Facile synthesis of Z -alkenes via uphill catalysis

Article abstract of DOI:10.1021/ja5019749

Catalytic access to thermodynamically less stable Z-alkenes has recently received considerable attention. These approaches have relied upon kinetic control of the reaction to arrive at the thermodynamically less stable geometrical isomer. Herein, we present an orthogonal approach which proceeds via photochemically catalyzed isomerization of the thermodynamic E-alkene to the less stable Z-isomer which occurs via a photochemical pumping mechanism. We consider two potential mechanisms. Importantly, the reaction conditions are mild, tolerant, and operationally simple and will be easily implemented.

Full text of DOI:10.1021/ja5019749

Communication
pubs.acs.org/JACS
Facile Synthesis of Z‑Alkenes via Uphill Catalysis
Kamaljeet Singh, Shannon J. Staig, and Jimmie D. Weaver*
Department of Chemistry, Oklahoma State University, Stillwater, Oklahoma 74078, United States
S
* Supporting Information
result of the catalysis. From a synthetic perspective, such a
process could take advantage of the many existing methods to
synthesize the E-alkenes for the synthesis of Z-alkenes which
could give it a distinct advantage when compared to other
methods. However, challenges to this strategy do exist. For
instance, the molecule must not decompose upon irradiation,
the thermal back reaction should not take place, and finally a
significant rate difference in the catalytic excitation of the E-
and Z-isomers must exist. The first issue is mostly overcome by
using a photocatalyst which absorbs in the visible region.
Solutions for the other issues are less obvious, and herein we
present one successful strategy.
Allylamines are an important class of biologically active
molecules,8 and in contrast to the E-isomer, accessing the Z-
isomers is particularly challenging. Most commonly, partial
hydrogenation, over a poisoned catalyst, of protected
propargylamines is employed. This requires starting from the
alkyne as well as a protection and deprotection step.9 Thus,
there is a real need for methods that can efficiently deliver Z-
allylamines in fewer steps which utilize more accessible starting
materials and consequently seemed like a good place to start
our investigation.
ABSTRACT: Catalytic access to thermodynamically less
stable Z-alkenes has recently received considerable
attention. These approaches have relied upon kinetic
control of the reaction to arrive at the thermodynamically
less stable geometrical isomer. Herein, we present an
orthogonal approach which proceeds via photochemically
catalyzed isomerization of the thermodynamic E-alkene to
the less stable Z-isomer which occurs via a photochemical
pumping mechanism. We consider two potential mecha-
nisms. Importantly, the reaction conditions are mild,
tolerant, and operationally simple and will be easily
implemented.
lefins are a fundamental building block for the synthetic
O
community, and many possess biological activity. Not
surprisingly, considerable effort has gone into developing
methods that can lead to stereodefined alkenes. While methods
to access thermodynamically more stable E-alkenes are
abundant, corresponding methods to access the less stable Z-
alkene are far less common. A number of conceptually diverse
strategies have been employed to access Z-alkenes1 (Figure 1)
Our initial efforts focused on cinnamyl derived amines
(Table 1). The reaction parameters were investigated utilizing
E-1a. Fortuitously, our initial conditions gave, within the
detection limits,10 almost exclusively the Z-isomer (Z-1a).
Controls showed that both the photocatalyst fac-tris[2-
phenylpyridinato-C2,N] iridium(III), [Ir(ppy)3], and light
were essential (entry 2). Remarkably, the isomerization worked
well in a number of solvents (entry 3) with the exception of
nitromethane (entry 4) which may be due to efficient
quenching of the excited photocatalyst by solvent.11 Interest-
ingly, the reaction works well over a range of concentrations
(entries 5−8). The effect of the external base was probed both
by removing it (entry 9) and by substituting it (entry 10−12).
While the effect is subtle, EtNiPr2 does seem to promote the
reaction and is easily removed in vacuo after reaction
completion. Other bases were less effective. The presence of
air prevents the isomerization from occurring but does not
cause degradation of the starting material (entry 13).
