Axially Chiral Enamides: Substituent Effects, Rotation Barriers, and Implications for their Cyclization Reactions

Article abstract of DOI:10.1021/acs.joc.6b00889

The barrier to rotation around the N-alkenyl bond of 38 N-alkenyl-N-alkylacetamide derivatives was measured (ΔG^∞ rotation varied between <8.0 and 31.0 kcal mol^-1). The most important factor in controlling the rate of rotation was the level of alkene substitution, followed by the size of the nitrogen substituent and, finally, the size of the acyl substituent. Tertiary enamides with four alkenyl substituents exhibited half-lives for rotation between 5.5 days and 99 years at 298 K, sufficient to isolate enantiomerically enriched atropisomers. The radical cyclizations of a subset of N-alkenyl-N-benzyl-α-haloacetamides exhibiting relatively high barriers to rotation round the N-alkenyl bond (ΔG^∞ rotation >20 kcal mol^-1) were studied to determine the regiochemistry of cyclization. Those with high barriers (>27 kcal mol^-1) did not lead to cyclization, but those with lower values produced highly functionalized γ-lactams via a 5-endo-trig radical-polar crossover process that was terminated by reduction, an unusual cyclopropanation sequence, or trapping with H₂O, depending upon the reaction conditions. Because elevated temperatures were necessary for cyclization, this precluded study of the asymmetric transfer in the reaction of individual atropisomers. However, enantiomerically enriched atropsiomeric enamides should be regarded as potential asymmetric building blocks for reactions that can be accomplished at room temperature.

Full text of DOI:10.1021/acs.joc.6b00889

Subscriber access provided by University of Birmingham
Article
Axially Chiral Enamides: Substituent Effects, Rotation
Barriers and Implications for their Cyclization Reactions
Andrew J. Clark, Dennis P. Curran, David J. Fox, Franco Ghelfi, Collette S. Guy, Benjamin Hay, Natalie
James, Jessica Margaret Phillips, Fabrizio Roncaglia, Philip Bryan Sellars, Paul Wilson, and Hanmo Zhang
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.6b00889 • Publication Date (Web): 06 Jun 2016
Downloaded from http://pubs.acs.org on June 9, 2016
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 67
The Journal of Organic Chemistry
1
2
3
4
Axially Chiral Enamides: Substituent Effects, Rotation
Barriers and Implications for their Cyclization Reactions
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
*,†
*,‡
†
§
†
Andrew J. Clark, Dennis P. Curran, David J. Fox, Franco Ghelfi, Collette S. Guy, Benjamin
‡
†
†
§
†
†
Hay , Natalie James, Jessica M. Phillips, Fabrizio Roncaglia, Philip B. Sellars, Paul Wilson and
‡
Hanmo Zhang.
†
Department of Chemistry, University of Warwick, Coventry, West Midlands, CV4 7AL, UK.
eꢀmail:a.j.clark@warwick.ac.uk
‡
Department of Chemistry, University of Pittsburgh, Pittsburgh Pennsylvania 15260, USA.
eꢀmail: curran@pitt.edu
§
Dipartmento di Scienze Chimiche e Geologiche, Univerita degli, Studi, di Modena e Reggio Emilia,
Via Campi 103, 41125 Modena, Italy.
TOC GRAPHIC
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 67
ABSTRACT
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
The barrier to rotation around the Nꢀalkenyl bond of thirtyꢀeight NꢀalkenylꢀN-alkyl acetamide
‡
ꢀ1
derivatives was measured (ꢀG rotation varied between <8.0 to 31.0 kcal mol ). The most important
factor in controlling the rate of rotation was the level of alkene substitution, followed by the size of the
nitrogen substituent and finally the size of the acyl substituent. Tertiary enamides with four alkenyl
substituents exhibited halfꢀlives for rotation between 5.5 days ꢀ 99 years at 298K, sufficient to isolate
enantiomerically enriched atropisomers. The radical cyclizations of a subset of NꢀalkenylꢀNꢀbenzyl αꢀ
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
‡
haloacetamides exhibiting relatively high barriers to rotation round the Nꢀalkenyl bond (ꢀG rotation >
ꢀ1
2
0 kcal mol ) were studied to determine the regiochemistry of cyclization. Those with high barriers
ꢀ1
(>27 kcal mol ) did not lead to cyclization but those with lower values produced highly functionalized
γꢀlactams via a 5ꢀendo-trig radicalꢀpolar crossover process that was terminated by reduction, an unusual
cyclopropanation sequence or trapping with H O depending upon the reaction conditions. Because
2
elevated temperatures were necessary for cyclization this precluded study of asymmetric transfer in the
reaction of individual atropisomers. However enantiomerically enriched atropsiomeric enamides should
be regarded as potential asymmetric building blocks for reactions that can be accomplished at room
temperature.
INTRODUCTION
In recent years, the bond rotational dynamics, asymmetric synthesis and reactions of nonꢀbiaryl
1
ꢀ3
atropisomers have received considerable attention. The majority of work has focussed on the
4
ꢀ9
10ꢀ14
chemistry of anilides 1a and benzamides 1b
where the amide group is perpendicular to the plane
3
4
of the aryl system, (Figure 1). In 2,6ꢀdisubstituted anilide derivatives 1a (R or R ≠ H) rotation around
the Nꢀaryl bond is slow enough for atropisomers to be separated at room temperature and axially chiral
3
anilides 1a have been shown to undergo a range of reactions with transfer of chirality. Secondary
1
5ꢀ16
enamides are valuable synthetic intermediates
but the chemistry of tertiary enamides 1c has
2
ACS Paragon Plus Environment

