Enigma: Failure to detect products of Ethanol oxidation

[Don_Hilton]

Ethanol can be oxidixed with oxidizing reagents. This is well known. And you probably do not want traces of chromium sallts in your GC inlet - so you would have to handle your sample differentlly than you have before. The question is where did your ethanol go.

I would suggest you repeat the experiment, with something other than parafilm as a seal. If you want to propose generation of ethyl chloride, use a headspace vial and sample the headspace for volatile components as well.
As a control place some ethanol at the same dilution in water and add acid. This should help to rule out other comounds in the hand sanitizer that would give rise to alternate explanations.

If the alcohol is denatured, take the denaturing agent into account.

[Ron]

There are certainly some interesting assumptions being made in this experiment, I would suggest rethinking some of the basic premises. First of all, the assumption is made that if there is loss of area in the treated sample compared to the untreated sample compounds other than ethanol should be detected in the GC run. The possibility of evaporative losses is dismissed since ethanol volatiiles at ~85 C, which is in conflict with the physical properties of ethanol. The boiling point of ethanol is 78C, but the vapor pressure at 20C is 9 kPa, so ethanol will evaporate rapidly at room temperature unless the container is well sealed. I share the sentiments expressed above about the sealing efficiency of Parafilm.

The statement is made that the evaporation of propanol is as likely as ethanol, in fact the vapor pressure of propanol is significantly lower than that of ethanol, making equal evaporative losses very unlikely.

I am not an expert on alcohol chemistry, and it has been many years since I took organic chemistry, but I don't remember primary alcohols being oxidized by dilute hydrochloric acid, especially at room temperature. Do you have any literature references that lead you to believe that this is a primary reaction pathway under your experimental conditions? My recollection is that the oxidation of ethanol generally is a catalytic process.

After a certain number of attempts to make the experimental data fit the model and not succeeding, it's time to carefully evaluate the experiment to see if the experiment is well designed or if changes in the experimental design need to be made. After repeated trials and modifications to the experiment to account for potential experimental issues, the possibility must be considered that the data don't fit the model because the model is incorrect.

[AICMM]

Jumpshooter,

Have you dissolved the two gels in a volume of water and pH'd them? Don is correct in that many of the hand gel formulas that I have looked at rely on acidification to get them from mix-able liquids to gel form. Thus, my first assumption, is that your acidified sample may have been quickly neutralized.

For what it is worth....

Best regards.

[krickos]

The possibility of evaporative losses is dismissed since ethanol volatiiles at ~85 C, which is in conflict with the physical properties of ethanol. The boiling point of ethanol is 78C, but the vapor pressure at 20C is 9 kPa, so ethanol will evaporate rapidly at room temperature unless the container is well sealed. I share the sentiments expressed above about the sealing efficiency of Parafilm.

The statement is made that the evaporation of propanol is as likely as ethanol, in fact the vapor pressure of propanol is significantly lower than that of ethanol, making equal evaporative losses very unlikely.

I am not an expert on alcohol chemistry, and it has been many years since I took organic chemistry, but I don't remember primary alcohols being oxidized by dilute hydrochloric acid, especially at room temperature. Do you have any literature references that lead you to believe that this is a primary reaction pathway under your experimental conditions? My recollection is that the oxidation of ethanol generally is a catalytic process.

Agree with Ron here. However if for instance there is ZnCl present it can serve as a catalyst and oxidation might occour. Also if there is an ingrediant that can pick up a chlorine atom temporaly you might get an substitution reaction (lower temp ie roomtemp generally favour substitution rather than elimination).

As stated previously, looking for stuff like methyl-ethylchloride is not easy as they have very lowboiling points, reminds me of an incident where I tried to explain the lost cause for a production engineer in looking for those ones when she accidendently got some methanol/ethanol in her touluene/tionylchloride mixture and raised temperature to 50°C before adding the primary alcohol that was supposed to react with the tionylchloride.

