Metathesis reaction of formaldehyde acetals: An easy entry into the dynamic covalent chemistry of cyclophane formation

Article abstract of DOI:10.1021/ja054362o

The acid-catalyzed transacetalation of formaldehyde acetals (formal metathesis) is a suitable reaction for the generation of well-behaved Dynamic Libraries of cyclophane formals. The composition of the equilibrated mixtures solely depends on concentration

Full text of DOI:10.1021/ja054362o

Published on Web 09/08/2005
Metathesis Reaction of Formaldehyde Acetals: An Easy Entry
into the Dynamic Covalent Chemistry of Cyclophane
Formation
Roberta Cacciapaglia, Stefano Di Stefano, and Luigi Mandolini*
Contribution from the Dipartimento di Chimica and IMC-CNR Sezione Meccanismi di
Reazione, UniVersita` di Roma La Sapienza, Box 34 - Roma 62, 00185 Roma, Italy
Received July 1, 2005; E-mail: luigi.mandolini@uniromal.it
Abstract: The acid-catalyzed transacetalation of formaldehyde acetals (formal metathesis) is a suitable
reaction for the generation of well-behaved Dynamic Libraries of cyclophane formals. The composition of
the equilibrated mixtures solely depends on concentration, and is totally independent of whether the
feedstock is any of the pure cyclic oligomers, or a mixture of oligomers/polymers. Effective Molarities related
to the formation of the lower cyclic oligomers were directly measured as their equilibrium molar concentrations
above the critical monomer concentration. The finding that silver cation acts as a selective binder toward
the cyclic dimer C2, coupled with the “proof reading and editing” capability of our quickly equilibrating system,
translated into significant amplifications of C2 when the equilibrated mixtures were exposed to the action
of the silver template. These results highlight the potential of Dynamic Combinatorial Chemistry as a powerful
tool for the preparation in synthetically useful amounts of an otherwise elusive macrocyclic compound.
The possibility of using a mixture of high molecular weight byproducts as feedstock adds considerably to
the practical value of the procedure.
Introduction
In a search for reactions useful for the synthesis of cyclo-
phanes8 under mild dynamic conditions, our attention was
The number of synthetically useful irreversible reactions
involving the formation of covalent bonds largely outweighs
that of reversible ones. Yet, the past decade has witnessed a
renewal of interest in the investigation of reactions carried out
under thermodynamic control (Dynamic Covalent Chemistry).1
Such an interest has been mainly motivated by the prospects of
“proof reading and editing” via repeated bond dissociation-
recombination processes, and of amplification of a desired
component of the equilibrated mixture via specific interaction
with a target entity (template). Various reactions leading to
covalent connection between reactants have been employed,
such as olefin metathesis,2 imine,3 and hydrazone4 formation,
transesterification,5 thiol-disulfide interchange,6 and transaceta-
lation.7
attracted by the acid-catalyzed transacetalation of 1,3-dioxacy-
cloalkanes and related compounds.9 Here, we report that the
acid catalyzed transacetalation of formaldehyde acetals (formals)
nicely serves to the purpose of generating dynamic families of
oligomeric cyclophanes Ci (Scheme 1), which are fully inter-
changeable under mild conditions.
Results and Discussion
Pure samples of the lowest cyclophane oligomers were
obtained in low yields from the irreversible reaction of 1,4-
benzenedimethanol with bromochloromethane in the presence
of NaH in boiling THF (eq 1) under Ziegler’s high dilution
conditions. The complexity of the crude reaction product is
1
shown by the H NMR spectrum in Figure 1. ESI-TOF-MS
* To whom correspondence should be addressed. Fax: +39 06 490421.
(1) Rowan, S. J.; Cantrill, S. J.; Cousins, G. R. L.; Sanders, J. K. M.; Stoddart,
J. F. Angew. Chem., Int. Ed. 2002, 41, 898-952.
analysis revealed the presence of cyclic oligomers ranging from
the dimer C2 up to at least the cyclic hexamer C6 (54 ring
atoms). The absence of the cyclic monomer C1 is not surprising,
as a five atom chain is obviously too short to span a p-phenylene
moiety. Column chromatography of the crude product afforded
(2) Marcella, M. J.; Mainard, H. D.; Grubbs, R. H. Angew. Chem., Int. Ed.
1997, 36, 1101-1103.
