Comprehensive Inorganic Chemistry EDITORIAL BOARD J. C. BAILAR JR., Urbana H. J. EMELÉUS, F.R.S., Cambridge tSIR RONALD NYHOLM, F.R.S., London A. F. TROTMAN-DICKENSON, Cardiff {Executive Editor) The Chemistry of OXYGEN E. A. V. Ebsworth, J. A. Connor and J. J. Turner Chapter 22 of Comprehensive Inorganic Chemistry PERGAMON PRESS OXFORD . NEW YORK . TORONTO SYDNEY . PARIS . BRAUNSCHWEIG Pergamon Press Offices: U.K. Pergamon Press Ltd., Headington Hill Hall, Oxford, OX3 OBW, England U.S.A. Pergamon Press Inc., Maxwell House, Fairview Park, Elmsford, New York 10523, U.S.A. CANADA Pergamon of Canada Ltd., 207 Queen's Quay West, Toronto 1, Canada AUSTRALIA Pergamon Press (Aust.) Pty. Ltd., 19a Boundary Street, Rushcutters Bay, N.S.W. 2011, Australia FRANCE Pergamon Press SARL, 24 rue des Ecoles, 75240 Paris, Cedex 05, France WEST GERMANY Pergamon Press GmbH, D-3300 Braunschweig, Postfach 2923, Burgplatz 1, West Germany Copyright © Pergamon Press 1973 All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means: electronic, electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without permission in writing from the publishers First edition 1973 Reprinted, with corrections, from Comprehensive Inorganic Chemistry, 1975 Library of Congress Catalog Card No. 77-189736 Printed in Great Britain by A, Wheaton & Co, Exeter ISBN 0 08 018858 3 (hard cover) ISBN 0 08 018857 5 (Flexicover) PREFACE The excellent reception that has been accorded to Compre- /tensive Inorganic Chemistry since the simultaneous publication of the five volumes of the complete work has been accompanied by the plea that sections should be made available in a form that would enable specialists to purchase copies for their own use. To meet this demand the publishers have decided to issue selected chapters and groups of chapters as separate editions. These chapters will, apart from the corrections of misprints and the addition of prefatory material and individual indices, appear just as they did in the main work. Extensive revision would delay publication and greatly raise the cost, so limiting the circulation of these definitive reviews. A. F. TROTMAN-DICKENSON Executive Editor vii 22. OXYGEN E. A. V. EBSWORTH University of Edinburgh J. A. CONNOR The University of Manchester and J. J. TURNER The University, Newcastle-upon-Tyne 1. OXYGEN 1.1. DISCOVERY OF OXYGEN* By the middle of the seventeenth century it was appreciated that air contained a compo­ nent associated with breathing and burning. In the first theory of burning to become widely accepted, this component was called phlogiston. When something burned, it was believed to release phlogiston to the air around it. If the burning took place in a sealed system, it stopped after a time because the air in the system became saturated with phlogiston. There were serious difficulties about this interpretation. For instance, metals such as tin gain in weight when they burn. However, the theory was widely accepted until the end of the eighteenth century. Both the experimentalists generally credited with the discovery of oxygen, Joseph Priestly and Carl Wilhelm Scheele, were believers in the phlogiston theory. Indeed, Scheele called his experiments "... proofs that Heat or Warmth consists of Phlogiston and Fire Air". Scheele obtained oxygen, which he called fire air, by heating nitrates, mercuric oxide, or manganese dioxide in retorts to the ends of which bladders had been fixed; Priestly heated mercuric oxide with a magnifying glass and collected the gas over water. Both found that the gas they had obtained would support combustion better than does common air; and after some experiments