Microfluidics and Nanofluidics
Online Resource: Electronic Supplementary Material for paper:
Fabrication of Layered Polydimethylsiloxane (PDMS) / Perfluoropolyether (PFPE) Microfluidic Devices with Solvent Compatibility and Valve Functionality
Marco Domenichinia, Ranjana Sahaib, Piero Castrataroa, Roberto Valsecchic,*, Claudio Tonellic, Francesco Grecod and Paolo Darioa
aThe BioRobotics Institute, Polo Sant’Anna Valdera, Viale Rinaldo Piaggio, 34, 56025, Pontedera (PI), Italy
bNEST Laboratory, Scuola Normale Superiore, Piazza San Silvestro, 12, 56127 Pisa (PI), Italy
cSolvay Specialty Polymers, V.le Lombardia 20, 20021 Bollate (MI), Italy
dCenter for MicroBioRobotics@SSSA, Istituto Italiano di Tecnologia, Viale Rinaldo Piaggio, 34, 56025 Pontedera (PI), Italy
* To whom correspondence should be addressed. E-mail: .
1. Polymer swelling and solubility background
While details on polymer swelling and solubility analysis can be found elsewhere (Lee et al. 2003; Nohilé et al. 2008) a brief summary of some basic concepts is here reported as a Supporting Information (Online Resource), useful in discussing and rationalizing the results obtained in the swelling ratio tests.
In order to characterize solubility of a given material, a commonly selected parameter is the Cohesive Energy Density, CED, (cal/cm3), often expressed in terms of the Hildebrand value or solubility parameter, δ, where CED = δ2. In a unit volume of material, this energy gives an indication of the attractive interactions between molecules or, in other words, the energy that should be exceeded to allow for the insertion of solvent molecules in the solute molecules. Therefore, for a given couple solvent-solute, they should have similar cohesive energy densities (similar δ) in order to be soluble. In the case of materials which do not dissolve (like cross-linked polymers), the metric for solubility is the degree of swelling; this leads to the solubility parameter being a predictive tool for whether a solvent will cause swelling of a polymer. Swelling is more likely to occur for solvents whose solubility parameters are similar to that of the polymer.
However, this relationship is nonlinear and is generally different for each polymer and solvent combination. In addition, two solvents that have virtually the same solubility parameter δ can have significantly different swelling effects on a given polymer. One way of trying to predict this behavior is by examining the expression for Hansen’s total solubility parameter; that is,
δ2 = δd2 + δp2 + δh2(1)
where δd, δp, and δh represent the dispersion, polar, and hydrogen-bonding forces, respectively. It is found that, for a given polymer, the solvent that has the most similar partitioning of the solubility parameter in terms of these forces will swell the polymer the most. Because the individual values of δd, δp, δh are not easily available, a previous study regarding the swelling induced in PDMS by various solvents (Lee et al. 2003) used the molecular electric dipole moment of a solvent, µ, to represent the effect of polar forces. While the examination of these parameters can provide a general likelihood of the swelling behavior of a polymer, a reliable prediction of swelling from the solvent/polymer’s physico-chemical characteristics is still lacking. As such, many alternative models have been investigated including considering the solvent/polymer system’s chemical potential and the influence of mechanisms like diffusion, absorption, and adsorption where the size and shape of the solvent molecule plays a role. Moving in that direction, Nohile, Dolez, and Vu-Khanh (Nohilé et al. 2008) looked at results of the swelling of butyl rubber with various solvents and determined that, if the solvents were first separated into their respective chemical classes (e.g., aromatic hydrocarbons), then a plot of volume swelling versus the solvent molar volume produced a linear relationship as did a plot of volume swelling versus the solvent saturation vapor pressure.
2. Chip Application tests
Experimental
Tests were conducted with the fabricated chips by delivering a solvent through the fluid channel while the valve was operating at 0.5 Hz. Three different solvents were tested: acetone, toluene, and tetrahydrofuran. In the case of testing of acetone and tetrahydrofuran, a 8% fluorescein isothiocyanate solution (F7250, Sigma-Aldrich) in DI water (0.05 g per 1 ml of DI water) was added to solvents and both fluorescent and bright field images were taken of the fluid channel and functioning valve. Bright field images alone were captured for toluene. Measurements of the channel width and valve function were made throughout (by averaging the results of 5 fluorescent and bright field images taken every thirty minutes) of at least 8 hours of operation. The data were acquired using an inverted microscope (Nikon TE2000, Nikon Inc.) with a connected digital CCD camera (ORCA ER, Hamamatsu) that was controlled using NIS Elements software (Nikon Inc.). The images were analyzed using Image J (National Institute of Health, USA), MATLAB (Mathworks, MA) and ORIGIN Pro 8.1 software (OriginLab Corp., Northampton, MA, USA).
