Abstract
The reaction of BAL with Sn(IV), SnCl2·2H2O, Pb(II), Ag(I), Cd(II) and Pd(II) gave water-insoluble solids where BAL coordinated using its two sulfur atoms, BAL(S,S)M, while anhydrous SnCl2 gave BAL(O,S)Sn. Apart from the Sn(IV) and Ag(I) which should be monomeric, the structure of the other solids (monomeric or polymeric) could not be established. Indium(III) chloride did not react with either thiophenol or BAL in the absence of a base. In the presence of sodium hydroxide under 1:1:1, 1:2:2 and 1:3:3 molar ratios of InCl3/PhSH/NaOH, the products incorporated the formed sodium chloride and In(SPh)3 was isolated after aqueous work up in the 1:3:3 case. Thallium(III) aceteate gave Tl-SPh and PhSSPh. Thallium(I) acetate and thallium(I) fluoride also gave Tl-SPh in 100% and 75% yields, respectively. The lower yields with TlF is attributed to the non thiophilic HTlF2 formed during the reaction. Theoretical calculations of TlF, (TlF)2 and TlF2- revealed a novel structure of (TlF)2 and shed light to the non thiophilicity of TlF2-. Thallium(I) acetate and BAL gave the expected bis salt while the reactions of TlF under various stoichiometries of BAL:TlF were complicated. Summarizing these and previous results it seems that BAL cannot detoxify heavy metal cations by forming water soluble complexes.
Keywords
Introduction
British Anti-Lewisite (BAL, dimercaprol, 2,3-dimercapto-1-propanol) is an oily α,β dithiol synthesized in Oxford under the direction of Sir Rudolf Peters in 1940 [32] in search of antidotes to various chemical weapons. BAL was introduced by Peters in 1945 against Lewisite (Cl-CH=CH-AsCl2) [35, 45]. Although the R and S isomers were later prepared [1], BAL was and is used as a racemic mixture. BAL is effective in vivo (rats, guinea-pigs and man) not only against Lewisite, but also it can reverse the function of mitochondria [36] and E.Coli [40] after treating them with sodium arsenite (NaH2AsO3).
Many heavy metal cations deactivate enzymes and can cause metal poisoning in animals and humans. Intensive research in the 1940s, summarized by Stocken and Thompson [41], on enzymes, animals and humans using Sb(III), Bi(III) // Pb(II) // Zn(II), Cd(II), Hg(II) // Cu(II), Ag(I), Au(I) // Ni(II) // Cr(VI) // V(IV) established that BAL can reverse the intoxication by e.g. Hg(II) and Au(I) or can moderate the effect of e.g. Pb(II) and Ni(II). In other cases, e.g. Cd(II), Sb(III), Bi(III) the effect of BAL was controversial or BAL had no effect as in argyria. Despite these uncertainties, BAL is regarded as a prototype in chelation therapy [9, 10] based on chelate formation of the products of BAL with As(III) [23, 41].
It is worth mentioning that BAL treatment increased the urinary excretion of Pb(II), Au(I) and Cu(II) and the urinary plus feces excretion of Cd(II) [41] implying that BAL formed water-soluble complexes with these metal cations.
There are only a few reports on the products formed or isolated from the reaction of BAL with heavy metal cations in vitro. Thus, the reaction of BAL with Ni(II) at 9 < pH < 10 in the absence of air afforded soluble complexes [Ni(BAL)2]2 - and [Ni2(BAL)3OH]3 - [28], with Mn(II) the extremely air sensitive water-soluble MnBAL and [Mn(BAL)2]2 - were detected [30], while Zn(II) gave the water insoluble ZnBAL that dissolved with excess BAL giving [Zn(BAL)2]2 - and polynuclear species [30]. Highly oxygen sensitive, water-soluble mononuclear, [Fe(BAL)2]2 -, and polynuclear complexes were obtained with excess BAL [29]. It is not quite clear whether the coordination of Ni(II), Mn(II), Zn(II) and Fe(II) to BAL involves only its sulfur atoms. The conditions used in Leusing’s works may not be biologically relevant because of the presence of dioxygen in tissues. Gold(III) reacted with BAL under various stoichiometries, giving highly insoluble solids in which both Au(III) and Au(I) were detected [5]. BAL reacted in water with Hg(II) giving the insoluble linear polymeric BAL(S,S)Hg [8]. The reactions of BAL with various As(III), Sb(III) and Bi(III) compounds under 1:1 and 2:3 molar ratios of metal/BAL were studied [23]. The metals coordinated to BAL at its sulfur atoms and the products were insoluble in water. When BAL and Cu(II) reacted in water, the insoluble solid was found, indirectly, to have a 1:1 stoichiometry of copper to BAL [47]. Further experiments (P.V. Ioannou et al., in preparation, 2017) revealed that Cu(II) oxidized BAL being reduced to Cu(I) and the products interacted to give a highly insoluble solid that could not be solubilized and the use of BAL for excreting copper in the urine of Wilson’s disease patients is at present poorly understood. Finally, the interaction of BAL with a variety of heavy metal cations [Au(III), Ag(I), Hg(II), Cu(II), As(III), Sb(III), Bi(III), Ni(II), Co(II), Zn(II), Tl(I), Pb(II), Cd(II), Fe(III), Sn(II)] has been studied polarographically and probable structures of the products were formulated [49], not always in accord with those proposed in the literature [5, 23] and the results showed that Tl(I) was unreactive towards BAL.
