Reacting HL4 with zinc acetate afforded a white crystalline powder that was recrystallized from hot methanol affording a trinuclear complex [Zn₃(L4)6(H₂O)6] which crystallizes in a trigonal space group (R-3c). In this complex three Zn^II ions are found in octahedral coordination (Figure 1(a)) with HL4 acting as a μ-N1,N2-bridging ligand. Indeed, the central Zn^II ion is surrounded by six nitrogen atoms belonging to six ligands with Zn1-N1 = 2.159(4) Å and forms a perfect octahedral geometry of ZnN₆ (N1–Zn1–N1, 180°(3)), while terminal Zn^II ions are coordinated by three nitrogen atoms (Zn2-N5, 2.145(4) Å) from three triazole ligands and three water molecules (Zn2–O14, 2.107(4) Å) forming a ZnN₃O₃ coordination sphere (with N5-Zn2–O14, 178.4(1)°).

A total of six HL4 ligands bridge in a bidentate fashion to three Zn^II ions which are equidistant (Zn⋯Zn = 3.848 Å). This distance coincidently matches with that of a known binuclear Zn^II metalloenzyme from B.cereus (3.848 and 4.365 Å) (Figure 1(b)) [59] and is typical of trinuclear zinc complexes, e.g., [Zn₃(etrz)₆(H₂O)₆](CF₃SO₃)₆ [60] with Zn⋯Zn = 3.815(1) Å. Across the central Zn^II, on either side, the tripod of 1,2,4-triazole planes adopt a ‘staggered’ arrangement due to steric requirements and interestingly three aqua ligands on each terminal zinc too adopt the same conformation. The three ligands bridging two Zn^II ions form a ‘paddle wheel’ like motif with triazoles planes as propellers. The amine derivatization keeps thioether and carboxylic group to lie on either side of the 1,2,4-triazole plane as pendent arms away from the coordination sphere due to favorable and dominant bridging mode by triazole. These dangling arms are only involved in supramolecular interactions. The torsion angle C10-C11-S12A-C13A is 54.8° which is completely different from that of β-DL methionine which is 174.9° [61,62], and all the carbon atoms and sulfur atoms form an almost-planar zigzag chain. But the torsion angle C7-C6-C10-C11 of 173.3° indicates that main chain carbon atoms lie in a plane away from thiomethyl group.

What makes this polynuclear complex distinctive is due to the dual contribution from its ligand. Here HL4 not only acts as a bridging ligand but also as anion due to the presence of deprotonable carboxylic group. Such ionized carboxylate group was reported for a Pt^II complex where coordination from nitrogen and thiomethyl sulfur is observed thus creating a new sterogenic centre at the sulfur atom [63]. A recent report indicates the use of soft-hard recognition principle in methionine to construct a 3D homochiral coordination network [64]. Choice of O/N/S donor preference for coordination from methionine largely depends on this hard-soft concept over steric encumbrance. In 1 the charge on each metal ion is balanced by the deprotonation of carboxylic group which are engaged in H-bonding. The bond length (C7–O9, 1.223(8) Å; C7–O8, 1.213(8) Å) indicates the resonance stabilized double bond after deprotonation. The ethyl(methyl)sulfane group (–CH₂–CH₂–S–CH₃) was found to be two-fold disordered. All coordinated water molecules are involved in H-bonding, with oxygen atom of carboxylate group belonging to different molecules [65]. The plane containing chiral centre carrying carboxylic moiety tilts ∼32° away from orthogonality with respect to triazole plane, favoring two strong hydrogen bonding (O8⋯H14A, 1.85 Å O9⋯H14B, 2.01 Å). Thus carboxylic group on each ligand and aqua ligands on terminal zinc act as chain binders establishing the supramolecular interaction (at the periphery) and form a 3D chain (Figure 1(c)). Thus a dense H-bonding network can be seen in this structure. View of crystal packing is shown in Figure 1(d). Recent reports highlight the importance of H-bonding of zinc bound water molecules on catalytic activity [66,67]. It was observed that functional mimics of metallo-β-lactamases catalytically hydrolyze β-lactam substrates, such as oxacillin and penicillin G. The high β-lactamase activity of such compounds was ascribed to the presence of zinc bound water molecules that is activated by being H-bonded to acetate substituent [66,67].

