Dicarboxylate-bridged (Mo2)n (n = 2, 3, 4) paddle-wheel complexes: Potential intermediate building blocks for metal-organic frameworks

Article abstract of DOI:10.1039/c1dt11249j

The treatment of the dimeric paddle-wheel (PW) compound [Mo ₂(NCCH₃)₁₀][BF₄]₄1 with oxalic acid (0.5 equiv.), 1,1-cyclobutanedicarboxylic acid (1 equiv.), 5-hydroxyisophthalic acid (1 equiv.) (m-bdc-OH) or 2,3,5,6- tetrafluoroterephthalic acid (0.5 or 1 equiv.) leads to the formation of macromolecular dicarboxylate-linked (Mo₂)_n entities (n = 2, 3, 4). The structure of the compounds depends on the length and geometry of the organic linkers. In the case of oxalic acid, the dimeric compound [(CH ₃CN)₈Mo₂(OOC-COO)Mo₂(NCCH ₃)₈][BF₄]₆2 is formed selectively, whereas the use of 2,3,5,6-tetrafluoroterephthalic acid affords the square-shaped complex [(CH₃CN)₆Mo₂(OOC-C ₆F₄-COO)]₄[BF₄]₈3. Bent linkers with a bridging angle of 109° and 120°, respectively, lead to the formation of the molecular loop [(CH₃CN)₆Mo ₂(OOC-C₄H₆-COO)]₂[BF ₄]₄4 and the bowl-shaped molecular triangle [(CH ₃CN)₆Mo₂(m-bdc-OH)]₃[BF ₄]₆5. All complexes are characterised by X-ray single crystal diffraction, NMR (¹H, ¹¹B, ¹³C and ¹⁹F) and UV-Vis spectroscopy.

Full text of DOI:10.1039/c1dt11249j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 11490
www.rsc.org/dalton
PAPER
Dicarboxylate-bridged (Mo2)n (n = 2, 3, 4) paddle-wheel complexes: potential
intermediate building blocks for metal–organic frameworks†
Mathias Ko¨berl,a Mirza Cokoja,b Bettina Bechlars,b Eberhardt Herdtweckb and Fritz E. Ku¨hn*a,b
Received 30th June 2011, Accepted 19th August 2011
DOI: 10.1039/c1dt11249j
The treatment of the dimeric paddle-wheel (PW) compound [Mo2(NCCH3)10][BF4]4 1 with oxalic
acid (0.5 equiv.), 1,1-cyclobutanedicarboxylic acid (1 equiv.), 5-hydroxyisophthalic acid (1 equiv.)
(m-bdc–OH) or 2,3,5,6-tetrafluoroterephthalic acid (0.5 or 1 equiv.) leads to the formation of
macromolecular dicarboxylate-linked (Mo2)n entities (n = 2, 3, 4). The structure of the compounds
depends on the length and geometry of the organic linkers. In the case of oxalic acid, the dimeric
compound [(CH3CN)8Mo2(OOC–COO)Mo2(NCCH3)8][BF4]6 2 is formed selectively, whereas the use
of 2,3,5,6-tetrafluoroterephthalic acid affords the square-shaped complex [(CH3CN)6Mo2(OOC–C6F4–
COO)]4[BF4]8 3. Bent linkers with a bridging angle of 109◦ and 120◦, respectively, lead to the formation
of the molecular loop [(CH3CN)6Mo2(OOC–C4H6–COO)]2[BF4]4 4 and the bowl-shaped molecular
triangle [(CH3CN)6Mo2(m-bdc–OH)]3[BF4]6 5. All complexes are characterised by X-ray single crystal
diffraction, NMR (1H, 11B, 13C and 19F) and UV-Vis spectroscopy.
bound bridging ligands, such as N¢N-di-p-anisylformamidinate
(DAniF). The remaining equatorial positions are occupied by
solvent molecules and thus are available for the substitution by
other ligands. The character of these accessible coordination
sites determines the geometry of the product: a PW complex
with two available, cis-oriented sites leads—dependent on the
linker—either to dimeric molecular loops, or to tetrameric squares.
The advantage of such precursors is their templating character,
allowing the controlled design of discrete metal–organic arrays.
The disadvantage, however, is the low reactivity of the product
for further reaction with dicarboxylates to form larger networks.
The synthesis of such coordination oligomers or polymers requires
PW precursors which only exhibit labile solvent ligands, such as
[Mo2(NCCH3)10][BF4]4.20 To date, there is only one report on
the use of this compound for the synthesis of dicarboxylate-
linked PW complexes. In 1997, McCann et al. investigated the
reactivity of [Mo2(NCCH3)10][BF4]4 with aliphatic dicarboxylic
acids to create the first molecular loop with dimolybdenum units.21
Yet, the geometry of this molecule, predetermined by the flexible
dicarboxylate ligand, does not allow the formation of ordered
networks.
Previously, we reported the synthesis of one-dimensional metal–
organic complexes22 and polymers23 containing Mo2 and Rh2 PW
complexes which are connected by various difunctional linkers
via the axial coordination sites of the M2 cores. This report
focuses on the selective synthesis of molecular dimers, triangles
and squares containing Mo2 PW units bridged by linear or bent
dicarboxylate ligands from [Mo2(NCCH3)10][BF4]4 as intermediate
building blocks for three-dimensional frameworks. Unlike Cot-
ton’s structural congeners, the complexes described here should
Introduction
In the last decade, a large number of macromolecular coordination
compounds and metal–organic frameworks (MOFs) containing
dinuclear paddle-wheel (PW) units connected by dicarboxylate
linkers have been synthesised.1,2 This can largely be ascribed to
the unique coordination behaviour of PW complexes, exhibiting
an octahedral coordination environment with four bridging equa-
torial ligands, e.g. carboxylates, and one axial ligand per metal
(usually solvent ligands). The angles between two cis-oriented
equatorial ligands are usually 90◦, which is why the linkage of PW
units with dicarboxylates leads to well-ordered lattice structures.
