Phosphinooxazolines derived from 3-amino-1,2-diols: Highly efficient modular P-N ligands

Article abstract of DOI:10.1002/adsc.200600599

A family of chiral phosphinooxazolines (12a-e) derived from modular, enantiopure β-amino alcohols has been prepared, and their palladium complexes have been used as chiral mediators in the asymmetric allylic alkylation reaction. The oxazoline moiety in 12 contains a C-4 aryl and a C-5 alkoxymethyl substituent that can be independently optimized for high catalytic activity and enantioselectivity. A methoxymethyl substituent at C-5 has beenfound to provide the best results in terms of enantioselectivity and activity in the alkylation of a diverse family of allylic substrates under both thermal and microwave-assisted activation. The palladium-phos- phinooxazoline complexes described in this work are remarkably robust, as the enantioselectivity recorded in the asymmetric allylic alkylation remains essentially unchanged in the temperature range between 20 and 130°C. An unprecedented reversal in enantioselectivity has been observed between 1, 3-diphe-nylallyl and 1,3-dimethylallyl alkylation substrates, and the origin of this behavior has beeh explained by means of ONIOM QM/MM calculations.

Full text of DOI:10.1002/adsc.200600599

FULL PAPERS
DOI: 10.1002/adsc.200600599
Phosphinooxazolines Derived from 3-Amino-1,2-diols:
Highly Efficient Modular P-N Ligands
Dana Popa,a Cristina Puigjaner,b Montserrat Gómez,c Jordi Benet-Buchholz,a
Anton Vidal-Ferran,a,d,* and Miquel A. Pericàsa,b,*
a
Institute of Chemical Research of Catalonia (ICIQ), Avinguda Països Catalans 16, 43007 Tarragona, Spain
Phone: (+34)-977-920-200; fax: (+34)-977-920-222; e-mail: avidal@iciq.es or mapericas@iciq.es
Departament de Química Orgànica, Universitat de Barcelona, 08028 Barcelona, Spain
UniversitØ Paul Sabatier, Laboratoire HØtØrochimie Fondamentale et AppliquØe, UMR CNRS 5069, 118, route de
b
c
Narbonne, 31062 Toulouse cedex 9, France
ICREA (Institució Catalana de Recerca i Estudis AvanÅats), Passeig Lluís Companya 23, 08010 Barcelona, Spain
d
Received: November 17, 2007; Revised: July 10, 2007
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: A family of chiral phosphinooxazolines phinooxazoline complexes described in this work are
(12a–e) derived from modular, enantiopure b-amino remarkably robust, as the enantioselectivity recorded
alcohols has been prepared, and their palladium in the asymmetric allylic alkylation remains essen-
complexes have been used as chiral mediators in the tially unchanged in the temperature range between
asymmetric allylic alkylation reaction. The oxazoline 20 and 1308C. An unprecedented reversal in enan-
moiety in 12 contains a C-4 aryl and a C-5 alkoxy- tioselectivity has been observed between 1,3-diphe-
methyl substituent that can be independently opti- nylallyl and 1,3-dimethylallyl alkylation substrates,
mized for high catalytic activity and enantioselectivi- and the origin of this behavior has been explained by
ty. A methoxymethyl substituent at C-5 has been means of ONIOM QM/MM calculations.
found to provide the best results in terms of enantio-
selectivity and activity in the alkylation of a diverse
family of allylic substrates under both thermal and Keywords: allylic alkylation; asymmetric catalysis;
microwave-assisted activation. The palladium-phos- microwave-induced alkylation; ONIOM calculations
Introduction
Although C2-symmetric ligands have dominated the
field of asymmetric catalysis because they facilitate
Phosphinooxazolines 1 and 2 belong to a non-C2-sym- the analysis of the catalyst-substrate interactions, for
metric class of ligands which combine the characteris- reactions where the two enantiomeric products arise
tics of a “soft” phosphorus donor group with p-ac- from the regioselectivity of the process, C1 ligands
ceptor properties and a “hard” nitrogen sigma donor possessing two different coordinating atoms should
group.[1] The oxazoline and phosphorus groups are allow a more effective regiocontrol.[5] The palladium-
separated by a linker which corresponded, in the catalyzed asymmetric allylic alkylation of symmetrical
seminal work by Pfaltz,[2] Helmchen[3] and Williams,[4] p-allyl ligands (3) falls into this category, since the
to a 1,2-disubstituted benzene ring, 2.
enantioselectivity of the reaction depends on the abil-
ity of the ligand in inducing a different reactivity in
the two terminal allyl carbons (Scheme 1).
Following the initial studies by Pfaltz[2], Helmchen[3]
and Williams,[4] a great variety of phosphinooxazo-
lines has been developed and studied in the asymmet-
ric allylic alkylation with excellent results.[1b,e–g,6] How-
ever, modular b-amino alcohols, suitable for the fine-
tuning of catalytic properties, have never been used
as chiral building blocks for the preparation of phos-
phinooxazolines.
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2265

Dana Popa et al.
FULL PAPERS
Results and Discussion
Synthesis of the Phosphinooxazolines and their
Palladium Complexes
The preparation of phosphinooxazolines 12 and of
their p-allyl palladium complexes 13 was planned
from Sharpless epoxy ethers 7 through well docu-
mented synthetic transformations (Scheme 2).
Scheme 1.
Synthetic, yet enantiopure amino alcohols 4 arising
Both the ready availability of enantiopure epoxy al-
from Sharpless epoxides 5 present several elements of cohols through Sharpless epoxidation and the ease for
structural diversity and have been developed by our selective derivatization of the primary alcohol in
group as very efficient mediators for the asymmetric these structures render this approach highly attractive
addition of diethylzinc onto carbonyl compounds[7] in terms of the introduction of molecular diversity.
and imines[8] and enantioselective transfer hydrogena- Moreover, the substitution scheme in 7 should pro-
tion.[9] Their derived oxazaborolidines have been used vide access to trans-4,5 disubstitued modular phosphi-
in the enantioselective borane reduction of prochiral nooxazolines, whose synthesis and catalytic properties
ketones[10] and the derived bisoxazolines in the asym- have remained practically unexplored.[2,13]
metric allylic alkylation.[11] In all these cases, the fine-
Amino alcohols 8 were readily prepared by aminol-
tuning of catalytic properties is greatly facilitated by ysis of epoxy ethers 7 under pressure either by micro-
the complete modular nature of 4 which allows inde- wave irradiation or by simple thermal activation.[14]
pendent optimization of the different structural ele- With microwave activation, the ring opening proceed-
ments. Along with this strategy, we report in this ed satisfactorily just by dissolving 7a (R=Me), 7b
paper the preparation of a new family of modular (R=CH2Ph) and 7e [R=(CH2)2OMe] in 25% aque-
phosphinooxazolines 6, incorporating an alkoxymeth- ous ammonia (50 equivs.) and irradiating the corre-
yl substituent at C-5 of the oxazoline ring and the per- sponding solution for 8 min at 1258C. Epoxy ethers 7c
formance of their p-allylpalladium complexes as cata- and 7d, which bear bulkier R groups, were not soluble
lysts for the asymmetric alkylation of a diverse family in aqueous ammonia and a co-solvent (i-PrOH) was
of allylic substrates.
required to facilitate the reaction. In those cases,
Whereas known phosphinooxazoline libraries incor- 1 equivalent of lithium perchlorate was also added to
porate modularity at the phosphorus atom or at the the reaction mixture in order to accelerate the ring-
opening reaction. Full conversion was achieved in this
manner after not more than 20 min. It is worth noting
that, under thermal conditions, these reactions re-
quired 5 h for complete conversion (entries 4 and 6,
Table 1). Independent of the ring-opening method
which was used, all the reactions took place stereo-
specifically and with complete regiocontrol.
