Effects of LiBF4 Addition on the Lithium-Ion Conductivity of LiBH4

Complex hydrides, such as LiBH4, are a promising class of ion conductors for all-solid-state batteries, but their application is constrained by low ion mobility at room temperature. Mixing with halides or complex hydride anions, i.e., other complex hydrides, is an effective approach to improving the ionic conductivity. In the present study, we report on the reaction of LiBH4 with LiBF4, resulting in the formation of conductive composites consisting of LiBH4, LiF and lithium closo-borates. It is believed that the in-situ formation of closo-borate related species gives rise to highly conductive interfaces in the decomposed LiBH4 matrix. As a result, the ionic conductivity is improved by orders of magnitude with respect to the Li-ion conductivity of the LiBH4, up to 0.9 × 10−5 S cm−1 at 30 °C. The insights gained in this work show that the incorporation of a second compound is a versatile method to improve the ionic conductivity of complex metal hydrides, opening novel synthesis pathways not limited to conventional substituents.


Introduction
In our current society, rechargeable lithium-ion batteries are crucial energy storage devices for a wide variety of applications, ranging from small portable electronics to electric vehicles. The performance and safety requirements for automotive electrification and grid-energy storage systems, however, cannot be met by the commercially available batteries [1,2]. Hence, the development of batteries with higher energy densities and improved safety is crucial. The performance limitations in commercial Li-ion batteries are caused by their organic liquid electrolytes that often suffer from poor electrochemical and thermal stability, which poses safety risks and limits storage capacity [3]. Replacement of the liquid electrolyte with a solid electrolyte could overcome these persistent problems. Solid-state electrolytes (SSEs) are inherently safer and, in many cases, compatible with high energy density electrodes, such as metallic lithium [4]. Consequently, the realization of allsolid-state batteries implementing SSEs can lead to safer batteries that store more energy.
The anticipated benefits of SSEs has led to a growing number of studies on inorganic lithium ion conductors [5], such as garnet-type [6,7], perovskite-type [8,9] and sulphidetype materials [10,11] The research scope on potential Li-ion conductors was recently extended to complex metal hydrides following the surprising discovery of fast Li-ion mobility in LiBH 4 [12]. Complex metal hydrides exhibit several unique material properties that are interesting for battery application, i.e., low molecular weight, easy deformability and high compressibility, and compatibility with a lithium metal anode [13][14][15][16]. Among others, these benefits have spurred research on complex metal hydrides as promising SSEs for all-solid-state Li-ion batteries.

Structural Characterization
X-ray powder diffraction (XRD) analysis was performed with a Bruker-AXS D8 Advance X-ray diffractometer (Co Kα 1 = 1.78897 Å). The samples were placed in an airtight sample holder, preventing air and moisture exposure. Diffractograms were recorded at RT from 20 • to 80 • 2θ range.
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) measurements were performed on a Perkin-Elmer 2000 spectrometer equipped with an MCT detector. Using an airtight sample holder with KBr windows, samples were measured without any air or moisture exposure. Spectra were acquired from 900 cm −1 to 4500 cm −1 with a resolution of 4 cm −1 .

Electrochemical Impedance Spectroscopy
Electrochemical impedance spectroscopy (EIS) measurements were performed using a Princeton Applied Research Parstat 2273 connected to a custom-made measurement cell in a Büchi B-585 glass oven. Two stainless steel cylinders with Li foil (Sigma-Aldrich, purity 99.9%, 0.38 mm thick) were placed on one side and configured as electrodes. Using a standard 13 mm pellet press, an approximately 80-150 mg sample was pressed between the electrodes with a pressure of approximately 1.5 ton cm −2 . During a typical conductivity measurement, the measurement cell was incrementally heated to 130 • C (∆T = 10 • C) and incrementally cooled to room temperature (∆T = 20 • C). At each increment the temperature was allowed to equilibrate before an EIS measurement was performed using a 10 mV of amplitude and with frequencies from 1 MHz to 1 Hz.

