Investigations on 1-metallaindanes

Article abstract of DOI:10.1002/ejic.200390088

1-Mercuraindane (6) was prepared in three steps from 1-bromo-2-(2-bromoethyl) benzene (3) and converted into 1-magnesaindane (1) by reaction with magnesium. In solution, 6 was observed to occur in a remarkable dimer/trimer equilibrium, whereas 1 forms a dimer. From 1 and dichlorozirconocene, the 1-zirconaindane 2 was obtained, which reacted with B(C₆F₅)₃ to form the adduct 13. According to spectroscopic data, 13 is formed by attack of the p-methylene carbon atom at the boron atom and has a zwitterionic structure with weak interactions between the zirconium ion and one of the ortho-fluorine atoms. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200390088

FULL PAPER
Investigations on 1-Metallaindanes
Marcel Schreuder Goedheijt,[a] Tom Nijbacker,[b] Andrew D. Horton,[c]
Franciscus J. J. de Kanter,[b] Otto S. Akkerman,[b] and Friedrich Bickelhaupt*[b]
Keywords: Metallacycles / Isomerization / Magnesium / Mercury / Zirconium
1-Mercuraindane (6) was prepared in three steps from 1-
bromo-2-(2-bromoethyl)benzene (3) and converted into 1-
magnesaindane (1) by reaction with magnesium. In solution,
6 was observed to occur in a remarkable dimer/trimer equi-
librium, whereas 1 forms a dimer. From 1 and dichlorozircon-
ocene, the 1-zirconaindane 2 was obtained, which reacted
with B(C6F5)3 to form the adduct 13. According to spectro-
scopic data, 13 is formed by attack of the β-methylene carbon
atom at the boron atom and has a zwitterionic structure with
weak interactions between the zirconium ion and one of the
ortho-fluorine atoms.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
Cyclic diorganylmagnesium compounds[1] often show in-
teresting structural features.[2,3] One aspect is the relatively
large CϪMgϪC bond angle, which for normal diorganyl-
magnesium compounds of the general type R2MgL2 (L ϭ
ligand, mostly an ether) is around 120Ϫ130°. This leads to
considerable strain if the two R groups are connected such
that magnesium is incorporated into a small ring. As the
carbonϪmagnesium bonds undergo a rapid exchange in
ethereal solution, the strained monomers can easily escape
from this situation and thus tend to form more stable
dimers or higher aggregates in which the large bond angle
Scheme 1
of interest as a precursor of a new class of single-site olefin
polymerization catalysts.[5Ϫ7]
can be accommodated free of strain; sometimes an equilib- Results and Discussion
rium between two oligomers is established. For the sake of
simplicity, the oligomers are indicated by the names of the
monomeric parent units where appropriate; furthermore,
Synthesis of Magnesaindane (1)
magnesium is assumed to be tetracoordinate including two
In order to obtain 1 in pure form, we decided to prepare
solvent molecules,[1Ϫ3] which in the present case is THF. We it by transmetalation of the corresponding cyclic organo-
have previously reported the investigation of a number of mercury compound 6 with metallic magnesium, as shown
such cyclic diorganylmagnesium compounds,[3,4] and in Scheme 2. For this purpose, the dibromide 3 was con-
wanted to extend this study to 1-magnesaindane (1) verted into 4, which is the known[8] di-Grignard reagent
(Scheme 1).
corresponding to the diorganylmagnesium 1. In an attempt
A further reason for our interest in 1 was that it is a to obtain 6 from 4 directly, we treated 4 with 1 mol-equiv.
difunctional organomagnesium compound and as such is of mercury dibromide but, according to the 199Hg NMR
attractive for the preparation of other metallacycles such as spectrum, a mixture of organomercury compounds was
the corresponding 1-zirconaindane 2, which is potentially formed which could not be separated. However, 6 was ob-
tained by reduction[9] of the corresponding organomercury
[a]
Diosynth bv,
halide 5 with stannous chloride in the presence of base. Re-
action of 6 with metallic magnesium in THF furnished 1,
Postbus 20, 5340 BH Oss, The Netherlands
Scheikundig Laboratorium, Vrije Universiteit,
De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
Fax: (internat.) ϩ 31-20/444-7488
[b]
which was characterized spectroscopically by its 1H and 13
C
NMR spectrum (see Exp. Sect.) and chemically by reaction
with deuterium oxide which gave [D2]ethylbenzene (7) ex-
clusively.
