X-ray Photoelectron Spectra of Copper( I) - ACS Publications

Mar 12, 1980 - Department of Chemistry, Purdue University, West Lafayette, Indiana 47907. ... Abstract: The X-ray photoelectron spectra (X-ray PES) of...
1 downloads 0 Views 582KB Size
GagnC, Walton, et al.

/

1905

X-ray PES of Copper(I) and Copper(II) Complexes

X-ray Photoelectron Spectra of Copper( I) and Copper( 11) Complexes Derived from Macrocyclic Ligands' J. L. Allison,2aC. A. Koval,ZaW. S. Mialki,2b R. R. T. J. Smith,2aand R. A. Walton*2b Contribution No. 6050 f r o m the Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91 125, and the Department of Chemistry, Purdue University, West Lafayette, Indiana 47907. Received August 3, I979

Abstract: The X-ray photoelectron spectra (X-ray PES) of pairs of mononuclear Cu(l1) and Cu(1) complexes of several macrocyclic ligands show that there is a chemical shift of -2.5 eV between the Cu 2p binding energies of these two oxidation states. A comparison of these data with that for mononuclear Cu(II1) complexes of biuret and the tetradentate tripeptide ligand derived from a-aminoisobutyric acid shows a shift of -2 eV between Cu(II1) and Cu(I1). These X-ray PES measurements do not support the alternative formulation of certain of the Cu(1) complexes as Cu(1l) or Cu(II1) species with reduced ligands. Thus a previous suggestion that Cu1(TAAB)(N03) is in reality [Cu"'(TAAB2-)](N03) is not in accord with the present X-ray PES results. The preceding information has been used in the analysis of the X-ray PES of binuclear copper complexes of the macrocycle (L) which is prepared by condensing 1,3-diaminopropane with 5-methyl-2-hydroxyisophthaldehyde.The oxidation-state formalism which has been used to describe the complexes Cu'CulL, [ C U W I ~ L ] C I Oand ~ , [CuilCui(CO)L]C104 is consistent with the observed Cu 2p binding energy spectra. However, the CuI1Cuiicomplex [ C U ~ ~ C U ~ ~ L ] ( C I O ~ ) ~ . ~ H ~ O doublets is attribreveals some unexpected and novel features in its Cu 2p spectrum. The appearance of two sets of Cu 2~112,312 uted to the X-ray induced one-electron reduction of -SO% of the Cu(I1)-Cu(I1) complex to a symmetrical delocalized species Cu(+l .S)-Cu(+l S)possessing Cu 2p binding energies which are intermediate between those of Cu(l1) and Cu(1). This novel species also displays a characteristic satellite structure. Similar behavior is exhibited by the chloride complex C U ~ ~ C U ~ ~ L C I ~ . 6 H 2 0 , but with the mixed metal complex [ C U ' ~ Z ~ ~ ' L ] ( C I O ~the ) ~one-electron ~H~O reduction is believed to produce the dIo-dlo Cu(1)-Zn(1l) species.

Introduction The four-coordinate copper(1) ~ o m p l e x2, , ~has been shown to react with a variety of monodentate ligands (Scheme I, L' = isonitriles, phosphines, CO, amines, nitrile^).^^^ The resulting diamagnetic adducts, 3, are five coordinate, an unusual coordination number for ~ o p p e r ( I ) . ~An J alternative formulation for both the adducts, 3, and the four-coordinate precursor, 2, regards these species as containing Cu(1I) dr Cu(II1) complexed to a reduced macrocyclic ligand. The latter explanation is comparable to that invoked to account for the chemical properties of certain iron and nickel dithiolate complexes.6 Scheme I

Oxidation state formalism questions also exist for the binuclear complex 5 (Scheme 11). Prepared by one-electron reduction of the CU"CU'~ complex 4, complex 5 is, formally, a mixed-valence, Cu"Cui ~ o m p l e xThe . ~ mixed-valence complex 5 forms a carbonyl adduct 6 , apparently another example of five-coordinate ~ o p p e r ( I )The . ~ mixed-valence species 5 can also be further reduced, by one electron, to give a black, crystalline, diamagnetic product, 7, presumably containing two copper(1) ions.7 Complexes 4-7 have been examined by electron paramagnetic resonance, electronic absorption spect r o ~ c o p yand , ~ crystallographic methods,* but continue to pose intriguing questions of bonding, intramolecular electron transfer, and metal oxidation states. This paper presents the results of X-ray photoelectron spectral studies on the complexes 1-7, Cu11(trans-diene)2+(8) and its copper(1) analogue 9, and CU"(TAAB)~+(10) and its

