Ion pair dynamics: Solvolyses of chiral 1,3-diarylallyl carboxylates as a case study

Article abstract of DOI:10.1021/ja308670g

Chiral ion pairs play a key role in modern enantioselective synthesis, though little is known about their properties. We have now used the special features of unsymmetrically substituted allyl derivatives to obtain unprecedented insight into ion pair dynamics. By employing chiral high-performance liquid chromatography, it was possible to follow the time-dependent concentrations of all four isomeric esters (two regioisomeric pairs of enantiomers) and all four isomeric alcohols generated during the hydrolysis of enantiopure 1-(4-chlorophenyl)-3-phenylallyl and 3-(4-chlorophenyl)-1-phenylallyl 4-nitrobenzoates. Combination of these results with the directly measured rate constant for the reaction of the laser-flash photolytically generated 1-(4-chlorophenyl)-3-phenylallyl cation with water provided a complete mechanistic scheme for allyl carboxylate solvolysis. It is demonstrated that solvolysis and internal return can be explained by the same intermediates. The correlation equation log k = s_N(N + E) was used to elucidate the variable importance of external and internal return in the solvolysis reactions. This information will be crucial for the interpretation of the ultrafast dynamics of ion pairs generated by femtosecond laser pulses.

Full text of DOI:10.1021/ja308670g

Subscriber access provided by Milner Library | Illinois State University
Article
Ion Pair Dynamics: Solvolyses of Chiral 1,3-
Diarylallyl Carboxylates as a Case Study
Konstantin Troshin, and Herbert Mayr
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja308670g • Publication Date (Web): 28 Nov 2012
Downloaded from http://pubs.acs.org on December 10, 2012
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.
Journal of the American Chemical Society 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 46
Journal of the American Chemical Society
1
2
3
4
5
6
Ion Pair Dynamics: Solvolyses of Chiral
7
8
9
1,3-Diarylallyl Carboxylates as a Case Study
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
Konstantin Troshin and Herbert Mayr*
Department Chemie, Ludwig-Maximilians-Universität München, Butenandtstrasse 5-13
(Haus F), 81377 München, Germany
herbert.mayr@cup.uni-muenchen.de
ABSTRACT. Chiral ion pairs play a key-role in modern enantioselective synthesis though
little is known about their properties. We have now used the special features of
unsymmetrically substituted allyl derivatives to obtain unprecedented insight in ion pair
dynamics. By employing chiral HPLC it was possible to follow the time-dependent
concentrations of all four isomeric esters (two regioisomeric pairs of enantiomers) and all four
isomeric alcohols generated during the hydrolysis of enantiopure 1-(4-chlorophenyl)-3-
phenylallyl and 3-(4-chlorophenyl)-1-phenylallyl 4-nitrobenzoates. Combination of these
results with the directly measured rate constant for the reaction of the laser-flash
photolytically generated 1-(4-chlorophenyl)-3-phenylallyl cation with water provided a
complete mechanistic scheme for allyl carboxylate solvolysis. It is demonstrated that
solvolysis and internal return can be explained by the same intermediates. The correlation
equation log k = sN(N + E) was used to elucidate the variable importance of external and
internal return in the solvolysis reactions. This information will be crucial for the
interpretation of the ultrafast dynamics of ion pairs generated by femtosecond laser pulses.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 46
1
2
3
4
5
Introduction
6
7
8
9
Control of enantioselectivity by noncovalent interactions has become a major tool in organic
synthesis. In particular hydrogen bonding, as in thiourea-catalyzed reactions,1 and ion-pairing2
have extensively been used for stereoselective transformations. An early example for the use
of ion-pairing in asymmetric synthesis was reported by Dolling et al.,3 who found that a
cinchona-derived quarternary ammonium ion salt can be employed as catalyst for the
asymmetric methylation of a substituted indanone under phase-transfer conditions. This
pioneering work opened the field of asymmetric phase-transfer catalysis, the scope of which
has widely been elaborated by Maruoka and associates.4 Protonation of organic substrates by
strong chiral Brønsted acids, in particular BINOL-derived phosphoric acid diesters and their
derivatives, gives rise to chiral ion pairs or similar structures, which served as chiral
electrophiles in a manifold of reactions,5 including enantioselective cycloadditions,6
electrocyclic reactions,7 1,4-additions,8 Friedel-Crafts allylations,9 reductions,10 and ene
reactions.11 Hydrogen bonding between the resulting cation and the chiral counteranion often
provides additional stabilization of the positively charged electrophilic intermediate and
accounts for the augmented enantioselectivity. In some cases it is difficult to differentiate
whether chirality is induced by ion-pairing of an achiral cationic electrophile with a chiral
counteranion or by activation of neutral electrophiles (e. g., imines or carbonyl compounds)
through hydrogen bonding with a chiral Bronsted acid.12
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
However, treatment of prochiral substrates with strong chiral Bronsted acids is not the only
method to generate chiral ion pairs for asymmetric counterion-directed synthesis.2,13 Ion-
pairing was also employed in enantioselective iminium-activated reactions by using achiral
ammonium ions with chiral counterions14 as well as in various asymmetric transition-metal-
catalyzed reactions, which utilized the directing effect of chiral counteranions.15 Ooi
demonstrated that transition metal complexes with achiral ligands carrying a quarternary
onium moiety can electrostatically be bound to chiral anions to induce asymmetrical
2
ACS Paragon Plus Environment

