This work has been digitalized and published in 2013 by Verlag Zeitschrift für Naturforschung in cooperation with the Max Planck Society for the Advancement of Science under a Creative Commons Attribution4.0 International License.

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschungin Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung derWissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht:Creative Commons Namensnennung 4.0 Lizenz.

Alkoxohydridoaluminates. I 27AI NMR Characteristics of (C4H9)4N+AlH4_„(OR)„" in Benzene Stanislav Heřmánek*, Jiří Fusek, Otomar Kříž, Bohuslav Čásenský, and Zbynĕk Černý Institute of Inorganic Chemistry, Czechoslovak Academy of Sciences, 250 68 Rez near Prague, Czechoslovakia Z. Naturforsch. 42b, 539-545 (1987); received October 1, 1986

Tetrahydridoaluminates, Alkoxohydridoaluminates, Tetraalkoxoaluminates, 27A1 NMR Spectra

The 27A1 NMR spectra of (C4H9)4N+AlH4_n(OR)„- prepared in the reaction of (C4H9)4NA1H4 with ROH (R - CH3, C2H5, /-C3H7, r-C4H9, «-C5HU , n-C9H19, cyclo-C6Un, C6H5, CH3OCH2CH2) in benzene show — in contrast to the Li+ and Na+ salts — a distinct H-splitting, which has allowed to ascertain: 1) the presence of all members (n = 0—4) in solution, 2) the Al chemical shift and AI —H coupling constant of all individual classes, namely (n = 0): 101.6 ppm, quintet 174 Hz; (n = 1): 114-123 ppm, quartet 184-193 Hz; (n = 2): 103-112 ppm, triplet 197-219 Hz; (n = 3): 75-91 ppm, doublet 218-237 Hz; (n = 4): 5 1 - 7 3 ppm, singlet, 3) in-creased shielding in the order: CH3 < n-R = i-R = sec-R < C6HU C6H5 < r-C4H9. The anomalous "sagging" pattern of the curve: d27Al vs. number of OR groups, is interpreted on the basis of significant differences among Al and H and OR in electronegativity.

Introduction

Alkoxohydridoaluminates (AHA), the 2—4 mem-bers of the AlH4_„(OR)„~ family (1 n = 0, 2 n = 1, 3 n = 2, 4 n = 3, 5 n = 4), are regarded either as intermediates in the reduction of the carbonyl group with tetrahydridoaluminates or as easily accessible reducing agents. The vision of tailoring reducing agents and modifying their reducing power for chemo-, regio-, stereo- or enantio-selective reduc-tions of selected groups in molecules has been the reason for great interest in AHA [1, 2]. These expec-tations have been, however, only partially fulfilled due to the inadequate knowledge of the stability and the corresponding disproportionations of AHA.

Alkoxohydridoaluminates are prepared by the use of one of the three general methods:

A. The alcoholysis of tetrahydridoaluminates [3-5]:

MA1H4 MAlH3(OR) MAlH2(OR)2 — H2 — rl2 — rl2 MAlH(OR)3 R S H > MAl(OR)4 (1) " " 2

B. The reaction of tetrahydridoaluminates with ketones R - C O - R ' (R, R' = alkyl, aryl) in solu-tions [6]:

* Reprint requests to Dr. S. Hermanek.

Verlag der Zeitschrift für Naturforschung, D-7400 Tübingen 0932-0776/87/0500-0539/$ 01.00/0

MA1H4 R C Q R> MAlH 3 (OCHRR') r c q r > R —PO —R '

MAlH2(OCHRR')2 > D _Pf)_R '

MAlH(OCHRR') 3 > MAl(OCHRR') 4 (2)

C. The exchange of ligands among the individual I—V members [3]:

MA1H4 + MAI (OR)4 * MAlH3OR + MAlH(OR)3 , etc. (3)

The mutual exchange of ligands (Method C), in contrast to the irreversible procedures A and B, has the character of an equilibrium reaction, and that is why the less stable AHA can disproportionate to more stable ones [3, 7, 8]. The stability of salts con-taining the 2—4 anions is relatively low, and depends on many factors and, thus e.g. in the LiT salts, it drops from 4 to 2 [7],

For the comprehension of the regio- and stereo-specificity of reductions with AHA the following must be known:

1) The actual composition and mutual ratio of components 1—5 in solutions of a given AI: H : OR ratio,

2) the rate of mutual ligand exchange among the present species,

3) the reducing power of individual 1—4/M+ in combination with the steric accessibility and reduci-bility of the given group in the substrate, and

4) the structure, i.e. steric requirements and, con-sequently, the stereospecificity of each reducing agent 1—4 present in the given solution.

