T. Vuong, A. J. L. Adam, J. M. Brok, P. C. M. Planken, and H. P.
Urbach, “Electromagnetic spin-orbit interactions via scattering of subwavelength apertures,” Phys. Rev. Lett., vol. 104, p. 083903, 2010.
DOI:10.1103/PhysRevLett.104.083903
[45] J. R. Knab, A. J. L. Adam, M. Nagel, E. Shaner, M. A. Seo, D. S. Kim, and
P. C. M. Planken, “Terahertz near-field vectorial imaging of subwavelength
apertures and aperture arrays,” Opt. Express, vol. 17, no. 17, pp. 15 072–
15 086, 2009.
DOI:10.1364/OE.17.015072
Bibliography
85
[46] J. R. Knab, A. J. L. Adam, R. Chakkittakandy, and P. C. M. Planken,
“Terahertz near-field microspectroscopy,” Applied Physics Letters, vol. 97,
no. 3, 2010. DOI:10.1063/1.3467192
[47] J. R. Knab, A. J. L. Adam, E. Shaner, H. J. A. J. Starmans, and P. C. M.
Planken, “Terahertz near-field spectroscopy of filled subwavelength sized
apertures in thin metal films,” Opt. Express, vol. 21, pp. 1101–1112, 2013.
DOI:10.1364/OE.21.001101
[48] A. Bitzer, H. Merbold, A. Thoman, T. Feurer, H. Helm, and M. Walther,
“Terahertz near-field imaging of electric and magnetic resonances of a planar metamaterial,” Opt. Express, vol. 17, no. 5, pp. 3826–3834, 2009.
DOI:10.1364/OE.17.003826
[49] A. Bitzer, A. Ortner, H. Merbold, T. Feurer, and M. Walther, “Terahertz near-field microscopy of complementary planar metamaterials: Babinet’s principle,” Opt. Express, vol. 19, no. 3, pp. 2537–2545, 2011.
DOI:10.1364/OE.19.002537
[50] Y. Liu and X. Zhang, “Metamaterials: a new frontier of science
and technology,” Chem. Soc. Rev., vol. 40, pp. 2494–2507, 2011.
DOI:10.1039/C0CS00184H
[51] L. A. Butler, “Design, simulation, fabrication and characteristics of terahertz metamaterial devices,” Ph.D. dissertation, University of Alabama,
USA, 2012.
[52] V. G. Veselago, “The electrodynamics of substances with simultaneously
negative values of and µ,” SOV PHYS USPEKHI, vol. 10, p. 509514, 1968.
DOI:10.1070/PU1968v010n04ABEH003699
[53] J. B. Pendry, A. J. Holden, W. J. Stewart, and I. Youngs, “Extremely low
frequency plasmons in metallic mesostructures,” Phys. Rev. Lett., vol. 76,
pp. 4773–4776, 1996. DOI:10.1103/PhysRevLett.76.4773
[54] J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J. Stewart, “Low frequency plasmons in thin-wire structures,” Journal of Physics: Condensed
Matter, vol. 10, no. 22, p. 4785, 1998. DOI:10.1088/0953-8984/10/22/007
[55] J. Pendry, A. Holden, D. Robbins, and W. Stewart, “Magnetism from conductors and enhanced nonlinear phenomena,” Microwave Theory and Techniques, IEEE Transactions on, vol. 47, no. 11, pp. 2075–2084, Nov 1999.
DOI:10.1109/22.798002
[56] D. R. Smith, W. J. Padilla, D. C. Vier, S. C. Nemat-Nasser, and
S. Schultz, “Composite medium with simultaneously negative permeability and permittivity,” Phys. Rev. Lett., vol. 84, pp. 4184–4187, 2000.
DOI:10.1103/PhysRevLett.84.4184
Bibliography
86
[57] R. A. Shelby, D. R. Smith, and S. Schultz, “Experimental verification of a
negative index of refraction,” Science, vol. 292, no. 5514, pp. 77–79, 2001.
DOI:10.1126/science.1058847
[58] A. Sihvola, Handedness in Plasmonics: Electrical Engineer’s Perspective.
Springer, 2009.
[59] W.
Withayachumnankul,
“Metamaterials
regime,” Photonics Journal, IEEE, vol. 1,
DOI:10.1109/JPHOT.2009.2026288
in
the
terahertz
pp. 99–118, 2001.
[60] K. Aydn, “Characterization and applications of negative-index metamaterials,” Ph.D. dissertation, Bilkent University, Turkey, June 2008.
[61] S. Linden, C. Enkrich, G. Dolling, M. Klein, J. Zhou, T. Koschny,
C. Soukoulis, S. Burger, F. Schmidt, and M. Wegener, “Photonic metamaterials: Magnetism at optical frequencies,” IEEE Journal of Selected Topics in Quantum Electronics, vol. 12, pp. 1097–1105, 2006.
DOI:10.1109/JSTQE.2006.880600
[62] B. Lahiri, “Characaterization and applications of negative-index metamaterials,” Ph.D. dissertation, University of Glasgow, Scotland, UK, March
2010.
[63] Y. Kawano and K. Ishibash, “An on-chip near-field terahertz
probe and detector,” Nat. Photon., vol. 2, pp. 618–621, 2008.
DOI:10.1038/nphoton.2008.157
[64] M. Burresi, D. van Oosten, T. Kampfrath, H. Schoenmaker, R. Heideman, A. Leinse, and L. Kuipers, “Probing the magnetic field of
light at optical frequencies,” Science, vol. 326, pp. 550–553, 2009.
DOI:10.1126/science.1177096
[65] E. Devaux, A. Dereux, E. Bourillot, J. C. Weeber, Y. Lacroute, and J. P.
Goudonnet, “Local detection of the optical magnetic field in the near zone
of dielectric samples,” Phys. Rev. B, vol. 62, pp. 10 504–10 514, 2000.
DOI:10.1103/PhysRevB.62.10504
[66] H. W. Kihm, S. M. Koo, Q. H. Kim, K. Bao, J. E. Kihm, W. S. Bak, S. H.
Eah, C. Lienau, H. Kim, P. Nordlander, N. J. Halas, N. K. Park, and D. S.
Kim, “Bethe-hole polarization analyser for the magnetic vector of light,”
Nat. Commun., vol. 2, p. 451, 2011. DOI:10.1038/ncomms1430
[67] C. C. Chen and J. F. Whitaker, “An optically-interrogated microwavepoynting-vector sensor using cadmium manganese telluride,” Opt. Express,
vol. 18, pp. 12 239–12 248, 2010.
DOI:10.1364/OE.18.012239
[68] J. Riordan and X. C. Zhang, “Sampling of free-space magnetic pulses,”
Optical and Quantum Electronics, vol. 32, pp. 489–502, 2000.