Solvothermal and classical synthesis with an identical recipe involving 4H-1,2,4-triazol-4-yl acetic acid (HL2) and CdCl₂ afforded rectangular blocks of 1D (2) and hexagonal crystals of 2D (3) CPs, respectively. [Cd(HL2L2)Cl]·2H₂O (2) crystallizes in the monoclinic space group P2₁/c. Each Cd atom is coordinated by two chloride atoms (Cd–Cl, 2.656(1), 2.616(6) Å), in axial position and four nitrogen atoms (Cd–N1, 2.343(3) Å, Cd–N2, 2.312(3) Å, Cd–N4, 2.374(3) Å and Cd–N5, 2.339(3) Å) belonging to four different ligands forming the equatorial plane, thus completing an octahedral geometry. The triple bridge that is made up of one μ-chloride and two μ-N1,N2-1,2,4-triazole of different L2/HL2 (Figure 2(a)) connect neighboring cadmium centers leading to a 1D coordination polymer with Cd···Cd distance of 3.845 Å.

There are a number of interesting similar Cd MOFs of 1,2,4-triazole derivatives [68]. Lu et al. reported a series of Cd coordination compounds ranging from 0 to 3D thanks to substitution of the 1,2,4-triazole or variation of the anion [69]. In a co-ligand modulated Cd-1,2,4-triazolate framework, tri and pentacadmium clusters were reported with a slightly longer Cd…Cd distance (4.049 Å) than for 2 [70]. Interesting cases of μ-OH₂ bridged 3,5-disubstituted 1,2,4-triazole systems and polymer chains containing μ₃-Cl were also reported [71]. It is worth mentioning too, the series of 3D Cd-triazolate complexes involving anions like F⁻, Cl⁻, Br⁻, I⁻, NO₃⁻, SO₄²⁻ which show diversity in coordination [27]. What makes 2 unique relates to the dual contribution of its ligand system similar to the one described in 1. Indeed, HL2/L2 not only acts as a bridging ligand, but also as anion due to the presence of a deprotonated carboxylate group in HL2. This shows predominance of bridging mode of 1,2,4-triazoles over carboxylic group under present experimental conditions.

The three planes containing ligands are mutually perpendicular around cadmium ions and are threaded by μ-Cl anions as intersecting points forming a 1D zig-zag chain. It was difficult to locate the hydrogen atoms in structure refinement on the carboxylic group of the ligand, but a comparison of bond length of –C–O and –C=O between ligands contributing to the two planes with that of free HL2 gave evidence of deprotonation. In the free ligand [53], the –C=O distance is 1.215(3) Å and the –C–O⁻bond length is 1.307(2) Å. It is interesting to note that the ligand exhibiting dual nature forms an exclusively separate plane (Figure 2(a), plane colored in pink) with shortened C(4)–O(1) = 1.255(5) Å and C(4)–O(2) = 1.251(4) Å distances, which indicates resonance double bond nature as a result of deprotonation. The ligand molecule in the other plane shows bond lengths of C8–O3 = 1.306(4) and C8–O4 = 1.206(4) Å similar to free ligand. Chloride forms two unequal bonds with cadmium and all the Cd-N bond lengths are different as indicated above. Thus the CdN₄Cl₂ octahedron is distorted with a distortion parameter of Σ = 20.44° [72]. The tail ends bearing carboxylic groups of the ligands of one plane forming this 1D network face in the same direction. Similarly this region is also shared by a plane from neighboring chain creating a super hydrophilic region (Figure 2(b)). The formation of triple μ-N1,N2-triazole bridging, as found for instance for [Cu(hyetrz)₃](CF₃SO₃)₂·H₂O [73], or double chloro-bridges plus a single μ-N1, N2-triazole bridging [68], was not favored in this network, due to the possibility for the ligand to act as monoanionic. A significant deviation of planes containing the carboxylic group with respect to the triazole planes in L2 of 2, was noted. In HL2, these planes are orthogonal but in 2 they deviate by nearly 21° (torsion angle (ø), C2–N3–C3–C4, 69.13°) and 25° (ø, C5–N6–C7–C8, 65.40°). This deviation is not only due to sterical reasons but also to secondary interactions in the super hydrophilic region. The hydrophilic region is rich in hydrogen bonds, the lattice water molecules being the epicenters of the H-bonding network. Carboxylic groups also greatly contribute to favor a dense H-bonding network. Two H-bonds with the carboxylic groups of different 1,2,4-triazoles from different planes (C4–O2⋯O6, 2.751 Å with an angle C4–O2–O6 of 105.08° and C8-O3⋯O6, 2.568 Å and an angle of C8–O3–O6 of 119.25°) are found. The deprotonated ligand forms additional H-bonds; one with 2.764 Å (O1⋯H(O5)) with lattice water which in turn forms an H-bond with the chloride bridge (Cl⋯H(O5), 3.228Å). As a consequence, the bond angle of 122.64° for –COO^- in the free ligand is stretched out to 127.02° in the complex. The CH₂ spacer is also involved in H-bonding with lattice water molecules with 2.628 Å (C3(H3A)⋯O1). The proton on the triazole ring C6(H6)⋯O5 forms a strong H-bond with 2.242 Å. Interestingly, the bridging mode of two ligands and one chloride around two Cd atoms create a bicyclo 7-membered metallomacrocycle ring [74,75].

