Synthesis and reactivity of N-heterocycle-B(C6F 5)3 complexes. 3. generation of N-methylpyrrol-2-yl and N-methylindol-2-yl borate zwitterions with acidic sp3 carbons

Article abstract of DOI:10.1021/om0499308

The stoichiometric reactions of N-methylpyrrole and N-methylindole with B(C₆F₅)₃ produce the zwitterionic species 2-[tris(pentafluorophenyl)borane]-5H-1-methylpyrrole (8) and 2-[tris-(pentafluorophenyl)borane]-3H-1-methylindole (4), in which a C(α)-B bond and an acidic sp³ methylene carbon are formed in the heterocyclic part of the molecule. Both derivatives present a restricted rotation around the C(α)-B and/or B-C₆F₅ bonds, and their rotational barriers (13.8 and 14.8 kcal mol^-1 for 3 and 4, respectively) were calculated from ¹H NMR experimental data. A kinetic study of the reaction, carried out by following the conversion by NMR, gave rate constant values (at 298 K in dichloromethane) of 3 × 10 ^-5 and 6 × 10^-5 M^-1 s^-1 for 3 and 4, respectively. Complexes 3 and 4 react quantitatively with triethylamine to give the corresponding triethylammonium salts 5 and 6. Both the zwitterionic complexes and their ammonium salts are efficient activators of Ind ₂ZrMe₂ for the polymerization of ethylene.

Full text of DOI:10.1021/om0499308

Organometallics 2004, 23, 5135-5141
5135
Syn th esis a n d Rea ctivity of N-Heter ocycle-B(C6F 5)3
Com p lexes. 3. Gen er a tion of N-Meth ylp yr r ol-2-yl a n d
N-Meth ylin d ol-2-yl Bor a te Zw itter ion s w ith
Acid ic sp 3 Ca r bon s
Francesca Focante,‡ Isabella Camurati,† Daniele Nanni,‡ Rino Leardini,‡ and
Luigi Resconi*,†
Basell Polyolefins, Centro Ricerche “G. Natta”, I-44100 Ferrara, Italy, and
Dipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna,
Viale Risorgimento 4, I-40136 Bologna, Italy
Received J anuary 27, 2004
The stoichiometric reactions of N-methylpyrrole and N-methylindole with B(C6F5)3 produce
the zwitterionic species 2-[tris(pentafluorophenyl)borane]-5H-1-methylpyrrole (3) and 2-[tris-
(pentafluorophenyl)borane]-3H-1-methylindole (4), in which a C(R)-B bond and an acidic
sp3 methylene carbon are formed in the heterocyclic part of the molecule. Both derivatives
present a restricted rotation around the C(R)-B and/or B-C6F5 bonds, and their rotational
1
barriers (13.8 and 14.8 kcal mol-1 for 3 and 4, respectively) were calculated from H NMR
experimental data. A kinetic study of the reaction, carried out by following the conversion
by NMR, gave rate constant values (at 298 K in dichloromethane) of 3 × 10-5 and 6 × 10-5
M-1 s-1 for 3 and 4, respectively. Complexes 3 and 4 react quantitatively with triethylamine
to give the corresponding triethylammonium salts 5 and 6. Both the zwitterionic complexes
and their ammonium salts are efficient activators of Ind2ZrMe2 for the polymerization of
ethylene.
Sch em e 1
In tr od u ction
We have previously reported that pyrrole and indole
react with tris(pentafluorophenyl)borane, B(C6F5)3, to
give quantitative and instantaneous conversion to the
5H-pyrrole-B(C6F5)3 complex 1 and the 3H-indole-
B(C6F5)3 complex 2, respectively (Scheme 1).1 Many
N-H-pyrroles and -indoles with different substituents
(e.g., R, RO, Ar, Cl) give the same reaction with a formal
N-to-C hydrogen shift. Some of these adducts are
efficient activators of group 4 metallocenes in the
polymerization of olefins.2 We now report the results of
the reaction between B(C6F5)3 and N-methylated pyrrole
and indole. With these substrates, a B-C(R) bond is
formed, and although the heterocycle lacks the poten-
tially movable hydrogen at the 1-position, the reaction
products again contain an acidic methylene group in the
heterocyclic moiety, generated by a formal vinylic C-H
activation.3
Sch em e 2
Resu lts a n d Discu ssion
B(C6F5)3 at room temperature, in toluene or dichlo-
romethane, gave high yields of compounds 3 and 4,
respectively (Scheme 2). In both cases the reaction was
rather slow and reached 95% conversion after 4-10
days, depending on the initial concentration.
Mass spectra obtained by ESI technique revealed for
both compounds a parent peak deriving from the mass
addition of borane and heterocycle; however the detailed
structures of the compounds were mainly characterized
by NMR analyses. In particular, 1H NMR in CD2Cl2
showed for 3 a broad singlet at 4.7 ppm for the
1. Syn th esis a n d Ch a r a cter iza tion . The reaction
of N-methylpyrrole or N-methylindole with 1 equiv of
* Corresponding author. E-mail: luigi.resconi@basell.com.