DOI:10.1023/A:1007066809933
Bibliography
87
[69] W. J. Padilla, A. J. Taylor, C. Highstrete, M. Lee, and R. D.
Averitt, “Dynamical electric and magnetic metamaterial response at
terahertz frequencies,” Phys. Rev. Lett., vol. 96, p. 107401, 2006.
DOI:10.1103/PhysRevLett.96.107401
[70] A. Sell, A. Leitenstorfer, and R. Huber, “Phase-locked generation and
field-resolved detection of widely tunable terahertz pulses with amplitudes exceeding 100 MV/cm,” Opt. Lett., vol. 33, pp. 2767–2769, 2008.
DOI:10.1364/OL.33.002767
[71] F. Falcone, T. Lopetegi, J. Baena, R. Marques, F. Martin, and M. Sorolla,
“Effective negative–ε stopband microstrip lines based on complementary
split ring resonators,” Microwave and Wireless Compon, vol. 14, pp. 280–
282, 2004.
DOI:10.1109/LMWC.2004.828029
[72] H. T. Chen, J. F. O’Hara, A. J. Taylor, and R. D. Averitt, “Complementary
planar terahertz metamaterials,” Opt. Express, vol. 15, pp. 1084–1095, 2007.
DOI:10.1364/OE.15.001084
[73] N. T. Messiha, A. M. Ghuniem, and H. M. ELHennawy, “Numerical studies
of transmission characteristics of omega and omega-like structures and their
dual counterparts,” Journal of Basic and Applied Physics, vol. 3, pp. 48–53,
2014.
[74] F. Falcone, T. Lopetegi, M. Laso, J. Baena, J. Bonache, M. Beruete, R. Marques, F. Martin, and M. Sorolla, “Babinet principle applied to the design of
metasurfaces and metamaterials,” Phys. Rev. Lett., vol. 93, p. 197401, 2004.
DOI:10.1103/PhysRevLett.93.197401
[75] H.-T. Chen, J. O’Hara, A. Azad, and A. Taylor, “Manipulation of terahertz
radiation using metamaterials,” Laser & Photonics Reviews, vol. 5, no. 4,
pp. 513–533, 2011. DOI:10.1002/lpor.201000043
[76] G. I. Likhtenshtein, J. Yamauchi, S. Nakatsuji, A. I. Smirnov, and
R. Tamura, Fundamentals of Magnetism. Wiley-VCH Verlag GmbH &
Co. KGaA, 2008.
DOI:10.1002/9783527621743.ch1
[77] J. P. Jakubovics, Magnetism and magnetic materials. Institute of Materials,
1994. DOI:10.1002/9783527621743.ch1
[78] M. Getzlaff, Fundamentals of Magnetism.
DOI:10.1007/978-3-540-31152-2
Springer,
2007.
[79] D. C. Jiles, Introduction to Magnetism and Magnetic Materials, Second Edition. CRC Press, 1998.
[80] N. A. Spaldini, Magnetic Materials: Fundamentals and Device Applications.
Cambridge University Press, 2003.
[81] J. Stohr and H. C. Siegmann, Magnetism: From Fundamentals to Nanoscale
Dynamics. Springer, 2006.
DOI:10.1007/978-3-540-30283-4
Bibliography
88
[82] R. J. Hicken, “Ultrafast nanomagnets: seeing data storage in a new
light,” Phil. Trans. R. Soc. Lond. A, vol. 361, pp. 2827–2841, 2003.
DOI:10.1098/rsta.2003.1285
[83] E. Beaurepaire, G. M. Turner, S. M. Harrel, M. C. Beard, J.-Y. Bigot,
and C. A. Schmuttenmaer, “Coherent terahertz emission from ferromagnetic
films excited by femtosecond laser pulses,” Applied Physics Letters, vol. 84,
no. 18, pp. 3465–3467, 2004.
DOI:10.1063/1.1737467
[84] M. B. Agranat, S. I. Ashitkov, A. B. Granovskii, and G. I. Rukman, “Interaction of picosecond laser pulses with the electron, spin, and phonon
subsystems of nickel,” JETP, vol. 86, pp. 1376–1379, 1984.
[85] A. Vaterlaus, T. Beutler, and F. Meier, “Spin-lattice relaxation time
of ferromagnetic gadolinium determined with time-resolved spin-polarized
photoemission,” Phys. Rev. Lett., vol. 67, pp. 3314–3317, 1991.
DOI:10.1103/PhysRevLett.67.3314
[86] A. Vaterlaus, T. Beutler, D. Guarisco, M. Lutz, and F. Meier,
“Spin-lattice relaxation in ferromagnets studied by time-resolved spinpolarized photoemission,” Phys. Rev. B, vol. 46, pp. 5280–5286, 1992.
DOI:10.1103/PhysRevB.46.5280
[87] E. Beaurepaire, J.-C. Merle, A. Daunois, and J.-Y. Bigot, “Ultrafast spin
dynamics in ferromagnetic nickel,” Phys. Rev. Lett., vol. 76, pp. 4250–4253,
1996. DOI:10.1103/PhysRevLett.76.4250
[88] J. Hohlfeld, E. Matthias, R. Knorren, and K. H. Bennemann, “Nonequilibrium magnetization dynamics of nickel,” Phys. Rev. Lett., vol. 78, pp.
4861–4864, 1997.
DOI:10.1103/PhysRevLett.78.4861
[89] A. Scholl, L. Baumgarten, R. Jacquemin, and W. Eberhardt, “Ultrafast
spin dynamics of ferromagnetic thin films observed by fs spin-resolved twophoton photoemission,” Phys. Rev. Lett., vol. 79, pp. 5146–5149, 1997.
DOI:10.1103/PhysRevLett.79.5146
[90] C. Stamm, T. Kachel, N. Pontius, R. Mitzner, T. Quast, K. Holldack,
S. Khan, C. Lupulescu, E. F. Aziz, M. Wietstruk, H. A. Durr, and W. Eberhardt, “Femtosecond modification of electron localization and transfer of
angular momentum in nickel,” Nat Mater, vol. 6, no. 10, pp. 740–743, 2007.
DOI:10.1038/nmat1985
[91] H. E. Elsayed-Ali, T. Juhasz, G. O. Smith, and W. E. Bron, “Femtosecond
thermoreflectivity and thermotransmissivity of polycrystalline and singlecrystalline gold films,” Phys. Rev. B, vol. 43, pp. 4488–4491, 1991.
DOI:10.1103/PhysRevB.43.4488
[92] G. P. Zhang and W. Hübner, “Laser-induced ultrafast demagnetization in
ferromagnetic metals,” Phys. Rev. Lett., vol. 85, pp. 3025–3028, 2000.
