The Facile Synthesis of Exogenous Lewis-Base-Free Amidoalanes: A Structural Comparison

Abstract

A simple one-pot reaction of LiAlH₄, AlCl₃, and a secondary amine HNR₂ [R = Et, ⁱPr, ⁱBu, cyclo-C₆H₁₁, (CH₂)₄, and (CH₂)₅] in hydrocarbon solvents results in the formation of exogenous Lewis-base-free amidoalanes [H₂Al(NR₂)]_n (n = 2 or 3) as crystalline solids (35–85% yield). In the solid state (seven X-ray structures), all the amidoalanes exist as dimers, with the exception of the pyrrolidine-derived alane which exists as a trimer. As solids, these amidoalanes exhibit significant kinetic stability towards oxygen/moisture allowing the brief (ca. 5 min.) handling of [H₂Al(NⁱPr₂)]₂ in air.

1. Introduction

The utilisation of organoaluminium reagents in both academia and industry is common practice, especially in reductive chemistries, such as the reduction of carbonyl groups and the metalation of unsaturated C=C/C≡C bonds. Despite their prevalence, these reagents (even ubiquitous LiAlH₄ to some extent) are frequently pyrophoric in nature, meaning special precautions need to be taken for their safe use. Over the last 30 years, there have been significant efforts made towards stabilizing such air-sensitive compounds, thus making them easier and safer to work with in the laboratory. The first reported examples of air-stabilised aluminium reagents were in 1996, where one of the alkyl groups in R₃Al was substituted by a group containing a chelating moiety, which imparts stability via the intramolecular formation of a Lewis acid/base pair (Figure 1, I and II) [1,2].

Since that seminal work, the stabilisation of organoaluminium reagents via Lewis complexation has been the go-to approach to in acquiring user-friendly reagents. In 2005, we popularised the stabilisation of AlMe₃ using DABCO (DABCO = 1,4-diazabicyclo[2.2.2]octane) to form DABAL-Me₃ (III) [3], which shows remarkable air stability, comparable to LiAlH₄. This complexation with DABCO was later expanded upon to furnish an equivalent PhAlI₂ reagent (IV) [4]. However, this complex displayed much lower stability in air, and it could only be handled on the bench for short periods of time. Whilst significant progress has been achieved stabilizing alkyl/aryl organoaluminium reagents, the stabilisation of aluminium hydrides has lagged behind. One reason for this is that LiAlH₄, the most ubiquitous aluminium hydride, already displays reasonable air stability and handling it in air (ca. 5–30 min) does not usually compromise its reactivity, unless aerobic moisture is present [6]. A breakthrough in developing other air-stabilised alanes was reported by us in 2012, where we described simple ethereal Lewis bases, such as THF, as stabilisers of dichloroalane, allowing it to be handled in air without any significant impact on reactivity for 30 min (V) [5]. Amidoalanes (1) constitute an alternative interesting class of aluminium hydrides as they contain an endogenous Lewis basic moiety. Early studies found that oligomeric species bridged through amido nitrogen atoms are obtained when small substituents are bound to the nitrogen centre. Increasing the steric hinderance enables the isolation of typically dimer species, unless the amine substituents are sterically compact furnishing trimeric species.

The most common routes to species 1 are either the direct reaction of the amine with AlH₃•NR₃ (R = Me, Et), the reaction of the amidoaluminium dichloride with LiH, or the generation of AlH₃ through the in situ protonation of LiAlH₄ via the ammonium salt of the amine [7,8,9,10,11,12,13,14]. Herein, we report a facile, one-pot synthesis of a range of exogenous Lewis-base free amidoalanes [H₂Al(NR₂)]_n (n = 2–3) and explore the properties of these compounds. Conveniently, multigram quantities of these reagents are easily attained, without the need to use pyrophoric AlH₃.

2. Results and Discussion

2.1. Synthesis of Amidoalanes

Based on our previous work on synthesizing dichloroalane from LiAlH₄ and AlCl₃ (1:3 stoichiometry), we reasoned that with appropriate LiAlH₄ and AlCl₃ stoichiometries, we could generate ‘AlH₃’ in situ and we found that this approach would be advantageous as the free volatile primary alane would not be lost from the reaction mixture if it was already datively bound to an amine donor. The choice of reaction solvent mixture for this procedure (Scheme 1) is crucial. The premixing of LiAlH₄ and AlCl₃ in petroleum solvents (n-hexane or n-pentane) leads to no reaction due to the insolubility of these species in such solvents. However, upon the slow addition of the sec-amine, even at 0 °C, the mixture solubilises and the rapid formation of the amidoalane is evidenced by the evolution of hydrogen gas. This approach completely minimises the formation of free AlH₃ in the reaction mixture, greatly simplifying the scale-up of the reaction. On completion of the reaction, workup is also facilitated by the use of the hydrocarbon solvent. The only by-product, LiCl, is completely insoluble and is easily separated via cannula filtration under an inert atmosphere, without the need for glovebox procedures. The results from six representative amines are shown in Scheme 1.

