Modular furanoside diphosphite ligands for Pd-catalyzed asymmetric allylic substitution reactions: Scope and limitations

Article abstract of DOI:10.1002/adsc.200505013

We have synthesized a library of furanoside diphosphite ligands for the Pd-catalyzed allylic substitution reactions of acyclic and cyclic allylic esters. The library has been designed to rapidly screen the ligands to uncover their important structural features and to determine the scope of diphosphite ligands in these catalytic reactions. After the systematic variation of the sugar backbone, the substituent at C-5 and the phosphite moieties, the diphosphite ligand 4c was found to be optimal in the Pd-catalyzed asymmetric allylic substitution of hindered (S1) and unhindered (S2-S5) substrates, yielding high activities [TOFs up to > 3000 mol x (mol x h)^-1] and enantioselectivities (ees up to 99%). In addition, the screening of the library enabled us to find other suitable ligands for hindered disubstituted linear substrate S1 (ligands 1b-d, g and 4b, d, g) and for unhindered cyclic substrates S3-S5 (ligands 6c and 7c).

Full text of DOI:10.1002/adsc.200505013

FULL PAPERS
Modular Furanoside Diphosphite Ligands for Pd-Catalyzed
Asymmetric Allylic Substitution Reactions: Scope and
Limitations
Montserrat Di e´ guez,* Oscar P a` mies,* Carmen Claver
Universitat Rovira i Virgili, Departament de Química Física i Inorganica, C/ Marcel lí Domingo s/n, 43007 Tarragona,
Spain
Fax: (þ34)-977-559-563, e-mail: montserrat.dieguez@urv.net, oscar.pamies@urv.net
Received: January 21, 2005; Accepted: March 21, 2005
Abstract: We have synthesized a library of furanoside (S2–S5) substrates, yielding high activities [TOFs up
À1
diphosphite ligands for the Pd-catalyzed allylic substi- to >3000 molÂ(molÂh) ] and enantioselectivities
tution reactions of acyclic and cyclic allylic esters. The (ees up to 99%). In addition, the screening of the li-
library has been designed to rapidly screen the ligands brary enabled us to find other suitable ligands for hin-
to uncover their important structural features and to dered disubstituted linear substrate S1 (ligands 1b–d,
determine the scope of diphosphite ligands in these g and 4b, d, g) and for unhindered cyclic substrates
catalytic reactions. After the systematic variation of S3–S5 (ligands 6c and 7c).
the sugar backbone, the substituent at C-5 and the
phosphite moieties, the diphosphite ligand 4c was Keywords: allylic substitution; asymmetric catalysis;
found to be optimal in the Pd-catalyzed asymmetric CÀC bond formation, P ligands, palladium, phosphite
allylic substitution of hindered (S1) and unhindered ligands
Introduction
of the possibilities offered by diphosphites as new li-
gands for Pd-catalyzed allylic substitution reactions is
Palladium-catalyzed asymmetric allylic substitution is a still needed. To fully investigate these possibilities, in
versatile, widely used process in organic synthesis for the this paper we expand the previous study of 2001 to other
enantioselective formation of carbon-carbon and car- diphosphite furanoside ligands (Figure 2) and also ex-
bon-heteroatom bonds. In recent years, many chiral li-
gands, mainly P- and N-containing ligands, which pos-
sess either C or C symmetry, have provided high enan-
2
1
[
1]
tiomeric excesses. However, these catalytic systems
generally have two important limitations (Figure 1).
Firstly, they have a high substrate specificity (i.e., high
ees are obtained in hindered disubstituted linear sub-
strates while low ees are obtained in cyclic substrates
and unhindered linear substrates, or vice versa). Second-
ly, the reaction rates are usually low [i.e., TOFs¼
À1 [1]
1
00 molÂ(molÂh) ]. In this context, the research
into faster and more versatile ligand systems from read-
ily available simple starting materials is nowadays of
great importance in this reaction.
In 2001, we reported the first diphosphite furanoside
ligands used for Pd-catalyzed asymmetric allylic alkyla-
tion. Results were only reported for the alkylationof 1,3-
[
2]
diphenyl-3-acetoxyprop-1-ene (S1). In addition to the
high enantiomeric excesses (ees up to 95%), the reaction
rates were extremely fast [TOFs up to 1200 molÂ
Figure 1. Summary of the best results with substrates S1–S3
with three of the most representative ligands developed for
À1
(
molÂh) ] in simple non-optimized reaction condi-
[2]
tions. Despite this success, the use of other diphosphite the Pd-catalyzed allylic substitution reactions (reactions usu-
ligands has not yet been reported and a systematic study ally carried out with 2–4 mol % of Pd).
Adv. Synth. Catal. 2005, 347, 1257–1266
DOI: 10.1002/adsc.200505013
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1257

FULL PAPERS
Pd-Catalyzed Asymmetric Allylic Substitution Reactions
Figure 2. Diphosphite ligand library.
tend the study of this family of ligands to other types of steric properties: rac-1,3-diphenyl-3-acetoxyprop-1-
substrates with different steric properties (Figure 3). ene (S1; widely used as a model substrate) and rac-1,3-
To do this, we have taken advantage of the high mod- dimethyl-3-acetoxyprop-1-ene (S2). Different nucleo-
[
3]
ularity of these ligands and have synthesized and philes were tested. In all the cases, the catalysts were
screened a library of 33 furanoside diphosphite ligands generated in situ from p-allyl-palladium chloride dimer
3
with the same underlying structure (Figure 2). The syn- [PdCl(h -C H )] and the corresponding ligand.
