The synthesis and use in asymmetric epoxidation of metal salen complexes derived from enantiopure trans-cyclopentane- and cyclobutane-1,2-diamine

Article abstract of DOI:10.1016/S0957-4166(02)00757-7

A complete synthesis of enantiopure trans-cyclopentane-1,2-diamine and trans-cyclobutane-1,2-diamine is described. These diamines have been used as components of novel chiral salen ligands whose chromium and manganese complexes were then evaluated as oxygen transfer agents in the asymmetric epoxidation of alkenes.

Full text of DOI:10.1016/S0957-4166(02)00757-7

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 127–137
The synthesis and use in asymmetric epoxidation of metal salen
complexes derived from enantiopure trans-cyclopentane- and
cyclobutane-1,2-diamine
Adrian M. Daly and Declan G. Gilheany*
Chemistry Department, Centre for Synthesis and Chemical Biology, Conway Institute of Biomolecular and Biomedical Sciences,
University College Dublin, Belfield, Dublin 4, Ireland
Received 16 October 2002; accepted 8 November 2002
Abstract—A complete synthesis of enantiopure trans-cyclopentane-1,2-diamine and trans-cyclobutane-1,2-diamine is described.
These diamines have been used as components of novel chiral salen ligands whose chromium and manganese complexes were then
evaluated as oxygen transfer agents in the asymmetric epoxidation of alkenes. © 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
92% using complex 3 (L=triphenylphosphine oxide
(Ph3PO)) in stoichiometric mode.9
There continues to be great interest in the development
of general reagents for catalytic asymmetric alkene
epoxidation,1–3 the most notable successes being the
systems of Sharpless and Jacobsen/Katsuki1,3,4 The chi-
ral manganese salen complex 1 developed by Jacobsen
and co-workers was the first catalyst to enable the
asymmetric epoxidation of unfunctionalised alkenes in
high ee. This and other similar manganese complexes
catalyse the epoxidation of Z- and tri-substituted alke-
nes, often with excellent selectivity (e.g. 92% ee in the
epoxidation of (Z)-b-methylstyrene). However, the
major limitation of 1 is the low selectivity displayed in
the epoxidation of E-1,2-disubstituted alkenes (e.g. 22%
ee in the epoxidation of (E)-b-methylstyrene).
We2,6,7,12 and others16–19 have suggested that the
stereoselection obtained using manganese and
chromium salen complexes is due to a non-planar com-
plex. This non-planarity can be observed in a number
of X-ray crystal structures16 and arises as a result of the
sp3 centres on the diimine bridge of the ligand. Thus,
the NꢀCꢀCꢀN dihedral angle (Fig. 1) between the two
nitrogen atoms of the ligand is a measure of the resul-
tant twist or step in the complex. For complexes
derived from trans-cyclohexane-1,2-diamine the rigid
nature of the cyclohexane ring, with a set NꢀCꢀCꢀN
dihedral angle, is a major influence on the complex
conformation. Since our group has previously sug-
gested that chromium salen complexes facilitate selec-
tive epoxidation of E-alkenes by their stepped nature6
we felt that investigation of smaller ring diamines might
lead to more enantioselective epoxidation catalysts by
increasing the NꢀCꢀCꢀN dihedral angle and thus alter-
ing the ligand conformation in a favourable manner.
For some time we have been working on the develop-
ment of epoxidation catalysts based on a chromium
salen template (2,3, Table 1).5–12 The stability of the
oxochromium(V) species13 3 allows the reaction to be
run in both stoichiometric and catalytic mode. The
most notable aspect of our system is that E-alkenes
give much better selectivity than Z isomers. This is in
contrast to the manganese salen based system and
indeed nearly all other metal oxo based oxidants.14,15
Very recently we have improved the ee in our standard
test reaction, the epoxidation of (E)-b-methylstyrene, to
Figure 1. Illustration of the NꢀCꢀCꢀN dihedral angle in
metal salen complexes.
* Corresponding author. E-mail: declan.gilheany@ucd.ie
0957-4166/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(02)00757-7

128
A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
Despite the widespread use of metal salen complexes in
asymmetric catalysis there has been no previous report
of ligands containing such diamines (Scheme 1).
from cyclopentene via the analogous dibromide. How-
ever, despite evaluating a variety of reaction conditions
to convert the dibromide to diazide 6 we found that
competing elimination led to a mixture of two products.
Therefore, the procedure of Minisci and co-workers for
the synthesis of 1,2-diazidocyclohexane 6, directly from
cyclopentene was followed.24 This produced a 15:1 mix-
ture of the trans- and cis-isomers of 6 in poor yield
(13%). Separation of the small amount of cis isomer
was possible by column chromatography; however, this
was not carried out in most runs of the experiment
since separation of the isomers was possible in the later
resolution step.
2. Results and discussion
We initially investigated synthetic routes to the chiral
diamines trans-cyclopentane-1,2-diamine 4 and trans-
cyclobutane-1,2-diamine 5. The easier to make
appeared to be the cyclopentyl species 4, although it is
not as conformationally rigid as the cyclobutyl species
5. Although we were particularly interested in the use
of these diamines as components of chiral salen com-
plexes they also have numerous other potential applica-
tions in asymmetric synthesis.20,21
Reduction of the diazide by hydrogenation over PtO2
proved unsuccessful, contrary to the report of Potter et
al.23 Instead we found that Pd/C was effective and
using hydrogen at 3 atm pressure gave trans-cyclopen-
tane-1,2-diamine 4 in 73% yield (Scheme 2). This
racemic material was resolved using tartaric acid
according to the method of Toftlund and Pedersen.25
The assignment of an (R,R) configuration to the
resolved diamine is on the basis of a crystal structure of
the tartrate salt obtained by Toftlund and Pedersen.25
A procedure from the same authors was used to release
the enantiopure diamine,25 which, due to its air sensitiv-
ity, was used immediately for the generation of salen
ligands. While this work was in progress an alternative
method for the preparation of 4 was reported; however,
this starts from the expensive trans-cyclopentane-1,2-
dicarboxylic acid.26
2.1. Synthesis of (R,R)-trans-cyclopentane-1,2-diamine,
4
Two syntheses of trans-cyclopentane-1,2-diamine 4 had
previously been reported. The route of Jaeger and
Blumendal involved the reduction of cyclopentane-1,2-
dione dioxime using sodium.22 Attempts to repeat this
procedure were unsuccessful and the yields obtained
were considerably lower than those reported in the
literature.
An alternative route reported by Potter et al.23 involved
the hydrogenation of diazide 6 over PtO2 to yield the
diamine 4. Our initial attempts to synthesise 6 were
Scheme 1.
Scheme 2.

A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
129
2.2. Synthesis of (+)-trans-cyclobutane-1,2-diamine, 5
found it more satisfactory to remove the excess
methanol by distillation at atmospheric pressure and
then flush the residue through a pad of silica gel using
CH2Cl2. In this manner a mixture of the two isomers of
nitrile diester 10 was obtained in 71% yield. It was
possible to separate these products since one crystallises
from the mother liquor, leaving the other behind as a
liquid. This separation was carried out initially; how-
ever, it only leads to loss of material and is unnecessary
since the distinction between isomers is immediately
lost in the subsequent hydrolysis to triacid 11, heating
of which (180°C at 20 mmHg) yields cyclobutane-1,2-
diacid as a 70:30 mixture of trans- and cis-isomers 12.
The combined hydrolysis and decarboxylation were
carried out in 65% yield.
Again, only two routes to trans-cyclobutane-1,2-
diamine 5 have been published. The most recent of
these was developed from a by-product of a different
reaction, which was being investigated by the same
workers.27 The starting iminonitrile species requires a
three step synthesis in its own right28 so this route was
not examined further.
