Hierarchic supramolecular interactions within assemblies in solution and in the crystal of 2,3,6,7-tetrasubstituted 5,5′-(anthracene-9,10-diyl) bis[pyrimidin-2-amines]

Article abstract of DOI:10.1002/hlca.200690037

The title compounds 6 and 7 were synthesized in good yield (Schemes 1 and 2), and their mode of assembly was studied both in solution, for the tetrakis(decyloxy) derivative 6, and in the crystal, for the tetramethoxy analogue 7. The pyrimidin-2-amine moieties of 6 and 7 can engage in three different supramolecular interactions: i) metal ligation via one of the pyrimidine N-atoms, ii) cooperative double H-bonding via the NH₂ group, and iii) π-π-stacking interactions. In solution, coordination of the central Zn-atom within the soluble porphyrinatozinc complex 19 leads to significant changes in the NMR and absorption spectra of 6. In the absence of metal ligation, the next strongest interaction is H-bonding which can operate in nonpolar or moderately polar solvents. In these cases, however, no stacking interaction or inclusion compounds could be put into evidence in the case of 6 by absorption, fluorescence, or NMR spectroscopies. The π-stacking interactions were only observed in the crystal of 7 in conjunction with double H-bonding. Slightly disordered DMSO molecules are also H-bonded to the NH ₂ groups of 7, perturbing the expected packing. The present study illustrates some of the challenges inherent to directing hierarchical assembly processes in the solid state.

Full text of DOI:10.1002/hlca.200690037

Helvetica Chimica Acta – Vol. 89 (2006)
333
Hierarchic Supramolecular Interactions within Assemblies in Solution and in
the Crystal of 2,3,6,7-Tetrasubstituted 5,5’-(Anthracene-9,10-
1
diyl)bis[pyrimidin-2-amines] )
a
2
a
3
b
4
a c
5
by Teodor Silviu Balaban ) ), Andreas Eichhöfer ) ), Michael J. Krische ) ), and Jean-Marie Lehn ) ) )
a
)
Forschungszentrum Karlsruhe, Institute for Nanotechnology, Postfach 3640, D-76021 Karlsruhe
b
)
University of Texas at Austin, Department of Chemistry and Biochemistry,
Welch Hall, Austin, TX 78712, USA
c
)
ISIS, Université Louis Pasteur, 8 Allée Gaspard Monge, BP 70028, F-67083 Strasbourg
The title compounds 6 and 7 were synthesized in good yield (Schemes 1 and 2), and their mode of
assembly was studied both in solution, for the tetrakis(decyloxy) derivative 6, and in the crystal, for
the tetramethoxy analogue 7. The pyrimidin-2-amine moieties of 6 and 7 can engage in three different
supramolecular interactions: i) metal ligation via one of the pyrimidine N-atoms, ii) cooperative double
H-bonding via the NH group, and iii) p–p-stacking interactions. In solution, coordination of the central
2
Zn-atom within the soluble porphyrinatozinc complex 19 leads to significant changes in the NMR and
absorption spectra of 6. In the absence of metal ligation, the next strongest interaction is H-bonding
which can operate in nonpolar or moderately polar solvents. In these cases, however, no stacking inter-
action or inclusion compounds could be put into evidence in the case of 6 by absorption, fluorescence, or
NMR spectroscopies. The p-stacking interactions were only observed in the crystal of 7 in conjunction
with double H-bonding. Slightly disordered DMSO molecules are also H-bonded to the NH groups
2
of 7, perturbing the expected packing. The present study illustrates some of the challenges inherent to
directing hierarchical assembly processes in the solid state.
Introduction. – Pyrimidin-2-amine (2PA) is a versatile recognition group which has
been used to decorate tectons with the scope of programed self-assembly [1–4]. It can
simultaneously engage in three different supramolecular interactions: i) metal ligation,
ii) cooperative double H-bonding, and iii) p-stacking. Porphyrinatozinc complex 1
which has a single 2PA recognition group assembles into tetramers both in solution
and in the crystal (Fig. 1) [4]. Tetramer formation is directed by the meta-position of
the two endocyclic N-atoms of the pyrimidine moiety, one of which coordinates the cen-
tral Zn-atom within the tetrapyrrole core of an adjacent porphyrinato ligand. Within
the tetramer’s interior, although energetically less stabilizing, two 2PA groups prefer
to p-stack instead of engaging in double H-bonding. In the crystal, the two other
1)
2)
For previous papers about pyrimidin-2-amine (=2-aminopyrimidine) as a recognition group, see
1–4].
Materials and methods or general questions: fax: +497247828298; e-mail: silviu.balaban@
int.fzk.de.
Crystallographic issues: fax: +497247826368; e-mail: eichhoefer@int.fzk.de.
Pyrimidin-2-amine directed self-assembly: fax: +1512471 8696; e-mail: mkrische@mail.utexas.edu.
Other issues: fax: +33390241117; e-mail: lehn@chimie.u-strasbg.fr.
[
3
)
4
)
5)
©
2006 Verlag Helvetica Chimica Acta AG, Zürich

3
34
Helvetica Chimica Acta – Vol. 89 (2006)
2PA groups, which point to the tetramer’s exterior, lead to an infinite strand of tetram-
ers stitched together by double H-bonding as shown in Fig. 1. The generality of tet-
ramer formation was more recently established with the related porphyrinatozinc com-
plexes 2 [5] and 3 [6] which both have a meso-pyridin-3-yl group, i.e., a meta-positioned
N-atom. When two 2PA groups are bound to the same porphyrinatozinc as in 4, oligom-
ers are preferentially formed in solution instead of tetramers [4].
From the earlier work involving 2PA-directed self-assembly [1–3], it is known that,
in the absence of metal ligation, double H-bonding occurs in several crystal structures
in conjunction with extensive p-stacking interactions. These were used to from clath-
rates with various aromatic guests or solvent molecules, as indicated schematically in
Fig. 2 by the gray rectangles in the case of the 2PA 5,5’-(anthracene-9,10-diyl)bis[pyri-
midin-2-amine] (5) [3].
Fig. 1. Porphyrinatozinc Complexes 1–4 and doubly H-bonded tetramers of 1 in the crystal [4].
DTBP stands for 3,5-di(tert-butyl)phenyl which is an excellent solubilizing group introduced to por-
phyrin chemistry by Crossley and Burn [7].