Thus, PW secondary building units have become important for
the construction of larger coordination networks.3–7 Yet, it has so
far not been attempted to synthesise coordination polymers from
molecular PW entities. The group of Cotton performed pioneering
work in creating oligomeric structures such as dimers,8,9 molecular
loops,10 triangles (Mo,11 Rh,12–14 Re15), propellers16 and squares
(Mo,12,17 Ru,18 Rh19) by connecting PW complexes with various
dicarboxylate linkers to form dimeric molecules, triangular or
square arrays. In Cotton’s reports, at least one equatorial coordi-
nation site (per PW metal) at the M2 centre is blocked by strongly
aMolecular Catalysis, Catalysis Research Center, Technische Universita¨t
Mu¨nchen, Ernst-Otto-Fischer-Strasse 1, D-85747, Garching, Germany.
E-mail: fritz.kuehn@ch.tum.de; Tel: +49 89 289 13081
bChair of Inorganic Chemistry, Catalysis Research Center, Technische
Universita¨t Mu¨nchen, Ernst-Otto-Fischer-Strasse 1, D-85747, Garching,
Germany
† Electronic supplementary information (ESI) available. CCDC reference
numbers 830421 and 830422. For ESI and crystallographic data in CIF or
other electronic format, see DOI: 10.1039/c1dt11249j
11490 | Dalton Trans., 2011, 40, 11490–11496
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
exhibit solvent molecules in the equatorial coordination sites of
the PW core, allowing further controlled substitution and thus a
framework growth along the coordination sphere.
Synthesis of [(CH3CN)6Mo2(OOC–C4H6–COO)]2[BF4]4 (4)
solution of equiv. [Mo2(NCCH3)10][BF4]4 (117 mg,
A
1
0.123 mmol) and 1 equiv. 1,1-cyclobutanedicarboxylic acid
(17.8 mg, 0.123 mmol) in 10 mL acetonitrile is stirred for 24
h at room temperature, whereupon the colour of the solution
changes from blue to purple. The solution is layered by a solution
of hexane/diethyl ether = 4 : 1 to yield pink crystals within 72 h.
The crystalline product is collected and washed with diethyl ether
(3 ¥ 4 mL) and hexane (3 ¥ 4 mL). Upon drying in vacuo, 47
mg (51% yield, 0.031 mmol) of pink crystals of 4 are isolated.
It is possible to increase the yield to 75% by precipitating 4
Experimental
All experiments were carried out under a dry nitrogen at-
mosphere using standard Schlenk techniques. The solvents
were dried and distilled under nitrogen following conventional
methods. Chemicals were purchased from Aldrich (tetrafluo-
roterephthalic acid, 97%), Alfa Aesar (oxalic acid, 98% and
1,1-cyclobutanedicarboxylic acid, 99%) and Acros Organics (5-
hydroxyisophthalic acid, 99%) and used as received. The precursor
[Mo2(NCCH3)10][BF4]4 1 was prepared according to literature
procedures.20 The NMR spectra were recorded on a Bruker Avance
DPX 400 spectrometer (1H: 400.13 Hz, 11B: 128.38 Hz, 13C: 100.61
Hz and 19F: 376.5 Hz). UV-Vis measurements were performed on
a JASCO photometer V-550.
1
as a non-crystalline solid from the initial reaction solution. H
NMR (d (ppm), in CD3CN): 3.35 (t, CH2, 8 H), 2.37 (m, CH2,
4 H) and 1.95 (s, CH3CN, ca. 30 H). 11B NMR (d (ppm), in
CD3CN): 7.25 (s, BF4 ). 13C NMR (d (ppm), in CD3CN): 187.9
-
(COO-), 147.6 (Cquat), 60.0 (CH2) and 32.7 (CH2). 19F NMR (d
(ppm), in CD3CN): -146.4 (s, BF4-) and -146.5 (s, BF4-). Anal.
Calcd. for C26H33B4F16Mo4N7O8 = [(CH3CN)2.5Mo2(OOC–C4H6–
COO)]2[BF4]4: C, 23.97; H, 2.55; N, 7.53. Found: C, 23.90; H, 2.71;
N, 7.16. UV-Vis (CH3CN): lmax = 533 nm.