11

copper(1) analogue 11, along with model complexes, in an attempt to better define the electronic structures of these species. 3

L' = (a) R N C , ( b ) C O , ( c ) R,P, ( d ) R , N , ( e ) l-methylimidazole, etc. 0002-7863/80/1S02-1905$01 .OO/O

Experimental Section The perchlorate salts of complexes 1,44,9 5,7 6,7 8,1° and 9 , l O the nitrate salts of 10 and 11,11and the neutral complexes 2,4 3 (L' =

C2 1980 American Chemical Society

1906 Scheme I1

Journal of the American Chemical Society

/

102.6

/ March 12, 1980

close proximity (these complexes contain the pairs Cu(1I)c u ( I I ) , C U ( ~ I ) - C U ( I )or, Cu(I)-Cu(I)). The other macrocycles, trans-diene and TAAB, are as shown in the Introduction. The appearance of very weak Cu 2p peaks a t -955 and -935 eV in the X-ray PES of Cu(LBF2), Cu(LBF2)(1,’), Cu(trans-diene) (C104), and Cu( TAA B) ( NO3) arises from small amounts of surface oxidation to copper(I1) during our X-ray PES sample preparations. This occurs in spite of our attempts to prevent such a decomposition by resorting to the use of a nitrogen-filled drybox. Upon deliberately exposing these complexes to the air (for periods of up to 30 min) and rerunning their X-ray PES, we observe a dramatic increase in the amount of copper(I1) contamination, the intensity of the copper(I1) peaks now exceeding that of their lower binding energy counterparts.