Helvetica Chimica Acta – Vol. 89 (2006)
335
Fig. 2. Planar tape-like double H-bonding of 5,5’-(anthracen-9,10-diyl)bis[pyrimidin-2-amine] (5) used
to assemble ca. 7.0-Å voids which in the crystal can bind various aromatic guests such as phenazine [3]
Because the 2PA-directed self-assembly of the porphyrinatozincs 1 and 4 worked
well also in solvents of medium polarity such as CHCl , we wanted to learn whether,
3
in the absence of metal ligation, similar assemblies as that depicted in Fig. 2 are formed,
not only in the crystal but also in solution, from more soluble (anthracenediyl)bis[pyr-
imidin-2-amine] derivatives than 5. Thus, we report here the syntheses of the 5,5’-(an-
thracene-9,10-diyl)bis[pyrimidin-2-amines] 6 and 7 which are 2,3,5,6-tetrakis(decyl-
oxy)- and 2,3,5,6-tetramethoxy-substituted, respectively, and studies of their self-

3
36
Helvetica Chimica Acta – Vol. 89 (2006)
assembly which allow us now to draw a hierarchy in 2PA-mediated supramolecular
interactions.
Results and Discussion. – Syntheses. The synthetic sequence for the preparation of
the soluble long-chain derivative 6 is presented in Scheme 1. The condensation of vera-
trol (=1,2-dimethoxybenzene; 8) with acetaldehyde mediated by H SO was per-
2
4
formed as described earlier [8], while the oxidation of the Me groups in 9 with
Na Cr O in AcOH to the 2,3,6,7-tetramethoxy-9,10-anthraquinone (10) and the subse-
2
2
7
quent hydrolysis of the MeO groups were slightly improved as compared to the meth-
ods described by Boldt [9] (Scheme 1). We consistently obtained 11 in different runs
with more than 70% yield over two steps. The alkylation of all four OH groups in 11
with 1-bromodecane was unproblematic, while the reduction of the anthraquinone 12
proved to be more efficient if the intermediate anthracenone 13 was isolated, rather
than to attempt a direct reduction to the anthracene 14.
Scheme 1. Synthesis of 6

Helvetica Chimica Acta – Vol. 89 (2006)
337
The four long alkoxy whiskers at the anthracene skeleton impart special solubility
properties to compound 14. We especially chose decyl chains because two long-chain
decyloxy substituents at C(2) and C(3) of an anthracene skeleton are known to form
excellent and fluorescent gelators for organic solvents such as MeOH in minute quan-
tities (<10 mM) [10]. With two hexyloxy chains, the 2,3-disubstituted anthracene could
be crystallized with a unit cell composed of 18 monomers which apparently associate
into triads which pack into a head-to-tail fashion [11]. Recently, due to its good solubil-
ity, compound 14 was used as a guest taking advantage of its p-stacking ability for nano-
mechanical molecular tweezers which can be conformationally switched by metal coor-
ꢁ1
dination [12]. A binding constant of 1020ꢀ60 M could be measured between 14 and
an acridine tweezer in CDCl containing 8% CD OD. This value allows an estimation
3
3
of the p–p interactions of a donor-substituted anthracene in the absence of H-bonding
(vide infra, discussion and conclusions).
Dibromination of 14 and a double Stille condensation with the protected stannyl
derivative 16 [3] followed by a final deprotection with methylhydrazine completed
the reaction sequence. Over nine steps, starting from commercial veratrol (8), the
global yield of the desired product 6 with long chains was thus 15%. Gratifying was
that the optimized Pd-catalyzed double Stille coupling and phthalimide deprotection
[
(
[
3] functioned so well that the two final steps of the sequence were very high-yielding
91% over two steps) and thus accounted for a pronounced atom and catalyst economy
13].
In the somewhat shorter synthesis of the less soluble tetramethoxy derivative 7
Scheme 2) we found that the alkaline zinc reduction of the anthraquinone 10 pro-
(
ceeded as described [14] giving directly the anthracene 17 so that the isolation of the
intermediate anthracenone could be avoided. Again we ascribe this striking different
reaction progress to the wettability difference imparted by the MeO substituents in
17 in comparison to the decyloxy substituents in 14. However, in this case, the double
Stille coupling of 18 could not be driven to completion due to the low solubility so that a
rather tedious and low-yielding chromatographic purification of 7 had to be performed
to obtain single crystals (see Exper. Part).
Self-Assembly Studies in Solution. With compound 6 in hand, which has a good sol-
13
ubility in solvents such as CH Cl , CHCl , toluene, or even warm benzene so that C-
2
2
3
NMR spectra can be recorded within a few hours, we attempted to study the pyrimidin-
-amine induced self-assembly in solution. Neither absorption combined with station-
2
ary fluorescence nor NMR spectra gave any indication that the anthracene p-systems
are somehow coming into close interactions. No concentration-dependent shifts
could be detected in a variety of solvents. Also no shifts were observed upon addition
of various guest molecules such as phenazine or 2,3,6,7-tetramethylphenazine [15].
Fluorescence-quenching experiments in CHCl showed that upon addition of phena-
3
zine, the anthracene fluorescence was quenched similarly in 6 and in 14 (Fig. 3),
again indicating that preorganization induced by the pyrimidin-2-amine moieties of
6
, does not occur in solution. No attempts were made to see if nonlinear Stern–Volmer
behavior is encountered.
To exclude that adventitious H O could inhibit the H-bonding of the pyrimidin-2-
2
amines, all solvents were thoroughly dried as described in the Exper. Part. In
1
(
D )benzene at 558, no chemical-shift changes could be observed in the H-NMR spec-
6

3
38
Helvetica Chimica Acta – Vol. 89 (2006)
a
Scheme 2. Synthesis of 7 )
a
)
The yield of the last two reactions in the sequence is not given due to incomplete conversion of
the starting material and difficult chromatographic separation. A pure sample of 7 could be isolated
1
in ca. 10% yield, although according to TLC and H-NMR an analytical yield of >50% could be
determined.
tra upon addition of phenazine. However, small shifts for the protons of 6 were
observed upon titrating with CD OD in the NMR tube, which efficiently disrupts H-
3
bonds. To exclude that these shifts are due to a perturbation of the putative H-bonding
network, an additional control experiment was carried out by titrating a solution of
pure 6 dissolved in dry CDCl with CD OD. As seen from Fig. 4, the effects are
3
3
quite small, and we ascribe them to the change in solvent polarity and not to a disrup-
tion of the H-bonding network. Note that upon CD OD addition, the NH protons
3
2
become deshielded while the anthracene and pyrimidine protons (data not shown)
are slightly shielded.
The above data show that the long chains in 6, probably due not only to their bulk
but also to their hydrophobic nature, hinder formation of an extended H-bonding net-
work in solution which would bring about p-stacking interactions, and of course no
association constant could be determined. Furthermore, in the absence of this preorga-
nization, no inclusion complexes are formed upon addition of aromatics with extended
p-systems.
Complexation Studies with Porphyrins and Charge-Transfer Acceptors. The ability
of the 2PA groups in 6 to coordinate the central Zn-atom within porphyrinato com-
plexes was probed by absorption and NMR spectroscopies in the presence of the
very soluble model porphyrinatozinc 19. We have shown earlier that 19, in contrast
to the isomeric porphyrinatozinc 20, shows only small variations in chemical shifts
over a wide concentration and temperature range [16]. This is because the 1,4-phenyl-

