α- and β-bis(1,1,1-trifluoro-5,5-dimethyl-5-methoxyacetylacetonato)copper(II): Transforming the dense polymorph into a versatile new microporous framework

Article abstract of DOI:10.1021/ja981443u

Bis(1,1,1-trifluoro-5,5-dimethyl-5-methoxy-acetylacetonato)copper(II) was prepared in two polymorphic modifications. The orthorhombic α-form is stable and densely packed, with four trans and four cis square bischelate building blocks per unit cell. These are connected through additional coordination bonds to form a dense polymer network. For the trigonal β-form, the square bischelate complex units are present exclusively as the trans isomers. The distinctive assembly of these units results in a lattice with an open pore volume of about 17% that is accessible to a wide range of guests. The compound has a remarkably strong affinity for the porous β-form as evident from the efficient α-to-β conversion on contact not only with liquid guests but also with organic vapors at pressures well below the saturation pressure. Although the open β-form is metastable, it has a remarkable kinetic stability, most likely because of the trans-to-cis isomerization that must accompany the β-to-α transformation. Many sorbents play a dual role as stabilizing guest and as catalyst promoting the α-to-β or β-to-α conversion. Because of its versatile sorption properties and relative robustness, the β-form of the complex can be classified as a novel organic zeolite mimic.

Full text of DOI:10.1021/ja981443u

J. Am. Chem. Soc. 1999, 121, 4179-4188
4179
R- and
â-Bis(1,1,1-trifluoro-5,5-dimethyl-5-methoxyacetylacetonato)copper(II):
Transforming the Dense Polymorph into a Versatile New
Microporous Framework†
D. V. Soldatov,‡,§ J. A. Ripmeester,*,‡ S. I. Shergina, I. E. Sokolov, A. S. Zanina,
S. A. Gromilov,§ and Yu. A. Dyadin§
Contribution from the Steacie Institute for Molecular Sciences, National Research Council of Canada,
Ottawa K1A 0R6, Canada, Institute of Inorganic Chemistry, Siberian Branch of the Russian Academy of
Sciences, NoVosibirsk 630090, Russia, and Institute of Chemical Kinetics and Combustion, Siberian
Branch of the Russian Academy of Sciences, NoVosibirsk 630090, Russia
ReceiVed April 27, 1998
Abstract: Bis(1,1,1-trifluoro-5,5-dimethyl-5-methoxy-acetylacetonato)copper(II) was prepared in two poly-
morphic modifications. The orthorhombic R-form is stable and densely packed, with four trans and four cis
square bischelate building blocks per unit cell. These are connected through additional coordination bonds to
form a dense polymer network. For the trigonal â-form, the square bischelate complex units are present
exclusively as the trans isomers. The distinctive assembly of these units results in a lattice with an open pore
volume of about 17% that is accessible to a wide range of guests. The compound has a remarkably strong
affinity for the porous â-form as evident from the efficient R-to-â conversion on contact not only with liquid
guests but also with organic vapors at pressures well below the saturation pressure. Although the open â-form
is metastable, it has a remarkable kinetic stability, most likely because of the trans-to-cis isomerization that
must accompany the â-to-R transformation. Many sorbents play a dual role as stabilizing guest and as catalyst
promoting the R-to-â or â-to-R conversion. Because of its versatile sorption properties and relative robustness,
the â-form of the complex can be classified as a novel organic zeolite mimic.
Introduction
microporous architectures with the ability to switch between
dense and open forms.
There is considerable interest in the design of novel organic
and metal-organic materials that exhibit zeolite-like behavior.
Materials that have a predisposition toward forming open
structures often show a rapidity and reversibility of guest
inclusion that facilitate their function as zeolite mimics. Potential
applications include processes such as separation and purifica-
tion, catalysis, stereoselective syntheses, sensor applications, or
the preservation of unstable or highly reactive species.1 Despite
extensive efforts, only a few host types forming stable porous
matrixes have been discovered so far,2 as most either collapse
when the templating guest is removed or lose their crystallinity.
Here we report a novel host matrix capable of supporting a
remarkably robust channel structure. Stable kinetically, it is
sufficiently labile that controlled transformation of the dense
to the open forms and vice versa can be contemplated as a design
feature. The host shows considerable versatility in sorption
behavior and an unusually strong affinity for the formation of
inclusions with aliphatic hydrocarbon gases and offers insight
into possible new design concepts for the development of
The crystal structure of the title complex in its open form
was recently reported.3 The presence of solvent species,
however, was not clearly established. Our work has revealed
that the quantities of included solvent must have been significant,
from both the unit cell volume of 5486 Å and the melting point
of 94 °C (cf. 5450 Å and 157 °C for the pure complex,
respectively, according to the present work). Here we report
the synthesis and structure of both dense and open forms, as
well as data on the versatility as inclusion host of the title
compound.
Results and Discussion
r-Form of [CuL2]. The reaction of CF3COCH2COC(OMe)-
Me2 ()HL) with Cu(II) salts in basic water/ethanol solution
(2) (a) Allison, S. A.; Barrer, R. M. J. Chem. Soc. A 1969, 1717-1723.
(b) Ibragimov, B. T.; Talipov, S. A.; Aripov, T. F. J. Inclusion Phenom.
1994, 17, 317-324. (c) Ung, A. T.; Gizachew, D.; Bishop, R.; Scudder,
M. L.; Dance, I. G.; Craig, D. C. J. Am. Chem. Soc. 1995, 117, 8745-
8756. (d) Venkataraman, D.; Gardner, G. B.; Lee, S.; Moore, J. S. J. Am.
Chem. Soc. 1995, 117, 11600-11601. (e) Mak, T. S. W.; Bracke, B. R. F.
In ComprehensiVe Supramolecular Chemistry; MacNicol, D. D., Toda, F.,
Bishop, R., Eds.; Pergamon: Oxford, 1996; Vol. 6, p 39. (f) Gdaniec, M.;
Ibragimov, B. T.; Talipov, S. A. Ibid. 139-140. (g) Reference 18c, pp 697-
698. (g) Brunet, P.; Simard, M.; Wuest, J. D. J. Am. Chem. Soc. 1997, 119,
2737-2738. (h) Kondo, M.; Yoshitomi, T.; Seki, K.; Matsuzaka, H.;
Kitagawa, S. Angew. Chem., Int. Ed. Engl. 1997, 36, 1725. (i) Yaghi, O.
M.; Li, G.; Li, H. Nature 1995, 378, 703. (j) Russell, V. A.; Evans, C. C.;
Li, W., Ward, M. D. Science 1997, 276, 575
* To whom correspondence should be addressed. E-mail: jar@
ned1.sims.nrc.ca. Fax: (613) 998-7833.
† Issued as NRCC No. 40910.
‡ National Research Council of Canada.
§ Institute of Inorganic Chemistry, Siberian Branch of the Russian
Academy of Sciences.
Institute of Chemical Kinetics and Combustion, Siberian Branch of
the Russian Academy of Sciences.
(1) ComprehensiVe Supramolecular Chemistry, Vol. 10, Supramolecular
Technology; Reinhoudt, D. N., Ed.; Pergamon: Oxford, 1996.
(3) Gromilov, S. A.; Baydina, I. A.; Zharkova G. I. Zh. Strukt. Khim.
1997, 38, 946-953 (in Russian).
10.1021/ja981443u CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/16/1999

4180 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
SoldatoV et al.