DOI:10.1103/PhysRevLett.85.3025
Bibliography
89
[93] F. Dalla Longa, J. T. Kohlhepp, W. J. M. de Jonge, and B. Koopmans, “Influence of photon angular momentum on ultrafast demagnetization in nickel,” Phys. Rev. B, vol. 75, p. 224431, 2007.
DOI:10.1103/PhysRevB.75.224431
[94] B. Koopmans, H. Kicken, M. van Kampen, and W. de Jonge, “Microscopic model for femtosecond magnetization dynamics,” Journal of Magnetism and Magnetic Materials, vol. 286, pp. 271 – 275, 2005, proceedings of the 5th International Symposium on Metallic Multilayers.
DOI:10.1016/j.jmmm.2004.09.079
[95] I. Radu, G. Woltersdorf, M. Kiessling, A. Melnikov, U. Bovensiepen, J.U. Thiele, and C. H. Back, “Laser-induced magnetization dynamics of
lanthanide-doped permalloy thin films,” Phys. Rev. Lett., vol. 102, p. 117201,
2009. DOI:10.1103/PhysRevLett.102.117201
[96] M. Battiato, K. Carva, and P. M. Oppeneer, “Superdiffusive spin transport
as a mechanism of ultrafast demagnetization,” Phys. Rev. Lett., vol. 105, p.
027203, 2010.
DOI:10.1103/PhysRevLett.105.027203
[97] A. Eschenlohr, M. Battiato, P. Maldonado, N. Pontius, T. Kachel, K. Holldack, R. Mitzner, A. Fhlisch, P. M. Oppeneer, and C. Stamm, “Ultrafast
spin transport as key to femtosecond demagnetization,” Nat Mater, vol. 12,
pp. 332–336, 2013. DOI:10.1038/nmat3546
[98] A. J. Schellekens, W. Verhoeven, T. N. Vader, and B. Koopmans, “Investigating the contribution of superdiffusive transport to ultrafast demagnetization of ferromagnetic thin films,” Applied Physics Letters, vol. 102, no. 25,
p. 252408, 2013. DOI:10.1063/1.4812658
[99] K. Sakai, Terahertz optoelectronics.
Springer, 2005.
[100] G. K. P. Ramanandan, G. Ramakrishnan, N. Kumar, A. J. L. Adam, and
P. C. M. Planken, “Emission of terahertz pulses from nanostructured metal
surfaces,” Journal of Physics D: Applied Physics, vol. 47, no. 37, p. 374003,
2014. DOI:10.1088/0022-3727/47/37/374003
[101] F. Kadlec, P. Kužel, and J.-L. Coutaz, “Optical rectification at
metal surfaces,” Opt. Lett., vol. 29, no. 22, pp. 2674–2676, 2004.
DOI:10.1364/OL.29.002674
[102] G. Ramakrishnan, N. Kumar, P. C. M. Planken, D. Tanaka, and K. Kajikawa, “Surface plasmon-enhanced terahertz emission from a hemicyanine
self-assembled monolayer,” Opt. Express, vol. 20, no. 4, pp. 4067–4073, 2012.
DOI:10.1364/OE.20.004067
[103] E. V. Suvorov, R. A. Akhmedzhanov, D. A. Fadeev, I. E. Ilyakov, V. A.
Mironov, and B. V. Shishkin, “Terahertz emission from a metallic surface
induced by a femtosecond optic pulse,” Opt. Lett., vol. 37, no. 13, pp. 2520–
2522, 2012.
DOI:10.1364/OL.37.002520
Bibliography
90
[104] G. H. Welsh and K. Wynne, “Generation of ultrafast terahertz radiation
pulses on metallic nanostructured surfaces,” Opt. Express, vol. 17, no. 4, pp.
2470–2480, Feb 2009.
DOI:10.1364/OE.17.002470
[105] G. Ramakrishnan and P. C. M. Planken, “Percolation-enhanced generation
of terahertz pulses by optical rectification on ultrathin gold films,” Opt. Lett.,
vol. 36, no. 13, pp. 2572–2574, 2011.
DOI:10.1364/OL.36.002572
[106] D. K. Polyushkin, E. Hendry, E. K. Stone, and W. L. Barnes, “THz generation from plasmonic nanoparticle arrays,” Nano Letters, vol. 11, no. 11, pp.
4718–4724, 2011.
DOI:10.1021/nl202428g
[107] K. Kajikawa, Y. Nagai, Y. Uchiho, G. Ramakrishnan, N. Kumar, G. K. P.
Ramanandan, and P. C. M. Planken, “Terahertz emission from surfaceimmobilized gold nanospheres,” Opt. Lett., vol. 37, no. 19, pp. 4053–4055,
2012. DOI:10.1364/OL.37.004053
[108] J. Shen, X. Fan, Z. Chen, M. F. DeCamp, H. Zhang, and J. Q. Xiao, “Damping modulated terahertz emission of ferromagnetic films excited by ultrafast
laser pulses,” Applied Physics Letters, vol. 101, no. 7, p. 072401, 2012.
DOI:10.1063/1.4737400
[109] Y. Martin and H. K. Wickramasinghe, “Magnetic imaging by force microscopy with 1000 Å resolution,” Applied Physics Letters, vol. 50, no. 20,
pp. 1455–1457, 1987. DOI:10.1063/1.97800
[110] C. Hsieh, J. Liu, and J. Lue, “Magnetic force microscopy studies of domain
walls in nickel and cobalt films,” Applied Surface Science, vol. 252, no. 5,
pp. 1899 – 1909, 2005.
DOI:10.1016/j.apsusc.2005.05.041
[111] M. Hehn, S. Padovani, K. Ounadjela, and J. P. Bucher, “Nanoscale magnetic
domain structures in epitaxial cobalt films,” Phys. Rev. B, vol. 54, pp. 3428–
3433, 1996.
DOI:10.1103/PhysRevB.54.3428
[112] J. Brandenburg, R. Hühne, L. Schultz, and V. Neu, “Domain structure of
epitaxial Co films with perpendicular anisotropy,” Phys. Rev. B, vol. 79, p.
054429, 2009.
DOI:10.1103/PhysRevB.79.054429
[113] N. Saito, H. Fujiwara, and Y. Sugita, “A new type of magnetic domain structure in negative magnetostriction Ni-Fe films,” Journal of
the Physical Society of Japan, vol. 19, no. 7, pp. 1116–1125, 1964.
DOI:10.1143/JPSJ.19.1116
[114] D. J. Hilton, R. D. Averitt, C. A. Meserole, G. L. Fisher, D. J. Funk, J. D.
