Cyanophosphine derivatives: Nitrile or cyanide functionality?

Article abstract of DOI:10.1021/om000704h

Treatment of dicyanophosphines R-P(CN)₂ with 1 equiv of Schwartz's reagent, [Cp₂Zr(H)Cl]_n, led: (i) via a substitution reaction to the formation of the stable hydrocyanophosphines R-P(H)CN (R = Me, ^tBu, Mes*) or (ii) through a hydrozirconation reaction to the synthesis of the aldiminocyanophosphine R-P(CH double bond NZrCp₂Cl)CN (R = NⁱPr₂). Further addition of 1 equiv of [Cp₂Zr(H)Cl]_n to R-P(H)CN (R = Me, ^tBu, Mes*) gave the primary phosphine R-PH₂ (R = Me) or the aldimino secondary phosphine R-P(H)(CH double bond NZrCp₂Cl) (R = ^tBu, Mes*) after the hydrozirconation reaction. Interestingly, the dialdiminophosphorus compound R-P(CH double bond NZrCp₂Cl)₂ (R = NⁱPr₂) was obtained after addition of [Cp₂Zr(H)Cl]_n to R-P(CH double bond NZrCp₂Cl)CN (R = NⁱPr₂). The difference in reactivity observed for the dicyanophosphines R-P(CN)₂ with the metal hydride [Cp₂Zr(H)Cl]_n have been studied with the help of experimental analysis (UV-photoelectron spectroscopy, NMR, IR, X-ray) and computational methods.

Full text of DOI:10.1021/om000704h

Organometallics 2001, 20, 25-34
25
Cya n op h osp h in e Der iva tives: Nitr ile or Cya n id e
F u n ction a lity?
Alexandrine Maraval,† Alain Igau,*,†,‡ Christine Lepetit,† Anna Chrostowska,§
J ean-Marc Sotiropoulos,§ Genevie`ve Pfister-Guillouzo,*,§ Bruno Donnadieu,† and
J ean Pierre Majoral*,†
Laboratoire de Chimie de Coordination, CNRS, 205 route de Narbonne,
31077 Toulouse Cedex, France, and Laboratoire de Physico-Chimie Mole´culaire, UMR 5624,
Universite´ de Pau & des Pays de l’Adour, Avenue de l’universite´, 64000 Pau, France
Received August 14, 2000
Treatment of dicyanophosphines R-P(CN)2 with 1 equiv of Schwartz’s reagent, [Cp2Zr-
(H)Cl]n, led: (i) via a substitution reaction to the formation of the stable hydrocyanophos-
t
phines R-P(H)CN (R ) Me, Bu, Mes*) or (ii) through a hydrozirconation reaction to the
synthesis of the aldiminocyanophosphine R-P(CHdNZrCp2Cl)CN (R ) NiPr2). Further
t
addition of 1 equiv of [Cp2Zr(H)Cl]n to R-P(H)CN (R ) Me, Bu, Mes*) gave the primary
phosphine R-PH2 (R ) Me) or the aldimino secondary phosphine R-P(H)(CHdNZrCp2Cl)
(R ) tBu, Mes*) after the hydrozirconation reaction. Interestingly, the dialdiminophosphorus
compound R-P(CHdNZrCp2Cl)2 (R ) NiPr2) was obtained after addition of [Cp2Zr(H)Cl]n
to R-P(CHdNZrCp2Cl)CN (R ) NiPr2). The difference in reactivity observed for the
dicyanophosphines R-P(CN)2 with the metal hydride [Cp2Zr(H)Cl]n have been studied with
the help of experimental analysis (UV-photoelectron spectroscopy, NMR, IR, X-ray) and
computational methods.
In tr od u ction
The chemistry of the cyano group has been exten-
sively studied and used in organic, inorganic, and
organometallic synthesis.1 In the course of our studies,
directed to the use of the interaction between main-
group elements and group 4 elements for the develop-
ment of new synthetic approaches for the synthesis of
organic or inorganic species, we have been interested
in the chemical behavior of cyanophosphines toward
metal hydride reagent (X ) Cl)5 or the hydrozirconation
reaction may be observed on the cyano group when X
) CN.6 It has long been established that changing
substituents on phosphine derivatives can dramatically
induce changes in the behavior of the free ligands and
of their transition-metal complexes. This can be ratio-
nalized in terms of electronic and steric effects.7 In a
first approach, we will try to analyze with the help of
experimental (photoelectronic spectra, molecular struc-
ture, IR and NMR spectroscopy) and computational
evidence how these effects on dicyanophosphines,
R-P(CN)2, are responsible for the reaction observed and
how one effect can dominate the other.
zirconocene complexes.2 Very few studies have appeared
on the coordination chemistry of dicyanophosphines,
R-P(CN)2.3 We recently focused our attention on the
study of the reactivity of dicyanophosphines R-P(CN)2
with Schwartz’s reagent [Cp2Zr(H)Cl]n (1). Dicyano-
phosphines, R-P(CN)2, may act on group 4 d0 com-
plexes,4 Cp2MXY (M ) Ti, Zr), as triple-sided L type
ligands which can coordinate to metal through both the
P (soft) and N (hard) donor atoms (A). Moreover, our
studies on the reactivity of mono(pseudo)halogenated
derivatives R2PX have shown that 1 could react as a
† Laboratoire de Chimie de Coordination, CNRS.
‡ Fax: 05 61 55 30 03. E-mail: igau@lcc-toulouse.fr.
§ Universite´ de Pau & des Pays de l’Adour.
Resu lts
(1) The Chemistry of the Cyano Group; Patai, P., Series Ed.,
Rappoport, Z., Vol. Ed.; Wiley: London, 1970.
(2) Cadierno, V.; Donnadieu, B.; Igau, A.; Majoral, J . P. Eur. J . Inorg.
Chem. 1999, 417. Igau, A.; Dufour, N.; Straw, T.; Dewey, M.; Bouton-
net, F.; Mahieu, A.; Zablocka, M.; Majoral J . P. Phosphorus, Sulfur
Silicon Relat. Elem. 1993, 77, 283.
(3) (a) Nazmutdinova, V. N.; Chirkova, L. P.; Mukhamadeeva, R.
M.; Plyamovatyi, A. Kh.; Gainullin, R. M.; Pudovik, M. A.; Polovnyak,
V. K.; Pudovik, A. N. Zh. Obshch. Khim. 1994, 64, 1617. (b) Wilkie, C.
A.; Parry, R. W. Inorg. Chem. 1980, 19, 1499. (c) Hoefler, M.; Schnitzler,
M. Chem. Ber. 1974, 107, 194. (d) Sabherwal, I. H.; Burg, A. B. Inorg.
Chem. 1972, 11, 3138. (e) Coskran, K. J .; J ones, C. E. Inorg. Chem.
1971, 10, 1664.
Addition at -78 °C of 1 equiv of Schwartz’s reagent
[Cp2Zr(H)Cl]n (1) to dichlorophosphine derivatives
R-PCl2 (2a -e) led to the formation of the corresponding
primary phosphorus compounds R-PH2 (3a -e) along
with some remaining starting compound 2a -e (Scheme
1). As expected, [Cp2Zr(H)Cl]n (1) acts as a hydride
(5) Zablocka, M.; Delest, B.; Igau, A.; Skowronska, A.; Majoral, J .-
P. Tetrahedron Lett. 1997, 38, 5997.
(4) Cardin, D. J .; Lappert, M. F.; Raston, C. L. Chemistry of Organo-
Zirconium and Hafnium Compounds; Wiley: New York, 1986; and
references therein.
(6) Boutonnet, F.; Dufour, N.; Straw, T.; Igau, A.; Majoral, J .-P.
Organometallics 1991, 10, 3939.
(7) Tolman, C. A. Chem. Rev. 1977, 77, 313.
10.1021/om000704h CCC: $20.00 © 2001 American Chemical Society
Publication on Web 12/09/2000

26 Organometallics, Vol. 20, No. 1, 2001
Maraval et al.