with mice Priestly ventured to breath some of it himself, with very pleasant results. Priestly isolated what he called "dephlogisticated air" on 1 August 1774; after further experiments he wrote about his results to Sir John Pringle, the President of the Royal Society, in March 1775, and his letter was read before the Society on 23 March, while a detailed account of his experiments was published in the same year in the second volume of his book Experiments and Observations on Different Kinds of Air. Scheele worked at about the same time, but he had difficulties with his publishers (as others have 1 M. E. Weeks, The Discovery of the Elements, 6th edn., published by the Journal of Chemical Education, New York (1956); D. McKie, Antoine Lavoisier, Gollancz, London (1935); J. G. Gillam, The Crucible, Robert Hale, London (1954); The Collected Papers of C. W. Scheele (translated by L. Dobbin), Bell, Edinburgh (1931). 685 686 OXYGEN: E. A. V. EBSWORTH, J. A. CONNOR AND J. J. TURNER had since then), and his book, Chemische Abhandlung von der Luft und dem Feuer, did not appear until 1777. Neither Priestly nor Scheele, however, seems fully to have understood the signifi­ cance of their discoveries. When Priestly was in Paris in October 1774, he mentioned some of his results to the distinguished French Academician, Antoine Lavoisier, who was interested in combustion. Lavoisier was also in touch with Scheele. He repeated and extended Priestly's experiments, and began to consider his results in the light of the deficiencies of the phlogiston theory. He burned tin in a sealed vessel, and showed that after combustion the weight of the vessel was effectively unchanged as long as the seal was not broken ; thus the increase in weight of the tin on combustion could not be derived from outside the vessel. If the tin had lost phlogiston, then the phlogiston must have negative mass. From these and other experiments he concluded that when the metal burned what really happened was that it combined with something in the air; that "something" was the gas discovered by Priestly and Scheele, which Lavoisier called "principe oxygine". His view was not absolutely correct. He thought he was talking of a principle of acidity, for "oxygine" comes from the Greek word oxus, meaning sharp, and hence acid. But it was Lavoisier's penetration of mind, coupled with the experiments of himself, of Priestly, and of Scheele, that led to the collapse of the phlogiston theory and to the development of modern chemistry. 1.2. GENERAL Oxygen is the eighth element in the Periodic Table. The electronic structure can be represented in terms of one-electron wave functions as ls22s22p4; the ground state and some of the lower excited state terms are given in Table 1, with their energies above the ground state, and some other important properties of oxygen are given in Table 2. Detailed calcula­ tions of the wave function for the oxygen atom have been made using SCF and other methods2. Oxygen forms compounds with all the elements of the Periodic Table except for the lightest rare gases. These compounds could in principle be formed if oxygen were to lose electrons (forming cations), to gain electrons (forming anions) or to share electrons (forming bonds). The first four ionization potentials and the first two electron affinities are given in Table 2. These values show that the loss of electrons from oxygen is a process requiring much energy. Compounds are known in which oxygen is formally cationic, such as O/PtF^, but in all known cases the cations are polyatomic, and in general the chemistry of oxygen is not cationic. The doubly charged anion