Photocatalytic E → Z isomerization of polarized alkenes inspired by the visual cycle: Mechanistic dichotomy and origin of selectivity

Article abstract of DOI:10.1021/acs.joc.7b01281

Iteratively executed with exquisite spatial and temporal control, the selective isomerization of polarized alkenes underpins a plethora of complex biological processes ranging from natural product biosynthesis through to the mammalian visual cycle. However, nature's proficiency conceals the inherent difficulties in replicating this contrathermodynamic transformation in the laboratory. Recently, we disclosed the first highly Z-selective isomerization of polarized alkenes, employing the cinnamoyl chromophore as a retinal surrogate under UV-irradiation (402 nm) with (-)-riboflavin (Vitamin B₂) as an inexpensive, organic photocatalyst (J. Am. Chem. Soc. 2015, 137, 11254-11257). This study was inspired by the propensity of crystalline (-)-riboflavin in the eyes of vertebrates to invert the intrinsic directionality of retinal isomerization. Herein, we extend this methodology to include a bioinspired, catalytic E → Z isomerization of α,β-unsaturated nitriles, thereby mimicking the intermediate Opsin-derived, protonated Schiff base in the visual cycle with simple polarized alkenes. Replacement of the iminium motif by a cyano group is well tolerated and gives an additional degree of versatility for postisomerization functionalization. Broad substrate scope is demonstrated (up to 99:1 Z:E) together with evidence of mechanistic dichotomy via both singlet and triplet energy transfer mechanisms. Kinetic studies, temperature dependent photostationary state correlations and investigation of substituent-based electronic perturbation of the alkene identified polarization combined with increased Z-isomer activation barriers as the selectivity governing factors in catalysis. This investigation demonstrates the importance of internal structural preorganization on photostationary composition and explicates the augmented Z-selectivity upon hydrogen-alkyl exchange at the β-position of the alkene.

Full text of DOI:10.1021/acs.joc.7b01281

Subscriber access provided by NEW YORK UNIV
Article
Photocatalytic E#Z Isomerization of Polarized Alkenes Inspired by
the Visual Cycle: Mechanistic Dichotomy and Origin of Selectivity
Jan B. Metternich, Denis Gennadievich Artiukhin, Mareike C Holland,
Maximilian von Bremen-Kühne, Johannes Neugebauer, and Ryan Gilmour
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.7b01281 • Publication Date (Web): 27 Jun 2017
Downloaded from http://pubs.acs.org on June 29, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 69
The Journal of Organic Chemistry
1
2
3
4
Photocatalytic E→Z Isomerization of Polarized Alkenes
Inspired by the Visual Cycle: Mechanistic Dichotomy
and Origin of Selectivity
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Jan B. Metternich,* Denis G. Artiukhin, Mareike C. Holland, Maximilian von BremenꢀKühne, Johannes
Neugebauer* and Ryan Gilmour*
Institute for Organic Chemistry, Westfälische WilhelmsꢀUniversität Münster, Corrensstrasse 40, 48149
Münster, Germany.
KEYWORDS. Biomimetic; Mechanism; Photocatalysis; Flavin.
ABSTRACT: Iteratively executed with exquisite spatial and temporal control, the selective isomerization
of polarized alkenes underpins a plethora of complex biological processes ranging from natural product
biosynthesis through to the mammalian visual cycle. However, Nature’s proficiency conceals the
inherent difficulties in replicating this contraꢀthermodynamic transformation in the laboratory. Recently,
we disclosed the first highly Zꢀselective isomerization of polarized alkenes, employing the cinnamoyl
chromophore as a retinal surrogate under UVꢀirradiation (402 nm) with (–)ꢀriboflavin (vitamin B ) as an
2
inexpensive, organic photocatalyst (J. Am. Chem. Soc. 2015, 137, 11254–11257). This study was
inspired by the propensity of crystalline (–)ꢀriboflavin in the eyes of vertebrae to invert the intrinsic
directionality of retinal isomerization. Herein, we extend this methodology to include a bioꢀinspired,
catalytic E→Z isomerization of
α,βꢀunsaturated nitriles, thereby mimicking the intermediate Opsinꢀ
derived, protonated Schiff base in the visual cycle with simple polarized alkenes. Replacement of the
iminium motif by a cyano group is well tolerated and gives an additional degree of versatility for postꢀ
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 69
isomerization functionalization. Broad substrate scope is demonstrated (up to 99:1 Z:E) together with
evidence of mechanistic dichotomy via both singlet and triplet energy transfer mechanisms. Kinetic
studies, temperature dependent photostationary state correlations and investigation of substituentꢀbased
electronic perturbation of the alkene identified polarization combined with increased Zꢀisomer
activation barriers as the selectivity governing factors in catalysis. This investigation demonstrates the
importance of internal structural preorganization on photostationary composition and explicates the
augmented Zꢀselectivity upon hydrogenꢀalkyl exchange at the βꢀposition of the alkene.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
Conceptualizing and consolidating biological structure – function interplay in a laboratory paradigm has
1
2
become a hallmark of contemporary organocatalysis. Inimitable in the mimesis of enzymatic catalysis,
this discipline’s legacy is the platform to translate Nature’s intricate reactivity modes into robust,
3
practical chemical methods. This juxtaposition, forged as traditional boundaries in the natural sciences
4
coalesce, ensures that rigorous mechanistic interrogation of biological catalysis continues to inspire and
inform this expansive discipline. Unsurprisingly, the most celebrated covalent and nonꢀcovalent
organocatalysts share common structural resemblances with the more complex biological ensembles
from which they are inspired. This has resulted in structureꢀfunction relationships that have been
definitively classified and are now widely exploited by practitioners of organic synthesis. With the
5
emergent success of organic photocatalysis comes the opportunity to capitalize on the rich tapestry of
biological photosystems and thus utilize Nature’s blueprints for reaction design. In contrast to covalent
and nonꢀcovalent organocatalysis, establishing structureꢀbased comparisons to biology in
photochemistry is more challenging: This can be attributed to the sensitivity of biological photosystems
when altered, or removed from the natural environment of the host protein.
ACS Paragon Plus Environment

Page 3 of 69
The Journal of Organic Chemistry
Scheme 1 Conceptual framework for photocatalytic isomerization of polarized alkenes.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Nonetheless, the conceptual foundations of a variety of biological phenomena are adaptable and have
been successfully translated to small molecule catalysis. Protonꢀcoupled electron transfer (PCET),
6
ubiquitous in enzymatic redox catalysis, beautifully exemplifies this notion. As part of a broader
7
interest in acyclic conformational control, we recently became interested in the structural and
8
photochemical prerequisites for retinal isomerization in the vertebrate visual cycle. This curiosity arose
from what was perceived to be a strategic gap in the catalysis ordnance, namely general strategies to
realize the selective contraꢀthermodynamic E → Z isomerization of polarized alkenes using inexpensive
small molecule photoꢀorganocatalysts. This fundamental transformation has significant appeal both in
9
terms of atom economy, and in accessing configurationally defined alkenes for subsequent application
1
0
in stereospecific transformations. Bioꢀinspired organic photocatalysis offers a unique paradigm to
5
redress this balance, by amalgamating established photochemical activation modes with the structural
and mechanistic precedents found in biology.
Laboratory precedent: It is pertinent to note that a survey of existing alkene isomerization catalysts
reveals a rich body of literature describing positional isomerization, in contrast to the limited reports
1
1
pertaining to geometric isomerization: In both scenarios, transition metal complexes predominate.
Organometallic catalysts based on Ir(III) have also proven to be highly effective in the photocatalytic
12
isomerization of styrenyl derivatives as reported by Weaver and coꢀworkers. This latter example is
particularly noteworthy on account of the efficiency and selectivity of catalysis, and reiterates the
potential of photochemical activation modes in addressing the strategic gap in catalytic Zꢀselective
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 69
alkene isomerization. Despite this seminal report, the historic mechanistic delineation of the E to Zꢀ
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
isomerization of simple alkenes has remained relatively unutilized in small molecule catalyst design.
1
3
14
15
Detailed investigations by Hammond, Arai and others, carefully delineated the sensitized
16
isomerization mechanisms of various stilbenes, styrenes and fumarates. Whilst these important works
provide valuable guidelines, scope, yields and translation to catalysis are conspicuously absent. Due to
the nature of the experiments performed, stoichiometric quantities of organic sensitizers were routinely
employed. Furthermore, examples of polarized, nonꢀsymmetrical alkenes are not disclosed despite their
synthetic versatility. Many of these limitations also persist in the photochemical isomerization of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7
alkenes facilitated by stoichiometric Lewis acids.
It was envisaged that by integrating the stoichiometric sensitization paradigm delineated in the midꢀ20
th
century for nonꢀpolarized alkenes, with pioneering studies on the visual cycle from around the same
18
time period, that a theoretical framework could be constructed to facilitate the Zꢀselective
isomerization of activated alkenes from the more abundant Eꢀconfigured precursors. Conceptually, this
would require an initial, substrateꢀselective sensitization of the starting material to confer directionality
(Scheme 1). This would result from biasing the sensitization efficiency (catalystꢀsubstrate > catalystꢀ
product), and could likely be preꢀdetermined at a structural level by perturbation of the inherent πꢀ
system following the catalysis event. This is achievable by inducing nonꢀbonding interactions in the
product Zꢀalkene to reduce conjugation and thereby alter the photophysical signature. Since the
ephemeral intermediate generated by photocatalyst activation lacks the initial πꢀbond, an orthogonal
arrangement of the substituent is preferred such that the system may collapse to either the Eꢀ or Zꢀ
isomer. Under a substrateꢀselective activation regime, iterative excitation would logically favor
formation of the Zꢀisomer. Photochemical activation would thus mitigate the risk of microscopic
reversibility and the formation of thermodynamic mixtures.
ACS Paragon Plus Environment

