Design of ionic phosphites for catalytic hydrocyanation reaction of 3-pentenenitrile in ionic liquids

Article abstract of DOI:10.1002/adsc.200505086

The synthesis and characterization of a novel class of ionic phosphites bearing either a single cationic group obtained by quaternization of aminophosphites or three cationic groups prepared by reaction of phosphorus trichloride with imidazolium phenols are reported. The catalytic hydrocyanation reaction of 3-pentenenitrile (3PN) into adiponitrile has been performed in the presence of Ni(0) with ionic phosphite ligands, and a Lewis acid in biphasic ionic liquid/organic solvent system. The screening of several original cationic phosphites was performed and the experimental conditions were optimized for the tricationic phosphite tris-4-[(2,3-dimethylimidazol-1-yl) methyl]phenyl phosphite tris[bis(trifluoromethylsulfonyl)amide]. It is possible to obtain performance similar to molecular systems and the catalyst and the Lewis acid were immobilized in the ionic phase.

Full text of DOI:10.1002/adsc.200505086

FULL PAPERS
DOI: 10.1002/adsc.200505086
Design of Ionic Phosphites for Catalytic Hydrocyanation
Reaction of 3-Pentenenitrile in Ionic Liquids†
a
a
a
a,
´
Christophe Vallee, Yves Chauvin, Jean-Marie Basset, Catherine C. Santini, *
Jean-Christophe Gallandb,*
Laboratoire de Chimie Organometallique de Surface, UMR 9986 CNRS – ESCPE Lyon, 43 bd du 11 Novembre 1918,
a
´
F-69626 Villeurbanne Cedex, France
Phone: (þ33)-4-7243-1794; Fax: (þ33)-4-7243-1795; e-mail: santini@cpe.fr
b
`
Rhodia Recherches , Centre de Recherches de Lyon, Service CAT, 85 avenue des Freres Perret BP 62, 69192 Saint-Fons
Cedex, France
Fax: (þ33)4-7289-6890, e-mail: jean-christophe.galland@eu.rhodia.com
Received: February 10, 2005; Revised: May 31, 2005; Accepted: June 13, 2005
Abstract: The synthesis and characterization of a nov- original cationic phosphites was performed and the
el class of ionic phosphites bearingeither a single cat-
experimental conditions were optimized for the tri-
ionic group obtained by quaternization of aminophos- cationic phosphite tris-4-[(2,3-dimethylimidazol-1-yl)
phites or three cationic groups prepared by reaction methyl]phenyl phosphite tris[bis(trifluoromethylsul-
of phosphorus trichloride with imidazolium phenols fonyl)amide]. It is possible to obtain performance
are reported. The catalytic hydrocyanation reaction similar to molecular systems and the catalyst and
of 3-pentenenitrile (3PN) into adiponitrile has been the Lewis acid were immobilized in the ionic phase.
performed in the presence of Ni(0) with ionic phos-
phite ligands, and a Lewis acid in biphasic ionic liq- Keywords: adiponitrile; hydrocyanation; ionic liquids;
uid/organic solvent system. The screening of several ionic phosphite; nickel; P ligands
Introduction
methylsulfonyl)imide [BMMI][NTF2]/heptane} in the
presence of Ni(cod)2 and (m-sulfophenyl)-diphenyl-
Adiponitrile [AdN], a nylon 6,6 intermediate is industri- phosphine sodium salt, [TPPMSNa], with a conversion
ally prepared by the hydrocyanation of butadiene[1] of 2M3BN of 96% and selectivity to 3PN of 93%. Parti-
which can be described in three steps.[2]A first hydrogen tion experiments proved that the catalyst was immobi-
cyanide molecule is added to butadiene to yield a mix- lized in the ionic phase. TON (1020) and turn-over fre-
ture of branched and linear pentenenitriles. The quency (TOF) (103 hÀ1) of the catalyst were measured.
branched isomer is then isomerized into the linear
one: this is the second step. Then the addition of a second
hydrogen cyanide molecule leads to AdN, Scheme 1.
The industrial hydrocyanation of butadiene into adi-
ponitrile is catalyzed by homogeneous zero-valent triar-
yl phosphate-nickel complexes. Triaryl phosphites are
also the subject of constant research for hydrocyanation:
new structures are regularly claimed, in particular by
DuPont[3–5] and BASF.[6,7] Other ligands containing
À
P O bonds were used for this reaction, like phosphin-
ites,[8] phosphonites[9,10] and bidentate phosphorus
amide ligands.[11,12]
We have recently reported the first example of the cat-
alytic isomerization of 2-methyl-3-butenenitrile
(2M3BN) into 3-pentenenitrile (3PN) (the second step
in the adiponitrile preparation) under biphasic condi-
Scheme 1. Industrial preparation of AdN by hydrocyanation
of butadiene catalyzed by homogeneous nickel(0)-phosphite
tions {1-butyl-2,3-dimethylimidazolium bis(trifluoro- complexes.[2]
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
1835

´
Christophe Vallee et al.
FULL PAPERS
Recyclingof the catalyst was possible but led to an im-
portant deactivation which could be related to a possible
evolution of the nature of the ligand by ion exchange be-
tween the cation (Naþ) of the ligand and the cation
[BMMIþ] of the ionic liquid. Such an exchange was re-
cently proved by 23Na solid state NMR.[13]
Although phosphite ligands are considered to be eas-
ier to synthesize and less prone to oxidation than phos-
phines, they are more sensitive to hydrolysis. They are
much cheaper than most phosphines and a wide variety
can be obtained. However, their use in ionic liquids was
rarely reported and limited to the hydroformylation of
methyl 3-pentenoate[14] and 1-hexene.[15,16] In order to
immobilize the catalyst in the ionic phase, ionic phos-
phites were prepared and characterized. Applications
of the catalytic systems involvingthese ligands to the hy-
drocyanation of 3PN are described herein.†
Scheme 3. Synthesis of a cationic phosphinite.[21]
Scheme 4. Synthesis of a cationic phosphite.[25]
Results
dium salts of the mono-, di-, and tri-sulfonated triphenyl
Ionic phosphines have largely been described and this phosphite and which were used as ligand for the rhodi-
field has been supported by the successes of aqueous cat- um-catalyzed hydroformylation of 1-hexene in the ionic
alysis.[18] Publications related to ionic phosphites are liquid [BMI][PF6].[16]
much more limited. Indeed these ligands can be used
A cationic phosphinite obtained by reaction of trime-
À
very rarely in water as the P O single bond is likely to thyloxonium tetrafluoroborate with an aminophosphin-
be easily hydrolyzed. However non-aqueous ionic liq- ite (Scheme 3) was prepared in situ and directly reacted
uids could enhance the attractiveness of such ligands. with rhodium in a reaction of catalytic asymmetric hy-
Recently, we have demonstrated that the presence of drogenation.[21]
an ionic group on the aromatic ring of a phosphine im-
The only cationic phosphite described in the litera-
proved the immobilization of the catalyst in the ionic liq- ture was obtained by reaction of methyl iodide with
uid phase.[17] As a consequence, the synthesis of triaryl the tris-[3-(N,N-dimethylamino)-phenyl] phosphite to
phosphites containingat least one ionic rgoup has
been developed to reach the same goal.
form the ammonium compound (Scheme 4).[25] This
compound was not characterized further than the deter-
mination of its meltingpoint.
To conclude, the field of ionic ligands containing P O
single bonds has received very little attention. While a
few examples of anionic phosphites have been reported
À
Synthesis of Ionic Aryl Phosphites
The anionic aryl phosphites reported in the litera- only a single example of a cationic phosphinite was
ture[19–23] were generally a mixture of ammonium or so- found. To our knowledge, cationic phosphites have nev-
er been mentioned in the literature as ligands for orga-
nometallic catalysis. The development of non-aqueous
ionic liquids could enhance the attractiveness of such li-
gands. Special efforts were therefore devoted to their
synthesis.
Note that all our attempts for the synthesis in large
quantities of pure anionic phosphite and cationic phos-
phinite failed. In contrast, several original tricationic
aryl phosphites were efficiently prepared.
