Investigation on LiBH4-CaH2 composite and its potential for thermal energy storage.

The LiBH4/CaH2 composite are firstly studied as Concentrating Solar Power Thermal Storage Material. The LiBH4/CaH2 composite according to the stoichiometric ratio are synthesized by high-energy ball milling method. The kinetics, thermodynamics and cycling stability of LiBH4/CaH2 composite are investigated by XRD (X-ray diffraction), DSC (Differential scanning calorimeter) and TEM (Transmission electron microscope). The reaction enthalpy of LiBH4/CaH2 composite is almost 60 kJ/mol H2 and equilibrium pressure is 0.482 MPa at 450 °C. The thermal storage density of LiBH4/CaH2 composite is 3504.6 kJ/kg. XRD results show that the main phase after dehydrogenation is LiH and CaB6. The existence of TiCl3 and NbF5 can effectively enhance the cycling perfomance of LiBH4/CaH2 composite, with 6-7 wt% hydrogen capacity after 10 cycles. The high thermal storage density, high working temperature and low equilibrium pressure make LiBH4/CaH2 composite a potential thermal storage material.

Solar energy is the most plentiful renewable and clean alternative to fossil fuels1. International Energy Agency (IEA) points out that solar energy will make up 22 percent of the global electricity, and it is possible that solar photovoltaics (PV) and concentrating solar thermal (CST) power technology will play roughly equal, but complementary roles by 20502. The CST power technology can store energy as heat that can be assessed on demand to generate electricity when PV technology is inefficient, such as at night or during rainy days.

As thermal storage (TS) material is the key element in the CST, improving the energy storage density and working temperature have great value on power generation efficiency and cutting back on the cost. There are three basic methods of thermal storage. Considering the condition of CST plants, sensible heat storage with molten salt is of low-efficiency and can be corrosive sometimes and latent heat storage using NaNO3 is of high flammability and reactivity and is uncertain over its longevity. Thermochemical heat storage materials have quite high energy storage density, in which hydrides’ exceed 1700–4000 kJ kg−1 (10~30 times more than molten salts’ energy storage density and 4~10 times more than phase change materials’ energy storage density)34. The characteristics of common and potential thermal storage materials are listed in Table 1.

Table 1

Energy stored per mass of different storage materials5678.

Material	Sensible		Latent		Thermochemical
	Rock	Concrete	Paraffin wax	NaNO3	CaCl2·H2O	MgH2
Specific heat capacity (kJ kg−1)	0.9	1.13	—	—	3.06	—
Latent heat of fusion (kJ kg−1)	—	—	174.4	172	—	—
Reaction enthalpy (kJ kg−1)	—	—	—	—	433.6	2860

Among all hydrides, complex hydrides such as LiBH4, NaAlH4 and NaBH4 possess a quite high forming enthalpy due to the transition to an ionic or covalent compound of metals upon hydrogen absorption9 It seems that complex hydrides are promising thermal storage materials in CST plants.

In fact, LiBH4 is mostly researched as a hydrogen storage material due to the second highest hydrogen content (18.4 wt.%) of all alanates and boranates10111213. The thermal hydrogen desorption of pure LiBH4 starts at ~320 °C and proceeds mainly in the temperature region 400–600 °C, which is in accordance with the working temperature of thermal storage material in CST plant.

However, the sluggish kinetics and poor reversibility of LiBH4 are the problems that limit its use in hydrogen storage111213. The destabilization was proposed to change the reaction process by adding a reactive additive14. Thus, many new systems have been proposed based on DFT calculation of reaction enthalpies in multi-component systems1516. Among all systems, LiBH4-CaH2 is one of promising composites that are suitable for thermal storage due to its onset temperature. And this reaction can produce around 11.7 wt% hydrogen. But Yang reported that LiBH4-CaH2 composite is irreversible under the condition tested (350 °C, 150 bar)17. The sluggish kinetics in LiBH4-CaH2 composite is another problem that need to be solved18.

Additives, such as TiCl31920212223, V2O519, TiF31920, TiO21920, LiNH219, NbF52425, NbCl524 were investigated their effects on the kinetics and cycling performance of LiBH4-CaH2 composite. Results show that TiCl3 additive is very effective in lowering activation energy of dehydrogenation and enhancing reversibility (about 9 wt% hydrogen reversibly)1920212223. Besides, Lim reported that LiBH4-CaH2 composite with NbF5 maintains a reversible hydrogen storage capacity of about 6 wt% at 450 °C25.

