Observation of reaction intermediates and kinetic mistakes in a remarkably slow self-assembly reaction

Article abstract of DOI:10.1039/b914750k

Several intermediates and oligomeric mistakes in a metal-ligand self-assembly reaction are identified by ¹H NMR, MALDI-MS, and XRD, providing evidence in support of multiple pathways in the "free-for- all" self-assembly process.

Full text of DOI:10.1039/b914750k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Observation of reaction intermediates and kinetic mistakes
in a remarkably slow self-assembly reactionw
Virginia M. Cangelosi, Timothy G. Carter, Lev N. Zakharov and Darren W. Johnson*
Received (in Austin, TX, USA) 21st July 2009, Accepted 1st August 2009
First published as an Advance Article on the web 20th August 2009
DOI: 10.1039/b914750k
Several intermediates and oligomeric mistakes in a metal–ligand
self-assembly reaction are identified by 1H NMR, MALDI-MS,
and XRD, providing evidence in support of multiple pathways in
the ‘‘free-for-all’’ self-assembly process.
Self-assembly is the most efficient route to prepare discrete
supramolecules,1,2 but the complexity of the dynamic self-assembly
process is still poorly understood.3 It is generally accepted that
Scheme 1 Ligands and As2L2Cl2 macrocycles.11
this process involves the correction of misconnections and
random oligomeric errors, quickly leading from kinetic inter-
mediates to the discrete, thermodynamic product. Despite this
widespread assumption, there exist only a few examples of the
observation of oligomeric errors (‘‘kinetic mistakes’’) and in
these cases they are not structurally characterized.4 Additional
examples involving the observation of self-assembly inter-
mediates exist, but because most of these reactions between
metals (M) and organic ligands (L) occur spontaneously and
quickly, it is difficult to observe kinetic intermediates that form
prior to the final thermodynamic product.5 Intermediates have
been observed during the titration of one component (M or L)
into the other, but this approach is limited in that only the
equilibrium product of each titration is observed in solution.6,7
Rarely are kinetic intermediates stable enough to isolate from
solution.7,8
Fig. 1 X-Ray crystal structure representations of (a) As2(Lb)2Cl2
(stick), (b) As2(Lb)2Cl2 (van der Waals radii), and (c) As2LcCl4 (stick).
A mixture of syn- and anti-As2L2Cl2 macrocycles forms in
solution over the course of several days (Scheme 1) when rigid
dithiol ligands H2La,9 H2Lb and H2Lc10 are individually
treated with arsenic trichloride. X-Ray quality crystals of
syn-As2Lb Cl2 were grown by diffusion of pentane into a
Since the speed by which self-assembly reactions occur
precludes the observation of intermediates and kinetic
mistakes, slowing down the process could allow a better
understanding of metal-directed assembly. This may
ultimately lead to the ability to incorporate design features
and specific properties into supramolecular assemblies with
more predictability. Paradoxically, for the self-assembly of
discrete species to occur in a reasonable amount of time, fast
kinetics are required in the forming and breaking of individual
supramolecular interactions (typically either labile metal–
ligand bonds or H-bonds). In this communication, we describe
the relatively slow self-assembly of M2L2 supramolecular
macrocycles. This reaction occurs over the course of several
days which allows for the observation and identification of
several intermediate species and ‘‘kinetic mistakes’’ along the
pathway to macrocycle formation.
2
solution of As2Lb Cl2 in chloroform. The crystal structure
2
(Fig. 1a and b) reveals an As–As distance of 4.45 A and a
distance of 4.26 to 7.45 A between the methyl carbons on
opposite ligands, leaving a small cavity that is devoid of a guest.
When As2La Cl2 (4a) and As2Lb Cl2 (4b) are prepared in
2
2
d-chloroform, the reaction progress can be monitored by
observing the changes to the methylene region of the
1H NMR spectrum (Fig. 2 for 4a and Fig. S1 (ESIw) for 4b).
In each case, as H2L is consumed, its resonances are replaced
with those corresponding to several different reaction inter-
mediates. These, in turn, are replaced with the resonances
for 4, the fully formed macrocycles. While several of the
methylene region resonances overlap, it was possible to
identify three of these intermediate species by symmetry. The
first species observed upon treatment of H2L with AsCl3 is
HL(AsCl2) (1), which appears as a doublet for the unbound
(CH2SH) end and a singlet for the bound (CH2S(AsCl2)) end.
The next observed species is HL(AsCl)LH (2) which appears
as an ABq for the bound ((CH2S)2AsCl) end and a doublet
for the unbound (CH2SH) end, which is coincidental with
the resonance for H2L or 1. The third species that can be
identified by 1H NMR is L(AsCl2)2 (3), which appears
downfield as a singlet since the ligand symmetrically spans
two As centers.
Department of Chemistry and Materials Science Institute,
University of Oregon, and the Oregon Nanoscience and
Microtechnologies Institute (ONAMI), Eugene, OR 97403-1253,
USA. E-mail: dwj@uoregon.edu; Fax: +1-541-346-0487;
Tel: +1-541-346-1695
w Electronic supplementary information (ESI) available: Experimental
procedures, 1H NMR, MALDI, and X-ray data. CCDC 741266,
741267. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/b914750k
ꢀc
This journal is The Royal Society of Chemistry 2009
5606 | Chem. Commun., 2009, 5606–5608