NMR spectrum at δCH 9.34 (2J HP ) 75.4 Hz) ppm and
in the 13C NMR spectrum with the imino carbon at δCd
Sch em e 1. Rea ctivity of R-P Cl2 (2a -e) w ith
Sch w a r tz’s Rea gen t [Cp 2Zr (H)Cl]n (1)
171.8 (1J CP ) 17.0 Hz) ppm. The signal at 120.8 (1J CP
N
) 101.2 Hz) ppm can be attributed, as expected, to the
nitrile carbon atom in cyano phosphine compounds.
Addition of 1 equiv of the Schwartz’s reagent 1 to the
secondary cyanophosphine 5a led to a substitution
reaction to give the corresponding primary phosphine
MePH2 (δ(31P) -162 (1J HP ) 193 Hz) ppm) along with
the loss of the metal fragment [Cp2Zr(CN)Cl] (Scheme
3). In marked contrast the secondary cyanophosphines
5b,c reacted as nitrile compounds to give, after 1,2-
addition of the zirconium-hydride bond to the CtN
group, the hydrozirconation products 7b,c respectively,
Sch em e 2. Rea ctivity of R-P (CN)2 (4a -e) w ith
Sch w a r tz’s Rea gen t [Cp 2Zr (H)Cl]n (1)
1
as one isomer (Scheme 3). The J HP coupling constants
for 7b,c of 213 and 234 Hz, respectively, are typical for
a P-H bond in phosphine compounds. The chemical
1
shift observed in H NMR at 9.70 and 9.20 ppm for 7b,c
are characteristic for the aldimido CHdN protons. Cp
groups of the metal fragment in the region of 6.00-6.20
ppm are inequivalent. At room temperature a 1,3-H
sigmatropic rearrangement is observed for 7c, giving
the corresponding N-zirconated phosphaalkene product
8. A COSY 1H-1H 31P-decoupled experiment allows us
to attribute a broad signal at 7.98 ppm to the NH group
and at 8.01 ppm to the dCH fragment for 8. Interest-
ingly, the same products, 7b and 8, were obtained in a
one-pot procedure starting from 4b,c and 2 equiv of 1.
Note that the dihydrozirconated compound 9 is quan-
titatively prepared after addition of 1 equiv of 1 to 6.8
1
Only one isomer is identified. The H NMR spectrum
exhibited, besides the signals corresponding to the
protons of the diisopropylamino groups, a doublet at
10.08 (2J HP ) 61.8 Hz) ppm, while the 13C NMR
spectrum displayed a doublet at 181.5 (1J CP ) 30.1 Hz)
ppm characteristic of the aldimido groups HCdN.
source5 but no secondary halogenated phosphine R-P(H)-
Cl was observed in the reaction mixture. Note that
addition of 2 equiv of the metallocene complex 1 to 2
gave the phosphine 3 as the sole phosphorus product of
the reaction.
Discu ssion
Phosphines of the type R-P(H)X with X ) haloge-
nated or pseudohalogenated groups are rare. These
derivatives are normally unstable with respect to the
loss of HX and consequent oligomerization to cyclophos-
phines, (RP)n.9,10 The stabilization of the phosphine unit
R-P(H)X (X ) Cl, Br, I) has been only accomplished
with the use ofvery bulky groups (R )Mes*, N(SiMe3)2)11
or with the strongly electronegative group (R ) CF3).12
Note that no phosphine of the type R-P(H)CN has yet
been isolated; only CF3-P(H)CN was observed by 31P
NMR, but this decomposes by loss of HCN.12 The
hydridocyanophosphines R-P(H)CN (5a-c) are remark-
ably thermally stable toward R-elimination at the
central phosphorus atom. Then, the study of the forma-
tion of these derivatives would be of interest with
In marked contrast with R-PCl2 (2a -c) the corre-
sponding dicyanophosphine compounds R-P(CN)2 (4a -
c) were reacted with 1 equiv of Schwartz’s reagent 1 in
toluene to give, after the loss of the metallic fragment
[Cp2Zr(CN)Cl], the secondary cyanophosphine R-P(H)-
CN (5a -c) in quantitative yield (Scheme 2). Compounds
5a -c were unambiguously characterized by the usual
1
spectroscopic methods. J PH coupling constants in the
range of 210-250 Hz are characteristic for the P-H
bond in phosphine compounds. Infrared spectroscopy
(νCtN ) 1581-1583 cm-1) and 13C NMR (δCtN ) 119-
1
122 ppm (d, J CP ) 41-75 Hz)) corroborated the
presence of the cyano PCN group in 5a -c. Note that
the reaction of 4d with 1 does not give the expected
primary phosphine PhP(H)CN. The behavior of 4d is
similar to that observed for 2d with 1; PhPH2 was
identified among other phosphorus derivatives.
(8) O’Neal, R. H.; Neilson, R. H. Inorg. Chem. 1984, 23, 1372 and
references therein.
(9) Phosphines of the type R-P(H)X with X ) F, Cl, Br, I stabilized
as complex ligands are already known: Marinetti, A.; Mathey, F.
Organometallics 1982, 1, 1488 and references therein.
(10) Cowley, A. H.; Kilduff, J . E.; Norman, N. C.; Pakulski, M. J .
Chem. Soc., Dalton Trans. 1986, 1801. Cowley, A. H.; Kilduff, J . E.;
Lasch, J . G.; Norman, N. C.; Pakulski, M.; Ando, F.; Wright, T. C. J .
Am. Chem. Soc. 1983, 105, 7751. Escudie´, J .; Couret, C.; Ranaivonja-
tovo, H. J .; Satge´, J .; J aud, J . Phosphorus, Sulfur Relat. Elem. 1983,
17, 221.
(11) Dobbie, R. C.; Gosling, P. D.; Straughan, B. P. J . Chem. Soc.,
Dalton Trans. 1975, 2368.
(12) Maraval, A.; Donnadieu, B.; Igau, A.; Majoral, J . P. Organo-
metallics 1999, 18, 3183.
Surprisingly, after 2 h in solution in toluene, the 1,2-
addition of the Zr-H bond of 1 occurs on one of the two
cyano groups in 4e to give quantitatively the hydrozir-
conation reaction product 6 as one isomer (Scheme 2).
The composition and constitution of 6 follow from mass
1
analysis and 31P, H, and 13C NMR spectra. The pres-
ence of the aldiminato moiety -CHdN- is proven by
the significantly deshielded chemical shift in the 1H

Cyanophosphine Derivatives
Organometallics, Vol. 20, No. 1, 2001 27
Sch em e 3. Rea ctivity of 5a -c a n d 6 w ith Sch w a r tz’s Rea gen t [Cp 2Zr (H)Cl]n (1)
respect to the difference in reactivity observed between
4a -c and 4e with 1.
expected four π
orbitals (b2, a2, b1, and a1) could be
CN
involved.19 The presence of the phosphorus atom leads
to an additional interaction with the b1 orbital of the
same symmetry. The calculated molecular orbital lo-
calization for 4a (Figure 1) shows these interactions
well. Considering this scheme, we can easily assign the
photoelectron spectrum of 4a (Figure 3a). The spectra
of the phosphine 4a gives rise to a first band at 10.8
eV, followed by one intense band at 12.7 eV and one
larger band, centered at 13.5 eV. The first band can be
assigned to the ejection of an electron of the phosphorus
lone pair orbital; this attribution is confirmed by the
decrease of the intensity using He II radiation. Note that
there is an obvious stabilization to more than 1 eV
relative to Me-PH2 derivative. The integrated areas of
the second band at 12.7 eV suggest that it may in fact
be due to overlapping of the expected b2, a1, and a2
ionizations. The intensity of the third band clearly
increases compared to the two others, on going from He
I to He II radiation, which corresponds well to the
ionization of the nN orbitals (b2 and a1). This broad band
at 13.5 eV could also reasonably contain the ionization
of the b1 orbital.
Even if no experimental geometries are available for
Schwartz’s reagent 1, theoretical studies and chemical
evidence from the hydrozirconation reaction showed
that a precoordination of the unsaturated molecule on
the vacant site of the zirconocene(IV) metal fragment
occurs in the transition states of the reaction process.13
Phosphines and nitriles are good L type ligands in group
4 metallocene(IV) chemistry.14 These two moieties are
present in the structure of the cyanophosphines deriva-
tives 4a -e. Thus, three chelating centers are available
for coordination on the vacant site of the zirconocene.
