Investigation of the transition-metal- and acid-catalyzed reactions of β-(N-tosyl)amino diazo carbonyl compounds

Article abstract of DOI:10.1021/jo0259818

A series of β-(N-tosyl)amino diazo carbonyl compounds have been prepared by nucleophilic condensation of N-tosylimines with acyldiazomethanes. The diazo decomposition of these diazo carbonyl compounds under various catalytic conditions, including Rh(II) carboxylates, Cu(I) and Cu(II) complexes, PhCO₂Ag/Et₃N, TsOH, and SnCl₂·2H₂O, has been investigated. It was found that, in most cases, the diazo decomposition gave preferentially 1,2-aryl migration product, but 1,2-hydride migration predominated when PhCO₂Ag/Et₃N was the catalytic system. Hammett correlation has been applied in the analysis of the electronic effects of 1,2-aryl migration. The factors that govern the migratory preference and the mechanistic aspects of the reaction are discussed.

Full text of DOI:10.1021/jo0259818

In vestiga tion of th e Tr a n sition -Meta l- a n d Acid -Ca ta lyzed
Rea ction s of â-(N-Tosyl)a m in o Dia zo Ca r bon yl Com p ou n d s
Nan J iang, Zhihua Ma, Zhaohui Qu, Xiaoyu Xing, Linfeng Xie,† and J ianbo Wang*
Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education,
Department of Chemical Biology, College of Chemistry, Peking University, Beijing 100871, China
wangjb@pku.edu.cn
Received May 28, 2002
A series of â-(N-tosyl)amino diazo carbonyl compounds have been prepared by nucleophilic
condensation of N-tosylimines with acyldiazomethanes. The diazo decomposition of these diazo
carbonyl compounds under various catalytic conditions, including Rh(II) carboxylates, Cu(I) and
Cu(II) complexes, PhCO2Ag/Et3N, TsOH, and SnCl2‚2H2O, has been investigated. It was found that,
in most cases, the diazo decomposition gave preferentially 1,2-aryl migration product, but 1,2-
hydride migration predominated when PhCO2Ag/Et3N was the catalytic system. Hammett
correlation has been applied in the analysis of the electronic effects of 1,2-aryl migration. The factors
that govern the migratory preference and the mechanistic aspects of the reaction are discussed.
In tr od u ction
1,2-Hydride, 1,2-alkyl, and 1,2-aryl migrations are
frequently encountered in the transition-metal-complex-
catalyzed decomposition of R-diazo carbonyl compounds.1,2
In some cases, these migration reactions can compete
with the typical reactions of R-diazo carbonyl compounds,
such as X-H insertions and cyclopropanations. On the
other hand, these reactions may be of synthetic utility
in organic chemistry. For example, 1,2-hydride migration
promoted by Rh2(O2CCF3)4 provides an efficient route to
(Z)-R,â-unsaturated carbonyl compounds,3 and SnCl2-
promoted 1,2-hydride migration of R-diazo-â-hydroxy
esters, known as Roskamp homologation, leads to the
formation of â-keto esters.4
â-Substitution of R-diazo carbonyl compounds has been
found to have marked influence on the reaction pathway
of diazo decomposition. For example, Rh2(OAc)4-catalyzed
decomposition of diazo compound 1 (Scheme 1) gives,
after tautomerization, â-keto ester 2 as a result of 1,2-
hydride migration,2a,b while acetoxy migration of 3 yields
4 as the major product under similar reaction conditions.2j,k
† Visiting professor from the Department of Chemistry, University
of WisconsinsOshkosh, Oshkosh, WI.
(1) For comprehensive reviews, see: (a) Doyle, M. P.; McKervey, M.
A.; Ye, T. Modern Catalytic Methods for Organic Synthesis with Diazo
Compounds; Wiley-Interscience: New York, 1998. (b) Ye, T.; McKervey,
M. A. Chem. Rev. 1994, 44, 1091.
(2) 1,2-Hydride migration: (a) Pellicciari, R.; Fringuelli, R.; Cec-
cherelli, P.; Sisani, E. J . Chem. Soc., Chem. Commun. 1979, 959. (b)
Ikota, N.; Takamura, N.; Young, S. D.; Ganem, B. Tetrahedron Lett.
1981, 22, 4163. (c) Hoffmann, K.-L.; Regitz, M. Tetrahedron Lett. 1983,
24, 5355. (d) Hudlicky, T.; Olivo, H. F.; Natchus, M. G.; Umpierrez, E.
F.; Pandolfi, E.; Volonterio, C. J . Org. Chem. 1990, 55, 4767. (e) Taber,
D. F.; Hoerrner, R. S. J . Org. Chem. 1992, 57, 441. (f) Ohno, M.; Itoh,
M.; Umeda, M.; Furuta, R.; Kondo, K.; Eguchi, S. J . Am. Chem. Soc.
1996, 118, 7075. 1,2-Alkyl or 1,2-aryl migration: (g) Nagao, K.; Chiba,
M.; Kim, S.-W. Synthesis, 1983, 197. (h) Ye, T.; McKervey, M. A.
Tetrahedron 1992, 48, 8007. (i) Kanemasa, S.; Kanai, T.; Araki, T.;
Wada, E. Tetrahedron Lett. 1999, 40, 5055. 1,2-Acetoxy migration: (j)
Lopez-Herrera, F. J .; Sarabia-Garcia F. Tetrahedron Lett. 1994, 35,
6705. (k) Lopez-Herrera, F. J .; Sarabia-Garcia F. Tetrahedron 1997,
53, 3325.
SCHEME 1
(3) (a) Taber, D. F.; Hennessy, M. J .; Louey, J . P. J . Org. Chem.
1992, 57, 436. (b) Taber, D. F.; Herr, R. J .; Pack, S. K.; Geremia, J . M.
J . Org. Chem. 1996, 61, 2908.
(4) (a) Holmquist, C. R.; Roskamp, E. J . J . Org. Chem. 1989, 54,
3258. (b) Holmquist, C. R.; Roskamp, E. J . Tetrahedron Lett. 1992,
33, 1131. (c) Padwa, A.; Hornbuckle, S. F.; Zhang, Z.; Zhi, L. J . Org.
