Konrad
Burkmann
a,
Franziska
Habermann
a,
Bianca
Störr
a,
Jürgen
Seidel
a,
Roman
Gumeniuk
b,
Klaus
Bohmhammel
a and
Florian
Mertens
*ac
aInstitut für Physikalische Chemie, TU Bergakademie Freiberg, Leipziger Straße 29, 09599 Freiberg, Germany. E-mail: florian.mertens@chemie.tu-freiberg.de; Tel: +49-3731-393737
bInstitut für Experimentelle Physik, TU Bergakademie Freiberg, Leipziger Straße 23, 09599 Freiberg, Germany
cZentrum für Effiziente Hochtemperatur-Stoffwandlung, TU Bergakademie Freiberg, Winklerstraße 5, 09599 Freiberg, Germany
First published on 6th May 2025
Over the last two decades, complex metal hydrides have attracted attention in various fields such as chemical hydrogen storage, solid state electrolytes and superconductivity. In this context, we determined thermodynamic properties of the complex metal hydride Y(BH4)3. Yttrium borohydride was prepared by solid state metathesis yielding a mixture containing three equivalents lithium chloride beside the boranate. The benefit of the mechanochemical preparation procedure compared to the classical wet chemical one is to avoid the desolvation step which can lead to a partial decomposition of the hydride. The heat capacity of the compound as a function of the temperature covering a temperature range from 2 K to 370 K was determined using two different calorimetric techniques. Based on the heat capacity data, the standard entropy at 298.15 K was obtained. From the evaluation of the low temperature heat capacity region the Sommerfeld coefficient and the Debye temperature were derived.
Several synthesis routes have been described for the synthesis of the α polymorph of Y(BH4)3 in the literature.2,7 Nevertheless, solvent mediated or mechanochemically driven metathesis reactions without solvents using a ball mill are the two most common methods to synthesise Y(BH4)3 from YCl3 and an alkali metal borohydride (LiBH4 or NaBH4).2,7 The reaction between yttrium chloride and lithium borohydride is given by eqn (1):
3LiBH4 + YCl3 → Y(BH4)3 + 3LiCl | (1) |
Jaroń and Grochala showed that the ball milling reaction did not take place by using sodium borohydride.18 Several authors used ball milling for the synthesis of Y(BH4)3 as well19–26 resulting in mixtures consisting of Y(BH4)3 and LiCl. For the solvent mediated metathesis DMS,27 THF18,28–30 or diethyl ether15,31 were used. Besides the formation of stable solvent adducts, another problem is given by the solubility of the by-product LiCl in the used solvent,15,32,33 except for DMS29 in which LiCl is insoluble. Therefore, when using THF or diethyl ether as a reaction medium, it is advised to extract the product mixture with DMS to obtain a pure DMS-solvent adduct29,31 or to use NaBH4 instead of LiBH4 as a reactant for the synthesis,28 because NaCl is insoluble in diethyl ether and THF.33 Certainly, the extraction step can also be applied when using the mechanochemical route. Any method that uses solvents must be complemented by a thermal desolvation step to obtain the pure Y(BH4)3.29 However, very often uncontrolled hydrogen desorption occurs during the decomposition of the solvent complexes. In addition, the solvent itself can decompose during desolvation and thus contaminate the sample.
Up to our best knowledge, only one publication addresses the determination of thermodynamic quantities of Y(BH4)3 so far. Lee et al. calculated the reaction enthalpies and reaction entropies for the decomposition of Y(BH4)3 into several yttrium borides using the Quantum-ESPRESSO density functional theory34 from which we derived the enthalpies of formation and the absolute entropies of α-Y(BH4)3 and β-Y(BH4)3, resulting in the following values: ΔFH (298.15 K) = −409.8 kJ mol−1 and S (298.15 K) = 149.4 J mol−1 K−1 for α-Y(BH4)3 and ΔFH (298.15 K) = −405.9 kJ mol−1 and S (298.15 K) = 153.4 J mol−1 K−1 for the high temperature β polymorph of Y(BH4)3. The experimental determined temperature of the phase transition is about 454 K.25,35,36 Nevertheless, no experimental values are available for both thermodynamic quantities.