Dihydrolipoic acid and its amide, being constituents of important mitochondrial multienzyme complexes such as pyrurate dehydrogenase, are α,γ dithiols and not many studies on their interactions with heavy metal cations have been reported [3, 21]. From the α,δ dithiols, dithioerythritol and dithiothreitol, their interaction with heavy metal cations did not receive much attention [22, 26]. Therefore there is a need for such studies and the results to be compared with those of the α,β dithiol BAL.
BAL being of biological importance for metal detoxification has been studied to a reasonable extent in vitro as described above. In this paper we report on the reaction of BAL with the remaining heavy metal cations of the p-block [Sn(II), Sn(IV), Pb(II), In(III), Tl(III) and Tl(I)] and some of the d-block [Cd(II), Ag(I) and Pd(II)]. The products isolated were studied and probable structures are proposed. Ab-initio theoretical calculations were made on TlF, (TlF)2 and TlF2- in order to explain the low yields of the reaction of TlF with thiols.
Experimental section
Materials
The following chemicals were obtained and used as received. From: Merck: anhydrous crystalline tin(II) chloride, crystalline tin chloride dihydrate, lead acetate trihydrate, thallium(I) acetate, cadmium acetate dihydrate, silver acetate, palladium acetate, chromium(III) chloride hexahydrate, 40% aqueous hydrofluoric acid, thiophenol, and racemic British Anti-Lewisite (2,3-dimercapto-1-propanol) containing by TLC (Et2O) traces of oxidized BAL at Rf 0.60, BAL running at Rf 0.75; Aldrich: thallium(III) acetate; Alfa: dibutyltin oxide; Alfa Aesar: vanadyl sulfate hydrated; Fluka: indium(III) chloride; Schuchard: thallium(I) fluoride; Ferak: vanadium pentoxide; Riedel - de Häen: chromium(III) chloride.
Silica gel H (Merck) was used for thin layer chromatography (TLC), and no dried solvents were used, except when stated, the drying agent being molecular sieves 4 Å.
Methods
TLCs were run on microslides and visualization of the spots was done by spraying with 35% sulfuric acid but more conveniently with iodine vapors. The reactions were run in centrifuge tubes, unless otherwise stated, while stirring. All reactions were followed by observing the disappearance of BAL by TLC in Et2O. Infrared spectra were obtained on a Perkin-Elmer, model 16 PC, FT-IR spectrometer, while 1H NMR (400 MHz) spectra were recorded on a Bruker DPX Avance spectrometer with internal TMS (0.00 ppm) as standard. Elemental analyses were obtained through the Centre of Instrumental Analyses, University of Patras, Patras, Greece.