Hexagonal colorless crystals of [Cd₂(L2)₂Cl₂] (3) crystallize in the non-centrosymmetric orthorhombic space group P2₁2₁2₁. The corresponding crystal is shown in Figure 3.

The self-assembling motif consists of two cadmium ions, five ligands and three chloride ions. Although both cadmium ions share the same central core of CdN₂O₂Cl₂, they differ in their bond distances and angles. Each Cd1 atom is coordinated by two bridging Cl atoms (Cd1–Cl1, 2.543(2) Å; Cd1–Cl2, 2.598(2) Å), two L2 molecules via the nitrogen atoms of the ring (Cd1–N1, 2.368(6) Å, Cd1–N2, 2.273(5) Å) and one L2 molecule via both O atoms of the carboxylate group (Cd1–O4, 2.274(4) Å Cd1–O3, 2.516(4) Å) in a bidentate fashion.

The Cd2 has a similar environment with slight variation in the Cd2–N, Cd2–O and Cd2–Cl bond distances. The single μ–N1,N2-triazole and μ–Cl bridging hold the neighboring cadmium atoms at a distance of Cd⋯Cd = 4.101 Å whereas across the full length of ligand the Cd⋯Cd distance is 8.907 Å. This pattern of dissimilar environment around each Cd1(a) and Cd2(b) propagates in abbaab order along the chain in c direction. The simultaneous use of 1,2,4-triazole and carboxylic groups in coordination with cadmium ions in a self-assembly process thus leads to a 2D network.