Thompson, and A. J. Taylor, “Terahertz emission via ultrashort-pulse excitation of magnetic metalfilms,” Opt. Lett., vol. 29, no. 15, pp. 1805–1807,
2004. DOI:10.1364/OL.29.001805
[115] T. L. Gilbert, “A phenomenological theory of damping in ferromagnetic materials,” IEEE Trans. Magn., vol. 40, 2004.
DOI:10.1109/TMAG.2004.836740
Bibliography
91
[116] E. Barati, M. Cinal, D. M. Edwards, and A. Umerski, “Calculation of Gilbert
damping in ferromagnetic films,” EPJ Web of Conferences, vol. 40, p. 18003,
2013. DOI:10.1051/epjconf/20134018003
[117] B. Koopmans, J. Ruigrok, F. Longa, and W. de Jonge, “Unifying ultrafast magnetization dynamics,” Phys. Rev. Lett., vol. 95, p. 267207, 2005.
DOI:10.1103/PhysRevLett.95.267207
[118] D. E. S. Stanescu, “Magnetization dynamics in nanostructures,” Ph.D. dissertation, Universite Joseph Fourier, France, 2003.
[119] E. Šimánek, “Gilbert damping in ferromagnetic films due to adjacent normal-metal layers,” Phys. Rev. B, vol. 68, p. 224403, 2003.
DOI:10.1103/PhysRevB.68.224403
[120] G. Malinowski, F. Dalla Longa, J. H. H. Rietjens, P. V. Paluskar, R. Huijink,
H. J. M. Swagten, and B. Koopmans, “Control of speed and efficiency of
ultrafast demagnetization by direct transfer of spin angular momentum,”
Nat Phys, vol. 4, pp. 855–858, 2008. DOI:10.1038/nphys1092
[121] A. Barman, S. Wang, O. Hellwig, A. Berger, E. E. Fullerton, and
H. Schmidt, “Ultrafast magnetization dynamics in high perpendicular
anisotropy [Co/Pt]n multilayers,” Journal of Applied Physics, vol. 101, no. 9,
2007. DOI:10.1063/1.2709502
[122] S. Jian, Z. Huai-Wu, and L. Yuan-Xun, “Terahertz emission of ferromagnetic
Ni-Fe thin films excited by ultrafast laser pulses,” Chinese Physics Letters,
vol. 29, no. 6, p. 067502, 2012.
DOI:10.1088/0256-307X/29/6/067502
[123] T. Kampfrath, M. Battiato, P. Maldonado, G. Eilers, J. Notzold,
S. Mahrlein, V. Zbarsky, F. Freimuth, Y. Mokrousov, S. Blugel, M. Wolf,
I. Radu, P. M. Oppeneer, and M. Munzenberg, “Terahertz spin current
pulses controlled by magnetic heterostructures,” Nat Nano, vol. 8, pp. 256–
260, 2013.
DOI:10.1038/nnano.2013.43
[124] Y. Tserkovnyak, A. Brataas, and G. Bauer, “Enhanced gilbert damping
in thin ferromagnetic films,” Phys. Rev. Lett., vol. 88, p. 117601, 2002.
DOI:10.1103/PhysRevLett.88.117601
[125] C. H. Back, R. Allenspach, W. Weber, S. S. P. Parkin, D. Weller, E. L.
Garwin, and H. C. Siegmann, “Minimum field strength in precessional
magnetization reversal,” Science, vol. 285, no. 5429, pp. 864–867, 1999.
DOI:10.1126/science.285.5429.864
[126] E. B. Myers, D. C. Ralph, J. A. Katine, R. N. Louie, and R. A.
Buhrman, “Current-induced switching of domains in magnetic multilayer devices,” Science, vol. 285, no. 5429, pp. 867–870, 1999.
DOI:10.1126/science.285.5429.867
Bibliography
92
[127] G. Ramakrishnan, “Enhanced terahertz emission from thin film semiconductor/metal interfaces,” Ph.D. dissertation, Delft University of Technology,
The Netherlands, 2012.
[128] B. Ferguson and X.-C. Zhang, “Materials for terahertz science and technology,” Nat Mater., vol. 1, pp. 26–33, 2002. DOI:doi:10.1038/nmat708
[129] Y. Jin, X. F. Ma, G. A. Wagoner, M. Alexander, and X.-C. Zhang, “Anomalous optically generated THz beams from metal/GaAs interfaces,” Appl.
Phys. Lett., vol. 65, pp. 682–684, 1994. DOI:10.1063/1.112267
[130] G. Ramakrishnan, G. K. P. Ramanandan, A. J. L. Adam, M. Xu, N. Kumar,
R. W. A. Hendrikx, and P. C. M. Planken, “Enhanced terahertz emission
by coherent optical absorption in ultrathin semiconductor films on metals,”
Opt. Express, vol. 21, pp. 16 784–16 798, 2013. DOI:10.1364/OE.21.016784
[131] H. Dember, “Photoelectromotive force in cuprous oxide crystals,” Phys. Z.,
vol. 32, pp. 554–556, 1931.
[132] D. A. Neamen, Semiconductor Physics And Devices: Basic Principles
(Fourth Edition). McGraw-hil, 2011.
[133] C. Wolfe, N. Holonyak, and G. Stillman, Physical properties of semiconductors, ser. Solid state physical electronics series. Prentice Hall, 1989.
[134] P. S. Parasuraman, “Mod-01 lec-09 metal-semiconductor junctions.” [Online]. http://textofvideo.nptel.iitm.ac.in/showVideo.php?v=CT6olzelSKQ
[135] K. Liu, J. Xu, T. Yuan, and X.-C. Zhang, “Terahertz radiation from InAs
induced by carrier diffusion and drift,” Phys. Rev. B, vol. 73, no. 155330,
2006. DOI:10.1103/PhysRevB.73.155330
[136] J. N. Heyman, N. Coates, A. Reinhardt, and G. Strasser, “Diffusion and
drift in terahertz emission at GaAs surfaces,” Appl. Phys. Lett., vol. 83, pp.
5476–5478, 2003.
DOI:10.1063/1.1636821
[137] A. Urbanowicz, R. Adomavicius, and A. Krotkus, “Terahertz emission from
photoexcited surfaces of Ge crystals,” Physica B, vol. 367, pp. 152–157, 2005.
DOI:10.1016/j.physb.2005.06.010
[138] R. Ascazubi, C. Shneider, I. Wilke, R. Pino, and P. S. Dutta, “Enhanced
terahertz emission from impurity compensated GaSb,” Phys. Rev. B, vol. 72,
no. 045328, 2005. DOI:10.1103/PhysRevB.72.045328
[139] X.-C. Zhang and D. H. Auston, “Optoelectronic measurement of semiconductor surfaces and interfaces with femtosecond optics,” J. Appl. Phys.,
vol. 71, no. 326, 1992.
