Methylene-bridged glycoluril dimers: Synthetic methods

Article abstract of DOI:10.1021/jo0258958

Methylene-bridged glycoluril dimers are the fundamental building blocks of cucurbituril (CB[6]), its homologues (CB[n]), and its derivatives. This paper describes three complementary methods for the synthesis of C- and S-shaped methylene-bridged glycoluril dimers (29-34 and 37-44). For this purpose, we prepared glycoluril derivatives (1a-d) bearing diverse functionalities on their convex face. These glycoluril derivatives were alkylated under basic conditions (DMSO, t-BuOK) with 1,2-bis(halomethyl)aromatics 6-15 to yield 4a-d and 16-24, which contain a single aromatic o-xylylene ring and potentially nucleophilic ureidyl NH groups. Glycoluril derivatives bearing potentially electrophilic cyclic ether groups (5a-f) and 25-28 were prepared by various methods including condensation reactions in refluxing TFA containing paraformaldehyde. The condensation reactions of 4a-d and 16-24 with paraformaldehyde under anhydrous acidic conditions (PTSA, ClCH₂CH₂Cl, reflux) give, in most cases, the C-shaped and S-shaped methylene-bridged glycoluril in good to excellent yields. In many cases, the C-shaped compound is formed preferentially with high diastereoselectivity. Cyclic ethers 5a,d-f and 25-26 undergo highly diastereoselective dimerization reactions to yield methylene-bridged glycoluril dimers with the formal extrusion of formaldehyde. Last, it is possible to perform selective heterodimerization reactions using both cyclic ethers and glycoluril derivatives bearing ureidyl NH groups. These reactions deliver the desired C- and S-shaped heterodimers with low to moderate diastereoselectivities. This heterodimerization route is the method of choice in cases where the homodimerization reactions fail. The formation of side products (±)-35b and (±)-35d helps clarify the electronic requirements for a successful CB[n] synthesis. The x-ray structures of 30C, 38C, and 38S allow for a discussion of the structural features of this class of compounds.

Full text of DOI:10.1021/jo0258958

Meth ylen e-Br id ged Glycolu r il Dim er s: Syn th etic Meth od s
Anxin Wu,†,‡ Arindam Chakraborty,† Dariusz Witt,†,§ J ason Lagona,† Fehmi Damkaci,†
Marie A. Ofori,† J essica K. Chiles,† J ames C. Fettinger,† and Lyle Isaacs*,†
Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742, and
National Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, P. R. China
LI8@umail.umd.edu
Received May 2, 2002
Methylene-bridged glycoluril dimers are the fundamental building blocks of cucurbituril (CB[6]),
its homologues (CB[n]), and its derivatives. This paper describes three complementary methods
for the synthesis of C- and S-shaped methylene-bridged glycoluril dimers (29-34 and 37-44). For
this purpose, we prepared glycoluril derivatives (1a -d ) bearing diverse functionalities on their
convex face. These glycoluril derivatives were alkylated under basic conditions (DMSO, t-BuOK)
with 1,2-bis(halomethyl)aromatics 6-15 to yield 4a -d and 16-24, which contain a single aromatic
o-xylylene ring and potentially nucleophilic ureidyl NH groups. Glycoluril derivatives bearing
potentially electrophilic cyclic ether groups (5a -f) and 25-28 were prepared by various methods
including condensation reactions in refluxing TFA containing paraformaldehyde. The condensation
reactions of 4a -d and 16-24 with paraformaldehyde under anhydrous acidic conditions (PTSA,
ClCH2CH2Cl, reflux) give, in most cases, the C-shaped and S-shaped methylene-bridged glycoluril
in good to excellent yields. In many cases, the C-shaped compound is formed preferentially with
high diastereoselectivity. Cyclic ethers 5a ,d -f and 25-26 undergo highly diastereoselective
dimerization reactions to yield methylene-bridged glycoluril dimers with the formal extrusion of
formaldehyde. Last, it is possible to perform selective heterodimerization reactions using both cyclic
ethers and glycoluril derivatives bearing ureidyl NH groups. These reactions deliver the desired
C- and S-shaped heterodimers with low to moderate diastereoselectivities. This heterodimerization
route is the method of choice in cases where the homodimerization reactions fail. The formation of
side products (()-35b and (()-35d helps clarify the electronic requirements for a successful CB[n]
synthesis. The X-ray structures of 30C, 38C, and 38S allow for a discussion of the structural features
of this class of compounds.
In tr od u ction
conditions (H2SO4, 135-145 °C) results in the formation
of CB[6] in high yield.4 This straightforward synthetic
method has allowed the use of CB[6] in many elegant
studies including molecular necklaces,5 bowls,6 polyro-
taxanes,7 DNA complexes,8 molecular switches,9 removal
of contaminants from aqueous waste streams,10 studies
of molecular polarizability,11 and ion and molecular
complexation studies.2,12
In efforts to expand the range of applications, several
groups have been investigating the preparation of con-
geners of cucurbituril that display enhanced properties.
This line of inquiry was first pursued by Stoddart, who
prepared Me10CB[5] by condensation of 1e with formal-
dehyde.13 More recently, Kim and co-workers14,15 as well
as Day and co-workers16,17 isolated homologues of cucur-
bituril comprising five, seven, eight, and ten glycoluril
Cucurbituril (CB[6]) is a an intriguing macrocyclic
compound comprising six glycoluril (1f) rings and twelve
methylene bridges whose structure was established by
Mock in 1981.1 CB[6] possesses a hydrophobic cavity with
carbonyl-lined portals that results in remarkable molec-
ular recognition properties (Chart 1). For example, Mock
and co-workers found that CB[6] binds tightly (Kd ≈ 1
µM) to alkyldiammonium ions in aqueous solution by a
combination of the hydrophobic effect and ion-dipole
interactions.2 It was also demonstrated that CB[6] is an
efficient enzyme mimic capable of catalyzing the dipolar
cycloaddition between azide and acetylene-derivatized
ammonium ions by their simultaneous binding within the
cavity of CB[6].3 The synthetic method used to prepare
cucurbituril is equally impressive; simply heating gly-
coluril (1a ) and formaldehyde under strongly acidic
(3) Mock, W. L.; Irra, T. A.; Wepsiec, J . P.; Manimaran, T. L. J . Org.
Chem. 1983, 48, 3619-3820. Mock, W. L.; Irra, T. A.; Wepsiec, J . P.;
Adhya, M. J . Org. Chem. 1989, 54, 5302-5308.
* To whom correspondence should be addressed.
† University of Maryland.
(4) Buschmann, H.-J .; Fink, H.; Schollmeyer, E. Preparation of
Cucurbituril. German Patent DE 196 03 377 A1, 1997.
(5) Whang, D.; Park, K.-M.; Heo, J .; Ashton, P.; Kim, K. J . Am.
Chem. Soc. 1998, 120, 4899-4900. Roh, S.-G.; Park, K.-M.; Park, G.-
J .; Sakamoto, S.; Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 1999,
38, 638-641. Lee, E.; Kim, J .; Heo, J .; Whang, D.; Kim, K. Angew.
Chem., Int. Ed. 2001, 40, 399-402.
‡ Lanzhou University.
§ Current address: Chemical Faculty, Technical University of
Gdansk, Narutowicza 11/12, 80-952 Gdansk, Poland.
(1) Freeman, W. A.; Mock, W. L.; Shih, N. Y. J . Am. Chem. Soc.
1981, 103, 7367-7368.
(2) Mock, W. L.; Shih, N. Y. J . Org. Chem. 1983, 48, 3618-3819.
Mock, W. L.; Shih, N. Y. J . Org. Chem. 1986, 51, 4440-4446. Mock,
W. L.; Shih, N. Y. J . Am. Chem. Soc. 1988, 110, 4706-4610. Mock, W.
L.; Shih, N. Y. J . Am. Chem. Soc. 1989, 111, 2697-2679.
(6) J eon, Y.-M.; Kim, J .; Whang, D.; Kim, K. J . Am. Chem. Soc. 1996,
118, 9790-9791. Whang, D.; Heo, J .; Park, J . H.; Kim, K. Angew.
Chem., Int. Ed. 1998, 37, 78-80.
10.1021/jo0258958 CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/03/2002
J . Org. Chem. 2002, 67, 5817-5830
5817

Wu et al.
CHART 1
Stoddart, Kim, and Day, we have chosen to pursue a
multistep synthetic approach. Such an approach, while
inherently more labor intensive, affords greater struc-
tural control, may generate mechanistic insights that
result in cucurbituril syntheses with enhanced scope, and
offers the opportunity to study the self-assembly and
molecular recognition properties of intermediates en-
route to congeners of cucurbituril.27 Our approach28 to
the synthesis of congeners of CB[6] relies on the identi-
fication of the methylene-bridged glycoluril dimer sub-
structure (2, bold in CB[6]) as the essential building block
for cucurbituril derivatives (Scheme 1). In this paper we
present three complementary synthetic routes to meth-
ylene-bridged glycoluril dimers. We also present the
X-ray crystallographic characterization of these two
diastereomers.
Resu lts a n d Discu ssion
Exp er im en ta l Design . To develop flexible methods
for the synthesis of derivatives of CB[n], we have initially
focused our attention on the preparation of methylene-
bridged glycoluril dimers (2C and 2S) which constitute
units (CB[5], CB[7], CB[8], and CB[10]) and detected
other homologues by performing the condensation reac-
tion under milder, kinetically controlled conditions. These
advances have already expanded the range of molecular
recognition applications18-20 of these systems to include
molecular Russian dolls,21 ball bearings,22 gyroscopes,17b
the catalysis of a [2 + 2] photoreaction,23 and the selective
recognition of a charge-transfer complex.24 Most recently,
Kim’s group has demonstrated that cyclohexyl-fused
glycoluril 1d is transformed into Cy5CB[5] and Cy6CB[6]25
whereas Nakamura’s group isolated the partially sub-
stituted Ph2CB[6].26
We have also been interested in tailoring the recogni-
tion properties of CB[n] by preparing derivatives func-
tionalized around their equator, at their methylene
bridges, or by substitution of an aromatic ring for a
glycoluril ring. In contrast to the one-step syntheses of
(12) Hoffmann, R.; Knoche, W.; Fenn, C.; Buschmann, H.-J . J . Chem.
