[Nomeru]

I recently completed an unknowns lab. I was given an unknown, and list of possible aldehydes and ketones. I am certain at this point as to what it is (4-hydroxy-3-methoxybenzaldehyde aka vanillin), but I cannot explain why it gave a negative result for both tollens and schiffs test when ran.

Evidance I have for it being vanillin: 1st, the strong scent of vanilla. 2nd: melting point of 81.9-83.6C (Lit. 82C), it's semicarbizone derivative MP was 224.7-225.5C, and it's 2,4-dinitrohydrozone derivative MP was 267.5-269.1C (Literature for both was 230C and 271C respectfully). IR also matches very well with literature IR.

Info on tollens test: 1ml 5% aqueous silver nitrate w/ 0.5ml NaOH, Added about 1.5-2mL 10% ammonia, mixedm abd dukyted to 5mL w/ DI H2O. Test tubes also sat a couple minutes with NaOH. started with 0.5mL tollens soln and 10-20mg vanillin. Heated, saw nothing. Added a little more tollens soln and more vanillin, still saw nothing. It was in warm water bath for about 30 minutes. At the end, it was perhaps slightly darkened, but certainly no silver mirror. Positive control had turned black+silver. Another positive was the same. Also something I found interesting: Our positive control, which worked fine, was 4-methoxybenzaldehyde. Not too unsimilar from vanillin, but a very different result.

Schiffs test: I didnt run this personally, but what I know is this: it initially turned yellow, and when heat was applied, it turned pink/salmon, but reverted to yellow when cooled.

All the data points very strongly to it being vanillin, but I cannot help but wonder why these tests would come out negative.

Thanks for any input or ideas.

[orgopete]

IR also matches very well with literature IR.

The wording on this is not what I asked of students. Does your spectrum match a literature spectrum "peak for peak"? (I cannot ask the spectra to be superimposable, the concentrations will differ, the instruments may be linear or log, etc. They should match peak for peak, everywhere the literature spectrum has a peak, shoulder, ratio in sizes, or any feature, your spectrum should match peak for peak.)

[discodermolide]

I would choose a different positive control, 4-hydroxy benzaldehyde perhaps.

[Nomeru]

The wording on this is not what I asked of students. Does your spectrum match a literature spectrum "peak for peak"? (I cannot ask the spectra to be superimposable, the concentrations will differ, the instruments may be linear or log, etc. They should match peak for peak, everywhere the literature spectrum has a peak, shoulder, ratio in sizes, or any feature, your spectrum should match peak for peak.)

What I see: Strong OH beak on both (mine's 3187, lit is 3184), mine has 2 peaks at about 2845 and 2735, while lit has them at 2836 and 2742. These indicate to me it's an aldehyde at least. Ignore the 2 peaks at about 2300, I'm told that's CO2. Lit next has a carbonyl peak at 1672 (mine's at 1660), 1590-1598 on lit vs my 1587.
It may not be superimposable (though it's not far off), but outside of those CO2 peaks, I dont see any differences. Peaks are in all the same places, and are all very similar in size.

How about I let you see and judge it yourself. I'm sorry about the quality of my IR, the scan wasn't perfect. Also, the labels are not all in the right spot, but they are close enough to get a pretty good idea.

My unknown (KBr)



Literature IR (KBr disk) from SDBS

[orgopete]

I compared your spectrum with the IR in SDBS and they do match, peak for peak. I began looking at the peak position and relative intensity from the longest wavelengths and up. Although the exact numbers differed, that is okay, the difference remained constant and the intensity patterns were the same. I too would conclude this was vanillin from this IR spectrum.

Back to the original question about the Tollens test. What I would be concerned with in the reaction is the NaOH and the phenolic functionality. I could imagine the phenoxide might form and alter the reaction. Your comparison compound cannot ionize and this could give a positive test while vanillin might not. If you draw the mechanism out, I think you may find the phenoxide may reduce the acidity of the CH of a needed
CH-O-Ag bond. It can also affect the formation of the hydrate. The carbonyl groups of aromatics are already less reactive as seen in the IR shift to lower frequency. A phenoxide should further reduce its reactivity. Just guessing.

[discodermolide]

This is why I suggested using a hydroxybenzaldehyde as a test system!

[Nomeru]

Well I cannot run the test again, but thank you for at least giving me a reasonable explanation as to why I would get these results.