Residual dipolar couplings as a powerful tool for constitutional analysis: The unexpected formation of tricyclic compounds

Article abstract of DOI:10.1002/anie.201007305

Analyze this! Residual dipolar couplings (RDCs) can be used to determine the constitution of a small molecule when traditional methods for structure elucidation fail. In a case study, a highly congested, tricyclic compound resulting from the electrophilic

Full text of DOI:10.1002/anie.201007305

DOI: 10.1002/anie.201007305
Structure Elucidation
Residual Dipolar Couplings as a Powerful Tool for Constitutional
Analysis: The Unexpected Formation of Tricyclic Compounds**
Grit Kummerlꢀwe, Benedikt Crone, Manuel Kretschmer, Stefan F. Kirsch,* and Burkhard Luy*
Anisotropic NMR parameters in partially aligned samples,
such as residual dipolar couplings (RDCs), residual chemical
shift anisotropy (RCSA), and residual quadrupolar couplings
(RQCs), contain valuable structural information.[1] As has
been shown on a multitude of examples, RDCs are a useful
tool to determine the configuration[2] and the conformation[3]
of small to medium-sized organic molecules. Herein, we now
add a further facet by demonstrating the power of RDCs in
the analysis of the constitution of an unknown small molecule.
The molecule we investigated in this case study is one of
the products obtained by reacting the azide-containing 1,5-
enyne 1 in the presence of electrophilic iodine sources
(Scheme 1). Recently, it was reported that enyne 1 (R1 =
led to the generation of the unknown compound 4 in low
yield.[5] These seminal studies, while clearly indicating that the
iodonium-induced carbocyclization of enynes shows great
promise, leave the unanswered question of what the structure
of 4 is. Assuming that the electrophilic cyclization of enynes
will become a growing field,[6] we felt that the structure
elucidation of compound 4 would be indispensable for a
better understanding of the mechanisms at work.
Classical methods for the structure determination of small
quantities of compounds, including mass spectrometry, IR
spectroscopy, and conventional NMR experiments like one-
dimensional (1D) 1H and 13C, and two-dimensional (2D)
COSY, HSQC, and HMBC experiments could not be used to
solve the constitution of compound 4. The molecular formula
C16H18IN and eleven fragments could be identified: a phenyl
group, a methyl group, five methylene groups (three forming
an isolated chain), a tertiary nitrogen atom, an iodine atom,
and four quaternary carbon atoms (see the Supporting
1
Information). The H,13C HMBC spectrum revealed 63 and
the 1H,15N HMBC spectrum 7 cross peaks, correlating almost
every fragment with every other fragment, and thus only
indicating a very compact structure. Even the acquisition of a
2D 1,1-ADEQUATE spectrum[7] could not solve the struc-
ture, although five additional 13C,13C correlations could be
identified, reducing the number of fragments to six (see the
Supporting Information).
Scheme 1. Reaction leading to the formation of unknown product 4.
NIS=N-iodosuccinimide.
Since classical NMR analysis failed, we decided to
approach the problem in an unconventional way by adopting
residual dipolar couplings. RDCs and other anisotropic NMR
parameters contain unique angular information of internu-
clear vectors with respect to the static magnetic field which
has been shown to be useful for the verification/falsification of
a proposed configuration[2] or conformation.[3] We therefore
assumed that as long as sufficient anisotropic parameters can
be measured and a large enough set of structural models can
be constructed, it should also be possible to identify the
correct constitution of our reaction product 4.
Me, R2 = Ph) can be selectively transformed into either aryl
2 or cyclohexadiene 3, depending on the exact conditions.[4]
Additionally, in studies of the reactivity of enyne 1 with I2 and
K3PO4, it was surprisingly found that temperatures above 08C
[*] Dr. G. Kummerlꢀwe,[+] Dr. B. Crone, M. Kretschmer,
Prof. Dr. S. F. Kirsch, Prof. Dr. B. Luy[$]
Department Chemie, Technische Universitꢁt Mꢂnchen
Lichtenbergstrasse 4, 85747 Garching (Germany)
E-mail: stefan.kirsch@ch.tum.de
burkhard.luy@kit.edu
[+] Current address: Institut fꢂr Biologische Grenzflꢁchen (IBG-2),
Karlsruher Institut fꢂr Technologie
Postfach 3640, 76021 Karlsruhe (Germany)
[$] Current address: Institut fꢂr Organische Chemie
For the measurement of RDCs, we used a stretched
polystyrene/CDCl3 gel[8] as an alignment medium for the
induction of anisotropy. CLIP-HSQC spectra[9] were acquired
for an isotropic sample as well as for the anisotropic gel
1
sample, and DCH RDCs were extracted as the difference
between the corresponding couplings measured. In addition,
2DHH RDCs between the geminal protons of methylene
groups were obtained from corresponding P.E.HSQC spec-
tra.[10] Altogether 17 RDCs were gained for the structural
analysis (see Supporting Information).