Two different routes were considered for the synthesis
of cationic phosphites. The first one required the syn-
thesis of an aminoaryl phosphite and subsequent quater-
nization of the amino group with an appropriate re-
agent. The second possibility begun with the prepara-
tion of a phenol with an ammonium group and was pur-
À
sued by reaction of this compound with a P Cl bond.
Tetraalkylammonium and imidazolium were chosen as
Scheme 2. Synthesis of
a mixture of sulfonated phos-
phites.[23,24]
1836
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
cationic groups, and our work was limited to triaryl
phosphites.
Quaternization of an Aminoaryl Phosphite
The biphenyl chlorophosphite 1 prepared in good yields
from 2,2’-biphenol and phosphorus trichloride reacted
with 3-N,N-dimethylaminophenol in the presence of
triethylamine. After filtration of triethylamine hydro-
chloride salt, the pure aminoaryl phosphite 2 was isolat-
ed (Scheme 5). A suspension of trimethyloxonium tetra-
fluoroborate in dichloromethane was added to this com-
pound. After solvent concentration, the ammonium
phosphite 3 was obtained quantitatively. The product
Scheme 5. Synthesis of a cationic phosphite.
was fully characterized by NMR spectroscopy (1H, 13
and 31P) and elemental analysis.
C
The versatility of this new synthetic route was assessed
by the preparation of several ligands having various
electronic and steric properties.
A sterically hindered 2,2’-(3,3’,4,4’-tetramethyl-6-6’-
di-tert-butyl)-biphenyl chlorophosphite 5 was prepared
from the correspondingsubstituted biphenol 4 and
phosphorus trichloride (Scheme 6). Bis-m-tolyl chloro-
phosphite 6 was furnished by Rhodia.
The third aryl oxide group of the phosphite was de-
rived from 4-imidazol-1-yl-phenol 7 and from either 3-
or 4-(N,N-dimethylaminomethyl)-phenol (8 or 9),
which were prepared from 3-or 4-hydroxybenzaldehyde
by a reductive amination with dimethylamine and so-
dium triacetoxyborohydride in THF[26] (Scheme 7).
Several aminoaryl phosphites were prepared by com-
bination of these chlorophosphites and aminophenols
and were obtained in modest to good yields (Scheme 8).
Compounds 10, 11 and 12 were purified through a thin
silica column (eluent dichloromethane) and the phos-
phite 13 was purified through a thin Florisilꢁ column
(eluent dichloromethane). Crude phosphites 14 and 15
were sufficiently pure to be used directly for the subse-
quent step. The amino groups in these compounds
were then transformed into ammonium groups by reac-
tion with trimethyloxonium tetrafluoroborate. The cat-
ionic phosphites were obtained in high yields
(Scheme 8).
Scheme 6. Chlorophosphites.
The ligands 16 to 21 were obtained as oils containing
small amounts of hydrolysis products [(RO)2P(O)H],
phosphonium salts [(RO)3PMeþBF4À] and compounds Scheme 7. Aminophenols.[26]
resultingfrom redistribution of the phenols around
phosphorus. All attempts at purification were unsuc-
cessful and mainly led to degradation products. As a 22 was obtained as a TF2NÀ salt by reaction of aminoaryl
consequence, crude products were engaged in catalytic phosphite 2 with LiTF2N in acetonitrile in the presence
tests and the purity of the compounds was estimated of dimethyl sulfate (Scheme 9).
1
by H, 13C and 31P NMR spectroscopy. For all of these
A family of phosphites with three amino groups was
synthesized by reactingphosphorus trichloride with 3-
phosphites, the anion was tetrafluoroborate.
As [BMMI][TF2N] was the best ionic liquid for the N,N-dimethylaminophenol in the presence of triethyl-
studied catalytic reaction, we developed ligands bearing amine (Scheme 10). The phosphite 23 was purified by
the anionic TF2NÀ moiety. The ammonioaryl phosphite filtration on silica and then submitted to various alkylat-
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1837

´
Christophe Vallee et al.
FULL PAPERS
Scheme 9. Cationic phosphite with a TF2NÀ anion.
Scheme 10. Synthesis of tri-ammonioaryl phosphites.
Scheme 8. Aminoaryl phosphates and cationic phosphates.
ingreagents. Trimethyloxonium tetrafluoroborate, tri-
ethyloxonium hexafluorophosphate and methyl tri-
fluoromethanesulfonate led to complex mixtures con-
tainingpartially quaternized phosphites and several
Scheme 11. Synthesis of tricationic phosphites.
lem, the quaternization of the startingphenol was per-
phosphonium derivatives. An equimolar dimethyl sul- formed before reaction with phosphorus derivatives.
fate/LiTF2N mixture gave the expected products 24.
Duringthis work, we were able to synthesize several
cationic phosphites bearing a single cationic group Reaction of a Cationic Phenol with a P Cl Bond
with good to excellent purity. On the other hand, at-
À
tempts at the synthesis of tri-cationic ligands led to Suitable phenols were prepared in two steps: 2-methyl-
only one ligand of moderate purity. To avoid this prob- imidazole was heated with 4-hydroxybenzyl alcohol to
1838
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
Scheme 12. Summary of the two routes of synthesis and list of synthesized phoshites (with purity and overall yield).
À
yield product 25 (Scheme 11)[27] which was then quater- TF2NÀ anions (28) or BF4 anions (29) were obtained
nised by reaction with LiTF2N (26) or NaBF4 (27)in the in good yields and characterized by NMR spectroscopy
presence of dimethyl sulfate. Reaction of phosphorus (1H, 13C and 31P in acetonitrile-d3) and elemental analy-
trichloride with these imidazolium phenols yielded tri- sis. It was possible to prepare 28 on a 20-gscale.
cationic phosphites and triethylamine hydrochloride
To conclude, two routes of synthesis of cationic phos-
salts (Scheme 12). Chloride salts were eliminated by phites were successfully developed. In the first one, an
successive washings with dichloromethane as the phos- aminoaryl phosphite was synthesized and the amino
phites were not soluble in this solvent. Phosphites with group was transformed into ammonium by reaction
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1839

´
Christophe Vallee et al.
FULL PAPERS
with an alkylation reagent: this method was particularly
suitable for the synthesis of a monocationic phosphite.
The second strategy began with preparation of phenols
bearingan imidazolium group. The reaction with phos-
phorus trichloride led to tricationic phosphites in high
yield and purity. These results are summarized on
Scheme 12.
(1)
(2)
Cationic Biphenyl Phosphites
Phosphites 3, 16, 17 and 22 which have the same struc-
Catalytic Studies
The hydocyanation of 3PN was realized in the presence ture, were tested both for isomerization of 2M3BN
of catalytic precursor Ni(0)-triphenyl phosphite and into 3PN and for the hydrocyanation of 3PN in adiponi-
Lewis acid promoter which affect the double bond iso- trile, but, in both reactions, no activity was observed.
merization of 3PN to 4-pentenenitrile (4PN) concur-
A31P NMR spectrum of the reaction mixture containing
rently with the selective addition of HCN to 4PN. By- Ni(cod)2/phosphite 3/2M3BN/[BMMI][TF2N] after reac-
products include 2-methylglutaronitrile (MGN), ethyl- tion and work-up was recorded at room temperature.
succinonitrile (ESN)arising, respectively, from Markov- Two peaks were observed at d¼141.6 and 157.5 ppm
nikov addition to 4PN, and direct addition of HCN to (Figure 1: spectrum III). The peak at d¼141.6 ppm could
3PN.[2]
be attributed to 3 (d¼138.8 ppm in CDCl3) (Fig ure 1:
The catalytic results were given in conversion of 3PN, spectrum I). On the other hand, when a mixture of
yield and linearity in dinitriles, data determined as de- Ni(cod)2 and 4 equivalents of ligand 3 was stirred at
picted in Equations (1) and (2):
room temperature in THF, a white precipitate was formed
whose 31P NMR spectrum (Figure 1: spectrum II) exhibit-
ed a single peak at d¼157.5 ppm attributed to an NiL4
complex. Thus the spectrum of the catalytic solution after
reactioncould be interpretedascontainingmainly an NiL4
complex and free ligand. This result supported the fact
that the complex was stable in the presence of penteneni-
triles and ionic liquid preventingany coordination of an
olefin to the nickel and preventingany catalytic reaction.
