2D ultrafast HMBC: A valuable tool for monitoring organic reactions

Article abstract of DOI:10.1021/ol902532f

Chemical Equation Presented Ultrafast 2D HMBC spectroscopy permits real-time monitoring of a reaction based on the structural changes produced in a carbonyl carbon atom. This new technique was used to study the reaction of ketones, nitriles, and Tf₂O, affording relevant information about new intermediates and kinetic data.

Full text of DOI:10.1021/ol902532f

ORGANIC
LETTERS
2010
Vol. 12, No. 1
144-147
2D Ultrafast HMBC: A Valuable Tool for
Monitoring Organic Reactions§
Antonio Herrera,*,† Encarnacio´n Ferna´ndez-Valle,‡ Eva M. Gutie´rrez,†
†
Roberto Mart´ınez-Alvarez, Dolores Molero, Zulay D. Pardo, and Elena Sa´ez
‡
†
‡
´
Departamento de Qu´ımica Orga´nica and CAI de RMN Facultad de Qu´ımicas,
UniVersidad Complutense de Madrid, 28040 Madrid, Spain
aherrera@quim.ucm.es
Received November 4, 2009
ABSTRACT
Ultrafast 2D HMBC spectroscopy permits real-time monitoring of a reaction based on the structural changes produced in a carbonyl carbon
atom. This new technique was used to study the reaction of ketones, nitriles, and Tf2O, affording relevant information about new intermediates
and kinetic data.
Traditional nD NMR experiments are collected as an array
of 1D scans and are intrinsically time-consuming.1 Different
proposals to speed up nD NMR spectroscopy have been
described in recent years.2 Among them, ultrafast-NMR
spectroscopy (UF-NMR) alone or coupled with other meth-
odological elements into a single scheme has shown its
capability of delivering any type of nD spectrum in a single
scan by spatially encoding the indirect-domain interactions.3
Real-time monitoring of dynamic chemical systems, i.e.,
organic reactions, can be done as they happen.4
Recently we have studied the synthesis of alkylpyrimidines
from aliphatic ketones by two-dimensional UF-TOCSY,
which has permitted the collection of a complete multidi-
mensional data set within a single continuous acquisition.
The evolution of the reactants from early stages of the
reaction, the presence of different nitrilium-type intermedi-
ates, and the generation of reaction products can be moni-
torized.5 Despite these important findings, no direct
information regarding structural changes in the carbonyl
carbon atom was obtained. The reason is clear. Since it
is a quaternary carbon, its evolution cannot be detected
using 2D homonuclear sequences such as COSY or
TOCSY or heteronuclear such as HSQC. On the other
hand, a UF-HMBC pulse sequence can offer direct and
more precise information about the structural changes that
take place on the carbon core and protons attached at
neighbor positions. Chemical shifts obtained from 2D
HMBC 2J/3J correlations should permit real-time observa-
tion of the changes produced. This information can be
§ Dedicated to Prof. Antonio García Martínez on the occasion of his
retirement.
† Departamento de Qu´ımica Orga´nica.
‡ CAI de RMN.
(1) (a) Jeener, J. Lecture at International Ampe`re Summer School, Basko
Polje, Yugoslavia, 1971. (b) Aue, W. P.; Bartholdi, E.; Ernst, R. R. J. Chem.
Phys. 1976, 64, 2229–2246.
(2) (a) Frydman, L. C. R. Chimia 2006, 9, 336. (b) Schanda, P.;
Brutscher, B. J. Am. Chem. Soc. 2005, 127, 8014. (c) Kupcˇe, E.; Freeman,
R. J. Biomol. NMR 2003, 25, 349.
(3) (a) Mishkovsky, M.; Frydman, L. Annu. ReV. Phys. Chem. 2009,
60, 429. (b) Tal, A.; Shapira, B.; Frydman, L. Angew. Chem., Int. Ed. 2009,
48, 2732.
´
(4) Gal, M.; Mishkovsky, M.; Frydman, L. J. Am. Chem. Soc. 2006,
128, 951.
(5) Herrera, A.; Ferna´ndez-Valle, E.; Mart´ınez-Alvarez, R.; Molero, D.;
Pardo, Z. D.; Sa´ez, E. Angew. Chem., Int. Ed. 2009, 48, 6274.
10.1021/ol902532f  2010 American Chemical Society
Published on Web 11/30/2009