The crude amino alcohols 8 were simply dried with
MgSO4 (in dichloromethane solution) and directly
substituent of the carbon atom next to nitrogen, the submitted to the next step without further purifica-
study of the effects of the substituents at C-5 of the tion. The N-acylation step was carried out using 2-flu-
oxazoline moiety (such as the R2OCH2 group in 6) re- orobenzoyl chloride as the limiting reagent and tri-
mains unexplored. Besides its primary effect on cata- ethylamine as the auxiliary base in dry tetrahydrofur-
lytic activity and on enantioselectivity, the alkoxy- an. At this stage, the intermediate hydroxy benza-
methyl substituent in 6 offers the additional interest mides 9 were purified and characterized. The overall
of allowing the heterogeneization of the phosphinoox- yields (referred to the ring opening and benzoylation
azoline ligand onto insoluble organic resins. From this steps) range from 53% to 80% and have been sum-
perspective, the optimization of the R2OCH2- moiety marized in Table 1.
in 6 is of paramount importance: On one hand, the
Conversion of 9a–e into oxazolines 11a–e was car-
primary hydroxy group in these structures is the obvi- ried out through a two-step sequence[11] involving: ac-
ous point for polymer anchoring with minimal pertur- tivation of the hydroxy group in 9 as a mesylate
bation of the reaction site.[12] On the other hand, the (MsCl, NEt3 as auxiliary base) and base-induced cycli-
optimal OR2 group will dictate the optimal nature for zation (5% KOH/MeOH) of the intermediate mesy-
the resin onto which 6 can be anchored.
lates 10a–e through an SN2 mechanism. Due to the
limited stability of mesylates 10, these compounds
were not stored, and the base-induced cyclization to-
2266
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
Scheme 2. Preparation of phosphinooxazolines 12a–e and of their p-allyl Pd complexes 13a–e from epoxy ethers 7a–e.
wards 11 was carried out within a few hours after
The nucleophilic displacement of fluoride from 11
their preparation. The overall activation-cyclization with potassium diphenylphosphide proceeded un-
sequence took place in a straightforward way, and eventfully in THFat À208C. Quenching of the reac-
high overall yields for the two steps (68–80%, see tion mixture with sodium sulfate decahydrate fol-
Table 2) were observed in all cases. The relative trans lowed by a short filtration through deoxygenated SiO2
stereochemistry of the oxazoline substituents could be afforded pure phosphinooxazolines 12 in 62–97%
confirmed by means of a NOESY experiment on 11b, yield (Table 2). 31P NMR data were in agreement with
where a cross-peak indicating NOE between the CH- the proposed structure, as a singlet at around À3 ppm
N and the CH2 units could be observed (See Support- was observed for the diphenylphosphino group along
ing Information).
the whole series of compounds.
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2267

Dana Popa et al.
FULL PAPERS
Table 1. Ring-opening and acylation of epoxy ethers 7a–e.
Entry Epoxy ether
Aminolysis conditions
2-Fluorobenz-
amide
Overall
yield
1
2
3
4
5
7a, R=Me
Microwave irradiation: 1258C, 8 min
Microwave irradiation: 1258C, 8 min
Microwave irradiation: 1258C, i-PrOH, 8 min
Thermal conditions: 1008C, i-PrOH, 5 h; 1 equiv. LiClO4
Microwave irradiation: 1308C, i-PrOH, 20 min; 1 equiv.
LiClO4,
9a, R=Me
53%
69%
61%
80%
67%
7b, R=CH2Ph
7c, R=CHPh2
7c, R=CHPh2
7d, R=CPh3
9b, R=CH2Ph
9c, R=CHPh2
9c, R=CHPh2
9d, R=CPh3
6
7
7d, R=CPh3
7e,
R=(CH2)2OMe
Thermal conditions: 1008C, i-PrOH, 5 h; 1 equiv. LiClO4
Microwave irradiation: 1258C, 8 min
9d, R=CPh3
9e,
R=(CH2)2OMe
60%
59%
Table 2. Transformation of hydroxybenzamides 9 into the palladium complexes 13.
Entry
Hydroxy amide
Overall yield steps iii and iv
Yield of step v
Yield of step vi
Palladium complex
1
2
3
4
5
9a, R=Me
76%
68%
70%
80%
68%
93%
97%
62%
76%
85%
76%
65%.
64%
63%
63%
13a, R=Me
9b, R=CH2Ph
9c, R=CHPh2
9d, R=CPh3
13b, R=CH2Ph
13c, R=CHPh2
13d, R=CPh3
9e, R=(CH2)2OMe
13e, R=(CH2)2OMe
The phosphinooxazolines were finally converted the palladium sigma complex followed by a final coor-
into the ionic p-allylpalladium complexes immediately dination of the olefin to the palladium center.
after isolation. The complexes were prepared by re-
A detailed NMR study (see Supporting Informa-
acting the palladium dimer [Pd
AHCTREUNG
appropriate phosphinooxazoline 12 in ethanol in the occur in solution as endo-isomers. In addition, the
presence of NH4PF6.[11] The metal complexes of gen- solid state structure of one of the complexes could be
eral formula [PdACHTREUNG
(h3-allyl)(L)]PF6 13 precipitated from determined by X-ray crystallography (Figure 1), since
the solution at 48C and could be isolated as crystal- crystals of anion exchanged complex 13a could be
line materials. Yields of isolated material are given in grown in methanol.[19] Selected geometrical bond dis-
Table 2.
As far as the structure of 13a–e in solution is con-
tances for the allyl moiety have been summarized in
cerned, it is well known that palladium h3-allyl com-
plexes normally exist as two isomeric species which
interconvert through a fluxional process of the allyl
group.[15] In the case of an unsymmetrical palladium
complex such as 13a–e, endo- and exo-isomers are
possible (Scheme 3).[16] The interconversion of endo-
and exo P-N-palladium h3-C3H5[15b–c,17] and allyl substi-
tuted[15e,f,16,18] species is a well studied process which
involves, in most of the cases, a h3-h1-h3 isomerization,
with rotation of the sp3-sp2 bond of the allyl moiety in
Scheme 3.
Figure 1. X-Ray structure of endo-13a.
2268
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
Table 4. Ligand fine-tuning in the catalytic asymmetric allyl-
ic alkylation with 13.
Table 3. As anticipated from what has been observed
in other palladium-allyl complexes,[17] the Pd C bonds
À
to the terminal allylic C-atom trans to the P-atom are
significantly longer (2.217 Š for the endo isomer and
Entry Catalyst
Time
[h]
Conversion
[%][a]
ee
[%][b]
À
2.225 Š for the exo one) than the Pd C bonds trans
1
2
3
4
5
13a (R=Me)
4
>99
>99
87
60
>99
96
85
87
83
86
to the N-atom (2.125 Š for both the endo and exo iso-
13b (R=CH2Ph)
13c (R=CHPh2)
13d (R=CPh3)
13e
4.3
27
27
4.5
mers).
Table 3. Selected bond lengths [Š] for the endo- and exo-
13a-bis.
(R=CH2CH2OMe)
À
Pd(1) N(1)
Pd(1) C(30)
2.109(3)
2.125(4)
2.137(7)
2.206(8)
2.217(7)
2.225(8)
2.289(8)
[a]
[b]
Conversion was measured by NMR.
À
The product had the R configuration, which was estab-
lished by comparison of the sign of the optical rotation
of 15a with the reported value.[20] Enantiomeric excesses
were measured by chiral HPLC.
À
Pd(1) C(31)
Pd(1)-C
(31’)
À
Pd(1) C(32)
À
Pd(1) C
A
À
Pd(1) P(1)
When 2.5 mol% of catalyst were used, total conver-
sion of 14a was observed at room temperature after
The allyl ligand was found to be disordered and re- 4 h with catalysts 13a, b and 13e (R=Me, CH2Ph and
fined with only one position for one terminal allyl CH2CH2OMe respectively, entries 1, 2 and 5 in
carbon (C-30, see Figure 1) and two positions for the Table 4). The activity of the catalysts in terms of con-
other terminal (C-32) and the central (C-31) allyl version decreased with the size of the protecting
carbon atoms. Refining the occupancy factors afford- group, and lower conversions were observed for the
ed an endo:exo ratio of 58:42. This result fits reasona- ligands with bulkier protecting groups (R=CHPh2
bly well with the endo:exo ratio measured in solution and CPh3, entries 3 and 4, Table 4). The best enantio-
by integration of the central allyl proton signals, selectivity (96% ee, entry 1 in Table 4) was observed
which was found to be 65:35.
in the case of the catalyst 13a which bears the less
bulky substituent at C-5 in the oxazoline ring. The
enantioselectivity for the rest of the members of the
family ranged from 83 to 87% ee, the lower value
being recorded for the catalyst with the bulkier sub-
stituent (83% ee, entry 4 in Table 4, R=CPh3). It is
Allylic Alkylation Catalyzed by Pd/
Phosphinooxazoline Complexes
To test the effect of structural variation in the modu- clear from these results that 13a, containing a me-
lar phosphinooxazolines (CH2OR group) on the cata- thoxymethyl substituent, is the optimal ligand among
lytic activity in the asymmetric allylic alkylation reac- those in the studied series.