Instead the synthesis carried out by Buchman et al. was
used as a starting point.29 The general scheme of this
synthesis involves the conversion of adipic acid 7 into
trans-cyclobutane-1,2-dicarboxylic acid 8. A Schmidt
rearrangement, using hydrazoic acid, is then used to
convert this to the analogous diamine 5. At approxi-
mately the same time, a similar synthesis was reported
by Shuikina, which involved a Hoffman rearrangement
for the conversion of the diacid to the diamine.30 Since
the publication of these works a small number of
reports have been made of their use, either as a whole,
or in part.31–33 Watson and O’Neill reported improved
yields for the initial step34 and Cory and co-workers
found that the Schmidt rearrangement was unsuccessful
and used the Curtius rearrangement instead.31 How-
ever, none of these authors described their synthesis in
detail, and we therefore feel it is appropriate to describe
the route we followed. This is shown in Scheme 3 and
was developed on the basis of all the work mentioned
above. The formation of the dibromo-dimethyl ester of
adipic acid 9 was carried out as a one pot procedure
giving 57% yield.34 The ring-closure step to yield 10 was
carried out using potassium cyanide with suitable pre-
cautions to ensure a basic environment. Buchman’s
procedure29 for this step calls for distillation of the
reaction mixture to obtain the product; however, we
This mixture of isomers of 12 was equilibrated by
heating in 12 M HCl at 120°C for 6 days.35 Subsequent
cooling led to crystallisation of pure trans-isomer 8 in
54% yield. After further heating of the mother liquor
under similar conditions, a further sample of 8 was
isolated. As mentioned above, Buchman and co-work-
ers converted this diacid 8 to the diamine 5 by the
Schmidt rearrangement in 55% yield.29 However, given
the difficulties in handling hydrazoic acid and the
reported failure of this method31 we instead used a
Curtius rearrangement. This necessitated initial forma-
tion of the foul smelling diacid chloride 13. Using
thionyl chloride this was accomplished in 96% yield.
Initially we attempted to isolate the diacyl azide deriva-
tive of 13 prior to the Curtius rearrangement, as previ-
ously reported for the cyclopropyl analogue,36 however
a small sample of the diazide exploded violently very
shortly after being removed from the rotary evaporator,
fortunately leading to no serious injuries. We therefore
abandoned this procedure and instead used a one-pot
Scheme 3.

130
A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
procedure37 in which neither the diacylazide nor its
diisocyanate rearrangement product were isolated. The
diamine dihydrochloride 5·2HCl was isolated in 46%
yield, in agreement with the previous report for the
cyclopropyl analogue by Sturm et al.31
enantiopurity of the resolved diamine was confirmed by
chiral HPLC analysis of the derived salen ligand 14a.
The mother liquor from the initial resolution step was
reacted to regenerate the remaining diamine, and this
was then reacted with unnatural (−)-tartaric acid. The
salt obtained from this procedure contained, after one
recrystallisation, the (−)-diamine in enantiopure form.
A
resolution of trans-cyclobutane-1,2-diamine has
never previously been reported. However, diacid 8 has
been resolved in reasonable yield using cinchonidine.35
The recent development of the ‘Dutch resolution’ tech-
nique,38 based on the use of mixtures of three resolving
agents, had given us confidence when originally
embarking on this synthesis, that resolution of the
diamine would not prove problematic. The authors
reported that addition of a suitable mixture or ‘family’
of resolving agents to a racemic substrate in the appro-
priate solvent led to the almost immediate formation of
a salt. This salt is composed of a mixture of the
substrate and each of the three resolving agents. In
many cases38 the salt contains just one substrate enan-
tiomer in 30–50% yield after one recrystallisation.
Given that tartaric acid has proved useful for resolving
both the cyclohexyl and cyclopentyl analogues of 5 we
chose the family of tartaric acid derivatives ‘T-mix’ for
our initial attempts.
2.3. Synthesis of chromium and manganese salen
complexes
The four salen ligands 14a,b and 15a,b were synthesised
by simple condensation of the relevant diamine and
salicylaldehyde. Formation of the (salen)Cr(III) and
(salen)Mn(III) complexes of these ligands was then
attempted by reaction of the appropriate salen ligand
with anhydrous chromium(II) chloride39 and mangane-
se(II) acetate respectively (Scheme 4).40 Using the
cyclopentyl ligands 14b, 15b both these reactions pro-
ceeded without difficulty yielding the desired complexes
16b, 17b, which were characterised by electrospray mass
spectroscopy and IR spectroscopy.
We then attempted to synthesise the cyclobutane-con-
taining (salen)Cr(III) complex 16a using chromium(II)
chloride (Scheme 4). For all the (salen)Cr(III) com-
plexes which have previously been synthesised by us,
addition of chromium(II) chloride to a solution of the
yellow solution of the ligand in THF leads to an
immediate colour change to dark brown. Unusually we
observed a green coloured solution instead. Neverthe-
less, the reaction was continued as before and a green
solid was isolated. This was reacted with silver nitrate
yielding a precipitate of silver chloride and leading to
the isolation of a further green solid. The IR spectrum
of this solid contained the characteristic imine stretch at
1624 cm−1. However, the electrospray mass spectrum
indicated that only a small proportion of the substance
had the correct m/z value of 480 for 16a. The other
components of the mixture are not readily identifiable
from their mass spectrum or IR spectrum.
We did observe the formation of a salt, but to our
disappointment and surprise, liberation of diamine 5
from this salt yielded only racemic material. This failure
of the ‘Dutch resolution’ method is notable. It may be
that we would have been successful using other
‘families’; however, these are not readily available. On
the other hand very few difunctionalised species and no
diamines were reported among the many successfully
resolved substrates in the original communication of
Vries et al.38
We thus resorted to tartaric acid as a resolving agent
and we followed Toftlund and Pedersen’s procedure for
trans-cyclopentane-1,2-diamine 4 to give enantiopure
(+)-trans-cyclobutane-1,2-diamine in 10% yield.25 The
Scheme 4.

A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
131
In order to further investigate the synthetic possibilities
of trans-cyclobutane-1,2-diamine the tetra tert-butyl
substituted salen 15a was also prepared. This was then
reacted with manganese(II) acetate tetrahydrate in an
attempt to synthesise 17a. A brown powder was
obtained as product of the reaction; however, the elec-
trospray mass spectrum of this material showed no
trace of a signal at the correct m/z of 571, instead a
number of other peaks were present, with the base peak
at m/z 631. We would suggest that in both the
attempted insertions described above the product con-
sisted of polymeric material. It is likely that in the
cyclobutyl case the NꢀCꢀCꢀN angle has become too
large and the ligand is no longer able to bind the metal
effectively.
Figure 2. Conformations of cyclopentane.
Table 1. Comparison of manganese salen catalysts 1 and
17b for epoxidation of E- and Z-b-methylstyrenes
2.4. Asymmetric epoxidation mediated by complexes 16
and 17
Substrate
Catalyst
ee
(%)
Yield
(%)
Product
configuration
E-b-Methylstyrene
E-b-Methylstyrene
Z-b-Methylstyrene
Z-b-Methylstyrene
17b
1
17b
1
28
22
84
90
67
24
25
55
(2R,3R)
(2R,3R)
(2R,3S)a
(2R,3S)a
We evaluated the epoxidation of (E)- and (Z)-b-
methylstyrene using the four complexes 16 and 17
described above. For the chromium complexes, Ph3PO
was used as a donor ligand additive and reactions were
carried out in stoichiometric mode using iodosylben-
zene as terminal oxidant. The manganese salen com-
plexes were used at a catalyst loading of 2.5 mol% with
bleach as terminal oxidant.41
a (2R,3S) refers to 2-methyl-3-phenyloxirane (b-methylstyrene oxide).
numerous other derivatives. Chromium and manganese
complexes of the salen ligands derived from trans-
cyclopentane-1,2-diamine were synthesised. The
chromium complex 16b epoxidises an E-alkene in high
ee, whereas the manganese complex 17b shows excellent
selectivity for a Z-alkene. Neither chromium nor man-
ganese were suitable metals for complexing to the salen
ligands derived from trans-cyclobutane-1,2-diamine.
