Polyhedron Vol. IO, No. 7, pp. 703-709, I991 Printed in Great Britain

0277V5387/91 $3.00+.00 0 1991 Pergamon Press plc

VOLUMES OF ACTIVATION FOR DISSOCIATION OF THE CATIONS OF [TRIS-2,2’-BIPYRIDYL]IRON(II), [TRIS-l,lO- PHENANTHROLINE]IRON(I) AND OF OTHER DIIMINE

IRON COMPLEXES IN AQUEOUS SOLUTION

JOHN BURGESS and SASKIA A. GALEMAt

Department of Chemistry, University of Leicester, Leicester LEl 7RH, U.K.

and

COLIN D. HUBBARD1

University of New Hampshire, Durham, NH 03824-3598, U.S.A.

(Received 1 October 1990 ; accepted 4 December 1990)

Abstract-The kinetics of dissociation of several iron(I1) diimine complexes by aqueous hydroxide ions have been studied spectrophotometrically at atmospheric pressure and elevated pressures, at 298.2 K. Values of AV* are between + 10 and + 16 cm3 mol- ‘, and are interpreted as arising from a significant loss of electrostricted water from the hydroxide ion as it reacts with the iron(I1) complex associatively. Dissociation of the [Fe(SNO,phen),]*+ and [Fe(5Brphen)3]2+ complexes in the presence of aqueous EDTA is reported. For both, AV* is +22 cm3 mol- ‘. These values can be understood by considering the reaction as a process comparable to aquation in acidic medium, a dissociative process. Activation enthalpy data support this idea.

Markedly positive activation volumes (A V*) have been found for hydroxide and/or cyanide attack at low-spin iron(H) tris-diimine cations, [Fe(bpy),12+ (bpy = 2.2’-bipyridyl), [Fe(phen),12+ (phen = l,lO- phenanthroline)’ and [Fe(4Mephen),12+ 2 in aque- ous solutions. For a bimolecular reaction a value of about --‘lo cm3 mol- ’ is expected for AV*.3 For the former two ions, Stranks and co-workers, ’ although recognizing the possibility of a dominant role for desolvation of the attacking ion on forming the transition state, argued that the values of + 19 to + 22 cm3 mall ’ pointed to a dissociative mode of activation. This interpretation was also proposed for cyanide ion displacement of ligand from [Fe(CN)5L]3--,4*5 where L is 3,5_dimethylpyridine, imidazole or pyrazine, and was based upon a com- bination of analysis of atmospheric pressure kin- etics results, and volume of activation data (AV* is

t Permanent address : Department of Organic Chem-

istry, University of Groningen, Groningen, The Nether- lands.

1 Author to whom correspondence should be addressed.

z + 20 cm3 mall ’ and independent of departing ligand). The dissociative mechanism has been re- iterated recently for base hydrolysis of several iron(H) complexes. 6 However, various features of some of these reactions are not yet resolved unam- biguously and require further study and discussion.

On the basis of thorough kinetic studies of nucleo- philic dissociation of [Fe(phen)3]2+, Margerum and Morgenthaler7 concluded that the mechanism was associative. Except at very high nucleophile concentrations the rate law reduces to a two-term expression ; one is for the aquation step and the second contains a first-order term in, for example, hydroxide ion concentration. Under typical reaction conditions, using a large excess of hydroxide ion over metal complex ion, the rate is dominated by the latter term, and thus the rate-determining step is second-order, and considered to be bimolecular and the mechanism associative. Dissociation of the complex ion can also occur in the presence of an acid or other reagents, for example, EDTA.8*9 In these cases the reaction rate is independent of the con- centration of the dissociating agent, and a ligand

703

704 J. BURGESS et al.

molecule which has begun to separate from the ion is prevented from resuming coordination by being scavenged rapidly by a proton, in the presence of acid, or when EDTA is present the iron is scavenged. Dissociation by aquation is considered to be a dissociative mechanism.

