CHAPTER

5
RATES AND MECHANISMS

OFREACTIONS OF COORDINATION COMPOUNDS

Having discussed the history and nomenclature (Chapter 2), structures (Chapter 3), and the bonding (Chapter 4) of coordination compounds, we turn now to a treatment of the reactions of these compounds. We wiill start with a discussion of the common types of reactions usuaiiy encountered, but this simple categorization wiil iead quite quickly into the more important questions of the how these reactions actually take piace. For example, we wiil start to consider whether reactions happen primariiy because (1) two reactant molecuies colhde to produce an intermediate that subsequently falis apart into the product moiecuies or (2) a reactant molecule falis apart first and a resulting fragment then coilides with another reactant tu make a product moiecuie. These are questions concerning the pathways, or mechanisms, for reactions involving coordination compounds. We wilhl then see that the favored mechanism, that is, the one which is the most apt tu happen, is the one characterized by the smaflest energy increase accompanying the formation of a reaction intermediate. An investigation of energy changes wiil involve our knowledge, albeit recently acquireci in the iast chapter, of the bonding in various reactant, product, and possible intermediate moiecuies.

5.1

A BRIEF SURVEY OF REACTION TYPES Figure 5.1 classifies the more common reactions of coordination compounds.. Starting in the center with a general compound {ML~]~~, substitution, dissociation, addition, redox or eiectron transfer reactions, and the reaction of a coordinated Iigand are indicated. Note that there is one major oversimplification in Figure 5.1, nameiy, that ali the hgands in the starting compound are shown to be identicai.

_________-

______-

89

90

COORDINATION CHEMISTRY

(These are known as homoleptic compounds, those iii which ali the ligands attached to the central metal atom or ion are identical.) Most reactions of coordination compounds actuaily involve heteroleptie compounds (those with more than one type of ligand), but for reasons of simplicity, these are not specificaliy reprcsented in the figure. Additionally, as we will see, there are a signfficant number of reactions that are a combination of more than one of the types shown in the figure. With these caveats (notes of warning), we can now turn to a brief discussion of each of these rcaction typcs. Substitution reactions, shown vertically at the top of the figure, are the most cominon. These involve the substitution of one ligand in a coordination sphere for another without any change in the coordination number or the oxidation state of the metal. Sometimes there is a wholesale replacement of ali the ligands of a homoleptic compound with a different ligand as shown in Equation (5.1), but more often only a fraction .of the original ligands are replaced as shown in Equations (5.2) and (5.3): [Cu(H2O)4]2~ + 4NH3(aq) [Cr(CO)6] + 3PPh3

[Cu(NH3)4]2~ + 4H20

(5.1) (5.2) (5.3)

[Cr(CO)3(PPh3)3j + 3C0(g) (Pt(NH3)ci3]

[PtCI4]2 + NH3

~*

+ CV(aq)

The tendency of such substitution reactions to take place is often tabulated in terms of stepwise and overail equilibrium constants. For example, the reaction of the tetraaquocopper(H) ion with ammonia in aqueous solution, shown in Equation (5.1) above, can be broken down into the stepwise replacement of one watcr ligand at a time as given in Equations (5.4) to (5.7): [Cu(H2O)4]2~ + NH3
K1=
.

[Cu(NH3)(H2O)3j2~ + H20

(5.4)

[{cu(NH3xH20)3}2j [{cu@2o)4}2~ [NH3]

[Cu(NH3)(H2O)3j2~ + NR3

[Cu(NH3)2(H2O)2j2~ + H20

(5.5)

[{cu(NH3)2(H2o)2}2~ [{cu(NH3)(H2o)3}2~ [NH3] [Cu(NH3)2(H2O)2j2~ + ~

K3=

[Cu(MT3)3(H2O)}2~ + H20

(5.6)

[(cu(NH3)3(H2o)}2~
2+

1
(5.7)

[{cu(IuIs)2(H2o)2}

j {NH3]

[Cu(NH3)3(H20)]2t + NH3
K4=

[Cu(NH3)4j2 + H20

[{cu(NH3)4}2+] [{cu(NH3)3(H2o))
2+

][~31

RATES AND MECHANISMS OF REACTIONS OF COORDINATION COMPOUNDS

91

Classifications of Reactions of Coordination Compounds [ML~_1L]4-+ L

substitutionj +iJ

[ML,,. 1]~~

diss~i~ion redox or 01ec[ron/

[ML,] +

adi~on

[ML,,Ll +
FIGURE 5.1

reaction o! coordinated
.

[ML~] (n+1)+ or

[ML0_ 1L] +

Five different types of reactions of co ordination compounds: substitution, dissociation, addition, redox or electron transfer, and the reaction of a coordi nated ligand.

The equilibrium constant K~ for eac&sixccessive replacement of one ligand for another is called, logically enough, atstepwise equilibrium constant. Since these reactions are carried out in aqueous soluohs Whre the concentration of water is very nearly constant, the concentration of water, [H20], does not appear in the mass action expressions for the equilibrium constants but rather is incorporated into the value of each K. (You should be familiar with this tactic from your previous studies, for example, the expressions for K0 and K,, used in discussing aqueous acid-base equilibrium problems.) The overail replacement of the four water ligands with animonia ligands [Equation (5.1)] can be described by an overali ~u~;~equi1ibrium constant, symbolized by a /3. Since four ligands are replaced, this is designated /34. The overail constant is just the product of the four stepwise constants as shown in Equation (5.8):
=

[{cu(~3)4}2~
[{cu(H2o)4}2~ [NH3]~

K1K2K3K4

(5.8)

As we have seen in Chapter 3 on the structures of coordination compounds, there are many more examples of heteroleptic complexes than homoleptic. These, too, undergo substitution reactions. Equations (5.9) and (5.10) show two common examples. The first, the reaction of a complex with water, is an aquation reaction, while the second, the reaction with an anion, is called an anation reaction. We will return to a more detailed discussion of these various reactions in Section 5.3. [co(NH3)5NO3]2~ + H20
~

[Co(NH3)5H2Oj3~ + N03

(5.9) (5.10)

[Co(NH3)5(H2O)j3~ + cr. [co(NH3)5cl]2~ + H20

Dissociation reaction.s, shown in the center left of Figure 5.1, involve a decrease in the number of ligands and sometimes, but not always, a decrease in the coordination number. An example involving a decrease in coordination number is the invisible ink reaction discussed in the last chapter and repeated below in Equation (5.11). A dissociation reaction in which the coordination number does not decrease because a bidentate ligand is replaced by two monodentate ligands is

92

COORDLNATION CHEMISTRY

given in Equation (5.12).

2[Co(H20)6]C12
Red

Co[CoCIJ
Bine

12H20
+

(5.11) (5.12)

[Cr(en)3jC13

120C

[CrCI2(en)2jC1

en

Addition reactions, shown in the center right of Figure 5.1, are accompanied by an inerease in the coordination number of the metal. For steric reasons, most addition reactions occur with complexes in which the metal has a low coordination number initially. For example, Equation (5.13) shows a tetrahedral titanium compound accepting two additional chioride ligands to produce an octahedral hexachlorotitanate(IV) complex. Equation (5.14) shows a square planar bis(acetylacetonato)copper(II) molecule accepting a pyridine (py) ligand to form a square pyramidal product. TiCI4 + 2C1 {TiCI6]2 Cu(acac)2
+

(5.13) (5.14)

py

Cu(acac)2py

Oxida tion-reduction or electron-transfer reactions, shown in the lower left of Figure 5.1, involve the oxidation or reduction of a coordinated transition metal atom or ion. The figure represents the simplest case of electron transfer, namely, one in which the coordination number remains constant. Equation (5.15) shows an example in which the hexaammineruthenium(HI) ion is reduced by the addition of chromium(II) ion, while Equation (5.16) shows the oxidation of a hexacyanofer rateO) complex by hexaehloroiridate(IV). In neither of these complexes is the coordination number altered. [Ru(NH3)6]3~ + Cr2~ [Ru(NH3)6j2~ [Fe(CN)514 + [1rCl6j2

Cr3~

(5.15) (5.16)

[Fe(CN)5]3

[IrCl6]3

There are examples, not represented in Figure 5.1, in which the oxidation of a complexed metal atom or ion is accompanied by the addition of one or more ligands so as to increase the coordination number. These are common enough reactions that they are designated oxidatiue-addition reactions. One of the very first reactions of this type was first characterized by L. Vaska in the 1960s and is shown iii Equation (5.17): trans-[Ir(CO)Cl(PPh3)2] + HCI

[Ir(CO)CI2H(PPh3)2]

(5.17)

Note that in the reactant, quite often referred to as Vaskas compound, the coordination number of the iridium(I) ion is 4, while in the product the iridium is formally of oxidation state +3 and has a coordination number of 6. The opposite of oxidative-addition, logically enough, is called reductive elimination. Equation (5.18) shows the reductive elimination reaction of carbonyl chlorodihydridobis(triphenYlPhoSPhifle)iridium(111) to form diatomic hydrogen and Vaskas compound: [Ir(CO)CIH2(PPh3)2j
...

[Ir(CO)Cl(PPh3)2]

H2

(5.18)

Reactions of coordinated ligands, shown in the lower right of Figure 5.1, are reactions of a ligand that take place without the breaking of the metal-ligand bond. Equation (5.19) shows a water Iigand in hexaaquochromium(III) reacting with a hydroxide ion to produce the corresponding hydroxo complex. Here the Cr0H2 bond has not been broken, but the water has lost a proton to produce the hydroxo ligand. Equation (5.20) shows the reaction of pentaamminecarbonatocobalt(IIJ)

Is__.