Table 1. Parameters for the X-ray Experiments, Crystal Data and Structure Analysis for the Two Polymorphs of the Complex and Its Benzene
Clathrate
R-[CuL2]
â-[CuL2]
â-[CuL2]‚2/3C6H6
Single-Crystal Experiment
C16H20CuF6O6
485.9
blue, massive
orthorhombic, Cmc21 (no. 36)
a ) 14.463(1) Å
b ) 15.827(1) Å
c ) 18.742(2) Å
4290(1) Å3
formula
C20H24CuF6O6
537.9
blue, prism
trigonal, R3h (no. 148)
a ) 24.492(1) Å
c ) 10.624(1) Å
formula weight
crystal color, habit
system, space group
unit cell dimensions
unit cell volume
5519(1) Å3
Z
8
9
calculated density
absorption coefficient
1.504 g cm-1
1.457 g cm-1
11.0 cm-1
9.7 cm-1
total reflections
22 996 (Rint ) 0.029)
5410
4708
13 098 (Rint ) 0.033)
number of unique reflections (I > 0)
number of unique reflections (I > 2σ(I))
number of parameters
2879
2408
288
438
unweighted agreement factor, R (data with I > 2σ(I))
weighted agreement factor
Rw2 (data with (I > 2σ(I))a
goodness of fit
absolute structure parameter
residual extrema
0.042
0.097
k ) 0.0469, m ) 3.06
1.038
0.040
0.098
k ) 0.0555, m ) 3.71
1.050
0.04(2)
+0.44, -0.33 e Å-1
+0.58, -0.17 e Å-1
Powder Experiment
20
orthorhombic
a ) 14.468(4) Å
b ) 15.842(5) Å
c ) 18.734(9) Å
4294(3) Å3
number of used reflections
system
unit cell dimensions
17
trigonal
a ) 24.417(2) Å
c ) 10.555(2) Å
17
trigonal
a ) 24.500(3) Å
c ) 10.623(3) Å
unit cell volume
5450(1) Å3
0.498(1)
5522(2) Å3
0.577(1)
Packing Analysis
0.565(1)
packing coefficient
2
2
2
2
2
a Rw ) ∑[w(Fo - Fc2)2]/{∑[w(Fo )2]}1/2; w ) 1/[σ2(Fo ) + (kP)2 + mP], where P ) (max(Fo ) + 2Fc2)/3.
Table 2. Lengths (Å) and Angles (deg) of Additional Bonds in the Compounds Studieda
compound
R-[CuL2]
bond type
atom D
atom A
D‚‚‚A
atom H
D‚‚‚H
D‚‚‚H-A
coordination, intermolecular
coordination, intermolecular
coordination, intermolecular
coordination, intermolecular
hydrogen, intermolecular
hydrogen, intermolecular
coordination, intermolecular
coordination, intermolecular
hydrogen, intermolecular
hydrogen, intermolecular
coordination, intermolecular
coordination, intermolecular
coordination, intermolecular
coordination, intermolecular
hydrogen, intermolecular
hydrogen, intermolecular
O15B*2
O15B*3
O25B*2
O25B*3
O15A
O25A
O25A
F23A*4
O15B
O25B
O15#2
O15#3
O25#2
O25#3
O15
Cu1A
Cu1A
Cu1A
Cu1A
C13A
C23A
Cu1B
Cu1B
C13B
C13B
Cu
2.804(3)
2.804(3)
2.702(8)
2.702(8)
2.615(6)
2.846(7)
2.695(5)
3.253(5)
2.834(5)
2.913(9)
2.800(2)
2.800(2)
2.879(6)
2.879(6)
2.892(3)
2.773(7)
H13A
H23A
2.220(4)
2.573(6)
104.6(3)
97.3(3)
H13B
H13B
2.505(4)
2.66(1)
101.1(3)
96.6(3)
â-[CuL2]‚2/3C6H6
Cu
Cu
Cu
C13
H13
H13
2.648(2)
2.429(8)
93.6(2)
102.8(2)
O25
C13
a Symmetry operators: (*2) ) (-1/2 - x, 3/2 - y, 1/2 + z); (*3) ) (1/2 + x, 3/2 - y, 1/2 + z); (*4) ) (-x, 1 - y, -1/2 + z); (#2) ) (1/3
- y, x - y - 1/3, z - 1/3); (#3) ) (2/3 + y, 1/3 - x + y, 4/3 - z).
gives blue [CuL2] bischelate. Ethanol and water are able to
coordinate to the molecules, thus giving green products, but
these solvents are easily removed by drying in air or by
warming. Vacuum sublimation of the blue complex results in
the pure crystalline R-form of [CuL2] (see Figure 5 for a powder
diffractogram). This is the stable polymorphic modification of
the complex without pore space for guest inclusion. It sublimes
at temperatures above 150 °C and melts congruently in a sealed
tube at 157 °C to give green liquor.
two chelate ligands (Figure 1). The Cu-O bond lengths are
quite short at 1.90-1.92 Å. The difference between A and B
molecules is in the arrangement of the ligands; molecule A has
a trans configuration, while that of unit B is cis. The crystal is
therefore an equimolar mixture of two isomers of the compound
situated in an ordered crystal lattice.4
The trans molecule A is located on a mirror plane perpen-
dicular to the crystallographic a direction. The copper atom and
chelate rings lie in the plane, while the trifluoromethyl and
From X-ray structure determination (Tables 1 and 2) R-[CuL2]
is orthorhombic, with two crystallographically independent
complex units (labeled A and B). Both have square-planar
coordination of the Cu(II) cation by four donor oxygens from
(4) This phenomenon is interesting itself, and is analogous to contact
conformational isomerism (coexistence of different conformations in the
same structure)5 and contact stabilization of molecules.6 It may be referred
to as “contact coordinational isomerism”.

Transforming Dense [CuL2] into a Microporous Framework
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4181
of specific interactions occur in the structure (Table 2). All
possible intramolecular C-H‚‚‚O hydrogen bonds are observed.
As for the additional apical coordination to copper, this
interaction is impossible within the same molecule and thus of
necessity becomes intermolecular. Except for atom O15A, all
methoxy oxygens participate in both types of interaction. It is
interesting that the methoxy-tert-butyl fragment of molecule B
is disordered over a main (89%) and a secondary (11%)
orientation. The oxygen atom associated with the main orienta-
tion, O15B, coordinates to Cu1A at 2.80 Å and is linked to
C13B at 2.83 Å by hydrogen bonding. The atom associated with
the second orientation, O25B, is coordinated more strongly to
Cu1A at 2.70 Å and forms a weaker connection to C13B at
2.91 Å. Ligation of atom O15A to copper is not observed, and
instead one of the apical positions on Cu1B is occupied by
fluorine F23A. Although the Cu-F bond is much weaker than
Cu-O, its presence is evident from the interatomic distance of
3.25 Å, the mutual arrangement of the atoms, and the fact that
the trifluoromethyl group containing F23A is the only one which
is not disordered. Thus, the apical coordination sites on copper
in molecule A are occupied by two oxygen donors, while in
molecule B there are one oxygen and one fluorine atom.
R-[CuL2] turns out to have a somewhat cumbersome,
complicated crystal structure for such a simple molecule: it has
a large unit cell with eight formula units, two crystallographi-
cally distinct molecules, isomerism within the same structure,
a distorted geometry of the bischelate fragment in molecule B,
and, finally, coordination of a trifluoromethyl fluorine atom to
copper in the presence of available oxygen. All of these
observations suggest that molecular packing is inefficient, and
assembly to give a dense structure is difficult.75-7 Such problems
of inefficient packing are overcome in the second, â-architecture
of the compound. Although the packing arguments do not tell
the entire story, they do give qualitative insight into the factors
contributing to the relative thermodynamic stabilities of the two
forms.
