Polyvalent interactions in unnatural recognition processes

Article abstract of DOI:10.1021/jo030283o

The synthesis of two cluster compounds, one containing six secondary dialkylammonium ion centers and the other possessing six benzo-m-phenylene[25] crown-8 (BMP25C8) macrocycles, both appended to hexakis(thiophenyl)benzene cores, is described. The binding of these clusters with complementary mono- and divalent ligands is investigated with NMR spectroscopy to probe polyvalency in these unnatural recognition systems. The ability of the two different families of clusters to bind complementary monovalent ligands is compared with that of the monovalent receptor pair, namely the dibenzylammonium ion and BMP25C8. This comparison is made possible by determining an average association constant (K_AVE) for the binding of each recognition site on the cluster with the corresponding monovalent ligand. We have found that the clustering of recognition sites together in one molecule is detrimental to their individual abilities to bind monovalent ligands. In the case of the polyvalent interaction between the hexakisBMP25C8 cluster and divalent dialkylammonium ions, an association constant, K_POLY, was calculated from the value of K _AVE determined for the complexation of the individual component recognition sites. This polyvalent interaction is significantly stronger than that associated with the averaged monovalent interactions.

Full text of DOI:10.1021/jo030283o

P olyva len t In ter a ction s in Un n a tu r a l Recogn ition P r ocesses
J ames N. Lowe, David A. Fulton, Sheng-Hsien Chiu, Arkadij M. Elizarov, Stuart J . Cantrill,
Stuart J . Rowan, and J . Fraser Stoddart*
California NanoSystems Institute and Department of Chemistry and Biochemistry,
University of California, Los Angeles, 405 Hilgard Avenue, Los Angeles, California 90095
stoddart@chem.ucla.edu
Received September 4, 2003
The synthesis of two cluster compounds, one containing six secondary dialkylammonium ion centers
and the other possessing six benzo-m-phenylene[25]crown-8 (BMP25C8) macrocycles, both appended
to hexakis(thiophenyl)benzene cores, is described. The binding of these clusters with complementary
mono- and divalent ligands is investigated with NMR spectroscopy to probe polyvalency in these
unnatural recognition systems. The ability of the two different families of clusters to bind
complementary monovalent ligands is compared with that of the monovalent receptor pair, namely
the dibenzylammonium ion and BMP25C8. This comparison is made possible by determining an
average association constant (KAVE) for the binding of each recognition site on the cluster with the
corresponding monovalent ligand. We have found that the clustering of recognition sites together
in one molecule is detrimental to their individual abilities to bind monovalent ligands. In the case
of the polyvalent interaction between the hexakisBMP25C8 cluster and divalent dialkylammonium
ions, an association constant, KPOLY, was calculated from the value of KAVE determined for the
complexation of the individual component recognition sites. This polyvalent interaction is
significantly stronger than that associated with the averaged monovalent interactions.
In tr od u ction
hence stronger binding with a lectin. In the case of the
chelate effect, the multiple binding moieties in a carbo-
hydrate ligand can bind cooperatively to the lectin,
leading to a strong interaction. Chemists4 are starting
to employ this concept of polyvalency in the synthesis of
artificial receptors and ligands. There have been a
number of striking examples5 of multi- and polyvalent
ligandssdesigned either rationally or randomlysthat
In natural systems it has been observed1 that the
binding interactions between multivalent clusters of
receptors and substrates can be extremely strong, even
though individually these ligands bind only weakly to
each other. An example of this concept of employing
polyvalent interactions between species is the phenom-
enon known2 as the “glycoside cluster effect”, in which
the relatively low binding affinity exhibited between
single carbohydrate ligands and lectins is overcome by
the clustering together of these ligands. These clusters
of carbohydrate ligands show enhanced binding affinities
to lectins and are involved in many important biological
processes by which cells, bacteria, and viruses recognize
one another. The two most important mechanisms3 that
are responsible for this enhancement in the binding
affinity between carbohydrate clusters and lectins are (i)
statistical effects and (ii) the chelate effect. Statistical
effects can express themselves in a cluster presenting a
high local concentration of, for example, carbohydrate
ligands, which may lead to lower dissociation rates and
(3) Statistical and chelate effects have been proposed to explain the
“glycoside cluster effect” that has been observed in nature. See: (a)
Pohl, N. L.; Kiessling, L. L. Synthesis 1999, 1515-1519. (b) Mann, D.
A.; Kiessling, L. L. In Glycochemistry: Principles, Synthesis and
Applications; Wang, P. G., Bertozzi, C. R., Eds.; Marcel Dekker: New
York, 2001; pp 221-277. (c) Cairo, C. W.; Gestwicki, J . E.; Kanai, M.;
Kiessling, L. L. J . Am. Chem. Soc. 2002, 124, 1615-1619. (d) Gestwicki,
J . E.; Cairo, C. W.; Strong, L. E.; Oetjen, K. A.; Kiessling, L. L. J . Am.
Chem. Soc. 2002, 124, 14922-14933. (e) Woller, E. K.; Walter, E. D.;
Morgan, J . R.; Singel, D. J .; Cloninger, M. J . J . Am. Chem. Soc. 2003,
125, 8820-8826.
(4) (a) Roy, R. Polym. News 1996, 21, 226-232. (b) Hamuro, Y.;
Clama, M. C.; Park, H. S.; Hamilton, A. D. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2680-2683. (c) Kanai, M.; Mortell, K. H.; Kiessling,
L. L. J . Am. Chem. Soc. 1997, 119, 9931-9932. (d) Page´, D.; Roy, R.
Bioconj. Chem. 1997, 8, 714-723. (e) J ayaraman, N.; Nepogodiev, S.
A.; Stoddart, J . F. Chem. Eur. J . 1997, 3, 1193-1199. (f) Dubber, M.;
Lindhorst, T. K. Carbohydr. Res. 1998, 310, 35-41. (g) Fujimoto, K.;
Miyata, T.; Aoyama, Y. J . Am. Chem. Soc. 2000, 122, 3558-3559. (h)
Roy, R.; Herna´ndez-Mateo, F.; Santoyo-Gonza´lez, F. J . Org. Chem.
2000, 65, 8743-8746. (i) Rao, J .; Lahiri, J .; Weis, R. M.; Whitesides,
G. M. J . Am. Chem. Soc. 2000, 122, 2698-2710. (j) Fazal, M. A.; Roy,
B. C.; Sun, S.; Mallik, S.; Rodgers, K. R. J . Am. Chem. Soc. 2001, 123,
6283-6290. (k) Rockendorf, N.; Lindhorst, T. K. Top. Curr. Chem.
2001, 217, 201-238. (l) Roy, R. J . Carbohydr. Chem. 2002, 21, 769-
798. (m) Arranz-Plaza, E.; Tracy, A. S.; Siriwardena, A.; Pierce, J . M.;
Boons, G.-J . J . Am. Chem. Soc. 2002, 124, 13035-13046. (n) Turnbull,
W. B.; Kalovidouris, S. A.; Stoddart, J . F. Chem. Eur. J . 2002, 8, 2988-
3000. (o) Dubber, M.; Fre´chet, J . M. J . Bioconj. Chem. 2003, 14, 239-
246.
* Address correspondence to this author. Phone: (310) 206-7078.
Fax: (310) 206-1843.
(1) (a) Varki, A. Glycobiology 1993, 3, 97-130. (b) Mammen, M.;
Choi, S.-K.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37, 2754-
2794. (c) Essentials of Glycobiology; Varki, A., Cummings, R., Esko,
J ., Freeze, H., Hart, G., Marth, J ., Eds.; Cold Spring Harbor Laboratory
Press: New York, 1999.
(2) (a) Lee, R. T.; Lee, Y. C. In Neoglycoconjugates: Preparation and
Applications; Lee, R. T., Lee, Y. C., Eds.; Academic Press: San Diego,
CA, 1994; pp 23-50. (b) Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995,
28, 321-327. (c) Drickamer, K. Nat. Struct. Biol. 1995, 2, 437-439.
(d) Lundquist, J . J .; Toone, E. J . Chem. Rev. 2002, 102, 555-578.
10.1021/jo030283o CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/20/2004
4390
J . Org. Chem. 2004, 69, 4390-4402