The necessity of achieving good orbital overlap with the
1a1 orbital of zirconium implies coordination in the
HZrCl plane; the angle between these three ligands
cannot exceed 140°.4 In a first approach, (i) a study of
the electronic effect of the substituent R on the poten-
tially chelating sites in 4a -e using UV-photoelectron,
NMR, and IR spectroscopy and (ii) an estimation of
steric hindrance may help to understand the effect
governing the chemical behavior of the dicyano com-
pounds 4a -e toward 1.
The He I and He II photoelectron spectra of the series
R-P(CN)2 (4a -e) have been recorded and interpreted.
The PE spectra assignment is based on the qualitative
estimation of interactions between the cyano group,
phosphorus lone pair, and substituent and have been
compared with theoretical studies by the single-refer-
ence-based MP2 method,15 in conjunction with the
6-311G(d,p) basis set, and the density functional theory
(DFT)16 with B3LYP17 functional (Gaussian 98)18 real-
ized for 4a ,e and Me2N-P(CN)2. Qualitatively, we can
consider the “through space ” interaction in which the
t
Replacement of a Me group by a Bu group causes a
decrease in IP. The spectra of 4b shows a first band at
(16) Parr, R. G.; Yang, W. Functional Theory of Atoms and Mol-
ecules; Oxford University Press: New York, 1989.
(17) (a) Becke, A. D. Phys. Rev. 1988, 38, 3098. (b) Becke, A. D. J .
Chem. Phys. 1993, 98, 5648. (c) Lee, C.; Yang, W.; Parr, R. G. Phys.
Rev. 1988, B37, 785.
(18) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratman, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .; Barone, V.;
Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma,
K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J . B.;
Cioslowski, J .; Ortiz, J . V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B.; Chen, W.;
Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.; Pople,
J . A. Gaussian 98, revision A.7; Gaussian, Inc.: Pittsburgh, PA, 1998.
(19) Stafast, H.; Bock, H. Z. Naturforsch. 1973, 28B, 746-753.
(13) Endo, J .; Koga, N.; Morokuma, K. Organometallics 1993, 12,
2777.
(14) As an example see: Walsh, P. J .; Hollander, F. J .; Bergman,
R. G. J . Am. Chem. Soc. 1990, 112, 894.
(15) Hehre, W. J .; Radom, L.; Schleyer, P. v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.

28 Organometallics, Vol. 20, No. 1, 2001
Maraval et al.
F igu r e 1. Correlation diagram resulting from “through space” and “through bond” interactions between the CN groups
and the phosphorus and/or amino nitrogen lone pair orbitals for 4a and Me2N-P(CN)2: B3LYP/6-311G** eigenvalues (in
eV).
between the lone pairs is no longer visible. The HOMO
is well localized on the nitrogen atom, whereas the
phosphorus lone pair corresponds to HOMO-1 (Figure
2). The geometrical parameters of the calculated struc-
tures are described in Table 1 and show well that the
nitrile groups are not equivalent in the calculated
structure of 4e.
Experimentally, the PE spectrum of 4e (Figure 3e)
shows two ionizations at 9.3 eV and at 10.35 eV followed
by a broad massif at 12.5 eV and a band at 13.0 eV.
The first bands are respectively assigned to the ejection
of an electron from the amino nitrogen and the phos-
F igu r e 2. HOMO and HOMO-1 B3LYP/6-311G** for 4e.
phorus lone pairs. The first orbital has a large localiza-
tion on the nitrogen lone pair and the second on the
10.4 eV, always assigned to the ionization of the
phosphorus lone pair. The relative intensity variations
phosphorus lone pair, followed by four better resolved
in the He II spectrum demonstrate experimentally the
next bands, described as b2 at 12.0 eV, a1 at 12.5, a2 at
presence at the same time of these combinations and
12.8, and one more intense, at 13.3 eV corresponding
the distinct participation. The massif at 12.2 eV con-
to the nN orbital ionizations and that of the b1 orbital.
tains, as expected, the b2, a2, and a1 orbitals mixed with
For the amino compound R2N-P(CN)2, a model
compound of 4e where the isopropyl substituents bonded
to the nitrogen amino atom were replaced by methyl
the same-symmetry CN orbital ionizations. Another
clearly distinguishable band is the one at 13.0 eV; this
intense peak can be assigned to the nitrogen atom lone
pair orbitals from the cyano moieties.
groups, Me2N-P(CN)2, was first considered. At the MP2
level, the calculated energetically privileged structure
In the case of 4c,d , the spectra (Figure 3c,d) show a
exhibits a slightly sp3 nitrogen atom hybridization with
first ionization at 8.8 and 10.0 eV and a second band at
a gauche lone pair orientation compared to the phos-
9.5 and 10.7 eV, respectively. Considering the rotation
phorus one. The geometry of 4e optimized at the B3LYP/
angle of the mesityl ring (vide infra), it seems reasonable
6-311G** level exhibits an sp2 nitrogen atom hybrid-
to take into account a weak interaction between the
ization (sum of the valence angles 359.8°) in agreement
aromatic ring and the phosphorus atom lone pair, but
with the geometry of the NiPr2 moiety generally ob-
a more important one between the aromatic ring and
served in the Cambridge Data Base.20 The orientation
the cyano groups. The broad shape of the first band,
is also slightly gauche (around 5°). The position of the
methyl substituents of the isopropyl group develops a
(20) Hydrio, J .; Gouygou, M.; Dallemer, F.; Daran, J . C.; Balavoine,
difference between the cyano moieties. The interaction
G. G. A. Private communication.

Cyanophosphine Derivatives
Organometallics, Vol. 20, No. 1, 2001 29
precoordination on the CN system followed by the
hydrozirconation reaction. However, due to the low
localization on the πCN orbitals, the latter reaction can
be considered because of the reduced accessibility of the
phosphorus lone pair.
Though this mechanism cannot be controlled by the
charges, we have examined the repartition of the
electronic population in order to estimate the charac-
teristics of the P-substituent bond for 4a -e.
The repartition of the electronic population in the di-
cyanophosphines 4a -e have been examined with the
help of the electron localization function (ELF).21 Ac-
cording to our studies, no delocalization is observed from
the lone pair of the phosphorus atom to the nitrile
moiety for the dicyanophosphines 4a -e. More surpris-
ingly, the topological analysis of ELF provides a new
description of the C-N chemical bond in nitrile com-
pounds. According to the electronic population found for
the C-N bond, the bond order is close to 2, whereas
the electronic population of the nitrogen lone pair is
larger than 3 (Table 5).22 Indeed, the isosurface of the
ELF is isotropic along the C-N axis and the analysis
of the orbital contribution to the C-N bond shows that
the πX and πY systems have the same weight but are
lower than two electrons. The picture given by AIM
(topological analysis of the electron density within the
theory of “atoms in molecules”) is in agreement with the
latter point, as the AIM charge resulting from integra-
tion of the population of the nitrogen core basin is close
to about -1.4.23 Then, as expected with the description
given by ELF, the ellipticity, ꢀ, which is a measure of
the anisotropy of the electron density along the bond
path in the AIM method, is found to be close to zero.24
The comparison of the orbital contribution to the lone
pair of N and P and of the P-C and C-N bonds is very
i
similar for the methyl and Pr2N substituents. However,
the nitrile groups are not completely equivalent in the
i
case of the Pr2N substituent. Finally, in the case of the
methyl substituent the contribution of the P-C’s orbital
to the electronic population of the phosphorus lone pair
i
F igu r e 3. Photoelectron spectra of 4a -e. For clarity, only
He II spectra for the compounds 4a ,e are presented.
is about twice that obtained for the Pr2N substituent,
suggesting that, in the case of an attack of Schwartz’s
reagent 1 at the phosphorus lone pair, the P-C bond
would be strongly weakened in the case of the methyl
substituent.
Photoelectron experiments and theoretical approaches
point out the only interaction b1πCN-nP. The cyano
group ionizations remain well-distinguished, and the
structure better corresponds to the resonance forms I
and II, which is in full agreement with the IR and NMR
analysis.
centered at 8.8 eV, is characteristic of the nondegener-
ated πaryl system ionizations. The following band at 9.5
eV corresponds to the phosphorus atom lone pair orbital
ionization. The same attribution can be proposed in the
case of the phenyl derivative 4d , in which the first band
corresponds to the πphenyl ionizations and the second
band at 10.7 eV corresponds to the ejection of the elec-
tron localized on the phosphorus atom lone pair orbital.
