Syntheses, Structures, and Corrosion Inhibition of Various Alkali Metal Carboxylate Complexes

Complexes of the alkali metals Li-Cs with 3-thiophenecarboxylate (3tpc), 2-methyl-3-furoate (2m3fur), 3-furoate (3fur), 4-hydroxycinnamate (4hocin), and 4-hydroxybenzoate (4hob) ions were prepared via neutralisation reactions, and the structures of [Li2(3tpc)2]n (1Li); [K2(3tpc)2]n (1K); [Rb(3tpc)(H2O)]n (1Rb); [Cs{H(3tpc)2}]n (1Cs); [Li2(2m3fur)2(H2O)3] (2Li); [K2(2m3fur)2(H2O)]n (2K); [Li(3fur)]n(3Li); [K(4hocin](H2O)3]n (4K); [Rb{H(4hocin)2}]n.nH2O (4Rb); [Cs(4hocin)(H2O)]n (4Cs); [Li(4hob)]n (5Li); [K(4hob)(H2O)3]n (5K); [Rb(4hob)(H2O)]n (5Rb); and [Cs(4hob)(H2O)]n (5Cs) were determined via X-ray crystallography. Bulk products were also characterised via XPD, IR, and TGA measurements. No sodium derivatives could be obtained as crystallographically suitable single crystals. All were obtained as coordination polymers with a wide variety of carboxylate-binding modes, except for dinuclear 2Li. Under conditions that normally gave coordinated carboxylate ions, the ligation of hydrogen dicarboxylate ions was observed in 1Cs and 4Rb, with short H-bonds and short O…O distances associated with the acidic hydrogen. The alkali-metal carboxylates showed corrosion inhibitor properties inferior to those of the corresponding rare-earth carboxylates.

Rare-earth carboxylates are normally prepared via metathesis reactions between sodium carboxylates, often prepared in situ, and rare-earth salts. Because it has been claimed that the use of alkali-metal salts of carboxylic acids prevent corrosion in water treatment facilities [10], the sodium salts are usually tested for comparison with the rare-earth carboxylates, and inevitably perform much worse [3,4,10,13,18]. Moreover, preliminary immersion studies performed by Wormwell and Mercer [19] demonstrated that lithium and potassium benzoates also exhibit inhibitory properties against mild-steel corrosion. sodium, none of the five ligands formed crystalline products suitable for X-ray crystallography. The bulk samples of sodium complexes were characterised by powder XRD, IR, and TGA in the ESI † , and the results suggest that they have dissimilar structural profiles compared to the other alkali-metal carboxylate complexes in the same group. Therefore, further characterisations were not carried out. In the case of 1Cs and 4Rb, acid salts [M{H(L) 2 }] (M = Rb, Cs) were obtained, despite intentionally maintaining a 1:1 M:LH reaction stoichiometry (M = alkali metal; L = ligand) to prevent the incorporation of the carboxylic acid into the product (see Introduction for previous similar issues). Among the structures, 1Cs, 2K, and 5Li exhibited slight hygroscopicity. This was evident from the thermal analysis of 1Cs and 2K, as well as the microanalysis of 5Li, which showed values consistent with water absorption. The compositions of the [M(3tpc)] n (M = Li, K, Rb, and Cs) complexes in the 1M series were determined through X-ray crystallography, elemental analysis, and thermogravimetric analysis (TGA). The elemental analysis agreed with the single-crystal compositions for all four compounds. The TGA results of 1Li, 1K, and 1Rb ( Figure S1) were consistent with the compositions from the single-crystal data. However, the TGA of 1Cs showed a loss of 0.5 water molecules in a single step below 150 • C. Nonetheless, the powder XRD analyses of all four compounds showed close similarities to the patterns generated from the single-crystal data ( Figure S2).