3
5
2
thesis and screening of the library were performed in
parallel using two parallel reactors each equipped with
24 different positions. With this library we fully investi-
gated the effects of systematically varying the configura-
tions at C-3 and C-5of the ligand backbone (1–6), differ-
ent substituents at C-5 [H, Me and CH OTBDPS
2
ð1Þ
(
TBDPS¼tert-butyldiphenylsilyl); ligands 1 and 2, 3–
6
and 7, respectively], and different substituents/config-
urations in the biaryl phosphite moieties (a–h). By care-
fully selecting these elements, we achieved high enantio-
selectivities and activities in different substrate types.
Results and Discussions
Allylic Alkylation of rac-1,3-Diphenyl-3-acetoxyprop-
-ene (S1)using Dimethyl Malonate as Nucleophile
Equation (1)]
1
[
Allylic Substitution of Disubstituted Linear Substrates
In this section, we report the use of the chiral diphos- We determined the optimal reaction conditions by con-
phite ligands 1–7 in the Pd-catalyzed allylic substitution ducting a series of experiments in which the solvent and
[
Equation (1)] of two linear substrates with different the ligand-to-palladium ratio were varied.
We first studied the effect of the solvent. Four solvents
tetrahydrofuran (THF), dichloromethane (DCM), tol-
[
uene and dimethylformamide (DMF)] and seven li-
gands (1c–7c) were tested. The best combination of ac-
tivity and enantioselectivity was achieved with dichloro-
methane (Figures 4 and 5). The enantiomeric excesses
obtained with tetrahydrofuran and toluene were compa-
rable to those of dichloromethane, but the activities
were the lowest. On the other hand, dimethylformamide
Figure 3. Substrates tested with diphosphite ligands 1a–h to yielded high relative conversions, similar to those of di-
7a–h.
chloromethane, but the ees were the lowest of the four
1258
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1257–1266

Montserrat Di e´ guez et al.
FULL PAPERS
Figure 4. Conversions of 8 after 5 minutes using ligands 1c–7c in different solvents (THF, DCM, toluene and DMF).
Figure 5. Enantiomeric excess of 8 obtained using ligands 1c–7c in four solvents. The positive numbers refer to the formation
of the S-isomer in excess and the negative numbers refer to the formation of the R-isomer.
solvents. In conclusion, we choose dichloromethane as
our solvent.
Table 1. Selected results for the Pd-catalyzed allylic alkyla-
tion of S1 using ligands 1c, 3c and 4c. Effect of the ligand-
to-palladium ratio.
We next studied the effect of varying the ligand-to-
palladium ratio. The conversion and enantioselectivity
when dichloromethane was used as a solvent are shown
in Table 1 (similar trends were observed for the other
solvents). These results show that excess ligand is not
needed to obtain high enantioselectivities.
In conclusion, the trade-off between enantioselectivi-
ties and reaction rates was optimum with dichlorome-
thane anda ligand-to-palladium ratioof 1.1:1. These op-
timal conditions were then used to test the catalytic per-
formance of the complete series of ligands. The activities
and enantiomeric excesses obtained are plotted in Fig-
ure 6. Note the high activities obtained with these cata-
lytic systems (e.g., 100% conversion is usually reached
after 5 minutes) and the excellent enantioselectivities
obtained with ligands 1b–d, g and 4b–d, g (ees up to
[a]
[
b]
[c]
Entry Ligand L/Pd % Conversion (t/min)
% ee
1
2
3
4
5
6
7
8
9
1c
1c
1c
3c
3c
3c
4c
4c
4c
0.9
1.1
2
0.9
1.1
2
0.9
1.1
2
100 (30)
100 (30)
100 (30)
100 (5)
100 (5)
100 (5)
100 (5)
100 (5)
100 (5)
97 (S)
97 (S)
97 (S)
45 (R)
45 (R)
44 (R)
98 (S)
98 (S)
98 (S)
[
a]
All reactions were run at room temperature: 0.5 mol %
3
[
PdCl(h -C H )] ; CH Cl as solvent.
3 5 2 2 2
[b]
1
Conversion percentage of acetate 8 determined by H-
NMR.
[c]
98%).
Enantiomeric excesses determined by HPLC on a Chiral-
cel-OD column. Absolute configuration given in parenth-
eses.
From Figure 6 (left) we can conclude that activities
were affected by the substituent at C-5 and the phos-
Adv. Synth. Catal. 2005, 347, 1257–1266
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1259

FULL PAPERS
Pd-Catalyzed Asymmetric Allylic Substitution Reactions
Figure 6. Left: Conversion of 8 after 5 minutes using the library presented in Figure 2. Right: Enantiomeric excess of 8 using the
different ligands. The positive numbers refer to the formation of the S-isomer in excess and the negative numbers refer to the
formation of the R-isomer.
phite moiety. When the substituent at C-5 was Me or sidering ligands 3–6, which contemplate the four possi-
CH OTBDPS (ligands 3–7), activities were better ble combinations of stereocenters C-3 and C-5, we found
2
than when the substituent was H (ligands 1 and 2). Ac- a cooperative effect between C-3 and C-5. [The opportu-
tivities were also better when the ligand contained bulky nity provided by this ligand library to study all the possi-
substituents at the ortho positions and electron-donat- ble combinations of stereocenters C-3 and C-5 comple-
[2]
ing substituents at the para position of the biphenyl moi- ments the preliminary results (Ref. )that erroneously in-
eties (i.e., bꢀc>d>a). dicated that enantioselectivities were mainly affected by
Like activities, enantioselectivities (Figure 6, right) the configuration of C-3.] Specifically, the best results
were affected by the substituent at C-5 and the phos- were obtained when the configuration of C-3 was S
phite moiety but they were also affected by the configu- and the configuration of C-5 was R (ligands 4b–d, g).
ration of carbon atoms C-3 and C-5 and the configura- The excellent ees obtained with ligands 1, which do
tions of the biaryl moieties. Since each of these param- not have a stereocenter in C-5, suggest that the spatial
eters interacted with the others, we could not assign a arrangement around C-5 adopted by these ligands in
certain effect on the enantiomeric excess to each indivi- the Pd-allyl intermediate must be similar to that of li-
dual parameter. However, we found that enantioselec- gands 4. A second cooperative effect between the ster-
tivities were best (ees up to 98%) with the following spe- eogenic centers of the ligand backbone and the configu-
cific combination: xylo- (1b–d, g) and gluco-furanoside ration of the binaphthyl phosphite moieties was ob-
(
4b–d, g) ligands containing bulky substituents at the or- served when binaphthyl-based ligands (e–h) were
tho positions of the biaryl moieties. Moreover, by con- used. This cooperative effect, and the previously ob-
1260
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1257–1266

Montserrat Di e´ guez et al.
