Divergent Reactivity in Pd-Catalyzed [3,3]-Sigmatropic Rearrangement of Allyloxy- and Propargyloxyindoles Revealed by Computation and Experiment

Article abstract of DOI:10.1021/acs.orglett.8b02864

Detailed computational (DFT) studies of the palladium(II)-catalyzed Claisen rearrangement of 2-allyloxy- and propargyloxyindoles revealed an unexpected divergent mode of reactivity. Subsequent experimental kinetic isotope effects are in accord with the me

Full text of DOI:10.1021/acs.orglett.8b02864

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Divergent Reactivity in Pd-Catalyzed [3,3]-Sigmatropic
Rearrangement of Allyloxy- and Propargyloxyindoles Revealed by
Computation and Experiment
Osvaldo Gutierrez,*,†,‡ Charles E. Hendrick,† and Marisa C. Kozlowski*,†
†Department of Chemistry, Roy and Diana Vagelos Laboratories, University of Pennsylvania, Philadelphia, Pennsylvania 19104,
United States
S
* Supporting Information
ABSTRACT: Detailed computational (DFT) studies of the palladium(II)-catalyzed Claisen rearrangement of 2-allyloxy- and
propargyloxyindoles revealed an unexpected divergent mode of reactivity. Subsequent experimental kinetic isotope effects are in
accord with the mechanism derived from the computations. The computational results led to the development of Pd(II)-
catalyzed [3,3]-sigmatropic rearrangement of 3-aryl substituted 2-propargylindoles.
Scheme 1. (A) Representative Natural Products Bearing the
Oxyindole Backbone with a C3-Quaternary Stereocenter;
(B) Enantioselective Palladium(II)-Catalyzed [3,3]-
Sigmatropic Rearrangement of 2-Allyloxyindoles (Top) and
2-Propargyloxyindoles (Bottom)
igmatropic shifts constitute an essential class of pericyclic
1
Sreactions in organic chemistry.
Initially discovered by
Cope and Hardy in 1940,2 the thermal [3,3]-sigmatropic
rearrangement of 1,5-dienes (e.g., Cope rearrangement) is the
prototypical reaction of this class taught in undergraduate
organic chemistry courses to showcase concerted (and
stereospecific) σ- and π-bond reorganization.3 Seminal work
by Overman showed that palladium(II) chloride salts could
4
́
catalyze the Cope rearrangement. In 2012, Gagne and co-
workers reported the first enantioselective variant using a chiral
Au(I) catalyst.5 More recently, Gleason and Kaldre reported
the first organocatalytic Cope rearrangement using a novel
chiral N-acyl diazepane as the catalyst.6 In contrast to the Cope
rearrangement, the [3,3]-sigmatropic rearrangement of allyl
vinyl ethers (Claisen rearrangement) has been widely used in
asymmetric catalysis and in the synthesis of complex natural
products.7,8 Our group became interested in utilizing a novel
[3,3]-sigmatropic shift to access the C3-quarternary carbon
oxyindole backbone found in many natural products (Scheme
1A). We discovered that axial chiral palladium(II) salts are able
to catalyze the asymmetric [3,3]-sigmatropic rearrangement of
3-ester-substituted 2-allyloxy- and propargyloxyindoles with
high yields and enantioselectivities (Scheme 1B).9,10 Herein,
we use a combined computational and experimental approach
to elucidate the mechanism of these transformations.
Surprisingly, DFT calculations and labeling experiments
support a divergent mode of reactivity. Implications for
rational reaction design are shown by engineering a new
Pd(II)-catalyzed [3,3]-sigmatropic rearrangement of a system
lacking the previously necessary chelating 3-ester group to
generate 3-aryl-2-propargyloxyindoles.
allyloxyindole with Pd(PH3)2Cl+ as a model catalyst (Figure
1).11,12 As shown in Figure 1, palladium(II) complexation to 3-
methylester-2-allyloxyindole can occur via Lewis acid type
coordination to the carbonyl and ether (1A) or via π-
coordination to the alkene moiety (1A0).13 DFT calculations14
show a strong preference (8.6 kcal/mol) for Lewis acid
coordination, which leads to a concerted [3,3]-sigmatropic
We initiated our mechanistic studies by modeling the
Received: September 7, 2018
palladium(II)-catalyzed rearrangement of 3-methylester-2-
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b02864
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
coordination (2A0) that activates the alkyne for nucleophilic
attack by the indole (via 2A0-TS) leading to low-lying
intermediate 2B. Finally, ring opening via 2B-TS leads to the
allenyl product (2C). Overall, the barrier for this process is
14.8 kcal/mol.