Synthesis of [(CH3CN)8Mo2(OOC–COO)Mo2(NCCH3)8][BF4]6
(2)
Synthesis of [(CH3CN)6Mo2(m-bdc–OH)]3[BF4]6 (5)
An acetonitrile solution (10 mL) of 1 equiv. [Mo2(NCCH3)10][BF4]4
(102 mg, 0.108 mmol) and 1 equiv. 5-hydroxyisophthalic acid
(m-bdc–OH) (19.6 mg, 0.108 mmol) is stirred for 24 h at room
temperature, whereupon the colour of the solution changes from
blue to red. The concentrated solution is layered by a solution of
hexane/diethyl ether = 4 : 1 to yield red crystals within 4 d. The
crystalline product is collected and washed with diethyl ether (3 ¥
3 mL) and hexane (3 ¥ 3 mL). Upon drying under vacuum, 41
mg (50% yield, 0.018 mmol) of red solid 5 are isolated. 1H NMR
(d (ppm), in CD3CN): 8.66 (s, OH), 7.99 (s, CH, 2 H), 7.82 (s,
CH, 1 H) and 1.96 (s, CH3CN, ca. 10 H). 11B NMR (d (ppm), in
An acetonitrile solution (17 mL) of 1 equiv. [Mo2(NCCH3)10][BF4]4
(150 mg, 0.158 mmol) and 0.5 equiv. oxalic acid (7.1 mg,
0.079 mmol) is stirred for 20 h at 60 C to yield a red solution.
◦
After solvent removal in vacuo, the residue is washed with diethyl
ether (3 ¥ 6 mL) and pentane (3 ¥ 6 mL). Upon drying under
vacuum, 122 mg (94% yield, 0.074 mmol) of compound 2 are
isolated as a red solid. By slow diffusion of pentane into a
concentrated acetonitrile solution of 2, red crystals are obtained
within one week. 1H NMR (d (ppm), in CD3CN): 1.95 (s, CH3CN).
11B NMR (d (ppm), in CD3CN): 7.39 (s, BF4 ). 19F NMR (d
-
(ppm), in CD3CN): -145.59 (s, BF4-) and -145.64 (s, BF4-).
Anal. Calcd. for C22H30B6F24Mo4N10O4 = [(CH3CN)3Mo2(OOC–
COO)Mo2(NCCH3)3][BF4]6: C, 18.83; H, 2.15; N, 9.98. Found: C,
18.87; H, 2.20; N, 9.59. UV-Vis (CH3CN): lmax = 424 nm.
CD3CN): 7.13 (s, BF4 ). 13C NMR (d (ppm), in CD3CN): 181.4
-
(COO-), 159.0 (COH), 147.5 (Cquat) and 132.6 (CH). 19F NMR
(d (ppm), in CD3CN): -146.68 (s, BF4-) and -146.74 (s, BF4-).
Anal. Calcd. for C44H42B6F24Mo6N10O15 = [(CH3CN)3.33Mo2(m-
bdc-OH)]3[BF4]6: C, 25.81; H, 2.07; N, 6.84. Found: C, 25.88; H,
2.21; N, 7.17. UV-Vis (CH3CN): lmax = 416 nm.
Synthesis of [(CH3CN)6Mo2(OOC–C6F4–COO)]4[BF4]8 (3)
A
solution of 1 equiv. [Mo2(NCCH3)10][BF4]4 (986 mg,
Crystallographic data
1.038 mmol) and 1.1 equiv. perfluoroterephthalic acid (271 mg,
1.138 mmol) in 90 mL acetonitrile is stirred for 20 h at 60 ◦C.
The colour of the solution changes from blue to red. After
solvent removal in vacuo, a red solid is isolated and washed with
diethyl ether (4 ¥ 10 mL) until the filtrate is clear to remove any
residual HBF4·Et2O. The solid is further washed with pentane
(3 ¥ 10 mL). Upon drying under vacuum, 721 mg (82% yield,
0.213 mmol) of 3 are obtained. Compound 3 can be further
purified by crystallisation by overlaying a concentrated solution
with pentane. 1H NMR (d (ppm), in CD3CN): 1.95 (s, CH3CN).
Single crystal X-ray structure determinations: 2: red prism,
C50H72B6F24Mo4N24O4, Mr = 1977.94; monoclinic, space group
˚
P21/n (No. 14), a = 14.9896(4), b = 18.3686(5), c = 15.5071(4) A, b =
◦
3
˚
˚
92.7229(14) , V = 4264.9(2) A , Z = 2, l(Mo-Ka) = 0.71073 A, m =
0.680 mm-1, rcalcd = 1.540 g cm-3, T = 123(1) K, F(000) = 1972, qmax
:
25.46◦, R1 = 0.0220 (7572 observed data), wR2 = 0.0537 (all 7880
-3
˚
data), GOF = 1.079, 536 parameters, Drmax/min = 0.76/-0.56 e A .
3: red prism, C38H33B4F24Mo4N11O8, Mr = 1654.75; triclinic, space
¯
group P1 (No. 2◦), a = 14.2639(6), b = 16.9759(8), c = 17.5316(7)
◦
◦
11B NMR (d (ppm), in CD3CN): 2.04 (s, BF4 ). 13C NMR (d
A, b = 96.067(2) , b = 98.424(2) , b = 99.904(2) , V = 4099.3(3)
-
˚
(ppm), in CD3CN): 172.1 (COO-) and 147.6 (Cquat). 19F NMR (d
A , Z = 2, l(Mo-Ka) = 0.71073 A. At R1 = 0.0571 and wR2 =
0.1721 the refinements were aborted due to unresolvable disorder
problems. 4: red fragment, C40H54B4F16Mo4N14O8, Mr = 1589.97;
3
˚
˚
(ppm), in CD3CN): -132.9 (s, 16 F, Far), -146.3 (s, 6 F, BF4-) and
-146.4 (s, 26 F, BF4-). Anal. Calcd. for C66H51B8F48Mo8N17O16
=
¯
[(CH3CN)4.25Mo2(OOC–C6F4–COO)]4[BF4]8: C, 25.54; H, 1.66; N,
triclinic, space group P1 (No. 2), a = 10.6698◦(4), b = 10.7944(5), c =
◦
◦
˚
7.67. Found: C, 24.53; H, 1.76; N, 8.19. UV-Vis (CH3CN): lmax
=
15.1873(6) A, a = 89.095(2) , b = 86.007(2) , g = 64.711(2) , V =
3
-1
˚
˚
432 nm.
