UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies in 57Fe and 121Sb Mossbauer spectorscopy Scott, James Charles Stewart 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1973_A1 S37_8.pdf [ 8.51MB ]
Metadata
JSON: 831-1.0060176.json
JSON-LD: 831-1.0060176-ld.json
RDF/XML (Pretty): 831-1.0060176-rdf.xml
RDF/JSON: 831-1.0060176-rdf.json
Turtle: 831-1.0060176-turtle.txt
N-Triples: 831-1.0060176-rdf-ntriples.txt
Original Record: 831-1.0060176-source.json
Full Text
831-1.0060176-fulltext.txt
Citation
831-1.0060176.ris

Full Text

STUDIES IN 5 7 F e AND 1 2 1 S b MOSSBAUER SPECTROSCOPY by JAMES CHARLES STEWART SCOTT B.Sc.(Hons.), University of Br i t i s h Columbia, 1965 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Chemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA JULY, 1973 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission fo r extensive copying of t h i s thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. The University of B r i t i s h Columbia Vancouver 8, Canada Department i ABSTRACT A number of iron carbonyl complexes of the general formula LFe2(CO)g (L = f luoroalicyclic-bridged di (tertiary arsine or phosphine)) have been investigated by Mossbauer spectroscopy and the results are consistent with the known structure of f^farsFe2(CO)g. As well, a number of derivatives of the general formulae L,mLFe2 ( C O ) ( L m = monodentate ligand) , L CLFe 2(CO) 4 (L° = bidentate ligand, chelating), and LbLFe2(CO)^ (L b = bidentate ligand, bridging) have been examined. The Mossbauer parameters are consistent with substitution trans to the iron-iron bond in LmLFe2(CO) ,j. Mossbauer and infra-red data show complexes of the type L LFe (CO), have structures similar to f.AsP f,AsPFe„(CO).. Mossbauer 2 4 4 4 2 4 spectroscopy shows that in complexes of the type ^U^CCO)^ the ligands bridge the two iron atoms and are coordinated cis to the iron-iron bond. The usefulness of the magnetic perturbation technique for removal of ambiguities in the assignment of Mossbauer spectral parameters in low-spin iron compounds having two iron sites has been demonstrated. 121 Sb Mossbauer spectra were obtained for the following compounds: (C6H5)4SbX (X = Cl, OH, NCS), (C 5H 5) 3SbX 2 (X = OCOCH3, NCS, N03, ^(OCrO^), (C 6H 5) 2SbCl 3 and (CgH^SbtOOH. The data for the (CgH^SbX and (CgH^)3SbX2 complexes are consistent with trigonal bipyramidal structures for these compounds (except for the acetate) with the X groups in the axial positions. The additive model for quadrupole splittings has been success-fully applied to these and some related compounds. The linear relationship i i 2 between the isomer shift and e qQ for many of the compounds of the type (CgH,.) ,jSbX2 suggests a-bonding plays a dominant role in determining their Mossbauer parameters. Possible structures are examined for both (C,Hc).SbCl_ and (C,HC)„Sb(0)0H. For (C.HC)„Sb(OCOCH„)„, octahedral structures with one ester-like and one bidentate acetate group are compatible with the observed parameters. + A number of cations of the general formula X Sb (Fe (CO) 0Tr-CcHc) . ° n 2 5 5 4-n (X = Cl, Br, I, CF^, CgH5, n-C^; n = 1, 2, 3, but not a l l combinations) have been studied by ^ 7Fe and ^ "''Sb Mossbauer spectroscopy. These compounds are nominally isoelectronic with the extensively studied neutral tin derivatives X Sn(Fe(CO)„(1T-C.-H,.)) . and a number of correlations n 2 5 5 4-n between their respective spectral parameters have been investigated. The 121 Sb isomer shifts were found to overlap with the ranges of isomer shifts characteristic of Sb(III) and Sb(V), hence the assignment of a formal oxidation state to antimony in these compounds has l i t t l e justification. + 121 For the complexes R3SbFe(CO)2 (TT-C^) (R = C 6 H 5 » n - 0 ^ ) t h e s b 2 quadrupole coupling constants e qQ are positive while the coupling constant in the corresponding tin derivative (n-C^Hg) ^SnFe (CO) 2 (FT-C^ -H,.) has been 57 119 121 found to be negative. Isomer shift data for Fe, Sn and Sb as well as the carbonyl stretching frequencies for the Fe(CO)2(iT-C^H^) group indicate Fe-Sb ir-bonding is more important than Fe-Sn TT-bonding, although a-bonding effects are the dominant factor in determining the Mossbauer spectral parameters. i i i TABLE OF CONTENTS Page ABSTRACT i ACKNOWLEDGEMENTS x INTRODUCTION 1 THE MOSSBAUER EFFECT 2 EXPERIMENTAL 31 RESULTS AND DISCUSSION PART 1 L F e 2 ( C O ) 6 COMPLEXES AND THEIR DERIVATIVES 39 (A) L F e 2 ( C O ) 6 39 (B) L m L F e 2 ( C O ) 5 . . . . . . 57 (C) L C L F e 2 ( C O ) 4 65 (D) L b L F e 2 ( C O ) 4 76 (E) THE IRON-OLEF IN BOND IN L F e „ ( C O ) , AND Z o THEIR DERIVATIVES 90 PART 2 P h c SbX DERIVATIVES . . . . 95 5-n n PART 3 R S b ( F e ( C 0 ) o C p ) . X COMPOUNDS . . 121 n 2 4-n 121 (A) Sb MOSSBAUER PARAMETERS . . . 123 (B) 5 7 F e MOSSBAUER PARAMETERS . . . 139 (C) THE CORRELATION OF 5 7 F e MOSSBAUER P A R A -METERS WITH THE CARBONYL STRETCHING FREQUENCIES IN THE I . R 145 BIBLIOGRAPHY 153 APPENDIX I 163 APPENDIX I I 173 i v LIST OF TABLES Table Page(s) I Mossbauer Parameters at,80°K for LFe 2(C0) & Compounds . 42 II Possible Assignments of the Mossbauer Parameters for f 6 f a r s F e 2 ( C 0 ) 6 44 III Mossbauer Parameters at 80°K for L mLFe 2(C0> 5 Compounds 59 IV Mossbauer Parameters at 80°K for L°LFe 2(C0) 4 Compounds • 68-69 V Mossbauer Parameters at 80°K for L bLFe 2(CO) 4 Compounds 77-78 VI Magnetic Perturbation Results . . . . 93 VII 1 2 1 S b Mossbauer Parameters at 9°K (This Work) . 97 VIII 1 2 1 S b Mossbauer Spectra at 4.2°K (Previous Studies) 98 IX Application of the Additive Model to Predict the 2 e qQ Values for R,. nSbX^ Compounds . . . 102 X NQR Data - At Room Temperature . . . . 103 121 XI Sb Mossbauer Parameters for Compounds of the Type R Sb(Fe(C0)„Cp) . X 124-125 J r n 2 4-n XII Isomer Shifts of Nominally Isoelectronic 136 Antimony and Tin Complexes V LIST OF TABLES (Continued)  Table Page(s) XIII ~*^ Fe Mossbauer Parameters for.Compounds of the Type RnSb(Fe(CO)2Cp)4_nX . . . . . . 140 XIV ~^Fe Mossbauer Parameters for Fe(C0)2Cp Groups Bonded to Tin and Antimony 141 XV "*^ Fe Mossbauer and Parameters for Some X3MFe(C0)2Cp Derivatives . 146 XVI "^Fe Mossbauer Parameters and Parameters of Some X2M(Fe(C0)2Cp)2 Derivatives . . . . 147 v i LIST OF FIGURES Figure Page 1 Schematic of a Typical Mossbauer Spectrometer Using Transmission Geometry and Absorber Cooling . . . 5 2 An Approximate Energy Level Diagram for a "*^ Fe Nucleus Subjected to an e.f.g. (n = 0) and Then to an Applied Magnetic Field at an Angle 8 to the Principal Component of the e.f.g. 12 121 3 Energy diagram for an Sb Nucleus Subjected to Non-Zero Axially Symmetric e.f.g. 14 4 Some Regular Structures and the Point Charge Expressions for the Components of Their EFG Tensors. Where the Principal Axes are Determined by Symmetry, the Components are Designated by Upper Case Subscripts. Otherwise, the Principal Axes can Only be Found by Diagonalization of the Tensor for Each Case Under Consideration 22 5 Schematic Diagram of the Apparatus Employed for 121 Obtaining Sb Mossbauer Spectra 33 6 Schematic of Magnetic Perturbation Apparatus . . 36 7 The Structures of Some Typical Ligands Used in this Work 40 8 The Structure of f.farsFe„(CO), 41 v i i LIST OF FIGURES (Continued)  Figure Page 9 "*7Fe Mossbauer Spectra of far s F e ^ ( C O ) ^ and (PhO) 3Pf 4AsPFe 2(CO) 5 . . . . . . . 45 10 Typical Spectra Produced i f Assignment (a) is Correct with Site I = 0.34 mm/sec, n = 0 and Site II = 0.30 mm/sec, n = 0 . . . . 49 11 Typical Spectra Produced i f Assignment (b) i s Correct with Site I = 1.00 mm/sec, n <= 0 and Site II = 1.06 mm/sec, n = 0 . . . . . 50 12 "*7Fe Mossbauer Spectrum of f ,f arsFe_ (CO) , i n an o z o Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical F i t . 51 13 ^ 7Fe Mossbauer Spectrum of f^AsPFe2(CO)^ in an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical F i t . . . 52 57 14 Fe Mossbauer Spectrum of f o s F e 2 ( C 0 ) ^ in an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical F i t . . . 53 15 5 7 F e Mossbauer Spectrum of (PhO) 3Pf 4AsPFe 2(C0) 5 in an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical F i t . . . 58 16 The Structure of f.AsPCf.AsPFe_(CO). . . . 66 4 4 2 4 v i i i L I S T OF FIGURES ( C o n t i n u e d )  F i g u r e Page 17 " ^ F e M o s s b a u e r S p e c t r a o f f ^ f o s ^ f ^ f o s F e , , ( C O ) ^ and f 4 A s P ° f 4 f a r s F e 2 ( C O ) 4 67 57 b 18 Fe M o s s b a u e r S p e c t r u m o f f ^ f o s f o s F e ^ ( C O ) ^ i n an A p p l i e d L o n g i t u d i n a l M a g n e t i c F i e l d o f 50kG S h o w i n g E x p e r i m e n t a l P o i n t s and T h e o r e t i c a l F i t . . . 83 121 19 The Sb M o s s b a u e r S p e c t r u m o f P h 2 S b C l 3 Show ing F i t 121 w i t h n = 0 and t h e Sb M o s s b a u e r S p e c t r u m o f a T y p i c a l Compound o f t h e Type R ^ S b X , ^ Name ly P h 3 S b ( N 0 3 ) 2 99 121 2 20 C o r r e l a t i o n o f t h e Sb I . S . and e qQ f o r a Number o f D e r i v a t i v e s o f t h e Type P h 3 S b X 2 . . . . 104 121 21 Sb M o s s b a u e r S p e c t r u m o f P h 3 S b ( 0 A C ) 2 Show ing Improvement o f F i t f o r n. = 0 . 4 6 (b) O v e r T h a t f o r n = 0 . 0 (a) 115 22 1 2 1 S b M o s s b a u e r S p e c t r u m o f C l 2 S b ( C p ( C O ) 2 F e ) 2 + P F 6 ~ I l l u s t r a t i n g Improvement o f F i t f o r f| = 0 . 4 6 (b) O v e r T h a t f o r n = 0 . 0 (a) 126 121 + — 23 Sb M o s s b a u e r S p e c t r u m f o r C I S b ( C p ( C O ) 2 F e ) 3 P F f i . 127 121 + — 24 Sb M o s s b a u e r S p e c t r u m o f P h 3 S b F e ( C O ) 2 C p P F 6 Show ing A l t e r n a t e F i t s t o t h e D a t a . I n (a) t h e F i t t i n g 2 P a r a m e t e r s w e r e 5 = - 6 . 7 m m / s e c , e qQ = + 9 . 4 m m / s e c , ix LIST OF FIGURES (Continued)  Figure Page r = 2.9 mm/sec, n = 0.0. In (b) , 6 = - 6.5 mm/sec, 2 e qQ = - 2.9 mm/sec, T = 4.0 mm/sec, n = 0.0. The 2 Fit with e qQ>0 is Clearly Preferable 128 25 The Correlation of ^ ^Sn and "^ S^b Isomer Shifts. The Straight Line, Based on the Assumption of Equivalent Electron Density at the Two Nuclei, is after Ruby. The Points for the Isoelectronic Pairs are Labelled Using the Notation M = Sn or Sb + and Fe = Fe(C0)2Cp . . . . . 137 57 + — 26 Fe Mossbauer Spectrum of PhSb(Cp(CO)_Fe). PF, in A J O an Applied Longitudinal Magnetic Field of 30kG. The 2 Sign of e qQ is Clearly Positive 144 X ACKNOWLEDGEMENTS I wish to express my thanks to Dr. J.R. Sams for his patience and guidance of this work. I am particularly grateful for his invaluable assistance in the past few months. I should like to extend my sincere appreciation to Dr. J.N.R. Ruddick for his advice and encouragement during the early course of the antimony work and for his synthesis of the five-coordinate antimony derivatives. As well, our discussions of some of the experimental problems were particularly stimulating. To Mr. L.S. Chia, a particular vote of thanks for his many hours of hard labour in the separation and purification of the iron carbonyl complexes and for free access to his unpublished experimental data. I am also indebted to Dr. D.J. Patmore for the contribution of his synthetic skills to the preparation of the antimony-iron compounds. Finally, I appreciate the contributions of the rest of the Mossbauer group, including Lia Sallos, Troy Lassau and Tsin Bik. INTRODUCTION One of the most important problems confronting the chemist is the elucidation of the structure and bonding in new or unusual compounds. In order to solve this problem there is recourse to many physical methods, but in particular the various forms of spectroscopy play a dominant role. The applications of nuclear magnetic resonance, infrared and electronic absorption spectroscopies to the elucidation of structure and bonding are well known''". Among the spectroscopic tech-niques which have been employed i s Mossbauer spectroscopy or nuclear 2 gamma resonance spectroscopy , with which this thesis i s concerned. Spectroscopic techniques rely on the interaction between atoms or molecules and electromagnetic radiation. This usually takes the form of the emission or absorption of one or more quanta of radiation (in Raman spectroscopy, this process is somewhat more complex since the energy corresponding to the transition i s added to or subtracted from a 3 4 scattered photon ). Owing to certain restrictions , the energies (and thus the frequencies) of these quanta are normally confined to a certain definite range of values. Thus, a l l forms of spectroscopy may be characterized by the particular type of transitions which may be studied (e.g. rotational, vibrational, electronic) and by the range of the electro-magnetic spectrum to which they are confined. For this reason the insights into chemical behaviour which can be gained by any given form of spectro-scopy are somewhat limited by the particular spectroscopic "window" which is employed. For example, one could not expect to use exactly the same spectro-2 scopic tools to study both Raman and infrared transitions. Although the type of information to be gained i s very similar, the techniques are very different^. Mossbauer spectroscopy, too, has i t s characteristic limitations. For example, only nuclei in solid materials or in very viscous liquids may be studied^', the number of isotopes to which the technique is applicable is limited since the effect has never been observed in isotopes lighter than ^K. As well, there are physical limitations such as very short half-lives of radioactive precursors, or extremely broad or extremely narrow linewidths. A l l these factors mean that there are only about a dozen isotopes which are currently in use for routine chemical applications and naturally an individual spectroscopist is somewhat limited in his access to those isotopes which are available. Nevertheless, for those isotopes to which the effect is applicable, Mossbauer spectroscopy provides some valuable information about such diverse phenomena as the chemical effects of nuclear trans-formations, magnetic phase transitions, electron densities, and molecular symmetry^. So although a Mossbauer spectroscopist's "window" on chemical behaviour i s a very small one, for those elenents to which i t can be applied, some unique structural and bonding information may be derived, particularly from measurements of the isoner shift and quadrupole s p l i t t i n g . The purpose of this thesis i s to ill u s t r a t e how Mossbauer spectroscopy can be employed to study structure and bonding.. In particular the techniques and theories which have been so successfully applied in 3 elucidating the structures of iron and tin compounds have been 121 applied to the interpretation of the Sb Mossbauer spectra in two series of antimony compounds. The first series of compounds of the type Ph^_nSbXn (Ph = phenyl, X = electronegative group, NCS, NO.j> etc., n = 1, 2, 3) is an g extension of the earlier work of Long e_t al. and represents the first explicit^ application of partial quadrupole splitting (p.q.s.) theory to the interpretation of the electric field gradients (e.f.g.) in these antimony compounds. In the second set of compounds of the type R4_nSb{Cp(CO)2Fe} X (R = Cl, Br, Ph, etc., Cp = 7T-cyclopentadienyl, X = large anion, PFg , etc., n = 1, 2, 3), the theory of p.q.s. has also been applied. 57 As well, the Fe resonance has also been studied in order to elucidate the nature of the antimony-iron bond. The nature of the iron-iron bond as well as the structures of a number of substituted iron carbonyls bridged by di(tertiary arsines and phosphines) has also been investigated. Magnetic perturbation techniques have been used as a check on the assignment of the Mossbauer parameters for the two iron sites in these compounds as well as to investigate the signs of the e.f.g. at both sites. 4 THE MOSSBAUER EFFECT The Mossbauer effect depends on the recoil-free emission and absorption of y-rays by suitable nuclei. Although there are many isotopes for which the effect has been demonstrated, most of the experimental results to date have dealt with either ^ 7Fe or ^ "^Sn. This is not surprising in view of the ease of working with these isotopes. In a conventional Mossbauer experiment in transmission geometry, the y-rays from an appropriate single-line source are Doppler modulated, and, after passing through a suitable absorber, are detected and counted. The range of source velocities is normally such that the Y-rays from the source go in and out of resonance with the absorber. The spectra consist of a plot of transmitted counts versus Doppler 9 velocity (Figure 1). In most cases, except for certain compounds of iron and tin, i t is necessary to cool both the source and absorber at least to liquid nitrogen temperature. This is because f , the resonance fraction or the probability Si of recoil-free absorption of a Y-ray from the source, is of the form^ 2 2 -E <X > (1) f = exp(-^ 5-) a ( * c r 2 where E^ is the Y-ray energy, <X > the mean square vibrational amplitude of the nucleus in the direction of the y-ray, sometimes referred to as the Debye-Waller factor, -fi is Planck's constant over 2TT, and c the velocity of light in vacuo. A similar equation holds for f , the probability of recoil-FIGURE 1. Schematic of a Typical Mossbauer Spectro-meter Using Transmission Geometry and Absorber Cooling. c 0 u N T S TYPICAL SPECTRUM ~ \ A / ~ VELOCITY HIGH VOLTAGE POWER SUPPLY DATA OUTPUTS: PAPER TAPE L_ © DISPLAY OSCILLOSCOPE ANALYZER DETECTOR AMPLIFIER PICKUP SIGNAL » * -< r DEWAR —» 1 y-SOURCE TRANSDUCER MOSSBAUER DRIVE UNIT '-DRIVE SIGNAL H OPTICAL BENCH 400 WORD MEMORY PULSE HEIGHT ANALYZER TIME BASE GENERATOR SYNCHRONIZATION SIGNAL M O % 6 free emission from the source. Now, as <XZ> depends on the firmness with which the Mossbauer nucleus is bound to its lattice site (this may be an anisotropic function), and on the temperature, i t is apparent that cooling the absorber to as low a temperature as is feasible will increase the resonance fraction. Similar considerations apply to the source, although in general i t is possible to choose a suitable source matrix so that the nuclei are very strongly bound and the temperature effect is correspondingly diminished. To understand how the Mossbauer effect can be used as a probe to investigate structure and bonding, i t is necessary to understand how a nucleus is influenced by its chemical environment. There are three important ways that a nucleus interacts with its external environment which may be manifested in a Mossbauer spectrum. These are the electric monopole interaction, the electric quadrupole interaction and the magnetic dipole interaction. These interactions give rise to the isomer shift (I.S.), the quadrupole splitting (Q.S.) and the magnetic hyperfine splitting, respectively. The Isomer Shift The isomer shift, due to the electric monopole interaction, arises because the nuclear volume is penetrated by some of the charge density of the electrons. The total energy of the nucleus then is subtly influenced by variations in the charge density of the s-electrons. This variation would be unimportant except that a nucleus undergoing a y-transition 7 will normally change in size and so the magnitude of the nucleus-s-electron interaction energy will be slightly different in each of the two states. Although these changes are only a very small part of the total nucleus-electron interaction energy and would be nearly impossible to measure on an absolute scale, a relative energy scale may be readily constructed. If a nucleus B is chosen as a reference nucleus and the relative energy of nucleus A is measured with respect to B then the difference in their relative energies or the chemical isomer shift, 6, may be approximated a s ^ (2) 6 = frr ZeV f k | ^  (0)J 2- | ^  <0) B | 2 } where Z is the nuclear charge, e the electronic charge constant, r the mean nuclear radius, 6r = r -r the difference in the nuclear radii of ex gr 2 2 the excited and ground states, and (0).| and |¥ (0) | are the prob-S A S U ability densities for s-electrons in the nuclear volume for A and B respectively. A 2 2 The isomer shift is the product of a nuclear term (!<jr Z e r «r) r which is essentially constant for a given nucleus, and a chemical term 2 2 {|Y (0) | -|y (0) | } which varies with the electronic environment of the A S i3 2 nucleus. Each of the s-orbitals of the atom will contribute to I ¥ (0)1 ' s 1 but in lesser amounts as the principal quantum number, n, increases. Normally i t is considered that the core electrons are only slightly influenced by the behaviour of the valence electrons, so their contribution to |y (0)| remains essentially constant. The principal contribution to the 8 isomer s h i f t comes from changes i n the number of s-electrons i n the outermost occupied s - o r b i t a l s . Since p, d and f electrons exert screening e f f e c t s on the outermost s-electrons, they also can influence the magnitude 2 12 of 1^(0)1 , but to a l e s s e r degree 6r The sign of — f o r a given nucleus determines the trend of isomer s h i f t with s-electron density. For both "*7Fe and ^^Sb i s n e g a t i v e 7 so that an increase i n s-electron density at the nucleus leads to a decrease i n the isomer s h i f t . In contrast, for ^ "^Sn — • i s p o s i t i v e 7 . The isomer s h i f t i s temperature independent except f o r two f a c t o r s . The f i r s t of these i s known as the second-order Doppler s h i f t and i s a consequence of the temperature dependence of the l a t t i c e v i b r a t i o n • 2 term, <X >. This gives r i s e to a small change i n the y-ray energy of the form"*""* (3) S . . ^ V 2c • 2 2 Since <X > i s exceedingly small r e l a t i v e to c t h i s term i s correspondingly very small. For a se r i e s of s i m i l a r compounds measured at the same . 2 temperature, <X > should be nearly constant and so the d i f f e r e n c e i n the second-order Doppler s h i f t w i l l be e s s e n t i a l l y zero. The second temperature dependent e f f e c t i s that a r i s i n g from chemical or p h y s i c a l changes i n the system i t s e l f . For example, depopulation of thermally a c c e s s i b l e energy states or phase changes i n the c r y s t a l l a t t i c e w i l l lead to changes i n the isomer s h i f t . (Both of these phenomena should lead to changes i n the Q.S. as we l l . ) 9 I n g e n e r a l , t h e i s o m e r s h i f t may be u s e d as a p r o b e f o r t h e e l e c t r o n c o n f i g u r a t i o n o r , i n t h e c a s e o f h i g h l y i o n i c compounds , f o r t h e o x i d a t i o n s t a t e o f a toms w h i c h c o n t a i n a M o s s b a u e r n u c l e u s . F o r 2+ e x a m p l e , h i g h - s p i n Fe compounds may be e a s i l y d i s t i n g u i s h e d f r om h i g h -3+ 7 s p i n F e compounds ow ing t o t h e more p o s i t i v e i s o m e r s h i f t s i n t h e f o r m e r . H o w e v e r , f o r l o w - s p i n i r o n compounds t h e changes i n i s o m e r s h i f t w h i c h o c c u r on a d e c r e a s e i n o x i d a t i o n number a r e v e r y much s m a l l e r , as an e x t r a 14 e l e c t r o n i s l a r g e l y d e l o c a l i z e d o n t o t h e l i g a n d s . S i m i l a r l y , f o r h i g h l y c o v a l e n t t i n compounds , t h e r e a r e many c a s e s i n w h i c h t h e i s o m e r s h i f t f a l l s i n t o t h e r e g i o n b e t w e e n t h e i s o m e r s h i f t s o f compounds w h i c h may d e f i n i t e l y be c h a r a c t e r i z e d as e i t h e r S n ( l l ) o r S n ( l V ) . T h i s i n d i c a t e s t h a t t h e e l e c t r o n d e n s i t y a t t h e t i n n u c l e u s i s i n t e r m e d i a t e b e t w e e n t h e s e two e x t r e m e s . The Q u a d r u p o l e S p l i t t i n g The s e c o n d way i n w h i c h a n u c l e u s i n t e r a c t s w i t h i t s c h e m i c a l e n v i r o n m e n t i s v i a i t s e l e c t r i c q u a d r u p o l e moment. N u c l e i w h i c h h a v e a s p i n quantum n u m b e r , I, g r e a t e r t h a n h h a v e a n o n - s p h e r i c a l n u c l e a r c h a r g e d i s t r i b u t i o n . The m a g n i t u d e o f t h i s c h a r g e d i s t o r t i o n i s m e a s u r e d b y Q , t h e n u c l e a r q u a d r u p o l e moment. A p o s i t i v e Q c o r r e s p o n d s t o an o b l a t e c h a r g e d i s t r i b u t i o n a b o u t t h e s p i n a x i s w h i l s t a n e g a t i v e Q c o r r e s p o n d s t o a p r o l a t e d i s t r i b u t i o n . The i n t e r a c t i o n o f a d i s t o r t e d n u c l e u s w i t h a n o n - c u b i c e x t r a n u c l e a r e l e c t r i c f i e l d g r a d i e n t , e . f . g . , g i v e s r i s e t o a s p l i t t i n g o f t h e n u c l e a r e n e r g y l e v e l s . F o r h a l f - i n t e g r a l n u c l e a r s p i n s t h e s e w i l l c o n s i s t o f K r a m e r s d o u b l e t s . 10 (4) The e.f.g. at the nucleus is defined by the tensor e.f.g. 15 V V V XX xy xz V V V yx y y yz V V V zx zy zz where V„ = 8 V/3i8j and V is the electrostatic potential. Since V has a continuous second derivative in the domain under consideration we have and the Laplacian vanishes; 2 3V 2 9i3j 3j3i (5) v2v V + V + V = 0, xx yy zz so that the tensor is symmetric traceless. Such a tensor may be diagonalized by suitable choice of axes (the so-called principal axes), and since the Laplacian is zero, only two parameters are necessary to specify the e.f.g. completely. In the principal axis system (denoted X, Y, Z) , the following convention has been adopted; (6) v z z l ^ | v Y Y ^ | v x x | , where and the asymmetry parameter (7) n = v -v YY XX ZZ ( o $ n a ) have been chosen as the independent parameters. The nuclear quadrupole coupling Hamiltonian for a nucleus of 15 spin I may be expressed as' 11 where eq = V z z, I z i s the spin operator and f , I are the shift operators. For both "^Fe and "^"^ Sn the ground state has spin I = "*7 2 and thus no quadrupole moment. The excited state has I = /2 s 0 that when the nucleus is subjected to a non-cubic e.f.g., two different substates arise. In the axially symmetric case (n = 0) these substates have I z = ± 12 and I z = ± ^~12 respectively. The energy separation between the two substates is called the quadrupole s p l i t t i n g , A E ^ , and is given i n the general case by (9) AEQ = 1 / 2 eZqQ (1 + § ) For a randomly oriented polycrystalline diamagnetic absorber containing either of these nuclei , and an unpolarized single-line source of Y -rays, the absorption spectrum w i l l consist of two lines of equal intensity (with the exception of a few special cases when, for example, the Gol'danskii-Karyagin effect is observed^ - see below). The separation of the two lines i s equal to A E ^ and the centroid corresponds to the centre of gravity of the unsplit case. In such a case, neither the sign of eq(=V z z) nor the magnitude of n can be deduced. Only in a series of experiments with oriented crystals or with magnetic perturbation techniques (see below) can the sign of eq or the magnitude of n be estimated (Figure 2 ) . 121 For Sb the quadrupole-split spectra are far more complex. Under the influence of a non-cubic e.f.g., the ground state which has I = "V2 s p l i t s into three sublevels, having I z = ± / 2» - 12 and ± 12* respectively, 12a FIGURE 2. An Approximate Energy Level Diagram for a "^Fe Nucleus Subjected to an e.f.g. (n = 0) and Then to an Applied Magnetic Field at an Angle 6 to the Principal Component of the e.f.g.. 12 FIGURE 2. 