540 St. H e r m ä n e k ^ a / . • Alkoxohydridoaluminates

Despite more than a thousand papers dealing with AHA [1,2] and the great effort of many teams [e.g. 3, 7 -20 ] , no contribution (except [21]) provide infor-mation on the real composition of the reducing solu-tions described by an appropriate stoichiometric M^AlH4_„(OR)„ - formula. At present, almost thirty years after the discovery of A H A only a rough knowledge of the disproportionation of some 2—4 exists, but the genuine composition of solutions showing definite A l : H : O R ratios are not known. This can be explained by the application of indirect physical methods for the analysis of AHA solutions, such as ebullioscopic molecular weight [9, 10] and conductance measurements [9, 11] or IR [12, 13] and

NMR spectroscopy [14, 15] which provided infor-mation on the properties of the bulk but not on the presence and ratio of individual components.

A promising method of monitoring the individual 1—5 species in solutions is 27AI NMR spectroscopy. Until now, however, only a few contributions exist dealing with the A H A species resulting from the reactions of LiAlH4 [15, 20, 21, 23, 24] or NaAlH4 [20, 21] with alcohols such as CH3OH [20, 23], C2H5OH [20], /-C3H7OH [20], r-C4H9OH [20, 23, 24], C9H1 9OH [20], CH3OCH2CH2OH [20, 25], with phenols [21] or with compounds containing the C = 0 group such as acetaldehyde [20], acetone [20], and ethyl acetate [20] in tetrahydrofuran. Sufficiently dis-tinct signals were observed only in the case of dispro-portionation products, namely the quintet of A1H4-

at approx. 100 ppm and the singlet of Al(OR)4~ in the region of 50—70 ppm [20, 21]. From the signals belonging to AHA only imperfectly resolved dou-blets of monohydridoaluminate 4 were observed in the case of methyl [20], ethyl [20], /-propyl [20] and t- butyl [20] derivatives. Other signals of the above-men-tioned compounds — if visible at all — were not H-split and, therefore, not assignable. Due to this fact, the real 27A1 chemical shifts of the 2 and 3 anions were unknown and, consequently, their presence in solutions was not determinable.

All 27A1 NMR spectra reported so far were ob-tained with instruments working at relatively low magnetic fields, i.e. at frequencies of 23.0 MHz [20, 21, 24] and 25.2 MHz [23],

The present study, therefore, focused on the prob-lem indicated in paragraph 1), i.e. on seeking a sys-tem which would allow unambiguously to reveal and specify all 1—5 members with basic alkyls R in solu-

tions by the 27A1 NMR spectroscopy at the frequency of 52.1 MHz.

Results and Discussion

A 27A1 NMR (52.1 MHz) study of the reaction products of LiAlH4 with alcohols in diethyl ether proved not to be very useful due to split signals — doublets — only with a few trialkoxohydridoalumi-nates (R = smaller «-alkyls, r-butyl, phenyl) (Fig. l a ) . In all other cases, the signals were broad and without lines splitting which did not allow any definite assignment. In contrast to the literature data [20, 21, 23], we have also found signals in the region above 100 ppm.

Similar, but somewhat better was the situation with the NaAlH4/nROH/monoglyme system in which the signals were sharper, better visible, but the H-splitting was observed only with the same signals as in the case of the lithium analogs (cf. Fig. lb ) .

a) LiAIH4 + 3 CH3OH

diethyl ether

b) NaAIH4 + 2 CH3OH monoglyme

Ai(H2o)6

160 120 80 40 ppm Fig. 1. 27A1 NMR (52.1 MHz) spectrum of the products of reaction: a) LiAlH4 + 3 CH3OH in diethyl ether at ambient temperature; b) NaAlH4 + 2 CH3OH in monoglyme at ambient temperature.