But the enthalpies of pure LiBH4-CaH2 composite and LiBH4-CaH2 composite with additives have not been reported exactly. Pinkerton21 reported the estimated enthalpy of reaction 2 at 400 °C is ΔH = 59.2 kJ/mol H2. Lim24 reported that the reaction enthalpy of LiBH4-CaH2-0.2NbF5 composite is estimated to be 56.5 kJ/mol H2 at 305 °C. But according to HSC thermochemical database21, the enthalpy is 66.2 kJ/mol H2 at room temperature.

In this research, the enthalpy ΔH and equilibrium pressure of desorption according to reaction 2 are determined by PCT (pressure, concentration, and temperature) measurements. Only LiBH4-CaH2 without catalysts was measured due to the fact that influence the thermodynamics of the mixture as Ti-doped NaAlH412. The additives such as TiCl3 and NbF5 are investigated their effect on kinetics and cycling performance of LiBH4-CaH2 composite.

Experimental Details

LiBH4 (≥95% pure), purchased from Acros Organics, CaH2 (≥98% pure) and NbF5 ( ≥ 99% pure), purchased from Alfa Aesar, and TiCl3 (≥95% pure), synthesized by the reaction of titanium tetrachloride with metallic titanium in molten CaCl2 and the enrichment process with HCl gas26, were utilized directly without any further purification. The mole ratio of LiBH4-CaH2 composite according to reaction 2 is 6:1. The pure LiBH4-CaH2 composite and LiBH4-CaH2 composite doped with different additives (1 mol% TiCl3, and 5 wt% NbF5) was ball-milled under argon atmosphere by using a QM-2B high energy mill (Nanjing NanDa Instrument Plant) at a rotating speed of 1200 rpm for 1 h. Two kinds of stainless steel balls with 4 mm and 8 mm diameters were added with a ball-to powder weight ratio of 12.5:1. Typically, 4 g mixture was sealed in the stainless steel vessel within a high purity argon atmosphere during milling. To avoid excess heating of the stainless steel vessel, there were 10 min intervals between each 5 min milling process.

The isothermal desorption was measured by using the Sieverts-type pressure-composition-temperature (P-C-T) apparatus (General Research Institute for Nonferrous Metals, China). The maximum pressure, maximum vacuum degree and maximum temperature of this apparatus is 10 MPa, 10−1 Pa and 800 °C, respectively. Typically, 60–100 mg sample was loaded into the vessel, and then heated up to 450 °C under 0.1 MPa hydrogen atmosphere. Following the dehydrogenation, the samples were subjected to rehydrogenation studies at 450 °C under 8 MPa hydrogen pressure for 16 h. It should be noted that the additional content was not taken into consideration when calculating the released hydrogen. The PCI (Pressure-composition isotherms) curves were measured at 405 °C, 420 °C, 435 °C, 450 °C and 465 °C, respectively.

The phase structure of the samples after milling and dehydrogenation was examined by an MXP21VAHF X-ray Diffractometer (XRD with Cu Kα radiation, 40 kV, 300 mA), with the 2θ angle ranged from 10° to 90° with a scanning rate of 10° min−1. X-ray photoelectron spectroscopy (XPS) was performed with the PHI-5300 spectrometer. The morphology and phase constitution of all samples after ball milling and desorption were observed by and transmission electron microscopy (Tecnai G2 F30 S-TWIN, FEI, USA). Simultaneous differential scanning calorimetry (DSC) and Thermogravimetric Analysis (TGA) experiments were conducted under 50 mL min−1 argon flow in a NETZSCH STA 449F3 Jupiter instrument between 50 °C and 500 °C with a heating rate of 5 °C min−1. The samples were transferred to Al2O3 crucibles under argon atmosphere for the DSC-TGA measurements.

All samples handling was performed under strictly inert conditions (≥99.99% Ar atmosphere) in the glove box (Mikrouna, Super-750) equipped with oxygen/humidity sensors and recirculation system to avoid oxidation and moisture. Oxygen and H2O levels were kept below 0.1 ppm.

