Tuning the structure and solubility of nanojars by peripheral ligand substitution, leading to unprecedented liquid-liquid extraction of the carbonate ion from water into aliphatic solvents

Article abstract of DOI:10.1039/c6dt00847j

Nanojars, a novel class of neutral anion-incarcerating agents of the general formula [Cu^II(OH)(pz)]_n (Cu_n; n = 27-31, pz = pyrazolate anion), efficiently sequester various oxoanions with large hydration energies from water. In this work, we explore whether substituents on the pyrazole ligand interfere with nanojar formation, and whether appropriate substituents could be employed to tune the solubility of nanojars in solvents of interest, such as long-chain aliphatic hydrocarbons (solvent of choice for large-scale liquid-liquid extraction processes) and water. To this end, we conducted a comprehensive study using 40 different pyrazole ligands, with one, two or three substituents in their 3-, 4- and 5-positions. The corresponding nanojars are characterized by single-crystal X-ray diffraction and/or electrospray-ionization mass spectrometry (ESI-MS). The results show that Cu_n-nanojars with various substituents in the pyrazole 4-position, including long chains, phenyl and CF₃ groups, can be obtained. Straight chains are also tolerated at the pyrazole 3-position, and favor the Cu₃₀-nanojar. Homoleptic nanojars, however, could not be obtained with phenyl or CF₃ groups. Nevertheless, if used in mixture with the parent non-substituted pyrazole, sterically hindered pyrazoles do form heteroleptic nanojars. With 3,5-disubstituted pyrazoles, only heteroleptic nanojars are accessible. The crystal structure of novel nanojars (Bu₄N)₂[CO₃?{Cu₃₀(OH)₃₀(3,5-Me₂pz)_y(pz)_30-y}] (y = 14 and 15) is presented. We find that in contrast to the parent nanojar, which is insoluble in aliphatic solvents and water, nanojars with alkyl substituents are soluble in saturated hydrocarbon solvents, whereas nanojars based on novel pyrazoles, functionalized with oligoether chains, are readily soluble in water. Liquid-liquid extraction of carbonate from water under basic pH is presented for the first time.

Full text of DOI:10.1039/c6dt00847j

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: B. M. Ahmed, B.
Calco and G. Mezei, Dalton Trans., 2016, DOI: 10.1039/C6DT00847J.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/dalton

Page 1 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
Tuning the structure and solubility of nanojars by peripheral ligand
substitution, leading to unprecedented liquid-liquid extraction of the
carbonate ion from water into aliphatic solvents
Basil M. Ahmed, Brice Calco and Gellert Mezei*
Department of Chemistry, Western Michigan University, Kalamazoo, Michigan, USA
*Correspondent author. Email: gellert.mezei@wmich.edu
Carbonate, an anion with large hydration energy, is extracted from water by solvent
extraction for the first time, using functionalized nanojars
ABSTRACT
Nanojars, a novel class of neutral anionꢀincarcerating agents of the general formula
[CuII(OH)(pz)]n (Cun; n = 27−31, pz = pyrazolate anion), efficiently sequester various oxoanions
with large hydration energies from water. In this work, we explore whether substituents on the
pyrazole ligand interfere with nanojar formation, and whether appropriate substituents could be
employed to tune the solubility of nanojars in solvents of interest, such as longꢀchain aliphatic
hydrocarbons (solvent of choice for largeꢀscale liquidꢀliquid extraction processes) and water. To
this end, we conducted a comprehensive study using 40 different pyrazole ligands, with one, two
or three substituents in their 3ꢀ, 4ꢀ and 5ꢀpositions. The corresponding nanojars are characterized
by singleꢀcrystal Xꢀray diffraction and/or electrosprayꢀionization mass spectrometry (ESIꢀMS).
The results show that Cunꢀnanojars with various substituents in the pyrazole 4ꢀposition, including
long chains, phenyl and CF3 groups, can be obtained. Straight chains are also tolerated at the
pyrazole 3ꢀposition, and favor the Cu30ꢀnanojar. Homoleptic nanojars, however, could not be
1

Dalton Transactions
Page 2 of 28
View Article Online
DOI: 10.1039/C6DT00847J
obtained with phenyl or CF3 groups. Nevertheless, if used in mixture with the parent nonꢀ
substituted pyrazole, sterically hindered pyrazoles do form heteroleptic nanojars. With 3,5ꢀ
disubstituted pyrazoles, only heteroleptic nanojars are accessible. The crystal structure of novel
nanojars (Bu4N)2[CO3⊂{Cu30(OH)30(3,5ꢀMe2pz)y(pz)30ꢀy}] (y = 14 and 15) is presented. We find
that in contrast to the parent nanojar, which is insoluble in aliphatic solvents and water, nanojars
with alkyl substituents are soluble in saturated hydrocarbon solvents, whereas nanojars based on
novel pyrazoles, functionalized with oligoether chains, are readily soluble in water. Liquidꢀ
liquid extraction of carbonate from water under basic pH is presented for the first time.
1. INTRODUCTION
Nanojars constitute a unique, novel class of anionꢀsequestering agents of unparalleled
efficiency.1−3 Comprised of threeꢀring assemblies of [cisꢀCuII(ꢁꢀOH)(ꢁꢀpz)]x metallamacrocycles
(x = 6−14, except 11), these supramolecular coordination assemblies bind diꢀ or trinegative
anions, such as sulfate,1,3 phosphate,2 arsenate2 and carbonate,3 by wrapping a multitude of
hydrogen bonds around and totally isolating the anion from its surrounding medium (as in the
sulfateꢀ4 and phosphateꢀbinding proteins5). Various nanojars of the general formula
2–
(Bu4N)2[EOa ⊂{Cu(OH)(pz)}n] (E = C, a = 3; E = S, HP or HAs, a = 4; n = 27−31) have been
discovered to date, containing the following ringꢀcombinations: n = 6+12+9 (Cu27), 6+12+10
(Cu28), 7+13+9 (Cu29), 8+13+8 (Cu29) and 8+14+9 (Cu31). The unprecedented anionꢀbinding
strength is demonstrated by the inability of an aqueous Ba2+ solution to precipitate barium
sulfate, phosphate or arsenate, respectively, when stirred with a solution of the corresponding
nanojars.1,2 Thus, nanojars could be exploited as highly efficient liquidꢀliquid extraction agents
for the removal of those anions from contaminated aqueous media.
Currently, there is a strong interest in developing novel chemical systems for the recognition
of anions, with an aim at efficient and selective removal and recovery of anions from
solution,6−16 anion sensing,17,18 and anion transport.19−23 Anions with large hydration energies,
3−
2−
2−
such as PO4 (ꢀGh° = −2773 kJ mol−1), CO3 (ꢀGh° = –1315 kJ mol−1) and SO4 (ꢀGh° =
−1090 kJ mol−1),24 are particularly difficult to extract from aqueous solutions, and to the best of
our knowledge, no liquidꢀliquid extraction method of carbonate is known. Nanojars based on the
parent pyrazole ligand bind carbonate strongly; however, these nanojars are not soluble in
organic solvents suitable for industrialꢀscale liquidꢀliquid extraction processes, i.e. longꢀchain
2

Page 3 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
aliphatic hydrocarbons. To address this issue, as well as to study the effect of pyrazole
substituents on the structure and stability of nanojars, we conducted a comprehensive study using
40 different pyrazole ligands with either one, two or three substituents in their 3ꢀ, 4ꢀ and 5ꢀ
positions (Scheme 1). The novel nanojars are characterized by singleꢀcrystal Xꢀray diffraction in
the solid state, and/or electrosprayꢀionization mass spectrometry (ESIꢀMS) in solution. The
structure of two different Cu30ꢀnanojars are described for the first time, which are comprised of
previously unknown ring combinations, n = 8+14+8 and n = 7+14+9. Peripheral decoration of
the nanojars with long hydrocarbon or oligoether chains expands the solubility of nanojars to
aliphatic solvents and water, respectively. For this purpose, novel pyrazole ligands with
oligo(ethylene glycol) methyl ether substituents have been synthesized. Furthermore, we use
highly hydrophobic nanojars, functionalized with multiple alkyl chains, to transfer the carbonate
ion from alkaline aqueous solutions to aliphatic solvents, and present the first liquidꢀliquid
method for the extraction of carbonate.
Scheme 1. Pyrazole ligands used in the present study. Homoleptic nanojars can be obtained with
pyrazoles shown in green; those in orange can only form heteroleptic nanojars, and the ones in red do not
form nanojars under similar conditions.
3

Dalton Transactions
Page 4 of 28
View Article Online
DOI: 10.1039/C6DT00847J
2. RESULTS AND DISCUSSION
2.1. Synthesis of the new ligands. Pyrazole ligands substituted in the 4ꢀposition with monoꢀ, diꢀ
or triethyleneglycol methyl ether chains are prepared in three steps (Scheme 2). First, the
oligo(ethylene glycol) monomethyl ether (OEG), CH3(OCH2CH2)zOH (z = 1, 2 or 3), is
condensed with excess bromoacetaldehyde diethyl acetal in the presence of NaH, to yield the
OEG diethyl acetals in quantitative yield. Next, the acetal is added to the Vilsmeier reagent
obtained from POCl3 and dimethylformamide, resulting in a mixture of OEGꢀsubstituted
dimethylaminoꢀ and ethoxyꢀpropenal, and smaller amounts of EtOꢀsubstituted dimethylaminoꢀ
and ethoxyꢀpropenal. This mixture is used without separation in the next step: treatment with
hydrazine hydrate in methanol yields the corresponding OEGꢀsubstituted pyrazole, along with 4ꢀ
ethoxypyrazole as sideꢀproduct. The pure OEGꢀpyrazoles are obtained in 20−34% overall yields,
based on the OEG starting material, after purification by column chromatography. 4ꢀnꢀ
Octylpyrazole is synthesized similarly, using decanal diethyl acetal. 3,5ꢀDimethylꢀ4ꢀnꢀ
octylpyrazole is prepared from hydrazine and 3ꢀ(nꢀoctyl)pentaneꢀ2,4ꢀdione, which in turn is
obtained from acetylacetone and 1ꢀiodooctane (Scheme 2).
Scheme 2. Synthesis of the novel OEGꢀpyrazole ligands, and new synthetic routes to 4ꢀnꢀoctylpyrazole
and 3,5ꢀdimethylꢀ4ꢀnꢀoctylpyrazole.
2.2. Synthesis of the nanojars. Nanojars of the general formula [CO32−⊂{Cu(OH)(Rpz)}n] (pz =
pyrazolate anion, R = substituent(s), n = 27−31) are synthesized using a oneꢀpot reaction, by
4

