X-ray crystal structure of tris(11-crown-3)triphenylene

Article abstract of DOI:10.1016/S0022-2860(01)00606-8

Benzo-11-crown-3 ether trimerizes in the presence of FeCl₃ and aqueous H₂SO₄ to produce tris(11-crown-3)triphenylene. This compound crystallizes, along with a molecule of diethyl ether and CH₂Cl₂ in the triclinic P1? space group; a = 8.8913(9) ?, b = 13.5790(14) ?, c = 17.5244(17) ?, β = 91.026(2)°, with Z = 2. Two of the three 11-crown-3 ether units of the trimer possess essentially identical geometries, while the third has a significantly different conformation. These stereochemical results are compared to those calculated for benzo-11-crown-3 in the gas phase.

Full text of DOI:10.1016/S0022-2860(01)00606-8

Journal of Molecular Structure 598 ꢀ2001) 145±151
www.elsevier.com/locate/molstruc
X-ray crystal structure of trisꢀ11-crown-3)triphenylene
a,
a
Gerald W. Buchanan , Mastaneh Azad , Glenn P.A. Yap
b
*
aDepartment of Chemistry, Ottawa Carleton Chemistry Institute, Carleton University, 1125 Colonel By Drive, Ottawa, Ont., Canada K1S 5B6.
bDepartment of Chemistry, Ottawa-Carleton Chemistry Institute, University of Ottawa, Ottawa, Ont., Canada K1A 6N5.
Received 12 January 2001; accepted 26 March 2001
Abstract
Benzo-11-crown-3 ether trimerizes in the presence of FeCl3 and aqueous H2SO4 to produce trisꢀ11-crown-3)triphenylene.
Ê
Å
This compound crystallizes, along with a molecule of diethyl ether and CH2Cl2 in the triclinic P1 space group; a ˆ 8.8913ꢀ9) A,
Ê
Ê
b ˆ 13.5790ꢀ14) A, c ˆ 17.5244ꢀ17) A, b ˆ 91.026ꢀ2)8, with Z ˆ 2. Two of the three 11-crown-3 ether units of the trimer
possess essentially identical geometries, while the third has a signi®cantly different conformation. These stereochemical results
are compared to those calculated for benzo-11-crown-3 in the gas phase. q 2001 Elsevier Science B.V. All rights reserved.
Keywords: Crown ether; Oxidative trimerization; Stereochemistry
1. Introduction
triphenylene 2.The single-crystal X-ray structure of 2
has been determined.
Hexaalkoxy-substituted triphenylenes show liquid-
crystal properties if the side chains are of suf®cient
length. Hence, these molecules are important models
for discotic liquid crystals [1,2]. A procedure for
preparing 2,3,6,7,10,11-hexamethoxytriphenylene in
near-quantitative yield was reported [3] using oxida-
tive trimerization of 1,2-methoxybenzene, and
recently we reported the application of this method
to the synthesis of a novel trisꢀ9-crown-3)triphenylene
in ca. 25% yield [4].
We are interested in exploring the scope of this
novel reaction to the synthesis and characterization
of other trisꢀcrown)triphenylene systems. Herein we
report the synthesis of benzo-11-crown-3 ether 1 for
the ®rst time and its trimerization to trisꢀ11-crown-3)-
2. Results and discussion
In the presence of FeCl3 and H2SO4, 1 trimerizes in
a low isolated yield ꢀ0.5%) to the triphenylene system
2 depicted in Fig. 1. Fig. 2 shows the ORTEP diagram
for 2. Table 1 contains selected atomic bond lengths
and angles. Selected torsion angles involving non-
hydrogen atoms are given in Table 2, while the crys-
tallographic data are presented in Table 3. Atomic
coordinates and equivalent isotropic displacement
factors are given in Table 4. Anisotropic temperature
factors are available as supplementary data as are
structure factor tables.
2.1. X-ray crystal structure
* Corresponding author. Tel.: 11-613-520-3841; fax: 11-613-
520-3749.
With respect to bond lengths, there are signi®cant
differences found in the aryl C±O units, ranging from
E-mail address: gwbuchan@ccs.carleton.ca ꢀG.W. Buchanan).