13 in the axially symmetric case. Similarly the excited state with I = splits into four sublevels, with = ± ^^2, - "12' - ~^ 2 a n c* ± ^^ 2» r e s P e c t i v e l y , in the axially symmetric case. As before, the energy separation of these sublevels is given by the quadrupole coupling Hamiltonian appropriate to each state (eqn. 8, Figure 3). In addition to the parameters of the e.f.g. (i.e., V z z and n), the relative energies of the transitions between ground and excited states are determined by two constants: the quadrupole moment of the ground state 17 Q (-0.28±0.1 barn) (one could just as well use Q except that the xgr J xex v value of Q is inherently more precise), and the ratio of the quadrupole § r 18 moments of the excited and ground states, R = Q /Q (+1.34±0.01) . The ' ex ^gr other parameters which determine the spectrum are the linewidths at half-height (assuming Lorentzian lineshapes), T, for each component, and the transition probabilities which connect the various substates. Normally the linewidths are assumed to be equal for a l l the components, and usually lie in the range 2.1 to 3.1 mm/sec. The transition probabilities are considered to be the theoretical ones as deduced from the Clebsch-Gordan coefficients and the appropriate angular factors (Appendix I). The magnitude of Q and R ensure that in a typical antimony spectrum most of the component lines will overlap. Thus the spectra have the appearance of one or more asymmetric lines due to this overlap (Figure 3), The information obtained on computer fitting a spectrum (Appendix I) consists 2 of the sign and magnitude of e qQ(Q = Q ), the linewidth T, the isomer 2 shift 6, and, in favourable cases (large e qQ), the value of n as well. If the e.f.g. is small, so that je qQ|-£ 5 mm/sec. then the sign and magnitude 14a FIGURE 3. Energy diagram for an Sb Nucleus Subjected to Non-Zero Axially Symmetric e.f.g.. FIGURE 3. m RELATIVE ENERGY 15 of e qQ may be in some doubt. There are two factors which can cause changes in the theoretical intensity ratios derived for randomly oriented polycrystalline 2 materials. The theoretical intensity ratios depend on terms like 1+ cos 6 Cwhere 8 i s the angle between the incoming y-ray and the Z axis of the e.f.g.) integrated over a l l the possible orientations (which are considered equally probable) of the crystallites (i.e. from 0 to TT) . If there is any preferential orientation of the crystallites in the sample then a l l orientations are no longer equally probable and there are departures from the theoretical intensity ratios. The second factor which may lead to changes in the intensity 2 ratios is the presence of anisotropics in the Debye-Waller factor, <X > (eqn. 1). The intensity ratio terms should really be written as say 2 f (1+ cos 8) but i f f (eqn. 1) is isotropic then the function is simplified 2 to (1+ cos 8) as above. However, i f the Debye-Waller factor i s anisotropic then f w i l l be described by some function f (6) which depends on the angle 2 8 as well. Thus the intensities w i l l now depend on terms like f (8)(1+ cos 8) a integrated over a l l possible orientations of the c r y s t a l l i t e s , and so departures from the intensity ratios calculated using the isotropic formulation are to be expected. This is the so-called Gol'danskii-Karyagin e f f e c t 1 ^ . The Gol'danskii-Karyagin effect is independent of the angle between the y-beam and the absorber since this dependence has been e x p l i c i t l y integrated out. On the other hand, the effects of preferential orientation 16 of the crystallites is not independent of the angle. This provides a simple means of distinguishing the two effects since a change in the intensity ratios observed on altering the absorber-y-beam angle means the Gol 1danskii-Karyagin effect may be ruled out - the converse i s not always true, of course. The other way in which these two effects are different i s in their temperature dependence. Orientation effects are expected to be temperature independent. The Gol'danskii-Karyagin effect which depends on the anisotropies in the Debye-Waller factor i s expected to be temperature dependent since these anisotropies should increase as the temperature is raised. To relate the magnitude and sign of the quadrupole s p l i t t i n g exhibited by a Mossbauer nucleus to the chemical environment which produces the e.f.g. is not a simple problem. Indeed, even i f the orien-tation of the e.f.g. with respect to the crystal axes (i.e. laboratory frame of reference) is determined i t i s apparently necessary to have detailed knowledge of the crystal structure in order to relate the e.f.g. to the 19 molecular axes . Nevertheless, a knowledge of the sign and magnitude of the e.f.g. w i l l in many cases allow us to make some deductions about the dispositions of the ligands in the molecule. Some assumptions as to how the ligands influence the distribution of electrons in a molecule and thus how the e.f.g. at the nucleus arises are essential in order to make structural predictions. In dealing with a molecule the e.f.g. may be separated to a 17 f i r s t approximation into two main contributions. One is the l a t t i c e contribution,  eQ.jjj<> due to charges on the ligands and other ions in the crystal. The second is the valence contribution, eq^^, due to the asymmetric distribution of electrons in bonding and non-bonding orbitals. Electrons in s-orbitals and electrons in f i l l e d shells are not considered to contribute to the e.f.g. except for induced polarization effects. 20 The conventional picture is thus (10) eq = ( l - Y j e q L A I + (l-R)eq^^ where R(0.2 > R > -0.2) and ym(-7 > y^ > -100) are Sternheimer factors 21 accounting for the induced polarization of inner electrons. Also , 2 /n\ (3cos 0. - 1) j 3 and / U c o s V - 1)' <12> et*VAL " "^PiC -L-where q^ is the charge on ion j whose polar coordinates are 0^, r^ (in this approximation ions are treated as point charges) while p. is the 2 3 population of the ith valence s h e l l o r b i t a l , <(3cos 8^  - l ) / r ^ > being the expectation value of this population over the electron coordinates 0 , r^. The summations are over a l l ions j and a l l valence shell orbitals i . For highly covalent molecular systems, the contribution from _3 external ions is expected to be small (owing to the r dependence) and thus 18 the q term to a good approximation arises solely from charges on LJA.L the ligands. Similarly, i f there are no non-bonding electron pairs*, or for transition metals with no p a r t i a l l y f i l l e d shells**, the e q v ^ L term w i l l consist of contributions from each ligand L. In this situation the so-called point charge model may be applied, in which the e.f.g. can 21 be written as (13) eq = Z [ L ] (3 cos 2 9 L-l) L where eq T ( l - Y o o ) ep T(l-R) (14) [ L ] = 3 . | 3. r L L > q is the charge on ligand L with coordinates r , 9 , and p the effective Li LI LI LI orbital population in the hybrid orbital directed towards L with effective electron coordinates 8 , r '. The l a t t i c e and valence contributions are Li LI opposite in sign and in most covalent compounds | eqyAL I > : > I ec*LAT ^  3 3 presumably due to the fact r >> <r ' >. Li LI There also exists a molecular orbital approach in terms of the relative a-donor and TT-acceptor strengths of the ligands. This approach has only been developed in detail for low-spin six-coordinate first-row transition-metal species (in particular for F e ( I I ) ) 2 2 . The £ L j value for * This may not be s t r i c t l y true i f there is only one lone pair which may be treated as a ligand. ** This restriction can sometimes be relaxed since both the dy and d £ subsets w i l l have spherical charge distributions i f they are either f i l l e d or empty (see below). 19 any particular ligand L is taken to be the sua of a a term for the contribution to the e.f.g. due to a-donation and a TT term due to the contribution from u-acceptance (TT then is opposite in sign to a ) . The L Li relative TT^ and values are assumed to be constant from compound to compound and, again, the contributions due to charges from external ions are neglected. The relative contribution of each cr to the total e.f.g. L i is assumed to be proportional to the squares of the coefficients for the appropriate hybrid orbitals directed towards the ligand. The hybridization is taken to be d^ sp"^  so that for "^Fe 3d „ _, 3d „, 4p , 4p , 4p and 4s x z_y z zl x y z orbitals will participate in a bonding. The TT contribution involves the relative TT bonding abilities of the d , d and d orbitals which taken ° xy' xz yz together are assumed to be equal in the three principal directions of the e.f.g.. The e.f.g. is considered to arise solely from d-orbital augmen-tation (or depletion) and so the contribution from the 4p orbitals is neg-3 3 lected on the basis <r >, » <r >_, (although no justification for this 4p 3d 22 assumption has been given ). Also, i t is expected that j a | > | TT | in most Li LI cases. The central feature of both the above models for the origins of the e.f.g. is the assumption that the e.f.g. at the nucleus may be treated as the sum of contributions L , one from each ligand. This is the 23 so-called additive approximation , or as i t is sometimes called the partial quadrupole splitting (p.q.s.) theory. To what extent this approximation is valid is governed by how well the contribution due to a given ligand remains essentially invariant for any series of compounds under 23 consideration 20 '•}. In this thesis, the value assigned to the contribution to the total e.f.g. from a ligand L will be expressed in terms of the value of £ as previously defined (eqn. 14) times e|oJ (or for "^Fe %e|Q|) so that £ L j has units of mm/sec. Thus, our values of £LJ will have the same significance as the (p.q.s.) values of Bancroft7. This also means that the V derived from p.q.s. calculations will be scaled up by e jQ1 (or ^ setQI) times the nuclear V^z' This procedure is adopted since the sign of is related to the equivalent ellipsoid of charge, so i f is negative the ellipsoid of charge is prolate, while i f i t is positive the ellipsoid of charge is oblate. 2 On the other hand, e qQ is what is determined experimentally and the 2 relationship between our derived V values and e qQ will just be 2 121 2 e qQ = sign(Q) V . Thus since Q is negative for Sb, e qQ = -V in 57 2 that case. For Fe Q is positive and so in this case he qQ = V _. As the |LJ values cannot be deduced from first principles (e.g. 3 for the q„,T term in eqn. 14, <r > is unknown) i t is necessary to deduce them from compounds of known structure. Since the components of the e.f.g. consist of sums and differences of values in such a manner that the addition of an arbitrary constant to each value of £LJ makes no difference 7 25 to the predicted result ' , the normal practice is to assume a value of 0.0 for one particular ligand and to derive values for the others from this arbitrary starting point. For molecules with regular geometry, the principal components of the e.f.g. tensor for various combinations of ligands (e.g. for tetra-21 hedral cases A^ BM, A2&2^> e t c-) have been tabulated in terms of the £LJ values^'2^ and also some of them are presented in Figure 4. The values of £LJ are also dependent on the coordination number of the complex, a fact which is not obvious from the derivations. For example, the value of QL^ j for the same ligand in octahedral coordina-tion has been found to be about 70% of its value in tetrahedral coordina-25 tion . This ratio has been deduced theoretically from consideration of the overlap integrals between metal and ligand orbitals in octahedral 25 and tetrahedral compounds . At the same time, the suggestion has been made that the £LJ value of a ligand in the equatorial positions is different from the ^ i f j value of that ligand in the axial positions in trigonal bipyramidal compounds. Thus the £LJ values for a ligand will be [ "1TET f~ "1 OCT Lj for tetrahedral coordination, L for octahedral r "1TBA coordination, I LJ for trigonal bipyramidal axial coordination, and t "l TBE Lj for trigonal bipyramidal equatorial coordination. The success of any model for a system is judged by how well the behaviour of the system may be predicted from the model. Certainly, the additive model for e.f.g.s has had noticeable success in predicting the sign and magnitude of the e.f.g. in many compounds (particularly of Sn(lV)) which have regular geometry''. In the converse case, namely the prediction of the disposition of the ligands about the metal from a knowledge of the sign and magnitude of the e.f.g., the success rate has not been quite so high. The reason for this is obvious since the problem is complicated by the fact that distortions from ideal geometry lead to changes in the relative £LJ contri-FIGURE 4. Some Regular Structures and the Point Charge Expressions for the Components of Their EFG Tensors. Where the Principal Axes Are Determined by Symmetry, the Components are Designated by Upper Case Subscripts. Otherwise, the Principal Axes Can Only be Found by Diagonalization of the Tensor for Each Case Under Consideration. CONTINUED/... 23 FIGURE 4 (CONTINUED) V, ZZ YY XX V. 'ZZ YY XX 0 0 0 CONTINUED/... FIGURE 4 (CONTINUED) 25 butions to the e.f.g. and by the fact that more than one possible structure may lead to essentially the same e.f.g. parameters. As yet, no completely satisfactory treatment has emerged for dealing with distortions from ideal geometry. There have been two approaches to this problem. One is just the straightforward application of the simple point charge model in which the value of £LJ is assumed to be invariant on distortion and only the changes in relative orientation of the ligands are c o n s i d e r e d " ^ . A second approach has been developed for tetrahedral systems assuming that the ligand-metal orbital overlaps 25 do not change on distortion . Both these methods have some value in predicting sign changes on distortion but have had much less success i n 25 predicting the magnitude of the e.f.g. (or conversely, the magnitude of Besides the additive models, another approach which i s sometimes of use is based on the Townes-Dailey approximation in which the effective 7 22 electron populations i n the p- and d-orbitals are considered ' . Thus i f the e.f.g. i s assumed to arise solely from p electrons the q T I A T contribution VAL may be written as ( 1 5 ) <VAL = Kp {" Np + 1 * ( N p + N P ) } _3 where K is a constant depending on <r > and N , the effective orbital P P P ± population in the appropriate p-orbital (this formulation i s only useful when the Z axis i s uniquely defined). Similarly for an e.f.g. arising solely from contributions due to d electrons 26 (16) q„ A T = K, {-N + N, + N, 1/o(N, + N, )} VAL d d ? d ? 2 d ' * d d x -y xy xz yz -3 where K. is a constant depending on <r >, and N. the effective orbital d d d. l population in the appropriate d-orbital. This approach is particularly useful for examining the effects due to partially f i l l e d d-orbitals. For example, i f d , d and d are fil l e d and d o and do ? are empty as xy xz yz z^- x -y in low-spin ferrous compounds, then the e.f.g. should be very small or zero. On the other hand, i f only d 2 is empty and the other d-orbitals f u l l , then a large e.f.g. is expected which will not vary too much with ligand substitution (provided the N^ for the other d-orbitals remains i essentially constant). This is apparently exactly the situation in 27 phosphine and arsine derivatives of Fe(CO)^ The Magnetic Hyperfine Interaction The interaction of the nucleus with a magnetic field is known as the nuclear Zeeman interaction. The magnetic moments of the ground and excited states interact with an effective magnetic field at the nucleus so that the degeneracy of the nuclear spin substates is lifted. In the case 2 15 when e qQ is zero, the Hamiltonian describing the interaction is given by ( 1 7 ) "MAG - S * n ^ where g is the gyromagnetic ratio for the state under consideration, $ n the nuclear magneton, I the nuclear spin operator, and H. the effective magnetic field at the nucleus. For this case, without loss of generality we may choose the z axis to lie along H and so the eigenvectors of the Hamiltonian are given by 27 (18) E M . r = - g 3 HmT with mT = - I , -1+1 . . . I MAG n I I where H is the magnitude of the magnetic f i e l d . Thus 21+1 equally spaced energy levels w i l l arise for both the ground and excited states. For "*7Fe the ground state, with I = ^ 12* ^ s s P ^ t into two levels whose separation is g oB nH, while the excited state, with I = 12> is s p l i t into four levels with a separation of gjj^H between adjacent pairs. The ratio between the ground and excited state g values, g and g^ 28 respectively, has been found to be g Q/g^ = 1.750 For "*7Fe the Y-transitions between ground and excited states are restricted by the dipole radiation selection rules"*'"' (Appendix II) so that only Ani-j. = 0, ±1 transitions w i l l occur. This restricts the allowed transitions to six at most and explains the familiar six line hyperfine pattern for iron f o i l . The line intensities depend on the squares of the Clebsch-Gordan coefficients and on the angular factors characteristic of 2 dipole radiation. The angular factors are 1+cos 9 for Am^. = ±1 transitions 2 29 and sin 9 for Am^. = 0 transitions . For a randomly oriented polycrystalline 30 absorber, averaging over a l l the orientations gives rise to integrals such as ( 1 9 ) 4 2fl = y- / / (sin 29) sin0d9d<j> = | sin 0 4TT o o 3 and ( 2 0 ) . . 2 F L = 7 - / / (l+cos 26) sin0d6dct) = | 1+cos 9 4 T T O O 3 and so the angular factors average to the same value. Thus the observed intensity pattern of 3:2:1:1:2:3 of iron f o i l reflects the magnitude of 28 31 the squares of the Clebsch-Gordan coefficients for the transitions connecting the various substates. 121 2 For a Sb nucleus in a magnetic field (e qQ = 0) there would 32 be a total of eighteen allowed transitions and the resultant spectrum would be quite complex. Since a l l the antimony compounds dealt with in this thesis are diamagnets in zero applied field there is no contribution from the nuclear Zeeman interaction and this interaction will not be considered further. The case which is of most interest here is that in which a nucleus is subjected simultaneously to an electric quadrupole interaction and to a magnetic hyperfine interaction. In particular for diamagnetic compounds of "^Fe and "^Sn with non-zero e.f.g.'s in moderate (25-50 kG) applied fields, 15 33 the sign of the e.f.g. and the magnitude of n may be determined ' . The Hamiltonian for such an interaction may be written (21) ^ Q • ^ G + HQ - + trfcy { ( 3 t z - + K2+f-2>> where H is the applied magnetic field. This is a fairly difficult problem to solve exactly in that the Z axis defined by the principal component of the e.f.g. will seldom be colinear with the Z1 axis defined by the applied 119 57 field. For Sn and Fe, the quadrupole moment of the ground state is zero, so that for this state we simply have = H^^. However, for the excited state the f u l l Hamiltonian (^+Q) must be used and the problem is further complicated by the fact that the case which is of most interest is that of a randomly oriented polycrystalline powdered sample subjected to an 29 external field. If the coordinate axes are taken to be the ones defined by the principal axes of the e.f.g. then the applied field will be at Z' defined 33 by the direction cosines 8, cj> . The magnetic part of the interaction may be written as (22) YL,.N = -g 3 H (sin8cos<t>I + sin8sin<J>I + cosSl ) MAG I n x y z ^ /\ <N where H is the magnitude of the applied field (H) and 1^, I , 1^ are spin 29 operators . For polycrystalline materials, a l l possible orientations of the Z' axis with respect to the Z axis as defined by the e.f.g. w i l l occur. Therefore the energy eigenvalues must be found for a l l possible values of 8 and <f>. This is usually done by computer diagonalization of the appropriate 4x4 matrices for each value of 8 and (j) under consideration. In order to simulate a spectrum, the transition probabilities connecting the various substates are needed as well as the differences in 33 the energy eigenvalues . This is a somewhat more tedious operation since one needs to know the eigenvectors in terms of the basis vectors for both ground and excited states, the Clebsch-Gordan coefficients for the appro-31 priate transitions , and then sum over the relative amounts of Y-quanta available with longitudinal and left- and right-hand circular polarizations, These computations are discussed in Appendix II. The result for an "^Fe compound with an axially symmetric e.f.g. is the appearance of a characteristic doublet-triplet pattern in moderate fields. For a positive V^, the triplet lies at low velocities and the 30 doublet at higher velocities. The order i s reversed i f V z z I s negative. When the value of the Q.S. f a l l s below about 1 mm/sec (or the applied f i e l d i s too large) the triplet-doublet observation no longer holds since some of the component lines may overlap. Similarly, as n becomes large, the spectrum in the applied f i e l d becomes more symmetrical (usually tending to a t r i p l e t - t r i p l e t ) . Nevertheless, the spectrum may be simulated and the sign of the e.f.g. and the magnitude of n estimated for most of the cases of interest. 119 Similar considerations hold for Sn except that the situation is not quite so favourable owing in part to the greater natural linewidth of t i n and in part to the larger magnetic moments of the ground and excited 119 states. Representative spectra of Sn in applied fields have been published for many cases of interest and serve as a guide in interpreting observed spectra 31 EXPERIMENTAL 121 The Sb MBssbauer spectra of two series of compounds were 121 obtained in transmission geometry using a 1 mCi Ba SnO^  source (New England Nuclear) cooled to liquid nitrogen temperature. The source, which 2 had a cross sectional area of 2.2 cm , was positioned by a phosphor-bronze drive spring mounted on a copper cold finger which dipped into liquid nitrogen. This assembly was carefully insulated with about 3 cm. of styrofoam except for a 1 cm. thick portion directly in front of the source. The source was driven via a 12 cm. long phenolic plastic drive rod, by an Austin Associates K.3-K linear motor and an S-3 spectrometer drive unit. The rest of the spectrometer consisted of a Reuter- Stokes RSG-61 proportional counter (Xe - 10% CO^ at two atmospheres), and standard Nuclear-Chicago modules. These included a model 40-9B high voltage power supply, a model 23805 preamplifier, a model 33-15 amplifier-single-channel analyzer, a model 23-4 analog-to-digital converter, a model 021308 time-base generator, and a model 24-2 400-word multichannel analyzer operating in time mode. The single-channel analyzer was set on the escape peak of the 121 32 37 keV Sb Y~ray . The absorbers consisted of finely powdered neat solids which 2 contained 8 to 10 mg. of antimony/cm . A copper cell with mylar windows which 2 had a cross-sectional area of 2.5 cm was employed. During the course of each experiment, the absorbers were maintained at 8.5 - 9.0°K in a Janis 32 model 6DT variable temperature cryostat. The temperature was monitored by a germanium resistance thermometer. A typical spectrum was run for 14 - 18 hours and consisted of two mirror image spectra containing 20 - 40,000 counts in each of the 400 channels (Figure 5). After each run, the velocity scale was calibrated using 28 35 57 metallic iron f o i l ' and a nominally 10 mCi Co(Cu) source. The zero of velocity was found by using a BaSnO^ absorber and a 5 mCi Ba^^^SnO^ 121 source mounted and run under the same conditions as the Ba SnO^ source. As a further check, a number of runs were made using the small amount of residual activity due to Ba^^^SnO^ impurity i n the Ba^^SnO^ source. The difference in the isomer shift of the BaSnO^ absorber relative to these two sources was found to be less than 0.01 mm/sec. The data points were computer fi t t e d to either a single Lorentzian or to an eight-line Lorentzian pattern appropriate for a non-zero, 32 axially symmetric e.f.g. . The program was kindly supplied by Dr. L.H. Bowen. 2 In this program for non-zero e qQ (n=0), the energies of the transitions are determined analytically and the intensities of the lines are assumed to be the appropriate combinations of the squares of the Clebsch-Gordan c o e f f i -cients for a polycrystalline absorber. The value of Q /Q = R = 1.34 was J ex ^gr 18 assumed throughout . In the f i n a l data reduction step, after the parameters for each of the two mirror image spectra were determined for comparison (this 2 is particularly important for small e qQ since any prejudice introduced by the folding procedure may have a marked effect on the f i n a l parameters), FIGURE 5. Schematic Diagram of the Apparati. 121 Employed for Obtaining Sb Mossbauer Spectra. LIQUID HELIUM RESERVOIR VACUUM SPACE STYROFOAM INSULATION PLASTIC DRIVE ROD K3K LINEAR MOTOR COPPER BLOCK LIQUID NITROGEN RESERVOIR LIQUID NITROGEN RESERVOIR OUTER WALL OF DEWAR HOLLOW TUBE FILLED WITH HELIUM EXCHANGE GAS DETECTOR (Xe-C02) ABSORBER SOURCE PHOSPHOR-BRONZE SPRING FIGURE 5. 34 the spectra were folded using the velocity scale determined from the best least-squares f i t to the iron f o i l calibration. This procedure generally led to an increase of approximately 0.05 to 0.10 mm/sec. i n the apparent linewidth. For spectra which gave poor f i t s for r\ = 0, i t was assumed that n 4 0, and they were f i t t e d using n as a variable. For this purpose, 121 36 a subroutine was written for calculating Sb spectra with non-zero n In this subroutine, the energies were found by machine diagonalization of the appropriate Hamiltonian matrices for the ground and excited states. The transition probabilities for the twelve possible transitions connecting the two states were assumed to be the appropriate combinations of the Clebsch-Gordan coefficients. The integrals for the angular-dependent part of the above were carried out assuming randomly oriented polycrystalline powders (Appendix I ) . 