The dihedral angle between 1,2,4-triazole plane and –CH₂COO⁻ planes connected to Cd1 (C40–N5–C24–C1) is 80.72° whereas across Cd2 (C34–C38–N3–C7) it is 114.0°, which is significantly larger than the value found for the free ligand HL2 [53]. Navigation along the different planes of this structure reveals that this 2D network has a spiral connectivity leading to helicates [76-80]. The propagating chains running along the a axis (Figure 4) are made up of head to tail coordinated L2 molecules connecting Cd1 atoms to form a helical chain. The resulting intrinsically spontaneous helicity reveals a skeletal structure with a single stranded helix consisting of ligand and cadmium in each turn. Only two cadmium and two ligands span one helical turn so that the helix is considerably compressed. The pitch height of the helix is equal to the length of the crystallographic a axis (7.579 Å). Helical arrangement implicitly results in chirality producing right (Δ = P) or left-handed (Λ = M) helicity around principal axes [76-80]. The cylindrical helices are arranged in an alternating right-handed and left-handed fashion in the intrachain ‘a’ direction with a distance of 10.3 Å (Figure 4) which are domains of opposite helicity. Thus the complex is racemic because the right and left handed chains are alternatively arranged but in b direction, it shows single handedness. The helical twist of the strand results from the torsion between the 1,2,4-triazole plane and the spacer connecting the chelating carboxylic group. The orientation of the –CH₂ [81] (winged, (w) or pinched, (p)) is crucial to direct the spirality. Both the –CH₂ of each turn adopt a favorable winged orientation. This spiral ‘staircase’ leaves behind a roughly polygonal cavity of 4.89 Å (O3–O3) × 6.224 Å (C24–C24). Helicity has previously been noticed on a few 1,2,4-triazole CPs. In a related system constructed from of 3-amino-1,2,4-triazolate, where additionally cadmium ions are bridged by bromide ions, two kinds of helical channels with opposite chirality were observed [74]. With 3-amino-1,2,4-triazole, a Cd^II coordination framework with a 2D helical channel was also recently reported [75].

Reaction of HL2 with aq Cu(BF₄)₂ in water afforded [Cu₃(μ₃–O)(HL2)₆·(H₂O)₃]BF₄·H₂O, (4) in low yield. The yield can be dramatically improved when HL2 was replaced by L1 and slowly hydrolyzing the ester functionality by a base. The dark blue cubic crystals thus obtained crystallize in a cubic space group (I 4̄ 3m). Mixed coordination environments of N and O are seen around copper. Three copper atoms are lodged at the vertices of an equilateral triangle around μ₃–O keeping Cu⋯Cu distance of 3.368 Å (Figure 5(a)). This value compares well with a similar triangular tricopper MOF built from unsubstituted 1,2,4-triazole (Cu⋯Cu = 3.439 Å) [82]. An overview of the crystallographic parameters of 1–4 is given in Table 1 for further comparison. The copper centre is in ‘4 + 1’ (CuN₂O₃) square pyramidal geometry and each copper binds three ligands: two ligands coordinate through nitrogen of 1,2,4-triazole in a N₁, N₂ bridging mode (Cu–N2 = 1.950(9) Å) while one ligand coordinates through a carboxylate group in an unidentate fashion (Cu–O9 = 1.943(1) Å) and a water molecule loosely coordinated to copper (2.32(2) Å). The Addison structural index parameter τ, which is relevant for five-coordinate structures as an index of the degree of trigonality [83], was evaluated as 0.014 thus confirming a square pyramidal geometry. This triangular, tricopper(II) unit supported by –μ₃–O bridging and six ligands, forms the secondary building unit (SBU). The μ₃–O atom is in nearly planar triangular coordination sphere with an average Cu⋯μ₃–O distance of 1.946(2) Å. The Cu–O–Cu angle is of nearly 120° (119.8(1)°) which is indicative of a sp² configuration of oxygen atom [82]. This SBB has a three-fold axis of symmetry passing through μ₃–O.

The self assembly of this rigid SBU in 3D generates a cage like structure. The interlinking arrangement of eight such SBUs generates a molecular cage/porous ball which can be considered as tertiary building unit (TBU) or ‘supercluster’ [84,85]. The axially bound 24 water molecules on each of 24 copper atoms of TBU cage, decorate a hydrophilic interior pointing towards the void. The total solvent accessible volume is 4,477.5 ų which accounts for 48% of the cell volume. The distance between opposite oxo-centers is 18.219 Å. Considering van der Waals radii, the void may admit a sphere of 1 nm of diameter (Figure 5(b)). Each nanoball is the artifact of 96 molecules (24 Cu²⁺ ions, 24 water molecules, and 48 ligands) self-assembling in three dimensions (Figure 6). Each nanoball is tunneled into six neighboring nanoballs through a roughly square window/aperture. These mutually perpendicular 3D channels represent a kind of undulating passages.

