Catalytic enantioselective [3 + 2]-cycloadditions of diazoketone-derived aryl-substituted carbonyl ylides

Article abstract of DOI:10.1021/jo026307t

An evaluation of α-aryl-α-diazodiones in tandem carbonyl ylide formation - enantioselective [3 + 2]-cycloaddition reactions is described. Such substrates were designed to allow investigation of the electronic characteristics of the dipole upon asymmetric induction. Intramolecular cycloadditions (with a tethered alkene dipolarophile) were found to occur in good to quantitative yields, with a difference in ee exhibited by the two electronically different diazodiones 8 and 9. Intermolecular cycloadditions using diazodiones 12 and 13 with DMAD and arylacetylenes 16-18 again demonstrated that electronics play a key role in determining the outcome of the cycloaddition reactions. Enantioselectivities of up to 76% were observed.

Full text of DOI:10.1021/jo026307t

Ca ta lytic En a n tioselective [3 + 2]-Cycloa d d ition s of
Dia zok eton e-Der ived Ar yl-Su bstitu ted Ca r bon yl Ylid es
David M. Hodgson,*,† Rebecca Glen,† Guy H. Grant,‡ and Alison J . Redgrave§
Dyson Perrins Laboratory, Department of Chemistry, University of Oxford, South Parks Road,
Oxford OX1 3QY, U.K., Physical and Theoretical Chemistry Laboratory, Department of Chemistry,
University of Oxford, South Parks Road, Oxford OX1 3QZ, U.K., and Medicinal Chemistry 1,
GlaxoSmithKline Medicines Research Centre, Gunnels Wood Road, Stevenage, SG1 2NY, U.K.
david.hodgson@chem.ox.ac.uk
Received August 12, 2002
An evaluation of R-aryl-R-diazodiones in tandem carbonyl ylide formation-enantioselective [3 + 2]-
cycloaddition reactions is described. Such substrates were designed to allow investigation of the
electronic characteristics of the dipole upon asymmetric induction. Intramolecular cycloadditions
(with a tethered alkene dipolarophile) were found to occur in good to quantitative yields, with a
difference in ee exhibited by the two electronically different diazodiones 8 and 9. Intermolecular
cycloadditions using diazodiones 12 and 13 with DMAD and arylacetylenes 16-18 again
demonstrated that electronics play a key role in determining the outcome of the cycloaddition
reactions. Enantioselectivities of up to 76% were observed.
SCHEME 1
An attractive approach to reduced furans (and other
five-membered oxygen-containing heterocycles) is 1,3-
dipolar cycloaddition of a π-bond with a carbonyl ylide.1
Intramolecular carbonyl ylide formation, usually by
metal-catalyzed decomposition of a diazocarbonyl com-
pound (e.g. 1), followed by cycloaddition, offers the
opportunity to control up to four stereocenters. Despite
the obvious value of this class of 1,3-dipolar cycloaddi-
tions to natural product synthesis2 (e.g. zaragozic acid3
and cis-nemorensic acids4), only recently have examples
been reported of enantioselective tandem carbonyl ylide
formation-cycloaddition.5 In contrast to other metal-
catalyzed transformations of diazocarbonyl compounds,
e.g. enantioselective insertions and cyclopropanations,
† Dyson Perrins Laboratory, University of Oxford.
‡ Physical and Theoretical Chemistry Laboratory, University of
Oxford.
where it is evident that an intermediate metal-complexed
carbene can directly exert an influence on selectivity, the
situation with potentially asymmetric ylide transforma-
tions is more complex.5a For instance, it could be argued
that once an ylide is formed, the catalyst is not involved
in the subsequent cycloaddition, and asymmetric induc-
tion would be unlikely. Alternatively, if a chiral, non-
racemic catalyst were to remain associated (e.g. as shown
in ylide 2, Scheme 1), then there is the possibility of
achieving an enantioselective cycloaddition reaction. We
reported the first example of enantioselective tandem
carbonyl ylide formation-cycloaddition in 1997, for the
intramolecular process shown in Scheme 1.6 Enantio-
meric excesses of up to 52% were achieved using Davies’
prolinate catalyst Rh2(S-DOSP)4 47 (Figure 1) in hexane.
After this initial result, the binaphthyl phosphate cata-
§ GlaxoSmithKline.
(1) (a) Huisgen, R. In 1,3-Dipolar Cycloaddition Chemistry; Padwa,
A., Ed.; Wiley-Interscience: New York, 1984; pp 1-176. (b) Padwa,
A.; Weingarten, M. D. Chem. Rev. 1996, 96, 223-269. (c) Cycloaddition
Reactions in Organic Synthesis; Kobayashi, S., J ørgensen, K. A., Eds.;
Wiley-VCH: Weinheim, 2002. (d) Synthetic Applications of 1,3-Dipolar
Cycloaddition Chemistry Toward Heterocycles and Natural Products;
Padwa, A., Pearson, W. H., Eds.; J ohn Wiley & Sons: New York, 2002.