Conveniently, the reaction did not display moisture sensitivity
(entry 14) adding to the operational simplicity of the method.
Having established the reaction conditions, we next sought to
evaluate the scope of the reaction (Table 2). The reaction
works well for a number of other methylene carbinamines (Z-
2a, 3a, 4a), as well as substrates with a methine carbinamine
(Z-5a, 6a, 7a). Reactions with aryl bromides (Z-8a, 9a, 10a)
Figure 1. A comparison of the current strategies to access Z-alkenes to
our proposed photochemically driven isomerization.
and include semihydrogenation of alkynes,2 the Wittig
reaction,3 stereospecific cross-coupling reactions,4 and more
recently Z-selective cross-metathesis5 and terminal-to-internal
olefin isomerization.6 Inherently, all of these strategies rely on
kinetic control of the reaction and use higher energy reagents to
access the Z-alkene.
A strategy missing from this collection of methods is the
photocatalytic E-to-Z isomerization,7 which does not rely on
kinetic control of the reaction to deliver the desired product.
Photocatalysis has the ability to “pump” or circumvent the
normal thermodynamic constraints of a reaction. In other
words, the more energetic product can actually be favored as a
Received: February 25, 2014
Published: March 28, 2014
© 2014 American Chemical Society
5275
dx.doi.org/10.1021/ja5019749 | J. Am. Chem. Soc. 2014, 136, 5275−5278

Journal of the American Chemical Society
Communication
Table 1. Effects of Reaction Parameters
Table 2. Scope of the Photochemical Isomerization
GC-MS
conversions
entry
deviation from standard conditions
1
none
>95%
0%
2
no Ir(ppy)3 or in the dark
DMA, THF, Tol, DCM, and DMF
nitromethane
3
>95%
0%
4
5
0.02 M in substrate
88%
93%
92%
90%
88%
94%
90%
84%
0%
6
0.1 M in substrate
7
1.0 M in substrate
8
2.0 M in substrate
9
no EtNiPr2, after 20 h
10
11
12
13
14
0.1 equiv of EtNiPr2, after 20 h
0.1 equiv of DBU, K2CO3, Cs2CO3, after 20 h
0.1 equiv of pyridine, imidazole, after 20 h
without degassing
with 1 drop of H2O
>95%
highlight the orthogonality of this chemistry to other late
transition metal chemistry (i.e., Pd-catalyzed cross-couplings).
Substitution is well tolerated at the 4-position (Z-11a, 13a), 3-
position (Z-12a), and the 2-position (Z-14a). Furthermore, the
reaction tolerates secondary aliphatic amines (Z-15a, 16a) and
aromatic amines (Z-17a). The reaction tolerates amides (Z-
18a) giving better than 99:1 Z/E after chromatography.
Additionally, acid labile silyl ether (Z-19a), which would be
difficult to maintain during a partial hydrogenation under acidic
conditions, is a particularly good substrate. Finally, several
common N-protecting groups were evaluated. Tosyl (Z-20a)
protected, N-acyl (Z-21a), and even acid sensitive N-Boc
protected amines (Z-22a) worked well. It is noteworthy that
the reactions are remarkably clean with essentially no other
chemistry occurring under these conditions and, in most cases,
the lower isolated yields are simply the result of the partial
resolution of the alkenes by chromatography.
Having explored the scope we next began to consider the
mechanism. Given the uphill nature of the reaction in the
ground state as well as the need for the photocatalyst and light,
we believed it was unlikely that the reaction could be
proceeding by a purely thermal process. Rather, the reaction
must be driven by light. Given that substrates do not absorb
light in the visible region as well as the need for the
photocatalyst, a direct photochemical isomerization is unlikely.
Instead, the selectivity of the reaction must result primarily
from a difference in rates of the photoquenching event of the
photocatalyst by the E- and Z-isomers.
