Typically, 0.11 g of LiBH₄ (5 mmol, 95%) was weighed into a glass frit-equipped Schlenk flask within a glovebox, and the flask was then sealed with a rubber septum before being removed from the glovebox. Subsequently, 20 mL of diglyme and 27.5 mmol (2.75 mL) of (CH₃)₂S·BH₃ were added outside the glovebox using standard Schlenk techniques [Supplementary Figure 2A]. The solution was then heated to 120 °C, and the initially colorless solution turned yellow after approximately 2 h, accompanied by the formation of a white precipitate. The reaction mixture was further stirred at 120 °C for 24 h. The reaction solution was allowed to cool naturally to room temperature and was then filtered (using a 10-16 μm fritted glass filter) and washed with diglyme (3 × 5 mL). The resulting sample was dried under vacuum at 80 °C for 2 h to remove residual solvent. The solvated product, Li₂B₁₂H₁₂·n C₆H₁₃O₃, was obtained as a white crystalline solid with a yield of approximately 40% Li₂B₁₂H₁₂ based on the amount of LiBH₄ used.

Typically, 0.11 g of LiBH₄ (5 mmol) was weighed into a quartz-lined reactor inside a glovebox and sealed with Parafilm®. Immediately before loading the quartz reactor into the autoclave, 20 mL of diglyme followed by 27.5 mmol (2.75 mL) of (CH₃)₂S·BH₃ were quickly added to the borohydride. The autoclave was then sealed and purged with argon for 30 s to remove air from the system. The reaction temperature was increased from room temperature to 120 °C at a rate of 2 °C/min and maintained at this temperature for 24 h. After the system had cooled to room temperature, the reaction mixture was filtered using a Schlenk filtration apparatus [Supplementary Figure 2A] and washed with diglyme (3 × 5 mL). The sample was finally dried under vacuum at 80 °C for 2 h. The yield of Li₂B₁₂H₁₂ was approximately 91%, based on the amount of LiBH₄ used.

Typically, 0.22 g of LiBH₄ (10 mmol) was weighed into a quartz-lined reactor inside a glovebox and sealed with Parafilm®. Immediately before loading the quartz reactor into the autoclave, 20 mL of monoglyme followed by 55 mmol (5.5 mL) of (CH₃)₂S·BH₃ were quickly added to the borohydride. The autoclave was then sealed and purged with argon for 30 s to remove air from the system. The reaction temperature was increased from room temperature to 120 °C at a rate of 2 °C/min and maintained at this temperature for 24 h. After the system had cooled to room temperature, the reaction mixture was transferred into a Schlenk flask equipped with a filtration glass frit [Supplementary Figure 2A] and cooled in a refrigerator overnight (10 °C). At this stage, part of the product crystallized from the solution. The sample was then washed with cold monoglyme (10 °C, 3 × 10 mL) under an inert atmosphere, and finally dried under vacuum at room temperature for 2 h. The resulting Li₂B₁₂H₁₂∙n DME (DME, 1,2-dimethoxyethane) was obtained with an approximately 89% yield of Li₂B₁₂H₁₂ based on the amount of LiBH₄ used.

Typically, 1 g of dried Li₂B₁₂H₁₂·n diglyme was suspended in 10 mL of dimethyl sulfoxide (DMSO) in a glass frit-equipped Schlenk flask. Subsequently, the suspension was heated at 100 °C under vacuum to remove residual volatile solvents, affording Li₂B₁₂H₁₂·n DMSO as a solid product. Final desolvation was achieved by thermal treatment at 200 °C for 12 h under dynamic vacuum, affording anhydrous Li₂B₁₂H₁₂ as a white crystalline solid.

Typically, 1 g of dried Li₂B₁₂H₁₂·n DME was suspended in 10 mL of deionized water in a glass frit-equipped Schlenk flask. Subsequently, the suspension was heated at 40 °C under vacuum to remove residual solvents, affording Li₂B₁₂H₁₂·n H₂O as a solid product. Final desolvation was achieved by thermal treatment at 200 °C for 12 h under dynamic vacuum, affording anhydrous Li₂B₁₂H₁₂ as a white crystalline solid.

More experimental details are available in the Supplementary Materials.

In the first step of our new synthetic method, LiBH₄ is reacted with DMS∙BH₃ in diglyme to produce Li₂B₁₂H₁₂, following Equation (1). This reaction was first investigated using a classical Schlenk setup at 120 °C with a molar ratio of 1:5.5 (10% excess DMS∙BH₃) [Supplementary Figure 4].