Because of our interest in these reactions we have also proceeded to study the kinetics in selected cases, over a range of pressure, with the objective of gaining additional insight into reaction mechanism. Aquation in the presence of sulphuric acid of

]Fe(phen) 31 2+ and the 5-nitro and 4,7-dimethyl ana- logues yielded, at 308.2 K, AV* values of + 15.4, + 17.9 and + 11.6 cm3 mall ‘, respectively. lo These results were rationalized in terms of a dissociative activation mode which consists of a two-ended partial dissociation of a phen bidentate ring, with different degrees of bond extension explaining the variation in AV* values. The D mechanism for cyanide replacement of ligand in [Fe(CN),L13- is predicated on the virtually constant degree of stretching of the Fe-N bond which gives rise to an invariant A V *. 5 The hydrophilic cyanide ion has an unchanged state of solvation in the rate-deter- mining step and the intrinsic component of AV*

completely dominates. In considering bimolecular nucleophilic dis-

sociation, the contribution of solvation effects to AV* must be included and this will be especially significant for charged nucleophiles. This was illuminated by high pressure studies” on the reac-

tions between [Fe(ppsa)3]4P and OH- (ppsa = l), and between [Mo(CO),(bpy)] and CN-,3 the latter in dimethylsulphoxide. For the former case, AV* is approximately zero, and A V* is - 9 cm3 mol- ’ for the reaction of the molybdenum complex. The latter value was explained as a consequence of an associ- ative process in which there is little or no desolv- ation contribution because the cyanide ion is very poorly solvated in DMSO. The reaction of the tris- iron(I1) complex of 1, which has a minus four charge, with OH- is characterized overall, by a modest degree of desolvation as the transition state is reached, since the intrinsic volume change for the associative process is negative. Dissociation of [Fe(4Mephen)3]2+ by aqueous cyanide ions yields a AV* value of + 10 cm3 mall ‘. These early exam- ples indicate desolvation contributions ranging from zero to about + 20 cm3 mall ‘.

We also measured AV* for attack by hydroxide ion for several closely related iron(I1) complex cations, [Fe(gmi),j2+,12 [Fe(Me2bsb),12+ I3 and [Fe(hxsb)12+ I3 (gmi = 2, Me,bsb = 3, hxsb = 4). The values of AV* were +16, +ll and +13 cm3 mall ‘, respectively, decidedly positive, but lower than those determined for nucleophilic dissociation earlier. Because of this we measured activation vol- umes for hydroxide attack at two closely related Schiff base complexes, the iron(I1) complexes of 3Mebsb (5) and 4MeObsb (6). Again values in the range + 10 to + 15 cm3 mol- I were obtained (vi& infra) .

Volumes of activation of diimine iron(I1) complexes 705

A dissociative mechanism for a given complex would, in principle, have a constant AV*, inde- pendent of the subsequent attack of the nucleophile. Since the rate-limiting step involves only length- ening of a ligand-metal bond, nucleophile solvation change is not relevant. For different complexes reacting with the same nucleophile, A V* could vary, reflecting the difference in partial molar volume of the departing ligand and different degrees of stretch- ing in the transition state. For a bimolecular reac- tion a negative value of AV* is expected in the absence of solvation. 3 However, significant changes in the solvation of both the metal complex ion and the nucleophile can occur for an associatively formed activated complex. For a small anionic nucleo- phile, loss of electrostricted solvent is larger than for a positively-charged complex ion (of much lower charge density). Added co-solvent will, it may be predicted, change solvation of both species giving rise to changes in AV* for a particular reactant pair. Such changes have been observed ’ 3-‘5 and explanations sought in terms of solvation variation both of the nucleophile, and as a consequence of the hydrophobic/hydrophilic complex ion being compatible or not, with the co-solvent. The focus in this paper is on aqueous solution, but these earlier results13-’ 5 should be borne in mind as they provide a framework of explanation of AV* variation. Aquation reactions for similar systems, vide supra, yielded variations in A V*. Consequently, variations or constancy of AI’*, alone, are of no particular assistance necessarily in an attempt to assign a mechanism unambiguously.