Page 3 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
palladium catalysis.16 The general importance of ion-pairing in transition metal-catalyzed
reactions has been reviewed by Macchioni who suggested that ion-pairing should be
considered as one of the instruments for tuning catalytic processes.17
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
Enantioselective catalysis was furthermore achieved by combining ion-pairing with hydrogen
bonding. Examples are reactions catalyzed by chiral ureas which are coordinated to achiral
counterions18 as well as cooperative catalysis, where hydrogen bonding was used to link the
complex counterion to the active intermediate, thus providing additional attractive forces
within the ion pair.19, Extensive studies on the role of ion-pairing in carbocationic
polymerizations have recently been reviewed by Bochmann. 20
While the targeted use of ion-pairing in enantioselective synthesis and polymerization
reactions thus reflects recent developments, the importance of ion-pairing for organic
reactivity has already been recognized in the 1950s by Winstein and coworkers. Their studies
of SN1 solvolyses21 provide the basis of our current understanding of ion pairs and their
reactivities.22 Because of their ability to undergo allylic rearrangements, allyl derivatives
turned out to be particularly valuable systems for gaining insight in the nature and reactivities
of ion pairs.21a,22,23 Goering’s pioneering investigations of the transformations of optically
active allyl derivatives by titrimetric and polarimetric methods as well as product analyses
including isotope exchange experiments have become text-book examples24 to demonstrate
the role of ion-pairing in SN1 reactions.
Since the cis/trans isomerization of optically active cis- and trans-5-methylcyclohex-2-enyl 2-
carboxybenzoates (A
B, LG = OPht, Scheme 1) in acetonitrile was found to proceed
(R)-B),25
significantly more slowly than the racemization ((S)-A
(R)-A; (S)-B
Goering proposed the formation of ion pair intermediates (C, D), in which the anion remains
on the same face of the allyl cation as in the substrate. The collapse of these ion pairs (internal
return) either regenerates the starting materials or leads to their enantiomers by attack of the
anion at the other allylic terminus. The cis/trans-isomerization, which requires the migration
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 46
1
2
3
4
5
6
7
8
9
of the leaving group to the other face of the allyl cation, was proposed to proceed via
dissociation of the initial ion pairs to an achiral carbocationic intermediate (E) that can be
attacked by the leaving group from both faces (Scheme 1).25
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
Scheme 1. Ion Pair Mechanism Proposed by Goering for Solvolyses and Rearrangements of
cis- and trans-5-Methyl-cyclohex-2-enyl Derivativesa
a The descriptors (R) and (S) for the configuration of C-5 (connected to the methyl group) are
omitted.
The hydrolyses of cis- and trans-5-methylcyclohex-2-enyl 2-carboxybenzoates (A, B, LG =
OPht)26 and 4-nitrobenzoates (A, B, LG = OPNB)27 in aqueous acetone as well as the
ethanolyses and acetolyses of cis- and trans-5-methylcyclohex-2-enyl chlorides (A, B, LG =
Cl)28 were rationalized on the basis of this mechanism. In none of these cases cis/trans
isomerization of the non-reacted substrates was detected, and the polarimetric rate constant
4
ACS Paragon Plus Environment

Page 5 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
(racemization plus solvolysis) was always larger (factor of 1.1 to 5.07) than the titrimetric rate
constant (solvolysis only).
7
8
9
In contrast to the situation described for symmetrical allyl systems in Scheme 1, racemization
of unsymmetrical allyl derivatives implies migration of the leaving group to the other face of
the allyl cation. As solvolyses of trans-3-methyl-1-phenylallyl and trans-1-methyl-3-
phenylallyl 4-nitrobenzoates in aqueous acetone were later observed to be accompanied by
approximately 70% racemization of the unsolvolyzed esters, Goering concluded that the
previously investigated stereochemical behavior of cyclohexenyl derivatives (Scheme 1) was
largely dominated by conformational phenomena, which are absent in acyclic derivatives.29
The stereochemistry of SN1 reactions of acyclic allyl derivatives, which is crucial for the
understanding of ion-pairing in general, has, therefore, not conclusively been clarified up to
the present time.
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
Winstein's ion pair mechanism (Scheme 2),30 which has commonly been used to rationalize
the course of solvolysis reactions through the intermediacy of contact ion pairs (CIP), solvent
separated ion pairs (SSIP), and free ions, has recently been employed to interpret the
picosecond dynamics of laser-flash-generated contact ion pairs.31
Scheme 2. Classical Winstein Scheme for Solvolysis Reactions.30
It is commonly assumed that such photolytically generated contact ion pairs, which may
either be formed directly as the initial cleavage products or through electron transfer in the
initially generated geminate radical pairs, are similar to the intermediates in solvolysis
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 46
1
2
3
4
5
6
7
8
9
reactions.31 Vice versa, the rates measured for geminate recombinations of laser-flash
photolytically generated ion pairs, were used to discuss structural effects on the rates of
internal return during the solvolytic reactions. 32
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
However, Winstein's solvolysis scheme includes many parameters which could not be
unambiguously differentiated with the analytical methods available at that time so that many
questions remained open. It is still unclear, for example, whether solvolysis and internal return
proceed through the same intermediates or are two independent processes involving different
types of ion pairs.33
A crucial step toward a quantitative description of solvolysis reactions, and of nucleophilic
aliphatic substitutions in general, were the investigations of Jencks, Richard, and Tsuji,34-36
who used clock-methods to determine rate constants for the attack of solvents on the
intermediate carbocations. These authors were not only able to identify the change from SN1
to SN2 mechanisms34 but also clarified the dynamics of ion pair dissociation35 and
recombination.36
By introducing stopped-flow techniques for determining rates of solvolysis reactions, which
occur in the millisecond to second time domain,37 and systematic extension of the data set for
the rates of the reactions of carbocations with solvents38 and other nucleophiles,39 we arrived
at linear free energy relationships,39,40 which allow one to predict changes of solvolysis
mechanisms as the substrates and solvents are altered.41 Under the same conditions,
increasing stabilization of the carbocations led to the change from SN2 reactions, over SN1
reactions without and with common ion return, to SN2C+ processes (formation of carbocation
occurs faster than its reaction with the solvent) and heterolytic cleavages of esters which
proceed with formation of persistent carbocations.41 Furthermore, we have recently employed
femtosecond spectroscopy to investigate the dynamics of free and paired carbocations on the
picosecond time scale in collaboration with the Riedle group.42
6
ACS Paragon Plus Environment