(3) (a) Horn, M.; Ihringer, J.; Glink, P. T.; Stoddart, J. F. Chem. Eur. J. 2003,
9, 4046-4054. (b) Leung, K. C.-F.; Arico´, F.; Cantrill, S. J.; Stoddart, J.
F. J. Am. Chem. Soc. 2005, 127, 5808-5810.
(4) Lam, R. T. S.; Belenguer, A.; Roberts, S. L.; Naumann, C.; Jarrosson, T.;
Otto, S.; Sanders, J. K. M. Science 2005, 308, 667-669.
(5) Rowan, S. J.; Reynolds, D. J.; Sanders, J. K. M. J. Org. Chem. 1999, 64,
5804-5814.
(8) (a) Diederich, F. N. “Cyclophanes” The Royal Society of Chemistry,
Cambridge, 1991. (b) Vo¨gtle, F. Cyclophane Chemistry; John Wiley &
Sons Ltd.: Chichester, 1993. (c) Gleiter, R.; Hopf, H. Modern Cyclophane
Chemistry; Wiley-VCH Verlag: Weinheim, 2004.
(9) (a) Yamashita, Y.; Mayumi, J.; Kawakami, Y.; Ito, K. Macromolecules
1980, 13, 1075-1080. (b) Schulz, R. C.; Albrecht, K.; Hellermann, W.;
Kane Tran Thi, Q. V. Pure Appl. Chem. 1981, 53, 1763-1776, and
references therein.
(6) (a) Otto, S.; Furlan, R. L. E.; Sanders, J. K. M. Science 2002, 297, 590-
593. (b) Otto, S.; Kubik, S. J. Am. Chem. Soc. 2003, 125, 7804-7805. (c)
Corbett, P. T.; Tong, L. H.; Sanders, J. K. M.; Otto, S. J. Am. Chem. Soc.
2005, 127, 8902-8903.
(7) Fuchs, B.; Nelson, A.; Star, A.; Stoddart, J. F.; Vidal, S. Angew. Chem.,
Int. Ed. 2003, 42, 4220-4224.
9
13666
J. AM. CHEM. SOC. 2005, 127, 13666-13671
10.1021/ja054362o CCC: $30.25 © 2005 American Chemical Society

Dynamic Covalent Chemistry of Formal Metathesis
A R T I C L E S
Scheme 1. Ring-Ring Equilibria of Cyclophane Formals
in the given order pure samples of C2, C3, and C4 in 1.2, 3.7,
and 2.7% yield, respectively.10
The singlets of the aromatic hydrogens of the lowest
oligomers are well separated signals (Figure 1), whereas the
intense signal in the range of δ 7.30-7.36 is assigned to higher
oligomers, whose nature should be largely cyclic, because no
1
sign of end groups is visible in the H NMR spectrum.
A simple version of the acid-catalyzed transacetalation
reaction involving open chain reactants served to the purpose
of demonstrating the full reversibility under mild conditions.
An equimolar 10 mM mixture of the formals of benzyl alcohol
and of p-methylbenzyl alcohol (1 and 2, respectively) was
exposed to the action of 0.5 mM triflic acid (TfOH) in anhydrous
CHCl3 at 25 °C (eq 2). After 4 h, an equilibrium was reached,
characterized by a gaschromatographically determined composi-
tion 1:2:3 ) 1:1:2, which amounts to a K value of 4 and is
exactly what predicted for a purely entropy driven equilibrium.
The reaction is conveniently described as a reversible formal
metathesis, in that interchange of alkoxy partners bound to the
formal methylene takes place.
Figure 2. 1H NMR spectra (CDCl3, 25 °C) of equilibrated solutions of
cyclophane formals: (a) from 5 mM C2, (b) from 3.33 mM C3, (c) from
25 mM C2, (d) from 16.7 mM C3. (CHCl3 marked with an asterisk.)
The ring opening oligomerization11 of dilute solutions of C2
in CDCl3 at 25 °C was initiated by the addition of catalytic
further changes took place.12 It appears that the composition of
the reaction mixtures closely resembles that of the irreversible
reaction of eq 1 (Figure 1). The reversible nature of the acid
catalyzed formal metathesis is not only demonstrated by the
time invariance of the product composition, but even more so
1
amounts of TfOH. The H NMR spectra of typical reaction
mixtures are shown in Figure 2, traces a and c. These spectra
were taken after about 4-5 h from start, after which time no
by the finding that the 1H NMR spectra of the reaction mixtures
(10) Although the cyclic pentamer C5 was not obtained in a pure form,
unequivocal assignment of the peak at δ 7.29 (Figure 1) as associated to
its aromatic hydrogens was based on a chromatographic fraction in which
C5 was shown by ESI-TOF-MS to be the major component. Unfortu-
nately, precise integration of this signal was prevented by overlap with
other signals.