(2) (a) Gothelf, K. V.; J ørgensen, K. A. Chem. Rev. 1998, 98, 863-
909. (b) Karlsson, S.; Ho¨gberg, H.-E. Org. Prep. Proced. Int. 2001, 33,
105-172.
(3) (a) Hodgson, D. M.; Bailey, J . M.; Villalonga-Barber, C.; Drew,
M. G. B.; Harrison, T. J . Chem. Soc., Perkin Trans. 1 2000, 3432-
3443. (b) Kataoka, O.; Kitagaki, S.; Watanabe, N.; Kobayashi, J .;
Nakumura, S.; Shiro, M.; Hashimoto, S. Tetrahedron Lett. 1998, 39,
2371-2374.
(4) Hodgson, D. M.; Avery, T. D.; Donohue, A. C. Org. Lett. 2002, 4,
1809-1811.
(5) (a) Hodgson, D. M.; Pierard, F. Y. T. M.; Stupple, P. A. Chem.
Soc. Rev. 2001, 30, 50-61. (b) Kitagaki, S.; Hashimoto, S. J . Synth.
Org. Chem. J pn. 2001, 59, 1157-1169. (c) Hodgson, D. M.; Stupple,
P. A.; Forbes, D. C. In Rodd’s Chemistry of Carbon Compounds, Topical
Volume, Asymmetric Catalysis; Sainsbury, M., Ed.; Elsevier: Oxford,
2001; pp 65-99. (d) Forbes, D. C.; McMills, M. C. Curr. Org. Chem.
2001, 5, 1091-1105.
(6) Hodgson, D. M.; Stupple, P. A.; J ohnstone, C. Tetrahedron Lett.
1997, 38, 6471-6472.
(7) (a) Davies, H. M. L.; Bruzinski, P. R.; Lake, D. H.; Kong, N.;
Fall, M. J . J . Am. Chem. Soc. 1996, 118, 6897-6907. (b) Davies, H.
M. L. Eur. J . Org. Chem. 1999, 2459-2469.
10.1021/jo026307t CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/24/2002
J . Org. Chem. 2003, 68, 581-586
581

Hodgson et al.
F IGURE 1.
lysts Rh2(R-BNP)4 58 and Rh2(R-DDBNP)4 6 (Figure 1)
were found capable of delivering improved ee values of
up to 90% for cycloadduct 3.9
imoto et al.12a and ourselves9,13 (Scheme 2) illustrates that
the steric and/or electronic nature of the dipole has an
important bearing upon the level of asymmetry induced
and warranted further investigation. R-Aryl-R-diazodi-
ones (e.g. Schemes 3 and 4)14 were selected as test
substrates, since substituent variation within the aryl
group (e.g., p-methoxy or p-nitro groups) allows electronic
perturbations on the system to be analyzed without
concomitant alteration in the steric environment of the
dipole. However, there is only minimal literature prece-
dence for the reactions of R-aryl-R-diazoketones under
rhodium(II) catalysis.15 C-H insertion and cyclopropa-
nation reactions of such substrates have been reported
by Mateos.16 The only observation to date of enantio-
selectivity was made by McKervey et al., in which an
intramolecular C-H insertion occurred in 45% ee.17 Until
recently, there were no reported examples of tandem
carbonyl ylide type formation-cycloaddition reactions of
R-aryl-R-diazoketones. In 1999, Padwa et al. published
studies with such substrates in which cycloadditions
occurred via “aromatic” carbonyl ylides (oxidopyryli-
ums).18 Although not strictly comparable to carbonyl
ylides formed from R-aryl-R-diazodiones, this was an
encouraging precedent.
Enantioselective tandem carbonyl ylide formation-
intermolecular cycloadditions have been reported by
research groups led by Doyle,10 Ibata,11 and Hashimoto.12
The asymmetric induction in these cycloadditions was
low (<30% ee), aside from the work of Hashimoto using
R-diazoketones with DMAD as the dipolarophile, where
ee values up to 92% were reported (Scheme 2; the
absolute sense of predominant asymmetric induction was
not determined).12a
SCHEME 2
SCHEME 3
The aim of the work described herein was to develop a
better understanding of the factors affecting asymmetric
induction in this emerging asymmetric process. This
study has involved the synthesis and examination of a
new type of diazocarbonyl substrate in both intra- and
intermolecular cycloadditions, which generate cyclo-
adducts in up to 76% ee.13 To begin to elucidate the
factors influencing enantioselectivity, we focused initially
on electronic effects within the dipole. Work by Hash-
Our studies on diazodiones commenced with a simple
tethered alkene as the dipolarophile (Scheme 3), designed
to undergo intramolecular cycloaddition, analogous to our
original work on R-diazo-â-keto esters (Scheme 1).6,9
(14) See Supporting Information for details.