Page 3 of 67
The Journal of Organic Chemistry
1
7ꢀ18
1
2
received much less attention.
Systems with small substituents 2 (e.g. R = R = Me) may be planar in
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
2
3
4
5
the ground state, but if substituents R and R are large enough or the alkene is substituted (R = R = R
≠ H) then enamides 1c, have the potential to exhibit axial chirality. Theoretically if the rotation barrier
around the Nꢀalkenyl bond is high enough then individual atropisomers may be separated and their
chemistry studied.
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
R1
O
O
R1
O
N
R1
R2
R2
R3
N
R2
R3
N
R4
R3
R4
R4
R5
1a anilides
1b benzamides
1c enamides
O
O
R2
R2
N
R1
N
R1
O
R1
(E)-syn 2
(E)-anti 2
N
R2
R1
R1
3
R2
R2
N
E-amide
rotamers
favoured
N
O
O
(Z)-syn 2
(Z)-anti 2
Figure 1. Structures of some axially chiral molecules and conformations of tertiary enamides 2.
The rotation dynamics of tertiary enamides can be complicated because of restricted rotation around
both the amide NꢀCO [(E)ꢀsyn 2 → (Z)ꢀsyn 2] and the Nꢀalkenyl bonds [(E)ꢀsyn 2 → (E)ꢀanti 2], (Figure
1
2
1
1
). The barrier to rotation around the amide NꢀCO bond for 2a (R = R = Me) has been measured at
ꢀ1
19
ꢀ1 20
4.0 kcal mol and is slightly lower than the general value for amides (15ꢀ20 kcal mol ).
Nꢀ
2
1
cycloalkenylꢀNꢀalkylꢀacetamides such a 3 generally prefer the amide NꢀCO E-rotamer and exhibit
transient axial chirality on the NMR timescale. This makes analysis of the rotational dynamics
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 67
1
21
around the Nꢀalkenyl bond relatively easy to study by VT H NMR. In this paper we report studies into
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
2
3
4
5
the effect of substitution (1c, R , R , R , R and R ) on the barriers to rotation around the Nꢀalkenyl bond
of tertiary enamides and show that it is possible to separate individual enamide atropisomers at room
temperature with halfꢀlives (t ) of up to 99 years at 298K. We probe the ability of a subset of sterically
½
‡
congested racemic αꢀhalogenated tertiary enamides with relatively high barriers to rotation (ꢀG
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
298 rot
ꢀ1
‡
>
20 kcal mol ) to undergo 5ꢀendo-trig radical cyclization. Those with high barriers (ꢀG
>27 kcal
98 rot
2
ꢀ1
ꢀ1
‡
ꢀ1
mol ) do not lead to cyclization but those with lower values (27 kcal mol > ꢀG
> 20 kcal mol )
2
98 rot
mostly lead to highly functionalised γꢀlactams (no βꢀlactam formation was observed) where the mode of
termination is controlled in part by steric factors and in part by the method of cyclization.
RESULTS AND DISCUSSION
1
2
Substitution at the alkene: Compound 2b (R = Me, R = Bn) was prepared by acetylation of the Nꢀ
2
2
23
benzylimine of acetaldehyde (13%). Computational analysis of 2b using the TZVP basis set and the
2
4
25
B3LYPꢀD3(BJ) functional using PCꢀGAMESS/Firefly 8.0 suggested that three out of the four
possible conformations were planar (the exception being (E)ꢀsyn 2b) with (E)ꢀanti 2b and (Z)ꢀanti 2b
conformations predominating at equilibrium, Figure 2. Electronic energies and zeroꢀpoint energies for
all conformations are provided in the supplementary information. Two diastereomeric versions of the
nonꢀplanar (E)ꢀsyn 2b conformation exist, where either the vinyl group or the phenyl group are angled
either in front of the plane of the amide or behind it. The energy and relative equilibrium constant of the
more stable conformer is shown. In all cases the minimized energy structures had imaginary frequencies
of zero. The E+ZPE energy was used to calculate the relative populations at 298K.
4
ACS Paragon Plus Environment

Page 5 of 67
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
1
Figure 2. Calculated relative energies, H NMR shifts and equilibrium contributions of conformers of
2
2
b.
1
Experimentally, a 2:1 ratio of two conformations was observed in the 600 MHz H NMR spectrum of
b in CDCl at 298K. The major isomer was confirmed as (E)ꢀantiꢀ2b and the minor isomer (Z)ꢀantiꢀ2b
3
1
based upon their calculated theoretical H NMR chemical shifts and nOe data. DFT groundꢀstate
2
6
27
structures were analysed using GIAO method Gaussian03 using mPW1PW91 functional and the 6ꢀ
28
3
11+G(2d,p) basis set and scrf=(solvent=chcl3,cpcm,read) radii=uaks nosymcav options. NMR shifts
were calculated using parameters specific to the functional, basis set and option combination described
2
9
Lodewyk et al. The Η−1 enamide proton resonates at 6.80 ppm in the major conformer and 7.50 ppm
in the minor conformer in good agreement with the theoretical calculations. This data is similar to that
1
2
30
reported for the related NꢀbenzylꢀNꢀvinylformamide 2c (R = H, R = Bn) and is consistent with a slow
rotation around the amide bond and a fast rotation around the CꢀN bond of the alkene with a higher
population of the (E)ꢀanti conformer at equilibrium at room temperature. Heating 2b (298K→373K) in
d8ꢀtoluene caused a broadening of all signals and coalescence to a single set of peaks consistent with
rapid rotation around the amide bond.
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 67
In order to assess the effect of increasing alkene substitution on the barrier to Nꢀalkenyl bond rotation
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
we prepared structures 4aꢀi, Figure 3. We chose to study the Nꢀbenzyl derivatives as measurement of
1
rotation rates should be possible using VT H NMR. Upon cooling, the benzylic CH singlet should
2
broaden and would be expected to decoalesce (caused by the two protons of the CH group being in an
2
asymmetric environment) and ultimately form two doublets due to the diastereotopic nature of the
benzylic protons. This behavior is not consistent with amide bond rotation as this would cause the
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
31
number of signals to double. Using the WINDNMR 7.1 lineꢀshape analysis program it was possible to
determine the rotational rate constant at each temperature and determine the thermodynamic parameters
3
2
via a standard Eyring plot. Compounds 4a, 4eꢀf were prepared by acetylation of the known benzyl
22
22
33
imines of acetone, 2ꢀmethyl propanal and 2ꢀphenyl cyclohexanone, while 4bꢀd and 4gꢀi were
prepared by benzylation (NaH, BnBr) or methylation (NaH, MeI) of the corresponding Nꢀacetyl
3
4ꢀ37
enamides 5 (16ꢀ80%).
The barrier to rotation around the CꢀN bond for 4a and 4e was too low to
1
‡
ꢀ1
measure by VT H NMR (ꢀG
estimated to be <8.5 kcal mol ). On cooling, the benzylic CH singlet
2
98
2
began to broaden as expected, but even at 179K decoalescence was not observed. On the other hand, the
‡
values of ꢀG
for rotation for the 1,2ꢀsubstituted derivatives 4bꢀd in d ꢀtoluene were determined to be
8
2
98
ꢀ1
16.3, 18.6 and 9.3 kcal mol . As expected, increasing steric congestion around the alkene increases the
barrier to rotation due to the inherent difficulty in passing through a hindered planar structure during
rotation.
6
ACS Paragon Plus Environment

