A microreactor with phase-change microvalves for batch chemical synthesis at high temperatures and pressures

Article abstract of DOI:10.1039/c3lc50939g

We present a simple microreactor with dimethyl sulfoxide (DMSO) phase-change valves suitable for performing batch organic chemistry under high temperature and pressure conditions. As a proof of principle, we demonstrate a radiofluorination reaction import

Full text of DOI:10.1039/c3lc50939g

Lab on a Chip
View Article Online
View Journal | View Issue
TECHNICAL INNOVATION
A microreactor with phase-change microvalves for
batch chemical synthesis at high temperatures
and pressures†
Cite this: Lab Chip, 2014, 14, 280
*
Xiaoxiao Ma, Wei-Yu Tseng, Mark Eddings, Pei Yuin Keng and R. Michael van Dam
We present a simple microreactor with dimethyl sulfoxide (DMSO) phase-change valves suitable for
performing batch organic chemistry under high temperature and pressure conditions. As a proof of
principle, we demonstrate
a new positron emission tomography biomarker for immune system monitoring and prediction of
chemotherapy response. We achieved high radioactivity recovery (97 1%, n = 3) and conversion
efficiency (83 1%, n = 3), comparable to that achieved with macroscale systems, but with a volume
a radiofluorination reaction important in the synthesis of [
18F]FAC,
Received 12th August 2013,
Accepted 8th November 2013
30× smaller. This platform overcomes the limitations of previously reported phase-change valves in
terms of compatibility with organic chemistry, and extends the range of reaction conditions for carrying
out harsh batch chemistry at the microscale.
DOI: 10.1039/c3lc50939g
www.rsc.org/loc
vapor pressures of solvents well in excess of 100 psi.9,10 Pres-
surized conditions may also be important more generally in
Introduction
Performing chemical reactions in a microscale format has
significant advantages for many applications.1 Batch micro-
reactors, as opposed to continuous flow microreactors, pro-
vide the ability to perform reactions in extremely small total
volumes. They are especially suited for working with scarce
reagents, such as natural products or isolated proteins, or
for producing short-lived radiolabeled molecular imaging
probes for positron emission tomography (PET),2 where it
is desirable to work with extremely low mass quantities.
Another advantage of batch microreactors is the ability to
implement multi-step processes in a single reactor, simplify-
ing the device and minimizing transfer losses.
Though several microfluidic batch reaction platforms have
been reported in the past decade3–7 they tend to be limited to
more mild reaction conditions (i.e., lower temperatures and
pressures) than their continuous flow counterparts. Handling
high pressures is important for many applications, however.
To ensure fast kinetics and high yields, the radiosynthesis of
short-lived imaging probes for PET is often performed in a
sealed vessel under super-heated conditions, resulting in the
generation of significant solvent vapor pressure. For example,
the probe 1-(2′-deoxy-2′-[18F]fluoroarabinofuranosyl)cytosine
([18F]FAC)8 and its analogs require the containment of the
organic chemistry as pressure can accelerate reactions that
are accompanied by a decrease in volume and shift the equi-
libria toward the side of the products.11 In some other
cases, high pressures are generated by the formation of
gaseous productions.
Current batch microreactors are typically incompatible with
harsh conditions due to several limitations. Pressure compati-
bility of poly(dimethylsiloxane) (PDMS)-based microreactors3,4
is severely restricted by the inter-layer bonding strength as
well as by vapor-permeability of the PDMS itself; furthermore,
PDMS is incompatible with many solvents and reagents.
Digital microfluidic chips,5,6 made of materials including glass
and fluoropolymers, have emerged as a more chemically inert
batch reaction platform, but these chips have an open struc-
ture and cannot perform pressurized processes unless the
entire chip is placed inside a pressure chamber.12 Recently,
using a closed geometry, and rigid, non-permeable polymers,
Lebedev et al.7 recently demonstrated a 50 μL batch reactor
for multi-step reactions that can withstand pressures up to
300 psi; however, the device assembly and fabrication are
complex and the large size and positioning of the off-chip
valve actuators increases the overall footprint of the
microreactor system and risk of mechanical failure.
Microvalve technologies reported capable of withstanding
high pressure can be potentially employed to develop batch
microreactors that do not suffer the above limitations.
