Geometrically restricted intermediates in the self-assembly of an M12L24 cuboctahedral complex

Article abstract of DOI:10.1002/anie.201409216

The self-assembly of a cuboctahedral M₁₂L₂₄ complex is traced by time-dependent NMR spectroscopy and mass spectrometry. The metastable intermediate structures that exist during the self-assembly process are not a chaotic mixture of numerous species, but instead are geometrically restricted. Short-lived M₈L₁₆ (D_4d) and relatively long-lived M₉L₁₈ (D_3h) are fully characterized as major intermediates. Employing a ligand with a smaller bend angle (112°) allows these two species to be kinetically trapped and more clearly observed by NMR spectroscopy. X-ray crystallography shows that M₉L₁₈ has the framework topology predicted by geometric discussion.

Full text of DOI:10.1002/anie.201409216

Angewandte
Chemie
DOI: 10.1002/anie.201409216
Self-Assembly
Geometrically Restricted Intermediates in the Self-Assembly of an
M12L24 Cuboctahedral Complex**
Daishi Fujita, Hiroyuki Yokoyama, Yoshihiro Ueda, Sota Sato, and Makoto Fujita*
Abstract: The self-assembly of a cuboctahedral M12L24 com-
plex is traced by time-dependent NMR spectroscopy and mass
spectrometry. The metastable intermediate structures that exist
during the self-assembly process are not a chaotic mixture of
numerous species, but instead are geometrically restricted.
Short-lived M8L16 (D4d) and relatively long-lived M9L18 (D3h)
are fully characterized as major intermediates. Employing
a ligand with a smaller bend angle (1128) allows these two
species to be kinetically trapped and more clearly observed by
NMR spectroscopy. X-ray crystallography shows that M9L18
has the framework topology predicted by geometric discussion.
the short-lived M8L16 and then the relatively long-lived M9L18.
No other species were identified as intermediates before the
assembly process reached the final M12L24 complex. Our
results thus demonstrate that the geometrical restraints even
apply to the intermediates of self-assembly, and facilitate the
self-assembly of giant polyhedra. By using a ligand with
a reduced bend angle (1128), the M9L18 intermediate was
directly crystallized, and crystallographic analysis showed
that the structure has the predicted framework topology.
To access the self-assembly intermediates, we synthesized
new M12L24 precursor ligand 1, which incorporates a triazine
core with a 2-methyl substituent as a probe for NMR studies.
We first confirmed that ligand 1 self-assembled into M12L24
spherical complex 2 by using the reported procedure.[6]
Ligand 1 (14.7 mmol) was treated with Pd(NO3)2 (11.7 mmol)
and stirred for three days at 808C in [D6]DMSO. The
1H NMR measurement gave a simple spectrum with down-
field shifting of the pyridine peaks and line broadening
S
elf-assembly of Platonic and Archimedean solids from
metal and small-ligand components is currently a topic of
considerable interest.[1–3] Using only simple procedures, the
components spontaneously find the right pathways that lead
to stable polyhedral structures. In most cases, only these
resultant stable polyhedral structures are characterized and
thoroughly discussed, and far less attention has been paid to
the metastable intermediate structures that exist during the
self-assembly process.[4] This is, in part, due to the difficulties
in characterizing such transient, short-lived species, but also
largely due to the prejudice that a number of intermediates/
pathways exist before the assembly process reaches the final
stable structures.[5]
(Figure 1b) that are characteristic of M12L24 assemblies.[6]
A
single diffusion band in the 1H diffusion-ordered NMR
spectroscopy (DOSY; see Figure S6 in the Supporting Infor-
n+
mation, SI) and a series of [Pd12L24(BF4)24Àn
]
ion peaks in
the cold-spray ionization mass spectrometry (CSI-MS)[7]
(Figure 2a) analyses supported the exclusive formation of 2.
X-ray crystallographic analysis of a single crystal grown from
the DMSO solution confirmed the formation of 2, which has
cuboctahedral symmetry (Figure 1d).
The polyhedral structures are, however, formed under
a geometrical restriction known as Euler’s polyhedron
theorem. We assumed that this geometrical restriction
would also apply to the intermediates of self-assembly, thus
limiting the number of metastable intermediate structures
and self-assembly pathways. Accordingly, the self-assembly of
M12L24 complex 2 from PdII ions and bent ligand 1 was
carefully traced by time-dependent NMR spectroscopy and
mass spectrometry. Surprisingly, both analyses consistently
revealed the formation of just two intermediate species: first
We then set out to examine the self-assembly intermedi-
ates that exist prior to complete convergence to the thermo-
dynamically stable M12L24 structure. 1H NMR spectra of
a nonheated (rt, 30 min) mixture of ligand 1 and Pd(NO3)2
showed, in addition to the known M12L24 peaks, other
noticeable signals (Figure 1c), which are presumably derived
from metastable intermediates. An important clue was
obtained from careful CSI-MS analysis. Comparison of the
spectra of the converged M12L24 and the nonconverged self-
assembly mixtures showed two distinct series of ions that were
only observed in the nonconverged spectrum (Figure 2a,b).
