Organic synthesis in soft wall-free microreactors: Real-time monitoring of fluorogenic reactions

Article abstract of DOI:10.1021/ac800855u

A new approach of "laboratory on a chip" (LOC) devoted to organic synthesis based on electrically operated ionic liquid microdroplets used as air-stable "soft" microreactors was recently introduced. A number of challenging issues have yet to be addressed

Full text of DOI:10.1021/ac800855u

Anal. Chem. 2008, 80, 6051–6055
Organic Synthesis in Soft Wall-Free Microreactors:
Real-Time Monitoring of Fluorogenic Reactions
G. Marchand,*,† P. Dubois,†,‡,§ C. Delattre,† F. Vinet,† M. Blanchard-Desce,‡,§ and M. Vaultier‡,§
Department of Technology for Biology and Health, CEA/LETI-Minatec, 17 rue des Martyrs 38054 Grenoble, France,
Molecular Chemistry and Molecular Photonics, UMR 6510, CNRS and Universite´ de Rennes1, 263 avenue Ge´ne´ral
Leclerc, 35042 Rennes, France
A new approach of “laboratory on a chip” (LOC) devoted
to organic synthesis based on electrically operated ionic
liquid microdroplets used as air-stable “soft” microreac-
tors was recently introduced. A number of challenging
issues have yet to be addressed to allow standard mac-
roscale organic chemistry to be directly transposed in
these nonclassical microreactors. In particular, since
standard (i.e., magnetical or mechanical) stirring methods
are prohibited in such wall-free microreactors, effective
alternatives have to be developed to circumvent mass-
transfer limitations. With this aim in mind, a fluorogenic
version of a click chemistry reaction was developed to
evaluate the efficiency of alternative mixing methods on
the reaction kinetics. We demonstrate that the combina-
tion of chaotic advection created by surface acoustic waves
combined with a temperature increase (Marangoni effect)
leads to the same kinetics regime as in standard macros-
cale conditions. This opens the route for application of
the new generation of LOC to efficient organic synthesis
in microscale.
electrostatic field pressure applied on the droplet6–9), thus avoiding
complex fluidic connections and void volumes as observed with
standard (i.e., channel) microfluidics10,11 and offering wider possibili-
ties of combinations than linear channels. Hence microdroplets made
of RTILs can be used as “soft” electrically operated microflasks for
reagents,1,12,13 thus allowing chemical reactions to be conducted with
such systems. This opens a new promising technological approach
for microscale organic synthesis. As a proof of concept, a three-
component reaction was carried out in RTIL droplets leading to the
expected product after 2 h.1
As for other microsystems, the real-time in situ monitoring of
the reaction kinetics in such microsystems raises a number of
questions. Contrary to standard (i.e., channel) microfluidics, the wall-
free microreactors allow for in situ reaction medium sampling, though
with inherent experimental uncertainty associated with the sampling
of small volumes.2 In contrast, fluorescence is a noninvasive and
quantitative analytic technique, which offers interesting prospects for
real-time monitoring of reactions within single microreactors as well
as parallel monitoring of many discrete microdroplets by scanning.
Such methodology would be of particular interest for various issues
including combinatorial synthesis of fluorophores14 as well as
optimization of reaction conditions and protocols. In particular,
monitoring and optimization of reagents mixing processes is a critical
issue for microfluidics. In particular, since standard (i.e., magnetical
or mechanical) stirring methods are prohibited in such wall-free
microreactors, effective alternatives have to be developed to circum-
vent mass-transfer limitations. In this paper, we demonstrate that
fluorescence can indeed be used for real-time monitoring of a
chemical reaction within single microdroplets, allowing testing and
defining experimental conditions for improved mixing efficiency. To
achieve this goal, the strategy has been to make use of a fluorogenic
Laboratory on a chip (LOC) systems dedicated to biotechnologi-
cal and chemical applications allow the handling of very small
quantities of reagents for optimization of reactions while usually
producing the desired products faster and in higher yields and purity.