The aim of our work is to determine experimentally the standard entropy of α-Y(BH4)3 using its measured heat capacity function. We used a solid state metathesis reaction leaving out the solvent extraction step that is often applied in complex metal hydride syntheses.7,29 The latter to avoid decomposition and contamination with the solvent of the hydride halide mixtures (Y(BH4)3 + 3LiCl). To the best of our knowledge, this procedure has not been shown to be applicable to boranates and our study is the first to confirm that, if the composition of the mechanochemically produced mixtures is known, accurate thermodynamic data for the boranate can be obtained without separating the by-product prior to the respective calorimetric measurements. In future, this will further establish mechanochemistry as a simple and environmentally friendly method for the production of various hydrides, if their thermodynamic data are to be determined. This procedure is justified, because no double salts, like in the case of lanthanum borohydride forming LaLi(BH4)3Cl,37,38 are formed and therefore the by-product LiCl can be assumed to be inert.
Lithium borohydride (LiBH4, Chemetall) and yttrium trichloride (YCl3, Alfa Aesar, ultra dry, 99.99%) were used as received. Sapphire, (aluminium oxide, Al2O3) was used as reference material for the heat capacity measurements. A 25 μm thick copper foil (Cu, Alfa Aesar, 99.999%) was used to fold copper crucibles as containers for the samples used in the PPMS measurement.
![]() | (2) |
Furthermore, the same approach is used to obtain the specific heat of the Y(BH4)3 from the measured sample mixture of Y(BH4)3 + 3LiCl by subtracting the specific heat of LiCl from it using the data published by Shirley47 and Moyer48 for LiCl for temperatures lower than 300 K and from NIST49 for temperatures higher than 300 K. The heat capacity measurements at temperatures higher than 280 K were performed in a DSC 111 from SETARAM, France, using the instrument software (Calisto, AKTS/SETARAM) to conduct the data evaluation and were inspired by the method already described elsewhere.4–6,8,9,42,50–54 In detail, about 100 mg of the sample were filled into standard stainless steel sample vessels. Using a nickel sealing the vessels were crimped tightly. The nickel sealings and the vessels used for the sample, the reference and the blank were brought to an identical mass by hand polishing to avoid variations in the blank effects. A CP-by-step method55 based on four temperature steps (5 K, 3 K min−1) each followed by an isothermal period of 30 min duration was conducted in the temperature range from 10 °C to 30 °C. After that, another step programme based on eight steps (10 K, 3 K min−1) each also followed by an isothermal period of 30 min duration in the temperature range from 30 °C to 100 °C was performed. This procedure was used for the measurements of the sample, the reference (granular sapphire for the verification of the overall accuracy of the experiments) and the blank. Using the recommended reference heat capacity of sapphire published by Della Gatta et al.,56 the specific heat of the sample were calculated viaeqn (3).
![]() | (3) |
![]() | ||
Fig. 1 PXRD of the product mixture after the metathesis reaction according to eqn (1). The reflections of the assigned phases were taken from the ICSD.57 |
Furthermore, there are no reflections of possible decomposition products such as YH2, YH3, B or other boron containing species visible indicating that there was no or negligible decomposition during milling, so the used procedure appears appropriate for the preparation of the compound.
