I'd like to know if there are any (free or paid) softwares to predict solubility of a compound in organic solvents based on its structure.
I know it's possible to calculate this by hand using Abraham solvent equations and water solubility tables, however I thought it'd be plausible that
this process was automated somehow but I don't know about it yet.
Thank you Nicodem - 9-2-2015 at 06:47
There is no such software that could be considered applicable (there are only some useless predictors), because no useful model yet exists.
Semiempirical models like the one you mentioned are of little use as they require measurements or model-based predictions that are more difficult to
make than a simple solubility determination. Which one do you think is easier?
a) determining the crystal lattice energy and the solvation energy, then calculating the solubility by thermodynamic equations;
b) using the unreliable solubility models and deriving the parameters from similarly unreliable models or difficult measurements;
c) measuring the solubility curve for a compound.
The option c gives actual experimental data and with the automatized experiment it can give reliable solubility curves overnight without any effort.
From the same experiment, you also get the metastability zone and with some additional effort you can also obtain some rudimentary data about the
dissolution kinetics. You can get similar information, even if you do the solubility measurement manually via old-fashion methods. Do you then wander
why nobody in experimental chemistry ever uses the computational models for solubility?kazaa81 - 9-2-2015 at 06:59
Your reply was indeed exhaustive.
Do you know any good database to look out solubilities of organic compounds in organic solvents?Nicodem - 9-2-2015 at 10:12
Reaxys is probably the best choice to look for solubility data.
Otherwise, there are compilations of data for some more common compounds (UTFSE to get more details - this topic has already been covered). The Merck
index commonly gives some basic solubility data for its indexed compounds.
The exact solubility information is rarely of use in experimental organic chemistry. Why would you want to know the exact solubility of a specific
compound in a specific solvent? The number of possible combinations is beyond practical use. Besides, this is commonly considered relatively
insignificant information for organic synthesis where you only want to know whether the compound is soluble enough for the solubility not to affect
the reaction (for this reason, descriptors like insoluble, poorly soluble, soluble, well soluble, miscible and similar ones are sufficient in
most cases). In practice, even the approximate solvents partition is a much more useful information (and the computational prediction models for this
are relatively good, at least for the n-octanol/water partition). On the other hand, you need solubility curves during process development
(for crystallization optimizations), but when in need of such data, it is gathered experimentally (predictions are of no use, as you would need to
confirm them empirically anyway). I seriously doubt any amateur chemist is going to get involved in process development for his hobby projects (as
then it would not be amateur chemistry any more), so I wonder what exactly are you looking for?blogfast25 - 9-2-2015 at 17:24
Reaxys is a paid for database. Nicodem, do you have any indication of cost per search (I don't know their pricing structure at all). I'm not willing
to 'ask inside' for fear of never getting rid of them after that.Nicodem - 10-2-2015 at 08:15
Reaxys is a paid for database. Nicodem, do you have any indication of cost per search (I don't know their pricing structure at all). I'm not willing
to 'ask inside' for fear of never getting rid of them after that.
Off course it is a paid service - kazaa81 explicitly said this is not an exclusion criteria.
It is not a "pay per search" service (as far as I know, no database offers any such payment conditions). It is a subscription service and the price
depends on the number of licences, academic vs. commercial application, and possibly also on your bargaining skills. I would estimate it is about a
dozen thousands EUR per year for non-commercial users, but I have no information on this as I never had to pay for it from my pocket.Mailinmypocket - 10-2-2015 at 08:49
In the references section there is a thread by kristofvagyok where he offers to do reaxys searches for members... Whether or not he still does this I
don't know. Not sure if I'd bother him just for solubility info though for the reasons Nicodem mentioned.blogfast25 - 10-2-2015 at 09:44
I would estimate it is about a dozen thousands EUR per year for non-commercial users, but I have no information on this as I never had to pay for it
from my pocket.
That's considerably more than I expected. Thank you.