In the [M(2m3fur)] n (M = Li, K) 2M series, the TGA and elemental analysis of compound 2Li aligns with the composition determined crystallographically as [Li 2 (2m3fur) 2 (H 2 O) 3 ]. Although the elemental analysis of compound 2K matches the composition of [K 2 (2m3fur) 2 (H 2 O)] n of the crystal structure, the TGA of the bulk precipitate indicated the presence of water of crystallisation ( Figure S4). The weight loss below 200 • C indicated the evolution of nearly two water molecules. This suggests that the compound may be hygroscopic after undergoing drying to a constant weight over silica gel. Nonetheless, the powder XRD analyses revealed similar d-spacings to those generated from the single-crystal data for both 2Li and 2K ( Figure S5).
The elemental analysis of lithium-3-furoate (3Li) aligns with the calculated percentage composition for [Li(3fur)] n (3Li), and the thermogravimetric analysis ( Figure S7) and IR analysis ( Figure S9) of the sample confirm the absence of any water. Additionally, the powder diffractogram obtained from the bulk precipitate matches the calculated diffractogram based on the single-crystal data ( Figure S8).
All three compounds in the M(4hocin) 4M series (M = K, Rb, and Cs) have different compositions. The TGA of compound 4K yielded results consistent with its crystal composition, revealing the loss of three water molecules in a single step below 100 • C ( Figure S10), but the elemental analysis of it indicated a loss of 1.5 ligated water molecules due to the drying of the samples over silica gel. Nonetheless, the powder diffractogram of the bulk sample matches the one generated from the crystal data ( Figure S11).
In the case of compound 4Rb, the structure was determined as [Rb{H(4hocin) 2 }] n .nH 2 O, but the analysis of the bulk product through microanalysis and TGA suggests that 0.75 4hocinH had been lost from the crystal composition. Unlike in 4K, the experimental powder diffractogram of the bulk sample 4Rb exhibited a noticeable difference when compared to the diffractogram generated from the crystal data attributed to the loss of 4hocinH. Upon drying the crystals of compound 4Rb over silica gel, they seemed to lose their crystalline form and appeared more amorphous.
By comparison, the caesium analogue prepared under similar conditions was determined via crystallography to be [Cs(4hocin)(H 2 O)] n , which was consistent with the elemental analysis and TGA results. The powder XRD analysis of the bulk sample of 4Cs closely resembles the powder diffractogram generated from the crystal data, but a few minor discrepancies can be observed ( Figure S11).
The TGA thermal data of 5Li ( Figure S13) indicated that it is an anhydrous compound, which aligns with the X-ray composition of [Li(4hob)] n . However, the elemental analysis of the complex indicated the presence of some water, suggesting it is slightly hygroscopic. The powder diffractogram closely resembles the generated powder diffractogram obtained from the crystal data ( Figure S14), indicating that the structural integrity is maintained.
The TGA of 5K revealed the loss of three water molecules in two steps, occurring below 75 • C and below 150 • C. However, the elemental analyses of the bulk sample yielded a molecular composition of [K(4hob)(H 2 O)] n , as opposed to the composition derived from the crystal data (namely, [K(4hob)(H 2 O) 3 ] n ), indicating efflorescence. Additionally, the experimentally obtained powder diffractogram differs from the powder diffractogram generated based on its crystal data, consistent with the water loss ( Figure S14).
The two complexes 5Rb and 5Cs have the general formulae [M(4hob)(H 2 O)] n (M = Rb or Cs) from elemental analyses (C, H). The TGA results for 5Rb are in support, but the bulk [Cs(4hob)(H 2 O)] n (5Cs) lost a small fraction of coordinated water during drying prior to the TGA. The powder diffractograms of both compounds exhibited some similarities and displayed peaks that matched the generated diffractograms based on their crystal data.
The IR spectra of all the complexes displayed peaks corresponding to carboxylate COOstretching bands in the range of 1560-1360 cm −1 ( Figure S15). These peaks included both asymmetric stretching bands (from 1509 to 1556 cm −1 ) and symmetric stretching bands (1361-1399 cm −1 ). Additionally, the presence of a broad band in the range of 3200-3550 cm −1 indicates the presence of water molecules and hydrogen-bonded alcohol O-H stretching (3400-3200 cm −1 ) in the compounds. Having absorption bands in the region of 1617-1682 cm −1 in 1Cs confirms the presence of an [H(3tpc) 2 ] ligand with the C(O)OHO(O)C group ( Figure S3). The IR spectra for all three compounds in the 4M series have bands of moderate intensity in the range of 1633-1636 cm −1 , which corresponds to the stretching of the propenyl C=C double bond. The consistent υ(C=C) frequencies indicate that the propenyl group is not engaged in coordination [28]. The carbonyl absorption band of [H(4hocin) 2 ] in the range of 1617-1682 cm −1 is not clear in the spectrum of compound 4Rb ( Figure S12). Table 1 presents a summary of the significant absorption bands observed in the IR spectra of all 14 compounds.