Since each TBU is roughly spherical, their specific spatial arrangement in 3D obviously generates smaller voids/interstices as can be seen in the body centered cubic arrangement of spheres. The larger channels open into smaller hydrophobic ones which are roughly square shaped (6.328 × 6.328 Å) and where the non coordinated BF₄^- anions are stranded. Indeed, anions are distributed in both periphery of hydrophilic and centre of hydrophobic channels. Water molecules of crystallization are lying within the cavity anchored by H-bonding. The –C=O of carboxylic group with a length of 1.215(3) Å in free ligand is lowered to 1.187(2) Å in the complex due to H-bonding with the water molecules, whereas the C–O length in the ligand is lowered by approximately 0.06 Å to 1.294(1) Å, upon coordination to copper.

The semi-rigid nature of the ligand framework was realized after closely studying the bond length and torsion angle changes upon complexation. The crooked ligand with an obtuse angle of 110.6(2)° between N4–C6–C7 has flexed out to 112.0(1)°, while that of C6–C7–O8 with 115.5(2)° bent to 122.2(1)° in 4. The dihedral angle between two planes involving N4–C6–C7–O9 in 4 is exactly 180° which means that the two planes are stiffened to lie parallel as opposed to an angle of 176(1)° in the ligand. The flexibility is also evident from the change in dihedral angle between triazole plane and plane containing –CH₂–O group from 81°(1) in HL2 to 87° in 4. These parameters may assist in redesigning the framework for specific applications like gas storage where a spacer length and bent angle as well as pore volume could be tuned.

The coordinating abilities of L1–L5 (Scheme 1) with Zn^II ions in solution, solid state and vs. three zinc-β-lactamases have been exploited through inhibition assays. The production of β-lactamases is the most common mechanism of bacterial resistance to β-lactam antibiotics, i.e., penicillins, cephalosporins, and carbapenems [86]. These defense enzymes pose serious medical problems, mainly in hospital environments because they rapidly hydrolyze drugs thus reducing curing efficiency. Most of β-lactamases are serine-proteases (classes A, C and D) against which selective β-lactam inhibitors, such as tazobactam, have been developed and marketed for coadministration with antibiotics [87]. Metallo-β-lactamases (i.e., zinc proteases) constitute the class B [86]; presently, none of the inhibitors of zinc-β-lactamases have emerged for therapeutical use. Thus the search for non-β-lactam compounds, susceptible to form Zn^II complexes, could be a valuable strategy to discover inhibitors of class B β-lactamases [88]. L1–L5 were screened for inhibition against three representative metallo-β-lactamases, namely BcII from Bacillus cereus [89], CphA from Aeromonas hydrophila [90] and L1 from Stenotrophomoncas maltophilia [91] representing B1, B2 and B3 sub-classes. CephA exists mainly in the mono-zinc form, while L1 forms a dinuclear active site [92]; the question of one or two Zn^II ions in the class of BcII is still open [93]. The tested compounds were pre-incubated with the enzymes (30 min 30°C) before addition of imipenem antibiotic. The enzymatic residual activity was determined by monitoring the drug hydrolysis at 300 nm using a spectrophotometer. Although this assay clearly indicates that tested ligands are very weak inhibitors of class B β-lactamases, one compound (HL4) can be retained as a modest inhibitor of BcII enzyme. As a matter of fact, the HL4 ligand is able to form a polynuclear Zn^II complex, as shown by the X-ray crystal structure (Section 3.1). The BcII inhibition may result from the complexation of one or two zinc ions of the active site by 4-(methylthio)-2-(4H-1,2,4-triazole-4-yl)butanoic acid molecules which could stimulate further developments in medicinal chemistry [54].