Although various microvalves have been reported,13–20 very
few of them have been extensively developed for practical
use. MEMS-based mechanical microvalves, where deformable
Department of Molecular and Medical Pharmacology, Crump Institute for
Molecular Imaging, University of California, Los Angeles (UCLA), Los Angeles,
CA 90095, USA. E-mail: mvandam@mednet.ucla.edu
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3lc50939g
280 | Lab Chip, 2014, 14, 280–285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Lab on a Chip
Technical Innovation
membranes are coupled to magnetic, electric, piezoelectric,
or thermal actuation methods, generally have low maximum
pressures with a couple of exceptions.13 In addition, these
valves generally have a relatively complicated structure and
require multi-step fabrication including multi-layer align-
ment, bonding and assembly, resulting in the difficult inte-
gration with an overall microfluidic reaction platform. Valves
based on mobile free-standing polymer elements have been
demonstrated to withstand pressures up to 4500 psi,14 how-
ever, the additional steps of monomer filling and patterned
in situ polymerization complicate the device fabrication. In
contrast, phase-change microvalves comprise a very straight-
forward and convenient alternative approach, relying on
reversible solidification of liquid in the channel itself to form
a solid plug to obstruct the fluid path. Impressive reports
demonstrate the capability of to withstand up to 1450 psi,
without the need for complex fabrication or moving parts.
Since the first report,15 a variety of materials including
water,16 paraffin,17 hydrogel,18 polymer monolith19 and metal
alloys20 have been chosen as the actuation liquid in these
valves, with phase transitions triggered thermally, chemically
or optically. Thermally-actuated valves based on water or par-
affin have been used in biochemistry for performing polymer-
ase chain reaction (PCR)16,17 and those with metal alloys have
been used for gas sampling applications.20 However, limited
by material incompatibility, none of them would be suitable
for radiochemistry or more general chemical reactions. For
example, paraffin and many polymers are soluble in com-
monly used organic solvents, which could lead to contamina-
tion of the reaction mixture and instability of the valves; water
and hydrogel materials present significant concerns for water-
sensitive reagents and reactions; and metals are likely to be
problematic in reactions involving strong acids and bases.
Using a new phase-change material, dimethyl sulfoxide
(DMSO), we extend the application of phase-change valves to
high-pressure organic chemistry in batch microreactors to
address the limitation of current phase-change valves. In the
field of radiochemistry, DMSO is a commonly used solvent
and is thus intrinsically favorable as a phase-change material.
The solvent compatibility allows direct translation of a known
chemical reaction onto a microfluidic platform, without the
need to change the chemistry to avoid the risk obtaining sub-
optimal yield due to contamination. Physically, DMSO has a
relatively high freezing point (18.9 °C) making it easier to
freeze than water and many other organic solvents. As a
demonstration of this microreaction platform, we performed
the high-temperature and pressure radiofluorination step in
the synthesis of the molecular probe [18F]FAC. Our new platform,
based on a simple architecture, broadens the range of reac-
tion conditions in batch microreactors, and enables increased
diversity of microscale batch chemistry applications.
Cole-Parmer). The central portion of the capillary (43 cm long
and 31 μL in volume), designed to contain the reactant mix-
ture, was coiled and sandwiched between two heating blocks
(each 5.0 × 5.0 × 1.2 cm); and the distal portions (each 4.4 cm
long and 3.2 μL in volume), designed to contain slugs of the
phase-change liquid DMSO (>99%, from Sigma-Aldrich,
colored by Nile Red (Sigma-Aldrich) for clear visualization),
were coiled and mounted on the cold side of a thermoelectric
Peltier cooler (CP854388, CUI Inc.). Coiling turns were packed
as flat as possible and held in place by Kapton® tape (Kapton
tape Inc.) with thermal paste to ensure uniform heating along
the length of the capillary. To improve cooling, the hot side of
the Peltier was affixed to an aluminum block with an internal
flow path connected to a recirculating liquid-cooling system
(Freezone Elite FZ-1003, CoolIT Systems). It took ~1 min after
activation of the Peltier and cooling system to cool the top sur-
face of the Peltier from room temperature down to a steady
state value of −19 °C. This temperature is sufficient to rapidly
solidify the two DMSO slugs, to seal the contents in the cen-
tral region of the capillary. To provide heating to activate
chemical reactions in this central region, each heating block
was equipped with a heater (C1E13, Waltlow), thermocouple
(K type, CHAL-002, Omega) and dedicated temperature con-
troller (CN7523, Omega). The entire experimental setup was
operated inside a lead hot cell to provide radiation shielding.