Figure 1. Two complex units found in R-[CuL2]: top, trans isomer;
bottom, cis isomer. Each unit cell contains four trans and four cis units.
â-Inclusion with Benzene, â-[CuL2]‚2/3C6H6. When R-[CuL2]
is contacted by an appropriate (see below) liquid or gaseous
organic guest, there is a significant increase in mass and a phase
transformation to the trigonal â-modification (see, e.g., Figure
5a,b). Among the adducts with aromatics (Table 3), the benzene
compound itself has remarkable stability: crystals of the
compound show no mass decrease in a vacuum (0.03 Torr) for
24 h. In a closed volume the compound is stable up to 141 °C,
at which point it melts incongruently into the solid R-form and
green liquor.
The single-crystal X-ray diffraction study (Table 1) illustrates
the inclusion character of the compound. The host complex
molecule is centrosymmetric (Figure 2), closely resembling the
trans configuration of molecule A in the R-form. There is square-
planar ligation by two pairs of chelating oxygens from within
the molecule. Donor oxygens associated with the methoxy
groups take part in the formation of intramolecular hydrogen
bonds with C-H groups of the chelate fragments, and at the
same time ligate to the apical coordination sites on copper atoms
of the adjacent molecules (Table 2). As for the R-form, disorder
of the methoxy-tert-butyl tail is present. The oxygen atom in
the main orientation (81%), O15, is coordinated to Cu at 2.80
Figure 2. Host molecule and oxygen atoms of neighboring hosts in
â-[CuL2]‚2/3C6H6.
methoxy-tert-butyl substituents are disordered on either side of
the plane. The molecule is asymmetric but roughly approximates
centrosymmetry (cf. the molecule in the benzene inclusion,
Figure 2). Molecule B is cleaved by a mirror plane through the
copper atom, the molecule thus consisting of two symmetrically
placed halves in a cis configuration. Also, the bischelate
fragment is bent, with the dihedral angle between chelate rings
as large as 20°.
Each isolated molecule has two extra coordination sites on
the copper cation and two donor oxygen atoms associated with
the methoxy groups. There also are two hydrogen atoms whose
acidity is promoted by the participation of adjacent carbons in
the π-system of the chelate ring and the shifting of electron
density toward the copper cation. Therefore, two additional types
(5) Zorkii, P. M.; Razumaeva, A. E. Zh. Strukt. Khim. 1979, 20 (3), 463-
466 (in Russian); Chem. Abstr. 1979, 91, 174710a.
(6) (a) Dyadin, Yu. A.; Kislykh, N. V. MendeleeV Commun. 1991, 134-
136. (b) Dyadin, Yu. A.; Soldatov, D. V., Logvinenko, V. A.; Lipkowski,
J. J. Coord. Chem. 1996, 37, 63-75.
(7) Kitaigorodsky, A. I. Mixed Crystals; Nauka: Moscow, 1983 (in
Russian).

4182 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
Table 3. â-[CuL2] and Its Inclusionsa
SoldatoV et al.
unit cell parameters
composition G:H
N
guest
a, Å
c, Å
V, Å3
method A
method B
1
2
(empty phase)
benzene, C6H6
24.417(2)
24.500(3)
24.492(1)
24.544(3)
24.572(5)
24.512(3)
24.458(6)
24.517(5)
24.456(5)
24.500(4)
24.488(6)
24.650(5)
24.505(4)
24.519(4)
24.462(5)
24.491(5)
24.455(4)
24.467(4)
24.513(6)
24.514(5)
24.483(5)
24.460(6)
24.652(6)
24.555(5)
24.542(6)
24.487(4)
24.500(5)
24.784(6)
24.541(6)
10.555(2)
10.623(3)
10.624(1)
10.655(2)
10.720(8)
10.618(2)
10.581(3)
10.539(4)
10.519(4)
10.555(3)
10.604(3)
10.720(4)
10.525(3)
10.522(3)
10.525(4)
10.546(2)
10.502(4)
10.503(2)
10.520(5)
10.531(2)
10.526(4)
10.521(5)
10.661(5)
10.564(5)
10.579(2)
10.566(3)
10.554(2)
10.741(4)
10.609(5)
5450(1)
5522(2)
5518(1)
5559(2)
5605(4)
5525(2)
5481(3)
5486(3)
5448(3)
5487(2)
5507(3)
5641(3)
5473(2)
5479(2)
5454(3)
5478(2)
5432(3)
5445(2)
5475(3)
5481(2)
5464(3)
5451(3)
5611(3)
5516(3)
5518(2)
5487(2)
5486(2)
5714(3)
5533(3)
does not react
0.661(4)
0.68
3
4
5
6
7
8
9
10
11
12
fluorobenzene, C6H5-F
0.659(4)
0.650(8)
0.68
0.66
0.58
0.48
0.50
0.39
0.53
ca. 0.3
0.66
0.46
chlorobenzene, C6H5-Cl
bromobenzene, C6H5-Br
iodobenzene, C6H5-I
trifluoromethylbenzene, C6H5-CF3
trichloromethylbenzene, C6H5-CCl3
nitrobenzene, C6H5-NO2
benzaldehyde, C6H5-CHO
toluene, C6H5-CH3
0.474(5)
0.657(9)
0.386(8)
ethylbenzene, C6H5-C2H5
13
14
15
cyclopropylbenzene, C6H5-CH(CH2)2
propylbenzene, C6H5-C3H7
0.36(1)
0.34(1)
0.34(1)
0.40
0.37
0.39
tert-butylbenzene, C6H5-C(CH3)3
16
17
18
19
4-fluorotoluene, 4-F-C6H4-CH3
4-fluoroanisole, 4-F-C6H4-OCH3
o-xylene, o-CH3-C6H4-CH3
m-xylene, m-CH3-C6H4-CH3
0.413(8)
0.38(1)
0.62(1)
0.34(1)
0.48
0.41
0.63
0.41
20
p-xylene, p-CH3-C6H4-CH3
0.36(1)
0.41
0.66
21
22
23
o-dichlorobenzene, o-Cl-C6H4-Cl
hexafluorobenzene, C6F6
mesitylene, 1,3,5-(CH3)3-C6H3
ca. 0.4
a Unit cell parameters from powder patterns are given in normal type with those from the single-crystal measurement (if available) in italics. For
an explanation of methods A and B see the text. The experimental error in method B is ∼3%.
Å and connected to C13 at 2.89 Å by hydrogen bonding,
whereas the atom of the secondary orientation (19%), O25, is
coordinated more weakly to Cu at 2.88 Å but connected more
strongly to C13 at 2.77 Å.
along the channel direction so that each cavity is occupied by
one benzene molecule (Figure 4).