Page 5 of 69
The Journal of Organic Chemistry
Scheme 2 Reaction design considerations. Left: the Z→E photoisomerization of retinal. Center:
Qualitative report that (–)ꢀriboflavin photocatalyzes the reverse reaction. Right: A bioꢀinspired catalytic,
Zꢀselective isomerization of cinnamonitriles.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Precedence in biology: In vertebrae, 11ꢀcisꢀretinal is isomerized, via the protonated Opsinꢀderived
1
9
Schiff base, by retinalꢀisomerase to furnish allꢀtransꢀretinal (Scheme 2, left). Iteratively performed
with exquisite spacial and temporal efficiency, this process is the ultimate blueprint for photocatalystꢀ
based alkene isomerization, but with one obvious caveat: In the visual cycle, the Z→E direction is
favored in the photochemical process. However, a study from Walker and Radda in the 1960s, reported
that crystalline (–)ꢀriboflavin (vitamin B2), present in the eyes of various mammals and fish, facilitates
1
8
the contraꢀthermodynamic E→Z isomerization of retinal (Scheme 2, center). Although of a qualitative
nature, this important study contained two critical details: (i) That an inexpensive, and commercially
available coꢀfactor can function as an effective Zꢀselective isomerization catalyst of a polarized alkene,
and (ii) that by virtue of its location in the eye, a photocatalytic mechanism was likely operational. This
important observation led to the conceptual framework for a generic photocatalytic isomerization of
polarized alkene as summarized in Scheme 1.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 69
Results and Discussion
This laboratory recently reported that (–)ꢀriboflavin is an efficient, inexpensive photocatalyst for the
isomerization of ꢀunsaturated carbonyl compounds under UVꢀirradiation (402 nm) with low
photocatalyst loadings (5 mol%). By extension, the transformation can also be telescoped to generate
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
α,β
2
0
2
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
coumarins directly from (E)ꢀcinnamic acids via two discrete photochemical activation modes. Thus, it
has been possible to augment the synthetic virtuosity of flavins in organocatalysis to complement
22,23
24
25
existing redox
cycloaddition and halogenation processes. In order to augment the scope of this
bioꢀinspired transformation, a detailed mechanistic interrogation of catalysis was devised using
cinnamonitrile as a model system. This versatile scaffold provides the requisite polarized olefin /
chromophore and mimics the Opsinꢀderived Schiff base that undergoes isomerization in the visual
cycle. Moreover, the nitrile group acts as a masked amine in this oxidized form and thereby alleviates
the risk of competing, physical or chemical quenching mechanisms. This organocatalytic approach
would complement the existing work of Weaver and coꢀworkers using Ir(III) catalysis, to provide rapid
1
2
entry to Zꢀconfigured allylic amine precursors.
It was envisaged that the catalysis activation mode would not be adversely affected by the carbonyl to
nitrile substitution in the substrate, such that the generic (E)ꢀcinnamonitrile πꢀsystem would likely
match the excitation energy of (–)ꢀriboflavin. Moreover, selectivity enhancement would be possible by
introduction of a βꢀsubstituent, thereby inducing nonꢀbonding interactions in the product isomer. This
would necessary tilt the πꢀsystem, leading to deconjugation and thus diminish activation efficiency
(Scheme 2, right).
Expanding Substrate Scope: As a starting point for catalysis, commercially available cinnamonitrile 1
and (–)ꢀriboflavin (5 mol%) were dissolved in MeCN and irradiated at 402 nm (Table 1). Gratifyingly, it
was possible to recover the material quantitatively by simple filtration through a silica plug. Analysis by
1
H NMR spectroscopy confirmed a Z:E ratio 63:37 thereby providing an initial proof of concept. To
2
substantiate the proposal that a βꢀsubstituent (R ) would enhance selectivity by augmenting strain in the
ACS Paragon Plus Environment

Page 7 of 69
The Journal of Organic Chemistry
product, a series of substituted cinnamonitriles were prepared (2ꢀ10) and submitted to standard
photocatalysis conditions. As expected, this led to a significant improvement in the geometric ratios
with excellent material recovery.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
[a]
Table 1 The (–)ꢀriboflavinꢀcatalyzed isomerization of cinnamonitriles
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[
a]
Standard reaction conditions: Reactions were performed with 0.1 mmol substrate (E:Z >20:1) at
ambient temperature in MeCN using 5 mol% catalyst loading under 12 hours of UVꢀlight irradiation
1
(
402 nm). Z:E ratios reported are the average of two experiments and were determined by H NMR
[b]
spectroscopy and HPLC analysis.
priority of F than C.
N.B. product Z:E ratio is inverted to reflect the higher IUPAC
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 69
1
Installation of the methyl group (2) led to a Z:E ratio of 91:9 as determined by H NMR spectroscopy,
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
and this could be increased further to Z:E 97:3 when R = Et (3). Extending the chain length to R = nPr
2
(4) and R = nBu (5) led to a plateau in selectivity (Z:E 96:4). This was also observed in the branched
2
2
cases where R = iPr (6) and R = iBu (7) led to Z:E ratios of 96:4 and 95:5, respectively. Substrate 8
2
where R = cyPr also gave a Z:E ratio of 95:5. Interestingly, isomerization of the tBu derivative 9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
proved to be challenging with only traces of the Zꢀisomer having been detected (Z:E 2:98). This is likely
a consequence of 1,3ꢀallylic strain in the starting material thus rotating the phenyl ring of the plane: This
would disrupt conjugation of the chromophore πꢀsystem and renders substrate activation challenging. In
contrast, substrates 2ꢀ8 contain at least one methine proton at the αꢀcarbon of the substituent to
minimize this unfavorable interaction. As a control experiment, the fluorinated species 10 was
investigated and behaved in a comparable manner to 1. To explore the influence of electronically
modulating the aryl unit, a series of pꢀCF and pꢀOMe derivatives were subjected to catalysis
3
conditions. The Z:E selectivities obtained with substrates 11ꢀ15 were found to be identical to those of 1ꢀ
2
5 (pꢀH): (R = H, 11, Z:E 62:38; Me, 12, Z:E 91:9; Et, 13, Z:E 97:3; nPr, 14, Z:E 96:4; nBu, 15, Z:E
9
6:4). A slight deviation in selectivity was observed in the pꢀOMe derivatives 16ꢀ20 when compared
2
with the parent series 1ꢀ5: pꢀOMe (R = H, 16, Z:E 44:56; Me, 17, Z:E 81:19; Et, 18, Z:E 91:9; nPr, 19,
Z:E 90:10; nBu, 20, Z:E 90:10). The origin of selectivity required clarification. Further examination of
the substrate scope revealed a general tolerance for electron deficient (21 – 23, 27 and 28; up to Z:E
96:4) and electron rich substituents (24, 25 up to Z:E 95:5). Heterocyclic aromatic substrates (30 Z:E
9
1:9) as well as larger πꢀsystems (26 Z:E 70:30 and 31 Z:E 91:9) were well tolerated with good to
excellent selectivities. Furthermore, deuteriumꢀhydrogen exchange was not observed for perꢀdeuterated
substrates (29) giving some mechanistic insight. Finally, the installation of a trifluoromethyl group at
the βꢀcarbon was well tolerated, albeit with a slight impact on yield (32).
Mechanistic Delineation: Having established substrate scope and the decisive influence of the βꢀ
substituent in governing stereoselectivity, efforts were focused on experimental mechanistic
interrogation of this transformation. This study, together with the initial photocatalytic isomerization
ACS Paragon Plus Environment

