摘要:

FARADAY SYMPOSIA OF THE ROYAL SOCIETY OF CHEMISTRY NO. 14 1980 Diatomic Metals and Metallic Clusters THE FARADAY DIVISION THE ROYAL SOCIETY OF CHEMISTRY LONDON A SYMPOSIUM ON Diatomic Metals and Metallic Clusters 8th and 9th January 1980 A SYMPOSIUM on Diatomic Metals and Metallic Clusters was held at Owens Park University of Manchester on 8th and 9th January 1980. The President of the Faraday Division Professor J. S. Rowlinson F.R.S. was in the chair; about 70 Fellows of the Faraday Division and visitors from overseas attended the meeting. Among the overseas visitors were Dr. E. J. Baerends The Netherlands Prof. H. Basch Israel Dr. M. Benard France Dr. K. J. Cavell The Netherlands Prof. M. H. Chisholm U.S.A.Prof. F. A. Cotton U.S.A. Dr. 0. Echt West Germany Prof. K. A. Gingerich U.S.A. Prof G. Granozzi Italy Dr. K. Hilpert West Germany Dr. P. H. Kasai U.S.A. Mr. R. A. Meuli Switzerland Dr. A. R. Miedema The Netherlands Prof. P. Montano U.S.A. Prof. A. Oskam The Netherlands Prof. G. Ozin Canada Dr. D. Post The Netherlands Dr. L. Scheire Belgium Dr. I. Shim Denmark Dr. J. G. Snijders Tlle Netherlands Dr. G. G. Stanley France Dr. D. J. Stufkens The Netherlands Dr. H. Van Dam The Netherlands Mr. J. Van Der Velden The Netherlands Prof. A. Veillard France Prof. K. G. Weil West Germany Prof. J. S. Winn U.S.A. Organising Committee Prof. H. A. Skinner (Chairman) Dr. R. F. Barrow Dr. E. R. Buckle Dr. J.A. Connor Mrs. Y. A. Fish Dr. R. Grinter Dr. I. M. Hillier Dr. J. S. Ogden Dr. P. B. Wells Dr. D. A. Young ISBN 0-85186-998-X ISSN 0301-5696 0The Royal Society of Chemistry and Contributors 1980 Printed in Great Britain by Fletcher & Son Ltd. Norwich CONTENTS Page 7 Spectroscopy Chemistry and Catalysis of Metal Atoms Metal Dimers and Metal Clusters by G. A. Ozin 65 Electron Spin Resonance Study of Intermetallic Molecules CUM and AuM (M = Mg Zn Cd and Hg) by P. H. Kasai and D. McLeod Jr 79 Matrix Isolation Studies of Bimetallic Molecules of Fe-M(3d) Metals by P. A. Montano 87 Optical and Vibrational Studies of Siloer Molecules and Microcrystallites Prepared by Matrix Aggregation and Gas Aggregation Techniques by W.Schulze and H. Abe 94 Application of Magnetic Circular Dichroism Spectroscopy to the Ident$ca- tion of Small Matrix-isolated Metal Clusters and the Assignment of their Electronic Spectra by R. Grinter S. Armstrong U. A. Jayasooriya J. Mc- Combie D. Norris and J. P. Springall. 102 Gas Phase Production and Chemistry of Transition Metal Atoms and Clusters from Polynuclear Metal Carbonyls by J. S. Winn 109 Experimental and Predicted Stability of Diatomic Metals and Metallic Clusters by K. A. Gingerich 126 Models for Calculation of Dissociation Energies of Homonuclear Diatomic Molecules by L. Brewer and J. S. Winn 136 Model Predictions of the Dissociation Energies of Homonuclear and Hetero- nuclear Diatomic Molecules of Two Transition Metals by A.R. Miedema 149 Electronic Structure of Heavy Metal Diatomics from ab Initio Relaticistic EfSectire Core Potential Studies by H. Basch 159 Theoretical Study of the Electronic Structure of the Transition Metal Dimers Se, Cr, Mo andNi by C. Wood M. Doran I. M. Hillier and M. F. Guest 170 Structure and Electronic Properties of Copper Clusters by C. Bachmann J. Demuynck and A. Veillard 180 Dependence of Stability Bond Strength and Electronic Structure of Dimetal Units upon Atomic Number Oxidation Number and Chemical Encironment by B. E. Bursten and F. A. Cotton 194 Dynamic and Static Stereochemistry in Dimolybdenum and Ditungsten Com- pounds containing a Central (M-M)6+ Unit by M. H. Chisholm 21 1 Electronic Structure of Binuclear Metal Carbonyl Complexes by W. Heijser E. J. Baerends and P. Ros 235 GENERAL DISCUSSION 251 AUTHORINDEX