Polyvalent Interactions
F IGURE 1. Dibenzylammonium ions (DBA+) form complexes
possessing pseudorotaxane superstructures with the crown
ethers dibenzo[24]crown-8 (DB24C8) and bis-m-phenylene[25]-
crown-8 (BMP25C8).
F IGURE 2. The binding in CDCl3/CD3CN (1:1) of (a) the
trisammonium salt 1‚3PF6 and tris(crown ether) 2 is consider-
ably stronger than the corresponding monovalent interaction
between (b) DB24C8 and DBA+.
show promising inhibitory behavior toward toxins, vi-
ruses, and bacteria.
than 106 M-1 in CDCl3/CD3CN (1:1), a Ka value that
corresponds to a free energy of binding (-∆G°) of more
than 7.9 kcal mol-1. For the corresponding monovalent
interaction between DB24C8 and the DBA+ ion in the
same solvent mixture, an association constant of 1700
M-1 has been determined.6b This value of Ka corresponds
to a free energy of binding of 4.2 kcal mol-1. Thus, the
polyvalent interaction between the two trivalent species
is in excess of 3.7 kcal mol-1 more favorable than the
The complexation (Figure 1) between dibenzylammo-
nium (DBA+) cations and dibenzo[24]crown-8 (DB24C8)
or benzo-m-phenylene[25]crown-8 (BMP25C8) derivatives
is an example of an extensively studied6-9 monovalent
interaction. Although there are examples10,11 where this
binding interaction is enhanced by the clustering together
of multiple copies of this recognition motif, they are few
in number. The clustering of binding sites (Figure 2) in
the tris(dialkylammonium) salt 1‚3PF6 and the tris(crown
ether) 2 results10 in an extremely stable complex being
formed when the two are mixed. The value12 for the
association constant (Ka) between 1‚3PF6 and 2 is greater
(8) For reviews, see: (a) Glink, P. T.; Schiavo, C.; Stoddart, J . F.;
Williams, D. J . Chem. Commun. 1996, 1483-1490. (b) Fyfe, M. C. T.;
Stoddart, J . F. Acc. Chem. Res. 1997, 30, 393-401. (c) Fyfe, M. C. T.;
Stoddart, J . F. Adv. Supramol. Chem. 1999, 5, 1-53. (d) Hubin, T. J .;
Kolchinski, A. G.; Vance, A. L.; Busch, D. H. Adv. Supramol. Chem.
1999, 5, 237-357. (e) Fyfe, M. C. T.; Stoddart, J . F.; Williams. D. J .
Struct. Chem. 1999, 10, 243-249. (f) Fyfe, M. C. T.; Stoddart, J . F.
Coord. Chem. Rev. 1999, 183, 139-155. (g) Hubin, T. J .; Busch, D. H.
Coord. Chem. Rev. 2000, 200, 5-52. (h) Cantrill, S. J .; Pease, A. R.;
Stoddart, J . F. J . Chem. Soc., Dalton Trans. 2000, 3715-3734. (i)
Takata, T.; Kihara, N. Rev. Heteroatom Chem. 2000, 22, 197-218. (j)
Kihara, N.; Takata, T. J . Synth. Org. Chem. J pn. 2001, 59, 206-218.
(9) For selected contributions from other research laboratories,
see: (a) J ohnson, B. F. G.; J udkins, C. M. G.; Matters, J . M.; Shephard,
D. S.; Parsons, S. Chem. Commun. 2000, 1549-1550. (b) Duggan, S.
A.; Fallon, G.; Langford, S. J .; Lau, V. L.; Satchell, J . F.; Paddon-Row,
M. N. J . Org. Chem. 2001, 66, 4419-4426. (c) Asakawa, M.; Ikeda, T.;
Yui, N.; Shimizu, T. Chem. Lett. 2002, 174-175. (d) Tokunaga, Y.; Seo,
T. Chem. Commun. 2002, 970-971. (e) Tokunaga, Y.; Kakuchi, S.;
Akasaka, K.; Nishikawa, N.; Shimomura, Y.; Isa, K.; Seo, T. Chem.
Lett. 2002, 810-811. (f) Clifford, T.; Abushamleh, A.; Busch, D. H. Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 4830-4836. (g) Gibson, H. W.;
Yamaguchi, N.; Hamilton, L.; J ones, J . W. J . Am. Chem. Soc. 2002,
124, 4653-4665. (h) J ones, J . W.; Gibson, H. W. J . Am. Chem. Soc.
2003, 125, 7001-7004. (i) Gibson, H. W.; Yamaguchi, N.; J ones, J . W.
J . Am. Chem. Soc. 2003, 125, 3522-3533. (j) Furusho, Y.; Oku, T.;
Hasegawa, T.; Tsuboi, A.; Kihara, N.; Takata, T. Chem. Eur. J . 2003,
9, 2895-2903. (k) Oku, T.; Furusho, Y.; Takata, T. J . Polym. Sci., Part
A 2003, 41, 119-123. (l) Iwamoto, H.; Itoh, K.; Nagamiya, H.;
Fukazawa, Y. Tetrahedron Lett. 2003, 44, 5773-5776. (m) Dvornikovs,
V.; House, B. E.; Kaetzel, M.; Dedman, J . R.; Smithrud, D. B. J . Am.
Chem. Soc. 2003, 125, 8290-8301. (n) Smukste, I.; Smithrud, D. B. J .
Org. Chem. 2003, 68, 2547-2558. (o) Smukste, I.; House, B. E.;
Smithrud, D. B. J . Org. Chem. 2003, 68, 2559-2571. (p) Kawano, S.
I.; Fujita, N.; Shinkai, S. Chem. Commun. 2003, 1352-1353. (q)
Frankfort, L.; Sohlberg, K. J . Mol. Struct. THEOCHEM 2003, 621,
253-260.
(5) (a) Gordon, E. J .; Sanders, W. J .; Kiessling, L. L. Nature 1998,
392, 30-31. (b) Kitov, P. I.; Sadawska, J . M.; Mulvey, G.; Armstrong,
G. D.; Ling, H.; Pannu, N. S.; Read, R. J .; Bundle, D. R. Nature 2000,
403, 669-672. (c) Mourez, M.; Kane, R. S.; Mogridge, J .; Metallo, S.;
Deschatelets, P.; Sellman, B. R.; Whitesides, G. M.; Collier, R. J . Nat.
Biotechnol. 2001, 19, 958-961.
(6) For publications that address the complexation of DBA+ ions by
DB24C8, see: (a) Ashton, P. R.; Campbell, P. J .; Chrystal, E. J . T.;
Glink, P. T.; Menzer, S.; Philp, D.; Spencer, N.; Stoddart, J . F.; Tasker,
P. A.; Williams, D. J . Angew. Chem., Int. Ed. Engl. 1995, 34, 1865-
1869. (b) Ashton, P. R.; Chrystal, E. J . T.; Glink, P. T.; Menzer, S.;
Schiavo, C.; Spencer, N.; Stoddart, J . F.; Tasker, P. A.; White, A. J . P.;
Williams, D. J . Chem. Eur. J . 1996, 2, 709-721. For some of our recent
publications in which the complexation of dibenzylammonium ions by
DB24C8 has been exploited for the construction of mechanically
interlocked molecules, see: (c) Elizarov, A. M.; Chiu, S.-H.; Stoddart,
J . F. J . Org. Chem. 2002, 67, 9175-9181. (d) Chiu, S.-H.; Elizarov, A.
M.; Glink, P. T.; Stoddart, J . F. Org. Lett. 2002, 4, 3561-3564. (e)
Elizarov, A. M.; Chang, T.; Chiu, S.-H.; Stoddart, J . F. Org. Lett. 2002,
4, 3565-3568. (f) Elizarov, A. M.; Chiu, S.-H.; Glink, P. T.; Stoddart,
J . F. Org. Lett. 2002, 4, 679-682. (g) Chiu, S.-H.; Rowan, S. J .; Cantrill,
S. J .; Ridvan, L.; Ashton, P. R.; Garrell, R. L.; Stoddart, J . F.
Tetrahedron 2002, 58, 807-814.
(7) For publications that address the complexation of DBA+ and
related ions by BMP25C8, see: (a) Ashton, P. R.; Bartsch, R. A.;
Cantrill, S. J .; Hanes, R. E., J r.; Hickingbottom, S. K.; Lowe, J . N.;
Preece, J . A.; Stoddart, J . F.; Talanov, V. S.; Wang, Z.-H. Tetrahedron
Lett. 1999, 40, 3661-3664. (b) Cantrill, S. J .; Fulton, D. A.; Heiss, A.
M.; Pease, A. R.; Stoddart, J . F.; White, A. J . P.; Williams, D. J . Chem.
Eur. J . 2000, 6, 2274-2287. (c) Rowan, S. J .; Cantrill, S. J .; Stoddart,
J . F.; White, A. J . P.; Williams, D. J . Org. Lett. 2000, 2, 759-762. (d)
Chiu, S.-H.; Rowan, S. J .; Cantrill, S. J .; Glink, P. T.; Garrell, R. L.;
Stoddart, J . F. Org. Lett. 2000, 2, 3631-3634. (e) Cantrill, S. J .; Youn,
G. J .; Stoddart, J . F.; Williams, D. J . J . Org. Chem. 2001, 66, 6857-
6872. (f) Chiu, S.-H.; Rowan, S. J .; Cantrill, S. J .; Stoddart, J . F.; White,
A. J . P.; Williams, D. J . Chem. Commun. 2002, 2948-2949. (g) Rowan,
S. J .; Stoddart, J . F. Polym. Adv. Technol. 2002, 13, 777-787. (h) Chiu,
S.-H.; Rowan, S. J .; Cantrill, S. J .; Stoddart, J . F.; White, A. J . P.;
Williams, D. J . Chem. Eur. J . 2002, 8, 5170-5183.
(10) (a) Fyfe, M. C. T.; Lowe, J . N.; Stoddart, J . F.; Williams, D. J .
Org. Lett. 2000, 2, 1221-1224. (b) Balzani, V.; Clemente-Leo´n, M.;
Credi, A.; Lowe, J . N.; Badjic, J . D.; Stoddart, J . F. Chem. Eur. J . 2003,
9, 5348-5360.
(11) Fulton, D. A.; Cantrill, S. J .; Stoddart, J . F. J . Org. Chem. 2002,
67, 7968-7981.
(12) Ion pairing was not taken into account for the calculation of
all the association constants in this paper.
J . Org. Chem, Vol. 69, No. 13, 2004 4391