O-2 is a common species, even though its formation from the gaseous atom involves substantial absorption of energy ; in ionic oxides the lattice energies are very high and more than compensate for the energy of formation of O-2. This must also explain why the singly charged 0~ is only known as an unstable species in irradiated solids and in the gas phase; in oxides it is unstable with respect to disproportionation to O2 and O-2. The coordination numbers of 0~2 ions in oxides are set out in Table 2, and are considered in more detail in section 3.3. Oxygen can form two additional bonds either by forming two σ-bonds with other atoms or groups, as in (ΟΗβ^Ο, or by forming a σ-bond and a π-bond with the same other atom 2 A. L. Merts and M. D. Torrey, /. Chem. Phys. 39 (1963) 694; C. C. J. Roothaan and P. S. Kelly, Phys. Rev. 131 (1963) 1177; E. Clementi, /. Chem. Phys. 40 (1964) 1944. GENERAL 687 TABLE 1. ATOMIC ENERGY LEVELS OF OXYGEN Electron Energy above configuration State / ground state (cm-1) "2/74 3p 2 0.0 1 158.5 * 0 226.5 2J22/>4 2/?4 ID 2 15 867.7 2/74 15 0 33 792.4 2522/?3(450)35 35 550 2 73 767.81 2522/73(450)35 35 350 1 76 794.69 2522/73(450)3^ 3/7 5p 1 86 625.35 2 86 627.37 3 86 631.04 2*22/73(450)3^ 3/7 3p 2 88 630.84 1 88 630.30 0 88 631.00 2522/>3(450)4y 45 550 2 95 476.43 2*22/73(450)45 45 350 1 96 225.5 2522/73(450)3ί/ 3d 5£>0 4 97 420.24 3,2 97 420.37 2,1,0 97 420.50 Ionization: 2522/73 2/73 45Ο 3/2 109 836.7 Data from C. E. Moore, Atomic Energy Levels, NBS Circular 467 (1949). TABLE 2. SOME PHYSICAL PROPERTIES OF THE OXYGEN ATOM Ionization potentialsa (eV) 1st, 13.614; 2nd, 35.146; 3rd, 54.934; 4th, 77.394 Electron affinities (eV) 1st, 1.478±0.002 b; 2nd, (O, -> 0,"2), -7.8±0.3 Atomic weight (O2 scale)d 15.9994 Atomic radiuse 0.73 A Ionic radius d 1.39 ±0.004 A Van der Waals radius r -1.50 A Electronegativitye 3.46 Coordination numbers at oxygen:β (i) In ionic or near-ionic compounds 2 (e.g. S1O2), 3 (e.g. rutile), 4 (e.g. ZnO), 6 (e.g. MgO), 8 (e.g. Na 0) 2 (ii) In molecular compounds 1 (e.g. CO), 2 (e.g. H 0), 3 (e.g. Me OBF ), 4 (e.g. Be OAc) 2 2 3 4 • C. E. Moore, Atomic Energy Levels, NBS Circular 467, 1949. b R. S. Berry, J. C. Mackie, R. L. Taylor and R. Lynch, /. Chem. Phys. 43 (1965) 3067. c From thermochemical cycles: M. F. C. Ladd and W. H. Lee, Acta Cryst. 13 (1960) 959. d A. E. Cameron and E. Wickers, /. Am. Chem. Soc. 84 (1962) 4175. e R. T. Sanderson, Chemical Periodicity, Reinhold (1960). ' A. Bondi, /. Phys. Chem. 68 (1964) 441. The value is about the same for -O- and for =0. • See section 3. C.I.C. VOL II—Y 688 OXYGEN: E. A. V. EBSWORTH, J. A. CONNOR AND J. J. TURNER or group, as in )C=0. Species in which oxygen forms just one σ-bond with another group (e.g. OH) have been detected spectroscopically, but they are free radicals and are not normally stable under chemical conditions. For once, all scales of electronegativity agree that oxygen is a very electronegative element. The electronegativity depends on the orbitals and the electron configuration; calculations have been made of the parameters associated with the valence state of oxygen, based on spectroscopic measurements3. As expected, the double bonds are shorter than the single bonds, and they have higher energies and stretching frequencies. Bonds of intermediate order are found in many compounds, including oxyanions such as RCO^, COj2, NOj, or SO42; relationships between bond length, bond order and stretching force constants have been described for BO 4, CO 5, NO <*, SiO 7, PO 7, SO 8 and CIO 7 bonds. Like fluorine, oxygen is a ligand which tends to promote oxidation of other elements to which it