Plasma Surface Modification of PDMS-Glass Microfluidic Chips for Oil Recovery Studies

Abstract

Wetting hysteresis is the most important characteristic of microfluidic chips for modeling multiphase flows in rocks, including for oil production problems. Plasma modification of surface wetting characteristics is well studied, but there is a problem of stabilizing the resulting surface for use in a liquid hydrocarbon media. In this work, a simple and accessible technology for modifying the surface of PDMS and glass using a dielectric barrier discharge in a chamber based on the d’Arsonval apparatus has been developed. The surface wetting hysteresis for PDMS and glass was studied as a function of the plasma treatment time. It is shown that with the help of plasma treatment it is possible to change the wetting angles of the walls of microfluidic chips in a very wide range, thereby simulating the conditions of both hydrophobic and hydrophilic rocks. At the same time, PDMS has the widest possible range of changes in the wetting angle; the advancing contact angle decreases from 120° to 10°; receding contact angle—from 70° to 0° during plasma treatment. It has been shown that plasma treatment of a microfluidic chip, together with a 30 min primary extraction with oil and salt water, leads to a significant change in the wetting characteristics of its surface. This in total leads to an increase in the oil displacement efficiency from the chip by about 10%. In general, the results of the study showed that plasma surface treatment for the hydrophilization of microfluidic chips is a simple and affordable technology for controlling the wetting characteristics of microfluidic chips. PDMS in this case is a promising material.

1. Introduction

Capillary fluid flows are widely used and studied in microfluidic devices, which are structures with one or more capillaries that perform certain manipulations with the flow of liquids or gases in them. For example, in microfluidic devices, one can obtain an emulsion [1], separate cells in a stream according to certain characteristics [2], conduct enzyme immunoassay [3], and much more [4]. The most popular manufacturing technology for microfluidic devices is lab-on-a-chip. Microfluidic devices made using this technology are referred to as “microfluidic chips”. A distinctive feature of microfluidic chips is that the capillary network is located in one plane—at the junction of two hermetically connected plates, on one of which (sometimes on both) a deep channeled pattern is formed, called the microfluidic chip topology. The walls of this pattern are the walls of the capillary network. Most often, transparent materials are used for the manufacture of microfluidic chips, since they allow visual control of processes in flows, as well as quantitative measurements by optical methods.

Microfluidic chips are made of glass [5], thermoplastic polymers [6], and elastomers [7]. The channeled pattern is formed by soft lithography [8], laser ablation [9], mechanical engraving [10], and taking an impression from a previously made master mold [11,12,13,14]. In addition to the formation of a channeled pattern, an important stage in the manufacture of microfluidic chips is their sealing, during which the topology of the chip is sealed and a capillary network is formed. Sealing is carried out by gluing with adhesives [15], solvents [16], thermal sintering [17], and chemical crosslinking [18,19,20,21]. Often, before sealing, the surfaces to be joined must be activated using plasma or ultraviolet light [22,23,24]. When sealing, a change in the geometric characteristics of the chip topology can occur. Thus, the choice of material is due to the presence of both technologies for it: micromachining and sealing.

In addition to the geometric properties of the capillary network of chips, the chemical and physical characteristics of the surface of the capillaries are of key importance: the chemical composition of the surface and wettability. For many microfluidic applications, the natural surface properties of the materials from which microfluidic chips are made are not suitable. For example, certain biochemical groups must be attached to the surface of enzyme immunoassay test strips to ensure the specificity of analyte binding in order to increase its local concentration and the possibility of visual detection [25]. Another example where the physical characteristic of wettability is important is the simulation in porous chips of multiphase flows that are observed in rocks (for example, when studying oil recovery during reservoir flooding). In such chips, it is necessary to accurately recreate the wetting angle of the capillary surface similar to that observed on natural rock grains [26]. It is known that in nature the wettability of natural rocks dominated by silicates is primarily determined by primary extraction [27]: if most nano- and micropores are filled with water, then the surface will be hydrophilic; if the pores are filled with non-polar liquids, then such a surface will be hydrophobic. Polydimethylsiloxane (PDMS), from which chips are made for modeling flows in rocks [28], is a porous material with a contact angle of about 120°. With this initial contact angle, the primary water extraction does not lead to an increase in hydrophilicity, since water does not enter the pores of the PDMS.

