Sunday, September 20, 2009

Viologen-Mediated Direct Carbohydrate Fuel Cell 1

Dear Dr. Dean Wheeler,

My name is David Brosnahan and I was a undergraduate research assistant with Dr. Gerald Watt from 1997 to 2000 and a graduate student from 2000 to 2002. After completing a masters degree in biochemistry, I went on to medical school and on to complete an emergency medicine residency. However, today while surfing the web, I came across the following abstract regarding Methyl Viologen and Glucose Fuel Cells at the website for an 2008 electrochemical meeting in Japan (see link below). I saw your name first of the abstract and I wanted to give you some background into how Dr. Watt got involved with this project.

Dr. Watt had two main projects. His primary project was Nitrogenase. He wanted to understand how Nitrogenase worked so he could replace the inefficient Haber-Bosh process to fix nitrogen into ammonia. His secondary project was ferritin, which is involved in iron metabolism in the body. When I started with the lab in 1997, our lab had national grant funding for both projects. But a year later, the government agency funding our nitrogenase research categorically withdrew funds in this area. Several years after that our ferritin funding ran out. We had been in a publishing drought because we had the correct but conflicting result with another lab who held the political and publishing high ground. The lean years that followed in the Watt Lab were very frustrating but resulted in our expanding our research into new areas. We had to evolve, publish, and compete for grants or die. The best way I and Dr. Watt saw to becoming competitive was to develop projects in the areas of nano-technology and fuel-cells. These were the emerging fields then as they still are today. Before I left,we began to collaborate with with the ChemE department and with other in our department on the nano-battery project. I was pleased that Dr. Watt had the opportunity to take a sabbatical at Langley Research Center to develop his idea. It was fun to brainstorm at the beginning of the project how the heat-stable and self-assembling ferritin protein could be arranged to work as a nano battery.

It was about this time when issues or global climate change and energy independence were rising in the public consciousness when a mishap in the lab got me thinking about fuel cells. We were anaerobically separating the nitrogenase protein using ion-affinity column chromatography when another hydrogenase enzyme in the mixture together with sodium dithionite catalyzed the electrolysis of water and the evolution of large amounts of hydrogen which completely interrupted the separation and the column. Upon observing the hydrogen evolution I kinda had an ah-hah moment and started thinking along the lines of a microbial or enzymatic (hydrogenase) fuel cell that would evolve hydrogen from dithionite. However, the idea of continually having to ferment bacteria, over-express and purify enzyme, and continually replace biological catalysts didn't seem to me like it was a system that would meet the global power demands of the planet (I could be wrong).

So, I started thinking about more direct methods to evolve hydrogen or directly reduce organic molecules. While working with Dr. Watt on a project related to the redox properties of ferritin and other metaloproteins, we routinely used a Coulometer to measure redox properties of redox-active metaloproteins. However, enzymes do not interact with metal electrodes directly, so Methyl Viologen or Paraquat was used to mediate the interaction. I thought, If methyl viologen could mediate between protein and anode, they maybe it could mediate between a hydrocarbon or partially oxidized hydrocarbon and the anode. Methyl viologen works because it its singly and doubly reduced states its heterocyclic, aromatic, conjugated, pi-electron system stabilizes a free radical state which then freely interacts with the anode. This is similar to how the body does it using biological mediators such as NADH, FAD, etc.

So, I started playing around in the lab and MacGyvered a table-top anaerobic fuel cell using methyl viologen added to various hydrocarbons. I tried hexanes, phenol, methanol, formaldehyde, and formic acid as well as glucose under various pH and temperature conditions. Glucose seemed to work the best. However, I didn't know at the time to what extent the glucose was being oxidized. Glucose is a reducing sugar and it is not difficult to get an electron pair out of it. I had tried to measure evolved carbon dioxide without success using GC. However, I stumbled upon the following abstract for a 2008 electrochemical meeting (see link below) and I was excited that your abstract mentioned that you and Dr. Watt had fully oxidized glucose to carbonate. I am excited that someone from the BYU ChemE Department took an interest in this problem. Before leaving BYU, I unsuccessfully tried to get Dr. John Harb involved with the project. However, I think Dr. Harb was more interested in collaborating with Dr. Watt on the ferritin nano-battery project then the fuel cell idea.