Reaction of BAL with group 14 metal ions
With anhydrous SnCl2
To a solution of crystalline tin(II) chloride (380 mg, 2 mmol) in methanol (5 mL) in a 10 mL round-bottomed flask was added BAL (200μL, 2 mmol) and the solution stirred at room temperature (RT) for 1.5 h. Evaporation and drying in vacuum gave a viscous oil (673 mg) that solidified when covered with ether and left overnight to give the solvated product
With SnCl2·2H2O
To a solution of crystalline tin chloride dihydrate (1 mmol) in methanol (3 mL) addition of BAL (1 mmol) and stirring at RT for 1.5 h, gave a clear colorless solution that was evaporated and dried in vacuum to give an oil (365 mg). After treating the oil with ether overnight the solvated BAL(S,S)Sn·1.5MeOH
With (Bu2SnO) x
In a 25 mL round-bottomed flask was suspended dibutyltin oxide (249 mg, 1 mmol) in methanol (10 mL), BAL (100μL, 1 mmol) was added and refluxed for 3 h giving a clear colorless solution. TLC (Et2O) showed the product (Rf 0.90) quite close to BAL. Evaporation and drying in vacuum gave the product
With Pb(AcO)2·3H2O
To a solution of lead acetate trihydrate (379 mg, 1 mmol) in water (3 mL), BAL (100μL, 1 mmol) was added and the yellow suspension stirred at RT for 6 h. Centrifugation, washing with MeOH (1×2 mL) and drying gave the solvated
Reaction of BAL with some metal ions of d-block
With Cd(AcO)2·2H2O
In a round-bottomed flask cadmium acetate dihydrate (267 mg, 1 mmol) was dissolved in water (2 mL) and BAL (100μL, 1 mmol) was added. The white emulsion that was formed, was stirred at RT for 4 h. Evaporation (rotary 45°C) and drying in vacuum gave the solvated
With AgAcO
The reaction of BAL (1 mmol) with a suspension of silver acetate (2 mmol) in water (3 mL) gave a yellow suspension that was centrifuged, after stirring for 4 h at RT. The yellow solvate was washed with water (2 mL) and the solvate dried in vacuum to give the product
With Pd(AcO)2
To a solution of palladium acetate (112 mg, 0.5 mmol) in dichloromethane (3 mL), BAL (50μL, 0.5 mmol) was syringed in and the dark brown suspension, formed at once, was stirred at RT for 4 h. Centrifugation and washing with dichloromethane (2×2 mL) gave the solvated product
Reaction of InCl3 with thiols
With PhSH in the absence of a base
InCl3 and xPhSH (x = 1, 2, 3) in dry methanol did not react by TLC (petroleum ether) after 1 h at RT.
With PhSH in the presence of NaOH
2.5.2.a Under a 1:1:1 molar ratio. To a solution of indium(III) chloride (278 mg, 1.26 mmol) in dry methanol (2 mL), thiophenol (129μL, 1.26 mmol) was syringed in and the solution stirred at RT for 1 h. Solid sodium hydroxide (1.26 mmol) was added and stirred at RT for 2 h giving a white light suspension the TLC (petroleum ether) of which showed a faint, elongated spot at Rf 0.40. Centrifugation and washing with dry methanol (1 mL) gave a white solid (23 mg) that was not completely soluble in water indicating that it was not sodium chloride. Evaporation and drying in vacuum of the methanolic supernatant and washing gave a white solid (386 mg, 94% as
2.5.2.b Under a 1:2:2 molar ratio. Done as in 2.5.2.a at 0.7 mmol InCl3 scale. The methanol insoluble solid (33 mg) was less than the NaCl expected (41 mg) and it was not completely soluble in water. The product (90% as
2.5.2.c Under a 1:3:3 molar ratio. Using InCl3 (0.57 mmol) and thiophenol (1.71 mmol) in dry methanol, no reaction took place in the solution. Adding solid NaOH (1.71 mmol) and stirring at RT for 2 h the methanol insoluble solid weighed 47 mg (expected 67 mg NaCl) and it was partially soluble in water indicating that it was not only NaCl. From the methanolic supernatant and washing a semi-solid
The system InCl3 (0.82 mmol)/3 PhSH/3 NaOH in dry methanol stirred at RT gave a white, methanol insoluble solid (63 mg) not completely soluble in H2O indicating that it was not only NaCl (expected for 2 NaCl 96 mg, for 1 NaCl 48 mg). Evaporation of the methanolic supernatant and washing gave a white solid (387 mg), that was triturated with H2O for 1 h. Centrifugation gave an opalescent supernatant having pH ∼3 (which on leaving at RT precipitated traces of a solid) and a solid (134 mg, 37 %) that was (PhS)3In
With BAL in the absence and in the presence of a base
These experiments were done under a 1:1, 1:2 and 1:3 molar ratios of InCl3 to BAL by adding neat BAL to a solution of InCl3 in dry MeOH. The solutions obtained were treated with 1, 2 and 3 equivalents of solutions of NaOH in MeOH. All reactions were followed by TLC and after evaporation and drying the solids obtained were examined by elemental analyses, IR (KBr) and 1H NMR spectra. The results are described in Section 3.3.2.