[Jumpshooter]

OK, given the critique of Chrom Forum writers, I thought it wise to re-think our approach to this "Stability Indicating" study. Pls identify any shortcomings or confounders in my experimental design. Here are the basics:

1. Purpose: To investigate the stability of the product and/or the product's active ingredient by exposing it to treatments according to the company SOP. The SOP verbatim said: 1) "expose Product to oxidizing conditions by adding an amount of Hydrogen Peroxide not in excess of 10% of the sample weight for 3 days"; 2) "expose Product to acid degradation by adding an amount of Hydrochloric acid not in excess of 10% of sample wt. for 3 days"; 3) "expose Prod. to alkaline degradation by adding an amount of sod.hydrox. not in excess of 10% sample wt for 3 days".

2. Procedure: Set up 5 clean erlenmeyr flasks, then weigh in 9.0+/-0.05 g of Product (gel hand sanitizer) and weigh in 1.0+/-0.05 g of either acid, base, or peroxide into each dedicated flask. For the Control, weigh in 9.0 g of gel plus 1.0 g of distilled water. For the reference STD weigh in 9 g of USP ethanol plus 1 g of water. Record all wts on the four place balance. Attach the ground glass stopper and let all flasks sit on lab bench for 3 days. After three days, add 1.0 g of ISTD to each flask, then assay the samples by diluting the volume to 100 mL mark, mixing well, then allocate to als vials.

3. Data analysis:
Get the peak area counts for Ethanol and ISTD for all samples, control, and STD. Compute the peak area ratio for ethanol (etoh area / istd area) for all samples.
A comparison of the STD (100% ETOH) to the Control (~65% ETOH, product label claim) and to the treatment Sx (~?? % ETOH) should yield an estimation of the effect of each treatment on the ETOH content in the test samples.
%ETOH in sample = 100% x (p.a.r Sx / p.a.r STD)x(STD wt./Sx wt.)
sx = sample std = reference std.


4. Results: In theory, the treatment of acid to a primary alcohol ought to yeild a halogenated alkane, so HCl + ethanol ---> chloro-ethane + water.
NaOH + ethanol ----> ethene + water. Peroxide + ethanol ---> ??

Thus, there should be a net decline in the ethanol concentration for treatment groups COMPARED to the Control. This decrease is not due to evaporation (why? b/c it is glass stopper on top and held at 22C) but is concievably due to chemical conversion of ethanol to other non-ethanol by products.

Chromatography: Because neither chloro-ethane or ethene will be retained on the 100% wax column, then peaks for them will not show up. But I should see a discernable decline in the ethanol peak area for the treatment group samples.

Delimitations: 1. We will concede that some (small amt) ethanol could be diffused into the headspace of the erlenmyr flasks---but even if so, then this ought to be cancelled out by the same process that occured in the Control. I am NOT doing a headspace analysis; only pulling samples from the solution. 2. The ISTD will be added after the treatments so that the presumed effects on ethanol degradation/interaction won't be confounded by the acid, base, or peroxide interacting with another alcohol species (n-propanol, the ISTD). This is a logistical departure from tradition wherein the ISTD is added prior to extraction/chemical workup, but necessary in this paradigm. 3. It is conceded that the acid, base, and peroxide will interact with other moieties in the Product (e.g., acrylates, etc..) but this is not the focus of the present study.

Does this sound kosher? Pls critique! Thanky for your Time

[Don_Hilton]

The problem with what I see is not necesarily the design but rather that you guess that ethanol has been converted into a volatile compund.

You weighed in all of the components in the mixture, so it should not be hard to detct the loss of about 0.1 to 0.2 g from the total added. If you get a loss in mass, then you know the material is gone. The way to prove or disprove a question of mass balance is to measure the mass.

If you believe the vapors to be trapped in the fllask, then it would be a good idea to sample the headspace and inject that in to the GC. 5 uL or so should be more than plenty. If you have ethyl chloride or ethene present and they are unretained, they will elute at the dead volume of the column. (My guess is that you are not likey to see much, as your conditions are rather mild for displacing the hydroxyl group on the ethanol unless there is something else in the mix that would help out. Alternatively you could see some methanol from transesterification.)