However, because values of A V* established for nucleophilic dissociation of several iron(I1) diimine complexes were in the range + 10 to + 16 cm3 mall ‘, it seemed incumbent on us to remeasure AV* for hydroxide attack at [Fe(phen)3]2+ and

[Wbpy)312+ as one step in an attempt to clarify the situation. In addition, because of our general interest in dissociation and other properties of iron(I1) diimine complexes, we have examined the kinetics of dissociation of two substituted tris- phen complexes, [Fe(SBrphen),]*+ and [Fe(SNO, pheN312+, in aqueous EDTA solution at elevated pressures. Thus we can compare Alf* values with those obtained for aquation in acid medium.

Materials

EXPERIMENTAL

Solutions of [Fe(bpy),]*+ and [Fe(phen)3]2f were prepared by dissolving appropriate amounts of their perchlorate salts in water. The solutions of the [Fe(5Brphen),12+ and [Fe(SNO,phen),]*+ cations were obtained by adding a small excess of ligand to

an aqueous solution of iron(I1) ammonium sulphate. Concentrations of these solutions were checked spectrophotometrically via published molar extinction coefficients : [Fe(bpy),12+ has E (522 nm) = 8700 mall’ dm3 cm-‘,‘6 [Fe(phen)3]2+ has E (510 nm) = 11,000 mol-’ dm3 cm-‘,‘7~‘8 [Fe(5N02phen)3]2+ has E (512 nm) = 12,100 mol-’ dm3 cm-“’ and [Fe(SBrphen),]*+ has E (510 nm) = 11,500 mol- ’ dm3 cm-‘.‘9 Iron(I1) com- plexes of the Schiff base ligands, 3Mebsb (5) and 4MeObsb (6), were prepared according to published procedure *’ in the form of their chloride salts, which were subsequently dissolved in water to give solutions of appropriate concentration. Solutions of iron(I1) cations were made up immediately prior to use in kinetic experiments. Sodium hydroxide, sodium chloride, potassium hydroxide, potassium chloride and Na,EDTA were Analar materials.

Methods

The experimental techniques and apparatus used were those described in detail in earlier pub- lications. ’ 3*2 ’

RESULTS

All kinetic runs were carried out at 298.2 K in the presence of a large excess of hydroxide ion, or of EDTA, and all followed first-order kinetics for at least three half lives. The experimental results are presented in the form of logarithms of ratios of rate constants under pressure to those for an aliquot of the same reactions at atmospheric pressure in Table 1. For each complex, plots of these ratios against pressure were linear within experimental uncer- tainty. Examples are shown for OH- +[Fe

(phen) 31* + and EDTA + [Fe(5N02phen),] 2+ in Fig. 1 and OH- +[Fe(bpy),]*+ and EDTA+[Fe

0m5-

Fig. 1. Plots of -log (k,,/k,) against pressure for dis- sociation of [Fe(phen),]‘+ in 0.0050 mol dmm3 NaOH (a), and for dissociation of [Fe(SNO,phen),]‘+ in the

presence of 0.010 mol dm-’ EDTA (w), at 298.2 K.

p / kbor

706 J. BURGESS et al.

Table 1. Dissociation of iron(II)-diimine complexes at 298.2 K ;” -log (k,,/kO) at specified pressures in aqueous solution

P WarI Complex 0.17 0.34 0.51 0.68 0.85 1.01 1.18 1.35

[Fe(bpy)3]2f ‘SC 0.080 0.149 0.157 0.186 0.245 0.305 [Fe(phen),12+ ‘,’ 0.088 0.149 0.173 0.238 0.284 0.364 [Fe(3Mebsb),]2+b,‘J 0.110 0.178 0.278 [Fe(4MeObsb)#+ h,eJ 0.0825 0.159 0.173 0.206 [Fe(5Brphen),]2fg 0.070 0.146 0.229 0.288 0.313 0.396 [Fe(5N02phen)3]2+g 0.155 0.227 0.269 0.385 0.446