Chem. 1990, 55, 5297. (d) Normura, K.; Iida, T.; Hori, K.; Yoshii, E. J .
Org. Chem. 1994, 59, 488.
In this paper, we report our detailed study on the
transition-metal-complex- or acid-catalyzed reaction of
R-diazo carbonyl compound 5, which bears an N-tosy-
10.1021/jo0259818 CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/28/2002
J . Org. Chem. 2003, 68, 893-900
893

J iang et al.
SCHEME 2
SCHEME 3
We found that a catalytic amount of DBU (10-20% mol)
was enough to promote the condensation efficiently in
MeCN at room temperature. Moderate to good yields of
5a -p were obtained, as shown in Table 1 (condition B).
Since DBU is a much weaker base than LDA or NaH,
the fact that the nucleophilic condensation can be ef-
ficiently promoted by this base suggests that the acidity
of the proton attached to the diazo carbon of 6 is much
greater than generally believed.9 A similar reaction
condition has also been employed to condense acyldiaz-
omethanes 6 with aromatic or aliphatic aldehydes 7, thus
constituting a mild route to diazoketols 9, which are
useful precursors for the preparation of â-keto esters.2,4
Having prepared 5a -p , we then proceeded to examine
the diazo decomposition of these compounds under vari-
ous reaction conditions, as depicted in Scheme 3. R-Dia-
zoketamine 5a (Ar ) C6H5, R ) H) was the first substrate
studied to examine the effect of the catalysts on the
reaction, and the results are summarized in Table 2.
lamino group on its â-position.5 The study reveals that,
contrary to the R-diazo-â-hydroxy carbonyl compound 1
(X ) OH) in which the diazo decomposition gives pref-
erentially 1,2-hydride migration product, the correspond-
ing reaction with 5 results in predominantly 1,2-aryl
migration in most cases. It is also found that the
migratory aptitude is affected by substituents on the aryl
group, the catalytic conditions employed, and the struc-
ture of the diazo substrates.
1
Three products have been isolated or identified with H
NMR spectra, the 1,2-phenyl migration product trans-
R-phenyl-â-enamino ester 10a , its cis isomer 11a , and
the 1,2-hydride migration product cis-â-phenyl-â-enamino
ester 12a (Scheme 3). From Table 2, it is clear that the
diazo decomposition could be effected under various
reaction conditions and the ratio of the three products
depends on the reaction conditions employed. It is
worthwhile noting that the combined yields, after column
chromatography, of the three products range from very
good to excellent. The ratio of (10a + 11a )/12a reflects
the migratory aptitude of Ph vs H, and shows noticeable
dependence on the metal as well as the ligands on the
catalyst. With Rh2(OAc)4 as the catalyst, 1,2-phenyl
migration was predominant (88:12, Table 2, entry 1).
Stronger electron-withdrawing ligands in Rh2(O2CCF3)4
resulted in less selectivity between Ph and H migration
(61:39, Table 2, entry 3) while the less electron-with-
drawing ligands in Rh2(acam)4 enhanced the selectivity
(91:9, Table 2, entry 2). These results are consistent with
the selectivity trends previously reported for C-H inser-
tion regioselectivity and cyclopropanation diastereose-
lectivity.10 A ligand-induced electronic effect is also
evident for the three copper catalysts studied, Cu(acac)2,
Cu(hfacac)2, and Cu(MeCN)4PF6. Of the three catalysts,
Cu(acac)2 gave the best 1,2-phenyl migratory selectivity
(88:12, Table 2, entry 4), while electron-withdrawing
Resu lts
N-Tosyl diazoketamines 5a -p were prepared by nu-
cleophilic condensation of acyldiazomethane with N-
sulfonylimines, as shown in Scheme 2. Deprotonation of
acyldiazomethanes is usually achieved with lithium
diisopropylamide (LDA).6 We found that sodium hydride
was a better base in our case because of its ease of
handling and clean reaction. Thus, 6 was treated with
NaH in anhydrous THF at 50 °C, followed by the addition
of N-sulfonylimine 7 (Scheme 2). N-tosyl diazoketamines
5a -p were obtained in good yields, as shown in Table 1.
Although nucleophilic addition to N-sulfonylimines has
been widely applied to organic synthesis,7 to our best
knowledge, there has been no systematic study on the
nucleophilic addition of R-diazocarbonyl anion to N-
sulfonylimines.8
During the course of this investigation, we discovered
a much milder reaction condition for this condensation.5c
(5) For the preliminary communications related to this investigation,
see: (a) J iang, N.; Qu, Z.; Wang, J . Org. Lett. 2001, 3, 2989. (b) J iang,
N. Wang, J . Synlett 2002, 149. (c) J iang, N.; Wang, J . Tetrahedron
Lett. 2002, 43, 1285.
(6) (a) Pellicciari, R.; Fringuelli, R.; Sisani, E. J . Chem. Soc., Perkin
Trans. 1 1981, 2566. (b) Pellicciari, R.; Natalini, B.; Sadeghpour, B.
M.; Marinozzi, M.; Snyder, J . P.; Williamson, B. L.; Kuethe, J . T.;
Padwa, A. J . Am. Chem. Soc. 1996, 118, 1. (c) Moody, C. J .; Taylor, R.
J . Tetrahedron Lett. 1987, 28, 5351.
(7) For a review, see: Weinreb, S. M. Top. Curr. Chem. 1997, 190,
131; For a recent example, see: Lu, W.; Chan, T. H. J . Org. Chem.
2001, 66, 3467.
(9) It has been reported that certain aldehydes can condense with
ethyl diazoacetate in the absence of base; see: Sarabia, F.; Lopez-
Herrera, F. J . Tetrahedron Lett. 2001, 42, 8801.
(10) Doyle, M. P.; Westrum, L. J .; Wolthuis, W. N. E.; See, M. M.;
Boone, W. P.; Bagheri, V.; Pearson, M. M. J . Am. Chem. Soc. 1993,
115, 958.
(8) For
a
condensation reaction of N-(2,2,2-trichloroethylidene)-
tosylamide with ethyl diazoacetate in a review paper, see: Fink, J .;
Regitz, M. Synthesis 1985, 569.