Discussion Copper 2p Binding Energies in Mononuclear Complexes. The Cu 2p binding energies of the copper(I1) complexes listed in the top half of Table I and those of complexes 1, 8, and 10, which have been formulated as derivatives of copper(II), are in good agreement with literature X-ray PES data for copper(1I) Furthermore, the occurrence of shake-up eV to the high binding energy side of ~ a t e l l i t e s l ~a .t ’-8-10 ~ the primary 2pl/2.3/2 peaks confirms this conclusion. The chemical shift of approximately -2.5 eV between the Cu 2p peaks of Cu(LBF2)(C104) ( 1 ) and Cu(LBF2) (2) (and its derivatives with carbon monoxide (3b) and I-methylimidazole (3e), together with the absence of satellites in the Cu 2p spectra of the latter complexes, is characteristic of the difference between structurally related copper( 11) and copper( I ) species.13116*18 Therefore, the f o r m ~ l a t i o of n ~Cu(LBF2) ~ and Cu(LBF2)(L’), where L’ = C O or I-methylimidazole, as copper(1) complexes is fully substantiated. Similarly, the Cu 2p binding energies of the complexes Cu(trans-diene)(CIOJ) (9) and Cu(TAAB)(N03) (11) are between 2.5 and 3.0 eV lower than those of the copper(I1) species 8 and 10, consistent with the formulation of the 9 and 11 as complexes of copper( I). This interpretation is further supported by the absence of the characteristic copper( 11) shake-up satellite structure.13~’ The results of the Cu 2p binding energy measurements on 2, 3, and 11 are of further significance insofar as they relate 7 to the q ~ e s t i o Inof~ whether ~ ~ ~ ~ these complexes are genuine CO),4 3 (L’ = I-Melm)? and 7’ were prepared by previously reported copper( I ) species or, alternatively, contain copper( 11) or copprocedures. All reduced, copper(1) complexes were stored under heper(lI1) bound to a reduced macrocyclic system. Our X-ray lium in a Vacuum Atmospheres controlled atmosphere chamber prior to use. Analytically pure samples of the complexes [ C ~ ( p y ) 2 C l 2 ] ~ , PES results clearly favor the former formulation and do not [Cu(bpy)CI*],,, [Cu(phen)Cl2],, Cu(terpy)Cl2, Cu(trien)(ClO4)2. support the contention of Katovic et a1.I’ that [Cu(TAAB)]+ H 2 0 , and Cu(tren)Clz were kindly provided by Drs. G. Cayley and ( 1 1 ) is in reality [Cu1I’(TAAB2-)]+. In fact, X-ray PES D. P. Murtha. measurements on the genuine copper(II1) complex of biuret, X-ray photoelectron spectra were recorded using a Hewlett-Packard KCu(H-zbi)2, shows that its Cu 2p energies (Cu 2~312at 936.7 Model 5950A ESCA spectrometer. Monochromatic aluminum K a l , ~ eV) are -2 eV higher than those of its copper(l1) analogue, radiation (1486.6 eV) was used as the X-ray excitation source and the KzCu(H-2bi)2 (Cu 2p3/2 at 934.8 eV).19 This is supported by powdered samples were dispersed on a gold-plated copper surface. An measurements on the copper(lI1) complex Cu(H-~Aib3), electron “floodgun” was used in conjunction with this instrument to where H-2Aib3 is the tetradentate tripeptide derived from eliminate, or at least reduce to a minimum, surface charging effects. Additional experimental details are described fully elsewhere.I2 a-aminoisobutyric acid,20which has a Cu 2 ~ 3energy , ~ of 936.2 eV.*I Accordingly, there is no X-ray PES evidence for 1 1 being Results a derivative of copper(I11).22 The pertinent X-ray PES data for monomeric copper(1) and Although it is not our intent to give a detailed consideration copper(l1) complexes derived from macrocyclic ligands, toto the N Is spectra of the macrocyclic ligands, a couple of gether with data for several standard copper(I1) complexes points are worth mentioning at this time. First, in the case of containing nitrogen ligands, are presented in Table I. Related Cu(TAAB)(N03)2 (10) and Cu(TAAB)(N03) ( l l ) , the N binding energies for the binuclear copper complexes of a Is intensity ratios for the two types of nitrogen (NO3 a t -406 macrocyclic ligand are listed separately in Table 11. In the case eV, TAAB at -400 eV) change from 1:2 to 1.4 in accord with of complexes 1-3, LBF2 is used as the abbreviation for the the stoichiometry differences between these two complexes, macrocyclic ligand l,l-difluoro-4,5,11,12-tetramethyl-l- thereby confirming the structural integrity of these complexes bora-3,6, IO, 13-tetraaza-2, 14-dioxacyclotetradeca-3,5,10,- during the X-ray PES measurements. Second, in the N Is 12-tetraenato, while L is used for the macrocycle in the binuspectrum of Cu(LBFZ)(C104) (I), two well-resolved peaks of clear complexes 4-7, wherein two copper atoms are held in equal intensity, are observed, a gratifying result in view of the

Gagn;, Walton, et al.

/

1907

X-ray PES of Copper(I)and Copper(l1) Complexes

Table I. X-ray Photoelectron Spectra of Copper Complexes of Nitrogen-Containing Ligands complexb

c u 2P312

Cu(tren)C12 C U (L B F2) ((2104) e Cu(LBF2) CU(LBF2) (CO) Cu( LB F2)( 1-MeIm) Cu(trans-diene)(C104)2 Cu(trans-diene) (CIO4) Cu(TAAB)(N03)2 Cu(TAAB) ( N 03)

934.5 (2.0) 934.0 (1.7)

399.4 (1.1) 399.3 (1.1)

934.3 (1.7) 934.9 (1.5)

399.4 (1.2) 399.9 (1.1)

93424 1.7)

399.9 (1.3)

934.8 (1.8)

400.0 (1.7)

1

935.4 (1 .7)c

2

9 10

932.9 (1.6)" 933.1 (I.6)d 932.9 ( 1 . 8 ) d 935.2 (1.5)c 932.5 ( 935.4 (2.0)C

11

932.9 (1.8)"