The reaction of Tl(III) and Tl(I) with thiophenol
Reaction of Tl(AcO)3 with PhSH
To a solution of thiophenol (1.5 mmol) in dry MeOH (2 mL) was added dropwise a solution of thallium triacetate (191 mg, 0.5 mmol) in dry methanol (6 mL) and the yellow to light orange suspension stirred at RT for 4 h. TLC (petroleum ether) showed the presence of PhSSPh at Rf 0.35. Centrifugation and washing with MeOH (2×2 mL) gave a yellow solid and colorless supernatants. The combined supernatants were evaporated and dried to give PhSSPh (105 mg, 96 %) m.p. 59–61°C; lit. [48] 61–62°C. The yellow solid (157 mg, 100 %) was Tl-SPh (see below, 2.6.2) by m.p. and elemental analyses.
Reaction of TlAcO with PhSH
Dropwise addition of thiophenol (52μL, 0.5 mmol) to a solution of thallium(I) acetate (132 mg, 0.5 mmol) in methanol (2 mL) precipitated Tl-SPh as a yellowish powder. After stirring at RT for 3 h, centrifugation gave an opalescent supernatant and a yellow precipitate. Washing with methanol (2 mL) gave an opalescent supernatant that settled by adding Et2O (5 mL) and leaving overnight. The product (150 mg, 96 %) was a yellow powder insoluble in Me2CO, MeOH, H2O and DMSO. Thermal behavior: at 225°C slight shrinking, at 235°C turns to light orange, at approx. 255°C turns orange and at 270°C melts to a red foam; lit. [17] 258–260°C. Calculated for C6H5STl (Mr 313.53): C 22.98, H 1.61, S 10.23 %; found C 23.22, H 1.50, S 9.63 %. IR (KBr):
Reaction of TlF with PhSH
To a solution of TlF (224 mg, 1 mmol) in water (2 mL) was added dropwise, during 30 min while stirring, a solution of thiophenol (102μL, 1 mmol) in methanol (2 mL). At once a white solid formed that at the end of the addition turned yellow. Stirring was continued for 4 h and TLC (petroleum ether) showed traces of PhSSPh and a faint spot at Rf 0.0. Centrifugation, washing with H2O (1×2 mL) and MeOH (1×2 mL) gave Tl-SPh (233 mg, 74 %) as a yellow powder insoluble in H2O, MeOH and DMSO. The supernatants were clear, colorless but when combined gave a white suspension that was not investigated. Thermal behavior: at 230°C darkens to light orange, at 260°C shrinks and melts to orange-red mass and at 275°C foams. Calculated for C6H5STl (Mr = 313.53): C 22.98, H 1.61, S 10.23 %; found C 23.18, H 1.55, S 10.30 %. IR (KBr):
The effect of HF in the reaction of TlF with PhSH
Solutions of thiophenol (41μL, 0.4 mmol) dissolved in methanol (3 mL) were added dropwise to solutions of TlF (0.4x mmol) and HF (0.4y mmol) in water (3 mL) and the yellowish suspensions stirred at RT for 4 h. Centrifugation and washing with water (1×2 mL) and methanol (1×2 mL) gave clear colorless supernatants that when combined gave opalescent solutions or light suspensions that were not investigated further. The isolated solids were examined by elemental analyses, melting point and IR (KBr). The molar ratios of the reactants, the yields and the purities of the isolated solids are shown in Table 1.
The effect of HF on the reaction of TlF with thiophenol PhSH. A methanolic solution of PhSH was added during 15 min to an aqueous solution of TlF containing HF and after 4 h stirring was centrifuged and the yellow solid washed with water and methanol
The effect of HF on the reaction of TlF with thiophenol PhSH. A methanolic solution of PhSH was added during 15 min to an aqueous solution of TlF containing HF and after 4 h stirring was centrifuged and the yellow solid washed with water and methanol
a: by m.p., IR and elemental analyses, b: by m.p. and IR, c: impure by m.p. and elemental analyses but pure by IR.