With E- and Z-isomers in hand, we were able to investigate
this question by performing photoquenching studies with both
isomers. In this study, the Ir(ppy)3 is excited by the absorption
of photons at the appropriate wavelength (370 nm) to facilitate
a metal ligand charge transfer-excited state which emits at 520
nm. Quenching of the excited photocatalyst can be observed by
monitoring the disappearance of the emissive wavelength (520
nm). From the slope of the line generated by plotting of I°/I (I
= emission intensity) vs concentration of the quencher, we can
glean information regarding the rate of the quenching event.
The slope of the lines in this Stern−Volmer plot is proportional
to the rate of the quenching event.12 Indeed, we found that E-
19a (>99:1, E/Z) quenches the emission of the excited catalyst
with a slope of 0.0915 while Z-19a (3:97, E/Z) quenches the
excited state of the catalyst with a slope of 0.012. These results
suggest that the E-isomer quenches the photoexcited state with
a relative rate 8-fold greater than that of the Z-isomer,13
confirming our suspicions and suggesting that the geometrical
isomers quench the photocatalyst at significantly different rates
(Figure 2).
Having established that the difference in the rate of
quenching of the photoexcited catalyst by the two isomers is
responsible for the buildup of the Z-isomer, we wanted to
understand the nature of the quenching event. Two conceivable
mechanisms that lead to photoquenching of the Ir(ppy)3*
catalyst are outlined below (Figure 3). In the first (eq 1),
reductive quenching by electron transfer from the amine to the
excited photocatalyst14 gives radical cation E-23a. Upon amine
oxidation, the αC−H becomes significantly more acidic15 and
could be reversibly deprotonated by an exogenous base to give
allylic radical E-23b. Radical E-23b is expected to have a
diminished barrier toward cis−trans isomerization16 and would
lead to Z-23b. Z-23b would undergo protonation and back
5276
dx.doi.org/10.1021/ja5019749 | J. Am. Chem. Soc. 2014, 136, 5275−5278

Journal of the American Chemical Society
Communication
Scheme 1. Mechanistically Insightful Reactions;
Stereochemical Probe and Reaction of Simple Styrene
Derivative
Figure 2. A Stern−Volmer photoquenching plot of E-19a and Z-19a.
If the direct ET mechanism is operative, then it would be
reasonable to assume that the amine may not be a necessary
structural component. Thus, we subjected several nonamine
containing styrenes, including hydrocarbons, alcohols, and
acetates (E-24a−29a) to the reaction conditions. Indeed, all of
the substrates undergo isomerization to give Z-24a−29a in Z/E
ratios ranging from 87:13 to 78:22, further supporting the
direct ET mechanism and suggesting a much broader scope for
this reaction should be possible.
In conclusion, we have demonstrated a novel approach to
access thermodynamically less stable alkenes directly from the
more stable and easily synthesized E-alkenes in synthetically
useful ratios and yields. Importantly, the starting alkenes are
readily available and the photocatalyst is both highly efficient
and commercially available. The reaction is remarkably clean
and allows for the incorporation of a number of functional
groups and maintains adjacent stereocenters in the substrates.
Our preliminary mechanistic experiments indicate that the
reaction proceeds by selective triplet sensitization of the E-
alkene, suggesting that the reaction should be general to the
styrenyl motif. This reaction serves as a rare example of uphill
catalysis, which arguably is an underutilized strategy in the
catalysis field, and provides useful insight into this under-
exploited strategy.
Figure 3. Two possible mechanisms.
electron transfer to afford Z-23, which quenches the excited
photocatalyst more slowly than E-23, resulting in buildup of Z-
23.
We next considered an energy transfer (ET) mechanism in
which enerergy transfer from the photoexcited catalyst leads to
a biradical intermediate (E-23, eq 2), which is expected to
rapidly lead to intermediate 3312 which has the critical
geometry in which a 90° rotation about the C−C has
occurred.17 33 can undergo intersystem crossing to give either
E-23 or Z-23. A buildup of Z-23 would result from a greater ktt
value.18 Interestingly, in contrast to what we observe, Yoon
reported a related triplet sensitized [2 + 2] of cinnamyl ethers
in which he observed rapid E/Z isomerization. One intriguing
potential explanation is the difference in emissive energy of the
different photocatalysts used.19 Specifically, the catalyst used by
Yoon has greater emissive energy (470 nm, 61 kcal/mol) than
the Ir(ppy)3 catalyst (520 nm in MeCN, 55 kcal/mol).