Intrigued by the apparent divergent mode of activation for
these similar substrates, we next investigated the choice of
computational method and counterion in the reaction barriers.
As shown in Table 1, a screening of commonly used DFT
Table 1. Effect of DFT Method and Counterion on Relative
Energetics between Competing Transtion States 1A-TS/
1A0-TS and 2A-TS/2A0-TS
HLA − Hπ
substrate entry
computational method
counterion
Cl−
(kcal/mol)
4.9
1A
1
2
3
4
5
6
7
8
9
B3LYP/6-31G(d)-
LANL2DZ-gas
B3LYP/6-31G(d)-
LANL2DZDCM(CPCM)
B3LYP/6-31G(d)-SDD-
DCM(CPCM)
B97D/6-31G(d)-SDD-
DCM(CPCM)
M06L/6-31G(d)-SDD-
DCM(CPCM)
B3LYP/LANL2DZ-
DCM(CPCM)
B3LYP/LANL2DZ-
DCM(CPCM)
B3LYP/LANL2DZ-
DCM(CPCM)
Figure 1. Competing reaction pathways for the Pd-catalyzed [3,3]-
sigmatropic rearrangement of 2-allyloxyindoles.
Cl−
Cl−
Cl−
Cl−
Cl−
F−
4.4
6.3
4.7
7.4
8.8
5.4
6.3
6.7
shift (via 1A-TS) leading to the product 1C. The overall
barrier is 19.8 kcal/mol, and the products lie 8.2 kcal/mol
downhill. In contrast, the alternative Pd(II)-catalyzed pathway
via π-coordination to the alkene (1A0) is much higher in
energy (28.2 kcal/mol via 1A0-TS) and stepwise, albeit via
shallow intermediate 1B. These results are in agreement with
the experiment and show the importance, and role, of the 3-
ester functionality in accelerating the reaction. That is,
experimentally, the reaction proceeded at low temperatures
(0 °C) and the substrate scope was limited to substrates
bearing a carbonyl substituent at the C3-position.9
−
SbF6
−
B3LYP/LANL2DZ-
DCM(CPCM)
BF4
Next, we examined the energetics for the related Pd(II)-
catalyzed [3,3]-sigmatropic shift of 3-ester-2-propargyl-
oxyindoles. As shown in Figure 2, similar to the 3-ester-2-
2A
10
11
12
13
B3LYP/6-31G(d)-SDD-
Cl−
Cl−
Cl−
F−
−8.8
−5.3
−8.0
−7.8
DCM(CPCM)
M06L/6-31G(d)-SDD-
DCM(CPCM)
B3LYP/LANL2DZ-
DCM(CPCM)
B3LYP/LANL2DZ-
DCM(CPCM)
functionals and counterions predicted the same divergent
reactivity. That is, for alkenyl substrate 1A the barrier for π-
activation (via 1A0-TS) is higher in energy by 4−9 kcal/mol
than the Lewis acid pathway (via 1A-TS). In contrast, for
alkynyl 2A substrate, the Lewis acid promoted rearrangement
(via 2A-TS) is higher in energy than the π-activation pathway
(via 2A0-TS) by 5−9 kcal/mol. A similar trend was observed
using the full catalyst system [PdCl(BINAP)]+ and a number
of the substrates examined experimentally as outlined in the
Supporting Information, SI). Overall, the relative energetics
between the two pathways is unaffected by the method or by
the use of truncated substrates/ligands.
To gain insight into the origin of the different reactivity, we
applied a distortion−interaction analysis on the pathway
determining transition state structures (e.g., 1A-TS/1A0-TS
and 2A-TS/2A0-TS). Distortion−interaction analyses has
been used to rationalize reactivities and selectivities of various
organic and organometallic transformations.15,16 As shown in
Scheme 2, the energies [Etotal (Etot), Edistortion (Edis), and
Einteraction (Eint)] for the Lewis acid activation (via 1A-TS and
2A-TS) are nearly identical for both substrates (Scheme 2,
red). However, the energies for the π-activation are
significantly different (Scheme 2, blue). For the alkenyl
Figure 2. Competing reaction pathways for the Pd-catalyzed [3,3]-
sigmatropic rearrangement of 2-propargyloxyindoles.