tion, palladium complexes 13 containing an unsubsti-
For this ligand (13a), the reaction temperature was
tuted allyl moiety were used as catalytic precursors in optimized, and alternative activation methods such as
model asymmetric allylic alkylation reactions. In this microwave irradiation were also studied.[21] The re-
way, rac-(E)-3-acetoxy-1,3-diphenyl-1-propene (14a) sults are summarized in Table 5.
was alkylated in dichloromethane at room tempera-
The reaction was carried out at different tempera-
tures (from À108C to 558C) and very good selectivi-
ties were observed throughout the whole temperature
range (94–97% ee). With 13a, the best performance
in terms of enantioselectivity was achieved at 358C
(97% ee, entry 4). The reaction was also studied
under microwave irradiation and, in this case, the re-
action time could be dramatically shortened (15 min
for complete conversion) without any loss in enantio-
selectivity (entry 7 in Table 5). All together, this
result represents a remarkable example of good per-
Scheme 4. Asymmetric allylic alkylation of rac-(E)-3-acet-
oxy-1,3-diphenyl-1-propene (14a).
ture in the presence of 2.5 mol% of 13 (Scheme 4), formance of the asymmetric allylic alkylation under
the nucleophile being generated from dimethyl malo- microwave irradiation. The remarkably good temper-
nate with N,O-bis(trimethylsilyl)acetamide (BSA) ature/selectivity profile renders catalyst 13a optimal
and a catalytic amount of potassium acetate. The re- for future applications in larger scale because almost
sults of the enantioselective allylic alkylation are col- no temperature control would be required in order to
lected in Table 4.
achieve high enantioselectivities.
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2269

Dana Popa et al.
FULL PAPERS
Table 5. Optimization procedure for catalyst 13a in the
asymmetric allylic alkylation of 14a.
Entry Temperature [8C] Time
Conversion
[%][a]
ee
[h]
[%][b]
1
2
3
4
5
À10
24
7
4
2.5
1
>99
>99
>99
>99
96
94
95
96
97
95
0
r.t.
35
35, MW irradia-
tion
45
6
7
2
>99
95
95
55[c], MW irradia- 15 min 99
tion
[a]
[b]
Conversion was measured by NMR.
The product had the R configuration, which was estab-
lished by comparison of the sign of the optical rotation
of 15a with the reported value.[20] Enantiomeric excesses
were measured by chiral HPLC.
[c]
The reaction was carried out in acetonitrile.
With the optimized catalyst in hand, the asymmet-
ric alkylation reaction of substrates 14b–h mediated
by the p-allylpalladium-phosphinooxazoline complex
13a was studied (Scheme 5). These results are collect-
ed in the Table 6.
Excellent results were obtained in terms of enantio-
selectivity in the alkylation of allylic systems bearing
a bulky substituent at the carbon atom undergoing al-
kylation, such as 14b, 14f, and 14g. For all these sys-
tems, the enantioselectivities recorded with catalyst
Scheme 5. Asymmetric allylic alkylation of substrates 14b–h
13a are amongst the highest recorded with phosphi- with 13a.
Table 6. Asymmetric allylic alkylation of substrates 14b–h with catalyst 13a.
Entry
Substrate
Product
Solvent/Temperature
Reaction time [h]
Conversion [%][a]
ee [%][b]
1
2
3
4
5
6
7
8
14b
14b
14b
14c
14c
14d
14d
14e
14e
14f
14g
14h
14h
14h
(R)-15b
(R)-15b
(R)-15b
(R)-15c
(R)-15c
(S)-15d
(S)-15d
(S)-15e
(S)-15e
(R)-15f
(R)-15g
(R)-15h
(R)-15h
(R)-15h
DMF/308C
THF/708C
96
38
2
2
2
24
3.5
24
3.5
4
42
90
99
64
94
94
94
72
70
43
35
56
34
82
92
98
91
93
MeCN/1308C, MW
MeCN/1008C, MW
MeCN/1308C, MW
CH2Cl2/r.t.
MeCN/608C, MW
CH2Cl2/r.t.
>99
87
>99
40
>99
>99[c]
54[d]
95
>99
>99
9
MeCN/608C, MW
CH2Cl2/rt
10
11
12
13
14
MeCN/808C, MW
THF/À108C
8
17
1.5
2.5
THF/r.t.
THF/858C, MW
[a]
Conversions were measured by NMR.
[b]
The configurations for 15–g were established by comparison with either reported chromatographic elution orders or opti-
cal rotations. Compound 15h was assumed to have the (R) configuration by comparing its elution order with the one for
(R)-15a in the Chiralcel OD-H and AD-H columns. Enantiomeric excesses were measured by chiral HPLC for 15b, 15f
and 15h, by GC for 15d and 15e and by NMR for 15c and 15g (see Experimental Section for details).
The linear to branched product ratio was 93/7.
[c]
[d]
It refers to isolated yield.
2270
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
nooxazoline ligands in the considered reaction.[1g,2,28] a similar reversal in enantioselectivity is not observed
Excellent enantioselectivities have also been achieved in the case of 14g, where the allyl system bears bulki-
in the allylic alkylation of substrate 14h, which had er alkyl substituents. Thus, this behavior appears to
not previously been studied with phosphinooxazoline- be related to the small size of the methyl substituents,
based catalysts. The very good results with this sub- and to the possible changes that this can provoke in
strate and 14a render catalyst 13a as very suited for the stereochemistry/reactivity of the reactive inter-
the allylic alkylation of 1,3-diaryl-substituted allyl ace- mediate.
tates.
The influence of the bulkiness of the ipso substitu-
ent on enantioselectivity becomes clear when the re- Mechanistic Rationalization
sults on 14b/14c or 14a (Table 5)/14d are compared.
In these cases, replacement of a phenyl substituent by The Pd-catalyzed enantioselective C C bond-forming
a less bulky methyl group at the reaction site is ac- reactions that have been studied in this work start
À
companied by an important decrease in enantioselec- from racemic substrates and proceed via symmetrical
tivity.
Pd-allyl complexes 16. It should be recalled that com-
The tolerance of 13a towards temperature changes plexes 16 containing the 1,3-disubstituted allyl moiety
is of paramount importance for the determination of can give rise to several isomers, depending on the ste-
optimal reaction conditions. Thus, while the enantio- reochemical relationship between the substituents of
selectivity was high for substrate 14b at room temper- the allyl group and the central allyl hydrogen: syn-
ature (94% ee, entry 1), conversion was rather low syn; syn-anti, anti-syn and anti-anti isomers.[15a,f,18a,c–e,23]
under these conditions even after long reaction times. Furthermore, each of those isomers can have an exo
Microwave-assisted allylic alkylation turned out to be or endo arrangement of the allyl moiety. These iso-
more effective for this kind of substrates, and com- meric species are interconverting due to several flux-
plete conversion was achieved in acetonitrile at ional processes (Scheme 6). The anti-anti isomers (not
1308C in only 2 h. Very gratifyingly, the enantioselec- depicted in Scheme 6) have a very limited abundance
tivity achieved under these conditions was the same given their lower stability and are normally not detect-
recorded at room temperature (compare entries 3 and ed.[18d] It has been reported in the literature that the
1 in Table 6). To conclude with, microwave activation most abundant isomer in the case of 1,3-diphenyl sub-
is the technique of choice for the allylic alkylation re- stituted allyl systems corresponds to the exo-syn-syn
action of hindered allyl substrates mediated by 13a isomer, followed by the endo-syn-syn one, the ratio
(see entries 5, 7, 11, and 14 in Table 6). In particular, among them ranging from 4:1 to 12:1.[15a,16,18e,23b–c] The
the results obtained with 13a for substrates 14a, at same trend has been found for the 1,3-dimethylallyl
558C, 14b and 14c, at 1308C, and for 14g and 14h, at precursor,[18d,23a] although syn-anti diastereomers are
80–858C, confirm the remarkably wide selectivity- stable enough as to be also observed as minor compo-
temperature profile of this catalyst.