Page 5 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
stirring copper(II) nitrate, the pyrazole ligand (or a 1:1 mixture of two different pyrazole ligands)
and a base (Cu:Rpz:OH in 1:1:2 molar ratio) in tetrahydrofuran for three days. Various
hydroxides, such as MOH (M = Li+, Na+, K+, Rb+, Cs+, Bu4N+, Et4N+) and M(OH)2 (M = Sr2+,
Ba2+) can be employed as bases, along with the corresponding carbonates, which provide the
2−
incarcerated CO3 ion. Alternatively, copper carbonate Cu2CO3(OH)2 can be used both as
copperꢀ and carbonateꢀsource, although it provides lower yields than the THFꢀsoluble copper(II)
nitrate.
Tetrabutylammonium was initially used as counterion in all cases, as it had been shown to
promote crystallization.1−3 In the case of highly hydrophobic nanojars based on pyrazoles with
butyl or longer alkyl substituents, however, no ESIꢀMS signals can be detected when the
counterion is Bu4N+. The lack of ionization is attributed to extensive ionꢀpair formation, due to
strong hydrophobic interactions between the long alkyl chains of the nanojars and the large
Bu4N+ counterions. Alkali metal counterions (Li+, Na+, K+, Rb+, Cs+) offer partial remedy for the
problem, although in this case the MS signal is weakened by another factor: highly hydrophobic
nanojars are insoluble in neat acetonitrile, therefore less polar tetrahydrofuran has to be added to
the MS solvent in a 1:1 ratio to achieve solubility, which further suppresses the MS signal. Et4N+
behaves similarly to Bu4N+, although it does induce noteworthy changes in the crystal structure
of the nanojar (see Crystallography section). Besides tetraalkylammonium and alkali or alkalineꢀ
earth metal cations, [K⊂18ꢀcrownꢀ6]+ has also been successfully employed as counterion, which
forms in-situ when 18ꢀcrownꢀ6 is added to the reaction mixture. The identity of a particular
counterion associated with the nanojar is confirmed by ESIꢀMS(+) (of the metal cations, only
Rb+ and Cs+ have m/z values within the mass spectrometer’s detection window).
The size distribution of the different [CO32−⊂{Cu(OH)(pz)}n] nanojars (n = 27−31) varies
slightly with the particular reaction conditions, such as the copper source, counterion and the
base. In contrast, a much more pronounced variation is produced by the different substituents on
the pyrazole ligands.
2.3. Effect of pyrazole substituents on nanojar structure. As seen in Figure 1, the pyrazole 4ꢀ
positions point radially away from the nanojar. Therefore, no steric hindrance is predicted to be
caused by substituents in the 4ꢀposition. Substitution at the pyrazole 3ꢀ and/or 5ꢀpositions,
however, is expected to be more critical for nanojar formation, due to the allꢀcis configuration of
the pyrazole ligands, which places substituents on the pyrazole units close to the neighboring
5

Dalton Transactions
Page 6 of 28
View Article Online
DOI: 10.1039/C6DT00847J
ligands. The steric hindrance is expected to be most significant for the two smaller sideꢀrings,
with pyrazole units pointing to the same side of the ring, while the larger, central ring can relieve
the hindrance by orienting the pyrazole units alternately above and below the ring meanꢀplane
(see also the Crystallography section).
All experiments discussed below are carried out in tetrahydrofuran as reaction medium, an
excellent solvent for all nanojars, but not for the NaNO3 byꢀproduct, which precipitates out of the
reaction medium. The product mixtures are analyzed by ESIꢀMS in acetonitrile solutions, or, in
the case of pyrazoles with butyl or longer chains, in a less polar CH3CN/THF (1:1) mixture.
Mass spectra are shown in the Supporting Information (Figures S1−S30). Although the relative
intensity of ESIꢀMS signals is dependent on individual ionization efficiencies, the various
[CO3⊂{Cu(OH)(pz)}n]2− nanojars (n = 27−31) have identical charge and very similar size and
shape, and are expected to have similar ionization efficiencies under identical ESIꢀMS
conditions. Therefore, the observed ESIꢀMS signals should provide reliable information about
the relative concentration of different nanojars based on the same ligand in a given solution.
Indeed, this has been confirmed for nanojars based on the parent pyrazole ligand, using 1H NMR
spectroscopy.3 In the case of the mixedꢀpyrazole nanojars, however, ionization efficiencies drop
drastically as the amount of the hydrophobic ligand component within a nanojar of a given size
increases.
Formic acid is commonly used as an additive for mobile phases in reversedꢀphase liquid
chromatography separations, as a buffer component, to improve peak shapes, and to promote
ionization by producing [M+H]+ ions. We find that nanojars are extremely sensitive to even
traces of formic acid in the mass spectrometer; consequently, substituted nanojar species
[CO32−⊂{Cun(OH)n(HCOO)y(pz)}n−y] (y = 1−3) are inevitably found in most spectra (at slightly
lower m/z values than the pure nanojar), unless the instrument is opened up and thoroughly
cleaned prior to injection of nanojar samples. In addition, adduct species are also observed
occasionally, at larger m/z values than the pure nanojar.
2.3.1. Substitution of the pyrazole 4-position. Nanojars [CO3⊂{Cu(OH)(4ꢀRpz)}n]2– (n = 27−31)
based on pyrazoles substituted in the 4ꢀposition with chains of varying lengths (up to 11 atoms)
have been successfully prepared (Table 1, Figures S6−S13), indicating that linear chains do not
interfere with nanojar formation, regardless of their length (Figure 1). In addition, nanojars with
R = F, Cl, Br, I, phenyl and CF3 have been observed (Figures S2−S5, S14, S15), showing that
6

Page 7 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
moderately bulky groups in the immediate vicinity of the pyrazole 4ꢀposition are also tolerated.
In the case of 4ꢀtrifluoromethylpyrazole, a hexanuclear species [Cu6O2(4ꢀCF3pz)9]− (m/z 1629) is
also formed beside the nanojars. Pyrazoles with potentially coordinating moieties (R = SO3H,
COOH, CH=O, OH) in the 4ꢀposition do not form nanojars: only insoluble, presumably
polymeric products are obtained in those cases. With 4ꢀnitropyrazole (LH), a soluble mixture of
monoꢀ to hexanuclear species (as indicated by the Cuꢀisotope pattern of the ESIꢀMS signals) is
obtained (Figure S29). Within this mixture, [CuL2]− (m/z 288), [Cu2L4]− (m/z 575),
[NaCu3OL3(NO3)2]− (m/z 690), [Cu3OL5]− (m/z 767), [Cu4O2L6]− (m/z 959), [(Bu4N)Cu3OL6]−
(m/z 1122) and [Cu6O2L9]− (m/z 1422) are identified as major components. The structure of
related diꢀ, triꢀ, tetraꢀ and hexanuclear copper(II)ꢀpyrazolate species has been established by Xꢀ
ray diffraction.25−29 Partial reduction of Cu2+ to Cu+ during ESIꢀMS is also common.30
Figure 1. Spaceꢀfilling models of [CO3⊂{Cu(OH)(4ꢀRpz)}27]2– nanojars (from left to right: R = H, Me,
nBu, nOct).
2.3.2. Substitution of the pyrazole 3(5)-position. Alkyl substituents of various lengths (e.g.,
methyl to nꢀoctyl) are also tolerated at the pyrazole 3(5)ꢀposition (Table 1). It is likely that the
longer alkyl chains turn away from the nanojar after the first CH2ꢀunit connected to the pyrazole
nucleus, thus avoiding interference from neighboring pyrazole units of the same Cuxꢀring. In
contrast to 4ꢀalkylpyrazoles, the Cu30ꢀnanojars [CO3⊂{Cu(OH)(3ꢀRpz)}30]2– are obtained almost
exclusively with 3(5)ꢀalkylpyrazoles (Figures S16−S20). No nanojars can be observed with
bulky phenyl or CF3 substituents at the pyrazole 3(5)ꢀposition, which yield only smaller,
Cu1−Cu6 species. However, a 1:1 molar mixture of 3(5)ꢀPhpzH and HpzH affords an
7

Dalton Transactions
Page 8 of 28
View Article Online
DOI: 10.1039/C6DT00847J
approximately Gaussian distribution of nanojars [CO3⊂{Cun(OH)n(3ꢀPhpz)y(pz)n−y}]2– (Figure
S23; major species: n = 29, y 5−19). Similarly, a 1:1 mixture of 3(5)ꢀF3CpzH and HpzH
produces a series of nanojars [CO3⊂{Cun(OH)n(3(5)ꢀF3Cpz)y(pz)n−y}]2– (Figures S21, S22; major
species: n = 30, y
6−17), beside lower nuclearity species [Cu(3ꢀCF3pz)2]− (m/z 334),
≈
≈
[Cu3O(pz)(3ꢀCF3pz)4]− (m/z 814), [Cu3O(3ꢀCF3pz)5]− (m/z 882), [Cu4O(pz)(3ꢀCF3pz)2(3ꢀ
CF3pzH)4]− (m/z 1152), [Cu4O(3ꢀCF3pz)3(3ꢀCF3pzH)4]− (m/z 1220). These latter two experiments
demonstrate that some but not all pyrazole units of a nanojar can tolerate bulkier substituents in
the 3(5)ꢀposition. As with 4ꢀsubstituted pyrazoles, no nanojars are formed by pyrazoles with
COOH, CH=O, NO2, NH2 or OH substituents at their 3(5)ꢀposition. In the case of 3ꢀ
nitropyrazole (HL), a complex, soluble mixture of products (similar to the one obtained with 4ꢀ
nitropyrazole), is characterized by ESIꢀMS (Figure S30): L− (m/z 112), [CuL2]− (m/z 288),
[CuL3]− (m/z 400), [Cu2L3]− (m/z 463), [Cu2OL3]− (m/z 479), [NaCuL4]− (m/z 535), [Cu2(OH)L4]−
(m/z 592), [NaCu2OL4]− (m/z 614), [Cu3O(OH)L4]− (m/z 672), [Cu2L5]− (m/z 687),
[NaCu2(OH)L5]− (m/z 727), [Cu3OL5]− (m/z 767), [NaCu2L6]− (m/z 822), [Cu4O2L5]− (m/z 847),
[NaCu3OL6]− (m/z 902), [Cu4O2L6]− (m/z 959), [(Bu4N)Cu3OL6]− (m/z 1122), [Cu6OL9]− (m/z
1406), [Cu6L10]− (m/z 1502), [NaCu6O2L10(NO3)]− (m/z 1619).
2.3.3. Simultaneous substitution of the pyrazole 3- and 5-positions. As expected, no nanojars are
obtained with 3,5ꢀdisubstituted pyrazoles as sole ligands, not even with the smallest methyl
substituent. When used in a 1:1 molar mixture with HpzH, however, 3,5ꢀdialkylpyrazoles and
3(5)methylꢀ5(3)trifluoromethylpyrazole do form mixed nanojars. The ESIꢀMS(−) spectrum of
the nanojars [CO3⊂{Cun(OH)n(3,5ꢀMe2pz)y(pz)n−y}]2–, obtained from a 1:1 molar mixture of 3,5ꢀ
Me2pzH and HpzH, and Cu2CO3(OH)2 as copper/carbonate source, shows three major groups of
peaks (n = 27, y
≈
0−9; n = 28, y
≈
≈
6−13; n = 30, y
≈ 10−14) and small amounts of nanojars with
n = 29, y 9−14, and n = 31, y
≈
10−14 (Figure 2). When Cu(NO3)2 and Na2CO3 are used as
copper and carbonate sources, respectively, a slightly different distribution is observed, with n =
30 as the major group, small amounts of n = 28, 29, and only traces of n = 27, 31 (Figure S24).
Different ratios of 3,5ꢀMe2pzH and HpzH also lead to different distributions: a 2:1 molar mixture
leads almost exclusively to n = 30, y
≈
13−20 (most abundant peak: y = 14), with small amounts
14−20 (most abundant peak: y = 20).
of n = 28, y 12−18, while a 5:1 ratio leads to n = 30, y
≈
≈
Using 3,5ꢀMe2pzH:HpzH ratios higher than 1:1 favors nanojars with larger value of y, but
increasing amounts of lowꢀnuclearity species are also produced at the same time. A 1:1 molar
8