Page 7 of 67
The Journal of Organic Chemistry
O
O
O
1
2
3
4
5
6
7
8
9
Ph
Ph
N
Ph
N
Ph
N
R
4
4
b R = Me (16.3)
c R = Ph (18.6)34
4d (9.3)
4
a (<8.5)
O
N
O
O
N
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
Ph
N
Ph
Ph
'
'
4
e (<8.5)
4g
4
f
O
O
O
R1
R
N
R2
HN
R1
N
R3
I
R3
'
'
R2
3
35
6a R = H, (< 24)
i R = Bn (25.6) 6b R =Me, (> 26)
5
4
4
h R = Me
3
36
‡
Figure 3. Effect of alkenyl substitution on barrier for rotation around Nꢀalkenyl bond for 4aꢀi (ꢀG 298
ꢀ1
in kcal mol for rotation around the Nꢀalkenyl bond is shown in brackets).
For the derivatives 4fꢀi the barrier to rotation was too high to measure with conventional VT
1
experiments. The H NMR showed a sharp set of diastereotopic signals for the CH benzyl protons even
2
at 373K. It was not possible to fully resolve the enantiomers of 4f and 4g by chiral HPLC on a variety
of columns, indicating that the atropisomers were either inseparable on the columns examined or that
they were rapidly interconverting on the HPLC timescale, evidence for the latter is provided vide infra.
While compound 4h was partially resolved, both enantiomers of 4i were fully resolved on a semiꢀ
®
‡
preparative WhelkꢀO column (30.1 and 37.4 mins). The ꢀG
for rotation of 4i was determined to be
98
2
ꢀ
1
43
25.6 kcal mol by measuring the kinetics of racemization at 82 ˚C. This suggests that a branching
3
substituent at the sp hybridised α’ꢀposition is required to generate suitable barriers to rotation for
atropisomers to be successfully separated at room temperature. This is a similar structural motif to
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 67
3
2
that found in anilides 6b, (R ≠ H, shown in red with an sp branching point) where barriers to rotation
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 40
were greater than 26 kcal mol . On the other hand, structures 4b, 4f and 4g with intermediate barriers
3
to rotation structurally resemble anilides, 6a (R = H) and as a consequence have lower barriers to
39
rotation.
To further study the effects of branching at the α’ꢀposition we prepared structures 7aꢀd,
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
investigated their resolution by chiral HPLC and calculated the barriers to rotation and halfꢀlives of
those that could be resolved, Figure 5.
O
O
O
Ph
Ph
N
N
Ph
N
'
'
'
7c (31.0)
7a (21.3)
7b (27.5)
74°
X-ray crystal structure 7b
Chiral HPLC 7b
O
Ph
N
'
65°
7d
X-ray crystal structure 7d
Figure 4. Structures 7aꢀd, Xꢀray crystal structures for 7b and 7d and chiral HPLC traces for 7b,
‡
ꢀ1
(ꢀG
in kcal mol for rotation around the Nꢀalkenyl bond is shown in brackets).
98
2
It was possible to fully separate the enantiomers of 7b (containing both an αꢀmethyl group and an α’ꢀ
®
‡
annulated phenyl group) on a semiꢀpreparative WhelkꢀO column (37.09 and 46.47 mins) and a ꢀG
ꢀ1
2
98K for rotation of 27.5 kcal mol was determined by measuring the kinetics of racemization of a
9
9:1 enantiomerically enriched sample at 82 ˚C. Compound 7b is a restricted conformational
8
ACS Paragon Plus Environment

Page 9 of 67
The Journal of Organic Chemistry
‡
ꢀ1 38
analogue of the acyclic derivative 7a first prepared by Ahlbrecht (7a ꢀG = 21.3 kcal mol ). The
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
difference in the measured barrier to rotation is significant, δꢀG = 6.2 kcal mol , and the relatively low
value for 7a indicates a likely coꢀoperative gearing of the α’ꢀphenyl substituent rotation during the key
rotation around the Nꢀalkenyl bond in 7a, Figure 4. The related derivative 7d gave a single peak on the
®
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
WhelkꢀO column and its rotation dynamics were not further analysed. On the other hand 7c provided
‡
ꢀ1
the highest barrier to rotation (ꢀG = 31.0 kcal mol , t = 99 years at 298K, enantiomers separated at
½
4
0
2
9.0 and 31.9 minutes) and this barrier is comparable to published values for oꢀiodoacrylanilides 6b.
We solved the Xꢀray structures of 7b and 7d which confirmed the preference for adoption of the amide
NꢀCO E-rotamer geometry in the solid state, Figure 4. The torsional angle of the key Nꢀalkenyl bond
4
1
21,42
C=CꢀNC(O) in 7b is 74º which is similar to that observed for related anilides and other enamides.
The sum of the angles around the nitrogen atom was 359.9º suggesting the nitrogen was planar as
1
expected. For 7d the torsional angle was slightly smaller, 65º. The 400MHz H NMR of both 7a and 7d
in d ꢀtoluene showed one pair of mutually coupled sharp doublets (for 7a J = 14.0 Hz at 5.27 and 3.48
8
ppm, for 7d J = 14.0 Hz at 5.37 and 3.37 ppm) which did not broaden upon heating. This is consistent
with the existence of largely a single Eꢀamide rotamer in solution with a high barrier to rotation around
the Nꢀalkenyl bond causing the benzyl protons to be diastereotopic as seen in the crystal structure of 7d.
1
2
3
5
4
These results suggest that enamides 1c (R , R , R , R ≠ H, R = branched substituent) are likely to have
sufficient barrier to rotation around the Nꢀalkenyl bond to be resolvable for useful periods at room
temperature.
Substitution at nitrogen: The effect of the size of the nitrogen substituent 8aꢀe on the energy barrier
for Nꢀalkenyl bond rotation was assessed by VT NMR, Figure 5. Deprotonation of Nꢀ1ꢀcyclohexenꢀ1ꢀ
4
3
ylꢀbenzeneacetamide with NaH in THF followed by the addition of either MeI (8a, 84%), BuBr (8b,
i
44
7
0%), and PrI (8d, 13%) furnished the desired enamides 8a, b or 8d. Compound 8c was prepared by
acylation of the Nꢀbenzylimine of cyclohexanone with phenylacetyl chloride according to a literature
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 67
4
2
procedure. We also synthesized compound 8e by reaction of Nꢀ1ꢀcyclohexenꢀ1ꢀylꢀbenzene acetamide
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
4
5
with 3 equivalents of 2,6ꢀlutidine and TBSOTf, Figure 5.
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 5. Effect of alkenyl substitution on the rotation barrier around Nꢀalkenyl bond for 8aꢀe (ꢀG
298
ꢀ1
in kcal mol for rotation around the Nꢀalkenyl 8aꢀd or Nꢀaryl 9aꢀd bond is shown in brackets).
The moisture sensitivity of 8e required its preparation in situ in d ꢀtoluene in an NMR tube and it was
8
not possible to isolate a pure sample. For compounds 8aꢀb and 8d a characteristic change in the
chemical shift of the alkene proton occurred upon Nꢀalkylation (δꢀ between 0.46ꢀ0.70 ppm)
highlighting the loss of conjugation with the nitrogen lone pair in 8aꢀd. The sense and magnitude of this
shift was also observed upon silylation in d ꢀtoluene (8e 4.84 ppm) providing evidence that silylation to
8
ꢀ1
give 8e had occurred. As expected, the barrier to rotation increased, 8a (10.1 kcal mol ) < 8bꢀc (11.5ꢀ
ꢀ1
ꢀ1
1
1.7 kcal mol ) < 8d (15.0 kcal mol ), as the size of the alkyl group increased in parallel with the trend
4
6
observed in anilides 9aꢀd. In both series there is relatively small increase on moving from the Nꢀ
‡
ꢀ1
‡
methyl to Nꢀ1° alkyl substituent (enamides 8a → 8b δꢀG = 1.6 kcal mol , anilides 9a → 9b δꢀG =
ꢀ1
1.0 kcal mol ) whereas there is a significantly larger increase for the iꢀpropyl substituent (enamides 8a
‡
ꢀ1
‡
ꢀ1
→
8d δꢀG = 4.9 kcal mol , anilides 9a → 9d δꢀG = 5.1 kcal mol ). This is likely due to the R
group in the NꢀCH R substituent being able to rotate away from the plane of the amide and alkene 10
2
ACS Paragon Plus Environment