TlAcO and BAL
To a solution of BAL (50μL, 0.5 mmol) in methanol (1 mL) was added dropwise, during 1 h, a solution of thallium(I) acetate (132 mg, 0.5 mmol) in methanol (3 mL). The orange suspension that formed at once, was stirred at RT for 18 h. TLC (Et2O) of the supernatant showed BAL that had not reacted. Centrifugation and washing with methanol (3×2 mL) gave the product
The supernatant and the washings were combined and AcOTl (132 mg, 0.5 mmol) was added. The orange suspension, after stirring at RT for 2 h, was worked up to give
TlF/HF and BAL
Methanolic solutions of BAL were added drop-wise to aqueous solutions of TlF under 1:1, 1:3 and 1:5 molar ratios BAL:TlF and the suspensions formed stirred at RT for at least 3 h. TLC (Et2O) revealed the presence of BAL that had not reacted and very small amounts of oxidized BAL. The orange solids after centrifugation and washing with water and methanol were examined by elemental analyses and IR (KBr). The results are discussed in Section 3.4.2.
To aqueous solutions of TlF containing added aqueous HF were added drop-wise methanolic solutions of BAL so as the molar ratios of BAL/TlF/HF were 1:2:2, 1:2:4, 1:2:6, 1:3:2 and 1:6:4 and the heterogeneous systems worked up as above. The results are discussed in Section 3.4.2.
Computational details
Molecular orbital ab-initio calculations based on Hartree-Fock (HF) and second order Møller-Plesset (MP2) perturbation theory as implemented in the Gaussian 09 program package [16], were utilized in the present study in order to calculate the structural details and study the intermolecular interactions/rections of TlF, (TlF)2, TlF2- and MeSH species. Calculations have been also performed in the presence of water solvent by placing the species (solutes) in a cavity within the solvent reaction field (SCRF). The SCRF calculation method used was based on the Polarizable Continuum Model (PCM) and creates a solute cavity via a set of overlapping spheres [43]. The basis sets that have been used are all-electron double zeta plus polarization functions [H, C, F basis sets [7]; Tl basis set [6]]. Specific basis sets have been obtained by use of the Basis Set Exchange (BSE) software and the EMSL Basis Set Library [14, 39]. The geometries of all structural models have been fully optimized at the MP2 all-electrons level of theory.
Results and discussion
The reaction of BAL with Sn(II), Sn(IV) and Pb(II) compounds
Both anhydrous and dihydrated SnCl2 reacted with BAL in methanol giving soluble products probably indicating coordination of the HCl produced to Sn(II) because the isolated solids were insoluble in methanol. Based on the IR (KBr) spectrum anhydrous SnCl2 gave solvated BAL(O,S)Sn

Products obtained from the reaction of BAL(OH)(SH)2 with Bu2SnO and AgAcO and products expected from its reaction with the divalent cations Sn(II), Pb(II), Cd(II) and Pd(II). Some products were solvated and/or hydrated.
Dibutyltin oxide reacted smoothly with BAL in refluxing methanol giving the monomeric
Most likely the monomeric
From the biological point of view the insolubility in water of the products of BAL with Sn(II) and Pb(II) indicates that BAL cannot detoxify these two metals by forming water soluble products, although it is claimed [41] that BAL increased the urinary excretion of Pb(II).
BAL and Cd(AcO)2·2H2O in water gave an emulsion containing very fine particles that could not be isolated either by centrifugation or filtration. The product
Silver acetate and BAL in water under 1:1 molar ratio gave the bis salt
The polymeric, solvated
BAL did not react (by color and TLC) in methanol with CrCl3 (suspension) or CrCl3·6H2O (solution), with VOSO4 in water in the absence and in the presence of NaOH to scavenge the expected H2SO4, while V2O5 as a suspension in water oxidized part of the BAL. Coordinated BAL was not detected in the isolated solids by IR. The non reactivity of BAL with these cations may be explained by the HSAB principle (hard cations, soft thiol).
Studies on detoxification of Cd(II) gave conflicting results but urinary excretion was noticed in certain cases and BAL was of no value in treating argyria in rats [41]. Our results indicate that no water-soluble complexes are formed between BAL and Cd(II), Ag(I) or Pd(II).