[*] Dr. D. Fujita, H. Yokoyama, Dr. Y. Ueda, Dr. S. Sato,
Prof. Dr. M. Fujita
By MS analysis at high resolution, these two ion series were
Department of Applied Chemistry, Graduate School of Engineering
The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656 (Japan)
E-mail: mfujita@appchem.t.u-tokyo.ac.jp
n+
reliably assigned to [Pd8L16(BF4)16Àn
]
(n = 8–11) and
n+
[Pd9L18(BF4)18Àn
]
(n = 8–12), respectively. The isotopic dis-
tribution patterns of each peak clearly match the simulated
patterns (Figures S10–S15).
[**] This work was supported by the Japanese Ministry of Education,
Culture, Sports, Science, and Technology (MEXT) with their Grant-
in-Aids for Scientific Research on Innovative Areas (25102007) and
for Young Scientists (A) (24685010). The synchrotron X-ray
crystallography was performed at the BL1A beamline at the KEK
Photon Factory (2011G522) and at the BL41XU beamline at SPring-8
(2014A0042).
To verify that these Pd8L16 and Pd9L18 species are trapped
as metastable intermediates rather than as side products, we
examined whether Pd8L16 and Pd9L18 would spontaneously
convert to the stable Pd12L24 upon heating. Figure 2c,d shows
magnified views of representative peaks assigned to Pd8L16
(11 +) and Pd9L18 (11 +). We observed that Pd8L16 disap-
peared selectively over Pd9L18 after heating for 1 h at 608C.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201409216.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Figure 1. a) Schematic representation of the self-assembly of M12L24
complex 2. b,c) 1H NMR spectra (500 MHz, [D6]DMSO, 300 K) after
mixing ligand 1 and PdII at b) 808C for 3 d, and c) rt for 30 min.
À
d) X-ray crystal structure of 2 (BF4 salt). Hydrogen atoms, counter
ions, and solvent molecules have been omitted for clarity.
Additional heating for 6 h at 808C caused the remaining
Pd9L18 species to disappear as well. The comprehensive time-
dependent behavior in the CSI-MS study (see Section 3.1 of
SI) demonstrates the existence of short-lived Pd8L16 and
relatively long-lived Pd9L18 as metastable intermediates in the
formation of Pd12L24.
Figure 2. CSI-TOF mass spectra of ligand 1 and PdII (BF4À salt) under
different self-assembly conditions: a) 808C, 6 h. b) rt, 30 min.
c,d) Comparison of expanded spectra of c) the 11+ peak of Pd8L16
([Pd8L16(BF4)5]11+) and d) the 11+ peak of Pd9L18 ([Pd9L18(BF4)7]11+
after 30 min at rt (i), 1 h at 608C (ii), and 6 h at 808C (iii).
)
The framework topologies of Pd8L16 and Pd9L18 inter-
mediates could be predicted by mathematical discussion.
Under the postulate that each vertex (M) connects four edges
(L), the possible polyhedral topologies for M8L16 and M9L18
are limited to one each. Figure 3 shows models of these
unique topologies, 3 for M8L16 (D4d) and 4 for M9L18 (D3h),
which do not express any significant distortion. Interestingly,
no allowed structures exist for MnL2n, in which n ꢀ 5 and n = 7.
A possible structure for n = 6 is unrealistically distorted as our
ligand bend angle is much larger than the ideal 908 required
for the n = 6 structure. M10L20 and M11L22 are geometrically
allowed but already very close to the final M12L24. Thus, they
may exist as momentary, unobservable intermediates. All the
possible topologies for MnL2n (n < 12) are summarized in
Figure S23.
