5230
Langmuir 2006, 22, 5230-5232
Stable Modification of PDMS Surface Properties by Plasma Polymerization: Application to the Formation of Double Emulsions in Microfluidic Systems Valessa Barbier,*,† Michae¨l Tatoulian,‡ Hong Li,† Farzaneh Arefi-Khonsari,‡ Armand Ajdari,† and Patrick Tabeling† Groupe Microfluidique, MEMS & Nanostructures, UMR 7083, ESPCI, 10 rue Vauquelin, 75005 Paris, France, and Groupe Ge´ nie des Proce´ de´ s Plasma et des Traitements de Surface, ENSCP, UniVersite´ Pierre et Marie Curie, 11 rue Pierre et Marie Curie, 75005 Paris, France ReceiVed December 5, 2005. In Final Form: February 10, 2006 We describe a method based on plasma polymerization for the modification and control of the surface properties of poly(dimethylsiloxane) (PDMS) surfaces. By depositing plasma polymerized acrylic acid coatings on PDMS, we succeeded to fabricate stable (several days) hydrophilic and patterned hydrophobic/hydrophilic surfaces. We used this approach to generate direct and (for the first time in this material) double emulsions in PDMS microchannels.
Poly(dimethylsiloxane) (PDMS) is a material especially well adapted for the fabrication of microsystems. It permits rapid prototyping using soft lithography,1 it is inert and rather biocompatible (at least for in-vitro applications), and its transparency is convenient for visualization and detection. The flexibility of PDMS is suitable for the integration of actuators such as microvalves and micropumps. This capability is crucial for increasing the number of functionalities in lab-on-a-chip devices2 and achieving flow control in microfluidic systems. As an illustration of this feature, it has recently been shown that, by using microactuators, droplet sizes produced in miniaturized systems can be varied by 1 order of magnitude without changing the imposed flow conditions.3 However, despite its advantages, the use of PDMS is limited because of the poor control of its surface properties that is currently achieved. This is obviously a limitation since in microfluidics, owing to the large surface-to-volume ratios, surfaces play a crucial role. Moreover, in practice, surfaces with specific properties are often needed (biospecific, hydrophilic, charged, etc.). The control of surface properties is thus essential for an appreciably wide range of applications. This has generated growth in research activity in recent years, specifically dedicated to modify the surface properties of PDMS.4 The most common method for modifying the wetting properties of PDMS is based on surface oxidation by an energetic source that is usually oxygen plasma.5,6 Adsorption of specific molecules such as surfactants, lipids,7 or polyelectrolyte multilayers8,9 has also been investigated. Nonetheless, neither exposure to oxygen plasma nor dynamic coatings lead to long term stable surfaces; * To whom correspondence should be addressed. † UMR 7083. ‡ Universite ´ Pierre et Marie Curie. (1) McDonald, J. C.; Whitesides G. M. Acc. Chem. Res. 2002, 35, 491-499. (2) Thorsen, T..; Maerkl, S. J..; Quake, S. R. Science 2002, 298, 580-584. (3) Willaime, H.; Barbier, V.; Kloul, L.; Maine, S.; Tabeling, P. PRL 2006, 96, 054501. (4) Makamba, H.; Kim, J. H.; Lim, K.; Park, N.; Hahn, J. H. Electrophoresis 2003, 24, 3607-3619. (5) Hillborg, H.; Ankner, J. F.; Gedde, U. W.; Smith, G. D.; Yasuda, H. K.; Wilstro¨m, K. Polymer 2000, 41, 6851-6863. (6) Ginn, B. T; Steinbock, O. Langmuir 2003, 19, 8117-8118. (7) Lenz, P.; Ajo-Franklin, C. M.; Boxer, S. G. Langmuir 2004, 20, 1109211099. (8) Ro, K. W.; Chang, W.-J.; Kim, H.; Koo, Y.-M.; Hahn, J. H. Electrophoresis 2003, 24, 3253-3259. (9) Wu, D.; Luo, Y.; Zhou, X.; Dai, Z.; Lin, B. Electrophoresis 2005, 26, 211-218.