ISSN:0301-5696    

DOI:10.1039/FS9801400001

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Spectroscopy Chemistry and Catalysis of Metal Atoms Metal Dimers and Metal Clusters BY GEOFFREY A. OZIN Lash Miller Chemical Laboratories and Erindale College University of Toronto Toronto Ontario Canada Received 8th January 1980 This paper starts with the experimental realization of trapped silver atoms in various matrix supports with emphzsis placed on optical absorption emission and e.s.r. spectroscopic properties. Special attention is devoted to ground- and excited-state silver atom interactions with the surrounding matrix cage of atoms particularly with respect to the recently discovered phenomenon of light-induced diffusion and aggregation of silver atoms to small precisely defined silver clusters. Various aspects and applications of photodiffusion methods are highlighted and some pertinent comparisons are made with metal concentration and bulk annealing approaches as alternatives to controlled metal nucleation.This will lead to the embryonic clusters Ag and Ag for which there now exists consider- able spectroscopic photochemical and theoretical information. Silver-containing bimetallic clusters generated either by metal concentration deposition or photoaggregation methods will be a natural extension to the Ag,, presentation; their relevance to bimetallic cluster catalysts will be briefly contemplated. The concept of generating anionic silver clusters by Na/Ag photoionization methods will be briefly described with reference to the parent Ag- anion. In considering the higher silver clusters one is now in a position to evaluate experimentally the genesis of silver nucleation from the stage of an isolated silver atom through to a six-atom array and to higher less well defined aggregates.With these data one can attempt to track the evolving optical and e.s.r. properties by reference to the build-up of cluster electronic states calculated by way of SCF-Xor-SW molecular- orbital procedures for assumed cluster geometries. The observation of rudimentary interband transitions of silver clusters in the range 6-13 atoms that absorb in a similar energy region to the collective electronic excitations associated with plasmon absorption of silver microcrystallites simul- taneously with the evolution of conduction electron spin resonance absorption whose observed line- widths and g-shifts conform to the theoretical predictions of Kubo and Kawabata can in principle provide a valuable criterion on which to judge the atomic composition at which optical-electronic characteristics of silver aggregates transform from those of the molecular to the bulk state.Future directions in diatomic-metal and metal-cluster chemistry are briefly contemplated in the light of recent break-throughs with ambient-temperature metal-vapour-liquid-polymer techniques and the discovery of polymer-supported very low nuclearity metal clusters generated and stable at or very close to room temperature. INTRODUCTION It is a great honour for me to be chosen to deliver the introductory lecture at the Faraday Division’s Symposium on Diatomic Metals and Metallic Clusters a field which interested Michael Faraday himself who was one of the first scientists to record the making of gold clusters electrostatically and of course much work has been done since.As the topic of the Symposium is very close to the heart of much of our work in the field of metal vapour chemistry over the past seven years or so I have elected to present a paper on some of our more recent research that focuses attention on the spectroscopic chemical and catalytic aspects of silver atoms silver dimers and silver clusters. Because of restricted time coverage of details will necessarily be somewhat limited although these are generally available in the cited references. To open the lecture let us begin with the concept of viewing a finite organo- SPECTROSCOPY CHEMISTRY AND CATALYSIS metallic cluster complex as a fragment of a metal surface interacting with a group of ligands a notion which is beginning to yield incisive details at the atomic level for chemisorption of the same ligand on metal surfaces.l In particular M~etterties,~-~ Ugo and others have presented an experimental perspective on the cluster-surface analogy and have attempted to establish boundary conditions for employing small clusters as chemisorption models ; some surface scientists6-lo have begun to explore the applicability of the cluster-surface analogy ; whilst Messmer,'l Johnson and co- workers,12 Goddard and Schaefferl4 and other cluster-surface quantum theorists promulgate the viewpoint that the field of metal cluster compounds is likely to have considerable impact on our understanding of chemisorption and catalysis by metal surfaces.One should however realise that the theoreticians' localised approach to the cluster-surface interface is invariably by way of " low-coverage '' computer models of the form M,L rather than the " high-coverage " models typified by M,L,,. From a purely experimental standpoint one would ideally like to focus attention on " superco-ordinatively unsaturated " cluster models of the M,L type (together with the M core itself) as they are likely to represent a more realistic picture of surface-ligand inter-actions (and bare metal surfaces); in particular they relate more closely to the theoreticians' visualisation of the chemisorption bond.Unfortunately M,L cluster compounds by their very nature are extremely unstable with respect to a variety of reactions (for example further reaction with ligand more pronounced agglomeration etc.) and cannot generally be synthesised and studied by room-temperature methods. A similar statement often applies to the naked metal cluster cores M,. Experimental access to this important class of cluster compounds can however be achieved through the recently developed tech- nique of cocondensing metal atoms with reactive ligand molecules or weakly inter- acting matrices under cryogenic condition^.'^ The success of the method for generat- ing and spectroscopically probing low-coverage chemisorption models can be seen from the work of Ozin Moskovits and coworkers with M,(C0),16u Mn(C2H4),17 M,(O,) l8 cluster complexes which have allowed a limited 166 yet intriguing evaluation of the relationship between the electronic and vibrational properties of finite " low-coverage " M,L models and M(Lchemisorbed) as a function of cluster size.Although much work still remains to be done in this area these early studies15-18 indicate the usefulness of a localised bonding viewpoint of the chemisorbed state for clusters com- prising as few as two to four metal atom centres. Another rather important issue at this point concerns the generation spectro- scopic detection chemistry and catalytic properties of the naked metal clusters them- selves and the appropriateness of employing such very small metal atom arrays as models for clean metal For interfacial discussions of this type one is confronted with the problem of choosing a chemistry or solid-state physics ap- proach to questions concerning the number of metal atoms required by a cluster to describe bulk properties such as cohesive energies work functions densities of states band widths and separations to name a Enquiries of this type present stringent demands on both experimental techniques and theoretical methods.For certain properties of bulk solids or surfaces it seems that an adequate physical model consists of quantum mechanically treating a finite group of metal atoms artificially removed from the infinite solid. The intention of such calculations is usually to probe the similarities and differences between clusters of a given size and the bulk metal for a particular property of the metal.In order to perform meaningful cluster calculations certain minimal requirements are demanded of the theoretical method. The mathematical method must at least be reliable for both the finite- G. A. OZIN cluster and infinite-surface extremes. Also the method has to be computationally capable of handling sufficiently large cluster calculations so that cluster-to-bulk convergence properties can be eva1~ated.l'~'~ The SCF-Xu-SW method appears to well satisfy these minimum criteria and has been successfully applied to finite clusters as well as to bulk solids remembering that SCF-Xa-SW theory is the discrete analogue of the KKR band theory of infinite solids.21 Clearly the property of the bulk metal under scrutiny will govern to a large extent the type and magnitude of the cluster calculation.However within the framework of currently popular cluster-surface analogues and the more localized aspects of metals such as chemisorptive and catalytic properties of very small groups of metal atoms (<10 A) it is likely that electronic and geometric properties play a prominent role and are therefore deemed more pertinent to the present discussion. Certainly the more dramatic excursions in these properties are likely to be exhibited in the cluster regime of just a few atoms or so. With these considerations in mind one can begin to appreciate the urgency for developing reliable experimental and theoretical methods for generating and studying very small unimetallic and bimetallic clusters of known composition.In this paper the present status of some of these topics will be surveyed. SILVER ATOM CRYOPHOTOAGGREGATION Let us begin by considering the recently discovered phenomenon of light-induced matrix diffusionlaggregation of metal atoms to precisely defined nuclearity metal clus- ter~,~~-~~*~~ illustrated for Ag atoms in Ar in fig. l(a). Preliminary studies of the kinetics of the clustering process have indicated that under certain conditions the rates of formation of diatomic and triatomic silver may usefully be approximated by simple second-order kinetics. A simple analysis23 predicts that the slope of a plot of log (Ag,/Ag) against log t where Ag and Ag represent the absorbances and t represents the irradiation time should have a value of 1.0 for n = 2 and 2.0 for n = 3.These plots are shown in fig. l(b). The observed slopes 0.9/1.0 and 2.1/2.2 support the Ag and Ag assignments which are indicated in fig. l(a). These nuclear- ity determinations correlate exactly with assignments based on Ag atom concentration experiment~.~~'~~ The photoaggregation effect is not limited to Ar matrices as can be seen from fig. 2(a) which shows controlled nucleation up to the trisilver stage in Ar Kr and Xe.25 Correlations between the optical excitations of Ag Ag and Ag in Ar Kr and Xe can be fairly easily pinpointed from such studies. Incidentally simple mass-balance considerations 23 in an experiment of the kind depicted in fig.l(a) lead to the following expression which relates the decrease in an atomic absorption to increases in diatomic and triatomic absorptions in terms of the appropriate extinction coefficients -(A; -A:) = 2~1/~2(AiA;) + 3&1/&3(A;-A;). (1) It is prearranged in these experiments that Ag and higher clusters are not produced in significant quantities. A; represents the absorbance due to Ag at time t'. A; is the absorbance due to Ag at time t" and E represents the molar extinction coefficients for Ag,. For very dilute conditions and short irradiation times one can arrange that only negligible quantities of Ag are formed so that eqn (I) may be solved directly for E~/E,. For longer irradiation times the complete eqn (I) may be used to obtain a value for E~/E~.We note here that a similar procedure using multiple depositions at different concentrations may also be used.26 However the advantage of the 2.4 e X I m a -= 1.6 m -f -0 0 -I 0.9,l .o d 0.8 I I I 1.5 1.7 1.9 log t 250 300 3 50 450 wavelength Inm FIG.1.-(a) U.v.-visible spectra of Ag,,,,,/Ar mixtures (Ag/Ar E 1/103)at 12 K.Note the growth of Ag and Ag clusters and loss of Ag atoms as a result of 305-nm Ag atom excitation. Spectra A B and C represent irradiation times of 0,l and 4 min re~pectively.'~ (6) Kinetic plots showing a linear dependence on irradiation time (305 nm) of the absorbance ratios A&63nm/Ag300nm and ~~390nm/~~300nm (0)and a linear dependence on the square of irradiation time of the absorbance and Agi40nm/Ag300nm ratios Ag$45nm/Ag300nm (O),as predicted from the simple kinetic analysis.The quantities X were chosen in order to shift the Ag2/Ag against t and Ag3/Ag against t2plots through the origin.23 I I 1 I I I I I I I 100 250 3 00 350 400 450 wavelengthlnm wavelenqthlnm -FIG.2.4) U.v.-visible spectra of Agl,2,3/Ar Kr and Xe mixtures generated by condensing silver atoms with the rare gases (Aglmatrix E 1/103) at 12 K and irradiating the deposits at the frequencies of silver atomic resonance absorption lines.25 (b) Optical absorption spectra of well isolated Aglinert gas z 1/105mixtures at 10-12 K.” SPECTROSCOPY CHEMISTRY AND CATALYSIS photoaggregation method is that only one deposition is required for each EJE~or EJE~determination so that the method is much more convenient and considerably more accurate since mass balance of total metal is maintained after each irradiation thus eliminating the need for multiple quantitative deposition^.^^ TABLE 1.-RELATIVEEXTINCTION COEFFICIENTS' FOR Ag AND Ag peak height peak area E~~~~/E~~~~(A~) 0.40& 0.05 0.8 j=0.2 &1315/&3245( Ar) 1.2 & 0.5 0.60 5 0.30 &1323/&2270(Kr) 0.43 & 0.05 a The corresponding wavelengths (nm) in Ar and Kr matrices are indicated as superscripts.The uncertainty limits represent estimated upper and lower bounds. The extinction coefficient results are summarised in table 1 where c1/e2 and EJE values for Ag/Ar matrices are given for both peak-height and peak-area measurements.The value of EJE~was determined by both the photoaggregation and deposition procedures and the results were in satisfactory agreement. The results in table 1 show that cl/e2 is essentially invariant within experimental error to the change from Ar to Kr matrices an important observation in view of the silver cluster photochem- istry described later. Since these early discoveries a variety of new experiments has yielded pertinent information about the selectivity of the photoaggregation phenomenon for atomic species,27a selective photoaggregation in bimetallic systems,22*28 the feasibility of naked cluster cryophotochemistry 25929 and the ability to photo-manipulate cluster distribu- t ions .30 These experiments bring to the forefront the need for obtaining photophysical detail associated with energy transfer from an excited metal atomic guest to the host environment.As photodiffusion photoaggregation effects are expected to be highly sensitive to the selection of metal atom and associated excited state choice and tem- perature of the host material duration and intensity of the excitation metal atom dispersion and pretreatment of the matrix one clearly needs to develop a fairly thorough understanding of the photophysics behind the effect in order to establish experimental guidelines for the generality of the technique as a route to minicluster systems. With the prior knowledge that not all metal atomic systems enjoy the ability to photoaggregate in a facile well-behaved manner,31 one might expect that a study of the excited state properties of metal-atom-matrix interactions by means of absorp- tion and emission spectroscopy coupled with a suitable form of theoretical analysis would provide a monitor of cage effects and energy dissipation channels for excited metal atoms in equilibrated lattice In the particular case of 2S+ 2P excited silver atoms embedded in noble gas solids (where all the interesting level splittings occur in the 2P excited state) one does not simply observe resonance fluorescence with minimal Stokes shifts indicating weak electronic coupling of the valence silver electrons to the lattice phonons.Instead one observes 27ap32933 drama-tically red-shifted emissions from the 2P -+2Sresonance absorption values an in- triguing discovery when one realises that no other atomic state of silver lies to lower energy.34 The well-structured form of these emissions alerts one to the existence of a sizeable " silver-atom-matrix cage " excited state interaction (some resemblance to gas-phase alkali-atom-inert-gas exciplexes should be noted at this stage).35 Whereas the 2S-+ 2P (4d105s1-+4dl05p1)optical transition of free silver atoms appears as a spin-orbit doublet27b(2Sl,z+ 2P1/2,2P3/2),the absorption spectra of G. A. OZIN 13 well-isolated silver atoms in Ar Kr and Xe23*27u [fig. 2(b)]generally consist of three major (thermally stable blue site) and one or two weaker (thermally unstable red site) maxima the blue/red site effect becoming less pronounced on moving to the heavier matrix materials where site overlap problems are more serious.Note that these slightly different matrix environments can be easily seen in the corresponding HIG HIG 2800 3 300 3800 2 0 3300 31 I0 II.I'II* F FIG.3.-Matrix e.s.r. spectra of (A) a freshly deposited Ag/Ar z 1/103 mixture onto an optically flat sapphire rod at 10-12 K; (B)-(H) after various narrow-band irradiations and thermal annealings at the following wavelengths and temperatures B 315 nm; C,385 nm; D 315 nm; E 442 nm; F 35 K; G 40 K and H 42 K. Hyperfine lines of lo7Ag/lo9Ag are indicated; various suspected secondary silver-atom trapping sites are labelled 6,c d e; hyperfine and silver isotope lines tentatively ascribed to either Ag isomers or different Ag sites are labelled f g g' respectively; higher molecular Ag cluster (n z 4-6) hypedine lines are depicted by h i; conduction electron spin resonance absorption of silver microcrystallites of z 10-20 8,dimension occurs around g = 2 and is indicated by (see p.53)j; the quartet hyperfine pattern of CH3impurity occurring aroundg = 2 is responsible for some spectral overlap problems; photolytically generated H-atom impurity is labelled H [see text and ref. (36)]. SPECTROSCOPY CHEMISTRY AND CATALYSIS e.s.r. e~periments~~p~~ (see for example Ag/Ar in fig. 3 about which more later) as weak isotropic lines with slightly smaller hyperfine coupling constants than the main 107A g(Z = 3; 51 %)/'09Ag(Z = 3 49%) doublet split atomic components.The monotonic approach of the main triplet optical absorption of Ag in Ar Kr and Xe matrices towards the gas phase atomic value [fig. 2(b)] reflects the decreasing repulsive interactions with the matrix atoms on changing to the more spacious substitutional cage of Xe. The origin of the major triplet structure for Ag in Ar Kr and Xe (actually a quartet in Xe under higher resolution condition^)^^ as well as a similar triplet observed for other 2s+1S ground-state atoms (e.g. Li,39 Cu,IJ9 Mg,40 etc.) has been assessed by several groups19,38-40 and turns out to be of pivotal significance to the appreciation of the photophysics responsible for light-induced diffusion of these matrix-entrapped atomic The two most popular recent proposals for the triplet structure focus attention on (i) a static crystal-field intera~tion~~ in which an assumed axial distortion of the ground-state cage (caused by a static or vacancy effect) can split the 2P state into three spin-orbit levels 2S1,2 -t -+ 2P3/2,j,=3,2; 2S1/2 2P3/2,j,= 1/2; 'S1,,+2P1/2, jz = 1/2 noting Kramer's degeneracy for each of the 'P states and (ii) Jahn-Teller considerations 39 *40 involving vibronic coupling of the 2P state with lattice vibration distortion modes of the appropriate symmetry.Although the crystal-field analysis provides a satisfactory explanation for the structure of the absorption spectra it is clear that different concepts are required to account for the luminescent and photolytic properties of silver atoms in rare-gas solids.In what follows we will show how these recently discovered properties can be understood in terms of the operation of the Jahn-Teller effect. The latter proposal is attractive in view of the e.s.r. evidence3' for a " cubic ground state " substitutional cage for Ag/Xe mixtures from the observation of isotropic 107Ag/109Ag hyperfine and 107*'09Ag/'29Xe superhyperfine coupling; m.c.d. evidence4' for the 'S-+'P Mg/inert gas triplet absorption a situation in which spin-orbit effects are expected to be minimal and Jahn-Teller effects dominate the excited-state splittings ; e.s.r. optical and theoretical evidence41 for a static elongated tetradecahedral MX12 coordination sphere for ground- state 2P A1 and Ga/inert gas Jahn-Teller van der Waals complexes; and 2S-+2P triplet absorption and significant Stokes-shifted luminescence for Li/inert gas mix- tures inter~reted~~ in terms of valence electron coupling of the 2Pstate of Li with E or T2,lattice cage vibrations active in the Jahn-Teller effect.On surveying a large number of emission and excitation spectra from freshly deposited (coexisting blue and red sites) and well annealed (blue sites predominating) Ag/Kr 2 1/104 matrices one can identify two major emissions at 415 and 495/525 nm which have their origin essentially in atomic excitation of the major (blue) triplet absorption at 309 313 and 323 nm.27a Typical emission and excitation spectra for this thermally stable site are shown in fig. 4(a) A (1 2 3) and 4(b) and in table 2 from which it can be ascertained that both major emissions can only be stimulated by excitation into the two high-energy components (309 and 313 nm) of the blue triplet fig.4(a) A (1 2) whereas the low-energy component (323 nm) yields only the long- wavelength emission fig. 4(a),A (3). In well annealed films these two major emis- sions essentially stand alone and can be unequivocally associated with the blue atomic silver site. Evidence for emissions originating from the thermally labile red atomic silver site can be clearly found in non-equilibrium quench-condensed films excited at 328 and 335 nm typical emission and excitation spectra being shown in fig. 4(a),A (4 5) and fig. 4(b). Table 2 serves to emphasise the similarity between the Ar Kr and Xe systems in emission for the blue and red atomic sites (less pronounced in Xe).A comparison of the major emission bands observed for Ar Kr and Xe matrices 318 328 I I 32 c 313 *In .-31 3091 f 3 313 5 328 \ -C ID > c .-In 311 C c .-C 335 .-0 In \ .-u) E -u111111111111111 Ill 290 310 330 300 320 wavelength Inm 3 00 LOO 500 600 wavelength Inm FIG.4.4~) Emission spectra (band pass 2 nm) observed from (A) freshly deposited Ag/Kr M 1/104 matrices (B) after 10-min 310-nm excitation of (A) and (C) after 10 min of 35 K annealing of (B) where (1) =309 (2) =313 (3) =323 (4) =328 and (5) =335 nm excitation wavelengths.*'" (b)Excitation spectra (band-pass 2 nm) of freshly deposited Ag/Kr FZ 1/104 mixtures at 10-12 K viewing at (A) 374 (B) 405 (C) 447 and (D) 500 nm.SPECTROSCOPY CHEMISTRY AND CATALYSIS TABLE 2.-cORRELATION OF Ar Kr and Xe EMISSION DATA ARISING FROM BLUE AND RED SILVER ATOMIC SITES~~~ [REF. (27a)l ArCvd KrcTd Xec*d blue red blue red blue red 360 (5.67) 375 (5.70) [5.17] [4.95] 367 (6.20) 415 (8.25) 480 (10.22) [5.70] [7.50] [9.22] 423 (9.80) 454 (10.34) 560 (13.20) E9.301 [9.59] [12.20] 457 (1 1.56) 495 (12.16) 550 (12.81) [11.06] [1 1.411 [11.81] 478 (12.52) 530 (13.50) 680 (16.35) [12.02] [12.751 [15.35] 478 (12.52) 525 (13.13) 670 (16.13) c12.021 [12.56] [15,131 ~ ~~~ a Tentative Ar-Kr-Xe correlations are indicated by emissions placed on same line. Overall (lo3cm-') energy shifts from highest energy absorption maxima are given in parentheses.Approxi- mate values corrected for excited state vibrational relaxation are also included in square brackets. Units of emissions in nm. All emissions were essentially depolarized. is shown in fig. 5(a). Note that in addition to the long-wavelength bands the Ag/Ar and Ag/Kr emission spectra each show the presence of a narrow band between 300 and 350 nm strong in Ar quite weak in Kr and entirely absent in Xe. The weight of evidence from the fluorescence excitation spectra indicates that the aforemen- tioned emission arises from a 2D-+2S transition where 2D denotes a slightly per- tubed silver atomic state in which matrix-induced 2P-2 D configuration mixing effects may be important* (the 2Dlevels are the only silver atomic states in the same-energy range as the 2P It seems likely that this type of transition is also res- ponsible for the fourth band seen in the Ag/Xe absorption spectrum so that co- valent interactions need not be invoked in this c~nnection.~' Because of the localized tightly bound character of the 4d orbitals in silver atoms the 2D(4d95s2) excited state does not interact strongly with the matrix cage and very small Stokes shifts are ob- served between the fluorescence emission and excitation maxima; this is to be sharply contrasted with the very large Stokes shifts observed for the 2P(4d105p1) excitations.A study of the emission spectra of copper atoms in Kr provides additional support for this assignment. The 'S -+ 2P Cu excitations produce a very intense emission band at 754nm corresponding almost exactly with the 2S1,2-2D3/2 separation of free copper atoms.27b Thus it seems certain that the 'D,,,state of copper atoms is populated from the 2P levels and that a radiative transition subsequently occurs between the 2D3/2and 2S1,2 states.The emission spectra of silver atoms isolated in Ar Kr and Xe matrices undergo pronounced changes when the matrix deposits are exposed to U.V. light centred near the atomic absorption bands. In the case of very dilute Kr matrices (Ag/Kr NY 1/104) these changes [fig. 4(a),B] can be exactly reversed by thermal annealing [fig. 4(a),C]. In more concentrated matrices the U.V. exposure causes an irreversible loss of emission intensity due to photo-induced diffusion/aggregation processes.Analysis of these u.v.-thermal-induced spectral alterations27a points to a process * We note that the 2D5/2+2S./2 parity and J forbidden transition has been observed in the gas- phase emission spectrum of atomic silver at 330.6 nm.42 h VI c .-C $ Y > 4-.-n UI aJ Y Ag/Kr C ? .-0 C N .-ul .-E hI* ul Ag/Xe k; / II 8 300 400 500 600 700 wavelength /nm FIG.5.-(a) Comparison of the major emission bands of silver atoms isolated in Ar Kr and Xe matrices. The corresponding excitations are illustrated on the absorption spectra shown on the left.27a (b)Emission spectra observed from freshly deposited Ag/Kr matrices at (A) 1/104,(B) 1/103and (C)1/102[same excitation nomenclature as used in fig.4(a)]."" SPECTROSCOPY CHEMISTRY AND CATALYSIS in which light-induced diffusion of silver atoms under " high-dispersion conditions '' leads to a " redistribution of trapping sites " favouring the thermally unstable over the thermally stable ones with only " minimal accompanying photoaggregation ". This photo-induced generation of a non-equilibrium distribution of atomic silver sites can be efficiently relaxed by 35 K thermal treatment of the matrix presumably rein- stating the original local structure of the trapped silver atoms. The fluorescence excitation spectra shown in fig. 4(b)provide direct evidence that secondary trapping sites are involved. These spectra show at least nine distinct absorption maxima of which only four are thermally stable (309 313 322 and 328 nm).Note that traces A and C in fig. 4(6) do not show the presence of the primary absorption maxima at 309 313 and 322 nm indicating that the corresponding emis- sion bands are associated exclusively with secondary trapping sites. Incidentally this effect has been simultaneously monitored by e.s.r. spectro~copy,~~*~~ confirming the proposed site redistribution. Other interesting observations concern the metal concentration dependence of the emission spectra ( 1/104-1/102) shown in fig. 5(b);27n the corresponding absorption spectra indicate that slightly larger numbers of absorbing atoms were present for the higher-concentration samples. The overall reduction in emission intensity and preferential quenching of the high-energy emissions at higher silver concentrations seems to depict a situation where either energy transfer or electron transfer from Ag to Ag,,,, becomes important or selective internal filtering of atomic emissions occurs by Ag,,,, [cf.fig.2(a)],the latter being especially pronounced in the 360-400 nm range.,'" It is hoped that future studies involving both steady-state and time-resolved fluorescence measurements will clarify these remarkable concentration quenching effects. A study of the degree of polarization of the luminescence was carried out using polarizers in the excitation and emission beams.27a The polarization P,is defined as P f(41 -IJ(41 + Il) where 41 and ZL are the emission intensities measured for parallel and perpendicular orientations of the emission polarizer relative to the excitation polarizer.The measured values of P,which were corrected for instru- mental factors,27c were close to zero for each of the emission bands observed for Ar Kr and Xe matrices (see later). Preliminary lifetime measurements have been made for the 550 and 680 nm Ag/Xe emission bands. Transient fluorescence signals were produced by pulsed 337-nm nitrogen laser excitation of very dilute (Ag/Xe z 1/104) matrices; the luminescence appeared instantaneously with the laser pulses and decayed exponentially with time. The characteristic decay times were found to be 28 & 3 and 56 5 5 ns for the 550 and 680 nm bands respectively. When the inherent frequency dependence of the transition probability is taken into account these decay times agree very well with the mean radiative lifetime of the 2P3,2level of free silver atoms27d (6.5 -J= 0.6 ns).This result suggests that the fluorescent states decay exclusively by the observed radiative pathways since lifetimes shorter than the free-atom value would be expected if concurrent non-radiative decay processes were important. The probable nature of the relaxation processes which occur between absorption and fluorescence will be dis-cussed after the vibronic coupling model for the Ag (,P)-rnatrix cage interaction has been described.274 As a structural model for the major trapping site of silver atoms in rare-gas solids we adopt the generally accepted twelve-fold tetradecahedral substitutional site having octahedral symmetry.The 2P excited state of silver transforms as T, in the Ohsymmetry site and is therefore a Jahn-Teller unstable state. The vibronic prob- lem involving a ,TlU electronic state subject to strong spin-orbit forces has been fully discussed by MoranZ7" and by Fulton and Fit~hen~~~ in connection with the optical G. A. OZIN absorption spectra of F-centres in caesium halides. In brief the spin-orbit interac- tion causes a splitting of the TI state into G3/2 and components in octahedral symmetry (crystal double-group notation appropriate to the description of half- integral angular momentum states). Although the Kramer's degeneracy of the level cannot be lifted by vibronic interactions a Jahn-Teller instability is expected for the four-fold degenerate G3,2ustate.27g The vibrational modes that can couple with a G3,2ustate are contained in the antisymmetric square27g {G3/2E) = A,, + E -+ T2,. The assumptions in the model are that spin-orbit coupling is much more important than the vibronic interaction and that vibronic coupling involves onlyia single (dominant) E distortion mode neglecting interactions with the T2,modes (fig. 6 A and B). The potential-energy surface for a doubly degenerate electronic state (in Oh symmetry) interacting with a doubly-degenerate vibrational mode assuming harmonic A B C Y*'Q3) D ::I (Q 0:); FIG.6.-Tetragonal (A) and trigonal (B) distortions of a tetradecahedral complex. The extremal geometries have D4h and symmetries respectively.(C) Schematic potential-energy surface for a doubly degenerate electronic state (in Oh symmetry) interacting with a doubly degenerate vibrational mode neglecting anharmonic effects. V represents the nuclear potential energy and Q2 and Q3 represent the degenerate components of an Eg vibrational m~de.~'~-g (D) Schematic representation of the ground- and excited-state potential-energy surfaces for Ag X12complexes (where X is Ar Kr or Xe). Q2 and Q3 are " quasi-molecular " normal coordinates corresponding to the degenerate components of an Eg vibrational mode. If anharmonicity is included (see text) the excited-state surfaces have trigonally disposed minima representing equivalent tetragonal distortions along the x y and z directions.The absorption and emission processes are illustrated. SPECTROSCOPY CHEMISTRY AND CATALYSIS potentials and only first-order vibronic interactions is illustrated in fig. 6C. The separation between the upper and lower branches at any point in the (Q, Q3)co-ordinate plane of this double-valued surface labelled U and U, refers to the vibronic splitting of the G3/2,, excited state of silver atoms. If anharmonic terms are included in the expansion of the nuclear potential then the surface of fig. 6C loses its cylindrical character and develops three equivalent trigonally disposed minima separated by saddle points corresponding to the three equivalent tetragonal distortions along the x y and z direction^.^'^ M~ran,~~ mixing effects by means takes account of 'P3/2-'P1/2 of a second-order perturbation calculation.A schematic representation of the ground- and excited-state potential-energy surfaces which includes all of the effects mentioned above (that is strong 'TIuspin-orbit splitting to give 2G3/2u,2E1/2u states 2G3/2u vibronic coupling with the E distortion mode with second-order corrections arising from the interaction between the 2E1/2u spin-orbit components) is and 2G3/2u shown in fig. 6D. Without going into any further detail we can use the vibronic coupling model and schematic configuration coordinate diagram of fig. 6D to explain the optical absorp- tion emission and photolytic properties of silver atoms entrapped in rare-gas solids. The manner in which the triplet structure arises is illustrated in fig.6D. Franck-Condon vertical transitions from the maximum in the probability distribution of the ground-state vibrational wavefunction cross three excited-state potential-energy surfaces resulting in a three-fold splitting of the optical absorption band. Note that a model in which vibronic coupling is much more important than the spin-orbit interaction would predict a single Gaussian absorption band,27f and is therefore inappropriate to the present problem. Optical excitation to the U, U2or U3levels should be followed by rapid vibra- tional relaxation to the minimum point of the respective potential-energy surface and subsequent radiative decay to the ground-state surface as illustrated in fig. 6D. The large Stokes shifts are seen to be a consequence of the tendency for distortion of the excited-state complex and the accompanying destabilization of the ground state.The destabilization caused by producing ground-state complexes in the relaxed excited-state configuration can be visualized as repulsive forces generated at the photolytic centres which can cause the entrapped silver atoms to undergo diffusive "jumps ''to adjacent lattice sites. The photodiffusion process can also be thought of as the result of a local heating effect caused by dissipation of the ground-state de- stabilization energy in the form of localized lattice (cage) vibrations (photoselective bi- metallic aggregation experiments,22~28 described later provide strong support for these proposals). The striking trend of larger Stokes shifts for silver atoms isolated in the heavier rare-gas matrices [fig.5(a)]can now be understood in terms of (a) increasing stability of the excited-state cage complexes along the series Ar to Xe arising from deeper potential-energy wells (stronger van der Waals-type binding) for the more polarizable rare gases and (b)accompanying destabilization of the ground state in the same order. According to the vibronic coupling model both the upper and lower excited-state potential-energy surfaces (labelled A and B in fig. 6D) should have trigonally dis- posed minima representing equivalent tetragonal distortions along the x y and z directions. Another important topological feature for the upper surface B is the continuity between the U and U branches.Thus vibrational relaxation following excitation to the U and U2branches must lead to fully equivalent structural con- figurations of the relaxed excited states. Therefore the same emission spectrum should be produced for both U and U excitation which should in general be differ- ent from the Uqexcitation. This is the proposed origin of the two major emission G. A. OZIN 21 band systems. The appearance of both emission band systems for the high-energy (Ul and U,) excitations in Ar Kr and Xe matrices implies that non-radiative transi- tions can occur between the upper (B)and lower (A)potential-energy surfaces. The potential for three equivalent tetragonal distortions of the excited-state com- plex along the x y and z directions suggests an explanation of the observed fluores- cence depolarization.Since the states produced by optical excitation are highly vi brationally excited levels of a distorted configuration there exists the possibility of fluxionality in the excited state allowing interconversions between differently oriented configurations (dynamic Jahn-Teller effect). In this way all memory of the polarization of the incident photons is lost and the fluorescence is completely de- polarized. In summary one can provide a satisfactory interpretation of the ,S+ 2P ab-sorption fluorescence and photolytic properties of silver atoms in rare-gas solids in terms of a Jahn-Teller effect operative in the degenerate electronic state of a 2P (energetics and dynamics) excited silver atom-cage complex.DISILVER CRYOPHOTOCHEMISTRY Besides giving an understanding of the photophysics of light-induced diffusion of silver atoms absorption and emission spectroscopy are also able to provide pivotal clues concerning the radiative and nonradiative processes accompanying the recently observed cryophotochemical transformations of Ag and Ag in noble-gas lid^.^^^^^,^^ In particular one would like to develop a photophysical model that simultaneously provides an explanation for the photochemical responses (e.g. photofragmentation photoisomerisation) and fluorescence behaviour following ultraviolet and visible excitation of selected silver cluster electronic states. Cluster-based SCF-Xcr-SW computations are proving to be an indispensable adjunct for interpreting the experimental data emerging from this highly novel form of naked cluster cryo- photochemi~try.~~*~~ Besides the fascinating photochemistry photophysics and spectroscopy that are beginning to evolve from such investigations it seems that one may be able to gain access to unique cluster distributions by exploiting the outcome of selective (sequential and/or simultaneous) photoexcitation of small cluster arrays.3o Recently we have been able to demonstrate that narrow-band irradiations into the visible absorptions of immobilised Ag (and Cu,) cause photofragmentation of the diatomics as seen in the absorption spectrum by decay of Ag (and Cu,) with simultaneous growth of Ag (and Cu) and at higher concentrations Ag (and CU~).,~ These effects for Ag,/Kr are illustrated in fig.7(a). Ag and Cu are particularly appropriate to such experiments for two reasons. Firstly their spectra have been rather thoroughly studied both in the gas phase at high temperature^^^ and in inert- gas matrices at low temperature^.^^*^^*^^ Vibration-rotation analyses of the gas- phase spectra have produced precise values for the 0 -+0 transition energies metal- metal stretching frequencies and for Cu, the internuclear distance (2.2195 A). Dissociation energies have been obtained from mass spectra.45 The assignments pro- posed for the observed electronic transitions from the experiments and from simple MO calculations of the extended Hiickel CNDO or Xcr type29*46 serve as a guide to selection of transitions for the photochemical studies.Secondly the inherently simple electronic structures expected for Cu and Ag further aid the design and interpretation of the photochemical experiments. The metal-metal bonds should largely involve simple pairing of the s electrons on the two dlOsl metal atoms. One of the lowest energy absorptions should there- fore involve excitation from the bonding sog to the antibonding so orbital. h, A h I JjD 43 E 43 L I I 300 L 00 500 wavelengthlnm FIG.7.-(a) Optical spectrum of Ag/Kr E 1/104 deposited at 12 K (A) showing isolated Ag atoms and Ag molecules; (B) after 30-min 312-nm narrow-band continuous photoexcitation of the Ag atomic resonance lines showing photoaggregation to the Ag stage; (C) after 70-min 390-nm excitation of Ag, showing mainly Ag photodissociation to Ag atoms.The Ag high-energy band labelled (a) is thought to be a site splitting effect. Note scale change between 325 and 400 nm.29 (b)Optical absorption and emission spectra for Ag, ,,JAr mixtures where traces A-E refer to excitation at energies corresponding to A-E respectively in the absorption spectrum shown in top trace (A =Ag; B C =Ag,; D E =Ag,).,’” G. A. OZIN The gas-phase spectra strongly suggest that the apparent so -+ sou excited states of Cu and Ag are bound with extensive vibrational structure being How-ever recent experiment^,^ indicate that the behaviour of these excited states in a condensed matrix phase is quite different from that which is observed in gas-phase studies.As will be seen later our interpretation of irradiation near 400 nm in this context suggests that the factors which are important in fig. 6 are also important in determining the fate of the excited molecular states of Cu and Ag in cryogenic rare- gas solids. It is clear from the observations described in fig. 7(a)and other experiments that the end result of Ag visible photoexcitation in terms of the observed yield of isolated atoms as the photodissociation products is strongly dependent upon the concentra- tion conditions employed but that the excitation invariably results in a reduction of the Ag concentration. A similar observation applies for the Cu photoexcitations. In the case of both Cu and Ag the observed yield of isolated atoms generally in- creases with decreasing total metal concentration.Typical yields are low being near 20% for Ag/Kr [fig. 7(a)]. Excitation at many of the absorption wavelengths of Ag (and Ag,) in for example Ar matrices results in intense part of which always corresponds closely with emission bands associated with the respective silver atomic species [fig. 7(b)] implying that the emitting species is the same for both Ag and Ag photo- excitation^.^^" Let us briefly describe some of the emission and excitation data for matrix-isolated Ag in Ar. Fig. 7(b) shows that the 260 and 390 nrn Ag excitations both produce the same 478 nm emission band which corresponds closely with the band produced by the 2S-f 2P excitation of Ar-entrapped silver atoms.27a Fig. 8(a) shows the effect on the Ag emission spectra of prolonged (40 min 20 nin slit) 305 nm atomic Ag excitation (1) an increase in the overall Ag emission intensity (2) for both 260 and 390 nm excitation the 454 nm band nearly disappears while the 478 nrn band increases in intensity (same as site effect seen for atomic excitations) (3) for the 260 nm excitation the strong 284 nm emission decreases in intensity relative to the 478 nm band.In fig. 8(a) on an expanded scale one observes that 260 nm Ag excitation produces emissions also at 326 365 and 423 nm which correlate exactly with the emission bands produced by atomic 299 nm excitation (table 2).27h Fig. 8(b) shows the Ag,/Ar excitation spectra in which the following points are noteworthy (1) The excitation maxima near 260 nm are displaced slightly in the 284 nm excitation spectrum relative to the corresponding maxima in the 480 nm excitation spectrum (photostable and photolabile Ag sites giving rise to molecular Ag,* emission and excited silver atom-cage complex emission re~pectively~’~).(2) The 480 nm excitation spectrum shows a broad maximum near 272 nm which does not appear in the 284 nm excitation spectrum but which does correlate with a shoulder in the absorption spectrum [fig. 2(a)]; this shoulder is believed to be due to a secondary Ag trapping (3) The 480 nm spectrum shows very intense excitation maxima at 390-410 and 440 nm in line with the high intensity of the 480 nm emission band due to 390 nm Ag excitation (some Ag contribution also see next section).(4) The 480 nm spectrum shows a distinct excitation maximum at 410 nm correlating with the partially resolved 410 nm Ag (secondary site) absorption band [fig. 2(a)]. A straightforward interpretation of the 390 nm emission data under high dilution conditions (Ag/Ar FZ 1/104) therefore involves a two-site Ag and two-site Ag pro- posal as indicated in scheme A below for Ar matrices:27h h) P h v) I.. .-C ? c) n 0 L CJ Y 261 265 250 LJ 1 I 250 300 250 300 350 400 wavelength I nm I I I 300 400 500 wavelength / nm FIG.&-(a) Emission spectra of Ag/Ar M 1/10 mixtures at 10-12 K showing (A) 390-nm and (B) 260-nm excitation of Ag (2~,-+20 and 20 -+ 2n transitions respectively) on deposition and (C) 390-nm and (D) 260-nm excitation of Ag after 40-min 305-nm irradiation [see text and ref.27h)l. (6)Excitation spectra of the Ag/Ar z 1/10 matrix shown in (a) A and B observed at the (A) 284-nm and (B) 480-nm emission frequencies ascribed respectively to vibrationally relaxed (roughly 20 vibrational quanta) molecular Ag and the excited silver atom-cage complex (see text). Some contributions to the 480-nm emission can be seen to originate from the 245-nm band of Ag (see trisilver section). Note a 430-nm filter was used in this e~periment.~’~ G. A. OZIN SCHEME A where Ag(Ar)* refers to the A potential energy surface (fig. 6D) of the excited silver atom-cage complex (I) and (11) to the two different sites of Ag and Ag and Ag+(,S) to a photomobile ground-state silver atom.The 260 nm Ag emission results can be rationalised as shown in scheme B below for Ar matrices 27h SCHEME B In essence one is obliged to propose that Ag formed on deposition resides in two distinct trapping sites Ag,(I) and Ag2(11) which can be photodissociated as well as grown and redistributed by silver atom photodiffusion and matrix annealing so as to favour Ag,(I) and Ag2(11) respectively (fig. 8). The details of selective quenching and energy-transfer phenomena which are thought to occur under higher concentra- tion conditions can be found in forthcoming publications.27h The important step of our proposed mechanism for Ag photodissociation involves Ag-Ag bond breaking in the excited state [O in the case of (J-J) coupling] and simultaneous Ag(Ar),* bond formation to form an excited state silver atom-cage complex.This step amounts to a transfer of electronic excitation from Ag to Ag(Ar),* and elimination of a ground-state Ag atom. The final step of the photo- dissociation following radiative decay of the excited silver atom-cage complex states would then be dissociation of the repulsive ground-state-cage complex thus producing a second (mobile) ground-state Ag atom (as described in the previous section). Although we have not yet undertaken quantum efficiency measurements the intensity of the emission bands resulting from 390 nm Ag excitation suggests that the quantum efficiency is high. The formation of mobile ground-state Ag atoms as a result of Ag visible photoexcitation may explain the low yield of isolated atoms since the mobile photodissociation products are potentially reactive in subsequent aggre- gation processes.Finally we note that the 'C,+ state of Ag correlates best with a 2S+ 2P separated atom limit,44b.f and that 2P atoms are produced as a result of the atomic excitations. Hence it is not unreasonable to suggest that the emitting species is the same for both 390 nm Ag and 300 nm Ag excitation^.^^" 27h I I1 Ag2 2w9 FIG.9.-(a) Ground-state SCF-Xcr-SW valence energy levels for Cu Cu, Ag and Ag. The bands of molecular levels most closely correlating with atomic d s and p levels are indi~ated.,~ (b)(I) Con-tour maps of the wavefunctions for the mainly s-bonding orbitals of Cu and Ag at 2.22 and 2.84 A respectively.The map for Ag at 2.47 A is similar. Note the significant amount of d character in the Cu case. The contour values for these maps are 0 1 2 3 4 5 6 7 = 0 0.02 0.03 0.04 0.05 0.06,0.09,0.15,re~pectively.~~ (11) Contours maps of the wavefunctions for the lowest unoccupied orbitals (mainly s antibonding) of Cu2 and Ag2 at 2.22 and 2.84 A respectively. The map for Ag at 2.47 A is similar.29 TABLEVA VALENCE ENERGY LEVELS (Hartrees) AND CHARGE DISTRIBUTIONFOR Cu2 AND Ag229 cu2 2.22 A Ag, 2.84 A Ag, 2.47 A Dmh % charge" cu sphereb Dmh % charge' Ag sphere' Dmh % chargeb Ag sphere" level energy level energy level energy Cu outer %d %s %p Ag outer %d %s %p Ag outer %d %s %p +0.0074d 20 81 5 6 89 30 -0.0040 20 75 2 2 96 30 -0.0007 6 93 3 20 77 -0.0309 19 48 7 93 2n -0.0329 21 38 3 97 27K -0.0363 16 46 4 96 -0.0732 46 36 8 74 18 20 -0.0843 55 24 2 86 12 20, -0.0685 40 38 4 79 17 -0.1611 96 1 95 4 1 20" -0.1652 68 6 4 89 7 2ogc -0.1684 56 11 5 89 6 -0.1627 98 1 99 1 lo -0.2575 98 0 98 1 1 lo -0.2467 96 0 96 2 1 -0.1743 69 8 49 49 2 l~ -0.2629 99 0 100 0 l?t -0.2555 97 0 100 0 -0.1755 98 0 100 1% -0.2712 98 0 100 16 -0.2729 97 0 100 -0.1836 96 1 100 16 -0.2747 98 0 100 16 -0.2812 96 0 100 -0.1992 97 0 99 1 ln -8.2833 98 0 100 0 ln -0.2991 96 0 100 0 -0.2380 100 0 63 28 9 lo -0.2975 100 0 95 3 2 lo -0.3245 100 0 93 4 3 Percentage of electrons within the metal and outside the outer spheres.The balance is in the intersphere region. Relative spherical-harmonic contributions to the charge density within the metal spheres.Note that this represents only a minor fraction of the charge for the 2n and 30 orbitals. The highest occupied levels. Position estimated from transition-state calculations where it becomes a bound level. SPECTROSCOPY CHEMISTRY AND CATALYSIS Let us briefly consider the ground-state bonding and assignments of the optical spectrum of Ag (and Cu,) from SCF-Xa-SW calculations at the calculated minimum energy geometries of 2.84 A (and 2.22 A) re~pectively.~~ The one-electron energies and orbital contributions are summarised in table 3 and the energies are compared with those of Ag and Cu atoms in fig. 9(a). The bound valence orbitals include in order of decreasing energy the six completely filled a and 6 bonding and antibond- ing d-band levels; the s-band with an occupied a bonding and empty a antibonding level; and empty 71 and a bonding levels from thep-band.The major difference between the two molecules is that the s-band overlaps the upper part of the d-band in Cu, while the two are very well separated in Ag,. The d-like la and In levels are thus the highest occupied in Cut rather than the 20 level as in Ag,. As can be seen from fig. 9(a) this is a consequence of the much smaller separation of the d and s atomic levels in Cu than in Ag. It leads in Cu to a signifi- cant mixing of 4s character into the formally 3d-derived la orbital and 3d character into the formally 4s-derived 20 orbital. In contrast the la orbital is nearly pure 4d in Ag and the 20 nearly pure 5s.Fig. 9(b) provides a pictorial representation of the difference between the 2a wave functions of the two molecules. The un- occupied orbitals are similar in the two; the lowest energy 20 is essentially s as seen pictorially in fig. 9(b),and the higher-lying 271 and 30 are essentially p. The symmetry of the occupied orbital manifolds alone dictates that there is a net single bond of a type in both molecules. The relative d-s-p character of this net bonding may be estimated by adding together the percentage contributions of the spherical harmonic in question to the bonding la and 20 orbitals and subtracting out the contribution to the antibonding la orbital. The result using the numbers in table 3 is 73% s 17% d and 10% p for Cu, compared with 91 % s 1% d and 8% p for Ag at 2.84 A (the results at 2.47 A are the same within 1 %).Table 4 shows our assignment of the electronic transitions of Cu and Ag ob- TABLE 4.-cALCULATED AND EXPERIMENTAL ELECTRONIC SPECTRA OF CUz AND Ag,29 experimental transit ion calculated” gasb Ar Kr Xe cu2 lng -+ 20 24.1(0.02) 20.4 25.0 25.0 25.0 20 -+ 20 26.5(0.32) 21.7 27.O 27.0 27.8 ln -+ 2n 35.2(0.15) 28.3 37.0 35.1 20 -b 2n 37.2( 1.02) 41.7142.4 41 Sl42.0 40.5141.O 16 -+ 2n 3930.1 3) 43.1 42.7 41.7 lo -b 20 39.8(0.36) 44.8 44.0 43.1 lo -+ 30 43.7(0.03) 20 -+ 20 22.0(0.64); 25.7(0.63) 23.0 24.3125.8 25.6 25.6 20 + 2n 33.2(1.33); 33.6(1.37) 36.6 37.8138.3 35.5137.0 34.5135.3 In + 20 43.5(0.03); 45.6(0.04) 40.2 44.1 45.0 46.1 (I All spin- and dipole-allowed transitions below (48 and 51) x lo3cm-I for Cu and Ag, respec-tively.For Ag, the first value is for 2.84 and the second for 2.47 A. Oscillator strengths are given in parenthesis. 0 +0 transitions from ref. (44e) for Cu2 and ref. (44f) for Ag,. The weak B tX and D +-X bands of Ag at 35.8 and 39.0 x lo3cm-’ believed due to forbidden transitions are omitted. G. A. OZIN served in the gas phase at 2000-2300 K and in solid inert-gas matrices at 12 K. The calculated positions specifically represent the singlet (spin-allowed) components and include orbital relaxation between ground and excited states [see ref. (29) for details]. TRISILVER CRYOPHOTOCHEMISTRY As mentioned earlier apart from subtle matrix site perturbations the photochemi- stry and emission spectroscopy of disilver are essentially common to Ar Kr and Xe However trisilver photochemistry 25933 and emission seem to be more sensitive to the matrix environment than di~ilver.’~ Let us briefly consider the data for trisilver.The optical absorption bands attributed to Agl,,, isolated in Ar Kr and Xe matrices 20*25 are illustrated in fig. 2(a). Because of the difficulties involved in arriving at an unambiguous correlation of absorption bands with specific cluster species the assignments for some of the weaker features of the absorption spectra must be re- garded as tentative although the correspondence with the work of Schulze and co- worker~~~ is gratifying (see paper at this Symposium). However it is our contention that the results of the trisilver photochemistry study2’ provide support for these assignments.A detailed analysis of the trisilver absorption spectrum is subject to a number of complications. The Ag band assignments for example must be based on a molecu- lar orbital model together with an assumed or predicted geometry48 since no gas- phase spectroscopic data are presently available for Ag,.* Additional complications arise from the possible coexistence of geometrical isomers or spectroscopically distinct matrix trapping sites.? In this connection we believe that the photochemical studies described below show potential for aiding spectral assignments and for elucidat- ing the optical spectroscopic consequences of structural rearrangements of matrix- entrapped trisilver molecules.The outcome of photoexcitation centred at the visible band of entrapped Ag in Ag1,2,3,4/Kr mixtures (Ag/Kr N 1/102)at 12 K is illustrated in fig. lO(a). A compari- son of spectrum B with spectrum A shows that major spectral changes result from the photolysis and that these alterations are restricted to the set of absorption bands believed to be associated with Ag,. Thus while the bands labelled Ag Ag and Ag remain essentially invariant each of the original Ag bands undergoes a substantial decrease in intensity and a number of new features appear the most prominent of which is centred near 445 nm. Spectrum C in fig. lO(a) illustrates the interesting result that the effects of the visible trisilver photolysis can be essentially exactly reversed by a brief 25 K thermal annealing period.Thus the photo-induced trans- formation is thermally reversible and can be recycled many times without major modifications to the optical spectrum. Notable also is the observation that the original Ag absorption spectrum can be similarly regenerated by irradiating the new band at 445 nm. Fig. lO(b) shows results for Xe matrices closely analogous to the Kr results illus- trated in fig. lO(a) indicating a similar photochemical behaviour of Xe- and Kr- entrapped trisilver. The results shown in fig. 10(a) and lo@) indicate that visible Ag photoexcitation in Kr and Xe matrices results in a highly selective thermally and photolytically reversible trisilver phototransformation. These results for Kr and Xe * Gingerich and co-workers have now obtained high-temperature mass spectroscopic data for Cu3 Ag3 and Au (personal communication).i We note with reference to the question of matrix “ site effects ” that two magnetically distinct trapping sites of Na3 in an argon matrix have been identified in an e.s.r. and that a rather complex picture of geometrical isomers and/or multiple trapping sites is emerging from our (unpub- lished) e.s.r. investigations of Ag3 isolated in Ar rnatri~es.~~.~~ 0 u C 0 13 L 0 Y n 0 & wavelength Inm 250 300 350 400 450 500 wavelength In m FIG.10.-(a) U.v.-visible spectra of Ag1,2,3,4/Kr mixtures (Ag/Kr z 1/102) at 12 K (A) after a 30-min irradiation centred at the atomic re- sonance absorption lines; (B) the outcome of a 10-min 423-nm Ag irradiation showing major decay of the bands associated with Ag3 (indicated by arrows) and the appearance of two new bands near 450 nm; (C) the result of a 5-min 25 K bulk thermal annealing period showing re- generation of the original Ag spectrum and loss of the new band near 445 nm.25 (6) U.v.-visible spectra of Ag1,2,3,4/Xe mixtures (Ag/Xe w 1/102)at 12 K (A) after a 30-min irradiation centred at the atomic resonance absorption lines; (B) the outcome of a 10-min 440-nm irradi- ation showing major decay of some of the bands associated with Ag3 and growth of other Ag3 bands (indicated by arrows); (note the appear- ance of the new band near 470 nm and the invariance of the Ag Ag2 and Ag bands); (C) the result of a 5-min 30 K bulk thermal annealing period showing regeneration of the original Ag spectrum.25 G.A. OZIN supports are in distinct contrast to the results for Ar supports where prolonged trisilver visible photoexcitation produces no observable changes in the optical spec- trum. Recent fluorescence spectroscopic studies ,''are similarly indicative of a differ-ent photochemical behaviour of trisilver entrapped in Ar as compared with Kr or Xe matrices. In the case of Ar matrices excitation centred at the visible (and u.v.) band of trisilver produces a very intense emission which corresponds exactly with the major emission band produced by atomic silver excitation [fig. 7(b) D and El. These results alert one to the possibility of an efficient Ag,Ar photo-dissociation- recombination process localized in the matrix cage.* Recall that no such emission is observed for the analogous trisilver excitations in Kr and Xe matrices.At the present time we wish only to point out that Ar-entrapped trisilver molecules behave differently from the Kr-and Xe-entrapped Ag species. Further discussion of this point must await the outcome of more detailed studies including cluster fluorescence lifetime and polarisation Apart from the possibility of a photoionisation process which we discount based on estimated Ag cluster ionisation potential^,^^ the occurrence of two different photolytic processes for Ag in Kr and Xe matrices can be suggested to explain our observations (1) a photodissociation process (Ag -+ Ag + Ag) and (2) " photoiso-merisation " of the trisilverlmatrix cage unit involving either different geometrical isomers of Ag or different orientations of a " rigid " Ag molecule within a deform-able matrix cage.The invariance of the Ag Ag and Ag absorption bands during the Ag visible photolysis/thermal (or photolytic) regeneration cycle and the absence of atomic silver emissions (cf. Ag,/Ar) provides a strong argument against the photodissociation proposal. Based on the relative extinction coefficient meas uments for matrix- isolated Ag,,2,3,23 described earlier we conclude that a photodissociation process could have been detected through substantial intensity variations of the Ag and Ag bands. Also inconsistent with the photodissociation proposal is the fact that visible Ag photolysis results in the appearance of new absorption bands.These bands seem to be associated with Ag rather than Ag or Ag. Similarly the observation of thermal and photolytic reversi bility would argue in favour of a non-dissociative process. We conclude from these considerations that a " photoisomerisation " process is responsible for the spectral changes illustrated in fig. lO(a) and (6). At this stage it is possible only to speculate on whether the proposed structural interconversion can be described in terms of the geometry of Ag alone or whether it is necessary to consider also the geometry of the matrix cage. In view of the orbital symmetry and energy correlations which can be deduced for an Ag linear (see Xa calculations later on4 and the papers by Schulze and Grinter in this Symposium) to non-linear to D,,,triangular structural interconversion it is at least feasible that selective 423 nm Ag3/Kr or 440 nm Ag,/Xe HOMO-LUMO population of the strongly antibonding molecular orbital of a linear three-level Ag cluster 43 (fig.23) that is could induce photoisomerisation to a " matrix-stabilised " (at 10-12 K) triangular (a3'(e')'D3,or acute isosceles (a,)*(b,)'C, form of Ag,.t$ Of course 25 K Kr or Xe * Similar photo-dissociative emission and cage-recombination processes have recently been observed for the vacuum U.V. 'C,+-t 'nu and 'Z,+ -+ 'E2 excitations of XeF,/inert gas films which lead to XeF(B2Z+)+ F('P) and XeF(D2n) + F(2P) emitting states of XeF.50 t Relevant to these considerations is an ab inifio study of the dynamic Jahn-Teller effect in the electronic ground state of Li3.52 1(al)'(al)' C2nfor obtuse isosceles.SPECTROSCOPY CHEMISTRY AND CATALYSIS matrix annealing could provide just sufficient thermal energy to permit cage softening and back-conversion of " metastable '' Ag to the ground-state linear form of Ag,. Further experimentation with for example combined f.t.i.r. Raman and e.s.r. spectral monitoring will be required to clarify the intriguing visible phototransforma- tions of Ag as well as a final decision on its ground-state geometry. PHOTO M AN IP ULA TI ON OF CLUSTER D I ST RI B UTI ON S It is worth mentioning briefly some recent experiments in which it has been demon- strated that cryophotoaggregation experiments involving matrix-entrapped silver atoms can be tailored to the point of generating almost pure disilver clusters as well as cluster distributions that are inaccessible by conventional deposition and bulk annealing proced~res.~' By carefully selecting the silver concentration and matrix support it can be arranged to convert substantial amounts of Ag atoms to Ag without too much conversion loss to Ag and higher silver clusters.Fig. 11 demon-strates this remarkable atom-diatom photoredistribution reaction where (aside from site redistribution effects) a matrix containing large quantities of Ag relative to Ag has been produced (the best photoconversion yields realised so far are 80%). The potential of the method for generating very narrow distributions of Ag,/Ag in the absence of Ag has been realised in cyclo-octane matrices30 and of Ag,/Ag,/Ag with very little Ag in methane matrices3' [fig.11 (b)]. The ramifications of these kinds of experiments in for example single-cluster EXAFS fabrication of diatom and diatom/triatom cluster catalysts and photosintering are clearly considerable. SILVER-CONTAINING BIMETALLIC CLUSTERS The interactive electronic architecture which has recently emerged from the very low nuclearity cluster systems and Cr,Agms5 entrapped within weakly interacting low-temperature supports has provided an interesting new perspective with which to view highly dispersed multimetallic cluster catalysts comprised of partially immiscible or wholly immiscible component^.^^^^^ In contradistinction the striking non-interactive characteristics observed at the atomic level for the Pd/Ag and P~/MO*',~~ combinations alert one to the fact that the simple mixing of metal atomic vapours under cryogenic conditions is not sufficient to ensure reactive en- counters and the formation of bimetallic clusters.Clearly more subtle factors of an electronic and kinetic nature can contribute to and markedly affect the early stages of cluster growth at low temperatures even though the respective bulk-phase dia- grams may imply total miscibility of the metallic components over the entire con- * cent ra t io n range. In order to determine the generality of the above observations and the scope of the method for synthesizing controlled-size multimetallic clusters for spectroscopic theoretical and chemical investigations a systematic study of a wide range of metallic combinations needs to be undertaken.The basic requirements for pro- gress in this area are first fairly well understood parent cluster systems and secondly minimal spectral interference of atomic and cluster species in the two-component systems. In this particular discussion we will focus attention on the rather unusual ato- mic combination silver-chromium two transition metals which in the bulk exhibit * Subsequent studies employing Mg/M (where M = Cu Ag Au) codeposition with e.s.r. detec- tionssb and Fe/M (where M = Co Ni Cu) codeposition with Mossbauer monitorings8' speak highly of the tremendous potential that the matrix-isolation technique offers for studying the embryonic stages of bimetallic-cluster formation.See also the paper by Kasai and McLeod in this Symposium 200 300 400 500 wavelength Inm 200 300 400 500 600 wavelength Inrn FIG.ll.-(a) Optical spectrum of the products of an Ag/Kr = 1/104cocondensation reaction (A) after deposition at 10-12 K and (B) after 60-min narrow-band (8-nm) 325-nm continuous irradiation from an Oriel 500-W xenon lamp-Schoeffel mono- chromator assembly.30 (b)Optical spectrum of the products of an Ag/CH4 = 1/103cocondensation reaction (A) after deposi- tion at 10-12 K and (B) after 30-min narrow-band (8-nm) 332-nm continuous irradiati~n.~' w w SPECTROSCOPY CHEMISTRY AND CATALYSIS f.c.c. and b.c.c.crystal structures respectively which are of quite different atomic sizes and whose phase diagrams demonstrate very low miscibility proper tie^.^^" Let us first consider the products of a Cr/Ag/Ar z 1/1/104 depositi~n~*~~~ at 10-12 K [fig. 12(a) A] remembering that under comparable Cr/Ar z 1/104[fig. 12(a) B] and Ag/Ar z 1/104[fig. 12(a) C] conditions only the diatomic cluster stage is achieved. The most striking observation in the Cr/Ag/Ar high-dispersion system is the appear- ance of a new spectral feature centred at 283 nm. Other absorptions which can be associated with a mixed Cr,Ag cluster are not observed under these conditions. Ag Ag Cr Cr Cr Cr A 8 C. -..-. ... 1 I I 250 300 350 400 450 500 wavel en gt h/ nm 1 (b) Ag Ag Cr Cr Cr Cr wavelength /nm FIG.12.-(a) U.v.-visible spectra of (A) Cr/Ag/Ar z 1/1/104,(B) Cr/Ar z 1/104 and (C) Ag/Ar z 1/104mixtures deposited at 10-12 K showing the formation of Cr, Ag and AgCr.28*55 (6) U.v.-visible spectrum of Cr/Ag/Ar NN 1/1/104 mixture (A) upon deposition at 10-12 K showing Cr Ag Cr2 Ag and AgCr (B) after Cr atom 350-nm photoexcitation and 30-min relaxation time showing the growth of Cr and AgCr and (C) after Ag atom 305-nm photoexcitation showing the growth of Ag, Ag, AgCr and Ag2Cr.zs*s5 G.A. OZIN Clearly the 283 nm absorption can be given the a priori assignment of AgCr where other absorptions associated with the heteronuclear diatomic cluster are either (i) too weak to observe under these conditions (ii) obscured by band overlap with the intense Cr Ag atomic or Cr, Ag cluster absorptions or (iii) outside the range of our instrument capability (200-900 nm).Cr/Ag/Ar concentration experiments confirm that the 283 nm absorption is indeed associated with the first stage of Cr/Ag clustering that is AgCr. Only in experiments involving large amounts of material do we ob-serve another very weak absorption at 220 nm which might be associated with AgCr. Let us turn our attention now to the Ag/Cr bimetallic photoaggregation experi- ment~.~~ Fig. 12(b) A shows the Ag/Cr/Ar x 1/1/104 system after deposition at 10-12 K. As in fig. 12(a) one observes mainly Cr, Ag and AgCr on deposition. Narrow-band Cr (3d54s1-f 3d44s14p1)excitation causes a large decrease in the Cr atomic resonance lines with a concomitant but much smaller loss of the Ag atomic resonance lines.Under these conditions the Cr species grows in as in the pure Cr/ Cr,/Ar system,, along with a notable increase in the intensity of the AgCr (283 nm) absorption. Significantly the Ag absorptions remain essentially unchanged during Cr atom photoexcitation and the characteristic Ag absorptions at 245 and 440 nm do not grow in. All the evidence therefore points in favour of photoselective bimetallic aggregation upon Cr photomobilisation in the presence of Ag Ag, AgCr and Cr species. Focusing attention on sequential Ag atom photoexcitation as depicted in fig. -2 5P 27cu .... -. 4P 3n ... -4 -6 2 k\ t FIG.13.-Extended Huckel molecular orbital energy level schemes calculated for Ag, Crz and AgCr at their minimum-energy geometries of 2.7 1.7 and 2.9 A respecti~ely.~~*~~ SPECTROSCOPY CHEMISTRY AND CATALYSIS 12(b) C one observes a marked decrease in the Ag atomic resonance absorbances with concominant but much smaller loss of the Cr atomic resonance lines.Especi- ally noteworthy is the observation of growth of Ag and AgCr together with the ap- pearance of Ag and a striking new spectral feature at 276 nm as a shoulder on the AgCr absorption at 283 nm. The results once again emphatically point to the exist- ence of a phososelective bimetallic aggregation phenomenon. Moreover from a series of sequential Cr Ag atomic photoexcitations and Cr/Ag concentration ratio varia- tions one can conclude that the new absorption at 276 nm is most probably associated with the triatomic mixed-metal cluster Ag,Cr.An absorption unequivocally ascribable to Cr,Ag has not yet been observed in this system. Additional experimental support for the concept of photoselectivity in this bimetallic system and for the spec- tral assignment of AgCr and Ag,Cr stems from a series of experiments similar to those described in fig. 12(b) except that the aggregation is initiated by Ag rather than Cr photoexcitation. The photoselectivity proposals are reinforced by broad-band (250-600 nm) control experiment^.