Page 11 of 67
The Journal of Organic Chemistry
4
7
during CꢀN alkenyl rotation while this is not possible for the methyl groups NꢀCH(Me) in 8d. No
2
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
broadening of the geminal protons of 8e was observed on cooling to 193K indicating that rotation is
rapid, presumably through the Oꢀsilyl imidate 8f, Figure 5. Silylation of related anilides has been
48
reported to proceed to give both Nꢀsilylated and Oꢀsilylated structures that are in rapid equilibrium and
this may be occurring for enamide 8e.
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
O
O
Br
Bu SnH
3
ACN
R
N
R
N
10
11
O
O
X
1) NaH, THF, RT
HN
N
2)
X
Y
Br
n
12
13 n = 1
Y
15 n = 2, X = Y = Br
O
Br
N
Ph
79°
14
X-ray structure of 14
Figure 6. Synthesis of compounds 13aꢀf.
The addition of radicals onto tertiary enamides (derived from aryl bromides 10) have been reported to
4
9
give tetrahydroisoquinolines 11 via a 6ꢀendo-trig radical cyclization. The bond rotation dynamics of Nꢀ
2
ꢀhalobenzyl derivatives has not been investigated. We briefly investigated the effect of orthoꢀ
halobenzyl substituents upon the rotation barrier around the Nꢀalkenyl bond in related systems 13aꢀg,
Table 1. While the effect of such substituents X and Y might be limited (as they are 3 atoms away from
the nitrogen atom), the relatively large size of bromine and iodine atoms (typically used to initiate
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 67
radical reactions or other metal mediated cyclizations) may be significant. The compounds 13aꢀg were
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
3
4
prepared by alkylating the known acetamide 12 with appropriately functionalized benzyl halides using
the same approach as for 8aꢀe.
Table 1. NꢀAlkenyl bond rotation rates from variable temperature NMR experiments for 13aꢀg.
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
‡
Comp. X Y
ꢀG
298
ꢀ
1 a
(
kcal mol )
38
1
3a H H
3b F H
3c I H
13d F F
3e Cl Cl
10.1
1
9.9
9.9
1
10.9
11.7
13.1
14.8
1
1
3f Br Br
b
15
Br Br
a
‡
ꢀ1
b
Estimated errors ꢀG
± 0.2 kcal mol these are inꢀline with related work. Enamide constrained in a seven membered
98
2
ring
The introduction of a single 2ꢀfluorine 13b or 2ꢀiodine 13c substituent had little effect on the rotation
barrier; the Xꢀray structure of the related 2ꢀbromide 14 clearly showed the halide orientated away from
the alkene group where it doesn’t interfere with the Nꢀalkenyl bond rotation, Figure 6. On the other hand
benzyl substituents containing two ortho substituents increase the barrier to rotation in line with their
ꢀ1
ꢀ1
ꢀ1
size (13d F = +0.8 kcal mol , 13e Cl = +1.6 kcal mol , 13f Br = +2.9 kcal mol ) despite their
12
ACS Paragon Plus Environment

Page 13 of 67
The Journal of Organic Chemistry
distance from the key Nꢀalkenyl bond rotation. The size of the ring in which the enamide alkene was
constrained also affected the barrier to rotation in the order 6ꢀmembered < 7ꢀmembered (compare 13f
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
and 15 δꢀG = +1.7 kcal mol ) which is a consequence of the bond angles of the different ring
2
1
systems.
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
Substitution at the acyl group: Previous work with tertiary enamides containing electron poor and
electron rich aromatic acyl substituents has shown that the electronic nature of the acyl substituent has a
34
negligible effect on the Nꢀalkenyl rotation barrier, suggesting that steric effects alone are important. It
has previously been reported that the size of the acyl group (R) in 3aꢀi only moderately affects the
2
1,42
‡
barrier to rotation in enamides, (Figure 7).
Hence, replacing an acetyl group 3a (ꢀG
= 10.1
98 rot
2
ꢀ1
‡
ꢀ1
kcal mol ) with the much larger trichloroacetyl group 3i (ꢀG
= 14.2 kcal mol ) increases the
298 rot
ꢀ1
barrier to rotation by 4.1 kcal mol . This steric effect is less significant for the acyl substituent than the
i
ꢀ1
nitrogen substituent (nitrogen substituent Me 8a→ Pr 8d δꢀG = 4.9 kcal mol , acyl substituent Me
i
ꢀ1 42
47
3
a→ Pr 3d δꢀG = 1.6 kcal mol ). Similar behaviour has also been reported for anilides.
‡
Figure 7. Effect of acyl substituent 3aꢀi upon the ꢀG
barriers around the Nꢀalkenyl bond (values
98 rot
2
ꢀ1
42
in kcal mol shown in brackets).
For small R substituents (such as 3a) the barrier to Nꢀalkenyl rotation is likely to be lower than amide
1
9
NꢀCO rotation, but for larger substituents (such as 3i) the barriers are likely to be similar.
Consequently, a number of potential mechanisms for enantiomerization (E,Mꢀ3 → E,Pꢀ3) are
possible, including simple enamide rotation (E,Mꢀ3 → E,Pꢀ3), or a coꢀoperative coupled rotation of
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 67
5
0
both amide and enamides rotation, a geared process, (E,Mꢀ3 → Z,Mꢀ3 → E,Pꢀ3). On plotting ln krot
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
5
1
values for 3aꢀi against the cone angle (θ values) of the acyl substituent we found a linear correlation
R
2
(R = 0.958) which suggests that the mechanism for rotation is likely to be the same for all of the series
and that the cone angle of the acyl substituent may be a useful tool in predicting rotational barriers for a
given series of enamides, Figure 8.
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
O
R
Bn
Bn
N
R
N
O
amide
rotation
E,M-3
Z,M-3
enamide
rotation
enamide
rotation
O
R
O
Bn
Bn
N
R
N
amide
rotation
E,P-3
Z,P-3
3h
3i
3f
3e
3g
3c
3d
3b
3a
2
Figure 8. Correlation between cone angle θ and ln k values for 3aꢀi (R = 0.958).
R
rot
In summary, the most important factors in controlling the rate of Nꢀalkenyl bond rotation in enamides is
the level of alkene substitution (particularly any branching substituent at the α’ꢀposition), followed by
14
ACS Paragon Plus Environment