Lowe et al.
corresponding monovalent interaction. It has also been
observed11 that cyclodextrin clusters bearing seven
DB24C8 or seven BMP25C8 rings bind monovalent
dialkylammonium salts more weakly than does DB24C8
or BMP25C8, respectively. It was found, however, that
the strength of binding of these clusters to a divalent bis-
(dialkylammonium) salt is much stronger than the
interactions between the corresponding monovalent spe-
cies. That is to say, these crown ether clusters show
reduced binding toward monovalent ionic species, but
enhanced binding to multivalent ionic compounds, rela-
tive to the strength of the parent monovalent systems.
To understand better these phenomena, we have inves-
tigated the binding of other multivalent clusters bearing
crown ether and dialkylammonium ion recognition sites
with monovalent and divalent ligands.
systems. A hexakis(phenylthio)benzene core was em-
ployed in the clusters since (i) it can be readily synthe-
sized and (ii) the recognition sites attached to this core
are spatially arranged in a precise manner with a large
degree of directionality and symmetry. Hexakis(phenyl-
thio)benzene derivatives usually possess, in both the solid
state15 and in solution,16 conformations in which neigh-
boring substituents of the benzenoid core point in op-
posite directionssthat is to say, the thiophenyl substit-
uents on the 1,3,5-positions of the central aromatic ring
point in one direction with respect to the plane of the
central benzenoid core and those on the 2,4,6-positions
point in the opposite direction.
Resu lts a n d Discu ssion
Syn t h esis of t h e H exa k is(p h en ylt h io)b en zen e
Clu ster s. Hexakis(phenylthio)benzene derivatives can be
synthesized17 readily by substituting the fluorine atoms
on hexafluorobenzene with aromatic thiolate ions in
aprotic, polar solvents, such as 1,3-dimethylimidazolidin-
2-one (DMI). This reaction proceeds well at room tem-
perature, generally requiring only a few hours to reach
completion.
Syn th esis of th e Hexa k is(cr ow n eth er ) Clu ster .
The hexakisBMP25C8 cluster 3 was synthesized as
outlined in Scheme 1. We employed a divergent strategy,
first preparing the hexathiobenzene core and then at-
taching the crown ether appendages to the periphery of
this core. Hexakis(p-methoxyphenylthio)benzene16 (5)
was obtained by reacting the thiolate anion of p-meth-
oxythiophenol and hexafluorobenzene in DMI. This com-
pound was demethylated by using boron tribromide to
afford hexakis(p-hydroxyphenylthio)benzene (6). The
BMP25C8 derivative 8 was prepared by macrocyclization,
under high dilution conditions, between the commercially
available 3,5-dihydroxybenzyl alcohol and the bis-
tosylate11 7. This crown ether was reacted subsequently
with methanesulfonyl chloride to afford the mesylate 9.
Hexakis(p-hydroxyphenylthio)benzene (6) was then re-
acted with the BMP25C8 derivative 9 in MeCN, in the
presence of K2CO3 and [18]crown-6, to produce a good
yield of the hexakisBMP25C8 cluster 3.
Syn th esis of th e Dia lk yla m m on iu m Clu ster . The
hexakis(dialkylammonium) cluster 4‚6PF6 was synthe-
sized (Scheme 2) starting from the commercially available
4-methylthiobenzaldehyde. The starting material was
converted to the dialdehyde disulfide 11, employing a
procedure similar to that reported in the literature.18
First, the thioether was oxidized to the sulfoxide 10 with
m-chloroperoxybenzoic acid (mCPBA) at 0 °C to avoid
over-oxidation. After purification by column chromatog-
raphy, the sulfoxide 10 was subjected to a Pummerer
rearrangement with trifluoroacetic anhydride, followed
by hydrolysis with a mixture of methanol and triethy-
As the recognition system involves the interaction of
multiply charged species, our observations may also be
of relevance to the self-assembly of other systems, e.g.,
the self-assembly of block copolymer micelles via the
aggregation of multiply charged amphiphilic block co-
polymers.13 Recent developments in this vast field of
research include the development of polyelectrolyte co-
polymers, which have potential for use as drug-delivering
microgels.14
F IGURE 3. The clusters 3 and 4‚6PF6, which present six
crown ether rings and six dialkylammonium ion centers,
respectively.
This paper describes the synthesis of two multivalent
such clusters (Figure 3), one bearing six crown ether rings
(3), and the other possessing six secondary dialkyl-
ammonium ion centers (4‚6PF6). We have probed the
binding of these clusters to complementary monovalent
and polyvalent ligands by NMR spectroscopy to evaluate
the contributions of statistical and chelate effects to
polyvalent interactions in such unnatural recognition
(13) (a) Harada, A.; Kataoka, K. Science 1999, 283, 65-67. (b)
Harada, A.; Kataoka, K. J . Am. Chem. Soc. 1999, 121, 9241-9242. (c)
Simmons, C.; Webber, S. E.; Zhulina, E. B. Macromolecules 2001, 34,
5053-5066. (d) Matj´ıcek, P.; Uhl´ık, F.; Limpouchova´, Z.; Procha´zka,
K.; Tuzar, Z.; Webber, S. E. Macromolecules 2002, 35, 9487-9496. (e)
Pergushov, D. V.; Remizova, E. V.; Feldthusen, J .; Zezin, A. B.; Mu¨ller,
A. H. E.; Kabanov, V. A. J . Phys. Chem. B 2003, 107, 8093-8096. (f)
Matj´ıcek, P.; Humpolickova´, J .; Procha´zka, K.; Tuzar, Z.; Sp´ırkovea,
M.; Hof, M.; Webber, S. E. J . Phys. Chem. B 2003, 107, 8232-88240.
(14) For some recent examples see: (a) Bromberg, L.; Temchenko,
M.; Hatton, T. A. Langmuir 2003, 19, 8675-8684. (b) Bromberg, L.;
Temchenko, M.; Hatton, T. A. Langmuir 2002, 18, 4944-4952.
(15) (a) MacNicol, D. D.; Hardy, A. D. U.; Wilson, D. R. Nature 1977,
266, 611-612. (b) MacNicol, D. D.; Downing, G. A. In Comprehensive
Supramolecular Chemistry; Atwood, J . L., Davies, J . E. D., MacNicol,
D. D., Vo¨gtle, F., Eds.; Pergamon: Oxford, UK, 1996; Vol. 6, Chapter
14.
(16) Stack, T. D. P.; Holm, R. H. J . Am. Chem. Soc. 1988, 110, 2484-
2494.
(17) MacNicol, D. D.; Mallinson, P. R.; Murphy, A.; Sym, G. J .
Tetrahedron Lett. 1982, 23, 4131-4134.
(18) Young, R. N.; Gauthier, J . Y.; Coombs, W. Tetrahedron Lett.
1984, 25, 1753-1755.
4392 J . Org. Chem., Vol. 69, No. 13, 2004

Polyvalent Interactions
SCHEME 1. Th e Syn th esis of th e Hexa k isBMP 25C8 Clu ster 3
SCHEME 2. Th e Syn th esis of th e Hexa k is(d ia lk yla m m on iu m ion ) Clu ster 4‚6P F 6
lamine to afford the corresponding thiol-bearing benz-
aldehyde. For ease of characterization and use, this thiol
was oxidized, using iodine, to form the disulfide 11. The
dialdehyde disulfide 11 was condensed with benzylamine
J . Org. Chem, Vol. 69, No. 13, 2004 4393

Lowe et al.
resonance (-112.7 ppm) corresponds to the fluorine
atoms of the [BMP25C8‚DBA-F4][PF6] complex and the
+
more upfield one (-113.1 ppm) to the free DBA-F4 ion.
Hence, by comparing the ratio of the integrals of these
two resonances, the relative concentrations of the com-
plexed and free species can be calculated and, thus, a
value for the binding constant (Ka) between the two
components can be determined.
19F NMR spectra of solutions of the hexakisBMP25C8
cluster 3 and DBA-F4‚PF6 reveal (Figure 5) a signal
(-113.1 ppm) corresponding to the fluorine atoms in free
+
DBA-F4 ions and a set of overlapping signals (-112.4
to -112.6 ppm) corresponding to fluorine atoms of bound
+
F IGURE 4. Partial 19F NMR spectrum [376 MHz, CD2Cl2/
CD3CN (1:1), 300 K] of an equimolar solution of BMP25C8 and
DBA-F4‚PF6.
DBA-F4 ions. These overlapping signals presumably
correspond to fluorine atoms of the inclusion complexes,
ranging from [2]- through to [7]pseudorotaxanes, formed
between the hexakis(crown ether) and the cation. Al-
though assigning these signals to the different [n]-
pseudorotaxane species is not an easy task, the total
to yield the diimine disulfide 12. In one step, the imino
and disulfide groups were reduced by using sodium
borohydride to afford the thiol 13. Sodium hydride was
employed to form the thiolate anion from the thiol 13.
This thiolate anion was then reacted, in DMI, with
hexafluorobenzene at room temperature to give the
hexaamine 14. The structure was confirmed by its 13C
NMR spectrum, which exhibited only one signal corre-
sponding to the carbon atoms of the thiobenzene core at
a characteristic16 chemical shift of 148 ppm. The hexa-
amine 14 was protonated with concentrated hydrochloric
acid to give the hexakis-hydrochloride salt. Counterion
exchange with ammonium hexafluorophosphate afforded
the desired hexakis(dialkylammonium) hexakis(hexafluo-
rophosphate) salt 4‚6PF6 as a yellow solid.
Stu d yin g Sta tistica l Effects in th e Bin d in g of
Clu ster Com p ou n d s w ith Mon ova len t Liga n d s. To
investigate the importance of statistical effects in the
interactions of multivalent crown ethers and multivalent
dialkylammonium salts with their respective ligands, we
studied their strength of binding with complementary
monovalent species. If statistical effects play an impor-
tant role in the polyvalent binding of these systems, an
increase in the strength of complexation of the individual
binding moieties with monovalent species would be
expected.
+
concentration of complexed and uncomplexed DBA-F4
ions can be determined readily by integration of the
spectrum. An average association constant (KAVE) for the
complexation can be calculated from these concentra-
tions. This average association constant11 is defined as a
measure of the strength of the binding interactions
between a single crown ether moiety and a single
secondary dialkylammonium moiety. This quantity can
be defined as
KAVE
)
[complexed NH2+ centers]
[uncomplexed NH2+ centers][uncomplexed crown ether units]
(1)
+
We define an “NH2 center” as a single secondary
dialkylammonium ion center within a compound contain-
ing one or more of these centers, and a “crown ether unit”
as a single crown ether moiety within a compound
containing one or more of these moieties.
Bin d in g betw een th e Hexa k is(cr ow n eth er ) Clu s-
ter a n d Mon ova len t Liga n d s. To investigate how the
clustering of the crown ether moieties in the hexakis-
BMP25C8 cluster 3 might affect the strength of the
binding to secondary dialkylammonium ions of each
moiety, we studied the complexation between this cluster
and bis(3,5-difluorobenzyl)ammonium hexafluorophos-
phate salt DBA-F4‚PF6. The dialkylammonium salt DBA-
F4‚PF6 is known7e to complex with BMP25C8 with an
association constant of 25 M-1 in CD3CN. In this binding
process, the rate of exchange19 between the free and
complexed species is slow on the 19F NMR time scale at
376 MHz. As a result, the 19F NMR spectrum of a solution
of BMP25C8 and DBA-F4‚PF6 reveals (Figure 4) two
resonances that correspond to the fluoro substituents of
the dialkylammonium ion. The relatively downfield
(19) In the interaction with the DBA+ ion, the rate of exchange
between complexed and uncomplexed BMP25C8 species is neither slow
nor fast on the NMR spectroscopic time scale at 300-600 MHz. 1H
and 19F NMR spectra containing broad resonances result from this
phenomenon and, hence, the value of Ka cannot be determined easily
by the single-point method.
F IGURE 5. Partial 19F NMR spectrum [376 MHz, CD2Cl2/
CD3CN (1:1), 300 K] of a solution of the hexakisBMP25C8
cluster 3 and DBA-F4‚PF6.
4394 J . Org. Chem., Vol. 69, No. 13, 2004