In conclusion, the photoelectron spectra of 4a -e show
that the bands assigned to the phosphorus lone pair
orbital ionization correspond to either the HOMO or the
HOMO-1 and are in the range of 9.7 and 10.8 eV (Figure
4). Then, in comparison to the nitrile nitrogen lone pair,
nN (up to 14 eV), the phosphorus lone pair site, when
accessible, can be reasonably envisaged as being coor-
dinated on the metal fragment.
(21) Silvi, B.; Savin, A. Nature 1994, 371, 683. Savin, A.; Nesper,
R.; Wengert, S.; Fa¨ssler, T. F. Angew. Chem., Int. Ed. Engl. 1997, 36,
1808.
(22) Note also that the bond order of the P-C bond is slightly greater
than 1.
(23) Aray, Y.; Murgich, J .; Luna, M. A. J . Am. Chem. Soc. 1991,
113, 7135.
(24) Up to now, a value of zero found for ꢀ was attributed to a single
or a triple bond. It is important to note that, as expected, the picture
derived from natural bond orbital analysis (NBO) is more classical
(triple C-N bond (σ + 2π) and a nitrogen lone pair). Indeed, as
mentioned by Bachrach (Bachrach, S. M. In Reviews in Compututa-
tional Chemistry; Lipkowitz, K. B., Boyds, D. B., Eds.; VCH: New York,
1994; pp 171-227), NBO rests on allocating electrons through an
orbital occupancy basis, not in terms of the actual locations of electrons.
The ELF analysis is expected to yield a more accurate picture of the
molecule, as it takes into account the possible resonance forms, while
NBO does not.
In comparison to 4e, steric hindrance would be
smaller for 4c. However, this hypothesis cannot explain
the behavior of 4c,e. Another explanation can be found
by considering the localization of the HOMO. The
HOMO of an amino derivative such as 4e shows a π
CN
orbital contribution (Figure 2), which could explain a

30 Organometallics, Vol. 20, No. 1, 2001
Maraval et al.
Ta ble 1. Geom etr ica l P a r a m eter s (An gles (d eg), Bon d Len gth s (Å)) of 4a , Me2N-P (CN)2, a n d 4ea
4a
MP2
Me2N-P(CN)2
4e
B3LYP
MP2
B3LYP
B3LYP
P-C1
1.789
1.789
1.179
1.179
1.844
174.3
174.3
96.5
1.795
1.795
1.156
1.156
1.860
173.4
173.4
97.5
P-C1
1.793
1.815
1.179
1.181
1.689
174.6
172.1
94.6
1.803
1.822
1.156
1.157
1.684
173.5
171.9
95.1
100.8
104.8
116.8
123.2
114.6
-52.4
45.9
P-C1
1.811
1.806
1.157
1.157
1.676
170.2
172.4
96.0
106.8
105.0
115.9
128.9
115.0
-59.0
41.0
P-C2
P-C2
P-C2
C1-N1
C1-N1
C1-N1
C2-N2
C2-N2
C2-N2
P-C3
P-N3
P-N3
N1-C1-P
N2-C2-P
C1-P-C2
C1-P-C3
C2-P-C3
N1-C1-P
N2-C2-P
C1-P-C2
C1-P-N3
C2-P-N3
P-N3-C3
P-N3-C4
C3-N3-C4
C4-N3-P-C1
C4-N3-P-C2
N1-C1-P
N2-C2-P
C1-P-C2
C1-P-N3
C2-P-N3
P-N3-C3
P-N3-C4
C3-N3-C4
98.8
99.8
99.2
98.8
99.8
103.5
114.5
120.4
112.7
-53.0
44.0
C3-N3-P-C1
C4-N3-P-C2
a
Calculations were carried out at the 6-311G** level in all cases.
F igu r e 4. Ionization potentials (IP) of the compounds R-P(CN)2 (4a -e).
In comparison to each other, the 31P and 13C NMR
data for compounds 4a -e do not exhibit any specific
changes. The trend in the 31P NMR chemical shift for
dicyanophosphines 4a -e follows the trend observed for
the corresponding dihydro and dichloro phosphines
RPX2 (X ) H, Cl).25 The signal of the carbon atom in
the P-C-N moiety of 4a -e (δ(13C) 113-119 ppm) is
shifted to the region expected for organonitrile deriva-
tives. In the infrared spectra the symmetric C-N
stretching frequencies for compounds 4a -e26 appear in
the close region of 2166-2182 cm-1. Moreover, it is

Cyanophosphine Derivatives
Organometallics, Vol. 20, No. 1, 2001 31
Ta ble 2. Va lu es of th e Liga n d Con e An gle θ
Ta ble 3. Cr ysta llogr a p h ic Da ta , Su m m a r y of Da ta
Collection , a n d Refin em en t Deta ils for 4c
empirical formula
fw
C20H29N2P
328.42
temp
160 (2) K
0.710 73 Å
wavelength
cryst syst, space group
unit cell dimens
monoclinic, Cc
a ) 17.708(4) Å
b ) 9.518(2) Å
c ) 11.491(2) Å
â ) 95.92(3)°
1926.4(7) Å3
4, 1.132 Mg/m3
0.145 mm-1
712
V
Z, calcd density
abs coeff
F(000)
cryst color
colorless
cryst form
parallelepiped
0.8 × 0.6 × 0.4 mm3
1.50 kW
cryst size
tube power
tube voltage
tube current
collimator size
detector dist
2θ range
50 kV
30 mA
0.5 mm
a
b
From the X-ray crystal structure. Determined from the
70.0 mm
calculated structure with Cerius 2. Mes* ) 2,4,6-tBu3C6H2.
3.3-52.1°
d(hkl) range
ψ movement mode
ψ start
12.453-0.809 Å
rotation
interesting to note that the symmetric C-N frequencies
of Ph2PCN, 4d , and Ph2P(S)CN are in the same range
and are equal to 2173, 2182, and 2180 cm-1, respec-
tively. Then NMR and IR spectroscopic data demon-
strate that in the ground state dicyanophosphines 4a -e
do not exhibit heterocumulene character. This is in
accord with what is observed with other phosphorus-
(III) compounds containing π-acceptor substituents, as
for example CHO.27 The structure of compounds 4a -e
is more in favor with resonance forms I and II rather
than the cumulenic form III.
The steric accessibility of the lone pair of the phos-
phorus atom is also one of the factors which determines
the coordination chemistry of the phosphine compounds.
The steric hindrance of phosphines has been rational-
ized by Tolman in terms of the cone solid angle θ (Table
2).7 In neutral zirconocene(IV) complexes, Cp2ZrXY,
phosphines are often used to stabilize the metal frag-
ment. To the best of our knowledge, Ph2PMe (θ ) 136°)
is the most hindered phosphine that has been coordi-
nated to neutral zirconocene(IV) complexes.28,29 Then,
the significant degree of steric hindrance provided by
the Mes* and the NiPr2 groups should prevent any
possible coordination of the lone pair of the phosphorus
atom in 4c,e on the zirconium fragment of 1.
To gain more insight into the structure of the dicy-
anophosphine 4c, X-ray crystallography was under-
taken (Table 3, Figure 5). The bond distances and bond
angles for 4c are given in Table 4. The phosphorus is
in a pyramidal environment (sum of the three bond
angles 312°). The average P-CN bond length is 1.79 Å,
and the average C-N bond is 1.14 Å. The bond angles
at the phosphorus (C1-P1-C2) and at the carbon atoms
of the cyano groups (P1-C1-N1 and P1-C2-N2) are
95.27, 171.3, and 168.5°, respectively. All these values
are in good agreement with those observed for phos-
phorus tricyanide, P(CN)3. The significant deviation
0.0°
ψ end
250.5°
ψ increment
no. of exposures
irradiation/exposure
index ranges
1.5°
192
5.00 min
-19 e h e 20, -11 e k e 11,
-13 e l e 13
6117/3072 (R(int) ) 0.0638)
99.2%
no. of rflns collected/unique
completeness to 2θ ) 24.41
refinement method
full-matrix least squares on F2
3072/2/217
no. of data/restraints/params
goodness of fit on F2
1.038
final R indices (I > 2σ(I))
R indices (all data)
R1 ) 0.0385, wR2 ) 0.0953
R1 ) 0.0415, wR2 ) 0.0972
-0.05(8)
absolute structure param37
largest diff peak and hole
0.223, -0.182 e/Å3
F igu r e 5. X-ray crystal structure of 4c (ZORTEP drawing
with thermal ellipsoids at 50% probability).