Linear Sweep Voltammetry
Linear sweep voltammetry (LSV) was used to determine the oxidative stability of the LiBH 4 -20% LiBF 4 sample. The sample was mixed with carbon black (CB, Ketjenblack EC600JD) in a weight ratio of 90:10 using an agate mortar. A pellet was obtained by pressing approximately 5 mg of electrolyte-carbon mixture and 50 mg of the electrolyte at 1.5 ton cm −2 . After compressing, a cell was formed with stainless-steel as the working electrode in contact with the electrolyte-carbon mixture and a lithium disk as the counter electrode in contact with the solid electrolyte. The linear sweep measurement was performed at 60 • C with a voltage range from 1.0 V to 5.0 V versus Li + /Li at a scanning rate of 0.1 mVs −1 .

Structural Characterization and Conductivity of LiBH 4 -LiBF 4 Mixtures
The impact of the addition of LiBF 4 on the crystal structure of the Li-B-H complexes has been determined using XRD. In Figure 1a, the diffractograms of a mixture consisting of 80 mol% of LiBH 4 and 20 mol% LiBF 4 prepared via solid-state reaction at 150 • C and 280 • C are shown. The XRD patterns of LiBH 4 and LiBF 4 are provided for comparison. In the diffraction pattern of the LiBH 4 -LiBF 4 mixture heated at 150 • C, the main diffraction peaks characteristic of both starting compounds are still observed. Additionally, two new peaks have appeared at 45 and 53 2θ • , that correspond to cubic (Fm-3m) LiF. This can be explained considering that the decomposition of LiBH 4 in the presence of LiBF 4 starts at 80 • C and LiF has been reported as a decomposition product [36]. During the reaction, hydrogen-fluorine exchange results in the formation of LiBH 4−x F x [36], which is not stable and decomposes to form LiF.  [36]. During the reaction, hydrogen-fluorine exchange results in the formation of LiBH4−xFx [36], which is not stable and decomposes to form LiF. Similarly, the diffraction pattern of the LiBH4-LiBF4 mixture prepared at 280 °C predominantly contains peaks related to LiF; however, in this case no clear diffraction peaks of the starting compounds can be identified. Thus, at a reaction temperature of 280 °C, the decomposition reaction proceeds further in agreement with results reported by Richter et al. [36] Note that with XRD only crystalline compounds can be identified. Therefore, it is not clear whether LiF is the only solid decomposition product or if an unidentified amorphous phase is formed as well. For example, Zhu et al. reported that highly conductive amorphous Li-B-H complexes can form during the LiBH4 decomposition [38].
Because with the XRD analysis the presence of possible amorphous Li-B-H complexes could not be confirmed, the structural properties of the LiBH4-LiBF4 mixtures were further analysed using DRIFTS measurements. This technique is not limited to crystalline compounds, which means that it can be used to identify the composition of the crystalline and amorphous phases present in the LiBH4-LiBF4 mixtures. The DRIFTS absorbance spectra obtained for LiBH4-LiBF4 synthesized at 150 °C and 280 °C are shown in Figure 1b, as well as the spectra for pristine LiBH4 and LiBF4. Pristine LiBH4 typically shows several characteristic absorption bands. The most predominant bands, corresponding to [BH4 − ] stretching vibrations, can be seen between 2000 and 2800 cm −1 . Other characteristic LiBH4 vibrations can be seen between 1000 and 1300 cm −1 ([BH4 − ] bending) and 3200 and 3700 cm −1 ([OH] stretching from adsorbed moisture) [40][41][42]. Though less intense, these characteristic bands are seen in the spectra of both LiBH4-LiBF4 mixtures, indicating that LiBH4 is still present in both mixed compounds, hence the starting compounds have not completely decomposed. Note that the bands related to LiBH4 vibrations are especially pronounced in the DRIFTS spectrum of LiBH4-LiBF4 prepared at 280 °C, even though no peaks related to crystalline LiBH4 were present in the diffractogram, suggesting that residual amorphous LiBH4 is present in the compound.