Page 7 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Combination of all these techniques has provided detailed information about the whole range
of carbocation reactivities37,43 – from very slow reactions to those occurring within ion pairs
which proceed at rates approaching vibrational frequencies. On the other hand, our knowledge
about structures and dynamics of intramolecular interconversions of ion pairs has remained
crude.
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
We, therefore, approached the dynamics of ion pair transformations by taking advantage of
the special properties of allylic derivatives described above. While titrimetry and polarimetry
were the most important analytical tools available to Goering for studying the course of the
solvolysis reactions, we now employed chiral HPLC techniques to obtain unprecedented
insights in the course of SN1 reactions and ion-pairing.
When the enantiopure esters (R)-1 or (S)-2 were dissolved in aqueous acetone (for structures
see Scheme 3), mixtures of four isomeric allyl alcohols (hydrolysis products) and of four
isomeric allyl 4-nitrobenzoates (starting material and products of ion recombination) were
obtained. As we succeeded in separating these eight compounds by HPLC, it was possible to
follow the concentrations of each of these individual compounds as a function of time and
develop a kinetic model, which quantitatively describes the whole mechanistic scheme. As we
will report individual rate constants for interconversions between covalent substrates, ion
pairs, and free ions, this work can be considered as a bridge between the classical solvolysis
studies of the 1950s to 1970s and modern applications of ion-pairing in stereoselective
organic synthesis and their role in photosolvolytic processes. We, thus, can not only answer
questions which remained open when our knowledge of ion pairing effects in solvolysis
reactions was mostly derived from the difference between polarimetric and titrimetric rate
constants, but also provide the basis for the interpretation of ultrafast dynamics of ion pairs
generated by femtosecond laser pulses.
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 46
1
2
3
4
5
Results
6
7
Synthesis and Chromatographic Separation of the Model Compounds
8
9
The regioisomeric alcohols rac-3 and rac-4 were obtained by NaBH4 reduction of the
corresponding chalcones according to ref 44. Sharpless kinetic resolution45 of (rac)-3 and
rac-4 using D- or L-diisopropyltartrate (D-/L-DIPT), respectively, followed by treatment with
4-nitrobenzoyl chloride in the presence of triethylamine and recrystallization gave the
enantiopure (ee > 99%, HPLC) allylic esters (R)-1 and (S)-2 (Scheme 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
Scheme 3. Synthesis of the Enantiopure Allylic Esters (R)-1 and (S)-2 (PNB = 4-
nitrobenzoyl, DIPT = diisopropyltartrate, DMAP = 4-(dimethylamino)pyridine).
As treatment of either (R)-1 or (S)-2 with aqueous acetone may give a mixture of four esters
[(R)-1, (S)-1, (R)-2, and (S)-2] and of four alcohols [(R)-3, (S)-3, (R)-4, and (S)-4], a complete
analysis of the solvolysis mechanism requires monitoring of the concentrations of eight
compounds.
8
ACS Paragon Plus Environment

Page 9 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 1 shows that mixtures of the racemic compounds 1–4 can be resolved using chiral
HPLC (eight peaks in total), which gives the unique possibility of following the time-
dependent concentrations of the individual enantiomers.
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
3000
Acetone
(R )-3
2500
(S )-4
2000
(R )-4
1500
(S )-1
5
1000
(S )-2
(S )-3
(R )-2
(R )-1
500
0
0
10
20
30
40
50
60
70
80
90
100
t R / min
Figure 1. Chromatogram of an artificial mixture of racemic 1–4 and (E)-1-(4-methylphenyl)-
3-phenylprop-2-en-1-one (5) (internal standard). Details of the HPLC method are described in
the Supporting Information.
Kinetic experiments
General. The 4-nitrobenzoates 1 and 2 were solvolyzed in aqueous acetone at 25 °C with and
without addition of various nucleophiles (the experiments are summarized in Table 1).
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 46
1
2
3
4
Table 1. Summary of the HPLC kinetic experiments.
5
6
7
8
entry
1
2
3
4
5
6
7
8
solvent
substrate [substrate]0 / M
additive
[additive]0 / M
-
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
60 % aq acetone
80 % aq acetone
90 % aq acetone
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(S)-2
(S)-2
(R)-1
(R)-1
1.53 × 10-3
8.04 × 10-4
7.82 × 10-4
8.04 × 10-4
1.02 × 10-3
8.51 × 10-4
8.70 × 10-4
8.61 × 10-4
7.85 × 10-4
7.90 × 10-4
7.72 × 10-4
6.82 × 10-4
-
Bu4NOPNBa
Bu4NOPNBa
NaN3
Bu4NCl
Bu4NCl
Piperidine
LiClO4
-
4.82 × 10-3
5.25 × 10-2
7.08 × 10-2
1.87 × 10-2
1.01 × 10-1
5.42 × 10-2
9.46 × 10-3
-
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
9
10
11
12
NaN3
7.66 × 10-2
-
-
-
-
a Tetrabutylammonium 4-nitrobenzoate
Aliquots of the reaction mixtures were extracted with dichloromethane or diethyl ether after
certain time intervals followed by HPLC analysis. A typical chromatogram obtained during
such an experiment is shown in Figure S1 of the Supporting Information. By using (E)-1-(4-
methylphenyl)-3-phenylprop-2-en-1-one (5) as internal standard, we have determined the
time-dependent yields for all compounds present in the mixture.
Solvolyses of (R)-1 and (S)-2 in 60% Aqueous Acetone. Figure 2 shows that both
regioisomeric alcohols 3 and 4 are formed as racemates during the solvolysis of (R)-1 in 60%
aq acetone. The marginal separation of the graphs for (R)-3 and (S)-3 can be explained by the
different shapes of their HPLC signals, and the drift of the baseline caused by the gradient
used to shorten the overall elution time. As no isomerization was detectable within 2.75 h
when (S)-4 was dissolved in 60% aq acetone containing 2 mM of 4-nitrobenzoic acid (the
highest concentration of the acid which can be present at the end of the solvolyses under the
experimental conditions used), one can exclude enantioselective formation of 3 and 4 and
subsequent isomerization. In line with this observation, the ratio [3]/[4] = 1.5 remained
constant throughout the reaction. An SN2 mechanism as well as nucleophilic trapping of the
10
ACS Paragon Plus Environment