0022-2860/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-2860ꢀ01)00606-8

146
G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
Fig. 1.
triphenylene in the 13C CPMAS spectrum [4]. Unfor-
tunately, the limited quantities of 2 available have
precluded a solid state 13C NMR investigation at
this time.
In examining the bond angle data, it is clear that
there are three anomalously large angles involving
aromatic carbons and three correspondingly smaller
angles involving the same central atoms. The central
carbons involved are C3, C15 and C34. In the case of
Ê
the shortest ꢀ1.361 A) for C3±O1 to the longest
Ê
ꢀ1.381 A) for the C10±O3 moiety. A similar range
of C±O bond lengths was noted for the related
trisꢀ9-crown-3)triphenylene studied previously [4].
A shorter C±O bond can be viewed as possessing
more double bond character which should lead to a
more deshielded 13C resonance in the solid state
spectrum. Indeed a range of ca. 5 ppm was found
for the six aryl C±O resonances in trisꢀ9-crown-3)
Fig. 2.

G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
147
Table 1
ꢀa) Atomic bond lengths ꢀA) and ꢀb) selected atomic bond angles ꢀ8)
Ê
Ê
Atomic bond lengths 2A)
O1±C3
1.361ꢀ3)
1.425ꢀ3)
1.364ꢀ3)
1.437ꢀ3)
1.383ꢀ3)
1.430ꢀ3)
1.412ꢀ3)
1.521ꢀ3)
1.364ꢀ3)
1.405ꢀ3)
1.405ꢀ3)
1.511ꢀ3)
1.410ꢀ3)
1.411ꢀ3)
1.510ꢀ3)
1.505ꢀ3)
O1±C4
1.431ꢀ3)
1.381ꢀ3)
1.416ꢀ3)
1.378ꢀ3)
1.444ꢀ3)
1.365ꢀ3)
1.419ꢀ3)
1.513ꢀ3)
1.409ꢀ3)
1.408ꢀ3)
1.513ꢀ3)
1.507ꢀ3)
1.460ꢀ3)
1.364ꢀ3)
1.508ꢀ3)
1.369ꢀ3)
O2±C6
1.418ꢀ3)
1.450ꢀ3)
1.419ꢀ3)
1.453ꢀ3)
1.428ꢀ3)
1.440ꢀ3)
1.466ꢀ3)
1.517ꢀ3)
1.469ꢀ3)
1.377ꢀ3)
1.514ꢀ4)
1.370ꢀ3)
1.411ꢀ3)
1.410ꢀ3)
1.520ꢀ3)
1.414ꢀ3)
O2±C7
O3±C10
O4±C16
O6±C22
O7±C28
O9±C34
C1±C2
O3±C9
O4±C15
O5±C18
O7±C27
O8±C30
C1±C12
C5±C6
O5±C19
O6±C21
O8±C31
O9±C33
C1±C36
C8±C9
C7±C8
C10±C11
C13±C24
C15±C22
C19±C20
C23±C24
C25±C36
C28±C29
C32±C33
C11±C12
C13±C14
C16±C17
C20±C21
C24±C25
C26±C27
C29±C30
C34±C35
C12±C13
C14±C15
C17±C18
C22±C23
C25±C26
C27±C34
C31±C32
C35±C36
2b) Selected atomic bond angles 28)
C3±O1±C4
120.2ꢀ2)
112.1ꢀ2)
115.1ꢀ2)
112.3ꢀ2)
119.5ꢀ2)
113.8ꢀ2)
114.5ꢀ2)
108.5ꢀ2)
110.9ꢀ2)
119.0ꢀ2)
115.2ꢀ2)
114.3ꢀ2)
109.2ꢀ2)
107.7ꢀ2)
121.1ꢀ2)
111.4ꢀ2)
107.4ꢀ2)
106.0ꢀ2)
113.9ꢀ2)
C6±O2±C7
114.6ꢀ2)
C10±O3±C9
C19±O5±C18
C27±O7±C28
C34±O9±C33
O1±C3±C10
C4±C5±C6
C15±O4±C16
C22±O6±C21
C31±O8±C30
O1±C3±C2
120.6ꢀ2)
116.3ꢀ2)
114.4ꢀ2)
126.4ꢀ2)
105.6ꢀ2)
115.5ꢀ2)
113.0ꢀ2)
121.5ꢀ2)
125.2ꢀ2)
107.4ꢀ2)
110.4ꢀ2)
114.1ꢀ2)
119.1ꢀ2)
119.2ꢀ2)
114.5ꢀ2)
115.2ꢀ2)
126.1ꢀ2)
O1±C4±C5
O2±C6±C5
O2±C7±C8
C7±C8±C9
C8±C9±O3
C11±C10±O3
O4±C15±C14
O4±C16±C17
O5±C18±C17
C19±C20±C21
C23±C22±O6
O7±C27±C34
C30±C29±C28
O8±C31±C32
O9±C34±C35
C3±C10±O3
O4±C15±C22
C16±C17±C18
O5±C19±C20
O6±C21±C20
O6±C22±C15
O7±C28±C29
O8±C30±C29
O9±C33±C32
O9±C34±C27
C3, the O1±C3±C2 angle is found to be 126.48, which
is ca. 68 larger than expected for an sp2 hybridized