Besides the LiBH4 vibrations, multiple unidentified peaks are present in both LiBH4-LiBF4 spectra, specifically around 2500 cm −1 and at 1800 cm −1 . Surprisingly, none of these Similarly, the diffraction pattern of the LiBH 4 -LiBF 4 mixture prepared at 280 • C predominantly contains peaks related to LiF; however, in this case no clear diffraction peaks of the starting compounds can be identified. Thus, at a reaction temperature of 280 • C, the decomposition reaction proceeds further in agreement with results reported by Richter et al. [36] Note that with XRD only crystalline compounds can be identified. Therefore, it is not clear whether LiF is the only solid decomposition product or if an unidentified amorphous phase is formed as well. For example, Zhu et al. reported that highly conductive amorphous Li-B-H complexes can form during the LiBH 4 decomposition [38].
Because with the XRD analysis the presence of possible amorphous Li-B-H complexes could not be confirmed, the structural properties of the LiBH 4 -LiBF 4 mixtures were further analysed using DRIFTS measurements. This technique is not limited to crystalline compounds, which means that it can be used to identify the composition of the crystalline and amorphous phases present in the LiBH 4 -LiBF 4 mixtures. The DRIFTS absorbance spectra obtained for LiBH 4 -LiBF 4 synthesized at 150 • C and 280 • C are shown in Figure 1b, as well as the spectra for pristine LiBH 4 and LiBF 4 . Pristine LiBH 4 typically shows several characteristic absorption bands. The most predominant bands, corresponding to [BH 4 − ] stretching vibrations, can be seen between 2000 and 2800 cm −1 . Other characteristic LiBH 4 vibrations can be seen between 1000 and 1300 cm −1 ([BH 4 − ] bending) and 3200 and 3700 cm −1 ([OH] stretching from adsorbed moisture) [40][41][42]. Though less intense, these characteristic bands are seen in the spectra of both LiBH 4 -LiBF 4 mixtures, indicating that LiBH 4 is still present in both mixed compounds, hence the starting compounds have not completely decomposed. Note that the bands related to LiBH 4 vibrations are especially pronounced in the DRIFTS spectrum of LiBH 4 -LiBF 4 prepared at 280 • C, even though no peaks related to crystalline LiBH 4 were present in the diffractogram, suggesting that residual amorphous LiBH 4 is present in the compound.
Besides the LiBH 4 vibrations, multiple unidentified peaks are present in both LiBH 4 -LiBF 4 spectra, specifically around 2500 cm −1 and at 1800 cm −1 . Surprisingly, none of these peaks is related to the presence of LiBF 4 or LiF ( Figure 1b) [40,43]. Clearly, decomposition of the unstable LiBH 4−x F x complex does not only result in the formation of LiF, but also in an unidentified amorphous compound. Similar to the Li-B-H complexes obtained during LiBH 4 decomposition described by Zhu et al., this unidentified compound could be beneficial for the overall conductivity of the LiBH 4 -LiBF 4 compounds [38]. Therefore, the nature of this compound has been studied in more detail, which will be discussed later in this work.
The impact of the decomposition reaction on Li-ion conductivity in the prepared LiBH 4 -LiBF 4 mixed compounds is determined using electrochemical impedance spectroscopy (EIS). The obtained Nyquist plots display single semicircles that could be analysed with a simple one-phase model, as is shown in Figure S1. Additionally, exemplary Bode plots corresponding to the data are shown in Figure S2. In this way, the resistance and corresponding conductivity are determined over a temperature range from room temperature to 130 • C. In Figure 1c and Figure S3, the temperature dependence of the ionic conductivity of the LiBH 4 -LiBF 4 mixtures prepared at 150 • C and 280 • C is displayed using an Arrhenius plot. Comparison to the ionic conductivity of LiBH 4 shows that ion transport in the LiBH 4 -LiBF 4 mixtures prepared at 280 • C has improved at temperatures lower than 100 • C, reaching 0.9 × 10 −5 S cm −1 at 30 • C, which is three orders of magnitude higher than that of LiBH 4 . In contrast, the mixture prepared with a lower reaction temperature exhibits a conductivity lower than that of LiBH 4 over the entire temperature range.