typically, the original hydrophobicity is regained after several minutes to several hours. It is generally believed that the physical mechanism driving the recovery process is the migration of low molar chains from the bulk of the material to the surface.5 An approach that may inhibit this process and thus favor the realization of stable coatings is the covalent bonding of polymer layers onto the PDMS surface.4 Such a layer could act as a physical barrier that limits PDMS chains migration toward the surface. The few methods proposed over the last years for binding polymers to the PDMS surface are based on silanization,10 chemical vapor deposition (CVD),11 or UV-polymerization.12-13 Plasma processes have also been explored. In particular PDMS surfaces have been recently functionalized by acrylonitrile14 and imidazole15 in a microwave excitation frequency system. Altogether, given the technological importance described above and the limitations of the methods explored so far, there is still a demand for expanding the number of techniques permitting control of the surface properties for this particular material. A “must” would consist of a robust method leading to stable and Versatile coating of the surfaces. This is the objective of the present work. The approach we develop here is based on the use of a lowpressure plasma reactor for depositing polymer films onto PDMS surfaces.16 This original method has not been explored for microfluidic applications despite its potential advantages. First, it is really dedicated to surface modification and does not alter bulk properties. Second, in contrast with other methods (except CVD), the coating is performed under vacuum. The monomer is the only reactant feeding the process, and the amount of byproducts is low (compared to wet chemistry). Moreover, when compared to the techniques presently mentioned, the coating process is fast, with deposition speeds on the order of 10 nm/s. Finally, the technique is versatile as different kinds of monomers (10) Hellmich, W.; Regtmeier, D. T. T.; Ros, R.; Anselmetti, D.; Ros, A. Langmuir 2005, 21, 7551-7557. (11) Lahann, J.; Balcelis, M.; Hang, L.; Rodon, T.; Jensen, K. P.; Langer, T. Anal. Chem. 2003, 75, 2117-122. (12) Hu, S.; Ren, X.; Backman, M.; Sims, C. E.; Li, G. P.; Allbritton, N. Electrophoresis 2003, 24, 3679-3688. (13) Hu, S.; Ren, X.; Backman, M.; Sims, C. E.; Li, G. P.; Allbritton, N. Anal. Chem. 2004, 76, 1865-1870. (14) He, Q.; Liu, Z.; Xiao, P.; Liang, R.; He, N.; Lu, Z. Langmuir 2003, 19, 6982-6986. (15) Bae, W.-S.; Urban, M. W. Langmuir 2004, 20, 8372-8378. (16) Arefi, F.; Andre, V.; Montazer-Rahmati, P.; Amouroux, J. Pure Appl. Chem. 1992, 64, 715-723.
10.1021/la053289c CCC: $33.50 © 2006 American Chemical Society Published on Web 05/05/2006
Letters
Langmuir, Vol. 22, No. 12, 2006 5231
Figure 1. Schematic drawing of the process used for the surface treatment of PDMS with PPAA by a low-pressure plasma polymerization; 1 hollow electrode; 2 cylinder: electrode & samples carrier (length 22 cm and diameter 3.5 cm); 3 plasma zone; 4 turbomolecular pump; 5 chemical pump; 6 monomer gas bubbling. Table 1. Plasma Polymerization; Optimized Parameters: Gas, Operating Pressure, Plasma Power and the Real Time Spent by the PDMS Substrate in the Plasma Discharge step
gas
operating pressure (mbar)
power (W)
time in the discharge (s)
1 2 3
Ar Ar/AA He
0.7 0.4 0.7
20 25 20
5 30 5
