A Simple and Efficient Way to Directly Synthesize Unsolvated Alkali Metal (M = Na, K) Salts of [CB₁₁H₁₂]⁻

Abstract

Alkali metal (M = Na, K) salts of the [CB₁₁H₁₂]⁻ anion have attracted increasing attention in many fields such as catalysis and all-solid-state batteries. However, tedious and low-yielding synthetic methods have seriously limited their application in these fields. We developed a facile method for the direct synthesis of the unsolvated potassium and sodium salts of the [CB₁₁H₁₂]⁻ anion in 66% and 68% yields, respectively, by reactions of Na[B₁₁H₁₄] with NaH/NaHMDS (sodium bis(trimethylsilyl)amide), CF₃SiMe₃ in THF and K[B₁₁H₁₄] with KH, CF₃SiMe₃ in DME. This method avoids the exchange of cation and significantly simplifies the reaction setup, thus enabling convenient large-scale synthesis. It is believed that this method will support further application of Na[CB₁₁H₁₂] and K[CB₁₁H₁₂] as solid electrolytes for an all-solid-state battery.

1. Introduction

The monocarborane [CB₁₁H₁₂]⁻ anion and dicarbaborane C₂B₁₀H₁₂ are two important icosahedral derivatives of [B₁₂H₁₂]²⁻ in which one or two BH vertices are replaced by CH [1]. The [CB₁₁H₁₂]⁻ anion was synthesized for the first time by Knoth in 1967 [2]. Recently, the study on complexes containing the [CB₁₁H₁₂]⁻ anion has been of interest because they are well known as weakly coordinating anions [3]. Monocarborane anions with halogen substituents, such as [CB₁₁Me₆X₆]⁻, [CB₁₁H₆X₆]⁻ (X = Cl, Br, I) and [CHB₁₁Cl₁₁] are amongst the least coordinating, least basic and most chemically inert anions presently known, and have been used to stabilize many new extremely reactive cations [4,5]. H(CHB₁₁Cl₁₁) is the strongest isolable Brønsted acid [6]. Additionally, they can also be applied in the fields of coordination [7], supramolecular chemistry [8] and medicine [9], as well as fluorescence [10] and materials [11]. More importantly, it has been found that M[CB₁₁H₁₂] (M = K, Na) have great application prospects as solid electrolytes [12,13]. However, compared to the large study on the application and functionalization of dicarbaborane C₂B₁₀H₁₂ [14,15,16], the research on the monocarborane [CB₁₁H₁₂]⁻ anion has been much more scarce [17,18,19]. This should be attributed to its complicated synthetic method and the fact that it is hard to commercialize. So the development of an efficient method to directly synthesize the [CB₁₁H₁₂]⁻ anion is highly desired.

The counter cation is [Me₃NH]⁺ in most synthetic methods reported for the synthesis of the [CB₁₁H₁₂]⁻ anion when considering the solubility of the salts of [CB₁₁H₁₂]⁻ in organic solvents. These methods have been improved since 1967 (Scheme 1). In early reported methods, decaborane (B₁₀H₁₄) was mainly used as a starting material, which is toxic and flammable. The first preparation of [CB₁₁H₁₂]⁻ was achieved by Knoth through the reaction of decaborane with sodium cyanide (Scheme 1A) [2]. A safer and more convenient modification of the synthesis was developed by the group of Hughes using alkyl isocyanides replacing toxic cyanide to form the initial C-B bond (Scheme 1B) [20]. Later, the group of Kennedy used formaldehyde and decaborane to form the B-C bond, which could simplify the synthetic route and save the trouble of a C-N bond fracture (Scheme 1C) [21]. However, the conditions for closing the cage by adding Me₂S·BH₃ require harsh reaction conditions such as 3 days of heating.

To avoid the use of toxic decaborane, the [B₁₁H₁₄]⁻ anion was introduced to the synthesis of [CB₁₁H₁₂]⁻ [22,23,24,25]. In 2001, Michl’s group reported a relatively safe way to synthesize [CB₁₁H₁₂]⁻ by deprotonation of [B₁₁H₁₄]⁻, followed by the insertion of dichlorocarbene, but the yield was very low due to the low activity of the selected carbene source (Scheme 1D). Agnes Kütt and co-workers found that [Me₃NH][CB₁₁H₁₂] could be obtained by reacting [B₁₁H₁₄]⁻ with NaH and CF₃SiMe₃ in up to 95% yield (Scheme 1E) [24]. By improving the above method, [CB₁₁H₁₂]⁻ was synthesized with common laboratory reagents such as NaOH, K₂CO₃ and CHCl₃, despite the yield being relatively low (40%) (Scheme 1F) [25]. Obviously, recent studies mainly focused on the synthesis for the [Me₃NH]⁺ salt of [CB₁₁H₁₂]⁻, while methods on the direct synthesis of alkali metal (M = Na, K) salts of the [CB₁₁H₁₂]⁻ anion are still very limited, which is inconsistent with the great application of alkali metal (M = Na, K) salts of the [CB₁₁H₁₂]⁻ anion. Recently, we improved the synthetic method of [Et₄N][closo-1-CHB₉H₉] [26], and on the basis of our previous work on the condensation reaction of the B-H bond for the synthesis of polyhedral boranes [27,28,29,30] and the application of a dihydrogen bond in amine boranes [31,32,33], we developed a straightforward method for the synthesis of unsolvated potassium and sodium salts of the [CB₁₁H₁₂]⁻ anion. This method avoids the exchange of cation, significantly simplifies the reaction procedure and can be easily scaled-up.

