A R T I C L E S
Mo¨ssinger et al.
33% yield).48 The best results for the test compound were
obtained by applying an in situ variant of Plenio’s ready-made
catalyst mixture (69% yield),49 that is, with free phosphine
-
instead of (t-Bu)3PH+BF4 and without (i-Pr)2NH2+Br- as a
catalyst matrix salt. These optimized conditions could be
transferred one-to-one to the current synthesis. Na2[PdCl4], CuI,
and (t-Bu)3P smoothly catalyzed the coupling of 7 and 8 in (i-
Pr)2NH at 80 °C and gave yields around 90%. Treatment of 9
with potassium carbonate selectively cleaved the CPDMS
protective group and produced the final spoke/rim module 10
in close to quantitative yields. In the critical 6-fold Sonogashira
coupling of the hub module 4 and the spoke/rim module 10
toward 11 (Scheme 3), we slightly changed the well-established
conditions from the previous synthesis.23 Although still 10 mol
% of palladium(II) catalyst was employed per reactive site (0.6
equiv in total), we further increased the overall concentration
of the reaction partners to approximately 0.15 M (in the case
of 10), omitted THF as solvent, raised the reaction temperature
to 80 °C, and stopped the reaction after one night. 11 could be
completely separated from small amounts of the 4- and 5-fold
coupling products by column chromatography and was obtained
in yields from 73% to 78%.50 If the reaction was conducted
with significantly lower catalyst load (0.1 equiv in total), 11a
could still be isolated in 57% yield, but the separation from
incomplete coupling side products became more tedious.51
Initially, the desilylation of 11 toward 12 also caused some
unexpected trouble. To completely remove all CPDMS groups,
we added a 1.0 M solution of TBAF in THF at room temperature
and stirred overnight. However, this harsh treatment produced
12 along with oligomeric impurities, which could be removed
by preparative gel permeation chromatography (GPC) only.
When a more dilute solution of TBAF (0.25 M in THF) was
added at 0 °C and the reaction was stopped after 3 h, pure 12
could be obtained, even though in suboptimal yields (86-92%).
According to the covalent-template principle, 12 bears 12
adjacent terminal acetylenes and supports its 6-fold intramo-
lecular oxidative coupling. Intuitively, we applied the previ-
ously reported copper(I)-catalyzed cyclization conditions on
12a.23 A solution of the crude product was filtered over a
silica gel column to remove the copper catalyst. GPC analysis
revealed a mixture of 85% of the desired spoke wheel 3a
(11 900 g mol-1) and 15% of the dimer and higher oligomers
(Figure 1a). However, purification by preparative recycling
GPC separation and precipitation with methanol afforded only
small amounts of the pure spoked wheel (yields around 20%).
Apart from substance losses during the purification procedure,
this poor recovery rate must be due to partial polymerization
of 12a during the cyclization reaction and retention of the
polymer and higher oligomers by the silica gel column.
However, under pseudo high-dilution conditions, intermo-
lecular side reactions can only occur if the intramolecular
reaction proceeds slowly. We assumed that pyridine is not
basic enough to deprotonate the terminal acetylenes of
electron-rich 12a. Consequently, the overall yield and the
cycle/oligomer ratio could be improved in consecutive
reactions by the addition of 1,8-diazabicyclo[5.4.0]undec-7-
ene (DBU) as strong auxiliary base (Figure 1b) and even
Figure 1. GPC analyses (in THF vs PS) of the crude cyclization products
of 3a (a-c) and 3b (d,e). Conditions were (a) CuCl, CuCl2, pyridine, room
temperature, 4 d; (b) CuCl, CuCl2, DBU (2%, v/v), pyridine, room
temperature, 4 d; (c) CuCl, CuCl2, DBU (2%, v/v), pyridine, 50 °C, 4 d;
(d) CuCl, CuCl2, DBU (2%, v/v), pyridine, 50 °C, 4 d; (e) [PdCl2(PPh3)2],
CuI, I2, THF, HN(i-Pr)2, 50 °C, 3 d; (f) pure 3b after recycling GPC
purification of the above crude products. See the Supporting Information
for GPC analyses of pure 3a, 12a, and 12b.
more when the latter reaction was performed at 50 °C (Figure
1c). Under optimized reaction conditions, subsequent column
chromatography, and precipitation with methanol, 3a was
obtained in 75% yield without detectable oligomer contami-
nation and did not require further purification steps such as
preparative GPC.
Oddly, the optimized conditions for the cyclization of 12a
could not be transferred to the hexadecyl precursor 12b, but
produced an oligomeric mixture that contained less than 55%
of 3b (16 200 g mol-1, Figure 1d). As a consequence, we
completely discarded our established set of reagents and
switched to palladium-catalyzed cyclizations in diisopropyl-
amine/THF with iodine as oxidant.52 This approach brought
considerable improvements toward a wheel/oligomer ratio of
about 17:3 (Figure 1e) when a catalyst system consisting of
[PdCl2(PPh3)2] and CuI was applied.53 Still, after column
chromatography, the product was not obtained as pure as 3a
and thus required recycling GPC separation. Subsequent pre-
cipitation afforded pure 3b in 64% yield (Figure 1f; for details,
see the Supporting Information). Despite the recycling GPC
separation during the last step, a 50 mg amount of substance
was easily obtained in a simple run of the complete synthesis
over seven steps (see TOC graphic). Still, all C-C coupling
reactions took place in miniature flasks (1-5 mL) aside from
the cyclization step, which was performed under pseudo high-
dilution conditions (20 mL reaction volume).
As for spoked wheel 2, the target compounds appeared as
slightly yellow solids, which can be easily handled in solution
and stored under air exposure. The large number of alkyl chains
afforded a better solubility than for 2. 3a’s hexyl chains lead to
(48) Bo¨hm, V. P. W.; Herrmann, W. A. Eur. J. Org. Chem. 2000, 3679–
3681.
(52) Marsden, J. A.; Miller, J. J.; Haley, M. M. Angew. Chem., Int. Ed.
2004, 43, 1694–1697.
(49) Ko¨llhofer, A.; Plenio, H. AdV. Synth. Catal. 2005, 347, 1295–1300.
(50) Note that 78% in a 6-fold reaction corresponds to 96% in each of the
six coupling reactions per molecule.
(53) [PdCl2(dppe)] was only tried once. The HT 1H NMR spectrum of the
GPC-separated main fraction showed a minor singlet peak at δ ) 3.34
ppm, possibly a hint at an incomplete terminal acetylene conversion
during the cyclization reaction.
(51) Those conditions were not applied for the synthesis of 11b.
9
1414 J. AM. CHEM. SOC. VOL. 132, NO. 4, 2010