† Basell Polyolefins.
‡ Universita` di Bologna.
(1) Guidotti, S.; Camurati, I.; Focante, F.; Angellini, L.; Moscardi,
G.; Resconi, L.; Leardini, R.; Nanni, D.; Mercandelli, P.; Sironi, A.;
Beringhelli, T.; Maggioni, D. J . Org. Chem. 2003, 68, 5445.
(2) Resconi, L.; Guidotti, S. (Basell Polyolefins) Int. Pat. Appl. WO
01/62764 (priority 24.2.2000). Bonazza, A.; Camurati, I.; Guidotti, S.;
Mascellani, N.; Resconi, L. Macromol. Chem. Phys.2004, 205 (3), 319.
(3) Dash, A. K.; J ordan, R. F. Organometallics 2002, 21, 777.
10.1021/om0499308 CCC: $27.50 © 2004 American Chemical Society
Publication on Web 09/30/2004

5136 Organometallics, Vol. 23, No. 22, 2004
Focante et al.
F igu r e 1. 1H NMR spectra of 3 at variable temperature
(in CD2Cl2).
R-methylene group, a singlet at 3.6 ppm for the N-
methyl group, and the signals for the two vinylic protons
(6.9 ppm, br s, H3, and 7.4 ppm, d, H4). DEPT, COSY,
NOESY, HMBC, and HSQC experiments gave further
support to the structure (see Experimental Section).
Analogously, structure 4 was deduced by the presence
in the 1H NMR spectrum in CD2Cl2 of a diagnostic broad
AB system at 4.6 ppm with a two-proton area value,
revealing not only the presence of a CH2 group, but also
that these two protons are diastereotopic. For the same
compound 13C NMR revealed for carbon 2 a signal at a
very low field (a multiplet, 214.3-217.4 ppm, likely due
to a scalar spin-spin coupling with 19F of the C6F5 rings,
through nonbonding orbital overlap).4 Variable-temper-
ature 1H NMR experiments confirmed for both com-
plexes the presence of a restricted rotation around some
of the C-B bonds, likely due to steric interaction with
the methyl group. Figure 1 shows the variable-temper-
F igu r e 2. 1H NMR spectra of 4 at variable temperature
(in C2D2Cl4).
1
ature H NMR spectra for compound 3. At 298 K the
methylene protons gave one slightly broad signal that,
below 250 K, decoalesced into an AB-type spectrum.
Compound 4 showed the presence of diastereotopic
protons at somewhat higher temperatures compared to
3, the methylene signal being a sharp singlet at 346 K,
a broad AB system at 298 K, and a well-defined one at
263 K (Figure 2).
these compounds are chiral molecules. The anisochro-
nous methylene signals are therefore the result of pairs
of stereolabile conformational enantiomers that can be
detected by NMR techniques through the diastereotop-
icity of the methylene protons. The interconversion of
the two atropisomers probably occurs by restricted
rotations around the B-C(R) and/or the B-C6F5 bonds,
as suggested for the previously reported B-N com-
plexes.1
As observed before,1 the spectroscopic behavior of
adducts 3 and 4 can be explained by assuming that
(4) Mallory, F. B.; Mallory, C. W.; Ricker, W. M. J . Am. Chem. Soc.
1975, 97, 4770.

N-Heterocycle-B(C6F5)3 Complexes
Organometallics, Vol. 23, No. 22, 2004 5137
F igu r e 4. Plot of 1/[A] vs time for reaction A + B f C at
298 K, where A is the heterocycle, B is B(C6F5)3, and C is
the final adduct (plot a: synthesis of 3 in CH2Cl2; b:
synthesis of 4 in C6D6, and b′ synthesis of 4 in CH2Cl2).
Sch em e 3
F igu r e 3. (Left) Experimental 1H NMR signals of the
methylene hydrogens of 4 as a function of temperature.
(Right) Computer simulations obtained with the rate
constants indicated.