Results and Discussion

XPS characterization

The XPS results of three LiBH4-CaH2 composites after milling are presented in Fig. 1(a–c), which confirms the existence of element Li, B and Ca in both composites. Element Nb, F and Cl are identified in the catalyst-doped composite, while Ti are not discovered due to the low amount addition. The XRD results are presented in Fig. S1. There are only two obvious peaks in both composites, which are characterized as CaH2. It can be inferred that the structure of LiBH4 after milling becomes amorphous. No peaks of LiBH4, TiCl3 or NbF5 are detected. The XPS narrow spectra of ball-milled LiBH4-CaH2 composite are showed in Fig. 1(d). The photo-emission spectrum of B 1 s at 187.8 eV corresponds to LiBH4, while the existence of LiCl and NbF5 are also convinced in Fig. S2. XPS results testify that LiBH4 and TiCl3 react during ball milling22.

Figure 1

XPS scan spectra of three LiBH4-CaH2 composites after ball milling: (a) pure composite (b) 1 mol% TiCl3 addition (c) 5 wt% NbF5 addition (d) B 1 s in pure composite.

Investigation on energy storage density

The energy storage density is the most significant factor when evaluating a material is suitable for TS. The energy storage density to weight is related with the reaction enthalpy and molar mass. For dehydrogenation reaction of hydrides, the Van’t hoff equation and DSC integration can be used to calculate the reaction enthalpy. Besides high energy storage density, low equilibrium pressure and high working temperature are also two important factors in selecting hydrides for TS.

PCI curves and Van’t Hoff Calculation

The dehydrogenation PCI curves of pure LiBH4-CaH2 composite are shown in Fig. 2(a). Due to the sluggish kinetics, the pressure value in each platform can only be read after 4-hours waiting. Even so, the platform inclination is quite a lot, especially in low temperature condition. Considering the platform is a slope to some extent, the equilibrium pressures from 405 °C to 465 °C are calculated as the mean of pressure values in the platform. The equilibrium pressures of pure LiBH4-CaH2 composite from 405 °C to 465 °C are 0.2458 MPa, 0.3208 MPa, 0.4018 MPa, 0.4820 MPa and 0.5967 MPa, respectively. They are lower than the equilibrium pressures of reported thermal storage metal hydrides, such as MgH2, Mg2FeH6 and Ce2Mg17Hx468. What’s more, the low equilibrium pressure makes LiBH4-CaH2 composite possible to be operated at higher temperature. The higher working temperature can increase overall solar to electricity conversion efficiency and reduce the cost in CST plants27. The dehydrogenation capacity of pure LiBH4-CaH2 composite is mostly ranging from 10.5 wt% to 11.6 wt%, which is close to their theoretical value. The sluggish kinetics resulting from the relatively low temperature (405 °C) may account for the lower capacity (9.5 wt%). Only LiBH4-CaH2 without catalysts was measured due to the fact that influence the thermodynamics of the mixture as Ti-doped NaAlH412. Liu22 reported that LiCl forms during ball milling of 6LiBH4/CaH2/xTiCl3. LiF and CaF2 are observed after the ball milling reaction of NbF5 and LiBH4 or CaH224. Thermodynamics of pure LiBH4-CaH2 composite might have changed due to the formation of LiCl or LiF and CaF2.

Figure 2

(a) Dehydrogenation PCI curves, (b) Dehydrogenation van’t hoff curves of pure LiBH4-CaH2 composite.

A plot of ln P against 1000/T in Fig. 2(b) results in a nearly straight line. Calculation of ΔH = R · (lnP2 − lnP1)/(1/T2 − 1/T1) from ln P and 1/T values at 405 °C and 465 °C provides a ΔH of 60.555 kJ mol−1 H2. According to the reaction 2, a thermal storage density value of 3504.6 kJ kg−1 is calculated. It shows a superior capacity to sensible and latent thermal storage materials, even to thermochemical thermal storage materials shown in Table 1.