T [K] | C P [J mol−1 K−1] | T [K] | C P [J mol−1 K−1] | T [K] | C P [J mol−1 K−1] |
---|---|---|---|---|---|
2.0606 | 0.014475 | 8.9878 | 0.91862 | 16.132 | 3.9476 |
2.1413 | 0.015631 | 9.0881 | 0.95080 | 16.232 | 4.0055 |
2.2274 | 0.017200 | 9.1884 | 0.97756 | 16.333 | 4.0576 |
2.3149 | 0.018878 | 9.2883 | 1.0041 | 16.433 | 4.1196 |
2.4037 | 0.021122 | 9.3890 | 1.0356 | 16.534 | 4.1757 |
2.4938 | 0.023054 | 9.4893 | 1.0617 | 16.635 | 4.2365 |
2.5843 | 0.025405 | 9.5898 | 1.0927 | 16.735 | 4.2967 |
2.6782 | 0.027826 | 9.6902 | 1.1235 | 16.836 | 4.3564 |
2.7706 | 0.030499 | 9.7908 | 1.1540 | 16.936 | 4.4154 |
2.8635 | 0.033377 | 9.8935 | 1.1841 | 17.135 | 4.5323 |
2.9572 | 0.036619 | 9.9941 | 1.2141 | 22.171 | 7.6974 |
3.0517 | 0.040245 | 10.095 | 1.2491 | 27.115 | 11.301 |
3.1464 | 0.043804 | 10.195 | 1.2839 | 32.150 | 14.834 |
3.2414 | 0.047587 | 10.296 | 1.3131 | 37.185 | 18.269 |
3.3370 | 0.051741 | 10.396 | 1.3474 | 42.228 | 21.756 |
3.4332 | 0.056436 | 10.496 | 1.3814 | 47.270 | 25.309 |
3.5298 | 0.061063 | 10.597 | 1.4151 | 52.310 | 29.000 |
3.6266 | 0.066121 | 10.697 | 1.4537 | 57.355 | 32.649 |
3.7233 | 0.071602 | 10.797 | 1.4920 | 62.394 | 36.422 |
3.8203 | 0.077023 | 10.897 | 1.5249 | 67.437 | 40.117 |
3.9177 | 0.083336 | 10.998 | 1.5627 | 72.478 | 43.696 |
4.0155 | 0.089187 | 11.098 | 1.5950 | 77.519 | 47.354 |
4.1131 | 0.096278 | 11.199 | 1.6322 | 82.560 | 50.983 |
4.2109 | 0.10351 | 11.299 | 1.6692 | 87.605 | 54.165 |
4.3114 | 0.11039 | 11.399 | 1.7110 | 92.645 | 57.312 |
4.4094 | 0.11826 | 11.499 | 1.7473 | 97.679 | 60.687 |
4.5078 | 0.12619 | 11.599 | 1.7834 | 102.72 | 64.238 |
4.6060 | 0.13486 | 11.700 | 1.8242 | 105.73 | 66.244 |
4.7047 | 0.14348 | 11.800 | 1.8649 | 115.76 | 73.770 |
4.8031 | 0.15253 | 11.901 | 1.9050 | 125.85 | 80.581 |
4.9019 | 0.16270 | 12.001 | 1.9450 | 135.90 | 86.421 |
5.0007 | 0.17256 | 12.101 | 1.9846 | 145.99 | 92.704 |
5.0996 | 0.18273 | 12.202 | 2.0290 | 156.08 | 98.812 |
5.1985 | 0.19347 | 12.302 | 2.0679 | 166.19 | 104.51 |
5.2978 | 0.20505 | 12.402 | 2.1117 | 176.30 | 110.05 |
5.3966 | 0.21646 | 12.503 | 2.1551 | 186.47 | 115.26 |
5.4956 | 0.22863 | 12.603 | 2.1930 | 196.59 | 120.25 |
5.5948 | 0.24118 | 12.704 | 2.2357 | 206.72 | 125.19 |
5.6945 | 0.25413 | 12.804 | 2.2832 | 216.82 | 131.33 |
5.7936 | 0.26769 | 12.904 | 2.3252 | 226.93 | 137.14 |
5.8931 | 0.28123 | 13.005 | 2.3720 | 237.02 | 142.90 |
5.9926 | 0.29621 | 13.105 | 2.4184 | 247.11 | 148.37 |
6.0921 | 0.31006 | 13.205 | 2.4645 | 257.17 | 153.60 |
6.1916 | 0.32517 | 13.306 | 2.5049 | 267.27 | 158.99 |
6.2915 | 0.33963 | 13.406 | 2.5554 | 277.36 | 165.66 |
6.3908 | 0.35666 | 13.507 | 2.6003 | 285.98 | 176.07 |
