Test method for measuring the solubilities of various compounds in organic solvents

This experiment is a test of the methods and ways of recording the data than actually focussing on measuring the solubilities of aldehydes, ketones and carboxylic acids. Once I have a method which I am happy with, I will move to actually testing compounds.


24 Eppendorfs were taken and split into two groups of 12. Group 1 were labelled with the planned contents and group 2 labelled with planned contents and weighed. [Do you have access to screw-cap tubes of some kind? As your adventures in wet gloves (below) illustrate, snap-cap tubes are fairly crap at keeping anything but aqueous solutions contained during mixing! --BH] I have swapped to screw cap tubes for later experiments. JH

Saturated solutions of glucose, mannitol, salt and vanillin were prepared by adding 1 mL of each solvent to it's correspondingly labelled Eppendorf and adding solid, mixing the contents by vortexing to ensure complete dissolution before addition of more solid until the solution was saturated. This gave 12 saturated solutions of glucose in ethanol, methanol and THF; mannitol in ethanol, methanol and THF; salt in ethanol, methanol and THF; and vanillin in ethanol, methanol and THF.
This method worked well for the barely soluble solutes, but did not work for the very soluble compounds as the volume of the solvent increased substantially as more solute was dissolved. This was particularly noticeable for the vanillin as my gloves became wet from lost solvent as the tube became full. When trying the experiment again, either a different tube size must be used or a different volume.

The saturated solutions were centrifuged for 1 minute at 13 200 rpm/16,100 g [How did you calculate the g? JCB ][unless I have the same machine you do, or know the radius of its rotor, I cannot use this number! best to give centrifugation values in multiples of g --BH] in order to ensure that the remaining solid was separate from the supernatant.
0.5 mL of each solution was transferred to it's corresponding weighed group 2 eppendorf.
The group 2 eppendorfs were transferred to the speedvac and dried under vacuum, initially for three hours. At this point, all the eppendorfs were weighed. The eppendorfs were returned to the speedvac to be dried to constant weight.
The eppendorfs were dried by speedvac overnight. This led to the drying of 7 of the 12 samples to constant mass. One sample had gained 0.1 mg in weight, but I think that this must be an error with the balance reading. The samples which had not dried to constant mass were again put on the speed vac. this tiime with the addition of heat, to drive the last traces of solvent off. A photo of the wet appearance of the vanillin was taken on my mobile phone. I will upload the photo to chemtools but I don't think the quality will be sufficient to see anything useful. Photos at http://chemtools.chem.soton.ac.uk/projects/blog/blogs.php/bit_id/23163

the experimental details can be found at http://spreadsheets.google.com/ccc?key=pUtF1N4vVD6dGvbeUgkLjfQ&hl=en_GB#v


Some of the other values are closer to those obtained previously in EXP207 - it would make sense to compare those in a discussion.

As mentioned above, this method seemed to work well for the solutes that were barely soluble in the given solvents. The method will need to be adapted again to ensure that there isn't a repeat of the spillage ensuing from the volume increasing beyond the capacity of the vessel.
Comparison of the solubilities gained in my experiment and those in EXP207 show mixed degrees of correlation. I used a greater amount of solvent (500 μL compared to 200 μL) but in several cases got a lower solubility reading. Perhaps this is due to adding solute to solvent compared to solvent to solute.
The results that gave good correlation were for sodium chloride in ethanol (insoluble), mannitol in THF (0.01M and 0.02 M), vanillin in ethanol (2.23M and 2.5 M) this value was marked as DONOTUSE for insufficient mixing and glucose in methanol (0.02 to 0.05 M). For vanillin in methanol and THF, I ended up with a lower solubility overall and a significantly different value. . Note that only boc-glycine was a gel in ethanol, as mentioned in the discussion of EXP207 - that was a mistake in the spreadsheet, based on double checking the images, and it is fixed now. So for vanillin in methanol we're looking at 24% different value (your 3.2 M vs. 4.2 in EXP207) and this value was marked as DONOTUSE for insufficient mixing JCB

The widely varying results are to be found in the negligibly soluble solutes. This perhaps is just due to a large error margin and will not completely correlate despite trying to address the issue with a larger volume of solvent.
In all, I think the method designed is sduitable for measuring solubilities with some slight adaptation.