Download Study on the Sensing Coating of the Optical Fibre CO2 Sensor

Article
Study on the Sensing Coating of the Optical Fibre
CO2 Sensor
Karol Wysokiński 1, *, Marek Napierała 1,2 , Tomasz Stańczyk 3,† , Stanisław Lipiński 1,† and
Tomasz Nasiłowski 1
Received: 31 August 2015; Accepted: 11 December 2015; Published: 17 December 2015
Academic Editor: Michael Tiemann
1
2
3
*
†
InPhoTech, 17 Słomińskiego St31, 00-195 Warszawa, Poland; [email protected] (M.N.);
[email protected] (S.L.); [email protected] (T.N.)
IPT Safety, Ceramiczna St 8A, 20-150 Lublin, Poland
Polish Centre for Photonics and Fibre Optics, Rogoźnica 312, 36-060 Głogów Małopolski, Poland;
[email protected]
Correspondence: [email protected]; Tel.: +48-533-779-177; Fax: +48-22-304-6450
These authors contributed equally to this work.
Abstract: Optical fibre carbon dioxide (CO2 ) sensors are reported in this article. The principle of
operation of the sensors relies on the absorption of light transmitted through the fibre by a silica
gel coating containing active dyes, including methyl red, thymol blue and phenol red. Stability
of the sensor has been investigated for the first time for an absorption based CO2 optical fiber
sensor. Influence of the silica gel coating thickness on the sensitivity and response time has also been
studied. The impact of temperature and humidity on the sensor performance has been examined
too. Response times of reported sensors are very short and reach 2–3 s, whereas the sensitivity of the
sensor ranges from 3 to 10 for different coating thicknesses. Reported parameters make the sensor
suitable for indoor and industrial use.
Keywords: optical fiber sensors; CO2 sensors; gas sensors; chemical sensors; sol-gel coatings; silica
gels; indicator dyes; absorption-based sensors
1. Introduction
1.1. An Overview of Current Carbon Dioxide Sensors Applications
Carbon dioxide is an important gas in ventilation, agriculture and many industrial branches.
The possibility of determining the concentration of CO2 enables one to control various processes and
improve safety standards. The increase of CO2 level is a major concern in a majority of buildings
where it may affect the living comfort and in extreme scenarios, can even cause health issues.
Agriculture and greenhouses are other fields where an appropriate level of CO2 content in the
atmosphere is crucial. Miscellaneous industries also require the monitoring of CO2 concentration
with fermentation plants and food storehouses being the most well-known ones. The control of
carbon dioxide level is also important in underground mining, especially in coal mines, in which
an increase of CO2 concentration may be an indicator of fire or ventilation malfunction.
In the latter application safety is of major importance. Even a small damage of an electronic
device may end in a formation of spark which can lead to explosion. Therefore, all the devices
brought underground need to be validated as anti-sparking ones. Optical remote sensing seems to
be an attractive way of achieving this goal, since no electric current is required, and thus they are
intrinsically explosion safe. Hence, the optical fibres are a promising candidate for a gas monitoring
in harsh environments.
Sensors 2015, 15, 31888–31903; doi:10.3390/s151229890
www.mdpi.com/journal/sensors
Sensors 2015, 15, 31888–31903
1.2. Commercially Available CO2 Sensors
Currently, there are several techniques available on the market, which are used for CO2
detection [1–3]. The most widespread carbon dioxide sensors are based on non-dispersive infrared
(NDIR) detection and do not utilize optical fibres. It is noteworthy that while the measurement
itself is very quick, the diffusion of gas through the protecting membrane is much slower. The
installation of membrane is omitted in sensors dedicated to applications, where dust and other
contaminants are not expected to be present. Capnographs do not utilize such membranes and
thus provide a very short response time. Another type of commercially available CO2 sensors are
semiconductor devices, which exhibit different resistance, capacitance or impedance at different
carbon dioxide concentrations [2–4]. They are very cheap, however, they may suffer from
cross-sensitivity to other gases and base signal shifting. Optical fibre fluorescent sensors are another
emerging technology already present on the market, which will be discussed more thoroughly in the
subsequent subsection.
1.3. Optical Fibre Gas Sensors
There are various ways of utilizing the optical fibres in various systems for chemical monitoring.
They can be used for sample illumination and signal reception or they can be sensors per se by using
special active fibre coatings, which are sensitive to the analysed substance.
The most straightforward method of chemical analysis is a measurement of the absorption. It is
especially useful for measuring small concentrations of substances e.g., CO [5,6] or CH4 [7]. Optical
fibres can be used to deliver light to the measurement chamber and subsequently to the detectors.
High sensitivity of the method can be further improved by the use of dielectric mirrors in cavity
enhanced spectroscopy [8,9] The absorption of chemicals can be measured by using hollow-core
fibres, where the light is guided in the air inside the fibre [10].
The measurement of the substance concentration can also be performed by investigating the
evanescent wave absorption, e.g., by microstructured optical fibres [10], D-shaped optical fibres [11]
or Plastic Clad Silica (PCS) fibres. Optical fibres may also be deformed e.g., by tapering to achieve
this goal [12]. Such tapers can be made by using either drawing at high temperature [13] or etching
in hydrogen fluoride solution [14].
Refractive index may also be used for an analysis, since it may change in some polymers and
silica gels when in contact with a certain chemical. This fact is widely used in humidity sensing by
employing Fabry-Perot nanocavities at the tips of optical fibres [15,16]. Refractive index change can
also be employed in the coated fibre tapers [17]. The chemical may be adsorbed on the surface of
optical fibre, which facilitates the detection thereof [18]. There is also a class of compounds, which
change their volume when exposed to an analyte e.g., water [19,20].
Fluorescence found many applications in sensing with optical fibres [21]. Usually, a PCS fibre
or optical fibre taper or optical fibre tip is coated with a matrix material with an active substance
incorporated in it. The possible matrix materials are usually chosen from polymer and sol-gel groups.
Measurement of the intensity of the emission peak can be used for determination of concentration of
the analysed substance [22,23]. However, such a measurement procedure requires the detection of
a narrow band of light, which can be achieved by using optical spectrum analysers, but it is also
very expensive. Alternatively, an optical filter can be used for detection, which, however can be
costly, too. Apart from the intensity-based fluorescent sensing, decay time may also be measured in
order to determine the analyte concentration, which is often called a lifetime-based method [22–24].
Lifetime-based sensors are currently commercially available [23]. Both luminescent methods are
widely used for sensing of many parameters such as pH and the concentration of CO2 or O2 [25].
In a similar way, colour changing dyes can also be used for sensing purposes [23,25,26]. PCS
fibres, fibre tapers, fibre tips or other structures can be coated with a polymer or silica gel containing
an organic indicator dye [27]. The analyte reacts with the indicator inside the sensing layer, which
31889
Sensors 2015, 15, 31888–31903
changes the colour of the latter. Each form of the dye has got different absorption spectrum. Sensors
based on such solutions have much simpler construction than fluorescent-based sensors.
Another method of chemical sensing is based on the inscription of a long period fibre grating
(LPFG) in the optical fibre. LPFGs are sensitive to many parameters like temperature, strain, refractive
index of surrounding media and other factors [28,29]. Chemical sensors employing LPFGs must be
coated with an analyte-sensitive substance [28,29].
The described methods of chemical detection can be used for monitoring different gases. Not
all of them, however, are suitable for each substance. Humidity levels have been measured by
detecting a change of absorption, refractive index or reflectance in interferometers, LPFGs and other
structures [28–31].
Ammonia sensing has also been thoroughly explored. Sensors based on optical fibres usually
utilize the fact that ammonia is a basic gas and it increases the pH [32,33], but other solutions have
also been reported [17].
Fluorescence has been employed for the detection of oxygen [25,34,35]. Optical fibre O2 sensors
utilize special noble metal complexes, which change their fluorescence yield when exposed to
different oxygen concentrations.
Optical fibre sensors of CO2 utilize the acidity of this gas. Many publications on CO2 sensors are
focused on fluorescence [25,35,36]. Few substances, like pyranine (also known as HPTS) or fluorescein
exhibit dependence between pH and fluorescence yield. Even more solutions have been reported for
optical sensors not utilizing optical fibres, however, such results can be easily reproduced in the fibre
optic field [24,37,38].
2. Proposed CO2 Sensor
The aim of the authors was to develop a low-cost, fast responsive optical fibre CO2 sensor.
Despite the availability of a great variety of reported fluorescent CO2 sensors, the authors recognized
that those solutions are too expensive due to the necessity of filtering the wavelengths. The
absorption-based solution with indicator dye doped coating [27] seems to allow one to conduct gas
sensing in a much simpler way. On the other hand, the reported sensors [27] suffer from long response
times, a low sensitivity to CO2 and a high cross-sensitivity to humidity.
Indicator substances usually operate only if they are dissolved in a matrix material. The
most frequently encountered ones are polymers and silica gels. Manifold polymers can be used
for this purpose [19,24,29,36,37,39–41]. Silica gels are made during a controlled hydrolysis of
alkoxysilanes and their derivatives. The most frequently used substrates for silica gel preparation
are tetraethoxysilane [42,43], triethoxymethylsilane [27], triethoxy-n-octylsilane [44] and other
derivatives. A comparison of both polymers and silica gels is presented in Table 1. Due to a high
porosity, a possibility of adjusting the parameters of the coating and other advantages, the authors
have chosen silica gels for the preparation of gas sensors. Other, inorganic matrix materials like ZnO
have also been reported [45], however, they provide less sensing possibilities.
In the reported solution, a fragment of PCS optical fibre acrylic coating is removed and then
it is recoated with silica gel containing pH sensitive dye. Indicator dyes, which can be used for
this purpose include e.g., methyl red, phenol red, phenolphthalein, thymol blue. These substances
change colour when exposed to environments with different pH levels. For example, methyl red
changes from yellow to red when pH decreases [46], phenol red changes from fuchsia to yellow [47]
and thymol blue changes from blue to yellow [48]. The ranges at which a colour changes occur are
quite narrow: 4.8–6.0 for methyl red, 6.4–8.0 for phenol red and 8.0–9.6 for thymol blue [49]. The silica
gel layer after solvent evaporation becomes porous, which facilitates the interaction between the dyes
and carbon dioxide. Some indicators, like thymol blue, do not work properly in a form of powder
or after annealing in silica gel [27]. The interaction between single molecules is responsible for this
phenomenon. Molecules need to be separated to be acceptors or donors of a proton.