894 J . Org. Chem., Vol. 68, No. 3, 2003

Reactions of â-(N-Tosyl)amino Diazo Carbonyls
TABLE 1. Na H- or DBU-P r om oted Con d en sa tion of Acyld ia zom eth a n es w ith N-Tosylim in es
N-tosylimine 7
yield (%)a
acyldiazomethane 6
condition
Ab
condition
Bc
entry
R1
Ar
R
product
1
2
3
4
5
6
7
8
8
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
C6H5
CH3
C6H5-
H
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
5o
5p
74
85
83
d
d
79
d
68
87
d
32
26
70
81
d
72
77
70
82
49
78
63
70
66
73
69
77
75
<20
76
57
p-MeOC6H4-
m-BrC6H4-
m-CF3C6H4-
m-CNC6H4-
p-PhC6H4-
p-FC6H4-
p-ClC6H4-
o-MeC6H5-
o-AllyloxyC6H4-
2,4-Cl2C6H3-
2,6-Cl2C6H3-
3,5-(MeO)2C6H3-
C6H5-
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
9
10
11
12
13
14
15
C6H5-
C6H5-
d
a
b
Yields after column chromatographic purification with silica gel. Condition A: NaH/THF, 50 °C. c Condition B: 10 mol % DBU/
d
MeCN, room temperature. Not tried.
TABLE 2. Effects of Ca ta lyst on 1,2-P h en yl vs 1,2-Hyd r id e Migr a tor y Selectivity
reaction
conditions
product ratioa
10a :11a :12a
overall
entry
catalyst
yieldb (%)
1
2
3
4
5
6
7
8
9
Rh2(OAc)4
Rh2(acam)4
Rh2(O2CCF3)4
Cu(acac)4
Cu(hfacac)4
Cu(MeCN)4PF6
PhCO2Ag
CH2Cl2, 0 °C
CH2Cl2, 0 °C
CH2Cl2, 0 °C
CH2Cl2, 25 °C
CH2Cl2, 25 °C
CH2Cl2, 25 °C
Et3N, THF, reflux
CH2Cl2, 0 °C
CH2Cl2, 0 °C
85:3:12
85:6:9
98
95
97
91
94
89
81
87
89
57:4:39
56:32:12
38:44:18
64:7:29
27:16:57
17:62:21
29:71:0
SnCl2‚2H2O
TsOH
a
b
The product ratio was determined by 1H NMR (400 MHz) of the crude product. Combined yields of 10a , 11a , and 12a after column
chromatography.
TABLE 3. Dia zo Decom p osition Ca ta lyzed by Rh 2(OAc)4
isolated yield (%)
ligands in Cu(hfacac)2 and Cu(MeCN)4PF6 reduced the
selectivity, especially in the case of Cu(MeCN)4PF6.
We also examined the diazo decomposition of 5a in
refluxing THF with PhCO2Ag/Et3N, a typical Wolff
rearrangement catalyst.11 We were astonished to find
that the 1,2-hydride migration product 12a was favored
(57%) in this case. No products due to Wolff rearrange-
ment were observed. For comparison, the diazo decom-
position was also carried out under photolysis conditions
(hν > 300 nm) in benzene at room temperature. The
photolysis gave 1,2 migration products 10a , 11a , and 12a
in 94% yield and with a ratio of 42:40:18. Again, no Wolff
rearrangement products were detected.
Since diazo decomposition could also be promoted with
catalytic Lewis acid and protonic acid, we next proceeded
to treat diazoketamine 5a with SnCl2‚2H2O and TsOH,
respectively. In both cases 1,2-phenyl migration products
are predominant, as shown in Table 2 (entries 8 and 9).
On the basis of these results, it is clear that the 1,2-
phenyl migration pathway dominates in most cases
during the diazo decomposition of 5a , except in the case
promoted by PhCO2Ag where 1,2-hydride migration is
favored. To assess the effect of substrates on product
selectivity, we set out to study reactions of a series of
N-tosyl diazoketamines under three different catalytic
N-tosyl R-diazo- product ratioa
entry
ketamine
10:11:12
10
71b
12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
5a
5b
5c
5d
5e
5f
5g
5h
5i
85:3:12
c
c
84:9:7
73:5:22
62:5:33
51:6:47
92:4:4
93:3:4
75:12:13
e
79b
53d
59d
48d
86d
87d
68b
81d
76d
84b
18b
24b
20b
c
c
c
c
c
c
5j
e
5k
5l
e
e
90d,f
5m
5n
71:6:23
e
68d
90d
c
c
a
The product ratio was determined by 1H NMR (400 MHz).
Yields after column chromatography. c 1,2-Hydride migration
b
products 12 and the cis isomer of the phenyl migration products
d
11 could not be isolated due to their minute amount. Yields after
single recrystallization of the crude product. e Ratio not deter-
mined. f The yield refers to the trans isomer (E)-ethyl 3-(2,6-
dichlorophenyl)-3-(N-tosyl)amino-2-propenoate.
conditions, Rh2(OAc)4, TsOH, and PhCO2Ag/Et3N. The
corresponding results are summarized in Tables 3-5,
respectively.
For most reactions catalyzed by Rh2(OAc)4, the trans
isomer 10 from 1,2-aryl migration was the only product
(11) Newman, M. S.; Beal, P. F., III. J . Am. Chem. Soc. 1950, 72,
5163.
J . Org. Chem, Vol. 68, No. 3, 2003 895

J iang et al.
TABLE 4. Dia zo Decom p osition Ca ta lyzed by P h CO2Ag
could also result from such isomerization promoted by
TsOH. This was indeed confirmed by carrying out the
reaction of 5a in the presence of a 10-fold amount of
TsOH (10% mol) at room temperature. As expected, the
ratio of cis to trans changed from 71:29 to 95:5. At this
point it is unclear as to what relative percentage of the
product selectivity results from the intrinsic cis/trans
migratory selectivity vs cis/trans isomerization.