If your qestion was is the mixture stable with the addition of acid, you have an answer: no.

To conclude where the ethanol went, you need to make some measurements to see where the remaining mass is or is not. The fact that you did not see a volatile product of the ethanol does not mean that there was no nonvolatile product formed.

To try to figure out what happened, the next step is to line up all the possible suspects. Ethanol is only one. The gelling agent is important and needs a name and percentage in the composition. And so on.

Going back to the mass balance issue and the idea of nonvolatile products, repeat the experiment and weigh the flask at the end. Draw out a small portion of the solution to dilute with water and analyze by GC. (0.1 mL at most weigh it to determine the correct proprtions of water and isopropanol to add) and evaporate the rest of the sample to dryness. See if it gives more or less mass than a sample handled the same way, but without acid. Drying will be a bit tricky because you will have residues of HCL, which may not leave the sample easily. If there is a way to quantitatively precipitate and rinse the gelling agent, all the better.

Find the increased mass to determine where the ethanol went.

[Peter Apps]

Forgive me if I do not actually do the calculations, but your reaction schemes produce gaseous (or at least very low boiling) products, and enough ethanol is vanishing for the pressure or volume increase of the headspace to be easily measurable. How about a simple S-tube manometer through a gas-tight bung in the top of each bottle ?

Peter

[HW Mueller]

Don, how can one get methanol here via transesterification?

General question: Is there anybody out there that has considerable experience with EtOH handling, especially preventing evaporation? I am a bit skeptical that this can be ruled out here, even if the ref. is seemingly handled the same as unknowns.

The three required reactions should all be well known, one would think, so that the literature would yield info on the possible products.
Also, I would play with the reactions themselves to get a "feeling" of what gives: Extent the reaction time, increase temp. . . .

[DanCooper]

Hi,

Intersting thread...

Your reaction schemes are wrong;

EtOH + HCl = no reaction unless there's a catalyst present (and if you've ever tried it, it's actually a fairly difficult reaction to get to work).

EtOH + NaOH = EtONa, this will be in equilibrium, the position of which will depend on the pH.

Also, in my experience, parafilm and nescofilm are more hassle than they're worth in many situations. A lot of solvents permeate through them and actually react/dissolve them.

Hope this helps!

[Don_Hilton]

Trans esterification would be methylacryate + ethanol -> ethylacrylate + methanol. When you make pesticide standards you do not want to disolve methyl esters in ethanol for this reason. As the standards stand, they change on you.

See http://en.wikipedia.org/wiki/Transesterification

[Ron]

If you are producing a volatile reaction product a simple manual headspace experiment should show this. Even though the column you are using is not ideal for ethyl chloride or ethene and they will elute very close to the dead time you should still be able to see a peak or peaks before the ethanol peak if you do a headspace analysis.

I am in agreement with Don, just because there is a loss of response does not mean a volatile product was formed, the mass balance would provide a lot of information as to what is going on.

HW, I don't have experience with a lot of experience with ethanol, but many years ago I was working with monoglyme/water and diglyme/water mixed solvent systems, and we made up the mixed solvent systems by weight, recorded the weight of the solvent and container, and when it was time to use the solvent we re-weighed to verify that we had not had evaporation of organic from the solution. Ah, the joys of working with volatile solvents in the summertime in a building without air conditioning in the summer.

[HW Mueller]

Don, ok, didn´t see the METHYLacrylate.

[Jumpshooter]

Reply to Chrom Writers:


EtOH + HCl = no reaction unless there's a catalyst present (and if you've ever tried it, it's actually a fairly difficult reaction to get to work).

EtOH + NaOH = EtONa, this will be in equilibrium, the position of which will depend on the pH.

Also, in my experience, parafilm and nescofilm are more hassle than they're worth in many situations. A lot of solvents permeate through them and actually react/dissolve them.