Cationic Tolyl Phosphites
In a typical run, acetone cyanohydrin (24 equivalents)
was slowly added at 708C in 3 h to a solution of Ni(cod)2
Scheme 13. Formation of dinitrile derivatives duringthe hy-
drocyanation reaction of 3PN.[2]
Figure 1. 31P NMR spectra registered at room temperature. (I): 3 in CDCl3; (II): Ni(cod)2 þ4 equivs. of 3 in CD3CN; (III):
Ni(cod)2 þ5 equivs. of 3þ3PNþ[BMMI][TF2N] after 3 h at 1008C.
1840
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
Table 1. Hydrocyanation of 3PN with cationic phosphites in [BMMI][TF2N].
Ligand
Conv. 3PN
Yield of dinitriles [%]
Linearity
84
TPP
18
49
46
24
15
28
21
70
75
19
20
11
7
30
38
76
77
77
35[a]
44
21
[a]
The mixture of nickel, phosphite, substrate and ionic liquids was stirred overnight at room temperature in order to form the
nickel-phosphite complex.
(40 mg), cationic phosphites 18–21 (5 equivalents), 3PN temperature of the catalytic reaction in the isomeriza-
(30 equivalents) and zinc chloride (1 equivalent) in tion of 2M3BN has been already mentioned.[17] The hy-
[BMMI][TF2N]. After 3 h of reaction, products were an- drocyanation reaction is also sensitive to experimental
alyzed by gas chromatography. Results obtained with parameters, i.e., the preparation of the catalytic mixture
the molecular ligand tris(meta-tolyl) phosphite (m- had a strongeffect on catalytic performances in the brief
TTP) are included as a base for comparison (Table 1). study of the monocationic ligands. As the phosphite 28
Note that the performance of the catalytic system in- was synthesized on a 20-gscale, it was decided to further
cludingthe cationic phosphite 21 was the closest to explore this parameter space usingan established design
that of the molecular m-TTP based system.
of experiments technique (DOE).[28]
With one ligand (Table 1), the hydrocyanation reac-
For this study, the influence of two different factors at
tion was performed after a different time of reaction be- different levels were retained, in first the ratio of ligand
tween 20 and Ni(COD)2 (20 min and 12 h); it appeared per nickel: a value of 2, which corresponds to a minimum
that when the ligand/nickel system was allowed to react number of ligands necessary to stabilize nickel(0), and
for a longer time before reaction, superior performance one of 5 (an excess of ligand). The second factor was
was observed. It thus appeared that disparities due to ex- the ratio of zinc chloride per nickel: one to three equiv-
perimental conditions were much more important that alents of Lewis acid per nickel were used. The factors
those due to the structure of the phosphites.
The cationic phosphite 21 yielded very promisingresults
which were studied and the level which were chosen
are depicted in Table 2. The responses of the DOE
for the hydrocyanation of 3PN in ADN. However, the pos- were the yield in dinitriles and the linearity.
sibility of the redistribution of phenol around the phos-
A full factorial design with a center point (L/Ni¼3.5
phorus atom could not be excluded. Such a redistribution and ZnCl2/Ni¼2) was chosen. Two reproductions of
might have a dramatic importance, as it was likely to gen- the DOE were performed, with a blockingon the prep-
erate in situ molecular m-TTP ligand which might be re- aration of the catalyst.
sponsible for the catalytic activity. In order to eliminate
this alternative, a tricationic ligand was tested in catalysis.
Table 2. Parameters of the experimental design.
Tricationic Phosphite 28
Level À1
Level þ1
Ratio L/Ni
Ratio ZnCl2/Ni
2
1
5
3
The importance of experimental parameters such as the
ligand to metal ratio, the presence of Lewis acid and the
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1841

´
Christophe Vallee et al.
FULL PAPERS
a detrimental impact on the yield and this factor effect
could explain 16% of the variations.
Globally, the two main factors (ZnCl2/Ni ratio and li-
gand/nickel ratio) were responsible for more than 80%
of yield variations and the restingpart could be attribut-
ed to background noise. It was an indication that the crit-
ical factors of the reaction had been identified. The yield
can be represented as a function of the two significant
factors in a three-dimensional graph (Figure 2).
The analysis of the DOE in terms of the second re-
sponse, the linearity, was also performed although the
variations of this response were limited. The ratio of
zinc chloride per nickel was the only statistically signifi-
cant factor and it explained only a small part (31%) of
the variation of linearity. As a consequence, it was not
Figure 2. Yield in dinitriles as a function of L/Ni and Zn/Ni.
Each reproduction was divided in two blocks depend- possible to establish a robust predictive model for the
ingon the catalyst preparation procedure. In the first
linearity.
block the trials were run after in situ catalyst preparation
To conclude, the DOE allowed the optimization of the
whereas those of the second block were run with a pre- reaction conditions with respect to two factors previous-
formed catalyst (see Table 1). If the variability between ly identified as havingan important effect. A low ratio of
the two blocks is significant, the preparation of the cat- ligand per nickel and a large excess of Lewis acid are
alyst will be identified as a potential factor. Otherwise necessary to obtain high yields. Furthermore, it has
the reproduction of the tests will provide information been proved that there was no significant interaction be-
about the reproducibility of the tests.
Concerningthe first response, dinitriles yields, neither
tween the two factors. Further reduction of the number
of ligands per nickelseemed useless, as it would certainly
the interactions between L/Ni and ZnCl2/Ni nor the lead to unstable catalysts. In contrast, it was possible to
block (the preparation of the nickel complex) had signif- increase the proportion of zinc chloride. A few tests
icant effects. The situation was therefore clearly differ- were then performed with 5 or 10 equivalents of Lewis
ent of that with the monocationic ligand 20, as the prep- acid per nickel (Table 4).
aration of the catalyst had a strongimpact on the catalyt-
When 5 equivalents of zinc chloride per nickel were
ic performance in the latter case. However, the origin of used (entry 401), the results were very similar to the
this difference was not understood and would require one with 3 equivalents (entry 398). When a further ex-
further investigation.
Concerningthe principal effects, the first factor,
cess of Lewis acid was employed (entries 402 and 403),
lower activity resulted. The large excess of Lewis acid
(ZnCl2/Ni ratio) had a significant consequence on the might be negative for the stability of the nickel complex.
yield which increased when an excess of Lewis acid By optimization of experimental conditions with zinc
was used, and 65% of the yield variations could be at- chloride, it has been possible to obtain high yields of di-
tributed to this factor effect (more precisely, the factor nitriles with phosphite 28. The highest linearity obtained
variance was 65% of total variance).
was 73% and thus further improvements were required.
The second factor, (ligand/nickel ratio) had a signifi- To this end, different Lewis acids were tested as co-cat-
cant but less important effect. An excess of ligand had alysts (Table 5).
Table 3. Hydrocyanation of 3PN with phosphite 30 in [BMMI][TF2N].
Experiments order
L/Ni
Lewis acid/Ni
Blocks[a]
Yield of dinitriles*
Linearity*
1
2
3
4
5
6
7
8
9
5
3.5
5
2
2
2
2
5
3.5
5
1
2
3
3
1
3
1
3
2
1
a
a
a
a
a
b
b
b
b
b
4
4
17
36
6
34
8
19
22
2
65
61
71
71
65
72
71
71
73
50
10
[a]
Block a means in situ formation of the catalyst while block b means reaction with preformed catalyst (see Table 1).
1842
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
Table 4. Hydrocyanation of 3PN with phosphite 28 in [BMMI][TF2N].
Entry
L/Ni
ZnCl2/Ni
Conv. 3PN
Yield of dinitriles
Linearity
398
401
402
403
2
2
2
5
3
5
10
10
40
39
12
17
36
34
7
71
70
67
71
13
Table 5. Hydrocyanation of 3PN with phosphite 28 and different Lewis acids.
