Improved synthesis of 1,2-bis(trimethylsilyl)benzenes using Rieke-magnesium or the entrainment method

Article abstract of DOI:10.1002/adsc.201000560

1,2-Bis(trimethylsilyl)benzene is the key starting material for the synthesis of efficient benzyne precursors and certain luminescent π-conjugated materials. We now report that it can be conveniently prepared in tetrahydrofuran from 1,2-dibromobenzene, chlorotrimethylsilane, and either Rieke-magnesium (Mg^R) or magnesium turnings in the presence of 1,2-dibromoethane as an entrainer (Mg^e). The most important advantages of these new protocols over the currently best-established procedure (1,2-dichlorobenzene, chlorotrimethylsilane, magnesium turnings, hexamethylphosphoramide) lie in the milder reaction conditions (Mg^R: 0 °C, 2 h; Mg^e: room temperature, 30 min vs. 100 °C, 2 days) and in the fact that the cancerogenic solvent hexamethylphosphoramide is avoided. Moreover, the improved protocols are also applicable for the high-yield synthesis of 1,2,4,5-tetrakis(trimethylsilyl)benzene, 4-fluoro-1,2- bis(trimethylsilyl)benzene, 4-chloro-1,2-bis(trimethylsilyl)benzene, and 4,5-dichloro-1,2-bis(trimethylsilyl)benzene. Copyright

Full text of DOI:10.1002/adsc.201000560

UPDATES
DOI: 10.1002/adsc.201000560
Improved Synthesis of 1,2-Bis(trimethylsilyl)benzenes using
Rieke-Magnesium or the Entrainment Method
Andreas Lorbach,a Christian Reus,a Michael Bolte,a Hans-Wolfram Lerner,a
and Matthias Wagnera,*
a
Institut fꢀr Anorganische und Analytische Chemie, Goethe-Universitꢁt Frankfurt, Max-von-Laue-Strasse 7,
60438 Frankfurt (Main), Germany
Fax : (+49)-69-798-29260; e-mail: Matthias.Wagner@chemie.uni-frankfurt.de
Received: July 16, 2010; Revised: October 12, 2010; Published online: December 7, 2010
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201000560.
Abstract: 1,2-Bis(trimethylsilyl)benzene is the key
starting material for the synthesis of efficient ben-
zyne precursors and certain luminescent p-conjugat-
ed materials. We now report that it can be conven-
iently prepared in tetrahydrofuran from 1,2-dibro-
mobenzene, chlorotrimethylsilane, and either
Rieke-magnesium (MgR) or magnesium turnings in
the presence of 1,2-dibromoethane as an entrainer
(Mge). The most important advantages of these new
protocols over the currently best-established proce-
dure (1,2-dichlorobenzene, chlorotrimethylsilane,
magnesium turnings, hexamethylphosphoramide)
lie in the milder reaction conditions (MgR: 08C, 2 h;
Mge: room temperature, 30 min vs. 1008C, 2 days)
and in the fact that the cancerogenic solvent hexa-
ides with triethoxysilane or triorganosilanes.[3] Howev-
er, especially the synthesis of sterically encumbered
arylsilanes still poses considerable challenges as illus-
trated by the case of 1,2-bis(trimethylsilyl)benzene (1;
Scheme 1), which is only accessible via pathways re-
quiring either highly toxic or very costly chemicals.
ACHTUNGTRENNUNGmethylphosphoramide is avoided. Moreover, the
improved protocols are also applicable for the high-
yield synthesis of 1,2,4,5-tetrakis(trimethylsilyl)ben-
zene, 4-fluoro-1,2-bis(trimethylsilyl)benzene, 4-
chloro-1,2-bis(trimethylsilyl)benzene, and 4,5-di-
chloro-1,2-bis(trimethylsilyl)benzene.
Keywords: arenes; entrainment method; Grignard
reaction; Rieke-magnesium; silanes; silylation
Scheme 1. 1,2-Bis(trimethylsilyl)benzene (1) is a key starting
material for the preparation of the benzyne precursor A, the
polymer building block B, and Lewis-acid catalysts C. (i) cf.