allyloxyindoles, Lewis acid coordination to carbonyl and
ethereal moieties (2A) is favored by ∼12 kcal/mol over π-
coordination (2A0). In addition, the Lewis acid promoted
pathway proceeds in a concerted manner, while the π-
activation pathway proceeds in a stepwise fashion. However,
for the propargyl substrates, the barrier for the Lewis acid
promoted rearrangement (via 2A-TS) is much higher in energy
(18.9 kcal/mol) than the π-activation, stepwise pathway (via
2A0-TS). Overall, the lowest energy pathway proceeds via π-
B
DOI: 10.1021/acs.orglett.8b02864
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 2 via deuterium labeling studies. We hypothesized that
the concerted versus stepwise pathways would yield distinct
secondary isotope effects.17 Specifically, a concerted process,
such as via 2A-TS or the thermal rearrangement of a substrate
deuterated at the propargylic position, would give rise to a
normal, secondary isotope effect. On the other hand, no
change in hybridization would occur at this center if reaction
proceeds via 2A0-TS and no isotope effect would be expected.
Competition experiments with 6 and 6-D2 under thermal
conditions (Scheme 3) give rise to a secondary kinetic isotope
Scheme 2. Distortion−Interaction Model Analysis of
a
Barriers
a
Energies (kcal/mol) calculated using B3LYP/6-31G(d)-SDD(for
Pd) in CH2Cl2 (CPCM) relative to complexed substrates 1A and 2A
(see Figure 1 and Figure 2).
Scheme 3. Saucy−Marbet Claisen Rearrangement of 3-Aryl-
2-propargylindoles
substrate 1A, while both the magnitudes of distortion and
interaction energies are greater in 1A0-TS, the higher overall
barrier for the π-activation pathway (e.g., 1A0-TS) is attributed
to a higher distortion energy (58.2 kcal/mol) versus only 42.1
kcal/mol for the concerted Lewis acid rearrangement (1A-TS).
In contrast, for alkynyl substrate 2A, the lower barrier of 2A0-
TS is attributed to a minor distortion in the transition state
structure (33.1 kcal/mol) compared to 2A-TS (42.1 kcal/
mol). Note that the interaction energies for 2A-TS and 2A0-
TS are nearly identical.
To gain insight into the origin of enantioselectivity, we
modeled the diastereomeric transition states of 3-ester-2-
propargyloxyindoles 2A with the full chiral palladium(II)
[Pd(R-BINAP)Cl]+ (Figure 3). In agreement with experiment,
effect (kH/kD = 1.36
0.05), which is consistent with the
concerted reaction pathway (via TS1) predicted by calcu-
lations for the uncatalyzed reaction (see SI). On the other
hand, we observed no secondary kinetic isotope effect under
catalytic conditions (kH/kD = 0.98
0.05). This result is
consistent with the free-energy landscape depicted in Figure 2
where the rate-determining step (2A0-TS) does not involve
changes of hybridization at the labeled methylene.
Previously, all substrates tested in this reaction possessed the
3-ester moiety. However, DFT calculations and kinetic isotope
effect experiments support a stepwise mechanism via
coordination to an alkyne (i.e., without the association of the
ester to the metal in the rate-determining step). As a
consequence, computations predict that 2-propargyloxyindoles
without the 2-ester moiety should react under palladium
catalysis. To probe this hypothesis, we designed a substrate 4
(Scheme 4) that would be electronically similar to substrate 6.
Thus, an electron-poor aryl ring was placed at the C3 position
and the same bulky propargyl chain appended to the C2
position to slow any potential background [3,3]-sigmatropic
reaction.18 The coupling of indole 1 with aryl bromide 2 using
Figure 3. Origin of enantioselectivity in Pd(II)-(R)-BINAP-catalyzed
[3,3]-sigmatropic rearrangement of 2-propargyloxyindoles. Relative
energies are in kcal/mol, and distances in Å.
Scheme 4. Saucy−Marbet Claisen Rearrangement of 3-Aryl-
2-propargylindoles
the lowest energy diastereomeric transition state places the
ester group away from one of the phenyls of the chiral ligand
leading to the major enantiomer 4A (Figure 3, left).