DOI:10.1063/1.350710
[140] R. Ascazubi, I. Wolke, K. Denniston, H. Lu, and W. Scha, “Terahertz emission by InN,” Appl. Phys. Lett., vol. 84, no. 4810, 2004.
DOI:10.1063/1.1759385
Bibliography
93
[141] M. Nakajima, Y. Oda, and T. Suemoto, “Competing terahertz radiation mechanisms in semi-insulating InP at high-density excitation,”
Applied Physics Letters, vol. 85, no. 14, pp. 2694–2696, 2004.
DOI:10.1063/1.1796532
[142] T. Dekorsy, H. Auer, H. J. Bakker, H. G. Roskos, and H. Kurz, “THz electromagnetic emission by coherent infrared-active phonons,” Phys. Rev. B,
vol. 53, pp. 4005–4014, 1996.
DOI:10.1103/PhysRevB.53.4005
[143] P. Gu, M. Tani, S. Kono, K. Sakai, and X.-C. Zhang, “Study of terahertz
radiation from InAs and InSb,” J. Appl. Phys, vol. 91, no. 5533, 2002.
DOI:10.1063/1.1465507
[144] M. B. Johnston, D. Whittaker, A. Corchia, A. Davies, and E. Lineld, “Simulation of terahertz generation at semiconductor surfaces,” Phys. Rev. B,
vol. 65, no. 165301, 2002.
DOI:10.1103/PhysRevB.65.165301
[145] G. Klatt, F. Hilser, W. Qiao, M. Beck, R. Gebs, A. Bartels, K. Huska,
U. Lemmer, G. Bastian, M. B. Johnston, M. Fischer, J. Faist, and T. Dekorsy, “Terahertz emission from lateral photo-Dember currents,” Opt. Express,
vol. 18, pp. 4939–4947, 2010.
DOI:10.1364/OE.18.004939
[146] Y. Zhao, Y. Xie, X. Zhu, S. Yan, and S. Wang, “Surfactant-free synthesis
of hyperbranched monoclinic bismuth vanadate and its applications in photocatalysis, gas sensing, and lithium-ion batteries,” Chemistry– A European
Journal, vol. 14, pp. 1601–1606, 2008. DOI:10.1002/chem.200701053
[147] Z. Zhao, Z. Li, and Z. Zou, “Electronic structure and optical properties of
monoclinic clinobisvanite BiVO4 ,” Phys. Chem. Chem. Phys., vol. 13, pp.
4746–4753, 2011.
DOI:10.1039/C0CP01871F
[148] R. M. Hazen and J. W. E. Mariathasan, “Bismuth vanadate: A
high-pressure, high-temperature crystallographic study of the ferroelasticparaelastic transition,” Science, vol. 216, pp. 991–993, 1982.
DOI:10.1126/science.216.4549.991
[149] X. Wang, J. D. G. Li, H. Peng, and K. Chen, “Facile synthesis and photocatalytic activity of monoclinic BiVO4 micro/nanostructures with controllable morphologies,” Mater. Res. Bull., vol. 47, pp. 3814–3818, 2012.
DOI:10.1016/j.materresbull.2012.04.082
[150] S. Tokunaga, H. Kato, and A. Kudo, “Selective preparation of monoclinic and tetragonal BiVO4 with scheelite structure and their photocatalytic properties,” Chem Mater., vol. 13, pp. 4624–4628, 2001.
DOI:10.1021/cm0103390
[151] B. Cheng, W. Wang, L. Shi, J. Zhang, J. Ran, and H. Yu, “One-pot templatefree hydrothermal synthesis of monoclinic BiVO4 hollow microspheres and
their enhanced visible-light photocatalytic activity,” Int. J. Photoenergy, no.
797968, 2012.
DOI:10.1155/2012/797968
Bibliography
94
[152] J. K. Cooper, S. Gul, F. M. Toma, L. Chen, P.-A. Glans, J. Guo, J. W. Ager,
J. Yano, and I. D. Sharp, “Electronic structure of monoclinic BiVO4 ,” Chemistry of Materials, vol. 26, pp. 5365–5373, 2014. DOI:10.1021/cm5025074
[153] F. F. Abdi, N. Firet, A. Dabirian, and R. van de Krol, “Spray-deposited copi catalyzed BiVO4 : a low-cost route towards highly efficient photoanodes,”
Mater. Res. Soc. Symp. Proc., vol. 1446, 2012. DOI:10.1557/opl.2012.811
[154] F. F. Abdi, N. Firet, and R. van de Krol, “Efficient BiVO4 thin film photoanodes modified with cobalt phosphate catalyst and W-doping,” ChemCatChem, vol. 5, pp. 490–496, 2013. DOI:10.1002/cctc.201200472
[155] F. F. Abdi, L. Han, A. H. M. Smets, M. Zeman, B. Dam, and R. de Krol,
“Efficient solar water splitting by enhanced charge separation in a bismuth
vanadate-silicon tandem photoelectrode,” Nat. Comm., vol. 4, no. 2195,
2013. DOI:10.1038/ncomms3195
[156] S. S. Dunkle, R. J. Helmich, and K. S. Suslick, “BiVO4 as a visible-light photocatalyst prepared by ultrasonic spray pyrolysis,” Mater. Res. Soc. Symp.
Proc., no. 1446, 2012.
DOI:10.1021/jp903757x
[157] P. Hoyer, M. Theuer, R. Beigang, and E.-B. Kley, “Terahertz emission
from black silicon,” Applied Physics Letters, vol. 93, p. 091106, 2008.
DOI:10.1063/1.2978096
[158] V. L. Malevich, R. Adomavicius, and A. Krotkus, “THz emission from semiconductor surfaces,” Comptes Rendus Physique, vol. 9, pp. 130–141, 2008.
DOI:10.1016/j.crhy.2007.09.014
Summary
Terahertz radiation is electromagnetic waves with frequencies from 0.1-10 THz.
THz radiation can pass through cardboard, paper, plastics, ceramics and many
other materials. Hence, it can be used for non-destructive imaging. Another important application of THz radiation is spectroscopy. Many organic molecules
absorb light at THz frequencies and these absorption lines can be used for the
identification of the molecules. This spectroscopic technique is called terahertz
time domain spectroscopy (THz-TDS). It is a valuable tool for studying the properties of the material. In THz-TDS we measure the amplitude and phase of the
THz pulse in the time domain using coherent detection techniques. Usually, in
THz-TDS technique we measure the THz electric field using electro-optic detention technique. However, in thesis, the main goal is to focus on the magnetic
aspect of THz generation and detection using THz-TDS.
