Molecular cartography - how's it done?
May 31, 2009 7:11 AM   Subscribe

What techniques do chemists use to map the molecular structure of a substance?
posted by shivohum to Science & Nature (15 answers total) 1 user marked this as a favorite
 
you mean like x-ray crystallography?
posted by sero_venientibus_ossa at 7:29 AM on May 31, 2009


NMR Spectroscopy and IR Spectroscopy as well.
posted by bitterpants at 7:32 AM on May 31, 2009


X-ray diffraction lets chemists determine the crystalline structure of a material. The short version of how it works: an x-ray beam with a known wavelength is aimed at the sample. The beam hits atoms in the crystal and is diffracted. At some angles, the diffracted beams from different atoms will be out of phase and interfere destructively and will not be detected at other angles, they will be in phase and will hit a detector plate. The crystal structure can be determined based on the angles that cause constructive interference.

I don't know as much about non-crystalline materials, but I know scientists can determine the composition of an unknown substance with mass spectrometry and then make predictions about the structure based on properties like density, solubility, melting and boiling points, etc.
posted by martinX's bellbottoms at 7:33 AM on May 31, 2009


I hadn't heard the term "molecular cartography" before (though I'm not a scientist) so I went a-Googlin' and came up with this 1986 journal article "Molecular Cartography of Globular Proteins with Application to Antigenic Sites" which goes into lots of depth. Though it appears that the term is referring to a very specific technique at least there; you might be asking about more general molecular imaging methods like other people are mentioning.
posted by XMLicious at 7:40 AM on May 31, 2009


NMR Spectroscopy is one of the most powerful techniques for structural analysis (especially for inorganic compounds). Other techniques such as IR Spectroscopy and UV-Vis Spectroscopy help elucidate functional groups. Mass spectrometry allows you to find the exact mass of the molecule and empirical analysis tells you how "much" of each element is in the molecule. And as said above, X-ray Diffraction allows you to determine the crystal structure of a crystalline material.
posted by stevechemist at 8:16 AM on May 31, 2009


I would say Mass Spectrometry and NMR are the two most powerful tools. Mass Spec helps you determine what elements are present and how many - when you get an m/z (mass-to-charge ratio) up to the 4-5th decimal place there are only so many combinations of atoms that are possible to have a total mass that is that exact. NMR then helps you see how all the atoms are positioned next to each other.

You can do C13 NMR to determine how many different species of carbons there are in the molecule (and how many of each type in relation to each other), can do H1 NMR to determine how many types of protons (and how many of each in relation to each other as well), F19 NMR to see if there are Fluorines. From the chemical shifts in H1 NMR you can tell whether the H is sitting on an alkyl chain, or if it's part of a double bond, part of a carbonyl group, etc. NMR also allows you to see which protons are coupled with each other, if you know how to interpret complicated NMR data like that.

Then from Mass Spec you can see the molecular weight of the molecule (usually... sometimes you see the molecular weight of the fragments or the dimer, but lets ignore that). From the MW you can tell if there is an odd or even # of Nitrogens. You can also tell from the isotope pattern how many carbons there are, how many (if any sulfurs there are - if there's a sulfur present you'd see an M+2 peak at around 6% i believe?), if there are Chlorines or Bromines present. There are really powerful softwares that can analyze the isotope pattern and give you a prediction of the chemical formula.

You can also use IR to see if there are specific functional groups present, such as OH groups, carbonyl groups, etc.

X-ray diffraction will let you see if the material is crystalline or amorphous (and what % crystalline and amorphous if you have a standard).

And of course, everything is more complicated then I just wrote above, and it doesn't always work that easy - hey lets do Mass Spec and it'll tell us the MW - you have to know how to set up the experiment parameters, know which peaks to look at, how to determine if something is an adduct, etc.

Fascinating stuff.
posted by KateHasQuestions at 8:21 AM on May 31, 2009


It depends partly on the size and the stability of the compound.

NMR is like an MRI for molecules. Basically, you're looking at all of the instances of a certain element in the structure (most commonly hydrogen and carbon); these atoms behave a little differently, depending on what other atoms and molecular groups are nearby. When you know what atoms are in the compound to begin with, you can use these signals to figure out which atoms are near which other sorts of atoms and groups, and thus what the basic structure is. NMR is increasingly used for large biomolecules like DNA and proteins, but it's still difficult because it is often hard to get enough material; traditionally, NMR has been particularly the tool of organic chemists. There are a lot of derivatives like varieties of 2D-NMR (NOSEY, COSY, HMBQ, etc.) that can let you get even more detailed information.

EPR is rather like NMR, but it looks for atoms with an unpaired electron: mostly atoms with radicals or transition metals. Related techniques include ENDOR & ELDOR, pulsed EPR, and probably more things I don't know about.

Mass spectrometry is used to figure out the overall size of the molecule. Essentially, molecules are ionized and shot past a magnet; they travel differently based on size, and so you can tell what the overall size of the molecule is. Under certain circumstances, groups of molecules can be knocked off; these groups have well-known sizes, and so you can also get an idea of what functional groups your molecule has. Gas chromatography/mass spectrometry can help differentiate between different materials of the same size (I'll get to chromatography in a second.) Everyone uses MS, from chemists to biologists. There are a lot of derivative methods - MALDI and ESI ionisation, MS/MS, GC/MS, LC/MS, etc.

