Photonic crystal fibre as an optofluidic reactor for the measurement of photochemical kinetics with sub-picomole sensitivity

Article abstract of DOI:10.1039/c2lc40062f

Photonic crystal fibre constitutes an optofluidic system in which light can be efficiently coupled into a solution-phase sample, contained within the hollow core of the fibre, over long path-lengths. This provides an ideal arrangement for the highly sensitive monitoring of photochemical reactions by absorption spectroscopy. We report here the use of UV/vis spectroscopy to measure the kinetics of the photochemical and thermal cis-trans isomerisation of sub-picomole samples of two azo dyes within the 19-μm diameter core of a photonic crystal fibre, over a path length of 30 cm. Photoisomerisation quantum yields are the first reported for "push-pull" azobenzenes in solution at room temperature; such measurements are challenging because of the fast thermal isomerisation process. Rate constants obtained for thermal isomerisation are in excellent agreement with those established previously in conventional cuvette-based measurements. The high sensitivity afforded by this intra-fibre method enables measurements in solvents in which the dyes are too insoluble to permit conventional cuvette-based measurements. The results presented demonstrate the potential of photonic crystal fibres as optofluidic elements in lab-on-a-chip devices for photochemical applications. This journal is

Full text of DOI:10.1039/c2lc40062f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Lab on a Chip
Cite this: Lab Chip, 2012, 12, 3356–3361
www.rsc.org/loc
PAPER
Photonic crystal fibre as an optofluidic reactor for the measurement of
photochemical kinetics with sub-picomole sensitivity{
Gareth O. S. Williams,a Jocelyn S. Y. Chen,b Tijmen G. Euser,b Philip St.J. Russellb and Anita C. Jones*a
Received 13th January 2012, Accepted 8th June 2012
DOI: 10.1039/c2lc40062f
Photonic crystal fibre constitutes an optofluidic system in which light can be efficiently coupled into a
solution-phase sample, contained within the hollow core of the fibre, over long path-lengths. This
provides an ideal arrangement for the highly sensitive monitoring of photochemical reactions by
absorption spectroscopy. We report here the use of UV/vis spectroscopy to measure the kinetics of the
photochemical and thermal cis–trans isomerisation of sub-picomole samples of two azo dyes within
the 19-mm diameter core of a photonic crystal fibre, over a path length of 30 cm. Photoisomerisation
quantum yields are the first reported for ‘‘push–pull’’ azobenzenes in solution at room temperature;
such measurements are challenging because of the fast thermal isomerisation process. Rate constants
obtained for thermal isomerisation are in excellent agreement with those established previously in
conventional cuvette-based measurements. The high sensitivity afforded by this intra-fibre method
enables measurements in solvents in which the dyes are too insoluble to permit conventional cuvette-
based measurements. The results presented demonstrate the potential of photonic crystal fibres as
optofluidic elements in lab-on-a-chip devices for photochemical applications.
optofluidic device that combines a long path-length with a small
Introduction
cross-sectional area, in which the entire (small) sample volume is
The application of photochemistry is seeing growth in many
exposed to both the photochemical excitation (pump) beam and
areas, including photo-medicine,1 artificial photosynthesis,2,3
the spectroscopic probe beam. These requirements are hard to
optically switchable media for data storage4 and the control of
molecular machines,5 stimulating the synthesis and characterisa-
tion of novel photoactive compounds. There is also a drive for
the miniaturisation and integration of sample preparation and
monitoring systems into lab-on-a-chip devices. This approach
can be applied to photochemical reactions by utilising opto-
fluidic systems, in which optical excitation and detection systems
are integrated with microfluidics.6 UV/visible (UV/vis) absorp-
tion spectroscopy is a convenient and commonly used method
for monitoring the progress of photochemical reactions in
conventional cuvette-based measurements, but is relatively
insensitive because of the short path-lengths that are typically
used. Increasing the path-length is often impractical because of
the constraints of commercial spectrometers or the limited
availability of sample. The use of glass capillaries reduces the
required sample volume but simple capillaries have very high
optical losses and are highly sensitive to bending, greatly limiting
the possible length that can be used.7 The ideal sample cell for
the spectroscopic study of photochemical reactions is an
achieve in conventional lab-on-a-chip designs.8,9,10
Integration of lasers and a micro-cuvette on a chip has been
achieved11 and provides small volumes, but, as the sample was
not contained within a waveguide, the distance over which it
could be excited was limited to y 1 mm. Other systems have
used liquid core waveguides with guidance through metal
jackets.12 These designs experience high loss due to surface
imperfections, reducing guidance and limiting the path length
that can be used.