Page 11 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
chiral contact ion pairs (CIPs) by water prior to racemization (steps with ksI and ksII in the
classical Winstein scheme30) can, therefore, be excluded for this system: both pathways would
result in the formation of enantioenriched products (complete or partial inversion of the
configuration).
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
30
25
20
15
(S)-3
(R)-3
(S)-4
(R)-4
10
5
0
0
50
100
150
t / min
Figure 2. Time-dependent yields of 3 and 4 during the solvolysis of (R)-1 (1.5 mM) in 60%
aqueous acetone at 25°C.
Figure 3 shows that the consumption of (R)-1 does not only lead to the formation of the
hydrolysis products 3 and 4 discussed above but also to the intermediate appearance of the
isomeric 4-nitrobenzoates (S)-2, (S)-1, and (R)-2.
The prevailing formation of (S)-2 indicates that the leaving group stays preferentially at the
same face of the plane of the allyl cation which implies that allylic rearrangement occurs
predominantly at the contact ion pair stage.
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 46
1
2
3
4
5
6
7
8
9
(a)
(b)
4
100
80
60
40
20
0
(S )-2
3
2
1
0
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
(S )-1
(R )-2
0
50
100
150
0
50
100
150
t / min
t / min
Figure 3. Time-dependent yields of (a) (R)-1 and (b) (S)-2, (S)-1, and (R)-2 during solvolysis
of (R)-1 (1.5 mM) in 60% aq acetone, 25 °C.
Solvolysis of (S)-2 in 60% aqueous acetone followed exactly the same pattern as the
solvolysis of (R)-1 in the same solvent. Both alcohols 3 and 4 were formed as racemates, and
the ratio [3]/[4] = 1.5 is the same as in the solvolysis of (R)-1. One can, therefore, conclude
that the same achiral intermediates are responsible for the formation of the alcohols 3 and 4
from both precursors 1 and 2. As addition of 9.5 mM LiClO4 had no noticeable effect on the
time-dependent concentrations of 1-4 during the solvolysis of (R)-1 in 60% aq acetone (see pp
S44-S46 of the Supporting Information), the participation of SSIPs cannot play a significant
role (otherwise special salt effect30 would be expected). Therefore, the alcohols 3 and 4 must
be formed via the free 1-(4-chlorophenyl)-3-phenylallylium ions (6) or through ion pairs
which interconvert more rapidly via 6 than they react with water. The preferred formation of 3
over 4 can be explained by the charge distribution in 6 and does not reflect the relative
thermodynamic stabilities of 3 and 4 (a ratio of [3]/[4] = 0.88 was obtained by equilibration in
the presence of p-toluenesulfonic acid, see p S57 of the Supporting Information for details).
According to NBO calculations (see p S96 of the Supporting Information for details), the
positive charge is greater on the phenyl-substituted allyl terminus of the 1,3-diarylallyl cation,
as phenyl stabilizes carbocations better than 4-chlorophenyl. These observations are
12
ACS Paragon Plus Environment

Page 13 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
consistent with results obtained by Easton et al. for other unsymmetrical allyl derivatives.46
The stereospecificity of the allylic rearrangement of (S)-2 in 60% aqueous acetone is also
analogous to the previous case: (R)-1 is the major isomerization product followed by (S)-1
and (R)-2 (Figure S30 of the Supporting Information).
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
Solvolysis of (R)-1 in 80 and 90% Aqueous Acetone. The rate of consumption of (R)-1 in
acetone-water mixtures decreases from 60% to 80% and 90% aq acetone (krel = 73, 7, and 1,
respectively) as expected from the solvent ionizing power Y.47 Figure 4 shows that the yield of
rearranged esters increases considerably with decreasing water content in the solvent and that
the sequence [(S)-2] > [(S)-1] > [(R)-2] does not change, which can be explained by
decreasing dissociation abilities of the solvents (εr)48 and increasing nucleophilicities of
–OPNB from 60% to 80% and 90% aq acetone (as observed for acetate anion37).
(a)
8
7
6
5
4
3
2
1
0
80 % aq acetone
(S )-2
(S )-1
(R )-2
0
5
10
15
20
25
30
t / h
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 46
1
2
3
4
(b)
12
10
8
90 % aq acetone
(S )-2
5
6
7
8
9
(S )-1
(R )-2
6
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
4
2
0
0
40
80
120
160
t / h
(c)
8
7
6
5
4
3
2
1
0
90% acetone
80% acetone
60% acetone
0
2
4
6
8
t / t 1/2
Figure 4. Time-dependent yields of (S)-2, (S)-1, and (R)-2 during solvolysis of (R)-1 in (a)
80% aq acetone ([(R)-1]0 = 0.77 mM) and (b) 90% aq acetone ([(R)-1]0 = 0.68 mM) as well as
(c) yields of (R)-2 during solvolyses of (R)-1 in 60 (([(R)-1]0 = 1.5 mM), 80 ([(R)-1]0 = 0.77
mM), and 90% ([(R)-1]0 = 0.68 mM) aq acetone.49
Solvolysis of (R)-1 in the Presence of External Nucleophiles. When the solvolysis of (R)-1
in 60% aqueous acetone was performed in the presence of piperidine (54.2 mM), 40% of the
allyl cations 6 were intercepted by the amine leading to the formation of the regioisomeric
(E)-1,3-diarylallylpiperidines in ca 1:1 ratio (1H NMR), and the total yield of the alcohols
decreased to ca 60%. Non-regioselective formation of 1,3-diarylallylpiperidines can be
explained by diffusion-controlled reaction of piperidine with both allylic termini of the cation,
14
ACS Paragon Plus Environment

Page 15 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
which is in agreement with the prediction based on reactivity parameters of piperidine and 6.50
The same product ratio was found for the reaction of piperidine with the free cation 6 in
dichloromethane (see p S6 of the Supporting Information for details).
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
In the presence of sodium azide (70.8 mM), allyl azide was the major product of the
solvolysis reaction of (R)-1 in 60% aqueous acetone, and the total yield of 3 and 4 was only
about 3%. These observations show that the intermediates, which give rise to the formation of
3 and 4, can almost quantitatively be intercepted by external nucleophiles, which is in
agreement with the hypothesis that these intermediates are free 1,3-diarylallyl cations 6. On
the other hand, the various isomerization pathways of (R)-1 were differently affected by
external nucleophiles. The yield of (S)-2, which reached a maximum of 3.2% in the absence
of nucleophiles, was only slightly reduced to 2.6% in the presence of piperidine (54.2 mM),
and to 2.4% in the presence of NaN3 (70.8 mM). In contrast, the same concentrations of NaN3
reduced the formation of (S)-1 by a factor of 2.2 (Figure 5), and [(R)-2] levels below the
detection limit (not shown in Figure 5, see Figure S24a). An analogous situation was observed
for the solvolysis of (S)-2 in the presence of 76.6 mM NaN3 (Figure S33 of the Supporting
Information).
70.8 mM NaN3 54.2 mM piperidine
4
no additives
3
2
(S )-2
1
0
}
(S )-1
}
0
50
100
150
t / min
Figure 5. Time-dependent yields of (S)-2 (filled points) and (S)-1 (open points) generated
during solvolysis of (R)-1 (0.80 – 1.5 mM) without and with added nucleophiles (60% aq
acetone, 25 °C).
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 46
1
2
3
4
5
6
7
8
9
Addition of chloride ions reduced the yields of the rearranged diarylallyl 4-nitrobenzoates
(S)-1, (S)-2 and (R)-2 to similar extents (Supporting Information, Figures S19 and S22). Since
the 1,3-diarylallyl chlorides formed by trapping of the free cation 6 by Cl– undergo fast
dissociation to regenerate 6, the yields of the hydrolysis products 3 and 4 were not affected,
however.
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
Solvolysis of (R)-1 in the Presence of Tetrabutylammonium 4-Nitrobenzoate (Common
Ion Return). Small amounts of Bu4NOPNB (4.8 mM) reduced the rate of consumption of
[(R)-1] in 60% aq acetone by only 7% (Figure S9a of the Supporting Information), while the
yields of (S)-2 (Figure 6a), (R)-2 (Figure 6b) and (S)-1 (Figure 6c) increased by factors of 1.3,
2.5 and 2, respectively. When a high concentration of Bu4NOPNB (52.5 mM) was present, the
consumption of (R)-1 became significantly slower (factor of 0.65, common ion rate
depression), and the yield of (S)-2 increased by a factor of 2.5 (Figure 6a), while the yields of
(R)-2 and (S)-1 were approximately 8 (Figure 6b) and 6.5 (Figure 6c) times higher than in the
absence of 4-nitrobenzoate. In summary, Bu4NOPNB additives increased the yields of (R)-2
and (S)-1 by a significantly higher factor than the yield of (S)-2.
(a)
10
8
52.5 mM Bu4NOPNB
6
4.8 mM Bu4NOPNB
no additives
4
2
0
0
50
100
150
t / min
16
ACS Paragon Plus Environment

