Importance of molecular-level contacts under solventless conditions for chemical reactions and self-assembly

Article abstract of DOI:10.1246/bcsj.80.1617

It has been shown that solventless reactions are accelerated when the reactants are mixed at molecular level. Accordingly, a solid film obtained by evaporating a solution of reactants mixture underwent facile reaction either upon grinding or upon simply standing. The effectiveness of these methods is evident from comparison with conventional solventless and solution reactions. It was concluded consequently that, once the molecular-level contacts are formed between reactants, the reaction occurs more efficiently under solventless conditions than in solution. Due to this pre-mixing method, efficient synthesis of [2]rotaxanes and rate acceleration of the Wittig reaction were achieved.

Full text of DOI:10.1246/bcsj.80.1617

Ó 2007 The Chemical Society of Japan
Bull. Chem. Soc. Jpn. Vol. 80, No. 8, 1617–1623 (2007) 1617
Importance of Molecular-Level Contacts under Solventless Conditions
for Chemical Reactions and Self-Assembly
ꢀ
Akihiro Orita, Junji Okano, Genta Uehara, Lasheng Jiang, and Junzo Otera
Department of Applied Chemistry, Okayama University of Science, Ridai-cho, Okayama 700-0005
Received January 22, 2007; E-mail: otera@high.ous.ac.jp
It has been shown that solventless reactions are accelerated when the reactants are mixed at molecular level. Ac-
cordingly, a solid film obtained by evaporating a solution of reactants mixture underwent facile reaction either upon
grinding or upon simply standing. The effectiveness of these methods is evident from comparison with conventional sol-
ventless and solution reactions. It was concluded consequently that, once the molecular-level contacts are formed be-
tween reactants, the reaction occurs more efficiently under solventless conditions than in solution. Due to this pre-mixing
method, efficient synthesis of [2]rotaxanes and rate acceleration of the Wittig reaction were achieved.
The solventless reactions have received increasing attention
and show promise for green chemical processes.1 However,
more significantly, from viewpoints of organic synthesis and
process chemistry, high efficiency and selectivity are frequent-
ly attained under solventless conditions. Many organic reac-
tions proceed faster under solventless conditions than in solu-
tion, and we have reported a dramatic rate enhancement for su-
pramolecular self-assembly as well.2 For example, a platinum
cage molecule, which originally had been prepared in H2O by
heating the reactants for 4 weeks at 100 ꢁC,3 can be obtained in
10 min at ambient temperature by simply grinding the reaction
components.4 Most reactions between solid reactants demand
mechanical forces, like grinding in a mortar with a pestle1 or
compounding in a ball mill.5 The reaction generally proceeds
via a liquid or melt phase through an intimate mixture of solid
macroscopic particles of the reactants, so that the molecules
inside the solid particles are mobile enough for frequent mo-
lecular collisions to occur.6 In a few cases, powdered, solid re-
actants liquefy simply by coming in contact each other. Once
the melt has been formed, reaction occurs very quickly. It has
been suggested that the high mobility can also be achieved in
co-crystal formation by adding a minor amount of appropriate
solvent (2 drops (ca. 0.05 mL) of organic solvent per ca. 200
mg of the solid mixture).7 On the other hand, it has been re-
ported that formation of voids and cracks are responsible for
the reaction between solids in case of no melt being formed.8
In our project involving rotaxane chemistry,9 we have pub-
lished a preliminary communication on new versions of sol-
ventless reaction.10 Namely, solid reactants are mixed in solu-
tion, and subsequently, the solvent is evaporated. Grinding of
the resulting solid film drives the reaction more readily than
the direct grinding of the solid reaction components, because
the reactant amalgam thus formed contains molecular-level
contacts. More remarkably, the reaction proceeds simply upon
standing the film without any mechanical forces. In these treat-
ments, both threading of the crown-ether beads and N-alkyla-
tion take place without solvent. A similar rate acceleration has
also been observed for carbon–carbon bond formation like the
Wittig reaction.11 Thus, we became interested in gaining more
insight into how the blending of reactants affects a reaction un-
der solventless conditions. In this paper, we report a full ac-
count on assessment of these pre-mixing methods by compar-
ing relevant protocols for N-alkylation reaction of pyridine
derivatives and Wittig reaction. In addition, in the context
of rotaxane chemistry, we tried to elucidate how easily self-
assembly between bead and thread molecules occurs under sol-
ventless conditions, because the mobility of these molecules is
another key factor in the efficient solventless end-capping
method for rotaxane synthesis.12 This issue also will be
discussed in detail.
