About Me: Hello, my name is Demetrios Kennedy and I am a WPI graduate from graduating class of 2025. While at WPI, I was a member of the Teacher Preparation Program while also being the Vice President of the Inspiring and Developing Educators Association.

About the Lab: The Energy Metals Research Group is the lab Jared and I are working with this summer. This group hosts a large amount of projects ranging from magnesium production and recycling (our group’s project) to other projects such as and . For our project, the research we are working on the the production and recycling of magnesium through two processes: electrolysis and distillation. For electrolysis, we are working with the lab to combine magnesium and tin in order to create an alloy that we can then separate further down the experiment chain while also separating magnesium from oxygen. For the distillation procedure, we are melting the tin and magnesium and then boiling the mixture to separate the magnesium from the tin. This is so we can record and recover the magnesium gained from the entire experiment.

The lab itself has a vision statement found here which states “The Energy Metals Research Group (EMRG) will serve society as a leader in practical applications for electrochemical technologies; pioneer in enhanced transportation systems and industrial safety; catalyst for reduced manufacturing waste and energy consumption; and key player in attaining a world with open international collaboration; 100% reliance on clean energy, expanded economic prosperity, healthier environment, and improved quality of life.”

Magnesium Production and Recycling for Clean Energy

Weekly Updates

Week 1

As mentioned above, the entire experiment follows a 2 step process of electrolysis and distillation. The first step is electrolysis, however due to schedule changes in the group, we are starting to learn about the distillation step first. This part of the experiment involves loading a crucible and a distiller piece into a heating apparatus along with evacuating oxygen to create a vacuum.

Below is a picture of how it looked on July 1st, 2025. The two tubes are connected to create a vacuum that pushes out oxygen so it doesn’t react with magnesium. Below is a furnace meant to heat up the metal alloy.

The apparatus is placed into a hole in a table and is contained within an insulating device to control heat loss. In order to create a vacuum, two tubes are connected to the apparatus. The tube on the left is connected to an argon tank and the tube on the right is a tube made for oxygen. Once the experiment is ready to run, the argon tank is turned on so argon can flow into the apparatus and force the oxygen out the other tube. The oxygen leaving the apparatus reduces the internal pressure which then causes the boiling temperature of the mixture inside to decrease.

Jared and I entered the lab in the middle of Daniel’s (The graduate student we are working with) current experiment which is trying to recover leftover magnesium from the crucible, which depending on the results will determine the next steps for the next couple of weeks. If Daniel is able to recover magnesium from the experiment, then we can continue on as originally planned. If Daniel is not able to recover magnesium, then this would mean the metal oxidized with the stainless steel walls of the container and a new container would have to be purchased. Luckily, we were able to generate magnesium. From there, our group sanded down the sample and used a machine to test the purity of the sample. The sample was roughly 98% pure, which was different from Daniel’s last sample so our group of three spent some time bouncing ideas back and forth as to how the sample purity could have changed. After this point, we prepared the next sample for analysis which was an alloy of 60% Tin and 40% Magnesium and brought this to the welders for next week.

Week 2

Most of Week 2 has been spent trying to analyze Daniel’s prepared sample from Week 1 (60% Tin and 40% Magnesium). We sent the sample in early-mid Week 1 and Daniel predicted that we wouldn’t get the sample back until Tuesday, which ended up being right. This meant that we had to go on some different quests while waiting for the sample to come in.

On Monday, Daniel introduced us to another topic that was heavily related to our research. This topic surrounds the walls of a potential furnace in use of the research. Most furnaces are made with Nickel as the material since Nickel can withstand high temperatures. While this benefits most processes, for the purpose of magnesium recycling, the Nickel from the furnace wall with alloy with magnesium and can cause corrosion which will damage the magnesium. To counteract this corrosion, the plan is to use a furnace made of steel since Iron bonding with Magnesium won’t cause the magnesium to corrode. The issue that then comes is that magnesium bonds with iron in the first place. A proposed plan is to somehow coat the inner walls of the furnace with Chromium Oxide (Cr2O3) so that the compound will bond to with magnesium. When researching, I offered up the idea that maybe we should do electrolysis after the furnace section. Jared offered up a great point that if we did do that, it would add another step to the process and that could cause even more issues to arise. After Monday, we put this on hold for if we ever had free time and were possibly caught up with RET work.

