Studies on the true catalyst in the phosphate-promoted desymmetrization of meso-aziridines with silylated nucleophiles

Article abstract of DOI:10.1016/j.tet.2012.10.068

Significant amounts of metal phosphate impurities have been detected by NMR spectroscopy and ICP-OES analysis of synthetic, as well as purchased, samples of VAPOL hydrogen phosphate purified on silica gel. The previously reported VAPOL hydrogen phosphate-catalyzed desymmetrization of meso-aziridines with different silylated compounds, has been re-examined. While metal-free phosphoric acid exhibited low activity and enantioselectivity, calcium and magnesium phosphate salts proved to be the true catalysts in these related processes. Reproducible high enantioselectivities were obtained in the ring-opening of cyclic meso-aziridines with Me₃SiX (X=SPh, SePh, N₃) employing a 1:1 mixture of calcium and magnesium VAPOL phosphates. The reaction has been successfully extended to Me₃SiSBn, Me₃SiSMe and Me₃SiNCS, giving ring-opened products in high yield and unprecedented moderate to good enantioselectivities.

Full text of DOI:10.1016/j.tet.2012.10.068

Tetrahedron 69 (2013) 50e56
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Studies on the true catalyst in the phosphate-promoted
desymmetrization of meso-aziridines with silylated nucleophiles
*
Giorgio Della Sala
ꢀ
Dipartimento di Chimica e Biologia, Universita degli Studi di Salerno, Via Ponte Don Melillo, 84084 Fisciano (SA), Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 August 2012
Received in revised form 1 October 2012
Accepted 22 October 2012
Available online 29 October 2012
Significant amounts of metal phosphate impurities have been detected by NMR spectroscopy and ICP-
OES analysis of synthetic, as well as purchased, samples of VAPOL hydrogen phosphate purified on sil-
ica gel. The previously reported VAPOL hydrogen phosphate-catalyzed desymmetrization of meso-azir-
idines with different silylated compounds, has been re-examined. While metal-free phosphoric acid
exhibited low activity and enantioselectivity, calcium and magnesium phosphate salts proved to be the
true catalysts in these related processes. Reproducible high enantioselectivities were obtained in the
ring-opening of cyclic meso-aziridines with Me3SiX (X¼SPh, SePh, N3) employing a 1:1 mixture of cal-
cium and magnesium VAPOL phosphates. The reaction has been successfully extended to Me3SiSBn,
Me3SiSMe and Me3SiNCS, giving ring-opened products in high yield and unprecedented moderate to
good enantioselectivities.
Keywords:
Asymmetric catalysis
Aziridines
Calcium
Desymmetrizations
Magnesium
Ó 2012 Elsevier Ltd. All rights reserved.
Phosphates
1. Introduction
enantioselectivities were achieved after appropriate modifications
of both the nitrogen activating group and experimental details. In
Since the early works of groups of Terada1 and Akiyama2 in
2004, phosphoric acids derived from axially chiral biaryls have
rapidly emerged as highly efficient and widely applicable enan-
tioselective organocatalysts.3 The universally recognized mode of
action of these catalysts is based on the dual acid/base function of
the phosphoric group. In fact, while the acidic OH functionality is
able to activate a wide range of electrophiles through the formation
of hydrogen bonds or ionic pairs, the P]O basic group has the
complementary function of activating protic pronucleophiles.
However, during the last years, chiral phosphoric acids have also
been reported to be involved in catalysis with silylated nucleo-
philes. A highly enantioselective ring opening desymmetrization of
activated meso-aziridines 2 with Me3SiN3, catalyzed by the VAPOL-
derived phosphoric acid 1, was described by Antilla and co-
workers.4 The authors proposed the mechanism depicted in
Scheme 1, with the phosphoric acid 1 acting as a Lewis base pro-
moter rather than a proton catalyst. The activation of the silane was
suggested to result in the formation of a Lewis acidic intermediate
silyl phosphate. During our studies on the desymmetrization of
meso-aziridines,5 we developed closely related processes based on
view of the similarity of the reaction with Me3SiN3, we believed
that the mechanism of Antilla could be involved with different
silylated nucleophiles as well. An analogous role for phosphoric
acids and other Brønsted acids in the activation of silanes, with the
in situ formation of a silylated catalyst, has also been invoked by
other groups.8
Recent studies indicated that phosphoric acids obtained by
previously established synthetic procedures as well as commer-
cially available materials may contain variable amounts of metallic
phosphate impurities, mainly produced during purification on sil-
ica gel columns.9e13 However, the acid form can be regenerated by
washing with aqueous HCl. These important discoveries have shed
new light on the previously reported phosphoric acids promoted
processes, since in most of them salt-contaminated catalysts were
employed. Since metallic phosphates are known to catalyze
a number of asymmetric reactions,11,12,14e16 it is important to re-
evaluate each process in order to verify if these impurities may
affect the catalytic activity and the enantioselectivity. To this end, it
could be useful to compare the results obtained in reactions cata-
lyzed by phosphoric acids purified on silica gel with those observed
by employing the catalyst washed with HCl. In cases where metallic
phosphates did not present any appreciable catalytic activity, the
treatment with HCl resulted only in improvement of the reaction
rate, with no changes in enantioselectivity. In other cases a strong
influence of metallic impurities has been observed. For example,
6
the activation of different silylated nucleophiles, such as PhSSiMe3
and PhSeSiMe3,7 with the same phosphoric acid. High
* Tel.: þ39 089 969554; fax: þ39 089 969603; e-mail address: gdsala@unisa.it.
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2012.10.068

G. Della Sala / Tetrahedron 69 (2013) 50e56
51
Scheme 1. VAPOL phosphate catalyzed desymmetrization of activated meso-aziridines with silylated nucleophiles as described in previous reports. Proposed catalytic cycle.
