Compatibility of poly(dimethylsiloxane) microfluidic systems with high viscosity hydrocarbons
- 35 Downloads
In this work, the compatibility of poly(dimethylsiloxane) (PDMS) with high viscosity petroleum liquids (bitumen) was evaluated to study the possibility of using PDMS microchannels in heavy oil and bitumen extraction research using solvent based methods such as Vapor Extraction. Three curing agent to base ratios (1:2, 1:10, and 1:20) were used to fabricate PDMS. Swelling ratio of these three samples in the vicinity of different organic solvents and diluted bitumen were evaluated by measuring the weight of sample every 15 min. It was found that 1:10 ratio PDMS was preferential ratio for minimizing absorption. We were particularly interested in the kinetic of swelling, deformation, and discoloration of PDMS. The hypothesis was that a PDMS microchip can still be used for experiments with heavy oil and bitumen if the experiment time is much shorter than the time it takes for PDMS discoloration, swelling, and deformation. The coating of PDMS slabs with trichloro (1 h 1 h 2 h 2 h-perfluorooctyl) silane was also tested and swelling ratios were again measured to evaluate the effect of surface coating. Swelling ratios were at the same order for uncoated PDMS which shows silane coating is not able to improve the solvent compatibility of PDMS. Moreover, solubility parameter was used to predict the swelling ratio of 1:10 PDMS sample in organic solvents. It predicts that hexane and toluene have the highest solubility that was in great agreement with experimental results. It seems that solubility parameter is a reliable factor to qualitatively predict the swelling ratio of PDMS. The effect of bitumen on transparency of PDMS was also studied using Ultraviolet–Visible spectrophotometer.
KeywordsMicrofluidics Poly(dimethylsiloxane) Polymers Swelling Enhanced oil recovery Chemical compatibility
Microfluidic devices have been applied in petroleum or energy [1, 2, 3, 4] for evaluation of wetting properties , single  and multiphase flow study , asphaltene deposition , enhanced oil recovery [9, 10] or other fields like fuel cells , microbioreactors , etc. Glass was the main substrate for micro fabrication early in 1990s due to its mechanical and optical properties, low chemical reactivity, and transparency and they are mostly fabricated by photolithography and wet etching methods . However, there are number of disadvantages for glass microchannels like high price and multistage fabrication process and irreproducible thermal bonding process . Poly(dimethylsiloxane) (PDMS) is one of the most popular materials for microfluidic fabrication that has been thoroughly studied in various fields of application [15, 16, 17]. PDMS microchannels are low-cost and easy to fabricate with low polarity and chemical inertness [18, 19, 20]. Moreover, they are transparent to visible light and suitable for detection systems like ultraviolet–visible (UV–Vis) and fluorescence . PDMS consists of base and curing agent that various curing agent to base ratios alter the properties of elastomer such as stiffness [22, 23, 24].
One of the main drawbacks of PDMS is that it swells with solvents such as toluene and low molecular weight organic solutes. In addition, bitumen as a complicated material with various fractions and large molecules such as asphaltene can affect the swelling ratio. Although accurate knowledge of the dimensions is very important in analysis of data captured from microchannel and fluid flow, swelling deforms the microchannel and change the rate and profile of the flow, consequently. Therefore, compatibility evaluation of PDMS with organic solvents along with heavy oil and bitumen  are the basic requirements in specific energy related fields such as capillary driven flow and pore scale study of effective forces on bitumen extraction where solvents are used for viscosity reduction. Noteworthy to mention that our hypothesis for this work was that if the swelling rate is slow, we can still use PDMS as a cheap and easily-made choice for microfluidic studies where the experimental time (i.e. contacting time of PDMS and bitumen) is shorter than swelling rate of the chip.