This thesis is divided into three research problems, in which THz-TDS plays
the key role. In the first part, the THz-TDS setup is used for characterising the
metamaterial elements. Metamaterials are artificially structured materials that
are used to control and manipulate light. A split-ring resonator is one of the most
common metamaterial elements. Usually these split-ring resonators are studied in
far-field but near-field interactions are important to understand the properties of
metamaterials. In the past electric near-field of SRRs are already studied. For the
first time, we have directly measured the magnetic near-field of SRRs.
The second research problem investigated in the thesis is generation of THz
radiation from ferromagnetic cobalt thin film. When femtosecond laser pulses
are incident on ferromagnetic metals prepared on a glass substrate, THz pulses
are emitted via ultrafast demagnetization. It is often forgotten that nonmagnetic
metals are also capable of emitting THz light and that such a contribution to
the emitted THz field cannot a priori be excluded for magnetic metals. Our work
clearly establishes a strong correlation between the magnetization and the emission
of THz light following the excitation of cobalt with a femtosecond laser pulse. It
does this by highlighting the role of the orientation of the magnetization in the
terahertz emission from ferromagnetic thin films. We find that as we increase the
cobalt film thickness, the polarisation direction of the emitted THz pulse changes,
correlating with the transition from a predominantly in-plane to a predominantly
out-plane magnetisation, as measured with magnetic force microscopy.
When femtosecond laser light is incident on a semiconductor thin film, emission
Bibliography
96
of THz radiation is observed. The emission can be enhanced if semiconductor
materials are deposited on metal surfaces. In the last research problem of this
thesis, in chapter 5, THz spectroscopy was used for studying the BiVO4/Au interfaces. BiVO4 is a semiconductor material which is widely used for water splitting.
When a BiVO4/Au interface is illuminated with ultrashort laser pulses, due to reflection from various interfaces, a standing wave pattern is observed. As a result,
a difference in carrier concentration builds up which gives rise to THz emitting
dipole.
In short, this thesis discusses the possibilities of using terahertz time domain
spectroscopy for studying the generation of THz radiation and using it for imaging
and material characterization.
Nishant Kumar, March 2015
Samenvatting
Terahertzstraling (THz-straling) is elektromagnetische straling met frequenties in
het gebied van 0.1 tot 10 THz. Er is recentelijk veel interesse in THz-straling vanwege haar toepassingspotentieel. THz-straling gaat door karton, papier, plastic,
keramiek en vele andere materialen die ondoorzichtig zijn voor zichtbaar licht heen.
In tegenstelling tot Rntgenstraling heeft THz-licht zeer weinig energie per foton
en heeft het dus geen invloed op biologische samples. Vele organische moleculen
absorberen licht bij specifieke THz-frequenties en deze absorptielijnen kunnen
gebruikt worden voor de identificatie van de moleculen. Hierom wordt THzstraling gebruikt voor non-destructieve beeldvormings- en spectroscopietoepassingen. Een belangrijke spectroscopische techniek die THz-straling gebruikt om de
eigenschappen van een materiaal te bestuderen wordt terahertztijddomeinspectroscopie (THz-TDS) genoemd. In THz-TDS meten we normaliter de amplitude en
fase van het elektrische veld van een THz-puls. Het hoofddoel in dit proefschrift
is echter om te focussen op de magnetischeveldaspecten van THz-straling, gebruik
makende van THz-TDS. Dit proefschrift is onderverdeeld in drie onderzoeksonderwerpen. In het eerste deel wordt de THz-TDS-opstelling gebruikt voor het
karakteriseren van split ring resonators. Een split ring resonator is een metallische
ringstructuur met n of meer onderbrekingen (vandaar split ring) op een dielektrisch
substraat. Deze split ring resonators worden gewoonlijk bestudeerd in het verreveldgebied. Het is om een fatsoenlijk begrip te verkrijgen van de eigenschappen van
de split ring resonators echter belangrijk om kennis te hebben van hun respons
in hun onmiddellijke omgeving, het zogenaamde nabijeveldgebied. Het elektrische
nabije veld van deze split ring resonators is al bestudeerd, maar het magnetische
nabije veld nog niet. Voor de eerste keer hebben wij het THz-magnetische nabije
veld gemeten met een diep subgolflengte ruimtelijke resolutie. We observeren dat
wanneer een split ring resonator wordt gexciteerd door een incidente elektromagnetische golf, er een ruimtelijk circulerende en temporeel oscillerende elektrische
stroom in de metallische ring wordt genduceerd. Deze stroom creert een tijdsafhankelijk magnetisch veld dat normaal op het vlak georinteerd is. In een enkele
split ring resonator is het magnetische nabije veld maximaal langs de lange arm,
vlakbij de hoek, en is volgens berekeningen ongeveer 200 keer sterker dan het
binnenvallende veld. We meten het magnetische veld op een afstand van 10 tot
20 µm van het oppervlak die veel kleiner is dan de afmeting van de ring en de
golflengte van de THz-straling die overeenkomt met de resonantiefrequentie van
de resonator. Dit laat zien dat we direct in het nabije veld meten met een sub-
Bibliography
98
golflengte ruimtelijke resolutie. Tevens hebben we metingen uitgevoerd aan een
dubbele split ring resonator. Voor een dubbele split ring resonator worden twee resonatoren tegen elkaar aan geplaatst zodat ze de middelste arm delen. De sterkste
stroom loopt langs de middelste arm en is veel sterker dan de stroom in de enkele
split ring resonator. Door toedoen van de structuur van de resonator bestaan er
in de linker en rechter ring oscillerende stromen met de klok mee en tegen de klok
in en daarom zijn de richtingen van de magnetische velden in beide resonatoren
tegengesteld. De magnetische nabijeveldcomponent wijst het vlak in voor de linker
ring en het vlak uit voor de rechter ring. We hebben ook het magnetische veld van
de dubbele split ring resonator in het nabije veld met een diep subgolflengte resolutie gemeten. De resultaten laten zien dat het THz-magnetische veld van de split
ring resonators in het vlak van de structuur hoofdzakelijk geconcentreerd is bij de
rand van het metaal. Naarmate we van de structuur weg bewegen (maar in het
nabijeveldgebied blijven), verandert de magnetischeveldverdeling geleidelijk in een
uniforme verdeling. Het tweede researchprobleem onderzocht in dit proefschrift is
de generatie van THz-straling aan ferromagnetische dunne lagen kobalt. Wanneer
femtosecondelaserpulsen binnenvallen op dunne lagen kobalt opgedampt op een
glazen substraat, wordt emissie van THz-straling waargenomen. We laten zien dat
de emissie van THz-licht gerelateerd is aan de magnetisatie van het sample. Voor
dunne samples, wanneer de magnetisatie in het vlak ligt, verandert de polariteit
van de gemitteerde THz-straling wanneer het sample 180 graden geroteerd wordt.
