Steric control of geminal Lewis pair behavior: Frustration induced dyotropic rearrangement

Article abstract of DOI:10.1055/s-0033-1339125

A series of methylene-linked boron/nitrogen geminal Lewis pairs were synthesized and the impacts of sterical effect on their chemical behavior were systematically investigated. Increasing the steric demand around the boron atom is manifested first by an incremental change in the structure of the resulting dative adducts. Accordingly, in the case of phenyl substituents (Alk ₂NCH₂BPh₂), formation of head-to-tail dimers/oligomers was observed, while such an intermolecular association was avoided when o-tolyl moiety was introduced [Alk₂NCH ₂B(o-Tol)₂], affording only an intramolecular dative complex. Further increase of sterical hindrance to a point (i.e. using mesityl substituents), however, caused a radical change in the structure; a dyotropic rearrangement occurred. Thus, steric interference induced a rearrangement in the geminal pair to relieve or minimize the frustration strain. Georg Thieme Verlag Stuttgart New York.

Full text of DOI:10.1055/s-0033-1339125

CLUSTER
▌1525
cluster
Steric Control of Geminal Lewis Pair Behavior: Frustration Induced Dyotro-
pic Rearrangement
Frustration Induced Dyotropic Rearrangement
Éva Dorkó,a Eszter Varga,a Tamás Gáti,b Tamás Holczbauer,a Imre Pápai,a Hasan Mehdi,a Tibor Soós*a
a
Hungarian Academy of Sciences, Research Centre for Natural Sciences, Institute of Organic Chemistry, Magyar tudósok körútja 2,
1117 Budapest, Hungary
E-mail: soos.tibor@ttk.mta.hu
b
Servier Research Institute of Medicinal Chemistry, Záhony Street 7, 1031 Budapest, Hungary
Received: 18.03.2014; Accepted after revision: 22.04.2014
geminal aminoborane FLPs such as 2 (Scheme 1) should
Abstract: A series of methylene-linked boron/nitrogen geminal
act as a superior hydrogenation catalyst.
Lewis pairs were synthesized and the impacts of sterical effect on
their chemical behavior were systematically investigated. Increas-
ing the steric demand around the boron atom is manifested first by
t-Bu
TMS
AlR2
H
Ph
an incremental change in the structure of the resulting dative ad-
ducts. Accordingly, in the case of phenyl substituents
(Alk2NCH2BPh2), formation of head-to-tail dimers/oligomers was
observed, while such an intermolecular association was avoided
when o-tolyl moiety was introduced [Alk2NCH2B(o-Tol)2], afford-
ing only an intramolecular dative complex. Further increase of ste-
rical hindrance to a point (i.e. using mesityl substituents), however,
caused a radical change in the structure; a dyotropic rearrangement
occurred. Thus, steric interference induced a rearrangement in the
geminal pair to relieve or minimize the frustration strain.
N
H
R2N
N
Al
N
Ph
B(C6F5)2
Ph
t-Bu
Al
N
H
Ph
N
Ph
P
Cp2Zr
PPh2
Ph
P
B(C6F5)2
Ph
Ph
t-Bu
t-Bu
Trip
Sn
(C6F5)2P
(C6F5)2P
B(C6F5)2
B(C6F5)2
Al
t-Bu
t-Bu
Al
Key words: Lewis acids, boron, steric hindrance, dyotropic rear-
rangement, frustrated Lewis pairs
Trip
P
t-Bu2P
BR2
Trip = 2,4,6-i-Pr3C6H2
Figure 1 Representative examples of recently introduced intramo-
lecular FLP systems
The activation of small molecules and unreactive bonds is
perhaps one of the most important areas of modern chem-
istry. The concept of frustrated Lewis pairs (FLP), intro-
duced by Stephan and co-workers, holds a promise as
unique and complementary bond activation approach that
has generated an immense scientific interest.1 A frustrated
Lewis pair is composed of bulky Lewis acid (e.g. borane)
and Lewis base (e.g. amine or phosphine) that are unable
to form a stable dative bond because of steric hindrance.