Polyvalent Interactions
Historically, the determination of an association con-
stant for the binding of a dibenzylammonium cation by
a DB24C8 receptor, undergoing slow exchange on the
NMR time scale, has been performed by the single-point
method. Indeed, since the complexation of dialkylammo-
nium cations by macrocyclic polyethers was first re-
ported,6a,20 many papers have utilized the single-point
method to determine association constants. Gibson et al.
has published9h,21 recently a more sophisticated analysis
for the complexation of ionic guests by neutral receptors.
This analysis takes into account the ion-pairing equilib-
rium between the guest cation and its associated coun-
terion. The effects of this equilibrium can manifest
themselves in concentration-dependent fluctuations in
the apparent Ka values, and their more sophisticated
equilibrium treatment takes this observation into ac-
count. Our attempts to reanalyze our data from the
experiments presented here, utilizing the method de-
scribed by Gibson et al., resulted in very limited successs
only the simplest experiments involving the complexation
of a monocationic guest by a monovalent crown ether
afforded sensible values for Ka. When the treatment was
applied to experimental data from our own systems
involving polyvalent recognition, spurious results were
obtained. Consequently, we have analyzed our own
experimental data using the method described6b,7a,b,11,22
previously by our own research group, one that is based
firmly upon the tried-and-tested single-point method for
the determination of Ka.
a 1:1 model for this system, was constructed (Figure 6)
by plotting ([complex]/[crown ether moieties]TOT) against
[free NH2+ centers]. This binding isotherm could then be
compared with others to obtain a qualitative assessment
of binding strength of the individual crown ethers to the
dialkylammonium ion centers. The mean of the values
of KAVE for the individual data points was calculated to
be 116 ( 35 M-1, a value that corresponds to a mean
free energy change, ∆G°AVE, of -2.8 ( 0.2 kcal mol-1
.
A similar set of 19F NMR titration experiments were
undertaken in CD3CN/CD2Cl2 (1:1) between the hexakis-
BMP25C8 cluster 3 and DBA-F4‚PF6. In these experi-
ments, the concentration of 3 was held constant at 0.5
mM, i.e., the concentration of crown ether sites cor-
responded to 3.0 mM, just as was the case in the titration
experiments between BMP25C8 and DBA-F4‚PF6. The
concentration of the salt was varied from 1.01 to 12.12
mM. From the relative integrals of all the resonances
+
corresponding to complexed DBA-F4 ions and the free
+
DBA-F4 ions, we determined the concentrations of
+
complexed NH2 centers. From these experiments, a
binding isotherm was plotted (Figure 6) and a mean value
of KAVE was calculated to be 48 ( 14 M-1, which
corresponds to a mean ∆G°AVE of -2.3 ( 0.2 kcal mol-1
.
It should also be noted from the separate data points that
the individual values of KAVE, between the hexakis-
+
BMP25C8 cluster 3 and DBA-F4 ions, decrease as the
concentration of the salt is increased.
From the comparison of the binding isotherms and the
mean values of the KAVE, it is apparent that BMP25C8
For a system in which there are two monovalent
ligands present that bind together in a 1:1 stoichiometry
(e.g., BMP25C8 and DBA-F4‚PF6), KAVE is simply the
standard association constant (i.e., Ka). In a system where
there are multivalent species present, however, KAVE is
the association constant for each crown ether/dialkylam-
monium ion binding interaction viewed independently
from each other. Thus, by comparing the values of KAVE
for the binding of DBA-F4‚PF6 with (i) BMP25C8 and (ii)
the hexakisBMP25C8 cluster 3, the effect of clustering
the crown ether moieties on their binding of DBA-F4+ ions
can be determined. First, to determine an accurate value
of KAVE for the binding between BMP25C8 and DBA-F4‚
PF6, 19F NMR spectra were obtained for a series of
solutions of the two species in CD3CN/CD2Cl2 (1:1). In
each solution, the concentration of BMP25C8 was held
constant at 3.0 mM, while the concentration of DBA-F4‚
PF6 was varied from 1.0 to 15.1 mM. From each of these
spectra, a value for KAVE was determined by using the
single-point method.6,11 A binding isotherm,23 based on
(20) Kolchinski, A. G.; Busch, D. H.; Alcock, N. W. J . Chem. Soc.,
Chem. Commun. 1995, 1289-1291.
(21) Huang, F.; J ones, J . W.; Slebodnick, C.; Gibson, H. W. J . Am.
Chem. Soc. 2003, 125, 14458-14464.
F IGURE 6. “Best-fit” binding isotherms for the complexation
between BMP25C8 and DBA-F4‚PF6 (9) and the complexation
between the hexakisBMP25C8 cluster 3 and DBA-F4‚PF6 (b).
(22) (a) Ashton, P. R.; Glink, P. T.; Stoddart, J . F.; Tasker, P. A.;
White, A. J . P.; Williams, D. J . Chem. Eur. J . 1996, 2, 729-736. (b)
Ashton, P. R.; Fyfe, M. C. T.; Hickingbottom, S. K.; Stoddart, J . F.;
White, A. J . P.; Williams, D. J . J . Chem. Soc., Perkin Trans. 2 1998,
2117-2128. (c) Cantrill, S. J .; Fyfe, M. C. T.; Heiss, A. M.; Stoddart,
J . F.; White, A. J . P.; Williams, D. J . Org. Lett. 2000, 2, 61-64. (d)
Chang, T.; Heiss, A. M.; Cantrill, S. J .; Fyfe, M. C. T.; Pease, A. R.;
Rowan, S. J .; Stoddart, J . F.; White, A. J . P.; Williams, D. J . Org. Lett.
2000, 2, 2947-2950. (e) Ashton, P. R.; Becher, J .; Fyfe, M. C. T.;
Nielsen, M. B.; Stoddart, J . F.; White, A. J . P.; Williams, D. J .
Tetrahedron 2001, 57, 947-956. (f) Chiu, S.-H.; Pease, A. R.; Stoddart,
J . F.; White, A. J . P.; Williams, D. J . Angew. Chem., Int. Ed. 2002, 41,
270-274. (g) Chiu, S.-H.; Liao, K. S.; Su, J . K. Tetrahedron Lett. 2004,
45, 213-216.
Next to each data point is displayed the value of KAVE (M-1
;
errors (30%) determined by using the single-point method.
The values on the y-axis represent the measured concentration
of complexed DBA-F4+ ions divided by the total concentration
of crown ether moieties. The values on the x-axis represent
the concentration of free DBA-F4+ ions in solution. Error bars
on each data point are set at 10% and 14% for the x- and
y-axes, respectively, and reflect experimental errors in the
weight and volume of the samples and in the integration of
peaks in spectra.
(23) Connors, K. C. Binding Constants; Wiley: New York, 1987.
J . Org. Chem, Vol. 69, No. 13, 2004 4395