31890
Sensors 2015, 15, 31888–31903
Table 1. Comparison of the most important properties of polymers and silica gels.
Porosity
Transparency
Surface quality
Mechanical behaviour
Cost of single deposition process
Cost of series coating process
Easiness of deposition
Easiness of the solution preparation
Coatings with different thickness
Solubility of organic dyes
Leakage of organic dyes
Polymers
Silica Gels
varied: low to moderate
varied: moderate to high
smooth
elastic
low
low
easy
easy
possible to deposit
limited to those soluble in a
solvent dedicated to the polymer
possible unless copolymerized or
immobilized otherwise
high
varied: low to high
pores
varied: moderately elastic to brittle
low
moderate (periodic gelling process)
easy
moderately difficult
possible to deposit
limited to those soluble
in water and alcohols
possible unless immobilized
Carbon dioxide can react with water according to the following equations:
CO2 ` 2H2 O Ñ HCO-3 ` H3 O+
(1)
+
HCO-3 ` H2 O Ñ CO23 ` H3 O
(2)
The formed hydronium ion may subsequently react with the indicator dye, which finally leads
to a colour change. The presented mechanism requires water to be present at the reaction side.
Therefore, sensors utilizing such principle of operation should not be annealed.
3. Experimental Section
Carbon dioxide sensors were prepared according to the procedure described below. A 12 cm
fragment of a plastic clad silica optical fibre was stripped of its acrylic coating by immersing it for
60 s in dichloromethane and subsequent manual taking off the softened polymer. Afterwards we have
coated the fibre with silica gel doped with indicator dye and then we have left it for curing for 24 h.
Silica gels were prepared from triethoxymethylsilane (TEMS) according to the procedure described
in [27]. All the substrates, solvent and a liquid detergent were placed in a plastic vial. The reaction
mixture was stirred for 6 min until a transparent liquid has been obtained. Then a stripped fragment
of PCS fibre has been immersed in the prepared solution. It is possible to control the thickness of the
silica gel layer by managing the fundamental solution parameters [50]. Nevertheless, since available
thickness values for single dip coating process are low, in this work thicker layers have been obtained
by repeating the immersion process several times. Dyes used for the sensors preparation included
thymol blue, phenol red, methyl red and bromothymol blue. The analysed dyes were chosen due to
their pH change range close to neutral, which corresponds to changes induced by CO2 gas, which
decreases the pH of water from neutral to weakly acidic. After the curing time, the fibre was attached
to a PMMA slide. One end of the fibre was connected to the light source and the other one was put
inside a detector. For temperature tests a Peltier module was placed under the slide to control the
temperature of the sensor. All the other tests were performed at a constant temperature.
The light source used for phenol red and methyl red samples was a 520 nm pigtailed laser
(Thorlabs, Newton, NJ, USA). Thymol blue and bromothymol blue sensors were illuminated with
650 nm laser, which was FIS visual fault locator. The mentioned wavelengths were chosen due to
big differences between acidic and basic spectra of the dyes. Every pH sensitive dye has at least two
forms specific for certain type of environments (e.g., acidic and basic). Each form of the dye has a
different absorption spectrum. For certain substances it is possible to choose the wavelength range,
within which, the difference of absorption is high. This is evident in the analysed dyes [46–49]. That
provides a possibility of working at a single wavelength instead of working with a broader spectrum
of light. One just needs to choose the wavelength at which there is a big difference between the spectra
31891
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
of different forms of a dye. A GL55 series photoresistor (Senba Optoelectronic, Shenzhen, China) was
used of
fordifferent
the detection
light.A GL55 series photoresistor (Senba Optoelectronic, Shenzhen, China) was
forms ofofa dye.
used
for
the
detection
of light.
The composition of an
atmosphere inside the chamber was regulated by dosing pure CO2 from a
The
composition
of
an atmosphere
inside the chamber
dosing
pure
CO2afrom
gas pressure bottle. For reducing
the concentration
of CO2 , was
the regulated
chamber by
was
purged
with
fresh air.
a
gas
pressure
bottle.
For
reducing
the
concentration
of
CO
2, the chamber was purged with a fresh
As a result, the concentration of carbon dioxide decreases slowly, which makes it possible to easily
air. As a result, the concentration of carbon dioxide decreases slowly, which makes it possible to
record the sensor readings at different concentrations. Response time measurements were performed
easily record the sensor readings at different concentrations. Response time measurements were
by a rapid filling the chamber with maximum CO2 gas flow and then a subsequent rapid purging
performed by a rapid filling the chamber with maximum CO2 gas flow and then a subsequent rapid
by compressed
The actual
of carbon dioxide
inside
the inside
chamber
monitored
purging by air.
compressed
air.concentration
The actual concentration
of carbon
dioxide
the was
chamber
was by
two commercial
CO
sensors
based
on
NDIR
and
electrochemical
principles
of
operation.
Humidity
2
monitored by two commercial CO2 sensors based on NDIR and electrochemical principles of
was monitored
by an electronic
meter. Temperature
was
measured
with a was
pyrometer
to ensure
operation. Humidity
was monitored
by an electronic
meter.
Temperature
measured
with a that
pyrometer
to ensureof
that
actualwas
temperature
of the sensor was being reported.
the actual
temperature
thethe
sensor
being reported.
The measurements
were
carriedout
outininaadedicated
dedicated gas
is depicted
in Figure
1. 1.
The measurements
were
carried
gaschamber,
chamber,which
which
is depicted
in Figure
Figure
1. 1. Measurement
for2 sensing.
CO2 sensing.
The redrepresents
fragmenttherepresents
optical
Figure
Measurement setset
up up
usedused
for CO
The red fragment
optical fibrethe
sensor.
fibre sensor.
4. Results and Discussion
4. Results and Discussion
4.1. Analysed Indicator Dyes
4.1. Analysed
Indicator
Dyes acting as CO2 indicators have been examined. Out of the four tested
Several
substances
dyes—thymol blue, phenol red, methyl red and bromothymol blue—except for bromothymol blue
Several substances acting as CO2 indicators have been examined. Out of the four tested
three were responsive to carbon dioxide concentration changes in air. Bromothymol blue has already
dyes—thymol
blue,
phenol
red,inmethyl
bromothymol
blue—except
for reported
bromothymol
been reported
as an
indicator
pH andred
gasand
sensors
[32,42], and the
lack of activity
by the blue
threeauthors
were responsive
to
carbon
dioxide
concentration
changes
in
air.
Bromothymol
blue
has
can be attributed either to the interaction between the dye and a sol gel basic catalyst oralready
to
been the
reported
pH due
and to
gas
sensors [32,42],
and
the
activity
reported
the
changeas
of an
pHindicator
sensitivityin
range
immobilization
of the
dye.
Alllack
the of
other
dyes yielded
COby
2
responsive
Sensors
incorporating
thymol blue
and phenol
redand
haveaalready
reported
in or to
authors
can be sensors.
attributed
either
to the interaction
between
the dye
sol gelbeen
basic
catalyst
the literature,
but their sensitivities
much lower [27].
An dye.
optical
2 sensor
the change
of pH sensitivity
range duewere
to immobilization
of the
Allfibre
the CO
other
dyes utilizing
yielded CO2
methyl
red
has
not
been
reported
yet.
Making
such
a
sensor
was
even
claimed
to
be
impossible
due
responsive sensors. Sensors incorporating thymol blue and phenol red have already been reported
to
a
too
low
pH
colour
change
range
of
this
dye
[27].
However,
due
to
the
decomposition
of
alkylin the literature, but their sensitivities were much lower [27]. An optical fibre CO2 sensor utilizing
substituted ammonia catalyst upon drying, which can take several days, the pH of the sensing layer
methyl red has not been reported yet. Making such a sensor was even claimed to be impossible
decreases. Therefore, a sensor incorporating thymol blue, which initially has a blue active layer, turns
due to a too low pH colour change range of this dye [27]. However, due to the decomposition of
green and subsequently turns yellow. This colour change is also associated with the decrease of the
alkyl-substituted
ammonia
catalyst upon drying, which can take several days, the pH of the sensing
sensitivity to CO
2. Eventually, after 5–10 days when the catalyst decomposes inside the active layer,
layerthe
decreases.
Therefore,
a sensor
incorporating
thymol blue,
which initially
has a bluea active
sensor ceases
its operation
and
is no longer working.
The authors
have encountered
similarlayer,
turnsissue
green
and
subsequently
turns
yellow.
This
change
is also
associated
with
theofdecrease
with
phenol
red. Due to
the lower
range
of colour
pH inducing
colour
change
it retained
little
its
after
few
weeks. The opposite
problem
was observed
for methyl
red. This
dye has
of thesensitivity
sensitivity
to aCO
after 5–10
days when
the catalyst
decomposes
inside
the aactive
2 . Eventually,
range
of colour
change
within theand
weakly
acidic
region.
Therefore,
initially
it was
not responsive
layer,pH
the
sensor
ceases
its operation
is no
longer
working.
The
authors
have
encountered a
to carbon
dioxide,
but after
of therange
catalyst,
the sensing
layerchange
decreased
and then little
similar
issue with
phenol
red.decomposition
Due to the lower
of pH
pHof
inducing
colour
it retained
it
was
able
to
detect
CO
2 within a full range of concentrations. This is why it is possible to prepare
of its sensitivity after a few weeks. The opposite problem was observed for methyl red. This dye has a
CO2 sensor with methyl red dye. This is also the first time, when the stability issue for an absorption
pH range of colour change within the weakly acidic region. Therefore, initially it was not responsive
based carbon dioxide optical fibre sensor is reported and a method of circumventing thereof
to carbon dioxide, but after decomposition of the catalyst, pH of the sensing layer decreased and then
is proposed.
it was able to detect CO2 within a full range of concentrations. This is why it is possible to prepare CO2
sensor with methyl red dye. This is also the first time, when the stability issue for an absorption based
carbon dioxide optical fibre sensor is reported and5 a method of circumventing thereof is proposed.
