Regiospecific and stereoselective alkylation of allylic substrates on Pd, Ir, and Rh complexes with chiral phosphite ligands

Full text of DOI:10.1134/S0012500807070026

ISSN 0012-5008, Doklady Chemistry, 2007, Vol. 415, Part 1, pp. 161–166. © Pleiades Publishing, Ltd., 2007.
Original Russian Text © A.A. Kabro, S.E. Lyubimov, N.S. Ikonnikov, S.V. Zheglov, M.G. Maksimova, S.K. Moiseev, V.A. Davankov, K.N. Gavrilov, V.N. Kalinin, 2007, published in
Doklady Akademii Nauk, 2007, Vol. 415, No. 1, pp. 61–66.
CHEMISTRY
Regiospecific and Stereoselective Alkylation
of Allylic Substrates on Pd, Ir, and Rh Complexes
with Chiral Phosphite Ligands
A. A. Kabroa, S. E. Lyubimova, N. S. Ikonnikova, S. V. Zheglovb, M. G. Maksimovab,
S. K. Moiseeva, V. A. Davankova, K. N. Gavrilovb, and V. N. Kalinina
Presented by Academician V.N. Charushin December 17, 2006
Received December 20, 2006
DOI: 10.1134/S0012500807070026
Substitution reactions in allylic systems by different is an important problem to reveal factors affecting the
regio- and stereoselectivity of allylation reactions.
nucleophiles catalyzed by transition metal complexes
have been widely used in recent years in organic syn-
thesis, including asymmetric synthesis [1]. Asymmetri-
cally substituted allylic systems are the most general
and important in synthetic practice. The substitution in
allylic substrates containing, along with a leaving
group, only one substituent at one of the terminal car-
bon atoms of the allylic system may be accompanied by
allylic rearrangement to give, in the general case, a
In this work, we carried out, for the first time, the
comparative study of dimethyl malonate allylation
reactions on Pd, Ir, and Rh complexes with chiral
ligands of phosphite and amidophosphite types and
established the decisive effect of the metal nature on the
regioselectivity of the process: the use of Ir and Rh
catalysis favors the formation of a branched product,
while Pd catalysis assists the production of a linear sub-
mixture of two products: linear and branched. The stitution product. In combination with a particular type
of ligand, Pd- and Ir-catalyzed reactions were shown to
ensure the regiospecific formation of the linear or
branched product, respectively. The main structural
features of chiral ligands facilitating the enhancement
of the stereoselectivity of formation of the branched
product were also revealed.
branched product contains an asymmetric center and,
therefore, can exist as two enantiomers. When com-
plexes with chiral ligands are used as catalysts, the
regio- and stereoisomeric composition of the reaction
products depends on the nature of both the chiral ligand
and the transition metal. In recent years, in addition to
palladium catalysis, catalysis by Rh, Ir, and other met-
als has been often used in allylation reactions [2–4].
The search for new types of chiral ligands, especially
As model reactions, we used the allylation of dime-
thyl malonate with methyl 3-phenylallyl carbonate (1)
and its p-chloro-substituted analogue 2 (Scheme 1) to
phosphite ones, is taking place in parallel. Therefore, it form compounds 3 and 5 or 4 and 6, respectively.
OCOOMe
CH(COOMe)2
*
COOMe
COOMe
CH(COOMe)2
M/L*
+
+
CH2Cl2, 48 h, 20°C
X
BSA, KOAc
X
X
1 (X = H)
2 (X = Cl)
3 (X = H)
4 (X = Cl)
Scheme 1.
5 (X = H)
6 (X = Cl)
In this work, we used chiral ligands of the following
types:
(1) phosphite ligands 7–10 based on the chiral frag-
ment of 1,1'-bis-2-naphthol (BINOL): both bidentate
P,N-ligands containing a cymantrenylimino (7, 8)
(cymantrenyl = η5-(C5H4)Mn(CO)3) or oxazoline frag-
ment (10, complexes 11, 12) and monodentate phos-
phite ligand 9;
a Nesmeyanov Institute of Organoelement Compounds,
Russian Academy of Sciences, ul. Vavilova 28, Moscow,
119991 Russia
b Ryazan State Pedagogical University, ul. Svobody 46,
Ryazan, 390000 Russia
(2) monodentate P*-diamidophosphites (13, com-
plexes 14–16);
161

162
KABRO et al.