centre. It appears that the O1±C3±C2 angle `opens
up' in an attempt to minimize the effect of the steric
repulsion resulting from the nearly coplanar
ꢀu ˆ 17.38) torsional C4±O1±C3±C2 unit. Such an
expansion does not occur in the O3±C10±C11 unit
since the torsion angle for the C9±O3±C10±C11
network is 2102.78. The expansion of the O1±C3±
C2 bond angle is accompanied by a contraction of the
O1±C3±C10 angle to 113.88 with the `internal' C2±
C3±C10 angle having a normal value of ca. 1208.
At C15, we attribute the anomalously large ꢀ125.28)
O4±C15±C14 angle to the same factors as described
before for C3, since the C14±C15±O4±C16 torsion
angle is 14.78. Similarly at C34, the expansion of the
O9±C34±C35 angle to 126.18 can be attributed to the
repulsive torsional interactions arising from the nearly
coplanar ꢀu ˆ 16.78) C33±O9±C34±C35 network.
With respect to the conformation of the 11-
membered rings in 2, there are no reports in the litera-
ture of any 11-crown-3 systems. A comparison of
torsion angles for the 11-membered rings is presented
in Table 5 and it is evident that two of the three

148
G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
Table 2
Selected torsion angles ꢀ8)
C4±O1±C3±C2
C1±C2±C3±O1
17.3ꢀ3)
177.5ꢀ2)
174.8ꢀ2)
49.3ꢀ3)
C4±O1±C3±C10
O1±C4±C5±C6
2 163.0ꢀ3)
2 77.3ꢀ2)
68.1ꢀ3)
C3±O1±C4±C5
C7±O2±C6±C5
C4±C5±C6±O2
O2±C7±C8±C9
C6±O2±C7±C8
2 173.3ꢀ2)
2 146.6ꢀ2)
2 102.7ꢀ3)
2 175.2ꢀ2)
2 2.8ꢀ2)
60.2ꢀ3)
C10±O3±C9±C8
C9±O3±C10±C11
O1±C3±C10±C11
C11±C12±C13±C14
C16±O4±C15±C22
C15±O4±C16±C17
C19±O5±C18±C17
C18±O5±C19±C20
C22±O6±C21±C20
C21±O6±C22±C23
C28±O7±C27±C26
C27±O7±C28±C29
C31±O8±C30±C29
C30±O8±C31±C32
C34±O9±C33±C32
C33±O9±C34±C35
C7±C8±C9±O3
71.3ꢀ2)
C9±O3±C10±C3
O3±C10±C11±C12
C16±O4±C15±C14
C13±C14±C15±O4
O4±C16±C17±C18
C16±C17±C18±O5
O5±C19±C20±C21
C19±C20±C21±O6
C21±O6±C22±C15
C28±O7±C27±C34
O7±C28±C29±C30
C28±C29±C30±O8
O8±C31±C32±C33
C31±C32±C33±O9
C33±O9±C34±C27
78.4ꢀ2)
178.6ꢀ2)
14.7ꢀ4)
2 164.7ꢀ2)
2 174.5ꢀ3)
2 129.5ꢀ3)
156.2ꢀ2)
2 179.6ꢀ2)
2 50.2ꢀ3)
81.6ꢀ3)
2 68.7ꢀ3)
63.4ꢀ3)
148.6ꢀ2)
120.8ꢀ2)
2 65.1ꢀ3)
79.4ꢀ2)
2 101.5ꢀ2)
2 144.3ꢀ2)
2 173.9ꢀ2)
68.2ꢀ3)
69.7ꢀ3)
60.4ꢀ3)
48.8ꢀ3)
2 75.7ꢀ3)
2 163.1ꢀ2)
173.1ꢀ2)
16.7ꢀ3)
Table 3
11-membered rings of 2 have essentially the same
geometries, while the third ꢀC15±C22) has a unique
conformation. Unfortunately, benzo-11-crown-3,1, is
an oil, however we have used the HyperChem
program to calculate the preferred geometry of the
11-membered ring of 1 in the gas phase. These data
are included in Table 5 for comparison. It is important
to include H atoms in the HyperChem structure of 1
and a fair agreement between observed solid state
geometries and the calculated gas phase structure for
the 11-membered ring of 1 is found. When viewing
the ORTEP plot of 2 ꢀFig. 2), it is clear that there is a
`northern' ring ꢀi.e. C3±C10), an `eastern' ring ꢀi.e.