Results and Discussion
N-Alkylation. In order to assess the effect of mixing, the
following five procedures were compared for three N-alkyla-
tion reactions of pyridine derivatives 1 with bromides 2
(Scheme 1). Procedure A is a typical conventional solventless
reaction, in which solid reactants are directly ground at ambi-
ent temperature in a mortar with a pestle. In Procedure B, the
conventional method is slightly modified by adding a small
amount of an organic solvent to the reactant mixture, accord-
ing to the literature method.7 In Procedure C, both reactants
are dissolved in acetone in a mortar, and then the solvent is
evaporated in vacuo in 5 min period. A film, which has been
cast on the surface of the mortar, is ground. Alternatively, in
Procedure D, the film is merely allowed to stand on the mortar
or more conveniently on an inner surface of a flask without
grinding. The high concentration solution method, Procedure
E, was used as the reference solution protocol, because of
its effectiveness for end-capping synthesis of rotaxanes.13 In
Procedure F, a static pressure (40 kg cmꢂ2) was applied to a
pellet of the solid reactants mixture for 1 h at ambient tempera-
ture to clarify the importance of mechanical forces for inter-
mingling solid reactants. The results of these operations for
the three N-alkylation reactions, depicted in Scheme 1, are
summarized in Table 1. All reactions were carried out using
0.03 mmol of 1 and 0.06 mmol of 2. The yields were deter-
1
mined by measuring H NMR spectra of the reaction mixture
or by isolation as PF6 salts.14 The typical mechanochemical
Published on the web August 10, 2007; doi:10.1246/bcsj.80.1617

1618 Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007)
Importance of Molecular-Level Contacts
Br
O
O
O
NH4PF6
+
O
O
O
+
N
N
O
O
O
+
+N
N
(1)
O
O
O
-
-
2PF6
PF6
1a
2a
3a
NH4PF6
+
O
O
O
N
O
O
O
O
O
O
Br
N
N+
+
+
N
N+
+
N
O
O
O
(2)
-
-
3PF6
2PF6
1b
2b
3b
NH4PF6
Br
Br
Br
+
+
+
N
N+
(3)
N
N
-
-
PF6
2PF6
1c
2c
3c
Scheme 1.
mixture solidified immediately upon grinding probably on ac-
count of rapid vaporization of the acetone, and the reaction
was not complete after 1.5 h. On the other hand, this protocol
was not effective for reaction (3). The reactants mixture did
not converted into a molten state by adding anisole, acetone
or acetophenone, and grinding of this heterogeneous mixture
did not cause the reaction to go to completion, yet a 73% yield
of 3c was obtained. Apparently, the choice of suitable solvent
and reactants is crucial for this method.
Table 1. Solventless N-Alkylation under Various Conditions
Yiled/%a)
3a
40
78(73)
79(78)f)
85(77)
83(79)
0
3b
60
3c
60
73
82(73)h)
81(75)j)
86(77)
0
Procedure Ab)
Procedure Bc)
Procedure Ce)
Procedure Di)
Procedure Ek)
Procedure F
80(72)d)
82(76)g)
80(74)
85(80)
0
By contrast, Procedure C constantly gave good yields of the
N-alkylation products 3 after grinding for 15–50 min. No melt
was formed during the manipulation. With Procedure D, the
reaction proceeded smoothly even without mechanical forces,
although the reaction time was longer than in Procedure C. No
change was observed again in the appearance of the film, ex-
cept for a slight deepening of the yellow color. It was con-
a) Determined by NMR; Isolated yileds are given in parenthe-
ses. b) Reaction conditions: for 3a, 3c: 1 (0.03 mmol); 2
(0.06 mmol); grinding for 1.5 h at r.t.: for 3b: 1b (0.06 mmol);
2b (0.03 mmol); grinding for 1.5 h at r.t. c) Reaction condi-
tions: for 3a, 3c: 1 (0.03 mmol); 2 (0.06 mmol); anisole
(0.02 mL); grinding for 30 min at r.t.: for 3b: 1b (0.06 mmol);
2b (0.03 mmol); anisole (0.02 mL); grinding for 30 min at r.t.
d) 40 min. e) Reaction conditions: for 3a, 3c: 1 (0.03 mmol); 2
(0.06 mmol); acetone (0.2 mL); grinding at r.t. after evapora-
tion: for 3b: 1b (0.06 mmol); 2b (0.03 mmol); grinding at r.t.
after evaporation. f) Grinding for 30 min. g) Grinding for
50 min. h) Grinding for 15 min. i) Reaction conditions: for
3a, 3c: 1 (0.03 mmol); 2 (0.06 mmol); acetone (0.2 mL); stand-
ing for 2 h at r.t. after evaporation: for 3b: 1b (0.06 mmol); 2b
(0.03 mmol); acetone (0.2 mL); standing for 2 h at r.t. after
evaporation. j) Standing for 40 min. k) Reaction conditions:
for 3a, 3c: 1 (0.03 mmol); 2 (0.06 mmol); acetone (0.2 mL);
stirring for 2 days at r.t.: for 3b: 1b (0.06 mmol); 2b
(0.03 mmol); acetone (0.2 mL); stirring for 2 days at r.t.
