Supramolecular porphyrinic prisms: Coordinative assembly and solution phase X-ray structural characterization

Article abstract of DOI:10.1039/b610025b

Supramolecular porphyrin prisms have been obtained via coordinative self-assembly and characterized by ¹H NMR, PFG NMR, electronic absorption spectroscopy and synchrotron-based measurements of solution phase X-ray scattering and diffraction. Th

Full text of DOI:10.1039/b610025b

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Supramolecular porphyrinic prisms: coordinative assembly and solution
phase X-ray structural characterization{
Suk Joong Lee,a Karen L. Mulfort,a Jodi L. O’Donnell,ab Xiaobing Zuo,c Andrew J. Goshe,c
Paul J. Wesson,d SonBinh T. Nguyen,a Joseph T. Hupp*a and David M. Tiedec
Received (in Berkeley, CA, USA) 13th July 2006, Accepted 15th September 2006
First published as an Advance Article on the web 18th October 2006
DOI: 10.1039/b610025b
Supramolecular porphyrin prisms have been obtained via
coordinative self-assembly and characterized by 1H NMR,
PFG NMR, electronic absorption spectroscopy and synchro-
tron-based measurements of solution phase X-ray scattering
and diffraction.
Highly conjugated porphyrin dimers, trimers, and larger oligomers
are often much broader blue absorbers and much stronger red
absorbers than simple monomers such as tetraphenylporphyrin.1
The striking spectral enhancements, which are due in part to
symmetry reduction and unidirectional electronic delocalization,
make the oligomers potentially very attractive as light harvesters
for energy conversion schemes. This is especially true if conjuga-
tion is introduced in ways that simultaneously sustain reasonably
Scheme 1
long (ca. ns) singlet excited-state lifetimes and inhibit access to
chemically unstable (in the presence of O2) triplet excited states.
Meso-connected ethynyl and butadiynyl groups are two examples
of conjugating linkages meeting these requirements.2
addition of L (essentially nonabsorbing) to 3 in CH2Cl2 as solvent.
The observed binding stoichiometry (inset) is consistent with the
formation of the proposed prism, 33–L3. The observed isosbestic
behavior points to all-or-nothing prism formation, while the
resistance to spectral change beyond the 3 : 3 stoichiometry
illustratesthestabilityoftheputativeprism. Asimilartitrationof2
reaches an end point with y0.67 equivalents of L, consistent with
the formation of 23–L2. All-or-nothing prism formation is again
indicated, but excess L causes further spectral changes, implying
further ligation and break-up of the prism (see ESI{). For both 2
and 3, prism formation is characterized by broadening in the
porphyrin Soret region and intensification and red-shifting in the
Q-band region beyond 700 nm. Similar effects have been noted
previously by Taylor and Anderson in their studies of the
For certain purposes, such as photo-electrode sensitization, it
may be desirable to further organize porphyrins into higher-order
structures. Such structures are potentially useful for minimizing
chromophore aggregation, controlling chromophore spacing, and
facilitating chromophore immobilization (e.g. by allowing for
redundant surface attachment and by enabling control of
orientation).3 They also can be useful for fundamental photo-
physical investigations.4 Herein we report on the formation of well
defined prism-shaped assemblies featuring three, six, or nine
porphyrins and comprising, respectively, triplicate sets of por-
phyrin monomers (1), dimers (2), or trimers (3). We also report on
the in situ X-ray structural characterization of the prisms via both
scattering (overall assembly size) and diffraction (intra-assembly
atom–atom spacing).
As illustrated in Scheme 1 the prisms form via reversible coordi-
nation of triethynylpyridylbenzene (L) by porphyrinic Zn(II) sites.
(For simplicity, solubilizing groups are omitted from the prisms.)