Page 9 of 69
The Journal of Organic Chemistry
based on the cinnamoyl motif, provided a rich body of data from which to probe catalysis activation and
the influence of internal structural preorganization on the photostationary composition. As a working
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3
model, the diffusion controlled energy transfer mechanism summarized in Scheme 3 was considered.
Scheme 3 Postulated, diffusion controlled, collisional energy transfer mechanism for the photocatalytic
isomerization of activated alkenes with (–)ꢀriboflavin (RF): S(E) Eꢀalkene; S(Z) Zꢀalkene; S(T) twisted
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13
intermediate; [RFꢁꢁS]* exciplex.
Initiation of the photocatalytic isomerization reaction would occur by direct excitation of (–)ꢀriboflavin
to the first excited singlet state at 402 nm (eq. 1), opening up the possibility of intersystem crossing to
the corresponding triplet state (eq. 2). Energy transfer subsequently results from collisional interactions,
1/3
likely proceeding via exciplex formation, between the excited photosensitizer ( RF*) and the groundꢀ
state substrate S(E) (eq. 3). Immediate relaxation to the twisted excited alkene S(T) is described by the
reaction rates k , respectively (eq. 4). Deactivation of the twisted intermediate results in statistical
5
formation (k vs k ) of the ground state Eꢀ and Zꢀconfigured alkenes. The photostationary state
6
7
composition (E/Z ratio) would consequently be determined by the E → Z isomerization reaction rates
compared to the inverse process (k + k + k versus k + k + k ). This gives rise to the question of the
3
4
5
8
9
10
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 69
2
0
selectivity determining step. In our initial study, correlation of the triplet energy (E ) of the
T
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
6
photocatalyst with isomerization selectivity reinforced the working hypothesis that isomerꢀselective
energy transfer was operational (k >> k ). To explore the dependence on the alkene substitution pattern,
3
8
2
SternꢀVolmer fluorescence quenching experiments of (–)ꢀriboflavin with substrates 1 (R =H), 2
2
2
(
R =Me) and 3 (R =Et) were performed (Figure 1). Inspection of this plot, and comparison with the Z:E
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
selectivities reported in Table 1, shows that selectivity does not correlate with quenching efficiency (full
2
2
2
details in the SI). The following trend was noted: 2 (R = Me) > 1 (R = H) > 3 (R = Et).
Figure 1 SternꢀVolmer fluorescence quenching of (–)ꢀriboflavin using substrates 1, 2 and 3.
In contrast to many popular transitionꢀmetal photocatalysts (e.g. Ru/Ir), phosphorescence of (–)ꢀ
riboflavin is not directly observed upon excitation.3 Thus a more important observation from the Sternꢀ
Volmer experiment is the interaction between the singlet excited state of the photocatalyst (S ) and the
1
substrate. This is not consistent with initial proposal invoking only triplet energy transfer as the
activation mode.
Confronted with the possibility of mechanistic dichotomy, additional quenching experiments were
considered to elucidate the nature of energy transfer. Initially, azulene was investigated as a quencher in
a model transformation (2, Figure 2) on account of its ability to quench both the singlet and triplet
states. Its quenching ability is known to be susceptible to changes in solvent effects, particularly
ACS Paragon Plus Environment

Page 11 of 69
The Journal of Organic Chemistry
2
7
viscosity, making it a strategically useful tool in delineating photochemical mechanisms. It was
observed that the addition of azulene to the reaction in MeCN caused a severe decrease in
photostationary state Zꢀselectivity, and clear conclusions regarding the activation mode could not be
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
28
extrapolated. This may be due to the low singlet and triplet state energies of azulene.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OH
(_
)ꢀriboflavin
(
5 mol%)
HO
OH
OH
photoꢀ
catalyst,
azulene
CH3
CH3
H C
3
N
N
N
O
(
Z)
CN
NH
H3C
(
E)
MeCN,
rt, 402 nm
CN
O
E-2
2
azulene
Figure 2 Exploring solvent dependence on the relationship between photostationary state product
composition and azulene concentration.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
[
a]
Table 2. Investigating the oxygen effect on the photostationary state composition
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Time (h) Z:E
Conditions
Yield
quant.
quant.
ratio
Air
atmosphere
12
12
12
91:9
Argon
atmosphere
92:8
91:9
O atmosphere quant.
2
[
a]
Standard reactions conditions: Reactions were performed with 0.1 mmol substrate (E:Z >20:1) at
room temperature with UVꢀlight irradiation (402 nm) under the atmosphere indicated. Z:E ratios are the
average of two experiments and were determined by H NMR spectroscopy and HPLC analysis.
1
Switching to the comparably viscous solvent pivalonitrile, did little to influence the quenching rate.
When employing polar protic solvents such as methanol and tꢀbutanol, a significantly lower nonꢀlinear
quenching phenomenon was observed. This nonꢀlinear behavior in protic solvents may be attributable to
solvation of (–)ꢀriboflavin, and thus disruption of the wellꢀdescribed chargeꢀtransfer complexes that
29
form between azulene and flavins. The appearance of a clear viscosity effect in protic solvents is in
2
7
line with a diffusion controlled, collisional singlet energy transfer mediated reaction.
To interrogate the involvement of a triplet activation pathway further, the effect of oxygen on selectivity
was examined (Table 2 and Figure 3). Oxygen is a highly efficient triplet quencher and thus the rigorous
30
presence or absence of oxygen would allow the question of a triplet pathway to be addressed.
Comparison of reactions performed under argon, oxygen and air atmospheres are summarized in Table
. Intriguingly, the reaction selectivity proved to be oxygen independent, reinforcing the SternꢀVolmer
2
analysis such that our mechanistic interpretation gravitated towards the notion of mechanistic (spin
ACS Paragon Plus Environment

