Boron trifluoride activation of ethylene oligomerization and polymerization catalysts

Article abstract of DOI:10.1016/S0020-1693(02)01345-2

Addition of BF₃·Et₂O to {[(C₆H₅)₂PC₆H₄(μ- CO₂)-κ³P,O,O′]Ni(η¹-CH ₂C₆H₅)}₂ gives [(C₆H₅)₂PC₆H₄C(O-BF ₃)O-κ²P,O]Ni(η³-CH₂C ₆H₅) (4). The single crystal structure of 4 is consistent with substantial positive charge at the nickel atom. Addition of ethylene to solutions of 4 in toluene provides ethylene oligomers. The average molecular weight of the products is larger than that of the oligomers obtained using [(C₆H₅)₂PC₆H₄C(OB(C ₆F₅)₃)O-κ²P,O]Ni(η ³-CH₂C₆H₅) (1). When 2 equiv. of Et₂O·BF₃ are added to {(H₃C)C[N(2,6-(CHMe₂)₂-C₆H ₃)]C(O)[N(2,6-(CHMe₂)₂-C₆H ₃)]-κ²N,O}Ni(η¹-CH₂C ₆H₅)(PMe₃) (5) one obtains {(H₃C)C[N(2,6-(CHMe₂)₂-C₆H ₃)]C[O-BF₃][N(2,6-(CHMe₂)₂-C ₆H₃)]-κ²N,N′}Ni(η ³-CH₂C₆H₅) (7). Two isomers of 7 are observed by NMR spectroscopy and are incorporated into the crystal lattice, as shown by X-ray diffraction studies. Polyethylene is obtained upon exposure of a solution of 7 to ethylene. Active species can also be generated in situ by addition of BF₃ gas, Et₂O·BF₃ or ^tBuMeO·BF₃ to 5.

Full text of DOI:10.1016/S0020-1693(02)01345-2

Inorganica Chimica Acta 345 (2003) 95ꢀ102
/
www.elsevier.com/locate/ica
Boron trifluoride activation of ethylene oligomerization and
polymerization catalysts
Zachary J.A. Komon, Guillermo C. Bazan *, Cindy Fang, Xianhui Bu
Departments of Chemistry and Materials, University of California, Santa Barbara, CA 93106, USA
Received 15 May 2002; accepted 23 September 2002
Dedicated in honor of Professor Richard R. Schrock
Abstract
Addition of BF3×
CH2C6H5) (4). The single crystal structure of 4 is consistent with substantial positive charge at the nickel atom. Addition of ethylene
/
Et2O to {[(C6H5)2PC6H4(m-CO2)-k3P,O,O?]Ni(h1-CH2C6H5)}2 gives [(C6H5)2PC6H4C(OꢀBF3)O-k2P,O]Ni(h3-
/
to solutions of 4 in toluene provides ethylene oligomers. The average molecular weight of the products is larger than that of the
oligomers obtained using [(C6H5)2PC6H4C(OB(C6F5)3)O-k2P,O]Ni(h3-CH2C6H5) (1). When 2 equiv. of Et2O×
/
BF3 are added to
C6H3)]-k2N,O}Ni(h1-CH2C6H5)(PMe3) (5) one obtains {(H3C)C[N(2,6-
C6H3)]-k2N,N?}Ni(h3-CH2C6H5) (7). Two isomers of 7 are observed by NMR
spectroscopy and are incorporated into the crystal lattice, as shown by X-ray diffraction studies. Polyethylene is obtained upon
{(H3C)C[N(2,6-(CHMe2)2ꢀ
/
C6H3)]C(O)[N(2,6-(CHMe2)2ꢀ
/
(CHMe2)2ꢀC6H3)]C[OꢀBF3][N(2,6-(CHMe2)2ꢀ
/
/
/
exposure of a solution of 7 to ethylene. Active species can also be generated in situ by addition of BF3 gas, Et2O×
/
BF3 or tBuMeO×
/
BF3 to 5.
# 2002 Published by Elsevier Science B.V.
Keywords: Ethylene polymerization; Homogeneous catalysis; Nickel; Borane activation; Branched polyethylene
1. Introduction
tions available for precatalyst development [4]. An
important consideration in metallocene chemistry is
that the final cost of the activator frequently exceeds
that of the transition metal reagent. The majority of co-
catalysts are derived from aluminum alkyls, alkylalumi-
noxanes or perfluoroaryl boranes. An interplay between
several kinetic and thermodynamic factors determines
the utility of a co-catalyst but, after activation, the co-
catalyst is transformed into an anion which is loosely
associated with the electrophilic metal center [5]. Be-
cause the structure and dynamics of alkylaluminoxanes
are complex and they must be used in large excess, the
highly Lewis acidic perfluoroaryl borane reagents, such
as B(C6F5)3, provide better defined catalysts. In a typical
reaction, B(C6F5)3 reacts with a neutral metallocene
Transition metal-mediated ethylene polymerization
and oligomerization catalysts are under intense investi-
gation in academic and industrial laboratories [1].