Ishihara and co-workers, during re-examination of a Mannich re-
action previously developed by the Terada group,1,17 observed for
the first time a dramatic difference in the catalytic behavior of
a BINOL derived phosphoric acid purified on silica gel with that of
the compound successively washed with HCl.11 Both the chiral
phosphoric acid and its calcium phosphate salt, in fact, proved to
catalyze the reaction but with the opposite sense of asymmetric
induction. Lately, Antilla and co-authors reported marked differ-
ences in some reactions catalyzed by HCl washed chiral phosphoric
acids and separately prepared metallic phosphates.12 These find-
ings prompted the group of Terada to re-consider some phosphoric
acid catalyzed enantioselective addition to imines.13 While in aza-
FriedeleCrafts and aza-ene-type reactions the metal-free phos-
phoric acid were shown to be the active catalyst, in substitution of
reactions with protic nucleophiles, 1 purified on silica gel showed
a different catalytic behavior from the same compound succes-
sively washed with HCl.12aec
Recently, List and co-authors demonstrated that basic phos-
phate impurities in a sample of phosphoric acid TRIP caused a shift
of some of the NMR signals. We tried an analogous approach in
order to evaluate the possible presence of phosphate salts in our
sample of phosphoric acid 1. The NMR spectrum of 1 in DMSO-d6 is
reported in Fig. 1(a). A nearly indistinguishable spectrum was ob-
tained with a commercially available sample. The addition of 4 N
NaOD in D2O caused only small shifts of the doublet at
d 9.95 and of
the singlet at 7.53 (Fig. 1(c)). These minor changes were due to the
d
presence of D2O rather than some acidebase reaction. The same
changes, in fact, were observed after addition of the same volume of
pure D2O (Fig. 1(b)). On the other hand, when the solution of 1 in
DMSO-d6 was treated with 4 N DCl in D2O, a clearly different
spectrum resulted. Two especially pronounced shifts could be no-
a
-diazoacetates, a not-well determined metallic impurity proved to
be involved.
In the light of the aforementioned accounts, we felt it useful to
re-examine phosphoric acid catalyzed asymmetric reactions with
silylated nucleophiles, with a special focus on the desymmetriza-
tion of meso-aziridines. During our previous studies, in fact, we
made use of (R)-VAPOL hydrogen phosphate 1 prepared following
the reported synthetic procedure,18 or else the commercially
available catalyst. Then, we wondered as to whether the true cat-
alytic species involved in the process is the phosphoric acid itself or
some adventitious salt impurity. In this regard, it should be con-
sidered that examples of asymmetric metal phosphate salt-
catalyzed reactions through activation of silylated nucleophiles,
are known.19 Here, we describe the results of our study on the
determination of the actual composition of synthetic phosphoric
acid 1 and of the true active catalyst in the ring opening desym-
metrization of meso-aziridines with different silylated
nucleophiles.
ticed, namely an upfield shift for a doublet, from
d 9.90 to d 9.58,
and a downfield shift for a singlet, from 7.51 to 7.65 (compare (b)
d
d
and (d) in Fig. 1). These observations confirmed that the VAPOL
hydrogen phosphate purified on silica gel, as well as the
2. Results and discussion
2.1. Analysis of VAPOL hydrogen phosphate prepared by the
literature procedure
In our previous work on the desymmetrization of meso-azir-
idines6,7 we used the catalyst 1 prepared according to literature
procedures.18 We observed similar enantioselectivities for products
3, by using the commercially available catalyst. However, since
these reports included the final purification of the product on silica
gel, we envisaged the presence of significant amounts of salt im-
purities. Indeed, Antilla and co-workers recently reported that, in
Fig. 1. 1H NMR spectra in DMSO-d6 of 1 purified on silica gel: (a) without additives; (b)
after addition of D2O; (c) after addition of 4 N NaOD in D2O; (d) after addition of 4 N
DCl in D2O.

52
G. Della Sala / Tetrahedron 69 (2013) 50e56
commercially available product, exists mostly in the form of basic
phosphate salts. The same conclusion could be drawn from the 31P
NMR spectra (Fig. 2). The addition of D2O to a solution of 1 in
DMSO-d6 caused a sharpening of the 31P NMR signal and a small
downfield shift from 1.03 to 1.24 ppm (compare (a) and (b) in
Fig. 2). A further downfield shift to 1.65 ppm was observed upon
addition of 4 N NaOD in D2O (Fig. 2(c)), but a much more marked
upfield shift to ꢀ0.33 ppm, resulted upon addition of 4 N DCl in D2O
(Fig. 2(d)).
model reaction. First, we used the catalyst 1 washed with 6 N HCl,
under the usual reaction conditions previously adopted by us.6
While the catalyst purified on silica gel readily promoted the
highly enantioselective complete conversion of 2a to (1R,2R)-3a in
CCl3CH3 at room temperature (Table 2, entry 1), washing with HCl
caused the catalytic activity todecrease. The conversionproved tobe
incompleteafter4 days, and3awasrecovered in racemicform (entry
2). This finding led us to attribute to phosphate salts, rather than to
phosphoric acid, the role of effective catalyst in this process.
Nevertheless, it was still possible to speculate on the mechanism,
analogous to that proposed by Antilla, with phosphate ion, in place
of phosphoric acid, involved in the activation of silane, with the
resulting formation of a chiral Lewis acidic silyl phosphate. In order
to assess this hypothesis, we treated a solution of 1 with DIPEA prior
to the addition of reagents, to form its ammonium salt. At any rate,
only a negligible increase of conversion was accomplished after four
days and a racemate was again obtained (entry 3).
On the basis of these experiments we deduced that the presence
of some metal species is essential to achieve high levels of activity
and enantioselectivity. Then, we pursued the determination of the
metal phosphate actually responsible for our best results. The most
common alkali and alkaline earth metal salts of phosphoric acid 1
were prepared and screened against catalyst purified on silica gel.
