Enigma: Failure to detect products of Ethanol oxidation

[Jumpshooter]

Control = untreated hand sanitizer gel. 5 g of gel. + 0.5 water, mixed an let sit for 5 days ''in parellel with the treatment sample".
Treatment = 1 N Hydrochloric Acid (0.5g) added to 5 g of gel; mixed and let sit for five days. During this time the Ethanol would theoretically be degraded/oxidized to acetaldehyde/ethanal and/or ethanoic acid.

After five days both control and treatment were brought to 100 mL with distilled water; injected 1 uL into GC.

GC parameters: 100% wax column, 40C isothermal for 6 minutes; ramped to 100C hold for 1 min; inlet psig= 10; inj = 200C; Det = 300C.

Results:
Control peak area: 500,000 for ethanol with no other peaks.
Treatment: 400,000 area count for ethanol with no other peaks.

Conclude: The ethanol was degraded/oxidized, as indicated by the ~20% drop in area counts, but no chromatographic evidence was observed.
Thus, the ethanol was converted to by products that evidently do not chromatograph. Why no by-product peaks?

[Jumpshooter]

Control = untreated hand sanitizer gel. 5 g of gel. + 0.5 water, mixed an let sit for 5 days ''in parellel with the treatment sample".
Treatment = 1 N Hydrochloric Acid (0.5g) added to 5 g of gel; mixed and let sit for five days. During this time the Ethanol would theoretically be degraded/oxidized to acetaldehyde/ethanal and/or ethanoic acid.

After five days both control and treatment were brought to 100 mL with distilled water; injected 1 uL into GC.

GC parameters: 100% wax column, 40C isothermal for 6 minutes; ramped to 100C hold for 1 min; inlet psig= 10; inj = 200C; Det = 300C.

Results:
Control: peak area of 500,000 for ethanol with no other peaks.
Treatment: 400,000 area count for ethanol with no other peaks.
Retention time: 2.5 mins.

Conclude: The ethanol was degraded/oxidized, as indicated by the ~20% drop in area counts, but no chromatographic evidence was observed.
Thus, the ethanol was converted to by products that evidently do not chromatograph. Why no by-product peaks?

[Don_Hilton]

1) Did you weigh the sameles before and after standing to demonstrate the ethanol did not evaporate?

2) Did you make multiple injections - to ensure against injetion errors. Without an internal standard, injection errors cause loss of peak area - and you don't get any indication why.

[chromatographer1]

Don has made a good point. Acetaldehyde is quite volatile. Also understand that it is also very reactive and will react with a great many things, including itself. You should determine the weight before and after and measure the loss of ethanol. But trying to measure acetaldehyde based on your experimental conditions and making the assumption that it does not react I suspect is misfounded. But I may be wrong.

best wishes,

Rodney George

Dear Jump, isn't the weather in KC this week unreal? I hate to have it like August in the beginning of June. My grass is growing like crazy !

[Don_Hilton]

A couple of other things crossed my mind here - What is the concentration of ethanol in the mix? And besides water and alcohol, what else is present?

I would not expect the addition of acid to particularly accelerate the air oxidation of ethanol at room temperature. But, I would expect it to accelerate things like transesterification reactions or other reactions that are catalyzed by the addition of protons. And, the question of what is the gelling agent becomes pertinent.

[Peter Apps]

Don's wet chemistry probably accounts for it, also with these small molecules oxidation causes a substantial loss in FID response per unit mass (I have to presume that you are using an FID).

Peter

[Jumpshooter]