There are a number of surface modification methods to improve the solvent compatibility of the PDMS. Abate et al.  presented a coating technique with a glass-like layer using sol–gel chemistry. It increases chemical resistance of the channel, but it changed the dimensions of the designed microchannel. Coating of the surface with other materials like Teflon  and silane  are reported by other authors. Laser surface modification  is another method to adjust the surface but considering this technique, PDMS is not a cheap and fast method of microfluidic device fabrication anymore. Application of self-assembled monolayer (SAM) with a high number of CF3 and CF2 groups alter the wettability of the surface . Presence of these functional groups on the surface of FOTS coated PDMS has been reported by other authors . There are a few limitations on PDMS surface treatment such as aggregation of fluorocarbon and surface structural defects that impacts the performance of the coating .
In this study, dynamic of swelling ratio of PDMS in pure solvents and diluted bitumen samples (10–50%) in small time-lapsed increments were measured and transparency of swelled PDMS as a function of time was evaluated using UV–Vis. Total Internal Reflection Fluorescence (TIRF) microscopy was also applied to detect the bitumen on PDMS samples due to the fluorescence nature of the bitumen . Moreover, PDMS slabs were coated with trichloro(1 h 1 h 2 h 2 h-perfluorooctyl) silane (FOTS) in order to evaluate the effect of silane coating on solvent and bitumen compatibility of PDMS.
3 Materials and method
Hexane (lab grade), acetone (lab grade), methanol (analytical grade), toluene (HPLC grade), and isopropanol (lab grade) were supplied from Sigma-Aldrich Co., USA. PDMS (Sylgard 184 Silicone Elastomer Kit) purchased from Dow Corning Co., USA. Trichloro(1 h 1 h 2 h 2 h-perfluorooctyl) silane (FOTS) was supplied from Sigma-Aldrich Co., USA. McKay River bitumen was provided by Innovates Alberta.
3.2 PDMS fabrication and bitumen sample preparation
According to the manufacturer (Dow Corning, Midland, MI), the recommended ratio for curing agent to base is 1:10. Here we have tried two other ratios with higher curing agent (1:2) and less (1:20) to evaluate the optimum ratio for PDMS fabrication. Therefore, PDMS samples were prepared by mixing a 1:2, 1:10, and 1:20 curing agent to base (Sylgard 184) ratios. Higher curing agent:base ratio means higher cross linking and rigid PDMS structure that breaks easily under shear strain. Mixture was degassed in a vacuum chamber for 60 min to remove the air trapped during the mixing. Then mixture was poured into a brass mold with two cylindrical reservoirs and a 20 mm × 2 mm × 100 µm (length × width × depth) rectangular channel connecting the reservoirs and cured for 24 h at 80 °C to add Si–H bonds to C=C bonds. PDMS was extracted and individual chips was washed and glass substrate (24 × 50 microscope coverslips from Fisher Scientific) was bonded to the PDMS forming an enclosed space. Glass substrates were cleaned using piranha solution (H2SO4 and H2O2 with 3:1 volume ratio) and rinsed using de-ionized water. Cleaned PDMS and glass substrate wafers were pressed together and baked in oven at 80 °C for 1 h for the reversible sealing using van der Waals forces at room temperature. A drop of fluid sample (~ 15 µl) was placed in the inlet reservoir and fluid flowed through the micromodel due to capillary forces. In addition to microchannels, PDMS samples were cut into small square pieces (2.5 cm × 2.5 cm) to submerge in different liquids. Pieces were washed with acetone and then DI water to remove the possible contaminants. Stock bitumen was used to prepare 10–50% bitumen dilutions in hexane. All the samples were aged for 24 h to settle down the large aggregates and precipitates.
3.4 UV–Vis spectrophotometer
3.5 TIRF microscopy
In this work, TIRF microscopy was applied to image absorbed bitumen on the surface layer of the PDMS due to its the natural fluorescence . This method excites fluorescence at solid surface without any other fluorescence background from deeper areas . Therefore, fluorescence nature of the bitumen enables us to capture absorption of bitumen on the surface of the sample. As mentioned earlier, TIRF microscopy eliminates background noises of the PDMS. In fact, PDMS has small pores on the surface and TIRF microscope illustrates bitumen distribution on specific spatial location. Based on previous study in our group , a 488 nm laser source and 500–550 nm (green) emission filter were selected for the microscope.