(15) This is in stark contrast with R-aryl-R-diazoacetates, which have
been comprehensively examined in most transformations of diazocar-
bonyl compounds, some with exceptionally high levels of enantio-
selectivity. Selected examples: (a) Doyle, M. P.; Zhou, Q. L.; Charn-
sangavej, C.; Longoria, M. A.; McKervey, M. A.; Garcia, C. F.
Tetrahedron Lett. 1996, 37, 4129-4132. (b) Davies, H. M. L.; Hansen,
T.; Churchill, M. R. J . Am. Chem. Soc. 2000, 122, 3063-3070. (c)
Kitagaki, S.; Kinoshita, M.; Takeba, M.; Anada, M.; Hashimoto, S.
Tetrahedron: Asymmetry 2000, 11, 3855-3859. (d) Doyle, M. P.;
Davies, S. B.; Hu, W. Org. Lett. 2000, 2, 1145-1147.
(8) Pirrung, M. C.; Zhang, J . Tetrahedron Lett. 1992, 33, 5987-5990.
(9) (a) Hodgson, D. M.; Stupple, P. A.; J ohnstone, C. Chem. Commun.
1999, 2185-2186. (b) Hodgson, D. M.; Stupple, P. A.; Pierard, F. Y. T.
M.; Labande, A. H.; J ohnstone, C. Chem. Eur. J . 2001, 7, 4465-4476.
(10) Doyle, M. P.; Forbes, D. C. Chem. Rev. 1998, 98, 911-935.
(11) Suga, H.; Ishida, H.; Ibata, T. Tetrahedron Lett. 1998, 39, 3165-
3166.
(12) (a) Kitagaki, S.; Anada, M.; Kataoka, O.; Matsuno, K.; Umeda,
C.; Watanabe, N.; Hashimoto, S. J . Am. Chem. Soc. 1999, 121, 1417-
1418. (b) Kitagaki, S.; Yasugahira, M.; Anada, M.; Nakajima, M.;
Hashimoto, S. Tetrahedron Lett. 2000, 41, 5931-5935.
(13) For a preliminary communication of part of this work, see:
Hodgson, D. M.; Glen, R.; Redgrave, A. J . Tetrahedron Lett. 2002, 43,
3927-3930.
(16) (a) Mateos, A. F.; Coca, G. P.; Alonso, J . J . P.; Gonza´lez, R. R.;
Herna´ndez, C. T. Synlett 1996, 1134-1136. (b) Mateos, A. F.; Alonso,
J . J . P.; Gonza´lez, R. R. Tetrahedron 1999, 55, 847-860.
(17) Ye, T.; Garc´ıa, C. F.; McKervey, M. A. J . Chem. Soc., Perkin
Trans. 1 1995, 1373-1379.
(18) Padwa, A.; Precedo, L.; Semones, M. A. J . Org. Chem. 1999,
64, 4079-4088.
582 J . Org. Chem., Vol. 68, No. 2, 2003

Catalytic Enantioselective [3 + 2]-Cycloadditions
Following confirmation of the viability of the substrates
8 and 9 to undergo the desired ylide formation-cyclo-
addition process on treatment with rhodium(II) acetate
in CH2Cl2 at room temperature (Scheme 3, Table 1,
entries 1, 13), representatives of the carboxylate and
binaphthyl phosphate classes of chiral rhodium(II) cata-
lysts previously shown to impart asymmetric induction
in cycloadditions with carbonyl ylides were screened. It
was anticipated that the presence of the electron-
withdrawing nitro group on diazodione 9 might lead to
higher enantioselectivity in the cycloaddition reaction
compared with diazodione 8, since the former would
resemble more electronically the successful ester-substi-
tuted diazodione 1 (Scheme 1).
the yield almost doubled to 81% (Table 1, entry 5);
however, there was a slight erosion in ee to 19%. With
diazodione 9, cycloadduct (+)-11 was produced with
essentially identical enantiomeric excess (20%, Table 1,
entry 15). A switch to Rh2(R-DDBNP)4 6 under the same
conditions demonstrated that the two cycloadducts could
form with different levels of asymmetric induction using
the same catalyst; phenyldiazodione 8 delivered cyclo-
adduct in 19% ee, compared with 28% ee from nitro-
phenyldiazodione 9 (Table 1, entries 7, 16). As catalyst
6 was designed for its solubility in hydrocarbon solvents,
ambient temperature could also be used. The difference
in cycloadduct ee was now more pronounced: 35% ee for
phenyl-substituted cycloadduct (+)-10 and 51% ee for the
nitrophenyl-substituted cycloadduct (+)-11 (Table 1,
entries 8, 17). However, changing the solvent to CH2Cl2
with Rh2(R-DDBNP)4 6 removed differences in ee be-
tween the two cycloadditions (Table 1, entries 9, 18).