Progress in implementing long path-length absorption spectro-
scopy in an optofluidic system has been made through the use of
photonic crystal fibre (PCF). In a PCF, light is trapped inside the
fibre core by a periodic wavelength-scale lattice of microscopic
holes in the cladding glass, a ‘‘photonic crystal’’.13 Initial studies
were performed using solid-core PCF (SC-PCF), a fibre with a
solid silica core surrounded by a lattice of air holes producing the
cladding crystal structure. The sample is infiltrated into the
cladding of the fibre and probed by the evanescent field of the
light guided through the core.14,15 For example, Cordeiro et al.
showed enhanced absorption sensitivity for Methylene Blue dye
using a micro-structured fibre core,16 whilst Euser et al.
demonstrated the use of a suspended-core fibre for absorption
sensing of aqueous NiCl2 with a sensitivity two orders of
magnitude greater than in conventional systems.17 Whilst SC-
PCF is effective in achieving detection over long path-lengths,
there is limited overlap between the guided light and the sample,
aEaStCHEM School of Chemistry, King’s Buildings, The University of
Edinburgh, Edinburgh, EH9 3JJ, United Kingdom.
E-mail: a.c.jones@ed.ac.uk; Fax: +44(0)1316504753
bMax Planck Institute for the Science of Light, Gu¨nther-Scharowsky-Str.
1/Bldg. 24, 91058 Erlangen, Germany
{ Electronic supplementary information (ESI) available. See DOI:
10.1039/c2lc40062f
3356 | Lab Chip, 2012, 12, 3356–3361
This journal is ß The Royal Society of Chemistry 2012

View Article Online
providing a less-than-ideal probe regime. Furthermore, the
probed sample is in very close proximity to the silica fibre
surface, where its properties may differ from those in the bulk
solution. These disadvantages are overcome by the use of
hollow-core PCF (HC-PCF), in which light is guided through the
sample solution in the hollow core of the fibre.
high quantum yield requires the use of a low probe intensity to
avoid probe-induced photochemistry.
We note that the photogeneration and detection of transient
species within a HC-PCF has been reported previously by
Khetani et al., using laser flash photolysis.21 However, they used
a perpendicular arrangement of pump and probe beams, such
that the sample solution was excited perpendicular to the fibre
core, through the cladding, and monitored by probe light guided
through the core. Thus, the path-length for pump beam was
extremely short, only a few microns (the diameter of the fibre
core), requiring the use of a concentrated (mM) sample solution,
and the effective probe path length was only y1 cm. This study,
therefore, failed to exploit the large gain in sensitivity offered by
coaxial guiding of both pump and probe beams through the
hollow core. Also, the cladding holes of the fibre used were
collapsed, affecting the guidance mechanism of the fibre. When
the cladding holes are collapsed, the core of the fibre has a higher
effective refractive index than the cladding and guides by the
mechanism of total internal reflection. The relatively large
diameter of the core means that a high number of modes are
supported and single-mode guidance is very hard to achieve.
Two azobenzene-based dyes, Disperse Red 1 (DR1), N-ethyl-
N-(2-hydroxyethyl)-4-(4-nitrophenylazo)aniline, and Disperse
Orange 1, 4-anilino-49-nitroazobenzene, were chosen for the
present study. The structures of the dyes, together with the cis–
trans isomerisation reaction, are shown in Fig. 2.