^^ One might expect that a molecule such as AgCr with metal atom components of quite disparate sizes and electronic type would display properties distinct from those of the parent homonuclear diatomic molecules.A qualitative insight into the problem has been obtained by direct transference of the optimised e.h.m.0. parameters derived for Ag and Cr, to AgCr. This leads to the molecular orbital scheme shown in fig. 13. (It is worth noting that this picture closely resembles the ground-state spin-restricted Xa description of AgCr.)59 In essence one finds that the 3d orbitals of Cr and 4d orbitals of Ag are non-interacting in AgCr. The Cr 4s and Ag 5s orbitals were found to be the main contributors to the silver-chromium bond. The minimum- energy geometry for AgCr was computed for both closed- and open-shell configura- tions at 2.9 A considerably elongated with respect to the multiply bonded Cr2,* and slightly longer than the singly bonded Ag,.From the e.h.m.0. energy-level scheme for AgCr three absorption bands can be expected within the 200-700 nm range of our instrument 20 -+ 4a,3n 50 (fig. 13). The 20 -f 40 transition is calculated around 415 nm in a region of overlap of atomic diatomic and triatomic features. The 20 -+ 3n transition is expected around 276 nm; we observe an AgCr absorption at 283 nm. The 20 -+ 50 transition calculated at 224 nm might possibly be associated with the weak AgCr absorption at 220 nm. Let us now make a brief excursion to some bimetallic work incorporating silver which is still in progress. We have seen that Cr/Mo and Cr/Ag cocondensations led to the bimetallic species C~MO~~ and AgCr? A natural extension involved a mixed Ag/Mo experiment.The results of a typical Ag/Mo/Ar z 1/1/104 cocon- densation reaction are shown in fig. 14(a). An absorption around 430 nm has grown in prominently and under the binuclear conditions chosen can be given the apriori assignment of AgMo. Details of the Ag/Mo concentration and photoaggregation studies can be found in ref. (62). * In the context of Cr2 we note that density functional theory with a local spin density approxima- tion for exchange-correlation energy brings forth the delicate balance between chemical bond forma- tion and the reduction in spin degeneracy which accompanies it.60 Near the centre of the 3d dimer series K2-CuZ the bonding contribution is at a maximum (all bonding orbitals occupied and all antibonding orbitals empty CrZ 'X;). However this implies a large spin energy as the infinitely separated atoms have high-spin ground states.In the case of the 'Z state of Cr2 the energy cost of six spin flips required to create this configuration has recently been calculated to make this state unbound and in fact the high-spin ''Z; configuration is favoured for the ground state.60 Thus the short bond length of 1.7 the multiply bonded closed-shell 'C; Xcr ground state6l* and 3E; e.h.m.0. ground state6'= for Crz must now be viewed with caution (cf. papers by Hillier et al. and Bursten and Cotton). G. A. OZIN :I ii \ r I s I C I I e 8 n 0 I 1 L 200 250 300 350 400 wavelength Inm FIG.14.4~) U.v.-visible absorption spectra of the matrix cocondensation products of (A) Ag/Ar = 1/104,(B) Mo/Ar = 1/104and (C) Ag/Mo/Ar = 1/1/104at 10-12 K showing the bands of Ag, Mo and AgMo in the presence of Ag and Mo atoms.62 (b)U.v.-visible spectra of the matrix co-condensation products of (A) Au/Ar = 1/5 x LO3 at 10-12 K (B) Au/Ar z 1/5 x 10' at 20 K and (C) Ag/Au/Ar z 1/1/5 x 10' at 20 K showing the presence of absorptions tentatively assigned to Au, Au3 and a suspected region of Au,Ag, bimetallic cluster formation.Two different (photo-interconvertible) matrix sites A and B of atomic gold are indicated in the figure.62 SPECTROSCOPY CHEMISTRY AND CATALYSIS We have seen from our cluster studies that both copper and silver aggregate upon deposition and photoexcitation into the 2S1,2 -+ 2P1/2,3/2 atomic resonance lines.With an atomic mass nearly double that of silver one would anticipate that the matrix diffusion characteristics of gold would be different from those of Cu and Ag. A typical result of cocondensation of gold atoms with argon (Au/Ar z 1/5 x lo3) at 10-12 K is displayed in fig. 14(b) A that is a simple three-line gold atom matrix spectrum. Only when employing higher concentrations (Au/Ar x 1/5 x lo2) and an enhanced deposition temperature of 15-20 K (the choice of temperature is crucial) can new absorptions be observed at 198,208 317 and 365 and 292 and 471 nm which have been tentatively associated with Au and Au gold clusters respectively [fig. 14(6) B]. An example of a preliminary Ag/Au/Ar cocondensation study is traced in fig. 14(b) C to indicate the kind of overlap problem that is being encountered in this system.Noteworthy here are the regions 260-295 and 325-340 nm which are presently suspected to contain new features associated with small bimetallic gold- silver clusters.28*62 A metal not yet discussed but highly interesting because of its known catalytically significant alloys with silver and other metals is palladium. The isolated palladium atom in a variety of inert-gas matrices has been a mystery for some time. Its spec- trum was first reported by Mann and Br~ida,~~ subsequently shown to be erroneous [in fact mistaken for Pd(N2)1,2,3] and rea~signed.~~ Further studies62 proved that the spectrum actually consisted of a doublet of triplets where one triplet would col- lapse on warm-up indicating that the switch-over was due to a matrix site change [fig.15(a)]. Atomic palladium has a ground-state electronic configuration of 4d1'5s0 with a likely Pd2 van der Waals dimer (Do z 22 10 kcal m01-l)~' absorb- ing at 265 and 283 nm in Pd/Ar and Pd/Kr x 1/102- matrices.62 From simple MO theory one might also expect Pd to form a weakly bonded dimer with silver (4d1O5s'). However neither Ag/Pd cocondensation experiments [fig. 15(6) B] nor photoexcita- tion of silver atoms in low-temperature Ag/Pd/Ar composite matrices [fig. 15(b) C)] has given any reason to indicate bimetallic cluster formation at the atomic level. Furthermore photoexcitation of the palladium ISo-+ 'P1, 3D1and 3P1atomic reso- nance transitions in Ag/Pd/Ar composite matrices although involving energies >40000 cm-' has not led to bleaching of the spectrum nor to any other sign of photoaggregation .62 The apparent ease with which certain atomic components participate in bimetallic cluster formation at the scale of just a few atoms especially for combinations such as Ag/Cr Ag/Mo Ag/Mn which from their bulk-phase diagrams comprise a class of immiscible alloys brings to the forefront the question of whether any two metal atomic components can form heteronuclear cluster molecules.In a practical sense these studies are of considerable significance in view of the work of Sinfelt56*57*64 and otherP with high-dispersion multicomponent particle catalysts (10-1000 A) com-posed of totally or partially miscible metal constituents in the bulk Sup-ported and unsupported catalysts of this type often display remarkably distinct chemical properties from those of the individual metals and led Sinfelt and others to view them as " bimetallic " rather than " alloy " cluster catalysts.In this context the term " bimetallic " implies an electronic interactive mode! rather than one com- prising a simple addition or superposition of the properties of the individual atom/ cluster components. The results of bimetallic cluster growth in a region of two or three metal atomic components for example CrMo CrMo, Cr2Mo,54 AgCr Ag2Cr,55 NbMq6 AgMo and AgMn,62 convey the image that these metal molecules are " normal " in an elec- tronic optical bonding and probably a chemical sense. The most significant aspect *... . .. .. .. .. .. .. .:: :'. . /-%; ...... ...... :: .: . -. :: . . . . . Q .... V C e Y, P 0 1 I I L I 1 200 250 300 wavelengt hlnm 0 C d n L 0 C B A FIG.15.40) U.v.-visible absorption spectra of the cocondensation of palladium atoms with krypton (A) after deposition at 10-12 K (B) after warmup to z17 K and (C)final structure after warmup to 20 K. Two different matrix sites A and B of atomic palladium are indicated with arrows.62 (6) U.v.-visible absorption spectra of the matrix cocondensation products at 10-12 K of (A) Pd/Ar = 1/104,(B) Ag/Pd/Ar z 1/1/104and (C) after subsequent prolonged irradiation in the 300-nm Ag atomic resonance line.62 SPECTROSCOPY CHEMISTRY AND CATALYSIS of these studies is probably that such bimetallic cluster molecules can actually be fabricated.Recall that the respective bulk alloy systems are of the immiscible or partially miscible and hence the small cluster results can be considered to lend further credence to Sinfelt’s proposal of a direct interaction between metallic com- ponents in the respective multicomponent particle systems. In contrast one must be alert to the fact that straightforward mixing of metal atomic components under matrix conditions does not necessitate bimetallic cluster growth at the few-atom level of operation as witnessed by the optical spectra of Ag/Pd Ag/Pt Pd/Mo and A~/Ru,~~*~~~~~ which display just the additive features of the atomic and diatomic components.Clearly a more thorough appreciation of metal atom/cluster nucleation phenomena under low-temperature conditions requires a detailed understanding of the electronic configurational alterations and kinetic bar- riers which make up the complex sequence of events leading up to the nucleation and growth of bimetallic molecules. Research is continuing along these lines using other spectroscopic probes. Extensions to polymer-supported unimetallic bimetallic and trimetallic cluster composition are already under development 66 and the stage is now set for an excursion to the selective chemical reactions of small M,Md cluster molecules with reactive ligands. The latter investigations are likely to generate valuable data on (i) preferred sites of ligand coordination (ii) alterations in cluster and ligand pro- perties on complexation and (iii) relationships to and differences from the situation of the ligand chemisorbed on bulk bimetallic particule or alloy surfaces.Studies of this type at the atomic scale of operation might illuminate present concepts of “ elec-tronic and geometrical ” factors in ~atalysis.~’ AN EXPERIMENTAL APPROACH TO METAL CLUSTER ANIONS Kasai 68 has demonstrated that the rare-gas matrix isolation e.s.r. technique is not only applicable to electrically neutral species. In a series of experiments a variety of charged species was generated and trapped in a facile manner by a process con- sisting of photoinduced electron transfer between suitable well-isolated electron- donating and electron-accepting species.68 Of particular interest to the present dis- cussion is a series of experiments using certain metals Na Cd Cr and Mn,68 as electron donors and HI as electron acceptor where the net reaction is envisaged to be M + HI~M+ + H.+ I-. The e.s.r. spectra showed that the starting metal atoms and the resulting singly ionized metal atoms can be considered isolated within the matrix. From Kasai’s work68 one also learns that electron transfer between metal atoms and HI involving dissociative electron capture of HI can be put to use in generating molecular anions. Early feasibility experiments used Na atoms as electron donors and tetracyanoethylene (TCNE) B2H6,and furan molecules as the acceptors.68 In this way e.s.r.evidence was obtained for TCNE- B2H6-and c\ the latter a-anion radical resulting 0 0 from the spontaneous opening of furan anions. The anion work of Kasai opens up the possibility of photogenerating metal anions M- and metal-cluster anions M; by Na/M codeposition and Na photoionisation techniques a particularly important area of research in view of the suspected involvement of very small silver cluster anions G. A. OZIN in the development of the photographic latent image,69 silver(r)/zeolite photocatalytic water splitting 70 and particle catalysis involving electron transfer interactions from the ligand to the metal upon chemisorption. 71 Very recently some macroscale bimetal atom cryochemical reactions 72 involving Group IA/IB ammonia cocondensations have pointed the way to solvated transition metal anions of the type M-.Thus in sharp contrast to Zmbova'~~~ Li/Au/Ar 10-12 K matrix reactions that led to the i.r. detection of molecular 6*7LiAu Lagowski and Peer 72 discovered that M/Au/NH cocondensation at 77 K (M = Li K Rb Cs) result in solutions at -65 "C which display gradual loss of the 1840 nm solvated electron absorption with concomitant growth of an intense U.V. absorption in the range 277-289 nm. The experimental evidence encouraged Lagowski and Peer to promote the notion of a W06s2Au- solvated gold anion." Recently Timms and co- worker~~~ have employed Na and K atom reactions for generating active metals for organo metallic synthesis THF/ -130°C e.g.CrC13.3THF+ K + naphthalene -+ (naphth),- Cr. ~ In this context a number of groups have recently studied the matrix aggregation properties of atomic So far there seems to be fairly close agreement on the optical assignments of Nal,,, [fig. 16(b) B-D]. In view of our earlier discussions of Ag atom photoaggregation properties it was not too surprising to discover that 2S-f 2P excitation of the blue or red triplet absorption of Na/inert gas mixtures res- sulted in facile growth of Na1,2,3.76a*77 However in the sodium system in addition to light-induced diffusion of Na atoms there appears to be a competing non-radiative channel involving photoionisation. 76a*77 This pathway seems particularly pronounced when Nan electron traps coexist in the matrix together with Na atoms presumably leading to Nay cluster anions (as yet undetected in the 200-800 nm range).77 When the experiment is conducted under high-dispersion conditions in the presence of silver atoms such that species higher than diatomics are in very low abundance [fig.16(a) A] one can detect on deposition Ag,, and Na,, together with a very weak new feature around 370 nm which is tentatively associated with molecular NaAg 77 (seen previously by high-temperature mass spectroscopy). 78 Remembering that photoexcitation at 337 nm is known to induce photoaggregation of Ag atoms and photoionisation of Na atoms in separate matrix experiments one finds that in a Na/Ag/Xe composite matrix aside from the growth of some Ag and Ag and photoionisation of Na atoms [fig.16(a) B] one also notices a concurrent photoprocess involving the creation of a totally new species displaying a doublet absorption at 220 and 213 nm which cannot be associated with any known Ag,,,,,, or Nal,,,,, optical feature. Also of signifi- cance here is that the 4dl05s2-+ 4d105s15p1 ('S -+ 'P)excitation of atomic Cd occurs at 229 nm., This and other evidence points towards an assignment in favour of the silver anion Ag- isoelectronic with and optically resembling atomic Cd. Modifica-tion of the concentration matrix support and ionization conditions could permit the method to be extended to the generation of naked silver cluster anions Age,,,. Work is continuing along these lines. In concluding this section it is pertinent to emphasise a remarkable resemblance between the optical spectra of Na1,2,3,4 and Ag,,,,,, in for example Kr matrices,76a not in the sense of absolute transition energies but rather in terms of the spectral distribution and intensities of the M2,3,4 cluster absorptions straddling the 2S,,2-f 2P1/2,3/2 parent atomic resonance lines [illustrated in fig.16(b)]. We believe that this is not a purely serendipitous event but instead a prerequisite of the ns' isovalence of * Related work involves the stabilization of alkali metal anions by complexation with crown ethers and crypt and^.^^ n I I I I I 1-I 300 LOO 500 600 700 300 LOO 500 600 700 wavelength/nm wavelength I nm FIG.16.-(a) Matrix u.v.-visible spectra of (A) Na/Ag/Xe x 1/1/1000mixture deposited at 10-12 K and (B) after 5-min 337-nm irradiation showing the decay of Nal.2 and Ag with growth of a little Ag and Ag3 and a new species labelled " a " which is tenta- tively ascribed to Ag- (see text).In the figure 1' Na atomic blue site 1" = Na atomic red site 2' =Na2 1 = Ag 2 = Ag2,3 = Ag3,b = NaAg.77 (6)(A) Matrix optical spectra of A&,2,3,4/Kr compared with (B)-(D) the matrix optical spectra of Na,,2,3,4/Kr recorded at 10-12 K the latter at decreasing metal concentrations re~pectively.~~" G. A OZIN Na and Ag the predominance of ns orbital contributions to the metal-metal bonding interactions the uninvolvement of low-lying Ag 4d orbitals in the silver-silver bonding the absence of Na 3d character in the sodium-sodium bond and the mainly 3s -+ 3p and 5s -f 5p localised character of the visibk absorptions of Na1,2,3,4 and ultraviolet- visible absorptions of Ag1,2,3,4 respectively.The bZue shifting of the Ag,,,,,, excita- tion energies with respect to Na1,2,3,4 is presumably an electronic manifestation of the more favourable 5s/% overlap properties and the larger 5s/5p energy separations for atomic Ag compared with the 3s/3p of atomic Na. SILVER ATOMS AND SILVER CLUSTERS IN HIGH-TEMPERATURE MATRIX SUPPORTS In the field of metal atom-metal cluster chemistry there has emerged an urgent need to establish a range of high-temperature matrix media for use in supported catalyst preparati~n.~~ The ultimate objective here is an evaluation of the catalytic properties of ultra-finely dispersed narrow-size distribution particle catalyst systems.Tmmo-bilisation on oxide zeolite polymer-type supports is envisaged for these very small cluster arrays and unique activity/selectivity patterns are anticipated. The usefulness of a support material for the envisioned preparative applications will be determined by (a) the reactivity of the support towards the metal vapour on deposition and during any subsequent annealing and photolysis treatment (b) the nature of the metal atom and metal cluster support interactions as these will inevitably influence both the degree to which metal atom aggregation processes occur and the extent to which the spectroscopic properties of the atoms and clusters are perturbed and hence disguised in the matrix (c) the thermal stability of metal atomic/small cluster dispersions indicated by the temperature at which metal atom diffusion- aggregation processes become noticeable and (d) the practicability of a macro-scale preparation and the availability of a convenient means of extracting and dispersing the metal clusters on a support material suitable for catalyst testing.With the wealth of knowledge for Ag,,2,3, in a variety of weakly interacting low- temperature supports one is in a strong position to assess the outcome of silver-atom depositions into selected quench-condensed films of high-molecular-weight paraffin and olefin waxes and ice.s0 (A) ICE MATRICES A representative optical spectrum obtained when silver vapour is cocondensed with H20at 10-12 K under high dilution conditions (Ag/H20 z 1/10,) is shown in fig.17(a) C. Two well-resolved band maxima (336 and 316 nm) a shoulder (301 nm) and a broad red-shifted absorption (380 nm) labelled B’ C’,D’ and A’ respectively are reproducibly observed on deposition. Fig. 17(a) B and (b) B show typical spectra obtained upon increasing the Ag/H20 ratio to zl/103 and 1/102 respectively and fig. 17(c) shows a series of deposition/warm-up spectra obtained for an Ag/H20 ratio of z 1/102. Noticeable especially in the spectra shown in fig. 17(c) is the appearance of absorption bands not associated with silver atoms. These bands are attributed to silver clusters; the assignments have been made by comparison with the known optical spectra of Ag1-4 in a variety of matrix support rnaterial~.~~*~’~~~*~’ The absorption spectra of Agl-,/CH4 mixtures (10-12 K) are shown in fig.17(a) A and (b) A for the purpose of comparison with the Ag,-,/H20 spectra.80 The effect of warming a matrix very dilute in silver (Ag/H20 z 1/10’) is illustrated in fig. 18(a). Note that the bands labelled A’-D’ behave very differently as the matrix 1 01 E 2 ln n I I 200 300 LOO 500 200 300 400 500 200 300 LOO 500 wavelength Inm wavelength I nm wavelength Inm FIG.17.-(a) Optical spectra obtained on depositing silver vapour with water vapour or methane at 10-12K (A) silver vapour with methane at Ag/CH4 z 1/102,(B) Ag/H20 z 1/103and (C) Ag/H20 z l/104. Throughout this study A’ B’ C’ D’ refer to slightly different Ag atom trapping sites in vapour-condensed ice films and 2 = Ag, 3 = Ag, 4 = Ag4.80 (6) Optical spectra obtained on depositing (A) silver vapour with methane at 10-12 K at Ag/CH4 z 1/102and (B) silver vapour with water vapour at 10-12 K at Ag/H20 “N 1/10’ brief warming to 77 K and recooling to 10-12K for spectral recording.*’ (c) Optical spectra obtained on depositing silver vapour with water vapour (A) at 10-12 K and Ag/H20 z 1/102and (B)-(F) showing the progress of thermal annealing at 77,130 170 170 and 170 K respectively.80 G.A.OZIN I I I J 200 300 400 500 00 300 400 500 wavelength/nm wavelength /n m FIG.18.-(a) Optical spectra obtained on depositing silver vapour with water vapour (A) at 10-12 K and Ag/H20 z 1/105and (B)-(G) showing the progress of thermal annealing at 60 100 125 150 170 K respectively.*’ (b) Optical spectra obtained on depositing silver vapour with water vapour (A) at 10-12 K and Ag/H20 = 1/104and (B)-(G) showing the progress of a series of 30-s narrow-band irradiations (500 W xenon lamp-monochromator assembly with 10-cm water filter) at 400 400 400 360 340 and 300 nm respectively.*’ is warmed and that these changes are irreversible suggesting the existence of at least four distinct trapping sites for silver atoms in vapour-condensed ice films.Moreover the thermal stability of the trapping sites is seen to decrease in the sequence D’ > C’ B’ > A’. As shown in fig. 18(b) silver-atom site interconversions in vapour-con- densed ice matrices can also be induced photolytically.Thus site A’ can be photo- lysed to disappearance with concomitant growth of the thermally stable site D‘ [fig. 18(b) A-D]. Irradiation at 360 nm has the effect of partially regenerating site A’ while depleting sites B’ and C’ [fig. 18(b) D and El. Photolysis centred near 300 nm corresponding to site D’ apparently results in silver-atom photomobilisation similar to that observed for CH4 and rare-gas rnatrice~.~~*~~*~~ SPECTROSCOPY CHEMISTRY AND CATALYSIS The occurrence of multiple trapping sites of Ag atoms isolated in ice supports has been demonstrated by means of detailed e.s.r. studies of radiation-produced Ag atoms in AgNO ices." The overall picture that emerges from these e.s.r. studies is one of magnetically distinct multiple trapping sites that are interconvertible by thermal and optical excitation with little loss of Agospins up to 77 K.The evidence of the present study indicates that the silver atomic trapping sites are quite different de- pending on the method of sample preparation i.e. vapour condensation or reduction of silver ions in pre-frozen aqueous solutions. They thus differ from the sites sug-gested by Kevan et aZ.'lU Thus when Ag atoms are condensed with H20 at 77 K the observed e.s.r. parameters are very similar to those of gas-phase Ag atomss2 but radiation-produced silver atoms trapped in frozen (77 K) aqueous solutions display e.s.r. parameters markedly different from the gas-phase values." Similarly the optical absorption spectra obtained for Ag atoms isoiated in vapour-condensed ice films at 77 K (the present study)80 are quite different from those obtained for radiation- produced silver atoms in 77 K aqueous solids.s3 In the latter case there is a complete lack of absorption bands near 300 nm where the most intense features were observed in the present study.We note here that 12 K vapour-condensed ice films are likely to be amorphous,s4 while frozen aqueous solutions of the type used for generhting the radiation-produced silver atoms are probably more ordered in structure. In any case the ordering influence of the electric field of Ag"ions is entirely absent for the vapour-condensed samples and this influence has been shown by e.s.r. studiess1 to be of dominant importance in determining the structures of the atomic silver trapping sites.In view of the complex phase behaviour of ice therefore and the ordering influence of the electric field of Ag",it is not surprising that the spectroscopic results indicate very different silver atomic trapping sites for the different methods of sample preparation. Notwithstanding this general conclusion it is conceivable that at least one of the A'-D' trapping sites observed in the present study does in fact correspond with one of the A-D trapping sites observed in the e.s.r. spectra.s1 Concurrent e.s.r. and optical absorption studies might provide a clarification of this point The results described above allow for an assessment of the isolation properties of vapour-condensed ice films. At low concentrations (Ag/H20 z 1/10') isolation of atomic silver predominates and the presence of a number of distinct trapping sites of widely varying thermal stabilities can be recognised.At temperatures above z125 K diffusion-aggregation processes become important probably as a result of the occurrence of exothermic phase changes in the ice In the context of the present study it is encouraging that a metal atomic dispersion can be stabilised in a weakly interacting support material at the relatively high temperature of 125 K [fig. lS(a)]. As illustrated in fig. 17(c) silver clusters in the size range Ag2-Ag can be isolated at higher Ag/H20 concentrations (z1/102). These silver atomic/cluster dispersions are stable to x130 K ; at higher temperatures diffusion-aggregation processes seem to lead to silver micro~rystallites.~~~~~ Further studies relating to the size distribution and spectroscopic properties of thzse ice-entrapped silver cluster species would be worthwhile as would methods for extracting these small silver clusters onto conventional room-temperature supports for catalyst testing.(B) PARAFFIN WAX MATRICES Methane and the short-chain alkanes have been explored as " weakly interacting " matrix media in metal atom-metal cluster experiments as alternatives to the noble gases.87 The initial objectives in the investigation of silver-paraffin codepositions were to establish the conditions under which silver atoms could be isolated and to G. A. OZIN determine the temperature range over which diffusion-aggregation events become pro- nounced.Samples of n-C22H46 and n-C&66 were chosen as representative of high molecular-weight paraffins. A four-electrode furnace was employed with dual quartz-crystal mass monitors for controlling the silver and wax vapour deposition rates from tantalum and stainless stell Knudssn cells. respectively. In order effi- ciently to isolate atomic silver one is obliged to employ extremely slow depositions of the wax (at least an order of magnitude less than with for example methane) with the appropriate proportion of silver vapour to maintain the 0.1 to 0.01% metal concentra- tion netessary for efficient atomic isolation. Only under these extremely low-flow high-dispersion conditions was it possible to establish the atomic spectrum of silver in for example the n-C22H46 wax (fig.19 A). 300 400 500 wavelength /nm FIG.19.-Optical spectra obtained on depositing silver vapour with n-C22H4,j vapour at 10-12 K and (A) Ag/n-C2,H4 x 1/104 (B) Ag/n-CZ2Ha6 x 1/103 and (C)-(E) showing the progress of annealing (B) at 30 40-and 80K respectively.s0 The divergence in the isolation properties of paraffin waxes from the trends noted for the shorter-chain alkanes (increasing isolation efficiency with increasing chain length)s7 was found also in the observed thermal stabilities of the Ag/wax dispersions. Thus while the short-chain alkanes appeared to follow the reasonably successful Tammann predictive diffusion temperature of 0.3 TmaP., bulk diffusion of silver atoms SPECTROSCOPY CHEMISTRY AND CATALYSIS in the paraffip wax supports was observed at temperatures as low as 30 K.This observation is not surprising in view of the likelihood that vapour-condensed wax films are highly disordered or glassy with the very open structures allowing for diffu- sion at relatively low temperatures.88 TABLE 5.-oPTICAL SPECTRA OF Ag ATOMS AND Ag2 DIMERS IN METHANE AND HIGH MOLECULAR- WEIGHT PARAFFIN WAX (n-C22H46 n-C32H66) MATRICES (nm) [REF. (SO)] CH4 (10-12 K) n-Cz2H4 (10-12 K) n-C32H66 (10-12 K) assignment 380 333 390 357 390 360 Ag2 Ag 320 334 338 Ag 310 312 315 Ag 278 432 Ag2 The ability to trap and retain Ag/Ag,/wax mixtures on deposition at 10-12 K is shown in fig. 19 which illustrates part of a typical Ag/n-C2&46 concentration study (fig.19 A and B). Silver atom and silver dimer lines are quite easily identified and correlated by only slight frequency shifts with the values found for Ag/Ag2 in solid CH4 (table 5). These data follow the .trend with increasing alkane chain length found previously for V/V2/CnH2n+2 for n = l-10s7(i.e. red-shifting towards the gas phase transition energies with increasing chain length). Representative portions of Ag/Ag2/n-C22H46 warm-up studies are depicted in fig. 19 C-E showing marked Ag atom diffusion at temperatures near 30 K with growth of silver microcry~tallite~~~~~~ becoming apparent around 40-80 K (fig. 19 D-E). The results of the present study indicate that very small naked metal clusters cannot be stabilised in high molecular-weight paraffin wax supports close to room tempera- ture.Instead pronounced diffusion-agglomeration processes dominate at very low temperatures although it is likely that the metal aggregates so formed are still ex- tremely small and therefore of interest to those concerned with the catalytic properties of metal clusters and metal colloids. Such extraneous agglomeration effects can however be avoided by trapping of atoms and clusters in certain liquid polymers.66 (c) OLEFIN MATRICES It is interesting to consider olefinic matrices as media for generating weakly com- plexed metal atomic and small cluster species (solvated in Klabunde's terminol~gy)~~ which can subsequently be manipulated to yield narrow cluster distributions possibly for use in fabrication of high-dispersion metal particulate catalysts.For example it is known that silver atoms react with ethylene at 10-12 K to yield a purple Ag(C2H4) n-olefinic complex90 characterised by i.r. 12C2H4/13C2H4/12C2D4 isotopic substi- tution,15 e.s.r. 91 and intense mainly metal-localised s -+ p and d -+ s visible and U.V. absorptions around 580 and 300 nm re~pectively'~~-~~ [fig. 20(a) B]. These optical assignments are based on a series of spin-unrestricted transition state calcula- tions for M(C2H4) 92 complexes [fig. 20(b)] which satisfactorily reproduce the optical trends on passing from Cu to Ag to Au [fig. 20(a)]. In the context of employing these weakly bonded systems for cluster fabrication one notes the stability trend93 / Ag-11 < Ag-11 < Ag-11 the former decomposing around 40 K and the latter around 80 K.93 This leads to the idea of using high-molecular-weight alkenes as 0 .o -0.2 6a1 -.-q-I -0.4 -& 5al -0.6 -0.8 z1 a \ L' -1.0 -1.2 'bz 'b? -1 .L -1.6 200 300 400 500 600 wavelength Inm -1.8 FIG.20.-(a) Optical spectra of matrix-entrapped (A) CU(C~H~)~,~,~ in Ar (B) Ag(C2H4)in C2H4 and (C)Au(C2H4)in C2H4 at 10-12 K where dotted lines indicate correlations between corresponding " low-energy " 60 -f 3b2 and " high-energy " 50 -f 6a excitations of M(C2H4)where M = Cu Ag Au [see text fig.20(6) and ref. (1 76) (90) and (92)]. (6) SCF-Xcc-SW molecular orbital energy level scheme for M(C2H4)where M = Cu Ag Au [see text and ref.(92) for details]. SPECTROSCOPY CHEMISTRY AND CATALYSIS wavelength Inm 250 300 350 LOO 450 500 550 0 V c n L 0 n 0 J I I I I i 200 300 LOO 500 600 700 wavelength Inm 300 400 500 wavelengthInm FIG.21.-(a) Optical spectra of the products formed from Ag/C2H4 = ]/lo3mixtures (A) codeposited at 10-12K showing mainly Ag(C2H4) (B)-(E) the results of warming the matrix in (A) to 40 60,70 and 80 K showing the gradual decomposition of Ag(C2H4) to Agl,z,3. possibly some Ag,(CZH4) and eventually higher molecular clusters (6 5 ri 13) (F) the result of having sublimed off the C2H4 showing an optical spectrum typical of a silver microcrystallite residue on the optical window [see text and ref.(93)]. (b)Matrix u.v.-visible spectra (A) Ag/Ar = 1/10' deposited at 10-12K showing isolated Ag atoms; (B) Ag/Ar 2 ]/lo3deposited at 10-12 K and photolysed at 315 nm for 60 min showing clustering into the range ii z 5; (C) same as (B) but with 10-min 20 K photolysis at 315 nm showing the loss of most of the Ag3 absorption with a little more clustering; (D) Ag/Ar z 1/300 deposited at 10-12K and warmed to 40 K showing still further clustering to at least n z 6; (E) Ag/ Ar z 1/50 deposited at 10-12K; (F) same as (D) but after warm-up to 40 K showing extensive aggre- gation approaching that of (G)Ag/Ar z 1 /20 deposited at 10-12K in which Ag clusters approaching colloidal dimensions (~10 A) seem to be present as witnessed by the broad plasmon resonance absorption centred at roughly 347 nm ved-shifting to around 370 nm on annealing to 100-150K.20 G.A. OZIN weakly complexing/aggregation media where a material such as dec- 1 -ene or higher may be regarded as a " hybrid matrix support " in which one essentially has an " al-kene trap" in an " alkane-like" matrix. Research along these lines is presently underway.93 Meanwhile the methodology can be exemplified by reference to the highly labile Ag(C,H,) system which at 40 K begins to fragment to ethylene and silver atoms aggregating to Ag,, around 70 K \fig. 2l(~z)].~~ These clusters nucleate further at 80 K (at which point the ethylene matrix is quite mobile) and a broad bulk- silver-like resonance appears around 350 nm which most probably originates from an envelope of mainly s +p " interband "-like transitions but in " molecular clusters " (see later) with nuclearities in the range roughly 6-13.93 Clearly one would envisage that a more rigid alkene matrix could promote this process in a more con- trolled manner and thereby effect isolation of a narrower distribution of the smaller more well-defined clusters in the range say two to six hopefully at higher tempera- tures.Incidentally further warming of these very small silver-cluster-ethylene ma-trices from 80 to 273 K induces more extensive agglomeration to a dimension which produces an optical spectrum reminiscent of bulk silver micro crystal^.^^,^^ These spectra display the characteristic high transmission edge around 320 nm and broad plasmon absorption in the U.V.around 400 nm (see later and the paper by Schulze in this Symposium). Similar growth of very small copper clusters has been realised by controlled thermal decomposition of Cu(C,H,). and Cu,(C,H,) in ethylene supports in the range 50-80 K.17' VARIATION OF THE ELECTRONIC PROPERTIES OF SILVER CLUSTERS WITH CLUSTER SIZE; THE CLUSTER-BULK-METAL PROBLEM The controlled clustering of silver atoms in a wide variety of matrices described in the earlier sections leads us to the general question of the fundamental interrela- tionship between metal cluster size and molecular and bulk particle properties as seen through the eye of their e.s.r. and optical spectra. A number of groups have recently directed their attention to the optical absorp- tion spectra of very small silver clusters generated by matrix cryochemical deposition and photoaggregation method^,^'*^^* 76c using the appearance of a surface plasmon absorption as one criterion for establishing the onset of bulk silver microcrystalline properties from that of molecular silver cluster behaviour.In the physics terminology of metallic crystals the optical absorption of light from the visible to the i.r. region can be discussed within the framework of the free electron gas Microcrystals of roughly 10-100 A on the other hand show size- and shape-dependent plasmon absorptions peaking in the U.V. or visible and dropping off with increasing wavelength associated with collective electronic excitation in the microcrystal. Note that in this cluster size regime the free electron model remains valid.Below 10 A mean diameter however quantum size effects remove the con- tinuous nature of the conduction band and a molecular-like model with discrete energy levels is a more appropriate description of the optical properties. In this situation continuous acceleration of the valence electrons by an external electric field is not possible; instead energy changes can only occur through transitions between quantized eigenstates. Broadening of these discrete electron eigenstates can also occur by lifetime limitation^^^**^ and may become more pronounced with diminishing particle dimensions. Thus quantum size effects in this regime are expected to be " washed out " when the energy level broadening overcomes the mean SPECTROSCOPY CHEMISTRY AND CATALYSIS energy spacing between the levels.* The method of preparation temperature and structure of the aggregates are likely to influence this critical dimension.Above this particle size the quasicontinuous nature of the conduction band becomes apparent and the electron mean free path assumes a central role becoming size-dependent in small particles by surface scattering and affecting electronic properties differently than in the quantum size region. It should be feasible to use these size-dependent phenomena as a criterion for assessing a discrete eigenstate or conduction band des- cription for a particular dispersion of metallic aggregate^.^^'^^ This " optical electronic image " of the conversion from bulk-like micro-crystals to molecular metal clusters has recently been observed by a number of groups studying controlled silver-atom-inert-gas matrix recombination reactions.20*47*76c Although the nuclearity determination above six silver atoms is not firmly established a rela- tively smooth conversion from a one-electron molecular eigenstate cluster picture to that of conduction-band bulk-like character is emerging from the analysis of the very early stages of silver cluster growth.Fig. 21(b) shows a collection of optical data20 (similar to those recently obtained by Schulze this Symposium) the general appear- ance of which can be seen to change from that of simple well-defined atomic silver through varying degrees of molecular spectral complexity with band-broadening and band-overlap effects to a relatively simple bulk-like picture where a single broad structureless absorption dominates the spectrum at roughly 347 nm.Based on these silver concentration-dependent spectral alterations one can see that the disappearance of molecular cluster absorptions and the onset of bulk-like plasmon absorption in metal microcrystals occur around an aggregate composition of roughly six silver atoms. It would appear that the collective electron excitation appearing around 347 nm cannot be described by a one-electron approximation applicable to silver clusters with n 7 6-13 atoms. Instead a continuum theory of dielectric response seems more appropriate to explain the origin and properties of the plasmon ab~orption.~~,~~ It is interesting to note that the strong dipole transitions of very small molecular silver clusters (n 76) roughly coincide with the plasmon fre- quency of silver microcrystals.In this context Kreibig and co-workersg6 have recently derived a dielectric func- tion appropriate for very small particles using a simple quantum-mechanical electron- in-a-box formulation. The model indicates that quantum size effects should be manifested for ultrafine particles below 40 A in the form of shifting and broadening of the plasmon resonance absorption which below roughly 15-20 A is expected to develop fine structure originating from one-electron excitation between discrete energy levels of the conduction band. Photosensitive glasses have proved to be particularly well suited rigid media for generating rather narrow size distribution almost spherical silver particle disper- sion~.~~*~~ Optical absorption spectra of the very early stages of silver particle nucleation (10-1008,) in these glasses have been monit~red~~~~~ and show a number of resemblances to the later (n > 6) aggregation stages of silver cluster growth from silver vapour depositions into various low-temperature matrices (inert gases alkanes ice and In particular the glass matrices of Kreibig and coworkers,86 for particles smaller than ~25 8 diameter display besides the plasmon resonance absorption around 400 nm a complicated absorption structure around 322 nm which is thermally unstable to tempering and shifts to lower energies and decays to zero at particle diameters above ~23 A.This behaviour seems reminiscent of very small * A statistical smearing of quantum size effects can be expected for a collection of varying size and shape particles as the energy spacings will not be some constant value but rather a random variable controlled by some probability distribution. G. A. OZIN molecular-like silver aggregates of the kind seen around 300 nm in for example quench-condensed " glassy " ice supports,8o which almost certainly represent a com-posite of lower nuclearity clusters than those giving rise to the plasmon resonance absorption at 400 nm. In the size regime below 25 A it is thoughts6 that the fine structure superimposed on the plasmon resonance absorption of silver microcrystals in glass matrices which is not unlike the features seen on the similar absorption around 370 nm in for example inert gas mat rice^,^^*^^,^^^ is connected with finite level spacing in the conduction band possibly a direct observation of a quantum size effect.95 Along with the optical absorption measurements of growing silver clusters from atom to bulk we have also conducted a series of e.s.r.measurements in a designed attempt to collect complementary structural and electronic information for the para- magnetic species present in the silver cluster Simultaneous monitoring of the optical and e.s.r. spectra from the AgJmatrix sample deposited onto a sapphire rod has recently been achieved in our lab~ratory,~~ and our early experiments have indicated that the onset of collective electronic excitations in silver clusters around a size of about six atoms is accompanied in the e.s.r.spectrum by the appearance of what appears to be best described as conduction electron spin resonanace 96-99 of silver micro- crystals (fig. 22). At high silver concentrations or after extensive annealing or photo- aggregation the matrix sample showed only weak lines corresponding to the hyper- H/ G 050 3300 355 (01 SPECTROSCOPY CHEMISTRY AND CATALYSIS 0.025 1; 68 20 -\ J-ShItt a \ \ 10 L d/i 'i 0 *t I Ag/inert gas (xlO-' 1 FIG.22.4~) Matrix e.s.r. spectra of Ag/Ar depositions at 10-12 K in the range 1/103-1/102 after photoaggregation and annealing pretreatment showing what is thought to be a composite picture of sharp lines associated with a range of aggregates with molecular properties superimposed on a broader line ascribed to c.e.s.r.of small silver microcrystallites [see text and ref. (36)]. (6)Collected data for the c.e.s.r. linewidths and g-shifts for Ag/inert gas z l/5 x 103-1/2 x lo2mixtures which had been deposited at 10-12 K annealed to 40-60 K and then recooled to 10-12 K for spectral re- cording purposes. Argon results shown in open circles and squares krypton results shown in solid circles and squares.36 fine spectra of isolated 107Ag/109Ag atoms (nuclear spin I = 1/2 in both cases) as well as a resonance around g = 2 whose intensity and associated structure depended on the matrix preparation and pretreatment mentioned above.According to earlier e.s.r. studies of small metal particles,96 it seems likely that this resonance is a composite of two contributions (i) sharp lines associated with a range of aggregates with mole- cular rather than metallic properties (4-6 suggested optically) and (ii) a broader line ascribable to small metal microcrystallites on which the sharp lines of (i) are super- i rnpo sed. The temperature dependence of these e.s.r. spectra indicated that atomic features tended to decay to zero in Ar around 30-35 K while the resonances associated with molecular clusters disappear around 40-45 K leaving behind the broader small- particle resonance whose linewidth and g-value varied between roughly 15 and 35 G and 2.013 and 2.028 respectively and whose intensity increased during the warming process.This type of behaviour seems to be mirroring the metal atom to cluster to bulk transformations observed in the corresponding optical spectra [fig. 21(b)] during silver nucleation. As mentioned earlier for small metal crystals the boundary condi- tions imposed on the conduction electrons cause the conduction energy levels to be- come discrete.95 K~bo,~~ Gorkov and Eliashberg 98 and KawabataY9 have principally developed the electronic theory for such metal microcrystallites. In particular Ka~abata~~ predicted that the quantum size effect could under certain conditions be observed in the conduction electron spin resonance (c.e.s.r.) spectrum which should differ markedly from the bulk. Particle size dependences of the c.e.s.r.line-width and shifts have been proposed and should be observable for crystals smaller than z30 A. G. A OZIN 55 In brief as the particle size decreases and the electron levels of the conduction band become discrete rather than quasicontinuous the average spacing of 6 between adjacent levels will eventually dominate the Zeeman energy haz. To understand the accompanying changes in the c.e.s.r. linewidth and g-shift one must enquire into the mechanism of the relaxation of conduction electron spins. Following Smithard,96" the spin relaxation time z2 of the bulk metal (originally discussed by Elliot100) can be related approximately to the resistivity scattering time zR by l/z C/ZR (2) ZZZ where C depends on the spin-orbit coupling contribution of the g-shift of the bulk metal (Le.C = AgSo2 and Ag, x A/AE where A is the spin-orbit coupling constant and AE is an appropriate energy band gap in the metal). However surface scattering effects as well as resistivity collisions must be considered for the total electron scat- tering time in small particles and following Ka~abata,~~ the spin-relaxation time becomes where V is the Fermi velocity and d the particle diameter assuming C the spin relaxation probability per scattering collision is the same for both processes. (Classic-ally speaking when an electron is scattered at the surface of a small particle transla- tional invariance is broken and a spin-flip occurs the probability of which depends on the spin-orbit coupling strength.) For particles sufficiently small to satisfy both the condition of 6 > hmz and z69 k Kawabatag9 showed that the normal relaxation effect would be quenched by a factor hmJ6 and quantum size effects would be mani- fested in a modified relaxation time given by Remembering that for arbitrary shape particles 6 is approximately given by EF/N (the Fermi energy divided by the total number of conduction electrons in the particle) and therefore 6 varies as d-3 one can easily see that the linewidth will increase at first from the bulk value as the particle size decreases because of the d-' surface effect [eqn (3)] and will then continue to decrease sharply (with the g-shift) due to the d2qEantum size effect [eqn (4)].99 With this basic information at hand the c.e.s.r.linewidths and g-shifts for Ag/Ar and Ag/Kr deposits in the range 1/5 x lo3 to 1/2 x lo2 have been investigated and the results are displayed in graphical form in fig. 22(b). The observed changes in c.e.s.r. bandwidths (15-35 G) and g-shifts (2.013-2.028) depicted in fig. 22(b),together with the accompanying growth in ampli- tude with silver concentration matrix annealing and silver-atom ph~toaggregation,~~ can be considered to represent one of the few direct experimental verifications of the existence of quantum size effects [see ref. (95) for an up-to-date account of the field] in fine metallic particles as first discussed by Frohlich some forty-three years ago lo' and later formalized by Ku~o,~' Ka~abata~~ Gor'kov and Elia~hberg,~~ and According to the theory of Ka~abata,~~ G a c.e.s.r.bandwidth of ~25 corresponds to particle diameters of roughly 10 A a dimension close to that expected for a close-packed 13-atom cluster. This estimate places the e.s.r. data in line with the conclusions of the corresponding optical experiments.20 The details of these spectra have yet to be de~iphered,~~ but it would appear that quantum size effects are being detected in the aforementioned e.s.r./optical experiments and can provide crucial information on the transition cluster to bulk metal. Along these lines we have performed preliminary SCF-XU-SWcalculations for Ag (n = 2-6) assuming linear geometries as a starting point,43 analogous to Baet- SPECTROSCOPY CHEMISTRY AND CATALYSIS zold's4* EHMOlCNDO studies.The genesis of the band structure density of states ionisation potentials and optical transition energies can be monitored from these calculations as the cluster size grows (fig. 23). Around z4-6 atoms one generates an electronic scheme which carries some of the features of bulk silver. An intriguing early indicator emerging from these computations is the expectation of s-to-p-like I- t? m. t".. m. FIG. 23.-SCF-Xa-SW molecular orbital energy level scheme for linear Ag clusters (n = 2-6) [see text and ref. (43) for details of these calculations]. G. A. OZIN interband transitions for a silver cluster containing as few as six atoms which may be the molecular origin of the broad optical absorption which develops initially in the 350-nm region (fig.21) that is " an envelope of molecular one-electron cluster excita- tions ". Past this nucleation stage one apparentlymoves into a size domain capable of supporting " both " plasmon and molecular absorptions arising from the coexistence of silver clusters and microcrystallites. However analysis of the data in this inter- facial region is complicated by the near degeneracy of the two different modes of exci- tation. Further silver aggregation transforms the optical picture into one displaying only collective excitations of silver micro crystal^.^^^^^ Other features of interest that derive from these Xcc calculations of linear Ag2_ clusters include the beginning of a sawtooth dependence of the ionisation potentials and electron affinities on n,43in which even clusters have the largest ionisation poten- tials and odd clusters the largest electron affinities a property first noted by Baetzold and HamiltonLo2 in their EHMO calculations of linear Ag chains in line with the normalised mass spectral intensity data for Ag; (n = 2-30)'03 which are thought to indicate a pronounced odd-even effect of cluster stability.In this context the XC~ calculations predict that the energy gap between the HOMO and LUMO from n = 2 to n = 6 decreases monotonically indicating a long-wavelength shift in light absorption as the particle size increases.43 Experimentally this is the trend which is observed for the lowest energy absorptions of small Ag clusters growing in solid Ar up to about n z 6.If one plots these AE values as a function of l/n a remark- able linear correlation is discovered [fig. 24(a)].20 As the cluster-size progresses to higher values one only observes a gradual broadening of the low-energy optical absorption and a tendency for the band maximum to red-shift towards a limiting value of ~555-570nm [fig. 24(b)]. Interestingly the low-energy band broadens to disappearance at a cluster value corresponding to n z 10-15 estimated from the AE against I/n extrapolation shown in fig. 24(a). For an assumed cubic close-packed structure (found in crystalline silver metal itself) this n-value calculates close to a 10 A microcrystal diameter (in line with the conclusions drawn from the corresponding e.s.r. studies). 36 Noteworthy is the spectral persistence and monotonic red-shifting of the lowest energy " molecular cluster absorption ",but to a seemingly limiting value of around 600 nm (rather than into the i.r.as predicted by the calculations indicating semiconductor character) for an estimated 6 Z n 215 [fig. 24(b)]. This seems to be a size range which also appears to display microcrystallite proper tie^.^^^^^ One might speculate that this behaviour implies a structural-electronic interconversion between linear chains for small " molecular '' Ag clusters and a close-packed three- dimensional array for Ag clusters larger than six. Finally we note recent Xa cal-culations for Ag6'O4 which claim that an octahedral six-atom cluster is large enough to reproduce trends in energy differences such as the width of the d-bands and the distance from the top of the d-bands to the Fermi level as found in experiment and bulk energy band calculations.Significant also is an X-ray structural report of a perfectly octahedral Ag clusterlo5 entrapped within the cage of zeolite A which can be considered to be the smallest possible fully developed single crystal of silver. In the same vein we note the recent discovery of growing silver clusters in the size range n z 2-6 in the octahedral sites of Ag+-exchanged zeolite Y.Io6 The optical spectra of these samples have been recorded (fig. 25) as a function of dehydration tempera- ture106 and show remarkable resemblances to those observed in our studies for very small Ag clusters (n = 2-6) growing in inert gas ice wax and olefin matrices The implication seems to be a weak cluster-support interaction for the Ag,/zeolite system not unlike that found in our low-temperature matrices.A related study recently claimed that silver ion-exchanged zeolites are sensitive to visible light and release 33m 31 -29 -cn 27 -cd m c, 4 c 25 -P I E a V f O mz 23 -n L 0 \ A% iii n a 21 -IS t I I 1 1 1 1 1 I 200 3 00 400 500 600 wavelength Inm FIG.24.-(a) Graphical representation of the energy (cm-') of the lowest energy absorption of Ag (where n = 1-7) as a function of l/n in solid argon. The dotted line on the Ag point represents the energy uncertainty because of the observed 390/410 nm site splitting.The Ag, to Ag, points represent the extrapolated cluster sizes corresponding to the lowest energy (HOMO-LUMO) Ag absorptions observed in the range 555-570 nm [see text and ref. (20)]. (b)Collection of optical spectra of silver clusters Ag (n = 2 3 4 5 6 .. .) isolated in solid Ar at 10-12 K showing the red shifting and broadening behaviour of the lowest energy ab- sorption. Note also the growth of broad plasmon resonance-like absorptions in the region of 350-400 nm with superimposed molecular cluster absorptions. The plasmon absorption becomes progressively more pronounced with increasing silver cluster size.20 G. A. OZIN 250 350 150 550 200 250 300 LOO 600 300 LOO 500 wavelength I nm wavelength Inm FIG.25.-Optical spectroscopic comparisons between the growth of very small silver clusters from silver atoms (a) in the cages of zeolite Y (tentative cluster assignments indicated) and (b)within the confines of solid argon [see ref.(20) and (106) and text for details]. oxygen from absorbed water while irradiated and form molecularly dispersed silver." Furthermore thermal reduction of zeolite water to hydrogen occurs above 600 "C with production of Ag+. It has been suggested that variable-size Ag clusters and cluster anions are responsible for the sequence of redox events that leads to photo- catalytic water splitting over Ag(1)j~eolites.~~ The silver clusters are envisaged to form a system of two "size-tunable "redox pairs Agr/Agr-l and AgT-f/AgF-2 con- nected by several processes which match the HzO,02,H2system and its intermediates.To realise a practical photochemical solar-energy storage system however the redox pairs must be carefully matched and cluster agglomeration to bulk metal must be prevented.70 CONCLUSIONS On a final note let us briefly contemplate the likely future of metal-atom diatomic- metal and metal-cluster chemistry. We would like to propose that new directions should focus attention on the chemistry and catalytic properties of these intriguing compositions of matter. The past decade has witnessed remarkable growth in the ability to manipulate metal-atom recombination reactions under cryochemical conditions to fabricate supercoordinatively unsaturated cluster chemisorption models and to probe spectroscopically and theoretically subtle details of the molecular architecture of naked clusters and cluster complexes.However of central concern SPECTROSCOPY CHEMISTRY AND CATALYSIS to this author has been the noticeable absence of any chemical procedure in the open literature for capturing and stabilizing these fascinating few-atom metal cluster com- binations for subsequent testing as real working catalysts. If existing predictions of catalytic activity/selectivity patterns for few-atom clusters are correctly founded then one can expect marked improvements in experimental operating conditions (tempera- ture pressure) with concomitant energy conservation by eliminating inefficient and uneconomical processes currently employed and held onto so dearly by industry.It is our contention that an extremely promising new direction for concerns the marriage of metal vapour chemistry and liquid polymer chemistry the concept being to wrap functionalized polymers around metal atoms and very small metal clusters,66 thereby providing a room-temperature chemical trapping mechanism for stabilizing the metal species rather than the less practical matrix-isolation method. Significant early observations in this new field of research are that (i) the experiments can be performed at either a matrix spectroscopic or macrosynthetic scale of opera- tion (ii) the experiments can be entirely conducted at or close to room temperature (iii) the resulting " polymer-stabilized metal cluster " compositions are homogeneous liquids which are stable at room temperature and (iv) the methodology is easily extended to bimetallic and trimetallic polymer combinations.Adjustable cavity size polymeric crown and cryptate-type interactions may be responsible for the remarkable stabilization of very low nuclearity clusters (Mn; n = 1-5) by polymers such as poly-ethers and polysiloxanes which have the common property of solvating backbones.66 In this context we note the recent report107 of silver atom stabilization* in the hetero- atom-crown species 2-tetradecyl-l,4,10,13-tetraoxa-7,16-diazacyclo-octadecane, by photoreduction of the Ag' precursor complex. n n ro Ag,") -HN NH hu HN Ago NH These early indicators of the feasibility of ambient temperature metal-atom and metal-cluster stabilization point to an exciting and promising future for the chemistry of these intriguing " little pieces of metal ".I wish to acknowledge the invaluable assistance of my graduate students and colleagues whose names appear in the cited articles who have made enormous contri- butions to the development of metal vapour cryochemistry particularly the pioneering explorations of that scientifically intriguing interdisciplinary " fuzzy '' interface between metal clusters and metal surfaces/coordination and chemisorption. would also like to express my special indebtedness to Mr. Ted Huber and to Mr. Alex Campbell Mr. Karl Molnar Mr. Martin Mittelstaedt and Mr. Bob Torbet for their * In this context caution needs to be exercised as the optical spectrum of the proposed crown ether stabilized silver atom shows a strong absorption at 415 nm,lo7 which should be critically com- pared with the 390 nm absorption of site D' silver atoms in-Ag/H20 matrices (four-fold oxygen solvation shell),80 the 410 nm absorption of Ag/polyphenylmethyl siloxane compositions,1o8 and the 370-450 nm plasmon absorption of 10-100 A silver microcrystallites in rare-gas solids 20*47*76c and photosensitive glasse~.~~*~~ G.A. OZIN expert machine shop contributions; and to Mrs. Elinor Foden for typing this manu- script. The generous financial assistance of the National Research Council of Canada Operating Grant Programme New Ideas Programme and Strategic Energy Programme is gratefully acknowledged. We are also indebted to the Atkinson Foundation the Connaught Fund Imperial Oil of Canada and the Lash Miller Chemical Laboratories and Erindale College for support at various stages of this work.E. L. Muetterties T. N. Rhodin E. Band C. F. Brucker and W. R. Pretzer Chem. Rev. 1979 79 91. E. L. Muetterties Bull. Soc. Chim. belg. 1975 84 959; 1976 85 451. E. L. Muetterties Angew. Chem. Int. Edn 1978 17 545. E. L. Muetterties Science 1977 196 839. ’R. Ugo Catalysis Rev. 1975 11 225. T. N. Rhodin and N. Rosch Faraday Disc. Chem. Soc. 1974,58,28. G. A. Somorjai Angew. Chem. Int. Edn 1977 16 92; Accounts Chem. Res. 1976 9 248 and references therein. E. W. Plummer W. Salaneck and and J. S. Miller Phys. Rev. B 1978 18 1673. R. Mason Israel J. Chem. 1976/77 15 174 and references therein.lo Y. Takasu and A. M. Bradshaw Chemical Physics of Solids and their Surfaces (Spec. Period. Rep. The Chemical Society London 1978) vol. 7 p. 59. R. P. Messmer in Nature of the Surface Chemical Bond ed. T. Rhodin and G. Ertl (North Holland Press Amsterdam 1979) chap. 2. l2 R. P. Messmer K. H. Johnson C. Y. Yang S. K. Knudson and J. B. Diamond Phys. Rev. B 1976 13 1396. l3 T. H. Upton W. A. Goddard I11 and C. F. Melius J. Vac. Sci. Techno/. 1979 16,531 and references cited therein. l4 H. F. Schaeffer 111 Accounts Chem. Res. 1977 10 287. Is G. A. Ozin Catalysis Rev. Sci. Eng. 1977 16 191 ; Accounts Chem. Res. 1977 10 21 and references cited therein. l6 (a) M. Moskovits and J. Hulse J. Chem. Phys. 1977 66 3988; Surface Sci. 1976 57 125; (6) M.Moskovits Accounts Chem. Res. 1979 12,229. l7 (a) G. A. Ozin W. J. Power T. H. Upton and W. A. Goddard 111 J. Amer. Chem. Soc. 1978 100,4750; (b)H. Huber G. A. Ozin and D. McIntosh Inorg. Chem. 1977,16 3070; (c) W. J. Power A. J. Lee Hanlan and G. A. Ozin Inorg. Chem. 1978 17 3648. A. J. Lee Hanlan and G. A. Ozin Inorg. Chem. 1977 16 2857. l9 M. Moskovits and J. Hulse J. Chem. Phys. 1977 67 4271. 2o H. Huber and G. A. Ozin Inorg. Chem. 1978,17 155. 21 J. Korringa Physica 1947 13 392; W. Kohn and N. Rostoker Phys. Rev. 1954 94 11 1 1. 22 W. Klotzbucher and G. A. Ozin J. Amer. Chem. Soc. 1978 100 2262; J. Mol. Catalysis 1977 3 195. 23 S. Mitchell and G. A. Ozin J. Amer. Chem. Soc. 1978 100 6776. 24 S. Mitchell and G. A. Ozin unpublished work on Agl,2,3,/Ar concentration studies based on the kinetic theory of M.Moskovits and J. Hulse J.C.S. Faraday II 1977,73,471. 25 S. Mitchell and G. A. Ozin Trisilver Cryophotochemistry Inorg. Chem. 1979 18 2932 first presented at the A.C.S. Cluster Symposium Anaheim March 1978. 26 T. A. Ford H. Huber W. Klotzbucher E. P. Kundig M. Moskovits and G. A. Ozin J. Chern. Phys. 1977 66 524. 27 (a) S. Mitchell J. Farrell G. A. Ozin and G. Kenney-Wallace J. Amer. Chem. Soc. in press; (b)C. E. Moore Nut. Bur. Stand. Circ. 467,vol. 11,111,1958; (c)T. Azumi and S. P. McGlynn J. Chem. Phys. 1962,37,2413; (d)J. Z. Klose Astrophys. J. 1975,198 229; (e)P. R. Moran Phys. Rev. 1965,137 A1016; (f) T. A. Fulton and D. B. Fitchen Phys. Rev. 1969,176,846; (g) R. Englman The Jahn-Teller Eflect in Molecules and Crystals (Wiley-Interscience New York 1972); M.D. Sturge Solid State Phys. 1967 20 91 ; A. D. Liehr J. Phys. Chem. 1963 67 389 (h) S. Mitchell G. A. Ozin and G. Kenney-Wallace in preparation. 28 W. Klotzbucher and G. A. Ozin Inorg Chem. 1979 18 2101 and N.B.S. Special Report on High Temperature Science Washington 1978. 29 S. Mitchell D. McIntosh G. A. Ozin J. G. Norman Jr and L. Noodleman J. Amer. Chent. Soc. 1979 101 3504. 62 SPECTROSCOPY CHEMISTRY AND CATALYSIS 30 S. Mitchell H. Huber and G.A. Ozin to be published. 31 S. Mitchell Metal Atom-Matrix Recombination Reactions MA. Thesis (University of Toronto 1978); W. Klotzbucher Bimetal Vapour Chemistry Ph.D. Thesis (University of Toronto 1979). 32 D. M.Kolb and D. Leutloff Chem. Phys. Letters 1978 55 264. 33 G. A. Ozin Naked Cluster Cryophotochemistry paper presented at Materials Sci. Conf. Boston November 1978. " C. E. Moore Nut. Bur. Stand. Circ. 467,1949 1; 1952 2; 1958,3; C. H. Corliss and W. L. Bozman Nut. Bur. Stand. Monograph 53 1962. 35 G. York R. Scheps and A. Gallagher J. Chem. Phys. 1975 63 1052; R. E. Smalley D. A. Auerbach P. S. H. Fitch D. H. Levy and L. Wharton J. Chem. Phys. 1978 66 3778 and references cited therein. 36 G. A. Ozin J. Amer. Chem. Sac. 1980 102 3301. 37 P. H. Kasai and D. McLeod Jr J. Chem. Phys. 1971,55 1566. 38 F. Forstmann D. M. Kolb D. Leutloff and W. Schulze J. Chem. Phys. 1977 66 2086; W. Schulze D. M. Kolb and H. Gerischer J.C.S. Faraday ZZ,1975 71 1763; F. Forstmann D.M. Kolb and W. Schulze J. Chem. Phys. 1976 64 2552. 39 A. A. Belyaeva Y. B. Predtechenski and L. D. Shcherba Opt. Spectrosc. 1973,34,21. 40 R. L. Mowery J. C. Miller E. R. Krausz P. N. Schatz S. M. Jacobs and L. Andrews J. Chem. Phys. 1979,70 3920. 41 J. H. Ammeter and D. C.Schlosnagle J. Chern. Phys. 1973 59 4784. 42 A. G. Shenstone Phys. Rev. 1940 57 894. 43 R. P. Messmer G. A. Ozin D. McIntosh and S. Mitchell SCF-Xu-SW Calcirlationsfor Ag Clusters n = 2-6 in preparation. 44 (a)B. Klemen and S. Lindkvist Arkiv. Fys. 1954 8 333; 1955 9 385; (b) J. Ruamps Ann. Phys. (Paris) 1959 4 1111; (c) R. C. Maheshwari Indian J. Phys. 1973 31 368; (d) N. hlund R. F. Barrow W. G. Richards and D. N. Travis Arkiu. Fys. 1965 30 171 ; (e) B. Rosen Spectroscopic Data Relative to Diatomic Molecules (Pergamon Press New York 1970); (f)C.M. Brown and M. L. Ginter J. Mol. Spectr. 1978,69,25. 45 K. A. Gingerich J. Crystal Growth 1971 9 31. 46 (a)C.R. Hare T. P. Sleight W. Cooper and G. A. Clarke Inorg. Chem. 1968,7,669; (b)R. C. Baetzold J. Chem. Phys. 1971,55,4355; (c)A. B. Anderson J. Chem. Phys. 1978 68 1744. 47 W. Schulze H. U. Becker and D. Leutloff J. Physique 1977 C-2,7; W. Schulze H. U. Becker and H. Abe Chem. Phys. 1978 35 177 and references therein. 48 R. C. Baetzold J. Chem. Phys. I97 1 55 4363; J. Catalysis 1975 51 1 and references cited therein. 49 D. M. Lindsay D. R. Herschbach and A. L. Kwiram Mol. Phys. 1976,32 I 199. 50 K. M. Monahan V. 0.Jones and V. Rehn J. Chem. Phys, 1979,71,2360. 51 R.C. Baetzold Photogr. Sci. Eng. 1973 17 78. " W. H Gerber and E. Schumacher J. Chem. Phys. 1978,69 1692. 53 See for example P. Montano Faraday Symp. Chern. Soc. 1980 14 79. s4 W. Klotzbucher G. A. Ozin J. G. Norman Jr and H. J. Kolari Inorg. Chem. 1977 16 2871. 55 W. Klotzbucher and G. A. Ozin Inorg Chem. 1976 15 292; 1979 18 2101. 56 J. H. Sinfelt et al. U.S. Patent 3871997 (Mar. 1975); 3901827 (Aug. 1975); 3429619 (Dec. 1975); 3953368 (Apr. 1976). '' (a) J. H. Sinfelt Accounrs Chem. Res. 1977 10 15 and references cited therein; (b) J. H. Sinfelt Ann. Rev. Mat. Sci. 1972 2 641; J. Catalysis 1973 29 308; (c) D. W. McKee Trans. Faraday SOC. 1965 61 2273; (d) R. Bouman and W. M. H. Sachtler J. Catalysis 1970 19 127; (e)R. J. Kokes and P. H. Emmett J.Amer. Chem. Soc. 1959,81 5032; (f)J. H. Sinfelt and J. A. Cusumano in Advanced Moterials in Catalysis ed. J. J. Burton and R. L. Garten (Academic Press New York 1977) p 1 ; J. J. Burton and R. L. Garten ibid. p. 33 and references cited therein; (g) R. T. K. Baker R. D. Sherwood and J. A. Dumesic J. Vcrc. Sci. Technol. 1979 16 493. '* (a) M. Hansen Constitution ofBinary AIIoys (McGraw Hill New York 1958); (b) P. Kasai and D. McLeod J. Phj9s. Chem. 1978 82 1554; (c) P. A. Montano J. Appl. Phys. 1978 49 1561. 59 D. McIntosh R.P. Messmer and G. A. Ozin to be published. 6o J. Harris and R. 0.Jones J. Chem. Phys. 1979 70 830. 6' (a) Effremov et a/. Opt. Spectr. 1974 36 381 ; (b) Norman et nl. Itiorg. Chem. 1977 16 2871 ; (c)Cooper et al. J. Phys. Chem. 1972 76 2268.62 W. Klotzbucher and G. A. Ozin Inorg. Chem. in press. 63 D. M. Mann and H. P. Broida J. Cheni. Phys. 1971 85 84. G. A. OZIN 63 64 (a)J. H. Sinfelt Y.L. Lam J. A. Cusumano and A. E. Barrett J. Catalysis 1976 47 227 and references cited therein; (b) E. P. Prestridge G. H. Via and J. H. Sinfelt J. Catalysis 1977,50 115. 65 (a)C. H. Bartholomew and M. Boudart J. Catalysis 1973,29,278; (6) R. L. Garten and D. F. Ollis J. Catulysis 1974,35,232; (c)R L. Garten J. Catalysis 1976,43 18; (d)M. A. Vannice and R. Id. Garten J. Mol. Catalysis 1975 1 201. 66 C. G. Francis H. Huber and G. A. Ozin J. Amer. Chem. SOC. 1979 101 6250; Inorg. Chern. 1980 19 219; Proc. EUCMOS Meeting Frankfurt Sept. 1979 in Spectroscopy in Chemistry and Physics Modern Trends (Elsevier Amsterdam) and J.Mol. Struct. 1980 59 55; Proc. Climax Inst. Molybdenum Conf. Ann Arbor Aug. 1979; Metal Vapor. Synthesis of Organo-metal Polymers and Polymer-Supported Metal Clusters in Organometal Polymers ec. C. E. Carraher (Academic Press New York 1980) and J. Macromol. Chem. in press; Angew. Chem. 1980 in press. 67 The Electron Factor in Catalysis N.B.S. Special Report No. 475 34 1977. P. H. Kasai Accounts Chem. Res. 1971,4 329 and references cited therein. 69 J. F. Hamilton J. Vac. Sci. Technol. 1976 13 319 and references cited therein. 70 S. Leutwyler and E. Schumacher Chimia 1977 31 475 and references cited therein. 71 The Physical Basis of Heterogeneous Catalysis ed. E. Drauglis and R. I. Jaffee (Plenum Press London and New York 1979 and references cited therein.7L W. J. Peer and J. J. Lagowski J. Amer. Chem. SOC. 1978,100,6260. 73 B. Zmbova H. R. Ihle and E. Langenscheidt J. Chem. Phys. 1977,66,5105. 74 J. L. Dye Angew. Chem. Int. Edn. 1979 18 587. 75 P. N. Hawker E. P. Kundig and P. L. Timms J.C.S. Chem. Comm. 1978,730. 76 (a)G. A. Ozin and H. Huber Inorg. Chern. 1979 18 1402; (b) M. Hofrnann S. Leutwyler and W. Schulze Chern. Phys. Letters 1979 40 145; (L') T. Welker and T. P. Martin J. Clzem. Phys. 1979 70 5683. 77 H. Huber and G. A. Ozin unpublished work. 78 V. Piacente and K. Gingerich High Temp. Sci. 1977 9 189. 79 (a) Advanced Materials in Catalysts ed. J. J. Burton and R. L. Garten (Academic Press New York 1977); (6) The Structure of Metallic Catalysts J.R. Anderson (Academic Press New York 1975) and references cited therein. *O H. Huber P. McKenzie and G. A. Ozin J. Amer. Chem. SOC. 1980 102 1548. (aj B. L. Bales and L. Kevan J. Chern. Phys. 1970 52 4644; 1971 55 1327; (b) L. Kevan H. Hase and K. Kawabata J. Chem. Phys. 1977 66 3834. 82 R. A. Zhitnikov N. V. Kolesnikov and V. I. Kosyakov Sou. Phys. J.E.T. P. 1963,17 8 15. 83 R. A. Zhitnikov and N. I. Melnikov Opt. Spectr. 1968 24 53. 84 F. Franks Water A Cotnprehensiue Treatise ed. F. Franks (Plenum Press New York 1972) vol. 1 chap. 4; (b)J. A. Ghormley J. Chem. Phys. 1968,48 503. U. Kreibig J. Phys. F 1974 4 999; Appl. Phys. 1976 10 255; Solid State Cornm. 1978 28 767; T. P. Martin and H. Schaber Phys. Stat. Sol. (b) 1977 81 lC41; L. Genzel and T.P. Martin Surface Sci. 1973 34 33; C. G. Granquist N. Calander and 0.Hunderi. Solid State Comm. 1979 31 249; and references cited therein. 86 L. Genzel T. P. Martin and U. Kreibig Z. Phys. B 1975 21 339. W. Klotzbucher S. A. Mitchell and G. A. Ozin Inorg. Chem. 1977 16 3063. W. R. Turner Ind. Eng. Chem. Prod. Res. Deuelop. 1971 10 238; M. G. Broadhurst J. Res. N.B.S. 1972 66A 241 ; and references cited therein. 89 K. J. Klabunde T. Groshens M. Brezinski and W. Kennelly J. Amer. Chem. SOC. 1978 100 4437 and references cited therein. 90 D. McIntosh G. A. Ozin and H. Huber J. Organomet. Chem. 1976,121 127. 91 P. H. Kasai and D. McLeod Jr J. Amer. Chem. SOC.,1975,97,6602. 92 D. McIntosh G. A. Ozin and R. P. Messmer SCF-Xa-SW Calculations for M(C2H4),where M = Cu Ag Au presented at C.I.C.Meeting Vancouver 1979 and Inorg. Chem. 1980 in press. 93 H. Huber M. Andrews and G. A. Ozin unpublished work. 94 J. Kittel Introduction to Solid State Physics (Wiley New York 1970). 95 R. F. Marzke Catalysis Rev. Sci. Eng. 1979 19 43. 96 (a)M. Smithard Solid State Comm. 1973 13 153; (b)W. K. Knight J. Physique 1977 C-2 110; (c) D. A. Gordon Phys. Rev. B 1976 13 3738; (d) R. Monot A. Chiitelain and J. P. Borel Phys. Letters 1971 34A 57 and J. Physique 1977 C-2 115; (e) R. Dupree C. T. Forwood and M. H. A. Smith Phys. Stat. Sol. 1967 24 525; (f)R. Monot C. Narbel and J. P. Borel Nuovo Cimento 1974 19 253. 97 R. Kubo J. Phys. SOC.Japan 1962 17 975. SPECTROSCOPY CHEMISTRY AND CATALYSIS 98 L. P. Gorkov and G.M. Eliashberg Soviet Phys. J.E.T.P. 1965,21,940. 99 A. Kawabata J Phys. SOC.Japan 1970 29 902. loo R. J. Elliot Phys. Reo. 1954 96 266. lo’ H. Frohlich Physicu 1937,6,406. lo’ R. C. Baetzold J. Chem. Phys. 1971,55 4363; J. Appl. Phys. 1976 47 3799; R. C. Baetzold and R. E. Mack J. Chem. Phys. 1975 62 1513 and references cited therein. lo3 G. Hortig and M. Muller 2.Phys. 1969 221 119. N. Rosch and D. Menzel Chem. Phys. 1976 13 243. lo’ Y. Khim and K. Seff J. Amer. Chem. Soc. 1977 99 7055. lo6 R. Kellerman and J. Texter J. Chem. Phys. 1979 70 1562. lo’ R. H. Humphry-Baker M. Gratzel P. Tuundo and E. Pilizzetti Angew. Chem. 1979,91,669. lo* C. G. Francis and G. A. Ozin unpublished work. lo9 C. G. Francis and P. L. Timms J.C.S. Chem. Comrn. 1977,466; J.C.S.Dalton 1980 in press; C. G. Francis Ph.D. Thesis (University of Bristol 1978).