On Tuesday we got the sample back from the welders and we were finally able to start the experiment, until we realized that we needed a vacuum gauge. It turns out that the gauge Daniel had been using stopped working, so we needed to find a new one. First we checked with the lab manager Tony, but he said he didn’t have one. Then we went to different plumbing stores until one recommended Brierly Lombard. After going to Brierly Lombard, we weren’t able to find the part there however they gave us another location to go to which was Omni Services in Auburn. They even called ahead to let them know we were coming, which was super helpful. Omni Services had the part, so we brought it back to WPI and used glue to seal up any gaps in between the vacuum gauge and the piece we were connecting it to. After this, we were done for the day as we had to let the glue dry.

Wednesday was spent setting up the experiment and we had professional development with Donna, so not much happened then. On Thursday we started the distillation experiment. This time, there were different parameters like with temperature and set up. Down below will be a picture of the set up, but essentially a crane was used to lift the distillation apparatus into a furnace. After this, three rods were placed into three holes on the setup in order to track heat throughout the experiment. Another rod was attached to a piece connected to a pipe to track the highest temperature change. Once all rods were properly connected, we then placed a lot of insulation to keep the heat in. The vacuum gauge worked so we attached this to a pole sticking out to keep the system in vacuum. The furnace was heated until roughly 900 degrees C in order for the inside to reach a temperature of around 700 degrees C so the magnesium could melt and then evaporate.

This is an image of the furnace setup for distillation. The white foam-like substance is insultation. The two half circles are also insultation meant to protect the device that reads the temperature.

Once the device was set up to start heating, we then set off to do other work while the furnace was heating up. This work involved cutting three different metals to test magnesium absorption, the three metals being 304SS, 310SS, and an unknown one. The numbers refer to the composition of the alloy. The process was to take small chunks of the metals and melt magnesium onto them to see how much magnesium would be absorbed. We used an abrasive saw to cut the chunks off and then used a larger saw to cut the magnesium into chunks that would be melted later.

On the last day of Week 2, we had a tour at Polar Beverages in Worcester. There we got to see how a drink was made along with comparisons between Polar and other companies. I think my favorite part of the tour was seeing the chemistry section where they were showing off the titrations as I don’t think many people would think chemistry like that is a part of making seltzer water. I wanted to ask a question about how reverse osmosis was a part of the entire process of making the drinks at Polar, but I forgot to ask at some point during the trip. Reverse osmosis is the process in which a solvent with varying concentrations of contaminants is forced through a semi-permeable membrane. The purpose of this membrane is to allow only the solvent to go through and to trap the contaminants. This method was almost used in my MQP during senior year, but my partner and I opted for a more cost-friendly approach. This is part of the reason why I would have asked this question, but the main reason would just be to figure out their process. My group didn’t look into how it would function as the price pushed us away immediately, so seeing how the actual process would have unfolded would’ve been interesting. To end Friday, our research group ran into a slight issue. The molten metal alloy had cooled down and left a solid chunk of metal, however it was a little stuck inside of the crucible. Jared, Daniel, and I tried our best to get the metal out but we weren’t able to and we had to wait until Monday in order to hear back from professor Powell. The reason for this is that we wanted to know if breaking the crucible was mandatory or if there was an alternative way to get the metal and analyze it.

Week 3

As mentioned at the end of Week 2, the question was whether or not we had to break the crucible. The answer to this question is yes, we needed to break the crucible in order to get to the metal alloy. On Monday, we tried to drill a hole into the metal and place a bolt in the hole to create leverage to pull on the metal. The metal bolt wouldn’t stay in place long enough and would actually be pulled out of the metal. Eventually after trying, we came to the conclusion that we had to break the crucible

Rest in Pieces Crucible

Here is a picture of the crucible after we smashed it with a hammer

Once we got the metal piece out, we needed to cut the metal in half and clean it to get a proper reading of its composition. After these two steps, we determined that the metal was 99.7% Magnesium with low amounts of Tin. This was an issue however, as it should have been the opposite. If we were to assume that the experiment went according to plan, then we should actually see 99% Tin and low amounts of magnesium. This would mean most of the magnesium was evaporated and landed in the condenser. Since we see the opposite, it either means we need to use the machine differently and change parameters or that we truly evaporated mainly tin instead.