Soc., Faraday Trans. 1994, 90, 1507-11. Meschke, C.; Buschmann,
H. J .; Schollmeyer, E. Thermochim. Acta 1997, 297, 43-48. Bus-
chmann, H. J .; J ansen, K.; Meschke, C.; Schollmeyer, E. J . Solution
Chem. 1998, 27, 135-140. Buschmann, H.-J .; J ansen, K.; Schollmeyer,
E. Acta Chim. Slov. 1999, 46, 405-411; Buschmann, H. J .; J ansen,
K.; Schollmeyer, E. Thermochim. Acta 2000, 346, 33-36. Buschmann,
H. J .; Cleve, E.; J ansen, K.; Wego, A.; Schollmeyer, E. J . Inclusion
Phenom. Macrocyclic Chem. 2001, 40, 117-120. Buschmann, H. J .;
Cleve, E.; J ansen, K.; Schollmeyer, E. Anal. Chim. Acta 2001, 437,
157-163. El Haouaj, M.; Young, H. K.; Luhmer, M.; Kim, K.; Bartik,
K. J . Chem. Soc., Perkin Trans. 2 2001, 2104-2107. El Haouaj, M.;
Luhmer, M.; Ko, Y. H.; Kim, K.; Bartik, K. J . Chem. Soc., Perkin Trans.
2 2001, 804-807. Marquez, C.; Nau, W. M. Angew. Chem., Int. Ed.
2001, 40, 3155-3160. Neugebauer, R.; Knoche, W. J . Chem. Soc.,
Perkin Trans. 2 1998, 529-534. Zhang, X. X.; Krakowiak, K. E.; Xue,
G.; Bradshaw, J . S.; Izatt, R. M. Ind. Eng. Chem. Res. 2000, 39, 3516-
3520. Wagner, B. D.; Fitzpatrick, S. J .; Gill, M. A.; MacRae, A. I.;
Stojanovic, N. Can. J . Chem. 2001, 79, 1101-1104.
(13) Flinn, A.; Hough, G. C.; Stoddart, J . F.; Williams, D. J . Angew.
Chem., Int. Ed. Engl. 1992, 31, 1475-1477.
(14) Kim, K.; Kim, J .; J ung, I.-S.; Kim, S.-Y.; Lee, E.; Kang, J .-K.
Cucurbituril Derivatives, Their Preparation and Uses. European
Patent Appl. EP 1 094 065 A2, 2001.
(7) Whang, D.; J eon, Y.-M.; Heo, J .; Kim, K. J . Am. Chem. Soc. 1996,
118, 11333-11334. Whang, D.; Kim, K. J . Am. Chem. Soc. 1997, 119,
451-452. Whang, D.; Heo, J .; Kim, C.-A.; Kim, K. Chem. Commun.
1997, 2361-2362. Lee, E.; Heo, J .; Kim, K. Angew. Chem., Int. Ed.
2000, 39, 2699-2701. Lee, J . W.; Ko, Y. H.; Park, S.-H.; Yamaguchi,
K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 746-749. Park, K.-M.;
Whang, D.; Lee, E.; Heo, J .; Kim, K. Chem.sEur. J . 2002, 8, 498-
508. Meschke, C.; Buschmann, H.-J .; Schollmeyer, E. Polymer 1998,
40, 945-949; Meschke, C.; Buschmann, H. J .; Schollmeyer, E. Mac-
romol. Rapid Commun. 1998, 19, 59-63. Buschmann, H. J .; Wego,
A.; Schollmeyer, E.; Dopp, D. Supramol. Chem. 2000, 11, 225-231.
Buschmann, H. J .; Cleve, E.; Mutihac, L.; Schollmeyer, E. Microchem.
J . 2000, 64, 99-103. Tuncel, D.; Steinke, J . H. G. Chem. Commun.
1999, 1509-1510; Tuncel, D.; Steinke, J . H. G. Chem. Commun. 2001,
253-254.
(15) Kim, J .; J ung, I.-S.; Kim, S.-Y.; Lee, E.; Kang, J .-K.; Sakamoto,
S.; Yamaguchi, K.; Kim, K. J . Am. Chem. Soc. 2000, 122, 540-541.
(16) Day, A. I.; Arnold, A. P.; Blanch, R. J . Method for Synthesis
Cucurbiturils. PCT Intl. Appl. PCT/AU00/00412, 2000.
(17) (a) Day, A.; Arnold, A. P.; Blanch, R. J .; Snushall, B. J . Org.
Chem. 2001, 66, 8094-8100. (b) Day, A. I.; Blanch, R. J .; Arnold, A.
P.; Lorenzo, S.; Lewis, G. R.; Dance, I. Angew. Chem., Int. Ed. 2002,
41, 275-277.
(18) Lorenzo, S.; Day, A.; Craig, D.; Blanch, R.; Arnold, A.; Dance,
I. CrystEngComm 2001, 49, 1-7.
(19) Ong, W.; Go´mez-Kaifer, M.; Kaifer, A. E. Org. Lett. 2002, 4,
1791-1794.
(20) Kim, H.-J .; J eon, W. S.; Ko, Y. H.; Kim, K. Proc. Natl. Acad.
Sci. U.S.A. 2002, 99, 5007-5011.
(8) Isobe, H.; Tomita, N.; Lee, J . W.; Kim, H.-J .; Kim, K.; Nakamura,
E. Angew. Chem., Int. Ed. 2000, 39, 4257-4260.
(9) Mock, W. L.; Pierpont, J . J . Chem. Soc., Chem. Commun. 1990,
1509-1511. J un, S. I.; Lee, J . W.; Sakamoto, S.; Yamaguchi, K.; Kim,
K. Tetrahedron Lett. 2000, 41, 471-475.
(21) Kim, S.-Y.; J ung, I.-S.; Lee, E.; Kim, J .; Sakamoto, S.; Yamagu-
chi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 2119-2121.
(22) Blanch, R. J .; Sleeman, A. J .; White, T. J .; Arnold, A. P.; Day,
A. I. Nano Lett. 2002, 2, 147-149.
(10) Buschmann, H. J .; Gardberg, A.; Schollmeyer, E. Textilvered-
lung 1991, 26, 153-7. Buschmann, H. J .; Schollmeyer, E. J . Inclusion
Phenom. Mol. Recognit. Chem. 1992, 14, 91-9; Buschmann, H. J .;
Schollmeyer, E. J . Inclusion Phenom. Mol. Recognit. Chem. 1997, 29,
167-174; Buschmann, H. J .; Schollmeyer, E. Textilveredlung 1998,
33, 44-47. Karcher, S.; Kornmuller, A.; J ekel, M. Water Sci. Technol.
1999, 40, 425-433; Karcher, S.; Kornmueller, A.; J ekel, M. Acta
Hydrochim. Hydrobiol. 1999, 27, 38-42; Kornmuller, A.; Karcher, S.;
J ekel, M. Water Res. 2001, 35, 3317-3324. Taketsuji, K.; Tomioka, H.
Nippon Kagaku Kaishi 1998, 670-678.
(23) J on, S. Y.; Ko, Y. H.; Park, S. H.; Kim, H.-J .; Kim, K. Chem.
Commun. 2001, 1938-1939.
(24) Kim, H.-J .; Heo, J .; J eon, W. S.; Lee, E.; Kim, J .; Sakamoto, S.;
Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 1526-1529.
(25) Zhao, J .; Kim, H.-J .; Oh, J .; Kim, S.-Y.; Lee, J . W.; Sakamoto,
S.; Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 4233-
4235.
(26) Isobe, H.; Sato, S.; Nakamura, E. Org. Lett. 2002, 4, 1287-
1289.
(27) Isaacs, L.; Witt, D.; Lagona, J . Org. Lett. 2001, 3, 3221-3224.
(28) Witt, D.; Lagona, J .; Damkaci, F.; Fettinger, J . C.; Isaacs, L.
Org. Lett. 2000, 2, 755-758.
(11) Marquez, C.; Nau, W. M. Angew. Chem., Int. Ed. 2001, 40,
4387-4390.
5818 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
SCHEME 1. Retr osyn th etic An a lysis of CB[n ]
the fundamental structural unit of cucurbituril (Scheme
1). To minimize the synthetic challenges posed by the
presence of four ureidyl NH groups in 2C and 2S, we
further restrict the present study to the preparation of
derivatives of 2C and 2S comprising a single set of
methylene bridges and bearing two o-xylylene groups (3C
and 3S). We use the suffixes C and S throughout this
paper to distinguish between these two diastereomers
because their three-dimensional structures resemble
those letters (Scheme 1). A successful synthesis of
congeners of cucurbituril requires control over the rela-
tive stereochemistry of each pair of methylene-bridged
glycoluril dimers. For example, consideration of the pair
of diastereomers 3C and 3S reveals that only the C-
shaped diastereomer 3C is capable of being transformed
into a derivative of cucurbituril, since the S-shaped
diastereomer 3S possesses the wrong relative stereo-
chemistry. An important objective of the present work,
therefore, is the development of methods that allow the
diastereoselective formation of 3C. Our retrosynthetic
analysis of 3C and 3S leads to ureidyl NH compound 4
and cyclic ether 5. We envisioned that condensation
reactions between 4 and 5 would proceed under acidic
conditions where the nucleophilic tautomer of 4 could
react with the iminium ion generated after protonation
of 5. To investigate the scope and limitations of the
dimerization reaction used to prepare methylene-bridged
glycoluril dimers, we needed to prepare derivatives of 4
and 5 bearing a range of solubilizing substituents on their
convex faces and on their aromatic rings.
of glycoluril derivatives typically encountered (alkyl,
carboxylic acid derivative, aromatic, and heteroaromatic),
and (3) they provided good solubility characteristics. The
majority of glycoluril derivatives used in this paper,
however, are derived from 1a , which possesses two ethyl
ester groups on its convex face. This choice is based on
our interest in preparing water soluble methylene-
bridged glycoluril dimers27,34 and the fact that glycoluril
derivatives bearing ethoxycarbonyl groups dimerize in
high yield.
Syn th esis of 1,2-Bis(h a lom eth yl)a r om a tic Com -
p ou n d s. In their pioneering studies of molecular recog-
nition, self-assembly, and catalysis, the groups of Nolte35,36
and Rebek37,38 have devised many practical synthetic
methods for the preparation of derivatives of glycoluril.
An important step in many of these syntheses involves
the nucleophilic addition of glycoluril anions to 1,2-bis-
(halomethyl)aromatics to generate glycoluril derivatives
bearing o-xylylene rings on one or both sides of the
glycoluril skeleton.39 Chart 2 shows the structures of 10
alkylating agents (6-15) that we have used in our
synthetic studies. Of these 10 alkylating agents, 6 and
14 were commercially available, 7,40 8,41,42 10,43 11,44 and
1545 were prepared by literature procedures, and 9, 12,
(34) Isaacs, L.; Witt, D. Angew. Chem., Int. Ed. 2002, 41, 1905-
1907.