“Ionic strength, Z, for each reaction due to the dissociating agent except as in ‘. h Nucleophilic attack. “[OH-] = 0.010 mol dm-3. ‘[OH-] = 0.0050 mol dmp3. ‘[OH-] = 0.073 mol dm- 3. ‘I = 0.33 mol dm-’ (KCl). 9 Aquation, in 0.010 mol dmm3 EDTA.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

p / kbor

Fig. 2. Plots of -log&,/k,) against pressure for dis- sociation of [Fe(bpy),12+ in 0.010 mol drnm3 NaOH (0) and for dissociation of [Fe(5Brphen),12+ in the presence

of 0.010 mol dmM3 EDTA (m), at 298.2 K.

(5Brphen)s12+ in Fig. 2. Activation volumes obtained from the slopes are reported in Table 2, together with published data for other iron(I1) diimine com- plexes. The results are separately grouped for dissociation by nucleophilic attack and for aqua- tion in acidic medium” or in the presence of EDTA. No correction was made for the con- tribution of aquation in the nucleophilic-dependent dissociation reaction as the former represents a very minor contribution to the observed rate constant. For dissociation by nucleophile most systems were studied at an ionic strength of 0.33 mol dmm3. This and other conditions chosen reflect the need for data comparison in aqueous solvent mixtures (not reported in this communication) and practical con- siderations regarding solubility and rate ranges for accurate and convenient measurement.

DISCUSSION

Dissociation by hydroxide ion

All A V* results for nucleophilic attack on iron(I1) complexes, except those reported earlier,’ are within the range + 10 to + 16 cm3 mall’. Our results for the reaction of hydroxide on [Fe(bpy)s12+

and [Fe(phen)s12+, +13 and +14 cm3 mol-‘,

respectively, are in the centre of the small range. The earlier A V* values obtained for both hydroxide and cyanide ion reaction with these two complexes are + 19 to +22 cm3 mol- ‘, and clearly much higher than those reported here. The measurement temperature for the two studies differed by 5°C ; this would give rise to a very small difference in A V* values. There is a difference of about 0.10 mol dm- 3 in the ionic strength of the solutions of the two sets of experiments, but again this would prob- ably not account for the whole difference in AV* values. Accordingly, since our values are within the range of + 10 to + 16 cm3 mall ’ found for six other diimine complexes of iron( we conclude that the differences in A V* remain unexplained and beyond experimental error. It is our position that the rate law for the reaction”“*” indicates that the reaction is associative, and that AV* can be attributed to the common domination of release of electrostricted solvent from hydroxide upon reaction with the com- plex ion in achieving the transition state. Since hydroxide desolvation predominates, the solvation changes at the complex ion may well be minor. Values of AV* differ by only k2 to 3 cm3 mall ‘, which is not much greater than experimental error

Volumes of activation of diimine iron(I1) complexes

Table 2. Activation volumes for dissociation of iron(I1) diimine cations in aqueous solution

707

Complex” I AV*

Reactant (mol dm- ‘) (cm3 mol- ‘) (KT) Reference

[Fe(bpy)31z+

FeWen)3l *+

[Fe(4Mephen),]*+ [Fe(Me2bsb)3]2+ [Fe(3Mebsb),12+ [Fe(4MeObsb),]‘+ [Fe(hxsb)]*+ ~Wgm9312+ [Fe(5Brphen)3]2+d [Fe(5N02phen)3]2fd [Fe(5N02phen)3]2+e [Fe(phen)3]2+’ [Fe(4,7Me2phen),12+’