This incapability of dative bond formation is the key ele-
ment to a series of unprecedented chemical transforma-
tions among which the heterolytic hydrogen activation
overtops.2
Alk
Ar
H
N
B
N
B
Alk
Ar
1
2
Scheme 1 Envisioned geminal N/B pairs for H2 activation
Since the envisioned geminal Lewis pairs were expected
to trap ethereal solvents, the geminal N/B Lewis pairs
were prepared through the metathesis of the R2NCH2SnR3
(Scheme 2), in analogy to a reported procedure.3j As out-
lined in Scheme 2, the 5a and 5b precursors were easily
formed in the alkylation of pyrrolidine or piperidine with
4, respectively.5 The trimethylstannyl groups were then
replaced with lithium in hexanes and the resulting lithium
salts were isolated (6a,b). Reaction of these lithium salts
with equimolar quantities of diarylboron halides (11–14)
bearing different steric demand around boron atom result-
ed in the formation of new N/B pairs. In case of diphenyl-
To increase further the cooperativity of the catalytic sites
in FLPs, several intramolecular systems have been devel-
oped using appropriate covalent linkers3 (see Figure 1).
Among the possible alternatives, the geminal P/Al3i and
P/B3f,g,j Lewis pairs have attracted considerable interests
due to their remarkable activity on small molecules acti-
vation. Therefore, our efforts have been devoted to pre-
pare boron- and nitrogen-containing (B/N) geminal
systems 1 to test the effect of sterical hindrance on their
chemical behavior. This study was also triggered by a re-
cent theoretical work of Wang group4 concluding that
1
boranes 7a,b, the H NMR spectrum of the isolated
product showed two groups of broad signals. We reasoned
that this feature of the solution-phase spectra might be due
to a complex and dynamic mixture of monomers, dimers
and oligomers. Finally, we could grow suitable crystals
SYNLETT 2014, 25, 1525–1528
Advanced online publication: 06.06.2014
DOI: 10.1055/s-0033-1339125; Art ID: st-2014-b0238-c
© Georg Thieme Verlag Stuttgart · New York
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6

1526
É. Dorkó et al.
CLUSTER
for single-crystal X-ray diffraction analysis. The crystal from the boron atom to the adjacent carbon, while simul-
structure determination verified the formation of a ‘head- taneously one hydrogen atom shifted from the carbon to
to-tail’ dimeric form of the desired product having inter- the boron atom (Scheme 2). The same rearrangement was
molecular boron–nitrogen dative bonds (Scheme 2). also observed when the less electron-rich 3-chloromesityl
Thus, a six-membered B2C2N2 ring was formed that ad- groups were attached to the boron atom affording product
opted a chair conformation despite the axial position of 10a and 10b.
phenyl rings. Although we could detect the monomers of
7a and 7b in HRMS, these dative complexes proved to be
rather stable in solution; they showed no capability of
hydrogen activation.
These experimental findings indicate that the geminal
frustrated N/B pairs are prone to rearrange to minimize
the frustration strain, a capacity that was not yet observed
in analogous geminal P/B and P/Al FLP pairs. During this
In order to avoid the self-quenching of geminal pairs, we process, the Lewis pairs gained stabilization, thus the en-
increased the steric congestion around the boron atom. By tatic state, the FLP reactivity, was lost.
reacting 6a and 6b with bromobis(2-methylphenyl)borane
These rearranged products have some interesting features
in the solid state that deserves a comment. DFT calcula-
tions carried out for the 9a complex predict notably differ-
ent N–C and N–B bond lengths (1.52 and 1.63 Å) in the
(12) we isolated 8a and 8b N/B pairs, respectively
(Scheme 2). The 11B NMR investigation confirmed an in-
tramolecular B–N dative bond, thus the steric shielding of
the tolyl groups shifted the position of the equilibrium of
strained three-membered ring. On the other hand, the X-
the dative complexes. Nevertheless, the dative bonds in
ray analysis revealed rather symmetrical structures having
these N/B pairs, despite their strained nature, turned out to
formally the same N–C and N–B distances (1.58 Å). The
be strong enough to make these species inactive towards
reason for this discrepancy is the inverse placement in the
asymmetric unit of 9a (and also in 10a and 10b). In other
words, the molecules are randomly placed in the crystal
structures concerning the pseudo symmetric molecular ar-
hydrogen molecule. The relative stabilities of various
forms of the synthesized geminal aminoboranes with pyr-
rolidine ring (7a,8a) were estimated via DFT calculations6
(see Supporting Information for details). Those results
were in accord with the observed structures.