Metallocenes and other single-site catalysts provide
access to new manufacturing opportunities and to
materials with properties not obtained using the more
established heterogeneous processes [2]. Oligomerization
catalysts have also recently appeared with extremely
high reactivity toward ethylene and with remarkable
specificity toward 1-alkenes [3]. A general feature of
these reactions is the generation of highly active species
by combination of a transition metal precatalyst with a
suitable co-catalyst.
The importance of co-catalysts has coupled their
of general type Cp2ZrMe2 (Cpꢁ
derivative) to give Cp2ZrMe(MeB(C6F5)3) [6]. It is
widely accepted that anionꢀcation separation or dis-
/
cyclopentadienyl
design and performance to new ligandꢀmetal combina-
/
/
placement of the borate anion (MeB(C6F5)3) by the
olefin is required for the insertion reaction to take
place [5].
* Corresponding author. Tel.: ꢂ
4120.
E-mail address: bazan@chem.ucsb.edu (G.C. Bazan).
/
1-805-893 5538; fax: ꢂ1-805-893
/
0020-1693/02/$ - see front matter # 2002 Published by Elsevier Science B.V.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 1 3 4 5 - 2

96
Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ102
/
We recently reported that attachment of B(C6F5)3 to a
complexes are useful precatalysts for the oligomeriza-
tion and polymerization of ethylene.
site removed from the monomer trajectory can be used
to activate nickel-based oligomerization and polymer-
ization catalysts [7]. For example, [(C6H5)2PC6H4C-
(OB(C6F5)3)O-k2P,O]Ni(h3-CH2CMeCH2) displays an
activity toward ethylene orders of magnitude greater
than the precursor [(C6H5)2PC6H4C(O)O-k2P,O]Ni(h3-
CH2CMeCH2) [8]. Faster initiation occurs by use of the
isoelectronic h3-CH2C6H5 ligand in [(C6H5)2PC6H4C-
(OB(C6F5)3)O-k2P,O]Ni(h3-CH2C6H5) (1), relative to
the methallyl counterpart. The two resonances shown
for 1 (Scheme 1) show the net withdrawal of electron
density from the metal atom. Conceptually, this mode of
activation is different from other reactions where the
counteranion needs to decoordinate in order to offer a
vacant site for the monomer or be displaced by the
incoming substrate [9]. Compound {(H3C)C[N(2,6-
2. Experimental
2.1. General remarks
All manipulations were carried out in an inert atmo-
sphere of Ar or nitrogen using standard glovebox and
Schlenk
B(C6F5)3)O-k2P,O]Ni(h3-CH2C6H5)
{[(C6H5)2PC6H4(m-CO2)-k3P,O,O?]Ni(h1-CH2C6H5)}2
[7b] and {(H3C)C[N(2,6-(CHMe2)2ꢀC6H3)]C(O)[N(2,6-
(CHMe2)2ꢀ
C6H3)]-k2N,O}Ni(h1-CH2C6H5)(PMe3) (5)
[7a] were prepared according to literature procedures.
BF3×Et2O and BF3×
MeOtBu were obtained from Al-
techniques
[15].
[(C6H5)2PC6H4C(Oꢀ
/
(1) [7b],
/
/
/
/
(CHMe2)2ꢀ
/
C6H3)]C[Oꢀ
/
B(C6F5)3][N(2,6-(CHMe2)2ꢀ
/
drich and used as received. Tris(pentafluorophenyl)bor-
ane was obtained from Boulder Scientific Company,
Mead, Colorado and was sublimed under high vacuum
prior to use. Benzene-d6 and C5H12 were distilled from
C6H3)]-k2N,N?}Ni(h3-CH2C6H5) (2) is highly active
and produces high molecular weight polyethylene.
The Lewis acidity of the activator, B(C6F5)3 is
comparable to that of BF3 [10ꢀ12], therefore it seemed
/
sodium benzophenone ketyl. Toluene was distilled from
1
Naꢀ
/
K alloy. H, 19F and 31P{1H} NMR spectra were
acquired using a Varian Unity 400 spectrometer at
plausible that the action of the simple trihaloborane
reagent could be used in generating initiators analogous
to 1 and 2. Despite this competitive acidity, the use of
BF3 in metallocene chemistry is prevented by facile
fluoride transfer to the transition metal [13]. In 1 and 2,
charge buildup at boron and the metal are not as
extreme as in the prototypical case of Cp2ZrMe(MeB-
(C6F5)3). This feature, coupled to the greater tolerance
toward functional groups displayed by late metals, is
expected to make fluoride transfer less pronounced.
Ethylene olgomerization catalysts have been prepared
by addition of BF3 to allyl-nickel alkoxides, though the
activated species were not clearly defined [14]. We note
also that certain cationic nickel olefin polymerization
catalysts have been shown to be inert towards anions
1
ambient temperature. H spectra are referenced intern-
ally to the solvent resonances, while 19F and 31P{1H} are
referenced externally to a,a,a-trifluorotoluene (0.0 ppm)
and 85% H3PO4 (0.0 ppm), respectively. 13C{1H} and
11B{1H} spectra were recorded on a Varian Unity 500
spectrometer and are referenced internally to solvent
resonances and externally to BF3×/Et2O, respectively. IR
data were collected on a Shimadzu FTIR-8300 spectro-
meter. GPC analyses were carried out by Equistar
Chemicals, Cincinnati, OH. Elemental analyses were
performed by Desert Analytics, Tucson, AZ.