Table 1. Allylation of dimethyl malonate with carbonate 1 in the presence of complexes with BINOL-derived ligands
Yield of
3 + 5, %
Ligand
Precatalyst
M/L
3/5
ee 3, %
7.
8.
[Pd(allyl)Cl]2
1/1
1/2
1/1
1/2
1/1
1/2
82
28
83
69
32
92
29/71
15/85
99/1
5(S)
25(R)
0
[Ir(COD)Cl]2
[Rh(COD)Cl]2
O
O
P O
N
80/20
86/14
90/10
15(R)
9(R)
0
Mn(CO)3
[Pd(allyl)Cl]2
[Ir(COD)Cl]2
[Rh(COD)Cl]2
1/1
1/2
1/1
1/2
1/1
1/2
47
98
37
28
32
75
33/67
19/81
96/4
10(R)
20(R)
4(S)
O
O
P O
P O
P O
N
99/1
1(S)
82/18
90/10
7(R)
Mn(CO)3
17(R)
[Ir(COD)Cl]2
[Rh(COD)Cl]2
1/1
1/2
1/1
1/2
79
20
99
13
99/1
93/7
95/5
91/9
6(S)
36(S)
8(S)
9.
CF3
CF3
O
O
23(S)
10.
[Pd(allyl)Cl]2
[Ir(COD)Cl]2
[Rh(COD)Cl]2
1/1
1/2
1/1
1/2
1/1
1/2
60
84
95
70
14
12
97
36
46/54
34/66
99/1
61(S)
17(S)
64(S)
45(S)
0
N
O
O
O
99/1
96/4
96/4
0
11. [Pd(allyl)L]+BF4– (L = 10)
12. [Rh(COD)L]+BF4– (L = 10)
63/37
96/4
59(S)
0
(3) a bidentate P,N-ligand of the diamidophosphite reaction rates. The corresponding results are given in
type with the N donor center in the ferrocenylimino Tables 1 and 2.
fragment (complexes 17, 18);
It is seen from these data that the use of Pd catalysis
(4) bidentate P,N-ligands of the phosphite type
based on phenols with the N donor center in the ferro-
cenylimino fragment (19, complexes 20, 21) or the
oxazoline fragment (complex 22).
leads, depending on the ligand, mainly or exclusively to
the formation of the linear products. In contrast, the use
of Rh and Ir catalysis allows one to obtain the branched
regioisomers (Table 1). Both metals ensure high regio-
selectivity. Iridium is slightly more efficient than rhod-
ium and, in the case of certain ligands (7–10, 13), leads
to the regiospecific formation of only the branched iso-
mer. In this respect, diamidophosphite ligand 13 is the
most striking example: depending on the metal, it per-
mits the regiospecific preparation of the linear isomer on
the palladium catalyst or the branched isomer on the irid-
ium catalyst. This feature is observed for both unsubsti-
tuted allyl carbonate 1 and its Ô-chloro-substituted ana-
logue 2 (Table 2).
A chiral catalyst was used as an individual complex
or was preliminarily obtained by the reaction of a chiral
ligand with a precatalyst, a complex of the correspond-
ing transition metal with an achiral ligand. All reactions
were carried out in an inert atmosphere (argon) for
48 h. The analysis of reaction mixture compositions by
1
TLC and H NMR showed only the presence of com-
pounds 3–6 and the unreacted initial compounds. No
marked amounts of by-products were detected. Thus,
the yield of allylation products 3–6 under these condi-
tions indicated the extent of conversion of the initial
The use of individual palladium complexes 14–18
allylic substrates (1 or 2) and, in the case of incomplete with diamidophosphite ligands results in the regiospe-
conversion, allowed us to semiquantitatively estimate cific preparation of linear products 5 and 6. The size
DOKLADY CHEMISTRY Vol. 415 Part 1 2007

REGIOSPECIFIC AND STEREOSELECTIVE ALKYLATION OF ALLYLIC SUBSTRATES
163
Table 2. Allylation of dimethyl malonate with carbonates 1 and 2
Yield of
Yield of
Complex, ligand
Precatalyst
M/L
3/5
ee 3, %
30(S)
4/6
ee 4, %
3 + 5, %
4 + 6, %
13.