Helvetica Chimica Acta – Vol. 89 (2006)
339
Fig. 3. Stationary fluorescence quenching experiments with phenazine added to solutions in dry CHCl3
of a) 6 and b) 14. Pathlength 1 cm.
ene moieties in position 5 and 15, which in homologous crystal structures have a dihe-
dral angle to the porphyrin plane of 658, allow only a marginal p-stacking interaction
[
16].
Fig. 5 and 6 present the changes in the NMR spectra of 6 in dry CDCl , recorded at
3
2
58 upon addition of the porphyrin 19. Significant shieldings of the pyrimidine (HꢁC(4)
and HꢁC(6)) and a-anthracene signals were observed upon titration with porphyrin 19.
In contrast, the ethereal CH O (aCH ) groups of the decyloxy chains show only a minor
2
2
upfield shift while the porphyrin protons remain constant.
As seen from Fig. 5, beside the upfield shift, the pyrimidine protons experience con-
siderable broadening and eventually become split at 258 and at a 1:1 stoichiometry
between 6 and 19 (upper most trace in Fig. 5,a and inset). Upon heating to 558, this sig-
nal shifts back downfield and sharpens (Fig. 5,b). These changes are perfectly reversi-
ble upon cooling. This spectral behavior can only be accounted for either by the forma-
tion of two different complexes which are in equilibrium and by the slow exchange limit

3
40
Helvetica Chimica Acta – Vol. 89 (2006)
Fig. 4. Titration of a solution of 6 in dry CDCl at 40ꢀ0.58 with CD OD.
3
3
at room temperature on the NMR time scale, or by the magnetic nonequivalence of Hꢁ
C(4) and HꢁC(6) of the pyrimidine moiety upon 1:1 complexation of the N(1)-atom.
From the crystal structure of 1, we exclude that complex formation occurs via the NH
2
group. With excess 19, evidently a 1:2 complexation will be favored. Both ‘syn’ and
‘
anti’ conformations are possible in such a 6·19 species. While rotation about the
2
(
1,4-phenylene)-prophyrin linkage of 19 should be unhindered, at room temperature,
the pyrimidine groups of 6 rotate slowly because of the neighboring anthracene H-
atoms. The two rotamers, which have different chemical shifts, may equilibrate on
the NMR time scale upon heating. Evidently this is not true for absorption or fluores-
cence where different excitonic interactions are encountered between the opposing
prophyrin rings in the two rotamers. Nakanishi, Berova, and coworkers have systemati-
cally studied such interactions in bis-porphyrins covalently linked to chiral skeletons

Helvetica Chimica Acta – Vol. 89 (2006)
341
using circular dichroism [17][18]. Very recently, Hunter, Ballesteros, and coworkers
have studied the complexation behavior of a (tetraarylporphyrinato)zinc with the
ditopic N-ligand DABCO (1,4-diazabicyclo[2.2.2]octane) [19]. We expect a very similar
and rather complicated spectral behavior in our case. A line-shape analysis which, in
principle, could give more information on the multiple equilibria involved was beyond
the scope of our present communication. Also the absorption spectra at 1:1 stoichiom-
etry between 6 and 19 show complicated line shapes, indicating the presences of multi-
ple species (data not shown).
Fig. 7 shows probable calculated molecular models of the complex 7·19 . More
2
complicated architectures, which we cannot exclude, are also envisageable, and we
are pursuing crystallization trials both at 1:1 stoichiometries of 7 and 19 and with excess
porphyrin 20. However, due to the presence of multiple long chains, chances to obtain
measurable X-ray diffraction patterns are quite small.
With ethenetetracarbonitrile (TCNE), compound 6 formed an interesting charge-
transfer (CT) complex of blue color but it could not be isolated in pure form and char-
acterized. While anthracene derivative 6 melts at 2488 (and decomposes over 3508), the
blue complex did not melt or change its appearance up to 4008. Unchanged 6 which was
pale tan-yellow could be recovered in good yields from the complex by washing with
acetone. Attempts to obtain mass spectra with a variety of ionization techniques and
matrices failed to evidence parent peaks with a molecular mass higher than that of 6.
Elemental analysis for H and N gave results which indicated that probably one mole-
cule of TCNE is present in the complex, although the C-content was consistently lower
than the required value. Anthracenes substituted at C(9) are known to form CT com-
plexes with TCNE [21–23]. In the case of the 9,10-[2.2]-para-anthracenophane, while
in CH Cl solution a l value of 865 nm was reported for the CT band [23], in the crys-
2
2
max
tal which is of green color, the TCNE molecules lie in an unparallel orientation to the
anthracene planes, at an angle of 1448 [21], bridging anthracenophanes along the stack-
ing axis. This is quite untypical for other arene·TCNE complexes where the overlap
between the HOMO of the arene and the LUMO of TCNE is maximized in a parallel
orientation. Apparently, no other crystal structure of an anthracene derivative and
TCNE is known. We note in the case of 6·TCNE that, due to the four alkoxy substitu-
ents, the stability of the complex should be increased, and a bathochromic shift of the
CT band was observed in comparison to all other known examples.
Co-crystallization experiments of 6 with phenazine from various solvents led in sev-
eral cases to the isolation of large yellow crystals which, however, proved to be pure
phenazine. Thus, no clathrates which could resemble the sponging ability of 5 were
formed. Furthermore, several attempts to grow single crystals suitable for X-ray diffrac-
tion from pure 6 – over half a decade – invariably failed although several methods, sol-
vent mixtures, and temperature protocols were employed. It is interesting to note that
also the parent compound 5 when co-crystallized with anthraquinone instead of phen-
azine, grows large needles (1–2 cm long) from a DMSO solution but which proved to
be only anthraquinone and not a clathrate.
Self-Assembly in the Crystal. Evidence that 2PA-mediated assembly does operate in
homologous constructs came from the less soluble compound 7. We could grow single
crystals of 7 suitable for X-ray diffraction by slow evaporation from DMSO. Of course,
as the previously synthesized 5,5’-(anthracene-9,10-diyl)bis[pyrimidin-2-amine] (5)

3
42
Helvetica Chimica Acta – Vol. 89 (2006)
1
Fig. 5. a) Low-field region of the H-NMR spectra (300 MHz, CDCl , 258) of 6 upon addition of por-
3
1
phyrin 19. b) Temperature dependence of the low-field H-NMR region (300 MHz, CDCl ) at a 1:1
3
stoichiometry between 6 and 19. Note that the pyrimidine-proton signals, which show a broad and par-
tially complicated splitting at 258, sharpen and shift to lower field at 358, but there is hardly any
change from 45 to 558.