Page 9 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
mixture of 3,5ꢀMe2ꢀ4ꢀnOctpzH and HpzH also provides nanojars [CO3⊂{Cun(OH)n(3,5ꢀMe2ꢀ4ꢀ
nOctpz)y(pz)n−y}]2–, among which the ones with n = 30, y
10−16 constitute the major group.
The ESIꢀMS(−) of [CO3⊂{Cun(OH)n(3,5ꢀEt2pz)y(pz) n−y }]2– shows three major groups, n = 27, y
0−3; n = 29, y 0−10; n = 30, y 9−14, and small amounts of n = 31, y 0−11 and n = 32, y ≈
≈
≈
≈
≈
≈
12−15 (Figure S25). In the case of [CO3⊂{Cun(OH)n(3ꢀnBuꢀ5ꢀnHexpz)y(pz)n−y}]2–, the major
peaks correspond to the parent nanojars (n = 27, 29−31; y = 0), but smaller amounts of
substituted species (y = 1−6) are also observed (Figure S26). In the case of a 3(5)ꢀMeꢀ5(3)ꢀ
CF3pzH:HpzH (1:1) mixture, the major species is [Cu3O(pz)3(3ꢀMeꢀ5ꢀCF3pz)2]− (m/z 706), along
with other lowꢀnuclearity species, and small amounts of nanojars are also observed:
[CO3⊂{Cun(OH)n(3ꢀMeꢀ5ꢀCF3pz)y(pz)n−y }]2– (n = 27, y
≈ 0−2; n = 29, y ≈ 0−2; n = 30, y ≈
1−11) (Figures S27, S28). In the case of pyrazoles with bulkier substituents, such as
bis(trifluoromethyl)ꢀ, diꢀtbutylꢀ and diphenylpyrazole, however, not even mixed nanojars can be
obtained. Diꢀtbutylꢀ or diphenylpyrazole mixtures with HpzH (1:1) lead to
[CO3⊂{Cu(OH)(pz)}n]2– (n = 27−31) as the only THFꢀsoluble products. In the case of 3,5ꢀ
(CF3)2pzH:HpzH (1:1), the main species observed by ESIꢀMS(−) is the trinuclear
[Cu3O(pz)3{3,5ꢀ(CF3)2pz}2]− (m/z 814), along with its adducts [NayCu3O(pz)3{3,5ꢀ(CF3)2pz}2+y]−
(y = 1, m/z 1040; y = 2, m/z 1266; y = 3, m/z 1492; y = 4, m/z 1718; y = 5, m/z 1944) and
[(Bu4N)Cu3O(pz)3{(3,5ꢀCF3)2pz}3]− (m/z 1260), as well as [Cu3O(pz)2{(3,5ꢀCF3)2pz}3]− (m/z
950), [Cu3OH(pz)3{(3,5ꢀCF3)2pz}3]− (m/z 1018) and its adduct [Na2Cu3OH(pz)3{(3,5ꢀ
CF3)2pz}5]− (m/z 1470). In addition, mononuclear [Cu{(3,5ꢀCF3)2pz}2]− (m/z 470), tetranuclear
[NayCu4O(pz)3{(3,5ꢀCF3)2pz}3+y]− (y = 1, m/z 1307; y = 2, m/z 1533; y = 3, m/z 1759),
hexanuclear [NayCu6O2(pz)6{(3,5ꢀCF3)2pz}3+y]− (y = 1, m/z 1651; y = 2, m/z 1877; y = 3, m/z
2103) as well as the free ligand and its Naꢀadducts, [Nay{(3,5ꢀCF3)2pz}1+y]− (y = 0, m/z 203; y =
1, m/z 429; y = 2, m/z 655), are also observed. Concurrently, the ESIꢀMS(+) spectrum of the
same solution shows, beside the dominant Bu4N+ ion, mostly trinuclear species
[NayCu3O(pz)3{3,5ꢀ(CF3)2pz}y]+ (y = 0, m/z 408; y = 1, m/z 634; y = 2, m/z 860; y = 3, m/z 1086;
y = 4, m/z 1312; y = 5, m/z 1538), and smaller amounts of hexanuclear species
[(Bu4N)yNazCu6O2(pz)6{(3,5ꢀCF3)2pz}1+y+z]+ (y = 0, z = 3, m/z 1697; y = 0, z = 4, m/z 1923; y =
0, z = 5, m/z 2149; y = 2, z = 0, m/z 1910; y = 2, z = 1, m/z 2136; y = 2, z = 2, m/z 2362),
[Na3Cu6O(OH)(pz)6{(3,5ꢀCF3)2pz}5]+ (m/z 1901) and [(Bu4N)3Na2Cu6O2(pz)7{(3,5ꢀCF3)2pz}5]+
(m/z 2672). In general, we observe that the larger the substituents at the pyrazole 3/5 positions,
9

Dalton Transactions
Page 10 of 28
View Article Online
DOI: 10.1039/C6DT00847J
the less favorable the formation of the corresponding nanojars becomes, and the more lowꢀ
nuclearity species are obtained.
Figure 2. ESIꢀMS(−) spectrum of [CO3⊂{Cun(OH)n(3,5ꢀMe2pz)y(pz)n−y}]2– (y:n−y & m/z values shown)
obtained from Cu2CO3(OH)2, 3,5ꢀMe2pzH/HpzH (1:1), NaOH and Bu4NOH.
10

Page 11 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
Table 1. [Nanojar]2− species with differently substituted pyrazoles detected by ESIꢀMS(−). Occasionally,
minor amounts of singlyꢀcharged [Counterion+(Nanojar2−)]− species are also observed (not shown).
Nanojar [CO3⊂{Cu(OH)L}n]2–
Entry Pyrazole (LH)
1
2
HpzH
m/z 2023 (n = 27), 2171 (n = 29), 2244 (n = 30), 2318 (n = 31)
m/z 2266 (n = 27), 2431 (n = 29), 2514 (n = 30), 2597 (n = 31)
m/z 2488 (n = 27), 2670 (n = 29), 2761 (n = 30), 2852 (n = 31)
m/z 3088 (n = 27), 3314 (n = 29), 3428 (n = 30), 3541 (n = 31)
m/z 3723 (n = 27), 3996 (n = 29), 4133 (n = 30), 4270 (n = 31)
m/z 2618 (n = 27), 2809 (n = 29), 3001 (n = 31)
m/z 2212 (n = 27), 2374 (n = 29), 2455 (n = 30), 2536 (n = 31)
m/z 2780 (n = 27), 2984 (n = 29), 3188 (n = 31)
m/z 3538 (n = 27), 3798 (n = 29), 4058 (n = 31)
m/z 2807 (n = 27), 2910 (n = 28), 3013 (n = 29), 3116 (n = 30), 3218 (n = 31)
m/z 3023 (n = 27), 3245 (n = 29), 3466 (n = 31)
m/z 3618 (n = 27), 3751 (n = 28), 3883 (n = 29), 4016 (n = 30), 4149 (n = 31)
m/z 4212 (n = 27), 4522 (n = 29), 4832 (n = 31)
m/z 2941 (n = 27), 3157 (n = 29), 3372 (n = 31)
m/z 3050 (n = 27), 3274 (n = 29), 3386 (n = 30), 3498 (n = 31)
m/z 2455 (n = 30)
4ꢀFpzH
3
4ꢀClpzH
4
4ꢀBrpzH
5
4ꢀIpzH
6
4ꢀEtOpzH
7
4ꢀMepzH
4ꢀnBupzH
4ꢀnOctpzH
8
9
10
11
12
12
14
15
16
17
18
19
20
4ꢀ(HOCH2CH2CH2)pzH
4ꢀ(CH3OCH2CH2O)pzH
4ꢀ(CH3(OCH2CH2)2O)pzH
4ꢀ(CH3(OCH2CH2)3O)pzH
4ꢀCF3pzH
4ꢀPhpzH
3ꢀMepzH
3ꢀEtpzH
3ꢀnPrpzH
3ꢀnBupzH
3ꢀnOctpzH
m/z 2665 (n = 30)
m/z 2876 (n = 30)
m/z 3086 (n = 30)
m/z 3928 (n = 30)
[CO3⊂{Cun(OH)n(3ꢀCF3pz)y(pz)n−y}]2– (n = 27, 29, 30; y ≈ 3−17)
3ꢀCF3pzH/HpzH
(1:1)
21
22
23
24
25
26
27
m/z 2125−2822
[CO3⊂{Cun(OH)n(3ꢀPhpz)y(pz)n−y}]2– (n = 27, 29; y ≈ 5−19)
m/z 2251−2893
[CO3⊂{Cun(OH)n(3,5ꢀMe2pz)y(pz)n−y}]2– (n = 27−31, y ≈ 0−14)
3ꢀPhpzH/HpzH
(1:1)
3,5ꢀMe2pzH/HpzH
(1:1)
3,5ꢀMe2ꢀ4ꢀnOctpzH/HpzH
m/z 2023−2515
[CO3⊂{Cun(OH)n(3,5ꢀMe2ꢀ4ꢀnOctpz)y(pz)n−y}]2– (n = 29−31, y ≈ 11−17)
m/z 3015−3366
[CO3⊂{Cun(OH)n(3,5ꢀEt2pz)y(pz)n−y}]2– (n = 27−32, y ≈ 0−15)
(1:1)
3,5ꢀEt2pzH/HpzH
(1:1)
3ꢀnBuꢀ5ꢀnHexpzH/HpzH
m/z 2023−2813
[CO3⊂{Cun(OH)n(3ꢀnBuꢀ5ꢀnHexpz)y(pz)n−y}]2– (n = 27, 29−31, y ≈ 0−6)
m/z 2023−2665
[CO3⊂{Cun(OH)n(3ꢀMeꢀ5ꢀCF3pz)y(pz)n−y}]2– (n = 27, 29, 30, y ≈ 0−11)
(1:1)
3ꢀMeꢀ5ꢀCF3pzH/HpzH
(1:1)
m/z 2023−2696
11