The enantiomerization process was quantitatively
studied for both 3 and 4. Simulation of their 1H NMR
spectra at various temperatures allowed determination
of the rate constant of interconversion at each temper-
ature. For adduct 3 a virtually free rotation was
observed above 300 K (k > 105 s-1), whereas its two
enantiomeric conformers do not interconvert on the
NMR time scale below 250 K; simulation of the 1H NMR
spectra at intermediate temperatures gave rate constant
values of 9 and 18 s-1 at 256 and 263 K, respectively,
from which we calculated a value of 13.8 ( 0.2 kcal
mol-1 for the activation barrier of the enantiomerization
(∆Gq). Compound 4 showed instead free rotation above
340 K and no interconversion below 260 K; simulation
of congested imine-B(C6F5)3 adducts has been recently
reported by Piers and co-workers.5
The methylene groups in compounds 3 and 4 are
acidic enough to protonate triethylamine: both com-
plexes reacted quantitatively and instantaneously with
1 equiv of NEt3, at room temperature in CH2Cl2, to give
the corresponding triethylammonium salts, tris(pen-
tafluorophenyl)-(1-methylpyrrol-2-yl)borate(1-) triethyl-
ammonium (5) and tris(pentafluorophenyl)-(1-methyl-
indol-2-yl)borate(1-) triethylammonium (6), respec-
tively (Scheme 3). Like in the case of 1 and 2,1 the
driving force for this reaction is clearly the restoration
of aromaticity in the five-membered heterocyclic ring.
1
of the H NMR spectra at intermediate temperatures
gave rate constants of 160, 58, and 32 s-1 at 305, 297,
and 284 K, respectively, from which we calculated a ∆Gq
value of 14.8 ( 0.2 kcal mol-1 for the enantiomerization
barrier. The simulation results for adduct 4 are shown
in Figure 3.
1
Both structures (5 and 6) were characterized by H
NMR experiments, and the structure of compound 6 was
also confirmed by X-ray diffraction analysis.6
2. Kin etic Tr ea tm en t of Da ta . Both heterocycles
(N-methylpyrrole or -indole) reacted slowly with B(C6F5)3
(4-10 days at room temperature). The concentration
values [A] for the starting heterocyclic substrate were
Since both 3 and 4 have an N-Me iminium group
linked to the boron atom and they differ only in the
presence of sp2-CH or sp3-CH2 moieties beta to the
borate group, it is evident that the two methylenic
protons of 4 bring about an increase as important as
ca. 1 kcal mol-1 with respect to the single vinylic proton
of 3 in the enantiomerization process. The interconver-
sion barrier is therefore originated mainly by the steric
hindrance between the fluorinated rings and the N-
methyl group, but the hydrogen atoms of the pyrrolic
â-carbon play a role as well, although a minor one. It is
also worth noting that the barrier calculated for 4 is
strictly comparable to that obtained for N-[tris(pen-
tafluorophenyl)borane]-2,4-dimethyl-5H-pyrrole (14.5
kcal mol-1),1 in which the borate moiety, although
linked to a nitrogen atom, is surrounded by the same
kind of bonds and groups. Another restricted rotation
1
deduced by the H NMR spectra recorded at different
times and were plotted as [A]-1 versus time. Linear
correlation curves, indicative of second-order reaction,
were obtained for both compounds. The reaction of
formation of 3 in CH2Cl2 at room temperature gave a
rate constant value of 3 × 10-5 M-1 s-1 (Figure 4, curve
a). The same value was calculated for compound 4 when
the reaction was carried out in C6D6 at room tempera-
ture (k ) 3 × 10-5 M-1 s-1, Figure 4, curve b), whereas
the reaction in CH2Cl2 was much faster (k ) 6 × 10-5
(5) Blackwell, J . M.; Piers, W. E.; Parvez, M.; McDonald, R.
Organometallics 2002, 21, 1400.
(6) Sironi, A.; Mercandelli, P. Unpublished results.

5138 Organometallics, Vol. 23, No. 22, 2004
Focante et al.
Sch em e 5
more feasible. Intermediate 4r can evolve to 4 either
by 1,2-hydride migration from C-2 to C-3 or, more likely,
by prior abstraction of the 2-proton by a base (probably
free N-methylindole itself) followed by re-protonation
at the nucleophilic â-position.
F igu r e 5. Eyring plot for the synthesis of 4 in C2D2Cl4.
The 3,2-borane shift (or the direct attack at the
2-position) is probably due to some stabilizing effect of
the type of charge separation: indeed, the negative
boron and the positive nitrogen atom are closer in
structure 4 compared to intermediate 4â. A thermody-
namic preference for the formation of the R-isomer is
also suggested by theoretical calculations (DFT), which
predict a difference of 8.4 kcal/mol between structures
4â and 4.
Sch em e 4
It is worth noting that 1,2-dimethylindole does not
react with B(C6F5)3. The lack of reaction may lend
support to the preference of 4r to undergo a base-
catalyzed proton transfer from C-2 to C-3, rather than
to rearrange by a 1,2-hydride migration. Indeed, if an
anionotropic rearrangement were involved in interme-
diate 7, the 2-methyl group would be expected to
migrate easily to the 3-position, giving the 3-methyl-
substituted analogue of 4. However, we cannot exclude
that the lack of any reaction could be due to some
additional steric hindrance that would prevent attack
of B(C6F5)3 to the 3-position and/or its translocation to
the 2-position.
