Thursday, July 7, 2016

Distinguishing chlorine and bromine by 1H-13C HSQC

Natural products often incorporate chlorine or bromine, but less rarely are both halogens found in the same molecule. When they are, the problem becomes how to distinguish them. Typically, chlorinated carbon resonances are assumed to be shifted further downfield than brominated ones. This rule of thumb was of no help to one of our users who recently isolated a compound containing both chlorine and bromine because the two halogenated carbons had very similar chemical shifts. How then could we distinguish the two halogens?

The table below shows some NMR parameters of chlorine and bromine, with the corresponding 1H parameters for comparison.

Nucleus 1H 35Cl 37Cl 79Br 81Br
Nuclear spin 12 32 32 32 32
Gyromagnetic ratio
(107 rad/T.s)
26.7519 2.624 2.1842 6.7228 7.2468
Quadrupole moment
(1028 Q/m2)
0 -0.1 -0.079 0.37 0.31
Resonance frequency
at 14.1 T(MHz)
600.0000 58.7892 48.9348 150.3204 162.0372
Natural abundance 99.98 75.53 24.47 50.54 49.46
Relative sensitivity 100.00 0.47 0.271 7.86 9.85

Since chlorine and bromine both have two NMR active nuclei a chlorinated or brominated carbon should show two peaks because the different isotopes will shield the carbon nucleus differently, thereby inducing different chemical shifts. Since the natural abundance ratio of the two chlorine nuclei is roughly 3:1, while for bromine its 1:1, the chlorinated carbon peaks will have different intensities while the brominated carbon peaks will be of equal intensity. Since the chemical shift difference is likely to be small, a high resolution 13C spectrum would be required to resolve the peaks.

A 1D 13C experiment offers the easiest way to obtain a high resolution 13C spectrum but in this case there was insufficient material to obtain a 1D 13C spectrum. As an alternative, the 13C resolution of the 2D 1H-13C HSQC experiment was optimized to obtain an indirectly detected 13C spectrum. Band selective 13C excitation enabled the 13C sweep width to be reduced to a few ppm. Reducing the 1H sweep width also allowed the entire dynamic range of the receiver to be applied to the peak of interest. Acquiring a 1H-13C HSQC with these optimized parameters, and using a 1.7 mm probe, gave sufficient sensitivity and resolution to resolve splitting of the resonances of interest.

The figure below shows the structure of the natural product including the halogenated carbons. The expansions of the band selective HSQC show the H8-C8 resonance (left) and H9-C9 resonance (right).

Clearly, the C8 resonance has a shoulder while the C9 does not. Fitting two peaks to a 13C slice running through the C8 resonance gave the expected 3:1 intensity ratio and a slight difference in linewidth that would be expected from the difference in the quadrupole moments. This indicates that C8 bears a chlorine atom. The lack of splitting in the C9 resonance, which must be brominated if C8 is chlorinated, could be due to a number of reasons. The larger quadrupole moment of bromine likely broadens the resonances more than chlorine does, making resolution of the peaks more difficult. It is also possible that the chemical shift difference induced by the bromine nuclei is less than that induced by the chlorine nuclei, making it harder to observe two peaks. Indeed, band-selective HSQC experiments acquired on a series of brominated test compounds failed to reveal any peak splitting. Chlorinated test compounds were also examined, including a dicholorinated compound for which the carbon resonance was split into three peaks (C-35Cl2, C-35Cl37Cl, and C-37Cl2). For more information please see the papers.

1. Xiao Wang, Brendan M. Duggan and Tadeusz F. Molinski.
Mollenynes B-E from the marine sponge Spirastrella mollis. Band-selective heteronuclear single quantum coherence for discrimination of bromo-chloro regioisomerism in natural products.
J Am Chem Soc. 2015 Sep 30;137(38):12343-51.

2. Xiao Wang, Brendan M. Duggan and Tadeusz F. Molinski.
Ultra-high resolution band-selective HSQC for nanomole-scale identification of chlorine-substituted 13 C in natural products drug discovery.
Magn Reson Chem. 2016 Mar 7

No comments:

Post a Comment