Gold(I) and gold(III) complexes of triazolyl-functionalised NHCs

Abstract

Novel triazolyl-functionalised imidazolium salts and their corresponding Au^I/Au^III complexes were synthesised and characterised by spectroscopic techniques and X-ray diffraction studies. The reaction of (mono and bis carbene) Au^I complexes with thionyl chloride gave Au^III complexes. These Au^III complexes are reduced to Au^I complexes by DMSO dissolution, and react with silver nitrate to afford chelating complexes, where the triazolyl N atoms bind to gold. The triazolyl-functionalised Au^I complexes showed potency against ovarian cancer cells (OVCAR-8 cells, IC₅₀ < 15 μM).

---

Introduction

N-Heterocyclic carbenes (NHCs) are well known as ligands having strong donor properties akin to those of phosphines. Generally, however, highly functionalised NHCs are more readily accessible than phosphines, since they can be conveniently generated from air-stable azolium salts^1,2 and an array of other stable precursor compounds.^3–6 The ready synthetic accessibility of functionalised NHCs has meant that properties of metal–NHC complexes, such as lipophilicity and steric bulk around the metal, can easily be tailored to a broad range of applications, including catalysis,^7–13 imaging,^14–20 and chemotherapy.^21–30 In cases where robust metal–NHC complexes are required, stability can be enhanced by incorporation of the NHC moiety into a multidentate or a pincer ligand.^31–33

An interesting type of NHC results from combining an imidazolium salt and a triazole in the same compound by applying an azide–alkyne cycloaddition or “click” reaction.³⁴ One application of triazoles obtained by the click reaction is their use as ligands in metal complexes.³⁵ 1,2,3-Triazole ligands have been used previously for the generation of well-defined abnormal/mesoionic complexes of palladium, ruthenium, iridium, copper, silver, gold and platinum in which the triazolyl ligand binds to the metal centre as a carbene.^33,35–40 1,2,3-Triazolyl ligands have also been used to form triazolyl–gold complexes in which the triazolyl unit binds to gold via C (as an abnormal carbene) or N, and these complexes have been tested for their catalytic activity.^{33,39,41–43} The triazolyl–Au complexes or triazolyl–Au complexes having hemilabile ligands have proven to be efficient catalysts for enone formation from propargylic esters and alcohols,⁴² for intermolecular hydroamination of amines with alkynes,⁴¹ for formation of oxazolines by condensation of aldehydes and isocyanides,³⁹ and for synthesis of γ-substituted γ-butyrolactones.⁴⁰

Combining triazolyl-type ligands with NHCs in the same ligand framework could lead to interesting properties and coordination modes of the resulting complexes.^38,44 For example, chelating triazolyl-functionalised NHC ligands (Tz-NHCs) have been used in combination with ruthenium in the transfer hydrogenation of ketones.³⁸ Similar ligand structures have been used in medicine,⁴⁵ metallopolymers,⁴⁴ and supramolecular chemistry.⁴⁶ Gold complexes derived from imidazolium salts that contain triazolyl units attached at the imidazolyl C4 or C5 positions have been reported.^45,47 Tubaro and coworkers have reported Au–NHC complexes of Tz-NHC ligands where the Tz units are pendant from the N of the NHC ring, but the Tz-NHC ligands did not chelate the Au centre.^48,49 In this work the synthesis of both neutral and cationic Au^I complexes of two new Tz-NHCs is reported. The Au^I complexes were then oxidised to give the corresponding Au^III complexes. The anticancer activity of some of these complexes was investigated in OVCAR-8 (ovarian cancer) cells. The new Tz-NHC ligands include a bromine atom to allow the possibility of using the NanoSIMS technique to track the fate of Au complexes in cells in future imaging studies.⁵⁰

Results and discussion

Synthesis of triazolyl-functionalised imidazolium salts

The synthetic pathway to the triazolyl-functionalised imidazolium salts is shown in Scheme 1. It relies on applying the copper catalysed azide–alkyne cycloaddition reaction to appropriate azide and alkyne coupling partners. The known azide 2 was obtained in high yield (83%) from 2-bromobenzyl bromide 1 by treatment with sodium azide at room temperature in dimethylformamide.⁵¹ Azide 2 was then reacted with 1-methyl-3-propargylimidazolium chloride 3 by adapting a literature procedure,³⁸ using catalytic amounts of a Cu/CuSO₄ mixture under heating at reflux. After workup, which included filtration through a column of silica gel, the triazolylimidazolium salt 5 was obtained as a brown oil in 45% yield. Using a similar procedure, reaction of azide 2 with 1,3-dipropargylimidazolium chloride 4 gave the new triazolylimidazolium salt 6 as a tan solid in 77% yield.

Scheme 1 Synthesis of triazolylimidazolium chloride salts 5 and 6.

Synthesis of Au–NHC complexes having pendant triazolyl groups

Two neutral Au^I–NHC complexes 7 and 8 were then prepared by reacting imidazolium salts 5 and 6 with silver oxide according to a well-known literature procedure (Scheme 2).⁵² The presumed intermediate Ag–NHC complexes derived from 5 and 6 are treated in situ with dimethylsulfide gold chloride [(Me₂S)AuCl] in one-pot reactions. Complex 7 was obtained as an off-white powder in relatively low yield (31%). Significant losses occurred due to the need for the material to be recrystallised twice, initially from acetonitrile/diethyl ether, and then from ethyl acetate at 0 °C, to achieve satisfactory purity. On the other hand, complex 8 was obtained as a white powder in 77–96% yield by filtration either through a plug of silica (4 : 1 DCM/MeCN) or Celite, followed by precipitation with DCM/hexanes.

Scheme 2 Synthesis of the Au^I–NHC complexes having pendant triazolyl groups.

The cationic bis-NHC Au^I complexes 9 and 10 were prepared by reaction of the salts 5 and 6 and with ∼0.5 equivalents of (Me₂S)AuCl in the presence of potassium carbonate⁵³ (8–10 equiv.) in acetonitrile (Scheme 2). Workup of 9 was straightforward and the compound was obtained as an off-white powder in 86% yield. Compound 10 was poorly soluble in acetonitrile, so during workup the acetonitrile was evaporated and the residue dissolved in ethanol. Filtration through Celite followed by evaporation of the solvent and washing with diethyl ether then gave 10 as a white powder in 72% yield.

In terms of solubility, it should be noted that both salts 5 and 6 are soluble in H₂O but differ in their solubility profile for less polar solvents, with salt 6 being soluble in DCM, while salt 5 is not. The neutral complexes 7 and 8 and the cationic complex 9 are all soluble in DCM but are insoluble in water. The cationic complex 10 is the only complex to show good solubility in water, but it is not soluble in DCM.

Oxidation of the triazolyl-functionalised Au^I–NHC complexes

Synthesis of Au^III–NHC complexes via oxidation of Au^I–NHC complexes with thionyl chloride. The reaction of Au^I–NHC complex 8 with thionyl chloride (220 equiv.) in DCM at reflux for ∼44 h produced Au^III complex 12 in nearly quantitative yield (Scheme 3). Attempts to oxidise 8 using only a small excess (12 equiv.) of thionyl chloride at room temperature were not successful. In this experiment, the starting material 8 was the only compound detected by ¹H NMR spectroscopy in aliquots taken from the reaction mixture after 16 and 40 hours, respectively. Adding further thionyl chloride (∼200 equiv.) to the reaction mixture, and heating at reflux for a further 26 h, resulted in a mixture of the Au^III complex 12 and the Au^I complex 8 in a ratio of 4 : 1 respectively, and the reaction had gone to completion after heating at reflux for a further 22 h (total time at reflux ∼48 h). Removal of volatiles under vacuum and washing of the residue with diethyl ether left the Au^III complex 12 as a yellow powder, in excellent yield (97%) and purity that is completely free from 8 (Fig. S5†). The choice of SOCl₂ oxidizing agent is vital, the ESI† describes complications encountered upon employment of KAuCl₄ and I₂.

Scheme 3 Oxidation of Au^I–NHC complex 8 to Au^III–NHC complex 12 by SOCl₂.

While the NMR spectrum of trichloride 12 (Fig. S5†) in CD₂Cl₂ suggests a single species in solution, 12 converts to the monochloride 8 in DMSO-d₆. For example, the ¹H NMR spectrum of 12 in DMSO-d₆ showed 12 as the primary complex present immediately after preparation, but with trace 8 also detected (Fig. S7, top spectrum†). After 44 h, about a third of complex 12 had been reduced back to 8, and after 25 days, the sample contained only 8 (Fig. S7, bottom spectrum†). The ESI† describes even greater apparent solution instability of related iodo analogues.

Curiously, 12 showed similar stability in acetone-d₆ and DMSO-d₆, even though DMSO is a more strongly coordinating solvent. A sample prepared by dissolving 12 in acetone-d₆ was monitored by ¹H NMR spectroscopy (Fig. S8†). After 4 days, the sample contained 8 and 12 in a ratio of ∼1 : 1. In this experiment, a drop of water was added to the sample after 5 days, to see if water made the reversion of 12 to 8 occur more rapidly. The ¹H NMR spectrum recorded immediately after addition of water showed a similar ratio of 8 and 12 seen in the ¹H NMR spectrum recorded a day previously, indicating that addition of water did not immediately cause 12 to be reduced to 8, but nevertheless after 13 days, no 12 remained in the sample.

In view of the encouraging results obtained using thionyl chloride for the oxidation of Au^I–NHC complex 8 to Au^III complex 12, similar oxidations were explored for Au^I complexes of the form [Au(NHC)₂]⁺, 9 and 10. Complex 9 was oxidised to Au^III complex 13 by treatment with excess thionyl chloride in DCM under reflux for 4 days. After workup complex 13 was obtained as a pale yellow powder in 90% yield (Scheme 4).

Scheme 4 Oxidation of bis Au^I–NHC complexes by SOCl₂.

In the same way, oxidation of the bis Au^I–NHC 10 with excess thionyl chloride gave the Au^III complex 14 as a yellow powder in 86% yield (Scheme 4). In this case, however, the starting material 10 was poorly soluble in DCM. When 10 was stirred in DCM a cloudy suspension was obtained, but when SOCl₂ was added a clear solution formed. Interestingly, 10 is soluble in thionyl chloride, so the oxidation of 10 to 14 could be achieved simply by addition of neat reagent without any DCM, and the yield of 14 was >80% in both variations.

The pure compounds 13 and 14 are not soluble in DCM or acetone, and while 13 has only very limited solubility in acetonitrile, 14 is insoluble in this solvent. Fortunately, 13 and 14 were both soluble and sufficiently stable in DMSO-d₆ that they could be characterised by NMR spectroscopy in freshly-prepared DMSO-d₆ solutions. Nevertheless, the complexes slowly reverted to their Au^I precursors in DMSO-d₆, and for 14, NMR spectra of even freshly prepared solutions showed signals due to traces of 10. Elemental analyses of both 13 and 14 indicated that both complexes were pure, suggesting that the traces of 10 seen in NMR spectra of 14 were due to reduction of 14 by reaction with the solvent.

Comparing different pathways to access Au^III complexes. The work described above shows that oxidation of Au^I–NHC complexes by thionyl chloride affords isolable Au^III compounds. Employment of SOCl₂ was much more successful than our attempts with I₂ or KAuCl₄ as oxidants (see Section 1.1 of the ESI†). There is literature precedence for the importance of a chlorinating oxidiser in stabilising Au^III centres.⁵⁴ Furthermore, the approach to Au^III–NHC complexes via oxidation of Au^I–NHC complexes by thionyl chloride is superior to an approach by directly reacting KAuCl₄ with Ag–NHC complexes. This finding is consistent with a number of previous studies that showed that reaction of Au^III salts with sources of NHCs is invariably complicated by reduction to Au^I (ref. 55–57) (see ESI†).