(25) It is well-established that cyanide shields the phosphorus atom,
as does the hydrogen atom; there is an opposite effect for the chloro
atom.3b
from linearity of the P-C-N linkage is surprising, and
up to now there does not appear to be any ready
explanation for it.30 The most striking feature in the
structure of compound 4c is the position of the aryl
group. The plane of the Mes* substituent deviates from
the P-C3 bond axis by 33.4°. However, due to the steric
hindrance of this bulky group, this deviation is com-
(26) The weak antisymmetric C-N stretching frequencies of 4a -e
have not been identified.
(27) Amsallem, D.; Gornitzka, H.; Baceiredo, A.; Bertrand, G. Angew.
Chem., Int. Ed. 1999, 38, 2201.
(28) Cardin, D. J .; Lappert, M. F.; Raston, C. L. Chemistry of
Organozirconium and Hafnium Compounds; Wiley: New York, 1986.
(29) Note that the cationic zirconocene complex [Cp2Zr(H)(Ph2-
PMe)2]+ has been characterized: J ordan, R. F.; Bajur, C. S.; Dasher,
W. Organometallics 1987, 6, 1041.

32 Organometallics, Vol. 20, No. 1, 2001
Maraval et al.
Ta ble 4. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for 4c
effect, the HOMO of the amino derivative presents a
π
orbital contribution, which could explain a preco-
CN
Distances
ordination on the CN system followed by a hydrozir-
conation process. However, to rationalize the difference
in reactivity of 4a -c on one hand and 4e on the other
hand, a theoretical study of the reaction path seems to
be required, especially of the first step yielding the
intermediate resulting from the coordination of the
dicyanophosphine to the metal center. The same theo-
retical investigations will be realized on hydrocyano-
phosphines 5a -c, which exhibited a different chemical
behavior with 1.
It is interesting to note that the topological analysis
of the electron localization function (ELF) in dicyano-
phosphine derivatives R-P(CN)2 provides a new de-
scription of the chemical bond C-N in nitrile com-
pounds. According to the electronic population found for
the C-N fragment, the ELF analysis suggests that the
resonance form P-C+dN- has a strong weight (72%).
This repartition of the electronic population in the CN
does not depend on the group linked to the cyano moiety;
the same electronic population is observed in the case
of acetonitrile.
P-C(1)
1.796(3)
1.838(2)
1.140(3)
1.421(3)
1.388(3)
1.382(3)
P-C(2)
1.800(2)
1.142(3)
1.413(3)
1.388(3)
1.394(3)
P-C(3)
C(1)-N(1)
C(3)-C(4)
C(4)-C(5)
C(6)-C(7)
C(2)-N(2)
C(3)-C(8)
C(5)-C(6)
C(7)-C(8)
Angles
P-C(1)-N(1)
C(1)-P-C(3)
C(1)-P-C(2)
P-C(3)-C(8)
171.2(2)
P-C(2)-N(2)
C(2)-P-C(3)
P-C(3)-C(4)
168.5(2)
114.08(10)
125.20(16)
102.42(10)
95.24(12)
110.97(14)
Ta ble 5. Aver a ge P op u la tion s (N) of Selected
Ba sin s Obta in ed fr om th e Top ologica l An a lysis of
ELF for 4a ,c,e (See Sch em e 2), MeP (O)(CN)2, a n d
MeCN
basin
model 4a model 4c model 4e MeP(O)(CN)2 MeCN
V(P, CN)
V(C,N)
V(P)
2.4
4.3
2.0
3.2
2.4
4.3
2.0
3.2
2.4
4.2
2.1
3.2
2.4
4.3
2.1a
4.5
V(N)
3.1
3.4
a
V(C,CN).
monly observed as soon as a phosphorus atom of a
phosphine is directly linked to the Mes* substituent.31
These structural data confirm the presence of a weak
interaction involving the phosphorus lone pair orbital
and the sp3 carbon C3 of the mesityl group.
Exp er im en ta l Section
All manipulations were performed under an argon atmo-
sphere, either on a vacuum line using standard Schlenk
techniques or in a Braun MB 200-G drybox. Solvents were
freshly distilled from dark purple solutions of sodium/ben-
zophenone ketyl (THF, toluene, benzene, diethyl ether), lithium
aluminum hydride (pentane), or CaH2 (CH2Cl2). Deuterated
NMR solvents were treated with LiAlH4 (C6D6, THF-d8) or
CaH2 (CD2Cl2), distilled, and stored under argon. Nuclear
magnetic resonance (NMR) spectra were recorded at 25 °C on
Bruker MSL-400, AM-250, AC-200, and AC-80 Fourier trans-
form spectrometers. Chemical shifts are given downfield
relative to Me4Si (1H, 13C) or H3PO4 (31P), respectively. 13C
NMR assignments were confirmed by inverse gradient δ(1H)-
δ(13C){1H,31P} HMQC and 31P-1H INEPT NMR experiments.
Elemental and mass spectrum analyses obtained on a Nermag
R10-10H were performed by the analytical service of the
Laboratoire de Chimie de Coordination (LCC) of the CNRS.
RPCl2 (R ) tBu,32 Mes*,33 NiPr2,34 [Cp2Zr(H)(Cl)]n35) were
prepared according to literature procedures. MePCl2, PhPCl2,
AgCN, and Me3SiCN were purchased from Aldrich and used
without further purification.
Con clu sion
We have found an easy way to prepare quantitatively
the cyano primary phosphanes R-P(H)CN (5a -c) from
the addition of Schwartz’s reagent [Cp2Zr(H)Cl]n (1) to
the dicyanophosphine compounds R-P(CN)2 (4a -c).
Surprisingly, it is a hydrozirconation reaction which
occurred on addition of 1 to iPr2N-P(CN)2 (4e). Adding
a further 1 equiv of 1 to 5a -c gave the substitution
product Me-PH2 or the hydrozirconation products 7b,c,
respectively. Delineation of the factorsselectronic and
stericsthat govern the reactivity of Schwartz’s reagent
1 with dicyanophosphine derivatives R-P(CN)2 (4a -
e) is by no means straightforward. Nevertheless, ex-
perimental analysis (NMR, IR, UV-PES, X-ray) showed
that, whatever the R group linked to the P(CN)2 moiety,
there is practically no interaction between the phos-
phorus lone pair and the cyano groups. In the photo-
electron analysis the bands attributed to ionization of
the phosphorus lone pair in 4a -e correspond to either
the highest occupied molecular orbital (HOMO) or
HOMO-1. Then coordination of dicyanophosphine de-
rivatives R-P(CN)2 on the zirconium fragment by the
lone pair of the phosphorus atom is largely expected for
4a ,b. However, for the phosphines 4c,e coordination of
the metal fragment should occur on the cyano groups,
which results in the formation of completely different
products with the metal hydride 1, the hydrocyanophos-
phine 5c and the aldimidophosphine 6, respectively. In
a first approach, one explanation can be found consider-
ing the molecular orbital, from the point of view of
localization and accessibility. Apart from the steric
Typ ica l P r oced u r e for th e P r ep a r a tion of P h osp h i-
n on itr iles 4a a n d P h P (CN)2. To a solution of 2a (21.20
mmol) in CH3CN (20 mL) at room temperature was added 2
equiv of AgCN (5.700 g; 42.50 mmol). The reaction mixture
was stirred for 2.0 h at room temperature. The resulting
suspension was filtered over Celite. The solvent was evapo-
rated under vacuum.