Lewis acid
Entry
Conv. 3PN
Yield of dinitriles
Linearity
Yield of
other products
ZnCl2
BPh3
CoCl2
In(CF3CO2)3
In(CF3SO3)3
ErCl3
398
404
407
405
409
406
408
40
11
21
24
7
36
5
15
19
2
71
74
65
59
53
80
81
4
3
3
3
3
4
3
32
19
29
15
YCl3
Triphenylborane led to a low activity and no signifi- high quantity with excellent purity, were original ones,
cant improvement of the linearity was observed (en- and they are likely to offer new opportunities for catal-
try 404). With cobalt chloride (entry 407), both yield ysis in ionic liquids.
and linearity were inferior to those obtained with zinc
The hydrocyanation of 3PN in AdN was considered in
chloride. Indium-based Lewis acids gave different re- this work. The dramatic importance of experimental
sults in terms of activity, but the linearity was systemati- conditions has been highlighted. The importance of
cally inferior to 60% (entries 405 and 409). Rare earth the experimental conditions is such that the comparison
chlorides yielded the best results, with linearities ex- of data which are not obtained in rigorously identical
ceeding80% (entries 406 and 408).
It was difficult to further interpret these results. In- ture of the ligand. The screening of several original cat-
deed, it is worth rememberingthat a variety of param- ionic phosphites was performed and the experimental
conditions might be misleading for the choice of the na-
eters have been shown to be important in a complex conditions were optimized for tris-4-[(2,3-dimethylimi-
manner. They were optimized for zinc chloride but these dazol-1-yl)methyl]phenyl phosphite tris[bis(trifluoro-
conditions were not necessarily optimal for other Lewis methylsulfonyl)amide]. It was possible to obtain per-
acids. Furthermore, the solubility of the Lewis acids in formance similar to molecular systems for the hydrocya-
[BMMI][TF2N] has not been studied, and the exact na- nation reactions [conversion of 3PNꢀ40% ;yield of
ture of the Lewis acid promoter in the reaction medium dinitrilesꢀ40% and linearity>70%]. The cationic
is unknown.
phosphite-based system might afford the supplementary
For all these reasons, it was not possible to interpret advantage of the immobilization of the catalyst and the
these data from a mechanistic point of view. It is, howev- Lewis acid in the ionic phase when the reaction is run un-
er, important to note that high activity and linearity have der these biphasic conditions.
been obtained with lanthanides chlorides. Even if these
co-catalysts are somewhat more expensive than zinc
chloride, the recyclingof the co-catalyst thanks to ionic
liquids might improve the economic attractiveness of
Experimental Section
the use of such compounds in an industrial context.
Syntheses and Characterization of Cationic Ligands
All the synthesis were carried out under dry argon using stand-
Conclusion
ard Schlenk techniques. Solvents were distilled usingthe ap-
propriate dryingagents: CaH 2 for acetonitrile and dichlorome-
By quaternization of aminoaryl phosphites, it was possi-
thane, sodium/benzophenone for diethyl ether and tetrahydro-
ble to isolate in good yields ligands bearing a single cat-
ionic group. Tricationic aryl phosphites were efficiently
prepared by reaction of phosphorus trichloride with imi-
furan, K2CO3 for chloroform, sodium for toluene and sodium/
potassium for pentane.
Phosphorus trichloride and diphenylchlorophosphine were
dazolium phenols. All these compounds, obtained in purchased from Aldrich and distilled before use. Bis-m-tolyl
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1843

´
Christophe Vallee et al.
FULL PAPERS
chlorophosphite was supplied by Rhodia and used as is. Trie-
thylamine was purchased from Aldrich and distilled over calci-
um hydride.
uum to afford a purple foam; yield: 640 mg(99%). 1H NMR
(CDCl3): d¼3.49 (s, 9H, CH3), 7.04–7.63 (m, 12 H, CH);
13C{1H} NMR (CDCl3): d¼57.2 (CH3), 112.6 (d, JCP
¼
Classical neutral alumina for chromatography (Fluka), silica
gel 60 (Merck) or Florisilꢁ were used for purification of the li-
gands. All other reagents were obtained from commercial sour-
ces and used as received.
5.1 Hz), 115.8, 122.3, 122.5 (d, JCP ¼9.2 Hz), 126.0, 129.8,
130.2, 130.8 (d, JCP ¼2.8 Hz), 132.2, 147.9, 148.7 (d, JCP
¼
5.1 Hz), 153.8 (d, JCP ¼4.1 Hz); 31P{1H} NMR (CDCl3): d¼
138.8; anal. calcd. for C21H21BF4NO3P: C 55.66, H 4.67, N
3.09; found: C 56.02, H 4.57, N 2.97; MS (ESI): m/z¼366
(C21H21NO3Pþ).
1H, 13C and 31P NMR were obtained on a Bruker AC
300 MHz spectrometer. Chemical shifts were measured rela-
tive to SiMe4 as internal standard for 1H and 13C. Positive 31
P
2,2’-(3,3’,4,4’-Tetramethyl-6,6’-di-tert-butyl)biphenyl
Chlorophosphite (5) by Procedure A: From phosphorus tri-
chloride (0.75 mL, 8.5 mmol), triethylamine (2.35 mL,
16.9 mmol) , 2,2’-(3,3’,4,4’-tetramethyl-6-6’-di-tert-butyl)-bi-
phenol (2.0 g, 5.64 mmol); slightly yellow solid; yield: 2.33 g
(99%); 31P{1H} NMR (CD3Cl): d¼165.7.
chemical shifts are downfield from external 85% H3PO4. The
solvents used (CDCl3, CD3CN, DMSO-d6) were purchased
from SDS and distilled usingthe appropriate dryingagent.
Elemental analyses for carbon, hydrogen and nitrogen were
`
`
performed by the “Laboratoire de Synthese et Electrosynthese
´
Organometallique” at Dijon. All other elemental analyses
were performed by the “Service Central dꢂAnalyse” of the
CNRS at Solaize.
3-[(N,N-Dimethylamino)-methyl]-phenol (8)
3-Hydroxybenzaldehyde (0.977 g, 8.00 mmol) and dimethyla-
mine (2.0 M solution in THF, 6 mL, 12.00 mmol) were dis-
solved in THF (10 mL). Sodium triacetoxyborohydride
(2.544 g, 12.00 mmol) was added and the mixture was stirred
at room temperature for 18 h. A saturated aqueous NaHCO3
solution was added and solvents were removed under vacuum
to yield a white solid. The solid was extracted with dichlorome-
thane and the solution acidified with hydrogen chloride (1.0 M
in diethyl ether, 8 mL, 8 mmol). A white precipitate appeared
and was washed with dichloromethane. The solid was then dis-
solved in saturated aqueous NaHCO3 solution and the basic
phase (pH¼9) obtained was extracted with dichloromethane.
Organic phases were dried on magnesium sulfate and solvent
was removed under vacuum to give a slightly yellow solid;
yield: 0.92 g(76%). 1H NMR (CDCl3): d¼2.27 (s, 6H, CH3),
General Procedure A: 2,2’-Biphenyl Chlorophosphite
(1)
2,2’-Biphenol (2.00 g, 10.7 mmol) was dissolved in toluene
(25 mL) and the solution was cooled to 08C. A solution of
phosphorus trichloride (2.21 g, 16.0 mmol) and triethylamine
(4.5 mL, 32.0 mmol) in toluene (10 mL) was slowly added.
The mixture was allowed to warm to room temperature and
stirred for 36 h. It was filtered and concentrated to afford a
slightly yellow oil; yield: 2.55 g (95%). 31P{1H} NMR (CDCl3):
d¼178.9.
General Procedure B : 2,2’-Biphenyl-3-(N,N-
dimethylamino)-phenyl Phosphite (2)
3.42 (s, 2H, CH2), 6.70–6.80 (m, 3H, CH), 7.14 (t, 1H, 3JHH
¼
7.8 Hz, CH); 13C{1H} NMR (CDCl3): d¼44.8 (CH3), 63.9
(CH2), 115.4, 117.1, 121.3, 129.4, 138.5 (CCH2), 157.0 (CO).