Page 15 of 67
The Journal of Organic Chemistry
the size of the nitrogen substituent (and the distance of any branching from the nitrogen atom) and
finally the size of the acyl substituent.
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
Radical cyclization substrate dynamics: The radical cyclization of αꢀhaloenamides 16ꢀ18 is well
documented and may proceed via a 4ꢀexo or 5ꢀendo cyclization depending upon the substrate, Figure
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
5
2ꢀ64
9
.
Rotational features of enamides can dictate the success or failure of radical cyclizations with the
3
0
Eꢀamide rotamer being required. In general, 1ꢀsubstituted enamides 16 cyclize via a 5ꢀendo pathway
2
52ꢀ55
56ꢀ58
(
R = Ph, or CO R)
while 2ꢀ or 2,2ꢀsubstituted enamides 17 proceed via a 4ꢀexo pathway,
2
although electronic factors and temperature can also play a part in controlling regiochemistry for these
5
8ꢀ60
substrates.
By far the majority of cyclization reactions reported involve 1,2ꢀsubstituted enamides 18,
59ꢀ70
which generally proceed via a 5ꢀendo pathway.
For cyclization of radicals derived from homolysis
6
6ꢀ70
of the CꢀBr bond in substrates 3c and 3h,
it is necessary for a twisting to occur in the transition state
so that the radical SOMO and alkene LUMO orbitals can overlap efficiently, and this has been
71
confirmed by calculations. As a consequence, the twisted ground state of molecules such as 3c and 3h
likely facilitate cyclization with a movement towards planarity occurring during the reaction, (Figure 9).
Far less is known about cyclizations of 1,2,2ꢀsubstituted enamides where movement towards planarity
59,60,72
during cyclization may be hindered due to steric effects.
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 67
O
O
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
R1
R2
R1
ref. 52-55
N
N
1
2
5-endo
R2
16
O
R1
R1
O
ref. 56-60
N
N
R3
4-exo
R3
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
2
R2
R2
17
O
O
R1
R2
R1
R2
ref. 59-70
N
N
1
2
5-endo
R3
R3
18
O
O
Bn
Br
Bn
O
N
N
R
Bn
N
R R CuI
or
R
5
-endo
R
R
H
Bu SnH
3
chirality
3
3
c R = H
h R = Me
transfer from
atropisomer?
Figure 9. Regiochemical modes of cyclization of αꢀhaloenamides.
Cyclization of 19 was reported to be unsuccessful using Bu SnH at 80 °C, while cyclization of the
3
72
analogue 21 gave both 4ꢀexo 23 and 5ꢀendo products 22 dependent upon the temperature, Figure 10.
This suggests that the 4ꢀexo cyclization process was reversible and at the higher temperature 5ꢀendo
cyclization followed by irreversible loss of the phenylthiyl radical predominated to give 22. From our
studies above it is apparent that these substrates are likely to have ground state Nꢀalkenyl bond rotation
ꢀ1
barriers less than 20 kcal mol .
16
ACS Paragon Plus Environment

Page 17 of 67
The Journal of Organic Chemistry
O
O
1
Bu SnH
Bn
Bn
3
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
N
CH3
N
CH Br AIBN
2
Ph
Ph
'
C H
6
6
reflux
8%
4
19
2
0
O
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
O
Bn
'
N
Bu SnH
3
N
CHCl2
SPh
AIBN
22 Bn
SPh
Cl
21
8
1
0°C, 78% 22:23 = 1.0:1.2
10°C, 57% 22:23 = 2.2:1.0
N
Bn
O
2
3
Figure 10. Regiochemical modes of cyclization of 1,2,2ꢀsubstituted αꢀhaloenamides.
To the best of our knowledge the cyclization of enamides containing further branching at the α’ꢀposition
have not been studied in detail, Figure 10. For these substrates where barriers to rotation around the Nꢀ
ꢀ1
alkenyl bonds are likely to be significant (> 20 kcal mol ) it is unclear if cyclization would be efficient
because steric interactions would develop as the radical and radical acceptor move towards planarity
during the cyclization process. In order to address these issues we prepared enamides 24ꢀ28, examined
their barriers to rotation and investigated their radical cyclization reactions, Figure 11.
The most commonly used protocols for mediating radical cyclizations of αꢀhaloenamides are (i) Bu SnH
3
5
2ꢀ54, 62ꢀ65
70
/
AIBN,
where chlorides provide higher yields of cyclized products than bromides or iodides
5
7
66
56ꢀ57, 67ꢀ
and (ii) copper(I) complexes of bipyridine, hexamethyltriethylenediamine, or tripyridylamine,
6
9
that work for trichloroacetamide or tertiary halide derivatives only, (Bu SnH: R = CH Cl, CH Br,
3
2
2
64
CH I; Cu(I) R = CCl , CMe Br). Consequently, we tested αꢀhaloenamides 24ꢀ28, under a range of
2
3
2
conditions and determined the regiochemistry of their radical cyclization reactions, Figure 11. It
17
was not possible to fully resolve the atropisomers of 24ꢀ26, despite the large acyl substituents. We
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 67
believe this is indicative of individual isomers interconverting on the HPLC timescale at room
temperature. Evidence for this assumption can be obtained from the HPLC chromatogram for
compound 26a which is indicative of the individual enantiomers interconverting at room temperature,
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
Figure 12d. The ꢀG value for rotation for 26a was estimated to lie between 21ꢀ23 kcal mol based
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
upon a halfꢀlife of minutes at 298K. In all these cases the coꢀoperative gearing of the phenyl substituent
7
3
during Nꢀalkenyl bond rotation is likely to lower the observed barrier to rotation.
O
O
O
Bn
Bn
Bn
N
R
N
R
N
R
'
Ph
Ph
'
'
2
2
2
6a R = CH Cl
6b R =CCl3
2
2
4a R =CCl3
4b R = CMe Br 25b R =CCl3
25a R = CH Cl
2
2
2
6c R = CMe Br
2
5c R = CMe Br
2
2
O
O
Bn
Bn
N
Cl
N
R
'
'
2
2
2
2
7a R = CH Cl (27.8)
2
28
7b R =CH Br (29.3)
2
7c R = CMe Br (25.6)
2
7d R = CH2I
‡
Figure 11 Structures 24ꢀ28 showing the ꢀG
barriers around the Nꢀalkenyl bond in brackets
98 rot
2
ꢀ1
(
values in kcal mol ).
For relatively large acyl substituents (25b, 26b, R = CCl , 25c, 26c, R = CMe Br) the barrier for Nꢀ(CO)
3
2
amide bond rotation was found to be lower than that of Nꢀalkenyl bond rotation, Figure 12b.
18
ACS Paragon Plus Environment