ISSN:0301-5696    

DOI:10.1039/FS9801400007

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Electron Spin Resonance Study of Intermetallic Molecules CUM and AuM (M=Mg Zn Cd and Hg) AND DONALD BY PAULH. KASAI~ MCLEODJR Union Carbide Corporation Tarrytown Technical Center Tarrytown New York 10591 U.S.A. Received 21st August 1979 A series of intermetallic diatomic molecules CUM and AuM (M = Mg Zn Cd and Hg) generated in argon matrices was examined by e.s.r. spectroscopy. For all the CUM and AuM examined the hyperfine coupling tensors to the Cu and Au nuclei were isotropic. The unpaired electron resides in an orbital given essentially by an antibonding combination of the valence s orbitals of Cu (or Au) and M atoms. Coupling interactions observed with magnetic Cd and Hg nuclei and the analysis of the g tensor indicate a small admixture ( % 10%) of the valence pz orbital of the atom M in each CUM and AuM examined.The nature of metal-metal bonds found between metal atoms in complex coordina- tion compounds1 as well as those existing between ligand-free metal has been the subject of many recent investigations. The homonuclear diatomic molecules of various metals have been examined by mass spectroscopy of the vapour phase3 and by optical spectroscopy using the matrix isolation techniq~e.~ Electron spin resonance (e.s.r.) spectra of heteronuclear intermetallic diatomic molecules if ob-served would be particularly elucidative of the interaction between the pair of metal atoms. When metal atoms M are condensed in inert gas matrices at liquid helium tem- perature the diatomic species MZ are found in a quantity much larger than that expected from the vapour phase compo~ition.~ The diffusion of metal atoms within the quasi-liquid surface layer existing during the deposition must be responsible for the extra dimerization process.It follows that a sufficient amount of heteronuclear diatomic molecules may be generated in rare gas matrices by cocondensation of two different atoms. Recently we reported an e.s.r. study of AgM (M = group I1 metal) generated by this cocondensation technique in argon mat rice^.^ The present paper describes the result of the extension of the matrix-isolation e.s.r. study to include CUMand AuM (M = Mg Zn Cd and Hg). The e.s.r. spectra of CUM and AuM are expected to be characterized by extremely large hyperfine interactions with the Cu and Au nuclei and hence to provide more sensitive probing of the semi-filled orbital.The observed e.s.r. spectra revealed that the hyperfine coupling interaction is indeed large but is essentially isotropic. As in the case of AgM the electronic configurations of CUM and AuM are concluded to be a;a*,'where a and a,*are the bonding and antibonding a orbitals arising essentially from the valence s orbitals of Cu (or Au) and the atom M. -f To whom correspondence should be addressed. Present address IBM (78V-906) Box 390, Poughkeepsie New York 12602 U.S.A. E.S.R. OF INTERMETALLIC MOLECULES EXPERIMENTAL The cryostat-spectrometer assembly that would permit trapping of high temperature vapour phase species in a rare-gas matrix at -4 K and observation of the resulting matrix by e.s.r.has been described earlier.6 In the present series of experiments two metal atoms were vaporized from two independently resistively heated tantalum cells. The tantalum cells contained the metals in granular form and were heated to a temperature at which the vapour pressure of the respective metal would be 0.1-0.5 Torr. In the case of Au in order to prevent alloying with the Ta cell Au metal was placed inside an alumina tube capped with a molybdenum plug and then placed inside the Ta cell. The gold vapour effused through an opening drilled through the Ta and alumina walls. The frequency of the e.s.r. spectrometer locked to the sample cavity was 9.42 GHz and all the spectra were obtaiiKd while the matrices were maintained at -4 K.RESULTS SPECTRA AND ANALYSES The e.s.r. spectra of Cu atoms (3d1'4s') and Au atoms (5dl06s1)isolated in argon matrices have been studied earlier.7 The spectra are characterized by extremely large hyperfine interactions with the nuclei of spin $ i.e. 63Cu (natural abundance = 69% I = 3,p = 2.2206&) 65Cu (natural abundance = 31% Z = 3,p = 2.3790&) and I9'Au (natural abundance = loo% I = 4,p = 0.1439pN). The splittings of energy levels of an isotropic system with S = 3and I = $ in a magnetic field Hcan be readily obtained from the Breit-Rabi solution of the spin Hamiltonian (1) X=gPH*S+AZ*S (1) and are illustrated in fig. l(a). Ar f2 ft FIG.1.-(a) Splitting of energy levels of a system with S = 4and I = $ in external magnetic field.Observable transitions are indicated for the two cases Av > 2A and hv < 2A. (b) Dependence of the resonance fields (gpH//Av)upon the hyperfine coupling coupling constant (A/hv)for the system given in (a). P. H. KASAI AND I). MCLEOD JR 67 As can be seen from fig. 1 the e.s.r. spectrometer frequency v must be larger than the zero-field splitting (I + +)A in order to observe the '' normal '' (21+ 1) hyperfine components of the e.s.r. transitions AMs = & 1 AMI = 0. If the zero-field splitting is larger than the spectrometer frequency only two transitions are observable; one corresponds to the highest field hyperfine component of the normal e.s.r. transition (AMs = &l MI = -I) and the other corresponds to an n.m.r.transition (M = _-i MI = -I ++ -I + 1). The Breit-Rabi solution also yields the following ex- pressions for the resonance positions of the normal e.s.r. transitions and the particular n.m.r. transitions discussed above. (21+ l)a -2 (3) q(n.m.r.) = 2-a where Dependencies of the reduced resonance fields y(Mz)and y (n.m.r.) upon the reduced hyperfine coupling constant a are illustrated in fig. l(b). The hyperfine coupling constants of Cu atoms isolated in an argon matrix are 6.151 and 6.587 GHz for 63Cu and 65Cu re~pectively.~ Thus for a typical X-band spectrometer frequency of 9.4 GHz a = 0.65 for 63Cuand a = 0.70 for 65Cu as indicated in the figure and only the transitions given by y (n.m.r.) and y(-*) are observed. Fig. 2 shows the e.s.1.spectrum observed from an argon matrix in which Cu and 63CuMg +I $. ' 1 4 P- I I I I I I 0 1 2 3 4 5 6 kG FIG.2.-E.s.r. spectrum of an argon matrix containing Cu and Mg atoms. Mg atoms were cocondensed. The signals indicated by the arrows are those of isolated Cu atoms. Two sets of quartets were recognized among the remaining E.S.R. OF INTERMETALLIC MOLECULES signals. They were unique to matrices containing Cu and Mg atoms and hence were assigned to (j3CuMg and TuMg as indicated. The overall spread the line shape and their positions indicate that the coupling tensors to the Cu nuclei of these species are nearly isotropic and correspond to the situation C( = 0.28 in fig. l(b). The molecular symmetry of CUM (or AuM) dictates that their e.s.r.spectra be compatible with an axially symmetric spin Hamiltonian [eqn (4)] 2-F = g,,PHzSz + glP(KSx + HYSY) + AllIz& + AI(IxSx + IySy) (4) where A,,and Al represent the hyperfine coupling tensor to the Cu (or Au) nucleus. When the hyperfine interaction is of such magnitude as that encountered here the usual second-order solution of the matrix cannot be used to determine the relevant parameters in eqn (4). However it can be shown that when the magnetic field is parallel to the symmetry axis the secular determinant derived from eqn (4) is tri- diagonal and hence can be expanded exactly using a continuous fraction technique.* From the eigenvalues expressed in the continuous fraction form the following " con-tinued expressions " were obtained for the resonance positions of the parallel com- ponents.(5) where hV and a If /A,,-All < gPH a condition which is surely met in the present case the secular determinant derived from eqn (4) can be treated as being tridiagonal when the mag- netic field is perpendicular to the symmetry axis also. It is thus possible to derive the following " continued expressions " for the perpendicular components of the resonance signals Hl(MI) = HY -MIA; -Fl -GI (6) where Q:[I(I + 1) -MI(MI+ 111 F_i = HdMI) + (MI + %)A; + F_L and P. H. KASAI AND D. MCLEOD JR A close examination of the CuMg spectrum observed in an expanded scale clearly resolved the pattern expected from an axially symmetric system. The measured parallel and perpendicular positions of each hyperfine component of 63C~Mg are listed in table 1.From the observed parallel and perpendicular resonance positions TABLE1 .-OBSERVED (G) RESONANCE POSITIONS OF 63C~M 1740 480 465 250 (n.m.r.) 273 (n.m.r.) ++{& 2392 971 940 986 965 -4{& 3335 2538 2520 3346 2570 2578 4587 5188 5203 5483 4603 5234 5306 5795 v/GHz 9.4255 9.4256 9.4237 9.4213 of the MI = 3-components and eqn (5) and (6) the consistent set of g and hyperfine coupling tensors can be readily determined through a computer-assisted iteration process. The g tensor and the hyperfine coupling tensor to the Cu nucleus thus deter- mined for 63C~Mg are given in table 2. 2.-sPIN HAMILTONIAN TABLE PARAMETERS OF 63C~M ~ ~ Mg 2.0031 (7) 1.9984 (7) 2.609 (2) 2.618 (2) Zn Cd Hg 2.0049 (7) 2.0001 (7) 1.996 (1) 1.9895 (7) 1.9659 (7) 1.890 (1) 4.186 (2) 4.200 (2) 4.968 (3) 4.195 (2) 4.218 (2) 4.976 (3) 2.34 (10) 2.83 (10) free Cu 1.9994 (2) 6.151 (1) a Coupling to "'Cd or 199Hg.Fig. 3-5 show respectively the e.s.r. spectra observed from argon matrices in which Cu and Zn Cu and Cd and Cu and Hg atoms had been cocondensed. In each trace the signals due to isolated Cu atoms are marked by solid straight arrows. The two sets of quartets readily recognized in fig. 3 and 4 were assigned to CuZn and CuCd. The overall spread of the quartets indicates that the hyperfine coupling constants to the Cu nuclei in these molecules correspond to the case a E0.45 in fig. l(b). In the case of CuHg (fig.5) the normal quartet signals were no longer observed; only the n.m.r. transition and the highest-field component of the e.s.r. transitions were recog- and 65C~Hg nized for both 63C~Hg as indicated. The spectral pattern of CuHg is that expected for a zz0.53 in fig. l(b). The measured parallel and perpendicular positions of each observed transition of 63CuZn (j3CuCd and 63C~Hg are listed in table 1. The g tensors and the hyper- fine coupling tensors to the 63Cu nuclei of these molecules determined from the parallel and perpendicular positions of the Mi = &-$ components (the n.m.r. and MI = -3 component in the case of WuHg) resorting to the iteration process described earlier are compiled in table 2. Comparison of the spectra seen in fig.2-5 shows that the E.S.R. OF INTERMETALLIC MOLECULES 65C"Z" 1 2 3 4 5 6 kG FIG.3.-E.s.r. spectrum of an argon matrix containing Cu and Zn atoms. 65CuCd 1 63CuCd I I kG FIG.4.-E.s.r. spectrum of an argon matrix containing Cu and Cd atoms. anisotropy (the separation between the parallel and perpendicular positions within each hyperfine component) of CUM (M = Mg Zn Cd Hg) increases with the in- creasing atomic number of M. The spin Hamiltonian parameters determined and given in table 2 reveal that the hyperfine coupling tensor to the Cu nucleus is essentially isotropic in each case and the noted increase in the anisotropy is due entirely to in- crease in the anisotropy of the g tensor. The parallel and perpendicular resonance positions of all the components of 63C~M and TuM computed based upon the para- meters given in the table the known ratio of the magnetic moments of 63Cu and Vu and the iteration process based upon eqn (5) and (6) were found to be in excellent agreement with the observed values.The representative result obtained for CuCd is shown in table 5. There are two magnetic Cd nuclei of significant abundance lllCd (natural abund- P. H. KASAI AND D. MCLEOD JR 65CuHg(nmr) 6sCu Hg n I,! 63CuHg (nmr) 63CuHg n T-O 1 kG FIG.5.-E.s.r. spectrum of an argon matrix containing Cu and Hg atoms. ance = 13% I = 4,p = -0.5922pN) and 'I3Cd (natural abundance = 12% I = 3 p = -0.6195PN) and two magnetic Hg nuclei of significant abundance 199Hg (natural abundance = 17% I = 3 p = 0.4979 pN)and "'Hg (natural abundance = 13% Z = $ p = -0.5513pN).The curved arrows in fig. 4 indicate the signals that were .recognized as the perpendicular components of the "'Cd and 'I3Cd satellites of the MI = -3 component of 63C~Cd. The curved arrows in fig. 5 indicate the per- pendicular components of the 199Hg satellites of the MI = -$ component of 63C~Hg. The parallel components of these satellites and in the case of CuHg the satellites due to 201Hg were too weak to be detected. Because of the coexistence of extremely large coupling interactions with two nuclei the coupling tensors to the magnetic Cd and Hg nuclei in CUM could not be determined completely from the resonance positions of the perpendicular components only.However in each case the peak-to-peak intensity ratio of the satellite to the main peak is close to that expected from the natural abundance of the isotopes involved. It thus appears that the coupling tensors to the magnetic Cd and Hg nuclei are nearly isotropic. The magnitudes of the coupling interactions with the "'Cd and 199Hg nuclei assessed from the resonance positions of the MI = -3 main peak and its satellites assuming an isotropic tensor are included in table 2. Fig. 6-9 show the e.s.r. spectra observed from argon matrices containing Au and Mg Au and Zn Au and Cd and Au and Hg atoms respectively. In each trace the quartet signals due to isolated Au atoms are indicated by solid straight arrows. The hyperfine coupling constant of Au atoms isolated in an argon matrix has been deter- mined to be 3.1379 GHz,~ and corresponds to the situation CL = 0.33 in fig.I@). The second set of quartets observed in each trace with substantially reduced spacings was unique to the matrix containing Au and the specific M(M = Mg Zn Cd and Hg) and was assigned to the diatomic species AuM. The parallel and perpendicular signals of all the hyperfine components of AuM are well resolved as indicated. The experimentally determined resonance positions of AuM are listed in table 3. The g tensors and the hyperfine coupling tensors to the 19'Au nucleus determined from the parallel and perpendicular positions of the MI = &-% components using eqn (5) and (6) are given in table 4. The resonance positions of all the components of 197A~M E.S.R.OF INTERMETALLIC MOLECULES 1 3 4 5 kG FIG6.-E.s.r. spectrum of an argon matrix containing Au and Mg atoms. / J I1 1I 1 2 3 4 5 kG FIG.7.-E.s.r. spectrum of an argon matrix containing Au and Zn atoms. P. H. KASAI AND D. MCLEOD JR Au Cd (I)Jl I 1 3 4 5 kG FIG.8.-E.s.r. spectrum of an argon matrix containing Au and Cd atoms. c4+ 1 I1 1 2 3 kG 4 5 FIG.9.-E.s.r. spectrum of an argon matrix containing Au and Hg atoms. E.S.R. OF INTERMETALLIC MOLECULES TABLE3.-oBSERVED RESONANCE POSITIONS OF AUM (G) 3003 2561 2535.5 2084 ++{$ 3033 2589 2578.9 2165 3220 2984 2968.2 2665 3255 3022 3025.9 2778 3452 3484 3483.4 3439 3492 3533 3556.6 3591 3703 4060 4081.9 4406 3748 41 21 4171.3 4602 v/GHz 9.4216 9.4228 9.42196 9.4238 TABLE4.-sPIN HAMILTONIAN PARAMETERS OF AUM MgZn Cd Hgfree Au 2.0003 (7) 1.9771 (7) 1.9996 (7) 1.9734 (7) 1.9983 (7) 1.9593 (7) 1.9932 (7) 1.9121 (7) 2.0012 (1) 0.650 (2) 0.658 (2) 1.390 (2) 1.403 (2) 1.433 (2) 1.447 (2) 2.132 (2) 2.146 (2) 3.1379 (2) 4.72 (10) 6.45 (10) Coupling to lilCd or 199Hg.TABLE5.-cOMPARISON OF OBSERVED AND CALCULATED RESONANCE POSITIONS OF CuCd AND AuCd (G) MI 63C~Cd 65C~Cd AuCd obs. calc. obs. calc. obs. calc. 465 465 191 193 2535.5 2535.5 187 2578.9 2578.9 940 943 (masked) 452 2968.2 2968.1 965 965 (masked) 470 3025.9 3026.0 2520 2522 21 57 2157 3483.4 3483.5 2578 2579 2214 221 3 3556.6 3556.8 5203 5203 (masked) 5308 4081.9 408 1.9 5306 5306 5414 5414 4171.3 4171.3 computed based upon the parameters given in the table and the iteration process based upon eqn (5) and (6) were in excellent agreement with the observed values.The representative result obtained for AuCd is shown in table 5. In the cases of AuCd (fig. 8) and AuHg (fig. 9) many minor signals are seen that are attributable to the satellites due to "'Cd and '13Cd and to I9'Hg and 201Hg respectively. The curved arrows in fig. 8 and 9 indicate the perpendicular components of the lllCd and 199Hg satellites of the main MI = -3 component of AuM. The parallel components of the satellites were not observed. Again because of the ex- treme magnitudes of the hyperfine interactions with two nuclei the coupling tensor to the magnetic Cd or Hg nucleus could not be determined completely from the resonance positions of the perpendicular components only.The magnitudes of the coupling P. H. KASAI AND D. MCLEOD JR interactions with the lllCd and 199Hg nuclei assessed from the resonance positions of the main MI = -3 peak and its satellites assuming an isotropic tensor are given in table 4. DISCUSSION As revealed in tables 2 and 4,the coupling tensors to the Cu and Au nuclei are essen- tially isotropic for all the CUM and AuM molecules examined in the present study. The electronic configurations of Cu and Au atoms are 38'4s' and 5d1°6s1 respectively and those of Mg Zn Cd and Hg are either ns2or (n -l)dl0ns2. The semi-filled orbital of CUM (or AuM) of the present series is thus expected to be given essentially by an antibonding combination of the valance s orbitals of Cu (or Au) and M atoms.Hence @ = aPcu(4s) -bPMW or The spin density at the Cu or Au atom is then given by p = a2 -abS where S here represents the overlap integral between the two valence s orbitals. The coefficients a and b in eqn (7) may be determined from the secular determinant 1%' -ESI = 0. The usual LCAO-MO calculations were carried out equating the Coulomb integrals Zrito the atomic ionization potentials of the respective atoms and approximating the resonance integrals Ziby the Wolfsberg and Helmholtz expres- sion~~ The spin densities p thus determined for the present series of CUM and AuM assum- ing the fixed values of K = 2.0 and Sij = 0.2 are compared with the experimental values in fig.10. The experimental values were given by A/Aowhere A and A. are respectively the observed coupling constants of the diatomic molecules and the isolated atoms. A reasonable agreement obtained here is a strong substantiation to the envisaged orbital interaction. As in the case of AgM a more accurate LCAO description of the semi-filled orbital of CUM or AuM would be of the following form 0= aPcu(4s)-bP&) -cPM(nPz)* (8) It reflects the fact that the valence state in which the unpaired electron resides on Cu is neutral while the valence state in which the unpaired electron resides on M is polarized c; . . . M++ cu . . . M. (-) (+I The coupling constants Ao(lllCd) or Ao(199Hg) expected from a unit spin density in the valence ns orbital of Cd or Hg are not known.They may be estimated from the known coupling constants of the Ag and Au atoms and the Goudsmit relation (9)" AisoK z(Eip)'(I/p) (9) where 2 Ei,and p represent the atomic number the ionization potential and the magnetic moment of the nucleus respectively. The resulting estimated values are A,("'Cd) = 12.5 and Ao(199Hg) = 40.0 GHz. Thus for CuCd CuHg AuCd and E.S.R. OF INTERMETALLIC MOLECULES I I I 1 1 I I ' ' 5 6 7 8 9 10 11 i.p./ eV FIG.10.-Correlations between the observed spin densities (AIA,,)and the ionization potentials of the atom M for the CUM (solid circles) and AuM (open circles) series. The lines indicate the correla- tions predicted by the LCAO-MO calculations (see text).AuHg the spin densities (AIA,) in the valence s orbitals of both atoms were assessed as follows molecules vcu(4s) or VAuWI % v,M(ns)/ % CuCd 69 19 CuHg 81 7 AnCd 46 38 AuHg 68 16 We are thus led to surmise that the contribution of pM(np,)in eqn (8) is probably in the range 10-15 % for all CUM and AuM of the present series. It has been shown that for a paramagnetic molecule having a non-degenerate ground state the deviation of the g value along a principal axis " i" from that of a free electron (2.0023) can be given by l1 Agi = -2AC (OILi In>(nl Lilo> n+~ En -Eo where A is the spin-orbit coupling constant of the relevant atom Li is the usual angu- lar momentum operator and En -Eo is the energy separation between the ground P.H. KASAI AND D. MCLEOD JR state 10) and the excited state In). Thus for the ground state given by eqn (8) it follows immediately that Agl(=Ag,) = 0 and that Agl(=Ag = Agy) would be caused solely by the np part of M. One electron spin-orbit coupling constant of atom M may be assessed from the fine-structure intervals of the atom in its 3P(ns'np') 10 1 / A /cm" FIG.1 1.-Correlations between the observed Agi of CUM (solid circles) and AuM (open circles) and the one-electron spin-orbit coupling constant of the atom M. state.I2 Fig. 11 shows the values of Agi for CUM and AuM plotted against the spin- orbit coupling constant A of M. Linear dependencies were obtained for both the CUM and AuM series.As was the case with the AgM series studied earlier the straight line of the CUM series goes through the origin indicating the absence of the contribution of the Cu atom to Agl. The spin-orbit coupling constants of Cu and Au are 165 and 2544 cm-' respectively. The straight line of the AuM series inter- sects the ordinate at Agl x 100 = 2.3. We do not believe it represents the contribu- tion of Au to Agi. As discussed below if the semi-filled orbital of AuM contains the 6pz orbital of Au it is expected that A,,> Ai for the coupling tensor with the Au nucleus contrary to the observed result. The non-vanishing Agi at A = 0 for AuM may be ascribed to the increased interaction between the neighbouring Ar atoms and large valence orbitals of Au.'~'~ For the diatomic molecules CUM and AuM having the semi-filled orbitals of the form such as that given in eqn (8) the hyperfine coupling tensor to the Cu Au or M nucleus should be given by l4 All == Aiso + 2Adip Al= Aiso -Adip where Aiso is the isotropic coupling constant arising from the spin density in the valence s orbital and Adiprepresents the anisotropic part due to the spin density in the E.S.R.OF INTERMETALLIC MOLECULES valence p orbital of the nucleus under consideration. It thus follows that A, 2Al. For all the coupling tensors to the Cu and Au nuclei of CUM and AuM examined here A is slightly less than Al. It signifies the absence of the valencep orbital of Cu or Au in the semi-filled orbital and the presence of small negative spin density gener- ated through polarization in the p section of the filled bonding orbital.F. A. Cotton Accounts Chem. Res. 1969 2 240. * K. A. Gingerich J. Cryst. Growth 1971 9 31. W. A. Cooper G. A. Clarke and C. R. Hare J. Phys. Chem. 1972,76,2268. E. P. Kundig M. Moskovits and G. A. Ozin Angew. Chem. Int. Edn 1975 14 292. P. H. Kasai and D. McLeod Jr J. Phys. Chem. 1978 82 1554. ti P. H. Kasai E. B. Whipple and W. Weltner Jr J. Chem. Phys. 1966 44 2581. P. H. Kasai and D. McLeod Jr J. Chem. Phys. 1971 55 1566. See for example M. W. P. Strandberg Microwave Spectroscopy (Methuen London 1954) p. 11. M. Wolfsberg and L. Helmholz J. Chem. Phys. 1952 20 837. lo See for example H. Kopfermann Nuclear Moments (Academic Press New York 1958) pp.123-128. l1 M. H. L. Pryce Proc. Phys. Soc. A 1950 63 25. l2 C. E. Moore Nut. Bur. Stand. Circ. No. 467 1949 1; 1952,2; 1958,3. l3 F. J. Adrian J. Chem. Phys. 1960 32 972. l4 See for example P. W. Atkins and M. C. R. Symons The Structure of Inorganic Radicals (Elsevier Amsterdam 1967).