1
firmed by monitoring H NMR spectra of reaction (1) that the
reaction actually proceeded in the solid film without grinding
(Fig. 1). In a fresh film which was obtained by mixing 1a
and 2a in solution followed by evaporation (5 min after the
start of mixing), the reaction started to occur to some extent
since 1a was detected in 75% yield.15 The amount of 1a was
67% after 30 min, 43% after 1 h, and 9.2% after 1.5 h. The re-
action was nearly complete after 2 h with 2% of 1a remaining.
Apparently, the reaction occurred without external mechanical
forces, probably because of the close contacts between or fac-
ile movement of the reactants molecules in the solid. To check
if any acetone remained in the film or not, freshly prepared
1
film in reaction (3) was extracted with CDCl3. The H NMR
spectrum of this solution exhibited no signal attributable to
acetone, an implication of no intervention of the acetone
between the reactant molecules in the film.
For comparison, the same reaction was conducted in solu-
tion (Procedure E). This procedure furnished comparative
yields of 3 after stirring an acetone solution (0.2 mL) of 1
(0.06 mmol) and 2 (0.03 mmol) at ambient temperature.16 Un-
fortunately, however, it took 2 days to reach the maximum
yield, and TLC analyses showed that the starting compounds
remained after shorter reaction times.
treatment, i.e., Procedure A, did not go to completion, and the
desired products 3 were obtained only in 40–60% yields. After
being ground for 1.5 h, most of the reactants mixture stayed
powdery, although small yellow pasty spots appeared, on
which presumably a biased pressure was applied locally. The
yield of the alkylation product was increased using Procedure
B. The reaction mixture of 1a/2a or 1b/2b formed a melt upon
addition of 0.02 mL of anisole and became to a yellow paste
by grinding for 5 min. Further grinding for 10 min caused the
mixture to solidify, and the reaction was complete in 30–40
min. When acetone was used in place of anisole, the reaction
The importance of the molecular-level mixing was support-
ed by the following experiments. A pellet of the well-shaken

A. Orita et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007) 1619
dure A, grinding may help change relatively intimate solid
reactants to a molten state. The formation of voids and cracks8
is presumably facilitated by mechanical forces for the solid
reactants in case of no molten mixture being formed.
a
b c
d
O
O
O
+
N
N
-
PF6
1a
End-Capping Synthesis of [2]Rotaxanes.
Because of
their high efficiency as found above, Procedures B and D
were applied to the end-capping synthesis of [2]rotaxane 5b
(Scheme 2). First, 2b (0.03 mmol) and crown ether 4 (0.06
mmol) were ground in the presence of anisole (0.02 mL) for
20 min. To this mixture were added 1b (0.06 mmol) and
anisole (0.02 mL). Grinding the mixture for 1.5 h afforded,
after treatment with NF4PF6, the desired rotaxane 5b in 72%
yield together with 3b in 17% yield (Procedure B-1). When
the three components with the same ratio and anisole
(0.06 mL) were mixed simultaneously, 5b and 3b were ob-
tained in 75 and 13% yields, respectively after grinding for
2 h (Procedure B-2). No appreciable difference was observed
between these two procedures, suggesting that the threading
between 2b and 4 is fast and exerts no influence on the ratio
of 5b and 3b. It should be noted that the high concentration
solution method (Procedure E) gave a similar outcome, yet a
much longer reaction time (3 days) was needed. Procedure D
was carried out using a mixture of acetone and CH2Cl2 be-
cause 4 is not soluble in acetone. Thus, 1b (0.06 mmol), 2b
(0.03 mmol), and 4 (0.06 mmol) were dissolved in a mixture
of acetone (0.1 mL)/CH2Cl2 (0.1 mL). The film which had
been cast from this solution was allowed to stand for 4 h to af-
ford 5b in 85% yield together with 3b in 8% yield. The yield
of 5b by using this procedure was higher than those by using
Procedures B and E. More significantly, the ratio of 5b/3b in-
creased, an implication that the difference in relative rate of
the threading reaction over N-alkylation is larger for Proce-
dure D than for the others.
The advantage of the pre-mixing protocol was also con-
firmed in the synthesis of [2]rotaxane 5a (Scheme 3). Howev-
er, even Procedure D resulted in a lower yield of 5a than that
of 5b in the previous case. This is reasonable because 2b in-
nately has a dipyridinium moiety that can interact with 4 to fa-
cilitate the threading prior to the N-alkylation. It is apparent
from these results that the self-assembly of the thread and bead
can take place as easily as or more easily than the N-alkylation
in the solid film.