E-mail: bicklhpt@chem.vu.nl
Shell Research and Technology Centre,
Badhuisweg 3, 1031 CM Amsterdam, The Netherlands
[c]
638
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ1948/03/0204Ϫ0638 $ 20.00ϩ.50/0 Eur. J. Inorg. Chem. 2003, 638Ϫ643

Investigations on 1-Metallaindanes
FULL PAPER
A dimer/trimer equilibrium is plausible since, in hypo-
thetical monomeric 6, the CϪHgϪC angle would deviate
too much from the ideal value of 180°, making the ensuing
ring strain prohibitive. This is less so for a dimeric species
6a, while a trimeric species 6b is strain-free as concluded
from CPK model studies. Species higher than a trimer are
not formed since oligomerization is entropically unfavor-
able.
The establishment of such an equilibrium is unexpected
because, in contrast to the kinetically highly labile
carbonϪmagnesium bond, the carbonϪmercury bond is
usually stable under ordinary conditions. Redistribution re-
actions of bifunctional organomercury compounds have
Scheme 2
been reported, but only in the presence of a catalyst such
as AlCl3
Structure of 1-Mercuraindane (6)
[10]
or an organomercury halide (RHgX).[10,11] For
Although initially 6 was only of interest as an interme-
diate for the preparation of 1, it showed spectral features
which led to a closer investigation. The 13C NMR spectrum
of pure 6 in CDCl3 shows evidence for the occurrence of
two species 6a and 6b, presumably differing in their degree
of oligomerization. For instance, four methylene carbon res-
onances were observed (δ ϭ 41.5 and 39.7 ppm for 6a; δ ϭ
41.7 and 38.4 ppm for 6b) as well as eight resonances of
aromatic carbon atoms and two of 199Hg (δ ϭ Ϫ273.2 ppm
for 6a; δ ϭ Ϫ305.5 ppm for 6b). The signal integral ratio of
the benzylic methylene protons of 6a and 6b in the 1H
NMR spectrum was found to be concentration-dependent,
which suggests an equilibrium between them in solution
example, an equimolar mixture of Me2Hg and Et2Hg gives
a random distribution of the three possible alkylmercury
compounds, Me2Hg, Et2Hg and MeEtHg, upon standing
at room temperature in the presence of AlCl3, but after re-
moval of the catalyst, MeEtHg can be distilled at 127 °C
and is stable for months at room temperature.[10] The pres-
ence of an organomercury halide, such as 5, as an impurity,
which might act as a catalyst, was excluded by a negative
halide test performed by evaporation of the solvent from a
sample followed by treatment with HNO3 and AgNO3; a
precipitate of AgCl was not observed.
To confirm the occurrence of an equilibrium for 6 and to
estimate the exchange rates between both species, the mag-
netization-transfer NMR technique[12] was applied between
25 and 95 °C. Below 83 °C, no transfer was observed, which
indicates that the establishment of the equilibrium is rela-
tively slow at these temperatures. At 83 °C and 95 °C, the
1
(Scheme 3). From H NMR spectroscopic data (Table 1),
the equilibrium constant at room temperature was deter-
mined. As, according to HRMS spectroscopy, 6 has a di-
meric structure (m/z ϭ 610.0642) in the gas phase, one of
the species involved in solution is presumably a dimer. As-
suming a dimer/trimer equilibrium, the equilibrium con-
stant was found to be in the range of K ϭ 0.4 Ϯ 0.2
(pseudo-first-order) rates obtained for the equilibrium 3 6a
Ǟ
ǟ
2 6b were in the range of 0.1Ϫ1 sϪ1. Thus, the exchange
between the two oligomers is rather fast, and even at room
temperature, one can estimate that it will be sufficient to
allow establishment of the equilibrium. However, it is a re-
markable process, and it remains unclear whether it is
caused by inherent structural features or by some kind of
undetected catalyst present in the system. Compound 6a is
probably dimeric (vide supra) and CPK models show that,
of the two possible structures, 6aЈ is less strained than 6aЈЈ
due to the presence of the arylϪHgϪaryl arrangement in
the latter (Scheme 4).
L·molϪ1
.
Scheme 3
Table 1. Data on the equilibrium of 6 at room temperature [in
0.5 mL of CDCl3 at room temperature; determined from the ratio
of the benzylic protons at δ ϭ 3.36 ppm (6a) and 2.95 ppm (6b)]
6[a]
[6a][b]
[6b][b]
K[c]
1.5
4.1
9.1
13.5
17.8
21.5
25.4
31.0
8.8·10Ϫ3
2.4·10Ϫ2
5.2·10Ϫ2
7.6·10Ϫ2
1.0·10Ϫ1
1.2·10Ϫ1
1.4·10Ϫ1
1.7·10Ϫ1
5.8·10Ϫ4
2.8·10Ϫ3
7.6·10Ϫ3
1.2·10Ϫ2
1.7·10Ϫ2
2.1·10Ϫ2
2.7·10Ϫ2
3.3·10Ϫ2
0.49
0.57
0.41
0.33
0.30
0.26
0.24
0.22
Scheme 4
Structure of 1 in Solution
The appreciable solubility of 1 in THF (ca. 0.1  at room
temperature) allowed investigations concerning its structure
in solution. The degree of association was determined by
the stationary isothermal distillation method,[13] which fur-
[a]
[b]
[c]
In mg. In mol·LϪ1 (6a as dimer, 6b as trimer). In L·molϪ1
.