Another important finding here is that the choice of solvent is significant because Au^III complexes 11–14 can be easily reduced, especially 11, which is reduced immediately in DMSO-d₆. By avoiding DMSO-d₆ and acetone-d₆ and utilising less polar and non-coordinating solvents such as dichloromethane-d₂, reduction of Au^III to Au^I is minimised. This finding is consistent with reports by Hirtenlehner et al.⁵⁸ of the ease of reduction of Au^III–NHC complexes to Au^I–NHC complexes, although those studies were primarily focused on photo-reductions and there was no comment on DMSO as a potential reductant. The present study emphasizes the importance of solvent on stability of Au^III–NHC complexes, even though some Au^III–NHC complexes are stable DMSO-d₆.⁵⁸

It is not surprising that 11–14 are easily reduced to their Au^I counterparts as previous studies have shown that Au^III–NHCs can be easily reduced to Au^I–NHCs by such reducing agents as phosphines, sulfides⁵⁹ or when chloropyridine is utilised as the solvent of the reaction.^55,56,60 In addition, it has been reported that when some Au^III–NHC complexes are utilised as precatalysts for hydration of alkynes, reduction to Au^I–NHC complexes occurred.⁶¹

Studies of Au–NHC complexes by NMR spectroscopy

¹H NMR spectroscopy was the primary tool used to monitor reactions and identify products of reactions explored in this work. ¹H and ¹³C NMR spectra of the new Tz-Im salts 5 and 6 and the corresponding Au^I and Au^III complexes are consistent with the structures determined by X-ray crystallography. ¹H and ¹³C NMR chemical shifts seen for 5 and 6 and the corresponding Au^I and Au^III complexes are in the ranges seen for similar compounds reported in the literature.^{6,22,30,56,58,62–64} The NMR signals associated with the imidazolyl and triazolyl moieties of the imidazolium salts and NHC complexes (e.g., Scheme 5) are particularly informative. Changes in signals associated with imidazolyl and triazolyl moieties were used as indicators of coordination of these moieties to Au. Signals due to other protons and carbons showed little variation in splitting patterns or chemical shifts whether they arose from an imidazolium salt, an Au^I complex, or an Au^III complex.

Scheme 5 Informative ¹H and ¹³C environments for characterisation.

Consider, for example, the ¹H NMR spectra for the Tz-Im salt 6 and the corresponding Au^I complex 8 and Au^III complex 12 shown in Fig. 1. The ¹H NMR spectrum of the salt 6 in DMSO-d₆ shows three downfield singlets, due to the imidazolium H2 proton (9.36 ppm), the triazole ring proton (8.27 ppm) and the H4/H5 protons (7.77 ppm). These signals are in the regions expected based on previous reports for similar compounds.³⁸ Signals for the protons of the aromatic ring are seen in the region 7.0–7.6 ppm and are well-resolved, and the protons of the two methylene groups occur as singlets at 5.7 ppm (C9 methylene, between the triazolyl and phenyl rings) and 5.5 ppm (C6 methylene, between triazolyl and imidazolyl rings). When 6 was treated with Ag₂O followed by (Me₂S)AuCl, formation of the corresponding Au^I complex 8 was indicated by the disappearance of the ¹H NMR signal at 9.36 ppm due to the imidazolyl proton (Fig. 1(b)). The other significant differences between the ¹H NMR spectra of 6 and 8 were the chemical shifts due to the H4/H5 protons (7.49 ppm, shifted upfield from 7.77 ppm for 6) and C6 methylene protons (5.46 ppm, shifted upfield from 5.5 ppm for 6). These upfield shifts can be attributed to the decrease in the Lewis acidity⁶⁵ of the substituent at the imidazolyl C2 position, H in 6 compared to Au^I in 8. When the Lewis acidity of the substituent on C2 decreases, there is less withdrawal of the electron density from the imidazolyl ring, resulting in increased electronic shielding of the H4/H5 protons and the C6 methylene protons. The other protons in 8, being more remote from C2 and the effects of the substituent on C2, have chemical shifts quite close to their counterparts in 6.

Fig. 1 ¹H NMR spectra (DMSO-d₆) of: (a) the imidazolium salt 6 showing the characteristic downfield signal for the H2 proton at 9.36 ppm; (b) the Au^I–NHC complex 8; and (c) the Au^III–NHC complex 12. Spectra (a) and (b) were recorded at 500 MHz, and spectrum (c) was recorded at 600 MHz.

In the ¹H NMR spectrum of 12 (Fig. 1(c)), the chemical shifts of the H4/H5 and C6 methylene protons are shifted significantly downfield of their counterparts in 8, and slightly further downfield than their counterparts in 6. These shifts for 12 can be attributed to the Au^IIICl₃ substituent at the imidazolyl C2 being a substantially stronger Lewis acid than the Au^ICl substituent in 8, due to the electron withdrawing effects of Au^III and the three attached electronegative chlorine atoms. Again, the other protons in 12, being more remote from C2 and the effects of the substituent on C2, show little difference in chemical shifts to their counterparts in 6 or 8. Similar trends in ¹H NMR chemical shifts were seen for the key protons in the other Au complexes discussed herein, and are summarized in Table S2.†

In the ¹³C NMR spectra of compounds 5 and 6 and the corresponding Au^I and Au^III complexes, for most carbons there is little variation in chemical shifts amongst the series imidazolium salt/Au^I–NHC complex/Au^III–NHC complex etc. For example, amongst all of compounds 5 and 6 and their corresponding Au^I and Au^III complexes, the ¹³C chemical shifts of the triazolyl C8 carbons all fall within the range 122–126 ppm and the ¹³C chemical shifts of the imidazolyl C4/C5 carbons all fall within the range 121–126 ppm (Table S3†). The C2 (carbene) carbons, however, show huge differences in chemical shifts within the series imidazolium salt/Au^I–NHC complex/Au^III–NHC complexes and interesting trends are evident (Table S3†). The chemical shifts of the carbene carbons fall into narrow but very distinct ranges depending on the class of compound: imidazolium salts ∼ 136 ppm; (NHC)Au^ICl ∼ 169 ppm; (NHC)Au^IIIX₃ 140–141 ppm; Au^I(NHC)₂ ∼ 183 ppm; Au^III(NHC)₂Cl₂ 151–152 ppm. These ranges show that oxidation state of Au, as well as the ligands attached to Au, cause significant differences in the chemical shift of the carbene carbon. In all cases, the carbene chemical shift for Au^III complexes is about 30 ppm upfield of the carbene chemical shift for Au^I complexes, which has been attributed to the increased Lewis acidity of Au^III compared to Au^I causing increased delocalisation of electron density from the imidazolyl C4–C5 double bond to the carbene carbon via the imidazolyl aromatic system.⁵⁶

Syn and anti isomerism of Au^III complex 13

Oxidation of the Au^I complex 9 by excess of thionyl chloride in DCM led to formation of the Au^III complex 13 (Scheme 4). The ¹H NMR spectrum of 13 (Fig. S6†) showed two distinct sets of signals, indicating the presence of two species in a ratio of 62 : 38, but elemental analysis was consistent with the elemental composition expected for 13 (with traces of occluded CH₂Cl₂). When the temperature was increased, various signals for the two species broadened and coalesced, resulting in an “averaged” spectrum at high temperature (Fig. S10†), indicating that the two species interconvert on the NMR timescale. The two species detected by NMR spectroscopy are therefore assigned as syn and anti conformations of 13 that differ by the orientation of the NHC moieties about the C–Au–C axis (Scheme 6).

Scheme 6 Syn and anti conformers of Au^III complex 13.

The existence of two observable conformations of 13 presumably results from the square planar geometry of Au^III complex 13 and steric bulk provided by the chlorido ligands on Au, which can restrict the rotation of NHC ligands around the C–Au^III–C axis. Two conformations are not seen in the case of the Au^I complex 9 because, without the bulk provided by additional chlorido ligands on the Au centre, there is no substantial barrier to rotation of the NHC ligands about the C–Au^I–C axis. This sort of conformational isomerism has been reported previously for other square planar complexes of the form M(NHC)₂X₂.^1,58,66,67 Intuitively, the dominant conformation is expected to be anti-13, since the syn conformation would be destabilised by unfavourable steric interactions between the bromobenzyltriazolyl moieties. The ¹H–¹H ROESY NMR spectrum for a sample containing anti-13 and syn-13 showed a correlation between the signals due to the CH₃ and H6 protons of the major isomer, but not for the minor isomer. The observed correlation is consistent with the assignment of the major isomer as anti-13, since in this conformation the CH₃ protons in one NHC ligand will be close in space to the H6 protons of the other NHC ligand. In syn-13, the CH₃ protons in one NHC ligand are remote from both the H6 protons in the same NHC ligand and the H6 protons in the other NHC ligand.

The variable temperature NMR study of 13 in DMSO-d₆ (Fig. S10†) shows various signals broadening and coalescing as the temperature is raised. This behaviour is reversible, as expected for an NMR exchange process involving interconversion of conformations. Two sets of signals, for the CH₃ groups and for the imidazolyl H4 protons in the syn and anti conformations, are sufficiently well-separated that they can be integrated in the room temperature spectrum. Integration of these signals indicated that ratio of the populations of nuclei in the anti and syn conformations is 62 : 38 respectively. At room temperature, the chemical shift separation for the signals due to the CH₃ protons of the two conformations is 50.0 Hz, and coalescence of those signals occurs at ∼338 K (Fig. S10†). Similarly, the chemical shift separation for the signals due to the imidazolyl H4 protons of the two conformations is 30.0 Hz, and coalescence of those signals occurs at ∼328 K (Fig. S10†).

For an exchange between equally populated sites, the rate constant k at the coalescence temperature (the temperature at which signals of the exchanging nuclei are just merged together) can be estimated as k = 2.22Δν, where Δν is the chemical shift separation (in Hz) of the nuclei in the absence of exchange.⁶⁸ In the case of the interconversion of anti-13 and syn-13, the populations of exchanging nuclei are unequal (see ESI†).

The results in (Table S1†) appear to be reasonable. The rate constants increase with temperature as expected, and the relative size of the rate constants for the anti → syn and the syn → anti processes are consistent with the observed ratio of anti-13 : syn-13 in solution. In principle, it should be possible to calculate values for ΔS^‡ and ΔH^‡69 from the results in (Table S1†), but the uncertainty in the coalescence temperature is probably too large for such calculations to be meaningful.

The results in (Table S1†) are in the range seen for the results of previous studies for the interconversion of syn and anti conformations of Au^III–NHC complexes. Hirtenlehner et al.⁵⁸ used quantum chemical calculations to estimate the barrier to rotation for interconversion of syn and anti conformations of a [(NHC)₂Au^IIIBr₂]⁺ complex as ∼60 kJ mol⁻¹, but noted that the result was not consistent with their ¹H NMR study (200 MHz, DMSO-d₆). In their NMR study, ¹H NMR signals separated by ∼40 Hz at room temperature were not near coalescence even when the temperature was raised to 90 °C, indicating a high rotational barrier, expected to be significantly greater than 60 kJ mol⁻¹. Huynh and Wu⁷⁰ estimated the rotational barrier for syn/anti interconversion in a [trans-(NHC)₂PdBr₂] complex to be 74 kJ mol⁻¹, from ¹H NMR experiments in which the coalescence occurred at high temperature (380 K). None of these studies appear to have taken into account of the population difference between anti and syn conformations in their analyses.