(11) Ring opening oligomerization is here preferred to the more widely used
ring opening polymerization because of the low concentrations used in our
experiments.
Figure 1. 1H NMR spectrum (CDCl3, 25 °C) of the crude product from
reaction 1. (CHCl3 marked with an asterisk.)
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13667

A R T I C L E S
Cacciapaglia et al.
Table 1. Metathesis of Cyclophane Formals, Equilibrium Yields
and Concentrationsa
yield, %c
[Ci], (mM)
[M]tot
,
entry
source
mMb
C2
C3
C4
C2
C3
C4
1
2
3
4
5
6
7
8
9
C2
C2
C3
C2
C3
C2
C3
C3
C4
C3
4.35
10.0
10.0
25.0
25.0
50.0
50.0
50.0
50.0
82.2
6.4
4.1
3.8
2.1
2.3
1.3
1.3
1.2
1.2
0.71
14.4
13.5
13.2
9.1
8.2
4.9
5.3
5.2
5.8
3.2
8d
0.14
0.21
0.19
0.27
0.29
0.33
0.32
0.31
0.30
0.29
0.21
0.45
0.44
0.77
0.69
0.81
0.88
0.87
0.96
0.89
0.09d
0.25
0.27
0.50
0.47
0.56
0.55
0.60
0.62
0.57
9.7
10.8
7.9
7.5
4.5
4.4
4.8
5.0
2.8
10
a Equilibration reactions carried out in CDCl3 at 25 °C in the presence
of 0.5 mM TfOH. Analytical data from integration of the 1H NMR signals
in the ArH region. b [M]tot is the equivalent concentration in monomer units.
c Calculated as (i[Ci]/[M]tot) × 100. d Affected by severe uncertainty because
of interference from chloroform signal.
obtained from 5 mM C2 (Figure 2a, entry 2 in Table 1) and
3.33 mM C3 (Figure 2b, entry 3 in Table 1) are practically
indistinguishable, so are the spectra of the reaction mixtures
obtained from 25 mM C2 (Figure 2c, entry 6 in Table 1) and
16.7 mM C3 (Figure 2d, entry 8 in Table 1). Thus, as expected
for a truly reversible system, the composition at equilibrium is
the same, no matter what oligomer is used as feedstock, on
condition that the equivalent concentration expressed in mono-
mer units ([M]tot) is the same. On the other hand, concentration
has a marked influence on the equilibrium composition. High
dilution favors the lower cycles at the expense of high molecular
weight materials, as shown by a comparison of traces a and b
with c and d in Figure 2, and on a more quantitative basis by
the equilibrium concentrations and yields of oligomers C2-C4
listed in Table 1.
Figure 3. 1H NMR spectra (CDCl3, 25 °C) of: (a) equilibrated solution
from 8.33 mM C3, (b) solution of trace a immediately after addition of
solid C3 so as to double the concentration, (c) solution of trace b after
equilibration, (d) equilibrated solution from 25 mM C2. The axis refers to
spectrum a. Other spectra are offset with respect to the next one by 0.4
ppm (CHCl3 marked with an asterisk).
An important characteristic of dynamic systems is the ability
to readjust the product distribution by changing the factors that
rule the equilibrium, even once the system has reached the
equilibrium composition dictated by the initial conditions. Trace
a in Figure 3 refers to an equilibrated reaction mixture in which
8.33 mM C3 ([M]tot ) 25 mM, entry 5 in Table 1) was the
starting material. The equilibrium was perturbed by adding an
amount of solid C3 such as to double the equivalent monomer
1
concentration ([M]tot ) 50 mM). Comparison of the H NMR
Figure 4. Computer generated molecular model of the C2‚Ag+ complex.
spectrum taken immediately after complete dissolution of C3
(trace b) with that taken after reequilibration (trace c, entry 8
in Table 1) shows that excess C3 was digested and mostly
transformed into high molecular weight materials, as the
equivalent concentration in the original solution was not far from
the critical monomer concentration (see below). That the
spectrum in trace c belongs to a fully equilibrated mixture is
demonstrated by the fact that it is indistinguishable from that
(trace d, entry 6 in Table 1) taken on an equilibrated reaction
mixture derived from 25 mM C2 ([M]tot ) 50 mM).
