Sciencemadness Discussion Board - The credibility of syntheses around?

The credibility of syntheses around?

Refinery - 1-5-2020 at 05:29

How much of the syntheses floating around the internet and classic media are credible? I've found out that many process descriptions are either misinterpreted or outright false, or they produce significantly lower yields or totally other compounds than prescribed. There are also a plethora of concept reactions which are highly unproved.

(Of course, not speaking of operator errors including impure/false reagents which in reality can be a much bigger factor)

Have you faced with this issue, and are there legitimate sources that are highly credible? The issue not only wastes time and reagents, but it can be outright dangerous to perform these reactions.

Apparently, highly professional and institutional chemists have better capability of dealing with good practices and they have resources to study even the most intrigue reactions, and therefore this is more of a matter for us amateurs that must many times rely on second hand processes.

mackolol - 1-5-2020 at 05:49

Most of theoretical syntheses, that are not tested are not credible. Usually, when you perform such synthesis, you come to the point when you have to change something and you must figure it out by yourself or by looking into literature.

Typical example of such synthesis is the one from erowid, which carries MPV reduction of cyclopentanone, which isn't even enough diluted, and then neutralises all stuff with 50% H2SO4...

I don't know how this synthesis were written, I assume that on some drugs, but in the most popular book that talks about MPV reduction it's clearly said, that particular cyclopentanone is very delicate and if one carries this reduction with it must dilute it extremely and even though the yields will be pretty low....

So if it's not clearly marked, that synthesis has been performed in reality, you'll probably end up with few times work than you predicted and even then its not guaranteed that you will even make it. Ideally if synthesis has some photos in it, but that's not always the case..

Sigmatropic - 1-5-2020 at 08:21

I'd consider orgsyn.org to have some of the more credible syntheses floating around on the internet. Presumably due to the fact that they have two independent labs carry out the synthesis before publication. Should become a standard practice imo.

So many scientific papers cannot be reproduced, it's a shame.

Impure reagents are a thing of all time and can be for the better and worse. I would recommend reading about the dispute about the first quinine total synthesis, hint Impure reagents make the reaction chooch). I surely wouldn't call that operator mistake, and it happens all too often to professional chemists.

mackolol - 1-5-2020 at 09:46

Oh yes and speaking about websites that are credible I can also recommend prepchem.com
Also youtube or any videos are very good as they show the whole process of synthesis, the ones that are specially noteworthy:
chemplayer, nilered, nurdrage, extractions&ire, cody's lab...

Texium - 1-5-2020 at 09:49

Oddly enough I actually encountered more problems with literature procedures in the university research lab I worked in than I ever did at home. This is likely because the reactions I ran at work were usually more sensitive and complicated than anything I’d do at home, but it still goes to show that even with optimal conditions, replicating what should be working procedures will not go as smoothly as expected.

Three years ago, I unsuccessfully tried to run a reaction several times based on a reputable journal article that claimed nearly quantitative yield for it, and I had no idea what was going on. I tried emailing the authors of the paper looking for some insight, and never got a reply. My PI and an old friend of his who worked in the lab at the time, both experienced veterans of synthetic chemistry, also attempted the reaction after me, and failed as well. The fact that we couldn’t get that reaction to work resulted in us having to scrap plans for making several analogs of the drug we were studying.

morganbw - 1-5-2020 at 14:26

I have come across a couple of synthesis that I could not replicate. I usually conclude that there was a step or reagent that was not mentioned in the paper.
I am also highly aware that my skill might not be up to par on some reactions.

I am pretty sure that some papers are worthless but I lack the confidence to state so definitely.
I have royally screwed up simple/basic stuff so it is hard for me to critique a serious synthesis.

Just because I cannot or you cannot make it work does not mean that it cannot happen/work.

Heptylene - 1-5-2020 at 15:02

I have struggled myself with syntheses that provided insufficient information to be reproduced. In fact, a lot of the synthesis papers I read nowadays seem to lack some experimental information and it infuriates me :mad:

. I wish people would write a detailled procedure everytime, as in orgsyn papers, instead of the usual non-descript bullshit like:
"1 (0.01 mmol) in THF (1 ml) was added to 2 (0.02 mmol) in THF (1 ml) and stirred overnight at RT. The product was isolated by filtration in quantitative yield."