The heating and cooling blocks were positioned 1.5 cm
apart, with the volume of gradient region between heating
Fig. 1 (A) Simplified schematic of batch microreactor. (B) Schematic
showing actual capillary configuration on the heating and cooling
blocks. (C) Top view and (D) side view photographs of the reaction
platform. (E) Estimated temperature profile along the capillary under
the assumption of one-dimensional linear heat transfer model.
Experimental
The batch reactor (Fig. 1) was implemented with capillary
tubing (polytetrafluoroethylene (PTFE), OD: 0.030′′, ID: 0.012′′,
This journal is © The Royal Society of Chemistry 2014
Lab Chip, 2014, 14, 280–285 | 281

View Article Online
Technical Innovation
Lab on a Chip
and cooling blocks as ~2.2 μL. On one side, the gradient
region needs to be long enough so that DMSO freezing was
not compromised when the capillary contained in the reac-
tion region is heated. On the other side, because the liquid
inside the capillary in the gradient regions is not uniformly
heated, it is desirable to minimize the ratio of this volume to
the volume in the reaction region. In our gradient region
design, a trade-off was made between these two consider-
ations: the volume of the gradient region accounts for ~7%
of the total reaction volume whose effect on the reaction per-
formance is discussion in the next section, and the effect of
heat convection on valve performance is included in ESI.†
The volume in the reaction region could easily be scaled up
or down by changing the capillary length (i.e. number of
coiling turns) for other potential applications.
The steps to perform a reaction inside the microreactor
are illustrated in Fig. 2. Slugs of DMSO (3.2 μL), reactant
mixture (31 μL), and DMSO (3.2 μL) are aspirated into the
capillary by a computer-controlled syringe pump (PSD/4,
Hamilton). To minimize cross-contamination between DMSO
slugs and the reactant mixture, small air gaps (0.3 cm, 0.2 μL)
were introduced between these liquids. The reaction mixture
is then heated to the desired temperature for the desired
period of time. Upon completion, the heater was turned off
and the capillary was lifted away from the heater, allowing the
reaction chamber to cool to room temperature. Next, the
cooler was deactivated to melt the DMSO and open the valves.
Finally, the product is sequentially collected from the capil-
lary into three collection vials (separate vials for DMSO slugs
and reaction mixture) for subsequent analysis.
As a demonstration of a reaction involving high pressure,
the radiofluorination step of the synthesis of [18F]FAC (Fig. 3)
was performed. High temperature and pressure are essential
for obtaining high yield. First,
[
18F]fluoride (5 μL of
cyclotron-bombarded [18O]H2O containing 3–4 mCi) in a solu-
tion of K2CO3 (0.71 mg mL−1) and K2.2.2 (7.14 mg mL−1) in a
1.4 mL mixture of acetonitrile (MeCN) and H2O (0.4 : 1 v/v)
was evaporated to dryness at 110 °C in a 5 mL vial. This was
followed by three cycles of azeotropic distillation, each using
0.5 mL MeCN, evaporated at 110 °C, to form an anhydrous
residue of the [18F]KF/K2.2.2 complex. K2CO3, K2.2.2, and anhy-
drous MeCN were purchased from Sigma Aldrich and used
as received. After cooling the complex to room temperature,
350 μL of 2-O-(trifluoromethylsulfonyl)-1,3,5-tri-O-benzoyl-a-D-
ribofuranose (ABX Advanced Biochemical Compounds) in
MeCN (14.3 mg mL−1) was added and mixed by bubbling
nitrogen. The radiofluorination reaction was then performed
in the batch microreactor as described above, under the reac-
tion condition of 165 °C for 15 min. Following the reaction,
the DMSO slugs and reactant mixture were separately col-
lected. Samples were analyzed by radio-thin-layer chromatog-
raphy (radio-TLC) in a mobile phase of 95 : 5 v/v MeCN/H2O
and the conversion efficiency calculated from the integrated
area under two peaks in the resulting chromatograms (obtained
from mini-GITA radio-TLC scanner, Raytest). One peak rep-
resented the unreacted [18F]fluoride ion (retention factor
Rf = 0.0), and the second corresponded to the desired fluori-
nated product (Rf = 0.97). The radioactivity in each vial was
measured using a calibrated dose calibrator (CRC-25PET,
Capintec). The initial amount of radioactivity loaded into
the capillary was calculated as the difference in radioactivity
Fig. 2 Schematic of batch microreactor operation. (A) DMSO, air, and
reagent slugs are aspirated by the syringe pump; (B) slugs are moved
to designated locations in the tubing; (C) cooler is turned on to close
valves, i.e., solidify DMSO (purple); (D) heater is activated to perform
reaction; (E) heater is deactivated to cool reaction mixture and reduce
pressure; (F) cooler is turned off to open DMSO valves (pink); (G) the
samples are pumped out into collection vials for subsequent analysis.