Similar results were obtained from an experiment in which
the equilibrium was perturbed by dilution (1H NMR spectra not
shown). Briefly, a 33 mM solution of C3 ([M]tot ) 100 mM)
was equilibrated in the usual way. After the equilibrium was
reached, the reaction mixture was diluted 1 to 10 by adding
CDCl3 containing 0.5 mM TfOH, so as to leave unchanged the
catalyst concentration. A new state of equilibrium was reached
(entry 3 in Table 1), whose composition was again indistin-
guishable from that obtained from 5 mM C2 as the starting
material (entry 2 in Table 1).
Amplification Experiments. Even more significant is the
equilibrium perturbation caused by the addition of a selective
template, which can lead to amplification of a desired component
of the dynamic library. The affinity of the silver cation toward
π-base receptors such as cyclophanes and calix[4]arenes is well-
known.13 Molecular models indicate that silver ion has the right
size to be hosted between the aromatic rings of C2 in a highly
distorted sandwich structure (Figure 4). Strong 1H NMR
evidence was obtained that C2, unlike its higher homologues,
forms a complex of definite stability with Ag+,14 but the very
(13) Lhota´k, P.; Shinkai, S. J. Phys. Org. Chem. 1997, 10, 273-285.
(14) ∆δ values of +0.42, +0.051, -0.046 for the ArH, -OCH2O-, and
ArCH2O- signals, respectively, were observed in the presence of excess
solid TfOAg.
(12) Occasional checks showed that the 1H NMR spectra of equilibrated solutions
remained unchanged even after more than one week. This indicates the
virtual absence of irreversible side reactions.
9
13668 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Dynamic Covalent Chemistry of Formal Metathesis
A R T I C L E S
Figure 6. Equilibrium concentration of C2-C4 as a function of total
monomer concentration (data from Table 1).
mercially available (CF3SO2)2NAg as a template. When an
amplification experiment was carried out in the NMR tube in
the presence of 1 mmol equiv of (CF3SO2)2NAg under
heterogeneous conditions in CDCl3, the yield of C2 raised to
60%, to be compared with the 11% yield obtained in the
analogous experiment with TfOAg, most likely due to a higher
solubility of the former salt.
Figure 5. 1H NMR spectrum of the ArCH2 region (CD2Cl2, 25 °C) of an
equilibrated solution from 6.25 mM C4, in the absence (top) and presence
(bottom) of solid TfOAg (aromatic region not shown because overlap of
the ArH signals of C2‚Ag+ with other signals).
low solubility of the parent salt TfOAg in chloroform prevented
a quantitative estimate of the binding affinity.
Effective Molarities. Plots of the molar concentrations of
macrocycles C2-C4 against equivalent monomer concentration
(Figure 6) display negative curvatures, with an unmistakable
tendency toward saturation in the high concentration region.
This behavior is well in keeping with what predicted by theory,16
and fully confirmed by experiment.17 According to the theory
of macrocyclization equilibria,16 as the monomer concentration
is raised, the concentration of each individual cyclic species
increases until a critical monomer concentration is reached.
Above such critical value, the concentration of each cyclic
species remains constant, and coincides with the effective
molarity EMi18 of the given cyclic oligomer. Although our data
do not allow the critical monomer concentration to be precisely
determined, inspection of the data plotted in Figure 6 strongly
indicates that the critical monomer concentration should lie
somewhere in the range from 50 to 100 mM. We note that the
EMi values related to the formation of cyclic oligomers C2-
C5 (Table 2) are surprisingly low. In fact, they are much lower
than the EMs usually recorded for the formation of a large
number of macrocycles of comparable sizes,19 which provides
a strong indication that oligomers C2-C5 are affected by
considerable strain. An admittedly crude estimate of the strain
energies involved can be obtained by a comparison of the
experimental EMi values with the EMi* values (Table 2)
estimated for strainless rings in which the ease of formation is
The outcome of the formal metathesis reaction changed
significantly in the presence of the silver template. When an
equilibrated mixture derived from C4 ([M]tot ) 25 mM, TfOH
) 0.5 mM, CDCl3, 25 °C) was treated with TfOAg (1 mol per
monomer equivalent) and left to stand at 25 °C for 4 h, the
yield of C2 raised from 2.2 to 11%. Clearly, the silver cation
selectively bound to C2 and shifted the ring-ring equilibrium
position in its favor. Changing the solvent to CD2Cl2 caused a
further increase to 20% of the yield of C2 (Figure 5), presumably
due to a higher solubility of the salt in this solvent.
