Chemistry in confined spaces: High-energy conformer of a piperidine derivative is favored within a water-soluble capsuleplex

Article abstract of DOI:10.1021/ol101841n

Propyloxy-substituted piperidine in solution adopts a conformation in which its alkoxy group is equatorially positioned. Surprisingly, two conformers of it that do not interconvert in the NMR time scale at room temperature have been found within an octa-acid capsule. The serendipitous finding of the axial conformer of propyloxy-substituted piperidine within a supramolecular capsule highlights the value of confined spaces in physical organic chemistry.

Full text of DOI:10.1021/ol101841n

ORGANIC
LETTERS
2010
Vol. 12, No. 20
4544-4547
Chemistry in Confined Spaces:
High-Energy Conformer of a Piperidine
Derivative is Favored Within a
Water-Soluble Capsuleplex
Mintu Porel,† Nithyanandhan Jayaraj,† S. Raghothama,‡ and V. Ramamurthy*,†
Department of Chemistry, UniVersity of Miami, Coral Gables, Florida 33124, United
States, and NMR Research Centre, Indian Institute of Science, Bangalore, 560 012, India
murthy1@miami.edu
Received August 6, 2010
ABSTRACT
Propyloxy-substituted piperidine in solution adopts a conformation in which its alkoxy group is equatorially positioned. Surprisingly, two
conformers of it that do not interconvert in the NMR time scale at room temperature have been found within an octa-acid capsule. The
serendipitous finding of the axial conformer of propyloxy-substituted piperidine within a supramolecular capsule highlights the value of confined
spaces in physical organic chemistry.
Confined spaces offer a unique opportunity to examine and
manipulate the properties of small molecules and reactive
intermediates.1,2 Pioneers Cram and co-workers established
in 1991 the storage of the highly reactive cyclobutadiene in
solution within a hemicarcerand at room temperature.3 Their
predicted possibility of synthesis and examination of a
number of reactive species within the inner phases of
appropriate hemicarcerands have been realized during the
last two decades; persistency when generated within a
hemicarcerand in solution of transient species such as
benzyne, cycloheptatetraene, enol, carbene, and nitrene has
been demonstrated.4 However, conformational isomers of
cyclohexanes trapped within the confined spaces of synthetic
cavitands at room temperature remain elusive.5,6 In this
context, it must be noted that within solid thiourea channels
the preferred conformation of a few monohalocyclohexanes
has the halogen axially positioned and differs from their
equatorial position in solution.7,8 This communication is
concerned with stabilization of a high-energy conformer of
a piperidine derivative within the confined space of a capsule
made of two octa-acid molecules (OA, Scheme 1), at room
temperature in aqueous solution.9-11 We have carried out
1H NMR experiments on capsuleplexes of piperidine deriva-
tives 2a-e (Scheme 1) within OA in aqueous solution.12
(4) Warmuth, R. In Molecular Encapsulation; Brinker, U. H., Mieusset,
J.-L., Eds.; John Wiley & Sons: Chichester, 2010; Chapter 9.
(5) O’Leary, B. M.; Grotzfeld, R. M.; Rebek, J., Jr. J. Am. Chem. Soc.
† University of Miami.
‡ Indian Institute of Science.
1997, 119, 11701
(6) Bao, X.; Reith, S.; Stojanovic, S.; Hadad, C. M.; Badjic, J. D. Angew.
Chem., Int. Ed. 2010, 49, 1
(7) Aliev, A. E.; Harris, K. D. M. J. Am. Chem. Soc. 1993, 115, 6369
(8) Muller, K. Magn. Reson. Chem. 1992, 30, 228
.
(1) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular
Photochemistry of Organic Molecules; University Science Books: Sausalito,
CA, 2010; Chapter 13.
.
.
(2) Molecular Encapsulation; Brinker, U. H., Mieusset, J.-L., Eds.; John
.
Wiley & Sons: Chichester, 2010
(3) Cram, D. J.; Tanner, M. E.; Thomas, R. Angew. Chem., Int. Ed.
Engl. 1991, 30, 1024.
.
(9) Yousif, G. A.; Roberts, J. D. J. Am. Chem. Soc. 1968, 90, 6428.
(10) Lambert, J. B.; Bailey, D. S.; Michel, B. F. J. Am. Chem. Soc.
1972, 94, 3812.
10.1021/ol101841n  2010 American Chemical Society
Published on Web 09/24/2010