(1) $$ \mathrm{2MBH_4+10(CH_3)_2S\cdot BH_3\overset{\mathit{diglyme} }{\rightarrow} M_2B_{12}H_{12}\cdot nC_6H_{14}O_3+10(CH_3)_2S+13H_2\ (M=Li,Na,K)} $$

After 24 h, the yield of B₁₂H₁₂^2- in the solid product reached 45%, as shown by the sole presence of a doublet at around -15.2 ppm in the proton-coupled ¹¹B solution-state NMR spectrum of the solid product dissolved in DMSO-d₆ [Figure 2A]^[31,52,53]. NMR analysis of the filtrate [Figure 2B] revealed residual intermediates such as B₃H₈^- and higher-nuclearity boron clusters, including B₉H₁₄^-, B₁₀H₁₃^-, and B₁₁H₁₄^-, all of which are soluble in diglyme^[42,54-56].

To optimize the reaction conditions, the DMS·BH₃ ratio was varied while maintaining all other parameters constant (5 mmol LiBH₄, 20 mL diglyme, and 24 h of heating at 120 °C). Experiments were carried out with LiBH₄/DMS·BH₃ molar ratios of 1:2 and 1:10. The 1:2 ratio, which is optimal for the synthesis of B₃H₈^-, resulted in a Li₂B₁₂H₁₂ yield of approximately 11% based on the amount of LiBH₄ used, with B₃H₈^- remaining as the main boron species in the filtrate [Supplementary Figures 4 and 5, Supplementary Table 2]. In contrast, the 1:10 ratio yielded 25% Li₂B₁₂H₁₂ and produced B₁₁H₁₄^- as the main boron species in the filtrate. Excess DMS·BH₃ thus slightly increases Li₂B₁₂H₁₂ yield but reduces BH₃ utilization efficiency due to the formation of higher-nuclearity boron clusters (B₉H₁₄^-, B₁₁H₁₄^-), which deplete BH₃ in the reaction mixture^[55].

Temperature is a crucial factor in the synthesis of polyhedral borates, and its influence was therefore also investigated in this work^[36,37]. At a lower temperature (85 °C), incomplete conversion of BH₄^- and DMS∙BH₃ into B₃H₈^-, B₉H₁₄^-, B₁₀H₁₃^-, and B₁₁H₁₄^- was observed after 24 h, as shown by ¹¹B NMR [Supplementary Figure 6], leading to a yellow reaction solution without precipitation. After extending the reaction time to 48 h, a very small amount of precipitate appeared, and the ¹¹B NMR spectrum [Supplementary Figure 6] showed weak signals of B₁₂H₁₂^2- and near-complete depletion of LiBH₄. At a higher temperature (160 °C) without a condenser, a poly-anionic species formed directly [Supplementary Figure 7]. We speculate that this species might arise from a reaction between B₁₁H₁₄^- and partially dehydrogenated B₁₂H_12-x^2- anions. Dehydrogenation of the icosahedral B₁₂H₁₂^2- cluster alters the charge and symmetry of the twelve BH vertices, leading to the formation of disubstituted anions when reacted with electrophilic compounds^[57]. Performing the same experiment at 160 °C but using a setup equipped with a condenser resulted in the formation of B₁₂H₁₂^2-, along with some B₁₀H₁₀^2- and partially dehydrogenated B₁₂H_12-x^2- anions [Supplementary Figure 8]^[58]. The use of a condenser also increased the yield compared to the reaction at 120 °C, as indicated in Supplementary Table 3. Additionally, the predominance of B₁₁H₁₄^- in the filtrate [Supplementary Figure 9] indicates that refluxing promotes the conversion of lower boranes to B₁₂H₁₂^2-.