N-tosyl R-diazo-
product ratioa
yield of
entry
ketamine
10:11:12
12b (%)
1
2
3
4
5
6
7
5a
5b
5c
5d
5e
5f
27:16:57
16:25:59
14:25:61
10:27:63
10:22:68
21:21:58
17:25:58
53
46
59
57
54
51
43
5g
Discu ssion
a
The product ratio was determined by 1H NMR (400 MHz).
b
Yields after column chromatography.
Our results clearly indicate that 1,2-migratory selectiv-
ity in the decomposition of 5 varies drastically with the
ligands and the type of metal contained in the catalysts.
It is generally accepted that in transition-metal-catalyzed
diazo decomposition, a metal carbene instead of a free
carbene is involved as an intermediate. The metal
carbene is a resonance hybrid of a formal metal carbene
(13) and a metal-stabilized carbocation (14), which
together describe the reactivity of the intermediate.12 It
seems reasonable then that the reactivity of such an
intermediate may be influenced by factors such as the
ligands, metal, and R group on the phenyl ring.
TABLE 5. Dia zo Decom p osition Ca ta lyzed by TsOH
N-tosyl R-diazo-
product ratioa
overall yield of
entry
ketamine
10:11c
10 + 11b (%)
1
2
3
4
5
5a
5b
5d
5f
29:71
38:62
25:75
22:78
5:95
89
90
85
95
87
5g
a
The product ratio was determined by 1H NMR (400 MHz).
b Yields after column chromatography. c The 1,2-hydride migration
product 12 was not detected in 1H NMR.
isolated from the reaction mixture. Although both 11 and
1
12 were detectable on TLC and in H NMR of the crude
products, they were not isolated due to their minute
quantity. In cases where a significant amount of 12 was
formed (entries 3-5), it was isolated and the yields were
determined after column chromatographic purification.
It is interesting to note that electron-withdrawing sub-
stituents on the phenyl ring (m-Br, m-CF3, and m-CN,
entries 3-5) seem to reduce the migratory aptitude of
the aryl group relative to the hydrogen. Remarkably, 1,2-
hydride migration dominates completely (entry 12) over
2,6-dichlorophenyl group migration, and the major prod-
uct 12l was isolated in excellent yield. Obviously, the
migratory aptitude of the aryl group is influenced to some
degree by the electronic effect imposed by the ring
substituents.
For PhCO2Ag/Et3N-catalyzed reactions, the predomi-
nant formation of 1,2-hydride migration product appears
to be general. As shown in Table 4, for substrates with
either electron-withdrawing groups (entries 3-5 and 7)
or an electron-donating group (entry 2), the main product
is â-aryl-â-(N-tosyl)enamino ester 12. Interestingly, the
product ratio of 1,2-aryl vs 1,2-hydride migration cata-
lyzed by silver benzoate is rather independent of the
electronic nature of the aryl substituents.
First, different ligands on Rh or Cu catalysts exert a
profound effect on the competitive 1,2-aryl vs 1,2-hydride
migratory selectivity, as shown in Table 1. There seems
to be a general trend that electron-withdrawing ligands
enhance 1,2-hydride migration for both Rh- and Cu-
catalyzed decompositions. This trend can be rationalized
in terms of the increase in electrophilicity (more contri-
bution from carbocation-like structure 14) of the inter-
mediate metal carbene, which results in a more reactive
carbene center and its reduced ability to discriminate
migrating substituents. Such observations are consistent
with the selectivity trends previously reported for C-H
insertion regioselectivity and cyclopropanation diaste-
reoselectivity, in which electron-withdrawing ligands on
Rh and Cu catalysts were found to decrease the selectiv-
ity.10
To understand the steric effect around the metal
carbene center, we prepared diazo ester 15 (Scheme 4)
and subjected it to the usual decomposition condition
with Rh2(OAc)4. Interestingly, 15 failed to decompose
with Rh2(OAc)4, even in refluxing ClCH2CH2Cl (83 °C).
We believe this is due to the bulky 2,6-di-tert-butyl-4-
methylphenyl group, which prevents the approach of the
negatively polarized diazo carbon to the axial position of
the hindered Rh(II) catalyst. Circumstantial evidence for
this explanation is given by the fact that 15 was rapidly
decomposed to give 1,2-phenyl migration products 16 and
17 (Scheme 4) with a catalytic amount of TsOH at 0 °C
in CH2Cl2.
When diazoketamines 5a , 5b, 5d , 5f, and 5g were
treated with 1 mol % TsOH at 0 °C in anhydrous CH2-
Cl2, the diazo compounds disappeared within 30 min to
give a mixture of 1,2-aryl migration products 10 and 11.
In contrast to the reactions catalyzed by Rh2(OAc)4, 1,2-
hydride migration product 12 could not be detected in
1
the H NMR of the crude reaction mixture in this case.
Moreover, cis-R-aryl-â-enamino ester 11 was formed as
the major product. The cis isomer is obviously more stable
than the corresponding trans isomer because of the
intramolecular hydrogen bonding in the former case.
Since cis/ trans isomerization of R-aryl-â-enamino esters
was observed during the silica gel column chromato-
graphic separation, it is likely that the predominant
formation of the cis isomer in TsOH-catalyzed reactions
(12) For a recent mechanistic investigation on Rh(II)-mediated diazo
decomposition, see: (a) Qu, Z.; Shi, W.; Wang, J . J . Org. Chem. 2001,
66, 8139. (b) Pirrung, M. C.; Liu, H.; Morehead, A. T., J r. J . Am. Chem.
Soc. 2002, 124, 1014.
896 J . Org. Chem., Vol. 68, No. 3, 2003

Reactions of â-(N-Tosyl)amino Diazo Carbonyls
SCHEME 4
SCHEME 5
F IGURE 1. Hammett correlation of the relative migratory
aptitude of the Rh2(OAc)4-catalyzed reaction.
SCHEME 6
On the basis of these results and those reported in the
literature,2,3 a possible reaction mechanism can be pro-
posed for Rh2(OAc)4-catalyzed reactions, as depicted in
Scheme 5. The first step involves the formation of
rhodium(II) carbene intermediate 13 or 14, as generally
believed for Rh(II)-mediated diazo reactions.12 The sub-
sequent aryl migration may proceed through a bridged
phenonium ion, 18, or through a process in which the
substituent migrates simultaneously with the dissocia-
tion of the rhodium(II) catalyst. The former pathway is
similar to the well-known 1,2-shift in carbonium ions.13
If the bridged ion intermediate 18 is involved in the
reaction pathway, the orbital system of the aryl group is
expected to assist the stabilization of the carbonium ion
or partial carbonium ion through delocalization. In this
case, a significant substituent electronic effect is ex-
pected, as in the case of 1,2-aryl shift in carbonium ions.