Wittig Reaction. An evaluation of Procedures A, C, D,
and E, for Wittig reactions of solid reactants is summarized
in Table 2. Obviously, solventless protocols (Procedures A,
C, and D) always gave rise to better outcomes in terms of re-
action rate and yield than the solution reactions (Procedure E).
In particular, Procedure C worked best, except Entry 4, indi-
cating the effectiveness of combination of film casting and
grinding. However, the relative significance between these
two effects is not straightforward. Comparison of Procedures
A and D shows that the former is more effective in Entries 1
and 4, whereas the reverse is true in Entries 2, 3, 5, and 6. No-
tably, no reaction occurred by using Procedure A in Entries 5
and 6, implying that the affinity of the surface of the solid par-
ticles plays a pivotal role. The same reactions did not go to
completion in Procedure C. This is attributed to disruption
of the molecule-to-molecule contacts once formed by grinding
as well as the low reactivity. The importance of the molecular-
α
β
O
O
O
+
N
+
N
O
O
O
-
2PF6
3a
(a)
a
c
d
b
9.5
9.0
8.5
8.0 ppm
(b)
(c)
d
a
c
b
9.5
9.0
8.5
8.0 ppm
c
a
b
b
d
9.5
9.0
8.5
8.0 ppm
(d)
(e)
c
a
d
9.5
9.0
8.5
8.0 ppm
d
a
b
c
8.0 ppm
9.5
9.0
8.5
Fig. 1. Monitoring of reaction between 1a and 2a by the
pre-mixing method D. (a) After 5 min. (b) After 30 min.
(c) After 1 h. (d) After 1.5 h. (e) After 2 h.
powdery mixture composed of 1 and 2 was subjected to a static
pressure (40 kg cmꢂ2) for 1 h (Procedure F). No reaction oc-
curred at all in all cases, indicating that the molecule-to-mole-
cule contacts are crucial for the reaction to occur. Obviously,
the contacts are best attained by mixing in solution followed
by evaporation (Procedures C and D) and next best by adding
a small amount of suitable solvent to a solid reactants mixture
(Procedure B), whereas the mechanical forces are modestly
useful (Procedure A). In Procedure B, anisole, though having
a rather high boiling point, was not detected after grinding
for 10 min, and thus the reaction mostly proceeded under sol-
ventless conditions. Acetone vaporized before the molecular-
level contacts could form sufficiently. Consequently, pre-mix-
ing protocols are highly dependent on the solvent. In Proce-

1620 Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007)
Importance of Molecular-Level Contacts
O
O
O
O
O
O
O
O
O
1b
NH4PF6
O
O
O
+
N
N
Br
+
+
-
O
2PF6
2b
4
O
O
O
O
O
O
O
O
+
N
O
O
O
+
N
O+
NO
3b
+
-
O
O
3PF6
O
5b
Procedure B-1 (2 h)
72%
17%
Procedure B-2 (2 h)
Procedure D (4 h)
75%
85%
78%
13%
8%
Procedure E (3 days)
11%
a For the reaction conditions, see the text.
Scheme 2.
O
O
O
O
O
O
O
O
O
O
Br
NH4PF6
+
+
+
N
N
O
O
O
O
O
O
1a
2a
4
O
O
O
O
O
O
O
O
O
O
O
N
N
O
+
3a
O
O
O
O
-
2PF6
5a
40%
64%
56%
36%
20%
32%
Procedure B
Procedure D
Procedure E
Scheme 3.
level homogeneity is apparent from the solvent effect in Pro-
cedure D. A large amount of CH3CN (70 mL) was employed
as a solvent for reaction between 6a (2.0 mmol) and 7b (2.2
mmol) because of poor solubility of 7b. When the resulting
homogeneous solution was concentrated to dryness, a white
precipitate of 7b appeared first, whereas yellow 6a remained
in solution. Complete removal of the solvent resulted in a light
yellow film with white spots. Standing this film at room tem-
perature for 24 h afforded only a 53% yield of 8b.
In conclusion, the superiority of the molecular-level interac-
tions of reactants to the solid particle-to-particle contacts are
evident from the high efficiency of Procedures B, C, and D,
although it was not clear in case of Procedure B if a homo-
geneous state emerges or if some solid particles remain. In any
case, once the molecular-level contacts are realized, the sol-
ventless conditions are more advantageous than the solution
conditions, because the contacts are preserved more easily
due to lack of thermal movements as in the case of solution
conditions. In addition, the desolvation energy, which is inevi-
table in solution reaction,17 is avoided. A further notable fea-
ture is the facile threading, which not only represents the use-
fulness of the solventless treatment in supramolecular chemis-
try but also implies that molecules move rather facilely in a
solid mixture.