View Article Online
Chart 1 Oligomeric kinetic mistakes.
are corrected during the course of the reaction because
everything remains soluble and all resonances except those
for As2L2Cl2 disappear eventually. These oligomeric species
observed by MALDI-MS are the most likely culprits.
Fig. 2 CH2 region of 1H NMR spectra of reaction of H2La
with AsCl3 after (a) 35, (b) 81, (c) 122, (d) 186, (e) 366, (f) 1378, and
(g) 5650 minutes. Structures of 1a–4a are shown in Scheme 2.
The isolation of compounds 1–3 was unsuccessful until
prepared from a ligand with additional coordinating groups
(methyl ether in H2Lc). While macrocycle was typically
isolated by crystallization of the reaction mixture of H2Lc
and AsCl3,10 one attempt led to the isolation of 3c (Fig. 1c).
The intermediate is stabilized by As–O secondary bonding
interactions14 and crystal packing is dictated by the As–p
interaction.15 This experiment further supports the idea that
structure 3 is an intermediate or a kinetic mistake in the
self-assembly of As2L2Cl2 (Scheme 2).
In conclusion, we have shown a metal-directed self-assembly
reaction in which several intermediates and kinetic mistakes
can be identified. This provides insight into the complexity of
the self-assembly process—as revealed by the multiple possible
pathways—even for a simple dinuclear species. While most
self-assembly reactions occur too quickly to observe without
stopped flow kinetics, the dynamic covalent16 nature of the
As–S bond and the steric bulk of the ligands make this
reaction slow enough to observe by 1H NMR, providing
exquisite structural detail. This shows that supramolecular
chemistry based on main group elements not only leads to
new structure types,17 but also can add valuable insight into
the nature of self-assembly, which could be applicable to
understanding the formation of nanoparticles,18 polymers19
and monolayers20 by self-assembly.
Scheme 2 Self-assembly of As2L2Cl2.
The identities of 1b–4b were also confirmed by MALDI-MS
([1b + Na + 2H]+, 394.9, calc. 394.9; [2b + Na + 4H]+,
587.0, calc. 587.1; [3b + Na]+, 536.8, calc. 536.8; [4b + Na]+,
690.9, calc. 690.9), when As2L2Cl2 was prepared from H2Lb.w
These species could either be intermediates or kinetic mistakes
that are corrected in the self-assembly of As2L2Cl2 (4).
It is possible that the self-assembly reaction is occurring
simultaneously through several competing pathways as
outlined in Scheme 2.12z
While the overlapping resonances of H2L, 1, 2, 3, and
As2L2Cl2 (4) make it difficult to accurately integrate the
NMR spectra, it is clear that at least one additional discrete
We gratefully acknowledge funding from the National
Science Foundation (CAREER award CHE-0545206).
D.W.J. is a Cottrell Scholar of Research Corporation. This
material is based upon work supported by the U.S.
Department of Education under Award No. P200A070436
(V.M.C.) and an NSF Integrative Graduate Education and
Research Traineeship No. DGE-0549503 (T.G.C.). The
purchase of the MALDI Mass Spectrometer was made
possible by a grant from the NSF (CHE-0639170).
species exists in solution,13 likely
a larger oligomeric
‘‘kinetic mistake’’. These kinetic mistakes could easily form
reversibly in the free-for-all chaos of the self-assembly reaction
and then equilibrate to As2L2Cl2. After 5 minutes, and under
the same conditions as the NMR experiments, several
oligomeric species (Chart 1) were identified by MALDI-MS
([5b + Na + 2H]+, 728.9, calc. 728.9; [6b + Na]+, 870.7, calc.
870.7; [7b + Na + 4H]+, 921.0, calc. 921.0; [8b + Na + 2H]+,
1062.8, calc. 1062.9; [9b + Na]+, 1206.7, calc. 1204.7).w
These species could be expected to have 1H NMR
resonances that overlap with the other intermediates. We do
not know for sure that every one of these oligomeric species is
present in solution during the reaction, as they could be
formed as a concentration effect upon evaporation of the
MALDI sample, yet after 30 minutes they are no longer
observed by MALDI. It is clear that some kinetic mistakes
Notes and references
z Unfortunately the slow kinetics and complexity of the reaction do
not allow for the measurement of rate constants. EXSY NMR
experiments were carried out on the reaction of AsCl3 with a
monofunctional model ligand (2-mercaptomethylnaphthalene) at
120 1C, but no ligand exchange was observed.
1 C. Piguet, G. Bernardinelli and G. Hopfgartner, Chem. Rev., 1997,
97, 2005; M. Albrecht, Chem. Rev., 2001, 101, 3457; M. Elhabiri
and A. M. Albrecht-Gary, Coord. Chem. Rev., 2008, 252, 1079.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 5606–5608 | 5607