This indicates that these ligands are most likely highly
non-planar, however perhaps other metals would be
suitable for complexation.
Unfortunately, but not entirely unexpectedly, epoxida-
tion of (E)-b-methylstyrene using the Cr(V)ꢁO deriva-
tive of cyclopentane diamine derived 16b in
stoichiometric mode led to a slightly reduced ee (87%)
compared to the cyclohexyl case 3 (92%). We suggest
that the conformational flexibility of the cyclopentane
ring, which can exist in either the half-chair or envelope
conformations (Fig. 2), is not favourable for achieving
high ee.
The results obtained using the manganese complex 17b
are presented in Table 1 and they are compared to the
results obtained with the cyclohexyl analogue 1. Inter-
estingly, there is a slight improvement in the selectivity
and yield with respect to Jacobsen’s catalyst 1 for
E-alkenes, but a deterioration for the Z isomers.
4. Experimental
4.1. General experimental
Despite the apparent failure to synthesise the complexes
16a and 17a, the products of the attempted complexa-
tion reactions were evaluated as epoxidation catalysts.
In the case of the chromium species the characteristic
black colour of the derived (salen)Cr(V)ꢁO species was
not generated and the reaction product showed only a
0.2% yield of trans-b-methylstyrene oxide. Similarly,
the product of the manganese insertion reaction gave
an insignificant yield of epoxide.
Melting points were determined either on a Gal-
lenkamp melting point block or a Reichert Thermovar
and are uncorrected. Optical rotation values were
obtained using a Perkin–Elmer 241 polarimeter. 1H
NMR spectra were recorded at 270 MHz on a Jeol
JNM-GX270 FT spectrometer and at 300 MHz on a
Varian INOVA 300 spectrometer. 13C NMR spectra
were recorded at 67 MHz (Jeol) or 75 MHz (Varian).
Chemical shifts are reported as l values in ppm relative
to internal standard tetramethylsilane (TMS) for 1H
spectra. Chiral gas–liquid chromatography (GC) was
performed on a Shimadzu GC-8A gas chromatograph
coupled to a Shimadzu C-R3A integrator. High-perfor-
mance liquid chromatography was performed using a
Waters 501 HPLC pump with a Waters 486 Tunable
Absorbance detector coupled to a Shimadzu C-R5A
integrator. All commercially available solvents were
3. Conclusion
1,2-Diamines have been used extensively as the source
of chiral information in asymmetric synthesis20,21 and
we have drawn together here a complete synthesis of
two such diamines. From these, four new chiral ligands
have been synthesised with the possibility existing of

132
A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
used as supplied, unless otherwise stated. Solvents were
dried according to standard procedures.42 Flash column
chromatography was performed on Merck silica 9385,
particle size 0.04–0.063 mm. Iodosylbenzene was pre-
pared according to a standard procedure.43 2-(Trifl-
uoromethyl)phenol was obtained from Fluorochem
Ltd. Z-b-Methylstyrene was obtained from Chemsam-
pco Inc., 40 Enterprise Avenue, Trenton, NJ 08638,
USA, tel.: +1-609-656-2440. All other chemicals were
obtained from the Aldrich Chemical Company and
used as received.
[h]D20=+10.0 (c 2, H2O, lit.25 +10.1°); 1H NMR (270
MHz, D2O) l 4.31 (s, 4H, CHOH), 3.60–3.56 (m, 2H,
CHN), 2.12–2.05 (m, 2H, cyclopentyl-H), 1.69–1.58 (m,
4H, cyclopentyl-H); 13C NMR (67 MHz, D2O) l 179.2,
75.6, 57.7, 32.3, 24.3.
4.5. (R,R)-(−)-trans-Cyclopentane-1,2-diamine, (R,R)-4
Following the procedure of Toftlund and Pedersen the
product of Section 4.4 was reacted to give a pale yellow
liquid (194 mg, 70%). Due to the known air sensitivity
of 1,2 diamines this was used immediately (see Sections
1
4.2. ( )-trans-1,2-Diazido-cyclopentane, 6
4.9 and 4.10): H NMR (270 MHz, CDCl3) l 2.77–2.70
(m, 2H, CHN), 2.05–1.93 (m, 2H, cyclopentyl-H), 1.72–
1.58 (m, 2H, cyclopentyl-H), 1.60 (br s, 4H, NH2),
1.39–1.28 (m, 2H, cyclopentyl-H).
Potassium permanganate (23.2 g, 0.147 mol) was added
portionwise over 4 h to a mixture of sodium azide (38.2
g, 0.587 mol) and cyclopentene (20.0 g, 0.294 mol) in
MeOH (150 mL). The reaction was stirred for a further
4 h and water (100 mL) was added. The reaction was
carefully acidified using conc. H2SO4 (20 mL was
required, caution, very exothermic). The resulting
brown suspension was extracted with Et2O (3×100 mL)
and the extracts were washed with saturated Na2HCO3
(100 mL), dried over Na2SO4 and the solvent was
removed in vacuo to yield an orange liquid (13.0 g,
29%). This was purified by flash chromatography to
yield a sample of trans-1,2-diazidocyclopentane (3.77 g,
8%) and a sample of a 9:1 mixture of the trans and
cis-1,2-diazidocyclopentane (2.04 g, 5%) In subsequent
repeats of this experiment the small amount of cis
isomer produced was not removed and the mixture was
used in the next reaction: IR (neat, cm−1) 2958, 2100
(N3), 1268; 1H NMR (300 MHz, CDCl3) l (trans
isomer) 3.78–3.52 (m, 2H, CHN3), 2.10–2.00 (m, 2H,
cyclopentyl-H), 1.84–1.66 (m, 4H, cyclopentyl-H); 13C
NMR (67 MHz, CDCl3) l (trans isomer) 67.0, 29.3,
20.9. Anal. calcd for C5H8N6: C, 39.47; H, 5.30; N,
55.23. Found C, 39.04; H, 5.18; N. 56.08%.
4.6. 2-(Methoxymethoxy)-1-(trifluoromethyl)benzene
Sodium hydride (6.40 g of a 60% dispersion in mineral
oil, 160 mmol) was washed with hexane and transferred
to a two-neck 250 mL round bottom flask under an
atmosphere of N2. After addition of anhydrous DMF
(50 mL) the slurry was cooled with stirring to 0°C. To
the resulting grey suspension was added dropwise a
solution of 2-(trifluoromethyl)phenol (24.00 g, 150
mmol) in anhydrous DMF (25 mL) at such a rate that
the evolution of hydrogen did not become too vigorous.
After complete addition the ice bath was removed and
the brown reaction mixture stirred for
1
h.
Chloromethylmethyl ether (12.4 mL, 160 mmol) was
added dropwise and the resulting white suspension
stirred overnight. Ice/water (100 mL) was added cau-
tiously and the mixture extracted with Et2O (3×100
mL). The combined Et2O extracts were washed with
NaOH (2 M, 100 mL), HCl (2 M, 100 mL), and brine
(100 mL). The solution was dried over MgSO4 and the
solvent was removed in vacuo to yield a colourless
liquid (28.2 g, 92%): IR (neat, cm−1) 2961, 2919, 2849,
1609, 1496, 1463, 1323, 1244, 1116, 1056, 989, 758, 648;
1H NMR (270 MHz, CDCl3) l 7.58 (d, J=7.6 Hz, 1H,
ArH), 7.50–7.44 (m, 1H, ArH), 7.26–7.21 (m, 1H,
ArH), 7.04 (d, J=7.9 Hz, 1H, ArH), 5.26 (s, 2H,
OCH2O), 3.49 (s, 3H, OCH3); 13C NMR (75 MHz,
CDCl3) l 155.0, 133.2, 127.1, 123.7 (q, JCꢀF=273 Hz),
121.1, 119.5 (q, JCꢀF=31 Hz), 115.2, 94.2, 56.3; MS
(EI) m/z (relative intensity) 206 (M+, 2), 145 (M−
OCH2CH3, 3), 127 (2), 95 (2), 63 (3), 45 (100). Anal.
calcd for C9H9F3O2: C, 52.43; H, 4.40; F, 27.65. Found:
C, 53.05; H, 4.62; F, 27.20%.