C15±C22) and a `western' ring ꢀC27±C34). The
northern and western rings have identical conforma-
tions. The primary differences between the conforma-
tion of the eastern ring and the other two crown units
are in the region most remote from the central core of
the molecule-i.e. in the C5±C6±O2±C7±C8 region,
where differences in torsion angles up to ca. 608 are
found.
Crystallographic data for 2
Formula
C41HO10
777.74
203ꢀ2)
Formula weight ꢀg/mol)
Temperature ꢀK)
Crystal system
Space group
Ê
a ꢀA)
Ê
b ꢀA)
triclinic
P-1
8.8913ꢀ9)
13.5790ꢀ14)
17.5244ꢀ17)
91.026ꢀ2)
1909.6ꢀ3)
2
Ê
c ꢀA)
b ꢀ8)
Volume ꢀA )
Ê 3
Z ꢀmolecules per unit cell)
Dc ꢀgm/cm3)
Crystal dimensions ꢀmm3)
Ê
Radiation ꢀl, A)
Octants measured
1.353
0.2 £ 0.1 £ 0.1
MoKa, 0.71073
hꢀ211 to 11),
kꢀ218 to 17),
lꢀ222 to 17)
1.66 to 28.68
8624
Theta range ꢀ8)
No. of unique re¯ections
No. of re¯ections measured
14,943
No. of re¯ections with I . 2sꢀI)
R1 ꢀI . 2sꢀI))
R1 ꢀall data)
7724
0.0547
0.0998
Conversely, the calculations for 1 in the gas phase
show the largest differences in torsion angles from
those found in 2 in the region surrounding the site
corresponding to O1 in 1. Of course, crystal packing
effects limit the value of detailed comparisons
wR2ꢀall data)
0.1491
wR2ꢀI . 2sꢀI))
GoF
0.1343
1.04
Largest diff. peak and hole
Ê 23
0.600 and 20.680
ꢀe A
)

G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
149
Table 4
Atomic coordinates ꢀ £ 104) and equivalent isotropic displacement
between gas phase calculations and experimental
X-ray data.
The four aromatic rings of 2 are essentially
coplanar, in contrast to the situation for trisꢀ9-
crown-3)triphenylene [4] where deviations from
coplanarity up to ca. 148 were found.