Recently, we proposed a new approach of LOC devoted to organic
synthesis based on the combination of tiny droplets of Room
Temperature Ionic Liquids (RTILs) used as air-stable, “soft” microre-
actors and electrowetting on dielectric (EWOD)1,2 as the fluidic
motor. This new generation of LOC systems presents several
advantages due to the specific features of the two building blocks:
(i) ionic liquids exhibit a negligible vapor pressure, which ensures
suitable microreactors “bench” lifetime thanks to the lack of evapora-
tion and have been successfully used in many fields, especially in
green chemistry and organic synthesis;3–5 (ii) the EWOD phenom-
enon allows one to manipulate (move, split, mix) single liquid
microdroplets on a two-dimensional electrodes matrix, (by using the
(3) Wasserscheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39, 3772
(4) Wasserscheid, P.; Welton, T. Ionic Liquids in Synthesis;Wiley-VCH: Wein-
heim, 2007
(5) Rogers, R. D.; Seddon, K. R. Science 2003, 302, 792
(6) Pollack, M. G.; Shenderov, A. D.; Fair, R. B. Lab Chip 2002, 2, 96
(7) Cho, S.-K.; Moon, H.; Kim, C.-J. JMEMS 2003, 12, 70
(8) Mugele, F.; Baret, J. C. J. Phys. Condens. Matter 2005, 17, 705
(9) Fouillet, Y.; Jary, D.; Chabrol, C.; Claustre, P.; Peponnet, C. Microfluidics
Nanofluidics 2008, 4, 159
(10) Kikutani, Y.; Hisamoto, H.; Tokeshi, M.; Kitamori, T. Lab Chip 2004, 4,
328
(11) Doku, G. N.; Verboom, W.; Reinhoudt, D. N.; van.den Berg, A. Tetrahedron
2005, 61, 2733
(12) Millefiorini, S.; Traczyck, A. H.; Sedev, R.; Efthimiadis, J.; Ralston, J. J. Am.
Chem. Soc. 2006, 128, 3098
(13) Chatterjje, D.; Hetayothin, B.; Wheeler, A. R.; King, D. J.; Garrell, R. L.
Lab Chip 2006, 6, 199
(14) Finney, N. S. Curr. Opin. Chem. Biol. 2006, 10, 238
.
.
.
.
.
.
.
* To whom correspondence should be adressed. E-mail: gilles.marchand@
cea.fr.
.
† CEA/LETI-Minatec.
‡ Universite´ de Rennes1.
.
§ CNRS, UMR 6510, Universite´ de Rennes1.
(1) Dubois, P.; Marchand, G.; Fouillet, Y.; Berthier, J.; Douki, T.; Hassine, F.;
.
Gmouh, S.; Vaultier, M. Anal. Chem. 2006, 78, 4909
(2) Dubois, P.; Marchand, G.; Gmouh, S.; Vaultier, M. Chem. Eur. J. 2007,
13, 5642
.
.
.
.
10.1021/ac800855u CCC: $40.75 2008 American Chemical Society
Published on Web 07/02/2008
Analytical Chemistry, Vol. 80, No. 15, August 1, 2008 6051

standard, J values in hertz; 13C chemical shifts relative to the central
peak of CDCl3 at 77.0 ppm. High-resolution mass spectra measure-
ments and elemental analyses were performed at the Centre Re´gional
de Mesures Physiques de l’Ouest (C.R.M.P.O., Rennes).
SynthesisN-Trimethyl-N-butylammonium bis(trifluorom-
ethylsulfonyl)imide [tmba][NTf2].