This synthetic route is attractive as it can be carried out with small solvent volumes on large scales (up to 20 g is achieved here). Additionally, in most cases, compounds 1a–f can also be prepared using LiAlH₄, AlCl₃, and petroleum solvents used directly as supplied, with only the simple (4 Å molecular sieve) drying of secondary amines. This approach typically results in a slight reduction in the isolated yield (10–15%) but does allow the isolation of large quantities of 1 used as a reagent in synthetic organic chemistry, without the requirement for specialist equipment. The structure of 1c is known in the Cambridge Structural Database (CSD), but no experimental data have been previously published [8]. Since it was reported to be an ‘unstable solid’ decomposing above 0 °C, it is unlikely that 1c constituted the bulk material in this report [8].

Attempts to extend the chemistry highlighted in Scheme 1 to include chelating diamines, RNHCH₂CH₂NHR (R = Me, tBu) were unsuccessful, resulting in only complicated mixtures. However, by changing the stoichiometry of LiAlH₄:AlCl₃:HN(ⁱBu)₂ to 3:1:8, we were able to prepare the dimeric monohydrido species [HAl(NⁱBu₂)₂]₂ (2) in a 58% yield (Scheme 2). In this case, extended heating (48 h vs. 1–2 h) was necessary to convert intermediate 1d to 2 in this moderate yield.

2.2. Spectroscopic Studies of 1a–f

The ¹H NMR spectra of compounds 1a–f in benzene-d₆ clearly show the presence of the amido ligands. The hydride resonance can be identified via very broad resonances in the 4.1–4.6 ppm range (

Table 1. Selected spectroscopic data for compounds 1a–f and 2.

Compound	Et (1a)	iPr (1b)	-(CH2)5- (1c)	iBu (1d)	Cy (1e)	-(CH2)4- (1f)	iBu (2)
1H NMR (δAl-H) (ppm)	4.13	4.30	4.14	4.31	4.47	4.14	4.31
27Al NMR (ppm)	150.2	143.5	150.3	147.0	-	144.9	131.1
IR (νAl-H) (cm–1)	1826	1816	1794	1823	1834	1807	1833

). All these signals are appreciably broadened due to the associated aluminium quadrupole. The ²⁷Al NMR spectra for 1a–f in benzene-d₆ all show a single broad resonance in the 145 ppm region. These resonances fall into the characteristic region for tetrasubstituted aluminium centres with nitrogen ligands. Interestingly, despite being able to observe an Al–H resonance in the ¹H NMR spectrum for 1e, we were unable to observe any signals in the ²⁷Al NMR spectrum. The ²⁷Al NMR spectrum for compound 2 exhibits a singlet which is shifted up-field by ca. 15 ppm compared to the dihydride analogue 1d, which is characteristic for diamidoalane compounds [9]. It should be noted that whilst the compounds can be isolated cleanly in the solid state, if they are left in solution for prolonged periods of time, degradation and potential ligand scrambling to H₃Al•N(H)R₂ can be observed.

In further support of their formulation as alanes, a signal assigned to an Al–H stretch can be detected in every case in the region 1835–1794 cm⁻¹ in their IR spectra. The Al–H stretching frequency correlates well with the degree of steric hindrance experienced by the localised Al–H bond.

2.3. Single-Crystal X-Ray Crystallographic Studies

An attractive feature of the amidoalanes 1a–e is their high crystallinity which greatly facilitates their isolation and purification. To confirm their identities, their purities (the complexes are sometimes too reactive for routine CHN combustion analyses) and to determine their degree of aggregation in the solid state, X-ray structural studies on all of the compounds isolated were carried out. The structures of the amidoalanes 1a–b, 1d–f, and 2 are shown in Figure 2. The X-ray structure of 1c is identical to that crystallographically reported (but without any supporting spectroscopic data) in the literature [8]. Comparative structural data for the dimeric compounds 1a–e and 2 are given in

Table 2. Comparison of selected amidoalane structures 1a–f and 2 ^a.

