Yesteray Dr. Hotz explained how the Gas Chromatograph worked and I’d like to share the specifics of the essential piece of equipment I worked with here.
Looking at the key, there are a multitude of essential parts to this machine. Distilled into a user’s guide, the two coils in the center (inside the oven) are constantly carrying gas through the machine and out the back. When the measurement is engaged, the sampler (1), which feeds the coils, switches off and prevents any more gas to enter. This gas is then heated. The coil is coated with an absorptive substance that the gasses stick to. As the gas mixture is heated and the particles separate, different gasses detatch from the coating at different times (once they have reached the heat necessary unique to the molecule). By measuring the time at which the gas molecules detatch in the detector (4), we can tell which gasses are in our final composition i.e. what reaction took place, how much reacted, and what the yield in terms of energy will be.
Yesterday, before leaving the lab, I ran a trial testing a catalyst another student had made, only using a higher concentration of methane in the reactants. We expected to see a proportionate change in the resulting methane concentration after running through the reactor, but saw almost no change. Thus begins our problem shooting.
While waiting for measurements to come in, I drew schematics of the GC to be explained in tomorrow’s post
There were a couple essential areas of fault here: method (experimentor systematic error), the measuring device (systematic error- is the GC insensitive to methane?) and the flow meters (random error- if the flow of one gas changes randomly as the regulators are wont to do, it will change the concentration of the reactant mixture randomly)
The rest of the day was spent trying to isolate these areas of error. The major method fault is that we cannot separate the independent variables of time (always changing) and temperature (what we want to change to evaluate the effectiveness of our catalyst). If the catalyst deteriorates with time (due to oxidization of the materials) then the consecutive trials will lower in reaction yield even with an increase in temperature, a change that generally spurs reactions to become more active. We were, however, able to rule out the sensistivity of our machine by running 100% methane through the GC and seeing a doubling of the output relative to the 50% mixture we left yesterday with. This means that the failure to change between yesterday’s two mixtures (going from 3% to 50%) was due to some other issue. We then took two more trials at the same heat and saw a regressive curve with consecutive trials. It appeared that our catalyst had a lifespan. In order to fix this, Dr. Hotz showed Maya how to pretreat the catalyst while I worked on collecting calibration points.
Having made a new catalyst yesterday, I spent my last day in the lab testing the reactor. After a few warm up measurements (which would be used as calibration points later on) I measured the output of the reactor when receiving a mixture of 50% CO2 and 50% methane (CH4) and….there was a leak. One of the pipes (potentially the pipe that we had to refit onto the tube housing the reactor, since there was very little insulation, only a heat-fitted tightness keeping the gas in) was leaking air and preventing us from seeing any coherent data or testing the reactor to see if it was actually better than the old one. Since our flow meters also seemed to not be working properly, I had abandoned this task. Undoubtedly a poor way to finish the experience, but I enjoyed every experiment, measurement, and lesson I learned along the way.
After many days of taking measurements and reading articles, today is the day for work. I was tasked with creating a fresh reactor by following the steps a previous student had detailed in her thesis (a process she had adopted from a different article). We had all the necessary materials, containers, and devices to create this catalyst, we only needed the time. Yesterday we started the process by calcinating a sample of ~200mg of zirconium oxide which would be our base for about 4 hours (we heated it up at 400 degrees C for 4 hours). Today I started by calculating the amount of Ruthenium Chloride Hydrate needed to add to the base. This was tricky. We wanted a final catalyst mixture that had 4% of the weight being consisted of pure ruthenium. We did not add pure ruthenium. I had to determine how much of the salt to add to get a hypothetical return of 4%. After being stuck in the lab for months, the zirconium oxide had acquired some moisture and lost a lot of mass after being calcinated, which means the calculation wasn’t made with the original 200mg but with ~120mg of zirconium oxide. After dissolving the salt into water, mixing, adding the base, mixing, and filtering for a solid half an hour (almost 15 runs through a quartz glass filter) I had distilled a solid nanoparticle catalyst. This was then stuffed into a glass tube between two packages of quartz glass fiber and, voila! We have a brand new reactor, ready to be tested tomorrow.
I used the relevant information published online by the company we ordered this material from to determine the amount we needed to add.
Today I set out on making a new reactor, sort of. I was going to find a new reactor to create, but by the time we would have ordered the materials I would be long gone. I did, however, read plenty of papers on various different experiments and had to determine which potential candidates were 1) reliable, 2) cheap, and 3) better at our desired temperature range than our current catalyst. To do this, I had to choose my search terms carefully and vet a lot of articles based on the reliability of the journal they were published in. Easiest way to do this is looking at the impact factor (average number of citations an article published in the journal gets in a year). Mechanical Engineering journals have fewer citations on average than the natural sciences, thus we are looking for something around Impact factor of 7 for a good journal and a likely good article.