3b 3e 8

binding energies, eVu N Is c 1s

CI 2PWz

284.5 284.5 285.5 sh 284.8 285.3 286.3 sh -285.0 sh 285.7 285.9

399.7 (1.4) 401.1 (1.5) -399.9 (3.0) -400.0 (4.0) 400.0 (3.3) 399.8 ( I .5) 400.0 (1.7) 400.2 (1.3) 406.3 (1.4) 399.6 (1.8) 406.2 (1.8)

F Is

197.4 197.6 207.9 198.0

285.4 286.4 sh 285.1 285.1 285.2 285.2 285.4 285.1

207.5

686.1 685.7

207.5

285.0

Full width half-maximum values (fwhm) for the Cu 2~312and N 1s peaks are given in parentheses. The Cu 2~112component is located at 20.0 f 0.2 eV above that of the 2~312peak. Ligand abbreviations: pyridine (py), 2,2'-bipyridyl (bpy), I,l0-phenanthroline (phen), 2,2',2"-terpyridyl (terpy), triethylenetetramine (trien), tris(2-aminoethy1)amine (tren). Weak peak located at -933.0 eV due to small amounts of Cu(1) produced by X-ray photoreduction. The intensity of this peak decreased further upon reduction in the X-ray power from 1 kW to 600 W. A weak peak centered between 934.8 and 935.2 eV arises from trace amounts of surface oxidation of this air-sensitive Cu(1) complex to Cu(I1). This occurs during sample preparation (see Discussion). e The complex was studied as its dioxane solvate Cu(LBF2)(C104).

Table 11. X-ray Photoelectron Spectra of Binuclear Copper Complexes of a Nitrogen Macrocycle binding energies, eVa CU 2P3/2 N Is

complex [CU"CU"L](CIOA,)~.~H~O

4

[ C U " C U ' L ] C I O ~ - O . ~ C H ~ O H5 [cu"cu'(co)L]c1o4

6

CU'CU'L

7

935.4 (-2.0) 934.2 (-2.0) 935.3 (2.6) 932.3 (1.9) 935.5 (2.5) 932.7 (2.0) 932.5 (1.6)

399.4 (1.2) 399.1 (2.0) 399.5 (1.9) 399.6 ( 1 S)

Full width half-maximum values (fwhm) for the Cu 2p3p and N Is peaks are given i n parentheses. The Cu 2p1/2 component is located at 20.0 f 0.2 eV above that of the 2p3/2 peak. I n all instances the C Is peak was close to 285 eV.

two different nitrogen environments present; we presume that the higher energy peak arises from the nitrogen atoms which are bound to the 0 atoms of the O2BF2 bridging unit. Unfortunately, with Cu(LBF2) and Cu(LBF2)(L'), the two nitrogen peaks are not resolved. However, the fwhm values (3.0-4.0 eV) of the resulting broad N 1s band (Table I) are comparable to the overall width (2.9 eV) of the resolved doublet in the case of the N Is spectrum of Cu(LBF2)(C104). Copper 2p Binding Energies in Binuclear Complexes. The Cu 2p chemical shift difference of approximately 2.5 eV observed for the pairs of structurally related mononuclear copper(1) and copper(1I) complexes in the preceding section provides a basis for considering the related X-ray PES data of the binuclear species [CuzL] n+ (n = 0, 1, or 2), 4-7. In the case of the binuclear complex Cu'Cu'L (7), a sharp Cu 2p3p peak at 932.5 eV (Table I1 and Figure 1 ) is characteristic of the Cu(1) oxidation state in a nitrogen macrocycle. The weak feature close to 935 eV (Figure Id) is due to a trace of Cu(I1) surface contaminant, an interpretation which was confirmed by deliberately exposing this air-sensitive complex to the atmosphere (Figure le). The dramatic increase in the intensity

L I

950

I

I

I

940

I

a

L

930

GV

Figure 1. X-ray photoelectron spectra in the region 950-925 eV showing , ~ and associated satellite structure: (a) the primary Cu 2 ~ 3 photolines [ C U " C U " L ] ( C I O ~ ) ~ . ~ H(b) ~ O[ C ; U ~ ' C U ' L ] C I O ~ O . ~ C H (c) ~OH [CUI'; Cu'(CO)L]Cl04; (d) Cu'Cu'L; (e) spectrum of Cu'Cu'L after exposure to the air for 30 min.