To substantiate this mechanism, the relative migratory
aptitude of the aryl group, kX/kH, was calculated from the
product distribution data in Table 3, by taking 1,2-
hydride migration as the internal reference for each
reaction.14,15 The relative migratory aptitude was further
analyzed with Hammett substituent constants.16 As
shown in Figure 1, although the data do not fit well to
the Hammett equation (F ) -1.43 with σ, r ) -0.85),
the obvious trend is that an aryl group with electron-
donating substituents has a higher migratory aptitude.
This trend suggests that the aryl group is migrating to
an electron-deficient carbon, in agreement with the fact
that rhodium(II) carbene is electrophilic. However, the
relatively small magnitude of the reaction constant seems
to disfavor a bridged ion pathway. If a bridged ion were
involved, a stronger substituent electronic effect should
have been observed, especially for the p-methoxy group
because it would have direct interaction with the devel-
oping positive or partial positive charge.13,17 Thus, the
moderate effect imposed by the substituents on the
migratory aptitude of an aryl group lends its support to
a concerted migration pathway, although a composite of
the two pathways cannot be ruled out.
It is useful to compare these electronic effects to those
of similar reactions involving R-diazo-â-hydroxy carbonyl
compounds and their derivatives. For R-diazo-â-hydroxy
carbonyl compounds 1 (X ) OH), literature information
and our study indicate that 1,2-hydride migration domi-
nates in all cases,2a,b even when the â-substituent is a
p-methoxyphenyl group (1, X ) OH, R ) p-MeOC6H4).
However, when the â-hydroxyl group is derivatized, 1,2-
aryl migration may dominate in some cases. For example,
Kanemasa has demonstrated that 1,2-migratory selectiv-
ity of competing aryl vs hydride shift could be reversed
by modifying the electronic property of the migrating aryl
group in diazo compound 19, as depicted in Scheme 6.18
Thus, 1,2-hydride shift in 19 is preferred over p-nitro-
phenyl migration to give 21, while changing p-NO2 to
(13) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper Collins Publishers: New York,
1987; p 434.
(14) 1,2-Hydride migration may also be affected by the substitution
on the phenyl ring; therefore, a question remains when taking the 1,2-
hydride migration as internal reference. For a theoretical investigation
of the influence of bystander substituents on the 1,2-H and 1,2-Ph
migration in free carbene, see: Keating, A. E.; Garcia-Garibay, M. A.;
Houk, K. N. J . Phys. Chem. A 1998, 102, 8467.
(15) The relative migratory aptitude data as well as the Hammett
correlation with σ+ and σ- have been submitted as Supporting
Information.
(17) Brown, H. C.; Kim, C. J . J . Am. Chem. Soc. 1971, 93, 5765.
(18) Kanemasa, S.; Araki, T.; Kanai, T.; Wada, E. Tetrahedron Lett.
1999, 40, 5059.
(16) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.
J . Org. Chem, Vol. 68, No. 3, 2003 897

J iang et al.
p-MeO reverses the aptitude and leads to the main
product 20. For diazo decomposition of R-diazo-â-acetoxy
carbonyl compound 3 (Scheme 1), the acetoxy migration
is predominant, probably through a five-membered ring
intermediate in which the acetoxy carbonyl oxygen
interacts with the positively polarized rhodium carbene.2b,j,k
To further understand the migratory preference of aryl
vs hydrogen, we need to consider a conformational
requirement involved in the migration process. We envi-
sion the migrating group to be perpendicular to the plane
of the sp2-hybridized carbene carbon, such that it can best
overlap with the empty p orbital on the carbene carbon.
Three representative conformations are depicted by 22-
24 that would lead to each migration product, respec-
tively.
reported.22 We then continued to build the remainder of
the carbenoid complex and carried out an optimized
conformational search using various geometry labels for
dihedral angles. The energy of each conformation was
optimized by the augmented MM2 parameters provided
in the CAChe package. The conformations represented
by 22-24 were optimized by locking the migrating H or
Ph group parallel to the empty p orbital (on the assump-
tion that this allows for the maximum overlap during the
migration) on the carbene center. The results show that
conformation 23 is the most stable conformation, in which
case Ph migration appears to be favored over H migra-
tion; however, this would lead to the cis isomer 11. The
conformation required for the phenyl migration (22) to
give the trans isomer product 10 would put NHTs on the
side of Rh and its ligands, an apparently sterically
hindered situation. We felt that this approach was
oversimplified because the C-Rh bond in the carbenoid
intermediate might have departed to some extent during
the actual reaction, thus reducing the steric interaction
between NHTs and the ligands on the rhodium. Never-
theless, the same argument could presumably apply to
24.
It appears that the conformational analysis could not
provide conclusive explanation for the preferential Ph
migration in the reaction of 5. Considering the fact that
for the similar R-diazo-â-hydroxy carbonyl compound 1
(X ) OH) the diazo decomposition almost exclusively gave
H migration products,2a the nature of the â-substitution
is obviously the crucial factor in determining which group
migrates preferentially. In free carbene chemistry, it has
long been recognized that the ease of migration is affected
by nonmigrating substituents (i.e., bystanders);14,19,23
however, the similar bystander effect in the transition-
metal-complex-catalyzed diazo reaction has rarely been
documented previously. Further systematic investigation
is required to elucidate the mechanism of this effect.
To further elucidate factors that may affect the migra-
tory aptitude, Rh2(OAc)4-catalyzed reaction of diazoke-
tones 5o and 5p and diazoamides 25 and 26 was
investigated (Scheme 7). The reaction of diazoketones 5o
and 5p gave predominately 1,2-phenyl migration prod-
ucts 27 and 28, just as the corresponding reaction of diazo
ester 5a . The reaction with diazoamides 25 and 26,
however, resulted in 1,2-hydride migration products 29
and 30 as the only isolated products. This result indicates
that other structural factors exist that may override the
effect of the â-NHTs group in dictating the preferential
1,2-aryl migration.