Sometimes I use Reaxys (organic reactions database) and it's filled with 99+% yields. That's a major red flag to me regarding credibility. Maybe I just suck at workups, or my reactions don't work, but a 99% yield sounds like their product is either not dry, not pure, or they are just lying.

While I'm at it, what's up with using novel, facile, serendipitous, mechanistic insight and other shitty buzzwords in every damn paper title?

karlos³ - 1-5-2020 at 15:15

Quote: Originally posted by Heptylene

99+% yields. That's a major red flag to me regarding credibility. Maybe I just suck at workups, or my reactions don't work, but a 99% yield sounds like their product is either not dry, not pure, or they are just lying.

It is likely they don't mean isolated yield, and simply have analytically measured a 99% yield.
Either this, or they are exaggerating, I would agree on that.
It does take quite much attention and good lab techniques to isolate a quantity corresponding to a yield of 90%, of course likewise to make.

Refinery - 1-5-2020 at 23:27

And then there are these outright foolish and even life-threatening synthesis, where "just simple add 2.5kg of this and 7.5kg of that, quickly place a bucket over the vessel and wait for the explosive reaction of cyanidious nitroglycerol to dissipate and reflux for few hours".

Like, a child could conclude that by slowly mixing the reagents could result in a safer, cleaner and better yield with much less losses.

mackolol - 2-5-2020 at 00:48

Quote: Originally posted by Heptylene

I wish people would write a detailled procedure everytime, as in orgsyn papers, instead of the usual non-descript bullshit like:
"1 (0.01 mmol) in THF (1 ml) was added to 2 (0.02 mmol) in THF (1 ml) and stirred overnight at RT. The product was isolated by filtration in quantitative yield."

Sometimes I use Reaxys (organic reactions database) and it's filled with 99+% yields. That's a major red flag to me regarding credibility. Maybe I just suck at workups, or my reactions don't work, but a 99% yield sounds like their product is either not dry, not pure, or they are just lying.

Quote: Originally posted by karlos³

It is likely they don't mean isolated yield, and simply have analytically measured a 99% yield.

Quote: Originally posted by Refinery

And then there are these outright foolish and even life-threatening synthesis, where "just simple add 2.5kg of this and 7.5kg of that, quickly place a bucket over the vessel and wait for the explosive reaction of cyanidious nitroglycerol to dissipate and reflux for few hours".

It's typical for patents. They just give incomplete information, workup at super big scale, often using some very fancy lab equipment. Usually they don't even bother to isolate the product and give the yield basing on chromatography or whatever of mother liquor contents.

What's even worse about patents is that they give very good and interesting routes to hard to reach compounds, but you must be very lucky or experiment a lot to get what you want by following the patent. Usually it just turns you that they don't work or you don't know what to do at some point and end up pissed out.
And they don't even bother if the chemical is nasty or sensitive. Just pure theory and theory is totally different thing than practice in chemistry.

[Edited on 2-5-2020 by mackolol]

karlos³ - 2-5-2020 at 07:33

A patent has to be that way, I mean it is also meant for protection, they can't give absolutely clear instructions.
Especially when some nations don't really care about patent law at all, then the vagueness of it are the only form to protect the ownership.
But we can be lucky with the patents nowadays, have you had a look at pre-WWII patents?
And for that matter, maybe ever seen a japanese patent out of that time?
Those are all rather vague, obviously obfusciated, not unusual you end up with more questions than you had before.

Sigmatropic - 2-5-2020 at 10:07

Pre ww2 patents from Germany are actually pretty good, maybe some pretty hardcore chemistry involved but it sure as hell worked.

Sigmatropic - 2-5-2020 at 10:16

Quote: Originally posted by mackolol

Oh yes and speaking about websites that are credible I can also recommend prepchem.com
Also youtube or any videos are very good as they show the whole process of synthesis, the ones that are specially noteworthy:
chemplayer, nilered, nurdrage, extractions&ire, cody's lab...