Dalton Transactions
Page 12 of 28
View Article Online
DOI: 10.1039/C6DT00847J
2.4. Crystallographic analysis. All three nanojar structures described herein consist of three
neutral [cisꢀCuII(ꢁꢀOH)(ꢁꢀpz)]x rings, with a larger one (x = 13 or 14) sandwiched by two
smaller ones (x = 7, 8 or 9), and they all incarcerate a carbonate ion in their central cavity. The
central, larger ring is approximately flat, with the pyrazolate units symmetrically alternating
slightly above and below the ring meanꢀplane, and they do not form hydrogen bonds to the
carbonate ion. The smaller sideꢀrings are bowlꢀshaped, with their pyrazolate moieties pointing
away from the central ring, while their OH groups point toward the center of the nanojar and
form multiple hydrogen bonds with the incarcerated CO32− ion. While there is no direct bonding
between the two smaller rings, they are both involved in multiple Hꢀbonds and weak axial Cu−O
interactions with the larger central ring. In the [Cu(OH)(pz)]x rings, Cu−O and Cu−N bondꢀ
lenghts are within normal ranges (1.872(4)−1.967(4) and 1.923(4)−2.025(4) Å, respectively).
Two Bu4N+ (or Et4N+) counterions balance the 2− charge of the carbonate ion in all nanojars.
2.4.1. (Et4N)2[CO3⊂{Cu(µ-OH)(µ-pz)}29(H2O)2] (Et4N-1·2H2O). The nanojar in Et4Nꢀ1ꢂ2H2O,
which consists of three [Cu(OH)(pz)]x rings with x = 8, 13 and 8, is similar to the one found in
Bu4Nꢀ1 (Figures S31−S33);3 however, the difference in counterion size (Et4N+ vs. Bu4N+) does
induce significant changes in the molecular structure of the nanojar (Figure S34). Furthermore,
despite a seemingly very similar overall crystal packing, the position of the nanojars relative to
the crystal’s symmetry elements is altered in Et4Nꢀ1 compared to Bu4Nꢀ1. In Bu4Nꢀ1, the nanojar
is located on a twoꢀfold rotation axis (C2), which runs through the Cꢀatom of the incarcerated
carbonate ion, symmetrically bisecting the latter. In Et4Nꢀ1, the nanojar is also located on a twoꢀ
fold rotation axis, but unlike in Bu4Nꢀ1, the C2 axis does not bisect the carbonate ion
symmetrically, although it still runs through its central Cꢀatom. As a result, the incarcerated
carbonate ion in Et4Nꢀ1 is disordered over two positions (50/50). Similarly to Bu4Nꢀ1,3 in Et4Nꢀ1
each Cu8ꢀring forms six Hꢀbonds to the carbonate ion (for one Cu8ꢀring, O···O: 2.81(1)−3.12(1)
Å; average: 2.90(1) Å; for the other Cu8ꢀring, O···O: 2.68(1)−3.01(1) Å; average: 2.90(1) Å).
As in Bu4Nꢀ1, only eight of the twelve Hꢀbonds to carbonate in Et4Nꢀ1 have O···O distances
<2.92(1) Å. The average Hꢀbonding distance to carbonate is identical in Bu4Nꢀ1 and Et4Nꢀ1
(2.90(1) Å). The central Cu13ꢀring forms five Hꢀbonds to each Cu8ꢀring (average O···O: 2.752(6)
Å in Et4Nꢀ1, 2.766(6) Å in Bu4Nꢀ1).
The presence of two water molecules (symmetryꢀrelated) in Et4Nꢀ1 has significant
consequences on the structure of the nanojar (no water is found in Bu4Nꢀ1). Each H2O molecule
12

Page 13 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
is hydrogenꢀbonded to two OH groups across the Cu8ꢀring (O19···O5: 2.851(6), O19···O2:
3.039(7) Å). As a result, those two OH groups are pulled closer to each other (O2···O5: 4.981(6)
Å), compared to the corresponding distance in Bu4Nꢀ1 (O···O: 5.596(6) Å), which lacks the
bridging H2O molecules. In fact, O2 and O5 are indeed the two most likely Oꢀatoms to bind a
bridging H2O molecule, since those two Oꢀatoms are not involved in significant Hꢀbonding with
the Cu13ꢀring (O···O: 3.326(6) and 3.317(6) Å in Et4Nꢀ1; 3.245(6) and 3.324(6) Å in Bu4Nꢀ1).
The distance between the two bonded Cuꢀatoms across the ring also decreases from 8.254(1) Å
in Bu4Nꢀ1 to 7.846(1) Å in Et4Nꢀ1.
In both Et4Nꢀ1 and Bu4Nꢀ1, seven Cuꢀatoms of each Cu8ꢀring form axial Cu−O interactions
shorter than 3.00(5) Å with the Cu13ꢀring (Et4Nꢀ1, Cu···O: 2.367(5)−3.041(5) Å, average:
2.575(4) Å; Bu4Nꢀ1, Cu···O: 2.340(4)−2.993(4) Å, average: 2.573(4) Å). The other Cuꢀatoms,
including the ones of the Cu13ꢀring, are at distances larger than 3.201(4) Å from the closest nonꢀ
bonding Oꢀatoms. Overall, there are fourteen Cu···O distances <3.00(5) Å in each nanojar, with
an average of 2.574(4) Å.
2.4.2. (Bu4N)2[CO3⊂{Cu30(µ-OH)30(µ-pz)15(µ-3,5-Me2pz)15(H2O)}] (Bu4N-2). Within the triclinic
(Pī) crystal lattice of 2, the nanojar is located on a general position (with pseudoꢀmirror
symmetry) and consists of three rings, [Cu(OH)(pz)]7, [Cu(OH)(3,5ꢀMe2pz)]14 and
[Cu9(OH)9(pz)8(3,5ꢀMe2pz)(H2O)] (Figures 3, S35−S45). The encapsulated carbonate ion is
found disordered over three positions (55/28/17), around the axis perpendicular to the central
carbon atom (Figure S35). In each orientation, the carbonate ion is bound by twelve hydrogen
bonds (average O···O: 2.90(2) Å), six each from the Cu7ꢀ and Cu9ꢀrings, four to each of the
three carbonate Oꢀatoms. The Cu7ꢀ and Cu9ꢀrings each accept seven Hꢀbonds from the Cu14ꢀring
(Cu7ꢀring, O···O: 2.787(5)−3.123(5) Å, average: 2.976(5) Å; Cu9ꢀring, O···O: 2.750(5)−3.050(5)
Å, average: 2.910(5) Å). Thus, the fourteen OH groups of the Cu14ꢀring form alternating Hꢀ
bonds with the Cu7ꢀ and Cu9ꢀrings (average O···O: 2.943(5) Å). The presence of a water
molecule within the nanojar is apparently the consequence of the steric hindrance caused by the
methyl groups of the 3,5ꢀMe2pz unit within the Cu9ꢀring. Thus, the H2O molecule fills the void
created between the Cu9ꢀ and Cu14ꢀrings, as the Cu2(ꢁꢀ3,5ꢀMe2pz) moiety moves away from the
two neighboring pyrazole units of the Cu9ꢀring (Figure 3, bottom). This H2Oꢀmolecule bridges
the pair of Cuꢀatoms linked by the only 3,5ꢀMe2pz unit of the Cu9ꢀring (Cu28···O31: 2.431(6),
Cu29···O31: 2.450(7) Å), and donates a very short hydrogen bond to an adjacent OHꢀgroup of
13

Dalton Transactions
Page 14 of 28
View Article Online
DOI: 10.1039/C6DT00847J
the Cu14ꢀring (O16−H16o···O31: 2.482(9) Å). The overall Hꢀbonding pattern in 2 is shown in
Figure S44.
Figure 3. Topꢀ and sideꢀviews of the crystal structure of 2. Color scheme: Cu7 – olive; Cu9 – red; Cu14 –
orange; CO32− − limeꢀgreen; Hꢀbonds and weak Cu···O interactions – green and black dashed lines,
respectively; Cu−O bonds to the coordinated H2O molecule – red/white stripes. Only the major
component of the disordered units, and no C−H and H2O hydrogens, counterions or solvent molecules are
shown.
The Cu7ꢀ and Cu9ꢀrings form seven and six Cu···O contacts <3.000(1) Å, respectively, with
Oꢀatoms of the Cu14ꢀring (Cu7ꢀring, Cu···O: 2.516(4)−2.809(4) Å, average: 2.618(4) Å; Cu9ꢀ
ring, Cu···O: 2.407(4)−2.984(5) Å; average: 2.541(4) Å). Two Cuꢀatoms of the Cu9ꢀring
(bridged by the 3,5ꢀMe2pz ligand) bind the bridging H2O molecule, which donates a Hꢀbond to
the Cu14ꢀring. All other Cuꢀatoms, including the ones of the Cu14ꢀring, are at distances larger
14

Page 15 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
than 3.545(4) Å from the closest nonꢀbonding Oꢀatoms. Overall, there are thirteen Cu···O
distances <3.000(1) Å inꢀbetween Cuxꢀrings, with an average of 2.548(4) Å.
2.4.3. (Bu4N)2[CO3⊂{Cu30(µ-OH)30(µ-pz)16(µ-3,5-Me2pz)14}] (Bu4N-3). Nanojar 3, consisting of
one [Cu(OH)(3,5ꢀMe2pz)]14 and two [Cu(OH)(pz)]8 rings, is located on a twoꢀfold rotation axis
(C2), which runs perpendicular to the encapsulated carbonate ion, disordered accordingly over
two positions (50/50). The position of the nanojar and encapsulated carbonate ion relative to the
C2 axis is clearly different than in Et4Nꢀ1 and Bu4Nꢀ1, in which the C2 axis is located in the plane
of the carbonate ion. The central Cu14ꢀring is made up by 3,5ꢀdimethylpyrazolate ligands, while
the two Cu8 side rings contain only nonꢀsubstituted pyrazole ligands.
2−
The CO3 ion is bound by a total of nine Hꢀbonds in each disordered position (O···O:
2.63(9)−2.97(9) Å, average: 2.87(8) Å), four from one Cu8ꢀring and five from the other, with
each Oꢀatom having three Hꢀbonds. Each Cu8ꢀring forms six Hꢀbonds with the Cu14ꢀring (O···O:
2.73(5)−2.96(5) Å, average: 2.87(5) Å). The two Cu8ꢀrings form six and eight Cu···O contacts
<3.00(7) Å, respectively, with Oꢀatoms of the Cu14ꢀring (one Cu8ꢀring, Cu···O: 2.52(3)−2.76(3)
Å, average: 2.56(3) Å; other Cu8ꢀring, Cu···O: 2.43(3)−3.10(3) Å; average: 2.67(3) Å). All other
Cuꢀatoms, including the ones of the Cu14ꢀring, are at distances larger than 3.13(3) Å from the
closest nonꢀbonding Oꢀatoms. Overall, there are fourteen Cu···O distances <3.00(7) Å, with an
average of 2.61(3) Å.
2.5. Solubilizing nanojars in hydrocarbon solvents and water. For liquidꢀliquid extraction
purposes, solubility of the nanojar in longꢀchain aliphatic solvents is desirable. Characteristics of
such solvents (typically C9−C18 alkanes/isoalkanes) include low toxicity, low vapor pressure, low
flammability (high flash point), negligible solubility in water, chemical inertness, no odor or
color, and low price. Nanojars comprised of the parent pyrazole ligand (HpzH), however, are
insoluble in aliphatic solvents.1 As illustrated in Figure 1, attachment of aliphatic chains of
increasing lengths to the pyrazole 4ꢀposition leads to a gradual increase of the solubility of the
corresponding nanojar in aliphatic solvents, such that nanojars based on 4ꢀnꢀoctylpyrazole are
readily soluble in nꢀhexadecane (C16, longest straightꢀchain alkane that is liquid at room
temperature; mp = 18 °C) and even in heavy mineral oil (C15−C50). The same is true for the 3ꢀnꢀ
octylpyrazole derivative, as well as the heteroleptic nanojars based on mixtures of nonꢀ
substituted pyrazole and 3,5ꢀdimethylꢀ4ꢀnꢀoctylpyrazole or 3ꢀnꢀbutylꢀ5ꢀnꢀhexylpyrazole.
15

