Head-on versus side-on [5-5]bitrovacenes featuring benzene and naphthalene units as spacers: How π-stacking affects exchange coupling and redox splitting

Article abstract of DOI:10.1021/om020520y

The biradicals 1,8-di([5]trovacenyl)naphthalene (6), 1,5-di([5]trovacenyl)naphthalene (7), and 1,3-di([5]trovacenyl)benzene (8) and the monoradical 1-[5]trovacenylnaphthalene (9) have been prepared and studied by means of single-crystal X-ray diffraction

Full text of DOI:10.1021/om020520y

5
810
Organometallics 2002, 21, 5810-5819
Hea d -on ver su s Sid e-on [5-5]Bitr ova cen es F ea tu r in g
Ben zen e a n d Na p h th a len e Un its a s Sp a cer s: How
π-Sta ck in g Affects Exch a n ge Cou p lin g a n d Red ox
†
Sp littin g
Christoph Elschenbroich,* Matthias Wolf, Olav Schiemann, Klaus Harms,
Olaf Burghaus, and J u¨ rgen Pebler
Fachbereich Chemie der Philipps-Universit a¨ t, D-35032 Marburg, Germany
Received J uly 2, 2002
The biradicals 1,8-di([5]trovacenyl)naphthalene (6), 1,5-di([5]trovacenyl)naphthalene (7),
and 1,3-di([5]trovacenyl)benzene (8) and the monoradical 1-[5]trovacenylnaphthalene (9) have
been prepared and studied by means of single-crystal X-ray diffraction (6, 8, 9), cyclic
voltammetry, EPR spectroscopy, and magnetic susceptometry. Comparison of the magnetic
properties of 6-9 reveals that π-stacking, as encountered in 6, largely enhances exchange
2
interaction between the two singly occupied V(3d
z
) orbitals. The effect of π-stacking on the
electrochemical properties, as manifested in the redox splitting between consecutive electron
transfer steps, is less pronounced. Redox splittings δE1/2 for consecutive reductions exceed
those for oxidations of binuclear trovacenes, δE1/2(2+/+, +/0) being apparent for π-stacked
2
2
6
z z
only. Because of the orthogonality of the metal-centered redox orbitals V1(d ) and V2(d )
5
and the η -cyclopentadienyl π-orbitals, electro- and magnetocommunication are indirect
processes. Electrocommunication rests on changes of the donor/acceptor properties of
V1,V2 caused by oxidation or reduction which govern metal-ligand charge distribution; the
latter changes are transmitted via the spacer. Magnetocommunication takes the form of
antiferromagnetic coupling, which can be traced to spin polarization of filled π-orbitals of
2
the bridge by orthogonal singly occupied vanadium 3d
z
orbitals.
In tr od u ction
Spectroscopic and electrochemical manifestations of
asset of binuclear metallocenes of type 3 is the possibil-
ity to use central-metal related spectroscopic techniques
to gauge the extent of stacking interactions; the respec-
tive methods have recently been reviewed.
Our own contributions to the study of intramolecular
π-stacking have been explored for organic as well as for
7
1
-2
3
organometallic molecules, the compounds 1,
2, and
4
3
,4 may serve as representative examples. Remarkably,
communication are based on the use of trovacene
π-π interactions between stacked porphyrin rings and
their metal complexes, which may lead to dimer forma-
tion, have aroused considerable interest because of their
possible relation to enzyme function. The nature of π-π
7
[
(
(tropylium)vanadium(cyclopentadienyl), (η -C7H7)V-
5
1,8
η -C5H5)] 5 as a probe. Its suitability derives from
the orbitally nondegenerate ground state A1, the pres-
ence of the magnetic nucleus V (I ) 7/2) in high
2
51
5
interactions in general and of dimeric cofacial porphy-
abundance (99.75%), the attendant favorable EPR prop-
erties, comparatively low sensitivity to air, and the ease
of functionalization. Therefore, access to a large variety
of di- and oligonuclear species featuring diverse spacers
is envisaged. Here we report on the synthesis and
characterization of two regioisomers in which [5]tro-
6
rins in particular has been discussed extensively. An
†
Dedicated to Prof. G. Huttner on the occasion of his 65th birthday.
*
To whom correspondence should be addressed. E-mail: eb@
chemie.uni-marburg.de.
1) Trovacene Chemistry. 4. Part 3: Elschenbroich, C.; Wolf, M.;
Burghaus, O.; Harms, K.; Pebler, J . Eur. J . Inorg. Chem. 1999, 2173.
(
(
(
2) Gerson, F. Topics Curr. Chem. 1983, 115, 57.
3) (a) Boekelheide, V. Pure Appl. Chem. 1986, 58, 1. (b) Voegeli, R.
(6) (a) Scheidt, W. R.; Geiger, D. K.; Lee, Y. L.; Reed, C. A.; Lang,
G. J . Am. Chem. Soc. 1985, 107, 5693. (b) Gupta, G. P.; Lang, G.;
Scheidt, R. W.; Geiger, D. K. J . Chem. Phys. 1985, 83, 5945. (c) Scheidt,
W. R.; Lee, J . L. Struct. Bonding (Berlin) 1987, 64, 1. (d) Song, H.;
Reed, C. A.; Scheidt, W. R. J . Am. Chem. Soc. 1989, 111, 6867. (e)
Song, H.; Rath, N. P.; Reed, C. A.; Scheidt, W. R. Inorg. Chem. 1989,
28, 1839. (f) LeMest, Y.; L’Her, M.; Hendricks, N. H.; Kim, K.; Collman,
J . P. Inorg. Chem. 1992, 31, 835. (g) Fletcher, J . T.; Therien, M. J .
Inorg. Chem. 2002, 41, 331. (h) Gupta, G. P.; Lang, G.; Scheidt, W. R.;
Geiger, D. K.; Reed, C. A. J . Chem. Phys. 1985, 83, 5945.