As the next step, we created a set of structures of
compound 4 to be tested (Scheme 2), which fulfill the
molecular formula C16H18IN, the basic fragments known
Karlsruher Institut fꢂr Technologie
Fritz-Haber-Weg 6, 76131 Karlsruhe (Germany)
[**] S.F.K. thanks the Deutsche Forschungsgemeinschaft (DFG) and the
Fonds der Chemischen Industrie (FCI) for support. B.L. thanks the
FCI and the DFG (Heisenberg program LU 835/2,3,4,7 and
Forschergruppe FOR 934).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201007305.
Angew. Chem. Int. Ed. 2011, 50, 2643 –2645
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2643

Communications
cantly from the experimental data.
We also did not expect a priori to
5
find two intensive JCH cross peaks
in the 1H,13C HMBC spectrum,
correlating two protons with the
ortho-carbons of the (rotating)
phenyl group (see the Supporting
Information).
Besides the interesting case
study for structure determination
by NMR spectroscopy, the identi-
fied constitution also reveals a novel
Scheme 2. Potential structures of 4.
from 1D, COSY, and HSQC spectra (see above), and, in
addition, make sense from the chemistꢀs point of view. To not
miss any unusual product, we also included several structural
models that were unlikely, either based on the reaction
mechanism or because of spectral data like individual HMBC
signals and chemical shift values.
To test whether the experimental RDCs are consistent
with the proposed structures, SVD fits using the program
PALES[11] (“bestFit” option) were made. Therefore, pdb files
of the 14 suggested input structures were created and energy-
minimized using the program Sybyl. Within the fitting
process, all prochiral assignments for the five methylene
groups were permutated, leading to 25 = 32 fits for each of the
14 structures. Furthermore, the assignment of the two isolated
methylene groups was varied as their position could not be
identified unambiguously on the basis of 1D, COSY, and
HSQC spectra. The number of fits therefore increased to 64
for each suggested molecule, leading to a total number of 64 ꢁ
14 = 896 SVD fits.
To compare the quality of the fits, n/c2 values[2e] were
calculated for each fit. In Figure 1a the resulting quality
factors for the best permutation of each of the 14 structures
are summarized (see the Supporting Information for all
values and a description of the quality factor). Clearly, only
the aziridine structure Ba has a n/c2 value significantly larger
than 1, indicating a good agreement with experimental data.
Also, the comparison of measured and back-calculated RDCs
(Figure 1b) demonstrates that this structural model is con-
sistent with the experimental data. As for all structures other
than Ba, the quality factor for the SVD fit is poor and
corresponding structures can be excluded, Ba can be consid-
ered to be the correct constitution of reaction product 4.
To make sure that the constitution determined by RDCs is
the correct one, efforts were made to independently verify the
obtained result with alternative methods. Therefore almost
100 mg of the reaction product was synthesized and a 2D
INADEQUATE spectrum[12] was acquired, verifying the
carbon skeleton of the substance. Additional evidence could
be obtained by labeling the starting material of the reaction
with 15N-azide and measuring 13C,15N couplings for the 15N-
labeled compound. Both additional experiments clearly
support the structure determined by RDCs (see the Support-
ing Information).
Figure 1. a) Comparison of quality factors n/c2 for the SVD fits done
with PALES (for each structure only the permutation with the best n/c2
value is shown). b) Plot of back-calculated RDCs, D(calcd), against
experimental RDCs, D(meas.), for the best permutation of assignment
for structure Ba. c) Molecular formula of the determined reaction
product. d) The structural model of Ba depicted with color-coded
bonds (red: negative; blue: positive RDCs) and the axis of the
corresponding alignment tensor next to it.
