2
|
Hi, I have synthesized coproporphyrin II tetramethyl ester and taken a 1H NMR in CDCl3: http://illumina-chemie.de/upload/5_11600286385136157fa7971.png Especially the peaks at 10.1 ppm, belonging to the CH groups connecting the pyrrole rings, show a strong concentration dependence. They should be two singlets, but in fact they are only singlets at very low concentrations and split up with increasing concentration: http://illumina-chemie.de/upload/5_14276994513616989db39.png To me that makes it look like it's an aggregation thing, probably monomer/dimer equilibrium, and that is in agreement with the literature (Abraham, R. J., et al., J. Chem. Soc. (B), 621 (1966)). It is, however, not in agreement with 1H DOSY spectra taken at 2.7 mM and 9.0 mM. These only show one porphyrin species, in both cases around D = 1E-5 m^2/s. Could it be such an equilibrium despite no signs for it in the DOSY spectra (if so, why doesn't it show?) and if not, what else could cause the concentration-dependent fine structure of the signals? Thanks for your help. Jan |
|
2
|
What about magnetic alignment? A large planar conjugated system like this may have a tendency to align, and to some extent it may be cooperative and therefore concentration-dependent. I noticed that the "center of gravity" of the 10 ppm peaks is concentration-dependent. (Center of gravity is also known as the first moment. The midpoint of the integral is one easy way to quantify the first moment.) In the case of alignment, a peak may shift due to the chemical shift anisotropy, and it may split due to through-space dipole-dipole couplings. In this system, the strongest couplings would be associated with the nearest protons. The couplings are mutual, of course, so you should look for splittings elsewhere in the spectrum, particularly of the CH2 site "A" and the CH3 on the rings. In comparison to J couplings in isotropic solutions, dipolar splittings in aligned media can be very complicated. Don't expect the same rules to apply. For example, the protons in a methyl group split each other. You can probably find more information about NMR of aligned solutions by searching for "residual dipolar couplings" and/or research groups which study that sort of thing. |
|
2
|
Jan, The diffusion constant you mention is extremely fast - almost certainly too fast. 1E-10 to 1E-09 m^2/s would be a more reasonable range. Is that just a typo, or an indication that something misfired in the DOSY experiment? |
|
2
|
I think Tony and Alexander are right, about dipolar splittings and diffusion respectively. This is the classic sort of species one would expect to show magnetic orientation, and it may be that association is causing the effect to increase with concentration. For a species of this size one would expect a much lower diffusion coefficient than 10^-9 m^2/s in chloroform at room temperature. Diffusion measurements in chloroform are notoriously difficult; because chloroform is so mobile, it convects very easily and signal attenuation in PFG experiments is often dominated by convection. For example, on both our Bruker and Varian 500 MHz spectrometers there is only a very narrow window of a degree or so within which convection effects can be neglected: with the VT set outside this range there is easily measurable convection and significant non-diffusional attenuation. If you want to work in chloroform I suggest you use a 3 mm tube and keep the VT setting close to the quiescent temperature of the probe (i.e. the temperature in the absence of VT airflow), and check for convection either using an unbalanced convection-compensating sequence or by simply doing a diffusion measurement on a solution containing a high MW species (e.g. a polydimethylsiloxane). |
|
1
|
The splittings are on the order of 0.01 ppm, and by the formula In the dimer was very short-lived, say, less than 1 ms, I think DOSY would show only one diffusion rate, which would be the mean diffusion rate, and concentration-dependent. With such a short dimer lifetime, the small chemical shift differences (on the order of 0.01 ppm) would also be averaged, and the spectrum would simply show 2 singlets. As you reported, D is not significantly concentration-dependent, and there are splittings, so I am guessing it is not a case of dimerization. |
|
1
|
Thanks, both of you. That helps a lot. Alexander, yes, the diffusion constant was wrong. The value is what MestreNova gave using the Bayesian DOSY transform. I don't know whether some program settings are wrong or where else the problem lies, but with the same data sets Topspin gave D = 10E-9.1 m^2/s (again in both cases). Tony, that sounds quite reasonable. The methyl group signal indeed shows splitting which I had already wondered about, sorry I forgot to mention that. It is also concentration-dependent. Signal A is two overlapping triplets, but very broad (?), so no further splitting is visible there (and no discernible concentration dependence) : http://illumina-chemie.de/upload/5_1413284676513f84ccb871e.png ("A" CH2's around 4.4, CH3's around 3.65) Thanks again! Jan |
