Reaction development: A checklist (Part 3)

Citation:

Tyler, J.L.; Trauner, D.; Glorius, F. Reaction development: a student's checklist. Chem. Soc. Rev. ASAP. https://pubs.rsc.org/en/content/articlelanding/2025/cs/d4cs01046a

Summary Figure:

Background:

Finishing up this series on what goes into a method! Finally, I'll cover the scope and applications part of the paper, including my thoughts on what substrate scope should look like. 

6. Scope

(a) Is the substrate scope diverse?

(b) Have you demonstrated functional group tolerance?

(c) Have you defined the parameters of reactivity?

I love the way they break this into 3 different questions. In general, I really like the way Frank Glorius approaches the subject of scope, and I am a fan of the robustness screen he developed. 

I want to first tackle the question of functional group tolerance. There are a lot of different relevant functional groups. If you begin to include all the different heterocycles, it gets even more overwhelming. A quick list of (highly relevant!) functional groups that might appear in your standard scope table:

alcohol, ether, ketone/aldehyde, ester, carboxylic acid, silyl ether, acetal, sulfonate, epoxide, furan

And that's just the general classes of oxygen containing functional groups!

Not to mention the differences that may pop up between classes- primary and tertiary alcohols can be very different, and there are many flavors of silyl ethers that can be orthogonally protected. 

The problem with testing functional group tolerance is that making compounds that have these functional groups on them takes a significant amount of time, money, and/or labor. 

I'm going to pick on this paper published today, because (a) it's a generic paper that illustrates my point and (b) I like carboranes. https://pubs.acs.org/doi/10.1021/acs.orglett.5c00059 

In general, this is a fairly standard Heck reaction, made interesting because the electrophile is the B-H bond of a carborane, and it is selective for usually one position of the 10 possible B-H bonds. (Carboranes are cool!) The scope mostly examines the different alkene partners that can be used in the Heck reaction. I'm going to go through this excruciatingly slowly. 

1. Styrene (default substrate)

2. m-methylstyrene

What does this substrate test? Is it really expected that adding a methyl group to the styrene will significantly change its reactivity?

3. p-tBu-styrene 

Charitably, this is an inductively electron donating substituent, and has steric bulk?

4. p-fluorostyrene

5. m-fluorostyrene

6. o-fluorostyrene

Is it truly necessary to test fluorination at all positions of the styrene? It is valuable to test at least one, as electron deficient styrenes may not react, but all 3?

7. o-chlorostyrene

Testing the halide series in a cross-coupling reaction is helpful, as aryl halides are competent electrophiles and can react. However, if you're going to test this, show where the reaction failed. If it doesn't work with the bromide, show it. 

8. nitro group

9. carboxylic acid

10. aldehyde

11. ester

12. isomeric ester

13. nitrile

17. amide

These are all valid functional groups that may or may not work in certain cross-couplings. No notes, great work. However, this still leaves out, from my earlier list:

alcohol, ether, silyl ether, acetal, sulfonate, epoxide, furan

They might have made it into JACS if they had done an additional 30 substrates, right?

I'll reiterate at this point that this is just checking for functional group compatibility. There's still another 2 sections to go for scope. 

The idea behind the robustness screen (https://www.nature.com/articles/nchem.1669) is that it does not matter if the functional group is actually attached to the molecule. If you make a substrate that has an arbitrarily long tether to different functional groups, that should react no differently than if you just threw in a separate small molecule with that functional group. There are 2 questions that need to be asked of each functional group:

1. Does the reaction take place in the presence of this functional group?

2. Does the functional group survive the reaction?

For example, you can have a functional group that shuts down the catalyst but is not consumed in the reaction. Nitro groups or terminal alkenes can bind to metal catalysts as ligands and then stop reactivity. They are not consumed in the reaction, but no product will be generated. On the other hand, you can have functional groups that react, but don't stop product from being formed. For a Kumada cross-coupling, if you have an aldehyde on your substrate it will react with the organomagnesium reagent. If you throw in enough equivalents of the Grignard reagent, then the cross-coupling will also take place. 

To answer these questions, the robustness screen measures (1) the yield of the product and (2) the recovery of the additive with the relevant functional group. 