Computational conformational analysis of 13 was also used to explore the experimental observation of two species in a ratio 62 : 38 in the NMR spectrum of 13 for anti : syn isomers, respectively. This suggests a significant preference for the anti-conformation under the studied conditions. Computational calculations support this finding, indicating that the anti isomer is more stable by 2.0 kcal mol⁻¹ compared to the syn form (see Fig. 3). While this energy difference is not exceptionally large, it is sufficient to largely suppress the population of the syn isomer at equilibrium. The reported ΔE was used only to qualitatively support the observation that the anti isomer is energetically favored, rather than to quantitatively reproduce the NMR-derived syn : anti ratio. According to Boltzmann distribution principles, even differences as small as 1–2 kcal mol⁻¹ can result in a dramatic shift in population at room temperature.⁷¹ This has been demonstrated in several experimental and computational studies. For example, in a study on azomethine derivatives, the anti isomer was favoured by only ∼1.8 kcal mol⁻¹.⁷² Similarly, in a study on losartan isomers, the anti conformation was more stable and dominant in NMR spectra.⁷³ These findings underscore that even moderate energy differences can dictate the dominant species in solution. Therefore, while 2.0 kcal mol⁻¹ might seem modest, it is more than sufficient to explain why the anti conformer is the dominant species form the ratio in the NMR spectrum of 13.

Like the Au^III complexes 11 and 12, the Au^III complex 13 was unstable in DMSO. For example, in DMSO-d₆ solution, ∼50% of the Au^III complex 13 had been reduced to the corresponding Au^I complex 9 after 4 days, and little 13 remained after 25 days (Fig. S9†).

Halide abstraction reactions of Au^III complexes of Tz-NHC ligands: the quest for chelate Au^III–NHC complexes

Au^III–NHC complexes in which the Au centre is chelated by a functionalised NHC ligand are interesting for their potential use as catalysts.⁷⁴ Chelation should stabilise the Au^III centre, allowing such complexes to participate in reactions as catalysts without the Au^III centre reverting to Au^I. In order to chelate the Au^III centre in complexes such as 12 and 13 to the N atoms of triazolyl rings, experiments to abstract chloride from the Au^III centres were explored. The goal was to facilitate displacement of the two chloride ligands of 12 and 13 with intramolecular coordination of the pendant triazolyl groups to from the proposed chelate complexes 15 and 16 (Scheme S3†).

Reaction of the (NHC)AuCl₃ complex 12 with AgNO₃

When complex 12 was treated with AgNO₃ in DMSO-d₆ solution in an NMR tube, a white precipitate (presumably AgCl) formed immediately. A new Au^III–NHC complex was formed and was assigned the structure 15, in which the metal centre is chelated by the NHC unit and the two triazolyl units in the organic ligand (Scheme 7). When only 2.2 equivalents of AgNO₃ were used, 15 was formed as a mixture with the Au^I–NHC complex 8 in a ratio of 19 : 1, respectively. The two complexes were separated during workup based on their solubilities (15 was insoluble in CH₂Cl₂ so 8 was easily removed by a simple washing step with CH₂Cl₂) and pure 15 was obtained as a white powder in 56% yield. Interestingly, when 3.3 equivalents of AgNO₃ was used (again in DMSO-d₆ in an NMR tube), 15 was the only complex formed. It may be that with excess Ag⁺, abstraction of chlorido ligands from 12 occurs rapidly, so that the Au^III centre is quickly stabilised by chelation by the triazolyl moieties, whereas when less Ag⁺ is used, the chelation occurs more slowly, allowing some time for 12 to be reduced to 8 as discussed above. Furthermore, the fact that the same product (15) was formed when 12 was treated with 2.2 or 3.3 equivalents of AgNO₃ indicates that Ag⁺ is only able to abstract two chlorido ligands from 12. Evidently 15 is resistant to abstraction of the final chlorido ligand from the Au centre. It may be that the result of such an abstraction, an Au^III centre chelated by an NHC and two triazolyl units, with the remaining coordination site either vacant or occupied by nitrate or DMSO, is unstable and so cannot be formed, or it may be that chelation provides kinetic stability to 15.

Scheme 7 Abstraction of chlorido ligands from 12 by Ag⁺, to form 15.

It is interesting to compare these results with those of Hirtenlehner et al.,⁵⁸ who treated complexes of form (NHC)AuBr₃ and [(NHC)₂AuBr₂]⁺ with Ag⁺ in the presence of various anionic ligands L⁻. In that study, the NHC ligand was 1,3-dibenzylimidazol-2-ylidene, which has no ability to chelate the Au^III centre. In every case, the starting Au^III–NHC complexes decomposed to form complicated mixtures containing Au^I species such as (NHC)AuL and [(NHC)₂Au]⁺, in addition to Au(0) and imidazolium salts as the major products.

Unfortunately, attempts to obtain single crystals of 15 were not successful, so structure determination was based on the results of elemental analysis and NMR studies. Elemental analysis indicated an empirical formula C₂₃H₂₀AuBr₂ClN₁₀O₆·(CH₂Cl₂)_1.2, which is consistent with the structure proposed for 15 with some entrapped CH₂Cl₂ (used during workup). The ¹H NMR spectrum (Fig. 2) was also consistent with the structure 15 in terms of number of signals seen, their chemical shifts, and their relative integrals. Key signals are a 2H singlet at 9.03 ppm (assigned to the triazolyl protons), a 2H singlet at 7.99 ppm (assigned to the imidazolyl H4/H5 protons), and two 4H singlets at 6.1 and 5.8 ppm, assigned to the methylene groups adjacent to the bromophenyl and imidazolyl groups, respectively. Assignments of all ¹H NMR signals for 15 were carried out with the aid of ¹H–¹³C HSQC and HMBC experiments. The number of signals (one signal for the H4/H5 protons, one for the two triazolyl protons, two signals for the four methylene groups, and one set of signals for the two bromophenyl groups) is consistent with a structure that has a plane of symmetry, as in the chelate complex 15 where both triazolyl groups are bound directly to the Au centre. The possibility of one or both triazolyl groups being bound as an abnormal carbene (C–Au bond) can be discounted by the presence of the 2H signal for the triazolyl protons.

Fig. 2 ¹H NMR spectra (600 MHz, DMSO-d₆) of chelate complex 15 (top spectrum), and Au^IIICl₃ complex 12 (bottom spectrum).

Comparison of the ¹H NMR spectra of 12 and 15 (both recorded in DMSO-d₆) is informative. The most significant difference between the two spectra is the chemical shift of the signal for the triazolyl protons (H8 in Fig. 2), which is shifted from ∼8 ppm in 12 to ∼9 ppm for 15, consistent with the triazolyl N donating electron density to Au centre in 15, leaving less electron density in the ring to shield the triazolyl proton. Both methylene signals are also shifted downfield in the spectrum of 15, again consistent with coordination of the triazolyl groups to Au, resulting in electrons being withdrawn from the triazolyl ring causing deshielding of the methylene groups relative to the situation in 12.

In the ¹³C NMR spectrum of 15, the key signals are those for the Au-bound carbon and the triazolyl carbon. In DMSO-d₆ solution the imidazolyl C2 carbon appears at 138.0 ppm and the triazolyl carbons (C8) appear at 128.1 ppm. These chemical shifts are quite close to those of the corresponding signals for 12, for which C2 occurs at 140.9 ppm and C8 occurs at 125.1 ppm, but the comparison is problematic because the instability of 12 in DMSO made it necessary to record the ¹³C NMR spectrum in CD₂Cl₂. Assignments of all ¹³C and ¹H NMR signals for 15 were aided by ¹H–¹³C HSQC and HMBC experiments.

To explore the stability of complex 15, an NMR sample of 15 prepared by adding 3.3 equivalents of AgNO₃ to 12 in DMSO-d₆ was monitored for change over the course of one week (Fig. S11†). ¹H NMR spectra showed new signals emerging that did not match the chemical shifts of Au^I complex 8, although new signals associated with protons of the bromophenyl moiety did appear in the region where the corresponding signals for 8 occurred. These observations suggest that decomposition might involve dissociation of one (or both) triazolyl groups from the Au centre, but not simply via some reduction process to form the Au^I complex 8. Excess Ag⁺ was present in the sample and may have been involved in additional reactions with 15. Nevertheless, 15 was quite stable in the solution, and much more stable than its non-chelate counterpart 12. Integration of the signal for the triazolyl proton (using the DMSO-d₆ solvent signal as an internal standard) showed that 86% of 15 remained after 22 h, and 55% remained after 7 days.

Furthermore, the origin of the pronounced stability observed for the closed form of the compound 15 was investigated, density functional theory (DFT) calculations were performed to compare two forms of compound 15: tridentate N,N,C-Tz-NHC Au^III monochloride complex (closed form) versus hypothetical monodentate Tz-NHC Au^III monochloride complex (open form). In the proposed open structure, the compound adopts a linear coordination geometry, while the experimentally observed compound 15 exhibits a square planar coordination involving intramolecular interactions with nitrogen atoms from the adjacent five-membered rings. The calculations in Fig. 3 indicate that the closed form is energetically favored by 70.4 kcal mol⁻¹ relative to the open form. This substantial energy difference not only underscores the thermodynamic preference for the closed conformation but also provides a compelling justification for why the open form has not been experimentally isolated. The strong stabilization arising from the Au–N chelation likely plays a pivotal role in stabilizing the closed geometry of compound 15.

Fig. 3 The calculated relative stability of the complexes 13, 15 and the attempted 16: the atomic colors: gray; carbon, blue; nitrogen, green; chlorine, brown; bromide, yellow; gold, white; hydrogen.

Reaction of the [(NHC)₂AuCl₂]⁺ complex 13 with AgNO₃

In preliminary attempts to abstract chlorido ligands from 13 to form chelate complexes, KPF₆ was used as the chloride abstracting agent. However, several attempts to abstract chloride by treating 13 with KPF₆ in DMSO-d₆, at room temperature and at 50 °C, were unsuccessful. No evidence of chloride abstraction was seen and no precipitate was formed, indicating that the salt KCl (which would form if abstraction occurred) was not present in the mixtures. The only change seen in the samples was partial reduction of 13 to the Au^I complex 9.

After these unsuccessful experiments to abstract the chlorido ligands by KPF₆, AgNO₃ was tested as the abstracting agent. When 3.4 equivalents of AgNO₃ was added to a solution of 13 in DMSO-d₆ in an NMR tube, a white precipitate (presumably AgCl) formed immediately. The ¹H NMR spectrum of the sample showed that none of 13 remained in the sample—both the syn and anti conformations of 13 had been consumed, and a complex mixture of products was formed (Fig. S12†).

Sadly, there were no ¹H NMR signals seen that could be confidently attributed to the desired product, the chelate complex 16 (Scheme S3†). Nevertheless, the ¹H NMR spectrum showed signals that are tentatively assigned to a chelate ligand in a species such as 17 (Scheme 8). One of the signals of interest here is a singlet at 9.11 ppm, which is tentatively assigned to the triazolyl proton H8 of a chelating triazolyl moiety, based on its strongly downfield chemical shift. The triazolyl H8 protons of the chelate complex 15 had a similar chemical shift. The other signals of interest are a group in the region 5.9–6.1 ppm. This group has the appearance of two doublets, but with small shoulders on either side of the group, suggesting that they may actually be two AB doublets. These signals are tentatively assigned to the C6 and C9 methylene protons, one AB pattern to each group. The integrals of the signal assigned to H8 and the C6 + C9 group are in the ratio 1 : 4 respectively, consistent with the tentative assignment.

Scheme 8 Abstraction of Cl⁻ from Au^III complex 13 to form Au^III complex 17, followed by reduction to form Au^I complex 9.