Although the pre-mixing protocols may not be solventless in
a strict sense if the prior dissolution step is counted, the reac-

A. Orita et al.
Table 2. Wittig Reaction under Various Conditions
Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007) 1621
+
Ph3P=CHR'
(2.2 mmol)
RCHO
(2.0 mmol)
RCH=CHR'
6a: 4-Nitrobenzaldehyde
7a:R' = COOEt
7b: R = COPh
8
6b: 2-Naphthalenecarbaldehyde
6c: 4-Chlorobenzaldehyde
6d: 3,5-Dibromobenzaldehyde
6e: 3,5-Dichlorobenzaldehyde
Yield of 8/%a)
Procedure Cb)
Entry
6
7
8
Procedure A
Procedure Dc)
Procedure Ed)
1
2
3
4
5
6
6a
6a
6b
6c
6d
6e
7a
7b
7a
7a
7b
7b
8a
8b
8c
8d
8e
8f
95 (30 min)
65e) (1 h)
94 (50 min)
91 (5 min)
nrf)
92 (5 min)
92 (20 min)
94 (5 min)
92 (5 min)
52e) (1 h)
93 (1 h)
92 (1 h)
93 (15 min)
93 (10 min)
84 (1.5 h)
87e) (24 h)
92e) (24 h)
93 (3 h)
93 (5.5 h)
91e) (24 h)
86 (3.5 h)
85 (8 h)
nrf)
80e) (1 h)
a) Isolated yileds after column chromatography. Reaction times are given in parentheses. b) 6: 2 mmol; 7: 2.2 mmol;
CH2Cl2: 4 mL; evaporation by rotary evaporator followed by pumping in vacuo (5 min); grinding. The time for
grinding after evaporation is given in parentheses. c) 6: 2 mmol; 7: 2.2 mmol; CH2Cl2: 4 mL; evaporation by rotary
evaporator followed by pumping in vacuo (5 min); standing. The time for standing after evaporation is given in pa-
rentheses. Time for standing after evaporation in parentheses. d) 6: 2 mmol; 7: 2.2 mmol; CH2Cl2: 1 mL; stirring at
27 ꢁC. e) The aldehyde remained, but no further reaction occurred even upon prolonged reaction. f) No reaction.
131.4, 131.2, 130.7, 128.6, 128.4, 128.3, 128.2, 126.7, 125.2,
125.1, 116.3, 115.3, 114.3, 114.1, 71.6, 70.5, 69.5, 68.6, 64.3,
34.9, 31.6; MS (MALDI-TOF) found m=z 1788.86 ½M ꢂ PF6ꢃþ,
calcd for C104H112F12N3O6P2 m=z 1788.78. 3c: 1H NMR (300
MHz, CD3COCD3): ꢀ 9.60 (d, J ¼ 6:9 Hz, 2H), 9.53 (d, J ¼ 6:9
Hz, 2H), 8.86 (d, J ¼ 6:9 Hz, 4H), 8.25 (s, 1H), 8.06–7.95 (m,
3H), 7.74–7.61 (m, 7H), 6.36 (s, 2H), 6.19 (s, 2H); 13C NMR
(75.5 MHz, CD3COCD3): ꢀ 151.4, 146.9, 134.2, 133.5, 133.4,
132.3, 131.4, 130.5, 130.4, 129.2, 128.7, 128.6, 128.5, 128.3,
127.9, 126.7, 124.7, 65.9, 64.9; MS (MALDI-TOF) found m=z
611.62 ½M ꢂ PF6ꢃþ, calcd for C28H23BrF6N2P m=z 611.07.
Preparation of 3a (Procedure B, Representative). In a mor-
tar, 1a (27.8 mg, 0.03 mmol), 2a (42.3 mg, 0.06 mmol), and an-
isole (0.02 mL) were placed, and the mixture was ground with
a pestle for 30 min at r.t. The mixture was dissolved in acetone
(5 mL), CHCl3 (2 mL), and water (1 mL). To the resulting solution
was added NH4PF6 (97.8 mg, 0.6 mmol). The organic layer was
separated and evaporated. 1H NMR spectrum of this crude mixture
indicated formation of 3a in 78% yield. Column chromatography
of this mixture on silica gel (70:16:11:3 CH3OH/CH2Cl2/CH3-
NO2/NH4Cl (2 mol dmꢂ3) furnished 3a (37.2 mg, 73%).
Preparation of 3a (Procedure C, Representative). In a mor-
tar, 1a (27.8 mg, 0.03 mmol) and 2a (42.3 mg, 0.06 mmol) were
dissolved in acetone (0.2 mL). The acetone solution became yel-
low film after 5 min of standing in open air, followed by evapo-
ration in vacuo. The film was ground with a pestle for 30 min
at r.t. The mixture was dissolved in acetone (5 mL), CHCl3 (2
mL), and water (1 mL). To the resulting solution was added NH4-
PF6 (97.8 mg, 0.6 mmol). The organic layer was separated and
evaporated. 1H NMR spectrum of this crude mixture indicated for-
mation of 3a in 79% yield. Column chromatography of this mix-
ture on silica gel (70:16:11:3 CH3OH/CH2Cl2/CH3NO2/NH4Cl
(2 mol dmꢂ3) gave 3a (39.7 mg, 78%).
tion itself proceeds under solventless conditions. In a practical
sense, however, reactants in solution can be charged into a re-
actor much more conveniently than a solid reactants mass. We
showed the usefulness of the new protocol via simple N-alkyl-
ation and the Wittig reaction in addition to self-assembly. Fur-
ther applications to other reactions are now in progress in our
laboratories.