The purpose of this email is to verify that you are working on the MV fuel cell project. And if it is, I just wanted to let you know where I left off, and some of the problems I had talking this project any further. As you already alluded to in your abstract, designing a fuel cell which runs on glucose would be a historic breakthrough. Glucose derived from switch grass and other cellulose-based sources is 100% renewable. In addition, as you alluded to in your abstract, designing a fuel cell without needing gold or platinum in the reforming system, anode, or membrane (PEM) is a major hurdle to making fuel cells a viable answer to global energy needs and energy independence here in the US.

However, the major hurdle I faced was developing a fuel cell with adequate current per area. I could get 0.5 V short circuit voltages, but no milliamps. And this is because the reaction in solution is rate limited by the concentration of the MV. For a current to be generated a molecule of glucose would have to collide with a MV molecule which would then need to collide with the anode. That would result in a second order reaction dependent on a small concentration of MV in solution. However, if the MV mediator could be somehow directly associated with the anode, then the reaction would become first-order and dependent only on the concentration of glucose in solution. And glucose is very soluble. It seems from the abstract that you and your team figured out how to make this work. I couldn't think of a way to make a MV anode. I didn't know how to create a MV thin film, or polymerize MV. I discussed with Dr. Harb talking solid MV salt and mixing it straight with plasticizer and seeing if that worked. According to the chemical equations, I'm not sure if water was involved in the redox reaction. Looking up in Google, there are several groups that have discovered how to polymerize MV. If the polymerized MV is conductive or can be directly associated with an anode material, this may be a breakthrough.

Also, one more interesting characteristic of MV is that it can be doubly reduced into a quinoid form is soluble in organic solutions. Both the singly and doubly reduced MV forms carry out the electrolysis of water and form hydrogen gas using a variety of catalysts and photo-catalysts.

I apologize for the length of this email. I would appreciate any reply or information on the progress of this project. I am 100% into medicine now, but like to follow the progress of the research I was once briefly a part of.

David D. Brosnahan

1. MV thin film patent:
2. MV doped silicate anode and methanol Optical Materials Volume 22, Issue 3, May 2003, Pages 221-225 Electroreduction of methyl viologen in methanol and silicate thin films prepared by the sol–gel method
3. MV polymers Journal of Polymer Science: Polymer Chemistry Edition Volume 21 Issue 1, Pages 293 - 300 Preparation of viologen polymers from alkylene dipyridinium salts by cyanide ion
4. MV vs. MV polymers and electroysis J. Chem. Soc., Faraday Trans. 2, 1982, 78, 1937 – 1943Viologen/platinum systems for hydrogen generation
5. MV nanoparticles Electrochimica Acta Volume 53, Issue 26, 1 November 2008, Pages 7655-7660 Electrochemical formation of viologen nano/microsized wires and tubes by potential sweep technique combined with micellar disruption method
6. MV polymer Bioelectrochemistry Volume 60, Issues 1-2, August 2003, Pages 57-64 Voltammetric and spectroelectrochemical characterization of a water-soluble viologen polymer and its application to electron-transfer mediator for enzyme-free regeneration of NADH