Da ta for 4a . 4a was purified by sublimation (0.1 mmHg,
55 °C) to give a gray powder in 65% yield. Mp: 71 °C. 31P{1H}
1
NMR (C6D6, 62.9 MHz): δ -81.1 ppm. H NMR (C6D6, 200.1
MHz): δ 0.30 (d, 2J HP ) 7.6 Hz, 3H, PCH3) ppm. 13C{1H} NMR
(C6D6, 62.9 MHz): δ 7.3 (d, 1J CP ) 9.3 Hz, CH3), 115.5 (d, 1J CP
) 66.4 Hz, PCN) ppm. IR: γ(CN) 2179, 2060 cm-1. Anal. Calcd
for C3H3N2P: C, 36.75; H, 3.08. Found: C, 36.62; H, 2.94.
(32) Voskuil, W.; Arens, J . F. Recl. Trav. Chim. Pays-Bas 1963, 82,
302.
(30) Emerson, K.; Britton, D. Acta Crystallogr. 1964, 17, 1134.
(31) Dubourg, A. Ph.D. Dissertation, Paul Sabatier University,
Toulouse, France, 1990, No. 117A. Andrianarison, M.; Couret, C.;
Declercq, J .-P.; Dubourg, A.; Escudie´, J .; Ranaivonjatovo, H.; Satge´,
J . Organometallics 1988, 7, 1545. Kandri-Rodi, A.; Declercq, J .-P.;
Ranaivonjatovo, H.; Escudie´, J . Organometallics 1995, 14, 1954.
(33) Cowley, A. H.; Norman, N. C.; Pakulski, M. Inorg. Synth. 1990,
28, 235.
(34) King, R. B.; Sadanani, N. D. Synth. React. Inorg. Met.-Org.
Chem. 1985, 15, 149.
(35) Buchwald, S. L.; LaMaire, S. J .; Nielsen, R. B.; Watson, B. T.;
King, S. M. Tetrahedron Lett. 1987, 28, 3895.

Cyanophosphine Derivatives
Organometallics, Vol. 20, No. 1, 2001 33
Da ta for P h P (CN)2. PhP(CN)2 was purified by distillation
under reduced pressure (0.1 mmHg, 80 °C) to give a white solid
in 78% yield. 31P{1H} NMR (C6D6, 62.9 MHz): δ -74.0 ppm.
1H NMR (C6D6, 200.1 MHz): δ 6.79-7.37 (m, 5H, Ph) ppm.
(1; 0.118 g, 0.46 mmol) and 4c (150 mg, 0.46 mmol) in toluene
(10 mL) was stirred for one night at room temperature. The
volatiles were removed in vacuo, and the product 5c was
extracted with pentane (2 × 20 mL) and isolated as an orange
powder in 98% yield. 31P{1H} NMR (CD2Cl2, 62.9 MHz): δ
-101.6 ppm. 1H NMR (CD2Cl2, 200.1 MHz): δ 1.28 (s, 9H,
1
13C{1H} NMR (C6D6, 62.9 MHz): δ 114.7 (d, J CP ) 60.9 Hz,
3
PCN), 130.4 (d, J CP ) 10.1 Hz, m-Ph), 133.4 (s, p-Ph), 135.1
2
1
(d, J CP ) 25.9 Hz, o-Ph) ppm. The i-Ph carbon atom was not
p-CH3), 1.57 (s, 9H, o-CH3), 1.60 (s, 9H, o-CH3), 5.95 (d, J HP
identified. IR: γ(CN) 2184, 2092 cm-1. Anal. Calcd for
C8H5N2P: C, 60.01; H, 3.15. Found: C, 59.78; H, 3.02.
P r ep a r a tion of 4b. To a solution of 2b (2.00 g, 12.6 mmol)
in CH2Cl2 (10 mL) at room temperature was added 2 equiv of
Me3SiCN (2.50 g; 3.35 mL, 25.10 mmol). The reaction mixture
was stirred for 4 days at room temperature. The volatiles were
removed under vacuum. The residue was distilled under
reduced pressure (5 mmHg, 92 °C) to give 4b as a yellow oil
in 83% yield. 31P{1H} NMR (C6D6, 62.9 MHz): δ -43.5 ppm.
) 249.7 Hz, 1H, PH), 7.51 (d, J HP ) 2.8 Hz, 2H, CHaryl) ppm.
4
13C{1H} NMR (CD2Cl2, 62.9 MHz): δ 30.7 (s, p-C(CH3)), 32.4
4
(d, J CP ) 6.5 Hz, o-C(CH3)), 34.3 (s, p-C(CH3)), 37.5 (s,
1
1
o-C(CH3)), 117.3 (d, J CP ) 23.5 Hz, i-Caryl), 121.2 (d, J CP
)
3
74.4 Hz, PCN), 123.0 (d, J CP ) 4.7 Hz, m-CHaryl), 149.4 (s,
p-Caryl), 155.4 (s, o-Caryl) ppm.
Rea ctivity of 4d w ith 1. A suspended solution of [Cp2Zr-
(H)Cl]n (1; 0.241 g, 0.94 mmol) and 4d (150 mg, 0.94 mmol) in
CH2Cl2 (8 mL) was stirred for one night at room temperature
to give PhPH2 (31P NMR (CH2Cl2, 62.9 MHz): δ -123.8 (t, 1J HP
) 195.0 Hz, PH2) ppm) and other unidentified phosphorus
derivatives.
P r ep a r a tion of 6. A suspended solution of [Cp2Zr(H)Cl]n
(0.282 g, 1.092 mmol) and 4e (0.200 g, 1.092 mmol) in CH2Cl2
(8 mL) was stirred for 1 h at 0 °C. Removal of the solvent in
vacuo of the clear brown reaction mixture gave 6 in nearly
quantitative yield. 31P{1H} NMR (CD2Cl2, 62.9 MHz): δ 15.2
ppm. 1H NMR (CD2Cl2, 200.1 MHz): δ 1.24 (d, 3J HH ) 6.6 Hz,
3
1H NMR (C6D6, 200.1 MHz): δ 0.85 (d, J HP ) 17.5 Hz, 9H,
2
CH3) ppm. 13C{1H} NMR (C6D6, 62.9 MHz): δ 26.3 (d, J CP
)
1
1
16.1 Hz, CH3), 33.3 (d, J CP ) 9.5 Hz, C(CH3)), 113.4 (d, J CP
) 68.5 Hz, PCN) ppm. IR: γ(CN) 2182 cm-1. Anal. Calcd for
C6H9N2P: C, 51.43; H, 6.47. Found: C, 51.27; H, 6.39.
P r ep a r a tion of 4c. To a solution of 2c (2.50 g, 7.20 mmol)
in CH3CN (20 mL) at room temperature was added 2 equiv of
Me3SiCN (1.92 g; 1.92 mL, 14.40 mmol). The reaction mixture
was stirred for 2 h at room temperature. The volatiles were
removed under vacuum, and 4c was obtained as a yellow
powder in 87% yield. Mp: 92.3 °C. 31P{1H} NMR (CD2Cl2, 62.9
MHz): δ -80.6 ppm. 1H NMR (CD2Cl2, 200.1 MHz): δ 1.34
3
3
6H, CH3), 1.28 (d, J HH ) 6.6 Hz, 6H, CH3), 3.46 (sept d, J HP
3
) 13.2 Hz, J HH ) 6.6 Hz, 2H, CHNP), 6.17 (s, 5H, Cp), 6.26
(s, 5H, Cp), 9.34 (d, 2J HP ) 75.4 Hz, 1H, CHdN) ppm. 13C{1H}
NMR (CD2Cl2, 62.9 MHz): δ 23.2 (s, CH3), 53.3 (s, CHNP),
110.7, 111.0 (s, Cp), 120.8 (d, 1J CP ) 101.2 Hz, PCN), 171.8 (d,
1J CP ) 17.0 Hz, CdN) ppm. IR (KBr, CH2Cl2): γ 2155 (CtN),
4
(s, 9H, p-C(CH3)), 1.67 (s, 18H, o-C(CH3)), 7.59 (d, J HP ) 4.2
Hz, 2H, CHAr) ppm. 13C{1H} NMR (CD2Cl2, 62.9 MHz): δ 32.5
(s, p-C(CH3)), 35.5 (d, 4J CP ) 5.8 Hz, o-C(CH3)), 36.9 (d, 3J CP
)
1660 (CHdN) cm-1
.