Similar reactions at 160 °C in Schlenk flasks equipped with a condenser [Supplementary Figure 10 and Supplementary Table 4], where LiBH₄ was substituted by NaBH₄ and KBH₄, produced Na₂B₁₂H₁₂ and K₂B₁₂H₁₂ in isolated yields of 92% and 88%, respectively. In this case, higher temperature favors the formation of B₁₂H₁₂^2-, which is in line with previous research showing that the thermal decomposition of BH₄^- and B₃H₈^- salts can generate closo-borates such as B₁₀H₁₀^2- and B₁₂H₁₂^2-[59-62]. Furthermore, in contrast to diglyme solvated Li₂B₁₂H₁₂, the solvates of Na₂B₁₂H₁₂ and K₂B₁₂H₁₂ are much more stable at this temperature^[36]. The small amounts of B₁₀H₁₀^2- impurities in Na₂B₁₂H₁₂ and K₂B₁₂H₁₂ can be completely removed by washing with diglyme. These results indicate that a Schlenk setup equipped with a condenser is suitable for high-yield synthesis of Na₂B₁₂H₁₂ and K₂B₁₂H₁₂ but is not ideal for Li₂B₁₂H₁₂ preparation.

Further optimization of the synthesis conditions was performed by varying the concentrations of reactants and reaction times. The resulting products consistently exhibited high purity under all tested conditions, as shown in Figure 2C and D. The reactions were performed either in a Schlenk setup without a condenser or in an autoclave, as the latter enables all generated gaseous species, such as B₂H₆ and DMS, to remain confined within the reaction volume. As a general trend, the yields of Li₂B₁₂H₁₂ obtained in the autoclave were much higher than those in the Schlenk setup. Based on the yield comparison presented in Figure 2E and Table 1, we speculate that B₂H₆ present in the autoclave atmosphere promotes the hydroboration reaction from BH₄^- to B₁₁H₁₄^-, thus improving the yield of B₁₂H₁₂^{2-[31,37,38,42,50,60,63]}. Detailed reaction conditions are summarized in Supplementary Tables 3-7 and Supplementary Figures 11 and 12. In addition, longer reaction times enabled complete conversion of the reactants, resulting in higher yields. Remarkably, an isolated yield of 96% pure Li₂B₁₂H₁₂ was achieved in the autoclave, representing, to the best of our knowledge, the highest yield achieved thus far for the preparation of metal dodecaborates [Supplementary Table 8]. When the reactant concentration in the autoclave was increased, a drop in yield was observed. This contrasts with the reactions in the Schlenk setup, for which increasing reactant concentrations or reaction times had no significant impact on the final yield of B₁₂H₁₂^2-.

Signals of B₁₀H₁₃^- were not observed in the ¹¹B NMR spectra [Supplementary Figure 13] of filtrates resulting from autoclave experiments. The autoclave keeps all gaseous species under high pressure within the reactor system, and the B₂H₆ dimer can thus be retained in the reaction mixture. We speculate that the pyrolysis of B₂H₆ directly leads to the formation of higher boranes, such as B₄H₁₀ or B₆H₁₀, resulting in an increased content of B₁₁H₁₄^- in solution and a slow reaction between B₃H₈^- and B₁₁H₁₄^-[64,65]. Remaining DMS·BH₃ can be observed in the NMR spectra of products obtained from the autoclave system but not in those from reactions performed in a Schlenk flask.

The synthesis of Li₂B₁₂H₁₂ in an autoclave can also be carried out using a lighter ether such as monoglyme (DME, boiling point of 85 °C) as the solvent. Supplementary Figures 14 and 15 show the ¹¹B NMR spectra of the solid precipitate and filtrate obtained from the reaction of 10 mmol LiBH₄ with 55 mmol DMS·BH₃ in 20 mL monoglyme. The ¹¹B NMR spectra [Supplementary Figures 14 and 15] demonstrate that under these conditions, pure B₁₂H₁₂^2- is obtained, with an isolated yield of 42%. However, a considerable amount of Li₂B₁₂H₁₂ remains as small particles in the filtrate [Supplementary Figure 14B], which makes full recovery of the solid product difficult. Cooling the reaction solution in a refrigerator at 10 °C overnight before filtration addresses this issue by increasing crystallite size. By doing so, and washing the product with cold monoglyme (10 °C), a monoglyme-rich solvated Li₂B₁₂H₁₂ was obtained with an isolated yield of 89%.

After initial drying, the crude products contained variable amounts of diglyme molecules per formula unit, as evidenced by NMR spectroscopy [Supplementary Figures 11 and 12]. Attempts to remove coordinated solvent from Li₂B₁₂H₁₂·n diglyme by thermal treatment were unsuccessful. Thermogravimetric analysis (TGA) was performed to investigate the thermal stability and composition of the Li₂B₁₂H₁₂·n diglyme adducts [Supplementary Figure 16]. The thermal decomposition proceeds in three steps: the removal of free diglyme starts at 25 °C (-9.5 wt.%), followed by a larger mass loss of 15.5 wt.% starting at around 150 °C and a third mass loss of 5 wt.% between 200 and 300 °C. The latter two steps are attributed to the removal of coordinated diglyme.