C1 positions (see Figure 2 and SI).
rangement with 50%-50% site occupation of the B1 and
(i)
(iii)
(iv)
N
Sn
I
Sn
I
I
(ii)
n
3
4
5a (n = 1)
5b (n = 2)
N
Li
(vi)
N
B
n
6a (n = 1)
6b (n = 2)
n
(v)
R
(vii)
8a (n = 1)
8b (n = 2)
Ph
n
H
R
N
B
Ph
N
B
Ph
n
B
N
+ monomers
+ oligomers
Ph
9a (n = 1, R = H)
9b (n = 2, R = H)
10a (n = 1, R = Cl)
10b (n = 2, R = Cl)
n
7a (n = 1)
7b (n = 2)
Scheme 2 Synthesis of geminal N/B Lewis pairs. Reagents and con-
ditions: (i) Cu/Zn, THF; (ii) Me3SnCl, THF; (iii) n = 1 (pyrrolidine),
n = 2 (piperidine), 3 d, r.t.; (iv) BuLi, hexane, 0 °C, 1 h; (v) Ph2BBr
(11), hexane, –78 °C, r.t., 12 h; (vi) (o-Tol)2BBr (12), hexane, –78 °C,
r.t., 12 h; (vii) Mes2BF (13, to form 9) or ClMes2BF (14, to form 10),
hexane, –78 °C, r.t., 12 h.
The attempt to synthesize a frustrated geminal N/B pair
led us to increase further the steric hindrance around the
boron atom. We then performed the analogous salt me-
tathesis reaction of 6a,b with the fluorodimesitylborane
(13). These reactions, however, gave unexpected datively
bound structures (9a,b) that were unambiguously con-
firmed by NMR and X-ray analysis. Accordingly, in a
dyotropic reaction,7 one of the mesityl groups moved
Figure 2 A. ORTEP plot of 9a molecule. Half of the molecule can be
found in the asymmetric unit. The other part of the molecule and the
common N atom are represented by orange color. The B1 and C1 at-
oms have symmetrical location with 50% occupation. B. Overlay
structures of 9a, determined by DFT calculations (blue) and by X-ray
analyses (red).
Synlett 2014, 25, 1525–1528
© Georg Thieme Verlag Stuttgart · New York

CLUSTER
Frustration Induced Dyotropic Rearrangement
1527
Subsequently, in a crossover experiment, we wished to uct (III′ in Figure 3). The free energy balance of the over-
find an answer whether the rearrangement occurs in an in- all I → III′ transformation is predicted to be close to zero,
tramolecular or an intermolecular fashion. Therefore, we which may imply that the intramolecular rearrangement
reacted one equivalent of 6a with a half equivalent of flu- might be reversible.10 Nevertheless, we did not observe
orodimesitylborane (13) and half equivalent of fluoro- such an equilibrium in the NMR experiments.
di(3′-chloromesityl)borane (14). The HRMS analysis of
the crude reaction mixture indicated that the rearrange-
ment is unequivocally intramolecular, as we could not de-
tect any crossover products.
The mechanism of the observed dyotropic rearrangement
which affords 9a was also examined computationally.8
Figure 4 Transition states of H and Mes transfer steps in the dyo-
tropic rearrangement which affords aminoborane 9a. Relative Gibbs
free energies (relative to complex I; in kcal/mol) are given in paren-
thesis.