Helvetica Chimica Acta – Vol. 89 (2006)
343
Fig. 5 (cont.)
[
1][2], 7 is completely insoluble in less polar solvents. Fig. 8 shows the crystal structure
of a monomer of 7 which binds two adventitious DMSO molecules to both 2-amino
groups which are in an ‘anti’-conformation. The length of the O···H bond is 2.161 Å
while the O···HꢁN angle is 168.68 with an OꢁN distance of 3.009 Å which document
quite tight H-bonding of the solvent.
The programmed double H-bonding self-assembly known to function in the previ-
ously described pyrimidin-2-amine crystal structures [1–3] is perturbed in the case of 7
due to the DMSO inclusion which blocks one of the NH H-bonding donor sites. Thus
only the other NH can bind to a neighboring N(1) atom of the pyrimidine moiety, form-
ing an extended H-bonding network in one direction which is depicted in Fig. 9. In the
perpendicular direction, p-stacking assembles the anthracenes in an off-setted geome-
try as displayed in Fig. 10.

3
44
Helvetica Chimica Acta – Vol. 89 (2006)
1
Fig. 6. Dependency of the H-NMR chemical shifts (300 MHz, 258) of 6 in dry CDCl on the concen-
3
tration of added porphyrin 19. The initial concentration of 6 was 3.0 mM. Initially 30 ml and then 50 ml
aliquots of a 6.3 mM solution of 19 were incrementally added to the NMR tube. Note that the y-axis
is fourfold expanded in comparison to Fig. 4.
Fig. 7. Probable molecular models for the rotamers of a complex 7·19 . Geometry optimization was
2
®
performed within the HyperChem package [20]. meso-Substituents of 19 were omitted from the cal-
culation, while the crystal geometry of 7 was used in the optimizations.

Helvetica Chimica Acta – Vol. 89 (2006)
345
Fig. 8. X-Ray structure of 7. Apparently, the S-atom (yellow) of the DMSO solvent molecules is
slightly disordered as models fitted with a 9 :1 distribution between two different position gives a sig-
nificant improvement of the R factor. Only the major orientation is depicted. N- and O-Atoms are
represented by blue and red ellipsoids, resp.
Fig. 9. H-Bonding network in the crystal of 7 which leads to a planar tape. S-Atoms are yellow,
O-atoms red, N-atoms blue, and C-atoms gray, while protons are magenta.
The plane-to-plane spacing of ca. 3.45 Å in the crystal of 7 indicates quite tight p-
stacking interactions. Interestingly, there exists an extensive overlap between the dime-
thoxy-substituted ends of the anthracene moieties. In the simple electrostatic model of
the p–p interactions proposed by Hunter and Sanders [24] which predict off-setted geo-
metries for extended p-systems such as porphyrins, an attractive interaction between
the s- and p-orbitals is balanced by the repulsion between the two p-sytems. In 7, we
note that two electron-rich regions are forced to overlap. Thus charge-transfer interac-
tions do not operate. In principle the electron-rich regions of the anthracene moieties
could have stacked with the electron-poor pyrimidine moieties. This, however would
have resulted in less favorable H-bonding interactions between two adjacent pyrimi-
din-2-amines. Previously it has been observed that charge-transfer interactions are
actually a particular case of p-stacking, which is the more general phenomenon.
Thus, in phenyl-substituted pyrylium salts [25], this most electron-deficient 6-mem-
bered heteroarene ring p-stacks in the crystal with the much more electron-rich phenyl
rings of adjacent molecules [26]. When the phenyl rings are electron poor, extended p-
stacking is encountered, while with electron-rich phenyl rings, dimers are formed which

3
46
Helvetica Chimica Acta – Vol. 89 (2006)
Fig. 10. p-Stacking interactions in the crystal of 7
also give more deeply colored crystals due to partial charge transfer from the phenyl to
the pyrylium ring [26]. It is also well known from dynamic NMR studies that the p–p
interaction between electron-deficient aromatics is stronger than the one between an
electron-poor and an electron-rich system, or the one between two electron-rich aro-
matic systems [27].
Conclusions. – Clearly a hierarchic self-assembly algorithm, beside dense crystal
packing, operates in the case of 7, with H-bonding, including also the crystallization sol-
vent, being the dominant interaction. p-Stacking is the recessive interaction, and hydro-
phobic or dispersive forces which may act between the Me groups are probably negli-
gible. The conformation of the MeO groups allows for close packing. If longer alkyl
chains are present at the ether O-atoms, as in the case of 6, probably dispersive forces
may become dominant, and their interdigitation would lead to a completely different
and more complicated crystal packing as shown for the 2,3-bis(hexyloxy)anthracene
gelators [10]. We have reproducibly obtained nano- to microcrystals of 6 which tend
to aggregate in fractal fashion but we have never observed gelating tendencies in our
solutions. This lends support for the hypothesis that for anthracene gelators, in the
absence of H-bonding groups, a dipole moment which packs the molecules (or higher
assembly of polarized molecules) in a head to tail fashion is essential.
That the p–p interactions do operate in the crystal of 7 is an indication that the elec-
tron-rich anthracene rings cannot overlap in solutions of 6 as seen from absorption,
fluorescence, and NMR studies. H-Bonding probably does occur between molecules
of 6 in nonpolar solvents such as dry benzene or even dry CHCl but these lead to
3
assemblies without stacking. Thus, the 2PA recognition group is useful for solution
self-assemblies only in the case when metal-ion ligation can operate, as was the case
with the 2PA-substituted porphyrinatozincs 1 and 4. Presently, it is not easy to estimate