[Pd(allyl)Cl]2
1/1
1/2
1/1
1/2
81
85
97
79
79
59
50
62
10/90
–/100
84
85
72
65
52
62
60
68
–/100
–/100
Ph
N
P
N
O
Ad
[Ir(COD)Cl]2
>99/1
0
99/1
2(S)
0
(Ad = adamantyl)
98/2
4(R)
0
99/1
[Rh(COD)Cl]2 1/1
1/2
78/22
92/8
72/28
89/11
–/100
–/100
11(R)
11(R)
18(R)
14. [Pd(allyl)L2]+BF4– (L = 13)
15. [Pd(allyl)L2]+BF4–
–/100
–/100
Ph
N
P
L =
N
O
16. [Pd(allyl)L2]+BF4–
69
95
–/100
–/100
71
90
–/100
–/100
Ph
N
P
L =
N
O
Me
17. [Pd(allyl)L]+BF4–
–
Ph
N
P
L =
N
O
N
Fc
(Fc = ferrocenyl)
18. [Pd(allyl)L2]+BF4–
Ph
84
–/100
71
–/100
L =
N
P
N
O
N
Fc
[Pd(allyl)Cl]2
1/1
1/2
94
81
38/62
39/61
50(R)
48(R)
69
78
41/59
41/59
80(R)
75(R)
19.
O
O
P O
N
Fc
20. [Pd(allyl)L]+BF4–
(L = 19)
87
50/50
39(R)
62
54/46
56(R)
DOKLADY CHEMISTRY Vol. 415 Part 1 2007

164
KABRO et al.
Table 2. (Contd.)
Yield of
Yield of
Complex, ligand
Precatalyst
M/L
3/5
ee 3, %
33(R)
4/6
ee 4, %
62(R)
3 + 5, %
4 + 6, %
21. [Pd(allyl)L]+BF4–
97
37/63
86
45/55
O
L =
P O
N
O
Fc
22. [Pd(allyl)L]+BF4–
93
54/46
59(S)
N
O
L =
O
P O
O
and nature of the exocyclic substituent in the ligand has
The addition of LiCl to iridium catalysts is known to
no effect on the regioselectivity. Thus, the variation of increase both regio- and stereoselectivity [4]. We used
metal and ligand makes it possible to achieve the this method for the first time for palladium catalysis for
regiospecific formation of any regioisomer.
alkylation of substrate 1 in the presence of Pd complex
22: 1 equiv of LiCl was added, which changed the ratio
of regioisomers 3 : 5 toward the formation of chiral
product 3 (from 54/46 to 63/37), the optical purity of 3
increasing from 59 to 75%.
Thus, the comparative study of the allylation reac-
tions of dimethyl malonate with asymmetrically substi-
tuted allyl carbonates on Pd, Ir, and Rh complexes with
chiral phosphites and amidophosphites revealed the
decisive factors (metal nature, ligand type) that provide
for the regiospecific formation of one of the possible
isomers when allylic substrates allowing the reaction to
proceed with nondegenerate allylic rearrangement are
used. We also found the major structural features of the
chiral ligands that improve the enantioselectivity of for-
mation of the product with an asymmetric center.
When a catalytic species results from ligand
exchange by the reaction of a chiral ligand with a pre-
catalyst, the use of two equivalents of the ligand per
equivalent of the metal (M/L = 1/2) usually improves
the stereoselectivity as compared with the ratio M/L =
1/1. However, this is often accompanied by a decrease
in the reaction rate. The variation of the M/L ratio for
the same ligand can accelerate or retard the reaction
depending on M (Table 1, ligands 7, 8, 10; Table 2,
ligand 13).