6.4906 | 0.37244 | 13.607 | 2.6447 | 286.21 | 179.95 |
6.5901 | 0.38877 | 13.707 | 2.6940 | 287.45 | 173.64 |
6.6898 | 0.40518 | 13.808 | 2.7376 | 291.13 | 176.92 |
6.7894 | 0.42601 | 13.908 | 2.7860 | 291.35 | 183.82 |
6.8891 | 0.44148 | 14.009 | 2.8339 | 296.08 | 177.07 |
6.9887 | 0.46201 | 14.109 | 2.8869 | 296.29 | 187.76 |
7.0887 | 0.48236 | 14.210 | 2.9339 | 297.56 | 177.53 |
7.1884 | 0.49736 | 14.310 | 2.9858 | 301.09 | 185.29 |
7.2885 | 0.51738 | 14.410 | 3.0373 | 301.30 | 189.24 |
7.3878 | 0.53729 | 14.511 | 3.0831 | 310.37 | 191.46 |
7.4878 | 0.55697 | 14.611 | 3.1336 | 310.38 | 192.50 |
7.5876 | 0.58171 | 14.712 | 3.1837 | 320.44 | 197.12 |
7.6876 | 0.60106 | 14.812 | 3.2385 | 320.45 | 199.20 |
7.7876 | 0.62543 | 14.913 | 3.2876 | 330.49 | 204.91 |
7.8876 | 0.64442 | 15.013 | 3.3414 | 330.50 | 206.57 |
7.9877 | 0.66842 | 15.127 | 3.4012 | 340.30 | 211.37 |
8.0875 | 0.69226 | 15.226 | 3.4544 | 340.53 | 216.02 |
8.1878 | 0.71588 | 15.327 | 3.5066 | 350.31 | 218.64 |
8.2872 | 0.73936 | 15.427 | 3.5584 | 350.32 | 224.28 |
8.3874 | 0.76780 | 15.528 | 3.6149 | 360.10 | 224.91 |
8.4874 | 0.78565 | 15.628 | 3.6656 | 360.33 | 229.90 |
8.5875 | 0.81370 | 15.729 | 3.7210 | 370.11 | 231.34 |
8.6873 | 0.84156 | 15.830 | 3.7759 | 370.12 | 239.35 |
8.7875 | 0.86919 | 15.930 | 3.8302 | ||
8.8876 | 0.89662 | 16.031 | 3.8892 |
The inset graph in Fig. 2 displays the relative deviation of the experimental data from the respective fit function. The relative deviation of the measured values from the fits is mostly within ±1.5%, but the deviations are rising to maximal ±4.5% below a temperature of about 5 K and above a temperature of about 280 K. In order to keep the degree of the fitting polynomials both as low as possible and to ensure still satisfying fits, the temperature range of our heat capacity measurements was divided into five appropriate intervals. Similarly as in our recently published works4–6,8,42,50–54 [Ca(AlH4)2, Sr(BH4)2], the heat capacities CP(T) were fitted then within these intervals using established polynomial functions.58 The applied functions and the corresponding temperatures are given below.
0 K to 5 K: CP = b·T + e·T3 + f·T5 | (4) |
5 K to 17 K: CP = a + b·T + c·T2 + d·T−2 + e·T3 + g·T−1 | (5) |
17 K to 100 K: CP = a + b·T + c·T2 + d·T−2 + e·T3 | (6) |
100 K to 230 K: CP = a + b·T + c·T2 + d·T−2 + e·T3 | (7) |
230 K to 380 K: CP = a + b·T + c·T2 + d·T−2 | (8) |
The results for the fit coefficients and the fit quality parameters R2 (coefficient of determination) and FitStdErr (fit standard error) are summarised in Table 2. The fits represent the experimental values with a deviation similar to the experimental standard uncertainty.