Helvetica Chimica Acta – Vol. 89 (2006)
347
the association constants for multiple supramolecular interactions which act in conjunc-
tion as either positive or negative cooperativity can be encountered.
Our study shows that predicting crystal packing(s) or polymorph formation with
concomitant supramolecular interactions, in spite of continuing progress [28][29], is
still a field with much room for improvement towards the goal of crystal engineering
[
30].
We are grateful to Professor Jean-Pierre Vigneron (Paris) for generously supplying us with a sample
of 2,3,5,6-tetramethylphenazine which was used in the co-crystallization experiments, absorption studies,
1
and H-NMR titrations. S. T. B. was supported at ISIS (Strasbourg), where the main part of the experi-
mental work was performed, by the Collège de France through a ‘maître de conférences’ position.
Experimental Part
General. CHCl and CDCl were dried by boiling overnight over P O and were freshly distilled
3
3
2
5
before spectroscopic measurements. C D was dried by distillation from NaꢁK alloy. TLC: silica gel
6
6
plates from Macherey-Nagel; elution with CH Cl stabilized with 0.2% EtOH. Column chromatography
2
2
(CC): Merck silica gel (40–63 mm). M.p.: Büchi B-540 apparatus; in open capillaries; not corrected. UV/
VIS Spectra: Varian Cary-3 and Cary-500 spectrometer, equipped with a Peltier variable-temp. unit; l
max
ꢁ1
in nm (log emax). FT-IR Spectra: Perkin-Elmer 1600 or Spectrum GX spectrometers; in cm . NMR Spec-
1
13
tra: Bruker AC-200 spectrometer, at 200 ( H) and 50.33 MHz ( C); Bruker DPX-300 spectrometer, at
1
3
00 MHz ( H); chemical shifts d in ppm rel. to residual nondeuterated solvent signal (for CHCl , d
.26 ( H) and 77.0 ( C); for (D )DMSO, d 2.50 ( H)), J in Hz. FAB-MS: by the Service de Spectrométrie
3
1
13
1
7
5
de Mass de l’Université Louis Pasteur in Strasbourg; Micromass-Autospec instrument; 2-nitrobenzyl
alcohol matrix; in m/z (rel. %). Elemental analyses; obtained either at the Service de Microanalyse de
l’Université Louis Pasteur, Strasbourg, or at the Institute for Nanotechnology, Karlsruhe; Carlo-Erba
CEA flash microanalyzer.
2,3,6,7-Tetrahydroxyanthracene-9,10-dione (11). A modification of the procedure reported by Boldt
[
9] was employed: 2,3,6,7-Tetramethoxyanthracene-9,10-dione (10; 7.51 g, 22.9 mmol) and 48% HBr
soln. were heated under reflux (oil bath temp. 1508) for 3 days. After the first 24 h, more 48% HBr
soln. (15 ml) was added through the reflux condenser to wash down some unreacted starting material
which had collected there due to intense foaming and to its poor wettability. After 48 h, the yellow
color had turned completely to ochre, and the foaming had stopped. After cooling, the precipitate was
collected by filtration, washed to neutral pH with dist. H O, and air-dried: 6.36 g of crude 11 (102%).
2
Purification was be accomplished after the next step. IR (KBr): 3406s, 1655m, 1580s, 1511m, 1321s,
1
1
232m, 1160m, 1070m, 778m, 749m, 598m. H-NMR ((D )DMSO; oven dried (1108) sample): 10.39
6
(
sharp s, OH; br. signal in wet samples); 7.43 (s, arom. H); 3.94 (residual s, MeO of some contaminating
unhydrolyzed product).
,3,6,7-Tetrakis(decyloxy)anthracene-9,10-dione (12). The crude 11 (5.0 g, 18.3 mmol) was finely
ground in a mortar and then transferred to a reaction flask. DMF (100 ml) and anh. K CO (17.64 g)
2
2
3
were added to the flask after rinsing the mortar. Then 1-bromodecane (40.34 g, 183 mmol) was added,
and the mixture was stirred for 5 min under Ar bubbling for degassing after which the flask was stoppered
with an Ar-filled balloon. Stirring was continued at r.t. for 90 min, and then the mixture was heated gently
to 608. After 1/2 h at 608, the mixture gelified completely, and an efficient magnetic stirrer was needed to
stir the mixture overnight at this temp. Then, dist. H O (80 ml) was added to dissolve the excess K CO
2
2
3
and to break up the cake. The mixture was cooled in an ice-water bath and then filtered. The kaki-colored
precipitate was washed with dist. H O (3×), 50% aq. MeOH (1×), and lastly with cold MeOH. Isolation of
2
additional product from the dark-colored filtrate by CC proved uneconomical, so the filtrate was dis-
carded. The crude product was recrystallized from boiling toluene (250 ml): 10.65 g (69.6% yield) of
pure 12, free of the MeO-substituted contaminant. R (CH Cl ) 0.76. M.p. 139–1408. UV (CH Cl ): 280
f
2
2
2
2
(sh., 4.84), 293 (5.11), 345 (4.40). IR (KBr): 2918s, 2850s, 1662m, 1575s, 1512m, 1466m, 1368m, 1329s,

