Hydrophobization of Inorganic Oxide Surfaces via Siloxane Equilibration

Kaleigh Ryan
Dr. Krumpfer
March 21, 2017

Blog #3

In the past couple of months, we have continued to investigate the reaction of silica (SiO2) wafers and hexamethylcyclotrisiloxane (D3) through the vapor phase. To do this, silica wafers were exposed to D3 vapor at 100°C and 150°C and then analyzed by ellipsometry (thickness) and contact angles. To better understand the reaction, we have performed kinetics at both of these temperatures by reacting them at different time intervals ranging from 15 minutes up to a week. The results were not completely surprising, as thickness grows rapidly in the first few hours (See Figure 1) and settled to thicknesses of approximately 11 nm after 72 hours. Additionally, the contact angles reach a minimum hysteresis (difference in values) around the same time with advancing contact angles of 105° and receding contact angles of 100°. This suggests that reaction is near completion after 72 hours under these conditions. While the two graphs below do not show it, each data point is the average of at least 3 samples, or 21 measurements, and so have statistical significance.

Figure 1. Kinetics of the reaction between D3 vapor and silica interfaces at 100 °C. The right graph shows the increase in thickness measured by ellipsometry while the left graph shows changes in the advancing and receding contact angles.

The above results were important in helping to understand the mechanism of this reaction. First, the low contact angle hysteresis tells us that the surface is almost completely covered and very smooth, since hysteresis increases with partial coverage and roughness. The ellipsometry results were at first puzzling. Why should the surface grow to about 17 nm and then suddenly get smaller again? How could the surface made be larger than the starting molecule? However, this would make sense if the layer was actually the ring-opening polymerization (ROP) of D3. Therefore, it would keep growing straight up until it falls over and can’t react anymore. Figure 2 shows a proposed mechanism, which is consistent with our experimental results.

Figure 2. Proposed mechanism for the reaction between D3 vapor and silica surfaces

We decided to also test the kinetics of this reaction at 150 °C. It was expected that the reaction would reach completion much faster, since reactions have faster rates with increasing temperature. From the data in Figure 3, thickness and contact angles reach the same values as they did at 100 °C, but they do so much more quickly. When the reaction was run at 150 °C, completion was reached after 6 hours with an average thickness of 10.5-10.7 nm, proving what I had anticipated. The average contact angles also reflected that of the 100°C reaction, average advancing contact angle being 101.2° and average receding contact angle being 96.6°. What is interesting about these data is that there is no “bump” in the thickness. This means the reaction occurred too quickly for the chains to grow as high, and likely fall over very early in the reaction.

Figure 3. Kinetics of the reaction between D3 vapor and silica interfaces at 150 °C. The right graph shows the increase in thickness measured by ellipsometry while the left graph shows changes in the advancing and receding contact angles.
The work done so far was able to result in the determination of this mechanism, which I see as a huge success, and the only challenge I have faced recently is time management. In the future, I would like to investigate this reaction using different cyclic siloxanes. It is expected that the ring-strain (how much it wants to open) will have a direct effect on the surfaces prepared.
Additionally, I was able to participate in two conferences over the last few months. The first was the Annual Dyson Society of Fellows Meeting. The second was the national PittCon conference in Chicago. Both were very rewarding experiences. I love the feeling of being a part of a larger community that epitomizes hard work and dedication that both of these events portrayed. At the Dyson Society meeting, I was able to see other students work and it was really impressive to see what students can accomplish with the help of funding from this program. PittCon was the first event I ever attended that encompassed not just work from fellow students, but also in the professional field. At PittCon, I also got to talk to people who were genuinely interested in my reactions and how they could use them. The following is an image of the board I presented at PittCon- seeing my work all come together the way that it did was such a rewarding experience.

Hydrophobicity of Cyclic Siloxanes

The first variable I altered in trying to find the optimal time and temperature to make a surface hydrophobic using cyclic siloxanes was time. I allowed the reaction to run at 9 different time intervals: 15 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, 24 hours, 48 hours, 72 hours, and 168 hours. I had hypothesized that the longer the reaction was allowed to run, the more hydrophobic the silicon surface would become, and the thicker the siloxane layer coating the silicon would become. This set of reactions was done at 100°C with about 150 mg. of hexamethylcyclictrisiloxane present in each vile.

