Tripodal gold(I) polypyridyl complexes and their Cu+ and Zn2+ heterometallic derivatives. Effects on luminescence

The synthesis of three gold(I) tripodal complexes containing the tris(2pyridylmethyl)amine (TPA) ligand coordinated to Au-PR3 moieties (PR3 = 1,3,5-triaza7-phosphatricyclo[3.3.1.13.7]decane, PTA (1), 3,7-diacetyl-1,3,7-triaza-5phosphabicyclo[3.3.1]nonane, DAPTA (2) and triphenylphosphane (3) was performed together with a cage-like structure containing the triphosphane 1,1,1tris(diphenylphosphinomethyl)ethane (4). The luminescence of these complexes has been studied and they show a red shift upon formation of heterometallic complexes by reaction with Zn(NO3)2, CuCl and [Cu(CH3CN)4]BF4. The different coordination motifs of the Zn2+ and Cu+ heterometallic species and the resulting changes on the recorded absorption, emission and NMR spectra were analysed and supported by TD-DFT calculations.


Introduction
Tris(2-pyridylmethyl)amine (TPA) which possesses four nitrogen atoms able for the ligation of a wide variety of metal ions, and its derivatives are a family of versatile ligands that has been extensively studied. [1] The mobility of its pyridine arms allows TPA to accommodate metal centers with different coordination geometries and electronic structures [2] with a wide variety of different application including optoelectronics, sensing [3][4][5][6] or catalysis [7,8] among others. The additional introduction of terminal alkynyl moieties allows the coordination with d 10 coinage-metals, such as gold(I), [9] that are very well known to favor the coordination by the establishment of metallophilic contacts.
Moreover, gold(I) complexes are an important class of complexes well known to facilitate a triplet excited state emission [10] and the formation of weak Au···Au interactions arising from relativistic effects, play a key role in the observed emission. [11] Additionally, the ability of the alkynes to bind late-transition metal ions in a bidentate mode through a combination of σ and π bonding has been extensively utilized in the synthetic chemistry of Cu, Ag, and Au. This coordination mode of alkynyl ligand provides additional opportunities for fine tuning of the photophysical characteristics of the resulting compounds. [12,13] This fact, together with the tripodal shape of the TPA ligand makes the resulting complexes particularly interesting due to their luminescent properties able to be modulated by the presence or absence of aurophilic contacts. The coordination of a second metal atom can play an important role in this process. However, controlling the formation of heterometallic arrays is more challenging in comparison with the formation of homometallic gold(I) species.
The second coordination position of the linear Au(I) complexes is frequently occupied by tertiary phosphanes, as ancillary soft donors. In general, they have less influence on the arrangement of the metal core and, consequently, on the photophysical properties of the complexes. Nevertheless, they have a particular influence on the solubility of the resulting complexes. [10,[14][15][16][17][18] The presence of several metal ions or atoms in one molecular entity often brings a cooperative effect and leads to the emergence of particular physical and chemical properties, which cannot be attributed to the sum of the properties of the monometallic components. Such synergistic functionality has significantly driven the experimental and theoretical efforts that have impressively advanced the fundamental chemistry of polymetallic structures. [19] Taking all of this into consideration, we have designed and synthesized different tripodal gold(I) alkynyl complexes derived from novel TPA ligands and containing three different monophosphanes and a triphosphane. The free coordinative positions of the pyridyl moieties of TPA have been explored in order to obtain Au(I)/Cu(I) or Au(I)/Zn(II) heterometallic structures with tunable photophysical properties.

Synthesis and Characterization
Synthesis of tris-pyridyl alkynyl ligand, L.
The synthesis of the tripodal complexes required the previous design and synthesis of the tripyrdidylmethylamine ligand L, containing three alkynyl moieties in ortho position with respect to pyridyl nitrogen as it is displayed in Scheme 1.
It is synthesized through the previous synthesis of L1 and L2 precursors via reductive amination of commercially available 6-bromo-2-pyridinecarboxaldehyde (A) with sodium borohydride triacetate and ammonium acetate (L1) [20] and the following three-fold Sonogashira coupling between L1 and dimethyl ethynyl carbinol and deprotection with sodium hydroxide gaining the desired tris-pyridyl ethynyl ligand L in 60% yield. The correct formation of the product was confirmed by 1    All complexes were obtained as a yellow solid stable to air and moisture. Their spectroscopic data are in accordance with the proposed stoichiometry. 1 H NMR spectra display the typical pattern of the phosphanes, the protons of the pyridine moiety and the disappearance of the terminal alkynyl protons upon formation of the complexes. 31 Figure S14) and 1 H and 31 P NMR spectra, although in this case, the solubility of the complex is considerably lower (Figures S15-S16).