NaCN 0.10 20.9 + 1.6 293.2 1 NaOH 0.10 21.5kO.4 293.2 1 NaOH 0.010 12.8kO.4 298.2 This study NaCN 0.10 19.8+ 1.0 293.2 1 NaOH 0.10 19.7f0.3 293.2 1 NaOH 0.0050 14.2kO.8 298.2 This study KCN 0.033 10+2 298.2 2 KOH” 0.33 ll.lk1.6 298.2 13 KOHh 0.33 13.6f 1.8 298.2 This study KOH* 0.33 12.0f2.1 298.2 This study NaOH’ 0.33 13.4f 1.9 298.2 13 NaOH’ 0.33 16.7 298.2 12 EDTA 0.030 22.3 + 1.0 298.2 This study EDTA 0.030 21.7f 1.0 298.2 This study H2S04 3.0 17.9f0.3 308.2 10 H2S04 3.0 15.4 308.2 10 H2SO4 3.0 11.6 308.2 8

” Ligands and abbreviations defined in text and formulae 16. ’ KC1 was used to maintain the ionic strength. ‘NaCl was used to maintain the ionic strength. “Aquation in the presence of 0.010 mol dm- 3 EDTA. ’ Aquation in the presence of 1 .O mol dm- 3 H,SO+

and an attempt at an explanation is therefore not warranted. If the changes are not minor they are quite common to each iron complex cation. This situation is dramatically different for dissociation in aqueous alcohol mixtures of complexes 2,3 and 4 ’ 2,’ 3 illustrating therein significant ligand influ-

ence upon solvation changes. Preliminary findings for dissociation by hydroxide of [Fe(phen),12+ and

PWw)312f in aqueous alcohol mixtures at high pressures indicate a significant participation by the hydrophobic ligands, more so for [Fe(phen)3]2f than for [Fe(bpy)3]2+, in the solvation contribution to AL’*.

Activation parameters from temperature depen- dence studies provide support for different mech- anisms for the aquation and nucleophilic attack routes of dissociation of [Fe(phen)3]2+. Aquation, discussed in more detail below, in the presence of 1.0 mol dmp3 H,SO, is characterized” by a AH*

value of 122.4 f 0.8 kJ mall ‘, while dissociation by 0.1 mol dm- 3 OH- has a AH* value of 96.3 + 0.5 kJ mol-‘. The values of AS* are 88f3 and 18+3 J mol- ’ K- ’ for aquation and dissociation by OH-, respectively. In the present work, we have measured the temperature dependence of the dissociation of [Fe(SBrphen),]‘+ in the presence of 0.10 mol dme3 HCl and 0.010 mol dmm3 EDTA at five tem- peratures and find AH* values of 115.6 and 119.0

kJ mol- ‘, and AS* values of 74.7 and 86.4 J mall ’ K- ‘, in the two media. (Table 3 and references therein.) In contrast, the dissociation in the presence of OH- yields AH* = 84.2 kJ mall ’ and AS* = 37.8 J mall ’ K- ‘. Such results argue against a common mode of dissociative release of diimine ligand.

In summary, the interpretation of a dissociative process based only on Al’* values of ref. 2 is not unreasonable ; however, the nature of the rate law, and other considerations outlined above, yield a preponderance of support for an associative process for nucleophilic attack.

Aquation

During aquation of [Fe(phen),12+ one ligand becomes extended to a certain distance from the central ion thus permitting protons to prevent re- attachment of a phen ligand and causing the sub- sequent complete complex dissociation. It has been shown previously,8 and in this study, that EDTA can perform an equivalent function to H+ in the aquation process. Both can prevent the partly dis- sociated phen ligand reattaching itself completely to the central iron ; EDTA can react with the cation, and the acid can protonate a phenanthroline nitro- gen. The rate constants at atmospheric pressure for

708 J. BURGESS et al.

Table 3. Activation parameters for dissociation of complexes

iron(I1) tris-phen

AH* AS* (kJmol_‘) (J mol-’ Km’) Reference

Base hydrolysis

[Fe(phen)j]z+a [Fe(SBrphen),]‘+’

Aquation

[Fe(phen)J*+ [Fe(5Brphen),]2+d [Fe(SBrphen),]*+’

96.3 18 Cited in ref. 10 84.2 37.8 This study

122.4 88 Cited in ref. 10 115.6 74.7 This study 119.0 86.4 This study

“0.1 mol dm-3 NaOH. h Parameters derived from second-order rate constants.