*OK, so we ought not expect any by-products from the Acid treatment b/c it is NOT spontaneous at room temp. But, isn't chloro-ethane also a by product? OK, so what do you suggest in lieu of parafilm to suitably prevent evaporation?

How about a simple S-tube manometer through a gas-tight bung in the top of each bottle ?
*This apparatus we do not have; so we are beset with using the GC.

You weighed in all of the components in the mixture, so it should not be hard to detct the loss of about 0.1 to 0.2 g from the total added. If you get a loss in mass, then you know the material is gone. The way to prove or disprove a question of mass balance is to measure the mass.
*This will undoubtedbly be very tedious and wrought with some uncertainty of measurement, but do-able.
We do not have a headspace [HS] setup in this lab; only an autosampler with tray. Thus, although it would be ideal to use HS procedure we are limited to what we have. In the end this may prove to be our undoing no?

If you are producing a volatile reaction product a simple manual headspace experiment should show this. Even though the column you are using is not ideal for ethyl chloride or ethene and they will elute very close to the dead time you should still be able to see a peak or peaks before the ethanol peak if you do a headspace analysis.
*We don't have a HS sampler, we are beset with using als vials with syringe drawing up 1 uL of test solution.

*Given the delimitations and limitations of our stability study, then it would seem that an appropriate hypothesis is that: "there will be no difference between the Control and Treatment samples". If this is the case, then I wonder if our method is even capable of indicating product stability?




2. we want to investigate some type of effect due to our treatments. From a perusal of the chrom writers, it appears that in order to "make" an effect happen we need to create conditions that would make the reaction proceed from ethanol to ethanol degradation products. The reaction as it stands now is not spontaneous at room temp and would need a "bump".

3. I acknowledge that I must consider gathering pre and post weight viz. mass measurements of the samples in the erlenmeyr flasks. But the weight differences are expected to be minute and wrought with error--sort of like this: What is the weight of one box of oranges on a semi-truck trailer filled with 100 boxes of oranges? By taking the weight of the trailer and subtracting out the weight of one box of oranges. The result will be still be the weight of the tractor trailer filled as if it were still loaded with 100 boxes of oranges. This is why we are proposing the use of peak area differences to account for "loss of ethanol".

4. Question: once I remove the stopper, any ethanol that may have volatized will be lost to the atmosphere[/quote]

[bhuvfe]

Reply to Chrom Writers:


If you are producing a volatile reaction product a simple manual headspace experiment should show this. Even though the column you are using is not ideal for ethyl chloride or ethene and they will elute very close to the dead time you should still be able to see a peak or peaks before the ethanol peak if you do a headspace analysis.
*We don't have a HS sampler, we are beset with using als vials with syringe drawing up 1 uL of test solution.

[/quote]

As quoted above you don't need an autosampler to perform a manual injection. Just scale down the size of your experiment to few grams in a 10 ml HS vial and wait your 5 days and inject 250 ul of the gas phase. Being a manual injection means that it will be qualitative but it might shed some light on this (or at least exclude something). Obviously a GC-MS would be preferred.

[Don_Hilton]

What to use for a stopper? I would suggest a rubber stopper. If you need to purchase that, I would suggest getting a one-hole stopper as well. put a glass tube in the hole and you can even use a length of clear tubing, like Tygon to make the s-shaped trap to look for gas evolution.

Ethanolic HCL can be purchased as a laboratory chemical. This is why I strongly suggest you look for a different suspect. If you are losing ethanol from the mixture, there is something else involved. The key to the mystery in in the list of other ingreedients.

As far as measureing differnces on the order of 0.1 g on 10 g material in an erlenmyer flask - this should be easily done. If you are using an electronic balance, zero it and have it perform the autocalibration. Now take a 250 g weight from your standard weight set and weigh it. Your weight should read correctly to the number of places specified for the class of weights you own and by the specification for the balance. The analytical balance is able to give the most accurate and precise measurement in the laboratory.