RATES AND MECHANISMS OF REACTIONS OF COORDINATI0N C0MPOUNDS

93

with acid to form pentaammineaquocobalt(III). This reaction proceeds without the breaking of the CoOCO2 bond. Another common ligand that reacts while coordinated is acetylacetonate. Equation (5.21) shows the replacement of the central hydrogen atom of acac with a bromine atom. [Cr(H2O)6]3~ + OH- [Cr(H2O)5(OH)]2~ + H20 [Co(NH3)5C03]
+

(5.19)

+ 2H3O~. [Co(NH3)5(H2O)]2~ + 2H20 + CO2 (5.20)

+ 3HBr

(5.21)

Before leaving this brief survey of reaction types, a word is in order concern ing how the course of these various reactions might be monitored. As we have already discussed in Chapter 4, different complexes have different uvvisible (or absorption) spectra. The substitution, dissociation, addition, or alteration of a ligand or ligands often changes the maximum wavelength of uvvisible light absorbed. The sarne can be said when the oxidation state of the metal changes. These changes in the uvvisible spectra (often accompanied by changes in the actual color of the compounds) can be used to monitor the rate of these reactions. Changes in the molar susceptibility can be used in a similar manner. Other physical techniques such as infrared (ir) and nuclear magnetic resonance (nmr) spectroscopy, not discussed in this text, can also be used to monitor the progress of a given reaction. 5.2

LABILE AND INERT COORDINATION COMPOUNDS

To classify the varying rates of reaction (most commonly with regard to substitu tion) of coordination compounds, Henry Taube, who received the 1983 Nobel Prize in chemistry for his work in the kinetics of coordination compounds, suggested the terms labile and inert. If we consider a 0.1 M aqueous solution, a labite coordination compound is one that under these circumstances, hasa half-life of less than a minute. (Recail that half-life is the amount of time required for the cifrceiitratioix of the reactant to decrease to half its jnitial concentration.) An inert compound, on the other hand, is one with a half-life greater than a minute. We niust clearly keep in mmd that the terms labile and inert are kinetic terms. They refer to the rate of a reaction that in turn is governed by its energy of activation E~. (Here, you may want to review your knowledge of kinetics as introduced in previous courses.) These terms relate to how fast a compound reacts rather than to how stable it is. Stable and unstable are thermodynamic terms. They are related to the changes in free energy, enthalpy, and entropy of the compound. Reactions with a large negative change in free energy and a correspondingly large positive equilibrium constant go spontaneously from left to right; the products of such reactions are considered more stable than the reactants. The change in free energy can in turn be related to the changes in enthalpy and entropy (recail that AG = AH TAS). To iliustrate the difference between kinetic labiiity and thermodynamic stability, we consider some specific exampies. Take the familiar Werner cornplex cation hexaamminecobalt(III), [Co(NH3)6]3t It reacts spontaneously with acid. In

94

COORDINATION

fact, the equilibrium constant for the reaction corresponding to Equation (5.22) is very large, in the vicinity of 1030: [Co(NH3)6]3~ + 6H3O~ [Co(H2O)6]3~ + NH4 (5.22)

Therefore, we would say that this cation is unstable toward reaction with acid. On the other hand, it takes several days at room temperature to get this reaction to go significantly from left to right even in 6 M HCI. ccordingly, the rate of this reaction is so slow that [Co(NH3)6]3~ must be classified as ineri under these circurnstances. This complex cation, then, is unstable (thermodynamically) but inert (kinetically) toward reaction with acid. In contrast, tetracyanonickelate(II), [Ni(CN)4]2, is exceptionally stable (ther modynaniically). The equilibrium constant for its formation, represented in Equa tion (5.23), is very large, also in the vicinity of

io~:

Ni2~+ 4CN [Ni(cN)4]2

io~

(5.23)

At the sarne time, this complex anion is labile; that is, the cyanide ligands in the coordination sphere exchange rapidly with those found free in au aqueous solu tion. This exchange rate can be measured when carbon 14labeled cyanide ions are placed in solution with the complex as represented in Equation (5.24): [Ni(CN)4] 2 + 4 14CN [Ni(14cN)4] 2 + 4CN (5.24)

The labeled cyanides exchange places with their unlabeled counterparts quite rapidly. Jn a rnatter of seconds, in fact, half the unlabeled cyanides are replaced with labeled ligands. The tetracyanonickelate(II) ion, then, is stable but labile. In summary, some coordination compounds are kinetically inert, while thers turn out to be labile. Furtherrnore, this lability seems to be unrelated to the thermodynamic stability of the compound. Now, being a veteran chemistry student trained to ask critical questions, you are about to ask how can we teu which complexes will be inert and which will be labile? As you might suspect, this is indeed a crucial question. It turns out that complexes of the first-row transition metal ions, with the exception of Cr3 + and Co3 ~, are generally labile, while rnost second- and third-row transition metal tons are znert But how, you ask, do we e~plain such a statement? Why, for example, should the rates of reactions involving Co3+ and Cr3+ be different from those involving other first-row transition metal atorns and cations? J4liat is it about these particular cations that makes them 50 inert? To start to answer such queries, we turn now to a discussion of some of the most extensively studied reactions of coordination compounds, those involving the substitution of octahedral complexes.

5.3

SUBSTITUTION REACTIONS OF OCTAREDRAL COMPLEXES

Possible Mechanisms
The overail substitution reaction of a homoleptic complex was represented in Figure 5.1. To discuss mechanisms, however, it will be helpful to consider a more specific situation. Take, for example, n octahedral coordination compound con taining a metal bound to five inert ligands (L) nd one labile ligand (X) that is about to be replaced by au incomihg ligand (Y). The overail equation for this reaction is ML5X + Y H- ML5Y + X (5.25)
-

RATES AND MECHANISMS OF REACTI0NS OF COORDINATION COMPOUNDS

95

(1)

ML5X

ML5+ X
five-coordinate niermediate

FIGURE 5.2

L
LML L
square pyramidal

L or

LM 1 L L
trigonal bipyramidal

The dissociative ) mechanism for Lhe substitution of one ligand for another ir an octahedral com plex, ML5X (L = inert ligands, X = jubile Iigand, and Y = incoming ligand). The mechanism assumes Lhat Lhe first step (1), the

(2)

ML5 + Y

fast
.~

Overail rate

breaking of the MX bond Lo form the flve-coordinate ML5 in termediate, is Lhe rate-determin rate of rate-determining step (1) ing step. The raLe of Lhe reaction is fourid lo be dependent only ~ = ,~ [ML XJ upon Lhe concentration of Lhe 1 5 original complex, [ML5X]. ML5Y

How might this substitution actually take place on te molecular levei? Or, to restate the question, what is the reaction ,nechanism, that is, the sequence of molecular-levei steps involved in the reaction? There weuld, at first blush, seem te be at least two major possibilities. One is calied the dissociative (13) mechanism, the other associatwe (A) In the dissociative mechanism, we picture the X ligand breaking off from the reactant te produce the five-coordinate ML5 intermediate. (This intermediate might weii assume a square pyramidai or perhaps a trigonal bipyramidal geometry.) In a second, faster reaction the intermediate and the incoming ligand join together te yieid the product. The complete mechanism is shown in Figure 5.2. Note that we have assumed that the siower first step, the dissociation, is the rate-determining step. (At this point, you may want to review your knowiedge of eiementary reaction mechanisms as covered in previous courses. Recail that the rate-determining step, sometimes called the bottleneck step, is the slowest elemen tary reaction in a mechanism. The rate of this step determines the overali rate of the reaction.) Note that the dissociative mechanism predicts that the rate of the overali substitution reaction is dependent only on the concentration of the ori ginal complex, [ML5X], and is independent of the concentration of the incoming ligand, [Y]. The other logicai pathway by which a substitution reaction might take place is the associative mechanism. In this case, the rate-determining step is the collision between the original complex, ML5X, and the incoming iigand, Y, to prod4ce a seven-coordinate intermediate, ML5XY. (This intermediate might assume a capped octahedral structure, in which the X and Y ligands share one of the normal octahedral sites, or possibiy a pentagonal bipyramidai geometry.) The second, faster step is the dissociation of the X ligand to produce the desired 1product. This mechanism is shown in Figure 5.3. Note that the associative mecha ~1nism predicts that the rate of the substitution reaction wiil be dependent upon the ~!concentration of both ML5X and Y.
-

So, you may be beginning to think, deciding whether a mechanism is. associative or dissociative shouid not be particuiarly difficult. It iooks like ali we need do is determine the rate iaw for the reaction: if the rate depends upon

96

COORDINATION CHEMISTRY

(1)

ML5X + Y

-~!, ML5XY

(seven-coordinate intermediate)

Y or

LLL

x
penlagonal bipyramid

bi-capped octahcdron

(2)

ML5XY

-~

ML5Y + )

Overail rate

rate of rate-determining step (1)

A[ML5Xj

k1[ML5X][Y]

FIGURE 5.3

li]

The assocjatjye (A) mechanism for lhe substitution of one ligand for another in an octabedral complex, ML5X (L = inert ligands, X = labile Iigand, and Y = incoming ligand). The mechanism assumes that lhe first step (fl, the formation of an M Y bond lo form lhe seven-coordinate ML5XY intermediate, is the rate-determining step. The rate of lhe reaction is found to be dependent upon both lhe concentration of lhe original complex, [ML5X], and that of lhe incoming ligand, [YJ.

[ML5X] oniy, it is dissociative, but if the rate also depends on [Y1, it is associative. Not surprisingly, coordination chemistry kinetics turns out not to be quite this simple. Two complications come immediately to mmd. First, the actual mecha nisms may be more complicated than the clearly differentiated D and mecha nisms outiined above. Second, there may be special experimental conditions that mask the dependence of a rate on the concentration of the incoming ligand. First, consider the idea that the pureiy associative and dissociative mecha nisms, while useful as starting points in a discussion of the kinetics of coordination compounds, are rather idealized and perhaps somewhat oversimplified possibili ties. Is it not just a little unreaiistic, for exampie, to suppose that the MX bond fuily breaks and the resuiting five-coordinate ML5 intermediate exists for a substantial period of time before it subsequently collides with Y to produce the final product? Or may we be expecting too much to suppose that the ML5XY intermediate in the associative pathway has a signfficant measurabie lifetime before the MX bond breaks and ML5Y results? After ali, if the real mecha nisms were this neat and simpie, we should have been abie to isolate a variety of ML5 and/or ML5XY intermediates and thereby add to the evidence that a given reaction is D or A. Unfortunateiy, isolation of such intermediates is rather rare. Given this discussion, we need to consider one other ~e of reaction pathway, ~ called the interchange (1) mechanism. ConsidSr iiitiaation where the entering ligand, Y, sits just outside the coordination sphere of the ML5X complex and, as the MX bond starts to break and X starts to move away, the MY bond simultaneously starts to form and Y moves into the coordination sphere. In this case, the attacking (Y) and ieaving (X) ligands gradually interchange places in the coordination sphere of the metal and no distinct five- or seven-coordinate intermediate wouid be formed (or, therefore, be available to be isolated). Such a reaction pathway, sometimes cailed a concerted

______________ -_

RATES AND MECL4ANISMS OF REACflONS OF COORDINATION COMPOUNDS

97

or an interchange mechanism, may well be more realistie than a purely D or A mechanism. Having described the above 1 mechanisrn, it is important to note that we do not have to totally discard the idea of dissociative and associative mechanisms. For example, if the MX bond breaks preferentially in the interchange, then the interchange is closer to a dissociative than to an associative mechanism. In this case, we could designate the mechanism d (interchange-dissociative), where the d subscript indicates the dissociative nature of the interchange. Similarly, when M Y bond formation is favored sornewhat, the mechanism would be designated as ~ (interchange-associa tive). So, in summary, we have described five types of mechanisms, D, A, 1, d and I~. Let us assume from here on out that while we will speak generally about dissociative and associative mechanisms, the terrns D and are reserved for situations in which five- and seven-coordinate intermediates have actually been isolated. II no interrnediates have been isolated, the designations 4, and I~ would seem to be more appropriate.