The optical yields of the branched products are low
irrespective of the metal and ligand nature and their
ratio. Exceptions are ligand 10 and complex 22, as well
as ligand 19 and complexes 20 and 21, containing
oxazoline and ferrocenylimino fragments, respectively,
which allowed us to obtain the branched allylation
products with satisfactory optical purity (ee 56–80%).
This is observed only for Pd and Ir. Rh catalysis usually
leads to a very low, if any, enantioselectivity of the pro-
cess. Obviously, it is the oxazoline and ferrocenylimino
fragments in the bidentate ligands used that are respon-
sible for the enantioselectivity of the allylation reac-
tion. It should be noted that this does not refer to the
cymantrenylimino fragment (compare the data for
ligands 7, 8, and 19). This is likely due to the difference
in the electronic properties of the ferrocenyl and
cymantrenyl groups (which are close in size) as substit-
uents in the imine fragment and, therefore, to the differ-
ences in the n-donor properties of the imine nitrogen
atom in these ligands.
EXPERIMENTAL
All reactions were carried out in a dry argon atmo-
sphere in anhydrous solvents. Ligands 9 [5], 13 [6], and
19 [7]; complexes [Pd(allyl)Cl]2 [8], [Ir(COD)Cl]2 [8],
[Rh(COD)Cl]2 [9], 14–16 [6], 17 and 18 [10], 20 and 21
[7]; and 22 [11]; and (Sax)- and (Rax)-2-chlorodinaph-
tho[2,1-d:1',2'-f ][1,3,2]dioxaphosphepin [12] and sub-
strates 1 and 2 [13] were obtained by published proce-
dures. N,O-Bis(trimethylsilyl)acetamide (BSA) and
dimethyl malonate (Acros Organics) were used without
additional purification.
General procedure for the synthesis of ligands
based on BINOL. Triethylamine (0.2 mL, 1.4 mmol)
DOKLADY CHEMISTRY Vol. 415 Part 1 2007

REGIOSPECIFIC AND STEREOSELECTIVE ALKYLATION OF ALLYLIC SUBSTRATES
165
and 1.4 mmol of the appropriate iminoalcohol or 69.1 (s, OCH2), 39.0 (s, CH), 25.3 (CH2), 14.4 (s, CH3),
11.4 (s, CH3). MS (EI, m/z (I, %)): 534 (2, [M]+), 408
(15), 332 (63), 268 (51), 219 (47), 162 (100).
For C33H28NO4P anal. calcd. (wt %): C, 74.29;
H, 5.29; N, 2.63.
Found (wt %): C, 74.54; H, 5.20; N, 2.89.
((Sax)-2-[2''-((4'''S)-4'''-sec-Butyl-2'''-oxazolin-
2'''-yl)phenoxy]dinaphtho[2,1-d:1',2'-f][1,3,2]dioxa-
oxazoline were added to a solution of 0.491 g (1.4
mmol) of (Sax)- or (Rax)-2-chlorodinaphtho[2,1-d:1',2'-
f][1,3,2]dioxaphosphepin in 15 mL of toluene with stir-
ring and cooling (ice–water). The reaction mixture was
heated to 110°C and cooled to ambient temperature, the
.
Et3N HCl precipitate was filtered off, and the solvent
was removed in vacuum (40 Torr). Twenty-five millili-
ters of heptane was added to the residue, the mixture
was heated to 98°C, the resulting solution was sepa-
rated from the insoluble residue by decantation, and the
solvent was removed in vacuum (40 Torr). The products
were dried in vacuum (1 Torr) for 1.5 h.
phosphepin-P,N)(p-allyl)palladium(II)
tetrafluo-
roborate (11). Yellow powder. Mp 123–125°C
(decomp.). 31P NMR (CDCl3, δP, ppm): 148.9 (s), 44%
and 147.9 (s), 56%. MS (electrospray, m/z (I, %)): 681
(100, [M–BF4]+), 640 (1, [M–allyl–BF4]+).
(Rax,2''S,3''S)-2-[3''-Methyl-2''-{(cymantre-
nylidene)amino}pentyloxy]dinaphtho[2,1-d:1',2'-
f][1,3,2]dioxaphosphepin (7). Pale yellow powder,
For C36H33BF4NO4PPd anal. calcd. (wt %): C,
56.31; H, 4.33.