Infrared spectroscopy can distinguish between different sorts of bonds between atoms, because those bonds have different strengths and energies. UV/visible light spectroscopy is used a lot by biochemists to determine the amount of DNA or protein in a sample, because DNA and certain amino acids absorb light at 260 and 280nm, respectively. There are other forms of spectroscopy - Raman, Mößbauer (uses gamma radiation, most useful for compounds containing iron), etc. Main uses of the method depend a lot on which method we're talking about; IR and Raman are used a lot by physical chemists, while the UV-vis range is mostly biologists.

X-ray crystallography is used particularly by inorganic chemists and biochemists, and a little less commonly by organic chemists. In inorganic chemistry, it's great for figure out the crystal structures of compounds. In biochemistry, large molecules like proteins and DNA can be manipulated into forming crystals under certain conditions, and then their structures can be analyzed via the same method. For all molecules, you're basically aiming an x-ray at your crystal; when it hits the crystal, it scatters, and by measuring strength and angle of the scattered beams, you can figure out the crystal structure. This gets very complicated for huge molecules like proteins, and so protein crystallography has really come into its own since the development of powerful computers and good software. Using X-ray crystallography on biomolecules in general remains a challenge, because it can be very, very hard to get them to crystallize (particularly membrane proteins, where part of the protein wants to be in a polar (water-like) environment, and part of it wants to be in a non-polar (oil-like) environment.)

Chromatography covers a wide range of techniques. In all of them, you have some sort of matrix - a column full of beads, for example - which some molecules will pass through easier than others; by marking how fast things pass through, you can figure out what your sample contains. Column chromatography includes particularly HPLC (great for organic compounds and smaller biochem ones like some sorts of DNA) and FPLC (great for larger things like proteins.) Other methods sort things based on molecular binding (affinity chromatography), ionic or polar nature of the compound (many varieties), solvent solubility (TLC), etc. Even the common bio technique of gel electrophoresis to sort DNA or proteins by size is much like chromatography.

There are also lots of microscope-based methods used for biological samples, but these are better for figuring out interactions between macromolecules than for figuring out the structure of a single molecule.

Those are some of the main methods, but there are more, of course, and many of these methods can be combined, or at least applied to different samples of the same molecule.
posted by ubersturm at 8:23 AM on May 31, 2009 [1 favorite]


Response by poster: Great, this is just what I was looking for, thanks!
posted by shivohum at 9:05 AM on May 31, 2009


Is this homework-filter? The above answers are great, but I want to add a few of the newer ones:

Recent advances in Cryoelectron microscopy, particularly the availability of high magnetic field strengths, has meant that it is increasingly being used for meso-scale objects, including biological and nano materials.

Atomic force microscopy has been used similarly for more than a decade, and now it is even being used to look at conformations of small molecules like stillbenes.

Multidimensional infrared spectroscopy has moved that field far from it's 'identifying functional group' roots. There was a recent PNAS issue devoted to it. It is not crazy to think that it will supplant NMR at some point, as the instruments are a lot less expensive, and do not require liquid helium, which is a big problem with NMR (we seem to be at peak helium already).
posted by overhauser at 10:38 AM on May 31, 2009


Oops, messed up the PNAS link.
posted by overhauser at 10:45 AM on May 31, 2009


Maybe it's a bit self-serving as a theoretical chemist, but don't forget molecular modelling, electronic structure, etc. In many ways it's a confirmatory technique, but it has its place even as a predictive technique.
posted by Fortran at 11:19 AM on May 31, 2009


And don't forget the totally old-school methods, either! If some reaction— chemical, or thermal, or whatever— produces a known substance, or a particular ratio of substances, or whatever, then you've learned a tiny bit about the structure of the mystery molecule. After a vast amount of experimentation, you can maybe work out the molecule's structure. This is how most small molecules' structures were investigated before techniques like x-ray crystallography were developed.
posted by hattifattener at 11:51 AM on May 31, 2009


I was really struck when I saw how powerful and direct H1 and C13 NMR was for getting at the structure of organic molecules. With really not a whole lot of training (a couple quarters of ochem and a couple of reference sheets, really) you can pick up an incredible amount of information from an NMR spectrum. Seriously cool stuff.
posted by devilsbrigade at 12:48 PM on May 31, 2009


Molecular biologists are also interested in how macromolecules move during reactions - for a beautiful example of this, check out ATP synthase.

Such dynamic motion can be inferred from sets or changes in interacting partners as captured by cross-linking studies.
posted by Jorus at 7:14 AM on June 1, 2009


In the case of pentacene, the structure has recently been shown directly by Atomic Force Microscopy (paper, BBC News article).

I should emphasize that pentacene's structure was already known, and that this is not a routine method finding a molecule's structure, but it is a beautiful proof nonetheless.
posted by James Scott-Brown at 12:38 PM on November 17, 2009


« Older Getting a Guitar and Computer Back to the US   |   How to save streamed programs permanently? Newer »
This thread is closed to new comments.