31892
Sensors 2015, 15,
15, page–page
31888–31903
Sensors
Figure 2. Dependence between the intensity of light transmitted through an optical fibre sensor
Figure 2. Dependence between the intensity of light transmitted through an optical fibre sensor
containing thymol blue as an indicator dye and the CO2 concentration.
containing thymol blue as an indicator dye and the CO2 concentration.
Figure 2 shows the dependence between the CO22 concentration and the intensity of the 650 nm
laser light transmitted by the optical fibre sensor comprising thymol blue. The series of experiments
have been performed 24 h after preparation of the sensor. Response of the sensor is logarithmic as a
function of carbon dioxide concentration. Dependence between the intensity of transmitted light and
concentration can
canbe
bedescribed
describedby
byan
anequation:
equation:I =I 2.48ln[CO
= 2.48ln[CO
+ 7.9,
where
CO22 concentration
2] +
where
[CO[CO
2] represents
the
2 ]7.9,
2 ] represents
2
2
the percentage
of CO
in air.
value
equaltoto0.99.
0.99.Character
Characterofofthe
the dependence
dependence between CO
percentage
of CO
2 in2 air.
R Rvalue
is is
equal
CO22
concentration and the transmitted light intensity is a result of a way, in which colour of the dye
changes. As the pH of the layer gradually changes, the absorption spectrum of the layer also changes
gradually. The ratio of the intensity of light transmitted at 100% CO22 concentration and the intensity
concentration equals
equals 8.3, which exceeds the value of 1.7 previously
of light transmitted at 0.1% CO22 concentration
reported for
for this
thisdye
dye[27].
[27].AnAn
increase
of
the
intensity
of transmitted
light occurs
alonga with
increase of the intensity
of transmitted
light occurs
along with
coloura
colour change
of an active
to bright
yellow,
which
consistentwith
withaa literature
change
of an active
layer layer
fromfrom
darkdark
blueblue
to bright
yellow,
which
is isconsistent
The response
response characteristics
characteristics of
of the
the sensor
sensor have
have been
been deteriorating after
spectroscopic data [48]. The
performing the experiments. After approximately two weeks it did not respond to CO22 any more.
The other examined sensor was based on phenol red dye. In the beginning it worked within a
concentration. The
The response
response characteristics
characteristics tended
tended to
to deteriorate
deteriorate after a few days.
full range of CO22 concentration.
The results obtained two weeks after sensor preparation
preparation are
are depicted
depicted in
in Figure
Figure 3.
3. The sensor was
illuminated with a 520 nm laser light.
The response of the sensor is
is linear
linear within
within aa low
low CO
CO22 concentration
concentration range.
range. The dependence
concentration and
and the
the intensity
intensity of
of light
light transmitted
transmitted by the fibre is equal to:
between CO22 concentration
value is equal to
I = 1.50[CO22] ++ 2.03, where [CO22] represents the percentage of
of carbon
carbon dioxide.
dioxide. R22 value
Sensor detects
detects the
the gas
gas only
only until the concentration reaches 0.65%. The
The ratio of maximum and
0.97. Sensor
However, if only a
minimum signal intensity equals 1.36, which is much less than for thymol blue. However,
is is
equal
to to
3.13.1
forfor
thymol
blue,
which
makes
the
0.65% limit is
is taken
takeninto
intoaccount,
account,than
thanthe
thesignal
signalratio
ratio
equal
thymol
blue,
which
makes
difference
smaller.
During
the operation
of theofsensor,
the active
layer changed
its colour
from red
to
the
difference
smaller.
During
the operation
the sensor,
the active
layer changed
its colour
from
paletoyellow,
whichwhich
standsstands
in accordance
with spectroscopic
data available
in theinliterature
[47]. [47].
The
red
pale yellow,
in accordance
with spectroscopic
data available
the literature
limitlimit
of 0.65%
of COof
is a result
a relatively
narrow narrow
pH range
forrange
whichfor
thewhich
colourthe
of
The
of 0.65%
CO2 concentration
is a of
result
of a relatively
pH
2 concentration
phenolof
red
changes.
As the basic
catalyst
the pH ofthe
anpH
active
layer
decreases,
thus the
colour
phenol
red changes.
As the
basicdecomposes,
catalyst decomposes,
of an
active
layer decreases,
available
CO2 detection
narrows
down.
Taking
into account
the given
analytical
expression
for
thus
the available
CO2 range
detection
range
narrows
down.
Taking into
account
the given
analytical
the responsefor
of the
phenol
red to of
carbon
dioxide
andcarbon
the experimental
uncertainty
(˘0.05 a.u.)
one can
expression
response
phenol
red to
dioxide and
the experimental
uncertainty
expecta.u.)
that one
the lower
limit of
detection
should
be highershould
than 0.03%
of higher
CO2 , which
is suitable
for
(±0.05
can expect
that
the lower
limit not
of detection
not be
than 0.03%
of CO
2,
indoorisuse.
which
suitable for indoor use.
31893
6
Sensors
2015,
15,15,
31888–31903
Sensors
2015,
page–page
Sensors
2015,
15,
page–page
Figure
3. Dependence
between
of light
lighttransmitted
transmitted
through
optical
sensor
Figure
3. Dependence
betweenthe
the intensity
intensity of
through
an an
optical
fibrefibre
sensor
Figure
3.phenol
Dependence
between
the intensity
of light concentration.
transmitted through an optical fibre sensor
containing
redred
as an
indicator
containing
phenol
as an
indicatordye
dyeand
andthe
the CO22 concentration.
containing phenol red as an indicator dye and the CO2 concentration.
The sensor containing methyl red has been operating in a stable fashion for a second week after
The
sensor
containing
methyl red
been operating
in a stable
stablefashion
fashion fora asecond
second
week
after
The
sensor
redhas
has
operating
week
after
preparation.
Itcontaining
was tested methyl
with a laser
lightbeen
source
workinginata520
nm. Figure 4for
shows the response
preparation.
It Itwas
tested
with
at 520
520nm.
nm.Figure
Figure4 4shows
shows
the
response
preparation.
was
tested
withaalaser
laserlight
light source
source working
working at
the
response
characteristics
of the
sensor.
characteristics
of
the
sensor.
characteristics
thesensor
sensor.
Signal ofof
this
decreased when CO2 concentration increased, which is an opposite situation
Signal
of of
this
sensor
decreased
when
CO
increased,
which
anopposite
opposite
situation
this
decreased
whenDependence
CO22 concentration
concentration
which
isisan
situation
in Signal
comparison
tosensor
the previous
sensors.
betweenincreased,
light intensity
and
concentration
is
in
in in
comparison
totothe
sensors.
Dependence
between
light
concentration
this
case logarithmic,
similarly
to the Dependence
thymol
blue between
sensor.
Itlight
can
be intensity
described
by an
equation:
comparison
theprevious
previous
sensors.
intensity
andand
concentration
is in is
2 value is equal to 0.96. The quotient of the maximum and minimum intensity
I = case
−0.63ln[CO
2] + 4.58.similarly
R
in this
this
case
logarithmic,
similarly
thethymol
thymolblue
bluesensor.
sensor.It Itcan
canbebedescribed
describedbybyananequation:
equation:
logarithmic,
totothe
2 value
the signal22]is
to2R10.
When
sensor
exposed
toThe
CO2,quotient
the
layer
changes
its
colour
from
=of
−0.63ln[CO
]++equal
4.58. R
value
is equal
tois0.96.
of the
maximum
and
minimum
intensity
I =I´0.63ln[CO
is
equal
to The
0.96.quotient
of the maximum
andyellow
minimum
to
red,
which
is
consistent
with
the
literature
data
[46].
The
lowest
CO
2
concentration,
which
was
of the signal
equal is
to equal
10. When
sensor
is exposed
2, the layer
changes
its colour
fromits
yellow
intensity
of theissignal
to 10.
When
sensor to
is CO
exposed
to CO
changes
colour
2 , the layer
tested
was
0.08%,
but
concentrations
down
to
0.01%–0.02%
should
also
be
possible
to
measure,
since
to red,
which
is consistent
the literature
[46]. The
lowest
2 concentration,
which was
from
yellow
to red,
which is with
consistent
with thedata
literature
data
[46]. CO
The
lowest CO2 concentration,
it was
reported
for
solutions [27].
tested
was
0.08%,was
butsimilar
concentrations
down to 0.01%–0.02%
also be possible
to measure,
since to
which
was
tested
0.08%,
but concentrations
down to should
0.01%–0.02%
should also
be possible
it
was
reported
for
similar
solutions
[27].
measure, since it was reported for similar solutions [27].
Figure 4. Dependence between the intensity of light transmitted through an optical fibre sensor
containing methyl red as an indicator dye and the CO2 concentration.
Figure
4. 4.Dependence
of light
light transmitted
transmittedthrough
throughananoptical
opticalfibre
fibresensor
sensor
Figure
Dependence between
between the
the intensity
intensity of
4.2.containing
Influence
of
the
Dye
Concentration
containing
methyl
red
as
an
indicator
dye
and
the
CO
concentration.
2
methyl red as an indicator dye and the CO2 concentration.
Since the light goes through the active layer deposited on an optical fibre, the concentration of
the
Dye
Concentration
4.2.4.2.
Influence
of of
the
Dye
Concentration
anInfluence
indicator
dye
is
crucial
to the response characteristics of the sensor. It is reasonable, that the more
dye is in the sensor, the stronger response should be reported. However, other factors may also affect
Since
lightgoes
goesthrough
throughthe
theactive
active layer
layer deposited
deposited on
of of
Since
thethe
light
on an
anoptical
opticalfibre,
fibre,the
theconcentration
concentration
the output signal of a sensor, which is presented in Figure 5. Figure 5 and all the other plots presented
an
indicator
dye
is
crucial
to
the
response
characteristics
of
the
sensor.
