I have some infrared mass spectrometry results, if I post them can someone here read them for me and interpret?

[Xorkoth]

Okay so I have a number of very rare drugs and a few I wanted to confirm. A friend of mine works with an organization that just developed a device that can read a tiny sample of a drug and print out the infrared mass spectra results, he claims it's a revolution is simplicity of identifying drugs. However basically none of the drugs I sent him have a reference stored into the machine so it just came up with guesses it considered close, but he told me not to worry, that it just didn't know, but he gave me the readouts as .png files. I'm hoping one of you intelligent souls can read this stuff and give me any insight. The three I care the most about are DOF, DOiP, and DOPr. The DOF in particular is, as far as I know, the only DOF anywhere. I wrote the first full trip report on it and, well, any other TRs forthcoming are likely to be from the same batch. There are also a couple of other drugs that are new to the scene that I don't know of anyone testing before. So, obviously, I'm hoping I can verify their identity.

What I'm going to do is post each in an NSFW with what it's supposed to be, what the machine said it was, and the spectra image. Thanks so much for any help. Keep in mind this is a brand new device so for all I know it might be inaccurate. It did get one of them correct and it was the one he said they had in their database already (escaline). This was free, I would have preferred to send off to a lab for GC/MS but being in the US, that's stupidly expensive. If anyone is able to help, you'd be doing a favor to the community as a whole because some of these things are extremely rare or just released for the first time from a single source.

Also my friend said that if, for example, their reference was for the HCl salt of a drug, and I have the fumarate or some other salt, it would come up wrong because of a different molecular weight.

NSFW:

Supposed to be DOF, the machine said 2C-E HCL - it is most absolutely clearly not 2C-E

NSFW:

Supposed to be BOD, the machine said 5,6-Methylenedioxy-2-aminoindane

NSFW:

Supposed to be proscaline, the machine said mescaline sulfate dihydrate

For reference, I had this one lab tested and it was actually good so if the readout with this one looks funny I think it means the machine doesn't work right.

NSFW:

Supposed to be escaline, the machine said escaline HCl so it matched, but would like to confirm

NSFW:

Supposed to be 4C-D (ARIADNE from PIHKAL), the machine said 2C-E HCl (same as it did for DOF). This is also EXTREMELY obviously not 2C-E.

NSFW:

Supposed to be 3C-P, the machine said 4-Chloropentedrone HCl. This is obviously not that as it matches duration, potency and style of effect of the 3C-P reports I've read, as opposed to some shitty cathinone.

NSFW:

Supposed to be 3C-E, the machine said 4-Fluoroamphetamine HCl. Suffice to say, it is not 4-FA by any stretch of the imagination

NSFW:

Supposed to be 5-MeO-MiPT fumarate, and the machine says it's 5-MeO-DPT HCl, which it couldn't be as the effects match 5-MeO-MiPT and 5-MeO-DPT is described as physically worrisome. But probably due to different molecular weight of fumarate salts.

Thanks again for any help or feedback. I really have to get some of these positively IDd and I'm hoping it's possible to do from the information in this post, but if not, let me know and I'll send to a lab.

 

[S.J.B.]

A friend of mine works with an organization that just developed a device that can read a tiny sample of a drug and print out the infrared mass spectra results, he claims it's a revolution is simplicity of identifying drugs.

Infrared spectra and mass spectra are completely different things. What you have here are infrared (IR) spectra. IR spectra give very little information that can be used to deduce the structure of an unknown sample unless you have a reference spectrum to compare it to. So, for example, you would never be able to determine if the first spectrum is of DOF unless you had an IR spectrum of pure DOF to compare it to (and usually, a computer program will do the comparison). It is certainly not nearly as powerful of an analytical technique as mass spectrometry, and it falls apart pretty quickly as soon as you start trying to look at impure/adulterated material.

Unfortunately, it seems like whatever system your friend has going is not very accurate. It likely matches to the closest compound in the database, but it seems unable to give a negative answer (i.e. "this compound does not match anything in our database") if there is something close that doesn't quite match, which means it is effectively useless when looking at esoteric designer psychedelics. You would never know if you have a true match or if the compound you actually have just isn't in the database. These are also pretty poor-quality IR spectra, which doesn't help.

I tried to look for published IR spectra of these compounds in order to compare them visually to the ones you posted, but unfortunately I could not find any of them in the scientific literature. For what it's worth, all these compounds are available from chemical suppliers, so your friend could order them and add them to his database.

I'm afraid you'll likely have to resort to mass spec if you want to be sure of what you have. Nuclear magnetic resonance (NMR) spectroscopy would work to, if you have any friends studying chemistry at a university. :D

 

[Nicomorphinist]

I was going to volunteer something similar to a quick summary of the above response, but still two questions:
I. Why are the spectra Not Safe For Work?
II. Can we continue this as a general-purpose GC/MS, FTIR &c spectra thread? Sort of like I Like To Draw Pictures of Random Molecules? That would be hell on wheels!

 

[Xorkoth]

I just didn't want the post to stretch forever so I put them in NSFWs.

Alright looks like I'll probably need to find some cost effective way to get these things tested via some other method. I don't think there's a free service in the US unfortunately, unless I just don't know about it.

 

[negrogesic]

Impossible to decipher without an IR reference for those. Maybe send to ecstacydata.org though at $100 a shot it will add up

 

[Bagseed]

it's pretty grand for someone to build an IR spectrometer and try to sell it as MS imho.... such a concept can only run on uniformed customers

 

[Nicomorphinist]

I just didn't want the post to stretch forever so I put them in NSFWs.