Viologen-Mediated Direct Carbohydrate Fuel Cell 2

Dear Dr. Dean Wheeler,

Thank you so much for sending the power point presentation. I thoroughly enjoyed every slide. I was very excited to see that you were able to measure respectable current densities. Several of your slides actually resolved some questions I had had at the time. I especially appreciated the Pourbaix diagram. I just wanted to pass on some more background info that came to mind after going through your power point presentation.In my first email, I described unsuccessfully measuring CO2 evolution by gas chromatography (GC). I understood that CO2 would not exist in significant amounts at basic pH but it was nice to see your potential/pH diagram map out the various phases in this electrochemical system. After seeing your presentation, I had a few additional ideas as possible models to characterize the chemistry better.
When I was playing around with MV and glucose, I was routinely using high-Molar and saturated glucose solutions and a limiting amount of MV. Using Dr. Watt's Schlenk line/manifold, I made a basic, aqueous glucose solution anaerobic and then added an anaerobic solution of MV. The solution immediately turned deep blue. I see that you also discovered that the reaction occurs with both reducing and non-reducing sugars. I think if I had acidified the solution, any carbonate would have been converted to CO2, which I could have then injected into a Hewlett Packard-CG and been able to make a quantitative estimate of the extent of the reaction.

I also noted that you monitored the MV-glucose reaction using a pressurized cell and indirectly calculating the O2 consumption based on pressure differences. A simple variation on this theme would be to measure O2 directly. Dr. Watt had a direct electrochemical dissolved-oxygen probe in the lab. The probe worked by covering the end of the electrochemical probe with a cellulose acetate film with a concentrated KCl solution between the film and the end of the probe. I believe there are several much simpler oxygen concentration probes on the market today.

What I did was created an ambient-pressure reaction cell using two pieces of mortised Plexiglas bolted together which created a central reaction well similar to your reaction apparatus. The oxygen probe fit snuggly into a bore hole drilled into one side of the block; the tip of the probe inserted into the central reservoir. Stop cock vacuum grease was used to make the probe insertion air-tight. Another deep but small-diameter hole was also cut into the Plexiglas. The small-diameter hole was just big enough to insert a long small-gauge syringe needle down into the central reaction reservoir. I found by leaving this deep bore hole open, I could normalize the reaction to atmospheric pressures but the bore hole didn't allow ambient oxygen to enter the reaction system to a significant degree. The bore hole was so deep and so small-gauge, that the diffusion of oxygen from outside was negligible (see figure).

The electrochemical oxygen probe with the cellulose-acetate membrane was very pressure sensitive; so this setup, which operated at atmospheric pressure, allowed me to add and remove small volumes of solution without altering the pressure of the system. And believe it or not, ambient oxygen didn't diffuse into the system. Part of this is because the rate of the diffusion of atmospheric oxygen into and out of solution is porportional to the exposed surface area. In my ambient pressure reaction vessel, the surface area exposed to the atmosphere was only the size of a small gauge needle.

Now that I think about it, Dr. Watt had a glass reaction vessel I found especially blown for the electrochemical oxygen probe. The peace of glassware was composed of inner and outer glass cylinders with 2 communicating ports from the outer to the inner cylinder. One port was sized for the oxygen probe and the second port was sized to fit a rubber stopper. Instead of a rubber stopper I used a cylindrical Plexiglas stopper that fit snug into this second port. This Plexiglas stopper was modified with a small-gauged bore hole drilled lengthwise; providing communication between the inner reaction reservoir and the outside.

This glassware piece had 2 more ports which emanated from the outer cylinder and opened up into the space between in the inner and outer cylinders allowing fluid circulation around the inner reaction reservoir for purposes of temperature control.I used this same cellulose-acetate membrane material used to cover the oxygen probe as a membrane to make a concentrated KCL salt bridge and as the membrane material for my fuel cell. I am not sure if the cellulose-acetate membrane allowed diffusion of MV and glucose molecules or not. I didn't have any Nafion so, it was the best I could come up with.When I used concentrated alkalinized methanol in the reaction, MV was reduced and the solution turned blue. But when the MV was exposed to oxygen again and allowed to oxidize, a yellow-orange species was produced. I am not sure what this species was. It may have been a MV dimer. We never characterized it. The more I reduced and reoxidized MV, the more orange the solution became. This yellow-orange species was not seen with glucose. Using glucose, aqueous MV could be reversibly reduced and reoxidized again and again without degradation.If you want to make doubly reduced MV, you add hexanes or other organic phase over water. You add Zinc powder to an aqueous MV and shake. doubly reduced MV diffuses and is soluble in the organic phase and forms a deep orange color.