57 57 The Fe spectra were recorded with the 10 mCi Co(Cu) source at room temperature and the absorbers at 80°K in transmission geometry 37 with a spectrometer previously described (Figure 1). The centroid of the sodium nitroprusside spectrum was used as an isomer shift standard and the Q.S. of this spectrum was used at the velocity calibration. This procedure was adopted since the velocity range scanned was approximately -4 to +4 mm/sec A number of runs carried out at higher velocity ranges using metallic iron f o i l , show that the drive i s sufficiently linear for this purpose . A l l the iron compounds used as absorbers were in the form of 35 finely powdered neat solids. A minimum of sample thickness was used consistent with the desire to obtain good signal to noise ratio. The success of this procedure may be judged from the linewidths of the fitted spectra. For compounds containing a large arsenic to iron ratio, a number of runs were usually made until the optimal sample thickness was obtained. A l l the ~^Fe spectra were least-squares fitted to Lorentzian components by a program based on one originally supplied by the National 39 Bureau of Standards The magnetic perturbation "^Fe measurements were carried out in a Janis model 11MDT Helium cryostat with a Westinghouse superconducting solenoid capable of generating magnetic fields of up to 50 kG. The finely powdered samples were placed in a brass cell located at the centre of the applied field. The vertically mounted "^Co(Cu) source was driven, via a drive rod made from thin-walled stainless steel tubing, by an Austin Science Associates K-3 linear motor, located in the common vacuum space with the absorber (Figure 6). With this geometry, the directions of the applied field and of the beam of y~rays from the source were colinear. Spectra run at both liquid nitrogen temperature and liquid helium temperature showed no significant differences in parameters. In fitting the magnetic perturbation spectra, theoretical spectra 33 were generated using a program generously supplied by Dr. George Lang The value of g ^ B n was taken as 0.0068 mm/sec/kilogauss^ while the value FIGURE 6. Schematic of Magnetic Perturbation Apparatus 36 FIGURE 6. DRIVE CAN K-3 DRIVE •« * TO S-3 DRIVE UNIT DEMOUNTING FLANGE OUTER WALL OF DEWAR COMMON VACUUM SPACE LIQUID NITROGEN RESERVOIR SUPERCONDUCTING SOLENOID INNER VACUUM SPACE TO VACUUM LINE ft »TO VACUUM LINE LIQUID HELIUM RESERVOIR STAINLESS STEEL DRIVE ROD SOURCE SUPPORT FOR ABSORBER CELL ABSORBER CELL MYLAR WINDOWS DETECTOR 37 -1.750 was taken for the ratio (g /g^) of the g values of the ground 28 and excited states In f i t t i n g magnetic perturbation spectra for compounds with more than one iron site (iron A and iron B), the assumption was made that the contributions to the area under the absorption spectrum would be equal for each si t e . This i s equivalent to assuming that the resonance fractions for each of the two sites are equal. That this i s a good approximation i s shown by the unperturbed spectra at liquid nitrogen and at liquid helium temperature where the differences in the total areas under the f i t t e d absorption peaks for each of the sites are less than 5% for those compounds where a l l the component lines for the two sites are resolved. In the f i r s t step of the f i t t i n g procedure, theoretical spectra for the magnetically perturbed samples were calculated using the observed Q.S. for each of the sites and using values of n, = 0.0, 0.6, and 0.8. Appropriate values for the relative isomer shifts were employed and a l l possible combinations for the signs of the e.f.g. were generated, i.e. Site A+, Site B+, Site A-, Site B+, Site A+, Site B-, and Site A-, Site B-. The relationship that a change in sign of the e.f.g. gives an identical spectrum but in reverse order was employed . At this stage, examination of the spectra showed which of the combinations was the best approximation to the observed spectrum. As a refinement, i f necessary, further spectra were calculated for different values of n. or for different values of the assumed linewidths u n t i l satisfactory agreement between theoretical and observed spectra was obtained. In each of the spectra examined, the sign of the 38 larger e.f.g. could be unequivocally assigned. The value of n. is some-what more in doubt and in particular for the site with the smaller quadrupole s p l i t t i n g the uncertainty is f a i r l y large. In at least one compound, the signs of the e.f.g.s were obvious from visual inspection of the spectrum and so no further calculations were carried out. Three series of compounds were employed in this work. The f i r s t series of compounds of the type n Sb(Cp(CO)2^ e) n + x ( R = halogen, etc.; Cp = TT-cyclopentadienyl, n = 1, 2, or 3 and X = large anion) were prepared by Dr. D.J. Patmore of this department and have been characterized by I.R. techniques and chemical analyses. The preparation of these compounds has been reported ' The second series of antimony compounds of the type Ph^ ^ ^^^ n (X = halogen, etc.; n = 1, 2, or 3) have been prepared by Dr. J.N.R.. Ruddick of the Mossbauer research group following published procedures. They have been characterized by I.R. spectra and by melting points . The iron compounds were synthesized by Mr. L.S. Chia of this department and have been characterized by I.R. techniques and by chemical 4 4 4 5 analyses ' . The preparation of some of these derivatives has been 4 6 recently reported 39 RESULTS AND DISCUSSION PART 1 LFe„(CO), COMPLEXES AND THEIR DERIVATIVES (A) LFe 2(CO) 6 The structures and bonding characteristics of transition-metal carbonyls with metal-metal bonds have been of interest for some time. In particular, a number of iron carbonyls with iron-iron bonds have been studied^ 7'^ 7 >^^. In most cases such systems involve some bridging group 49 50 51 such as CO , H , SME , etc. although systems without such bridges are 2- 52 53 49 not entirely unknown (e.g. Fe 2(C0)g * and Fe.j(C0)^2 both contain a non-bridged iron-iron bond). In the past, a number of novel iron carbonyls containing unsaturated fluoroalicyclic-bridged di(tertiary arsines and phosphines) have been 37 6^ synthesized which have the general formula LFe o(C0), ' . The structures of 2. O some of these ligands L, as well as a number of other di(teriary arsines and phosphines) are illustrated i n Figure 7. Details of the preparation of some 54 of the fluoroalicyclic ligands have been published and as well, they have 47 been the subject of a recent review . The fluo r o a l i c y c l i c o l e f i n i c ligands illustrated are very versatile in that they may act as bridging groups, as for example in f^farsCo^CO)^^, as monodentate ligands as in f ^AsPFe (CO) ^  where the ligand i s coordinated via phosphorus, or as chelating ligands, as for example in f^fosFe(NO) 2^ 7. Of course, the unsaturated flu o r o a l i c y c l i c ligands also have the potential of coordinating via the o l e f i n i c double bond to form a FIGURE 7. The Structures of Some Typical Ligands Used in this Work. 1 f4fos f 6fos f 4 AsP f 6 AsP f4fars tfars 6 s(chy2 ^ ^ A s ( C H 3 ) 2 H 2 (./As(C 6H 5) 2 ^P(C 6H 5) 2 diars arphos P(C6H5) h 2 C^P(C 6H 5) 2 dppm dppe=diphos f8fos D=P(CgH 5 ) 2 , E = P ( C 6 H 5 ) 2 — D=P (C 6 H 5 ) 2 , E = As(CH 3 ) 2 f8fars D = As (CH 3 ) 2 , E = As(CH 3 ) 2 H2C As(C6H5)2 dpam H2C-P(C6H5)2 H2C H 2C-PqH 5) 2 dppp FIGURE 7. FIGURE 8. The Structure of f,farsFe 0(CO) 41 TABLE I. f.fars 4 f. fos 4 f 6 f o s f .AsP 4 f 6AsP fgfars fgfars MOSSBAUER PARAMETERS AT 80^ FOR LFe 2(CO) 6 COMPOUNDS * AEq(mm/sec) t 6(mm/sec) T(mm/sec) IRON SITE REFERENCE 1.44 0.32 0.25 B 37,47 0.64 0.28 0.25 A ** + 1.32^ - 0.66 0.32 0.26 B 37,47 0.23 0.26 A 1.19 0.32 0.23 B 37 0.65 0.22 0.23 A ** 0.31 0.23 B 44,47 - 0.83 0.27 0.23 A 1.19 0.32 0.24 B 45 0.73 0.26 0.30 A ** + 1 - 4 1 * * 0.33 0.28, 0.26 B This work - 0.67 0.30 0.24, 0.28 A 1.38 0.32 0.26, 0.28 B This work 0.62 0.28 0.28, 0.28 A f,fos Fe (C0) 7 1.55 0.32 0.23 B 48 0.17 0.20 0.25 A * Experimental uncertainty ±0.01 mm/sec. Relative to sodium nitroprusside, experimental uncertainty ±0.01 mm/sec. ** o Sign determined with source and absorber at 4.2 K in this work (See Table VI). 43 TT-complex with suitable metals. In fact, the X-ray structure of 58 f 4farsFe2(CO)g , as il l u s t r a t e d in Figure 8, reveals that the ligand is not only chelating to one iron via the two arsenics in the usual manner but that the perfluorocyclobutenyl double bond is linked to the second iron atom. In this compound, one iron atom labelled Fe i n the figure, is approximately octahedrally coordinated being surrounded by three carbonyls, two arsenics from the ligand, and the other iron atom, B B labelled Fe . The coordination about Fe is somewhat more d i f f i c u l t to A. describe. It i s obviously surrounded by three carbonyls, Fe , and the "double-bond". The coordination may be regarded as distorted trigonal bipyramidal i f the "double-bond" is considered as a single entity forming 58 a TT-bond with the iron moiety . It may also be regarded as distorted octahedral, i f the carbon atoms at each end of the "double-bond" are considered to form two normal cf-bonds with Fe and thus to occupy two sites in the octahedron^. From the crystal structure data of this"^ and 60 related derivatives i t appears that perhaps the true situation l i e s between these two extremes. In any case the "double-bond" character of the C - C link w i l l be severely reduced. The Mossbauer spectra of a number of these derivatives IJF^CCO)^ 37 44 48 have been investigated in the past ' ' and these data as well as the * 59,60 This i s not the only formulation of olefin-iron bonding . A more detailed discussion is given below in Section E. 44 present results are presented i n Table I. As well, a typical Mossbauer spectrum of one of these derivatives, namely fgfarsFe2(C0)g, i s repro-duced in Figure 9. In this figure, the absorption lines may be numbered 1, 2, 3, 4 reading from l e f t to right and the two iron sites w i l l be arbi t r a r i l y labelled I and II without distinction at this point as A B to which is Fe and which i s Fe . Using this procedure, there are three possible assignments of the lines, v i z . : (a) lines 1 and 2 to site I, lines 3 and 4 to site IIj (b) lines 1 and 3 to site I, lines 2 and 4 to Site II* (c) lines 2 and 3 to site I, lines 1 and 4 to site II. Parameters derived from these assignments are given in Table II. TABLE II. Possible Assignments of the Mossbauer Parameters for f 6farsFe 2(CO) 6. ASSIGNMENT LINES 6(mm/sec) AEQ(mm/sec) IRON SITE (a) 1,2 -0.20 0.34 I 3,4 +0.83 0.30 II (b) 1,3 +0.13 1.00 I 2,4 +0.50 1.06 II (c) 2,3 +0.30 0.67 I 1,4 +0.33 1.41 II The usual method of assigning the lines in iron carbonyl derivatives which contain two different iron sites has been to use an empirical approach so that the assignment of the Q.S. and I.S. of the two FIGURE 9. "*7Fe Mossbauer Spectra of f 6farsFe 2(CO) 6 and (PhO) 3Pf 4AsPFe 2(CO) f 45 f 6farsFe 2(CO) 6 o o ° ( P h 0 ) 3 P f 4 A s P F e 2 ( C 0 ) 5 -2 0 2 VELOCITY (mm/sec) FIGURE 9. 46 sites gives parameters which are i n reasonable agreement with the Q.S. and I.S. of related molecules containing only one site. The effects on the Mossbauer parameters of ligand variation may also be employed as an aid in assigning the lines. For example, i n the previous work of 37 Cullen et a l . on LFe2(CO)g compounds the following approach was used. F i r s t , the decision was made that the magnitude of the Q.S. of Fe was A 58 B A greater than that of Fe based on the known geometry about Fe and Fe . Secondly, assignment (a) was rejected since the Q.S. was not large enough B for Fe and the I.S. values l i e outside the range of I.S. which i s 37 characteristic of iron carbonyls . Finally, assignment (b) was rejected since the results did not accord with the predicted magnitude of the changes in the I.S. in going from f 4 f a r s F e 2 ( C 0 ) 6 to f^fosFe^CO)^ and i n going from f ^ fosFe^CO)^ to f^fosFe 2(CO)g. Thus only assignment (c) gave 37 results in accord with predicted behaviour 61 A similar empirical approach was employed by de Beer ejt a l . in assigning the four Mossbauer peaks arising from the inequivalent iron atoms in some Fe2(CO)^L(SR>2 compounds (R = alkyl, aryl, L = PPh 3, P(OMe)3, AsPh^, etc.). One possible assignment was rejected on the basis of i t s unreasonable I.S. values while a second was rejected on this basis as well, although the I.S. evidence was somewhat weaker in the latter case. The fact that the third assignment produces one set of parameters which corres-ponds to the parent compound Fe 2(CO)^(SR) 2, as would be expected, i s good evidence that this choice i s the correct one. Nevertheless, the assignment of Mossbauer parameters by these 47 e m p i r i c a l a p p r o a c h e s c a n some t imes l e a d t o d i f f i c u l t i e s and p o s s i b l y even t o i n c o r r e c t r e s u l t s . The m a g n e t i c p e r t u r b a t i o n t e c h n i q u e p r o v i d e s a me thod o f a v o i d i n g t h e s e d i f f i c u l t i e s . E a c h o f t h e p o s s i b l e a s s i g n m e n t s o f t h e l i n e s w i l l l e a d t o a d i s t i n c t i v e p a t t e r n when a m a g n e t i c f i e l d ^ i s a p p l i e d t o t h e s a m p l e . I f t h e d i f f e r e n c e s i n t h e Q . S . a n d / o r I . S . p r e -d i c t e d f o r e a c h o f t he p o s s i b l e c h o i c e s a r e l a r g e enough t h e n t h e s p e c t r u m p r o d u c e d by a s a m p l e s u b j e c t e d t o a m a g n e t i c f i e l d w i l l be u n i q u e and t h e a s s i g n m e n t of t h e l i n e s c a n be c o n f i r m e d . The m a g n e t i c p e r t u r b a t i o n t e c h n i q u e h a s a p p a r e n t l y n e v e r b e f o r e b e e n u s e d f o r t h i s p u r p o s e . F o r f g f a r s F e 2 ( C 0 ) g i n a m a g n e t i c f i e l d o f 50kG t y p i c a l s p e c t r a w h i c h w o u l d be p r o d u c e d i f a s s i g n m e n t (a) i s c o r r e c t a r e r e p r o d u c e d i n F i g u r e 1 0 . S i m i l a r l y , i f a s s i g n m e n t (b) i s c o r r e c t , t y p i c a l s p e c t r a w h i c h w o u l d be p r o d u c e d a r e shown i n F i g u r e 1 1 . T h e s e p a t t e r n s may be compared w i t h t h e f i n a l f i t t e d e x p e r i m e n t a l r e s u l t u s i n g a s s i g n m e n t ( c ) as r e p r o d u c e d i n F i g u r e 1 2 . A s c a n be s e e n c l e a r l y b y a c o m p a r i s o n o f t h e s e f i g u r e s , o n l y a s s i g n m e n t ( c ) g i v e s a p a t t e r n w h i c h i s s i m i l a r t o t h a t o f t h e e x p e r i m e n t a l r e s u l t . Thus t h e e a r l i e r a s s i g n m e n t o f c a s e ( c ) t o s i m i l a r s p e c t r a made on a more e m p i r i c a l b a s i s 3 7 i s c o n f i r m e d . A t t h i s s t a g e , h a v i n g c o n f i d e n c e t h a t t h e r i g h t p a i r i n g o f t h e l i n e s h a s been m a d e , i t i s n e c e s s a r y t o use q u a l i t a t i v e a r g u m e n t s i n o r d e r A t o f i n d t h e c o r r e s p o n d e n c e b e t w e e n s i t e s I and I I and i r o n atoms Fe and B A Fe . F i r s t l y , t h e c o o r d i n a t i o n a b o u t Fe i s r o u g h l y o c t a h e d r a l and c e r t a i n l y i n s i m i l a r c a s e s w h i c h h a v e s i x - c o o r d i n a t i o n and i r o n - i r o n bonds t h e r e a r e 52 62 no c a s e s o f a n o m a l o u s l y l a r g e s p l i t t i n g s ' e ven f o r s y s t e m s c o n t a i n i n g 48 arsines and phosphines^. Thus, a relatively small Q.S. is expected for A Fe . Secondly, by analogy to other complexes containing iron-iron bonds and olefinic l i g a n d s ^ 3 ' ^ the Q .S. for Fe^ is expected to be large and certainly for these compounds the relation that the |Q.S.| for Fe** i i A >IQ.S.I for Fe should hold. These arguments imply site I corresponds A B to Fe and site II to Fe . As a confirmation of this assignment, inspection of Table I shows that the I.S. of Fe varies markedly as the ligands are changed from f^fos to f n f a r s while the I.S.:of Fe hardly varies. This i s exactly the result expected since substitution of groups on Fe would have only a second-order effect on the parameters of Fe so that the assignment of the B A large Q.S. to Fe and the small Q.S. to Fe is confirmed. As a byproduct of f i t t i n g the magnetic perturbation spectra, the sign of the Q.S. and the value of n at each of the sites i s also found. As a further check on the assignment of the lines and to investigate the systematics of the signs of the Q.S. and the values of r| with ligand substitution the magnetic perturbation technique was also applied to the mixed ligand complex f^AsFFe^iCO)^ and to a phosphine only complex, f 4fosFe2(CO)g. These spectra are shown i n Figures 13 and 14. As can be seen from inspection of these results, the sign of i s positive and ri =0.6 for iron B while V ^ is negative for iron A. The spectra are not too sensitive to ri for iron A so not too much significance should be attached to the value derived other than n. i s either large or small. Unfortunately, the origins of the e.f.g.'s at the two iron sites in these FIGURE 10. Typical Spectra Produced i f Assignment (a) Correct with Site I = 0.34 mm/sec, n = 0 and Site II = 0.30 mm/sec, n = 0. 49 FIGURE 10. SITE I = + 0.34 mm/sec, n = SITE II = + 0.30 mm/sec, n SITE I = + 0.34 mm/sec, n = SITE II = - 0.30 mm/sec, n SITE I.= - 0.34 mm/sec, n SITE II = + 0.30 mm/sec, n SITE I = - 0.34 mm/sec, n SITE II = - 0.30 mm/sec., n I -2.0 T I 0.0 -1.0 0.0 +1.0 VELOCITY (mm/sec.) +2.0 FIGURE 11. Typical Spectra Produced i f Assignment (b) Correct with Site I = 1.00 mm/sec, n = 0 and Site II = 1.06 mm/sec, n = 0. FIGURE 1 1 . VELOCITY - m m / s e c . FIGURE 12„ Fe Mossbauer Spectrum of f^farsFe2(CO) an Applied Longitudinal Magnetic Field 50kG Showing Experimental Points and Theoretical Fit. FIGURE 12. Fe Mossbauer Spectrum of f6fars Fe2(C0)6 in a parallel magnetic field of 50 kG -1.6 -0.8 0.0 0.8 1.6 2.4 VELOCITY (mm/sec) FIGURE 13. "*7Fe Mossbauer Spectrum of f^AsPFe2(CO)g an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical Fit. Fe Mossbauer Spectrum of f4 AsPFe2 (C0)6 FIGURE 13. in a parallel magnetic field of 50 kG Site A, Q.S. = -0.83 mm/sec ,77=0.8 VELOCITY (mm/sec) FIGURE 14. Fe Mossbauer Spectrum of f 4fosFe2(CO)g i n an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical F i t . FIGURE 14. 57Fe Mossbauer Spectrum of f4fos Fe2(CO)6 in a parallel magnetic field of 50 kG Site A, Q.S. = -0.66 mm/sec,77= 0.6 -0.8 0.0 0.8 VELOCITY (mm/sec) 54 compounds are not known. This problem is discussed in more detail in section E. There are some interesting trends in Q.S. as examination of the data in Table I shows. F i r s t l y , the magnitude of the Q.S. at Fe in compounds of the type f^AsPFe^CCO)^ is larger than that of the symmetrical derivatives f fosFe_(C0), and f farsFe„(CO),. This sort of behaviour is n Z o n z o not unexpected i n view of the differences in a-donor and Tr-acceptor g a b i l i t i e s of phosphines and arsines. Secondly, the Q.S. of Fe decreases as the value of n increases in each series of compounds containing f fars, f AsP, and f fos. This trend correlates with the fact that as n . n n the size of the fl u o r o a l i c y c l i c ring is increased the steric requirements g at the carbons bonded to Fe are reduced. For example in a four membered ring, the angles C - C - C are expected to be close to 90° so when the ring size i s expanded the angles C - C - C are opened out. Thus changes g in the geometry at the carbons bonded to Fe are expected to occur as the ring size increases and this may well be reflected in the Mossbauer para-B meters at Fe . Of course, changes in the hybridization at the carbon atoms g bonded to Fe may also be influencing the Q.S. in such cases. The Q.S. of fgfosFe2(C0)^ i s also of interest. In this compound, g the ligand i s apparently linked to Fe via the double-bond in the usual A 48 manner while only one of the phosphines is bonded to Fe . Inspection of g Table I shows the Q.S. of Fe remains at about 1.5 mm/sec. as expected while that of Fe is only about 0.2 mm/sec. The decrease in the Q.S. i n the latter case i s presumably due to the less severe steric requirements at 55 A 48 Fe although the nature of the iron-iron bond may change as well 60 since the ligand constraints may not be as severe in this case. g In a l l LFe 2(C0) 6 complexes studied to date, the I.S. of Fe is A A greater than that of Fe . This implies the s-electron density on Fe i s g greater than that on Fe . To satisfy the effective atomic number (EAN) rule, Fe^ is considered 3 7 as forming a dative bond to Fe**, so i t is not surprising g to find that the electron density on Fe approaches, but i s not equal to, that on Fe . As mentioned above, there are only small variations i n the g I.S. of Fe since there is no substitution of groups on this iron atom. In the compound f^fosFe,,(CO)j only one phosphine is bonded to A 48 Fe and i n going to fgfosFe 2(C0)g where one carbonyl is replaced by a A. phosphine there i s an increase of + 0.02 mm/sec. in the I.S. of Fe . This is approximately the value usually observed on substitution of a carbonyl by a phosphine in similar compounds^ and reflects the difference in C-donor and TT-acceptor strengths between a phosphine and a carbonyl group. Similarly, there i s an increase of about + 0.04 mm/sec. i n replacing a phosphine by an arsine in going from f n f o s to f nAsP derivatives. Again 61 this Is the usual effect observed , and reflects the differences in bonding characteristics between arsines and phosphines. The only trend which may be a b i t anomalous i s that observed on replacing the second phosphine by an arsine in going from ^AsF' to f^fars complexes, where the A, increase i n I.S. at Fe varies from + 0.01 to + 0.04 mm.sec. In this g case, there seems to be a correlation between the Q.S. of Fe and the I.S. 56 A A change at Fe , since a small change in I.S. at Fe is accompanied by a B A small change in Q.S. at Fe while a large change in I.S. at Fe i s B accompanied by a large change in Q.S. at Fe . Whether this i s a consequence A B of the bonding between Fe and Fe or of the geometry adopted by the ligand A B at both Fe and Fe is not clear. The CO stretching frequencies in these compounds are also of interest and the values for the f^fos, f^fos, and f^fars derivatives have 37 been reported previously .. As expected, in a l l LFe2(C0)^ derivatives there are six CO stretching frequencies. Three of these are relatively insensitive to ligand substitution, the change in frequency ( A V ^ Q ) being -1 45 less than 7cm in a l l the compounds reported , while the other three vary markedly with ligand substitution (Av = 12 - 22cm "*"). The f RAsP and f fars derivatives have V n at lower frequencies than the equivalent V for f fos derivatives. This correlates with the differences i n a-donor n and ir-acceptor characteristics of phosphines and arsines. In summary, a generally useful application of the magnetic perturbation technique to the removal of ambiguities in the assignment of Mossbauer spectral parameters in iron carbonyl complexes having two different iron sites has been investigated. As a consequence the signs of the Q.S. and magnitude of the asymmetry parameters of some LFe^CCO)^ derivatives have been measured. The Mossbauer spectral parameters of some new LJi^CCO)^ derivatives have also been measured and the results interpreted in terms of the known structure of f 4farsFe 2(CO)g. 57 (B) L mLFe 2(CO) 5 The next class of compounds which is of interest i s that of the general type L mLFe 2(CO)^, where L™ is a monodentate ligand such as Ph^P or (PhCO^P or a potentially bidentate or terdentate ligand such as f^AsP or dppp (Figure 7) which is coordinating v ia only one of i t s possible sites. The compounds are regarded as derivatives of LFe o(C0), z o where one carbonyl is replaced by a monodentate group. These compounds may be recognized as a class by chemical analysis and by their distinctive I.R. patterns in the carbonyl stretching region which show three bands in the ranges v^ ^ = 2035 - 2044 cm - 1, v 2 = 1975 - 1990 cm"1 and v 3 = 1954 -1 Q 7 n -1 44,45 1970 cm ' A typical Mossbauer spectrum of one of these derivatives, namely (PhO) 3Pf 4AsPFe 2(C0) 5, is ill u s t r a t e d in Figure 9, and the similarity of the Mossbauer spectra of these derivatives to those of the LFe„(C0), complexes z o is readily apparent. For the two iron sites, I and II, the possible assignments of the spectral parameters are the same as for those discussed for LFe 2(C0)g complexes (Table II). Again, the magnetic perturbation spectrum (Figure 15) confirms that (c) is the valid assignment. On the basis of the similarity between the Mossbauer parameters of iAFe^CO^ and LFe 2(C0)g derivatives, lines 1 and 4 are assumed to arise from Fe and lines 2 and 3 from Fe . Using this as a basis, the spectral parameters of a number of such derivatives are presented i n Table III. FIGURE 15. 5 7Fe Mossbauer Spectrum of (PhO)3Pf4AsPFe2(CO)5 in an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical Fit. F I G U R E 15. Fe Mossbauer Spectrum of [(Ph0)3P]f4AsPFe2(C0)5 in a parallel magnetic field of 50 kG -1.6 -0.8 0.0 0.8 1.6 VELOCITY (mm/sec) TABLE III. MOSSBAUER PARAMETERS AT 80°K FOR L^Fe^CO^ COMPOUNDS L mL „m. Ph.P f.AsP 3 4 dpppmf^AsP f.AsP mf,fars 4 4 (PhO) 3P mf 4AsP (PhO) 3P mfgfars (PhO) 3P mf 4fars (Ph) 3Sb mf 4fos AE„ (mm/sec.) <5(mm/sec.) r(mm/ ** sec.) IRON SITE REFERENCE —Q 1.38 ± 0.03 0.30 + 0.02 0.27, 0.31 B 44,47 0.56 ± 0.03 0.35 + 0.02 0.27, 0.32 A 1.28 ± 0.04 0.31 + 0.02 0.25, 0.39 B This work 0.59 ± 0.04 0.32 + 0.02 0.33, 0.35 A 1.35 + 0.02 0.30 + 0.01 0.22, 0.26 B This work 0.44 ± 0.02 0.34 + 0.01 0.24, 0.26 A +1.51+ ± 0.03 0.29 + 0.02 0.31, 0.30 B This work -0.58 ± 0.03 0.29 + 0.02 0.29, 0.32 A 1.26 ± 0.02 0.31 + 0.01 0.26, 0.25 B This work 0.50 + 0.02 0.29 + 0.01 0.23, 0.26 A 1.45 ± 0.03 0.31 + 0.02 0.25, 0.28 B This work 0.64 ± 0.03 0.32 + 0.02 0.29, 0.25 A 1.27 ± 0.02 0.30 + 0.01 0.24, 0.23 B This work 0.17 ± 0.03 0.38 + 0.02 0.27, 0.28 A ** t Relative to sodium nitroprusside. Experimental uncertainty, ± 0.02 mm/sec Sign determined with source and absorber at 4.2°K (see Table VI). 60 Examination of the I.S. data in Table III shows that the I.S. of Fe increases by approximately + 0.01 to + 0.08 mm/sec. on substitution of a carbonyl by a phosphine group. Since the I.S. of Fe is less sensitive to ligand substitution (the change is - 0.01 to - 0.03 mm/sec.) i t i s reasonable to conclude that substitution takes place on Fe . As expected, the increase i n the I.S. of Fe on substituting a carbonyl by a (PhO),J? group is somewhat less (+ 0.01 to + 0.04 mm/sec.) than that observed on substituting a Ph^P or dppp group (+ 0.05 to + 0.08 mm/sec). This is reasonable since the a, TT characteristics of the triphenyl phosphite group are expected to be more similar to CO than those of Ph^P or dppp and thus the I.S. should show less change. The much larger increases i n the I.S. of Fe observed on replacement of a CO by a phosphine in these compounds, relative to that observed between f,fosFe_(CO) n and b I I fgfosFe2(C0)g, may be ascribed to a saturation of the total u-bonding capacity of the ligands i n the LmLFe2(C0),- complexes, as has been suggested for some sulphur bridged iron carbonyls^''". This is not unreasonable since in going from LFe 2(C0)^ to LJj^CCO)^ any buildup of d-electron density at A Fe due to substitution of a carbonyl may be partially compensated by increased dere a l i z a t i o n of this d-electron density onto the three remaining carbonyl groups via metal*ligand TT-donation. Thus the increase i n the I.S. of Fe^ is small. However, in the L mLFe2(C0)^ complexes, there are only two A. CO groups remaining on Fe so the degree to which excess d-electron density may be delocalized i s somewhat reduced and the increase in d-density at Fe is larger than for the LJi^CCO)^ complexes. This accounts for the larger increase i n the I.S. at Fe . 61 Normally, the increase in iron I.S. for arsenic and antimony complexes are expected to be similar to and larger than those for the corresponding phosphorus d e r i v a t i v e s 7 ' ^ . The increase in the I.S. of Fe A in going from f^farsFe 2(C0> 6 to f 4AsP mf^farsFe 2(C0> 5 is only + 0.06 mm/sec., compared to an increase of + 0.08 mm/sec. for Ph.jPmf ^ AsPFe^CO) relative to f^AsPFe^CCOg, while the increase for the Ph^Sb derivative i s + 0.15 mm/sec. On this basis, i t i s reasonable to suppose f^AsP i s coordinated through phosphorus and not arsenic. Additional support for this conclusion comes from the X-ray study"'*' of f 4AsP mFe(C0) 4 which has a trigonal bipyramidal structure in which the ligand i s monodentate and coordinated through phosphorus. Also, the chemical evidence from this and other s t u d i e s ^ shows that ditertiary arsines form weaker chelates than ditertiary phosphines so that any reaction involving competition between these products i s expected to favour formation of the phosphorus-bonded derivative. Unfortunately, i t was not possible to prepare enough of a monodentate arsenic derivative for i t to be characterized by Mossbauer spectroscopy, although I.R. spectra of two such complexes were recorded 45 and are very similar to the corresponding phosphorus compounds . The inst a b i l i t y of such complexes presumably arises because the Fe - As bond is much weaker than the Fe - P bond i n these compounds. Interestingly enough, i t was possible to characterize a monodentate stibine derivative, namely Ph.jSb mf 4fosFe 2(C0), whose I.R. pattern corresponded closely to the other derivatives. In Ph^Sb"^^fosFe,,(CO),-, the Q.S. i s much reduced over 62 A the corresponding Ph^P derivative and the increase in the I.S. of Fe on substitution is much larger (Table III). The increase of electron density on Fe in l!\Fe2 ( CO) 5 complexes over that in IJ^faO)^ derivatives should make i t a more effective a-donor. Molecular orbital calculations on such systems as M ^^O)^^ have shown that the metal-metal bond is predominantly 0" in character and i t is reasonable to suppose that such a situation might hold for the Fe — Fe bond in the present series of compounds. If indeed a-bonding is important then the hybrid orbitals involved on Fe should 3 2 3 A be dsp or d sp or some intermediate hybridization, while for Fe they 2 3 should be close to d sp . Thus, an increase in the a-donor strength of A B Fe should lead to an increase in the s-electron density on Fe . The decrease in the I.S. of Fe then follows as a natural consequence of the increase in electron density on Fe . The Q.S. parameters are also of interest. From the results of the magnetic perturbation experiment (Figure 15) i t is seen that the sign A B of the Q.S. for sites Fe and Fe remains unchanged from the previous case LFe2(C0)g. In general the magnitude of the Q.S. of Fe i s reduced over the corresponding complexes of the type LFe2(C0)g and in particular the reduction is very large in the Pb^Sb derivative. These results imply that the change in the Q.S. of Fe arises chiefly from the differences in O - and TT-bonding a b i l i t i e s of the substituents relative to CO and although geometry changes which follow on ligand substitution play some role in determining 63 the precise value of AE^ they are not the major factor. For most of B B the complexes the Q.S. of Fe is somewhat smaller than that of Fe in the corresponding LFe^CCO)^ complexes and since i s positive and corresponds to electron deficiency along the Z axis, any increase i n electron density along the Fe - Fe bond should lead to a decrease i n g the Q.S. at Fe (provided that i s directed more or less along the Fe - Fe bond). Thus the expected increase i n a-donor power of Fe owing to i t s greater electron density should lead to a decrease i n the Q.S. at g Fe as observed i n most cases. However, geometrical factors may also B play some role since the Q.S. of Fe i s increased slightly i n two cases. Thus while the I.S. acts as a f a i r l y sensitive indicator as to which iron is being substituted, the Q.S. parameters are a much less reliable guide. A similar situation apparently exists i n t r i i r o n cluster complexes with ditertiary arsines of the type LFe^CCO).