Interestingly the various salts exhibited a completely different be-
havior, suggesting the active involvement of the metal center in the
reaction mechanism. Sodium and potassium salts Na[1] and K[1],
catalyzed the ring-opening with low levels of enantioselectivity
favoring the opposite enantiomer of 3a (1S,2S) (entries 4 and 5).
The alkaline earth metal phosphates Ca[1]2 and Mg[1]2 were more
enantioselective, affording the same enantiomer obtained with the
catalyst purified on silica gel (entries 6 and 7). The two salts showed
a complementary effectiveness, being Ca[1]2 less enantioselective
but more active compared with Mg[1]2. Unfortunately, even with
these salts, the % ee values did not reach the high level obtained
with 1 purified on silica gel. Then, we speculated that in the latter,
actually a mixture of different VAPOL phosphate species, a co-
operative effect of two or more catalysts might be working. The
presence of the calcium salt looked to be essential, otherwise the
reaction was too slow. In view of this, mixtures of two phosphate
constituents were tested. A 2:1 mixture of Ca[1]2 and the phos-
phoric acid 1 washed with HCl, afforded the product with a lower %
ee (entry 8). To our delight a 1:1 mixture of Ca[1]2 and Mg[1]2
worked much better than single salts, furnishing a reaction rate and
an enantioselectivity very close to those observed with 1 purified
on silica gel (entry 9). Apparently, the properties of the two salts
complement each other, so as to maximize the catalytic activity and
the stereoselectivity. We performed analogous experiments with
the five-membered meso-aziridine 2b. The ring-opening desym-
metrization promoted by Ca[1]2 was incomplete after 23 h leading
to 3b with poor enantioselectivity (entry 10). The mixture Ca[1]2/
Mg[1]2 1:1, conversely, led to the same product in 4 h and with very
high levels of enantioselectivity (entry 11), as previously reported
by us with the catalyst purified on silica gel.6
Fig. 2. 31P NMR spectra in DMSO-d6 of 1 purified on silica gel: (a) without additives;
(b) after addition of D2O; (c) after addition of 4 N NaOD in D2O; (d) after addition of 4 N
DCl in D2O.
As reported in the literature,9e13 the pure acidic form of 1, could
be easily regenerated by washing with 6 M HCl.20 As expected, the
elution through a silica gel column led to the formation of a mate-
rial displaying a 1H NMR spectrum identical to that of the synthetic
compound 1 (Fig. 1(a)). No changes in the 1H NMR spectrum were
noticed after elution through a metal-free silica gel column.21
The relative contents of the various metal elements in a sample
of phosphoric acid 1 purified on silica gel, were determined by
Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-
OES). As shown in Table 1, the most abundant metal elements
proved to be calcium and magnesium. Much lower levels of sodium
were detected, although this element was previously found to be
easily captured by purification of phosphoric acids on silica gel.
Only traces of other elements could be detected. From the collected
data, compound 1 purified on silica gel could be roughly estimated
to exist at least in 80% of its amount in the form of metal salts.
Table 1
Trace metal analysis of 1 purified on silica gel, by ICP-OES
Metal element
Na
K
Ca
Mg
Cd
Pb
Contenta,b
0.288
0.0069
19.36
8.70
0.00078
0.0109
We next diverted our attention to the efficiency of the phos-
phate catalysts with different silylated nucleophiles (Table 3). The
desymmetrization of 2a with Me3SiSBn and Me3SiSMe in CCl3CH3
at room temperature, promoted by Ca[1]2/Mg[1]2 1:1, proceeded
with good enantioselectivity (entries 1 and 2), even if the latter
reagent gave better results when used in larger excess. These re-
sults are especially noteworthy since low enantioselectivities have
so far been reported in the few examples of desymmetrization of
meso-aziridines with benzylthiols and alkylthiols.22 Moreover, ac-
a
Values are in ppm.
Cr, Fe, Al, Cu, Ni, As, Mn were not detected.
b
2.2. Determination of the true catalyst in the desymmetri-
zation of meso-aziridines with different silylated nucleophiles
After having established that VAPOL hydrogen phosphate 1
synthesized by literature procedures consists in a mixture of phos-
phoric acid and metal salts, the next step was to determine, which is
the active species in the desymmetrization of meso-aziridines with
silylated nucleophiles. Forthis purpose, thedesymmetrization of the
cyclic six-membered aziridine 2a with Me3SiSPh was chosen as the
ylated
thetic point of view, because the removal of the benzyl group,
furnished useful
-aminothiols.23 Me3SiNCS proved to be a re-
markably more active silane. Under the usual conditions, the ring-
b-amino(benzyl)sulfides are very interesting from a syn-
b

G. Della Sala / Tetrahedron 69 (2013) 50e56
53
Table 2
Comparison of the efficiency of different VAPOL phosphates in the desymmetrization of aziridines 2a,b with Me3SiSPha
Entry
Catalyst
Substrate 2
Time/h
Yieldb/%
eec/% (absolute configuration)
1
2
3
4
5
6
7
8
1 purified on silica gel
1 washed with HCl
1 washed with HCl, DIPEA
Na[1]
K[1]
Ca[1]2
2a
2a
2a
2a
2a
2a
2a
2a
2a
2b
2b
2
96
96
12
3
1
24
3
100
72
77
92 (1R,2R)
0
0
95
40 (1S,2S)
7 (1S,2S)
72 (1R,2R)
80 (1R,2R)
56 (1R,2R)
90 (1R,2R)
37 (1R,2R)
92 (1R,2R)
100
100
97
90
94
Mg[1]2
Ca[1]2/1 2:1
Ca[1]2/Mg[1]2 1:1
Ca[1]2
9
10
11
3
23
4
84
98
Ca[1]2/Mg[1]2 1:1
a
Reactions performed at 0.08 mmol scale using 2 (1 equiv), Me3SiSPh (1.5 equiv), and catalyst (0.10 equiv for 1, Na[1] and K[1]; 0.05 equiv for metal phosphate salts
Ca[1]2 and Mg[1]2) in CCl3CH3 (0.1 M concentration), at room temperature, unless otherwise stated.
b
Isolated yields.
