1
$\begingroup$

I am a first year PhD student in the UK, and without giving too many details of my work as it is industrially funded and my sponsors are a bit touchy over IP, I am struggling to isolate the products of a series of reactions I have done. The general reaction is as follows:

Reaction scheme. Alcohols are treated with a source of a methylene (-CH2- unit) in 1,4-dioxane to produce dialkoxymethanes (R-O-CH2-O-R).

I have found that yields (by NMR) are best with a significant excess of alcohol compared to the methylene source, so typical conditions are 8.3 mmol of methylene source, 83 mmol of alcohol in 5 mL of 1,4-dioxane. The rub is that my supervisors want me to get isolated yields, which I am struggling to do. Here is a few examples of the kind of products I am talking about.

Series of products as examples. Includes: dimethoxymethane, diethoxymethane, diisopropoxymethane, bis(cyclohexanemethoxy)methane, bis(2-methoxyethoxy)methane, dibutoxymethane.

Here is what I have tried so far:

  • Simple aqueous workups. This generally gives me a pure fraction of product at the end to get a nice NMR spectrum of, but an awful lot of product is lost in the process. I think the -O-CH2-O- bridge gives the products a certain level of solubility in water, so while I am washing out the alcohol, I am taking a lot of product with it.
  • Distillations. I've worked out the boiling point gaps between the alcohols and those of the products that are well characterised. Methanol and dimethoxymethane have an azeotrope so can't get it pure, ethanol and diethoxymethane have only a 10 C difference in BP and I can't get pure product with my equipment (have tried a few times, best was like 80% product). IPA and butanol should work, just haven't done it yet. For the larger alcohols, such as the cyclohexanemethanol and some of the ones I haven't included, this probably isn't feasible, however. The presence of 1,4-dioxane (BP 101 C) complicates this, but a lot of these reactions will proceed (if significantly more slowly) neat.
  • Column chromatography. This is the obvious answer, but the problem I am having is that the products don't seem to be very visible, or at all visible, by TLC. There is just no functionality in a lot of cases for a stain to react with and visualise. Dimethoxymethane (the methanol product) boils at 42 C, so I don't know how I would concentrate the fractions without losing product, assuming the column worked in the first place.

These reactions have been shown in literature. I have found dozens of references for the synthesis of dialkoxymethanes, but the vast majority of examples are of various substituted benzyl alcohols which obviously have a UV chromophore. The few papers that use small alcohols generally only report NMR yield, and those that do column simple alcohols (never methanol, ethanol or IPA, generally just butanol or longer) do not report the visualisation method used aside from 'visualised by UV' which can't be possible as there is no UV chromophore. No clue how they did the column for these, its infuriating.

Paper that shows large scope but gives no visualisation method aside from 'UV'- https://pubs.acs.org/doi/full/10.1021/acs.joc.2c02637

Paper that shows simpler alcohols but no isolated yield - https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cctc.201902332

Can give more examples if you like.

Does anyone have any advice? I do not have a wealth of experience with columns or TLC so please let me know if I am missing something obvious. My supervisors don't have any immediate suggestions and no one in my group can think of anything either.

Improve this question
$\endgroup$
10
  • $\begingroup$ What TLC stains have you tried? $\endgroup$
    – Waylander
    Commented Dec 7, 2023 at 12:46
  • $\begingroup$ @Waylander Have only tried KMnO4, although I have looked through a few different sites that describe the best stains to use for specific functional groups (reachdevices.com/TLC_stains.html for example). It was my understanding that KMnO4 was the best choice, and I can get a faint spot from my crude product mixture that is more than likely product. When it comes to TLCing the fractions from the column, the product is now so dilute that I can't see it in any of the fractions. $\endgroup$
    – Matthew
    Commented Dec 7, 2023 at 13:17
  • $\begingroup$ 2,4-DNP may show it as the dioxymethylene products are masked aldehydes. Also CAM might be worth trying see pcwww.liv.ac.uk/~jxiao/links/TLC%20Stains.pdf $\endgroup$
    – Waylander
    Commented Dec 7, 2023 at 13:44
  • $\begingroup$ I think it is just too difficult to get the dimethoxy methylene product clean. Quote the HMR yield and move onto to reacting other alcohols whose products will be easier to purify. If your industry supervisors want that result they can find a purification method. And yes, I have been an industrial PhD supervisor in the UK $\endgroup$
    – Waylander
    Commented Dec 7, 2023 at 13:47
  • $\begingroup$ If it's not necessary to isolate a product to go on with successive steps of a synthesis route, then you don't do it. Do you need the isolate? That being said: have you tried to crystallise the stuff in a -20 or -40 °C fridge? That doesn't give you a very high yield (and you need a few tries to work out the best solvent/temp. combination), but likely a good amount of pure substance, reducing the amount that needs workup by destillation/chromatography. Btw. while you're standing close to the fridge: have you tried the aqueous workup with ice cold water? $\endgroup$
    – Karl
    Commented Dec 7, 2023 at 20:53

0

Reset to default