DSC Calculation

The DSC and TGA curves of pure LiBH4-CaH2 composite are shown in Fig. 3. There are mainly three endothermic peaks during the heating process. The endothermic effect at 108–112 °C is reversible and corresponds to polymorphic transformation of LiBH4. The second peak at 268–286 °C corresponds to the fusion of LiBH4. The third peak corresponds to the dehydrogenation behavior of LiBH4. The onset temperature is 392 °C and the peak temperature is 446 °C. According to TGA results, dehydrogenation reaction ends at 497 °C. The integration of DSC on temperature from 392 °C to 497 °C is calculated as enthalpy of reaction 2, with a value of 60.706 kJ mol−1 H2.

Figure 3

DSC and TGA curves of pure LiBH4-CaH2 composite.

The DSC and TGA curves of LiBH4-CaH2 composites with TiCl3 and NbF5 addition are shown in Fig. S3. There are both three endothermic peaks in these two composites. NbF5 addition shows a more remarkable influence on the decrease of onset temperature than TiCl3. The onset temperature, dehydrogenation reaction enthalpy and thermal storage density of three composites and other potential TS system are shown in Table 2. The TiCl3 doped composite and NbF5 doped composite shows similar reaction enthalpy as the pure LiBH4-CaH2 composite. It can be speculated that catalyst additions in LiBH4-CaH2 composite do not have a remarkable influence on the dehydrogenation reaction enthalpy. The TS density of pure composite is the highest (3511.45 kJ kg−1), while TiCl3 and NbF5 doped composites possess nearly 3300 kJ kg−1 TS density, with a little reduction. The DSC calculation results are in accordance with the Van’t hoff calculation results. Comparing with actual TS density of MgH2 (2147 kJ kg−1)2829, LiBH4-CaH2 composites shows a clear superiority.

Table 2

TS properties of three LiBH4-CaH2 composites and other potential TS composites468930.

	H2 content (wt%)	Enthalpy (kJ mol−1 H2)	TS density (kJ kg−1)	Equilibrium temperature at 1 bar [°C]
MgH2	7.6	74.4	2860	280
NaMgH3*	4.0	86	1700	380
Mg2FeH6	5.5	77.4	2106.5	320
Mg2NiH6	3.6	62	1116	250
CaH2	4.8	186.2	4422.8	950
Ce2Mg17Hx	5.0	75.5	1926.3	310
LiBH4-CaH2	11.7	60.706	3511.45	350
LiBH4-CaH2-1 mol% TiCl3	11.6	59.354	3316.64	—
LiBH4-CaH2-5 wt% NbF5	10.4	60.011	3307.71	—

*Calculated for the reaction .

Investigation on kinetics

The Fig. 4 shows the desorption behavior of three LiBH4-CaH2 composites. The addition of TiCl3 significantly improves the dehydrogenation kinetics of LiBH4-CaH2 composite, while NbF5 influence it in an opposite way. Both composites can release 9–10 wt% hydrogen in an hour. After 4 hours, the pure composite and TiCl3 doped composite shows a nearly theoretical hydrogen capacity (11.7 wt%), while NbF5 doped composite only desorbs around 10 wt% hydrogen. TiCl3 shows a remarkable impact on improving the desorption kinetics and maintaining the hydrogen capacity. Liu22 reported that LiCl formed through replacement reaction between LiBH4 and TiCl3 during ball milling can be incorporated into LiBH4 to form solid solution LiBH4·LiCl. It favorably changes viscosity, preserving the nano-sized phase arrangement formed after milling, leading to fast kinetics.

Figure 4

Isothermal dehydrogenation curves of different LiBH4-CaH2 composites under 450 °C and H2 atmosphere (0.1 MPa).

The XRD results of three LiBH4-CaH2 composites after dehydrogenation are shown in Fig. 5. The main phase of both composites is LiH and CaB6, which is in accordance with reaction 2. The existence of phase CaO, LiOH!H2O and LiBO2 is due to the oxidation during the experiments. The remaining CaH2 and B are identified in the pure composite after dehydrogenation. B is the product of LiBH4 after dehydrogenation (shown in reaction 1), which also explains why a little CaH2 remains. The LiF and CaF2 phase are detected in the NbF5 doped composite, while no peaks of chlorides are identified in the TiCl3 doped composite.

Figure 5

XRD patern of three LiBH4-CaH2 composites after desorption: (a) pure composite (b) 1 mol% TiCl3 addition (c) 5 wt% NbF5 addition.