3
Ê 2
parameters ꢀA £ 10 ). Uꢀeq) is de®ned as one third of the trace of
the orthogonalized Uij tensor
x
y
z
Uꢀeq)
O1
4517ꢀ2)
4380ꢀ2)
4238ꢀ2)
8207ꢀ2)
9090ꢀ2)
2 1156ꢀ1)
1070ꢀ1)
1611ꢀ1)
2625ꢀ1)
4580ꢀ1)
5839ꢀ1)
3844ꢀ1)
2 227ꢀ1)
35ꢀ1)
36ꢀ1)
29ꢀ1)
41ꢀ1)
35ꢀ1)
36ꢀ1)
29ꢀ1)
35ꢀ1)
35ꢀ1)
24ꢀ1)
27ꢀ1)
28ꢀ1)
33ꢀ1)
33ꢀ1)
39ꢀ1)
37ꢀ1)
37ꢀ1)
33ꢀ1)
25ꢀ1)
26ꢀ1)
24ꢀ1)
24ꢀ1)
29ꢀ1)
31ꢀ1)
35ꢀ1)
39ꢀ1)
42ꢀ1)
36ꢀ1)
35ꢀ1)
35ꢀ1)
29ꢀ1)
29ꢀ1)
24ꢀ1)
25ꢀ1)
25ꢀ1)
27ꢀ1)
31ꢀ1)
34ꢀ1)
33ꢀ1)
39ꢀ1)
35ꢀ1)
33ꢀ1)
27ꢀ1)
27ꢀ1)
24ꢀ1)
O2
2 2919ꢀ1)
2 51ꢀ1)
5573ꢀ1)
7656ꢀ1)
6404ꢀ1)
3436ꢀ1)
1777ꢀ1)
1322ꢀ1)
1443ꢀ2)
348ꢀ2)
2.2. 13C NMR spectra
O3
O4
The solution 13C spectrum ꢀCDCl3 solvent)
contains only six resonances, as expected from the
three-fold symmetry element present in the planar
depiction of 2, which arises from suf®ciently rapid
conformational interconversions at room temperature.
O5
O6
O7
10513ꢀ2)
11755ꢀ2)
10255ꢀ2)
10504ꢀ1)
7378ꢀ2)
6508ꢀ3)
5450ꢀ2)
4862ꢀ3)
3688ꢀ3)
4038ꢀ2)
3103ꢀ3)
3685ꢀ3)
4912ꢀ3)
5249ꢀ2)
6033ꢀ2)
7093ꢀ2)
7907ꢀ2)
7606ꢀ3)
8416ꢀ3)
7368ꢀ3)
7213ꢀ3)
8723ꢀ3)
10612ꢀ3)
10640ꢀ3)
9758ꢀ3)
9569ꢀ3)
9856ꢀ3)
9025ꢀ2)
9331ꢀ2)
10396ꢀ2)
10722ꢀ2)
10979ꢀ3)
11859ꢀ3)
11837ꢀ3)
10020ꢀ3)
10342ꢀ3)
9599ꢀ3)
10009ꢀ3)
8925ꢀ3)
8544ꢀ2)
O8
2 2386ꢀ1)
2 936ꢀ1)
1391ꢀ2)
998ꢀ1)
O9
C1
C2
C3
2 129ꢀ2)
2 1949ꢀ2)
2 3008ꢀ2)
2 3584ꢀ2)
2 2549ꢀ2)
2 1747ꢀ2)
2 758ꢀ2)
461ꢀ2)
1395ꢀ1)
336ꢀ2)
3. Experimental
C4
C5
C6
202ꢀ2)
759ꢀ2)
3.1. Materials
C7
1999ꢀ2)
2871ꢀ2)
2904ꢀ2)
2220ꢀ1)
2597ꢀ1)
2198ꢀ1)
2612ꢀ1)
3406ꢀ1)
3810ꢀ1)
5110ꢀ2)
5872ꢀ2)
6264ꢀ2)
5600ꢀ2)
4882ꢀ2)
4092ꢀ2)
3426ꢀ1)
2648ꢀ1)
2214ꢀ1)
1382ꢀ1)
957ꢀ1)
C8
C9
The synthetic sequence leading to 2 begins with the
synthesis of the previously unknown benzo-11-
crown-3 ether,1. Preparation of 1 was carried out
starting with 3-chloropropanol ꢀAldrich) which was
subsequently converted to 1,7-dichloro-4-oxaheptane
via the following method.
A 250 ml round-bottom ¯ask was charged with
3-chloropropanol ꢀ100.0 g, 1.1 mol) and the ¯ask
was ®tted with a condenser and a dropping funnel.