It
is
reasonable,
that
the
more
an indicator
dye
is crucial
to the
response
characteristics
the sensor. It is reasonable, that the more
henceforth
show
the results
obtained
for methyl
red-basedofsensors.
is in
the
sensor,
thestronger
strongerresponse
responseshould
should be
be reported.
reported. However,
other
factors may
also
affect
dyedye
is in
the
sensor,
the
However,
other
also
Sensitivity in Figure 5 refers to a ratio of the transmitted signal
intensity
at factors
0.1% of may
CO2 to
theaffect
the
output
signal
of
a
sensor,
which
is
presented
in
Figure
5.
Figure
5
and
all
the
other
plots
presented
the output
of at
a sensor,
is presented
in Figure
Figure 5 and
all the
plotsin
presented
signal signal
intensity
100% ofwhich
CO2. This
term is present
in 5.
subsequent
figures
and other
is defined
the
henceforth show the results obtained for methyl red-based sensors.
henceforth
show the results obtained for methyl red-based sensors.
same way.
Sensitivity in Figure 5 refers to a ratio of the transmitted signal intensity at 0.1% of CO2 to the
Sensitivity in Figure 5 refers to a ratio of the transmitted signal intensity at 0.1% of CO2 to the
signal intensity at 100% of CO2. This term is present
in subsequent figures and is defined in the
7
signal intensity at 100% of CO2 . This term is present
in subsequent figures and is defined in the
same way.
same way.
7
31894
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
Sensors 2015, 15, page–page
One can notice that initially an increase of the methyl red concentration improves the sensor
One
notice that
an
of the
the methyl
methylred
red concentrationimproves
improves
sensor
Onecan
can
thatinitially
initially rises,
an increase
increase
of
thethe
sensor
sensitivity.
As notice
the concentration
the dependence
becomesconcentration
less inclined and eventually
goes
sensitivity.
As
the
concentration
rises,
the
dependence
becomes
less
inclined
and
eventually
goes
sensitivity.
As the concentration
the dependence
lessofinclined
goes
down.
This phenomenon
can be rises,
explained
by a limitedbecomes
solubility
methyl and
red eventually
in silica gel.
The
down.
phenomenon
can be
by aa limited
limited solubility
solubilityofofmethyl
methylred
redininsilica
silica gel.
The
down.This
This
phenomenon
be explained
explained
by
The
response
characteristics
getcan
to the
point in the
dye concentration
domain, when
no furthergel.
response
response
characteristics
get
to
the
point
in
the
dye
concentration
domain,
when
no
further
response
response
to theincrease
point inthe
theamount
dye concentration
when
no further
response
increase
is characteristics
possible. If oneget
would
of the dye domain,
in the active
layer,
the excess
of this
increase
onewould
would
increase
amount
theindye
in thelayer,
activethelayer,
the
increaseisispossible.
possible. IfIfone
increase
the the
amount
of theofdye
the active
excess
of excess
this
substance would form a very fine powder during the drying. This was observed as a red, opaque
of substance
this substance
very
fine powder
during
the This
drying.
was as
observed
as a red,
wouldwould
form aform
very afine
powder
during the
drying.
was This
observed
a red, opaque
appearance of an active layer on optical fibre in contrast to yellow, transparent coatings for low
opaque
appearance
of an active
layer
on optical
in contrast
to yellow,
transparent
coatings
appearance
of an active
layer on
optical
fibre infibre
contrast
to yellow,
transparent
coatings
for lowfor
concentrations. What is more, dye in a powdered form may behave in a different way than dissolved
concentrations.
What
is
more,
dye
in
a
powdered
form
may
behave
in
a
different
way
than
dissolved
low concentrations. What is more, dye in a powdered form may behave in a different way than
powder. As one can see in Figure 5, after reaching the maximum, the dependence not only decreases,
powder.powder.
As one can
Figure
reaching
maximum,
dependence
not only decreases,
dissolved
Assee
oneincan
see 5,
inafter
Figure
5, afterthe
reaching
the the
maximum,
the dependence
not only
but even goes below sensor response value equal to 1. This means that after reaching this point, the
but even but
goeseven
below
sensor
response
value
equal to
1. This
means
thatThis
aftermeans
reaching
this
point,
the
decreases,
goes
below
sensor
response
value
equal
to
1.
that
after
reaching
sensor works the opposite way i.e., during the increase of the CO2 concentration, the intensity of
sensor
works
the
opposite
way
i.e.,
during
the
increase
of
the
CO
2
concentration,
the
intensity
ofthe
this
point, thelight
sensor
worksThe
the most
opposite
way i.e.,
during thefor
increase
of the CO
concentration,
transmitted
increases.
reasonable
explanation
this behaviour
is 2that
an acidic form
transmitted
light
increases.
The
most
reasonable
explanation
for
this
behaviour
is
that
an
acidic
form
intensity
of red
transmitted
light increases.
most reasonable
explanation
for this
behaviour ismay
thatbe
an
of methyl
is more soluble
in silica The
gel matrix
than its basic
form. Such
a phenomenon
of methyl
red
is more
soluble
insoluble
silica gel
matrix
than
its basic
form.
Suchform.
a phenomenon
may be
acidic
form
of
methyl
red
is
more
in
silica
gel
matrix
than
its
basic
Such
a
phenomenon
peculiar
simply reach
reach the
the maximum
maximum response
responseatata acertain
certain
peculiar toto methyl
methyl red,
red, since
since other
other dyes
dyes simply
may
be peculiar
to methyl
red, since
other
dyes simply
reach the
maximum response
at a certain
concentration
[27].
concentration [27].
concentration [27].
Figure
5. 5.
Dependence
and theconcentration
concentration of methyl
red in the
sensing
layer.
Figure
Dependencebetween
betweenthe
thesensitivity
layer.
Figure
5. Dependence
between
the
sensitivity and
and the
the concentrationofofmethyl
methylred
redininthe
thesensing
sensing
layer.
Each
sample
has
the
same
geometrical
parameters
and
the
same
layer
thickness.
Eachsample
samplehas
hasthe
thesame
same geometrical
geometrical parameters
parameters and
Each
and the
thesame
samelayer
layerthickness.
thickness.
4.3.
Influence
of
theSensor
Sensor
4.3.
Influenceofofofthe
theActive
ActiveLayer
LayerThickness
Thickness on
on the
Response
4.3.
Influence
the
Active
Layer
Thickness
on
the Response
Response of
of the
the
Sensor
One
canexpect
expectthat
thatsimilarly
similarly to
to the
the concentration,
concentration, the
layer
should
also
One
can
thethickness
thicknessofof
ofan
anactive
active
layer
should
also
One
can
expect
that
similarly
to
the
concentration,
the
thickness
an
active
layer
should
also
play
an
important
role
in
the
sensor
operation.
It
is
possible
to
check
the
influence
of
thickness
on
the
play
an
important
role
in
the
sensor
operation.
is
possible
to
check
the
influence
of
thickness
on
the
play an important role in the sensor operation. It is possible to check the influence of thickness on the
sensor
signal
by
preparing
samples
with
different
thickness
of
a
sensing
layer.
This
can
be
achieved
sensor
of aa sensing
sensing layer.
layer. This
Thiscan
canbe
beachieved
achieved
sensorsignal
signalby
bypreparing
preparing samples
samples with
with different
different thickness
thickness of
by
depositionof
ofdifferent
differentnumber
numberof
oflayers
layers on
on
an
experiment
are
by
deposition
an
optical
fibre.The
Theresults
resultsofof
ofsuch
such
an
experiment
are
by
deposition
of
different
number
of
layers
on an
an optical
optical fibre.
fibre.
The
results
such
an
experiment
are
depicted
in
Figure6.6.The
Theplot
plotshows
shows that
that the
the response
of
the
sensor
(defined
the
same
way
like
in the
depicted
in
Figure
response
of
the
sensor
(defined
the
same
way
like
in
the
depicted in Figure 6. The plot shows
sensor (defined the same way like in the
previous subsection)rises
rises whenthickness
thickness of aa sensing
sensing layer rises.
previous
previoussubsection)
subsection) riseswhen
when thickness of
of a sensing layer
layer rises.
Figure 6. The influence of a sensing layer thickness on the sensitivity. Measurements have been
Figure
The
Figure
6.6. The
of a sensing
layer thickness on the sensitivity. Measurements
Measurementshave
havebeen
been
performed
on influence
scanning electron
microscope.
performedon
onscanning
scanningelectron
electronmicroscope.
microscope.
performed
8
8
31895
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
4.4. Repeatability of the Sensor
Sensor Response
Response
4.4.
Sensor
aptitude
for work
work should
should be
be tested
tested not
not only
only during
during aa short
short test,
test, but
but also
also during
during aa longer
longer
Sensors
2015,
15, page–page
Sensor
aptitude
for
experiment. Figure
fibre
experiment.
Figure 77 presents
presents a plot of the intensity of light transmitted through the optical fibre
4.4.during
Repeatability
of the Sensor The
Response
sensor
the
experiment.
sensor
was
consecutively
exposed
to
100%
CO
2
and
fresh
air.
This
sensor during the experiment. The sensor was consecutively exposed to 100% CO2
air. This
aptitude
for work
shouldoperation
be tested not
only
during a short test, but also during a longer
results in
inSensor
number
of dips
dips
in aa sensor
sensor
operation
time
plot.
results
aa number
of
in
time
plot.
experiment. Figure 7 presents a plot of the intensity of light transmitted through the optical fibre
sensor during the experiment. The sensor was consecutively exposed to 100% CO2 and fresh air. This
results in a number of dips in a sensor operation time plot.
Figure 7.
7. Intensity of the transmitted light for a sensor
sensor consecutively
consecutively exposed
exposed to
to CO
CO22 and fresh
fresh air
air
Figure
with
CO
2
content
below
0.1%.
with Figure
CO2 7. Intensity of the transmitted light for a sensor consecutively exposed to CO2 and fresh air
with CO2 content below 0.1%.