Investigation on reversibility and cycling stability performance

A test over 10 cycles was performed under very severe cycling conditions (desorption: 450 °C, 0.1 Mpa,4 h; absorption: 450 °C, 8 MPa, 16 h). The results are shown in Fig. 6. The initial hydrogen capacity of the pure composite and TiCl3 doped composite shows a nearly theoretical hydrogen capacity (11.7 wt%), while NbF5 doped composite only desorbs around 10 wt% hydrogen. The hydrogen capacity of both composites declines during cycling. It is worth mentioning that TiCl3 doped composite can reversibly store 9 wt% hydrogen during first three cycles. After 10 cycles, the remaining hydrogen capacity of pure composite, NbF5 doped composite and TiCl3 doped composite is 3.8 wt%, 6.4 wt% and 7.1 wt%, respectively. TiCl3 and NbF5 seems effectively raise the cycling stability performance of LiBH4-CaH2 composite.

Figure 6

Cycling curves of three LiBH4-CaH2 composites.

The TEM images of pure LiBH4-CaH2 composite after 10 cycles are shown in Fig. 7(a–d). The main phases are small particles with a diameter of 3–6 nm, separately scattering. Particle aggregation shown in Fig. 7(c), which may result from the sintering, is also found. The diffraction ring in Fig. 7(d) is very obvious, indicating that amorphous structure is formed. The particle aggregation and amorphous structure of products accounts for the dramatic loss of hydrogen capacity of pure LiBH4-CaH2 composite during cycling. TEM images of TiCl3 doped composite and NbF5 doped composite after 10 cycles are shown in Fig. 7(e,f,g and h). The small particles with a diameter of 3–6 nm are both observed. However, the results of electron diffraction indicate that the TiCl3 doped composite after 10 cycles is crystal structure, while NbF5 doped composite after 10 cycles is amorphous structure. By analyzing the diffraction ring diameter, the crystal structure is assumed to be CaB6. The amorphous structure of B is not good for the reverse reaction to produce LiBH4, while the crystal structure of CaB6 is in favor of the reverse reaction293132. This explains why TiCl3 plays a more effective role in raising the cycling stability performance of LiBH4-CaH2 composite than NbF5. Moreover, it is noteworthy that a graphene-like lamellar structure are found in NbF5 doped composite after 10 cycles. The value of interlamellar spacing (d) is 0.3364 nm, which is corresponding to NbF5. But the lamellar structure of NbF5 is never reported. Thus, the d value of NbB2 is 0.3321 nm, which is close to the 0.3364 nm. Minella reported that NbB2 nanoparticles was observed after milling or upon sorption reactions of Nb-based Ca(BH4)2 doped composites33. It is reasonable to be assume that a small amount of NbB2 can also be formed in Nb-based LiBH4-CaH2 doped composites. It needs more research work to identify this graphene-like lamellar structure in NbF5 doped composite.

Figure 7

TEM and SAED images: (a–d) pure LiBH4-CaH2 composite after 10 cycles; (e,f) TiCl3 doped composite after 10 cycles; (g,h) of NbF5 doped composite after 10 cycles.

Conclusion

The reaction enthalpy of LiBH4/CaH2 composite is almost 60 kJ/mol H2 and equilibrium pressure is 0.482 MPa at 450 °C. The thermal storage density of LiBH4/CaH2 composite is 3504.6 kJ/kg. XRD results show that the main phase after dehydrogenation is LiH and CaB6. The exsience of TiCl3 and NbF5 can effectively enhance the cycling perfomance of LiBH4/CaH2 composite, with 6–7 wt% hydrogen capacity after 10 cycles. The high thermal sotrage density, high working temperature and low equilibrium pressure make LiBH4/CaH2 composite a potential thermal storage material.

Although the high price of starting materials, such as LiBH4, will limit its usage, the LiBH4/CaH2 composite could serve as the additives for Magnesium-based alloys in TS. The research will be continued on the pair study of LiBH4/CaH2 composite with another metal hydride working at lower temperature.

Additional Information

How to cite this article: Li, Y. et al. Investigation on LiBH4-CaH2 composite and its potential for thermal energy storage. Sci. Rep. 7, 41754; doi: 10.1038/srep41754 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.