The structure of the kagome´ HC-PCF used in this study,
shown in Fig. 1(C) and (D), allows the guidance of light, in a
single well-defined mode, over long path-lengths through the
core region. When the fibre microstructure is filled with a sample
solution there is near perfect light-sample overlap within the
fibre core.18,19 The dimensions of the kagome´ PCF are compared
to a standard 1-cm cuvette in Fig. 1(A) and (B). Irradiation
across the full width of the cuvette, over a path-length of 1 cm,
requires a minimum sample volume of 1 ml, whereas the 19-mm
core of the HC-PCF holds only 3 nl per cm of path-length. In the
present work, a 30-cm length of fibre was used, containing a
sample volume of 90 nl. Both excitation and broadband
measurement sources can be co-coupled into the fibre core,
giving excellent overlap of both with the sample (Fig. 1(E) and
(F)). Furthermore, precise knowledge of the excitation intensity
and probed volume facilitates quantitative modelling of the
kinetics of a photochemical reaction taking place within the
fibre. The kagome´ HC-PCF used in these experiments exhibits
losses of y1 dBm21 allowing, in principle, several metres of fibre
to be used before loss hinders measurements. Moreover, this
fibre type supports broadband guidance, permitting measure-
ment of the complete visible absorption spectra of the azo dyes
studied here.
The isomerisation of azobenzene and its derivatives is
considered a paradigm of unimolecular photochemistry, and
the kinetic model is well-established, as briefly summarised here.
The trans isomer is the thermally stable species and the un-
irradiated dye exists exclusively in this form. Irradiation, by
either UV or visible light, induces reversible photoisomerisation,
such that an equilibrium between trans and cis forms, known as
the photostationary state (PSS), is attained. There is a high
thermal barrier to the forward process which, therefore, only
occurs photochemically, but the reverse process occurs both
photochemically and thermally. The photoisomerisation of azo
dyes is typically accompanied by a colour change (photochro-
mism), allowing the process to be monitored by absorption
spectroscopy.22,23,24,25,26,27 The dyes studied here belong to a
particular class of azobenzene derivative known as ‘‘push-pull’’
azobenzenes. These contain electron donating and electron
accepting substituents at the 4 and 49 positions, respectively,
para to the azo bond. This substitution increases electron
delocalisation, affecting the ordering of the excited state energy
Previously, we have demonstrated the use of HC-PCF as a
photochemical reactor to monitor the photolysis of vitamin B12,
cyanocobalamin, an irreversible reaction with very low quantum
yield (y 1024).20 The strong confinement of both sample and
light in the hollow core resulted in a 1000-fold increase in
reaction rate over that in a conventional cuvette. We now report
the extension of this approach to the measurement of the kinetics
of thermally reversible photochemical reactions with relatively
high quantum yield, the photoisomerisation of azo dyes. The
spectroscopic monitoring of such reactions poses challenges that
were not encountered in our previous study. The thermal
reversibility demands that the spectroscopic measurement is
near-simultaneous with the photochemical excitation and the
Fig. 1 Schematic comparison of the irradiated sample volumes in a
conventional cuvette (A) and a HC-PCF (B). High-resolution scanning
electron micrographs of a kagome´ HC-PCF with a core diameter of
19 mm (C–D). Transverse irradiance profile measured using the 488-nm
laser (E) and the broadband xenon lamp (F).
Fig. 2 (A) Structure of trans Disperse Orange 1; (B) the reversible
isomerisation of Disperse Red 1, from trans (left) to cis (right); hn
indicates a photochemical process and D a thermal process.
This journal is ß The Royal Society of Chemistry 2012
Lab Chip, 2012, 12, 3356–3361 | 3357

View Article Online
levels and shifting the most intense electronic transitions into the
visible part of the spectrum. A further consequence of this
substitution is an acceleration of the rate of thermal cis-to-trans
isomerisation and a strong dependence of this rate on solvent
polarity.28 Measuring the photoisomerisation quantum yields of
push-pull azobenzenes is challenging because of the fast thermal
isomerisation and we are not aware of any previous measure-
ments in solution at room temperature. A very recent review28
cites only one study,29 carried out at low temperature to prevent
the thermal process.
Azo dyes have long been used as dyes and colorants,30 but
more recent applications seek to exploit the ability to reversibly
photoswitch the geometric, electronic and optical properties of
the molecule. Such applications include memories and switches
in new electronic and photonic systems,31,32 optical control of
various physical properties of materials33,34 and optically
switchable components in molecular machines35 and biomole-
cules.36 Measurement of the kinetics of azo isomerisation is
central to the design and optimisation of optical switching
characteristics for specific applications.
maintaining optical alignment and minimising dead volume. An
optical window on the cell allows coupling of light into the fibre.