Structural Parameter	Et (1a)	iPr (1b)	-(CH2)5- (1c) a	iBu (1d)	Cy (1e)	-(CH2)4- (1f)	iBu (2)	CSD Ave. (N = 17) b
Al—N(1)/Å	1.946(2)	1.9631(12) 1.9640(12)	1.94	1.964(2) 1.968(2) 1.970(2) 1.971(2)	1.961(3) 1.978(3)	1.862(2) 1.9343(5) 1.9354(5) 1.991(2)	1.975(3) 1.991(3)	1.95
Al—N(2)/Å	-	-	-	-	-	-	1.820(3)	1.89
Al—H/Å	1.56(4) 1.60(4)	1.46(2) 1.46(2)	1.75	1.42(3) 1.47(3) 1.78(3) 1.84(3)	1.48(4) 1.55(3)	1.490(8) 1.497(11) 1.520(7) 1.520(10)	1.43(3)	1.67
Al··Al/Å	2.756(2)	2.7499(9)	2.76	2.7442(10)	2.759(2)	3.2442(3) 3.2452(3) 3.2452(3)	2.8197(19)	2.84
R–N(1) c/Å	1.488(2) 1.492(3)	1.515(2) 1.517(2)	1.49	1.491(3) 1.493(3) 1.496(3) 1.498(3)	1.508(4) 1.509(4)	1.5203(8) 1.5298(7)	1.497(4) 1.498(4)	1.50
N(1)–Al–N(1)/o	89.86(8)	91.10(4)	89.5	91.48(8) 91.72(8)	91.07(10)	108.48(3) 107.17(3) 110.84(3)	89.39(11)	84.9
Al–N(1)–Al/o	90.14(8)	88.90(4)	90.5	88.47(8) 88.33(8)	88.93(10)	113.99(3) 114.64(3)	90.61(11)	93.5
N(2)–Al–H/o	-	-	-	-	-	-	116.7(13)	117.5
R–N(1)–R/o	111.9(2)	119.38(10)	109.1	112.2(2) 112.5(2)	120.4(2)	105.19(4) 105.99(6)	113.5	110.1
H–Al–H/o	102.1(17)	122.3(10)	121	113.2(14) 118.0(14)	122(2)	116.7(4) 118.6(5)	-	108.3

^a CSD structure QAGNAO. ^b Average of all 17 structures containing H₂Al(μ-NY₂)AlH₂ fragments in the CSD database (Y = any substituent). ^c Measured at the α-C atom.

.

The more sterically encumbered members of the amidoalanes, 1b, 1d–e, 1f, and 2, are all essentially planar. The bridging amido geometries of these dimeric species are all highly similar to one another. The monohydride dimer (2) shows the longest Al^…Al interatomic distance indicative of the greater steric congestion in this compound. This idea is reinforced in the ¹H NMR spectrum of this compound which shows restricted rotation around the N-CH₂ bond of the bridging amide (as these appear as separate signals at 22 °C). Remarkably, pyrrolidine-derived 1f was found to be trimeric in the solid state in our studies. This is in line with expectations for this class of alane derivatives, as sterically compact amido bridges are known to favour higher degrees of oligomerisation [15]. The change in aggregation has little effect on the H–Al–H angle. The values attained in trimeric 1f (ca. 117–118°) being mid-range to that observed in Table 2.

Technical issues aside (difficulties in accurate detection and the small number of examples of this motif in the CSD database), the Al–H distances in this family of molecules deserve comment. Firstly, a wide range of Al-H distances (1.37–1.84 Å) are observed. The average Al–H for Table 2 (1.54 Å) is shorter than pre-existing CSD database averages, both for the most related molecules (1.67 Å) [10] and for all Al–H distances in AlH₂ motifs (1.66 Å). The observed distances are, in most cases, similar to those in [H₂Al(N(SiHMe₂)₂]₂ (1.49, 1.52 Å) [11]. In one case, significant differences in the Al–H bond distances within the AlH₂ unit have been detected (1d). Control experiments and checks excluded the following sources for this irregularity: the presence of small amounts of a co-crystallised H₂Al(μ-NⁱBu₂)₂AlHX (X = Cl or other atom) through partial Al–C displacement; a non-representative crystal; the partial conversion of 1d to 2 in the reaction mixture and its subsequent co-crystallisation; and an insufficiently modelled minor disorder in the structure. At present, we are unable to explain the origin of this effect, and further investigations are continuing in our laboratory in light of this interesting observation.