1577.52(11) A , Z = 1, l(Mo-Ka) = 0.71073 A, m = 0.880 mm ,
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 11490–11496 | 11491
©

View Article Online
rcalcd = 1.674 g cm-3, T = 123(2) K, F(000) = 788, qmax: 26.01◦,
R1 = 0.0231 (5979 observed data), wR2 = 0.0628 (all 6159 data),
GOF = 1.058, 395 parameters, Drmax/min = 0.90/-0.68 e A . 5:
analysis is obtained for [2 – 6 MeCN]. Attempts to synthesise a
square-shaped compound by either treating 1 with an excess of
oxalic acid, or reacting 2 with 1 equiv. oxalic acid were so far not
successful. The molecular structure of compound 2 is presented in
Fig. 1.
-3
˚
red fragment, C70H78B6F24Mo6N23O15, Mr = 2578.05; monoclinic,
space group P21/c (No. 14), a =◦ 14.7364(6), b = 32.6112(13),
3
˚
˚
c = 22.5349(9) A, b = 104.233(2) , V = 10497.2(7) A , Z = 4,
˚
l(Mo-Ka) = 0.71073 A. At R1 = 0.0658 and wR2 = 0.1523 the
refinements were aborted due to unresolvable disorder problems.
CCDC-830421 (2) CCDC-830422 (4). For more details, see the
ESI.†
Results and discussion
The blue dimolybdenum complex [Mo2(NCCH3)10][BF4]4 (1) se-
lectively reacts with oxalic acid in acetonitrile to give the dimeric
complex 2 in a yield of 94% (Scheme 1a). The 1H NMR spectrum
shows one singlet at 1.95 ppm due to free acetonitrile, which is
a consequence of a fast scrambling with CD3CN. The 11B NMR
-
spectrum shows one resonance at 7.39 (BF4 ). According to the
isotope effect of the boron atom (10B, 19.9% and 11B, 80.1%), two
adjacent singlets in the 19F NMR with an integral ratio of ~1 : 4
are observed at -145.59 and -145.64 ppm.
-
Fig. 1 Molecular structure of 2 in the solid state. The BF4 anions are
omitted for clarity.
It crystallises in the monoclinic space group P21/n with an
inversion centre located in the middle of the oxalate C–C bond.
Two dimolybdenum units are linked by the oxalate linker and
the Mo2 axes are assembled parallel to each other (torsion angle
Mo1–Mo2–Mo1¢–Mo2¢ = 0◦). The Mo2 units and the linker atoms
are all situated almost in one plane. In fact, compound 2 can
be regarded as the smallest metal–organic unit, i.e. an ‘edge’
of a three-dimensional network such as [Zn2(bdc)2(bipy)]24 or
[Zn2(ndc)2(dabco)]25 (bdc = 1,4-benzenedicarboxylate, bipy = 4,4¢-
bipyridine, ndc = 1,4-naphthalenedicarboxylate and dabco = 1,4-
diazabicyclo[2.2.2]octane). As shown in Table 1, all molybdenum
atoms exhibit an octahedral coordination environment. The Mo–
Mo, Mo–Neq and Mo–Nax bond lengths are very similar to those
found in the precursor [Mo2(CH3CN)10][BF4]4 (deviations of 0.01
20
˚
A). The Mo–Mo bond lengths of similar dimers described by
8
˚
Cotton et al. (on average 2.09 A), which are also listed in Table
Scheme 1 Synthesis of (a) [(CH3CN)8Mo2(OOC–COO)Mo2(NCCH3)8]-
[BF4]6 2 and (b) [(CH3CN)6Mo2(OOC–C6F4–COO)]4[BF4]8 3.
1, are shorter than those in 2. In comparison to Cotton’s oxalate-
bridged dimer, the elongation of the Mo–Mo bond distance in
2 is probably due to axial coordination of electron-donating
acetonitrile ligands. The Mo–O distance in 2 is about 0.05 A
shorter than in Cotton’s oxalate-bridged dimer, whereas the Mo–
Compound 2 is well soluble in benzonitrile (56 g L-1) and
dimethylformamide (50 g L-1) and fairly soluble in acetonitrile (3.3
g L-1). Due to the low solubility of 2, 13C NMR measurements in
CD3CN were not successful. UV-Vis experiments of the red dimer
show a maximum absorbance at 424 nm. Considering the lability20
of the coordinating acetonitrile ligands, an adequate elemental
˚
Neq bond lengths are similar. Comparing the distances d between
˚
centroids of Mo2 units within compound 2 (d = 6.86 A) and
˚
˚
Cotton’s dimer (d = 6.95 A), a difference of 0.09 A is found. The
20
Table 1 Comparison of bond distances (A) in the precursor 1, compounds 2, 4 and 5, Cotton’s (FAC) oxalate-linked dimer8 and malonate-bridged
˚
loop10
dmin-max
1
2
FAC’s dimer
3a
4
FAC’s loop
5a
Mo–Mo
Mo–O
Mo–Neq
Mo–Nax
Mo–F
2.187(1)
—
2.11(1)–2.14(1)
2.60(1)
—
2.1760(2)
2.090(1)
2.115(3)–2.145(3)
2.119(4)–2.168(5)
—
—
—
2.159–2.160
2.072–2.093
2.145–2.155
2.533–2.636
2.661
2.1529(2)
2.088(1)
2.130(6)–2.144(6)
2.098(8)–2.124(7)
—
—
—
2.145–2.152
2.053–2.106
2.134–2.164
2.507–2.701
2.624
2.075(1)–2.082(1)
2.130(2)–2.151(2)
2.606(2)–2.613(2)
—
2.079(1)–2.093(1)
2.150(2)–2.159(2)
2.648(2)–2.700(1)
—
Mo–OH
—
—
—
2.902
a Data are taken from the aborted refinements.