M-1 s-1, Figure 4, curve b′). The difference in reaction
rate between the two solvents is in agreement with their
different polarity; in fact the dielectric constants7 are
2.28 and 8.93 respectively for benzene and dichlo-
romethane. Figure 5 shows the temperature dependence
of the rate constant of 4 in C2D2Cl4 in the range 298-
335 K. The Eyring plot gave the following activation
parameters: ∆Hq ) 6.6 kcal mol-1 and ∆Sq ) -60 eu.
3. Mech a n ism of th e Rea ction betw een N-Meth -
ylp yr r ole or N-Meth ylin d ole a n d Bor a n e. The only
species detectable by NMR techniques were the starting
heterocycle and the final borate complex. It was not
possible to isolate and characterize any reaction inter-
mediate, and for this reason we assume that the first
reaction step is a slow electrophilic attack that follows
second-order kinetics and determines the global rate of
the process. In the case of the formation of 4, we can
assume either an initial attack of the borane to C-3 (4â)
followed by 1,2-migration to the R-position (4r) or a
direct attack at C-2 yielding the intermediate 4r in one
step (Scheme 4). Although it has been demonstrated
that direct attack at an R-position can also occur,
electrophilic reagents strongly prefer to attack indoles
at their â-carbons, even when that position carries a
substituent;8 therefore, the migration route seems
We also found that the anion 6a (the indolyl-borate)
was instantaneously and quantitatively protonated to
4 by 3H-2-methylindolium (the cation of indolium salt
8, Scheme 5), with concomitant formation of free 2-me-
thylindole. Since the cation of 8 can be considered a
species analogous to the generic BH+ indicated in
Scheme 4, this experiment demonstrates that indoles
and the corresponding indolium cations can efficiently
act as proton-transfer catalysts, according to our as-
sumption on the last step of Scheme 4.
In the case of N-methylpyrrole, we can apply the same
mechanism, but without the need of a borane rear-
rangement (Scheme 6). Indeed, B(C6F5)3 can directly
attack the R-position, which is the preferred one in
electrophilic additions to pyrroles, giving intermediate
3r; this can lose and recover its proton, owing to the
catalytic action of free N-methylpyrrole, eventually
yielding borate 3. Analogously to 1,2-dimethylindole,
1,2,5-trimethylpyrrole did not react either with B(C6F5)3.
(7) Lide, R. D. Handbook of Chemistry and Physics, 76th ed.; CRC
Press: London, 1995, section 6.
(8) J oule, J . A.; Mills, K. Heterocyclic Chemistry, IV ed.; Blackwell:
London, 2000; Chapter 16.

N-Heterocycle-B(C6F5)3 Complexes
Organometallics, Vol. 23, No. 22, 2004 5139
Ta ble 1. Eth ylen e P olym er iza tion Resu ltsa
cocatalyst
yield (g)
kgPE/(mmolZr × h)
kgPE/(gcat × h)b
comments
TIBA
0
0
0
blank test
2
3
4
5
6
9
10
3.21 ( 0.36
2.43 ( 0.37
1.60 ( 0.16
7.61 ( 0.88
1.74 ( 0.08
5.71 ( 0.11
7.05 ( 0.88
12.9 ( 1.4
9.74 ( 1.5
6.4 ( 0.7
30.5 ( 3.5
6.93 ( 0.3
22.85 ( 0.45
28.2 ( 3.5
12.2
9.8
6.4
27.4
5.9
16.0
22.9
benchmark test average of three tests
average of four tests
average of three tests
average of three tests
average of two tests
average of two tests
average of two tests
a
Applied conditions: heptane 100 mL, scavenger TIBA 0.1 mmol, P(C2) ) 4 bar-g, Tp ) 80 °C, time ) 15 min, Ind2ZrMe2 ) 1 µmol,
B/Zr ) 1.1 mol/mol, premix in 4 mL of toluene, no aging. bcat ) zirconocene + activator.
Sch em e 6
that includes a slow electrophilic attack of B(C6F5)3 to
the heterocycle followed by a fast structural rearrange-
ment to the final products, which, in the case of 4, also
entails a 1,2-borane migration from C(3) to C(2). Further
experiments are in progress to investigate this reaction.
Exp er im en ta l Section
Gen er a l P r oced u r es a n d Sta r tin g Ma ter ia ls. All opera-
tions were performed under nitrogen by using conventional
Schlenk-line techniques. Solvents were purified by degassing
with N2 and passing over activated (8 h, N2 purge, 300 °C)
Al2O3 and stored under nitrogen. Indole (Aldrich, purity 98%
or Fluka, purity 99%), N-methylindole (Aldrich, purity 97%),
2-methylindole (Aldrich, purity 98%), 1,2-dimethylindole (Al-
drich, purity 99%), 3-methylindole (Fluka, purity 98%), N-
methylpyrrole (Aldrich, purity 99%), NEt3 (Aldrich, purity
99.5%), and B(C6F5)3 (Boulder) were used as received.