Lowe et al.
DBA+ ions, at the same concentrations of crown ether
rings, by a factor of between 2 and 30. The values of KAVE
for those polymer/DBA+ interactions also decrease with
increasing dialkylammonium ion concentration.
In the cases of BMP25C8 and hexaBMP25C8 3 binding
with DBA-F4+, the general trend is a decrease in KAVE
as the concentration of the ammonium ions increases.
This general decrease is probably a consequence of ion
pairing between the ammonium ion and its hexafluoro-
phosphate anion, a situation that consequently diverges
the system away from ideal behavior. Research by Gibson
and co-workers9h has explored this phenomenon with a
variety of counterions, with the hexaflurophosphate anion
exhibiting the weakest ion pairing with dibenzylammo-
nium cations. The use of hexafluorophosphate as a
counterion in our own experiments helps to minimize ion-
pairing effects, as does the relatively narrow concentra-
tion range over which these experiments are performed.
Indeed, the relatively narrow range of experimentally
determined KAVE values (129 to 106 M-1 and from 77 to
33 M-1: Figure 6) supports the idea that although our
system exhibits ion-pairing effects, they do not affect
significantly the validity of the conclusions.
Bin d in g betw een th e Hexa k is(d ia lk yla m m on iu m )
Clu ster a n d Diben zo[24]cr ow n -8. The binding of the
hexakis(dialkylammonium) cluster 4‚6PF6 with DB24C8
was investigated (Figures 8 and 9) to discover how the
clustering of six NH2+ centers in a single species affects
each unit’s ability to bind to crown ethers. If a value for
KAVE is determined for the binding between these two
species, then this value can be compared to that found6,22e
for the complexation between DBA+ and DB24C8. Thus,
a similar comparison can be made between these two
F IGURE 7. A comparison of the free energy of binding
between (a) BMP25C8 and DBA-F4‚PF6 and (b) the hexakis-
BMP25C8 cluster 3 and DBA-F4‚PF6.
+
is a better host for DBA-F4 ions than is the hexa-
BMP25C8 cluster 3. Indeed, the crown ether rings of the
hexakisBMP25C8 cluster seem to bind (Figure 7) the
dialkylammonium thread DBA-F4+ about 0.5 kcal mol-1
less effectively.24 Hence, the clustering of the crown ether
rings has a detrimental effect on their ability to bind
monovalent secondary dialkylammonium ions. This effect
is probably a consequence of unfavorable electronic and
steric factors. A hexakis(crown ether) cluster that binds
a single dialkylammonium ion will possess a single
positive charge. Any additional binding between the
remaining free crown ether macrocycles in the 1:1
complex and other dialkylammonium ions will have to
overcome the electrostatic repulsion between this posi-
tively charged complex and the approaching unbound
dialkylammonium ions. Steric hindrance from each bound
cation and its associated anion is also expected to inhibit
further binding between the hexakis(crown ether) cluster
and dialkylammonium ions. Both effects are detrimental
to binding and they will increase in significance as the
cluster binds further dialkylammonium ions. This in-
creased repulsion of DBA-F4 cations may be the reason
for the decrease in the observed KAVE for hexaBMP25C8
cluster 3 as the concentration of the dialkylammonium
ion increases. At low concentrations of cations, lower
order complexes (i.e., [2]- and [3]pseudorotaxanes) are all
that can be expected to form and, since these complexes
are not expected to suffer the steric and electronic effects
to the degree that the higher order complexes do, the
value of KAVE is relatively high. At higher concentrations
of dialkylammonium ions, higher order [n]pseudorotax-
anes are inhibited from forming, and so these complexes
having lower binding affinities for dialkylammonium ions
will cause the value of KAVE to decrease.
Gibson and co-workers25 have also observed a similar
phenomenon between DBA+ ions and polymethacrylates
bearing DB24C8 appendages. They found that the values
of KAVE for the binding of these polymers with DBA+ ions
were lower than those for the binding of DB24C8 with
F IGURE 8. 1H NMR spectra [500 MHz, CD2Cl2/CD3CN (1:
1), 300 K] of a mixture of 4‚6PF6 and DB24C8. The concentra-
tion of 4‚6PF6 is 4.88 mM, while the concentration of DB24C8
is (a) 0.00, (b) 2.32, (c) 5.04, (d) 14.8, (e) 29.2, (f) 44.9 , and (g)
59.9 mM. For peak labeling, see Figures 2 and 3.
(24) We have observed a similar effect in the binding of (i) a
â-cyclodextrin-based heptakisBMP25C8 derivative and (ii) BMP25C8
with DBA-F4‚PF6 in CD3CN. The cluster compound binds this ion less
effectively by ca. 0.4 kcal mol-1. See ref 11.
(25) Yamaguchi, N.; Gibson, H. W. Macromol. Chem. Phys. 2000,
201, 815-824.
4396 J . Org. Chem., Vol. 69, No. 13, 2004

Polyvalent Interactions
systems as was made above for the corresponding binding
of the hexakisBMP25C8 cluster 3 and BMP25C8 with
DBA-F4+ ions. To establish a value of KAVE and a binding
isotherm between DBA‚PF6 and DB24C8, 1H NMR
spectra were obtained for a series of CD3CN/CD2Cl2 (1:
1) solutions of DBA‚PF6 and DB24C8. In these experi-
ments, the concentration of DBA‚PF6 was held constant
at 29.3 mM, while the concentration of DB24C8 was
varied from 3 to 58 mM. The relative amount of free and
complexed DB24C8 was determined by comparing the
integrals for the resonances of the γ-OCH2 protons
(Figure 8) in the bound and unbound states of the crown
ether. For the individual data points, a binding isotherm
was constructed (Figure 9), values of KAVE were deter-
mined, and a mean value for KAVE of 478 ( 143 M-1 was
calculated.26 This value of KAVE corresponds to a mean
monium) cluster 4‚6PF6 was held constant at 4.88 mM,
+
i.e., the concentration of NH2 centers is 29.3 mM. The
concentrations of DB24C8 in these solutions were varied
1
from 0 to 59.9 mM. The H NMR spectra of the mixture
of the two compounds become increasingly complicated
on addition of increasing amounts of DB24C8. As the
concentration of DB24C8 is increased, the intensities of
the resonances corresponding to the protons in the
uncomplexed dialkylammonium ion sites decrease and
new resonances appear for the protons of the complexed
species. As there are many geometries possible for the
higher order complexes, exact assignments to particular
species are not possible. In the region of the 1H NMR
spectra representing signals of the aliphatic protons (3.2-
4.8 ppm), however, it is possible to assign some of the
resonances to protons in the complexed and uncomplexed
species. As the concentration of DB24C8 is increased, the
signals for the benzylic methylene protons adjacent to
value for ∆G°AVE of -3.7 ( 0.2 kcal mol-1
.
+
the NH2 centerssnamely HA and HB (see Figure 3 for
their assignment)sin the complexed DBA+ ions appear
in a characteristic region22e (4.3-4.6 ppm) of the spec-
trum. Also, the resonances corresponding to the R-, â-,
and γ-OCH2 protons (see Figure 2 for their assignment)
in the free and the complexed states of DB24C8 are
observed. By comparing the integrals of the resonances
corresponding to the γ-OCH2 protons in the bound
DB24C8 macrocycles and the â-OCH2 protons in the
uncomplexed DB24C8 rings, the relative concentrations
of DB24C8 in the free and complexed states were
determined.
1
From the data obtained from these H NMR spectra,
a binding isotherm was plotted (Figure 9) and a mean
value for KAVE of 55 ( 33 M-1 was determined. This value
of KAVE corresponds to a mean ∆G°AVE of -2.4 ( 0.5 kcal
mol-1. Since the individual values of KAVE vary consider-
ably with the concentration of DB24C8, a large error is
27
assumed for the mean value of KAVE
.
At a DB24C8
concentration of 2.3 mM, the observed value of KAVE is
10 M-1 while at a concentration of 14.8 mM, the value of
KAVE is 94 M-1. This change in the observed value of the
association constant is large and is probably outside the
error limits of the experiment. At low concentrations
(2.3-14.8 mM) of DB24C8, the observed values of KAVE
increase quite dramatically upon increasing the concen-
tration of the crown ether. At higher concentrations
(14.8-59.9 mM), however, the observed values of KAVE
decrease as the concentration of DB24C8 increases. This
latter observation can be explained by unfavorable steric
interactions between crown ethers bound to the same
hexakisammonium cluster. The increase, however, in the
value of KAVE upon increasing the concentration of
DB24C8 (between 2.3 and 14.8 mM) is not easy to
rationalize.28 It is possible that it is a result of experi-
mental error in determining integrated values of signals
at low concentrations.29
F IGURE 9. “Best-fit” binding isotherms for the complexation
between DB24C8 and DBA‚PF6 (9) and the complexation
between DB24C8 and the hexakis(dialkylammonium) cluster
4‚6PF6 (b). Next to each data point is displayed the value of
KAVE (M-1; errors (60%) determined by using the single-point
method. The values on the y-axis represent the measured
concentration of complexed DB24C8 divided by the total
concentration of dialkylammonium ion centers (error in y-axis
for upper isotherm: 14%; lower isotherm: 28%). The values
on the x-axis represent the concentration of free DB24C8 in
solution (error in upper isotherm 10%, lower isotherm 20%).
To draw a comparison between the complexation of
DB24C8 with DBA+ ions and the cluster 4‚6PF6, 1H NMR
spectra were obtained (Figure 8) for a series of solutions
of 4‚6PF6 and DB24C8 in CD3CN/CD2Cl2 (1:1). In these
solutions, the concentration of the hexakis(dialkylam-
From a comparison of the binding isotherms and the
mean free energy changes (Figure 10) upon complexation,
the DBA+ moieties in the hexakis(dialkylammonium)
cluster seem to bind DB24C8 ca. 1.3 kcal mol-1 less
effectively than does DBA‚PF6. This reduction in the
(26) Because of difficulties in integrating these spectra, we believe
that the error limits are large. Additionally, there is probably some
divergence from the ideal 1:1 binding isotherm because the δ values
of some of the signals shift upon addition of increasing amounts of the
crown ether. This divergence from the ideal behavior could be a
consequence of competing equilibria, such as, for example, ion pairing
between the dialkylammonium cation and its hexafluorophosphate
anion.
(27) We assume an error of (60% for values of Ka, which results
primarily from an assumed error of 10% in integrating signals in these
complicated NMR spectra.
J . Org. Chem, Vol. 69, No. 13, 2004 4397