Eur. J. Inorg. Chem. 2003, 638Ϫ643
639

F. Bickelhaupt et al.
FULL PAPER
nished a concentration-independent degree of association nately, data on the degree of oligomerization of 10 in solu-
of 2.08 (see Table 2).
tion are not available.
In analogy to other magnesacycles,[1Ϫ3] one may safely
assume that the dimeric 1 has a ten-membered ring struc-
ture. As both the magnesium atom and the benzene ring
prefer large bond angles (Ն 120°), the structure of 1Ј as
shown in Scheme 5 (which is analogous to 6aЈ, Scheme 4),
Table 2. Degree of association of 1 in THF solution at 301 K[13]
[Mg][a]
Sth
Sa
i[d]
[b]
[c]
1.24
2.42
3.53
4.57
5.55
6.49
0.052
0.101
0.148
0.192
0.233
0.272
0.028
0.045
0.068
0.092
0.117
0.126
1.84
2.26 is suggested to be energetically more favorable than the al-
2.15
2.08
1.99
2.16
ternative arrangement 1ЈЈ (analogous to 6a") in which a
C(Ar)ϪMgϪC(Ar) unit is present which probably would
cause a higher ring strain in the molecule, as discussed for
6a.
[a]
[b]
[Mg] ϭ Formal concentration of magnesium in mmol·gϪ1
.
Sth ϭ Theoretical rate of evaporation (in mm·hϪ1) ϭ [Mg]·SS,
1,1-Dicyclopentadienyl-1-zirconaindane (2)
where SS (in mm·hϪ1·mmolϪ1·L) ϭ standard rate of evaporation
Having 1 available, it appeared to be an attractive starting
material for the preparation of the 1-zirconaindane 2 which
has previously been obtained by thermolysis of diphenylzir-
conocene in the presence of ethene.[17] It has been shown
that the reaction of 11, the 2-zircona analogue of 2, with
B(C6F5)3 furnished the intramolecular ion pair 12
(Scheme 6).[5] It was therefore of interest to investigate the
reaction of 2 with B(C6F5)3 in order to generate the zwitter-
ionic complex 13. Study of the effect of anionϪcation
bonding interactions in 13 on the reactivity of unsaturated
molecules was expected to improve our understanding of
the factors influencing the activity of cationic metallocene
catalysts in olefin polymerization.[18Ϫ22]
found by calibration of the apparatus with triphenylmethane, SS
(301 K) ϭ 0.0119 mm·hϪ1·mmolϪ1·L.
Sa ϭ Apparent rate of
Association number i ϭ Sth/Sa.
[c]
evaporation in mm·hϪ1
.
[d]
The structure of two benzomagnesacycloalkanes in solu-
tion has been investigated by Freijee et al., who determined
the degree of association in solution of the 1-magnesatetra-
lin 8 and the tetrahydrobenzomagnesepin 9 (Scheme 5).[4] It
was established that 8 was involved in a monomer/dimer
equilibrium as its concentration-dependent degree of asso-
ciation was 1.26Ϫ1.49 in THF solution at room temper-
ature, whereas the higher homologue 9 was completely
monomeric. It was assumed that the higher ring strain in
monomeric 8 (about 46 kJ·molϪ1) than in 9 (about 11
kJ·molϪ1) is largely responsible for the observed difference.