FULL PAPERS
served cooperative effect between the backbone stereo- as for the alkylation reaction, though the CIP descriptor
centers C-3 and C-5, control the enantioselectivity.
In addition to the effect of structural parameters on
enantioselectivity, reaction parameters can also be con-
was inverted due to the change in priority of the groups.
trolled to further improve enantioselectivity. In this Allylic Alkylation of rac-1,3-Dimethyl-3-acetoxyprop-
case, ees were further improved by lowering the reaction 1-ene (S2)using Dimethyl Malonate as Nucleophile
temperature to 58C. For example, enantioselectivities
up to>99% were obtained with ligands 1c and 4c while We also screened the ligand library in the allylic alkyla-
activity was almost unaffected (100% conversion after tion of the linear substrate S2 [Equation (1)]. This sub-
35 and 7 minutes using ligands 1c and 4c, respectively). strate is less sterically demanding than the previously
We also performed the reaction at a low catalyst concen- used substrate S1. Enantioselectivity for S2 is therefore
tration (S1/Pd¼2000) using ligand 4c. In this way, excel- more difficult to control than with hindered substrates
lent enantioselectivity [98% (S) ee] and activity [100% such as S1. To obtain high ees, the ligand must create a
conversion after 40 min at room temperature, TOF> small chiral pocket (the chiral cavity where the allyl is
À1
3
000 mol (mol·h) ] were obtained.
embedded) around the metal center, mainly because
of the presence of less sterically methyl syn substitu-
[
1]
ents.
Allylic Amination of rac-1,3-Diphenyl-3-acetoxyprop-
-ene (S1)using Benzylamine as Nucleophile
The results of using the diphosphite ligand library un-
der the optimized conditions are summarized in Table 3.
Activities followed the same trends as for the alkylation
1
We also evaluated the ligand library in the allylic substi- of substrate S1. The best reaction rates were obtained
tution process of S1 using benzylamine as nucleophile with ligands 3b and c to 7b and c, which contained Me
[
Equation (1)]. This situation is different from that de- or CH OTBDPS groups at C-5, and bulky tert-butyl sub-
2
scribed before for the alkylation of S1 because a new stituents at the ortho positions and electron-donating
stereogenic CÀN bond is created rather than a CÀC substituents at the para position of the biphenyl moieties
À1
bond. The most remarkable results are shown in Table 2. [TOF>300 mol (mol·h) ]. However, the general
In general, they follow the same trends as for the allylic trends that controlled enantioselectivity were different
alkylation of S1. However, the enantiomeric excesses from those that controlled S1, and only ligand 4c
were higher (ees up to 99% at room temperature). Al- reached high enantioselectivities for this substrate (en-
though, as expected, the activities were lower than in try 15). This result is amongst the best reported for this
[
4]
the alkylation reaction of S1, they were much higher type of unhindered substrates. Note also that these li-
[
1c]
than those obtained with other homodonor ligands.
Again, the catalyst precursor containing the diphosphite
ligands 1b–d, g and 4b–d, g provided the best trade-off
Table 3. Selected results for the Pd-catalyzed allylic substitu-
between activity and enantioselectivity (entries 1–4 and tion of S2.[
a]
7). The stereoselectivity of the amination was the same
[
b]
[c]
Entry
Ligand
% Conversion (t/min)
% ee
1
2
3
4
5
6
7
8
9
1c
2c
3c
4a
4b
4c
4d
4e
4f
4g
4h
5c
6c
7c
4c
100 (30)
100 (30)
100 (10)
100 (30)
100 (10)
100 (10)
90 (30)
54 (30)
61 (30)
51 (30)
49 (30)
100 (10)
100 (10)
100 (15)
12 (60)
59 (R)
52 (S)
45 (S)
4 (R)
67 (R)
78 (R)
71 (R)
8 (R)
Table 2. Selected results for the Pd-catalyzed allylic amina-
tion of S1.
[a]
[
b]
[c]
Entry
Ligand
% Conversion (t/h)
% ee
1
2
3
4
5
6
7
8
9
1b
1c
1d
1g
2c
3c
4c
5c
6c
7c
100 (1)
92 (1)
95 (1)
53 (1)
81 (1)
100 (0.75)
100 (0.75)
100 (0.75)
100 (0.75)
67 (1)
98 (R)
98 (R)
98 (R)
98 (R)
16 (S)
61 (S)
99 (R)
13 (S)
55 (S)
81 (R)
4 (S)
10
11
12
13
14[
72 (R)
21 (S)
34 (S)
43 (S)
58 (R)
85 (R)
d]
1
0
15
[
a]
[a]
All reactions were run at room temperature: 1 mol %
All reactions were run at room temperature: 1 mol %
3
3
[
PdCl(h -C H )] ; CH Cl as solvent.
[PdCl(h -C H )] ; CH Cl as solvent.
3
5
2
2
2
3
5
2
2
2
[
[
b]
1
[b]
[c]
1
Conversion percentage of 9 determined by H-NMR.