In material sciences, metal threaded “soft” nonaromatic networks like 3 give tangible perception for facile generation of oxide material upon controlled breakdown of their network. Such methods where surface energy and growth kinetics can be fine tuned offer tailoring architecture of nano- and microcrystals oxide. Various synthetic methods were explored to prepare cadmium oxides owing to their shape- and size-dependent technological applications such as photovoltaic cells, gas sensors, transparent conductive oxides, phototransistors, photodiodes, transparent electrodes, in the form of rods and wires [94], whiskers [95], disk [96], cubes, belts [97,98], most of which are based on physical and chemical technique such as sol-gel, spray pyrolysis, sputtering method, Langmuir-Blodgett deposition, solvothermal synthesis, and controlled decomposition of MOF [98,99]. Thermogravimetric analyses (TGA) of 1 (Figure 7(a)) provides an overview of thermal stability and decomposition patterns. After the initial weight loss around 45 °C, which is due to surface moisture, the compound was stable until 170 °C and then decomposed into three ill-defined broad steps from 200 to 900 °C. At 850 °C a complete destruction of the ligand framework is evidenced. In a typical pyrolysis procedure, 3 was heated at a rate of 10 °C/min until 850 °C and allowed to rapidly cool to r.t, which led to isolation of a white residue. Scanning electron microscopy (SEM) images (Figure 7(b)) on this residue showed the presence of nearly cubic rods with thickness of 800 nm. The poor quality of the powder X-ray diffraction (PXRD) pattern did not allow us to assign it to any known phase. As mild sputtering was frequently observed with this procedure, the isolation of sufficient quantity of residue was difficult. This led us to use a mild thermal treatment of 1 °C/min until 850 °C instead, followed by rapid cooling to r.t. The obtained reddish-brown residue was considered for further investigation for phase identification, purity, and morphological characterization.

The PXRD pattern shown in Figure 7(f) was indexed to the powder pattern of the pure cubic phase of CdO (β-form) (JCPDS file No. 05-0640) [100,101]. The diffractogram shows prominent peaks at 20(°) = 32.99, 38.29, 55.27, 65.94, 69.26, which are associated with (111), (200), (220), (311) and (222) planes, respectively. It clearly shows that CdO crystallizes in rock-salt cubic structure. The pattern does not reveal any other peaks thereby confirming excellent phase purity. SEM images of this oxide (Figure 7(c)) shows truncated octahedron shaped crystals of 1–2 μm size. The shaping of these oxides can be further tuned and the same results of phase purity (based on XRPD pattern) were observed for the reddish brown residue obtained when pyrolysis was extended up to 850 or 900 °C, and subsequently allowed to be slowly cooled to r.t. In the former case, hexagonal blocks were obtained (Figure 7(d)), whereas a treatment at 900 °C afforded well-defined octahedral crystals (Figure 7(e)), both of them in the range of 1–2 um. This demonstrates that such ‘soft’ non-aromatic CPs could be a suitable host to carve desired metal oxide with control over morphology and phase selectivity [55].

The nanoporous void found in 4 prompted us to investigate gas adsorption capacity and to study the porosity partitioning by mercury porosimetry. A gas sorption study for 4 was carried out using N_2(g) and H_2(g). Before measurements, the crystalline sample was vacuumed either with a gradual increase of temperature or rapid heating until 150 °C, the temperature corresponding to the end of dehydration according to TGA analysis. Within the relative pressure range investigated, the N_2(g) adsorption measurements did not show any gas uptake, thus making impossible the determination of the specific surface area by routine BET (Brunauer, Emmett, Teller) calculation. An adsorption was however observed for H_2(g). This adsorption difference between these two molecules could be related to their size, their kinetic diameter being 2.8 Å (H_2(g)) and 3.6 Å (N_2(g)) [102]. The H_2(g) adsorption of 1.8 cm³/g and 2.7 cm³/g for the fast and slow out-gassing, respectively, is irreversible, which was confirmed by several consecutive runs. These low values could account for little surface adsorption as observed for a related SBU system [29]. These poor adsorption properties could also result from the breakdown of the network after thermal pre-treatment. This prompted us to revisit the gas adsorption measurement with mild pre-treatment at ambient condition. The fresh crystals were evacuated overnight under vacuum without any warming treatment.