At temperatures above 110 • C, the same temperature-dependent Li-ion conductivity behaviour as LiBH 4 has been measured for the mixture treated at 150 • C, i.e., a drastic increase of the Li-ion conductivity. This behaviour indicates that the sample contains unreacted LiBH 4 that undergoes a phase transition from the orthorhombic to the hexagonal phase, increasing the Li-ion conductivity. These results agree with both the diffraction and IR data reported in Figure 1. The lower ion conductivity with respect to pure LiBH 4 can be explained considering that the sample also contains non-conductive phases, i.e., unreacted LiBF 4 and LiF. The LiBH 4 -LiBF 4 mixture prepared at 280 • C also exhibits a conductivity enhancement above 110 • C related to the phase transition of LiBH 4 , though it is not as pronounced. The conductivity above 110 • C is lower than that of pristine LiBH 4 , which again can be attributed to the presence of non-conductive LiF.
The difference in conductivity between the mixtures prepared with different reaction temperature (see also Figure S4) indicates that the reaction temperature of the prepared LiBH 4 -LiBF 4 system plays an important role. It is not likely that LiF predominantly contributes to the enhanced conductivity, as it does not exhibit high room temperature conductivity by itself, typically below 10 −7 S cm −1 [44,45]. On the other hand, the incorporation of LiF could lead to the formation of a LiF-substituted LiBH 4 solid solution in which the hexagonal LiBH 4 phase is stabilized at lower temperatures. However, previous studies on halide substituted LiBH 4 demonstrated that the RT stability range of the hexagonal phase decreases with the anionic radius. In fact, it has been reported that h-Li(BH 4 ) 1−α (I) α solid solutions are stable at RT in the range of 0.18 ≤ α ≤ 0.50, [46,47] [29], while Cl − is not able to stabilize the high temperature phase at RT [46]. The ionic size of F − is considerably smaller with respect to BH 4 − as well as to the other halides, i.e., r(F − ) = 1.33 Å, r(Cl − ) = 1.81 Å, r(Br − ) = 1.96 Å, r(I − ) = 2.20 Å and r(BH 4 − ) = 2.03 Å [48], which makes stabilization of the hexagonal phase at ambient temperature unlikely. This hypothesis is further confirmed by the absence of reflections related to hexagonal LiBH 4 in the LiBH 4 -LiBF 4 diffraction patterns (Figure 1a). Consequently, it is not expected that the formation of LiF contributes to conductivity enhancement in the LiBH 4 -LiBF 4 compound.
Nevertheless, it is clear that a highly conductive compound forms after the solid-state reaction occurring at 280 • C, during which hydrogen-fluoride substitution and subsequent decomposition of LiBH 4−x F x occur. Because an improved conductivity is only seen in this case, it seems that the decomposition reaction that results in the presence of amorphous LiBH 4  Their analysis illustrates that the formation of LiBH3F from hydrogen-fluor stitution does not solely result in the formation of LiF. The highly unstable LiBH pound decomposes to form both LiF and diborane (B2H6). Finally, diborane can re an excess of LiBH4 to form Li2B12H12. Because the studied composition consists of 8 of LiBH4, it is possible for this final reaction (Figure 2c) to occur. This decompositio way can support our observations on the effect of hydrogen-fluorine exchange composition of the LiBH4-LiBF4 mixtures together with the enhancement of the Liductivity.