4-[(N,N-Dimethylamino)-methyl]-phenol (9): Usingthe
same procedure as for 8: from 4-hydroxybenzaldehyde
(0.977 g, 8.00 mmol), dimethylamine (2.0 M solution in THF,
6 mL, 12.00 mmol), THF (10 mL), sodium triacetoxyborohy-
dride (2.544 g, 12.00 mmol) to afford a slightly yellow solid
which was recrystallized in a toluene-hexane mixture (1/1) to
2,2’-Biphenyl chlorophosphite (1; 2.55 g, 10.2 mmol) and trie-
thylamine (4.3 mL, 31 mmol) were dissolved in toluene
(10 mL) and the solution was cooled to 08C. A solution of 3-
(N,N-dimethylamino)-phenol (1.40 g, 10.2 mmol) in toluene
(10 mL) was slowly added. The mixture was allowed to warm
to room temperature and stirred for 12 h. The product was pu-
rified through a short silica column (eluent: dichloromethane)
1
1
to afford a purple solid; yield: 1.89 g(53%). H NMR (CDCl3):
give a white solid; yield: 400 mg (33%). H NMR (CDCl3):
d¼2.95 (s, 6H, CH3), 6.47–6.63 (m, 2H), 7.14–7.53 (m, 10H);
d¼2.29 (s, 6H, CH3), 3.42 (s, 2H, CH2), 6.0 (br s, 1H, OH),
6.60–6.70 (m, 2H, CH), 7.05–7.15 (m, 2H, CH); 13C{1H}
NMR (CDCl3): d¼44.7 (CH3), 63.5 (CH2), 115.6, 128.3
(CCH2), 130.9, 156.2 (CO).
13C{1H} NMR (CDCl3): d¼40.6 (CH3), 104.7 (d, JCP
¼
7.7 Hz), 107.1 (d, JCP ¼9.2 Hz), 108.7, 122.3, 125.5, 129.3,
130.1, 130.1, 131.4 (d, JCP ¼3.2 Hz), 149.1 (d, JCP ¼5.1 Hz),
152.1, 152.7 (d, JCP ¼8.3 Hz); 31P{1H} NMR (CDCl3): d¼
143.6. HRMS (CI): m/z¼352.1129; calcd. for C20H19NO3Pþ:
352.1125.
2,2’-Biphenyl-4-(imidazol-1-yl)-phenyl Phosphite (10) by
Procedure B: From 4-(imidazol-1-yl)-phenol (1.07 g,
6.66 mmol), triethylamine (2.8 mL, 20 mmol), toluene
(5 mL), 2,2’-biphenyl chlorophosphite (1; 2.01 g, 7.00 mmol)
solution in toluene (10 mL); white solid; yield: 1.25 g(50%).
1H NMR (CDCl3): d¼7.15–7.81 (m, 15H); 13C{1H} NMR
(CDCl3): d¼118.7, 122.0 (d, JCP ¼7.4 Hz), 122.1, 123.2, 125.8,
129.5, 130.3, 130.5, 131.1 (d, JCP ¼3.4 Hz), 133.9, 135.8, 148.9
(d, JCP ¼5.5 Hz), 151.1 (d, JCP ¼7.1 Hz); 31P{1H} NMR
(CDCl3): d¼141.6.
General Procedure C: 2,2’-Biphenyl-3-(N,N,N-
trimethylammonio)-phenyl Phosphite
Tetrafluoroborate (3)
A solution of 2,2’-biphenyl-3-(N,N-dimethylamino)-phenyl
phosphite (2; 500 mg, 1.42 mmol) in dichloromethane (5 mL)
was quickly added to a suspension of trimethyloxonium tetra-
fluoroborate (210 mg, 1.42 mmol) in dichloromethane. The
mixture was stirred for 1 h and solvent was removed under vac-
2,2’-(3,3’,4,4’-Tetramethyl-6,6’-di-tert-butyl)-biphenyl-3-
(N,N-dimethylamino)-phenyl Phosphite (11) by Procedure B:
From 3-(N,N-dimethylamino)-phenol (751 mg, 5.45 mmol),
triethylamine (2.27 mL, 16.35 mmol), toluene (5 mL), 2,2’-
1844
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
(3,3’,4,4’-tetramethyl-6,6’-di-tert-butyl)-biphenyl chlorophos-
phite (5; 2.36 g, 5.64 mmol) solution in toluene (10 mL); purple
oil; yield: 2.59 g(91%). 1H NMR (CDCl3): d¼1.48, 1.49 [2 s,
18H, C(CH3)3], 1.84, 185, 2.25, 2.26 (4 s, 12H, CCH3), 2.91 (s,
6H, NCH3), 6.40–6.55 (m, 2H), 7.05–7.30 (m, 4H); 13C{1H}
NMR (CDCl3): d¼16.8, 17.1, 31.4, 31.9, 34.8, 35.0, 40.6, 104.6
(d, JCP ¼8.4 Hz), 108.0 (d, JCP ¼9.7 Hz), 108.2, 127.9, 128.4,
129.9, 130.6, 132.2, 132.1, 132.9, 134.4, 135.3, 138.2, 138.2,
144.9, 144.6, 151.9, 153.6 (d, JCP ¼10.2 Hz); 31P{1H} NMR
(CDCl3): d¼133.5.
JCP ¼7.9 Hz), 124.1, 124.6, 126.0, 129.6, 130.3, 130.6, 131.0 (d,
JCP ¼2.8 Hz), 135.4, 148.7 (d, JCP ¼5.1 Hz); 31P{1H} NMR
(CDCl3): d¼141.3; HR-MS (LSI): m/z¼419.1526; calcd. for
C24H24N2O3Pþ: 419.1525.
2,2’-(3,3’,4,4’-Tetramethyl-6,6’-di-tert-butyl)-biphenyl-3-
(N,N,N-trimethylammonio)-phenyl Phosphite Tetrafluorobo-
rate (17) by Procedure C: From 2,2’-(3,3’,4,4’-tetramethyl-6,6’-
di-tert-butyl)-biphenyl-3-(N,N-dimethylamino)-phenyl phos-
phite (11; 520 mg, 1.00 mmol), trimethyloxonium tetrafluoro-
borate (148 mg, 1.00 mmol), dichloromethane (10 mL), brown
solid. 1H NMR (CDCl3): d¼1.38, 1.40 [2 s, 18H, C(CH3)3], 1.81,
1.83, 2.25, 2.27 (4 s, 12H, CCH3), 3.54 (s, 9H, NCH3), 6.90–7.55
(m, 6H); 31P{1H} NMR (CDCl3): d¼134.4.
Bis-(m-tolyl)-3-(N,N-dimethylamino)-phenyl
Phosphite
(12) by Procedure B: From 3-(N,N-dimethyamino)-phenol
(930 mg, 6.78 mmol), triethylamine (2.83 mL, 20.3 mmol), tol-
uene , bis-(m-tolyl) chlorophosphite (2.00 g, 7.12 mmol) solu-
tion in toluene (10 mL); viscous purple oil; yield: 580 mg
(22%); 1H NMR (CDCl3): d¼2.32 (s, 6H, CCH3), 2.89 (s, 6H,
NCH3), 6.40–7.25 (m, 12H, CH); 13C{1H} NMR (CDCl3): d¼
21.5, 40.5, 104.9 (d, JCP ¼6.5 Hz), 108.3 (d, JCP ¼7.3 Hz),
108.5, 117.9 (d, JCP ¼7.8 Hz), 121.6 (d, JCP ¼6.9 Hz), 125.0,
129.4, 130.0, 139.9, 151.7 (d, JCP ¼3.2 Hz), 152.0, 152.8 (d,
JCP ¼5.1 Hz); 31P{1H} NMR (CDCl3): d¼128.9.
Bis-(m-tolyl)-3-(N,N,N-trimethylammonio)-phenyl Phos-
phite Tetrafluoroborate (18) by Procedure C: From bis-(m-tol-
yl)-3-(N,N-dimethylamino)-phenyl phosphite (12; 290 mg,
0.76 mmol), trimethyloxonium tetrafluoroborate (112 mg,
0.76 mmol), dichloromethane (10 mL); purple product; yield:
329 mg(89%). 1H NMR (CDCl3): d¼2.31 (s, 6H, CCH3),
3.58 (s, 9H, NCH3), 6.85–7.65 (m, 12H, CH); 13C{1H} NMR
(CDCl3): d¼21.5 (CCH3), 57.3 (NCH3), 112.5 (d, JCP
¼
5.1 Hz), 115.6, 117.4 (d, JCP ¼7.4 Hz), 121.2 (d, JCP ¼6.9 Hz),
125.6, 129.5, 129.8, 132.2, 140.4, 147.8, 151.2 (d, JCP ¼4.1 Hz),
152.8; 31P{1H} NMR (CDCl3): d¼134.4; HR-MS (LSI): m/z¼
410.1877; calcd. for C24H29NO3Pþ: 410.1885.