Scheme 5
The fact that 1 is completely dimeric in THF solution at
room temperature is in line with the results obtained from
8 and 9 since ring strain in the hypothetical monomer of 1,
with magnesium in a five-membered ring, will be even
higher than in monomeric 8. As the trimer of 1 is not ob-
served, we assume that entropy protects the system from
further oligomerization as is also the case for other cyclic
organomagnesium compounds.[3,4,14,15] Note, however, that
the results presented so far pertain to the situation in solu-
tion where an equilibrium can be established; in the solid
Scheme 6
Compound 2 was treated with 1 equiv. of B(C6F5)3 in
state, the degree of oligomerization may be different. Thus, bromobenzene at Ϫ30 °C, and the product was isolated as
Lappert et al. reported the crystal structure of 2-magnesain- a pale yellow powder in 81% yield by precipitation with n-
dane (10), the regioisomer of 1, which turns out to be a pentane. Elemental analysis confirmed the composition of
trimer with a central fifteen-membered ring.[16] Unfortu- 13. The 1H NMR spectrum in C2D2Cl4 shows a broad trip-
640
Eur. J. Inorg. Chem. 2003, 638Ϫ643

Investigations on 1-Metallaindanes
FULL PAPER
let signal at δ ϭ 2.30 ppm which was assigned to the in 12) hinders coordination of the substrate. Nevertheless,
methylene protons next to the borate group; in the 13C the fluorineϪzirconium interaction in 13 is not strong
NMR spectrum, a signal at δ ϭ 178.2 ppm indicates that enough to resist cleavage by PMe3 in CD2Cl4 with forma-
the ZrϪC(Ar) bond is still intact. Furthermore, the 19F tion of 15. The yellow color of 13 disappeared, and 15 was
NMR spectrum (C2D2Cl4, Ϫ25 °C) shows three resonances isolated as an off-white solid in 96% yield by precipitation
at δ ϭ Ϫ134.8 (o-F), Ϫ160.1 (p-F), and Ϫ165.0 ppm (m-F). with n-pentane. The structure of 15 was assigned on the
1
These data are in line with the presence of a tetracoordin- basis of the H and 19F NMR spectra. In particular, the
ate, negatively charged boron substituent[5] [RB(C6F5)3]Ϫ in small value of ∆δ(m,p-F) ϭ 3.0 ppm is consistent with the
13. The presence of an anionϪcation bonding interaction donor having displaced the anionϪcation bonding interac-
is indicated by the large value of ∆δ(m,p-F) ϭ 4.9 tion;[23] an aromatic proton signal at δ ϭ 4.48 ppm strongly
(Scheme 6).[23]
suggests an agostic Zr···H interaction.[26]
Solutions of 13 in tetrachloroethane are stable at room
temperature for several hours. We ascribe this stabilization
to intramolecular coordination of an ortho-fluoro substitu-
ent of one of the C6F5 groups to zirconium, as shown. This
assumption was confirmed by variable-temperature 19F
NMR experiments in C6D5Br/[D8]toluene. At Ϫ60 °C, the
19F NMR spectrum shows eight resonances (Table 3) as ex-
pected when one of the C6F5 groups is coordinating to the
zirconium atom and hence becomes rigid and unsymmet-
Conclusion
1-Magnesaindane (1) was obtained via the corresponding
1-mercuraindane (6) and found to occur as a dimeric ten-
membered ring species in THF solution; in line with the
behavior of other magnesacycles, this is due to an interplay
between ring strain and entropy. In contrast, intermediate
6 forms a dimer/trimer equilibrium, which is remarkable as,
at room temperature and in the absence of a catalyst, dior-
ganomercury compounds usually do not establish equilibria
by exchanging ligands.
The 1-zirconaindane 2 was obtained from 1 and dichloro-
zirconocene. It reacts with B(C6F5)3 to form the adduct 13
which, according to spectroscopic data, has a zwitterionic
structure with transannular interactions between the zirco-
nium atom and one of the ortho-fluorine atoms. Although
this interaction may contribute to the inertness towards in-
sertion of the zirconiumϪaryl bond in 13, it is kinetically
unstable at room temperature and is easily cleaved by the
Lewis base trimethylphoshane, which coordinates to the zir-
conium atom with formation of 15.
rical. One fluorine resonance is shielded to
δ ϭ
Ϫ180.2 ppm, more than 50 ppm from the region typical for
an ortho-fluorine nucleus in an [RB(C6F5)3]Ϫ anion. A sim-
ilar shielding (to δ ϭ Ϫ190.3 ppm at Ϫ88 °C) was observed
for one of the ortho-fluorine nuclei in the zirconoxyborane
Cp*2ZrOB(C6F5)3 in which the CϪF···Zr interaction was
proven by X-ray analysis;[24] more recently, a number of
zwitterionic complexes with similar Zr···F bonding interac-
tions have been spectroscopically and structurally charac-
terized.[6,7] The 19F resonances of 13 broaden above Ϫ50
°C, and at room temperature, only the initial three 19F res-
onances are discernible. Note that, presumably for steric
reasons, such a Zr···F interaction does not occur in 12; in-
stead, the X-ray structure reveals an interaction between
the zirconium atom and the methylene group carrying the
borate substituent.[5]
Experimental Section
General: All manipulations involving organometallic compounds
were carried out in fully sealed glassware using standard high va-
cuum techniques. Solvents were dried by distillation from liquid
Na/K alloy after predrying with NaOH. Magnesium was sublimed
twice before use. NMR spectra were measured with a Bruker AC
200 spectrometer (1H NMR: 200 MHz. 13C NMR: 50.3 MHz), a
Varian VXR-300 spectrpmeter (1H NMR: 300 MHz. 13C NMR:
75.4 MHz, 31P NMR: 121.6 MHz. 19F NMR: 283.23 MHz), or a
Bruker MSL 400 spectrometer (1H NMR: 400.1 MHz. 199Hg
NMR: 71.6 MHz). GCMS analyses were performed with an HP
5890 GC/5970 MS combination, operating at 70 eV and equipped
with a Chrompack BP1 (QSGE) 50 m/0.25 mm column. HRMS
measurements were performed with a Finnigan MAT 90 mass spec-
trometer under EI conditions using a direct introducing interface
for the introduction of air- and moisture-sensitive compounds. Ele-
mental analyses were carried out at the Microanalytical Laborat-
ory, Rijksuniversiteit Groningen, The Netherlands.