Ref.[5]; (ii) cf. Refs.[6,7]; (iii) cf. Ref.[8]
Introduction
Arylsilanes and their hypervalent derivatives are val-
Compound 1 is the key starting material for the
uable intermediates in organic synthesis, for example, synthesis of (i) (phenyl)[2-(trimethylsilyl)phenyl]iodo-
in borylation[1] or Pd-catalyzed Hiyama cross-coupling nium triflate (A; Scheme 1), one of the most efficient
reactions.[2] Accordingly, various different methods benzyne precursors available to date,[4,5] (ii) 9,10-dihy-
for the preparation of (functionalized) arylsilanes dro-9,10-diboraanthracene (B; Scheme 1), a versatile
have been worked out, ranging from the arylation of building block of luminescent boron-doped p-conju-
halosilanes with organolithium or Grignard reagents gated polymers,[6,7] and (iii) 9,10-dimethyl-9,10-dihy-
to the transition metal-mediated silylation of aryl hal- dro-9,10-diboraanthracene, a powerful ditopic Lewis-
Adv. Synth. Catal. 2010, 352, 3443 – 3449
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3443

Andreas Lorbach et al.
UPDATES
Table 1. Comparison of the yields of 1,2-bis(trimethylsilyl)benzenes obtained with different Grignard methods.
Starting Material
Product R1
R2
Entrainment Approach[a] Rieke-Mg Approach[a] Fe-Catalyzed Approach[b] HMPA Approach
X=Br
X=Br
X=Br
X=Cl
1
2
4
5
8
11
12
H
X
H
X
Cl
H
H
62%
54%
67%
56%
59%
65%[c]
80%
70%
41%
38%
–
–
–
75%[d]
49%[e]
Cl
F
Cl
Me
t-Bu
–
–
–
53%
60%[f]
10–40%[h]
0%
Me <10%[g]
19%
14%
50%[i]
–
H
<10%[g]
[a]
[b]
[c]
[d]
[e]
[f]
This work.
Mg powder/Me3SiCl/DIBAL-H, TMEDA, FeCl3/THF, ꢀ108C to 08C, 1 day; cf. Ref.[15]
Isolated yield after redistillation of several combined forerunnings.
Mg turnings/Me3SiCl/HMPA, 1008C, 2 days; cf. Ref.[9]
Mg powder/Me3SiCl/HMPA-THF, 1008C, 2 days; cf. Ref.[26]
X=I.
[g]
[h]
[i]
Yields were estimated from the 1H NMR spectra of the crude product mixtures.
Yields suffer from poor reproducibility.
Mg/Me3SiCl/HMPA-THF; cf. Ref.[24]
acid catalyst (cf. C for the activation of 1,2-diazines;
The synthesis protocols for 1 reviewed thus far
demonstrate that substantial efforts have already
Scheme 1).[8]
The currently best-established synthesis protocol been spent on optimizing the reagents and the reac-
for compound 1, which starts from 1,2-dichloroben- tion conditions. However, in the case of the Grignard
zene, Mg turnings, and Me3SiCl, suffers from two approaches, only one study is known which focuses on
major disadvantages: (i) the use of the toxic and can- the arguably most influential parameter, i.e., the ac-
cerogenic
solvent
hexamethylphosphoramide tivity of the magnesium employed: Wegner et al. ap-
(HMPA) is necessary, and (ii) the transformation re- plied the diisobutylaluminium hydride (DIBAL-H)
quires high temperature (1008C) and a long reaction activation procedure[14] to magnesium powder and
time (2 days).[9] To avoid the use of HMPA, 1 can al- showed that a subsequent Grignard reaction with
ternatively be prepared from 1,2-dibromobenzene, Me3SiCl in THF provided 1 in 37% yield (reflux,
tert-butyllithium (4 equiv.), and an excess of 45 min).[15] They have also shown that a catalytic
Me3SiOTf (OTf=triflate; Et2O-THF, ꢀ1208C).[10] amount of anhydrous FeCl3 (3 mol%)[16] allows one to
Here, the use of Me3SiCl does not give 1 in apprecia- conduct the silylation at ꢀ108C (1 day; yield of 1:
ble amounts, which, together with the fact that tert- 41%) and thereby to prepare a number of alkyl-,
butyllithium is employed instead of Mg, renders this alkoxy- or fluorine-substituted 1,2-bis(trimethylsilyl)-
methodology rather expensive. Given this back- benzenes.