In summary, in this study the chemical behavior of gemi-
nal Lewis pairs has been extended to easily accessible
N/B pairs. We carried out a systematic study on the im-
pact of the steric effect on chemical behavior. The inves-
tigation indicated that while phenyl derivatives (7a,b)
associated through intermolecular B–N bond, the o-tolyl
derivatives (8a,b) afforded an intramolecular B–N dative
complex. More interestingly, when mesityl groups were
attached to the boron precursor (9a,b) a dyotropic rear-
rangement took place to gain stabilization in the geminal
pair. Thus, the steric congestion was released by the ex-
change of the mesityl group with the neighboring H atom.
We examined this intramolecular rearrangement by DFT
calculation, which suggested a quasi-stepwise mecha-
nism, which is at the borderline with a concerted/asyn-
chronous mechanism. These findings demonstrate the
inherent capacity of geminal FLPs to promote dyotropic
rearrangement and draw also the attention to the impor-
tance of the careful structural design of bifunctional frus-
trated Lewis pairs if one’s goal is to develop efficient
hydrogenation catalysts.
Figure 3 Gibbs free energy diagram computed for the dyotropic re-
arrangement which affords aminoborane 9a (denoted as III′). Rela-
tive Gibbs free energies of intermediates and transition states are
given in parenthesis (in kcal/mol). All structures are reported in the
SI.
Possible reaction pathways towards the rearranged prod-
uct were explored by locating transition states for envi-
sioned H and Mes migration steps along with the
corresponding reaction intermediates. As illustrated in
Figure 3, the calculations revealed that the rearrangement
of the putative, first formed aminoborane (I) is initiated by
hydride transfer to the boron center facilitated by the N-
lone pair, yielding an iminium-hydroborate intermediate
(II), which lies 9.9 kcal/mol above I. The structure of the
transition state of this step is depicted in Figure 4 (TSI–II),
and the computed barrier towards II is fairly low. This
zwitterionic species can easily revert to the reactant state
(via barrier of 1.6 kcal/mol) implying that complex II
should be considered as a transient intermediate in this re-
action. Species II can also undergo B-to-C mesityl migra-
tion via transition state TSII–III (see Figure 4).
Acknowledgment
We are grateful for the financial support of the Lendület Program
E13/11/2010 and the Hungarian Scientific Research Fund (OTKA,
K-81927). We thank László Drahos for measuring the high-resolu-
tion mass spectra and Bianka Kótai for assistance in the initial phase
of the computational work. The X-ray diffractometer purchase grant
from the National Office for Research and Technology (MU-
00338/2003) is gratefully acknowledged.
This migration process represents the rate-limiting step of
the overall transformation. The computed overall activa-
tion barrier (22.5 kcal/mol with respect to I) is rather low
which is consistent with the fact that no formation of I was
detected. The aminoborane species formed upon Mes mi-
gration (III in Figure 3) lies fairly high in free energy (at
9.3 kcal/mol) pointing to a notable strain in this mole-
cule,9 however, the reorientation of the bulky Mes groups
gives rise to a thermodynamically more favored isomeric
form, which is actually identical to the observed 9a prod-
Supporting Information for this article is available online
at
http://www.thieme-connect.com/products/ejournals/journal/
10.1055/s-00000083.SunogIpimrfiantoSuIpg
n
fonirtat
ori
References and Notes
(1) Welch, G. C.; San Juan, R. R.; Masuda, J. D.; Stephan, D. W.
Science 2006, 314, 1124.
© Georg Thieme Verlag Stuttgart · New York
Synlett 2014, 25, 1525–1528

1528
É. Dorkó et al.
CLUSTER
(5) Cox, P. J.; Doidge-Harrison, S. M. S. V.; Howie, R. A.;
(2) For recent reviews, see: (a) Stephan, D. W.; Erker, G.
Angew. Chem. Int. Ed. 2010, 49, 46. (b) Topics in Current
Chemistry: Frustrated Lewis Pairs I; Stephan, D. W.; Erker,
G., Eds.; Springer: New York, 2013, Vol. 332. (c) Topics in
Current Chemistry: Frustrated Lewis Pairs II; Stephan, D.