Inclusions with Benzene Derivatives. Data for inclusions
of the complex with benzene derivatives are summarized in
Table 3. These were prepared and analyzed by two independent
methods. In method A (isopiestic method),2a,8 the reactions take
place by gas-phase guest transport and result in a complete phase
transformation into the â-form (cf. Figure 5a,b). It should be
noted that reproducible results were obtained only with the
thoroughly purified complex, and two or three sorption-
desorption cycles often were necessary to avoid oversaturation
and wetting of the samples. This requirement probably arises
because of a side process by which R-[CuL2] dissolves in excess
liquid guest. Taking the process through a number of cycles
also is likely to produce crystals that are not too dissimilar in
size and therefore capable of a uniform response to the solvation
or desolvation process. The isopiestic method also served for
the determination of equilibrium compositions. Method B
involved crystallization of R-[CuL2] from neat guests. Unit cell
parameters were obtained from powder patterns in an atmo-
sphere of the corresponding guest. For inclusions that were
stable enough, unit cell parameters were also determined on a
single-crystal diffractometer. Inclusions of benzene and m- and
p-xylenes are examples of the most stable compounds in the
series. The majority of inclusions lost guest material in air,
revealing the type of decomposition that is typical of solid
solutions:9 the crystals cracked, retaining their habit and
transparency. Often it was even possible to observe a decrease
The remarkable advantage of the â- over the R-framework
is that all of the additional interactions possible involving the
methoxy oxygen indeed occur in the former material. The
polymer-like assembly results in a trigonal porous structure with
straight channels along the c direction (Figure 3). The inner
surface of each channel is formed by methyl and trifluoromethyl
groups (Figure 3 (bottom)). The channel consists of alternating
large and small cavities (Figures 4 and 6): the large cavity at
z/c ) 0.3-0.7 has two openings of more than 6.4 Å each, and
the small cavities have one wide part at z/c ) 0.8-1.2 and two
constrictions at 5.3 Å. When each cavity is filled with one guest
molecule, the guest-to-host molar ratio equals 2/3, a value that
is actually observed for the benzene inclusion (see Table 3).
Two orientations for the guest benzene molecule have been
found (Figure 3). The first orientation is almost perpendicular
to the channel direction in the large cavity, and there are two
such sites symmetrically located on either side of the cavity
center. Its low thermal parameters (see the Supporting Informa-
tion) suggest that this benzene is particularly tightly bound and
may be responsible for the high stability of the compound by
tightly blocking the channel. The second orientation corresponds
to a longitudinal displacement of the tilted guest into the small
cavities, and again there are two such sites symmetrically
displaced from the cavity center; the thermal parameters are
twice those of the first orientation. Since the sites overlap, the
cavity can be occupied by only one molecule. This and the
approximately 50% distribution between the perpendicular and
tilted orientations imply alternation of these two orientations
(8) (a) Angla, B. Ann. Chim. 1949, 4, 639-698; Chem. Abstr. 1950, 44,
3442c. (b) Schlenk, W. Justus Liebigs Ann. Chem. 1949, 565, 204-240;
Chem. Abstr. 1950, 44, 3900a.
(9) (a) Kemula, W.; Lipkowski, J.; Sybilska, D. Rocz. Chem. 1974, 48,
3-12. (b) Reference 18a, p 94.

Transforming Dense [CuL2] into a Microporous Framework
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4183
Figure 4. Possible variants of local ordering of benzene molecules in
the channels of â-[CuL2]‚2/3C6H6.
the fact that for compact, sterically simple guests the composi-
tions coincide within experimental error.
From Table 3, the organic guests may be divided into three
groups with regard to their ability to be included. First is the
group of small, relatively compact molecules that are included
with molar ratios close to the ideal value of 2/3. These are
benzene, its F, Cl, and Me derivatives, and the ortho isomers
of xylene and dichlorobenzene. The second group consists of
molecules that are small enough to be included, but that are
too long to fill all available cavities stoichiometrically. The
resulting inclusions are typical solid solutions with nonstoichio-
metric compositions that depend on the guest size and the P-T
conditions during synthesis. The final group consists of com-
pounds that are too large to be included: mesitylene is an
example of this and thus is a convenient choice as a nonincluded
solvent for the title complex.
Sorption of Gases, Phase Conversion Catalysis, and
Empty â-[CuL2]. In the course of our experimental work we
encountered an interesting phenomenon: the complex in its
R-form reacts with hydrocarbons that are gases under ambient
conditions. Reactions occur at room temperature and atmo-
spheric pressure and result in a complete transformation to the
â-form.10 With ∼0.5 g samples the reactions with neopentane
(bp +9.5 °C), n-butane (bp -0.5 °C), isobutylene (bp -7 °C),
and isobutane (bp -11.7 °C) were complete after several hours,
the rates correlating with boiling points and decreasing in the
given sequence. During the reaction, the samples revealed mass
increases of about 7 wt % due to the absorbed gas, the first
portions of which disappeared again quickly when the material
was removed from the gas stream. In air, the samples showed
a gradual return to the original R-form.
Figure 3. Molecular packing and mode of assembly in â-[CuL2]‚2/
3C6H6. Top: the host framework along the channel direction, showing
the included benzene molecules. For clarity, two independent orienta-
tions of the guest molecules are shown in the two channels; hydrogen
atoms are not shown. Bottom: schematic showing how the copper
chelate building blocks are linked to form the channel. The rectangles
represent the roughly square-planar ligation around the copper atoms
located at the center of the rectangles. Each copper chelate uses one
donor (O-Cu) and one acceptor (Cu-O) link (shown as dashed lines)
to connect to other building blocks forming the channel, and the other
donor and acceptor links connect to building blocks forming neighboring
channels.
These observations are exciting as they have bearing on the
potential of the host material for applications involving gases
where conventional host materials are not effective. In addition,
the transformation itself demonstrates an amazingly strong
affinity of the host for the channel form. Finally, the results
provide convincing evidence of the relative robustness of the
empty framework. Indeed, we have developed an elegant method
of preparing pure, empty â-[CuL2] by finding a suitable
in unit cell parameters on repeating single-crystal X-ray experi-
ments every 2 h. The least stable in the series were inclusions
of hexafluorobenzene and o-xylene. Guest-to-host ratios were
1
determined with H NMR analysis.
As one can see from Table 3, all inclusions are isostructural
and have a trigonal R cell. There is some difference in
composition for the same inclusions obtained by methods A
and B. Taking into account different synthetic conditions for
the methods, we suggest that method A defines the lower limit
of the equilibrium guest content, whereas method B gives the
maximum (but not necessarily the equilibrium) guest content
in the crystalline inclusions. This assumption is supported by
(10) This should be clearly distinguished from other types of processes
with participation of gaseous guests: (a) trapping of gas molecules into
the closed cages of clathrate hydrates, â-hydroquinone, Dianin’s com-
pound,11 and calixarenes.12 (b) Sorption into ready-made empty matrixes.2a,13
(c) Reactions with guest vapor at full saturation.14

4184 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
SoldatoV et al.
cis isomer, whereas in the â-form they are all trans isomers.
The phase change therefore depends on the isomerization of
half of the chelate building blocks and implies that at least two
coordination bonds around the Cu center must break. This
process is clearly affected by the presence of a sorbent, perhaps
acting as a catalyst. Typically, sorbents play a double role in
the inclusion compound formation process: stabilizing the
empty â-form (thermodynamic action) and catalyzing the cis-
trans isomerization necessary for phase transformation (kinetic
aspect). However, even if the sorbent does not stabilize the
â-lattice sufficiently as a guest molecule, apparently it can still
trigger the â-to-R conversion (see, e.g., eq 3) or assist in the
R-to-â transformation. This latter ability is demonstrated very
well by the following example. The R-form does not react with
paraffin oil upon wetting, and only small quantities of the
â-phase form in the course of prolonged grinding of the mixture.
At the same time, addition of a little mesitylene (which cannot
be included as guest) completes the reaction within minutes.