ground,
[2-(hydroxydimethylsilyl)phenyl]-
Herein, we will show that the reaction time of the
ACHTUNGTRENNUNG(phenyl)iodonium triflate has been suggested as a iron-catalyzed approach and the associated yield of 1
substitute for A,[11] because the synthesis of this new can be improved further by using (i) Rieke-Mg
benzyne precursor proceeds via 1,2-bis(dimethylsilyl)- (MgR)[17] or (ii) Mg turnings in the presence of 1,2-di-
benzene, which is sterically less encumbered than 1 bromoethane as an entrainer for continuous activa-
and
accessible
from
1,2-dibromobenzene/Mg/ tion (Mge;[18] Table 1). Both protocols also provide
Me2(H)SiCl in THF.[12] For the synthesis of B, the silyl convenient high-yield routes to 1,2,4,5-tetrakis(trime-
derivative 1 could, in principle, be replaced by 1,2- thylsilyl)benzene and first time access to 4-fluoro-1,2-
bis(trimethylstannyl)benzene.[13] However, in this case bis(trimethylsilyl)benzene, 4-chloro-1,2-bis(trimethyl-
4 equivalents of Me3SnCl are generated as a side silyl)benzene as well as 4,5-dichloro-1,2-bis(trimethyl-
product of the assembly of the 9,10-dihydro-9,10-di- silyl)benzene. Especially the latter two compounds
boraanthracene scaffold with BCl3. Me3SnCl is not are relevant, because they offer the possibility of ex-
only toxic but also difficult to remove quantitatively tensive functionalization via transition metal-mediat-
[19]
ꢀ
from the resulting 9,10-dichloro-9,10-dihydro-9,10-di- ed C C coupling reactions.
boraanthracene intermediate.
In those cases where
the substitution pattern on the phenylene ring is not
3444
asc.wiley-vch.de
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 3443 – 3449

Improved Synthesis of 1,2-Bis(trimethylsilyl)benzenes using Rieke-Magnesium
an issue (e.g., numerous applications of B; Scheme 1),
the readily available 1,2-dibromo-4,5-dimethylben-
zene[20] could be an economically attractive alterna-
tive to the parent 1,2-dibromobenzene as a starting
material. Thus, we also investigated the reaction be-
tween MgR/Mge, Me3SiCl, and 1,2-dibromo-4,5-dime-
thylbenzene.
Results and Discussion
The slurry of MgR in THF used in our investigations
was prepared from MgCl2 and 1.5 equivalents of K;
excess MgCl2 as well as the generated KCl were not
removed (cf. the Experimental Section). In a previous
report, the presence of KI in the MgR-forming step
has been described to produce a particularly reactive
metal powder.[21] For the synthesis of the compounds
discussed here, such a beneficial effect could not be
Scheme 2. Synthesis of the 1,2-bis(trimethylsilyl)benzenes 1
confirmed.
and 2 in THF; reaction of hexabromobenzene with Me3SiCl
and MgR in THF. (i) MgR: +5Me3SiCl, +2.5MgR, 08C,
1.5 h, 65%; Mge: +8Me3SiCl, +3Mg, +0.21,2-C2H4Br2,
room temperature, 30 min, 62%. (ii) MgR: +11Me3SiCl,
+9MgR, 08C!room temperature, 3.5 h, 80%; Mge:
+12Me3SiCl, +8Mg, +1.51,2-C2H4Br2, room temperature,
2 h, 54%. (iii) MgR: +30Me3SiCl, +15MgR, 08C!room
temperature, 1.5 h; 3 could be detected as one of several
components in the product mixture.
In the entrainment approach, continuous activation
at room temperature was achieved by dropwise addi-
tion of 1,2-dibromoethane to a mixture of Mg turn-
ings, the respective 1,2-dibromobenzene derivative,
and excess Me3SiCl in THF. An amount of 0.2–
0.8 equivalents of the entrainer was sufficient for
quantitative conversion. In contrast to the Rieke-Mg
protocol, which requires carefully dried solvent and
the strict maintenance of inert conditions, the THF
(p.a. grade) employed in the entrainment method was et al., who started from hexabromobenzene and de-
used as received from the commercial supplier and veloped a three-step synthesis sequence via hexakis-
the glassware was just briefly flushed with N2.
(dimethylsilyl)benzene.[28] Previous attempts of
ACHTUNGTRENNUNG
The reaction of MgR/Mge with 1,2-dibromobenzene Gilman et al. at the direct trimethylsilylation of hexa-
and excess Me3SiCl in THF at 08C/208C gave 1,2-bis- bromobenzene (Mg/Me3SiCl/THF, reflux, 12 h) have
ACHTUNGTRENNUNG
(trimethylsilyl)benzene (1) in 65%/62% yield failed.[29,30] The only well-defined product that could
(Scheme 2; Table 1).[22] In terms of solvent toxicity as be isolated in small amounts from the reaction mix-
well as the required reaction temperatures and times, tures, was 1,1,3,4,6,6-hexakis(trimethylsilyl)-1,2,4,5-
both protocols offer major advantages over the estab- hexatetraene (3; Scheme 2).[29,30] Since hexakis(trime-
lished method[9] while comparable product yields are thylsilyl)benzene is known to undergo a rearrange-
achieved (Table 1).