H.; Kang, H. C.; Finke, R. G.; Boekelheide, V. J . Am. Chem. Soc. 1986,
08, 7010. (c) Schr o¨ der, A.; Mekelburger, H.-B.; V o¨ gtle, F. Top. Curr.
Chem. 1994, 172, 41. (d) Kang, H. C.; Plitzko, K.-D.; Boekelheide, V.;
Higuchi, H.; Misumi, S. J . Organomet. Chem. 1987, 321, 79.
1
(4) Lee, M.-T.; Foxman, B. M.; Rosenblum, M. Organometallics 1985,
4
, 539. (b) Arnold, R.; Foxman, B. M.; Rosenblum, M.; Euler, W. B.
Organometallics 1988, 7, 1253. (c) Herber, R. H.; Nowik, I.; Rosenblum,
M. Organometallics 2002, 21, 846.
(
5) (a) Misumi, S.; Otsubo, T. Acc. Chem. Res. 1978, 11, 251. (b)
Heilbronner, E.; Yang, Z. Top. Curr. Chem. 1983, 115, 1. (c) Hunter,
C. A.; Meah, M. N.; Sanders, J . K. M. J . Am. Chem. Soc. 1990, 112,
(7) Barlow, S.; O’Hare, D. Chem. Rev. 1997, 97, 637.
(8) (a) Elschenbroich, Ch.; Bilger, E.; Metz, B. Organometallics 1991,
10, 2823. (b) Elschenbroich, Ch.; Schiemann, O.; Burghaus, O.; Harms,
K. J . Am. Chem. Soc. 1997, 119, 7452. (c) Elschenbroich, Ch.;
Schiemann, O.; Burghaus, O.; Harms, K.; Pebler, J . Organometallics
1999, 18, 3273. (d) Elschenbroich, Ch.; Plackmeyer, J .; Harms, K.;
Burghaus, O.; Pebler, J . Organometallics, submitted.
5
773. (d) Cozzi, F.; Cinquini, M.; Annunziata, R.; Siegel, J . S. J . Am.
Chem. Soc. 1993, 115, 5330. (e) Hunter, C. A. Chem. Soc. Rev. 1994,
7, 101. (f) Hunter, C. A. J . Chem. Soc., Perkin Trans. 2 2001, 651. (g)
2
Tsuzuki, S.; Honda, K.; Uchimaru, T.; Mikami, M.; Tanabe, K. J . Am.
Chem. Soc. 2002, 124, 104.
1
0.1021/om020520y CCC: $22.00 © 2002 American Chemical Society
Publication on Web 11/20/2002

Head-on versus Side-on [5-5]Bitrovacenes
Organometallics, Vol. 21, No. 26, 2002 5811
Sch em e 1
Sch em e 2
vacenyl units are linked to naphthalene in 1,8-(6) and
,5-(7) positions, respectively. Whereas 6 is a biradical
analogue of the stacked 1,8-dimetallocenylnaphthalenes
and 4, π-stacking is absent in 7. Obviously, a com-
bearing vanadium atoms are bridged by C5 units. The
mononuclear complex 9 was prepared as a reference
molecule with the aim of eventually detecting leakage
of spin density into the 1-naphthyl substituent. Fur-
thermore, 9 constitutes the standard against which the
redox potentials of the dinuclear complexes 6-8 should
be assessed.
1
3
parison of the electrochemical and magnetic properties
of 6 and 7 should reveal to what extent face-to-face
interactions between aromatic decks govern intramo-
lecular communication. To this end, the measurement
of exchange coupling J has been applied only rarely,
X-r a y Cr ysta llogr a p h y. The solid-state structures
of compounds 6, 8, and 9 are displayed in Figures 1-3;
selected bond distances and angles are shown in the
captions. The exceedingly low solubility of 7 in all
common solvents prevented the growth of crystals
suitable for a diffraction study. The dimensions of the
trovacenyl groups in 6, 8, and 9 do not present notice-
able differences from that of parent 5. With regard to
1-[5]trovacenylnaphthalene 9 the only structural pa-
rameter worth mentioning is the dihedral angle Θ(C H /
+
inter-heme spin coupling in {[Fe(OEP)(2-MeHIm)] }2
dimers6h constituting an example (OEP ) octaethylpor-
phyrinate; 2-MeHIm ) 2-methylimidazole).
Resu lts a n d Discu ssion
The new trovacene derivatives 6-9 were prepared via
the Negishi-Kumada route9 (Scheme 2). Since the
isomers 6 and 7 differ in the number of spacer atoms,
5
4
C H ) ) 79.3°, which largely exceeds the corresponding
1
0
7
1
0
we also synthesized 1,3-di([5]trovacenyl)benzene 8,
(9) Negishi, E. I. Acc. Chem. Res. 1982, 15, 340. Negishi, E. I. In
2
which, like 6, features a (C, sp )3 unit linking the two
trovacenyl cores. Alternatively, if the ipso-carbon atoms
of the cyclopentadienyl rings are included in the count,
the view may be taken that in 6 and 8 the two spin-
Organozinc Reagents. A Practical Approach; Knochel, P., J ones, P.,
Eds.; Oxford University Press: Oxford, U.K., 1999; Chapter 11.
(10) The numbers in brackets indicate the site of substitution: [5]-
trovacenyl is functionalized at the cyclopentadienyl and [7]trovacenyl
at the cycloheptatrienyl ligand.

5
812 Organometallics, Vol. 21, No. 26, 2002
Elschenbroich et al.