(35) Rowan, A. E.; Elemans, J . A. A. W.; Nolte, R. J . M. Acc. Chem.
Res. 1999, 32, 995-1006.
(36) Sijbesma, R. P.; Nolte, R. J . M. Top. Curr. Chem. 1995, 175,
25-56.
(37) Rebek, J ., J r. Chem. Soc. Rev. 1996, 25, 255-264.
(38) Rebek, J ., J r. Acc. Chem. Res. 1999, 32, 278-286.
(39) (a) Valde´z, C.; Spitz, U. P.; Toledo, L. M.; Kubik, S. W.; Rebek,
J . J . J . Am. Chem. Soc. 1995, 117, 12733-12745. (b) Wyler, R.; de
Mendoza, J .; Rebek, J ., J r. Angew. Chem., Int. Ed. Engl. 1993, 32,
1699-1701. (c) O’Leary, B. M.; Szabo, T.; Svenstrup, N.; Schalley, C.
A.; Luetzen, A.; Schaefer, M.; Rebek, J ., J r. J . Am. Chem. Soc. 2001,
123, 11519-11533. (d) Reek, J . N. H.; Kros, A.; Nolte, R. J . M. Chem.
Commun. 1996, 245-247. (e) Elemans, J . A. A. W.; de Gelder, R.;
Rowan, A. E.; Nolte, R. J . M. Chem. Commun. 1998, 1553-1554. (f)
J ansen, R. J .; Rowan, A. E.; de Gelder, R.; Scheeren, H. W.; Nolte, R.
J . M. Chem. Commun. 1998, 121-122.
Syn th esis of Glycolu r il Bu ild in g Block s. Com-
pounds 1a-d were prepared by literature procedures.29-33
We chose these four building blocks because (1) they were
easily prepared, (2) they broadly represented the range
(29) Branda, N.; Grotzfeld, R. M.; Valde´z, C.; Rebek, J . J . J . Am.
Chem. Soc. 1995, 117, 7.
(30) Reek, J . N. H.; Kros, A.; Nolte, R. J . M. Chem. Commun. 1996,
245-247.
(31) Butler, A. R.; Leitch, E. J . Chem. Soc., Perkin Trans. 2 1980,
103-105.
(40) Diederich, F.; J onas, U.; Gramlich, V.; Herrmann, A.; Ringsdorf,
H.; Thilgen, C. Helv. Chim. Acta 1993, 76, 2445-2453.
(41) Ardecky, R. J .; Kerdesky, F. A. J .; Cava, M. P. J . Org. Chem.
1981, 46, 1483-1485.
(42) Kang, J .; Hilmersson, G.; Santamar´ıa, J .; Rebek, J . J . J . Am.
Chem. Soc. 1998, 120, 3650-3656.
(43) Shahak, I.; Bergmann, E. D. J . Chem. Soc. C 1966, 1005-1009.
(32) Gompper, R.; No¨th, H.; Rattay, W.; Schwarzensteiner, M.-L.;
Spes, P.; Wagner, H.-U. Angew. Chem., Int. Ed. Engl. 1987, 26, 1039-
1041.
(33) Kutepov, D. F.; Potashnik, A. A.; Khokhlov, D. N.; Tuzhilkina,
V. A. J . Gen. Chem. USSR 1959, 29, 840-842.
J . Org. Chem, Vol. 67, No. 16, 2002 5819

Wu et al.
CHART 2
their aromatic rings, we turned to the problem of creating
a series of potentially electrophilic glycoluril derivatives
(5a -f and 25-28). For this purpose, we turned to the
work of Nolte, who has developed a methodology utilizing
glycoluril-derived cyclic ethers, chloromethyl groups,
acetoxymethyl groups, and hydroxymethyl groups for the
generation of iminium ions that undergo efficient elec-
trophilic aromatic substitution reactions.48-52 For our
purposes, the more stable cyclic ethers (Chart 4) were
preferable. The Nolte cyclic ether synthesis53 calls for
sequential treatment with NaOH and formaldehyde in
aq DMSO, followed by reflux in HCl at pH 1. We
anticipated that these basic and aqueous acidic conditions
might pose problems with substrates bearing ethoxycar-
bonyl groups. We, therefore, developed a one-step pro-
cedure that proceeds under anhydrous acidic conditions
(TFA, reflux) using paraformaldehyde (Table 2). These
reactions proceed in moderate to good yield and offer an
alternative to Nolte’s procedure when working with
compounds containing potentially sensitive functional
groups. The lowest yield (20%) was obtained for 2-pyridyl-
substituted glycoluril 5c. This result is not surprising,
since the pyridyl ring is protonated in TFA, which
probably raises the energy of the intermediates leading
to 5c, resulting in a reduced reaction rate or side
reactions.
Syn th esis of Meth ylen e-Br id ged Glycolu r il Dim -
er s. After preparing a series of glycoluril derivatives
bearing potentially nucleophilic ureidyl NH groups (4a -d
and 16-24) and potentially electrophilic cyclic ether
groups (5a -f and 25-28), we turned our attention to-
ward their condensation reactions that lead to methylene-
bridged glycoluril dimers. Chart 5 gives a summary of
the compounds (29C-44C and 29S-44S) that are dis-
cussed in this paper. There are three synthetic methods
that lead from the two sets of building blocks to meth-
ylene-bridged glycoluril dimers: (1) the reaction of 2 equiv
of 4 with a source of formaldehyde, (2) the condensation
of 4 with cyclic ether 5, and (3) the reaction of 2 equiv of
5 with the formal extrusion of formaldehyde. In each
case, we propose that the reaction proceeds through a
common set of intermediates (Scheme 2), although many
subtle variations are possible and we do not have evi-
dence to exclude those possibilities in this discussion.
Compound 4 can tautomerize into nucleophile (()-45,
which after reaction with formaldehyde, proton transfer,
loss of water, and tautomerization leads to the racemic
mixture (()-47. Similarly, protonation of cyclic ether 5
followed by extrusion of formaldehyde also leads to race-
mic mixture (()-47. At this stage, two different scenarios
are possible: 47 can react with a molecule of like handed-
ness (45) or it can react with a molecule of opposite
handedness (ent-45). Reaction between 47 and 45 gener-
and 13 were prepared by bromination of the correspond-
ing o-xylylene derivatives with N-bromosuccinimide in
CCl4.46,47
Syn th esis of Glycolu r il Der iva tives Bea r in g Tw o
Ur eid yl NH Gr ou p s a n d a Sin gle o-Xylylen e Gr ou p .
For the preparation of glycoluril derivatives bearing a
single substituted o-xylylene sidewall, we adapted chem-
istry developed by the groups of Rebek and Nolte.39
Treatment of glycoluril derivatives 1a -d with t-BuOK
in DMSO results in nucleophilic species that react with
bis(halomethyl)aromatics 6-15 yielding glycoluril de-
rivatives 4a -d and 16-24 (Chart 3 and Table 1) in low
to moderate yields. Entries 1-4 (Table 1) illustrate the
effect of the four different solubilizing groups (CO2Et, Ph,
2-pyridyl, and (CH2)4) on the alkylation reaction with a
single alkylating agent (6). The nature of the solubilizing
group significantly affects the yield of the alkylation
reaction, and in our hands, the alkylation of 1a proceeds
most smoothly, since the anion generated by treatment
with t-BuOK is nicely soluble in DMSO and shows a
lower tendency to form gels which lower yields signifi-
cantly. Entries 1 and 5-13 illustrate the effect of the
alkylating agent on the alkylation reaction; compounds
16-24 have been arranged from electron rich to electron
poor. The electronic nature of the substituents on the
aromatic ring does not have a discernible effect on the
efficiency of the alkylation reaction, although we have
noticed that extended reaction times lead to decreased
yields in the case of the more electron deficient alkylating
agents. Compounds (()-18, (()-20, and (()-21 are chiral
because of the unsymmetrical arrangements of functional
groups on their aromatic rings; these compounds are
synthesized and used in this paper as the racemic
mixture.
Syn th esis of Cyclic Eth er s by Acid -Ca ta lyzed
Con d en sa tion w ith P a r a for m a ld eh yd e. Having se-
cured a range of potentially nucleophilic glycoluril de-
rivatives (4a -d and 16-24) bearing a range of solubi-
lizing groups on their convex faces and substituents on
(48) Smeets, J . W. H.; Sijbesma, R. P.; Niele, F. G. M.; Spek, A. L.;
Smeets, W. J . J .; Nolte, R. J . M. J . Am. Chem. Soc. 1987, 109, 928-
929.
(49) Niele, F. G. M.; Zwikker, J . W.; Nolte, R. J . M. Tetrahedron
Lett. 1986, 27, 243-246.
(50) Smeets, J . W. H.; Sijbesma, R. P.; van Dalen, L.; Spek, A. L.;
Smeets, W. J . J .; Nolte, R. J . M. J . Org. Chem. 1989, 54, 3710-3717.
(51) Sijbesma, R. P.; Nolte, R. J . M. Recl. Trav. Chim. Pays-Bas
1993, 112, 643-647.
(52) Reek, J . N. H.; Elemans, J . A. A. W.; Nolte, R. J . M. J . Org.
Chem. 1997, 62, 2234-2243.
(53) Niele, F. G. M.; Nolte, R. J . M. J . Am. Chem. Soc. 1988, 110,
172-177.
(44) Goldberg, Y.; Bensimon, C.; Howard, A. J . Org. Chem. 1992,
57, 6374-6376.
(45) Coe, P. L.; Croll, B. T.; Patrick, C. R. Tetrahedron 1967, 23,
505-508.
(46) Kno¨lker, H. J .; Bauermeister, M.; Pannek, J . B. Tetrahedron
1993, 49, 841-862.
(47) Lai, Y.-H.; Yap, A. H.-T. J . Chem. Soc., Perkin Trans. 2 1993,
1373-1377.
5820 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
CHART 3
TABLE 1. Syn th esis of 4b-d a n d 16-24
CHART 4
entry
R
alkylating agent
product
yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
CO2Et
Ph
6
6
6
6
7
4a
4b
4c
4d
16
17
18
19
20
21
22
23
24
84a
62
43
31
45
47
22
68
62
32
37
26
45
2-pyridyl
(CH2)4
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
8
9
10
11
12
13
14
15
a
Ref 39a.