ment reaction to afford 3 upon thermal treatment,[28]
When applied to 1,2,4,5-tetrabromobenzene, both we reckoned that the lower reaction temperature re-
procedures also work faithfully for the preparation of quired for the Rieke-Mg protocols might offer an op-
1,2,4,5-tetrakis(trimethylsilyl)benzene (2; Scheme 2), portunity to prepare hexakis(trimethylsilyl)benzene in
a versatile building block for linear polycyclic aromat- one step. However, the reaction with MgR took a sim-
ic compounds.[23,24] The obtained yields of 2 (MgR: ilar course as the Gilman experiments and 3 was ob-
80%, Mge: 54%; Table 1) are substantially higher tained as one of several products (Scheme 2).
than the yields of alternative synthesis methods (Mg/
1,2,4,5-Br4C6H2/Me3SiCl/THF, reflux, 2 days: 20%;[25] zene (1), the halogenated derivatives 4, 5, and 8 were
Mg/1,2,4,5-Cl4C6H2/Me3SiCl/HMPA-THF, 1008C, also readily prepared from the corresponding 1,2-di-
2 days: Mg/1,2,4,5-Br4C6H2/Me3SiCl/ bromo- and 1,2-diiodobenzenes (Scheme 3; Table 1).
49%;[23,24,26]
DIBAL-H, TMEDA, FeCl3/THF, ꢀ108C to 08C, The use of 1,2-dibromo-4-chlorobenzene in the syn-
1 day: 38%[15]). thesis of 8 with MgR led to an undesired partial Cl/
Apart from the parent 1,2-bis(trimethylsilyl)ben-
Encouraged by these positive results, we next SiMe3 exchange even though the reaction tempera-
tested whether the use of MgR also provides a route ture was lowered to ꢀ408C. This problem was solved
for the six-fold silylation of hexabromobenzene with by using 4-chloro-1,2-diiodobenzene instead. With the
Me3SiCl.[27] The target compound, hexakis(trimethyl- Mge method, we were surprised to find that 4-chloro-
silyl)benzene, has already been described by Sakurai 1,2-diiodobenzene turned out to be far less reactive
Adv. Synth. Catal. 2010, 352, 3443 – 3449
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3445

Andreas Lorbach et al.
UPDATES
zene starting from 4-bromo-1,2-diiodobenzene and
MgR failed due to pronounced Br/SiMe3 exchange.
In summary, selectivity was not an issue in the case
of the transformations 1,2-dibromo-4,5-dichloroben-
zene!4 (MgR, Mge), 1,2-dibromo-4-fluorobenzene!5
(MgR, Mge), and 4-chloro-1,2-diiodobenzene!8
(MgR).
Compounds 4, 5, and 8 have been characterized by
(heteronuclear) NMR spectroscopy and combustion
analysis (cf. the Supporting Information). Moreover,
in the case of 4 an X-ray crystal structure analysis was
performed (Figure 1). Considerable steric strain
within this molecule can be inferred from the elon-
ꢀ
AHCTUNTGERNgGNUN ated Si Cipso distances, which possess an even larger
ꢀ
value [1.907(5) ꢃ] than the average Si CH3 bond
ꢀ
ꢀ
length [1.872(6) ꢃ]. Moreover, the Si(1) C(1) C(2)
ꢀ
ꢀ
and Si(2) C(2) C(1) bond angles are expanded to
126.7(4)8 and 129.2(4)8, respectively, and the torsion
angle Si(1) C(1) C(2) Si(2) of ꢀ10.4(8)8 deviates
ꢀ
ꢀ
ꢀ
from the ideal value of 08.
In order to find out whether the behavior of the
bromo (iodo) substituent in the para position to the
fluorine (chlorine) atom is different from the one in
the meta position, we repeated the reaction of 1,2-di-
bromo-4-fluorobenzene (4-chloro-1,2-diiodobenzene)
with only 1.1 equivalents of Mg. In this case, only the
Rieke-Mg approach was applied, because it allows
easy control of the actual amount of active metal
even on a very small scale (Scheme 3). About one-
quarter (one-half) of the organic starting material re-
mained unreacted. Consequently, significant amounts
of the halogenated 1,2-bis(trimethylsilyl)benzene 5
Scheme 3. Synthesis of the halogenated 1,2-bis(trimethylsi-
lyl)benzenes 4, 5, and 8 in THF; product distribution of the
corresponding reactions with 1.1 equivalents of MgR in
THF; drawings of 11 and 12. (i) MgR: +10Me3SiCl, +3MgR,
ꢀ408C!room temperature, 2 h, 70%; Mge: +8Me3SiCl,
+4Mg, +0.41,2-C2H4Br2, room temperature, 45 min, 67%.