An in situ synchrotron radiation powder X-ray diffraction (SR-PXRD) experiment was performed to further investigate the behavior of Li₂B₁₂H₁₂·n diglyme during thermal treatment [Supplementary Figure 17]. The only crystalline phase observed in the diffraction patterns corresponds to Li₂B₁₂H₁₂·2 diglyme, which is thermally stable up to 179 °C and then decomposes into an amorphous phase. The structure of Li₂B₁₂H₁₂·2 diglyme was determined by single crystal X-ray diffraction (SC-XRD; Figure 3A, Supplementary Figure 59, Supplementary Table 9). In this solvate, two Li ions are threefold coordinated by three oxygen atoms from two diglyme molecules, evidencing strong interactions with the solvent molecules. It is noteworthy that upon reaching the decomposition temperature, the solid sample starts melting and bubbling. Hydrogen loss from the boron cages was confirmed by ¹H NMR and ¹¹B NMR [Supplementary Figures 18-20], as the characteristic asymmetric multiplet of B₁₂H₁₂^2- (1.6-0 ppm) disappears after thermal treatment at 170 °C under vacuum for 12 h, indicating competition between desolvation and dehydrogenation^[66]. These results suggest that direct removal of diglyme by thermal treatment is not possible.

We investigated a new way to obtain anhydrous Li₂B₁₂H₁₂ from glyme solvates, eliminating the need for the previously used and inconvenient ion exchange strategy, by using DMSO or N, N-diethylformamide (DEF) as solvent exchange agents. This approach aims to overcome the strong chelation of diglyme with dodecaborate salts containing small cations such as Li⁺ and Na⁺. The methodology consists of adding 10 mL of DMSO or DEF to 1 g of the diglyme-solvated product, followed by drying the solution under vacuum at 100 °C for 12 h. Upon this treatment, completely exchanged DMSO- or DEF-solvated Li₂B₁₂H₁₂ is achieved, as demonstrated by powder X-ray diffraction (PXRD) and Fourier transform infrared (FTIR) spectroscopy [Figure 3A-C and Supplementary Figures 21-23]. The ¹H NMR analysis [Supplementary Figure 24] of Li₂B₁₂H₁₂·n DMSO before and after desolvation indicates that DMSO completely replaces diglyme. Additionally, DMSO can be removed from the solvate by thermal treatment without decomposition of the B₁₂H₁₂^2- anion. This suggests that DMSO can be used as an effective solvent for the removal of diglyme in Li₂B₁₂H₁₂. Subsequently, drying the solvent-exchanged products at 200 °C under vacuum results in the formation of unsolvated Li₂B₁₂H₁₂, as shown by in situ PXRD [Supplementary Figure 23] and NMR [Supplementary Figures 24-26]. The choice of DMSO and DEF for solvent exchange is justified by their higher boiling points and weaker coordinating ability compared to diglyme^[67]. Although both solvents can effectively substitute diglyme, DMSO is the more economically viable option given the much higher cost of DEF.

When monoglyme is used in the synthesis, the crystalline phase of the obtained solvate corresponds to Li₂B₁₂H₁₂·1.5 monoglyme, as evidenced by its structure determined by SC-XRD [Figure 3D, Supplementary Figure 60, Supplementary Table 10]. In this structure, monoglyme shows a coordinating ability toward Li⁺ similar to that of diglyme in solvated Li₂B₁₂H₁₂, and thus may pose similar challenges for desolvation. However, in situ PXRD data [Supplementary Figure 27] reveal that monoglyme-solvated Li₂B₁₂H₁₂ undergoes different phase changes during thermal desolvation compared to diglyme-solvated Li₂B₁₂H₁₂, ultimately yielding Li₂B₁₂H₁₂. Diffraction peaks of the monoglyme-solvated Li₂B₁₂H₁₂ disappear at around 150 °C, while Li₂B₁₂H₁₂ forms at this temperature. At 310 °C, a phase transition from the α-Li₂B₁₂H₁₂ to β-Li₂B₁₂H₁₂ is observed, followed by decomposition into a hydrogen-poor γ-Li₂B₁₂H_12-x phase at higher temperatures^[68,69]. This last phase transition occurs at a temperature approximately 50 °C lower than previously reported, likely due to the experiment being performed under vacuum. An attempt to directly remove monoglyme from the solvated Li₂B₁₂H₁₂ by heating at 180 °C under vacuum for 12 h resulted in a small amount of Li₂B₁₂H_12-x, as observed in the ¹¹B NMR spectrum [Supplementary Figure 28].