^^. One of the most interesting questions about the structures of in, A L LJi^CCO)^ derivatives is which of the carbonyl groups on Fe has been replaced. There are two carbonyl groups cis to the iron-iron bond and one carbonyl group trans to this bond. At this stage, we would not expect to be able to distinguish between a substituent replacing either cis carbonyl but we might expect to be able to distinguish between cis and trans substitution. In a case in which cis substitution i s presumed to occur, namely b A for L LFe2(C0) 4 complexes (see below) the Q.S. of Fe is dramatically increased and i s positive in sign. Although there may be geometric factors 64 at work as well in L bLFe2(CO) 4 complexes nevertheless cis substitution in (CO),- could probably be ruled out on these grounds since the A Q.S. of Fe remains small and negative. The Mossbauer evidence is not c unequivocal on this point, however, since in the case of L I J ^ ^ O ) ^ complexes (see below) there is both cis and trans substitution of carbonyls at Fe^ ^ but no large increase in the Q.S. at this iron atom. Other evidence would seem to favour trans substitution. In particular, there are only three CO bands in the I.R. spectra of these derivatives in 44 45 CH2CI2 solution * and, for those compounds which have been studied, 45 only three in the solid state as well . This implies that the two A. carbonyls on Fe are equivalent or nearly so, and the structure with the highest symmetry puts the two CO's cis to the iron - iron bond. The structural study of f 4AsP Cf 4AsPFe2(CO) 4^ shows that phosphorus in the c f^AsP ligand i s substituted trans to the iron - iron bond while arsenic is c i s . Since i t is unlikely that there i s any major rearrangement in going from f 4AsP mLFe 2(CO) 5 to f 4AsP CLFe 2(CO) 4 this implies that the phos-phorus substitutes trans (as stated above) and then the arsenic displaces an additional CO to form the tetracarbonyl complex. Thus, chemical evidence also favours trans substitution. In summary, Mossbauer spectroscopy has established that i n m A complexes of the type L LFe2(C0)^ one carbonyl on Fe i n the parent LFe2(C0)g is replaced by a monodentate group. Further, both Mossbauer and I.R. evidence favour substitution trans to the iron - iron bond and both Mossbauer and chemical evidence establish that in f 4AsP mf 4farsFe2(CO)^ the f^AsP i s bonded' through phosphorus, 65 (C) L°LFe 2(CO) 4 The next class of derivatives of LFe„(CO), which has been Z b c c studied i s of the general formula L LFe 2(CO)g where L is a chelating bidentate ligand (Figure 7) which i s not necessarily the same as L. The structure of one member of this class of compounds, namely f 4AsP°f 4AsPFe 2(CO) 4, has been determined by X-ray d i f f r a c t i o n ^ and is illustrated i n Figure 16. A comparison of this structure to that of 58 f 4farsFe 2(CO)g (Figure 8) is of interest. In both compounds, the iron -iron bond i s f a i r l y long, being 2.869 A* in the f 4AsP°f^AsPFe,, (CO)^ derivative*^ compared to 2.89 A* in f ^ f a r s F e 2 ( C O ) T h e equivalence of the iron - iron bond lengths i n these two derivatives has been inter-preted as indicating that ligand constraints play a major role in deter-mining the iron - iron distance The major structural difference, other than the substitution of an f^AsP for an f^fars bridge between the two irons, is the replacement of two carbonyls on Fe by a chelating f^AsP. As mentioned above and as can be clearly seen in Figure 16 the phosphorus i s bonded trans to the iron -iron bond, while the arsenic is cis to this bond. Compounds of the type L LFe^CO)^ may be recognized as a group by their chemical analyses and by their distinctive I.R. patterns i n the 45 CO stretching region . There are two closely related classes of I.R. frequencies one of which shows three absorption bands and one of which shows four absorption bands. Each class has two high frequency bands i n FIGURE 1 6 . The S t r u c t u r e o f f . A s P C f , A s P F e , ( C O ) FIGURE 16. FIGURE 17. 57 b Fe Mossbauer Spectra of f^fos f^fosFe^ (CO)^ and f 4 A s P C f 4 f a r s F e 2 ( C O ) ^ 67 (f4fos)f4fosFe2(CO)4 o OCD o J?o°cfL&oo o CO CO CO < cr (tAsP) t f a rsFeXOl 1 . 0 0 0 1 VELOCITY (mm/sec) FIGURE 17. M'6SSBAUER PARAMETERS A L CL f,AsP Cf,AsP 4 4 f.AsP Cf.AsP 6 4 f 6AsP Cf 6AsP A,EQ (nm/sec) cS (mm/sec) 1.07 + 0.01 0.28 + 0.01 0.61 + 0.01 0.50 + 0.01 • 1.05 + 0.02* 0.28 ± 0.01 0.61 + 0.02T 0.50 + 0.01 1.07 + 0.01 0.29 + 0.01 0.67 + 0.01 0.49 + 0.01 0.97 + 0.01 0.28 + 0.01 0.60 + 0.01 0.47 + 0.01 f,AsP Cf,fars 4 6 f,fars°f.AsP 6 4 fgfars cfgfars diars f .AsP 0.96 ± 0.02 0.59 ± 0.02 1.15 ± 0.02 0.73 ± 0.02 0.91 ± 0.03 0.40 ± 0.03 0.99 ± 0.02 0.65 + 0.02 0.86 ± 0.02 0.45 ± 0.02 0.29 ± 0.01 0.47 ± 0.01 0.30 ± 0.02 0.52 ± 0.02 0.27 + 0.02 0.53 ± 0.02 0.32 ± 0.02 0.49 + 0.02 0.28 ± 0.02 0.50 ± 0.02 TABLE IV. 80°K FOR L° L Fe o(C0). COMPOUNDS r (nm/sec) LINE ASSIGNMENT IRON SITE REFERENCE 0.28, 0.28 1,3 B 45 0.27, 0.28 2,3 A 0.30, 0.30 1,3 B This work 0.29, 0.30 2,3 A 0.27, 0.27 1,3 B 44,47 0.27, 0.27 2,3 A 0.28, 0.33 1,3 B 45 0.28, 0.33 2,3 A 0.23, 0.28 1,3 B This work 0.27, 0.28 2,3 A 0.24, 0.29 1,3 B This work 0.26, 0.29 2,3 A 0,31, 0.36 1,3 B This work 0.26, 0.36 2,3 A 0.25, 0.28 1,3 B This work 0.22, 0.28 2,3 A 0.29, 0.32 1,3 B This work 0.32, 0.32 2,3 A CONTINUED/.... L CL dppm f^AsP dppmCf 4AsP + + f.AsP f Q f a r s 4 8 diphos f^AsP fgfos f 4AsP f.fos f.AsP 4 4 TABLE IV (CONTINUED) MOSSBAUER PARAMETERS AT 80°K FOR L° L Fe 2(CO) 4 COMPOUNDS A E Q (mm/sec) * 6 (mm/sec) ** r(mm/sec) LINE ASSIGNMENT IRON SITE REFERENCE 0.95 + 0.02 0.31 + 0.01 0.25, 0.22 B This work 0.36 + 0.02 0.46 + 0.01 0.22, 0.30 2,3 A 0.80 + 0.02 0.24 + 0.01 0.25, 0.30 1,3 B This work 0.51 + 0.02 0.53 + 0.01 0.22, 0.22 2,4 A 1.01 + 0.02 0.30 + 0.02 0.23, 0.26 1,3 B This work 0.64 + 0.02 0.46 + 0.02 0.24, 0.26 2,3 A 1.08 + 0.04 0.27 + 0.03 0.34, 0.46 1,3 B This work 0.59 + 0.04 0.51 + 0.03 0.36, 0.46 2,3 A 1.06 + 0.02 0.28 + 0.02 0.21, 0.32 1,3 B This work 0.64 + 0.02 0.49 + 0.02 0.27, 0.32 2,3 A 1.10 + 0.03 0.25 + 0.02 0.26, 0.36 1,3 B This work 0.62 + 0.03 0.49 + 0.02 0.24, 0.36 2,3 A Relative to sodium nitroprusside. Experimental uncertainty, ±0.02 mm/sec. Sign determined with source and absorber at 4.2°K, see text and Table VI. Other line assignment. 70 the range v^ ^ = 2002 - 2013 cm"1 and v 2 = 1939 - 1952 cm 1 4 5 . For the majority of compounds, there are only three absorption frequencies with the third one lying in the range V 3 = 1903 - 1920 cm"'1". However, there are a few compounds which have four absorption frequencies, the two components at low frequency lying in the range = 1917 - 1921 cm 1 —1 B and = 1906 - 1908 cm . Since the local symmetry at Fe is not very high and should be easily broken i t i s not surprising that this sort of behaviour i s observed. The Mossbauer spectra of these derivatives are also very distinctive since for most of the compounds there are three component peaks of relative area 1:1:2. A typical spectrum is il l u s t r a t e d i n Figure 17 while the Mossbauer parameters are tabulated in Table IV. For some of the compounds, the most intense peak i s somewhat broadened and in one case (dppm f^AsPFe^CO)^) the two components have been resolved. For those compounds with three peaks there are two possible assignments for the three lines, v i z . : (a) lines 1 and 2 to s i t e I, line 3 to site II, i f line 3 i s assumed to be a singlet; (b) lines 1 and 3 to site I, lines 2 and 3 to site II, i f line 3 consists of two over-lapping components. Again, magnetic perturbation techniques allow us to decide which of these p o s s i b i l i t i e s i s valid. From the appearance of the magnetic perturbation spectrum of f^AsP f^AsTFe^iCO)^ i t was possible to reject assignment (a) and the parameters in Table IV have been derived on the basis of assignment (b). Unfortunately, i t was not possible to achieve a very good f i t of the spectrum. For some reason, two of the components of 71 the magnetic perturbation spectrum appear to be too intense relative to the theoretical ones, so i t was not possible to find values of the Q.S., T and n which would f i t the spectrum exactly. However, the f i t became noticeably worse except when the sign of the larger Q.S. was positive and n. for this site was zero. Similarly, for the smaller Q.S., the sign was apparently negative although the f i t of the spectrum did not improve significantly as X] was varied. Despite these problems, the decision as to whether the para-A B meters of site I should be assigned to Fe or Fe is relatively easy to make. Since i n f^AsP f^AsPFe2(CO)^ we know that only the carbonyls on A. Fe are being replaced then in view of the changes in the parameters of A m Fe in ^ ^ ( C O ) ^ complexes and L LFe2(C0)^ complexes on ligand substitution we expect rather large changes in Fe , particularly i n the I.S., while for Fe , the changes i n the parameters should be somewhat smaller. Thus, the assignment of the larger Q.S. to Fe gives A E ^ = + 1.05 mm/sec. and 6 = 0.28 mm/sec. and so both the I.S. and Q.S. are i n good agreement with our previous results for LFe2(C0)g and L mLFe2(C0)^ compounds, A i i The assignment of the smaller Q.S. to Fe gives | A EQ| = 0.61 mm/sec. and 6 = 0.50 mm/sec. with a large increase i n I.S. as expected. The Mossbauer parameters of these compounds, excluding dppm Cf 4AsPFe 2(C0) 4, f a l l in the range | A E | = 0.40 - 0.67 mm/sec, 6 = 0.46 - 0.53 mm/sec. for Fe A while for Fe B, | A E | = 0.86 - 0.15 mm/sec, 6 = 0.25 - 0.32 mm/sec. Interestingly, for dppm f 4AsPFe 2(CO) 4 where the four component lines could be resolved, the parameters are such that i t i s 72 d i f f i c u l t to choose between the two possible assignments of the lines since both give parameters (Table IV) which are in reasonable agreement with the ranges given above. A number of features of these spectra are worthy of comment. A. At this stage, since there i s only one CO le f t on Fe a rather large increase in the isomer shift of Fe over the corresponding iron atom in related LFe2(C0)g and L mLFe2(C0)^ derivatives is expected and indeed this is observed. This phenomenon arises because the CO groups, which are very effective TT-acceptors, act as electron "sinks" in withdrawing electron density from the metal. When CO i s replaced by a less effective TT-acceptor, this electron density now becomes localized on the metal atom. This sort of behaviour accounts for the failure of a par t i a l isomer shift model for 7 A compounds of this type . It i s of interest to note that the I.S. of Fe seems to be independent of whether the ligands are phosphines or arsines. A c c For example, the I.S.'s of Fe in fgfos f 4AsPFe 2(CO)^ and fgfars fgfars Fe2(C0) 4 are identical despite the fact that there are three phosphines and one arsine bonded to Fe in the former while there are four arsines bonded to Fe in the latter. The large increase in I.S. (+.10 to .15 mm/sec.) expected on replacement of the penultimate carbonyl in going from L mLFe o(C0) 1. to L°LFe2(C0) 4, as. mentioned above, may swamp out the smaller variations due to the differences i n a- and TT-bonding characteristics of phosphines and arsines. The I.S. of Fe in this series of compounds varies much more than i t does i n either I J ^ ^ O ) ^ or if^LFe^W)$ compounds. The values 73 range from 0.32 mm/sec. in fgfars fgfarsFe^(CO)^(practically unchanged from LFe 2(C0)g compounds), to 0.25 mm/sec. in f^fos f^AsPFe^CO)^. Since the decrease in 6 g in LmLFe„(C0)[. compounds was attributed to a-bonding Fe . between the metal moities i t i s rather surprising to find that there are compounds of the type L LFe^CCO)^ which have practically no decrease i n g the I.S. at Fe relative to the corresponding LFe„(C0),. compound even L o though the I.S. of Fe becomes very large and i t s a-donor power i s increased. There are a number of possible rationalizations for this behaviour, but as Table IV shows, for every compound for which this. apparent behaviour i s observed, there is significant line broadening of the most intense component. If the effect of this line broadening is to g increase the apparent value of the I.S. of Fe then correction for this B behaviour gives I.S. for Fe which are less than or equal to 0.30 mm/sec. This argument also implies that the second assignment for the spectral parameters for dppp f 4AsPFe 2(CO)^ is the correct one since i t gives an g I.S. for Fe which is less than 0.30 mm/sec.. The constraints on the Fe -Fe bond length imposed by the bridging ligand limits the effectiveness of the orbital overlap between the two irons and probably plays a major role g in determining the extent of the I.S. decrease at Fe . As well, there B A may be hybridization changes in the orbitals Fe i s using to bond to Fe . B 58 A closer inspection of the geometry at Fe i n f 4farsFe 2(C0)^ (Figure 8) and f^AsP^AsPFe^CO)^ 6 0 (Figure 16) reveals that there are no 74 major changes in the bond angles of the groups bonded to Fe • For A B o example, the angle Fe - Fe - midpoint C = C which differs by only 3 between these two compounds represents one of the major changes in B B geometry at Fe . The angle C - Fe - C, where the C's originate from the double bond, remains v i r t u a l l y unchanged between these two compounds 37 and so the assumption that these angles play a major role i n deter-g mining the Q.S. at Fe is apparently untrue. „ , J 58,60 . . From these crystal structure data then i t i s apparent g that the changes in geometry at Fe are too small to account for the g large reduction in the Q.S. at Fe in these compounds relative to those in equivalent complexes of the type LFe^CCO)^ and LmLFe2(CO),.. As m B discussed above for L LFe2(C0)^ complexes, since i s positive for Fe , and i f the z-axis defined by V„„ is more or less i n the direction of the Fe - Fe bond then any increase in electron density along the Fe - Fe bond g w i l l tend to reduce the Q.S. at Fe . This would seem to be the most li k e l y explanation of the observed behaviour. The Q.S. at Fe in these compounds has about the same range of values as observed for L UJ^CCO),. complexes (except for Ph3Sb f 4fosFe 2(C0) 5) A c Again, the sign of the Q.S. at Fe is apparently negative i n f^AsP f^AsP Fe 2(C0) 4 as i t is in LFe 2(C0) 6 and L mLFe 2(C0)^ complexes. Interestingly, I i A c c A E Q | S 0.4 mm/sec. in complexes L f4AsPFe2(G0)^ when L is of the type As"~A^ s. In these same complexes | A EQ|^ i s also quite small. The explanation for this behaviour is not clear especially when i t is considered that in 75 fgfars fgfarsFe 2(CO)g, | A E Q | = 0.65 mm/sec. There i s also a notice-I i A c A E , . i n complexes of the type L f AsPFe_(C0), relative Q1 n / 4 to complexes of the type f nAsPFe 2(CO)g. The explanation of this phenomenon probably l i e s in the somewhat anomalously high values for the corresponding f nAsPFe 2(CO)^ derivatives. Considering the Q.S. at Fe for dppm f 4AsPFe 2(CO)^ we see that a l l the complexes of the type L Cf nAsPFe 2(CO) 4 where L° is a P"P or As~P then | A E ^ | A l i e s in the range i i A 0.59 - 0.67 mm/sec. so i t i s extremely unlikely that | A EQ| for dppm f 4AsPFe 2(CO)^ could be 0.36 mm/sec.. Thus, the other assignment of the spectral lines must be the valid one - the same conclusion reached above on the basis of I.S. data. So, although the origin of the Q.S. at Fe in these compounds is not clear there are certainly no large changes in the Q.S. at Fe on substitution both cis and trans to the iron - iron bond. In summary, the Mossbauer and I.R. data show that a l l the complexes of the type L LFe 2(CO) 4 have structures similar to the known structure of f 4AsP Cf 4AsPFe 2 (CO)^^. Magnetic perturbation techniques have helped to establish which iron site gives rise to which absorption 58 peaks. The known crystal structures of f^farsFe^CO)^ and f 4 A s P C f 4 A s P F e 2 ( C 0 ) 4 ^ have been used to demonstrate that i t i s the buildup A B of electron density at Fe and not geometry changes at Fe which deter-B A mines the decrease in Q.S. at Fe . The origins of the Q.S. at Fe remain unclear, although a few trends are noted. 76 (D) L bLFe 2(CO) 4 There remains one other type of derivative of LFe„(CO)> / o which has not yet been discussed. These compounds have the general formula l^LFe^iCO) ^ and may be distinguished from complexes of the type L LFe^CCO)^ by the fact that they have four I.R. active CO stretching frequencies which occur at distinctly different frequences from those of L LTe^iCO)^. The ranges of the CO frequencies are as follows: v = 1982 - 1988 cm"1, v £ = 1932 - 1948 cm"1, V 3 = 1914 - 1923 cm"1, and = 1870 - 1902 cm 1. The Mossbauer spectra of these complexes (Table V) are also very different from those of any other derivatives reported here. Most of the spectra consist of two broadened lines as illustrated in Figure 17 for f^fos bfosFe2(CO)^ although occasionally three or four lines could be resolved. Details of the line assignments in Table V will be discussed below. Since these fluoroalicyclic olefinic ligands are so versatile 47 and may act as monodentate, bidentate, or terdentate groups i t is necessary to examine these possibilities in detail. Firstly, since i t is difficult to establish the composition of carbonyls of this type exactly by chemical analysis, i t is necessary to establish whether alternative formulations such as L^LFe^CCO)^ or LbLFe2(C0)3 are possible. Firstly, since there are four CO stretching frequencies, i t is unlikely that the L b ligands could be acting as terdentate groups. Moreover, ligands such as dpam and arphos, which have similar I.R. and Mossbauer parameters to the rest of the TABLE V. MOSSBAUER PARAMETERS AT 80°K FOR L bLFe 2(C0) 4 COMPOUNDS. L bL AE^ (mm/sec) 6" (mm/sec) T (mm/sec) ' , T m REFERENCE Q(m ( , * ^sec) + r(mm/sec) L I N E ASSIGNMENT IRON SITE 1.48 ± 0.02 0.41 + 0.01 0.28, 0.26 1,3 B 1.26 ± 0.02 0.30 + 0.01 0.28, 0.29 2,3 A 1.21 ± 0.04 • 0.36 + 0.03 0.32, 0.45 1,2 B 1.21 ± 0.04 0.36 + 0.03 0.32. 0.45 1,2 A 1.30 ± 0.04 0.35 + 0.03 0.34, 0.44 1,2 B 1.30 ± 0.04 0.35 + 0.03 0.34, 0.44 1,2 A 1.48 ± 0.03 0.39 + 0.02 0.29, 0.26 1,3 B 1.35 ± 0.03 0.32 + 0.02 0.29, 0.30 1,2 A 1.84 ± 0.02 0.39 + 0.01 0.23, 0.24 1,4 B 1.28 ± 0.02 0.31 + 0.01 0.25, 0.25 2,3 A 1.48 ± 0.02 0.21 + 0.01 0.23, 0.25 1,3 B 1.64 ± 0.02 0.49 ± 0.01 0.25, 0.24 2,4 A 1.09 ± 0.03 0.38 + 0.02 0.28, 0.31 1,2 B 1.09 $ 0.03 0.38 + 0.02 0.28, 0.31 1,2 A f,AsPbf,AsP ± This work ± 44,47 1.21 ± 0.04 0.36 ± 0.03 0.32. 0.45 1,2 A f 4AsP bf 6AsP ± 1,2 B 45 f 4 A s P b f 4 f o s 1.48 ± 0.03 0.39 ± 0.02 0.29, 0.26 1,3 B This work dppp bf 4AsP 1.84 ± 0.02 0.39 ± 0.01 0.23, 0.24 1,4 B This work dppp f 4AsP 1.48 ± 0.02 0.21 ± 0.01 0.23, 0.25 1,3 B This work f 4 f a r s b f 4 A s P 1.09 ± 0.03 0.38 ± 0.02 0.28, 0.31 1,2 B This work CONTINUED/ ... TABLE V (CONTINUED) L bL arphos f^AsP dppp bf^fos dpam^f^AsP f^fos f^fos f , f o s b f , f o s + + 4 4 MOSSBAUER PARAMETERS AT 80°K FOR L bLFe 2(C0) 4 COMPOUNDS. AEQ(mm/sec) 6 (mm/sec) t T (mm/sec) LINE ASSIGNMENT IRON SITE REFERENCE 1.43 + 0.02 0.43 + 0.01 0.24, 0.23 1,3 B This work 1.18 + 0.02 0.30 + 0.01 0.24 0.26 2,3 A 1.80 + 0.02 0.42 + 0.01 0.25, 0.23 1,3 B This work 1.54 + 0.02 0.29 + 0.01 0.25, 0.25 1,2 A 1.22 + 0.07 0.38 + 0.04 0.39, 0.33 1,2 B This work 1.22 + 0.07 0.38 + 0.04 0.39, 0.33 1,2 A 1.31 + 0.05 0.34 + 0.03 0.29, 0.38 1,2 B This work . 1.31 + 0.05 0.34 + 0.03 0.29, 0.38 1,2 A + 1.35 0.38 B This work + 1.26 0.30 A Relative to sodium nitroprusside. * Experimental uncertainty i s ± 0.02 mm/sec. ** Other possible assignment of lines, see text, t t From magnetic perturbation results, see text and Table VI. 79 derivatives, can only act as bidentate or monodentate moieties. Thus the possibility of terdentate L b ligands leading to the formulation L bLFe2(C0) 3 can be ruled out. The second possibility i s that the L b ligand i s acting i n a b b monodentate fashion to give L LFe^CCO)^ complexes. This means that L g i s either bonded to Fe or that i t is bonded cis to the iron - iron bond on Fe A, since compounds of the type L mLFe2 (CO)have been established to have trans substitution. In the case of substitution on Fe , we have the two series of complexes L LFe2(C0),_ and L ^ ^ ( C O ) ^ in which there is g substitution on this iron atom, and in both series the I.S. of Fe lie s in the range 0.30 - 0.24 mm/sec. Since the the compound f^fars bf^AsPFe,,(CO)^ B A there i s l i t t l e i f any line broadening and the I.S. of Fe = I.S. of Fe = 0.38 mm/sec. no matter which of the possible line assignments is employed, monodentate substitution for CO on Fe may be ruled out. The second p o s s i b i l i t y , i f L b is acting i n a monodentate fashion, i s substitution on Fe^. In view of the fact that in L^Fe^iCO)^ where A Fe i s being substituted there are only small changes in the Mossbauer B B parameters of Fe we would expect conversely when Fe is being substituted that there would not be any dramatic change in the Mossbauer parameters of A A Fe . So neither the I.S. nor the Q.S. of Fe should be too much changed from LFe^iCO)^. Since this i s not the case, the possi b i l i t y of monodentate g substitution at Fe must be discounted. Another possibility for L bLFe 2(C0)^ structures could arise i f the g o l e f i n i c bond to Fe were broken. However, in that case possible structures 80 g would be based on five coordination of Fe . For a l l known five coordinate derivatives of Fe(CO),. which do not involve o l e f i n i c linkages the Q.S. 27 67 li e s in the range 2.0 - 3.5 mm/sec. * and so this p o s s i b i l i t y too can be discarded. We have therefore established that the ligands L b are acting as normal bidentate ligands and that the possible structures for the products must be based on the L bLFe 2(C0) 4 formulation with the basic LFe^CO)^ g skeleton intact. Possible modes of ligation at Fe now must take the form A B of intermolecular bridging, intramolecular bridging between Fe and Fe , A B or chelation at either Fe or Fe . Chelation at Fe would have to be between the two positions cis to the iron - iron bond since the other possible structure is adopted by L LFe 2(CO) 4 complexes, This type of chelation would be expected to produce a large increase in the I.S. of Fe - the range of I.S. observed A c B for Fe in L LFe„(C0). i s 0.46 - 0.53 mm/sec. while the I.S. of Fe lie s 2 4 in the range 0.24 - 0.30 mm/sec The examination of the Mossbauer para-meters of f 4 f a r s b f 4 A s P F e 2 ( C 0 ) 4 (Table V) shows the I.S. i s 0.38 mm/sec. for both iron atoms which is well outside the above ranges for either Fe B A or Fe , and so chelation at Fe can be ruled out. Similarly, chelation at B A Fe should have only a minimal effect on the parameters of Fe . As there are large increases in the Q.S. at Fe (no matter which line assignment is employed - see for example f ^ AsP^f ^ AsPFe^CO) ^  (Table V)) and in particular, b A for f^fars f^AsP there is a large increase in the I.S. of Fe as well, such a poss i b i l i t y must be discounted. 81 Intermolecular bridging between different IJ^CCO)^ moieties remains a pos s i b i l i t y . It is unlikely that this i s true since these compounds are readily soluble in CI^C^, and a molecular weight determina-tion in this solvent gave values very close to the formula weight. I . R . spectra of (Ph 3Sb)2 bLFe2(C0) 4 derivatives gave values for which l i e in the ranges reported above for the remaining L^IJ^CCCO^ derivatives, so these data tend to substantiate the absence of intermolecular bridging. Owing to their Instability, i t was not possible to purify sufficient amounts of the bis(triphenylstibine) derivatives for them to be characterized by Mossbauer spectroscopy. Thus, at this point, the possible structures for these complexes are narrowed down to ones in which the ligands are bridging intramole-A B cularly between Fe and Fe . We w i l l refer to the carbonyls trans to the Fe - Fe bond in LFe2(C0)g as the apical carbonyls and the ones cis to the A B Fe - Fe bond as the equatorial carbonyls for both Fe and Fe . Consideration of the structure of a ligand such as fgfars (Figure 7) shows that i t i s f a i r l y r i g i d owing to the perfluorocyclobutene ring and there is no possibility of bridging from either apical position to any equatorial position on the other iron atom nor from one apical position to another since the distances to be spanned are too large. Thus, there are only four possible bridging structures remaining - those between the various equatorial positions. If we denote the two equatorial CO*s on Fe as C0(1) and CO(2) respectively, then the equivalent carbonyls on Fe are; (i) the one which 82 A B l i e s roughly in the plane C0(1) - Fe - Fe , denoted CO(l') (Figure 8); A B and ( i i ) the one which l i e s roughly in the plane CO(2) - Fe - Fe , denoted C0(2') (Figure 8). There are four possible ways in which the carbonyls may be replaced, v i z . : (a) C0(1), CO(l'); (b) C0(1), CO(2'); (c) C0(2), CO(l'); (d) C0(2), C0(2'). Of course, (b) and (c) are much less l i k e l y than (a) and (d) but they cannot be discounted even though the distortions at the iron atoms are expected to be quite large and such structures should be less favourable. The magnetic perturbation technique was applied to two of these derivatives namely f^fos bf^fosFe 2(CO)^ (Figure 18) and f^AsP bf 4AsPFe 2(CO)^. In both cases the sign of the Q.S. was positive for both sites, so there i s A b a sign reversal for the Q.S. of Fe in going from LFe 2(CO)^ to L LFe 2(CO) 4 derivatives. To discover which Q.S. belongs to which iron atom in those cases where three or four lines were resolved, i t is necessary to investigate the I.S. of the two sites. From our work with L mLFe 2(CO) 5 and L CLFe 2(CO) 4 A B we have a good idea how the I.S. of Fe and Fe w i l l vary. F i r s t l y , since A B both Fe and Fe are being substituted we expect that there w i l l be an increase i n the I.S. of both iron atoms. Secondly, the change i n the I.S. on substituting an arsenic should be somewhat greater than that on substituting a phosphorus, i.e. Al.S.(As) > Al.S.(P) on the same iron atom. This i s based on the observations above for LFe o(C0), and L mLFe_(CO) c complexes. FIGURE 18. 57 b Fe Mossbauer Spectrum of f^fos f^fosFe2(CO)^ in an Applied Longitudinal Magnetic Field of 50kG Showing Experimental Points and Theoretical Fit, oo FIGURE 18. Fe Mossbauer Spectrum of (f4fosD) f 4fos Fe 2 (C0) 4 in a parallel magnetic field of 5 0 kG S i t e A , Q . S . = +1.29 m m / s e c , 77=0 Site B , Q . S . = +1.35 m m / s e c , 77 = 0 0.8 VELOCITY (mm/sec) 84 With these two facts in mind i t i s instructive to examine the I.S. of these compounds (Table V) relative to the I.S. of LFe o(C0), L o derivatives.(Table I). Denoting the change in I.S. as A l . S . and the mode of linkage of the L b ligand as (As) or (P) depending on the situation. For f 4fars bf 4AsPFe2(CO)^ there i s only one possible line assignment and one mode of linkage of the ligand L b, so we have A l . S . (As) = + 0.07 mm/sec, for Fe , and A A l . S . (As) = + 0 . 1 1 mm/sec. for Fe . Similarly, for dpam^f^AsPFe2(CO)^ we have B A l . S . (As) = + 0.07 mm/sec. for Fe , and A l . S . (As) = + 0.11 mm/sec. for Fe A. The latter values are somewhat less accurate owing to the line broadening in this compound. For f 4fos bf 4fosFe2(C0) 4, although only two lines could be resolved in the zero-field spectrum, detailed f i t t i n g of the magnetic perturbation spectrum indicated that the relative I.S. of Fe (assumed to have the B smaller Q.S.) was 0.08 mm/sec. lower than the I.S. of Fe . If we take the (common) I.S. obtained from the f i t of the zero f i e l d spectrum to be A B the arithmetic mean of the I.S. values of Fe and Fe , the following para-meters are obtained for f,fos bf.fosFe_(CO).: 4 4 2 4 A A E q = + 1 . 