nents of the equilibrium mixture in this case. Further-
On the other hand, substrates 14d and 14e showed more, the relative abundance of the different diaste-
low enantioselectivities in the alkylation reaction (en- reomers of the 1,3-dimethylallyl system depends on
tries 6 and 8 in Table 6). Conversion could be im- the substitution pattern of the sp3 carbon next to ni-
proved under microwave irradiation but selectivities trogen in the oxazoline ring: if this carbon is a quater-
remained low (entries 7 and 9 in Table 6). In any case, nary one, the syn-anti isomers predominate.[23c]
it should be recalled that the selectivities observed for
In the asymmetric allylic alkylation, the enantiose-
these two substrates in the present work are similar to lectivity of the process is determined by the regiose-
the best ones achieved using well established phosphi- lectivity of the attack of the nucleophile onto the pal-
nooxazolines.[22] Also on the negative side, it is to be ladium allylic intermediate. The higher ability a
noted that with the unsymmetrical substrate 14f, al- ligand has to differentiate the two allyl termini to-
though the branched product is formed with high wards the incoming nucleophile, the higher selectivity
enantioselectivity, the ratio of branched to linear will be induced. In this complex scenario of reaction
product is low. Thus, the challenge of simultaneously pathways arising from the different precursors and
achieving high regiocontrol in favour of the branched attack positions, it is normally accepted that the
product and high enantiocontrol in the major regio- major ongoing process arises from the exo-syn-syn-16
isomer posed by non-symmetrical substrates does not precursor by attack of the nucleophile onto the allyl
find a definitive solution with the use of 13a.
Among these results, the reversal in the sign of The R configuration of the final product 15a for the
enantioselectivity observed when passing from sub- 1,3-diphenyl-substituted starting material 14a
carbon trans to the phosphorus atom (Scheme 7).[1e,15a]
strate 14a (Table 5) to 14d is particularly noteworthy, (Scheme 4) would be in agreement with this assump-
since a parallel behavior has never been observed tion. In an analogous way, the attack of the nucleo-
before in this reaction. It is interesting to realize that phile onto the allyl carbon trans to the phosphorus
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2271

Dana Popa et al.
FULL PAPERS
Scheme 6.
In order to understand the reversal in the sign of
the enantioselection when moving from the 1,3-diphe-
nylallyl to the 1,3-dimethylallyl substrate, we have
performed and report here the results of a theoretical
study on the relevant syn:anti isomers of the p-allyl-
palladium complexes of phosphinooxazoline 12a
(16aMe and 16aPh). A two-layer ONIOM approach
has been used for these studies,[25] with the partition-
ing shown in Figure 2. The high-level layer was treat-
Scheme 7.
Figure 2. Molecular partition for ONIOM calculations.
substituent in precursor endo-syn(P)-anti(N)-16 ed at the highest, B3LYP/LanL2DZ level. The rest of
would lead to the major enantiomer of the substitu- the molecule was included into the low-level layer
tion product 15a as well. However, this reaction path- and treated at the semi-empirical AM1 level. There-
way is probably less important as the previous one, fore, the two-layer ONIOM approach used in the
since the endo-syn(P)-anti(N)-16 precursor is predict- present studies is denoted as ONIOM (B3LYP/
ed to be much less abundant than the exo-syn-syn LanL2DZ: AM1). The considered geometries were
one.[24] In principle, the substitution product with the optimized with respect to all geometrical parameters,
R configuration could also arise from attack of the and the ONIOM total energy was computed. All cal-
nucleophile onto the allyl carbon trans to nitrogen in culations were performed without symmetry con-
precursors endo-syn-syn-16 and exo-anti(P)-syn(N)- straints using the Gaussian 03 program.[26]
16 as well. However, it is normally accepted that
The syn-syn and the two possible syn-anti diastereo-
these last two reaction pathways are irrelevant, as the mers were calculated for both exo and endo arrange-
allyl carbon trans to phosphorus is much more reac- ments of 16aMe and 16aPh.[27] Results on ONIOM
tive towards nucleophiles (The chemical shift of the total energy and relative energy have been summar-
carbon atom trans to phosphorus was observed at ized in Table 7.
103.1 ppm in 13C NMR for exo-syn-syn-16 (R=Ph),
The molecular structures of the minimal energy iso-
whereas the chemical shift of the carbon atom trans mers of 16aPh and 16aMe are shown in Figure 3
to the nitrogen was observed at 69.1 ppm). along with some key atomic distances. From the
2272
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
Table 7. ONIOM total energy (in atomic units) and relative energy (in kcal·molÀ1) of the relevant stereoisomers of 16aPh
and 16aMe.
16aPh (1,3-Diphenylallyl)
16aMe (1,3-Dimethylallyl)
ONIOM Total Energy [a.u.]
Rel. Energy
ONIOM Total Energy [a.u.]
Rel. Energy
[kcal·molÀ1]
[kcal·molÀ1]
endo-anti(P)-syn(N)[a]
endo-syn(P)-anti(N)
endo-syn-syn
exo-anti(P)-syn(N)
exo-syn(P)-anti(N)
exo-syn-syn
À576.5771334
À576.5670563
À576.5793477
À576.5706163
À576.5738638
À576.5819787
3.04
9.36
1.65
7.13
5.09
0.00
À576.6357478
À576.6285414
À576.6310965
À576.6312656
À576.6407325
-576.6336009
3.13
7.65
6.05
5.94
0.00
4.48
[a]
This denotes that the substituent in the allyl terminus proximal to P is in anti arrangement and the substituent in the allyl
terminus proximal to N is syn.
Figure 3. ONIOM optimized molecular structures of the minimal energy stereoisomers of 16aPh and 16aMe.
methodological point of view, it is important to note through intermediate exo-syn-syn-16aPh and will lead
that the optimized distances show a good overall to an R product, as experimentally observed. In the
agreement with those found in the crystal structure of same manner, the calculations predict that the alkyla-
13a (see Table 3), only the Pd-P distance in the p-allyl tion of 1,3-dimethylallyl acetate will take place
complexes being slightly overestimated by the calcula- through intermediate exo-syn(P)-anti(N)-16aMe and
tions.
will lead to an S product, as experimentally observed.
With respect to the stereochemical outcome of the Thus, the ONIOM calculations provide an explana-
reactions, and assuming a reactant-like character of tion for the reversal of enantioselectivity in the alky-
the transition states for the malonate attack to the lations of 1,3-diphenylallyl and 1,3-dimethylallyl ace-
cationic complexes 16a, the energetic ordering of the tate with the same catalytic system based on the
different stereoisomers and the well-based[1e,15a] as- nature of the most stable p-allylpalladium intermedi-
sumption that the nucleophilic attack will take place ates containing these systems. The stability of the dif-
at the allyl terminus trans to phosphorus dictate the ferent stereoisomers contributing to the description of
stereochemistry of the reaction product.
these intermediates appears to be importantly con-
Very interestingly, the calculations predict that the trolled by the size of the 1,3-substituents in the allyl
alkylation of 1,3-diphenylallyl acetate will take place system.[28]
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2273

Dana Popa et al.
FULL PAPERS
vacuum. The raw material was dissolved in CH2Cl2 (5 mL)
and washed with water (3”5 mL). The organic phase was
dried over anhydrous MgSO4 and the solvent was removed
under reduced pressure. The corresponding amino alcohols
8a–e were used in the next step without further purification.
Noteworthy, the ONIOM calculations provide also
a clue for the understanding of the different reactivity
recorded with the 1,3-diphenylallyl and 1,3-dimethyl-
allyl substrates. Taking the 1,3-diphenylallyl system
(16aPh) as a reference (see Figure 3), a comparison
of the distance between palladium and the carbon
Epoxide Ring Opening with Aqueous Ammonia
under Thermal Conditions; Synthesis of Amino
Alcohols 8a–e; General Procedure
À
atom trans to phosphorus (Pd Cc) in the exo-syn-syn
stereoisomer with that found in the minimal energy
stereoisomer of 16aMe (2.321 Š vs. 2.269 Š) tends to
indicate that substrate 14a should exhibit a higher re-
activity than 14d.
The corresponding epoxide (1.0 mmol), 30% aqueous NH3
(50 mmol), LiClO4 (1 mmol) and isopropyl alcohol (5 mL) if
required were heated in a sealed pressure at 1008C for the
period of time indicated in Table 1. The mixture was left to
reach room temperature and the solvent was removed under
reduced pressure. The raw material was dissolved in CH2Cl2
(5 mL) and washed with water (3”5 mL). The organic
phase was dried over anhydrous MgSO4 and the solvent was
removed at reduced pressure. The corresponding amino al-
cohols 8a–e were used in the next step without further pu-
rification.