ISSN:0301-5696    

DOI:10.1039/FS9801400065

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Matrix Isolation Studies of Bimetallic Molecules of Fe-M(3d) Metals BY PEDROA. MONTANO Department of Physics West Virginia University Morgantown West Virginia 26506 U.S.A. Received 4th July 1979 A systematic study of the iron-M(3d) molecules (M = Mn Fe Co Ni Cu) was carried out using matrix isolation techniques in conjunction with Mossbauer spectroscopy. From the Mossbauer parameter of the isolated species it was possible to determine their electronic ground state. The bonding between the various transition metal atoms and iron varies appreciably. The strongest bond is associated with the molecules having an odd number of electrons. The isomer shift of FeNi is very close to that obtained for iron monomers in rare gas solids (weakest bond molecule). In contrast FeCo has a more positive isomer shift closer to that expected for 3d74s configuration at the iron atom.The study of diatomic molecules and higher aggregate species of metal atoms is of great interest in such areas as nucleation surface physics and chemistry the photo- graphic process and heterogeneous catalysis.' The diatomic molecules of the first transition period are very important in astronomy and high temperature processes. Recently rare-gas matrix isolation (RGMI) techniques have been used to study dia- C~MO,~ tomic molecules like Ni2,2 Ag2,3 CU~,~ FeMn5 and others. The correct identification of the species and their electronic ground state is one of the major tasks confronted by the spectroscopist. The use of the Mossbauer effect on 57Fe allows us to study a wide range of heteronuclear diatomic molecules of great scientific and technological significance.A systematic study was carried out of the electronic structure of the diatomic molecules of Fe-M (M = Mn Fe Co Ni Cu) using Moss- bauer spectroscopy in conjunction with RGMI. A summary of this work is presented here. EXPERIMENTAL The samples were prepared in a liquid-helium cryostat evacuated to a pressure < Torr. The 3d-metal and iron (90% enriched 57Fe) atomic beams were produced in two alumina crucibles contained in resistance-heated tantalum furnaces. The 3d-metal and iron atomic beams were co-deposited with a stream of rare gas onto a Be disc at ~4.2 K. The stream was introduced into the cryostat through a needle valve on the side.The metal deposition rates and concentrations were calculated using previously determined collection efficiencies. The rare-gas deposition rate was continuously monitored by the attenuation of the 6.3 keV X-ray of a 57Co:Pd source. An argon matrix was used for the study of the diatomic heteronuclear molecules since in this matrix the diatomic species are found in a quantity much larger than expected from simple random occupation of the lattice site. One important mechanism for the formation of aggregates is the fact that when the metal atom collides with the cold surface it does not stick immediately but moves about before finding an equilibrium position. During this process the condensing atom can react with other atoms forming diatomics or higher multimem6 A similar effect occurs in a neon MATRIX ISOLATION OF BIMETALLIC MOLECULES matrix where surface diffusion takes place more rapidly than in argon with the consequent appearance of dimers quadrimers and larger aggregates.Typical deposition times in these experiments ranged from 2 to 5 h depending on the degree of dilution. The metal deposi- tion rate was between 1 x 1014and 4 x 1014atoms cm-* s-I and the argon deposition rate was between 1.0~ 10l6and 3 x 1OI6 atoms cm-2 s-‘. The Mossbauer spectra were obtained with a constant acceleration spectrometer. An iron foil was used for calibration purposes and the zero velocity is given with respect to this absorber. All the spectra were analysed using a non-linear least square fitting programme and assuming Lorentzian line shapes.RESULTS AND DISCUSSION The major Mossbauer parameters used in the study of diatomic molecules are the isomer shift electric quadrupole interaction and the magnetic hyperfine interaction. Any differences in electron configuration between the source and the absorber in a Mossbauer experiment will influence the electron density at the 57Fe nucleus. The following expression is obtained for the electrostatic energy shift between a source and an absorber IS = ~7cze’A(rz)[[IY(0)~~ -/Y(0)1:] where IS is the isomer shift; A(r’) is the difference in the expectation value of the square of the nuclear radius between the excited and ground state (a characteristic quantity for a specific Mossbauer transition like 57Fe); IY(0)l; and IY(0)l; are the squares of the absolute values of the electron densities at the nucleus for the absorber and source respectively; e is the elementary charge and 2 is the atomic number.In fig. 1 a calibration curve is given for the electron density as a function of IS. The electron density calculations are from Dirac-Fock self-consistent calculations carried out by G. Shenoy. The squares are reference points from standard matrix isolation experiment^.^-" One observes in fig. 1 that a more negative isomer shift means a higher electron density at the 57Fe nucleus. The highest electron density is obtained for the [Ar]3d64s2Fe0 configuration. The second important parameter used in this study is the quadrupolar interaction or quadrupole splitting (QS).This interaction between the nuclear electric quadrupole moment Q and the gradient of the electric field is expressed for the excited state of 57Fe(I= 3)as” where q = VJe and V, is the electric field gradient (e.f.g.) principal component; q is the asymmetry parameter and is equal to zero for an axially symmetric e.f.g. The nuclear ground state of 57Fe has a spin of Z = 4and consequently is not split because there is no spectroscopic quadrupole moment in nuclei with Z = 3. The excited state is split into two doubly degenerate substates. In a standard Mossbauer experiment (single line source) one would observe a quadrupole doublet (fig. 2). The intensities of the two lines are equal for random absorbers (neglecting vibrational anisotropies).For oriented samples the two lines have different intensities depending on the angle between the e.f.g. principal axis and the y-ray direction. For an axially symmetric field finding which substate is higher in energy I&-> or I&+> will allow the determination of the sign of q. The magnitude and sign of the e.f.g. will be used to characterize the electronic ground state of the diatomic molecules. The magnetic hyperfine structure arises from the interaction of the nuclear magnetic dipole moment with a magnetic field external or due to the atom’s own P. A. MONTANO IS/mm s-1 FIG.1 .-Electron density at the nucleus IY(R,)IZin units of l/d taken from Dirac-Fock-Slater SC calculations by G. Shenoy for a finite nucleus (R,= 8.005 x a.u.) as a function of isomer shift.The squares are reference points. IS = uAIY(R)Jz,where u = 0.183 a; mm s-'. r -2 0 2 mm s-1 FIG.2.-Mossbauer spectrum of iron monomers and dimers in argon at 4.2 K 0.1 atom % iron. electrons. In 57Fe the effect of this interaction is to split the I = 3excited state into four substates and the ground state I = 0 into two substates.'l The allowed gamma transitions between the substates of I = 3 and those of the ground state are found according to the magnetic dipole selection rules AI = 1 AMI = 0 1. In the absence of an e.f.g. at the nucleus the Mossbauer spectrum of a magnetically ordered absorber will show a six-level spectrum. Two conditions are necessary to 82 MATRIX ISOLATION OF BIMETALLIC MOLECULES observe a resolved hyperfine magnetic splitting in a Mossbauer experiment T/hco < 1 and x-'/w < 1 where LL) is a frequency parameter characteristic of the hyperfine interaction w N" A/h; z is the electronic relaxation time; and r is the linewidth of the nuclear level.In the presence of rapid electronic spin relaxation the hyperfine field at the nucleus averages to zero and the Mossbauer spectrum collapses to one line in a paramagnetic substance. At low temperatures and in the presence of a large external magnetic field one can produce a large magnetic hyperfine field (m.h.f.) at the nucleus. Such a technique was used for measuring the m.h.f. for iron monomers in rare gas solids. Values ranging from 70 T for xenon12 and 83 T for argon13 to 90 T in nitro- gen14 were obtained (including an external 3 T field).The reduction of the m.h.f. from its free atom value of 110 TI5 is due to crystal field effects with the largest reduc- tion in xenon and the weakest in nitrogen. A similar study was carried out for iron dimers in argon with the application of an external magnetic field it was possible to determine the sign of the e.f.g. and the magnitude of the m.h.f. In what follows the techniques described will be applied to identify the electronic structure of the diatomic molecules of iron-M (M = Mn Fe Co Ni Cu). HOMONUCLEAR MOLECULE Fe At iron atomic concentrations ranging from 0.1 to 2% in argon one observed the appearance of two narrow lines with an IS = -0.14 & 0.02mm s-l and QS = 4.05 st 0.04 mm s-l (mm s-I = 0.48 x eV) fig.2. These values are matrix inde- pendent;* however in xenon the dimers are difficult to observe by Mossbauer spectroscopy due to a very low Debye-Waller factor. Fe molecules have been de- tected in the gas phase before by mass spectrometry;16 and electronic absorption bands in the visible spectral region were observed by de Vore et aL2in argon matrices. These experimental facts suggest that that the Fez dimer is a bound state with a constant internuclear separation. The IS of Fe is more negative than that of metallic iron indicating a higher elec- tron density at the nucleus. One can observe in fig. 1 that the IS of Fez is far from the value expected for the 3d74satomic configuration. It seems more appropriate to consider that in the iron dimer the bonding is mainly 4s-4s weak and that the atomic- like configuration is partially retained (the electronic configuration between 3d64s2 and 3d64s).A delocalization of the 4s electrons by ~50% is necessary in order to reduce the electron density from that of the iron monomer (Fe'). The Mossbauer spectrum of the dimer in an external magnetic field shows the presence of a magnetic hyperfine field at the 57Fe nucleus. If the Fez molecule has an electronic angular momentum the magnetic field associated with this angular mo- mentum interacts strongly with the nuclear magnetic moment giving rise to magnetic hyperfine interactions comparable in size with those found in atoms. This is the case for Fe molecules where a m.h.f. of 66.0 T was measured (including a 3 T ex-ternal field).13 The sign of the e.f.g.was found to be negative q< 0. From these two experimental results one can infer the electronic ground state of Fez. The homo- nuclear diatomic molecule has a symmetry Dmh. For this symmetry the splitting of the d-orbitals is given in table 1. In this table the symmetry type and the e.f.g. produced by the atomic orbitals of the iron atom are given as well. One observes that only the d3z2-r*gives the right sign and magnitude €or the e.f.g. This means that Fe is a sigma molecule with a large spin at the iron atoms (large m.h.f.). Since there is no orbital contribution the m.h.f. is due to the Fermi contact interaction. The numerical value obtained for the Fermi term at the iron nucleus for Fe2+(3d6) S = 2 is -55.0 T;I7 this value is not far from our measured m.h.f.(however the P. A. MONTANO 83 IS of Fez is not that of Fe2+). If one uses the suggested electronic configuration for Fe2,18 '~(o,'.u"s~a,s:,o=.~,) one obtains S = 3per atom. With a value of S = 3 the expected m.h.f. is -41.3 T. The difference between the two values can be accounted for by a contribution from the 4s electron [lY4,,(0)12 -IY4sr(0)l]. All the experimental evidence indicates that the ground state of Fe is 'C (note that the interatomic distance given in the reference for Fe is not in agreement with the experimental measurements). HETERONUCLEAR DIATOMIC MOLECULES FeMn FeMn Mossbauer measurements of Fe Mn (1 :1) mixtures in argon show the presence of a doublet with an IS = 0.24 & 0.03 mm s-l and a QS = 1.93 & 0.03 mm s-~.~ This doublet was identified as due to FeMn from concentration dependence studies.The heteronuclear diatomic molecule has a symmetry of Cmo. From table 1 it can be seen that only d,, d,, gives the correct magnitude of the e.f.g. (about half the value obtained for Fe,). From the known ground-state electronic configura- tion of Fe and Mn (IC,o,'n~o,'B~a~2),19 one can suggest the following configuration for FeMn 4~, 02n402846*2n*.In this molecule iron and manganese are antiferro- magnetically coupled. The more positive IS for FeMn as compared with Fe is interpreted as meaning a stronger bonding. If one considers the bond orders of Fe, Mn and FeMn FeMn falls between the values of Fe and Mn2.The IS for FeMn suggests that the atomic configuration of iron is not 3d74s. From the difference in the IS between the iron monomer and FeMn one can infer that this corresponds to about one 4s electron less at the iron atom in FeMn. This suggests that the 4s electrons are actively participating in the bonding between the two atoms. FeCo Diatomic molecules of iron and cobalt are very difficult to observe due to the high diffusivity of Co in argon. Higher metal aggregates are easily formed. Moss-bauer measurements at a concentration of 0.1 atom % metal and Fe Co ratio of 1.2 show the presence of iron monomers FeCo molecules and a doublet with IS = 0.55 & 0.05 mm s-l and QS = 3.60 & 0.05 mm s-'. At higher concentrations this doublet tends to disappear due to the formation of larger multimers.The sign of the e.f.g. was determined from the angular dependence of the amplitudes of the quadrupolar doublet (due to partial orientation of the molecules during deposition with e.f.g. principal axis parallel to the substrate). Consequently the only symmetry type possible is C (table 1). The smaller value of the QS 3.60 mm s-l compared TABLESYMMETRY TYPE AND ELECTRIC FIELD GRADIENT OF d ORBITALS IN DmhAND C, SYMMETRIES symmetry type electric orbit a1 Dca h Cmv field gradient MATRIX ISOLATION OF BIMETALLIC MOLECULES with the value of Fe, 4.05 mm s-l can be produced by expansion of the 3d-orbitals ((Y-~)~~~~ < (r-3)Fe2).The large positive IS also indicates a stronger bond and participation of d-electrons in the bonding.Using the known electronic ground state of Fe the following ground state is suggested for FeCo ‘C,a2n464a26*2a*7r*2. This ground state is consistent with the symmetry type and gives the correct magnitude and sign of the e.f.g. This configuration indicates that the coupling between iron and cobalt is ferromagnetic. The observed IS for the FeCo molecule is larger than any of the heteronuclear diatomics listed in table 2 indicating less total ,+electron density TABLE 2.-sYMMETRY TYPE AND GROUND STATE OF Fe-M DIATOMIC MOLECULES at the 57Fe nucleus. This can be accomplished by an increase in the 3d electron population and/or a decrease in the 4s contribution. FeNi This heteronuclear diatomic molecule was observed in runs at an atomic % concentration of metal of w0.7and Fe Ni ratio of 1 :1 fig.3. The doublet identified . Fe Ni -0 . .:+. . Feo I. -99 -3 -2 -1 0 1 2 mm s-1 FIG. 3.-Mossbauer spectrum of iron-nickel in solid argon. The positions of Fez FeO and FeNi are indicated in the figure. P. A. MONTANO as FeNi has an IS = -0.54 & 0.05 mm s-l and QS = 1.95 -J-=0.05 mm s-~.~'The IS is not far from the values for the iron monomer in an argon matrix (-0.75 mm s-l) indicating that the atomic configuration of iron in FeNi is close to 3d64s2 and that consequently it is a very weakly bound molecule. According to the value of the QS the symmetry type should be 7c only d,, d,, give the right magnitude for the e.f.g.From the electronic ground states of Fe and Ni2 one can suggest a '7~ ground state for FeNi (possible Configuration a27r464a26*4a *n*). FeCu The heteronuclear diatomic molecule FeCu was identified from several measure- ments in a concentration range from 0.18 to 4 atom % metal in argon.21 A doublet with an IS = 0.46 & 0.07 mm s-' and QS = 1.63 & 0.07 mm s-l is observed for FeCu. The ground state can be determined from the electronic ground states of Fe and CuZ2 molecules. The QS of FeCu is approximately one half the value for Fe,; consequently the ground state must be of the symmetry type n. The sug- gested electronic ground state of FeCu consistent with this symmetry is 2~(n4a264a2-d4a*,n*). The QS in FeCu is smaller than half the value of Fe, indicating that an expansion of the radial wave functions of the 3d orbitals is taking place (Y-')~~~> (Y-~)~~~.similar to FeCo. The IS of FeCu is also more positive than Fe, FeNi FeMn and closer to the value of FeCo. This suggests a strong bond between Fe and Cu comparable to FeCo. This value of the IS also suggests some participation of the d-orbitals in the bonding. In the FeCu molecule the magnetic moment seems to be localized at the iron atom. In summary one observes that the bond strength for the Fe-3d (M) molecules is 6 5 4 3 2 €0 E 3; -1 c( -2 -3 -4 -5 -6 FIG.4.-IS plotted against total number of 3d and 4s electrons for the molecules of Fe-M (M = Mn Fe Co Ni Cu).MATRIX ISOLATION OF BIMETALLIC MOLECULES FeCo (strongest) >FeCu > FeMn > FeFe > FeNi. A summary of the ground states for the diatomic molecules is given in table 2. An interesting result of the Mossbauer measurements of the diatomic molecules is that for an odd total number of electrons the IS is positive and consequently leads to stronger bonds than in molecules with an even number of electrons. In fig. 4 this result is schematically represented by plotting IS against total number of 3d and 4s electrons in the molecules. The author acknowledges the helpful collaboration of his students Maria Angela Talarico and William Dyson and of Drs. G. Shenoy and S. Ruby of the Argonne National Laboratory. This work was supported by the U.S.National Science Foundation. R. van Hardeveld and F. Hartog Adv. Catalysis 1972 22 75. T. C. deVore A. Ewing H. F. Fransen and V. Calder Chem. Phys. Letters 1975,35 78. G. A. Ozin Appl. Spectr. 1976 30 573. W. Klotzbucher G. A. Ozin J. G. Norman and H. J. Kolari Inorg. Chetn. 1977 16 2871. W. Dyson and P. A. Montano J. Amer. Chem. Soc. 1978 100 7439. J. E. Hulse and M. Moskovitz Surface Sci. 1976 57 125. 0.C. Kistner and A. W. Sunyar Phys. Rev. Letters 1960 4,412. T. K. McNab H. Micklitz and P. H. Barrett Phys. Rev. €3 1971 4 3787. H. Micklitz and P. H. Barrett Phys. Rev. Letters 1972 28 1547. lo H. Micklitz and F. J. Litterst Phys. Rev. Letters 1974 33 480. 11 G. K. Wertheim Applications of the Mossbauer Efect in Chemistry and Solid State Physics Technical Report No.50 IAEC 1966 p. 1. l2 P. A. Montano P. H. Barrett and Z. Shanfield Solid State Comm. 1974 15 1675. l3 P. A. Montano P. H. Barrett and Z. Shanfield J. Chem. Phys. 1975 64 2896. l4 P. H. Barrett and P. A. Montano J.C.S. Faraday ZZ,1977 73 378. W. J. Childs and L. S. Goodman Phys. Rev. 1966 48 74. l6 Sin-Shong Lin and Arthur Kant J. Phys. Chem. 1969 73 2450. l7 R. E. Watson and A. J. Freeman Phys. Rev. 1961 123 2927. l8 A. B. Anderson Phys. Rev. B 1977 16 900. l9 R. K. Nesbet Phys. Rev. A 1964 135 460. 'O P. A. Montano J. Appl. Phys. 1978 49 1561. 21 P. A. Montano and M. A. Talarico J. Appl. Phys. 1979 50 2405. 22 P. Joyes and M. Leleyter J. Phys. B 1973 6 150.

ISSN:0301-5696    

DOI:10.1039/FS9801400079

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Optical and Vibrational Studies of Silver Molecules and Microcrystallites Prepared by Matrix Aggregation and Gas Aggregation Techniques SCHULZE BY WILFRED AND HITOSHIABE Fritz-Haber-Institut der Max-Planck-Gesellschaft Dahlem Berlin Germany Received 29th August 1979 Silver molecules Ag, n < 10 have been prepared by matrix aggregation techniques under various experimental conditions i.e. metal concentration speed of layer growth condensation temperature and type of gas (Xe-Ne). The optical absorption spectra of these molecules show several distinct absorption bands within the range 200-550 nm and in the corresponding laser Raman spectra several lines below 220 cm-' are present. The spectral features of silver molecules Ag, n < 6 could be identified by comparison with experimental data from the gas phase and/or by comparing the relative intensities of pairs of bands observed for matrices formed under different experimental conditions allowing the remaining bands to be assigned to larger silver clusters.Silver microcrystallites with a mean diameter 2R 2 10 A could be easily prepared by a gas aggre- gation technique. In contrast to the results of other authors we observed a shift of the surface plasmon from 3800 A for particles with 2R % 100 A to 3600 8 for particles with 2R z 10 A i.e. a shift to higher energies. The reported increase in the halfwidth with decreasing cluster size from 0.2 to 1 eV and more could also not be found. An increase of only 0.1 eV could be measured. The importance of the properties of small metal particles with respect to many problems in physics and physical chemistry has often been In this paper we demonstrate the advantages of the different techniques which we use for particle preparation namely the matrix aggregation technique for the preparation of mole- cules mainl~,~-~ and the gas aggregation technique for the production of metal microcrystallites with mean diameters as small as 10 A.'v9 Furthermore we will show the optical and vibrational (laser Raman spectroscopy) properties of silver molecules in dependence on size and the optical properties of silver microcrystallites > 10A in diameter.EXPERIMENTAL The experimental arrangements used in matrix aggregation studies have been widely rep~rted,~*~J~ therefore a brief summary of the most important elements of these techniques will be sufficient.The silver atomic beam generated from a resistance heated tantalum Knudsen cell was simultaneously condensed with an excess of noble gas on a helium cooled target a sapphire plate for the optical studies and a copper target for the laser Raman studies. The silver condensation rate measured by a quartz microbalance and the gas admission rate measured by using a capillary could be independently adjusted allowing an adjustment of the metal concentration as well as the speed of layer growth. A spectrophotometric arrangement for the optical studies allowed us to measure the absorbance A directly A = log (IsigiIrer) where Irefis the intensity of light passing through a part of the matrix target onto which SILVER AGGREGATES only matrix gas was condensed whereas I,, is the residual intensity of monochromatic light passing through that part of the target containing metal and matrix gas.For the Raman measurements either an Ar or Kr laser could be used as excitation source. The laser beam entering the vacuum chamber from the bottom hits the matrix target at an angle of x 20". The point of incidence covered the image of the entrance slit of the Raman spectrometer (Varian Cary 82). The apparatus used for gas aggregation studies consists mainly of two components the gas aggregation cell for particle preparation and the above-mentioned liquid helium cooled cryostat (TK22.0 K) suitable for optical measurements. Silver evaporation was achieved by an electrically heated tantalum boat (T = 1000-1450 "C)located inside the gas aggre- gation cell.The evaporating silver atoms are effectively cooled by collisions with inert gas atoms (the gas pressure is in the range 0.1-10 Torr) and microcrystallites are formed in a region of sufficientlyhigh supersaturation. The crystallites are then transported through an aperture in this cell into the high vacuum chamber in which the cryostat is situated. For the optical studies these particles are embedded in a matrix at a filling factor c <0.1 %. A sample consisting of an amorphous carbon film 20 A thick supported by a copper grid is mounted on a rotary motion drive allowing the microcrystallites to be sampled by intro-ducing the sample into the crystallite stream.A characterisation of the size distribution as well as the shape of the crystallites was achieved by evaluating the corresponding electron micrographs. RESULTS AND DISCUSSION OPTICAL AND VIBRATIONAL PROPERTIES OF SILVER MOLECULES PREPARED BY THE MATRIX AGGREGATION TECHNIQUE The aggregation process for silver atoms has been systematically studied as a function of condensation conditions i.e. metal concentration condensation tempera- ture speed of deposition and the nature of the gas (Xe-Ne). The influence of the condensation parameters is roughly the same for the different gases if the layers are prepared at temperatures of the same ratio with respect to the matrix triple point temperature." The only obvious difference is that the tendency of silver atoms to form aggregates increases along the series Xe to Ne the difference being greatest between Xe and Kr.The influence of increasing the metal concentration from 0.2 to 3 % on the resulting matrix spectra of Ar and Ne condensed at 16 and 3 K respectively (w0.2Tt,,) in the u.v.-vis. range is shown in fig. l(a) and (b). Several distinct bands in the spectral range 5500-2000 A appear and grow successively. Besides these distinct features a broad background absorption extending from w2500 8 to longer wavelengths with a maximum at 3400 A for Ar and 3300 A for Ne is present. As already shown for Kr matrices the yield of higher clusters (Ag, n 3 3) can be considerably increased if the condensation is performed at w0.3 of the matrix triple-point temperat~re.~ Fig.2 showing the spectrum of a Ne matrix at an Ag atomic concentration of w1 %condensed at 4.2 K'(0.25Tt,,),clearly demonstrates this effect e.g. by comparison with the corresponding spectrum in fig. l(b). The speed of deposition was of only minor influence on the aggregation thus allowing the deposition of a defined quantity of silver in shorter periods without affecting the relative yield of aggregates very much. Annealing matrix layers with a defined metal content at a temperature of w0.4Tt, of the corresponding gas results in the growth of larger aggregates at the expense of the smaller ones. Fig. 3 shows the energetic position of all absorption bands found in the inert gas W. SCHULZE AND H. ABE 5000 4000 5000 4000 3000 2000 h /ic 3000 A /ii FIG.1 .-Spectra of Ag atom and molecules in (a) Ar and (6) Ne matrices condensed at (a) 16 and (b) 3 K (~0.2.T~~) in the u.v.-vis.range. The silver concentration c is as follows (i) 0.3 (ii) 0.5 (iii) 0.8 (iv) 1.5 (v) 2.5 and (vi) 4.0 %. Apart from molecular bands a broad plasmon band is to be seen. matrices in the energy range 2-6 eV. The spectral features of silver molecules Ag, n < 64could be identified by comparison with experimental data from the gas phase and/or by comparing the pairs of bands observed for matrices under different experi- mental conditions. Many unassigned bands which are either too weak or overlap with other bands remain however. A definite assignment of the remaining bands to silver molecules Ag, n > 6 seems to be impossible so long as one does not have an independent measure of the mass.It is worthwhile to mention that generally the trend in energy shift of the aggre- gate bands is the same as for atoms i.e. within the series Xe-Ar a blue shift is found and compared with Ar a red shift is found for Ne matrices. The rate of formation of silver molecules can be made sufficiently high for one to study the vibrational properties of these molecules by laser Raman spectroscopy. This could be achieved by increasing the deposition time of the matrix layer from z 1 min sufficient for optical absorption studies to several h. As already reportedlo a band at 120.5 cm-l has been found as well as the Ag vibrational mode at 194 cm-'. These matrix layers were prepared at low temperatures at which as has been men- tioned above only dimeric and trimeric silver molecules can be grown.We therefore assigned the band at 120.5 cm-' to Ag,. The existence of only one strong Raman SILVER AGGREGATES 5000 4 000 3000 h /i FIG.2.-Spectrum of a Ne matrix with an Ag atomic concentration % 1% condensed at 4.2K (0.25Tt,,). E/eV FIG.3.-Peak position of absorption bands of silver molecules in inert gas matrices (Xe-Ne) in the range 2.0-6.0 eV. The spectral features of silver molecules Ag, n d 6 could be identified either by comparison with experimental data from the gas phase and/or by comparing the relative intensities of pairs of bands observed for matrices formed under different experimental condition^.^ active mode due to Ag gives a strong indication that Ag3 belongs to the point group Dmh(j.e.,Ag is linear) excluding the other two possibilities D3h and C2", i.e.the equi- lateral and isosceles triangles. Because of insufficient time for these measurements apart from the presence of an impurity exhuded from the oven material giving extremely strong Raman spectra in the spectral range of interest we could not obtain a definite assignment of all the other Raman bands found in these studies for larger clusters Ag, n > 3. The additional bands centred at z 150 110 and 75 cm-l respectively,1° which could be grown in a Kr matrix at 25 K result however from at least two overlapping bands. W. SCHULZE AND H. ABE Fig. 4(a) shows three Raman spectra of different atomic metal concentrations ; the growth of these additional bands with increasing metal concentration can be clearly seen.In fig. 4(b) spectrum B is shown at a resolution of 2 and 1 cm-l Ly-J 'A 200 100 2 00 100 cm-1 cm-1 FIG.4.-Laser Raman spectra below 240 cm-' of silver molecules in krypton matrices condensed at 21 K. (a)All three spectra were recorded with a resolution of 2cm-' and the concentration of silver increases from A to C 0.2 0.6 and 2.0 % respectively. (b) Spectra of the matrix B at a resolution of 1 and 2 cm-' respectively. The upper spectrum in this figure shows the spectrum of the annealed matrix (50 K). respectively demonstrating that to overcome the overlap problem a still higher reso- lution would be desirable.Even so the bands at 75 105 151 and 203 cm-I can be assigned to silver molecules Ag,, n > 3 and the bands at 156 108 and 77 cm-' which can be grown at the expense of all other bands by annealing the matrix layer at x0.4 of the corresponding matrix triple-point temperature are independent and can therefore be assigned to still larger clusters. Because of the difficulties in characterizing the size distribution of the metal clus- ters prepared by matrix techniques we began a different series of experiments by preparing molecules and/or clusters in the gas phase thus allowing a characterisation of the size distribution in the gas phase either by mass spectrometry or by electron micrography before embedding these particles in a matrix.We began these experi- ments making use of the gas aggregation technique first. SILVER AGGREGATES OPTICAL PROPERTIES OF SILVER MICROCRYSTALS Crystallite formation using the gas aggregation technique has been studied in detail as a function of the experimental parameters ie. inert gas pressure within the cell silver vapour pressure and hole diameter.9 In accordance with literature data7q8 the following results were obtained. Increasing the inert gas pressure (0.1-10Torr) as well as the silver vapour pressure at a fixed hole diameter (3 mm) results in the growth of larger silver particles. Changing these parameters however did not allow one to reduce the mean diameter of the crystals below 50 A; furthermore larger particles are 3.50 3.40 3.30 3.20 0 50 100 150 200 2 &ass 12i > a \ k 0.1 0.2 0.3 0.4 (I/ZR)/nrn-l FIG.6.4~) Peak position of the plasmon band of silver microcrystallites as a function of crystallite size.(b)Full halfwidth of the plasmon band as a function of crystallite size. The dashed line shows results reported in the 1iterat~re.I~ (2) ' 1000% ' R c (1) (2) d ~-I r'I, 4 I a, d 0 .-c 20 a 0 100 200 0 100 200 0 0 100 200 0 100 200 diameter 2 R /% FIG.5.-Optical absorption spectra of silver microcrystallites in argon matrices produced by the gas aggregation technique as a function of crystallite size the corresponding micrographs and size distribution histograms. (--) Number distribution; (--) mass distribution.[Toface page 92 W. SCHULZE AND H. ABE often formed by coagulation of smaller ones. Spherical crystals with mean diameters 200 32R/A > 10 could be grown by changing the aperture diameter from 3 to 16 mm at an inert gas pressure of 0.5 Torr and silver pressure of 1 .O Torr approximately. By so doing the gas exchange velocity varies drastically thus allowing one greatly to change the temperature of the nucleation and growth zones. Fig. 5 shows selected micrographs the evaluated number and mass distributions as well as the corresponding optical spectra. This figure clearly demonstrates that the peak position of the plasmon bands shifts to higher energies with decreasing cluster size. Furthermore under experimental conditions allowing the preparation of the smallest crystals a considerable number of atoms are also matrix-isolated.The atoms give rise to the well-known triplet at ~3200-30008 and the additional two absorption bands at z 3400 8 are also due to Ag atoms isolated in a less stable matrix site.I2 Fig. 6(a) shows the peak position and fig. 6(b)the halfwidth of the plasmon band as a function of size. We have evaluated these dependences using the best mass distribution available as the absorption contribution of each crystallite is proportional to its mass. The shift of the plasmon band to higher energies within the size range 200 32R/A 310from 3.3 to 3.5 eV is in agreement with various quantum mechanical calculation^^^ and with experimental res~lts.'~ The increase in the halfwidth from 0.1 to 0.15 eV within the given size range is in disagreement with experimental results where an increase in halfwidth from 0.16 eV for 2R z 100 8 to 1.0 eV for 2R z 10 A was reported and also disagrees with all existing theoretical res~lts.'~-l~ The dashed line represents the theoretical and experimental size dependence after Genzel et aZ.14 The physical reason for the discrepancy between theoretical predictions and our experimental results is not clear to us at the moment.B. Meyer Ber. Bunsenges. phys. Chem. 1978 82,24. R. C. Baetzold J. Chem. Phys. 1971 55 4355 4363. A. B. Anderson J. Chem. Phys. 1975 62,1185. W. Schulze H. U. Becker and H. Abe Chem. Phys. 1978 35,177. M. Moskovits and J. E. Hulse J.Chem. Phys. 1978 66,3988; 1977 67,4271. G. A. Ozin and H. Huber Iriorg. Chem. 1978 17,155. S. Yatsuya S. Kasukabe and R. Uyeda Japan J. Appl. Phys. 1973 12,1675. C. G.Grandqvist and R. A. Buhrmann J. Appl. Phys. 1976 47 2200. H. Abe W. Schulze and B. Tesche to be published; similar measurements are also being per- formed by T. Welker and T. R. Martin (Stuttgart). lo W. Schulze H. U. Becker R. Minkwitz and K. Manzel Chem. Phys. Letters 1978 55 59. W. Schulze H. U. Becker and H. Abe Ber. Bunsenges. phys. Chent. 1978 82,138. W. Schulze D. M. Kolb and H. Gerischer J.C.S. Fuvuduy II 1975 71,1763. l3 A. A. Lushnikov V. V. Masimenko and A. J. Simonov 2.Phys. 1977 B27,321. l4 L.Genzel T. P. Martin and V. Kreibig Z. Phys. 1975 B21,339. l5 A. Kawabata and R. Kubo J. Phys. SOC. Japan 1966 21,1765. l6 R. Ruppin Phys. Rev. 1975 B11,2871. M. A. Smithard Solid State Comm. 1973 13,153. l8 R. Kubo J. Phys. (Paris) Colloq. 1977 2 69.

ISSN:0301-5696    

DOI:10.1039/FS9801400087

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Application of Magnetic Circular Dichroism Spectroscopy to the Identification of Small Matrix-isolated Metal Clusters and the Assignment of their Electronic Spectra BY ROGERGRINTER ARMSTRONG, STEPHEN UPALIA. JAYASOORIYA JUNE MCCOMBIE AND JON P. SPRINGALL DAVID NORRIS School of Chemical Sciences University of East Anglia Norwich NR4 7TJ Received 1st August 1979 Those principles of magnetic circular dichroism which are particularly relevant to the assignment of the spectra of matrix-isolated metal atoms and small molecules formed from such atoms are briefly reviewed. These principles are illustrated through a discussion of the m.c.d. spectra of Ag and Cu isolated in argon and methane matrices. The analysis of the spectra of the silver species presents few problems and for Ag2 and Ag the results strongly confirm the assignments of species and electronic states which have been made hitherto.For the copper spectra the situation is much less satisfactory and further work is indicated. The study of matrix-isolated metal atoms and small aggregates of such atoms is a subject of much current re~earch.'-~ For many such interesting species the matrix phase would appear to be the only one in which they can be prepared and studied. The assignment of the electronic spectra of these molecules is an important task for the attribution of a particular spectral band to a specific molecule M, will only be really certain when detailed assignments of the bands have been made. Furthermore the process of assignment will also help to reveal the details of the electronic and geo- metrical structure of the molecules.Assignment of the electronic spectra of matrix-isolated molecules particularly those which have not been observed in other media presents difficulties over and above those normally encountered in the assignment of the spectra of molecules in solution or the gas phase. For example if the symmetry of the matrix site in which the guest species finds itself is low then the degeneracy of electronic states may be lifted and two or more bands observed where only one is expected. A well-known example of this phenomenon is found in the spectra of matrix-isolated copper silver and gold atoms for which the absorption band corresponding to the 2S1/2 -+2P3/z transition is invariably split into at least two It should be pointed out that there is not universal agreement on this subject and the formation of exciplexes has been suggested as the origin of the splitting.' Also in some cases silver in xenon for = example further splittings which cannot be explained by site-symmetry effects are also observed.6 A further cause of difficulty in assignment arises from the possibility of different matrix sites.If multiple sites exist in the host then the absorption bands of the guest may be correspondingly multiplied. Such effects are commonly seen in the spectra of metal atoms in matrices and can frequently be identified as such by the fact that R. GRINTER et al. they disappear on annealing.',' However there is no guarantee that all such bands will be removed by an annealing process.The purpose of the present paper is to illustrate the use of magnetic circular di- chroism (m.c.d.) spectroscopy in the assignment of the electronic spectral bands of matrix-isolated species. This subject has received rather little attention to date8v9 though the potential of m.c.d. in other fields has been amply demonstrated." EXPERI M E NTAL Matrices were prepared on a sapphire window within a matrix-isolation/m.c.d. cryostat which has been described elsewhere." Where they were not formed during deposition molecular species were produced within the matrix by thermal annealing or by irradiation at the wavelengths of the strong atomic absorption bands. Absorption spectra were measured with a Cary 14 spectrophotometer and m.c.d.spectra with a Cary 61 spectro-polarimeter. RESULTS AND DISCUSSION In using m.c.d. spectroscopy to assign electronic spectral bands and identify species two aspects of the technique play particularly important roles. This can best be illustrated by the following simple energy-level diagrams fig. 1 and 2. B=O I II I II FIG.1.-A diagrammatic illustration of the origin of a Faraday A-term. In fig. 1 we consider an atomic transition from a 'S state to a 'P which in the absence of a magnetic field would appear as a single line or band. In the presence of a magnetic field the triple degeneracy of the excited state is lifted and transitions to the states having ML = -1 and Mt= +1 are allowed for right circularly polarised (r.c.p.) light and left circularly polarised (1.c.p.) light respectively if the direction ofthe M.C.D.OF METAL CLUSTERS I FIG.2.-A diagrammatic illustration of the origin of a Faraday C-term. field is the same as that of the propagation of the light. Thus a plot of AA (absorbance of 1.c.p. light -absorbance of r.c.p. light) against wavelength gives rise to a sigmoid curve the properties of which depend critically upon the quantitative details of the excited state. This type of spectral feature is known as a Faraday A-term.I2 In fig. 2 the reversed situation a 'P+ 'S transition is illustrated. Again im- position of the magnetic field lifts the degeneracy of the 'Pstate and a sigmoid curve results from a plot of AA against the wavelength or frequency of the light.How-ever in this case the two lobes of the curve will not be equal in magnitude. At low temperatures in particular the Boltzmann population of the lowest level will exceed that of the highest and the population difference will grow as the temperature decreases until at very low temperatures (below 5 K in practice) the single band remaining will cease to grow and the signal is said to be saturated. Before the onset of saturation there is a temperature region in which the area of the band is linearly dependent upon the inverse temperature. The observation of this type of band which is known as a Faraday C-term,12 is an unambiguous indication of ground-state degeneracy for the absorbing species. For atomic systems and to a lesser extent for molecules analysis of the effects described above is a powerful way of obtaining quantitative estimates of ground or excited state g-factors spin-orbit coupling etc.In the present paper we shall be more concerned to show that qualitative observation of A and C term behaviour is quite sufficient for assignment purposes in many cases. We begin our discussion with some aspects of the spectra of Ag in argon matrices. Ag IN ARGON MATRICES Several laboratories have reported absorption spectra attributed to Ag (n < 10) in noble-gas matrice~.~-~*~ To date attribution of the bands observed has been based largely upon kinetic considerations as far as the species responsible for the band is R. GRINTER etal. concerned and upon theoretical work with respect to assignment.The work of Ozin and Huber' is the most relevant for comparison with our results since they have also used argon matrices. Table 1 is taken from their paper.2 TABLEBAND ASSIGNMENTS FOR Ag and Ag ISOLATED IN ARGON MATRICES~ d/nm Ag2 assignment Alnm Ag3 assignment 227 ? 245 ? 261 1264 1z; 3mu (244)" 387/412 (257.5/262)" 12; 3Iz; 440 HOMO -+LUMO a The figures in parentheses are the wavelengths recorded in the present work where they differ from those of Ozin and Huber.2 Fig. 3 shows our m.c.d. and absorption spectra of an argon matrix containing molecular silver species in the region between 235 and 270 nm. Though our wave- length data for the major atomic transitions agree exactly with those of Ozin and Huber we find differences of up to 3.5 nm for the peak positions of the bands in the spectral region shown in fig.3. These discrepancies which we do not regard as im- 2L0 250 260 wave Ie ng t h / nm FIG.3.-Absorption and m.c.d. spectra of silver in an argon matrix. The m.c.d. spectra were measured at a field of 6 T. Absorption spectrum at 10 K (.-.--.); m.c.d. spectrum at 10.0 (-) and 20.0 K (--). portant in the present context may arise from the lower temperatures of our measure- ments. The most notable feature of the m.c.d. spectra in fig. 3 is the temperature dependence of AA in the 246 nm region as opposed to its virtual absence between 254 and 270 nm. The temperature dependence does not have the most simple C-term form a subject to which we return but it is strong evidence for the assignment of the 244 nm absorption band to an electronic transition of the paramagnetic Ag molecule.M.C.D. OF METAL CLUSTERS Conversely the effective absence of temperature dependence suggests that the bands at 258 and 262 nm be assigned to a diamagnetic species in particular Ag,. There is a wavelength shift with changing temperature in the 254-270 nm region and this we believe is responsible for the slight change in the amplitude of the negative lobe of the m.c.d. at ~258 nm. However this is not a C-term phenomenon and we can further substantiate Ozin and Huber's assignment of the two bands in this spectral region. First we observe that each of the two partially-resolved absorption bands gives rise to an m.c.d.signal of sigmoid form crossing the AA = 0 line very close to the position of the absorption maximum. This is clear evidence of excited state de- generacy for both absorption bands. The two sigmoid features are of extremely similar form and their relative magnitudes parallel those of the corresponding ab- sorption bands. It appears therefore that we are dealing here with two transitions of a very similar nature but at slightly different energies. The suggestion2 that these two bands be assigned to 'C -f 'nutransitions at different matrix sites is therefore in excellent agreement with our m.c.d. evidence. We can further confirm the above conclusion by considering the signs and mag- nitudes of the m.c.d. features in fig. 3.The sign of the m.c.d. A-terms i.e. long-wavelength lobe negative short-wavelength lobe positive is quite in accord with the 'nuassignment for the excited state. The magnitude of the A-terms as measured by their first moment divided by the zero'th moment of the corresponding absorp- tion band 9,-,,13 can be shown to be equal to minus one half of the g-factor for the orbital magnetic moment of the excited state. Because the bands are not fully re- solved only approximate g-values can be obtained but the values of 0.14 for the long- wavelength band and 0.17 for the other provide good supporting evidence for the Illu assignment. When we examine the longer-wavelength regions of the spectrum we find clear absorption bands at 440 and 388 nm and some evidence of a weak shoulder around 412 nm.All these figures are in good agreement with Ozin and Huber's observations.2 However we can find no detectable m.c.d. under these bands even with a magnetic field of 6 T. For Ag, at least this negative result is to be expected if the 388/412 nm bands are assigned as 'ZC,+-f 'Xi. Theoretical analysis shows that for such a transi- tion the only m.c.d. to be expected is the so-called B-term12 which arises from the mix- ing of the %,+ excited state with other states notably the In, by the magnetic field. Since the latter lies some 13 000 cm-l away from the former this mixing will be small and a very weak B-term will result in agreement with our observations. In the case of the 440 nm band attributed to Ag the questions of molecular geo- metry and electronic structure both arise.Both theoretical14" and experimental 14' work indicates that a linear or nearly linear geometry is preferred for this molecule and if this is indeed the case then theoretical considerations suggest that the m.c.d. of the HOMO 3 LUMO band is likely to be weak and featureless as we observe. In that case the band at 245 nm might well correspond to a transition to an excited state of II symmetry. A state under spin-orbit coupling would give rise to ,113/,and components. Temperature dependence is expected on account of the 2X ground state degeneracy and a characteristic sigmoid m.c.d. signal should be found under the ,II3/ band but not under the 2111/2. The possibility of B-terms is always present. Though the temperature dependence is not as simple as that shown in fig.5 probably on account of an underlying background of temperature-dependent plasmon ab~orption,~ the m.c.d. spectra are quite consistent with the assignment of the 244 nm absorption band to a 211state of the Ag molecule. The m.c.d. in this spectral region is clearly not R. GRINTER etal. simple and a more detailed investigation may well reveal the positions of both the 2n3,2 components of the 211state. and 2111,2 CU IN ARGON AND METHANE MATRICES The position with regard to the spectra of copper species is by no means as clear cut as with silver. Nevertheless a number of interesting points may be made. One of the problems which makes the certain identification of copper aggregates difficult is the large number of atomic bands which occui in the ultraviolet between 200 and 300 nm.' Furthermore a large number of formally forbidden atomic transitions lie in this regionI5 and it appears to us quite possible that perturbations due to matrix- site asymmetry or adjacent copper atoms might make some of these transitions weakly allowed.We therefore begin our analysis of copper-cluster spectra by comparing the atomic transitions which occur in the ultraviolet (particularly those which arise from transitions to the configuration 3d9 4s' 4p') in methane and argon matrices. Fig. 4 and table 2 summarize our results on this subject. The m.c.d. spectra 220 230 240 250 wavelength 1 nm FIG.4.-M.c.d. spectra of copper atoms in argon (--) and methane (-) matrices at 4.5 K TABLE2.-wAVELENGTHS AND ASSIGNMENTS FOR THE ABSORPTION AND M.C.D.SPECTRA OF COPPER ATOMS ISOLATEDINARGON AND METHANE MATRICES IN THE SPECTRALREGION 21 5-250 nm assignmentUpb A(argon)/nm abs.b m.c.d.C I(methane)/nm abs.b m.c.d.C 4p3,2 238.3 239.5 242 243 4p112 (233.6 229.3 234.5 227 ? ? 238 222.5 4D1,~ 220.7 221.5 ? 21 8.5 2p1/2 218.0 218.5 21 6 216.5 a Electron configuration 3d94s' 4p' in all cases. Wavelengths of absorption maxima and assign- ments as given by Moskovits and Hulse.' = Wavelengths of the m.c.d. maxima see fig. 4. M.C.D. OF METAL CLUSTERS shown were readily obtained from matrices whose absorption spectra showed only very weak features which could be ascribed to the atomic transitions detailed in table 2; all are T-dependent.The power of a differential technique is clear in these spectra and taking the assignments of Moskovits and Hulse' as our basis we suggest the following additional assignments relating the spectra in argon to those in methane. The strong sigmoid features in the region of 240 nm are interpreted as two oppo- sitely signed C-terms on account of their marked temperature dependence (not shown for the sake of clarity) and in analogy with the m.c.d. spectra of other atomic transi- tions in copper silver and gold.16 The similarity in the form of these bands for the two different matrices suggests that they be assigned to the same transitions; the 4P3,2 according to Moskovits and Hu1se.l Thus the second component of this transi- tion is found to lie at 238 nm in a methane matrix.The remarkable resemblance of the two m.c.d. spectra in this region in particular the very similar separation of the bands (800 cm-l in argon 850 cm-l in methane) raises the important question of the origin of this splitting. It is rather difficult to believe that a crystal field effect of two different matrices could give rise to so similar a splitting and the exciplex explanation presents a similar problem. Further detailed analysis of the m.c.d. spectra of these and other atomic transitions should throw some light on these problems and is in progress.16 As we move towards higher energies further comparisons between the two m.c.d. spectra may be made and assignments suggested as in table 2.Apart from the dis- crete bands two inflexions are visible in the spectra at 235 nm in methane and 242 nm in argon. These features might be due to atomic transitions or to molecules of which we otherwise see no sign in this spectral region. The absence of bands definitely ascribable to molecules is surprising since Moskovits and Hulse find absorption bands which they attribute to Cu2 and Cu in this region for both argon and methane matrices. However we do find very strong evidence for both paramagnetic and diamagnetic molecules in the spectral region 340-600 nm fig. 5. The distorted sigmoid curve I I I I I I 2 2t c Q L I I I I I I 350 400 450 500 550 600 wavelength / nm FIG.5.-Absorption and m.c.d. spectra of copper in a methane matrix.Upper absorption spectrum at 4.5 K; lower m.c.d. spectra at 4.5 (-) 6.6 (--) 10.0 (.-*--.-. ) and 15.0 K (-.*--..-). Note that the wavelength scale changes at 430 nm. R. GRINTER et al. centred z 390 nm clearly belongs to a paramagnetic species with a degenerate excited state and a linear Cu molecule seems most likely although a triangular shape is also possible. The m.c.d. beneath the absorption bands peaking at 510 and 550 nm is quite weak and independent of temperature clearly suggesting the presence of diamagnetic species. The two bands could be due to different transitions of the same species a matrix site splitting or to two different species Cu2 and Cu for example. There is no evidence for a band due to Cu in this region but if such a molecule is linear and the absorption corresponds to a 2C-+2C transition then the m.c.d.might be very weak though mixing with the nearly 211state should be quite significant and moderately strong B-terms could result. More surprising is the absence of a strong signal due to Cu3 in methane at 227 nm (fig. 4) in view of Mosko- vits and Hulse’s assignment. In summary the assignment of the spectra of polymeric copper species presents considerable problems and further work is required on almost all aspects of the subject. In this task more theoretical work would be of great assistance to the experimentalist. Predictions concerning molecular shapes energies of electronic states and the effects of spin-orbit coupling would be most useful not only for copper but also for silver molecules and radicals.We are grateful to the S.R.C. for their support of this work by provision of apparatus liquid helium and Research Fellowships and Studentships. The m.c.d./m.i. cryostat was constructed with the aid of a grant from the Paul Instrument Fund of the Royal Society. M. Moskovits and J. E. Hulse J. Chem. Phys. 1977 67 4271. ’ G. A. Ozin and H. Huber Znorg. Chem. 1978 17 155. S. A. Mitchell and G. A. Ozin J. Amer. Chem. SOC., 1978 100 6776. W. Schulze H. U. Becker and H. Abe Chem. Phys. 1978 35 177. F. Forstmann D. M. Kolb D. Leutloff and W. Schulze J. Chem. Phys. 1977 66,2806. F. Forstmann D. M. Kolb and W. Schulze J. Chem. Phys. 1976 64 2552. L. Andrews and G. C. Pimentel J. Chem. Phys. 1967 47 2905.* I. N. Douglas R. Grinter and A. J. Thomson Mol. Phys. 1974 28 1377. R. L. Mowery J. C. Miller E. R. Krausz P. N. Schatz S. M. Jacobs and L. Andrews J. Chem. Phys. 1979,70,3920. lo P. J. Stephens Ann. Rev. Phys. Chem. 1974 25 201. T. J. Barton R. Grinter and A. J. Thomson J.C.S. Dalton 1978 608. l2 P. N. Schatz and A. J. McCaffery Quart. Reu. Chem. SOC., 1969,23,552; 1970,24,329. l3 P. J. Stephens J. Chem. Phys. 1970,52,3489. l4 (a) R. C. Baetzold J. Chenz. Phys. 1971 55 4363; R. C. Baetzold and E. Mack J. Chem. Phys. 1975 62 1513; (6) W. Schulze H. U. Becker R. Minkwitz and K. Manzel Chem. Phys. Letters 1978 55 59. l5 C. E. Moore Atomic Energy Levels (United States National Bureau of Standards 1958). l6 R. Grinter S. Armstrong U. A. Jayasooriya J. McCombie D. Norris and J. P. Springall unpublished observations.