Prepchem is some sort of unreferenced compendium similar to Vogels, most of it works but not everything should be taken at face value.

YouTube videos show the process but all too often fail to follow up on product characterization (and also isolation and purification). I seem to remember a dispute about one of Chem players videos of making succinic acid from glutamate, so even there it's not always crystal clear( http://www.sciencemadness.org/talk/viewthread.php?tid=69635&... )

mackolol - 2-5-2020 at 13:22

Quote: Originally posted by karlos³

nd for that matter, maybe ever seen a japanese patent out of that time?
Those are all rather vague, obviously obfusciated, not unusual you end up with more questions than you had before.

Lately, I have performed deethylation of ethylvanillin to protocatechualdehyde following japanese patent. It was quite brief and understandable, but the thing that pissed me off was, that they ended the synthesis at the state of post reaction mixture and stated that yield is quantitative taking in mind chromatography of mother liquor.
And they didn't mention single way of purification or extration. I had to take that knowledge from another english patent. Even though it turned out that this particular aldehyde has much poorer solubility in Ethyl acetate than expected

(strange thing because the first extraction with EA was very abundant in color, but following extractions were light in color and I couldn't extract the water layer in such way that it became decolorised. After evaporation of solvent it turned out that the product is in 50% yield, which would mean that another half is left in mother liquor). Anyways, when I'll figure it out better I can share results.

AvBaeyer - 2-5-2020 at 19:30

There appears to be general misunderstanding here about experimental procedures in patents. First off, patents are not scientific papers. A patent only has to teach a person "skilled in the art" how to reproduce what is in the claims. For example, if a product yield for a reaction is not a part of any claim, it is not considered a critical piece of information. It assumed that enough information is provided that a "skilled practitioner" will know how to make the reaction product even if not every detail is provided. Depending on the "skill" of the practitioner, a reported yield may or may not be achieved.

I was an expert witness for several years for lawyers challenging or defending chemistry patents in court. In the case of chemical procedures, all the patent owner had to show was that the reaction product could be "found" irrespective of any reported yield. Typically, this was demonstrated by both parties in the case agreeing to have a person "skilled in the art" run the disputed reaction and determine the product(s) formed.

A bigger problem is the irreproducability of procedures in published papers. I was a reviewer/referee for several journals years ago. Many times experimental sections were sloppily written and had to be sent back for rewriting and clarification. More often, experimental procedures in papers describe only a single run of the experiment so there can be no information about yield ranges or side products. There is not much that can be done about these problems unless paper authors learn to take some pride in providing well written experimental sections and referees pay real attention to experiment descriptions (that is a tedious exercise.)

Just my nickels worth.

AvB

mackolol - 3-5-2020 at 02:47

Quote: Originally posted by AvBaeyer

There appears to be general misunderstanding here about experimental procedures in patents. First off, patents are not scientific papers. A patent only has to teach a person "skilled in the art" how to reproduce what is in the claims. For example, if a product yield for a reaction is not a part of any claim, it is not considered a critical piece of information. It assumed that enough information is provided that a "skilled practitioner" will know how to make the reaction product even if not every detail is provided. Depending on the "skill" of the practitioner, a reported yield may or may not be achieved.

Okay, maybe I said it in wrong words, patents are credible, but they aren't the easiest and most practical source of knowledge for amateur chemist.

beeology - 17-1-2022 at 02:00

Quote: Originally posted by mackolol

hello, do you have any update of this work and can you give us the number of the other patent? thank you very much.

valeg96 - 19-1-2022 at 12:14

I've personally found so far that many literature procedures aren't really false or exaggerated, they are just oversimplified. I myself have tried repeating "official" procedures handed down to me multiple times to no avail, only to find out from looking at the actual lab books that they had been attempted a lot of times and with a huge series of cul-de-sacs and changes of mind during the work-up and the setup, which in the end are sheared off so it all fits into two lines of text.

Of course when you carry out a synthesis where you change 5 times the solvent during work-up with all the included evaporations, filtrations and changes of vessel, or a weird emulsion forms and you try 10 different ways to break it down, you can't write it all in detail in the article. On the other hand it's very frustrating when they just trim it down to "the product was purified by column chromatography, yield 97%".

