Every morning I deal with my editorial duties for Organic and Biomolecular Chemistry, an RSC publication (http://pubs.rsc.org/en/journals/journalissues/ob#!recentarticles&all). There is one particular type of manuscript I reject without even thinking twice about it and certainly without needlessly bugging the already overwhelmed refereeing system. I want to talk about it today because this problem correlates with a continuing issue I have and will always have while training my own students. Here is a dramatic oversimplification of what I saw today (I removed the structural details, the only real thing is the yield):
There was no other information supplied, just a table with similarly low yields. There was also no attempt to discuss the factors affecting the reaction efficiency. So what does this number (28%) mean? What are we to learn from it? I do not have any OBJECTIVE parameters in front of me and I cannot, therefore, trust this chemistry. The removal of subjective factors from research is critical. Here is the main rule I have been fighting hard for in my own lab: have a good handle on the assay yield, selectivity, and conversion. Let me explain this in more detail. In short, selectivity and conversion are the metrics that enable a simple calculation of the assay yield:
I will give an example. If 90% of the starting material is converted in a given reaction, but only 80% of it actually goes on to form the desired product, we have:
Assay yield: 90% x 80% = 72%
It is called “assay” because conversion and selectivity come from an internal standard-calibrated measurement, or assay (using GC, HPLC, you name it). While every process chemist in industry knows about the significance of an assay yield, academics tend to turn a blind eye on this critical parameter. It is important to know conversion and selectivity, both of which can be obtained by monitoring product formation/starting material disappearance vs an internal standard. Without knowing these metrics, we do not know what a given number means. Say we see a reported 28% isolated yield. What does this number mean? No one knows. But, if you can: a. get a number for conversion (just look at the analytics using an appropriate internal standard) and b. get a number for selectivity (just calculate the amount of the starting material that went into the desired product by measuring its molar amount, using an internal standard), you will get an assay yield which is calculated as conversion x selectivity. If the assay yield is 70% and you get an isolated yield of 28%, you know what to do – improve your isolation skills. If you are reporting an isolated yield is 80%, you know that you are lying. On the other hand, if the assay yield is 25%, then you are doing as good a job as you can while dealing with a tough problem to crack. However, your reaction sucks because the selectivity is likely low (or conversion, or both – regardless, go back to lab). By measuring the assay yield, we confront the clear and present danger of purification inadequacies in our efforts. We might also have a “to do list” in terms of understanding the mechanism and improving the reaction. As John Hartwig puts it, “yield is a relative rates phenomenon”. So – let’s remove subjective factors out of chemistry. Otherwise there will always be people, who will report great isolated yields and say “…well, it worked in my hands” while no one else can repeat their work. There aren’t many luminaries with unique and magical hands in chemistry who have a patented “green thumb” technique that others lack. So – show me the numbers.
Pingback: Amphoteros – A Blogging Professor | Chemtips
Since I read this post more than one year ago, I’ve been thinking of it every time I got unexpectedly low yield. Is it conversion or selectivity? Or maybe troubles with isolation? Seriously, it haunts me all the time, but I guess it’s for my own good.
Anyway, I had a colleague who had always reported isolated yield 5-10% lower than the actual one. He was really afraid that someone repeating his synthesis in future gets lower yield and then thinks of his work as of ‘unfair’ or at least ‘requiring mojo’.
Right… But gathering objective metrics is what will always keep you safe. So get that calibration curve and measure chemoselectivity!