Both 4r (or 4â) and 3r prefer to isomerize to 4 and 3,
respectively, to achieve structures with more substituted
double bonds and lower internal steric hindrances;
indeed, species 4 and 3 have the C-B bond in the same
plane as the heterocycle.
4. Meta llocen e Activa tion . Compounds 3-6 are
efficient activators of bis(indenyl)zirconium dimethyl in
the polymerization of ethylene, likely generating, upon
loss of methane, a zirconocene methyl cation-borate an-
ion pair, in which the anion is sterically bulky and pro-
bably only very weakly coordinating. Table 1 shows the
preliminary results obtained for ethylene polymerization
1
The H and 13C NMR spectra were obtained using a Bruker
DPX 200 spectrometer operating in the Fourier transform
mode at 200.13 and 50.33 MHz, respectively. The samples were
dissolved in CD2Cl2 or C6D6. As reference the residual peak of
CHDCl2 or C6HD5 in the 1H spectra (5.35 and 7.15 ppm,
respectively) and the peak of the solvent in the 13C spectra
(53.80 ppm for CD2Cl2 and 128.00 ppm for C6D6) were used.
Proton spectra were acquired with a 15° pulse and 2 s of delay
between pulses; 32 transients were stored for each spectrum.
The carbon spectra were acquired with a 45° pulse and 6 s of
delay between pulses; about 512 transients were stored for
each spectrum. CD2Cl2 (Aldrich, 99.8% atom D) was used as
received, whereas C6D6 (Aldrich, 99% atom D) was dried over
activated 4 Å molecular sieves before use. Preparation of the
samples was carried out under nitrogen using standard inert
atmosphere techniques. Mass spectra (MS) were obtained by
ESI technique from acetonitrile (3) or methanol (4) solutions
of suitable samples. The melting points were obtained by using
a capillary Electrothermal instrument. Bis(indenyl)zirconium
dimethyl was prepared as previously described.10 The H2O-
B(C6F5)3 complex was prepared as described in ref 11.
with complexes 3-6. Results using the reference activa-
-
tors [CPh3+][B(C6F5)4-] (9) and [HNMe2Ph+][B(C6F5)4
]
(10) are also reported for comparison.9 Remarkably, the
most active complex 5 is also more active than 9 and
10. 3 is more active than 4, likely because of the lower
steric hindrance of the reactive CH2 in 3 versus 4.
Con clu sion s
N-Methylpyrrole and N-methylindole react with
B(C6F5)3 (1:1) in a novel one-pot reaction giving rise to
unexpected zwitterionic adducts in which a C-B bond
is generated and a formal R-R′ (or R-â for N-meth-
ylindole) hydrogen shift takes place. The reaction causes
the loss of aromaticity in the five-membered ring, owing
to the formation of an acidic methylene group. Both
derivatives present a restricted rotation around the
C(R)-B and/or B-Ar bonds, due to steric hindrance
between the fluorinated rings and the N-methyl group,
with the hydrogen atoms of the pyrrolic â-carbon playing
a minor role. Their rotational barriers (13.8 and 14.8
kcal mol-1 for 3 and 4, respectively) were calculated
2-[Tr is(p en t a flu or op h en yl)b or a n e]-5H -1-m et h ylp yr -
r ole (3). A light yellow solution of N-methylpyrrole (99%, 0.503
g, 6.1 mmol) (dichloromethane, 10 mL) was added at room
temperature to a white-gray suspension of B(C6F5)3 (99.4%,
3.18 g, 6.2 mmol) (dichloromethane, 10 mL). The resulting
orange cloudy solution was stirred at room temperature for 3
days, and then the solvent was removed under reduced
pressure. The obtained orange powder was suspended with a
1:2 dichloromethane/pentane mixture and filtered. The filtrate
gave a dark pink solid (the desired product together with
unidentified species), whereas the residue on the frit was a
very light yellow powder (product 3, 2.42 g, yield 66.5%), mp
1
from H NMR experimental data. Both adducts are able
to protonate triethylamine, recovering their aromaticity
and leading to the corresponding borate ammonium
salts. The acidic character of the methylene moiety in
3 and 4, as well as the stability of the borate anion in 5
and 6, has been advantageously used for metallocene
activation in ethylene polymerization. To explain the
outcome of this novel reaction, we propose a mechanism
1
) 116.6-118.2 °C; H NMR (CD2Cl2) δ 3.55 (s, 3 H, N-CH3),
4.70 (br s, 2 H, H5,H5′), 6.91 (br s, 1 H, H3), 7.42 (d, J ) 5.87
(10) Balboni, D.; Camurati, I.; Prini, G.; Resconi, L.; Galli, S.;
Mercandelli, P.; Sironi, A. Inorg. Chem. 2001, 40, 6588.
(11) Siedle, A. R.; Miller, J . A.; Lamanna, W. M. Int. Pat. Appl. WO
96/26967 to 3M, 1996.