Enantiomeric excesses determined by HPLC on a Chiral-
cel-OJ column. Absolute configuration given in parenth-
eses.
Conversion percentage of 10 determined by H-NMR.
Enantiomeric excesses determined by GC. Absolute con-
figuration given in parentheses.
c]
[d]
Reaction carried out at À208C.
Adv. Synth. Catal. 2005, 347, 1257–1266
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1261

FULL PAPERS
Pd-Catalyzed Asymmetric Allylic Substitution Reactions
gands provided higher activities [TOFs up to >300 molÂ
Again, enantioselectivities (Figure 7,right) were high-
À1
(
molÂh) ] than those for other successful ligands for ly affected by the substituent at C-5, the configuration of
[4]
this substrate.
carbon atoms C-3 and C-5 and the phosphite moieties.
However, the effect of these parameters was different
from those observed in the alkylation of S1 and S2.
Enantioselectivity was best with ligands 4c, 6c and 7c.
To obtain high enantioselectivities, therefore, the ligand
Allylic Alkylation of Cyclic Substrates
As for the unhindered substrate S2, enantioselectivity in must have a substituent at C-5, bulky tert-butyl groups at
cyclic substrates is difficult to control. Although high both the ortho and para positions and an S configuration
enantioselective catalysts have been developed for cy- at C-3. These results also indicate that the configuration
clic substrates, they generally provide low enantiocon- at C-5 controls the senseof the enantioselectivity. There-
[
1]
trol in hindered linear substrates. Presumably this fore, the R enantiomer of 11 was obtained using ligand
lack of generality is derived from different ligand struc- 6c and the S enantiomer of 11 was obtained using ligand
tural requirements for both types of substrates (see 4c, both with excellent enantioselectivities and activi-
above). Therefore, the development of enantioselective ties. For these two ligands, enantioselectivities up to
catalysts for both unhindered (S2–S5, Figure 3) and hin- 95% were obtained by lowering the reaction tempera-
[
5]
dered substrates (S1) is still a challenge.
ture to À208C (24% conversion after 12 hours using li-
This section reports that the chiral diphosphite ligands gand 4c and 26% conversion after 12 hours using ligand
–7 applied in the previous section to the Pd-catalyzed 6c). These results are amongst the best reported for cy-
allylic alkylation of linear substrates, can also be used for clic substrates.
cyclic substrates. In this case, three cyclicsubstrates were
tested [Equation (2)]: rac-3-acetoxycyclohexene (S3;
1
[
4a,6]
which is widely used as a model substrate), rac-3-acet- Allylic Alkylation of rac-3-Acetoxycyclopentene (S4)
oxycyclopentene (S4) and rac-3-acetoxycycloheptene and rac-3-Acetoxycycloheptene (S5)[Equation (2)]
(
S5).
To further study the potential of ligand 4c (the only li-
gand that reaches excellent values of enantioselectivi-
ties for both linear and cyclic substrates), we also tested
it in the allylic alkylation of 5- and 7-membered cyclic
substrates S4 and S5, respectively (Table 4). Interesting-
ly, for these sterically undemanding substrates, high re-
ð2Þ
À1
action rates [TOFs up to >200 molÂ(molÂh) ] and
enantioselectivities (ees up to 96%) were also obtained.
Allylic Substitution of Monosubstituted Linear
Substrates
Allylic Alkylation of rac-3-Acetoxycyclohexene (S3)
Equation (2)]
[
Encouraged by the excellent results obtained for disub-
Preliminary investigations into the solvent effect and li- stituted linear substrates and for cyclic substrates, we ex-
gand-to-palladium ratio provided the same trends as amined the regio- and stereoselective allylic alkylation
those with the previously tested linear substrate S1. of 1-phenylallyl acetate (S6) and 1-(1-naphthyl)allyl
The trade-off between enantioselectivities and reaction acetate (S7) with dimethyl malonate. For both sub-
rates was optimum with dichloromethane and a ligand- strates, the development of highly regio- and enantiose-
[
7]
to-palladium ratio of 1.1:1. The results of using the di- lective Pd-catalysts is still a challenge. The results ob-
phosphite ligand library under the optimized conditions tained with the diphosphite ligand library are summar-
are shown in Figure 7.
ized in Table 5. Unfortunately, the regioselectivity for
Activities (Figure 7, left) followed the already ob- the branched products was not high in this case. Howev-
served trends for the alkylation of substrates S1 and er, high enantioselectivities can be obtained by increas-
S2. The best reaction rates, therefore, were obtained ing the size of the substrate substituent. Ligand 4c pro-
with ligands 3b and c to 7b and c, which contain Me or vided 95% ee for substrate S7 but only 29% ee for sub-
CH OTBDPS groups at C-5, and bulky tert-butyl sub- strate S6 (entry 10 vs. 6). Note also, the high activities
2
À1
stituents at the ortho positions and electron-donating [TOFs up to>600 molÂ(molÂh) ] observed for
substituents at the para position of the biphenyl moieties these substrates.
À1
[
TOF>200 mol (mol·h) ]. As expected, the activities
were lower than in the alkylation reaction of S1.
[
4a,6]
1262
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1257–1266

Montserrat Di e´ guez et al.
FULL PAPERS
Figure 7. Left: Conversion of 11 after 30 minutes. Enantiomeric excess of 11 using the library presented in Figure 2. Right:
Enantiomeric excess of 11 using the different ligands. The positive numbers refer to the formation of the R-isomer in excess,
and the negative numbers refer to the formation of the S-isomer.