Given the superiority of methylene chloride over chloroform,
a preparative scale experiment was carried out in the latter
solvent. As feedstock we used the mixture of oligomeric/
polymeric material with i g 5, recovered from the reaction of
1,4-benzenedimethanol with CH2BrCl (eq 1) as waste material
after the chromatographic isolation of oligomers C2-C4. A
solution of 500 mg of the above material (3.3 monomer mequiv)
in 140 mL of dry CH2Cl2 was treated at room temperature with
1 g of TfOAg (3.9 mol) and TfOH (0.06 mmol) for 4 h. Workup
with aqueous ammonia, followed by column chromatography
afforded pure C2 (170 mg) in 34% yield.15 All the remaining
material was recovered by elution with acetone and reequili-
brated under the same conditions to give additional 75 mg of
pure C2, thus leading to a 49% isolated overall yield of C2 after
two templated equilibration steps. The latter experiment il-
lustrates well the high synthetic potential of the “proof reading”
and “error checking” capability of the dynamic formal metath-
esis, when carried out in the presence of such selective template
as a silver cation. Clearly, the templating efficiency is greatly
influenced by the heterogeneous nature of the reaction mixture
caused by the sparing solubility of the salt used as a source of
silver ions. In fact, in the late part of this work we discovered
that even better results could be obtained using the noncom-
(16) (a) Jacobson, H.; Stockmayer, W. H. J. Chem. Phys. 1950, 18, 1600-
1606. (b) Ercolani, G.; Mandolini, L.; Mencarelli, P.; Roelens, S. J. Am.
Chem. Soc. 1993, 115, 3901-3908.
(17) (a) Ito, K.; Hashizuka, Y.; Yamashita, Y. Macromolecules 1977, 10, 821-
824. (b) ten Cate, A. T.; Kooijman, H.; Spek, A. L.; Sijbesma, R. P.; Meijer,
E. W. J. Am. Chem. Soc. 2004, 126, 3801-3808. (c) Paulusse, J. M. J.;
Sijbesma, R. P. Chem. Commun. 2003, 1494-1495.
(18) Polymer chemists usually refer to the macrocyclization equilibrium constant
Ki (see Flory, P. J. “Statistical Mechanics of Chain Molecules” Wiley
Interscience, New York, 1969). It was shown that Ki ) EMi within the
usual approximation of reactivity of end groups independent of chain length
(see ref 19). It should be noted that, by definition, the critical monomer
concentration is equal to the quantity Σi iEMi.
(15) The higher yield of C2 in the preparative experiment, carried out under
efficient stirring, might indicate incomplete equilibration in the NMR tube
where the experiment of Figure 5 was carried out.
(19) (a) Mandolini, L. AdV. Phys. Org. Chem. 1986, 22, 1-111. (b) Galli, C.;
Mandolini, L. Eur. J. Org. Chem. 2000, 3117-3125.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13669

A R T I C L E S
Cacciapaglia et al.
Table 2. Experimental EMi Values and Calculated Strain Energies
of Cyclic Oligomers C2-C5
the availability of a selective template for the exploitation of
dynamic libraries in synthetic work. The finding that oligomeric/
polymeric materials obtained as undesirable and otherwise
useless byproducts can be recycled and converted into syntheti-
cally useful amounts of C2 adds considerably to the practical
value of the method.
As an important “byproduct” of the work, the equilibration
experiments carried out above the critical concentration provided
direct access to the Effective Molarities of the lower cyclic
oligomers, which are the fundamental thermodynamic quantities
in any physicochemical discussion of ring-chain and ring-
ring equilibria.
strain energy
(kcal/mol)c
i
ring atoms
EMi (M)a
EMi* (M)b
2
3
4
5
18
27
36
45
3.0 × 10-4
9.0 × 10-4
5.9 × 10-4
4.2 × 10-4 d
8.5 × 10-2
1.6 × 10-2
7.0 × 10-3
4.0 × 10-3
3.3
1.7
1.5
1.3
a Error limits in the range of (10% unless otherwise stated. b Taken from
the compilation of entropic components of EM as a function of the number
r of rotatable bonds reported in Table 1 of ref 19b. For the present system
r ) 6i - 1. The listed EMi* data have been corrected for the symmetry
number of Ci, namely, σi ) 2i. c Calculated as RTln(EMi*/EMi). d Ap-
proximate value, due to the poor precision of the integrated intensity of the
1H NMR signal at δ ) 7.29 (see ref 10). A conservative estimate of the
uncertainty is in the range of (25%.