The results suggest that the processes involving the title material
may be effectively controlled by applying certain “switching”
substances, making the material a candidate for “smart” ap-
plications.15
Helium pycnometry provides further evidence of the kinetic
stability of the guest-free framework. The effective density found
for the empty â-[CuL2] exposed to He gas is relatively high at
1.604(5) g cm-1 (cf. 1.333 g cm-1 from powder diffraction),
which leads to a pore volume per unit cell (for He) of 925(15)
Å3, or 17%.16 From Table 3, one finds that this “effective”
volume may increase at least by up to 20% as a result of
framework expansion during guest inclusion. Assuming that the
pore space is a uniform channel, the average channel diameter
is 6.10(5) Å,16 similar to that in medium-pore zeolites such as
ZSM-12 and mordenite.17 Due to the properties outlined above,
â-[CuL2] can be classified as a novel metal-organic zeolite
mimic. Longer-term stabilization of the empty framework may
be possible by the introduction of a small quantity of a suitable
“pillaring” guest. At the same time, chemical modification of
the title complex may well lead to a series of similar structures
with different pore sizes and shapes.
Figure 5. Selected powder diffractograms (λ ) 1.7902 Å): (a)
R-[CuL2] after vacuum sublimation (sample A); (b) sample A after
saturation with benzene for 24 h (the mass increase corresponds to
0.67 mol of benzene); (c) sample A after treatment with methyl bromide
(2 h), followed by evacuation (3 h); the final mass equals that of the
original sample (sample B); (d, e) sample B after 2 and 20 days,
respectively.
“transforming” gas. Methyl bromide (bp +3.6 °C), with a
compact molecule of maximum van der Waals diameter of 5.6
Å, is easy to remove from the framework. After R-[CuL2]
(Figure 5, “sample A”) is treated with methyl bromide for 2 h
(eq 1) followed by 3 h of evacuation (eq 2), the product (“sample
B”) is the empty â-framework. The weight gain on transforming
to the methyl bromide-â-[CuL2] compound varied from 16.8%
to 19%, which amounts to about 0.9 mol of guest for each host
molecule. The compound quickly loses methyl bromide on
exposure to air, and exposure to vacuum leads to a quantitative
removal of the guest species as judged by the weight loss. X-ray
powder diffraction confirmed that the product was the â-form;
also the stability of the empty â-[CuL2] framework (Figure 5)
and He pycnometry confirmed the presence of void space
accessible to helium (see below). Some structural parameters
of â-[CuL2] are given in Table 1. The very low packing
coefficient of 49.8% is possible only because of sufficiently
strong intermolecular bonds that sustain the porous structure.
Kept in air in a closed container, the empty â-[CuL2] is stable
for some days, eventually returning again to the R-form (eq 3).
The rate of transformation depends on sample purity. For
instance, treatment with propane accelerates the reversion to
the R-form (eq 3).
It is not entirely clear whether the mechanism of conversion
of the R- to the â-form, which occurs both for the isopiestic
(11) Inclusion Compounds; Atwood, J. L., Davies, J. E. D., MacNicol,
D. D., Eds.; Academic Press: London, 1984. (a) Jeffrey, G. A.; Vol. 1, pp
135-190. (b) MacNicol, D. D.; Vol. 2, pp 1-46. (c) Davies, J. E. D.; Vol.
3, pp 37-68. (d) Davidson, D. W.; Ripmeester, J. A.; Vol. 3, pp 69-128.
(12) (a) Brouwer, E. B.; Enright, G. D.; Ripmeester, J. A. J. Am. Chem.
Soc. 1997, 119, 5404-5412. (b) Brouwer, E. B.; Enright, G. D.; Ripmeester,
J. A. Chem. Commun. 1997, 939-940.
(13) Manakov, A. Yu.; Lipkowski, J. J. Inclusion Phenom. 1997, 29,
41-55.
(14) Kaupp, G. In ComprehensiVe Supramolecular Chemistry; Davies,
J. E. D., Ripmeester, J. A., Eds.; Pergamon: Oxford, 1996; Vol. 8, p 398.
(15) Newnham, R. E. Mater. Res. Soc. Bull. 1997 (May), 22 (5), 20-
34.
The reversion is direct evidence that the empty â-form is
thermodynamically unstable with respect to the R-form. The
relative kinetic stability, however, is remarkable, and can be
explained as follows. According to the above structural data,
the R-form has half of the Cu chelate molecules present as the
(16) For details see the Experimental Section.
(17) Atlas of Zeolite Structure Types; Meier, W. M., Olsen, D. H., Eds.;
Butterworth-Heinemann: London, 1992.

Transforming Dense [CuL2] into a Microporous Framework
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4185
method and for the determination of composition and the
synthesis of the empty form as described above, is identical in
these two instances. It is likely that surface adsorption of guest
material occurs, followed by capillary condensation, dissolution
of the guest, and recrystallization, especially for guests that are
rather high boiling. However, this seems less certain for low-
boiling guests, especially light hydrocarbons in which the host
is essentially insoluble. If the process is seen as the reaction of
a gas with a disorganized surface (e.g., such as also takes place
for ice to give clathrates), the concepts of phase rebuilding14
might be more appropriate. Kaupp14 discusses various possibili-
ties for gas-surface reactions involving enclathration in light
of information gained from application of scanning microscopy.
Comparison with Other Classes of Inclusion Compounds.
The host framework is built up of metal-chelate building blocks
connected to each other by secondary coordination bonds. From
this viewpoint, the hosts are intermediate between Werner
hosts18 with their pure van der Waals interactions and typical
coordination polymers such as the host complexes with 4,4′-
dipyridyl19 or cyanide20 ligands. It is peculiar in that despite
the simplicity of the CuL2 building blocks, the mode of their
assembly is far from obvious (Figure 3 (bottom)). The “4 × 2”
mode of assembly (coordination by four strong intramolecular
and two weak intermolecular ligands, all with a single neutral
building block) is worthy of investigation as a new strategy for
the assembly of novel open frameworks. Note that coordination
polymers based on chelated Cu(II) centers have been described
elsewhere,21 and although no open channel frameworks were
observed among them, these systems exemplify possible direc-
tions in such design.
Figure 6. Characterization of the channel shape. The diameter of the
inscribed sphere centered at (0, 0, z/c) versus z/c for the host
frameworks: 1, â-[CuL2]‚2/3C6H6; 2, urea; 3, thiourea; 4, Werner
clathrate Mg(4-MePy)4(NCS)2. All are trigonal (hexagonal), with the
channel along the z-axis; c parameters are between 10.6 and 12.6 Å.
thiourea26 and that of the connected large cages typical of the
trigonal Werner clathrates22c (Figure 6).26 This results in two
qualitatively different types of inclusions: inherent nonstoichio-
metric ones such as classic inclusions of urea with hydrocarbons,
and inclusions with compositions near the maximum structural
stoichiometry. As one can see from Table 3, the â-[CuL2] lattice
is capable of some “adjusting” to a certain guest within a small
range. Comparison of the data for â-[CuL2] and the inclusions
with X-C6H5 (X ) H, F, Cl, Br, I) shows that elongation of
the guest in one dimension results in a slight expansion of the
framework at constant (maximum) stoichiometry. However,
when the substituent is larger (bromo- or iodobenzene), the guest
content drops sharply and the cell parameters of the framework
tend to approach those of the empty lattice.
Regarding symmetry, the â-[CuL2] framework belongs to a
very large group of trigonal (hexagonal) inclusion lattices, many
of which have the same space group and similar unit cell
dimensions and channel sizes. They may be found among the
Werner clathrates,22 porphyrin-based frameworks,23 Dianin’s
compound clathrates,24 alicyclic diols,2c “trigonal-symmetry”
organic hosts,25 etc.