can be activated by plasma17 opening pathways for realizing a wide range of surface modifications with a reduced number of limitations if compared to current approaches (e.g., thermal constraints in CVD and photochemical ones in UV polymerization). To create hydrophilic poly(dimethylsiloxane) surfaces, we deposit plasma polymerized acrylic acid (PPAA) coatings on PDMS substrates. We use a bell-jar reactor as described in Figure 1. The reactor is evacuated by a turbo molecular pumping system, and a base pressure of 10-3 Pa is reached. Then, the operating pressure is maintained at a prescribed value by a chemical pump. The plasma is produced in the gap between two electrodes less than 1 cm apart: a stainless steel blade (hollow) and a grounded stainless steel cylinder. The first one serves for the introduction of gases, and the second one is used as a rotating sample carrier. The gaseous monomer is produced by bubbling gas (10 cm3/ min) in a 10 mL round-bottom flask containing the acrylic acid monomer heated at 40 °C. We have varied operational parameters to optimize the process (this optimization will be described elsewhere18) which has led us to the following 3-step protocol (see Table 1 for parameters): 1. Pretreatment. This step activates the surface before deposition since the oxygen present as traces in the reactor is excited by Ar plasma which leads to a partial oxidation of the topmost layer. That leads to a better adhesion of the deposited film. The Ar pretreatment is also used to cross-link the PDMS surface in order to (i) decrease the ablation of Si during plasma deposition and further contamination of the deposited film and (ii) limit the migration of the PDMS chains and thus the surface regeneration. 2. Polymer Deposition. In a second step, a plasma of Ar/ (17) Lee, W. W., d’Agostino, R., Wertheimer, M. R., Eds.; Plasma Deposition and Treatment of Polymers; MRS Symp. Proc. Vol.544, MRS, 1999. (18) Tatoulian, M.; Barbier, V.; Li, H.; Arefi-Khonsari, F. Chem. Mater. 2006, under preparation.
Figure 2. XPS spectra, PDMS coated with PPAA by plasma polymerization; Nature and composition of observed bonds: COOH, 23.1% (288.727 eV); C-OH, 2.9% (286.375 eV); C-C, 42% (284.775 eV); C-Si, 32% (284,075 eV). Insert: XPS spectra of uncoated PDMS.
acrylic acid is used to deposit an acrylic acid coating on the pretreated PDMS. 3. Posttreatment. This last step is used to cross-link the deposited polymer film and to reinforce its cohesive strength. As a consequence, the resulting film is less soluble in water, which limits the hydrophobic recovery of coated PDMS with time. Parallel studies with the same deposition process on other substrates found deposition speeds of 15 nm/s,19 which suggests that the thickness of the films we obtain should be typically 450 nm. The deposited films were characterized by X-ray photoelectron spectroscopy (XPS), contact angle measurements (CA), and scanning electron microscopy (SEM). XPS spectra have shown the presence of carboxylic acid groups (288.7 eV) on the surface, indicating that acrylic acid has actually been deposited (Figure 2). Low resolution XPS spectra also revealed the presence of Si on the surface. The atomic percentage of Si measured in the resulting PPAA coatings decreases from 9% without pretreatment to 4% after an Ar pretreatment. The conclusions on the presence of C-Si bonds in the deposited film are still under discussion and will lead to further experiments. To quantify the hydrophilicity of the coated surface, we performed static CA measurements. These show a high hydrophilicity before rinsing with water (CA < 10°, to be compared to 110-120° for untreated PDMS). After rinsing with distilled water, the contact angle increases up to 50°-60°, still much lower than untreated PDMS. Interestingly, we find that full wettability is reached by adding tiny amounts of surfactant to the water used for CA measurements (Figure 3). This is in sharp contrast with uncoated PDMS for which a similar amount of added surfactant has no significant influence on the contact angle. SEM experiments were performed and showed that the PPAA deposited film is quite homogeneous at the micron scale. Moreover, the coating is stable and resistant to rinsing with water, which is crucial for applications where the surface is exposed to flowing aqueous solutions. To conclude, PPAA was successfully deposited on a PDMS surface. That leads to an homogeneous hydrophilic polymer layer (19) Jafari, R.; Tatoulian, M.; Arefi-Khonsari, F. React. Func. Polym. 2006, submitted.