It is clear that the wettability of a surface depends on its chemical composition. Therefore, to change the wettability, the surface of the capillary walls undergoes modification, during which the chemical composition of the thin surface layer changes. Such modification is carried out by photochemical methods [29], surface lamination with scaffold [30], treatment with active solutions of alkalis [31], acids [32], and alcohols [33], as well as by plasma-chemical methods [34].

However, plasma chemical modification often results in an unstable surface that contains many polar radicals. These polar radicals make the surface more hydrophilic, but also more chemically active. Crude oil components can react with the plasma-activated surface, making it hydrophobic again. Additionally, the mobility of low molecular weight chains, which also impair hydrophilicity [35], can increase significantly in the oil environment.

The aim of this work was to develop and study a simple and cheap method for crude oil-resistant hydrophilization of the surface of microfluidic chips.

2. Materials and Methods

2.1. Materials

To measure the contact angle, we used distilled water, 3 µS/cm, GOST 58144-2018. Plates for measuring the contact angle were made of Sylgard 184 PDMS (Dow Europe GmbH, Wiesbaden, Germany) and 0.18 mm thick cover slips (MiniMed LLC, Bryansk, Russia); microfluidic chips were made from the same PDMS and glass slides 1–1.2 mm thick CAT.No 217102X (Citotest Labware Manufacturing Co. Ltd., Haimen, China). The chips were filled with crude oil with a viscosity of 17.8 MPa∙s and a density of 852 kg/m³. To modify the surface, we used an aqueous solution composed of high-purity sodium chloride art. 533000 (Lenreaktiv LLC, Saint Petersburg, Russia) in a sample of 100 g/L.

2.2. Plasma Laboratory Instrument and Chip Plasmaization

To modify the surface, a plasma chamber was made (Figure 1), in which a dielectric barrier discharge was ignited. The design of this chamber was based on a typical scheme for the industry of sealing chips from PDMS: a grounded cathode is placed inside a vacuum flask with a glass wall, outside of which an anode is placed, to which voltage is pulsed, which leads to ignition of the plasma in the flask between the anode and cathode. To create a high voltage, we used the d’Arsonval apparatus DE-212 Karat (SMP LLC, Moscow, Russia) (Figure 2a). The plasma flask was made from a 175 mm diameter microwavable glass pan with a wall thickness of 5 mm. A copper anode 1 mm thick was glued to its cover from the outside using KER-215 and ESO-15 epoxy resin (Himalyans LLC, Nizhny Novgorod, Russia), with a screw soldered in advance for connecting a high-voltage current source. The design involved inserting a cathode inside the chamber (a 2 mm thick stainless steel plate) and placing a fitting for air evacuation. To avoid mechanical processing of glass, it was decided to make the lower part of the pan by extrusion printing from PETG plastic (Bestfilament LLC, Tomsk, Russia) using the existing Hercules 3D printer (Imprinta LLC, Krasnoyarsk, Russia). Elite Double 22 silicone (Zhermack S.p.A., Badia Polesine, RO, Italy) was used to manufacture the sealing ring and seal the threaded connections. Air was evacuated from the chamber using a Laboport N820 membrane vacuum pump (KnF Technology Co. Ltd., Shanghai, China) to a pressure of 5 kPa.

Plasma treatment of PDMS and glass plates took place at a temperature of 25 °C in humidified air. To do this, the sample was placed on the cathode with the treated side up. Water was poured to the bottom of the chamber under the cathode, the chamber was closed with a lid with an anode, and a vacuum pump with a vacuum manometer was started. Humidification of the atmosphere in the chamber was carried out by natural evaporation of water. When the pressure in the chamber was reached, at which the water boiled, the source of high-voltage pulsed voltage was turned on, the plasma was ignited in the chamber, and the processing time was recorded. After the time had elapsed, the plasma setup was turned off and air was let into the chamber to open it and remove the sample.