Reaction of InCl3 with thiols in methanol
With PhSH in the absence and in the presence of NaOH
There seems to be no reports on the nature of the species formed when InCl3 is dissolved in methanol but there are with water as solvent [33]. Thus, InCl3 dissolves exothermally in water practically without hydrolysis indicating coordination of water to InCl3, and the tetrahydrate, InCl3·4H2O, can be crystallized.
We found no literature reports on the reaction of InCl3 with thiols and our experiments showed that InCl3 and xPhSH (x = 1, 2, 3) in methanol did not react (TLC and smell). Evaporation and drying in vacuum over P2O5 gave back InCl3 (100% recovery).
In the presence of a base or with alkoxides, InCl3 gives the indium(III) alkoxides, In(OR)3 [46]. However, there seems to be no reports on the products obtained with less than 3 equivalents of a base.
Even less information exists on the reaction of InCl3 with thiolates. Thus, the reaction of equivalent amount of InCl3 and PhSNa in refluxing methanol and non-aqueous work up gave the air-stable In(SPh)3 that, being a Lewis acid, can coordinate Cl- and other Lewis bases [11]. Substituted thiophenolates of In(III) have also been prepared [38] and investigated as building block for polymers [19], and other uses of indium(III) organothiolates can be envisioned [31]. From the intermediates, Cl-In(SPh)2 and Cl2In-SPh, only the former has been prepared by insertion of InCl into PhSSPh [34].
All the reactions of InCl3/PhSH/NaOH under 1:1:1, 1:2:2 and 1:3:3 molar ratios in methanol gave traces to very little insoluble solids that were also mostly insoluble in water thus indicating that the NaCl produced was coordinated to the soluble products. These, in turn, when evaporated and dried in vacuum, revealed the evaporation of HCl/H2O, H2O and H2O, respectively. The elemental analyses, solubilities and thermal behavior of the solids pointed to compounds tentatively formulated as

Proposed simplistic structures of the products obtained from the reaction of InCl3 with PhSH in the presence of NaOH in methanol under 1:1:1
Compounds with an In-O-In linkage should exist but probably they will be hydrolytically unstable. During the preparation of
The different yields of In(SPh)3 obtained with non-aqueous work up, 60% [11], and with aqueous work up, 37% [this work], tend to indicate that the In(SPh)3Cl- complex cannot keep tenaciously the Cl- and rearrangement can take place giving In(SPh)3 and non-identified complexes of In(SPh)3.
BAL, being an α,β dithiol capable of forming thermodynamically stable chelates, can probably react with InCl3 in the absence of a base. However, the reactions of InCl3 with BAL in dry methanol under 1:1, 1:2, 1:3 and even 2:1 molar ratios revealed, by TLC, that BAL had not reacted. Evaporation and drying in vacuum gave solids soluble in organic solvents but insoluble in water, the IR (KBr) of which showed the -SH group and their 1H NMR spectra in various solvents revealed broad signals for the BAL backbone protons not allowing any assignments. Thus, these reactions of InCl3 with BAL gave mixtures insoluble to sparingly soluble in water.
Adding NaOH to methanolic solutions of InCl3 and BAL so as the molar ratios to be 1:1:1, 1:2:2 and 1:3:3 InCl3/BAL/NaOH gave solutions (opalescent in 1:3:3 case) that by TLC contained BAL. All BAL had reacted only in the 2:1:2 case. Evaporation and drying in vacuum gave solids composed of compounds that could not be identified by elemental analyses and spectra.
Overall, BAL and InCl3 in the absence and in the presence of a base did not react to a significant extent and the solids obtained after evaporation and drying in vacuum could not be characterized, and their biological significance, if any, remains abscure.
The reaction of Tl(III) and Tl(I) with thiols
With PhSH
Thallic acetate oxidized thiophenol to its disulfide, PhSSPh, with concomitant production of thallous thiophenolate [17, 25] obtained in 100% yield.
Thallous acetate and thiophenol in methanol gave the yellow Tl-SPh in 100% yield, while aqueous TlF and methanolic PhSH produced only ∼75% Tl-SPh.
