The molecules AlO2, Al(O2)2, and Al(O 2)3: Experimental and quantum-chemical investigations on the oxidation of aluminum atoms

Article abstract of DOI:10.1002/anie.200500512

(Chemical Equation Presented) On the route to alumina, oxygen-rich oxides such as Al(O₂)₃ (see picture) are detected spectroscopically during the stepwise oxidation of aluminum atoms and confirmed structurally by quantum-chemical cal

Full text of DOI:10.1002/anie.200500512

Angewandte
Chemie
Oxygen-Rich Compounds
The Molecules AlO2, Al(O2)2, and Al(O2)3:
Experimental and Quantum-Chemical
Investigations on the Oxidation of Aluminum
Atoms**
Gregor Stꢀßer and Hansgeorg Schnꢀckel*
Dedicated to Professor Klaus Krogmann
on the occasion of his 80th birthday
We have already mentioned the significance of peroxo- and
hyperoxo complexes with unusual spin states as intermediates
in oxidation processes in biology and synthetic chemistry
when we reported the oxidation of AlX molecules with O2
under matrix-isolation conditions.[1][2] The reactions of metal
atoms with oxygen discussed herein also prove to be complex,
since electrons are taken up stepwise by the O2 molecules
with retention of symmetry and spin rules (O2!O2ꢀ!O22ꢀ).
On the preparative scale, this process results exclusively
2ꢀ
ꢀ
cleavage of the O O bond and the formation of O ions;
with metal cations these ions form particularly stable oxides,
such as Al2O3. To investigate the primary steps of this reaction
we have treated aluminum atoms with an increasing excess of
O2 in the matrix-gas argon under matrix-isolation conditions
and have obtained the molecules AlO2 (1), Al(O2)2 (2), and
Al(O2)3 (3). These result are reported herein, while details of
the spectra and the results of quantum-chemical calculations
are given in the Supporting Information.
The formation of C2v symmetrical AlO2 (1) in solid argon
from aluminum atoms with approximately 0.1% O2 has been
reported elsewhere.[3]. In our experiments, and in agreement
with the literature, we have observed the symmetrical AlO2
vibration in the IR spectrum at 496 cmꢀ1 and measured its
16O/18O shifts.[4] A new band at 1070 cmꢀ1 is observed at
higher O2 concentrations in argon gas (1–3%) after reaction
of aluminum atoms, again in agreement with the literature.[5]
16O/18O substitution experiments indicate that this new band
belongs to a species with four equivalent oxygen atoms.[6] As
will be shown below, this species is Al(O2)2 (2), which is
assigned the structure (D2d symmetry) illustrated in Figure 1.
At still higher O2 concentrations in the matrix and finally in
pure solid O2 a band at 688 cmꢀ1 is observed in the IR
spectrum. This band shifts to 671 cmꢀ1 in 18O2 enriched O2 and
occurs at 679 cmꢀ1 in the presence of 16O/18O species.[7] Since,
[*] Dr. G. Stꢀßer, Prof. Dr. H. Schnꢀckel
Institut fꢁr Anorganische Chemie
Universitꢂt Karlsruhe (TH)
Engesserstrasse 15, Geb. 30.45, 76128 Karlsruhe (Germany)
Fax: (+49)721-608-4854
E-mail: hansgeorg.schnoeckel@chemie.uni-karlsruhe.de
[**] This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie. We thank Prof. Dr. Wim
Klopper for his help with the quantum-chemical calculations and
Priv.-Doz. Dr. Dr. Hans-Jꢀrg Himmel for valuable suggestions.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 4261 –4264
DOI: 10.1002/anie.200500512
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4261