domino reaction leading to the unexpected formation of
tricyclic compound 4.[13] As hypothesized in Scheme 3,
iodonium activation of the starting 1,5-enyne 1 gives cyclic
cation 5, which undergoes proton abstraction to give cyclo-
hexadiene 3 under the reaction conditions. Most likely,
aziridine 4 then results from intramolecular 1,3-dipolar
cycloaddition followed by loss of nitrogen.[14] This view that
aziridine formation proceeds via diene intermediate 3 and not
through direct cyclization[15] of the cationic intermediate 5 is
supported by the following observation: upon treatment with
I2 and K3PO4 in CH2Cl2 at room temperature, 3 smoothly
transforms into a mixture of aziridine 4 and aromatic
compound 2. It is particularly significant that simple heating
of cyclohexadiene 3 results in only traces of 4, thus indicating
that I2 might be involved in the aziridine-forming step. Our
Interestingly, the constitution Ba determined by RDCs
was almost excluded from the set of structures, because the
13C chemical shifts predicted by ChemDraw differed signifi-
2644 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 2643 –2645

Mo, M. J. Shapiro, T. A. Smitka, G. A. Stephenson, E. R. Zartler,
J. Am. Chem. Soc. 2004, 126, 5008 – 5017.
[3] a) J. Klages, C. Neubauer, M. Coles, H. Kessler, B. Luy,
ChemBioChem 2005, 6, 1672 – 1678; b) A. Schuetz, T. Mura-
kami, N. Takada, J. Junker, M. Hashimoto, C. Griesinger, Angew.
Chem. 2008, 120, 2062 – 2064; Angew. Chem. Int. Ed. 2008, 47,
2032 – 2034; c) C. Farꢄs, J. Hassfeld, D. Menche, T. Carlomagno,
Angew. Chem. 2008, 120, 3782 – 3786; Angew. Chem. Int. Ed.
2008, 47, 3722 – 3726; d) M. B. Schmid, M. Fleischmann, V.
DꢀElia, O. Reiser, W. Gronwald, R. M. Gschwind, ChemBio-
Chem 2009, 10, 440 – 444; e) V. M. Sꢅnchez-Pedregal, R. Santa-
marꢆa-Fernꢅndez, A. Navarro-Vꢅzquez, Org. Lett. 2009, 11,
1471 – 1474; f) R. S. Stoll, M. V. Peters, A. Kuhn, S. Heiles, R.
Goddard, M. Bꢇhl, C. M. Thiele, S. Hecht, J. Am. Chem. Soc.
2009, 131, 357 – 367; g) M. U. Kiran, A. Sudhakar, J. Klages, G.
Kummerlꢂwe, B. Luy, B. Jagadeesh, J. Am. Chem. Soc. 2009, 131,
15590 – 15591.
[4] B. Crone, S. F. Kirsch, K.-D. Umland, Angew. Chem. 2010, 122,
4765 – 4768; Angew. Chem. Int. Ed. 2010, 49, 4661 – 4664.
[5] Broadening the scope of this unexpected result was difficult
since, even after attempted optimization compounds of type 4
were typically formed in only low yields (R1 = Me, R2 = 2-
naphthyl, 23%; R1 = Me, R2 = 4-MeO2C-(C6H4), 23%; R1 = Me,
R2 = Me, 32%; R1 = Bu, R2 = Ph, 30%) along with arenes of
type 2.
Scheme 3. Possible mechanism for the formation of aziridine 4.
results on the synthesis of tricyclic aziridine 4 represent a
remarkable example of how easily molecular complexity can
be generated from simple acyclic precursors. Moreover, our
results are in perfect agreement with the previous report[4]
that in the electrophilic cyclization of 1,5-enyne 1 the ultimate
products (2 or 4) are typically formed via key intermediate 3.
In summary, we have shown that RDCs are a valuable tool
to determine the constitution of unknown compounds, thus
offering a novel approach to the structure elucidation of small
molecules. Our method provided evidence for the unexpected
formation of a previously unidentified tricyclic compound by
which novel insights into the electrophilic cyclization of 1,5-
enynes were revealed. Although additional investigations are
warranted to define the scope of RDCs more precisely, this
study constitutes an important step toward a prolific use of
RDCs as a powerful alternative and complementary approach
to commonly used spectroscopic methods for structure
determination.
[6] For recent reports on electrophilic enyne cyclizations, see: a) J.
Barluenga, J. M. Gonzꢅlez, P. J. Campos, G. Asensio, Angew.