For the PhCO2Ag/Et3N-catalyzed reaction, similar
electronic effect analysis has been performed using the
product distribution data collected in Table 4. A smaller
magnitude of the reaction constant was obtained (Figure
2, F ) -0.20, r ) -0.82). The Hammett linear correlation
for this catalytic system is not so good, but the general
trend of the electronic effects is the same as in the Rh-
(II)-mediated reaction. The PhCO2Ag and Et3N combina-
tion has been widely applied in the Wolff rearrangement
of R-diazo carbonyl compounds; however, the exact mech-
anism of how this catalytic system decomposes diazo
The results summarized in Tables 2 and 3 show that,
for the Rh(II)-catalyzed reaction, in all cases the ther-
modynamically less stable trans isomer 10 from 1,2-aryl
migration is the major product. The product ratios were
determined by 1H NMR of the crude products before
column chromatography, and we have confirmed that the
trans/ cis isomerization does not occur under the Rh(II)-
mediated reaction condition. Therefore, the product
distributions obtained in this case originate from the
diazo decomposition reaction itself but not isomerization.
The predominate formation of trans 1,2-aryl migration
products implies that the conformational factors may play
a role in determining the migratory preference.19,20 In
other words, the rhodium(II) carbene intermediate may
preferentially take the formation of 22 that is in favor of
aryl migration to give the trans product.
To assess the relative stability of the three conforma-
tions 22-24 shown above, we undertook computational
calculations on the rhodium-complexed carbenoid (rep-
resented by resonance structures 13 and 14 in Scheme
5) using augmented MM2 parameters in the CAChe
package.21 We first built the model catalyst Rh2(OAc)4
and optimized its structure using MM2. All O-Rh bond
lengths after optimization were equal, and the Rh-Rh
bond length was in good agreement with that previously
(19) A recent study of the 1,2-migration in singlet free carbene
concludes that the molecular geometry of the carbene intermediate is
responsible for the higher 1,2-alkyl migration. See: Farlow, R. A.;
Thamattoor, D. M.; Sunoj, R. B.; Hadad, C. M. J . Org. Chem. 2002,
67, 3257.
(20) Aggarwal, V. K.; Sheldon, C. G.; Macdonald, G. J .; Martin, W.
P. J . Am. Chem. Soc. 2002, 124, 10300.
(21) Structure optimization was conducted by an augmented version
of Allinger’s MM2 force field provided in the CAChe system, version
3.1. The optimized structures have been submitted as Supporting
Information.
(22) Pirrung, M. C.; Morehead, A. T., J r. J . Am. Chem. Soc. 1994,
116, 8991.
(23) For a review, see: Nickon, A. Acc. Chem. Res. 1993, 26, 84.
898 J . Org. Chem., Vol. 68, No. 3, 2003

Reactions of â-(N-Tosyl)amino Diazo Carbonyls
SCHEME 8
reaction mixture, it was not possible to conduct similar
analysis of the electronic effects. A possible pathway to
account for the 1,2-phenyl migration is presented in
Scheme 8. The initial protonation occurs on the nega-
tively polarized diazo carbon.25 The aryl migration to the
R-carbon may proceed with simultaneous extrusion of
nitrogen gas to generate a cationic intermediate, 32,
which is deprotonated to give 1,2-phenyl migration
product 11a . The migration may also follow a stepwise
pathway, with the dinitrogen being extruded to give a
cationic intermediate,26 followed by the phenyl group
migrating to the cationic center.
F IGURE 2. Hammett correlation of the relative migratory
aptitude of the AgO2CPh/Et3N-catalyzed reaction.
SCHEME 7
Con clu sion
It is obvious that, among the factors that influence the
migratory aptitude in the reaction of R-diazo carbonyl
compounds, the substituents at the â-position have the
prime importance. Other factors include the catalysts and
the structure of the carbonyl group. This investigation
demonstrates that the diazo decomposition of N-tosyl-R-
amino â-diazo carbonyl compounds gives either 1,2-aryl
migration or 1,2-hydride migration products as the major
products, depending on the reaction conditions. In the
Rh2(OAc)4-catalyzed reaction, 1,2-aryl migration domi-
nates and the thermodynamically less stable trans-R-
aryl-â-enamino ester is the major product in most cases,
while in the PhCO2Ag/Et3N-catalyzed reaction, 1,2-
hydride migration is the major reaction pathway. Al-
though the exact picture of the reaction mechanism is
still not clear at present, we prefer a concerted migration
for the Rh(II)-mediated reaction on the basis of the
information now available.
compounds is still not clear.11,24 It is worthwhile to
compare this catalytic system with Rh2(O2CCF3)4. In the
Rh2(O2CCF3)4-catalyzed reaction of 5a , the amount of 1,2-
hydride migration products increased significantly com-
pared to that in Rh2(OAc)4- and Rh2(acam)4-catalyzed
reactions (Table 2, entry 3). Interestingly, Wolff rear-
rangement products were observed in the Rh2(O2CCF3)4-
catalyzed reaction of R-diazo carbonyl compounds,22
although Rh(II) catalysts generally promote C-H inser-
tions, cyclopropanations, and ylide formations. It appears
that these two catalysts share some similarities in their
reactions with diazo compounds. The exact reason 1,2-
hydride migration has higher migratory aptitude under
these reaction conditions, however, is unclear at present.
Finally, in the TsOH-catalyzed reaction, since 1,2-
hydride products could not be detected in the crude
Exp er im en ta l Section
Gen er a l In for m a tion . See ref 12a.