1-Chlorobutane (6.5 mL, 63.1 mmol) and trimethylamine (16.5
mL, 126.2 mmol, 45 wt % aqueous solution) were dissolved in
acetonitrile (10 mL) in a sealed flask. The flask was heated to 70
°C overnight. After elimination of volatiles and solvents under
vacuum, the residue was washed with diethyl ether and recrystal-
lized from acetone, leading to a white solid. This intermediary
salt (9 g, 59.3 mmol) was dissolved in the minimum of distilled
water and 1.1 equiv of lithium bis(trifluoromethylsulfonyl)imide
(25 g, 87.4 mmol) added. After 2 h of stirring at room temperature,
the [tmba][NTf2] was extracted with methylene chloride and the
organic phase dried over Na2SO4. Solvent was removed under
vacuum giving a water-like free flowing liquid, which was dried
under high vacuum at 60 °C for 3 h. Yield: 85%. 1H NMR (300
MHz, acetone-d6) δ: 0.95 (t, 3H); 1.40-1.59 (m, 2H); 1.88-2.11
(m, 2H); 3.42 (s, 9H); 3.60-3.75 (m, 2H). 13C NMR (75 MHz,
acetone-d6) δ: 14.1; 20.6; 25.8; 53.9 (t, JC-N ) 4.1 Hz); 67.7; 121.8
(q, JCF ) 321.0 Hz). 19F NMR (376 MHz, acetone-d6) δ: -80.3.
HRMS (FAB) for (2C+, NTf2-)+: calcd 512.2051, found 512.2069
Synthesis of 1-Azido-4(trifluoromethanesufonyl)benzene 1. 4-(Tri-
fluoromethylsulfonyl)aniline (2.8 g, 11 mmol), water (2 mL), and
2 M HCl (15 mL) were added to a sealed flask. After cooling at 0
°C, 1.0 g of sodium nitrite (14.5 mmol) was added to the solution,
incubated for 10 min, and 1.65 g of urea (2.75 mmol) was added.
After 1 h at 0 °C, an aqueous solution of sodium azide (1.4 g, 21.5
mmol) and potassium acetate (3.2 g, 32.60 mmol) in water (10
mL) was slowly added at 0 °C. After 2 h at 0 °C, the organic phase
was separated, dried over Na2SO4, and filtered, and the solvent
was removed under vacuum. The obtained oil was purified by
chromatography on silica gel (80% heptane/20% ethyl acetate).
Yield: 82%.1H NMR (300 MHz, CDCl3) δ: 7.27 (d, 2H, 3J ) 8,5
Hz); 8.01 (d, 2H, 3J ) 8.5 Hz). 13C NMR (75 MHz, CDCl3) δ: 120.19
Figure 1. Principle of the real-time monitoring of a fluorogenic
Huisgen reaction in ionic liquids microreactors handled by EWOD.
The two droplets, containing respectively compounds A and B, were
merged, producing fluorescence whose intensity is directly correlated
to the generation of final product C.
reaction. Recently, authors reported15–17 efficient synthesizes of
fluorescent 1,2,3-triazoles by copper-catalyzed 1-3 dipolar cycload-
dition from reaction of azides with terminal acetylenes.18–20 This well-
known so-called click chemistry21 reaction leads to 1,4-disubstitued
1,2,3-triazoles without byproduct and with excellent yields in various
organic solvents, including ionic liquids.22 In this paper, such a
reaction, which can produce fluorescent probes starting from non-
fluorescent reagents, was thus implemented in ionic liquid micro-
droplets (Figure 1). This was meant to allow for simple real-time
monitoring of the reaction by following the fluorescence signal of
the dropletswhich is directly correlated to the generation of the
productswithout requiring a complex set of optical filters or acquisi-
tions at different wavelengths to eliminate the contribution of other
species to the fluorescence signal. As a first step, the fluorescent
monitoring of the formation of 1,4-disubstituted 1,2,3-triazole, 4, at
both the macroscale (flask) and microscale (ionic liquid microdrop-
lets) was demonstrated. With this model reaction in hand, the
efficiency of different mixing methods and their influence on the
reaction kinetic rates in soft IL microreactors were compared and
evaluated. As a result, suitable mixing conditions are proposed as a
general protocol for common organic reactions in IL microreactors.