2.2. Preparation of [(C6H5)2PC6H4C(Oꢀ
/BF3)O-
ꢃ
ꢃ
k2P,O]Ni(h3-CH2C6H5) (4)
such as BF4 and PF6 [1d,1e].
In this contribution we show that, indeed, the above
considerations are appropriate and that BF3 can be used
to generate complexes analogous to 1 and 2. These
Boron trifluoride diethyl etherate (70 mg, 490 mmol)
was added with rapid stirring to a slurry of 3 (172 mg,
Scheme 1. Ethylene oligomerization and polymerization initiators.

Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ
/
102
97
378 mmol) in 3 ml CH2Cl2. The reaction mixture quickly
turned dark red and homogeneous and was stirred for
an additional hour. The solvent was removed in vacuo.
The residue was redissolved in a few drops of CH2Cl2,
precipitated by addition of C6H5CH3 (approximately 5
ml) and was triturated overnight. The solid was
collected by filtration, washed once with C6H5CH3
and twice with C5H12. Drying in vacuo yields 126 mg
(64%) of an orange powder. The structure of 4 was
confirmed by X-ray crystallography [16]. Anal. Calc. for
C26H21BF3NiO2P: C, 59.72; H, 4.05. Found: C, 59.50;
3J(H,H)ꢁ
1.38 (d, 6H, 3J(H,H)ꢁ
3J(H,H)ꢁ
6.9 Hz; MeCHCH3), 1.37 (s, 2H; CH2Ph),
1.20 (d, 6H, 3J(H,H)ꢁ
6.7 Hz; MeCHCH3), 1.09 (d, 6H,
3J(H,H)ꢁ6.7 Hz; MeCHCH3), 0.96 (d, 6H, 3J(H,H)ꢁ
6.9 Hz; MeCHCH3), 0.94 (d, 6H; 3J(H,H)ꢁ
6.9 Hz;
6.7 Hz;
/
6.7 Hz; MeCHCH3), 1.50 (s, 2H; CH2Ph),
/6.7 Hz; MeCHCH3), 1.37 (d, 6H,
/
/
/
/
/
MeCHCH3), 0.85 (d, 6H, 3J(H,H)ꢁ
/
MeCHCH3); 13C{1H} NMR (C6D6, 125 MHz): d
180.3, 177.6, 141.0, 140.8, 138.52, 138.47, 136.0, 135.4,
129.2, 127.8, 127.5, 124.6, 124.3, 124.1, 123.7, 104.3,
104.0, 39.7, 38.8, 29.8, 29.7, 29.4, 29.3, 24.8, 24.5, 24.3,
24.1, 23.93, 23.85, 23.5, 21.0, 14.6; 19F NMR (C6D6, 376
1
H, 3.98%; H NMR (CD2Cl2, 400 MHz): d 8.38 (ddd,
3
4
4
1H, J(H,H)ꢁ
/
7.9, J(H,P)ꢁ
7.20 (m, 16H; Ph and h3-benzyl-
m,p), 7.14 (d, 2H, J(H,H)ꢁ
7.7 Hz; h3-benzyl-o), 1.63
(d, 2H, 3J(H,P)ꢁ3.8 Hz; CH2Ph); 13C{1H} NMR
/
4.8, J(H,H)ꢁ
/
1.0 Hz; o-
MHz, CF3Ph): d ꢃ
/
81.7, ꢃ
/
82.0; 11B{1H} NMR (C6D6,
0.71 (br).
to carboxylate), 7.88ꢀ
/
160 MHz, BF3×OEt2): d ꢃ
/
/
3
/
/
2.4. Exchange reaction with 4 and B(C6F5)3
(CD2Cl2, 125 MHz): d 170.3 (COO), 135.8ꢀ128.8 (m;
/
C-arom.), 116.6 (d, 2J(C,P)ꢁ
/
5.5 Hz; h3-benzyl-o), 27.9
Inside the glovebox, a resealable NMR tube was
charged with 14.0 mg (26.8 mmol) 4 and approximately
500 ml CD2Cl2. 1H, 31P, 19F and 11B NMR spectra were
obtained as references for the starting material. The tube
was then brought back into the box and a solution of
13.7 mg (26.9 mmol) B(C6F5)3 in approximately 100 ml
CD2Cl2 was added to the solution of 4 with no
observable changes in color. 1H, 31P, 19F and 11B
NMR spectra were recollected and showed only com-
pound 1. The solution was then degassed on a high
2
(d, J(C,P)ꢁ
/
5.5 Hz; CH2Ph); 19F NMR (CD2Cl2, 376
MHz, CF3Ph): d ꢃ
MHz, H3PO4): d 22.1; 11B{1H} NMR (CD2Cl2, 160
MHz, BF3×OEt2): d ꢃ0.82.