Keep in mind that I am comparing the "acid treatment" hand sanitizer gel to the "untreated control" hand sanitizer gel in this study. Procedure: A 4 oz sample of product was mixed then divided into equal parts (by wt.)--part 1 became the control, part 2 became the acid treatment. Entire experiment was conducted at room temp (22C; ethanol evaporates ~85C). We did add a equivalent volumes of internal standard (n-propanol) to each sample.
So, we are looking for comparative differences between the ethanol content in control vs acid treatments.
Observations: 1) a 20% decrease in the peak area ratio of ethanol occured (the ISTD was n-propanol and its' peak area was constant between control and acid treatment samples, so it was appropriate to compute peak area ratio as the quantifier); 2) the pH of the control was 6.97, the pH of the acid-trt sample after the 5 day period was 1.75.
The fact that we observed a difference in ethanol content was consistent with theory that conversion/degradation of ethanol did occur in the presence of acid. How did it occur? This is the thesis argument in-play here. We know that it was not due to evaporative loss because 1) the study was conducted at 22C and 2) previous stability data indicated that our product has 'no change' in ethanol content after sitting on shelf for 18 months at room temp.
In theory, this conversion was due to the oxidation of ethanol into ethanal and/or ethanoic acid; however, we have no chromatographic evidence of this speculation. What we do have is ''correlative data" that shows a strong pH change was correlated to a decrease in the ethanol content. Given these findings we can conclude that: So, either this wax column does not chromatograph aldehydes so the by-products are there but not detectible or the by products are not there. This is the current state.

Reply to forum writers:
1. Samples were not weighed before and after the acid treatment ended; this was not neccesary because any wt difference would be cancelled out by the comparing it to the controls. We did take before and after pH readings and the value of Control wa 6.97 vs Acid trt was 1.75 pH.
2. five injection replicates were made on each sample and %RSD was 0.95%, so instrument error was not a factor.
3. Regarding "per unit mass" of FID response for small molecules. Yes I will concede that this is a factor, yet it is countermanded by the fact that a 20% loss in ethanol signal would create an approximately 20% gain in the signal for the presumptive by-products (ethanal/ethanolic acid). To theorize: our Control had ethanol peak area of 500,000 and acid treatment had area count of 400,000. This 10,000 area count difference --if attributable to the by-products--would be easily detectible via FID. But we saw no such peaks other than the ethanol and ISTD peaks. So, it is still a mystery what is going on?
3. Concentration of ethanol in the product was 65%w/w which corresponds to ~ 32.5 mg/mL ethanol once worked up in the product testing assay. This is sufficient analyte to generate an easily discernable signal as peak areas are on order of 500,000.
4. The gelling agents are acrylates and they are irrelevant as a reactants in this paradigm. The central issue remains: were there any by -products of ethanol degradation created in the test sample, and are they detectible?
5. Why we do NOT see any peaks preceding the ethanol peak that would indicate the presence of ethanal or ethanoic acid is still a mystery.
6. Oh and yes Chrom#1, the weather in KC this wknd was great! Ate my first funnel cake of the summer at the town festival Saturday.

[HW Mueller]

What makes you think that EtOH doesn´t evaporate at 25°? Because you had a closed container stand for a longer time?
Why are acrylates irrelevant as possible reaction partners?
The oxidation of ethanol has been extensively studied probably 100 years ago, you can look it up.
You have never bothered to see what acetic acid does under your analytical conditions? Out of memory I would say that if you get acetal under your conditions you should be due for some sort of big award (someone correct me if I am confusing something).
Incidentally, no matter what I did I got 100 000 for the diff. between your peak areas.

[Peter Apps]

Hi Jumpshooter

There is another conundrum - any simple chemistry that made ethanol disappear would surely have a similar effect on the propanol that you used as internal standard. Ethanol will evaporate more quickly than propanol, unless the container is sealed, and in a sealed container there may not be enough oxygen to get the chemistry that you are predicting.

Due to the vagaries of FID response to small molecules with relatively large fractions of oxygen, a loss of X units in ethanol response does not imply that that response will be visible in other peaks - in the end it would all end up as CO2 and water, neither of which will give you any FID response at all.

I agree with Hans that you need to prove that you can detect the two expected breakdown products at the concentrations that you expect to be generated from the ethanol that has vanished - you should anticipate some problems; both of them are frustratingly intractable by GC.

Peter

[Jumpshooter]

What makes you think that EtOH doesn´t evaporate at 25°? Because you had a closed container stand for a longer time?

--> This is again a moot point because even if it does evaporate (miniscule as it might be) we must compare it to the CONTROL (which would've had equal and very low probability of evaporation) thus any presumed levels of evaporation would've cancelled out. The samples sat under a sealed container with parafilm over the opening and screw cap on top of that and were kept at 22C. LeChatlier would concur with me in saying that these conditions would not favor evaporation. So, this is a fringe explanation.