Table 3. 19F NMR chemical shifts of 12 {in C6D5Br/[D8]toluene at
Ϫ60 °C (283.23 MHz; ref. CF3C6H5 in C6D6: δ ϭ Ϫ64 ppm)}
F(2)[a]
F(3)[b]
F(4)[b]
F(5)[b]
F(6)[b]
Ring A
Ring B, C
Ϫ180.2
Ϫ131.5
Ϫ155.6
Ϫ162.6
Ϫ156.3
Ϫ157.5
Ϫ161.8
Ϫ162.6
Ϫ123.3
Ϫ131.5
[a]
Fluorine positions are numbered clockwise, starting from the
boron-substituted position.
The reactivity of 13 contrasts sharply with that of the
related complex 12. Whilst 12 inserts 2-butyne to furnish
14 (Scheme 6) and shows moderate catalytic activity in the
polymerization of ethylene (about 20Ϫ40 times less active
than simple cationic Group-4 catalysts[18,19]), 13 is inactive
in both types of reactions. This may be partly due to the
low intrinsic reactivity of the metalϪaryl bond towards in-
sertion as compared to metalϪalkyl bonds.[25] Additionally,
1-(Bromomercurio)-2-[2-(bromomercurio)ethyl]benzene (5): In
a
Schlenk vessel, 1-bromo-2-(2-bromoethyl)benzene (3)[8] (3.17 g,
12 mmol) in 20 mL of THF was slowly added during 5 h to a
stirred suspension of magnesium (1.54 g, 64 mmol) in 60 mL of
the fluorineϪzirconium interaction in 13 (which is absent THF at room temperature. After the addition was complete, an
Eur. J. Inorg. Chem. 2003, 638Ϫ643 641

F. Bickelhaupt et al.
FULL PAPER
aliquot was hydrolyzed. Titration (HCl, EDTA)[27] showed an
Reaction of 1 with Deuterium Oxide: An excess of D2O was added
at room temperature to a solution of 1 (about 0.1 mmol) in THF.
Mg2ϩ/OHϪ ratio of 1.1:1 and 78% yield of 1-(bromomagnesio)-2-
[2-(bromomagnesio)ethyl]benzene. Subsequently, mercury bromide Diethyl ether and water were added and the organic layer was dried
(8.70 g, 24.2 mmol) in 30 mL of THF was added and the solution
was stirred overnight, after which a grey suspension was obtained.
After hydrolysis, the resulting suspension was filtered and the res-
idue was washed twice with THF and H2O and subsequently dried
in vacuo to give 5 as a white powder. Yield: 6.0 g (75% based on 3).
1H NMR (200 MHz, [D6]DMSO, ref. [D5]DMSO: δ ϭ 2.50 ppm;
NOESY): δ ϭ 7.47 (m, 1 H, 4-H), 7.30 (m, 1 H, 7-H), 7.21 (m, 2
(MgSO4). By GCMS, only [D2]ethylbenzene (7) was observed.
GCMS: m/z (%) ϭ 108 (100) [M]ϩ, 107 (18), 106 (6), 105 (7), 104
(6), 103 (2).