3
1
48
Helvetica Chimica Acta – Vol. 89 (2006)
1
227m, 1077m, 1063m, 892w, 742m, 616w. H-NMR (200 MHz, CDCl ): 7.64 (s, HꢁC(1), HꢁC(4), Hꢁ
3
C(5), HꢁC(8)); 4.18 (t, J=6.5, 4 CH O), 1.89 (quint., 8 H); 1.28 (sharp m, 56 H); 0.88 (t, 4 Me). FAB-
2
+
MS: 833.8 (100, [M+H] ), 693.6 (51, [M+HꢁC H ]), 553.5 (20, [M+Hꢁ2 C H ). Anal. calc. for
10
20
10 20
C H O : C 77.84, H 10.64; found: C 77.51, H 10.51.
54
88
6
2,3,6,7-Tetrakis(decyloxy)anthracen-9(10H)-one (13). Zn-Powder (16.9 g) was activated by stirring
for 10 min with a soln. of CuSO (0.4 g) in H O (250 ml). The aq. soln. was decanted, and 10% NaOH
4
2
soln. (160 ml), 12 (8.67 g, 10,4 mmol), and toluene (150 ml) were added to the Zn-powder. The mixture
was heated overnight on an oil bath at 1208 maintaining vigorous boiling and magnetic stirring (the aq.
layer became reddish while the yellow toluene layer discolored slowly as the reaction proceeded). Then,
the toluene layer was decanted while still warm from the orange-red aq. layer and washed once with
warm H O and separated by decantation from all the unreacted Zn-powder. Refrigeration and seeding
2
produced crystals which were filtered off and dried: 7.35 g (86%) of crude 13, contaminated by a small
amount of the anthracene (TLC, NMR). Longer reaction times did not increase the yield in anthracene.
Extraction with CH Cl of the aq. layer and combining it with the toluene filtrate yielded an additional
2
2
1
.31 g of material which was subjected to CC (silica gel, CH Cl ). TLC (CH Cl ): R 0.68; spots became
2 2 2 2 f
yellow by and by probably due to back conversion to the anthracenedione. UV (CH Cl ): 284, 337. IR
2
2
1
(KBr): 2918s, 2848s, 1648w, 1596s, 1513s, 1467m, 1382m, 1334s, 1262s, 1228s, 1091s. H-NMR (200
MHz, CDCl ): 7.79 (s, HꢁC(1), HꢁC(8)); 6.82 (s, HꢁC(4), HꢁC(5)); 4.13 (s, CH (10)); 4.08 (t,
3
2
13
J=6.6, 4 CH O); 1.87 (br. quint., 8 H); 1.48 (br. m, 8 H); 1.28 (sharp m, 48 H); 0.88 (t, 4 Me). C-
2
NMR (50 MHz, CDCl ): 182.5 (CO); 153.1; 148.40; 135.0; 125.4; 111.1; 110.3; 69.2 (OCH ); 32.0; 29.7;
3
2
+
+
2
(
9.5; 29.24; 29.17; 26.1; 22.8; 14.2. FAB-MS: 819.9 (100, [M+H] ), 791.8 (13, [M+HꢁCO] ), 679.7
+
23, [M+HꢁC H ] ).
10
20
2
,3,6,7-Tetrakis(decyloxy)anthracene (14). By warming gently, 13 (7.35 g, 8.98 mmol) was dissolved in
CH Cl (200 ml). Ar was then bubbled through the soln. for 5 min. Powdered NaBH (5.25 g, 135 mmol)
2
2
4
was added, and then MeOH (30 ml) was added dropwise under stirring at r.t. After the addition was com-
pleted, foaming started. After 90 min, when the foaming subsided, a fresh portion of NaBH (1.5 g) was
4
added. The reaction was quenched after 7 h by adding carefully and dropwise AcOH (10 ml). Extensive
foaming could be controlled by adding small portions of CH Cl . The mixture was then left to stand over-
2
2
night. Then AcOH (5 ml) was added until no more foaming occurred, and then carefully conc. HCl soln.
10 ml). At the end of the addition, the mixture became slightly greenish. Stirring was continued for 5 min
after which H O (50 ml) was added. The CH Cl layer, which contained a fine precipitate, was washed
(
2
2
2
with H O (200 ml) and then filtered. The colorless precipitate was washed repeatedly with H O until
2
2
the filtrate had neutral pH and dried: 5.4 g (75%) of 14, pure by TLC and NMR. Interestingly, this
long-chain-substituted anthracene is completely unwettable. Additional 0.24 g of 14 could be obtained
from the filtrate: Thus, the yellowish CH Cl filtrate was washed once with H O, then with aq.
2
2
2
NaHCO soln., then again with H O, dried (Na SO ), concentrated, and deposited onto silica gel. CC
3
2
2
4
(
hexane/CH Cl 1:1) separated neatly 14 from some corresponding 9,10-dihydroanthracene (180 mg)
2 2
and unreacted 13. An anal. sample of 14 was obtained by recrystallization from toluene.
TLC (CH Cl /hexane 1:1): R 0.63 (14); 0.47 (dihydro derivative), 0.18 (13). 14: M.p. 160–1618. UV
2
2
f
1
(
CH Cl ): 271 (5.23), 372 (4.30). IR (KBr): 2956m, 2918s, 2852s, 1492s, 1467s, 1241s, 1167m, 893m. H-
2 2
NMR (200 MHz, CDCl ): 7.95 (s, HꢁC(9), HꢁC(10)); 7.11 (s, HꢁC(1), HꢁC(4), HꢁC(5), HꢁC(8));
3
4
.12 (t, J=6.6, 4 CH O); 1.91 (br. quint., 8 H); 1.50 (br. m, 8 H); 1.28 (sharp m, 48 H); 0.88 (t, 4 Me).
2
+ +
EI-MS: 802.9 (100, M ), 662.6 (8, [MꢁC H ] ), 241.7 (20). FAB-MS (a variety of matrices): no signals.
10
20
Anal. calc. for C H O : C 80.74, H 11.29; found: C 80.50, H 11.03.
54
90
4
9,10-Dibromo-2,3,6,7-tetrakis(decyloxy)anthracene (15). Compound 14 (2.41 g, 3.0 mmol) was dis-
solved in boiling CCl (100 ml) and filtered while hot. After cooling to r.t., Br (1.20 g, 0.38 ml, 7.5
4
2
mmol) was added via syringe. The mixture was heated (1008 bath temp.) to a gentle boil when HBr
started to evolve. The reaction was stopped after 35 min at reflux, the mixture cooled, and the precipitate
filtered and dried: 1.10 g of 15. An additional crop (0.885 g) was obtained by evaporating the filtrate and
recrystallizing the residue from boiling CCl (15 ml), increasing thus the yield to 69%. R (CH Cl /hexane
4
f
2
2
1
:1) 0.77. UV (CH Cl ): 282, 384. IR (KBr): 2954m, 2921s, 2850s, 1505s, 1496s, 1464s, 1251s, 1167s, 831m.
2 2
1
H-NMR (200 MHz, CDCl ): 7.67 (s, HꢁC(1), HꢁC(4), HꢁC(5), HꢁC(8)); 4.21 (t, J=6.6, 4 CH O); 1.96
3
2