In conclusion, I really like the robustness screen for testing functional group compatibility. As a specific tool, it does a good job of saying if a given functional group is likely to survive the reaction without interfering with the reaction. 

It does NOT answer the question, "Will this work with my substrate?"

In general, when you need to use an existing method for your chemistry, you go to the literature to find the closest analogue and trust that it will work as long as it is similar enough. For that, you need structural similarity. For example, a recent OPRD article lamented that there were very few examples of cross-electrophile coupling reactions of benzylic halides with cyclic benzylic halides. (Paper is: https://pubs.acs.org/doi/10.1021/acs.oprd.4c00497, I strongly considered doing a full blog post on this before going on this long tangent). There are many, many XEC reactions with benzylic halides. The fact that when people in industry needed to use one they could not find a relevant precedent is a massive failure. 

Returning to the carborane example: 

How many different structures are examined here?

18 examines a vinyl phosphonate, which is fantastic.

19 examines an acrylate, also fantastic. These cover the electron deficient alkenes, in comparison to the electron rich styrenes. 

21 examines an allyl ester, which is very intersting. 

22 examines a 1,1-disubstituted system

23 examines a 1,2-disubstituted system

26 examines a 1,2,2-trisubstituted system. These systems all have different sterics, so I do think it is good to include all 3. 

The goal for every substrate is to answer a question someone might ask. If the substrate does not answer a question, then the question needs to be, "why include this?"

The best parts of substrate scopes are ones that "define the parameters of reactivity" and I really like how the authors use that phrase. If you can say with confidence "electron rich substituents work, electron poor ones do not" that is an incredibly useful statement for anyone who wants to use your work. Returning to the halide screen I discussed earlier, being able to say "aryl fluorides and chlorides work, aryl bromides and iodides do not" is a powerful statement because it lets people plan around it. 

Finally, we get to part 7 of the paper, applications:

(a) Can it be used to access target compounds?

(b) Can it be used for late-stage transformations?

(c) Can the products be derivatised?

(d) Can it be used for bioconjugation?

(e) Can it be used in materials chemistry?

Ultimately, this entire section boils down to "Can other chemists use this chemistry to make things of benefit to the world."

Therefore, I think that a, b, and c all are trying to say the same thing: can this reaction make drugs. The target compounds? Drugs. The late stage transformations? Late stage transformations of drugs. Derivatized? Derivatized INTO DRUGS. 

There are thousands of currently existing drugs or drug candidates. If you cannot find one of them that seems related to your chemistry, then you may have made a useless motif. I am absolutely guilty of this. I'll admit it: I derivatized a couple steroids because they look fancy. The products that I made are not and will never be useful. We need to stop encouraging this behavior. 

I agree that finding interesting and more wild applications of the chemistry (bioconugation, materials) is a valid thing to show off. If the chemistry can be used in an application far away from what it was designed for, then it is possibly useful in many more situations. Obviously, these two are insanely hard to do- when your example is Nobel prize winning chemistry, it's hard for your average grad student to replicate that. 

Summary and Takeaways:

The goal of a substrate scope is to allow other people to use your chemistry. It is not to show off how many different molecules you can make, it is so other people can look for molecules that look like the ones they are trying to make. It does not matter that you made 20 compounds if they all look the same.

To conclude this 3 part story, here is my blueprint for what the process of making a methods paper should look like. 

1. Discover a reaction through careful thought and hard work using a combination of studying the literature, designing a reaction from elementary steps, identifying unusual side reactivity, and a healthy dose of luck

2. Identify the failure points of the reaction to figure out what is truly necessary and optimize to improve the yield and selectivity. 

3. Probe how the reaction works so you and other researchers can develop similar reactions by analogy. This will also help to improve the reaction and define what is necessary for the reaction to succeed. 

4. Make a library of example compounds so other researchers can identify if the reaction will be useful for them. 

The purpose of a scientific communication is to tell other researchers about your work so they can use it.  

I'm glad I broke this up into 3 parts, which feels mostly like a rant about what a methods paper looks like. I think Trauner and Glorius have some very good ideas about how to go about designing a methods project (no shit), and identifying places where I agree or disagree with the specifics of their argument leads to an interesting conversation. If you have thoughts on any part of this series, leave a comment and join in!