GCMS Services - Comparisons

[jfrater]

I was going to message you, but figured it best to ask in the open where it might also help someone else with the same question, but you didn't mention the difference in SPEED. I also remember you're not based in the same country as me, so wondering if that will play into it as well.

How fast were the samples processed?

Wb2Lab is extremely fast- within a week from receipt of samples. Phytochemia is currently just under a month.

 

[ScentAle]

Would be interesting to have directly this gcms machine eheh

 

[Lucofborg]

Just a quick update. I sent 5 samples to Wb2Lab last week. They received it last Friday, and today (Tuesday) they sent me the report for the 5 samples. They were extremely fast. I decided to send some samples with lots of naturals like Bvlgari's Man in Black and Guerlain Santal Royal and the reports were very very well done. I had Man in Black done by Phytochemia in the past but was never really happy and with this new report I finally understood why.
1-Lots of missed molecules by phytochemia
2-Different percentages of some molecules

 

[jfrater]

Would be interesting to have directly this gcms machine eheh

It's on the shopping list

 

[jfrater]

Just a quick update. I sent 5 samples to Wb2Lab last week. They received it last Friday, and today (Tuesday) they sent me the report for the 5 samples. They were extremely fast. I decided to send some samples with lots of naturals like Bvlgari's Man in Black and Guerlain Santal Royal and the reports were very very well done. I had Man in Black done by Phytochemia in the past but was never really happy and with this new report I finally understood why.
1-Lots of missed molecules by phytochemia
2-Different percentages of some molecules

Having spent time analysing some of the other samples I tested with both companies I think that Phytochemia can be good with certain naturals but Wb2Lab wins in the overall quality stakes.

 

[pkiler]

It is very odd and a bit disconcerting to see my own fragrance GCMS for sale with WB2 Lab.
At least SOMEBODY wants to know about it... LOL!

 

[downmemorylane]

It is very odd and a bit disconcerting to see my own fragrance GCMS for sale with WB2 Lab.
At least SOMEBODY wants to know about it... LOL!

That's a quality stamp Paul ✨

 

[ScentAle]

It is very odd and a bit disconcerting to see my own fragrance GCMS for sale with WB2 Lab.
At least SOMEBODY wants to know about it... LOL!

I imagine that it is also a good sensation that someone pay for analyze your art doing after hard months eheh
Ps: just yesterday i readed an Italian long good review about one your perfume!

It's on the shopping list

Wow and big wow! 2000 gcms everyday ahah.
It need many times for complete an analysis? Is it hard to learn?

 

[Phytochemia]

I stumbled upon this post at random this morning. I am very please by what I am reading and thank you. We know that we still have work to do, and the comments that I read are very encouraging.

Regarding the CAS.
I will be honest that I have a very personnal negative opinion of CAS. They are, in my opinion, a good idea that turned bad. I admit that I am countercurrent with this. but the way that I see them is that they are more confusing than helping, specially with natural product. For example, Limonene can have many CAS:

• From lemon oil, it can has the lemon oil CAS - 8008-56-8
• From orange oil, the orange oil CAS - 8028-48-5
• Racemic limonene is 138-86-3
• d-Limonene is 5989-27-5
• l-Limonene is 5989-54-8

As you go down in precision, the higher number of CAS there can be. Alternatively, a compound which has multiple derivative can have only one CAS. Iso E Super is a mixture of more than 10 compounds and is all under the same CAS 54464-57-2.

When I started the development of our database I needed a unique identifier for all compounds, and CAS left a bad taste due to their ambiguity even though they are "marketed" as unique identifier. However, we did listen and the next version of our database will have the CAS (or at least some CAS), even if there are multiple options available. And we currently are able to give them on request, we send an additional JSON file that contains the CAS and InchiKey of the compounds.

Regarding the different percentage.
This can be explained quite easily that we do not use the GC-MS trace for quantification, but the GC-FID trace. This is following recomendation of the ISO TC 54 (for essential oil) and IOFI recommendation (for flavor). MS linearity and response factor are more variable than FID so they tend to give a larger error in the end. With the IOFI recommendation we can also go from FID trace to mg/ml using their algorithm. It seems that the resulting number are not closer to the blend from the comments here. We will try to investigate eventually.

On this note, I would like to point out that we sadly do not have anyone trained in perfumery on staff. This is not due to the lack of will, more of time and local ressource. We would love to eventually have someones (or two) trained so we could do this kind of check in house and be able to increase of accuracy based on organoleptic properties.