St. Hermanek et al. • Alkoxohydridoaluminates 541

An outstanding improvement has been achieved, however, by changing (C4H9)4N+ for the Li+ or Na+ cation in alkoxohydridoaluminates. All [AlH4_„(OR)„]N(C4H9)4 ion pairs are soluble not only in tetrahydrofuran (THF) or 1,2-dimethoxy-ethane (monoglyme) but also in aromatic hydrocar-bons in which the Al atom is not affected by the donicity of the solvent. Moreover, the (C4H9)4N+

cation, in contrast to Li+ or Na+ , cannot catalyze on OR/H exchange [26], thus increasing the stability and lowering the possibility of disproportionation of the individual 2—4 anions. The expected weak interac-tion of the 1—5 anions with NR4+ cation were con-firmed by the individual 27A1 NMR signals which were substantially narrower than those of Li+ or Na+

salts and showed a distinct H-splitting, correspond-ing to the number of H-atoms on the central aluminum atom (Fig. 2). This fact has made possible the revelation of all the 27Al signals present, to deter-mine their 27Al chemical shifts, as well as H-splittings and, consequently, to assign the signals to individual anions.

For following the effect of alkyls on the 27A1 NMR chemical shifts and on chemical equilibria of 1—5 [M+ = (C4H9)4N+] in solutions, basic series of alkoxohydridoaluminates with R = CH3 (a), C2H5 (b), /-C3H7 (c), r-C4H9 (d), n-C5H„ (e), n-C9H19 (f), cyclo-C,H„ (g), C6H5 (h), CH3OCH2CH2 (i), and

n = 1—4 were prepared by procedures A and C and their 27A1 NMR spectra were measured in benzene at 52.1 MHz using 10 mm NMR tubes sealed under ni-trogen.

Our2 7AI NMR study clearly shows that in the case of the (C4H9)4N+ cation, all five 1—5 members are always present in benzene solutions, irrespective of the alkyl R ( a - i ) , the ratio of A l : H : O R and the preparation method. The mutual ratios of 1—5 mem-bers, as determined from the proton decoupled spectra, did not change significantly with increasing temperature (20-80 °C), but in the case of bulky alkyls they were found to depend on the method of preparation. A detailed study on the effects influenc-ing the ratio of the 1—4 members will be a matter of a forthcoming paper.

Chemical shifts d27Al of the 1—5 members [M+ = (C4H9)4N+] depend negligibly on temperature and occur at 22 °C in the following regions: 1: 101.6 ± 0.3, 2: 114-123, 3: 103-112, 4: 75-91, 5: 51-73 ppm (cf. Table I).

The 2 < 3 < 1 < 4 < 5 sequence of signals ordered according to the increased shielding of the aluminum nucleus has been found to differ distinctly from those of the boron, carbon and silicon ZH4_„(OR)„ analogs, with which the tetrahydrido derivatives ( B H r [27], CH4 [28], SiH4 [29]) are strongly shielded, i.e. they resonate at lower frequencies than

(C4H9)4N AIH4 + 2 CH3OH

benzene

140 120 I

100 80 60 1 r

ppm

Fig. 2. 27AI NMR (52.1 MHz) spectrum of the products of reaction [(C4H9)4N]A1H4 + 2 C H 3 O H in benzene at ambient temperature.

542 St. Hermänek et al. • Alkoxohydridoaluminates

Table I: 27AI Chemical shifts of [(C4H9)4N]AlH4_„(OR)„, present at 22 °C (70 °C) in benzene solutions, prepared according to the Methods A and C. Standard: external A1(H 2 0) 6 3 + C1 3 /H 2 0.

R 1 2 3 4 5

CH3 101.6 121.1 111.5 90.6 72.6 100.9 120.9 111.5 90.6 72.0

C2H5 101.6 118.2 108.3 87.8 68.3 100.9 117.9 107.7 87.3 69.5

St. Hermanek et al. • Alkoxohydridoaluminates 543

adjacent substituents (H, OR) but also between the H (EN = 2.2) and OR (3.55) substituents them-selves. To what extent the latter interaction is in-volved in the AI NMR product is under investiga-tion.

With mixed haloaluminates AlX4_nY„~, it was found earlier [39] that a stepwise exchange of Y for X brings no regular changes in the 27Al chemical shifts. A good agreement ( < 1 ppm) was, however, achieved by calculating AlH3OR~ > AlH2(OR)2~ > AlH(OR)3~, and the stereospecific-ity will increase with growing steric requirements of the alkoxy group: OCH 3 < OC2H5 < 0 - / - C 3 H 7 < 0 - r - C 4 H 9 .