2.4. Air Stability Studies

The air stability of compounds 1a–f is improved by the presence of large NR₂ groups, but these compounds are still rather air-sensitive when in powder forms and should be stored in a glovebox. By preparing the samples in a glovebox, we could attain accurate CHN analyses on 1d and 1e. Despite repeated attempts, the other compounds showed significant deviations (~2%) on carbon analyses. Nevertheless, we investigated the use of these species in air as reagents and measured their air stability by way of gas evolution analysis using a gas burette (see ESI). Some of these compounds were proven to be unusable in air; for example, 1c is rather pyrophoric in the crystalline solid state. Usefully, however, compound 1b crystallises in larger colourless tablets, and, in this form, it shows some degree of air tolerance (Figure 3). While this is not as good as in DABAL-Me₃, it does usefully allow for its brief weighing in the open lab [3]. Some (~5%) surface decomposition is evidenced, but the use of 1b as a potential organic reagent under these conditions is clearly practical, and this is being investigated further.

2.5. Reactivity Studies

Having found that compound 1b exhibits some air stability, we wanted to test the reactivity of this amidoalane in reduction chemistry (Scheme 3). Compound 1b readily reduced aldehydes and ketones to the corresponding alcohols in good yields and was also able to reduce benzoic acid to benzyl alcohol when two equivalents of the alane were employed. Our groups have a long-standing interest in transition-metal-catalysed hydroalumination reactions and reasoned that 1b would be a viable aluminium hydride source for this transformation. To our delight, when 1b was reacted with a series of olefins under Cp₂TiCl₂ catalysis, the corresponding alkanes were furnished in a good-to-excellent yield, upon quenching with aqueous HCl. If the electrophilic quenching reagent is changed from H₂O/HCl to molecular oxygen, the corresponding alcohol product can be obtained in a moderate yield.

3. Materials and Methods

General: All procedures were carried out under purified argon in dried solvents (n-hexane and n-pentane dried over LiAlH₄). Lithium aluminium hydride (Alfa Aesar, 97%, a fine white powder, as supplied) was recrystallised from diethyl ether. For the preparation of diisopropylamidoaluminium dihydride and diisobutylamidoaluminium dihydride, it was found that the same LiAlH₄ source could be used directly in the preparation, leading to only a slight (10–15%) reduction in the yield compared with recrystallised LiAlH₄. Aluminium trichloride (Fluka, >99%, cat. number: 06220, pale yellow) was used as purchased. The secondary amines (diethylamine—Alfa Aeser; diisobutylamine, piperidine—Sigma Aldrich, diisopropylamine—Acros, and pyrrolidine—Fluka) were distilled from molecular sieves (4 Å) before use.

Instrumentation: ¹H, ¹³C{¹H}, and ²⁷Al NMR spectra were recorded on a Bruker advance III HD 400 spectrometer (operating frequencies: 400.20 MHz, 128.25 MHz, and 104.27 MHz, respectively). ¹H and ¹³C{¹H} NMR chemical shifts were internally referenced to the following residual solvent resonances C₆D₆ (benzene-d₆): ¹H δ = 7.16 ppm and ¹³C{¹H} δ = 100.64 ppm. NMR samples were prepared under an inert atmosphere in 5 mm J. Youngs NMR tubes. Data were analysed using MestReNova V14.0.0 software. IR spectra were recorded as Nujol mull in KBR disks using a Bruker Tensor 27 spectrometer. The assignment of NMR spectra was based on HMBC, HMQC, ROESY, and DEPT135 techniques as appropriate.

The CCDC deposition numbers are as follows: 1a—783514; 1b—783515; 1d—783516; 1e—783517; 1f—783518; and 2—783519.