Figure 3. Geometrically allowed topologies for M8L16 and M9L18 compo-
sitions with D4d and D3h symmetries, respectively. Both M8L16 and
M9L18 have two edge environments (colored in red and blue).
ligand environments, illustrated by red and blue edges in
Figure 3. The ratios of the two ligands are 1:1 for 3 (D4d) and
2:1 for 4 (D3h), respectively. These environmental differences
were clearly distinguishable from the signals of the methyl
Based on both the experimental and theoretical discus-
sions, we went over the 1H NMR experiments to confirm the
intermediate structures. Both 3 and 4 have two different
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Figure 4. Time-dependent 1H NMR spectra (500 MHz, [D6]DMSO,
300 K) for the self-assembly of ligand 1 and PdII: a) rt, 30 min,
b) 608C, 1 h, and c) 808C, 6 h. The peak labeled A corresponds to the
methyl protons of Pd12L24. Pairs of B, B’ (1:1) and C, C’ (2:1) are
assigned to Pd8L16 and Pd9L18, respectively. Unassignable minor peaks
were detected at 2.57, 2.70, 2.71, and 2.83 ppm.
Figure 5. X-ray crystal structure of Pd9L18 complex 7. View along the C3
À
(left) and C2 (right) axes. Hydrogen atoms and counter ions (BF4
)
have been omitted for clarity. Crystal data: space group P21/m,
a=53.598(11) ꢀ, b=43.824(9) ꢀ, c=53.599(11) ꢀ, b=120.01(3)8,
V=109022(38) ꢀ3, Z=6, R1 =0.1244, wR2=0.3667, GOF=1.177.
Experimental Section
probe on ligand 1 (Figure 4). The single peak A was assigned
to 2, whereas peaks B and B’ were assigned to 3 and peaks C
and Cꢀ to 4. The ratio of B and B’ is 1:1, and that of C and Cꢀ is
2:1; this is in good agreement with the topologies and
symmetries of 3 and 4, respectively. Peaks B and B’
disappeared after heating for 1 h at 608C, and peaks C and
Cꢀ subsequently disappeared upon additional heating for 6 h
at 808C, which is fully consistent with the time-dependent MS
study.
Preparation of M12L24 complexes 2: Ligand 1 (14.7 mmol) was treated
with Pd(NO3)2 or Pd(BF4)2·(CH3CN)4 (11.7 mmol) in DMSO
(0.74 mL) at 808C for 3 d. 1H NMR spectroscopy confirmed the
quantitative formation of M12L24 spherical complex 2. 1H NMR
(500 MHz, [D6]DMSO, 300 K, NO3À salt) d = 9.77 (br, 96H), 8.81 (br,
96H), 2.71 ppm (br, 72H). Diffusion coefficient ([D6]DMSO, 300 K,
NO3 salt): D = 5.0 ꢀ 10À11 m2 sÀ1
.
13C NMR (150 MHz, [D6]DMSO,
À
300 K, BF4À salt) d = 179.3 (C), 168.3 (C), 152.8 (CH), 146.1 (C), 126.1
(CH), 26.2 ppm (CH3). The following signals are the signals with the
highest intensity among the isotope patterns of each ion peak. m/z
calcd for [MÀ14(BF4À)]14+ 580.5945, found 580.5942; calcd for
[MÀ13(BF4À)]13+ 631.8714, found 631.8708; calcd for [MÀ12-
(BF4À)]12+ 691.7777, found 691.7774; calcd for [MÀ11(BF4À)]11+
762.4851, found 762.4844; calcd for [MÀ10(BF4À)]10+ 847.4340,
found 847.4339; calcd for [MÀ9(BF4À)]9+ 951.3717, found 951.3718;
calcd for [MÀ8(BF4À)]8+ 1081.0437, found 1081.0453; calcd for
[MÀ7(BF4À)]7+ 1247.9076, found 1247.9085.
When ligand 5 with a smaller bend
angle (1128) was employed, its M8L16 and
M9L18 complexes, 6 and 7, were kinetically
trapped because their framework distor-
tions are considerably reduced. Complex-
ation at 808C for 14 h afforded a mixture of
6 and 7, from which 7 was isolated as single
crystals. The crystal structure confirmed
Received: September 17, 2014
the predicted D3h framework of 7 (Figure 5). The sum of the
cis-N-Pd-N angles at every PdII ion is close to 3608 (359.88–
360.18), which indicates that there are no distortions around
any of the PdII centers.
Published online: && &&, &&&&
Keywords: geometrical restriction · metastable intermediates ·
.
palladium · self-assembly
In summary, we have designed and performed experi-
ments to uncover self-assembly intermediates that exist in the
course of the formation of an M12L24 cuboctahedral complex.