Chem. 1988, 100, 1604 – 1605; Angew. Chem. Int. Ed. Engl. 1988,
27, 1546 – 1547; b) S. Shin, C. Lim, M. S. Rao, Synlett 2010, 368 –
373; c) R. Sanz, A. Martꢆnez, P. Garcꢆa-Garcꢆa, M. A. Fernꢅn-
dez-Rodrꢆguez, M. A. Rashid, F. Rodrꢆguez, Chem. Commun.
2010, 46, 7427 – 7429; d) A. Pradal, A. Nasr, P. Y. Toullec, V.
Michelet, Org. Lett. 2010, 12, 5222 – 5225.
[7] B. Reif, M. Kꢂck, R. Kerssebaum, H. Kang, W. Fenical, C.
Griesinger, J. Magn. Reson. A 1996, 118, 282 – 285.
[8] a) B. Luy, K. Kobzar, H. Kessler, Angew. Chem. 2004, 116, 1112 –
1115; Angew. Chem. Int. Ed. 2004, 43, 1092 – 1094; b) B. Luy, K.
Kobzar, S. Knꢂr, J. Furrer, D. Heckmann, H. Kessler, J. Am.
Chem. Soc. 2005, 127, 6459 – 6465; c) G. Kummerlꢂwe, S. Knꢂr,
A. O. Frank, T. Paululat, H. Kessler, B. Luy, Chem. Commun.
2008, 44, 5722 – 5724.
[9] A. Enthart, J. C. Freudenberger, J. Furrer, H. Kessler, B. Luy, J.
Magn. Reson. 2008, 192, 314 – 322.
[10] P. Tzvetkova, S. Simova, B. Luy, J. Magn. Reson. 2007, 186, 193 –
200.
[11] M. Zweckstetter, Nat. Protoc. 2008, 3, 679 – 690.
[12] A. Bax, R. Freeman, T. A. Frenkiel, M. H. Levitt, J. Magn.
Reson. 1981, 43, 478 – 483.
[13] For recent reviews on aziridine synthesis, see: a) A. Padwa, S. S.
Murphree, ARKIVOC 2006, 6 – 33; b) G. S. Singh, M.
DꢀHooghe, N. De Kimpe, Chem. Rev. 2007, 107, 2080 – 2135.
[14] For a leading reference, see inter alia: P. A. S. Smith, S. P. Chou,
J. Org. Chem. 1981, 46, 3970 – 3977.
[15] For selected examples of the reaction of azides with carbocations
see: a) W. H. Pearson, R. Walavalkar, J. M. Schkeryantz, W. K.
Fang, J. D. Blickensdorf, J. Am. Chem. Soc. 1993, 115, 10183 –
10194; b) G. L. Milligan, C. J. Mossman, J. Aube, J. Am. Chem.
Soc. 1995, 117, 10449 – 10459.
Received: November 20, 2010
Published online: February 18, 2011
Keywords: electrophiles · enynes · NMR spectroscopy ·
.
residual dipolar couplings · structure elucidation
[1] a) C. M. Thiele, Eur. J. Org. Chem. 2008, 5673 – 5685; b) G.
Kummerlꢂwe, B. Luy, Annu. Rep. NMR Spectrosc. 2009, 68,
193 – 232; c) G. Kummerlꢂwe, B. Luy, TrAC Trends Anal. Chem.
2009, 28, 483 – 493; d) C. M. Thiele, Concepts Magn. Reson. Part
A 2007, 30, 65 – 80.
[2] a) C. M. Thiele, A. Marx, R. Berger, J. Fischer, M. Biel, A.
Giannis, Angew. Chem. 2006, 118, 4566 – 4571; Angew. Chem.
Int. Ed. 2006, 45, 4455 – 4460; b) C. Aroulanda, V. Boucard, F.
Guibꢃ, J. Courtieu, D. Merlet, Chem. Eur. J. 2003, 9, 4536 – 4539;
c) R. R. Gil, C. Gayathri, N. V. Tsarevsky, K. Matyjaszewski, J.
Org. Chem. 2008, 73, 840 – 848; d) G. Kummerlꢂwe, S. Schmitt,
B. Luy, Open Spectrosc. J. 2010, 4, 16 – 27; e) D. Intelmann, G.
Kummerlꢂwe, G. Haseleu, N. Desmer, K. Schulze, R. Frꢂhlich,
O. Frank, B. Luy, T. Hofmann, Chem. Eur. J. 2009, 15, 13047 –
13058; f) J. L. Yan, F. Delaglio, A. Kaerner, A. D. Kline, H. P.
Angew. Chem. Int. Ed. 2011, 50, 2643 –2645
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2645