/
87.4; 31P{1H} NMR (CD2Cl2, 162
/
/
2.3. Preparation of {(H3C)C[N(2,6-(CHMe2)2ꢀ
C6H3)]C[OꢀBF3][N(2,6-(CHMe2)2ꢀC6H3)]-
k2N,N?}Ni(h3-CH2C6H5) (7)
/
/
/
Boron trifluoride diethyl etherate (37 mg, 260 mmol)
in 250 ml C6H6 was added to compound 5 (80 mg, 130
mmol) in 1 ml C6H6 and mixed thoroughly. The solution
vacuum line at ꢃ78 8C and backfilled with BF3 gas.
Reacquisition of multinuclear NMR did not show the
formation of 4; only 1 was observed.
/
color changed from orangeꢀred to dark red. This
/
solution was allowed to stand for 2 additional hours,
during which a colorless solid precipitates and the
solution becomes less intensely colored. Filtration
through oven-dried glass wool and removal of volatiles
in vacuo gave an orange powder which was redissolved
in approximately 1 ml of C6H5CH3 and placed in a
2.5. Ethylene oligomerization procedure
Inside the glovebox, the appropriate mass of a stock
solution (1.00 mmol gꢃ1) of 4 or powder form of 1 was
weighed into a 5 ml vial, transferred to a FisherꢀPorter
/
bottle and diluted with 30 ml (25.8 g) of C6H5CH3. Once
the apparatus was assembled, it was brought out of the
box and placed in a room temperature (r.t.) water bath.
Ethylene was introduced at the desired pressure and the
oligomerization was allowed to proceed for 15 min
monitoring the monomer consumption with a mass flow
controller to determine the activity of the catalyst. At
the end of the reaction time, the vessel was vented and
the reaction quenched by addition of 5 ml H2O and
stirring rapidly for approximately 5 min. The C6H5CH3
layer was then collected and analyzed by gas chromato-
graphy to determine the distribution of oligomers, as
well as by proton NMR to determine the composition of
a-olefins, internal olefins and dimers [18].
C5H12 filled chamber at ꢃ35 8C for 2 days. Red crystals
/
were isolated, after removal of the mother liquor and
drying in vacuo to yield 20 mg (25%) of 7. A second crop
obtained in the same manner gave 12 mg (15%) of the
desired product. The structure of 7 was confirmed by X-
ray crystallography [17] Anal. Calc. for C34H44BF3-
NiN2O: C, 65.53; H, 7.12; N, 4.49. Found: C, 65.47; H,
7.21; N, 4.49; two pseudo-rotamers in a 1:1 ratio are
1
observable in the NMR spectra. H NMR (C6D6, 400
i
6.82 (m, 12H; both rotamers Pr2C6H3),
MHz): d 7.20ꢀ
/
6.63 (t, 1H, 3J(H,H)ꢁ
/
7.5 Hz; h3-benzyl-p), 6.57 (t, 1H,
3J(H,H)ꢁ
3J(H,H)ꢁ
3J(H,H)ꢁ
3J(H,H)ꢁ
/
7.6 Hz; h3-benzyl-p), 6.19 (pst, 2H,
7.8 Hz; h3-benzyl-m), 6.13 (pst, 2H,
7.7 Hz, h3-benzyl-p), 5.86 (d, 4H,
7.3 Hz; both rotamers h3-benzyl-o), 3.65
/
/
/
2.6. Ethylene polymerization procedure
3
(sept, 2H, J(H,H)ꢁ
3J(H,H)ꢁ
6.9 Hz; CHMe2), 3.18 (sept, 2H, J(H,H)ꢁ
6.8 Hz; CHMe2), 2.81 (sept, 2H, 3J(H,H)ꢁ
6.8 Hz;
CHMe2), 2.28 (s, 3H; Me), 2.24 (s, 3H, Me), 1.50 (d, 6H,
/
6.9 Hz; CHMe2), 3.23 (sept, 2H,
3
/
/
Inside the glovebox, a FisherꢀPorter bottle was
/
/
charged with 14 mmol of 7 or 5 with the appropriate
amount of BF3 and diluted with 30 ml of C6H5CH3. A

98
Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ102
/
stainless steel reactor head was assembled and the
system brought out of the glove box. Ethylene was fed
continuously for 15 min under 100 psig pressure with a
r.t. external water bath to control the reaction tempera-
ture. The ethylene flow rate was monitored by a mass
flow controller. The reaction was quenched by release of
ethylene pressure and addition of C3H6O (30 ml).
Polyethylene powder was collected by filtration and
dried in vacuo. The activity of the catalyst was
determined by the mass of the polymer obtained and
correlated against the amount of ethylene measured by
the mass flow controller. Branch numbers were calcu-
lated from the integration values of methyl, methylene
and methine proton regions in 1H NMR spectra
collected at 115 8C in C6H6-d6 and 1,2,4-trichloroben-
zene (v/v, 1:4). The polymers were further characterized
by GPC and DSC.
Fig. 1. ORTEP view of 4, showing the atom-numbering scheme.