There is however a difference between nonstoichiometric
inclusions of urea and â-[CuL2]. In urea inclusions guest species
are tightly packed, resulting in the appearance of an independent
crystallographic period for the guest subsystem which does not
coincide with the respective unit cell period of the urea host
framework.27a This was attributed to the thermodynamic and
kinetic instability of the empty urea framework.28 During
experiments with the â-[CuL2] inclusions we did not observe
extra reflections in the powder diffraction patterns that could
be assigned to independent “guest” periodicity. Most likely, the
void space in our nonstoichiometric compounds is filled with
disordered species, and if so, their inclusion into â-[CuL2]
resembles the sorption inside capillary tubes rather than the
formation of densely packed supramolecular architectures.
The void space in the material is intermediate in character
between that of true channel inclusions such as urea and
(18) (a) Lipkowski, J. In Inclusion Compounds; Atwood, J. L., Davies,
J. E. D., MacNicol, D. D., Eds.; Academic Press: London, 1984; Vol. 1,
pp 59-103. (b) Hanotier, J.; Radzitzki, P. Ibid. pp 105-134. (c) Lipkowski,
J. In ComprehensiVe Supramolecular Chemistry; MacNicol, D. D., Toda,
F., Bishop, R., Eds.; Pergamon: Oxford, 1996; Vol. 6, pp 691-714.
(19) (a) Robson, R.; Abrahams, B. F.; Batten, S. R.; Gable, R. W.;
Hoskins, B. F.; Liu, J. In Supramolecular Architecture; Bein, T., Ed.; ACS
Symposium Series 499; American Chemical Society: Washington, DC,
1992; pp 256-273. (b) Fujita, M.; Kwon, Y. J.; Washizu, S.; Ogura, K. J.
Am. Chem. Soc. 1994, 116, 1151-1152 and refs 2-4 therein. (c) Zaworotko,
M. J. Chem. Soc. ReV. 1994, 23, 283-288. (d) Subramanian, S.; Zaworotko,
M. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 2127-2129.
(20) Iwamoto, T. In ComprehensiVe Supramolecular Chemistry; Mac-
Nicol, D. D.; Toda, F.; Bishop, R., Eds.; Pergamon: Oxford, 1996; Vol. 6,
pp 643-690.
(21) (a) Saalfrank, R. W.; Danion, D.; Hampel, F.; Hassa, J.; Struck,
O.; Toupet, L. Chem. Ber. 1994, 127, 1283-1286. (b) Saalfrank, R. W.;
Struck, O. Chem. Mater. 1994, 6, 1432-1436. (c) Saalfrank, R. W.; Harbig,
R.; Struck, O.; Peters, E.-M.; Peters, K.; Schnering, H. G. Z. Naturforsch.
B 1996, 51, 399-408 (in German). (d) Saalfrank, R. W.; Harbig, R.; Struck,
O.; Hampel, F.; Peters, E.-M.; Peters, K.; Schnering, H. G. Z. Naturforsch.
B 1997, 52, 125-134 (in German).
(22) Pervukhina, N. V.; Podberezskaya, N. V.; Davydova, I. V.; Kislykh,
N. V.; Dyadin, Yu. A. J. Inclusion Phenom. 1992, 13, 9-16. (b) Lipkowski,
J.; Soldatov, D. V.; Kislykh, N. V.; Pervukhina, N. V.; Dyadin, Yu. A. J.
Inclusion Phenom. 1994, 17, 305-316. (c) Lipkowski, J.; Soldatov, D. V.
J. Inclusion Phenom. 1994, 18, 317-329.
(23) Krupitsky, H.; Stein, Z.; Goldberg, I. J. Inclusion Phenom. 1994,
18, 177-192.
(24) (a) Ref 11b, 15. (b) Abriel, W.; Bois, A.; Zakrzewski, M.; White,
M. A. Can. J. Chem. 1990, 68, 1352-1356.
(25) (a) Gerdil, R. In ComprehensiVe Supramolecular Chemistry; Mac-
Nicol, D. D., Toda, F., Bishop, R., Eds.; Pergamon: Oxford, 1996; Vol. 6,
pp 240, 243-252. (b) Allegra, G.; Farina, M.; Immirzi, A.; Colombo, A.;
Rossi, U.; Broggi, R.; Natta, G. J. Chem. Soc. B 1967, 1020-1028. (c)
Freer, A. A.; MacNicol, D. D.; Mallinson, P. R.; Vallance, I. Tetrahedron
Lett. 1992, 33, 261-264.
(26) For channel geometry in urea and thiourea see also: (a) George,
A. R.; Harris, K. D. M. J. Mol. Graphics 1995, 13, 138-141. For Figure
3 X-ray data on clathrates of urea with n-hexadecane,27a thiourea with
adamantane,27b and [Mg(4-MePy)4(NCS)2] with 4-MePy and water 22c were
used.
(27) (a) Harris, K. D. M.; Thomas, J. M. J. Chem. Soc., Faraday Trans.
1990, 86, 2985-2996. (b) Gopal, R.; Robertson, B. E.; Rutherford, J. S.
Acta Crystallogr., Sect. C 1989, 45, 257-259.
(28) (a) Belosludov, V. R.; Lavrentiev, M. Yu.; Dyadin, Yu. A. J.
Inclusion Phenom. 1991, 10, 399-422. (b) Dyadin, Yu. A.; Bondaryuk, I.
V.; Aladko, L. S. J. Struct. Chem. 1995, 36, 995-1045.

4186 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
SoldatoV et al.
It is interesting to compare the title material with others that
show similar zeolite-like behavior. Although a number of such
systems are known,2 currently only one has been studied in detail
and continues to attract interest because of its high selectivity.
This is the so-called â-phase of [Ni(4-MePy)4(NCS)2] and some
related complexes.18a,c It is formed by van der Waals packing
of coordinatively saturated units, and therefore it is rather
flexible, expanding by as much as 9% during the inclusion
process.29 The sorption behavior of the material was studied
and explained well in relation to its structure,13,18a and attractive
applications in chromatography have been suggested.30 At the
same time, a number of properties of the material, such as the
thermal dissociation of the complex or the absence of changes
in chemical bonding during the R-â-phase interconversions,
place limits on its general application. In this sense, [CuL2] with
the remarkable stability of the molecule and the possibility of
catalytic switching between polymorphic forms gives real
possibilities for its potential as a useful material.
Versatility, Chemical Stability, and Regeneration of the
Host Material. The wide variety of guests (∼100 guests have
now been checked in addition to those given in Table 3) includes
normal, branched, and cyclic hydrocarbons, their F, Cl, Br, I,
and S derivatives, alcohols, ethers, esters, ethers of mineral acids,
acid anhydrides, aldehydes, ketones, and compounds with nitro,
amido, and nitrile functions.31 The material is thus versatile and
is also stable in the presence of many classes of organic
compounds. Inclusion of monomers such as isoprenes or
lactones has potential for directed polymer syntheses inside the
matrix. We also note the marked selectivity of inclusion. For
example, the difference in stability and composition for inclu-
sions of xylene isomers, as mentioned above, may well be useful
for separation and purification processes.
systems by adding solvent or changing pH may be seen as an
advantage in the design of future molecular-scale processes. This
study demonstrates the extremely high potential for molecular
complexes; the material presented here reveals novel properties
such as the remarkable stability of the empty inclusion
framework, catalytically induced switching between dense and
open forms, and transformation from the dense to the porous
form in an atmosphere of aliphatic hydrocarbon gases. At the
same time, the mode of molecular assembly resulting in the
open form of the title complex is neither obvious nor trivial. It
is hoped that detailed analysis of such systems may well lead
to the formulation of new approaches to supramolecular
chemistry and crystal engineering for the creation of a new
generation of novel materials.