11492 | Dalton Trans., 2011, 40, 11490–11496
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Mo–Neq–C bond angles in 2 and [Mo2(NCCH3)10]4[BF4]4 are all
close to 180◦, while the Mo–Nax–C angles in 2 (average of 175◦)
are about 20◦ larger than in 1. The angle Mo1–Mo2–N7ax is 169.6◦
and about 7◦ smaller than the angle in the starting material. Labels
are given in the ESI.†
The treatment of 1 equiv. [Mo2(NCCH3)10][BF4]4 with 1 equiv.
perfluoroterephthalic acid (Scheme 1b) in acetonitrile at 60 ◦C
leads to the formation of the square-shaped complex 3 in 82%
yield. Below 60 ◦C, the reaction rate is significantly slower.
Surprisingly, the treatment of 2 equiv. [Mo2(NCCH3)10][BF4]4
with 1 equiv. dicarboxylic acid does not afford the terephthalate-
bridged analogue of compound 2, but the tetramer 3 instead
(1 equiv. of unreacted [Mo2(NCCH3)10][BF4]4 can be isolated).
The reaction of 1 equiv. of 3 with 20 equiv. perfluoroterephthalic
acid at 60 ◦C results in the precipitation of a red solid which
could not be convincingly identified. Obviously, the product
structure is dependent on the length of the linker; in the case
of the oxalate-bridged PW complexes, a square-shaped molecule
is thermodynamically less favourable than the single-bridged
complex 2.
The purity of the compound was confirmed by 1H NMR
spectroscopy in CD3CN. The spectrum shows only one singlet at
1.95 ppm originating from free acetonitrile. Compound 3 exhibits
a higher solubility in acetonitrile (577 g L-1) than the starting
material 1 (11 g L-1), but is barely soluble in other organic solvents
such as benzonitrile (3.6 g L-1) and pyridine (2.3 g L-1). In the
13C NMR, two shifts at 172.1 (COO-) and 147.6 ppm (Cquat) can
Fig. 2 Molecular structure of 3 in the solid state with a pore aperture
-
˚
width of around 9.5 A which is partially coordinated by BF4 anions in
-
axial positions. The remaining six BF4 anions are omitted for clarity.
(di)carboxylates should be possible. According to 1H NMR spec-
troscopy, the acetonitrile ligands can be replaced by benzonitrile.
The design of cubes or 1D-tubes via stacking molecules of 3 along
the Mo–Mo bond, however, was not yet successful. Similarly,
the passivation of the equatorial coordination sites via replacing
acetonitrile molecules by carboxylates, as well as the extension
of 3 into a two-dimensional plane did so far not succeed. In the
structure of compound 3, the area of the square formed by the
be observed. The11B NMR signal of BF4 is found at 2.04 ppm.
-
The 19F NMR shows one singlet at -132.9 ppm (attributed to 16
aromatic fluorine atoms of the linker) and two singlets at -146.3
-
and -146.4 ppm (arising from the 32 fluorine atoms of 8 BF4 ). The
ratio of integral intensities of C6F4 and BF4 is 1 : 2 and confirms
the cis-configuration of 3. UV-Vis experiments of the red tetramer
show a maximum absorbance at 432 nm. Since the coordinating
acetonitrile ligands are very weakly bound to the Mo centre and
thus easily removed from the complex as has been documented
before,20 elemental analysis values are not in good agreement with
the calculated values.
2
˚
four Mo2 cores is estimated to be around 11 ¥ 11 A . Cotton’s
tetramers have areas from 7 ¥ 7 (oxalate-bridged square) to 15 ¥ 15
2
17
˚
A (4,4¢-biphenyldicarboxylate square). There is no substantial
difference between the length of the lateral edge of 3 (Mo2 ◊ ◊ ◊ Mo2
˚
The molecular structure of compound 3 is shown in Fig. 2. It
is ca. 11 A) and the distance between centroids of Mo2 units
¯
crystallises in the triclinic space group P1 with one formula unit
within the perfluoroterephthalate-bridged dimer of Cotton et al.
8
˚
in the unit cell and an inversion centre located at the centre of the
cationic molecule. Each dimolybdenum unit represents a corner
of the square with two cis-arranged linking perfluoroterephthalate
dianions as equatorial ligands. All remaining equatorial and most
axial positions are occupied by acetonitrile molecules; BF4- anions
occupy two of the eight axial positions. The Mo2 axes are arranged
parallel to each other. Likewise compound 2, 3 can also be re-
garded as a macromolecular subunit of typical three-dimensional
PW-MOFs, such as Kim’s [Zn2(bdc)2(dabco)]n.4 The molybdenum
atoms exhibitanoctahedralcoordination environment. The planes
of the carboxylic group and the aromatic ring of the linker are
twisted relative to each other. The cis-coordinated carboxylic
groups form a right angle and enforce the square array of 3.