Experimental Complications In Figures 5.2 and 5.3 we considered the rate laws that would result from dissociative and associative mechanisms, respectively. In the end, you will recali, it appeared that these pathways could be readily differentiated by the dependence or lack of dependence of their rate laws on the concentration of the entering ligand, Y. As noted earlier, we need to discuss the possibility that this dependence may be masked by certain experimental conditions. Common examples of this masking of a concentration dependence are substitution reactions carried out in aqueous solution. Again considering the sarne overali substitution of Y for X in the ML5X reactant, two probable rate-determin ing steps are shown in Figure 5.4. Note that the dissociative mechanism still predicts a rate dependence only on the concentration of the initial complex, IML5X]. This result is shown in Equation (5.26). For the associative rnechanism, the reaction of the starting complex with a water molecule is assumed to be the rate-determining step. Such an assurnption is certainly not unexpected because the concentration of water in a dilute aqueous solu~ion is very large (approximately 55.6 M), much larger than the concentration of Y (perhaps about 0.1 M). The likelihood of the starting complex colliding with a water molecule is therefore much greater than a collision with Y. As shown in Equation (5.27), the rate of this step is dependent upon both the concentrations of ML5X and 1120; But now we have to recognize, as noted previously (p. 91), that [1120] is so large that, in this situation, it is essentially a constant. Therefore, as shown, the two constants in Equation (5.27), k and tH2O], are cornbined to yield Equation (5.28). Note that while the result appears to be first-order only in [ML5X], we how that it is also first-order in [H20]. This latter dependence, however, has been rnasked or hidden by the nature of the experimental situation. For this reason, Equation (5.28) is often said to represent a pseudo first-order reaction. In the cases discussed then, the rate laws for both the dissociative and associative mechanisms are, for practical purposes, identical. In this situation we cannot decide between the two mechai~isms through an analysis of the observed rate laws.

98

COORDINATION CHEMISTRY

(a)

Dissociative:

L
Rate determining step.

L
L~L

x
Rate[= overali L rate j

k[ML5x]

(5.26)

(first order in[ML5XJ)

(b) Associative:

L
Rate determining step

L
+

H20
L

x
Rate~= overaili [ rate j
= =

k[ML5x][u2o]

(5.27)

(k[H2oj}[ML5x] k[ML5XJ (pseudo first order in [ML5XJ)

(5.28)

FIGURE 5.4 The masking of concentration dependence in aqueous solution. (a) A dissociative mechanism is flrst-order in lhe concentration of the complex reactant. (b) An asspciative mechanism is first-order in both lhe complex reactant and lhe incoming water Iigand. However, since the concentration of waler in aqueous solution is so large and very nearly constant, [H201 is combined with k lo produce an overail rale constant k. The resulling rale taw is pseudo first-order. The dependence of the associalive mechanism ora the concentralion of lhe incoming water Iigand has been masked.

Evidence for Dissociative Mechanisms Throughout the above discussion, we still have not answered the fundamental guestion: What is the preferred mechanism of substitution reactions of octahedral coordination compounds? The answer is that disso da tive mechanisms are preferred. In this section we investigate some of the evidence that supports such a conclusion. We will cite three different types of reactions: (1) the rates of exchange of water molecules, (2) anation reactions, and (3) aquation reactions. We start with a strange set of measurements. These involve the exchange between the ordinary water molecules in the hydration spheres of various metal ions and isotopically labeled bulk water (H207). (Section 11.2 presents a more detailed description of water structure.) A general equation representing such reactions is given in Equation (5.29): M(H20)r + H2 017

ii

M(H20)5(H2 o~)~ + H20

(5.29)

The measurements of the rates of these reactions are carried out by a variety of methods, some very sophisticated, depending upon the rate of the exchange. Such

RATES AND

MECHANISMS OF

REAC1IONS

OF COORDINATION COMPOUNDS

99

methods demonstrate that these reactions are first-order in the concentration of the original hydrated cation, M(H20)r, that is, the rate Iaw for these reactions is as given in Equation (5.30). As discussed earlier, the fact that these reactions are first-order is consistent with but certainly not definitive proof of a dissociative mechanism.

[M(H2o)r]
t

k1[M(H20)r]

(5.30)

The relative rates for a large number of metal ions are given in Figure 5.5 in terms of the log of the rate constant for the reaction. Note that these rate constants (and therefore the rates themselves) vary over approximately 16 orders of magnitude, an amazing range of rates. Close examination of this data reveals that there is a definite increase in the rate of exchange (reflected in an increase in the rate constant) going down various groups of the periodic table. Look, for example, at the results for Group lA (Li~, Na~, K~, Rb~, and Cs~) or for Group 2A (Be2~, Mg2~, Ca2~, Sr2~, and Ba2~). Comparison of these two groups also reveals that the higher + 2 charged ions of Group 2A have slower rates than the + 1 charged Group lA ions. Can we expIam

(M2

10

Characteristic water exchange rate constants, log k (25C, s~ l)

FIGURE 5.5 RaLe constants for water exchange for various ions.
nt+

[M(H2o)~j~~ + 112017 ._~i~ [M(112o)5(112o91

+ 1120

Data are tabulated as Lhe Iog of Lhe iate constant at 2SGC. Jnert hydrated ions, those that only slowly exchange water molecules between Lhe hydration sphere and bulk water structure, are given on Lhe left, while labile hydrated ions are on the right. [Ref. 71

100
such results and is this explanation consistent with a dissociative mechanism for these exchange reactions? The explanation revolves around the effect of the charge and the size of a metal cation on the strength of the ~ OH2 bond. It shouid make senseto you that the higher the charge, the greater the eiectrostatic attraction between the positively charged metal and the electron pair of the water iigand. Therefore the strength of the ~ OH2 bond shouid increase with increasing charge. The larger the metal ion, on the other hand, the farther the center of its positive charge is from the ligand electron pair and, consequentiy, the weaker the bond. These two factors can be combined in what is known as the charge density, defined as the ratio of te charge on a metal cation over its ionic radius. By definition, the charge density of a metal cation increases with increasing charge and decreasing size. It follows from the above discussion that as charge density increases 80 also should the strength of the M0H2 bond. (Charge density is futher discussed in Sections 8.1 and 9.2.) Now we turn to a consideration of whether the data for these exchange reactions are consistent with a dissociative mechanism. Within a given group, ali the cations are the same charge and only the size varies. So, in Group lA, as we go down the group, the cations are getting larger and the charge density decreases. This in turn means the M~ OH2 bond is gettingweaker and more easily broken. li is, of course, the breaking of the M + OH2 bond that is the rate-determining step in a dissociative mechanism. Therefore, as the bond gets weaker and easier to break, it makes sense that the rate of the water exchange reaction should increase. When we turn te the doubly charged Group 2A ions, the charge density shouid be larger and therefore the strength of the M2~ OH2 bonds should be greater. Accordingly, it foHows from a dissociative mechanism that the rate of exchange shouid be slower. Comparison of the data for Groups lA and 2A shows that this is exactly the case. So, we see that these data are consistent with (but not necessarily proof ofl a dissociative mechanism for these reactions. We will post pone a discussion of the remainder of this table, specifically, the trends within the transition metais, until the next section. Anation reactions were defined eariier as those in which an anion replaces another ligand lii a given compiex. Equation (5.31) represents the anation of a hexaaquo metal cation with a 1 anion:

[M(H2O)6f~

+ x-~ [M(H2O)5x]~ + H20

(5.31)

Such reactions presumably could proceed by either a dissociative or an associative mechanism, but most evidence seems, again, te favor the dissociative. For exampie, Tabie 5.1 shows the rate constants for three anations of [Ni(H2O)6]2~ as weli as the reactions with neutral ammonia and water ligands. Note that these constants and therefore the rates of reaction show little or no dependence (iess than one order of magnitude) on the identity of the entering hgand. Furthermore, the vaiues are alI very similar to that of the rate of water exchange. (See the vaiue of log k for L = H2O in Tabie 5.1; also note that this vaiue is consistent with that obtained by reading Figure 5.5.) These data are consistent with a dissociative rate-determining step in which a water molecule breaks away from the nickei(II) and, in a succeeding fast step, is repiaced by L. Aquation reactions invoive the replacement of a Jigand (other than water itseif) with a water molecule. A considerabie amount of work has been done on reactions of this type invoiving a variety of inert cobaitOli) complexes. For

RATES AND

MECHANISMS OF REACTIONS OF COORDINATION COMIOUNDS

101

TABLE 5.1 Rate constants for substitution reactions of [Ni(H2O)612~ [Ni(H20)6]2+ L L F SCN CH3COO NH3 1120

.L

[Ni(H.O)5L]+ 1120
Iogk

k,s~

8x103 6 x io~ 30 X i~ 3x103 25 X i~

3.9
3.8
4.3

3.5
4.4.

Data taken from R. O. Willcins, Acc. Chem. Res., 3: 408(1970).

example, Equation (5.32) represents the aquation reactions of various penta ammineQigand)cobalt(III) cations:

{Co(NH3)5L]~ + H20

[Co(NH3)5(H2O)]~ + L

(5.32)

Table 5.2 gives some kinetic and thermodynaniic data related to this reaction. First note that the rate constants, unlike those for the anation reactions just discussed, now do seem to vary quite significantly with various ligands, L. Such a variation is consistent with a rate-determining step in which ML bonds of varying strength are broken.