Gen er a l P r oced u r e for th e P r ep a r a tion of N-Tosyl
Dia zok eta m in es 5a -p . With Na H a s th e Ba se (Con d ition
A). To a solution of N-tosylimine 7 (1.0 mmol) in anhydrous
THF (20 mL) at 50-55 °C was added sodium hydride (1.0
mmol) under N2. The mixture was stirred for 5 min, and the
solution of acyldiazomethane 6 (1.2 mmol) in anhydrous THF
(5 mL) was added dropwise during a period of 30 min. The
mixture was then stirred until TLC analysis indicated the
disappearance of the starting material. The solution was then
cooled to -30 °C, and saturated aqueous NaHCO3 (20 mL) was
(24) (a) Yukawa, Y.; Tsuno, Y.; Ibata, T. Bull. Chem. Soc. J pn. 1967,
40, 2613. (b) Yukawa, Y.; Tsuno, Y.; Ibata, T. Bull. Chem. Soc. J pn.
1967, 40, 2618.
(25) Malherbe, R.; Dahn, H. Helv. Chim. Acta 1977, 60, 2539
(26) Creary, X. J . Am. Chem. Soc. 1984, 106, 5568.
J . Org. Chem, Vol. 68, No. 3, 2003 899

J iang et al.
carefully added. Usual workup gave a crude product, which
was purified by column chromatography with silica gel.
With DBU a s th e Ba se (Con d ition B). To a solution of
acyldiazomethane 6 (1.2 mmol) in anhydrous CH3CN (5 mL)
at room temperature under N2 were added successively a
solution of DBU (0.1 mmol) in anhydrous CH3CN (1 mL) and
N-tosylimine 7 (1 mmol) in anhydrous CH3CN (4 mL) via
syringes. After being stirred at room temperature for 8 h, the
reaction was quenched with saturated aqueous NaHCO3, and
then extracted with CH2Cl2 (20 mL). The solvent was removed
by evaporation under reduced pressure, and the crude product
was purified by flash chromatography to yield N-tosyl diaz-
oketamines 5a -p .27
Gen er a l P r oced u r e for Cop p er -Com p lex-Ca ta lyzed
Dia zo Decom p osition of N-Tosyl Dia zok eta m in e. A solu-
tion of N-tosyl diazoketamine (30 mg, 0.08 mmol) in 20 mL of
CH2Cl2 was stirred at room temperature under N2, followed
by the addition of copper catalyst (10% mmol). The reaction
mixture was stirred until the disappearance of the starting
material. The solvent was removed under vacuum, and the
residue was analyzed by 1H NMR and purified by column
chromatography.
Gen er a l P r oced u r e for Dia zo Decom p osition of N-
Tosyl Dia zok eta m in e Ca ta lyzed by P h CO2Ag. To a solu-
tion of N-tosyl diazoketamine (100 mg, 0.26 mmol) in anhy-
drous THF (50 mL) was added a solution of PhCO2Ag (10 mg)
in Et3N (0.2 mL). The mixture was heated under gentle reflux
for 4 h. The solvent was removed under vacuum, and the
residue was analyzed by 1H NMR and purified by column
chromatography.
Gen er a l P r oced u r e for th e TsOH-Ca ta lyzed Dia zo
Decom p osition . To a solution of TsOH (1 mg) in anhydrous
CH2Cl2 (20 mL) at 0 °C under N2 was added dropwise a
solution of diazo compound in anhydrous CH2Cl2 (5 mL). The
reaction was complete within 30 min as indicated by TLC. The
solvent was removed by evaporation, and the residue was
purified by flash column chromatography to give cis- and trans-
R-aryl-â-enamino esters as a mixture.
Da ta for 2,6-d i-ter t-bu tyl-4-m eth ylp h en yl 2-d ia zo-3-(N-
tosyl)a m in o-3-p h en ylp r op a n oa te (15): yield 71% (condi-
tion B); mp 160 °C dec; IR 3422.5, 2092.6, 1637.5, 1335.0 cm-1
;
1H NMR (200 MHz, CDCl3) δ 1.22 (d, 18H), 2.29 (s, 3H), 2.41
(s, 3H), 5.35 (d, J ) 6.2 Hz, 1H), 5.68 (d, J ) 6.2 Hz, 1H), 7.07
(s, 2H), 7.25-7.28 (m, 7H), 7.75 (d, J ) 8.2 Hz, 2H); 13C NMR
(50 MHz, CDCl3) δ 21.40, 21.46, 31.37, 34.96, 35.04, 53.63,
61.81, 126.10, 126.87, 126.90, 128.43, 128.79, 129.74, 134.74,
136.42, 137.93, 142.19, 142.40, 143.80, 165.36; MS (FAB) m/z
554 [(M + Li)+, 18], 528 (30), 397 (8), 157 (17), 115 (100). Anal.
Calcd for C31H37N3O4S: C, 67.98; H, 6.81; N, 7.67. Found: C,
67.80; H, 7.04; N, 7.43.
Data for (Z)-3-ph en yl-4-(N-tosyl)am in obu tan -2-on e (28):
mp 147-148 °C; IR 3451, 2924, 1640, 1594 cm-1; 1H NMR (200
MHz, CDCl3) δ 2.06 (s, 3H), 2.43 (s, 3H), 6.99 (d, J ) 9.0 Hz,
1H), 7.14-7.42 (m, 7H), 7.75 (d, J ) 8.2 Hz, 1H), 11.74 (d, J
) 10.6 Hz, 1H); 13C NMR (50 MHz, CDCl3) δ 21.49, 29.66,
118.68, 126.70, 127.63, 128.55, 129.39, 129.64, 129.90, 129.99,
130.12, 137.10, 137.35, 139.06, 144.41, 201.29; MS (EI) m/z
315 (M+, 58), 251 (6), 160 (50), 84 (92), 43 (100). Anal. Calcd
for C17H17NO3S: C, 64.74; H, 5.43; N, 4.44. Found: C, 64.30;
H, 5.38; N, 4.19.
N,N-Diben zyl 2-Dia zo-3-(N-tosyl)a m in o-3-p h en ylp r o-
p a n a m id e (25). Reaction condition B was followed, except 20
mol % DBU was used in this case: yield 54%; yellow solid;
mp 113 °C dec; IR 3244.5, 2061.6, 1619.7, 1414.0 cm-1 1H
;
NMR (200 MHz, CDCl3) δ 2.40 (s, 3H), 4.27 (d, 2H), 5.40 (d, J
) 8.2 Hz, 1H), 5.55 (d, J ) 8.2 Hz, 1H), 6.96-7.30 (m, 17H),
7.74 (d, J ) 8.2 Hz, 2H); 13C NMR (50 MHz, CDCl3) δ 21.49,
49.78, 50.53, 126.24, 127.13, 127.26, 127.63, 127.99, 128.63,
128.64, 129.62, 135.74, 137.39, 138.38, 143.42, 165.94; MS
(FAB) m/z 531 [(M + Li)+, 22], 505 (57), 157 (14), 115(100), 91
(61). Anal. Calcd for C30H28N4O3S: C, 68.68; H, 5.38; N, 10.68.