Taken together, the different mode of reactivity is distortion-
controlled wherein the alkynyl substrates proceed via π-
activation (i.e., without coordination to the 3-ester moiety),
and the enantioselectivity arises from lesser steric interactions
between the ester and the ligand. As such, we hypothesize that
the palladium(II)-catalyzed [3,3]-rearrangement of 2-prop-
argylindoles can be expanded beyond the 3-ester-substituted
substrates. To test these computational predictions exper-
imentally, we probed the mechanistic picture presented in
C
DOI: 10.1021/acs.orglett.8b02864
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
conditions as reported by Bellina and Rossi,19 followed by a
two-step procedure to install the propargyloxy fragment,
delivered 4 (Scheme 4). Notably, compound 4 was stable at
room temperature for over a week without any detectable
thermal [3,3]-sigmatropic rearrangement. As a proof-of-
principle, we tested our computational predictions by
subjecting 4 to 20% mol of PdCl2(PhCN)2 at room
temperature. Under these conditions we found 4 delivered
the rearrangement product 5 cleanly in 57% yield. These
results provide experimental validation for the computations,
which predicted a stepwise, low-temperature palladium-
catalyzed rearrangement of propargyloxyindoles lacking the
C3-ester functionality.
In conclusion, we have examined the mechanism of Pd(II)-
catalyzed sigmatropic shifts of 3-allyloxy- and 3-propargylox-
yindoles using density functional theory. These calculations
revealed a divergent mode of reactivity which is surprising
given that the same catalyst gives the same level of the
enantioselectivity for the two similar substrates. Lower
distortion energies control the reaction pathway, where a
pathway involving Lewis acid coordination predominates for 3-
allyloxyindole substrates and a pathway involving π-coordina-
tion predominates for 3-propargyloxyindoles. Labeling experi-
ments are consistent with computational predictions. Further,
for the first time, 3-propargyloxyindoles lacking the 3-ester
functionality were found to be functional substrates for a
palladium(II)-catalyzed rearrangement. This study represents a
successful example using computational and experimental tools
to elucidate the mechanism of a transformation and to broaden
the substrate scope. Studies continue on the development of
the asymmetric rearrangement of 3-aryl-2-propargyloxyindoles.
REFERENCES
■
(1) Sankararaman, S. Pericyclic Reactions-A Textbook: Reactions,
Applications, and Theory; Wiley-VCH: New York, 2005.
(2) Cope, A. C.; Hardy, E. M. The Introduction of Substituted Vinyl
Groups. V. A Rearrangement Involving the Migration of an Allyl
Group in a Three-Carbon System. J. Am. Chem. Soc. 1940, 62, 441−
444.
(3) For a recent, and entertaining, historical perspective, see:
Borden, W. T. With a Little Help from My Friends: Forty Years of
Fruitful Chemical Collaborations. J. Org. Chem. 2011, 76, 2943−2964.
(4) Overman, L. E.; Knoll, F. M. Catalyzed sigmatropic rearrange-
ments. 5. Palladium(II) chloride catalyzed Cope rearrangements of
acyclic 1,5-dienes. J. Am. Chem. Soc. 1980, 102, 865−866.
(5) Felix, R. J.; Weber, D.; Gutierrez, O.; Tantillo, D. J.; Gagne, M.
R. A gold-catalysed enantioselective Cope rearrangement of achiral
1,5-dienes. Nat. Chem. 2012, 4, 405−409.
́
(6) Kaldre, D.; Gleason, J. L. An Organocatalytic Cope Rearrange-
ment. Angew. Chem., Int. Ed. 2016, 55, 11557−11561.
(7) Hiersemann, M.; Nubbemeyer, U. The Claisen Rearrangement;
Wiley-VCH: Weinheim, 2007.
(8) (a) Corey, E. J.; Kurti, L. Enantioselective Chemical Synthesis:
̈
Methods, Logic, and Practice; Direct Book Publishing: Dallas, TX,
2010; pp 110−113. (b) Nubbemeyer, U. Recent Advances in
Asymmetric [3,3]-Sigmatropic Rearrangements. Synthesis 2003,
2003, 961−1008.
(9) (a) Linton, E. C.; Kozlowski, M. C. Catalytic Enantioselective
Meerwein−Eschenmoser Claisen Rearrangement: Asymmetric Syn-
thesis of Allyl Oxindoles. J. Am. Chem. Soc. 2008, 130, 16162−16163.
(b) Cao, T.; Deitch, J.; Linton, E. C.; Kozlowski, M. C. Asymmetric
Synthesis of Allenyl Oxindoles and Spirooxindoles by a Catalytic
Enantioselective Saucy−Marbet Claisen Rearrangement. Angew.