(ii) MgR: +7Me3SiCl, +3MgR, room temperature!408C,
1 h, 53%; Mge: +8Me3SiCl, +4Mg, +0.41,2-C2H4Br2, room
temperature, 45 min, 56%. (iii) MgR: +5Me3SiCl, +1.1MgR,
room temperature!408C, 1 h. (iv) MgR: +9Me3SiCl,
+2.5MgR, ꢀ408C!room temperature, 2 h, 60%; Mge:
+10Me3SiCl, +15Mg, +101,2-C2H4Br2, room temperature,
4 h, 70%. (v) MgR: +5Me3SiCl, +1.1MgR, ꢀ408C!room
temperature, 2.5 h.
Figure 1. Molecular structure and numbering scheme of
compound 4. Displacement ellipsoids are drawn at the 50%
probability level. H-atoms have been omitted for clarity. Se-
lected bond lengths (ꢃ), bond angles (deg), and torsion
than
1,2-dibromo-4-chlorobenzene,
requiring
15 equivalents of Mg and 10 equivalents of entrainer
for full conversion. At the same time, the chemoselec-
tivity of the halogen/SiMe3 exchange was comparable
to that of 1,2-dibromo-4-chlorobenzene. Attempts at
ꢀ
ꢀ
angle (deg): Si(1) C(1)=1.908(5), Si(2) C(2)=1.905(5);
ꢀ
ꢀ
ꢀ
ꢀ
Si(1) C(1) C(2)=126.7(4),
the synthesis of 4-bromo-1,2-bis(trimethylsilyl)ben- Si(1) C(1) C(2) Si(2)=ꢀ10.4(8).
Si(2) C(2) C(1)=129.2(4);
ꢀ
ꢀ
ꢀ
3446
asc.wiley-vch.de
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 3443 – 3449

Improved Synthesis of 1,2-Bis(trimethylsilyl)benzenes using Rieke-Magnesium
(8) were formed in both cases; the product distribu- nylene (5%) and 4-tert-butyl-1,2-bis(trimethylsilyl)-
tion of both reactions was determined by NMR spec- benzene (12; <10%)/di-tert-butylbiphenylenes (5%),
troscopy and is shown in Scheme 3. Compared to 4- respectively (Scheme 3; Table 1).
chloro-1,2-diiodobenzene, 1,2-dibromo-4-fluoroben-
zene thus shows (i) a higher degree of overall conver-
sion, (ii) a lower degree of 1,2-disilylation, and (iii) a Conclusions
considerably lower selectivity between the two mono-
silylated isomers.
1,2-Bis(trimethylsilyl)benzene and 1,2,4,5-tetrakis(tri-
The assignment of the structures 6 and 9 to the methylsilyl)benzene have been conveniently prepared
major monosilylated isomers is based on 1H-1H- from the corresponding bromobenzenes and Me3SiCl
1
1
COSY, H-1H-ROESY, and H-29Si-HETCOR NMR in Grignard-type reactions, i.e., via the Rieke-Mg
experiments and will be explained for 1-bromo-4- method (MgR) and the entrainment method (Mge; en-
fluoro-2-trimethylsilylbenzene/2-bromo-4-fluoro-1-tri-
trainer: 1,2-dibromoethane). One important improve-
methylsilylbenzene (6/7), but similar arguments apply ment with respect to the currently best-established
to 9/10 (for plots of the spectra see the Supporting In- synthesis protocol for 1,2-bis(trimethylsilyl)benzene
formation). Both isomers 6/7 give rise to three signals lies in the fact that the syntheses can be carried out in
1
in the aromatic region of the H NMR spectrum. The THF rather than in the cancerogenic solvent HMPA.
major isomer shows: (i) one multiplet at 6.90 ppm, (ii) Moreover, the reactions readily proceed at room tem-
one doublet of doublets at 7.13 ppm, and (iii) one perature or below, as opposed to 1008C in the HMPA
3
doublet of doublets at 7.47 ppm. Values of JF,H
=
method, and the reaction times are reduced from
8.6 Hz and JF,H =4.8 Hz are evident from the major 2 days to 1.5 hours (MgR) to 30 minutes (Mge). The
4
19F NMR signal. Since a JH,F =4.8 Hz coupling only yields obtained are comparable for all three protocols.