TABLE 2. Syn th esis of Cyclic Eth er s u n d er An h yd r ou s
Acid ic Con d ition s
ates intermediate 48S, which is transformed into 49S
and ultimately into the S-shaped product 3S. Conversely,
reaction between 47 and ent-45 generates intermediate
48C, which leads to the C-shaped methylene-bridged
glycoluril dimer via intermediate 49C. Since a highly dia-
stereoselective reaction is not expected between inter-
mediates 45 and 47, one would expect to isolate a mixture
of the S-shaped and C-shaped methylene-bridged gly-
coluril dimers if the reaction is run under kinetically con-
trolled conditions. If, however, the reaction is run under
thermodynamically controlled conditions, the ratio of the
two products will be dictated solely by the relative free
energies of the S- and C-shaped diastereomers.
Hom odim er ization Reaction s of Ur eidyl NH Com -
p ou n d s. The most straightforward synthesis of methyl-
ene-bridged glycoluril dimers involves the condensation
reaction between 2 equiv of 4 and 2 equiv of formalde-
hyde. Our standard procedure (Scheme 3) involves heat-
ing the reactants at reflux in 1,2-dichloroethane contain-
ing p-toluenesulfonic acid as acid catalyst under anaddition
funnel filled with molecular sieves for at least 1 day.
Table 3 summarizes 12 homodimerization reactions that
we have performed. In all cases, we observe a moderate
entry
R
starting material
product
yield (%)
1
2
3
4
5
6
7
CO2Et
2-pyridyl
(CH2)4
CO2Et
CO2Et
CO2Et
CO2Et
4a
4c
4d
16
17
19
20
5a
5c
5d
25
27
26
28
44
20
63
52
56
55
34
to large preference for the formation of the C-shaped
diastereomers. Such a preference, if general, would
explain the high yields obtained in the synthesis of
CB[n].
Entries 1-4 (Table 3) illustrate the pronounced influ-
ence of the solubilizing groups on the convex face of the
glycoluril ring on the dimerization reaction. For example,
J . Org. Chem, Vol. 67, No. 16, 2002 5821

Wu et al.
CHART 5. Ch em ica l Str u ctu r es of Meth ylen e-Br id ged Glycolu r il Dim er s
the dimerization reaction with ethoxycarbonyl-substitut-
ed glycoluril 4a (entry 1) furnishes only the C-shaped
diastereomer 29C, whereas cyclohexyl- and phenyl-
substituted glycolurils 4d and b (entries 4 and 2) proceed
in lower yield and with the formation of side products
35 and 36. For 4c, with 2-pyridyl solubilizing groups, we
did not detect either 31C or 31S by 1H NMR. These
changes in yield and product distribution cannot be
explained by steric differences. Below, we present a
mechanistic rationale for these changes based on the
electronic nature of the solubilizing groups.
The remaining entries in Table 3 focus on the chem-
istry of glycoluril derivatives bearing ethoxycarbonyl
solubilizing groups, since those substrates result in
efficient dimerization reactions. Entries 5 and 7-12
illustrate that the dimerization reaction tolerates a
variety of different substituents on their aromatic rings
(OMe, Br, NO2, F, and heteroaromatics). The nature and
location of substituents can, however, significantly influ-
ence the rate and yield of the reaction. For example,
simply changing the location of two methoxy groups
(compare entries 5 and 6) resulted in decomposition
rather than dimerization, and the presence of a quinoxa-
line ring (entry 11) lowers the yield and greatly reduces
the reaction rate presumably because of the protonation
of the quinoxaline ring N-atom.
As mentioned above, compounds (()-20 and (()-21 are
chiral and racemic because of the unsymmetrical ar-
rangements of substituents on their o-xylylene rings. In
the dimerization reaction of (()-21 (Table 3, entries 9),
two C-shaped and two S-shaped diastereomers were
isolated and characterized. We denote these compounds
as 41CC (C-shaped, OMe groups cis), (()-41CT (C-
shaped, OMe groups trans), (()-41SC (S-shaped, OMe
groups cis), and 41ST (S-shaped, OMe groups trans).
Compounds 41CC and 41ST result from the dimerization
of two molecules of opposite handedness and are achiral
meso compounds, whereas the racemic mixtures (()-
5822 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
SCHEME 2. Ou tlin e of Th r ee P a th w a ys to Meth ylen e-Br id ged Glycolu r il Dim er sa
a
Mechanistic steps: (a) tautomerization, (b) nucleophilic addition (+CH2O) and proton transfer, (c) loss of water and tautomerization,
(d) proton transfer.
SCHEME 3. Dim er iza tion of Glycolu r il Der iva tives w ith Ur eid yl NH Com p ou n d s
TABLE 3. Dim er iza tion of Glycolu r il Der iva tives
Bea r in g Ur eid yl NH Gr ou p s
the C-shaped diastereomers 40CC and (()-40CT in 94%
combined yield.
starting C-shaped
material (yield, %)
S-shaped
(yield, %)
side product(s)
(yield, %)
Sep a r a tion , Id en tifica tion , a n d X-r a y Cr ysta l-
logr a p h ic Ch a r a cter iza tion of th e C-Sh a p ed a n d
S-Sh a p ed Dia ster eom er s. Gratifyingly, the separation
of the crude reaction mixtures described in Table 3 was
possible using simple silica gel chromatography. The
conformationally rigid C-shaped diastereomers have
higher dipole moments, lower Rf values, and lower
solubilities in common organic solvents than the corre-
sponding S-shaped diastereomers (µ ) 0 D by symmetry
arguments), which facilitate their purification. Spectro-
scopic identification of the C-shaped and S-shaped dia-
stereomers is based on a combination of 1H and 13C NMR
spectroscopy and symmetry arguments. Consider, for
example, the C-shaped and S-shaped diastereomers 38C
and 38S (Table 3, entry 7). The C-shaped diastereomer
38C is C2v-symmetric, whereas 38S has time-averaged
C2h-symmetry. These symmetry differences manifest
themselves in the number of resonances expected for the
newly formed methylene bridges; for 38C we expect and
observe a pair of doublets for the diastereotopic methyl-
ene protons (6.02 and 4.58 ppm), whereas for 38S we
expect and observe a singlet (5.00 ppm) for the chemically
equivalent methylene protons. These symmetry consid-
erations are sufficient to allow complete spectroscopic
identification of methylene-bridged glycoluril dimers
prepared from two achiral building blocks. A peculiar but
entry
R
1
2
3
4
5
6
7
8
CO2Et
Ph
4a
4b
4c
4d
16
17
19
20
29C (88) 29S (nd)
30C (19) 30S (nd) 35b (26), 36b (2)
2-pyridyl
(CH2)4
CO2Et
CO2Et
CO2Et
CO2Et
31C (nd)a 31S (nd)
32C (57) 32S (nd) 35d (9), 36d (5)
37C (87) 37S (nd)
39C (nd) 39S (nd)
38C (75) 38S (6)
40CC (46) 40SC (nd)
40CT (48) 40ST (nd)
9
CO2Et
21
41CC (28) 41SC (21)
41CT (24) 41ST (22)
10 CO2Et
11 CO2Et
12 CO2Et
22
23
24
42C (47) 42S (10)
43C (35) 43S (18)
44C (44) 44S (3)
a
nd ) not detected.
41CT and (()-41SC result from the dimerization of 21
of like handedness. In this example, a nearly statistical
distribution of the four stereoisomers was obtained. In
contrast, the dimerization of (()-20 yielded exclusively
1
particularly diagnostic feature of the H NMR spectra of
J . Org. Chem, Vol. 67, No. 16, 2002 5823

Wu et al.
all of the S-shaped diastereomers that we have prepared
to date is the significant upfield shift observed for only
one of the two chemically nonequivalent CH2 groups of
the CO2CH2CH3 solubilizing groups. For example, for 38S
the two methylene groups resonate at 4.19 and 3.58 ppm,
whereas for 38C both resonate at 4.17 ppm. The X-ray
crystal structure of 38S (Figure 1) provides an explana-
tion for this observation. In each of two rapidly equili-
brating S-shaped conformations of 38S, one of the
methylene groups of the internal CO2Et groups is in the
shielding region of the aromatic ring of the opposing
sidewall, leading to the observed upfield shift. A similarly
diagnostic feature of the S-shaped versus the C-shaped
diastereomers is the 13C NMR chemical shifts of the
central methylene bridges. For 38C these carbon atoms
resonate at 47.8 ppm, whereas for 38S they resonate at
51.8 ppm. In general, the C-shaped diastereomers reso-
nate at ≈47-48 ppm, and the S-shaped diastereomers
resonate at 51-52 ppm.17 These three criteria and
symmetry arguments allow complete structural assign-
ments of even the most complicated Cs-, C2-, and C1-
symmetric methylene-bridged glycoluril dimers (ex-
amples: 41CC, (()-41CT, (()-41SC, and 41ST).
The structural assignments of the C-shaped and S-
shaped diastereomers based on 1H and 13C NMR and
symmetry arguments described above have been further
corroborated by X-ray crystallography of many of our
compounds. Figure 1 shows the X-ray structures deter-
mined for 30C, 38C, and 38S. Compounds 30S and 38C
assume C-shaped conformations with their o-xylylene
rings roughly parallel. All four solubilizing groups (Ph
and CO2Et) are displayed on one face of the molecule,
resulting in an amphiphilic topology. Compound 38C
crystallized as the CH3CN solvate; one of the solvating
CH3CN molecules fills its cleft with the CH3 group
oriented toward the glycoluril rings. The distances be-
tween the centers of the o-xylylene rings of 30C and 38C,
defined as the C1-C2-C3-C4-C4A-C18A and C9A-
C10-C11-C12-C13-C13A centroids, are 7.366 and
7.588 Å, respectively. Because of the slight tapering of
the cleft, the distances between the tips of the o-xylylene
rings, defined as the distance between the centroids of
the C2-C3 and C11-C12 bonds, are 6.951 and 7.258 Å
for 30C and 38C, respectively. The mean planes of the
aromatic rings of 30C and 38C intersect each other with
angles of 21.8° and 17.1°, respectively. There is a slight
overall end-to-end twist of the C-shaped molecules of 30C
(-4.0°) and 38C (-3.3°), as measured by the dihedral
angle through the centroids of the C2-C3, C4A-C18A,
C9A-C13A, and C11-C12 bonds. The substituents on
the convex face of 30C and 38C are nearly eclipsed; the
C5B-C5A-C6A-C6B and C7B-C7A-C8A-C8B dihe-
dral angles measure 9.6° and 3.4° (30C) and -1.6° and
1.4° (38C), respectively. The substituents at C6A and
C7A on the convex face of the molecules are nearly
collinear; the C6B-C6A-C7A and C7B-C7A-C6A angles
amount to 95.0° and 94.7° (30C) and 96.5° and 95.0°
(38C), respectively. The separations between these sub-
stituents, as measured by the C6B-C7B distance, are
3.985 Å (30C) and 3.984 Å (38C), indicating that they
are not in van der Waals contact. The N6-C7-N7 and
N16-C16-N15 bond angles of the methylene bridges
amount to 116.2° and 118.3° (30C) and 114.8° and 115.0°
(38C). These values are larger than the tetrahedral bond
F IGURE 1. X-ray crystal structures of 30C, 38C, and 38S.