1
19.7 Hz, o-C(CH3)), 41.6 (s, p-C(CH3)), 118.9 (d, J CP ) 69.1
Hz, PCN), 127.3 (d, 3J CP ) 10.3 Hz, m-Ar), 157.8 (s, p-Ar), 162.3
(d, 2J CP ) 18.9 Hz, o-Ar) ppm. IR: γ(CN) 2166 cm-1. Anal. Calcd
for C20H29N2P: C, 73.14; H, 8.89. Found: C, 72.97; H, 8.85.
P r ep a r a tion of 4e. To a solution of 2e (4.30 g, 21.20 mmol)
in CH3CN (20 mL) at room temperature was added 2 equiv of
AgCN (5.700 g; 42.50 mmol). The resulting solution was heated
under reflux. After 3.5 h, AgCl was eliminated by filtration
and the volatiles were removed under vacuum. 4e was purified
from the residue after distillation under reduced pressure (0.01
mmHg, 125 °C), leading to the final product as a white solid
at room temperature in 86% yield. Mp: 34 °C. 31P{1H} NMR
P r ep a r a tion of 7b. A suspended solution of [Cp2Zr(H)Cl]n
(0.224 g, 0.869 mmol) and 5b (0.100 g, 0.869 mmol) in CH2Cl2
(10 mL) was stirred for 1 h at 0 °C. Removal of the solvent in
vacuo of the brown reaction mixture gave 7b in nearly
quantitative yield. 31P{1H} NMR (CD2Cl2, 62.9 MHz): δ 3.3
3
ppm. 1H NMR (CD2Cl2, 200.1 MHz): δ 1.33 (d, J HP ) 10.6
1
3
Hz, 9H, CH3), 3.61 (dd, J HP ) 213.0 Hz, J HH ) 4.8 Hz, 1H,
2
PH), 6.14 (s, 5H, Cp), 6.31 (s, 5H, Cp), 9.70 (dd, J HP ) 50.1
Hz, J HH ) 4.8 Hz, 1H, CHdN) ppm. 13C{1H} NMR (CD2Cl2,
3
2
1
62.9 MHz): δ 29.9 (d, J CP ) 10.6 Hz, CH3), 31.6 (d, J CP
10.8 Hz, C(CH3)), 110.7 (s, Cp), 112.8 (s, Cp), 176.6 (d, J CP
)
)
1
1
32.4 Hz, CHdN) ppm.
(C6D6, 62.9 MHz): δ -23.0 ppm. H NMR (C6D6, 200.1 MHz):
3
δ 0.71 (d, J HH ) 6.7 Hz, 12H, CH3), 3.09 (m, 2H, CHN) ppm.
P r ep a r a tion of 8. A suspended solution of [Cp2Zr(H)Cl]n
(0.127 g, 0.490 mmol) and 5c (0.150 g, 0.490 mmol) in toluene
(10 mL) was stirred at room temperature. 31P and 1H NMR
showed the formation of 7c along with the formation of 8.
Da ta for 7c. 31P NMR (CD2Cl2, 62.9 MHz): δ -44.0 (d, 1J PH
13C{1H} NMR (C6D6, 62.9 MHz): δ 22.5 (s broad, CH3), 52.6 (s
1
broad, CHN), 117.7 (d, J CP ) 69.7 Hz, PCN) ppm. IR: γ(CN)
2175 cm-1. Anal. Calcd for C8H14N3P: C, 52.45; H, 7.70.
Found: C, 52.35; H, 7.58.
1
P r ep a r a tion of 5a . A suspended solution of [Cp2Zr(H)(Cl)]n
(1; 0.395 g, 1.53 mmol) and 4a (150 mg, 1.53 mmol) in CH2Cl2
(10 mL) was stirred for 1 h at 0 °C. 5a and the solvent were
transferred in a trap immersed in nitrogen liquid. The final
product 5a was used without further treatment. In an identical
) 233.9 Hz, PH) ppm. H NMR (CD2Cl2, 200.1 MHz): δ 1.25
(s, 9H, p-CH3), 1.55 (s, 9H, o-CH3), 1.61 (s, 9H, o-CH3), 5.45
1
(d, J HP ) 233.9 Hz, 1H, PH), 6.04 (s, 5H, Cp), 6.20 (s, 5H,
4
2
Cp), 7.48 (d, J HP ) 1.6 Hz, 2H, CHaryl), 9.20 (d, J HP ) 37.6
Hz, 1H, CHdN) ppm. After 1 day at room temperature, 7c
was totally transformed into 8. The volatiles were removed in
vacuo, and the product 8 was extracted with pentane (2 × 20
mL) and isolated as an orange powder in 90% yield.
1
experiment in CD2Cl2, 5a was identified by H and 31P NMR
experiments as the sole product of the reaction. 31P{1H} NMR
1
(CD2Cl2, 62.9 MHz): δ -119.9 ppm. H NMR (CD2Cl2, 200.1
2
3
Da ta for 8. 31P{1H} NMR (CD2Cl2, 62.9 MHz): δ 95.5 ppm.
1H NMR (CD2Cl2, 200.1 MHz): δ 1.27 (s, 9H, p-CH3), 1.60 (s,
9H, o-CH3), 1.63 (s, 9H, o-CH3), 6.06 (s, 5H, Cp), 6.20 (s, 5H,
MHz): δ 1.47 (dd, J HP ) 5.3 Hz, J HH ) 7.9 Hz, 3H, CH3),
4.15 (dq, J HP ) 227.5 Hz, J HH ) 7.9 Hz, 1H, PH) ppm. 13C-
1
3
{1H} NMR (CD2Cl2, 62.9 MHz): δ 1.6 (d, 1J CP ) 8.0 Hz, PCH3),
1
4
3
119.6 (d, J CP ) 70.9 Hz, PCN) ppm.
Cp), 7.50 (d, J HP ) 0.6 Hz, 2H, CHaryl), 7.98 (dd, J HH ) 16.5
3
2
3
P r ep a r a tion of 5b. A suspended solution of [Cp2Zr(H)(Cl)]n
(1; 0.276 g, 1.07 mmol) and 4b (150 mg, 1.07 mmol) in CH2Cl2
(8 mL) was stirred for 1 h at 0 °C. 5b was isolated from the
reaction mixture by trap-to-trap transfer and used without
further treatment. 31P{1H} NMR (CD2Cl2, 62.9 MHz): δ -59.8
ppm. 1H NMR (CD2Cl2, 200.1 MHz): δ 1.34 (d, 3J HP ) 14.5 Hz,
9H, CH3), 3.98 (d, 1J HP ) 213.0 Hz, 1H, PH) ppm. 13C{1H} NMR
Hz, J HP ) 6.0 Hz, 1H, NH), 8.01 (dd, J HP ) 49.7 Hz, J HH )
16.5 Hz, 1H, PCH) ppm. 13C{1H} NMR (CD2Cl2, 62.9 MHz): δ
4
31.6 (s, p-C(CH3)), 32.5 (d, J CP ) 6.9 Hz, o-C(CH3)), 33.6 (s,
p-C(CH3)), 38.3 (s, o-C(CH3)), 111.9 (s, Cp), 122.0 (s, m-CHaryl),
133.2 (d, J CP ) 58.7 Hz, i-Caryl), 149.9 (s, p-Caryl), 156.4 (s,
o-Caryl), 177.7 (d, J CP ) 61.0 Hz, PdC) ppm.
1
1
P r ep a r a tion of 9. A suspended solution of [Cp2Zr(H)(Cl)]n
(0.282 g, 1.092 mmol) and 6 (0.482 g, 1.092 mmol) in CH2Cl2
(10 mL) was stirred for 1 h at 0 °C. Removal of the solvent in
vacuo of the clear brown reaction mixture gave 9 in nearly
2
(CD2Cl2, 62.9 MHz): δ 31.4 (d, J CP ) 13.6 Hz, CH3), 37.2 (d,
1
1J CP ) 24.6 Hz, C(CH3)), 121.4 (d, J CP ) 41.0 Hz, PCN) ppm.