Fig. 3
[
18F]FAC synthesis scheme. The radiofluorination step (dashed
box) from precursor 1 to the fluorinated sugar 2 was chosen as a
demonstration reaction.
282 | Lab Chip, 2014, 14, 280–285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Lab on a Chip
Technical Innovation
of the premixed reagent vial before and after loading.
The total radioactivity after the reaction is the sum of radio-
activity of samples in all collection vials. All radioactivity
measurements were decay-corrected to the start time of the
experiment. As a control experiment, the same procedure
was followed except that the cooler was not activated to
freeze the DMSO. The capillary was disconnected from
syringe pump after loading and its two ends were inserted
into two empty collection vials. Samples collected were mea-
sured for radioactivity and their composition analyzed by
radio-TLC.
comparable to typical results attained using a macroscale
radiosynthesizer.10 A representative radio-TLC chromatogram
is shown in Fig. 4. Taking into account the small volume
fraction (~7%) contained within the temperature gradient
region, the conversion efficiency is expected to be slightly
higher if the microreactor design was further optimized, or if
it was implemented in a microfluidic chip with tiny thermo-
electric junctions16 where the fraction of gradient region is
reduced. The high conversion efficiency also suggests that
the residue DMSO left behind into the reactant mixture
during sample loading and collection, does not adversely affect
the reaction, and/or the level of this contamination is low.
The observation that only 3 2% (n = 3) of the total recovered
radioactivity present in the collection vials containing the
DMSO slugs, verifies the overall cross-contamination at a low
level. Besides the contamination occurring during sample load-
ing and collection and a possible trace amount of reaction
mixture redistributed through the air gaps by evaporation/
condensation during the reaction, the main contamination
comes from the imprecise nature of the manual operations
involved in DMSO–crude product–DMSO sequential sample
collection. An automated sample collection system employing
liquid sensors could improve the accuracy and reproducibility
of sample collection. In future applications that are more sen-
sitive to cross-contamination, it may also be possible to place
intervening “washing” slugs between the reactant mixture
and the DMSO slugs.21
To confirm the need for high-pressure valves in the batch
reactor for our particular model reaction, the reaction was
repeated without first freezing the DMSO valves. In three
separate attempts, it was observed during heating that the reac-
tion mixture “burst” and pushed all slugs, into the collection
vials at the two ends of the capillary. Radioactivity measure-
ment and radio-TLC showed 98 2% (n = 3) of the radioac-
tivity initially loaded into the capillary was lost into these
collection vials and conversion efficiency was 0% (n = 3).
These results suggest the importance of properly closed
DMSO valves to contain the high pressure generated during
the batch reaction process. In the multi-step synthesis of PET
probes, the radiofluorination step is often the most difficult,
requiring the highest temperatures and pressures. One could
envision performing additional reaction steps in this batch
Results and discussion
Our DMSO phase-change microvalves have been character-
ized in terms of maximum withstanding pressure and
response time. Based on our investigation of the effects of
slug length, capillary diameter, and curvature of the capillary
“turns”, on valve withstanding pressure (Fig. S1 and ESI†),
a 0.012′′ ID capillary with DMSO slug of length 4.4 cm was
chosen for our demonstration reaction. The maximum with-
standing pressure of DMSO valves with this particular geome-
try is characterized to be at least 450 psig, governed by the
limit of our pressure source. This pressure is significantly
higher than 121 psi – the anticipated pressure during the
reaction estimated according to the vapor pressure of the
reaction solvent MeCN at the reaction temperature 165 °C.
Freezing of the DMSO slugs in the capillary was found to
occur within 2 s of contacting the capillary with the cooling
block pre-cooled to −19 °C. The DMSO colored by Nile Red
turned from pink to dark purple as the phase changed from
liquid to solid. The valve re-opened in 25 s after the Peltier
was deactivated. These response times are more than ade-
quate for most chemical applications, including the synthesis
of short-lived radioisotopes which require fast operations.