The AB patterns arise because the two protons in each methylene group are rendered inequivalent by the asymmetry in 17 associated with the orientation of the non-chelating NHC ligand with respect to the C–Au–C axis. Interestingly, when the sample was heated (Fig. S13†), the two AB patterns collapsed to two singlets, which is consistent with interconversion of two isomeric conformations of 17 by rotation of the non-chelating NHC ligand about the C–Au–C axis (Scheme 8). Unfortunately, 17 was not stable at elevated temperatures, a significant proportion being decomposed when the sample was heated to 318 K and none remaining after the sample had been heated to 378 K (Fig. S13†), so the behaviour of 17 could not be explored in detail.

The species 17 is interesting as a potential precursor of 16 (Scheme S3†). Unfortunately, the species was formed as part of a complex mixture. The ¹H NMR spectrum of the mixture containing 17 and other products formed by treatment of 13 with AgNO₃ shows an envelope of signals at 5.5–5.9 ppm. These signals fall into the range seen for the methylene signals for anti- and syn-13 (Fig. S6 and S10†), and are tentatively assigned to methylene signals associated with NHC ligands in which the triazolyl groups are not involved in chelation of the Au centre. Comparison of integrals for these signals and the signals assigned to the methylene groups of the chelating NHC ligand in 17 suggests that about 46% of 13 was converted to 17 and the remaining 54% was converted to other unidentified products. The mixture slowly underwent further change over time, to form the Au^I complex 9. As noted above, formation of 9 was evident in the VT NMR study (Fig. S13†), but 9 was also formed in significant quantities in samples left at room temperature. Similarly, Hirtenlehner et al.⁵⁸ reported that attempts to abstract bromido ligands from [(NHC)₂Au^IIIBr₂]⁺Br (NHC = 1,3-dibenzylimidazolin-2-ylidene) by AgNO₃ resulted in partial or complete reduction to Au^I.

A comparative computational analysis of compound 13 with the attempted compound 16 is illustrated in Fig. 3. Although compound 16 has not been experimentally observed, the reaction energy calculations indicate a striking difference of over 500 kcal mol⁻¹ in favour of compound 13. This substantial stability suggests that compound 13 occupies a deep thermodynamic minimum, likely due to favourable electronic and structural factors that are absent in the structure of 16.

X-ray diffraction studies

The structures of the imidazolium salt 6·2H₂O, its (NHC)Au^ICl derivative 8·0.5Et₂O, its (NHC)Au^IIIClI₂ derivative 11 (see ESI†) and its (NHC)Au^IIICl₃ derivative 12 were characterised by X-ray crystallographic studies. Crystals suitable for X-ray analysis were grown by diffusion of vapours between pentane and a dichloromethane solution of the compound (for 6, 11, and 12) or diethyl ether and a solution of the compound in acetone (for 8).

In the solid-state structure of 6·2H₂O, the cation was associated with one chloride counterion and two molecules of water (see Fig. S14,† which includes the atom numbering). The chloride ion shows weak interactions with an imidazolyl hydrogen (Cl(1)⋯H(3) 2.769 Å) and the triazolyl hydrogen (Cl(1)⋯H(25) 2.764 Å), smaller than the sum of the van der Waals radii as calculated by Bondi (2.95 Å)⁷⁵ and Alvarez (3.02 Å).⁷⁶ One water is involved in hydrogen bonding with an imidazolyl hydrogen (O(2)⋯H(4) 2.159 Å) and the other water is involved in hydrogen bonding with the triazolyl hydrogen (O(1)⋯H(55) 2.512 Å), both O⋯H distances being smaller than sum of the van der Waals radii as calculated by Bondi and Alvarez (2.72 and 2.70 Å respectively).^75,76 Interestingly, there appear to be no hydrogen bonding interactions involving the imidazolium N–CH–N hydrogen. The cation is in a trans-extended zig-zag conformation, with the three methylene carbons approximately coplanar, and the five rings being oriented approximately perpendicular to that plane (Fig. S15†). The cations are stacked into columns, perhaps stabilised by π-stacking interactions; the inter-centroid distance between opposing rings in adjacent cations is 4.467 Å (Fig. 4). Such arrangements can indicate weak parallel π–π interactions.⁷⁷

Fig. 4 Two adjacent cations in 6·2H₂O, showing the inter-centroid spacings (4.467 Å) between opposing rings.

The structure of 8·0.5Et₂O contains two independent molecules of 8, one with disorder associated with one of its phenyl rings (Fig. 5). The Au–C distances are (Au(1)–C(11) 2.04(2), Au(2)–C(21) 2.03(2) Å), and Au–Cl distances are (Au(1)–Cl(1) 2.295(5), Au(2)–Cl(2) 2.294(5) Å). Coordination about Au is approximately linear (C(11)–Au(1)–Cl(1) 179.0(6)°, C(21)–Au(2)–Cl(2) 178.8(7)°). These distances and angles are in the ranges typically seen for (NHC)Au^ICl complexes.^64,78–80 The N-substituents on the NHC moiety are oriented anti to one another, with their N–C bonds in planes approximately orthogonal to the imidazolyl plane (Fig. 6).

Fig. 5 Crystal structure of the 8·0.5 Et₂O (50% probability level for the thermal displacement of ellipsoids). Selected bond distances (Å) and angles (°). Molecule 1: Au(1)–C(11), 2.04(2); Au(1)–Cl(1), 2.295(5); C(11)–Au(1)–Cl(1), 179.0(6); molecule 2: Au(2)–C(21), 2.03(2); Au(2)–Cl(2), 2.294(5); C(21)–Au(2)–Cl(2), 178.8(7).

Fig. 6 One of the (NHC)AuCl molecules in 8·0.5Et₂O, viewed close to the NHC plane, showing the pendant triazolyl groups with their triazolyl planes approximately orthogonal to the imidazolyl plane.

In the structures of both 11 and 12 there are two independent molecules and the coordination modes for Au in the (NHC)Au^IIIX₃ seen in the structures of 11 and 12 are similar (see ESI, Fig. S16–S18†). The Au–C_carbene and Au–X bond distances (see figure captions and Table S4†) are in the ranges seen in related compounds.^{56,58,81–84} In 12, the chlorido ligands trans to the NHC group has a longer Au–Cl bond (Au–Cl_trans 2.3078, 2.3215 Å) than for the chlorido ligands cis to the NHC group (Au–Cl_cis 2.2851, 2.2828, 2.2835, 2.2806 Å),⁸³ as expected due to the strong trans effect of the NHC ligand. In 12, the Au is in a square planar coordination motif, with angles C_carbene–Au–Cl_trans ∼ 176° and C_carbene–Au–Cl_cis in the range 87–89°. The Cl_cis–Au–Cl_cis angles are in the range 175.7–176.9°, with the chloride atoms pointing towards the carbene carbon. It has been suggested that this arrangement may be a consequence of an interaction between a chloride lone pair and the empty p-orbital on the carbene carbon.^81,85 Consistent with this suggestion, structures show C_carbene⋯Cl contacts (C⋯Cl_cis = 2.975–3.003 Å) that are substantially less than the sum of the van der Waals radii as calculated by Bondi (C⋯Cl 3.45 Å)⁷⁵ and Alvarez (C⋯Cl 3.59 Å).⁷⁶ The suggestion should be treated with some scepticism though, since the contacts between the Cl atom trans to the carbene C and the adjacent Cl atoms (Cl_trans⋯Cl_cis = 3.279–3.318 Å) are also less than the sum of the van der Waals radii as calculated by Bondi (Cl⋯Cl 3.50 Å)⁷⁵ and Alvarez (Cl⋯Cl 3.64 Å).⁷⁶

In 12, the triazolyl pendants are mutually anti about the imidazolin-2-ylidene moiety, creating a groove within which the Cl_cis–Au–Cl_cis moiety is located. The plane of the imidazolin-2-ylidene ring is inclined at an angle to the CAuCl₃ plane (θ in Fig. 7). The angle of inclination for the (NHC)AuCl₃ complex 12 (∼69°).

Fig. 7 View of the structure of the cation in 12 along the Cl_transAu–C_carbene axis. The plane defined by the imidazolyl ring in each case is horizontal and perpendicular to the page with the Cl_trans ligand in front of the page. The coordination plane for Au (defined by the Au centre and the Cl_trans, C_carbene) is also perpendicular to the page and inclined at an angle θ with respect to the plane of the imidazolyl ring. In 12, θ = 68.68° (molecule 1) and 68.87° (molecule 2).

Activity of triazolyl-functionalised Au–NHC complexes against OVCAR-8 cells

Gold-based compounds, especially Au–NHC complexes, are receiving much attention due to their promising anti-cancer activity against a wide variety of cancer cell lines.^26,86,87 The dominant proposed mechanism for the activity of Au–NHCs is that Au inhibits the activity of mitochondrial thioredoxin reductase (TrxR) by binding to the soft Se and S atoms in the cysteine and selenocysteine residues at the active site.^24,26 TrxR is an important enzyme that regulates the cellular redox system and inhibition of TrxR can lead to a build-up of reactive oxygen species, which sets off a cascade of processes that leads to apoptosis (controlled cell death).^26,87,88 In this study, Au–NHC complexes were tested against OVCAR-8 ovarian cancer cells, a cisplatin-resistant cell line. Previous genetic profiling on OVCAR-8 suggested that it resembles high-grade serous ovarian carcinoma in patients (PMID 23839242), which is the most common histological subtype of ovarian cancer. A few classes of Au complexes, including auranofin and [Au(dppe)₂]ClO₄,⁸⁹ and imidazole-based complexes⁹⁰ have been tested in OVCAR-8 cells previously.

The activity of the triazolyl functionalised Au–NHC complexes 7–9 against OVCAR-8 cells is summarised in Table 1 and Fig. 8. The Au^III complexes 12 and 14 were not tested due to their instability in DMSO. The chelate Au^III complex 15 was not tested as it was only synthesised in small amounts, insufficient for biological testing.

Table 1 IC₅₀ values for triazolyl-functionalised Au–NHC complexes against OVCAR-8 cells^a

Compound	IC50 (μM) ± SEM^c	n
a OVCAR-8 cells were exposed to test compounds during incubation for 72 h and activity of the compounds on cell viability was evaluated using an MTT assay. Data are shown as mean ± SEM of at least 3 independent experiments (n) each of which used 2 to 4 replicates per concentration for the relevant test compound.b Imidazolium salt 6 used as a negative control.c IC50 refers to the concentration of the drug that is required to inhibit the growth of the cancerous cell by 50%.
7	7.9 ± 1.3	4
8	13.2 ± 1.7	4
9	6.5 ± 0.1	3
6^b	>100	3

---

Fig. 8 The inhibition activity of the triazolyl-functionalised Au–NHC complexes against OVCAR-8 cells, measured by MTT assay after 72 h of treatment. The imidazolium salt 96 was used as negative control. Data are shown as mean ± SEM of at least 3 independent experiments, each of which used 2 to 4 replicates per concentration for the relevant test compounds.

The triazolyl functionalised Au–NHCs were active against the OVCAR-8 cells but showed higher IC₅₀ values than the (NHC)Au(SCOR) complexes and [(Pr₂Im)₂Au]Br. From the IC₅₀ values, the cationic bis-NHC complex 9 and the least hindered neutral (NHC)AuCl complex 7 are the most potent inhibitors, the more hindered neutral (NHC)AuCl complex being somewhat less active. Several factors likely contribute to activity in this series—the cationic nature of 9 (which would help it to target mitochondria), the hydrophilicity of the triazolyl groups, and different steric bulk around Au. The hydrophilic characteristic of the triazole moiety may decrease inhibitory activity, as Au compounds with greater hydrophobicity are typically more active than comparable hydrophilic ones.²⁴ In 8, the steric bulk of the two bromophenylmethyltriazolyl groups may shield the Au centre from its biological targets, leading to decreased inhibitory activity of 8 compared to 7; this shielding effect has been suggested previously.^23,91 Further work and a larger library of compounds are needed to provide further insights into the relative activities within the triazolyl-functionalised Au–NHC complexes.