Tests results
Validation tests were conducted on the fabricated microfluidic chips in order to demonstrate the feasibility of the proposed approach. During these tests it has been possible to determine the pressure levels the chip could withstand without any delamination of the layers and the pressure values needed to close the valves. No delamination between the layers of FOMBLIN® MD40 PFPE or between the FOMBLIN® MD40 PFPE and the PDMS layers was observed when delivering pressure up to 35psi to the control channels layer that was the maximum pressure setting on the compressed air pressure regulators used in this study. Even if the pressure actually being applied to the chip could be lowered because of minor losses in the connecting tubing, test demonstrated reliability of the chip assembly under normal working conditions.
Efficient closing of the valves by pressurizing control channels was visualized under the microscope by injecting a fluorescent dye solution inside the fluid channel and applying pressure to the control channel. Complete darkening of the fluid channel was found to occur at the valve (where the two channels intersect) when the fluid channel was fully closed. The full closing of the valve typically occurred at about 25psi, as read on the pressure regulator. The pressure needed to close the valves in these chips are typically slightly higher than those needed for equivalent PDMS ones since the FOMBLIN® MD40 PFPE is less elastic than PDMS, having a higherYoung's modulus (9.6 vs. 2.4 MPa).
A series of tests were conducted on the application chip with three different solvents (acetone, toluene, and tetrahydrofuran)to determine if any swelling occurs in the fluid channels and if the valve function remains unaffected after several hours of continuous operation. The choice of these three solvents for testing was made for three main reasons. First, they are all classified in “Group 2” ofswelling behavior for PDMS, as assessed in the reported swelling tests, and therefore their use in PDMS in microfluidic devices must be avoided due to large swelling. Testing FOMBLIN® MD40 PFPE as PDMS replacement in these cases is therefore crucial. Second, they usefully represent the whole range of swelling observedin FOMBLIN® MD40 PFPE material; as reported in Table 1, indeed, S = 1.00 for acetone, S = 1.01 for toluene, S = 1.04 for THF. Third, and final, while the maximum swelling ratio observed in the series (S = 1.05, obtained in the case of diisopropylamine) was slightly larger, testing this latter case has been considered less significant for applications with respect to testing more common solvents as they are normally employed in a large variety of reactions, as in the case of the selected solvents. As stated previously, the three solvents were tested by delivering them through the fluid channel at a pressure of 3-5 psi while operating the valve at 0.5Hz. In order to assess the resistance of the assembled microfluidic chip to solvent swelling over the time a measurement of the width of the channel was taken at the beginning. Similar measurements were then taken every thirty minutes over 8 hours of continuous operation and then compared. Results showeda negligible change in the width of the fluid channels up to 8 hours of continuous operation for all the three solvents tested.Valve function also continued unimpaired throughout for all three solvents cases. In order to visualize this, images were taken in correspondence of valve, i.e. at crossing of control and fluidic channels. As an example, Figure S1(g) shows images of normal operation of the valve (open- closed), while images (a-f) demonstrates the functioning of the valvein the presence of acetone at three different times of operation (3, 3.5, 4 hours of operation), and then 4 hours later (7, 7.5, 8 hours of operation). Corresponding light intensity profiles taken along the fluidic channel widthare also reported in the graph in Figure S1. As it can be seen from the figure, the valve function remains essentially unchanged. Also in the case of continuous operation (up to 8 hours) in the presence of solvents no delamination of the layers occurred.
Figure S1.Valve function after continuous operation with acetone-filled fluid channels. Up Left: Comparison between valve function (valve open) after 3 (a), 3.5 (b) and 4 (c)hours of operation at 0.5 Hz and then after 7 (d), 7.5 (e) and 8 (f) hours of operation as obtained in fluorescence microscope imaging. Scale bar: 50 µm. Up Right: Corresponding light intensity profiles taken along the fluidic channel width. Bottom: g) normal operation of valve at beginning of operation: valve opened (left) and closed (right). Scale bar: 50 µm
REFERENCES
Lee JN, Park C, Whitesides GM (2003) Solvent compatibility of poly(dimethylsiloxane)-based microfluidic devices. Anal Chem 75:6544-6554. doi:10.1021/ac0346712 [doi]
Nohilé C, Dolez PI, Vu-Khanh T (2008) Parameters controlling the swelling of butyl rubber by solvents. Journal of Applied Polymer Science 110:3926-3933. doi:10.1002/app.29004