Assembly of high-nuclearity Sn26, Sn34-oxo clusters: solvent strategies and inorganic Sn incorporation

Unprecedented tin-oxo clusters with record high-nuclearities and electrocatalytic CO2 reduction applications have been prepared via solvent dependent assembly strategies.


Introduction
Tin oxide (SnO 2 ) has attracted increasing research attention due to its application in a variety of areas, including gas sensors, 1 catalysis, 2,3 lithium batteries, 4,5 solar cells 6,7 and transparent electrodes. 8 Some important factors, such as the size, composition, and structure, greatly inuence the electronic and physicochemical properties of SnO 2 . [9][10][11] Thus it is crucial to understand the binding mode and atomic connectivity of tin oxide materials at the molecular level, which will be benecial for exploring the structure-property relationship and further achieving precise tuning of the physicochemical properties.
As molecular models of tin oxide materials, crystalline tinoxo clusters (TOCs) can provide precise atomic structural information by X-ray diffraction analysis. Accordingly, they are efficient molecular tools for building bridges between theoretical modeling, crystallography, and physical applications. Many research groups have made great efforts to explore the synthesis and structures of TOCs. [12][13][14][15][16][17][18][19][20][21][22] One common way is to use organotin as precursors to react with carboxylate ligands in organic solvents, such as benzene, toluene, and alcohols. [23][24][25][26][27] [31][32][33] However, despite these advances, the structural diversity of tin-oxo clusters is still less developed compared to transition-metal oxo clusters. And the TOC nuclearity also remains quite low, with the highest one being Sn 14 . 34 Therefore, it is an attractive but challenging goal to synthesize TOCs with larger core nuclearity and more structural diversity. Moreover, although most of the driving forces of the research on TOCs originate from understanding the physical attributes of tin oxide, the studies on their applications (e.g. in catalysis) still remain very rare. Thus it is rather urgent to explore the physical application of TOCs to acquire the important Sn-O connectivity-activity relationship.
The assembly of tin-oxo clusters highly depends on the reactivity of organotin precursors, which is further inuenced by organic ligands, reaction temperature, solvent and so on. Therefore, to prepare a diverse range of high-nuclearity TOCs, it is necessary to develop new synthetic approaches. During our recent research in titanium-oxo clusters, [35][36][37] high-temperature assembly reactions were widely used for the preparation of high nuclearity Ti-O clusters. Herein, we introduce this method into the eld of tin-oxo clusters, and make the following indispensable modications: (1) mixed alcohol-water solvents were applied to inuence the aggregation rate of Sn atoms and the nucleation rate of tin-oxo clusters; (2) the aprotic solvent CH 3 CN was then used in some cases instead of the general protic solvent to affect the conguration of basic building blocks, as well as their way of connecting; (3) nally, inorganic Sn atoms without any alkyl group or phenyl group were introduced into the reaction system to supply more bridging Sn atoms to promote the formation of high nuclearity TOCs. With these synthetic strategies, a series of unprecedented high-nuclearity TOCs have been successfully obtained whose atomic structures were characterized by single crystal X-ray diffraction analysis (Table 1). They possess a much higher number of Sn atoms (26,34) than the known TOCs (#14). Meanwhile, they also present new structural types different from previous TOCs, including the rod-shaped Sn 26 and cage-dimer Sn 34 . Furthermore, the application of Sn 26 or Sn 34 derived electrode in electrocatalytic CO 2 reduction was investigated for the rst time. Nuclear magnetic resonance (NMR) spectroscopy analysis indicated that formate was obtained as the only liquid reduction product. And the corresponding formate faradaic efficiency (FE) was found to be cluster dependent, with the highest value of 41.90% on the Sn 26 derived electrode.