(Mo2 ◊ ◊ ◊ Mo2 = 11.3 A). Hence, the number of carboxylic groups
coordinating to one Mo2 unit does not seem to influence the
Mo–O bond length (see Table 1). This fact is also confirmed by
comparison of Mo–O bond lengths within Cotton’s dimers8 and
tetramers.17
The reaction of 1 equiv. [Mo2(NCCH3)10][BF4]4 with 1 equiv.
1,1-cyclobutanedicarboxylic acid leads to the formation of the
molecular loop [(CH3CN)6Mo2(OOC–C4H6–COO)]2[BF4]4 4 in
yields of 75% (Scheme 2a). Compound 4 is characterised by
1H, 11B, 13C and 19F NMR spectroscopy. As expected, the 11B
-
NMR shows one singlet at 7.25 ppm (related to BF4 ) and the 19
F
NMR exhibits two adjacent singlets at -146.4 and -146.5 ppm
(originating from BF4 ). The 1H NMR shows three signals at 3.35
-
-
The coordination of BF4 anions raises the question whether the
ppm (t, –CquatCH2–, 8H), at 2.37 ppm (m, –CH2CH2CH2–, 4 H)
and at 1.96 ppm (s, free CH3CN, ca. 30 H) with an integral ratio
of 2 : 1 : 7.5, indicating the removal of two axially coordinating
acetonitrile ligands upon drying under vacuum. In the case of
removing five acetonitrile ligands, calculated values for elemental
analysis concur with [4 – 5 MeCN]. In comparison to the free
reactivity of 3 is influenced by axially coordinating BF4- anions.26–28
This assumption is supported by the unusually poor reactivity
of 3. For instance, 3 is inert to an excess of (di)carboxylates, or
to linear linkers such as 1,4-dicyanobenzene. This is contrary
to the chemical intuition, a substitution of nitrile ligands by
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 11490–11496 | 11493
©

View Article Online
Table 2 Selected NMR shifts in compounds 3, 4, 5 and in corresponding free linkers in CD3CN. In comparison to free acids, a general trend to low field
shifted peaks can be observed
d (ppm)
1H
3
Free acid 3
—
4
Free acid 4
5
Free acid 5
—
3.35 CH2
2.37 CH2
—
2.48 CH2
1.94 CH2
—
8.66 OH
7.99 CarH
—
7.45 OH
7.61 CarH
—
19F
-132.9 Far
-135.1 Far
2
˚
and the quaternary C atoms is 17.7 A . The Mo2 units are
oriented parallel to each other and form a rectangle with four
right angles. Since the average Mo–O bond length in 4 (2.09
˚
A) is shorter and stronger than in Cotton’s malonate loop [cis-
˚
Mo2(DAniF)2]2(O2CCH2CO2)2 (2.14 A), the Mo2 entities in 4 (d =
˚
˚
6.3 A) are about 0.2 A closer to each other than in Cotton’s
10
˚
loop (d = 6.5 A). Due to the bent nature of the linker (angle
C6–C2–C1¢ of 103.3◦ in 4), the geometry of 4 is predetermined.
Labels are given in the ESI.† Similar to compounds 1–3, the four
Mo atoms exhibit an octahedral coordination environment. The
parallel oriented Mo2 units are surrounded by two cis-oriented
carboxylate groups and six acetonitrile ligands. The molecular
rhombus 4 is the first example of a molecular loop containing a
1,1-cyclobutanedicarboxylate linker. Cotton et al. synthesised a
similar loop by using malonate as the linker.10
Scheme 2 Synthesis of (a) [(CH3CN)6Mo2(OOC–C4H6–COO)]2[BF4]4 4
and (b) [(CH3CN)6Mo2(m-bdc–OH)]3[BF4]6 5.
So far, the passivation of 4 with stoichiometric amounts of
glacial acetic acid was not successful. When an excess of glacial
acetic acid is used, 4 reacts to [Mo2(OAc)4] as confirmed by a
single resonance at 2.59 ppm related to the methyl groups of
dicarboxylic acid (see Table 2), the methylene signals are shifted to
low field. Four signals in the 13C NMR spectrum can be assigned to
compound 4, i.e. the carboxyl group (187.9 ppm), the quaternary
carbon atom (147.6 ppm) and the methylene groups (60.0 and
32.7 ppm). Complex 4 is pink in colour and exhibits a maximum
absorbance at 533 nm in the UV-Vis spectrum.
1
the coordinating acetates in the H NMR spectrum (in THF-
d8). Similarly to 3, it is possible to substitute all acetonitrile with
benzonitrile ligands.
The reaction of equimolar amounts of
1
and 5-
hydroxyisophthalic acid (m-bdc–OH) yields the bowl-shaped
molecular triangle [(CH3CN)6Mo2(m-bdc–OH)]3[BF4]6 5 in yields
of 50% (see Scheme 2b). The purity of compound 5 is confirmed
by NMR spectroscopy. The 11B and 19F NMR spectra show
The molecular structure of 4 is shown in Fig. 3. The compound
crystallises in the triclinic space group P1 with one formula unit
¯
per unit cell and an inversion centre located at the centre of
the molecule. Two dimolybdenum units are linked by two bent
1,1-cyclobutanedicarboxylate linkers forming a molecular loop 4.