To this was added drop wise with stirring, concen-
trated H2SO4 ꢀ70 ml) and the reaction mixture was
re¯uxed overnight. After cooling, the mixture was
poured onto ice ꢀ300 g), neutralized with 3 M NaOH
and extracted with CH2Cl2 ꢀ4 £ 300 ml). The
combined organic layers were dried over anhydrous
sodium sulfate and the drying agent was removed
by ®ltration. The solvent was removed via roto
evaporation to yield 39.5 g of crude 1,7-dichloro-4-
oxaheptane, as a dark green oil, which was used
without further puri®cation in the next step.
For preparation of 1, a three-necked 3 l round
bottom ¯ask, equipped with a dropping funnel and
condenser, was charged with catechol ꢀAldrich,
25.0 g, 0.23 mol), distilled water ꢀ2.5 l) and
LiOH´2H2O ꢀ20.0 g, 0.48 mol). This mixture was
then stirred under re¯ux for 1 h. Subsequently 1,7-
dichloro-4-oxaheptane ꢀ38.9 g, 0.23 mol) was added
drop wise with stirring and the mixture was re¯uxed
for 5 days. After cooling, the mixture was acidi®ed to
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
C35
C36
1527ꢀ2)
2054ꢀ2)
3208ꢀ2)
3839ꢀ2)
4906ꢀ2)
5137ꢀ2)
6088ꢀ2)
6923ꢀ2)
7798ꢀ2)
8165ꢀ2)
7305ꢀ2)
5379ꢀ2)
4775ꢀ2)
3687ꢀ2)
3056ꢀ2)
3536ꢀ2)
2953ꢀ2)
3677ꢀ2)
3519ꢀ2)
2346ꢀ2)
626ꢀ2)
186ꢀ1)
2 845ꢀ1)
2 1593ꢀ2)
2 2104ꢀ2)
2 2812ꢀ2)
2 2281ꢀ2)
2 1488ꢀ1)
2 192ꢀ1)
190ꢀ1)
11ꢀ2)
272ꢀ2)
1830ꢀ2)
1350ꢀ2)
1949ꢀ2)
980ꢀ1)

150
G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
Table 5
Comparison of calculated 11-membered ring torsion angles ꢀ8) in 1 with those in the crystal structure of 2
Network
1
C3±C10 ꢀ2)
C15±C22 ꢀ2)
C27±C34 ꢀ2)
C3±C10±O3±C9
C10±O3±C9±C8
O3±C9±C8±C7
C9±C8±C7±O2
C8±C7±O2±C6
C7±O2±C6±C5
O2±C6±C5±C4
C6±C5±C4±O1
C5±C4±O1±C3
C4±O1±C3±C10
O1±C3±C10±O3
61.4
2 148.7
98.7
78.4
2 146.6
71.3
2 65.1
148.6
79.4
2 144.3
69.7
63.4
2 85.9
161.1
60.2
2 68.7
156.2
60.4
2 173.3
68.1
2 173.9
68.2
2 74.3
2 50.1
100.6
2 129.5
81.6
49.3
48.8
2 77.3
174.8
2 163.0
3.7
2 50.2
2 174.5
2 164.7
6.4
2 75.7
173.1
2 163.1
3.2
2 166.6
109.5
2 10.9
pH 3 using concentrated H2SO4 and then extracted
with CH2Cl2 ꢀ3 £ 300 ml). The combined organic
layers were washed with 5% NaOH ꢀ3 £ 300 ml)
and water ꢀ2 £ 200 ml) and dried over anhydrous
sodium sulfate. Rotary evaporation yielded an oil,
which was then treated with boiling hexane
ꢀ5 £ 300 ml). The hexane extracts were decanted
into a beaker, whereupon a small quantity ꢀ0.8 g) of
white crystals precipitated. This material was
removed by ®ltration and subsequently identi®ed as
dibenzo-22-crown-6 ether. The hexane ®ltrate was
concentrated by rotary evaporation to yield a dark
yellow oil. This oil was subjected to column chroma-
tography on silica gel using CH2Cl2: diethyl ether
ꢀ95:5) as eluent and yielded 9.0 g ꢀ19.2%) of benzo-
11-crown-3 ether 1, as a pale yellow oil, which was
identi®ed by 13C NMR ꢀCDCl3 solvent) with
resonances at 29.4, 66.4, 69.5, 118.4, 122.6 and
stirring under N2. Re¯ux was carried out for 4 days.