One can
thethe
sensor
works
in a stable,
repeatable
way. The
sequential
CO2 exposures
One
can observe
observethat
that
sensor
works
in a stable,
repeatable
way.
The sequential
CO2
do
not
affect
the
base
level
of
the
intensity
neither
for
low
nor
for
high
CO
2 2concentrations.
One
can
observe
that
the
sensor
works
in
a
stable,
repeatable
way.
The
sequential
CO
exposures
exposures do not affect the base level of the intensity neither for low nor for high CO2 concentrations.
do not affect
the base
level of7 the
for low nor for
CO2 concentrations.
Fluctuations
present
in Figure
Figure
are intensity
resultneither
of aa non-uniform
non-uniform
gashigh
distribution
during those
those
Fluctuations
present
in
7 are
aa result
of
gas
distribution
during
Fluctuations
present
in
Figure
7
are
a
result
of
a
non-uniform
gas
distribution
during those
fast-paced
tests.
fast-paced tests.
fast-paced tests.
4.5. Response
Response Time
Time of
of the
the Sensor
Sensor
4.5.
4.5. Response Time of the Sensor
The most
most important
important parameter
parameter of the sensor,
sensor, which
which affects
affects the
the response
response time,
time, is
is the
the thickness
thickness of
of
The
The most
important parameterofofthe
the sensor, which
affects the
response time,
is the thickness
of
the sensing
sensing
layer.
TheThe
thicker
the
layer,
the
more
timeisisisneeded
needed
for
CO
2 to
diffuse
through
Figure
the
layer.
The
thicker
the
layer,
the
for
CO
to
diffuse
through
it.it.
Figure
8
the sensing
layer.
thicker
the
layer,
themore
moretime
time
needed
for
CO
2 2to
diffuse
through
it. Figure
8 presents
response
time
plots
for
four
different
layer
thicknesses.
presents
response
time
plots
for
four
different
layer
thicknesses.
8 presents response time plots for four different layer thicknesses.
(a)
(b)
(a)
(b)
(c)
(d)
Figure 8. Response time characteristics for four different sensing layer thicknesses, 0.75 μm, 3 μm,
Figure 8. Response time
0.75 µm, 3 µm,
(c) characteristics for four different sensing layer thicknesses,
(d)
3.5 μm and 9 μm ((a), (b), (c), (d) respectively).
3.5 µm and 9 µm ((a), (b), (c), (d) respectively).
Figure 8. Response time characteristics for four different sensing layer thicknesses, 0.75 μm, 3 μm,
9
3.5 μm and 9 μm ((a), (b), (c), (d) respectively).
31896
9
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
The plots presented in Figure 8 follow the rule mentioned above. What is more, one can notice
Sensors 2015, 15, page–page
that the
time needed for the sensor's response to 100% CO2 is much lower than the time needed to
The plots presented in Figure 8 follow the rule mentioned above. What is more, one can notice
get back
to time
the base
reading
after theresponse
end of to
the
COCO
The
difference
increases
significantly
2 pulse.
that the
needed
for the
100%
2 is much
lower
thanis
the
time one
needed
get
The plots
presented
insensor's
Figure 8 follow the
rule mentioned
above.
What
more,
can to
notice
whenback
the
thickness
increases.
The
response
time
for
the
exposure
to
CO
changes
slightly
2 thesignificantly
to the
base
reading
after
the end
of the CO
pulse.
The
increases
that the
time
needed
for the
sensor's
response
to 2100%
CO
2 is difference
much lower
than
time neededwhen
toover
get the
whole
range
of
thickness.
Figure
9
shows
the
dependence
between
the
response
time
and
the
sensing
the
thickness
increases.
response
forCO
the2 pulse.
exposure
CO2 changes
slightly
over the whole
back
to the base
readingThe
after
the endtime
of the
Thetodifference
increases
significantly
when
of thickness.
FigureThe
9 shows
the dependence
between the
response
timeslightly
and theover
sensing
layer
layerrange
thickness.
the
thickness
increases.
response
time for the exposure
to CO
2 changes
the whole
thickness.
rangelowest
of thickness.
Figure
9 showstimes
the dependence
between
the response
timefrom
and the
layer
The
reported
response
are very short.
When
switching
lowsensing
to high
carbon
The
lowest
reported
response
times
are
very
short.
When
switching
from
low
to
high
carbon
thickness.
dioxide concentration, the response time is equal to 2 s for 0.75 µm thick silica layer. For the same
dioxide
concentration,
the response
response times
time isare
equal
toshort.
2 s for
0.75 μm
thick silica
thecarbon
same
The
lowest
reported
very
switching
from layer.
low toFor
high
thickness,
the
response
time
during switching
from
highWhen
to
low
CO
2 concentration is equal to 3 s.
thickness,
the response the
timeresponse
during switching
from
high
to 0.75
low μm
CO2thick
concentration
is For
equal
tosame
3 s.
dioxide
concentration,
time
is
equal
to
2
s
for
silica
layer.
the
These values are much lower than for other similar sensors reported in the literature [27,30]. Such
These
values
much lower
than forswitching
other similar
the literatureis[27,30].
Such
thickness,
theare
response
time during
fromsensors
high toreported
low CO2in
concentration
equal to
3 s.
shortshort
response times
make
it possible
totouse
such sensor
sensorfor
foron-line
on-line
CO2 monitoring.
Apart
from
the
times
make
it possible
use such
CO2inmonitoring.
Apart
fromSuch
the
Theseresponse
values are
much
lower
than for
other
similar sensors
reported
the literature
[27,30].
low response
times,
thin
layers
exhibit
satisfactory
sensitivity
(see
Figure
6)
equal
to
3.
Therefore
thin
low
response
times,
layers
exhibit satisfactory
sensitivity
(see Figure
equal to 3. Apart
Therefore
short
response
timesthin
make
it possible
to use such sensor
for on-line
CO2 6)
monitoring.
fromthin
the
layers
areresponse
expected
to be
anan
optimum
choice
for
versatile
applications.
layers
are expected
tothin
be
optimum
choice
for versatile
applications.
low
times,
layers
exhibit
satisfactory
sensitivity
(see Figure 6) equal to 3. Therefore thin
layers are expected to be an optimum choice for versatile applications.
(a)
(b)
(b) during exposure to
Figure 9. Dependence(a)
between the sensing layer thickness and (a) response time
Figure 9. Dependence between the sensing layer thickness and (a) response time during exposure to
(b) response between
time during
to airthickness
after COand
2 impulse.
100%
CO
Figure
9.2 Dependence
the exposure
sensing layer
(a) response time during exposure to
100% CO2 (b) response time during exposure to air after CO2 impulse.
100% CO2 (b) response time during exposure to air after CO2 impulse.
4.6. Influence of Temperature
4.6. Influence
of Temperature
4.6. Influence
of Temperature
Temperature is a factor which can substantially change during a measurement. Therefore, its
impact
on the sensor's
should
besubstantially
thoroughly inspected.
The temperature
of the
sensor
was
Temperature
is a isfactor
which
can
change
during
measurement.
Therefore,
Temperature
a response
factor
which
cansubstantially
change during
a ameasurement.
Therefore,
its its
controlled
by
Peltier
module
and
it
has
been
measured
by
a
pyrometer
at
the
centre
of
the
optical
impact
on sensor's
the sensor's
response
shouldbe
bethoroughly
thoroughly inspected.
temperature
of the
was was
impact
on the
response
should
inspected.The
The
temperature
of sensor
the sensor
fibre
sensor.
intensity
ofand
transmitted
through
optical atfibre
increases
when
the fibre
controlled
by The
Peltier
module
it has
been
measured
athe
pyrometer
atthe
thecentre
centreof
ofthe
theoptical
optical
controlled
by Peltier
module
and
itlight
has
been
measured
byby
a pyrometer
temperature
increases
as it is shown
in the
Figure 10.through the optical fibre increases when the
fibre
sensor.
The of
intensity
of light
transmitted
sensor.
The
intensity
light
transmitted
through
the optical fibre increases when the temperature
The dependence
the intensity
of transmitted
light and the temperature is weak below
temperature
increasesbetween
as it is shown
in the Figure
10.
increases as it is shown in the Figure 10.
25 °C.The
Above
this temperature,
significantly
until
55 °C–60 is
°Cweak
range.
The
dependence
between the
the intensity
intensity rises
of transmitted
light
andreaching
the temperature
below
The dependence
between
the intensity
of 10)
transmitted
light
and than
the temperature
is weak
below
intensity
at thethis
highest
point (55the
°Cintensity
in Figure
is 3.85 timesuntil
higher
at5520°C–60
°C. Therefore
itThe
is
25
°C.
Above
temperature,
rises
significantly
reaching
°C
range.
˝
˝
˝
25 C.
Above this temperature,
the intensity
rises significantly
until while
reaching
55 such
C–60optical
Cit is
range.
important
or control
the times
temperature
using
intensity attothemeasure
highest point
(55 °C simultaneously
in Figure 10) is 3.85
higher than
at 20
°C. Therefore
˝ C in Figure 10) is 3.85 times higher than at 20 ˝ C. Therefore it is
The intensity
at
the
highest
point
(55
fibre
sensor.to measure or control simultaneously the temperature while using such optical
important
important
to measure or control simultaneously the temperature while using such optical fibre sensor.
fibre sensor.
Figure 10. Dependence between the light intensity transmitted by an optical fibre sensor containing
methyl
red
and the temperature.
Measurement
performed
at 0.1%
ofanCO
2 in air.
Figure
10. Dependence
between
thelight
lightintensity
intensity
transmitted
byby
optical
fibre
sensor
containing
Figure
10. Dependence
between
the
transmitted
an
optical
fibre
sensor
containing
methyl
red and
the temperature.
Measurementperformed
performed at
2 in air.
methyl
red and
the temperature.
Measurement
at0.1%
0.1%ofofCO
CO
10
2 in air.
10
31897
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
The phenomenon described above can be a result of an increased solubility of the indicator dye.