Page 19 of 67
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
a)
1
400 MHz H NMR of 25c in d -toluene
8
3
73 K
5
.5
5.0
5.0
Ph
4.5
f1(ppm)
4.0
4.0
2
98 K
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
5.5
4.5
f1(ppm)
b)
Br
O
N
Br
O
N
Ph
Ph
Ph
(
Z)-25c
(E)-25c
d)
c)
26a
Chiral HPLC of 26a
showing interconversion
on HPLC timescale
X-ray structure of (Z)-25c
1
Figure 12 a) Expansion of the benzylic region of the 400 MHz H NMR spectra of 25c in d toluene
8
at 298K and 373K, b) Coꢀoperative rotation of amide Nꢀ(CO) bond during Nꢀalkenyl bond rotation, c)
Xꢀray crystal structure of 25c, and d) chiral HPLC chromatogram of 26a.
1
Analysis of the 400MHz H NMR of 25bꢀc and 26bꢀc at 298K showed a doubling of all peaks (Figure
12a), indicative of both (E)ꢀ and (Z)ꢀamide rotamers at room temperature (while 25a or 26a only showed
1
one set of peaks). The ratio of amide rotamers obtained from the 400 MHz H NMR at RT was similar
for the two trichloroacetyl derivatives (25b (E):(Z) = 1.0:0.25 and 26b (E):(Z) = 1.0:0.26) and the two
bromo derivatives (25c (E):(Z) = 1.0:0.65 and 26c (E):(Z) = 1.0:0.62), indicating that the
population of the two rotamers was dictated by the acyl substituent. Heating either 25bꢀc or 26bꢀc
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 67
at 373K led to coalescence of the four sets of doublets to two broad singlets, indicative of a rapid
interconversion on the NMR timescale between the (E)ꢀ and (Z)ꢀamide rotamers with a slower rotation
around the enamide NꢀC bond, Figure 12a. The barrier to rotation around the Nꢀ(CO) amide bond is
known to be influenced by the acyl substituent with sterically demanding (tBu) groups lowering the
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
74
barrier by 4ꢀ5 kcal mol compared to simple acetamides derivatives. The Xꢀray crystal structure 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
25c clearly shows the (Z)ꢀamide rotamer in the solid state, Figure 12c. A similar doubling of signals
1
was also observed in the H NMR of the related structure 27c (1:0.17). Thus, for enamides containing
four alkenyl substituents, the (E)ꢀgeometry is favoured in both solution (CDCl ) and the solid state for
3
relatively small primary acyl groups (R = Me, CH Cl, CH Br, cone angles 112ꢀ130°) but mixtures of
2
2
both (E)ꢀ and (Z)ꢀamide rotamers can be detected in solution if larger acyl groups or strongly electron
withdrawing groups (R = CMe Br, CCl , cone angles 153ꢀ160°) are present.
2
3
®
It was possible to fully separate the enantiomers of 27aꢀc on a semiꢀpreparative WhelkꢀO column
‡
ꢀ1
allowing for barriers to rotation to be calculated (27a er = 99:1, ꢀG = 27.8 kcal mol , t = 163 days at
½
‡
ꢀ1
‡
298 K, 27b er = 98:2, ꢀG = 29.3 kcal mol , t = 5.6 years at 298K, 27c er = 87:13, ꢀG = 25.6 kcal
½
ꢀ1
mol , t = 4 days at 298K) suggesting the that these compounds would make interesting substrates with
½
which to investigate chirality transfer during 5ꢀendo trig radical cyclisation reactions at room
temperature. If cyclization is significantly more rapid than Nꢀalkenyl bond rotation then chirality
1
transfer from individual atropisomers is theoretically possible. Radical cyclizations of related axially
chiral oꢀhaloanilides have been shown to proceed with high levels of chirality transfer from the chiral
3
9ꢀ40,75ꢀ79
axis to the newly formed stereocentre.
If however, elevated temperatures are necessary for
cyclization (e.g. 80 °C) then only bromide 27b or iodide 27d are likely to exhibit a suitable barrier to
rotation (27b t = 18.5 h at 353K) compared to 27a (2 h) and 27c (7 min). It is interesting to observe
½
that 27c has a lower barrier to Nꢀalkenyl bond rotation than either 7b or 27aꢀb despite having a larger
acyl substituent. The fact that it exists as a mixture of (E)ꢀ and (Z)ꢀamide isomers in solution
20
ACS Paragon Plus Environment

Page 21 of 67
The Journal of Organic Chemistry
suggests that the lower Nꢀalkenyl barrier may be due to a coꢀoperative gearing with rotation of the amide
functional group. However, in anilides it has been reported that the steric repulsion between bulky
substituents causes pyramidalization of amide nitrogen and twisting of amide bond (destabilization of
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
73
the ground state) to bring about the decrease in the rotational barrier around an NꢀC chiral axis. While
we cannot disprove this is the reason for the lowered barrier of 27c it is unlikely as the related substrate
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
5c shows no such pyramidization in its Xꢀray structure, Figure 12c.’
Figure 13 Formation of products from 5ꢀendo cyclisation of 3c, 3h, 3j and 3k.
Radical cyclization reactions: We have previously shown that a slow syringe pump addition of Bu SnH
3
/
AIBN to the chloride 3j produces mainly 33 (92%) from reduction of the cyclised radical 30, Figure
66
1
3. Reactions of the bromide 3c and iodide 3k proceed differently. While the bromide 3c gave 33 as
the major product (55%), the alkene regioisomers 34 (11%) and 35 + 36 (11%) were also isolated. The
iodide 3k gave the uncyclized material 32 (68%) as the major product with 34 (11%) and 35 + 36
(13%). The different product ratios were explained via a competing electron transfer from the
intermediate radical 30 to the starting halides 3c and 3k giving the acyl iminium ion 31. Elimination
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 67
of a proton from 31 produces the three alkene regioisomers 34ꢀ36. The higher ratio of reduced product
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
7
0
3
2 from the iodide 3k was shown to be due to a competing nonꢀradical deiodination process.
By analogy, we initially chose to investigate the Bu SnH mediated cyclization of the primary chlorides
3
2
6a, 27a and 28 representing varying levels of hindrance to rotation around the Nꢀalkenyl bond. The
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
chlorides were chosen to suppress products arising from electron transfer from the cyclized radical to the
starting chloride.
O
Bn
X
0
.02 M 1.5 eq,
Bn
Ph
X
N
O
N
Bu3SnH, 0.2 eq
ACN, toluene,
reflux 26 h
Ph
26a X = Cl (21%)
37a X = H (27%)
7d X = H (30%)
37b X = OH (17%)
O
O
Bn
Cl
Bn
Cl
N
N
1
.5 eq Bu SnH
3
(
syringe pump)
20 mol% eq AIBN
reflux, 48 h
27a
38a
(
47%)
AIBN
O
O
Bn
I
Bn
I
N
N
reflux,
24 h
52%
38b
27d
Figure 14 Reaction of 26a and 27a with Bu SnH.
3
‡
ꢀ1
When a 0.02M solution of 26a (ꢀG 2
~ 21ꢀ23 kcal mol ) was treated with 1.5 eq of Bu SnH and 0.2
3
98 rot
eq of ACN at reflux for 26 hours the major product was the uncyclized compound 7d (30%), suggesting
a relatively slow cyclization which is in line with that previously reported for 19. Significant
7
2
starting material 26a was also recovered (21%). The expected cyclized product 37a was obtained 22
ACS Paragon Plus Environment