5232 Langmuir, Vol. 22, No. 12, 2006
Letters
Figure 3. Static contact angle measurements of a water drop immersed in oil (tetradecan). Contact angles are given for both uncoated PDMS (×) and PDMS coated with PPAA by plasma polymerization (O) and with increasing concentration of surfactant (sodium dodecyl sulfate, SDS).
with (1) full wettability for water with surfactant (SDS around its cmc) and (2) a rather good resistance to rinsing. We now demonstrate the potential of our plasma polymerization method by generating emulsions in PPAA coated PDMS microchannels. Although inverse emulsions are currently produced in PDMS, direct and double emulsions, requiring nonhydrophobic surfaces, have thus far been obtained in glass only, where versatile surface treatments are available.20-22 The reason is that, as shown in a previous study,23 the nature of the droplets that can be formed in a microsystem depends on the wetting properties of the working fluids with respect to the walls. In systems of practical interest where well defined droplets are produced, the fluid that wets preferentially the walls must form the continuous phase. As PDMS is hydrophobic, it is naturally dedicated and restricted to water in oil (W/O) emulsions. By making PDMS hydrophilic by plasma polymerization as described before, we could generate O/W single emulsions at a simple T-junction (Figure 4a). More remarkably, we could also generate W/O/W double emulsions in a PDMS device (Figure 4b-e). Double emulsions were formed using two T-junctions in cascade, in a scheme similar to that used by Okushima et al., in glass.21 Here we used an advantage of our method: the possibility to pattern the surface properties of the surface using masking techniques during the plasma polymerization. Indeed a waterin-oil emulsion is produced at the first T-junction in a zone that has not been coated and remains thus strongly hydrophobic. This primary W/O emulsion is further encapsulated by an aqueous phase at the second T-junction, treated hydrophilic by deposition of a PPAA film. The experiments could run continuously for more than 3 weeks without any significant evolution. For comparison, the stability of PDMS microsystems treated with O2 (30′′, plasma cleaner) for the same flow conditions, is below 1 day. To the best of our knowledge, this is the first time that double emulsions are stably produced in a PDMS device. This illustrates the possibilities offered by plasma polymerization for microfluidics applications. In summary, this letter reports a robust way to master PDMS surface properties in a stable manner. The method described (20) Joanicot, M.; Ajdari, A. Science 2005, 309, 887-888. (21) Okushima, S.; Nisiksako, T.; Torii, T.; Higuchi, T. Langmuir 2004, 20, 9905. (22) Utada, A. S.; et al. Science 2005, 308, 537. (23) Dreyfus, R.; Tabeling, P.; Willaime, H. Phys. ReV. Lett. 2003, 90 (14).
Figure 4. Single O/W (a) and double W/O/W (b-e) emulsions made in a PDMS microsystem; Droplets are produced at T-junctions where both continuous and dispersed phases meet. The cross section of microchannels is 200 × 50 µm2. Fluids are driven by syringe pumps with flow rates varying from 0.1 to 20 µL/min. Observations are made using fluorescence video microscopy (water phases are labeled with fluorescein). For single O/W emulsions (a) tetradecan was used as dispersed phase and water with sodium dodecyl sulfate (SDS, 0.4%) as continuous phase. The whole microsystem is coated with PPAA. For O/W/O double emulsions (b-e), primary droplets of water in tetradecan (+SPAN 80, 0.75%) are produced at a first T-junction (uncoated, hydrophobic) and are then encapsulated with water and Tween (2%) at a second T-junction (coated with PPAA, hydrophilic). The hydrophobic/hydrophilic patterning was achieved by masking the area of the first T-junction with a glass slide according to the dotted line on picture b).
here is versatile. Its relevance for microfluidics is illustrated here for multiphase flow operation; it could be used as well in situations where surface functionalization is essential. The variety of polymers that can be used is large: we thus have, potentially, a tool-box of properties/functional groups, permitting the ondemand generation of new surfaces. Finally, this plasma polymerization method naturally lends itself to the creation of patterns. The microfluidic controlled generation of double emulsions reported here directly takes advantage of such a (hydrophilic/hydrophobic) pattern, with features at the millimeter scale. LA053289C