(9) Chen, E.; Marks, T. J . Chem. Rev. 2000, 100, 1391.

5140 Organometallics, Vol. 23, No. 22, 2004
Focante et al.
Hz, 1 H, H4); 13C NMR (CD2Cl2) δ 38.12 (CH3), 69.81 (C5),
137.33 (C3), 146.06 (C4); NOESY (CD2Cl2) δ1H/δ1H 7.42/6.91
(H4/H3), 7.42/4.70 (H4/H5), 4.70/3.55 (H5/N-CH3); COSY (CD2-
Cl2) δ1H/δ1H 7.42/6.91 (H4/H3), 7.42/4.70 (H4/H5), 4.70/3.55
7.23-7.29 (m, 1 H, H7), 7.41-7.48 (m, 1 H, H4); 13C NMR (CD2-
Cl2) δ 8.56 (N(CH2CH3)3), 31.84 (N-CH3), 47.32 (N(CH2CH3)3),
104.75 (C3), 109.35 (C7), 118.29 and 118.35 (C4 and C5),
119.19 (C6), 128.79 (C3a), 139.47 (C7a); NOESY (CD2Cl2) δ1H/
δ1H 6.19/7.48 (H3/H4), 3.51/7.23-7.29 (N-CH3/H7); COSY δ1H/
δ1H 7.23/6.19 (H7/H3); HMBC (CD2Cl2) δ1H/δ13C 3.51/139.47
(N-CH3/C7a); 7.41-7.48/119.19 (H4/C6); 6.19/128.79 (H3/C3a).
Anal. Calcd for C33H24BF15N2: C, 53.25; H, 3.25; N, 3.76.
Found: C, 53.50; H, 3.24; N, 3.75.
1
(H5/N-CH3); H NMR analysis at variable temperature, T )
-22 °C, 1H NMR (CD2Cl2) δ 3.52 (s, 3 H, CH3), 4.71 (AB
system, 2 H, H5,H5′), 6.87 (t, J ) 4.50 Hz, 1 H, H3), 7.43 (d,
J ) 5.87 Hz, 1 H, H4); MS 592 (M - 1)-. Anal. Calcd for C23H7-
BF15N: C, 46.58; H, 1.19; N, 2.36. Found: C, 46.70; H, 1.19;
N, 2.35.
Tr ea tm en t of 1,2-Dim eth ylin d ole w ith B(C6F 5)3. A light
pink solution of 1,2-dimethylindole (99%, 0.513 g, 3.5 mmol)
(dichloromethane, 10 mL) was added at room temperature to
a white-gray suspension of B(C6F5)3 (99.4%, 1.797 g, 3.5 mmol)
(dichloromethane, 10 mL) in a 25 mL Schlenk flask. The
resulting orange suspension was stirred at room temperature
2-[Tr is(p e n t a flu or op h e n yl)b or a n e ]-3H -1-m e t h ylin -
d ole (4). A yellow solution of N-methylindole (97%, 0.78 g,
5.8 mmol) in 10 mL of dichloromethane was added at room
temperature to a white suspension of B(C6F5)3 (99.4%, 2.97 g,
5.8 mmol) in 10 mL of dichloromethane in a 25 mL Schlenk
flask. The resulting orange solution was stirred at room
temperature for 10 days and analyzed by 1H NMR at different
times. During this time the color of the solution turned from
orange to dark bordeaux; NMR analyses showed a slow
conversion of the starting N-methylindole to product 4. The
solvent was evaporated in vacuo, and the solid obtained was
suspended in a 9:1 pentane/dichloromethane mixture and
filtered. The residue on the frit was a fuchsia solid (3.27 g,
1
for 10 days and analyzed by H NMR at different times, but
no reaction occurred, neither after 8 h of stirring at 40 °C in
toluene solution. Only a small percentage of decomposition
products was present.
Tr ea tm en t of 1,2,5-Tr im eth ylp yr r ole w ith B(C6F 5)3. A
very light yellow solution of 1,2,5-trimethylpyrrole (99%, 0.246
g, 2.2 mmol) (CH2Cl2, 5 mL) was added at room temperature
to a white-gray suspension of B(C6F5)3 (99.4%, 1.165 g, 2.3
mmol) (CH2Cl2, 5 mL) in a 10 mL Schlenk flask to give
immediately an orange suspension. The reaction mixture was
stirred at room temperature for a week and analyzed by 1H
NMR at different times: many unidentified byproducts were
formed, but the main species was yet the starting trimeth-
ylpyrrole.