31
yield 69%. Mp 78–80°C. P NMR (CDCl3, δP, ppm):
Found (wt %): C, 56.64; H, 4.44.
142.4 (s). 13C NMR (CDCl3, δC, ppm, JC,P, Hz): 223.8
((Sax)-2-[2''-((4'''S)-4'''-sec-Butyl-2'''-oxazolin-
2'''-yl)phenoxy]dinaphtho[2,1-d:1',2'-f][1,3,2]diox-
aphosphepin-P,N)([1,2:5,6-h-(1,5-cyclooctadi-
ene)]rhodium(I) tetrafluoroborate (12). Dark yellow
powder. Mp 146–148°C (decomp.). 31P NMR (CDCl3,
2
(s, CO), 155.1 (s, C=N), 148.4 (d, JC,P = 6.1), 147.3,
132.3, 131.2, 130.7, 130.1, 129.7, 128.1, 128.0, 126.7,
126.6, 125.9, 124.8, 124.6, 123.8, 122.4, 121.6, 121.3,
120.6 (ëAr), 94.9 (s, CCp ipso), 85.6, 83.6, 82.8, 81.8 (s,
2
δP, ppm, 1JP,Rh, Hz): 149.2 (d, 291.6). MS (electrospray,
m/z (I, %)): 745 (100, [M–BF4]+).
CCp), 75.5 (s, NCH), 66.1 (d, JC,P = 5.3, OCH2), 36.2
(s, CH), 24.9 (CH2), 15.4 (s, CH3), 10.8 (s, CH3). IR
(CHCl3, ν, cm–1): 2026 and 1947 (CO). MS (electro-
spray, m/z (I, %)): 646 (46, [M]+), 547 (4), 314 (100).
For C35H29MnNO6P anal. calcd. (wt %): C, 65.12;
H, 4.53; N, 2.17.
For C41H40BF4NO4PRh anal. calcd. (wt %): C,
59.23; H, 4.85.
Found (wt %): C, 59.43; H, 4.72.
Catalytic allylation of dimethyl malonate with
carbonates 1 and 2 (general procedure). A solution
of a precatalyst (0.01 mmol) and a corresponding
ligand (0.02–0.04 mmol) in 5 mL of CH2Cl2 was stirred
for 40 min (or 0.02 mmol of individual complexes were
Found (wt %): C, 65.42; H, 4.39; N, 2.39.
(Sax,2''S,3''S)-2-[3''-Methyl-2''-{(cymantre-
nylidene)amino}pentyloxy]dinaphtho[2,1-d:1',2'-
f][1,3,2]dioxaphosphepin (8). Pale yellow powder,
yield 72%. Mp 71–73°C. 31P NMR (CDCl3, δP, ppm): dissolved in 5 mL of CH2Cl2). Allylic substrate 1 or 2
144.6 (s). 13C NMR (CDCl3, δC, ppm, JC,P, Hz): 223.8
(0.5 mmol) was added, and the mixture was stirred for
2
20 min; then dimethyl malonate (0.75 mmol), BSA
(s, CO), 155.2 (s, C=N), 148.3 (d, JC,P = 6.0), 147.4,
(0.75 mmol), and KOAc (2 mg) were added. The reac-
132.6, 132.4, 131.3, 130.8, 129.7, 128.3, 128.0, 126.7,
tion mixture was stirred for 48 h at 20°C. The solvent
126.6, 125.9, 124.8, 124.6, 123.8, 122.5, 121.7, 121.4,
was evaporated and the products were isolated by chro-
120.8 (CAr), 95.1 (s, CCp ipso), 85.7, 83.5, 82.9, 81.8 (s,
matography on a SiO2 column (200 × 17 mm, hexane–
CCp), 75.7 (s, NCH), 66.4 (d, 2JC,P = 5.1, OCH2), 36.8 (s,
EtOAc (9 : 1) as eluent). The solvent was removed and
CH), 25.0 (CH2), 15.3 (s, CH3), 10.2 (s, CH3). IR
the fraction composed of a mixture of regioisomers 3
(CHCl3, ν, cm–1): 2028 and 1950 (CO). MS (electro-
1
and 5 (or 4 and 6) was analyzed by H NMR (CDCl3,
spray, m/z (I, %)): 646 (39, [M]+), 547 (9), 314 (100).
