Decomposition of LiAlH

In Fig. 1.27 and Table 1.5 a typical DSC peak position shift related to various heating rates for thermal decomposition of LiAlH is presented. 

The effect of catalytic metal chloride additives on the kinetics of isothermal decomposition of LiAlH in a Sieverts-type apparatus has been studied by a few research groups and the results seem to be rather contradictory. 

Figure 3.30 shows the DSC traces for the (MgH + 20, 30, 50 and 70 wt%LiAlH ) composites. Only single endothermic peak centered at 350°C is visible in DSC traces for the (MgH + 20 wt%LiAlH ) composite (Fig. 3.30a). This peak corresponds to the decomposition of MgH. The first low temperature exothermic effect observed in Fig. 3.9 for a pure LiAlH (both unmilled and milled), which is usually assigned to the interaction of LiAlH with hydroxyl impurities [67], is not observed in Fig. 3.30a-c but it appears in Fig. 3.30d for (MgH + 70 wt%LiAlH ). Four endothermic events occur for (MgH + 30, 50 and 70 wt%LiAlH ) (Fig. 3.30b-d). The first endothermic peak at 174-182°C has almost exactly the same temperature range as (Rla) in Fig. 3.9. No exothermic peak (Rib) of melting from Fig. 3.9 is seen in Fig. 3.30a-d. It seems that the addition of just 30 wt%MgH suppresses melting of LiAlH and its first decomposition into LijAlH and Al ((Rib) in Fig. 3.9) occurs from a solid phase and is endothermic. This is supported by the observation of partial decomposition of LiAlH into (LijAlH + Al) during milling as discussed before. The second endo peak in Fig. 3.30b-d at 198,193 and 223°C, respectively, corresponds to the decomposition...

As mentioned before in order to determine whether or not the free A1 formed upon decomposition of LiAlH /LijAlH in the composite could act as a catalyst, we also prepared composites with the content of A1 equivalent to the content of A1 in the Awt%LiAlH. Their DSC desorption peak temperature maxima are also plotted in Fig. 3.31. The composites with the equivalent content of A1 do not seem to follow the ROM behavior. Therefore, one can tentatively conclude that the underlying physical mechanism for the ROM behavior is not related to the catalytic effect of free Al. However, this possibility, however remote, cannot be completely ruled out of hand because the particle size of free Al formed upon decomposition might be much smaller than that obtained by ball milling of Al metal powder added to MgH powder. Nanosized free Al could aquire catalytic behavior. However, at the moment we do not have any evidence for that. 

Based on the DSC cnrve in Fig. 3.9b a pnre LiAlH milled for 20 h shonld decompose at 300°C throngh the reactions (Rib) and (R2) of (3.12) and (3.13), respectively, to LiH and A1 and in doing so, desorb a theoretical purity-corrected (97%) amount of eqnal to 7.66 wt%. However, it is seen in Fig. 3.34 that a pure LiAlH desorbs about 7 wt%H at 300°C. This amount is 0.7 wt% deficient with respect to the theoretical valne. This missing amount of H is lost, most likely, owing to the partial decomposition of LiAlH occurring dnring milling as evidenced by the presence of the... 

J.W. Wiench, V.P. Balema, V.K. Pecharsky, M. Pruski, Solid-state 27A1 NMR investigation of thermal decomposition of LiAlH , J. Solid State Chem. 177 (2004) 648-653. 

It is well established in the literature [67, 70, 85-89] that hydrogen desorbs from a pure, uncatalyzed LiAlH hydride in a three-step decomposition very similar to that of NaAlH, the first of which goes through the melting of LiAlH ... 

Figure 3.32 shows XRD patterns of (MgH -i-LiAlH ) composites after DSC testing up to 500°C. The primary phases present are Mg and Al. Peaks of MgO and (LiOH) HjO arise from the exposure of Mg and Li (or possibly even some retained LiH) to the environment during XRD tests. Apparently, XRD phase analysis indicates that a nearly full decomposition of original MgH and LiAlH hydride phases has occurred to the elements during a DSC experiment. In addition, no diffraction peaks of any intermetallic compound are observed in those XRD patterns. That means that no intermetallic compound was formed upon thermal decomposition of composites in DSC. Therefore, the mechanism of destabilization through the formation of an intermediate intermetallic phases proposed by Vajo et al. [196-198] and discussed in the beginning of this section seems to be ruled out of hand. 

Nonetheless, we have ealculated the lattice parameters and t e unit ce vo ume of free Mg which is formed from the decomposition of MgH an oun decreases with increasing content of LiAlH in a composite up to 30 wt c an t en more or less saturates as shown in Fig. 3.33. Such lattice shrinkage is most i e y... 

Fig. 3.33 Unit ceU volume of Mg formed upon decomposition of MgH in the composites during DSC experiments as a function of LiAlH content...