is bound (cf. OsVIII04). Bonds from 2-coordinated oxygen are usually considered as formed from (roughly) sp* hybrid orbitals, leading to bond angles at oxygen near the tetrahedral value. This leaves two lone pairs, which are also regarded as being in roughly sp*-orbita\s. Thus in each of the compounds H2O, Μβ2θ and F 0, the angles are near 109° (Table 3). However, 2 oxygen can also use its lone pairs to form either intermolecular σ-bonds or intramolecular π-bonds additional to the normal σ-bonds. Additional σ-bonds. Despite its high electronegativity, oxygen is a lone pair donor. Compounds like Μβ 0 form complexes with acceptors such as BF3, and values for the 2 energies of some donor-acceptor bonds involving oxygen are given in Table 4, with some values for other elements for comparison. Water is a well-known donor ligand in transition metal chemistry; oxygen is also a hydrogen bond acceptor, as in carboxylic acid dimers and in ice. All these interactions involve the lone pairs. In ice and in basic beryllium acetate, the angles at the 4-coordinated oxygen atoms are roughly tetrahedral, which also implies that the lone pairs are in approximately spi-orbitals; similarly, in HaO+, which is isoelec- tronic with NH3, the angle is near 109°. In compounds, where it forms two σ-bonds (e.g. Me20), oxygen is a hard (class A) base, though in compounds like Me2C=0 it has some "soft" character. Water comes near the "small Δ" end of the spectrochemical series; water and OH~ have small trans-eïïects, and in the nephelauxitic series water and OH~ come close to F~ as ligands with the smallest effect. Internal π-bonding. If an attached atom or group Q has empty orbitals of π-symmetry relative to the Q-0 σ-bond, these will overlap the lone pair orbitals at oxygen, and the overlap may lead to the formation of a donor π-bond. Thus in carbon monoxide the CO bond is of higher order than 2, and the strength of the bond can be attributed at least partly to an interaction of the form (0+ = C-). In compounds in which oxygen is forming two or more σ-bonds, donor π-bonding like this will affect bond angles at oxygen. The donor π-overlap will be greatest when the lone pairs are in pure /?-orbitals. If one lone pair is in a pure /^-orbital, the two σ-bonds and the other lone pair must be built from one s~ and two /7-orbitals. This would lead (if all three σ-orbitals are equivalent) to an angle of 120°. If both lone pairs are in pure /7-orbitals, the σ-bonds must be built from 5/7-orbitals, 3 G. Pilcher and H. A. Skinner, /. Inorg. Nucl. Chem. 24 (1962) 937. 4 J. Krogh-Moe, Acta Chem. Scand. 17 (1963) 843. 5 J. P. Fackler and D. Coucouvanis, Inorg. Chem. 7 (1968) 181. * Yu. Ya. Kharitonov, Izv. Akad. Nauk SSSR, Otdel Khim. Nauk 1962, 1953. 7 E. A. Robinson, Can. J. Chem. 41 (1963) 3021. * P. Haake, W. B. Miller and D. A. Tyssee, /. Am. Chem. Soc. 86 (1964) 3577. GENERAL TABLE 3. ANGLES AT OXYGEN How Angle measured Phase Reference In species QOZ 1. Neither Z nor Q π-acceptors: H 0 104.52° vib. vap. a 2 F 0 103.1 ±0.05° μ wave vap. b 2 (CH ) 0 111.5±1.5° ED vap. c 3 2 CH3OH 109° ±3° ED vap. c RbOH 180° μ wave vap. d 2. Q, π-acceptor; Z, not: S1H3OCH3 120.6 ±0.9° ED vap. e 3. Q and Z both π-acceptors : C1 0 110.8±1° ED vap. f 2 CI2O7 118.6±0.7° ED vap. g SiH OC H 121 ±1° ED vap. h 3 6 5 (SiH ) 0 144.1 ±0.9° ED vap. i 3 2 SiOSi in silicates 140-180° X-ray solid j (GeH ) 0 126.5 ±0.3° ED vap. k 3 2 [0 POP0 ]-4 133.5° X-ray solid 1 3 3 [0 SOS0 ]-2 124° X-ray solid m 3 3 [0CrOCr0]"2 115° X-ray solid n 3 3 [CI5MOMCI5]-4 180° X-ray solid 0 (M = Ru, Re) 4. Q 0+: 3 H30+ 112° X-ray solid P (ClHg)30+ 120° X-ray solid q a W. S. Benedict, N. Gailar and E. K. Plyler, /. Chem. Phys. 24 (1956) 1139. b Y. Morino and S. Saito, /. Mol. Spectrosc. 