The time-dependent NMR and CSI-MS analyses consistently
showed the exclusive formation of just two metastable
intermediates: M8L16 and M9L18. This suggests that the
geometrical restriction begins from the early stages of the
self-assembly process, and may serve to guide the self-
assembly into the final giant polyhedral structures. The
framework topology of the intermediates was mathematically
predicted. This newly disclosed self-assembly pathway is
reminiscent of the folding funnel hypothesis for protein
folding,[8] in which proteins drop down to their ground state
guided by a number of trappings in local minima along the
way.
[1] Earlier examples: a) M. Fujita, D. Oguro, M. Miyazawa, H. Oka,
K. Yamaguchi, K. Ogura, Nature 1995, 378, 469 – 471; b) D. L.
Caulder, R. E. Powers, T. N. Parac, K. N. Raymond, Angew.
Chem. Int. Ed. 1998, 37, 1840 – 1842; Angew. Chem. 1998, 110,
1940 – 1943; c) B. F. Abrahams, S. J. Egan, R. Robson, J. Am.
Chem. Soc. 1999, 121, 3535 – 3536; d) A. Mꢁller, V. P. Fedin, C.
Kuhlmann, H. BÅgge, M. Schmidtmann, Chem. Commun. 1999,
927 – 928; e) B. Olenyuk, M. D. Levin, J. A. Whiteford, J. E.
Shield, P. J. Stang, Nature 1999, 398, 796 – 799; f) M. Ni, A. Tassar,
T. Antoun, O. M. Yaghi, J. Am. Chem. Soc. 2005, 127, 12752 –
12753.
[2] Recent examples: a) M. Liu, W. Liao, C. Hu, S. Du, H. Zhang,
Angew. Chem. Int. Ed. 2012, 51, 1585 – 1588; Angew. Chem. 2012,
124, 1617 – 1620; b) A, Stephenson, D. Sykes, M. D. Ward,
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Dalton Trans. 2013, 42, 6756 – 6767; c) Q. F. Sun, S. Sato, M.
Fujita, Nat. Chem. 2012, 4, 330 – 333; d) K. Xiong, F. Jiang, Y. Gai,
D. Yuan, L. Chen, M. Wu, K. Su, M. Hong, Chem. Sci. 2012, 3,
2321 – 2325; e) R. A. Bilbeisi, J. K. Clegg, N. Elgrishi, X. d.
Hatten, M. Devillard, B. Breiner, P. Mal, J. R. Nitschke, J. Am.
Chem. Soc. 2012, 134, 5110 – 5119; f) M. Otte, P. F. Kuijpers, O.
2002, 21, 3552 – 3561; f) Y.-R. Zheng, P. J. Stang, J. Am. Chem.
Soc. 2009, 131, 3487 – 3489; g) V. M. Cangelosi, T. G. Carter, L. N.
Zakharov, D. W. Johnson, Chem. Commun. 2009, 5606 – 5608;
h) A. Marquis, J. Kintzinger, R. Graff, P. N. Baxter, J.-M. Lehn,
Angew. Chem. Int. Ed. 2002, 41, 2760 – 2764; Angew. Chem. 2002,
114, 2884 – 2888; i) S. J. Dalgarno, N. P. Power, J. E. Warren, J. L.
Atwood, Chem. Commun. 2008, 1539 – 1541; j) W. Jiang, A.
Schꢂfer, P. C. Mohr, C. A. Schalley, J. Am. Chem. Soc. 2010, 132,
2309 – 2320; k) N. Fatin-Rouge, S. Blanc, E. Leize, A. Van Dors-
selaer, P. Baret, J. L. Pierre, A. M. Albrecht-Gary, Inorg. Chem.
2000, 39, 5771 – 5778; l) M. Elhabiri, J. Hamacek, J. C. G. Bunzli,
A. M. Albrecht-Gary, Eur. J. Inorg. Chem. 2004, 51 – 62; m) B.
Hasenknopf, J. M. Lehn, N. Boumediene, E. Leize, A. Van Dors-
selaer, Angew. Chem. Int. Ed. 1998, 37, 3265 – 3268; Angew.
Chem. 1998, 110, 3458 – 3460; n) M. Albrecht, S. Dehn, R.
Frohlich, Angew. Chem. Int. Ed. 2006, 45, 2792 – 2794; Angew.
Chem. 2006, 118, 2858 – 2860; o) C. Su, Y. Cai, C. Chen, F. Lissner,
B. Kang, W. Kaim, Angew. Chem. Int. Ed. 2002, 41, 3371 – 3375;
Angew. Chem. 2002, 114, 3519 – 3523; p) J. A. Sheikh, H. S. Jena,
A. Adhikary, S. Khatua, S. Konar, Inorg. Chem. 2013, 52, 9717 –
9719; q) C. G. Claessens, M. J. Vicente-Aranab, T. Torres, Chem.