of this latter feature, together with the appearance of the attendant characteristic Cu(I1) satellite structure a t -945 eV, attests to the rapid surface oxidation of this complex. The mixed-valence complexes 5 and 6 both contain the expected pair of Cu 2p1/2,3/2 doublets whose energies closely match those expected for mixed Cu(I1)-Cu(1) species (Figures lb,c). Like the related X-ray PES data for 7,the spectra of 5 and 6 showed no dependence on X-ray flux (X-ray power

1908

Journal of the American Chemical Society

/ I02:6 / March 12, I980

within experimental error the same, so that this complex could be an example of a delocalized mixed-valence complex containing very weakly coupled metal centers. Consequently, we were left with the problem of distinguishing between a genuine "trapped-valence" mixed oxidation state complex and one which has a delocalized symmetric ground state. Accordingly, we have recently embarked on the solution of the crystal .!=6 structure of [Cu"Cu'L]Cl04. This structure solution clearly shows8 that this molecule has an unsymmetrical ground state, in accord with our initial interpretation of the X-ray PES results. This is similar to the crystallographic results for the mixed Co( 1I) -Co( I I I ) complex [CoI1Co'"LBr2( H20)2] +Brwhich was previously investigated by Hoskins and W i l l i a m ~ . ~ j I I 940 935 930 Therefore, to our knowledge, there are still no examples of mixed-valence metal complexes which are known to possess eV a symmetrical delocalized ground state and for which two sets Figure 2. Cu 2~312primary photolines of [Cu"CuiL]C10~0.5CH~OH of metal core binding energies have been observed, despite recorded using different X-ray fluxes: (a) 600 W; (b) 1000 W. Both spectra efforts to search for such examples.26 were recorded for ca. 20 min, and their deconvolution using a Du Pont 310 curve resolver (employing a Gaussian-shape fit) showed that the intensity The final complex of interest to us in the present study was ratio /[Cu(ll)]/l[Cu(l)] was 1.0 f 0.1. Full width half-maximum values the binuclear Cu(I1) species 4. To our surprise, rather than the were 2.6 f 0.1 eV for the Cu(l1) component and 1.9 f 0.1 eV for expected single Cu 2pi/2,3/2 spin-orbit doublet and its assoCu(1). ciated satellite structure, two sets of primary photolines were observed (Table I1 and Figure la). The higher energy 2p1/2.3/2 varied between 600 W and 1 kW), but unlike 7 these spectra doublet was characteristic of Cu(I1) while the lower energy were unaffected by prior exposure of the samples to the atset was intermediate in value between that expected for momosphere, an observation which is in keeping with the relative nonuclear Cu(I1) and Cu(1). Deconvolution of the spectrum stability of these two complexes to aerial oxidation. Deconshowed that the two sets of peaks were of similar intensity and volution of the Cu 2p spectra of mixed-valence 5 and 6 (Figure possessed very similar fwhm values (1 3-2.0 eV). Furthermore, 2) confirmed ( I ) that the Cu(1) component peaks were sigan additional feature in the X-ray PES of 4 was the presence nificantly narrower than those due to Cu(I1) (this is expected of a second set of satellite structure (at -6 eV above the pridue to the broadening of the latter via multiplet splitting efmary Cu 2p photolines). This spectrum was reproducible under f e c t ~ and ) ~ ~(2) that the I[Cu(II)]/I[Cu(l)] intensity ratio was a variety of different experimental conditions. These included 1.O f 0.1, The presence of two different copper environments variations in the probe temperature (from room temperature in 5 and 6 is supported by their N I s binding energy spectra. to -40 "C), the time of irradiation (varied between 5 min and The N Is peaks (Table 11) are significantly broader than that I O h), and the X-ray power (varied from 400 to 1000 W).27 for the symmetric [CulCulL] species in which the four nitrogen The one minor spectral variation was the presence of a weak shoulder a t -932.5 eV in the Cu 2p spectrum of some of the atoms are indistinguishable. samples. This was due to the formation of small amounts of The similarity of the X-ray PES of 5 and 6 argues that a t "normal" Cu(1) species. Different synthetic batches of the least on the X-ray PES time scale (-IO-l7 s) we are dealing complex showed the same spectral features. Monitoring the with genuine mixed-valence copper(I1)-copper( I) complexes. This is certainly the case with 6 , for which EPR data show7 that C 1 s, N 1s, 0 Is, and CI 2p spectra provided no evidence for radiation damage involving the ligands. For example, the CI the odd electron is localized on a single copper center. With 5 , 2p binding energy spectrum (CI 2~312at 207.7 eV) showed that electronic and EPR spectra provide evidence7 of interaction the only chlorine-containing anion present was C104-. between the Cu(1) and Cu(I1) centers, although this is markedly temperature dependent. In the case of the E P R experiTo check whether the unusual Cu 2p X-ray PES of 4 was ments, a seven-line isotropic spectrum (in CH2C12) a t 25 O C specific to this complex or typical of other binuclear Cu(l1) arises because of the interaction of the odd electron with both species of this same ligand, we investigated the X-ray PES of the analogous chloride complex Cu"Cu"LC12.6H20, a species copper centers ( I = Y2), and the intramolecular electron which contains coordinated chloride and square-pyramidal transfer therefore occurs a1 a rate which is rapid compared to Cu(I1) centers,28 and of the mixed Cu(I1)-Zn(I1) complex the relatively slow EPR experiment ( 10-8-10-4s). However, in frozen solutions the observed four-line (gl,)anisotropic [ C U ~ ~ Z ~ I ~ L ] ( C I OThe ~ ) Cu ~ .2p ~ Hspectrum ~ O . ~ ~of Cu"spectrum shows that, as in 6 , the odd electron is now localized Cu"LCIy6H20 revealed features in common with 4, although the two sets of Cu 2p doublets were unresolved. However, the on a single copper center. overall width of the overlapping Cu 2p3/2 components (fwhm Up to this point we have assumed that the appearance of two = 3.8 eV) was identical with that of 4 (Figure la). In addition, sets of peaks near the binding energies of isolated Cu(l1) and the lower energy satellite structure ( A E 6 eV) seen i n the Cu(1) species could be taken as evidence that the compound spectrum of 4 (Figure la) is clearly present i n the X-ray PES does not have a delocalized ground state. Such an assumption has been criticized by Hush,Z4 who pointed out that a symof C ~ ~ ~ C u ~ l L C 1 2 . 6 HThe 2 0 . spectrum of the mixed metal complex [ C U ~ ~ Z(C104)2.2H20 ~ ~ ~ L ] is different from that of metrical delocalized mixed-valence compound will have two the Cu(I1)-Cu(I1) derivatives in two important respects. First, accessible unsymmetrical photoionized states, due to electron while two Cu 2p doublets are observed, they are at energies (Cu relaxation in the strong field of the core hole, and these will be 2p3/2 at 934.7 and 932.7 eV) which are typical of mononuclear localized. Accordingly, the peak separation for a complex with Cu(I1) and Cu(1) species, respectively (Table I ) . Second, the a delocalized ground state could be close to that for isolated lower energy satellite structure, which is present in the X-ray Cu(1) and Cu(I1). From this theory,24the intensity of the peak PES of 4 and C ~ ~ ~ C u ~ ~ L C l 2 . 6isHabsent 2 0 , in the related a t low binding energy to that at high binding energy is prospectrum of [ C U " Z ~ ~(C104)2-2H20.30 ~L] portional to the electronic coupling IJ I , so that for weakly While the above observations concerning the X-ray PES of coupled but delocalized metals (i.e., 1J( is small) we should the Cu(l1)-Cu(I1) and Cu(Jl)-Zn(lI) species can be rationobserve two peaks of nearly equal intensity. As we have menalized in terms of a fairly constant level of X-ray induced retioned earlier, the intensities of the two sets of peaks for 5 are 1

-

GagnC, Walton, et al.