Sometimes it's also poor bookkeping from the bottom rows, though. Happened to me recently with an article procedure that said to use DCM for the column, in which the product (I discovered at my demise) didn't move an inch (cue 2+ liters of DCM to flush out everything and literally every solvent in the laboratory employed to remove my expensive compound from the silica). Found the lab books and it turned out that it had been attempted like 20 times by multiple people and nobody bothered to write the solvent used in the column or any fundamental details. Also, I've noticed that inconsistent reporting introduces mistakes in the procedure, like the solvent suddenly changes or the temperature is different from one run to another.

A simple synthesis ended up being two pages long on my notebook because of inconsistent reporting and/or excessive trimming of the actual attempts. And unless I invest the time and money to repeat it again from scratch just to obtain a clean, linear path, the cycle will continue. Some people are more interested in having the compound and once it's done, they don't care about all this. It's the next person's problem, I guess! /j

WGTR - 19-1-2022 at 12:57

I developed a procedure once that required a washing step in DI water. My colleague attempted the same procedure and the process failed. Come to find out that he was washing the crap out of the sample for several minutes in the sink, and I was merely rinsing it with a few spritzes of DI water from a bottle. The faint contamination left over from my brief rinsing step actually made the procedure work. Of course this necessitated revisiting the procedure and a lot was learned during the process.

Sometimes while reading through the experimental section of a journal article, I will notice that some steps are described very briefly. If I am familiar enough with the topic then questions will develop while reading through the article. Sometimes these questions will not be answered by the article itself; it may take knowledge of other similar procedures to achieve success. It is as valeg96 described, the procedure is just oversimplified. It might seem like anyone should be able to replicate a result simply by working from a journal article, but it often takes someone "skilled in the art" that is already familiar with that type of procedure to be able to fill in the missing blanks and the "gotchas".

S.C. Wack - 19-1-2022 at 19:42

Not every author is wrong, unaware, or lying. There are nuances that are either difficult to put into concise form, or appear trivial when put into concise form. Patents and journals are not cookbooks. OS and IS are cookbooks, actual books...cookbook is a derisive term in chemistry, which kind of explains why things are the way they are.

The discoverer of hydrazine from urea was published in 2 patents (in 1903) and a 1905 journal article (with different conditions in each); the article (his last?) gives the only claim of yield, 60%. I can only find 3 people that mention those within the next 25 years; they do so in passing, and with a wave of their hand they say it's BS.

Then, another patent uses urea, claiming 60-65%. Then another, 60-70%. AFAIK no one here or in books or videos has said that they get 60% (based on urea); Mr. Anonymous is closest with 55.6%, everyone else gets 51% or less to much less. IS tossed it into the dustbin of history in their preparation from ammonia, "The only satisfactory method"; to them, the article was only a historical reference for the preparation and decomposition of benzalazine, and the urea patents to that point weren't mentioned at all.

In my hands, the average yield of sparkling clean (>99.5% volatile upon ignition) sulfate has been 61% lately, which can likely still be improved. IS got 35% based on hypochlorite with ammonia, which is easily doubled with urea.

When I first tried it, it did not go as well.

In any case, authors have little interest in helping other people/corporations make money off of their experimentation.

BTW a lot of the videos y'all worship (because it's on TV?) are garbage too.

Texium - 19-1-2022 at 20:08

Quote: Originally posted by valeg96

Of course when you carry out a synthesis where you change 5 times the solvent during work-up with all the included evaporations, filtrations and changes of vessel, or a weird emulsion forms and you try 10 different ways to break it down, you can't write it all in detail in the article. On the other hand it's very frustrating when they just trim it down to "the product was purified by column chromatography, yield 97%".

This is frustrating to be sure, and I can confirm that it can be frustrating for the one writing the procedure too... I have written up experimental procedures for publication, and my PI mandated that they follow a very particular, concise format. Details that I wanted to include had to be left out. Even if my PI didn't mandate that format, and we had submitted my preferred version, it's likely that the journal would have demanded similar revisions.