4 Result and discussion
4.1 Pure solvents swelling
The first part was related to the effects of PDMS preparation method on solubility in different solvents. Therefore, weight of swollen networks at different reagent to base ratios (1:10, 1:2, 1:20) were recorded. All sides of PDMS slabs should be covered with solvent for a uniform absorption of the solvent. Swelling ratio of the PDMS in solvent should be measured at equilibrium condition. This condition was considered at the time that there are no more changes in the mass of swollen network. Based upon this, a procedure was developed whereby small (5 cm × 5 cm) squares of PDMS were immersed in solvent in airtight bottles. Due to the volatile nature of the exposed to air solvents at room temperature (22 °C), the evaporation of the solvent could affect the measurements. This was resolved by measuring the weight of samples while it was submersed in the solvent plate. Weight of samples were measure in different time intervals until a constant weight was observed.
As mentioned earlier, 1:20 PDMS slab had highest swelling ratio where submerged slabs in toluene and hexane with 128% and 105% increase in weight had the largest ratios, respectively. It means low cross linked elastomer (1:20 curing agent:base PDMS) swells more than other ratios where solvent molecules detach non-crosslinked oligomers from PDMS. Toluene and hexane swelled over 85% after 80 min while other solvents have less than 20% of swelling. In fact, PDMS with low polarity  contribution had higher swelling ratio in liquids with lower polarity index (i.e. toluene and hexane) while PDMS is more compatible with polar liquids (i.e. methanol, acetone, isopropanol). Solvents with high solubility in PDMS are not compatible with PDMS microfluidic platforms even for fast experiments. As can be seen, higher rate of swelling is related to the first 60 min and was decreased until equilibrium condition was achieved. For example, swelling was stopped after 350 min for 1:20 PDMS in toluene and 250 min for 1:20 ratio PDMS in hexane. Comparison between 1:10 and 1:2 PDMS slabs revealed that higher reagent to base ratio sample (1:2) with a rigid, fragile structure, had higher swelling ratio. Ratio of the curing agent to base affects the number of cross-links in cured PDMS and higher swelling ratio of PDMS in solvent, consequently . It seems that 1:10 cross-linker to base reagent ratio was the preferential ratio for minimizing organic solvent absorption. Neither increasing nor decreasing the amount of cross-linker had any beneficial effects for solvent compatibility. This is in good agreement with previous studies that showed minimum swelling ratio for 1:10 PDMS sample .
Solubility parameters and interaction radius of solvents and PDMS
4.2 Diluted bitumen
4.2.1 Swelling ratio
4.2.2 Transparency and fluorescence intensity
Fluorescent images showed that bitumen content of the PDMS surface is increasing fast that makes PDMS inappropriate for micro scale monitoring of heavy oil and bitumen. Moreover, bitumen is increasing is size on the surface that could be related to bridging and aggregation of asphaltene and other components of the bitumen on the PDMS surface. However, larger size of bitumen molecules compared with pure solvents did not reduce the swelling rate of bitumen in PDMS slab and PDMS microchannel. All in all, it seems dynamic of the swelling is fast and it is not possible to use PDMS either for pure solvent or high viscosity fluids such as bitumen and heavy oil.
Application of PDMS microchannel as a fast and cheap approach can improve our understanding of fluid flow in small scales such as porous media. Solubility parameters are good factors for swelling correlation of PDMS. It has been reported that PDMS has high swelling of pure solvents that makes it unsuitable for solvent related studies. This was confirmed with polar factor in solubility where similar polar factor leads to high swelling of hexane and toluene in PDMS. Swelling dynamic of pure liquids and large molecules such as bitumen along with its swelling ratio has not thoroughly evaluated yet. On the other hand, if dynamic of swelling is not very fast, PDMS microchannel can still be applicable for investigation of fluid flow. In this study, it was found that swelling of immersed PDMS into toluene and hexane was very fast with deformation of the PDMS slab (i.e. curved edges). Degree of swelling in other solvents was relatively small with faster steady state swelling. In addition, experimental solvent swelling of different PDMS (base to curing agent ratio) was measured to find the best ratio for PDMS fabrication and 1:10 ratio PDMS had the minimum swelling ratio. The improvement of solvent compatibility of PDMS with silanization was evaluated using FOTS as a fast and chip method of surface modification. However, high level of swelling ratios was reported for coated PDMS slabs. This means that silanization is not a suitable approach for solvent compatibility enhancement.