19 (1966) 435. c K. Kimura and M. Kubo, /. Chem. Phys. 30 (1959) 151. d C. Matsuma and D. R. Lide, /. Chem. Phys. 50 (1969) 71. e C. Glidewell, D. W. H. Rankin, A. G. Robiette and G. M. Sheldrick (to be published). f J. D. Dunitz and K. Hedberg, /. Am. Chem. Soc. 72 (1950) 3108. 8 B. Beagley, Trans. Faraday Soc. 61 (1965) 1821. h C. Glidewell, D. W. H. Rankin, A. G. Robiette, G. M. Sheldrick, B. Beagley and J. M. Freeman, Trans. Faraday Soc. 65 (1969) 2621. 1 A. Almenningen, O. Bastiansen, V. Ewing, K. Hedberg and M. Traetteberg, Acta Chem. Scand. 17 (1963) 2455. J D. W. J. Cruickshank, /. Chem. Soc. 1961, 5486; D. W. J. Cruickshank, H. Lynton and G. A. Barclay, Acta Cryst. 15 (1962) 493. k C. Glidewell, D. W. H. Rankin, A. G. Robiette, G. M. Sheldrick, B. Beagley and S. Cradock, /. Chem. Soc. (A), 1970, 315. 1 D. M. Macarthur and C. A. Beevers, Acta Cryst. 10 (1957) 428. m H. Lynton and M. R. Truter, /. Chem. Soc. 1960, 5112. n C. A. Brystrom and K. A. Wilhelmi, Acta Chem. Scand. 5 (1951) 1003. 0 A. M. Mathieson, D. P. Mellor and N. C. Stephenson, Acta Cryst. 5 (1952) 185; J. C. Morrow, Acta Cryst. 15 (1962) 851. p C. E. Nordman, Acta Cryst. 15 (1962) 18. q S. Séavniëar and D. Grdenié, Acta Cryst. 8 (1955) 275. 690 OXYGEN: E. A. V. EBSWORTH, J. A. CONNOR AND J. J. TURNER TABLE 4. GAS-PHASE DISSOCIATION ENTHALPIES FOR SOME MOLECULAR COMPLEXES (kcal mol-1) Complex AH Complex AH MeOBF 13.3 MeSGaMe ~8 2 3 2 3 Et2OBF3 10.9 Me2SBH3 5.2 THFBF3 13.4 Me NBF b 3 3 Me2OBMe3 a Me3NAlMe3 b MeOAlMe b Me NGaMe 21 2 3 3 3 MeOGaMe 9.5 MePBF 18.9 2 3 3 3 MeOBH a 2 3 Data from F. G. A. Stone, Chem. Rev. 58 (1958) 101. a Too unstable to study. b Too stable to determine. c For dissociation into Me S and B HÔ. 2 2 and the angle at oxygen will be 180°. These possibilities, which are represented below in valence-bond terms, are extremes : intermediate angles might be expected, deriving from a balance between π-bonding (widening the angle, and removing electrons from oxygen) and charge distribution (which will tend to keep electrons on the oxygen atom) : Internal π-bonding will lead to shorter and stronger bonds, and should weaken the donor properties of the oxygen atom; the shortness and strength of bonds between oxygen and borono, silicon™ or transition elements11 have been accounted for in terms of this type of interaction. Several compounds are known in which oxygen is bound to one or two π-acceptors and in which the angle at oxygen is unusually wide (see Table 3). A similar argument could be used to explain why angles in ChO+ might be nearer 120° than 109°; π-interactions between silicon and Q would be greatest if the lone pair at Q were in a pure /7-orbital. However, this argument should be used with some caution; the angle at oxygen in RbOH is 180°, yet there is unlikely to be significant π-bonding between rubidium and oxygen. Oxygen is known in a variety of formal oxidation states, from +2 to —2. Of these, the positive states are (by definition) only found when oxygen is bound to a more electro­ negative element—which must be fluorine—or forms part of a cation such as 0}. In other oxidation states greater than —2 the oxygen atom concerned must either be bound to at least one other oxygen atom or form part of a cation or free radical .A potential diagram for the redox chemistry of oxygen in aqueous solution is given in Table 5. » C. A. Coulson and T. W. Dingle, Acta Cryst. B24 (1968) 153. !0 D. W. J. Cruickshank, /. Chem. Soc. 1961, 5486. 11 F. A. Cotton and R. M. Wing, Inorg. Chem. 4 (1965) 867.