• Syntheses of dialkylamidoaluminium dihydrides [H₂Al(NR₂)]_n 1

Solid LiAlH₄ (1.00 g, 26.4 mmol) and AlCl₃ (1.17 g, 8.8 mmol) were suspended in 60 mL of n-hexane or n-pentane. The mixture was cooled to 0 °C using an ice bath. Dry secondary amine (34.0 mmol) was added slowly via a rubber septum using a syringe. During the addition, gas formation (H₂) was observed. The connection to the Schlenk line vent was kept open in order to avoid a build-up of pressure. After the complete addition of the amine, the rubber septum was replaced (under a flow of argon) by a condenser equipped with a gas bubbler. The mixture was heated at reflux until the gas evolution stopped (typically 1–2 h). After cooling the supernatant, liquids were removed from the solid residue via cannular filtration. The solvent was removed under vacuum to give the crude products. In some cases (1b and 1d), the products began to crystallise upon the partial removal of the petroleum solvents. Alternatively, the crude dialkylamidoaluminium dihydrides were purified by being recrystallised from n-hexane or n-pentane. Remarkably, diethylamidoaluminium dihydride 2 was purified by way of vacuum sublimation (T_sub 50–60 °C, 1 mmHg). The preparative scale could be increased without problems to at least 20 g and LiAlH₄ and AlCl₃ could be used as supplied with only a small reduction in yield.