Given that radical inversion of sp3 hybridized carbons is rapid
(i.e., E/Z-23b, Figure 3),20 we expected that if the reductive
quenching mechanism were operative, any stereochemical
information at a carbinamine stereocenter would be lost.
Thus, we synthesized a chiral, nonracemic alkene E-(R)-10a
(Scheme 1) and subjected it to the reaction conditions. Upon
photocatalytic E, Z-isomerization, no racemization was
detected, indicating that the reductive quenching mechanism
is not likely operative and suggesting direct ET as the operative
mechanism.
ASSOCIATED CONTENT
* Supporting Information
Complete experimental procedures and product character-
ization. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
Jimmie.weaver@okstate.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by startup funds from Oklahoma
State University.
■
REFERENCES
■
(1) For a nice review of strategies to access Z-alkenes, see: Siau, W.-
Y.; Zhang, Y.; Zhao, Y. Top. Curr. Chem. 2012, 327, 33.
5277
dx.doi.org/10.1021/ja5019749 | J. Am. Chem. Soc. 2014, 136, 5275−5278

Journal of the American Chemical Society
Communication
(2) Lindlar, H. Helv. Chim. Acta 1952, 35, 446.
unity and thus the overall equilibrium to be primarily a reflection of
the rate of quenching. This is consistent with the findings of
Hammond, which suggest that the rates of intersystem crossing (ktrans
and kcis) approach unity for the isomerization of 1,3-dienes and
stillbenes by high energy sensitizers; see: Hammond, G. S.; Saltiel, J.;
Lamola, A. A.; Turro, N. J.; Bradshaw, J. S.; Cowan, D. O.; Counsell,
R. C.; Vogt, V.; Dalton, C. J. Am. Chem. Soc. 1964, 86, 3197.
(19) One intriguing potential explanation for the difference in the
observed reactivity of our system and that of Yoon’s is the emissive
energy of the different photocatalysts. The photocatalyst used by Yoon
Ir(dF(CF3)ppy)2dtbbpy+ has an emission wavelength (470 nm) (61
kcal/mol) while fac-Ir(ppy)3 has a peak emission at 520 nm in MeCN
(55 kcal/mol): Lu, Z.; Yoon, T. P. Angew. Chem., Int. Ed. 2012, 51,
10329.
(3) Bergelson, L. D.; Shemyakin, M. M. Tetrahedron 1963, 19, 149.
(4) Hegedus, L. S. Transition Metals in the Synthesis of Complex
Organic Molecules; University Science Books: Mill Valley, CA, 1994.
(5) For recent examples of Z-selective cross-metathesis, see: (a) Koh,
M. J.; Khan, R. K. M.; Torker, S.; Hoveyda, A. H. Angew. Chem., Int.
Ed. 2014, 53, 1968. (b) Quigley, B. L.; Grubbs, R. H. Chem. Sci. 2013,
5, 501. (c) Gottumukkala, A. L.; Madduri, A. V. R.; Minnaard, A. J.
ChemCatChem 2012, 4, 462. (d) Meek, S. J.; O’Brien, R. V.; Llaveria,
J.; Schrock, R. R.; Hoveyda, A. H. Nature 2011, 471, 461. (e) Endo, K.;
Grubbs, R. H. J. Am. Chem. Soc. 2011, 133, 8525.
(6) (a) Zhuo, L.-G.; Yao, Z.-K.; Yu, Z.-X. Org. Lett. 2013, 15, 4634.
(b) Chen, C.; Dugan, T. R.; Brennessel, W. W.; Weix, D. J.; Holland,
P. L. J. Am. Chem. Soc. 2014, 136, 945.
(20) Houk found a barrier of only 1.2 kcal/mol for tBu radical
inversion, suggesting racemization would be fast if the radical was
formed; see: Paddon-Row, M. N.; Houk, K. N. J. Am. Chem. Soc. 1981,
103, 5046.