2 9 mm/sec, I.S. = 0.30 mm/sec. for Fe , and g A E Q = + 1.35 mm/sec, I.S. = 0.38 mm/sec. for Fe . Using this assignment, we have A l . S . (P) = + 0.06 mm/sec. for Fe , and A l . S . (P) = + 0.07 mm/sec. for Fe A, 85 A B Interchanging the assignment of the lines to Fe and Fe leads to a g decrease in the I.S. at Fe and so may be eliminated. For dpppkf4fosFe2(CO)^, i f the assignment i n Table V i s employed, we have AI.S. (P) = + 0.10 mm/sec. for Fe , while AI.S. (P) = +0.06 mm/sec. for Fe A which is i n only moderate agreement with the results for f^fos^f^fosFe^CCO)^. A B Again, interchanging the line assignments to Fe and Fe gives a g negative value for AI.S. (P) for Fe and is rejected. For dpppbf^AsPFe2(CO)4, four lines were resolved and there are two possible combinations of spectral data. Using the f i r s t set of data from Table V, we have AI.S. (P) = + 0.07 mm/sec. for Fe B and AI.S. (P) = + 0.04 mm/sec. for Fe A. A B Interchange of the roles of Fe and Fe leads to a decrease in the I.S. at Fe and is rejected. The second combination of the spectral lines for dppp^f^AsPFe2(C0)^ (Table V) gives AI.S. (P) = - 0,10 mm/sec. for Fe B, and AI.S. (P) = + 0.22 mm/sec. for Fe A, A B which i s clearly unreasonable. If the roles of Fe and Fe are interchanged then we find a decrease i n I.S. at Fe so this solution may also be rejected. Thus, the f i r s t combination of the spectral lines i s the correct one, 86 We consider now the cases where L b is a mixed ligand such as arphos or f^AsP, so that there is a possi b i l i t y of either phosphorus or arsenic being bonded to Fe**. For arphos^f^AsPFe^(CO)^ using the assign-g ment in Table V and assuming arsenic i s bonded to Fe we have g Al.S. (As) = + 0.12 mm/sec. for Fe , and Al.S. (P) = + 0.03 mm/sec. for Fe A. Another possible assignment for the structure puts the phosphorus of B A arphos on Fe and the arsenic on Fe . This gives the following parameters: Al.S. (P) = + 0.11 mm/sec. for Fe , and Al.S. (As) = + 0.03 mm/sec. for Fe A. This assignment i s rejected since AI.S.(P) i s greater than Al.S.(As) B b on Fe ( c f . Al.S. (As) for fgfars f^AsPFe^CO)^) , while Al.S. (As) for A ID. Fe i s unrealistically small i f the I.S. values for the L LFe2(C0),-complexes are considered. Another po s s i b i l i t y exists, namely that the assignment of the spectral lines i s incorrect. However, interchanging A B the roles of Fe and Fe i n Table V leads to a decrease in the I.S. of g Fe relative to i t s value in f4AsPFe2(C0)g and so this possibility can be discounted. For f 4AsP bf 4AsPFe 2(C0) 4 following the assignment in Table V b B and considering that arsenic in f^AsP is bonded to Fe we have g Al.S. (As) = +0.10 mm/sec. for Fe , and Al.S. (P) = + 0.03 mm/sec. for Fe A. Arguments similar to those for arphos^f^AsP lead us to reject other p o s s i b i l i t i e s . 87 For f^Aspk^AsPFe^CO)^ using the assignments from Table V B and assuming As i s bonded to Fe we have T J AI.S. (As) = +0.09 mm/sec. at Fe , and AI.S. (P) =- + 0.03 mm/sec. at Fe A. g However, i f phosphorus is bonded to Fe we have AI.S. (P) = + 0.09 mm/sec. at Fe B, and AI.S. (As) = + 0.03 mm/sec. at Fe A i n poor agreement with our previous results, and so this solution i s unacceptable. In this compound, there i s substantial line broadening and so the uncertainties involved in the parameters are f a i r l y large. However, in a l l three of these "mixed L^" complexes i t i s clear that the A B phosphorus end of the ligand i s bonded to Fe and the arsenic end to Fe . Finally, for f^AsP^f^fosFe2(CO)^ we have, i f the arsenic of b B f^AsP i s bonded to Fe , AI.S. (As) = +0.07 mm/sec. at Fe , and AI.S. (P) = +0.09 mm/sec. at Fe A. g If we consider phosphorus bonded to Fe we get AI.S. (P) = +0.07 mm/sec. at Fe B, and AI.S. (As) = +0.09 mm/sec. at Fe A. Either of these p o s s i b i l i t i e s is acceptable. Again, interchange of the A B assignments of the spectral parameters to Fe and Fe leads to a decrease in g the I.S. at Fe and can be rejected. In view of the results above for other f^Aspk compounds i t might be reasonable to suppose arsenic i s bonded to B Fe but in this case the conclusion i s by no means definite. 88 The Q.S. at Fe in these compounds i s somewhat of a puzzle. Not only are there substantial increases in the magnitude of the Q.S. but, in those cases in which i t has been measured, there is a sign change as well. Some of this behaviour may be rationalized by the fact that substitution occurs cis to the iron-iron bond. If is directed more or less along this bond then increases in a-donor strength and decreases in TT-acceptor strength on ligand substitution w i l l tend to change the sign of V„„. As well, there must be other factors at work to explain such large changes in A V^2« In. particular, the geometry at Fe should be quite sensitive to the nature of the bridging ligand L b, although there does not appear to be any A b anomalous increase in the Q.S. at Fe for the L ligands which contain the perfluorocyclobutene ring. As well, the strength of the iron-iron bond may also be changing. There is evidence for the weakening of the iron-iron bond in at A least some of these complexes. For example, the I.S. at Fe i n these compounds does not seem to increase as much relative to LFe2(C0)g on ligand substitution as i t does in the I^LJ^CCO)^ complexes (compare the I.S. data of Table V to that in Table III). This could arise i f less s-electron B 2 3 density is donated to Fe via the d sp hybrids and consequently more s-density A i s localized on Fe . Again, i f there is a weakening of the metal-metal bond then the Mossbauer parameters of Fe A should be dependent on the nature of l}3 and L while those of Fe^ should be quite insensitive to changes at Fe A. Moreover, as X-ray structural studies show^*^, the geometry of the ole f i n -g Fe linkage is not too sensitive to the nature of the ligand L. Both these g facts imply that the Q.S. at Fe should be nearly independent of the rest of 89 the molecule and should be primarily dependent on the nature of ligand L^. B b As an i l l u s t r a t i o n of this, the Q.S. at Fe in both f^AsP f^AsPFe^CO)^ and f 4Aspkf 4fosFe2(CO) 4 i s equal to 1.48 mm/sec. (Unfortunately, the b 45 spectrum of f^AsP fgAsPFe 2(CO) 4 was not sufficiently well resolved for i t to be f i t t e d to three peaks, although there is substantial line broadening of one of the component lines which implies the Q.S. at Fe could be somewhat larger than reported.) For dppp^f4AsPFe2(CO)^ and b B dppp f 4fosFe2(CO) 4 nearly identical Q.S. at Fe are found as well (Table V) while the Q.S. parameters of Fe are substantially different. In summary, the Mossbauer spectra of the l^LFe^CCO) ^ complexes are quite different from those of the other derivatives of LFe„(CO)<-. In 2. o particular, the Q.S. parameters of Fe are much larger and opposite in sign A B to those of Fe in other derivatives. The Q.S. of Fe shows much more variation in magnitude than i t does in other complexes. This is consistent B with substitution at Fe . We have demonstrated that only structures with ligand bridging the two iron atoms are consistent with the Mossbauer parameters and other physical evidence such as I.R. spectra. For cases in which the bridging ligand contains both arsenic and phosphorus, the I.S. data lead to a A B formulation in which phosphorus is bonded to Fe and arsenic to Fe . Distortions from ideal geometry, substitution cis to the iron-iron bond, and a weakening of the iron-iron bond have been postulated to explain the sign and magnitude of the Q.S. in these compounds. A number of conse-quences of this behaviour have been investigated and are consistent with this interpretation. 90 (E) THE IRON-OLEFIN BOND IN LFe_(C0), AND THEIR DERIVATIVES. z o The nature of the iron - olefin bond in these complexes is not well understood. There are two extreme formulations of such bonding. 59 In the first there is assumed to be a transfer of TT-electron density from the olefinic double bond to iron via essentially a a-bond with a 3 dsp hybrid orbital, accompanied by TT back-donation from f i l l e d iron * d-orbitals of appropriate symmetry to vacant TT -orbitals of the olefinic linkage. As a result of this transfer of electron density into anti-* 59 bonding TT -orbitals the carbon - carbon.bond will be lengthened . In B this description, Fe in our compounds would be considered to be essen-tially five-coordinate. In our present series of complexes, the X-ray structural studies have shown that the plane of the perfluorocyclobutane ring is not at a particularly favourable angle for TT-bonding to Fe^ The second formulation of the bonding implies that iron is six-coordinate in these complexes and that the olefin forms two normal 68 CT-bonds from carbon to iron . The carbon - carbon bond would therefore be lengthened to approximately the normal single bond distance. The first problem encountered in trying to discover which of these bonding schemes gives the most adequate description in this particular case is the fact that similar systems containing iron - iron dative bonds 37 are not at a l l common . In most examples of dinuclear iron carbonyl -olefin complexes which have been reported the iron - iron bonds are consid-ered to be "normal" shared pair a-bonds and the olefinic linkages to be 91 three electron donors. An example of such a system is cycloheptatri-enyldiiron hexacarbonyl*'9. The magnitudes of the Q.S. in such systems*'4 are not very different from those in the present cases. However, the best systems with which to compare the present compounds would appear to be o l e f i n i c derivatives of Fe(CO),.. These compounds have Q.S. of approx-imately the same magnitude as in the present case*^' 7^. Moreover, the bonding should be quite similar to that at Fe except for the formation of a o~-bond along the Fe - Fe bond axis in our case, versus the a- and TT-bonding in the case of axial carbonyls. Interestingly enough, in our compounds there does not seem to be any significant shortening of the iron - axial carbonyl bond length"^'*^ which one might have anticipated A i f there i s no TT-bonding to Fe . 27 67 Fe(CO),. and i t s non-olefinic derivatives such as Ph^PFeCCO)^ ' have | A EQ| in the range 2.0 - 3.5 mm/sec. The principal contribution to the s p l i t t i n g has been attributed to the fact that d Z2 is nominally empty. Variations in the magnitude but not the sign of AE^ are expected on axial substitution and are attributed to some participation of d Z2 in 27 ' the formation of a-molecular orbitals . In addition, the effects of equatorial substitution have been investigated both theoretically and 27 experimentally. The results of this work show that for ligands which are sufficiently like CO in their bonding characteristics that the substi-tution may be treated as a perturbation, then the effect i s to mix d Z2 with varying amounts of the other d-orbitals, so that although the magnitude of A E Q remains approximately constant, V ^ may have either sign and n any value 92 in i t s permitted range. Thus, for our compounds we can see that the B Q.S. at Fe should be dominated by the contribution from the essentially non-bonding orbit a l and the sign of the e.f.g. w i l l be dependent on the B exact details of the bonding at Fe . The model in which the two carbons from the olefin form only a-bonds with the i r o n ^ can also be used to make predictions about the e.f.g.. Again, for this theory the e.f.g. w i l l be dominated by the fact that the d-orbitals are not f i l l e d and the exact ordering of the energies of the d-orbitals w i l l depend on the details of the O and IT contributions from each of the ligands. Thus, again, the sign of the e.f.g. and the magnitude of H w i l l vary depending upon which of the d-orbitals is highest in energy and essentially unoccupied. That the e.f.g. i s really dominated by the "hole" i n the d-orbitals is amply demonstrated by a series of experiments with olefin -Fe(CO)complexes, where controlled reduction gives rise to radical anions whose Q.S. i s approximately one half the value for the neutral complex7^. This i s exactly the result one would predict i f one electron were introduced into the non-bonding o r b i t a l 7 . g It i s apparent from the foregoing that the e.f.g. at Fe is not going to be a sensitive probe for the nature of the iron - olefin bond since both extreme formulations of this linkage give rise to qualitatively g similar results, namely the e.f.g. at Fe is dominated by the unfilled d-orbitals. One thing which might be of some value in 'this connection, 93 TABLE VI. MAGNETIC PERTURBATION RESULTS* COMPOUND f 4AsPFe 2(CO) 6 f AfosFe 2(CO) 6 f 6farsFe 2(CO) 6 (PhO) 3P mf 4AsPFe 2(CO) 5 f 4AsP f 4AsPFe 2(CO) 4 f 4AsP bf 4AsPFe 2(CO) 4 f 4 f o s b f 4 f o s F e 2 ( C O ) 4 (mm/sec) n SITE + 1.45 0.6 B - 0.83 0.8 A + 1.32 0.6 B - 0.66 0.6 A + 1.41 0.6 B - 0.67 0.0 A + 1.51 0.8 B - 0.58 0.0 A + 1.07 0.0 B - 0.61** o.ot A + 1.48tt o.ot B + 1.26tt o.ot A + 1.35 o.ot B + 1.29 o.ot A Source and absorber at 4.2 K, 50 kG magnetic f i e l d applied parallel to y-beam. t Not e x p l i c i t l y f i t t e d as a parameter. ** Apparent sign, but not conclusive. t t Sign determined on inspection. 94 however, would be an investigation of the sign of the e.f.g. and the value of n for iron in a series of olefin - FeCCO)^ and olefin -Fe(CO),jL complexes where the nature of both the o l e f i n i c substituents and the ligand L were systematically varied. The results of our present series of magnetic perturbation experiments show that is positive for Fe , in every case we have measured, although there appears to be some variation in n. (Table VI). This result i s not unexpected for LFe 2(CO) 6, L LFe 2(CO) 5 and L LFe^CO)^ A complexes since substitution occurs at Fe , and available X-ray data Indicate there are only small structural changes at Fe^ ^8,60^ t ^ e L bLFe 2(CO) 4 complexes, a sign change might conceivably occur since there i s substitution at Fe . However, no such change was observed in either case in which the sign was measured. This may imply that the particular substituents employed do not perturb the system su f f i c i e n t l y for a sign change to occur. 95 PART 2 Ph c SbX Derivatives, 5-n n X-ray crystallographic studies of compounds such as SbCl,j 7 1, P h 3SbCl 2 7 2, Me 3SbCl 2 7 3, Ph^bOMe74, Ph^b(OMe) £ 7 4 and Ph^SbOH75 have shown that these compounds adopt trigonal bipyramidal structures with the electronegative groups in the axial positions. I.R. and Raman data are also consistent with this type of structure for compounds of the type R 3SbX 2 and R^SbX 7 6' 7 7 , 7 8. There are a few exceptions such as Ph^SbClO^, 77 8 which i s apparently ionic, but both I.R. and Mossbauer studies show when such exceptions occur. The lack of a Q.S. (as expected for a tetra-hedral Ph^Sb+ cation) and the high resonance fractions support ionic 8 79 structures for both Ph.SbClO. and Ph.SbBF. . 4 4 4 4 In particular, the structural parameters of Ph^SbOMe and Ph3Sb(OMe)2 are of interest since they show there are no large deviations in the 0 - Sb bond lengths or in the equatorial Ph - Sb bond lengths 74 between these two compounds . These data lend some confidence in the application of an additive model for the Q.S. to related compounds since the theory depends c r i t i c a l l y on the assumption that the ligand parameters do not change appreciably from one compound to another. The interpretation of the Mossbauer parameters for compounds of 8 18 80 this type has been f a i r l y well established ' ' and recently there has been some attempt to apply the additive model to some of these compounds7. The 121 results of the Sb Mossbauer measurements are summarized in Table VII 96 and a typical spectrum i s shown in Figure 19b. In general, our present results for the Ph.jSbX2 and Ph^SbX 8 18 compounds are in good agreement with the values previously reported ' for compounds of this type which have been summarized in Table VIII. 2 In particular, the magnitudes of e qQ and the zero n values (except for Ph^Sb(OCOCH^)2 - see below), show that these compounds adopt trigonal bipyramidal structures with effective or C^v symmetry about antimony. 76 81 82 Previous I.R. studies on these compounds ' * and the structural study of Ph^SbOH7^ have come to similar conclusions. From the data in Tables VII and VIII i t is possible to derive p.q.s. values for various ligands and then to employ these p.q.s. values 2 to predict e qQ for other derivatives as a test of the additive model. As a starting point for these p.q.s. calculations, the parameters of Ph^SbCl and Ph^SbC^ w i l l be examined. This was selected as a starting 72 point because the structure of Ph^SbC^ i s known . A l l the bond angles about Sb are within 3° of those for a regular trigonal bipyramid, so we can safely treat this molecule as being undistorted. f i TBA Cl in TBA TBH 7 Ph^SbC^ and use the relation for a B2 A^ M case (Figure 4, #6) 2 and the value of e qQ for Ph^SbC^ from Table VIII, i t i s found that VZZ = A ~ 3 [ P b ] = + 20.6 Wsec, Ph = - 6.9 mm/sec. The positive sign for V ^ indicates that the electronegative Cl groups TABLE VII. Sb MOSSBAUER PARAMETERS AT 9 K (THIS WORK). COMPOUND I.S, . (mm/sec.) 2 e qQ (mm/sec.) nt LINEWIDTH (mm/sec.) 2 ** X Ph.SbCl 4 - 5.2 + 0.1 - 6.4 + 0.7 0.0 2.8 151 Ph.SbOH 4 - 4.1 + 0.1 ~ 5.3 + 0.5 0.0 2.9 172 Ph,SbNCS 4 - 5.2 + 0.1 - 6.4 + 0.6 0.0 2.9 188 Ph 3Sb(NCS) 2 - 5.6 + 0.1 - 20.4 + 0.7 0.0 2.6 147 Ph 3Sb(N0 3) 2 - 5.7 + 0.1 - 21.3 + 1.0 0.0 3.0 147 (Ph 3Sb) 2OCr0 4 - 4.3 + 0.1 - 16.6 + 0.8 0.0 2.8 143 Ph3Sb(OAC)2 - 5.1 + 0.1 - 21.6 + 0.5 0.0 2.6 192 - 5.2 + 0.1 - 20.1 + 0.5 6.46 ± 0.04 2.7 153 Ph2Sb(O)OH - 1.8 + 0.1 - 10.9 + 0.5 0.0 3.1 148 Ph 2SbCl 3 - 7.0 + 0.1 +. 25.9 + 0.7 0.0 3.0 194 7.0 + 0.1 + 25.2 + 0.7 0.22 ± 0.05 3.0 189+T Relative to source Ba Sn0 3 < t Constrained to n. = 0.0, except where otherwise noted. ** Approximately 180 degrees of freedom, t t No significant improvement in f i t for n 4 0» TABLE VIII. 1 2 1 S b MOSSBAUER SPECTRA AT 4.2°K (PREVIOUS STUDIES). COMPOUND I.S. (mm/sec.) 2 e qQ (mm/sec.) n LINEWIDTH (mm/sec.) REFEI Ph.SbF 4 - 4.56 - 7.2 0.0 2.62 8 Ph^SbCl - 5.26 - 6.0 0.0 2.73 8 Ph.SbBr 4 - 5.52 - 6.8 0.0 2.75 8 Ph.SbNO-4 3 - 5.49 - 6.4 0.0 2.57 8 Ph 3SbF 2 - 4.69 - 22.0 0.0 2.66 8 Ph 3SbCl 2 - 6.02 - 20.6 0.0 2.55 8 Ph 3SbBr 2 - 6.32 - 19.8 0.0 2.75 8 Ph 3 sb i 2 - 6.72 - 18.1 0.0 2.58 8 (CH 3) 3SbCl 2 - 6.11 - 24.0 0.0 2.74 8 (CH 3) 3SbBr 2 - 6.40 - 22.1 0.0 2.58 8 (Ph.Sb) (- 9.69) 7 (+ 17.5) (0.0) (2.82) 8 * 121 Relative to source Ca Sn03. 99a FIGURE 19. 121 The Sb Mossbauer Spectrum of Ph 2SbCl 3 Showing F i t 121 with n. = 0 and the Sb Mossbauer Spectrum of a Typical Compound of the Type R3SbX2, Namely Ph 3Sb(N0 3) 2. 100 withdraw charge along the Z axis and so the equivalent e l l i p s o i d of charge is oblate owing to the excess of electron density in the XY plane 7, 2 Turning now to Ph^SbCl i f we assume i t s e qQ value to be the g mean of the present measurement and the one due to Long e_t al , and use the applicable equations (Figure 4, # 5), we have V z z - 2 [ c i ] T B A - 3 [ p h ] T B E + 2 [ p h ] T B A = + 6.2 mm/sec. [ 1 TBA f "1TBE Cl] and Ph are substituted, i t i s found that jj?hJ T B A = - 7.2 mm/sec.. t ~lTBA X I values can be found from compounds [ 1TBA 2 XJ values e qQ for R^SbX (same X) derivatives can be calculated as a check. The results are summarized in Table IX. With the exception of the Br compounds the agreement between 2 observed and calculated e qQ values is very satisfactory considering the g quoted experimental error limits of about ± 0.3 to ± 0.7 mm/sec. for compounds of the type Ph^SbX (Tables VII and VIII). For Ph^SbBr, the agreement is somewhat less satisfactory, and i n view of the fact that I e qQ | for R^SbB^ i s less than that for R.jSbCl2 i t i s surprising to find 2 2 Ie qQ] Ph^SbBr is greater than |e qQ| Ph^SbCl. 2 A comparison of the e qQ values derived from nuclear quadrupole 83 84 85 2 resonance (NQR) studies ' ' (these give values of e qQ/h) and those derived from Mossbauer spectroscopy is of interest. The Mossbauer 101 parameters are consistently + 0.5 to + 2.0 mm/sec. larger in magnitude than those derived from NQR studies (compare Tables VII and VIII to Table X). This systematic difference could arise from some error in 18 the Mossbauer velocity calibration, or from the value of R , or perhaps from the conversion factor from NQR frequencies to mm/sec. (Table X -86 E could be wrong). The most likely explanation, however, is a temperature dependence of the quadrupole coupling constant.since the NQR values are measured at room temperature while the Mossbauer values are measured at 4.2°K. For the compounds of the type Ph.jSbX2 studied in this work and 8 18 for those previously reported ' , there is a fairly good linear corre-lation between |e qQ| and the I.S. (Figure 20), with the exception of the point due to (Ph^Sb^OCrO^. The best least squares f i t to these points (excluding (Pb^Sb^OCrO^) gives e qQ = - 1.74 (I.S.) - 30.5 mm/sec. From the slope of this line 7 i t is possible to t e l l that in this series of compounds O bonding effects are dominant and the ligand polarities 2 govern the trends in both I.S. and e qQ. Thus F which has the highest electronegativity produces the most positive I.S. (most like Sb+~* ) and the largest |e qQ| since its cf-donor ability should be the least, relative to a phenyl group. Similarly, I has the smallest electronegativity of the X groups studied here and leads to the most negative I.S. and the smallest |e2qQ|. The large deviation of (Ph^Sb)£0(^0^ from this linear correlation is probably due to the fact that i t is the only compound in the series which 102 TABLE IX. APPLICATION OF THE ADDITIVE MODEL TO PREDICT THE e qQ VALUES FOR Rc SbX COMPOUNDS, 5-n n TBE Values Assuming ^ C l J ^ B A = 0.0 mm/sec. Value (mm/sec.) - 6.9 - 8.0 ± 0.9 Source Compound Ph 3SbCl 2 Me 3SbCl 2 SbCl r * TBA C "1 TBA Values Assuming Cl = 0.0 mm/sec. Value (mm/sec.) Source Compound - 0.2 + 0.3 - 0.7 - 0.1 + 0.2 - 7.2 - 0.3 Ph 3SbBr 2 Ph 3SbF 2 Ph 3SbI 2 Ph3Sb (NCS)2 Ph 3Sb(N0 3) 2 Pb^SbCl Ph.SbOH 4 3. Predicted and Observed Values of e qQ. Compound Ph.SbBr 4 Ph.SbF 4 Ph.SbNO, 4 3 Ph.SbNCS 4 (CH.)_SbBr, 3 3 i Predicted e qQ (mm/sec.) - 5.9 - 6.9 - 6.7 - 6.5 - 23.2 Observed e qQ (mm/sec.) - 6.8 - 7.2 - 6.4 - 6.4 - 22.1 * See Text. TABLE X. NQR • DATA - AT ROOM TEMPERATURE. COMPOUND e2qQ (MHz) 1 2 1Sb EQUIVALENT e2qQ + REFERENCE (mm/sec.) (Signs Assumed) Me 3SbCl 2 660.393 - 22.04 83 Me 3SbCl 2 662.18 - 22.10 84 Me 3SbBr 2 631.127 - 21.06 83 Me3SbBr2 630.57 - 21.04 84 Ph 3SbCl 2 592.35 - 19.77 84 Ph 3SbBr 2 565.78 - 18.88 84 Ph 3SbF 2 603.91 - 20.15 84 (Ph3Sb) (509.00) (+ 17.0) 84 SbCl 5 at 210°K 84.67 2.83 85 SbCl 5 at 249°K 84.63 2.82 85 1 MHz = 3.337x10" mm/sec. using E ( Sb) = 37.15 Kev? and the relationship 6E = E. FIGURE 20. 121 Correlation of the Sb I.S. and 2 e qQ for a Number of Derivatives of the Type Pb^SbX^ 104 105 may be formulated as Ph^SbXY. Thus i f one of the groups (oxygen in this case) is more effectively able to compete for electron density than the other (CrO^) then there w i l l be deviations from regular behaviour. That this i s probably the situation is illustrated by the structure of (Ph 3SbN 3) 20, where the Sb - 0 distance of 1.985A 8 7 is significantly shorter than the Sb - 0 distances in Ph3Sb(0Me)2 (2.033^ , a v g . ) 7 \ Ph^SbOMe (2.061A ) 7 4 and Pb^SbOH (2.0481 ) 7 5 . Moreover, the I.S. and 2 e qQ values for (Ph 3SbCl) 20 and (Ph3SbBr>20 measured at liquid nitrogen g temperature also show significant departures from regular behaviour (the e qQ values at 78 K are inherently less accurate, however). As well, departures from predicted I.S. behaviour may be characteristic of Sb - 0 bonding (see discussion of Ph2Sb(0)0H below). The Ph^SbX derivatives also 2 f a i l to show a linear correlation of the I.S. and e qQ. There i s , however, a general trend to more positive I.S. with increasing electronegativity in Ph.SbX derivatives. 4 There are two other I.S. trends in these compounds which are worthy of comment. The f i r s t i s that the I.S. of Ph 3SbX 2 is more positive than Me3SbX2 as expected i n view of the better o donating power of alkyl [ 1 TBG Mel t l TBE Ph . Similar behaviour has been found i n t i n chemistry where the p.q.s. value for Me is more negative than that for Ph 25 in both tetrahedral and octahedral complexes , Further, this trend in I.S. has also been observed for Ph^bCp (C0) 2FePF 6 and Bu 3Sb(CO) 2FePF 6 (Bu = n-C^Hg) (see Part 3, below). The second trend is that for a given X the 106 I.S. of R^SbX is more positive than R^SbX^. Again, a similar trend to higher s-electron densities at the nucleus in R^SnX2 compounds relative to R^SnX compounds has been observed in tin chemistry 7. This phenomenon is discussed in more detail below. If the structure of Ph^SbCl^ is assumed to be trigonal + — 88 bipyramidal like the similar t i n compound, Et^N Me2SnCl3 , then in 2 order to calculate a value for e qQ"using the additive model, i t i s f "IT BE necessary to have a value for Cl A reasonable choice for a model compound from which to derive [ "1TBE o ClJ might appear to be SbCly Crystal structure data at -30 C show SbCl^ to adopt a regular trigonal bipyramidal structure 7 1, although the fact that the axial Sb - Cl bond lengths of 2.34A , avg., are somewhat - 72 shorter than those in Ph^SbC^ (2.48A. , avg.) might make the application of the additive model to this compound somewhat questionable. The Mossbauer parameters at 77°K for SbCl,. were measured by Bowen 89 2 et a l . and the value of -4.4 mm/sec. was found for e qQ on f i t t i n g the data to a quadrupole-split pattern. However, there is some evidence from 85 NQR studies that SbCl,. undergoes a phase transition, and in fact i t may 85 — + exist as a dimer or as an ionic species such as SbCl, SbCl, at low 6 4 90 temperatures. In view of this fact, Stevens and Bowen have reinterpreted 121 their earlier Sb data for SbCl^ and found that an equally good f i t could be obtained for two single line Lorentzian components, consistent with the existence of two slightly different sites i n the compound. 80 Surprisingly, Bancroft et a l . have recently employed the 107 89 original values of Bowen et a l . , citing the NQR data of Schneider 85 and Di Lorenzo as showing that there are no appreciable changes in the NQR frequencies above and below the transition temperature (presumably o 80 the NQR frequencies at 210 and 249 K form the basis of this statement ). In fact, i n the NQR study, the authors were unable to obtain consistent values of n and eQq (sic) for antimony below 195°K, and the six resonances which were obtained imply the presence of at least two different kinds of antimony^. Since the data of Bowen et al.^ were obtained at 77°K, this 2 — 80 means that the assignment of a negative sign for the e qQ of SnCl,. 2 based on i t s equivalence with the originally reported sign of e qQ of SbCl^ must be considered as doubtful at best. It i s possible to convert the NQR frequencies of Schneider and 85 2 Di Lorenzo to give a value of 2.82 mm/sec. for the |e qQ| (the sign i s indeterminate) of SbCl^ above 249°K where i t i s trigonal bipyramidal 7 1. 