Experimental Section
Instruments and General Methods. NMR spectra were recorded
on either a 200- or 300-MHz spectrometer. Chemical shifts are reported
as δ values in ppm from tetramethylsilane added as an internal standard.
Equilibration reactions were carried out in the NMR tube in the
thermostated probe of the spectrometer. High-resolution mass spectra
(HR-MS) were performed by an Electrospray Ionization Time-of-Flight
spectrometer.
Materials. CF3SO3H and CF3SO3Ag were commercial samples and
used without further purification. (CF3SO2)2NAg was prepared as
described in the literature.21 THF was dried by distillation from sodium
benzophenone ketyl. CDCl3 was dried over activated molecular sieves
(4 Å).
solely determined by the conformational entropy loss upon
cyclization of open-chain precursors composed of r rotatable
bonds.19
As shown by the data in Table 2, the strain energy, which is
quite high for cyclic dimer C2, shows the tendency to decrease
on increasing the ring size, but is still appreciable for the cyclic
pentamer C5 (45 ring atoms). Although too much emphasis
cannot be placed on exact figures, in view of the approximate
nature of the treatment involved, the observed trend is believed
to be real. As to the origin of strain in our cyclophane formals,
electron diffraction studies and ab initio calculations20 show that
the gauche (g) conformation of the C‚‚‚OsC‚‚‚O bond of
dimethoxymethane is more stable than the trans (t) conformation
(anomeric effect). A number of different sources20 indicate that
the most stable conformation, gg (denoting two gauche bonds
of the same sign), corresponds to a deep and narrow energy
well. Examination of CPK molecular models of the cyclic
oligomers shows that geometrical constraints imposed by the
cyclic structures force the CsOsCsOsC chains to adopt
conformations somewhat different from the most stable gg
conformation, and that such geometrical constraints become less
severe when the ring size gets larger. In conclusion, the torsional
strain arising from deviations of the CsOsCsOsC chains
from the ideal gg conformation is believed to be responsible
for the low EMi values of oligomers C2 - C5.
1,5-Diphenyl-2,4-dioxapentane (1). Bromochloromethane (0.480
mL, 7.4 mmol) and benzyl alcohol (0.380 mL, 3.7 mmol) were added
to a suspension of NaH (60% w/w, 0.450 g, 11.2 mmol) in dry THF
(70 mL). The mixture was refluxed for 14 h, cooled to room
temperature, and sodium hydroxide (1 M) was added to quench the
excess of NaH. Water (70 mL) was added and the mixture was extracted
with Et2O (3 × 100 mL). The combined organic phases were dried
over Na2SO4 and evaporated to give the pure product as a colorless
oil. Yield: 0.340 g, 81%. 1H NMR (200 MHz, CDCl3): δ ) 7.34 (m,
10H), 4.84 (s, 2H), 4.65 (s, 4H); 13C NMR (50 MHz, CDCl3) δ )
137.79, 128.41, 127.91, 127.70, 93.90, 69.48; HR-MS calcd for
C15H16O2+Na+: 251.1048; found: 251.1054.
1,5-Di-p-tolyl-2,4-dioxapentane (2). Bromochloromethane (0.480
mL, 7.4 mmol) and 4-methylbenzyl alcohol (0.450 g, 3.7 mmol) were
reacted as above to give the pure product as a colorless oil. Yield:
0.408 g, 86%. 1H NMR (200 MHz, CDCl3): δ ) 7.24 (ABq, J ) 8.02
Hz, 8H), 4.85 (s, 2H), 4.64 (s, 4H), 2.38 (s, 6H); 13C NMR (50 MHz,
CDCl3) δ: 137.37, 134.77, 129.06, 128.05, 93.60, 69.25, 21.13; HR-
MS calcd for C17H20O2+Na+: 279.1361; found: 279.1362.