Bis-(m-tolyl)-4-(imidazol-1-yl)-phenyl Phosphite (13) by
Procedure B: From 4-(imidazol-1-yl)-phenol (1.01 g,
6.32 mmol), triethylamine (2.63 mL, 18.96 mmol), toluene
(5 mL), bis-(m-tolyl) chlorophosphite (1.86 g, 6.64 mmol) sol-
ution in toluene (10 mL); colorless oil; yield: 1.35 g(53%); 1H
NMR (CDCl3): d¼2.33 (s, 6H, CH3), 6.85–7.80 (m, 15H,
CH). 13C{1H} NMR (CDCl3): d¼21.5, 117.6 (d, JCP ¼7.4 Hz),
118.7, 121.3 (d, JCP ¼6.9 Hz), 122.3 (d, JCP ¼6.4 Hz), 123.1,
Bis-(m-tolyl)-4-(3-methylimidazol-1-yl)-phenyl
Phosphite
Tetrafluoroborate (19) by Procedure C: From bis-(m-tolyl)-4-
(imidazol-1-yl)-phenyl phosphite(13;1.305 g, 3.23 mmol),trime-
thyloxonium tetrafluoroborate (430 mg, 2.90 mmol), dichloro-
methane (10 mL); colorless oil; yield: 1.47 g(99%). 1H NMR
(CDCl3): d¼2.32 (s, 6H, CCH3), 4.03 (s, 3H, NCH3), 6.85–7.55
(m, 15H, CH); 13C{1H} NMR (CDCl3): d¼21.5, 36.7, 117.5 (d,
JCP ¼6.9 Hz), 121.3 (d, JCP ¼7.0 Hz), 122.7 (d, JCP ¼6.4 Hz),
121.4, 123.9, 124.7, 125.5, 129.7, 130.6, 135.1, 140.3, 151.3 (d,
JCP ¼4.6 Hz), 153.1; 31P{1H} NMR (CDCl3): d¼126.3; HR-MS
(LSI): m/z¼410.1887; calcd. for C24H29NO3Pþ: 410.1885.
Bis-(m-tolyl)-3-[(N,N,N-trimethylammonio)-methyl]-phe-
nyl Phosphite Tetrafluoroborate (20) by Procedure C: From
bis-(m-tolyl)-3-((N,N-dimethylamino)-methyl)-phenyl phos-
phite (14; 1.526 g, 3.86 mmol), dichloromethane (10 mL), a
suspension of trimethyloxonium tetrafluoroborate (243 mg,
3.67 mmol) in dichloromethane (10 mL); colorless oil; yield:
1.83 g(99%). 1H NMR (CDCl3): d¼2.32 (s, 6H, CCH3), 3.04
(s, 9H, NCH3), 4.44 (s, 2H, CH2), 6.90–7.40 (m, 12H, CH);
31P{1H} NMR (CDCl3): d¼127.7.
125.4, 129.6, 130.4, 133.7, 135.8, 140.2, 150.9, 151.5 (d, JCP
4.2 Hz); 31P{1H} NMR (CDCl3): d¼126.6.
¼
Bis-(m-tolyl)-3-[(N,N-dimethylamino)-methyl]-phenyl
Phosphite (14) by Procedure B: From 3-[(N,N-dimethylami-
no)-methyl]-phenol (0.616 g, 4.08 mmol), triethylamine
(1.7 mL, 12.2 mmol), toluene (10 mL), bis-(m-tolyl) chloro-
phosphite (1.203 g, 4.28 mmol) solution in toluene (5 mL); col-
orless oil; yield: 1.526 g(94%). 1H NMR (CDCl3): d¼2.28 (s,
6H, NCH3), 2.37 (s, 6H, CCH3), 3.45 (s, 2H, CH2), 6.95–7.35
(m, 12H, CH); 13C{1H} NMR (CDCl3): d¼21.4, 45.3, 63.9,
117.7 (d, JCP ¼6.1 Hz), 119.4 (d, JCP ¼7.1 Hz), 121.4 (d, JCP
¼
7.1 Hz), 121.5 (d, JCP ¼7.1 Hz), 124.8, 125.0, 129.4, 129.5,
139.9, 140.8, 151.6 (d, JCP ¼3.3 Hz), 151.7 (d, JCP ¼3.3 Hz);
31P{1H} NMR (CDCl3): d¼129.0.
Bis-(m-tolyl)-4-[(N,N-dimethylamino)-methyl]-phenyl
Phosphite (15) by Procedure B: From 4-[(N,N-dimethylami-
no)-methyl]-phenol (0.438 g, 2.90 mmol), triethylamine
(1.2 mL, 8.70 mmol), toluene (10 mL), bis-(m-tolyl) chloro-
phosphite (0.856 g, 3.05 mmol) solution in toluene (5 mL); col-
orless oil; yield: 0.977 g(85%); 1H NMR (CDCl3): d¼2.27 (s,
6H, NCH3), 2.36 (s, 6H, CCH3), 3.43 (s, 2H, CH2), 6.95–7.35
(m, 12H, CH); 13C{1H} NMR (CDCl3): d¼21.4 (CCH3), 45.2
(NCH3), 63.6 (CH2), 117.6 (d, JCP ¼7.2 Hz, CH), 120.6 (d,
JCP ¼6.6 Hz, CH), 121.4 (d, JCP ¼7.1 Hz, CH), 125.0 (CH),
129.4 (CH), 130.4 (CH), 134.6, 139.8, 150.7 (d, JCP ¼2.7 Hz),
151.6 (d, JCP ¼3.9 Hz); 31P{1H} NMR (CDCl3): d¼128.7.
2,2’-Biphenyl-4-(3-methylimidazol-1-yl)-phenyl Phosphite
Tetrafluoroborate (16) by Procedure C: From 2,2’-biphenyl-
4-(imidazol-1-yl)-phenyl phosphite (10; 398 mg, 1.06 mmol),
trimethyloxonium tetrafluoroborate (157 mg, 1.06 mmol), di-
chloromethane (5 mL); white paste; yield: 499 mg(99%). 1H
NMR (CDCl3): d¼3.99 (s, 3H, CH3), 7.15–7.57 (m, 15H);
13C{1H} NMR (CDCl3): d¼36.9 (CH3), 121.4, 122.2, 122.4 (d,
Bis-(m-tolyl)-4-[(N,N,N-trimethylammonio)-methyl]-phe-
nyl Phosphite Tetrafluoroborate (21) by Procedure C: From
bis-(m-tolyl)-4-[(N,N-dimethylamino)-methyl]-phenyl phos-
phite (15; 756 mg, 1.93 mmol), dichloromethane (5 mL), a sus-
pension of trimethyloxonium tetrafluoroborate (272 mg,
1.84 mmol) in dichloromethane (5 mL); colorless oil; yield:
690 mg(75%). 1H NMR (CDCl3): d¼2.32 (s, 6H, CCH3),
3.06 (s, 9H, NCH3), 4.44 (s, 2H, CH2), 6.90–7.50 (m, 12H,
CH); 31P{1H} NMR (CDCl3): d¼127.5.
General Procedure D: 2,2’-Biphenyl-3-(N,N,N-
trimethylammonio)-phenyl phosphite
Bis(trifluoromethylsufonyl)amide (22)
A solution of dimethyl sulfate (220 mg, 1.74 mmol) and lithium
bis(trifluoromethylsulfonyl)amide (500 mg, 1.74 mmol) in
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1845

´
Christophe Vallee et al.