X-ray Crystal Structures
The carboxylate-binding modes observed in the complexes are shown in Figure 1, and a summary of the features of the structures is given in Table 2. The increase in the average metal-oxygen bond distances, as well as the increase in the coordination number, can be observed as the ionic radius increases from Li to Cs [29] within each series. However, 4K stands out as an outlier in this trend ( Table 2). The ESI † provides detailed information regarding the bond distances and angles.    The Li2#1-Li1-Li2-Li1#4 chain of the complex crosslinks with neighbouring chains through bridging 3tpc ligands, forming a 2D network. Li2 is connected to the Li1#3 metal atom of the neighbouring chain through oxygen atoms (O2 and O3#3), while Li1 is bound to Li2#2 through bridging oxygen atoms (O3 and O2#2). In both cases, there is a separation of 2.712(9) Å, forming four-membered rings that are perpendicular to each other ( Figure 3). The longest metal-oxygen bond in the structure is between Li2 and O3#3, while the shortest bond is between Li1 and O3.
The longest Rb-O bond is from a water oxygen, Rb1-O3#8 (3.278(3) Å), and the shortest bond is for a bridging carboxylate Rb1-O1 (2.828(2) Å). The coordinated water oxygens O3 and O3#4 appear to connect the wave-like 1D polymers into a two-dimensional (2D) chain ( Figure 7). All four water oxygen atoms found in the Rb coordination environment are involved in hydrogen bonding to carboxylate-O ( Figure 6, Table S12), forming an extensive network of hydrogen bonding and further linking the 2D network together.
The polymeric complex [Cs{H(3tpc) 2 }] n (1Cs) crystallises in the space group C2/c. Although the reaction stoichiometry was intentionally maintained at 1:1 Cs:3tpcH to avoid the incorporation of excess unreacted proligands in the structure, only partial deprotonation of the carboxylic acid group occurred. As a result, a hydrogen atom is symmetrically shared between two carboxylate groups, making a (HL 2 ) − ligand. The Cs atom is eight-coordinate with cubic geometry. In the asymmetric unit, the ligand bridges Cs1 and Cs1#2 via O1, and the only binding mode observed for the structure is The observed partial deprotonation results in both the Cs + cation and O-H hydrogen atom being located on crystallographic inversion centres. In the asymmetric unit, both entities carry an effective charge of 0.5+, effectively counterbalancing the negative charge of the 3tpcanion. The oxygen atoms of the carboxylate groups from the two 3tpc ions are involved in forming symmetrical bonds with the hydrogen atom located on the inversion centre ( Figure 8). This results in O-H bond lengths of 1.232(4) Å, which are longer than those of a typical O-H covalent bond (0.97 Å) [34], but shorter compared to the lengths observed in O..H hydrogen-bond interactions. The symmetrical hydrogen bond that is observed is commonly accompanied by a significantly short O...O distance ( 2.460 Å here). The coordinated O from the 3tpc − anion bridges the 1D polymers, forming a two-dimensional (2D) chain ( Figure 9).  highlighting the presence of a centrosymmetric hydrogen atom located between the two carboxylate groups. Symmetry code: #1 + X,1 + Y, + Z; #2 + X,-1 + Y, + Z; #3 1-X,2-Y,1-Z; #4 1-X,1-Y,1-Z; #5 1-X, + Y,1/2-Z; #7 1-X,1 + Y,1/2-Z; #8 + X,2-Y,1/2 + Z. Despite multiple attempts to refine the hydrogen atom as disordered, appearing on either side of the inversion centre, it consistently reverted back to the inversion centre. Determining the precise location of a hydrogen atom solely based on X-ray diffraction data presents a challenge. Nonetheless, by examining the distance between oxygen atoms and analysing the hydrogen atom behaviour during refinement, we have confidence in its location. Therefore, we can conclude that only the partial deprotonation of 3tpcH occurs in 1Cs.