’ 1.0 mol dme3 H,SO,. ‘0.1 mol dmp3 HCl. ‘0.01 mol dm-’ Na, EDTA.

aquation in the presence of acid and EDTA are the same and virtually independent of concentration. The associated enthalpies of activation for these two reactions are the same,9 as discussed earlier, confirming the idea of a scavenging function. Aquation occurs through a dissociative mode of activation giving rise to similar AV* values for similar ligands. The differences found” for [Fe(5N02phen),]*+, [Fe(4,7diMephen)J*+ and

[Fe(phen) A *+, +17.9, +11.6 and + 15.4 cm3 mol- ‘, respectively, can be rationalized in terms of a spin-state change of the iron( ‘o.22 but also with differing degrees of bond lengthening (Fe-N bonds) in the transition state. The value of AI’* for aquation of [Fe(5Brphen),12+, +22.3 + 1.0 cm3 mall ’ (Table 2) is consistent with the variation in AV*, caused by ligand substituent, that was found earlier lo (Table 2). However, practical reasons ruled out study of the aquation kinetics at high pressure in the directly comparable acidic medium (1 .O mol dmA3 H2S04). Where a parallel study was possible, we report AI/* values of + 21.7 f 1 .O cm3 mall ’ in the presence of EDTA at 298.2 K, and 17.9 +0.3 cm3 mall ’ lo for aquation of [Fe(5N02phen),12+ in 1.0 mol dmP3 H2S04 at 308.2 K. This difference is larger than could be explained by experimental un- certainty or the different temperature of measure- ment. At present we are unable to examine the generality (or not) of this result. However, it would appear that in addition to differences in bond exten- sion within complexes containing similar ligands, solvation effects by virtue of reaction in very differ- ent media may yield a significant difference in A V* for aquation with the same ligand ;23-25 the ionic strength is 3.0 mol dm- 3 in the acid aquation study,

much higher than in the presence of 0.010 mol dm- 3 EDTA. This gives rise to different structural charac- teristics of water in general and in the vicinity of the activated complex in the two reactions. Exactly how this gives rise to the particular solvation differ- ence, and whether ionic strength effects are as sig- nificant as in redox reactions23-25 remain to be explained. The complicated dependence upon [H+] for dissociation of [Fe(bpy),]*+ 26-28 precludes a parallel proton/EDTA aquation investigation.

Dissociation of [Fe(phen),]*+ in the presence of 0.1 M CF3S03H in anhydrous methanol yields2*” an activation volume of + 14.1 cm3 mol- ‘, close to the value for aquation in aqueous acid solution (+ 15.4 cm3 mol- ‘). But both are well below the reported values for dissociation in the presence of the nucleophiles, OH- (AI’* = + 19.7 cm3 mall ‘) and CN- (AV* = + 19.8 cm3 mall’). The close- ness of the two AV* values for the aquation is perhaps surprising in view of the pronounced differences in the reaction media, and bearing in mind the possibility that ionic strength variation may cause different volume changes upon achieving the transition state. However, it has been suggested ‘o,22 that AV* for a dissociative reaction of this type is dominated by the spin-change pre- equilibrium and is only modestly sensitive to aro- matic substitution, and therefore perhaps also to medium changes. It is almost ironical that A V* for dissociation of [Fe(phen)3]2f by OH- according to the new results reported here and for aquation are numerically the same, although we have argued that the mechanisms are different. Therefore, based upon the interpretation presented here, the agree- ment is fortuitous.