T interval [K] | a [J mol−1 K−1] | b [J mol−1 K−2] | c [J K mol−1] | d [J mol−1 K−3] | e [J mol−1 K−4] | f [J mol−1 K−6] | g [J mol−1] | R 2 | FitStdErr |
---|---|---|---|---|---|---|---|---|---|
0 to 5 | 0 | 1.4814·10−3 | 0 | 0 | 1.2216·10−3 | 4.0961·10−6 | 0 | 1.0000 | 2.6436·10−4 |
5 to 17 | 6.0668·100 | −8.0424·10−1 | 6.0859·10−2 | 3.8147·101 | −8.6660·10−4 | 0 | −2.4058·101 | 1.0000 | 3.1382·10−3 |
17 to 100 | −6.5064·100 | 5.9847·10−1 | 2.2764·10−3 | 5.4218·101 | −1.3670·10−5 | 0 | 0 | 0.9999 | 1.5525·10−1 |
100 to 230 | −1.4167·102 | 2.7555·100 | −1.1015·10−2 | 2.0107·105 | 1.8536·10−5 | 0 | 0 | 0.9999 | 2.6848·10−1 |
230 to 370 | −2.2501·102 | 1.5227·100 | −9.3269·10−4 | 3.3004·106 | 0 | 0 | 0 | 0.9879 | 3.0988·100 |
CP,BH4 = CP,MBH4 − CP,M with M = Li, Na | (9) |
CP,M(BH4)x = CP,M + x·CP,BH4 | (10) |
Table 3 lists the coefficients for the Maier Kelley58 heat capacity function polynomials of Y, Li and LiBH4 shown below:
CP = A + B·T + C·T−2 + D·T3 + E·T3 + F·T−1 | (11) |
Compound | T range [K] | A [J mol−1 K] | B [J mol−1 K−2] | C [J K mol−1] | D [J mol−1 K−3] | E [J mol−1 K−4] | F [J mol−1] |
---|---|---|---|---|---|---|---|
Y | 20 to 300 (ref. 64) | 2.831·10−1 | 4.842·10−2 | 1.098·104 | −2.899·10−4 | 4.886·10−7 | −1.093·103 |
300 to 1755 (ref. 65) | 2.390·101 | 7.620·10−3 | 3.130·104 | 0 | 0 | 0 | |
Li | 20 to 105 (ref. 66) | −1.615·100 | 2.728·10−2 | 2.814·102 | 2.085·10−3 | −8.382·10−6 | 0 |
105 to 170 (ref. 66) | 4.839·102 | −7.341·100 | −7.569·105 | 4.272·10−2 | −8.641·10−5 | 0 | |
170 to 300 (ref. 66) | −4.353·101 | 5.752·10−1 | 2.167·105 | −1.791·10−3 | 2.024·10−6 | 0 | |
300 to 453 (ref. 65) | 3.895·101 | −7.098·10−2 | −3.202·105 | 1.193·10−4 | 0 | 0 | |
LiBH4 | 20 to 300 (ref. 67) | −4.903·101 | 9.717·10−1 | −6.683·103 | −3.169·10−3 | 4.538·10−6 | 9.822·102 |
300 to 400 (ref. 65) | −4.671·101 | 6.779·10−1 | 1.728·105 | −8.409·10−4 | 0 | 0 |
The coefficients for the CP range below 300 K for Y and LiBH4 are derived from the fits of given literature values and the ones for Li are obtained from its G function. Above room temperature all values were taken from the HSC 5.1 database. All data used were validated by comparing various literature sources (as far as possible) and seem to posses high reliability. Respective references are shown in Table 3.