Shown in Fig. 1 are a series of electronic absorption spectra for the
aNorthwestern University, Department of Chemistry, 2145 Sheridan
Road, Evanston, IL 60208, USA. E-mail: j-hupp@northwestern.edu;
Fax: 847-491-7713; Tel: 847-467-3347
bDepartment of Chemistry, Reed College, 3203 SE Woodstock
Boulevard, Portland, OR 97202, USA
cChemistry Division, Argonne National Laboratory, Argonne, IL 60439,
USA
dDepartment of Chemical and Biological Engineering, Northwestern
University, 2145 Sheridan Road, Evanston, IL 60208, USA
{ Electronic supplementary information (ESI) available: Details of
synthesis and characterization, including solution phase X-ray scattering.
See DOI: 10.1039/b610025b
Fig. 1 Spectrophotometric titration of trimer 3 (conc. = 2.0 6 1025 M)
with trigonal ligand L in dichloromethane as solvent. Inset: absorbance
change at 780 nm, showing stoichiometric coordination of L and
illustrating resistance of prism to fragmentation by excess L.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 4581–4583 | 4581

View Article Online
formation of ladder structures from ditopic nitrogen donors and
oligomeric Zn(II) porphyrins.2b These effects are consequences of
porphyrin dimer or trimer planarization and the resulting
enhancements of porphyrin–porphyrin electronic coupling.
A similarly quantitative titration for the formation of 13–L was
unattainable, reflecting prism instability at the low concentrations
used for absorption measurements. This is consistent with the
generally weak interaction of porphyrinic Zn(II) with pyridyl
nitrogen atoms. The fact that prisms based on 2 and 3 do form at
low concentrations, despite weak Zn–N bonding, is illustrative of
beneficial pre-organization effects and the cumulative nature of
assembly stabilization by metal–ligand interactions.
Table 1 Summary of Guinier analysis of free porphyrin monomer,
dimer, and trimer species, and corresponding prism assemblies
˚
Rg/A
Monomer (1)
Prism (13–L)
Dimer (2)
Model
Experiment
Model
Experiment
Model
4.14
4.12 ¡ 0.22
10.63
10.37 ¡ 0.16
7.70
Experiment
Model
Experiment
Model
7.68 ¡ 0.14
12.76
12.87 ¡ 0.21
11.35
Prism (23–L2)
Trimer (3)
Experiment
Model
Experiment
11.38 ¡ 0.24
15.72
15.73 ¡ 4.21
Prism (33–L3)
1
Prism formation was further examined by H NMR spectro-
scopy. Fig. 2 compares spectra for L, 2, and 23–L2 in toluene as
solvent. Salient features are: (a) substantial upfield shifts of peaks
for the phenyl, b-pyridyl, and especially, a-pyridyl protons of L,
evidencing the effects of large porphyrin ring currents,5 and (b)
single sets of signals for each of the three types of protons of
ligated L. The second observation, while consistent with the
formation of a highly symmetrical prism structure is also consistent
with rapid exchange between bound and free versions of L.
Additional spectra with excess L (not shown) showed progressive
downfield shifts, consistent with rapid exchange between bound
and free environments. Spectra for sub-stoichiometric amounts of
L, on the other hand, showed full upfield shifts from the outset,
consistent with the observation in Fig. S3 of the ESI of all-or-
nothing 23–L2 prism formation.
I(q) = I(0)exp(2q2R2 /3)
(1)
g
where I(0) is proportional to the electron-density-contrast
weighted square of the number of electrons in the scattering
object.9 Table 1 compares modeled and experimental measured
porphyrin panel and prism sizes. As expected, the scattering
experiments return larger radii for prisms than component
panels, with prism sizes increasing in the order 13–L , 23–L2 ,
33–L3. The observed close agreement between experiment and
model provides compelling evidence for prism formation and
rules out the formation of larger aggregates (Fig. 3). (Note also
that the electron-density-weighted Rg values are also expected ,
and observed, to be smaller than the hydrodynamic radii for
these systems.)
For 1 + L, analysis of proton shifts indicates that y93% of the
pyridyl nitrogens are ligated under stoichiometric conditions when
[1] = 0.001 M; y96% are ligated (under stoichiometric conditions)
when [1] = 0.01 M.