Dalton Transactions
Page 16 of 28
View Article Online
DOI: 10.1039/C6DT00847J
Figure 4. Topꢀ and sideꢀviews of the crystal structure of 3. Color scheme: Cu8 – magenta; Cu13 – orange;
CO32− − limeꢀgreen (only one position of the disordered carbonate, and no H−atoms, counterions or
solvent molecules are shown).
Although all nanojars prepared so far are soluble in tetrahydrofuran, the solubility of the ones
based on 4ꢀhalopyrazoles in toluene is drastically reduced compared to the parent nanojar and the
alkylꢀ or etherꢀsubstituted ones. Thus, only the 4ꢀfluoroꢀderivative is sparingly soluble in
toluene, while the other haloꢀderivatives are practically insoluble. It is also noteworthy to
mention that the solubility of the parent nanojar (but not of the ones with alkyl chain
substituents) in toluene is lower when the counterion is a metal cation, compared to the Bu4N+ꢀ
salt.
16

Page 17 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
Some future applications of nanojars might require solubility in water. Again, nanojars based
on the parent nonꢀsubstituted pyrazole are insoluble in water.1 Polar, hydrophilic substituents,
such as sulfonate, carboxylate or hydroxyl groups, would be expected to increase the solubility
of the corresponding nanojars in water. However, those groups are also good donors (especially
in their deprotonated forms), and, in the presence of Cu2+, lead to the formation of polymeric
products instead of nanojars.31−33 To impart solubility in water while maintaining nanojar
structure, we prepared novel pyrazole ligands with oligo(ethylene glycol) methyl ether chains
attached to the 4ꢀposition, 4ꢀCH3(OCH2CH2)zOpzH (z = 1−3) (Scheme 2). While
Na2[CO3⊂{Cu(OH)(4ꢀCH3(OCH2CH2)zOpz)}n] (n = 27–31) nanojars are negligibly soluble in
water when z = 1, the corresponding ones with z = 2 have noticeable solubility, and become
readily soluble when z = 3.
2–
2.6. Liquid-liquid extraction of carbonate from water. The carbonate ion (CO3 ) is in a pHꢀ
dependent equilibrium with the bicarbonate ion (HCO3 ) and CO2.34 At neutral pH, virtually no
–
2–
CO3 ion is found in aqueous solution, while above pH~12, carbonate is found exclusively as
2–
CO3 . Given its very large hydration energy (ꢀGh° = –1315 kJ/mol), the CO32– ion is difficult to
–
extract from water, compared to HCO3 (ꢀGh° –368 kJ/mol).24 Herein we show that nanojars are
excellent hosts for the solvent extraction of carbonate at pH 8−14. The industrial solvent Isoparꢀ
L is chosen as the organic extraction medium, which is a relatively harmless, colorless and
odorless mixture of C11−C13 isoalkanes, with low vapor pressure (0.3 mmHg at 20 °C, bp
~189−209 °C, compared to hexane: 132 mmHg, bp 69 °C). Carbonate is extracted quantitatively
from a 1.0 mM aqueous Na2CO3 solution at pH 8−14 by stirring with a mixture of 4ꢀRꢀpyrazole
(R = nꢀbutyl or nꢀoctyl), copper nitrate and NaOH (1:1:2 molar ratio) in IsoparꢀL, whereby
nanojars Na2[CO3⊂{Cu(OH)(4ꢀRꢀpz)}n], identified by ESIꢀMS(−), are separated into the organic
phase. As the nanojar forms, it creates a highly hydrophilic cavity that encloses the carbonate
ion, surrounded by a highly hydrophobic outer layer, which fully isolates the ion and provides
solubility in the hydrophobic solvent. The incarcerated anion can be released as bicarbonate
and/or CO2 upon stripping the organic phase with a dilute nitric acid solution.
3. CONCLUSIONS
The present study, based on 40 differently substituted pyrazoles with one, two or three
substituents of varying length, bulkiness and coordinating ability in the 3ꢀ, 4ꢀ and 5ꢀpositions,
17

Dalton Transactions
Page 18 of 28
View Article Online
DOI: 10.1039/C6DT00847J
provides a comprehensive picture of the effect of ligand substitution on nanojar structure. We
find that reasonably bulky, nonꢀcoordinating substituents (such as alkyl chains, phenyl, CF3) are
tolerated at the pyrazole 4ꢀposition, while steric hindrance at the 3ꢀ and 5ꢀpositions will prevent
the formation of nanojars. Although long, straight chains are also tolerated at the pyrazole 3(5)ꢀ
position, no homoleptic nanojars can be obtained with 3ꢀphenylꢀ or 3ꢀCF3ꢀpyrazole, nor with any
3,5ꢀdisubstituted pyrazoles, regardless of the size of the two substituents. In those cases,
however, heteroleptic nanojars can be obtained in the presence of nonꢀsubstituted pyrazole (or,
presumably, 4ꢀsubstituted pyrazoles), with the exception of 3,5ꢀR2pyrazoles with bulky R groups
(tBu, CF3, Ph), which are unable to fit within any nanojar. Xꢀray crystallographic studies on two
of the various possible (Bu4N)2[CO3⊂{Cu30(OH)30(3,5ꢀMe2pz)y(pz)30ꢀy}] nanojars (with y = 14
and 15), obtained using a 1:1 mixture of 3,5ꢀdimethylpyrazole and nonꢀsubstituted pyrazole,
demonstrate in both cases that the disubstituted ligands are preferentially incorporated into the
larger, central [Cu(OH)(3,5ꢀMe2pz)]14 ring (Cu14). This approximately flat ring allows the
ligands to symmetrically alternate above and below the ring meanꢀplane, thus avoiding steric
hindrance between substituents on neighboring pyrazole units. In the nanojar with y = 14, the
central Cu14ꢀring is sandwiched inꢀbetween two smaller Cu8ꢀrings, which are bowlꢀshaped and
place the pyrazole ligands much closer to each other, leaving little room for substituents. In the
case of the nanojar with y = 15, the two smaller rings are [Cu(OH)(pz)]7 (Cu7) and
[Cu9(OH)9(3,5ꢀMe2pz)(pz)8] (Cu9). While the Cu7ꢀring is comprised only of nonꢀsubstituted
pyrazoles, the Cu9ꢀring does incorporate one disubstituted pyrazole. The crystal structure shows
that this bulkier ligand turns sharply away from the neighboring pyrazole units of the Cu9ꢀring, to
accommodate the substituents.
The limited ability of the smaller Cuxꢀrings to incorporate disubstituted pyrazoles is also
evident from the results of the ESIꢀMS studies. These studies show that even in the case of 3,5ꢀ
dimethylpyrazole/pyrazole ratios larger than 14:16 (such as 15:15 or 20:10), the most abundant
nanojar in the mixture is still [CO3⊂{Cu30(OH)30(3,5ꢀMe2pz)14(pz)16}]2−, in which the central
ring is fully substituted, and the two smaller sideꢀrings are not substituted at all. Not only is the
nanojar with y = 14 the most abundant species in these mixtures, but there is also an abrupt drop
in the abundance of the nanojars with y > 14, reflecting the weaker tendency of the smaller Cuxꢀ
rings to incorporate disubstituted pyrazoles. Nevertheless, homoleptic Cu30ꢀnanojars based solely
on 3ꢀsubstitued pyrazoles can be obtained, in which case the monosubstituted ligands of the
18

Page 19 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
smaller Cuxꢀrings are likely found in a propellerꢀtype arrangement, minimizing the repulsion
between substituents and neighboring pyrazole units. Regardless of the position on the pyrazole
ligand, COOH, SO3H, CH=O, OH, NH2 and NO2 groups prevent the formation of nanojars, and
either insoluble, presumably polymeric, or soluble, small complexes (Cu1−Cu6, in the case of 3ꢀ
and 4ꢀnitropyrazole) are obtained instead. This study also led to new synthesis methods of 4ꢀnꢀ
octylpyrazole, 3,5ꢀdimethylꢀ4ꢀnꢀoctylpyrazole and 4ꢀethoxypyrazole, and inspired the
preparation of novel pyrazole ligands with oligoether chain substituents, which render nanojars
soluble in water.
We have demonstrated that the solubility of nanojars in various solvents can be conveniently
tuned by varying the substitution of the pyrazole ligands. Thus, alkyl chains of increasing lengths
impart increasing solubility in aliphatic solvents, rendering nanojars amenable for the largeꢀscale
liquidꢀliquid extraction of anions with large hydration energies from aqueous solutions. The
extraction process is exemplified here by the efficient transfer of carbonate from water to an
aliphatic solvent, which is unprecedented for an anion with such large hydration energy. We are
currently exploring the removal of toxic anions, such as arsenate, chromate and selenate, from
water using nanojars as extracting agents.
4. EXPERIMENTAL SECTION
4.1. Materials and methods. 4ꢀHydroxypyrazole,35 3ꢀhydroxypyrazole,36 4ꢀfluoropyrazole,37 4ꢀ
chloropyrazole,38 4ꢀbromopyrazole,38 4ꢀiodopyrazole,29 4ꢀnitropyrazole,39 4ꢀformylpyrazole,40 3ꢀ
formylpyrazole,41 pyrazoleꢀ4ꢀsulfonic acid,42 3,5ꢀdiethylpyrazole,43 3(5)ꢀnꢀbutylꢀ5(3)ꢀnꢀ
hexylpzH,44
3,5ꢀdiꢀtertꢀbutylpyrazole,45
3,5ꢀbis(trifluoromethyl)pyrazole,46
3,5ꢀ
diphenylpyrazole,47 4ꢀphenylpyrazole,48 4ꢀnꢀbutylpyrazole,49 4ꢀ(3ꢀhydroxypropyl)pyrazole50 and
3(5)ꢀnꢀalkylpyrazoles51 (alkyl = Me, Et, Pr, Bu, Oct) are prepared according to published
procedures. All other commercially available chemicals are used as received. Reactions are
carried out in the air, unless stated otherwise. Hexane (mixture of isomers) and
dimethylformamide are dried with molecular sieves.
Mass spectrometric analysis of the nanojars is performed on a Waters Synapt G1 HDMS
instrument, using an electrospray ionization source. 10–4−10–5 M sample solutions are prepared
in CH3CN, except for the nanojars with butyl or longer alkyl chain substituents, in which case a
1:1 (vol.) mixture of CH3CN/THF is used. Samples are infused by a syringe pump at 5 ꢁL/min
19