Voor toenemende diktes ontwikkelt zich er een magnetisatiecomponent die uit het
vlak wijst en groeit naarmate de dikte van de kobaltlaag toeneemt. Voor dikke
kobaltfilms, dikker dan 250 nm, wijst de magnetisatie van het sample hoofdzakelijk uit het vlak. Daarom verandert de polariteit van de gemitteerde THz-straling
niet wanneer dikke kobaltlagen geroteerd worden. Metingen met magnetic force
microscopy ondersteunen onze hypothese dat veranderingen in de magnetisatie
waarschijnlijk verantwoordelijk zijn voor de THz-emissie. Deze resultaten zijn belangrijk omdat het bekend is dat niet-magnetische metalen ook in staat zijn tot
emissie van THz-pulsen wanneer ze belicht worden door femtosecondelaserpulsen.
Laten zien dat er een sterke correlatie is tussen magnetisatie en THz-emissie is
daarom essentieel om andere mechanismen die mogelijk verantwoordelijk zijn voor
emissie van THz-licht uit te sluiten. In het laatste deel van dit proefschrift, in
hoofdstuk 5, wordt THz-spectroscopie gebruikt om BiVO4 /Au-grensvlakken te
bestuderen. BiVO4 is een halfgeleidermateriaal dat wijdverbreid gebruikt wordt
in de pigmentindustrie en potentile toepassingen heeft op het gebied van watersplitsing met zonlicht. We observeren emissie van THz-pulsen van een BiVO4 /Audunnelaaggrensvlak wanneer het belicht wordt door femtosecondelaserpulsen. We
stellen voor dat het generatiemechanisme van de THz-emissie het longitudinale fotodembereffect is. Wanneer laserlicht invalt op de dunne laag BiVO4 , opgedampt
op goud, ontstaat er een staandegolfpatroon door de interferentie van licht gereflecteerd van de lucht/BiVO4 - en de BiVO4 /Au-grensvlakken. Als gevolg hiervan
kunnen we een hogere intensiteit (antiknoop) bij het lucht/BiVO4-grensvlak en een
lagere intensiteit (knoop) bij het BiVO4/Au-grensvlak hebben. Hierdoor hebben
we een hogere absorptie nabij het lucht/BiVO4-grensvlak en als gevolg meer ladingsdragers dan nabij het BiVO4/Au-grensvlak. Op deze manier krijgen we een
Bibliography
99
concentratiegradint die, gecombineerd met een verschil in de mobiliteit van elektronen en gaten, aanleiding geeft tot transinte dipolen die THz-licht uitzenden. Kortom, dit proefschrift benadrukt dat terahertztijddomeinspectroscopie (THz-TDS)
een krachtige techniek is met vele toepassingen in wetenschap en technologie. Het
bespreekt de mogelijkheden van het gebruik van THz-TDS voor de beeldvorming
van het magnetische nabije veld van metamateriaalelementen, voor het bestuderen
van de generatie van THz-straling aan ferromagnetische dunne lagen en voor het
onderzoeken van metaal-halfgeleidergrensvlakken.
Nishant Kumar, May 2015
Acknowledgements
Many people have helped me during my PhD research. First of all, I would like
to thank my advisor Professor dr. Paul Planken for his guidance, support and
encouragement. Without his guidance and persistent help this thesis would not
have been possible. Paul gave me freedom to do whatever I wanted to do but at the
same time he has also been guiding me in the right direction throughout my PhD.
He has always been very patient, encouraging and enthusiastic while supervising
me. His door has always been open for questions and problems. I would also
like to thank him for investing effort and time with reading and correcting the
manuscripts and thesis.
I am also thankful to my co-promotor Dr. Aurèle J. L. Adam. I have received
continuous support and guidance from him during my PhD. I knew that I could
always ask him for advice and opinions on lab related issues. He was always ready
with brilliant ideas, honest advice and encouraging words whenever I needed them.
My special thanks to him for performing simulations presented in my thesis.
A special thanks to the group leader, Prof. dr. Paul Urbach, for his warm and
cheerful support.
I would also like to thank my committee members, Prof. dr. H. J. Bakker, Prof.
dr. L. D. A. Siebbeles, Prof. dr. Ir. L. J. van Vliet, Dr. W. A. Smith for their
time and valuable feedback on a preliminary version of this thesis.
There are two people I need to mention especially, Gopakumar Ramakrishnan
and Gopika Ramanandan. I am thankful to Gopakumar, who invested a lot of
time in training me nanofabrication skills. I enjoyed doing experiments with him
which sometimes ran till late evening. I would like to thank Gopika Ramanandan
for her help in the lab and the cleanroom. I thank her for the scientific advice
that she gave, for suggestions and many insightful discussions. These two friends
formed the core of my research time in the Optics research group. Thank you for
great discussions and constant friendly support.
I could not have completed all the required paperwork and delivered it to the
correct places without Yvonne van Aalst. Thank you for your smiling face and
your warm and friendly heart. I would also like to thank Roland Horsten, Thim
Zuidwijk and Rob Pols for providing necessary and timely technical support. You
guys were always ready to help with a smile.
I would like to express my sincere thanks to my collaborators Fatwa F. Abdi,
Bibliography
102
Bartek Trzesniewski and Wilson A. Smith, from the Department of Chemical
Engineering (TU Delft) for preparing BiVO4 samples. I would also like to thank
Ruud Hendrikx at the Department of Materials Science and Engineering (TU
Delft) for the X-ray analysis of the samples.
I would like to acknowledge the help provided by the members of VSL nanofacility with the nanofabrication. Marco van der Krogt, Marc Zuiddam, Charles de
Boer, Roel Mattern, Hozanna Miro, Anja van Langen-Suurling, Arnold van Run
and Ewan Hendriks, thanks a ton to all of you.
I have been very lucky to be part of an extremely friendly and cheerful group.
Prof. Joseph Braat, Nandini Bhattacharya, Silvania Pereira, Florian Bociort, Peter Somers, Omar El Gawhary, Jeffrey Meisner, Man Xu, Adonis Reyes Reyes,
Andreas Hansel, Nitish Kumar, Luca Cisotto, Wouter Westerveld, Lei Wei, Katsirina Ushukova, Mahsa Nemati, Sarathi Roy, Matthias Strauch, Edgar Rojas
Gonzales, Zhe Hou, Hamed Ahmadpanahi, Gerward Weppelman, Young Mi Park,
Priya Dwivedi, Yifong Shao, Fellipe Grillo Paternella, Alberto da costa Assafrao,
Alessandro Polo, Mounir Zeitouny, Thomas Liebig, Olaf Janssen, Axel Wiegmann,
Sven van Haver, Pascal van Grol, Marco Mout, Liesbeth Dingemans, Gyllion
Loozen, Olav Grouwstra, Hui-Shan Chan, Wioletta Moskaluk, Rik Starmans, I
would like to thank all the members of Optica for creating a really good atmosphere in the group.