Wide-angle X-ray scattering (WAXS) was used to obtain
information about interatomic spacings within assemblies.
21
,
˚
Measurements were made in the q-range from 0.09 to 2.7 A
˚
Assembly sizes (hyrodynamic radii) were estimated from self-
diffusion coefficients by using the Stokes–Einstein relation. The
needed coefficients were obtained by the pulsed-field-gradient
NMR method (see ESI).{6,7 The values obtained were 20 ¡ 2,
corresponding to a spatial resolution of 2.4 A (d = 2p/q). These
were then Fourier transformed into real space using the program
GNOM.10 Fig. 4 shows experimental and calculated pair-
distribution plots for a representative prism and panel. Since
scattering intensity scales as the product of the number of electrons
comprising a scattering (diffracting) pair, the most intense peaks
should be those for Zn–Zn pairs. The data, shown in Fig. 4, reveal
˚
35 ¡ 2, and 42 ¡ 7 A, respectively, for 13–L, 23–L2, and 33–L3.
Importantly, identical values were obtained for porphyrin and
linker protons, confirming in each case that the units comprised a
common assembly.
˚
an intense peak at 13.5 A, assignable to scattering by the pair of
Definitive evidence for prism formation was obtained from
solution phase X-ray scattering measurements. These methods are
emerging as a powerful means for structurally characterizing
supramolecular assemblies, particularly those that cannot be
studied by traditional crystallographic methods.8 Small-angle
X-ray scattering (SAXS) data were used to obtain measures of
overall assembly size—specifically, electron-density-weighted radii
of gyration (Rg). The Guinier equation relates q-dependent
(reciprocal-distance dependent) scattering intensities to Rg :
zinc atoms embedded in the porphyrin dimer. Additionally,
however, the data reveal a second Zn–Zn interaction, unique to the
˚
prism, at 18.5 A. In all the prism assemblies, it is possible to detect
the presence of the Zn–Zn scattering interactions which are unique
Fig. 3 Experimental and modelled Guinier plots for 2 (upper) and 23–L2
(lower). Intensities are normalized by Io.
Fig. 2 1H NMR spectra of L, 2, and 23–L2 in toluene.
4582 | Chem. Commun., 2006, 4581–4583
This journal is ß The Royal Society of Chemistry 2006

View Article Online
NMR, UV-Vis, and high-resolution mass spectrometry. The trigonal
ligand, 1,3,5-triethylnyl-pyridylbenzene (L), was prepared in high yield by
Sonogashira coupling of 1,3,5-triethynylbenzene and 4-bromopyridine.12
§ Solution phase X-ray scattering measurements: The assembly solutions
(prisms or panels in toluene) used in the SAXS measurements were diluted
to sufficiently low concentrations such that monodisperse assemblies were
obtained. Consequently, the concentration of 33–L3 (7.6 6 1026 M) was
substantially less than that of the other assemblies (monomer prism 5.2 6
1024 M; dimer prism 1.7 6 1024 M). This is reflected in the greater
uncertainty for Rg for the trimer-based prism. Aggregation of assemblies
was detected in more concentrated solutions as evidenced by nonlinearity
of Guinier scattering plots (upward curvature in small q region).13
" PFG NMR measurements: All PFG NMR experiments were performed
on a Varian Inova 400 spectrometer equipped with an ultra-shielded Doty
PFG probe. The 13-interval PFG NMR pulse sequence with bipolar-
gradient pair suggested by Cotts et al. was used to measure the self-
Fig. 4 Right: Simplified model of dimer prism assembly (minus
solubilizing groups), illustrating metal–metal separation distances. Left:
Experimental and modeled pair-distribution plots, obtained by transform-
ing corresponding wide-angle scattering intensity versus q plots. Maroon
trace: Experimental pair-distribution function (PDF) plot for 23–L2. Red
trace: Experimental plot for free dimeric porphyrin panel (2). Black trace:
Modeled PDF for free dimeric panel based on contributions only from
Zn–Zn interactions. Gray trace: Modeled PDF for 23–L2 based on
contributions only from Zn–Zn interactions. The low amplitude oscilla-
tions in the calculated PDFs are Fourier transform truncation artifacts
that are expected to contribute to the experimental plots as well.
