MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2006; 44: 107–109
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1766
Rapid Communication
Long-range carbon–carbon connectivity via
unsymmetrical indirect covariance processing of HSQC
and HMBC NMR data
Kirill A. Blinov,1 Nicolay I. Larin,1 Antony J. Williams,2 Mark Zell3 and Gary E. Martin4∗
1 Advanced Chemistry Development, Moscow Department, 6 Akademik Bakulev Street, Moscow 117512, Russian Federation
2 Advanced Chemistry Development, 110 Yonge Street 14th Floor, Toronto M5C 1T4, Ontario, Canada
3 Pfizer Global Research and Development, Analytical Research and Development, 2800 Plymouth Road, Ann Arbor, Michigan 48105
4 Pfizer Global Research and Development, Analytical Research and Development, 7000 Portage Road, Kalamazoo, Michigan 49001-0199
Received 27 June 2005; Accepted 11 November 2005
It was recently demonstrated that an IDR- (Inverted Direct Response) HSQC-TOCSY data set could be
decomposed into a negatively phased direct response spectrum and a positively phased relayed response
spectrum that could then be subjected to unsymmetrical indirect covariance processing for the removal
of artifacts due to response overlap in the proton NMR spectrum of the molecule. Using experimentally
discrete HSQC and HMBC data sets, it is shown that unsymmetrical indirect covariance processing of the
pair ofNMRspectra affords a presentation containing long-range carbon–carbon connectivity information.
The method is demonstrated using strychnine as a model compound. The resulting data are largely free of
artifacts although artifacts can arise due to proton response overlap, as previously reported. Copyright 
2005 John Wiley & Sons, Ltd.
KEYWORDS: NMR; HSQC; HMBC; covariance processing; long-range coupling
Bru¨schweiler and co-workers have published a series of
reports dealing with various aspects of covariance NMR
spectroscopy.1 – 6 In their communication dealing with indi-
rect covariance NMR, Zhang and Bru¨schweiler4 noted that
cross correlation responses can appear between 13C reso-
nances of different spin systems if the carbons are cou-
pled to strongly overlapping protons. Pursuing Zhang and
Bru¨schweiler’s observation, we recently described two dis-
crete types of artifact responses that can be observed in
the case of IDR- (Inverted Direct Response) HSQC-TOCSY
spectra subjected to indirect covariance processing. Through
a modification of the covariance processing software, we
have also demonstrated how the artifact responses iden-
tified in our previous report can be eliminated.7 We now
wish to communicate a further extension of the capabil-
ities of the modified indirect covariance processing soft-
ware with experimentally discrete data sets. Rather than
decomposing a single data set into a pair of data sets, as
was done with the IDR-HSQC-TOCSY data to facilitate
the removal of artifacts,7 we have instead subjected dis-
cretely acquired HSQC and HMBC spectra to unsymmetrical
ŁCorrespondence to: Gary E. Martin, Pfizer Global Research and
Development, Analytical Research and Development, 7000 Portage
Road, Kalamazoo, Michigan 49001-0199.
E-mail: lighthousephoto@gmail.com
indirect covariance processing to obtain a presentation of
long-range 13C–13C correlation data.
Zhang and Bru¨schweiler4 observed in their communica-
tion dealing with indirect covariance NMR that the method
is, ‘. . .particularly useful for heteronuclear 2D experiments
that involve J-coupling, dipolar or cross-relaxation magne-
tization transfer from spin S evolving during t1, to multiple
spins I resonating during t2.’ Their observation, in conjunc-
tion with the software modification we reported previously,7
opens the possibility for the unsymmetrical indirect covari-
ance processing of an HSQC and HMBC spectrum to afford
the equivalent of a long-range carbon–carbon homonuclear
correlation spectrum.