3H-2-Meth ylin doliu m Hydr oxytr is(pen taflu or oph en yl)-
bor a te(1-). A pink solution of 2-methylindole (0.15 g, 3.1
mmol) (pentane, 12 mL) was added at room temperature to a
white suspension of H2O-B(C6F5)3 (1.66 g, 3.1 mmol) (dichlo-
romethane, 12 mL) in a 25 mL Schlenk flask. The resulting
mixture immediately became a pink suspension, which was
stirred for 1 h and then dried in vacuo to give a light pink
solid as product (7, 1.97 g, purity 90.4%, yield 86.2%): 1H NMR
(CD2Cl2) δ 2.72 (s, 3 H, CH3), 7.43 (m, 3 H, Ar), 7.60 (m, 1 H,
Ar); N-H and H3,H3′ are not visible because of an acidic
exchange reaction.
1
yield 87.8%), mp ) 126.7-127.8 °C; H NMR (CD2Cl2) δ 3.77
(s, 3 H, N-CH3), 4.59 (br AB system, 2 H, H3,H3′), 7.45-7.53
(m, 1 H, H7), 7.54-7.62 (m, 2 H, H5 and H6), 7.63-7.71 (m,
1
1 H, H4); H NMR (C6D6) δ 2.84 (s, 3 H, N-CH3), 4.04 (br AB
system, 2 H, H3,H3′); 6.42-6.51 (m, 1 H, H7), 6.81-6.97 (m,
3 H, H4, H5, and H6); 13C NMR (CD2Cl2) δ 36.55 (CH3), 48.33
(C3), 113.59 (C7), 125.22 (C4), 128.97 (C5 or C6), 129.22 (C6
or C5), 134.18 (C3a), 147.06 (C7a), 214.34-217.43 (m, C2);
NOESY (CD2Cl2) δ1H/δ1H 3.77/6.47 (N-CH3/H7), 3.77/7.62 (N-
CH3/H6), 4.59/7.69 (H3/H4); NOESY (C6D6) δ1H/δ1H 6.81-
6.97/6.42-6.51 (H6/H7), 6.42-6.51/2.84 (H7/N-CH3); HMBC
(CD2Cl2) δ1H/δ13C 3.77/147.06 (N-CH3/C7a), 7.46-7.70/134.18
(HAr/C3a), 3.77/214.34-217.43 (N-CH3/C2); 1H NMR analysis
at variable temperature, T ) -35 °C, 1H NMR (CD2Cl2) δ 3.79
(s, 3 H, N-CH3), 4.61 (AB system, 2 H, H3,H3′), 7.48-7.69 (m,
4 H, Ar). MS 642 (M - 1)-. Anal. Calcd for C27H9BF15N: C,
50.42; H, 1.41; N, 2.18. Found: C, 50.66; H, 1.41; N, 2.17.
Tr is(p en ta flu or op h en yl)-(1-m eth ylp yr r ol-2-yl)bor a te-
(1-) Tr ieth yla m m on iu m (5). A colorless solution of triethyl-
amine (99.5%, 0.167 g, 1.6 mmol) (dichloromethane, 6 mL) was
added at room temperature to an orange solution of 3 (1.048
g, 1.6 mmol) (dichloromethane, 6 mL) in a 25 mL Schlenk
flask. The resulting light yellow solution was stirred at room
temperature for 1 h. Then the solvent was removed under
reduced pressure to give a white-gray solid as product (0.892
g, yield 97.9%): 1H NMR (CD2Cl2) δ 1.22 (t, 9 H, J ) 7.34 Hz,
N(CH2CH3)3), 3.03 (q, 6 H, J ) 7.34 Hz, N(CH2CH3)3), 3.32 (s,
3 H, N-CH3), 5.76 (br d, 1 H, J ) 2.64 Hz, H3), 5.96 (d, 1 H, J
) 3.23 Hz, H4), 6.10 (br s, 1 H, NH), 6.64 (br s, 1 H, H5); 13C
NMR (CD2Cl2) δ 8.82 (N(CH2CH3)3), 35.94 (N-CH3), 47.32
(N(CH2CH3)3), 104.40 (C4), 111.85 (C3), 122.95 (C5), 146.20
(C2); NOESY (CD2Cl2) δ1H/δ1H 6.64/3.32 (H5/N-CH3), 6.64/
5.96 (H5/H4), 5.96/5.76 (H4/H3); COSY (CD2Cl2) δ1H/δ1H 5.96/
5.76 (H4/H3); HMBC (CD2Cl2) δ1H/δ13C 3.32/146.20 (N-CH3/
C2). Anal. Calcd for C29H22BF15N2: C, 50.17; H, 3.19; N, 4.03.
Found: C, 50.28; H, 3.18; N, 4.02.