Chem., Int. Ed. 2012, 51, 2448−2451. (c) Cao, T.; Linton, E. C.;
Deitch, J.; Berritt, S.; Kozlowski, M. C. Copper(II)- and Palladium-
(II)-Catalyzed Enantioselective Claisen Rearrangement of Allyloxy-
and Propargyloxy-Indoles to Quaternary Oxindoles and Spirocyclic
Lactones. J. Org. Chem. 2012, 77, 11034−11055.
(10) For a recent report, see: Liu, Y.; Liu, X.; Hu, H.; Guo, J.; Xia,
Y.; Lin, L.; Feng, X. Synergistic Kinetic Resolution and Asymmetric
Propargyl Claisen Rearrangement for the Synthesis of Chiral Allenes
Angew. Angew. Chem., Int. Ed. 2016, 55, 4054−4058.
(11) All calculations were carried out using Gaussian09. Frisch, M. J.
Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2013.
(12) For related computational studies, see: (a) Siebert, M. R.;
Tantillo, D. J. Transition-State Complexation in Palladium-Promoted
[3,3] Sigmatropic Shifts. J. Am. Chem. Soc. 2007, 129, 8686−8687.
(b) Watson, M. P.; Overman, L. E.; Bergman, R. G. Kinetic and
Computational Analysis of the Palladium(II)-Catalyzed Asymmetric
Allylic Trichloroacetimidate Rearrangement: Development of a
Model for Enantioselectivity. J. Am. Chem. Soc. 2007, 129, 5031−
5044.
(13) For a review, see: Gutierrez, O.; Kozlowski, M. C.
Computational Studies on Sigmatropic Rearrangements via π-
Activation by Palladium and Gold Catalysis. In Understanding
Organometallic Reaction Mechanisms and Catalysts; Ananikov, V. P.,
Ed.; Wiley-VCH: Weinheim, 2014.
(14) Unless otherwise specified, all calculations were carried out
using Gaussian 09. For Figures 1 and 2, computed with M06L/6-
31G(d)-SDD (for Pd) in CH2Cl2 (dielectric = 8.93) using the CPCM
solvation model. For full model system (i.e., Figure 3), calculations
were performed using B3LYP/LANL2DZ in CH2Cl2 (dielectric =
8.93) using the CPCM solvation model. Calculations using B3LYP/6-
31G(d)-LANL2DZ-CH2Cl2(CPCM) yield similar results. See SI for
additional computational details.
(15) For recent examples, see: (a) Barber, J. S.; Styduhar, E. D.;
Pham, H. V.; McMahon, T. C.; Houk, K. N.; Garg, N. K. Nitrone
Cycloadditions of 1,2-Cyclohexadiene. J. Am. Chem. Soc. 2016, 138,
2512−2515. (b) Green, A. G.; Liu, P.; Merlic, C. A.; Houk, K. N.
Distortion/Interaction Analysis Reveals the Origins of Selectivities in
Iridium-Catalyzed C−H Borylation of Substituted Arenes and 5-
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b02864.
Experimental procedures and spectroscopic data for all
new compounds; structures of all minimized ground
states and transition states (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*marisa@sas.upenn.edu
ORCID
Osvaldo Gutierrez: 0000-0001-8151-7519
Marisa C. Kozlowski: 0000-0002-4225-7125
Present Address
‡Department of Chemistry and Biochemistry, University of
Maryland, College Park, Maryland 20742, United States
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Institutes of Health
(GM-087605 to M.C.K.) and the National Science Foundation
(CHE1464778, CHE1764298 to M.C.K.). Computational
support was provided by XSEDE on SDSC Gordon (TG-
CHE120052).
D
DOI: 10.1021/acs.orglett.8b02864
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Membered Heterocycles. J. Am. Chem. Soc. 2014, 136, 4575−4583
and references therein .
(16) For a recent review, see: Bickelhaupt, F. M.; Houk, K. N.
Analyzing Reaction Rates with the Distortion/Interaction-Activation
Strain Model Angew. Angew. Chem., Int. Ed. 2017, 56, 10070−10086.
(17) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic
Chemistry; University Science Books: Sausalito, CA, 2006.
(18) See refs 7 and 9b for a discussion of how substituents affect the
rate of this Claisen rearrangement.
(19) Bellina, F.; Benelli, F.; Rossi, R. Direct Palladium-Catalyzed C-3
Arylation of Free (NH)-Indoles with Aryl Bromides under Ligandless
Conditions. J. Org. Chem. 2008, 73, 5529−5535.
E
DOI: 10.1021/acs.orglett.8b02864
Org. Lett. XXXX, XXX, XXX−XXX