Determined by chiral HPLC.
c
Table 3
Desymmetrization of aziridines 2a,c catalyzed by VAPOL phosphate salts with different silylated nucleophilesa
Entry
Nucleophile
Catalyst
Substrate 2
Product 3
Solvent
Time/h
Yieldb/%
eec/%
1
Me3SiSBn
Me3SiSMe
Me3SiNCS
Me3SiSePh/PhSeH
Me3SiSePh/PhSeH
Me3SiSePh/PhSeH
Me3SiN3
Me3SiN3
Me3SiN3
Ca[1]2/Mg[1]2 1:1
Ca[1]2/Mg[1]2 1:1
Ca[1]2/Mg[1]2 1:1
1 purified on silica gel
1 washed with HCl
Ca[1]2/Mg[1]2 1:1
1 purified on silica gel
1 washed with HCl
Ca[1]2/Mg[1]2 1:1
2a
2a
2a
2a
2a
2a
2c
2c
2c
3c
3d
3e
3f
3f
3f
3g
3g
3g
CCl3CH3
CCl3CH3
CCl3CH3
toluene
toluene
toluene
DCE
24
48
8
0.5
96
0.5
21
72
24
53
67
100
93
35
90
97
49
94
85
82
42
92
0
96
95
6
2d
3e
4f
5f
6f
7g,h
8g
9g
DCE
DCE
91
a
Reactions performed at 0.08 mmol scale using catalyst (0.10 equiv for 1; 0.05 equiv for metal phosphate salts), 2a,c (1.0 equiv) and silylated nucleophile (1.5 equiv) in the
appropriate solvent (0.1 M concentration) at room temperature, unless otherwise stated.
b
Isolated yields.
Determined by chiral HPLC.
c
d
Me3SiSMe (5.0 equiv) was used.
e
Reaction performed at ꢀ20 ꢁC.
f
Me3SiSePh (0.5 equiv) and PhSeH (1.0 equiv) were used.
g
Compound 2c (1.5 equiv) and Me3SiN3 (1.0 equiv) were used.
Ref. 4.
h
opening catalyzed with Ca[1]2/Mg[1]2 1:1 was very fast, and low-
ering of the temperature to ꢀ20 ꢁC was found to be useful. The
Based on the results described above, we intended to re-
evaluate the desymmetrization of meso-aziridines with silylated
selenium nucleophiles previously described by our group.7 Ring-
opening of 2a with a 1:2 mixture of Me3SiSePh and PhSeH in tol-
uene at room temperature, catalyzed by 1 purified on silica gel, was
accomplished in 30 min with 92% ee (entry 4). As expected, the
outcome changed dramatically when 1 washed with HCl was
employed. Indeed, the reaction proved to be very slow, and was far
from being complete after four days. Many byproducts were
nucleophile reacted cleanly at the sulfur atom giving the b-ami-
nothiocyanate 3e in high yield and moderate enantioselectivity
(entry 3). The result obtained, even if not fully satisfactory, is worth
considering as the aziridine opening with thiocyanate nucleophiles
has been extensively described,24 but the only reported attempt of
meso-aziridine desymmetrization promoted by a chiral catalyst
afforded a racemic product.25

54
G. Della Sala / Tetrahedron 69 (2013) 50e56
formed, and product 3a was recovered as a racemate in poor yield
(entry 5). Fortunately, as observed with sulfur nucleophiles, the
former high enantioselectivity was reproduced by using the mix-
ture of phosphate salts Ca[1]2/Mg[1]2 1:1 (entry 6).
To complete our study we became interested in re-examining
the highly enantioselective desymmetrization reaction with
Me3SiN3 previously reported by Antilla and co-workers.4 In fact,
although these authors did not clearly state if the phosphoric acid 1
had been purified or not on silica gel, we hypothesized, given the
close similarities with the process developed by us, that the catalyst
used might contain phosphate salts impurities as well. In our
hands, ring-opening of the aziridine 2c, promoted by 1 washed
with HCl, in 1,2-dichloroethane (DCE) at room temperature, oc-
curred in a strikingly different way from that described. A low
conversion was achieved after long reaction time, many byproducts
were formed, and the product 3g was recovered in a nearly racemic
form (cfr. entries 7 and 8). Once more, we were able to achieve high
enantioselectivities and a clean transformation, by employing the
mixture of catalysts Ca[1]2/Mg[1]2 1:1 (cfr. entries 7 and 9).
The data collected during our study demonstrated that phos-
phoric acids are not able to promote efficiently the ring-opening
desymmetrization of meso-aziridines with silylated nucleophiles.
The catalytic cycle proposed by the group of Antilla and later
quoted by us (Scheme 1), must be ruled out. The actual mechanism
should take into account that calcium and magnesium phosphate
salts are indeed the catalytic active species. Although further in-
vestigations are necessary to propose an exact mechanism, the
Lewis acidity of the metal center and the Lewis basicity of the
phosphate ligands are likely to be involved. In this regard, Ishihara
and co-authors, postulated the Lewis-base catalytic ability of lith-
ium phosphate salts is in the cyanosilylation of ketones with
Me3SiCN.19b Along these lines we suggest the catalytic cycle
depicted in Scheme 2 as a possible working hypothesis.
Recently, earth alkaline metal bis(phosphate) salts have been
found to exist mainly as dimers or more complex oligomers, in
equilibrium with the catalytic active monomer.11,16b The reaction of
monomer with silane, is supposed to force the equilibrium away
from the oligomers. We, in fact, observed that the sparingly soluble
phosphate salts were dissolved completely in CCl3CH3 after addi-
tion of silanes. The metal phosphate/Me3SiX activated complex
presents two Lewis acidic sites on metal and silicon atoms. The N-
acyl aziridine 2 might be activated by coordination to one of these
Lewis acidic centers (alternative transition states B and C, Scheme
1) or even to both of them (transition state A), acting as a biden-
tate ligand. Then, the thiolate may conduct the nucleophilic attack
either from inside or outside the coordination sphere of the metal.