Conclusions
In summary, the synthesis of a family of enantiopure
modular phosphinooxazolines, containing a trans-4-
phenyl-5-alkoxymethyl-disubstituted oxazoline ring
and arising from purely synthetic, yet enantiopure
precursors has been achieved, and their p-allylpalladi-
um complexes have been evaluated as ligands in the
asymmetric allylic alkylation reaction. From a practi-
cal point of view, we have been able to identify a
phosphinooxazoline ligand (12a) that efficiently con-
trols the enantioselectivity of the alkylation of allylic
substrates over a very broad range of temperatures,
thus allowing performing the reaction in extremely
short times (minutes). From a more general perspec-
tive, the presence of an additional substituent at C-5
in the oxazoline ring leads to improved catalytic char-
acteristics in the phosphinooxazoline ligands with re-
spect to those bearing a phenyl substituent at C-4, but
lacking the alkoxymethyl substituent at C-5, and this
could be related to conformational control of the
phenyl group by the alkoxymethyl substituent at the
level of the reactive intermediate. The remarkably
good performance of the catalysts makes them attrac-
tive for future applications. In view of immobilization,
the results of the benchmark-guided structure optimi-
zation suggest that linear, non-bulky linkers should
lead to minimal deterioration of catalytic properties
in polymer-supported phosphinooxazolines 12.
Synthesis of (2-Fluoro)phenyl Hydroxy Amides 9a-e;
General Procedure
Et3N (1.95 mmol) was added under Ar to a solution of the
corresponding amino alcohol 8a–e (1.0 mmol) in anhydrous
THF(1.5 mL) and the mixture was cooled at 0 8C. A solu-
tion of 2-fluorobenzoyl chloride (0.95 mmol) in anhydrous
THF(1.3 mL) was slowly added to the previous solution.
The mixture was left to reach room temperature and further
stirred for two hours. The solvents were removed under
vacuum. The raw material was diluted with CH2Cl2 (5 mL)
and washed with brine (3”5 mL). The organic phase was
dried over anhydrous Na2SO4, the drying agent was filtered
off and the solvent removed under reduced pressure. The
raw material was purified by chromatography over SiO2
eluting with a mixture of hexanes/AcOEt of increasing po-
larity: from 100:0 to 60:40.
See Supporting Information for the physical and spectro-
scopical data of compounds 9a–e.
As a representative example, 2-fluoro-N-[(1R,2R)-2-hy-
droxy-3-methoxy-1-phenylpropyl]-benzamide (9a): [a]2D3:
1
+32.9 (c 0.74 in CHCl3); H NMR: d=8.22 (bs, 1H, NH),
8.08 (dt, J=7.9, 1.9 Hz, 1H), 7.50–7.10 (m, 8H), 5.45 (ddd,
J=8.4, 4.8, 2.4 Hz, 1H), 4.14–4.07 (m, 1H), 3.45 (dd, J=9.5,
3.3 Hz, 1H), 3.38 (s, 3H), 3.28 (dd, J=9.9, 5.1 Hz, 1H), 2.58
(d, J=6.6 Hz, 1H); 13C NMR: d=162.8 (C), 160.7 (d, JC,F
=
246.0 Hz, C), 138.5 (C), 133.2 (d, JC,F =9.6 Hz, CH), 132.1
(d, JC,F =1.8 Hz, CH), 128.6 (CH), 127.6 (CH), 127.1 (CH),
124.7 (d, JC,F =3.2 Hz, CH), 120.9 (d, JC,F =11.0 Hz, C), 115.9
(d, JC,F =24.5 Hz, CH), 73.4 (CH2), 72.1 (CH), 59.2 (CH3),
57.1 (CHNH); IR (film): n=3340, 1638, 1615, 1534, 1482,
Experimental Section
General remarks about the experimental section can be
found in the Supporting Information.
1453, 1100, 754, 702 cmÀ1
; MS (CI, NH3): m/z=304
(C17H19NO3F·H+, 100); HRMS (CI): m/z=304.1371, calcd.
Microwave Assisted Epoxide Ring Opening with
Aqueous Ammonia; Synthesis of Amino Alcohols
8a-e; General Procedure
for C17H19NO3F·H+: 304.1349.
Synthesis of (2-Fluoro)phenyloxazolines 11a–e;
General Procedure
The corresponding epoxide (1.0 mmol), 30% aqueous NH3
(50 mmol), LiClO4 (1 mmol) and of isopropyl alcohol
(5 mL) if required were irradiated at 1258C for the period
of time indicated in Table 1. The mixture was left to reach
room temperature and the solvent was removed under
Methanesulfonyl chloride (1.1 mmol) was slowly added to a
solution of compounds 9a–e (1.0 mmol) and Et3N
(2.2 mmol) in CH2Cl2 (8.2 mL) under an argon atmosphere
2274
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
at 08C. The mixture was left to reach room temperature and
further stirred for two hours. This mixture was added over a
saturated aqueous solution of NH4Cl (9 mL), the two phases
were separated and the aqueous phase was extracted with
CH2Cl2 (3”6 mL). The organic phases were washed with
brine (5 mL), dried over anhydrous Na2SO4. The drying
agent was filtered off and solvent removed under reduced
pressure. This raw material was added to a 5% solution of
KOH in MeOH (9 mL, 2.0 mmol), and the mixture was
stirred 15 h at room temperature. The reaction was
quenched with water (5 mL) and extracted with CH2Cl2 (3”
5 mL). The organic phases were dried over anhydrous
Na2SO4. The drying agent was filtered off and solvent re-
moved under reduced pressure. The purification of the prod-
ucts 11a–e was carried by flash column chromatography.
See Supporting Information for the physical and spectro-
scopical data of compounds 11a–e.
added. The resulting solution was further stirred for 14 h
and allowed to stand at À208C for several hours. The palla-
dium complexes precipitated were filtered off, washed with
cold ethanol (3 mL) and dried under vacuum.
See Supporting Information for the physical and spectro-
scopical data of compounds 13a–e.
As a representative example, ACTHERUGN ACHTRE[UGN (4R,5S)-5-methoxy-
(h3-allyl)
methyl-4-phenyl-2-(2-diphenyl-phosphinophenyl)-4,5-dihy-
drooxazol]palladium(II) hexafluorophosphate (13a): [a]2D3:
À70.4 (c 1.0, CHCl3); 1H NMR: d=8.31 (ddd, J=8.0,
4.4,1.2 Hz, 2H), 7.71–7.06 (m, 32H), 4.63 (bs, 2), 3.83–3.67
(m, 5H); endo: 6.96 (d, J=7.3 Hz, 2H), 5.73 (tt, J=13.0,
7.0 Hz, 1H), 5.34 (d, J=7.5 Hz, 1H), 4.60 (bs, 1H), 3.39 (s,
3H), 3.22 (d, J=6.3 Hz, 1H), 2.74 (d, J=11.0 Hz, 1H), 2.61
(d, J=12.5 Hz, 1H); exo: 6.92 (d, J=7.5 Hz, 2H), 5.38 (d,
J=7.5 Hz, 1H), 5.15 (tt, J=13.3, 6.7 Hz, 1H,), 4.22 (bs, 1H),
3.38 (s, 3H), 3.31 (d, J=6.3 Hz, 1H), 2.78 (t, J=12.0 Hz,
1H); 13C NMR: d=134.4 (CH), 134.3 (CH), 133.7 (CH),
133.6 (CH), 133.5 (CH), 133.4 (CH), 133.0 (CH), 132.2
(CH), 132.1 (CH), 132.1 (CH), 129.9 (CH), 129.8 (CH),
129.7 (CH), 129.6 (CH), 129.3 (CH), 129.0 (CH), 128.9
(CH), 128.9 (CH), 126.9 (CH), 87.5 (CH), 76. (CH), 72.2
(CH2), 59.5 (CH3); endo: 165.2 (CN), 139.3 (C), 121.2 (CH),
79.2 (d, J=29.0 Hz CH2), 57.2 (CH2); exo: 165.2 (CN), 139.1
(C, Ph), 121.3 (CH), 80.9 (d, J=30.3 Hz CH2), 56.2 (CH2);
31P NMR: d=25.0 (s, PPh2 exo), 24.5 (s, PPh2 endo:), À141.2
(h, J=712.5 Hz, PF6); IR (KBr): n=3064, 2933, 1627, 1482,
1437, 1121, 1100, 839, 731, 698 cmÀ1; MS (FAB): m/z=598
(M+ÀPF6, 100%); anal. calcd. for C32H31NO2PPdPF6: C
51.66, H 4.20, N 1.88; found: C 51.17, H 4.30, N 1.96.