Although complete removal of monoglyme from Li₂B₁₂H₁₂ through direct heating is difficult, the solvent exchange strategy can in this case be implemented with an even lighter, more abundant, and greener solvent, namely water (H₂O), compared to DMSO and DEF used for the diglyme-solvated dodecaborate. Indeed, upon exchange with water and applying vacuum at 40 °C, monoglyme can be completely removed, as shown in Figure 3D-F and ¹H NMR spectra [Supplementary Figures 29-31], and pure hydrated Li₂B₁₂H₁₂ can be obtained. Water can then be easily removed from hydrated Li₂B₁₂H₁₂ without damaging the B₁₂H₁₂^2- clusters. It is worth noting that Li₂B₁₂H₁₂ is commercially available in hydrated form, which can be conveniently obtained by our solvent exchange strategy without the last heating step.

Single crystals of diglyme-solvated Na₂B₁₂H₁₂ and K₂B₁₂H₁₂ were also isolated from the reaction mixtures during synthesis, and their structures were determined by SC-XRD to be Na₂B₁₂H₁₂·4 diglyme [Figure 4A, Supplementary Figure 61, Supplementary Table 10] and K₂B₁₂H₁₂·4 diglyme [Figure 4B, Supplementary Figure 62, Supplementary Table 11]. Unlike Li⁺, Na⁺ and K⁺ possess lower charge densities and larger cation sizes, resulting in higher cation-to-anion size ratios^[70]. This leads to different coordination environments for Na⁺ and K⁺, with M⁺∙∙∙O bonds (Na⁺∙∙∙O = 2.378-2.446 Å, K⁺∙∙∙O = 2.728-2.844 Å) longer than Li⁺∙∙∙O = 1.958-2.039 Å, meaning that less energy is needed to break these coordination bonds. This finding explains why direct heat treatment of diglyme-solvated Na₂B₁₂H₁₂ (150 °C) and K₂B₁₂H₁₂ (200 °C) under vacuum effectively removed diglyme in our previous work, whereas the same approach was ineffective for the Li⁺ salt^[36].

Interestingly, drying crystals of diglyme-solvated Na₂B₁₂H₁₂ at 80 °C for 2 h induced a single-crystal-to-single-crystal transformation, yielding a high-temperature (HT) diglyme-solvated form of Na₂B₁₂H₁₂ [Supplementary Figure 63 and Supplementary Table 12]. Decomposition of the HT form, Na₂B₁₂H₁₂·2 diglyme, directly yielded anhydrous Na₂B₁₂H₁₂ [Supplementary Figures 32-36]. In contrast, decomposition of diglyme-solvated K₂B₁₂H₁₂ proceeded steadily in a single step, leading to the formation of K₂B₁₂H₁₂, as illustrated in Supplementary Figures 37 and 38.

We also attempted synthesis in toluene, a non-coordinating solvent, with the aim of directly obtaining unsolvated M₂B₁₂H₁₂ (M = Li, Na). The ¹¹B NMR spectra [Supplementary Figures 39-41] confirmed the formation of Li₂B₁₂H₁₂ along with BH₄^-, B₁₁H₁₄^-, B₁₁H₁₃^2-, and the thiomethyl-substituted cluster B₁₂H₁₁S(CH₃)^-[71]. PXRD analysis [Supplementary Figures 42-44] revealed Li₂B₁₂H₁₂ as the only crystalline phase when the reaction was performed at 100 °C, while additional impurity peaks appeared at 120 °C. Performing the reaction in an autoclave at 120 °C achieved complete conversion of BH₄^- to B₁₂H₁₂^2-. We also ball-milled LiBH₄ prior to synthesis in toluene to further improve the contact surface, as this reactant is insoluble in the solvent. This pretreatment significantly enhanced its reactivity, leading to the formation of BH₄^- and B₁₁H₁₃^2- together with B₁₂H₁₂^2- [Supplementary Figures 45-50]. While high-purity Li₂B₁₂H₁₂ could be obtained from toluene, post-synthetic solvent extraction was necessary to remove residual byproducts. In contrast, replacing LiBH₄ with NaBH₄ primarily yielded NaB₁₂H₁₁S(CH₃)₂ rather than Na₂B₁₂H₁₂ [Supplementary Figures 51 and 52]. These results suggest that the coordination and polarization characteristics of the metal cation play critical roles in directing closo-borate formation, particularly under conditions where the borohydride precursor exhibits limited solubility in the reaction medium.