80 This result indicates that the correlation of Q„ and Q„, by Bancroft et a l . Sn Sb J  2 is not necessarily in error, provided that the sign of e qQ for SnCl5 i s the same as that for SbCl,.. f ~\ TBE Thus, the situation for deriving a ^ClJ value from SbCl,. i s not too favourable. Nevertheless, an estimate for the possible range of r XBE values for Clj can be made i f the two possible signs and the magnitude 2 2 of e qQ from NQR studies are employed. For e qQ = + 2.82 mm/sec, ^ C l J ^ B E = + 0.9 mm/sec. while for e^qQ = - 2.82 mm/sec, £ciJ^ B E = -0.9 mm/sec.. 108 Using these values i t i s possible to calculate the e.f.g. components for Ph^SbCl^. An earlier X-ray structural study had character-ized Ph^SbCl^ as being trigonal bipyramidal with two axial chlorines, two 91 equatorial phenyl groups, and one equatorial chlorine atom but refine-92 ment of the structure showed the. compound under study was really Ph^SbCl.j'l^O which i s six coordinate. In our present case the compound has been properly characterized as Ph2SbCl.j . Thus, as previously mentioned, we shall assume the structure of Pb^SbCl-j i s very much like + — 88 that of the corresponding tin species, Et^N Me^nCl^ If this i s indeed the structure then application of the additive model gives the following components of the e.f.g. for Ph^SbCl^ (Figure 4, # 8) using £ciJ T B E = - 0.9 mm/sec: vzz " I WTBE " [C1JTBE " 2[ C 1] T B A = " 16-* V Y V = - 2 [ p h ] T B E - [ C l ] ™ + 4 [ d ] T B A = + 14.7 mm/sec. VXX " -iH"'' + 2 [ d ] T B E - 2 [ d ] T B A - + 1.7 mm/sec. 2 Thus, V z z = - 16.4 mm/sec, H. = 0.8 and e qQ = + 16.4 mm/sec. * The analysis of Ph2SbCl.j was performed by P. Borda of this department. Anal. Calculated for Ph 2SbCl 3: C, 37.67; H, 2.49; Cl, 27.85. Found; C, 37.52; H, 2.42; Cl, 27.78. 109 f" 1TBE Similarly, i f Cl = + 0 . 9 mm/sec. is employed: = - 18 .2 mm/sec. = + 12 .9 mm/sec. Vjg£ = + 5 . 3 mm/sec, 2 whence V z z = - 18 .2 mm/sec, n = 0 . 4 and e qQ = + 18.2 mm/sec. These values may be compared with the experimental result 2 which shows e qQ = + 2 5 . 2 mm/sec. (Table VII). Although the sign of 2 e qQ is predicted correctly, there is a lack of agreement in the magnitude. [ 1 TBE Cl values but is most li k e l y a result of distortions from regular geometry. In Me2SnCl.j the o 88 angle C - Sn - C i s about 140 and so i f Pl^Shd^ i s also trigonal bipyramidal a similar distortion of the C - Sb - C angle would not be unreasonable. Considering the effect of such distortions using the simple point charge approach shows that as the angle C - M - C i s increased from 120° (assuming the distortion i s in the C^SbCl plane), the magnitude of 21 V z z becomes larger . The equation describing this behaviour i s VZZ = " 2 [ C 1 ] T B A ~ [ c i ] ™ + 2 ( 3 s i n V l ) [ p h j T B E , where <f> i s the angle C - Sb - Cl. If we use cj> = 110° corresponding to C - Sb - C = 140° and \C1J = - 0 . 9 mm/sec, we find V z z = - 2 1 . 8 mm/sec, n = 0 . 3 . On the f ~l TBE Clj = + 0 . 9 mm/sec, then V z z = - 2 3 . 6 mm/sec, n = 0 . 1 , in more satisfactory agreement with the experimental result V z z = - 2 5 . 2 mm/sec. , ( r w 0 . 2 ? ) . 119 A recent Sn Mossbauer study using the magnetic perturbation 2 technique has established that e qQ is positive and n l i e s in the range 110 + - 93 0.5 - 0.7 in Et.N Me0SnCl0 . 4 2 3 In the present work, i t was not possible to establish the value of n with certainty for Ph^SbCl^ (see below), although i t is possible to estimate that O^ n^ O.6. The structure of Pt^SbCl^ cannot be definitely assigned as trigonal bipyramidal from the present work, however, although the Mossbauer parameters are consistent with this formulation. The reason for this discrepancy is that the formulation of the structure in terms of a six-coordinate moiety with bridging Cl groups and trans phenyl groups is also consistent with the observed parameters. The possibility of strong interactions via chlorine bridges and a six-coordinate structure has been 94 2 recently advanced for Me2SbCl3 • A very crude estimate of the e qQ value [ 1 OCT f" 1 TBE Ph = Ph and [C1]°CT = [ c i ] B R I D G I N G = [ c i ] T B A . We have (from Figure 4, # 12), VZZ = 4 ( j k ) 0 C T ~ 4 [ c i ] 0 C r - -27.6 mm/sec. If the bridging Cl's are cis to each other then, n. ^  0 (but small) , while i f they are trans, r) = 0. This result is not inconsistent with the experimental data. To this time, there has been no Mossbauer I.S. data published for organoantimony compounds with six-coordinate structures nor indeed any data on other organoantimony compounds of the type R^ SbX^  and so i t is not possible to choose between the five- or six-coordinate structures on the basis of the I.S. data. Thus, no definite conclusions may be reached regarding the structure of Pt^SbCl^ at 9°K on the present Mossbauer results. Unfortunately, we were unable to establish with certainty the value of n for Ph^SbCl^. When the spectrum was fitted with n as an adjustable I l l 121 parameter, the value derived was 0.22 ± 0.05 which for Sb Mossbauer 90 spectra is l i t t l e different from zero . Further, there was essentially 2 no improvement in the goodness of f i t (x dropped from 194 to 189) and visually, there was also no apparent improvement. It is possible that the (visually) poor f i t of the spectrum (Figure 19a) is a result of departures from theoretical intensity ratios due either to the Gol'danskii-18 Karyagin (G.K.) effect as previously reported by Stevens and Ruby for compounds of the type R.jSbX2, or to non-random orientation of the crystallites in the powdered sample. It is surprising that none of the other compounds of the type R^ SbX^  which were examined in this work and for which one might expect the G.K. effect to be at a maximum, showed any evidence for G.K. asymmetries (see for example Figure 19b). Indeed, the whole question of observing the G.K. effect in antimony compounds at 4.2°K is very intriguing. Firstly, one might expect that the anisotropics in the Debye-Waller factors at 4.2°K would be very small in any case, and certainly for any tin compounds in which the effect has been observed, i t is quite temperature dependent and is nearly absent at or below 77°K. However, the G.K. effect is also somewhat dependent on 16 the y-energy and this could conceivably account for differences between tin and antimony compounds. The other possible explanation for deviations from theoretical intensity ratios is non-random orientation of the crystallites in the powdered sample. This phenomenon can be distinguished from a genuine G.K. effect by changing the relative angle of incidence of the Y-beam and the 112 absorber. If the intensity ratios are observed to vary then the G.K, effect may be ruled out - the converse i s , of course, not necessarily true. Stevens and Ruby do not report whether such a test was carried out on their compounds'^. It i s interesting that for the antimony compounds where G.K. 18 effects have been claimed to occur the published spectra show a super-f i c i a l resemblance to those in which the asymmetry parameter, n, takes on 90 small, non-zero values. From the work of Stevens and Bowen , i t is obvious that i f n>0.4 then the energies and the transition probabilities are sufficiently different from the case when n = 0 that there w i l l be no doubt that these two effects may be distinguished. From the work of 18 Stevens and Ruby i t is not clear whether n was allowed to vary as a parameter before or after correction for the apparent G.K. effect and so i t i s not possible to judge whether they could distinguish unequivocally between these two effects. Unfortunately, i t is not possible for us to examine this problem in more detail since there is only one compound in this work for which the anomalous behaviour was observed and at present we have no computer subroutine to hand for the analysis of possible G.K. effects. Also, our present experimental set up (Figure 5) does not allow for any alteration of the source-absorber geometry and so i t was not possible to carry out an orientation study. I.R. studies on Ph3Sb(0C0CH3)2=(Ph3Sb(0AC)2) and Pb^Sb(0C0CD3)2 as well as their trimethylantimony analogs show the presence of ester-like 113 95 acetate groups . Further, there is no appreciable change in the frequencies v(C=0) and v(C-O) between solution and solid state spectra nor any significant differences i n the behaviour of v(C=0) and v(C-O) between triphenyl- and trimethylantimony derivatives. The I.R. spectra of the trimethyl derivatives show only an asymmetric Sb-CH^ stretching frequency (v (Sb-CH^)); the symmetric Sb-CH_ mode (v (Sb-CH_)) is absent in the I.R. spectrum but appears as a strong band in the Raman^. Thus trigonal bipyramidal structures with planar SbC^ groups have been assigned to these compounds^'9"'. The I.R. and Raman spectra of the triphenyl analogs show both V (Sb-C) and V (Sb-C) vibrations, but breakdown of the a s local symmetry approximation was invoked to rationalize these results and 76 planar SbC^ groups have been assigned to these compounds as well . Since V(C=0) and V(C-O) are very similar i n both series of compounds such a structural assignment does not seem unreasonable. If Ph^SbCOAC^ were trigonal bipyramidal with axial acetates and equatorial phenyl groups at 9°K then the value of n from the Mossbauer spectrum should be zero, or at least very small. Thus the f a i r l y large value of n (0.46) derived for Ph.jSb(0AC)2 (compare Figures 21a and 21b) implies that i t does not have this structure at 9°K. The most l i k e l y sort of structural change would be to a six coordinate structure, probably with one bidentate acetate group and one ester-like acetate group. A bidentate acetate group has been postulated to occur in Ph^SbOAC^1'9*' and the existence of both bidentate and ester-like acetate groups has been postulated i n 97 f ~|OCT Ph2Sb(OAC) 3 . A p.q.s. calculation for Ph 3Sb(0AC) 2 assuming [PhJ 114 [ P h ] T B E = - 6.9 mm/sec. and [ 0 A C ] B I D E N T = [ o A c ] E S T E R = 0.0 mm/sec. and a regular mer - octahedral structure (Figure 4, // 15) gives | e qQ | = 20.7 nm/sec, ri = 1.0. Small distortions from regular geometry could 2 conceivably lead to the observed values of e qQ = - 20.1 mm/sec. and ri = 0.46. Of course, a similar six-coordinate structure with bridging 2 acetate groups would also be in good agreement with the observed e qQ and 98 r i . Such bridging carboxylates are commonly found in organotin chemistry Calculations for the corresponding fac - octahedral structures (Figure 4, 2 # 14) give e qQ - 0 and so such a formulation must be rejected as incompatible with the observed values. Another po s s i b i l i t y would be a trigonal bipyramidal structure with one axial and one equatorial acetate group, as in Figure 4, # 8. If f "1TBA f~ "1 TJ3FJ we use OACI = OAC] = 0.0 mm/sec, and the appropriate phenyl values, then we derive |e qQ| - 11 mm/sec. and n - 1. It i s d i f f i c u l t to 2 conceive of distortions from this structure which would lead to e iQ - 20 mm/sec. and n, - 0.5 and so this structure, too, must be rejected as unreasonable. 99 Further studies are being conducted in these laboratories i n order to elucidate more clearly the bonding interactions in these carboxylates, In particular, low temperature I.R. studies of v(C=0) and v(C-0) should distinguish bidentate from ester-like carboxylates. FIGURE 21. Sb Mossbauer Spectrum of Ph3Sb(OAC)2 Showing Improvement of F i t for n = 0.46 (b) Over That for n = 0.0 (a). TRANSMISSION 116 The Mo'ssbauer parameters f o r Ph^SbCOjOH are quite d i f f e r e n t from those of Ph^SbCl^, whereas i f these compounds were i s o s t r u c t u r a l i t might be a n t i c i p a t e d that the data would be f a i r l y s i m i l a r . One might expect the structure of Ph^Sb^OH to consist of t r i g o n a l bipyramidal moieties with b r i d g i n g 0-atoms. The ready s o l u b i l i t y of the r e l a t e d 97 a l k y l compounds i n organic solvents i n d i c a t e s that i f such a bridged structure were adopted then i t would probably consist of oligomers perhaps being the five-coordinate antimony analog of BuSn(0)OH (Bu = n-C^Hg). The structure of t h i s t i n species i s proposed to be a c y c l i c trimer with Sn - 0 - Sn l i n k a g e s 1 ^ . I f indeed the i n d i v i d u a l Pb^SbCOjOH units have a t r i g o n a l bipyramidal structure then there are a number of p o s s i b l e arrangements of the ligands about antimony. A p p l i c a t i o n of the add i t i v e model f o r the Q.S. shows that some of these arrangements are incompatible with the experi-L "IBRIDGE 0 f "I TBE From our previous discussion of the C l values and the knowledge that [ O H ] ™ - [ci]™ we expect that [ o ] B R I D G E w i l l l i e i n the range - 1.0 to OH w i l l l i e i n the same range. We also w i l l assume that £OHJ w i l l have a value i n i t s range such that w i l l be at a maximum or a minimum as the case may be, i n order that the l a r g e s t range of ^ZZ v a l u e s ke produced. Using the appropriate phenyl values from 2 Table IX we f i r s t examine the range of p o s s i b l e e qQ values f o r a s t r u c t u r e with a x i a l phenyl groups. From Figure 4, # 6, we have, V Z Z - 4 [ p h ] T B A - { o ] B R I D G E - [ O H ] T B E 117 Thus, V z z should l i e in the range - 31.8 mm/sec. < 4 - 25.8 nm/sec. while n i s small. From this treatment we see that V i s far too large in magnitude and has the wrong sign, so this structure may be rejected. A structure with equatorial phenyl groups i s also possible and this gives (Figure 4, #8) '» - ! M I B E - M™.- W B A In this case has values in the range - 20.2 mm/sec. ^  •< - 14.2 mm/sec. while n l i e s in the range 0.5 - 0.9. Departures from regular trigonal bipyramidal geometry could conceivably lead to the observed result of V^z = +10.9 mm/sec , n = 0 but in view of. our previous discussion of distortions in such systems the angle C - Sb - C i s anticipated to become larger than 120° on distortion and so the value of should become even more negative. Thus, although this structure cannot be eliminated, i t must be regarded as extremely unlikely. Finally, for a structure with one axial and one equatorial phenyl group (Figure 4, #9) we have, v z z = | [ o ] B R I D G E - [ p h ] T B E - [ p h ] T B A - [OH]TBA which gives + 17.6 mm/sec. ^ V z z > + 10.6 mm/sec, n. = 0.3 - 0.4. This in much better agreement with the experimental results. Thus, i t i s possible to find a five-coordinate structure for 2 Pl^Sn^OH whose calculated e qQ is in reasonable agreement with the observed value. In light of the distortions which should be produced by the bridging oxygen atoms, i t i s unlikely that a regular structure w i l l 118 be adopted by this compound. The large difference in magnitude and the opposite sign observed for V^z in Ph^SbCl^ and Ph^SbCCOOH may thus be rationalized by assuming these compounds are not isostructural. It i s the I.S. of Ph^SbCOOOH, however, which is really somewhat anomalous as compared to that of Ph^SbCl^. In addition to Ph^SbtCOOH, other compounds containing antimony-oxygen linkages have I.S. values which are somewhat more positive than might have been anticipated on the basis of electronegativity arguments. For example, the I.S. of Ph^SbOH is more positive than that of Ph^SbF while the I.S. of (Ph^b^OCrO^ is more positive than that of Ph.jSbF2. These data suggest that i t i s probably the Sb - 0 linkage which is responsible for at least part of the large increase in the I.S. In this regard, i t would be of interest to measure the I.S. of the polymer Ph^SbO. There are other factors which may influence the I.S. as well. From t i n chemistry, we may extrapolate a number of trends. F i r s t l y , an increase in the coordination number is generally associated with a decrease in the I.S. at the tin nucleus"^ 1. Since 6r/r i s opposite in sign for 119 121 Sn and Sb, we would expect a more positive I.S. at the antimony 119 nucleus on increasing the coordination number. Secondly, the Sn I.S. values of compounds of the type R^^SnX^ i n i t i a l l y increase as n is increased, reaching a maximum at n = 2, and then decrease again 1^ 1. On simple electronegativity grounds, however, one might expect the I.S. to decrease monotonically. The explanation advanced for this phenomenon is that when the organic R group i s replaced by a more electronegative ligand, 119 some rehybridization takes place, leading to more s-character i n the Sn - C bonds and more p-character in the Sn - X bonds. Thus, although the X groups are more electron withdrawing, the electrons withdrawn are primarily p in character and in fact the net result i s less shielding of the s-electrons by p-electrons and an increase in the L S . ^ . Of course, this process must reach a maximum at some stage and then a decrease in the I.S. should be observed. Thus, for antimony compounds of the type ^SbX^ we would expect an i n i t i a l decrease in the I.S. as n increases and indeed this i s observed for n = 1, 2 (Tables VII and VIII). For antimony, however, the value of n which minimizes the I.S. has not been established but i t i s obvious from our discussion of the mechanism involved that i t may well depend on the electronegativity of the X-group and should be particularly sensitive to the hybridization at antimony. For example, in compounds of the type R^SbX^ which adopt trigonal bipyramidal structures, the I.S. should be different for the case when the two organic groups are equatorial relative to the cases in which they are both axial or one axial and one equatorial. Thus, the much more positive I.S. of Pb^Sb^OH relative to Pl^Shd^ may be explained in part by the nature of the Sb - 0 linkage and in part by assuming that these two compounds are not isostructural. In summary, the Mossbauer data for compounds of the type R^SbX and R,SbX are consistent with trigonal bipyramidal structures with the 120 X groups in axial positions. The additive model for Q.S. was applied 2 to these derivatives and the calculated e qQ values were in good agreement with the experimental ones. For compounds of the type Ph.jSbX2 a linear 2 relation was found between the I.S. and e qQ, the slope of which was consistent with a-bonding effects being the dominant factor in determining the Mossbauer parameters. (Ph^Sb^OCrO^ was found to be an exception to the above linear relationship. Possible structures were suggested for Pt^SbCl^* but on the basis of present evidence no choice could be made between a five-coordinate structure with equatorial phenyl groups and six-coordinate structures with trans phenyl groups and bridging chlorines. The Mossbauer parameters of Ph^SbCOAc^ were found to be inconsistent with the formulation of this compound as a trigonal bipyramidal species. Octa-hedral structures with one ester-like and one bidentate acetate group are 2 compatible with the observed parameters. For Pb^SbtWOH both the e qQ value and the I.S. are quite different from those of Pt^SbCl^ but these differences may be rationalized by assuming that the two compounds are not isostructural. It has been found that i n compounds containing the Sb - 0 linkage the I.S. values are apparently more positive than would be predicted on the basis of electronegativity arguments. 121 PART 3 RnSb(Fe(CO)2Cp)4_nX COMPOUNDS. The nature of heteronuclear metal-metal bonds has been the subject of much recent discussion, particularly in compounds where a group IV element is bonded to a transition metal such as manganese, iron or cobalt. Compounds containing Fe-Sn bonds have been rather extensively - A v v . 102-109 , _ , 57^ , 119_ studied by Mossbauer spectroscopy , where use of both Fe and Sn resonances has led to valuable insights concerning the nature of the Fe-Sn a-bond and the question of possible (d-d)" bonding. From data on a number of derivatives of the types CpFe(CO)L.SnR.j and CpFeL2.SnR.j (Cp = 7T-C^H^; R = Cl, CH^ , C,H-; L = tertiary phosphine, arsine, stibine) , i t has been suggested1^2 that the Fe-Sn bond is essentially pure a in character. Greenwood1^, Donaldson1^7 and their respective co-workers have 2 reported signs of the quadrupole coupling constants e qQ for both iron and 2 tin in CpFe^O^SnCly The positive e qQ for tin indicates an excess of ? z electron density on the tin atom, from which Donaldson1^7 has concluded that the Fe-Sn bond must have predominantly G g character. Bryan11*"1 has arrived at similar conclusions on the basis of X-ray crystallographic studies. To date the only Mossbauer study of the Fe-Sb bond using both "*7Fe and 1 2 1Sb resonances is of the two compounds Ph.jSbFe(C0)4 and (Ph^Sb^Fe^O) ^> i n which the group V element acts as a two-electron donor 1 1 1. The 1 2 1Sb data were interpreted 1 1 1 as indicating very l i t t l e , i f any, TT back-donation from fi l l e d 3d orbitals on iron to vacant 5d orbitals on antimony. 122 The preparation of a number of group V derivatives of transition metal carbonyls in which the group V element acts as a one-41 electron donor has recently been reported and i t seemed attractive to investigate the Mossbauer spectra of a f a i r l y extensive series of such 42 compounds containing Fe-Sb a-bonds . Since isoelectronic t i n and antimony derivatives having identical ligands are expected to show essentially linear correlations for both isomer shifts and quadrupole splittings, we have chosen for the present study cations of the type X nSb(Fe(CO) 2Cp) 4 n + (X = c l , Br, I, CF^, Cg^, n - C ^ ; n = 1, 2, 3, but not a l l combinations). The corresponding neutral tin species have been widely studied by Mossbauer 102-107,112,112 . „ . , _ . , v , • spectroscopy ' and X-ray structural data are available in ,,110,114-116 _ „. , t. . many cases as well ' . Furthermore, although there is a paucity of X-ray data for compounds with Fe-Sb bonds, the structures of two of the present derivatives, v i z . Cl 2Sb(Fe(CO) 2Cp) 2 SbCl 7 and ClSb(Fe(CO) 2Cp) 3 2 118 FeCl 4.CH 2Cl 2 , have been published. Both compounds consist of discrete cations in which the antimony atom i s in a very distorted tetrahedral environment, associated with large, replaceable anions (Sb-jCl^ and 2-FeCl^ , respectively). The corresponding t i n compounds consist of four-coordinate neutral s p e c i e s 1 1 4 11*'. In the case of Cl 2Sb(Fe(CO) 2Cp) 2SbCl^ there are two independent cations in the unit c e l l , but they have essentially the same geometry about antimony 1 1^. 123 121 .. 121 (A) Sb Mossbauer parameters. The Sb Mossbauer spectra of compounds containing tetracoordinate antimony species such as those reported here are expected to follow certain trends, which may be illustrated by a pre-liminary consideration of the data given in Table XI. F i r s t l y , since these derivatives should be isoelectronic with the corresponding neutral 21 25 t i n species, an additive model ' as discussed above for the quadrupole 2 24 coupling constants e qQ should be applicable , and there should be a 119 more or less direct correlation with Sn quadrupole spli t t i n g s . Thus, on the basis of 1 1 9 S n data [e2qQ| in the series XSb(Fe(CO) 2Cp> 3 + i s expected to increase with changes in X in the order Bu, Ph < I < Br, C l , as observed. Secondly, in compounds containing cations of the type X 2Sb(Fe(CO) 2Cp) 2 + i f the X^Sb^X angle is considerably less than 109.5° (as indeed i t is i n Cl 2Sb(Fe(CO) 2Cp) 2 S b 2 C l y 1 1 7 , the asymmetry parameter 2 H should be less than unity but considerably greater than zero, and e qQ should be opposite i n sign but of approximately the same magnitude as i n + 19 25 the corresponding XSb(Fe(C0) 2Cp) 3 derivatives ' . Table XI shows that this i s i n fact the case (see also Figures 22 and 23). It should be noted that the n values observed (0.44, 0.46) for the Cl 2Sb(Fe(CO) 2Cp) 2 + species 19 are slightly smaller than that reported (0.65) for the corresponding t i n compound Cl 2Sn(Fe(C0) 2Cp) 2. This is consistent with the fact that Fe^-Sb^Fe > Fe^Sn^Fe 1 1 4 , 1 1 7. Thirdly, the e2qQ value for Sb(Co(CO) 3PPh 3)^ is zero within experimental error, as anticipated for a tetrahedral antimony derivative with four identical ligands. 121 A closer look at the systematics of the Sb quadrupole coupling TABLE XI. 1 2 1 S b MOSSBAUER PARAMETERS FOR COMPOUNDS OF THE TYPE R Sb(Fe(CO)„Cp), X* n I 4-n COMPOUND [ci2Sb(Fe(C0)2Cp)2J|PF^| (ci 2Sb(Fe(CO) 2Cp)[Cr(SCN) 4(NHj) [fir2Sb (Fe(CO) 2Cp) ^  [PFg] ((CF 3) 2Sb (Fe(CO) 2Cp) Jj jjCr(SCN) 4(NH 3) J (ciSb(Fe(CO)2Cp) £ J JFeCl^ (BrSb(Fe(CO) 2Cp) 3j[PF 6J "lSb(Fe(CO)2Cp)3~|[l3j _ISb(Fe(CO)2Cp) J j [ l » F g J fphSb(Fe(CO) 2C P g [ P F ^ J [ph2Sb (Fe (CO) 2Cp) J P F ^ j 6*b (mm/sec.) 2 e qQ (mm/sec.) n T(mm/sec.) 2 c X -9.3 + 0.2 +29.0 ± 1.0 0.46 ± 0.05 2.8 + 0.1 206 (-9.4 + 0.2) (+31.4 ± 1.1) (0.0) (2.7 + 0.1) (263) -9.1 + 0.2 +28.7 ± 0.5 0.44 ± 0.04 2.9 + 0.1 183 (-9.2 + 0.2) (+30.2 ± 0.6) (0.0) (2.9 + 0.1) (213) -9.6 + 0.3 +26.6 ± 0.4 0.44 ± 0.06 3.0 + 0.1 203 (-9.7 + 0.3) (+28.6 ± 0.4) (0.0) (3.0 + 0.1) (334) -8.3 + 0.2 +18.3 ± 0.7 0.68 ± 0.08 2.4 + 0.1 187 (-8.4 + 0.2) (+20.1 ± 0.9) (0.0) (3.4 + 0.2) (265) -8.8 + 0.1 -23.9 ± 1.7 0.0 3.1 + 0.1 198 -8.6 + 0.3 -23.8 ± 0.7 0.0 3.1 + 0.1 225 -8.8 + 0.3 -20.5 ± 1.0 0.0 3.1 + 0.1 191 -8.8 + 0.2 -22.4 ± 0.2 0.0 2.9 + 0.1 205 -7.9 + 0.1 Unresolved 0.0 3.2 + 0.1 205 -7.0 ± 0.2 -7.0 ± 0.4 0.43 ± 0.16 2.8 + 0.1 116dd (-7.0 ± 0.2) (-6.8 ± 0.4) (0.0) (2.9 + 0.1) (118)° (-7.0 + 0.2) (+3.4 ± 1.4) (0.0) (3.4 + 0.1) (158) G CONTINUED/. to TABLE XI(CONTINUED). 121, Sb MOSSBAUER PARAMETERS FOR COMPOUNDS OF THE TYPE R^Sb^e (CO) 2Cp) ^  X COMPOUND [jh 3SbFe(C0) 2Cp] (PFg'J (Run 1^ (Run 2 ) f [Bu3SbFe(CO)2Cp)j [ P F J [sb(Co(CO)3PPh3)4J jPFg] 6^  (mm/sec.) 2 e qQ (nun/sec.) n r(mm/sec.) 2C X -6.7 ± 0.2 +9.6 ± 0.5 0.0 2.8 ± 0.1 149 (-6.5 ± 0.2) (-4.5 ± 1.3) (0.0) (3.6 ± 0.2) (208) -6.7 ± 0.2 +9.4 ± 0.4 0.0 2.9 ± 0.1 72 d (-6.5 ± 0.2) (-2.9 ± 0.8) (0.0) (4.0 ± 0.2) (147)' -6.9 ± 0.1 +6.9 ± 0.5 0.0 2.9 ± 0.1 169 (-6.8 ± 0.1) (-4.2 ± 0.8) (0.0) (3.3 ± 0.1) (200) -8.7 ± 0.2 0.0 0.0 2.9 ± 0.1 185 8L 2 o Samples contained ca. 10 mg. Sb/cm and gave absorption intensities of ca. 10-25% at 9 K. Values in parentheses are alternate f i t s of the data. In most cases we feel there is a, 2 significant reduction i n X f° r o u r preferred solution. b Isomer shifts are relative to a Ba 1 2 1 S n 0 3 source at 80°K. c d e f Approximately 180 degrees of freedom unless otherwise noted. Approximately 100 degrees of freedom. No significant improvement in f i t for n 4 0. Two independent measurements. 126a FIGURE 22. 121 Sb Mossbauer Spectrum of Cl2Sb(Cp(CO)2Fe)2+PFg- Illustrating Improvement of Fit for n = 0.46 (b) Over That for n = 0.0 (a). i 126 FIGURE 22. 1 .00-.98i .96-.94-Oj o Cl 2Sb(Cp(C0) 2Fe)2 PF 6 with 77 = 0.0 Q) 0 0 0 0 0 o° Cl 2Sb (Cp(C0^Fe^P^ with 77 = 0 .46 -20 -10 0 10 Velocity (mm/sec) FIGURE 23. 121 Sb McSssbauer Spectrum for CISb(Cp(CO) 2Fe) 3 +PF 6~• 127 CISb (Cp(CO)2FeSPfT -24 -16 -8 0 8 16 Velocity (mm/sec) FIGURE 2 3 . FIGURE 24. Sb Mossbauer Spectrum of Ph^SbFe(CO)2Cp PFfi Showing Alternate Fits to the Data. In (a) the Fitt i n g 2 Parameters Were 6 = -6.7 mm/sec, e qQ = +9.4 mm/sec, T = 2.9 m/sec, n. = 0.0. In (b) , <5 = -6.5 mm/sec, 2 e qQ = -2.9 mm/sec, T = 4.0 mm/sec, n = 0.0. 2 The F i t with e qQ>0 i s Clearly Preferable. Ph 3 SbCp(CO) 2 Fe + PI^" Ph 3 SbCp(CO) 2 Fe + PF6' e2qQ = -f 9.4 mm/sec e 2qQ = - 2.9 mm/sec FIGURE 24. 129 constants i s of interest. As discussed above, there are two major contributions to the el e c t r i c f i e l d gradient at antimony, and ^y^-In the compound Sb (Co (CO) ^ PPh^) ^ +PF^ one expects that - 0 so that 2 only q should contribute to the e.f.g.. Since e qQ = 0, with a l i n e -J_iA-L width of 2.9 mm/sec, an upper limit of a few mm/sec. may be set for any contribution from q j ^ j * Similar conclusions may be drawn from data for •+• — 2 8 Ph^Sb ClO^ , where again e qQ = 0 with a linewidth of 2.6 mm/sec. . It i s thus clear that in the present derivatives q, T A T makes the dominant contribution to the e.f.g at the antimony nucleus. 121 2 Perhaps the most interesting Sb e qQ results reported here are those for the organoantimony derivatives. For both R^SbFe(CO)^Cp+ 2 complexes (R = Bu, Ph), e qQ is apparently positive (see Figure 24). In contrast, the sign of e^qQ(11^Sn) in Bu^SnFe(CO)2Cp is reported"*"*^ to be 2 119 109 112 negative, and a negative e qQ( Sn) has been predicted ' for 119 Ph3SnFe(CO)2Cp also. Since the sign of the quadrupole moment Q for Sn 121 i s the same as the sign of Q (and Q ) for Sb, these results indicate gr ex that the principal component of the e.f.g. tensor, V„„, i s opposite i n sign i n the antimony and t i n compounds. For the antimony derivatives reported here, i t i s apparent from trends in the isomer shift (vide infra) that the electron density i n the region of the bonds decreases in the order Sb-M > Sb-R > Sb-X, and the same ordering i s deduced from data on corres-ponding t i n compounds1^. On this basis alone one would expect e^qQ to be positive for both antimony and t i n in the triorgano derivatives. However, because of the f a i r l y small difference in electron density at Sb between 130 Sb-M and Sb-R bond directions (and at Sn between Sn-M and Sn-R bond directions), relatively minor changes in p-electron donor and acceptor properties between R. and M groups vis-a-vis tin and antimony could presumably account for the observed sign reversal. 