Conclusions

The CuAAC click reaction of two propargyl-functionalised imidazolium salts with 2-bromobenzyl azide proceeded smoothly to form the corresponding triazolyl-functionalised imidazolium salts. The triazolyl-functionalised imidazolium salts were in turn converted into stable Au^I complexes of form (NHC)AuCl (e.g., 8) or [(NHC)₂Au]Cl (e.g., 10). The triazole-functionalised (NHC)AuCl complexes could be oxidised by thionyl chloride to yield Au^III complex (NHC)AuCl₃. The complex was relatively stable in CD₂Cl₂ solution, but when dissolved in DMSO-d₆ the (NHC)AuCl₃ complex was reduced to the corresponding (NHC)Au^ICl complex, further studies of Au^III–NHC complexes focussed on chlorides rather than iodides.

Oxidation of [(NHC)₂Au]Cl complexes with thionyl chloride gave the corresponding [(NHC)₂Au^IIICl₂]Cl complexes, for example 13, which was obtained as a mixture of syn and anti isomers. The [(NHC)₂Au^IIICl₂]Cl complexes could be characterised by NMR spectroscopy in fresh DMSO-d₆ solutions, but were slowly reduced to [(NHC)₂Au^I]Cl, and poor solubility issues prevented their study in other solvents. The stability of Au^III–NHC complexes was improved by chelation. Treatment of the (NHC)Au^IIICl₃ complex 12 with AgNO₃ in DMSO-d₆ resulted in abstraction of the chloride ligands and formation of the chelated complex 15, which showed good stability in DMSO-d₆.

The chelate compound 15 is a novel compound and it would be worthwhile to confirm the structure by X-ray diffraction. Further work could also be devoted to achieving the synthesis of the chelate complex 16, which so far remains elusive. Efforts to react the Au^III–NHC complex 13 with AgNO₃ to form the doubly-chelate complex 16 have so far only led to a product tentatively assigned as 17. Longer reaction times, more forcing reaction conditions, or the use of other chloride abstraction agents may permit 16 to be obtained.

Experimental section

General procedures

NMR spectra were recorded using Bruker AV500 III HD (500.13 MHz for ¹H, 125.77 MHz for ¹³C, 202.45 MHz for ³¹P), Bruker AV600 III HD (600.13 MHz for ¹H, 150.90 MHz for ¹³C), or Bruker AV NEO (400.15 MHz for ¹H) spectrometers at ambient temperature, unless otherwise indicated. ¹H and ¹³C NMR chemical shifts were referenced to the residual signal of the solvent (DMSO-d₆: ¹H 2.50 ppm; ¹³C 39.52 ppm; CD₂Cl₂: ¹H 5.32 ppm; ¹³C 53.8 ppm; acetone-d₆: ¹H 2.05 ppm; ¹³C 29.84 ppm). Assignments of signals were confirmed with the aid of ¹H–¹³C HSQC (heteronuclear single quantum coherence) and ¹H–¹³C HMBC (heteronuclear multiple bond correlation) spectra when necessary. Microanalyses were performed by Microanalytical Service, School of Chemistry & Molecular Biosciences, The University of Queensland, Australia. High resolution mass spectra were measured using Agilent LCMS 6510 Q-TOF and Waters LCT Premier XE spectrometers, using the APCI or ESI methods, with 9 : 1 CH₃CN : H₂O as solvent. All the reactions were performed under a nitrogen atmosphere, in oven dried glassware. (Me₂S)AuCl was prepared via the literature method.⁹² Gold complexes were wrapped in aluminium foil to exclude light. Solvents were of analytical or reagent grade and used without further purification.

Computational methodology

All structures were fully optimized using Density Functional Theory (DFT) with the unrestricted hybrid B3LYP functional^93,94 and the all-electron def2-TZVP basis set,⁹⁵ as implemented in the Gaussian 16 software package.⁹⁶ Grimme's D3 dispersion correction with Becke–Johnson damping (D3-BJ)⁹⁷ was applied to account for long-range dispersion interactions. Geometry optimizations were followed by analytical frequency calculations to confirm that the stationary points correspond to true minima on the potential energy surface, as indicated by the absence of imaginary frequencies in the Hessian matrix.

Single X-ray crystal diffraction

Crystallographic data were measured at 100(2) K on an Oxford Diffraction Gemini or an Oxford Diffraction Xcalibur X-ray diffractometer using Mo Kα or Cu Kα radiation. Following analytical absorption corrections and solution by direct methods, the structures were refined against F² with full-matrix least-squares using the SHELX program suite.⁹⁸ Unless stated differently below, all hydrogen atoms were added at calculated positions and refined by use of riding models.

Anti-cancer activity

Cell culture reagents. Rosewell Park Memorial Institute (RPMI-1640) medium (with L-glutamine and sodium bicarbonate), fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Sigma-Aldrich. Cell culture medium consists of RPMI-1640 medium (with L-glutamine and sodium bicarbonate) supplemented with 10% v/v FBS and 1% v/v penicillin–streptomycin. Cells were dissociated during passaging using trypsin solution 0.05% in Hank's balanced salt solution (HBSS), and ethylenediaminetetraacetic acid (EDTA), without calcium or magnesium (Hyclone GE Healthcare Life Sciences). MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was obtained from Thermo Fisher. Phosphate buffered saline (PBS) was obtained from Astral Scientific. Stock solutions (10 mM) of test compounds were prepared in DMSO (cell culture grade), obtained from Bio-Strategy. DMSO (to dissolve the MTT after incubation step in the MTT assay) was obtained from Sigma-Aldrich.

Cell culture. Human ovarian cancer cells (OVCAR-8, from ATCC) were cultured in 75 cm² tissue culture flasks in cell culture medium (10 mL) at 37 °C under air (5% humidity, 5% CO₂). After being incubated for 2 days, the cells had reached ∼80% confluency and were then passaged as follows. The medium was poured off, and the cells were detached by incubation and agitation with trypsin (3 mL) for 3–5 min at 37 °C. To neutralise the trypsin solution, cell culture medium (3 mL) was added, and the cell suspensions were transferred to Falcon tubes and centrifuged for 5 min at 1500 rpm. The supernatant was drawn off by pipette, and the cell pellet was resuspended in culture medium (10 mL). The cells were re-plated into three tissue culture flasks (2 mL cell suspension and 8 mL culture medium per flask) and incubated for two days. The cell culture medium was replaced after 1 day if the cells had shown substantial proliferation and confluency.

Cell viability assays. Stock solutions of Au–NHC compounds (10 mM) were prepared in DMSO and were stored at −20 °C until required. These stock solutions were serially diluted in cell culture medium. For cell viability assays, OVCAR-8 cells that had reached 80% confluency in the tissue culture flasks were used. For cell seeding, the cells were dissociated as described above, the cell number was counted using a hemocytometer, and the cells were suspended in culture medium at a final concentration of 2000 cells/100 μL. Cell viability was determined using the MTT assay following the previously published protocol.⁹⁹ Briefly, OVCAR-8 cells were seeded in polystyrene 96-well plates (Costar; 2000 cells in 100 μL culture medium per well) and incubated for 24 h. Aliquots of solutions of Au–NHC compounds in cell culture medium (100 μL) were added to the wells so that the final concentrations of Au–NHC complex in wells was 0.39, 0.78, 1.56, 3.125, 6.25, 12.5, 25, 50, or 100 μM. The cells with the Au–NHCs were then incubated for 72 h. The medium was then carefully and completely removed by multichannel pipette and replaced by fresh cell culture medium (90 μL per well) and MTT solution (5 mg per mL MTT dissolved in PBS, 10 μL), and the plates were agitated to ensure homogeneity of solutions. The plates were incubated at 37 °C for 2 h, then the solution was removed by multichannel pipette and replaced by DMSO (200 μL per well) to solubilise the formazan purple colour from the live cells. The absorbance at 570 nm (A₅₇₀) was measured using a plate reader (Ensight Multimode Plate Reader, PerkinElmer). Relative cell viability (%) was determined by comparison of A₅₇₀ for wells that had contained cells treated with Au–NHC complexes with A₅₇₀ for wells that contained cells in culture medium alone. GraphPad Prism 8 software was used to determine IC₅₀ values from cell viability data and to plot graphs of cell viability as a function of concentration of the Au–NHC complexes. Imidazolium salts (precursors of NHC ligands) were used as negative controls for the cytotoxicity of the compounds to OVCAR-8 cells.

Synthesis

Triazolyl-functionalised imidazolium salt (5). 2-Bromobenzyl azide (0.27 g, 1.27 mmol), CuSO₄ (27.0 mg, 0.169 mmol) and copper powder (4.7 mg, 0.075 mmol) were added to a solution of 1-methyl-3-propynylimidazolium chloride (200 mg, 1.27 mmol) in 1 : 1 H₂O/tert-butanol (5 mL each). The mixture was heated at reflux for 4 h, then cooled to room temperature and the solvents were evaporated to dryness. The residue was dissolved in 1 : 3 EtOH/MeOH (200 mL) and filtered through a plug of silica. The filtrate was evaporated to dryness and to the oily residue was added EtOH. The resulting precipitate was filtered off and the green filtrate was evaporated to give a yellow/brown oil (210 mg, 45%) which was dried under a flow of N₂. ¹H NMR (500 MHz, DMSO-d₆): δ 9.24 (br s, 1H, H2), 8.29 (s, 1H, H8), 7.76 (1H, apparent t, splitting 1.7 Hz, H4), 7.71–7.69 (m, 2H, H5 and H14), 7.42 (1H, apparent td, “triplet” splitting 7.5 Hz, ⁴J_H,H = 1.2 Hz, H12), 7.33 (1H, apparent td, “triplet” splitting 7.7 Hz, ⁴J_H,H = 1.8 Hz, H13), 7.24 (1H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.7 Hz, H11), 5.71 (s, 2H, H9), 5.55 (s, 2H, H6), 3.80 (s, 3H, ). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 140.6 (C7), 136.7 (C2), 134.5 (C10), 133.0 (C14), 130.7 (C11), 130.6 (C13), 128.3 (C12); 125.0 (C8), 123.8 (C5), 123.0 (C15), 122.4 (C4), 53.3 (C9), 43.5 (C6), 35.8 (C̲H₃). Microanalysis: found: C, 39.74; H, 4.82; N, 16.02%. C₁₄H₁₅BrClN₅ (H₂O)_3.2 requires C, 39.44; H, 5.06; N, 16.43. HRMS (ESI⁺): calcd for C₁₄H₁₅BrN₅ [M − Cl⁺], m/z 332.0511. Found, m/z 332.0507.