• [H₂Al(NEt₂)]₂ 1a. Yield 37%; colourless crystals (blocks); mp. (sealed capillary) 41–42 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 0.86 (t, 12 H, CH₃, ³J_HH = 7.1 Hz), 2.84 (q, 8 H, CH₂, ³J_HH = 7.1 Hz), and 4.13 (br., 4 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 12.3 (CH₃) and 42.3 (CH₂); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 150.2; IR (NaCl plates, nujol) ν/cm⁻¹ = 3446 (vs, br.), 2932 (vs, nujol), 1826 (vs, ν_AlH), 1453 (vs), 1382 (vs), 1315 (w), 1291 (m), 1261 (w), 1175 (s), 1111 (vs), 1046 (s), 1004 (s), 903 (m), 854 (s), and 732 (s); MS (EI, 70 eV): m/z 201 (38%) (M⁺–H), 187 (13%) (M⁺–CH₃), 173 (50%) (M⁺–C₂H₅), 171 (63%) (M⁺–C₂H₅–2H), 157 (77%) (M⁺–CH₃–C₂H₅), and 100 (100%) (monomer–H). CARE! This compound is somewhat pyrophoric in the solid state.
• [H₂Al(NⁱPr₂)]₂ 1b. Yield 88%; colourless crystals (blocks); mp. (sealed capillary) 133–134 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 1.29 (d, 24 H, CH₃, ³J_HH = 7.0 Hz), 3.56 (sept., 4 H, CH, ³J_HH = 7.0 Hz), and 4.30 (br., 4 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 24.7 (CH₃) and 50.7 (CH); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 143.5; IR (NaCl plates, nujol) ν/cm⁻¹ = 3585 (m), 3413 (m), 3318 (m), 2925 (vs, nujol), 2600 (m), 2303 (w), 1816 (vs, ν_AlH), 1592 (m), 1448 (vs), 1390 (vs), 1314 (m), 1261 (w), 1175 (vs), 1126 (vs), 977 (w), 956 (s), 915 (s), and 840 (s); MS (EI, 70 eV): m/z 257 (23%) (M⁺–H), 243 (8%) (M⁺–CH₃), 227 (20%) (M⁺–2CH₃–H); 213 (20%) (M⁺–3CH₃); 129 (6%) (monomer), 128 (100%) (monomer–H), 126 (30%) (monomer–3H), 114 (42%) (monomer–CH₃), and 86 (57%) (monomer–C(CH₃)₂).
• [H₂Al(NC₅H₁₀)]₂ 1c. Yield 69%; colourless crystals (blocks); mp. (sealed capillary) 91–92 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ ppm = 1.11–1.17 (m, 4 H, C4-H₂), 1.33–1.38 (m, 8 H, C3-H₂), 2.72–2.75 (m, 8 H, C2-H₂), and 4.14 (br., 4 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 24.1 (C4), 27.0 (C3), and 51.6 (C2); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 150.3; IR (NaCl plates, nujol) ν/cm⁻¹ = 3436 (vs, br.), 2923 (vs, nujol), 2499 (m), 2385 (w), 2314 (w), 2257 (w), 2209 (w), 2077 (w), 1990 (w), 1823 (vs, br., ν_AlH), 1643 (m), 1504 (s), 1452 (vs), 1373 (s), 1312 (m), 1281, (m), 1257 (m), 1190 (s), 1152 (s), 1086 (s), 1030 (vs), 940 (s), 856 (s), and 806 (s); MS (EI, 70 eV): m/z 226 (45%) (M⁺), 225 (52%) (M⁺–H), 224 (24%) (M⁺–2H), 197 (68%); 195 (46%), 113 (8%) (monomer); 112 (50%) (monomer–H), 85 (51%) (piperidine), and 84 (100%) (piperidine–H), 57 (43%), 56 (57%).
• [H₂Al(NⁱBu₂)]₂ 1d. Yield 90%; colourless crystals (blocks); mp. (sealed capillary) 66–67 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 0.90 (d, 24 H, CH₃, ³J_HH = 6.7 Hz), 1.80–2.00 (m, 4 H, CH), 2.90 (d, 8 H, CH₂, ³J_HH = 6.9 Hz), and 4.31 (br., 4 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 22.0 (CH₃), 27.2 (CH), and 57.0 (CH₂); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 147.0; IR (NaCl plates, nujol) ν/cm⁻¹ = 3384 (vs, br.), 2958 (vs, nujol), 1834 (vs, ν_AlH), 1465 (vs), 1391 (s), 1314 (m), 1268 (m), 1156 (s), 1134 (s), 1082 (vs), 1018 (vs), and 940 (s); MS (EI, 70 eV): m/z 313 (14%) (M⁺–H), 311 (13%) (M⁺–3H), 283 (19%) (M⁺–2CH₃–H), 271 (23%) (M⁺–C(CH₃)₂), 269 (20%) (M⁺–CH(CH₃)₂–H), 156 (80%) (monomer –H), 154 (100%) (monomer–3H), and 112 (34) (monomer–CH(CH₃)₂–H); elemental analysis: anal. calc. for C₁₆H₄₀Al₂N₂: C 61.1, H 12.8, N 8.9, found C 61.4, H 12.5, N 9.0%.
• [H₂Al(NCy₂)]₂ 1e. 85%; colourless crystals (blocks); mp. (sealed capillary) 193 °C (decomposition, yellow); ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 0.95 (dtt, 4 H, C4-H_ax, ²J_HH = 13.0 Hz, ³J_HH = 13.0, 3.1 Hz), 1.14–1.29 (m, 8 H, C3-H_ax), 1.47 (app. d, 4 H, C4-H_eq, ²J_HH = 13.2 Hz, the small ³J_HH was not fully resolved), 1.57–1.71 (m, 16 H, C3-H_eq overlapped by C2-H_ax), 2.23 (app. d, 8 H, C2-H_eq, ²J_HH = 11.6 Hz, the small ³J_HH couplings were not fully resolved), 3.21 (dt, 4 H, C1-H_ax, ³J_HH = 11.7, 2.6 Hz), and 4.47 (br., 4 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 25.8 (C4), 26.5 (C3), 35.7 (C2), and 61.1 (C1); IR (NaCl plates, nujol) ν/cm⁻¹ = 3428 (vs, br.), 2924 (vs, nujol), 2664 (s), 1806 (vs, br., ν_AlH), 1630 (m), 1448 (vs), 1368 (s), 1345 (m), 1259 (s), 1109 (vs), 1025 (vs), 888 (s), and 800 (vs); MS (EI, 70 eV): m/z 417 (10%) (M⁺–H), 387 (13%), 333 (34%) (M⁺–cyclohexyl–H), 208 (100%) (monomer–H), 206 (50%) (monomer–2H), and 138 (50%); elemental analysis: anal. calc. for C₂₄H₄₈Al₂N₂: C 68.9, H 11.6, N 6.7, found C 69.3, H 11.4, N 6.8%.
• [H₂Al(NC₄H₈)]₃ 1f. Yield 62%; colourless crystals (needles); mp. (sealed capillary) 106–107 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 1.44–1.69 (m, 12 H, NCH₂), 2.95–3.40 (m, 12 H, CH₂), and 4.14 (br., 6 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 25.7 (NCH₂) and 51.7 (CH₂); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 144.9; IR (NaCl plates, nujol) ν/cm⁻¹ = 3515 (vs, br.), 2923 (vs, nujol), 2359 (m), 1794 (vs, br., ν_AlH), 1456 (vs), 1342 (w), 1291 (w), 1256 (w), 1178 (m), 1102 (s), and 1036 (vs); MS (EI, 70 eV): m/z 297 (5%) (M⁺), 296 (44%) (M⁺–H), 295 (11%) (M⁺–2H), 268 (11%) (M⁺–AlH₂), 266 (16%) (M⁺–AlH₄), 264 (10%) (M⁺–AlH₆), (11%) (M⁺–2H), 219 (37%), 198 (33%) (dimer), 197 (79%) (dimer–H), 196 (28%) (dimer–2H), 169 (68%) (dimer–AlH₂), 167 (70%) (dimer–AlH₂), 131 (22%), 99 (8%) (monomer), 98 (100%) (monomer–H), 70 (18%) (pyrrolidine), and 69 (50%) (pyrrolidine–H).