The midpoints of the Mo2 bonds and the quaternary C atoms
C2 and C2¢ of the linker molecules define a rhombus having
-
typical shifts related to the BF4 anion. Low field shifted singlets
(compared to free 5-hydroxyisophthalic acid, see Table 2) in
the 1H NMR can be assigned to the hydroxyl group (8.66
ppm) and to aromatic hydrogen atoms (7.99 and 7.82 ppm) of
the 5-hydroxyisophthalate linker. Since the hydroxyl group is
interacting with Mo in solution, a sharp and low field shifted
singlet is observed. As usual, a singlet at 1.96 ppm can be
observed, originating from free acetonitrile molecules. The 13C
NMR spectrum shows signals at 181.4, 159.0, 147.5 and 132.6
ppm which are assigned to COO-, COH, Cquat and CarH. Again,
the coordination of the ligand to the Mo2 units causes low field
shifts. UV-Vis experiments of the red complex reveal a maximum
absorbance at 416 nm. According to elemental analysis, the drying
of 5 in vacuo leads to the removal of eight acetonitrile ligands.
The molecular structure of compound 5 is shown in Fig. 4. It
crystallises in the monoclinic space group P21/c with four formula
units per unit cell. Each dimolybdenum unit represents a corner of
the triangular bowl-shaped array with two cis-oriented bridging
5-hydroxyisophthalate linkers and either five or six coordinating
acetonitrile molecules.
˚
four almost equidistant sides (d = 4.35 A). The rhombic triangle
exhibits two angles 99.04(4)◦ for Mo2–C2–Mo2¢ and 80.81(4)◦
for C2–Mo2–C2¢ whose opposite angles are equal. The area
of the rhombus defined by the midpoints of the Mo2 bonds
The Mo2 units define an equilateral triangle exhibiting Mo2–
Mo2–Mo2 angles of ~60◦ and average distances d between
centroids of Mo2 bonds of ~9.5 A. Thus, the distances be-
-
˚
Fig. 3 Molecular structure of 4 in the solid state. The BF4 anions are
tween the midpoints of the Mo2 bonds in Cotton’s triangle
omitted for clarity.
11494 | Dalton Trans., 2011, 40, 11490–11496
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
a weak intermolecular HO ◊ ◊ ◊ Mo interaction (Mo3–O15 bond
˚
length of 2.902 A) between the triangles leading to the formation
of molecular chains in the solid state and in solution (Fig. 5). Since
the Mo2 corners are surrounded by different ligands, the Mo–Nax
bond lengths vary from 2.507 for Mo4–N11, 2.585 for Mo6–N16,
˚
2.630 for Mo2–N6 to 2.701 A for Mo1–N5. This variation is due to
electronic effects arising from the different nature of coordinating
ligands. In general, molecular triangles based on linear linkers
such as oxalate or terephthalate exhibit bridging angles of 180◦.
The bent nature of m-bdc–OH in 5 leads to the connection of
the Mo2 units with an angle of 120◦ and an unusual bowl-shaped
structure. Labels are given in the ESI.†
Conclusions
The compound [Mo2(NCCH3)10][BF4]4 serves as a building unit for
defined oligomeric structures 2–5. The molecular loop 4 and the
triangle 5 are novel compounds with regard to the linker and the
shape. Compounds 2–5 formally represent the first steps toward
the synthesis of more complex architectures such as polygons or
1D-, 2D- and 3D-network structures. In comparison to Cotton’s
‘passivated’ compounds (due to bulky and strongly coordinating
ligands), the substitution of labile acetonitrile ligands in complexes
2–5 by dicarboxylic acids should be much easier in order to create
highly ordered frameworks. The role of the anion, especially the
Fig. 4 Molecular structure of 5 in the solid state with a pore width of 6.3
-
˚
A. One BF4 anion coordinates to the axial position of one Mo2 corner.
-
The remaining five BF4 anions are omitted for clarity.
11
˚
[cis-Mo2(DAniF)2]3(trans-1,4-O2CC6H10CO2)3 are about 1.7 A
longer than in 5. The Mo–Mo bond lengths are 2.145, 2.150
-
˚
and 2.152 A and therefore similar to those in 4 and about 0.04
influence of larger, less coordinating anions such as B(C6F5)4 on
˚
A smaller than in 1 (see Table 1). Again, the coordination of
the reactivity of 3 is currently being investigated. In general, the
possibility of substituting weakly bound acetonitrile ligands leads
to numerous connection possibilities. Compounds 2–5, especially
2 and 3, are potential building blocks for MOFs. So far, in nearly
all cases, MOFs are prepared under solvothermal conditions and
in the case of PW MOFs the M2 unit is formed in situ. Hence,
the controlled assembly of the ‘lego bricks’ 2 and 3 to larger
architectures would represent a novel access to MOFs.
carboxylic groups to the Mo2 centre causes a shortening of the
Mo2 bond. The Mo–O bonds in 5 (average bond length of 2.08
˚
A) are shorter than in Cotton’s triangle (average bond length of
˚
2.15 A). In two of the three Mo2 units one axial position is not
-
occupied by acetonitrile. In one case, a BF4 anion coordinates
˚
to the Mo atom (Mo5–F7 bond length of 2.624 A) as in 3.
The unoccupied axial position of the second Mo2 unit allows
Fig. 5 Bowl-shaped triangles 5 are interacting by weak HO ◊ ◊ ◊ Mo-bonding to form molecular chains bearing O ◊ ◊ ◊ Mo bonds which are also present in
solution.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 11490–11496 | 11495
©

View Article Online
15 J. K. Bera, P. Angiridis, F. A. Cotton, M. A. Petrukhina, P. E. Fanwick
and R. A. Walton, J. Am. Chem. Soc., 2001, 123, 1515.