The pump wavelength lies within the probed spectral region,
so saturation of the spectrometer by the relatively intense pump
laser must be avoided. Therefore, electronic shutters, controlled
by the PC, are used to implement a sequential pump–probe
technique. A reference spectrum of the lamp is acquired for each
measurement to correct for any instability in the probe source.
Measurements and shutter control are automated using a
Matlab computer program, allowing the photoisomerisation
process to be monitored by acquisition of the complete
absorption spectrum on the 10s of millisecond timescale.
The experimental photoisomerisation data are fitted numeri-
cally by a kinetic model with three variable parameters, the rate
constants for the three isomerisation processes: trans-to-cis
photoisomerisation, cis-to-trans photoisomerisation and cis-to-
trans thermal isomerisation. The precisely defined excitation
conditions make measurement of the irradiation intensity
straightforward, without the use of actinometry, which is
necessary in cuvette-based experiments. The kinetic model is
described in detail in the Supplementary Information. The model
takes account of the fall-off in excitation intensity along the
length the fibre, due to both absorption and the intrinsic fibre
losses. For measurements in the pump–probe mode, the model
also allows for the occurrence of thermal relaxation from cis to
trans isomer during the probe periods, when the excitation laser
is off. The rate constant for thermal cis to trans isomerisation is
determined independently by monitoring the reverse reaction
from the PSS in the absence of irradiation; this value is then fixed
in the model. Application of the model to the photoismerisation
data returns the value of the photoisomerisation quantum yield.
Measurements were carried out on both dyes in cyclohexane
and toluene. DR1 was also measured in pentane, a solvent in
which the dye is highly insoluble, precluding conventional
cuvette-based measurement. The 30-cm path-length permitted
use of dye concentrations as low as 1 mM corresponding to less
than 100 fmol of dye in the measurement volume of 90 nL.
Experimental
The photonic crystal fibre was fabricated in-house at the Max
Planck Institute for the Science of Light. Disperse Red 1 and
Disperse Orange 1 were purchased from Sigma Aldrich and used
as received. All solvents were of HPLC grade.
The experimental system is shown in Fig. 3. A 488-nm diode
laser (photochemical pump) and a broadband xenon lamp
(spectroscopic probe) are co-coupled into a 30-cm length of the
HC-PCF, via an endlessly-single-mode fibre, to ensure both have
the same modal properties. This enables spectroscopic measure-
ments of precisely the same sample volume as that being excited
by the laser source. Each end of the hollow core fibre is mounted
in a custom-built liquid cell (Fig. 3 inset) to allow introduction of
the sample solution into the fibre, via a syringe pump, whilst
Results and discussion
As shown in Fig. 4, photoisomerisation of trans DR1 results in a
decrease in intensity of the visible absorption band between 400
and 550 nm. Similar behaviour is seen for DO1. Fig. 4 compares
spectra recorded for a 100-mM solution in a 1-cm cuvette (in a
conventional UV/vis spectrometer) with those measured for a
5-mM solution in a 30-cm PCF. The spectral profiles are in good
agreement; however, the absorbance measured in the PCF is
somewhat lower than that predicted on the basis of the Beer–
Lambert law. This can be attributed to adsorption of dye
molecules to the fibre walls, where they are not exposed to
guided light, which reduces the effective sample concentration.
Allowing the sample to flow through the fibre saturates the
surface and the recorded absorbance approaches that predicted.
There is the potential to use surface-functionalised fibres to
inhibit adsorption.
Fig. 3 Schematic diagram of the experimental setup. The 488-nm diode
laser (L) and the broadband xenon lamp (X) are co-aligned and coupled
into a short piece of endlessly single mode fibre (ESM) for spatial
filtering, before being coupled into the HC-PCF containing the sample
(SF). The transmitted light and a reference beam (directly from the xenon
lamp) are delivered, via multimode fibres (MMF), to the dual-channel
spectrometer (SPEC) interfaced to a PC. The laser and lamp outputs and
the input to the spectrometer are controlled by electronic shutters (S).
The inset shows the fluid cell through which the sample solution is
introduced into the core of the fibre.