F igu r e 1. Molecular structure of 1,8-di([5]trovacenyl)naphthalene (6) in the crystal (SHAKAL drawings, numbering scheme,
and stereoplot of the unit cell). Selected bond lengths (Å) and bond angles (deg): V1-C1 2.306(2); V1-C2 2.269(2); V1-C3
2
.250(3); V1-C4 2.247(3); V1-C5 2.266(3); V1-C21 2.179(3); V1-C22 2.182(3); V1-C23 2.178(3); V1-C24 2.192(3);
V1-C25 2.187(3); V1-C26 2.178(3); V1-C27 2.170(3); C-C(C
(centroid)-V1 1.923(1); C (centroid)-V1 1.466(1); dimensions of trovacene 2 are identical within the error margin;
intercentroid distance C (trovacene 1)‚‚‚C (trovacene 2) 3.432; dihedral angles C (trovacene 1), naphthalene (best plane)
2.5; C (trovacene 2), naphthalene (best plane) 40.5.
5
, mean) 141.1(4); C-C(C
7
, mean) 139.9(4); C1-C6 148.6(3);
C
5
7
5
5
5
4
5
angle Θ(C5H4/C6H5) ) 24.2° in [5]trovacenylbenzene.1
Not surprisingly, the structure of 6 largely resembles
its congener 3. It may be described as two stacked but
the Cp-ipso carbon atoms C1 and C16 from the best
naphthalene plane amount to 0.54 and -0.53 Å, respec-
tively. This translates into an angle of 31° between the
vectors C1-C6 and C14-C16. Yet, whereas the tro-
vacene units in 6 deviate markedly from an ideal face-
to-face disposition, considerable stacking interaction
nonetheless exists, which will be used in the sequel to
rationalize the physical data.
4
a
laterally slipped trovacene units which are linked by a
1
,8-naphthalenediyl lever. The twist angle between the
cyclopentadienyl(Cp) and naphthalene(Naph) planes
amounts to 40.5° compared with 42.5°; it is somewhat
smaller than that in 3 (45.0° compared with 47.0°). Since
the immediate environment at the Cp-Naph junction is
identical for 3 and 6, the small difference in the dihedral
angles must be traced to packing effects. The mean
1,3-Di([5]trovacenyl)benzene 8 in the solid state adopts
a syn conformation approaching an ideal side-on dispo-
sition of the sandwich units (Figure 2). The bridging
5
5
interplanar distance of the η -Cp rings in 6 amounts to
benzene and the two η -Cp rings deviate only marginally
7
3
43 pm, and repulsive interactions cause sizable puck-
(7.4°) from coplanarity. The η -C7H7 rings in 8 display
ering of the bridging naphthalene unit as well as
deviations of the CNaph-CCp bond vectors from copla-
narity with the naphthalene frame: the deviations of
rotational disorder.
Red ox P r ocesses. All newly synthesized trovacene
derivatives were subjected to cyclic voltammetry; elec-

Head-on versus Side-on [5-5]Bitrovacenes
Organometallics, Vol. 21, No. 26, 2002 5813
F igu r e 2. Molecular structure of 1,3-di([5]trovacenyl)benzene (8) in the crystal (SHAKAL drawing, numbering scheme,
and stereoplot of the unit cell). Selected bond lengths (Å) and bond angles (deg): V1-C(C , mean) 2.256(4); V1-C(C
mean) 2.160(20); rotational disorder; C-C(C , mean) 1.413(2); C-C(C , mean) 1.407(20); C1-C13 1.479(5); C (centroid)-
V1 1.9098(6); C (centroid)-V1 1.4402(6); dimensions of trovacene 1A are identical within the error margin. V1‚‚‚V1A 7.2535-
11); dihedral angle C /C 7.4(2).
5
7
,
5
7
5
7
(
5
6
number of electrons transferred cannot be assessed
reliably. This derives from the fact that electrocommu-
nication, which is clearly discernible for the directly
8
c
8d
linked [5-5]bitrovacene and its [7-7]isomer, is at-
tenuated in the case of spacered bitrovacenes. Further-
more, the reduction potentials E1/2(0/-) are very similar
for trovacene (-2.55 V) and for naphthalene (-2.48 V
vs SCE). Secondary and tertiary reductions close to the
cathodic limit of the medium also distort the shapes of
the waves and complicate analyses. Therefore, only
qualitative considerations will be offered.
Metal-centered primary oxidations of the complexes
7
, 8, and 9 give rise to unresolved single waves. Only
in the case of the stacked binuclear complex 6 is the
presence of two overlapping waves discernible, differing
by δE1/2(2+/+, +/0) e 70 mV. Clearly, therefore, the
head-on disposition of two trovacene units in 6 effects
somewhat more extensive electrocommunication than
the side-on arrangement in 7. The gradation is surpris-
ingly small, however. Although the one-electron oxida-
tions for 6-9 all take place in a 50 mV bracket, a
consistent interpretation of the individual values may
be proposed. Accordingly, the anodic shift of the poten-
tial E1/2(+/0, 9), relative to E1/2(+/0, 5), may be traced
to electron delocalization into the naphthyl substituent,
whereby electron density in the trovacenyl unit is
somewhat decreased. The same effect is operative in 7.
In 9 and 7, rotation about the naphthyl-trovacene bond
F igu r e 3. Molecular structure of 1-[5]trovacenylnaphtha-
lene (9) in the crystal (SHAKAL drawing and numbering
scheme). Selected bond lengths (Å) and bond angles (deg):
V-C(C
mean) 142.0(4); C-C(C
dihedral angle C /naphthalene 79.3.
5
, mean) 226.5(3); V-C(C
7
, mean) 218.5(3); C-C(C
5
,
7
, mean) 140.6(6); C1-C6 148.7(4);
5
trochemical traces are depicted in Figure 4, and the
corresponding data are found in Table 1. Due to con-
secutive redox processes of small potential differences,
the degree of electrochemical reversibility and the

5
814 Organometallics, Vol. 21, No. 26, 2002
Elschenbroich et al.