2.3. Measuring the Contact Angle of Advancing and Receiding

After plasma treatment of the sample, the wettability of its surface was studied by the sessile drop technique. To measure the contact angle hysteresis, the droplet volume was changed by adding and removing water. Observations were recorded using a camera whose optical axis was at a small angle to the plane of the sample.

2.4. Problems and Solutions

Preliminary observations showed that when large, thick dielectric objects are placed (PDMS plate 4 mm thick), the plasma goes around them and burns most intensively in the inner corners and along the edges of the object (Figure 2).

Therefore, the samples were made from thinner PDMS plates that were 2 mm thick and had a square shape with a side of ~1 cm thickness. In order to remove the edge effects of plasma flow around sharp corners, the edges of the material were cut off (Figure 3).

2.5. Topology and Chip Fabrication

The chips were made from PDMS by pouring into a silicon master mold. The channelized pattern on the master mold was made by laser lithography and reactive ion etching to a depth of 40 µm. The chip topology was a single channel form with pressure-controlled flow (Figure 4). A microfluidic chamber 7.6 mm wide and 20.5 mm long was formed in the central part of the channel. The inlet and outlet of the chamber were equipped with parallel laminators to create a straight fluid flow front. To imitate rock grains, an array of square columns with a side of 100 µm was formed in the chamber, located at a distance of 100 µm from each other.

Pouring PDMS into the master mold was carried out with a layer about 2 mm thick. After polymerization of PDMS, the chip was removed from the master mold and holes were made in it for connecting tubes. For sealing, the chip and the glass slide were processed in a plasma chamber for 2 min, after which they were pressed against each other and kept for 5 min at a temperature of 130 °C to form covalent siloxane bonds between the glass and PDMS.

2.6. Chip Surface Modification

The modification of the chips was carried out immediately after their sealing, for which the surfaces were subjected to plasma treatment for 120 s. Chip modification was carried out using primary extraction with oil and salt water with an exposure of 30 min. To improve the quality of water extraction, the chip after sealing and before injection of salt water was additionally processed in a plasma chamber for 4 min. After exposure to salt water, the chip was washed with distilled water, after which it was filled with oil for subsequent displacement studies.

2.7. Oil Displacement

The oil-filled chips were connected to an SPLab02 syringe pump (Baoding Shenchen Precision Pump Co. Ltd., Baoding, China) with a 1 mL 1001LT glass syringe (Hamilton Company, Reno, NV, USA) and pumped with water at a rate of 3 μL/min for 30 min. The pressure was recorded using an MPS 100 psi pressure sensor (Elveflow, Paris, France) mounted on a fluoroplastic inlet tube with an inner diameter of 1/32 inch, and an OB1 MK3+ flow controller (Elveflow, Paris, France). Displacement observations were made using a DSC-RX100M4 camera (Sony, Tokyo, Japan). After extrusion, the chips were examined under a Micromed 2 microscope (Micromed LLC, Saint Petersburg, Russia) with an FMA050 camera (Hangzhou Touptek Photonics Co. Ltd., Hangzhou, China).

2.8. Photometric Estimation of Oil Displacement Efficiency (K_dis)

Photos of chips with oil emulsion were subjected to pixel-by-pixel processing. To do this, all image pixels were converted from the RGB palette (R—red, G—green, B—blue) to the HSV palette (H—tone, S—saturation, V—value). From the photo of the chip with oil, pixels were determined with the values of the S and V numbers corresponding to the dark color. Thus, the shape of the capillary network was set. Later, in the process of displacement of oil by water, the parameters S and V changed for some pixels, since the water is transparent, and the background behind the chip is white. K_dis was determined as the ratio of the number of brightened pixels to the total number of pixels that make up the image of the capillary network.

3. Results and Discussions

3.1. Surface Modification

3.1.1. PDMS

Initially, the effect of plasma treatment on the wettability characteristics of the PDMS substrate was studied. Figure 5 shows a photograph of water droplets on the original and processed substrate during the advancing process.

As can be seen, plasma treatment makes it possible to radically change the wetting characteristics. The dependence of the water wettability characteristics of the PDMS surface on the treatment time in the plasma chamber in a humidified atmosphere was obtained (Figure 6). It can be seen that the advancing contact angle decreases from 120° to 10° in 400 s, and the receding one decreases from 70° to <1°. The data obtained indicate the possibility of changing the wetting properties of PDMS using electric plasma in a fairly wide range, from hydrophobic to superhydrophilic.