W.; Erker, G., Eds.; Springer Press: New York, 2013, Vol.
334.
Nowell, I. W.; Taylor, O. J.; Wardell, J. L. J. Chem. Soc.,
Perkin Trans. 1 1989, 2017.
(6) In the present study, the DFT calculations were carried out
using the ωB97X-D functional. The energy values reported
in the paper correspond to solution-phase Gibbs free
energies that are based on ωB97X-D/6-311++G(3df,3pd)
electronic energies and all additional terms (thermal
contributions and solvent effects) obtained at the ωB97X-
D/6-311G(d,p) level. For further details, see Supporting
Information.
(7) For a review on dyotropic reactions, see: Fernández, I.;
Cossío, F. P.; Sierra, M. A. Chem. Rev. 2009, 109, 6687.
(8) For a selection of recent theoretical mechanistic studies of
dyotropic rearrangements, see: (a) Davis, R. L.; Leverett, C.
A.; Romo, D.; Tantillo, D. J. J. Org. Chem. 2011, 76, 7167.
(b) Leverett, C. A.; Purohit, V. C.; Johnson, A. G.; Davis, R.
L.; Tantillo, D. J.; Romo, D. J. Am. Chem. Soc. 2012, 134,
13348. (c) Fernández, I.; Bickelhaupt, F. M.; Cossío, F. P.
Chem. Eur. J. 2012, 18, 12395. (d) Gutierrez, O.; Tantillo,
D. J. J. Org. Chem. 2012, 77, 8845.
(3) (a) Theuergarten, E.; Schlüns, D.; Grunenberg, J.; Daniliuc,
C. G.; Jones, P. G.; Tamm, M. Chem. Commun. 2010, 46,
8561. (b) Chernichenko, K.; Nieger, M.; Leskelä, M.; Repo,
T. Dalton Trans. 2012, 41, 9029. (c) Holtrichter-Rößmann,
T.; Rösener, C.; Hellmann, J.; Uhl, W.; Würthwein, E.-U.;
Fröhlich, R.; Wibbeling, B. Organometallics 2012, 31,
3272. (d) Zheng, W.; Pi, C.; Wu, H. Organometallics 2012,
31, 4072. (e) Freitag, S.; Henning, J.; Schubert, H.;
Wesemann, L. Angew. Chem. Int. Ed. 2013, 52, 5640.
(f) Stute, A.; Kehr, G.; Fröhlich, R.; Erker, G. Chem.
Commun. 2011, 47, 4288. (g) Rosorius, C.; Kehr, G.;
Fröhlich, R.; Grimme, S.; Erker, G. Organometallics 2011,
30, 4211. (h) Xu, X.; Kehr, G.; Daniliuc, C. G.; Erker, G.
J. Am. Chem. Soc. 2013, 135, 6465. (i) Bertini, F.;
Hoffmann, F.; Appelt, C.; Uhl, W.; Ehlers, A. W.; Slootweg,
J. C.; Lammertsma, K. Organometallics 2013, 32, 6764.
(j) Bertini, F.; Lyaskovskyy, V.; Timmer, B. J. J.; de Kanter,
F. J. J.; Lutz, M.; Ehlers, A. W.; Slootweg, J. C.;
(9) Short HH bond distances measured in the equilibrium
structure of III are clear indications of the internal strain (see
Supporting Information).
(10) The computed free energy balance cannot be regarded as a
very accurate prediction due to the inaccuracy of the present
computational approach. The error bar on the relative Gibbs
free energies reported in the present work is expected to be
about 3–4 kcal/mol.
Lammertsma, K. J. Am. Chem. Soc. 2012, 134, 201.
(4) Zhao, L.; Lu, G.; Huang, F.; Wang, Z.-X. Dalton Trans.
2012, 41, 4674.
Synlett 2014, 25, 1525–1528
© Georg Thieme Verlag Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.