Dalton Transactions
Page 20 of 28
View Article Online
DOI: 10.1039/C6DT00847J
and nitrogen is supplied as nebulizing gas at 500 L/h. The electrospray capillary voltage is set to
–2.5 or +3.0 kV, respectively, with a desolvation temperature of 110 °C. The sampling and
extraction cones are maintained at 40 V (100 V for nanojars with long alkyl chain substituents)
and 4.0 V, respectively, at 80 °C.
4.2. Decanal diethyl acetal. Decanal (47.1 mL, 39.1 g, 250 mmol), triethyl orthoformate
(41.6 mL, 37.1 g, 250 mmol) and pꢀtoluenesulfonic acid monohydrate (476 mg, 2.50 mmol) are
1
dissolved in 250 mL absolute ethanol. After standing for 1 hour at room temperature, H NMR
indicates that the reaction is practically complete. The solution is left to stand overnight, then it is
poured into 500 mL of saturated sodium bicarbonate solution (~96 g NaHCO3/L H2O at 20 °C),
and is extracted with ethyl acetate (5 × 100 mL). The combined organic extracts are dried
overnight with anhydrous Na2SO4, followed by evaporation of the solvent and drying under high
1
vacuum to yield 56.4 g (98%) of decanal diethyl acetal. H NMR (400 MHz, CDCl3): δ 4.46 (t,
3
1H, 3J = 6 Hz), 3.62 (dq, 2H, 3J = 7 Hz), 3.47 (dq, 2H, J = 7 Hz), 1.58 (m, 2H), 1.25 (m, 14H),
3
1.19 (t, 6H, J = 7 Hz), 0.86 (t, 3H, 3J = 7 Hz) ppm.
4.3. General method of preparation for 7-ethoxy-2,5,8-trioxadecane, 10-ethoxy-2,5,8,11-
tetraoxatridecane and 13-ethoxy-2,5,8,11,14-pentaoxahexadecane. NaH (60% dispersion in
mineral oil, 4.000 g, 100 mmol) is placed in a 250 mL threeꢀneck, roundꢀbottom flask, which is
connected to a Schlenk line. A condenser is attached, the third neck is sealed with a rubber
septum, and the system is evacuated and purged with nitrogen three times. The mineral oil is
washed away from the sodium hydride by rinsing with anhydrous hexane. The hexane (50 mL) is
introduced with an N2ꢀpurged syringe, and the mixture is stirred for 3 minutes. After the NaH has
settled, the supernatant is removed with a syringe, and the procedure is repeated three more
times. The residue is then dried thoroughly by applying high vacuum. Neat bromoacetaldehyde
diethyl acetal (100 mL, 131 g, 0.664 mol) is added by syringe under stirring, followed by the
dropwise addition of the oligo(ethylene glycol) monomethyl ether, CH3(OCH2CH2)zOH (50.00
mmole; z = 1, 3.950 mL, 3.8045 g; z = 2, 5.800 mL, 6.007 g; z = 3, 8.000 mL, 8.210 g). The
reaction is exothermic and is accompanied by a color change from light grey to brown. After
heating to 100 °C for 48 hours under stirring, the reaction mixture is left to cool down to room
temperature. Water (20 mL) is added to quench the unreacted NaH, and the excess
bromoacetaldehyde diethyl acetal is distilled off under high vacuum at ~30 °C. The dark brown
viscous residue is taken up into dichloromethane (200 mL) and is washed with water (40 mL)
20

Page 21 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
twice, followed by drying the organic layer with MgSO4 overnight. After filtration, the solvent is
removed under vacuum at ~ 30 °C and the product is obtained as redꢀbrown viscous oil in close
to quantitative yield. Although the product is of sufficient purity to be used in the next step,
vacuum distillation can be employed if further purification is needed.
4.3.1. 7-Ethoxy-2,5,8-trioxadecane. Obtained as colorless oil after distillation (b.p. 47–49 °C at
1
3
0.005 mmHg). H NMR (400 MHz, CDCl3): δ 4.63 (t, 1H, J = 5 Hz, CH(OCH3CH2)2), 3.65–
3.68 (m, 4H, OCH2CH3), 3.52–3.59 (m, 6H, OCH2CH2 and OCH2CH), 3.36 (s, 3H, OCH3), 1.20
13
(t, 6H, OCH2CH3) ppm. C NMR (101 MHz, CDCl3): δ 101.3, 72.03, 70.9, 70.96, 62.4, 59.1
ppm. HRMS (ESIꢀTOF) m/z: [M+Na]+ calcd for C9H20NaO4 215.1259; found 215.1254.
4.3.2. 10-Ethoxy-2,5,8,11-tetraoxatridecane. Obtained as colorless oil after distillation (b.p. 72–
1
73 °C at 0.005 mmHg). H NMR (400MHz, CDCl3): δ 4.67 (t, 1H, CH(OCH3CH2)2), 3.50–3.79
(m, 14H, OCH2 and OCH2CH3), 3.35 (s, 3H, OCH3), 1.19 (t, 3H, CH2CH3) ppm. 13C NMR (101
MHz, CDCl3): δ 101.4, 72.0, 71.7, 70.9, 70.7, 70.6, 65.7, 62.5, 59.2, 15.4 ppm. HRMS (ESIꢀ
TOF) m/z: [M+Na]+ calcd for C11H24NaO5 259.1521; found 259.1514.
4.3.3. 13-Ethoxy-2,5,8,11,14-pentaoxahexadecane. Obtained as brown viscous oil. 1H NMR (400
3
MHz, CDCl3): δ 4.62 (t, 1H, J = 6 Hz, CH(OCH3CH2)2), 3.62–3.66 (m, 12H, CH2CH3 and
13
OCH2CH2), 3.51–3.59 (m, 6H, OCH2CH2), 3.36 (s, 3H, OCH3), 1.20 (t, 6H, CH2CH3) ppm. C
NMR (101 MHz, CDCl3): δ 101.3, 72.012, 71.967, 70.965, 70.705, 70.667, 70.664, 70.613, 62.4,
59.1 ppm.
4.4. General method of preparation for 3-R-2-octyl-prop-2-enal, 3-R-2-(2-methoxyethoxy)-
prop-2-enal,
3-R-2-(2-(2-methoxyethoxy)ethoxy)-prop-2-enal
and
3-R-2(2-(2-(2-
methoxyethoxy)ethoxy)ethoxy)-prop-2-enal (R = NMe2 and OEt).
POCl3 (6.38 ml, 10.5 g, 68.4 mmol) is added to a 100 mL Schlenkꢀflask equipped with a
pressureꢀequalizing addition funnel, sealed with a rubber septum. The system is evacuated and
purged with nitrogen a few times while cooling in an ice bath. DMF (5.86 ml, 5.53 g, 75.7
mmol) is injected into the additional funnel via N2ꢀpurged syringe, and is added dropwise to
POCl3 under stirring. Decanal diethyl acetal (7.26 g, 31.5 mmol) or
CH3(OCH2CH2)zOCH2CH(OEt)2 (31.5 mmol; z = 1, 6.05 g; z = 2, 7.43 g; z = 3, 8.82 g) is added
via syringe to the colorless crystalline Vilsmeier reagent obtained. The ice bath is replaced with a
water bath and the reaction mixture is allowed to warm up to room temperature. The color of the
reaction slowly turns yellow. The temperature of the water bath is raised to 70−75 °C, causing
21

Dalton Transactions
Page 22 of 28
View Article Online
DOI: 10.1039/C6DT00847J
the color to quickly turn dark orange and then dark red. A vigorous reaction ensues,
accompanied by a large increase in pressure (relieved through the Schlenk line’s oil bubbler).
The reaction mixture is kept at the same temperature for 2 hours, then the dark brownꢀred,
viscous solution is poured into 50 g crushed ice and is left to stand overnight. Next day, the
mixture is extracted with CH2Cl2 (2 × 25 mL), followed by Et2O (25 mL). The aqueous phase is
neutralized with solid K2CO3 (20 g), which is added in small portions, under stirring, until pH ≈
8. The orangeꢀred solution is extracted again with CH2Cl2 (5 × 25 mL) and Et2O (2 × 20 mL).
The combined organic extracts are dried overnight with anhydrous MgSO4. After filtration, the
solvent is evaporated under vacuum and DMF is distilled off in vacuum on a water bath at 80 °C.
The resulting dark redꢀbrown oil consists of a mixture of 3ꢀ(dimethylamino)ꢀ2ꢀnꢀoctylꢀpropꢀ2ꢀ
enal and 3ꢀethoxyꢀ2ꢀnꢀoctylꢀpropꢀ2ꢀenal in the case of CH3(CH2)8CH(OEt)2, and OEGꢀ
substituted dimethylaminoꢀ and ethoxyꢀpropenal, as well as smaller amounts of EtOꢀsubstituted
dimethylaminoꢀ and ethoxyꢀpropenal in the case of CH3(OCH2CH2)zOCH2CH(OEt)2. These
mixtures are used without separation in the next step.
4.5. General method of preparation for 4-octylpyrazole, 4-(2-methoxyethoxy)pyrazole, 4-(2-
(2-methoxyethoxy)ethoxy)pyrazole and 4-(2-(2-(2-methoxyethoxy)ethoxy)ethoxy)pyrazole.
The mixtures obtained above are dissolved in methanol (60 mL), hydrazine monohydrate (1.95
ml, 1.98 g, 39.5 mmol) is added and the mixture is refluxed for 2 hours. After cooling to room
temperature, the solvent is removed under vacuum. The dark redꢀbrown oil is subjected to
purification by vacuum distillation or column chromatography.
4.5.1. 4-Octylpyrazole. Vacuum distillation using a Vigreux column provides pure 4ꢀ
octylpyrazole as a colorless liquid (b.p. 143 °C at ~0.1 mmHg), which solidifies on cooling.
Yield: 4.21 g (75% based on decanal). 1H NMR (400 MHz, CDCl3): δ 12.07 (s, br, 1H), 7.40 (s,
2H), 2.49 (t, 2H, 3J = 8 Hz), 1.56 (m, 2H), 1.30 (m, 10H), 0.90 (t, 3H, 3J = 7 Hz) ppm. 13C NMR
(101 MHz, CDCl3): δ 132.5, 121.2, 32.0, 31.2, 29.6, 29.4, 24.2, 22.8, 14.2 ppm.
4.5.2. 4-(2-methoxyethoxy)pyrazole. Purified by column chromatography on silica gel (300 g)
using neat ethyl acetate as eluent (Rf 0.49). The product (1.52 g) is obtained as an orange oil in
1
34% overall yield (based on ethylene glycol monomethyl ether). H NMR (400MHz, CDCl3): δ
7.29 (s, 2H, 3,5ꢀHꢀpz), 4.01–4.03 (m, 2H, OCH2), 3.67–3.70 (m, 2H, OCH2), 3.42 (s, 3H, OCH3)
ppm. 13C NMR (101 MHz, CDCl3): δ 145.5, 121.1, 71.1, 59.2 ppm. HRMS (ESIꢀTOF) m/z:
[M+Na]+ calcd for C6H10N2NaO2 165.0640; found 165.0643.
22