TAELE 5.2

Rate constants for the aquation of pentaammine(ligand)cobalt(llI) complexes and equilibrium constants for the anation of pentaammineaquocobalt(ffl) with various anions L
Strongest Slowest raLe of reaction

NCS
J-12P04

5.0 X 10_lo 8.6 ~ 108 2.6

470 20

M L bonds

1 1 Pastest
iate of reaction

10

Cl
Br 1 N3~

l.7x106
6.3 x 106 8.3 )< lo_6 2.7 < io~

7.4 1.25
0.37 0.16 0.077

1
Weakest MLbonds

The raLe constants k refer Lo Lhe following aquation reactions: [Co(NH3)5L]2+ 1120

L [Co(NH3)5(112O)]3~+
r

The equilibrium constants K, refer to Lhe following anation reactions:

[Co(NH3)5(H20)]3~+

K,, [Co(NH3)5L12+ 1120

The slowest rates of aquation correspond Lo Lhe largest equilibrium constants for anation.
Data Laken from F. Basolo and E. O. Pearson, Mechanisnis of fnorganic Reactions, A Study of Metal Compiexes in Solution, 2d cd., Wiley, New York, 1968, pp. 164166.

102

COORDINATION CHEMI5TRY

Weakest ML bonds 2 Slowest iate of reaction 1 O

Iog I<~ 1 2

Strongest ML bonds 3

log k

Fastest iate of reaction

ir.

FIGURE 5.6 A pIoL of Iog K, versus Iog k for a variety of pentaammine(Iigand)cobalt(III) coinplex cations. The iate constants k refer Lo Lhe following aquation reactions:

[Co(N~H3)sL12~ + H20 ~L, [Co(NH3)5(H2O)]3~

The equilibrium constants K, reter Lo Lhe following anation reactions: [co(NH3)5(H20)]3
+

4~= [Co(NH3)5L]2~ +

The value of K, is a measure of the ML bond strength with the strongest bonds having Lhe Iargest values of K3. The value of k is a measure of Lhe iate of the aquation. The smallest values of Iog k have the slowest iates of reaction. The pIoL shows that Lhe stronger Lhe ML bond, the slower the iate of aquation. -

Can we correlate the above aquation rate constants with a measure of the M L bond strength? The third column of Table 5.2 shows the equilibrium constants for the reactions in which a water ligand is replaced by the L anion in each case. Notice that the only major difference in these various reactions would seem to be the strength of the ML bond. In fact, it can be shown that log Ka for these reactions is directly proportional to the ML bond strength. (See Problem 5.30 for an opportunity to demonstrate this relationship.) The relationship between log Ka and k is in tum plotted in Figure 5.6. It demonstrates a straight1ine relationship between these two parameters and therefore a strong correlation between the rate constant and the ML bond strength. That is, as the ML bond strength increases, it becomes more difficult to remove the L, and the rate decreases. The evidence for dissociative pathways for octahedral substitution reactions continues to mount.

RATES AND MECHANISMS OS REACTIONS OF COORDINATION COMPOUNDS TAELE 5.3

103

Rate constants for the aquation of trans-bis(substituted-ethylenediamine) dichlorocobalt(HI) complexes:

trans-[CoCI2(H2NCR1R2CR3R4NH2)2]~+ H20

L
CI

trang-((H2O)CICo(H2NCR1R2CR3R4NH2)2]2~+ Groups on bidentateamine R1 Increasing bulk of bidentate amine ligand H CH3 CH3 CH3 CH3 R2 H H H C113 CH3 H II CH~ H C1-13 H H H 14 CH3 32 x io5 6.2 )< iO~ 4.2 x io4 2.2 x iO~ 3.2 x 1O_2

Increasing rate of reaction

Data taken trem 5. Basolo and R. O. Pearson, Mechanisms Coniplexes iii Solution, 2d cd., Wiley, New York, 1968, p. 162.

of

Inorganic Reactions, Study of Metal

Other aquation reactions give further evidence concerning the most favored mechanism of octahedral substitution reactions. Consider the data for the aquation of various bidentate amine complexes of cobaltUil) as shown in Equation (5.33) and Table 5.3:

[CoCI2(H2NCCNH2)2]4+ H20
R2R4
~1~3

[Co(H2O)C1(H2NCCNH2)2]2~+ C1 R2 R4

(5 33)

Notice that the more methyl groups (and therefore the bulkier the bidentate amine), the faster the rate of aquation. To see if this is consistent with either a dissociative or associate mechanism, consider the rate-determining step in each case as shown in Figure 5.7. In Figure 5.7a, the associative rate-determining step is shown. What would happen to the rate of such a reaction as the bulk of the bidentate amine increased? Simply put, tlie incoming water ligand would find it more difficult to make its way into the metal to form thd seven-coordinate intermediate, and therefore the rate would decrease. This is not what the data in Table 5.3 indicate happens, and therefore such data are not consistent with an associative mechanism. In Figure 5:7b, on the other hand, the dissociative rate-determining step is shown. Here the increased bulk of the bidentate amine would increase the steric hindrance around the metal and help to force the chioride ligand out of the coordination sphere. The dissociation of the chloride ligand would be more favorable as the bulk of the amine increased, and therefore the rate of the reaction would increase. This increase with the bulk of the ligand is exactly what Table 5.3 indicates happens. Therefore, the data are consistent with a dissociative mecha nism.

104

COORD1NATI0~ CHEMISTRY

CI R2H2 Rl%!~N R4% R3

/
N
112

/1

7 Co3~
ci

N CR1
+ H20

11R3

CI OH2 H2 R2H2\ / N R2 R1 N 7 CR1 ~ \A/ R4%// /1 C~ / N 112 R3 H2 Cl

(a) Associative rate-determining step

R211z Rl%T~N

z
ci

112

~CCR, -ci-

R4%c
R3

/ /
H2

/
H2 R3 N

R2 CR1

N~R4 H2 R3

(b) Dissociative rate-determining step

FIGURE 5.7

(a) Associative rate-determining step. The entering water ligand attacks the Co~ to form the seven-coordinate intermediate of the associative mechanism. As the bulk of tbe bidentate amine increases, the water ligand will find it more difficult to get Lo the metal cation. Therefore, Lhe raLe of Lhe rate-determining step should decrease with the increasing bulk of Lhe amine. (b) Dissociative rate-determining step. The chloride ligand dissociates Lo form Lhe five-coordinate intermediate of the dissociative mechanism. As Lhe bulk of Lhe bidentate amine increases, Lhe chloride ligand wiii experience more steric hindrance and will be pushed out of Lhe coordination sphere more easiiy. Therefore, Lhe rate of Lhe rate-determining step should increase with Lhe increasing bulk of Lhe amine.

Finally, consider the effect of changing the overali charge on a complex. Again, mnch work has been done with cobalt(IH) compounds. For the two reactions shown ia Equations (5.34) and (5.35), note that the faster rate (larger rate constant) goes with the smaller net charge: {Co(NH3)5Clj2~ + H20

[Co(NH3)5(H2O)13~ + Cl k 6.7 x 106 s

(5.34)

:rans-[Co(NH3)4C121 + + H20

[Co(NH3)4(H2O)Cl12~ + C1

(5.35)

1.8 >< ~ 5~~1

Is this consistent with a dissociative or associative mechanism? Jn the dissociative step, the greater the charge on the complex cation, the more difficult it wouid be to remove a chioride anion from it and therefore the slower rate of dissociation. This is consistent with the rates that are given. In the associative mechanism, on the other hand, the greater the net charge, the more attracted the incoming water Iigand would be and the faster the rate-determining step. Such a mechanism is jnconsistent with the data.

RATES AND

MECI-LANISMS

OF

REAaIoNs

OF COORDINATION COMPOUNOS

105

To conclude this section, we have seen that various pieces of data from (1) rates of water exchange, (2) anations, and (3) various aquation reactions ali seem to generaily favor a dissociative mechanism for substitution reactions of octahedral compounds.

ExplanatiOn of Inert versus Labile Complexes Now that we have fairly wefl estabhshed that the dissociative mechanism generaily appiies for the substitution reactions of octahedral complexes, we are in a good position to begin to answer some of our earlier (p. 94) criticai questions about inert versus labile complexes. As defined earlier, labile and inert are kinetic terms describing the rates of reactions of coordination compounds. Rates, in turn, you should recai! from earlier courses, are dependent on the magnitude of the energy of activation Ea of the rate-determining step. Figure 5.8 shows a typicai reaction profi!e, that is, a piot of potentiai energy versus reaction pathway. Recail that the r~actants must be converted to a transition state or activated complex before they can be transformed into the products. The
energy difference between the reactants and the transition state is cai!ed the energy of activation and must be attained before a reaction can take piace. In general, the rate-determining steps of slow reactions are characterized by high activation energies and, conveisely, those of fast reactions by 10w E0s. The Arrhenius equation, k = Ae_Ea/fl, gives the exact dependence of the rate con stant (and therefore the rate itself) on the energy of activation. (R is the gas constant, T is the temperature in degrees kelvin, and A is a constant often ca!led the coilision frequency.) Now for a substitution reaction in which the rate-determining step is the

dissociation of a iigand, what are the various factors that contribute to the activation energy? Some of these were discussed earlier but not in the context of Ea5 For examp!e, we said ear!ier that the size and charge of the metal cation
influence the strength of the ML bond. Now we can say that, since this bond is broken during the rate-determining step, it directiy affects the magnitude of Ea. Meta!s with !arger charge densities [or charge-to-radius (Z/r) ratios] have stronger M L bonds, higher energies of activation, and therefore slower rates of substitu tion. We a!so discussed steric hindrance about the metal and the overafl charge on a comp!ex as having a bearing on the rate of substitution reactions. Compiexes with litt!e steric hindrance or a high overai! charge will have iarger M L bond strengths, higher energies of activation, and s!ower rates of reaction. Given these factors, can we start to exp!ain why, as noted earlier, the first-row transition metais (with the exception of Co3~ and Cr3~) are generaliy

Activated complex or transition state flGURE 5.8 A reaction profile: a plot of potential energy versus reaction pathway. Reactants must ac guire the energy of activation in order Lo achieve Lhe transition state or activated cora plex before they can be transformed into products. The higher Lhe energy of activation, the slower the reactiOn.