Lowe et al.
strength of complexation is probably a consequence of
unfavorable steric interactions between crown ethers
bound to the same hexakis(dialkylammonium) cluster.
Also, the three dialkylammonium ion sites on each face
of the cluster may form a cavity that could bind favorably
with a hexafluorophosphate anion, producing an inter-
action that could compete with the binding of the crown
ethers.30 Taken together, these factors support the view
that the strength of binding between DB24C8 and the
hexakis(dialkylammonium) cluster 4‚6PF6 is expected to
be lower than that between DB24C8 and DBA+ ions in
the same solvent.
F IGURE 11. Partial 19F NMR spectrum [376 MHz, CD2Cl2/
CD3CN (1:3), 300 K] of an equimolar solution of BMP25C8 and
bis-DBA-F4‚PF6.
Stu d y of Ch ela te Effects in th e Bin d in g of th e
Hexa k is(cr ow n eth er ) Clu ster w ith Mon o- a n d Bis-
(d ia lk yla m m on iu m ) Ion s. To study the contribution of
chelate effects to multivalent interactions, we investi-
gated the binding of the bis(dialkylammonium) ion11 bis-
DBA-F4‚2PF6 (Figure 11) to the hexakisBMP25C8 cluster
3 and BMP25C8. An 19F NMR spectrum of a solution of
the bis(dialkylammonium) salt bis-DBA-F4‚2PF6 and
BMP25C8 displays (Figure 11) three resonances. These
resonances are a sharp singlet (-113.55 ppm) corre-
sponding to unbound bis-DBA-F42+ ions, a broad signal31
centered on -113.45 ppm, corresponding to the 1:1
complex, and another sharp singlet (-113.36 ppm),
corresponding to the 2:1 complex. From the integrals
associated with these resonances, it is possible to deter-
mine the concentrations of each species and hence a value
for KAVE can be calculated by using eq 1. Hence, from a
set of 19F NMR titration experiments32 in CD3CN/CD2-
Cl2 (3:1) with the concentration of BMP25C8 held at a
constant 6.0 mM and the concentration of salt varied
from 0 to 8 mM, a binding isotherm was plotted (Figure
12) and a mean value for KAVE of 48 ( 14 M-1 was
determined.33 This value of KAVE corresponds to a mean
F IGURE 10. A comparison of the free energy of binding
between (a) DBA+ and DB24C8 and (b) the hexakis(dialkyl-
ammonium) cluster 4‚6PF6 and DB24C8.
Thus, in both cases, clustering the crown ether rings
or dialkylammonium ion recognition sites reduces the
individual strength of binding of these moieties toward
complementary monovalent ligands. Thus, it seems that
steric and electronic factors play a more important role
in the binding of these clusters than do statistical effects.
(28) Ion pairing between the ammonium cluster and the hexafluo-
rophosphate anion may contribute to the changes in values of KAVE
.
At lower concentrations of the crown ethers, lower order complexes
are formed, which may weaken ion-pairing interactions between the
hexacationic host and the counterions, thus increasing the apparent
K
AVE. The weakening of the ion-pairing interaction upon binding of
the cationic centers could be a consequence of a bound cationic center
having a more dispersed positive charge than an unbound one. This
charge dispersal would result in reduced electrostatic charges between
the hexacation and any anions. This changing strength of ion pairing
may be a component of the effect of clustering multivalent cationic
receptors.
(29) The Weber rule of thumb (see, for instance: Tsukube, H.;
Furuta, H.; Odani, A.; Takeda, Y.; Kudo, Y.; Inoue, Y.; Liu, Y.;
Sakamoto, H.; Kimura, K. in Comprehensive Supramolecular Chem-
istry; Davies, J . E., Ed.; Pergamon Press: Oxford, UK, 1996; Vol. 8,
pp 425-482) states that accurate association constants can be obtained
only at total concentrations of host and guest at which the minor
component in the system is between 20% and 80% complexed. It may
appear at first glance that some of the data in the lower isotherm of
Figure 9 (i.e., those with values of Ka of 10 and 29 M-1) fall outside of
the Weber range. In actual fact, these data represent a situation in
which ca. 23% and 44%, respectively, of the crown ethersthe minor
componentsis complexed, and so the derived values of Ka seem to be
valid from the viewpoint of the Weber rule.
value for ∆G°AVE of -2.3 ( 0.2 kcal mol-1
.
(31) This signal is broad presumably because the crown ether moiety
+
shuttles back and forth between the two NH2 centers of the bis-
(dialkylammonium) ion at a rate commensurate with that of the NMR
time scale at 376 MHz. That is to say, the crown ether exists in discrete
“bound” and “unbound” states (i.e., slow exchange) because its rate of
passage over the 3,5-difluorophenyl units is slow, but within the 1:1
complex, the crown ether has less of a steric barrier for shuttling
between the binding sites. This lower shuttling barrier, however, is
not low enough to cause this process to be fast on the NMR time scale,
and so at 376 MHz the two termini of the bis(dialkylammonium) ion
are somewhat nonequivalent and, hence, their signals are broad.
(32) A ratio of 3:1 for CD3CN/CD2Cl2 was used as the solvent mixture
in these experiments because the bisammonium salt bis-DBA-F4‚2PF6
does not have the required solubility in CD3CN/CD2Cl2 (1:1) solvent
mixture.
(30) We have observed such interactions between dialkylammonium
ions and hexafluorophosphate anions (usually [C-H‚‚‚F] hydrogen
bonds) in many solid-state structures of similar host-guest systems.
See, for example: Ashton, P. R.; Fyfe, M. C. T.; Glink, P. T.; Menzer,
S.; Stoddart, J . F.; White, A. J . P.; Williams, D. J . J . Am. Chem. Soc.
1997, 119, 12514-12524.
4398 J . Org. Chem., Vol. 69, No. 13, 2004

Polyvalent Interactions
F IGURE 12. “Best-fit” binding isotherms for the complexation
between BMP25C8 and bis-DBA-F4‚2PF6 ([), the complexation
between the hexakisBMP25C8 cluster 3 and DBA-F4‚2PF6 (0),
and the complexation between BMP25C8 and DBA-F4‚PF6 (b).
The values on the y-axis represent the measured concentra-
tions of complexed DB24C8 divided by the total concentration
of dialkylammonium ion centers (errors (14%). The values on
the x-axis represent the concentration of free NH2+ centers in
solution (errors (10%). Some error bars have been omitted
for the sake of clarity.
F IGURE 13. Partial 19F NMR spectrum [376 MHz, CD2Cl2/
CD3CN (1:3), 300 K] of an equimolar solution of the hexakis-
BMP25C8 cluster 3 and bis-DBA-F4‚2PF6.
value for ∆G°AVE of -2.2 ( 0.2 kcal mol-1. From a
comparison of the binding isotherms and the mean values
for ∆G°AVE, it appears that the strength of binding
The 19F NMR spectrum of a solution of bis-DBA-F4‚
2PF6 and the hexakisBMP25C8 cluster 3 in CD3CN/CD2-
Cl2 (3:1) displays (Figure 13) a sharp resonance (-113.30
ppm), corresponding to the fluorine atoms associated with
+
between individual crown ether moieties and NH2
centers in these three binding systems differs only
marginally.35,36
+
2+
free NH2 centers of the bis-DBA-F4
ions, and a
Ca lcu la t in g Mu lt iva len t Bin d in g Affin it ies. For
the binding between the hexakisBMP25C8 cluster 3 and
the bis(dialkylammonium) salt bis-DBA-F4‚2PF6, a bind-
ing constant (KPOLY),1,11 which corresponds to the binding
constant in the complexation of 1 equiv of bis-DBA-F4
ions with 1 equiv of the hexakisBMP25C8 cluster 3 to
complicated set of overlapping signals in the range
-112.8 to -113.2 ppm that corresponds to fluorine atoms
of these dications having complexed NH2+ centers. Once
again, some of the signals most likely represent fluorine
atoms of the dication existing in 1:1 and 2:1 complexes,
i.e., there is a broad signal centered at -113.1 ppm, which
suggests complexes displaying shuttling, and sharp
signals between -112.8 and -113.0 ppm, which suggest
[3]pseudorotaxane-like states. By comparing the integrals
of the two main sets of resonances (i.e., total complexed
and total uncomplexed), the concentrations of complexed
2+
afford the 1:1 complex, can be determined from the mean
value of KAVE
.
The average free energy change for an individual
interaction is defined as
+
∆G°AVE ) - RT lnKAVE
(2)
and free NH2 centers were determined and values for
KAVE were calculated with eq 1. Hence, from a set of 19F
NMR titration experiments in CD3CN/CD2Cl2 (3:1) with
the concentration of the BMP25C8 cluster 3 held constant
(1.0 mM) and the concentration of salt varied from 0 to
8 mM, a binding isotherm was plotted (Figure 12) and a
mean value for KAVE of 39 ( 12 M-1 was determined. This
value of KAVE corresponds to a mean value for ∆G°AVE of
The free energy of binding between two multivalent
systems that have N individual component interactions
between them can be estimated as the sum of these
component interactions:
∆G°POLY ) N∆G°AVE
(3)
-2.2 ( 0.2 kcal mol-1
.
This equation can be expressed in terms of association
constants KPOLY and KAVE as follows:
For comparison, a binding isotherm was plotted (Fig-
ure 12) and the mean values for KAVE and ∆G°AVE were
found for the binding of BMP25C8 with DBA-F4‚PF6 over
a similar range of concentrations in this solvent system
[CD3CN/CD2Cl2 (3:1)].34 A mean value for KAVE of 39 (
12 M-1 was calculated, which corresponds to a mean
N
KPOLY ) (KAVE
)
(4)
Thus, a value of KPOLY can be calculated for the binding
2+
between one Bis-DBA-F4 dication and one hexakis-
(33) The corresponding value of KAVE has been calculated in CD3-
CN to be slightly lower, as expected for a more polar solvent, at 30 (
(34) Recall that we had calculated this association constant in CD3-
9 M-1 (∆G° ) -1.8 ( 0.2 kcal mol-1). See ref 11.
CN/CD2Cl2 solution (1:1) to be 116 ( 35 M-1. See Table 1, entry 3.
AVE
J . Org. Chem, Vol. 69, No. 13, 2004 4399