Page 17 of 46
Journal of the American Chemical Society
1
2
3
(b)
4
5
6
7
10
8
52.5 mM Bu4NOPNB
8
9
6
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
4
4.8 mM Bu4NOPNB
no additives
2
0
0
50
100
150
t / min
(c)
10
8
52.5 mM Bu4NOPNB
6
4
4.8 mM Bu4NOPNB
no additives
2
0
0
50
100
150
t / min
Figure 6. Time-dependent yields of (a) (S)-2, (b) (R)-2, and (c) (S)-1 generated during
solvolysis of (R)-1 (0.78– 1.5 mM) in the presence of various amounts of Bu4NOPNB (60%
aq acetone, 25 °C).
Reaction of the Diarylallyl Cation 6 with Water in Aqueous Acetone. In order to put the
trapping reactions of the intermediate diarylallyl cation 6 on an absolute scale, we have
directly measured the rate of consumption of laser flash photolytically generated 6 in 60% aq
acetone. As acetone has strong absorption in the UV region, excitation at λ < 310 nm, as used
in routine laser flash experiments,42 was not possible. For that reason, a modified procedure51
using a mixture of SuperDMAP-derived salts 8 and 9 which can be excited at 355 nm was
applied (Scheme 4).
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 46
1
2
3
4
5
6
Scheme 4. Laser Flash Photolytic Generation of the 1,3-Diarylallyl Cation 6 in 60% aq
7
8
9
Acetone.
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
N
N
N
HN
NH
Cl
HN
Cl
NH
ksolv
+
N
O
H2O,
355 nm
Cl
8
9
6
Cl
Cl
The reaction with water was followed spectrophotometrically at the absorption maximum of 6
(510 nm), and the rate constant for the reaction of 6 with water (ksolv= 1.34 × 107 s−1) was
obtained by fitting the time-dependent absorbance to the monoexponential function At =
A0e−k + C. As the ksolv values obtained at three different precursor concentrations agreed
t
solv
within the experimental error range (± 5%), the influence of the photoleaving group or
impurities, which might be present in the stock solution of the precursors, on the reaction
kinetics can be excluded.
Discussion
Let us now develop a mechanistic scheme which accounts for the experimental findings. The
observation that (S)-2 is the preferred rearrangement product during the solvolysis of (R)-1,
and vice versa, (R)-1 is the preferred rearrangement product during the solvolysis of (S)-2
indicates that there is a special pathway interconnecting these two isomers. Can it be a 1,3-
sigmatropic rearrangement that avoids the intermediate formation of allyl cations?
If this were the case, decreasing solvent ionizing power,47 i.e., changing from 60% aqueous
acetone to 80 and 90% aqueous acetone should decrease the yields of esters formed by ionic
pathways relative to those generated by a sigmatropic rearrangement. Comparison of Figures
3b, 4a, and 4b shows that the ratio [(S)-2]/([(S)-1] + [(R)-2]) even decreases in less ionizing
18
ACS Paragon Plus Environment

Page 19 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
solvents, which clearly rules out a sigmatropic rearrangement of (R)-1 into (S)-2 and vice
versa.
7
8
9
In line with Goering´s observations for cyclic allyl cations, we, therefore, conclude that the
preferred isomerization (R)-1 → (S)-2 proceeds via suprafacial migration of the carboxylate
anion at the contact ion pair stage.
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 5 shows that the addition of NaN3 (70.8 mM) and piperidine (54.2 mM) reduces the
yield of (S)-2 to a much smaller extent (factors of 1.3 and 1.2, respectively) than that of (S)-1
(factors of 2.2 and 1.5, respectively). (R)-2 shows a similar behavior as (S)-1 (Figure S24a of
the Supporting Information). Cl− ions exert similar effects (Figure S19a of the Supporting
Information). Vice versa, the addition of 4.8 mM Bu4NOPNB increases the maximum
concentration of (S)-2 by a factor of only 1.3 (Figure 6a), while the concentrations of (R)-2
(Figure 6b) and (S)-1 (Figure 6c) grow by factors of 2 to 2.5.
These observations indicate that the isomerizations of the 4-nitrobenzoates (R,S)-1 as well as
of (R,S)-2 proceed via two different pathways: one, which is affected by external nucleophiles
(including common ions, external return52), and one which is not affected by the presence of
external nucleophiles (internal return). The preferred rearrangement of (R)-1 into (S)-2, where
the leaving group stays on the same face of the allyl cation, is rationalized by internal return,
which is not 100% stereospecific, however, because part of (S)-1 and (R)-2 must also arise
from internal return. The latter conclusion is derived from the observation that 70.8 mM NaN3
reduces the yield of (S)-1 at the maximum of the curve in Figure 5 to 45% of the value
observed in the absence of additives, while the same concentration of azide ions reduces the
total yield of the hydrolysis products 3 and 4 from 100% to 3%. If (S)-1 would exclusively be
–
formed through external return, i.e., by trapping of 6 by OPNB, the yield of (S)-1 should be
reduced by a factor of 30, as the yields of the alcohols 3 and 4.
The isomer (R)-2 shows a similar behavior as (S)-1, but a precise evaluation of the small
quantities of (R)-2 is problematic because of the broadness of the HPLC peak of this isomer
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 46
1
2
3
4
5
6
7
8
9
(Figure 1). Complementary observations were made for the solvolysis of (S)-2 (see pp S50-
S51 of the Supporting Information for details). We, therefore, conclude that the
rearrangements through ion pairs do not only proceed via suprafacial migration of carboxylate
anion but also via migration of the carboxylate anion to the other face of the allyl cation
without dissociation to the free ions.
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
Table 2 shows that the time-dependent difference Δt = [(S)-2]t – [(R)-2]t is independent
(within experimental error) of the nature and concentration of the external nucleophile. If
external nucleophiles were able to attack the CIPs, the value of Δt could not be nucleophile-
independent. One can, therefore, conclude that the CIPs generated in this system are inert to
−
any additive used in the present work, including strong nucleophiles such as N3 .
Table 2. Difference Δt = [(S)-2]t – [(R)-2]t at Certain Reaction Times During Solvolysis of
(R)-1 in the Presence of Various Additives (60% Aq Acetone, 25 °C).
Additive
Δt
at 10 min at 20 min at 40 min at 50 min at 60 min
None
1.2
1.0
1.3
1.1
1.1
1.9
1.8
1.9
1.8
1.8
2.5
2.1
2.7
2.5
2.4
2.4
2.1
2.6
2.6
2.4
2.4
2.0
2.5
2.4
2.4
5 mM Bu4NOPNB
52.5 mM Bu4NOPNB
54 mM piperidine
70.8 mM NaN3
These observations exclude the solvolysis mechanism proposed by Dvorko et al.53 who
assumed that the azide anion generally attacks the contact ion pairs rather than SSIPs or
CSIPs (cavity separated ion pairs), which are proposed by Dvorko to be intermediates on the
way from CIP to SSIP in Scheme 2.
As discussed above, the formation of (S)-1 and (R)-2 from (R)-1 can be suppressed by strong
nucleophiles, such as NaN3, by more than 50 %. Internal return, therefore, is not the major
pathway for the formation of these isomers, and external return52 must be their main source,
particularly when the reactions are carried out in the presence of Bu4NOPNB. A similar
20
ACS Paragon Plus Environment

Page 21 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
situation is observed when (S)-2 is used as a substrate, where internal return favors the
formation of (R)-1 (suprafacial migration product). Table 3 shows that the ratio [(S)-1]/[(R)-2]
obtained by solvolysis of either (R)-1 or (S)-2 is the same within experimental accuracy,
indicating common intermediates from both precursors. This observation definitely excludes
that the rearrangements (R)-1 → (S)-1 and (S)-2 → (R)-2 are preferred over the
rearrangements (R)-1 → (R)-2 and (S)-2 → (S)-1, respectively. In other words, when the
leaving group migrates to the other face of the allyl cation, it has no preference for the carbon
from which it departs. The observation that the ratio [(S)-1]/[(R)-2] is comparable to the ratio
of the hydrolysis products ([3]/[4] = 1.5), supports the suggestion that external return of
–OPNB (the major source of (S)-1 and (R)-2 in the presence of Bu4NOPNB) and hydrolysis
proceed via the same key intermediates, i.e., the free 1,3-diarylallyl cations 6.
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
Table 3. The Ratios [(S)-1]/[(R)-2] at Certain Reaction Times During Solvolysis of (R)-1 and
(S)-2 in the Presence of Various Additives (60% Aq Acetone, 25 °C).
substrate, additive
[(S)-1]/[(R)-2]
at 50 min at 60 min
(R)-1, no additive
(R)-1, 4.8 mM Bu4NOPNB
(R)-1, 52.5 mM Bu4NOPNB
(S)-2, no additive
1.40
1.28
1.37
1.35
1.41
1.29
1.38
1.50
A summary of these observations is presented in Scheme 5. The ioniziation step (k1 starting
from 1 or k´1 starting from 2) provides contact ion pairs ((R)-CIP1, (S)-CIP1, (R)-CIP2, (S)-
CIP2) which retain the stereochemical and regiochemical information of the covalent
substrates, i.e., the 4-nitrobenzoate anion is still on the same face of the carbocationic plane,
close to the carbon to which it was covalently bound in the starting material. While these
unsymmetrical structures of the ion pairs are in agreement with previous suggestions by
Goering54 and Thibblin,55 the distinction of four different ion pairs is in line with, but not
inevitably required by our experimental data. Instead of assuming the rapidly equilibrating
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 46
1
2
3
4
5
6
7
8
9
pairs (R)-CIP1 (S)-CIP2 and (S)-CIP1 (R)-CIP2, one might also assume that the same
chiral ion pair is formed from (R)-1 and (S)-2, which is enantiomeric to that generated from
(S)-1 and (R)-2. The latter alternative is kinetically equivalent to the mechanism in Scheme 5
(four different CIPs) with kr = k´r = ∞. According to Scheme 5, the unsymmetrical contact ion
pairs can undergo recombination (k−1 and k´−1), suprafacial migration (kr and k'r), inversion (ki,
k'i, the anion migrates to the opposite face of the allyl cation), and dissociation with formation
of free cations 6 (k2, k'2). The free cations can re-associate with 4-nitrobenzoate anions
regenerating the contact ion pairs with the second-order rate constants k−2 and k´−2, or react
with other nucleophiles (represented in Scheme 5 with the effective first order rate constant
kNu which corresponds to the sum of reactions with all external nucleophiles and water) to
produce the racemic products 3, 4, 7, and 7´. As the alcohols 3 and 4 are formed as racemates,
nucleophilic trapping of the chiral CIPs can be excluded in Scheme 5. Trapping of ion pairs
by external nucleophiles must be taken into account, however, when ion pairs of less
stabilized carbocations are involved.56
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
22
ACS Paragon Plus Environment

Page 23 of 46
Journal of the American Chemical Society
1
2
3
4
Scheme 5. Mechanism for the Solvolysis of Allyl 4-Nitrobenzoates 1 and 2.a
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
a The rate constants shown in parentheses result from the fit described below. All first-order
rate constants as well as the pseudo-first-order rate constant ksolv are given in s−1 and
correspond to 60% aq acetone, 25 °C. Second-order rate constants k−2 and k´−2 are given in
M−1 s−1.
The gross rate constant of the diffusional encounter of 6 with –OPNB ( kdOifPfNB ) is expressed by
equation 1.
diff
OPNB
k
= 2(k−2 + k'−2 )
(1)
−
The factor of 2 implies that the encounter of 6 with OPNB is split into two equal pathways
leading to the enantiomeric ion pairs.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 46
1
2
3
4
The interconversions depicted in Scheme 5 can be described by the rate laws in eq 2 to eq 10.
5
6
7
8
9
d[(R) -1]
= −k1[(R) -1]+ k−1[(R) -CIP ]
(2)
(3)
(4)
(5)
1
dt
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
d[(S) -1]
= −k1[(S) -1]+ k−1[(S) -CIP ]
1
dt
d[(R) - 2]
= −k'1 [(R) - 2]+ k'−1 [(R) -CIP2 ]
dt
d[(S) - 2]
= −k'1 [(S) - 2]+ k'−1 [(S) -CIP2 ]
dt
d[(R) - CIP ]
= k1[(R) -1]+ k'r [(S) - CIP2 ]+ ki [(S) - CIP ]+ k−2[6][− OPNB]− (k−1 + kr + ki + k2 )[(R) - CIP ]
1
(6)
(7)
1
1
dt
d[(S) -CIP ]
= k1[(S) -1]+ k'r [(R) -CIP2 ]+ ki[(R) -CIP ]+ k−2[6][− OPNB]− (k−1 + kr + ki + k2 )[(S) -CIP ]
1
1
1
dt
d[(R) -CIP2 ]
= k'1 [(R) - 2]+ kr [(S) -CIP ]+ k'i [(S) -CIP2 ]+ k'−2 [6][− OPNB]− (k'−1 +k'r +k'i +k'2 )[(R) -CIP2 ]
(8)
1
dt
d[(S) -CIP2 ]
= k'1 [(S) - 2]+ kr [(R) -CIP ]+ k'i [(R) -CIP2 ]+ k'−2 [6][− OPNB]− (k'−1 +k'r +k'i +k'2 )[(S) -CIP2 ]
(9)
1
dt
d[6]
= k2[(R) - CIP ]+ k2[(S) - CIP ]+ k'2 [(R) - CIP2 ]+ k'2 [(S) - CIP2 ]− (2k−2[− OPNB] + 2k'−2 [− OPNB] + kNu )[6]
(10)
1
1
dt
The solution of the system of linear ordinary differential equations 2 to 10 provides the
calculated values for the time-dependent concentrations of (R)-1, (S)-1, (R)-2, and (S)-2 for a
given set of parameters (k1, k´1, k–1, k´–1, kr, ki, k´r, k´i, k2, k´2, k–2[–OPNB], k´–2[–OPNB],
kNu).57 In order to determine the individual rate constants shown in Scheme 5, we have
simulated the time-dependent concentrations of the four isomeric esters (R)-1, (S)-1, (R)-2,
and (S)-2 during the solvolysis in 60% aq acetone of:
• (R)-1 (0.80 mM) in the presence of 70.8 mM NaN3, which provides reliable data for
–
internal return because external return of OPNB is almost completely suppressed
(k–2[−OPNB] << kNu= kN3[N3 ] + ksolv).
–
• (S)-2 (0.79 mM) in the presence of 76.6 mM NaN3.
• (R)-1 (0.80 mM) in the presence of 4.8 mM Bu4NOPNB (to keep k–2[–OPNB] and
k´–2[–OPNB] constant during the reaction, which provides the reliable data for external
return).
24
ACS Paragon Plus Environment

Page 25 of 46
Journal of the American Chemical Society
1
2
3
4
The directly measured rate constant of the reaction of 6 with water in 60% aq acetone (ksolv
=
5
6
7
8
1.34 × 107 s–1) and the second-order rate constant kN3 (6.1 × 109 M–1s–1), which was derived
from ksolv and the allyl azide/allyl alcohol ratio, were introduced as fixed quantities.
As the free energies of the ion pairs CIP1 and CIP2 are closely similar, the corresponding rate
constants for diffusion and ion pair reorganization were set equal, i.e., k´2 = k2, k´–2 = k–2, k´r =
kr, and k´i = ki. The small errors introduced by these assumptions are compensated by the
relative magnitudes of the recombination rate constants k´–1/k–1.
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
Minimization of the sum of squared deviations (SSD) between calculated and experimental
time-dependent concentrations of all isomeric esters (R,S)-1 and (R,S)-2 yielded the values of
the rate constants which fit the experiments most correctly.
As shown in Table SN1 (page S67) of the Supporting Information k2 = k´2 was arbitrarily set
at values between 108 and 1011 s–1, while the remaining parameters were optimized. The last
column of the Table SN1 shows that equally good fits between calculated and experimental
concentrations were obtained for the different values of k2 = k´2. While k1, k´1, k–2, and k´–2
were found to be independent (± 4%) of the choice of k2 in the specified range, k–1, k´–1, and ki
= k´i were found to be directly proportional to k2 (Figures SN1 to SN8 of the Supporting
Information), which allowed us to express these rate constants as multiples of k2 in Scheme 5.
For a fixed value of k2 = k'2 = 2 × 1010 s–1, comparable SSDs were obtained for different
values of kr as long as they were greater than 10k2, and the values of k–1, k´–1, ki = k´i changed
insignificantly (< 7%) when kr = k'r was varied from 10k2 to 500k2, (Figures SN9 to SN15 of
the Supporting Information). As a consequence, each value of k2 entails certain values (±7%)
of k–1, k´–1, ki = k´i and lower limits for kr = k´r. The same results were obtained when the
steady state approximation was applied to (R)-CIP1, (S)-CIP1, (R)-CIP2, (S)-CIP2, and 6
(Figures SN1a to SN8a of the Supporting Information). The resulting absolute and relative
rate constants are presented in Scheme 5.
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 46
1
2
3
4
5
6
7
8
9
The good agreement between calculated and experimental time-dependent concentrations
(Figure 7 and Figures SN18 to SN19 of the Supporting Information) demonstrates that the
solvolyses of 1 and 2 can adequately be described by the mechanism presented in Scheme 5.
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)
(b)
100
80
60
40
20
0
4
3
2
1
0
(S )-2
(S )-1
(R )-1
(R )-2
(S )-2
0
50
100
150
0
50
100
150
t / min
t / min
Figure 7. Calculated (solid lines) and experimental time-dependent concentrations of
(a) (R)-1 (0.80 mM) and (S)-2 (0.79 mM) during their solvolysis in 60% aq acetone in the
presence of 4.8 mM Bu4NOPNB and 76.6 mM NaN3, respectively; and (b) of (S)-1, (S)-2, and
(R)-2 during solvolysis of (R)-1 (0.80 mM) in 60% aq acetone in the presence of 4.8 mM
Bu4NOPNB, 25 °C.
According to Scheme 5, the slowest step of the solvolysis is the initial ionization leading to
the CIPs (k1 = 3.5 × 10–4 s–1, k'1 = 3.8 × 10–4 s–1). Suprafacial migration of the 4-nitrobenzoate
anion (kr = k´r > 10k2) is the most likely transformation of the CIP, followed by dissociation
(k2 = k'2) and, finally, inversion (ki = k'i = 0.23k2) and ion pair collapse (k−1 = 0.21k2, k´–1 =
0.18k2), which have almost equal rates. This sequence explains the partial stereospecificity of
internal return, i.e., the fact that (S)-2 is the major product among the rearranged esters during
solvolysis of (R)-1 and vice versa.
Because of the availability of the directly measured rate constant ksolv for the reaction of 6
with water (in aq acetone), our experiments provide an accurate value for the diffusional
26
ACS Paragon Plus Environment

Page 27 of 46
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
process generating ion pairs from the free ions 6 (k–2 = k'–2). On the other hand, the rate
constants for the diffusional separation of the ion pairs (k2 = k'2) cannot be derived directly
from the experimental data. As values of 108 < k2/s–1 < 1011 give equally good fits, the value
of
k2 ≈ 1.6 × 1010 s–1, which was proposed by Richard and Jencks35 and is mostly used in the
literature, appears to be a good choice also for this system.
From the value of k−2 = k´–2 = 9.1 × 108 M–1s–1 one can calculate the second-order rate
constant of diffusional migration of 4-nitrobenzoate anion to the free cation 6 using eq 1
OPNB
diff
(k
= 4k–2 = 3.6 × 109 M–1s–1), which is similar to the value of 1.5 × 109 M–1s–1 reported by
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
Tsuji, Richard, and co-workers for the diffusion of carboxylate anions to the 1-(4-
methylphenyl)ethyl cation in 50% v/v TFE-water mixture.36a
Earlier analyses, which were based on titrimetric, polarimetric, and 18O-exchange rate
constants, left the question open whether solvolysis and internal return are two independent
processes involving different types of ion pairs.33 As simulations based on Scheme 5
accurately describe the distribution of the products generated by internal and external return,
it is now clear that solvolysis and internal return can be explained by the same intermediates.
A Comprehensive View on Solvolyses of Allyl Carboxylates.
The kinetic and stereochemical investigations of the solvolyses of the enantiopure allyl
carboxylates (R)-1 and (S)-2 provided detailed information on relative and absolute rates of
the individual steps for the system described in Scheme 5. Can one use these results to derive
a general scheme of solvolyses of allyl carboxylates?
In previous work,44 we have determined the electrophilicity parameters E of the symmetrical
1,3-diarylallyl cations X-10, which are listed in Table 4. According to equation 11, the
electrophilicity parameters E can be combined with the solvent-dependent nucleophile-
specific parameters N and sN to calculate second-order rate constants for the reactions of
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 46
1
2
3
4
5
carbocations with neutral and anionic nucleophiles39 as well as the first-order rate constants
for the reactions of carbocations with the solvents.38
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
log k (20 °C) = sN (E + N)
(11)
Table 4. Symmetrical 1,3-Diarylallyl Cations X-10 and Their Electrophilicity Parameters E.44
X
X
10
X -
X-10
X
E
F2-10
F-10
m,m-F2
m-F
6.11
4.15
2.85
2.69
2.70
1.23
-1.45
Br-10
Cl-10
H-10
p-Br
p-Cl
H
Me-10
MeO-10
p-Me
p-MeO
Me2N-10
p-Me2N
-7.50
As the 1,3-diphenylallyl cation H-10 and its dichloro-substituted analogue Cl-10 have almost
the same values of E, the same electrophilicity (E = 2.70) can also be assumed for the
monochlorinated system 6.
Rates of Reactions of the Allyl Cations 10 with Aqueous Acetone. Investigations of the
nucleophilic reactivities of solvents have shown that acetonitrile/water mixtures with 20% to
90% content of water (v/v) react with equal rates with benzhydrylium ions,38 in accordance
with earlier reports by McClelland.58 The same relationship seems to hold also for
acetone/water mixtures, as 90% (N = 5.70, sN = 0.85) and 80% aq acetone (N = 5.77, sN =
0.87) were reported to react with similar rates.59 Accordingly, the first-order rate constant for
the reaction of 6 (E = 2.70) with 80% aq acetone calculated by eq 11 (2.3 × 107 s–1) agrees
well with the directly measured rate constant for the reaction of 6 with 60% aqueous acetone
28
ACS Paragon Plus Environment

Page 29 of 46
Journal of the American Chemical Society
1
2
3
4
5
(1.34 × 107 s–1, see above). Equation 11 can thus be employed also to calculate the rate
constants for the reactions of 10 with aqueous acetone.
6
7
8
9
It should be noted that the agreement within a factor of 2 between the calculated and
experimental rate constant for the reaction of 6 with aqueous acetone cannot a priori be
expected, because deviations up to factors of 10–100 have to be tolerated for predictions of
absolute rate constants by equation 11, which covers a reactivity range of 40 orders of
magnitude with only three parameters.41,60 On the other hand, equation 11 allows one to
predict relative reactivities within reaction series, e.g., the relative reaction rates of X-10,44
with an accuracy better than factor of 2.43,60
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
Rates of Reactions of the Allyl Cations 10 with the 4-Nitrobenzoate Anion. In order to
apply equation 11, let us first derive the nucleophilicity parameter N for the 4-nitrobenzoate
anion (–OPNB) in 60% aq acetone. At low concentrations of the substrates, as they are usually
employed in solvolysis experiments, the concentration of ion pairs (corresponding to
encounter-complexes in ion-molecule or molecule-molecule reactions) is small compared
with the concentrations of the non-paired reactants, and the rate constant for the
–
recombination of 6 with OPNB to the covalent products 1 and 2 is given by eq 12, which
expresses the rate constants for ion recombination krec by multiplying the constant of the
diffusional association (k–2) with the partitioning factor (forward reaction, k–1, divided by the
sum of forward and backward reactions, k–1 + k2) and the corresponding term for attack at the
other allyl terminus.
k−1
k'−1
krec = 2k−2
+ 2k'−2
(12)
k−1 + k2
k'−1 +k'2
Substitution of k–1, k´–1, k–2, and k´–2 by the absolute values or multiples of k2 presented in
Scheme 5 yields krec = 5.91 × 108 M–1s–1 for the reaction of 6 with –OPNB in 60% aq acetone.
29
ACS Paragon Plus Environment