As a result of the tight confinement of the guided light in the
core of the sample fibre, very low source powers are needed to
produce sufficient power density to induce the photochemical
process. Consequently, the low-power broadband xenon lamp,
3358 | Lab Chip, 2012, 12, 3356–3361
This journal is ß The Royal Society of Chemistry 2012

View Article Online
absorption is saturated and no further increase in rate is seen with
increasing power. As the power is decreased, the rate of
attainment of the PSS decreases as expected. The time taken to
reach the PSS is plotted as a function of incident power in
Fig. 5(B), also shown for comparison is the result of a
conventional cuvette measurement on the same solution. Almost
three orders of magnitude more power was needed to reach the
PSS in a comparable time in the cuvette: irradiation with 100 mW
produced the PSS in 20 s, compared with ,0.1 mW in the PCF.
Although the total power entering the fibre is three orders of
magnitude lower, the power density in the 19-mm fibre core is
approximately three orders of magnitude higher, 60 mWcm22
compared to y 50 mWcm22 for the cuvette measurement.
,
Using the xenon lamp as both pump and probe, the
photoisomerisation of DO1 was monitored on the millisecond
timescale by recording the complete absorption spectrum, with
an integration time of 10 ms, at 100 ms intervals. The measured
temporal evolution of the absorbance at 470 nm is shown in
Fig. 6 (A), together with the fitted kinetic function. Whilst this
offers an attractively simple method of simultaneously exciting
and probing the photoswitching process, the use of polychro-
matic excitation hinders modelling of the kinetics because of the
need to take account of the wavelength-dependence of the
irradiation intensity and the absorption cross-sections of the two
isomers. In the present case, an approximate model was used
in which the total incident intensity and spectrally averaged
cross-sections were used. This gave an estimate of 0.23 for the
trans-to-cis photoisomerisation quantum yield of DO1. To our
knowledge this is the first reported measurement of this
parameter.
Fig. 4 Absorption spectra of Disperse Red 1 in toluene measured
during photoisomerisation in (A) a 1-cm cuvette, at 100 mM (in a UV/vis
spectrometer, scan time 45 s) and (B) in a 30-cm kagome´ PCF at 5 mM
(acquisition time 500 ms). The arrow indicates the direction of
photoisomerisation from the trans to the cis isomer.
intended for use as the spectroscopic probe source, can produce
photoisomerisation. This is illustrated in Fig. 5(A), which shows
the effect of low-power irradiation of a 5-mM cyclohexane
solution of DO1. At a lamp power of approximately 1 mW, the
PSS is attained in less than 10 s. At this power, the sample
Fig. 6 (A) The photoisomerisation kinetics of DO1 in toluene at a
concentration of 2 mM. The experimentally measured temporal evolution
of the absorbance at 470 nm (&), using the broadband output of the
xenon lamp as both pump and probe, and the fitted theoretical function
(–). (B) The photoisomerisation kinetics of DR1 in toluene at 5 mM.
The experimentally measured temporal evolution of the absorbance at
488 nm (&), using the diode laser as pump and the xenon lamp as probe,
and the fitted theoretical function (–).
Fig. 5 The effect of incident excitation power on the time taken to reach
the photostationary state for Disperse Orange 1 in cyclohexane at 5 mM,
in a 30-cm kagome´ fibre. (A) Temporal evolution of absorbance at
470 nm for excitation powers of 1.15 mW, 0.2 mW and 0.02 mW. (B) Time
taken to reach the photostationary state as a function of excitation
power, I0. The solid line is a guide only. The isolated point is the value for
an equivalent cuvette measurement.
This journal is ß The Royal Society of Chemistry 2012
Lab Chip, 2012, 12, 3356–3361 | 3359

View Article Online
To facilitate quantitative kinetic modelling, it is desirable to
reduce the probe lamp intensity to 0.02 mW or less, so that its
influence becomes negligible over the time interval (y 500 ms)
needed to acquire the absorption spectrum (see Fig. 5A), and use
an independent monochromatic (or narrow-band) pump source.