Lowe et al.
TABLE 1. Va lu es of Mea n KAVE a n d Mea n ∆G°AVE for th e In ter a ction s betw een Va r iou s Cr ow n Eth er Hosts a n d
Dia lk yla m m on iu m Gu ests a t 300 K
entry
host
guest
solvent
K
AVE (M-1
)
∆G°AVE (kcal mol-1
)
1a
2b
3
4
5b
6
BMP25C8
BMP25C8
BMP25C8
BMP25C8
BMP25C8
BMP25C8
3
DBA-F4‚PF6
DBA-F4‚PF6
DBA-F4‚PF6
DBA-F4‚PF6
bis-DBA-F4‚2PF6
bis-DBA-F4‚2PF6
DBA-F4‚PF6
bis-DBA-F4‚2PF6
DBA‚PF6
CD3CN
CD3CN
25
20 ( 6
-1.9
-1.8 ( 0.2
-2.8 ( 0.2
-2.2 ( 0.2
-1.8 ( 0.2
-2.3 ( 0.2
-2.3 ( 0.2
-2.2 ( 0.2
-3.6
CD3CN/CD2Cl2, 1:1
CD3CN/CD2Cl2, 3:1
CD3CN
CD3CN/CD2Cl2, 3:1
CD3CN/CD2Cl2, 1:1
CD3CN/CD2Cl2, 3:1
CD3CN
116 ( 35
39 ( 12
30 ( 6
48 ( 14
48 ( 14
39 ( 12
420
7
8c
9d
10
11
3
DB24C8
DB24C8
DB24C8
DBA‚PF6
CD3CN/CD2Cl2, 1:1
CD3CN/CD2Cl2, 1:1
478 ( 143
55 ( 33
-3.7 ( 0.2
-2.4 ( 0.5
4‚6PF6
a
b
Values taken from ref 7e. Values taken from ref 11. c The value of KPOLY for this interaction is 1500 ( 1100 M-1 (∆G°POLY ) 4.4 (
0.5 kcal mol-1). Value taken from ref 6a.
d
BMP25C8 cluster 3. When these two species bind to-
gether, there are two individual component interactions
(i.e., N ) 2) because when one dication is complexed by
and ligands, any potential for enhanced binding expected
from statistical effects is overwhelmed by unfavorable
electronic and steric factors, when considering the bind-
ing of monovalent species to a complementary multi-
valent cluster. Although the clustering of binding sites
reduces the strength of binding of the hexakisBMP25C8
cluster to monovalent cations (∆G°AVE ) ca. -2.3 kcal
mol-1) relative to the monovalent species (∆G°AVE ) -2.8
kcal mol-1), it increases the strength of the binding of
the cluster to multivalent species, such as the bis-
+
one cluster, two NH2 centers are bound. Hence, KPOLY
is calculated to be 1500 ( 1100 M-1 [i.e., (39 ( 12)2] in
CD3CN/CD2Cl2 (3:1), which corresponds to a value for
∆G°POLY of -4.4 ( 0.5 kcal mol-1. The corresponding
monovalent interaction between BMP25C8 with DBA-
F4‚PF6 has a value of Ka, coincidentally, of 39 M-1 in CD3-
CN/CD2Cl2 (3:1). Hence, the strength of interaction
between the two species acting multivalently is signifi-
cantly stronger than the component interactions between
any of the structurally similar monovalent species ex-
amined in this study. This increased strength of binding
between polyvalent species indicates that strong chelate
effects in multivalent systems can overcome adverse
electronic and steric effects. Table 1 summarizes the
binding constants obtained in this study.
2+
(dialkylammonium) ion Bis-DBA-F4 (∆G°POLY ) ca.
-4.4 kcal mol-1), in similar solvents. This result dem-
onstrates the large influence that chelate effects can have
in the binding of two polyvalent species. This study
demonstrates that multivalency can be employed ef-
fectively in unnatural systems and emphasizes the
importance of understanding the mechanisms through
which it occurs. By borrowing the concepts of multi- and
polyvalency from the natural world, chemists can solve
problems caused by monovalent species interacting only
relatively weakly by clustering many such ligands to-
gether to obtain strong interactions between polyvalent
versions of the same species.1,4,5 Although this and other
related10,11 studies provide some clues as to how chemists
can use multivalency in designing strongly binding
systems, optimizing the components in the systemss
recognition centers, core units, and linkerssremains a
challenge still to be met.
Con clu sion s
We have synthesized two cluster compounds, one
containing six BMP25C8 rings and the other possessing
six dialkylammonium ion centers attached to a hexakis-
(thiophenyl)benzene core. The clustering of the crown
ether rings and the NH2+ centers in these compounds is
detrimental to the strengths of binding by the individual
+
crown ethers or NH2 centers to monovalent ligands
(compare entries 3 and 7, and 10 and 11, in Table 1). In
the case of the binding between DB24C8 and the hexakis-
(dialkylammonium) cluster 4‚6PF6, the reduction in
binding strength is likely a consequence of unfavorable
steric interactions between adjacent crown ether units
bound to the same core unit. In addition to steric factors,
electronic repulsion between two or more dialkylammo-
nium cations reduces the strengths of binding of DBA-
F4+ ions to the hexakisBMP25C8 cluster 3. These results
demonstrate that, in the case of these simple receptors
Exp er im en ta l Section
Hexa k is(4-m eth oxyp h en ylth io)ben zen e (5). NaH (1.0 g,
43 mmol) was suspended in degassed 1,3-dimethyl-2-imida-
zolidinone (10 mL). The suspension was again degassed and
cooled to 0 °C, and then 4-methoxythiophenol (1.8 g, 13 mmol)
was added carefully. The reaction mixture was placed under
an Ar atmosphere and hexafluorobenzene (134 mg, 0.7 mmol)
was added. The mixture was stirred under Ar for 6 h at room
temperature, during which time a yellow suspension formed.
This suspension was partitioned between H2O (100 mL) and
PhMe (100 mL). The organic layer was then washed with H2O
(10 × 100 mL) and dried (Na2SO4) and the solvent was
removed in vacuo to give a yellow oil. EtOH (5 mL) was added
to this yellow oil and after 1 h a yellow solid had formed. The
yellow solid was filtered off and washed with EtOH to afford
the title compound 5 (0.65 g, 95%). Mp 158-161 °C; FAB-MS
m/z 906 [M]+; 1H NMR (500 MHz, CDCl3) δ 3.75 (s, 18H), 6.67
(d, J ) 8 Hz, 12H), 6.89 (d, J ) 8 Hz, 12H); 13C NMR (125
MHz, CDCl3) δ 55.1, 114.3, 128.5, 130.8, 147.6, 158.2. Anal.
(35) In an earlier study, we found that a cluster of seven BMP25C8
units appended to a â-cyclodextrin core exhibits a significantly higher
value of KAVE (59 ( 18 M-1) toward bis-DBA-F42+ than did BMP25C8
itself (20 ( 9 M-1). See ref 11.
(36) With this divalent cationic salt, we would expect ion pairing to
play a role in binding behavior. However, given that the values of KAVE
vary slightly (62 to 42 M-1 for BMP25C8 with bis-DBA-F4‚2PF6 and
43 to 33 M-1 for 3 with bis-DBA-F4‚2PF6), this effect does not change
the conclusion that the binding strength of these systems differs only
marginally.
4400 J . Org. Chem., Vol. 69, No. 13, 2004

