Dr. Yu is a keynote speaker at the 20th Annual Green Chemistry and Engineering Conference in Portland, Oregon. As a chemist at the Scripps Research Institute, his work has focused on discovery and rational design of new reactions using C-H activation.
Q: In what ways do you see your research accelerating the adoption of renewable feedstocks, such as biomass?
A: That’s a great question. In C–H activation, in my opinion, there are two sub-fields. One has largely been attributed to the alkanes, like methane activation. Of course, this sub-field is unlikely to have a lot to do with the activation of intermediates in synthetic chemistry, so what we are trying to do is to make more complicated molecules. We consider this the second sub-field. As you mentioned, the question becomes: how do you connect to biomass? So we’ve also thought a lot about this in our group.
To facilitate the discussion, we could divide the problem of C–H activation into “first functionalization” and “second functionalization.” While these two problems are fundamentally related, challenges and approaches for catalyst design differ significantly. The first functionalization is essentially alkane/methane activation, and by activating the C–H bonds on these substrates you introduce the first functional group to the molecule. What we’re focused onis the following problem: can you pick up a feedstock chemical that already has a functional group, like from biomass, and from there insert a second functional group and start building complexity? As you can imagine, unlike ethane or methane, many C–H bonds are different now due to their distal and geometric relationship with respect to existing functional groups, therefore, achieving the site-selectivity is challenging. Achieving such selectivity is essential as the distal relationship between the existing functional group and the new functional group largely defines the function of a molecule; i.e. how drugs, agrochemicals or whatever it may be, works.
Usually, we take whatever functional group we get to begin. In our group, we’re specifically aiming for those substrates that are abundant. Whether that’s to be found in biomass or not, I’m not quite sure. But, usually it does go back to biomass. Basically, it has to be something that’s abundant.
I’ll give you one example that we’ve been working on for many, many years now – about 15 years. We started this project in Cambridge University. What we’re really trying to do is to use stereochemistry to gain the basic understanding needed to solve what we call the isobutyric acid problem. One of the aims is to take isobutyric acid, and that can be viewed as biomass or an abundant bulk chemical, and then we try to functionalize this molecule at the beta-position enantioselectively. Then, you can create incredibly useful alpha-chiral centers that are cornerstones in asymmetric synthesis. The interesting part is if you are doing feedstock chemicals, your substrate scope is going to be relatively narrow, but we can use a handful of isobutyric acid analogues including pivalic acid.
The point is, that the product scope is very broad. This is because once you break the C–H bond, you insert the palladium, and then you can convert that C–H bond to many, many other bonds like C-N bonds, C-B, C-C bonds, etc. and you can make many connections. One of the famous analogies to that is found in nature. Nature takes isobutyric acid and converts it to the Roche ester. That’s only one substrate, and nature has done it very well. The Roche ester has served as one of the key chiral synthons for the synthesis of polyketides as well as many drugs and related molecules.
Nature, of course, does this very differently. Instead of going through the asymmetric activation of the gem-dimethyl group which is too small to differentiate, enzymes perform dehydrogenation of isobutyric acid and subsequent asymmetric hydration to give the Roche ester. In our case, the organometallic approach does have one advantage of being able to make a wide range of carbon-carbon and carbon-heteroatom bonds in addition to carbon-oxygen bond. As such, we can make a diverse range of compounds rapidly. I guess that’s one example. I don’t want to bombard you with many examples, but that’s one of my favorites. From isobutyric acid, you can also think about pivalic acid. It’s the same idea, and then you can make many chiral quaternary carbon centers. So the beauty is that the substrate scope is very limited – because it has to be if you’re talking about biomass feedstocks – but the scope of the utility is not. You can make so many products with many C–H bonds present in substrates. In isobutyric acid alone you have six C–H bonds that you could potentially turn into all kinds of functional groups.
Even if the number of biomass feedstocks is low, you can get a huge variety of molecules if you can cleave the C–H bonds. If you can’t do C–H activation, biomass is very limited in terms of synthesis because you’ll always have the six C–H bonds from the dimethyl sticking there, and that prevents diversification. Even if you can make a million compounds, they’ll all look basically like isobutyric acid. So it’s very important that you actually break those C–H bonds and make them unrecognizable. Then you access true diversity.
Q: In your research, you often use palladium catalysts and volatile organic solvents. How necessary are these kinds of sustainability trade-offs or are you looking for alternative materials (i.e. could you reduce critical material use or solvent-less or aqueous reactions)?