Triazolyl-functionalised imidazolium salt (6). 2-Bromobenzyl azide (0.318 g, 1.50 mmol) was added to a solution of 1,3-dipropynylimidazolium chloride (135 mg. 0.75 mmol) in 1 : 1 H₂O/tert-butanol (4 mL each), followed by addition of CuSO₄ (16.0 mg, 0.10 mmol) and copper powder (2.70 mg, 0.044 mmol). The mixture was heated at reflux for 4 h, then cooled to room temperature and the solvent evaporated under vacuum. The residue was dissolved in DCM (40 mL) and filtered; the filtrate then passed through plug of silica eluting with 4 : 1 DCM/MeOH. The solvents were evaporated and the sticky brown residue that was obtained was washed with diethyl ether (3 × 30 mL) and precipitated from DCM/hexane with cooling (ice bath) to leave a tan powder (350 mg, 77%). ¹H NMR (500 MHz, DMSO-d₆): δ 9.36 (1H, br t, ⁴J_H,H = 1.6 Hz, H2), 8.27 (s, 2H, H8), 7.77 (2H, d, ⁴J_H,H = 1.6 Hz, H4/H5), 7.70 (2H, dd, ³J_H,H = 7.9 Hz, ⁴J_H,_H = 1.3 Hz, H14), 7.41 (2H, apparent td, “triplet” splitting 7.5 Hz, ⁴J_H,H = 1.3 Hz, H12), 7.33 (2H, apparent td, “triplet” splitting 7.7 Hz, ⁴J_H,H = 1.7 Hz, H13), 7.22 (2H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.8 Hz, H11), 5.71 (s, 2H, H9), 5.55 (s, 2H, H6). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 140.4 (C7), 136.3 (C2), 134.5 (C10), 132.9 (C14), 130.7 (C11), 130.6 (C13), 128.3 (C12); 125.1 (C8), 123.0 (C15), 122.8 (C4/C5), 53.1 (C9), 43.6 (C6). Microanalysis: found: C, 42.61; H, 3.85; N, 17.07% C₂₁H₂₃Br₂ClN₈·(H₂O)_2.7 requires C, 42.28; H, 4.07; N, 17.15%. HRMS (ESI⁺): calcd for C₂₃H₂₁Br₂ClN₈ [M − Cl⁺], m/z 567.0256. Found, m/z 567.0277.

Crystals suitable for X-ray diffraction studies were grown by diffusion of vapours between diethyl ether and a solution of the compound in acetonitrile.

The triazolyl-functionalised (NHC)AuCl complex 7. A mixture of Ag₂O (80.0 mg, 0.34 mmol) and the triazolyl-functionalised imidazolium salt 5 (160 mg, 0.43 mmol) in DCM (20 mL) was stirred at room temperature in the absence of light for two days. (Me₂S)AuCl (100 mg, 0.34 mmol) was added and the mixture was stirred overnight then filtered through a plug of silica. The filtrate was evaporated to dryness. The residue was triturated with diethyl ether (3 × 10 mL), recrystallised from acetonitrile/diethyl ether, then recrystallised again from EtOAc at 0 °C and dried under vacuum to afford the product as an off white powder (60 mg, 31%). ¹H NMR (500 MHz, DMSO-d₆): δ 8.18 (s, 1H, H8), 7.68 (1H, dd, ³J_H,H = 7.8, ⁴J_H,H = 1.2 Hz, H14), 7.48 (1H, d, ³J_H,H = 1.8 Hz, H5), 7.43 (1H, d, ³J_H,H = 1.8 Hz, H4), 7.40 (1H, apparent td, “triplet” splitting 7.5 Hz, ⁴J_H,H = 1.2 Hz, H12), 7.31 (1H, apparent td, “triplet” spitting 7.8 Hz, ⁴J_H,H = 1.7 Hz, H13), 7.16 (1H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.7 Hz, H11), 5.67 (s, 2H, H9), 5.42 (s, 2H, H6), 3.74 (s, 3H, ). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 169.0 (C2), 142.4 (C7), 134.7 (C10), 132.8 (C14), 130.4 (C11), 130.4 (C13), 128.3 (C12), 124.5 (C8), 123.0 (C4), 122.8 (C15), 121.6 (C5), 52.9 (C9), 45.4 (C6), 37.7 (C̲H₃). Microanalysis: found: C, 30.32; H, 2.72; N, 11.89% C₁₄H₁₄AuBrClN₅ (CH₃CH₂OCOCH₃)_0.2 requires C, 30.53; H, 2.70; N, 12.03%.

The triazolyl-functionalised [(NHC)₂Au]Cl complex 9. K₂CO₃ (316 mg, 2.28 mmol) and the triazolyl-functionalised imidazolium salt 5 (84.4 mg, 0.228 mmol) in acetonitrile (20 mL) was stirred at room temperature for 15 min, then (Me₂S)AuCl (29.5 mg, 0.10 mmol) was added. The mixture was stirred 3 days, and examination of an aliquot by ¹H NMR spectroscopy showed that the reaction was complete. The mixture was filtered through Celite and the filtrate was evaporated to dryness. The residue was washed with diethyl ether, then crystallised from acetonitrile/diethyl ether to afford an off-white powder (77.5 mg, 86%). ¹H NMR (500 MHz, DMSO-d₆): δ 8.22 (s, 2H, H8), 7.65 (2H, dd, ³J_H,H = 7.8, ⁴J_H,H = 1.2 Hz, H14), 7.60 (2H, d, ³J_H,H = 1.8 Hz, H5), 7.49 (2H, d, ³J_H,H = 1.8 Hz, H4), 7.34–7.27 (m, 4H, H12 and H13), 7.15 (2H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.8 Hz, H11), 5.64 (s, 4H, H9), 5.51 (s, 4H, H6), 3.82 (s, 6H, ). ¹³C NMR (125.70 MHz, DMSO-d₆): δ 183.1 (C2), 142.8 (C7), 134.5 (C10), 132.9 (C14), 130.5 (C11), 130.4 (C13), 128.2 (C12), 124.3 (C8), 123.4 (C4), 122.9 (C15), 122.2 (C5), 53.0 (C9), 45.1 (C6), 37.5 (C̲H₃). Microanalysis: found: C, 35.48; H, 3.32; N, 14.66% C₂₈H₂₈ AuBr₂ClN₁₀·(H₂O)₃ requires C, 35.37; H, 3.60; N, 14.73%. HRMS (ESI⁺): calcd for C₂₈H₂₈Br₂N₁₀ Au [M − Cl⁺], m/z 589.0531. Found, m/z 589.0594.

The triazolyl-functionalised [(NHC)₂Au^IIICl₂]Cl complex 13. Thionyl chloride (1.5 mL, 20.7 mmol) was added to a solution of the triazolyl-functionalised [(NHC)₂Au]Cl complex 9 (44.0 mg, 0.049 mmol) in DCM (15 mL), and the mixture heated at reflux for 4 days under N₂. The solvent was removed under vacuum and the residue was triturated with diethyl ether (3 × 5 mL) to leave the product as a pale yellow powder (42.5 mg, 90%). ¹H NMR (600 MHz, DMSO-d₆) two isomers: δ 8.22 (s, 2H, H8, anti), 8.25 (s, 1.37H, H8 syn), 7.86 (s, 2H, H4 anti), 7.80 (s, 1.35H, H4 syn), 7.71–7.68 (br m, 7H, H5 and H14, anti and syn), 7.35–7.29 (br m, 7H, H12 and H13, anti and syn), 7.14–7.09 (br m, 4H, H11 anti and syn), 5.80 (s, 2.8H, H6, syn), 5.70 (s, 4H, H9, anti), 5.67 (s, 2.8H, H9, syn), 5.64 (s, 4H, H6, anti), 4.07 (s, 6H. anti), 3.97 (s, 4H, ). ¹³C NMR (150.90 MHz, DMSO-d₆) Two isomers: δ 151.84 (C2, syn), 151.76 (C2, anti), 141.71 (C7, anti), 141.5 (C7, syn), 134.5 (C10), 132.9 (C14), 130.5 (C13), 130.4 (C11), 128.1 (C12), 125.4 (C5, syn), 125.1 (C5, anti), 125.0 (C8, syn), 124.8 (C8, anti), 124.5 (C4, anti), 124.1 (C4, syn); 122.8 (C15), 53.1 (C9, anti), 44.5 (C6, anti), 44.3 (C6, syn), 37.5 (C̲H₃, anti), 37.3 (C̲H₃, syn).

Microanalysis: found: C, 33.88; H, 2.90; N, 13.78% C₂₈H₂₈AuBr₂C₁₃N₁₀. (CH₂Cl₂)_0.5 requires C, 33.89; H, 2.89; N, 13.87%.

The triazolyl-functionalised (NHC)AuCl complex 8. A mixture of Ag₂O (28.0 mg, 0.12 mmol) and the bis(triazolyl)imidazolium salt 6 (100 mg, 0.165 mmol) in DCM (10 mL) was stirred at room temperature in the absence of light for 33 h. (Me₂S)AuCl (40.0 mg, 0.136 mmol) was added and the mixture was stirred overnight. The mixture was eluted through a plug of silica using 1 : 4 MeCN/DCM (500 mL). The filtrate was evaporated, the product precipitated by dissolving it in ∼2 mL DCM and excess of hexanes. The product was collected and dried under vacuum to afford a white powder (105 mg, 96%). ¹H NMR (500 MHz, DMSO-d₆): δ 8.19 (s, 2H, H8), 7.68 (2H, dd, ³J_H,H = 7.8, ⁴J_H,H = 1.2 Hz, H14), 7.49 (s, 4H, H4/H5), 7.39 (2H, apparent td, “triplet” splitting 7.5 Hz, ⁴J_H,H = 1.2 Hz, H12), 7.30 (2H, apparent td, “triplet” splitting ³J_H,H = 7.8 Hz, ⁴J_H,H = 1.7 Hz, H13), 7.16 (2H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.7 Hz, H11), 5.67 (s, 4H, H9), 5.42 (s, 4H, H6). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 169.1 (C2), 142.2 (C7), 134.7 (C10), 132.8 (C14), 128.2 (C11), 130.4 (C13), 128.4 (C12), 124.5 (C8), 122.8 (C15), 121.9 (C4/C5), 53.0 (C9), 45.6 (C6). Microanalysis: found: C, 34.42, H, 2.32, N, 13.60% C₂₃H₂₀AuBr₂ClN₈ requires C, 34.50, H, 2.52, N, 13.99%. Crystals suitable for X-ray diffraction studies were grown by diffusion of vapours between pentane and a solution of the complex in dichloromethane.

Also, the work up can be by filtration through Celite using dichloromethane as a solvent to afford a white powder 85%.

The triazolyl-functionalised [(NHC)₂Au]Cl (10). A mixture of K₂CO₃ (106 mg, 0.76 mmol) and the bis(triazolyl)-functionalised imidazolium salt 6 (52.6 mg. 0.087 mmol) in acetonitrile (20 mL) was stirred at room temperature for 15 min. (Me₂S)AuCl (11.0 mg, 0.037 mmol) was added, and the mixture was stirred for two days. The solvent was removed under vacuum, the residue was dissolved in EtOH, and the mixture was filtered through Celite using EtOH. Removal of solvent and trituration of the residue with diethyl ether afforded the product as an off white powder (37 mg, 72%). ¹H NMR (500 MHz, DMSO-d₆): δ 8.24 (s, 4H, H8), 7.63 (4H, dd, ³J_H,H = 7.6, ⁴J_H,H = 1.4 Hz, H14), 7.60 (s, 4H, H4/H5), 7.29–7.28 (m, 8H, H12 and H13), 7.11 (4H, dd, ³J_H,H = 7.5, ⁴J_H,H = 2.1, H11), 5.61 (s, 8H, H9), 5.54 (s, 8H, H6). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 182.8 (C2), 142.5 (C7), 134.5 (C10), 132.9 (C14), 130.5 (C11), 130.4 (C13), 128.1 (C12), 124.5 (C8), 122.9 (C15), 122.5 (C4/C5), 53.0 (C9), 45.2. Microanalysis: found: C, 37.48; H, 2.58; N, 14.72% C₄₆H₄₀ AuBr₄ClN₁₆·(H₂O)_0.5 (CH₂Cl₂)₂ requires C, 37.25; H, 2.93; N, 14.48%. HRMS (ESI⁺): calcd for C₄₆H₄₀Br₄N₁₆Au [M − Cl⁺], m/z 1329.0021. Found, m/z 1329.0056.