Experimental Section
General Information. Reagent or analytical grade chemicals were
used. For step (ii) of the HL synthesis all reagents and solvents were
purified and dried by conventional methods. GLC analyses were
performed on a Khrom-5 chromatograph with a flame ionization
detector and a 360 × 0.4 cm column with 5% SE-30 on Inerton AW
with nitrogen as the carrier gas. Elemental analyses were carried out
at the Institute of Organic Chemistry, Novosibirsk. IR spectra were
recorded on a BIO-RAD FTS-40A spectrometer. 1H NMR spectra were
obtained with a Bruker DRX-400 instrument and analyzed with the
XWIN-NMR 2.0 program package.
CF3COCH2COC(OMe)Me2 ()HL). This diketone was synthesized
according to the scheme
Compounds with strong donor groups such as amines, or
compounds with oxygen-containing functions capable of coor-
dinating to copper centers, cause the loss of the capacity for
inclusion of the host material. In many cases this type of reaction
is easily reversed and may also be used for process control by
a temporary “switching off” of the inclusion properties.
Although the synthesis of the host material requires consider-
able effort, the material itself is remarkably convenient, as it is
completely recoverable. Unless highly reactive guest species
such as aldehydes or weakly volatile guests such as iodobenzene
are to be used, recovery can be quantitative. The regeneration
of the pure host is based on two or more cycles of two processes.
First is a recrystallization from hexane/benzene, where the
former is the highly preferred guest and the latter is added to
increase the solubility of the complex and to keep impurities in
solution. The second process is a vacuum sublimation that makes
it possible to separate the pure host from the volatile guest and
nonvolatile impurities. In the course of this work, the loss of
material after about 40 regenerations did not exceed 10%. Some
of the coordination properties mentioned above might make even
more effective recovery possible.
Step (i), adapted from Merz’s method,32 involved methylation of
commercially available ethynyldimethylcarbinol; alternatively, reaction
between acetylene and acetone followed by methylation of resulting
dimethylcarbinol in situ was used.33 Steps (ii),34 (iii), and (iv)35 were
reported before without details. (i) HCtCC(OH)Me2 (49 mL, 0.5 mol;
Aldrich), tetrabutylammonium iodide (1 g), and 50% aqueous NaOH
(104 g, 1.3 mol) were equilibrated by stirring for 15 min. Then dimethyl
sulfate (57 mL, 0.6 mol) was added dropwise for 1 h, and the
temperature of the mixture was kept below 40 °C. Next, this was stirred
vigorously (mechanically) for 3 h, and then for 30 min after adding
concentrated aqueous NH3 (10 mL). After the addition of excess water,
the organic phase was separated, washed with water, and dried over
MgSO4. A little xylene was used to avoid losses during the manipula-
tions. Distillation at 80-82 °C gave 44 g (90%) of pure product, HCt
Conclusion
20
CC(OMe)Me2 (98.1) [registry no. 13994-57-5, supplied by author], nD
Molecular metal complexes are very attractive as building
units for supramolecular architectures. This is because a wide
possible variability of the units makes it possible to provide
delicate control over selectivity, porosity, and sorption capability
of these host materials. On the other hand, the simplicity of
initiating self-assembly or the reversible destruction of such
1.4005. (ii) Into a flask equipped with a good mechanical stirrer was
(32) Merz, A. Angew. Chem., Int. Ed. Engl. 1973, 12, 846-847.
(33) Sokolov, I. E.; Kotlyarevskii, I. L. IzV. Akad. Nauk SSSR, Ser. Khim.
1980, 202-203 (in Russian); Chem. Abstr. 1980, 92, 214840x.
(34) Shergina, S. I.; Sokolov, I. E.; Zanina, A. S. IzV. Akad. Nauk SSSR,
Ser. Khim. 1992, 158-160 (in Russian); Chem. Abstr. 1992, 116, 255138
g.
(29) Lipkowski, J.; Majchrzak, S. Rocz. Chem. 1977, 49, 1655-1660.
(30) Kemula, W.; Sybilska, D. Nature 1960, 185, 237-238.
(31) The details of that study will be the subject of our next paper.
(35) Sokolov, I. E.; Shergina, S. I.; Zanina, A. S. IzV. Akad. Nauk SSSR,
Ser. Khim. 1992, 1390-1392 (in Russian); Chem. Abstr. 1993, 118,
124090d.

Transforming Dense [CuL2] into a Microporous Framework
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4187
suspended powdered CuCl (20 g, 0,2 mol) in dry benzene (150 mL)
under flowing argon. Triethylamine (30.3 g, 0.3 mol) was added, and
the mixture was stirrred under argon for 10 min. HCtCC(OMe)Me2
(19.6 g, 0.2 mol) was then added, and CF3COCl (28 g, 0.2 mol) was
fed to the mixture instead of argon. (The latter was prepared from
benzoyl chloride and trifluoroacetic acid.36) The reaction mixture was
stirred vigorously during the addition and for 2 h afterward, and its
temperature was kept below 40 °C. Finally, the resulting solution was
separated from the sticky solid, and the rest of the product was extracted
from the solid with three portions of 50 mL of benzene. The combined
volumes of benzene solution were washed three times with 5% HCl
(75 mL each) and then with water to neutral pH and dried over MgSO4.
After benzene was removed on a rotary evaporator, vacuum distillation
gave 24 g (60%) of the product, CF3COCtCC(OMe)Me2: bp 62 °C
5), the treatment continued for another hour (2 h total). Then the sample
was allowed to decompose in air (0.5 h) and in a vacuum (3 h, up to
5 × 10-4 Torr) until it returned to its original mass.
Single-Crystal Diffraction Analysis. Data were collected on a
Siemens SMART CCD diffractometer with Mo KR radiation (λ )
0.7107 Å) and a graphite monochromator. Unit cell parameters were
measured using all collected data. Absorption corrections were made
using SADABS-32. Solution and refinement were performed using
direct methods (SHELXS-86) and full-matrix least squares on F2
(SHELXL-93) using all positive data. All non-hydrogen atoms of
complex molecules were refined anisotropically, whereas hydrogen
atoms were fixed in the calculated positions with temperature factors
1.2 times exceeding those of adjacent carbon atoms. More experimental
details are given in Table 1 and in the Supporting Information.
Single crystals of R-[CuL2] were prepared by slow sublimation. For
this, about 400 mg of pure complex was sealed under vacuum inside
a 10 cm long Pyrex tube. One end of the tube was held at room
temperature while the temperature of the end containing the complex
was heated gradually from 150 to 200 °C over a 24 h period. The
crystals that formed in the cold end were shapeless but quite suitable
for crystallographic measurements.
(42 Torr); nD21 1.3880; IR (neat) 2218 (ν CtC), 1742 (ν CdO) cm-1
.
Anal. Calcd for C8H9F3O2 (194.2): C, 49.48; H, 4.67; F, 29.36.