As a representative example, ACHTER(UNG 4R,5S)-2-(2-fluorophenyl)-
5-methoxymethyl-4-phenyl-4,5-dihydrooxazole (11a): [a]2D3:
+47.6 (c 0.68 in CHCl3); 1H NMR: d=8.04 (ddd, J=8.0,
1.6, 1.6 Hz, 1H), 7.55–7.13 (m, 8H), 5.17 (d, J=7.6 Hz, 1H),
4.66 (ddd, J=7.6, 4.7, 4.7 Hz, 1H), 3.73–3.71 (m, 2H), 3.47
(s, 3H); 13C NMR: d=161.2 (d, JC,F =257.0 Hz, C), 160.8 (d,
J
C,F =5.4 Hz, C), 141.6 (C), 133.0 (d, JC,F =8.6 Hz, CH), 131.2
(d, JC,F =1.3 Hz, CH), 128.6 (CH), 127.6 (CH), 126.6 (CH),
123.8 (d, JC,F =4.1 Hz, CH), 116.6 (d, JC,F =21.0 Hz, CH),
115.8 (d, JC,F =10.4 Hz, C), 85.6 (CH), 73.5 (CH2), 72.3
(CH), 59.5 (CH3); IR (film): n=1652, 1497, 1457, 1055, 764,
700 cmÀ1; MS (CI, NH3): m/z=304 (M+ +18, 2%), 286
(M+ +1, 100%); HRMS (CI): m/z=286.1241, calcd. for
C17H17NO2F(M + +1): 286.1243.
(h3-1,3-Diphenylallyl)
ACHTRE[UGN (4R,5S)-5-methoxymethyl-4-
phenyl-2-(2-diphenylphosphinophenyl)-4,5-
dihydrooxazol]palladium(II) Hexafluorophosphate
(16)
Synthesis of the Phosphinooxazoline Ligands 12a–e;
General Procedure
A solution of Ph2PK (1.45 mmol, 2.8 mL of 0.5m solution in
THF) was slowly added under argon at À788C via cannula
to an oven-dried Schlenk flask which contained the corre-
sponding fluorooxazoline 11a–e (1.0 mmol) in THF(2 mL).
The temperature was allowed to reach À208C. The reaction
mixture was further stirred for 2 h at this temperature, al-
lowed then to reach room temperature, further stirred for
12 h at this temperature, quenched with Na2SO4·10H2O in
order to hydrolyze the excess of diphosphine, and filtered
through a short SiO2 pad eluting with CH2Cl2. The residue
after removing the solvents was further purified by column
chromatography over SiO2 under argon. SiO2 and the sol-
vents which were involved in the chromotagraphic purifica-
tions were previously deoxygenated either by flushing or
bubbling argon. Compounds 12 were not fully characterized
Following the general method from ligand 12a (0.06 g,
0.13 mmol),
complex
[Pd
(C3H5)(C6H5)2Cl]2
(0.047 g,
0.07 mmol) and NH4PF6 (0.033 g, 0.2 mmol), compound 16
was obtained after filtration as a yellow solid; yield: 0.060 g
1
(0.06 mmol, 50%). [a]D23: À325.3 (c 0.34, CHCl3); H NMR:
d=7.70–6.60 (m, 56H), 4.34–4.31 (m, 2H); endo: 8.22 (ddd,
J=8.0, 4.2, 1.0 Hz, 1H), 6.53–6.48 (m, 1H), 4.78 (t, J=
11.2 Hz, 1H), 4.68 (d, J=12 Hz, 1H), 4.46 (d, J=5.5 Hz,
1H), 3.29–3.20 (m, 2H), 3.22 (s, 3H); exo: 8.26 (ddd, J=8.0,
4.0, 1.2 Hz, 1H), 6.29 (dd, J=14.0, 10.5 Hz, 1H), 6.20 (d, J=
8.5 Hz, 1H), 4.26 (d, J=11.0 Hz, 1H), 4.23 (d, J=5.0 Hz,
1H), 3.53 (dd, J=11.0, 5.2 Hz, 1H), 3.50 (dd, J=11.0,
4.5 Hz, 1H), 3.47 (s, 3H); 13C NMR: d=165.1 (C), 139.2
(C), 138.9 (C), 137.9 (C), 135.8 (C), 135.3 (CH), 133.4 (CH),
133.2 (CH), 133.0 (CH), 132.1 (CH), 131.6 (CH), 131.2
(CH), 129.6 (CH), 129.3 (CH), 129.2 (CH), 128.6 (CH),
128.4 (CH), 128.3 (CH), 127.8 (CH), 127.1 (CH), 126.8
(CH), 126.7 (CH), 125.9 (CH), 125.7 (CH), 86.1 (CH), 72.3
(CH2), 59.6 (CH3); endo: 111.2 (CH), 95.7 (CH), 74.8 (CH),
1
(only H and 31P NMR were recorded, see Supporting Infor-
mation) and were immediately transformed into the corre-
sponding palladium complexes 13.
Synthesis of the Phosphinooxazoline Allylpalladium
Complexes 13a–e. General Procedure
71.2 (CH); exo: 110.7 (d, JC,P =5.2 Hz, CH), 103.1 (d, JC,P
=
A solution of compound 12a–e (1.0 mmol) in deoxygenated
ethanol (2 mL) was added via cannula to a Schlenk flask
which contained a solution of [PdACHTRE(UGN C3H5)Cl]2 (0.53 mmol) in
ethanol (5 mL). The reaction mixture was stirred 1 hour
under argon at room temperature, NH4PF6 (1.02 mmol) was
22.1 Hz, CH), 70.2 (CH), 69.1 (d, JC,P =7.0 Hz, CH);
31P NMR: d=22.5 (s, PPh2 endo), 17.7 (s, PPh2 exo), À140.6
(h, J=713.2 Hz, PF6); IR (KBr): n=3062, 2929, 1628, 1491,
1437, 1119, 1100, 837, 735, 695 cmÀ1; MS (FAB): m/z=750
(M+ÀPF6, 100%).
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2275

Dana Popa et al.
FULL PAPERS
pared in the appropriated solvent (see Table 6) in an analo-
gous way as described before for the thermal conditions.
The reaction mixture was heated in a microwave reactor for
the reaction time and at the temperature indicated in
Table 6. Enantiomeric excesses for 15a, 15b, 15d, 15e and
15h were measured as described before. The enantiomeric
excesses for 15c and 15g were determined by NMR using
General Procedure for Palladium-Catalyzed Allylic
Alkylation; Thermal Conditions
Allylic alkylation of 14a:
A solution of 14a (0,25 g,
1.0 mmol) in CH2Cl2 (2 mL) was added at room temperature
to a solution 13a–e (0.02 mmol) in CH2Cl2 (2 mL) under
argon. Dimethyl malonate (0.396 g, 0.350 mL, 3.0 mmol)
and BSA (0.610 g, 0.740 mL, 3.0 mmol) were syringed into
the above solution and a catalytic amount of KOAc (0.004 g,
0.04 mmol) was lastly added under argon. The mixture was
stirred at room temperature for 48 h (unless stated other-
wise stated). The reaction mixture was then diluted with di-
ethyl ether, filtered over Celite, and washed with water (4”
10 mL). The organic phase was dried over anhydrous
Na2SO4. The drying agent was filtered off and the solvent
was removed under reduced pressure. The crude mixture
was filtered through a short SiO2 pad eluting with ethyl ace-
tate. The conversion of the reaction was measured after re-
the Eu
of solutuins in CDCl3 were taken using 6 equivalents of 15c
and 15g with respect to one equivalent of the Eu(hcf)3. The
ACHTRE(UNG hcf)3 chiral chemical shift reagent. The spectra were
AHCTREUNG
enantiomeric excess was measured by integrating suitable
signals from the two diastereomeric complexes.