1
(q, JC-F ) 352.1 Hz); 120.7; 126.9; 133.2; 149.8.
Synthesis of N-(5-Diethylamino-pentyl)-N-trimethylammonium
Bistrifluoromethylsulfonimide 3. Trimethylamine gas generated by
heating at 40 °C a water solution (41 mL, 312.1 mmol, 45 wt %
aqueous solution) was condensed, after passing through KOH
pellets, in an anhydrous THF dibromopentane (48.1 mL, 200
mmol) solution (100 mL). After 8 h, the ammonium salt was
filtered, washed with ether (3 × 25 mL), and dried under high
vacuum. This solid was then dissolved in the minimum of distilled
water and 1.1 equiv of lithium bis(trifluoromethylsulfonyl)imide
added. After 2 h of stirring at room temperature, the solution was
extracted with methylene chloride and the organic phase dried
over Na2SO4. Solvent was removed under vacuum giving a water-
like free flowing liquid, which was dried under high vacuum at
60 °C for 3 h. The obtained liquid (36.6 mmol, 17.9 g), ethanol
(40 mL), and diethylamine (183.6 mmol, 19 mL of a 40% aqueous
solution) were added to a sealed flask. The mixture was heated
at 78 °C overnight. After cooling and addition of 146.4 mmol (20.2
g) of solid potassium carbonate, the solution was heated to 78 °C
for 4 h. Solvents and excess diethylamine were removed under
vacuum. After addition of 50 mL of CH2Cl2, filtration to remove
EXPERIMENTAL SECTION
Reagents and Characterization. Ethynylanisole 2 and
Cu(CH3CN)4PF6are commercially available from Sigma-Aldrich and
were used without further purification. Fluorescent beads (Fluores-
cent Yellow Green CML latex, reference 2-FY-10000) are com-
mercially available from Interfacial Dynamics Corp. NMR: 1H
chemical shifts (δ) are given in ppm relative to TMS as internal
(15) Parent, M.; Mongin, O.; Kamada, K.; Katan, C.; Blanchard-Desce, M. Chem.
Commun. 2005, 2029
(16) Zhou, Z.; Fahrni, C. J. J. Am. Chem. Soc. 2004, 126, 8862
(17) Sivakumar, K.; Xie, F.; Cash, B. M.; Long, S.; Barnhill, H. N.; Wang, Q.
Org. Lett. 2004, 6, 4603
.
.
.
(18) Tornøe, C. W.; Meldal. M. Peptidotriazoles Peptides 2001; Proc. Am. Pept.
Symp. Lebl, M., Houghten, R. A. , Eds.; American Peptide Society and
Kluwer Academic Publishers: New York, 2001; pp 263-264
(19) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057
(20) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem.,
Int. Ed. 2002, 41, 2596
(21) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001,
40, 2004
.
.
.
.
(22) Zhao, Y. B.; Yan, Z. Y.; Liang, Y. Mesult. Tetrahedron Lett. 2006, 47, 1545
.
6052 Analytical Chemistry, Vol. 80, No. 15, August 1, 2008

Fluorescence quantum yield was measured using standard meth-
ods on air-equilibrated samples at room temperature. Quinine
bisulfate in 0.05 M H2SO4 (Φ ) 0.546) was used as a reference.25
The reported fluorescence quantum yields are within ±10%.
RESULTS AND DISCUSSION
The 1,2,3-triazole 4 was obtained by a 1,3-dipolar cycloaddition,
catalyzed by a copper (I) salt, i.e., Cu(CH3CN)4PF6, between azide
1 and the alcyne 2 in [tmba][NTf2] (N-butyl-N-trimethylammo-
nium bistrifluoromethylsulfonylimide) in the presence of the
supported amine 3 used as a base (Scheme 1).
