musty odor in drinking water

[BMU_VMW]

I'm looking for some advise in earthy - musty odors in drinking water.
We are experiencing a odor problem at one of our production sites.
The water clearly has a musty smell but analysis with SPME-GC-MSMS on our target analytes does not yield any result. We analyse: IBMP, IPMP, MIB, Geosmin and the TCA's.
Full scan analysis after SPME and full scan of a CH2Cl2-extract did not come up with any relevant peaks.

Has any of you experience with other musty odors? It would be very helpfull to at least know what compounds to look for.

thanks
BMU

[rb6banjo]

I am a little surprised that the root cause is not one of those. What are your spme conditions? Most of those have an odor threshold that is lower than what a scanning method will allow you to see. I find TCA as the root of nearly all of my musty problems. I can see 5 parts-per-trillion with the CX/PDMS/DVB-2cm phase IF i only monitor the 2 largest masses in the mass spectrum.

[BMU_VMW]

I'm sure that the odor is not caused by one of the 'standard' compounds since I've analysed the sample and the same sample added with 2.5ng/l of standard.

For each sample I can see all compounds at this 2.5ng/l level, some of them have a recovery that is a bit high (150%), but that does not concern me to much since the pure sample shows a more or less flat baseline.

Taken in to account that the odor is pretty strong I don't think it will be caused by a compound at a concentration lower than 2.5ng/l.

We never find TCA's; if there is a compound that pops up its geosmin or MIB.

conditions are:
10ml sample with 1.5g Na2SO4 and 25µl 1M NaOH in 20ml HS-vial
SPME: 50/30µm DVB/CAR/PDMS 1cm
incubation: 70°C 2 min
extraction: 70°C 30 min
desorption: 250°C 2 min split 5:1

analysis GC-MSMS

[rb6banjo]

Very interesting. Looks like you have very good detection limits. That's a puzzle. Every musty problem I've ever come across could be attributed to one of those you listed in your first post. Mostly, TCA is the source of my problems. I have seen tribromoanisole in water samples. Did you check for those isomers? The most common one is the 2,4,6 isomer.

[BMU_VMW]

No I did not check for any others than the one listed and borneol.
I'll check the bromoanisoles.

thanks for your help. It anythings comes to mind, let me know

[GOM]

Hi

That is indeed interesting and surprising that you haven't detected your target analytes.

The human nose is very sensitive to micro-organism metabolites for good survival/evolutionary reasons.

Even so, I have analysed for musty odours by SPME/GCMS -SIM before with a lot of success in various samples.

Most are micro-organism metabolites. e.g. It is that typical smell released from the ground after summer rain =fungal metabolites

Geosmin, and methyl iso borneol are favourites,but you have checked for that

If it is drinking water I would have suspected a halogenated (from chlorinated water) species. e.g. 2,4,6 trichloro anisole. but you have checked for that

Octen-3-ol is another candidate

I have a summary document of musty taints that I wrote.

If you are interested please contact me on ralph.calvert@gmail.com and I can make a pdf file of it for you. i would be happy to share it with you or anyone else.

Kind regards

Ralph

[Peter Apps]

If you can do GC-sniffing it would at least show you where the peak is eluting (or that is is not making it through the column).

Peter

[BMU_VMW]

Ralph,

We are dealing with drinking water that has been made from surface water.
The Octen-3-ol is a new one for me. I'll check if I can find it in the Full-scans I have.
I'll send you an email regarding the paper.

Peter:
We don't have a sniffing port. But if we can't find the component causing the smell we might end up with a little DIY. using a PTV to inject and a SSL equiped with a glass funnel to improvise a sniffing-port.

Last thing I've done so far is to have approx 1L of water run over a RF-C18 cartridge (we used to use these for our odor analysis befor we turned to SPME). If I can elute this with 0.5-1ml of MeOH I can spike it in an empty vial and use the SPME to analyse. Other possibility is to elute it with CH2Cl2 and inject 50µl on an other GC in FS. 'Problem' is that both of them have simmilar columns (Rxi5_sil MS 30m 0.25 0.25)

BMU

[GOM]

Hi Bart

I replied to your email.