4
appears in the 7.47 ppm 1H NMR resonance, we We are aware that bromobenzenes are more costly
assign this signal to the H-6 proton. The H-1H-COSY than the chlorobenzenes employed in the HMPA
1
spectrum, in turn, allows us to assign the 6.90 ppm route, however, this argument is more than outweigh-
and the 7.13 ppm resonances to H-5 and H-3, respec- ed by the gain in safety and time efficiency.
tively. The H-3 signal shows a cross peak to the major
As a result of the milder reaction conditions, func-
SiMe3 resonance in the 1H-1H-ROESY and in the tionalized 1,2-bis(trimethylsilyl)benzenes bearing
1H-29Si-HETCOR NMR spectrum. We therefore con- fluoro or chloro substituents at their aromatic rings
clude that the SiMe3 group of the major isomer occu- are also accessible with high selectivity and yields.
pies the 2-position of the benzene ring (cf. 6). Consis- Thus, a broad application of these 1,2-bis(trimethylsi-
tent with that, the SiMe3 resonance of the minor lyl)benzene derivatives in the field of benzyne
isomer 7 shows cross-peaks exclusively to the signal chemistry, organic optoelectronic materials and catal-
of the H-6 proton.
ysis can be envisaged.
1,2-Dibromo-4,5-dimethylbenzene
reacts
with
Me3SiCl and MgR at room temperature in THF to
give 1,2-bis(trimethylsilyl)-4,5-dimethylbenzene (11; Experimental Section
Scheme 3) in amounts of 10%–40%. The following
features of this transformation are remarkable: (i) General Remarks
The yields of 11 are low, even though the related
compound 1,2,4,5-tetrakis(trimethylsilyl)benzene is
formed in 80% yield under similar reaction condi-
tions. (ii) The yields of 11 suffer from poor reproduci-
bility, which is most likely due to different and hard-
to-control surface structures of the MgR employed
(this is not an issue in the cases discussed before). (iii)
We do not observe major by-products, but after a
period of clean formation of 11, the reaction comes to
a halt. Application of the Rieke-Mg protocol to 1,2-
dibromo-4-tert-butylbenzene did not lead to the target
compound 11, even at elevated temperatures.
All reactions using MgR were carried out under a nitrogen
atmosphere using Schlenk techniques and carefully dried
solvents. Me3SiCl was stored over CaH2 and transferred by
pipette into the reaction flask or dropping funnel. All reac-
tions using Mge were carried out in glassware that had been
briefly flushed with N2. THF (p.a. grade, stabilized with
0.025% BHT) was purchased from Acros Organics (Geel,
Belgium) and stored over KOH. Me3SiCl was obtained from
Apollo Scientific Ltd. (Stockport, UK) and used as received.
1,2-Dibromo-4,5-dichlorobenzene,[31] 4-chloro-1,2-diiodoben-
zene,[32] and 4-bromo-1,2-diiodobenzene[32] were synthesized
according to literature procedures.
The entrainment method, in contrast, results in a
quantitative consumption of 1,2-dibromo-4,5-dime-
thylbenzene as well as 1,2-dibromo-4-tert-butylben-
zene already at room temperature. In both cases,
complex product mixtures are formed, the major con-
Preparation of Rieke-Magnesium (MgR)
In a representative procedure, a stirred mixture of THF
(800 mL), potassium (36.5 g, 934 mmol), and anhydrous
stituents being 11 (<10%)/2,3,6,7-tetramethylbiphe- MgCl2 (61.0 g, 641 mmol) was carefully heated to reflux for
Adv. Synth. Catal. 2010, 352, 3443 – 3449
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3447

Andreas Lorbach et al.
UPDATES
3 h. The resulting dark salt-containing MgR slurry was di-
rectly used for further transformations.
Acknowledgements
M.W. gratefully acknowledges financial support by the Beil-
stein-Institut, Frankfurt/Main, Germany, within the research
collaboration NanoBiC. A.L. wishes to thank the Fonds der
Chemischen Industrie for a Ph.D. grant. C.R. has been sup-
ported by the Studienstiftung des deutschen Volkes.
Synthesis of 1,2-Bis(trimethylsilyl)benzene (1) via
MgR
A
mixture of 1,2-dibromobenzene (24.5 mL, 48.0 g,
203 mmol), Me3SiCl (135.0 mL, 115.6 g, 1064 mmol), and
THF (150 mL) was added dropwise at 08C over 1 h to a
freshly prepared stirred suspension of MgR in THF (800 mL,
0.584M, 467 mmol). The slurry was stirred for further
30 min at 08C and then carefully quenched under nitrogen
with a saturated aqueous solution of NaHCO3 (300 mL).