5) The calculations of 27Al chemical shifts of 2—4 by means of pairwise additivities and trigonal addi-tive parameters fail. A similar failure can be ex-pected in general with all tetra-substituted com-pounds having two or more substituents which differ mutually in electronegativity or in the ability to form jr-back-donation interactions with the central atom.

Experimental

The 27A1 NMR spectra (52.128 MHz) were meas-ured using a Varian XL-200 NMR spectrometer. The

544 St. Hermanek et al. • Alkoxohydridoaluminates 544

samples were sealed in 10 mm tubes under argon with a capillary containing the standard, an aqueous solu-tion of [Al(OH2)6]Cl3. The spectra of 0.2 M benzene solutions of 1—5 were measured without lock, the field stability varied in the range of ± 5 Hz. The quadrupole broadening was reduced partially by measuring spectra at elevated temperatures (70 or 80 °C).

All reactions and manipulations were carried out under dry argon. Alcohols and solvents were dried over molecular sieves and distilled over CaH2 prior to use. NaAlH4 for the preparation of (C4H9)4NA1H4 contained 97.15% of the theoretical amount of Al (Al: Na: H~ = 1:1.03:3.97).

Preparation of [(C4H9)4N]AIH4 was a slight modifi-cation of a method described by Bakum [44]: A solu-tion of 7.89 g NaAlH4 (0.142 mol) in 200 ml of THF was added during 10 min to a stirred solution of (C4H9)4NBr (45 g, 0.139 mol) in 700 ml of dry THF at 40 °C. The reaction mixture was allowed to cool to rooms temperature and stirred for 4 h. After 1 d storing the precipitated sodium chloride was filtered

off, THF was evaporated in vacuo and white crystals of (C4H9)4NA1H4 were dried for 3 h at 50 °C and 50 Pa. Yield >95%. Analysis

Calcd Al 9.87 H 1.48 N 5.12, Found Al 9.85 H 1.46 N 5.12.

Preparation of benzene solutions of the (C4H9)4N+ salts of 2—4

Method A: A calculated amount of an appropriate alcohol (ROH) was added dropwise to the 0.2 M benzene solution of [(C4H9)4N] A1H4 at room temper-ature, the mixture was heated to 80 °C and kept at this temperature for 1 h. The resulting solution was transferred to the 10 mm NMR tube, sealed under argon and stored at room temperature till NMR measuring (2—6 d).

Method B: Calculated amounts of 0.2 benzene solutions of [(C4H9)4N]A1H4 and [(C4H9)4N]Al(OR)4 were mixed at room temperature and the mixture was treated analogously as described in method A.

J. Mälek, Org. React. 34, 1 (1985). J. Mälek and M. t e r n y , Synthesis 1972, 217. H. C. Brown and C. J. Shoaf, J. Am. Chem. Soc. 86, 1079 (1964). N. Ya. Turova, M. I. Karpovskaya, and A. V. Novoselova, Zhur. Obshch. Khim. 47, 1338 (1977). Ref. [1], p. 7. H. Haubenstock, J. Org. Chem. 38, 1765 (1973). H. Haubenstock and E. L. Eliel, J. Am. Chem. Soc. 84, 2363 (1962). H. Haubenstock, J. Org. Chem. 38, 1765 (1973). E. C. Ashby, F. R. Dobbs, and H. P. Hopkins (Jr.), J. Am. Chem. Soc. 97, 3158 (1975). B. täsensky, J. Machäcek, and K. Abrham, Coll. Czech. Chem. Commun. 36, 2648 (1971). E. C. Ashby, J. P. Sevenair, and F. R. Dobbs, J. Org. Chem. 36, 197 (1971). E. C. Ashby and J. R. Boone, J. Org. Chem. 41, 2890 (1976). M. H. Abd ElKader, Chem. Ber. 103, 1225 (1970). J. Fusek, F. Hanousek, and F. Haruda, Coll. Czech. Chem. Commun. 41, 187 (1976). M. H. Abd ElKader, Tetrahedron Lett. 1969, 2301. P. Stilbs and S. Forsen, Org. Magn. Reson. 8, 382 (1976). K. E. Wiegers and S. G. Smith, J. Am. Chem. Soc. 99, 1480 (1977). K. E. Wiegers and S. G. Smith, J. Org. Chem. 43, 1126 (1978). R. J. McMahon, K. E. Wiegers, and S. G. Smith. J. Org. Chem. 46, 99 (1981). V. V. Gavrilenko, M. I. Vinnikova, V. A. An-tonovich, and L. I. Zakharkin, Izv. Akad. Nauk SSSR, Ser. Khim. 1979, 1273.