The results of measurement of the photoswitching kinetics of
DR1 in toluene using this regime, with the 488-nm diode laser as
pump, are shown in Fig. 6B. Exposure of the sample to the
488-nm beam for 300 ms was followed by acquisition of the
absorption spectrum over a 500-ms interval, giving a total of
800 ms for each pump–probe measurement. Excitation at a
single wavelength makes determination of the photoisomerisa-
tion quantum yield straightforward and a value of 0.19 was
obtained for the trans to cis isomerisation of DR1. This is in
good agreement with the value of 0.2 which was reported
recently for DR1 deposited as a solid film (in which thermal
isomerisation is inhibited).37
perturbed by containment of the solution in the PCF and
confirms that the sample volume that is probed is the bulk
solution, not surface-bound molecules. The low solubility of
push-pull azobenzenes in non-polar solvents is problematic for
cuvette-based measurements and we are not aware of any
previous studies in very low polarity solvents, such as pentane;
the solubility of DR1 in pentane is only 1.3 mM.
Conclusion
Hollow-core PCF is a highly effective optofluidic system for the
measurement of photochemical kinetics by absorption spectro-
scopy. The extrusion of sub-microlitre sample volumes over
path-lengths of tens of centimetres enables highly sensitive
measurements on sub-picomole quantities of sample. The power
of this methodology has been illustrated by its application to the
reversible photoisomerisation of two push-pull azobenzenes.
Although the compounds used in this study are readily available,
the measurement of the isomerisation kinetics of this technolo-
gically important class of azobenzenes, using conventional
cuvette-based methods, has been hindered by the rapid thermal
isomerisation in polar solvents and low solubility in non-polar
solvents. Judicious choice of solvent has enabled some previous
conventional measurements of isomerisation rate constants
which have been valuable in validating the PCF technique for
quantitative kinetic measurements. Having demonstrated excel-
lent agreement of rate constants measured for the thermal cis–
trans isomerisation of DR1 and DO1 with available literature
values, we have extended our experiments to conditions that are
intractable to traditional methods.
We now turn to the measurement of the rate of thermal
conversion of the photochemically-formed cis isomer to the more
stable trans form. The temporal evolution of the absorbance, for
DO1 in three solvents, during the thermal isomerisation process
is shown in Fig. 7. The rate constants obtained, together with
those measured for DR1, are given in Table 1. Values reported
from previous conventional measurements (where available) are
also presented in Table 1, for comparison.
As observed in previous studies,24,28,38,39 the rate of thermal
isomerisation decreases with the polarity of the solvent. The
values obtained in the present experiments are in very good
agreement with those from previous, conventional measure-
ments. This demonstrates that the isomerisation process is not
We have exploited the ease of photometry afforded by the
near-perfect overlap of both pump and probe light with the
sample, within the hollow core of PCF, to make the first
measurement of the trans–cis photoswitching quantum yield of a
push-pull azobenzene in solution at room temperature. The very
few, previous measurements of quantum yield have used low
temperature or solid samples to inhibit thermal isomerisation.
The photochemical quantum yield is a crucial parameter and the
ability to measure this on a small quantity of sample under
ambient conditions is highly advantageous.
The sensitivity afforded by the long (30-cm) path-length of the
PCF enabled the first measurement of the thermal rate constant
of a push-pull azobenzene in a highly non-polar solvent, pentane,
in which the solubility is very low. A concentration of 1 mM was
used, equivalent to less than 100 femtomoles of DR1 in the
measurement volume. Knowledge of the isomerisation rate
constant under conditions of minimal solvent interaction is
essential for understanding the influence of molecular environ-
ment on the mechanism of the reaction. Experimental values of
thermal isomerisation rate constants in non-interacting solvents
are particularly valuable for validating quantum chemical
calculations, which typically consider isolated molecules in the
gas phase. Recent calculations26,40 predict rate constants in the
region of 5–10 s21 for push-pull azobenzenes, more than an
order of magnitude greater than the value measured here for
DR1 in pentane. This highlights the need for the inclusion of
realistic solvation effects in computational models.
Fig. 7 Kinetics of the thermal isomerisation of cis-DO1 in toluene,
cyclohexane and pentane at room temperature. The temporal evolution
of the absorbance at 470 nm, measured in the absence of pump
irradiation, following attainment of the PSS. Experimental data are
shown as points and the fitted theoretical functions as solid lines.