Page 23 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
4.5.3. 4-(2-(2-methoxyethoxy)ethoxy)pyrazole. Purified by column chromatography on silica gel
(300 g) using neat ethyl acetate as eluent (Rf 0.35). The product (1.50 g) is obtained as an
1
orangeꢀred oil in 26% overall yield (based on diethylene glycol monomethyl ether). H NMR
(400 MHz, CDCl3): δ 7.28 (s, 2H, 3,5ꢀHꢀpz), 4.02 (t, 2H, OCH2, 3J = 4 Hz), 3.78 (t, 2H, OCH2,
3
3
3J = 4 Hz), 3.68 (t, 2H, OCH2, J = 4 Hz), 3.56 (t, 2H, OCH2, J = 4 Hz), 3.37 (s, 3H, OCH3)
13
ppm. C NMR (101 MHz, CDCl3): δ 145.5, 120.9, 71.9, 71.2, 70.7, 69.9, 59.1 ppm. HRMS
(ESIꢀTOF) m/z: [M+Na]+ calcd for C8H14N2NaO3 209.09021; found 209.0902.
4.5.4. 4-(2-(2-(2-methoxyethoxy)ethoxy)ethoxy)pyrazole. Purified by column chromatography on
silica gel (300 g) using an eluent gradient as follows: 95:5 EtOAc:hexanes first, neat EtOAc
second, 95:5 EtOAc:MeOH third, and 90:10 EtOAc:MeOH last. The product (1.44 g) is obtained
as an orangeꢀyellow oil in 20% overall yield (based on triethylene glycol monomethyl ether). 1H
NMR (400 MHz, CDCl3): δ 7.28 (s, 2H, 3(5)ꢀHꢀpz), 4.00–4.02 (m, 2H, OCH2), 3.76–3.79 (m,
2H, OCH2), 3.62–3.71 (m, 6H, CH2OCH2CH2), 3.53–3.55 (m, 2H, OCH2), 3.36 (s, 3H, OCH3).
13C NMR (100 MHz, CDCl3): δ 145.5, 121.0, 71.9, 71.3, 70.7, 70.6, 70.5, 69.9, 59.1 ppm.
HRMS (ESIꢀTOF) m/z: [M+Na]+ calcd for C10H18N2NaO4 253.1164; found 253.1173.
4.6. 4-Ethoxypyrazole. Obtained as a sideꢀproduct during the column chromatography of the
monoꢀ, diꢀ and triethyleneglycol methyl etherꢀsubstituted pyrazoles described above, in 5:8, 5:10
1
and 5:7 molar ratios, respectively (4ꢀEtOpyrazole:4ꢀOEGpyrazole). H NMR (400 MHz,
3
3
CDCl3): δ 7.28 (s, 2H, 3,5ꢀHꢀpz), 3.94 (q, 2H, J = 7 Hz, CH2CH3), 1.37 (t, 3H, J = 7 Hz,
CH2CH3). 13C NMR (101 MHz, CDCl3): δ 145.6, 120.8, 67.5, 15.0 ppm. HRMS (ESIꢀTOF) m/z:
[M+H]+ calcd for C5H9N2O 113.0715; found 113.0695.
4.7. 3,5-Dimethyl-4-octylpyrazole. First, 3ꢀ(nꢀoctyl)pentaneꢀ2,4ꢀdione is prepared as follows. A
mixture of acetylacetone (30.0 ml, 29.2 g, 0.292 mol) and 1ꢀiodooctane (53.0 ml, 70.5 g, 0.293
mol) was added dropwise to a stirred suspension of anhydrous K2CO3 (50.0 g, 0.362 mol) in 50
mL acetonitrile. The reaction mixture was refluxed overnight with stirring. After cooling to room
temperature, the solid was filtered off and washed with acetone. The organic solvents were
evaporated on a rotavap, water was added to the residue and it was extracted with diethyl ether.
After evaporation of the ether, the residue (60.7 g) was fractionally distilled under vacuum using
a Vigreux column, and the fraction distilling at 104 °C (43.5 g) was redistilled. The fraction
distilling at 117 °C (41.0 g, yield: ~66%) was found to be a mixture of the keto and enol forms of
23

Dalton Transactions
Page 24 of 28
View Article Online
DOI: 10.1039/C6DT00847J
3ꢀ(nꢀoctyl)pentaneꢀ2,4ꢀdione, with a small amount of Oꢀalkylated enol. This mixture was used in
the next step without further purification.
Hydrazine hydrate (N2H4 50−60% in H2O, 12.414 g) is added dropwise to a solution of 3ꢀ(nꢀ
octyl)pentaneꢀ2,4ꢀdione (38.851 g) in 150 mL EtOH with stirring. The mixture is refluxed for 14
hours and the solvent is evaporated on a rotavap. The residue (38.664 g) is fractionally distilled
under vacuum, using a Vigreux column. The fraction distilling at 155−160 °C (32.721 g) is
redistilled, yielding 32.125 g (56% based on acetylacetone) of pure 3,5ꢀdimethylꢀ4ꢀnꢀ
1
3
octylpyrazole boiling at 160 °C. H NMR (400 MHz, CDCl3): 2.31 (t, 2H, J = 8 Hz), 2.19 (s,
3
6H), 1.41 (m, 2H), 1.27 (m, 10H), 0.87 (t, 3H, J = 7 Hz) ppm. 13C NMR (101 MHz, CDCl3):
142.0, 115.8, 32.0, 30.8, 29.6, 29.5, 29.4, 23.1, 22.8, 14.2, 11.0.
4.8. General methods of preparation for nanojars. In each case, the identity of the resulting
nanojar mixtures is determined using ESIꢀMS (Table 1).
Method A. Cu(NO3)2·2.5H2O (1.000 g, 4.30 mmol), pyrazole ligand (4.30 mmol), NaOH (332
mg, 8.30 mmol), Na2CO3·H2O (533 mg, 4.30 mmol) and Bu4NOH (1M in H2O, 318 mg, 0.318
mmol) are stirred in THF (20 mL) for 3 days. After filtration, the deep blue solution is left to
evaporate. The dark blue residue is stirred with THF (15 mL), then filtered and slowly added to
deionized water (100 ml) under stirring. The blue precipitate is filtered, washed with water and
dried in vacuum to yield (Bu4N)2[CO3⊂{Cu(OH)(pz)}n] (n = 27−31). In the case of mixedꢀ
pyrazole nanojars, 2.15 mmol of each pyrazole ligand is employed. Et4NOH can also be used
instead of Bu4NOH.
Method B. Same as above, except that Cu(NO3)2·2.5H2O and Na2CO3·H2O are replaced by
CuCO3·Cu(OH)2 (475 mg, 2.15 mmol).
Method C. Same as method A, except that MIOH (8.60 mmol; MI = Li+, Na+, K+, Rb+, Cs+) or
MII(OH)2·8H2O (4.30 mmol; MII = Sr2+, Ba2+) are used as base and counterionꢀsource instead of
NaOH/Bu4NOH. The corresponding metal carbonate is employed as carbonateꢀsource.
Depending on the base used, MI2[CO3⊂{Cu(OH)(pz)}n] or MII[CO3⊂{Cu(OH)(pz)}n] (n =
27−31) nanojars are obtained. In the case of KOH, 18ꢀcrownꢀ6 (137 mg, 0.518 mmol) can be
added to the reaction mixture to obtain (K+⊂18ꢀcrownꢀ6)2[CO32−⊂{Cu(OH)(pz)}n] (n = 27−31).
4.9. Liquid-liquid extraction of carbonate from water. To a 1.0 mM aqueous solution of
Na2CO3 at pH 8−14 (7.9 mg anhydrous Na2CO3 in 75 mL NaOH 10−6−1 M) is added
Cu(NO3)2·2.5H2O (500 mg, 2.15 mmol), 4ꢀRꢀpyrazole (R = nꢀbutyl, 267 mg, 2.15 mmol; R = nꢀ
24

Page 25 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
octyl, 385 mg, 2.15 mmol), NaOH (172 mg, 4.30 mmol) and ISOPARꢀL (75 mL), and the
mixture is stirred vigorously under N2. The organic phase immediately starts turning blue, and
the aqueous phase gradually turns colorless. ESIꢀMS of the deepꢀblue organic phase indicates
the presence of [CO3⊂{Cu(OH)(4ꢀRꢀpz)}n]2−, at m/z 2780 (Cu27), 2984 (Cu29), 3188 (Cu31) (R =
nꢀbutyl), and m/z 3538 (Cu27), 3798 (Cu29), 4058 (Cu31) (R = nꢀoctyl), respectively.
4.10. X-ray crystallography. Single crystals are grown by hexane vapor diffusion to a toluene
solution of the nanojars. Once removed from the mother liquor, the crystals are extremely
sensitive to solvent loss at ambient conditions and are mounted quickly under a cryostream (100
K) to prevent decomposition. Xꢀray diffraction data are collected at 100 K from a singleꢀcrystal
mounted atop a glass fiber under ParatoneꢀN oil, with a Bruker SMART APEX II diffractometer
using graphiteꢀmonochromated MoꢀKα (λ = 0.71073 Å) radiation. The structures are solved by
employing SHELXTL direct methods and refined by fullꢀmatrix least squares on F2, using the
APEX2 v2014.9ꢀ0 software package.52 All nonꢀH atoms are refined with independent anisotropic
displacement parameters, except the disordered molecules. C–H hydrogen atoms are placed in
idealized positions and refined using the riding model. O–H hydrogen atoms are located from the
difference Fourier maps; their displacement parameters are fixed to be 20 % larger than those of
the attached O atoms. For the disordered molecules, geometrical restraints are used and Hꢀatoms
are not assigned. Crystallographic details are summarized in Table 2, and thermal ellipsoid plots,
as well as structural details, are shown in the Supporting Information. Crystallographic data have
been deposited at the Cambridge Crystallographic Data Centre (deposition numbers: CCDC
1455325 and 1455326). Copies of the data can be obtained free of charge via
http://www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data
Centre, 12, Union Road, Cambridge, CB2 1EZ, UK; fax: +44 1223 336033; eꢀmail:
deposit@ccdc.cam.ac.uk).
2–
(Et4N)2[CO3 ⊂{cisꢀCuII(ꢁꢀOH)(ꢁꢀpz)}8+13+8](C7H8)4(H2O)2: Et4Nꢀ1(C7H8)4(H2O)2. One
Cu(OH)(pz)Cu unit of the Cu13ꢀring is disordered (50/50) abount the twoꢀfold rotation symmetry
axis running through the nanojar. The toluene solvent molecules are disordered over two
positions (50/50); one of them is disordered about the C2 rotation axis.
2–
II
(Bu4N)2[CO3 ⊂{cisꢀCuII(ꢁꢀOH)(ꢁꢀpz)}7{cisꢀCuII(ꢁꢀOH)(ꢁꢀ3,5ꢀMe2pz)}14{cisꢀCu9 (ꢁꢀOH)9
(ꢁꢀpz)8(ꢁꢀ3,5ꢀMe2pz)(H2O)}](C7H8)5: Bu4Nꢀ2(C7H8)5. One pyrazolate unit of the Cu7ꢀring, one
dimethylpyrazolate unit of the Cu14ꢀring and one pyrazolate unit of the Cu9ꢀring are disordered
25