Experimental
General Remarks. All reactions were carried out under an
atmosphere of nitrogen with freshly distilled solvents, unless oth-
erwise noted. All solvents were distilled from CaH2. Silica gel
(Daiso gel IR-60) was used for column chromatography. NMR
spectra were recorded at 25 ꢁC on JEOL Lambda 300 and JEOL
Lambda 500 instruments and calibrated with tetramethylsilane
(TMS) as an internal reference. Pelletized mixture of starting
compounds was compressed by using Riken Power P-18.
Starting compounds 2c was commercially available, and crown
ether 4 was prepared according to the reported procedure.18
Pyridine derivatives 1a–1c and benzyl bromide derivatives 2a
and 2b were prepared by conventional methods.19 Experimental
details are described in the Supporting Information.
Preparation of 3a (Procedure A, Representative). In a mor-
tar, 1a (27.8 mg, 0.03 mmol) and 2a (42.3 mg, 0.06 mmol) were
placed, and the mixture was ground with a pestle for 1.5 h at r.t.
The mixture was dissolved in acetone (5 mL), CHCl3 (2 mL),
and water (1 mL). To the resulting solution was added NH4PF6
(97.8 mg, 0.6 mmol). The organic layer was separated and evapo-
rated. Column chromatography of this mixture on silica gel
(70:16:11:3 CH3OH/CH2Cl2/CH3NO2/NH4Cl (2 mol dmꢂ3) fur-
nished 3a (20.4 mg, 40%).19 Compounds 3b and 3c were obtained
similarly. 3b: 1H NMR (300 MHz, CD3COCD3): ꢀ 9.46 (d, J ¼
5:1 Hz, 4H), 8.93 (d, J ¼ 7:2 Hz, 2H), 8.80–8.76 (m, 4H), 7.78–
7.62 (m, 8H), 7.34–7.06 (m, 32H), 6.87–6.81 (m, 4H), 6.20 (s,
2H), 6.10 (s, 2H), 5.88 (s, 2H), 4.63 (m, 2H), 4.20–4.13 (m, 6H),
4.00 (m, 2H), 3.88 (m, 6H), 1.30 (s, 36H); 13C NMR (75.5 MHz,
CD3COCD3): ꢀ 149.3, 148.3, 147.0, 146.0, 132.8, 132.1, 131.7,
Preparation of 3a (Procedure D, Representative). In a mor-
tar, 1a (27.8 mg, 0.03 mmol) and 2a (42.3 mg, 0.06 mmol) were
dissolved in acetone (0.2 mL). The acetone solution became a yel-
low film after 5 min of standing in open air, followed by evapora-
tion in vacuo. The film was kept standing in open air for 2 h at r.t.

1622 Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007)
Importance of Molecular-Level Contacts
The mixture was dissolved in acetone (5 mL), CHCl3 (2 mL), and
water (1 mL). To the resulting solution was added NH4PF6
(97.8 mg, 0.6 mmol). The organic layer was separated and evapo-
rated. 1H NMR spectrum of this crude mixture indicated forma-
tion of 3a in 85% yield. Column chromatography of this mix-
ture on silica gel (70:16:11:3 CH3OH/CH2Cl2/CH3NO2/NH4Cl
(2 mol dmꢂ3) gave 3a (39.2 mg, 77%).
1H NMR showed complete consumption of 6a. The crude products
were subjected to column chromatography on silica gel (30%
AcOEt/hexane) to give 8a in a pure form (93% yield). 8a:
1H NMR (500 MHz, CDCl3, E:Z = 99:1) E-isomer: ꢀ 1.36 (t,
J ¼ 7:1 Hz, 3H), 4.30 (q, J ¼ 7:1 Hz, 2H), 6.55 (d, J ¼ 16:1 Hz,
1H), 7.68 (d, J ¼ 8:9 Hz, 2H), 7.71 (d, J ¼ 16:1 Hz, 1H), 8.25
(d, J ¼ 8:9 Hz, 2H); Z-isomer: ꢀ 1.25 (t, J ¼ 7:1 Hz, 3H), 4.17 (q,
J ¼ 7:1 Hz, 2H), 6.13 (d, J ¼ 12:6 Hz, 1H), 7.02 (d, J ¼ 12:6 Hz,
1H), 7.68 (d, J ¼ 8:9 Hz, 2H), 8.25 (d, J ¼ 8:9 Hz, 2H).