Moreover, compatibility of PDMS with bitumen as a highly viscous and large molecule fluid was evaluated to see if PDMS is applicable for heavy oil and bitumen flow evaluation. Diluted bitumen (10–50 wt%) was injected in PDMS fabricated microchannel and the shape of the microchannel was disturbed after injection of diluted bitumen quite fast and glass, PDMS bonding was broken. Deformation of microchannel along with fast dynamic of swelling mean that PDMS is not suitable for evaluation of flow and rheology of heavy oil and bitumen. Furthermore, transparency of 1:2 and 1:10 PDMS slabs were measured that higher concentration of bitumen led to lower transparency and poor bitumen flow visualization consequently. TIRF microscopy was the last step for evaluation of bitumen swelling dynamic. It clearly showed that bitumen concentration on the PDMS surface was increasing fast. High level of swelling, low transparency and high concentration of bitumen on the surface are solid reasons to believe that PDMS microchannels are limited to certain solvents and are not a potential replacement for glass microchannels for capillary flow evaluation of bitumen and heavy oil. Equilibrium swelling ratio of coted PDMS revealed that surface treatment with short chain silane did not improve the resistivity toward swelling.
The authors would like to thank InnoTech Alberta and Natural Sciences and Engineering Research Council of Canada (NSERC) for funding. The authors also acknowledge Dr. Hyun-Joong Chung for his insightful comments and the use of silanization unit in his lab. The authors would like to thank Nicholas Moore for helping with experimental data.
This study was funded by Alberta Innovates - Alberta Research Council Core Industry (AACI) PhD Program (Grant No. RES0024693).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- 6.Keshmiri K, Mozaffari S, Tchoukov P, Huang H, Nazemifard N (2016) Using microfluidic device to study rheological properties of heavy oil. Presented at American Institude of Chemical Engineering (AIChE) annual meeting, San Francico, USA, 14 NovGoogle Scholar
- 19.Ng J, Gitlin I, Stroock A, Whitesides G (2002) Components for integrated poly(dimethylsiloxane) microfluidic systems. Electrophoresis 23(20):3461–3473. https://doi.org/10.1002/1522-2683(200210)23:20%3c3461:AID-ELPS3461%3e3.0.CO;2-8 CrossRefGoogle Scholar
- 21.Xiang C, Yang N, Zhang R, Guo J, Huang H (2017) An ultraviolet–visible (UV) photometry system based on the PDMS-based microfluidic chip. Presented at 13th global congress on manufacturing and management (GCMM 2016), Zhengzhou, China. https://doi.org/10.1051/matecconf/201710003036 CrossRefGoogle Scholar
- 22.Khanafer K, Duprey A, Schlicht M, Berguer R (2009) Effects of strain rate, mixing ratio, and stress–strain definition on the mechanical behavior of the polydimethylsiloxane (PDMS) material as related to its biological applications. Biomed Microdevice 11(2):503–508. https://doi.org/10.1007/s10544-008-9256-6 CrossRefGoogle Scholar
- 23.Wang Z (2011) Polydimethylsiloxane mechanical properties measured by macro-scopic compression and nanoindentation techniques. PhD diss., University of South FloridaGoogle Scholar
- 33.Du Y, Xue Y, Frisch HL (1996) Physical properties of polymers handbook. AIP Press, WoodburyGoogle Scholar
- 34.Ougizawa T, Inoue T (1999) Miscibility and interfacial behavior in polymer–polymer mixtures, in polymer blends and alloys, chapter 3. Marcel Dekker Inc, New YorkGoogle Scholar
- 40.Bruke J (1984) Solubility parameters: theory and application. AIC book pap, group annual vol 3, pp 13–54. https://cool.conservation-us.org/coolaic/sg/bpg/annual/v03/bp03-04.html. Accessed 20 May 2017