The solvating CHCl3 and toluene molecules have been removed
from the structure of 30C.
angle, as are those observed for Me10CB[5] (114.0°-
115.4°)13 and CB[5] (113.2°-114.7°), CB[6] (112.9°-
5824 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
SCHEME 4. Dim er iza tion Rea ction s of Glycolu r il-Der ived Cyclic Eth er s
TABLE 4. Dim er iza tion Rea ction s of Glycolu r il-Der ived Cyclic Eth er s
entry
R
starting material
C-shaped (yield, %)
S-shaped (yield, %)
side product(s) (yield, %)
1
2
3
4
5
6
7
8
9
CO2Et
Ph
2-pyridyl
(CH2)4
CONH(CH2)3NMe2
CO2Li
CO2Et
5a
5b
5c
5d
5e
5f
25
26
27
29C
30C
31C
32C
33C
34C
37C
38C
39C
(92)
(nd)a
(nd)
(34)
(85)
(85)
(93)
(87)
(nd)
29S
30S
31S
32S
33S
34S
37S
38S
39S
(nd)
(nd)
(nd)
(nd)
(nd)
(nd)
(nd)
(3)
35b (42)
35d (16), 36d (12)
CO2Et
CO2Et
(nd)
a
nd ) not detected.
115.0°), CB[7] (112.7°-114.5°), and CB[8] (113.0°-
113.6°).1,15 The ureidyl N atoms involved in the central
eight-membered ring do not show significant deviations
from planarity; the sum of the three bond angles around
N6A, N7A, N15A, and N16A of 30C amount to 358.3°,
359.9°, 359.5°, and 359.8°, whereas those around N6, N7,
N15, and N16 of 38C amount to 359.0°, 359.9, 359.0°,
and 360°, respectively. The distances between the car-
bonyl oxygens (O6-O17 and O8-O15) of a single gly-
coluril ring are 5.611 and 5.653 Å (30C) and 5.755 and
5.785 (38C), distances that are slightly larger than those
observed for related molecules containing a single gly-
coluril ring48-50 but smaller than those observed for CB[5]
(6.176-6.217 Å), CB[6] (5.98-6.042 Å), CB[7] (5.913-
6.114 Å), and CB[8] (6.041-6.171 Å).1,15 The distances
between oxygen atoms of adjacent glycoluril rings (O6-
O8 and O15-O17) are 3.405 and 3.309 Å (30C) and 3.424
and 3.389 Å (38C), respectively. These distances are
larger than those observed in the crystal structures of
Me10CB[5] (average, 3.177 Å; range, 3.141-3.218 Å)13
and CB[5] (average, 3.310 Å; range, 3.184-3.602 Å),
comparable to those of CB[6] (average, 3.4025 Å; range,
3.138-3.624 Å), and shorter than those of CB[7] (aver-
age, 3.627 Å; range, 3.405-3.859 Å) and CB[8] (average,
3.810 Å; range, 3.695-3.906 Å).1,15
angles around ureidyl nitrogen atoms N6, N7, N15, and
N16 of 38S amount to 350.1°, 357.9°, 359.9°, and 346.3°.
N atoms N6 and N16 are decidedly nonplanar, suggesting
the presence of strain relative to C-shaped diastereomers
30C (average, 359.4°; range, 358.3°-359.9°) and 38C
(average, 359.5°; range, 359.0°-360°) and cucurbiturils
Me10CB[5] (average, 359.7°; range, 359.5°-360), CB[5]
(average, 358.7°; range, 357.3°-359.9°), CB[6] (average,
358.8°; range, 356.0°-360°), CB[7] (average, 358.1°;
range, 354.9°-359.9°), and CB[8] (average, 357.3°; range,
355.3°-358.2°).1,13,15 Other structural features are com-
parable between the C- and S-shaped diastereomers. For
example, the N6-C7-N7 and N15-C16-N16 bond
angles measure 110.9° and 113.9°, values only slightly
smaller than those observed for 30C and 38C. The
substituents on the convex face of the glycoluril rings are
once again nearly eclipsed with C5C-C5B-C6B-C6C
and C7C-C7B-C8B-C8C dihedral angles of -5.0° and
-5.5°.
Hom od im er iza tion Rea ction s of Cyclic Eth er s.
The mechanistic rationale proposed in Scheme 2 suggests
that cyclic ethers should also participate in this dimer-
ization reaction (Scheme 4). Table 4 shows the results of
the dimerization reactions from cyclic ethers that we have
performed to date. Entries 1-6 illustrate the influence
of the solubilizing groups on the convex face of the
glycoluril skeleton on the dimerization reaction. Sub-
strates 5a , e, and f (entries 1, 5, and 6) that bear electron
withdrawing carboxylic acid derivatives on their convex
face are efficient substrates yielding only the C-shaped
diastereomers in high yield. As in the case of the
dimerization from the ureidyl NH compounds (Table 3),
the substrates bearing Ph, fused-cyclohexyl, and 2-pyridyl
substituents are poor substrates for the reaction (Table
4, entries 2-4), and both 5b and 5d lead to side products
(()-35 and (()-36. Compounds bearing functionalized
o-xylylene rings are also acceptable substrates for this
reaction (Table 4, entries 1, 7, and 8). Interestingly, we
could not detect either 39C or 39S in the dimerization
reaction with 27. Similar behavior was observed in the
dimerization reaction of 17 (Table 3, entry 6).
In constrast to 30C and 38C, diastereomer 38S crys-
tallizes in an S-shaped conformation that displays two
ethoxycarbonyl groups on each face of the molecule. One
of the most interesting features of the crystal structure
of 38S is the close proximity of methylene carbon atom
C6D to the centroid of the aromatic ring defined by C9A-
C10-C11-C12-C13-C13A (3.713 Å). One of the protons
attached to C6D is a mere 2.951 Å from the centroid of
this aromatic ring. The close proximity of this proton to
the center of the aromatic ring places it in its shielding
region, which provides an explanation for the observation
of the significant upfield shifts observed for these protons
1
in the H NMR of the S-shaped diastereomers. The most
notable structural effect of the relative stereochemistry
of the S-shaped diastereomers is present in the central
eight-membered ring. For example, the sums of the bond
J . Org. Chem, Vol. 67, No. 16, 2002 5825

Wu et al.
SCHEME 5. Dim er iza tion Rea ction s fr om Ur eid yl NH a n d Cyclic Eth er Com p ou n d s
TABLE 5. Dim er iza tion Rea ction s fr om Ur eid yl NH a n d
Cyclic Eth er Com p ou n d s
The survey of the substrates that participate effectively
in this reaction is not as extensive as that described in
Table 3 because the cyclic ethers themselves must be
derived from the corresponding compounds containing
ureidyl NH groups. Additionally, a comparison of the
results obtained by these two methods (Table 3, entries
1-7 versus Table 4, entries 1-4 and 7-9) indicates that
dimerization occurs in comparable yield in most cases.
The single exception is the dimerization reaction of
phenyl glycoluril (5b), a poor substrate for our reaction,
which yields 30C only from 4b (Table 3, entry 2 versus
Table 4, entry 2). These considerations suggest that the
method described in Table 3 is preferable, since the cyclic
ether substrates are themselves derived from the ureidyl
NH compounds in only moderate yield (Chart 4, Table
2).
Het er od im er iza t ion R ea ct ion s of Ur eid yl NH
Com p ou n d s a n d Cyclic Eth er s. The dimerization
reactions described in Tables 3 and 4 offer two routes to
the preparation of methylene-bridged glycoluril dimers.
Of these two methods, the direct dimerization of the
ureidyl NH compounds is preferable. On the basis of the
mechanism of the dimerization reaction proposed in
Scheme 2, we considered the possibility of performing a
selective heterodimerization reaction by the reaction
between 1 equiv of ureidyl NH compound and 1 equiv of
cyclic ether (Scheme 5). The success of this method, the
selective synthesis of a dimer comprising two different
o-xylylene rings, requires a fast reaction between inter-
mediates 45 and 51 (Scheme 2) and that the equilibria
connecting those intermediates (via 46 and 47) that
result in the scrambling of the locations of the methylene
bridges are slow relative to methylene-bridged glycoluril
dimer formation.
Initially, we choose to study reactions between ureidyl
NH compounds and cyclic ethers that would result in
homodimeric species to limit the potential complexity of
the reaction. Table 5 summarizes the results of the
experiments that we performed. The effects of the solu-
bilizing groups on the convex face of the glycoluril on the
dimerization reaction (Table 5, entries 1-4) are similar
to those observed for the direct dimerization of 4a -d
(Table 3) or 5a -d (Table 4). Glycoluril derivatives
bearing electron withdrawing ethoxycarbonyl groups
(Table 5, entry 1) dimerized much more readily than
those bearing phenyl or fused cyclohexyl groups (entries
2 and 4), and those bearing the readily protonated pyridyl
substituents (entry 3) were resistant to dimerization.
Those glycoluril derivatives bearing ethoxycarbonyl solu-
starting C-shaped S-shaped side product(s)
entry
R
material (yield, %) (yield, %)
(yield, %)
1
2
3
4
5
6
7
CO2Et
Ph
4a + 5a 29C (89) 29S (2)
4b + 5b 30C (16) 30S (nd) 35b (70)
2-pyridyl 4c + 5c 31C (nd)a 31S (nd)
(CH2)4
CO2Et
CO2Et
CO2Et
4d + 5d 32C (30) 32S (nd) 35d (12), 36d (10)
16 + 25 37C (91) 37S (2)
17 + 27 39C (56) 39S (nd)
19 + 26 38C (90) 38S (3)
a
nd ) not detected.
bilizing groups that undergo smooth homodimerization
also yield dimers by the heterodimerization route (Table
5, entries 5 and 7; Table 4, entries 7 and 8; Table 3,
entries 5 and 7). There are situations, however, where
the heterodimerization reaction is preferable to either of
the two homodimerization pathways. For example, even
though neither homodimerization pathway allowed the
detection of either 39C or 39S (Table 3, entry 6; Table 4,
entry 9), the heterodimerization pathway (Scheme 5)
allowed the isolation of 39C in good yield (Table 5, entry
6). In those cases, where direct dimerization reactions
fail, the heterodimerization route offers a viable alterna-
tive.
