Sciencemadness Discussion Board

Reaction scales

prole - 14-10-2008 at 05:23

I'm trying to understand the scaling of reactions. If one has a procedure that yields, say, 200 g (theoretical) of Rochelle salt, but one only wishes to produce 10 g (theoretical), then to scale it down, one would divide the smaller amount by the larger amount, giving a decimal, and then multiply each amount of reactant and solvent in the reaction by this decimal, giving the amount of each to use in the newly adjusted reaction. And for scale up from 10 g to 200 g, I'd divide the larger by the smaller and multiply each reactant. Do I have it right? The textbook I have (for a class I'm not taking yet) vaguely describes scaling up in one sentence, and it seems to me that it's probably more complicated than this. For instance, I understand that scaling down usually produces smaller yields than expected, so excess reagents are required to drive the reaction to completion, or to make it happen at all. What else is there to consider when scaling reactions?

Oops, I meant to post this in Short Questions.

[Edited on 10/14/2008 by prole]

12AX7 - 14-10-2008 at 08:00

Ratios, almost everything is ratios. You don't need moles, you don't need weights, all you have are proportions. 200/10 = 20, so you scale down to 1/20th and you scale up by 20x.

I don't think smaller reactions give smaller yields. That sounds like carelessness on the part of the experimenter. Excess with respect to some reactant is probably more common with small reactions in order to increase what yield is there.

Tim

vulture - 14-10-2008 at 08:54

Smaller reaction scale usually means losses due to transfer and product sticking to vessels become more significant.

Nicodem - 15-10-2008 at 00:06

In most cases you can scale down just by factoring down.
However, only few reactions can be scaled up by just factoring up. In some cases, doing this can result in serious mishaps and injuries. Scaling up reactions is a tedious process.

prole - 15-10-2008 at 05:38

Nicodem, when you say 'serious mishaps and injuries', are you talking about the dangers in handling larger quantities of dangerous solvents and reagents, and the potential products and side reactions they produce, and possible runaways? I know that going from lab scale to industry scale involves some serious consideration, and even going from mg to multigram scale can get risky. What are some things to consider in a scale-up that can't be done by merely factoring?

I googled this first, and didn't notice anything helpful on the first couple pages. If there's a site you know of that explains this in detail, I would love to have a link to it.

@12AX7 and vulture: I know what you mean by losses and sloppy technique. In lab at uni, I'm usually among the last to finish, and often get the highest actuals. Nobody else seems to want to be there, and their work and results show it. School lab is the only time I get to spend in a lab these days, so I try to make the most of it. (On a sad note, last night we were testing electrical conductance on a variety of solutions, including abs. EtOH. More than a litre was trashed for the experiment. What a waste, it doesn't even conduct electricity).

Thanks for your replies.

Nicodem - 15-10-2008 at 05:49

Scaling up is the science behind chemical technology. Since I'm a chemist and not a technologist, I'm not the person competent to explain these things. There are however a few obvious things that need to always considered when scaling up. One is the heat flow which depends, among other things, of the ratio between surface and volume (just calculate the surface to volume ratio in dependence of the radius for a sphere or a cube, and you will see what I mean). This problem is generally the main source of runaway reactions - reactions which are innocuous at a mmol scale can just blow up on mol scale. The next problem is the mixing thing which is particularly important for heterogenous reaction. A good stirring is extremely important in scaling up to prevent local heating up and good mass transfer and contact between reactants. I guess some technologist will explain in more details also about other issues.

Panache - 15-10-2008 at 19:30

your attitude appears to be perfect, too gungho and its dangerous, not enough gungho and you never do anything a little bit unknown and hence learn far far less, especially from a practical point of view.
I have made so many many many mistakes, been covered so often in everything from liquified liver to wine to being embedded with glass shrapnel, however i have never been at risk in my work, (well depends on your definition of risk) as i always tried to assess the worst thing that could happen and have a plan for it, and never perform chemistry you don't understand from first principles. in reality my first job in industry ended with an explosion, two fatalities and some serious shit for Hoechst Aus, luckily i wasn't working that saturday, i say it because this was a brand new purpose built current (well 1990's) rigorous site and it still happened. You cannot be immune, learn to run fast Ha!
The ratio thing is pointy/non-pointy
as in 2g------>20, divide the pointy end(20g) by the non-pointy end(2g) and it will be correct, i can thank my first chemistry teacher or that little thing that has never left my head.