Results and discussion
In the initial stage of our research, mixed methanol-isopropanol was used as the solvent for reactions between butyltin hydroxide oxide and 2,6-pyridinedicarboxylic acid/NaOH or phenylphosphonic acid, resulting in the formation of two Sn 6 clusters (TOC-12 and TOC-13). Structural analysis indicated that they were both composed of two O-capped {Sn 3  In order to obtain different building units and change the aggregation/nucleation rate of tin atoms, water which is crucial for the growth of oxo clusters was directly introduced into the reaction system. Consequently, a new Sn 12 TOC was prepared namely [(n-BuSn) 12 (OH) 18 propionic acid). As shown in Fig. 1c, TOC-14 is constructed from two different building blocks, the O-capped {Sn 3 O 4 } and ladder {Sn 4 O 4 } units, indicating that the introduction of water is an effective strategy to prepare different structural types of TOCs with higher nuclearity.
By further optimizing the reaction conditions, especially the amount of used water, TOC-15 and TOC-16 with nuclearities of Sn 26 were successfully obtained in a mixed solvent system of methanol-isopropanol-water (Table 1, Fig. 2). To the best of our knowledge, the numbers of Sn atoms (26) in these TOCs far exceed the value (#14) in reported tin-oxo clusters to date. The core size of these nanoclusters is $1.6 Â 1.0 nm. As shown in Fig. 2, different from the usual cage structures of reported TOCs, TOC-15 presents a rod-shaped cluster core that can be    (Fig. 2c and d). By changing the functional ligands, other Sn 26 clusters of TOC-16 were obtained with the six labile coordination sites decorated with IANO ligands (Fig. S6 †). Although TOC-15 and TOC-16 present interesting highnuclearity layered structures, their yields are unfortunately quite low (2-4%), which greatly limits further applications. To obtain some mechanistic information of such synthetic shortages, we analyzed their structures in more detail. It is interesting to nd that TOC-15 and TOC-16 contain pure inorganic SnO 6 nodes without butyl groups. Such completely O-coordinated Sn atoms should be derived from the Sn-C bond cleavage of the applied butyltin hydroxide oxide. Considering the difficulty of Sn-C bond cleavage under the applied low-temperature conditions, it will benet the assembly of such Sn-O cores if isolated Sn ions could be incorporated into the reactions. For this aim, the inorganic SnCl 4 precursor was further introduced into the synthetic reaction of TOC-16. As expected, TOC-17 was successfully isolated in a much higher yield ($60%). As presented in Fig. S8, † TOC-17 is composed of the same Sn 26 moiety in TOC-16 and two additional {Sn 2 } dimers which are made up of two Sn atoms bridged by two oxygen atoms and one IANO ligand. The successful preparation of TOC-17 demonstrates that the strategy of introducing additional inorganic Sn atoms is indeed helpful for the formation of high-nuclearity TOCs.
Based on the above results, we can clearly see that the assembly of Sn-O clusters is greatly inuenced by the applied solvent conditions. Pure alcohol environments gave rise to small clusters of Sn 6 , and the introduction of water could signicantly increase the nuclearities to Sn 26 (Table 1). 38 However, as a whole, the above used solvents are all protic ones. If an aprotic solvent, e.g. CH 3 CN, could be applied, the different solvent environment may change the conguration of basic Sn-O building units, as well as their way of connecting. Following this consideration, the reaction of butyltin hydroxide oxide with 2-picolinic acid and NaOH was carried out in pure CH 3 CN. As a consequence, TOC-18 with a nuclearity of Sn 34 and a core size of $2.6 Â 1.1 nm was successfully synthesized, which is the largest Sn-O cluster reported to date. As exhibited in Fig. 3a, TOC-18 is made up of two unprecedented {Sn 12 } and {Sn 22 } cages linked together by a Na atom and PA ligands. Compared with the typical football cage {(RSn) 12 O 14 (OH) 6 } 2À , the {Sn 12 } in TOC-18 displays an asymmetric cage-like structure (Fig. 3b). The {Sn 22 } cage captures a central Na heteroatom via six oxygen atoms (Fig. 3c). From an architectural point of view, the {Sn 22 } cage can be considered to consist of four subunits with different  numbers of Sn atoms, including a top Sn 3 moiety, two middle Sn 6 circles and a bottom Sn 7 base.
Recently, SnO 2 has shown interesting energy conversion applications through electrocatalytically reducing CO 2 to formate. 2 Considering the similar layered characteristics of the Sn 26 cluster core and rutile SnO 2 , the obtained Sn 26 may also have potential application in the electrocatalytic CO 2 reduction reaction (CO 2 RR). Therefore, carbon paper with TOC-17 modi-cation was used as the working electrode to study its catalytic activity towards the CO 2 RR. A linear sweep voltammetry (LSV) test was conducted in Ar or CO 2 saturated 0.5 M KHCO 3 solution, respectively. As shown in Fig. 4a, the LSV curve of the TOC-17 derived electrode in the CO 2 saturated electrolyte exhibits the onset potential at approximately À0.69 V. Beyond this onset potential, the current density continuously increases and reaches 6.73 mA cm À2 at À1.159 V, which is obviously higher than that in the Ar saturated electrolyte. This indicates that the TOC-17 derived electrode may possess high catalytic activity towards the CO 2 RR. In order to identify and quantify the reduction products, gas chromatography (GC) and nuclear magnetic resonance (NMR) spectroscopy were applied to analyze the gas and liquid phase products during the CO 2 RR process. The obtained results indicate that formate was the only liquid CO 2 RR product, with the highest faradaic efficiency (FE) of 41.90% at À1.196 V (Fig. 4). Meanwhile, only a small amount of CO was detected in the gas phase products. Powder X-ray diffraction analysis further conrmed that the structures of TOC-17 remained rather intact aer electrolysis (Fig. S46 †). For comparison, the CO 2 RR application of the cage-dimer structure TOC-18 was also studied. Although presenting higher nuclearity and also producing formate as the main product, the activity of the TOC-18 derived electrode was signicantly lower than that modied with TOC-17. Therefore, these results indicate that the layered Sn-O structures might be benecial for electrocatalytic CO 2 RR applications.

Conclusions
In summary, we have successfully synthesized a series of unprecedented high-nuclearity tin-oxo clusters by solvent dependent synthetic strategies. These obtained high-nuclearity TOCs, with core sizes ranging from 1.6 to 2.6 nm, possess a much higher number of Sn atoms (26,34) than previously known ones (#14). Moreover, new structural types of layered nanorods and cage-dimers were also prepared. The applied solvent environments and Sn sources have proven to play crucial roles in the assembly of these high-nuclearity tin-oxo clusters. The introduction of water into alcohol greatly increased the cluster nuclearity; the incorporation of inorganic Sn ions signicantly increased the yields of layered structures, while the application of aprotic CH 3 CN produced the largest Sn 34 to date. Moreover, electrocatalytic CO 2 reduction studies conrmed that the electrodes derived from the layered Sn 26 cluster presented better performance than those derived from the Sn 34 cage-dimer. Therefore, these results afford effective synthetic strategies for the assembly of high-nuclearity TOCs and also extended their potential applications.

Conflicts of interest
There are no conicts to declare.