Page 13 of 69
The Journal of Organic Chemistry
multiplicity) dichotomy. Whilst the selectivity of catalysis was unaffected, the efficiency of the
transformation was inextricably linked.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
OH
(_
)ꢀriboflavin
CH3
CH
3
(
HO
(5 mol%)
photoꢀ
catalyst
OH
OH
Z)
CN
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H C
3
N
N
N
O
(
E)
MeCN,
rt, 402 nm,
O or Argon
CN
NH
3
H C
2
O
Figure 3 Oxygen effect on reaction rate of the isomerisation of 2.
Rigorous exclusion of oxygen (Ar atmosphere) resulted in almost quantitative formation of the desired
Zꢀisomer after only 2 h. In contrast, the oxygen saturated solution significantly inhibits productive
isomerization with only ca. 50% conversion observed after 2 h (Fig. 3). This observation may explain
the absence of an oxygen effect on the photostationary state after 12 h.
Overall, these quenching studies converge on two conceivable activation pathways: The first implicates
singlet energy transfer from (–)ꢀriboflavin to the starting material to generate the singlet excited alkene.
Following efficient intersystem crossing (ISC) based on spinꢀorbit coupling with the nitrile nonꢀbonding
orbitals to generate the triplet state. A second scenario invokes both singlet and triplet energy transfer
from (–)ꢀriboflavin to furnish a mixture of triplet and singlet excited state alkene. Since the singlet
alkene would likely undergo intersystem crossing to the corresponding triplet alkene, a common
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 69
intermediate would derive from both excitation pathways. This finding renders the isomerization of
activated cinnamoyl and cinnamonitrile derived alkenes mechanistically distinct from those of the
electron rich systems studied in the midꢀ1960s. Furthermore, this series of quenching experiments
eliminated the possibility that the initial excitation event from the photocatalyst to the alkene was
selectivity determining (Figure 1). Initially, it was postulated that the directionality of catalysis would
have a structural basis, and that the avoidance of nonꢀbonding interactions would dominate selectivity.
To support this supposition, and explore a likely correlation between the βꢀsubstituent and the Zꢀisomer
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
molar fraction, the Taft steric parameter was considered as a descriptor (Figure 4). Since the activated
alkenes explored in this study (EWG=CN) behaved in an analogous manner to the cinnamoyl containing
esters and amides reported in the initial communication, a set of distinct substrate classes was employed
in this analysis. In the case of the cinnamonitriles, the influence of the paraꢀsubstituent was also
addressed by comparing pꢀH with pꢀCF and pꢀOMe. To probe a tentative correlation between steric
3
2
size and Zꢀselectivity in catalysis, a series of subꢀstructures were prepared with varying R substituents
and are indicated as follows: a = βꢀH, b = βꢀMe, c = βꢀEt, d = βꢀnPr, e = βꢀnBu (Figure 4, A and B). A
plot of the respective Taft parameter versus photostationary state Zꢀisomer molar fraction revealed a
2
linear correlation for R = H (1), Me (2) and Et (3) (Figure 4, upper left; A). At this point, the Zꢀ
selectivity plateaued, as previously noted during the establishment of catalysis scope. This trend was
2
found to persist in all of the substrate groups studied when R = Et. Figure 4B shows the linear area of
the Taft parameter / photostationary state Zꢀisomer molar fraction correlation.
To explore the structural origins of isomerization selectivity further, computational analysis of the
3
2
substrate in MeCN was performed (Figure 4 C). From this theoretical study, it was possible to
2
extrapolate the C ꢀC ꢀC ꢀC torsion angle and, by extension the extent to which the R substituent twists
1
2
3
4
and deconjugates the πꢀsystem prior to photocatalyst activation. Again, this analysis was performed with
the substrate classes described above and showed a clear and consistent trend linking torsion angle and
1
,3
2
Taft steric parameter. This may be considered as the manifestation of 1,3ꢀallylic strain (A ) between R
and the orthoꢀproton of the aryl ring.
ACS Paragon Plus Environment

Page 15 of 69
The Journal of Organic Chemistry
Full computational details of the computational analysis are provided in the supporting information.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
Expectedly, an increase in the arylꢀalkene torsion angle resulted from systematically replacing R = H
by Me and then Et. As observed in both catalysis experiments and also in the Taft correlation, the Et
2
2
substituent appears to be a plateau and further size augmenting of the R had little effect (R = Et, nPr,
nBu).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Finally, a correlation of the torsion angle and the photostationary state Zꢀisomer molar fraction was
considered (Figure 4 D). This revealed a direct linear relationship over all of the βꢀsubstituted alkenes (2
2
0
– 6, 12 – 15, 17 – 20, 44 – 47 and esters). These data indicate that the conformation of the ground state
starting material influences the photostationary state composition. This notion is further reinforced by
consideration of starting materials that lack this conformational constraint. Contrastingly, these
βꢀ
2
unsubstituted substrates (1, 11, 16, 43, and methyl cinnamate, R = H) exhibit very different
photostationary state compositions (Z:E ratios), whilst having comparable torsional angles (for full
details see SI). General inspection of the computed Zꢀisomer conformations revealed that the
unsubstituted products have planar πꢀsystems and can thus be reꢀexcited by the photocatalyst. The
β
β
ꢀ
ꢀ
substituted substrates, on the other hand, are fully deconjugated (torsional angle 33ꢀ55°) and thus cannot
undergo photocatalyst activation (for full details see SI). This subtle structural effect is thus a major
determinant in endowing the process with an intrinsic directionality and is ultimately responsible for
high Zꢀselectivities observed with activated alkenes. Whilst the origin of selectivity can be broadly
2
interpreted at the structural level, R = Et is conspicuous in all of these analyses and remains an
anomaly. Evidently, a simple steric model based on ground state conformation is insufficient to account
for selectivity and it is likely that an interplay of several factors is operational.
Cognizant of the intrinsic flexibility of the substrates in this study, a series of bicyclic probes were
synthesized to minimize conformational freedom (Table 3; 33ꢀ36). It was envisaged that this may be
addressed by simply preꢀorganizing the chromophore through an alkyl tether. This would
simultaneously preserve the planarity of the πꢀsystem and mimic the steric footprint of the Et
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 69
substituent. It would also, however, subtly contract the C ꢀC ꢀC ꢀC torsional angle in the product (5ꢀ
1
2
3
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ring 33 > 6ꢀring 34 > 7ꢀring 35 > 8ꢀring 36) thereby forcing the πꢀsystem to deconjugate, and thus
augmenting unfavorable nonꢀbonding interactions (Table 3). The selectivities summarized in Table 3
proved to be instructive in confirming the importance of this interaction in ensuring directionality.
Whilst low to mediocre selectivities were measured for 33 and 34 (Z:E 46:54 and 66:34, respectively), a
substantial increase was noted for the sevenꢀ and eightꢀ membered ring containing substrates leading to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
5 and 36 (Z:E 97:3 and 99:1, respectively). Based on these empirical observations, two possibilities
appeared reasonable to account for the selectivity in the acyclic cases summarized in Table 1, and
required further exploratory studies: i) The first is the assumption that deconjugation of the alkene and
the phenyl ring must occur in the excited state. This is reminiscent of a “hulaꢀtwist”ꢀtype isomerization
3
3
mechanism, which would require rotation of the aromatic ring during the isomerization event (see
Scheme 3, k and k ), resulting in higher excited state activation barriers for less preꢀorganized systems
4
8
such as 1 (torsion angle, φ = 0°), or ii) As proposed by Hammond and coꢀworkers for stilbene
isomerization, substitution could potentially yield a “deeper”, more stabilized excited state with higher
13c,34
lifetimes and therefore more productive transitions from the Eꢀ to Zꢀisomer.
ACS Paragon Plus Environment

Page 17 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4 (A) Correlation of Taft parameter and photostationary state Zꢀisomer molar fraction,
(
B) Linear area of Taft parameter and photostationary state Zꢀisomer molar fraction correlation,
C) Correlation of Taft parameter and calculated arylꢀalkene torsional angle, (D) Correlation of
(
calculated arylꢀalkene torsional angle and photostationary state Zꢀisomer molar fraction of
substituted alkenes. (a = ꢀH, b = ꢀMe, c = ꢀEt, d = ꢀnPr, e = ꢀnBu)
βꢀ
β
β
β
β
β
It was reasoned that since rotational barriers are temperature dependent, activation parameters
derived from an Eyring analysis might provide an insight into the excited state rotational
processes. Ultimately, this may lead to a fuller understanding of the origins of selectivity. An
2
Eyring analysis of the isomerization of various substituted α,βꢀunsaturated nitriles (R = H, Me,
Et, nPr, nBu, 1ꢀ5) was therefore attempted. To that end, the catalytic isomerization was
performed at various temperatures and monitored by HPLCꢀUV under internal standard
calibration. Unfortunately, the kinetic data extrapolated showed a complex reaction order, and
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
thus exact reaction rates and activation parameters could not be derived. As might be expected, a
significant enhancement in the reaction rate was observed with increasing temperature. This
implicates longer excited state lifetimes and faster state and rotational transitions (see SI for full
details). Approaching this same issue from a slightly different perspective led us to determine the
effect of temperature on the photostationary state composition. This, in turn, would allow the
existence of activation barriers in the isomerization process to be derived. All reactions were
performed at 30 °C and 50 °C, and the photostationary state product composition was analyzed
by HPLCꢀUV (Table 4).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As summarized in Table 4, the photostationary state composition of the unsubstituted alkene 1 is
slightly enriched with Zꢀisomer at 50 °C compared to 30 °C, suggesting an activation barrier for
the transition from Eꢀ1 → Zꢀ1. This observation is fully in line with Hammond and coꢀworkers
13c
mechanistic investigation on electron rich stilbene isomerization.
Interestingly, for the
2
polarized alkenes bearing a substituent at R , the quotient is notably reduced on account of the
increased Eꢀisomer molar fraction at higher temperatures.
This contrasting behavior indicates a higher activation barrier for the transition from the Zꢀ → Eꢀ
isomer. Considering the torsional angle range between φ = 40.7° and 43.5° for 2, 3, 4 and 5,
respectively, an inꢀplane rotation of the aromatic ring is necessary to achieve conjugation and
consequently allow catalyst excitation of the Zꢀisomer. Conversely, the unsubstituted Zꢀisomer
2
(R = H) is fully conjugated (φ = 0 °C) and thus the rotational barrier for the Eꢀ to Zꢀisomer
transition must proceed via an excited state process.
ACS Paragon Plus Environment
18