PDMS is a porous material, but its untreated surface contains mainly only methyl groups and, as a result, its surface is hydrophobic. That is, filling the pores of the raw PDMS is only possible with non-polar liquids such as oil or crude oil. In the plasma of a water-moistened atmosphere, methyl groups are replaced by methanol and hydroxyl groups, which are polar [23]. Nearby hydroxyl groups can form a vicinal complex, that is, they will have a common hydrogen and such a complex will have a charge.

We associate the decrease in advanced contact angles with the ionization of the surface: the higher the charge, the greater the distance from which water molecules begin to be attracted to the surface; receding contact angles are related to the adhesion of liquid molecules to the surface. The strength of this compound is largely affected by the number of polar molecules; ions, on the other hand, form layers of bound water around themselves, and the strength of the connection with the “large” drop will be due to the cohesion of water. Thus, between 60 and 120 s, we observe a slight increase in the receding angle and, at the same time, a decrease in the advancing angle. This may be due to the fact that some of the polar molecules formed vicinal complexes. That is, the number of polar molecules decreased (the receding angle increased), and the number of ions increased (the advancing angle decreased). These considerations require verification by modeling and analysis of the chemical composition of the surface.

3.1.2. Glass

Photos of water droplets on glass before and after treatment are shown in Figure 7.

The dependence of the wetting angles during the advance and recede of a water drop on the glass surface is shown in Figure 8. It can be seen that already after 1 min of treatment, the wetting angle of the advancing liquid drops from 80° to 10°, and the receding one from 55° to <1°. For 2 min of treatment, the advancing angle drops to 5–6° and does not change any more during further processing for 5 min. The experiment showed that, using plasma treatment, a superhydrophilic surface can be obtained for both angles in less than 30 s. We explain this by the fact that the number of hydrophilic groups increases rapidly during plasma treatment, which does not contradict the known data from XPS-data analysis [36].

Glass is mainly silicon dioxide, in which silanol-bound water, silanol, siloxane, geminal −OH and vicinal groups can be found on the surface. Siloxane groups predominate on the untreated glass surface. We explain the rapid hydrophilization of glass compared with PDMS by the fact that there are no hydrophobic groups on glass, unlike PDMS, and therefore the effect of the appearance of even a small amount of such groups can be much more pronounced.

3.2. Modification of Chip Capillaries

As noted above, PDMS, from which the chips were made, is a porous hydrophobic material. After plasma treatment, polar groups are formed on its surface, which have a short lifetime, since they are chemically very active and can react with crude oil components. To stabilize the surface, we used salt water, which forms Helmholtz adsorption layers and thus can protect the modified surface while maintaining its hydrophilicity. In addition, we assume that the water fills the pores of the PDMS and makes them inaccessible to oil wetting.

To test the performance of the technology, two chips with the same topology were made from PDMS and glass, imitating the porous structure of rocks, and they were hydrophilized according to the method described in the previous section. Both chips were filled with oil and this oil was displaced by water, thus simulating the process of development by waterflooding of oil-saturated rocks with conditionally hydrophobic and hydrophilic properties. Observations of the displacement process (

Table 1. Patterns of oil displacement in chips depending on the duration of this displacement.

Time, min	Chip Primary Extracted in oil (O) *	Chip Primary Extracted in Water after Plasma Treatment (W) *
1
5
10
20
30

* In all photos, the pumping occurs from left to right.

) showed that the W chip after washing contained less oil than the O chip. At the same time, in the first minutes of washing, many thin, fast water jets formed at the beginning of the W chip chamber, which later merged into one wider one. Over time, the channel of the water jet expanded due to the displacement of oil from the far part of the chip, but many oil “islands” at the beginning of the chip did not move. In chip O, a different displacement pattern was observed: the front of the water flow was even, and the intensive expansion of the resulting water channel occurred in the part of the chip closer to the inlet channel; fewer oil “islands” were formed, and they were smaller.