4.3. ( )-trans-Cyclopentane-1,2-diamine, 4
( )-trans-1,2-Diazido-cyclopentane 6 (6.0 g, 39 mmol)
was hydrogenated at 45 psi for 24 h over Pd/C (400 mg
of 10% w/w, 0.38 mmol). The suspension was then
filtered and the filtrate was evaporated in vacuo to give
a yellow liquid (2.88 g, 73%). This was not purified
further but instead was resolved immediately: IR (neat,
1
cm−1) 2955, 2361, 1600; H NMR (270 MHz, CDCl3) l
2.77–2.70 (m, 2H, CHN), 2.05–1.93 (m, 2H,
cyclopentyl-H), 1.72–1.58 (m, 2H, cyclopentyl-H), 1.60
(br s, 4H, NH2), 1.39–1.28 (m, 2H, cyclopentyl-H).
4.7. 2-(Methoxymethoxy)-3-(trifluoromethyl)benz-
aldehyde
4.4. (R,R)-(−)-trans-Cyclopentane-1,2-diamine di-(+)-
tartrate
To
a
solution of 2-(methoxymethoxy)-1-(trifl-
( )-trans-Cyclopentane-1,2-diamine
4
(2.88 g, 28.8
uoromethyl)benzene (31.9 g, 155 mmol) in anhydrous
THF (200 mL) at −78°C (acetone/dry ice) under an
atmosphere of N2 was added n-butyllithium (68.0 mL
of a 2.5 M solution, 170 mmol) dropwise with stirring.
After an additional hour stirring at this temperature, a
mixture of anhydrous DMF (13.2 mL, 170 mmol) and
anhydrous THF (10 mL) was added to the green mix-
ture and the resulting solution was allowed to warm to
mmol) and (+)-tartaric acid (10.1 g, 67.3 mmol) were
reacted according to the procedure of Toftlund and
Pedersen25 to give an off-white precipitate which was
filtered and dried (5.21 g, 90% based on one enan-
tiomer). This solid was recrystallised 3 times from
water/MeOH to give white crystals (1.23 g, 22% based
on one enantiomer): mp 126–127°C (lit.25 128–130°C);

A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
133
rt and stirred overnight. The yellow solution was
hydrolysed by the addition of water (150 mL) and the
mixture extracted with Et2O (3×150 mL). The com-
bined Et2O extracts were then washed with 2 M HCl
(100 mL) and brine (100 mL), dried over MgSO4 and
the solvent was removed in vacuo to yield a viscous
yellow liquid. This was distilled in vacuo to yield a pale
yellow liquid (29.3 g, 81%): bp 153°C (0.5 mmHg); IR
(neat, cm−1) 2956, 2923, 2866, 2360, 1672 (CꢁO), 1623,
Hz and 1.1 Hz, 2H, ArH), 6.90 (apparent t, J=7.8 Hz,
2H, ArH), 3.82–3.75 (m, 2H, N-CH), 2.28–2.15 (m, 2H,
cyclopentyl H), 2.05–1.91 (m, 4H, cyclopentyl-H); 13C
NMR (67 MHz, CDCl3) l 164.5, 160.2, 135.1, 129.9,
123.6 (q, JCꢀF=283 Hz), 119.1, 118.4, 117.7, 76.0, 33.0,
21.9; MS (EI) m/z (relative intensity) 444 (M+, 5), 255
(100), 190 (20), 170 (28), 155 (20), 127 (25), 67 (30), 41
(29). Anal. calcd for C21H18F6N2O2: C, 56.76; H, 4.08;
N, 6.30. Found: C, 56.57; H, 4.05; N, 6.15%.
1
1454, 1335, 1117, 1075, 985, 923, 757; H NMR (270
MHz, CDCl3) l 10.31 (s, 1H, CHO), 8.07 (dd, J=1.7
Hz, 7.9 Hz, 1H, ArH), 7.90–7.87 (m, 1H, ArH), 7.38
(apparent t, J=7.9 Hz, 1H, ArH), 5.13 (s, 2H,
OCH2O), 3.63 (s, 1H, OCH3); 13C NMR (75 MHz,
CDCl3) l 189.6, 158.6, 132.7, 132.5, 131.5, 124.8, 123.0
(q, JCꢀF=273 Hz), 118.4, 102.5, 58.1; MS (EI) m/z
(relative intensity) 234 (M+, 9), 203 (M−OCH3, 4), 188
(3), 132 (2), 113 (3), 69 (6), 63 (5), 45 (100), 29 (47).
Anal. calcd for C10H9F3O3: C, 51.29; H, 3.87; F, 24.34.
Found: C, 51.13; H, 4.03; F, 24.14%.
4.10. (R,R)-(−)-N,N%-Bis(3,5-di-tert-butylsalicylidene)-
trans-cyclopentane-1,2-diamine, 15b
(R,R)-(−)-trans-Cyclopentane-1,2-diamine (R,R)-4 (188
mg, 1.88 mmol) and 3,5-di-tert-butyl-2-hydroxybenz-
aldehyde (880 mg, 3.75 mmol) were refluxed in EtOH
(10 mL) for 2 h. On cooling to rt crystals formed
which were filtered and dried (741 mg, 74%): mp 166–
168°C; [h]2D0=−365 (c 1, CH2Cl2); 1H NMR (270
MHz, CDCl3) l 13.67 (s, 2H, OH), 8.31 (s, 2H,
NꢁCH), 7.34 (d, J=2.5 Hz, 2H, ArH), 7.02 (d, J=2.5
Hz, 2H, ArH), 3.76–3.72 (m, 2H, CꢁN-CH), 2.27–2.13
(m, 2H, cyclopentyl-H), 2.05–1.91 (m, 4H, cyclopentyl-
H), 1.44 (s, 9H, C(CH3)), 1.26 (s, 9H, C(CH3)); 13C
NMR (67 MHz, CDCl3) l 165.7, 158.0, 140.1, 136.5,
126.9, 126.2, 117.8, 76.5, 35.1, 34.2, 33.3, 31.5, 29.5,
22.2; MS (EI) m/z (relative intensity) 533 (M+1, 9) 532
(M+, 27), 299 (100), 251 (25), 234 (13), 218 (10), 57
(26). Anal. calcd for C35H52N2O2: C, 78.90; H, 9.84;
N, 5.26. Found: C, 78.54; H, 9.47; N, 5.40%.