David D. Brosnahan MD MS

Viologen-Mediated Direct Carbohydrate Fuel Cell 3

Dear Dr. Dean Wheeler,

I stumbled upon the abstract of your latest paper on on the viologen direct carbohydrate fuel cell in the J. Electrochem Soc. I am very interested how you solved the problem of directly associating the viologen with the electrode to make possible more significant current densities. I am also happy that you have demonstrated that carbohydrate is fully oxidized to formate and carbonate and the reaction is not just the one electron pair activity of a reducing sugar.

Other than paying 24$ is that any other way I can get a copy of the paper? Also, I am curious if there are any other groups working on this project and what kind of response the paper has generated. Also, I am curious if the paper has demonstrated enough of a proof of concept and reduction to practice that BYU is interested in a patent. At the time I spoke with them about it in 2001 after building the first viologen direct-carbohydrate/
formaldehyde/etc voltaic cells they were not interested. BYU seemed to me to be an unusually tough place to file a patent. It seems you have to have whatever packaged with buyers waiting before they take you seriously. Dr.Daniel Simmons found that out with the COX-2 work.

Also, it seems to me that the trick with the viologen electrode is that the viologen is able to stabilize the separation of electron pairs in the carbohydrate into radicals which then can then interact and be conducted by metal. My view has been that the reasons hydrocarbons and carbohydrates fail to interact with metal is the stability of the electron pairing and the inability of metals to conduct electron pairs. Biological physiology uses carbohydrates and fats to store energy and uses mediators such as NADPH and FAD, which act like viologen, to stabilize the single electron radical and single electron transfers.

So, in the case with viologen we are stabilizing single electron transfers and the separation of molecular electron pairing. On the other side, I wonder is a fuel cell could be made that would conduct electron pairs directly using a superconductor as an electrode. I don't know much about superconductors, and if cooper pairs in a superconductor are at all relate to molecular electron pairs. I have wondered if a cryogenic fuel cell could be made using liquid methanol and a superconductor. Is there a salt bridge or PEM that would work at those temperatures.

I am so excited that Dr. Watt was able to pass this project on. This project deals with, I think, the most significant and important scientific problem of our generation. The mediated direct carbohydrate fuel cell gets past many of the hangup of other fuel cell designs. A hydrogen economy has difficulties with storage and hydrogen leaking out of fittings as well as the conversion of current energy infrastructure. Hydrogen, reforming hydrogen, and direct methanol systems also has a problem at the platinum catalysts due to cost and CO poisoning. Enzymatic and bacterial fuel cell systems are just not robust enough for practical use. Direct mediated carbohydrate fuel cells is exactly how biologic systems do it. Cellulose and other waste organics could be used as a renewable source of carbohydrate fuel. Viologen or paraquat, or "roundup" is already produced and used worldwide (not currently in the US) as a common herbicide.

Due to the potential significance of this project to science, geo-politics, and humanity in general; I hope you can understand my continued interest in the project and I hope you are not bothered by an occasional email on this subject. While the ferritin nano-battery was originally Dr. Watt's idea this viologen-carbohydrate fuel cell was originally mine. But I never got past producing the first voltaic cell and associating the viologen with the electrode. But we did work to partially demonstrate that glucose and other carbohydrates were fully oxidized to formate and carbonate and that we were not just seeing the activity of a common reducing sugar.

I am grateful for your work on this project and wish you well in your future work.

David Brosnahan

Viologen Catalysts for a Direct Carbohydrate Fuel Cell J. Electrochem. Soc., Volume 156, Issue 10, pp. B1201-B1207 (2009)