At any rate, the involvement of a hypervalent silicon intermediate
cannot be ruled out (alternative transition state C). Finally, a sily-
lation step has to occur, accordingly to the isolation of 4 reported by
Antilla and co-workers.4
The rough mechanistic model outlined above, does not clarify the
higher performance of the Ca/Mg combined catalyst compared to
the single salts. Nevertheless, one may speculate that both metals
play a key role in the catalytic cycle, whether as part of a bimetallic
species, or as distinct monometallic molecules; one metal center
might act as a Lewis acid, activating the aziridine, while the ligand
attached to the other generates the reactive nucleophile. In this
regard, some examples of aziridine desymmetrizations with sily-
lated nucleophiles, promoted by bimetallic complexes, are known.26
Scheme 2. Proposed catalytic cycle, and alternative transition state models, for the
earth alkaline metal VAPOL phosphate catalyzed desymmetrization of meso-aziridines.
by us and others can actually be ascribable to the combined action of
calcium and magnesium phosphate salts, present as adventitious
impurities in the phosphoric acid purified on silica gel. Reproducible
results could be obtained by using a 1:1 mixture of calcium and
magnesium VAPOL phosphate salts as the catalyst. This process has
been extended toa broad rangeof silylated nucleophiles. We suggest
metal phosphates might operate through a dual Lewis acidebase
activation mechanism. We believe that our findings may offer
valuable clues in the study of processes with silylated nucleophiles
promoted by phosphoric acid and other Brønsted acids. Specific
experiments will be required, addressed to confirm the capacity of
these compounds in activating the nucleophile, and precluding
significant involvement of metal salt impurities. Further studies
along these lines are currently under way in our laboratory.
3. Conclusions
4. Experimental section
4.1. General information
In summary, we have clearly demonstrated that metal-free
phosphoric acid 1 is not able to promote efficiently the ring-
opening desymmetrization of meso-aziridines with silylated nu-
cleophiles. The activity and enantioselectivity previously observed
All the aziridine desymmetrizations were performed in flame-
dried Schlenk tubes, under a dry nitrogen atmosphere and with

G. Della Sala / Tetrahedron 69 (2013) 50e56
55
magnetic stirring. Reactions were monitored by thin layer chro-
matography (TLC) on Merck silica gel plates (0.25 mm) and visu-
alized by UV light or by phosphomolybdic acid/ethanol spray test.
Flash chromatography was performed on Merck silica gel (60,
particle size: 0.040e0.063 mm). 1H NMR, 13C NMR, and 31P NMR
spectra were recorded on Bruker DRX 400 and 300 spectrometers
at room temperature in CDCl3, DMSO-d6 or acetone-d6 as solvents.
Chemical shifts for protons are reported using residual CHCl3
mixture was stirred for 10 min and then the aziridine 2aec was
added (0.080e0.120 mmol). After stirring for the appropriate time
(see Tables 2 and 3), CH2Cl2 was added until dissolution of the
precipitate, and then the solution was passed through a plug of
silica gel eluting with CH2Cl2 (3 mL) and then ethyl acetate (3 mL).
The solvent was removed under reduced pressure and the residue
was purified by FC (silica gel, eluent: CHCl3 for 3a,b; petroleum
ether/ethyl acetate mixtures from 95:5 to 70:30 for 3c,d,g; CH2Cl2/
acetone mixtures from 100:0 to 98:2 for 3e). The characterization
data of the opening products 3a,b6 and 3g4 were identical to those
previously reported and the enantiomeric purity was determined
by chiral HPLC as therein described.
(
d
¼7.26), CD3SOCHD2 (
d
¼2.50) or CD3COCHD2 ( ¼2.04) as internal
d
references. Carbon spectra are referenced to the shift of the 13C
signal of CDCl3
(d
¼77.0) or acetone-d6
(
d
¼206.7). 13P NMR spectra
were recorded using H3PO4 as an external standard. Optical rota-
tion measurements were performed on a Jasco DIP-1000 digital
polarimeter using the Na lamp. FT-IR spectra were recorded on
a Bruker Vertex 70 spectrophotometer. ESI-MS was performed us-
ing a Bio-Q triple quadrupole mass spectrometer (Micromass,
Manchester, UK) equipped with an electrospray ion source. Optical
Emission Spectrometry (ICP-OES) analysis was performed using an
Optima 7000DV (PerkineElmer) instrument. Elemental analyses
were carried out by using Flash EA 1112 (Thermo Electron Corpo-
ration) analyzer. Enantiomeric excesses were measured by HPLC
(Jasco PU-2089 quaternary gradient pump equipped with an MD-
2010 plus adsorbance detector) using Daicel Chiralcel AD-H and
AS-H chiral columns.