A special thanks to my officemates, Andreas Hansel, Ying Tang and Daniel
Nascimento Duplat, for the necessary distraction during work. I would like to
thank Andreas and Ying once more for accepting to be my paranymphs.
I am also thankful to my current officemates at ARCNL, Nick Spook, Niklas
Ottosan and Vanessa Verrina for their support. I would like to thank Nick specially
for kindly translating the summary and propositions into Dutch.
I am thankful to my current housemates, Afonso Henriques Graca, Karolina
Jinova and Maounis Konstantinos for providing a friendly atmosphere at home.
I am indebted to all my friends back home who have supported me over the
last few years: Abhishek, Sunil, Atish Srivastawa, Abhishek Kumar, Rose Mary,
Radhakant Singh, Alok Bharti and Kush Tiwari. You guys are amazing, I love
you all.
I had a really nice time at Delft and it was made enjoyable in large part due
to many friends that became a part of my life. I would like to thank my friends
Dhariyash Rathod, Gaurav Panchanan, Menal Lunawat, Vishwas Jain, Venkatraman Krishnaswami, Akshay Sharma, Ankit Verma, Anirban Saha, Nijesh James
and Tittu Varghese Mathew, for all the great times that we have shared. I would
like to specially thank Indunil Ruhunuhewa for her constant support and faith in
me.
Nitish and Gopika have been my family away from home. I cannot thank them
enough for their help and support till now from the moment I arrived in Netherlands. Having homemade Indian food and celebrating Indian festivals with them
was always a pleasure and luxury which I cherished.
Bibliography
103
Finally, a special thanks to my loving and caring family. I thank my brother
and sister for their unconditional love and affection. I am grateful to my parents
for endless love and support. They are the most important people in my world
and I dedicate this thesis to them.
Nishant Kumar, May 2015
Biography
Introduction
Nishant Kumar was born on September 23, 1985 in Patna, Bihar, India. In 2010,
he completed his five-year integrated Master’s degrees in Photonics from Cochin
University of Science and Technology, India, with distinction. His master thesis
was setting up a near-field scanning optical microscope at the Tata institute of
Fundamental Research, Mumbai, India. In October 2010, he joined the Optics
Research Group, Delft University of Technology, in The Netherlands, as a PhD
candidate. His work has been presented at several international conferences and
published in refereed journals.
Publications
Journals
“Investigating terahertz emission from BiVO4 /gold thin film interface,”
N. Kumar, F. F. Abdi, B. Trzesniewski, W. Smith, P. C. M. Planken and A. J. L.
Adam,
(Submitted ).
“Thickness dependent terahertz emission from cobalt thin films,”
N. Kumar, R. Hendrikx, A. J. L. Adam and P. C. M. Planken,
(Accepted for publication).
“Emission of terahertz pulses from nanostructured metal surfaces,”
G. K. P. Ramanandan, G. Ramakrishnan, N. Kumar, A. J. L. Adam, and P. C.
M. Planken,
Journal of Physics D: Applied Physics, 47, 374003 (2014).
“Plasmon-enhanced terahertz emission from Schottky interfaces,”
G. Ramakrishnan, N. Kumar, G. K. P. Ramanandan, R. Hendrikx, A. J. L. Adam,
and P. C. M. Planken,
Applied Physics Letters, 104, 071104 (2014).
Bibliography
106
“Enhanced terahertz emission from semiconductor by coherent optical absorption
in ultrathin semiconductor films,”
G. Ramakrishnan, G. K. P. Ramanandan, A. J. L. Adam, M. Xu, N. Kumar, R.
Hendrikx, and P. C. M. Planken,
Optics Express 20, 11277-11287 (2012).
“THz near-eld Faraday imaging in hybrid metamaterials,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Optics Express 20, 11277-11287 (2012).
“Surface plasmon-enhanced terahertz emission from a hemicyanine self-assembled
monolayer,”
G. Ramakrishnan, N. Kumar, P. C. M. Planken, D. Tanaka, and K. Kajikawa,
Opt. Express 20, 4067–4073 (2012).
“Terahertz emission from surface-immobilized gold nanospheres,”
K. Kajikawa, Y. Nagai, Y. Uchiho, G. Ramakrishnan, N. Kumar, G. K. P. Ramanandan, and P. C. M. Planken,
Opt. Lett. 37, 4053–4055 (2012).
2014 J. Phys. D: Appl. Phys. 47 374003
Conference contributions
“Terahertz generation From monoclinic BiVO4 /Au thin film Interfaces,”
N. Kumar, F. F. Abdi, W. Smith, P. C. M. Planken, A. J. L. Adam,
Poster presentation, The 38th International Conference on Infrared, Millimeter
and Terahertz Waves IRMMW-THz 2013, Mainz, Germany, Sep 1-6, 2013.
“Plasmon enhanced terahertz emission from a Schottky interface,”
G. Ramakrishnan, N. Kumar, G. K. P. Ramanandan, A. J. L. Adam, and P. C.
M. Planken,
Poster presentation, International workshop on optical terahertz science and technology (OTST), Kyoto Terrsa, 2013.
“THz near-field Faraday imaging in hybrid metamaterials,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Oral presentation, 3rd EOS Topical Meeting on Terahertz Science & Technology
(TST 2012), Prague, Czech Republic, 17-20 June, 2012.
“Surface-plasmon enhanced terahertz emission,”
G. Ramakrishnan, N. Kumar, P. C. M. Planken, D. Tanaka, and K. Kajikawa,
Poster presentation, 3rd EOS Topical Meeting on Terahertz Science & Technology
(TST 2012), Prague, Czech Republic, 17-20 June, 2012.
“Direct measurement of the THz near-magnetic field of metamaterial elements,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Bibliography
107
Oral presentation, The 36th International conference on Infrared, Millimeter and
Terahertz waves (IRMMW-THz 2011), Houston, Texas, USA, October 2-7, 2011.
“Direct measurement of the THz near-magnetic field of metamaterial elements,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Oral presentation, Physics@FOM, Veldhoven, Netherlands, Jan 22-23, 2012
“THz near-field Faraday imaging in hybrid metamaterials,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Oral presentation, Workshop on novel trends in optics and magnetism of nanostructures, Augustow, Poland, July 2-7, 2011.
“Direct measurement of the THz near-magnetic field of metamaterial elements,”
N. Kumar, Andrew C. Strikwerda, Kebin Fan, Xin Zhang, Richard D. Averitt, P.
C. M. Planken, and A. J. L. Adam,
Oral presentation, Magnetics and Optics Research International Symposium for
New Storage Technology, (MORIS), Nijmegen, Netherlands, June 21 - 24 2011.