Commun. 2008, 6378 – 6380; r) Y. Tsujimoto, T. Kojima, Shuichi
Hiraoka, Chem. Sci. 2014, 5, 4167 – 4172; s) M. Yoneya, T.
Yamaguchi, S. Sato, M. Fujita, J. Am. Chem. Soc. 2012, 134,
14401 – 14401.
´
´
Troeppner, I. Ivanovic-Burmazovic, J. N. H. Reek, B. d. Bruin,
Chem. Eur. J. 2013, 19, 10170 – 10178; g) S. Pasquale, S. Sattin,
E. C. Escudero-Adan, M. Martinez-Belmonte, J. d. Mendoza,
Nat. Commun. 2012, 3, 785; h) Q. F. Sun, J. Iwasa, D. Ogawa, Y.
Ishido, S. Sato, T. Ozeki, Y. Sei, K. Yamaguchi, M. Fujita, Science
2010, 328, 1144 – 1147; i) D. Fujita, K. Suzuki, S. Sato, M. Yagi-
Utsumi, Y. Yamaguchi, N. Mizuno, T. Kumasaka, M. Takata, M.
Noda, S. Uchiyama, K. Kato, M. Fujita, Nat. Commun. 2012, 3,
1093.
[3] Reviews: a) M. Han, D. M. Engelhard, G. H. Clever, Chem. Soc.
Rev. 2014, 43, 1848 – 1860; b) M. M. J. Smulders, I. A. Riddell, C.
Browne, J. R. Nitschke, Chem. Soc. Rev. 2013, 42, 1728 – 1754;
c) K. Harris, D. Fujita, M, Fujita, Chem. Commun. 2013, 49,
6703 – 6712; d) T. R. Cook, Y. R. Zheng, P. J. Stang, Chem. Rev.
2013, 113, 734 – 777; e) A. Mꢁller, P. KÅgerler, A. W. M. Dress,
Coord. Chem. Rev. 2001, 222, 193 – 218; f) D. L. Caulder, K. N.
Raymond, Acc. Chem. Res. 1999, 32, 975 – 982.
[4] a) A. V. Davis, R. M. Yeh, K. N. Raymond, Proc. Natl. Acad. Sci.
USA 2002, 99, 4793 – 4796; b) J. M. Lehn, Proc. Natl. Acad. Sci.
USA 2002, 99, 4763 – 4768; c) J. M. Lehn, Chem. Soc. Rev. 2007,
36, 151 – 160.
[5] a) M. D. Levin, P. J. Stang, J. Am. Chem. Soc. 2000, 122, 7428 –
7429; b) N. Fatin-Rouge, S. Blanc, A. Pfeil, A. Rigault, A.-M.
Albrecht-Gray, J.-M. Lehn, Helv. Chim. Acta 2001, 84, 1694 –
1711; c) S. Sato, Y. Ishido, M. Fujita, J. Am. Chem. Soc. 2009,
131, 6064 – 6065; d) H. N. Miras, G. J. T. Cooper, D.-T. Long, H.
Bçgge, A. Mꢁller, C. Streb, L. Cronin, Science 2010, 327, 72 – 74;
e) J. J. Pak, J. Greaves, D. J. McCord, K. Shea, Organometallics
[6] M. Tominaga, K. Suzuki, M. Kawano, T. Kusukawa, T. Ozeki, S.
Sakamoto, K. Yamaguchi, M. Fujita, Angew. Chem. Int. Ed. 2004,
43, 5621 – 5625; Angew. Chem. 2004, 116, 5739 – 5743.
[7] a) S. Sakamoto, M. Fujita, K. Kim, K. Yamaguchi, Tetrahedron
2000, 56, 955 – 964; b) K. Yamaguchi, J. Mass Spectrom. 2003, 38,
473 – 490.
[8] R. H. Pain, Mechanisms of Protein Folding, 2nd ed., Oxford
University Press, Oxford, 2000.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Self-Assembly
Geometrically dictated: Due to the geo-
metrically restricted number of possible
structures, the intermediates of self-
assembly are rather relatively ordered
than a chaotic mixture of numerous
species. Two prominent metastable
intermediates, M8L16 and M9L18, are well
characterized during the self-assembly of
an M12L24 cuboctahedral complex.
D. Fujita, H. Yokoyama, Y. Ueda, S. Sato,
M. Fujita*
&&&& — &&&&
Geometrically Restricted Intermediates in
the Self-Assembly of an M12L24
Cuboctahedral Complex
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!