Thermal ellipsoids are shown at 30% probability level. Hydrogen
˚
atoms have been omitted for clarity. Selected bond distances (A):
Ni(1)ꢀO(2), 1.905(2); Ni(1)ꢀP(1), 2.1270(7); Ni(1)ꢀC(37), 1.957(3);
Ni(1)ꢀC(31), 2.029(2); Ni(1)ꢀC(36), 2.228(2); O(2)ꢀC(47), 1.241(3);
O(1)ꢀC(47), 1.291(2). Selected bond angles (8): O(2)ꢀNi(1)ꢀP(1),
94.92(5); Ni(1)ꢀC(37), 100.18(8); C(37)ꢀNi(1)ꢀC(36),
P(1)ꢀ
70.73(10); C(36)ꢀNi(1)ꢀO(2), 95.93(8).
/
/
/
/
/
/
3. Results and discussion
/
/
/
/
/
/
/
/
/
3.1. Synthesis and characterization of [(C6H5)2-
BF3)O-k2P,O]Ni(h3-CH2C6H5) (4)
PC6H4C(Oꢀ
/
isometric to that of 1. The only statistically significant
difference lies within the CꢀO bond distances. In 1, the
two CꢀO distances are equal in length within statistical
error (2.7s). In 4, the C(47)ꢀ
shorter than the C(47)ꢀO(1) distance (1.291(2) A) which
suggests a stronger contribution from resonance struc-
ture of type A in Scheme 1 and a more cationic nickel
center in 4 relative to 1.
Addition of excess BF3×
/
Et2O (approximately 1.5
/
equiv.) to
a
slurry of {[(C6H5)2PC6H4(m-CO2)-
/
k3P,O,O?]Ni(h1-CH2C6H5)}2 (3) in dichloromethane
results in the formation of a red homogeneous solution.
Addition of toluene to this solution precipitates an
orange powder that can be isolated by filtration. The 1H
NMR spectrum of this solid in CD2Cl2 is consistent with
˚
O(2) length (1.241(3) A) is
/
˚
/
the structure [(C6H5)2PC6H4C(Oꢀ
CH2C6H5) (4) as shown in Eq. (1).
/
BF3)O-k2P,O]Ni(h3-
IR spectroscopy in CH2Cl2 can be used to gauge
ð1Þ
Single crystals of 4 suitable for X-ray diffraction
studies were obtained by layering pentane on top of a
CD2Cl2 solution, and allowing the mixture to slowly
diffuse overnight. The results of this study, shown in
Fig. 1, show a pseudo square planar arrangement of
ligands around nickel. With the exception of the boron
substituents, the solid state structure of 4 is nearly
differences in Cꢀ
1617 cmꢃ1, which is higher than the n(Cꢁ
cmꢃ1. However, the relative Cꢁ
O bond strengths
determined by IR spectroscopy do not indicate which
of the two possible resonance structures dominates, only
that 4 lies further towards one of the resonance extremes
than 1. If the bond order observed in the solid structure
/
O bond strengths. For 4, n(Cꢁ
/
O) is
/
O) for 1, 1602
/

Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ
/
102
99
Table 1
Oligomerization of C2H4 by 1 and 4
a
b
c
d
d
d
Entry
Precatalyst
P (psig)
[Ni] (mM)
Activity
K
% a-Olefin
% 2-Alkyl-1-alkene
% Internal
18
e
1
2
3
4
5
4
4
4
4
1
100
100
100
50
470
220
90
1066
1210
1290
590
0.26
0.21
0.18
0.19
0.07
60
98
22
2
100
99
220
90
1
14
100
4500
79
7
a
Reaction conditions: toluene, 15 min, 20 8C.
Kg ethylene consumed per mole precatalyst per hour.
Determined by GC.
Mole percent determined by 1H NMR spectroscopy.
The exothermicity of the reaction prevented temperature control.
b
c
d
e
is maintained in solution, then the higher stretching
frequency observed in 4 supports the notion of a more
positively charged nickel atom.
transfer are first order in monomer. Similar dependence
of rates on monomer concentration occurs for the
catalyst derived from 1 [7b]. Because the reaction is
highly exothermic, it is not possible to control tempera-
ture at higher nickel concentrations and a higher K
factor is observed (entry 1). Interestingly, reactions with
4/C2H4 have a higher K factor than the 1/C2H4 system,
indicating that they produce products with larger
average molecular mass. That 1 and 4 are nearly
isostructural near the vicinity of the nickel atom
indicates that the difference in activity and chain length
selectivity is largely electronic in origin.
A competition experiment was set up by adding 1
1
equiv. of B(C6F5)3 to a solution of 4 in CD2Cl2. H
NMR spectroscopy showed quantitative formation of 1.
Formation of 4 was not reestablished, even after
addition of excess (approximately 5 equiv.) of BF3 gas
to the solution. Therefore, the borane adduct is stronger
in 1, relative to 4 and B(C6F5)3 is a stronger Lewis acid
towards 3 than BF3. This observation coupled with the
higher activity of 1 vis-a`-vis 4 is the strongest evidence
that the nickel center in 1 is more positively charged
than in 4.