Dalton Transactions
Page 26 of 28
View Article Online
DOI: 10.1039/C6DT00847J
over two positions (50/50, 40/60, 40/60). A terminal CH3 group of one of the two TBA
counterions is disordered over two positions (50/50), and the carbonate ion is disordered over
three positions (55/28/17). Four of the five toluene solvent molecules are disordered. Hydrogens
for the water molecule could not be located form the residual electron density map.
2–
(Bu4N)2[CO3 ⊂{cisꢀCuII(ꢁꢀOH)(ꢁꢀpz)}8{cisꢀCuII(ꢁꢀOH)(ꢁꢀ3,5ꢀMe2pz)}14{cisꢀCuII(ꢁꢀOH)(ꢁꢀ
pz)}8]: Bu4Nꢀ3. Only very small crystals of Bu4Nꢀ3 could be obtained, which diffracted poorly;
therefore, only preliminary data are presented here (monoclinic, P2/n, a = 23.864(4) Å, b =
14.830(2) Å, c = 31.070(4) Å, β = 95.138(10), V = 10952(3) Å3, Z = 2). Although the severely
disordered solvent molecules and Bu4N+ counterions are not modeled, the identity of the nanojar
is established unambiguously, and all Cuꢀatoms are refined anisotropically (R1 = 16%, Rw =
20%).
Table 2. Crystallographic data of nanojars Et4Nꢀ1 and Bu4Nꢀ2.
Et4Nꢀ1(H2O)2(C7H8)4
Bu4Nꢀ2(C7H8)5
Formula
FW
C132H192Cu29N60O34
5006.10
C188H294Cu30N62O34
5873.03
Crystal system
Space group
a/Å
monoclinic
C2/c
triclinic
P ī
30.0687(6)
26.6625(5)
26.2115(5)
90.000
22.5542(3)
23.0479(3)
26.4003(3)
90.645(1)
102.627(1)
118.805(1)
b/Å
c/Å
α/deg
117.373(1)
90.000
β/deg
γ/deg
V/Å3
18661.0(6)
4
11625.0(3)
2
Z
Dcalc/g cm−3
1.782
1.678
/mm−1
3.304
2.754
ꢁ
Reflns collected/unique
R(int)
225664/23223
0.1191
305982/41130
0.0587
14010
30705
Obsd reflns [I > 2σ(Ι )]
Data/parameters/restrains
GOF (on F2)
23223/1205/34
1.063
41130/2867/57
1.034
0.0556, 0.1180
0.0483, 0.1296
R(F), Rw(F) [I > 2σ(Ι )]
26

Page 27 of 28
Dalton Transactions
View Article Online
DOI: 10.1039/C6DT00847J
R(F), Rw(F) [all data]
0.1201, 0.1481
0.0754, 0.1508
ACKNOWLEDGEMENTS. This material is based upon work supported by the National
Science Foundation under Grant No. CHEꢀ1404730. B.C. thanks the American Chemical Society
for financial support through the Project SEED endowment.
REFERENCES
1 I. R. Fernando, S. A. Surmann, A. A. Urech, A. M. Poulsen and G. Mezei, Chem. Commun.,
2012, 48, 6860−6862.
2 G. Mezei, Chem. Commun., 2015, 51, 10341−10344.
3 B. M. Ahmed, B. R. Szymczyna, S. Jianrattanasawat, S. A. Surmann and G. Mezei, Chem. Eur.
J., 2016, DOI: 10.1002/chem.201600271.
4 B. L. Jacobson and F. A. Quiocho, J. Mol. Biol., 1988, 204, 783−787.
5 (a) M. Elias, A. Wellner, K. GoldinꢀAzulay, E. Chabriere, J. A. Vorholt, T. J. Erb and D. S.
Tawfik, Nature, 2012, 491, 134–137; (b) D. Gonzalez, M. Richez, C. Bergonzi, E. Chabriere and
M. Elias, Sci. Rep., 2014, 4, 6636.
6 M. J. Langton, C. J. Serpell and P. D. Beer, Angew. Chem. Int. Ed., 2016, 55, 1974−1987.
7 N. Busschaert, C. Caltagirone, W. Van Rossom and P. A. Gale, Chem. Rev., 2015, 115,
8038−8155.
8 R. Dutta and P. Ghosh, Chem. Commun., 2014, 50, 10538−10554.
9 R. Custelcean, Chem. Commun., 2013, 49, 2173−2182.
10 I. Ravikumar and P. Ghosh, Chem. Soc. Rev., 2012, 41, 3077–3098.
11 N. H. Evans and P. D. Beer, Angew. Chem. Int. Ed., 2014, 53, 11716–11754.
12 R. Custelcean, Chem. Soc. Rev., 2014, 43, 1813−1824.
13 A. Frontera, Coord. Chem. Rev., 2013, 257, 1716−1727.
14 S. Lee, C.ꢀH. Chen and A. H. Flood, Nat. Chem., 2013, 5, 704−710.
15 N. Lopez, D. J. Graham, R. McGuire Jr., G. E. Alliger, Y. ShaoꢀHorn, C. C. Cummins, D. G.
Nocera, Science, 2012, 335, 450−453.
16 R. B. P. Elmes and K. A. Jolliffe, Chem. Commun., 2015, 51, 4951−4968.
17 P. A. Gale, C. Caltagirone, Chem. Soc. Rev., 2015, 44, 4212−4227.
18 A. Caballero, F. Zapata, P. D. Beer, Coord. Chem. Rev., 2013, 257, 2434−2455.
19 S.ꢀK. Ko, S. K. Kim, A. Share, V. M. Lynch, J. Park, W. Namkung, W. Van Rossom, N.
Busschaert, P. A. Gale, J. L. Sessler and I. Shin, Nat. Chem., 2014, 6, 885−892.
20 P. A. Gale, R. PérezꢀTomás and R. Quesada, Acc. Chem. Res., 2013, 46, 2801–2813.
21 J. A. Cooper, S. T. G. Street and A. P. Davis, Angew. Chem. Int. Ed., 2014, 53, 5609–5613.
22 N. Busschaert and P. A. Gale, Angew. Chem. Int. Ed., 2013, 52, 1374–1382.
23 A. Vargas Jentzsch and S. Matile, Top. Curr. Chem., 2015, 358, 205–240.
24 Y. Marcus, Ion Properties, Marcel Dekker, New York, 1997.
25 P. A. Angaridis, P. Baran, R. Boča, F. CervantesꢀLee, W. Haase, G. Mezei, R. G. Raptis and
R. Werner, Inorg. Chem., 2002, 41, 12219−2228.
27

Dalton Transactions
Page 28 of 28
View Article Online
DOI: 10.1039/C6DT00847J
26 G. Mezei, M. RiveraꢀCarrillo and R. G. Raptis, Inorg. Chim. Acta, 2004, 357, 3721–3732.
27 G. Mezei, M. RiveraꢀCarrillo and R. G. Raptis, Dalton Trans., 2007, 37−40.
28 L. Mathivathanan, M. RiveraꢀCarrillo and R. G. Raptis, Inorg. Chim. Acta, 2012, 391, 201–
205.
29 G. Mezei and R. G. Raptis, Inorg. Chim. Acta, 2004, 357, 3279−3288.
30 W. Henderson and J. S. McIndoe, Mass Spectrometry of Inorganic and Organometallic
Compounds, Wiley, 2005, pp. 117−119.
31 I. R. Fernando, N. Daskalakis, K. D. Demadis and G. Mezei, New J. Chem., 2010, 34, 221–
235.
32 E. Quartapelle Procopio, F. Linares, C. Montoro, V. Colombo, A. Maspero, E. Barea and J. A.
R. Navarro, Angew. Chem. Int. Ed., 2010, 49, 7308 –7311.
33 S. Su, Y. Zhang, M. Zhu, X. Song, S. Wang, S. Zhao, S. Song, X. Yang and H. Zhang, Chem.
Commun., 2012, 48, 11118–11120.
34 W. G. Mook, Chemistry of carbonic acid in water, in Environmental Isotopes in the
Hydrological Cycle: Principles and Applications, Vol. 1, IAEA / UNESCO: Paris, 2001, pp.
87−98.
35 T. A. Crowell, C. D. Jones and A. J. Shuker, US Patent 6046227, 2000.
36 G. A. Eller and W. Holzer, Molbank, 2006, 2006, M464.
37 K. England, H. Mason, R. Osborne and L. Roberts, Tetrahedron Lett., 2010, 51, 2849–2851.
38 M. K. Ehlert, S. J. Rettig, A. Storr, R. C. Thompson and J. Trott, Can. J. Chem., 1991, 69,
432−439.
39 K. P. Maresca, D. J. Rose and J. Zubieta, Inorg. Chim. Acta, 1997, 260, 83−88.
40 K. Takagi, A. Bajnati and M. HubertꢀHabart, Bull. Soc. Chim. France, 1990, 127, 660−666.
41 H. Bredereck, R. Sell and F. Effenberger, Chem. Ber., 1964, 97, 3407−3417.
42 C. S. Rondestvedt, Jr. and P. K. Chang, J. Am. Chem. Soc., 1955, 77, 6532–6540.
43 T. Ruman, Z. Ciunik and S. Wolowiec, Polyhedron, 2003, 22, 581−586.
44 B. M. Ahmed, H. Zhang, Y. Mo and G. Mezei, J. Org. Chem., 2016, 81, 1718–1722.
45 G. Yang and R. G. Raptis, Inorg. Chim. Acta, 2003, 352, 98−104.
46 A. Maspero, G. B. Giovenzana, D. Monticelli, S. Tagliapietra, G. Palmisano and A. Penoni, J.
Fluorine Chem., 2012, 139, 53–57.
47 N. Kitajima, K. Fujisawa, C. Fujimoto, Y. Moroꢀoka, S. Hashimoto, T. Kitagawa, K. Toriumi,
K. Tatsumi and A. Nakamura, J. Am. Chem. Soc., 1992, 114, 1277−1291.
48 J. Olguín and S. Brooker, New J. Chem., 2011, 35, 1242–1253.
49 D. L. Reger, J. R. Gardinier, T. C. Grattan, M. R. Smith and M. D. Smith, New J. Chem., 2003,
27, 1670–1677.
50 R. G. Jones and M. J. Mann, J. Am. Chem. Soc., 1953, 75, 4048−4052.
51 B. M. Ahmed and G. Mezei, RSC Adv., 2015, 5, 24081–24093.
52 APEX2 v2014.9ꢀ0; Bruker AXS Inc.: Madison, WI, 2014.
28