Figure 8(a) shows improved adsorption behavior of 4. H_2(g) adsorption gradually raises to 21 cm³/g with increase of pressure. It not only bettered the earlier value but also shows narrow hysteretic reversibility. But again no preference for N_2(g), which strengthens the hypothesis of size discrepancy discussed above. Consequent runs do not follow the same trend. It reduces to ~6–7 cm³/g in second and third runs. Thus adsorption efficiency is not only reduced but also becomes irreversible. It is possible that, after the initial H_2(g) sorption where the increasing pressure flexes-out the network to open the ‘gate’ to accommodate the guest, it keeps the network under stress and the subsequent desorption probably partially collapses the network while shrinking back. The retention of almost half of the adsorbed gas (3 cm³/g) in second or third run after depressurization at low temperature although less, is interesting and such retentions are significant in systems which show high sorption-desorption property [103]. Square shaped openings (Figure 8(b)) discovered on the cubic crystal surface of 4 stimulated us to investigate the ‘porosity partitioning’ by Hg porosimetry [104,105]. Both techniques of gas adsorption and MIP rely upon balances between surface tension forces, capillary forces and pressure but their outcome may not be comparable due to distinctly different physical interactions involved in the two methods [106]. Figure 8(b) shows the volume of intruded Hg as a function of applied pressure. We recall that pressure is inversely proportional to the pore size according to Washburn equation (high pressures being necessary to fill small pores by inverted capillary effect) [107]. Mercury fills largest pores first, entering the smallest pores last under high pressure unlike gas adsorption. A fine analysis of Figure 8(b) allows identification of several steps. Indeed, as the pressure is gradually increased and as soon as it corresponds to a pore size, Hg enters the pores and a new step is recorded. When considering the pore size distribution (by applying the Washburn equation assuming constant contact angle and surface tension), the recorded steps can be understood as follows: the step between 1 and 5 psi corresponds to pores of ~50 µm, between 10 and 25 psi to pores of about 6,000 nm, between 30 and 80 psi to pores of about 2,000 nm and the remaining uptake between 100 and 500 psi reveals pores with sizes smaller than 350 nm. After that, the curve reaches a plateau indicating that there are no smaller pores present. The major contributions at 50 µm, 6,000 nm and 2,000 nm would rather originate from interparticular porosity, whereas the pores of 350 nm could correspond to the voids identified by SEM. These results thus illustrate the usefulness of Hg porosimetry for a complete porosity characterization in porous MOFs [53].

We have selected chiral and achiral derivatized amino acids in the construction of CP/MOFs. In the preceding sections we have seen how chiral HL4 resulted in a trinuclear network in centrosymmetic space group and achiral HL2 resulted in a 1D CP and a chiral helicate. Although the use of chiral molecules is the most common strategy [108,109] for obtaining products with a non-centrosymmetric arrangement in the crystal lattice, it is not a rule. It is found that only 10–15% of achiral organic molecules crystallize in noncentrosymmetric space groups [108,109]. In contrast to the chiral amino acids, achiral glycine (as a structural unit) cannot enforce non-centrosymmetry [108] of a crystal structure. But its electronic and structural asymmetry together with consolidated effort of noncovalent interactions with suitable metal ions as connectors may lead to crystallization in a non-centrosymmetric space group. By using bridging ligands with a significant kink in the topology instead of linear rigid molecules, the formation of networks with a lower symmetry and helicity is possible [80,108]. The spontaneous helicity observed in 3 is due to the hierarchical self-assembly of the tiny synthon HL1 originating in non-centrosymmetric space group which falls into this category. The simple torsion angle count around the spacer in such molecules forms the basis of the assessment of molecular flexibility in them. Such conformationally flexible backbone and electronic asymmetry (push-pull effect) have high chances of ending up in a noncentrosymmetric space group, and such compounds have an enormous significance because of technological properties such as single harmonic generator (SHG), piezo-pyro-ferro electricity, triboluminescence [110]. Despite sharing common synthetic recipe 2 and 3 adopt totally different network formation. In 2, the carboxylic group is kept away from the coordination sphere unlike in 3 thus demonstrating role of synthetic method in directing network construction through self-assembly. This is a crucial point for the design of 1D coordination polymer involving switchable Fe^II spin crossover (SCO) centers with a MN₆ core [15-18]. A similar situation is also noticed in 1. Complex 1 is the first crystal structure of a triazole derivatized methionine complex, and also represents the first example of a neutral zinc trinuclear complex with three N1,N2-1,2,4-triazole bridges and without non-coordinated anion. It provides promising perspectives for computational studies in SCO research as it provides a model of a trinuclear diamagnetic complex without counter-anion while keeping the same size as high-spin Fe^II ions [111].