Table 4. Pd-catalyzed allylic alkylation of S4 and S5 using li- tion reactions of several substrate types. By carefully de-
[
a]
gand 4c.
signing this library we were able to systematically inves-
tigate the effects of varying the configurations at C-3 and
C-5 of the ligand backbone, the type of substituents at
[
b]
[c]
Entry Substrate T [8C] % Conversion (t/min)
% ee
1
2
3
4
S4
S5
S4
S5
20
20
100 (30)
100 (30)
12 (90)
13 (90)
68 (R) C-5 (H, Me and CH OTBDPS) and the type of substitu-
2
87 (S)
81 (S)
96 (R)
ents/configurations in the biaryl phosphite moieties.
Particularly, for the hindered disubstituted acyclic sub-
strate S1, we found that activities and enantioselectivi-
ties were highest with glucofuranoside ligands 4b and
c, which have an S configuration of C-3 and an R config-
uration of C-5, a methyl substituent at C-5 and bulky
tert-butyl substituents in the ortho positions of the bi-
À20
À20
[
[
[
a]
All reactions were run at room temperature: 0.5 mol %
PdCl(h -C H )] ; CH Cl as solvent.
Conversion percentage of 12 and 13 determined by H-
NMR.
Enantiomeric excesses determined by H-NMR using phenyl phosphite moieties. By screening the library we
Eu(hfc)3.
3
[
3 5 2 2 2
b]
c]
1
1
were able to find other suitable ligands for this substrate
S1 (ligands 1b–d, g and 4d, g). For the unhindered cyclic
substrate S3–S5, activities followed the same trends as
those for the alkylation of substrate S1, but enantiose-
lectivities were highest with ligands 4c, 6c and 7c. Re-
Conclusion
A library of readily available diphosphite ligands has markably, ligands 4c and 6c offer the S and R enantiom-
been synthesized for the Pd-catalyzed allylic substitu- ers of the product, respectively, in high enantiomeric ex-
Adv. Synth. Catal. 2005, 347, 1257–1266
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1263

FULL PAPERS
Pd-Catalyzed Asymmetric Allylic Substitution Reactions
[a]
Table 5. Selected results for the Pd-catalyzed allylic alkylation of S6 and S7.
[
b]
[b]
[c]
Entry
Substrate
Ligand
% Conversion (t/min)
Branched to linear ratio (b/l)
% ee
1
2
3
4
5
6
7
8
9
S6
S6
S6
S6
S6
S6
S6
S6
S6
S7
1a
1b
1c
2c
3c
4c
5c
6c
7c
4c
42 (30)
100 (30)
100 (30)
26 (30)
100 (5)
100 (5)
100 (5)
100 (5)
80 (5)
7/93
24/76
14/86
29/71
30/70
24/76
29/71
24/76
17/83
34/66
15 (S)
23 (S)
7 (S)
33 (R)
5 (R)
29 (S)
16 (S)
9 (R)
19 (S)
95 (S)
1
0
100 (10)
[
[
[
a]
3
All reactions were run at room temperature: 1 mol % [PdCl(h -C H )] ; CH Cl as solvent.
3
5
2
2
2
b]
1
Conversion percentage and linear-to-branched ratio determined by H-NMR.
Enantiomeric excesses determined by HPLC on a Chiralcel-OJ column. Absolute configuration given in parentheses.
c]
3
,5-Bis[(3,3’-bis-tert-butyl-5,5’-dimethoxy-1,1’-
cesses. For the monosubstituted acyclic substrateS6 and
S7, this ligand library proved to be inadequate in terms
of regioselectivities. However, we obtained high enan-
biphenyl-2,2’-diyl)phosphite]-1,2-O-isopropylidene-6-
tert-butyldiphenylsilyl-glucofuranose (7b)
[13]
tioselectivities by increasing the size of the substrate Phosphorochloridite (2.2 mmol) produced in situ was dis-
substituent. In this way, ligand 4c provided 95% ee for solved in toluene (5 mL) and pyridine (0.36 mL, 4.6 mmol)
substrate S7.
was added. 1,2-O-Isopropylidene-6-tert-butyldiphenylsilyl-
To sum up, by carefully selecting the ligand compo- glucofuranose (0.46 g, 1 mmol) was azeotropically dried with
nents we obtained high enantioselectivities (ees up to toluene (3Â1 mL) and then dissolved in toluene (10 mL), to
9
(
9%) and high activities [TOFs up to >3000 molÂ
which pyridine (0.18 mL, 2.3 mmol) has been added. The diol
À1
molÂh) ] in substrates with different steric proper- solution was transferred slowly over 30 minutes at room tem-
perature to the solution of phosphorochloridite. The reaction
mixture was stirred overnight at 808C, and the pyridine salts
were removed by filtration. Evaporation of the solvent gave
a white foam, which was purified by flash chromatography (tol-
ties. We highlight the excellent activities and enantiose-
lectivities obtained with ligand 4c for disubstituted acy-
clic and cyclic substrates with dimethyl malonate or ben-
zylamine as nucleophiles. This is therefore an exception-
al ligand that competes favorably with the few other li-
gands that also provide high ees for both types of
uene, R ¼0.42) to produce a white powder. Yield: 0.52 g
f
1
3
(
42%); P NMR: d¼145.3 (d, J ¼38 Hz), 147.3 (bd, J
¼
P-P
P-P
1
[5]
38 Hz); H NMR: d¼0.85 (s, 9H, CH , t-Bu-Si), 1.13 (s, 3H,
3
substrates. Mechanistic and computational studies
are currently under way.