To fully demonstrate the synthetic utility of the het-
erodimerization reaction (Scheme 5), it was necessary to
prepare true heterodimers, methylene-bridged glycoluril
dimers comprising two different o-xylylene rings, and
show that these heterodimers are produced selectively
at the expense of the corresponding homodimers. Table
6 shows the results of three heterodimerization reactions
that we have performed. Entry 1 shows the heterodimer-
ization of dimethoxyxylylene ureidyl NH compound 16
and xylylene cyclic ether 5a . In theory, six dimers might
be formed (homodimers 29C, 29S, 37C, and 37S and
heterodimers 52C and 52S); in practice, we isolate the
two heterodimers and the two C-shaped homodimers. The
desired heterodimers 52C and 52S were obtained in high
combined yield (81%), with a modest level of diastereo-
selectivity favoring the C-shaped diastereomer 52C. This
level of diastereoselectivity was particularly surprising,
considering the fact that the relative stereochemistry of
the product is set during the first covalent bond forming
reaction between the two reaction partners (Scheme 2,
(()-45 + (()-51). In a separate report, we present a
mechanistic rationale for the enhanced yield of the
5826 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
TABLE 6. Heter od im er iza tion Rea ction s
SCHEME 6. P r op osed Mech a n ism for th e
F or m a tion of (()-35 in th e Dim er iza tion Rea ction s
C-shaped heterodimer.54 Table 6 (entries 2 and 3) shows
the reactions of the racemic ureidyl NH compounds (()-
21 and (()-18 with the dimethoxyxylylene cyclic ether
25. In these cases, of the six possible products, we observe
only two, the desired heterodimers, and isolate them in
excellent combined yield. In contrast to entry 1 (Table
6), these two heterodimerization reactions produce the
S-shaped diastereomers with modest diastereoselectivi-
ties. These three experiments and the results of Tables
3-5 that demonstrate a preference for the C-shaped
diastereomers suggest that the formation of the mixture
of C- and S-shaped diastereomers occurs under kinetic
control (Table 6, entries 2 and 3) and that the preference
for the C-shaped heterodimer (Table 6, entry 1) reflects
thermodynamic control.54
Su bstitu en t Effects on th e Mech a n ism of th e
Dim er iza tion Rea ction a n d Im p lica tion s for CB[n ]
Syn th esis. To date, CB[n] and fully substituted deriva-
tives have been synthesized using glycolurils 1d -
f.1,4,13-17,21 These three glycoluril derivatives represent
only one of the four main classes of commonly encoun-
tered glycolurils (alkyl, aromatic, heteroaromatic, and
carboxylic acid derivative). We were surprised by the lack
of success in the synthesis of CB[n] derivatives using
other glycoluril derivatives. The formation of cyclohexyl-
and phenyl-substituted dimers 32C and 30C, in sharp
contrast to the dimerization reactions involving ethoxy-
carbonyl-substituted glycolurils, proceeds in modest yields
and with the formation of side products 35 and 36. These
side products provide clues for the lack of success in the
synthesis of fully substituted CB[n] from 1b. Scheme 6
shows a mechanistic proposal for the formation of side
products (()-35. This mechanistic proposal is illustrated
for the dimerization reactions of cyclic ethers 5, but the
mechanisms of all three types of dimerization reactions
potentially involve common intermediates (Scheme 2).
(54) Chakraborty, A.; Wu, A.; Witt, D.; Lagona, J .; Fettinger, J . C.;
Isaacs, L. J . Am. Chem. Soc. 2002, 124, 8297-8306.
J . Org. Chem, Vol. 67, No. 16, 2002 5827

Wu et al.
Cyclic ether 5 undergoes protonation on its ether oxygen
(55), followed by ring opening to yield iminium ion (()-
56. Formally, the conversion of (()-56 into the observed
products (()-35 occurs by a hydride shift. Formulation
of the reaction as a hydride shift from (()-56 ignores the
fact that only phenyl and cyclohexyl glycolurils generate
these side products, since one would also expect ethoxy-
carbonyl-substituted glycolurils to generate similar side
products. We have never observed aldehydic side prod-
ucts in the dimerization of ethoxycarbonyl-substituted
glycoluril derivatives. Alternatively, one can postulate the
formation of intermediates (()-57. One would expect that
intermediates (()-57 would be favored when R,R )
(CH2)4 and R ) Ph because of the ability of these groups
to stabilize adjacent carbocations but disfavored when R
) CO2Et, since this group would destabilize an adjacent
positive charge. The conversion of (()-57 into (()-35 is
then formulated as an ene reaction of a cationic N-acyl
iminium ion with an imine.55 The lower yield obtained
in the synthesis of 30C than in that of 32C can be
explained by the fact that the phenyl substituents are
better able to stabilize the adjacent positive charge in
(()-57, thereby leading to the enhanced yield of (()-35b
compared to (()-35d . These observations have implica-
tions for the synthesis of derivatives of CB[n]. They
suggest that the CB[n] synthesis is likely to be most
successful in the case of glycoluril derivatives bearing
electron withdrawing substituents on their convex face
and least successful in the case of electron donating
substituents that are able to stabilize adjacent positive
charges. The recent report of the synthesis of Cy6CB[6]
and Cy5CB[5] in 2% and 16% yields, respectively, sug-
gests that alkyl groups may be borderline substituents
for the synthesis of CB[n].21 However, if it is desired to
use a glycoluril derivative containing electron donating
groups on its convex face (e.g. 1b or d ) for CB[n]
synthesis, then it would be prudent to perform such a
reaction using heterodimerization conditions (Tables 5
and 6),16,26 involving the reaction of 1b or d with, for
example, 58 which is not prone to form aldehydic side
products.
cyclic ethers 5a -f and 25-27 (Table 4), and the het-
erodimerization reactions (Tables 5 and 6) of cyclic ethers
and ureidyl NH compounds, all deliver the C- and
S-shaped diastereomers in good to excellent yields with
glycoluril derivatives bearing electron withdrawing car-
boxylic acid derivatives on their convex face. The C-
shaped compound is usually formed preferentially in
diastereoselective reactions. Of these three synthetic
methods, we prefer the direct dimerization reaction of
the ureidyl NH compounds (Scheme 3, Table 3), since it
produces the C- and S-shaped dimers in similar yields
and diastereoselectivities to those of the dimerization of
cyclic ethers (Scheme 4, Table 4) but involves fewer
synthetic steps. The heterodimerization reactions (Tables
5 and 6) are most useful when it is necessary to access
methylene-bridged glycoluril dimers bearing differen-
tially substituted rings or when substrates undergo low
yielding homodimerization reactions. Glycoluril deriva-
tives bearing phenyl and fused cyclohexyl groups are poor
substrates for the dimerization reactions because they
are able to stabilize adjacent positive charges leading to
aldehydic side products. The development of synthetic
methods for the synthesis of methylene-bridged glycoluril
dimers offers the opportunity to study the fundamental
steps in CB[n] synthesis54 and the potential to expand
the range of CB[n] homologues and derivatives.
Exp er im en ta l Section
Gen er a l. Starting materials were purchased from com-
mercial suppliers and were used without further purification.
Compounds 5a , b, e, and f, 25, 27, 29C, 33C, 34C, 37C, 37S,
39C, 52C, 52S,28 29S,54 as well as 12, (()-21, (()-53S, and
(()-53C34 were prepared according to literature procedures.
THF and toluene were distilled from sodium benzophenone
ketyl, and methylene chloride was distilled from CaH2 im-
mediately before use. TLC analyses were performed using
precoated glass plates from Analtech or E. Merck. Column
chromatography was performed using silica gel (230-400
mesh, 0.040-0.063 µm) from E. Merck using eluents in the
indicated v/v ratio. Melting points were measured on
a
Meltemp apparatus in open capillary tubes and are uncor-
rected. IR spectra were recorded on a spectrophotometer as
KBr pellets or thin films on NaCl plates and are reported in
inverse centimeters. NMR spectra were measured at 400 MHz
for 1H and 100 MHz for 13C. Fast atom bombardment (FAB)
mass spectra were obtained using the indicated matrix. The
matrix “magic bullet” is a 5:1 (w/w) mixture of dithiothreitol-
dithioerythritol. Elemental analyses were performed by Mid-
west MicroLab (Indianapolis, IN).
Con clu sion s
The outstanding molecular recognition properties of
cucurbituril, its homologues CB[n], and its derivatives
(Me10CB[5], Cy5CB[5], Cy6CB[6], and Ph2CB[6]) have
prompted several groups to broaden the scope and define
the limitations of cucurbituril synthesis. We have taken
a multistep synthetic organic approach based on the
identification of methylene-bridged glycoluril dimers 2C
and 2S as the fundamental building blocks in CB[n]
synthesis. We examined condensation reactions between
glycoluril derivatives bearing one o-xylylene wall and
either free ureidyl NH groups (4a -d and 16-24) or cyclic
ether (5a -f and 25-27) groups. Three different methods,
the condensation reactions of 4a -d and 16-24 with
paraformaldehyde (Table 3), the homodimerization of
Rep r esen ta tive Exp er im en ta l P r oced u r e fr om Ta ble
1 (19). Glycoluril 1a (8.00 g, 28.0 mmol) was dissolved in anhyd
DMSO (100 mL) under N2, and t-BuOK (5.91 g, 52.7 mmol)
was added. After stirring for 15 min, 1,2-bis(chloromethyl)-
4,5-dimethylbenzene (1.26 g, 6.20 mmol) was added in one
portion, and stirring was continued for 3 h. The reaction
mixture was poured into 0.1 N HCl (1 L) and extracted with
EtOAc (3 × 400 mL). The extracts were washed with brine (2
× 300 mL) and dried over anhyd MgSO4. After filtration and
rotary evaporation, the residue was purified by flash chroma-
tography (SiO2, CHCl3/MeOH 25:1) to give 19 (1.76 g, 4.22
mmol, 68%) as a white solid. Mp 236 °C. TLC (CHCl3/MeOH,
25:1) Rf 0.23. IR (KBr, cm-1): 3217 s, 3019 m, 2940 m, 1710 s,
1
(55) Borzilleri, R. M.; Weinreb, S. M. Synthesis 1995, 347-360. The
conversion of (()-57d into (()-35d could also conceivably deliver the
compound with the trans-fused glycoluril derivative, whereas this
transformation would be precluded by the cyclohexyl ring of (()-35c.
In either case, the comparatively high energy trans-fused ring system
has never, to the best of our knowledge, been observed in glycoluril
derivatives.