Communications
For 2, the D2d isomer shown in Figure 1 is approximately
55 kJmolꢀ1 more stable than a molecule with D2h symmetry.
The other isomers (C4v and Td) also lie energetically much
higher (266 and 329 kJmolꢀ1, respectively). Although only
one band is observed for 2, the assignment seems to be correct
since five 16O/18O shifts were observed for this band which are
in very good agreement with those calculated, and indicates
that there are four equivalent oxygen atoms in this molecule.
Moreover, this interpretation is plausible since the bands
assigned to 2 appear after the formation of 1 and prior to the
formation of 3. The special bonding situation in 2 arises
because in the highest singly occupied molecular orbital
(SOMO; 1a2, Figure 3) one electron must be shared by both
Figure 1. Minimum structure of Al(O2)2 (2) calculated with B3LYP/
TZVPP (see Table 1).
under identical conditions, a new band appears at 1065 cmꢀ1
in the Raman spectrum,[8] we have assigned these two intense
bands as well as a weak signal in the Raman spectrum at
406 cmꢀ1 to the molecule Al(O2)3 (3), whose calculated
molecular structure (D3 symmetry)is illustrated in Figure 2.
Figure 3. Graphical representation of the SOMO (1a2) of Al(O2)2 (2).
O2 groups, that is, compared to neutral O2, each O2 group is
weakened because it carries an additional 1.5 electrons.
Hence 2 may not be described as O2 Al3+O2 (that is, as
ꢀ
2ꢀ
hyperoxide with an additional peroxo group). Thus it appears
ꢀ
ꢀ
plausible that the O O bond in 2 is weaker and the Al O
bond stronger than in 1.
Figure 2. Minimum structure von Al(O2)3 (3) calculated with B3LYP/
TZVPP (see Table 1).
Unlike the unusual bonding relationships in 2, those in 3
appear analogous to the bonding relationships in the recently
described XAl(O2)2 species (X = F, Cl, Br),[1] which means
that in 3 there are three O2ꢀ groups attached to the Al3+ ion by
interactions that are predominantly ionic in character. This
situation becomes especially clear through the absence of
vibrational coupling between the three O2 vibrations, since
almost identical frequencies of 1158 and 1154 cmꢀ1 are
calculated for the E and A1 vibrations. In this respect, as
well as with the O O and Al O distances and also the force
constants, the expected parallels to the XAl(O2)2 species
emerge. The unusual spin situation in the XAl(O2)2 com-
pounds (triplet electronic ground states) becomes even more
unusual in 3 for which a quartet ground state is found. To our
knowledge this is unique for molecular compounds free of
transition metals.[11] Admittedly, the energy difference
between the doublet and quartet states is very small. This
finding and the absence of stabilization in 3 from a Jahn–
Teller distortion also provide support for an essentially ionic
bonding mode. In spite of the unusual bonding situations in 1,
2, and 3 a consistent picture emerges for the three com-
pounds:
To confirm the assignment of the cited vibrational
frequencies to molecules 1, 2, and 3 extensive quantum-
ꢀ
chemical calculations were carried out; these gave the Al O
ꢀ
and O O distances and force constants listed in Table 1.
Table 1: Distances r [pm] and force constants f [Nmꢀ1] of AlO2 (1),
Al(O2)2 (2), and Al(O2)3 (3) calculated with B3LYP/TZVPP.
ꢀ
ꢀ
rAlꢀ
O
rOꢀ
O
fAlꢀ
O2
fOꢀ
O
1
2
3
195
175
185
134
148
136
210
386
353
648
473
632
The special bond properties in 1, its stability, and the
kinetics of the conversion of its different isomers have been
reported in detail recently.[9] We have been able to confirm
these results according to which the observed C2v hyperoxide
1 has a stability similar to that of the linear D1h isomer.
Presumably the energy barrier between the two species is very
high so that in our spectra only the somewhat less stable
(according to calculation[9]) C2v isomer is observed.[10]
The following distances and force constants were calcu-
lated for the isolated ions O2 and O22ꢀ: O2 135 pm,
ꢀ
ꢀ
2ꢀ
660 Nmꢀ1 and O2 163 pm, 190 Nmꢀ1 [1]
.
In comparison the
force constants determined from the observed frequencies for
4262
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 4261 –4264