2. Materials and Methods

2.1. Starting Materials

All manipulations were carried out on a Schlenk line or in a glovebox filled with high-purity nitrogen. Dry THF and DME were obtained by distillation from Na/benzophenone. K[B₁₁H₁₄] and Na[B₁₁H₁₄] were purchased from ZhengzhouYuanli technology. Potassium hydride, (trifluoromethyl)trimethylsilane, sodium hydride, sodium bis(trimethylsilyl)amide, tetrahydrofuran, 1,2-dimethoxyethane, DMSO-d₆ (D, 98%) were purchased from Energy Chemicals, Aladdin, Heowns or Royaltech. The ¹H and ¹H{¹¹B} NMR spectra were obtained using a Bruker Advance NEO 400 MHz instrument from Germany. The ¹³C NMR spectra were recorded at 101 Hz. The ¹¹B, ¹¹B{¹H} and ¹¹B-¹¹B cosy NMR spectra were recorded at 128 MHz. All ¹¹B chemical shifts were referenced to BF₃·OEt₂ in C₆D₆ (0.0 ppm), with a negative sign indicating an upfield shift. All ¹H chemical shifts were measured relative to internal residual hydrogens from the lock solvents (98% DMSO-d₆).

2.2. Synthesis of Na[CB₁₁H₁₂]

Na[B₁₁H₁₄] (1.58 g, 10 mmol), NaH (0.72 g, 30 mmol) and NaHMDS (1.84 g, 10 mmol) were added to a 100 mL Schlenk flask, which was equipped with a reflux condenser. The flask was connected with a Schlenk line and 40 mL 1,2-dimethoxyethane was injected. The reaction mixture was stirred under 0 °C for 15 min and (trifluoromethyl)trimethylsilane (4 mL, 30 mmol) was added. Then, the mixture was stirred at 60 °C for 3 days and a large amount of yellow precipitate was generated. After cooling down to room temperature and quenched with water (1 mL). The residue was subjected to extractive workup with ether (4 × 60 mL) and H₂O (4 × 60 mL). The organic phases were combined, and the solvent was evaporated to give a yellow oily product. Then, 1,4-dioxane was added into the oil residue until a large amount of white solid precipitate. The precipitate was filtered and then dried under dynamic vacuum to produce a 1,4-dioxane solvated Na[CB₁₁H₁₂] white powder. Solvent free Na[CB₁₁H₁₂] (1.14 g, 68% yield) was obtained by dissolving in 20 mL water and drying it first in a rotary evaporator and then under dynamic vacuum at 100 °C for 2 h. ¹¹B NMR (128 MHz, DMSO-d₆) δ −6.95 (d, J = 140.0 Hz, 1B), −13.27 (d, J = 136.3 Hz, 5B), −16.15 (d, J = 150.6 Hz, 5B). ¹¹B{¹H} NMR (128 MHz, DMSO-d₆) δ −7.01 (1B), −13.27 (5B), −16.16 (5B). ¹H NMR (400 MHz, DMSO-d₆) δ 2.36 (s, 1H), 2.24–0.70 (m, 11H). ¹H{¹¹B} NMR (400 MHz, DMSO-d₆) δ 2.36 (s, 1H), 1.55 (s, 6H), 1.40 (s, 5H). ¹³C{¹H} NMR (101 MHz, DMSO-d₆) δ 50.65 (s).