LiBH4-Li2B12H12 Composites
To investigate if the high ionic conductivity is related to the formation of L like or LiBH4-Li2B12H12 composites during the decomposition reaction, mixtures o with 10 to 75 mol% Li2B12H12 (see Table S1) were prepared following the syntheti dure used for the mixture LiBH4-LiBF4 synthetized at 280 °C. The composition b 25 mol% Li2B12H12 will be discussed below because it resembles the compositio LiBH4-LiBF4 mixture. The DRIFTS and XRD spectra of the LiBH4-25% Li2B12H12 prepared are shown in Figure 3, together with those obtained from the LiBH4-Li tem, while in Figure S5 the other LiBH4-Li2B12H12 samples are presented. In Figure  DRIFTS spectra of both LiBH4-LiBF4 and LiBH4-Li2B12H12 contain strong absorptio at 2535 cm −1 and 2480 cm −1 . The previously unidentified peaks in the LiBH4-LiBF4 sp can thus be associated with the presence of Li2B12H12, which displays characterist stretching vibrations exactly at these wavenumbers [49][50][51]. Other similarities DRIFTS spectra of LiBH4-LiBF4 and LiBH4-Li2B12H12 are also observed. The com reveals that, in addition to the pronounced bands around 2500 cm −1 , smaller pe tween 3700 and 3000 cm −1 , at 1835 cm −1 and at 1015 cm −1 , can also be explained presence of Li2B12H12 [49][50][51]. Together with the XRD results discussed previously 1b), this illustrates that the prepared LiBH4-LiBF4 mixtures decompose to form cry LiF and likely amorphous closo-borate-related compounds with residual LiBH4.

LiBH 4 -Li 2 B 12 H 12 Composites
To investigate if the high ionic conductivity is related to the formation of Li 2 B 12 H 12 -like or LiBH 4 -Li 2 B 12 H 12 composites during the decomposition reaction, mixtures of LiBH 4 with 10 to 75 mol% Li 2 B 12 H 12 (see Table S1) were prepared following the synthetic procedure used for the mixture LiBH 4 -LiBF 4 synthetized at 280 • C. The composition based on 25 mol% Li 2 B 12 H 12 will be discussed below because it resembles the composition of the LiBH 4 -LiBF 4 mixture. The DRIFTS and XRD spectra of the LiBH 4 -25% Li 2 B 12 H 12 sample prepared are shown in Figure 3, together with those obtained from the LiBH 4 -LiBF 4 system, while in Figure S5 the other LiBH 4 -Li 2 B 12 H 12 samples are presented. In Figure 3a, the DRIFTS spectra of both LiBH 4 -LiBF 4 and LiBH 4 -Li 2 B 12 H 12 contain strong absorption bands at 2535 cm −1 and 2480 cm −1 . The previously unidentified peaks in the LiBH 4 -LiBF 4 spectrum can thus be associated with the presence of Li 2 B 12 H 12 , which displays characteristic [B-H] stretching vibrations exactly at these wavenumbers [49][50][51]. Other similarities in the DRIFTS spectra of LiBH 4 -LiBF 4 and LiBH 4 -Li 2 B 12 H 12 are also observed. The comparison reveals that, in addition to the pronounced bands around 2500 cm −1 , smaller peaks between 3700 and 3000 cm −1 , at 1835 cm −1 and at 1015 cm −1 , can also be explained by the presence of Li 2 B 12 H 12 [49][50][51]. Together with the XRD results discussed previously (Figure 1b  This assertion is further investigated by comparing the XRD patterns of LiBH4-LiBF4 to that of LiBH4-Li2B12H12 shown in Figure 3b. Notably, the pattern collected for the LiBH4-Li2B12H12 contains characteristic diffraction peaks related to both starting compounds, which clearly demonstrates that both phases are still present and crystalline. This is different from the pattern of LiBH4-LiBF4, where, besides LiF, no other crystalline phase is detected. Note that a small peak indicative of Li2B12H12 can be observed at 24 2θ° in the LiBH4-LiBF4 mixture pattern, strengthening the possibility that Li2B12H12 