2 The additive model for e qQ should apply to these derivatives as i t does to the Rc SbX„ derivatives previously discussed. Since for 5-n n t- j these compounds, X-ray structural studies 1 1 show large deviations from regular tetrahedral bond angles we shall not carry out p.q.s. cal-culations explicitly since the theories for describing such distortions 7 25 are inadequate at present ' From data on a variety of tetrahedral tin compounds the magnitudes 25 112 of the p.q.s. values (which are negative quantities) are found ' to decrease in the order Bu > Ph > Fe^O^Cp, which is consistent with the 2 negative e qQ in Bu^SnFe^O^Cp and the fact that the magnitude of the Q.S. is greater for this compound than for the corresponding triphenyltin species. For our R^SbFe^O^Cp"*" derivatives, the positive coupling constants require that the p.q.s. value for Fe^O^Cp is more negative than that for either R group. This, together with the observation that 2 2 e qQ(Ph) > e qQ(Bu), means that the p.q.s. values for the antimony complexes decrease in magnitude in the order Fe^O^Cp > Bu > Ph, the same order expected from isomer shift data, since the I.S. of the butyl derivative is more negative than that of the phenyl derivative. It is interesting to note that in the compounds Ph.jSbFe(CO)4 131 2 121 111 and (Ph^Sb^FeCCCO-j, e qQ( Sb) is also positive . In fact, the values 121 2 2 of Sb e qQ and isomer shift 6 for these compounds (e qQ = + 9.0, + 10.9 mm/sec; & = - 6.62, - 6.65 mm/sec, respectively) are nearly identical v/ith those of Ph^SbFe(CO)2Cp+. Thus, while i t is usual to consider that in one case antimony i s acting as a two-electron donor to iron and in the other as a one-electron donor, the overall electron configuration at Sb and the electron density i n the Sb-Fe bonds are essentially the same in both cases and the distinction is a purely formal one. In the series Ph S b ( F e ( C O ) 2 C p ) ( n = 1, 2, 3), |e qQ j at Sb decreases with decreasing n. If we assume the bond angles are such that Fe-Sb-Fe > Fe-Sb-R > R^ -Sb^ R, as expected on the basis of X-ray structural data for similar t i n 1 1 4 and phosphorus 1^ derivatives and the known 121 structure of (CH^) 3 sbF e(CO)^ , we find that the molecular orbital 25 2 treatment of Clark et a l . predicts just the reverse order of |e qQ| 25 values. It should be noted that Clark's model assumes that p.q.s. values of the ligands are constant throughout such a series. However, there is some evidence from "^Fe Q.S. data discussed below that this i s not the case, since (AE^)^ is found to decrease i n the order Ph„SbFe(CO)_Cp+ > Q Fe 3 2 Ph 2Sb(Fe(CO) 2Cp) 2 + > PhSb(Fe(CO) 2Cp) 3 + (Table XIII), indicating changes in the electron distribution i n the Fe-Sb bonds. The observed trend in |e qQ| values is probably best rationalized in terms of this effect which would produce a decrease in the difference of relative p.q.s. values of Fe(C0)2Cp and Ph. 121 The second parameter of interest i s the Sb isomer shift ^g^' 132 Typical ranges of values for 6 ^ are -19 to -9 mm/sec. for Sb(lII) compounds and -7 to +4 mm/sec. for Sb(V) compounds7 (relative to 121 Ca SnO^). Normally, 6 ^ for organoantimony (III) derivatives f a l l at the most positive end of the range for Sb(III) while those for organoantimony (V) f a l l at the most negative end of the range for Sb(V) . The 6 values for the present series of compounds l i e between -9.6 and -6.7 mm/sec, neatly spanning the range of values from Sb(III) to Sb(V). 103 122 If the arguments of Zuckerman and co-workers * in assigning oxidation states to the corresponding isoelectronic t i n species as Sn(IV) are used, our compounds must be regarded as derivatives of Sb Sb(V). However, <Sg^  of organoantimony compounds such as (p-ClC^H^)^ and (p-CH^OC^H^^Sb56 which certainly would be regarded as Sb(III) derivatives, are more positive (i.e., more "Sb(V)-like") than those of X 2Sb(Fe(CO) 2Cp) 2 + (X = Cl, Br). Similarly at the other end of the scale i f we were to regard the present species as derivatives of Sb(III), where for example R<jSb: acts as a two-electron donor toward iron, we find some g compounds such as Ph.jSbI2 and Ph2SbCl.j (Table VII) with shifts more negative than those of R.jSbFe(CO)2Cp+ (R = Ph, Bu). This apparent dichotomy is perhaps not surprising in view of the fact that the isomer shift results from the particular electronic configuration about the metal nucleus which is essentially a continuous function, while the oxidation number i s a discontinuous function. Thus the assignment of a particular oxidation state for antimony in compounds of this type has l i t t l e j u s t i f i c a t i o n . 133 It has been suggested 1^ that 6g n should be used to assign the valency of tin rather than the formal oxidation state. On this basis the isomer shifts of our tetravalent antimony compounds span the range from trivalent to pentavalent antimony derivatives, behaviour which in no way seems unusual. In the compounds X2Sb(Fe(CO)2^p)2 o n e m -^§b.t expect that on substituting a more electronegative group X the s-electron density at the antimony nucleus would decrease due to more eff i c i e n t electron with-drawal. Since the change in nuclear charge radius between excited and 121 ground states (6r/r) i s negative for Sb, 6g^ should then increase towards more positive values i n the order Ph < CF^ < Br, Cl, whereas the reverse order i s that observed (Table XI). The explanation for this effect parallels that for tin-transition metal complexes1^3»107 where the relative amounts of p-character used in the M-X bonds increase i n the order Ph < CF^ < Br, Cl, and thus the amount of s-character in the M-Fe bonds w i l l increase in the same order. The fact that the s-electron density at the Sb nucleus increases with increasing s-character of the Sb-Fe bond shows that the Fe(CO)2Cp group i s a better donor than halogen, CF^ or Ph groups. For a series such as X^_ nSb(Fe(CO) 2Cp) n + the greater donor strength of the Fe(CO)2Cp moiety i s expected to dominate the trend in s-electron density so that should become more negative as n increases This trend i s observed for Ph^ nSb(Fe(CO) 2Cp) n +, but both X2Sb(Fe(CO)2Cp) (X = Cl, Br) complexes have 6_, values more negative than those of the OD 134 corresponding XSb (Fe(CO) 2Cp) . j + derivatives. This suggests that in the dihalides the s-character in the X-Sb bonds is already so low that replacement of an X by Fe(CO)2Cp leads to l i t t l e i f any increase in the total s-character in the Fe-Sb bonds, and the isomer shift i s thus dominated by an increase in p-shielding. Paralleling this there is a general trend to smaller increases in s-electron density at t i n for each increment in n as n increases in ~~ i * <-u - v 103,105,107,125,126 , , • • ' complexes of the type X^ nSnM n (where M represents a transition-metal carbonyl group). Indeed, for the series C l ^ nSn|Mn(C0)^j^ as n increases from 2 to 3, and for I 4_ nSn^Co(C0)^ j n ^ ^ ^ as n increases from 3 to 4, the s-electron density at t i n actually f a l l s , an effect similar to that observed here. While i t i s clear from the above discussion that the system-121 atics of Sb isomer shifts show many parallels to the corresponding t 119 Sn systems, i t is worth examining the correlation in more detail. Using the data of Ruby et a l . 1 7 (with a correction of + 0.20 mm/sec. 119 127 for a systematic error in their Sn parameters ) one obtains the linear correlation between and 5 1 1 Q for isoelectronic antimony 1 1Sb ± i y S n 7 and t i n compounds shown in Figure 26. Ruby's values for <S 1 9 1 were 121 S b converted to a scale relative to Ba Sn0„ (= < 5 , 9 , relative to 121 17 127 128 Ca ^nO^) using the value ' * - 8.62 mm/sec. for the isomer shift 121 of InSb relative to Ba SnO^- Similarly, a value of 1.95 mm/sec. for 119 119 the isomer shift of a- Sn relative to Ba SnO^ (= Sn02) as derived 129 7 from ref. 17 (accepted values are 2.0 to 2.1 mm/sec. ) was used to 135 convert the corrected <5^ g values of ref. 17 from relative 119 S n 119 S n to a- Sn to 3-^9 relative to Ba SnO^* This procedure was adopted ^ n l l 9 121 119 121 since the points 6- Sn, 8- Sn and SnC^, ^20^ were 128 available independently as a cross-check . We have gathered i n Table XII the available isomer shift data for nominally isoelectronic pairs of antimony and tin compounds of the types X^ nSb(Fe(C0)2Cp) n + and X 4_ nSn(Fe(C0)2Cp) n, and these points are also displayed i n Figure 25. It should be noted that in the region below the straight line 2 121 2 119 ¥ ( Sb) > ¥ ' ( Sn), while the opposite i s true i n the region above s s the line. Consider f i r s t the dichloro and dibromo derivatives. As can be seen from Figure 26 both points l i e below the isoelectronic line. That this should be so is a consequence of the slig h t l y different structures of Cl 2Sb(Fe(CO) 2Cp) 2 + 1 1 7 and Cl 2Sn(Fe.(CO) 2Cp) 2 1 1 4. The Fe-Sb-Fe angle (134.7°, avg.) is significantly greater than the Fe-Sn-Fe angle (128.6°) in their respective compounds. This indicates that the Fe-Sb bond should have more s-character than the Fe-Sn bond so that the 121 s-electron density at the Sb nucleus should be somewhat greater than 119 that at the Sn nucleus. Note that although the compounds are isoelectronic this does not mean that the charge densities at the nuclei 128 w i l l be equal only that they w i l l be equivalent. For the compounds Ph^ nSb(Fe(CO) 2Cp) n + and Ph^ nSn(Fe(CO) 2Cp) n i t appears (Figure 26) that there i s a trend toward somewhat less than 121 equivalent electron densities at the Sb. nucleus as n increases. It i s TABLE XII. ISOMER SHIFTS OF NOMINALLY ISOELECTRONIC ANTIMONY AND TIN COMPLEXES. 6 , (mm/sec.) 5_ (mm/sec) COMPOUND / n i * . T3 1 2 1 c n ^  KEF« COMPOUND n ^ _ 119c n . REF. (Rel. to Ba SnOy (Rel. to Ba SnO^) -9.3 -9.1 -5.9 -6.0 Cl 2Sn(Fe(CO) 2Cp) 2 1.95, 1.98 d,e Jci 2Sb (Fe (CO) 2C P) 2 J Jcr(SCN) 4(NHg) 2J ^Cl 2Sb(Fe(CO) 2Cp) 2J p F ^ j Br 2Sb(Fe(CO) 2Cp) 2JpF 6J -9.6 a Br 2Sn(Fe(CO) 2Cp) 2 1.99J f PhSb(Fe(CO) 2Cp) 3J [PF 6J -7.9 a PhSn(Fe(CO)2Cp)3 2.00^ g jPh2Sb(Fe(C0)2Cp)2j p F ^ j -7.0 a Ph 2Sn(Fe(CO) 2Cp) 2 1.74J g |l>h3SbFe(CO)2Cp| Q>F^j -6.7 a Ph3SnFe(CO)2Cp 1.43J, 1.41 g.h p h A s b ] [ c i o 4 J [Ph4Sb] [BF 4] JBu3SbFe(CO)2Cpj |PF^J -6.9 a Bu^nFe(CO)2Cp Ph4Sn 1.20J, 1.22 g,l 1.47 a This work. b Ref. 111. C Ref. 56. d Ref. 104. 6 Ref. 106. f Ref. 107. 8 Ref. 105. h Ref. 102. }3 i 119 H.A. Stockier and H. Sano, Trans. Faraday Soc, 64, 577 (1968). J Converted to Ba SnO- scale assuming 6(a-Sn) =• +2.10 mm/sec..7 FIGURE 25. The Correlation of Sn and Sb Isomer Shifts. The Straight Line, Based on the Assumption of Equivalent Electron Density at the Two Nuclei, is after Ruby. The Points for the Isoelectronic Pairs Are Labelled Using the Notation M = Sn or Sb+ and Fe = Fe(C0)9Cp. FIGURE 25. 2.00- Br* MFe< CUMFe: PhMFe3 Ph2MFe2 o CD 1.50-1.00-Bu 3MFe^ H PhJVIFe PK.M 4 o _o .50-a) Sb isomer shifts relative to Ba Sn 03 source 119 119 b) Sn isomer shifts relative to Ba SnO* 0.0--10 -8 -6 - 4 -2 a) Velocity (mm/sec) i—* o 138 d i f f i c u l t to establish i f this apparent trend is really significant i n view of the uncertainties involved in the slope and position of the isoelectronic line. It should be noted however that the error bars for the ^219 values are rather misleading, since the major source of error Sn arises from a possible systematic error in converting from a-Sn to BaSnO^ reference value. Thus the apparent gradient of the data points has somewhat more significance than the individual values. If the trend i s a real one i t implies that the difference in the amount of s-density donated by phenyl and by FeCCO^Cp i s less for 121 119 Sb than for Sn. This effect could arise in two ways. If there is a change in hybridization between the two series of compounds such that there is more s-character in the Sb-C than Sn-C bonds (and correspondingly less in the Sb-Fe than Sn-Fe bonds), then since Fe^O^Cp is a better 121 donor than phenyl the s-electron density at the Sb nucleus w i l l not 119 change as rapidly as that at the Sn nucleus. This seems unlikely i n view of the results for the dihalogeno complexes discussed above. The other possibility i s an increase in any (d-d)TT interaction in the Sb-Fe bond over that in the Sn-Fe bond. This would be expected since the greater % e££ of antimony w i l l contract i t s 5d orbitals and lower them in energy relative to t i n , so that any Fe -»• M back-n-donation should be enhanced for M = Sb. This interaction should lead to a lowering of the 121 119 s-electron density at Sb relative to Sn by increased shielding as the number of Fe^O^Cp groups i s increased. Clearly the best pair of compounds to distinguish between these two effects is Sn(Fe(C0)2Cp)4 and Sb(Fe(C0)2Cp) 4 +, but a l l attempts to prepare the latter derivative have 139 42 thus far been unsuccessful . For this pair any hybridization effects should be minimized so that i f the trend is due to a-bonding d i f f e r -ences only the point should f a l l below or on the isoelectronic line. Conversely, i f (d-d)TT interactions are responsible they should be maximized for this pair and the point should l i e above the isoelectronic line. While there thus remains some uncertainly on the basis of 121 Sb Mossbauer data concerning the ir-character of the Fe-Sb bond (see 121 below), i t is clear that the Sb isomer shifts are determined primarily by a-bonding effects and that any possible 7T-interactions play a purely secondary role. 57 •• 57 (B) Fe Mossbauer Parameters. The Fe isomer shifts i n the present derivatives (Table XIII) f a l l in the narrow range 0.38 - 0.42 mm/sec. (relative to sodium nitroprusside). The narrow linewidths (0.23 - 0.27 mm/sec.) indicate that in any given compound a l l the iron atoms are in essentially identical environments. The range of 6^^ values is nearly the same as that reported for FeXCO^Cp groups bonded to tin 1^ 2,104,106^ There may be a trend to slightly higher values in the antimony derivatives (see Table XIV) but a lack of published information on the "*7Fe resonances 130 in the t i n complexes and the inherent d i f f i c u l t y of comparing small differences i n isomer shifts derived from different sources precludes detailed analysis. It i s worth considering i n some detail just what changes in 6^^ 140 TABLE XIII. 57 •• a Fe MOSSBAUER PARAMETERS FOR COMPOUNDS OF THE TYPE R Sb(Fe(CO)„Cp), X n 2 4-n COMPOUND 5b (mm/sec) AEgC (mm/sec.) (mm/sec.) Cl2Sb(Fe(CO)2Cp)2J JCr(SCN)4(NH3)2J ~Cl2Sb (Fe (C0)2Cp) 2 J J P F 6 J |cl 2Sb(Fe(CO) 2Cp ) 2 J ^ Sb4Cl14J [Br2Sb (Fe (CO) 2Cp) 2J [pF6*j (CF3)2Sb(Fe(CO)2Cp)2J jPFgj ClSb(Fe(CO)2Cp)3J peCl^j [BrSb(Fe(CO)2Cp)3J JPF^ j JlSb(Fe(CO)2Cp)3JjPF6J [lSb(Fe(CO) 2Cp) 3j^I 3J [phSb(Fe(CO)2Cp)^| j~PF^  Ph2Sb(Fe(CO)2Cp)2J [PF^ Ph3SbFe(CO)2CpJ JPF^J JBu3SbFe(CO)2Cp| JPF 6 J 0.40 1.83 0.26 0.40 1.86 0.25 0.40 1.81 0.26 0.40 1.83 0.26 0.42 1.80 0.23 0.39 1.73 0.25 0.38 1.72 0.26 0.39 1.71 0.26 0.40 1.74 0.27 0.41 +1.736 0.26 0.39 1.74 0.26 0.41 1.86 0.23 0.38 1.87 0.24 c d A l l measurements on neat solids with absorbers at 80°K and "^Co(Cu) source at room temperature. Isomer shift relative to sodium nitroprusside; estimated error ±0.01 mm/sec.. Quadrupole s p l i t t i n g ; estimated error ±0.01 to ±0.02 mm/sec. Fu l l width at half-maximum; average of the two resonance lines. The sign of e^qQ was determined with both source and absorber at 4.2°K, and the absorber in a longitudinal magnetic f i e l d of 30 kG (Figure 26). TABLE XIV. Fe MtfSSBAUER PARAMETERS FOR Fe(CO)„Cp GROUPS BONDED TO TIN AND ANTIMONY COMPOUND 6(mm/sec.) (mm/sec.) REF. COMPOUND 6(mm/sec.) AE„ (mm/sec.) REF. Cp(CO) 2FeSnCl 3 0.41 +1.86 a —<J 0.40 1.86 b 0.39 1.84 c CCp(CO)2Fe)2SnCl2 0.39 +1.66 a (Cp(CO) 2Fe) 2SbCl 2 + 0.40 1.81-1.86 d 0.36 1.68 b (Cp(CO) 2Fe) 2Sn(NCS) 2 0.39 +1.69 a Cp(CO)2FeSnBu3 0.38 +1.75 a Cp(CO) 2FeSbBu 3 + 0.38 1.87 d Cp(CO)2FeSnMe3 0.36 1.75 c Cp(CO)2FeSnPh3 0.37 1.83 b Cp(CO) 2FeSbPh 3 + 0.41 1.86 d 0.35 1,82 c Ref. 106. Isomer shift values from this reference have been converted to the sodium nitroprusside scale, by the addition of 0.27 mm/sec. b Ref. 104. C Ref. 102. ^ ^ This work. 142 between isoelectronic t i n and antimony complexes would be expected depending upon whether or not there i s significant Fe-Sb TT-bonding. In these compounds antimony may be regarded either as forming a normal (shared) electron pair cr-bond to a neutral Fe(CO)2Cp group, in which case one i s essentially comparing R.jSb' + with R^Sn* (say), or as forming a dative bond to a positively charged Fe(C0) 2Cp + species, where now the comparison would be R^Sb: with R^Sn:-. In either case antimony should be a poorer CT-donor than t i n or conversely a better 0-acceptor. This means that i f the Fe-Sb bond is essentially pure a in character, the augmentation of 4s-electron density at iron w i l l be smaller i n the antimony derivatives, which should consequently show higher ^ 7Fe isomer shifts than the corresponding t i n complexes. Any Fe-Sb Tr-bonding, assuming this to be between f i l l e d 3d orbitals on iron and vacant antimony 5d orbitals, would affect 6 in the opposite direction since the decrease in d-electron density at iron would decrease 5 by deshielding. The fact that there is only a very small increase ( i f any) i n 6^ e in the antimony complexes appears to argue in favour of some 7T-interactions in these compounds. The "*7Fe quadrupole sp l i t t i n g s , (^EQ)pe> show two trends which we feel are interrelated. F i r s t l y , in the compounds R^Sb (Fe (CO) 2Cp) ^ _ n + (R = Ph, n = 1, 2, 3; R = Cl, Br, n = 1, 2), ( A E Q ) P E increases as n increases, and hence as the a-donor ab i l i t y of the RnSb moiety decreases. Similarly, in the compounds R 2Sb(Fe(CO) 2Cp) 2 + (^ Eg) F e increases in the order Ph < CF^ < Br, Cl, paralleling the increasing electronegativity of R. In both cases a decrease in Sb Fe cr-donation and thus in the charge 143 density at Fe along the Fe-Sb bond direction (Z axis) is accompanied by an increase in (AE ) . Although the nature of the iron bonding Q r e orbitals in these compounds is poorly understood, these results clearly imply a deficiency of electron density at iron along the Z axis and that V^^ should be positive (oblate charge distribution). The "^Fe Mossbauer spectrum of PhSb(Fe(CO)2Cp) .3 ( P Fg) w a s measured in an applied longitud-inal magnetic field of 30 kG, and V^ ^ was found to be positive as expected (Figure 26) . It might also be noted that ( A E ^ ) ^ is essentially constant for the compounds XSb(Fe(CO)2Cp)3+ (X = Ph, I, Br, Cl). In these cases any differences due to alterations in a-donor and TT-acceptor properties of the XSb group are distributed amongst three iron atoms and effectively masked. That the ^Fe quadrupole interaction in these compounds, particularly those containing only one or two Fe(C0)2Cp moieties, is reasonably sensitive to details of the electron distribution about iron is apparent from the data in Table XIII. This, together with the fact that there are no large changes in (^Eq)pe between corresponding tin and antimony compounds (Table XIV). implies that the electron distribution in the Fe-Sb bond is quite similar to that in the Fe-Sn bond. What small differences there are in ( ^ E ^ ) ^ are consistent with the expected changes This is exactly the inverse effect to the decrease in (AE ) ? e seen m c above in L LFe„(C0),- and L LFe_(C0), complexes as the a-donor power A B Fe •*• Fe increased. In both cases V„„ is predicted to be positive. 144a FIGURE 26. 57 Fe Mossbauer Spectrum of PhSb(Cp(CO)2Fe)3+PFg" in an Applied Longitudinal Magnetic Field of 30kG0 2 The Sign of e qQ is Clearly Positive. Fe Resonance of PhSb (Cp (CO)2 Fe )| P f | " in a parallel magnetic field of 30 kG 1.001 •2 98-CO 'E CO § -96-.94^  o CD -2.0 o0 ef t CD O O 66 o CP o o o °>o aP o o o o a ° - ° < t f c o o o o o Oo0> o o o o o •1.0 T 0.0 o °o o o o oo -o<fcocP o o 1.0 2.0 Velocity (mm/sec) 3.0 FIGURE 26. 145 i n a-donor and ir-acceptor properties between Sn and Sb. This r e s u l t r e i n f o r c e s the conclusions drawn above from "^Fe, 1 1 9 S n and 1 2 1 S b isomer s h i f t data, namely that antimony and t i n are nearly i s o e l e c t r o n i c i n these compounds. Thus i n the i o n i c antimony d e r i v a t i v e s most of the p o s i t i v e charge must reside on antimony rather than being d e l o c a l i z e d onto the ligands. This agrees with a s i m i l a r conclusion reached on the 118 basis of c r y s t a l s tructure data (C) The C o r r e l a t i o n of "*7Fe Mossbauer Parameters with the Carbonyl Stretching Frequencies i n the I.R. As we have discussed above, i n our 57 present s e r i e s of compounds containing iron-antimony bonds the Fe Mossbauer parameters have been found to be quite i n s e n s i t i v e to the nature of the other groups bonded to antimony (Table X I I I ) . To explain t h i s lack of s e n s i t i v i t y we have invoked a- and TT-bonding e f f e c t s since 6p e i s decreased by increasing donor strength (L ->• M) and increased by decreases i n TT back-bonding (M -> L) . The Q.S. f o r i r o n i s also a f f e c t e d by these fa c t o r s and we expect the Q.S. to become more negative with increasing a-donor power and more p o s i t i v e with increasing TT back-bonding. The e f f e c t s of these phenomena on V may be appreciated i f i t i s noted that has been found to be p o s i t i v e i n a l l the d e r i v a t i v e s , X Fe^O^Cp, i n which i t has been measured 1*^. I t i s obvious from the range of measured values of the Q.S. i n such compounds (1.80 ± 0.20 mm/sec.) 1^ 2 106,131 133 7 27 that there i s a large c o n t r i b u t i o n from an u n f i l l e d d - o r b i t a l ' and that the valence contribution (q__._) i s only 10-20% of the t o t a l e.f.g.. Since VAJ-i H l i e s i n the range 0 < n < 1 and can give contributions to (AE^)^ of up to TABLE XV. Fe MOSSBAUER AND v „ n PARAMETERS OF SOME X.MFe(CO)_Cp DERIVATIVES. COMPOUND V 1 ( c m 1 ) v 2(cm~ 1) SOLVENT REF. 6 Fe(mm/sec.) Q.S.(mm/sec.) REF. Ph 3PFe(CO) 2Cp +PF 6" 2070 2030 NUJOL 119 0.32a 1.92 132 Ph 3AsFe(CO) 2Cp +PF 6" 2062 2017 NUJOL 119 - - -Ph3SbFe(CO)2Cp+PF6"" 2050 2005 NUJOL 119 0.41 1.86 THIS WORK Bu3SnFe(CO)2Cp 1972 1923 CHC13 106 0.38a 1.75 106 Cl 3SnFe(CO) 2Cp 2048.0 2008.3 CHC13 137 0.40 1.85 b Me3SnFe(CO)2Cp 1985.4 1930.1 CHC13 137 0.36 1.75 102 Ph3SnFe(CO)2Cp 1995.3 1943.5 CHC13 137 0.36 1.82 c ClFe(CO)2Cp 2057.6 2012.0 CHC13 137 0.50d 1.88 131 BrFe(CO)2Cp 2052.9 2007.3 CHC13 137 0.51d 1.87 131 IFe(CO)2Cp 2043.8 2000.0 CHC13 137 0.49d 1.83 131 Converted to sodium nitroprusside by addition of 0.27 mm/sec. b Average value of refs. 102, 104, 106 (Table XIV). C Average value of refs. 102, 104 (Table XIV). d Converted to sodium nitroprusside by addition of 0.11 mm/sec. TABLE XVI. Fe MOSSBAUER PARAMETERS AND PARAMETERS OF SOME X 2M(Fe(CO) 2Cp) 2 DERIVATIVES. COMPOUND V ]_(cm 1) V 2(cm" 1) V 3(cm~ 1) v 4(cm" "1) SOLVENT REF. 5Je(mm/sec,) Q.S.(mm/sec. ) REF. Cl 2Sb(Fe(CO) 2Cp) 2PF 6" 2065 2053 2020 CH2C12 41 0.40 1.86 THIS WORK Br 2Sb(Fe(CO) 2Cp) 2PF 6" 2062 2048 2018 CH2C12 42 0.40 1.83 THIS WORK (CF 3) 2Sb(Fe(CO) 2Cp)2?F~ 2063 2050 2021 CH2C12 42 0.42 1.80 THIS WORK Ph 2Sb(Fe(CO) 2Cp) 2PF 6" 2045 2027 1994 1985 CH2C12 42 0.39 1.74 THIS WORK Cl 2As (Fe(CO) 2Cp) 2FeCl 4~ 2071 2057 2030 CH2C12 41 0.38 1.79 THIS WORK (NCS) 2Sn(Fe(CO) 2Cp) 2 2022 2007 1983 1970 CHC13 106 0.39 1.69 106 Cl 2Sn(Fe(C0) 2Cp) 2 2026 1999 1972 1963 CHC13 106 0.38 1.67 a Ph 2Sn(Fe(CO) 2Cp) 2 1998 1980 1947 1933 Cyclohexane 142 Cl 2Ge(Fe(CO) 2Cp) 2 2036 2010 1985 1962 Cyclohexane 141 0.36 1.66 104 a Average value ref. 104 and 106 (Table XIV). 148 15% of i t s total value (eqn. 9 with n, = 1) then small variations in ( A E Q ) ^ could equally well arise from changes in n as they could from changes in q__._. As well, there could be contributions to ( A E ) from changes in the geometry about iron. These facts mean that the Q.S. of Fe in these compounds is not a good diagnostic tool for separating possible a- and TT-bonding effects. 133 However, as reported by Burlitch and Ferrari , the combination of Mossbauer spectral parameters with data on carbonyl stretching frequen-cies is often informative. Similarly, both Mossbauer and I.R. techniques have been employed by Dessy and co-workers7^ in their study of olefin-Fe(CO) 4 complexes. In the analysis of carbonyl stretching frequencies, models in 13^ 133 136 which TT-bonding and models in which a-bonding ' are considered to predominate have both been employed. However, now, i t i s generally consid-70 137-139 ered that both a- and TT-bonding effects are important ' , although in certain cases either a or TT effects may be the predominant factor i n determining CO stretching frequencies. In fact, our present Mossbauer data and the reported CO stretching frequencies used in concert are incompatible with a a-only or a TT-only model. In the 0 + TT model, in every case where the electron density on 138 the metal is increased, V should decrease . For the TT contribution, the interaction is not a simple one, since the effect of TT-bonding on V is normally considered to be very directional and quite different i f the TT-bonding substituent is cis or trans to the carbonyl group. However, in 149 our present complexes, only cis substitution can take place and so there Is only the one type of interaction to consider. So, increases in a-donor strength lead to decreases in v r 1^8'''"4^ while decreases in TT back-bonding also lead to decreases in V ^ Q . From the foregoing discussion, we may recognize two extreme cases i f a ligand L ' is substituted for a ligand L in these complexes assuming the O and TT effects to be comparable in magnitude. Case A . If L ' is a better a-donor than L then 6p £ decreases and V ^ Q decreases. If L 1 is a better TT-acceptor than L then 6„ decreases and Fe V ^ Q will increase. The net effect of both these interactions is a relatively large decrease in the I.S. but l i t t l e i f any change in v r n . In the converse situation, where L ' is both a worse a-donor and TT-acceptor than L, the I.S. will increase but will hardly change. Case B. If L' is a better CT-donor than L then 6_ decreases and V„ n Fe CO decreases. If L' is a worse TT-acceptor than L then 5„ increases and v„„ Fe CO decreases. The net effect will be l i t t l e i f any change in the I.S. but a relatively large decrease in V ^ Q * In the converse situation, where the a-donor power is worse and the TT-acceptor power is greater then the I.S. will hardly change but V will increase. For most cases, either a or TT effects will dominate so both Mossbauer and I.R. parameters will change. Nevertheless, i f the two tech-niques are employed in concert the qualitative picture which emerges is' very satisfying, as illustrated below. 150 F i r s t , there are some experimental problems to be discussed. For example, the values of V which we are employing normally have been measured in solution at room temperature whilst the Mossbauer parameters have been measured in the solid state at 80°K. Nevertheless, since the effects which we are looking for are relatively large changes in the Mossbauer parameters (the change must be more than 0.02 mm/sec. to be significant) or in V we may have some confidence that things like solvent 138 —1 137 effects on V (shifts of ± 10 cm are not uncommon ) w i l l not affect our f i n a l conclusions. As well, compounds of the type R^MFe^O^Cp (M = P, As, Sb, Sn, etc.) normally have two I.R. frequencies in the carbonyl region (v^ and v^) while those of the type R2M(Fe(C0)2Cp)2 have either three I.R. frequencies (v^, , and v^) or four I.R. frequencies (^> v2> v^, and V^) in the carbonyl region. The appearance of three or more bands in the latter compounds arises because rotation of one Fe^O^Cp group about the M-Fe bond can give rise to three distinct structures with symmetries C^, C g, or C^; these correspond to the relative positions adopted by the 141 Fe(C0)2Cp moieties on rotation . Group theoretical treatments show that for symmetry there are three bands (A, B^, and B 2 modes), for C g symmetry there are four bands (2A' and 2A"), and for symmetry there are also four 141 bands (4A) . Evidence from band intensities shows the organotin derivatives f a l l into the class with C symmetry, while the halogen derivatives of s germanium and t i n have symmetry. It is apparent from the data that the halogen derivatives of antimony probably have C 2 v symmetry while + 42 Pl^SbtFeCCO^Cp^ probably has C g symmetry like the corresponding t i n derivatives. 