After running these reactions and gathering data through ellipsomerty and goniometry, I discovered that generally speaking, my hypothesis was right in regards to the change in hydrophobicity. The 15 minute reaction resulted in an average advancing contact angle of 68° and an average receding contact angle of 63°. By the time the reaction ran for 6 hours, the hydrophobicity increased to new advancing and receding contact angles of 96° and 89°, respectively. The additional 162 hours the reaction proceeded for only resulted in a 10° increase for both advancing and receding contact angles. What this data tells me is that most of the increase in hydrophobic qualities occurs within the first 6 hours under these conditions, and gradually plateaus soon after.

The results gathered regarding the change in thickness of the siloxane layer was different than what I had hypothesized. What I found is that, over the first 72 hours, the thickness has sharp increases and decreases. The thickness changed from 9.63Å to 3.27Å in 15 minutes, than went up to 7.67Å in just an additional half hour. This rise and fall in thickness continued until the third day, and after 72 hours there was a steady decline in thickness, which began to plateau.

The next variable I would like to alter after gathering this data is the temperature of the oven in which the reaction takes place. I am interested to see how the patterns from this experiment being performed at 150°C for the same time intervals will compare and differ to the patterns I found at 100°C.

Hydrophobic Qualities of Cyclic Siloxanes

In this project, I am reacting cyclic siloxanes with inorganic oxide surfaces via ring opening polymerizations at solid-vapor interfaces. A siloxane is composed of silicon, oxygen, carbon, and hydrogen atoms. The skeleton of siloxane is made up of silicon and oxygen, and the other two bonds that silicon can make are each with a methyl group (CH3). Depending on the number of siloxane groups present in the cyclic molecule, it will have either high or low ring strain. Ring strain, or instability of the molecule, will exist when the angles in ring are less than 112°. The higher the ring strain, the more easily the ring will open and therefore be able to polymerize. Hexamethylcyclotrisiloxane (D3) is the cyclic siloxane I am using in this project, and it has highest ring strain of all cyclic siloxanes. As a result, I expect it to open easily and bond with the silicon surface. My intention is to investigate the effects of time and temperature on the reactions of D3 with silicon surfaces to create the most hydrophobic (water repelling) surfaces.

There are many surfaces that we come across in daily life that would be most effective if they were hydrophobic, such as windshields, surfaces of electronics, and cooking materials. In this reaction, I am using silicon as the solid that the siloxanes will bond to, but in principle, the polymerized siloxanes can be used to coat any inorganic oxide-containing surface. Some examples of this would be aluminum, tin, and nickel. An advantage of this technique is that it can create conformal films on textured surfaces. In other words, it bonds to the surface evenly and will maintain the texture of the surface. I’m initially hoping to find the best time and temperature to run the reaction at to create the most hydrophobic siloxane coating through polymerization of hexamethylcyclotrisiloxane. I then intend to see how using different cyclic siloxanes would change these variables and results. I expect to see significant effects from the time and temperature of the reaction on the quality of the hydrophobic surfaces I prepare.

In order to find the best parameters to run the reaction at, I am performing many trials of the same reaction, adjusting one variable at a time. In all of them, a silicon disc will be the source of the inorganic oxide surface and the siloxane will be in vapor phase during the reaction. I’m starting off with hexamethylcyclotrisiloxane as the source of the siloxane polymer and 100 °C as the temperature to run the reaction, as well as keeping the size of he silicon discs and quantity of the cyclic siloxane uniform. The variable first being altered is the length of time the reaction will run for, starting with 15 minutes and ending at 1 week. In order to measure the results, I’m using two different techniques: ellipsometry and dynamic contact angle analysis. Through ellipsometry, I can measure how thick a siloxane coating is, and through contact angle analysis I’m able to measure what the contact angles of a droplet of water are on the surface, both advancing and receding, to determine its hydrophobicity. Through these two techniques, I can gain much information about the nature of the chemistry of the surfaces I am analyzing.