The two-dimensional polymer [K 2 (2m3fur) 2 (H 2 O)] n (2K) crystallises in the monoclinic P2 1 /c space group. Despite the crystal structure being a monohydrate, thermal analysis of the bulk precipitate indicated the presence of an extra water molecule. The asymmetric unit of the compound contains two distinct potassium metal centres. K1 is eight-coordinated; but it can be considered as pseudo-seven-coordinated if the carboxylate takes one coordinated site, resulting in a distorted pseudo-pentagonal bipyramidal stereochemistry. Thus, O4 and a chelating carboxylate are in axial positions, with an O4-K1-C7#5 angle of 168.53 (4) • , while O1, O1#5, O2#4, O7#2, and O7#6 are in equatorial positions. The six-coordinated K2 can be regarded as distorted pseudo-square pyramidal if the carboxylate is considered to occupy one coordination position. In this case, four basal oxygen atoms (O2#3, 4#1, 5, 7) form a square plane, while the carboxylate is in the axial position.
The carboxylate ligand binds in two unique modes:  The adjacent polymeric chains are oriented in opposite directions, and they further extend to form a 2D network through the bridging of the water oxygen atoms, as well as the carboxylate oxygen, from within the chain to the neighbouring chain ( Figure 12).   The 2D polymeric complex [K(4hocin)(H 2 O) 3 ] n (4K) crystallises in the space group P2 1 /c with a distorted tricapped trigonal prismatic geometry. The nine-coordinate potassium atoms are each attached to two µ-1κ(O);2κ(O) carboxylates and seven oxygen atoms from aqua ligands. The water oxygen atom O4 is bound to three potassium atoms (K1, K1#2, K1#3), while O5 and O6 are each bound to two potassium atoms: K1 and K1#3 and K1 and K1#1, respectively. The chain propagation exhibits a twist, with an angle of 126.48(2) • between K1#1, K1 and K1#2, and the distance between K1 and K1#1 is 3.9478 (7)Ǻ. Four ligated aqua molecules link neighbouring polymeric chains that run in opposite directions, forming a 2D network through O4, O4#3, O5, and O5#3 ( Figure 15). µ 3 -H 2 O coordination is known. For recent examples, see [35][36][37].
The longest metal-oxygen bond occurs between K1 and O5 (3.3762(14) Å), while the shortest bond is between K1 and O6 (2.7326(15) Å), both originating from ligated water molecules. The ligated water molecules form an extensive network of hydrogen bonding ( Figure 16 and Table S12), along with the carboxylate oxygen atoms of the ligands, which further interconnect the adjacent 1D chains into a 2D intermolecular framework.
The three-dimensional polymeric complex [Rb{H(4hocin) 2   The asymmetric unit of 4Rb consists of a single type of Rb atom and is expanded to depict the complete Rb coordination sphere in Figure 17. The arrangement of the sixcoordinated Rb metal centre is distorted from any regular polyhedron. Contrary to 1Cs, in the asymmetric unit of 4Rb, the ligand is connected to Rb1 by the carboxylate oxygen atom (O1), which also binds to an H atom on the inversion centre. Within the structure, the water molecule functions as a focus for crystallisation through hydrogen bonding, and it does not form coordination bonds with the rubidium centre. The coordination sphere of The adjacent chains in the 2D network are further connected by the bridging carboxylate oxygen atoms. The hydrogen bonding involving the phenolic-hydroxy groups and the carboxylate oxygen atoms of the ligand with the lattice water of crystallisation holds the neighbouring chains of polymers together, forming an extended three-dimensional packing arrangement of the polymeric chains ( Figure 18 and Table S12).   packing arrangement. #1 + X,3/2-Y,-1/2 + Z; #2 1-X,1-Y,1-Z; #3 1-X,1/2 + Y,3/2-Z; #4 1-X,1-Y,2-Z; #5 1-X,-1/2 + Y,3/2-Z; #6 + X,1/2-Y,1/2 + Z. #7 1-X,1 + Y,1/2-Z; #8 + X,2-Y,1/2 + Z. The X-ray crystal structure of [Cs(4hocin)(H 2 O)] n (4Cs) reveals a monoclinic crystal system with a P2 1 space group. The eight-coordinate caesium atoms have two bridging oxygen atoms (O1, O2#2) and two chelating oxygen atoms (O1#1,2#1) from three µ 3 -1κ(O);2κ(O,O );3κ(O ) ligands, two water molecules (O4, O4#3), and two phenolic oxygen atoms of the ligand (O3#4, O3#5). Four oxygen atoms are close to the plane with the Cs centre, while the other four oxygen atoms sit on one side of the plane. This arrangement creates an irregular eight-coordinate polyhedron ( Figure 19).