3.1. Synthesis of [HAl(NⁱBu₂)₂]₂ 2

All chemicals were used as supplied. Solid LiAlH₄ (3.0 g, 79.1 mmol) and AlCl₃ (3.51 g, 26.4 mmol) were suspended in 150 mL of n-hexane. The mixture was cooled to 0 °C using an ice bath. Dry diisobutylamine (36.7 mL, 210.8 mmol) was added slowly via a rubber septum using a syringe. During the addition, gas formation (H₂) was observed. The connection to the argon line was kept open in order to avoid a build-up of pressure. After complete addition of the amine, the rubber septum was replaced (under a flow of argon) using a condenser equipped with a gas bubbler. The mixture was heated at reflux for 48 h. After cooling, the supernatant liquids were removed from the solid residue by way of cannular filtration. The solvent was removed in a vacuum to produce the crude product. The product began to crystallise upon the partial removal of n-hexane. The crude bis(diisobutylamido)aluminium hydride was purified by being recrystallised from n-hexane.

• Yield 17.5 g, 58%; colourless crystals (blocks); mp. (sealed capillary) 107–108 °C; ¹H NMR (400.20 MHz, 298 K, C₆D₆): δ/ppm = 0.96 (d, 12 H, CH₃ bridged, ³J_HH = 6.6 Hz), 1.04 (d, 24 H, CH₃ term., ³J_HH = 6.5 Hz), 1.18 (d, 12 H, CH₃ bridged, ³J_HH = 6.7 Hz), 2.07 (m, 4 H, CH term.), 2.25 (m, 4 H, CH bridged), 2.97 (d, 8 H, CH₂ term., ³J_HH = 6.0 Hz), 3.04 (dd, 4 H, CH₂ bridged, ²J_HH = 14.9 Hz, ³J_HH = 5.6 Hz), 3.12 (dd, 4 H, CH₂ bridged, ²J_HH = 13.3 Hz, ³J_HH = 6.7 Hz), and 4.31 (br., 2 H, Al-H); ¹³C NMR (100.64 MHz, 298 K, C₆D₆): δ/ppm = 21.7 (CH₃), 22.9 (CH₃), 23.1 (CH₃), 27,4 (CH), 27.8 (CH, bridged), 55.7 (CH₂), and 57.4 (CH₂, bridged); ²⁷Al NMR (104.27 MHz, 298 K, C₆D₆): δ/ppm = 131.1; IR (NaCl plates, nujol) ν/cm⁻¹ = 3378 (vs, br.), 3175 (vs), 2951 (vs, nujol), 1833 (vs, νAlH), 1464 (vs), 1383 (vs), 1314 (m), 1298 (m), 1262 (s), 1154 (vs), 1102 (vs), 1064 (vs), 1011 (vs), 977 (s) 950 (s), 925 (m), and 862 (s); MS (EI, 70 eV): m/z 526 (13%), 442 (13%), 441 (38%) (M⁺–NⁱBu₂), 399 (23%), 398 (16%), 314 (45%), 313 (49%), and 154 (100%).

3.2. X-Ray Data Collection, Solution, and Refinement

Crystals were mounted in the cold stream of an Oxford Cryosystems open-flow cryostat on a Bruker SMART APEX CCD area detector diffractometer; all data were collected at 150 K. The structures were solved using direct methods and refined by full-matrix least-squares on F2. Crystal and refinement data are summarised in

Table 3. Crystal and refinement data for 1a, 1b, 1d–1f and 2.