There have been a scant few reports of studies involving
long-range carbon–carbon correlation spectroscopy using
INADEQUATE-based experiments. The earliest report of
which we are aware is that of Mueller and Bigler8 which dealt
with selective measurement of long-range carbon–carbon
coupling constants. Much more recently, Hori, et al.,9 used
a selective long-range INADEQUATE experiment to detect
a three-bond carbon–carbon coupling in hibarimicin dou-
bly labeled with 1,2-13C2 acetate. In contrast, there have
been a number of studies reported using variants of the 1,1-
ADEQUATE experiment first reported by Griesinger and co-
workers.10,11 A summary of the variants of the ADEQUATE
experiment, including 1,1-, n,1-, 1,n- and m,n-ADEQUATE
Copyright  2005 John Wiley & Sons, Ltd.

108 K. A. Blinov et al.
is provided in the 2003 report of Ko¨ck, Kerssebaum, and
Bermel.12 In a sense, the data in the indirect covariance NMR
spectrum obtained by the unsymmetrical coprocessing of a
multiplicity-edited HSQC spectrum and HMBC spectrum
provide correlation information analogous to that of the n,1-,
1,n- and m,n-ADEQUATE experiments albeit in long-range
carbon–carbon COSY form.12 In addition, the multiplicity-
edited nature of the long-range carbon–carbon connectivity
information in the unsymmetrical covariance presentation
is analogous to the multiplicity-edited ADEQUATE experi-
ments of Parella and Sa´nchez-Ferrando.13
For convenience, the multiplicity-edited HSQC and
HMBC spectra of strychnine (1) were acquired with identical 1
Figure 1. Carbon–carbon long-range correlation spectrum obtained by unsymmetrical indirect covariance processing of a
multiplicity-edited HSQC and an HMBC spectrum. Methines are denoted by red contours; methylene resonances are denoted by
black contours. Diagonally symmetric responses designated by solid red lines perpendicular to the diagonal, arise from mutual
long-range coupling; diagonally asymmetric long-range correlations are denoted by solid black lines and correspond to responses
where only the protonated carbon for which the correlation was detected is long-range coupled to the remote carbon resonance.
Hence, the correlations corresponding to C12-C23, C12-C8, C12-C11, and C12-C14 represent mutually coupled long-range
coupled resonances. The contour plot was plotted with a sufficiently low threshold to show the weaker correlations in the F1
frequency domain at the 13C shift of C12 in F2. The weaker long-range correlations include the 2JCH C12-C11 correlation as well as
the 4JCH correlations for C12-C7 and C12-C15. A single artifact response (C20*) is observed and arises through the overlap of the
shoulders of the H20b and H11b resonances in a manner analogous to artifacts discussed in the previous work.4,7 A projection
through the F2 frequency domain is plotted horizontally above the contour plot to show multiplicity; a 13C reference spectrum is
plotted vertically along the F1 axis.
Copyright  2005 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2006; 44: 107–109

Long-range carbon–carbon connectivity 109
digitization and spectral widths in both frequency domains.
The processed spectra were then subjected to unsymmetrical
indirect covariance processing using the modified software
described in our previous report.7 The aliphatic region of the
indirect covariance processed data is shown in Fig. 1. The
processed data has a diagonal typical of indirect covariance
processed data.4,7 Mutually long-range coupled resonant
pairs are diagonally symmetric; in the case of resonances
that are not mutually long-range coupled, responses are
diagonally asymmetric. The latter feature of these data
precludes the use of symmetrization to remove artifacts.7
Long-range correlation responses are observed to both
protonated and quaternary carbons ((1) and Fig. 1).
Methine resonances are represented by red contours;
methylene resonances are defined by black contours in
Fig. 1. For simplicity, only the long-range correlations
associated with C12 (horizontal solid black line) will be
discussed in the analysis of the long-range carbon–carbon
presentation shown in Fig. 1. The C12 and C23, C8, C11,
and C14 resonances are mutually long-range coupled and
give diagonally symmetric correlations identified by solid
red lines. A diagonally asymmetric correlation is observed
for C12-C7 and is denoted by a solid black line in Fig. 1. All
of these long-range correlations are legitimate, confirmed
in a 2–25 Hz optimized ACCORD-HMBC experiment in
our previous work.14 There is a single artifact response
observed among the long-range correlations to the 12-
position, this response (labeled C20* in Fig. 1) arises through
the overlap of the H20b and H11b proton resonances; the
latter coupled to C12.4,7 The artifact response is denoted
by the asymmetric dashed black line. Clearly, since C7 and
C20 have closely similar chemical shifts, care must be taken
when interpreting these data to make certain that artifact
responses are not mistakenly interpreted as legitimate long-
range carbon–carbon correlation responses.