16 F. A. Cotton, L. M. Daniels, C. Lin and C. A. Murillo, Inorg. Chem.
Commun., 2001, 4, 130.
17 F. A. Cotton, C. Lin and C. A. Murillo, Inorg. Chem., 2001, 40, 478.
18 P. Angaridis, J. F. Berry, F. A. Cotton, C. A. Murillo and X. Wang, J.
Am. Chem. Soc., 2003, 125, 10327.
Acknowledgements
M. K. acknowledges financial support of TUM Graduate School.
The authors thank Johannes Wandt and Stephanie Weicker for
their valuable help in experiments.
19 F. A. Cotton, L. M. Daniels, C. Lin, C. A. Murillo and S.-Y. Yu, J.
Chem. Soc., Dalton Trans., 2001, 502.
Notes and references
20 F. A. Cotton and K. J. Wiesinger, Inorg. Chem., 1991, 30, 871.
21 E. Whelan, M. Devereux, M. McCann and V. McKee, Chem. Commun.,
1997, 427.
1 F. A. Cotton, C. Lin and C. A. Murillo, Acc. Chem. Res., 2001, 34,
759.
2 M. Ko¨berl, M. Cokoja, W. A. Herrmann and F. E. Ku¨hn, Dalton Trans.,
2011, 40, 6834.
3 S. S.-Y. Chui, S. M.-F. Lo, J. P. H. Charmant, A. G. Orpen and I. D.
Williams, Science, 1999, 283, 1148.
4 D. N. Dybtsev, H. Chun and K. Kim, Angew. Chem., Int. Ed., 2004, 43,
5033.
5 M. Kramer, U. Schwarz and S. Kaskel, J. Mater. Chem., 2006, 16,
2245.
6 B. Chen, M. Eddaoudi, T. M. Reineke, J. W. Kampf, M. O’Keeffe and
O. M. Yaghi, J. Am. Chem. Soc., 2000, 122, 11559.
7 B.-Q. Ma, K. L. Mulfort and J. T. Hupp, Inorg. Chem., 2005, 44, 4912.
8 F. A. Cotton, J. P. Donahue, C. Lin and C. A. Murillo, Inorg. Chem.,
2001, 40, 1234.
9 F. A. Cotton, J. P. Donahue and C. A. Murillo, J. Am. Chem. Soc.,
2003, 125, 5436.
10 F. A. Cotton, C. Lin and C. A. Murillo, Inorg. Chem., 2001, 40, 472.
11 F. A. Cotton, C. Lin and C. A. Murillo, Inorg. Chem., 2001, 40, 575.
12 F. A. Cotton, L. M. Daniels, C. Lin and C A. Murillo, J. Am. Chem.
Soc., 1999, 121, 4538.
13 F. A. Cotton, C. A. Murillo, X. Wang and R. Yu, Inorg. Chem., 2004,
43, 8394.
22 (a) J.-L. Zuo, F. Fabrizi de Biani, A. M. Santos, K. Ko¨hler and F. E.
Ku¨hn, Eur. J. Inorg. Chem., 2003, 449; (b) J.-L. Zuo, E. Herdtweck and
F. E. Ku¨hn, J. Chem. Soc., Dalton Trans., 2002, 1244; (c) W.-M. Xue
and F. E. Ku¨hn, Eur. J. Inorg. Chem., 2001, 2041; (d) W.-M. Xue, F.
E. Ku¨hn, G. Zhang, E. Herdtweck and G. Raudaschl-Sieber, J. Chem.
Soc., Dalton Trans., 1999, 4103; (e) F. E. Ku¨hn, I. S. Gonc¸alves, A. D.
Lopes, J. P. Lopes, C. C. Roma˜o, W. Wachter, J. Mink, L. Hajba, A. J.
Parola, F. Pina and J. Sotomayor, Eur. J. Inorg. Chem., 1999, 295.
23 (a) W.-M. Xue, F. E. Ku¨hn, E. Herdtweck and Q. Li, Eur. J. Inorg.
Chem., 2001, 213; (b) W.-M. Xue, F. E. Ku¨hn and E. Herdtweck,
Polyhedron, 2001, 20, 791.
24 B. Chen, C. Liang, J. Yang, D. S. Contreras, Y. L. Clancy, E. B.
Lobkovsky, O. M. Yaghi and S. Dai, Angew. Chem., Int. Ed., 2006,
45, 1390.
25 H. Chun, D. N. Dybtsev, H. Kim and K. Kim, Chem.–Eur. J., 2005, 11,
3521.
26 M. Majumdar, S. K. Patra, M. Kannan, K. R. Dunbar and J. K. Bera,
Inorg. Chem., 2008, 47, 2212.
27 H.-Y. Weng, Y.-Y. Wu, C.-P. Wu and J.-D. Chen, Inorg. Chim. Acta,
2004, 357, 1369.
28 E. F. Day, J. C. Huffman, K. Folting and G. Christou, J. Chem. Soc.,
Dalton Trans., 1997, 2837.
14 F. A. Cotton, C. A. Murillo, S.-E. Stiriba, X. Wang and R. Yu, Inorg.
Chem., 2005, 44, 8223.
11496 | Dalton Trans., 2011, 40, 11490–11496
This journal is
The Royal Society of Chemistry 2011
©