A: There are many layers to that question. As you can imagine, it’s very complicated. Let’s talk about E-factor; there are so many parameters that go into that equation. The first question you should ask when you think about palladium catalysis in C–H activation is how many steps are you really cutting out from the synthesis? Recently, we have a process route designed by Novartis cutting eleven steps to five steps. So what that means is that the big stuff – whatever solvents or even water or cheap catalyst you’re using – are now reduced to zero in those removed six steps. You’re not using any of it. That’s one big-picture thing that green chemists or any synthetic chemists are thinking about. Cutting steps is much harder than anything else. If you cut one step out, you will improve E-factor of the synthesis significantly. Indeed, C–H activation is focusing on that exact aspect. One of the biggest goals of using C-H activation in synthesis is probably cutting steps out.
The second perspective is that even if the palladium catalyst is not as environmentally friendly as something like iron, which is cheap and readily available, I would say there are ways out for palladium catalysis which have been demonstrated in industry already. First, the palladium catalyst loading is a key issue. We have to really, really reduce the catalyst loading. The second is that the ligand can be just as expensive as anything else, if not more. It’s very important that we develop very cheap, readily available ligands. One of our trademarks is the amino acid ligands. As you can imagine, it meets these criteria. So the ligand has to be considered when you’re designing a catalyst.
Third, we’re trying to immobilize the catalysis. Recently we submitted a paper in collaboration with Chris Jones at Georgia Tech through an NSF center collaboration. We immobilized the C–H activation catalyst. This way, the palladium can be re-used.
The biggest problem for palladium catalysis in C–H activation is not necessarily what I just said. The biggest problem is that when you do C–H activation you have to use oxidants. That oxidant is the most problematic from a green perspective. Quite often, they are expensive and they are used in large amounts. We focus a lot on reducing the catalyst loading as well as really trying to develop new redox chemistry that doesn’t involve expensive oxidants. This is done either by designing new redox catalysis where the oxidant is just thrown out of the equation, and we’ve done that with palladium (II) catalysts, or by trying to use air as the oxidant. In my opinion, that’s the future of many of the reactions; either using air as the oxidant, or removing the oxidant entirely. We’re also trying to further develop palladium (0) catalysts.
I hope that’s not an overwhelming amount of information, but it’s a problem I’ve been thinking about for a few years and my students are working on many ideas to solve these problems.
Q: The palladium must at least be significantly more efficient than a stoichiometric reaction, which is something we talk about, switching from stoichiometric to catalytic. And I’ve read that sometimes nickel can work well for catalysis.
A: If you can use nickel, that’s great. But for the transformations we want to achieve, it will take some time. We also published something on copper – using copper to replace palladium in C-H activation - but in reality it’s a long way before anyone gets even close to doing with nickel or copper what palladium can do. We can work on mimicking palladium, and eventually maybe palladium can be replaced in certain cases. On the other hand, palladium is not necessarily the problem. It’s comes down to how can we do this better.
As for the solvent, that’s a big question and something I have to admit that we haven’t really studied very hard. Although we do pay attention because we work closely with industry, and we know what kind of solvents they want. Our group has probably been the one to use many solvents for the first time for C–H activation. So we’ve developed a wide range of solvents for this application. Some of them are green, and some of them are not green. Most of the time, C-H activation reactions are stuck with one solvent like dichloroethane, and nobody wants to use that. It’s not at all green on a large scale. So we’re developing C–H activations that work with many solvents, and that gives industry a range to choose from. Industry can be very picky in what solvent they can use in process and that presents a challenge for reaction development in academic settings.
Q: What if you find that a reaction is very good for catalysis, but it’s impractical to scale-up?
A: The approach to that depends on the reason it’s not good for scale-up. Maybe the reaction transformation is great, but maybe it’s dangerous or the catalytic loading is too high. My honest approach would be that if a reaction transformation is very valuable and for some reason -- like safety or expense at the ton-scale -- it’s not practical, we would still be happy to publish it. Someone may still find it very valuable, especially medicinal chemists who love transformations that lead to a new compound. Then, I have to ask myself questions. What is the problem? Is the chemistry valuable enough to try and work out a solution to make it scalable? If it’s important enough, then I would work on it. I would develop better ligands and better catalysts to overcome the limitations.
Q: When you’re making new compounds, like the ligands you’ve mentioned, do you assess their toxicity or potential environmental impacts?