Fig. 3.36 (a) Desorption curves for the (MgHj + Xwt%Al) composites with the content of A1 additive equivalent to the content of free A1 formed after decomposition of Awt%LiAlH in (MgH + Xwt%LiAlH ) composites (300°C 0.1 MPa Hj). The (MgH + Xwt%Al) composites were baU milled for 20 h. (b) Dependence of an effective kinetic parameter, k, in the JMAK equation on the equivalent content of A1 metal additive... 

In order to find out if the decomposition of NaAlH destabilizes the MgH constituent in a composite, the peak temperatures of the last peak in a doublet/triplet in Fig. 3.38b-d which supposedly corresponds to the decomposition of MgH are plotted as a function of vol.%NaAlH in Fig. 3.39. It is quite clear that the ROM behavior for the MgH temperature is not obeyed for this composite system in contrast to the (MgH + LiAlH ) composite system. There are two factors which may be responsible for this behavior. 

Aluminum tris(tetrahydridoborate), A1(BH4)3, a colourless liquid, f.p. -64.5 °C, b.p. 44.5 C (extrap.), is the most volatile known compound of aluminum. It is formed by reactions of McsAl, AIH3 or LiAlH with diborane. An alternative synthesis is from AICI3 + 3NaBH4. Thermal decomposition of A1(BH4)3 occurs above 70 °C with loss of diborane to form HA1(BH4)2. Although stable below 25 °C under vacuum, A1(BH4)3 is a hazardous substance, being oxidized explosively by air and reacting violently with water or any protic agent. 

Silicon hydrides are prepared either by acid decomposition of magnesium silicide or by reduction of SiCl with LiAlH. Only a relatively modest yield of silane (20-30%) is obtained by the aqueous acid decomposition of silicide, but the proportion of the higher silanes is somewhat greater. If the reaction is carried out in liquid NHg or NSH4 instead of water, appreciably higher yields (over 80%) are obtained besides, up to 90% of the product consists of SiH, if one neglects the easily separable Hg. 

The Balz-Schiemann reaction has been used for the preparation of the 4-iluoro-derivatives of pyridine and 2,5-, 2,6-, and 3,5-lutidine it has also been used to obtain 5-fluoronicotinic acid, required for conversion into the corresponding pyridylmethanol via LiAlH reduction of the ethyl ester. The introduction of F into fluoroaromatic compounds has been achieved via isotopic exchange in diazonium tetrafluoroborates. U.v. irradiation of aqueous solutions of the appropriate diazonium tetrafluoroborates has been used to procure the first ring-fluorinated imidazoks, e.g. photolysis of the diazonium solution obtained by adding sodium nitrite to 2-amino-imidazole in aqueous fluoroboric acid provides 2-fluoroimidazole contaminated with only a small amount of 2-azidoimidazole, the sole product of thermal decomposition of imidazole-2-diazonium tetrafluoroborate. 

It is sometimes necessary to reduce a functionalized site to a CH or CH group. Starting from the acid level of oxidation, this is usually done in two stages. For example, an ester may be reduced to an alcohol with LiAlH, the alcohol converted to a tosylate, and then reduced again to a CH group. Ketones and aldehydes can be reduced completely via the tosylhydrazone using catecholborane followed by decomposition of the hydroboration intermediate as in Equation 6.71 [116]. 

Fig. 1.28 Kissinger plot for dehydrogenation of as-received LiAlH. Energy of activation for the decomposition LiAlH —> /4 LijAlH + 2/3A1 + (peak III in Fig. 1.27)...

Fig. 3.10 Kissinger analysis of the activation energy, E, of decomposition for the reaction (a) (R2) and (b) (R3) for pure LiAlH ...

Andreasen et al. [86] also found that ball milling increased the rate constant, k, in the JMAK equation (Sect. 1.4.1), of reaction (Rib) in solid state but virtually had no effect on the rate constant of reaction (R2). They also showed that the reaction constant, k, of reaction (Rib) in solid state increases with decreasing grain size of ball-milled LiAlH within the range 150-50 mn. Andreasen et al. concluded that the reaction (Rib) in solid state is limited by a mass transfer process, e.g., long range atomic diffusion of Al while the reaction (R2) is limited by the intrinsic kinetics (too low a temperature of decomposition). In conclusion, one must say that ball milling alone is not sufficient to improve the kinetics of reaction (R2). A solution to improvement of the kinetics of reaction (R2) could be a suitable catalytic additive. 

With Ag no binary hydride is known, but several thermally unstable complex hydrides are isolated that contain H bonded to Ag. For example, LiAlH, LiBH or LiGaH react with Ag perchlorate in ether to give white or yellow precipitates of AgAlH, AgBH and AgGaH. Thermal decomposition occurs below RT and is catalyzed by Ag ... 

---