Preparation of 3a (Procedure E, Representative). In a 1-mL
vial, 1a (27.8 mg, 0.03 mmol) and 2a (42.3 mg, 0.06 mmol) were
dissolved in acetone (0.2 mL), and the vial was sealed with a
plastic cap. The acetone solution was kept standing for 2 days
at r.t. The mixture was dissolved in acetone (5 mL), CHCl3
(2 mL), and water (1 mL). To the resulting solution was added
NH4PF6 (97.8 mg, 0.6 mmol). The organic layer was separat-
Supporting Information
Preparation of 1a–1c and 2a and 2b. This material is available
free of charge on the web at http://www.csj.jp/journals/bcsj/.
1
ed and evaporated. H NMR spectrum of this crude mixture indi-
References
cated formation of 3a in 83% yield. Column chromatography of
this mixture on silica gel (70:16:11:3 CH3OH/CH2Cl2/CH3NO2/
NH4Cl (2 mol dmꢂ3) gave 3a (40.2 mg, 79%).
1
a) F. Toda, Synlett 1993, 303. b) F. Toda, Acc. Chem. Res.
1995, 28, 480. c) K. Tanaka, F. Toda, Chem. Rev. 2000, 100,
1025. d) G. W. V. Cave, C. L. Raston, J. L. Scott, Chem. Commun.
2001, 2159. e) K. Tanaka, Solvent-free Organic Synthesis,
Wiley-VCH, Weinheim, 2003.
2 A. Orita, L. Jiang, T. Nakano, N. Ma, J. Otera, Chem.
Commun. 2002, 1362.
Attempted Preparation of 3a (Procedure F, Representative).
In a 1-mL vial were placed 1a (27.8 mg, 0.03 mmol) and 2a (42.3
mg, 0.06 mmol), and the mixture was shaken. The mixture was
pelletized and compressed in 40 kg cmꢂ2 for 1 h at r.t. The mixture
was dissolved in acetone (5 mL), CHCl3 (2 mL), and water (1 mL).
To the resulting solution was added NH4PF6 (97.8 mg, 0.6 mmol).
3 a) M. Fujita, J. Yazaki, K. Ogura, J. Am. Chem. Soc. 1990,
112, 5645. b) M. Fujita, J. Yazaki, K. Ogura, Chem. Lett. 1991,
1
The organic layer was separated and evaporated. H NMR spec-
trum of this crude mixture indicated no formation of 3a.
1031.
4
Preparation of 5b (Procedure D, Representative). In a mor-
tar, 1b (36.8 mg, 0.06 mmol), 2b (37.7 mg, 0.03 mmol), and 4
(32.2 mg, 0.06 mmol) were dissolved in acetone (0.1 mL) and
CH2Cl2 (0.1 mL). The solution became a red film after 5 min
of standing in open air. The film was kept standing in open air
for 4 h at r.t. The mixture was dissolved in acetone (10 mL) and
water (2 mL). To the resulting solution was added NH4PF6
(97.8 mg, 0.6 mmol). The organic layer was separated and evapo-
rated. 1H NMR spectrum of this crude mixture indicated formation
of 5b in 85% yield. Column chromatography of this mixture
on silica gel (70:16:11:3 CH3OH/CH2Cl2/CH3NO2/NH4Cl (2
mol dmꢂ3) gave yellow solid. The resulting solid was dissolved
in acetone (10 mL) and water (2 mL). To the resulting solution
was added NH4PF6 (97.8 mg, 0.6 mmol). The organic layer was
separated, and the solvent was evaporated to afford 5b (58.6 mg,
79%): 1H NMR (300 MHz, CD3COCD3): ꢀ 9.23 (d, J ¼ 6:6 Hz,
2H), 9.18 (d, J ¼ 6:6 Hz, 2H), 8.93 (d, J ¼ 7:3 Hz, 2H), 8.30–
8.25 (m, 4H), 7.89 (d, J ¼ 8:4 Hz, 2H), 7.79–7.72 (m, 4H), 7.61
(d, J ¼ 7:5 Hz, 2H), 7.33–7.06 (m, 34H), 6.85–6.80 (m, 4H),
6.17 (s, 2H), 6.05 (s, 8H), 5.90 (s, 2H), 4.54 (mc, 2H), 4.22 (mc,
2H), 4.14–4.08 (m, 4H), 3.96–3.65 (m, 40H), 1.294 (s, 36H);
13C NMR (75.5 MHz, CD3COCD3): ꢀ 171.9, 161.1, 157.8, 157.7,
152.9, 149.2, 149.1, 148.2, 146.8, 146.7, 146.4, 145.0, 140.2,
140.1, 136.6, 135.3, 132.9, 132.8, 132.3, 131.6, 131.5, 131.4,
130.7, 128.2, 128.1, 126.6, 126.5, 126.3, 126.1, 125.1, 125.0,
116.1, 115.5, 115.1, 114.1, 114.0, 71.5, 71.3, 71.0, 70.7, 70.5,
70.4, 70.3, 69.4, 68.6, 68.3, 68.1, 68.0, 65.3, 65.0, 64.2, 62.7,
34.8, 31.6; MS (MALDI-TOF) found m=z 2180.78 ½M ꢂ 2PF6ꢃþ,
calcd for C132H152F6N3O16P m=z 2180.08. Compound 5a19 was
prepared similarly.