4.8. 2-Hydroxy-3-(trifluoromethyl)benzaldehyde
2-(Methoxymethoxy)-3-(trifluoromethyl)benzaldehyde
(15.0 g, 64.0 mmol) was dissolved in MeOH (300 mL)
and AcOH (0.2 M, 200 mL). The mixture was heated
under reflux for 8 h, at which stage TLC analysis (silica,
CH2Cl2) indicated the complete disappearance of start-
ing material. After cooling overnight white needle like
crystals formed in the reaction. These were filtered off,
and dried in a dessicator until of constant weight (5.44
g, 45%). Drying under vacuum had to be avoided as it
led to loss of material by sublimation. The filtrate was
concentrated to a yield a further crop which were dried
similarly (3.27 g, 27%): mp 58–59°C; IR (KBr, cm−1)
3103, 2867, 1680 (CꢁO), 1622, 1487, 1451, 1337, 1190,
4.11. meso-Dimethyl 2,5-dibromohexane-1,6-dioate, 9
A combination of the procedures of Buchman et al.29
and Watson and O’Neill34 were used. Thionyl chloride
(323 g, 2.71 mol) was added in 70 mL portions over 2
h to adipic acid 7 (197 g, 1.35 mol) heated at 80°C in
a two-neck 1 L round bottom flask equipped with a
reflux condenser and a pressure equalised dropping
funnel. After heating until gas evolution ceased some
solid (adipic acid) still remained in the reaction. There-
fore, a further 100 mL of thionyl chloride was added
and heating continued until gas evolution ceased. The
addition took 7 h in total. Following this bromine
(473 g, 2.96 mol) was added dropwise to the pale
yellow hot reaction mixture over 12 h and heating was
then continued for a further 3 h. After cooling to rt,
N2 was passed through the reaction to remove excess
bromine. The resulting brown reaction mixture was
added dropwise to MeOH (275 mL) in a 1 L round
bottom flask cooled in an ice bath. A white precipitate
formed during the addition and this was filtered as
soon as addition was complete and recrystallised from
MeOH (201 g, 45%). On standing further precipitate
formed in the mother liquor which was also recrys-
tallised from MeOH (52 g, 12%). The two crops were
1
1118, 747, 671, 486; H NMR (270 MHz, CDCl3) l
11.73 (s, 1H, OH), 9.96 (s, 1H, CHO), 7.85–7.76 (m,
2H, ArH), 7.15–7.09 (m, 1H, ArH); 13C NMR (67
MHz, CDCl3) l 196.3, 159.7, 137.4, 134.2, 122.9 (q,
J
CꢀF=273 Hz), 121.2, 119.2, 118.9 (q, JCꢀF=32 Hz);
MS (EI) m/z (relative intensity) 191 (M+1, 18), 190
(M+, 100), 189 (M−1, 33), 172 (29), 171 (19), 170 (4),
169 (66), 144 (23), 143 (7), 142 (65), 141 (47), 114 (42),
63 (37), 29 (39). Anal. calcd for C8H5F3O2: C, 50.54; H,
2.65; F, 29.98. Found: C, 50.21; H, 2.67; F, 30.04%.
4.9. 4.8(R,R)-(−)-N,N%-Bis(3-trifluoromethylsalicylidene)-
trans-cyclopentane-1,2-diamine, 14b
(R,R)-(−)-trans-Cyclopentane-1,2-diamine (R,R)-4 (194
mg, 1.94 mmol) was dissolved in EtOH (70 mL) and
2-hydroxy-3-trifluoromethylbenzaldehyde (737 mg, 3.87
mmol) added to the solution. The resulting bright yel-
low mixture was refluxed for 2 h, cooled and the
solvent was removed in vacuo to yield a solid which
was recrystallised from hexane to give bright yellow
needles (765 mg, 89%): mp 138°C; [h]2D0=−437; IR
(KBr, cm−1) 2969, 2882, 1633, 1497, 1457, 1333, 1190,
1
identical: mp 76–77°C (lit.34 75–76°C); H NMR (270
MHz, CDCl3) l 4.30–4.23 (m, 2H, CHBr), 3.80 (s, 6H,
OCH3), 2.36–2.35 (m, 2H, CH-CBr), 2.10–2.02 (m,
2H, CH-CBr); 13C NMR (67 MHz, CDCl3) l 170.2,
53.7, 44.8, 33.0.
1
1160, 1077, 842, 752, 685, 609; H NMR (270 MHz,
CDCl3) l 14.48 (s, 2H, OH), 8.34 (s, 2H, HCꢁN), 7.59
(dd, J=7.9 Hz and 1.1 Hz, 2H, ArH), 7.38 (dd, J=7.8

134
A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
4.12. 1-Cyano-cyclobutane-1,2-dicarboxylic acid
dimethyl ester, 10
4.15. trans-Cyclobutane-1,2-dicarbonyl dichloride, 13
trans-Cyclobutane-1,2-dicarboxylic acid 8 (7.99 g, 55.4
mmol) and thionyl chloride (15 mL, 206 mmol) were
heated at 70°C in benzene (35 mL) for 24 h. Benzene
and excess thionyl chloride were distilled off and the
brown residue dried under water pressure and purified
by Kugelrohr distillation to yield a colourless liquid
(9.62 g, 96%): bp ꢀ65°C at 0.1 mmHg, (lit.44 51°C at
Using a modification of the procedure of Buchman et
al.,29 meso-dimethyl 2,5-dibromohexane-1,6-dioate 9
(200 g, 0.602 mol) and potassium cyanide (88.3 g, 1.36
mol) were added to a 1 L round bottom flask contain-
ing MeOH (400 mL). The reaction mixture was heated
at 75°C for 56 h. After cooling to rt the MeOH was
distilled at atmospheric pressure and the residue was
flushed through a pad of silica with CH2Cl2. The sol-
vent was removed in vacuo to yield a yellow liquid
1
0.2 mmHg); H NMR (270 MHz, CDCl3) l 3.96–3.86
(m, 2H, CHCꢁO), 2.47–2.27 (m, 4H, cyclobutyl-H); 13C
NMR (67 MHz, CDCl3) l 173.2, 50.4, 22.1.
1
(84.1 g, 71%). H NMR indicated this to be a 43:57
mixture of the two isomers of product. On standing
crystals formed in the liquid and these were filtered and
dried (21.4 g, 18%). Distillation of the mother liquor
yielded a colourless liquid (bp 94°C at 0.2 mmHg (lit.29
119–120°C at 2 mmHg), 30.9 g, 26%): Crystalline iso-
mer: mp 92–93°C (lit.29 89.5–90°C) from MeOH; 1H
NMR (270 MHz, CDCl3) l 3.85 (s, 3H, OCH3), 3.79–
3.72 (m, 1H, cyclobutyl-H), 3.72 (s, 3H, OCH3), 2.75–
2.53 (m, 3H, cyclobutyl-H), 2.42–2.26 (m, 1H,
cyclobutyl-H); 13C NMR (67 MHz, CDCl3) l 170.2,
167.8, 118.7, 53.7, 52.4, 45.3, 42.2, 28.4, 21.4; liquid
isomer: bp 94°C at 0.2 mmHg (lit.29 119–120°C at 2
4.16. trans-Cyclobutane-1,2-diamine dihydrochloride,
5·2HCl
A solution of trans-cyclobutane-1,2-dicarbonyl dichlo-
ride (13) (9.50 g, 52.5 mmol) in benzene (60 mL) was
added dropwise over 5 min to a solution of sodium
azide (11.9 g, 184 mmol) in water (60 mL) cooled to
0°C. The resulting two phase mixture was stirred vigor-
ously for 2 h after which time the phases were separated
and the organic phase was washed with 5% Na2HCO3
(20 mL), water (20 mL) and dried over CaCl2. This
benzene solution of diacyl azide (CAUTION: explosion
risk) was decanted into a fresh flask equipped with a
reflux condenser and oil bubbler and heated slowly to
50°C. Initial slow evolution of gas became vigorous at
ꢀ40°C so heat was removed for a period. After gas
evolution had ceased heat was reapplied at 50°C for a
further hour. After cooling to rt 20% HCl (20 mL) was
added and the reaction heated to 90°C for 4 h, then
allowed to cool to rt overnight. The benzene layer was
separated and washed with water (50 mL). The aqueous
layers were combined, washed with benzene (100 mL)
and the water removed in vacuo to yield a brown solid
which was recrystallised from MeOH/Et2O to give a
white solid (6.29g, 46%): mp 256–257°C (lit.27 248–
1
mmHg); H NMR (270 MHz, CDCl3) l 3.88 (s, 3H,
OCH3), 3.83–3.76 (m, 1H, cyclobutyl-H), 3.80 (s, 3H,
OCH3), 2.69–2.53 (m, 3H, cyclobutyl-H), 2.35–2.27 (m,
1H, cyclobutyl-H); 13C NMR (67 MHz, CDCl3) l
169.9, 167.8, 117.5, 54.1, 52.6, 43.9, 43.6, 29.0, 20.2.