Table 1 The thermal isomerisation rate constant, kD, for DR1 and DO1
in various solvents. Values obtained in the present study are compared
with those from previous, conventional measurements
kD / 1023 21
s
Dye
Solvent
This Study
Literature
DR1
DO1
Toluene
30
35,25 2826
N/A
N/A
N/A
0.9,27 0.6,24 0.823
Cyclohexane
Pentane
Toluene
1.0
0.6
20
The impact of photochemistry in the physical, biological and
medical sciences, and in many modern technologies has increased
Cyclohexane
0.8
3360 | Lab Chip, 2012, 12, 3356–3361
This journal is ß The Royal Society of Chemistry 2012

View Article Online
12 P. Measor, B. S. Phillips, A. Chen, A. R. Hawkins and H. Schmidt,
Lab Chip, 2011, 11, 899–904.
greatly in recent years. New photoactive compounds are
emerging in increasing numbers from synthetic chemistry
laboratories, with a view to exploiting light-driven molecular
processes in applications ranging from diagnostics and photo-
therapy to the harnessing of solar energy for green synthesis,
water splitting, molecular photovoltaics and the capture and
reduction of CO2. There is undoubtedly a demand for micro-
analysis systems capable of delivering comprehensive, quantita-
tive, photochemical characterisation of the small sample
quantities typically available from research-scale synthesis. The
incorporation of PCFs as photochemical reactors in lab-on-a-
chip devices could satisfy this need and revolutionise high-
throughput screening of photoactive targets. There is also the
attractive prospect of implementing photochemistry-based
assays in micro-total-analysis systems; for example, combining
light-controlled release of photocaged probes or analytes with
highly sensitive spectroscopic detection. On the technical side,
the development of a robust, easy-to-use, all-fibre optofluidic
photoreactor can be envisaged, in which fibre lasers and/or PCF-
based supercontiuuum sources, as pump and probe, are
integrated with the sample-containing PCF.
13 P. S. J Russel, J. Lightwave Technol., 2006, 24, 4729–4746.
14 C. M. Cordeiro, M. A. Franco, C. J. Matos, F. Sircilli, V. A. Serra˘o
and C. H. Cruz, Opt. Lett., 2007, 32, 3324–3326.
15 Y. L. Hoo, W. Jin, H. L. Ho, D. N. Wang and R. S. Windeler, Opt.
Eng., 2002, 41, 8–9.
16 C. M. B. Cordeiro, M. A. R. Franco, G. Chesini, E. C. S. Barretto, R.
Lwin, C. H. B. Cruz and M. C. J. Large, Opt. Express, 2006, 14,
13056–13066.
17 T. G. Euser, J. S. Y. Chen, M. Scharrer, P. S. J. Russell, N. J. Farrer,
P. J. Sadler, P. St and J. Russell, J. Appl. Phys., 2008, 103, 103108.
18 L.W. Kornaszewski, N. Gayraud, J. M. Stone, W. N. MacPherson,
A. K. George, J. C. Knight, D. P. H and D. T. Reid, Opt. Express,
2007, 15, 11219–11224.
19 T. Ritari, J. Tuominen, H. Ludvigsen, J. C. Petersen, T. Sørensen,
T. P. Hansen and H. R. Simonsen, Opt. Express, 2004, 12,
4080–4087.
20 J. S. Y. Chen, T. G. Euser, N. J. Farrer, P. J. Sadler, P. St. and J.
Russell, Chem. Eur. J., 2010, 16, 5607–5612.
21 A. Khetani, M. Laferrie`re, H. Anis and J. C. Scaiano, J. Mater.
Chem., 2008, 18, 4769–4774.
22 K. M. Tait, J. A. Parkinson, S. P. Bates, W. J. Ebenezer and A. C.
Jones, J. Photochem. Photobiol., A, 2003, 154, 179–188.
23 S. R. Hairi, G. A. Taylor and L. W. Schultz, J. Chem. Educ., 1990,
67(8), 709–712.
24 S. Kobayashi, H. Yokohama and H. Kamei, Chem. Phys. Lett., 1987,
138, 333.