/ X-ray PES of Copper(I) and Copper(II) Complexes

1909

(1971); M. M. Rising, J. S.Hicks, and G. A. Moerke. J. Biol. Chem., 1, 1 duction of Cu(I1) (approximately 50% of the material is re(1930); F. J. Hollander, M. L. Caffery, and D. Coucouvanis, J. Am. Chem. duced), the spectral differences between these two species are SOC.,96, 4682 (1974): J. J. Bour and J. J. Steggerda, Chem. Commun., 85 (1967);S.0. Grim, L. J. Matienzo, and W. E. Swartz, Inorg. Chem., 12, best explained in terms of the nature of the molecular orbital 2762 (1973). which is involved in the reduction. In the case of the two (7) (a) R. R. Gagne, C. A. Koval, and T. J. Smith, J. Am. Chem. SOC., 99, 8367 Cu(I1)-Cu(I1) complexes, the chemical shift of the lower en(1977):(b) R. R. Gagne, C. A. Koval, T. J. Smith, and M. C. Cimolino, ibid., 101, 4571 (1979). ergy set of Cu 2p1/2,3/2 peaks is intermediate between that of (8) R. R. Gagne and L. Henling, manuscript in preparation. localized Cu(I1) and Cu(I), an observation which may be (9) N. H. Pilkington and R. Robson, Aust. J. Chem., 23, 225 (1970). consistent with the electron being delocalized over both metal (10) D. C. Olson and J. Vasilevskis, Inorg. Chem., 10, 463 (1971). (11) V. Katovic, L. T. Taylor, F. L. Urbach, W. H.White, and D. H. Busch, lnorg. centers in a genuine Cu(+1.5) dimer. The satellite at AE Chem., 11, 479 (1972). $ 6 eV would then be associated with this reduced species, (12) (a) A. D. Hamer and R. A. Walton, Inorg. Chem., 13, 1446 (1974);(b) A. D. Hamer, D. G. Tisley, and R. A. Walton, J. Inorg. Nucl. Chem., 36, 1771 while the higher energy satellite structure (that between 8 and (1974). 10 eV) is due to that proportion of the Cu(I1)-Cu(I1) dimers (13) D. C. Frost, A. Ishitani, and C. A. McDowell, Mol.Phys., 24, 861 (1972). which has not been reduced. The reason that this delocalized (14) H. Kolind-Andersen, S.-0. Lawesson, and B. Folkesson, Recl. Trav. Chim. Pays-@as, 93, 123 (1974). reduced binuclear complex can be generated in the solid state (15) J. G. Dillard and L. T. Taylor, J. Electron Spectrosc. Relat. Phenom.. 3,455 whereas solution electrochemical reductions produce localized (1 974). (16) D. A. Edwards, Inorg. Chim. Acta, 18, 65 (1976). Cu( 11)-Cu( 1)' is attributed to favorable lattice energy effects. (17) S.A. Best and R. A. Walton, lsr. J. Chem., 15, 160 (1977). With the heteronuclear complex [ C ~ ~ ~ Z n ~ ~ L ] ( C 1 0 4 ) 2 - 2 H 2(18) 0 , K. S.Kim, J. Electron Spectrosc. Relat. Phenom., 3, 217 (1974). the one-electron X-ray induced reduction leads to diamagnetic (19) D. W. Margerum, T. Neubecker, V. Srinivasan, and R. A. Walton, unpublished observations. Cu( I)-Zn( 11), rather than a paramagnetic delocalized species, (20) S. T. Kirksey. Jr., T. A. Neubecker. and D. W. Margerum, J. Am. Chem. Soc., a reflection on the difficulty of reducing Zn(1I). Such a re101, 1631 (1979). (21) Other binding energies for this complex are as follows: Cls,284.8, -286.0 duced species, possessing dIo metal centers, would not be exsh and 287.4; N Is, -399.2 br: 0 Is, 531.4 eV. pected to exhibit any satellite structure and, with the exception (22) Since our manuscript was first submitted for publication, Professor Jeremy K. Burdett has kindly informed us of the results of some extended Huckel of that due to unreduced Cu(Ir)-Zn(II) which is still present, MO calculations he has carried out on the complexes 1, 2, 3, and 11 (J. no additional satellite structure is observed. K. Burdett and P. D. Williams, manuscript in preparation).Calculations for