On the other hand HL2 undergoes stunning self-assembly to form a 3D array of nanoball architecture in 4. The nanoporous void obtained in the present system is the artifact of angular framework of the ligand, thus hinting that fine tune of the void shape and size should be possible with variation of methylene ankle in the ligand and even impose network interpenetration. Further work is in progress to increase the void volume by replacing α-amino acid by β-amino acid and γ-amino acid. The gas adsorption profile in 4 is not promising as the ‘soft’ skeleton of the ligand is not robust enough to withstand high temperature treatment of degassing. But in such systems in which axially bound water molecules can be easily removed, not necessarily thermally, that could possibly allow access to the Lewis acid sites for catalytic applications as nano-reactors for catalysis or anion/cation recognition. Another appealing perspective would concern the magnetic study of 4 in the frame of its potential spin frustration [112-115].

The family of thermochromic 1D iron(II) 1,2,4-triazole spin crossover (SCO) coordination polymers have attracted our attention in recent years due to their promising prospects in the thriving field of molecular electronics [15-18]. Such triazole based precursors not only provide appropriate ligand field strength but also favors the formation of CPs due to its dominant N1,N2 bridging mode ending up in high dimensional networks which are recognized for their ability to efficiently propagate elastic cooperative effects associated to the SCO event [116]. Although SCO is well understood in 3d⁴–3d⁷ transition metal ions in octahedral surrounding, the 3d⁶ iron(II) complexes received greater attention [117]. In these compounds, the entropy driven reversible intra-ionic electron transfer from a diamagnetic low-spin (LS, ¹A_1g) state to a thermally populated paramagnetic high-spin (HS, ⁵T_2g) state can be induced by external stimuli like temperature, pressure and light irradiation. In fact in the solid state, the presence of short and long range interactions acts as communication media between iron centers promoting cooperative first order spin transitions that can leads to a large memory domain that can be suitable for potential applications [118,119]. Indeed, a SCO compound meeting display and data processing requirements would have in addition a good shelf life and an easily detectable optical response, and would operate near room temperature [120]. In this context we recently introduced ethyl-4H-1,2,4-triazol-4-yl-acetate (L1) (Scheme 1) as a novel neutral bidentate ligand to synthesize a family of 1D SCO polymers [17]. Reaction of L1 with [Fe(H₂O)₆] (ClO₄)₂ in methanol afforded a white powder, which was dried under vacuum to lead to a desolvated compound [Fe(L1)₃] (ClO₄)₂ (5) whereas the same powder dried under stream of N_2(g) afforded [Fe(L1)₃] (ClO₄)₂·MeOH (6). The 1D polymeric nature of these compounds was confirmed by spectroscopic and analytical results. For 5, the temperature dependence of the high-spin molar fraction derived from ⁵⁷Fe Mossbauer spectroscopy reveals an exceptionally abrupt single step transition between LS and HS states with a hysteresis loop of width 5 K (T_c^↑ = 298 K and T_c^↓ = 293 K). This spin transition operating around room temperature presents striking reversible thermochromism from white at 295 K to pink at ice temperature (Figure 9), thus behaving as an optical alert towards temperature variations [118]. Interestingly, the transition temperature of 6 is shifted to below room temperature (T_c^↑ = 273 K and T_c^↓ = 263 K), thus showing a remarkable influence of solvent inclusion on the spin state of these chain compounds. Efforts are underway to introduce β and γ amino acids and esters to investigate the influence of spacer on the SCO properties [56,57].