The macroscale reaction (0.16 mmol, 800 µL of a magnetically
stirred solution at 400 rpm) showed a 100% conversion rate after
14 min at 25 °C. Using different temperatures and initial concen-
trations in 1 (100-500 mM), 2 (100-500 mM), and CuI catalyst
(2-10 mM), the experimental data were correctly fitted over the
whole experimental range using eq 1 as the kinetics rate
expression:
Figure 2. Absorption spectrum (-) and fluorescence emission
spectrum ( · · ·) of chromophore 4 in [tmba][NTf2].
KBr and excess K2CO3, and washing with water to remove residual
K2CO3, active charcoal was added and the suspension stirred for
2 h at room temperature and then filtered on a Celite pad. Solvent
was then removed under vacuum leading to an oily product. Yield:
60%. 1H NMR (300 MHz, acetone-d6) δ: 1.00 (t, 6H, 3J ) 7.0 Hz);
1.39-1.63 (m, 4H); 1.89-2.16 (m, 2H); 2.38-2.54 (m, 6H); 3.38
(s, 9H); 3.54-3.66 (m, 2H). 13C NMR (75 MHz, acetone-d6) δ:
12.3; 23.8; 25.1; 27.5; 48.1; 53.5; 54.0 (t, 1JC-N) 4.0 Hz); 67.9; 124.6
ν ) k0e(-Ea ⁄RT)[1]n[2]m[Cu+]p
(1)
where v is the reaction rate, k0 the pre-exponential factor, Ea the
activation energy, n the partial order in reagent 1, m the partial
order in 2, and p the partial order in Cu+.
1
(q, JC-F ) 319.9 Hz). FAB HRMS for C12H29N2 (M+.): calcd
Note that the partial orders are averaged over the whole range
of reagent concentrations and thus cannot represent a chemical
mechanism. However, the values of the partial orders are
consistent with previous work, as they fall between the maximal
and minimal values obtained by Fediorov et al.26 in using initial-
rate kinetic measurements with different reagents concentrations.
The best fit was obtained for the following values: k0 ) 1.76
201.2331, found 201.233.
Synthesis of 1-Trifluoromethanesulfonyl-4-(4-p-methoxyphenyl-1H-
1,2,3-triazol-1-yl)benzene 4 in Droplet. Solution 1: 1 (0.2 mol/L)
and Cu(CH3CN)4PF6 (0.01 mol/L) are dissolved in [tmba][NTf2].
Solution 2: 2 (1 mol/L) and 3 (0.03 mol/L) are dissolved in
[tmba][NTf2]. The 0.2-µL droplets of solutions 1 and 2 were
deposited on the chip with an Eppendorf micropipet. The droplets
were displaced and merged with a tension of 55 V rms. The
actuation of the electrodes was switched off and the reaction
carried out.
L
0.5 · mol-0.5 · s-1, Ea ) 42 kJ · mol-1, n ) 0, m ) 0.6, and p ) 0.9.
These values indicate that (i) only 2 and Cu+ concentrations affect
the kinetic rate and (ii) the temperature influence on the kinetics
rate is well accounted for by the Arrhenius law with a high value
of Ea showing that the measured kinetics data were not biased
by the mass-transfer limitation. As shown in Figure 2, chro-
mophore 4 shows an absorption band in the near-UV region,
peaking at 318 nm (extinction coefficient εmax ) 7.2 103
The following data were obtained for pure isolated compound
4 (synthesized in standard macroscale conditions). 1H NMR (300
3
MHz, DMSO) δ: 3.83 (s, 3H); 7.11 (d, 2H, J ) 8.8 Hz); 7.90 (d,
3
2H, J ) 8.8 Hz); 8.44 (m, 4H); 9.50 (s, 1H). 13C NMR (75 MHz,
DMSO) δ: 55.2; 114.5; 119.0; 121.0; 122.1; 126.8; 127.0, 128.2; 133.1;
142.8; 147.9; 159.5. ESI+/HRMS for C16H12F3N3O2S (M+.): calcd
383.05515; found 383.0549. Elemental analysis calcd (%) for
C16H12F3N3O2S C 50.13, H 3.16, N 10.96, S 8.36; found C 50.37, H
3.65, N 10.66, S 8.28.