Another factor which
bebe
anan
increased
affinity
to
which could
couldalso
alsoinfluence
influencethe
theoperation
operationofofthe
thesensor
sensorcould
could
increased
affinity
COCO
the
dye
to
2 of
the
dyeororsilica
silicagel.The
gel.Thealternative
alternativeexplanation
explanationfor
forthe
thesensor’s
sensor’sbehaviour
behaviour could
could be a change
2 of
A decrease
decrease of
of the
the refractive
refractive index would increase the amount of
of the active layer refractive index. A
light in the silica core.
A series of thermal experiments has been performed in order to determine the temperature at
C. Then,
which the drop in transmission occurs. It has been noted that the intensity increases until 55 ˝°C.
˝C
˝C
˝
drops abruptly
abruptly and
and decreases by
between 55 °C and 60 °C it does not rise any more.
more. At 60 ˘
± 22 °CC itit drops
Further increase
increase of
of the
the temperature makes the transmission even weaker (see Figure 10).
58%. Further
10). The
decrease of the intensity is linear in time, which is presented
in
Figure
11.
presented in Figure 11.
Figure
Figure 11.
11. Dependence
Dependence between
between the
the intensity
intensity of
of light
light transmitted
transmitted through
through the
the sensor
sensor and
and the
the time
time
˝
after
reaching
60
°C.
Measurement
performed
at
0.1%
of
CO
2
in
air.
after reaching 60 C. Measurement performed at 0.1% of CO2 in air.
It is important to highlight that before reaching the temperature of 60 ˝°C, the sensor was
It is important to highlight that before reaching the temperature of 60 C, the sensor was
operational. After reaching 60 °C
it did not respond to CO2 any more. The time needed for the whole
operational. After reaching 60 ˝ C it did not respond to CO2 any more. The time needed for the whole
change is approximately equal to 100 s at 0.1% of CO2. The reason for such an abrupt transition is
change is approximately equal to 100 s at 0.1% of CO2 . The reason for such an abrupt transition is
probably related to the nature of methyl red indicator dye, since such behaviour has not been
probably related to the nature of methyl red indicator dye, since such behaviour has not been reported
reported for other silica gel-based optical fibre sensors [35]. Therefore, the most probable explanation
for other silica gel-based optical fibre sensors [35]. Therefore, the most probable explanation for the
for the analysed phenomenon is that methyl red in silica gel ˝at 60 °C losses its ability to bind hydrogen
analysed phenomenon is that methyl red in silica gel at 60 C losses its ability to bind hydrogen ion.
ion. This can be caused either by a decrease of humidity in the silica gel and subsequent precipitation
This can be caused either by a decrease of humidity in the silica gel and subsequent precipitation
of the dye or by molecular properties of methyl red. The decay time of 100 s is probably needed to
of the dye or by molecular properties of methyl red. The decay time of 100 s is probably needed to
achieve the threshold temperature for the whole optical fibre sensing region.
achieve the threshold temperature for the whole optical fibre sensing region.
The described process was repeatable. However, the optical fibre regained its ability to sense
The described process was repeatable. However, the optical fibre regained its ability to sense
carbon dioxide after cooling to temperatures lower than 60 ˝°C. For instance, the sensor resumed
carbon dioxide after cooling to temperatures lower than 60 C. For instance, the sensor resumed
operation after cooling to 43 °C,
which occurred approximately 4 min after the start of cooling from
operation after cooling to 43 ˝ C, which occurred approximately 4 min after the start of cooling from
60 °C.
This
speaks
in
favour
of
the
humidity-based hypothesis described in the previous paragraph.
60 ˝ C. This speaks in favour of the humidity-based hypothesis described in the previous paragraph.
It is important to notice that the drop of the transmission is non-differentiable as a function of
It is important to notice that the drop of the transmission is non-differentiable as a function of
temperature. This could indicate that a high order phase transition may occur, since the imaginary
temperature. This could indicate that a high order phase transition may occur, since the imaginary
part of the dielectric constant (responsible for absorption) is non-differentiable.
part of the dielectric constant (responsible for absorption) is non-differentiable.
4.7.
4.7. Cross-Sensitivity
Cross-Sensitivity to
to Other
Other Gases
Gases
Equations
(1)
and
(2)
indicate
employing CO
CO2 inside
the sensing layer require
Equations (1) and (2) indicate that
that the
the processes
processes employing
2 inside the sensing layer require
water
water to
to be
be present
present at
at the
the reaction
reaction site.
site. Thereby,
Thereby, humidity
humidity may
may also
also affect
affect the
the operation
operation of
of the
the sensor.
sensor.
We
expected
that
the
humidity
dependence
should
also
be
relatively
low,
similar
to
the
We expected that the humidity dependence should also be relatively low, similar to the solution
solution
reported
in [30].
[30]. The
shown in
in Figure
Figure 12.
12.
reported in
The results
results are
are shown
When
increasing
the
relative
humidity
from
When increasing the relative humidity from 36%
36% to
to 84%,
84%, the
the intensity
intensity of
of light
light increases
increases only
only by
by
13%.
It
is
consistent
with
the
results
reported
for
methylene
blue
[30].
Measured
range
of
relative
13%. It is consistent with the results reported for methylene blue [30]. Measured range of relative
humidity is typical for majority of applications. It is necessary to highlight the fact, that the response
of the sensor to humidity can also be dependent on temperature.
31898
11
Sensors 2015, 15, 31888–31903
Sensors 2015, 15, page–page
humidity
is typical for majority of applications. It is necessary to highlight the fact, that the response
of the sensor to humidity can also be dependent on temperature.
Although the cross-sensitivity to water vapour is low, the sensor responds strongly to liquid
Although the cross-sensitivity to water vapour is low, the sensor responds strongly to liquid
water and dew. For instance, the transmitted intensity strongly increases when the sensing part of
water and dew. For instance, the transmitted intensity strongly increases when the sensing part of
the fibre is immersed in water. Therefore during the operation the sensor should be protected against
the fibre is immersed in water. Therefore during the operation the sensor should be protected against
water and other liquids.
water and other liquids.
Carbon dioxide is not the only gas which can influence the pH of a sensing layer. There are
Carbon dioxide is not the only gas which can influence the pH of a sensing layer. There are
several other gases, e.g., NO2 and SO2, which also react with water and produce H3O++ ions. Similarly,
several other gases, e.g., NO2 and SO2 , which also react with water and produce H3 O ions. Similarly,
NH3 is a basic gas, which increases pH of the sensing layer. All the mentioned gases may influence
NH3 is a basic gas, which increases pH of the sensing layer. All the mentioned gases may influence
the readings of the reported sensor. However, they are usually present in the atmosphere at low
the readings of the reported sensor. However, they are usually present in the atmosphere at low
levels, which may be too low to affect the sensor operation.
levels, which may be too low to affect the sensor operation.
Figure 12. Dependence between relative humidity and the intensity of transmitted light. The
Figure 12. Dependence between relative humidity and the intensity of transmitted light. The
measurements were carried out at 18 ˝ C with a sensor containing methyl red.
measurements were carried out at 18 °C with a sensor containing methyl red.
It is possible to design and prepare more selective sensors for solely detecting gases more acidic
than carbon dioxide or more basic than this gas. For instance, a sensor similar to the reported ones,
but also containing a buffer, which would keep a pH approximately around 4 would be less affected
by typical fluctuations of CO22 and it would mostly respond to more
more acidic
acidic gases.
gases. However, a dye
which changes its colour within lower pH values should be used for this purpose instead of methyl
red or phenol red.
red. In a similar way ammonia can also be detected. Other solutions like the use of
certain polymers and not using a matrix material at all [32] have already been reported and also may
help solving the problem of cross-sensitivity.
cross-sensitivity.
Thus aa sensor
sensormatrix
matrixincorporating:
incorporating:(a) (a)
reported
(b) sensor
of strongly
the the
reported
CO2CO
sensor;
(b) sensor
of strongly
acidic
2 sensor;
acidic
gases;
(c)
sensor
of
basic
gases;
(d)
humidity
sensor
and
(e)
temperature
sensor
should
gases; (c) sensor of basic gases; (d) humidity sensor and (e) temperature sensor should be
self-consistent and should provide the option
option of
of multigas
multigas sensing.
sensing.
4.8.
Final Recommended
Recommended Solution
Solution
4.8. Final
The
authors have
have tested
tested several
several sensors
sensors with
with different
different indicator
indicator dyes,
dyes, different
different response
response times
times
The authors
and
Thymol blue,
blue, due
due to
to its
its instability,
instability, was
useful for
further
and different
different sensitivities.
sensitivities. Thymol
was found
found not
not to
to be
be useful
for further
work.
Phenol
red
yielded
a
sensor
with
0.65%
maximum
limit
of
detection,
which
is
suitable
work. Phenol red yielded a sensor with 0.65% maximum limit of detection, which is suitable for
for
ventilation
Methyl red
red provided
provided aa sensor
sensor which
which could
could operate
ventilation monitoring.
monitoring. Methyl
operate at
at low
low and
and high
high CO
CO22,,
concentrations,
Taking into
account
concentrations, and
and therefore
therefore it
it is
is recommended
recommended dye
dye for
for versatile
versatile applications.
applications. Taking
into account
the
examined
silica
gel
thickness,
thin
layers
(approximately
0.75
µm)
are
suitable
for
the
majority
the examined silica gel thickness, thin layers (approximately 0.75 μm) are suitable for the majority of
of
applications,
sensitivity reaches
and response
response times
times are
are very
(2–3 s).
s). To
To the
applications, since
since the
the sensitivity
reaches aa value
value of
of 33 and
very low
low (2–3
the
best
our knowledge,
knowledge, these
these response
response times
for thin
thin silica
silica gel
gel layers
layers are
are the
the lowest
lowest ones
ones reported
reported for
for
best of
of our
times for
an
absorption-based
optical
fibre
CO
gas
sensor.
Similarly,
the
reported
sensitivity
for
a
thick
silica
an absorption-based optical fibre CO22 gas sensor. Similarly, the reported sensitivity for a thick silica
gel
gel coating
coating is
is the
the highest
highest for
for such
such aa sensor
sensor type.
type.