Volumes of activation of diimine iron(I1) complexes 709

Acknowledgements-We thank Royal Dutch She11 Oil Company, H. J. Backer fund, the Henriette Beck Fund and the University of Groningen fund for support

(S.A.G.), the University of New Hampshire for a sabbati- 13. cal leave during which some of the research was con- ducted in Leicester (C.D.H.) and the National Science 14. Foundation for a grant towards the purchase of the Cary 219 spectrophotometer (CHE-7908399). 15.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

G. A. Lawrance, D. R. Stranks and S. Suva- chittanont, Znorg. Chem. 1979, 18, 82. J. Burgess, A. J. Duffield and R. Sherry, J. Chem.

Sot., Chem. Commun. 1980,350. T. Asano and W. J. LeNoble, Chem. Rev. 1978,73, 407. M. J. Blandamer, J. Burgess and R. I. Haines, J. Chem. Sot., Dalton Trans. 1976, 1203. T. R. Sullivan, D. R. Stranks, J. Burgess and R. I. Haines, J. Chem. Sot., Dalton Trans. 1977, 1460. R. van Eldik, T. Asano and W. J. LeNoble, Chem.

Rev. 1989,88, 549. D. W. Margerum and L. P. Morgenthaler, J. Am.

Chem. Sot. 1962,&I, 706. J. Burgess and M. V. Twigg, Transition Met. Chem.

1978,3, 88. S. A. Galema, Research Report. University of Leicester, Leicester, U.K. (1988). J. M. Lucie, D. R. Stranks and J. Burgess, J. Chem.

Sot., Dalton Trans. 1975, 245. M. J. Blandamer, J. Burgess, P. P. Duce, K. S. Payne, R. Sherry, P. Wellings and M. V. Twigg, Transition Met. Chem. 1984, 9, 163.

12. M. J. Blandamer, J. Burgess, P. Guardado, C. D.

REFERENCES

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26 27.

28.

Hubbard, S. Nuttall, L. J. S. Prouse, S. Radulovic and D. R. Russell, Inorg. Chem., to be sub-

mitted. J. Burgess and C. D. Hubbard, J. Am. Chem. Sot.

1984,106, 1717. J. Burgess and C. D. Hubbard, J. Chem. Sot., Chem.

Commun. 1983, 1482. J. Burgess and C. D. Hubbard, Znorg. Chem. 1988, 27, 2548. W. M. Banick and G. F. Smith, Anal. Chim. Acta.

1958,19,304. M. L. Moss, M. G. Mellon and G. F. Smith, AnaIyt.

Chem. 1942,14,931. D. H. Wilkins and G. F. Smith, Anal. Chim. Acta.

1953, 9, 538. G. F. Smith and F. P. Richter, Phenanthroline and

Substituted Phenanthroline Indicators. G. F. Smith Chemical Co., Columbus, Ohio (1942). P. Krumholz, Inorg. Chem. 1965,4, 609 ; J. Burgess, J. Chem. Sot. 1968,497. N. Hallinan, P. McArdle, J. Burgess and P. Guardado. J. Organomet. Chem. 1987,333,77. (a) G. A. Lawrance and D. R. Stranks, Znorg. Chem. 1978, 17, 1804; (b) G. A. Lawrance, D. R. Stranks and T. R. Sullivan, Aust. J. Chem. 1981, 34, 1763 ; (c) G. A. Lawrance, Aust. J. Chem. 1985, 38, 1117.

S. Wherland, Inorg. Chem. 1983, 17,2349. L. Spiccia and T. W. Swaddle, Znorg. Chem. 1987,

26, 2265. R. van Eldik, T. Asano and W. J. LeNoble, p. 591, ref. 7. P. Krumholz, Nature 1949, 163, 724. J. H. Baxendale and P. George, Trans. Faraday Sot.

1950,46, 736. F. Basolo, J. C. Hayes and H. M. Neumann, J. Am.

Chem. Sot. 1954,76,3808.