With optimization of the geometry and the heating and
cooling systems, response times could likely be decreased
further. Response times down to 80 ms have been reported
in other configurations.16 As in the reaction the batch sealed
by two DMSO valves is heated (Fig. 1), studies to examine the
effect on valve stability of heat conduction from heaters to
DMSO slugs is included in ESI.† The DMSO slugs remained
frozen for 15 min and no leakage was observed, when the
temperature of the heater region was elevated and kept at up
to 200 °C, ensuring that the valve performance is more than
sufficient for the reaction at 165 °C.
As a demonstration of the utility of a batch microreactor
capable of withstanding high pressures, the radiofluorination
step in the synthesis of [18F]FAC, a new PET probe for
immune system monitoring and prediction of chemotherapy
response,4 was performed. The percentage of radioactivity
recovered after each experiment was nearly ideal (97 1%,
n = 3), consistent with the observation that no leaking was
visible during the experiments. Radio-TLC analysis of the
collected samples showed the conversion of [18F]fluoride ion
to the fluorinated product to be 83 1% (n = 3), which is
Fig.
4 Representative radio-TLC analysis of crude product of
radiofluorination reaction demonstrated in capillary reactor with
phase-change valves.
This journal is © The Royal Society of Chemistry 2014
Lab Chip, 2014, 14, 280–285 | 283

View Article Online
Technical Innovation
Lab on a Chip
microreaction platform, and integrating with microscale tech-
nologies for upstream fluoride drying and postreaction purifi-
cation for a fully-automated platform to synthesize 18F-labeled
molecular probes such as [18F]FAC and its analogs.
Acknowledgements
We thank the Dr. Saman Sadeghi and the UCLA Biomedical
Cyclotron facility for generously providing [18F]fluoride for
this study, Dr. David Stout and Jeffrey Collins of the Crump
Preclinical Imaging Center for assistance with [18F]fluoride
drying, and Huijiang Ding, Darin Williams and Dirk Williams
for assistance with the experimental setup. This work was
supported in part by the Department of Energy, Office of
Biological and Environmental Research (DE-SC0001249).
This platform could also be suitable for high-throughput
screening or combinatorial chemistry applications where
high temperatures and pressures are needed. Though we have
shown a single microreactor with DMSO phase-change valves
implemented in a section of a capillary, this concept could
readily be extended to larger numbers of reactors and phase-
change slugs by leveraging significant advances in spatially-
defined slug-based systems for high-throughput screening.22
Despite our implementation of the microreaction platform
in a capillary for simplicity, with substantially more effort,
we see no technical hurdles in extending this work into a
microfluidic chip, as several previous efforts have successfully
implemented phase-change valves in a microchip format.16
Because phase-change valves themselves require no fabrica-
tion, tremendous flexibility of chip material and geometry
can be considered to meet the needs of different applications.
With the possibility of multiple individually-addressable
phase-change valves on a single chip,16 sophisticated liquid
handling functionality could be integrated on-chip, simplify-
ing the preparation and collection of slugs and reducing
the size and cost of the overall microreaction platform.
Implementation in a microfluidic chip may also enabled
geometries with improved heating and mixing efficiency. It
should be appreciated, however, that batch-type microreactors
have inherently inferior mixing and thermal performance
compared to continuous flow and continuous droplet
microreactors and thus may be less suitable for certain types
of very fast reactions.
References
1 J. P. McMullen and K. F. Jensen, Annu. Rev. Anal. Chem.,
2010, 3, 19–42.
2 G. Pascali, P. Watts and P. A. Salvadori, Nucl. Med. Biol.,
2013, 40, 776–787.
3 C.-C. Lee, G. Sui, A. Elizarov, C. J. Shu, Y.-S. Shin,
A. N. Dooley, J. Huang, A. Daridon, P. Wyatt, D. Stout,
H. C. Kolb, O. N. Witte, N. Satyamurthy, J. R. Heath,
M. E. Phelps, S. R. Quake and H.-R. Tseng, Science, 2005,
310, 1793–1796.
4 A. M. Elizarov, R. M. van Dam, Y. S. Shin, H. C. Kolb,
H. C. Padgett, D. Stout, J. Shu, J. Huang, A. Daridon and
J. R. Heath, J. Nucl. Med., 2010, 51, 282–287.
5 P. Y. Keng, S. Chen, H. Ding, S. Sadeghi, G. J. Shah,
A. Dooraghi, M. E. Phelps, N. Satyamurthy, A. F. Chatziioannou,
C.-J. ‘CJ’ Kim and R. M. van Dam, Proc. Natl. Acad. Sci. U. S. A.,
2012, 109, 690–695.