The configurations of the final products were established
by comparison either with with either reported chromato-
graphic elution orders or optical rotations: (+)-(R)-15a;[20]
(+)-(R)-15b;[29] (+)-(R)-15c;[29] (À)-(S)-15d;[20] (À)-(S)-
15e;[20] 15f (Chiralcel OJ);[30] (S)-15f Rt=24.9 min, (R)-15f
Rt=27.7 min and (À)-(S)-15g.[31]
1
moving the solvent by H NMR of the crude mixture. Di-
methyl malonate was mostly removed under vacuum
(1308C, 0.4 mbar) and the enantiomeric excesses were deter-
mined from the residue by HPLC on a OD-H column
(0.5 mLminÀ1 n-hexane/isopropyl alcohol, 99:1): (R)-15a
Rt=23 min, (S)-15a Rt=25 min.
Allylic alkylation of 14b: The procedure was analogous to
the one described for 14a but using 13a as the catalyst, 14b
as starting material, and DMFor THFas solvents at the
temperatures indicated in Table 6. The enantiomeric excess-
Acknowledgements
We thank DGI-MCYT (Grant No. CTQ2005–02193/BQU),
DURSI (Grant. No. 2005SGR225),Consolider Ingenio 2010
(CSD2006–0003),ICIQ Foundation and Ramón Areces
Foundation for financial support.
es were determined by HPLC on
a AD-H column
(0.3 mLminÀ1 n-hexane/isopropyl alcohol, 97:3): (R)-15b
Rt=49.4 min, (S)-15b Rt=51.6 min.
References
Allylic alkylation of 14d: The procedure was analogous to
the one described for 14a but using 13a as catalyst and 14d
as starting material, and CH2Cl2 at room temperature. The
enantiomeric excesses were determined by GC on a FS-cy-
clodex-b-Dex column (808C): (R)-15d Rt=68.0 min, (S)-15d
Rt=69.2 min.
Allylic alkylation of 14e: The procedure was analogous to
the one described for 14a but using 13a as catalyst and 14e
as starting material, and CH2Cl2 at room temperature. The
enantiomeric excesses were determined by GC on a FS-cy-
clodex-b-Dex column (10 min at 908C, 18CminÀ1 up to
1308C, 15 min at 1308C, 58CminÀ1 up to 1508C, 15 min at
508C): (S)-15e Rt=57.0 min, (R)-15e Rt=57.4 min.
Allylic alkylation of 14f: The procedure was analogous to
the one described for 14a but using 13a as the catalyst, 14 f
as starting material, and CH2Cl2 at room temperature. The
enantiomeric excesses were determined by HPLC on a OJ-
H column (0.7 mLminÀ1 n-hexane/isopropyl alcohol 97:3):
(S)-15f Rt=26.7 min, (R)-15f Rt=28.4 min.
[1] a) C. G. Frost, J. Howart, J. M. J. Williams, Tetrahedron:
Asymmetry 1992, 3, 1089–1122; b) B. M. Trost, D. L.
Van Vranken, Chem. Rev. 1996, 96, 395–422; c) J. M. J.
Williams, Synlett 1996, 705–710; d) B. Bartels, G.
Helmchen, Chem. Commun. 1999, 741–742; e) G.
Helmchen, A. Pfaltz, Acc. Chem. Res. 2000, 33, 336–
345; f) H. A. McManus, P. J. Guiry, Chem. Rev. 2004,
104, 4151–4202; g) A. Pfaltz, W. J. Drury, III, Proc.
Natl. Acad. Sci. U.S.A. 2004, 101, 5723–5726.
[2] P. von Matt, A. Pfaltz, Angew. Chem. Int. Ed. Engl.
1993, 32, 566–568.
[3] J. Sprinz, G. Helmchen, Tetrahedron Lett. 1993, 34,
1769–1772.
[4] G. J. Dawson, C. G. Frost, J. M. J. Williams, S. J. Coote,
Tetrahedron Lett. 1993, 34, 3149–3150.
[5] G. Consiglio, R. M. Waymouth, Chem. Rev. 1989, 89,
257–276.
[6] The preparation and use of phosphinooxazolines and
other related P-N ligands in asymmetric catalysis has
been widely reviewed (see ref.[1]). For additional, rele-
vant references, see: a) O. Pàmies, M. DiØguez, C.
Claver, J. Am. Chem. Soc. 2005, 127, 3646–3647; b) A.
Froelander, S. Lutsenko, T. Privalov, C. Moberg, J. Org.
Chem. 2005, 70, 9882–9891; c) J. M. Benito, C. A.
Christensen, M. Meldal, Org. Lett. 2005, 7, 581–584;
d) D. Liu, Q. Dai, X. Zhang, Tetrahedron 2005, 61,
6460–6471; e) D. Franco, M. Gómez, M.; F. JimØnez,
G. Muller, M. Rocamora, M. A. Maestro, J. Mahia, Or-
ganometallics 2004, 23, 3197–3209; f) N. End, C. Stoes-
sel, U. Berens, P. di Pietro, P. G. Cozzi, Tetrahedron:
Asymmetry 2004, 15, 2235–2239; g) H. Danjo, M. Hi-
Allylic alkylation of 14h: The procedure was analogous to
the one described for 14a but using 13a as the catalyst, 14h
as starting material, and THFas solvent at the temperatures
indicated in Table 6. The enantiomeric excesses were deter-
mined by HPLC on a AD-H column (0.8 mLminÀ1 n-
hexane/isopropyl alcohol 85:15): (R)-15h Rt=19.3 min, (S)-
15h Rt=29.5 min.
General Procedure for Palladium-Catalyzed Allylic
Alkylation; Microwave-Assisted Conditions
The reaction mixture for racemic allylic acetates 14a, 14b,
14c, 14d, 14e, 14g, and 14h using 13a as the catalyst was pre-
2276
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278

Phosphinooxazolines Derived from 3-Amino-1,2-diols
FULL PAPERS
guchi, M. Yada, T. Imamoto, Tetrahedron Lett. 2004,
45, 603–606.
ggelin, G. Helmchen, Angew. Chem. Int. Ed. Engl.
1997, 36, 2108–2110.
[16] N. Baltzer, L. Macko, S. Schaffner, M. Zehnder, Helv.
[7] a) A. Vidal-Ferran, A. Moyano, M. A. Pericàs, A.
Riera, Tetrahedron Lett. 1997, 38, 8773–8776; b) A.
Vidal-Ferran, A. Moyano, M. A. Pericàs, A. Riera, J.
Org. Chem. 1997, 62, 4970–4982; c) C. Jimeno, M.
Pastó, A. Riera, M. A. Pericàs, J. Org. Chem. 2003, 68,
3130–3138.
Chim. Acta 1996, 79, 803–812.
[17] S. Schaffner, L. Macko, M. Neuburger, M. Zehnder,
Helv. Chim. Acta 1997, 80, 463–471.
[18] a) A. Togni, U. Burckhardt, V. Gramlich, P. S. Pregosin,
R. Salzmann, J. Am. Chem. Soc. 1996, 118, 1031–1037;
b) R. Malet, M. Moreno-MaÇas, F. Pajuelo, T. Parella,
R. Pleixats, Magn. Reson. Chem. 1997, 35, 227–236;
c) S. Liu, J. F. K. Muller, M. Neuburger, S. Schaffner,
M. Zehnder, Helv. Chim. Acta 2000, 83, 1256–1267;
d) M. Kollmar, B. Goldfuss, M. Reggelin, F. Rominger,
G. Helmchen, Chem. Eur. J. 2001, 7, 4913–4927; e) P.
Dotta, P. G. A. Kumar, P. S. Pregosin, Magn. Reson.
Chem. 2002, 40, 653–658.
[19] Crystals of anion exchanged 13a (13a-bis) could be ob-
tained by slow evaporation of a solution of the com-
pound in methanol. The crystals which were obtained
corresponded to the hydrogen sulfate palladium com-
plex. Traces of sulfate present in the solution produced
a more insoluble solid which crystallized out from
methanol after anionic exchange. Crystal data for 13a-
bis at 100 K: C32H31NO2PPd1·HSO4, 696.02 gmolÀ1,
monoclinic, P21, a=9.4197(11) Š, b=16.032(2) Š, c=
10.0162(12) Š, b=96.836(2)8, V=1501.9(3) Š3, Z=4,
1calcd. =1.539 Mg/m3, R1 =0.0637 (0.0756), wR2=0.1653
(0.1750), for 12654 reflections with I>2s(I) (for 14475
reflections [Rint: 0.0412] with a total measured of 28292
reflections), goodness-of-fit on F2 =1.099, largest diff.
peak (hole)=4.541 (À4.060) eŠÀ3. CCDC 626167 con-
tains the supplementary crystallographic data for this
paper. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the
Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB21EZ, UK; (fax: (+44) 1223-336-
033; or e-mail: deposit@ccdc.cam.ac.uk).