Page 19 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
[
a]
Table 3 The (–)ꢀriboflavinꢀcatalyzed isomerization of cyclic α,βꢀunsaturated nitriles
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[
a]
Standard reaction conditions: Reactions were performed with 0.1 mmol substrate (E:Z
20:1) at ambient temperature in MeCN using 5 mol% catalyst loading under 12 hours of UVꢀ
>
light irradiation (402 nm). Z:E ratios are the average of two experiments and were determined by
H NMR spectroscopy analysis.
1
[
a]
Table 4 Temperature effect on photostationary state composition
2
R
30 °C
50 °C
[E/Z]30°C
Z:E
Z:E
[
E/Z]50°C
1.137
0.873
0.788
0.820
0.816
H (1)
53.9 : 46.1 57.1 : 42.9
Me (2) 88.1 : 11.9 86.6 : 13.4
Et (3)
95.2 : 4.8
93.9 : 6.1
92.7 : 7.3
92.7 : 7.3
nPr (4) 93.9 : 6.1
nBu (5) 94.0 : 6.0
ACS Paragon Plus Environment
19

The Journal of Organic Chemistry
Page 20 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
[a]
Standard reaction conditions: Reactions were performed with 0.1 mmol substrate (E:Z
>
20:1) at the specified temperature under 12 hours of UVꢀlight irradiation (402 nm) and an air
atmosphere. Z:E ratios are the average of two experiments and were determined by HPLC
analysis.
A correlation of the ratio of temperature dependent photostationary states, reflecting the
activation barriers for isomerization, with the photostationary state composition at room
temperature showed a linear relationship (Figure 5). This observation, which is distinct from the
mechanistic nuances of stilbene isomerization, demonstrates that the selectivity and
directionality of isomerization catalysis has a structural origin. Ultimately, the presence or
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
absence of R is significant in determining the activation barrier for the forward and reverse
transformations, and in turn in governing the E/Z selectivity quotient.
Figure 5 Correlation of ratio of photostationary state composition at 30 °C and 50 °C with
photostationary state composition at room temperature; ; a =
e = ꢀnBu.
βꢀH, b = βꢀMe, c = βꢀEt, d = βꢀnPr,
β
Whilst it seems reasonable that increased activation barriers for the transition from Zꢀ to Eꢀ
isomer should augment the photostationary state Zꢀisomer molar fraction, the torsion angle data
ACS Paragon Plus Environment
20

Page 21 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
2
does not satisfactorily account for deviation in the selectivities observed for R = Et (3), nPr (4)
and nBu (5). It is therefore likely that a secondary temperature dependent effect, such as excited
state stability, must be considered to fully describe the origin of selectivity. Consequently, the
second hypothesis was considered. The previous reports by Saltiel et al. reported notably
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
enhanced lifetimes of excited state stilbene following substitution of one C(sp )ꢀH of the alkene
3
4
by a methyl group. It logically follows that a direct consequence of extended lifetimes would
be a more productive Eꢀ to Zꢀisomerization.
Since the steric effect of the substituent alone could be discarded based on previous
experiments, the electronic contribution of substitution on the alkene was considered. It was
1
3
envisaged that comparison of the Cꢀshifts of the alkene carbon atoms would give direct
quantification of electronic changes (Figure 6). Moreover, since conformational changes and
orbital hybridization (cf. Figure 4C) also manifest themselves in NMR shift behavior, this
3
5
13
spectroscopic approach was appealing. Gratifyingly, a direct linear correlation between the C
shift difference (ꢀδ [ppm]) of the βꢀ and αꢀcarbon atoms and the logarithm of the Z/E ratio was
observed (Figure 6).
ACS Paragon Plus Environment
21

The Journal of Organic Chemistry
Page 22 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
13
Figure 6 Correlation of C shift
β
ꢀα
carbon shift difference ꢀδ with photostationary state Z/E
1
3
selectivity ( C shifts were determined in CDCl ; a = ꢀH, b = ꢀMe, c = ꢀEt, d = ꢀnPr, e =
β
β
β
β
3
β
ꢀnBu, f = βꢀiBu).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
These data indicate that selectivity increases directly with the electron deficiency of the
β
ꢀcarbon
(
i.e. polarization influences selectivity). An intrinsic consequence of lowering electron density on
the βꢀcarbon atom is to enhance polarization of the alkene unit. Considering the zwitterionic
contribution to the excited state, stabilization by polarization of the double bond would explain
the prolonged lifetimes observed for methyl substituted stilbenes and can be logically extended
to the cinnamoyl scaffold. Furthermore, polarization of the alkene by decreasing
πꢀconjugation
and substituent effects logically results in an increased bond length and thus lower barrier for
3
5a,b
rotation around the central carbonꢀcarbon double bond.
A selectivity model invoking two
aspects is therefore postulated and includes: (i) alkene polarization, resulting in extended excited
state alkene lifetimes and lower rotational barriers for Eꢀ to Zꢀisomerization processes, and (ii)
increased activation barriers for the inverse Zꢀ to Eꢀisomer transitions.
To explore the implications of utilizing the arylꢀalkene torsional angle to enhance alkene
polarization, and thereby increase Zꢀselectivity, the effect of orthoꢀsubstitution was explored.
Halogenation provides a convenient handle by systematically increasing steric congestion but
maintaining the electronic signature of the substrate. A series of structurally modified Eꢀisomers
were synthesized (37ꢀ41) and exposed to the standard reaction conditions (Table 5).
A clear trend linking Zꢀselectivity with increasing halogen size was noted. Whilst introduction of
an orthoꢀfluorine substituent resulted in a slight decrease in selectivity (37, 56:44), chlorine 38,
bromine 39 and iodine 40 substitution led to a significant increase in Zꢀselectivity (68:32, 72:28
ACS Paragon Plus Environment
22

Page 23 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
and 74:26, respectively). Combining effects by adding a
β
ꢀmethyl substituent in the case of 41
furnished a Z:E ratio of 99:1 (cf. 91:9 for 2).
[a]
Table 5. Effect of orthoꢀsubstitution on Zꢀisomer selectivity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[
a]
Standard reaction conditions: Reactions were performed with 0.1 mmol substrate (E:Z
>
20:1) at ambient temperature in MeCN using 5 mol% catalyst loading under 12 hours of UVꢀ
light irradiation (402 nm). Z:E ratios are the average of two experiments and were determined by
H NMR spectroscopy analysis.
1
Conclusions
The ubiquity of Zꢀselective alkene isomerization in Nature belies the intrinsic challenges of
replicating this process in a catalysis paradigm. Consequently, a general solution to this
rudimentary process remains visibly absent from the catalysis arsenal. Close inspection of a
seminal report describing the qualitative isomerization of retinal in fish and mammals by (–)ꢀ
riboflavin allowed crucial data to be distilled and translated to address this conspicuous absence
in the organocatalysis armory.
ACS Paragon Plus Environment
23

The Journal of Organic Chemistry
Page 24 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7 Photochemical activation and origin of selectivity in the (–)ꢀriboflavin catalyzed
isomerization of polarized alkenes.
Reinterpretation of the biological photosystem facilitated the development of a highly Zꢀ
selective, catalytic isomerization of activated alkenes based on the cinnamoyl and cinnamonitrile
chromophores. By replacing retinal by a truncated chromophore, the degree of πꢀconjugation in
the starting material and product is determined by the alkene geometry and the avoidance of nonꢀ
bonding interactions; this results in a selective excitation manifold whereby only the substrate
and photocatalyst undergo efficient energy transfer: This ensured directionality in the reaction
design by circumventing microscopic reversibility and the inevitable product distributions that
result under thermal / nonꢀirradiative conditions.
The optimized transformation proved to be highly efficient (Z:E up to 99:1), quantitative and
operationally simple requiring only filtration. It also validated the synthetic utility of (–)ꢀ
riboflavin as a powerful antenna for photocatalytic transformations. Moreover, in the case of
cinnamonitriles, the masked amine mitigates the risk of physical and chemical quenching that
often inhibits photochemical processes involving amines. In addition to expanding substrate
ACS Paragon Plus Environment
24