has formed. However, in general it seems that the reaction between LiBF4 and LiBH4 leads to partial decomposition forming crystalline LiF and amorphous Li2B12H12. Several studies have shown that the incorporation of Li2B12H12 in the LiBH4 matrix could lead to enhanced lithium-ion mobility [38,[52][53][54]. Similar to the Li-B-H compound proposed by Zhu et al. [38], after reaction the LiBH4-LiBF4 mixture consists of residual LiBH4 in contact with LiF and amorphous Li2B12H12. Accordingly, the conductivity enhancement could originate from the in-situ formation of Li2B12H12. To further investigate this hypothesis, electrochemical impedance spectroscopy was used to determine the conductivity of LiBH4-Li2B12H12. Interestingly, in Figures 3c and S6, it can be seen that the conductivity of LiBH4-Li2B12H12 (0.5 × 10 −6 S cm −1 at RT) is improved compared to both pristine LiBH4 and Li2B12H12 (2.5 × 10 −8 S cm −1 at RT) [55]. The improvement in ionic conductivity can be attributed again to a partial LiBH4 decomposition, to form some amorphous Li2B12H12 or [Li2B12H11+1/n]n and undecomposed LiBH4 (see also next paragraph). Consequently, the improved cation mobility observed in the LiBH4-LiBF4 mixture prepared at 280 °C can be attributed to the formation of Li2B12H12 and its interaction with residual LiBH4, while the presence of LiF likely has a minor or no impact. Table 1 shows the values of the activation energies (Ea) obtained by a linear fit of the Arrhenius plot. The Ea of the orthorhombic LiBH4 is in good agreement with the calculated average value using all values reported in the literature by Gulino et al. [56], i.e., 0.75 ± 0.07 eV. The Ea obtained for the mixtures LiBH4-LiBF4 and LiBH4-Li2B12H12 are lower compared to that of pure LiBH4, indicating a more facile conduction mechanism, in agreement with the higher Li-ion conductivity. In addition, these values are in agreement with the data reported by Zhu et al. [38], e.g., 0.43 and 0.44 eV for the samples containing amorphous Li2B12H12 or [Li2B12H11+1/n]n, further indicating that the highly conductivity phase formed during reaction with LiBF4 could be amorphous Li2B12H12 or a related compound. It is also good to note that the oxidative stability of 4.0 V determined for the LiBH4-LiBF4 mixture ( Figure S7) is similar to the value reported for other closo-borate-related compounds [13,38,57]. This assertion is further investigated by comparing the XRD patterns of LiBH 4 -LiBF 4 to that of LiBH 4 -Li 2 B 12 H 12 shown in Figure 3b. Notably, the pattern collected for the LiBH 4 -Li 2 B 12 H 12 contains characteristic diffraction peaks related to both starting compounds, which clearly demonstrates that both phases are still present and crystalline. This is different from the pattern of LiBH 4 -LiBF 4 , where, besides LiF, no other crystalline phase is detected. Note that a small peak indicative of Li 2 B 12 H 12 can be observed at 24 2θ • in the LiBH 4 -LiBF 4 mixture pattern, strengthening the possibility that Li 2 B 12 H 12 has formed. However, in general it seems that the reaction between LiBF 4 and LiBH 4 leads to partial decomposition forming crystalline LiF and amorphous Li 2 B 12 H 12 . Several studies have shown that the incorporation of Li 2 B 12 H 12 in the LiBH 4 matrix could lead to enhanced lithium-ion mobility [38,[52][53][54]. Similar to the Li-B-H compound proposed by Zhu et al. [38], after reaction the LiBH 4 -LiBF 4 mixture consists of residual LiBH 4 in contact with LiF and amorphous Li 2 B 12 H 12 . Accordingly, the conductivity enhancement could originate from the in-situ formation of Li 2 B 12 H 12 .