Hopefully i haven't been redundant in my information here.

Magpie - 15-10-2008 at 21:35

Nicodem says:
Quote:

Scaling up is the science behind chemical technology. Since I'm a chemist and not a technologist, I'm not the person competent to explain these things. There are however a few obvious things that need to always considered when scaling up. One is the heat flow which depends, among other things, of the ratio between surface and volume (just calculate the surface to volume ratio in dependence of the radius for a sphere or a cube, and you will see what I mean). This problem is generally the main source of runaway reactions - reactions which are innocuous at a mmol scale can just blow up on mol scale. The next problem is the mixing thing which is particularly important for heterogenous reaction. A good stirring is extremely important in scaling up to prevent local heating up and good mass transfer and contact between reactants. I guess some technologist will explain in more details also about other issues.


Nicodem, I think you have given an excellent summary of the scale-up problem for chemical reactions.

Scale-up analyses are indeed a critical aspect of the design of large scale equipment based on lab scale or even pilot plant scale data. In their effort to save time and money plant design project managers often skip the aquistion of sufficient intermediate scale data. Then the full-scale plant ends up not working, or working poorly, and millions of dollars are lost.

I checked my old text books for sections on dynamic similitude and model theory but didn't find much. Even Perry's Chemical Engineers' Handbook, 3rd, Ed, doesn't have a section on the subject, to my surprise.

One book I have says: "Model theory is dependent upon several criteria of similarity....Model theory may be stated as follows: If two models are geometrically, kinematically, and dynamically similar, all velocities and forces are in a constant relationship at counterpart positions."

Similarity is usually established by ensuring that the model and full scale design have the same "dimensionless numbers." One such dimensionless number is the Reynold's number, (rho)Dv/(mu), which is the density times the diameter times the velocity divided by the viscosity. This is used for fluid flow. There are several other such numbers, eg, Nusselt and grashof numbers, which apply to heat transfer, if IRCC. There is also such a number for molecular mass transport via diffusion.

Much of wind tunnel testing is based on dynamic similitude, I believe.

I think a good example of where similitude is poorly applied is in grade B movies, especially naval battles where models are used. They've scaled down the ships correctly, but not the water, and gravity. So it looks fakey.

I ran into a scaleup problem once in my career. We were trying to set parameters for a 1 gallon batch steam jacket heated starch converter so it would produce a converted starch with the same properties as a 3000 gallon batch steam sparged converter that the plant was using. Variables we could change were the agitator speed, amount of enzyme used, and the time-temperature schedule. We never did find conditions that gave good similitude.

As far as giving any advice to prole for scaling reactions, I really can't add anything to what has already been said.

[Edited on 15-10-2008 by Magpie]

[Edited on 16-10-2008 by Magpie]

watson.fawkes - 16-10-2008 at 09:47

Quote:
Originally posted by Nicodem
Scaling up is the science behind chemical technology.
You know, this would make a great little four-page introductory document for beginners, as part of an advocacy to start with microscale experiments.

@Nicodem: Not that I'm advocating <i>you</i> write it, but it comes up often enough that a referable document would be a boon.

Panache - 16-10-2008 at 23:50

may i just add a qualifier to my previous post in that i have never worked with energetics and my comments need to be considered in that context.
Gung-ho and energetics i imagine make a terrible mix.

Nicodem - 21-10-2008 at 00:01

Quote:
Originally posted by watson.fawkes
Quote:
Originally posted by Nicodem
Scaling up is the science behind chemical technology.
You know, this would make a great little four-page introductory document for beginners, as part of an advocacy to start with microscale experiments.

@Nicodem: Not that I'm advocating <i>you</i> write it, but it comes up often enough that a referable document would be a boon.

Yes, maybe someone should write something about it. Seems like too much beginners thoughtlessly scale up reactions that are not meant to be scaled up and this way risk mishaps. But I will not write about it because nobody takes my warnings seriously (but most importantly because I don't have the time to do so). Besides, I see there are technologists among us with great experience who are actually competent in describing the issue to the beginners who had no accidents yet, but are eager to get hurt.

I will just add this:

- Never scale up a reaction before really understanding the theory behind it (the heat of the reaction, the kinetics and activation energy, the mechanism and side products, the solvent and concentration effects, the addition order, the stirring, etc.).

- Never scale up exothermic reactions by just factoring up the reaction components. Such reactions need to be seriously modified for scale up.

- Never scale up autocatalytic or radical chain reactions.