.
Angewandte
Communications
cally) more-stable cis-[(L2Pt)2(4,4’-bipy)Cl2] (Scheme 2,
route 1, step a).[8] After halide extraction using AgCSA,
treatment with tpt, and then anion exchange with NH4PF6,
cis-[(L2Pt)6(4,4’-bipy)3(tpt)2](PF6)6 was isolated in virtually
one of the PF6 counteranions (Figure 2a, inset). MS3 experi-
ments show that the disappearance of this 1093 m/z ion results
from the dissociation into a low-intensity ion at 873 m/z,
which matches [(L2Pt)2(4,4’-bipy)F]+, and subsequent MS4
experiments show that this fragments into a singly charged
ion at 505 m/z, by loss of the neutral [(L2Pt)F] from
[(L2Pt)2(4,4’-bipy)F]+. For the trans isomer, the fragmentation
pathway appears to first involve loss of a single 4,4’-bipy,
again either promoted or stabilized by fluoride abstraction
from the counter anion, to give [[(L2Pt)6(4,4’-bipy)2(tpt)2F]-
(PF6)2]3+ (1114 m/z), which subsequently looses another 4,4’-
bipy by a fluoride promoted/stabilized route to produce
[[(L2Pt)6(4,4’-bipy)(tpt)2F2](PF6)2]3+ (1020 m/z). Again, the
selective loss of 4,4’-bipy ligands from the weaker trans-to-
phenylato coordination sites supports the formation of trans-
[(L2Pt)6(4,4’-bipy)3(tpt)2](PF6)6.
1
quantitative yield (Scheme 2, route 1, step b). The H NMR
spectrum of this product (see the Supporting Information,
Figure S2d) again suggested the formation of a single isomer.
When the sequence of addition of 4,4’-bipy and tpt to
[(L2Pt)2Cl2] was switched (Scheme 2, route 2) a single, yet
different product was obtained. A comparison of the 1H NMR
spectra of the two isomers showed significant differences (see
Supporting Information, Figure S2d and S2e), particularly in
resonances HA–F
.
The nESI mass spectra of cis and trans-[(L2Pt)6(4,4’-
bipy)3(tpt)2](PF6)6 showed identical peaks at 1208 and 1884 m/
z, which matched the predicted isotope patterns for the intact
+ 3 and + 2 charged prisms, respectively (not shown).
However, in contrast to the analogous experiments with cis
and trans-[(HL1Pt)6(4,4’-bipy)3(tpt)2](PF6)6, the CID of the cis
and trans isomers of [(L2Pt)6(4,4’-bipy)3(tpt)2](PF6)6 showed
mainly peaks that did not correspond to any sensible
combinations of L2Pt, 4,4’-bipy, tpt, and PF6. Instead, the
dominant CID pathway for both cis and trans-[[(L2Pt)6(4,4’-
bipy)3(tpt)2](PF6)3]3+ involves fluoride abstraction from the
PF6 counteranions. With the cis isomer, the disappearance of
the 1208 m/z peak is initially accompanied by a dominant
species at 1093 m/z, which corresponds to the formula
[[(L2Pt)4(4,4’-bipy)2(tpt)F]PF6]2+ (Figure 2a). In an analogous
manner to cis isomer of [(HL1Pt)6(4,4’-bipy)3(tpt)2](PF6)6, it
appears that the initial fragmentation pathway involves the
The four metallosupramolecular stereochemical isomers
do not undergo isomerization, or reassemble to generate
other assemblies (e.g. Pt4 squares or Pt6(tpt)4 cages) at room
temperature in solution.[9] We attribute this stability to the
1
2
À
unlabile Pt N bonds trans to the nitrogen donors of HL /L ;
although these bonds form readily at room temperature (or
just above in the case of L2) from the corresponding solvato/
halide complexes, the activation barrier for de-coordination is
such that this step is essentially irreversible under ambient
conditions. Therefore the sequence in which the N-donor
bridging ligands are added to the starting platinum complexes
determines the stereochemical outcome of the reaction. In
this regard, the synthesis of these metallosupramolecular
isomers combines elements of covalent (irreversible) syn-
thesis and noncovalent (reversible) thermodynamically con-
trolled assembly. It could also be expected that the isomers
would show some thermodynamic bias towards either the cis
or the trans form, however, heating samples of either cis or
trans-[(HL1/L2Pt)6(4,4’-bipy)3(tpt)2](PF6)6 at 808C for
24 hours results in a complex mixture with no obvious
preference for a single species. This is perhaps unsurprising
as to gain greater than 95% selectivity for a single species, an
energy difference greater than 1.74 kcalmolÀ1 would be
required. This is in marked contrast to the sequential,
kinetically controlled syntheses described herein; these
syntheses give greater than 95% selectivity for single
stereochemical isomers, thus highlighting the potential bene-
fits of exploiting differences in rates of assembly, rather than
simply considering ground-state energies, for the preparation
of multicomponent systems.
À
cleavage of the weaker Pt tpt bonds from adjacent panels
(also supported by a smaller intensity ion at 1311 m/z, which
corresponds to [[(L2Pt)2(4,4’-bipy)(tpt)]PF6]+), a route that is
either promoted or stabilized by abstraction of fluoride from
The vast difference in labilities of cis exchangeable
platinum coordination sites have been exploited to selectively
synthesize multicomponent stereoisomeric assemblies using
a template-free, sequential, kinetically controlled approach.
We anticipate that this approach to noncovalent, and in
particular coordination driven self-assembly will facilitate the
generation of multicomponent, ultimately functional systems.
Received: December 20, 2011
Published online: March 5, 2012
Figure 2. Partial nESI mass spectra with increasing collisional energy
of the +3 charged intact prism (1208 m/z) to illuminate the fluoride-
induced CID pathways (see insets) for a) cis- and b) trans-[(L2Pt)6(4,4’-
bipy)3(tpt)2](PF6)6.
Keywords: cyclometalated complexes · kinetic control ·
.
platinum · self-assembly · supramolecular chemistry
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 4194 –4197