Page 23 of 67
The Journal of Organic Chemistry
in 27% yield arising from a 5ꢀendo trig cyclization followed by Bu SnH mediated reduction (no 4ꢀexo
3
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
products were isolated). A significant amount of the hydroxyl terminated compound 37b (17%) was also
detected. Presumably, alcohol 37b arises via a radicalꢀpolar crossover reaction with trapping of the
intermediate acyl iminium ion with water (upon workꢀup), there being no elimination pathway
analogous to 31 → 34ꢀ36 available for 26a, Figure 14. Attempts to cyclize derivatives with higher
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
‡
ꢀ1
barriers to rotation (27a ꢀG
= 27.8 kcal mol , 28 ꢀG
>3 1.0 kcal mol ) failed, presumably
298 rot
2
98 rot
due to the increased difficulty in moving to planarity during the cyclization. Reaction of the chloroꢀ
7
9
derivative 27a with Bu SnH (0.01M) and Et B at room temperature led to recovered starting material
3
3
2
7a (44%) and the reduced substrate 7b being isolated (70% based upon recovered starting material).
On the other hand, addition of 1.5 eq Bu SnH and 20 mol% AIBN (via a syringe pump addition for 2
3
hrs, initial concentration 0.01M) led to oxidation to the naphthalene 38a in 19% yield (47% based upon
recovered starting material). Complete oxidation to the naphthalene 38a could also be accomplished if
2
7a was reacted with 2 equivalents of AIBN in the absence of Bu SnH indicating that the radical
3
initiator was responsible for the oxidation and that this process was more rapid than homolytic fission of
80ꢀ82
the CꢀCl bond.
The same oxidation to give 38b was observed for the iodide 27d. In order to
enhance the rate of radical initiation over that of oxidation we investigated the reaction of the bromide
7b. Over 18 hours, 1.5 eq of Bu SnH and 0.2 eq of ACN were added via a syringe pump to a 0.12M
2
3
solution of the bromide 27b under nitrogen in dry toluene at reflux. Thin layer chromatography revealed
that significant amounts of starting material remained after 24 hours, indicative of a slow initiation, and
so further Bu SnH/ACN was added. Three further aliquots of the initiator (1 eq in total) over 48 hours
3
were required for complete consumption of the starting material. Upon workꢀup, three products were
isolated: the uncyclized acetamide 7b (42%); the naphthanilide 38c (7%); and the ketone 39a (10%)
arising from oxidative cleavage of the tetralone ring, Figure 15. Repeating the reaction using degassed
toluene, in a Schlenk tube under argon led to suppression of 39a (trace amounts) and only the
naphthanilide 38c was isolated (along with recovered starting material). Although the mechanism
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 67
for the formation of 39a remains unclear, the oxidative α−CꢀC bond cleavage of 2ꢀsubstituted 1ꢀ
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
8
3
tetralones to give 40 under radical conditions (TEMPO) has been reported. The mechanism has been
reported to involve addition of oxygen to give an α−hydroperoxide. Radical or anionic fragmentation,
BaeyerꢀVilliger/Criegee fragmentation or baseꢀinduced fragmentation of the peroxide can all potentially
lead to the observed products (although a theoretical study indicates a radical fragmentation is most
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
83
likely).
In order to remove the complication of undesired tetralone cleavage the reaction of 28 was investigated
where the analogous oxidation was not possible. Reaction with Bu SnH and ACN lead to formation of
3
uncyclized 7c in 88% yield as the sole product.
O
208 ppm
1
730 cm-1
Bn
O
N
3
eq Bu SnH
3
5
x 0.2 eq
ACN
O
27b
1
67 ppm
7
b
O
O
reflux,
166 ppm
-1
O
1664 cm
-1
48 h
1707 cm
Bn
N
NHBn
39a
38c
O
O
0.02 M 1.5 eq,
Bn
Cl Bu SnH, 0.2 eq
Bn
3
N
N
ACN, toluene,
reflux 26 hr
88%
7c
28
OH
O
R
40
O
Figure 15. Attempted cyclization of 27b with Bu SnH.
3
24
ACS Paragon Plus Environment

Page 25 of 67
The Journal of Organic Chemistry
While it is most common to carry out Bu SnH mediated reactions at elevated temperatures (typically
3
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
benzene or toluene at reflux), efficient 5ꢀendo radicalꢀpolar crossover reactions (3h → 35 and 36 R =
Me) mediated by Cu(I) complexes have been reported at room temperature for tertiary halides 3h, Figure
66
1
3.
The ability to conduct such cyclizations at lower temperature is attractive if chirality transfer
during reaction of individual atropisomers is to be studied. Another advantage of generating the
reactive radical from homolysis of a CꢀX bond via a Cu(I) mediated atom transfer is that this initiating
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
process is reversible, effectively providing longer ‘lifetimes’ and less reduction of preꢀcyclized radicals
‡
than alternative Bu SnH mediated processes. As before the cyclization of substrates with ꢀG
<26
298 rot
3
ꢀ1
kcal mol proceeded smoothly, albeit at elevated temperatures. Reaction of 24a with either 30 mol% of
Cu(Me ꢀtren)Cl or Cu(TPMA)Cl in toluene or DCM at room temperature only led to trace amounts of
6
41a (<5%) after 48 hours. On the other hand, heating of 24a with one equivalent of Cu(TPMA)Cl in
toluene at reflux led to complete conversion to 41a in 70% yield. We next investigated the cyclizations
of 25b and 26b using the optimum conditions determined for 24a (100 mol% Cu(TPMA)Cl), Figure 16.
Cl
60-100 mol%
Cu(TPMA)X
Cl
Ph
24a
O
toluene, 110 C
N
Bn
41a
toluene,
reflux, 2 h
Bn
HO
Bn
HO
O
O
N
N
25b
Cu(TPMA)Cl
61%)
Cl
Cl
42a (48%)
Ph
Ph
Cl
(
H
42b (13%)
Bn
HO
Ph
Bn
HO
Ph
7.8%
O
O
N
N
toluene,
reflux, 2 h
Cl
Cl
H
26b
Cl
Cu(TPMA)Cl
3.4%
(
39%)
43a (23%)
43b (16%)
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 67
Figure 16. Cu(TPMA)Cl mediated 5ꢀendo cyclisation of 24a, 25b and 26b, showing nOe
enhancements (in %) for 43b.
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
Cyclization of both substrates was complete in 2 hours and produced the expected dichlorides 42a
1
and 43a along with the monochlorides 42b and 43b as single diastereomers (assigned from the H NOe
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
difference spectrum of 43b). Both sets of products arose from trapping of the intermediate acyl iminium
ion by water (as seen for the Bu SnH reaction of 26a). The monochloride 42b is likely formed by
3
reduction of the αꢀamide radical obtained from a second atom transfer from 42a to Cu(TPMA)Cl or by
electron transfer from the cyclized radical to 42a, (the driving force being the relief of the eclipsing
interactions between the gem dichloride group and the neighbouring quaternary centre).
O
Ph
OH
Bn
4 (36%)
O
N
Ph
45 (15%)
N
toluene,
reflux, 10 h
Bn
4
25c
Cu(TPMA)Br
(
55%)
O
N
Ph
Bn
46 (4%)
46
O
O
-H
N
Ph
Br
N
Ph
Bn
Bn
47
48
Me
Me
4
N
Ph
O
Me
3
Me
49
Figure 17. Cu(TPMA)Br mediated 5ꢀendo cyclisation of 25c.
Reaction of 25c with one equivalent of Cu(TPMA)Br in toluene at reflux for 10 hours produced
three products 44, 45 and 46 in 36%, 15% and 4% isolated yields respectively, Figure 17. The
26
ACS Paragon Plus Environment