The comparison between our experimental values and those derived from the described method using LiBH4 are presented in Fig. 3. The use of NaBH4 as reported by Dematteis et al.61 was not reasonably applicable because of the presence of a phase transition at 189.9 K.68
![]() | ||
Fig. 3 Comparison between the heat capacity of Y(BH4)3 derived from our measurements (black solid line) and the one computed by applying Neumann–Kopp's rule according to ref. 61 and 63 (red dashed line: values below room temperature; blue dashed line: values above room temperature). The inset displays the relative deviation between the experimental values and the ones from the Neumann–Kopp rule. |
The inset in Fig. 3 displays the deviation of the values of the experimental fit curve from the one calculated by the Neumann–Kopp rule. The deviation between both curves is significant and exceeds in a wide temperature range 10%. Towards high temperatures the slope of the Neumann–Kopp curve decreases so that both curves intersect at about 335 K. Similar behaviour has been found in other systems59,60,62 and was attributed to factors that are not reflected in the Neumann–Kopp rule such as magnetism and anharmonic effects60 or the formation of Schottky defects.62
In general Neumann–Kopp's rule is reasonably applicable for simple compounds (alloys), whose elements show in pure form the same crystal structure as the compounds to be described. In cases where the compound possesses complex subunits, it is necessary not only to consider the binding situation within the subunit but also that of the subunit in the target compound.59,61 In the case of the calculation of the heat capacity of Y(BH4)3 according to eqn (9) and (10), the requirements that the crystal structures of the components and the binding situation of the subunits should be the same is not fulfilled. Li crystallises in the body centred cubic W type structure (space group 229),69 Y in the hexagonal Mg type structure (space group 194),70 LiBH4 in a orthorhombic structure (space group 62)31 and Y(BH4)3 in a cubic structure (space group 205).31
In the literature, the influence of the thermal expansion of the compounds, among other things, was used for the extension of the Neumann–Kopp approximation in order to obtain more realistic CP values. Kumar et al. and Leitner et al. applied the extended Neumann–Kopp method to CaHCl and CaHBr59 and mixed oxides,60 respectively and obtained significant improvements. Unfortunately, this procedure is currently not applicable because of the missing thermal expansion coefficient for Y(BH4)3.
By the integration of the conventional Neumann–Kopp CP function divided by the temperature in the range from 5 K to 300 K, a value of the standard entropy of S° (300 K) = 183.8 J mol−1 K−1 was derived. This value differs about 9% from the experimental one determined in this study (vide infra). This result is a further indication of the limitation in regard to the use of the Neumann–Kopp approximation.
Based on the given explanations, the Neumann–Kopp rule can be seen as a rough estimation method and cannot replace reliable experimental data. From our findings for Y(BH4)3 its application appears also not recommended for precise thermodynamic calculations regarding other transition metal boranates that have not yet been prepared/investigated.
![]() | (12) |
The value for the standard entropy S° at 298.15 K was determined to be S° (298.15 K) = (168.9 ± 5.1) J mol−1 K−1, which is considerably higher than the value reported by Lee et al. for the room temperature α-phase of Y(BH4)3 (149.4 J mol−1 K−1) derived via DFT calculations.34 All calculated values of S° are presented in Table 4. In addition, the molar standard entropy of formation ΔFS° has been calculated from eqn (13) resulting in a value of ΔFS° (298.15 K) = −677.7 J mol−1 K−1. The values for the standard entropy of yttrium, boron and hydrogen were taken from the HSC database.65