2.3. Synthesis of K[CB₁₁H₁₂]

K[B₁₁H₁₄] (1.74 g, 10 mmol) and KH (1.20 g, 30 mmol) were added to a 100 mL Schlenk flask. The flask was connected with a Schlenk line and 40 mL of tetrahydrofuran was injected. The reaction mixture was stirred under 0 °C for 15 min and then (trifluoromethyl)trimethylsilane (4 mL, 30 mmol) was added. The Schlenk flask was equipped with a reflux condenser. The mixture was stirred at 60 °C for 3 days and a large amount of yellow precipitate was generated. After cooling down to room temperature and quenched with water (1 mL), the residue was subjected to extractive workup with ether (4 × 60 mL) and H₂O (4 × 60 mL). The organic phases were combined, and the solvent was removed under reduced pressure to give a yellow oily product. Then, 1,4-dioxane was added into the filtrate until a large amount of white solid precipitate. The precipitate was filtered and then dried under dynamic vacuum to produce 1,4-dioxane solvated K[CB₁₁H₁₂] white powder. The precipitate was washed with dichloromethane (3 × 30 mL) and hexane (3 × 30 mL), and then dried under dynamic vacuum to produce an unsolvated K[CB₁₁H₁₂], white powder (1.21 g, 66% yield). ¹¹B NMR (128 MHz, DMSO-d₆) δ −6.95 (d, J = 140.0 Hz, 1B), −13.27 (d, J = 136.3 Hz, 5B), −16.15 (d, J = 150.6 Hz, 5B). ¹¹B{¹H} NMR (128 MHz, DMSO-d₆) δ −7.01 (s), −13.27 (s), −16.16 (s). ¹H NMR (400 MHz, DMSO-d₆) δ 2.36 (s, 1H), 2.24–0.70 (m, 11H). ¹H{¹¹B} NMR (400 MHz, DMSO-d₆) δ 2.36 (s, 1H), 1.55 (s, 6H), 1.40 (s, 5H). ¹³C NMR (101 MHz, DMSO-d₆) δ 50.69 (s) (in Supplementary Materials).

3. Results and Discussion

The [CB₁₁H₁₂]⁻ anion was discovered by Knoth using toxic decaborane B₁₀H₁₄ as starting material in 1967 [2]. Since B₁₀H₁₄ was synthesized from NaBH₄ via the [B₁₁H₁₄]⁻ ion, to obviate the toxicity and tedious synthetic steps, the group of Michl [19] developed an alternative approach, using [B₁₁H₁₄]⁻ as a starting material and dichlorocarbene as the carbon source, but in a poor yield. On the basis of this innovative method, Kütt’s group reported an improved method to obtain [Me₃NH][CB₁₁H₁₂] from boron cluster [B₁₁H₁₄]⁻ in up to 95% yield using difluorocarbene as the carbon source [21]. In 2021, the research group of Mark Paskevicius reported a cost-effective method to synthesize the [CB₁₁H₁₂]⁻ anion in 40% yield from [B₁₁H₁₄]⁻ using common laboratory reagents [22]. Generally speaking, the synthetic procedure of [CB₁₁H₁₂]⁻ from [B₁₁H₁₄]⁻ includes four steps represented by Equations (1)–(4), as reported in the literature, in which the cation exchange reactions are carried out (Equations (1) and (4)) considering the solubility of different salts. It is known that Na₂[B₁₁H₁₃] can be formed by the deprotonation of [Me₃NH][B₁₁H₁₄] with NaOH or NaH. The formed Me₃N must be removed because the presence of NMe₃ has previously been found to cause the formation of an excessive amount of 2-Me₃NCB₁₁H₁₁ byproduct (Equation (2)) [19]. The Na₂[B₁₁H₁₃] can then be used to synthesize the [CB₁₁H₁₂]⁻ anion, by insertion of :CCl₂ or :CF₂ sourced from CHCl₃ or CF₃SiMe₃ (Equation (3)). However, because of the limitation in the availability of suitable purification, the ion exchange through the metathesis reaction of Na[CB₁₁H₁₂] with Me₃N·HCl is carried out at room temperature in the aqueous medium (Equation (4)).

Na[B₁₁H₁₄] (aq) + Me₃N·HCl (s) → [Me₃NH][B₁₁H₁₄] (s) + NaCl (s)

(1)

[Me₃NH][B₁₁H₁₄] (s) + NaH (s) or NaOH (s) → Na₂[B₁₁H₁₃] (aq) + H₂ or H₂O + NMe₃

(2)

Na₂[B₁₁H₁₃] (aq) + NaH (s) or NaOH (s) + :CX₂ → Na[CB₁₁H₁₂] (aq)

(3)

Na[CB₁₁H₁₂] (aq) + Me₃N·HCl (s) → [Me₃NH][CB₁₁H₁₂] (s) + NaCl (s)

(4)

Considering the great importance of the alkali metal salts of the [CB₁₁H₁₂]⁻ anion in the fields of catalysis and all-solid-state batteries, we developed a simple, quick, and one-step procedure to directly synthesize unsolvated M[CB₁₁H₁₂] using M[B₁₁H₁₄] (M = K, Na) as starting materials. The overall yields of M[CB₁₁H₁₂] were up to 66–68%. M[B₁₁H₁₄] is commercially available and we first screened different bases with K[B₁₁H₁₄] as substrate. The results showed that KH was the best choice. Then, we examined different molar ratios of K[B₁₁H₁₄] to KH (

Table 1. Screening of the molar ratio of K[B₁₁H₁₄] to KH ^a.