4.13. Cyclobutane-1,2-dicarboxylic acid, 12
Using a modification of the procedure of Buchman et
al.29 a mixture of the two isomers of 1-cyano-cyclobu-
tane-1,2-dicarboxylic acid dimethyl ester 10 (46.3 g,
0.234 mol) and 6 M HCl (117 mL) were refluxed for 12
h. The mixture was concentrated in vacuo until a white
solid precipitated. Et2O (200 mL) was added to the
residue and the mixture filtered. The filtrate was washed
with water (50 mL), dried over Na2SO4 and the solvent
was removed in vacuo to yield a pale yellow liquid (33.4
g). This liquid was heated at 180°C under vacuum
(water pump) until gas evolution ceased (ꢀ2.5 h). On
1
251°C); H NMR (270 MHz, D2O) l 3.82–3.74 (m, 2H,
CHN), 2.20–2.09 (m, 2H, cyclobutyl-H), 1.87–1.72 (m,
2H, cyclobutyl-H); 13C NMR (75 MHz, D2O) l 48.6,
20.7.
4.17. ( )-trans-Cyclobutane-1,2-diamine, 5
1
cooling a brown solid was formed (21.0 g, 62%). H
NMR indicated this to be a 70:30 mixture (12) of the
trans and cis diacids: 1H NMR (270 MHz, D2O) l
(trans isomer) 3.09–3.02 (m, 2H, CHCꢁO), 1.79–1.76
trans-Cyclobutane-1,2-diamine
dihydrochloride,
5.2·HCl (1.22 g, 7.67 mmol) was dissolved in water (3
mL) and added to a separatory funnel containing
freshly ground KOH (1.72 g, 30.7 mmol). The mixture
was shaken vigorously and then extracted with CHCl3
(4×10 mL). The CHCl3 extracts were dried over
Na2SO4 and the solvent was removed in vacuo to a pale
yellow liquid (612 mg, 93%): 1H NMR (300 MHz,
CDCl3) l 2.90–2.79 (m, 2H, CHN), 2.09–1.96 (m, 2H,
cyclobutyl-H), 1.66 (br s, 4H, NH2), 1.39–1.16 (m, 2H,
cyclobutyl-H); 13C NMR (75 MHz, CDCl3) l 59.8,
26.1.
1
(m, 4H, cyclobutyl-H); H NMR (270 MHz, D2O) l
(cis isomer) 3.17–3.11 (m, 2H, CHCꢁO), 1.87–1.84 (m,
4H, cyclobutyl-H).
4.14. trans-Cyclobutane-1,2-dicarboxylic acid, 8
The mixture of cis- and trans-cyclobutane-1,2-dicar-
boxylic acid 12 (18.9 g, 0.131 mol) from the previous
reaction was subjected to the equilibration procedure of
Neibecker and co-workers35 to give grey crystals (10.3
1
g, 54%). H and 13C NMR indicated these to be the
4.18. Resolution of trans-cyclobutane-1,2-diamine using
tartaric acid
1
pure trans-isomer: mp 133–135°C (lit.29 130–131°C); H
NMR (270 MHz, D2O) l 3.09–3.02 (m, 2H, CHCꢁO),
1.79–1.76 (m, 4H, cyclobutyl-H); 13C NMR (67 MHz,
D2O) l 178.0, 40.2, 21.4.
( )-trans-Cyclobutane-1,2-diamine (5) (2.11 g, 24.5
mmol) was reacted with (+)-tartaric (8.64 g, 57.6 mmol)

A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
135
1
acid according to the procedure of Toftlund and Peder-
son,25 to give a white precipitate (1.95 g, 20%) with
1185, 1147, 1114, 1060, 842, 754, 618; H NMR (300
MHz, CDCl3) l 14.48 (s, 2H, OH), 8.34 (s, 2H, HCꢁN),
7.62 (dd, J=7.6 Hz and 0.9 Hz, 2H, ArH), 7.43 (dd,
J=7.6 Hz and 1.5 Hz, 2H, ArH), 6.90 (dd, J=7.6 Hz
and 7.6 Hz, 2H, ArH), 4.19–4.14 (m, 2H, N-CH),
2.40–2.35 (m, 2H, cyclobutyl H), 2.14–2.11 (m, 2H,
cyclobutyl-H); 13C NMR (67 MHz, CDCl3) l 163.9,
160.4, 135.6, 130.3 (q, JCꢀF=5 Hz), 123.9 (q, JCꢀF=273
Hz), 119.3, 118.5, 117.9, 68.7, 33.0, 24.2. Anal. calcd for
C20H16F6N2O2 C, 55.82; H, 3.75; N, 6.51. Found: C,
55.54; H, 3.69; N, 6.39. MS (EI) m/z (relative intensity)
430 (M+, 1), 411 (3), 241 (100), 200 (28), 186 (43), 168
(26), 80 (24), 67 (23).
1
[h]2D0=+23.0 (c 1, H2O). H NMR analysis suggested
that this substance is a salt containing a 2:1 ratio of
1
(+)-tartaric acid and trans-cyclobutane-1,2-diamine: H
NMR (300 MHz, D2O) l 4.46 (s, 4H), 4.0–3.9 (m, 2H),
2.38–2.26 (m, 2H), 2.02–1.88 (m, 2H); 13C NMR (67
MHz, D2O) l 179.2, 75.6, 50.9, 23.3. Three recrystalli-
sations of this material from MeOH/H2O (1:1) gave
white crystals (961 mg, 10%); [h]2D0=+28.0 (c 1, H2O)
mp 124–126°C; IR (KBr, cm−1) 3318, 3269, 2975, 1565,
1
1474, 1408, 1304, 1263, 1214, 1135, 1066, 677, 483; H
NMR (300 MHz, D2O) l 4.46 (s, 4H), 4.0–3.9 (m, 2H),
2.38–2.26 (m, 2H), 2.02–1.88 (m, 2H); 13C NMR (67
MHz, D2O) l 179.2, 75.6, 50.9, 23.3. Further recrys-
tallisation failed to change the specific rotation value
any further. The enantiomeric excess of the diamine
component of the salt was determined using a modifica-
4.20.
(+)-N,N%-Bis(3,5-di-tert-butylsalicylidene)-trans-
cyclobutane-1,2-diamine, 15a
(+)-trans-Cyclobutane-1,2-diamine,
5 (32 mg, 376
tion of the method of Jacobsen and co-workers:40
a
mmol) and 3,5-di-tert-butyl-2-hydroxybenzaldehyde
(176 mg, 752 mmol) were refluxed in EtOH (10 mL) for
2 h. On cooling to rt crystals formed which were filtered
and dried (160 mg, 82%): mp 165–166°C; [h]2D0=+400 (c
1, CH2Cl2); IR (KBr, cm−1) 2955, 2900, 1627 (CꢁN),
small sample of the salt (25 mg) was suspended in
CH2Cl2 (2 mL) in a test tube and a NaOH solution (4
M, 0.5 mL) was added. This mixture was mixed vigor-
ously using a ‘whirlimixer’. 2-Hydroxy-3-(trifluoro-
methyl)benzaldehyde (10 mg) was added and after fur-
ther vigorous mixing a 250 mL aliquot was removed,
diluted to 5 mL with hexane and analysed by chiral
HPLC (Chiralcel AD column, hexane/iPrOH 99:1, 0.6
mL/min). This indicated the presence of only one enan-
tiomer of the diamine derivative 14a.