ISSN:0301-5696    

DOI:10.1039/FS9801400094

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Gas Phase Production and Chemistry of Transition Metal Atoms and Clusters from Polynuclear Metal Carbonyls BYJOHNS. WINN Department of Chemistry University of California Berkeley California 94720 U.S.A. and Materials and Molecular Research Division Lawrence Berkeley Laboratory Berkeley California 94720 U.S.A. Received 7th August 1979 The volatility of metal carbonyls and a unique dissociative energy transfer mechanism are ex- ploited to generate gas-phase transition metals at low temperatures. A flowing afterglow of a meta- stable rare gas produces a flame with metal carbonyls which when spectrally analysed is found to consist of metal atom lines. Analysis of the kinetics of this reaction shows it to be bimolecular. Detailed analysis of the spectra shows that the rate of production of a given excited metal state is predicted by a statistical model in which the metal-ligand bonds rupture simultaneously along a radial reaction coordinate.There is little spin differentiation among the metal states produced and previ- ously unknown transitions from low-lying quintet states of Ni are observed. The possible extensions of this method tometal atom excited metal atom and metal cluster chemistry are discussed in terms of the electronic excitations produced by metastable atom energy transfer. As a class of molecules transition metal diatoms and larger clusters suffer from the lack of readily available data on even their most rudimentary chemical and physical properties. In large measure this lack of data is due to the large atomic binding energy of the bulk metals making studies of even the atoms' physico-chemical properties a difficult exercise for the high-temperature chemist.Our laboratory has taken the approach that the collection of these data is not likely to come from high temperature studies alone. Consequently we have under- taken a programme aimed at exploiting the low-temperature volatility of transition metal carbonyls and polynuclear metal carbonyls using these compounds as starting materials for gas phase transition metal atom and cluster chemistry and finding physical methods of breaking metal-ligand bonds selectively. In this article some of our recent results towards this goal are described. Even though certain results are preliminary to more extensive investigations they are included here to give examples of the types of measurements these techniques allow.These results are due to the continuing efforts of Dr. D. C. Hartman and Messrs. W. E. Hollingsworth and D. V. HorBk. EXPERIMENTAL The experimental method is a flowing afterglow technique with either chemiluminescence or mass spectrometric detection. The flowing afterglow uses a fast flow of a rare gas (typically Ar) at a pressure near 1 Torr (1 Torr 21 133 Pa). A d.c. hollow cathode discharge excites a small fraction of the rare gas atoms to metastable electronic states which are neither radia- tively nor collisionally relaxed during several cm of flow. The energies carried by these meta- stables is considerable ranging from He* 2lS at 20.62eV (1989 kJ mol-') to Ar* 4s[3/2]g3P2 J.S. WINN 103 at 11.55 eV (1114 kJ mol-l). The carbonyls and other reagents can be injected into the flow at various points depending on the type of experiment. In certain cases the carbonyl is injected upstream of the discharge and is subjected to the discharge; however the usual mode of operation is to inject the carbonyl downstream of the discharge where the only inter- action will be with metastables. The chemiluminescence afterglow apparatus has been described previously.'Y2 Briefly chemiluminescence is collected at right angles to the flow and spectrally analysed over the 200-820 nm region by a 1.5 m grating monochromator equipped with a photomultiplier which is cooled and operated in the pulse counting mode.Monochromator control data collection and signal averaging are under microcomputer control. The mass spectrometer detector is a commercial quadrupole mass filter with a particle multiplier detector. The mass spectrometer is mounted in a bakeable liquid-nitrogen trap- ped chamber with a base pressure of lo-' Torr. The flow is sampled through a 100 pm pinhole. Ionisation is by conventional electron impact and the particle multiplier signal is recorded as a current. The instrument has unit mass resolution over the range 1-350 a.m.u. As with the chemiluminescence apparatus provisions are made for a variety of injection ports along the flow and on either side of the discharge. RESULTS Some of our results on the chemiluminescent flames produced by metal carbonyl- metastable rare gas collisions have appeared in the literature.lV2 We describe them briefly here.With the compounds Ni(CO), Co2(C0), Fe(CO), Mn2(CO)lo W(CO)6 and Mo(CO), the only features in the chemiluminescence flames were atomic metal emissions. With Ni(C0)4 and Fe(CO), this general characteristic of the spec- tra was found to hold for Ar Ne and He metastables. For the other compounds only Ar was investigated but presumably the effects is general. Co,(CO) and Mn2- (CO), were investigated only briefly. They were introduced into the interaction region from a heated metal crucible located below the interaction region. These compounds were entrained in a second flow of pure Ar and thereby transported to the interaction region. This method proved unsatisfactory for extensive study due to thermal decomposition of the compounds in the metal crucible.Detailed studies of the Ni(CO) and Fe(CO) flames were made to determine the molecularity of the dissociation process and to determine the dynamical processes responsible for the ligand-metal dissociation. By varying the carbonyl concentration at constant metastable densities and by varying the metastable concentration at con- stant carbonyl densities it was determined that the emission features were first-order in each reagent. The elementary process M(CO) + Rg* -+M* + nCO + Rg (1) was established. Analysis of the intensities of the various atomic features and their assignments to transitions between known atomic energy levels was straightforward but not without surprise.The greater the energy in the metastable the higher was the highest excited state of the metal observed to luminesce. The energy of this highest state plus the energy required to break the metal-ligand bonds was essentially equal to the energy of the metastable implying that very little ligand excitation accompanied excitation of the highest observed metal state. Even though the carbonyl is a spin singlet and the metastable a triplet no spin differentiation among the various spin manifolds of the metals was observed. For example the most intense feature of the Ar* + Ni(CO) flame was the z5DG -+ a3F4 transition at 3881.86 A. This transition connects the lowest level of the quintet manifold in Ni to the ground-state term.Even more sur- TRANSITION METAL ATOMS FROM METAL CARBONYLS prising is the fact that this and two other easily assigned Ni lines had never been observed previously in either absorption or emission measurements on pure Ni. By measuring the intensities of the various atomic lines and from knowledge of the oscillator strengths for these lines one can deduce the relative rates of formation of the various excited states of the metal atoms. The relevant expressions are based on the assumption of no collisional deactivation of the emitting species (an assumption which was found not to be true for certain low-lying states with exceptionally long radiative lifetimes) and the assumption of a steady state. For a state with energy E which is observed to emit to a lower state with a measured intensity (in counts s-I) denoted by I one has the rate of depletion of E, being proportional to the intensity sum of all lines radiating from this level i.e.rate of depletion cc 2 IUjcc N,,Z A, I i where Nuis the steady-state concentration and AUjis the Einstein coefficient for spon- taneous emission. For any single line Nu K I/Aul,and it follows that rate of depletion cc I -. (3) (7;) Plots of the logarithm of the rate against E showed a monotonic decline to the rate as E increased. A typical plot is shown in fig. 1 for the Ar* + Ni(C0)4 flame. Col-\ \ \ \ \ l r i 1 i Y i I l I 25 26 27 28 29 30 31 32 33 34 35 36 37 E,/ 103~~-1 FIG.1.-Logarithm of the rate of production of various Ni excited states (in arbitrary units) plotted against the energy of the excited state for the Ni(C0)4 + Ar* reaction.Quintet states are denoted by * triplet states by x and singlet states by +. The dashed line is the prediction of a statistical dissociation model in which all degrees of freedom are allowed while the solid line which reproduces all the carbonyl + rare-gas combinations investigated is the result of a restricted statistical model described in the text. lisional deactivation* of certain low-lying quintet states is evident in this figure. The solid and dashed lines in fig. 1 are the predictions of two statistical models.2 In these models it is assumed that the rate of producticn of a given Ni state is propor- tional to the density of states of the products at that energy.The dashed line uses a * Krenos3 has observed Fe emission in a molecular beam experiment with Ar* and Fe(CO)5. His data agree superbly with ours except for fine-structure levels of low-lying septet states with long radiative lifetimes where our data show intramultiplet collisional relaxation. J. S. WINN 105 density of states function appropriate for the three-dimensional translation rotation and vibration of the CO fragments. This assumption does not predict the observed Fe* or Ni* kinetic data. The solid line is based on a restricted density of states wherein CO motion is restricted to vibrational and one-dimensional translation motion with no rotational motion. This assumption models the physical picture of a simul- taneous metal-ligand bond rupture along a radial reaction coordinate.Of the many choices of density of states functions only this one is able to reproduce the data for all rare gas +metal carbonyl combinations. We note that this technique produces appreciable populations of excited metal atoms many of long radiative lifetime (such as z5D00,for which we have measured2 a radiative lifetime of 3 x s) or of optical metastability (such as z5G;) and that the full range of spin multiplicities is generated. This suggests that lifetime-separated reaction chemistry can be studied by using the flow rate to separate spatially excited states of short radiative lifetime from long radiative lifetime states optically meta- stable states and ground states. We have generated Ni and Fe from Ar* and have looked for chemiluminescence from metal hydrides oxides and chlorides by injecting H2 N20 or C12 at appropriate points along the flow to allow reactions with either ground or excited metal atoms but we have found no chemiluminescence attributable to the species sought.An instructive comment about these experiments is in order. The NiCl spectrum which is incompletely anaiysed and ~nassigned,~ falls in the 400-470 nm region. If Ni(CO) and C12 are injected into an Ar* flow at the same point directly in the ob- servation region a flame is observed which is visually different in colour from that produced by Ni(C0)4 alone. The spectrum of this flame from 3900 to 4700 A is shown in fig. 2. The bands evident in this spectrum do not however appear to be m .cI .-C 3 4 .-m C c .-C I I I I I I I I I 3900 4700 wavelength /8 FIG.2.-Spectrum of a Ni(C0)4 + C12 i-Ar* flame at 1 A resolution over the region 3900-6700 A.The majority of this spectrum is due to CO bands discussed in the text. attributable to NiCl. They are instead bands of the CO Asundi system a’ 3C+ -+ a 311, with very high d. The source of CO is evidently reaction of C1 with Ni(C0)4 directly. Ar* excites many band systems5 of CO and this source of interference must be kept in mind. Fortunately the problem can be eliminated by changing the metastable from Ar* to Ne*. The energy of Ne* (16.6 eV) is above the ionisation potential of CO and the only reactions of Ne* with CO are Penning ionisation TRANSITION METAL ATOMS FROM METAL CARBONYLS u' forming CO+ and excitive Penning ionisation forming CO+ (A2Hi,= 0) a nearly resonant process.Thus with Ne* the only interfering CO emissions2Jj are the (0,O) and (0 1) bands of the A -+ X comet tail system of CO+. While chemiluminescent detection is very sensitive it is of course useless for detec- ting products formed in states which do not luminesce. Laser-induced fluorescence is a sensitive probe of ground-state products but only when the spectroscopy of these products is well understood. For most simple transition diatomics this is not the case. We have therefore turned to mass spectroscopic product identification. The mass spectra of metal carbonyls show extensive fragmentati~n.~ By monitoring these fragments as the carbonyl is alternately subjected and not subjected to meta- stables we have verified that eqn (1) is the only observable reaction i.e.no fragments such as Fe(CO), etc. are produced. DISCUSSION The results of these experiments give the following qualitative picture of metal carbonyl dissociation by metastable rare gases. In a single collision the carbonyl is energized to an excited state (or states) which is totally metal-ligand repulsive. Relaxation of this excited state leads to electronically excited metal atoms and to ligands with at most only a small amount of vibrational excitation. Easy spin-orbit mixing leads to metal product states which are not spin differentiated. It is important to understand the detailed energy transfer dynamics of the simple metal carbonyl dissociation before considering polynuclear metal carbonyl dissociation possibilities.The simplest view of energy transfer from a metastable rare gas to a target molecule considers energy exchange between two electrons. The target molecule transfers an electron from a high energy occupied molecular orbital to the orbital vacancy of the rare gas while the rare gas transfers its excited electron to a previously unoccupied target MO. If this is a continuum MO then Penning ionisation results; it is the rare gas electron which is ionised in this simplified picture. If the MO is a bound-state MO then the target is left in an excited but un-ionised state. To apply this mechanism to a metal carbonyl and to compare this mechanism to photolysis and electron im- pact dissociation it is instructive to understand the salient features of a metal car- bonyl's electronic structure.We use Ni(CO) as an example. Calculations of the electronic structure of Ni(CO) show the effects of a-donation and n-back-bonding on the electronic environment of the metal quite clearly.8 From a population analysis one can consider Ni(CO) to have the primitive configuration Ni[3d8.134po.88] 2n0*44])4 (CO[l n450'~~~ which should be contrasted with ground state Ni (a3F,3d84s2) the excited singlet valence state of Ni (a's 3d1') and the ground state of CO (ln4 5a2). The Ni 4p population is due to 50 donation the CO 271.population is due to 3d back-bonding and the highest occupied MOs are the 9t (3d and 2n in character) and the 2e (3d in charac- ter).The 9t orbital is not only the highest in energy it also has the greatest spatial extent of the high energy orbitals. On collision with a metastable an electron from the 9t2orbital will fill the rare gas vacancy. The rare gas electron will populate a metal centred orbital due to the positive effective charge of the metal. Note that even in the ground state Ni has a formal charge of +1 and that the Ni 4s orbital is empty. Whether the rare gas elec- tron populates the 4s or the 3d Ni vacancy is difficult to decide but the result would be the same. One would have effectively taken a strongly bonding electron from the n-back-bonding framework and promoted it via electron exchange to a previously J.S. WINN 107 unoccupied and antibonding orbital. The metal-ligand bonding would become repulsive and the Ni atom would be left in states derived from either the 3d9 4p or the 3ds 4s 4p excited configurations. These are the states we observe to luminesce. Charge flow during dissociation from the Ni 4p orbital back to the CO 50 orbital would produce the 3d8 4s2 ground configuration of Ni or the low-lying 3d9 4s con-figuration. Contrast this mechanism with the electron rearrangement produced by photolysis. The strongly-allowed optical excitations' are either d to d metal excitations or charge transfer excitations in which charge flows from the metal to the ligand. This flow of charge is in an opposite sense to that proposed in the metastable energy transfer mechanism above and photolysis even multiple photon dissociation," leads to quite different products.The sequential degradation of carbonyls observed in electron impact mass spectra which have been shown to be due to a series of unimolecular fragmentations,l' is also explained by noting that ionisation of a 9t2 or 2e electron would disrupt the x-back-bonding but not nearly to the extent that promotion of the bonding electron to a totally anti-bonding orbital would. These arguments suggest that metastable energy transfer will be a useful technique for generating metal clusters of a particular size from polynuclear metal compounds. Other techniques for producing gas-phase metal clusters in particular adiabatic expansion of a supersonic molecular beam suffer from the need for high temperature sources (which are not attacked by the molten metal of interest) and from the diffi- culty of observing the properties of only one metal cluster size in a distribution of many sizes.While it is clear from the known values of bond energies that pyrolysis of a poly- nuclear metal carbonyl will rupture the metal-metal bonds before rupturing the metal -1igand bonds it is important to remember that metastable energy transfer produces dissociation from an excited electronic state and that our observations on the mono- metal carbonyls indicate an excited-state metal configuration which is not only inetal- ligand repulsive but also of a type which favours metal-metal bonding. Metastable energy transfer populates just those configurations known to be important to bulk metal bonding as exemplified by the Brewer-Engel l2 theory.At the present time we have inconclusive but encouraging mass spectral evidence that one can produce for instance Fe from Fe2(C0)' or Fe from Fe,(CO),,. Other research groups are applying this technique along the same lines discussed here,13 and the main experimental difficulty seems to be in avoiding thermal dissociation of the carbonyl before it enters the afterglow flow rather than an iiiherent failure of the energy transfer mechanism. We look forward to the rapid development of gas phase metal cluster chemistry once the appropriate compounds (and techniques for manipulating them) are discovered. The author acknowledges partial support for this research from an Alfred P.Sloan Research Fellowship and partial support from the U.S. Department of Energy Office of Basic Energy Sciences Division of Chemical Sciences. D. C. Hartman and J. S. Winn J. Chem. Phys. 1978 68,2990. D. C. Hartman W. E. Hollingsworth and J. S. Winn J. Chem. Phys. 1980,72 833. J. Krenos personal communication. K. P. Huber and G. Herzberg Constants of Diatomic Molecules (Van Nostrand-Reinhold New York 1979) p. 462. D. H. Stedman and D. W. Setser J. Chem. Phys. 1970 52 3957. W. E. Hollingsworth and J. S. Winn unpublished data. TRANSITION METAL ATOMS FROM METAL CARBONYLS M. R. Litzow and T. R. Spalding Mass Spectrometry of Inorganic and Organornetallic Com-pounds (Elsevier New York 1973) chap. 11 p. 471. E. J. Baerends and P. Ros Mol. Phys. 1975 30 1735. M. Dartinguenave U. Dartinguenave and H. B. Gray Bidl. SOC.chim. France 1969 12,4223. lo Z. Karny R. Naaman and R. N. Zare Chem. Phys. Letters 1978,59,33. l1 R. E. Winters and J. H. Collins J. Phys. Chem. 1966 70,2057. l2 L. Brewer in Phase Stability in Metals and Alloys ed. P. Rudman J. Stringer and R. I. Jaffee (McGraw Hill New York 1967) p. 39. l3 E. Schumacher personal communication.