Helvetica Chimica Acta – Vol. 89 (2006)
349
1
3
(
1
quint., 8 H); 1.50 (br. m, 8 H); 1.28 (sharp m, 48 H); 0.89 (t, 4 Me). C-NMR (50 MHz, CDCl ): 150.6;
3
+
26.5; 106.5; 101.6; 68.91; 29.6; 29.44; 29.37; 29.0; 26.1; 22.7; 14.1. EI-MS: 961.5 (100, M , isotopic pat-
tern), 933.4 (50, isotopic pattern), 400.1 (60).
,5’-[2,3,6,7-Tetrakis(decyloxy)anthracene-9,10-diyl]bis[pyrimidin-2-amine] (6). An oven-dried flask
5
was charged with 15 (1.092 g, 1.136 mmol), 2-[5-(trimethylstannyl)pyrimidin-2-yl]-1H-isoindol-1,3(2H)-
dione [3] (16; 1.102 g, 2.842 mmol), and toluene (25 ml) freshly distilled from Na dispersion. Ar was bub-
bled through the mixture for 5 min after which tetrakis(triphenylphosphine)palladium catalyst (65.6 mg,
0
.056 mmol) was added. Ar bubbling was continued for 2 min after which the flask was stoppered with an
Ar-filled balloon. The mixture was heated to 100ꢀ58 as it became homogeneous, translucent pale yellow.
Heating was continued for 22 h. Cooling to ꢁ188 produced a fine precipitate which was filtered with dif-
ficulty and washed with two portions of CH Cl . Deprotection was performed by suspending this precip-
2
2
itate in CH Cl (400 ml) and treatment with methylhydrazine (8 ml). The mixture which became blueish
2
2
fluorescent contained a fine precipitate which was stirred for 2 h at r.t. Silica gel (8 g) was added and the
mixture evaporated (Caution: the dry silica gel with excess methylhydrazine, after removing from the
rotary evaporator, may overheat so that careful handling (insulating gloves) and cooling is recom-
mended). CC (silica gel, 10% EtOH/CHCl ) produced a yellow band which yielded 1.021g (90.9%
3
over two steps) of pure 6. An anal. sample was recrystallized from 1,2-dichloroethane. R (EtOH/
f
CHCl 2 :10) 0.82. M.p. 2488 (yellow melt). UV (CHCl ): 276, 383 (see also Fig. 11). IR (KBr): 3340m,
3
3
1
3
186m, 2924s, 2853ms, 1659s, 1595ms, 1496vs, 1465s, 1351m, 1244s, 1208m, 1142m, 841m, 806m. H-
NMR (200 MHz, CDCl ): 8.41 (s, 4 H, HꢁC(4), HꢁC(6) (py)); 6.81 (s, HꢁC(1), HꢁC(4), HꢁC(5),
3
HꢁC(8) (anth)); 5.24 (s, D O exchange, 2 NH ); 3.91 (t, J=6.4, 4 CH O); 1.81 (quint., 8 H); 1.43 (br.
2
2
2
13
m, 8 H); 1.27 (sharp m, 48 H); 0.88 (t, 4 Me). C-NMR (50 MHz, CDCl , 508): 162.4; 160.2; 149.9;
3
1
2
27.0; 123.5; 105.3; 69.1 (CH O); 31.9; 29.62; 29.55; 29.43; 29.32; 29.1; 26.1; 22.6; 14.0 (Me); the seventh
arom. C (probably C(2) of pyrimidine) resonated at d ca. 126 (br., due to the proximity of the 3 N-atoms).
+
+
FAB-MS: 989.8 (100, [M+H] ), 848.6 (16, [MꢁC H ] ). Anal. calc. for C H N O : C 75.26, H 9.78, N
10
20
62 96
6
4
8
.49; found: C 75.24, H 9.59, N 8.19.
Complex of 6 with TCNE. Ethenetetracarbonitrile (TCNE) was added to a warm CHCl soln. of 6
3
when a blue precipitate formed rapidly which could be filtered with difficulty under Ar. The starting 6
was recovered unchanged after addition of acetone which bleached the blue powder. Blue powder:
1
13
M.p. >4008 (dec.). H-NMR: no differences from the spectrum of uncomplexed 6. C-NMR (50 MHz,
Fig. 11. Absorption and fluorescence spectra of 6 in CH Cl (excitation wavelength 380 nm)
2
2

3
50
Helvetica Chimica Acta – Vol. 89 (2006)
CDCl , 508): 160.2; 149.8; 127.0; 123.5; 105.1; 69.0; 31.94; 29.64; 29.57; 29.44; 29.34; 29.1; 26.2; 22.7; 14.0;
3
only 5 arom. C were visible, the d(C) 162.4 of 6 was absent, and no TCNE C-signal was detected. Anal.
calc. for C H N O (6·TCNE): C 73.08, H 8.66, N 12.53; found: C 70.89, H 8.64, N 12.40.
68
96 10
4
9,10-Dibromo-2,3,6,7-tetramethoxyanthracene (18). Bromination of 2,3,7,8-tetramethoxyanthracene
(17) [14] which was recrystallized from hot nitrobenzene on a gram scale, was performed as described for
1
5, except that the suspension of 17 in CCl was not filtered. The product was purified by boiling with
4
1
toluene and triturating and was used in the next step. UV (CDCl ): 278, 380 (rel. extinct. 7.8 :1). H-
3
1
NMR (200 MHz, CDCl ): 7.68 (s, HꢁC(1), HꢁC(4), HꢁC(5), HꢁC(8)); 4.11 (s, 4 MeO). H-NMR
3
13
(
200 MHz, (D )DMSO): 7.65 (s, HꢁC(1), HꢁC(4), HꢁC(5), HꢁC(8)); 4.01 (s, 4 MeO). C-NMR (50
6
MHz, CDCl , 508): 150.7; 126.6; 105.6 (CH); 56.0 (MeO); due to the low solubility, the quaternary C
3
+
was not visible after an overnight accumulation. FAB-MS: 456.0 (100, [M+H] ), 442.0 (50,
+
+
[
MꢁMe] ), 426 (38, [Mꢁ2 Me] ), 413.2 (38); abundant ions below m/z 350 due to extensive fragmen-
tation of the anthracene skeleton; the low signal to noise ratio indicates the high stability of the parent
ion.
5,5’-(2,3,6,7-Tetramethoxyanthracene-9,10-diyl)bis[pyrimidin-2-amine] (7). Stille coupling of the
stannyl derivative 16 (669 mg, 250 mol-%) with 17 (315 mg) and tetrakis(triphenylphosphine)palladium
catalyst (40 mg, 5 mol-%) was performed in dry toluene (20 ml) by heating at 1008 for 56 h. Due to the
low solubility of 17, the mixture was heterogeneous, and the reaction could not be driven to completion.
After deprotection with methylhydrazine similarly as described for 6, CC (silica gel, 10% EtOH/CHCl3)
yielded several fractions. After impure fractions (100 mg) containing also starting material, the monocou-
pled product was eluted which was recrystallized from CHCl /EtOH 98 :2. After mixed fractions, pure 7
3
(250 mg) was obtained. No optimization of the reaction was attempted. Single crystals of 7 were grown
from DMSO upon dissolution under gently heating, cooling to r.t., and slow evaporation over 12 months.
Co-crystallization experiments with phenazine or 1,4,5,8-tetramethylphenazine from either hot DMSO
or MeCN/DMSO did not yield inclusion complexes but the same crystalline hexagonal plates of 7
were obtained. In some trials, yellow needles of phenazine even crystallized preferentially instead.
1
Data of 7: M.p. >4008 (dec.). H-NMR (200 MHz, (D )DMSO): 8.32 (s, 4 H, HꢁC(4), HꢁC(6) (py));
6
6
.86 (br.), 6.84 (2s, 2 NH , HꢁC(1), HꢁC(4), HꢁC(5), HꢁC(8) (anth)); 3.71 (s, 6 CH O); solvent (2.50),
2
2
þ
4
+
H O (3.30), and a broad peak at d ca. 1 were also visible. FAB-MS: 484.3 (C H N O ;[M+H] ; calc.
2
26 24
6
4
84.19); no other peaks over 35%.
Monocoupled Product 5-(10-Bromo-2,3,6,7-tetramethoxyanthracen-9-yl)pyrimidin-2-amine6) H-
1
NMR (200 MHz, (D )DMSO): 8.38 (s, HꢁC(4), HꢁC(6) (py)); 7.75 (s, HꢁC(1), HꢁC(8) (anth)); 6.78
6
(
s, HꢁC(4), HꢁC(5) (anth)); 5.23 (br. s, 2 H); 4.11 (s, MeOꢁC(2), MeOꢁC(7)); 3.84 (s, MeOꢁC(3),
MeOꢁC(6)); solvent (7.26), H O (1.550), and a broad peak at d ca. 1.6 were also visible. FAB-MS:
2
79
þ
+
4
69, 471.2 (isotopic pattern, C H BrN O , [M+H] ; calc. 470.1); no other peaks over 6%.
22 21 3
4
Crystallographic Data of 7. The data collections were carried out on a Stoe-IPDS-I diffractometer
equipped with a Schneider rotating anode by using graphite-monochromated Mo-Ka (l 0.71073 Å) radi-
2
ation at 200 K. The structure solutions and full matrix least-square refinements based on F were per-
formed with the SHELX-97 program package [31]. Molecular diagrams were prepared with the program
7
ꢀ
Diamond [32]. Data for 7 ): Rhombic plates, 0.22×0.2×0.05 mm, M 784.5, triclinic, space group P1 (No.
r
3
2
), a=8.206(2), b=8.633(2), c=12.681(3) Å, a=71.26(3), b=85.91(3), g=67.47(3)8, V=784.5(3) Å ,
Z=1, D =1.356 g cm , m(Mo-Ka)=0.222 mm . All O-, N-, C-, and S-atoms were refined with aniso-
ꢁ3
ꢁ1
c
tropic displacement parameters with a 10% disorder of the S-atom. H-Atoms were added in calculated
positions to give a final R value of 0.0617 for 208 parameters and 2479 unique reflections with Iꢂ2s(I)
1
and wR of 0.1804 for all 7012 reflections (R =0.1409).
2
int
6)
7)
This by-product may be useful in a second Pd-catalyzed aryl–aryl coupling with a different recogni-
tion group, or with the same pyrimidin-2-amine but protected otherwise than with phthalimide to
assure orthogonality.
CCDC-285563 contains the supplementary crystallographic data for this paper. These data can be
obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.