Reactants

Products Reaction pathway

106

CO0RDINATION CHEMISTRY

iabiie while the second- and third-row metais are inert? Based on the above factors, we can appreciate, ia part, why the larger metais are more apt to be inert. Because they are significantiy larger than the first-row eiements, there is iess steric hindrance among the iigands about them and therefore less tendency for a given hgand to be forced from the coordination sphere. Ia addition, these metais are often more highly charged than their lighter congenrs, ieading to stronger ML bonds that must be broken during the rate-determining step.72 Assuming a signifieant covalent contribution to the ML bond interaction (see the discussion in Chapter 4, p. 72, for further details), an additionai reason for the inertness of the second- and third-row metais can be offered. These iarger metais use 4d and 5d orbitais in their sigma bnd interactions with a iigand. These larger 4d and 5d orbitais extend farther out toward the iigand and overiap better with its orbitais. Therefore, this additional covaient contribution to the ML bond strength makes these ML (where M = second- or third-row transition metais) bonds stronger and more difficuit to break in the rate-determining step of a dissociation reaction. So now that we have a fair understanding of why the second- and third-row metais are inert, we can turn to a discussion of the reiative labilities of the first-row ions. RecaU that most of these are iabiie except for Co~ and Cr3~, whieh are inert. Why shouid this be? The key to this mystery, it turns out, iies in the change ia the crystai fieid stabiiization energy on going from the octahedrai reactant to the five-coordinate activated compiex. Tabie 5.4 shows CFSES for octahedrai and square pyramidai fieids. Most importantiy to the argument presented here is the
TABLE 5.4

Changes in crystal fieM stabilization energies

The CESEs (in units of Ar,) for octahedrai (ocO and square pyramidal (sp) fields are shown followed by the change in CFSE for the process ME6 (octabedral) , ME5 (square pyramid). A (+) indicates a gain in CFSE during the process and a () a Ioss in CESE
.

d d d2 d3

CESE ML6 (oct) 0.40 0.80 1.20

CESE ML5 (sp) 0.46 0.91 1.00

ACFSE
+ 0.06

+0.11 0.20
High.spinwealc field

Low.spinstrong field

CFSE ML6 (oct) d4 cI~

~j6

CFSE ML5
(sp)

CFSE ML6 ACFSE 0.14 0.09 0.40 +0.11 CFSE ME6 (ccl) CESE ME5 (sp) 1.00 0.91 O
(ccl)

CESE ME5
(si,)

ACESE +0.31 0 +0.06 +0.11

d7

1.60 P 2.002P 2.40 2P 1.80 P

1.46 P 1.912P 2.00 2P 1.91 P

0.60 O 0.40 0.80

0.91 O 0.46 0.91

ACFSE 0.20 + 0.31 O

d8 d9 d1

1.20 0.60 O

RATES AND MECHANISMS OF REACEIONS OF COORDINATION COMPOUNDS

107

change in CFSE shown in each case. A plus sign means there is a gain in CFSB on going from the octahedral reactanl to the square pyramidal intermediale, while a negative sign represents a loss in CFSE. Whal is lhe consequence of a gain or loss of CFSE on going from lhe oclahedral reactant to the square pyramidal transition state? It makes sense that if there is additional crystal field slabilization energy in lhe transilion slate, then its formation is favored and the rate-determining step is fasler. On the olher hand, if there is less CFSE in the transition stale than in the reactants, this would make il less stable (higher in energy) and more difficult to achieve. Therefore, the reaction would be slower. To see lhe effect of lhe change ~f crystal field stabilization energy (ACFSE) more clearly, consider the general substilution reaction shown in Equation (5.36): [CrL5X]3~ + Y
~

[CrL5Yj3~ + X

(5.36)

The reaction profile for the rate-determining step, lhe dissociation of lhe X ligand, is shown schematically in Figure 5.9. The lop curve (dashed une) shows lhe energy

Rate-determining step: [CrL5X]3~

[CrL5~~

[CrL5]3

1,
o e o e o

CFSE A 2090 kJ

[CrL5Y]3

Reaction profile FIGURE 5.9 The reaction profile for the rate-determining step of a substitution reaction of [CrL5X]3+ assuming a dissociative mechanism. The top curve (dashed tine) shows aD energy of activation of E, without considering the effect of crystal field stabilization energy (CFSE). The Iower curve (solid une) is the result of subtracting the CFSEs for thecomptex reactant, product, and transition state. lf X = L F120, Lhe energy of activation E3 is 420 Ia/moi greater in the Iower curve as the resutt of the Ioss of CESE on going from the octabedral reactant to the square pyramidai transition state.

108

C00RDINATION CHEMISTRY

changes without CFSEs. Now consider Lhe effect of subtracting the CFSEs for both the starting and transition state complexes. If ~, of Lhe octahedral reactant is about 2090 kJ (Lhe value for X = L H20), then Lhe CFSE 1.20A0 2510 kJ/mol. (See the d3 entry in Table 5.4 for CFSEs in terms of tx~ for Cr3~ complexes.) Therefore, Lhe starting complex is 2510 kJ/mol more stable due to the effect of its octahedral crystal fieid. For Lhe [Cr(H2O)5J3~ transition state, CFSE = 2090 kJ/mol, meaning that Lhe Lransition state is 2090 kJ more stable due to the effect of its square pyr~ida1 c~sLa1 field. fle resulting change in CFSE (.CFSE = 0.20~~ = 420 kJ) on going from Lhe reactant to Lhe transition state is a Ioss of 420 kJ. The iower curve (solid une) is lhe resulting reaction profile. Carefully note Lhat Lhe energy of activation of Lhe lower curve is now 420 kJ/moi higher than that shown in the top curve. The resuit is that the loss of CFSE has been added directly orno the ener~y of activation for the reaction. Can we use Lhe resuits given in Table 5.4 Lo expiam why Cr3~ and Co3~ complexes are inert? The table indicates that d3, d6 (low-spinstrong-field), and d5 have the most negative CFSES. Of Lhese cases, Lwo also occur in + 3 charged meLais, Cr3~ (d3) and Co3~ (d6). (Due Lo Lhe +3 charge, Co3~ complexes are generaIly low-spinsLrong-field.) These Lwo cases, Lhen, involve sLrong M3 L bonds and a Ioss of CFSE on going to Lhe Lransition state. Accordingly, Lhe dissociaLive raLe-determining steps involving Cr3~ and iow-spin Q~3~ have high energies of acLivaLion and are comparaLively slow. Here we have a rationale for Lhe fact LhaL Cr3~ and Co3~ compiexes are inert. WhaL abouL other metal cations LhaL have a charge of only +2 but invoive a ioss of CFSE? Figure 5.5 shows thaL V2~ (d3) and Ni2~ (d8), whilc noL considered inerL, have some of Lhe lowest rate constants for waLer exchange among Lhe labile metais. lso noLe thaL Cu2~ (d9) and Cr2~ (d4), with only +2 charges and significant gains in CFSE on going to the transition state, have among the fastesL rates of water exchange. So we see LhaL data from sLudies of the raLe of waLer exehange of various meLais correiaLe well wiLh Lhe results of Table 5.4.
= = =

5.4

REDOX OR ELECTRON-TRANSFER REACTIONS In the above substituLion reactions, the metal oxidation sLates remam consLant; no elecLrons are transferred Lo or from the metal ions. Now we Lurn our attention to reacLions in which electrons are, in fact, transferred from one metal Lo anoLher. Recali LhaL if an aLom or ion, usually a complexed metal cation in Lhe cases we are about Lo consider, loses elecLrons, iLs oxidaLion sLaLe increases and iL is said Lo be oxidized. If Lhe metal gains electrons, its oxidaLion staLe decreases and iL is reduced. (Some sLudents remember Lhese definiLions by Lhe mnemonic LEO goes GER that stands for Loses Electrons Oxidized, Gains Electrons Reduced. The Lheoretical basis of redox reacLions is more fully developed in Chapter 12, where we consider Lhe sLrength of various oxidizing and reducing agents using the concept of standard reduction poLentiais.) What is Lhe specific sequence of molecular-levei sLeps Lhat results in an eiecLron being transferred from one coordinated metal ion Lo anoLher? ThaL is, what are Lhe possible mechanisms for Lhese redox reacLions? There appear Lo be two possibiliLies, firsL delineaLed by Taube in Lhe early 1950s. In Lhe outer-sphere mechanism Lhe coordination spheres of Lhe meLais sLay inLacL, while in Lhe inner-sphere mechanism Lhey are aiLered in some way.

L~

RATES AND MECHANISMS OF REACTIONS OS COORDINATION COMPOIJNDS

109

TABLE 5.5

Some outer-sphere electron-transfer reactions

SeIf-exchange reactions k, M s~

[Mn(CN)6J3+ [Mn(CN)6]4. [Mn(CN)614+[Mn(CN)613 [IrCl6]3+[JrCI6j2. [1rCI6J2+[lrCl6]3 [Ru(NH3)6]3+ [Ru(N113)6]2 [Ru(NH3)6]2+ [Ru(NH3)613 [Fe(CN)514+[Fe(CN)613. [Fe(CN)6]3+[Fe(CN)6]4 (Ru(H2O)6J2+[Ru(H2O)5]3~ [RuQ-12O)6]3+ [Ru(H2O)512~ [Co(H2O)6J3~+[Co(H2O)6)2 [Co(H2O)612+ [Co(ll2O)6J3~ [Co(en)3]3+[Co(en)3]2. [Co(en)3j2~+[Co(en)3j3 [Co(C204)3]3+[Co(C204)314. [Co(C204)3)4
+

i0~ ia3 8 X 102 7 x 102

44 5 1 X io~ 1 x i0~ 2 X io~ 1 x 106

[Co(C204)3]3
=

[Cr(H20)613+[Cr(H20)612 [Cr(H2O)6]2~+ [Cr(HzO)613t [Co(NH3)6)3+ [Co(NH3)6]2~. [Co(NH3)6]2~+[Co(NH3)6]3 Cross reattions [Fe(CN)6]4+ [1rCIj2_~~,[Fe(CN)s]l_+[IrCI613 [Cr(H2O)6)2~+[Fe(H2O)6]3~* [Cr(H20)6]3+[Fe(H20)5]2 (CrIJ2O)6)2+[Ru(H2O)6]3~ [Cr(H2O)6]3~+[Ru(H2O)6]2~ [Ru(NI43)6]3+[V(H2Q)612~ [Ru(N~H3)6]2~+[V(H2O)6]3 [Co(NH3)5]3+[Ru(N}13)612 [Co(NJ13)5J2+[Ru(tffl3)6]3 [Co(NH3)61~+[V(H2O)6}2~. [Co(NH3)5]2~+[V(H2O)6J3~ [Co(en)3J3+W(H2O)j2~ [Co(en)3]2~+LVO~2O)613~ [Co(N113)6]3+[Cr(H20)612
,

4 X i~ 2 x ia3 2 x 102 8 x 101 1 X 10_2 4X 2 x i04 9 X i0~

[Co(NH3)6]2+ [Cr(H20)5]3
=

[Co(en)3J~~+[Cr(ll2O)j2 [Co(en)3]2~+[Cr(H2O)6]3

2 x i0~

Outer-Sphere Mechanisms To make it possible for an electron to move from one metal ion to another, it makes sense that they should be as dose to each other as possible. Assuming, for the moment, that their two coordination spheres stay intact, the distance between two metal ions will be at a minimum when the two coordination spheres are in contact. When an electron is transferred among metal ions whose intact coordina tion spheres are in contact at their outer edges, this is referred to as an outer-sphere meehanism. A variety of reactions are thought to occur via such a mechanism. Some examples with their corresponding (second-order) rate constants are shown in Table 5.5. The entries at the top of table are self-exchange reactions in which two coordinated ions, identical in every way except for the oxidation state of the metal ion, simply exchange an electron. One metal ion is oxidized, the other reduced, but no net reaction actually takes place because the products are indistinguishable from the reactants. The Iower set of reactions, labeled cross reactions, involve a transfer (or a crossing over) of an electron between different coordinated metal ions. These examples, as shown, do result in net reactions. Consider as an example for discussion the self-exehange reaction between the hexaaquoruthenium(III) and the hexaaquoruthenium(II) ions shown in Equation

--

110
(5.37): [Ru(H2oh7)6]3+ + [Ru(H2O)6j2~
.