1464 m, 1368 m, 1270 m, 1145 m, 1034 s. H NMR (400 MHz,
DMSO-d6): 8.38 (s, 2H), 6.99 (s, 2H), 4.48 (d, J ) 15.8, 2H),
4.32 (d, J ) 15.8, 2H), 4.19 (q, J ) 7.1, 2H), 4.09 (q, J ) 7.1,
2H), 2.12 (s, 6H), 1.19 (t, J ) 7.1, 3H), 1.16 (t, J ) 7.1, 3H).
13C NMR (100 MHz, DMSO-d6): 167.0, 166.6, 157.5, 135.6,
135.0, 130.9, 82.9, 74.4, 63.1, 62.9, 43.9, 19.2, 14.2, 14.2. MS
5828 J . Org. Chem., Vol. 67, No. 16, 2002

Methylene-Bridged Glycoluril Dimers
(FAB, magic bullet): m/z 417 (100, [M + H]+). HRMS (FAB,
magic bullet): m/z 417.1774 ([M + H]+, C20H25N4O6, calcd
417.1774). Anal. Calcd for C20H24N4O6 (416.17): C, 57.68; H,
5.81. Found: C, 57.66; H, 5.78.
CF3COOH, D2O-cap.): 7.40-7.35 (br m, 8H), 6.00 (d, J ) 16.6,
2H), 4.92 (d, J ) 16.2, 4H), 4.73 (d, J ) 16.2, 4H), 4.67 (d, J )
16.6, 2H), 2.60-2.50 (m, 8H), 2.00-1.85 (br m, 8H). 13C NMR
(100 MHz, CF3COOH, D2O-cap.): 160.8, 136.4, 131.5, 131.1,
82.4, 79.9, 46.3, 46.1, 26.0, 25.6, 16.5, 16.0. MS (FAB, magic
bullet): m/z 621 (100, [M + H]+). HRMS (FAB, magic bullet):
m/z 621.2968 ([M + H]+, C34H37N8O4, calcd 621.2938). Com-
pound (()-35d : mp 223-225 °C. TLC (CHCl3/MeOH 50:1) Rf
0.37. IR (KBr, cm-1): 2948 w, 2875 w, 1738 s, 1706 s, 1457 s,
Rep r esen ta tive P r oced u r e fr om Ta ble 2 (5d ). A mixture
of 4d (435 mg, 1.46 mmol) and paraformaldehyde (438 mg,
14.6 mmol) in TFA (5 mL) was stirred and heated at reflux
for 20 h. After rotary evaporation, the residue was dissolved
in EtOAc (150 mL), washed with saturated aq Na2CO3, dried
over anhyd MgSO4, and concentrated. The residue was purified
by flash chromatography (SiO2, CHCl3/MeOH 50:1) to give 5d
(314 mg, 0.924 mmol, 63%) as a white solid. Mp 245-246 °C.
TLC (CHCl3/MeOH 50:1) Rf 0.33. IR (KBr, cm-1): 2949 m, 2911
m, 2876 m, 1707 s, 1472 s, 1446 s, 1239 s, 1005 s, 740 s. 1H
NMR (400 MHz, CDCl3): 7.40-7.35 (m, 2H), 7.25-7.20 (m,
2H), 5.30 (d, J ) 11.3, 2H), 4.67 (d, J ) 15.8, 2H), 4.54 (d, J )
11.3, 2H), 4.35 (d, J ) 15.8, 2H), 2.19 (br m, 2H), 2.08 (br m,
2H), 1.66 (br m, 4H). 13C NMR (100 MHz, CDCl3): 157.4, 136.9,
129.8, 128.1, 71.8, 70.9, 44.0, 25.3, 24.3, 14.9, 14.8. (12
resonances expected, 11 observed). MS (FAB, magic bullet):
m/z 341 (100, [M + H]+). HRMS (FAB, magic bullet): m/z
341.1626 ([M + H]+, C18H21N4O3, calcd 341.1614).
1
1417 m, 1306 m, 758 m. H NMR (400 MHz, CDCl3): 8.93 (s,
1H), 7.40-7.30 (m, 2H), 7.30-7.20 (m, 2H), 4.79 (d, J ) 15.6,
1H), 4.72 (d, J ) 15.8, 1H), 4.41 (d, J ) 15.6, 1H), 4.30 (d, J )
15.8, 1H), 2.86 (s, 3H), 2.50-2.20 (m, 3H), 2.10-2.00 (m, 1H),
1.70-1.50 (br m, 4H). 13C NMR (100 MHz, CDCl3): 160.8,
156.8, 152.9, 136.8, 135.6, 130.0, 129.3, 128.5, 128.2, 76.8, 76.2,
43.6, 43.4, 27.4, 24.7, 23.9, 14.9, 14.5. MS (FAB, magic
bullet): m/z 341 (60, [M + H]+), 55 (100). HRMS (FAB, magic
bullet): m/z 341.1601([M + H]+, C18H21N4O3, calcd 341.1614).
Compound (()-36d : mp 314 °C dec. TLC (CHCl3/MeOH 25:1)
Rf 0.27. IR (KBr, cm-1): 3219 m, 2948 m, 1718 s, 1697 s, 1483
1
s, 1416 m, 764 m. H NMR (400 MHz, CDCl3): 7.35-7.10 (m,
4H), 6.06 (br s, 1H), 4.66 (d, J ) 15.7, 1H), 4.64 (d, J ) 15.7,
1H), 4.30 (d, J ) 15.7, 1H), 4.28 (d, J ) 15.7, 1H), 2.59 (s,
3H), 2.25-2.15 (m, 1H), 2.10-1.70 (m, 3H), 1.70-1.50 (m, 4H).
13C NMR (100 MHz, CDCl3): 158.2, 157.4, 137.3, 136.9, 129.8,
129.2, 127.9, 127.8, 77.5, 74.1, 43.6, 43.1, 27.9, 26.0, 24.5, 16.5,
16.2. MS (FAB, magic bullet): m/z 313 (100, [M + H]+). HRMS
(FAB, magic bullet): m/z 313.1660 ([M + H]+, C17H21N4O2,
calcd 313.1664).
Rep r esen ta tive P r oced u r e for Ta ble 3 (38C a n d 38S).
A mixture of PTSA (0.168 g, 0.884 mmol) and ClCH2CH2Cl
(10 mL) was heated under N2 at reflux for 30 min under an
addition funnel filled with molecular sieves (4 Å). Compound
19 (92.0 mg, 0.221 mmol) and paraformaldehyde (20.0 mg,
0.663 mmol) were added, and reflux was continued for 48 h.
The reaction mixture was diluted with EtOAc (150 mL),
washed with saturated Na2CO3, dried over anhyd MgSO4, and
concentrated. Flash chromatography (SiO2, CHCl3/CH3CN 20:
1) gave 38C (67.0 mg, 0.0782 mmol, 75%) and 38S (5.2 mg,
0.0061 mmol, 6%) as white solids. Compound 38C: mp > 300
°C. TLC (CHCl3/CH3CN 15:1) Rf 0.22. IR (KBr, cm-1): 2965
w, 1747 s, 1456 m, 1256 m, 1021 m. 1H NMR (400 MHz,
CDCl3): 6.99 (s, 4H), 6.02 (d, J ) 16.0, 2H), 4.79 (d, J ) 16.0,
4H), 4.58 (d, J ) 16.0, 2H), 4.32 (d, J ) 16.0, 4H), 4.17 (m,
8H), 2.12 (s, 12H), 1.29 (t, J ) 7.2, 6H), 1.24 (t, J ) 7.2, 6H).
13C NMR (100 MHz, CDCl3): 165.7, 165.0, 154.6, 136.4, 133.5,
131.1, 80.1, 78.7, 63.7, 63.3, 47.8, 44.9, 19.2, 13.9, 13.9. MS
Rep r esen ta tive P r oced u r e for Ta ble 5 (30C a n d (()-
35b). A mixture of PTSA (5.000 g, 26.3 mmol) in ClCH2CH2Cl
(25 mL) was heated at reflux for 30 min under an addition
funnel filled with molecular sieves (4 Å). Compounds 4b (1.042
g, 2.63 mmol) and 5b (922.7 mg, 2.11 mmol) were added, and
reflux was continued for 5 days. The reaction mixture was
diluted with EtOAc (100 mL), washed with saturated aq Na2-
CO3, dried over anhyd MgSO4, and concentrated. The residue
was purified by flash chromatography (SiO2, CHCl3/MeOH 50:
1) to yield 30C (282.7 mg, 0.35 mmol, 16%) and (()-35b (642
mg, 1.47 mmol, 70%) as white solids. Compound 30C: mp 384
°C dec. TLC (CHCl3/hexanes/EtOAc/MeOH 25:10:2:1) Rf 0.16.
IR (KBr, cm-1): 3062 w, 3034 w, 2962 w, 1734 s, 1450 s, 1426
(FAB, magic bullet): m/z 857 (27, [M + H]+), 174 (100, [C11H12
NO]+). HRMS (FAB, magic bullet): m/z 857.3440 ([M + H]+,
42H49N8O12, calcd 857.3470). X-ray crystal structure. Anal.
-
1
m, 1286 m, 753 m, 697 m. H NMR (400 MHz, CDCl3): 7.40-
C
7.30 (m, 8H), 7.10-6.95 (m, 6H), 6.95-6.80 (m, 10H), 6.55-
6.50 (m, 4H), 5.89 (d, J ) 15.4, 2H), 4.70 (d, J ) 15.5, 4H),
4.03 (d, J ) 15.5, 4H), 3.76 (d, J ) 15.4, 2H). 13C NMR (100
MHz, CDCl3): 156.5, 136.4, 133.4, 132.0, 129.7, 128.8, 128.6,
128.4, 127.8, 127.6, 127.6, 86.1, 84.0, 47.7, 45.0 (16 resonances
expected, 15 observed). MS (FAB, magic bullet): m/z 817 (30,
[M + H]+), 91 (100, C7H7+). HRMS (FAB, magic bullet): m/z
817.3226 ([M + H]+, C50H41N8O4, calcd 817.3251). Compound
(()-35b: mp 310-312 °C dec. TLC (hexanes/EtOAc 4:1) Rf
0.23. IR (KBr, cm-1): 3061 w, 3024 w, 2925 w, 1746 s, 1711 s,
1461 m, 1450 m, 1303 m, 1285 m. 1H NMR (400 MHz,
CDCl3): 9.19 (s, 1H), 7.40-6.90 (m, 13H), 6.75-6.60 (m, 1H),
4.95 (d, J ) 15.6, 1H), 4.86 (d, J ) 15.6, 1H), 4.29 (d, J ) 15.6,
1H), 4.18 (d, J ) 15.6, 1H), 2.93 (s, 3H). 13C NMR (100 MHz,
CDCl3): 160.3, 157.8, 154.2, 136.6, 135.3, 132.6, 131.4, 130.0,
129.4, 129.3, 129.0, 128.8, 128.7, 128.6, 128.5, 128.2, 127.6,
127.5, 126.6, 85.9, 84.3, 45.3, 45.2, 29.5. MS (FAB, magic
bullet): m/z 439 (72, [M + H]+), 91 (100, C7H7+). HRMS (FAB,
magic bullet): m/z 439.1812 ([M + H]+, C26H23N4O3, calcd
439.1770).