25 M. Poprawa-Smoluch, J. Baggernam, H. G. Zhang, H. P. A. Maas,
L. De Cola and A. M. Brouwer, J. Phys. Chem. A, 2006, 110,
11926–11937.
Acknowledgements
This work forms part of a project funded by the Koerber
Foundation. G.O.S.W. and A.C.J. also acknowledge the UK
EPSRC for funding. G.O.S.W. acknowledges programming
support from G.J.H. Williams.
26 N. A. Wazzan, P. R. Richardson and A. C. Jones, Photochem.
Photobiol. Sci., 2010, 9, 968–974.
27 K. Gille, H. Knoll and K. QuitzschInt, Int. J. Chem. Kinet., 1999, 31,
337–350.
28 H. M. D. Bandara and S. C. Burdette, Chem. Soc. Rev., 2012, 41,
1809, DOI: 10.1039/c1cs15179g.
29 G. Gabor and E. Fisher, J. Phys. Chem., 1971, 75, 581–583.
30 M. W. Mosher and J. M. Ansell, J. Chem. Educ., 1975, 52, 195–196.
31 K. G. Yager and C. J. Barrett, J. Photochem. Photobiol., A, 2006, 182,
250–261.
References
1 D. Phillips, Photochem. Photobiol. Sci., 2010, 9, 1589–1596.
2 D. Gust, T. A. Moore and A. L. Moore, Acc. Chem. Res., 2009, 42,
1890–1898.
3 J. S. Lee, S. H. Lee, J. H. Kim and C. B. Park, Lab Chip, 2011, 11,
2309–2311.
4 F. Li, J. Zhuang, G. Jiang, H. Tang, A. Xia, L. Jaing, Y. Song, Y. Li
and D. Zhu, Chem. Mater., 2008, 20, 1194–1194.
5 V. Balzani, A. Credi and M. Venturi, Chem. Soc. Rev., 2009, 38,
1542–1550.
6 H. Hamish and J. Wilkinson, Microfluid. Nanofluid., 2008, 4, 53–79.
7 E. A. J. Marcatili and R. A. Schmeltzer, Bell Syst. Tech. J., 1964, 43,
1783–1809.
32 M. Irie, Chem. Rev., 2000, 100, 1683–1684.
33 M. B. Sponsler, in Optical Sensors and Switches, Ed. V. Ramamurthy,
K. S. Schanze, Marcel Dekker, New York and Basel, 2001, Chapter 8.
34 R. H.E. l. Halabieh, O. Mermut and C. J. Barrett, Pure Appl. Chem.,
2004, 76, 1445–1465.
35 V. Balzani, A. Credi and M. Venturi, Pure Appl. Chem., 2003, 75,
541–547.
36 A. Beharry Andrew and G. Andrew Woolley, Chem. Soc. Rev., 2011,
40, 4422–4437.
37 J. Shan, S. Xu, W. Shi, L. Liu and L. Xu, Control of Photoisomerization
Quantum Efficiency by Metallic Nanostructures,in Laser Science, OSA
Technical Digest (CD), Optical Society of America, 2010, paper
LTuA6.
8 C. Monat, P. Domachuk and B. J. Eggleton, Nat. Photonics, 2007, 1,
106–114.
9 P. Anzenbacher and M. A. Palacios, Nat. Chem. Biol., 2009, 5, 80–86.
10 L. Rindorf, J. B. Jensen, M. Dufva, L. H. Pedersen, P. E. Høiby and
O. Bang, Opt. Express, 2006, 14, 8224–8231.
11 V. Horowitz, D. Awschalom and S. Pennathur, Lab Chip, 2008, 8,
1856–1863.
38 P. D. Wildes, J. G. Pacifici Irick Jr and D. G. Whitten, J. Am. Chem.
Soc., 1971, 93, 2004–2008.
39 T. Asano and T. Okada, J. Org. Chem., 1984, 49, 4387–4391.
40 J. Dokic, M. Gothe, J. Wirth, M. V. Peters, J. Schwarz, S. Hecht and
P. Saalfrank, J. Phys. Chem. A, 2009, 113, 6763–6773.
This journal is ß The Royal Society of Chemistry 2012
Lab Chip, 2012, 12, 3356–3361 | 3361