The formation of two phases was observed, which were sep-
arated with a separation funnel. The aqueous phase was ex-
tracted with hexane (5ꢄ40 mL), the THF phase and the ex-
tracts were combined, washed with water (5ꢄ40 mL), dried
over anhydrous MgSO4, and filtered. All volatiles were re-
moved from the filtrate under vacuum to obtain a mixture
of 1 (70%; NMR spectroscopic control) and PhSiMe3 (30%)
as a pale orange oil. Product 1 was isolated as a colorless
liquid by fractional distillation under reduced pressure (60–
658C, 10ꢀ2 Torr); yield: 23.53 g (52%). Note: The overall
yield can be increased to approx. 65% when the combined
forerunnings of several distillations are redistilled.
References
[1] a) D. Kaufmann, Chem. Ber. 1987, 120, 853–854; b) D.
Kaufmann, Chem. Ber. 1987, 120, 901–905; c) U.
Gross, D. Kaufmann, Chem. Ber. 1987, 120, 991–994;
d) Y. Qin, G. Cheng, A. Sundararaman, F. Jꢁkle, J. Am.
Chem. Soc. 2002, 124, 12672–12673; e) M. C. Habe-
AHCTUNGTREGrNNUN echt, J. B. Heilmann, A. Haghiri, M. Bolte, J. W. Bats,
H.-W. Lerner, M. C. Holthausen, M. Wagner, Z. Anorg.
Allg. Chem. 2004, 630, 904–913.
[2] For selected reviews, see: a) Y. Hatanaka, T. Hiyama,
Synlett 1991, 845–853; b) T. Hiyama, E. Shirakawa,
Top. Curr. Chem. 2002, 219, 61–85.
[3] a) D. Hꢁbich, F. Effenberger, Synthesis 1979, 841–876;
b) M. Murata, K. Ota, H. Yamasaki, S. Watanabe, Y.
Masuda, Synlett 2007, 1387–1390 and references cited
therein.
Synthesis of 1,2-Bis(trimethylsilyl)benzene (1) via
Mge
[4] T. Kitamura, M. Yamane, J. Chem. Soc., Chem.
Commun. 1995, 983–984.
Mg turnings (3.03 g, 125 mmol) were covered with THF
(100 mL) and treated with neat Me3SiCl (42.2 mL, 36.1 g,
332 mmol) and 1,2-dibromobenzene (5.00 mL, 9.78 g,
41.5 mmol). A temperature of approximately 208C was
maintained by means of a water bath throughout the reac-
tion time, because otherwise the mixture warms perceptibly
after an induction period of about 5 min. A solution of 1,2-
dibromoethane (0.716 mL, 1.56 g, 8.30 mmol) in THF
(10 mL) was added dropwise to the vigorously stirred slurry
over 30 min. All volatiles were removed under vacuum and
the remaining solid residue was treated at 08C with hexane
(60 mL) and water (60 mL). The two phases were separated
and the aqueous phase was extracted with hexane (2ꢄ
20 mL). The combined organic layers were washed with
water (4ꢄ20 mL), dried over anhydrous MgSO4, filtered,
and all volatiles were removed from the filtrate under
vacuum. The remaining orange oil was fractionally distilled
under reduced pressure (60–658C, 10ꢀ2 Torr) to afford 1 as a
colorless oil; yield: 5.69 g (62%).
[5] T. Kitamura, M. Yamane, K. Inoue, M. Todaka, N. Fu-
katsu, Z. Meng, Y. Fujiwara, J. Am. Chem. Soc. 1999,
121, 11674–11679.
[6] A. Lorbach, M. Bolte, H. Li, H.-W. Lerner, M. C.
Holthausen, F. Jꢁkle, M. Wagner, Angew. Chem. 2009,
121, 4654–4658; Angew. Chem. Int. Ed. 2009, 48, 4584–
4588.
[7] A. Lorbach, M. Bolte, H.-W. Lerner, M. Wagner, Orga-
nometallics 2010, 29, 5762–5765.
[8] a) S. N. Kessler, H. A. Wegner, Org. Lett. 2010, 12,
4062–4065; b) A. Lorbach, M. Bolte, H.-W. Lerner, M.
Wagner, Chem. Commun. 2010, 46, 3592–3594.
[9] T. Kitamura, M. Todaka, Y. Fujiwara, Org. Synth. 2002,
78, 104–108.
[10] H. F. Bettinger, M. Filthaus, J. Org. Chem. 2007, 72,
9750–9752.
[11] T. Kitamura, Z. Meng, Y. Fujiwara, Tetrahedron Lett.
2000, 41, 6611–6614.
[12] W. Fink, Helv. Chim. Acta 1974, 57, 1010–1015.