ISSN:0301-5696    

DOI:10.1039/FS9801400102

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Experimental and Predicted Stability of Diatomic Metals and Metallic Clusters BY KARLA. GINGERICH Department of Chemistry Texas A & M University College Station Texas 77843 U.S.A. Received 17th October 1979 The experimental bond energies of atomic and polyatomic metals and intermetallic compounds are reviewed and discussed in terms of various empirical models of bonding such as the Pauling model of a polar single bond the valence bond approach for certain multiply bonded intermetallic molecules and the atomic cell model. Illustrations emphasize recent Knudsen effusion mass spectro- metric results. For ligand-free diatomic metals a maximum dissociation energy of 640 & 40 kJ mol-I is predicted. 1. INTRODUCTION The investigation of diatomic metals and intermetallic compounds has been a major objective of our research for more than a decade.Our early emphasis was to test the applicability of the Pauling model of a polar single bond' to diatomic inter- metallic compounds. The determination of the dissociation energies of previously unknown homonuclear diatomic molecules permitted application of the Pauling model to the intermetallic molecules of these elements since these values are needed in the models' application. In our experimental work on testing the Pauling model we emphasized diatomic intermetallic compounds for which a sizeable ionic contribution to the dissociation energy was expected because of a large electronegativity difference in the component atoms e.g. rare-earth-gold intermetallic compounds.These compounds also in- cluded the largest measured bond energies ~340 kJ mol-' between two metal atoms known at that time. Other large dissociation energy values were measured for AlAu and UAu and for the gold compounds with the metalloids boron and silicon.2 The applicability of the Pauling model was soon extended to polyatomic intermetallic molecules.39 The Pauling model is not expected to apply to multiply bonded molecules. Mul-tiple bond formation between two different metal atoms and thus bond energies of possibly more than 420 kJ mol-' was according to a suggestion by Bre~er,~ expected for molecules formed between a group IV transition metal e.g. zirconium and platinum. The corresponding condensed intermetallic compounds are known to be very stable because in Brewer's view the transfer of one or more of the paired d-electrons of the platinum to the vacant d-orbitals of the group IV transition metal will make more valence electrons available on both atoms and thus permit strong mul- tiple bonding.An experimental test to confirm the expected large dissociation energies was difficult particularly because of the tremendously reduced activity of the group 1V metal in the corresponding condensed alloy and because of the high chemical reactivity of the group IV transition metals. Thus at first we confirmed Brewer's expectation qualitatively through a spark source mass spectrometric st~dy,~ the results DIATOMIC METALS AND METALLIC CLUSTERS of which also led us to the choice of the Ti + Rh system for measuring the first inter- metallic multiply bonded transition metal molecule under equilibrium conditions.6 Next the high temperature mass spectrometric equilibrium studies were extended to combinations between a rare earth metal or an actinide metal and a platinum metal.7 The results of these investigations could be interpreted in terms of an empirical valence bond approach for multiply bonded platinum metal intermetallic molecules with Group 111 to Group V transition metals or lanthanide or actinide metals.8 This valence bond approach accounts for many molecules with assumed double and triple bonds.Quadrupole bonding appears partially achievable but also restricted by the directional requirements of orbital overlap in the diatomic m~lecules.~ As a conse-quence of this limitation the maximum bond energy between any two ligand-free metal atoms could be predicted as 670 j-84 kJ mol-'.' Most compounds of platinum however cannot be quantitatively treated for lack of a suitable valence state.The range of validity and the weakness of the model have recently been critically discussed.1o The model has guided us in finding a score of additional very stable intermetallic molecules. It can account for or be used to predict the dissociation energies of z 200 diatomic intermetallic molecules with multiple bonds. The atomic cell model by Miedema and collaborators" for quantitative description of heats of formation of solid alloys and heats of mixing of liquid alloys has recently been extended to diatomic molecules.12 Miedema will discuss the model in detail at the present Symposium.lo3 The atomic cell model appears to be more general than either the Pauling model of a polar single bond or the valence bond model for certain multiply bonded intermetallic molecules; it encompasses both of them.It may be ap- plied with certain reservations to all metallic molecules including metalloids. Certain refinements of the model will however be needed and we are therefore engaged in an experimental programme to determine the bond energies of homonuclear and hetero- nuclear metallic molecules in order to provide a broader basis for the refinements of the various empirical models. Parallel to our efforts in measuring and understanding the bond energies of diatomic metals we have been extending our work on small metallic and intermetallic clusters with emphasis on group IV and group V homonuclear molecules and on small intermetallic clusters involving these elements.This work has the purpose of con- tributing to the understanding of certain catalylic processes in which such small metal clusters are believed to play an important role. In the following discussion some comments concerning the Knudsen cell mass spectrometric method will be made since it has been by far the most commonly used experimental technique for determining bond energies of diatomic metals and metallic clusters. Then a brief review and a tabulation of the dissociation energies of diatomic metals and the atomization energies of metal clusters will be given.Using these experimental data the various empirical models of calculating bond energies will be discussed and compared. 2. EXPERIMENTAL The bond energies of diatomic metals and small metal molecules have been determined mainly by the well established Knudsen cell mass spectrometric method. In our laboratory the measurements have been performed with a Nuclide Corporation 12-90 HT mass spectro- meter that resembles the one described by Chupka and 1ngh~am.l~ The mass spectrometer and the experimental technique have been described previously.'" The Knudsen cell containing the sample is heated by radiation from a non-inductively wound thoriated tungsten spiral resistance heater. This heater permits attainment of K.A. GINGERICH temperatures up to % 3000 K. Temperatures are measured by focusing a calibrated optical pyrometer on a black-body hole in the bottom of the Knudsen cell. Measured temperatures are corrected for the optical absorption of the viewing window and deflection prism used. The optical pyrometer can be calibrated at the freezing point of gold under in situ experimental conditions. Identification of the ions with the corresponding molecular precursors is usually done by measurement of their mass-to-charge ratio isotopic abundance distribution shutter profile appearance potential and ionization efficiency. The measured ion currents are correlated with the corresponding partial pressure^.'^*'^ From these partial pressures reaction enthalpies are obtained by the third law method using the equation AH' = -RTln Kp -T[A(G; -H;)/TJ and when a large enough temperature range can be covered in addition by the second-law method from the relation dln K,/d(l/T) = -AH,/R.Further details of the instrument experimental procedure and data evaluation are given elsewhereY7*l4 and in pertinent reviews 15*16 and in the individual papers quoted. 3. REVIEW OF EXPERIMENTAL BOND ENERGIES The dissociation or atomization energies of ligand-free gaseous diatomic metals and small metal clusters are presented in tables 1-4. Unless stated otherwise the compilation by Gurvich et a1.I' has been taken as a basis for the values listed. These authors present the latest critical evaluation available for the types of molecules covered.Other reviews have been by Siegel," Gaydon,I9 Drowart,*' Drowart and TABLE.-DISSOCIATION ENERGIES D& OF HOMONUCLEAR DIATOMIC MOLECULES (IN kJ mol -l). I [FROM REF. (17) INTEGRATED WITH NEW VALUES]. 101.o 21.2 DIATOMIC METALS AND METALLIC CLUSTERS TABLE2.-ATOMIZATION ENERGIES Do",OF POLYATOMIC METALS (IN kJ mOI-') molecule Do" ref. molecule Do" ref. Bi 365 f 13 27,30 IPbj 221 .+ 16 45 Bi 595 f 8 27-29 Pbj 415 f 16 45 Ge 639 I20 32-34 Sn 480.7 f 17 44 Ge 999 i-25 32-34 Sn 757.7 f 20 44 Ge 1343 & 42 34 Sn 1024 & 25 44 Ge6 1703 h 54 34 Sn6 1329 & 30 44 Ge7 2013 63 34 Sn7 1612 & 35 44 Li 173.6 * 16.7 51 TABLE 3.-DISSOCIATION ENERGIES D OF DIATOMIC INTERMETALLIC COMPOUNDS (IN kJ mOl -') molecule Do0 ref.molecule Do" ref. AgAl 172 f 17 17 CdIn 134 17 AgAu 200.8 f 10.5 17 CeIr 580.7 I 25 86 AgBi 192 42 17 CeOs 503 L-33 64 AgCu 169.5 f 10.5 17 CePd 318.4 f 17 76 AgDy 124 f 19 52 CePt 551.0 & 25 86 AgEu 123.0 f 12.5 47 CeRh 545.6 f 25 86 AgGa 177.8 f 6.3 2 CeRu 527 & 25 64 AgGe 170.7 i-21 35 cocu 163 & 21 17 AgHo 119.7 5 17 48 CoGe 230 f 21 17 AgIn 163.2 f 6 2 CrCu 155 f 25 17 AgLi 173.6 -C 6.3 53 CrGe 165.7 & 29 65 AgMn 96 f 21 17 CsHg 4.8 26 AgNa 136.0 f 10.5 54 CsLi (69) 26 AgSn 134 f 21 17 CsK (47) 26 AlAu 322.2 f 6.3 55 CsNa (41) 26 AlCu 209 f 17 17 CuDy 140 i 19 31 AlLi 172.0 f 14.6 56 CuGe 197 i 17 17 AlPd 250.6 f 12.0 57 CuHo 139 f 19 31 AuBa 251.1 f 10 58 CuLi 189.3 f 8.8 53 AuBe 280 17 CuNa 172.4 & 16.7 66 AuBi 293 f 8.4 17 CuNi 201 f 21 17 AuCa 238 17 CuSn 168 & 10 77 AuCe 335 f 21 17 CuTb 187 -i 19 31 AuCo 218 f 17 17 EuRh 231.8 f 34 47 AuCr 209 I 17 17 GaLi 129.3 f 14.6 56 AuCu 224.3 & 5.1 32 GeNi 276 13 17 AuDy 254 5 20 52 GePd 259 f 17 17 AuEu 238.9 f 10.5 47 HgK 5.8 26 AuGa 230 f 38 17 HgLi 10.1 26 AuGe 270.4 f 5.0 32,35 HgNa 5.3 26 AuFe 188 4 21 17 HgRb 4.7 26 AuHo 263.6 f 33 42 InLi 86.6 f 13.4 56 AuLa 335 f 21 17 IrLa 573 & 12 67 AuLi 280.8 f 6.5 53 IrTh 570.7 f 42 8a AuLu 328.4 i 17 50 IrY 452.8 + 16 62 AuMg 243 f 42 17 KLi 78.2 i 4.2 68 AuMn 188 f 13 17 KNa 61.0 f 4.2 68 AuNa 212.1 f 12.6 54 LaPt 496 & 15 69 K.A. GINGERICH 113 TABLE3.-continued molecuIe Do ref.molecule D ref. AuNd 297 f 21 17 LaRh 524.7 f 16.7 7c AuNi 251 f 21 17 LaY 197 f 17 17 AuPb 126 & 42 17 LiNa 86.6 & 6.3 68 AuPd 151 & 21 17 LuPt 397.5 3I 34 7a AuPr 305 f 21 17 MoNb 448 & 25 70 AuRh AuSc 228.9 & 29 276.6 & 17 47 59 NaRb PtTh 54.8 4 3.8 546.6 * 42 17 8a AuSn 251 & 8 60 PtTi 394 2c 11 71 AuTb 289.5 f 33 42 PtY 470 3 8 37 AuU 318 & 29 17 RhSc 440.3 f 10.5 72 AuV 238 & 12 61 RhTh 510 4 21 73 AuY 304.1 i 8.2 62 RhTi 387.0 f 14.6 41 BaPd 220.0 f 5.0 58 RhU 516 i17 73 BaRh 257.4 5 25 58 RhV 360 i29 74 BiGa 155 f 17 63 RhY 441.8 & 10.5 72 BiPb 134 & 8 17 RuTh 587.9 f 42 9 BiSn 206.3 f 8.4 27 RuV 410 f 29 74 BiTl 117 & 13 17 TABLE4.-ATOMIZATION ENERGIES D OF POLYATOMIC INTERMETALLIC COMPOUNDS (IN kJ mol -I) molecule D; ref.molecule Do0 ref. AlAu2 506.3 i 25.1 3 AuSn 486 -& 18 60 A12Au 460.2 Zt 20.9 3 AuSn 786 f. 25 60 A1 2Pd 492.4 &-24 57 Au,Sn 542 i 18 60 Au2Ba 552 & 21 75 Au2Snz 871 f 25 60 Au~Eu 549.4 f 16.7 47 Au2Tb 582 & 42 42 AuGe 531.9 f 10 32 Bi2Sn 427.6 i 10.5 27 AuGeJ 897 & 20 32 Bi,Sn 644.3 & 16.7 27 AuGe4 1295 I30 32 CuGe 506.0 i25 76 Au2Ge 534.9 5 12 32 CuSn2 391 & 25 77 Au2Ge2 927 & 14 32 CuzSn 452 & 25 77 Au~Ho 533 & 42 42 RhTiz 996 =t42 41 AuZLU 602.1 & 33.5 50 Goldfinger,” Gingerich,2*22 Cheetham and Barrow,23 Ro~en,~~ Barrow 25 and Huber and Herzberg. 26 4. EMPIRICAL CORRELATIONS OF BOND ENERGIES AND COMPARISON WITH EXPERIMENT Bond energies of metal molecules are among the most difficult properties to calcu- late from first principles.Ab initio calculations are rather scarce and limited to a few examples involving metals with low atomic numbers. Approximate theoretical treatments such as density functional theory and the SCF-Xor-SW method or semi- empirical calculations such as the extended-Huckel (EH) and CNDO techniques have no useful predicting power as far as bond energies are concerned. Therefore one has to rely on empirical methods for the prediction of unknown bond energies of diatomic metals and metallic clusters. In this section the results of various empirical methods DIATOMIC hl ET ALS AND METALLIC CLUSTERS are compared with the corresponding experimental values and relative merits and limitations of the empirical approaches are discussed.4.1. HOMONUCLEAR DIATOMIC METALS As can be seen from table 1 the dissociation energies of diatomic representative metals are more or less well known. Gaps or large uncertainties exist for the diatomic molecules of the group IIA and IIB metals and the transition metals. The group IIA and group IIB molecules are believed to be of the van der Waals type and will not be further discussed here. Some of the transition metal molecules are also of the van der Waals type e.g. Yb, very likely Eu and possibly Mn,. The others have chemical single bonds or muItiple bonds. Several early empirical correlations for interpreting or estimating dissociation energies of homonuclear diatomic metals relate the dissociation energy to the heat of vaporization of the atom and the excitation energy (promotion energy) of the ground state atoms to their valence state.Griffith78 introduced the concept of excitation energy (promotion energy) into the discussion of the cohesive energy of transition metals. Verhaegen et al. 79 related the atomization energy of the metal and the dis- sociation energy of the dimer through cc = AH&p(M)/DE(M2). Kant and associates have used a correlation between dissociation energies valence state promotion ener- gies and empirical valence state dissociation energies to account semiquantitatively for the observed dissociation energies of the 3d transition metals" and the lanthanide metals.46 Krasnovsl first predicted the dissociation energies of the 4d and 5d transition metal molecules using the empirical relation AH;,, -D8 = const -const = C (1) with an empirical value of C = 65 -l 10 kcal mol-' for the Y to Pd series based on the experimental and DG of the end members Y and Pd.His value C = 45 kcal mo1-l for the La to Pt series was solely based on La,. He suggested that possibly C > 45 which appears to be confirmed by the recent determination of DE(Pt,) in our lab~ratory,~~ which leads to C = 50. In an alternative approach Miedema and Gingerich have recently demonstrated that there is a relation between the dissociation enthalpies of homonuclear diatomic molecules D:,the heat of vaporization of the metals in the condensed phase AH:,, and the surface energy of the pure metals y and have also predicted dissociation energies for the 4d(Y-Pd) and 5d(La-Pt) transition metals.They have used these predicted values as a basis for calculating the dissociation energies of diatomic inter- metallic compounds by the atomic cell on which Miedema will report elsewhere at this Symposium.103 A third set of 4d and 5d transition metal dimer dissociation energies has been calculated by Brewers2 using a modification of the bonding energies he has used in the bulk In table 5 the estimated values by the three methods for the dissociation energies of 4d and 5d transition metal dimers are compared. Also included have been the recently measured values by Gupta et al. which were not available at the time of the predicted values li~ted,~~,~~*~~ as well as other estimates.Gupta and Gingeri~h~~ 71984*85 also predict the dissociation energy of Ta to be larger than that of Nb2 and possibly larger than that of W2. It should be pointed out that the dissociation energies of Mo and Nb2 are probably K. A. GINGERICH 115 TABLE 5 .-COMPARISON OF VARIOUS ESTIMATED DISSOCIATION ENERGIES OF SYMMETRIC 4d AND 5d TRANSITION METAL DIMERS ~~ M estimated D:(Mz)/kJ mo1-l Miedema and other exp. Do"(Md /kJ m0l-l a Krasnov 81 Gingerich12" Brewer 82 estimates ~~~~ Zr 335-293 309 293 Hf 431-314 304 335 Nb 448-360 371 335 502 503 -I 1036 Ta 590-393 395 377 Mo 385-322 325 293 41gS4 397 A 6385 404 f2040 W 661-423 452 397 535 & 6367 Tc - 3 30 272 Re 586-385 3 76 410 Ru - 327 293 0s 598-392 405 397 Ir 481-335 333 385 Pt 377 -28 5 278 358 & 737 a Experimental values obtained subsequent to estimated values listed.slightly lower than reported because of electronic contributions of possible low-lying excited states which had not been considered in arriving at the experimental values. Brewer 86 has reevaluated the experimental value for D;(Mo2)40 assuming population of excited states to correspond to an electronic degeneracy of 5 at 2900 K which leads to 365 40 kJ mol-' for D"(M0,). While there is no doubt about some electronic contribution to the free energy functions of Mo and Nb at the temperatures of measurement Brewer may have somewhat overestimated this contribution in case of Mo,(g) in view of the high experimental second-law value,4o Dg = 407 & 33 kJ mol-I.Even if one takes a small lowering of the dissociation energies for Mo and Nb into account the recent experimental values by Gupta et al. for these two molecules and for Pt show that the estimated values by the three empirical methods12"*8'*82 listed in table 5 appear to be too low but give the trends within each period cor- rectly. 4.2. DIATOMIC INTERMETALLIC COMPOUNDS In this section the application of empirical models of bonding to diatomic inter- metallic compounds is illustrated and discussed. The models considered are the Pauling model of a polar single bond,' the valence bond model for certain inter- metallic transition molecules with multiple bonds,' and the atomic cell model.lZb 4.2.1.PAULING MODEL OF A POLAR SINGLE BOND The Pauling model of a polar single bond' can generally be applied to interpret the dissociation energies of intermetallic molecules ,0p8' except in two cases (a)The values calculated by the model are lower than the experimental values for multiply bonded molecules. For such intermetallic molecules the valence bond approach and the atomic cell model to be discussed below are applicable. (b) It has been found experimentally that combinations of the copper subgroup transition metals with alkali or alkaline earth metals have lower dissociation energies than the values calculated DIATOMIC METALS AND METALLIC CLUSTERS by the model. These findings parallel the observation by Pauling' for the alkali hydrides.Similar can also be shown for mercury compounds with alkali metals using the data in tables 1 and 3. According to the model the bond energy D(A-B) of a diatomic molecule AB may be expressed by the relation D(A-B) = +[(D(A-A) + D(B-B)] + 96(xa -xB)'(in kJ mol-') (2) where D(A-A) and D(B-B) are " single bond " energies of component elements A and B respectively and x,,and xBare the respective electronegativities. In the appli- cation of the Pauling model in our laboratory we have preferred the Pauling scale' of electronegativities since it is based on bond energy measurements. The first term in eqn (2) gives the covalent contribution to the bond energy; the second term gives the "ionic resonance " energy which is always positive.Pauling also gives an alternate " geometric mean " version D(A-B) = [D(A-A) x D(B-B)]f + 125.5(~ -xB)2(in kJ mol-I). (3) A set of single bond energies for all non-transition elements has been calculated by Sanderson.8s Pauling' also gives values for the single-bond energies (in kJ mol-') for Ge (157.3) Sn (143.0) and Bi (105) which are pertinent to the present discussion and which differ from the corresponding dissociation energies for Ge, Sn and Biz respectively listed in table 1. Usually the dissociation energies of the transition metal dimers are taken as single-bond energies for lack of a reliable knowledge of the latter. This places an additional restriction on the application of the model especially where 4d and 5d transition metals are involved for most of which the dimers are expected to have multiple-bond character.Additional uncertainties in the use of the model come from the availability of different electronegativity scales and the dependency of the electronegativity of the valence ~tate.'*~'*~~ The range of applicability of the Paul- ing model to intermetallic molecules has been greatly extended during the last decade as a consequence of the experimental determination of the bond energies of many new diatomic metal molecules. The application of the Pauling model has also been demonstrated for triatomic intermetallic molec~les.~~~~ In this case the model is applied to each individual heteronuclear bond and the contributions of the various bonds are added to yield the atomization energies.The results for triatomic molecules are illustrated in section 4.4 below. For a comparison of calculated values with experimental dissociation energies of diatomic intermetallic compounds (see table 3) reference is made to the review 2.8 7 and to the original 31935A2 ,52.54,55,59,69.91-97 literature The intermetallic compounds of gold have been most thoroughly studied both by mass spectrometry and by optical spectroscopy. Since gold is the most electro- negative metal on the Pauling scale a sizeable ionic contribution to the bond energy is expected especially for combinations with the more electropositive metals. In table 6 the values obtained using eqn (2) are shown in brackets together with the experimental values (rounded to integers) taken from table 3.The calculated values for the alkali and alkaline earth compounds have been taken from Gingerich and In calculating the remaining values the Pauling electronegativities Finkbeir~er.~~~ were used adjusted for common valence" (e.g. to 1.7 for Ge Sn and Pb from 1.8 and 1.4 for U instead of 1.7). The use of 1.4 for Sc and 1.3 for the remaining rare- earth elements instead of 1.5 for Sc and 1.1-1.2 for the lanthanides has been discussed el~ewhere.~~"?~ The values in table 6 given in parenthesis have been calculated by the atomic cell K. A. GINGERICH TABLE 6.-cOMPARISON OF EXPERIMENTAL VALUES FOR DIATOMIC INTERMETALLIC COMPOUNDS OF GOLD WITH THOSE CALCULATED AFTER THE PAULING MODEL [ ] AND THE ATOMIC CELL MODEL 0 - Li 281 13491 Na 212 Mg243 1362 1 1251 I (207) CaK - Sc Ti Mn Fe 1 Co I Ni I Cu I Zn 238 277 188 I3041 12861 [2521 (405) (332)- (2101 (1273 Rb Sr Y Zr Tc 304 13051 -cs Ba (393) La (452) Hf (234) Re 251 335 :337] - 13481 (421) 1420) (255) 269) (244) (234) 11391 (6 9) Fr Ra Ac 318 13161 model IZa and will be discussed below.For Ge and Sn Pauling's single bond energies were used to calculate the covalent contribution. In all other cases the experimental dissociation energies were taken. The latter likely correspond to a bond order higher than 1 if they are >200 kJ mol-' and in these cases one notes from table 6 that the calculated values are consistently higher than the experimental ones (AuV AuLa AuCe AuRh). The agreement observed for AuNi is trivial since there is only a very small ionic component to the bonding.According to unpublished data obtained by Dr. G. D. Blue the dissociation energy of AuMg is considerably lower than the spectroscopic estimate listed in table 6. From the estimated data for AuPb it also appears that the spectroscopically derived experimental value is much too low. 4.2.2. EMPIRICAL VALENCE BOND METHOD FOR CERTAIN MULTIPLY BONDED TRANSITION ELEMENT MOLECULES A discussion of the applicability and limitations of the empirical valence bond method for certain multiply bonded molecules between electronegative and electro- positive transition metals,s has recently been presented.I0 In this method each of the two atoms forming the molecule is promoted to a valence state with two to four un- paired electrons that is suitable for multiple bond formation.Electron pair bonds are then allowed to form between the unpaired electrons of the two atoms in their valence states. The resulting bond energy per electron pair per mole is then taken the same as the determined valence state bonding enthalpy per mole per electron for the corre- sponding type of electrons in the respective condensed metal. The individual values for the latter are taken from the Brewer curvess3 which show the variation of s orp and DIATOMIC METALS AND METALLIC CLUSTERS d valence state bonding enthalpies per mole of electrons with atomic number and with principal quantum number and in the case of d-electrons also with the number of d-electrons participating in the bonding.Wengert 95 has conveniently tabulated such values obtained from the Brewer curves. The bonding energies of all bonds formed are then added. From this sum the sum of the valence state promotion energies is subtracted. The resulting value represents the calculated dissociation energy. The necessary valence state promotion energies are taken from the literat~re.~~*~' They are evaluated in a manner analogous to the one shown by Bre~er.*~*~~ Sample calculations have been given else~here.~*'~~~~ The experimental values of diatomic molecules of platinum metals with groups 111 and IV (including thorium) transition metals and with vanadium are compared in table 7 with the values calcu- TABLE 7.-cOMPARISON OF EXPERIMENTAL DISSOCIATION ENERGIES WITH VALUES CALCULATED USING EMPIRICALVALENCE BOND AND ATOMIC CELL MODELS.(VALUESARE IN kJ mol-I.) molecule (exp.1(see table 3) Do"(calculated) valence atomic bond model8 cell model1zb IrLa 573 * 12 565 537 IrTh 571 5 42 556 555 IrY 453 rt 16 451 522 LaPt 496 i15 423 573 LaRh 525 & 17 540 470 PtTh PtTi 547 342 394 * 11 456 360 588 494 PtY 470 & 8 316 555 RhSc RhTh 440+ 11 510 + 21 387 527 467 484 RhTi 387 f 15 427" 397 RhV 360 i29 460 362 RhY 442f 11 424 452 RuTh 588 & 42 657 496 RuV 410 5 29 569 382 ~~~ @ Revised value see ref. (71). lated by the empirical valence bond model. For these molecules values calculated by the atomic cell model12b have also been included.Additional comparisons with experimental values have been given elsewhere for BaRh BaPt LuPt RhU and the cerium compounds with all platinum metals." The experimental data in ref. (10) and in table 7 show that the calculated dis- sociation energies for the intermetallic molecules with platinum which have been based on an assumed double bond are lower than the experimental values. The latter are closer to those expected from an assumed triple bond but since platinum has no suitable valence state for triple-bond formation they have to be compared with the experimental values for the corresponding iridium compounds. Molecular orbital considerations lo can be used to explain the lower than calculated dissociation energies for RhV and RuV (for RuV a quadruple bond was assumed; with an assumed triple bond the same value as given for RhV would have been obtained).Thus optimum bonding can be expected for RhSc by assuming all electrons being paired with none in antibonding orbitals. One additional electron has to go into an antibonding K. A. GINGERICH orbital in case of RuV and two additional electrons in case of RhV weakening the bond in the observed sequence. 4.2.3. ATOMIC CELL MODEL The atomic cell model proposed by Miedema and GingerichlZb for calculating enthalpies of formation of diatomic intermetallic molecules is more general than either the Pauling model of a polar single bond or the valence bond method for intermetallic molecules with multiple bonds discussed above. It is based on a model description by Miedema and associates that was originally designed to account for the heat of formation of metallic alloys both in the solid and the liquid state.The enthalpies of formation of the diatomic intermetallic molecules are derived from the difference in electronegativity AVO*and that in the electron density at the boundary of the Wigner-Seitz atomic cell An,, using the symmetric dimers as reference states. In simplified form if the metals are of approximately equal size and do not contain p valence electrons the dissociation energy D(A-B) may be expressed by D(A-B) = -$[D(A-A) + D(B-B)] + P(Av*)~-Q(An$D2 (4) where P and Q are empirical constants. This relation may be compared with that of Pauling for a polar single bond [eqn (2)]. The second term corresponds to Pauling's ionic resonance contribution but uses the same electronegativity values as Miedema and associates used for the condensed alloys.Particularly noteworthy is the third (repulsive term) lacking in Pauling's formula. For diatomic molecules between transi- tion metal atoms and polyvalent main group atoms a fourth constant term must also be added (as in the formula for condensed alloys) to account for bond strengthening due to hybridization. Miedema will give a more detailed account of this model at this Symposium.103 Comparisons of model calculations with experimental values have been presented in our original paper.12b In table 6 a comparison of experimental dissociation energies with those calculated using the atomic cell model or Pauling model is made for intermetallic compounds with gold.The atomic cell model as has been noted for the Pauling model yields over-large dissociation energies for molecules of Group I B transition metals with alkali or alkaline earth metals. However if the value of the ionicity parameter P [eqn (4)] is reduced by a factor of 0.65 the agreement with the experimental values for alkali compounds with group IB metals becomes good.12b For the molecules with one of the magnetic metals Cr Mn Fe Co and Ni the atomic cell model yields over-small dissociation energies whereas the Pauling model gives quite good agreement in these cases. The atomic cell model has not yet been applied to the lanthanide and actinide-gold compounds (except thorium) for which the Pauling model gives good agreement with experiment.In table 7 values for dissociation energies calculated by the atomic cell model are compared with experimental values and those calculated by the valence bond model. As can be seen the values calculated for the platinum compounds are higher than the experimental values. This has been attributed to a barrier for charge transfer for metals of high electronegativity at the end of a transition metal series such as Pt and Au (see table 6) when these metals are combined with much more electropositive partners such as the group I11 transition metals.12b For RhV and RuV the atomic cell model reproduces the experimental values well unlike the valence bond model. A significant advantage of the atomic cell model is seen in its applicability to intercombinations of transition metals in the central region of the transition series to which the Pauling model and the valence bond model cannot be applied.In com- DIATOMIC METALS AND METALLIC CLUSTERS menting on the relative merits of the three empirical models discussed here the atomic cell model is most generally applicable to intermetallic compounds whereas the Pauling model and the valence bond approach may give more accurate predictions in their specific areas of applicability. Another advantage of the atomic cell model is seen in its direct relation to the bonding in the corresponding condensed alloys which can be expressed in quantitative terms. Further experimental results are expected to permit refinements of all models but especially of the atomic cell model.4.3. MAXIMUM BOND ENERGY BETWEEN TWO METAL ATOMS Both the valence bond model and the atomic cell model predict similar maximum bond energies between two metal atoms. The highest values calculated by the valence bond model assuming quadruple bond formation are between 630 and 670 kJ mol-1.8" The observation that quadruple bonding is not fully achieved led to the conclusion that the maximum possible bond energy between two ligand-free metal atoms is 670 & 84 m01-I.~ The atomic cell model yields maximum bond energies for intermetallic molecules of iridium (IrZr 588; IrHf 555; IrTh 555) osmium (OsZr 574) and platinum (PtSc 571 ; PtZr 621 ; LaPt 573; HfPt 584; PtTh 588).12* Here the values are given for molecules with a predicted dissociation energy of more than 550 kJ mol-'.As discussed earlier12' and as confirmed for LaPt,69 the calcu- lated values for the platinum-containing molecules with the more electropositive metals tend to be too high. On the other hand there will be a small upwards revision for the bond energies since the new experimental evidence quoted in table 5 indicates that the dimer dissociation energies of the 4d and 5d transition metals are higher than the estimated values1*" used as a basis for the calculations. However the downward correction for D"(PtZr) is expected to be larger than this upward correction thus 620 kJ mo1-' the highest value obtained by the atomic cell model may be con- sidered an upper limit value and is in agreement with the upper limit value based on the valence bond model.For the homonuclear transition metal molecules the highest estimates (see table 5) are 66lS1and 538'2a kJ mo1-' respectively both for W2where the latter value was based on eqn (4) in ref. [12(a)l. The experimental evidence for the isoelectronic molecules RhY and Mo (see table 4) suggests that D;(W,) is not larger than that of the isoelectronic molecule IrLa (573 -& 12 kJ mol-I). The dissociation energy of Ta is expected to be possibly larger than that of W2,36but not by much. Thus 600 & 40 kJ mol-' is considered to be a safe upper limit for the maximum dissociation energy of a symmetric diatomic metal molecule. Current knowledge from experi- ment and from the empirical models of bonding suggests that the actual upper limit for a diatomic intermetallic molecule is lower than the upper limit of 670 & 84 kJ mo1-l given earlier' and a revised upper limit of 640 rf 40 kJ mol-' (or 150 & 10 kcal mol-I) is proposed.It is of interest to compare this upper limit for the bond energy between two ligand- free metal atoms with the maximum bond energy expected for quadruply bonded pairs of metal atoms in dinuclear transition metal complexes of Moil Cr" TclI1 and Re1('." In early estimates of the strength of quadruply bonded metals e.g. for Re,Cli- Cotton99u had suggested that the bond energy holding the pair of metal atoms together may be as high as 400 kcal mol-' (1674 kJ mol-I). In all early discussions9'" the strongest argument for the high bond energy had been the very short observed bond distance in conjunction with the observed geometrical arrange- ment of the ligands.Cotton had qualitatively related the bond shortening to the bond strength.'" The experimental calorimetric determination of the bond strength K. A. GINGERICH between such pairs of metal atoms in a complex compound is difficult and the results quite uncertain since attribution of the measured heat effects has to be made to all bonds in the corresponding complex compounds. About ten years ago this author had therefore graphically related the bond strength to the bond shortening."' In these estimates the known dissociation energies D;,and equilibrium separations re,of the molecules Cu, Au2 AlAu RhC and PtC had served as a basis for an empiri- cal plot of bond shortening against bond energy.The bond shortening was based on the Pauling' metallic single bond radii or 12-coordinate radii respectively as a reference. For the largest bond shortening known at that time,Io2 in molybdenum(rr) acetate [Mo(O,CCH,),], a bond energy of 740 kJ mol-' resulted from both methods. For the first established quadruply bonded complex compound (Re,C1,)2-,100 360 kJ mol-l resulted using the 12-coordinate metallic radii as a reference or 520 kJ mo1-I using the metallic single bond radii. These results showed that such quadruple bonds were not the strongest chemical bonds known as had been widely assumed at that time. The recent experimental determination of the bond length r(Mo-Mo) = 1.929 Ag5and dissociation energy Dg(Mo,) = 404 -& 20 kJ m~l-',~' indicate that the bond energies between quadruply bonded metal atoms in complex compounds are actually smaller than the largest bond energies found in ligand-free metal dimers provided the experimental bond distance for gaseous Mo is correct.4.4. POLYATOMIC METAL MOLECULES For polyatomic metals (table 2) and intermetallic compounds (table 4) the bond additivity concept has frequently been used to interpret the experimental atomization energies. In case of the intermetallic compounds the Pauling model has been success- fully used in addition (see table 8). The electronic structures of each family of poly- 8.-COMPARISON OF EXPERIMENTAL AND CALCULATED VALUES USING THE PAULING TABLE MODEL FOR TRIATOMIC SYMMETRIC MAu2 INTERMETALLIC MOLECULES.(VALUESARE IN kJ mol-l.) DE(MAu2) molecule calc." exp.' AlAu2 528 506.3 f25.1 Au2Ba 674 552 i21 Au~Eu 482 549.4 k 16.7 Au2Ge 472 534.9 rt 12 AuZHO 534 533 f 42 Au~Lu 592 602.1 & 33.5 Au2Sn 502 542 f18 AuzTb 580 582 f42 = - a Using Do0 D(M-M) + D(Au-Au) + 192(xA x~)~; Taken from table 4. atomic metals (and metalloids) give rise to special trends in bond energies (and geometries) for each family e.g. group IV or group V polyatomic molecules. For specific details reference is made to the individual literature quoted in tables 2 and 4. The lack of spectroscopic data concerning their electronic and molecular structure limits the insight into the bonding that can be gained from available atomization energies.Theoretical and semi-empirical calculations have been contributing to qualitative or semi-qualitative estimates of stabilities geometry and spectra of small clusters. It is expected that in the near future the experimental and theoretical DIATOMIC METALS AND METALLIC CLUSTERS methods will mutually aid in gaining a deeper understanding of the nature of small metal clusters. Note added in proof In his paper at this Symposium,103 Miedema has treated the metals Ni Pd Pt or Au differently in the prediction of the dissociation energies of heteronuclear diatomic transition metal molecules from the original paper by Mie- dema and Gingerich.lzb The values listed in table 1 of Miedema's paper103 include a correction for resistance to charge transfer of the metals Ni Pd Pt and Au when combined with the more electropositive transition metals.In table 9 the values for those intermetallic molecules of gold and platinum for which a correction has been made by Miedernalo3 are listed together with the experimental values (table 3) and previously estimated values (tables 6 and 7). As can be seen from table 9 the refine- ments to the atomic cell model by Miedema have led to an improvement in all the calculated values. This is particularly the case for the platinum compounds for which the values calculated by the atomic cell model now agree with the experimental values except for TiPt for which it almost agrees. Thus for diatomic platinum com- pounds with electropositive transition elements the atomic cell model appears to be superior to the empirical valence bond model.TABLE9.-cOMPARISON OF EXPERIMENTAL VALUES FOR THE DISSOCIATION ENERGIES (IN kJ mol -l) OF DIATOMIC INTERMETALLIC COMPOUNDS WITH THOSE CALCULATED BY VARIOUS EMPIRICAL MODELS D;,calculated molecule D& exp. table 3) (see -atomic cell model ref. (126) ref. (103) model Pauling bond model valence AuSc 277 i17 405 369 286 AuY 304 & 8 393 359 305 AuLa 335 3= 21 421 362 348 AuTi 332 314 252 AuZr 452 420 AuHf 420 393 AuTh 423 392 369 AuV 238 12 322 314 252 AuNb 381 373 AuTa 383 376 LaPt 496 i15 573 484 423 PtTh 547 & 42 588 508 456 PtTi PtY 394 f11 470 + 8 494 555 441 470 360 316 HfPt 584 516 PtSc 571 488 PtZr 621 544 Also included in table 9 are values for platinum intermetallic molecules that have been mentioned in section 4.3 in connection with the estimation of the maximum bond energy between two metal atoms and for which Miedema has presented refined K.A. GINGERICH estimates. These refined values are ir,support of the qualitative conclusions drawn in section 4.3 above. In table 2 of their paper Brewer and Winn104 present different estimated dissocia- tion energies from those values listed in table 5 of this paper given by Brewer.' Expressed in kJ mol-' for comparison with Brewer's values quoted in table 5 these are Zr, 333; Hf, 333 & 50; Ta, 349 & 50; W, 482 & 66; Re, 312 & 82; Ru, 308; Os, 366 & 50; and Ir, 349 & 50.Except for the lower estimates for Ta, Re and Os, these more recent estimates by Brewer and Winn are closer to the corresponding estimates by Miedema and Gingerich 12b and by Krasnov.81 This article has the nature of a summary and revicw of more than a decade's effort. A number of colleagues have contributed at the various stages as can be seen from the list of references. In particular the author wishes to thank Drs. G. D. Blue U. V. Choudary D. L. Cocke S. K. Gupta R. Hague J. Kordis A. R. Miedema B. M. Nappi M. Pelino and V. Piacente Mr. J. E. Kingcade Jr and his wife. Dr. Gupta had in addition contributed to this manuscript with valuable discussions. The ex- perimental work at Texas A & M University has been generously supported through research grants by the National Science Foundation and the Robert A.Welch Foundati on. L. Pauling The Nature of the Chemical Bond (Cornell University Press Ithaca N.Y. 3rd edn 1960). K. A. Gingerich J. Cryst. Growth 1971 9 31. K. A. Gingerich D. L. Cocke H. C. Finkbeiner and C.-A. Chang Chem. Phys. Letters 1973 18,102. L. Brewer personal communication 1968. ' K. A. Gingerich and R. D. Grigsby Metallurg. Trans. 1971 2 917. K. A. Gingerlich and D. L. Cocke Chem. Cotnm. 1972 536. K. A. Gingerlich High Temp. Sci. 1971 3 415; D. L. Cocke and K. A. Gingerich J. Phys. Chem. 1972 76 2332 4042; D. L. Cocke K. A. Gingerich and J. Kordis High Temp. Sci. 1973 5 474. a (a)K. A. Gingerich Chem. Phys. Letters 1973 23 270; (6) K. A.Gingerich J.C.S. Faraday II 1974 70 471. K. A. Gingerich Chem. Phys. Letters 1974 25 526. lo K. A. Gingerich Int. J. Quantum Chern. Symp. 1978 12 489. l1 A. R. Miedema R. Boom and F. R. de Boer J. Less Common Metals 1975,41 283; A. R. Miedema J. Less Common Metals 1976,46,67; R. Boom F. R. de Boer and A. R. Miedema J. Less Common Metals 1976 45 237; 1976 46 271. (a) A. R. Miedema and K. A. Gingerich J. Phys. B 1979 12 2081; (6) A. R. Miedema and K. A. Gingerich J. Phys. B 1979 12 2255. l3 W. A. Chupka and M. G.Inghram J. Phys. Chem. 1954,59 100. l4 K. A. Gingerich J. Chem. Phys. 1968 49 14. '' M. G. Inghram and J. Drowart in High Temperature Technology (McGraw-Hill New York 1960) pp. 219-240. R. T. Grimley in Characterization of High Temperature Vapors ed.J. L. Margrave (Wiley- Interscience New York 1967) pp. 195-243. l7 L. V. Gurvich G. V. Karachevstev V. N. Kondrat'yev Y. A. Lebedev V. A. Mendredev V. K. Potapov and Y.S. Khodeev Bond Energies Ionization Potentials and Electron Afinities (Nauka Moscow 1974) in Russian. B. Siegel Quart. Rev. 1965 19 77. l9 A. G. Gaydon Dissociation Energies and Spectra of Diatomic Molecules (Chapman and Hall London 3rd edn 1968). 2o J. Drowart in Phase Stability in Metals and Alloys ed. P. S. Rudman J. Stringer and R. I. Jaffee (McGraw-Hill New York 1967) pp. 305-317. 21 J. Drowart and P. Goldfinger Angew. Chem. 1967,79 589; Angew. Chem. Int. Edn 1967 6 581. 124 DIATOMIC METALS AND METALLIC CLUSTERS 22 K. A. Gingerich Current Topics Mater.Sci. 6 to be published. 23 C. J. Cheetham and R. F. Barrow Adv. High Temp. Chem. 1967 1 7. 24 Spectroscopic Data Relative to Diatomic Molecules ed. B. Rosen (Pergamon Oxford 1970). 25 R. F. Barrow Diatomic Molecules A Critical Bibliography of Spectroscopic Data (Centre National de la Recherche Scientifique Paris 1973). 26 K.-P. Huber and G. Herzberg Molecular Spectra arid Molecular Structure IV. Constants of Diatomic Molecules (Van Nostrand Reinhold New York 1979). 27 K. A. Gingerich U. V. Choudary D. L. Cocke and K. Krishnan to be published 28 L. Rovner A. Drowart and J. Drowart Trans. Faraday SOC. 1967 63 2906. 29 F. J. Kohl 0. M. Uy and K. D. Carlson J. Chem. Phys. 1967 47 2667. 30 C. L. Sullivan J. E. Prusaczyk and K. D. Carlson High Temp.Sci. 1972 4 212. 31 K. Hilpert Ber. Bunsenges. phys. Chenr. 1979 83 161. 32 J. E. Kingcade Jr U. V. Choudary and K. A. Gingerich Inorg. Chem. 1979,18 3094. 33 J. Drowart G. DeMaria A. J. H. Boerboom and M. G. Inghram J. Chem. Phys. 1959 30 308. 34 A. Kant and B. H. Strauss J. Cheni. Phys. 1966 45 822. 35 A. Neckel and G. Sodeck Monatsh. Chem. 1972 103 367. 36 S. K. Gupta and K. A. Gingerich J. Chem. Phys. 1979 70 5350. 37 S. K. Gupta B. M. Nappi and K. A. Gingerich to be published. 38 W. C. Stwalley J. Chem. Phys. 1976 65 2038 39 JANAF Thermocheniical Tables J. Phys. Chem. Ref. Data 1978 7 pp. 793-940. 40 S. K. Gupta R. M. Atkins and K. A. Gingerich Inorg. Chem. 1978 17 3211. 41 D. L. Cocke and K. A. Gingerich J. Chern. Phys. 1974 60 1958.42 J. Kordis K. A. Gingerich and R. J. Seyse J. Chem. Phys. 1974 61 51 14. 43 C. A. Stearns and R. J. Kohl High Temp. Sci. 1973 5 113. 44 K. A. Gingerich A. Desideri and D. L. Cocke J. Cheni. Phys. 1975 62 731. 45 K. A. Gingerich D. L. Cocke and F. Miller J. Chem. Phys. 1976 64 4027. 46 A. Kant and S. S. Lin Monatsh. Chem. 1972 103 757. 47 D. L. Cocke K. A. Gingerich and J. Kordis High Temp. Sci. 1975 7 61. 48 D. L. Cocke and K. A. Gingerich J. Phys. Chent. 1971 75 3264. 49 M. Guido and G. Balducci J. Chem. Phys. 1972 57 561 1. 50 K. A. Gingerich Chenr. Phys. Letters 1972 13 262. 51 C. H. Wu J. Chent. Phys. 1976 65 3181. 52 K. Hilpert Ber. Bunsenges. phys. Chem. 1977 81 30; 348. 53 A. Neubert and K. F. Zmbov Trans. Faraday Soc. 1974 70 2219. 54 V.Piacente and K. A. Gingerich High Temp. Sci. 1977 9 189. 55 K. A. Gingerich and G.D. Blue J. Chem. Phys. 1973 59 185. 56 D. J. Guggi A. Neubert and K. F. Zmbov in Proc. 4th Int. Conf. Chem. Thermodynamics ed. M. Laffitte Montpellier France August 1975 paper III/23. 57 D. L. Cocke K. A. Gingerich and C. Chang J.C.S. Faraday I 1976,72,268. 58 K. A. Gingerich and U. V. Choudary J. Cheni. Phys. 1978 68 3265. 59 K. A. Gingerich and H. C. Finkbeiner in Proc. 9th Rare Earth Conf. Oct. 10-14 1971 Blacksburg Virginia ed. P. E. Field (CONF-711001 Chemistry [TID-45001,National Informa- tion Service U.S. Dept. Commerce Springfield VA22151) vol. 2 pp. 795-803. 6o K. A. Gingerich D. L. Cocke and U. V. Choudary Inorg. Chim. Acta 1975,14 L47. 61 S. K. Gupta M. Pelino and K.A. Gingerich J. Chem. Phys. 1979 70 2044. 62 R. Haque M. Pelino and K. A. Gingerich to be published. 63 V. Piacente and A. Desideri J. Chem. Phys. 1972 57 2213. 64 K. A. Gingerich and D. L. Cocke Inorg. Chitn. Acta 1978 28 L171. 65 A. Kant and B. H. Strauss J. Chem. Phys. 1968 49 523. 66 V. Piacente and K. A. Gingerich 2.Naturforsch. 1973 28a 316. 67 R. Haque M. Pelino and K. A. Gingerich J. Phys. Chem. 1979 71 2929. 68 K. F. Zmbov C. H. Wu and H. R. Ihle J. Chem. Phys. 1977 67,4603. 69 B. H. Nappi and K. A. Gingerich unpublished data. 70 S. K. Gupta and K. A. Gingerich J. Chem. Phys. 1978 69 4318. 71 S. K. Gupta M. Pelino and K. A. Gingerich J. Phys. Chem. 1979 83 2335. 72 R. Haque and K. A. Gingerich J. Chem. Thermodynamics in press. 73 K.A. Gingerich and S. K. Gupta J. Chem. Phys. 1978 69 505. 74 K. A. Gingerich and U. V. Choudary unpublished data. 75 U. V. Choudary K. Krishnan and K. A. Gingerich unpublished data. 76 J. E. Kingcade K. A. Gingerich and U. V. Choudary J. Phys. Chem. 1978 82 49. K. A. GINGERICH 125 77 J. E. Kingcade Jr. D. C. Dufner S. K. Gupta and K. A. Gingerich High Temp. Sci. 1978 10 213. 78 J. Griffith J. Inorg. Nuclear Cheni. 1956 3 15. 79 G. Verhaegen F. E. Stafford P. Goldfinger and M. Ackerman Trans. Faradrry SOC. 1962 58 1926. 8o A. Kant and B. Strauss J. Chem. Phys. 1964 41 3806. " K. S. Krasnov Teplofz. Vys. Temp. 1975 13 441. L. Brewer personal communication Oct. 1978. 83 L. Brewer in Phase Stability of Metals aiid Alloys ed. P. S. Rudman J.Stringer and R. I. Jaffee (McGraw-Hill New York 1967) pp. 39-61 241-249 344-346 and 560-568. 84 K. A. Gingerich 1977 as quoted in Yu. M. Efremov A. N. Samoilova V. B. Kozhukhovsky and L. V. Gurvich J. Mol. Spectr. 1978 73 430. Yu. M. Efremov A. N. Samoilova V. B. Kozhukhovsky and L. V. Gurvich J. Mol. Spectr. 1978,73,430. 86 L. Brewer personal communication March 1979. 87 K. A. Gingerich Chitnia 1972 26 619. R. T. Sanderson J. Inorg. Nuclear Chern. 1966 28 1553. 89 W. Gordy and W. J. 0. Thomas J. Chem. Phys. 1956 24 439. 90 H. 0. Pritchard and H. A. Skinner Chem. Rev. 1955 55 745. 91 M. Ackerman J. Drowart F. E. Stafford and G. Verhaegen J. Chem. Phys. 1962 36 1557. 9L G. D. Blue and K. A. Gingerich Ado. 16th Amer. Conf. Mass Spectrometry and Allied Topics May 12-17 1968 Pittsburgh PA.93 A. Kant J. Cheni. Phys. 1968 48 523; 49 5144. 94 (a) K. A. Gingerich and H. C. Finkbeiner Clienz. Cornm. 1969 901 ; (6) K. A. Gingerich and H. C. Finkbeiner J.Chetn. Phys. 1970,52,2956; 1971,54,2621. 95 P. R. Wengert Thermodynamic Stabilities of Certain Intermetallic Contpoiinds Mude From Traiisitiott Elenients UCRL-I 8727 Berkeley CA 1969. 96 C. E. Moore Atotnic Etiergy Levels Natl. Bur. Stand. Circ. 467 vol. 1-3 1949 1952 1958. 97 L. Brewer J. Opt. SOC. Anzer. 1971 61 1101. 98 L. Brewer High Stretigth Materials ed. V. F. Zackay (Wiley New York 1965) pp 12-103. 99 (a)F. A. Cotton Accounfs Chert?. Rex 1979 2 240; (b) F. A. Cotton Chem. Soc. Rev. 1975 4 225; (c) F. A. Cotton Accounts Clietn. Res. 1978 11 25; and literature quoted therein.loo F. A. Cotton Inorg. Chetn. 1965 4 335. K. A. Gingerich invited paper presented at the Joint Conf. of the Chem. Inst. of Canada with the Amer. Chem. SOC. Symp. on Structure and Chemistry of Compounds with Metal-Metal Bonds Toronto Canada (1970). Also quoted in ref. (8a). lo' D. Lawton and R. Mason J.Amer. Chem. SOC.,1965,87,92I. Io3 A. R. Miedema Faraday Disc. Chetn. SOC.,1980 14 136. Io4 L. Brewer and J. S. Winn Faraday Disc. Chent. SOC.,1980 14 127.