Then 120 ml of methanol was added and the mixture
was re¯uxed for one more day. After cooling, 100 ml
of distilled water was added and the crude product
was extracted with CH2Cl2 ꢀ3 £ 200 ml). The
combined organic layers were dried over anhydrous
Na2SO4 and the solvent was removed by rotary
evaporation to yield 4.0 g of a dark brown oil. H
NMR indicated that this was a mixture of 1 and 2 in
a ratio of ca. 8:1.
The dark brown oil was boiled in hexane
ꢀ5 £ 400 ml) to dissolve the unreacted 1 and a
brown precipitate resulted. This precipitate was
collected by vacuum ®ltration and washed with a
further 300 ml of hot hexane. Subsequently, the
resulting brown powder was transferred to a beaker
and 30 ml of acetone was added to produce a light
yellow precipitate. This light yellow solid was then
dissolved in 40 ml of CH2Cl2 and boiled with
activated charcoal. After gravity ®ltration, the volume
of solution was reduced to ca. 10 ml by boiling and
then anhydrous diethyl ether ꢀ20 ml) was added. After
3 days, 0.007 g ꢀ0.5% yield) of 2, mp 285±2908C
were obtained as light yellow needles. 13C NMR
ꢀCDCl3):150.4, 124.9, 111.6, 69.8, 66.5 and
29.6 ppm. Mass spectrum ꢀM 1 1) m/z ˆ 619.
1
150.9 ppm.
calculated m/e 208.2573, observed 208.1085.
High-resolution
mass
spectrum:
For the preparation of 2, a 500 ml double-necked
round bottom ¯ask was equipped with a dropping
funnel and a condenser with a gas inlet tube. To a
mixture of FeCl3 ꢀ10.5 g, 0.065 mol), CCl4 ꢀ120 ml)
and concentrated H2SO4 ꢀ0.75 ml) in this ¯ask was
added 1, ꢀ4.5 g, 0.022 mol) drop wise with vigorous

G.W. Buchanan et al. / Journal of Molecular Structure 598 22001) 145±151
151
The low isolated yield of 2, ꢀ0.5%) is most likely
due to the dif®culty in freeing the trimeric material
from its complex with FeCl3. The crude yield is
estimated to be ca. 12% based on the above mentioned
1H NMR integration results.
completed with difference Fourier syntheses and
re®ned with full-matrix least-squares procedures
based on F2. All non-hydrogen atoms were re®ned
with anisotropic displacement parameters and
hydrogen atoms were treated as idealized contribu-
tions. All scattering factors and anomalous dispersion
factors are contained in the shextl 5.1 program
library.
3.2. 13C NMR spectra
13C solution spectra were recorded at 100.6 MHz
using a Bruker AMX 400 instrument with complete
1H noise decoupling. An Aspect 3000 processing
controller was employed and all standard micro
programs used are in the Bruker Software Library.
Acknowledgements
G.W. Buchanan thanks the Natural Sciences and
Engineering Research Council of Canada ꢀNSERC)
for continued ®nancial support via Research Grants.
We also thank Majid Rastegar for his invaluable
assistance in this work.
3.3. X-ray crystallographic studies and structure
re®nement
Suitable crystals were selected and mounted on thin
glass ®bres using viscous oil. Data were collected on a
Bruker AX SMART 1k CCD diffractometer using 0.38
scans at 0, 90 and 1808. Unit cell parameters were
determined from 60 data frames collected at different
sections of the Ewald sphere. Semi-empirical absorp-
tion corrections based on equivalent re¯ections were
applied [5].
References
[1] A.M. Levelut, J. Chim. Phys. 80 ꢀ1983) 149.
[2] S. Chandrasekhar, G.S. Ranganath, Rep. Prog. Phys. 53 ꢀ1990)
57.
[3] H. Naarmann, M. Hanack, R. Mattmer, Synthesis 477 ꢀ1994).
[4] G.W. Buchanan, M.F. Rastegar, G.P.A. Yap, Can. J. Chem.
ꢀ2001).
Unit cell parameters and systematic absences in the
diffraction data are consistent with the space group
Å
P1. The structure was solved by direct methods,
[5] R. Blessing, Acta Crystallogr. A51 ꢀ1995) 33.