The Cs-O bond lengths for the bridging carboxylates are consistently shorter than their chelating counterparts (Table S8). The angle between the bridging water oxygen atoms O4 and O4#3, which originates from the adjacent caesium atom, has an O4-Cs1-O4#3 angle of 138.84(10) • . The carboxylate-O1,2, as well as the phenolic oxygen atom O3#5, link the metal centres into a 1D chain with a Cs1 . . . Cs1#6 distance of 4.3870(9) Å and a Cs1#1...Cs1 . . . Cs1#6 angle of 180 • . A phenolic oxygen atom, O3, bridges an adjacent metal centre in neighbouring chains, forming a 2D-layer network in the ac-plane (Figure 20a). A projection along the a-axis (Figure 20b) illustrates the interconnection of these 2D networks via a bridging polydentate ligand (via O1, 2), a coordinated aqua molecule (O4), as well as hydrogen-bonding interactions (Table S12) into an overall 3D structure. The complex [Li(4hob)] n (5Li) crystallises in the monoclinic P2 1 /c space group and exhibits similar ligand-binding modes and coordination environments to the 3Li structure ( Table 2). Analogous to the 3Li structure, the structure of 5Li also forms a 1D polymeric chain through the two alternate six-membered rings involving the bridging oxygen atoms O1, O2, O2#4, O1#4, and O2#5 ( Figure 21).     The average K-O bond length observed in complex 5K is 2.9123 Å, with the shortest bond being K1-O4 (2.734(13) Å), and the longest bond being K1-O5#4 (3.172(5) Å). Both the longest and shortest bonds involve aqua ligands rather than any of the bridging carboxylate oxygens. The adjacent polymeric chains connect to form a 2D network through the bridging of the water oxygen atoms (O5, O5#3). There is a significant presence of hydrogen bonding in the system, involving the carboxylate oxygen atoms and the OH groups of all sixcoordinated water molecules. Figure 24 illustrates the hydrogen-bonding pattern observed, while Table S12 provides the corresponding hydrogen-bond lengths. These hydrogen bonds play a crucial role in holding the neighbouring chains of polymers tightly together, further connecting the extended two-dimensional packing. It is noteworthy that none of the phenolic-hydroxy groups participate in these hydrogen-bonding interactions. [Rb(4hob)(H 2 O)] n (5Rb) crystallises in the monoclinic P2 1 /c space group with a square antiprismatic donor atom arrangement around Rb. Figure 25 illustrates the structure of the extended asymmetric unit of 5Rb.