TABLE 1. Effect of Exp er im en ta l Con d ition s on th e
Cycloa d d ition s of Un sa tu r a ted Dia zod ion es 8 a n d 9
diazo
entry dione
temp
(°C) solvent
yield eea
(%)
(%) [R]T
b
catalyst
D
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
8
8
8
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
Rh2(OAc)4
25
25
25
25
CH2Cl2 77
CH2Cl2 71
hexane 75
hexane 41
Application of Hashimoto’s optimum conditions12a to
the aryl diazodione systems delivered good to excellent
yields of cycloadduct (74% with 8, Table 1, entry 10 and
98% with 9, entry 19). However, ee values were again
low and at the same level. Other solvents screened in
conjunction with Rh2(S-BPTV)4 7 and diazodione 8 (Table
1, entries 11, 12) resulted in lower yields and ee values.
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(R-BNP)4
Rh2(R-BNP)4
Rh2(R-BNP)4
Rh2(R-DDBNP)4 reflux hexane 87
Rh2(R-DDBNP)4 25
Rh2(R-DDBNP)4 25
Rh2(S-BPTV)4
Rh2(S-BPTV)4
Rh2(S-BPTV)4
Rh2(OAc)4
Rh2(S-DOSP)4
Rh2(R-BNP)4
Rh2(R-DDBNP)4 reflux hexane quant 28 +21.7
Rh2(R-DDBNP)4 25
Rh2(R-DDBNP)4 25
5
12
+3.1
+4.6
30 +14.8
19
6
reflux hexane 81
25 PhCF3 67
+8.3
-3.2
+7.9
19
hexane 87
CH2Cl2 51
PhCF3 74
hexane 49
CH2Cl2 50
CH2Cl2 quant
hexane quant
35 +14.0
10
14
11
7
+3.8
-6.3
-4.8
-2.4
25
25
25
25
25
From the results recorded in Table 1 it is apparent that
most of the catalysts deliver similar levels of ee for both
systems. However, it was interesting to find the antici-
pated difference in levels of asymmetric induction be-
tween the two systems, although it was somewhat
surprising that only one of the catalysts screened [Rh2-
(R-DDBNP)4 6] exhibited this effect. The absolute values
of ee are modest, but a difference in ee between cycload-
ducts arising from the two catalyst-associated ylides with
the same tethered alkene dipolarophile exists. This
provides the first unambiguous demonstration that elec-
tronic effects can play a role in determining the level of
asymmetric induction in such cycloaddition processes.
8
+7.0
reflux hexane 97
20 +10.8
hexane quant 51 +36.1
CH2Cl2 63
PhCF3 98
8
13
+7.4
-7.7
Rh2(S-BPTV)4
25
a
b
Determined by chiral HPLC. c ) 0.95-1.15 in CHCl3, T )
24-26 °C.
Proline derivative Rh2(S-DOSP)4 4 (Figure 1), the first
catalyst shown to deliver more than 50% ee in the
carbonyl ylide formation-intramolecular cycloaddition
reaction of R-diazo-â-ketoester 1 (Scheme 1),9b delivered
good yields of the cycloadduct (+)-10 (Table 1, entries 2,
3) in both CH2Cl2 (71%) and hexane (75%), but the ee
was low in both cases (5% and 12%, respectively). The
absolute sense of asymmetric induction was tentatively
assigned as that shown in Scheme 3 by analogy with the
corresponding ester system (Scheme 1).9 Enantioselec-
tivity was only marginally better in hexane; ee improve-
ments on changing to a hydrocarbon solvent are usually
much more significant with other asymmetric transfor-
mations utilizing Rh2(S-DOSP)4 4.7,9b With diazodione 9
and Rh2(S-DOSP)4 4, a quantitative yield of cycloadduct
(+)-11 was obtained (Table 1, entry 14), but with simi-
larly low ee (8%) to the unsubstituted system 8.