capsuleplex of two OAs and a guest is symmetrical, only a
single signal each for all eight H-d and all eight H-f
hydrogens is expected. However, the host NMR signals of
capsuleplexes of 2b-e@OA2 reveal the presence of inde-
pendent signals for some of the identical OA hydrogens
present on the top and bottom halves of the capsule (e.g.,
signals due to He in 2b and signals due to Ha-f for 2c-e in
Figure S2, SI). This suggests that 2b-e do not tumble freely
within the capsule, which makes the two halves of the
capsule identical, in the NMR time scale at room temperature.
Although 2a, similar to 2b-e discussed above, formed a
1:2 capsuleplex (Figure S9, SI), it exhibited a distinctly
different behavior. Examination of 1H NMR spectra of OA,
2a, and 2a@OA2 displayed in Figure 1 reveals that for the
Scheme 1
.
Chemical Structures of Octa-Acid (1) and Piperidine
Derivatives 2a-e Examined in This Studya
a Two possible conformations of 2a with respect to the C-O-alkyl
substitution are also shown. These two conformations are color coded in
accordance with the NMR spectra displayed in Figures 2 and 3.
These spectra revealed that 2a exists in two distinctly
different noninterconverting conformations at room temper-
ature within the OA capsule, while 2b-e adopt the same
conformation both in the OA capsule and in solution.
First, we present the results on 2b-e that set the stage to
understanding the unique behavior of 2a. 1H NMR titration
experiments of 2b-e with OA established that they form
1:2 (guest:host) complexes in water at pH 8.5 (sodium
tetraborate buffer). Inclusion of the guest is revealed by the
significant upfield shift of the C-O-alkyl hydrogen signals,
especially the CH3 group of the alkyl chain (Figures S1 and
Figure 1.
Partial 1H NMR spectra (500 MHz, D2O) of (i) OA, (ii)
2a@OA2 ([OA] ) 1 mM; [2a] ) 0.5 mM; 10 mM sodium
tetraborate buffer), and (iii) 2a. Host resonances are labeled in letters
“a-f”, and guest resonances are labeled in numbers.
capsuleplex 2a@OA2 four signals for each of the H-d and
H-f of OA and two signals for each of the guest methyl
groups marked 2a, 2e, and CH3-7 are present. Note that for
2b-e@OA2 only two signals for each of H-d and H-f of
OA and one signal each for 2a, 2e, and the terminal alkyl
methyl group were present (Figures S1 and S2, SI). This
difference suggested that 2a forms two types of complexes
with OA. The presence of four distinct signals for H-d and
H-f of OA is consistent with the existence of “two indepen-
dent unsymmetrical complexes in solution” with one set of
signals (two H-d, two H-f, and one CH3-7) belonging to one
complex and the other set to the second complex.
One possibility for the above hypothesis of “two inde-
pendent unsymmetrical complexes” is that the guest 2a is
captured in two different conformations within the capsule
(Scheme 1). The 2D DQF COSY NMR spectrum of 2a in
D2O confirmed that this molecule exists in only one
conformation in solution similar to 2b-e (Figure S10, SI).
Given the existence of 2a in a single conformation in solution
(C-O-alkyl at equatorial position), the possibility of
trapping an alternate conformer with the C-O-alkyl group
at the axial position seemed exciting. The observed two
distinct signals suggested that there may indeed be two
conformers trapped within the OA capsule and that they do
not interconvert between the two conformers in the NMR
time scale at room temperature. To examine if the two
conformers interconvert at a longer time scale, we recorded
S2 in Supporting Information, [SI]).13 2D DQF COSY H
1
NMR correlation spectra of 2b-e@OA2 helped us identify
the conformation of guest molecules within the capsule
(Figures S3-S6 in SI). Strong cross peaks between diaxial
and geminal-hydrogens and weak or negligible cross peaks
between axial-equatorial hydrogens are expected in DQF-
COSY spectra.14,15 In Figures S3-S6 (SI), strong cross peaks
between geminal hydrogens at C-3 (marked as 3a and 3e)
and between vicinal hydrogens at C-4 (marked 4) and one
of the two hydrogens at C-3 (marked 3a) are observed. The
absence of a cross peak between C-4 and C-3e hydrogens is
noteworthy. These data suggest that the conformation
adopted by 2b-e within the OA capsule has the O-alkyl
group placed equatorially (Scheme 1). By examining Figures
S7 and S8 (SI), we came to the conclusion that in D2O both
2b and 2c adopt the same conformation as in a capsule.
The single set of H-d and H-f hydrogens in each of the
four symmetrical panels of the OA cavitand results in an
identical chemical shift of all these hydrogens. When the
(11) Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126, 11408.
(12) Kulasekharan, R.; Jayaraj, N.; Porel, M.; Choudhury, R.; Sundare-
san, A. K.; Parthasarathy, A.; Ottaviani, F.; Jockusch, S.; Turro, N. J.;
Ramamurthy, V. Langmuir 2010, 26, 6943.
(13) Jayaraj, N.; Zhao, Y.; Parthasarathy, A.; Porel, M.; Liu, R. S. H.;
Ramamurthy, V. Langmuir 2009, 25, 10575.
(14) Columbus, I.; Biali, S. E. J. Org. Chem. 1993, 58, 7029
.
(15) Keeler, J. Understanding NMR Spectroscopy; Wiley: New York,
2005
.
Org. Lett., Vol. 12, No. 20, 2010
4545