Regarding the decimals
This is the chemist in me commenting, but I would argue that more than 3 decimals start to loose sense since the error on the significant number at this level are very hight due to instrumental variation.

---

You are probably right that we are better with natural compounds as this is our primary specialisation, we have over 25 thousands essential oils analysis in our database so we were able to do a lot of correlation regarding essential oil composition and biochemistry which is harder to do with fragrance since they rarely follow a set of rules (or do they.....). But we have been stepping up our game lately and are planing on continuing to improve as we are aware that synthetic materials are our weak spot and want to solve that.

Thank you Jamie for sharing your experiences with those three labs. I'm mostly used to Phytochemia. I think I've sent 3 samples there. I was mostly satisfied but I found that they missed too many molecules. For example I've seen two analysis from 2 fragrances that I know contains veramoss/evernyl, and not in traces, but never appears in there reports.

I've noticed that the Creative Accord raw data doesn't appear to go as deep as I thought. Only 2 decimals, compared to WB2LAB, which goes to 4. So I'm wondering, is the $1000 because Creative Accord also makes a final formula for us?

We report Evernyl/veramoss as Atraric acid, its common name. Maybe it was listed under this name? I will admit that this is probably one of the best example of the need for CAS (CASes?)...

In any case, I really enjoyed reading this thread and I learned a lot. If you have any request, questions or suggestions, I would love to help!

 

[David Ruskin]

This can be explained quite easily that we do not use the GC-MS trace for quantification, but the GC-FID trace.

The GC part of GC/MS is used to quantify the mixture being analysed. The MS part is used to give a qualitative analysis. FID (Flame ionization detector) is the most common system used in gas chromatography to quantify the components of the mixture being analysed.

When I was working the GC was linked to the MS. GC measured amount and separated individual chemicals which were then passed into the MS, for identification.

 

[jfrater]

I stumbled upon this post at random this morning. I am very please by what I am reading and thank you. We know that we still have work to do, and the comments that I read are very encouraging.

Regarding the CAS.
I will be honest that I have a very personnal negative opinion of CAS. They are, in my opinion, a good idea that turned bad. I admit that I am countercurrent with this. but the way that I see them is that they are more confusing than helping, specially with natural product. For example, Limonene can have many CAS:

• From lemon oil, it can has the lemon oil CAS - 8008-56-8
• From orange oil, the orange oil CAS - 8028-48-5
• Racemic limonene is 138-86-3
• d-Limonene is 5989-27-5
• l-Limonene is 5989-54-8

As you go down in precision, the higher number of CAS there can be. Alternatively, a compound which has multiple derivative can have only one CAS. Iso E Super is a mixture of more than 10 compounds and is all under the same CAS 54464-57-2.

When I started the development of our database I needed a unique identifier for all compounds, and CAS left a bad taste due to their ambiguity even though they are "marketed" as unique identifier. However, we did listen and the next version of our database will have the CAS (or at least some CAS), even if there are multiple options available. And we currently are able to give them on request, we send an additional JSON file that contains the CAS and InchiKey of the compounds.

Regarding the different percentage.
This can be explained quite easily that we do not use the GC-MS trace for quantification, but the GC-FID trace. This is following recomendation of the ISO TC 54 (for essential oil) and IOFI recommendation (for flavor). MS linearity and response factor are more variable than FID so they tend to give a larger error in the end. With the IOFI recommendation we can also go from FID trace to mg/ml using their algorithm. It seems that the resulting number are not closer to the blend from the comments here. We will try to investigate eventually.

On this note, I would like to point out that we sadly do not have anyone trained in perfumery on staff. This is not due to the lack of will, more of time and local ressource. We would love to eventually have someones (or two) trained so we could do this kind of check in house and be able to increase of accuracy based on organoleptic properties.

Regarding the decimals
This is the chemist in me commenting, but I would argue that more than 3 decimals start to loose sense since the error on the significant number at this level are very hight due to instrumental variation.

---

You are probably right that we are better with natural compounds as this is our primary specialisation, we have over 25 thousands essential oils analysis in our database so we were able to do a lot of correlation regarding essential oil composition and biochemistry which is harder to do with fragrance since they rarely follow a set of rules (or do they.....). But we have been stepping up our game lately and are planing on continuing to improve as we are aware that synthetic materials are our weak spot and want to solve that.



We report Evernyl/veramoss as Atraric acid, its common name. Maybe it was listed under this name? I will admit that this is probably one of the best example of the need for CAS (CASes?)...