Formation of Zwitterion 13: The known complex 2[17] was obtained
from the reaction of 1 with Cp2ZrCl2. To a mixture of 2 (53.4 mg,
0.16 mmol) and B(C6F5)3 (84.0 mg, 0.16 mmol), 1 mL of C6H5Br
was added at Ϫ30 °C. Addition of 3 mL of n-pentane gave a light
yellow microcrystalline solid. The supernatant was pipetted off and
the precipitate was washed twice with n-pentane (3 mL) and dried
at 25 °C in the glovebox (1 h) to yield 13 (108.9 mg, 81%) as a pale
yellow solid. 1H NMR (300 MHz, C2D2Cl4, ref. C2DHCl4: δ ϭ
5.91 ppm; Ϫ25 °C): δ ϭ 6.83 (m, 3 H, ArH), 6.40 (s, 10 H, Cp),
3
3
H, 5,6-H), 3.00 (t, JH,H ϭ 7.5 Hz, 2 H, ArCH2), 2.06 (t, JH,H
ϭ
7.5 Hz, 2 H, CH2Hg) ppm. 13C NMR (50.3 MHz, [D6]DMSO, ref.
[D6]DMSO: δ ϭ 39.5 ppm): δ ϭ 154.0 (br. s, C-1), 149.9 (s, C-2),
1
3
1
136.4 (dd, JC,H ϭ 161.9, JC,H ϭ 5.6 Hz, C-6), 128.1 (dd, JC,H
ϭ
159.6, 3JC,H ϭ 7.8 Hz, C-4 or -5), 127.8 (dd, 1JC,H ϭ 156.0, 3JC,H ϭ
5.8 Hz, C-3), 125.5 (dd, JC,H ϭ 159.6, JC,H ϭ 7.0 Hz, C-4 or -5),
38.8 (t, JC,H ϭ 126.1 Hz, ArCH2), 35.2 (bt, JC,H ϭ 139.8 Hz,
CH2Hg) ppm. 199Hg NMR (71.6 MHz, DMSO, ref. Me2Hg ϭ 0):
δ ϭ Ϫ1107, Ϫ1231 ppm.
1
3
3
3
6.32 (d, JH,H ϭ 5.6 Hz, 1 H, ArH), 2.70 (bt, JH,H ϭ 7.5 Hz, 2
H, ArCH2), 2.30 (bm, 2 H, CH2B) ppm. 13C NMR (75.43 MHz,
C2D2Cl4, ref. C2D2Cl4: δ ϭ 74.35; Ϫ25 °C, cation signals only;
those of the anion part were too broad for identification): δ ϭ
178.2 (CZr), 152.2, 138.7, 129.4, 125.3, 124.8 (d, 1JC,H ϭ 157.9 Hz),
116.5 (d, 1JC,H ϭ 170.6 Hz, Cp), 37.8 (t, 1JC,H ϭ 126.5 Hz, ArCH2),
14.6 (CH2B) ppm. 19F NMR (283.23 MHz, C2D2Cl4, ref. CF3C6H5
in C6D6: δ ϭ Ϫ64; Ϫ25 °C): δ ϭ Ϫ134.8 (o-F), Ϫ160.1 (p-F),
Ϫ165.0 (m-F) ppm. C36H18F15Zr (826.73): calcd. C 51.62, H 2.17;
found C 52.02, H 2.37.
1
1
1-Mercuraindane (6): A solution of SnCl2·2H2O (5.42 g, 24 mmol)
in 120 mL of 20% aqueous NaOH was slowly added over 5 h to a
suspension of 5 (8.0 g, 12 mmol) in a mixture of 120 mL of 10%
aqueous NaOH and 24 mL of CHCl3 at room temperature.[9] After
the addition was complete, a grey precipitate (Hg) was filtered off
and the residue was extracted twice with CHCl3. The combined
organic phases were dried with MgSO4 and the solvents evaporated
to give 6 as a light-grey solid (3.46 g, 95%). 6: M.p. 71 °C. HRMS:
(12C16H16200Hg202Hg): calcd. 610.0642; found 610.0642. C8H8Hg
(304.74): calcd.C 31.53, H 2.65; found C 31.91, H 2.69. 6a: 1H
NMR (200 MHz, CDCl3, ref. CHCl3: δ ϭ 7.22 ppm): δ ϭ
7.30Ϫ7.16 (m, 4 H, ArH), 3.36 (m, 2 H, ArCH2), 1.73 (m, 2 H,
Reaction of 13 with PMe3. Formation of Adduct 15: PMe3 was ad-
ded (0.046 mmol, 4.8 µL) to a solution of 13 (38.5 mg, 0.046 mmol)
in 0.5 mL of C2D2Cl4 and an immediate color change from yellow
to almost colorless was observed. After addition of 1 mL of n-pent-
ane, a light yellow precipitate was obtained, which was filtered off
and washed twice with n-pentane (2 mL). Evaporation of the n-
pentane solution in the glovebox gave 15 as an off-white solid
CH2Hg) ppm. 13C NMR (50.3 MHz, CDCl3, ref. CDCl3: δ ϭ
1
1
(40.2 mg, 96%). H NMR (300 MHz, C2D2Cl4, ref. C2DHCl4: δ ϭ
76.8 ppm): δ ϭ 180.1, C-7a), 153.2 (s, C-3a), 136.7 (dd, JC,H
ϭ
3
3
1
3
5.91; Ϫ25 °C): δ ϭ 7.09 (d, JH,H ϭ 7.1 Hz, 1 H, ArH), 6.92 (m, 2
158, JC,H ϭ 7 Hz, C-3), 130.4 (dd, JC,H ϭ 156, JC,H ϭ 6 Hz, C-
H, ArH), 5.53 (s, 10 H, Cp), 4.48 (dd, 3JH,H ϭ 10.4, 4JH,H ϭ 7.0 Hz,
1
3
1
6), 127.7 (dd, JC,H ϭ 159, JC,H ϭ 7 Hz, C-5), 125.5 (dd, JC,H
ϭ
2
3
1
1 H, ArH···Zr), 2.06 (m, 2 H, ArCH2), 1.35 (d, JP,H ϭ 7.4 Hz, 9
159, JC,H ϭ 7 Hz, C-4), 41.5 (t, JC,H ϭ 131 Hz, C-3), 39.7 (t,
1JC,H ϭ 125.5 Hz, C-2) ppm. 199Hg NMR (71.6 MHz, CDCl3, ref.