Found: C, 49.80; H, 4.75; F, 28.92. (iii, iv) The mixture of CF3COCt
CC(OMe)Me2 (19.5 g, 0.1 mol) and aniline (11 g, 0.12 mol) in 120
mL of dioxane was stirred at 60 °C for 3 h and, after addition of 25
mL of 15% aqueous HCl, for a further 3 h. After cooling, the organic
phase was extracted with three portions of toluene (50 mL each), washed
with water, and dried over CaCl2. Vacuum distillation yielded 12 g
(55%) of the product, CF3COCH2COC(OMe)Me2 ()HL): bp 80 °C
Prismatic crystals of â-[CuL2]‚xC6H6 (x ) 2/3) were prepared directly
by crystallization. No precautions were needed to prevent decomposition
during data collection. Because of strong disordering of the guest
subsystem, benzene molecules were assigned their ideal geometry and
were refined isotropically with the same thermal factor for all atoms
of each orientation. In the course of the refinement the value of
0.70(1) was first obtained for the mole ratio x. In the last cycles however
it was fixed at 2/3 exactly. For this situation, site occupancy factors of
52.4(7)% and 47.6(7)% were obtained for the perpendicular and tilted
orientations, respectively.
A single-crystal diffraction measurement of the unit cell was also
performed on the CCD diffractometer. The actual values were derived
from 100-200 reflections randomly selected at different angles.
Powder Diffraction Measurements. The unit cell dimensions of
both complex forms and all inclusions, obtained by method B, were
determined from powder patterns recorded on a Rigaku Geigerflex
diffractometer (Co KR radiation, λ ) 1.7902 Å). The powder patterns
were cross-checked with those calculated from the single-crystal results.
Inclusions were studied in an atmosphere of the corresponding guest.
NaCl was added for the zero angle correction. Least-squares refinements
were carried out using 15-20 reflections in a 5-40° 2θ range. The
data are summarized in Table 3.
Analysis of Packing and Empty Space. Packing coefficients37 and
the geometry of the channel and sections for â-[CuL2]‚2/3C6H6 were
calculated on the basis of the crystallographic data using the CLAT
software package and the Zefirov-Zorkii system of van der Waals
radii (C, 1.71; H, 1.16; Cu, 1.40; F, 1.35; O, 1.29 Å).38 Changes of
molecular geometry due to the disordering of fragments were found
not to exceed declared experimental errors. For calculating the packing
coefficient for empty â-[CuL2], the volume of the complex molecule
as it was found in the benzene inclusion was used.
20
1
(35 Torr); nD 1.4080; H NMR (CDCl3) δ 6.39 (∼1H, enol H-O),
3.23 (3H, OMe), 1.37 (6H, Me), 1.31 (∼1H, enol HCdC) ppm. Anal.
Calcd for C8H11F3O3 (212.2): C, 45.29; H, 5.23; F, 26.86. Found: C,
45.21; H, 5.16; F, 26.21.
r-[CuL2]. To the solution of HL (2.2 g, 10 mmol) and Cu(NO3)2‚
2.5H2O (1.2 g, 5.2 mmol) in ethanol (15 mL) was added the solution
of NaOH (0.4 g, 10 mmol) in 10 mL of water/ethanol (1:1) for 5 min.
After the solution was mixed for 30 min at 40 °C, water (100 mL) was
added. The resulting green precipitate was washed with water and dried
in air in a warm place until it turned blue. Sublimation (160-180 °C,
10-3 Torr) resulted in 2 g (80%) of powder, or a fine crystalline blue
product. X-ray powder diffraction was used for product identification.
The presence of the â-modification was revealed by its strong reflection
(-1,2,0) (see Figure 5). Depending on the degree and the nature of
contamination, the product could be improved by holding it in a warm
place, by repeated sublimation, or by recrystallization from benzene/
hexane (1:1) followed by a double sublimation.
Synthesis and Analysis of Clathrates by the Isopiestic Method
(Method A). For this method [CuL2] was thoroughly purified by two
cycles of recrystallization and double sublimation as described above.
The product was then recrystallized from mesitylene (Aldrich, 99%),
dried, and ground into a fine powder. Samples of 100-200 mg were
placed in closed vessels along with a saturated solution of the complex
in the corresponding guest solvent. After the sample reached constant
mass, it was allowed to decompose in air for some time, and the
experiment was repeated until reproducible results were obtained. Two
independent determinations were made. One cycle took from several
hours to several weeks depending on the guest. Compositions x of the
resulting â-[CuL2]‚xG inclusions were calculated from the formula x
) (∆m/mo)(MH/MG), where ∆m and mo are the mass increase and
starting sample mass and MH and MG are host and guest molecular
masses, respectively.
Helium Pycnometry. A sample of 400 mg of â-[CuL2] was used
for the experiment (AccuPyc 1330 pycnometer); three determinations
were made. The pore volume was obtained from
Synthesis and Composition of Crystalline Clathrates (Method
B). [CuL2] is quite soluble in all aromatics. Cooling of warm green
solutions gave blue crystals of the inclusions. They were of different
habit depending on the guest used, usually isometric prisms or blocks
but sometimes needles (with o-dichloro- and trichloromethylbenzenes).
ZM × 1024
Vemp(He) ) Vu.cell
-
) 925(15) A3
(5)
d(He)NA
(He)
1
where Vemp
is the volume of â-[CuL2] accessible to He (Å3), Vu.cell
Compositions were obtained from H NMR spectra of their 0.05 M
) 5450(1) Å3 is the unit cell volume, Z ) 9 is the number of molecules
per unit cell, M ) 485.9 g mol-1 is the molar mass of the complex,
solutions in CDCl3. Host protons appeared in the spectra as broad bands
at 1.75-1.97 (>CH), 3.35 (Me), and 5.34 (OMe) ppm, whereas guest
resonances were sharp. The integration errors were ∼3%.
1024 is the conversion factor of centimeters to angstroms, d(He)
)
1.604(5) g cm-1 (at 30.3 °C) is the value from the pycnometry
experiment, and NA ) 6.02 × 1023 mol-1 is Avogadro’s number.
(Empty) â-[CuL2]. Ground R-[CuL2] (500 mg) and analogous
control samples were placed in an atmosphere of methyl bromide (0.1
L/min flow of gas) at 20 °C. After the control samples revealed the
full transformation to the â-form (powder diffraction control, cf. Figure
(37) Kitaigorodsky, A. I. Molecular Crystals and Molecules; Academic
Press: New York, 1973; p 18.
(38) (a) Grachev, E. V.; Dyadin, Yu. A.; Lipkowski, J. J. Struct. Chem.
1995, 36, 876-879. (b) Zefirov, Yu. V.; Zorkii, P. M. Russ. Chem. ReV.
1995, 64, 415-428.
(36) Henne, A. L.; Alm, R. M.; Smook, M. J. Am. Chem. Soc. 1948, 70,
1968.

4188 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
SoldatoV et al.
(He)
The average channel diameter dav
formula
was determined from the
tions on the crystallographic studies, Drs. C. I. Ratcliffe and G.
Shimizu for helpful discussions, Dr. A. Yu. Manakov for useful
suggestions on analytical procedures, E. V. Grachev for the
CLAT software package, and Dr. T. V. Rodionova and N. V.
Udachina for facilitating contact between the working groups.
Vemp(He) ) 3cπ(dav(He)/2)2
(6)
where 3 is the number of channels going through the unit cell and c )
10.555(2) Å is the period of the unit cell along the channel axis.
Supporting Information Available: Full tables of crystal
data and structural refinement, bond lengths and angles,
anisotropic displacement parameters, and hydrogen atom pa-
rameters for R-[CuL2] and â-[CuL2]‚2/3C6H6 (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. This work was supported by the NATO
Science Program through a Research Visit Grant for D.V.S.
(SRG 951535). We thank Professor I. K. Igumenov for
suggesting the possibility of using the title compound as a host
material, Professor M. S. Shvartsberg for continuing support
of this work, Drs. K. A. Udachin and G. Enright for consulta-
JA981443U