1
1464, 1436, 1360, 1272, 1250, 1171, 802, 763; H NMR
(300 MHz, CDCl3) l 13.65 (s, 2H, OH), 8.32 (s, 2H,
NꢁCH), 7.37 (d, J=2.3 Hz, 2H, ArH), 7.07 (d, J=2.3
Hz, 2H, ArH), 4.16–4.08 (m, 2H, CꢁN-CH), 2.37–2.27
(m, 2H, cyclopentyl-H), 2.19–2.03 (m, 2H, cyclopentyl-
H), 1.45 (s, 9H, C(CH3)), 1.29 (s, 9H, C(CH3)); 13C
NMR (67 MHz, CDCl3) l 164.8, 158.0, 140.2, 136.6,
127.1, 126.2, 117.6, 69.0, 35.0, 34.1, 31.5, 29.7, 24.1; MS
(EI) m/z (relative intensity) 518 (M+, 1), 285 (69), 244
(23), 234 (20), 57 (100). Anal. calcd for C34H50N2O2: C,
78.72; H, 9.71; N, 5.40. Found: C, 78.63; H, 9.76; N,
5.44%.
Further precipitate subsequently appeared in the
mother liquor of the initial reaction. This was recrys-
tallised twice from MeOH/H2O (1:1) to give white
crystals (304 mg, 3%) with [h]2D0=+12.2 (c 1, H2O).
Application of the procedure for determination of ee of
the diamine indicated that this salt contained the oppo-
site enantiomer to the initial precipitate, in enantiopure
form.
4.21. [(R,R)-(−)-N,N%-Bis(3-trifluoromethylsalicylidene)-
trans-cyclopentane-1,2-diamine chromium(III)] nitrate,
16b
The remaining mother liquors from the reaction were
combined and treated with KOH to regenerate 5 (220
mg, 2.55 mmol). This was reacted, in the same manner
as before, but with (−)-tartaric acid, to give a white
solid (300 mg, 30%). 1H NMR indicated this to be
bis-salt. One recrystallisation of this material from
MeOH/H2O (1:1) gave white crystals (115 mg, 12%)
with [h]2D0=−27.9 (c 1, H2O). Given that this shows
opposite specific rotation to the initial salt above it is a
confirmation of the success of the resolution. HPLC
analysis by the method above confirmed the presence of
enantiopure (−) diamine.
(R,R)-(−)-N,N%-Bis(3-trifluoromethylsalicylidene)-trans-
cyclopentane-1,2-diamine (350 mg, 787 mmol) and
chromium(II) chloride (97 mg, 787 mmol) were reacted
according to the procedure of Jacobsen and co-workers 39
to give a brown solid (250 mg, 60%). This was redis-
solved in methanol (25 mL) and a solution of silver
nitrate (88 mg, 1.1 equiv.) in water (2 mL) was added.
The resulting precipitate of silver chloride was filtered
and the filtrate concentrated to yield a brown solid (126
mg, 46%): mp >230°C; IR (KBr, cm−1) 2948, 2749,
1643, 1561, 1439, 1384, 1301, 1180, 1120, 1077, 860,
762, 647. Anal. calcd for C21H16CrF6N3O5.H2O.MeOH:
C, 43.57; H, 3.66; N, 6.93. Found: C, 43.15; H, 3.70; N,
6.32%. MS (Electrospray) m/z (relative intensity) 557.6
(M++1, 10), 525.6 (32), 511.5 (65), 494.3 (M−NO3, 100).
4.19. (+)-N,N%-Bis(3-trifluoromethylsalicylidene)-trans-
cyclobutane-1,2-diamine (14a)
(+)-trans-cyclobutane-1,2-diamine, (+)-5 (53 mg, 613
mmol), obtained from its bis-tartrate salt using the
procedure in Section 4.5, was reacted with 2-hydroxy-3-
(trifluoromethyl)benzaldehyde (233 mg, 1.23 mmol)
according to the method in Section 4.9 to give a yellow
oil. Recrystallisation from hexane gave a yellow solid
(198 mg, 75%): [h]D20=+485 (c 1, CH2Cl2); mp 115°C;
IR (KBr, cm−1) 2995, 2957, 2895, 1629, 1452, 1331,
4.22. [(R,R)-(−)-N,N%-Bis(3,5-di-tert-butylsalicylidene)-
trans-cyclopentane-1,2-diamine manganese(III)] chloride,
17b
Following the method of Jacobsen45 manganese(II) acet-
ate tetrahydrate (897 mg, 3.66 mmol) and (R,R)-(−)-N,
N%-bis(3,5-di-tert-butylsalicylidene)-trans-cyclopentane-

136
A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
1,2-diamine (650 mg, 1.22 mmol) were reacted to yield
a brown oil. This was purified by flash chromatography
(silica, CH2Cl2 then EtOH) to yield unreacted ligand
(189 mg, 29%) and the desired complex as a brown
solid (37 mg, 8%): mp>300°C; IR (KBr, cm−1) 2958,
1620, 1534, 1432, 1358, 1344, 1305, 1271, 1250, 1174,
839, 749, 542; MS (Electrospray) m/z 585 (M−Cl, 100).
Anal. calcd for C35H50ClMnN2O2·1/2H2O: C, 66.71; H,
8.16; Cl, 5.63; N, 4.45. Found: C, 66.70; H, 8.01; Cl,
5.81; N, 4.32%.
Acknowledgements
We acknowledge support from Enterprise Ireland
(Grant SC/97/536) and University College Dublin for a
Demonstratorship (AMD). We also thank Professor
Per-Ola Norrby for helpful discussions and Marie
Renehan for initial investigations of the trans-cyclopen-
tane-1,2-diamine synthesis.
References
4.23. Procedures for analysis of epoxidation reaction
product mixtures
1. Katsuki, T. In Catalytic Asymmetric Synthesis; 2nd ed.;
Ojima, I., Ed. Asymmetric epoxidation of unfunction-
alised olefins and related reactions; Wiley-VCH: New
York, 2000; pp. 287–325.
Typically the worked up reaction mixture (see below)
was dissolved in ꢀ1 mL of Et2O and 1 mL of n-decane
was added as an internal standard. 1 mL of this solution
was injected onto a GC column. The ee of trans-b-
methylstyrene oxide was determined with a Supelco
cyclodextrin-a capillary column (alphadex 120), 30 m×
0.25 mm i.d., 0.25 mm film operated at an injection
temperature of 230°C and a column temperature of
93°C, with a column pressure of 18 psi. The ee of
2. Dalton, C. T.; Ryan, K. M.; Wall, V. M.; Bousquet, C.;
Gilheany, D. G. Top. Catal. 1998, 5, 71–90.
3. Jacobsen, E. N.; Wu, M. H. In Comprehensive Asymmet-
ric Catalysis; Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H.,
Eds. Epoxidation of alkenes other than allylic alcohols;
Springer: New York, 1999; Vol. II, pp. 649–677.
4. Johnson, R. A.; Sharpless, K. B. In Catalytic Asymmetric
Synthesis; 2nd ed.; Ojima, I., Ed. Catalytic asymmetric
epoxidation of allylic alcohols; Wiley-VCH: New York,
2000; pp. 231–280.
cis-b-methylstyrene oxide was determined with
a
Supelco cyclodextrin-b capillary column (betadex 120),
30 m×0.25 mm i.d., 0.25 mm film. This was operated at
an injection temperature of 230°C and a column tem-
perature of 77°C with a column pressure of 20 psi. The
absolute configuration of trans-b-methylstyrene oxide
was assigned by comparison of a sample with the data
of Witkop and Foltz46 and of Shi and co-workers.47
The absolute configuration of cis-b-methylstyrene oxide
was assigned by comparison of the GC retention times
to those of a sample made according to the Jacobsen
method.41
5. Bousquet, C.; Gilheany, D. G. Tetrahedron Lett. 1995,
36, 7739–7742.