M
-1 · cm-1). Interestingly, chromophore 4 shows a broad emis-
sion band in the visible regionscharacteristic of a polarized
emittingexcitedstatewithstronginteractionswiththeenvironments
peaking at 530 nm (quantum yield Φ ) 0.15) in ionic liquid
Theoretical Calculations. Natural transitions orbitals23 of
chromophore 4 (Figure 2) were calculated at the TD-B3LYP/6-
31G//HF/6-31G level using the Gaussian 98 package.24 They hint
to a clear charge transfer from the electron-donating moiety
(methoxyphenyl) to the electron-withdrawing moiety (which
includes both the triazole and the trifluoromethanesulfonylphenyl
units) that occurs upon excitation.
Absorption and Emission. UV/vis spectra were recorded on
a Jasco V-570 double-beam spectrophotometer. Steady-state fluo-
rescence measurements were performed at room temperature
using an Edinburgh Instruments (FLS 920) spectrometer working
in photon-counting mode, equipped with a quantum counter for
excitation correction. Fully corrected emission spectra were
obtained at λex with A(λex) < 0.1 to minimize internal absorption.
(24) Gaussian 98, Revision A.11 and Gaussian 03, Revision D.02. Frisch, M. J.;
Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman,
J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N. ; Burant, J. C.; Millam,
J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara,
M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda,
Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.;
Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann,
R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.;
Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith,
T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; and Pople,
J. A.; Gaussian, Inc., Wallingford CT, 2004.
(25) Eaton, D. F. Pure Appl. Chem. 1988, 60, 1107
(26) Rodionov, V. O.; Fokin, V. V.; Finn, M. G. Angew. Chem., Int. Ed. 2005,
44, 2210
.
(23) Martin, R. L. J. Chem. Phys. 2003, 118, 4775
.
.
Analytical Chemistry, Vol. 80, No. 15, August 1, 2008 6053

Scheme 1. Fluorogenic 1,3 Dipolar Cycloaddition (Click) Reaction Used as a Model Reaction in Ionic
Liquid Solventa
a The task-specific ionic liquid 3 is used as the base.
Figure 3. Natural transitions orbitals of chromophore 4 calculated
at the TD-B3LYP/6-31G//HF/6-31G level using the Gaussian 98
package.
Figure 5. (A) Time evolution of the normalized fluorescence signal
generated in the microdroplet reactor by the fluorogenic reaction
without mixing activation. (B) Initial reaction rates at different tem-
peratures in the bulk (stirred solution in a flask) and in microdroplets
using three different mixing activation methods. The initial reaction
rates were determined by recording the fluorescence signal during
reaction and by measuring the slope at the origin in each condition.
Values of mass-transfer efficiency (ꢀ) are indicated.
Figure 4. (A) On-chip EWOD merging of microdroplets containing
reagents. Droplet 1 (0.2 µL) contains reagent 1 (0.2 mol/L; 10 µg)
and catalyst Cu(CH3CN)4PF6 (0.01 mol/L; 0.7 µg) while droplet 2 (0.2
µL) contains reagent 2 (1 mol/L; 26 µg) and base 3 (0.03 mol/L; 2.9
µg). (B) Real time monitoring of the formation of product 4 (microre-
actor top view).
After coalescence, the fluorescence signal of 4 generated by
the reaction was recorded (Figure 4B) and the fluorescent
intensities were normalized by the minimum and maximum levels
[tmba][NTf2] (Figure 2). The corresponding transition can be
related to a charge transfer between donor and acceptor units.
Indeed, transition orbitals calculations (shown in Figure 3) clearly
hint to a charge transfer from an electron-donating moiety
(methoxyphenyl) to an electron-withdrawing moiety of the mol-
ecule (which includes both the triazole and the trifluoromethane-
sulfonylphenyl units) upon excitation. The emission of product 4
in the visible region can thus be used for in situ real-time
monitoring of the model click reaction since none of the starting
materials are fluorescent.