a 2D ROESY spectrum with 300 ms mixing time at room
temperature; partial spectra are displayed in Figure 2 and
gained from examination of the correlation between H-d
signals of the host and 2,2′ CH3 signals of the guest.
Examination of Figures S12 and S13 (SI) reveals that the
methyl groups at the 2-positions of both isomers interact with
the H-d signal of the host that is split into four, one set for
the axial and one for the equatorial isomer complex. This
suggests that the relative locations of the two isomers are
similar within the capsule.
Confirmation that the above two sets of signals to be
indeed due to two conformers of 2a whose C-O-alkyl
groups are positioned axially and equatorially, respectively,
1
came from DQF COSY H NMR spectrum of 2a@OA2
(Figure 3 and Figure S14 in SI). Cross peaks between vicinal
Figure 2. Partial 2D ROESY spectrum (500 MHz, mixing time )
300 ms) of 2a@OA2 ([OA] ) 5 mM; [2a] ) 2.5 mM in 50 mM
sodium tetraborate buffer). The color represents the proton of the
conformer of the same color in Scheme 1 (bottom) for 2a. In these
spectra, we are unable to unequivocally assign the four signals seen
for the four methyl groups at 2,2′ positions to the exact conformer.
The color codes shown for the methyl groups should be taken as
tentative assignments.
Figure S11 (SI), respectively. Since the diagonal and cross
peaks have different signs, the exchange peaks (EXSY) and
NOESY peaks are easily distinguished in the ROESY
spectrum (see boxes Y and X in Figure S11; SI).16-18
Importantly, the EXSY cross-peaks have the same sign as
the diagonal. Examination of Figure 2 showing the signals
due to methyl groups of the guest (marked CH3-7, CH3-7′,
CH3-2, and CH3-2′) as well as others (see boxes A, B, C, D,
and E) suggests that the corresponding signals exchange in
the time scale (300 ms) of the ROESY experiment. The two
sets of peaks have thus been identified based on 2D ROESY
cross peaks to be due to two species (possibly conformers)
that exchange in 300 ms and accordingly are color coded
green and red in Figure 2.
In Figure S12 (SI), a partial 2D ROESY spectrum
displaying the NOESY interaction between the host and guest
signals is displayed. From this, it is clear that H-g signals of
the two halves of the capsule interact with the CH3-7 and
OCH3 signals of both isomers of the guest suggesting that
the two ends of the two conformers are anchored at the two
tapered ends of the capsule. Very important information is
Figure 3. Partial 2D DQF COSY NMR spectrum (500 MHz,
mixing time ) 300 ms) of 2a@OA2 ([OA] ) 5 mM, [2a] ) 2.5
mM in 50 mM sodium tetraborate buffer). Chemical structures of
two conformers are represented in two different colors. The color
of the assigned protons in the 2D DQF COSY NMR spectrum
represents the proton of the conformer with that particular color.
hydrogens H-4 and H-3 (axial) and H-3 (equatorial) of the
guest in the DQF COSY NMR spectrum provided the most
important information. In the green set of signals, H-4
correlates strongly with one of the two H-3 hydrogens
(marked 3a) and poorly with the other H-3 (marked 3e). Such
a correlation is consistent with the hypothesis that the green
and red signals correspond to the guest with the C-O-alkyl
group positioned equatorially and axially, respectively. It is
important to note that according to the DQF COSY 1H NMR
spectrum of 2a in D2O only one conformer with the
C-O-alkyl group at the equatorial position is present in
solution (Figure S10 in SI).
To examine the possibility of interconversion between the
1
two 2a@OA2 complexes, we recorded H NMR spectra at
different temperatures. As illustrated in Figure 4 (and Figure
S15 in SI) at 70 °C, the four signals due to H-d and H-f of
host OA coalesced into two, while the guest methyl signals
remained unaffected. The change in only the host signals
(16) Yamanaka, M.; Shivanyuk, A.; Rebek, J. J. J. Am. Chem. Soc. 2004,
126, 2939
.
(17) Basilio, N.; Rio, L. G.; Pastor, M. M. J. Phys. Chem. B 2010, 114,
4816
.
(18) Perrin, C. L.; Dwyer, T. J. Chem. ReV. 1990, 90, 935
.
4546
Org. Lett., Vol. 12, No. 20, 2010