![]() | (13) |
T [K] | C P [J mol−1 K−1] | ΔS0T [J mol−1 K−1] | ΔH0T/T [J mol−1 K−1] | Φ [J mol−1 K] | T [K] | C P [J mol−1 K−1] | ΔS0T [J mol−1 K−1] | ΔH0T/T [J mol−1 K−1] | Φ [J mol−1 K] |
---|---|---|---|---|---|---|---|---|---|
5 | 0.17291 | 0.06087 | 0.044012 | 0.016855 | 195 | 119.53 | 106.40 | 59.589 | 46.807 |
10 | 1.2194 | 0.44076 | 0.32539 | 0.11538 | 200 | 122.135 | 109.45 | 61.120 | 48.335 |
15 | 3.3375 | 1.3011 | 0.94792 | 0.35321 | 205 | 124.767 | 112.50 | 62.640 | 49.863 |
20 | 6.3997 | 2.6616 | 1.9117 | 0.74991 | 210 | 127.434 | 115.54 | 64.151 | 51.391 |
25 | 9.7512 | 4.4478 | 3.1428 | 1.3051 | 215 | 130.150 | 118.57 | 65.654 | 52.918 |
30 | 13.188 | 6.5283 | 4.5295 | 1.9988 | 220 | 132.927 | 121.60 | 67.151 | 54.444 |
35 | 16.687 | 8.8233 | 6.0157 | 2.8076 | 225 | 135.778 | 124.61 | 68.644 | 55.970 |
40 | 20.234 | 11.283 | 7.5708 | 3.7117 | 230 | 138.715 | 127.63 | 70.136 | 57.495 |
45 | 23.815 | 13.872 | 9.1765 | 4.6957 | 235 | 141.089 | 130.63 | 71.615 | 59.019 |
50 | 27.421 | 16.568 | 10.821 | 5.7474 | 240 | 144.024 | 133.64 | 73.093 | 60.543 |
55 | 31.039 | 19.351 | 12.494 | 6.8570 | 245 | 147.061 | 136.64 | 74.571 | 62.065 |
60 | 34.659 | 22.207 | 14.190 | 8.0167 | 250 | 150.189 | 139.64 | 76.052 | 63.586 |
65 | 38.271 | 25.124 | 15.904 | 9.2201 | 255 | 153.397 | 142.64 | 77.537 | 65.107 |
70 | 41.863 | 28.092 | 17.630 | 10.462 | 260 | 156.675 | 145.65 | 79.027 | 66.627 |
75 | 45.426 | 31.102 | 19.364 | 11.737 | 265 | 160.016 | 148.67 | 80.524 | 68.147 |
80 | 48.949 | 34.146 | 21.104 | 13.042 | 270 | 163.410 | 151.69 | 82.027 | 69.666 |
85 | 52.423 | 37.218 | 22.844 | 14.374 | 275 | 166.851 | 154.72 | 83.538 | 71.185 |
90 | 55.836 | 40.311 | 24.582 | 15.729 | 280 | 170.331 | 157.76 | 85.057 | 72.704 |
95 | 59.178 | 43.420 | 26.316 | 17.105 | 285 | 173.846 | 160.81 | 86.584 | 74.223 |
100 | 62.440 | 46.539 | 28.041 | 18.498 | 290 | 177.389 | 163.86 | 88.119 | 75.742 |
105 | 65.914 | 49.668 | 29.760 | 19.908 | 295 | 180.956 | 166.92 | 89.662 | 77.261 |
110 | 69.443 | 52.816 | 31.483 | 21.332 | 298.15 | 183.213 | 168.86 | 90.638 | 78.219 |
115 | 72.934 | 55.980 | 33.210 | 22.770 | 300 | 184.541 | 169.99 | 91.213 | 78.781 |
120 | 76.367 | 59.157 | 34.937 | 24.220 | 305 | 188.141 | 173.07 | 92.773 | 80.302 |
125 | 79.729 | 62.342 | 36.662 | 25.681 | 310 | 191.751 | 176.16 | 94.340 | 81.823 |
130 | 83.012 | 65.534 | 38.381 | 27.152 | 315 | 195.369 | 179.26 | 95.915 | 83.345 |
135 | 86.211 | 68.727 | 40.094 | 28.633 | 320 | 198.990 | 182.37 | 97.497 | 84.868 |
140 | 89.326 | 71.919 | 41.797 | 30.122 | 325 | 202.611 | 185.48 | 99.087 | 86.392 |
145 | 92.358 | 75.106 | 43.488 | 31.618 | 330 | 206.231 | 188.60 | 100.683 | 87.917 |
150 | 95.310 | 78.287 | 45.167 | 33.121 | 335 | 209.845 | 191.73 | 102.28 | 89.443 |
155 | 98.189 | 81.460 | 46.831 | 34.629 | 340 | 213.453 | 194.86 | 103.89 | 90.970 |
160 | 101.00 | 84.622 | 48.480 | 36.142 | 345 | 217.050 | 198.01 | 105.51 | 92.498 |
165 | 103.75 | 87.772 | 50.113 | 37.659 | 350 | 220.636 | 201.15 | 107.13 | 94.028 |
170 | 106.45 | 90.909 | 51.731 | 39.179 | 355 | 224.209 | 204.31 | 108.75 | 95.559 |
175 | 109.11 | 94.034 | 53.332 | 40.701 | 360 | 227.766 | 207.47 | 110.38 | 97.092 |
180 | 111.74 | 97.144 | 54.918 | 42.226 | 365 | 231.305 | 210.64 | 112.01 | 98.625 |
185 | 114.34 | 100.24 | 56.489 | 43.752 | 370 | 234.826 | 213.81 | 113.65 | 100.160 |
190 | 116.93 | 103.32 | 58.045 | 45.279 |
For the calculation of the change in molar enthalpy ΔH0T = H(T) − H(0) the heat capacity functions were integrated according to eqn (14).