The triazolyl-functionalised (NHC)Au^IIICl₃ complex 12. The triazolyl-functionalised (NHC)AuCl complex 8 (50.0 mg, 0.062 mmol) dissolved in DCM (10 mL), thionyl chloride (1.00 mL, 13.8 mmol) was added, and the mixture heated at reflux for 2 days under N₂. The solvent was removed under vacuum and the residue was washed with diethyl ether (3 × 5 mL) to leave the product as a yellow powder (53.5 mg, 97%). ¹H NMR (500 MHz, CD₂Cl₂): δ 7.86 (s, 2H, H8), 7.63 (2H, dd, ³J_H,H = 7.8, ⁴J_H,H = 1.2 Hz, H14), 7.42 (s, 2H, H4/H5), 7.34 (2H, apparent td, “triplet” splitting 7.5 Hz, ⁴J_H,H = 1.2 Hz, H12), 7.26 (2H, apparent td, “triplet” splitting 7.8 Hz, ⁴J_H,H = 1.7 Hz, H13), 7.22 (2H, dd, ³J_H,H = 7.5 Hz, ⁴J_H,H = 1.7 Hz, H11), 5.64 (s, 4H, H9), 5.51 (s, 4H, H6). ¹³C NMR (125.7 MHz, CD₂Cl₂): δ 140.8 (C2), 140.5 (C7), 134.0 (C10), 133.6 (C14), 131.1 (C11), 131.0 (C13), 128.6 (C12), 125.1 (C8), 124.5 (C4/C5), 124.0 (C15), 54.5 (C9), 46.3 (C6). Microanalysis: found: C, 31.84; H, 2.30; N, 12.53% C₂₃H₂₀AuBr₂Cl₃N₈ requires C, 31.69; H, 2.31; N, 12.86%. Crystals suitable for X-ray diffraction studies were grown by diffusion of vapours between pentane and a solution of the complex in DCM.

The triazolyl-functionalised [(NHC)₂Au^IIICl₂]Cl complex 14. Method 1: the triazolyl-functionalised (NHC)AuCl complex 10 (9.0 mg, 6.60 μmol) was suspended in DCM (10 mL), thionyl chloride (0.12 mL, 1.64 mmol) was added, and the mixture heated at reflux for 4 days under N₂. The solvent was removed under vacuum, the residue was stirred with diethyl ether, and the diethyl ether was decanted off, to leave the product as a yellow solid (8.3 mg, 82%).

Method 2: the triazolyl-functionalised (NHC)AuCl complex 10 (15.8 mg, 11.5 μmol) was dissolved in thionyl chloride (3 mL) and the mixture heated at reflux for 5 days under N₂. The solvent was removed under vacuum, the residue was stirred with diethyl ether, and the diethyl ether decanted off, to leave the product as a yellow solid (15 mg, 86%). ¹H NMR (600 MHz, DMSO-d₆) NHC–Au^III: δ 8.21 (s, 4H, H8), 7.82 (s, 4H, H4/H5), 7.64 (4H, br d, ³J_H,H = 7.5 Hz, H14), 7.29 (8H, m, H12 and H11), 7.08 (4H, br d, ³J_H,H = 7.3 Hz, H13), 5.80 (s, 8H, H6), 5.64 (s, 8H, H9). ¹³C NMR (150.9 MHz, DMSO-d₆) NHC–Au^III. δ 151.7 (C2), 141.3 (C7), 134.4 (C10), 132.9 (C14), 130.4 (C11), 130.4 (C13), 128.2 (C12), 124.9 (C8), 124.5 (C4/C5), 122.8 (C15), 53.0 (C6), 44.6 (C9). Microanalysis: found: C, 34.07; H, 2.85; N, 12.76% C₄₆H₄₀AuBr₄Cl₃N₁₆. (CH₂Cl₂)_3.5 requires C, 34.23; H, 2.73; N, 12.90%.

Attempted synthesis of triazolyl-functionalised Au^III–NHC complexes directly from KAuCl₄ (formation of 8, 12, 14). A mixture of Ag₂O (10.4 mg, 0.045 mmol) and the bis(triazolyl)imidazolium salt 6 (50 mg. 0.081 mmol) in DCM (10 mL) was stirred at room temperature in the absence of light overnight. KAuCl₄ (30.0 mg, 0.081 mmol) was added and the mixture was stirred for 10 days. The mixture was filtered through Celite. The filtrate was evaporated under vacuum and the residue was crystallised from DCM/diethyl ether to afford a yellow powder (42 mg). The ¹H NMR spectrum of a sample of the powder in DMSO-d₆ showed signals due 8, 12, 14 in a ratio of 6 : 31 : 25 respectively.

The chelate Au^III complex 15.
(i) Synthesis using three equivalents of AgNO₃. A solution of AgNO₃ (3.20 mg, 18.7 μmol) in DMSO-d₆ (0.25 mL) was added to a solution of the Au^III complex 12 (5.00 mg, 5.70 μmol) in DMSO-d₆ (0.25 mL) in an NMR tube. A white precipitate (presumed to be AgCl) formed immediately. The compound remaining in solution was identified as by its ¹H and ¹³C NMR spectrum.
(ii) Synthesis using two equivalents of AgNO₃. A solution of AgNO₃ (3.40 mg, 19.98 μmol) in DMSO-d₆ (0.25 mL) was added to a solution of the Au^III complex 12 (8.00 mg, 9.18 μmol) in DMSO-d₆ (0.25 mL) in an NMR tube. A white precipitate (presumed to be AgCl) formed immediately and the presence of Au^III complex 15 in solution was confirmed by ¹H NMR spectroscopy. The clear solution was separated from the precipitate using a pipette and the solvent was evaporated under a flow of N_2. The solid was swirled with DCM (3 mL), the DCM was drawn off by pipette and the process was repeated. The white powder that remained was dried under vacuum. Yield: 4.8 mg, 56%. ¹H NMR (500 MHz, DMSO-d₆): δ 9.03 (s, 2H, H8), 7.99 (s, 2H, H4/H5), 7.75 (2H, dd, ³J_H,H = 8.0 Hz, ⁴J_H,H = 1.0 Hz, H14), 7.54–7.49 (4H, m, H11 and H12), 7.40 (2H, ddd, ³J_H,H = 8.0 Hz ³J_H,H = 7.1 Hz, ⁴J_H,H = 2.1 Hz, H13), 6.08 (s, 4H, H9), 5.82 (s, 4H, H6). ¹³C NMR (125.7 MHz, DMSO-d₆): δ 137.9 (C2), 133.2 (C14), 132.2 (C7), 132.0 (C10), 131.9 (C11), 131.6 (C13), 128.5 (C12), 128.1 (C8), 123.8 (C15), 123.7 (C4/C5), 56.5 (C9), 44.5 (C6). Microanalysis: found: C, 28.09; H, 2.21; N, 13.98%. C₂₃H₂₀AuBr₂ClN₁₀O₆. (CH₂Cl₂)_1.2 requires C, 28.31; H, 2.20; N, 13.64%.

Data availability

The data supporting this article have been included as part of the ESI.† More raw data are available from the corresponding author on reasonable request.

Author contributions

Hawraa S. Al-Buthabhak: synthesis and characterisation, formal analysis; investigation; methodology; writing – original draft. Karrar Al-Ameed: computational calculations; formal analysis; investigation Yu Yu: methodology. Alexandre Sobolev: data curation; formal analysis; investigation. Stephen Moggach: data curation; formal analysis; investigation. Hani Al-Salami: resources. Vito Ferro: supervision; writing – review and editing. Murray V. Baker: supervision; resources; writing – review and editing.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

The authors acknowledge the facilities and the scientific and technical assistance of the Australian Microscopy & Microanalysis Research Facility at the Centre for Microscopy, Characterisation & Analysis, the University of Western Australia. Hawraa S. Al-Buthabhak thanks the Higher Committee for Education Development in Iraq (HCED) for financial support. Al-Salami is supported by the European Union Horizon 2020 Research Project and Innovation Program under the Marie Skłodowska-Curie Grant Agreement No. 872370, Curtin Faculty ORS-WAHAI Consortium, and the Australian National Health and Medical Research Council (NHMRC, APP9000597). Stephen A. Moggach thanks the Australian Research Council (ARC) for a Future Fellowship (FT200100243). We thank the University of Western Australia and the University of Queensland for a UWA-UQ Partnership Research Collaboration Award (to VF and MVB).