Angewandte
Chemie
excited with a argon ion laser (Coherent), the selected wave length
was 458 nm.
1, 2, and 3 and the calculated distances confirm that in 1 and 3
(as with the XAl(O2)2 species) the hyperoxide ions are
essentially ionically bonded to the positively charged Al
centers. In 2, in agreement with the electronic situation, the
O2 bond is clearly elongated, similar to the bond an O22ꢀ ion.
The quantum-chemical DFT calculations (functional B3-LYP)[16]
were carried out with the Turbomole program package.[17] The
ab initio calculations were carried out at the CASSCF level with
Dalton,[18] Gaussian98[19] was used for calculation of the CCSD(T)
energies. TZVPP basis sets were used for all calculations.[20]
ꢀ
On the other hand, the Al O bond also clearly lengthens in
the order 2 < 3 < 1, which correlates very well with the
calculated Al-O2 force constants (Table 1). These findings are
illustrated in Figure 3 for the special bonding situation of
compound 2. For compound 3 and 1 this bond lengthening
may be attributed to the reduced formal charge at the Al
center (Al1+ in 1 and Al3+ in 3).
Received: February 10, 2005
Published online: June 1, 2005
Keywords: ab initio calculations · aluminum oxides ·
.
matrix isolation · oxidation · Raman spectroscopy
The energetic ordering of the three hypervalent com-
pounds is shown in Figure 4 and emerges from experimental
[1] a) J. Bahlo, H.-J. Himmel, H. Schnꢀckel, Angew. Chem. 2001,
113, 4820; Angew. Chem. Int. Ed. 2001, 40, 4696; b) J. Bahlo, H.-
J. Himmel, H. Schnꢀckel, Inorg. Chem. 2002, 41, 2678; J. Bahlo,
H.-J. Himmel, H. Schnꢀckel, Inorg. Chem. 2002, 41, 4488.
[2] a) M. Moskovits, G. A. Ozin, Cryochemistry, Wiley, New York,
1976; b) B. Meyer, Low Temperature Spectroscopy, American
Elsevier, New York, 1971; c) H. E. Hallam, Vibrational Spec-
troscopy of Trapped Species, Wiley, London, 1973.
Figure 4. Energy scheme [kJmolꢀ1] for the stepwise exothermic
oxidation of Al atoms with O2.
[3] a) L. V. Serebrennikov, S. B. Osin, A. A. Malꢁtsev, J. Mol. Struct.
1982, 81, 25; b) S. M. Sonchik, L. Andrews, K. D. Carlson, J.
Phys. Chem. 1983, 87, 2004; c) I. L. Rozhanskii, L. V. Serebren-
nikov, A. F. Shevelꢁkov, Zh. Fiz. Khim. 1990, 64, 521; d) L.
Andrews, T. R. Burkholder, J. T. Yustein, J. Phys. Chem. 1992,
96, 10182.
[4] The following frequencies were detected in the IR spectrum for
the isotopologues of AlO2 (cmꢀ1): Al16O2 496.4, Al16O18O 488.6,
Al18O2 480.1.
[5] L. V. Serebrennikov, A. A. Malꢁtsev, Vestn. Mosk. Univ. Ser. 2
Khim. 1985, 26(2), 137.
[6] The following frequencies were detected in the IR spectrum for
the isotopologues of Al(O2)2 (cmꢀ1): Al16O4 1070.2, Al16O318O
1062.4, Al16O218O2 1054.1, Al(16O18O)2 1055.6, Al16O18O3 1043.9,
Al18O4 1035.4.
[7] This experimental observation is in agreement with the results of
earlier work, although in the earlier work no conclusive
explanation could be made: L. Andrews, T. R. Burkholder,
J. T. Yustein, J. Phys. Chem. 1992, 96, 10182.
[8] The following frequencies of the different isotopologues were
observed in Raman (a) and IR spectra (b) (cmꢀ1): a) Al(16O2)3
1065, Al(16O/18O)3 1035, Al(18O2)3 1005; b) Al(16O2)3 686.4,
Al(16O2)2(18O2) 681.3, Al(16O2)(18O2)2 675.1, Al(18O2)3 670.7; in
experiments with a 1:2:1 mixture of 16O2/16O18O/18O2 broad
bands appear at 685, 679, and 672 cmꢀ1 which arise from the
superimposition of ten different isotopologues.
thermodynamic data[12] and the quantum-mechanical results
obtained in this work. Relative to 1 mol aluminum atoms, in
the stepwise formation of 1, 2, and 3 more energy is released
than for any other molecular AlxOy compound.[13] Never-
theless, the Al(O2)n (n = 1, 2, 3) molecules described herein
are certainly only obtainable in measurable quantities under
matrix conditions, since the formation of alumina from
Al(O2)3 is favored by 587 kJmolꢀ1
The results on the oxidation of aluminum atoms with
excess O2 reported herein demonstrate the complexity of such
apparently trivial reactions. On the other hand, this process is
an extreme situation, in the reverse case, that is, the reaction
of excess aluminum atoms with few O2 molecules is a more
realistic model when it comes to the investigation of the
primary products of oxidation, for example, on aluminum
surfaces. This problem is at the center of current gas-phase
investigations of Aln clusters and their oxidation by O2,
reactions which are being studied with the aid of FT mass
spectroscopy.[14]
Experimental section
[9] M. V. Pak, M. S. Gordon, J. Phys. Chem. 2003, 118(10), 4471.
[10] In contrast, with the homologous species GaO2, the photolyti-
cally induced (254 nm) isomerization of the C2v structure to the
D1h structure with an activation barrier of about 300 kJmolꢀ1 is
readily observed: A. Kꢀhn, B. Gaertner, H. Himmel, Chem. Eur.
J. 2005, in press.
[11] E. C. Brown, W. T. Borden, J. Phys. Chem. A 2002, 106, 2963.
[12] “NIST-JANAF Thermochemical Tables, Fourth Edition”:
a) M. W. Chase, Jr., J. Phys. Chem. Ref. Data Monogr. 1989, 9;
b) National Institute of Standards Web Based Chemical Data,
http://webbook.nist.gov/chemistry/.
[13] G. Stꢀßer, Dissertation, Universitꢂt Karlsruhe, 2004.
[14] R. Burgert, H. Schnꢀckel, unpublished results.
[15] H. Schnꢀckel, S. Schunck, Chem. Unserer Zeit 1987, 21, 73.
[16] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[17] a) R. Ahlrichs, M. Bꢂr, M. Hꢂser, H. Horn, C. Kꢀlmel, Chem.
Phys. Lett. 1989, 162, 165; b) http://www.turbomole.com.
A detailed description of the experimental construction used may be
found in ref. [15]. Aluminum was evaporated in high vacuum
(10ꢀ7 mbar) at 1400 K from a resistance-heated boron nitride cell
and cocondensed with the matrix gas (Ar/O2) for 2 h onto a copper
block cooled to 10 K with a “closed cycle” cryostat (Leybold LB510).
Each hour approximately 150 mg Al and 50 mmol matrix gas were
evaporated. Isotope substitution experiments were carried out with
18O2, a 1:1 mixture of 16O2/18O2, and a 1:2:1 mixture of 16O2/16O18O/
18O2. The IR and Raman spectra were recorded immediately after
sample deposition. The following chemicals were used in the matrix
reactions: Al (Merck, 99.9999%), Ar (Messer 99.999%), 16O2
(Messer 99.9998%), 18O2 (Linde 99.9998%, isotopic purity 99%).
The IR spectra were taken with Bruker (Karlsruhe) FTIR
spectrometers, types 113v and 66v. DTGS and MCT detectors were
used, the resolution was 1 cmꢀ1. Raman spectra were taken with a
Dilor grating spectrometer, type 800. The Raman vibrations were
Angew. Chem. Int. Ed. 2005, 44, 4261 –4264
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4263