In any case, I really enjoyed reading this thread and I learned a lot. If you have any request, questions or suggestions, I would love to help!

Thank you for posting here - it is really valuable and your honesty and clear explanations are a great addition to the thread.

As to 3 decimal places, the reason this can be key for us is that it is frequently the lowest amount at which materials are direct-added to a fragrance. I add materials at 0.001 and upwards in my 1kg formulas so this can still be a very useful number for aromachems. Obviously once you get into essential oils those numbers can have less value (but not always - sometimes trace elements unlock clues to the use of rare materials).

 

[Phytochemia]

As to 3 decimal places, the reason this can be key for us is that it is frequently the lowest amount at which materials are direct-added to a fragrance. I add materials at 0.001 and upwards in my 1kg formulas so this can still be a very useful number for aromachems.

Yes, I agree with you, especially with aromachemical that are very potent. By sense I did not meant usefulness in the blend, more than the exactitude of the value.

In analytical chemistry there is something called the Horwitz Trumpet

What this indicate is that as you go lower in concentration, the relative error on the mesurement increase significatively. While you could expect that a compound that is measured at 50% can in reality be between 48-52%, a compound measured at 0.001% has a larger variation. For all intent and purpose, a 0.001% compound could be at anything between 0.005 and 0.0005%. We tend to indicate everything that is under 0.005% but detected to be "tr" (or trace) since the quantification value is too inexact to be reliable at this level. Their presence is, however, still important in the blend. The chance for a false positive identification also increase exponentially at these level.

Obviously once you get into essential oils those numbers can have less value (but not always - sometimes trace elements unlock clues to the use of rare materials)

This is a very good an interesting point. We have been playing around with clustering and statistical models on larger essential oils dataset and what we found out is that there are a lot of "latent" chemotypes that exist that can be determined on minor compounds.

 

[tensor9]

For the cutting edge, refer to this application note. I have requested it be sent in full.

Basically it’s GCxGC (2D GC), which gives better separation, along with MS (mass spectrometry) AND FID (flame ionization detection) in parallel, so you get both the identification and quantification benefits.

https://www.chromatographytoday.com...ion-for-improved-analysis-of-fragrances/50274

 

[Phytochemia]

Basically it’s GCxGC (2D GC), which gives better separation, along with MS (mass spectrometry) AND FID (flame ionization detection) in parallel, so you get both the identification and quantification benefits.

We have investigated this technic a bit, our conclusion is that the gain is minor since you are always limited by the column load, detector sensitivity and databases. It requires a lot more work for a minor return. It is not without merit though, and it can be a good tool in a more complex proccess.

 

[giftmischer]

These two approaches, GCxGC and the 2 polar/apolar columns of Phytochemia, are very appealing. Phytochemia does an excellent job with essential oils and perfume blends. But as H.Marceau from Phytochemia points out, it is also the sensitivity and above all the database(s).
First, there are so many olfactory very important compounds that you will never see on a list - of some others I'm glad when I find a trace.
Second, percentages are incorrect snapshots. There are no 100% reports in GC unless you calculate it like w2lab to make it fit.
(I would even say it's coincidence if the perfume formula fits better)
Third, the NIST database, for example, is nice but very patchy; The FFNSC is much nicer, but also inadequate. There are many compounds known as "single compound" that give (in reality) three or more peaks on the GC and you will not find a CAS number or RI for them. So IsoE, galaoxlide and sandela are just prominent examples, but also givescone, cashmeran, sandalore, tabanone, helvetolide, lilial, romandolide, hedione, ...
It needs many MS and RI and experience to get a good database. So what I was going to say: It's the database, folks...

 

[Big L]

So what I was going to say: It's the database, folks...

I wanted to write this reply when someone here mentioned the possibility of buying a GC machine, but it is much more credible coming from you. Thank you, giftmischer.

 

[minorstep]

So do they have a website after all? Where are they selling this? Struggling to find anything about them.

It is very odd and a bit disconcerting to see my own fragrance GCMS for sale with WB2 Lab.
At least SOMEBODY wants to know about it... LOL!

 

[Eric Anthony]

So do they have a website after all? Where are they selling this? Struggling to find anything about them.

You'll have to email him, he sent me a PDF with their wares. Not a shabby list, I purchased a few in curiosity.

 

[DarioSiegel]

Hello.

someone interested in sharing /exchanging formulation? (if allowed)

 

[Lucofborg]

I'm mostly interested in gcms analysis. If you have some we can discuss.