6. Dalton, C. T.; Ryan, K. M.; Coyne, E. J.; Wall, V. M.;
Bousquet, C.; Gilheany, D. G. Reported as a series of
oral presentations at the 211th National Meeting of the
American Chemical Society, New Orleans, 1996; (ORGN
161–165).
7. Daly, A. M.; Dalton, C. T.; Renehan, M. F.; Gilheany,
D. G. Tetrahedron Lett. 1999, 40, 3617–3620.
8. Ryan, K. M.; Bousquet, C.; Gilheany, D. G. Tetrahedron
Lett. 1999, 40, 3613–3616.
4.24. Epoxidation of (E)-b-methylstyrene by the
Cr(V)ꢁO derivative of 16b
9. Daly, A. M.; Renehan, M. F.; Gilheany, D. G. Org. Lett.
2001, 3, 663–665.
10. O’Mahony, C. P.; McGarrigle, E. M.; Renehan, M. F.;
Ryan, K. M.; Kerrigan, N. J.; Bousquet, C.; Gilheany, D.
G. Org. Lett. 2001, 3, 3435–3438.
11. Kerrigan, N. J.; Langan, I. J.; Dalton, C. T.; Daly, A.
M.; Bousquet, C.; Gilheany, D. G. Tetrahedron Lett.
2002, 43, 2107–2110.
Iodosylbenzene (1.2 equiv.) was added to a stirred
solution of 16b (30 mg, 1 equiv.) in CH3CN (5 mL). A
deep blue/black colour appeared almost immediately.
After stirring for 30 min this solution was filtered and
the filtrate cooled to 0°C using an ice/water bath.
Ph3PO (1 equiv.) was added, followed 5 min later by
(E)-b-methylstyrene (1 equiv.). The reaction mixture
was stirred at 0°C until the brown colour of the 16b
complex returned completely (ꢀ1.5 h). The solvent was
removed in vacuo and the residue treated with Et2O.
The Et2O washings were flushed through a short alu-
mina column using Et2O as eluant. The eluant was
concentrated in vacuo to a small volume (ꢀ1 mL) and
this sample was analysed by GC as described above.
12. Brandt, P.; Norrby, P.-O.; Daly, A. M.; Gilheany, D. G.
Chem. Eur. J. 2002, 8, 4299–4307.
13. Samsel, E. G.; Srinivasan, K.; Kochi, J. K. J. Am. Chem.
Soc. 1985, 107, 7606–7617.
14. Takeda, T.; Irie, R.; Shinoda, Y.; Katsuki, T. Synlett
1999, 1157–1159.
15. Nishikori, H.; Ohta, C.; Katsuki, T. Synlett 2000, 1557–
1560.
16. Houk, K. N.; DeMello, N. C.; Condroski, K.; Fennen, J.;
Kasuga, T. Origins of Stereoselectivity in Jacobsen Epox-
idations, at Electronic Conference on Heterocyclic Chem-
istry, ECHET96; Rzepa, H. S., Snyder, J. P.; Leach, C.,
Eds., London, 24/6–22/7, 1996; (http://www.ch.ic.ac.uk/
ectoc/echet96/).
17. Ito, Y. N.; Katsuki, T. Tetrahedron Lett. 1998, 39, 4325–
4328.
18. Hamada, T.; Fukuda, T.; Imanishi, H.; Katsuki, T. Tet-
rahedron 1996, 52, 515–530.
4.25. Epoxidation of alkenes catalysed by manganese
complex 17b
Epoxidation of both (E)-b-methylstyrene and (Z)-b-
methylstyrene was carried out according to the proce-
dure of Jacobsen and co-workers.41 The worked up
reaction mixture was analysed by GC as described
above.

A. M. Daly, D. G. Gilheany / Tetrahedron: Asymmetry 14 (2003) 127–137
137
34. Watson, H. A.; O’Neill, B. T. J. Org. Chem. 1990, 55,
2950–2952.
35. Brunet, J.-J.; Herbowski, A.; Neibecker, D. Synth. Com-
19. Irie, R.; Hashihayata, T.; Katsuki, T.; Akita, M.; Moro-
oka, Y. Chem. Lett. 1998, 1041–1042.
20. Bennani, Y. L.; Hanessian, S. Chem. Rev. 1997, 97,
3161–3195.
mun. 1996, 26, 483–493.
36. Witiak, D. T.; Lee, H. J.; Hart, R. W.; Gibson, R. E. J.
Med. Chem. 1977, 20, 630–635.
37. Overberger, C. G.; Nishiyama, T. J. Polym. Sci. Polym.
21. Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem.,
Int. Ed. 1998, 37, 2580–2627.
22. Jaeger, F. M.; Blumendal, H. B. Z. Anorg. Chem. 1928,
175, 161–231.
Chem. Ed. 1981, 19, 311–330.
38. Vries, T.; Wynberg, H.; van Echten, E.; Koek, J.; ten
Hoeve, W.; Kellogg, R. M.; Broxterman, Q. B.; Min-
naard, A.; Kaptein, B.; van der Sluis, S.; Hulshof, L.;
Kooistra, J. Angew. Chem., Int. Ed. 1998, 37, 2349–2354.
39. Martinez, L. E.; Leighton, J. L.; Carsten, D. H.; Jacob-
sen, E. N. J. Am. Chem. Soc. 1995, 117, 5897–5898.
40. Larrow, J. F.; Jacobsen, E. N. Org. Synth. 1998, 75,
1–11.
41. Jacobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J. R.;
Deng, L. J. Am. Chem. Soc. 1991, 113, 7063–7064.
42. Leonard, J.; Lygo, B.; Procter, G. Advanced Practical
Organic Chemistry; 2nd ed.; Blackie: London, 1995.
43. Saltzmann, H.; Sharefkin, J. G. Org. Synth. Coll. Vol.
1973, 5, 658–659.
23. Potter, G. W. H.; Coleman, M. W.; Monro, A. M. J.
Heterocyclic Chem. 1975, 12, 611–614.
24. Minisci, F.; Galli, R.; Cecere, M. Gazz. Chim. Ital. 1964,
94, 67–90.
25. Toftlund, H.; Pedersen, E. Acta Chem. Scand. 1972, 26,
4019–4030.
26. Ongeri, S.; Aitken, D. J.; Husson, H.-P. Synth. Commun.
2000, 30, 2593–2597.
27. Vergne, F.; Partogyan, K.; Aitken, D. J.; Husson, H.-P.
Tetrahedron 1996, 52, 2421–2428.
28. O’Donnell, M. J.; Bruder, W. A.; Eckrich, T. M.; Shul-
lenberger, D. F.; Staten, G. S. Synthesis 1984, 127–128.
29. Buchman, E. R.; Reims, A. O.; Skei, T.; Schlatter, M. J.
J. Am. Chem. Soc. 1942, 64, 2696–2700.
44. Woods, C. W.; Borkovec, A. B.; Hart, F. M. J. Med.
30. Shuikina, Z. J. Gen. Chem. USSR 1943, 13, 373–381.
31. Sturm, P. A.; Cory, M.; Henry, D. W.; McCall, J. W.;
Ziegler, J. B. J. Med. Chem. 1977, 20, 1327–1333.
32. Natansohn, A.; Yang, H.; Murti, D. K.; Popovic, Z. D.
Chem. Mater. 1993, 5, 1370–1371.
Chem. 1964, 7, 371–373.
45. Larrow, J. F.; Jacobsen, E. N.; Gao, Y.; Hong, Y.; Nie,
X.; Zepp, L. M. J. Org. Chem. 1994, 59, 1939–1942.
46. Witkop, B.; Foltz, M. J. Am. Chem. Soc. 1957, 79,
197–201.
47. Tu, Y.; Wang, Z.-X.; Shi, Y. J. Am. Chem. Soc. 1996,
118, 9806–9807.
33. Yang, H.; Jin, A.; Natansohn, A. J. Polym. Sci. Polym.
Chem. Ed. 1992, 30, 1953–1959.