FULL PAPERS
acetonitrile (5 mL) was slowly added to a solution of 2,2’-bi-
N-(4-Hydroxybenzyl)-2,3-dimethylimidazolium
Tetrafluoroborate (27)
´
phenyl-3-(N,N-dimethylamino)-phenyl phosphite (3; 611 mg,
1.74 mmol) in acetonitrile (5 mL). The mixture was stirred
for 72 h at room temperature and filtered. The solvent was re-
moved under reduced pressure and a purple oil was obtained;
yield: 1.11 g(99%). 1H NMR (CDCl3): d¼3.47 (s, 9H, CH3),
7.15–7.55 (m, 12H, CH). 13C{1H} NMR (CDCl3): d¼57.4
(CH3), 112.4 (d, JCP ¼5.5 Hz), 115.3, 119.9 (q, JCF ¼320 Hz),
122.0, 122.7 (d, JCP ¼8.7 Hz), 126.0, 129.6, 130.1, 130.6 (d,
JCP ¼3.3 Hz), 132.2, 147.2, 148.6 (d, JCP ¼5.2 Hz), 153.1 (d,
JCP ¼3.3 Hz); 31P{1H} NMR (CDCl3): d¼138.2.
N-(4-Hydroxybenzyl)-2-methylimidazole
(25;
2.50 g,
13.3 mmol) and [Na][BF4] (1.46 g, 13.3 mmol) were dissolved
in acetonitrile (12 mL). Dimethyl sulfate (1.68 g, 13.3 mmol)
was rapidly added and the solution stirred at room temperature
for 24 h. It was filtered, concentrated and washed repeatedly
with diethyl ether (8ꢁ20 mL). It was purified through a short
alumina column to afford N-(4-hydroxybenzyl)-2,3-dimethyli-
midazolium tetrafluoroborate as a yellow oil; yield: 2.82 g
1
(73%). H NMR (CD3CN): d¼2.53 (s, 3H, CCH3), 3.71 (s,
Tris-[3-(N,N,N-dimethylamino)-phenyl] Phosphite (23) by
Procedure A: From 3-(N,N-dimethylamino)-phenol (6.57 g,
48.00 mmol), phosphorus trichloride (2.20 g, 16.00 mmol),
triethylamine (6.46 g, 64 mmol), toluene (20 mL); viscous
dark oil; yield: 2.96 g(43%). 1H NMR (CDCl3): d¼2.90 (s,
18H, NCH3), 6.45–6.60 (m, 9H), 7.15 (m, 3H); 13C{1H} NMR
(CDCl3): d¼40.6, 105.1 (d, JCP ¼6.9 Hz), 108.4, 108.6 (d,
JCP ¼7.9 Hz), 129.9, 152.0, 152.9 (d, JCP ¼4.6 Hz); 31P{1H}
NMR (CDCl3): d¼129.7.
3H, NCH3), 5.16 (s, 2H, CH2), 6.85–6.90 (m, 2H, CH), 7.15–
7.30 (m, 4H, CH); 13C{1H} NMR (CD3CN): d¼9.4 (CCH3),
34.8 (NCH3), 51.1 (CH2), 115.8 (CH), 120.8, 122.4 (CH Im-
4,5), 124.7 (CCH2), 129.8 (CH), 144.6 (CH Im-2), 157.6
(COH); anal. calcd. for C12H15BF4N2O: C 49.69, H 5.21, N
9.66; found: C 49.40, H 7.11, N 9.71.
Tris-4-[(2,3-dimethylimidazol-1yl)methyl]phenyl
Phos-
phite Tris[bis(trifluoromethylsulfonyl)amide] (28) by Proce-
dure A: From N-(4-hydroxybenzyl)-2,3-dimethylimidazolium
bis(trifluoromethylsulfonyl)amide (6.521 g, 13.50 mmol), tri-
ethylamine (1.82 g, 18 mmol), PCl3 (618 mg, 4.50 mmol), ace-
tonitrile; yellow oil; yield: 4.37 g(66%). 1H NMR (CD3CN):
d¼2.54 (s, 9H, CCH3), 3.75 (s, 9H, NCH3), 5.27 (s, 6H, CH2),
7.15–7.35 (m, 18H, CH); 13C{1H} NMR (CD3CN): d¼9.4
(CCH3), 34.9 (NCH3), 50.8 (CH2), 120.5 (q, JCF ¼321 Hz,
CF3), 121.0, 121.2 (d, J¼7.2 Hz, o-CH), 122.7 (CH), 129.9
(m-CH), 130.2 (CCH2), 144.9 (CH Im-2), 151.5 (d, J¼2.7 Hz,
CO); 31P{1H} NMR (CD3CN): d¼128.00; anal. calcd. for
C42H42F18N9O15PS6: C 34.13, H 2.86, N 8.53, P 2.10; found:
C 34.52, H 2.74, N 8.19, P 2.43; MS (ESI): m/z¼1197
Tris-[3-(N,N,N-trimethylammonio)-phenyl]
Phosphite
Tris[bis(trifluoromethylsufonyl)amide] (24) by Procedure D:
From dimethyl sulfate (231 mg, 1.83 mmol) ,tris[3-(N,N-dime-
thylamino)-phenyl] phosphite (23; 268 mg, 0.61 mmol), aceto-
nitrile (10 mL) [Li][TF2N] (525 mg, 1.83 mmol) in acetonitrile
(10 mL); brown oil; yield: 791 mg(98%). 1H NMR (CD3CN):
d¼3.56 (s, 27H, NCH3), 7.40–7.70 (m, 12H, CH); 13C{1H}
NMR (CDCl3): d¼57.0, 113.4 (d, JCP ¼5.5 Hz), 116.5, 119.9
(q, JCF ¼320 Hz), 122.5 (d, JCP ¼8.2 Hz), 131.9, 147.9, 151.9
(d, JCP ¼3.3 Hz); 31P{1H} NMR (CDCl3): d¼128.7.
(C40H42F12N8O11PS4þ),
(C36H42N6O3P3þ).
458
(C38H42F6N7O7PS22þ),
212
Tris-4-[(2,3-dimethylimidazol-1-yl)methyl]phenyl
phite Tetrafluoroborate (29) by Procedure A: From N-(4-hy-
droxybenzyl)-2,3-dimethylimidazolium tetrafluoroborate
(1.749 g, 6.03 mmol), triethylamine (0.812 g, 8.04 mmol), PCl3
(276 mg, 2.01 mmol); white powder; yield: 1.46 g (80%); H
NMR (CD3CN): d¼2.54 (s, 9H, CCH3), 3.74 (s, 9H, NCH3),
5.28 (s, 6H, CH2), 7.15–7.35 (m, 18H, CH); 13C{1H} NMR
(CD3CN): d¼9.4 (CCH3), 34.9 (NCH3), 50.7 (CH2), 121.0
(CH), 121.2 (d, J¼6.6 Hz, o-CH), 122.6 (CH), 129.9 (m-CH),
130.3 (CCH2), 144.9 (CH Im-2), 151.5 (d, J¼2.7 Hz, CO);
31P{1H} NMR (CD3CN): d¼128.15; MS (ESI): m/z¼362
(C36H42BF4N6O3P2þ), 212 (C36H42N6O3P3þ).
Phos-
N-(4-Hydroxybenzyl)-2-methylimidazole (25)
4-Hydroxybenzyl alcohol (5.0 g, 40.3 mmol) and 2-methylimi-
dazole (3.31 g, 40.3 mmol) were heated at 1608C for 1 h. The
resultingsolid was washed with dichloromethane to afford N-
(4-hydroxybenzyl)-2-methylimidazole as a yellow powder;
yield: 6.9 g(92%). 1H NMR (DMSO-d6): d¼2.22 (s, 3H,
CH3), 4.97 (s, 2H, CH2), 6.70–7.10 (m, 6H, CH); 13C{1H}
NMR (DMSO-d6): d¼13.2 (CH3), 48.7 (CH2), 115.8 (CH),
120.5 (CH Im-4), 126.6 (CH Im-5), 128.0 (CCH2), 129.0
(CH), 144.1 (CH Im-2), 157.3 (CO); anal. calcd. for
C11H12N2O: C 70.19, H 6.43, N 14.88; found: C 70.37, H 6.92,
N 14.11.