A most likely hypothesis for the lower yields obtained from TlF and PhSH is that the HF produced, coordinated to TlF giving species not electrophilic towards PhSH. Table 1 indicates that the more HF initially present the less the yield of the (impure) product and implies that equilibria involving (TlF)2, TlF, TlF·HF, TlF·2HF and TlF·3HF can exist in solution with added HF and the dominating species is probably (F-Tl-F)-H+, its concentration increasing with increasing amounts of HF due to mass effect. The above compounds seems reasonable because in the literature there are analogous species, e.g. (HF)2 in concentrated and (F-H-F)-H3O+ in dilute aqueous solutions of HF [37], KF·HF (Frémy’s salt), KF·2HF to KF·12HF [13], AgF·HF and AgF·3HF the structures of the latter being unknown [37].
Theoretical calculations on thallium(I) fluoride indicate that TlF is a stable molecule having a considerable amount of covalent bond (2.05 Å) that when hydrated the bond elongates to 2.09 Å, Fig. 3, and this is in accord with the notion [27] that thallium(I) halides are classical examples of “incompletely dissociated” 1:1 electrolytes, i.e. form ion pairs [15] that are not hydrates [44]. Thallium(I) fluoride in the solid state has crystal structures depending on temperature and pressure and are not composed of monomers [2]. In the vapor state TlF forms monomers and oligomers, the monomer exhibiting ionic bonding [12].

Structures of TlF, (TlF)2 and TlF2- in the absence and in the presence of water, i.e. having a hydration environment. Bond lengths in Å.
The dimer (TlF)2 predominating in the vapor phase has been proposed to have a linear structure F-Tl-Tl-F, with a Tl-Tl bond, where the Tl atom has two sp and two p orbitals [24]. However, from photoelectron spectroscopic studies a rhombic dimer has been suggested having a cited Tl-Tl bond of 3.678 Å and four Tl-F bonds of 2.29 Å [42]. Our theoretical calculations indicate that the dimer is more stable by 27.1 kJ/mol compared to 2 TlF monomers and in water environment is slightly destabilized, Fig. 3. The dimer is neither linear or rhombic but has a bend structure implying a stereochemically active s2 electron pair. The distance Tl···Tl, being ∼2.85 Å and unaffected by solvation is quite long for a covalent bond. The geometry around each Tl(I) (F, Tl, s2) is approximately trigonal as the 114 angle tends to indicate the dihedral angle being 180.
As HF adds F- producing (F-H-F)-, TlF adds F- to give the anion (F-Tl-F)- with a bend structure, Fig. 3, implying that the s2 electrons of Tl(I) are stereochemically active but, as Table 1 indicates, chemically unreactive towards thiols. When hydrated, its structure changes a little implying that water molecules interact with both the s2 electrons and the fluorine atoms.
In the absence of HF, the interaction of TlF with the model thiol CH3SH seems to favor the arrangement shown in Fig. 4 where the optimized distance Tl···S (∼4.0 Å) is too long for a concerted four centered reaction. The reaction is probably aided by the partial ionization of fluorine which then abstracts the hydrogen from the thiol to give HF or aqueous HF followed by attack of the incipient RS- to Tl+. In the presence of HF where (F-Tl-F)-H+ is most likely the predominant species, MeSH interacts as shown in Fig. 4. In this case too, the distance Tl···S is again ∼4.0 Å but when HF is expelled, the resultant TlF in the presence of HF does not develop a positive charge to attract the RS-.

The interaction of TlF and TlF2- (anhydrous and hydrated) with the model thiol CH3-SH.
BAL did not coordinate to Tl(AcO)3 to give Tl2(BAL)3 under a 3:2 molar ratio in methanol. The supernatant contained, by TLC, BAL and oxidized BAL and the insoluble orange solid could not be characterized. However, TlAcO precipitated
The reaction of aqueous TlF with methanolic BAL under 1:1, 2:1, 3:1 and 5:1 molar ratios of TlF/BAL gave suspension and very small amounts of orange solids. BAL and oxidized BAL were detected by TLC in the 1:1 but not in the other cases. The orange solids could not be characterized by elemental and IR (KBr) analyses. In the presence of added aqueous HF the weights of the solids diminished even further, their colors were not orange and their analytical data did not point to a specific molecule.
Thallium is the most toxic element of group 13 of the Periodic Table, and its action has been mentioned a few times even the fiction literature, e.g. Agatha Christie’s “The pale horse”. Our results indicate that Tl(III) oxidize monothiols to their disulfides giving insoluble thallium(I) mercaptides also obtained from Tl(I) compounds. If Tl(I) compounds are found in a cell, then BAL produces the insoluble disalt