To gain insight into the mechanism of B₁₂H₁₂^2- formation, ¹¹B NMR spectroscopy was employed to monitor the composition of the reaction mixture throughout the synthesis. Figure 5A shows the ¹¹B NMR spectra of aliquots collected from a mixture of LiBH₄ and DMS·BH₃ in diglyme before heating (0 h) and at different time points during the solvothermal reaction at 120 °C.

In the initial mixture at room temperature, a signal at -24 ppm was observed together with those of the reactants and was assigned to B₂H₇^-[42]. This intermediate was then quickly converted into B₃H₈^-, together with B₉H₁₄^-, during the first half hour of the solvothermal reaction, in agreement with previous observations^[42]. Notably, the intensities of the BH₄^- and DMS·BH₃ signals decreased rapidly after initiating heating, accompanied by increasing concentration of B₉H₁₄^- and B₁₁H₁₄^- during the first hour of the reaction. This behavior is consistent with earlier reports on the dehydrocondensation of B₃H₈^- with B₂H₆^[72]. Furthermore, the characteristic doublet of B₁₂H₁₂^2- appeared after only 2 h of reaction, and its intensity increased continuously as the reaction progressed. In contrast, the intensities of the B₉H₁₄^- and B₁₁H₁₄^- signals remained almost constant throughout the reaction, while gradual consumption of B₃H₈^- was observed. The sustained concentrations of B₉H₁₄^- and B₁₁H₁₄^- may indicate that these species act as intermediates in a dynamic equilibrium. Alternatively, their behavior could reflect a kinetic steady state in which the rates of formation and consumption are closely balanced, thereby preventing accumulation despite continuous turnover. Based on these ¹¹B NMR observations, Figure 5B proposes a stepwise formation pathway for B₁₂H₁₂^2- involving the key intermediates B₂H₇^-, B₃H₈^-, B₉H₁₄^-, and B₁₁H₁₄^-.

The above experiment demonstrates that B₃H₈^-, B₉H₁₄^-, and B₁₁H₁₄^- are stable intermediates in the sequential buildup of B₁₂H₁₂^2- from BH₄^- and BH₃. To understand the final step in the formation mechanism of Li₂B₁₂H₁₂, isolation of the intermediate B₁₁H₁₄^- was critical. After several attempts [Supplementary Figures 53-55], this intermediate was successfully isolated by subjecting the reaction filtrate obtained from the initial synthetic step to thermal treatment at 160 °C [Figure 5C, Supplementary Figures 56 and 57]. To identify which species react with B₁₁H₁₄^- to form dodecaborate, a solution of isolated LiB₁₁H₁₄ in diglyme was reacted with various lower boranes, including DMS·BH₃ and BH₄^-, at 120 °C in a Schlenk flask for 3 h. As shown in Figure 5D and Supplementary Figure 58, direct reaction of LiB₁₁H₁₄ with BH₄^- did not yield B₁₂H₁₂^2-; instead, it produced B₁₁H₁₃^2- through deprotonation of B₁₁H₁₄^- by BH₄^-. On the other hand, reaction of B₁₁H₁₄^- with DMS·BH₃ successfully generated B₁₂H₁₂^2-, whereas thermal decomposition of DMS·BH₃ under the same conditions did not produce detectable amounts of B₁₂H₁₂^2- [Supplementary Figure 58]. Previous reports have demonstrated the isolation of the B₁₂H₁₂^2- anion through similar reactions, such as that between B₁₁H₁₄^- and triethylamine borane at 150 °C. It has also been reported that Na₂B₁₂H₁₂ can be obtained by reacting NaB₁₁H₁₄ with NaBH₄ in boiling diglyme (boiling point = 162 °C)^{[8,37,39,41,55,73]}. Our results indicate that at the lower reaction temperature of 120 °C, deprotonation is likely favored over direct hydroboration. Therefore, we propose that B₁₁H₁₃^2- serves as the terminal intermediate in this sequential pathway, and that its conversion to B₁₂H₁₂^2- proceeds only in the presence of sufficient BH₃.