Me2Hg: δ ϭ 0): δ ϭ Ϫ237.2 ppm. 6b: Due to the lower concentra-
tion of 6b and to partial overlap of its NMR signals with those of
6a, not all signals could be identified. 1H NMR (200 MHz, CDCl3,
ref. CHCl3: δ ϭ 7.22 ppm): δ ϭ 7.30Ϫ7.16 (undiscernible from
those of 6a, ArH), 2.95 (m, ArCH2), 1.7 (shoulder on signal of
6a, CH2Hg) ppm. 13C NMR (50.3 MHz, CDCl3, ref. CDCl3: δ ϭ
76.8 ppm): δ ϭ 178.8, C-7a), 154.5 (C-3a), 136.9 (C-3), 129.2 (C-6),
128.4 (C-5), 125.8 (C-4), 41.7 (C-3), 38.7 (C-2) ppm. 199Hg NMR
(71.6 MHz, CDCl3, ref. Me2Hg: δ ϭ 0): δ ϭ Ϫ305.5. ppm.
H, PMe3), 1.06 (m, 2 H, CH2B) ppm. 19F NMR (283.23 MHz,
C2D2Cl4, ref. CF3C6H5 in C6D6: δ ϭ Ϫ64 ppm; Ϫ25 °C): δ ϭ
Ϫ131.9 (o-F), Ϫ163.0 (p-F), Ϫ166.0 (m-F) ppm. 31P{H} NMR
(121.6 MHz, C2D2Cl4, ref. PPh3: δ ϭ Ϫ6.0 ppm): δ ϭ Ϫ15.8 ppm.
Acknowledgments
The investigations were supported by Shell International Chemicals
B. V.
[1]
F. Bickelhaupt in Grignard Reagents, New Developments (Ed.:
H. G. Richey, Jr.), Wiley, Chichester, 2000, chapter 11.
1-Magnesaindane (1): Compound 6 (1.78 g, 5.9 mmol) and magnes-
ium (2.5 g, 103 mmol) in 70 mL of THF were stirred for 4 d at 70
°C. A brown solution and a fine black precipitate were obtained.
The excess magnesium and the black precipitate were removed by
careful decantation. Crystallization by cooling a saturated THF
solution gave 1 as colorless needles (4.1 mmol, 70% based on 6;
determined by titration[27]). Attempts to obtain crystals suitable for
an X-ray crystal structure determination were not successful. 1H
NMR (400 MHz, [D8]THF, ref. [D7]THF: δ ϭ 1.75 ppm): δ ϭ 7.71
[2]
H. L. Uhm in Handbook of Grignard Reagents (Ed.: G. S. Sil-
verman, P. E. Rakita), Marcel Dekker, New York, 1996, chap-
ter 9.
F. Bickelhaupt in Grignard Reagents, New Developments (Ed.:
H. G. Richey, Jr.), Wiley, Chichester, 2000, chapter 9.
F. M. J. Freijee, G. Schat, O. S. Akkerman, F. Bickelhaupt, J.
Organomet. Chem. 1982, 240, 217Ϫ227, and references cited
therein.
[3]
[4]
[5]
J. Karl, G. Erker, R. Fröhlich, F. Zippel, F. Bickelhaupt, M.
Schreuder Goedheijt, O. S. Akkerman, P. Binger, J. Stannek,
Angew. Chem. 1997, 109, 2914Ϫ2917; Angew. Chem. Int. Ed.