Page 25 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
scope, this study intended to delineate both the photocatalyst activation mechanism and the
origin of selectivity for a range of substrate classes. A combination of SternꢀVolmer quenching
experiments and secondary external quencher effects were informative in this regard. Evidence
of mechanistic dichotomy has been demonstrated whereby both singlet and triplet energy
intermediates are operational and likely converge on a common species. The scenario invoking
exclusive singlet energy transfer to the alkene followed by ISC to the triplet state cannot be fully
discounted. In contrast to previous studies detailing the stoichiometric isomerization of electron
rich alkenes, this complexity in substrate activation arises from the polarized nature of the
alkene. It is likely that the nonꢀbonding orbitals of the electron withdrawing carbonyl and nitrile
units facilitate intersystem crossing (El Sayed’s Rule) thus rendering catalysis efficient. The
polarized nature of the alkene unit is also intimately linked to the origin of selectivity.
Conformational analysis of the Eꢀ and Zꢀisomers demonstrated a clear effect of arylꢀalkene
torsional preorganization on the photostationary composition. Kinetic studies, temperature
dependent photostationary state correlations and investigation of substituentꢀbased electronic
perturbation of the alkene identified polarization combined with increased Zꢀisomer activation
barriers as the selectivity governing factors in catalysis (Figure 7).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
All chemicals were purchased as reagent grade and used without further purification. Solvents
for purification (extraction and chromatography) were purchased as technical grade and distilled
on the rotary evaporator prior to use. For column chromatography SiO (40ꢀ63 ꢁm for Flashꢀ
2
Chromatography, VWR Chemicals) was used as stationary phase. Analytical thin layer
chromatography (TLC) was performed on aluminum foil preꢀcoated with SiO ꢀ60 F254 (Merck)
2
and visualized with a UVꢀlamp (254 nm) and KMnO solution. Concentration in vacuo was
4
ꢀ2
performed at ~10 mbar and 45 °C, drying at ~10 mbar and room temperature. NMR spectra
were measured by the NMR service of the OrganischꢀChemisches Institut, Westfälische
WilhelmsꢀUniversität Münster on a Bruker AV300, Agilent DD2 500 or an Agilent DD2 600
spectrometer at room temperature. The chemical shifts are referenced to the residual solvent peak
as internal standard. The resonance multiplicity is abbreviated as: s (singlet), d (doublet), t
(triplet), q (quadruplet), quint (quintet), sext (sextet), sep (septet), m (multiplet) and b (broad).
ACS Paragon Plus Environment
25

The Journal of Organic Chemistry
Page 26 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Assignments of unknown compounds are based on DEPT, COSY (HH and FF), HMBC, HSQC
and NOESY spectra. Alkene configuration is assigned based on coupling constants and NOESY
spectra. Melting points were measured on a Büchi Bꢀ545 meltingꢀpoint apparatus in open
capillaries. IR spectra were recorded on a PerkinꢀElmer 100 FTꢀIR spectrometer, selected
–1
adsorption bands are reported in wavenumbers (cm ) and intensities are reported as: w (weak),
m (medium), s (strong) and b (broad). Highꢀresolution mass spectra (HR ESI with
microTOF/orbitrap or Quad detection) were measured by the MS service of the Organischꢀ
Chemisches Institut, Westfälische WilhelmsꢀUniversität Münster. Isomerization reactions were
performed utilising a ACULED VHL (ACL01ꢀSCꢀUUUUꢀE05ꢀC01ꢀLꢀ0000) by LED Solutions
(emission spectrum see SI Figure S1). The forward current per chip was set to 350 mA, the
resulting forward voltage was 14 V while the resulting radiant flux was 1150 mW. Further
isomerization reactions were performed utilizing a setꢀup of 4 Winger WEPUV3ꢀS2 UV Power
LED Star (Schwarzlicht) 1.2 W lamps (emission spectrum see SI Figure S2). The forward
current per chip was set to 700 mA, the resulting forward voltage was 3.4 V while the resulting
radiant flux was 1200 mW. The distance between the reaction vessels and the UVꢀlamp was set
at approximately 0.5 cm for all reactions.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Preparation of the starting materials
(E)ꢀ3ꢀphenylacryl acid, (E)ꢀ3ꢀphenylbutꢀ2ꢀenoic acid, (E)ꢀ3ꢀphenylpentꢀ2ꢀenoic acid, (E)ꢀ3ꢀ
phenylhexꢀ2ꢀenoic acid, ethyl cinnamate (E-53), ethyl (E)ꢀ3ꢀphenylbutꢀ2ꢀenoate (E-54), ethyl
(E)ꢀ3ꢀphenylpentꢀ2ꢀenoate (E-55), ethyl (E)ꢀ3ꢀphenylhexꢀ2ꢀenoate (E-55), ethyl (Z)ꢀ
phenylacrylate (53), ethyl (Z)ꢀ3ꢀphenylbutꢀ2ꢀenoate (54), ethyl (Z)ꢀ3ꢀphenylpentꢀ2ꢀenoate (55)
and ethyl (Z)ꢀ3ꢀphenylhexꢀ2ꢀenoate (56) were synthesized according to our previous
20,21
procedure.
Synthesis of N-Methoxy-N-methyl-4-(trifluoromethyl)benzamide (48) 4ꢀTrifluoromethylꢀ
benzoic acid (2.30 g, 9.85 mmol, 1.0 eq.) was dissolved in DMF (100 mL) before N,Oꢀ
dimethylhydroxylamine (1.00 g, 10.4 mmol, 1.1 eq.), EDCI·HCl (1.99 g, 10.4 mmol, 1.1 eq.)
and Et N (1.45 mL, 10.4 mmol, 1.1 eq.) were added. The resulting white suspension was stirred
3
for 18 h at room temperature, the DMF was removed under high vacuum and the resulting
residue was dissolved in EtOAc (30 mL), washed with sat. NaHCO solution (aq., 20 mL) and
3
brine (20 mL), before the organic phase was dried over MgSO . The solvent was removed in
4
vacuum. Purification by column chromatography (SiO , 25% EtOAc/ cyclohexane) yielded the
2
1
product 48 as a colourless oil (1.15 g, 50%). R = 0.42 (EtOAc/cyclohexane 2:3); H NMR (400
f
MHz, CDCl ): δ = 7.77 (d, J = 7.8 Hz, 2H; H5), 7.65 (d, J = 8.0 Hz, 2H; H6), 3.51 (s, 3H; H1),
3
1
3
3.36 (s, 3H; H2) ppm; C NMR (101 MHz, CDCl ): δ = 168.6 (C3), 137.7 (q, J = 1.2 Hz; C4),
3 CF
1
32.3 (q, J = 32.5 Hz; C7), 128.6 (C5), 125.1 (q, J = 3.8 Hz; C6), 123.8 (q, J = 272.4 Hz;
CF
CF
CF
1
9
C8), 61.3 (C1), 33.4 (C2) ppm; F NMR (282 MHz, CDCl ): δ = ꢀ63.00 ppm; IR (ATR):
3
ῦ = 2966(w), 2941(w), 2352(w), 1645(m), 1620(m), 1578(w), 1516(w), 1462(w), 1407(w),
1
383(w), 1322(s). 1260(w), 1215(w), 1165(m), 1122(s), 1111(s), 1069(s), 1019(s), 980(w),
ꢀ1
889(w), 851(m), 797(m), 779(m), 764(m), 701(m), 662(w) cm ; GCꢀEIꢀMS (Quad): m/z: 233.2
3
6
([M], calcd. for C H F NO : 233.1902); analytical data in agreement with literature.
10 10 3 2
ACS Paragon Plus Environment
26