P r ep a r a tion of 5c. A suspended solution of [Cp2Zr(H)(Cl)]n

34 Organometallics, Vol. 20, No. 1, 2001
Maraval et al.
quantitative yield. 31P{1H} NMR (C6D6, 62.9 MHz): δ 52.0. 1H
NMR (C6D6, 200.1 MHz): δ 1.43 (d, 12H, 3J HH ) 6.6 Hz, CH3),
3.42 (m, 2H, CHNP), 6.15, 6.10, 6.03, 5.96 (s, 5H, Cp), 10.08
molecular orbitals interactions. ELF calculations and topologi-
cal analysis were carried out with TopMoD.42 Visualizations
of ELF isosurfaces and basins were done with the freeware
SciAn.43
2
(d, J HP ) 61.8 Hz, 2H, CHdN) ppm. 13C{1H} NMR (CD2Cl2,
62.9 MHz): δ 24.0 (s, CH3), 53.4 (s, CHNP), 110.5 (s, Cp), 181.5
Top ologica l An a lysis of ELF . The bonding in molecules
may be investigated by different techniques. The topological
methods investigate well-defined local functions. In the atoms
in molecules theory of Bader, criteria are proposed in order to
decide the existence and nature of the bonds from the electron
density analysis. Two atoms are bound if a bond path may be
found between them. The value of the Laplacian of the charge
density at bond critical points allows one to discriminate
between closed-shell and electron-shared (covalent) interac-
tions. In the latter case, however, it is not possible to
distinguish between the formation of an electron pair of
opposite spin and the sharing of only one electron.
1
(d, J CP ) 30.1 Hz, CHdN). IR: γ(CHdN) 1647, 1624 cm-1
.
X-r a y An a lysis of 4c. Single crystals of 4c were obtained
in acetonitrile. X-ray diffraction analysis was carried out at
160 K on a STOE IPDS (imaging plate diffraction system)
equipped with an Oxford Cryosystems cooler device. The
crystal-to-detector distance was 70 mm with the crystal
oscillated in æ. Crystal decay was monitored by measuring 200
reflections per image. The final unit cell was obtained by the
least-squares refinement of 5000 reflections using Mo KR
radiation. (λ ) 0.710 73 Å). The structure was solved by direct
methods (SIR9236) and refined by least-squares procedures on
F2 (SHELXL97).37 H atoms were located on a difference
Fourier map, but they were introduced in calculations in
idealized positions (d(C-H) ) 0.96 Å), and their atomic
coordinates were recalculated after each cycle of refinement.
They were given isotropic thermal parameters 20% higher
than those of the carbon to which they are connected. All non-
hydrogen atoms were refined anisotropically, and in the last
The same type of analysis may be performed using the
electron localization function (ELF), but it will additionally
allow measurement of the pairing of the electrons. ELF is
defined as ELF ) 1/[1 + (Dσ/Dσ0)2], in which Dσ and Dσ are
0
the excess local kinetic energies due to the Pauli repulsion (the
difference of the kinetic energy of the actual Fermionic system
with respect to the one of the bosonic system having the same
density) for the actual system and for a homogeneous electron
gas of the same density, respectively. ELF takes values
between 0 and 1. When electrons are alone or form pairs of
opposite spin, the Pauli principle has little influence, and the
ELF is therefore close to 1. However, when parallel-spin
electrons are close to one another, ELF comes close to 0.
The topological analysis of the ELF gradient field yields a
partition of the molecular space in basins and attractors (i.e.,
the local maxima of ELF). The basins are classified into core,
valence bonding, and nonbonding basins. A core basin contains
the nuclei X (except for a proton) and will be referred to as
C(X). A bonding basin lies between two or more core basins.
Valence bonding basins are further distinguished, depending
on their connectivity to the core basins. The topological
analysis of the gradient field of ELF assigns a region of spaces
the basinsto each attractor, providing a partition of the
molecular space analogous to that made in hydrography to
define river basins and watersheds. Each valence basin is
characterized by its synaptic order, which is the number of
core basins with which it shares a common boundary. There-
fore, the monosynaptic basins correspond to nonbonded pairs
(referred to as V(X)), whereas the di- and polysynaptic basins
are related to bi- and multicentric bonds (referred to as
V(X1,X2,X3, ...). A localization domain is a space region limited
by an iso-ELF surface. It has been shown that ELF performs
partitions of the molecular space which are consistent with
the Lewis description and with the VSEPR model.
cycles of refinement
a weighting scheme was used. The
absolute configuration has been assigned on the basis of the
refinement of the Flack parameter X,38 which is sensitive to
the polarity of the structure; the value of this parameter has
been found to be close to zero (-0.05(8)), clearly indicating the
correctness of the enantiomer choice. The drawing of the
molecule was realized with ZORTEP39 with thermal ellipsoids
at the 50% probability level for non-hydrogen atoms. Further
details on the crystal structure investigation are available on
request from the Director of the Cambridge Crystallographic
Data Centre, 12 Union Road, GB-Cambridge, U.K., on quoting
the full journal citation.
P h otoelectr on Sp ectr oscop y. The photoelectron spectra
were recorded on a Helectros 0078 instrument, using a He I/He
II source, equipped with a 127° cylindrical analyzer and
monitored by a microcomputer supplemented with a digital
analogic converter. The spectra were calibrated on the known
autoionization of helium at 4.98 eV and nitrogen ionization
at 15.59 eV.
Com p u ta tion a l Deta ils. Ab initio calculations were per-
formed using Gaussian 98.18 The optimization and vibrational
analysis were carried out at a 6-311G(d,p) level of theory with
inclusion of dynamic electron correlation at the MP2 level. Cal-
culations were also realized using the density functional theory
method,15 using the exact exchange functional Becke 3LYP16
with the basis set 6-311G(d,p). Equilibrium molecular geom-
etries were fully optimized, and on the resulting geometry,
second derivatives were calculated to find out if a minimum
or a saddle point was obtained. Graphical representation of
the nature of the molecular orbitals were obtained using a
MOLDEN program.40 Natural bond orbital analysis,41 which
represents ab initio molecular wave functions in terms of
localized Lewis structures, was realized in order to provide a
quantitative analytical framework for the interpretation of
Ack n ow led gm en t. This work was supported by the
CNRS (France) and the Ministe`re de la Recherche et
de l’Enseignement Supe´rieur (Ph.D. fellowship for
A.M.). Acknowledgment is made to the institut du
De´veloppement de Ressources en Informatique Scien-
tifique (IDRIS: Orsay, France), administrated by the
CNRS, for the calculation facilities for this research and
Dr. Gijs Schaftenaar for allowing us to use his graphical
program Molden. We also thank Prof. B. Silvi for fruitful
discussions.
(36) Altomare, A.; Cascarano, G.; Giacovazzo, G.; Guargliardi, A.;
Burla, M. C.; Polidori, G.; Camalli, M. SIR92: a Program for Automatic
Solution of Crystal Structures by Direct Methods. J . Appl. Crystallogr.
1994, 27, 435.
(37) Sheldrick, G. M. SHELXL97: Program for the Refinement of
Crystal Structures; University of Go¨ttingen, Go¨ttingen, Germany,
1997.
(38) Flack, H. Acta Crystallogr. 1983, A39, 876.
(39) Zolna¨ı, L. ZORTEP: Graphical Program for X-ray Structure
Analysis, University of Heidelberg, Heidelberg, Germany, 1998.
(40) See: http://www-srs.caos.kun.nl/∼schaft/molden.
(41) (a) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988,
88, 899. (b) Foster, J . P.; Weinhold, F. J . Am. Chem. Soc. 1980, 102,
7211.
Su ppor tin g In for m ation Available: Tables giving atomic
coordinates, bond lengths and angles, and anisotropic thermal
parameters for 4c. This material is available free of charge
via the Internet at http://pubs.acs.org.
OM000704H
(43) Pepke, E.; Murray, J .; Lyons, J .; Hwu, T.-Y. SciAn; Supercom-
puter Research Institute of the Florida State University at Tallahassee,
Tallahassee, FL, 1996.
(42) Noury, S.; Krokidis, X.; Fuster, F.; Silvi, B. TopMoD Package;
Universite´ Pierre et Marie Curie, 1997.