Attention next turned to binaphthyl phosphate cata-
lysts 5 and 6 (Figure 1), which had given the highest
levels of asymmetric induction with R-diazo-â-ketoester
1 (Scheme 1). Rh2(R-BNP)4 5 in hexane at room temper-
ature furnished cycloadduct (+)-10 with 30% ee but in a
much decreased yield of 41% (Table 1, entry 4), which is
most likely attributable to the poor solubility of this
catalyst in hydrocarbon solvent. Indeed, at reflux tem-
perature (where there are no catalyst solubility problems)
Removal of the tethered alkene from the substrates
provided ready opportunity for variation of the dipolaro-
phile electronics. Only a small range of dipolarophiles
have been examined in asymmetric intermolecular cyclo-
additions of carbonyl ylides, and of these, only DMAD,
used by Hashimoto, has met with any real success.12
However, with this dipolarophile and R-aryl-R-diazodione
12, only modest levels of asymmetric induction were
observed using a range of catalysts [up to 36% ee with
Rh2(S-BPTV)4 7 in PhCF3], and essentially no enantio-
induction was seen with R-aryl-R-diazodione 13.14 These
contrasting results led us to consider that when the dipole
and dipolarophile have similar electronic characteristics,
the ee is lower than when they oppose. To pursue this
analysis further, we considered it important to examine
cycloaddition reactions using the same substrates, but
with a less electron deficient dipolarophile. Arylacet-
ylenes were selected (16-18, Scheme 4) since, similarly
to the diazo substrates examined, introduction of sub-
stituents at the 4-position on the aryl group could allow
investigation of solely electronic variations. With these
unsymmetrical dipolarophiles, interest also lay in the
regiochemical outcome of the cycloaddition reactions.
J . Org. Chem, Vol. 68, No. 2, 2003 583

Hodgson et al.
TABLE 3. Cycloa d d u ct P a r tition in g in Com p etition
Rea ction s Em p loyin g Ar yla cetylen es 16-18
SCHEME 4
cycloadduct ratiob
diazo
entrya dione
catalyst
none
Rh2(OAc)4
Rh2(R-BNP)4 toluene
Rh2(S-BPTV)4 PhCF3
none
Rh2(OAc)4
Rh2(R-BNP)4 toluene
Rh2(S-BPTV)4 PhCF3
solvent 19 or 22 20 or 23 21 or 24
1
2
12
12
12
12
13
13
13
13
toluene
toluene
1
1
1
1
1
1
1
1
1
1
4
9
8
3c
4c
5
1.5
1
2.5
1.5
2
11
2
toluene
toluene
6
1
1
2.5
7c
8c
2.5
a
Carried out at room temperature, apart from entries 1 and 5
b
at reflux. Entries 1-4 refer to 19:20:21; entries 5-8 refer to 22:
23:24. c The ee values obtained were the same as when each
acetylene was used in isolation (see Table 4).
ments were then carried out (at room temperature) with
selected dirhodium catalysts.
Both substrates 12 and 13 underwent the tandem
catalytic ylide generation-intermolecular cycloaddition
process with Rh2(OAc)4 in toluene to deliver the cycload-
ducts from 4-methoxyphenylacetylene (16), phenylacet-
ylene (17), and 4-nitrophenylacetylene (18) in good to
excellent yields (Scheme 4, Table 2). Only single regio-
isomers were observed in the crude reaction mixtures
(one singlet in the crude 1H NMR spectra at approxi-
mately δ 6.5 for the olefinic proton). Once isolated, the
cycloaddition regiochemistry was determined to be the
same for all cycloadducts (that shown in Scheme 4) by
2D NOE NMR experiments, which showed a strong
correlation between the methyl group and the olefinic
proton in each case. Aromatic π-π interactions may
contribute to the origin of the regiochemistry.19 In the
absence of catalyst and using heat to decompose the diazo
diones, the cycloadducts were formed with the same
regioselectivity, but in much reduced yields (Table 2,
yields in parentheses).
For phenyldiazodione 12, in the absence of a catalyst,
reaction with nitrophenylacetylene 18 dominated (Table
3, entry 1), whereas for nitrophenyldiazodione 13, reac-
tion with methoxy- and nitrophenylacetylenes (16 and
18) was slightly favored over phenylacetylene 17 (Table
3, entry 5). Cycloaddition competition experiments with
phenyldiazodione 12 under dirhodium catalysis gave a
significantly greater proportion of cycloaddition from
nitrophenylacetylene 18. All three catalysts showed a
similar product profile (Table 3, entries 2-4). These
results indicate that the catalyst may be influencing the
energies of the molecular orbitals of the ylide. However,
with nitrophenyldiazodione 13, there was no significant
change in cycloadduct profile under dirhodium catalysis,
suggesting that during cycloaddition the catalysts do not
strongly perturb the molecular orbitals energies of the
comparatively more electron deficient ylide derived from
13. The possibility of interaction between the catalyst and
the alkynes also cannot be discounted.20
To account for the observations in the above competi-
tion reactions, at least in the absence of catalyst, the
HOMO-LUMO energy gaps between the two ylides 14
and 15 (Scheme 4) and the acetylenic dipolarophiles were
obtained from density functional theory (DFT) calcula-
tions (geometries optimized at the B3LYP/6-31G(d) level).14
The trend in reaction rates was well-reproduced by the
calculations in that the energy gap for nitrophenylacet-
ylene 18 and ylide 14 is much smaller than for the other
two alkynes, explaining the greater amount of cyclo-
adduct derived from this reactant combination in the
competition experiment. With ylide 15 the energy gaps
are fairly similar with all three alkynes, as reflected by
the product ratio.
