Non-Poisoning effect in Fuel Cells Or How would you discover something?

The article figure presents the influence of occluded hydrogen on the electro-oxidation of carbon monoxide, CO to CO2 on palladium electrodes. The y-axis is the absorbance of the CO2 infrared band, the x-axis is the potential vs the saturated calomel electrode (SCE). The open circles is the blank. This is the electro-oxidation of carbon monoxide in the absence of occluded hydrogen on the palladium surface. As observed, it has an exponential grow. This means that the CO electro-oxidation depends on the available surface. The presence of increasing amounts of occluded hydrogen in the palladium matrix (closed circles and triangles) change this behavior completely. First, it shifted the oxidation to lower potentials, meaning it is easier to oxidize (burn) the CO. And second, it does not depend on the surface coverage. This is a non-poisoning effect on the palladium catalyst! In turns, this can potentially solve the Fuel Cell poisoning.

I would like to share with you how I arrive to that figure. On 1987, I went back to my University to look for a Job. I got hired to work in the project Electrochemical conversion of methane to methanol. This was directed by my future PhD supervisor. We started studying the ideas of a german patent. This patent described how natural gas (methane) was converted into methanol by using a palladium catalyst and ferric ions (Fe+3). Thus, if ferric ions did have enough potential to oxidize natural gas into methanol, then this oxidant can be replaced by a potential imposed with a potentiostat. Ultimately, a electrolysis-like process could be make to produce methanol from a very inexpensive source. So far, so good! We started looking at the electrochemistry of ferric ions, the adsorption of methane on palladium electrodes, etc.

In parallel, we also started preparing an electrochemical cell and analytic techniques to detect the production of methanol after a promising electrolysis. We bought a large palladium medal from a jewelry store. It was a rectangle 5 x 2.5 cm. Big enough to fit into an electrolysis cell. One of the things I started to do was to deposit palladium particles (Pd plating) onto this electrode. This was to increase the area and had more chances the promising electrolytic reaction would produce enough methanol to be detected. We selected 0.5 M NaOH as an electrolyte. This was to avoid possible dissolution of the palladium in acidic media.

We were forbidden to do any electrolysis until we would better understand the ferric ions role in the electro oxidation of methane. This was somewhat frustrating because we were doing that but the research always produced more questions than answers.....The electrolytic cell to do the potential electrolysis was already assembled. The analytical method to detect methanol from the possible reaction was also in place, ready to do something useful. By the way, the method consisted in chromatography with a stationary phase resistant to water.

One day, We disobeyed our supervisor and started doing electrolysis of methane on palladium in 0.5 M NaOH. For 2 hours we passed a current that made the potential of the high area palladium electrode positive. After that, we distilled the electrolyte, took the first water drop of the distillate and injected it in the chromatograph. Surprise, surprise...methanol showed up! We were so excited about this discovery, we repeated the experiment twice. Then, we told our supervisor. He was very happy too and disregarded the prohibition.

For me, this was the start of a 7 years journey to understand why this did happen. We started performing methane electrolysis on a daily basis. Results were completely inconsistent. One day it produced 50 ppm of methanol, another day 100, other day 0. Some other day isopropanol, ethanol and acetone showed up, just to disappear the next day. What the heck?

One day, I was doing the palladium plating to increase the area of the electrode. The solution to deposit palladium is yellow. I always waited for all the palladium to deposit and you know that because the color disappeared. That particular day, the yellow color did not disappeared. I turned the equipment off, thinking to continue with the deposition the next day. Upon arrival the next day, the color disappeared! How come? I turned everything off! there were no electron source to deposit more palladium......wrong! Palladium is a hydrogen sponge. It can store upto 900 times its own volume in hydrogen. While the process of palladium plating was going on, inadvertently hydrogen was also being absorbed into the metal matrix.

I realized I had never controlled this unknown variable during the methane electrolysis. Next electrolysis, we did so. The results were way more consistent. It turns out methane was being oxidized by the potential. This oxidation stopped in carbon monoxide, CO. Then, the occluded hydrogen reacted with the CO, producing methanol. I could demonstrate that using surface FTIR at Purdue University (Figure). We started trying to convert methane into methanol and we ended finding a method to avoid CO poisoning in fuel cells!