On Tuesday we continued to test the composition of Magnesium to Tin. We decided to use a device that analyzes the sample composition. Looking at the slide show below, you can see that the amount of Magnesium increases as we get closer to the top. In the image of blue (Tin) and grey (Magnesium), you can see the grey surface increase in surface area as the images go from the bottom of the sample to the top of the sample. Even though the data from the device is inaccurate (I’ll explain shortly), what data we did gather does line up with our hypothesis that magnesium should be closer to the top (I’ll explain that too)

Inaccurate Device: The device we used was inaccurate for measuring composition due to its reading. For the more accurate setting, we ended up only seeing Tin as a metal. According to Daniel, we only recovered roughly 30 grams of magnesium out of 270.8, meaning that the rest of the magnesium would be in the crucible with the tin. If we only had tin in the reading it would mean that we lost over 200 grams of magnesium which is highly unlikely. Using the more inaccurate reader, we were able to determine that magnesium was inside of the alloy chunk, along with other metals that should be there. For example, there were metals like Calcium and Potassium along with nonmetals like Phosphorus. There is no way that those metals listed could have been introduced into the system. Having said that, the readings of strictly Tin and Magnesium do make sense when compared to the image of the metal surface, but the data can’t be used to definitively comment on the composition.

Metal Arrangement: The reason why Magnesium should be closer to the top of the alloy is due to the density of Magnesium compared to pure tin and a magnesium tin alloy. In the case of Magnesium, the density of the metal is around 1.7 g/cm^3 while Tin is essentially 7 g/cm^3 so when they both turn to liquid we should see Magnesium rise to the top of the solution while Tin sinks to the bottom. If a magnesium tin alloy were to be created, the density is roughly 3.5 g/cm^3 so it would still be above pure tin. If you look at the slide show above, you should see this effect going from top to bottom.

On Wednesday we had a meeting with professor Powell about the rest of our time for RET, mainly about starting the second half of the research. This part including massing the amount of salts we were using in the experiment in which we used Magnesium Fluoride (MgF2), Calcium Fluoride (CaF2), and Magnesium Oxide (MgO). At the bottom of the crucible was roughly 300 grams of tin. The target mass for the salt solution was 700 grams and this was measured by dividing the needed amounts in half as we didn’t have a container that could hold 700 grams of salt. We measured half and once the amount needed was measured, we placed the amount into the crucible on top of the tin and remeasured another half. After the crucible was filled with everything needed, we then set up the experiment for electrolysis and waited to start the experiment on the next day.

For this part of the week, I’m combining Thursday and Friday since not much was done on Thursday. The main event on Thursday was really the PowerPoint going over our group’s research during the first half of RET. We gave a brief overview of our progress and after that we went to the lab in order to start the electrolysis part of the experiment

Here are the images of the screen we used to analyze the sample along with the surface that we analyzed. In the slides are pictures of the surface zoomed in to see the distribution of magnesium and tin.

On Friday we came in a little early to start heating up the machine and then we ran the electrolysis. The process for electrolysis involved cycling the machine to make sure it works, then we let it wait for 5 minutes to cool down. To ensure that the machine worked properly, we ran the machine and raised the voltage by 0.5 volts every 12 seconds. Using an application on the computer, we should see a graph that looks like a hockey stick, with the foot of the hockey stick then leading into the pole. After letting the test run and letting the machine cool down, we ran the first step of electrolysis which was running the machine at 1 volt for 1 hour. Once the first test was completed, we repeated the steps from before: raising the volts by 0.5 every 12 seconds until we reached 5 volts and then letting the machine cool down. After this last step we ran the final trial which was to run the machine at 3 volts for 2 hours. During the last trial, the current was fluctuating between 0.6 amps and 1 amp which Professor Powell said was low so at the beginning of Week 4 we are going to discuss with him what may have happened.