References

1. W. Zhao, V. Ferro and V. M. Baker, Coord. Chem. Rev., 2017, 339, 1–16.
2. M. N. Hopkinson, C. Richter, M. Schedler and F. Glorius, Nature, 2014, 510, 485–496.
3. P. Mathew, A. Neels and M. Albrecht, J. Am. Chem. Soc., 2008, 130, 13534–13535.
4. R. Jothibasu and H. V. Huynh, Chem. Commun., 2010, 46, 2986–2988.
5. J. Turek, I. Panov, P. Svec, Z. Ruzickova and A. Ruzicka, Dalton Trans., 2014, 43, 15465–15474.
6. M. V. Baker, P. J. Barnard, S. J. Berners-Price, S. K. Brayshaw, J. L. Hickey, B. W. Skelton and A. H. White, Dalton Trans., 2006, 3708–3714,  10.1039/b602560a.
7. N. Hamdi, I. Slimani, L. Mansour, F. Alresheedi, N. Gürbüz and I. Özdemir, New J. Chem., 2021, 45, 21248–21262.
8. A. Martínez, J. L. Krinsky, I. Peñafiel, S. Castillón, K. Loponov, A. Lapkin, C. Godard and C. Claver, Catal. Sci. Technol., 2015, 5, 310–319.
9. B. Jacques, D. Hueber, S. Hameury, P. Braunstein, P. Pale, A. Blanc and P. de Frémont, Organometallics, 2014, 33, 2326–2335.
10. M. Bouhrara, E. Jeanneau, L. Veyre, C. Coperet and C. Thieuleux, Dalton Trans., 2011, 40, 2995–2999.
11. A. S. Hashmi, T. Lauterbach, P. Nösel, M. H. Vilhelmsen, M. Rudolph and F. Rominger, Chem.–Eur. J., 2013, 19, 1058–1065.
12. A. S. K. Hashmi, A. M. Schuster and F. Rominger, Angew. Chem., Int. Ed., 2009, 48, 8247–8249.
13. A. G. Nair, R. T. McBurney, M. R. D. Gatus, S. C. Binding and B. A. Messerle, Inorg. Chem., 2017, 56, 12067–12075.
14. C. Hemmert and H. Gornitzka, Dalton Trans., 2016, 45, 440–447.
15. P. J. Barnard, L. E. Wedlock, M. V. Baker, S. J. Berners-Price, D. A. Joyce, B. W. Skelton and J. H. Steer, Angew. Chem., Int. Ed., 2006, 45, 5966–5970.
16. Ö. Karaca, S. M. Meier-Menches, A. Casini and F. E. Kühn, Chem. Commun., 2017, 53, 8249–8260.
17. T. Zou, C. T. Lum, S. S. Chui and C. M. Che, Angew. Chem., Int. Ed., 2013, 52, 2930–2933.
18. G. Gasser, I. Ott and N. Metzler-Nolte, J. Med. Chem., 2011, 54, 3–25.
19. L. E. Wedlock and S. J. Berners-Price, Aust. J. Chem., 2011, 64, 692–704.
20. J. J. Soldevila-Barreda and N. Metzler-Nolte, Chem. Rev., 2019, 119, 829–869.
21. S. Y. Hussaini, R. A. Haque and M. R. Razali, J. Organomet. Chem., 2019, 882, 96–111.
22. H. S. Al-Buthabhak, Y. Yu, A. Sobolev, H. Al-Salami and M. V. Baker, J. Inclusion Phenom. Macrocyclic Chem., 2021, 101, 227–242.
23. T. Zou, C. N. Lok, P. K. Wan, Z. F. Zhang, S. K. Fung and C. M. Che, Curr. Opin. Chem. Biol., 2018, 43, 30–36.
24. J. L. Hickey, R. A. Ruhayel, P. J. Barnard, M. V. Baker, S. J. Berners-Price and A. Filipovska, J. Am. Chem. Soc., 2008, 130, 12570–12571.
25. G. Achar, S. A. Patil, J. G. Małecki and S. Budagumpi, New J. Chem., 2019, 43, 1216–1229.
26. S. J. Berners-Price and A. Filipovska, Metallomics, 2011, 3, 863–873.
27. G. Moreno-Alcántar, P. Picchetti and A. Casini, Angew. Chem., Int. Ed., 2023, 62, e202218000.
28. B. S. Murray and P. J. Dyson, Curr. Opin. Chem. Biol., 2020, 56, 28–34.
29. G. Augello, A. Azzolina, F. Rossi, F. Prencipe, G. F. Mangiatordi, M. Saviano, L. Ronga, M. Cervello and D. Tesauro, Pharmaceutics, 2023, 15, 466.
30. H. S. Al-Buthabhak, V. Falasca, Y. Yu, A. N. Sobolev, B. W. Skelton, S. A. Moggach, V. Ferro, H. Al-Salami and M. V. Baker, Appl. Organomet. Chem., 2022, 38, e6645.
31. M. G. Gardiner and C. C. Ho, Coord. Chem. Rev., 2018, 375, 373–388.
32. M. A. W. Lawrence, K.-A. Green, P. N. Nelson and S. C. Lorraine, Polyhedron, 2018, 143, 11–27.
33. M. Navarro, A. Tabey, G. Szalóki, S. Mallet-Ladeira and D. Bourissou, Organometallics, 2021, 40, 1571–1576.
34. B. Schulze and U. S. Schubert, Chem. Soc. Rev., 2014, 43, 2522–2571.
35. K. F. Donnelly, A. Petronilho and M. Albrecht, Chem. Commun., 2013, 49, 1145–1159.
36. M. Albrecht, Adv. Organomet. Chem., 2014, 62, 111–158.
37. M. Delgado-Rebollo, D. Canseco-Gonzalez, M. Hollering, H. Mueller-Bunz and M. Albrecht, Dalton Trans., 2014, 43, 4462–4473.
38. M. Hollering, M. Albrecht and F. E. Kühn, Organometallics, 2016, 35, 2980–2986.
39. D. Canseco-Gonzalez, A. Petronilho, H. Mueller-Bunz, K. Ohmatsu, T. Ooi and M. Albrecht, J. Am. Chem. Soc., 2013, 135, 13193–13203.
40. P. Font, H. Valdés, G. Guisado-Barrios and X. Ribas, Chem. Sci., 2022, 13, 9351–9360.
41. H. Duan, S. Sengupta, J. L. Petersen, N. G. Akhmedov and X. Shi, J. Am. Chem. Soc., 2009, 131, 12100–12102.
42. D. Wang, Y. Zhang, A. Harris, L. N. S. Gautam, Y. Chen and X. Shi, Adv. Synth. Catal., 2011, 353, 2584–2588.
43. Y. Yang, W. Hu, X. Ye, D. Wang and X. Shi, Adv. Synth. Catal., 2016, 358, 2583–2588.
44. C. Lang, K. Pahnke, C. Kiefer, A. S. Goldmann, P. W. Roesky and C. Barner-Kowollik, Polym. Chem., 2013, 4, 5456–5462.
45. M. Baron, S. Bellemin-Laponnaz, C. Tubaro, M. Basato, S. Bogialli and A. Dolmella, J. Inorg. Biochem., 2014, 141, 94–102.
46. V. Lewe, M. Preuss, E. A. Woźnica, D. Spitzer, R. Otter and P. Besenius, Chem. Commun., 2018, 54, 9498–9501.
47. M. Monticelli, S. Bellemin-Laponnaz, C. Tubaro and M. Rancan, Eur. J. Inorg. Chem., 2017, 2017, 2488–2495.
48. M. Monticelli, M. Baron, C. Tubaro, S. Bellemin-Laponnaz, C. Graiff, G. Bottaro, L. Armelao and L. Orian, ACS Omega, 2019, 4, 4192–4205.
49. A. Longhi, M. Baron, M. Rancan, G. Bottaro, P. Sgarbossa and C. Tubaro, Molecules, 2019, 24, 2305.
50. T. Eybe, T. Bohn, J. N. Audinot, T. Udelhoven, H. M. Cauchie, H. N. Migeon and L. Hoffmann, Chemosphere, 2009, 76, 134–140.
51. D. Rodríguez-Hernández, A. J. Demuner, L. C. A. Barbosa, L. Heller and R. Csuk, Eur. J. Med. Chem., 2016, 115, 257–267.
52. H. M. J. Wang and I. J. B. Lin, Organometallics, 1998, 17, 972–975.
53. A. Collado, A. Gomez-Suarez, A. R. Martin, A. M. Slawin and S. P. Nolan, Chem. Commun., 2013, 49, 5541–5543.
54. M. Pažický, A. Loos, M. J. Ferreira, D. Serra, N. Vinokurov, F. Rominger, C. Jäkel, A. S. K. Hashmi and M. Limbach, Organometallics, 2010, 29, 4448–4458.
55. S. Zhu, R. Liang, L. Chen, C. Wang, Y. Ren and H. Jiang, Tetrahedron Lett., 2012, 53, 815–818.
56. P. de Fremont, R. Singh, E. D. Stevens, J. L. Petersen and S. P. Nolan, Organometallics, 2007, 26, 1376–1385.
57. A. H. Mageed, B. W. Skelton and M. V. Baker, Dalton Trans., 2017, 46, 7844–7856.
58. C. Hirtenlehner, C. Krims, J. Holbling, M. List, M. Zabel, M. Fleck, R. J. Berger, W. Schoefberger and U. Monkowius, Dalton Trans., 2011, 40, 9899–9910.
59. S. Ahmad, A. A. Isab, H. P. Perzanowski, M. S. Hussain and N. M. Akhtar, Transit. Met. Chem., 2002, 27, 177–183.
60. S. Zhu, R. Liang and H. Jiang, Tetrahedron, 2012, 68, 7949–7955.
61. H. Ibrahim, P. de Frémont, P. Braunstein, V. Théry, L. Nauton, F. Cisnetti and A. Gautier, Adv. Synth. Catal., 2015, 357, 3893–3900.
62. R. Jothibasu, H. V. Huynh and L. L. Koh, J. Organomet. Chem., 2008, 693, 374–380.
63. H. V. Huynh, S. Guo and W. Wu, Organometallics, 2013, 32, 4591–4600.
64. M. V. Baker, P. J. Barnard, S. J. Berners-Price, S. K. Brayshaw, J. L. Hickey, B. W. Skelton and A. H. White, J. Organomet. Chem., 2005, 690, 5625–5635.
65. W. A. Herrmann, O. Runte and G. Artus, J. Organomet. Chem., 1995, 501, C1–C4.
66. Y.-P. Huang, C.-C. Tsai, W.-C. Shih, Y.-C. Chang, S.-T. Lin, G. P. A. Yap, I. Chao and T.-G. Ong, Organometallics, 2009, 28, 4316–4323.
67. H. V. Huynh, C. H. Yeo and Y. X. Chew, Organometallics, 2010, 29, 1479–1486.
68. H. Günther, NMR Spectroscopy, Basic Principles, Concepts, and Applications in Chemistry, Wiley, 2nd edn, 1995, vol. 343.
69. K. D. Zimmer, R. Shoemaker and R. R. Ruminski, Inorg. Chim. Acta, 2006, 359, 1478–1484.
70. H. V. Huynh and J. Wu, J. Organomet. Chem., 2009, 694, 323–331.
71. P. W. Atkins and J. de Paula, Physical Chemistry, Oxford University Press, 2010.
72. M. Hricovíni, J. Asher and M. Hricovíni, RSC Adv., 2020, 10, 5540–5550.
73. J. Kujawski, K. Czaja, T. Ratajczak, E. Jodłowska and M. K. Chmielewski, Molecules, 2015, 20(7), 11875–11190.
74. M. Boronat, A. Corma, C. González-Arellano, M. Iglesias and F. Sánchez, Organometallics, 2010, 29, 134–141.
75. A. Bondi, J. Phys. Chem., 1964, 68, 441–451.
76. S. Alvarez, Dalton Trans., 2013, 42, 8617–8636.
77. D. Piovesan, G. Minervini and S. C. Tosatto, Nucleic Acids Res., 2016, 44, W367–W374.
78. P. de Fremont, N. M. Scott, E. D. Stevens and S. P. Nolan, Organometallics, 2005, 24, 2411–2418.
79. M. V. Baker, P. J. Barnard, S. K. Brayshaw, J. L. Hickey, B. W. Skelton and A. H. White, Dalton Trans., 2005, 37–43,  10.1039/b412540a.
80. H. M. J. Wang, C. Y. L. Chen and I. J. B. Lin, Organometallics, 1999, 18, 1216–1223.
81. H. Sivaram, J. Tan and H. V. Huynh, Organometallics, 2012, 31, 5875–5883.
82. B. Jacques, J. Kirsch, P. de Frémont and P. Braunstein, Organometallics, 2012, 31, 4654–4657.
83. S. Gaillard, A. M. Z. Slawin, A. T. Bonura, E. D. Stevens and S. P. Nolan, Organometallics, 2010, 29, 394–402.
84. M. Baron, C. Tubaro, M. Basato, A. Biffis, M. M. Natile and C. Graiff, Organometallics, 2011, 30, 4607–4615.
85. H. V. Huynh, Y. Han, J. H. H. Ho and G. K. Tan, Organometallics, 2006, 25, 3267.
86. M. Porchia, M. Pellei, M. Marinelli, F. Tisato, F. Del Bello and C. Santini, Eur. J. Med. Chem., 2018, 146, 709–746.
87. T. Zou, C. T. Lum, C. N. Lok, J. J. Zhang and C. M. Che, Chem. Soc. Rev., 2015, 44, 8786–8801.
88. C. Marzano, V. Gandin, A. Folda, G. Scutari, A. Bindoli and M. P. Rigobello, Free Radic. Biol. Med., 2007, 42, 872–881.
89. J. H. Kim, E. Reeder, S. Parkin and S. G. Awuah, Sci. Rep., 2019, 9, 12335.
90. P. C. Kunz, M. U. Kassack, A. Hamacher and B. Spingler, Dalton Trans., 2009, 37, 7741–7747.
91. X. Zhou and L. Zhou, Theor. Chem. Acc., 2016, 135, 30.
92. M.-C. Brandys, M. C. Jennings and R. J. Puddephatt, J. Chem. Soc., Dalton Trans., 2000, 24, 4601–4606.
93. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
94. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
95. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305.
96. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman and D. J. Fox, Gaussian 16, Revision C.01, Wallingford, CT, 2016.
97. H. Schröder, A. Creon and T. Schwabe, J. Chem. Theory Comput., 2015, 11, 3163–3170.
98. G. M. Sheldrick, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2015, 71, 3–8.
99. T. Mosmann and J. lmmun, Methods, 1983, 65, 55–63.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 2416356–2416359. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d5ra02655e

---