F igu r e 4. Cyclic voltammograms for 1-[5]trovacenylnaphthalene (9), 1,5-di([5]trovacenyl)naphthalene (7), 1,3-di([5]-
trovacenyl)benzene (8), and 1,8-di([5]trovacenyl)naphthalene (6). For the electrochemical data see Table 1.
is only slightly hindered; therefore, conformations con-
ducive to π-conjugation are feasible. Contrarily, for the
π-stacked complex 6, steric hindrance will prevent the
trovacenyl units from adopting a conformation in which
the cyclopentadienyl and naphthyl π-electron systems
are coplanar. It may be argued that for 6 the twisted
form found in the crystal is caused by packing forces
and that, in fluid solution, a conformation in which the
naphthyl and cyclopentadienyl rings are orthogonal will
be favored. Transfer of π-electron density into the
naphthyl unit is then hampered, and 5 and 6 are
oxidized at almost identical potential. The argument put

Head-on versus Side-on [5-5]Bitrovacenes
Organometallics, Vol. 21, No. 26, 2002 5815
Ta ble 1. Cyclovolta m m etr ic Da ta for Tr ova cen e 5
a
a n d Its Der iva tives 6-9
5
6
7
8
9
E1/2(+/0)/V
0.26
64
1
0.24
57
≈0.9
0.30
61
0.31
68
0.9
0.31
68
0.9
0.99
1.36
-2.30
64
0.27
71
1
0.27
71
1
0.31
76
1
∆
Ep/mV
r
E1/2(2+/+)/V
∆
Ep/mV
r
≈0.9
Epa,1/V
Epa,2/V
1.0
1.0
1.0
1.47
-2.32
66
E1/2(0/-)/V
-2.55
66
-2.43
63
-2.35
57
∆Ep/mV
a
DME/(n-Bu)4NClO4, -40 °C, 100 mV/s, versus SCE. r ) ia/ic
ratio of peak currents). Additional peak potentials Ep (V) (revers-
(
ibility criteria not applicable): 6, Epc ) -2.65 (large peak current);
Epa ) -2.59, -2.46. 7, Epc ) -2.60 (large peak current), -2.78;
Epa ) -2.47, -2.70. 8, Epc ) -2.59; Epa ) -2.48. 9, Epc ) -2.68
(large peak current); Epa ) -2.53.
forward to explain the anodic shift of E1/2(+/0, 7) also
applies to 8; due to the reduced electron-accepting
capacity of benzene, compared to naphthalene, the
cathodic shift of E1/2(+/0, 8) is only marginal, however.
An interpretation of the electrochemical data for
reductions is impeded by the aforementioned problem
in assigning the close lying waves to specific ET-steps
(metal- or bridge-centered?). Reduction of the separate
building blocks occurs at E1/2(0/-, trovacene 5) ) -2.55
V and E1/2(0/-, naphthalene) ) -2.48 V. Consideration
of the electron-donating nature of 5 and the electron-
accepting property of naphthalene and of the anodic
shift of E1/2(-/0) for all derivatives containing both units
suggests that the first reduction step is trovacene-
centered. In the case of 9 the second wave must then
be assigned to naphthalene reduction, the cathodic shift
relative to free naphthalene being caused by the
F igu r e 5. EPR spectra of 1,8-di([5]trovacenyl)naphthalene
(6) in fluid and rigid solution (toluene) and their simulated
-
traces; parameters are collected in Table 2.
trovacene substituent. Why the cathodic peak current
seems to imply a 2e process here is not clear in view of
the fact that 9 contains one naphthalene and one
trovacene unit only. In the cathodic regime for 7, three
cathodic and three anodic peaks separated by ∼200 mV
are barely resolved. Whereas the first reduction prob-
ably involves the vanadium atom, assignment of the
second and third step is impractical. Very similar redox
behavior is displayed by 6, while the first reduction step
for 8 exhibits a cathodic shift of 110 mV relative to 6
and 7. Thus, for 6 and 7 two effects appear to cancel
each other: 6 features a shorter intermetal distance and
a π-stacking contact but impaired π-conjugation via the
naphthalenediyl bridge because of near orthogonality.
In 7, on the other hand, the intermetal distance and
compared directly because of differing frontier orbital
4
1
characteristics. Whereas trovacene (e2 a1 ) possesses a
2
SOMO V(3dz ) practically devoid of metal-ligand mixing
and therefore may undergo metal-centered oxidation
4
2
and reduction, ferrocene (e2g a1g ) upon oxidation yields
3
2
the configuration e2g a1g with considerable metal-
ligand mixing of the redox orbital, which explains the
larger redox splitting in the oxidation of 3. Furthermore,
the 18 VE configuration of ferrocenes is not prone to
reduction, at least not within the “normal” electrochemi-
cal window (cathodic limit ≈ -3.0 V).
At this point one may ask how electrocommunication
between the two sites actually arises, in view of the fact
that the redox orbitals are metal-centered and nearly
orthogonal to the cyclopentadienyl π-electron systems
which form the links between the metals and the spacer.
Conceivably, changes in the occupation of the redox
2
the number of intervening sp -carbon atoms are larger,
but ease of rotation about the C(trovacene)-C(naph-
thalene) bonds contributes conformations that are fav-
orable to π-conjugation. In 8, relative ease of conforma-
2
2
2
tional change and the presence of a short (C, sp )3 unit
orbitals V1(dz ) and V2(dz ) and the attendant change
in local charge modify the donor/acceptor characteristics
of the central metals and therefore influence the metal-
ligand charge distribution, in this way providing a way
by which redox processes at V1 may modulate redox
processes at V2. This is an indirect mechanism, how-
ever, and therefore redox splittings are expected to be
small. This contrasts with the considerably larger redox
splittings that are encountered when the metal redox
orbitals possess correct symmetry for d(π)-p(π) overlap
separating the trovacenes are combined with the effect
that the reduction of the first trovacene moiety is
rendered somewhat more difficult by the presence of the
second electron-releasing trovacene.