P r oton a tion of Tr is(p en ta flu or op h en yl)-(1-m eth ylin -
d ol-2-yl)bor a te(1-) An ion . A light pink solution of 3H-2-
methylindolium hydroxytris(pentafluorophenyl)borate (7) (12.8
mg, 19 µmol) (dichloromethane, 4 mL) was added at room
temperature to an orange solution of 6 (13.8 mg, 19 µmol)
(dichloromethane, 4 mL) to give a light yellow solution. 1H
NMR analyses showed an instantaneous and complete conver-
sion of reagents 7 and 6 in free 2-methylindole, 4, and
1
hydroxytris(pentafluorophenyl)borate triethylammonium. H
NMR of compound [HNEt3+]HO-B(C6F5)3-] (CD2Cl2): δ 1.16
(t, 9 H, J ) 7.24 Hz, CH3), 2.42 (br s, 1 H, N-H or O-H), 2.88
(q, 6 H, J ) 7.24 Hz, CH2), 9.76 (br s, 1 H, O-H or N-H).
Deter m in a tion of th e Kin etics for Syn th esis of 3 or 4.
The reactions of formation of compounds 3 and 4 follow the
scheme A + B f C, where A is the starting heterocycle (N-
methylpyrrole or -indole), B is B(C6F5)3, and C is complex 3 or
4. The syntheses were performed starting from solutions of
known concentration of A (where [A]0 ) [B]0). Aliquots of the
reaction mixture were taken at different times, dried, and
analyzed by 1H NMR. The only detectable species were the
heterocycle A and the final complex C. Depending on the NMR
area ratio of the methyl signals of both species and on the
starting [A]0 value, the concentration of the heterocycle A was
therefore calculated at the corresponding time. Figure 4
collects plots of [A]-1 versus time for reaction of formation of
3 and 4 at room temperature. Kinetics at higher temperature
were calculated following the same methods, but the experi-
ments were performed directly in deuterated solvent, and to
keep the temperature constant in time, the NMR tubes were
left at the temperature of analysis inside the instrument, and
the spectra were recorded approximately every half hour. The
Tr is(p en t a flu or op h en yl)-(1-m et h ylin d ol-2-yl)b or a t e-
(1-) Tr ieth yla m m on iu m (6). A colorless solution of triethyl-
amine (99.5%, 0.167 g, 1.6 mmol) (dichloromethane, 6 mL) was
added at room temperature to a bordeaux solution of 4 (1.048
g, 1.6 mmol) (dichloromethane, 4 mL) in a 10 mL Schlenk
flask. The resulting solution was stirred at room temperature
for 1 h, and its color turned from the initially orange to yellow.
Then the solvent was removed in vacuo to give a yellow solid
1
as product 6 (1.11 g, yield 93.2%), mp ) 130.6-132.8 °C; H
NMR (CD2Cl2) δ 1.03 (t, 9 H, J ) 7.24 Hz, N(CH2CH3)3), 2.60
(q, 6 H, J ) 7.24 Hz, N(CH2CH3)3), 3.51 (s, 3 H, N-CH3), 4.40
(br s, 1 H, NH), 6.19 (s, 1 H, H3), 6.95-7.13 (m, 2 H, H5,H6),

N-Heterocycle-B(C6F5)3 Complexes
Organometallics, Vol. 23, No. 22, 2004 5141
resulting kinetic constants calculated from each experiment
were then plotted as ln(k/T) versus 1/T in Figure 5.
by degassing the reactor and by adding 2 mL of methanol. The
polymer was precipitated with 200 mL of acetone, filtered,
washed with acetone, and dried overnight at 60 °C under
reduced pressure.
E t h ylen e P olym er iza t ion . Bis(indenyl)zirconium di-
methyl (0.1 µmol) was dissolved in 2 mL of toluene in a 10
mL Schlenk flask under a nitrogen atmosphere, and then the
cocatalyst (solution in 2 mL of toluene) indicated in Table 1
was quickly added (Zr/cocat molar ratio 1:1.1). Ethylene
polymerizations were carried out in a 260 mL Bu¨chi glass
autoclave equipped with magnetic stirrer, thermocouple, and
feeding line for the monomer, purified with nitrogen, and kept
in a thermostatic bath. Under ethylene purge, heptane (100
mL) and Al(i-Bu)3 (0.1 mmol) were introduced, the temperature
was brought to 80 °C, and the reactor was vented to remove
residual nitrogen, then pressurized with ethylene up to 0.5
bar-g. The catalytic system, prepared as described above, was
siphoned into the reactor by means of a Teflon cannula, and
the ethylene partial pressure was raised to 4 bar-g. The
polymerization was carried out at 80 °C for 15 min, by
maintaining a constant ethylene partial pressure, then stopped
Ack n ow led gm en t. We thank Dr. Pierluigi Mercan-
delli and Prof. Angelo Sironi (University of Milano) for
the X-ray structure determination of 6, Marcello Col-
1
onnesi for the H NMR spectra, and Simona Guidotti
for discussions. We also thank Prof. F. Gasparrini,
Universita` La Sapienza, Roma, for the PC version of
the DNMR-6 program.
Su p p or tin g In for m a tion Ava ila ble: 2D NMR charac-
terization of 3-6. This material is available free of charge via
the Internet at http://pubs.acs.org.
OM0499308