EXPERIMENTAL
The NMR spectra for a sample of approximately 5 mg
of strychnine (1) dissolved in 160 µl of deuterochloroform
in a 3 mm NMR tube were acquired using a 500 MHz
Varian INOVA NMR spectrometer operating at a proton
observation frequency of 500.034 MHz and equipped with
a Varian 5 mm Cold Probe, the rf coils of which were
maintained at 25 K. Standard gradient multiplicity-edited
HSQC and HMBC experiments from the vendor-supplied
pulse sequence library were used without modification. The
one-bond delays in the HSQC and HMBC experiments were
optimized for 140 Hz; the long-range delay in the HMBC
spectrum was optimized for 6 Hz. Spectral widths in both
frequency domains of both experiments were held constant.
Both experiments were digitized with 1 K points in the F2
frequency domain and with 128 increments of the evolution
time, t1. The data sets were identically processed with the
exception of weighting factors, which were independently
optimized for the individual experiments.
The indirect covariance NMR spectrum shown in Fig. 1
was computed using the 2D NMR processing module of
a modified beta version of ACD/SpecManager developed
initially to operate on an IDR-HSQC-TOCSY data set
as described in our previous report.7 The processing
was performed using a PC with a 2.8 GHz Pentium IV
processor with 1 GB of RAM. Processing time for the
unsymmetrical indirect covariance spectrum shown in Fig. 1
was approximately 4 s.
CONCLUSIONS
We are continuing to explore the utilization of indirect
covariance NMR data as input for the ACD/Structure
Elucidator Computer-Assisted Structure Elucidation (CASE)
software package in both the form reported in this study
as well as in our previous work.7,8 The results of those
investigations will be the subject of a future report.
REFERENCES
1. Bru¨schweiler R, Zhang F. J. Chem. Phys. 2004; 120: 5253.
2. Bru¨schweiler R. J. Chem. Phys. 2004; 121: 409.
3. Trbovic N, Smirnov S, Zhang F, Bru¨schweiler R. J. Magn. Reson.
2004; 171: 277.
4. Zhang F, Bru¨schweiler R. J. Am. Chem. Soc. 2004; 126: 13 180.
5. Zhang F, Bru¨schweiler R. Chem. Phys. Chem. 2004; 5: 794.
6. Zhang F, Trbovia N, Wang J, Bru¨schweiler R. J. Magn. Reson.
2005; 174: 219.
7. Blinov KA, Larin NI, Kvasha MP, Moser A, Williams AJ,
Martin GE. Magn. Reson. Chem. 2005; 43: 999.
8. Mueller C, Bigler P. J. Magn. Reson. 1993; 102A: 42.
9. Hori H, Kajiura J, Igarashi Y, Furumai T, Higash K, Ishiyama T,
Uramoto M, Uehara Y, Oki T. J. Antibiot. 2002; 55: 46.
10. Reif B, Ko¨ck M, Kerssebaum R, Kang H, Fenical W, Griesinger C.
J. Magn. Reson. 1996; 118A: 282.
11. Ko¨ck M, Reif B, Fenical W, Griesinger C. Tetrahedron Lett. 1996;
37: 363.
12. Ko¨ck M, Kerssebaum R, Bermel W. Magn. Reson. Chem. 2003; 41:
65.
13. Parella T, Sa´nchez-Ferrando F. J. Magn. Reson. 2004; 166: 123.
14. Martin GE, Hadden CE, Crouch RC, Krishnamurthy VV. Magn.
Reson. Chem. 1999; 37: 517.
Copyright  2005 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2006; 44: 107–109