Photophysical characterization
Absorption and emission spectra of all complexes were recorded in DMSO and acetonitrile at 10 -5 M concentration and the results are summarized in  An increase on the baseline in the absorption spectra can be also detected in the case of 4, which results from scattering effects due to the presence of small aggregates [15] in agreement with the lower observed solubility of the cage-like complex in this solvent.
The same scattering is observed in all cases when the spectra are recorded in acetonitrile ( Figure S17), where samples are not perfectly dissolved. In order to provide further evidences of these aggregates, UV-vis absorption spectra at different DMSO-water contents mixtures were examined ( Figure S18). The absorption bands at ca. 300 nm become broader when the water content is increased, and the shoulder at ca. 350 nm becomes the main band together with a clear increase of the baseline. All these profiles are in agreement with aggregates' formation. [15] Excitation of the samples at 350 nm in DMSO solutions displays only significant emission in the case of L and the emission is very weak in the presence of gold, probably due to non-radiative decays from a previously populated triplet state due to the heavy atom effect of the gold atom. On the other hand, the aggregated samples display intense emission centered at ca. 530 nm both in DMSO/water mixtures ( Figure S18

Tuning luminescence by the formation of heterometallic complexes.
One way to modulate the luminescence properties of the complexes is the formation of heterometallic structures, which will introduce changes on the threedimensional conformation of the molecules upon coordination with a second metal and to the additional intrinsic transitions resulting from the new complexes. We would like to exploit this methodology in our work by the reaction of 1-4 with Zn(II)-and Cu(I)-salts that have been chosen due to the well-known ability of these metals to coordinate amines (Zn(II) and Cu(I)) and alkynyl groups (in the case of Cu(I)), both present in L. [30][31][32] Preliminary analyses of the possible formation of heterometallic complexes were performed by absorption and emission titrations that allowed us to identify the final stoichiometry of the resulting heterometallic structures as 1:1 (Figures 2 and S20-S28).
In this way, absorption and emission spectra were recorded after addition of increasing amounts of Zn(II) and Cu(I) salts into a 1·10 -5 M solutions of complexes 1-3. The very low solubility of 4 precluded to carry out this type of studies.  Figure S24). These changes are in agreement with modifications in the chemical structure involving the tris-amine part of the molecule. [1,30] A progressive quenching of the emission band is observed in all cases, which is more significant in the case of Cu(I) salts, supporting a stronger interaction with this metal ( Figure 2). The analysis of the observed spectral variations prompted us to calculate the association constants using the HypSpec 1.1.33 software for Windows [33] and revealed a different behaviour for the heterometallic complexes involving Au(I)/Zn(II) and Au(I)/Cu(I). In the first case, the data fit well to a complex with 1:1 stoichiometry (see  Table   2 and Figures S21-S22 for analysis of the titrations involving 1 and Cu(I) salts).    The observed differences may be ascribed to the fact that chloride counterion is expected to be also coordinated to copper while BF 4is acting as external counterion and does not influence on the heterometallic species ( Figure S54). [12,19,31] The mass spectra of compounds xb ( x = 1-4) confirms the presence of chloride in the structure and supports the different coordination of the two copper salts ( Figure S33).
Significant differences can be also observed in the absorption spectra recorded for the different heterometallic complexes, that support the different coordination motifs.  The observed changes in the emission are due to the different coordination motifs of the second metal ion and the change in the metallophilic interactions since, in the case of copper derivatives, Cu(I)···Au(I) interactions can take also place. [31,43] CuCl coordinates to the alkynyl moiety of two of the three units of the complex and the aurophilic/metallophilic interactions are more effective than for the respective gold homometallic complex. A global approximation of the three gold atoms is expected to be even stronger for the reaction with [Cu(NCMe) 4 ]BF 4 , where Cu(I) coordinates to the three alkynyl moieties. On the other hand, coordination of the zinc cation is observed through the pyridyl moieties. Thus, the three arms of the tripodal unit become closer to each other for the copper heterometallic complexes and, in particular, for those having BF 4as external counterion. Although several attempts to grow single crystals suitable for X-ray diffraction failed, these statements are supported by the experimental data and comparison to literature. [12,31,43,44] [45] DAPTA, [46] PPh 3 [47] ), [Au(PPh 3 )(acac)], [48] tetrakis(acetonitrile)copper(I) tetrafluoroborate. [49] Physical measurements Infrared spectra have been recorded on a FT-IR 520 Nicolet Spectrophotometer. 1

Dalton Transactions Accepted Manuscript
Published

Synthesis of [Au 3 (PTA) 3 LCuCl] (1b)
Solid CuCl (0.51 mg, 0.005 mmol) was added to a solution of 1 (5 mg, 0.004 mmol) in CH 3 CN (10 mL). The solution was stirred during 2h at room temperature and then evaporated to dryness under vacuum to yield an orange solid that was washed with acetonitrile. Yield: 80%. 31

Computational details
All compounds were minimised with the Gaussian16 program package. [50] at the DFT level of theory with a hybrid density functional PBE0. [51] The basis set consisted of a quasi-relativistic effective core potential basis set def2-TZVPPD for gold atoms and 6-31G(d,p) for all atoms. [52] Frequencies were calculated to verify the minima.