In order to perform the reaction within a single droplet, a
microdroplet of [tmba][NTf2] containing reagent 1 and catalyst
Cu(CH3CN)4PF6 (droplet 1) was converged by EWOD at 55 V
toward a second droplet of the same ionic liquid containing
reagent 2 and base 3 (Figure 4A).
to describe the reaction kinetics (Figure 5A). The fluorescence
signal is observed to increase gradually in agreement with ongoing
formation of product 4. A 100% conversion rate is reached after
40 min (as confirmed by HPLC) in the case of a stationary
microdroplet. This slow reaction rate, as compared to the rate of
the macroscale reaction performed with standard magnetic stir-
ring, is due to the high viscosity of the RTIL used as solvent, i.e.,
[tmba][NTf2]27 resulting in slow diffusion of reagents within the
microreactor droplet and consequently in diffusion-limited reaction.
To circumvent this critical limitation, different mixing methods
were then investigated in order to enhance the molecules’ mobility
and therefore approach the kinetics rates obtained for macroscale
reactions when using standard magnetic stirring. The first mixing
(27) The [tmba][NTf2] viscosity with 2950 ppm water was measured to be 81
cP.
6054 Analytical Chemistry, Vol. 80, No. 15, August 1, 2008

method consists of keeping stationary the microdroplet obtained
after coalescence and heating it at 40 °C using a resistance kept
underneath the EWOD chip in order to create an additional
convectional flow and/or a Marangoni effect.28 Indeed, if the
substrate is warmer than the air, the liquid droplet sitting on the
substrate chip will experience a warmer region near the substrate
and a cooler region near the apex of the droplet. Thus, the surface
tension will be lower near the substrate and larger at the apex,
resulting in a tangential force on the interface. If sufficient, this
tangential force induces a circulation inside the droplet even if
the external shape of the droplet does not change. A preliminary
experiment with fluorescent beads (Fluorescent Yellow Green
CML latex with a 9.8-µm diameter, λexc ) 490nm, λem ) 520 nm),
meant to visualize the mixing in the microdroplet, reveals that,
above 40 °C, fluorescent beads move along circular trajectories
within [tmba][NTf2] droplets (see Supporting Information) and
that the beads velocity increases with the temperature.29 This
experiment shows that a simple heating, which produces a
temperature gradient and, consequently, a surface tension gradi-
ent,30 can activate mixing in the droplet. The second mixing
activation method tested is based on moving the droplet to and
fro along a row of five electrodes at a frequency of 0.85 Hz (55 V)
creating a flow inside the droplet.2 For the third mixing activation
test, the droplet was positioned on a module generating surface
acoustic waves (SAWs from Advalytix AG), which leads to a flow
created by the longitudinal pressure wave diffracted inside the
droplet.31,32 It should be noted that the experiment conducted with
fluorescent beads shows that, in that case, the mixing process
follows a chaotic advection,33,34 which accelerates the mixing by
decreasing the diffusion length.
The model click chemistry reaction was used as a tool to study
the influence of these different mixing activation methods on the
apparent reaction rate by comparing the initial reaction rates in
the microreactor obtained after merging the two initial droplets.
The three mixing activation methods lead to a reduction of the
time required for reaction completion with different ranges. The
fastest reaction completion is obtained under SAW activation,
where the reaction is completed within 21 min at room temper-
ature. The initial reaction velocities in droplet, viapp, determined
for the three mixing activation methods were compared to the
initial rate vi determined from a standard macroscale experiment
(magnetically stirred solution). In all cases, the measured viapp
values were smaller than vi indicating that, whatever the mixing
activation method, the reaction is mass-transfer limited in the
microdroplet. Based on this observation, eq 1 was modified by
including a correction factor (ε) that represents the efficiency of
the mass transfer (eq 2)
Efficiency factors of 0.53, 0.53, and 0.69 were obtained at 25
°C for respectively the first (heating), second (to and fro move),
and third (SAW) mixing activation methods.