1
diffusivity.14 The resonance frequencies of H nuclei was 400.6 MHz and
the corresponding p/2 pulse widths were 19 ms. The delays before and after
gradient pulses varied between 400 and 500 ms, depending on the width of
the applied gradient pulse. Preliminary experiments indicated that the
residual eddy current was negligible under these conditions. The gradient
pulse widths ranged from 500 to 1250 ms, and gradient intensities ranged
from 5 to 250 G cm21. Diffusion times between 600 and 800 ms were used,
with 16–64 scans on each of nine gradient intensities per sample. Prior to
the acquisition of NMR signals, the adsorbed sample was equilibrated for
at least 20 min at the desired temperature. The error in temperature was less
than 0.1 K.
1 (a) V. S.-Y. Lin, S. G. Dimagno and M. J. Therien, Science, 1994, 264,
1105–1111; (b) H. L. Anderson, Chem. Commun., 1999, 2323–2330; (c)
D. H. Kim and A. Osuka, Acc. Chem. Res., 2004, 37(10), 735–745.
2 (a) H. L. Anderson, Inorg. Chem., 1994, 33, 972–981; (b) P. N. Taylor
and H. L. Anderson, J. Am. Chem. Soc., 1999, 121, 11538–11545; (c)
S. M. LeCours, H.-W. Guan, S. G. DiMagno, C. H. Wang and
M. J. Therien, J. Am. Chem. Soc., 1996, 118, 1497–1503.
3 (a) A. B. F. Martinson, A. M. Massari, S. J. Lee, R. W. Gurney,
K. E. Splan, J. T. Hupp and S. T. Nguyen, J. Electrochem. Soc., 2006,
153(3), A527–A532; (b) K. E. Splan, A. M. Massari and J. T. Hupp,
J. Phys. Chem. B, 2004, 108, 4111–4115; (c) J. Libera, R. Gurney,
C. Schwartz, H. Jin, T.-L. Lee, S. T. Nguyen, J. T. Hupp and
M. Bedzyk, J. Phys. Chem. B, 2005, 109, 1441–1450.
4 (a) R. V. Slone and J. T. Hupp, Inorg. Chem., 1997, 36, 5422–5423; (b)
Y. Kobuke, Struct. Bonding, 2006, 121, 49 and references therein; (c)
L. Flamigni, V. Heitz and J.-P. Sauvage, Struct. Bonding, 2006, 121, 217
and references therein; (d) A. Prodi, M. T. Indelli, C. J. Kleverlaan,
E. Alessio and F. Scandola, Coord. Chem. Rev., 2002, 229, 51–58.
5 G. S. Wilson and H. L. Anderson, Chem. Commun., 1999, 1539–1540.
6 W. Price, Concepts Magn. Reson., 1998, 10, 297.
7 (a) P. Timmerman, J.-L. Weidmann, K. A. Jolliffe, L. J. Prins,
D. N. Reinhoudt, S. Shinkai, L. Frish and Y. Cohen, J. Chem. Soc.,
Perkin Trans. 2, 2000, 2077–2089; (b) B. Olenyuk, M. D. Levin,
J. A. Whiteford, J. E. Shield and P. Stang, J. Am. Chem. Soc., 1999, 121,
10434–10435; (c) W. H. Otto, M. H. Keefe, K. E. Splan, J. T. Hupp and
C. K. Larive, Inorg. Chem., 2002, 41, 6172–6274.
to prism assembly and do not appear for the porphyrin panel
alone. Additional peaks, including negative-going peaks clearly
absent from the simplified modelling, are likely caused by solvent
ordering that creates regions surrounding the porphyrinic
assemblies of higher and lower electron density compared to bulk
solvent. Further modelling of solvent ordering by the prisms and
related assemblies is in progress.