4.5. Spectral data of products 3cee
4.5.1. (1R,2R)-1-Benzylthio-2-[N-(3,5-dinitrobenzoyl)amino] cyclo-
hexane (3c). White solid; mp 152e157 ꢁC (dec); [
a
]
20 ꢀ40.4 (c 0.8,
D
CHCl3, 85% ee); 1H NMR (400 MHz, CDCl3):
d
9.14 (t, J¼1.9 Hz, 1H),
8.84 (d, J¼1.9 Hz, 2H), 7.33e7.22 (m, 4H), 7.19 (m, 1H), 6.22 (d,
J¼6.8 Hz, 1H), 3.87 (m, 1H), 3.81e3.66 (m, 2H), 2.52 (dt, J¼11.2,
3.4 Hz, 1H), 2.40e2.19 (m, 21H), 1.91e1.72 (m, 2H), 1.64 (m, 1H), 1.45
(m, 1H), 1.37e1.17 (m, 2H); 13C NMR (100 MHz, CDCl3):
d
162.4,
148.6, 138.6, 138.3, 128.8, 128.7, 127.3, 127.2, 120.9, 52.9, 48.4, 33.8,
33.6, 33.4, 26.0, 24.6; FT-IR (KBr): n
/cmꢀ1: 3272, 3106, 3086, 2922,
All the solvents used in moisture sensitive reactions were dis-
tilled over calcium hydride and stored over activated molecular
sieves, except for anhydrous 1,2-dichloroethane, which was pur-
chased from Aldrich. (R)-VAPOL hydrogen phosphate 1 purified on
silica gel,18 1 washed with HCl,12aec as well as salts Na[1], K[1], Ca
[1]2, and Mg[1]2,12aec were prepared as described in literature. The
2855, 1648, 1539, 1343, 915, 730, 706; MS (ESI): m/z: 454 [MþK]þ;
elemental analysis calcd (%) for C20H21N3O5S: C 57.82, H 5.09, N
10.11, S 7.72; found: C 58.12, H 5.37, N 9.97, S 7.63; HPLC (Chiralcel
AD-H, hexane/iPrOH 80:20,
tminor¼11.6 min, tmajor¼13.1 min.
l
¼220 nm, 1.0 mL minꢀ1):
nucleophiles
(phenylthio)trimethylsilane,
(methylthio)trime-
4.5.2. (1R,2R)-1-Methylthio-2-[N-(3,5-dinitrobenzoyl)amino] cyclo-
thylsilane, trimethylsilyl isothiocyanate, and azidotrimethylsilane,
were purchased from Aldrich and used without further purifica-
tion. (Benzylthio)trimethylsilane27 and (phenylseleno)trimethylsi-
lane28 were prepared according to literature procedures. The
acylaziridines 2aec were synthesized as described in literature, by
opening of epoxides with sodium azide, followed by Staudinger
rearrangement of azido alcohols and then in situ acylation of
unsubstituted aziridines with aroyl chlorides and triethyl amine.4,29
hexane (3d). White solid; mp 173e175 ꢁC; [
a
]
20 ꢀ80.2 (c 0.8, CHCl3,
D
82% ee); 1H NMR (400 MHz, CDCl3):
d
9.16 (t, J¼2.1 Hz, 1H), 8.96 (d,
J¼2.1 Hz, 2H), 6.54 (d, J¼6.9 Hz, 1H), 3.92 (m, 1H), 2.57 (dt, J¼11.2,
3.7 Hz, 1H), 2.39 (m, 1H), 2.20 (m, 1H), 2.06 (s, 3H), 1.91e1.75 (m,
2H), 1.70e1.55 (m, 1H), 1.54e1.17 (m, 3H); 13C NMR (100 MHz,
CDCl3):
24.7; FT-IR (KBr):
d 162.4, 148.6, 138.3, 127.2, 121.0, 52.4, 48.9, 33.5, 32.4, 26.0,
n
/cmꢀ1: 3320, 3097, 2930, 2858, 1644, 1543, 1341,
1076, 917, 730, 722; MS (ESI): m/z: 378 [MþK]þ; elemental analysis
calcd (%) for C14H17N3O5S: C 49.55, H 5.05, N 12.38, S 9.45; found: C
49.81, H 5.24, N 12.29, S 9.54; HPLC (Chiralcel AD-H, hexane/iPrOH
4.2. NMR analysis of VAPOL hydrogen phosphate purified on
silica gel, after treatment with acid or base (Figs. 1 and 2)
90:10,
l
¼220 nm, 1.0 mL minꢀ1): tmajor¼17.3 min, tminor¼20.1 min.
The sample for the reference 1H and 31P NMR spectra was pre-
pared by dissolving 1 purified on silica gel (5 mg) in DMSO-d6
(0.5 mL). The other spectra were recorded after treatment of this
4.5.3. (1R,2R)-3,5-Dinitro-N-(2-thiocyanatocyclohexyl) benzamide
(3e). White solid; mp 167e168; [
a
]
20 ꢀ82.4 (c 0.6, acetone, 42% ee);
D
1H NMR (400 MHz, acetone-d6):
d 9.08 (app. s, 3H), 8.66 (d,
sample with D2O, 4 N NaOD in D2O and 4 N DCl in D2O (40
respectively.
m
L),
J¼7.8 Hz, 1H), 4.18 (m, 1H), 3.54 (dt, J¼11.5, 4.0 Hz, 1H), 2.37 (m, 1H),
2.18 (m, 1H), 1.93e1.75 (m, 3H), 1.74e1.61 (m, 1H), 1.60e1.39 (m,
2H); 13C NMR (100 MHz, acetone-d6):
122.3, 112.0, 54.7, 53.9, 35.0, 34.5, 27.1, 25.8; FT-IR (KBr): :
d
163.5, 150.1, 138.8, 128.8,
4.3. Preparation of the sample for trace element analysis
(ICP-OES)
n
/cmꢀ1
3379, 3333, 3110, 3092, 2944, 2925, 2859, 2153, 1662, 1592, 1345,
1074, 916, 730, 721; MS (ESI): m/z: 373 [MþNa]þ; elemental anal-
ysis calcd (%) for C14H14N4O5S: C 47.99, H 4.03, N 15.99, S 9.15;
found: C 48.05, H 4.14, N 15.91, S 9.06; HPLC (Chiralcel AD-H,
A solution of VAPOL hydrogen phosphate 1 purified on silica gel
(21.5 mg) in CH2Cl2 (1.0 mL) was treated with 3.7 M aqueous HNO3
(10.0 mL), and the mixture was stirred overnight. An aliquot of the
aqueous phase (2.45 mL) was removed and diluted with deionized
H2O (up to 5.0 mL). The resulting solution was submitted to ICP-OES
analysis.