very important since the carbonyl carbon atom plays a
decisive role in numerous organic reactions.
Figure 2 shows the scheme of the 2D UF-HMBC sequence
used. This sequence consists basically of a continuous spatial
encoding ultrafast HSQC sequence in which the delay d is set
to 25 ms in order to monitor 2J and 3J couplings of 10 Hz.7
In this UF scheme, contrary to traditional approaches, the
spectral widths in both indirect and direct dimensions are
dependent on each other. The greater the spectral width that
must be covered in both dimensions, the poorer the digital
resolution in the indirect domain.8 Furthermore, the spectral
width and therefore the resolution in the indirect domain
depends on the strength Ga and duration Ta of the decoding
gradient, and both of these are limited. Because a wide
spectral range in the indirect domain needed to be covered
to detect the possible intermediates, we decided to monitor
two different spectral windows in this dimension. For this
purpose we alternated the same sequence with different
indirect dimension frequency offsets. The acquisition of these
two data sets was repeated during the progress of the reaction.
A total of 219 UF-HMBC spectra for each spectral window
were recorded over a total time of 112 min. Each single
spectrum was measured for 5.37 s, and a measurement was
repeated every 10 s. For acquisition conditions see Support-
ing Information.
Although the chemistry of triflic anhydrides and their
reactions with carbonyl compounds is well-known, some
mechanistic questions still remain open.6 Thus, it has been
accepted for many years that the reaction consists of the
electrophilic attack of the anhydride on the carbonyl oxygen;
this results in the formation of the short-lived (trifluoromethane-
sulfonyl)carbenium ion (2) (Figure 1). Despite this no data about
Figure 1. Reaction possibilities from ketones and triflic anhydride.
the structure and kinetic behavior of the crucial intermediate 2
are yet known. Depending on the nature of the carbonyl
compound and the reaction conditions, there are different
possibilities for 2; it can lead directly to vinyl triflates or to
different products after trapping by an external nucleophile.
In this paper we studied the organic reaction described above
and monitored how the carbonyl carbon atom evolves during
the process. Our aim is to identify possible intermediates
formed. We chose a simple aliphatic-aromatic ketone, acetophe-
none, as the model compound and studied its reaction with triflic
anhydride in the presence of an excess of acetonitrile acting as
nuclephile and the solvent. This leads to pyrimidines as main
products. Labeled 13C-carbonyl-acetophenone and [D3]acetoni-
trile were used as reactants. 13C-Carbonyl-labeled ketone ensures
the detection of short life intermediates and overcomes sensitiv-
ity problems. Attempts done with unlabeled ketones failed. The
use of [D3]acetonitrile eliminates the neccessity of solvent
suppression. The reaction was carried out inside a standard 500
MHz spectrometer with conventional hardware and a 5 mm NMR
tube. The reactants were added with a fast mixing device. For
experimental details and procedure see Supporting Information.
Figure 3 shows a series of six representative 2D UF-
HMBC spectra recorded (spectra 1-6). These spectra
correspond to the first window studied. The ranges examined
were 210-150 ppm for 13C, which covers the carbonyl
1
region and 9-2 ppm for H. Spectra were numbered 1-6
(referred to as HMBC-1, etc). Cross-peaks assigned to
starting and final products were confirmed using standard
1D and 2D spectra (see Supporting Information).
Spectrum HMBC-1 (0.0 min) corresponds to the solution of
ketone 1 in [D3]acetonitrile before the addition of Tf2O.
HMBC-1 shows a cross-peak at 198.9-2.56 ppm due to the 2J
coupling between the carbonyl carbon and the methyl protons
of 1 (red arrow). The addition of Tf2O produced a small change
in the position of the cross-peak from 1. In HMBC-2, taken
1.02 min after the addition, a new cross-peak appeared at
195.2-3.26 ppm (green arrow). This peak intensified in
HMBC-3 (1.53 min), decreased clearly in HMBC-5 (33.2 min),
and remained close to the detection limit in HMBC-6 (76.4
min). Simultaneously, the cross-peak (167.4-8.38, blue arrow)
that belongs to the pyrimidine 5 increased its intensity (HMBC-3
to HMBC-6). The signal from 1 decreased (HMBC-1 to
HMBC-5) and is absent in HMBC-6.
In our opinion, this new cross-peak (green arrow) depicted
in Figure 3 belongs to the complex 2 formed from ketone 1
and triflic anhydride. The rise and fall of 2 during the reaction
clearly shows its character as intermediate. Structures of 1 and
2 are similar and therefore produce cross-peaks close to each
(6) (a) Garc´ıa Mart´ınez, A.; Herrera Ferna´ndez, A.; Moreno Jime´nez,
F.; Garc´ıa Fraile, A.; Subramanian, L. R.; Hanack, M. J. Org. Chem. 1992,
´
57, 1627. (b) Herrera, A.; Mart´ınez-Alvarez, R.; Ramiro, P.; Molero, D.;
´
Almy, J. J. Org. Chem. 2006, 71, 3026. (c) Herrera, A.; Mart´ınez-Alvarez,
R.; Chioua, M.; Chioua, R.; Sa´nchez, A. Tetrahedron 2002, 58, 10053. (d)
Baraznenok, I. L.; Nenajdenko, V. G.; Balenkova, E. S. Tetrahedron 2000,
56, 3077.
Figure 2. Ultrafast two-dimensional HMBC implemented to
monitor changes on the carbonyl carbon atom.
(7) Shrot, Y.; Shapira, B.; Frydman, L. J. Magn. Reson. 2004, 171, 162.
(8) Frydman, L.; Lupulescu, A.; Scherf, T. J. Am. Chem. Soc. 2003,
125, 9204.
Org. Lett., Vol. 12, No. 1, 2010
145