View Article Online
2 R. W. Saalfrank, H. Maid and A. Scheurer, Angew. Chem., Int.
Ed., 2008, 47, 8794; S. Leininger, B. Olenyuk and P. J. Stang,
Chem. Rev., 2000, 100, 853; G. F. Swiegers and T. J. Malefetse,
Chem. Rev., 2000, 100, 3483; D. L. Caulder and K. N. Raymond,
J. Chem. Soc., Dalton Trans., 1999, 1185.
3 A. V. Davis, R. M. Yeh and K. N. Raymond, Proc. Natl. Acad.
Sci. U. S. A., 2002, 99, 4793; J. M. Lehn, Proc. Natl. Acad. Sci.
U. S. A., 2002, 99, 4763; J. M. Lehn, Chem. Soc. Rev., 2007, 36,
151.
4 M. Fujita, S. Nagao and K. Ogura, J. Am. Chem. Soc., 1995, 117,
1649; P. Baxter, J. M. Lehn, A. Decian and J. Fischer, Angew.
Chem., Int. Ed. Engl., 1993, 32, 69.
5 J. J. Pak, J. Greaves, D. J. McCord and K. J. Shea, Organometallics,
2002, 21, 3552; S. Sato, Y. Ishido and M. Fujita, J. Am. Chem. Soc.,
2009, 131, 6064.
6 A. Pfeil and J. M. Lehn, Chem. Commun., 1992, 838; M. D. Levin
and P. J. Stang, J. Am. Chem. Soc., 2000, 122, 7428;
N. Fatin-Rouge, S. Blanc, E. Leize, A. Van Dorsselaer, P. Baret,
J. L. Pierre and A. M. Albrecht-Gary, Inorg. Chem., 2000, 39,
5771; N. Fatin-Rouge, S. Blanc, A. Pfeil, A. Rigault,
A. M. Albrecht-Gary and J. M. Lehn, Helv. Chim. Acta, 2001,
84, 1694; A. Marquis, J. P. Kintzinger, R. Graff, P. N. W. Baxter
and J. M. Lehn, Angew. Chem., Int. Ed., 2002, 41, 2760;
J. Hamacek, S. Blanc, M. Elhabiri, E. Leize, A. Van Dorsselaer,
C. Piguet and A. M. Albrecht-Gary, J. Am. Chem. Soc., 2003, 125,
1541; M. Elhabiri, J. Hamacek, J. C. G. Bunzli and
A. M. Albrecht-Gary, Eur. J. Inorg. Chem., 2004, 51; E. Leize,
A. V. Dorsselaer, R. Kramer and J. M. Lehn, Chem. Commun.,
1993, 990.
7 N. Takeda, K. Umemoto, K. Yamaguchi and M. Fujita, Nature,
1999, 398, 794.
8 B. Hasenknopf, J. M. Lehn, N. Boumediene, E. Leize and A.
Van Dorsselaer, Angew. Chem., Int. Ed., 1998, 37, 3265;
M. Albrecht, S. Dehn and R. Frohlich, Angew. Chem., Int. Ed.,
2006, 45, 2792.
9 W. J. Vickaryous, E. R. Healey, O. B. Berryman and
D. W. Johnson, Inorg. Chem., 2005, 44, 9247.
10 V. M. Cangelosi, L. N. Zakharov, S. A. Fontenot, M. A. Pitt and
D. W. Johnson, Dalton Trans., 2008, 3447.
11 Both chlorine atoms are on the same side of the macrocyclic cavity
in the syn macrocycle and on opposite sides in the anti-macrocycle.
12 2 and 3 could be both intermediates and kinetic mistakes.
13 The ABqs in the 1H NMR corresponding to 2a and 4a overlap.
However, before the signal for 4a becomes prominent, the
asymmetry of the signal is clear, suggesting an overlapping
resonance. The signal for 2b is clearly an ABq (see Fig. S1, ESIw).
14 T. G. Carter, E. R. Healey, M. A. Pitt and D. W. Johnson, Inorg.
Chem., 2005, 44, 9634.
15 H. Schmidbaur and A. Schier, Organometallics, 2008, 27, 2361.
16 S. J. Rowan, S. J. Cantrill, G. R. L. Cousins, J. K. M. Sanders and
J. F. Stoddart, Angew. Chem., Int. Ed., 2002, 41, 898; V. E. Campbell,
X. d. Hatten, N. Delsuc, B. Kauffmann, I. Huc and J. R. Nitschke,
Chem.–Eur. J., 2009, 15, 6138.
17 M. A. Pitt and D. W. Johnson, Chem. Soc. Rev., 2007, 36, 1441.
18 M. A. Watzky, E. E. Finney and R. G. Finke, J. Am. Chem. Soc.,
2008, 130, 11959.
19 C. Alberto, Macromol. Rapid Commun., 2002, 23, 511.
20 D. K. Schwartz, Annu. Rev. Phys. Chem., 2001, 52, 107.
ꢀc
This journal is The Royal Society of Chemistry 2009
5608 | Chem. Commun., 2009, 5606–5608