1
N-(4-Hydroxybenzyl)-2,3-dimethylimidazolium
Bis(tri-
fluoromethylsufonyl)amide (26) by Procedure C: From N-(4-
hydroxybenzyl)-2-methylimidazole (25; 3.514 g, 18.7 mmol)
and [Li][TF2N] (5.366 g, 18.7 mmol) dissolved in acetonitrile
(15 mL). Dimethyl sulfate (2.35 g, 18.7 mmol) was added rap-
idly and the solution stirred at room temperature for 24 h. It
was filtered, concentrated and washed repeatedly with diethyl
ether (8ꢁ20 mL). Purification through a short alumina col-
umn afforded 22 as a yellow oil; yield: 7.72 g(85%). 1H NMR
(CD3CN): d¼2.53 (s, 3H, CCH3), 3.71 (s, 3H, NCH3), 5.15 (s,
2H, CH2), 6.85–6.90 (m, 2H, CH), 7.15–7.25 (m, 4H, CH);
13C{1H} NMR (CD3CN): d¼9.4 (CCH3), 34.8 (NCH3), 51.1
(CH2), 115.8 (CH), 120.5 (q, JCF ¼321 Hz, CF3), 120.7, 122.4
(CH Im-4,5), 124.7 (CCH2), 129.8 (CH), 144.6 (CH Im-2),
157.5 (CO); anal. calcd. for C14H15F6N3O5S2: C 34.78, H 3.13,
N 8.69; found: C 34.80, H 3.51, N 8.38.
Catalysis, Hydrocyanation Experiments and Analysis
Ni(cod)2 was purchased from Strem Chemicals, stored under
argon and used without further purification. Triphenyl phos-
phite was obtained from Aldrich and tri-m-tolyl phosphite
was furnished by Rhodia. They were stored under argon and
used as is. Anhydrous zinc chloride (99.999%, H2O<
100 ppm) was purchased from Aldrich.
Catalysis Runs: In a typical reaction, 40 mgof Ni(cod) , 5
2
equivalents of ligand, 400 mg of 3PN (30 equivalents), 20 mg
of zinc chloride (1 equivalent) and 2.0 gof ionic liquid were
weighed in a Schott tube under argon. The tube was then heat-
ed up for 3 h at 708C with slow continuous addition of acetone
cyanohydrin (24 equivalents).
1846
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1835 – 1847

Catalytic Hydrocyanation Reaction of 3-Pentenenitrile
FULL PAPERS
Work-up: At the end of the run, tubes were cooled to room
temperature and acetone (20 mL) was added. The solutions
were stirred for 30 minutes, and the tubes were opened (be-
ware: hydrogen cyanide concentration level has to be moni-
tored). The reaction mixture was diluted with 10 mL of THF
and butylbenzene was added as an internal standard. The sol-
ution was filtered through a Millipore filter and analyzed by
gas chromatography.
[11] W. Tam, T. Foo, J. M. Garner, (E. I. Du Pont de Nemours
and Company), World Patent WO 9952632, 1999.
[12] C. J. I. Van der Vlugt, A. C. Hewat, S. Neto, R. Sablong,
A. M. Mills, M. Lutz, A. L. Spek, C. Mueller, D. Vogt,
Adv. Synth. Catal. 2004, 346, 993–1003.
[13] V. Lecocq, C. C. Santini, F. Lefebvre, Y. Chauvin, J.-M.
Basset, J.-C. Galland, Book Abstract of 1st COIL, Salz-
burg, 2005.
Analyses: The products were quantitatively analyzed by gas
chromatography on an HP 6890 chromatograph equipped with
TCD detector and a Stabilwaxꢁ-DA [(poly(ethylene glycol)]
column (L¼30 m, fint ¼0.53 mm, film thickness¼1 mm). In-
jector and detector temperatures were 2508C. The tempera-
ture program was from 508C (3 min) to 2008C at a heating
rate of 48C/min and then to 2208C (25 min) at a heatingrate
of 28C/min. The relative response factors of the reaction prod-
ucts were determined from pure samples.
[14] W. Keim, D. Vogt, H. Waffenschmidt, P. Wasserscheid, J.
Catal. 1999, 186, 481–484.
[15] M. F. Sellin, P. B. Webb, D. J. Cole-Hamilton, Chem.
Commun. 2001, 781–782.
[16] F. Favre, H. Olivier-Bourbigou, D. Commereuc, L. Saus-
sine, Chem. Commun. 2001, 1360–1361.
´
´
[17] a) C. Vallee, C. Valerio , Y. Chauvin, G. P. Niccolai, J.-M.
Basset, C. C. Santini, J.-C. Galland, B. Didillon, J. Mol.
´
´
Catal. A 2004, 214, 71–81; b) C. Vallee, C. Valerio , Y.
Chauvin, G. P. Niccolai, J.-M. Basset, J.-C. Galland,
(Rhodia Polyamide Inter.), World Patent WO 03024919,
2003.
Statistical analysis of the DOE was performed with Minitab
V13.0.[28]
[18] E. Kuntz, Chem. Tech. 1987, 17, 570–575; F. Joo, in:
Aqueous organometallic catalysis, (Eds.: B. R. James,
P. W. N. M. van Leeuwen), Kluwer Academic Publishers,
Dordrecht, 2001.
[19] A. G. Abatjoglou, D. R. Bryant, (Union Carbide Chem-
icals and Plastics Co.), European Patent EP 353770, 1990.
[20] H. Bahrmann, B. Fell, G. Papadogianakis, (Hoechst
A. G.), German Patent DE 3942954, 1991.
[21] B. Fell, G. Papadogianakis, W. Konkol, J. Weber, H.
Bahrmann, J. Prakt. Chem. 1993, 335, 75–82.
[22] T. Fujii, W. Ueno, K. Nakamura, (Mitsubishi Gas Chem-
ical Co.), Japanese Patent JP 2000281689, 2000.
[23] H. Bahrmann, M. Haubs, T. Muller, N. Schopper, B. Cor-
nils, J. Organomet. Chem. 1997, 545–546, 139–149.
[24] Y.-Y. Yan, T. V. RajanBabu, J. Org. Chem. 2001, 66,
3277–3283.
[25] H. M. Fitch, British Patent GB 681102, 1952.
[26] S. J. A. Grove, J. Kaur, A. W. Muir, E. Pow, G. J. Tarver,
M.-Q. Zhang,. Bioorg. Med. Chem. Let. 2002, 12, 193–
196.
References
† French Patent Fr.04.08440, 2004.
[1] K. Weissermel, Industrial Organic Chemistry, 3rd edn,
Wiley-VCH, Weinheim, 1997.
[2] C. A. Tolman, R. J. McKinney, W .C. Seidel, J .D. Drulin-
er, W . R. Stevens, Adv. Catal. 1985, 33, 1–46.
[3] H. S. M. Lu, N. S. Radu, (E. I. Du Pont de Nemours and
Company), World Patent WO 0121579, 2001.
[4] J. M. Garner, K. A. Kruetzer, W. Tam, (E. I. Du Pont de
Nemours and Company), World Patent WO 9906358,
1999.
[5] K. A. Kreutzer, W. Tam, J. M. Garner, J. R. Boyles, (E. I.
Du Pont de Nemours and Company), World Patent WO
0121580, 2001.
[6] R. Baumann, M. Bartsch, D. P. Kunsmann-Keitel, G. Ha-
derlein, D. Siegel, D. Vogt, A. Hewat, (BASF A. G.),
German Patent DE 10046025, 2002.
[7] M. Bartsch, D. P. Kunsmann-Keitel, R. Baumann, G. Ha-
derlein, W. Siegel, (BASF A. G.), World Patent WO
0213964, 2002.
[8] A. I. Breikss, T. Foo, (E. I. Du Pont de Nemours and
Company), World Patent WO 9723446, 1997.
[9] J. Fischer, W. Siegel, (BASF A. G.), German Patent DE
19740180, 1999.
[10] M. Bartsch, R. Baumann, D. P. Kunsmann-Keitel, G. Ha-
derlein, T. Jungkamp, M. Altmayer, W. Siegel, (BASF
A. G.), German Patent DE 10136488, 2003.
[27] M. Wakselman, J. C. Robert, G. Decots, M. Vilkas, M.
Bull. Soc. Chim. Fr. 1973, 1179–1183.
[28] S. Godbert, in: Design and Analysis on Chemical Re-
search, (Ed.: R. L. Tranter), Sheffield Academic Press
Ltd, Sheffield, 1999, pp. 188–236.
Adv. Synth. Catal. 2005, 347, 1835 – 1847
ꢀ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1847