It remains to emphasize that the redox properties of
4
a,12
spacered diferrocenes
and of ditrovacenes cannot be
(
11) McCleverty, J . A.; Ward, M. D. Acc. Chem. Res. 1998, 31, 842.
Campagna, S.; Giuffrida, G. Coord. Chem. Rev. 1994, 136, 517.
12) Patoux, C.; Coudret, C.; Launay, J .-P.; J oachim, C.; Gourdon,
A. Inorg. Chem. 1997, 36, 5037.
(
1
1
with the bridging ligand.

5
816 Organometallics, Vol. 21, No. 26, 2002
Elschenbroich et al.
F igu r e 6. EPR spectra of 1,5-di([5]trovacenyl)naphthalene
7) in fluid and rigid solution (toluene) and their simulated
traces; parameters are collected in Table 2.
F igu r e 7. EPR spectra of 1,3-di([5]trovacenyl)naphthalene
(
(8) in fluid and rigid solution (toluene) and their simulated
traces; parameters are collected in Table 2.
EP R Sp ectr a a n d Ma gn etic Su scep tibility. Ex-
change interactions can be studied on a molecular level
by EPR and in the bulk by magnetic susceptometry.
Apart from the vastly higher sensitivity of the former,
the two methods differ in that EPR describes the
magnetic properties of individual molecules, eventually
reflecting their conformational freedom in the absence
of intermolecular interactions, whereas magnetic sus-
ceptibility deals with the conformation dictated by
packing forces and may also have to consider intermo-
lecular magnetic exchange.
Ta ble 2. EP R Da ta for Tr ova cen e 5 a n d th e
a
Der iva tives 6-9
a(51V)/mT |J |/cm-1
-6.98
T/K
〈g〉
b
c
d
5•
295 1.987
9• 295 1.980
6••
-7.22
250 1.9784
-7.17
-7.17
-7.18
2.15
2.22
2.16
-0.580
-0.415
-0.301
0.298 -0.285
0.060 -0.272
0.0056 -0.202
2
3
95 1.9789
30 1.9788
7••
240 1.9800
90 1.9782
1.9805
-7.23
-7.22
-7.22
0.72
0.63
0.44
-1.685
-0.608
0.029
0.056
-0.413 -0.08
1.89
1.31
1.07
2
Temperature-dependent EPR spectra of 6-8 in fluid
solution and their simulations are shown in Figures
3
0
5
-7, and the EPR parameters are collected in Table 2.
8•• 250 1.982
-7.20
-7.20
-7.20
0.61
0.48
0.35
-1.99
0.12
-0.82 -0.11
-0.37 -0.14
0.70
0.55
0.24
The coefficients a-d of the line width equation
295 1.982
3
60 1.982
2
a
∆
B ) a + b[m (1) + m (2)] + c[m (1) + m (2)] +
Solvent: toluene. a, b, and c are the coefficients in the line
I
I
I
I
width equation which, in the spectral simulations, allow for m(-
dependence of the line widths.
2
d[m (1) - m (2)]
I
I
are reproduced in Figure 8, and the parameters are
given in the captions. The interaction is antiferromag-
netic in all cases; exchange coupling, as extracted from
the EPR spectra recorded in fluid solution at ambient
temperature, follows the gradation
are also given. This relation treats the dependence of
the line widths on the anisotropies of the g-tensor and
the hyperfine interaction (a, b, c), and its consideration
is indispensable for a satisfactory fit of the experimental
spectra. Fluctuations of the exchange interaction J may
give rise to alternating line widths. This effect is
1
3
J (6, |2.22|) > J (7, |0.63|) ≈ J (8, |0.48|) [cm-1]
handled by inclusion of the parameter d.
-
1
Plots of ø versus temperature for 6-8 and the fits
of the Bleaney-Bowers equation
Compound 6 stands out in possessing a π-stacked
structure to which the comparatively large value of J
therefore must be attributed. The interpretation is
2
2
2
N µ g 3 exp(2J /kT)
A B
T
ø )
‚
3
kT[1 + 3 exp(2J /kT)] (T - Θ)
(13) Mabbs, F. E. Chem. Soc. Rev. 1993, 22, 313.