TABLE 2. Ra cem ic Cycloa d d ition s of Dia zod ion es 12
a n d 13 w ith Ar yla cetylen es 16-18
yield
entry
diazodione
dipolarophile
cycloadduct
(%)a
1
2
3
4
5
6
12
12
12
13
13
13
16
17
18
16
17
18
19
20
21
22
23
24
76 (8)
60 (2)
82 (8)
89 (17)
82 (19)
58 (17)
a
Yield in the absence of catalyst is indicated in parentheses.
To examine any effect a catalyst might have on the
relative rates of the cycloadditions and to establish if
there is a correlation between the rate of cycloaddition
and asymmetric induction, competition experiments were
devised. Each diazodione was individually heated in the
presence of equal quantities (5 equiv) of all three acet-
Having determined the relative rates of cycloaddition
between diazodiones 12 and 13 with three arylacetylenes
16-18, a series of experiments were carried out to
determine the influence of the dipole and dipolarophile
on asymmetric induction. Cycloadditions were performed
1
ylenes and the product ratio determined. The H NMR
(C6D6) of the crude reaction mixture displayed clear,
distinct singlets for the olefinic proton of each cyclo-
adduct. Integration then gave the relative amounts
generated from each acetylene (Table 3). Similar experi-
(20) For UV studies demonstrating association between electron-
deficient (perfluorinated) rhodium carboxylate catalysts and alkenes,
see: (a) Doyle, M. P.; Colsman, M. R.; Chinn, M. S. Inorg. Chem. 1984,
23, 3684-3685. (b) Doyle, M. P.; Mahapatro, S. N.; Caughey, A. C.;
Chinn, M. S.; Colsman, M. R.; Harn, N. K.; Redwine, A. E. Inorg. Chem.
1987, 26, 3070-3072. However, with acetylenes 16-18 and Rh2(R-
BNP)4, UV titrations indicated that association is not significant, see:
Glen, R. D. Phil. Thesis, University of Oxford, 2002.
(19) Ibata, T.; Motoyama, T.; Hamaguchi, M. Bull. Chem. Soc. J pn.
1976, 49, 2298-2301.
584 J . Org. Chem., Vol. 68, No. 2, 2003

Catalytic Enantioselective [3 + 2]-Cycloadditions
TABLE 4. Effect of Exp er im en ta l Con d ition s on th e
Cycloa d d ition s of Dia zod ion es 12 a n d 13 w ith
Ar yla cetylen es 16-18
not quite reach the level seen with the phosphate catalyst
5 (Table 4, entries 6, 15). For phenylacetylene (17) it can
be seen that all the catalysts screened [with the exception
of Rh2(S-DOSP)4 4] display higher levels of ee for diazo
dione 12 compared with 13. These results are consistent
with the suggestion made earlier invoking contrasting
dipole/dipolarophile electronics for favorable ee.
In contrast to phenylacetylene (17), where some ee was
observed in CH2Cl2 at least for the nitrophenyl substrate
13, Rh2(R-BNP)4 5 gave asymmetric induction with
methoxy-substituted phenylacetylene 16 only when used
in toluene (Table 4, entries 23, 27). Yields were in the
same range as for unsubstituted phenylacetylene (17),
but the ee values were reduced. This was contrary to
expectation, as it was considered that the increased
electron richness of the dipolarophile might lead to higher
ee. A difference between the two diazo substrates was
still exhibited, with the electron-poor diazodione 13
giving higher ee. This difference was more pronounced
when Rh2(S-BPTV)4 7 was employed. A 40% ee was
observed (Table 4, entry 28), the same as for phenyl-
acetylene (17) with a similar yield.
For 4-nitrophenylacetylene (18) the use of Rh2(R-BNP)4
5 in CH2Cl2 resulted in the formation of enantioenriched
product 21 with phenyl-substituted diazodione 12. The
use of toluene as solvent raised the ee slightly from 24%
to 31% (Table 4, entries 30, 31). Surprisingly, with
diazodione 13 and Rh2(R-BNP)4 5, the two solvents
delivered cycloadduct 24 with very similar ee values (35
and 31% ee, Table 4, entries 34, 35); furthermore, both
substrates 12 and 13 demonstrated the same level of
asymmetric induction (31% ee) with the phosphate
catalyst 5 in toluene (Table 4, entries 31, 35), whereas
CH2Cl2 did exhibit the expected difference with 24% ee
compared to 35% ee (Table 4, entries 30, 34). Rh2(S-
BPTV)4 7 also showed a difference in ee, dependent upon
the substrate used. Racemic cycloadduct 21 was obtained
for the phenyldiazodione 12, but 38% ee was achieved
for the nitrophenyldiazodione 13 (Table 4, entries 32, 36).