Polyvalent Interactions
Calcd for C48H42O6S6: C, 63.55; H, 4.67; S, 21.21. Found: C,
63.49; H, 4.68; S, 21.18.
with 1 M Na2S2O3 (100 mL) and dried (MgSO4), and then the
solvent was evaporated in vacuo. The resulting yellow solid
was subjected to column chromatography (SiO2; PhMe) to
afford the title compound 11 as a white solid (3.3 g, 50%). Mp
Hexa k is(4-h yd r oxyp h en ylth io)ben zen e (6). A solution
of BBr3 in CH2Cl2 (1 M, 7.9 mL, 7.9 mmol) was added over 30
min to a solution of hexakis(4-methoxyphenylthio)benzene (5)
(0.22 g, 0.24 mmol) in dry CH2Cl2 under Ar at -78 °C. The
solution was then warmed to room temperature and stirred
under Ar overnight. H2O (10 mL) was added carefully to the
reaction mixture and then EtOAc (20 mL) was also added. The
organic phase was washed with H2O (35 mL) and saturated
aqueous NaCl (35 mL) and dried (MgSO4), and then the solvent
was evaporated in vacuo to give a yellow solid. This solid was
subjected to column chromatography (SiO2; EtOAc/PhMe, 1:1
to 3:1) to afford the title compound 6 as a yellow solid (0.10 g,
1
98-103 °C; EI-MS m/z 274 [M]+; H NMR (400 MHz, CDCl3)
δ 7.62 (d, J ) 8 Hz, 4H), 7.82 (d, J ) 8 Hz, 4H), 9.96 (s, 2H);
13C NMR (100 MHz, CDCl3) δ 126.3, 130.4, 135.1, 143.8, 191.0.
Anal. Calcd for C14H10O2S2: C, 61.29; H, 3.67; S, 23.38.
Found: C, 60.94; H, 3.61; S, 23.18.
4,4′-Dith iobis(ben zylid en eben zyla m in e) (12). A solution
of 4,4′-dithiobis(benzaldehyde) (11) (3.0 g, 10.9 mmol) and
benzylamine (2.3 g, 21.8 mmol) in PhMe (200 mL) was heated
overnight under reflux in a Dean-Stark apparatus. The
reaction mixture was then cooled to room temperature and
the solvent was removed under reduced pressure to afford the
title compound 12 as a yellow solid (4.9 g, 99%). FAB-MS m/z
453 [M + H]+; 1H NMR (400 MHz, CDCl3) δ 4.81 (s, 4H), 7.10-
7.50 (m, 10H), 7.52 (d, J ) 8 Hz, 4H), 7.71 (d, J ) 8 Hz, 4H),
8.35 (s, 2H); 13C NMR (100 MHz, CDCl3) δ 65.0, 127.0, 127.1,
128.0, 128.5, 128.9, 135.1, 139.1, 139.6, 160.9.
1
63%). Mp 240 °C dec; FAB-MS m/z 822 [M]+; H NMR (500
MHz, CD3COCD3) δ 6.73 (d, J ) 8 Hz, 12H), 6.88 (d, J ) 8
Hz, 12H), 8.52 (s, 6H); 13C NMR (125 MHz, CD3COCD3) δ
115.9, 127.3, 130.7, 147.6, 156.2. Anal. Calcd for C42H30O6S6:
C, 61.29; H, 3.67; S, 23.38. Found: C, 61.15; H, 3.68; S, 23.15.
Hexa (ben zo-m -p h en ylen e[25]cr ow n -8) Clu st er (3). A
solution of mesylate 9 (1.1 g, 1.98 mmol) in dry MeCN (7.5
mL) was added to a hot suspension of hexakis(4-hydroxyphen-
ylthio)benzene (6) (90 mg, 0.11 mmol), K2CO3 (270 mg, 1.98
mmol), and [18]crown-6 (29 mg, 0.11 mmol) in dry MeCN (7.5
mL) under Ar. The reaction mixture was heated under reflux
overnight. After the reaction mixture had cooled to room
temperature, it was partitioned between EtOAc (40 mL) and
H2O (40 mL). The organic phase was dried (MgSO4) and the
solvent was removed in vacuo. The resulting yellow oil was
subjected to HPLC (reverse phase C18 silica; MeCN/H2O, 1:1
to 1:0 over 30 min, then MeCN, 10 min) to yield the title
compound 3 as a yellow oil (0.16 g, 40%). MALDI-TOF-MS m/z
4-(Ben zyla m in om eth yl)th iop h en ol (13). NaBH4 (2.1 g,
55 mmol) was added portionwise over 40 min to a solution of
4,4′-dithiobis(benzylidenebenzylamine) (12) (4.9 g, 10.9 mmol)
in degassed dry MeOH (50 mL) and degassed dry THF (50
mL) under Ar. The reaction mixture was stirred overnight.
The solvents were then evaporated and the resulting yellow
residue was partitioned between degassed CH2Cl2 (50 mL) and
degassed H2O (50 mL). The aqueous layer was washed with
degassed CH2Cl2 (50 mL). The organic layers were combined
and dried (MgSO4) under Ar, and then the solvent was
evaporated in vacuo to afford the title compound 13 as a yellow
1
oil (4.9 g, 99%). H NMR (500 MHz, CDCl3) δ 3.78-3.81 (m,
1
3069 [M + Na]+; H NMR (500 MHz, CDCl3) δ 3.67-3.71 (m,
4H), 7.24-7.34 (m, 7H), 7.45 (d, J ) 8 Hz, 2H); 13C NMR (125
MHz, CDCl3) δ 52.5, 53.0, 126.9, 128.0, 128.3, 128.7, 131.0,
133.1, 139.5, 140.3.
48H), 3.77-3.80 (m, 24H), 3.82-3.85 (m, 24H), 4.09-4.14 (m,
48H), 4.88 (s, 12H), 6.51 (d, J ) 2 Hz, 12H), 6.62 (t, J ) 2 Hz,
6H), 6.72 (d, J ) 8 Hz, 12H), 6.86-6.90 (m, 36H); 13C NMR
(125 MHz, CDCl3) δ 68.0, 68.9, 69.7, 69.8, 70.8, 70.9, 101.9,
106.7, 115.1, 115.3, 121.6, 128.8, 130.7, 138.8, 147.3, 148.9,
157.5, 160.1.
H exa k is[4-(b en zyla m in om et h yl)p h en ylt h io]b en zen e
(14). 4-(Benzylaminomethyl)thiobenzene (13) (4.9 g, 10.7
mmol) was dissolved in degassed 1,3-dimethyl-2-imidazolidi-
none (10 mL). This solution was then degassed further with a
stream of Ar. NaH (0.8 g, 33 mmol) was added portionwise.
After each portion of NaH had been added, the reaction
mixture was carefully degassed under vacuum. The reaction
mixture was then placed under an Ar atmosphere and
hexafluorobenzene (0.11 g, 0.6 mmol) was added. The reaction
mixture was stirred under Ar for 3 d before being partitioned
between PhMe (100 mL) and H2O (100 mL). The aqueous layer
was separated and washed with PhMe (100 mL). The organic
layers were combined and washed with H2O (5 × 100 mL).
The solvent was removed in vacuo to give a dark orange oil.
This oil was subjected to column chromatography (SiO2; CH2-
Cl2/THF/NEt3, 99:0:1 to 90:9:1 to 50:50:1) to afford the title
compound 14 as a yellow oil (0.51 g, 58%). FAB-MS m/z 1440
[M]+; 1H NMR (500 MHz, CDCl3) δ 3.61 (s, 12H), 3.78 (s, 12H),
6.93 (d, J ) 8 Hz, 12H), 7.16 (d, J ) 8 Hz, 12H), 7.28-7.37
(m, 30H); 13C NMR (125 MHz, CDCl3) δ 52.4, 53.0, 126.9, 128.0,
128.1, 128.3, 128.6, 136.0, 138.1, 139.9, 147.7.
H exa k is[4-(b en zyla m m on iom et h yl)p h en ylt h io]b en -
zen e Hexa k is(h exa flu or op h osp h a te) (4‚6P F 6). Hexakis-
[4-(benzylaminomethyl)phenylthio]benzene (14) (87 mg, 0.6
mmol) was dissolved in THF (2 mL). MeOH (10 mL) was added
to this solution to form a cloudy suspension and then concen-
trated HCl (2 mL) was added. The reaction mixture was stirred
overnight to form a yellow precipitate, which was filtered off,
washed with THF (5 mL), and then dissolved in hot H2O (50
mL). Saturated aqueous NH4PF6 (1 mL) was added to this
solution and the yellow suspension that formed was extracted
with MeNO2 (50 mL). The organic phase was washed with H2O
(2 × 40 mL) and dried (Na2SO4), and then the solvent was
evaporated in vacuo to afford the title compound 4‚6PF6 as a
yellow solid (0.12 g, 88%). Mp 150 °C dec; ES-MS m/z 941 [M
- HPF6 - 2PF6]2+; 1H NMR (400 MHz, CD3CN) δ 4.08 (s, 12H),
4.12 (s, 12H), 6.99 (d, J ) 8 Hz, 12H), 7.33 (d, J ) 8 Hz, 12H),
4-(Meth a n esu lfin yl)ben za ld eh yd e (10). 3-Chloroperoxy-
benzoic acid (13.0 g, 75 mmol) was added portionwise to a
solution of 4-(methylthio)benzaldehyde (11.4 g, 75 mmol) in
CHCl3 (200 mL) at 0 °C. The suspension was then stirred at
0 °C for 90 min. After the solution was warmed to room
temperature, Ca(OH)2 (8.3 g, 110 mmol) was added and the
suspension was stirred for a period of 30 min before being
filtered. The solvent was removed from the filtrate in vacuo,
leaving a white solid. This solid was subjected to column
chromatography (SiO2; CH2Cl2/EtOAc, 1:0 to 1:1) to yield the
title compound 10 as a white solid (7.5 g, 59%). Mp 85-86 °C
(lit.37 mp 81-82 °C); EI-MS m/z 169 [M]+; 1H NMR (200 MHz,
CDCl3) δ 2.77 (s, 3H), 7.81 (d, J ) 8 Hz, 2H), 8.06 (d, J ) 8
Hz, 2H), 10.11 (s, 1H); 13C NMR (125 MHz, CDCl3) δ 43.7,
124.2, 130.4, 138.1, 152.4, 191.1. Anal. Calcd for C8H8O2S: C,
57.12; H, 4.79; S, 19.06. Found: C, 56.98; H, 4.68; S, 19.07.
4,4′-Dith iobis(ben za ld eh yd e) (11). 4-(Methanesulfinyl)-
benzaldehyde (10) (8.1 g, 48 mmol) was heated under reflux
in freshly distilled trifluoroacetic acid anhydride (100 mL)
under Ar for 15 min. The solvent was then removed in vacuo
and an ice-cooled mixture of NEt3 (100 mL) and MeOH (100
mL) was added carefully to the resulting residue. The solvent
was removed in vacuo from the resulting solution to afford a
yellow residue. This residue was partitioned between CHCl3
(100 mL) and saturated aqueous NH4Cl (100 mL), and the
aqueous phase was washed with CHCl3 (100 mL). The organic
fractions were combined and dried (MgSO4), and the solvent
was evaporated in vacuo to give a yellow oil. The oil was
dissolved in CH2Cl2 (100 mL) and I2 (12.2 g, 48 mmol) was
added. This mixture was stirred for a period of 4 h then washed
(37) Creary, X.; Mersheikh-Mohammadi, M. E. J . Org. Chem. 1986,
51, 1110-1114.
J . Org. Chem, Vol. 69, No. 13, 2004 4401

Lowe et al.
7.40-7.50 (m, 30H); 13C NMR (125 MHz, CD3CN) δ 50.5, 51.1,
Su p p or tin g In for m a tion Ava ila ble: The 1H and 13C
NMR spectra of compounds 3, 4‚6PF6, 12, 13, and 14. This
material is available free of charge via the Internet at
http://pubs.acs.org.
127.3, 128.4, 129.0, 129.7, 130.1, 130.2, 131.1, 139.0, 147.5.
Ack n ow led gm en t. This research was supported by
the National Science Foundation (CHE-9910199). Dr.
Peter T. Glink is acknowledged for his considerable help
in the preparation of this manuscript.
J O030283O
4402 J . Org. Chem., Vol. 69, No. 13, 2004