Cyclic Oligomers C2-C4. Bromochloromethane (8.28 mL, 0.135
mol) was added to a suspension of NaH (60% w/w, 5.4 g, 0.135 mol)
in dry THF (450 mL). The mixture was heated to reflux and
1,4-benzenedimethanol (3 g, 0.022 mol) in THF (50 mL) was added
dropwise by a syringe during 24 h under an argon atmosphere. The
mixture was subsequently refluxed for 2 days, then cooled to room
temperature, and sodium hydroxide (1 M) was added to quench the
excess of NaH. After addition of water (150 mL) the mixture was
extracted with CH2Cl2 (1 × 400 mL and 2 × 200 mL). The combined
organic phases were dried over Na2SO4 and evaporated to give 3.1 g
of crude product. Pure samples of C2, C3, and C4 were obtained by
column chromatography on silica gel. After elution of a colored impurity
with CH2Cl2/heptane 11:8, elution with CH2Cl2/heptane/acetone 11:8:
0.7 gave the pure title compounds in the given order.
Concluding Remarks
In summary, the acid-catalyzed formal metathesis is a suitable
reaction for the generation of rapidly equilibrating and long-
lived libraries of cyclophane formals. Depending on whether
the feedstock is pure C2 or a mixture of high molecular weight
materials, the process can be formally described either as a ring-
opening cyclooligomerization/polymerization or as a cyclode-
polimerization, but it is clear that the composition of the
equilibrated mixtures solely depends on the total concentration
of monomer units, and is independent of the source of such
units. The thermodynamic sink featured by the selective
complexation of the cyclic dimer C2 with silver salts results in
high degree of amplifications of C2 under silver(I) template
action. These results establish Dynamic Combinatorial Chem-
istry as a convenient method for the synthesis of an otherwise
elusive macrocycle, and emphasize the crucial importance of
2,4,13,15-Tetraoxa[5,5]paracyclophane (C2). Yield: 40 mg, 1.2%.
1
Mp 161.5-163 °C; H NMR (200 MHz, CDCl3): δ ) 6.88 (s, 8H),
4.89 (s, 4H), 4.49 (s, 8H); 13C NMR (50 MHz, CDCl3) δ ) 137.33,
(20) (a) Jeffrey, G. A.; Pople, J. A.; Binkley, J. S.; Vishveshwara, S. J. Am.
Chem. Soc. 1978, 100, 373-379. (b) Smith, G. D.; Jaffe, R. L.; Yoon, D.
Y. J. Phys. Chem. 1994, 98, 9072-9077. (c) Smith, G. D.; Jaffe, R. L.;
Yoon, D. Y. J. Phys. Chem. 1994, 98, 9078-9082.
(21) Vij, A.; Zheng, Y. Y.; Kirchmeier, R. L.; Shreeve, J. M. Inorg. Chem.
1994, 33, 3281-3288.
9
13670 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Dynamic Covalent Chemistry of Formal Metathesis
A R T I C L E S
126.75, 96.54, 70.70. HR-MS calcd for C18H20O4+H+: 301.1440;
found: 301.1428; calcd for C18H20O4+Na+: 323.1255; found 323.1259.
2,4,13,15,24,26-Hexaoxa[5,5,5]paracyclophane (C3). Yield: 120
) 7.23 (s, 16H), 4.84 (s, 8H), 4.58 (s, 16H); 13C NMR (50 MHz, CDCl3)
δ ) 137.46, 128.11, 94.63, 69.77; HR-MS calcd for C36H40O8+Na+:
623.2621; found: 623.2635; calcd for C36H40O8+K+: 639.2360;
found: 639.2386.
1
mg, 3.7%. Mp 104.5-105.5 °C; H NMR (200 MHz, CDCl3): δ )
7.14 (s, 12H), 4.88 (s, 6H), 4.50 (s, 12H); 13C NMR (50 MHz, CDCl3)
δ: 137.69, 128.31, 96.01, 70.71; HR-MS calcd for C27H30O6+Na+:
473.1940; found: 473.1929; calcd for C27H30O6+K+: 489.1679;
found: 489.1677.
Acknowledgment. Financial contribution from MIUR COFIN
2003 Progetto Dispositivi Supramolecolari is acknowledged.
2,4,13,15,24,26,35,37-Octaoxa[5,5,5,5]paracyclophane
(C4).
Yield: 90 mg, 2.7%. Mp 102-104 °C; 1H NMR (200 MHz, CDCl3) δ
JA054362O
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13671