Introduction of the integration values in the EXSY CALC
program20 provided the magnetization exchange rate con-
stants that are related to the conformer exchange rate
constants k1 and k-1 (for definition see Scheme 1). The
activation parameters (∆Hq and ∆Sq) calculated from the
Eyring plot shown in Figure S19 (SI) for the forward process
(k1; Scheme 1) were 17.7 kcal mol-1 and 1.75 eu, respec-
tively, and for the reverse process (k-1; Scheme 1) were 18
kcal mol-1 and 2.52 eu, respectively. On the basis of
1
integration of the H NMR signals due to conformers, we
conclude that the two isomers are present at a ratio of 53:
47, indicative of a significant lowering of the energy of the
axial conformer within the capsule. No signals due to free
1
Figure 4. Partial H NMR spectra (400 MHz, D2O) of 2a@OA2
([OA] ) 1 mM, [2a] ) 0.5 mM, in 10 mM sodium tetraborate
buffer) at (i) 25 °C and (ii) 70 °C.
1
host and guest molecules were present in H NMR spectra,
both at 25 and 70 °C, suggesting that in the NMR time scale
the complex is stable (Figures S15 and S20, SI). On the basis
of this, we conclude that the kinetic parameters calculated
above are not complicated by dissociation of the complex.
In this study, we have demonstrated the possibility of
trapping a high-energy conformer of a piperidine derivative
within a supramolecular assembly in aqueous solution. It is
important to note that although only one of the five molecules
we investigated exhibited the unusual behavior it was neither
the smallest nor the largest in the series. Successful trapping
of this molecule, albeit not an ideal or a general example, in
a high-energy conformation within a capsule is suggestive
of the possibility of yet to be discovered examples.
suggested that the capsule was becoming symmetrical at
higher temperatures.
We interpret the absence of influence of temperature on
the guest signals to indicate that at 70 °C it tumbles freely
within the capsule without any interconversion between the
two chair forms. To gain support to this thought, we recorded
1H NMR spectra at various temperatures for 2b and 2c (smaller
and larger molecules, respectively, than 2a; i.e., R ) CH2CH3
and CH2CH2CH2CH3). It is clear that coalescence of signals
for 2b occurred between 45 and 55 °C (Figure S16 in SI), while
it was absent for 2c even at 70 °C (notice the H-a, H-c, and
H-f peaks in Figure S17, SI). From these observations, we
conclude that the rate constants for the tumbling motion of the
guest that would make the top and bottom halves of the capsule
identical vary with the size of the guest molecule (2b@OA2 >
2a@OA2 > 2c@OA2).
Although there was no change in the guest signals with
respect to temperature, 2D ROESY experiments carried out
at 300 ms and zero mixing times at various temperatures
with 2a@OA2 enabled us to estimate the activation param-
eters for chair-chair interconversion of 2a.19 In these
experiments, the areas under cross and diagonal peaks (Figure
S18 (SI)) were integrated for a particular set of protons that
exchanged. Examination of the data provided in Figure S18
(SI) reveals, as expected, the absence of cross peaks at zero
mixing time suggesting the absence of magnetization transfer.
Acknowledgment. V.R. is grateful to the National Science
Foundation, USA. for generous financial support (CHE-
0848017). The authors thank Prof. R. S. H. Liu (University
of Hawaii) and Dr. Yaopeng Zhao (University of Miami)
for useful discussions.
Supporting Information Available: Experimental pro-
cedure and additional 1H NMR, 2D DQF COSY NMR, 2D
ROESY, variable-temperature 1H NMR spectra, Eyring plot,
and 1H NMR of compounds 2a-e as mentioned in the text.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL101841N
(19) Abel, E. W.; Coston, T. P. J.; Orrell, K. G.; Sik, V.; Stephenson,
(20) EXSY CALC, Mestrelab Research Appliction Note, www.mestrelab.com.
D. J. Magn. Reson. 1986, 70, 34.
Org. Lett., Vol. 12, No. 20, 2010
4547