Dialysis System Update
Henry and I just got back from our meeting with Dean over at RIT to talk with him about the production process for making our proposed flow dialysis system. While we were there Dean described in detail the entire production process including the anodic bonding and the parylene deposition. We also returned with membrane samples which have been anodically bound to the glass cover slips. I think there was some miscommunication at the meeting last week. When reported his pressure tests the pressure reported were the pressure needed to force water through the capillary tubes they use in their systems. At the time we thought these values were for fluid flow through an assembled membrane-cover-slide system (this prompted Dave’s question about where the membrane was placed face down or face up during testing). No such tests were preformed, and since the capillary tubes we will be using are different material we will have to retest the necessary pressure needed to induce water flow. Due to the larger inner diameter the pressure that will be need to drive the flow through the tube will be smaller.
Anodic Bonding:
-The glass “cover slips” used are composed of borofloat glass. This is necessary because the borofloat and the membrane chip have the same thermal expansion coefficient ensuring that once the anodicaly bound membrane cools it does not shatter the glass.
-The anodic bonding system they have is homemade, consisting of a 4V DC power supply which feeds a high voltage source that outputs 700V. This is fed into a simple circuit board which leads to two capacitor plates. The simple circuit board mainly consists of two resistors one in series with one of the plates to limit current flow, and the other in parallel between the plates to allow the plates to safely discharge.
-One note Dean made is that when the pnc-Si membrane chip and the glass cover slip is sandwiched between the two metal plates they sometimes slide slightly, meaning that the two pieces might not be fully aligned.
Parylene Deposition:
-Dean has observed that the parylene can travel a distance of 250um down between the channel made for the tube and the tube itself (the gap is roughly 20-40 um). Based one Deans suggestion the longer the channel the better to prevent the parylene from entering the inner cavity of the membrane thus contaminating the membrane.
-The mean free path of parylene is dependent on temperature and pressure. Since this process is done at room temperature we could try low temperature deposition or increase pressure during deposition, either one would decrease the monomers mean free path. Currently Dean deposits parylene at 35mtorr and room temperature during his experiments.
-the deposition system runs as follows: 1. The parylene in grabular form were placed into an aluminum boat and sent into the system. 2. the pressure in the sample (to be parylene vapor treated) holding chamber are then held at 35 mTorr. 3. a cylinderical chamber were heated to 690 degrees Fahrenheit. 4. parylene in the aluminum boat were then vaporized and flown into the cylindrical chamber, where the parylene dimers (in its vapor form) were breaking down into monomers through the heat. 5. a cycle of heating, cooling, and pressure adjustment were then initiated to provide a constant feeding of parylene monomer vapor into the sample holding chamber with the desired pressure.
-to cap the end of capillary tubes Dean used aluminum foil and tape
-typically, 1 gram of parylene can coat the standard 4″ wafer to a thickness of 1 um.
How did the anodic bonding go? Does it look like a good seal? Did the membranes survive?
Also, what pressure do you think this is going to work at? At the meeting that I went to, David would not give a number.
Dean gave us the test membranes that he bound to the glass shards and there is not a 100% seal on any of the three samples. He did mention that he was not as rigorous (times ranged from 10 – 30 min) with the duration of the anodic bonding stage as he would have been if this had not been a test run. He also mentioned that the areas were the membrane chip did not successfully bind to the glass might be a result of the chips not being clean enough. He suggested that we might want to do a cleaning step with the chips prior to anodic binding (he suggested something like a piranha solution).
*I have these samples in the lab if you want to seen them
As for the membranes surviving, Deal did not seem very sure. The samples that were given to him were not all fully intact and he forgot to take picture of them before anodic bonding.
The anodic bonding process occurs on a bench top open to the atmosphere while the parylene deposition is currently done at 35mtorr.
As for the pressure needed to get water to flow through the capillary tubes Dean reported only “a few psi” are needed. But the tubing that he tested is not made of the same material that we are going to be using. I ordered some fused silica capillary tubing (that should be in today or tomorrow) that has hydrophilic properties hopefully allowing us to fill them through simple capillary action. Once these come in we can get some pressure numbers posted.