Finally, the three mixing activation methods were studied at
different temperatures (Figure 5B). Whatever the temperature,
the fastest initial reaction rates is obtained under SAW activation
while stationary or displaced droplets exhibit comparatively poor
results, especially at 60 °C. Moreover, the initial reaction velocities
obtained with stationary or displaced droplet being quite similar,
we can assume that the benefit of mixing activation by droplet
displacement is moderate. In contrast, whatever the mixing
method, the initial reaction rate significantly increases with the
temperature, demonstrating that mixing is accelerated by heating.
Interestingly, whereas linear increases are obtained with stationary
or displaced droplets, an exponential increase is obtained with
SAW activation as is the case for standard macroscale reaction.
In addition, the reaction becomes less and less mass-transfer
limited as the temperature increases with ꢀ values approaching 1
(at 60 °C, ꢀ ) 0.88). This is not the case with stationary or
displaced droplet for which temperature increases at opposite lead
to decreasing ꢀ values (Figure 5). The present study thus clearly
shows that SAW activation offers, for soft IL microreactors, an
efficient substitute to common macroscale magnetic stirring. This
is an important step in the implementation of general and efficient
operating protocols, demonstrating that electrically operated IL
microdroplets indeed open a promising route toward microscale
organic synthesis.
CONCLUSSION
This paper describes the synthesis and the real-time monitor-
ing of 1,2,3-triazole by a 1,3-dipolar cycloaddition in ionic liquid
droplets acting as soft wall-free microreactors. This click chemistry
reaction was used as a powerful tool to evaluate the efficiency of
different mixing activation methods on the reaction progress by
comparing the initial reaction velocities. This study provided
demonstration that the combination of chaotic advection created
by SAW and temperature increase allows us to approach a kinetic
regime similar to that for macroscale standard organic reactions
in solution (i.e., magnetically stirred solutions of organic reagents
in regular “hard” flasks).
ACKNOWLEDGMENT
This work was partially supported by the Ministe`re de
l’enseignement supe´rieur et la Recherche. We thank the members
of Biosystems On Chip laboratory. We also thank S. Gmouh
(UMR6510) for technical assistance in the synthesis and D.
Drouin-Kucma (UMR6510) for spectroscopic characterizations.
We gratefully acknowledge C. Katan (UMR6510) for transition
orbitals calculation.
νiapp ) ενi
(2)
SUPPORTING INFORMATION AVAILABLE
Description and fabrication of the EWOD chip; instrumentation
and optical setup for monitoring; visualization of Marangoni effect.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(28) Daviaud, F.; Vince, J. M. Phy. Rev. E 1993, 48, 4432
.
(29) Toduka, H.; Hayamizu, K.; Ishii, K.; Bin Hasan Susan, M.; Watanabe, M. J.
Phys. Chem. B 2006, 110, 2833
(30) Fowler, J.; Moon, H.; Kim, C. J. Micro Electro Mech. Syst. 2002, 97–100
(31) Wixforth, A.; Strobl, C.; Gauer, C.; Toegl, A.; Scriba, J.; Guttenberg, Z. Anal.
Bioanal. Chem. 2004, 379, 982
(32) Sritharan, K.; Strobl, C. J.; Schneider, M. F.; Wixforth, A.; Guttenberg, Z.
Appl. Phys. Lett. 2006, 88, 54102
(33) Aref, H. J. Fluid Mech. 1984, 143, 1
(34) Ottino, J. M. Chem. Eng. Sci. 1994, 49, 4005
.
.
.
Received for review April 28, 2008. Accepted June 3,
2008.
.
.
.
AC800855U
Analytical Chemistry, Vol. 80, No. 15, August 1, 2008 6055