With both diazodiones there was a large improvement
in yield compared to when using Rh2(R-BNP)4 5 as
catalyst.
The differences in ee between cycloadducts arising from
the same catalyst-associated ylide with the three differ-
ent dipolarophiles provide further support that electronic
effects play a role in determining the level of asymmetric
induction in such cycloaddition processes. Moreover, for
the phenyl-substituted dipole, where the competition
reactions conclusively ascertained that nitrophenylacet-
ylene 18 was the most reactive dipolarophile (Table 3,
entries 3, 4), it has been shown that higher rates of
cycloaddition do not lead to better asymmetric induction
(e.g. Table 4, entries 5, 23, 31). Thus, a simple reaction
scheme whereby a reduction in ee is due to loss of catalyst
from the ylide, followed by cycloaddition on the achiral
ylide, is shown not to be valid, at least in these
experiments.5a
dipolar- diazo
entry ophile dione
yield ee
b
catalyst
solvent
(%) (%)a [R]T
D
1
2
3
4
5
6
7
8
9
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
17
16
16
16
16
16
16
16
16
18
18
18
18
18
18
18
18
12 Rh2(OAc)4
toluene
CH2Cl2
toluene
CH2Cl2
toluene
PhCF3
PhCtCH 60
PhCtCH 25
PhCtCH 39
toluene
CH2Cl2
toluene
CH2Cl2
toluene
PhCF3
60
12 Rh2(S-DOSP)4
12 Rh2(S-DOSP)4
12 Rh2(R-BNP)4
12 Rh2(R-BNP)4
12 Rh2(S-BPTV)4
12 Rh2(OAc)4
12 Rh2(R-BNP)4
12 Rh2(S-BPTV)4
13 Rh2(OAc)4
13 Rh2(S-DOSP)4
13 Rh2(S-DOSP)4
13 Rh2(R-BNP)4
13 Rh2(R-BNP)4
13 Rh2(S-BPTV)4
13 Rh2(OAc)4
13 Rh2(R-BNP)4
13 Rh2(R-DDBNP)4 PhCtCH 80
13 Rh2(R-DDBNP)4 PhCtCH 83
13 Rh2(S-BPTV)4
12 Rh2(OAc)4
12 Rh2(R-BNP)4
12 Rh2(R-BNP)4
12 Rh2(S-BPTV)4
13 Rh2(OAc)4
15
29
25
37
47
3
5
3
38
28
+2.4
+8.7
-63.1
+48.7
38
36
-
2
3
-47.5
+67.5
10
11
12
13
14
15
16
17
18
19c
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
82
34
34
22
54
69
+5.3
+6.2
29
-57.4
48 -105.5
43
+97.0
PhCtCH 98
PhCtCH 73
66 -132.0
66 -114.6
76 -154.4
PhCtCH 49
45
+88.3
toluene
CH2Cl2
toluene
PhCF3
toluene
CH2Cl2
toluene
PhCF3
toluene
CH2Cl2
toluene
PhCF3
toluene
CH2Cl2
toluene
PhCF3
76
22
41
43
89
41
48
65
60
43
50
73
82
26
33
72
3
10
3
+3.0
+12.2
+2.7
13 Rh2(R-BNP)4
13 Rh2(R-BNP)4
13 Rh2(S-BPTV)4
12 Rh2(OAc)4
12 Rh2(R-BNP)4
12 Rh2(R-BNP)4
12 Rh2(S-BPTV)4
13 Rh2(OAc)4
0
18
40
-6.7
-43.8
+70.7
24
31
1
+22.0
+27.6
+2.0
13 Rh2(R-BNP)4
13 Rh2(R-BNP)4
13 Rh2(S-BPTV)4
35
31
38
+50.1
+49.0
+54.4
a
b
Determined by chiral HPLC. c ) 0.4-1.12 in CHCl3, T )
20-25 °C. c At 0 °C.
with each diazo dione 12 and 13 and dipolarophile 16-
18 combination (Scheme 4, Table 4).