hexane/iPrOH 90:10,
tminor¼31.1 min.
l
¼220 nm, 1.0 mL minꢀ1): tmajor¼23.8 min,
4.6. Desymmetrization of meso-aziridine 2a with Me3SiSePh/
PhSeH catalyzed by Ca[1]2/Mg[1]2 1:1
4.4. General procedure for the desymmetrization of meso-
aziridines 2aec with silylated nucleophiles catalyzed by Ca
[1]2/Mg[1]2 1:1
To a suspension of phosphates Ca[1]2 (2.5 mg, 0.0020 mmol)
and Mg[1]2 (2.5 mg, 0.0020 mmol) in anhydrous toluene (0.8 mL),
Me3SiSePh was added (9.1 mg, 0.040 mmol). The mixture was
stirred for 10 min, after which the solid dissolved. Then PhSeH
(12.7 mg, 0.080 mmol) and the aziridine 2a (23.2 mg, 0.080 mmol)
were added. After stirring for 30 min, CH2Cl2 was added until
To a solution of phosphates Ca[1]2 (2.5 mg, 0.0020 mmol) and
Mg[1]2 (2.5 mg, 0.0020 mmol) in the appropriate anhydrous sol-
vent (0.8 mL), the silane was added (0.080e0.40 mmol). The

56
G. Della Sala / Tetrahedron 69 (2013) 50e56
H. L.; Antilla, J. C. Org. Lett. 2011,13, 2188e2191; (d) Ingle, G. K.; Liang, Y.; Mormino,
M.; Li, G.; Fronczek, F. R.; Antilla, J. C. Org. Lett. 2011, 13, 2054e2057.
13. Terada, M.; Kanomata, K. Synlett 2011, 1255e1258.
dissolution of the precipitate, and then the solution was passed
through a plug of silica gel eluting with CH2Cl2 (3 mL) and then
ethyl acetate (3 mL). The solvent was removed under reduced
pressure and the residue was purified by FC (silica gel, eluent:
CHCl3). The characterization data of the opening product 3f were
identical to those previously reported and the enantiomeric purity
was determined by chiral HPLC as therein described.7
14. For a specific review on metal phosphates as enantioselective catalysts see:
ꢁ
Parra, A.; Reboredo, S.; Martín Castroa, A. M.; Aleman, J. Org. Biomol. Chem.
2012, 10, 5001e5020.
15. For reviews including parts on the combination of phosphoric acids with
metals in asymmetric catalysis see: (a) Rueping, M.; Koenigs, R. M.; Atodiresei,
I. Chem.dEur. J. 2010, 16, 9350e9365; (b) Hatano, M.; Ishihara, K. Synthesis
2010, 3785e3801; (c) Zhong, C.; Shi, X. Eur. J. Org. Chem. 2010, 2999e3025; (d)
Shao, Z.; Zhang, H. Chem. Soc. Rev. 2009, 38, 2745e2755; (e) Hashmi, A. S. K.;
Hubbert, C. Angew. Chem., Int. Ed. 2010, 49, 1010e1012; (f) de Armas, P.; Tejedor,
D.; García-Tellado, F. Angew. Chem., Int. Ed. 2010, 49, 1013e1016.
Acknowledgements
16. For additional recent reports on metal phosphates as asymmetric catalysts: (a)
Alix, A.; Lalli, C.; Retailleau, P.; Masson, G. J. Am. Chem. Soc. 2012, 134,
10389e10392; (b) Drouet, F.; Lalli, C.; Liu, H.; Masson, G.; Zhu, J. Org. Lett. 2011,
13, 94e97; (c) Rueping, M.; Bootwicha, T.; Sugiono, E. Synlett 2011, 323e326;
(d) Lv, J.; Li, X.; Zhong, L.; Luo, S.; Cheng, J. P. Org. Lett. 2010, 12, 1096e1099; (e)
Shi, S. H.; Huang, F. P.; Zhu, P.; Dong, Z. W.; Hui, X. P. Org. Lett. 2012, 14,
2010e2013; (f) Han, Z. Y.; Guo, R.; Wang, P. S.; Chen, D. F.; Xiao, H.; Gong, L.-Z.
ꢀ
The author thanks M.I.U.R. (Ministero Italiano Universita e
Ricerca) for financial support. Mr. Stefano Tommasone and Mr.
Aniello Cioffi for their valuable experimental work, and Dr. Patrizia
Iannece for ESI-MS spectra and for elemental and trace element
(ICP-OES) analyses, are gratefully acknowledged.
ꢁ
Tetrahedron Lett. 2011, 52, 5963e5967; (g) Barbazanges, M.; Auge, M.; Moussa,
J.; Amouri, H.; Aubert, C.; Desmarets, C.; Fensterbank, L.; Gandon, V.; Malacria,
M.; Ollivier, C. Chem.dEur. J. 2011, 17, 13789e13794; (h) Jiang, G.; Halder, R.;
Fang, Y.; List, B. Angew. Chem., Int. Ed. 2011, 50, 9752e9755; Yue, T.; Wang, M. X.;
Wang, D. X.; Masson, G.; Zhu, J. J. Org. Chem. 2009, 74, 8396e8399.
17. Gridnev, I. D.; Kouchi, M.; Sorimachi, K.; Terada, M. Tetrahedron Lett. 2007, 48,
497e500.
Supplementary data
Copies of 1H NMR and 13C NMR spectra for the unreported
products. Supplementary data related to this article can be found
online at http://dx.doi.org/10.1016/j.tet.2012.10.068.
18. (a) Desai, A. A.; Huang, L.; Wulff, W. D.; Rowland, G. B.; Antilla, J. C. Synthesis
2010, 2106e2109; (b) Rowland, G. B.; Zhang, H.; Rowland, E. B.; Chenna-
madhavuni, S.; Wang, Y.; Antilla, J. C. J. Am. Chem. Soc. 2005, 127, 15696e15697.