To further investigate this hypothesis, electrochemical impedance spectroscopy was used to determine the conductivity of LiBH 4 -Li 2 B 12 H 12 . Interestingly, in Figure 3c and Figure S6, it can be seen that the conductivity of LiBH 4 -Li 2 B 12 H 12 (0.5 × 10 −6 S cm −1 at RT) is improved compared to both pristine LiBH 4 and Li 2 B 12 H 12 (2.5 × 10 −8 S cm −1 at RT) [55]. The improvement in ionic conductivity can be attributed again to a partial LiBH 4 decomposition, to form some amorphous Li 2 B 12 H 12 or [Li 2 B 12 H 11+1/n ] n and undecomposed LiBH 4 (see also next paragraph). Consequently, the improved cation mobility observed in the LiBH 4 -LiBF 4 mixture prepared at 280 • C can be attributed to the formation of Li 2 B 12 H 12 and its interaction with residual LiBH 4 , while the presence of LiF likely has a minor or no impact. Table 1 shows the values of the activation energies (E a ) obtained by a linear fit of the Arrhenius plot. The E a of the orthorhombic LiBH 4 is in good agreement with the calculated average value using all values reported in the literature by Gulino et al. [56], i.e., 0.75 ± 0.07 eV. The E a obtained for the mixtures LiBH 4 -LiBF 4 and LiBH 4 -Li 2 B 12 H 12 are lower compared to that of pure LiBH 4 , indicating a more facile conduction mechanism, in agreement with the higher Li-ion conductivity. In addition, these values are in agreement with the data reported by Zhu et al. [38], e.g., 0.43 eV and 0.44 eV for the samples containing amorphous Li 2 B 12 H 12 or [Li 2 B 12 H 11+1/n ] n , further indicating that the highly conductivity phase formed during reaction with LiBF 4 could be amorphous Li 2 B 12 H 12 or a related compound. It is also good to note that the oxidative stability of 4.0 V determined for the LiBH 4 -LiBF 4 mixture ( Figure S7) is similar to the value reported for other closo-boraterelated compounds [13,38,57]. Table 1. Activation energies (eV) calculated from a linear plot of ln(σT) and 1000/T of the EIS data reported in Figure 3c. The standard deviation is based on the 99% confidence interval of the linear fit.

Alternative Decomposition Pathway
Although it seems clear that the formation of a Li 2 B 12 H 12 -like compound is responsible for the enhancement in ionic conductivity for LiBH 4 -LiBF 4 , this could not explain the behaviour of the system completely. Surprisingly, comparison of the respective conductivities of LiBH 4 -Li 2 B 12 H 12 and LiBH 4 -LiBF 4 mixtures (Figure 3c) illustrates that the latter has the highest conductivity. If the conductivity enhancement is solely related to the formation of Li 2 B 12 H 12 , this would not be the case. However, as our structural analysis did not yield any other compounds or phases, the unidentified conductive phase is likely very similar to amorphous Li 2 B 12 H 12 .
Interestingly, it has been proposed that the combination of different closo-borate complex anions could lead to improved ion transport in metal hydride-based ion conductors [38,54,58,59]. Toyama [60,61]. The in-situ generated B 10 H 14 can then react with LiBH 4 to form several closo-borate anions, similar to the synthesis performed by Toyama et al. [54] The reaction scheme discussed here is presented in Figure 4.  Table 1. Activation energies (eV) calculated from a linear plot of ln(σT) and 1000/T of the EIS data reported in Figure 3c. The standard deviation is based on the 99% confidence interval of the linear fit.

Alternative Decomposition Pathway
Although it seems clear that the formation of a Li2B12H12-like compound is responsible for the enhancement in ionic conductivity for LiBH4-LiBF4, this could not explain the behaviour of the system completely. Surprisingly, comparison of the respective conductivities of LiBH4-Li2B12H12 and LiBH4-LiBF4 mixtures (Figure 3c) illustrates that the latter has the highest conductivity. If the conductivity enhancement is solely related to the formation of Li2B12H12, this would not be the case. However, as our structural analysis did not yield any other compounds or phases, the unidentified conductive phase is likely very similar to amorphous Li2B12H12.