Head-on versus Side-on [5-5]Bitrovacenes
Ta ble 3. Cr ysta llogr a p h ic Da ta for Com p ou n d s 6 ‚C
6•• 8••
thin plate, dark
Organometallics, Vol. 21, No. 26, 2002 5817
•
•
••
•
7
8
H , 8 , a n d 9
‚C7H8
9•
habitus, color
thin plate, red-green
0.16 × 0.15 × 0.07
monoclinic
P21/c, Z ) 4
a ) 18.2179(17)
b ) 20.6897(12)
c ) 8.3505(8)
â ) 103.040(11)
3066.3(4) Å3
5000 reflns
C41H36V2
nugget, violet
0.21 × 0.15 × 0.09
orthorhombic
Fdd2, Z ) 16
a ) 13.5510(10)
b ) 67.397(6)
c ) 6.9222(6)
3
cryst size [mm ]
cryst syst
space group
0.50 × 0.30 × 0.03
orthorhombic
Pnma, Z ) 8
a ) 10.5710(10)
b ) 24.873(3)
c ) 8.5640(10)
unit cell [Å, deg]
volume
2251.8(4) Å3
25 reflns
C30H26V2
488.39
1.441
0.846
6322.0(9) Å3
5000 reflns
C22H18V
333.30
1.401
0.624
cell determination
empirical formula
fw
density(calcd) [Mg/m ]
abs coeff [mm
F(000)
630.58
1.366
0.638
1312
3
-
1
]
1008
2768
θ range [deg]
2.28 to 25.92
-22/22, -25/23, -10/10
21 855
5922 [0.0604]
99.1
2.52 to 25.00
0/12, -29/0, 0/10
2031
2030 [0.0006]
89.0
2.42 to 26.00
-16/16, -82/82, -8/8
9397
3090 [0.0460]
99.7
index ranges
no. of collected reflns
no. of ind reflns [R(int)]
completeness [%]
observed reflns [I>2σ(I)]
no. of reflns for refinement
abs corr
3480
5922
analytical
1258
2030
2625
3090
Tmax and Tmin
0.9458 and 0.8977
0.293 and -0.215
located, isotropic
refinement
532
largest diff peak and hole [e Å-3]
treatment of hydrogen atoms
0.306 and -0.362
calculated positions,
fixed isotropic U’s
210
0.280 and -0.304
located, isotropic
refinement
280
no. of params
Flack param (abs struct)
wR2 (all data)
R1 [I>2σ(I)]
0.00(3)
wR2 ) 0.0807
R1 ) 0.0330
wR2 ) 0.0651
R1 ) 0.0340
wR2 ) 0.1132
R1 ) 0.0470
based on the premise that the singly occupied a1 orbital
reasons spin transfer into the stacked π-ligands is
vanishingly small. Rather, spin polarization of filled
π-orbitals generated in the stacking interaction between
2
in trovacene 5 is a function of almost exclusive V(3dz )
1
4
character. This is because the basis orbitals of the
2
2
benzene π set lie in the nodal region of the V(dz ) orbital,
the Cp rings, effected by the singly occupied V(dz )
orbitals, can lead to antiferromagnetic coupling.18 In
Figure 9 only the linear combination formed from the
and metal-ligand orbital overlap therefore vanishes.
2
Direct interaction between the two V(dz ) orbitals in
-
binuclear 6 can also be neglected because of the large
metal-metal distance of almost 700 pm. Therefore, an
indirect coupling path must be devised. A mechanism
that comes to mind is sketched in Figure 9.
a2u C5H5 π MOs is shown; the MOs from interactions
-
between e1g orbitals of C H will be spin polarized as
5
5
well, thereby generating an additional coupling path.
The coupling mechanism proposed for 6 gains plausibil-
ity from a comparison with the binuclear complex 8. In
both 6 and 8, the trovacenyl units are separated by a
Nonconjugated π-sytems which are forced into close
proximity are encountered in the cyclophanes such as
the prototypical 1; transannular through-space π-π
interactions will result in the generation of bonding and
2
3
(C sp ) chain; they differ in the accessible conforma-
tions, however. Whereas for 6 only π-stacked forms are
attainable, in 8 rotation about the C(trovacene)-
C(arene) bonds is sterically hindered only to a small
extent, enabling a syn conformation with side-on dis-
1
5
antibonding sets. Covalent interactions between stacked
π-systems manifest themselves in the EPR spectra of
cyclophane radical ions, which point at delocalization
1
6,17
5
of the unpaired electron over both arene moieties.
posed trovacenes and coplanarity of the η -cyclopenta-
2
This type of communication is inoperative in the V(dz )-
Cp(π)/Cp(π) V(dz ) unit, though, since for symmetry
dienyl and the bridging m-phenylene units to be real-
ized. Nevertheless, J (8), derived from EPR in fluid
solution, falls short of J (6) to a considerable extent. For
2
(
(
14) (a) Rettig, M. F.; Stout, C. D.; Klug, A.; Farnham, P. J . Am.
2
the reason given for 6 (orthogonality of V(dz ) and the
Chem. Soc. 1970, 92, 5100. The M(dz2) character of the a1g frontier
ligand π-orbitals), antiferromagnetic coupling in 8 may
equally be traced to spin polarization of filled ligand
π-orbitals. This effect is less pronounced for 8, however,
since it is transmitted via the bottleneck of the C(tro-
vacene)-C(arene) bond, whereas in 6, due to the face-
orbital for sandwich complexes in general is born out by numerous
experimental studies as well as by quantum chemical calculations: (b)
Muetterties, E. L.; Bleeke, J . R.; Wucherer, E. J .; Albright, T. A. Chem.
Rev. 1982, 82, 499. (c) Cloke, F. G. N.; Dix, A. N.; Green, J . C.; Perutz,
R. N.; Seddon, E. A. Organometallics 1983, 2, 1150. Green, M. L. H.;
Ng, D. K. P. Chem. Rev. 1995, 95, 439. (d) Andrews, M. P.; Mattar, S.
M.; Ozin, G. A. J . Phys. Chem. 1986, 90, 1037. (e) Pandey, R.; Rao, B.
K.; J ena, P.; Blanco, A. M. J . Am. Chem. Soc. 2001, 123, 3799. (f) Lein,
M.; Frunzke, J .; Trimoshkin, A.; Frenking, G. Chem. Eur. J . 2001, 7,
5
to-face orientation of the η -cyclopentadienyl rings, all
carbon pπ-orbitals contribute.
4
155.
15) Gleiter, R.; Sch a¨ fer, W. Acc. Chem. Res. 1990, 23, 369. Gleiter,
R.; Kratz, D. Acc. Chem. Res. 1993, 26, 311.
16) Gerson, F.; Martin, W. B., J r. J . Am. Chem. Soc. 1969, 91, 1883.
Gerson, F. Top. Curr. Chem. 1983, 115, 57.