Within the polymeric structure, there are eight-coordinate rubidium atoms bonded by four bridging carboxylate oxygen atoms (O1, O1#1, O2#4, O2#5). Additionally, two phenolic oxygen atoms (O(3#2, 3#3)) from the µ 4 -1κ(O);2κ(O);3κ(O );4κ(O ) 4hob ligand and two coordinated water oxygen atoms (O4 and O4#6) contribute to the coordination. The metal centre achieves a total coordination of eight by incorporating another ligated water oxygen atom (O4#6) bridging from the adjacent metal centre. The binding of the phenolic oxygen atom to the metal centre is also observed in the 4Cs structure. Each 4hydroxybenzoate ligand in the chain forms bonds with four Rb atoms, following the same coordination pattern. For 5Rb, the average Rb-O bond distance is 3.0486 Å. Among the Rb-O bonds, the bridging Rb1-O1 bond is the shortest (2.844(14) Å). In contrast, one of the ligated water molecules exhibits the longest Rb-O bond distance of 3.570 (16) Å (Rb1-O4#6). The two ligated water oxygen atoms form an O4-Rb1-O4#6 angle of 82.89(4) • .  (Figure 26a). When projected along the a-axis (Figure 26b), it becomes apparent how these 2D networks are linked together by alternative four-membered rings (through coordinated aqua molecules (O4 and O4#4)) and eight-membered rings (through two bridging ligands), resulting in an overall 3D structure.  [Cs(4hob)(H 2 O)] n (5Cs) crystallises in the monoclinic P2 1 /c space group and exhibits square antiprismatic eight-coordination. Although the crystal structures of 5Rb and 5Cs have different unit cells due to the inequivalent b-angles and a slight variation in the c-axis, the ligand-binding modes within the asymmetric units remain the same. Furthermore, the eight-coordinated metal atoms in both cases have similar coordination environments ( Table 2). The 1D polymeric chain of Cs atoms is nonlinear, with Cs1#1-Cs1-Cs1#7 at an angle of 168.391(8) • . The separation of the metal atoms in the 1D chains is longer in 5Cs (4.3322(9) Å) compared to the 4.0584 (8) Å in 5Rb. In contrast to 5Rb, the average Cs-O bond distance in 5Cs is slightly longer (3.2165 Å). The disparity in the metal-metal distance and the average metal-oxygen bond distances between 5Rb and 5Cs aligns with the increase in the ionic radii of eight-coordinate monovalent caesium compared to rubidium [29]. Similar to 5Rb, the bridging Cs1-O1 bond in 5Cs is the shortest (3.012(3) Å), while the coordinated water oxygen O4#6 exhibits the longest Cs-O bond distance of 3.761(3) Å (Cs1-O4#6). The two ligated water oxygen atoms form an O4-Cs1-O4#6 angle of 81.89(4) • .
Additionally, the interactions that join the 1D networks and contribute to the overall 3D structure are similar in both 5Rb and 5Cs. However, the 5Cs structure exhibits a network of H-bonding (Table S12) that holds the 3D network together (Figure 27).

Corrosion Inhibition Immersion Tests
The immersion tests were conducted as an initial screening, lasting for seven days in a 0.01M NaCl control solution. After the completion of the test, digital photographs were taken of the sample coupons ( Figure S16). Because the primary objective of the current study was to observe and compare the effectiveness of alkali-metal complexes of various carboxylate derivatives with rare-earth-metal carboxylate corrosion inhibitors, detailed immersion analysis was not pursued for any of the compounds.  Table 3 provides a summary of the weight-loss measurements obtained after 168 h of immersing the samples in specific inhibitor solutions. Upon examination of the results, it was found that the mild-steel coupons immersed in 800 ppm [K(4Hocin)(H 2 O) 3 ] n (4K) exhibited the highest corrosion inhibition rates, reaching 80%. [K 2 (2m3fur) 2 (H 2 O)] n (2K) displayed the lowest inhibition efficiency, with only 9% corrosion inhibition. The remaining tested compounds demonstrated inhibition efficiencies ranging from 25% to 31%. Table 3. Observed weight loss, corrosion rates (µgm −2 s −1 ), and percentage inhibition (η) for mild steel. Coupons immersed in specific solutions at 800 ppm concentrations in 0.01 M NaCl for seven days.

Concentration
Corrosion Rate (µgm −2 s −1 ) % Inhibition (η) ppm mM Previous studies involving 4-hydroxycinnamate as the ligand have indicated that the corrosion rates calculated from the weight-loss data were generally higher for sodium salts compared to their rare earth metal-substituted analogues [18]. In our most recent studies with RE 3-thiophenecarboxylate [14] and 3-furoate [13], we demonstrated the inhibition properties of the rare-earth-metal complexes. By comparing both sets of results (Tables 3 and 4), it is evident that the corrosion rates of alkali-metal carboxylates on steel are generally much higher than those of rare-earth analogues, indicating the value of rare-earth carboxylates in inhibition.