CH ), 1.32 (s, 9H, CH , t-Bu), 1.33 (s, 9H, CH , t-Bu), 1.36 (s,
3
3
3
1
8H, CH , t-Bu), 1.40 (s, 28H, CH , t-Bu), 1.43 (s, 3H, CH ),
3
3
3
1
.53 (s, 9H, CH , t-Bu), 1.57 (s, 9H, CH , t-Bu), 3.81 (s, 3H,
3
3
OMe), 3.84 (b, 9H, OMe), 4.10 (m, 2H, H-6’, H-6), 4.24 (d,
3
1
4
7
H, H-2, J ¼3.6 Hz), 4.55 (m, 1H, H-5), 4.85 (m, 1H, H-
H-H
Experimental Section
3
), 5.01 (m, 1H, H-3), 5.53 (d, 1H, H-1, J ¼3.6 Hz), 7.0–
H-H
1
3
.8 (m, 18H, CH¼); C NMR: d¼26.9 (CH , t-Bu-Si), 27.0
3
General Considerations
(CH
3
), 27.2 (CH
), 31.4 (CH
3
3
, t-Bu), 31.6 (CH , t-Bu), 31.7
3
(
CH , t-Bu), 31.8 (CH , t-Bu), 31.9 (CH , t-Bu), 34.8 (C, t-
3
3
3
All reactions were carried out using standard Schlenk tech-
niques under an atmosphere of argon. Solvents were purified
and dried by standard procedures. Diphosphite ligands 1–6
were prepared using previously described methods. 1,2-O-
Isopropylidene-6-tert-butyldiphenylsilyl-glucofuranose was
Bu), 34.9 (C, t-Bu), 35.3 (C, t-Bu), 35.4 (C, t-Bu), 35.6 (C, t-
Bu), 35.7 (C, t-Bu), 35.9 (C, t-Bu), 55.3 (OMe), 55.4 (OMe),
55.5 (OMe), 64.8 (C-6), 73.2 (b, C-5), 76.3 (C-3), 77.8 (C-4),
[
3,8]
8
5.6 (C-2), 104.9 (C-1), 112.8 (CMe ), 124.3 (CH¼), 124.4
2
[
9]
(CH¼), 124.5 (CH¼), 124.7 (d, CH¼, J ¼6.1 Hz), 125.1
C-P
prepared as previously reported. Racemic substrates S1–S7
[
10–12] 1
13
1
(CH¼), 126.1 (d, CH¼, J ¼3.2 Hz), 126.7 (d, CH¼, J
¼
¼
were prepared as previously reported.
H, C{ H}, and
C-P
C-P
31
1
¼
¼
¼
4
.2 Hz), 127.6 (CH ), 127.9 (CH ), 129.2 (d, CH , J
P{ H} NMR spectra were recorded using a 400 MHz spec-
C-P
1
trometer. Chemical shifts are relative to that of SiMe ( H
8.0 Hz), 134.5 (C), 134.8 (C), 134.9 (C), 136.0 (C), 136.2
(CH
(C), 142.7 (C), 142.9 (C), 145.1 (C), 145.6 (C), 145.8 (C),
4
13
31
¼), 136.3 (CH¼), 136.5 (C), 138.1 (C), 142.0 (C), 142.4
and C) as internal standard or H PO ( P) as external stand-
3
4
ard.
1264
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1257–1266

Montserrat Di e´ guez et al.
FULL PAPERS
1
56.0 (C), 156.7 (C), 156.8 (C), 156.9 (C); anal. calcd. (%) for Typical Procedure for the Allylic Amination of rac-
C H O P Si: C 67.30, H 7.20; found: C 67.56, H 7.11.
69
88 14
2
1,3-Diphenyl-3-acetoxyprop-1-ene (S1)
3
A degassed solution of [PdCl(h -C H )] (1.8 mg, 0.005 mmol)
3
5
2
and the diphosphite ligand (0.011 mmol) in dichloromethane
0.5 mL) was stirred for 30 min. Subsequently, a solution of
rac-8 (126 mg, 0.5 mmol) in dichloromethane (1.5 mL) and
benzylamine (131 mL, 1.5 mmol) were added. The reaction
mixture was stirred at room temperature. After 1 hour the re-
(
3
,5-Bis[(3,3’,5,5’-tetra-tert-butyl-1,1’-biphenyl-2,2’-
diyl)phosphite]-1,2-O-isopropylidene-6-tert-
butyldiphenylsilyl-glucofuranose (7c).
action mixture was diluted with Et O (5 mL) and saturated
[13]
2
Treatment of phosphrochloridite (2.2 mmol) produced in
situ and 1,2-O-isopropylidene-6-tert-butyldiphenylsilyl-gluco-
furanose (0.46 g, 1 mmol), as described for compound 7b, af-
forded diphosphite 7c, which was purified by flash chromatog-
raphy (toluene) to produce a white powder. Yield: 0.64 g
aqueous NH Cl solution (25 mL) was added. The mixture
4
was extracted with Et O (3Â10 mL) and the extract dried
2
over MgSO . The solvent was removed and conversion was
4
1
measured by H-NMR. To determine the ee by HPLC (Chiral-
cel-OJ, 13% 2-propanol/hexane, flow 0.5 mL/min), a sample
3
1
(
6
48%); P NMR: d¼147.2 (d, J ¼62 Hz), 148.1 (bd, J
¼
P-P
P-P
was filtered over silica using 10% Et O/hexane mixture as
1
2
2 Hz); H NMR: d¼0.83 (s, 9H, CH , t-Bu-Si), 1.12 (s, 3H,
3
the eluent.