Alright looks like I'll probably need to find some cost effective way to get these things tested via some other method. I don't think there's a free service in the US unfortunately, unless I just don't know about it.

Given that there is an NSFW posting method, do folks post pr0n here? Are they supposed/allowed to do so?

Since smartphones and other small devices are so common these days, there are, most assuredly also things, many of them in the same category, which are Not Safe For Church/Synagogue/Mosque/Temple and so on.

Not Safe For Drug Deal . . .

 

[vecktor]

Infrared provides a wealth of information if you know how to interpret it, which it seems no posters in this thread do. The more concerning issue is that the promoter of this machine doesn't either.

For example the carbonyl stretch absorption at 1700 +/- tells you if there is a carbonyl C=O it also tells you what carbonyl type is present, so it is totally unacceptable to identify a hit on a carbonyl containing cathinone when there is no carbonyl absorption. PE have already got algorithms and a library that work pretty well and they have experimental 'Duh; filters which reduce mis-identifications where the lack of a diagnostic absorbance means it cannot be that compound.
I can also tell what the substitution pattern is on the phenyl ring by infrared but not using MS, In fact anyone reasonably skilled in the art can assign particular absorbances to particular bonds, if they have the time or inclination. There are wavenumber tables and interpretation of IR spectra books most from the 60s and 70s which are as correct today as they were then.
For simple structures of reasonably pure materials and be realistic these are all really simple molecules, a combination of IR and mass spectrometry can determine a semi unknown with a very high degree of certainty, much higher than any technique alone and much higher than NMR alone.
For example IR struggles to tell the difference between chloro and bromo and alkyl homologs but MS mass and more importantly isotopic ratios clear that up.

The maker/promoter of this machine is to use the technical scientific term full of shit. The different salts have different absorbances in the fingerprint region of the spectrum but this has nothing to do with molecular weight, they are different because they have different IR absorbing bonds.

The other major issue here is the quality of the spectrum and the questionable competence of whoever is designing this machine and the spectral comparison algorithm, on the right hand side of each spectrum at 500 cm-1 and below there is what looks like transmission cut off noise which is similar in every single sample, matrix multiplying peak matching algorithms and difference matching algorithms will disproportionately match the high intensity random spiky noise peaks which are clean sharp spikes. which then degrades the algorithm. why bother trying to record data in a region that is beyond the capabilities of the machine??
Rubbish in rubbish out.

 

[Xorkoth]

Ah bummer, well I'll let him know what I've found out here. Thanks for the input.

 

[vecktor]

Ah bummer, well I'll let him know what I've found out here. Thanks for the input.

The idea itself is a good one, but implementation isn't so great. This falls a bit short of the star trek Tricorder in its current implementation.

There are IR spectrum simulation models around they are reasonable for doing some things, for example the probablility of one group or another being present. This is also a classic application for AI, give it a commercial IR library (Hint Perkin Elmer) or non comercial library for example Japanese AIST SDBS and NIST as a training set with the functional groups outlined, and let the AI learn. Use the known strong wavenumbers out of a reference book as a filter, hint dot mutiply the known strong absorbances out of a book with the strong absorbances from the sample... either that or find some fossil of an analytical chemist that can still read and interpret IR spectra on sight and install that chemist in the machine.

Getting better data:
Cut out the regions where the machine cannot see properly, wavenumber <500 cut out the near IR at the other end because that doesn't seem to have much either. neither of these regions are going to help if the machine is blind.
adjust the signal using the sample prep so that the highest absorbance is 90% full scale the detail will then be much improved. plus weaker diagnostic signals will show.

I haven't got the time to go through the spectra properly and see if they are plausible for what you think they are, most likely they are correct, A quick glance the signal is a bit weak to show some the really useful information, like for example the relative amounts of aromatic and aliphatic C-Hs around 3000 cm-1, All of them are showing water but none of them are showing radically odd things.

 

[Hodor]

Supposed to be DOF, the machine said 2C-E HCL - it is most absolutely clearly not 2C-E

A peak at 3100-2800 indicates C-H bonds, with aromatic C-H bonds typically around 3050, while alkyl ones are at ~2900. Both 2C-E and DOF have aromatic and aliphatic C-H bonds. The C-O bonds of the methoxy groups typically produce two sharp peaks at 1200 and 1050 (which this spectrum does have), but again, that applies to both molecules. The only way to differentiate the two would be to look for the specific signal of a C-F bond.

Unfortunately, C-F stretches also tend to appear somewhere around 1200 so it is hard to assign any single peak to the C-F stretch without having a reference for comparison, or having a lot of experience with polysubstituted aromatic compounds with fluorine on them.
That is, unfortunately, the crux with IR spectroscopy: At low wavenumbers, in the so-called "fingerprint region", you get a pretty complex pattern comprised of a lot of information that can be used to identify a compound if you have a good reference spectrum of said compound; but just like a fingerprint, there is little you can infer about the person who left the print if you don't have a reference.

In general, I agree with everything vecktor posted.

At the very least, we can rule out your 3C-P being 4-Chloropentedrone, because the lack of a C=O stretch at 1700 cm^-1 (which we would expect from a cathinone) is a dead giveaway.

I might post more about this topic in the following days.

 

[Xorkoth]

Thanks for all the info, guys. vecktor, I'll pass along your feedback. I know they're an HR organization that is pretty passionate about what they're doing and this device is brand new, so if they can improve it I'm sure they'll want to.