151 As inspection of Table XV shows, there i s a wide range of V ^ Q as well as a f a i r l y large range in 6 e^ for the R^MFeCCO^Cp compounds. We shall pick some typical comparisons and discuss them in detail. For 57 .. + example, comparing the Fe Mossbauer parameters and of Ph^SbFeCCO^Cp cation with those of Ph^SnFe(CO)2Cp we see there are large changes in the spectral parameters. Both the decrease in \) and i n the I.S. for the Ph.jSnFe(C0)2Cp derivative may be explained by assuming that Ph^Sn in a better a-donor than Ph^ Sb"*". This i s exactly the conclusion which was 121 119 reached on comparison of the Sb and Sn Mossbauer parameters above. Similarly, comparing the parameters Ph^SbFe^O^Cp* and XFe(C0)2Cp (X = halogen) derivatives we see that hardly changes yet there is a large increase in I.S. for the halogens. Since the halogens should be worse TT-acceptors than Ph^ Sb"*" these results indicate this is the converse of case A above and so the halogens must be somewhat worse a-donors than Ph.jSb+. Comparison of the data for Ph.jPFe (CO) 2Cp + with that of Ph^SbFe^O^Cp* shows only small changes in V.^Q but a large decrease in I.S. for the Ph^PFe^O^Cp* derivative. This i s an example of case A above and shows Ph^P i s both a better Tr-acceptor and a better a-donor than Ph.jSb. Similarly, comparing the spectral parameters of the FeCCO^Cp group i n Cl 3SnFe(CO) 2Cp and Ph 3SbFe(CO) 2Cp + shows there i s l i t t l e or no change in either V or 6 and this implies the Ph,Sb+ group and Cl„Sn CU Fe 5 J are very similar in their effects on the electron density at iron in these 152 two complexes. This also means that the a-donor power of the Ph^ Sb"*" group is roughly equivalent to that of Cl^Sn. The worse a-donor power of the halogens relative to Ph.jSb as observed above does not seem unreasonable in view of this situation. In general, both the "*7Fe Mossbauer parameters and CO stretching frequencies for the R2M(Fe(CO)2Cp)2 derivatives (Table XVI) would seem to be less sensitive to the nature of R and M. This is exactly the behaviour which would be predicted since the effects of ligand changes are being distributed between two Fe(CO)2 groups. However, detailed analysis of the data for the R2M(Fe(CO)2Cp)2 derivatives is precluded at this stage both by the lack of published information on the ^ 7Fe Mossbauer parameters of some of the more interesting compounds (Ph2Sn(Fe(CO)2Cp)2 for one) and by the fact that various compounds apparently belong to different , 141,142 symmetry classes . 153 BIBLIOGRAPHY (1) R.S. Drago, "Physical Methods in Inorganic Chemistry," Reinhold Publishing Corp., New York, N.Y., 1965. C2) R.L. Mossbauer in "The Mossbauer Effect," H. Frauenfelder, Ed., W.A. Benjamin, Inc., New York, N.Y., 1963, pp 127-129. (3) Ref. 1, p 199. (4) R. P. Feynman, R.B. Leighton, and M. Sands, "The Feynman Lectures on Physics," Vol. I l l , Addison-Wesley, Inc., Reading, Mass., 1965, p 9-14. (5) N.N. Greenwood and T.C. Gibb, "Mossbauer Spectroscopy," Chapman and Hall, Ltd., London, 1971, p 6. (6) G.K. Wertheim, "Mossbauer Effect: Principles and Applications," Academic Press Inc., New York, N.Y., 1965, p 94. (7) G.M. Bancroft and R.H. Piatt, Adv. Inorg. Chem. Radiochem., 15, 59 (1972). (8) G.G. Long, J.G. Stevens, R.J. Tullbane, and L.H. Bowen, J. Amer.  Chem. Soc, £2, 4230 (1970). (9) Ref. 5, p 15. (10) Ref. 5, p 10. (11) Ref. 6, p 52. (12) Ref. 6, p 55. (13) Ref. 5, p 51. (14) Ref. 5, p 95. (15) R.L. Collins and J.C. Travis in "Mossbauer Effect Methodology," Vol. 3, I.J. Gruverman, Ed., Plenum Press, New York, N.Y., 1967. 154 (16) V.I. Gol'danskii and E.F. Makarov in "Chemical Applications of Mossbauer Spectroscopy," V.I. Gol'danskii and R.H. Herber, Eds., Academic Press, New York, N.Y., 1968. (17) S.L. Ruby, G.M. Kalvius, G.B. Beard, and R.E. Snyder, Phys. Rev., 159, 239 (1967). (18) J.G. Stevens and S.L. Ruby, Phys. Lett. (A), 32, 91 (1970). (19) T.C. Gibb, R. Greatrex, and N.N. Greenwood, J.C.S. Dalton, 238 (1972). (20) B.D. Dunlap in "Mossbauer Effect Data Index Covering the Year 1970," J.G. Stevens and V.E. Stevens, Eds., IFI/Plenum Data Corp., New York, N.Y., 1972, pp 25-41. (21) R.V. Parish and R.H. Piatt, Inorg. Chim. Acta, _4, 65 (1970). (22) G.M. Bancroft, M.J. Mays, and B.E. Prater, J. Chem. Soc. (A), 956 (1970). (23) M.G. Clark, Mol. Phys., 20, 257 (1971). (24) J.R. Sams i n "MTP International Review of Science," Phys. Chem. Ser. 1, Vol. 4, CA. McDowell, Ed., Ch. 3, Butterworths, London, 1972. (25) M.G. Clark, A.G. Maddock, and R.H. Piatt, J.C.S. Dalton, 281 (1972). (26) R.A. Mazak and R.L. Collins, J. Chem. Phys., 51_, 3220 (1969). (27) M.G. Clark, W.R. Cullen, R.E.B. Garrod, A.G. Maddock, and J.R. Sams, Inorg. Chem., 12, 1045 (1973). (28) R.H. Herber i n "Mossbauer Effect Methodology," Vol. 6, I.J. Gruverman, Ed., Plenum Press, New York, N.Y., 1971. (29) R.L. Collins, J . Chem. Phys., 42, 1072 (1965). (30) Ref. 5, p 68. 155 (31) A. Messiah, "Quantum Mechanics," Vol. II, J.O. Potter, Transl., John Wiley and Sons, Inc., New York, N.Y., 1964, p 560. (32) S.L. Ruby in "Mossbauer Effect Methodology," Vol. 3, I.J. Gruverman, Ed., Plenum Press, New York, N.Y., 1967. (33) G. Lang, J. Chem. Soc. (A), 3245 (1971). (34) T.C. Gibb, J. Chem. Soc. (A), 2503 (1970). (35) Ref. 5, p 90. (36) J.C. Scott, unpublished results. (37) W.R. Cullen, D.A. Harbourne, B.V. Liengme, and J.R. Sams, Inorg. Chem., j5, 95 (1969). (38) R.T. Lassau, M.Sc. Thesis, University of British Columbia, 1972. (39) J.J. Spijkerman, D.K. Sneidiker, F.C. Ruegg, and J.R. Devoe, Nat. Bur. Stand. (U.S.), Spec. Publ. No. 260-13 (1967). (40) G. Lang, personal communication. This value may also be derived 2 using the relationship that g^ = — u^, where = 0.15463 nm (Ref. 20, p 70), 6^ = 5.05038 x 10~ 2 1 ergs/kG (R.B. Leighton, "Principles of Modern Physics," McGraw-Hill, Inc., New York, N.Y., 1959) and the conversion factor for ~*7Fe, 1 mm/sec. = 7.68 x 10 ^ ergs (assuming E ( 5 7Fe) = 14.41 keV from Ref. 20, p 70). Y (41) W.R. Cullen, D.J. Patmore, and J.R. Sams, Inorg. Chem., 12, 867 (1973). (42) W.R. Cullen, D.J. Patmore, J.R. Sams, and J.C. Scott, submitted for publication. (43) J.N.R. Ruddick, J.R. Sams, and J.C. Scott, to be published. (44) L.S. Chia, M.Sc. Thesis, University of British Columbia, 1971. 156 (45) L.S. Chia and W.R. Cullen, unpublished results. (46) L.S. Chia, W.R. Cullen, and D.A. Harbourne, Can. J. Chem.,50, 2182 (1972). (47) W.R. Cullen, Adv. Inorg. Chem. Radiochem., 15, 323 (1972). (48) W.R. Cullen, D.A. Harbourne, B.V. Liengme, and J.R. Sams, Inorg. Chem., 9_, 702 (1970). (49) CH. Wei and L.F. Dahl, J. Amer. Chem. Soc., 91, 1351 (1969). (50) L.F. Dahl and J.F. Blount, Inorg. Chem., 4_, 1373 (1965). (51) T.C. Gibb, R. Greatrex, and N.N. Greenwood, J. Chem. Soc. (A), 1663 (1967). (52) K. Farmery, M. Kilner, R. Greatrex, and N.N. Greenwood, J. Chem.  Soc. (A), 2339 (1969). (53) F.S. Stevens, Acta Cryst., 21, A154 (1966). See also footnote 55 in L.B. Handy, J.K. Ruff, and L.F. Dahl, J. Amer. Chem. Soc, 9_2, 7312 (1970). (54) L.S. Chia and W.R. Cullen, Can. J. Chem., 50, 1421 (1972). (55) W. Harrison and J. Trotter, J. Chem. Soc. (A), 1607 (1971). (56) F.W.B. Einstein and R.D.G. Jones, J.C.S. Dalton, 442 (1972). (57) W. Harrison and J. Trotter, J. Chem. Soc. (A), 1542 (1971). (58) F.W.B. Einstein and J. Trotter, J. Chem. Soc. (A), 824 (1967). (59) J.A. McGinnety and J.A. Ibers, Chem. Comm., 235 (1968). (60) F.W.B. Einstein and R.D.G. Jones, Inorg. Chem., 12, 255 (1973). (61) J.A. de Beer, R.J. Haines, R. Greatrex, and N.N. Greenwood, J. Chem. Soc. (A), 3271 (1971). (62) W.R. Cullen, D.A. Harbourne, B.V. Liengme, and J.R. Sams, 157 (63) R.L. Collins and R. Pe t t i t , J. Chem. Phys., 39, 3433 (1963). (64) G.F. Emerson, J.E. Mahler, R. Pe t t i t , and R.L. Collins, J. Amer. Chem. Soc., 86, 3590 (1964). (65) D.A. Brown, W.J. Chambers, and R.M. Rawlinson, J. Chem. Soc. (A) , 720 (1971). (66) L.S. Chia, W.R. Cullen, J.R. Sams, and J.E.H. Ward, Can. J. Chem., in press. (67) W.E. Carroll, F.A. Deeney, J.A. Delaney, and F.J. Lalor, J.C.S. Dalton, 718 (1973). (68) G.W. Parshall and F.N. Jones, J. Amer. Chem. Soc, 87, 5356 (1965). (69) F.A. Cotton, B.G. De Boer, and T.J. Marks, J. Amer. Chem. Soc, 93, 5069 (1971). (70) R.E. Dessy, J.C. Charkoudian, T.B. Abeles, and A.L. Reingold, J. Amer. Chem. Soc., 92, 3947 (1970). (71) S.M. Ohlberg, J. Amer. Chem. Soc. , 80, 811 (1959). (72) T.N. Polynova and M.A. Porai-Koshits, J. Struct. Chem., (USSR), (Engl. Transl.) ]_t 742 (1966). (73) A.F. Wells, Z. Kr i s t . , 99, 367 (1938). (74) K. Shen, W.E. McEwan, S.J. La Placa, W.C. Hamilton, and A.P. Wolf, J. Amer. Chem. Soc, 90, 1718 (1968). (75) A.L. Beauchamp, M.J. Bennett, and F.A. Cotton, J. Amer. Chem. S o c , 91, 297 (1969). (76) R.G. Goel, E. Masowsky, Jr., and CV. Senoff, Inorg. Chem., 10, 2572 (1971). 158 J.B. Orenberg, M.D. Morris, and T.V. Long, II, Inorg. Chem., 10, 933 (1971). G.G. Long, G.O. Doak, and L.D. Freeman, J. Amer. Chem. Soc., 86, 209 (1964). S.E. Gukasyan and V.S. Shpinel, Phys. Stat. Sol., 29, 49 (1968). G. M. Bancroft, K.D. Butler, and E.T. Libbey, J.C.S. Dalton, 2643 (1972). R. Goel, Can. J. Chem., 41, 4607 (1969). H. C. Clark and R.G. Goel, Inorg. Chem. , _5_, 998 (1966). D.L. Macalady, Ph.D. Dissertation, University of Wisconsin, Madison, Wis., 1969, as quoted in Ref. 8. T.B. B r i l l and G.G. Long, Inorg. Chem. , 9_, 1980 (1970). R.F. Schneider and J.V. Di Lorenzo, J. Chem. Phys., 47, 2343 (1967). Ref. 5, p 608. G. Ferguson, R.G. Goel, F.C. March, D.R. Ridley, and H.S. Prasad, Chem. Comm., 1547 (1971). F.W.B. Einstein and B.R. Penfold, J. Chem. Soc. (A), 3019 (1968). L.H. Bowen, J.G. Stevens, and G.G. Long, J. Chem. Phys., 51, 2010 (1969). J.G. Stevens and L.H. Bowen i n "Mossbauer Effect Methodology," Vol. I. J. Gruverman, Ed., Plenum Press, New York, N.Y., 1970. T.N. Polynova and M.A. Porai-Koshits, J. Struct Chem., (USSR), (Engl. Transl.) 2, 445 (1961). T.N. Polynova and M.A. Porai-Koshits, J. Struct. Chem., (USSR), (Engl. Transl.), j3, 112 (1967). 159 (93) J.N.R. Ruddick, private communication. (94) I.R. Beattie, F.C. Stokes, and L.E. Alexander, J.C.S. Dalton, 465 (1973). (95) R.G. Goel and D.R. Ridley, J. Organomet. Chem., 38, 83 (1972). (96) H. Schmidbauer, K. -H. Mitschke, J. Weidlein, and St. Craddock, Z. Anorg. Allg. Chem., 386, 139 (1971). (97) H.A. Meinema and J.G. Noltes, J. Organomet. Chem., 36, 313 (1972). (98) N.W. Alcock and R.E. Timms, J. Chem. Soc. (A), 1873 (1968). (99) R.G. Goel, J.N.R. Ruddick, and J.R. Sams, unpublished results. (100) B.F.E. Ford, B.V. Liengme, and J.R. Sams, J. Organomet. Chem., 19, 53 (1969). (101) R.V. Parish, Progr. Inorg. Chem., 15, 101 (1972). (102) W.R. Cullen, J.R. Sams, and J.A.J. Thompson, Inorg. Chem. , 10, 843 (1971). (103) D.E. Fenton and J.J. Zuckerman, J. Amer. Chem. Soc., 90, 6226 (1968); Inorg. Chem., 8, 1771 (1969). (104) R.H. Herber and Y. Goscinny, Inorg. Chem., _7> 1 2 9 3 (1968). (105) V.I. Gol'danskii, B.V. Borshagovskii, E.F. Makarov, R.A. Stukan, K.N. Anisimov, N.E. Kolobova, and V.V. Skripkin, Teor. Eksp. Khim., Akad. Nauk. Ukr. SSR, 3, 478 (1967). (106) B.A. Goodman, R. Greatrex, and N.N. Greenwood, J. Chem. Soc. (A), 1868 (1971). (107) S.R.A. Bird, J.D. Donaldson, A.F. LeC. Holding, B.J. Senior, and M.J. Tricker, J. Chem. Soc. (A), 1616 (1971). (108) B.V. Liengme, M.J. Newlands, and J.R. Sams, Inorg. Nucl. Chem. Lett., 2, 1223 (1971). 160 (109) G.M. Bancroft, K.D. Butler, and A.T. Rake, J. Organometal. Chem., 34, 137 (1972). (110) R.F. Bryan, P.T. Greene, G.A. Melson, and P.F. Stokely, Chem. Commun. , 722 (1969). (111) L.H. Bowen, P.E. Garrou, and G.G. Long, Inorg. Chem., 11, 182 (1972). (112) G.M. Bancroft, K.D. Butler, A.T. Rake, and B. Dale, J.C.S. Dalton, 2025 (1972). (113) B.V. Liengme, J.R. Sams, and J.C. Scott, Bull. Chem. Soc. Japan, 45, 2956 (1972). (114) J.E. O'Connor and E.R. Corey, Inorg. Chem., 6_, 968 (1967). (115) (a) R.F. Bryan, J. Chem Soc. (A), 192 (1967); (b) P.T. Greene and R.F. Bryan, i b i d . , 1969 (1970); (c) G.A. Melson, P.F. Stokely, and R.F. Bryan, i b i d . , 2247 (1970); (d) P.T. Greene and R.F. Bryan, i b i d . , 2261 (1970). (116) B.P. Bir'yukov, Yu. T. Struchkov, K.N. Anisimov, N.E. Kolobova, and V.V. Skripkin, Chem. Commun., 750 (1967); i b i d . , 1193 (1968). F.W.B. Einstein and R. Restivo, Inorg. Chim. Acta, _5, 501 (1971). (117) F.W.B. Einstein and R.D.G. Jones, Inorg. Chem., 12, 1690 (1973). (118) Trinh-Toan and L.F. Dahl, J. Amer. Chem. Soc, 93, 2654 (1971). (119) A. Davison, M.L.H. Green, and G. Wilkinson, J. Chem. Soc, 3172 (1961). (120) H.J. Plastas, J.M. Stewart, and S.O. Grim, Inorg. Chem. , 12, 265 (1973). (121) J.J. Legendre, C. Girard, and M. Huber, Bull. Soc. Chim. France, 1998 (1971). 161 (122) Y.K. Ho and J.J. Zuckerman, J. Organometal. Chem., 49, 1 (1973). (123) S.R.A. Bird, J.D. Donaldson, S.A. Keppie, and M.F. Lappert, J. Chem. Soc. (A), 1311 (1971). (124) G.M. Bancroft and K.D. Butler, J.C.S. Dalton, 1209 (1972). (125) S.R.A. Bird, J.D. Donaldson, A.F. LeC. Holding, B. Ratcliff, and S. Cenini, Inorg. Chim. Acta, j6 , 379 (1972). (126) S. Ichiba, M. Katada, and H. Negita, Bull. Chem. Soc. Japan, 45, 1679 (1972). (127) S.L. Ruby and G.K. Shenoy, Phys. Rev., 186, 326 (1969). (128) S.K. Ruby, H. Montgomery, and CW. Kimball, Phys. Rev. B, JL, 2948 (1970). (129) Ref. 5, p 374. (130) A number of tin derivatives of Fe^O^Cp have been reported in Ref. 105, where the ranges 0.30 - 0.37 mm/sec. (relative to sodium nitroprusside) for "*7Fe isomer shifts and 1.60 - 1.83 mm/sec. for "*7Fe quadrupole splittings are given. However, these authors do not quote values for individual compounds. (131) R.H. Herber, R.B. King, and G.K. Wertheim, Inorg. Chem., 3_, 101 (1964). (132) K. Burger, L. Korecz, P. Mag, U. Belluco, and L. Busetto, Inorg. Chim.  Acta, 5_, 362 (1971). (133) J.M. Burlitch and A. Ferrari, Inorg. Chem. , 9_y 563 (1970). (134) F.A. Cotton and CS. Kraihanzel, J. Amer. Chem. Soc, 84, 4432 (1962). (135) R.J. Angelici, J. Inorg. Nucl. Chem., 28, 2627 (1966). (136) R. Ugo, S. Cenini, and F. Bonati, Inorg. Chem. Acta, 1, 452 (1967). 162 (137) J. Dalton, I. Paul, and F.G.A. Stone, J. Chem. Soc. (A), 2744 (1969). (138) W.A.G. Graham, Inorg. Chem., _7» 315 (1968). (139) T.B. B r i l l , J. Organomet. Chem., 40, 373 (1972). (140) R.E. Dessy and L. Wieczorek, J. Amer. Chem. Soc, 91, 4963 (1969). (141) N. F l i t c r o f t , D.A. Harb ourne, I. Paul, P.M. Tucker, and F.G.A. Stone, J. Chem. Soc. (A), 1130 (1966). (142) K.N. Anisimov, B.V. Lokshin, N.E. Kolobova, and V.V. Skripkin, Iz. Akad. Nauk., SSSR, Ser. Khim., _5» 1024 (1968), (Engl. Transl.). (143) Nucl. Data Sheets. 6, 90 (1971). (144) Ref. 4, pp 18-1 - 18-3. 163 APPENDIX I 121 The analysis of the quadrupole-split Sb spectra in the case when the asymmetry parameter n is not zero is interesting since i t illustrates in detail the procedures which are used in the general case for pure Ml y-transitions between states with different nuclear spins. In particular, the cases of interest w i l l involve randomly oriented polycryst-alline absorbers and single-line ysources. The Hamiltonian for a nucleus with spin I and quadrupole moment Q subjected to a non-axially symmetric e.f.g. is HQ " 41(21-1) < 3 1Z - + 5 ( V + * 3 > using the same conventions as before. 121 For the ground state of Sb with spin I the Hamiltonian matrix may be written as = 5 / 2 , denoting A = (AI-2) <m» 5/ 2 |H |5/2m> = 10A 0 lOriA 0 0 0 0 -2A 0 i8nA 0 0 10nA 0 -8A 0 18nA 0 0 18nA 0 -8A 0 lOnA 0 0 18nA 0 -2A 0 0 0 0 10nA 0 10A 164 121 7 Similarly, for the excited state of Sb with spin / 2 e2qQ* denoting B = — — , where Q* is the quadrupole moment of the excited state, the Hamiltonian matrix may be written as (AI-3) <m,7/2|HQA|7/2m> = 21B • 0 21nB 0 0 0 0 0 0 3B 0 45n.B 0 0 0 0 21nB 0 -9B 0 60nB 0 0 0 0 45nB 0 -15B 0 60nB 0 0 0 0 60nB 0 -15B 0 45nB 0 0 0 0 60nB 0 -9B 0 21riB 0 0 0 0 45nB 0 3B 0 0 0 0 0 0 21nB 0 2 IB The energy eigenvalues and eigenvectors would be found by machine diagonalization of each of the above Hamiltonian matrices for 2 appropriate values of e qQ and n. The energy eigenvalues may be represented A l , A2, A3 for the ground state and BI, B2, B3, B4 for the excited state. The corresponding eigenvectors in terms of the basis kets, j l e mg> and |lg nig> for the excited and ground states respectively are of the form (AI-4) - —' -» — b l l b21 b l 2 b22 b 1 3 b23 b l 4 b24 b15 b25 b16 b26 b17 b27 bl8 _ — > b28 etc. for the excited state and 165 (AI-5) l l l *12 l13 *14 »15 he l21 l22 *23 l24 l25 l26 etc. for the ground state. To calculate the intensities i t is necessary to have some 121 knowledge of the Y-rradiation available. An unpolarized source of Sb 143 Y-rays since the Y _transition is pure Ml , w i l l emit equal numbers of l e f t - and right-hand circularly polarized Y - r a y s , denoted |l,-l>' and |l,l>' respectively. The number of longitudinally polarized Y-quanta i 144 |1,0>' should be zero . In general the Y -ray direction and the Z axis, defined by the principal components of the e.f.g., w i l l not be colinear. Rather the incoming y-ray w i l l be at an angle 6,<{> relative to that Z axis, This is equivalent to observing the Y -radiation at an angle Q,$ relative to i t s direction of propogation. The numbers of Y-quanta with polarizations |1,-1>,|1,0> and |l,l> available in the basis coordinate system are given by the sum over the amounts of Y _quanta with l e f t - and right-hand circular polarization emitted by the source (since.11,0>' = 0), i.e. |l,-l>'+ |l,l> where (AI-6) |l,l>' = ^(l+cose)e i*|l,l> - %(l-cose)e" 1 < P|l,-l> - ^ s i n 6 | l , -i<J>l J2 0> and (AI-7) |l,-l>» = -!s(l-cose)e i < f >|l,l> + Ji(l+cose)e _ : L*|l,-l> - ^psin9|l,0> 166 relative to the basis coordinate system1"*. These terms may be rewritten as (AI-8) |l,l>' = A + ( l ) |l,l> + A + ( - l ) | l , - l > + A +(O)|l,0> and (AI-9) |l,-l>' = A_( l ) | l , l > + A_ ( - l ) | l , - l > + A_(0)|l,0> re A + ( l ) = %(l+cos9)e 1*, etc. whe The excited state eigenvectors are written in terms of the basis vectors | 7 ^ 2 m e > » t n e ground state eigenvectors are written i n terms of the basis vectors |^ I2m >, and the absorbed photons |l,M> connect the S „ 31 two sets of basis vectors via the Clebsch-Gordan coupling coefficients of the type (AI-10) < 5/ 2 1 mg M|7/2me> where M = m -m takes on the values 1. 0. -1. e g * •' Thus the relative intensities of the |l,l>' transitions may be found from consideration of matrices of the form 167 f H CN co <r in vo r-» oo r H r H r H r H r H i - H r H r H , O L O | c\l 1 ." CO CN LO|CM 1 i ~ m CN CO|cN 1 i * C0|CN I , * CO CN COlcN 1 * LO|Csl •-) I CM —<|CM <|CM r-l|cM I CM CO I CM I CJ CJ CJ —<|cN CO | CM <|CM < | CM r-<|CM |-<|CM CJ CJ CJ CO I CM «» m|cM CO | CM CO I CM CO I CM r - l I CM CJ CJ lO|CN m|cN m|cN CO | CM CJ I 3 r - | c M A ITllcN C J + I * r - t 1 CO 4C r H cd * r H co * r H « CN •K r H 1 « J J 168 where (AI-12) C. (m ,m ) = < 5 / o 1 m (m -m ) | 7 / 9 m >AJ(m -m ) + g e' z g v e g' 1 2 e + v e g' = < 5/ 0 1 m M|7/0 m >A. (M) 2 g 1 I e + and A+(M) are the angular parts A + ( l ) , A +(0) and A + ( - l ) as defined i n (AI-8). Similar matrices arise for |l,-l>' quanta except C (m , m ) are 8 e replaced by C (m , m ). These have identical Clebsch-Gordan coefficients - g e but the A+(M) terms are replaced by the corresponding A_(M) terms (AI-9). At this stage, after the matrix multiplication i s carried out and the results squared for each of |l,-l>' and |l,l>' those two terms would be summed and the result would be the relative intensity for some value of 9,<i>. Since the results for a polycrystalline randomly oriented powdered absorber are desired, integrations over the values of 8 and $ are required. In fact i t i s only necessary to integrate over the angular parts such as |Aj_(M)|2 since none of the other coefficients has any angular dependence. Hence only integrals of the form i 2ir TT (AI-13) "4TT O O |A+(l)rsined9d< r, etc., need be considered. Closer examination of the integrals arising from|l,l>' quanta shows some interesting results. F i r s t l y , evaluating the integrals e x p l i c i t l y , 1 2TT TT (AI-14) Jv £ & |A+(l)|sineded<f> 169 2TT .TT "(1 + 2 cos6 + cos^e) sinGdedcj) and 4TT O o 4 1 6 1 2 7 r 7 , 2 CAI-15) j / / |A (0) I sinedBdcj) 1 2 7 r 1 1 1 2 = 4rr o J I ( s i n 9 ) s i n 6 d 9 d (t> JL 3 CAI-16) ^ { ^ |A+(l)r sinGdedcj) , 2TT TT = "4TT O £ 4^1 " 2 c o s 0 + cos 8) sinBdedc}) 1 6 There are a number of other integrals which represent the cross terms such as 2TT TT * (AI-17) I f f V1) V°) sinSdSdc}) 4TT o o 2TT TT JY _. A= "4TT o o 4 + c o s 9 ) e 1 sine}sin8d9dc|) = 0 170 2TT TT $S CAI-18) i / / A + ( 1 ) A + ( _ 1 ) s i n9 d9 d (l> 4TT O O 1 ^ 1 9 9-A a T« L L -T(l-cos Ze)e~ Z l ( Psineded(}> = 0 and (AI-19) 1 / / A (-1) A (0) sinBdedt}) 4TT o o . 2TT TT r — . , = t l T O O - • f ( 1 - C O S e ) s i n 9 e S i n e d 9 d ( r = 0 In fact, a l l the cross terms are zero since a l l the integrals involve terms like 2TT . , (AI-20) / e±ml% = 0 For |l,-l>' quanta the integrals give the same results, 1 i 2 2 1 2 namely for |A_(1)| and |A_(-1) | , and for |A_(0) | . This means that i t is not necessary to carry out the matrix multiplication e x p l i c i t l y for |l,-l>' quanta since the f i n a l results after performing the integrations w i l l be identical to the | l , l > ? case. Thus i t i s only necessary to multiply the results of the |l,l>' case by two (after squaring) to account for the contribution of the |l,-l>' case. (AI-21) E * * * * * l' a12' a13 , al4' a15 , a16j 6 C, (1) 2C.(0) O, (-1) 0 0 0 0 0 C 0 ( l ) 2Co(0) C 0(-l) 2 0 0 0 0 0 0 c 3 (D 0 0 0 2Co(0) C 0(-l) c 4 d ) 0 0 where (^(1) = <-| 1 -| 1 |-^- -^->, etc., are the Clebsch-Gordan coefficients from the terms C^C-j, J") etc., in AI-11. 0 0 0 2C.(0) C.(-l) 0 0 0 0 C 5(l) 2CC(0) Cc(-1) 0 0 0 0 0 5 s ' w5 c 6 (D 2c6co) c 6 ( - i y 11 12 13 14 15 16 17 172 So the transition between the states Al and BI which has relative energy BI - Al (ignoring the y-ray energy) has intensity equal to the sum of four terms (owing to the degeneracies involved) each of which is the sum of three terms like (AI-22) i | au (^(1) b u + a12 C 2 ( l ) b 1 2 + a13 C 3 ( l ) b 1 3 + etc. | (AI-23) +| | a ^ C 1(0)b 1 2 + a* 2 C 2 ( 0 ) b 1 3 + a* 3 C 3 ( 0 ) b l 4 + etc. | 2 (AI-24) +| | a n C^ -Db^  + a12 C 2 ( - l ) b 1 4 + a13 C 3 ( - l ) b 1 5 + etc. | . From consideration of the foregoing, solutions to other cases involving pure Ml transitions may be developed. Programs have been written for the solution of the general case, but i t is sometimes more convenient and more eff i c i e n t to develop subroutines for a specific case such as this one. 173 APPENDIX II The following discussion of magnetically perturbed Mossbauer spectra follows the treatment of Collins and Travis 1"'. The case of a nucleus with ground state with nuclear spin 1 3 I = 12 and an excited state with nuclear spin I = 12 subjected to both an internal non-zero e.f.g. and to an external applied f i e l d i s considered. To be more specific, the case of a diamagnetic, poly-crystalline absorber i s considered. The f i r s t step i s to solve the energy eigenvalue problem. For the ground state, since Q^r = 0 the solution i s t r i v i a l . The Hamiltonian operator for the ground state interaction i s (AII-1) H ^ = = -go$nH(sinecos<i)ix + sin6sinc}>Iy + cosQI z) The Hamiltonian matrix i s (AII-2) <!Sm,|HMAG|laiii> = h h o n -h -Jig 3 HcosG -hg 3 HsinGe" 1* o n -hg 3 HsinGe + i* hg 3 HcosG o n o n and the eigenvalues are (AII-3) E l = -hg 3 H o n ( A I I-4) E2 = +hg 3 H ^o n 174 with eigenvectors (AII-5) |B1> = -sin | e~U\h,h> + cos | \h,-h> (AII-6) |B2> = cos | \h,h> + sin J e+±<i>\h,~h> Similarly, the Hamiltonian operator for the excited state 3 with 1 = is (AII-7) H ^ = ^  {3I Z - 1(1+1) + £ (l+2 + I_ 2)> -g.3 H (sin6cosd)I + sinOsinctl + cos6l ). I n x y z 2 If we define A = and = -g,3 H XI 1 n then the excited state Hamiltonian matrix i s (AII-8) <|-|H M + Q||m > 3 1 1 2 2 2 2 "2 3 2 1 2 1 2 3 3A + Ja cos6 Ja sin9e - i* 3nA Ja sinSe 1* -3A + Ja cos9 a sinSe - 1* 3nA 3r)A a sinOe 1^ -3A - Ja cos8 Ja sinGe | 0 3nA Ja sinSe 1* 3A - Ja cos9 175 Solutions to this matrix eigenvalue problem may be found 2 for any case of e qQ, n and H by computer diagonalization of the matrix for appropriate values of 8 and <f). The energy eigenvalues may be represented as E^', E^', E^', E^' and the corresponding eigenvectors in terms of the basis ket vectors as (AII-9) a12 "*13 — —. a14 a21 a22 a23 a24 a31 a32 a33 a34 * _a42. 9 *A3__ > _a44_ Similarly, the energy eigenvectors for the ground state are designated (AII-10) 11 21 in terms of the basis kets. and 12 22 Now, the exact solutions for the energies of the transitions would involve integrations over a l l values of 8 and cf). It is convenient to replace these integrals by sums over a f i n i t e number of values of 8 and cf>. The real energies may be simulated to any degree of exactness required by inclusion of sufficient terms in these sums. If there are n increments in <j>, then for each of these increments, there w i l l be n increments i n cos8 in order to approximate the elemental area -d(cos8)dc)> over the unit sphere. In fact, since the e.f.g. axes have been chosen as the coordinate 176 axes and since the e.f.g. has mirror symmetry about i t s principal planes, only values of 8 and $ lying in one octant need be chosen in approximating 33 the solutions Having found the energy eigenvalues for each increment i n 8 and <j>, one must have the transition probabilities as well in order to simulate a spectrum. To do this, i t i s necessary to know the eigenvectors for both the ground and excited states in terms of the basis kets (i.e. AII-9 and AII-10), the Clebsch-Gordan coefficients connecting the various substates, and then sum over the relative amounts of y-quanta available with appropriate polarizations. Usually, there are two choices for the direction of application of the y-beam; either pa r a l l e l to or perpendicular to the direction of 15 33 3A application of the magnetic f i e l d , H_ ' ' Our experimental situation is the former so only the results for that case w i l l be considered. Since H and the Y-beam are pa r a l l e l , they w i l l be at an angle 8, 4> relative to the e.f.g. coordinate system. Thus the amounts of radiation available with right- and left-hand circular polarizations and longitudinal polarization (|l,l>, |l,-l>, |l,0> respectively) are given by (AII-11) |l,l>' = A + ( l ) |l,l> + A + ( - l ) |l,-l> + A +(0) |l,0> (AII-12) |l,-l>» = A_(l) |l,l> + A_(-l) |l,-l> + A_(0) |l,0> which are entirely equivalent to equations AI-8 and AI-9 since the Y-transitions for "*7Fe and 1 1 9 S n are also pure Ml. 177 Thus, at each increment of 9 and <J> the transition probability connecting any two of the energy states obtained on diagonalization of the Hamiltonian matrices, say the states El and E l ' with eigenvectors 11 J12 and l l l l12 l13 *14 respectively, is the square of a term for |l,l>' quanta like (AII-13) Br bi 2 J C +( 2,2) S - ^ ^ C+ (2» 2? 0 ,-1 1 -1-1N •1-3, c +( 2,2) c +( 2,2) c +( 2,2) where, l l l l12 *13 l14 (AII-14) C+(mg,me) = <- 1 mg(me-mg) \j m^A^-m^) = <| 1 mg M|| me> A+(M) 1 3 m , m are the values of I„ in the basis representation, <•=• m M m > are g e Z ' 2' g 12 e 31 the Clebsch-Gordan coefficients and A+(M) are the coefficients for relative amounts of radiation with appropriate polarizations originating from source quanta of the type |l,l>' (eqn. AII-11). The square of the above term (AII-13) for |l,l>' quanta i s added to the square of a term for |l,-l>' quanta (AII-15) b l l ' b12 c-<2»2* ^ ^ ' ^ Q - ( i , 2 ) 0 C_( "J*"?) C_( 2>T) '-'_( ~2t~y) "11 l12 '13 414 178 where (AII-16) C (m ,m ) = <^ 1 m M|J- m >A (M) - g e 2 g ' 2 e -and the terms are as in (AII-14) except A (M) replaces A+(M) since they are the appropriate coefficients for the relative amounts of radiation with appropriate polarizations originating from the source quanta |l,-l>' (AII-12). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060176/manifest

Comment

Related Items