19. (a) Shen, K.; Liu, X.; Cai, Y.; Lin, L.; Feng, X. Chem.dEur. J. 2009, 15, 6008e6014;
(b) Hatano, M.; Ikeno, T.; Matsumura, T.; Torii, S.; Ishihara, K. Adv. Synth. Catal.
2008, 350, 1776e1780.
References and notes
1. Uraguchi, D.; Terada, M. J. Am. Chem. Soc. 2004, 126, 5356e5357.
2. Akiyama, T.; Itoh, J.; Yokota, K.; Fuchibe, K. Angew. Chem., Int. Ed. 2004, 43,
1566e1568.
20. The diagnostic 1H NMR signals for the HCl washed catalyst 1 in DMSO-d6 re-
sulted to be:
d 9.73 (d) and d 7.65 (s).
3. For recent reviews see: (a) Rueping, A.; Kuenkel, A.; Atodiresei, I. Chem. Soc. Rev.
2011, 40, 4539e4549; (b) Terada, M. Curr. Org. Chem. 2011, 15, 2227e2256; (c)
Terada, M. Synthesis 2010, 1929e1982; (d) Zamfir, A.; Schenker, S.; Freund, M.;
Tsogoeva, S. B. Org. Biomol. Chem. 2010, 8, 5262e5276; (e) Terada, M. Chem.
Commun. 2008, 4097e4112.
21. Metal-free silica gel was prepared as described in literature: Youard, Z. A.;
Mislin, G. L. A.; Majcherczyk, P. A.; Schalk, I. J.; Reimmann, C. J. Biol. Chem. 2007,
282, 35546e35553.
22. (a) Larson, S. E.; Baso, J. C.; Li, G.; Antilla, J. C. Org. Lett. 2009, 11, 5186e5189; (b)
Hayashi, M.; Ono, K.; Oshimi, H.; Oguni, N. Tetrahedron 1996, 52, 7817e7832.
23. (a) Enders, D.; Moll, A. Synthesis 2005, 1807e1816; (b) Zhu, Y.; Gieselman, M. D.;
Zhou, H.; Averin, O.; van der Donk, W. A. Org. Biomol. Chem. 2003, 1, 3304e3315;
(c) Nagasawa, H. T.; Elberling, J. E.; Roberts, J. C. J. Med. Chem.1987, 30,1373e1378.
24. (a) Das, B.; Ramu, R.; Ravikanth, B.; Reddy, K. R. Tetrahedron Lett. 2006, 47,
779e782; (b) Das, B.; Reddy, V. S.; Ramu, R.; Kumar, D. N. Catal. Commun. 2006,
7, 997e999; (c) Reddy, M. S.; Narender, M.; Nageswar, Y. V. D.; Rao, K. R. Tet-
rahedron Lett. 2005, 46, 6437e6439; (d) Yadav, J. S.; Reddy, B. V. S.; Jyothirmai,
B.; Murty, M. S. R. Tetrahedron Lett. 2005, 46, 6385e6387.
25. Prusinowska, N.; Gawronski, J. Synth. Commun. 2009, 39, 2795e2803.
26. (a) Mita, T.; Fujimori, I.; Wada, R.; Wen, J.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc.
2005,127,11252e11253; (b) Fujimori, I.; Mita, T.; Maki, K.; Shiro,M.; Sato, A.; Furusho,
S.; Kanai, M.; Shibasaki, M. Tetrahedron 2007, 63, 5820e5831; (c) Wu, B.; Gallucci, J.
C.; Parquette, J. R.; RajanBabu, T. V. Angew. Chem., Int. Ed. 2009, 48, 1126e1129.
27. Mai, K.; Patil, G. J. Org. Chem. 1986, 51, 3545e3548.
4. Rowland, E. B.; Rowland, G. B.; Rivera-Otero, E.; Antilla, J. C. J. Am. Chem. Soc.
2007, 129, 12084e12085.
5. Lattanzi, A.; Della Sala, G. Eur. J. Org. Chem. 2009, 1845e1848.
6. Della Sala, G.; Lattanzi, A. Org. Lett. 2009, 11, 3330e3333.
7. Senatore, M.; Lattanzi, A.; Santoro, S.; Santi, C.; Della Sala, G. Org. Biomol. Chem.
2011, 9, 6205e6207.
8. See for example: (a) Zamfir, A.; Tsogoeva, S. B. Org. Lett. 2009, 12, 188e191; (b)
García-García, P.; Lay, F.; García-García, P.; Rabalakos, C.; List, B. Angew. Chem.,
Int. Ed. 2009, 48, 4363e4366.
9. Xu, S.; Wang, Z.; Zhang, X.; Zhang, X.; Ding, K. Angew. Chem., Int. Ed. 2008, 47,
2840e2843.
10. Klussmann, M.; Ratjen, L.; Hoffmann, S.; Wakchaure, V.; Goddard, R.; List, B.
Synlett 2010, 2189e2192.
11. Hatano, M.; Moriyama, K.; Maki, T.; Ishihara, K. Angew. Chem., Int. Ed. 2010, 49,
3823e3826.
28. (a) Miyoshi, N.; Ishii, H.; Kondo, K.; Murai, S.; Sonoda, N. Synthesis 1979,
300e301; (b) Detty, M. R. J. Org. Chem. 1979, 44, 4528e4531.
29. Fukuta, Y.; Mita, T.; Fukuda, N.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2006,
128, 6312e6313.
12. (a) Zhang, Z.; Zheng, W.; Antilla, J. C. Angew. Chem., Int. Ed. 2011, 50,1135e1138; (b)
Zheng, W.; Zhang, Z.; Kaplan, M. J.; Antilla, J. C. J. Am. Chem. Soc. 2011, 133,
3339e3341; (c) Larson, S. E.; Li, G.;Rowland, G. B.; Junge, D.;Huang, R.;Woodcock,