Interestingly, it has been proposed that the combination of different closo-borate complex anions could lead to improved ion transport in metal hydride-based ion conductors [38,54,58,59]. Toyama [54]. They observed that with increasing amount of [B11H11] 2− and [B10H10] 2− , ionic conductivity improves. Remarkably, under the reaction conditions used during our synthesis, pyrolysis of B2H6 to B10H14 can occur [60,61]. The in-situ generated B10H14 can then react with LiBH4 to form several closo-borate anions, similar to the synthesis performed by Toyama et al. [54] The reaction scheme discussed here is presented in Figure 4. The conductivity enhancement achieved in decomposed LiBH4 in the presence of LiBF4 that forms multiple closo-borate phases would be higher than expected for a mixture consisting solely of LiBH4 and Li2B12H12. Consequently, the reaction process described in Figure 4 provides a possible explanation for an increased conductivity in LiBH4-LiBF4 compared to LiBH4-Li2B12H12 through the introduction of different lithium closo-borate compounds. A detailed investigation of this hypothesis will require solid-state 11 B-MAS-NMR measurements, which is the subject of an ongoing study.
Another explanation for the enhanced conductivity could be the formation of highly conductive interfaces between the amorphous LiBH4-Li2B12H12 solid solution and the insitu formed nanocrystalline LiF particles. The melting of the undecomposed LiBH4 will likely lead to strong interface interactions with LiF and Li2B12H12, thereby enhancing the The conductivity enhancement achieved in decomposed LiBH 4 in the presence of LiBF 4 that forms multiple closo-borate phases would be higher than expected for a mixture consisting solely of LiBH 4 and Li 2 B 12 H 12 . Consequently, the reaction process described in Figure 4 provides a possible explanation for an increased conductivity in LiBH 4 -LiBF 4 compared to LiBH 4 -Li 2 B 12 H 12 through the introduction of different lithium closo-borate compounds. A detailed investigation of this hypothesis will require solid-state 11 B-MAS-NMR measurements, which is the subject of an ongoing study.
Another explanation for the enhanced conductivity could be the formation of highly conductive interfaces between the amorphous LiBH 4 -Li 2 B 12 H 12 solid solution and the in-situ formed nanocrystalline LiF particles. The melting of the undecomposed LiBH 4 will likely lead to strong interface interactions with LiF and Li 2 B 12 H 12 , thereby enhancing the ionic conductivity of the composite. Recent studies by Yan and Zhao reported on similar systems in which LiBH 4 xNH 3 was stabilized in an amorphous (molten-like) state on (inert) Li 2 O nanoparticles that were formed in-situ via a decomposition pathway [62,63]. Even in compounds with a high weight percentage (78 wt%) of the non-conducting metal oxide, a high room temperature conductivity has been obtained. Likewise, in our case, the amorphous LiBH 4 -Li 2 B 12 H 12 phase could benefit from the presence of LiF particles, for example by stabilizing the amorphous form or even by beneficial interface interactions, thereby forming a conductive space-charge layer.

Conclusions
In conclusion, the impact of LiBF 4 addition on the ionic transport in LiBH 4 was explored by correlating the physical properties and ionic conductivity of LiBH 4 -LiBF 4 mixtures. After synthesis at 280 • C, the LiBH 4 -LiBF 4 sample contains LiF and amorphous lithium closo-borate, such as Li 2 B 12 H 12 . The presence of this phase is in line with the fact that partial substitution of the BH 4 anion in LiBH 4 with BF 4 anion yields unstable LiBH 3 F, which decomposes into LiF and B 2 H 6 . The latter may react further to form B 10  Interestingly, a conductivity enhancement of over two orders of magnitude (to 0.9 × 10 −5 S cm −1 at 30 • C) is achieved for the LiBH 4 -LiBF 4 mixture reacted at 280 • C. The formation of LiF does not seem to lead to any improvement in conductivity, while the presence of solely Li 2 B 12 H 12 leads to a minor enhancement. Thus, the improved lithium-ion mobility in LiBH 4 through decomposition reaction with LiBF 4 may be attributed to the formation of a lithium closo-borate-like phase, i.e., amorphous Li 2 B 12 H 12 or [Li 2 B 12 H 11+1/n ] n . The insights provided in this work highlight the versatility of LiBH 4 as a promising solidstate electrolyte, considering that several different strategies can be exploited to increase the ionic conductivity of the compound. Our work demonstrates that decomposition reactions is a promising approach to enhancing ionic conductivity in LiBH 4 . This approach might be applicable to other complex hydrides and other classes of inorganic solid electrolytes.