Your liquid injection approach sounds good.

If you don't have a dedicated sniff port but do have a GC with FID then, providing the FID is in the right position, use that as your sniff port - set at 130-150°C with the fuel and make up gasses off.

The narrow bore of your column will mean that you have to be quick to catch the narrow taint peak(s).

I had one case of a musty odour taint complaint in a dried powder food product relating to stagnant water

SPME GCMS SIM identified the taint (usual suspects) as being in the plastic lid, and not the container tub or product. At the time I could only say that it was in the lid and was consistent with stagnant water.

It was only months later that the supplier told me that the source was tracked down to the metal die used in stamping out the lid. It had cracks in it and, after washing, water was left in the cracks. The musty odour was then transferred to the plastic during stamping.

Regards

Ralph

[BMU_VMW]

Hi Ralph,

I've seen your paper and there are some interesting compounds in it that we will defenitly look for.
We only have GC-MS (1 headspace) and GC-MSMS (1 SPME setup; 1 liquid) systems so using an FID as sniffing-port is not an option.
As always when looking for difficult things Murphy paid us a visit and the liq GCMSMS broke down today.

In your experience will (most) odor compound elute from a standard 30m Rxi5 sil ms column? Not seeing the peak might be caused by the column or even by breakdown at injection???????

Our main goal now is to find what is causing the smell. Were the smell is comming from is a secundary problem

[GOM]

Hi Bart

I would prefer to use a polar column but your phase should work. You have already implied that "the usual suspects" elute well on your column.

In my experience injection port breakdown has never been an issue.

Looking back through my notes - 3-octanone is another contender as well as octen-3-ol for a mushroom odour (sometimes associated with a musty odour)

Regards

Ralph

[Peter Apps]

Probably stating the obvious, but are your checking your SPME fibers and extracts to check that they have the musty odour ? - it might be something that is not even extracting properly.

Peter

[GOM]

Peter's suggestion is a good one

My instinct says stick with the obvious musty taints associated with water.

The musty descriptor only applies to a few molecules.
There are few others that I have come across that meet that descrptor.

in my experience these are the ones most commonly associated with "musty".

Geosmin, 2-methyl isoborneol, 2-isopropyl-3-methoxypridine, 2-isobutyl-3-methoxypridine 3-octanone , octan-3-ol, 2,4,6 TCA

Regards

Ralph

[BMU_VMW]

Let me mix two of your replys in one answer:

Ralph:
The musty descriptor only applies to a few molecules.
There are few others that I have come across that meet that descrptor.

in my experience these are the ones most commonly associated with "musty".

Geosmin, 2-methyl isoborneol, 2-isopropyl-3-methoxypridine, 2-isobutyl-3-methoxypridine 3-octanone , octan-3-ol, 2,4,6 TCA

Peter:
Probably stating the obvious, but are your checking your SPME fibers and extracts to check that they have the musty odour ? - it might be something that is not even extracting properly.

The usual suspect are extracting fine since I'm able to see a 2.5ng/l spike in all the samples.

I can't conclude on 3-octanone and octan-3-ol since I don't have a standard of these compounds. But since these are chemicly not so different than the others I guess it should work.

I have not checked the SPME-fiber nor the extract. In fact so far I have never tried that. Do the SPME-fibers actualy smell?
The extract is a bit more difficult when eluting the SPE with CH2Cl2 there was no smell (other than DCM). I have yet to try an LLE with CH2Cl2


Once my routine analysis are done for today and the GC-MSMS is up and running again I'll do some more tests with LLE, SPME (e.g. different fiber types), and if I find a lot of time I might even DIY a sniffing port.

Thanks a lot for the advise.
It is greatly appreciated.

[Peter Apps]

When checking extracts for odour you need to put a couple of drops onto filter paper, or a perfumers sniffing strip, let the solvent evaporate, and then sniff immediately.

I have done very little SPME so I am not sure whether they smell or not. With serious taint compounds in analytically detectable quantities I would be surprised if they did not. If they don't smell at room temperature you could try warming one up with a gentle stream of hot air or radiant heat from a hot surface.

Peter