-

Acknowledgments. Support from the National Science Foundation (Grant CHE74-12788A04 to R. A. Walton) is gratefully acknowledged. W e thank Professor Jeremy K. Burdett, University of Chicago, for the information supplied in ref 22. References and Notes (1) This paper is part 25 in the series "The X-ray Photoelectron Spectra of Inorganic Molecules". For part 24, see P. Brant, W. S . Mialki. and R. A. Walton, J. Am. Chem. Soc., 101, 5453 (1979). (2) (a) California Institute of Technology; (b) Purdue University. (3) R. R. Gagne, J. L. Allison, and G. C. Lisensky, Inorg. Chem., 17, 3563 (1978). (4) (a) R. R. Gagne, J. Am. Chem. Soc., 98, 6709 (1976): (b) R. R. Gagne, J. L. Allison, R. S . Gall, and C. A. Koval, ibid., 99, 7170 (1977). (5) R. R. Gagne, J. L. Allison, and D. M. ingle, Inorg. Chem., 18, 2767 (1979). (6) (a) s. 1. Shupack, E. Billig. R. J. H. Clark, R. Williams, and H. B. Gray, J. Am. Chem. Soc., 86,4594 (1964);(b) H. B. Gray, R. Williams, I. Bernal, and E. Billig, ibid., 84, 3596 (1962); (c) H. B. Gray and E. Billig. ibid., 85, 2019 (1963): R. Williams, E. Billig, J. H. Waters, and H. B. Gray, ibid., 88, 43 (1966);F. Rohrscheid, A. L. Baich, and R. H.Holm, Inorg. Chem., 5, 1542 (1966): A. L. Balch and R. H. Holm, J. Am. Chem. Soc., 88, 5201 (1966); J. J. Bour, P. J. M. W. L. Birker, and J. J .Steggerda, Inorg. Chem., 10 1202

(23) (24) (25) (26)

[Cu(TAAB)]+ (11) are in accord with thls complex being a Cu(l) species in which the HOMO is Cu dp- and the LUMO is ligand A ' . This result agrees with our Cu 2p X-ray PEgresults which favor the formulation of this complex as Cu(l). Similarly, agreement is found between the interpretations of the X-ray PES results and the MO calculations for Cu(LBF2HC104)(1)and Cu(LBF2XCO)(3b), Insofar as the formal description of the copper oxidation states is concerned (Cu(ll) and Cu(l) species, respectively). However, for Cu(LBF2) (2), whose Cu 2p spectrum is similar to that observed for other authentic Cu(l) complexes (3, 9, 11, etc., in Table i), the MO calculations point to the metal d,+-yz orbital lying above that of the normally unoccupied ligand A* orbital. This would lead to internal electron transfer to the ligand, a situation where the metal is now formally Cu(ll1). This conclusion is not supported by the results of our X-ray PES studies on 2, which clearly favor it being an authentic Cu(l) complex. C. S. Fadley and D. A. Shirley, Phys. Rev. A, 2, 1109 (1970). N. S.Hush, Chem. Phys., 10, 361 (1975). B. F. Hoskins and G. A. Williams, Aust. J. Chem., 28, 2593 (1975). See for example, D. E. Sherwood, Jr., and M. B. Hail, inorg. Chem., 17,

3397 (1978). (27) The spectra were also independent of the use of the electron "floodgun".

(28) B. F. Hoskins, N. J. McLeod, and H. A. Schaap, Aust. J. Chem., 29, 515 (1976). (29) The synthesis and characterization of this complex and other mixed metal derivatives will be reported in detail at a later date.

(30) Other relevant binding energies of this complex are Zn 2 ~ 3 1 2= 1022.2 eV, ~ 207.7 2 eV, and C Is = 285.0 eV. The related N 1s = 399.3 eV, CI 2 ~ 3 = binuclear Zn(ll) complex Zn1'ZnliLCIp2H20has Zn 2p312= 1022.2 eV, N 1s = 399.1 eV, and C 1s = 285.0 eV.