Communications
[18] a) Dalton, a molecular electronic structure program, Release 1.2
M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.
Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J.
Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, P.
Salvador, J. J. Dannenberg, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G.
Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres,
C. Gonzalez, M. Head-Gordon, E. S. Replogle, and J. A. Pople,
Gaussian, Inc., Pittsburgh, PA, 2001.
(2001), T. Helgaker, H. J. Aa. Jensen, P. Joergensen, J. Olsen, K.
Ruud, H. Aagren, A. A. Auer, K. L. Bak, V. Bakken, O.
Christiansen, S. Coriani, P. Dahle, E. K. Dalskov, T. Enevoldsen,
B. Fernandez, C. Haettig, K. Hald, A. Halkier, H. Heiberg, H.
Hettema, D. Jonsson, S. Kirpekar, W. Klopper, R. Kobayashi, H.
Koch, K. V. Mikkelsen, P. Norman, M. J. Packer, T. B. Pedersen,
T. A. Ruden, P. Salek, A. Sanchez, T. Saue, S. P. A. Sauer, B.
Schimmelpfennig, K. O. Sylvester-Hvid, P. R. Taylor, O. Vahtras.
[19] Gaussian98 (RevisionA.11), M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C.
Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin,
[20] A. Schꢂfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100,
5829.
4264
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 4261 –4264