[Ru(H2o17)6]2+ + {Ru(H20)6j3~

(5.37)

Reactions like this one are followed by labeling one of the complexes with a radioactive isotopie tracer. In this case lhe water molecules in the reactant Ru3 + complex contam lhe O~ isolope instead of lhe normal 016. This reaction has been determined to be first-order in lhe concentrations of both reactants (second-order overail) and lo have a rale constant (sce Table 5.5) of aboul 44 M~s. On the other hand, the rate of watr exchange iii [Ru(H20)612 is considerably slower (with a rale constant between 10_2 and io~~ s~ according to Figure 5.5). The rale of water exchange in [Ru(H20)6]3~ would be still slower because of the +3 charge of the central metal ion. Given lhe relative slowness of these water exchange reactions, il follows that the above reaclion cannot proceed by a mechanism involving lhe breaking of the Ru~ OH2 bonds. FAlso note lhal if the water molecules did dissociale from lheir respective calions, lhe H2017 molecules would be randomly distribuled among lhe products and not concen trated in the hexaaquoruthenium(II) producl complex.] We are left wilh an electron transfer belwecn inlacl coordination spheres as the most plausible mecha nism for this reaction. So what is the mechanism of this outer-sphere eleclron transfer? The first slep, represented in Equation (5.38), would be lhe coilision of the two reactant coordinaled ions to form an outer-sphere complex, or what is somelimes called an ion pair. Next the electron transfer, represenled in Equation (5.39), takes place instantaneously within this ion pair. Finally, as shown in Equation (5.40), lhe two product ions separate. [Ru(H2017)j3 + [Ru(H20)6j2~

[Ru(HzO17)~j3 /[Ru(H20)6]2~ (5.38) (539)

[Ru(H2O~)51~~/[Ru(H2O)612

[Ru(H2017)612+ARu(H20))3+

[Ru(H2ojj2~/[Ru(H2Oh?~

[Ru(H2o~)6j2~ + (5 40)

There is one additional poinl to be made concerning this mechanism. The electron transfer in lhe outer-sphSre complex will be extremely fast lo lhe poinl that we may view il as occurring inslanlaneOusly. (Very Iighl electrons move much fasler than lhe heavier and therefore more lumbering and cumbersome complex ions.) Bul there is a problem here. If lhe ML dislances (Ru~OH2 distances in lhe above example) are very differenl in lhe two complexes, lhere will be a large energy barrier, lhal is lo say, energy of aclivation involved in bringing about lhe electron transfer. Consequenlly, il makes sense lhal lhe grealer lhe difference in lhe ML dislances in lhe reaclanls, lhe slower lhe reaclion. Incidenlally, lhe besl mental image of lhis process is lo lhink of lhe lwo ML dislances coming lo an inlermediale poinl aI which lhe eleclron transfer lakes place. The addilional

~iner~neededldJuslEL disla:ces lo lhis intermediale valuecontribules~j

RATES AND MECHANISMS OF REAcrioNs OF COORDINATION COMPOUNDS

111

directly to the energy of activation. Now the Ru2~ OH2 and Ru3~ OH2 bond lengths are fairly similar, 2.03 and 2.12 A, respectively, so this self-exchange reaction, as we have noted, takes place fairly rapidly. What would happen if the ML bond lengths in the two complex ions were more radically different? Examples of this situation are the self-exchange reactions involving Co3~ and Co2~ ions. (Note how these reactions are concentrated toward the lower ends of both parts of Table 5.5.) Specifically, the self-exchange reaction involving the hexaammine complexes, as shown in Equation (5.41), is a case where the ML bond lengths are different enough that this reaction is very slow:

[Co(NH3)6j
=

3+

[Co(NH3)6]
=

2+
.

[Co(NH3)6]
k

2+

3+

+ [Co(NH3)5] 10.6 M s~ (5.41)

Co3~NF13 1.94

2.11

A word is in order as to why the Co~NH3 distances are so dissimilar. Octahedral Co3~ (low-spin) has an electron configuration of 4g with ali the metal electrons pointing in between the hgands. Octahedral Ca2~ has an electron configuration of ~ (if low-spin) or t~9e~ (if high-spin). lxi either case, not oniy is the charge acting on the water hgands smaller in Co2~ but the seventh electron (and the eighth if high-spin) points directiy at the ligands. Therefore, ML distances are considerably greater in Co(II) complexes. Au exampie of a rapid seif-exchange reaction in which the change in ML distances is particulariy smali is that involving hexacyanoferrate complexes as shown in Equation (5.42): [Fe(CN)6j4
Fe2~CN
=

[Fe(CN)6]3[Fe(CN)6]3 k
=

+ [Fe(CN)6]4

Fe3~CN
=

1.92

1.95

700 M s~

(5.42)

Note that in these two cations tFe2~ (low-spin): t~g; Fe3~ (low-spin): t~8] an electron is mereiy transferred among the sets of orbitais (that point between the ligands), and therefore the ML distances do not change appreciably.

Inner-Sphere Mcchanisms As we have just discussed, the coordination spheres of both reactants remam intact during an outer-sphere electron-transfer reaction. The same is not true, however, of th inner-sphere reactions we now start to consider. Inner-sphere electron-trans fer reactions invoive the formation of a bridged complex in which the two metal ions are connected by a bridging ligand that heips to promote the electron transfer. Often, but not always, the bridging ligand itself is transferred from one metal center to the other. Ligand transfer, then, is a good sign (but not absolute proof) that an mnner-sphere mechanism is in operation. Of course, if there is not an available bridging iigand, then the inner-sphere mechanism cannot be correct. li a bridging hgand is available but not transferred, either an inner- or an outer-sphere mechanism may be possibie. The first and now classic set of reactions invoiving inner-sphere electron transfer mechanisms was reported in 1953 by Taube and his group. The overali

112

COORDINATION CI-IEMISTRY

reaction is given in Equation (5.43):

ICo(NH3)5X]2~+ [Cr(H2O)6]2~+ 5H~*

[Cr(H2o)5x]2~ + [Co(a2o)6}2~ + 5NH4~

X= F, C1, Br, 1, NCS, NOZ, CN,

(5.43)

Note that Co3~ is reduced to Co2~, while Cr2~ is oxidized to Cr3~. A bridging

ligand (Xi is transferred from the cobalt coordination sphere to that of the
chromium.

As we did for the outer-sphere case, we start by showing the sequence of the steps of the mechanism. First .we note that of the two complex reactants, one, the Co(IH) complex, is inert while the other is labile. This Ieads us to postulate that the first step [shown in Equation (5.44)] is the dissociation of a water molecule
from the labile complex. The second step [Equation (5.45)] is the formation of the

bridged complex connecting the two metal ions. The third and rate-deterinining step [Equation (5.46)] is the actual electron transfer along the bridge set up by the X ligand. Fourth, as shown in Equation (5.47), the Co2~ is now the labile metal, and it dissociates the bridging Iigand forming two ~eparate complexes again. Lastly [Equation (5.48)], the ammine ligands in the Co2~ complex are protonated and replaced by waters. [Cr(H2O)6]2~
Labile

[Cr(H2O)5]2~+ H20

(5.44)

[Co(NH3)5X]2~ + [Cr(H2O)5]2~. [(NH3)5cox Cr(H20)5]

Inert

(5.45) [(NH3)5Co11X-Cr(H2o)5]~~

[(Na3)5cox Crhhl(H20)514+ (5.46)

[(NH3)5CoX Cr(H2O)5]4~
Labile Tnert

[Cr(H2o)5xJ2~

[Co(NH3)5J2~ (5.47)

+ 5H~
Labile

[Co(H2O)6]2~+ 5NH

(5.48)

There are a number of variations on this basic seheme that, taken together, lend a great deal of support to the inner-sphere or bridging-ligand meehanism. For example, if X is the chioride ion and it is isotopically labeled with C136, the label is always transferred to the chromium coordination sphere. Conversely, if the reaction is carried out in a solution containing free 36Cl ions, none of these labeled ions are found in the products. If X is S-thiocyanate, SCN, the product

RATaS AND MECHANISMS OF

REACFIONS OF COORDINATION COMPOUNDS

113

TABLE 5.6

Comparative rate constants for the inner-sphere electron-transfer reaction [Co(NH3)5X]2~+ [Cr(H2O)6]2~
X F k, M s

CI
Br I

2.5 6.0 1.4 3.0

io~ x io~
X X io~ X 106

Data taken from F. Basolo and E. G. Pearson, Mccl, anisms of Jnorganlc Reactions, A Study of Metal Co, plexes ir, Solulion, 2d ad., Wiley, New York, 1968, p. 481.

contains predominantly N-thiocyanate, NCS, as one would predict if Lhe SCN acts as a bridging ligand. (See Problem 5.50 for an opportunity to work through this mechanism.) Table 5.6 shows Lhe variation of Lhe rates of the above reaction as the X bridging ligand varies among the four halides. Can we make any onclusions about the relative abilities of Lhese ligands in facilitating the rate-determining eIectrc~n transfer step? It appears Lhat the larger the halide, the faster the reaction. Why would this be? The reason seems Lo be connected to the polarizability of the halide. (Polarizability is Lhe ease with which Lhe electron cloud of an atom, molecule, or ion can be distorted so as to set up a dipole moment. Generally, large and diffuse species whose electron clouds are held relatively weakly by their nuclear charges are more polarizable than small and compact ones. This property is discussed in somewhat greater detail in Chapter 9, pp. 225226.) Once the bridged intermediate is formed as shown in Figure 5.10, the halide can be polarized by the more highly charged Co3~ cation. The induced dipole in Lhe halide Lhen attracts Lhe electron from Lhe Cr2~ and facilitates Lhe electron transfer. It follows that the larger and more polarizable Lhe halide is, the greater Lhe dipole moment that can be induced in iL, and Lhe easier Lhe elecLron Lransfer.