General Consideration
The experiments were carried out utilising commercially available chemicals and reagents of standard quality, which were used without undergoing any additional purification steps. The elemental Analysis Service Team at the Science Centre, London Metropolitan University, England, performed the elemental analyses. Melting points were determined using glass capillaries, and the values were reported without calibration. To obtain infrared (IR) spectra, a Nicolet™ iS™ 5 FTIR Spectrometer (ThermoFisher Scientific, Waltham, MA, USA) was employed in ATR mode, covering the range of 4000-500 cm −1 . Powder diffraction patterns were conducted at room temperature using a Bruker D2 PHASER diffractometer (Bruker corporation, Billerica, MA, USA). The measurements were conducted within the range of 2-60, using a 0.2 • divergence slit and with increments of 0.02 • . To generate X-ray powder simulations, the Mercury program provided by the Cambridge Crystallographic Data Centre [38] was employed. These simulations utilised the singlecrystal X-ray diffraction data. Thermogravimetric analysis (TGA) was performed on a TA instrument SDT 650 (TA Instruments, New Castle, DE, USA), using standard 90 µL alumina metal pans. The analysis was conducted at a heating rate of 10 • C/min under a nitrogen atmosphere with a flow rate of 50 mL/min.

Corrosion Testing
In order to assess the general corrosion and inhibition characteristics of the synthesised compounds, corrosion immersion experiments were conducted following the standard method ASTM G31-72 [44]. To conduct weight-loss tests, mild-steel alloy AS 1020 coupons were prepared by cutting them into approximately 20 × 20 × 1.5 mm dimensions and progressively abrading them using sanding sheets of grits ranging from 80 to 2000. The specimens were subsequently rinsed with distilled water, followed by ethanol, and dried under flowing N 2 gas. These coupons were immediately used for a series of immersion tests lasting up to 168 h (7 days). Both the sample and the control coupons were placed in separate beakers containing 0.01 M NaCl solutions, with and without the presence of 800 ppm of the inhibitor compounds. In each setup, coupons were fully immersed at mid-depth using Teflon strings. After completing the tests, the corrosion product adhering to the substrate was initially removed by subjecting it to mild sonication in clean, distilled water. Subsequently, the finest sanding papers were used with minimum force to prevent the removal of intact material. The experiments were carried out in duplicate to ensure reliability. Finally, the coupons were washed with ethanol and dried using N 2 gas. The corrosion rates (Rs) of the inhibitor solutions were determined by applying Equation (3), outlined below: where K is the A constant; W is the weight loss (g); A is the coupon area (cm 2 ); T is the time of exposure (168 h); D is the density of the test metal. The percentage corrosion inhibition (η) was determined using Equation (4):

Conclusions
Fourteen new complexes of the alkali metals with 3-thiophenecarboxylate (3tpc), 2methyl-3-furoate (2m3fur), 3-furoate (3fur), 4-hydroxycinnamate (4hocin), and 4-hydroxybenzoate (4hob) were synthesised. The preparations gave moderate yields of crystalline materials, and the complexes were characterised via IR spectroscopy, powder XRD, TGA, and microanalyses. Furthermore, X-ray crystallographic examination of the single crystals of all the compounds revealed unique structures for each of them. In the series of compounds, only the [Li 2 (2m3fur) 2 (H 2 O) 3 ] 2Li complex crystallised as a coordination dinuclear complex with four-coordinate lithium atoms. In contrast, all other complexes were found to be polymeric, with varying coordination numbers and stereochemistry. Compounds 1Cs and 4Rb incorporate [HL 2 ] − ligands rather than simple carboxylate groups and exhibit centrosymmetric O...H...O interactions between the carboxylate ions.
To investigate the potential application of the complexes as corrosion inhibitors, weight-loss measurements were conducted, but the compounds were less effective than their rare-earth counterparts.  Table S1: Crystal data and structural refinement for the alkali-metal carboxylate complexes; Table S2