ISSN:0301-5696    

DOI:10.1039/FS9801400109

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Models for Calculation of Dissociation Energies of Homonuclear Diatomic Molecules BY LEOBREWERAND JOHN S. WINN Materials and Molecular Research Division Lawrence Berkeley Laboratory and Department of Chemistry University of California Berkeley California 94720 U.S.A. Received 29th August 1979 The variation of known dissociation energies of the transition metal diatomics across the Periodic Table is rather irregular in a manner similar to the irregular variation of the enthalpies of sublimation of the bulk metals. This has suggested that the valence-bond model used for bulk metallic systems might be applicable to the gaseous diatomic molecules as well as to the various clusters intermediate between the bulk and the diatomic molecules. The available dissociation energies were converted to valence-state bonding energies considering various degrees of promotion to optimize the bonding.It was found that the model used for the bulk metals was applicable to the diatomic molecules. The degree of promotion of electrons to increase the number of bonding electrons is smaller than for the bulk but the trends in bonding energy parallel the behaviour found for the bulk metals. Thus using the established trends in bonding energies for the bulk elements it was possible to calculate all un- known dissociation energies to provide a complete table of dissociation energies for all M2molecules from H2to Lr2. The details of the calculations and final values are presented. For solids such as Mg Al Si and most of the transition metals large promotion energies are offset by strong bonding between the valence state atoms.The main question is whether bonding in the diatomics is adequate to sustain extensive promo- tion. The most extreme example for which a considerable difference would be expected between the bulk and the diatomics would be that of the Group IIA and IIB metals. The first section of this paper which deals with the alkaline earths Mg and Ca will demonstrate a significant influence of the excited valence state even for these elements. The next section will then expand the treatment to transition metals. THE ALKALINE EARTHS While most of the diatomic metals have at least one unpaired electron per atom to contribute towards bonding the Group IIA and IIB metals do not With ground state configurations ns2 and (n -l)d1'ns2 respectively the diatomics of these metals should be van der Waals molecules analogous to the rare gas diatomics with very small dissociation energies.However the first excited states of the rare gases involve excitation to a shell of the next higher principal quantum number but the Group IIA and IIB atoms have nsnp and ns(n -l)d excited configurations available at considerably lower energy. These low-lying configurations are certainly of importance in the bulk metal bonding. In this section we show how these low-lying states influence even the weakly bound diatomics and how spectroscopic data on weakly bound species may be treated to yield accurate estimates of the dissociation energy. We restrict the spectroscopic analysis to Mg and Ca, the only two diatomics of these groups for which detailed spectral constants of the ground electronic state have L.BREWER AND J. S. WINN been measured. The approach is to invert spectroscopic constants (ae,LO,^, Be a, etc.) to the parameters of a potential function expansion. We use the expansion where ;1= 1 -(R,/R)P. This potential function has been applied to several weakly bound diatomics with very good The parameters are e (with units of energy) the correction coeffi- cients en the equilibrium bond length Re and the parameter p which need not be integral. Note however if p = 6 and en = 0 n = 1 2 3 . . . then eqn (1) is the familiar Lennard-Jones (6-12) potential. The expressions relating these parameters to spectroscopic constants have appeared in the literat~re.~*~ For Mg, we have used the constants obtained by Vidal and Scheingraber6 in their analysis of the spectrum reported by Balfour and Douglas7 For Ca, we used the constants by Balfour and Whitlock.8 The parameters one obtains for the potential functions of the ’C,+ ground states are given in table 1.Dissociation energies are TABLE 1 .-PARAMETERS OF EQN (1) FOR X’Zc,+GROUND STATES OF Mg2 AND Ca2 eo = 785.94 K eo = 2570.8 K p = 3.59 p = 3.57 ela = 0 ela = 0 e2 = 0.058 99 e2 = -0.2317 e3 = 0.079 66 e3 = -0.1200 e4 = -0.122 9 e4 = 0.059 7 e5 = -0.147 6?6 = 0.108 Re = 3.890A Re = 4.277 4 A a The constant el is identically zero by our choice for determiningp as discussed in ref.(1)-(5). obtained by setting A = 1 in eqn (1). The predictions are dissociation energies of 768 K for Mg and 1820 K for Ca,. These values are 20 and 15 % respectively Iarger than spectroscopic estimates6*s of the dissociation energy and are in all likeli- hood truly in error by these amounts. The source of this error can be traced to the very informative parameter p. Note from table 1 that p x 3.6 for Mg and Ca, which means eqn (1) approaches the separated atom limit at large R like R-3-6. In contrast one knows that the proper large R behaviour should be R-6,in accordance with dispersion theory. Thus eqn (1) rises toward the dissociation plateau too slowly and thereby overestimates the dissocia- tion limit. The parameter p (as well as the others) is evaluated from equilibrium properties of the diatomic and perhaps should not be expected to give the proper long-range behaviour to the full potential.Yet in many case~l~~*~ as diverse as Ar, NaAr and BeAr+ the value ofp is large enough to give the proper long-range behaviour. (Actually theoretical arguments predict3 that p will be closer to the value n -1 than DISSOCIATION ENERGY MODEL to n where n is the expected long-range exponent. This prediction is observed in the previously reported molecules.) Therefore the small value of p for Mg and Ca is informative. For most chemically bound diatomics p is in the range 0.4-2.5 [and parenthetically eqn (1) does not converge at all well for these molecules]. Thus the alkaline earth diatomics have potential functions with a shape near Re which is inter- mediate between that of truly non-bonded diatomics such as Ar and NaAr and that of ordinary chemically bound diatomics.Perturbation theory expressions for p indicate the role of excited state mixing in determining the value ofp. It is clear that one is observing the effects of this mixing in the alkaline earth ground states even though the bonding remains very weak. TRANSITION METAL DIATOMICS The alkaline earth example illustrates that promotion from the ground atomic state plays a small but definite role in the bonding of even the weakly bound Group I1 element diatomics. Most atoms have a filled valence s orbital in the ground state; promotion of an s electron to provide two bonding electrons is important for the bulk metals.In addition promotion of inner shell d or felectrons can play an important role. The lanthanide elements provide a clear illustration of the role of promotion of 4f electrons in the homonuclear diatomic gases. Kant and Lin noted that the dis- sociation energies of the diatomic lanthanides decreased steadily from cerium to europium with a large increase for gadolinium with again a steady decrease to ytter- bium. They pointed out that the trends were parallel to those for the enthalpies of sublimation of the bulk metals and that the trends were due to the increasing difficulty of promotion of 4f electrons with increasing nuclear charge. Examination of the experimental values given in table 2 indicates similar parallel trends for the 3d transi-TABLE 2.-vALENCE STATE BONDING ENTHALPIES OF DIATOMICS element (AHi/R)/kK" reference valence state valence bonding in kKper electron d SP H 51.967 * 0.001 1s 52.0 He 0.Li 12.16 f0.1 2s J 2. Be B (<0.3) 35. f3 33. C 72.0 &1 36. N 113.25 f0.1 38. 0 59.36 & 0.02 29.7 F 18.59 f.0.07 18.6 Ne 0.025 f0.004 Na 8.36 Z!I 0.1 3s 8. Mg 0.5814 f.0.002 A1 20.0 rt 2 3P 20.0 Si 37.3 f.1 3P2 19. P 58.41 410.03 3P3 19.5 S 50.704 * 0.03 3P4 25. CI 28.774 0.001 3P5 29. Ar 0.122 30.002 L. BREWER AND J. S. WINN 129 TABLE2.-continued ~~ valence bonding element (AH,"/R)/kK" reference valence state in kK per electron d SP K 6.0 -4 0.1 4s 6.0 Ca 1.5 f 0.2 sc 19.1 & 3.0 3d24s 17.18. Ti 16. f3 3d '54s4p * 8. 19. V 28.6 rt 2.0 3d3''4s4po* 7.2 20. Cr 18. f 3.0 3d4'54s4po*5 2.5 21. Mn 5. f 3.0 '* Fe 14.6 f 2.5 3d6*'4.~4~ 5.5 22.0 co 20. f 3.0 3d 54s4po* 10.0 22.5 Ni 26. f 2.5 3d8* 54s4p0. 19.5 23.0 cu 23.5 12 4s 23.5 Zn 2.0 f 0.5 Ga 16.6 &fl 4P 16.6 Ge 32.6 i. 1.5 4P2 16.3 As 45.95 f 0.01 4P3 15.3 Se 39.58 f 0.03 4P4 19.8 Br 22.873 f 0.001 4P5 22.9 Kr 0.182 f 0.002 Rb 5.7 f 0.5 5s 5.7 Sr (1.7) Y 18.8 -4:3 4d25s 8.7 13 Zr (40) 4d35s 3.3 14 Nb 56 35 4d45s 0.3 15 Mo 44. 35 4d55s 5.6 16 Tc (34) 4d65s 6.1 17 Ru '(37) 4d75s 6.3 18 Rh 32.8 k3 4d85s 7.2 18.5 Pd 12.6 f 2.5 4d95s 2.5 19 Ag 19.3 f 0.8 5s 19.3 Cd 1.1 i. 0.2 In 12.0 It1 5P 12 Sn 23.O k2 5p2 12.5 Sb 35.9 f.0.5 5P3 12 Te 31.07 f 0.1 5P4 15.5 I 17.899 k 0.001 5P5 18 Xe 0.266 f 0.003 cs 4.57 f 0.1 6s 4.6 Ba (3) f2 5d6s 14 15 La 29. f3 Sd26s 10 17 Ce 29. f3 4f5d26s 9 18 Pr 18 & 3.5 4f25d26s 8.5 19 Nd 10 f 3.5 4f35d26s 8.5 20 Pm (8.5) 4f45d26s 8 21 Sm 7 f3 4f65do*56s6p0'5 19 22 Eu 4 &2 4f75do'56s6po'5 17 23 Gd 20.5 +4 4j'5d26s 7 24 Tb 15 f3 4J"5d26s 7 25 DY 8 f4 4f96s6p 26 DISSOCIATION ENERGY MODEL TABLE2.-continued valence bonding element (AH,"/R)/kK" reference valence state in kK per electron d SP Ho f3 4f O6s6p 27 Er +3 4fI26s6p 27 Tm f2 4f l36s6p 27 Yb f2 4f 146s6p 26 Lu 5d6s6p 18 26 Hf f6 5d26s6p 14 26 Ta f6 5d46s 11 26 W 58 5d56s 8 26 Re + 10 5d56s6p 8 26 0s +6 5d76s 11 26 Ir &6 5d86s 12 26 Pt &5 5d96s 13.5 26.5 Au stl 6s 26.7 Hg 0.15 T1 f3 7 Pb f3 5 Bi fl 8 Po f3 9 At 10 Rn Fr f1 7s 4 Ra fl Ac f7 6d 27s 11 16 Th f.4 6d37s 11 17 Pa 5f6d37s 10.5 17.5 U +6 5j36d27s 10 18 NP &7 5f46d27s 10 19 Pu 13 Am Cm f7 5f 76d27s 10 22 Bk It2 Cf Es Fm Md No Lr f10 7s27p 20 tion metals.l09l1 However the quantitative analysis of the data to be illustrated below shows that there are substantial differences between the bonding in the M2 gas and in the bulk solid for many elements.The second column of table 2 presents values of AHi/R for M,(g) = 2M(g). Calculated or estimated values are given in parentheses. Uncertainties are listed for all experimental values based on a critical evaluation of the literature.When a review paper adequately covers the literature and arrives at a value considered ac- ceptable only a reference to the review paper is given. Otherwise references are L. BREWER AND I. S. WINN given to the original papers. As the experimental values were used to calibrate the variation of bonding with nuclear charge across the Periodic Table the calculated values obtained by interpolation of bonding values have uncertainties close to those of adjoining elements but generally larger by "1 kK. Thus uncertainties are not indicated as they can be obtained from the uncertainties given for neighbouring experimental values. However where extrapolations are necessary or if there is reason to suspect the accuracy of the bonding trends uncertainties are also indicated for the calculated values.The fourth column of table 2 gives the electronic con- figuration of the atomic valence state selected as illustrated below. No configuration is shown for van der Waals molecules. The trends in bonding are shown in the last column where AHi/R for dissociation of M,(g) in its ground state to the atoms in the indicated valence state has been divided by the number of bonding electrons per atom with a separation into bonding per d-or per sp-electron for the transition elements. The method of determining the effective electronic configuration in the valence state is quite straightforward for most elements. The enthalpy of dissociation of M,(g) to two M(g) in their ground state is given by AH,"/R= (n -l)Ed/R+ E,/R -2P for a transition metal with ground state P2s2 and a valence state d"-'s.The promotion of a ground state atom to the valence state requires P kK for one atom or 2P kK for two atoms. Ed is the bonding energy (more strictly enthalpy but at 0 K they are essentially identical) per d electron and E is the bonding energy per s electron. The promotion energies to levels of each electronic configuration for elements other than the lanthanides and actinides are tabulated by Moore.35 Due to lack of data for the lanthanides and actinides a model for prediction of promotion energies had previously been de~eloped.~~ The recent review3' of values for the lanthanides has confirmed the reliability of the model and where experimental data are still lacking the predictions of the model can be confidently used.As noted earlier," the energy corresponding to the lowest state of each configuration can be accurately used in place of a weighted mean of all the levels of a configuration if the valence state bonding energies are obtained from experimental data using the same basis for the promotion energies. For the transition metals there are often two configurations e.g.,4dnW15sor 4dn-25s5p that might contribute significantly. One can differentiate the energy eqn (10) to obtain the optimum mix but the data are not accurate enough to specify more closely than one-half electron as in 3d2*54s4p0.s for Ti. Table 3 gives the TABLE 3 .-PROMOTION ENERGIES TO VALENCE STATES element ground state promotion energy /kK dfl-1 dn-2sp sc 3d4s2 16.575 22.550 Ti 3d24s2 9.434 22.844 V 3d 24s2 3.039 23.541 Cr 3d54s 0.0 35.929 Fe 3d64s2 9.968 27.842 co 3d74s2 5.011 33.973 Ni 3ds4s2 0.295 37.054 Y 4d5s2 15.737 21 SO9 Zr 4d25s2 7.008 21.270 Nb 4d45s 0.0 23.988 Mo 4d55s 0.0 40.094 132 DISSOCIATION ENERGY MODEL TABLE3.-continued element ground state promotion energy /kK Tc 4d55s2 d“-ls 3.702 d“-’sp 23.638 Ru 4d75s 0.0 36.278 Rh Pd 4d85s 4d’O 0.0 9.444 (48.0) <73.0 Ba 6s’ 12.998 17.648 f’-3d2s f”-3d5p La 5d6s2 3.867 19.078 Ce 4f5d6s2 3.409 19.444 Pr 4f 36s2 9.660 26.080 Nd 4f46s2 12.661 29.167 Pm 4f56s2 (14.4) (31) f”-2ds j”-2sp Sm 4f66s2 15.540 19.850 ELI 4f76s2 18.595 20.241 f n-3d2~ f” -3dsp Gd 4f75d6s2 9.177 20.195 Tb 4f 96s2 1 1.784 21.6 J’”-’ds f”-2sp DY 4f lo6s2 25.201 22.398 Ho 4f 116s2 27.146 22.812 Er 4f”6s’ 27.858 23.483 Tm 4f 136s2 29.362 24.088 Yb 4f146s2 35.235 24.875 d”-ls d”-’sp Lu 5d6s2 27.123 25.074 Hf 5d26s2 20.276 20.169 Ta 5d36s2 14.041 25.01 3 W 5d46s2 4.246 27.897 Re 5d56s2 16.912 27.265 0s 5d66s2 7.401 33.758 Ir 5d76s2 4.079 37.851 Pt 5d96s 0.0 43.390 Ac 6d7s2 13.261 19.730 Th 6d27s2 8.004 20.812 Pa U NP Pu Am Cm 5f 26d7s2 5f36d7s2 5f46d 7s2 5f67s2 5f 77s2 5f 76d7s2 8.991 (10.8) 21.455 (21.O) 14.597 (10.0) [f3d2s] [f4d2s] [f5d2s] [f 7ds] [f 7d2~] [fd3s] (23.0) 21.070 (21-0) 22.300 22.457 21.945 [fd2SPl [f 3dsp] Cf4dSP1 [f6sp] [f 7sp] [f 7dsp] promotion energies for those transition-metals lanthanides and actinides where one might have to consider the contribution of two configurations.With the various promotion energies available the procedure for calculation of L. BREWER AND J. S. WINN unknown dissociation energies involves the combination of the promotion energy for a given valence state with the interpolated bonding energies. For some elements with no unpaired electrons in the ground atomic state one calculates that no reasonable bonding energies could offset the promotion to even the lowest excited state and the cohesion of the atoms must be due primarily to van der Waals interactions. The noble gases the Group I1 elements Zn to Hg Be to Sr and Ra and the actinides Bk to No and probably Pu and Am fall into the van der Waals class.Most of these actinides have unpaired 5f electrons but the 5f electrons are so localized particularly for the second half of the series that they contribute insignificantly to the bonding. Ba is an exception among the Group I1 elements in that the 5d6s configuration is close enough to the ground 6s2 configuration to allow substantial contribution al- though the net contribution to AH,"/Ris still only 3 & 2 kK. For Eu and Yb like- wise the calculations indicate that they are not van der Waals molecules. For transition-metals of Groups 111-VI the valence state configurations are essen- tially the same for the diatomic and the solids in consisting of a mixing of the a"-ls and d"-2sp configurations with less p contribution for diatomic Zr Ta W and Group I11 and more p contribution for diatomic Hf than for the solid.A much more dramatic difference is found for Fe to Cu which can promote to a dn-2*5sp1.5 valence state in the solid but can only achieve dfl-1*5sp0.5 for Fe Co and Ni diatomics and no substantial promotion for Cu which uses the ground dlOs configuration. For the 4d and 5d Groups VII-XI all use the d'*-ls valence state for the diatomic with the exception of Re which is able to promote to d5sp. Examination of the bonding energies given in table 2 shows that the irregular behaviour of the dissociation energies of the diatomic is due to three contributions that change in different ways with variation of position in the Periodic Table. There is first the contribution from promotion energies which are known quite accurately for most elements.Secondly there is the increase of the s p bonding with increasing nuclear charge for a given period with a reduction in bonding per electron for multiple bonding and a reduction in p bonding when the core includes the closed s subshell of the outer shell. Thirdly there is the reduction in d bonding with nuclear charge for a given period up to the d5 configuration and an increase in bonding per d electron beyond the d5 configuration as the most localized orbitals are used by non-bonding electrons and the most extended orbitals are used by bonding electrons. The contri- bution of d bonding is greatly increased from 3d to 4d to 5d due to the contraction of the ns2np6subshell with increasing nuclear charge relative to the nd orbital.These same trends are found for the bulk metals and the simple smooth trends found for each of these factors makes the prediction of bonding energies and therefore dissociation energies quite straightforward and reasonably accurate. We thank J. H. Goble for his help in the analysis of the Mg and Ca potential functions. J. S. Winn acknowledges partial support from an Alfred P. Sloan Research Fellowship. This work was supported by the Division of Materials Sciences Office of Basic Energy Sciences U.S. Department of Energy under contract No. W-7405-Eng-48. J. H. Goble D. C. Hartman and J. S. Winn J. Chem. Phys. 1977 67 4206. J. H. Goble and J. S. Winn J. Chem. Phys. 1979 70,2051.J. H. Goble and J. S. Winn J. Chem. Phys. 1979 70 2058. J. H. Goble S. M. Walsh and J. S. Winn to be published. A. J. Thakkar J. Chem. Phys. 1975 62,1693. C. R. Vidal and H. Scheingraber J. Mol. Spectr. 1977 65 46. DISSOCIATION ENERGY MODEL 'W. J. Balfour and A. E. Douglas Canad. J. Phys. 1970,48,901; W. C. Stwalley Chem. Phys. Letters 1970 7,600. W. J. Balfour and R. F. Whitlock Canad. J. Phys. 1975 53,472. A. Kant and S. S. Lin Monatsh. 1972 103 757. loL. Brewer Viewpoints of Stability of Metallic Structures in Phase Stability in Metals and Alloys ed. P. Rudman J. Stringer and R. L. Jaffee (McGraw-Hill New York 1967) pp. 39-61 241-9 344-6 560-8. L. Brewer Science 1968 161 115. l2 The Committee on Data for Science and Technology (CODATA) of the International Council of Scientific Unions has critically evaluated the Doand AH of formation values for atomic and diatomic states of H N P 0 S C1 Br and I.The Brz and l2values were slightly revised to correspond to the recent values reported by Barrow et al. (R. F. Barrow D. F. Broyd L. B. Pederson and K. K. Yee Chem. Phys. Letters 1973 18 357). The S2value was changed in acknowledgement of the objection raised by Huber and Herzbergl3 to the use of a Do value that does not relate to the actual lowest rotational level of S2. The Do values in cm-' given in CODATA Report Part I Bulletin 5 (Dec. 1971) and Part V (Sept. 1975) were multiplied by hc/k = 1.4388 cm K to obtain the values in K. Brz and Clz differ from the others in not having a predominant isotope thus resulting in a small difference between Doand AH5 of dissociation of the dimer.The AH values reported by CODATA in J. Chem. Thermodynamics 1976 8,603 were converted to AH and divided by R = 8.314 33 J K-' but the uncertainties are those of the original D0" values. l3 K. P. Huber and G. Herzberg Molecular Spectra and Molecular Structure. IV Constants of Diatomic Molecules (Van Nostrand Reinhold New York 1979). l4 D. D. Konowalow and M. L. Olson J. Chem. Phys. 1979,71,450. l5 K. A. Gingerich J. Cryst. Growth 1971 9 31. l6 C. A. Stearns and F. J. Kohl High Temp. Sci. 1973 5,113. l7 C. Chatillon A. Michel and A. Pattoret Compt. rend. C 1975 280 1505. W. J. Balfour J. Chem. Educ. 1979 56 452. l9 A. Kant J. Chem. Phys. 1964 41 1872 and 1968 49 5144.E. Rutner and G. L. Haury J. Chem. Eng. Data 1974 19 19 obtained a different value upon repeating Kant's third-law calculations but they used the atomic weight of nickel rather than twice the atomic weight of nickel for the molecular weight of Ni2. 2o A Neckel and G. Sodeck Monatsch. 1972,103,367. 21 K. D. Carlson and K. R. Kushnir J. Phys. Chem. 1964,68 1566. 22 J. Drowart and S. Smoes J.C.S. Faraday 11 1977 73 1755. 23 S. K. Gupta and K. A. Gingerich Inorg. Chem. 1978 17 321 and J. Chem. Phys. 1979 70 5350 report values of D(Nbz) and D(Mo,) from third-law calculations which include no elec- tronic contributions in the calculation of -(Go -H&)/RT. Brewer and LamoreauxZ4 have pointed out that even with a IC ground state low-lying electronic states of higher multiplicity are expected to be populated at the temperature range of measurements and electronic terms must be included.The value given for Nbz has been corrected in a manner similar to the correction for M02.24 24 L. Brewer and R. H. Lamoreaux Atomic Energy Reuiew Molybdenum Part I Physicochemical Properties of Its Compounds and Alloys (International Atomic Energy Agency Vienna 1980). 25 D. L. Cocke and K. A. Gingerich J. Chem. Phys. 1974 60 1958. 26 V.Piacente G. Balducci and G. Bari J. Less-Common Metals 1974 37 123. 27 J. Kordis K. A. Gingerich and R. J. Seyse J. Chem. Phys. 1974 61 51 14. 28 D. L. Cocke and K. A. Gingerich J. Phys. Chem. 1971,75 3264. 29 M. Guido and G. Balducci J. Chem. Phys. 1972 57 561 1. 3o G.D. Blue R. S. Carbonara and C. A. Alexander Proc. 18th Annual ConJ on Mass Spectro- metry and Allied Topics San Francisco June 1970 and personal communiation reporting upper limit of 90 kcal mol-I for D(Pt2). 31 J. Drowart and R. E. Honig J. Phys. Chem. 1957 61,980. 3z D. R. Stull and G. C. Sinke Thermodynamic Properties of the Elements Adv. Chem. Ser. 1956 No. 18. 33 K. A. Gingerich High Temp. Sci. 1969 1 258. 34 R. Stern and N. Lang Lawrence Eivermore Laboratory San Francisco Bay Area Conference on High-Temperature Science and Technology 8th March 1979 report an upper limit of 15 kK for D(U2)compared to 20 41 5 kK reported by K. A. Gingerich and G. D. Blue J. Chem. Phys. 1967,47,5447 and 26 & 3 kK reported by L. N. Gorokhov A. M. Emel'yanov and Yu.S. Khodeev High Temperature 1974 12 1 156. L. BREWER AND J. S. WINN 35 C. E. Moore Atomic Energy Levels (U. S. Government Printing Office Washington D.C. 1949 1952 1958) VOI. 1-3. 36 L. Brewer J. Opt. SOC. Amer. 1971 61 1101. 37 W. C. Martin R. Zalubas and L. Hagan Atomic Energy Levels-The Rare-Earth Elements NSRDS-NBS 60 (U.S. Government Printing Office Washington D.C. 1978).

ISSN:0301-5696    

DOI:10.1039/FS9801400126

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Model Predictions of the Dissociation Energies of Homo-nuclear and Heteronuclear Diatomic Molecules of Two Transit ion Metals BY ANDRIES R.MIEDEMA Philips Research Laboratories Eindhoven The Netherlands Received 29th August 1979 It is demonstrated that by taking the solid metal at zero temperature as the reference state the enthalpy of formation of homonuclear diatomic molecules can be obtained as the surface energy of a piece of metal having the size of two atoms. It is suggested that the energy of larger homonuclear metal clusters can be derived similarly. The heat of formation of heteronuclear diatomic molecules relative to the two pure metal dimers is derived by means of a model description for the interaction energy which is very similar to that used for interfacial energies and alloy heats of solution in the condensed phase.1 INTRODUCTION In a series of papers1-’ we have recently demonstrated that the energies of metal- metal combinations can be described in terms of contact energy effects generated at the interface between dissimilar metal atoms. In our energy considerations the reference state we normally take is the pure condensed metal at zero temperature. If atomic cells of metal A are no longer completely surrounded by similar A-cells there is a change of energy which can be related to the change in atomic cell boundary condi- tions. It is assumed that this energy change is proportional to the fraction of the surface area of an atomic cell for which the boundary conditions have become differ- ent.This means that atomic cells are treated as if they were comparable with macroscopic pieces of metal. It follows that the atomic size is characterized by the value of Vg3,where V is the molar volume. The interfacial energy on a microscopic scale which forms the basis of our model description 1-3 of alloy heats of formation contains two terms (here we limit ourselves to combinations of two transition metals). The first term which is negative is introduced as the energy of an electrostatic dipole layer analogous to the dipole layer that would be generated at the interface between two macroscopic pieces of metal. The energy of the dipole layer can be expressed in terms of the difference in work function p of the two metals. The second term which is positive is derived from the difference in electron density nws,that would exist at the interface if the two types of metal atom cells were taken from the respective pure metals without any charge rearrangement.The electron density is required to become continuous across the interface. This is achieved by either6 changing the mixed electron configuration of the metallic atoms (for transition metals an interchange of d-electrons to s-electrons will increase the cell boundary A. R. MIEDEMA electron density) or by a volume change expanding the metal which originally had the higher value of nws and compressing the metal with the lower nws (or both). One would expect that the positive surface energy of a metal-vacuum interface could also be derived from the discontinuity in the electron density between a bulk metal atomic cell and empty space.Indeed it has been demonstrated4 that the surface energy of a solid metal at T =IZ 0 yo is related approximately linearly to nws the proportionality constant being a weakly varying function of the number of valence electrons per atom of the metal. For a metal-vacuum interface the macroscopic atom approach which is similar to that used in the description of alloy heats of formation suggests that the heat of vaporization of a metal is comparable with the surface energy of a piece of metal having the size of a single atom. Indeed such a relation does make sense3ys (see also below). In addition the heats of formation of homonuclear dimers are obtained when this type of approach is used as the difference in surface energy of pieces of metal consisting of a pair of atoms as compared with two single atoms (see below).Also the heat of formation of monovacancies in solid metals can be dealt with quite easily3y4 in terms of the surface energy of a hole of atomic size. A number of situations in which metal-metal and metal-vacuum interfaces play a role are surveyed in fig. 1 which shows the boundary conditions for atomic cells FIG.1.-Metallic adhesion on microscopic and macroscopic scales. The figure is intended to suggest that there are relations between the heat of formation of intermetallic molecules AB (relative to pure dimers) the heat of formation of intermetallic compounds AB (relative to pure metals) the heat of solution of A in B the interfacial energy between two crystals of A and B and the heal of adsorption of B on a substrate of A.in a variety of situations i.e. a heteronuclear AB molecule an ordered AB intermetallic compound the heat of solution of A in liquid B the interfacial energy between two crystals (ignoring the dislocation-determined density deficit) and the heat of adsorp-tion of an atom of B metal on a metallic substrate of A. In this example of A atoms that are clearly smaller than B atoms there is little difference between the heat of formation of the AB compound (per mol AB) and the heat of solution of A in B when the macroscopic atom approach is used. Once DISSOCIATION ENERGIES OF DIATOMIC MOLECULES assumptions are made about the geometry of atomic cells and interfaces the inter- facial energy between two crystals is directly related to the heat of solution.s The adsorption energy of B on A can be expressed in terms of the surface energies of the two metals and once again the heat of solution of A in B.One must make an assumption with regard to the fraction by which the adsorbed atom is in contact with the substrate. It has been found' that a fairly accurate description of experimental heats of adsorption of metals on metallic substrates is obtained by taking this fraction to be 0.35. The energy of the diatomic molecule is derived analogously. Compared with the average of the pure metal dimers the intermetallic molecule presents an interface between dissimilar A and B atoms and a slight increase in the B-vacuum interface (B larger than A).By expressing the interfacial energy in terms of the same para- meters (differences in work function differences in bulk cell boundary electron den- sities) that were used in the condensed phase we have been able to calculate the disso- ciation energies of diatomic molecules of two transition metals. Predicted values are presented in tables l(a)and l(b). TABLE1.-PREDICTEDVALUES OF THE DISSOCIATION ENTHALPIES OF DIATOMIC MOLECULES OF TWO TRANSITION METALS. DiB IN UNITS OF kJ mol-I. The starred values for homonuclear molecules are experimental. (4 Sc Ti V Cr Mn Fe Co Ni Cu 3d metals sc 158" Ti 128 125" V 213 183 238" Cr 207 163 191 151" Mn 139 96 133 94 42" Fe 230 177 185 131 80 100* co 292 232 229 167 123 129 167" Ni 316 262 272 205 163 163 199 229" cu 222 166 188 145 98 125 166 205 196" 4d metals Y 143 114 200 195 128 217 280 295 214 Zr 228 194 267 255 185 276 334 360 260 Nb 266 249 279 243 179 248 295 341 238 Mo 312 274 279 212 159 193 227 265 202 Tc 461 399 363 251 215 205 223 252 244 Ru 479 414 382 263 230 214 228 256 254 Rh 467 397 362 244 213 193 210 237 242 Pd 353 310 270 157 128 103 121 148 164 Ag 231 178 170 111 68 93 139 178 5d metals La 166 131 216 209 145 232 295 300 236 Hf 226 193 256 239 172 256 312 341 244 Ta 282 263 287 248 185 25 1 297 343 243 W 405 363 343 262 212 237 265 301 249 Re 513 450 395 276 242 227 241 268 264 0s 522 458 415 294 260 244 257 284 282 Ir 538 468 411 281 252 225 236 261 272 Pt 488 441 391 261 234 206 215 239 257 Au 369 314 258 159 127 116 144 175 Th 201 160 236 227 159 248 308 333 241 A.R. MIEDEMA 139 TABLE1-continued I___ (b) Y Zr Nb No Tc Ru Rh Pd Ag Y 156" Zr 212 309 Nb 250 333 371 NO 296 370 360 325 Tc 446 512 455 356 330 Ru 463 529 467 362 325 327 Rh 452 514 447 339 302 303 281" Pd 339 404 373 263 211 204 185 104" Ag 223 278 252 198 244 250 239 172 158" La 186 231 266 312 462 479 470 356 253 Hf 210 303 327 352 483 499 483 395 259 Ta 266 350 383 366 456 467 447 373 259 W 388 464 444 385 398 400 376 299 261 Re 496 567 504 395 349 343 317 231 273 0s 505 574 512 405 368 361 336 244 287 Ir 522 588 514 391 337 328 304 210 279 Pt 470 544 502 391 308 298 275 188 275 Au 357 420 373 258 234 231 216 139 Th 223 268 296 335 479 496 484 374 256 La Hf Ta W Re 0s Ir Pt Au La 241 * Hf 230 304 Ta 283 342 395 W 404 443 449 452 Re 511 535 505 436 376 0s 521 543 513 445 386 405 lr 537 555 513 427 352 371 333 Pt 484 516 502 425 328 339 299 278 Au 362 393 376 308 255 269 244 234 220" Th 251 263 312 427 531 540 555 508 392 The present paper is a summary of 3 earlier papers dealing with the heats of forma- tion of homonuclearsa and heteronuclear8b diatomic molecules and the special properties9 of molecules of divalent metals.In addition we shall extend the previous paper on homonuclear dimers to include predictions of the dissociation energies of larger homonuclear clusters.2. HOMONUCLEAR DIATONIC MOLECULES The starting point of the present description is that the difference between the energy of a homonuclear cluster of n atoms and that of n atoms in the bulk metal can be derived as the surface energy of a metallic particle. Defining Do as the zero ternpera- ture dissociation energy of a cluster into isolated free atoms we expect for diatomic homonuclear clusters that 2AHoVap -Do is linearly related to both the zero tempera- ture surface energy of the solid metal yo and the atomic size parameter VZ3(AHovap is the zero temperature heat of vaporization.). The validity of this linear relation is illustrated by means of fig. 2 in which all available experimental information" on 0' 140 DISSOCIATION ENERGIES OF DIATOMIC MOLECULES 800 600 7 z ? Gd" a 400 Ino / 0' X 3-/Pb -4 N 200 0 I I 0 0.4 0.8 1.2 y0V,2'3/ rnJ FIG.2.-Relation between the dissociation enthalpy * into free atoms of metallic dimers Do,the heat of vaporization" of the bulk metal AH!, and the surface energy4 per unit molar surface area ~O?ffn/~.dimers of transition metals and those of monovalent and trivalent non-transition metals has been included. Divalent metals have been omitted; the s2 outer free atom electronic configuration with large s -+p promotion energies leads to exceptional behaviour as will be dis- cussed at the end of this section. Tetravalent metals Si Ge Sn have also been omit- ted; they are considered to be at least partly covalent (i.e.? have more directional bonding than the average metal).Lead is considered to be a borderline case. The slope of the straight line in fig. 2 corresponds to (2AH:, -D0)/70V$3= 0.78 x lo9. It is of interest to compare this result with that derived for the surface energy of N spherical particles of volume np(Vm/N)for np = 2 (N is Avogadro's number np is the number of atoms in the cluster) which equals AHs,,f/~oVm= (36nN)1/3ni/3= 0.65 x lo9. (2.1) The agreement is quite convincing suggesting (1) that dimer interatomic separa- tions and bulk metal volumes are related and (2) that the surface energy remains a meaningful parameter for metal-vacuum interfaces of microscopically small pieces of metal. In fig. 3 we go one step further and investigate to what extent the energy difference between a pair of atoms in a dimer and two separated atoms can be regarded as a difference in surface energy.Again we find an approximate linear relation between Do and yoVz3. However due to the relatively small numerical values of Do [much smaller than (2AH,0ap-DO)],the scatter in the data points for the individual ele- ments is considerably larger than in fig. 2. The slope of the straight line drawn corre- sponds to D0/y0Vk/3= 0.265 x lo9. For spherical particles we would expect this factor to be (36nN)ll3(2 -22/3)= 0.17 x lo9. A. R. MIEDEMA 300 1 I 4 0 Rh 200 I + 0 E 7 Y \ -c ::100 Lio /oIn om Q 0 0.4 1.2 <Vi'3/ rnJ 0.8 FIG.3.-Relation between the dissociation enthalpy Do, of metallic dimers (in units kJ mol-' M2)and the product of the surface energy and molar surface area yoVz3,at zero temperature.Note however that the 3d metals Ti V Cr Mn Fe Co and also Pd that were represented in fig. 2 are omitted from fig. 3. It is suggested that the atomic closed shell d10electron configuration of Pd and magnetic energy contributions in the above 3d metals lead to the irregular behaviour of Do. In the case of a free atom the opti- mum magnetic state makes a larger contribution than in the dimer (or the solid). The reason for accepting that the above 3d metals (and Pd) as free atoms represent an exceptional situation while suggesting that the remaining metals of fig. 3 are the " normal" case is as follows.Combining the linear relations of fig. 2 (for 2AHta -Do and yoVg3)and fig. 3 (for Do and y0VZ3)we find that for the metals included in fig. 3 AHtapalone must also be linearly related to yoYz3. This is seen in fig. 4 (the open circles). The linear relation thus obtained also holds for a large number of 4d and 5d transition metals for which there is no information on Do,but for which surface energies4' and heats of vaporization" are known. The slope of the straight line in fig. 4 corresponds to AH$,,/y0Vm = 0.52 x lo9. This value has to be compared with 0.41 x lo9 the value expected for a pseudo-macroscopic spherical particle of volume VJN. It will be clear that having once accepted that the additional transition metals of fig. 4 represent the normal situation and are thus comparable with the other metals of fig.3 and 4 we can easily derive a predicted value for the missing transition metal dimer dissociation energies. They can be derived from the heat of vaporization the product of surface energy and molar surface parameter or the difference between the two. A suitable average result see ref. 8(a),is included in table 1. The results of fig. 3 and 4 imply that for a large number of metallic elements the dissociation enthalpy of the homonuclear dimer can be related to the heat of vaporiza- tion by Do(M2) = AHt,Ja (2.2) with a = 1.96. This value implies that the effective area of contact between two atoms in a homonuclear dimer equals about one fourth (0.255)of the surface area per atomic I42 DISSOCIATION ENERGIES OF DIATOMIC MOLECULES FIG.4.-Relation between the heat of vaporization of " normal metals " and the surface energy per unit molar surface area YOY:'~.The open circles represent the metals included in fig. 3; the closed circles represent transition metals for which there is no experimental information on Do. cell which seems a reasonable fraction. Relation (2.2) was suggested by Verhaegen et aZ.I2 These authors expected to find the same CI within a family of elements i.e. within a given column of the Periodic Table. It can be concluded from this section that a is an approximate constant for a large fraction of the metallic elements only the 3d metals Ti V Cr Mn Fe Co the divalent metals and Pd being clear exceptions. We emphasize at this point that the relation between 2AH,OaP-Do and y0Yg3is of a more general validity; only divalent metals are exceptions here.The exceptional properties of divalent mstals are illustrated by fig. 5. On the left-hand side we have indicated the position of free atom and dimer for the normal metals M in an energy scale in which the unit of energy is y0Vg3for the metal under consideration and where the solid metal is the zero of energy. Free atoms of the divalent metals are unusually stable as can be observed for instance in the sequence of first ionization potentials of elements as a function of atomic number. The free atom of Cd which in fig. 5 may also represent Hg Zn and Mg has an energy which is lower than that of the normal metallic dimer i.e.the dimer with a mixed electronic configuration comparable with that of the bulk solid. As a conse- quence admixture of 5p electronic wave functions is scarcely possible in the Cd-dimer. The pure s2 configuration of the free atom is retained leading to a van der Waals noble-gas-like molecule with for dimers an unusually large interatomic separation. The alkaline-earth metals Ca Sr and Ba are a borderline case. The free atom is clearly more stable than in the average situation but the free atom is slightly less stable than the " normal " metallic dimer would be. Hence whether we are concerned with a van der Waals type (large interatomic separation) molecule of Ca Sr or Ba or a more ordinary metallic one of mixed s-d configuration the dissociation enthalpy will in any case be small.A. R. MIEDEMA 0.75 0.50 C31?-0.25 0 FIG.5.-Exceptional and normal positions of free atom and dimer in an energy scale in which the unit is 109y0V:’3. The levels for M represent the majority of the transition metals and mono- and tri- valent non-transition metals. For the divalent metals Hg Cd Zn and Mg (represented there by Cd) the free atom energy level lies unusually low clearly below that of a “ normal metallic ” dimer. For Ca (also representing Sr and Ba) the free atom level is near to that of the normal metallic dimer so that metallic and van der Waals type dimers differ little in energy. For Fe (also representing other magnetic 3d-metals and Pd) the free atom energy level lies somewhat too low but the dimer is “ normal ” metallic.We expect the interaction potentials for a pair of alkaline-earth atonis (and to a minor extent also that for Mg,) to be quite unusual being only weakly dependent on interatomic separation over a wide range of distances. For both Mg and Ca there is evidence that the pair potential is indeed of such an unusual In fig. 5 we have included Fe which also represents other irregular 3d-metals. The dimer takes up a normal position in the energy scale but the dissociation en- thalpies are too low. It is of interest to note that the divalent metal Be falls into the Fe class. 3. LARGER HOMONUCLEAR CLUSTERS We have demonstrated that the energy of a cluster of two atoms relative to the bulk metal can be described fairly accurately as the surface energy of a spherical piece of metal of volume 2 VJN the difference being a factorf = 1.2.There is little reason to suppose that dimers are exceptional in this respect. Hence energies of larger clusters can also be derived from surface energies the above fact0r.f being expected to decrease fromf = 1.2tof = 1 with increasing n,. Effects of odd/even npare assumed to be relatively unimportant for larger n values. Intuitively we assumefto decrease asf= 1 + an;+; this assumption immediately gives us the energy of homonuclear clusters relative to the bulk solid in units of 0.41 x 109ni’3y0VZ3(see table 2). The dissociation energy relative to free atoms is obtained by subtracting the cluster surface enthalpy from n,AH&,. For the smaller clusters (dimers trimers) this means that we subtract two large terms so that uncertainties in the solid metal surface energies4 are multiplied.However for the larger clusters the uncertainties are much smaller. As an example we have included calculated values of the dissocia- ion energies (per mole of atoms) for clusters of Cu Pt or A1 in table 2. 1 44 DISSOCIATION ENERGIES OF DIATOMIC MOLECULES TABLE2.-PREDICTED VALUES OF THE HEAT OF VAPORIZATION OF HOMONUCLEAR METALLIC CLUSTERS RELATIVE TO THE BULK METAL AHsurrAND THE DISSOCIATION ENTHALPIES Do OF CLUSTERS OF Cu Pt AND A1 FOR VARIOUS NUMBERS OF ATOMS PER PARTICLE, nP. UNITSOF AHsurf ARE 0.41 x lo9 ni/3y0VA/3,i.e. THE SURFACE ENERGY OF A MACROSCOPIC SPHERICAL PARTICLE. UNITSOF DoARE' kJ mol-' atom-'.3 1.17 108 191 143 4 1.16 131 229 162 6 1.14 160 276 185 8 1.125 178 305 199 10 1.115 191 326 210 20 1.09 223 379 236 50 1.07 255 431 261 100 1.055 272 460 276 200 1.04 286 482 287 500 1.03 300 505 298 1000 1.025 307 517 304 Note the difference from fig. 2 and 3 and table 1 where units of Do are per mol of two atoms. There are few experimental data with which to make comparisons. Wu15 has studied the dissociation enthalpy of Li,. The experimental value of Do is 58 kJ mol-1 atom-' which can be compared with a calculated value of 75 kJ mol-' atom-l. Gingerich et a1.16 have reported on Pb3 and Pb,. The experimental values of Doare 73 and 104 kJ mol-' atom-' to be compared with predictions based on the macro- scopic atom considerations of 76 and 87 kJ mo1-I atom-' respectively.If one accepts the macroscopic atom approach as a correct method of estimating cluster dissociation energies one has a powerful criterion for evaluating theoretical calculations of cluster energies. For instance Anderson l7 has calculated that the binding energy per atom for CuL3clusters is much smaller than that of smaller clusters. (76 kJ mol-' atom-'). Anderson's prediction strongly disagrees with present macroscopic atom considerations. As another example we point to theoretical investigations of the binding energy of hydrogen with metals by means of cluster calculations18 with and without added hydrogen. One should correct calculated binding energies for the increase in the cluster surface energy by adding z2 cm3 per H atom to the cluster molar volume.In principle the atomic volume of small metal particles will be reduced because the surface energy represents an appreciable fraction of the binding energy. However when the volume compression is estimated from surface energy and solid metal bulk modulus K the effect is found to be not very large. For macroscopic spherical particles -AAH~~~~/AH~~~~ = 0.09 x 109n;1/3y0~,/~~,,,. Values of y0Vi'3/KVm are assembled in fig. 6. For transition metals they vary around y0V$3/KVm= 0.5 x corresponding to energy corrections of a few percent even for np = 2. Note that if the change of atomic volume with cluster size plays a role it will be seen first of all for Li clusters. We note that the calculations of table 2 cannot be used to derive dissociation ener- gies of relatively small molecules of the divalent metals Mg Zn Cd Hg.It can be A. R. MIEDEMA 0 li oNa,K .Zr oAs Ti Hf .V oFe Mn .Co oNi snsPb .Ru a Pd Rh Ir .Pt *Au 8 *Os 1 I 1 1 I 1 I 1 I ? t 2 3 4 5 6 7 8 9 10 z FIG.6.-Parameter yo V23/KVmwhich characterizes the relative volume contractions of small particles generated to reduce the total energy of the particle-vacuum interface. K is the bulk modulus yo the surface energy and V the molar volume. For the majority of transition metals y0V,?3/KVmw 0.5 x 10-9. shown’ that Hg particles of up to 20 atoms will be van der Waals type molecules. For Mg Zn Cd this critical number lies around np = 5.Also the more covalently bonded molecules of Si Ge or Sn cannot be expected to be described by the present surface energy considerations. 4. DIATOMIC INTERMETALLIC MOLECULES In a recent papeeb we have suggested a method for calculating dissociation energies of intermetallic molecules. The interactions between two atoms in a molecule are treated in a similar way to those between dissimilar nearest neighbour atoms in solid alloys. There are two differences in the heat of formation of condensed alloys the reference states are the two pure metals. Consequently the heat of formation of intermetallic molecules will be derived relative to the two pure metal dimers. Secondly the contact area between dissimilar atoms in a molecule is only a fraction of the atomic surface area.From the observed relation between the dissociation energy and the heat of evaporation of “normal metals” we derive this fraction as 1/2a=0.255 for atoms of equal size. Where atoms are of different size (V&I3> VZ3),we assume that the contact surface area is determined by the smaller atoms so that there is a positive contribution to the heat of formation of the molecule reflecting an increase in the B-atom-vacuum interface. As a result AH^^^^,^^ = 0.13 x 109~;(v;/3 -~23). (4.2) For AHchem,AB we start out from the corresponding relation for the case of condensed alloys; from the effective contact surface between A and B (0.255 of that of atom A) DISSOCIATION ENERGIES OF DIATOMIC MOLECULES we would expect AHchem,AB to be about one fourth of the heat of solution of A in €3 AHsol,Ain B.This quantity can be calculated from Both terms in relation (4.3) are proportional to the A-B interfacial area per atom A i.e. Vy3. The first term represents the energy of an interfacial electrostatic dipole layer; the average electrostatic shielding length is found in the denominator. The second term is due to the difference in the electron densities at the boundaries of the atomic cells in the two pure metals which must be equalized in the alloy. P and Q are empirical constants. Values for the effective work function q* and cell boundary electron densities nws can be found in previous paper^.^^'^ Previously we have combined the two terms to Analysing experimental data for diatomic we found that AHchem,AB could not be obtained by simply reducing AHso,,Ain by the above factor of 0.255 (accounting for the difference in A-B interface area).Instead the two terms in rela- tion (4.4)must be treated differently Q' is indeed reduced by a factor of z 0.255 but P is only slightly reduced. As a consequence the ionic term in the heat of formation is very much the same for molecules as for an ordered AB compound. Values of P and Q' used in the present calculations of AHche,,,,AB for molecules can be found in ref. [8(b)]. Results for DiBhave been assembled in table 1. Some additional comments are (a)In calculating DiBfor molecules in which one of the metals is Cu Ag or Au we have not used the experimental value of DiAin relation (4.1).It is suggested that the symmetric dimer of a metal with a single s electron and otherwise closed electron shells will have some additional s2 stability. This stability is observed in fig. 3 where data points for the dissociation energies of Li, Na, Cu, Ag, Au are all above the average line. We have introduced " ordinary metallic " Cu, Ag and Au with dis- sociation energies' of 165 140 and 180 kJ mol" respectively (points on the line in fig. 3). The correction is not very substantial. (b) In the present paper we treat the metals at the end of a transition metal series Ni Pd Pt and Au slightly differently from other metals. In the Periodic Table the electronegativity increases as a rule with increasing total number of electrons. There are exceptions there is a steep fall in q* from Ni to Cu from Pd to Ag and from Pt to Au to Hg.Since in the case of metals in a molecule which have an appreciable difference in q* the electron transfer may well become of the order of 1 electron atom" the resistance to charge transfer is higher than average if the electronegative metal is Ni Pd Pt or Au. In order to include this effect in our predictions of DABwe have used the following approximation. The value of P as given previously,' is reduced by 7 13 or 19% if one of the metals in the diatomic molecule is Ni Pd Pt or Au and Ay* lies between 1 and 1.5 1.5 and 2 or 2 and 2.6 V respectively. (c) In fig. 7 the calculated values of DOAB of table 1 are compared with experimental data. On the whole the agreement is quite satisfactory although in the region of small dissociation energies the relative deviations are fairly large.In the deviating cases we are concerned with molecules of Au with a magnetic 3d metal. For molecules such as RhLa IrTh and RhTh for which Aq* is large the dissocia- tion energies range up to very high values. It appears that whereas for homonuclear diatomic molecules the binding energy per atom varies around 1/4 of that of the solid metal this fraction may increase to above 0.4 for the strongly ionic intermetallic A. R. MIEDEMA t 600 RuThe IrTt/' PtTho / / 400 RuYy ePtLu RhV/ RhTi c /# AuLa 'd 0 € /-AuLu 7 Y I AuNi /-\ -.. AuCre. AuCo Ii".Rh :200 AuFewAuM,!?' CuNi 0 0 t .cucrdCuCo AgMn /AuPd I // '/ I 1 I I I 0 0 200 400 600 Doca,c /kJ mol-' FIG.7.-Comparison of experimental and calculated values of the dissociation energy of the diatomic molecules of two transition metals.Experimental values are from Gingerich,lo calculated values from table 1. The straight line corresponds to D",,,, = molecules. Consequently it is possible that molecules consisting of a metal from the left-hand side of the series of transition metals with one from the right-hand side have a higher dissociation energy than the homonuclear dimer of a transition metal in the middle of the series (Mo W). (d) In principle it is also possible to calculate heats of formation of dimers of a transition metal with a polyvalent non-transition metal. In this case the corre-sponding heat of formation of condensed alloys contains a large negative term from d-p hybridization which does not depend much on which transition metal is com- bined with which p-metal partner.It has been foundsb that apparently such a hybrid- ization term also plays a role in molecules. Experimental results mainly on transi- tion metal silicides suggest that the hybridization term like the ionic term is relatively larger in molecules. Since experimental data are restricted mainly to silicides and germanides quantitative predictionssb for d-p intermetallic molecules are less reliable than the predictions for molecules of two transition metals of table 1. 5. CONCLUSIONS We have argued that dissociation energies of metallic clusters can be related to surface energies and interfacial energies such as occur in metallurgy.Predictions of the dissociation energies of diatomic molecules of two transition metals have been tabulated. It is suggested that the binding energy of large homonuclear clusters can be easily derived relative to the bulk metal by comparing it with the surface energy of a metal sphere of the same volume. DISSOCIATION ENERGIES OF DIATOMIC MOLECULES A. R. Miedema F. R. de Boer and R. Boom J. Less-Common Metals 1976,45,237. A. R. Miedema J. Less-Common Metals 1976 46 271. ’A. R. Miedema and P. F. De Chatel Proc. Symp. Theory of Alloy Phase Formation ed. L. Bennett (Amer. Soc. for Metals New Orleans 1979). A. R. Miedema 2.Metallkunde 1978 69 287; 1979 70 345.A. R. Miedema and F. J. A. den Broeder 2.Metallkunde 1979,70,46. A. R. Williams Proc. Symp. Theory of Alloy Phase Formation ed. L. Bennett (Amer. SOC. for Metals New Orleans 1979). ’J. A. Alonso and L. A. Girifalco J. Phys. F 1978 8 2455. A. R. Miedema and K. A. Gingerich J. Phys. B 1979,12,2081 and 2255. A. R. Miedema and J. W. F. Dorleijn Phil. Mag. 1979 submitted. lo K. A. Gingerich Current Topics Muter. Sci. to be published. l1 R. Hultgren P. D. Desai D. T. Hawkins M. Gleiser and K. K. Kelley Selected Values of the Thermodynamic Properties of Metals and Alloys (Amer. Soc. Metals Ohio 1973). l2 G. Verhaegen F. E. Stafford P. Goldfinger and M. Ackermann Trans. Faraday Soc. 1962 58 1926. l3 R. 0.Jones J. Chem. Phys. 1979,71 1300. l4 W. J. Stevens and M.Krauss J. Chem. Phys. 1977 67 1977. l5 C. H. Wu J. Chem. Phys. 1976 65 3181. K. A. Gingerich D. L. Cocke and F. Miller J. Chem. Phys. 1976 64,4027. l7 A. B. Anderson J. Chem. Phys. 1978 68 1744. la A. Lodder personal communication 1979.

ISSN:0301-5696    

DOI:10.1039/FS9801400136

出版商:RSC    

年代:1980

数据来源: RSC