Page 27 of 67
The Journal of Organic Chemistry
major product 44 arises via the intermediate acyl iminium ion 47 as previously observed for 25b while
the reduced product 45 may arise from abstraction of a hydrogen atom from toluene by the cyclised
radical. This would indicate that electron transfer in the radical–polar crossover step was slower than in
the trichloroacetyl derivatives 25b and 26b. The formation of the cyclopropyl compound 46 deserves
comment. There is a significant steric clash between the Cꢀ3 and Cꢀ4 gemꢀdimethyl groups in the acyl
imimium ion intermediate 49. In order to relieve these clashes, torsion of the Cꢀ3 to Cꢀ4 CꢀC bond can
occur which places the Cꢀ4 pseudoꢀaxial methyl group almost parallel to the pꢀorbital of the acyl
iminium ion 49, initiating a rearrangement to form a protonated cyclopropane intermediate 48.
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
Formation of a cyclopropane by loss of a proton during a 1,2ꢀmigration of a methyl group has been
84ꢀ85
previously reported
and this process would give rise to the observed product 46. More of the
cyclopropyl derivative was isolated from the analogous reaction of 26c, (50, 51 and 52 being formed in
3%, 21% and 5% yields respectively). Presumably, the more electron rich nature of methylene group
3
in 26c compared to the methyl group in 25c is responsible for the greater yield of trapped migration
product 51, Figure 18.
Figure 18. Products from Cu(TPMA)Br mediated 5ꢀendo cyclisation of 26c.
Reaction of the tetralone derivative 27c was relatively slow compared to the other substrates (25c and
2
6c) and required heating at reflux for 44 hours with 1.2 eq of Cu(TPMA)Br before all the starting
material was consumed. This substrate exhibited the highest barrier to Nꢀalkenyl bond rotation of the
substrates cyclized using Cu(TPMA)Br. Two products were isolated, the expected cyclized product 53
and the oxidatively ringꢀopened compound 39b (in a 2:1 ratio). Repeating the reaction in a Schlenk
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 67
tube with degassed solvent (3 x freezeꢀthaw cycles) suppressed the formation of 39b to trace levels, (<
%), indicating that dissolved O was most likely responsible for mediating the unusual ringꢀopening
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
process to give 39b. The reaction was more efficient, requiring less copper reagent (0.6 eq) and a shorter
reaction time (15 hours). In addition, two further products 54 and 55 were isolated in 31% and 7%
respectively (although it was not possible to obtain 55 completely free of impurities), Figure 19.
Unfortunately, attempts to conduct the reaction of 27c at a lower temperature (353K) provided trace
levels of cyclized products only, while the barrier to rotation around the Nꢀalkenyl bond of 27c at 383K
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
was too low (t = 18 s) to warrant investigating whether chirality transfer from one atropisomer to the
½
cyclized products was possible.
O
O
1.2 eq
Cu(TPMA)Br
toluene, reflux
BnN
HO
O
44 h
27c
O
O
53
NHBn
9b
53:39b = 1.0:0.5
3
0.6 eq
Cu(TPMA)Br
toluene
53 (19%) + 39b (<2%)
(
degassed),
O
O
27c
BnN
BnN
reflux, Ar
15 h
54 (31%)
55 (7%)
Figure 19. Reactions of 27c with Cu(TPMA)Br.
As with the formation of 39a, from the Bu SnH mediated reaction of 27b, the mechanism for the
3
formation of 39b remains unclear, although the oxidative α−CꢀC bond cleavage of 2ꢀsubstituted 1ꢀ
8
3
tetralones with either CuCl or CuCl and amine bases in the presence of O has been reported.
2
2
Conclusions
28
ACS Paragon Plus Environment

Page 29 of 67
The Journal of Organic Chemistry
In conclusion, we have prepared 38 different enamides 1c varying in the substitution around the acyl
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
1
3
4
5
group 1 (R ≠H), the nitrogen substituent (R ≠H), and alkene substituents (R , R , R ≠H). The majority
‡
show slow rotation around the Nꢀalkenyl bond of the enamide with ꢀG
barriers varying between
98 rot
2
ꢀ1
<
8.5 to 31.0 kcal mol . The most important factor in controlling the rate of rotation is the level of
3
4
5
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
alkene substitution (R , R , R ), followed by the size of the nitrogen substituent (R ) and finally the size
2
of the acyl substituent (R ). Electronic effects are small for substituents positioned on the acyl group
2
(
R ) and the rate of rotation was linearly correlated with the cone angle of the substituent for the studied
series 3aꢀi, indicating that the mechanism of rotation was likely to be the same for all the compounds.
For Nꢀbenzyl groups 13aꢀf, ortho haloꢀsubstitution has little effect on the barrier to rotation except
where two ortho substituents are present, indicating that coꢀoperative gearing of the nitrogen substituent
is likely during the Nꢀalkenyl rotation. A similar gearing for freely rotatable αꢀsubstituents (Ph in 7a) is
also indicated. Tetraꢀsubstituted alkenes possessing rigid α,α’ꢀsubstituents (7bꢀc) have high enough
barriers to rotation at room temperature to be separated by chiral HPLC with halfꢀlives of up to 99 years
at 298K. This should theoretically enable future investigations into asymmetry transfer from
enantiomerically pure enamides in a range of synthetic processes. For enamides 25bꢀc, 26bꢀc, 27c
containing large or electron withdrawing acyl groups (cone angles θ > approximately 130 ˚) both (E)ꢀ
R
and (Z)ꢀamide rotamers were detected in solution at room temperature. Interconversion of these two
rotamers was found to be fast on the NMR timescale at 80 ˚C for all the examples studied. An
examination of the barriers to rotation for the series 7b and 27aꢀc shows that while there is a steady
‡
ꢀ1
increase between with increasing size of the acyl substituent as expected (7b ꢀG
27.5 kcal mol , 27a
2
98
‡
ꢀ1
‡
ꢀ1
‡
ꢀ
G
27.8 kcal mol , 27b ꢀG
29.3 kcal mol ) there is a significant drop for 27c (ꢀG
25.6 kcal
298
2
98
298
ꢀ1
mol ) indicative of a coꢀoperative gearing effect between rotation around the (E)ꢀ and (Z)ꢀamide and
(M)ꢀ and (P)ꢀenamide rotamers with larger acyl substituents. This provides important temperature
constraints when attempting asymmetric reactions of similar enamides.
29
ACS Paragon Plus Environment