Why are acrylates irrelevant as possible reaction partners?
--> The thermodynamics hugely favor reaction with ethanol as it has a concentration that is 100x greater than the acrylates etc...

You have never bothered to see what acetic acid does under your analytical conditions? Out of memory I would say that if you get acetal under your conditions you should be due for some sort of big award (someone correct me if I am confusing something).
-->There would be nothing exceptionally remarkable about getting an "aldehyde" from the oxidation of its' companion "primary alcohol" in this paradigm. So awarding me a Prize is neither needed nor desired.

We are just doing a simple test of a hypothesis: If conditions that favor the chemical degradation of the active ingredient (i.e., ethanol oxidation) are imposed on the Product, then can a change in the ethanol concentration be detected and quantified, and, can the likely by-products of this chemical oxidation be detected and quantified? This was the thesis upon which our empiricism was founded. Nothing more elaborate. We created a paradigm wherein this hypothesis could be tested.

Thus far we have a "detected" a difference in ethanol concentration of acid treated samples relative to the control, however, we have not been able to assign this FACT as being caused by oxidation of ethanol.

Incidentally, no matter what I did I got 100 000 for the diff. between your peak areas.

---> Yes, that was correct, so then that makes the issue even more enigmatic. Since a 100,000 difference in peak area counts in the acid treatment sample should have yielded "something in terms of additional peaks" that accounted for the differential and apparant loss of ethanol. Furthermore, the magnitude of such peaks would render them easily discernable--yet we saw nothing.

To date: still a mystery. Maybe I will try a stronger degradation trt such as 5N acid for five days just to bump up the differential and consequently bump up the titer of supposed by-products to a higher/detectible concentration level.

Still a mystery though So, at this point we are concluding that:
1) a pH change in Product occured as a result of the acid insult (acid sample pH = 1.75 vs. Control sample pH = 6.97).

2) a 20% decrease in the ethanol signal (which computed to a 10% change in ethanol concentration) was correlated to the acid insult.

3) Whether this change in ethanol concentration was due to chemical conversion of ethanol to other non-ethanol by-products is inconclusive because no chromatographic peaks that would indicate the likely by-products (i.e., ethanal nor ethanolic acid ) were observed.

[Don_Hilton]

I am still wondering about trans esterification - which acrylate is used in the gel? I would look to see if that alcohol is in the GC trace. I would suggest injecting methanol, ethanal, acetic acid, and ethyl acetate to be sure that you could locate likely products. I am suggesting methanol as methyl esters are comon comercially.

From way back in the days I took organic chemistry - typically aldehydes oxidize to acids faster than the alcohol oxidizes to form the aldehyde. While I have observed ethanal in a bottle of absolute alcohol, my recollection is that I have found much more acetic acid and ethyl acetate present. (This has been in the examination of a blank to identify background contaminants.)

[bhuvfe]

[quote=

GC parameters: 100% wax column, 40C isothermal for 6 minutes; ramped to 100C hold for 1 min; inlet psig= 10; inj = 200C; Det = 300C


:[/quote]


Have you tried to increase the final T to the maximum allowed for this column (and the hold time)?

[bhuvfe]

Do you have any way to verify ethyl chloride formation? Possibly before your sample preparation (ethyl chloride is a gas at room T)?

[HW Mueller]

I was talking about EtOH reaction with acrylate.
The use of parafilm as you describe is highly suspect.
You would not be the first one to inadvertendly have large handling differences between ref. and unknown.
I still don´t have any clue as to why you conclude that you have oxidation.

[Jumpshooter]

After conducting a few additional tests I have determined that our problem (i.e., failure to detect ethanol oxidation products) is not instrument related. The problem seems like it inheres in the actual chemical rxns. For one, I will treat the "acid" sample with acidified potass.dichromate to see if the ethanol can truly be oxidized into ethanal and ethanoic acid. However, this is a bit of a 'deviation' from the standard product stability testing protocol; yet, it seems necessary to try? Inject the sample and inspect that baseline chrom for peaks.