CH ), 1.30 (s, 9H, CH , t-Bu), 1.31 (s, 9H, CH , t-Bu), 1.32 (s,
3
3
3
1
8H, CH , t-Bu), 1.35 (s, 9H, CH , t-Bu), 1.36 (s, 9H, CH , t-
3
3
3
Bu), 1.39 (s, 3H, CH ), 1.49 (s, 9H, CH , t-Bu), 1.54 (s, 9H,
3
3
2
3
CH , t-Bu), 4.05 (dd, 1H, H-6’, J ¼9.3 Hz, J ¼1.5 Hz),
3
H-H
H-H
2
3
)
, 4.14 (dd, 1H, H-6, J ¼9.3 Hz, J ¼1.2 Hz), 4.27 (d,
H-H
H-H
3
Typical Procedure for the Allylic Alkylation of Cyclic
Substrates S3–S5
1
H, H-2, J ¼3.3 Hz), 4.65 (m, 1H, H-5), 4.85 (dd, 1H, H-4,
H-H
2
3
JH-H ¼9.3 Hz, J ¼2.7 Hz), 4.96 (m, 1H, H-3), 5.49 (d, 1H,
H-H
3
3
H-1, J ¼3.3 Hz), 6.97 (t, 2H, CH¼, J ¼5.4 Hz), 7.1–7.4
3
H-H
H-H
A
degassed solution of [PdCl(h -C H )]
(0.9 mg,
3
5
2
3
(
(
2
3
m, 12H, CH¼), 7.51 (d, 2H, CH¼, J ¼5.4 Hz), ¼), 7.65
H-H
0
.0025 mmol) and the diphosphite ligand (0.0055 mmol) in di-
3
3
13
dd, 2H, CH¼, J ¼5.1 Hz, J ¼1.2 Hz); C NMR: d¼
H-H
H-H
chloromethane (0.5 mL) was stirred for 30 min. Subsequently,
a solution of substrate (0.5 mmol) in dichloromethane
6.7 (CH , t-Bu-Si) 26.9 (CH ), 27.0 (CH ), 31.4 (CH , t-Bu),
3 3 3 3
1.5 (CH , t-Bu), 31.6 (CH , t-Bu), 31.7 (CH , t-Bu), 31.8
3
3
3
(
1.5 mL), dimethyl malonate (171 mL, 1.5 mmol), N,O-bis(tri-
(
CH , t-Bu), 31.9 (CH , t-Bu), 34.7 (C, t-Bu), 34.8 (C, t-Bu),
3
3
methylsilyl)acetamide (370 mL, 1.5 mmol) and a pinch of
KOAc were added. The reaction mixture was stirred at room
temperature. After 30 min the reaction mixture was diluted
with Et O (5 mL) and aqueous saturated NH Cl solution
(
(
S3, conversion and enantiomeric excess were determined by
GC using an FS-Cyclodex b-I/P 25 m column, internal diame-
ter 0.2 mm, film thickness 0.33 mm, carrier gas: 100 kPa He,
FID detector). For substrates S4 and S5, conversion was deter-
3
3
1
4.9 (C, t-Bu), 35.3 (C, t-Bu), 35.4 (C, t-Bu), 35.6 (C, t-Bu),
5.7 (C, t-Bu), 35.8 (C, t-Bu), 63.8 (C-6), 71.6 (d, C-5, JC-P
6.2 Hz), 76.5 (C-3), 77.4 (C-4), 84.7 (C-2), 105.1 (C-1), 112.2
¼
2
4
(
CMe ), 124.1 (CH¼), 124.3 (CH¼), 124.4 (CH¼), 124.5 (d,
2
25 mL) was added. The mixture was extracted with Et O
2
CH¼, J ¼6.1 Hz), 125.5 (CH¼), 126.5 (d, CH¼, J
¼
C-P
C-P
3Â10 mL) and the extract dried over MgSO . For substrate
4
3
.8 Hz), 126.8 (d, CH¼, J ¼4.5 Hz), 127.6 (CH¼), 127.8
C-P
(
1
1
CH¼), 129.4 (d, CH¼, J ¼8.3 Hz), 132.5 (C), 132.8 (C),
C-P
33.9 (C), 134.0 (C), 135.8 (CH¼), 135.9 (CH¼), 136.4 (C),
38.1 (C), 140.0 (C), 140.4 (C), 140.7 (C), 140.8 (C), 145.1
(
C), 145.6 (C), 145.8 (C), 146.0 (C), 146.7 (C), 146.8 (C),
1
mined by GC and enantiomeric excess was determined by H
NMR using Eu(hfc)3.
1
7
46.9 (C), 150.0 (C); anal. calcd. (%) for C H O P Si: C
81 112 10 2
2.83, H 8.45; found: C 73.01, H 8.31.
Typical Procedure for the Allylic Alkylation of
Disubstituted Linear Substrates S1 and S2.
Typical Procedure for the Allylic Alkylation of
Monosubstituted Linear Substrates S6 and S7
3
3
A degassed solutionof [PdCl(h -C H )] and the diphosphite li-
A degassed solution of [PdCl(h -C H )] (1.8 mg, 0.005 mmol)
3
5
2
3
5
2
gand (1.1 equivs. to Pd) in dichloromethane (0.5 mL) was stir-
red for 30 min. Subsequently, a solution of substrate
and the diphosphite ligand (0.011 mmol) in dichloromethane
(0.5 mL) was stirred for 30 min. Subsequently, a solution of
substrate (0.5 mmol) in dichloromethane (1.5 mL), dimethyl
malonate (171 mL, 1.5 mmol), N,O-bis(trimethylsilyl)-acet-
amide (370 mL, 1.5 mmol) and a pinch of KOAc were added.