Figure 4 again shows a series of six representative 2D UF-
HMBC spectra recorded corresponding to the second window
Figure 3. A series of 6 UF-HMBC experiments (1-6) taken at
different times throughout the reaction (scale is in ppm) and shifts
from the participanting reactant 1, product 5, and intermediate 2.
Figure 4. A series of 6 UF-HMBC experiments (7-12) obtained
at increasing times along the reaction (scale is in ppm) and shifts
from product 5 and participanting nitrilium intermediates 3/4.
other. The structure of benzylic intermediate 2 agrees with its
stability (relative to 1) and also with the observed small shielding
of the carbonyl carbon9 together with a displacement to higher
frequencies of the neighbor methyl protons. Benzyl cations like
2 have been shown to have relatively long lifetimes since the
phenyl ring stabilizes the positive charge formed on the carbonyl
carbon. They have been generated thermally and trapped in
organic solvents such as acetonitrile.10 The calculated chemical
shifts (ACD/Laboratories 8.00 Release) are in good agreement
with this new cross-peak.
studied (spectra 7-12). The ranges examined now are 180-120
ppm for 13C, covering the olephinic-aromatic zone and 9-2
ppm for 1H. Spectra are numbered 7-12 (referred to as HMBC-
7, etc). Obviously no HMBC signals were produced by the
starting ketone 1 (HMBC-7, 0.0 min). After 1.28 min a new
cross-peak at 138.3-5.56 is observed in HMBC-8. We assign
this cross-peak (orange arrow) to the CdCH2 correlation from
a mixture of the olephinic nitrilium salt intermediates 3/4. Their
calculated chemical shifts (ACD/Laboratories 8.00 Release) are
in agreement with the observed new signal.
After 1.53 min, a new cross-peak is detected in HMBC-3
which belongs to the final pyrimidine 5 (167.4-8.38, blue
arrow). Its intensity increases in HMBC-4 to HMBC-6.
The intensity of intermediates 3/4 decreases (HMBC-9 to
HMBC-11) and finally disappears in HMBC-12. The inter-
mediate character of (3/4) is shown again with their rise and
fall during the reaction. The presence of pyrimidine 5
increases (HMBC-9 to HMBC-12). The presence of vinyl
triflates formed by elimination of triflic acid from intermedi-
ate 2 was not detected in the UF-HMBC spectra. However
(9) 13C NMR data for carbocations: (a) Breitmeier, E.; Voelter, W.
Carbon-13 NMR Spectroscopy, 3rd ed.; VCH: New York, 1987. (b) Stadler,
D.; Goeppert, A.; Rasul, G.; Olah, G. A.; Prakash, S.; Bach, T. J. Org.
Chem. 2009, 74, 312.
(10) (a) Darbeau, R. W.; Gibble, R. E.; Pease, R. S.; Bridges, D. E.;
Siso, L. M.; Heurtin, D. J. J. Chem. Soc., Perkin Trans. 2001, 2, 1084. (b)
White, E. H.; DePinto, J. T.; Polito, A. J.; Bauer, I.; Roswell, D. F. J. Am.
Chem. Soc. 1988, 110, 3708.
146
Org. Lett., Vol. 12, No. 1, 2010

this cannot be ruled out below the detection limit of the
sequence applied. The reaction was additionally monitored
by 1H NMR, and the formation of a minor amount of vinyl
triflates was observed. An animation of this process was
made using the 219 experiments of the double set of HMBC
spectra registered and the combined mechanistic information
(see Supporting Information).
(2 and 3/4) were obtained by fitting the data points to the
equation I(τ) ) I0 exp(-τ/t) + I∞.
Values shown in Figure 5 gave reaction rate measurements,
which permits an evaluation of the reactive behavior of
intermediates.
In conclusion, a two-dimensional UF-HMBC experiment
is able to monitor an organic reaction and afford important
structural and mechanistic information about the chemical
environment and evolution of a reactive quaternary center.
This technique allows the detection of different intermediate
species. In the present case it has permitted the detection of
an intermediate (trifluoromethanesulfonyl)carbenium ion 2
and has confirmed the presence of nitrilium salt intermediates
3/4. Work is in progress to extend this methodology to studies
on unlabeled carbonyl compounds and to apply the ultrafast
schemes to other dynamic systems.
The variation in intensity of participant species detected
along the reaction studied is clearly shown in Figure 5. Here
Acknowledgment. We acknowledge a research grant from
MICINN (CTQ2007-61973) and Prof. Lucio Frydman (De-
partment of Chemical Physics, Weizmann Institute of Sci-
ence, 76100 Rehovot, Israel) for his scientific support. We
would like to thank Prof. John Almy (Faculte´ d’Enologie,
Universite´ de Bordeaux II, 33405 Talence cedex, France)
for his helpful discussions.
Figure 5. Averaged integrated peak intensity as a function of time
for starting ketone (1), intermediates (2 and 3/4), and final
pyrimidine (5).
Supporting Information Available: Experimental pro-
cedures and spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
are displayed the integrals of the cross-peaks versus time.
The t lifetimes of starting ketone (1) and intermediates
OL902532F
Org. Lett., Vol. 12, No. 1, 2010
147