A: Honestly, for the field of C–H activation, it’s a luxury to have any ligand that actually helps the activation. Take hydrogenation for example; In the 1960s, hydrogenation didn’t have any ligands. Nobody knew what ligand will help hydrogenation before the discovery of triphenylphosphine ligands. It took several decades and remarkable efforts to eventually develop highly efficient chiral ligands for asymmetric hydrogenation. What I’m saying is that in C–H activation, we’re still at the level of hydrogenation in the 1960s. Any ligand for accelerating C–H activation is great; otherwise, we are simply doing background reaction from the viewpoint of asymmetric catalysis. Very few ligands are available to really help C–H activation and that sounds crazy, but it’s true. By really help, I mean significantly accelerate C–H activation. It’s just hard. At the moment, we don’t really think about testing for toxicity and things like that. We just want to look for new ligands, new understanding, seeing what kinds of ligand scaffolding can help. Then we’ll worry about environmental issues when we have really powerful and potentially broadly useful ligands. This will take a few more years, but they are on the way.
We have been extremely lucky, though. So far, all the ligands, have turned out to be environmentally very easy to handle. As I mentioned, they’re amino acids which are obviously not a problem, and they can be recycled and they’re biodegradable. So, we do think a little about it, but it’s not our priority at this stage. Moving the field forward is what we’re much more worried about. When I started it was already a big field, but in the last ten years it’s gotten to much busier, to say the least. Especially the last three years. But, in terms of solving big problems and finding broad applications, it’s just beginning. That’s my humble view, but I am not sure whether that is the consensus of the community.
Q: You’ve said that we need to do more to engage young people in chemistry. What do you think is needed to make that happen, and do you see green chemistry playing a role?
A: Green chemistry is a convincing concept for society to accept. But practically we need to do more that will directly drive young students into the lab. One approach is to remove the limitations that would prevent a potential young chemist from coming to the lab. So what are these limitations that are preventing young people from coming to the lab and doing green chemistry? In the end of the day, it’s funding. It would be great if someone – some charity, non-profit, or other organization – could set up more fellowships for funding students to do green chemistry. We have turned away so many talents because we don’t have the funding; Post-docs, undergrad, and graduate students. Many undergraduate students want to do a summer internship in my lab, and I can’t even afford more than one at the most.
People asked, how do we deal with things like infectious disease? And Bill Gates came in and said he’d set up a foundation to fund research into that. So for green chemistry, it’s the same. How many foundations are there funding green chemistry and giving fellowships? Without funding, it just won’t happen. At the moment the funding is so tight that not many people are doing this kind of basic chemistry. Green chemistry is just a concept or principle, but it will take basic, fundamental chemistry research to solve the problems.
Q: What do you think of using biosynthetic methods in chemistry, like engineering microbes? It seems to be growing very quickly.
A: Oh, that’s very interesting because I was a biosynthetic Ph.D. student so I have seen both sides. The new genetic tools are providing unprecedented opportunities to explore biosynthetic pathways. New tools help us to understand these biosynthetic routes much faster and better. But what those tools can do is unlikely going to replace what synthetic organic chemists can do. It’s very different. Enzyme catalysis has a long history for process. If you have a specific known compound that you want to process you can use enzyme catalysis for a key step in some cases. These days, enzymatic hydrogenation and reductive amination are very powerful. But that’s not to say we can replace many, many other transformations like those that organic synthetic chemistry can create, such as metathesis. They don’t have the diversity or reaction scope. The biosynthetic reaction scope is complimentary to what chemists can do. Fundamentally, the disconnections organic chemists create are often different from what nature does.
Q: What’s your advice to students who are frustrated or feeling stuck with their research?
A: I wish I knew the answer. Of course you could advise them to take a break, go to the bar and have a beer, but that’s not necessarily going to help. When someone’s frustrated that obviously means the project’s not working. I would take a step back and ask hard questions. Why isn’t it working? I’d re-evaluate the approach or the problem that I was working on. If a student has doubts after serious thinking, I’d advise them to change projects. It’s not a bad thing to do, particularly in the early stages like the first three years. I don’t usually advise my students to stick with one project forever. I always give them a couple of choices. When they stop working on one project and start working on another, it’s interesting because a couple of months later they might come up with a different idea for the first one. Some other colleague will have an idea or something will happen in the lab that will benefit their project. It’s a good strategy to take a break from a project for a few months or a few weeks, but not really giving up, rather continuing to think about it. That’s usually how it works in my lab. Sometimes a new idea will just show up without warning.
The 20th Annual GC&E Conference will be held on June 14-16, 2016 in Portland, Oregon. For more information, visit www.gcande.org.
The opinions expressed in this interview are Dr. Yu’s alone and don’t necessarily represent the views of the ACS GCI.