Eng. l 1997, 36, 2771Ϫ2774.
M. Dahlmann, G. Erker, M. Nissinen, R. Fröhlich, J. Am.
Chem. Soc. 1999, 121, 2820Ϫ2828.
N. Kleigrewe, T. Brackemeyer, G. Kehr, R. Fröhlich, G. Erker,
3
3
(d, JH,H ϭ 6.4 Hz, 1 H), 6.73Ϫ6.86 (m, 3 H), 3.06 (t, JH,H
ϭ
3
7.2 Hz, 2 H, ArCH2), 0.02 (t, JH,H ϭ 7.2 Hz, 2 H, CH2Mg) ppm.
13C NMR (50.3 MHz, ref. [D8]THF ϭ 67.4): δ ϭ 168.7 (bs), 167.2
(bs), 142.2 (d, 1JC,H ϭ 149.3 Hz), 125.4 (d, 1JC,H ϭ 154.5 Hz), 124.9
(d, 1JC,H ϭ 150.3 Hz), 122.2 (d, 1JC,H ϭ 153.6 Hz), 43.3 (t, 1JC,H ϭ
[6]
[7]
1
119.3 Hz, C-3), 17.6 (t, JC,H ϭ 108.6 Hz, C-2) ppm.
Organometallics 2001, 20, 1952Ϫ1955.
642
Eur. J. Inorg. Chem. 2003, 638Ϫ643

Investigations on 1-Metallaindanes
FULL PAPER
X. Yang, C. L. Stern, T. J. Marks, J. Am. Chem. Soc. 1994,
116, 10015Ϫ10032.
L. Lia, X. Yang, A. Ishihara, T. J. Marks, Organometallics
1995, 14, 3135Ϫ3137.
B. Temme, G. Erker, J. Karl, H. Luftmann, R. Fröhlich, S.
Kotila, Angew. Chem. 1995, 107, 1867Ϫ1869; Angew. Chem.
Int. Ed. Engl. 1995, 34, 1755Ϫ1757.
A. D. Horton, J. de With, A. J. van der Linden, H. van de Weg,
Organometallics 1996, 15, 2672Ϫ2674.
A. R. Siedle, R. A. Newmark, W. M. Lamanna, Organometal-
lics 1993, 24, 1491Ϫ1492.
S. L. Borkowsky, R. F. Jordan, G. D. Hinch, Organometallics
[8]
[20]
[21]
[22]
H. Gilman, O. L. Marrs, J. Org. Chem. 1965, 30, 325Ϫ328.
[9]
H. Sawatzky, G. F. Wright, Can. J. Chem. 1958, 36, 1555Ϫ1569.
[10]
G. Calingaert, H. Soroos, V. Hnizda, J. Am. Chem. Soc. 1940,
62, 1107Ϫ1110.
[11]
H. B. Charman, E. D. Hughes, C. Ingold, J. Chem. Soc.
1959, 2523Ϫ2530.
[12]
J. J. Led, H. Gesmar, J. Magn. Reson. 1982, 49, 444Ϫ463.
[13]
A. Van Vulpen, J. Coops, Recl. Trav. Chim. Pays-Bas 1966,
[23]
[24]
[25]
[26]
[27]
85, 203Ϫ215.
[14]
H. C. Holtkamp, C. Blomberg, F. Bickelhaupt, J. Organomet.
Chem. 1969, 19, 279Ϫ285.
[15]
F. Bickelhaupt, Angew. Chem. 1987, 99, 1020Ϫ1035; Angew.
Chem. Int. Ed. Engl. 1987, 26, 990.
M. F. Lappert, T. R. Martin, C. L. Raston, B. W. Skelton, A.
1991, 10, 1268Ϫ1274.
[16]
G. G. Hlatky, H. W. Turner, R. R. Eckman, J. Am. Chem. Soc.
1989, 111, 2728Ϫ2729.
A. D. Vreugdenhil, C. Blomberg, Recl. Trav. Chim. Pays Bas
H. White, J. Chem. Soc., Dalton Trans. 1982, 1959Ϫ1964.
[17]
G. Erker, K. Kropp, J. Am. Chem. Soc. 1979, 101, 3659Ϫ3660.
[18]
M. Bochmann, Angew. Chem. 1992, 104, 1206Ϫ1207; Angew.
1963, 83, 453Ϫ460.
Chem. Int. Ed. Engl. 1992, 31, 1181Ϫ1182.
[19]
J. L. Kiplinger, T. G. Richmond, C. E. Osterberg, Chem. Rev.
Received September 18, 2002
1994, 94, 373Ϫ431.
[I02526]
Eur. J. Inorg. Chem. 2003, 638Ϫ643
643