300 MHz) to determine the isomer ratio. The optical
For C35H29MnNO6P anal. calcd. (wt %): C, 65.12;
H, 4.53; N, 2.17.
yield of compounds 3 and 4 was determined by HPLC
(Diacel Chiracel OJ-H chiral column for 3 and Diacel
Chiracel OD-H for 4).
Found (wt %): C, 65.28; H, 4.59; N, 2.02.
(Sax)-2-[2''-((4'''S)-4'''-sec-Butyl-2'''-oxazolin-
2'''-yl)phenoxy]dinaphtho[2,1-d:1',2'-f][1,3,2]dioxa-
phosphepin (10). White powder, yield 74%. Mp 65–
67°C. 31P NMR (CDCl3, δP, ppm): 142.7 (s). 13C NMR
(CDCl3, δC, ppm, JC,P, Hz): 160.1 (s, C=N), 149.9 (d,
ACKNOWLEDGMENTS
This work was supported by the Presidium of the
RAS (Basic Research Program no. 8).
2JC,P = 6.4), 146.7 (d, 2JC,P = 3.3), 133.5, 133.0, 132.8,
132.1, 131.3, 130.8, 130.5, 130.4, 128.8, 128.3, 128.0,
127.8, 127.2, 126.9, 126.7, 126.5, 125.2, 124.3, 123.5,
122.8, 121.7, 120.6, 119.8, 118.3 (CAr), 71.4 (s, NCH),
REFERENCES
1. Trost, B.M. and Crawley, M.L., Chem. Rev., 2003,
vol. 103, pp. 2921–2943.
DOKLADY CHEMISTRY Vol. 415 Part 1 2007

166
KABRO et al.
2. Polet, D. and Alexakis, A., Org. Lett., 2005, vol. 7, no. 8,
(Metal Complex Analysis in Organic Synthesis), Mos-
cow: Khimiya, 1999.
pp. 1621–1624.
3. Hayashi, T., Okada, A., Suzuka, T., and Kawatsura, M.,
9. Giordano, G. and Crabtree, R.H., Inorg. Synth., 1990,
Org. Lett., 2003, vol. 5, no. 10, pp. 1713–1715.
vol. 28, pp. 88–89.
4. Alexakis, A. and Polet, D., Org. Lett., 2004, vol. 6,
10. Tsarev, V.N., Lyubimov, S.E., Bondarev, O.G., et al.,
no. 20, pp. 3529–3532.
Eur. J. Org. Chem., 2005, pp. 2097–2105.
5. Gavrilov, K.N., Lyubimov, S.E., Zheglov, S.V., et al.,
11. Bondarev, O.G., Lyubimov, S.E., Shiryaev, A.A., et al.,
Tetrahedron: Asymmetry, 2002, vol. 13, no. 15,
pp. 1587–1588.
J. Mol. Catal. A: Chem., 2005, vol. 231, pp. 255–260.
6. Tsarev, V.N., Lyubimov, S.E., Shiryaev, A.A., et al., Eur.
J. Org. Chem., 2004, no. 10, pp. 2214–2222.
12. Francio, G., Arena, C.G., Faraone, F., et al., Eur. J. Inorg.
Chem., 1999, p. 1219.
7. Gavrilov, K.N., Tsarev, V.N., Maximova, M.G., et al.,
J. Mol. Catal. A: Chem., 2006, vol. 259, pp. 267–274.
8. Dzhemilev, U.M., Popod’ko, N.R., and Kozlova, E.V., 13. Lehmann, J. and Lloyd-Jones, G.C., Tetrahedron, 1995,
Metallokompleksnyi kataliz v organicheskom sinteze
vol. 51, no. 32, pp. 8863–8874.
DOKLADY CHEMISTRY Vol. 415 Part 1 2007