The reaction mixture was stirred at room temperature. After
the desired reaction time, the reaction mixture was diluted
with Et O (5 mL) and aqueous saturated NH Cl solution
(0.5 mmol) in dichloromethane (1.5 mL), dimethyl malonate
(171 mL, 1.5 mmol), N,O-bis(trimethylsilyl)acetamide
(370 mL, 1.5 mmol) and a pinch of KOAc were added. The re-
action mixture was stirred at room temperature. After the de-
sired reaction time, the reaction mixture was diluted with Et O
2
(
5 mL) and saturated aqueous NH Cl solution (25 mL) was
4
2
4
added. The mixture was extracted with Et O (3Â10 mL) and
(25 mL) was added. The mixture was extracted with Et O
2
2
the extract dried over MgSO . For substrate S1, conversion
(3Â10 mL) and the extract dried over MgSO . The solvent
4
4
1
was measured by H NMR and enantiomeric excess was deter-
was removed and conversion and regioselectivity were meas-
1
mined by HPLC (Chiralcel-OD, 0.5% 2-propanol/hexane, flow
ured by H NMR. To determine the ee by HPLC (Chiralcel-
0
.5 mL/min). For substrate S2, conversion and enantiomeric
OJ, 3% 2-propanol/hexane, flow 0.7 mL/min), a sample was fil-
tered over basic alumina using dichloromethane as the eluent.
excess were determined by GC.
Adv. Synth. Catal. 2005, 347, 1257–1266
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1265

FULL PAPERS
Pd-Catalyzed Asymmetric Allylic Substitution Reactions
[
4a]
Acknowledgements
[5] For successful applications, see: a) Ref.
b) G. Helm-
chen, A. Pfaltz, Acc. Chem. Res. 2000, 33, 336; c) B. M.
Trost, J. D. Oslob, J. Am. Chem. Soc. 1999, 121, 3057.
[6] See also: a) D. A. Evans, K. R. Campos, J. S. Tedrow,
F. E. Michael, M. R. Gagn e´ , J. Am. Chem. Soc. 2000,
This work was supported by the Spanish Ministerio de Educa-
ci o´ n, Cultura y Deporte (BQU2001-0656), the Spanish Minis-
terio de Ciencia y Tecnologia (Ramon y Cajal fellowship to
O. P.)and the Generalitat de Catalunya (Distinction to M. D.
and C. C.).
1
22, 7905; b) B. M. Trost, R. C. Bunt, J. Am. Chem.
Soc. 1994, 116, 4089.
[
7] For recent successful applications of Pd-catalysts, see:
a) S.-L.You, X.- Z. Zhu, Y.-M. Luo, X.-L. Hou, L.-X.
Dai, J. Am. Chem. Soc. 2001, 123, 7471; b) R. Pr e´ t oˆ t,
A. Pfaltz, Angew. Chem. Int. Ed. 1998, 37, 323; c) R. Hil-
graf, A. Pfaltz, Synlett 1999, 1814.
References and Notes
[
1] For recent reviews, see: a) J: Tsuji, Palladium Reagents
and Catalysis, Innovations in Organic Synthesis, Wiley,
New York, 1995; b) B. M. Trost, D. L. van Vranken,
Chem. Rev. 1996, 96, 395; c) M. Johannsen, K. A. Jorgen-
sen, Chem. Rev. 1998, 98, 1869; d) A. Pfaltz, M. Lautens,
in: Comprehensive Asymmetric Catalysis, (Eds.: E. N. Ja-
cobsen, A. Pfaltz, H. Yamamoto), Springer-Verlag, Ber-
lin, 1999, Vol. 2, Chapter 24; e) B: M. Trost, M. L. Craw-
ley, Chem. Rev. 2003, 103, 2921.
[8] a) G. J. H. Buisman, M. E. Martin, E. J. Vos, A. Kloot-
wijk, P. C. J. Kamer, P. W. N. M. van Leeuwen, Tetrahe-
dron: Asymmetry 1995, #6#8, 719; b) O. P a` mies, G. Net,
A. Ruiz, C. Claver, Tetrahedron: Asymmetry 1999,
#10#11, 1097; c) O. P a` mies, M. Di e´ guez, G. Net, A.
Ruiz, C. Claver, Tetrahedron: Asymmetry 2000, 11,
4377; d) M. Di e´ guez, A. Ruiz, C. Claver, Tetrahedron:
Asymmetry 2001, 12, 2895; e) M. Di e´ guez, A. Ruiz, C.
Claver, J. Org. Chem. 2002, 67, 3796; f) M. Di e´ guez, O.
P a` mies, A. Ruiz, C. Claver, New J. Chem. 2002, 26, 827.
[9] K. Akerfeldt, P. A. Barlett, Carbohydrate. Res. 1986, 158,
137.
[10] P. R. Auburn, P. B. Mackenzie, B. Bosnich, J. Am. Chem.
Soc. 1985, 107, 2033.
[11] C. Jia, P. Mꢁller, H. Mimoun, J. Mol. Cat. A: Chem. 1995,
101, 127.
[12] J. Lehman, G: C. Lloyd-Jones, Tetrahedron 1995, 51,
8863.
[13] Phosphorochloridites are easily prepared in one step
from the corresponding bisphenol as described in:
G. J. H. Buisman, P. C. J. Kamer, P. W. N. M. van Leeu-
wen, Tetrahedron: Asymmetry 1993, 4, 1625.
[
[
2] M. Di e´ guez, S. Jansat, M. Gomez, A. Ruiz, G. Muller, C.
Claver, Chem. Commun. 2001, 1132.
3] Ligands 1–7 were synthesized in parallel from the corre-
sponding diols in one step following the general proce-
dure described in: M. Di e´ guez, O. P a` mies, A. Ruiz, S.
Castill o´ n, C. Claver, Chem. Eur. J. 2001, 7, 3086.
[
4] For successful applications, see: a) P. Dierkes, S. Rande-
chul, L. Barloy, A. De Cian, J. Fischer, P. C. J. Kamer,
P. W. N. M. van Leeuwen, J. A. Osborn, Angew. Chem.
Int. Ed. 1998, 37, 3116; b) B. M. Trost, A. C. Krueger,
R. C. Bunt, J. Zambrano, J. Am. Chem. Soc. 1996, 118,
6
1
520; c) B. Wiese, G. Helmchen, Tetrahedron Lett.
998, 39, 5727.
1266
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1257–1266