Page 27 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Synthesis of 1-(4-(Trifluoromethyl)phenyl)butan-1-one (49) 48 (412 mg, 1.8 mmol, 1.0 eq.)
was dissolved in dry THF (10 mL) under Argon, the solution was cooled to 0 °C before
nPrMgCl (2.7 mL, 5.4 mmol, 3.0 eq.) was added dropwise. The mixture was gradually warmed
to room temperature over 3 h and the resulting suspension was added to 1 N HCl (30 mL). The
aqueous phase was extracted with EtOAc (3x 15 mL), the combined organic phases were dried
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
over MgSO and the solvent was removed in vacuo. Purification by column chromatography
4
(
SiO , 20% EtOAc/ cyclohexane) yielded the product 49 as a colourless oil (342 mg, 88%).
2
1
R = 0.78 (EtOAc/cyclohexane 1:4); H NMR (400 MHz, CDCl ): δ = 8.15 – 7.96 (m, 2H; H6),
f
3
7
.79 – 7.69 (m, 2H; H7), 2.97 (t, J = 7.3 Hz, 2H; H3), 1.79 (sext, J = 7.4 Hz, 2H; H2), 1.02 (t,
13
J = 7.4 Hz, 3H; H1) ppm; C NMR (101 MHz, CDCl ): δ = 199.5 (C4), 139.9 (C5), 134.4 (q,
3
J = 32.6 Hz; C8), 128.5 (C6), 125.8 (q, J = 3.8 Hz; C7), 123.8 (q, J = 272.7 Hz; C9), 40.9
CF
CF
CF
19
(
C3), 17.7 (C2), 13.9 (C1) ppm; F NMR (282 MHz, CDCl ): δ = ꢀ63.10 ppm; IR (ATR):
3
ῦ = 2963(m), 1698(w) 1513(w), 1410(w), 1326(m), 1259(s), 1210(w), 1172(w), 1066(s),
ꢀ
1
1
(
015(s), 864(m), 796(s), 703(w), 662(w) cm ; HRꢀESIꢀMS (MicroTOF): m/z: 239.0655
+
+
37
[M+Na] , calcd. for C H F ONa : 239.0660); analytical data in agreement with literature.
11 11 3
Synthesis of 1-(4-(Trifluoromethyl)phenyl)pentan-1-one (50) 48 (466 mg, 2 mmol, 1.0 eq.)
was dissolved in dry THF (10 ml) under Argon, the solution was cooled to 0 °C before nBuMgCl
(3 mL, 6 mmol, 3.0 eq.) was added dropwise. The mixture was gradually warmed to rt over 3 h
and the resulting suspension was added to 1 N HCl (30 mL). The aqueous phase was extracted
with EtOAc (3x 15 mL), the combined organic phases were dried over MgSO and the solvent
4
was removed in vacuo. Purification by column chromatography (SiO , 20% EtOAc/
2
cyclohexane) yielded the product 50 as a colourless oil (395 mg, 86%).R = 0.78
f
1
(
EtOAc/cyclohexane 1:4); H NMR (400 MHz, CDCl ): δ = 8.06 (dq, J = 7.9, 0.9 Hz, 2H; H7),
3
7
.73 (m, 2H; H8), 3.09 – 2.88 (t, J = 7.4 Hz, 2H; H4), 1.73 (p, J = 7.4 Hz, 2H; H3), 1.42 (tq,
13
J = 14.8, 7.3 Hz, 2H; H2), 0.96 (t, J = 7.3 Hz, 3H; H1) ppm; C NMR (101 MHz,
CDCl ): δ = 199.6 (C5), 139.9 (C6), 134.4 (q, J = 32.6 Hz; C9), 128.5 (C7), 125.8 (q,
3
CF
J = 3.8 Hz; C8), 123.8 (q, J = 272.6 Hz; C10), 38.8 (C4), 26.4 (C3), 22.6 (C2), 14.1 (C1)
CF
CF
19
ppm; F NMR (282 MHz, CDCl ): δ = ꢀ63.09 ppm; IR (ATR): ῦ = 2980(w), 2965(w), 2934(w),
3
2863(w), 1680(s), 1580(w), 1511(w), 1468(w), 1411(m), 1384(w), 1320(s), 1308(s), 1269(m),
1203(m), 1165(s), 1126(s), 1110(s), 1015(m), 979(m), 924(w), 904(w), 852(s), 799(m), 756(m),
ꢀ
1
+
7
26(m), 676(w) cm ; HRꢀESIꢀMS (MicroTOF): m/z: 253.0819 ([M+Na] , calcd. for
+
38
C H F ONa : 253.0816); analytical data in agreement with literature.
12
13 3
3
9
Synthesis of 7,8,9,10-Tetrahydrobenzo[8]annulen-5(6H)-one (51) 6ꢀPhenylhexanoic acid
500 mg, 2.6 mmol, 1.0 eq.) was dissolved in DCM (10 ml) and the solution was cooled to 0 °C
(
before oxalyl chloride (0.446 mL, 5.2 mmol, 2.0 eq.) was added. The resulting mixture was
gradually warmed to rt and stirred for 14 h at this temperature before being concentrated in
vacuo. The residue was dissolved in DCM (25 mL) under Argon and the solution was added to a
suspension of AlCl (1386 mg, 10.4 mmol, 4.0 eq.) in dry DCM (125 mL) under Argon over 3 h.
3
The reaction mixture was stirred for 17.5 h, washed with sat. NH Cl solution (aq., 50 mL) and
4
ACS Paragon Plus Environment
27

The Journal of Organic Chemistry
Page 28 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
dried over MgSO before being concentrated in vacuo. Purification by column chromatography
4
(
SiO , 10% EtOAc/ cyclohexane) yielded the product 51 as a colourless oil (340 mg, 75%).
2
1
R = 0.35 (EtOAc/cyclohexane 1:9); H NMR (400 MHz, CDCl ): δ = 7.65 (dd, J = 7.7, 1.5 Hz,
f
3
1H; H8), 7.39 (td, J = 7.5, 1.5 Hz, 1H, H10), 7.27 (td, J = 7.6, 1.4 Hz, 1H; H9), 7.18 (dd,
J = 7.7, 1.2 Hz, 1H; H11), 3.08 – 3.00 (m, 2H; H1), 2.96 – 2.89 (m, 2H; H5), 1.90 – 1.75 (m, 4H;
1
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H2/4), 1.57 – 1.49 (m, 2H; H3) ppm; C NMR (101 MHz, CDCl ): δ = 207.7 (C6), 140.1 (C12),
3
140.0 (C7), 131.6 (C10), 131.1 (C11), 127.7 (C8), 126.5 (C9), 44.3 (C5), 34.5 (C1), 28.0 (C2),
25.3 (C3), 24.2 (C4) ppm; IR (ATR): ῦ = 2928(m), 2854(w), 1664(s), 1597(m), 1485(w),
1
1
485(w), 1444(m), 1344(w), 1328(w), 1305(m), 1287(m), 1255(s), 1185(m), 1158(m), 1132(m),
115(m), 1040(m), 999(m), 997(m), 947(w), 914(w), 896(w), 787(m), 774(m), 751(s), 716(w),
ꢀ1
+
+
673(m) cm ; HRꢀESIꢀMS (MicroTOF): m/z: 197.0943 ([M+Na] , calcd. for C H ONa :
1
2
14
197.0937).
(E)-3-Phenylhept-2-enoic acid (52) 42 (400 mg, 1.72 mmol, 1.0 eq.) was dissolved in NaOHꢀ
solution (3.44 mL, aq. 1.0 N, 3.44 mmol, 2.0 eq.) and EtOH (2.5 mL). The resulting mixture was
stirred at room temperature for 24 h before the pH was adjusted to pH 1 with HCl (aq., 1.0 M).
The aqueous phase was extracted with EtOAc (3x 5 mL), the combined organic phases were
dried over MgSO4 and the solvent was removed in vacuo. Purification by column
chromatography (SiO , 10 – 30% Et O/nꢀpentane) yielded the product 52 as a white amorphous
2
2
1
solid (93 mg, 27%). R = 0.1 (Et O/nꢀpentane 1:9); M.p.: 52.5 – 53.4 °C; H NMR (600 MHz,
f
2
CDCl ): δ = 7.47 – 7.43 (m, 2H; H9), 7.41 – 7.37 (m, 3H; H10/11), 6.06 (d, J = 0.5 Hz, 1H; H2),
3
3
.14 – 3.11 (m, 2H; H4), 1.45 – 1.40 (m, 2H; H5), 1.36 (ddd, J = 14.8, 7.4, 1.3 Hz, 2H; H6), 0.88
1
3
(t, J = 7.2 Hz, 3H; H7) ppm; C NMR (151 MHz, CDCl ): δ = 170.3 (C1), 164.0 (C3), 141.4
3
(
(
C8), 129.3 (C11), 128.7 (C10), 126.9 (C9), 116.2 (C2), 31.3 (C5), 31.1 (C4), 22.9 (C6), 14.0
C7) ppm; IR (ATR): ῦ = 3059(w), 2954(m), 2927(m), 2870(m), 2751(w), 2690(w), 2600(w),
1683(s), 1612(s), 1576(m), 1497(w), 1464(w), 1448(m), 1407(m), 1378(w), 1349(w), 1304(m),
1
279(m), 1250(w), 1211(s), 1162(w), 1119(w), 1107(w), 1081(W), 1029(w), 1000(w), 987(w),
ꢀ1
924(m), 871(m), 832(w), 767(m), 722(m), 703(s), 693(s) cm ; HRꢀESIꢀMS (MicroTOF): m/z:
+
+
227.1045 ([M+Na] , calcd. for C H O Na : 227.1043); analytical data in agreement with
13 16 2
40
literature.
General Procedure A for the preparation of α,β-unsaturated carbonyl compounds
NaH (60% in mineral oil, 1.1 eq.) was dissolved in dry THF, the mixture was cooled to 0 °C,
diethylcyanomethylphosphonoacetate (1.05 eq.) was added dropwise to the solution before the
resulting solution was stirred for 1 h at 0 °C. The specified aromatic ketone (1.0 eq.) was added
to the reaction mixture, before the mixture was stirred for 1 h at 0 °C and subsequently being
gradually warmed to room temperature. The resulting mixture was stirred until full conversion
(by TLC) or for 24 h before water and Et O were added. The organic phase was separated, the
2
aqueous phase was extracted with Et O (3x), the combined organic phases were dried over
2
MgSO , concentrated in vacuo and purified by column chromatography (SiO , specified
4
2
combination of Et O/nꢀpentane).
2
ACS Paragon Plus Environment
28