Example article I would have to look at and determine whether to readIf I didn’t trust the data from research gate, I could look up the journal elsewhere and see that this journal has a different estimation on springerResearchgate calculates Impact factor determined on the times an article is cited from their website, and this way I could vet this article
Today we tested for a potentially large error in analyzing the efficiency of our reactors: whether the desired reaction was occuring. Since we only see the final composition of the gas from the GC, we cannot be certain if Hydrogen is being turned into water and condensing onto the pipes (therefore not being represented) or even turning back into Hydrogen. Since we have a gas mixtures of CO, CO2, and H2 we could be seeing a spike in CO without the original gas being used for any reaction if an auxiliary reaction (2CO2 + 2H2= 2H2O + 2CO) produces more CO. To test for this and account for the error, we ran pure H2 and CO2 through the machine with the current reactor to determine the rate that this reaction would occur in the reactor at varying temperatures. We can then use this as a maximum baseline error, since under the normal conditions there will be less likelihood of this reaction occuring since the desired reaction is taking place and using the necessary gas molecules.
Today we welcomed a new intern to Dr. Hotz’s team: Maya. Knowing that we would be starting around the same time, Dr. Hotz delayed our mandatory safety debrief until we were together.
Every lab has its fair share of safety hazards, but in many ways the combination of hazards is what defines each lab. After all, one can only say they know what they’re doing in the space once they are certain of what they are not doing, or rather must not do. This way of defining a space– using the safety procedures as negative outlines of the experimental procedures– is written into the department itself. Before working with any materials, I had to review countless standard operating procedure (SOP) debriefs and sign my name in compliance. Strangely, more than merely learning the rules, I gained a better understanding of the role of researchers in the greater university web. Suddenly the lab assistants weren’t just an assortment of people typing away at a computer, but part of a greater agreement with a shared responsibility to one another and to the university. Regardless of the individual motivations, projects, and goals of interns, assistants, and doctoral researchers, it was interesting to find a common thread of purpose– that is, carrying out experiments in an efficient, non-wasteful, and safe manner– that would remain long after I left and even after Dr. Hotz left too.
Some of the safety hazards included: Oxidizers (gasses that can add to or intensify an open fire), nanoparticles (particles that are so small they can diffuse through skin, enter the blood stream, and potentially cause cancer. Largely unresearched health detriments), glass (potentially broken/sharp), and noxious/flammable gasses.
The fume hood prevents us from breathing in loose nanoparticles when making catalystsOne of the safety agreements I had to sig
I took the bus to campus today– a first for me– and was pleasantly surprised at how manageable the walk from my stop to Hudson Hall was. All but assured I would get lost finding my way, I began the work day feeling more comfortable than I started– and not just because I had finally found air conditioning.
The first thing I did today was meet with Dr. Hotz to debrief on the direction of the internship. He made it clear from the beginning that the work I would experience as part of the Mechanical Engineering and Materials Science (MEMS) Department was unique in that it is funded by an academic institution, and as such would be very different than the work of my peers. For him, there are no deadlines, corners to cut in production, and a failed experiment isn’t a setback like it would be as a commercial researcher. We discussed the project his lab is working on, current experiments and material optimization research they are doing, and general lab safety procedure.
Essentially, Dr. Hotz is trying to optimize hydrogen fuel cells as a renewable energy source. He does this not by researching fuel cells, but by researching the way we get the fuel: hydrogen itself. Hydrogen is a gas that must be pressurized (requiring energy) to be stored, at which point it still takes up so much space it is impractical for single decentralized systems (power for your home or office). The solution to this problem, for many mechanical engineers, is methane steam reforming.
(Methane Steam reforming {formula 1} turns methane {CH4} and steam {H2O} into Hydrogen gas and Carbon Monoxide)
Methane steam reforming, however, can be incredibly inefficient. It requires temperatures greater than 500 degrees Celsius, temperatures for which most energy sources have to burn half of their methane supply to achieve. To combat this inefficiency, Hotz and his students have developed a solar powered, nanoparticle catalyzed solution. Using Biomethanol and a prototype flat solar collector, without solar concentration Hotz’s team can produce electricity at an efficiency of 45% with net neutral carbon emissions. The team is now working on optimizing the process. With greater temperatures the team will be able to use methane and not methanol, but it’s not so simple as cranking up the heat. At greater temperatures: materials malfunction, heat transfer (ambient loss) becomes easier, and the solar collection plate becomes more emmissive (radiative loss). I’ve already started collecting callibration data for a new catalyst built by other students, and tomorrow I’ll continue experimenting with new materials. I will also explain a little bit more about why higher temperatures screw up the current prototype, but until then, enjoy these pictures of the lab.
Vaccuum Chamber for testing energy efficiencySome of the catalyst (far right) and solar absorption materials (far left) Hotz’s students were testing