![]() | (14) |
The calculated values of ΔH0T divided by the temperature are given in Table 4 as well as the Φ parameter, which can be computed by eqn (15).
![]() | (15) |
CP = CP,elec + CP,vib = γ·T + β·T3 + δ·T5 | (16) |
Usually, the linear term with the Sommerfeld coefficient γ is attributed to the electrical conductivity of metals and related compounds.50–54,71,72 However, there are also publications taking into account the electronic contribution to the heat capacity for non conductive materials. Loos et al.42 observed this behaviour for lithium iron phosphate and explained it with the existence of electronic states in structural defects. Defects in the crystal structure of the synthesised Y(BH4)3 that may potentially contribute to the heat capacity even at low temperatures can be assumed to be present because of the preparation by ball milling.71,73,74 Habermann et al. found that the linear term is also required for a sufficient fit of the heat capacity values at low temperatures for various metal alanates produced by ball milling.4,5,8 The need of a linear term for nonmetallic, insulating compounds in the low temperature range due to the existence of vacancies in the lattice was also postulated by Schliesser and Woodfield.74 The parameters β and δ in eqn (16) are related to the lattice vibration contribution to the heat capacity in terms of the Debye theory.71,72
A value of γ = (1.48 ± 0.01)·10−3 J mol−1 K−2 is obtained for the Sommerfeld coefficient, which is comparable to that one of metals.71 As mentioned above the phenomenon is common if structural defects are present in the sample.42,74 For the lattice vibration parameter, the value of β = (1.22 ± 0.01)·10−3 J mol−1 K−4 was determined. A Debye temperature of ΘD = (294.2 ± 1.1) K was calculated using eqn (17) with the given number of atoms per formula unit of n = 16 for Y(BH4)3. This value represents the excitation of low energy acoustic phonons (lattice vibrations), shown in the phonon density of states diagram calculated by Lee et al.34
![]() | (17) |
The coefficient of the T5 term of the lattice contribution in eqn (16) was found to be δ = (4.09 ± 0.44)·10−6 J mol−1 K−6.
The measurement at low temperatures allows the determination of the absolute standard entropy resulting in a value of S° (298.15 K) = (168.9 ± 5.1) J mol−1 K−1 and a Debye temperature of ΘD = (294.2 ± 1.1) K. Despite the fact that the material is non-metallic, a comparably high value of γ = (1.48 ± 0.01)·10−3 J mol−1 K−2 was obtained for the Sommerfeld coefficient. This fact seems to be not unusual for mechanochemically prepared complex metal hydrides.
The absolute standard entropy derived from the measured data deviates significantly from the value of this quantity determined via DFT calculations in the literature. Given the determined uncertainty of the measurements of less than 5%, the obtained values appear to be more convincing. Although not surprising, however often applied, the estimate of the heat capacity via the modified Neumann–Kopp rule with the inclusion of complex subunits, in the presented case BH4, did not yield a satisfactory agreement with experimental findings.
Further investigations should address the determination of the enthalpy of formation as well as the understanding of the complex decomposition process. It is necessary to obtain thermodynamic values for possible decomposition intermediates to perform thermodynamic calculations on their existence during the decomposition process of Y(BH4)3. Applying this knowledge hopefully a deeper understanding of the potential use of Y(BH4)3 for reversible hydrogen storage concerning thermodynamic tuning strategies will be reached.
This journal is © The Royal Society of Chemistry 2025 |