How to determine chemical structure

[gregxy]

Hi

There seem to be many people with knowlege of chemestry here
so I thought Id ask this.

Back when I was taking organic chem in college we did the
following to determine the structure of an unknown organic.

1. chemical tests to look for things like keto or aledhyde groups
2. Mass spectrometer to get overall atomic mass
3. NMR to identify different types of hydrogen in the sample.
4. optical polaization for sterio isomers.
5. Chromatography to separate different components.

There was alot of guessing involved and I think it would have
been impossible to distinguish between something like
normal testosterone and 1-test that differ only in the position
of the double bond.

Of course if you had a known sample it's much easier since you
only need to figure out if the spectra match. (Sledge talks
about this in his posts) But what if you don't have a sample?

Can you do something like generate the NMR spectrum via computer simulation (something like TINKER). Or is there
some "magic" box like they show on CSI where you put the
sample in and out comes a complete breakdown of what's in
it.

Thanks

[carcinogen]

Originally posted by gregxy
Hi

There seem to be many people with knowlege of chemestry here
so I thought Id ask this.

Back when I was taking organic chem in college we did the
following to determine the structure of an unknown organic.

1. chemical tests to look for things like keto or aledhyde groups
2. Mass spectrometer to get overall atomic mass
3. NMR to identify different types of hydrogen in the sample.
4. optical polaization for sterio isomers.
5. Chromatography to separate different components.

There was alot of guessing involved and I think it would have
been impossible to distinguish between something like
normal testosterone and 1-test that differ only in the position
of the double bond.

Of course if you had a known sample it's much easier since you
only need to figure out if the spectra match. (Sledge talks
about this in his posts) But what if you don't have a sample?

Can you do something like generate the NMR spectrum via computer simulation (something like TINKER). Or is there
some "magic" box like they show on CSI where you put the
sample in and out comes a complete breakdown of what's in
it.

Thanks

CSI has a lot of libraries with a thousands of compounds. Theoritcal IR spectra can be generated using programs as old as Fortran 77 (using quatum mechnics, employing atomic weights and bond strengths).

Chromatography is useful with a MS because of retention times. If you dont have a standard, there are published analytical methods that descibe relative retention times with an internal standard (to me relative retention times are more useful than absloute retention times). MS obviously gives MW and frament profile in mass/charge ratios (m/z).

To be quite honest, the three instruments needed to find orgainic molecular structure are NMR (proton or carbon), GC/MS and FT-IR. I say FT-IR because the resolution is so much better than a single scan IR. It really isnt that hard to interpret data from these on simple compounds (if your used to it).

[Androgenic]

Fourier Transform Infrared Spectroscopy is awesome for organic materials (possibly some inorganic). It measures the absorption of different (IR) infrared light wavelengths by the given material.Those infrared absorption bands can match and specify molecular components and structures which allows computerized data searches to be performed against reference libraries to identify a material (e.g. CSI stuff). Bands in the range of 4000 - 1500 are typically from functional groups (e.g. -OH, CH3, etc.). Between 1500 - 400 is the "fingerprint region" which are usually highl specific and is what is used to indentify them in the aforementioned computer databases. Essentially it just matches the bands and says this is...DHT, for example. Testosterone would have a different fingerprint. With so many organic compound specificity is essential and therefor the database is monsterous.

[carcinogen]

Originally posted by Androgenic
Fourier Transform Infrared Spectroscopy is awesome for organic materials (possibly some inorganic). It measures the absorption of different (IR) infrared light wavelengths by the given material.Those infrared absorption bands can match and specify molecular components and structures which allows computerized data searches to be performed against reference libraries to identify a material (e.g. CSI stuff). Bands in the range of 4000 - 1500 are typically from functional groups (e.g. -OH, CH3, etc.). Between 1500 - 400 is the "fingerprint region" which are usually highl specific and is what is used to indentify them in the aforementioned computer databases. Essentially it just matches the bands and says this is...DHT, for example. Testosterone would have a different fingerprint. With so many organic compound specificity is essential and therefor the database is monsterous.

I would agree with the fingerprint region of the FT-IR. Something that is sometimes forgotten is overtone (resonance overtones). The molecule may have a primary absorbtion for the functional group (4000-1500 cm-1), but the overtones (some happen in the fingerprint region as well as Near-IR region) really can help in discerning the structure and how the functional groups are connected to the molecule...

But with unknowns it is essential to have more tools at hand, esp. when researching new chemicals

[REN]

Originally posted by carcinogen
I would agree with the fingerprint region of the FT-IR. Something that is sometimes forgotten is overtone (resonance overtones). The molecule may have a primary absorbtion for the functional group (4000-1500 cm-1), but the overtones (some happen in the fingerprint region as well as Near-IR region) really can help in discerning the structure and how the functional groups are connected to the molecule...

But with unknowns it is essential to have more tools at hand, esp. when researching new chemicals

Yeah, IR will get you nowhere without a reference...

nowadays, though, i suppose just about every biologically active compound has been charactereized