H20

H3N

OH2

FIGURE 5.10 The electron-transfer step in the inner-sphere complex, KNH3)5Co1 X Cr(1120)5]. The more highly charged Co3~ polarizes the bridging ligand X. The resulting induced dipole of X helps to facilitate the transfer of an electron from the Cr2+ to Lhe Co3h

114

cooarnNATloN CHEMISTRY

5.5 SUBSTITUTION REACTIONS IN SQUARE PLANAR COMPLEXES: THE KINETIC TRANS EFFECT A general substitution reaction of square planar complexes is shown in Equation (5.49): +L (59) L L L L Whereas, as we discussed in Section 5.3, the overwhelming bulk of evidence for substitution reactions of octahedral complexes indicates that they most often proceed via a dissociative mechanism, most of the evidence gathered in working with square planar complexes indicates that they are substituted via an associative pathway. This difference is not particuiarly surprising. After ali, a square planar complex has two places (above and below the plane of the molecule) where an incoming ligand can readily attack the metal to form a five-coordinate inter mediate. As we have seen, square planar complexe~ usually occur with the d8 metais such as Pt(II), Pd(II), Ni(II), and Au(III). The platinum(H) complexes are particu iariy inert and, because these rather siow reactions can be foilowd by fairly traditional and straightforward methods, have been extensively studied and ana lyzed. A particular feature of these substitution reactions is the kinetic trans effect, defined as the relationship between the rate of substitution of square planar complexes and the nature of the species trans to the ligand being displaced. To understand this effect more clearly, consider the general substitution reactions given in Figure 5.11. The nonlabile ligands A and B can be ranked in order of their abilities to dispiace the ligands trans to them. For example if the nonlabile A iigand is a better trans director than B, then reaction 1 occurs. Conversely, if B is the better transdirector, then the ligand trans to B is preferentially displaced and reaction II occurs. After a large number of comparative experiments have been carried out, a trans-directing series, such as the one shown below, can be constructed: CN CO > N02> 1, SCN> 13r> Cl> py NH3 > H20 Such a series, although empirically derived, can be most useful in preparing high-purity isomers of square planar complexes. To demonstrate this, suppose we start with the tetrachloroplatinate(ll) ion, Pt042, and examine some of its typical substitution reactions. First, consider two additional empirical facts. (1) It is generally easier to replace a chloride ligand bound to Pt(II) than it is to replace other hgands. That is, other ligands will most always displace chlorides, and the only way to carry out the reverse is to swamp the system with a large excess of Cl

LN~L

+X*

L~~X

A~ /
Reaction 1 +E/

~L \
Reaction II

FJGURE 5.11 The kinetic trans effect. An entering Iigand (E) can dispiace either of two Ieaving Iigands (L) depending UPOfl the trans-directing isability of the nonlabile Iigands (A and B). II A a better trans director than B, reaction 1 occurs. 1f B is the better trans director, reaction II occurs.

/
A B + L L A~ E B~ ~L + L

RATES AND MECHANISMS OF REAaI0N5 OF COORDINATION COMPDIJNDS

115

CI%%..pt~_NH31i~. C~NH3

c(
C .Ci

CI

cK~
(Note: C1

~NO2

ci~~~tN:i

is higher in the trans-directing series than NH3) ~NO~

CNO2 pu-3 C CV ~CI

N4 t%~~
(Note: NO~ is higher in the trans-directing seriesthanCF)

FIGURE 5.12
Cis- and trans.amminedichloronitrOplatifliUm(II) can be prepared in higb purity by varying Lhe order in which the ammine and nitrite Iigands are added Lo tetrachloropiatinate(Ifl. Charges on the various complexes are omitted for clarity.

(an example of Le Chatehers principie) as indicated by the unequai arrows in Equation (5.50: [PtCi4]2 + L~[PtCi3Lf + Ci (5.50)

(2) In a substitution reaction in which there is more than one possibiiity as to which chioride will be repiaced, the trans-directing series is used to predict which possibiiity wili be realized. For example, suppose we wish to prepare eis- and trans-ammine dichioronitroplatinate(II). How wouid we proeeed? The procedure is outlined in Figure 5.12. Jf we start with the tetrachloroplatinate(ll) and react it first with anunonia and then with nitrite, the cis isomer is prepared in high purity. Note that when amminetrichioropiatinate(II) is treated with nitrite, a chloride trans to another chioride is replaced preferentialiy to the chioride trans to ammine. This takes advantage of the fact that chloride is higher in the trans-directing series than is ammonia. On the other hand, when trichioronitroplatinate(JI) is treated with ammonia, the chioride trans to the nitro iigand is repiaced preferentialiy. Here, the nitrite ion is higher in the series than is chioride. How can we rationallze the kinetic trans effect? One stili popular but only partial explanation was devised by A. Grinberg in the 1930s. It maintains that the trans-directing ability of a Iigand is related to its polarizability. As noted above in the section on inner-sphere electron-transfer reactions, polarizability can be thought of as the ease with which a dipole moment can be induced in a species. Figure 5.13 iliustrates the poiarization theory. The first step is the induction of an instantaneous dipole in the trans-directing ligand by the platinum cation. Second, this latter dipole in turn induces a dipole in the iarge, polarizabie platinum cation. The Pt2~ C1 bond is somewhat weakened by the repuision between the negatively charged ligand and the negative end of the dipole induced in the cation. Therefore, the chioride trans to A is preferentially replaced. Support for this theory is demonstrated by looking at the trans-directing series. The more po~ariz able ligands, such as SCN and 1 and the ligands containing ir clouds (for example, CO and CN~j, are high in the series, whiie the less polarizable ligands

116

COORDINATION CHEMISTRY

trans-directing ligand

(a)

(1,)

(c)

FiGURE 5.13 The polarization theory for explaining lhe kinetic trans effecl in sguare planar platinum(II) complexes. (a) The platinumQj) cation induces a dipole in lhe polarjzabje trans-direcling Iigand A. (b) The induced dipole in Iigand A induces a dipole iii the polarizable Pt2~ calion. (c) The chioride anion Irans to A is more easily released due to the extra repulsive forces belween ils negative charge and lhe induced dipole of lhe platinum(ll) calion.

such as ammonia and water are lower in the series. Additional support comes from the observation that platinum complexes demonstrate a more pronounced trans effect than those of the less polarizable palladium(JJ) and nickel(II) cations.

SUMMARY
This chapter starts with a brief survey of five types of reactions of coordination compounds: (1) substitution, (2) dissociation, (3) addition, (4) redox or electron transfer, and (5) reactions of coordinated ligands. The tendency for a coordination compound to undergo a substitution reaction is tabuiated by means of stepwise and overail equilibrium constants. Aquation and anation reaetions are special examples of substitution reactions. Addition and dissociation reactions involve adding or removing ligands from the coordination sphere, respectively, while

RATES AND MECDHANISMS OF REACIIONS OF COORDINATION COMPOUNDS

117

reduction and oxidation reactions involve adding ar removing electrons from the metal. Oxidative-addition and reductive-elimination reactions, each combining two of the above reaction types, are logical opposites of each other. Reactions aI coordinated iigands occur without breaking or forming any metai-ligand bonds. Labile and ineri are kinetic terms that classify coordination compounds by how fast they react. Labile compounds react quickly and inert compounds slowly. These kinetic terms should not be confused with the thermodynamic terms stable and unstable. Compounds can be thermodynamicaliy unstable but kineticaliy inert or, conversely, stable but labile. Complexes aI the first-row transition metais, with the exception of Cr3 + and Co3 ~, are generally iabile, while coordination com pounds of second- and third-row transitjon metal ions are inert. There are two possible starting points for the study of the mechanisms of octahedral substitution reactions. The rate-determining step of the dissociative (D) mchanism involves the breaking of an M L bond to form a five-coordinate intermediate which then rapidly adds a new hgand. In the associative () mecha nism, the rate-determining step is the coilision of the new ligand with the octahedral reactant to produce a seven-coordinate intermediate from which an original Iigand is then rapidly expelled. The D mechanism, which predicts the rate of the reaction to be first-order only in the original coordination compound, should be clearly distinguishable from an mechanism that is consistent with a rate law which is first-order in both the reactants. There are, however, two experimental complications that blur the above distinctions. First, the D and mechanisms turn out to be idealized and oversimplified extremes. Reactions more likeiy proceed by an interchange (1) mechanism in which bond brealdng and forming happen nearly simultaneousiy. If bond breaking is favored (even if only slightly), the mechanism is termed d (interchange-dissociative), whiie if bond making is favored, an a (interchange associative) is indicated. These last two terms are particularly favored II no five- ar seven-coordinate intermediates, respectively, can be isolated. The second experi mental complication involves the masking of the rate dependence on the concen tration of the incoming ligand. This masking effect is comnion, particularly in reactions carried out in aqueous solution. Despite these complications, many pieces of evidence (drawn from the analysis of water exchange, anation, and aquation reactions) favor a dissociative mechanism. First, rates of water exchange are consistent with a dissociative mechanism in which the rate-determining step is the breaking of the M OH2 bond. Aniong other factors, this bnd-breaking step is dependent upon the charge density of the metal ion. The larger the charge density, the stronger the M0H2 bond and the slower the bond-severing step. The rates of anation reactions of hydrated complexes are consistent with the breaking of the M OH2 bond as the initial and rate-determining step. Finally, the rates of aquation reactions are dependent upon the strength of the bond between the metal and ligand being replaced. The rates of aquation reactions vary considerably with the identity of the displaced iigand, the buik of the inert ligands remaining in the coordination sphere, and the overail charge of the complex. With the dissociation mechanism firmly established for octahedral substitu tion reactions, an explanation of the labiity of complexes can be formuiated in terms of the energy of activation of the bond-breaking step. The size and charge of a metal ion as well as any changes in the crystal field stabiiization energy (CFSE) upon bond dissociation influence the energy of activation. Using these parameters, the trends in the iabiiity of coordination compounds can be rationaiized. Specifi