Entry	K[B11H14]:KH	t/d	Yield (%)
1	1:1	3	no product
2	1:2	3	58
3	1:3	3	66
4	1:4	3	65
5	1:5	3	64
6	1:6	3	66

^a reaction conditions: K[B₁₁H₁₄] (1 mmol), KH (1–6 mmol), CF₃SiMe₃ (3 mmol) in THF.

). It was found that the optimized ratio of K[B₁₁H₁₄] to KH to CF₃SiMe₃ was 1:3:3 in the overall preparation of K[CB₁₁H₁₂].

On the other hand, we attempted to synthesize Na[CB₁₁H₁₂] under similar conditions but failed. By using the same molar ratio of Na[B₁₁H₁₄]:NaH:CF₃SiMe₃ = 1:3:3, no reaction was observed and the starting material was monitored by ¹¹B NMR spectroscopy. The reaction could not be observed even if we increased the amount of NaH up to six times. These results indicated that NaH could not abstract hydrogen from Na[B₁₁H₁₄] to form the Na₂[B₁₁H₁₃] intermediate in this reaction. The observation further revealed that the reaction of Na[B₁₁H₁₄] with NaH was different from that of [Me₃NH][B₁₁H₁₄] with NaH in which the deprotonation occurred. Thus, it was concluded that the counter cation influenced the reactions of the salts of the [B₁₁H₁₄]⁻ anion. Then, we examined the reactivity of NaHMDS in this reaction in consideration with its strong basicity and good solubility in organic solvent. However, when three equivalent NaHMDS were used instead of NaH in the reaction, the decomposition of the boron cage was observed but no product was obtained. If only one equivalent NaHMDS was used in the reaction, Na₂[B₁₁H₁₃] was monitored by ¹¹B NMR. Based on these results, we speculated that NaHMDS could abstract hydrogen from Na[B₁₁H₁₄] to form Na₂[B₁₁H₁₃], but NaHMDS would continually react with the formed Na₂[B₁₁H₁₃], resulting in a decomposition to unidentified small boron cages. In considering that NaH can react with CF₃SiMe₃ to provide carbene:CF₂, thus, we selected the combination of NaHMDS and NaH, and screened the different molar ratios of Na[B₁₁H₁₄] to NaHMDS to NaH (

Table 2. Screening of the molar ratio of Na[B₁₁H₁₄] to base ^a.

Entry	Na[B11H14]:NaHMDS:NaH	t/d	Yield (%)
1	1:3:0	3	0 b
2	1:1:0	3	0 c
3	1:1:1	3	33
4	1:1:2	3	58
5	1:1:3	3	68
6	1:1:4	3	64
7	1:1:5	3	64
8	1:1:6	3	66

^a reaction conditions: Na[B₁₁H₁₄] (1 mmol), NaHMDS (1–3 mmol), NaH (0–6 mmol), CF₃SiMe₃ (3 mmol) in DME. ^b NaB₁₁H₁₄ degrade. ^c Na₂B₁₁H₁₃ was formed.

). After a series of screenings, the 1:1:3:3 molar ratio of Na[B₁₁H₁₄] to NaHMDS to NaH to CF₃SiMe₃ was used in the overall preparation of Na[CB₁₁H₁₂]. Furthermore, it is worth noting that when we examined DME as a solvent, the yields were relatively higher than those in THF, probably because the slight polarity of DME. Therefore, the reactions of sodium salts were carried out in DME, differentiating from the reaction of potassium salts.

In summary, the highly pure potassium and sodium salts of the [CB₁₁H₁₂]⁻ anion can be directly synthesized with M[B₁₁H₁₄] as starting materials (M = K, Na) (Figure 1) without converting the alkali cation into amine by a cation exchange reaction. The ¹¹B and ¹¹B{¹H} NMR spectra of the reaction mixture indicated that the reaction highly selectively converted to the [CB₁₁H₁₂]⁻ anion as shown in Figure 1a,b,e,f. These results make the purification of the crude product simple and efficient.

4. Conclusions

We developed a simple and effective method to directly synthesize the unsolvated potassium and sodium salts of the [CB₁₁H₁₂]⁻ anion in one step with 66–68% yields. Differentiating from the previous synthesis of [CB₁₁H₁₂]⁻ anion, this method avoided the exchange of cation and significantly simplified the reaction procedure, thus enabling a convenient large-scale synthesis. It paves the way for the application of K[CB₁₁H₁₂] and Na[CB₁₁H₁₂] in many fields such as in solid ionic conductors.