3.3. Synthesis and characterization of {(H3C)C[N(2,6-
(CHMe2)2ꢀ
/
C6H3)]C[Oꢀ
/
BF3][N(2,6-(CHMe2)2ꢀ
/
3.2. Ethylene oligomerization by [(C6H5)2PC6H4C(Oꢀ
/
C6H3)]-k2N,N?}Ni(h3-CH2C6H5) (7)
BF3)O-k2P,O]Ni(h3-CH2C6H5) (4)
Addition of 2 equiv. BF3×
/
OEt2 to {(H3C)C[N(2,6-
C6H3)]-
Rapid consumption of ethylene is observed upon
exposure to a toluene solution of 4. The results of these
(CHMe2)2ꢀ
/
C6H3)]C(O)[N(2,6-(CHMe2)2ꢀ
/
k2N,O}Ni(h1-CH2C6H5)(PMe3) (5) in C6D6 intensifies
the red color of the solution. H NMR spectroscopy
1
experiments are summarized in Table 1. Entries 1ꢀ
/
3
show the effect of concentration. Comparison against
entry 4 shows that reducing ethylene pressure by a factor
of two decreases the activity by a half. The activity of 4
towards ethylene is lower than for 1 (entries 3 and 5).
The activity versus time profiles of the two catalysts
measured by use of a mass flow controller are similar in
shape, with the activity dropping with reaction time.
The selectivity for a-olefins is enhanced by lowering the
shows complete consumption of 5 and the clean forma-
tion of a product. Since 31P NMR signals are observed
which do not correspond to F3Bꢀ
/
PMe3, we suspect that
this initial product is compound 6 in Eq. (2). An
alternative coordination of BF3 to amide nitrogen
cannot be ruled out at this stage. After 45 min, a fine
dispersion of F3Bꢀ
product exhibits resonances indicative of an h3-benzyl
ligand in the range of 5.8ꢀ6.7 ppm, significantly upfield
from the h1-benzyl species 5 (7ꢀ
8 ppm). Two species
/
PMe3 forms. The new organometallic
concentration of nickel (entries 1ꢀ
ethylene pressure (entries 2ꢀ4). Similar trends are
observed for 1 [7b].
The K factor, defined by Kꢁ
(moles Cn olefin)ꢁRP/(RPꢂRCT), where RP and RCT
/3) or by increasing the
/
/
/
form and are assigned to coordination isomers of the
benzyl ligand about nickel. The product does not
contain a coordinated phosphine, as determined by 31P
NMR spectroscopy. These data are consistent with the
/
(moles Cnꢂ2 olefin)/
/
/
are the rates of propagation and chain termination,
respectively, is readily determined by GC analysis of the
oligomer products and can be used to probe how
structural variations on the active species influence the
propagation and insertion processes. Entries 2, 3, and 5
show that the K factor remains constant within experi-
mental certainty. Thus, both propagation and chain
formation of {(H3C)C[N(2,6-(CHMe2)2ꢀ
/
C6H3)]C[Oꢀ
/
BF3 ][N(2,6-(CHMe2 )2 ꢀ
/
C6 H3 )]-k2 N ,N ?}Ni(h3 -
CH2C6H5) (7), as shown in Eq. (2). Compound 7 exists
as a mixture of interconverting isomers.
X-ray quality crystals were obtained by dissolving the
product in a small amount of toluene, layering pentane

100
Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ102
/
ð2Þ
on top and letting the solvents diffuse at ꢃ
/
35 8C. The
dered benzyl fragment. The error associated with the
structural model prevents a detailed analysis of metrical
parameters. Nonetheless, the connectivity is unambig-
uous and clearly shows that the BF3 has been coordi-
nated by the carbonyl functionality and that the two
isomers are due to benzyl ligand orientation, rather than
N,O bonding variations from the carboxamide function-
ality.
structure of this compound is shown in Fig. 2. The two
possible coordination isomers of the h3-benzyl ligand
are incorporated into the lattice, resulting in a disor-
3.4. Ethylene polymerization by {(H3C)C[N(2,6-
(CHMe2)2ꢀ
/
C6H3)]C[Oꢀ
/
BF3][N(2,6-(CHMe2)2ꢀ
/
C6H3)]-k2N,N?}Ni(h3-CH2C6H5) (7)
A series of polymerization reactions were studied in
which compound 7 was isolated and used or generated
directly from 5 by addition of a suitable BF3 source.
This study is summarized in Table 2. Entry 1 demon-
strates that 7 is an active initiator for ethylene poly-
merization. Generation of 7 in situ, either by addition of
BF3 gas or an etherate adduct, prior to introduction of
ethylene also yields active polymerization systems (en-
tries 2ꢀ5). This procedure allows one to save the time
/
Fig. 2. ORTEP view of 7 showing the average structure of the two
isomers. Thermal ellipsoids are shown at 20% probability level.
Hydrogen atoms have been omitted for clarity.
and effort required to isolate and purify 7. Highest
activity is obtained when 10 equiv. of BF3 gas is added
to 5 (entry 2). A branched structure is observed for all
Table 2
Activity of 7 towards C2H4
a
b
c
Entry
Source
[B]/[Ni]
Activity
Mw/1000
PDI
Branches
Tm (8C)
d
1
2
3
4
5
6
7
200
270
170
140
150
350
769
45.7
343
259
82.2
302
22
20
32
22
16
31
20
129
115
126
114
106
123
e
5/BF3
5/BF3
10
f
7.76
18.6
4.25
4.22
4.60
e
5/Et2O×
5/tBuMeO×
5/B(C6F5)3
/
BF3
13
13
13
/BF3
a
Polymerization conditions: [Ni]ꢁ
/
0.47 mM in toluene, 100 psig ethylene, 15 min.
b
c
d
e
f
Kg product (mol Ni h)ꢃ1
.