(
(18) Supportive evidence for polarization of the filled MO e2g by the
5
(
singly occupied orbital a1g in a (arene)
2
M(d ) unit is provided by a
comparison of the sign and magnitudes of the hyperfine coupling
1
6
•+
(
17) Interestingly, electron absorption and EPR spectroscopy has
constants a( H, CH
2
) in [bis(cyclobuta-η -benzene)chromium] and
•
+
12
•+
revealed the formation of the dimer radical cations (benzene)
2
and
[(η -[2.2]paracyclophane)chromium] and an interpretation in terms
of conformation-dependent and conformation-independent contribu-
tions: Elschenbroich, Ch.; Koch, J .; Schneider, J .; Spangenberg, B.;
Schiess, P. J . Organomet. Chem. 1986, 317, 41.
•+
2
, attesting to bonding interaction between stacked
(
naphthalene)
π-perimeters: Badger, B.; Brocklehurst, B. Trans. Faraday Soc. 1970,
6, 2939. Lewis, I. C.; Singer, L. S. J . Chem. Phys. 1965, 43, 2712.
6

5
818 Organometallics, Vol. 21, No. 26, 2002
Elschenbroich et al.
remains to comment on the low values of J for 7. For
this positional isomer of 6, J obtained from EPR is small
and in susceptometry too small to be determined. In the
spirit of the aforesaid, this is plausible because in 7
π-stacking of the trovacenyl units is impossible for
5
geometric reasons and conformations in which the η -
cyclopentadienyl and the naphthalene π-systems are
coplanar would suffer from severe strain caused by
compression of the cyclopentadienyl(C-H)ortho and the
naphthalene(C-H)peri bonds.
Interestingly, conformational preferences also seem
to govern the temperature dependence of the fluid
solution EPR spectra of 6-8. Whereas for the fairly rigid
complex 6, temperature dependence is insignificant, 7
and 8 exhibit an increase in J by 63% and 74%,
respectively, upon lowering the temperature by 100 °C.
This points at the population, at lower temperatures,
of conformations with smaller twist angles Θ and
increased π-overlap, the limit of Θ ) 0 being reached
by 8 in the crystal. Since J ø(7) was found to be very
small, it is tempting to predict that 7 in the crystal will
assume a highly twisted structure.
By way of conclusion it may be stated that the
influence the mutual disposition of two paramagnetic
trovacene unitsshead-on (6) or side-on (8)sexerts on
the degree of interaction depends on the physical
mechanism of interaction. Whereas π-stacking, as en-
countered in 6, boosts exchange coupling J to a consid-
erable extent relative to nonstacked 8, redox-splittings
for 6 and 8 differ only marginally. Thus, magnetocom-
munication responds more sensitively to structural
changes here than electrocommunication.
Exp er im en ta l Section
F igu r e 8. Temperature dependence of the inverse mag-
netic susceptibility ø for 6, 7, and 8. For parameters and
fitting procedures see text.
Chemical manipulations and physical measurements were
carried out using techniques and instruments specified previ-
-
1
ously.1 Trovacene 5 was prepared from (C
H
)V(CO)
19a
ac-
1,5-
5
5
4
1
9b
cording to the procedure given in the literature.
2
0
20
Diiodonaphthalene, 1,8-diiodonaphthalene, and the catalyst
2
1
Pd(dppf)Cl
1,1′-bis(diphenylphosphino)ferrocene). Anhydrous ZnCl
generated by degassing molten ZnCl ‚4H O in vacuo. Cyclic
voltammetry was performed under Ar protection in dimeth-
oxyethane as a solvent, (n-Bu) NClO serving as supporting
2
were also obtained as described previously (dppf
)
2
was
2
2
4
4
electrolyte. A glassy carbon working electrode, a Pt counter
electrode, and SCE reference electrode were employed.
+
/0
E
2
1/2(Cp Fe ) ) 0.49 vs SCE under these conditions.
1
•
•
,8-Di([5]tr ova cen yl)n a p h th a len e (6 ). To a solution of
trovacene (5; 435 mg, 2.1 mmol) in 130 mL of diethyl ether
was added at room temperature 2.1 mL of a solution (1.55 M)
of n-butyllithium in hexane. After stirring the mixture for 14
F igu r e 9. Antiferromagnetic coupling between two singly
occupied V(3dz2) orbitals, mediated by a π-stacked pair of
h, 3.7 mL of a solution (0.9 M) of anhydrous ZnCl
2
in
cylopentadienyl rings with vanishing V(3dz2)/Cp(a2u) over-
lap as proposed for 1,8-di([5]trovacenyl)naphthalene (6).
tetrahydrofuran was added. Transmetalation was effected
during 30 min, and 1,8-diiodonaphthalene (198 mg, 0.51 mmol)
2
and Pd(dppf)Cl (74 mg, 0.1 mol) in THF were added. The
Somewhat disturbingly, the gradation of J -values is
reversed if, instead of EPR spectroscopy, the data from
mixture was stirred at room temperature for 6 h and finally
refluxed for 30 min. After removing the solvent in vacuo, the
product was dissolved in toluene and chromatographed over
magnetic susceptometry are considered: J ø(6) ) -1.04,
J ø(8) ) -1.66 cm-1. Note, however, that the results refer
Al O (0% H O, column 2 × 15 cm). Unreacted trovacene was
2
3
2
followed by a green zone, from which 6 was obtained. Yield:
to differing states of aggregation. 6 in the solid experi-
ences considerable slippage with attendant decrease of
π-stacking interaction, whereas 8 forms a structure for
which π-conjugation is maximal. This conformation, in
which the two cyclopentadienyl and the meta-phenylene
rings are coplanar, is tolerated by the small C-H
compression strain inherent in this connectivity. It
(19) (a) Floriani, C.; Mange, V. Inorg. Synth. 1991, 28, 263. Floriani,
C. J . Chem. Soc., Dalton Trans. 1976, 1046. (b) King, R. B.; Stone, F.
G. A. J . Am. Chem. Soc. 1959, 81, 5263.
(
20) House, H.; Koepsell, D. G.; Campbell, W. J . J . Org. Chem. 1972,
7, 1003.
