[Guitarmaniac86]

I have been performing a few direct CH activation reactions lately and the aryl bromide was added to a toluene so that its concentration was 0.2M. When I performed test reactions at 50-75mg I got a yield of 56% which was not great so I scaled up to 250mg making sure that the concentration was the same, 0.2M. I was expecting a drop off in yield because of the scale up but I got a massive jump in yield, the best being 89%

I have tried the reactions several times, and at 50 - 75mg I get a poor yield but at 250 - 500mg I get a very high yield. I haven't tried a gram scale reaction but I suspect I will get a drop in yield if I tried this.

I am at a loss for an explanation. I would expect at higher scales you get less efficient heating of the reaction mixture so the reaction becomes less efficient which affects yields. By this logic, the smaller scale reaction should have the highest yield as it is easier to get up to temperature.

Anyone suggest a good book or paper that explains reaction scale ups. I am sure I am missing something fundamental.

Many thanks.

[discodermolide]

Why should you expect a drop of yield upon scale-up. Everything is the same. I would suggest that the 56% yield is the one that is wrong, and you did something different upon scale-up.
Remember that weighing and measurement in general is easier and more accurate on a larger scale.

[Guitarmaniac86]

Why should you expect a drop of yield upon scale-up. Everything is the same. I would suggest that the 56% yield is the one that is wrong, and you did something different upon scale-up.
Remember that weighing and measurement in general is easier and more accurate on a larger scale.

I would expect a drop off in yield because of less efficient heating of the solvent and greater degrees of temperature variation from top to bottom of the reaction vessel. However at the scales I am using these probably aren't an issue (I have no idea what I am talking about lol).

[discodermolide]

At your scale this is certainly not an issue. Even on a large scale this is not an issue. I always got better yields in the plant than in the lab. Because the equipment was better and validated.

Something went wrong with the first experiment to give you 56% yield.

[Guitarmaniac86]

At your scale this is certainly not an issue. Even on a large scale this is not an issue. I always got better yields in the plant than in the lab. Because the equipment was better and validated.

Something went wrong with the first experiment to give you 56% yield.

I agree with you that there was a weighing error because I did struggle to weigh such a tiny amount of mass.

[NotExactly]

It is not uncommon that reactions in the scale you are describing (50-75 mg) are lower simply due to the inherent challenges associated with working with less material.  More material can be lost simply in the workup and purification process.  Furthermore, reaction conditions need to be much more strict for a small scale reaction because the influence of things such as trace amounts of water or impurities from your glassware, reagents, gas, etc. is much higher.  A personal example:  whenever I do enolate chemistry on a small scale (20-100 mg) I take great care to make sure my glassware is completely dry, but on larger scale, in general, this becomes less important.  Just my two cents.

[discodermolide]

At your scale this is certainly not an issue. Even on a large scale this is not an issue. I always got better yields in the plant than in the lab. Because the equipment was better and validated.

Something went wrong with the first experiment to give you 56% yield.

I agree with you that there was a weighing error because I did struggle to weigh such a tiny amount of mass.

Well there is your answer. Do larger scale experiments.

[curiouscat]

I always got better yields in the plant than in the lab. Because the equipment was better and validated.

That statement was very surprising to me (not doubting your assertion; but doubting my knowledge). I always assumed mixing, heating, cooling etc. are Unit Operations easier to do in the lab than in pilot and easier in pilot than on scale.

I don't have many years of experience with scale up; yet as an Engineer the processes I've seen were mostly always performing better in the lab than at pilot or scale.

Did you get better yields in the actual process plant than the pilot-plant too? That'd really upset my world-view. 

[discodermolide]

The yields were always better in the plant and manufacturing. Simply because the conditions are constant. I mean the equipment is validated, and when you run a lab optimised reaction in the plant it runs better. Heating, cooling, stirring etc are all the same for each batch so you get a consistency that you can't achieve in the lab. Have a look at the post by Quintus at http://www.chemistry-blog.com

Anyway you must be consistent, the FDA demands it. Too much variation in yield shows that you don't have the reaction parameters under control. The process must be validated i.e. be able to accept variations in reaction parameters without affecting quality or quantity.

[curiouscat]

Anyway you must be consistent, the FDA demands it. Too much variation in yield shows that you don't have the reaction parameters under control. The process must be validated i.e. be able to accept variations in reaction parameters without affecting quality or quantity.

Thanks for the link to the blog. Interesting reading.

I've never dealt with the FDA. Most of our processes (bulk chem / petro) were surely not very robust. A parameter variation mean we were losing serious money somewhere.

[discodermolide]

I'm sure the processes are robust but you may not have complete control over some of them. It is a bit easier in the pharma business to control the reaction parameters and you can quantify them. In the petrochemical world it is very different, but as you say you still need robust processes. But I think you have more variables to control, making it more difficult.