Page 29 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
(
E)-3-Phenylbut-2-enenitrile (E-2) Prepared according to General Procedure A, acetophenone
(5 mmol) was converted to Eꢀ2 in 24 h yielding a colourless oil (521 mg, 73%) after purification
1
by column chromatography (SiO , 1 ꢀ 5 % Et O/nꢀpentane). R = 0.39 (Et O/nꢀpentane 1:9); H
NMR (600 MHz, CDCl ): δ = 7.48 – 7.45 (m, 2H; H6), 7.43 – 7.40 (m, 3H; H7/8), 5.62 (q,
J = 1.1 Hz, 1H; H2), 2.47 (d, J = 1.1 Hz, 3H; H4) ppm; C NMR (151 MHz, CDCl ): δ = 159.9
2
2
f
2
3
1
3
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
C3), 138.4 (C5), 130.4 (C8), 128.9 (2C; C7), 126.0 (2C; C6), 117.7 (C1), 95.7 (C2), 20.3 (C4)
ppm; IR (ATR): ῦ = 3062(w), 2981(w), 2213(s), 2116(w), 1991(w), 1608(m), 1575(m), 1496(w),
1446(m), 1381(m), 1333(w), 1261(w), 1159(w), 1079(w), 1027(w), 1001(w), 921(w), 826(w),
7
ꢀ1
+
56(s), 691(s), 658(w) cm ; HRꢀESIꢀMS (MicroTOF): m/z: 166.0648 ([M+Na] , calcd. for
+
41
C H NNa : 166.0627); analytical data in agreement with literature.
10
9
4
2
(E)-3-Phenylpent-2-enenitrile (E-3)
propiophenone (4 mmol) was converted to E-3 in 24 h yielding a colourless oil (314 mg, 50%)
after purification by column chromatography (SiO , 1 ꢀ 5 % Et O/nꢀpentane). R = 0.7 (Et O/nꢀ
pentane 1:9); H NMR (600 MHz, CDCl ): δ = 7.41 (m, 5H; H7ꢀ9), 5.49 (s, 1H; C2), 2.90 (q,
J = 7.6 Hz, 2H; C4), 1.13 (t, J = 7.6 Hz, 3H; C5) ppm; C NMR (151 MHz, CDCl ): δ = 166.5
(C3), 137.5 (C6), 130.2 (C9), 129.0 (2C; C8), 126.4 (2C; C7), 117.4 (C1), 95.1 (C2), 27.4 (C4),
1
1
Prepared according to General Procedure A,
2
2
f
2
1
3
13
3
3.4 (C5) ppm; IR (ATR): ῦ = 3060(w), 2975(w), 2938(w), 2877(w), 2213(m), 1603(m),
574(m), 1495(w), 1465(m), 1445(m), 1378(w), 1342(w), 1308(w), 1284(w), 1235(w), 1159(w),
ꢀ1
1081(w), 1030(w), 1001(w), 907(w), 829(m), 806(w), 760(s), 695(s) cm ; HRꢀESIꢀMS
+
+
(MicroTOF): m/z: 180.0786 ([M+Na] , calcd. for C H NNa : 180.0784).
11 11
(E)-3-Phenylhex-2-enenitrile (E-4) Prepared according to General Procedure A, butyrophenone
(4 mmol) was converted to E-4 in 24 h yielding a colourless oil (363 mg, 53%) after purification
by column chromatography (SiO , 1 ꢀ 5 % Et O/nꢀpentane). R = 0.72 (Et O/nꢀpentane 1:9); H
NMR (600 MHz, CDCl ): δ = 7.41 – 7.39 (m, 5H; H8ꢀ10), 5.52 (s, 1H; H2), 2.92 – 2.83 (m, 2H;
H4), 1.51 (sext, J = 7.4 Hz, 2H; H5), 0.95 (t, J = 7.4 Hz, 3H; H6) ppm; C NMR (151 MHz,
CDCl ): δ = 165.1 (C3), 137.8 (C7), 130.1 (C10), 129.0 (2C; C9), 126.4 (2C; C8), 117.6 (C1),
96.0 (C2) 35.9 (C4), 21.9 (C5), 13.7 (C6) ppm; IR (ATR): ῦ = 3060(w), 2962(m), 2933(w),
2
1
1
2
2
f
2
3
13
3
873(w), 2214(m), 1679(w), 1603(m), 1574(w), 1494(w), 1457(m), 1445(m), 1381(w), 1335(w),
255(w), 1213(w), 1159(w), 1081(w), 1031(w), 1001(w), 918(w), 878(w), 829(m), 784(w),
ꢀ1
+
758(s), 695(s), 657(w) cm ; HRꢀESIꢀMS (MicroTOF): m/z: 194.0946 ([M+Na] , calcd. for
C H NNa : 194.0940); analytical data in agreement with literature.
+
41
12
13
(E)-3-Phenylhept-2-enenitrile (E-5) Prepared according to General Procedure A,
valerophenone (4 mmol) was converted to E-5 in 24 h yielding a colourless oil (482 mg, 65%)
after purification by column chromatography (SiO , 1 ꢀ 5 % Et O/nꢀpentane). R = 0.72 (Et O/nꢀ
pentane 1:9); H NMR (600 MHz, CDCl ): δ = 7.40 (m, 5H; H9ꢀ11), 5.50 (s, 1H; H2), 2.90 (t,
J = 7.4 Hz, 2H; H4), 1.50 – 1.42 (m, 2H; H5), 1.41 – 1.35 (m, 2H; H6), 0.90 (t, J = 7.2 Hz, 3H;
H7) ppm; C NMR (151 MHz, CDCl ): δ = 165.3 (C3), 137.9 (C8), 130.1 (C11), 129.0 (2C;
2
2
f
2
1
3
13
3
C10), 126.4 (2C; C9), 117.6 (C1), 95.8 (C2), 33.8 (C4), 30.7 (C5), 22.5 (C6), 13.9 (C7) ppm; IR
ACS Paragon Plus Environment
29