[8] C. Jimeno, A. Vidal-Ferran, A. Moyano, M. A. Pericàs,
A. Riera, Tetrahedron Lett. 1999, 40, 777–780.
[9] M. Pastó, A. Riera, M. A. Pericàs, Eur. J. Org. Chem.
2002, 2337–2341.
[10] C. Puigjaner, A. Vidal-Ferran, A. Moyano, M. A. Peri-
càs, A. Riera, J. Org. Chem. 1999, 64, 7902–7911.
[11] M. A. Pericàs, C. Puigjaner, A. Riera, A. Vidal-Ferran,
M. Gómez, F. JimØnez, G. Muller, M. Rocamora,
Chem. Eur. J. 2002, 8, 4164–4178.
[12] a) A. Vidal-Ferran, N. Bampos, A. Moyano, M. A. Peri-
càs, A. Riera, J. K. M. Sanders, J. Org. Chem. 1998, 63,
6309–6318; b) J. M. Fraile, J. A. Mayoral, J. Serrano,
M. A. Pericàs, L. Solà, D. Castellnou, Org. Lett. 2003,
5, 4333–4335; c) M. A. Pericàs, D. Castellnou, I. Rodrí-
guez, A. Riera, L. Solà, Adv. Synth. Catal. 2003, 345,
1305–1313; d) D. Castellnou, M. Fontes, C. Jimeno, D.
Font, L. Solà, X. Verdaguer, M. A. Pericàs, Tetrahedron
2005, 61, 12111–12120; e) D. Castellnou, L. Solà, C.
Jimeno, J. M. Fraile, J. A. Mayoral, A. Riera, M. A.
Pericàs, J. Org. Chem. 2005, 70, 433–438; f) D. Font, C.
Jimeno, M. A. Pericàs, Org. Lett. 2006, 8, 4653–4655;
g) A. Bastero, D. Font, M. A. Pericàs, J. Org. Chem.
2007, 72, 2460–2468; h) D. Font, A. Bastero, S. Saya-
lero, C. Jimeno, M. A. Pericàs, Org. Lett. 2007, 9,
ASAP.
[13] B. Glaeser, H. Kunz, Synlett 1998, 53–54.
[14] a) M. Pastó, B. Rodríguez, A. Riera, M. A. Pericàs, Tet-
rahedron Lett. 2003, 44, 8369–8372; b) E. Medina, A.
Vidal-Ferran, A. Moyano, M. A. Pericàs, A. Riera, Tet-
rahedron: Asymmetry 1997, 8, 1581–1586; c) M. Alcón,
A. Moyano, M. A. Pericàs, A. Riera, Tetrahedron:
Asymmetry 1999, 10, 4639–4651; d) U. M. Lindstrom,
B. Olofsson, P. Somfai, Tetrahedron Lett. 1999, 40,
9273–9276; e) G. Sabitha, B. V. Subba Reddy, S. Abra-
ham, J. S. Yadav, Green Chem. 1999, 1, 251–252; f) B.
Charrada, A. Hedhli, A. Baklouti, Tetrahedron Lett.
2000, 41, 7347–7349; g) E. Medina, A. Moyano, M. A.
Pericàs, A. Riera, Helv. Chim. Acta 2000, 83, 972–988;
h) L. Favretto, W. A. Nugent, G. Licini, Tetrahedron
Lett. 2002, 43, 2581–2584; i) B. Olofsson, P. Somfai, J.
Org. Chem. 2002, 67, 8574–8583; j) K. B. Lindsay, M.
Tang, S. G. Pyne, Synlett 2002, 731–734.
[15] a) J. Sprinz, M. Kiefer, G. Helmchen, M. Reggelin, G.
Huttner, O. Walter, L. Zsolnai, Tetrahedron Lett. 1994,
35, 1523–1526; b) C. Breutel, P. S. Pregosin, R. Sal-
zmann, A. Togni, J. Am. Chem. Soc. 1994, 116, 4067–
4068; c) J. Herrmann, P. S. Pregosin, R. Salzmann, A.
Albinati, Organometallics 1995, 14, 3311–3318; d) P. S.
Pregosin, R. Salzmann, A. Togni, Organometallics
1995, 14, 842–847; e) P. Barbaro, P. S. Pregosin, R. Sal-
zmann, A. Albinati, R. W. Kunz, Organometallics 1995,
14, 5160–5170; f) P. S. Pregosin, R. Salzmann, Coord.
Chem. Rev. 1996, 155, 35–68; g) H. Steinhagen, M. Re-
[20] R. M. Moreno, A. Bueno, A. Moyano, J. Organomet.
Chem. 2002, 660, 62–70.
[21] For previous work on microwave-promoted allylic alky-
lation with phosphinooxazoline ligands, see: a) N.-F. K.
Kaiser, U. Bremberg. , M. Larhed, C. Moberg, A. Hall-
berg, J.Organomet. Chem. 2000, 603, 2–5; b) U. Brem-
berg, S. Lutsenko, N.-F. K. Kaiser, M. Larhed, A. Hall-
berg, C. Moberg, Synthesis 2000, 1004–1008; c) M.
Larhed, C. Moberg, A. Hallberg, Acc. Chem. Res. 2002,
35, 717–727.
[22] For another kind of P,N -ligands leading to better re-
sults with these systems, see ref.[6a]
[23] a) S. Liu, J. F. K. Muller, M. Neuburger, S. Schaffner,
M. Zehnder, J. Organomet. Chem. 1997, 549, 283–293;
b) S. Schaffner, J. F. K. Mueller, M. Neuburger, M.
Zehnder, Helv. Chim. Acta 1998, 81, 1223–1232; c) M.
Zehnder, S. Schaffner, M. Neuburger, D. A. Plattner,
Inorg. Chim. Acta 2002, 337, 287–298.
[24] The ionic palladium complex 16 with R=Ph was pre-
pared from standard diphenylallyl palladium dimer and
phosphinooxazoline 13a in the presence of NH4PF6.
The 31P NMR spectra of the complex showed six signals
at d=81.7, 40.5, 34.9, 27.8, 24.7, and 23.1 ppm with rel-
ative intensities 0.09:0.06:0.05:0.26:0.1:1.00. The most
abundant isomer was assigned as exo-syn-syn-16 by
Adv. Synth. Catal. 2007, 349, 2265 – 2278
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2277

Dana Popa et al.
FULL PAPERS
comparison of selected 1H-, 13C- and 31P NMR data
with reported values in the literature. The second most
abundant one was assigned as the endo-syn-syn-16 by
the same arguments as before. No further assignments
were possible due to heavy signal overlapping and low
intensity of the signals of the rest of the isomers.
[25] S. Dapprich, I. Komiromi, K. S. Byun, K. Morokuma,
M. J. Frisch, Theochem 1999, 461–462, 1–21.
[28] DFT calculations have also been used to rationalize
changes in enantioselectivity in Pd-catalyzed allylic al-
kylation with hydroxyl-containing ligands. Hydrogen
bonds with Pd(0) as the proton acceptor in the inter-
mediate complexes appear to play an important role in
this case. See ref.[6b]
[29] G. J. Dawson, J. M. J. Williams, S. J. Coote, Tetrahe-
dron: Asymmetry 1995, 6, 2535–2546.
[30] S. L. You, X. Z. Zhu, Y. M. Luo, X. L. Hou, L. X. Dai,
[26] M. J. Frisch et al. Gaussian 03, Rev. C-02; Gaussian,
Inc.: Pittsburgh PA, 2003.
J. Am. Chem. Soc. 2001, 123, 7471–7472.
[31] D. A. Evans, K. R. Campos, J. S. Tedrow, F. E. Michael,
M. R. GagnØ, J. Am. Chem. Soc. 2000, 122, 7905–7920.
[27] The anti-anti diastereomers are predicted to be of very
high energy content and have not been considered (see
ref.[18d]).
2278
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2265 – 2278