When compared to commercially available NDIR sensors, the reported sensor has much faster
response time (2–3 s instead of 60 s [1]) and can be used for measuring higher CO2 concentrations (up
31899hand, the resolution of NDIR devices within the
to 100% instead of 0.5%–1% for NDIR). On the other
low concentration range is much higher. The reported sensor has a faster response than commercially
12
Sensors 2015, 15, 31888–31903
When compared to commercially available NDIR sensors, the reported sensor has much faster
response time (2–3 s instead of 60 s [1]) and can be used for measuring higher CO2 concentrations
(up
to 2015,
100%15,instead
of 0.5%–1% for NDIR). On the other hand, the resolution of NDIR devices
Sensors
page–page
within the low concentration range is much higher. The reported sensor has a faster response than
available electrochemical
sensors and it sensors
has similar
(measurement
from 0.035%
to
commercially
available electrochemical
andresolution
it has similar
resolution range
(measurement
range
1%
or
5%
or
from
0.2%
to
95%,
response
time
of
several
minutes
[2]).
from 0.035% to 1% or 5% or from 0.2% to 95%, response time of several minutes [2]).
4.9. Self-Referencing Arrangement
The described sensor can
We propose a simple
can possibly
possibly find
find application
application in
in manifold
manifold areas.
areas. We
self-referencing
of the
thesensing
sensingsystem,
system,ininwhich
whichlight
light
from
source
passes
through
self-referencing arrangement of
from
thethe
source
passes
through
an
an
optical
fibre
coupler
then
it passes
independently
through
the sensor
the reference
optical
fibre
coupler
andand
then
it passes
independently
through
the sensor
and and
the reference
arm.arm.
The
The
arrangement
is shown
in Figure
arrangement
is shown
in Figure
13. 13.
Figure
of the
the sensor.
sensor. Lines
Lines represent
represent optical
optical fibres.
fibres.
Figure 13.
13. Self-referencing
Self-referencing arrangement
arrangement of
In such a sensing system the concentration of CO2 would be determined from the intensity ratio
In such a sensing system the concentration of CO2 would be determined from the intensity ratio
of both signals passing through the sensor and the reference arm. What is more, the described sensor
of both signals passing through the sensor and the reference arm. What is more, the described sensor
arrangement enables one to omit the problem of the light source power instability.
arrangement enables one to omit the problem of the light source power instability.
5. Conclusions
5.
Conclusions
Broad range
range CO
CO2 optical
fibre sensors have been presented. The reported sensors are made by
Broad
2 optical fibre sensors have been presented. The reported sensors are made by
deposition of
of an
an active
active silica
silica gel
gel coating
coating onto
onto aa plastic
plastic clad
clad silica
silica optical
optical fibre.
fibre. Considering
Considering the
the tested
tested
deposition
indicator
dyes,
methyl
red
was
found
to
be
the
most
utilitarian
one,
since
it
operates
over
a
whole
indicator dyes, methyl red was found to be the most utilitarian one, since it operates over a whole
range of
of CO
CO2 concentrations.
It is noteworthy that a methyl red CO2 optical fibre sensor has not been
range
2 concentrations. It is noteworthy that a methyl red CO2 optical fibre sensor has not been
reported before
before and
and it
it was
was even
even claimed
claimed to
to be
be impossible
impossible to
to manufacture
manufacture one.
one. The
The stability
stability of
of these
these
reported
absorption-based
carbon
dioxide
optical
fibre
sensors
has
been
examined
for
the
first
time.
A
method
absorption-based carbon dioxide optical fibre sensors has been examined for the first time. A method
of obtaining
obtaining aastable
stablesensor
sensorhas
hasbeen
beenproposed
proposedand
andexperimentally
experimentallyverified
verified
with
good
results.
What
of
with
good
results.
What
is
is
more,
the
phenol
red-containing
sensor
also
operated
stably
within
the
low
range
of
carbon
dioxide
more, the phenol red-containing sensor also operated stably within the low range of carbon dioxide
concentrations. This
red ones
ones to
to
concentrations.
This may
may destine
destine phenol
phenol red-based
red-based sensors
sensors to
to indoor
indoor use
use and
and methyl
methyl red
industrial applications.
applications. It
layer thickness
thickness increases
increases the
the
industrial
It has
has been
been noted
noted that
that an
an increase
increase in
in the
the active
active layer
sensitivity
of
the
sensor,
but
it
also
substantially
increases
the
response
time.
Humidity
was
found
to
sensitivity of the sensor, but it also substantially increases the response time. Humidity was found
have
a
weak
influence
on
the
response
of
the
sensor.
Reported
sensors
exhibit
high
sensitivities
to have a weak influence on the response of the sensor. Reported sensors exhibit high sensitivities
reaching II0.1%/I/I
100% of 10 for the 9 μm thick layers. For an active layer thickness of 0.75 μm the response
reaching
0.1% 100% of 10 for the 9 µm thick layers. For an active layer thickness of 0.75 µm the
time
can
be
as
lowbe
asas
2 slow
for as
switching
from lowfrom
to high
concentration
and 3 s for
the3 opposite
response time can
2 s for switching
lowCO
to 2high
CO2 concentration
and
s for the
process.
The
reported
sensor
thus
exhibits
much
lower
response
time
and
higher
sensitivity
than
opposite process. The reported sensor thus exhibits much lower response time and higher sensitivity
other
absorption-based
solutions
presented
elsewhere
in
the
literature.
than other absorption-based solutions presented elsewhere in the literature.
Author Contributions: Tomasz Nasiłowski conceived of the project and Marek Napierała coordinated it. Karol
Author
Contributions:
Tomasz Nasiłowski
conceived
of the project
andStańczyk
Marek Napierała
coordinated
Wysokiński
designed the experiments
and performed
measurements.
Tomasz
and Stanisław
Lipiński
it. Karol Wysokiński designed the experiments and performed measurements. Tomasz Stańczyk and
designed
and
prepared
the
measurement
chamber
and
prepared
the
light
sources.
The
manuscript
was
written
Stanisław Lipiński designed and prepared the measurement chamber and prepared the light sources.
The
by Karol Wysokiński
and
revised
byński
Marek
andby
Tomasz
manuscript
was written
bywas
Karol
Wysoki
andNapierała
was revised
Marek Nasiłowski.
Napierała and Tomasz Nasiłowski.
Conflicts
Conflicts of
of Interest:
Interest: The
The authors
authors declare
declare no
no conflict
conflictof
ofinterest.
interest.
References
1.
2.
Yasuda, T.; Yonemura, S.; Tani, A. Comparison
of the Characteristics of Small Commercial NDIR CO2
31900
Sensor Models and Development of a Portable CO2 Measurement Device. Sensors 2012, 12, 3641–3655.
García Mandayo, G.; Herrán, J.; Castro-Hurtado, I.; Castaño, E. Performance of a CO2 Impedimetric Sensor
Prototype for Air Quality Monitoring. Sensors 2011, 11, 5047–5057.
Sensors 2015, 15, 31888–31903
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Yasuda, T.; Yonemura, S.; Tani, A. Comparison of the Characteristics of Small Commercial NDIR CO2
Sensor Models and Development of a Portable CO2 Measurement Device. Sensors 2012, 12, 3641–3655.
[CrossRef] [PubMed]
García Mandayo, G.; Herrán, J.; Castro-Hurtado, I.; Castaño, E. Performance of a CO2 Impedimetric Sensor
Prototype for Air Quality Monitoring. Sensors 2011, 11, 5047–5057. [CrossRef] [PubMed]
Fine, G.F.; Cavanagh, L.M.; Afonja, A.; Binions, R. Metal Oxide Semi-Conductor Gas Sensors in
Environmental Monitoring. Sensors 2010, 10, 5469–5502. [CrossRef] [PubMed]
Tiemann, M. Porous Metal Oxides as Gas Sensors. Chem. Eur. J. 2007, 13, 8376–8388. [CrossRef] [PubMed]
Zhu, S.; Chen, Y.; Zhang, G.; Sa, J. An Optical Fiber Sensor based on Absorption Spectroscopy for Carbon
Monoxide Detection. In Proceedings of the 2010 International Conference On Computer Design and
Appliations, Qinhuangdao, China, 25–27 June 2010.
Mulrooney, J.; Clifford, J.; Fitzpatrick, C.; Chambers, P.; Lewis, E. A mid-infrared optical fibre sensor for the
detection of carbon monoxide exhaust emissions. Sens. Actuators A Phys. 2008, 144, 13–17. [CrossRef]
Liu, T.; Wang, C.; Wei, Y.; Zhao, Y.; Huo, D.; Shang, Y.; Wang, Z.; Ning, Y. Fibre optic sensors for mine
hazard detection. J. Phys: Conf. Ser. 2009, 178. [CrossRef]
Nakaema, W.M.; Hao, Z.Q.; Rohwetter, P.; Wöste, L.; Stelmaszczyk, K. PCF-Based Cavity Enhanced
Spectroscopic Sensors for Simultaneous Multicomponent Trace Gas Analysis. Sensors 2011, 11, 1620–1640.
[CrossRef] [PubMed]
Langridge, J.M.; Laurila, T.; Watt, R.S.; Jones, R.L.; Kaminski, C.F.; Hult, J. Cavity enhanced absorption
spectroscopy of multiple trace gas species using a supercontinuum radiation source. Opt. Express 2008, 16,
10178–10188. [CrossRef] [PubMed]
Cordeiro, C.M.B.; Franco, M.A.R.; Chesini, G.; Barretto, E.C.S.; Lwin, R.; Brito Cruz, C.H.; Large, M.C.J.
Microstructured-core optical fibre for evanescent sensing applications. Opt. Express 2006, 14, 13056–13066.
[CrossRef] [PubMed]
Zaatar, Y.; Zaouk, D.; Bechara, J.; Khoury, A.; Llinaress, C.; Charles, J.P. Fabrication and characterization
of an evanescent wave fiber optic sensor for air pollution control. Mater. Sci. Eng. 2000, B74, 296–298.
[CrossRef]
Singh, C.D.; Gupta, B.D. Detection of gases with porous-clad tapered fibers. Appl. Opt. 1995, 34, 1019–1023.