In summary, highly chromophoric porphyrin prisms have been
1
obtained via coordinative self-assembly and characterized via H
NMR, PFG-NMR and electronic absorption spectroscopy. The
largest of the prisms resists break up in the presence of excess
linker ligand. Unequivocal evidence for prism formation in
solution has been obtained from synchrotron-based measurements
of X-ray scattering and diffraction.{§"
We thank the Office of Science, U. S. Department of Energy for
support of our work (Grants DE-FG02-87ER13808 to
Northwestern and contract W-31-109-ENG-38 to the Argonne
National Laboratory). KLM and JLO gratefully acknowledge
fellowships from the Argonne Lab-Grad Program. We are also
especially appreciative of the expert help of the Sector 12 staff at
the Advanced Photon Source, and in particular that from Drs
Soenke Seifert and Nadia Leyarovska.
8 Two other wide-angle X-ray studies (solution phase diffraction studies)
of coordinatively assembled supramolecular systems have been
˚
described: (a) a 6 A resolution synchrotron study from our laboratory
of a hexaporphyrin host–guest system (D. M. Tiede, R. T. Zhang,
L. X. Chen, L. H. Yu and J. S. Lindsey, J. Am. Chem. Soc., 2004, 126,
14054–14062), and; (b) a report on platinum-coordinated rectangle,
square, and cage species (T. Megyes, H. Jude, T. Grosz, I. Bako,
T. Radnai, G. Tarkanyi, G. Palinkas and P. J. Stang, J. Am. Chem.
Soc., 2005, 127, 10731–10738).
Notes and references
{ Synthesis of porphyrin oligomers: Zinc porphyrin monomer, dimer, and
trimer panel species were synthesized according to modified literature
procedures.2b Briefly, a stepwise approach, entailing desilylation with
TBAF followed by monoprotection with trihexylsilane and Cu meditated
oxidative coupling, was used to synthesize the dimer and trimer from the
monomer. For the monomer, trihexylsilane endgroups facilitate chromato-
graphic separation of [5,15-bis(2,6-di-hexoxy)-10,20-(bis-ethynyl)porphyr-
inato]zinc (1a), [5,15-bis(2,6-di-hexoxy)-10-(ethynyl)-20-(trihexylsilyl-
ethynyl)porphyrinato]zinc (1b), and [5,15-bis(2,6-di-hexoxy)-10,20-bis(tri-
hexylsilylethynyl)porphyrinato]zinc (1). They also increase the solubilities of
dimers, trimers and prisms. The trimeric zinc porphyrin (3) was prepared in
reasonable yield by Glaser–Hey coupling11 with CuCl in DCM, treating 1a
with 1b. Dimeric porphyrin (2) was obtained by Glaser–Hey coupling of 1a
in DCM and as a side product during the synthesis of trimeric porphyrin.
The oligomers (i.e. panel units) have been characterized by 1H and 13C{H}
9 A. Guinier and G. Fournet, Small-Angle Scattering of X-rays, Wiley,
New York, 1955.
10 D. I. Svergun, J. Appl. Crystallogr., 1992, 25, 495.
11 P. N. Taylor, A. P. Wylie, J. Huuskonen and H. L. Anderson, Angew.
Chem., Int. Ed., 1998, 37, 986–989.
12 O. Mongin, C. Papamicae¨l, N. Hoyler and A. Gossauer, J. Org. Chem.,
1998, 63, 5568–5580.
13 A. Ducruix, J. P. Guilloteau, M. RiesKautt and A. Tardieu, J. Cryst.
Growth, 1996, 168, 28–39.
14 R. M. Cotts, M. J. R. Hoch, T. Sun and J. T. Markert, J. Magn. Reson.,
1989, 83, 252.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 4581–4583 | 4583