Helvetica Chimica Acta – Vol. 89 (2006)
351
REFERENCES
[
[
1] M. J. Krische, J.-M. Lehn, N. Kyritsakas, J. Fischer, Helv. Chim. Acta 1998, 81, 1909.
2] M. J. Krische, J.-M. Lehn, N. Kyritsakas, J. Fischer, E. K. Wegelius, M. J. Nissinen, K. Rissanen, Helv.
Chim. Acta 1998, 81, 1921.
[
3] M. J. Krische, J.-M. Lehn, N. Kyritsakas, J. Fischer, E. K. Wegelius, K. Rissanen, Tetrahedron 2000,
56, 6701.
[4] T. S. Balaban, R. Goddard, M. Linke-Schaetzel, J.-M. Lehn, J. Am. Chem. Soc. 2003, 125, 4233.
[5] A. Tsuda, S. Sakamoto, K. Yamaguchi, T. Aida, J. Am. Chem. Soc. 2003, 125, 15722.
[6] M. Vinodu, Z. Stein, I. Goldberg, Inorg. Chem. 2004, 43, 7582.
[7] M. J. Crossley, P. L. Burn, J. Chem. Soc., Chem. Commun. 1987, 39.
[8] A. Müller, M. Raltschewa, M. Papp, Chem. Ber. 1942, 75, 692.
[9] P. Boldt, Chem. Ber. 1967, 100, 1270.
[
[
10] T. Brotin, R. Utermöhlen, F. Fages, H. Bouas-Laurent, J.-P. Desvergne, J. Chem. Soc., Chem. Com-
mun. 1991, 416.
11] J.-L. Pozzo, J.-P. Desvergne, G. M. Clavier, H. Bouas-Laurent, P. G. Jones, J. Perlstein, J. Chem. Soc.,
Perkin Trans. 2 2001, 824.
[
[
[
[
[
[
[
[
12] A. Petitjean, R. G. Khoury, N. Kyritsakas, J.-M. Lehn, J. Am. Chem. Soc. 2004, 126, 6637.
13] B. M. Trost, Science (Washington, D.C.) 1991, 254, 1471.
14] P. W. Rabideau, J. Org. Chem. 1971, 36, 2723.
15] J.-M. Lehn, F. Schmidt, J.-P. Vigneron, J. Heterocycl. Chem. 1990, 27, 1633.
16] T. S. Balaban, A. Eichhöfer, J.-M. Lehn, Eur. J. Org. Chem. 2000, 4047.
17] S. Matile, N. Berova, K. Nakanishi, J. Am. Chem. Soc. 1995, 117, 7021.
18] S. Matile, N. Berova, K. Nakanishi, J. Fleischhauer, R.W. Woody, J. Am. Chem. Soc. 1996, 118, 5198.
19] P. Ballester, A. Costa, A. M. Castilla, P. M. Deyà, A. Frontera, R. M. Gomilla, C. A. Hunter, Chem.–
Eur. J. 2005, 11, 2196.
®
[
[
[
[
[
[
20] ‘HyperChem ’, Release 7 Professional Version, HyperCube Inc., Gainesville, Florida, 2002.
21] J. M. Masnovi, J. K. Kochi, E. F. Hilinski, P. M. Rentzepis, J. Phys. Chem. 1985, 89, 5387.
22] J. M. Masnovi, J. K. Kochi, E. F. Hilinski, P. M. Rentzepis, J. Am. Chem. Soc. 1986, 108, 1126.
23] A.-F. Mourad, A. Akkermann-Kubillus, J. Hucker, J. Prakt. Chem. 1988, 330, 15.
24] C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5525.
25] T. S. Balaban, A. T. Balaban, in ‘Science of Synthesis, Vol. 14’, E. J. Thomas, Thieme, Stuttgart,
2003, p. 11–200.
[
[
[
[
[
[
26] T. S. Balaban, N. Furmanova, Z. Naturforsch., B 1997, 52, 485.
27] F. Cozzi, J. S. Siegel, Pure Appl. Chem. 1995, 67, 683.
28] A. Gavezzotti, Acc. Chem. Res. 1994, 27, 309.
29] J. D. Dunitz, A. Gavezzotti, Acc. Chem. Res. 1999, 32, 677.
30] G. Desiraju, ‘Crystal Engineering: The Design of Organic Solids’, Elsevier, New York, 1989.
31] G. M. Sheldrick, ‘SHELX-97, Program for X-Ray Crystal Structure Determination and Refine-
ment’, Göttingen University, Göttingen, 1997.
[
32] ‘Diamond, V. 3.0’, Crystal Impact GbR, Bonn, 15. October, 2004.
Received October 19, 2005