[21] V. V. Gavrilenko, M. I. Vinnikova, V. A. An-tonovich, and L. I. Zakharkin, Izv. Akad. Nauk, Ser. Khim. 1984, 865.

[22] An exception was the attempt of Home (see Ref. [24]) to discern LiAlH4, LiAlH2(0—f-Bu)2 and LiAlH(0-r -Bu) 3 in THF by means of a 27AI NMR kinetic approach.

[23] J. Huet, J. Durant, and Y. Infarnet, Org. Magn. Reson. 8, 382 (1976).

[24] D. A. Home, J. Am. Chem. Soc. 102, 6011 (1980). [25] S. Hermanek, O. Knz, J. Fusek, B. Cäsensky, and

Z. t e rny , Z. Naturforsch. 40b, 582 (1985). [26] N. Ahmad, F. K. Cartledge, M. C. Day, and J. H.

Medley, Organometallics 3, 389 (1984). [27] H. Nöth and B. Wrackmeyer, NMR Spectroscopy of

Boron Compounds, in P. Diehl, E. Fluck, and R. Kös-feld (eds): NMR Basic Principles and Progress 14, Springer Verlag, Berlin - Heidelberg - New York 1978.

[28] P. C. Lauterbur, in F. C. Nachod and W. D. Phillips (eds): Determination of Organic Structure by Physical Methods. Vol. 2, p. 513, Academic Press, New York 1970.

[29] E. A. Williams and J. D. Cargioli, in G. A. Webb (ed.): Ann. Reps on NMR Spectroscopy 9, 221, Academic Press, London, New York, Toronto, Syd-ney, San Francisco 1979.

[30] O. Knz, B. Cäsensky, A. Lycka, J. Fusek, and S. Her-manek, J. Magn. Reson. 60, 375 (1984).

[31] J. T. Fenwick and J. W. Wilson, Inorg. Nucl. Chem. Lett. 7, 341 (1971).

[32] N. Ya. Turova, M. I. Karpovskaya, A. V. Novoselo-va, G. A. Kirakosyan, V. P. Tarasov, and N. I. Kozlo-va, Koord. Khim. 3, 1299 (1977).

St. Hermanek et al. • Alkoxohydridoaluminates 545

[33] G. Engelhardt, R. Radeglia, H. Jancke, E. Lippmaa, and M. Mägi, Org. Magn. Reson. 5, 561 (1973).

[34] W. Büchner and W. Wolfsberger, Z. Naturforsch. 32 b, 967 (1977).

[35] J. H. Letcher and J. R. Van Wazer, J. Chem. Phys. 44, 815 (1966); 45, 2916, 2926 (1966).

[36] R. Radeglia and H. Teichmann, Z. Chem. 14, 249 (1974).

[37] R. Radeglia and G. Engelhardt, Z. Chem. 14, 319 (1974).

[38] B. Wrackmeyer, Z. Naturforsch. 37b, 1524 (1982).

[39] R. G. Kidd and D. R. Truax, J. Am. Chem. Soc. 90, 6867 (1968).

[40] E. R. Malinowski, J. Am. Chem. Soc. 91, 4701 (1969). [41] G. R. Somayajulu and B. J. Zwolinski, J. Magn.

Reson. 30, 51 (1978). [42] G. R. Somayajulu, J. R. Kennedy, T. M. Vickrey, and

B. J. Zwolinski, J. Magn. Reson. 33, 559 (1979). [43] G. E. Maciel, J. W. Mclver (Jr.), N. S. Ostlund, and

J. A. Pople, J. Am. Chem. Soc. 92, 1 (1970). [44] S. I. Bakum, Izv. Akad. Nauk SSSR, Ser. Khim. 1974,

2138.