Determined by 1H NMR and corresponds to the number of branches/1000 carbons.
Bimodal distribution with centers at 31.6 and 631ꢄ
103 g molꢃ1
Obtained from aliquots of a saturated solution of the gas in toluene.
/
.
Saturated solution of BF3 in toluene (approximately 27 mM) was used as reaction solvent.

Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ
/
102
101
(g) G.J.P. Britovsek, V.C. Gibson, D.F. Wass, Angew. Chem. 111
(1999) 448; Angew. Chem., Int. Ed. Engl. 38 (1999) 429.;
(h) M. Bochmann, J. Chem. Soc., Dalton Trans. 3 (1996) 255.
[2] (a) Y.V. Kissen, in: J.I. Kroschwitz, M. Howe-Grant (Eds.), Kirk-
Othmer Encyclopedia of Chemical Technology, 4th ed., Wiley-
Interscience, New York, 1996;
polymers and the amount of branching is controlled to a
slight extent by the source of BF3. For example, higher
t
branching occurs when using BuMeO×
/
BF3 instead of
Et2O×
/
BF3 (entries 4 and 5). In comparison to 2/C2H4 in
which 2 is generated in situ (entry 6), the polymerization
system produced by 7/C2H4 under these conditions
exhibits a decreased activity towards ethylene.
(b) A.A. Montagna, Chemtech (1995) 45;
(c) P.M. Morse, Chem. Eng. News 76 (1998) 11;
(d) A.H. Tullo, Chem. Eng. News 78 (2000) 35;
(e) P. Short, Chem. Eng. News 78 (2000) 22.
[3] (a) J.S. Rogers, X. Bu, G.C. Bazan, J. Am. Chem. Soc. 122 (2000)
730;
4. Conclusions
(b) J.S. Rogers, G.C. Bazan, C.K. Sperry, J. Am. Chem. Soc. 119
(1997) 9305;
In summary, we have shown that it is possible to
activate single site oligomerization and polymerization
catalysts by use of the inexpensive reagent BF3 [14,19].
This co-activator may be used directly as a gas or in the
form of etherate adducts. The activated precatalysts
have been fully characterized by NMR spectroscopy and
X-ray crystallography. The key reactivity design consists
of attaching the Lewis acid at a site away from monomer
approach and the absence of a counteranion, which
must dissociate to allow monomer insertion.
(c) B.L. Small, M. Brookhart, J. Am. Chem. Soc. 120 (1998) 7143;
(d) R.W. Barnhart, G.C. Bazan, T. Mourney, J. Am. Chem. Soc.
120 (1998) 1082;
(e) G.J.P. Britovsek, V.C. Gibson, B.S. Kimberley, P.J. Maddox,
S.J. McTavish, G.A. Solan, A.J.P. White, D.J. Williams, Chem.
Commun. 7 (1998) 849.
[4] E.Y. Chen, T.J. Marks, Chem. Rev. 100 (2000) 1391.
[5] Considerable mechanistic research continues on the relationship
between the two charged species. For relevant discussion see: S.
Beck, S. Lieber, F. Schaper, A. Geyer, H.-H. Brintzinger, J. Am.
Chem. Soc. 123 (2001) 1483.
[6] X. Yang, C.L. Stern, T.J. Marks, J. Am. Chem. Soc. 113 (1991)
3623.
[7] (a) B.Y. Lee, G.C. Bazan, J. Vela, Z.J.A. Komon, X. Bu, J. Am.
Chem. Soc. 123 (2001) 5352;
5. Supplementary material
(b) Z.J.A. Komon, X. Bu, G.C. Bazan, J. Am. Chem. Soc. 122
(2000) 12379;
Crystallographic data (excluding structure factors) for
the structures reported in this paper have been deposited
with the Cambridge Crystallographic Data Centre,
(c) Z.J.A. Komon, X. Bu, G.C. Bazan, J. Am. Chem. Soc. 122
(2000) 1830.
[8] M.C. Bonnet, F. Dahan, A. Ecke, W. Keim, R.P. Schultz, I.
Tkatchenko, J. Chem. Soc., Chem. Commun. 5 (1994) 615.
[9] M.S.W. Chan, T. Ziegler, Organometallics 19 (2000) 5182.
[10] 1H NMR experiments have shown that BF3 is of similar acidity to
B(C6F5)3 towards trimethyl- and triethylamine. See: A.G. Massey,
A.J. Park, J. Organomet. Chem. 5 (1966) 218.
CCDC Nos. 169902ꢀ169903 for compounds 4 and 7.
Copies of the data may be obtained free of charge on
application to The Director, CCDC, 12 Union Road,
/
Cambridge, CB2 1EZ, UK (fax: ꢂ44-1223-336-033;
/
e-mail: deposit@ccdc.cam.ac.uk or www: http://
www.ccdc.cam.ac.uk).