After processing the obtained data, the dynamics of the oil displacement efficiency for both chips were obtained (Figure 9). It can be seen that the displacement of oil occurs most actively in the first minutes of the experiment. In chip W, the initial water jet broke through the oil layer faster (compared with chip O) and had a smaller width, thus forming a channel in a shorter period of time. This resulted in a comparatively lower K_dis in the first 5 min of displacement. However, the boundaries of the formed channel in the W chip further expanded faster, which led to an increase in K_dis. Thus, it was shown that plasma treatment of a PDMS microfluidic chip followed by primary aqueous extraction leads to an increase in the oil displacement efficiency from the chip by about 10%. We explain this by the fact that plasma treatment makes the nanopores of PDMS water-wettable, and thus available for filling with water. Because of this, hydrophilization of the surface, which is resistant to crude oil, occurs. To obtain quantitative characteristics of the effect of PDMS plasma treatment on the volume and depth of water absorption into the surface, an additional study of the surface topographic characteristics, including the analysis of XPS data, is necessary.

After the oil was displaced by water, the chips were examined under a microscope (Figure 10). The meniscus of the oil remaining between the columns shows that the wetting pattern in the W chip is significantly different. Although, as can be seen in general, the surface remains hydrophobic. In the O chip, oil slicks are visible in the water channels adhering to the lower and upper walls of the chip. In the W chip, such spots are either not observed, or they had a less pronounced color, which indicates a significantly smaller thickness of this drop compared with those observed in the O chip. Additionally, in the W chip, a significant number of columns are observed on all sides surrounded by water; in chip O, all columns have at least one oil bridge with neighboring columns. The thickness of the oil film enveloping the pillars in the W chip is also much smaller. All of this indicates hydrophilization of the microchip surface after its treatment in plasma and extraction in salt water.

4. Conclusions

A simple and accessible technology for modifying the surface of PDMS and glass using a dielectric barrier discharge in an air atmosphere saturated with water vapor at a temperature of 25 °C and a pressure of 5 kPa has been developed. The wettability characteristics (advance and recede angles) are measured experimentally as a function of plasma treatment time for one of the most common materials used in microfluidic technologies: PDMS and glass. To test the efficiency of the technology, microfluidic chips were made to study oil recovery and hydrophilization of their surface was carried out. The effect of hydrophilization of the surface of microfluidic chips on the pattern of two-phase flow in them during oil displacement has been studied.

As a result, the following conclusions were obtained:

• It is shown that with the help of plasma treatment it is possible to change the wetting angles of the walls of microfluidic chips in a very wide range, thereby simulating the conditions of both hydrophobic and hydrophilic rocks;
• PDMS has a very wide contact angle range from hydrophobic to super-hydrophilic: the contact angle decreases from 120° to 10°; receding angle—from 70° to 0° during plasma treatment;
• Glass in its original state is a hydrophilic material. It is shown that even short-term plasma treatment makes it possible to achieve superhydrophilicity of the glass surface. For 1 min of treatment, the contact angle decreases by more than 50º. This material is recommended for use at high pressures and relatively high temperatures due to its high strength, chemical resistance and the possibility of plasma-chemical surface modification;
• The study of the processes of oil displacement from hydrophilized chips showed that the patterns of water-oil flow in the treated and untreated chip differ significantly on the macro- and microscales. Despite the fact that the overall surface of the chips remains hydrophobic. It was shown that plasma treatment of a PDMS microfluidic chip leads to a significant change in the wetting characteristics of its surface. This in total leads to an increase in the oil displacement efficiency from the chip by about 10%.

In general, the results of the study showed that plasma surface treatment for the hydrophilization of microfluidic chips is a simple and affordable technology for controlling the wetting characteristics of microfluidic chips. This is a very important aspect, since the inability to obtain wetting characteristics close to real rocks is one of the significant issues that are often presented to microfluidic technologies for studying oil recovery. The presented method of surface modification makes it possible to partially remove these limitations. Additionally, for these purposes, PDMS is one of the most promising materials for modeling rocks, since it is porous at the nanoscale and, similar to a natural rock, its wettability depends on the primary extraction.