(7) For an early example of 1,3-diene photocatalyzed isomerization,
see: Ramamurthy, V.; Liu, R. S. H. J. Am. Chem. Soc. 1976, 98, 2935.
For another example of a net uphill reaction on 1,3-dienes, see:
Punner, F.; Schmidt, A.; Hilt, G. Angew. Chem., Int. Ed. 2012, 51, 1270.
̈
For a study using classic triplet sensitizers on a few related substrates,
see: Zhao, Y.-P.; Yang, L.-Y.; Liu, R. S. H. Green Chem. 2009, 11, 837−
842.
(8) The fungicidal drugs Lamisil and Naftin are both allylamines.
(9) For access to Z-1,2-disubstituted allylamines a semi-hydro-
genation approach is taken using propargylic N-phthalamides or
ammonium salts using Lindlar’s catalyst, which necessitates use of the
alkyne and the added complications of using the phthalamide
protecting group or running the reaction under acidic conditions;
for an example, see: Tomassy, B.; Zwierzak, A. Synth. Commun. 1998,
28, 1201.
(10) No special measures were taken to increase the sensitivity of the
GCMS towards our samples or to quantify these results but rather
were used to illuminate the relevant trends.
(11) Currently, it is not clear why nitromethane prevents the
isomerization, given that its redox potentials 2.4 V vs NHE suggest
it is unlikely to oxidatively or reductively quench the photocatalyst.
Furthermore, the lowest lying electronic transition is ∼3.4 eV which is
significantly above the 2.5 eV energy of the excited photocatalyst.
(a) Voorhies, J. D.; Schurdak, E. J. Anal. Chem. 1962, 34, 939.
(b) Arenas, J. F.; Otero, J. C.; Pelae
119, 7814.
́
z, D.; Soto, J. J. Chem. Phys. 2003,
(12) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular
Photochemistry of Organic Molecules; University Science Books:
Sausalito, CA, 2010; pp xxxiii, 1084.
(13) Given that the Z-isomer is not geometrically pure, we are
hesitant to assume any greater precision.
(14) For an example of reductive quenching of the photocatalyst by
an amine, see: Lalevee, J.; Tehfe, M.-A.; Dumur, F.; Gigmes, D.;
Blanchard, N.; Morlet-Savary, F.; Fouassier, J. P. ACS Macro Lett.
2012, 1, 286.
(15) Stephenson has estimated the pKa of the triethylamine radical
cation to be 26.7; see: (a) Freeman, D. B.; Furst, L.; Condie, A. G.;
Stephenson, C. R. J. Org. Lett. 2012, 14, 94. (b) Singh, A.; Arora, A.;
Weaver, J. D. Org. Lett. 2013, 15, 5390.
(16) The barrier to isomerization of such an allylic radical is ∼15
kcal/mol; see: Dibble, T. S.; Sha, Y.; Thornton, W. F.; Zhang, F. J.
Phys. Chem. A 2012, 116, 7603.
(17) For studies involving the geometry, see: (a) Bonneau, R.;
Herran, B. Laser Chem. 1984, 4, 151. (b) Caldwell, R. A.; Chi Van, C.
J. Am. Chem. Soc. 1981, 103, 3594.
(18) Technically, a buildup of Z-23 should be expressed as Z-23/E-
23 = (ktt*kcis)/(ktc*ktrans). We have experimentally observed a 93:7
photostationary state. Our photoquenching studies indicated a 7.6-fold
rate difference, which amounts to an 88:12 product ratio. This
difference is relatively small but could be attributed to the less than
geometrically pure nature of the alkene (97:3, Z/E) which would make
the Z-alkene appear to quench with a greater rate due to the presence
of the E-alkene. Alternatively, it could reflect the rates of ISC (kcis and
ktrans), but more likely the difference is a result of a combination of
these two factors. For now we assume these relative rates to be near
5278
dx.doi.org/10.1021/ja5019749 | J. Am. Chem. Soc. 2014, 136, 5275−5278