Calcd for C42H48N8O12 (856.88): C, 58.87; H, 5.65. Found: C,
58.74; H, 5.60. Compound 38S: mp 297-299 °C. TLC (CHCl3/
CH3CN 20:1) Rf 0.20. IR (KBr, cm-1): 2980 w, 1766 s, 1742 s,
1720 s, 1456 s, 1424 m, 1387 s, 1308 m, 1250 m, 1157 m, 1020
m. 1H NMR (400 MHz, CDCl3): 7.07 (s, 4H), 5.00 (s, 4H), 4.74
(d, J ) 16.0, 4H), 4.26 (d, J ) 16.0, 4H), 4.19 (q, J ) 7.1, 4H),
3.58 (q, J ) 7.1, 4H), 2.16 (s, 12H), 1.25 (t, J ) 7.1, 6H), 1.04
(t, J ) 7.1, 6H). 13C NMR (100 MHz, CDCl3): 165.4, 164.2,
155.3, 136.4, 133.4, 131.1, 81.8, 78.5, 63.6, 51.8, 44.9, 19.1, 13.9,
13.6 (15 resonances expected, 14 observed). MS (FAB, magic
bullet): m/z 857 (30, [M + H]+), 174 (100, [C11H12NO]+). HRMS
(FAB, magic bullet): m/z 857.3490 ([M + H]+, C42H49N8O12
calcd 857.3470). X-ray crystal structure. Anal. Calcd for
42H48N8O12 (856.88): C, 58.87; H, 5.65. Found: C, 58.69; H,
5.58.
Rep r esen ta tive P r oced u r e for Ta ble 4 (32C, (()-35d ,
,
C
a n d (()-36d ). A mixture of PTSA (760 mg, 4.00 mmol) and
ClCH2CH2Cl (5 mL) was heated at reflux for 30 min under an
addition funnel filled with molecular sieves (4 Å). Compound
5d (136 mg, 0.400 mmol) was added in one portion, and reflux
was continued for 20 h. The reaction mixture was diluted with
EtOAc (50 mL), washed with saturated aq Na2CO3, dried over
anhyd MgSO4, and concentrated. The residue was purified by
flash chromatography (SiO2, CHCl3/MeOH 25:1) to yield 32C
(42.5 mg, 0.0685 mmol, 34%), (()-35d (21.2 mg, 0.0624 mmol,
16%), and (()-36d (14.8 mg, 0.0474 mmol, 12%) all as white
solids. Compound 32C: mp 370 °C dec. TLC (CHCl3/MeOH
25:1) Rf 0.09. IR (KBr, cm-1): 2948 w, 2875 w, 1709 s, 1464 s,
R ep r esen t a t ive P r oced u r e for Ta ble 6 ((()-54C a n d
(()-54S). A mixture of PTSA (0.410 g, 2.15 mmol) in ClCH2-
CH2Cl (20 mL) was heated under N2 at reflux for 30 min under
an addition funnel filled with molecular sieves (4 Å). Com-
pound 25 (0.210 g, 0.43 mmol) and (()-21 (0.240 g, 0.43 mmol)
were added, and heating was continued for 24 h. The reaction
mixture was diluted with EtOAc (100 mL), washed with
saturated Na2CO3, dried over anhyd MgSO4, and concentrated.
Flash chromatography (SiO2, CHCl3/ CH3CN 4:1) gave (()-
1
1422 m, 1308 m, 1225 m, 759 m, 737 m. H NMR (400 MHz,
J . Org. Chem, Vol. 67, No. 16, 2002 5829

Wu et al.
54S (0.24 g, 0.23 mmol, 53%) and (()-54C (0.19 g, 0.19 mmol,
43%). Compound (()-54S: mp > 330 °C dec. TLC (CHCl3/CH3-
CN 4:1) Rf 0.40. IR (KBr, cm-1): 2983 w, 2936 w, 2835 w, 1717
X-r a y Cr ysta l Str u ctu r es for 30C, 38C, a n d 38S. De-
tailed descriptions of the data collection, solution, and refine-
ment of the structures can be found in the Supporting
Information. Crystal data for 30C: [C50H40N8O4][CHCl3]-
[C7H8]2 (1120.54); triclinic, space group P1; colorless block, a
) 14.4877(12) Å, b ) 14.8574(12) Å, c ) 15.1182(12) Å; V )
2787.7(4) Å3; Z ) 2; T ) 193(2) K; R(F) ) 0.0654; GOF on F2
) 1.082. Crystal data for 38C: [C42H48N8O12][NCCH3]2 (938.99);
orthorhombic, space group Pna21; colorless block, a ) 10.7933(9)
Å, b ) 26.350(2) Å, c ) 16.7931(14) Å; V ) 4776.1(7) Å3; Z )
4; T ) 193(2) K; R(F) ) 0.0586; GOF on F2 ) 1.065. Crystal
data for 38S: [C42H48N8O12] (856.88); monoclinic, space group
P21/c; colorless block, a ) 20.9952(17) Å, b ) 16.8963(13) Å, c
) 11.9470(9) Å; V ) 4083.8(6) Å3; Z ) 4; T ) 193(2) K; R(F) )
0.0774; GOF on F2 ) 1.087.
1
s, 1452 s, 1390 s, 1270 s. H NMR (400 MHz, CDCl3): 8.00-
7.95 (m, 1H), 7.90-7.80 (m, 1H), 7.80-7.70 (m, 2H), 7.13 (d,
J ) 8.8, 1H), 6.96 (d, J ) 8.8, 1H), 6.82 (s, 1H), 6.79 (s, 1H),
5.75 (d, J ) 16.2, 1H), 5.16 (d, J ) 13.6, 1H), 5.07 (d, J ) 13.6,
1H), 4.97 (d, J ) 13.6, 1H), 4.75-4.60 (m, 4H), 4.34 (d, J )
16.2, 1H), 4.30-4.00 (m, 6H), 3.95-3.85 (m, 3H), 3.89 (s, 3H),
3.82 (s, 3H), 3.80 (s, 3H), 3.60-3.50 (m, 1H), 3.40-3.30 (m,
1H), 1.30-1.00 (m, 9H), 0.96 (t, J ) 7.1, 3H). 13C NMR (100
MHz, CDCl3): 168.0, 167.0, 165.2, 165.1, 164.0, 163.9, 157.4,
155.3, 155.1, 154.8, 147.8, 136.3, 134.5, 134.3, 131.9, 131.3,
129.2, 128.6, 126.2, 124.1, 123.6, 122.5, 113.0, 112.9, 111.8,
81.6, 81.0, 78.8, 78.2, 64.1, 63.6, 63.4, 56.4, 55.9, 51.8, 44.7,
39.6, 36.3, 13.8, 13.6, 13.5, 13.4 (49 resonances expected, 42
observed). MS (FAB, magic bullet): m/z 1036 (100, [M + H]+).
HRMS (FAB, magic bullet): m/z 1036.3284 ([M + H]+,
Ack n ow led gm en t. We thank the National Insti-
tutes of Health (GM61854), the donors of the Petroleum
Research Fund, administered by the American Chemi-
cal Society (33946-G4), and the University of Maryland
for generous financial support. We thank the Howard
Hughes Medical Institute (M.O. and J .C.) and the
Dreyfus Foundation (M.O., J ean Dreyfus Boissevain
Undergraduate Scholarship for Excellence in Chemis-
try) for undergraduate research fellowships. L.I. is a
Cottrell Scholar of Research Corporation.
C
49H50N9O17, calcd 1036.3325). Compound (()-54C: mp > 330
°C dec. TLC (CHCl3/CH3CN 4:1) Rf 0.19. IR (KBr, cm-1): 2975
w, 2940 w, 2839 w, 174e, 1717 s, 1456 s, 1270 s, 1014 m, 909
m. 1H NMR (400 MHz, CDCl3): 7.95-7.90 (m, 1H), 7.85-7.80
(m, 1H), 7.70-7.60 (m, 2H), 7.13 (d, J ) 8.8, 1H), 6.88 (d, J )
8.8, 1H), 6.74 (s, 1H), 6.71 (s, 1H), 6.05 (d, J ) 16.0, 1H), 5.92
(d, J ) 16.0, 1H), 5.62 (d, J ) 16.0, 1H), 4.78 (d, J ) 16.0,
1H), 4.73 (d, J ) 16.5, 1H), 4.72 (d, J ) 16.0, 1H), 4.64 (d, J )
16.0, 1H), 4.48 (d, J ) 16.0, 1H), 4.34 (d, J ) 16.0, 1H), 4.33
(d, J ) 16.0, 1H), 4.25 (d, J ) 16.5, 1H), 4.25-4.00 (m, 9H),
3.80 (s, 3H), 3.79 (s, 3H), 3.78 (s, 3H), 1.30-1.10 (m, 12H). 13
C
NMR (100 MHz, CDCl3): 167.8, 166.5, 165.5, 165.4, 164.7,
164.6, 157.0, 154.5, 154.4, 154.4, 154.2, 147.6, 147.5, 135.7,
134.1, 134.0, 131.5, 131.1, 129.2, 128.3, 128.2, 125.3, 123.9,
123.3, 122.1, 112.7, 112.6, 111.4, 79.9, 79.5, 78.6, 63.5, 63.3,
63.2, 56.1, 55.6, 55.5, 47.7, 47.5, 44.8, 39.8, 36.4, 13.7, 13.7,
13.6, 13.6 (49 resonances expected, 46 observed). MS (FAB,
magic bullet): m/z 1036 (100, [M + H]+). HRMS (FAB, magic
bullet): m/z1036.3359 ([M +H]+, C49H50N9O17, calcd 1036.3325).
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectral data for all new compounds and details
of the X-ray crystal structures of 30C, 38C, and 38S. This
material is available free of charge via the Internet at
http://pubs.acs.org.
J O0258958
5830 J . Org. Chem., Vol. 67, No. 16, 2002