Week 4

After going over the data and analyzing the sample we got (see below), we decided that we needed to add more tin. Jared and I met with Daniel and Professor Powell in order to discuss the next steps and we found out we supposedly did not have any magnesium in with the tin. We tried the sample that we thought would have magnesium so we decided to add more tin compared to the first trial. The first trial was around 300 grams of tin, so this time around we are trying 1000 grams of tin to ensure that there is magnesium and tin contact. In the image of the alloy we created, you can see that tin does not completely cover the bottom of the surface. The conclusion that us 4 (Jared, Daniel, Professor Powell, and I) came to is that as the salts melted, the crucible was at an angle which caused the tin to unevenly melt. Rather than continuing with a new experiment, we modified the first one as mentioned above with 1000 grams of tin along with the normal amount of salt mass (700 grams). On Tuesday (everything above so far has taken place on Monday), we will set up the apparatus and potentially start the pre-bake.

On Tuesday we partially set up the experiment as seen below. This involved making sure there wasn’t an electrical connection between the crucible and the rest of the setup which would cause the device to short circuit. We split the connection through the usage of insulation (the white fluffy parts in the image). We tested the connection multiple times to make sure that there was no connection between the crucible and set up. We then debated on whether or not we should do the pre-bake on Wednesday or on Tuesday as if we did the prebake on Tuesday we would leave at 8 PM and then leave earlier on Wednesday. We made the decision to hold off until Wednesday to make sure we had plenty of time for the pre-bake

On Wednesday we had our professional development and then went straight to the pre-bake. This process involves carefully controlling the temperature on the furnace so that it does not go over the target temperature of 400 degrees C as this would require us to turn on the argon gas so that oxygen can’t react with the magnesium and cause a hazard. Jared and I were able to keep the temperature under control with the help of Daniel’s notes that he gave before he left for India. We kept the furnace at the target temperature for 3 hours to dry the salts and then we went home for the day.

On Thursday, we started the electrolysis! With Daniel’s help, we were able to heat the furnace itself to a 1150 degrees C because he programed the furnace like that and Jared and I made sure nothing went wrong with the preheat. Nothing went overly wrong with the experiment, so we finished up and went home. We couldn’t analyze on Friday, so our analysis had to be during Week 5.

Week 5

 

We got the results back from the electrolysis experiment and going based on the graphs (which will be attached below) we did the experiment correctly. The shapes of the graphs match what professor Powell said would be a successful experiment, however our actual metal sample did not have Magnesium with the Tin.

Above is a picture of the bottom of the sample. The dark gray is the salt layer that somehow got trapped underneath the tin which added weight to the overall mass. This was an issue for us as we thought we had successfully trapped Magnesium in the Tin, however when we shaved out the salt, the mass dropped below 1000 grams. Jared went to test the composition of the salt and ended up finding magnesium trapped in the salt. One thought I’ve had is that we should try putting the tin on top of the salt solution. Between Tin, Magnesium Fluoride, Calcium Fluoride, and Magnesium Oxide, Tin is the most dense so when everything melts, Tin should sink to the bottom and react with the Magnesium from Magnesium Oxide. I think this would increase the chance of having Magnesium inside of the tin so that we can properly run the experiment, however we might not get to test this idea as we ran out of argon.

Here is the XRF data (left) and the salt sample analyzed (right). The location where Jared used the gun to analyze the salt was on the darker gray area, which would make sense considering every substance’s density. The tiny pockets are due to oxygen, as the electrolysis only separates Magnesium and Oxygen. LE stands for “Light Elements”, which is basically the machine saying there’s something there but it can’t identify it.

We were able to get the SEM (Scanning Electron Microscope) data from Daniel this week which was surrounding the distillation experiment we started on Week 1 and finished on Week 2. The dark grey is the Magnesium while the light gray is the Tin. Similarly to the XRF gun, we can see the Magnesium increase in percentage as we go from bottom to top, but the difference is that in the SEM data the Magnesium and Tin are basically at the same percentage.

Week 6

For the final week, we spent time working on the poster that you’ll see below while also working on our lesson plans. For my lesson, I plan on teaching students about the conservation of mass through a baking soda and vinegar experiment while teaching about the Hindenburg disaster. While doing research for a real world connection, I found theories stating that the disaster was due to an air leak on top of bad weather. I wanted to take advantage of this and design an experiment around generating enough carbon dioxide to fill a balloon or a nitrile glove finger. Students would measure the volume and calculate how much CO2 could fit in there. From this, students would use stoichiometry to calculate how much of baking soda and vinegar are needed to generate the CO2 calculated and then they would test their hypothesis.

Final Poster:

Lesson Plan: