Clarification of the stereochemical course of nucleophilic substitution of arylsulfonate-based nucleophile assisting leaving groups

Article abstract of DOI:10.1021/jo900991z

(Chemical Equation Presented) Secondary alcohols modified as tosylates, PEG-sulfonates, or quisylates undergo inversion of configuration at the reacting center when treated with lithium halide in acetone at reflux, where the PEG-sulfonates and quisylates are substantially more reactive. In sterically hindered cases, elimination is a competing process. In contrast, when treated with TiCl₄, simple secondary sulfonates give chloride products with partial inversion of configuration. Any observed retention of configuration in a given alkyl sulfonate substrate under these conditions is likely due to neighboring group participation or diastereoselective attack on a carbocation (or ion pair) rather than an S_Ni mechanism.

Full text of DOI:10.1021/jo900991z

pubs.acs.org/joc
Clarification of the Stereochemical Course of Nucleophilic
Substitution of Arylsulfonate-Based Nucleophile Assisting
Leaving Groups
D. Christopher Braddock,*,† Rebecca H. Pouwer,† Jonathan W. Burton,*,‡ and
Phillip Broadwith‡
†Department of Chemistry, Imperial College London, London SW7 2AZ, U.K., and ‡Chemistry Research
Laboratory, 12 Mansfield Road, Oxford OX1 3TA, U.K.
c.braddock@imperial.ac.uk; jonathan.burton@chem.ox.ac.uk
Received May 13, 2009
Secondary alcohols modified as tosylates, PEG-sulfonates, or quisylates undergo inversion of
configuration at the reacting center when treated with lithium halide in acetone at reflux, where
the PEG-sulfonates and quisylates are substantially more reactive. In sterically hindered cases,
elimination is a competing process. In contrast, when treated with TiCl4, simple secondary sulfonates
give chloride products with partial inversion of configuration. Any observed retention of configura-
tion in a given alkyl sulfonate substrate under these conditions is likely due to neighboring group
participation or diastereoselective attack on a carbocation (or ion pair) rather than an SNi
mechanism.
Introduction
pairing under any conditions employed. This is especially
true when planning syntheses of complex targets where
an inversion, (partial) racemization, or retention could
dramatically alter the entire strategy for the synthe-
sis. Accordingly, the development of new leaving groups
for nucleophilic substitution processes with predictable
stereochemical control remains an important research
area.3
Recently, Leporehasshownthat2-(2-methoxyethoxy)ethyl
2-(alkoxysulfonyl)benzoates (PEG-sulfonates, ROPs) 1, pre-
pared in one step from an alcohol, ROH, undergo nucleo-
philic displacement at carbon with halide anion.4 Notably,
Nucleophilic substitution at an sp3 hybridized carbon
bearing a leaving group represents a fundamental transfor-
mation in organic chemistry, where the stereochemical
signatures of Walden inversion for the SN2 mechanism or
partial racemization via a planar carbocation and asso-
ciated counterion in the SN1 mechanism are fundamentally
associated with each process.1 The third possible stereo-
chemical outcome for substitution is retention of con-
figuration, where this mechanism is designated SNi.2 It is
clearly of the highest importance to be able to predict the
stereochemical outcome of a given nucleophile-substrate
(3) For an excellent recent review on “recent advances in heterolytic
nucleofugal leaving groups”, see: Lepore, S. D.; Mondal, D. Tetrahedron
2007, 63, 5103–5122 and references therein.
(1) Smith, M. B.; March, J. March’s Advanced Organic Chemistry;
Wiley: New York, 2007.
(2) For a recent example, see: Moss, R, A.; Fu, X.; Tian, J.; Sauers, R.;
Wipf, P. Org. Lett. 2005, 7, 1371–1374 and references therein.
(4) Lepore, S. D.; Bhunia, A. K.; Cohn, P. J. Org. Chem. 2005, 70, 8117–
8121.
6042 J. Org. Chem. 2009, 74, 6042–6049
Published on Web 07/10/2009
DOI: 10.1021/jo900991z
r
2009 American Chemical Society

Braddock et al.
JOCArticle
these PEG-sulfonates show acceleration of reaction rate
compared to the electronically similar tosylates and show
selectivity for lithium halides compared to sodium or potas-
sium salts.
with lithium bromide to give 2-bromoadamantane but the
2-adamantyl tosylate did not react was taken as evidence of
an SNi mechanism involving a macrocyclic transition state
where the SN2 pathway is “precluded due to steric shield-
ing”.4 In further work, Lepore reported that selected PEG-
sulfonates give secondary alkyl chlorides with retention of
configuration on treatment with titanium tetrachloride and
suggested that these reactions also proceed via an SNi
mechanism.5 Interestingly, under these conditions, a neo-
pentyl PEG-sulfonate was found to give a rearranged pro-
duct. This method has since been utilized by D. Kim et al.6 as
the key step to convert alcohol 3 to bromide 4 using titanium
tetrabromide with observed retention of configuration in the
total synthesis of (+)-microcladallene B. In 2008, Lepore
reported that the use of selected PEG-sulfonates 1 or addi-
tionally alkyl quinoline-8-sulfonates (quisylates, ROQs) 5
also allowed for introduction of azide or bromide with
retention of configuration using titanium(IV) salts via a
proposed SNi mechanism.7 Thus, the possible conclusion
drawn from the current literature is that PEG-sulphonates
and quisylates of any secondary alcohol should undergo
substitution (by an SNi mechanism) with retention of con-
figuration. In connection with both the London and Oxford
groups’ interest in the synthesis of complex halogenated
natural products, we became interested in the use of these
PEG-sulfonate groups for the potential introduction of
chloride or bromide to access the obtusallene family of
natural products8 and elatenyne,9 respectively. As a result
of that interest, herein we clarify the stereochemical course of
nucleophilic substitution of arylsulfonate-based nucleophile
assisting leaving groups.
At the outset of our investigations, we considered that the
two different types of reaction conditions reported by
Lepore, metal halide in acetone4 versus titanium(IV) salt in
a chlorinated solvent,5,7 could result in different stereochem-
ical outcomes. In particular, it was notable that there were no
reports on the use of PEG-sulfonates or quisylates of simple
aliphatic enantiopure secondary alcohols under either of
these conditions in the Boca Raton papers,4,5,7 thereby
defining the stereochemical outcomes for the simplest possi-
ble system with a single stereocenter. The finding that the
PEG-sulfonate of 2-adamantanol (2) underwent reaction
with lithium bromide in acetone to give the corresponding
bromide4 does not necessarily preclude inversion of config-
uration at the reacting center, since epimeric 4,4-dimethyl-2-
adamantyl tosylates have been reported to yield products
with a high degree of inversion on treatment with azide as the
nucleophile.10 When using titanium(IV) salts in conjunction
(5) Lepore, S. D.; Bhunia, A. K.; Mondal, D.; Cohn, P. C.; Lefkowitz, C.
J. Org. Chem. 2006, 71, 3285–3286.
(6) Park, J.; Kim, B.; Kim, H.; Kim, S.; Kim, D. Angew. Chem., Int. Ed.
2007, 46, 4726–4728.
(7) Lepore, S. D.; Mondal, D.; Li, S. Y.; Bhunia, A. K. Angew. Chem., Int.
Ed. 2008, 47, 7511–7514 and references therein.
(8) For recent synthetic work on the obtusallenes, see: (a) Braddock,
D. C.; Bhuva, R.; Millan, D. S.; Perez-Fuertes, Y.; Roberts, C. A.; Sheppard,
R. N.; Solanki, S.; Stokes, E. S. E.; White, A. J. P. Org. Lett. 2007, 9, 445–448.
ꢀ
(b) Braddock, D. C.; Millan, D. S.; Perez-Fuertes, Y.; Pouwer, R.; Sheppard,
R. N.; Solanki, S.; White, A. J. P. J. Org. Chem. 2009, 74, 1835–1841.
(9) For recent synthetic work directed towards the synthesis and structure
determination of elatenyne, see: (a) Sheldrake, H. M.; Jamieson, C.; Burton,
J. W. Angew. Chem., Int. Ed. 2006, 45, 7199–7202. (b) Sheldrake, H. M.;
Jamieson, C.; Pascu, S. I.; Burton, J. W. Org. Biomol. Chem. 2009, 7, 238–
252.
Lepore proposed that the polyether chain coordinates the
metal cation, simultaneously stabilizing the negative charge
on the leaving sulfonate and attracting the nucleophile to
effect overall substitution. The finding that the correspond-
ing PEG-sulfonate of 2-adamantanol (2) underwent reaction
(10) Banert, K.; Kurnianto, A. Chem. Ber. 1986, 119, 3826–3841.
J. Org. Chem. Vol. 74, No. 16, 2009 6043

Braddock et al.
JOCArticle
with PEG-sulfonates,5,7 no stereochemical information can
be extracted from the employment of the reported PEG-
sulfonates of achiral 2-adamantanol (2), 2-cycloheptanol (6),
1,3-diphenyl-2-propanol (7), and cyclododecanol (8) sub-
strates. On the other hand, the PEG-sulfonate of neopentyl
alcohol 9 was found to react with TiCl4 to give a rearranged
chloride,5 implicating the intermediacy of a carbocation
followed by a Meerwein-Wagner shift. It was also apparent
that the reported PEG-sulfonates and quisylates derived
from cholesterol (10)11 and (S)-1-phenylbutan-2-ol (11)
could possibly react by neighboring group participation
under these conditions, via cation 1212 and phenonium ion
13,13 respectively, leading to the observed retention of con-
figuration in these substrates by double inversion. The final
group of reported PEG-sulfonate and quisylate modified
substrates, menthol (14),14 isomenthol (15),14 and (()-trans-
2-methylcyclohexanol (16) (the corresponding neomenthol
substrate was not reported), could all be subject to diaster-
eoselective halogenation/aziridination under these condi-
tions via carbocations (or ion pairs) and attack at the least
hindered face leading also to the observed retention of
configuration. The above considerations could also apply
to the conversion of alcohol 3 into bromide 4 under the
reported conditions. Interestingly, the PEG-sulfonate and
quisylate of S-ethyl lactate 17 was found to undergo inver-
sion of configuration when treated with titanium(IV) bro-
mide, and it was noted that this suggests a change in
mechanism.7,15 Herein, we report on our investigations from
London and Oxford where our first objective was to deter-
mine the stereochemical course of reaction under each of the
different reaction conditions. Additionally, we desired to
compare and contrast tosylates’, PEG-sulfonates’, and qui-
sylates’ behavior for a given alcohol substrate. Accordingly,
we show that tosylates, PEG-sulfonates 1, and quisylates 5 of
enantiopure secondary alcohols (S)-octan-2-ol (18) and (R)-
decan-2-ol (19) all undergo inversion of configuration at the
reacting center when treated with lithium chloride to give
their corresponding chlorides. The same chlorides are ob-
tained from the same substrates when treated with titanium
tetrachloride but with partial racemization. Alkenes and the
products of hydride shifts are also obtained under these latter
conditions. We also present our results of nucleophilic dis-
placement of the tosylates, PEG-sulfonates, and quisylates
of (1R,2S,5R)-menthol (14), (1S,2S,5R)-neomenthol (20),
and 3β-chlolestanol (21) using LiCl or TiCl4, and tosylates,
PEG-sulfonates, and quisylates derived from the hydroxy
tetrahydrofurans 22 and 23 on treatment with lithium bro-
mide in acetone at reflux.
Results and Discussion
Tosylates 24,16 27, 30,17 33,17 36,18 39, and 42 were
prepared by standard procedures from alcohols 18, 19, 14,
20, 21, 22, and 23, respectively; the racemic cis-alcohol 2219
was prepared from cis-2-(benzyloxymethyl)-3-(methoxyme-
thyl)oxirane according to a literature procedure19 with the
racemic trans-alcohol 23 being prepared from the cis-alcohol
22 by a Mitsunobu/hydrolysis sequence. PEG-sulfonates 25,
28, 31,5 37, 40, and 43 were prepared from the same alcohols
using the method of Lepore.4 The corresponding quisy-
lates 26,20 32,20 38, 41, and 44 were prepared by analogous
procedures. Quisylate 35 was prepared by the method of
(11) In ref 4 cholesterol is shown incorrectly in Table 1 with a -C6H12 side
chain but is shown correctly with a -C8H17 side chain in the structures in
Supporting Information.
(12) For selected chlorinations of 3β-cholesterol with retention of con-
figuration, see the following. (a) In two steps via the mesylate: Sun, Q.; Cai, S.;
Peterson, B. R. Org. Lett. 2009, 11, 567-570 and references therein. (b) With
aluminium and titanium halides: Broome, J.; Brown, B. R.; Summers, G. H. R.
J. Chem. Soc. 1957, 2071–2073. (c) Via a xanthate and sulfuryl chloride:
Kozikowki, A. P.; Lee, J. Tetrahedron Lett. 1988, 29, 3053–3056. (d) Directly
with ferric chloride: Liu, F.-W.; Liu, H.-M.; Zhang, Y.-B.; Zhang, J.-Y.; Tian,
L.-H. Steroids 2005, 70, 825–830.
ꢀ
ꢀ
(16) Juaristi, E.; Jimenez-Vazquez, H. A J. Org. Chem. 1991, 56, 1623–
1630.
(17) Scianowski, J.; Rafinski, Z.; Wojtczak, A. Eur. J. Org. Chem. 2006,
3216–3225.
(18) Naito, K.; Miura, A.; Azuma, M. J. Am. Chem. Soc. 1991, 113, 6386–
6395.
(13) Cram, D. J. J. Am. Chem. Soc. 1964, 86, 3767–3772.
(14) In ref 4 (1S,2R,5S)-menthol is shown in Table 1. The PEG-sulfonate
is shown as (1R,2S,5R), and the chloride product is shown as (1R,2S,5R) in
Supporting Information. For isomenthol the (1S,2R,5R) enantiomer is
shown in Table 1. The spectra shown on pages S4 and S10 of Supporting
Information1 are annotated as derivatives of (1S,2R,5S)-menthol.
(15) The related 2-pyridylsulfonate group had already been reported to
give inversion of configuration in nucleophilic bromination reactions using
magnesium dibromide as the halide source: Hanessian, S.; Kagotani, M.;
Komaglou, K. Heterocycles 1989, 28, 1115–1120.
(19) Schomaker, J. M.; Pulgam, V. R.; Borhan, B. J. Am. Chem. Soc.
2004, 126, 13600–13601.
(20) The quisylates of (()-2-octanol, (()-neomenthol, and (1R,2S,5R)-
menthol have been reported: Corey, E. J.; Posner, G. H.; Atkinson, R. F.;
Wingrad, A. K.; Halloran, D. J.; Radzik, D. M.; Nash, J. N. J. Org. Chem.
1989, 54, 389–393.
6044 J. Org. Chem. Vol. 74, No. 16, 2009

Braddock et al.
JOCArticle
Corey.20 This method was adapted for the preparation of
quisylate 29 and PEG sulfonate 34.
TABLE 1. Nucleophilic Displacement of Tosylates, PEG-Sulfonates,
and Quisylates 24-38 Using LiCl in Refluxing Acetone
entrya
substrateb
productc
45 (93)e
[R]D
d
1
2
3
4
5
6
7
8
24 (48; >90)
25 (3.5; >90)
26 (3.25; 80)
27 (36; >90)
28 (4; >90)
29 (3.5; 70)
30 (72; <5)
31 (72; 60)
32 (36; >90)
33 (24; 89)
34 (2.5; 92)
35 (3.5; 65)
36 (48; 30)
-28.3
-28.4
-28.1
+28.3
+27.6
+26.3
45 (92)e
45 (90)f
46 (100)
46 (85)e
46 (88)e
47 (71); 48 (14); 49 (15)
47 (59); 48 (26); 49 (15)
48 (2); 49 (98)
9
10
11
12
13
14
15
48 (3); 49 (97)
48 (3); 49 (97)
50 (87); 51 (13)
50 (85); 51 (15)
50 (71); 51 (29)
37 (48; 82)
38 (72; 64)
aAll reactions performed at 0.2 M in acetone with 4 equivalents of
LiCl; bReaction time (h) followed by isolated yield in parentheses. Where
yields are listed as >90 the reaction was essentially quantitative, but
small quantities of the volatile chloride product were lost in the
c
evaporative removal of acetone; Product distribution in parentheses;
d[R]D values are based on mass of mixture, and are not corrected for any
other components; eTraces of elimination products were observed by 1H
NMR. Some unreacted alcohol (ca. 9%) was carried through from
quisylate formation.
f
clearly superior leaving groups under these conditions. Evi-
dently, the PEG-ether and quinoline nitrogen lone pair can
both sequester lithium cation and activate the leaving group
as previously postulated. This is also apparent from inspec-
tion of the results from menthol- and neomenthol-derived
substrates 30-35 (entries 7-12). Tosylate 30 failed to react
under the reaction conditions even after extended heating
(entry 7). Presumably, the further steric encumbrance of the
β-isopropyl group in this secondary alcohol is sufficient to
prevent reaction under these conditions. However, the evi-
dently more reactive PEG-sulfonate 31 underwent slow
reaction to give neomenthyl chloride 47 as the major pro-
duct, along with elimination products menth-2-ene 48 and
menth-1-ene 49 (entry 8).26 Notably, menthyl chloride was
not observed. Quisylate 32 gave a similar product distribu-
tion (entry 9). In these cases the extra activation given by the
coordinated metal is apparently sufficient to overcome any
steric encumbrance. The stereochemical outcome from these
reactions under these conditions is also inversion of config-
uration at the reacting center.25 It is also apparent that
additional activation for substitution for the PEG-sulfonates
and quisylates goes hand-in-hand with additional propensity
to eliminate (i.e., they are simply better leaving groups for
either process). Thus, the neomenthyl tosylate 33, which is
perfectly set up for antiperiplanar elimination to give a
trisubstituted olefin, did not undergo substitution but gave
elimination product 49 and traces of regioisomer 48 (entry
10).26,27 In line with previous relative reactivities, the PEG-
sulfonate 34 and quisylate 35 reacted substantially more
rapidly than the tosylate 33 but did not undergo substitution
either, giving instead elimination products with essentially
identical product distributions (entries 11 and 12). Finally, in
cholestanol substrates 36-38, where there is no possibility of
Substrates 24-38 were subjected to nucleophilic displace-
ment using LiCl in acetone at reflux as per Lepore’s original
conditions (Table 1).4 Tosylates S-24 and R-27 (entries 1 and
4) gave 2-chlorooctane 45 and 2-chlorodecane 46 in high
isolated yield with inversion of configuration as judged by
comparison of the optical rotation21-23 with that of authen-
tic (2R)-chlorooctane and (2S)-chlorodecane prepared by
Appel chlorination24 of (S)-octan-2-ol and (R)-decan-2-ol,
respectively. PEG-sulfonates 25 and 28 (entries 2 and 5) and
quisylates 26 and 29 (entries 3 and 6) also gave the chlorides
with inversion of configuration, although some elimination
products were also seen. Thus for simple secondary alcohols,
under these conditions, tosylates, PEG-sulfonates, and qui-
sylates all lead to the same stereochemical outcome: inver-
sion of configuration.25 However, by inspection of the
reaction times, the PEG-sulfonate and quisylate groups are
(21) (2R)-Chlorooctane has a negative optical rotation; as prepared by
Appel chlorination: [R]D -32.7 (CH2Cl2, c 0.32). (2S)-Chlorodecane has a
positive optical rotation; as prepared by Appel chlorination: [R]D +33.0
(CH2Cl2, c 0.97).
(22) A control experiment shows that (2R)-chlorooctane is configura-
tionally stable to these reaction conditions (LiCl, refluxing acetone) at least
up to 17 h. Recovered 45: [R]D -31.9 (CH2Cl2, c 0.29).
(23) There is some small disparity in the recorded optical rotations
(Table 1, entries 1-6) with the expected rotations from 100% inversion of
configuration for these products (see ref 21). Since we have shown that the
products are configurationally stable under these conditions (see ref 22) and
that the diastereomerically pure substrates 30-32 and 36-38 (Table 1) show
only inverted chloride-containing products (as easily monitored by 1H NMR;
see Supporting Information), we surmise that these products are all enantio-
merically pure and there is up to 10% error in the optical rotation readings.
The source of error probably results from the volatility of the products and
small amounts of alkene impurities from competing elimination. However,
we were unable to separate the 2-chloroalkane enantiomers either by chiral
HPLC or chiral GC at Imperial or Oxford, respectively, to determine exact
ee’s.
(26) All products were identified by comparison with an authentic
sample.
(27) Halide anions in acetone have been previously noted to function as
bases: Biale, G.; Parker, A. J.; Smith, S; Stevens, I. D. R.; Winstein, S. J. Am.
Chem. Soc. 1970, 92, 115–122.
(24) Appel, R. Angew. Chem., Int. Ed. Engl. 1975, 14, 801–811.
(25) This is the expected stereochemical outcome of an SN2 mechanism
(see ref 1), but this term has a strict kinetic relationship associated with it that
has not been measured.
J. Org. Chem. Vol. 74, No. 16, 2009 6045

Braddock et al.
JOCArticle
neighboring group participation (cf., cholesterol 10), tosy-
late 36 reacted only sluggishly (entry 13). In contrast, the
more reactive PEG-sulfonate 37 and quisylate 38 gave the
inverted chloride 5025 as the major product, with some
elimination product 51.26 It is manifest from the above
results that under these conditions (LiCl, acetone) tosylate,
PEG-sulfonate, and quisylate derivatives of secondary alco-
hols all undergo nucleophilic substitution with inversion of
configuration at the reacting center.25,28 In sterically hin-
dered substrates, elimination is a competing process. More-
over, the PEG-sulfonates and quisylates are significantly
more reactive than the corresponding tosylates under these
conditions.
TABLE 2. Nucleophilic Displacement of Tosylates, PEG-Sulfonates,
and Quisylates 39-44 Using LiBr in Refluxing Acetone
entrya
substrateb
productc
recovered substrate
1d
2
39 (30; 59)
40 (8.5; 84)
41 (8.5; 81)
42 (30; <5)
43 (30; 77)
44 (30; 70)
52:53 (>20:1)
52 (100)
52 (100)
52:53 (ca. 1:6)
52:53 (ca. 1:7)
52:53 (ca. 1:8)
30
0
0
75
12
tr
3
4d
5d
6d
aAll reactions performed at 0.05 M in acetone with 4 equiv of LiBr.
c
bReaction time (h) followed by isolated yield in parentheses. Product
distribution in parentheses. dReaction stopped after 30 h.
of significant amounts of the trans-bromide 52 in the reaction
of the trans-sulfonates 42-44 with bromide anion. Moreover,
aswiththe results reported in Table1, the PEG-sulfonatesand
quisylates are significantly more reactive than the correspond-
ing tosylates to lithium bromide in acetone.
In connection with the Oxford’s group work on the
synthesis and structure determination of elatenyne,9 we were
interested in investigating the bromination of 3-hydroxy-
tetrahydrofuan derivatives with the aim of converting a
single 3-hydroxy-tetrahydrofuran into the corresponding
bromide with either inversion or retention of configuration.
As a model study, we elected to investigate the bromination
of the tosylates, PEG-sulfonates, and quisylates derived
from diastereomeric tetrahydrofurans 22 and 23, which
would allow us to determine the stereochemical course of
the bromination reactions in these systems. The six sulfonate
esters 39-44 were treated with lithium bromide in acetone at
reflux using the conditions described by Lepore, and the
1
crude reaction mixtures were analyzed by H NMR after
filtration through a silica plug (Table 2). The cis-sulfonate
esters all gave the corresponding trans-bromide 52, which
was identical to material prepared by Appel bromination of
1
the cis-alcohol 22. Analysis of the H NMR spectra from
these three reactions indicated that the bromide derived from
the cis-tosylate 39 contained a small amount of the cis-
bromide 53. Control experiments demonstrated that both
the cis- and trans-bromides 53 and 52 were converted into
one another under the reaction conditions, with the cis-
bromide 53 reacting faster than the corresponding trans-
bromide 52. On extended reaction times (48 h) traces of the
cis-bromide 53 were formed from both the reaction of the cis-
PEG-sulfonate 40 and the cis-quisylate 41. The trans-sulfo-
nate esters (42-44) derived from the alcohol 23 reacted
significantly more slowly than the corresponding cis-sulfo-
nate esters, and none of the reactions had reached comple-
tion after 30 h. The PEG-sulfonate 43 and the quisylate 44
both gave the corresponding cis-bromide 53 as the major
product, which was identical to material prepared by Appel
bromination of the trans-alcohol 23. Significant amounts of
the trans-bromide 52 were also formed under the reaction
conditions. The trans-tosylate 42 gave only a very small
amount of the corresponding bromides as an approximate
6:1 mixture of 53:52. The above results are entirely in accord
with the mechanism of these reactions being nucleophilic
substitution with inversion of configuration at the reacting
center.25 In keeping with this stereochemical scenario, the
trans-sulfonates (42-44) and the trans-bromide (52) react
more slowly as a result of the bromide anion having a more
hindered trajectory in the trans-substrates then in the corre-
sponding cis-substrates. This readily accounts for the formation
Having established that tosylates, PEG-sulfonates, and
quisylates all undergo substitution with inversion of config-
uration at the reacting center with lithium halides in reflux-
ing acetone, substrates 24-38 were also subjected to
nucleophilic displacement using TiCl4 in dichloromethane
at -78 °C as per Lepore’s conditions (Table 3).5 In our hands
these reactions were not complete after 2 min as originally
reported and were allowed to react for 1 h at this temperature
before quenching.29 The isolated products were analyzed by
1H NMR and by polarimetry (Table 3). Inspection of the
results for the sulfonate esters derived from simple secondary
alcohols (Table 3, entries 1-6) shows that multiple products
are formed. These include the expected 2-chloroadducts 45
(entries 1-3) and 46 (entries 4-6) as well as 3- and 4-
chloroalkanes 54 and 55 (entries 1-3) and 57 and 58
(28) Control experiments with 3-β- and 3-R-chlorocholestanes and
menthyl chloride involving extended reflux (48 h) with LiCl (4 equiv) in
acetone showed them to be configurationally stable, and they were recovered
unchanged in quantitative yield.
(29) We found the tosylates to be comparably, if not more reactive than
the corresponding quisylates and PEG-sulfonates to reaction with TiCl4 by
inspection of the evolution of the TLC profile for these reactions, although
this was not quantified.
6046 J. Org. Chem. Vol. 74, No. 16, 2009

Braddock et al.
JOCArticle
(entries 4-6).26 Small quantities of the internal C8 (56)
(entries 1 and 2) and C10 (59) (entries 4 and 5) alkenes were
also observed.26 Thus, under these conditions, a different
product profile is observed compared with that when em-
ploying lithium halide in refluxing acetone (vide supra,
Table 1, entries 1-6). The products can all be postulated to
be derived from a common secondary carbocation (or
corresponding ion pair)30 at the 2-position. Direct trapping
of the carbocation yields 2-chloroalkanes 45 or 46. Alterna-
tively, 1,2-hydride shifts and subsequent trapping allow the
formation of the 3- and 4-chloroalkanes 54 and 55, or 57 and
58.26,31 Elimination under Saytzev orientation explains the
presence of alkenes 56 or 59. Analysis of the optical rotations
of these mixtures reveals (Table 3) that substitution has
proceeded with partial inversion of configuration.32 Thus,
the simplest model that can be formulated to account for all
of these observations involves a TiCl4-induced formation of
a carbocation (or ion pair)30 where the leaving group par-
tially shields front side attack from a returning chloride
nucleophile.33,34 It also seems reasonable to suggest that
other secondary sulfonate substrates (vide infra) will react
via TiCl4-induced carbocations (or ion pair) under these
conditions.30,34 Interestingly, the extent of racemization as
judged from the optical rotations of the mixtures is signifi-
cantly lower in the PEG-sulfonate and quisylate cases,
indicating more tightly bound ion pairs.35
Thus, for simple secondary alcohols modified as tosylates,
PEG-sulfonates, or quisylates, whereas reaction with LiCl in
refluxing acetone leads to inversion of configuration25 at the
reacting center, the use of TiCl4 leads to chlorides with
partial inversion of configuration,34 and retention of config-
uration is not observed. In contrast, the menthol-derived
sulfonates 30-32 (entries 7-9), including the simple tosy-
late, all gave menthyl chloride 6026 with overall retention of
configuration in good yield. A survey of the literature reveals
that this is not a new phenomenon, and solvolysis of menthol
tosylate with predominant retention of configuration has
previously been reported.36,37 In this case, on the basis of a
significantly reduced β,β0-d3 isotope effect, the authors pro-
posed that a chairlike conformation persists in a rate-deter-
mining transition state and an intermediate ion pair, where
the substituents at the 2- and 5-positions (relative to the
TABLE 3. Nucleophilic Displacement of Tosylates, PEG-Sulfonates,
and Quisylates 24-38 Using TiCl4 in Dichloromethane
entrya
substrateb
productc
[R]D
d
1
2
3
4
5
6
7
8
24 (75)
45 (40); 54 + 55 (56); 56 (4)
45 (62); 54 + 55 (35); 56 (3)
45 (43); 54 + 55 (57); 56 (0)
46 (29); 57 + 58 (68); 59 (3)
46 (57); 57 + 58; (42); 59 (1)
46 (42); 57 + 58; (58); 59 (0)
60 (ca. 90)
60 (ca. 90)
60 (ca. 90)
61 (47); 62 (53)
61 (46); 62 (54)
61 (46); 62 (54)
-4.7
-13.6
-7.5
+3.6
+11.7
+9.5
25 (>90)
26 (>90)
27 (88)
28 (>90)
29 (>90)
30 (>90)
31 (>90)
32 (>90)
33 (79)
34 (>90)
35 (>90)
36 (65)
37 (82)
38 (70)
9
10
11
12
13
14
15
0.0
0.0
0.0
50 (42); 51 (40); 63 (18)
50 (57); 51(1); 63 (42)
50 (66); 51 (8); 63 (26)
aAll reactions performed at 0.05 M in dichloromethane at -78 °C for
1 h with 2 equiv of TiCl4. bIsolated yield in parentheses. cProduct
d
distribution in parentheses. [R]D values are based on mass of mixture
and are not corrected for any other components.
reacting center) prevent the formation of a bent chair or half
chair conformation where the “bottom-face” would become
exposed. Nucleophilic capture then occurs on the exposed
“top-face”, leading to the observed retention of configura-
tion. We suggest that the same effect is in operation with
menthol sulfonates 30-32, and under these conditions a
TiCl4-induced carbocation (or ion pair)30,34 is formed where
it is the conformation of the menthyl framework and not the
leaving group that determines the resultant stereochemistry
of the product. Notably Kim’s substrate 3 also has this
cyclohexyl substitution pattern, where the reaction proceeds
with retention of configuration (vide supra). In further
contrast, whereas neomenthol tosylate 33, PEG-sulfonate
34, and quisylate 35 all underwent elimination reactions with
LiCl in refluxing acetone, on treatment with TiCl4 tertiary
chlorides 61 and 62 were isolated (entries 10-12).26 Evi-
dently, the incipient p-orbital forming from loss of any of the
axially disposed sulfonates by TiCl4-induced carbocation (or
ion pair)30 formation is perfectly aligned to induce an
immediate 1,2-hydride shift from the neighboring axial
hydrogen38 at the 2-position leading to a tertiary carbocation
and thence 61 after trapping by chloride. A second shift to
give another tertiary carbocation can lead to chloride 62.39
Finally, the sulfonate derivatives of cholestanol 36-38
(entries 13-15) underwent reaction under these conditions
to give epimeric chlorides 50 and 63,26 as well as elimination
product 51 in some cases. Whereas with LiCl in refluxing
acetone only chloride 50 with inversion of configuration was
observed, the observation of both epimeric chlorides 50
and 63 under these conditions is consistent with TiCl4-
induced carbocation (or ion pair)30 formation where the
leaving group partially shields the top face. Unlike the menthol
sulfonates, these sulfonates lack a 2-substituent on the cyclo-
hexyl ring, which must allow sufficient conformational flex-
ibility to accommodate bottom face attack. It is therefore
apparent that under these conditions (TiCl4, CH2Cl2,
(30) It is not our intention to become embroiled in the long-standing
debate regarding precise mechanistic distinctions or the SN1/SN2 mechanistic
borderline but simply to clarify the stereochemical outcomes using these
leaving groups under given conditions such that other workers may make
accurate predictions of the expected stereochemical outcomes.
(31) An authentic sample of 5-chlorodecane was prepared in anticipation
of its observation. However, 5-chlorodecane was not observed in the product
mixtures from reaction of 27, 28, and 29.
(32) Assuming that all products arising from hydride shifts are racemic,
the optical purities of the 2-chloroalkanes in Table 3, entries 1-6 are 36%
(R), 67% (R), 53% (R), 37% (S), 61% (S), and 68% (S), respectively.
(33) One referee suggested that a mixed SN1-SN2 mechanism could be
operating.
(34) This is the expected stereochemical outcome of an SN1 mechanism
(see ref 1), but this term has a strict kinetic relationship associated with it that
has not been measured.
(35) A control experiment with (2R)-chlorooctane shows that it is con-
figurationally stable under these conditions (TiCl4, CH2Cl2, -78 °C, 1 h).
Recovered 45: [R]D -32.5 (CH2Cl2, c 0.32).
(36) Hiral-Starcevic, S.; Majerski, Z.; Sunko, D. E. J. Am. Chem. Soc.
1974, 96, 3659–3661.
(38) Hiral-Starcevic, S.; Majerski, Z.; Sunko, D. E. J. Org. Chem. 1980,
45, 3388–3393.
(39) An authentic sample of these compounds was prepared as a mixture
by treating menth-1-ene 49 with concentrated aqueous HCl. The identity of
these compounds was also confirmed by readings of zero for their optical
rotations reflecting the internal mirror plane that now exists in these
compounds.
(37) The observation of menthyl chloride as a product (i.e., with retention
of configuration) in the chlorination of menthol with PCl5 and various
additives has been long known: Smith, J. G.; Wright, G. F. J. Org. Chem.
1952, 17, 1116–1121 and references therein.
J. Org. Chem. Vol. 74, No. 16, 2009 6047

Braddock et al.
JOCArticle
-78 °C), tosylate, PEG-sulfonate, and quisylate substrates all
undergo nucleophilic substitutions consistent with TiCl4-
induced carbocation (or ion pair)30 formation, leading to
partial inversion of configuration34 in simple secondary
systems. In the case of menthol-derived sulfonates, the steric
effect of the bulky isopropyl group dominates, and substrate-
controlled retention of configuration is observed.40
General Procedure for Appel Chlorination of Alcohols. To a
solution of alcohol (1 equiv) in carbon tetrachloride (0.4 M) was
added triphenylphosphine (2 equiv), and the reaction mixture
was heated to reflux for 24 h. The reaction mixture was allowed
to cool to rt, diluted with pentane, and filtered through a plug of
silica, and the solvent was evaporated to a minimum by atmo-
spheric distillation and finally removed by Kugelrohr distilla-
tion.
(R)-2-Chlorooctane 45. Prepared by Appel chlorination from
(S)-octan-2-ol: [R]D -32.7 (CH2Cl2, c 0.32); IR (thin film) υmax
2928, 2959 cm-1; 1H (400 MHz, CDCl3) δ 4.10-4.01 (m, 1H),
1.77-1.69 (m, 2H), 1.53 (d, J=6.7 Hz, 3H), 1.52-1.26 (m, 8H),
0.91 (t, J=6.9 Hz, 3H); 13C (100 MHz, CDCl3) δ 59.0, 40.4, 31.7,
28.8, 26.6, 25.4, 22.6, 14.1; GCMS 112 (13%, M - HCl), 83 (62),
70 (100), 55 (72).
General Procedure for Nucleophilic Substitution of Sulfonates
39-44 with LiBr. To a solution of sulfonate (1 equiv) in acetone
(0.05 M) was added lithium bromide (4 equiv), and the reaction
mixture was heated to reflux until no further reaction was
evident by TLC (or for 30 h, whichever was the shorter time).
The mixture was allowed to cool, diluted with light petroleum,
and filtered through a plug of silica washing with ether. The
solvent was removed in vauco, and the residue was analyzed by
1H NMR. The residue was then purified by flash chromatogra-
phy (4:1, light petroleum/ethyl acetate) (Table 2).
Conclusion
We have clarified the stereochemical course of nucleophi-
lic substitution of arylsulfonate-based leaving groups, show-
ing that tosylates, PEG-sulfonates, and quisylates of
secondary alcohols are all subject to inversion of configura-
tion at the reacting center25 when treated with lithium halides
in refluxing acetone. The PEG-sulfonates and quisylates are
considerably more reactive than the corresponding tosylates
in their reactions with metal halide salts in acetone and are a
significant addition to the armory of reagents available for
the activation of hydroxy groups. The increased activity is
consistent with the positively charged metal being chelated
by the PEG group or coordinated by the quinoline lone pair
and stabilizing the negative charge on the leaving sulfonate
as originally proposed by Lepore. In contrast, TiCl4-induced
substitutions of tosylates, PEG-sulfonates, and quisylates
leads to product distributions consistent with carbocation
(or ion pair)30 formation and in simple secondary substrates
leads to partial inversion of configuration.29,34 Any observed
retention of configuration5-7 is likely due to neighboring
group participation41 or diastereoselective attack on a car-
bocation (or ion pair) rather than an SNi mechanism.
Bromide 52. To a stirred solution of (2R*,3R*)-2-(benzyl-
oxymethyl)tetrahydrofuran-3-ol (50 mg, 0.24 mmol) in toluene
(3 mL) were added triphenylphosphine (126 mg, 0.48 mmol) and
carbon tetrabromide (159 mg, 0.48 mmol). The reaction mixture
was heated to 80 °C for 1 h, cooled to rt, diluted with CH2Cl2
(10 mL), adsorbed onto silica, and purified by flash column
chromatography (4:1, light petroleum/ethyl acetate) to give the
title compound as a colorless oil (48 mg, 0.17 mmol, 74%); Rf
0.64 (1:1 petroleum ether 40-60°/ethyl acetate); IR (CDCl3)
νmax 3030, 2983, 2944, 2858, 1497, 1451 cm-1 1H NMR
;
(400 MHz, CDCl3) δ 7.38-7.28 (m, 5H), 4.61 (d, J=12.1 Hz,
1H) 4.56 (d, J=12.1 Hz, 1H), 4.31-4.24 (m, 2H), 4.08-3.98 (m,
2H), 3.61-3.55 (m, 2H), 2.56-2.48 (m, 1H), 2.25 (tdd, J=4.3,
6.5, 13.6 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ 137.9, 128.4,
127.7, 127.6, 86.3, 73.5, 69.8, 67.3, 46.6, 36.8; MS (ES+) m/z
288.0/290.0 (M + NH4)+ (100%), 293.0/295.0 (M + Na)+,
563.0/565.0/567.0 (2M + Na)+; HRMS (ES+) m/z calcd for
C12H1579BrNaO2 (M+Na)+ 293.0148, C12H1581BrNaO2 (M+
Na)+ 295.0128, found 293.0145, 295.0128.
General Procedure for Nucleophilic Substitution of Sulfonates
24-38 with TiCl4. To a solution of sulfonate (1 equiv) in
dichloromethane (0.05 M) at -78 °C was added TiCl4 (2 equiv)
dropwise. The reaction mixture immediately became yellow and
was stirred at -78 °C for 1 h. The mixture was then quenched
with saturated sodium hydrogen carbonate solution and ex-
tracted with dichloromethane, and the combined organic phase
was dried over magnesium sulfate and then passed through a
plug of silica. The solvent was evaporated to a minimum by
atmospheric distillation and finally removed by Kugelrohr
distillation. The resulting product mixtures were analyzed by
1H NMR spectroscopy and polarimetry (Table 3). Products 45,
46, 50, 51, and 54-63 were identified by the following char-
acteristic 1H and 13C NMR shifts, and product ratios were
obtained by integration of the same resonances in the 1H NMR
spectra (see Supporting Information). δH (45) 4.05 (m, 1H); (46)
4.06 (m, 1H); (50) 4.5544 (br s, 1H); (51) 5.6245 (br s, 2H); (54)
3.87 (m, 1H); (55) 3.93 (m, 1H); (56) 5.43 (m, 2H); (57) 3.88 (m,
1H); (58) 3.93; (59) 5.4346 (m, 2H); (60) 3.804 (m, 1H); (63) 3.90
(m, 1H) ppm. δc (45) 59.0; (46) 59.0; (54) 65.9; (55) 64.0; (57)
66.0; (58) 64.1; (60) 63.9 ppm. Authentic samples of simple
Experimental Section
General. See Supporting Information.
General Procedure for Nucleophilic Substitution of Sulfonates
24-38 with LiCl. To a solution of sulfonate (1 equiv) in acetone
(0.2 M) was added lithium chloride (4 equiv), and the reaction
mixture was heated to reflux until no further reaction was
evident by TLC. The mixture was allowed to cool, diluted with
pentane, and filtered through a plug of silica. The solvent was
evaporated to a minimum by atmospheric distillation and
finally removed by Kugelrohr distillation. The resulting product
1
mixtures were analyzed by H NMR spectroscopy and polari-
metry (Table 1). Products 45-51 were identified by the follow-
1
ing characteristic H and 13C NMR shifts, and product ratios
were obtained by integration of the same resonances in the 1H
NMR spectra (see Supporting Information). δH (45) 4.05 (m,
1H); (46) 4.06 (m, 1H); (47) 4.5442 (br s, 1H); (48) 5.5543 (br s,
2H); (49) 5.3843 (br s, 1H); (50) 4.5544 (br s, 1H); (51) 5.6245 (br s,
2H) ppm. δc (45) 59.0; (46) 59.0; (47) 63.5 ppm. Authentic
samples of simple aliphatic chlorides 45 and 46 were prepared
by Appel chlorination of the corresponding secondary alcohol.
(40) Control experiments with 3-β- and 3-R-chlorocholestanes and
menthyl chloride show them to be configurationally stable under these
conditions (TiCl4, CH2Cl2, -78 °C, 1 h), and they were recovered unchanged
in quantitative yield.
(41) For bromine as a NGP on a leaving PEG-sulfonate, see: Braddock,
D. C.; Hermitage, S. A.; Kwok, L.; Pouwer, R.; Redmond, J. M.; White, A. J.
P. Chem. Commun. 2009, 1082–1084.
(42) Drabowicz, J.; Luczak, J.; Mikolajczyk, M. J. Org. Chem. 1998, 63,
9565–9568.
(43) Elson, K. E.; Jenkins, I. D.; Loughlin, W. A. Org. Biomol. Chem.
2003, 1, 2958–2965.
(44) Ho, P. T.; Davies, N. J. Org. Chem. 1984, 49, 3027–3029.
(45) Tal, D. M.; Keinan, E.; Mazur, Y. Tetrahedron 1981, 37, 4327–4330.
(46) Pelter, A.; Smith, K.; Elgendy, S. M. A. Tetrahedron 1993, 49, 7119–
7132.
6048 J. Org. Chem. Vol. 74, No. 16, 2009

Braddock et al.
JOCArticle
aliphatic chlorides 54, 55, 57, and 58 were prepared by Appel
chlorination of the corresponding racemic alcohol (see Support-
ing Information). An authentic sample of 61 and 62 as a mixture
was prepared by the action of concentrated aqueous HCl on
menthene 49. An authentic sample of chloride 63 was prepared
by hydrogenation of the corresponding cholesterol chloride.
Chlorides 61 and 62. To 4-methyl 1-isopropylcyclohexene
(8 mg, 0.06 mmol) was added 6 drops of conc aqueous HCl,
and the mixture was allowed to stir for 24 h. The mixture was
diluted with CDCl3 (1.5 mL), and the resulting solution was
washed with saturated NaHCO3 solution and dried over mag-
of silica gel, and the solvent was removed in vacuo to give
the title compound as a white powder (142 mg, 94% yield).
Mp 110-114 °C; [R]D +5.0 (CH2Cl2, c 0.3); IR (thin film) υmax
2933, 2865, 2850, 1467, 1373 cm-1; 1H NMR (400 MHz, CDCl3)
δ 3.94-3.84 (m, 1H), 2.09-1.95 (m, 2H), 1.88-0.84 (m, 40H),
0.70-0.61 (m, 1H), 0.67 (s, 3H); 13C (100 MHz, CDCl3) δ 60.3,
56.4, 56.3, 54.2, 46.8, 42.6, 40.0, 39.6, 39.5, 38.7, 36.2, 35.8, 35.4,
35.3, 33.2, 32.0, 28.5, 28.2, 28.0, 24.2, 23.8, 22.8, 22.6, 21.1, 18.7,
12.2, 12.1; MS (EI+) m/z 408 (M)+, 406 (Mþ); HRMS (EI+) m/
z calcd for C27H47Cl (M)+ 406.3366, found 406.3371.
1
nesium sulfate. H NMR showed complete conversion of the
Acknowledgment. We thank the EPSRC (Grant no. EP/
E058272/1 to D.C.B.) and the Royal Society (URF to J.W.
B.) for financial support and Jennifer Lloyd for preliminary
experiments.
starting material to give 61 and 62 as an approximately 2:1
mixture. 1H NMR (400 MHz, CDCl3) δ 2.08-2.02 (m, 2H, 61),
2.00-1.92 (m, 2H, 62), 1.83-0.87 (m, 22H), 1.05 (d, J=6.7 Hz,
6H, Cl-1), 0.95 (d, J=6.6 Hz, 3H), 0.90 (d, J=6.6 Hz, 3H); 13
C
(100 MHz, CDCl3) δ 80.1, 75.4, 50.2, 40.6, 37.2, 35.1, 32.5, 31.9,
30.7, 30.4, 29.7, 28.0, 22.5, 22.3, 17.6; GCMS (61) 138 (21%,
M - HCl), 130 (19), 95 (100), 81 (19), 67 (17), 55 (17); GCMS
(62) 138 (56%, M - HCl), 123 (42), 97 (75), 95 (84), 81 (100), 67
(39), 55 (76).
Supporting Information Available: General experimental
procedures, synthetic procedures, and characterizing data for
sulfonates 24-44 and bromide 53; characterizing data for 46, 54,
1
55, 57, 58, and 5-chlorodecane; copies of H spectra and 13C
NMR spectra for 24-46, 52-55, 57, 58, 5-chlorodecane, and 63;
and copies of 1H NMR spectra of the product mixtures for all
substitution reactions recorded in Tables 1-3. This material is
available free of charge via the Internet at http://pubs.acs.org.
Chloride 63. To a stirred solution of 3β-cholesteryl chloride
(150 mg) in diethyl ether (5 mL) was added Pd/C (15 mg). The
reaction mixture was purged with hydrogen gas and allowed to
stir for 4 h. The reaction mixture was then filtered through a plug
J. Org. Chem. Vol. 74, No. 16, 2009 6049