6 M. J. Jebrail, A. H. C. Ng, V. Rai, R. Hili, A. K. Yudin
and A. R. Wheeler, Angew. Chem., Int. Ed., 2010, 49,
8625–8629.
7 A. Lebedev, R. Miraghaie, K. Kotta, C. E. Ball, J. Zhang,
M. S. Buchsbaum, H. C. Kolb and A. Elizarov, Lab Chip,
2012, 13, 136–145.
Conclusions
We have developed a batch microreactor using phase-change
valves for high temperature and pressure chemical reactions.
With the use of DMSO as a new phase-change material, this
platform extends the application of phase-change valves from
applications in biology to applications in organic synthesis
involving harsh conditions or moisture-sensitive reagents.
The ability for this simple microreactor to hold high pressures
enables batch reactions to be performed under a wider range
of conditions (temperatures, pressures, reagent diversity) than
possible in previous batch devices.
As an example reaction, we successfully demonstrated the
radiofluorination step in the multi-step synthesis of [18F]FAC.
The synthesis was performed with high conversion efficiency
comparable to what can be typically achieved under sealed,
pressurized conditions at the macroscale. This platform
would be suitable for performing many other batch chemical
reactions at the microscale, especially those that involve
high pressures due to high vapor pressure under super-
heated conditions, or due to production of gaseous products
during reactions.
8 R. E. Laing, M. A. Walter, D. O. Campbell, H. R. Herschman,
N. Satyamurthy, M. E. Phelps, J. Czernin, O. N. Witte and
C. G. Radu, Proc. Natl. Acad. Sci. U. S. A., 2009, 106,
2847–2852.
9 T. J. Mangner, R. W. Klecker, L. Anderson and A. F. Shields,
Nucl. Med. Biol., 2003, 30, 215–224.
10 M. Lazari, K. M. Quinn, S. B. Claggett, J. Collins, G. J. Shah,
H. E. Herman, B. Maraglia, M. E. Phelps, M. D. Moore and
R. M. van Dam, EJNMMI Res., 2013, 3, 52.
11 F. Benito-López, R. J. M. Egberink, D. N. Reinhoudt and
W. Verboom, Tetrahedron, 2008, 64, 10023–10040.
12 W. C. Nelson, M. Yen, P. Y. Keng, M. R. van Dam, and
C.-J. Kim, in Proc. Int. Conf. Solid State Sensors, Actuators
and Microsystem (Transducer), Beijing, China, 2011.
13 E.-H. Yang, C. Lee, J. Mueller and T. George,
J. Microelectromech. Syst., 2004, 13, 799–807.
14 Hasselbrink, T. J. Shepodd and J. E. Rehm, Anal. Chem.,
2002, 74, 4913–4918.
15 C. D. Bevan and I. M. Mutton, Anal. Chem., 1995, 67,
1470–1473.
284 | Lab Chip, 2014, 14, 280–285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Lab on a Chip
Technical Innovation
16 J. Anderson and R. P. Welle, JALA, 2008, 13, 65–72.
17 R. Pal, M. Yang, B. N. Johnson, D. T. Burke and M. A. Burns,
Anal. Chem., 2004, 76, 3740–3748.
18 D. J. Beebe, J. S. Moore, J. M. Bauer, Q. Yu, R. H. Liu,
C. Devadoss and B.-H. Jo, Nature, 2000, 404, 588–590.
19 G. Chen, F. Svec and D. R. Knapp, Lab Chip, 2008, 8,
1198–1204.
20 R. P. Manginell, M. W. Moorman, J. A. Rejent, P. T. Vianco,
M. J. Grazier, B. D. Wroblewski, C. D. Mowry and
K. E. Achyuthan, Rev. Sci. Instrum., 2012, 83, 031301.
21 V. Linder, S. K. Sia and G. M. Whitesides, Anal. Chem., 2004,
77, 64–71.
22 R. R. Pompano, W. Liu, W. Du and R. F. Ismagilov, Annu.
Rev. Anal. Chem., 2011, 4, 59–81.
This journal is © The Royal Society of Chemistry 2014
Lab Chip, 2014, 14, 280–285 | 285