21) Hayashi, T.; Konishi, M.; Kumada, M. Tetrahedron Lett. 1979,
21, 1871.
3
(

Head-on versus Side-on [5-5]Bitrovacenes
75 mg (0.33 mmol, 31%) of red-green dichroic crystals. MS
Organometallics, Vol. 21, No. 26, 2002 5819
1
0.013 mmol) in 20 mL of THF. After refluxing for 1 h, the
solvent was removed in vacuo and the residue taken up in 5
mL of toluene. Chromatography over Al O (0% H O, column
2 3 2
+
+
(
C
EI, 70 eV): m/z (relative intensity) 538 (M , 87.3), 447 (M -
2
+
+
7
H
7
, 100), 269 (M , 23), 142 (C
7
H
7
V , 54.7). MS (high
1
2
resolution] calcd 538.10703, found 538.10747 ( C). IR (KBr):
2.5 × 30 cm) effected the separation of unreacted 5 (first
zone, elution by benzene) from the product 8 (second zone,
elution by benzene/THF, 15:1). Recrystallization from THF at
0 °C afforded 8 as violet platelets. Yield: 158 mg (0.32
3
1
081, 3044(w), 1760-1580(w) 1500-1310(w), 1261, 1091, 1054,
031, 1015(m), 823, 802, 779, 762(m), 488, 451, 432, 418(w).
Anal. Calcd for C34
28 2
H V (538.48): C, 75.84; H, 5.24. Found:
+
C, 75.21; H, 5.13. For EPR and CV see text.
mmol, 40%). MS (EI, 70 eV): m/z (relative intensity) 488 (M ,
•
•
2+
+
1
,5-Di([5]tr ova cen yl)n a p h th a len e (7 ). By a procedure
100), 244 (M , 15), 51 (V , 10). IR (KBr): 3041(w), 1700-
identical to that given for 6, except for the use of 1,5-
diiodonaphthalene instead of the 1,8-isomer, 7 was obtained
in 16% yield as a green, microcrystalline product. MS (EI, 70
1638(w), 1602(m), 782(s), 491(w), 468(w), 442(m), 424(m). Anal.
Calcd for C30
H, 5.29.
26 2
H V (488.42): C, 73.77; H, 5.37. Found: C, 73.65;
+
2+
eV): m/z (relative intensity) 538 (M , 100), 269 (M , 23.2).
••
••
•
7 8
X-r a y Cr yst a llogr a p h ic St u d y of 6 ‚C H , 8 , a n d 9 .
Anal. Calcd for C34
C, 74.91; H, 5.72.
H
28
V
2
(538.48): C, 75.84, H, 5.24. Found:
Data were collected at 193 K on a STOE IPDS or ENRAF
NONIUS CAD4 (8 ) diffractometer using Mo KR radiation. The
structures were solved using direct methods and refined on
••
•
1
-[5]Tr ova cen yln a p h th a len e (9 ). The preparation of 9
followed the instructions given for 6. From trovacene (280 mg,
.35 mmol), n-butyllithium (1.4 mL of a 1.55 M solution in
hexane), ZnCl (0.8 mL of a 0.9 M solution in THF), 1-iodo-
naphthalene (170 mg, 0.67 mmol), and Pd(dppf)Cl (51 mg,
.07 mmol) a product mixture was obtained, which was
separated by column chromatography (Al , 0% H
O, 2 × 25
cm). Elution by means of 1:1 toluene/petroleum ether delivered
in the second zone. Layering a solution of 9 in toluene with
n-pentane produced red-green dichroic cubes. Yield: 189 mg
2
F
values using the full matrix least-squares method. For
1
••
compound 8 a 2-fold disorder of the seven-membered rings
2
has been refined.
2
Programs used: SHELXS-97 and SHELXL-97 (G. M. Sheld-
rick, University of G o¨ ttingen, Germany, 1997), STOE IPDS
software (STOE & CIE GmbH, Darmstadt, Germany, 1999),
CAD4 EXPRESS (Enraf Nonius, Delft, The Netherlands,
994), SHELXTL 5.02 (Siemens Analytical X-ray Instruments
Inc., Madison, WI, 1996).
0
2
O
3
2
9
1
(
0.57 mmol, 42%). MS (EI, 70 eV): m/z (relative intensity) 333
+
2+
+
(M , 100), 166 (M , 3.4), 51 (V , 14). MS (high resolution):
1
2
Ack n ow led gm en t. This work has been supported
by the Volkswagen Foundation and the Fonds der
Chemischen Industrie.
calcd 333.08436, found 333.08363 ( C). IR (KBr): 3101, 3089-
w), 2956(m), 1765-1600(w), 1591(w), 1393(m), 1052(s), 796,
88, 774(s), 470, 439, 419(w). Anal. Calcd for
333.33): C, 79.27; H 5.44. Found: C, 78.18; H 5.80.
(
7
(
22 18
C H V
•
•
1
,3-Di([5]tr ova cen yl)ben zen e (8 ). Trovacene (300 mg,
Su p p or tin g In for m a tion Ava ila ble: Tables giving crys-
tal data and details of the structure determinations, positional
and thermal parameters, and all bond distances and angles
for 6, 8, and 9. This material is available free of charge via
the Internet at http://pubs.acs.org.
1
.4 mol) in 50 mL of diethyl ether was lithiated during 10 h
at room temperature by 0.9 mL of a solution of n-butyllith-
ium (1.6 M) in hexane. Transmetalation was effected by
addition of 1.6 mL of a solution (0.9 M) of anhydrous ZnCl
THF. To the resulting mixture was added a solution of 1,3-
diiodobenzene (115 mg, 0.35 mmol) and Pd(dppf)Cl (10 mg,
2
in
2
OM020520Y