Found: C, 68.62; H, 5.37; N, 10.52.
Da ta for (Z)-N,N-d iben zyl 3-(N-tosyl)a m in o-3-p h en yl-
p r op en a m id e (29): mp 114-117 °C; IR 3461.8, 1610.1, 1575.5
1
cm-1; H NMR (200 MHz, CDCl3) δ 2.40 (s, 3H), 4.38 (s, 2H),
3-[2′-Dia zo-3′-(N-tosyl)a m in o-3′-p h en ylp r op ion yl]-2-ox-
a zolid in on e (26). Butyllithium (1 mL, 1.1 M in hexane, 1.1
mmol) was added dropwise to diisopropylamine (1.2 mmol) in
anhydrous THF (6 mL) at -78 °C, under a nitrogen atmo-
sphere. After the solution was stirred for about 15 min at the
same temperature, 2-oxazolidinone diazoacetamide (155 mg,
1 mmol) in anhydrous THF (6 mL) was added dropwise. The
solution was stirred for 30 min, and then a solution of N-tosyl
benzaldimine 7a (0.285 g, 1.1 mmol) in anhydrous THF (6 mL)
was slowly added at -78 °C. The mixture was stirred until
TLC indicated no starting diazoacetamide left (about 2 h). The
reaction mixture was quenched with saturated aqueous NH4-
Cl solution at the same temperature, and the mixture was
extracted with CH2Cl2. The combined organic layers were dried
over anhydrous Na2SO4. Purification by column chromatog-
raphy with silica gel gave 0.332 g of a yellow solid (86%): mp
85 °C dec; IR 3451, 2104, 1776, 1642 cm-1; 1H NMR (200 MHz,
CDCl3) δ 2.40 (s, 3H), 3.74-3.90 (m, 2H), 4.27-4.36 (m, 2H),
5.50 (d, J ) 8.0 Hz, 1H), 6.04 (d, J ) 7.6 Hz, 1H), 7.25-7.28
(m, 7H), 7.25-7.28 (m, 7H), 7.74 (d, J ) 7.8 Hz, 2H); 13C NMR
(50 MHz, CDCl3) δ 21.43, 43.26, 54.86, 62.79, 126.34, 127.16,
128.27, 128.75, 129.58, 136.78, 137.01, 143.70, 152.64, 162.87;
MS (FAB) m/z 421 [(M + Li)+, 18], 395 (7), 157 (20), 115 (100),
49 (31). Anal. Calcd for C19H18N4O5S: C, 55.07; H, 4.38; N,
13.51. Found: C, 55.38; H, 4.72; N, 12.95.
4.59 (s, 2H), 5.45 (s, 1H), 6.99-7.46 (m, 19H), 12.37 (s, 1H);
13C NMR (50 MHz, CDCl3) δ 21.46, 48.48, 50.12, 101.07,
126.25, 127.32, 127.46, 127.53, 127.63, 128.20, 128.50, 128.61,
128.73, 129.06, 129.99, 134.99, 135.96, 136.78, 136.81, 143.36,
153.76, 168.24; MS (EI) m/z 496 (M+, 41), 341 (100), 196 (97),
149 (99), 91 (80). Anal. Calcd for C30H28N2O3S: C, 72.56; H,
5.68; N, 5.64. Found: C, 72.01; H, 5.74; N, 5.25.
Da ta for (Z)-3-[3′-(N-tosyl)a m in o-3′-p h en ylp r op en oyl]-
2-oxa zolid in on e (30): mp 103-105 °C; IR 3448, 2922, 2852,
1
1781, 1625 cm-1; H NMR (200 MHz, CDCl3) δ 2.44 (s, 3H),
4.04 (t, J ) 8.0 Hz, 3H), 4.40 (t, J ) 8.0 Hz, 3H), 6.65 (s, 3H),
7.15-7.45 (m, 9H), 11.40 (s, 3H); 13C NMR (50 MHz, CDCl3) δ
21.54, 29.63, 42.36, 61.88, 99.72, 125.92, 127.48, 127.79,
129.08, 129.37, 129.79, 130.65, 133.77, 144.18, 157.16, 166.72;
MS (EI) m/z 386 (M+, 12), 322 (9), 178 (20), 155 (94), 91 (100);
HRMS m/z calcd for C19H18N2O5S 386.0936, found 386.0932.
Ack n ow led gm en t. This project is generously sup-
ported by the Natural Science Foundation of China
(Grant No. 29972002, 20172002, 20225205) and by the
Trans-Century Training Programme Foundation for the
Talents by the Ministry of Education of China. L.X.
thanks the Chunhui Project of the Ministry of Education
of China.
Gen er a l P r oced u r e for Rh 2(OAc)4-Ca ta lyzed Dia zo
Decom p osition of N-Tosyl Dia zok eta m in es. A solution of
N-tosyl diazoketamine (0.5 mmol) in CH2Cl2 (20 mL) was
stirred at 0 °C under N2, and then Rh2(OAc)4 (1 mg) was added.
The reaction mixture was stirred at 0 °C for 10 min. The
solvent was then removed under vacuum, and the residue
purified by chromatography or recrystallization.27
Su p p or tin g In for m a tion Ava ila ble: Relative migratory
1
aptitude data and Hammett plots with σ- and σ+, H and 13C
NMR spectra for compounds 10c, 27, and 30, characterization
data for compounds 5a -p , 10a -n , 11a , and 12a -m , and the
optimized structures for rhodium carbene intermediates. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(27) Characterization data for compounds 5a -p , 10a -n , 11a , and
12a -m have been submitted as Supporting Information.
J O0259818
900 J . Org. Chem., Vol. 68, No. 3, 2003