[13] J. J. Eisch, B. W. Kotowicz, Eur. J. Inorg. Chem. 1998,
761–769.
1H and 13C{1H} NMR data were in accord with published
values.[10]
[14] U. Tilstam, H. Weinmann, Org. Process Res. Dev. 2002,
6, 906–910.
[15] S. L. Bader, S. N. Kessler, H. A. Wegner, Synthesis
2010, 2759–2762.
Supporting Information
Synthesis details and NMR spectroscopic characterization of
2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12, H NMR spectrum of the
1
´
[16] a) B. Bogdanovic, M. Schwickardi, Angew. Chem. 2000,
crude product 1 (Mge), 1H-1H-COSY, 1H-1H-ROESY, and
1H-29Si-HETCOR NMR spectra of product mixtures ob-
tained in the experiments with substoichiometric amounts of
MgR as well as details of the X-ray crystal structure analyses
and key crystallographic data of 4-chloro-1,2-diiodobenzene,
4-bromo-1,2-diiodobenzene, and 4 are available in the Sup-
porting Information.
112, 4788–4790; Angew. Chem. Int. Ed. 2000, 39, 4610–
4612; b) W. M. Czaplik, M. Mayer, A. Jacobi von Wan-
gelin, Angew. Chem. 2009, 121, 616–620; Angew.
Chem. Int. Ed. 2009, 48, 607–610.
[17] a) R. D. Rieke, P. M. Hudnall, J. Am. Chem. Soc. 1972,
94, 7178–7179; b) R. D. Rieke, Acc. Chem. Res. 1977,
10, 301–306; c) R. D. Rieke, S. E. Bales, P. M. Hudnall,
3448
asc.wiley-vch.de
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 3443 – 3449

Improved Synthesis of 1,2-Bis(trimethylsilyl)benzenes using Rieke-Magnesium
T. P. Burns, G. S. Poindexter, Org. Synth. 1979, 59, 85–
92; d) R. D. Rieke, Science 1989, 246, 1260–1264.
[18] D. E. Pearson, D. Cowan, J. D. Beckler, J. Org. Chem.
1959, 24, 504–509.
[25] H. Bock, W. Kaim, Z. Anorg. Allg. Chem. 1979, 459,
103–110.
[26] H. Bock, M. Ansari, N. Nagel, Z. Havlas, J. Organomet.
Chem. 1995, 499, 63–71.
[19] a) A. F. Littke, G. C. Fu, Angew. Chem. 2002, 114,
4350–4386; Angew. Chem. Int. Ed. 2002, 41, 4176–
4211; b) R. B. Bedford, C. S. J. Cazin, D. Holder,
Coord. Chem. Rev. 2004, 248, 2283–2321.
[20] P. R. Ashton, U. Girreser, D. Giuffrida, F. H. Kohnke,
J. P. Mathias, F. M. Raymo, A. M. Z. Slawin, J. F. Stod-
dart, D. J. Williams, J. Am. Chem. Soc. 1993, 115, 5422–
5429.
[21] R. D. Rieke, S. E. Bales, J. Chem. Soc., Chem.
Commun. 1973, 879–880.
[22] Attempts to conduct these silylation protocols with the
low-priced starting material 1,2-dichlorobenzene failed,
even when higher reaction temperatures and/or longer
reaction times were applied.
[27] Since the amount of entrainer required for the synthe-
sis of 1,2,4,5-tetrakis(trimethylsilyl)benzene (2) is al-
ready one order of magnitude higher than the amount
required for the preparation of 1,2-bis(trimethylsilyl)-
benzene (1), this experiment was not conducted with
Mge.
[28] H. Sakurai, K. Ebata, C. Kabuto, A. Sekiguchi, J. Am.
Chem. Soc. 1990, 112, 1799–1803.
[29] T. Brennan, H. Gilman, J. Organomet. Chem. 1968, 11,
625–626.
[30] D. Ballard, T. Brennan, F. W. G. Fearon, K. Shiina, I.
Haiduc, H. Gilman, Pure Appl. Chem. 1969, 19, 449–
472.
[31] F. Raymo, F. H. Kohnke, F. Cardullo, Tetrahedron 1992,
48, 6827–6838.
[23] C.-Y. Yick, S.-H. Chan, H. N. C. Wong, Tetrahedron
Lett. 2000, 41, 5957–5961.
[24] S.-H. Chan, C.-Y. Yick, H. N. C. Wong, Tetrahedron
[32] R. J. Perry, S. R. Turner, J. Org. Chem. 1991, 56, 6573–
6579.
2002, 58, 9413–9422.
Adv. Synth. Catal. 2010, 352, 3443 – 3449
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3449