Other reports on self-assembly by solventless reactions:
a) D. Braga, L. Maini, M. Polito, L. Mirolo, F. Grepioni, Chem.
Eur. J. 2003, 9, 4362. b) V. Chandrasekhar, V. Baskar, R.
Boomishankar, K. Gopal, S. Zacchini, J. F. Bickley, A. Steiner,
Organometallics 2003, 22, 3710. c) S. Mann, Chem. Commun.
2004, 1.
5
a) K. Komatsu, G.-W. Wang, Y. Murata, T. Tanaka, K.
Fujiwara, K. Yamamoto, M. Saunders, J. Org. Chem. 1998, 63,
9358. b) V. P. Balema, J. W. Wiench, M. Pruski, V. K. Pecharsky,
Chem. Commun. 2002, 724.
6 G. Rothenberg, A. P. Downie, C. L. Raston, J. L. Scott,
J. Am. Chem. Soc. 2001, 123, 8701.
7
8
N. Shan, F. Toda, W. Jones, Chem. Commun. 2002, 2372.
R. P. Rastogi, N. B. Singh, R. R. Singh, J. Solid State
Chem. 1977, 20, 191.
9 L. Jiang, J. Okano, A. Orita, J. Otera, Angew. Chem., Int.
Ed. 2004, 43, 2121.
10 A. Orita, J. Okanao, Y. Tawa, L. Jiang, J. Otera, Angew.
Chem., Int. Ed. 2004, 43, 3724.
11 A. Orita, G. Uehara, K. Miwa, J. Otera, Chem. Commun.
2006, 4729.
12 For the end-capping method: Y. Furusho, H. Sasabe, D.
Natsui, K. Murakawa, T. Takata, T. Harada, Bull. Chem. Soc.
Jpn. 2004, 77, 179, and references cited therein.
13 a) K. Nikitin, B. Long, D. Fitzmaurice, Chem. Commun.
2003, 282. b) The high concentration effect for asymmetric
catalysis was also reported: Y. Yuan, X. Zhang, K. Ding,
Angew. Chem., Int. Ed. 200ꢂ3, 42, 5478.
14 The mixed Brꢂ/PF6 salt is sparingly soluble in organiꢂc
solvents, and thus the product was converted to a simple PF6
salt.
Wittig Reaction (Procedure D, Representative). In a round-
bottom flask were placed 6a (302.2 mg, 2.0 mmol) and 7a (766.4
mg, 2.2 mmol), and CH2Cl2 (4 mL) was added. The clear solution
was evaporated at 27 ꢁC for 5 min, and, during evaporation, a solid
film formed. After the film had been kept in vacuo at rt for 1 h, a
portion of the film was dissolved in CDCl3. Analysis of the film by
15 As described in Ref. 14, it is not possible to determine the
yield of the primary product because of poor solubility of the
mixed salt. Accordingly, the reaction was monitored on the basis
of consumption of 1a.
16 The concentration of the substrates was set to nearly the

A. Orita et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 8 (2007) 1623
same as reported.13a
M. Pietraszkiewicz, L. Prodi, M. V. Reddington, A. M. Z. Slawin,
N. Spencer, J. F. Stoddart, C. Vicent, D. J. Williams, J. Am. Chem.
Soc. 1992, 114, 193.
17 a) C. Reichardt, Solvents and Solvent Effects in Organic
Chemistry, Third, Updated and Enlarged Edition, Wiley-VCH,
Weinheim, 2004, Chap. 5. b) W. L. Jorgensen, Acc. Chem. Res.
1989, 22, 184.
18 P. L. Anelli, P. R. Ashton, R. Ballardini, V. Balzani, M.
Delgado, M. T. Gandolfi, T. T. Goodnow, A. E. Kaifer, D. Philp,
ˇ
´
19 P. R. Ashton, R. Ballardini, V. Balzani, M. Belohradsky,
M. T. Gandolfi, D. Philp, L. Prodi, F. M. Raymo, M. V.
Reddington, N. Spencer, J. F. Stoddart, M. Venturi, D. J.
Williams, J. Am. Chem. Soc. 1996, 118, 4931.