[11] More recently, BF3 was reported to be a stronger Lewis acid than
B(C6F5)3 towards OPEt3. See: M.A. Beckett, D.S. Brassington,
S.J. Coles, M.B. Hursthouse, Inorg. Chem. Commun. 3 (2000)
530.
Acknowledgements
[12] Our own experiments have shown B(C6F5)3 to be the stronger
Lewis acid towards trimethyl phosphine. The measured chemical
31P{1H} shifts of Me3P0
/
BF3 and Me3P0
/
B(C6F5)3 are ꢃ28.3
/
This work was supported by the Department of
Energy. The authors would like to acknowledge Mr.
Javier Vela for technical assistance.
(m) and ꢃ5.4 (br, m), respectively, relative to 85% H3PO4. Both
/
31P signals are multiplets due to 19F, 11B and 10B coupling.
[13] S.H. Strauss, Chem. Rev. 93 (1993) 929.
[14] H. Bonnemann, J.D. Jentsch, Appl. Organomet. Chem. 7 (1993)
¨
553.
[15] B.J. Burger, J.E. Bercaw, in: A.L. Wayda, M.Y. Darensbourg
(Eds.), Experimental Organometallic Chemistry, ACS Sympo-
sium Series 357, American Chemical Society, Washington DC,
1987, p.79.
References
[1] (a) H.H. Brintzinger, D. Fischer, R. Mulhaupt, B. Rieger, R.M.
¨
Waymouth, Angew. Chem. 107 (1995) 1255; Angew. Chem., Int.
Ed. Engl. 34 (1995) 1143.;
[16] Crystal data for 4: C26H21BF3NiO2P, Mrꢁ522.92, crystal
/
dimensionsꢁ
P21/c, aꢁ8.967(2), bꢁ
104.130(4)8, Vꢁ
Tꢁ185 K, absorption coefficient mꢁ
collected on a Bruker SMART CCD diffractometer using Mo Ka
/
0.20ꢄ
/
0.13ꢄ
/
0.08 mm3, monoclinic, space group
˚
(b) W. Kaminsky, H. Sinn (Eds.), Transition Metals and
Organometallics as Catalysts for Olefin Polymerization, Springer,
Berlin, 1988;
/
/
15.782(4), cꢁ
/
16.847(4) A, bꢁ
/
3
˚
/2311.8(9) A , Zꢁ
/
4, rcalc
ꢁ
/
1.502 Mg mꢃ3
,
/
/
0.955 mmꢃ1. Data were
(c) G. Fink, R. Mulhaupt, H.H. Brintzinger (Eds.), Ziegler
¨
Catalysts, Springer, Berlin, 1995;
˚
radiation (lꢁ
independent (Rint
Final indices R1ꢁ
/
0.71073 A), 23 719 reflections collected, 5366
0.0534). Refinement on F2 with SHELXTL
0.0369, wR2ꢁ0.0800 with I ꢀ2s(I). The
(d) S.D. Ittel, L.K. Johnson, M. Brookhart, Chem. Rev. 100
(2000) 1169;
ꢁ
/
/
.
/
/
(e) S. Mecking, Coord. Chem. Rev. 203 (2000) 325;
(f) A. Togni, R.L. Halterman (Eds.), Metallocenes, Wiley-VCH,
New York, 1998;
largest peak and hole in a final difference electron density map
ꢃ3
˚
0.324 e A
were 0.893 and ꢃ
/
.

102
Z.J.A. Komon et al. / Inorganica Chimica Acta 345 (2003) 95ꢀ102
/
[17] (a) Crystal data for 7: C34H44BF3N2NiO, Mrꢁ
/
623.23, crystal
Final indices R1ꢁ
/
0.0967, wR2ꢁ
/
0.2266 with I ꢀ2s(I). The
/
dimensionsꢁ
P21/n, aꢁ10.614(5), bꢁ
99.627(9)8, Vꢁ
165 K, absorption coefficient mꢁ
collected on a Bruker SMART CCD diffractometer using Mo
/
0.20ꢄ
/
0.20ꢄ
/
0.02 mm3, monoclinic, space group
˚
largest peak and hole in a final difference electron density map
ꢃ3
˚
0.310 e A
/
/
20.812(9), cꢁ
/
15.025(6) A, bꢁ
/
were 0.578 and ꢃ
/
.
[18] For details see supplementary material for Ref. [7c].
3
˚
/
3272(2) A , Zꢁ
/
4, rcalc
ꢁ
/
1.265 Mg mꢃ3, Tꢁ
Data were
/
/
0.638 mmꢃ1
.
[19] The use of BF3 as cocatalyst in the heterogeneous, Zieglerꢀ
/Natta
type polymerization of butadiene is known, see: M.C. Throck-
morton, J. Appl. Polym. Sci. 42 (1991) 3019.
˚
Ka radiation (lꢁ
/
0.71073 A), 10 399 reflections collected, 3060
.
independent (Rint
ꢁ
/
0.1137). Refinement on F2 with SHELXTL