[CrossRef] [PubMed]
J˛edrzejewski, K. Biconical fused taper—A universal fibre devices technology. Opto-Electron. Rev. 2000, 8,
153–159.
Stewart, G.; Jin, W.; Culshaw, B. Prospects for fibre-optic evanescent-field gas sensors using absorption in
the near-infrared. Sens. Actuators B Chem. 1997, 38–39, 42–47. [CrossRef]
Arregui, F.J.; Liu, Y.; Matias, I.R.; Claus, R.O. Optical fiber humidity sensor using a nano Fabry–Perot cavity
formed by the ionic self-assembly method. Sens. Actuators B Chem. 1999, 59, 54–59. [CrossRef]
Pustelny, T.; Maciak, E.; Opilski, Z.; Bednorz, M. Optical interferometric structures for application in gas
sensors. Opt. Appl. 2007, 37, 187–194.
Shadaram, M.; Espada, L.; Martinez, J.; Garcia, F. Modeling and performance evaluation of ferrocene-based
polymer clad tapered optical fiber gas sensors. Opt. Eng. 1998, 37, 1124–1129. [CrossRef]
Shabaneh, A.; Girei, S.; Arasu, P.; Mahdi, M.; Rashid, S.; Paiman, S.; Yaacob, M. Dynamic Response of
Tapered Optical Multimode Fiber Coated with Carbon Nanotubes for Ethanol Sensing Application. Sensors
2015, 15, 10452–10464. [CrossRef] [PubMed]
Arregui, F.J.; Ciaurriz, Z.; Oneca, M.; Matias, I.R. An experimental study about hydrogels for the fabrication
of optical fiber humidity sensors. Sens. Actuators B Chem. 2003, 96, 165–172. [CrossRef]
Michie, W.C.; Culshaw, B.; Konstantaki, M.; McKenzie, I.; Kelly, S.; Graham, N.B.; Moran, C. Distributed
pH and Water Detection Using Fiber-Optic Sensors and Hydrogels. J. Light. Technol. 1995, 13, 1415–1420.
[CrossRef]
Posch, H.E.; Leiner, M.J.P.; Wolfbeis, O.S. Towards a gastric pH-sensor: An optrode for the pH 0–7 range.
Fresenius Zeitschrift Anal. Chem. 1989, 334, 162–165. [CrossRef]
Wolfbeis, O.S. Fiber-Optic Chemical Sensors and Biosensors. Anal. Chem. 2004, 76, 3269–3284. [CrossRef]
[PubMed]
31901
Sensors 2015, 15, 31888–31903
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
McDonagh, C.; Burke, C.S.; MacCraith, B.D. Optical Chemical Sensors. Chem. Rev. 2008, 108, 400–422.
[CrossRef] [PubMed]
Sipior, J.; Bambot, S.; Romauld, M.; Carter, G.M.; Lakowicz, J.R.; Rao, G. A Lifetime-Based Optical CO2 Gas
Sensor with Blue or Red Excitation and Stokes or Anti-Stokes Detection. Anal. Biochem. 1995, 227, 309–318.
[CrossRef] [PubMed]
Hernaez, M.; Zamarreño, C.R.; Goicoechea, J.; Matias, I.R.; Arregui, F.J. Nanostructured Materials in Optical
Fiber Sensing. Open Opt. J. 2013, 7, 84–94. [CrossRef]
Rovati, L.; Fabbri, P.; Ferrari, L.; Pilati, F. Plastic Optical Fiber pH Sensor Using a Sol-Gel Sensing Matrix. In
Fiber Optic Sensors, 1st ed.; Yasin, M., Harum, S.W., Arof, H., Eds.; InTech: Rijeka, Croatia, 2012; pp. 415–438.
Segawa, H.; Ohnishi, E.; Arai, Y.; Yoshida, K. Sensitivity of fiber-optic carbon dioxide sensors utilizing
indicator dye. Sens. Actuators B Chem. 2003, 94, 276–281. [CrossRef]
Panwar, N.; Tiwari, U.; Nidhi Khan, M.M.; Jain, S.C.; Garg, R.; Kapur, P. Long Period Fiber Grating
Humidity Sensor with Gelatin/Cobalt Chloride coating. In Proceedings of the International Conference
on Fiber Optics and Photonics, Madras, India, 9–12 December 2012.
Konstantaki, M.; Pissadakis, S.; Pispas, S.; Madamopoulos, N.; Vainos, N.A. Optical fiber long-period
grating humidity sensor with poly(ethylene oxide)cobalt chloride coating. Appl. Opt. 2006, 45, 4567–4571.
[CrossRef] [PubMed]
Zhao, Z.; Duan, Y. A low cost fiber-optic humidity sensor based on silica sol-gel film. Sens. Actuators B
Chem. 2011, 160, 1340–1345. [CrossRef]
Kolpakov, S.A.; Gordon, N.T.; Mou, C.; Zhou, K. Toward a New Generation of Photonic Humidity Sensors.
Sensors 2014, 14, 3986–4013. [CrossRef] [PubMed]
Yimit, A.; Itoh, K.; Murabayashi, M. Detection of ammonia in the ppt range based on a composite optical
waveguide pH sensor. Sens. Actuators B Chem. 2003, 88, 239–245. [CrossRef]
Timmer, B.; Olthuis, W.; van den Berg, A. Ammonia sensors and their applications—A review.
Sens. Actuators B Chem. 2005, 107, 666–677. [CrossRef]
Wolfbeis, O.S.; Weis, L.J. Fiber-optic Fluorosensor for Oxygen and Carbon Dioxide. Anal. Chem. 1988, 60,
2028–2030. [CrossRef]
Chu, C.S.; Lo, Y.L.; Sung, T.W. Review on Recent Developments of Fluorescent Oxygen and Carbon Dioxide
Optical Fiber Sensors. Photonic Sens. 2011, 1, 234–250. [CrossRef]
Wolfbeis, O.S.; Kovacs, B.; Goswami, K.; Klainer, S.M. Fiber-Optic Fluorescence Carbon Dioxide Sensor for
Environmental Monitoring. Mikrochim. Acta 1998, 129, 181–188. [CrossRef]
Amao, Y.; Nakamura, N. Optical CO2 sensor with the combination of colorimetric change of
α-naphtholphthalein and internal reference fluorescent porphyrin dye. Sens. Actuators B Chem. 2004, 100,
347–351. [CrossRef]
Malins, C.; MacCraith, B.D. Dye-doped organically modified silica glass for fluorescence based carbon
dioxide gas detection. Analyst 1998, 123, 2373–2376. [CrossRef]
Miao, Y.; Liu, B.; Zhang, H.; Li, Y.; Zhou, H.; Sun, H.; Zhang, W.; Zhao, Q. Relative Humidity Sensor
Based on Tilted Fiber Bragg Grating With Polyvinyl Alcohol Coating. IEEE Photonics Technol. Lett. 2009, 21,
441–443. [CrossRef]
Zhang, Z.; Zhang, Y.; Ma, W.; Russell, R.; Shakhsher, Z.M.; Grant, C.L.; Seitz, W.R.; Sundberg, D.C.
Poly(viny1 alcohol) as a Substrate for Indicator Immobilization for Fiber-Optic Chemical Sensors.
Anal. Chem. 1989, 61, 202–205. [CrossRef]
Corres, J.M.; Arregui, F.J.; Matias, I.R. Sensitivity optimization of tapered optical fiber humidity sensors
by means of tuning the thickness of nanostructured sensitive coatings. Sens. Actuators B Chem. 2007, 122,
442–449. [CrossRef]
El Nahhal, I.M.; Zourab, S.M.; Kodeh, F.S.; Qudaih, A.I. Thin film optical BTB pH sensors using sol-gel
method in presence of surfactants. Int. Nano Lett. 2012, 2. [CrossRef]
Abdelghani, A.; Chovelon, J.M.; Jaffrezic-Renault, N.; Lacroix, M.; Gagnaire, H.; Veillas, C.; Berkova, B.;
Chomat, M.; Matejec, V. Optical fibre sensor coated with porous silica layers for gas and chemical vapour
detection. Sens. Actuators B Chem. 1997, 44, 495–498. [CrossRef]
Yeh, T.S.; Chu, C.S.; Lo, Y.L. Highly sensitive optical fiber oxygen sensor using Pt(II) complex embedded in
sol–gel matrices. Sens. Actuators B Chem. 2006, 119, 701–707. [CrossRef]
31902
Sensors 2015, 15, 31888–31903
45.
46.
47.
48.
49.
50.
Renganathan, B.; Sastikumar, D.; Gobi, G.; Rajeswari Yogamalar, N.; Chandra Bose, A. Nanocrystalline ZnO
coated fiber optic sensor for ammonia gas detection. Opt. Laser Technol. 2011, 43, 1398–1404. [CrossRef]
El-Nasser, H.M. Effects of methyl red acidity and UV illumination on absorption coefficient of MR/PVA
thin films. Phys. B 2011, 406, 1940–1943. [CrossRef]
De Meyer, T.; Hemelsoet, K.; Van Speybroeck, V.; De Clerck, K. Substituent effects on absorption spectra of
pH indicators: An experimental and computational study of sulfonphthaleine dyes. Dyes Pigment. 2014,
102, 241–250. [CrossRef]
Balderas-Hernandez, P.; Ramirez, M.T.; Rojas-Hernandez, A.; Gutierrez, A. Determination of pKa’s for
thymol blue in aqueous medium: Evidence of dimer formation. Talanta 1998, 46, 1439–1452. [CrossRef]
Lide, D.R. CRC Handbook of Chemistry and Physics, 86th ed.; CRC Press: Boca Raton, FL, USA, 2005.
Hołowacz, I.; Podbielska, H.; Bauer, J.; Ulatowska-Jarża, A. Viscosity, surface tension and refractive index of
tetraethylorthosilicate-based sol-gel materials depending on ethanol content. Opt. Appl. 2005, 35, 691–699.
© 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open
access article distributed under the terms and conditions of the Creative Commons by
Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
31903