For phenylacetylene (17), Rh2(S-DOSP)4 4 gives low
yields of essentially racemic cycloadduct 20 with both
diazo substrates 12 and 13 in chlorinated and hydrocar-
bon solvents (Table 4, entries 2, 3, 11, 12). Rh2(R-BNP)4
5 and Rh2(R-DDBNP)4 6 were both shown earlier to give
ee values in the range of 30-50% with the substrates 8
and 9 containing a tethered alkene as the dipolarophile
(Scheme 3, Table 1). It was considered that phenylacet-
ylene (17) might possess similar dipolarophile character
to a terminal alkene and therefore favor asymmetric
induction with the phosphate catalysts. Hexane as
solvent gave the best results for the intramolecular
cycloaddition; however, nitrophenyldiazodione 13 proved
only sparingly soluble in this solvent. Using CH2Cl2 as
solvent with diazo substrate 12 and Rh2(R-BNP)4 5 gave
essentially racemic product 20 in low yield (Table 4, entry
4). The yield remained poor but enantioselectivity im-
proved from 3% to 29% upon changing substrate to 13
(Table 4, entry 13). Switching to the hydrocarbon solvent
toluene gave better yields and ee values, as anticipated,
for both substrates 12 and 13 with nitro-substituted
diazodione 13 delivering cycloadduct 23 in 48% ee (Table
4, entries 5, 14). Rh2(S-BPTV)4 7 was found to be the most
efficient chiral catalyst in terms of yield for each sub-
strate. However, the asymmetric induction observed did
From the results in Table 4 discussed so far, it is
evident that the highest ee values were generally ob-
tained using phenylacetylene (17), and so further experi-
mentation was carried out using 17 as both the dipolaro-
phile and reaction solvent. The use of phenylacetylene
(17) as solvent had little effect with phenyldiazodione 12.
The yield with Rh2(OAc)4 remained at the 60% observed
J . Org. Chem, Vol. 68, No. 2, 2003 585

Hodgson et al.
with toluene (Table 4, entries 4, 7). Rh2(R-BNP)4 5 gave
the same 38% ee in neat phenylacetylene (17) as for
toluene (Table 4, entries 5, 8), whereas Rh2(S-BPTV)4 7
gave a reduced yield but slightly better ee (Table 4,
entries 6, 9). The yield with Rh2(OAc)4 was significantly
improved to near quantitative when the reaction was
carried out in neat phenylacetylene (17) with nitro-
phenyl-substituted substrate 13 (Table 4, entry 16). Also,
phenylacetylene (17) appeared to be a better solvent than
toluene for inducing enantioselectivity with the binaph-
thyl catalysts and nitrophenyldiazodione 13. Both phos-
phate catalysts 5 and 6 gave the same 66% ee (compared
with 48% ee for catalyst 5 in toluene), but the yield of
cycloadduct 23 was better for the more soluble Rh2(R-
DDBNP)4 6 at 80% (Table 4, entries 14, 17, 18). Lowering
the temperature to 0 °C, in combination with Rh2(R-
DDBNP)4 6, led to the highest ee observed in the current
study: 76% with good yield (83%) maintained (Table 4,
entry 19). Rh2(S-BPTV)4 7 gave essentially the same ee
values in PhCF3 and neat phenylacetylene (17) with
precursor 13 (Table 4, entries 15, 20) but with a 20%
reduction in yield in the latter case.
induction in such cycloaddition processes. Intermolecular
cycloaddition competition experiments with aryl-substi-
tuted carbonyl ylides show that nitrophenylacetylene (18)
reacts fastest in the cycloaddition process, and DFT
calculations14 concur with this observation. The presence
of a Rh(II) catalyst can affect the product ratio. The first
asymmetric induction in intermolecular cycloadditions of
carbonyl ylides with aryl acetylenes has been observed.
Higher rates of cycloaddition have been shown not to lead
to better asymmetric induction. The binaphthyl catalysts
gave the best enantioselectivities (up to 76% ee), extend-
ing their usefulness.21 These results suggest that a
complex blend of electronic effects from the dipole and
dipolarophile, together with the nature of the catalyst,
contribute to the origin of asymmetric induction. Further
studies are underway to investigate enantioselectivity in
such cycloaddition processes.
Ackn owledgm en t. We thank the EPSRC and Glaxo-
SmithKline for a CASE award (to R.G.), Dr. B. Odell
for NMR experiments and the EPSRC National Mass
Spectrometry Service Centre for mass spectra.
R-Aryl-R-diazodiones have been shown to undergo
tandem ylide formation-cycloaddition processes with
both internal and external dipolarophiles. A difference
in ee between the two electronically differing substrates
8 and 9 was displayed in intramolecular cycloadditions
with a tethered alkene, although only when Rh2(R-
DDBNP)4 6 was used as catalyst in hexane. As antici-
pated, the more electron-deficient cycloaddition precursor
9 (therefore bearing a closer resemblance to the R-diazo-
â-ketoester substrate 1 for which high levels of ee were
previously achieved in intramolecular cycloaddition)
delivered the higher enantioselectivity. Electronic effects
clearly play a role in determining the level of asymmetric
Su p p or tin g In for m a tion Ava ila ble: Full experimental
details of syntheses and characterization of cycloaddition
substrates and cycloadducts, chiral HPLC conditions for all
cycloadducts, copies of 1H and 13C spectra for all novel
compounds, and DFT studies and Gaussian 98 archive files
containing minimized structures and vibrational frequencies
for compounds 14-18. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O026307T
(21) Hodgson, D. M.; Petroliagi, M. Tetrahedron: Asymmetry 2001,
12, 877-881.
586 J . Org. Chem., Vol. 68, No. 2, 2003