When I started working in the wet lab again last winter, I noticed one of the drawers in the back was newly labeled as “Michael’s Manifest Destiny”1. Michael Wisser, a sixth year in the D-Lab and a mainstay of the upconversion subgroup, is known for his sense of humor (see his colorful Day in the Life posts), so I figured he was synthesizing a nation of samples to take over our wet lab. While he hasn’t managed to do so yet, his nanoparticles have broken record upconversion efficiency values, and resulted in a recently published paper in ACS Photonics. I sat down to talk with him about his research, his paper results, and survival techniques for debugging a nanoparticle synthesis.
What is your research on?
I work on upconversion, the process of converting lower energy light into higher energy light. Upconversion has a lot of exciting applications, but the one that motivates me the most is photovoltaics. The sun emits light at a broad range of wavelengths in varying amounts. However, our current solar cells have a cutoff energy in terms of the light they’re able to use, so some of the light from the sun will get wasted no matter which solar cell you use. Upconverters could convert the unusable lower energy light into useful higher energy light. However, upconverting materials are currently very inefficient, with nanoparticle upconverters only converting about 0.1% of incoming light. My work aims to improve the efficiency of lanthanide-based upconverters by modifying the structure of their host material.
Why modify the host material?
The bulk of the upconversion process in these materials is accomplished by lanthanide ions, but these ions can’t just float in space. You need to put them in some sort of solid material, which we refer to as a host lattice or host matrix. The precise structure of the host material influences the probability of a lanthanide absorbing the incoming low energy light or emitting high energy light. These processes are not very probable in our control material, which is why upconversion is typically very inefficient. By distorting the host structure, we can make the transitions more probable.
So in your most recent paper, how do you distort the host matrix?
We use this host material called sodium yttrium fluoride (NaYF4). Previously, we took sort of a brute force approach by putting these upconverting nanoparticles in a diamond anvil cell2. This allows us to compress the particles to very high pressures, which essentially pushes all the atoms together and introduces some degree of distortion in the material. However, this wasn’t something we could incorporate if upconverters were to be used in applications, so we had to be a little more clever.
In our most recent paper, we instead modify the composition of the host matrix. We start with sodium yttrium fluoride (NaYF4) and replace yttrium with elements that are slightly different in size, which should mess up or distort the host matrix.
Why replace yttrium, as opposed to, let’s say, the fluoride?
Changing the fluorine would change a lot about the synthesis and sample because the fluoride makes up the bulk of how the material behaves chemically. Yttrium is not technically a lanthanide, but it behaves very much like one. Lanthanides are this group of 14 elements that have extremely similar chemical properties, and it’s quite easy to take one out and replace it with something else.
So once you’ve modified the host lattice, how did this affect the upconversion process?
When we removed 10% of the yttrium and replaced it in equal parts with gadolinium (slightly bigger) and lutetium (slightly smaller), we were able to improve the upconversion efficiency of the particles by a factor of 1.6. This is the record efficiency value for this type of upconversion process.
Wow, that’s awesome!
Yeah, so awesome. I think it’s a good technique because it only requires a slight adjustment to the synthesis procedure. This substitution method is also not specific to the lanthanides we used; this approach would work for any lanthanide ion. And it doesn’t matter if the particle is 5 nanometers, or 50 nanometers, or 500 – this is not a size dependent effect, so we expect the same enhancement for any size. There are a lot of alternative approaches to enhancing upconversion, but most of them sort of live with the problem of inefficient and improbable transitions and try to find a way around it rather than addressing it specifically. Since we do address that specifically, we can combine our approach with others that focus on different problems to get really unprecedented efficiencies.
These are amazing results! I’m also interested in hearing about some of the behind the scenes of your research – why don’t we start with the ‘Manifest Destiny’ drawer?
I didn’t know anyone had even noticed that.
That was one of the first things I noticed! I thought it was hilarious.
I’m glad. Well, I initially had a pretty small drawer. I then did a lot of syntheses, which led to a lot of samples, so I quickly ran out of space. That drawer was only partially occupied by people who had since graduated, so I saw it as territory to which I could expand and was my divine right to colonize.
What was your favorite part of this project?
It was especially gratifying to be successful with my idea. Before this, I was trying to reproduce the diamond anvil cell effects synthetically, so shrinking or expanding the host lattice. I was kind of daydreaming, and tried to figure out if it would be totally ridiculous to just remove yttrium and throw in two other ions that would cause local distortions but balance out on average. I don’t remember if I asked Jen before I started, or if I had a free afternoon and just tried it, but I got really bright particles on my first try, which made me want to explore it more thoroughly. It then took a lot of tweaking to synthesize all the samples to make the study valid.
How many syntheses did you do?
Wow that’s a lot! How did you get through all those syntheses?
Morale was definitely hard. I never really knew how the particles would turn out until the end when I looked at them under an electron microscope. The synthesis takes 7-8 hours so I might spend the whole day thinking, “Yeah, this is going to be my last synthesis. My particles are going to be great” but then I would find out that they’re terrible. You just need to reset your frame of mind for the next day and try to be happy again. I reminded myself that every failed synthesis gave me some type of information, like that specific recipe yielded particles that were a little too big or too small or whatever but now I know so I can move on. It’s not like anything was really useless.
That’s really good advice. I’m currently in the synthesis state of my research and it’s the same thing where I keep having to remind myself that every failure is still helpful. How long did all of this take you, and what was your schedule like?
I first tried my idea around August 2015, and the syntheses mainly ended in February of 2016, so it took about six months. I was in the lab doing synthesis a lot and scheduling was rough. There were many days where I would do a synthesis, clean the glassware, put it in the base bath, come back at like 11pm to take it out of the base bath, put it in the oven, and then come back at 8:30am in the morning to start a synthesis again. But I had to give this priority in my life in terms of scheduling, so less time for video games and my girlfriend.
Jen (our advisor) did give me bonus points, though, for being here on New Years, so there’s that. Got those bonus points.
Would you do this many syntheses again?
If I knew there was a meaningful end, I would. I was also doing syntheses before this project, but my attempts at synthetically reproducing the diamond anvil cell effects didn’t really lead to anything, and that was a little sad. I don’t really feel sad that I spent the time, but I feel like something could be there and it feels wasteful to just have the samples and the data hanging out.
But, every time the synthesis gave me something I wanted, it was very encouraging. I felt like I actually made progress and could do it forever. On the other hand, when it didn’t go well, it can really hard to turn it around. It was somewhat an exercise in mental strength.
I guess you find a way to make it work, like I still went to the gym a lot. At some point, I also decided that when I would do a synthesis, I would save some time if I bought breakfast from CoHo3. I only saved like a marginal amount of time, but now every time I do a synthesis, I get food from there.
Yeah actually, I do. There’s one woman there who knows what I get. I’ve probably spent a thousand plus at CoHo.
Well, if it’s your good luck breakfast and it works for you…
Yeah. I recently switched to the veggie breakfast sandwich (from the regular one) though, trying to be healthy. It’s not as good.
1For those who might not remember all of high school US history, manifest destiny is the 19th century belief that the United States was destined to expand and colonize across the continent.
2Check out Michael’s previous work here
3CoHo is a COffeeHOuse (get it?) on campus that has the best hours (it’s almost always open)