Compound	1a	1b	1d	1e	1f	2
formula	C8H24Al2N2	C12H32Al2N2	C16H40Al2N2	C24H48Al2N2	C12H30Al3N3	C32H74Al2N4
Fw	202.25	258.36	314.46	418.60	297.33	568.91
crystal system	monoclinic	monoclinic	orthorhombic	triclinic	monoclinic	monoclinic
space group	P21/c	C2/c	P212121	Pī	P21/m	P21/n
wavelength [Å]	0.71073	0.71073	0.71073	0.71073	0.71073	0.71073
temperature/K	150(2)	150(2)	150(2)	150(2)	150(2)	150(2)
a, Å	7.351(1)	14.514(3)	9.1662(7)	6.613(1)	6.332(1)	10.377(2)
b, Å	12.859(1)	15.439(3)	14.3798(10)	12.546(1)	15.113(1)	8.685(1)
c, Å	7.219(1)	8.661(2)	15.8971(11)	15.162(1)	9.274(1)	21.34(1)
α, deg	90	90	90	98.28(1)	90	90
β, deg	90.27(1)	122.23(1)	90	92.08(1)	103.33(1)	99.29(1)
γ, deg	90	90	90	90.12(1)	90	90
V, Å3	682.3(1)	1641.8(6)	2095.4(3)	1244.0(1)	863.6(1)	1898.1(5)
Z	2	4	4	2	2	2
crystal dimensions/	0.10	0.17	0.40	0.12	0.19	0.04
mm	0.30	0.27	0.49	0.19	0.44	0.18
	0.40	0.57	0.63	0.32	0.50	0.20
crystal shape	block	tablet	block	tablet	tablet	plate
ρcalc, Mg/m3	0.984	1.045	0.997	1.118	1.143	0.995
Θ range/°	2.77–25.04	2.12–27.48	2.56–25.1	2.28–27.26	2.26–27.53	1.93–25.05
μabs [mm−1]	0.177	0.160	0.023	0.129	0.209	0.100
F(000) [e]	224	576	704	464	324	640
index ranges	−8 < h < 8	−13 < h < 18	−10 < h < 10	−8 < h < 8	−7 < h < 7	−12 < h < 12
	−15 < k < 15	−19 < k < 20	−15 < k < 17	−16 < k < 16	−17 < k < 17	−10 < k < 10
	0 < l < 8	−11 < l < 9	−16 < l < 18	−19 < l < 19	−11 < l < 11	−24 < l < 25
no. meas. refl.	6599	5118	10,851	14,659	9008	13,559
no. indep. reflns	1215	1861	3707	5591	1587	3372
Rint	0.032	0.022	0.023	0.081	0.019	0.062
no. reflns. observed	1027	1578	3424	3894	1502	2700
parameter/restraints	64/0	85/0	202/0	269/0	116/0	184/0
GooF (F2)	1.06	1.09	1.05	1.12	1.08	1.09
R [I > 2σ(I)]/wR2	0.048/0.125	0.061/0.169	0.0466/0.128	0.085/0.171	0.030/0.085	0.079/0.163
R1 (all)/wR2	0.060/0.137	0.067/0.176	0.0497/0.131	0.118/0.184	0.031/0.087	0.101/0.172
ρrest (max/min) [e·nm−3]	582/−131	1038/−212	1020/−160	502/−348	263/−142	414/−270

and full details are given in the Supplementary Data.

• Hydroalumination of carbonyl compounds: In a nitrogen-filled glovebox, an oven-dried Schlenk tube was charged with amidoalane 1b (0.225 mmol), carbonyl compound (0.205 mmol), and THF (1 mL). The flask was removed from the glovebox and stirred for 2 h at room temperature. After 2 h, the reaction was quenched with HCl_(aq) and the layers separated, and the aqueous phase was extracted with diethyl ether (3 × 5 mL). The yield was determined by ¹H NMR spectroscopy using 1,1,2,2-tetrachloroethane as the internal standard.
• Hydroalumination of olefins: In a nitrogen-filled glovebox, an oven-dried Schlenk tube was charged with amidoalane 1b (0.225 mmol), olefin (0.205 mmol), Cp₂TiCl₂ (2.55 mg, 0.01 mmol), and THF (1 mL). The flask was removed from the glovebox and stirred for 3 h at room temperature. After 3 h, the reaction was either quenched with HCl_(aq) or O₂-bubbled through the reaction and then quenched with HCl_(aq). The layers were separated, and the aqueous phase was extracted with diethyl ether (3 × 5 mL). The yield was determined by ¹H NMR spectroscopy using 1,1,2,2-tetrachloroethane as the internal standard.

4. Conclusions

The direct reaction of LiAlH₄, AlCl₃, and secondary amines constitutes a highly effective and economical route to dialkylamidoalanes. The isolation and handling of these materials are facilitated by their high crystallinity. This preparative route offers advantages over traditional routes to these compounds using H₃Al•NR₃ adducts that have led, in some cases in the past, to reports of partially or even erroneously characterised materials. Some of these hydrides have great potential to be used as reagents in organic chemistry due to their high solubility and, in some cases (1b), usable levels of air stability. Compound 1b was found to reduce a range of carbonyl containing substrates and undergo a titanium-catalysed olefin hydroalumination.