118

COORDINATION CHEMJSTRY

cally, complexes of Cr3~ (d3) and Co3~ (d6), are inert because of their high charge and signfficant Ioss of CFSE upon bond dissociatiOn. Menor Taube formulated two principal mechanisms for electron transfer reactions: outer-sphere and inner-sphere. A variety of self-exchange reactions and cross reactions occur by an outer-sphere mechanism. After the formation of an ion-pair composed of the intact coordination spheres of the two reacting metais, an electron is transferred and the ion pair dissociated. The faster outer-sphere redox reactions are those whose ML distances need only be adjusted slightly in the course of the reaction. An inner-sphere mechanism involves the formation of a bridged complex in which the bridging ligand facilitates the electron transfer and is often transferred from one metal to the other. The more polarizable the potential bridging ligand, the faster the reaction. Substitution reactions in square planar complexes take place via associative mechanisms. In the kinetic trans effect, a series can be constructed in which species are put in order of their abiiity to labilize ligands trans to themselves. Using this series, high-purity square planar isomers can be readily synthesized. The trans-directing ability of a ligand is also directly reiated to its polarizability. PKOBLEMS 5.1. Demonstrate that the product of K1 to K4 for the stepwise repiacement of waters in ~Cu(M2O)4]2~ by ammonias [given in Equations (5.4) to (5.7)] results in the expres sion for /34 given in Equation (5 8) 5.2. The water Jigands in the hexaaquonickelOl) cation can be replaced by ammonia. (a) Show chemical equations and write rnass action expressions for the stepwise replacement of each water with one ammonia. (b) Write an equation for the overali replacement of the six water Iigands by six ammonia ligands and then demonstrate that j~6 the overali equilibrium constaflt, is given by the product of the six stepwise constants. 5.3. Give the stepwise and overali equilibrium constants for the reaction Cr(CO)6 + 3PPh3 Cr(CO)3(PPh3)3 + 3C0

5.4. Give the stepwise and overail equilibrium constants for the reaction 3+ 3+ [Fe(1120)&j + 3en [Fe(en)3] + 6H20

5.5. If you were to write the mass action expression to represent the reaction of dilute aqueous acetic acid flC211302(aq), with water to yieid aqueous acetate and hydro nium ions, would water generally be included? Why or why not? Where would the value of [H20) be found? 5.6. Write the mass action expression for the equilibrium constant of the weak base ammonia, NH3, in aqueous solution corresponding to the following reaction NH3+H2O~M~4~0~ 5.7. C1assi~ each of the following reactions. More than one of the five basic classifications may apply. (a) W(CO)6 W(CO)5 + co (b) [Co(CN)5?+ 1 [Co(CN)51]3 (c) [PdCl4]2+ c2114 [PdC13(C2H4)1+ cr (cl) jrBr(CO)(PPh3)2 + lEr IrBr2(CO)H(PPh3)2 (e) [Co(NH3)5C1]24 +[Cr(H2O)6]2~ . [Co(NH3)s(}I2O)]2~+ [Cr(H2O)sCI12~ 5.8. Classify each of the following reactions. More than one of the five basic classifica tions may apply.
.~

:1

(a) ~Co(NH3)5NOi]24 + [Ru(NH3)6]2~ (b) [PtCl6r jPtCI4]2+ C12

[Co(NH3)5NO3V+ [Ru(N~3)j3~

RATES AND MECHANISMS OF REACFIONS OP COORDINATION COMPOUNDS

119

(e) [Cr(H2O)6}3~+ ~ , [Cr(H2O)5OH]2~+ H3O~ (d) Cr(CO)6 + 3PPh3 , Cr(CO),(PPh3)3 + 3C0 (e) Zn(NH3)r+ 2NH, , Zn(NF13)~~ 5.9. In Equation (5.19) (repeated below), suppose that the hydroxide reactant were to be labeled with o7, where in the products would the tabel end up? Briefiy expiam your answer. [Cr(H2O)6]3~+ 0H [Cr(H2O)5(OH)]2~+ H20 5.10. In Equation (5.20) (repeated below), suppose that the oxygen atoms of the carbonate Jigand were labeled with o~, where in the products would the labeis end up? Briefly expiam your answer. [Co(NH3)5C03]
+

+ 2H3O~ [co(NH3)5(H,O)j2~ + 2H2o +

Co2

5.11. Briefly speculate on why many cobatt(IIJ) compounds are prepared by oxidizing CoQJ) salts rather than by substituting Iigands mn cobaltO!!) complexos. 5.12. Briefly speculate on why many chromiumOlJ) compounds are prepared by reducing chromates and dichromates rather than by substituting ligands in chromiumOJl) complexes. 5.13. Classify the following complex ions as inert or labile: (a) [Co(NH3)6]2~

(b) [Co(NH,)5NO21~
(e) [Co16]3

(d) [Fe(H2O)5(NCS)]2~
(e) [Ni(en)3]2~ (f) [IrCI6]3 Classify the following complex ions as inert or labile: (a) [Ru(NH3)5py]3~ (b) [Cr(acac)3] (e) [Cr(en)3]C12 (d) [Mo(CN)6]3 (e) [V(H2O)6~3~ (f) [MnF6]3 Of the following cyanide complexes, [Ni(CN)4]2, [Mn(CN)6]3, and [Cr(CN)613, which one wouid you most expect to be (a) labile, (b) inert? Briefiy justify your answer. In terms of Figure 5.2, would you expect the concentration of ML5 to be reiatively low or high (compared to the concentrations of the reactants and products) through out the reaction? Briefly expiam your answer in terms of its relative production and consumption during the reaction. Suppose that the rate-determining step in Figure 5.2 was the second step in which the intermediate ML5 and Y come together to form the product. In this case, what rate law is predicted? Suppose that the rate.determining step in Figure 5.3 was the second step in which the seven-coordinate intermediate breaks apart to form ML5Y and X. In this case, what rate law is predicted? Suppose that the rate-determining step in Figure 5.3 was the second step in which the seven-coordinate intermediate breaks apart to ML5Y and X. Would you expect the concentration of ML5XY to be relative low or high (compared to the concentra tions of the reactants and products)? Briefly expiam your answer in terms of its relativo production and consumption during the reaction.

5.14.

5,15.

5.16.

5.17.

5.18.

5.19.

5.20. Draw a diagram showing an interchange (1) mechanism in the following three stnges: (a) The resulting situation after an incoming Y ligand has taken its place just outside the coordination sphere of the starting ML5X complex

__-

120

COORDINATION CHEMISTRY

(b) The midpoint of the interchange between the attacking Y and the leaving X ligand when the M X and M Y bonds are approximately at half strength

(c) The situation after the MX bond has been fully replaced by the MY bond

5.21.

5.22.

5,23.

5.24.

5.25.

but the X ligand now resides just outside the coordination sphere of the product ML5Y complex. Assuming no experimental complications, expiam, ia your own words, some ways in which one might expect to differentiate between an associative (A) and dissociative (D) mechanism for the substitution of octahedral complexes. Suppose the exchange of water reactions of Figure 5.5 were carried out by an associative mechanism rather than a dissociative mechanism. Would you expect, in this case, the rates to increase or decrease with (a) increasing charge of the metal or (b) increasing size of the metal? Briefly expIam your answer. In one well-written paragraph, summarize the evidence that favors a dissociative or interchange-dissociative mechanism for the substitution of octahedral coordination compounds. The anation of [Co(H2OXCN)5]2 in which the water ligand is replaced by a variety of anions (X) has been shown to proceed by a dissociative (D) mechanism. (In this case the five-coordinate intermediate has been identified.) Write the steps for this mechanism. Do you expect the rates of these reactions to depend upon the identity of the X ligand? Why or why not? The rate constants for some anations of [Cr(NH3)5FI2O}3~ are shown below. Are these data consistent with a dissociative or associative mechanism? Write an equa tion for the rate-determining step of these reactions and briefly justi~ your answer. [Cr(NH3)5H2O]3~+ L
L

ICr(NH3)sL12~+ H20
logk

k,Ms 4.2 x 0.7 x

NCS

cr Br~ CF3C00

ir4 i0~ 3.7 x ir4 1.4 x i0~

3.4 4.1
3A

3.9

Data from the work of O. Thusius, Inorg. Citem., 10:1106 (1971).

5.26. The rate constants (at 45C) for the anation of [Co(NH3)5H2O]3~ with three anions are shown below. Are these data coilsistent with a dissociative or associative inechanism? Write an equation for the rate-determining step of these reactions and briefly justify your answer.
[Co(NH3)5F120]3 + L~

[Co(NH3)5L1~

H20

L~ NCS

k,s

logk 4.8

1.6 x iO~

cr

2.1 x iO~
2.4>< 1O~

4.7
4.6

Sor

From R. G. Wilkins, lhe Study of Kinettcs and Meclwnism of Reactions of T,ansttion Metal Complexes, AlIyn and Bacon, Boston, 1974, p. 188.

RATES AND MECHANISMS OF REAaloNs OF COORDINATION COMPOUNDS

121

5.27. The rate constant for the exchange of the water ligand in [Co(NH3)5H2O]3~ is 1.0 x i0~ s~ at 45C. Compare this value with the data shown in Problem 5.26. What conelusions can be drawn from the comparison? 5.28. The rate constants at the sarne temperature for the anation of [Cr(H20)6]3~ with a series of four anions are shown below. (a) Are these data consistent with a dissociative or associative rnechanism? (b) Write an equation for the rate-determining step of these reactions and briefly justify your answer. (e) Rationalize the trends in the rate constant data. [Cr(H2O)6j3~ + L
L

[Cr(H2O)5L]2~
Iogk

H20

k,M_LS_t 1.8 x 7.3 x 2.9 X 8.0 X

NCS
N03

io6 io~
iO_8 10_lo

5.7
6.1 7.5 9.1

Cr
1

From the work of 8. T. D. Lo and O. W. Waue, Aust. J. Chem., 28:491, 501(1975). *5.29.

The rate of exchange of the water ligand in [Cr(H20)6]3~ is 3 x 106 s~ at 25C. Compare this value with the data shown in Problem 5.28. What conclusions can be drawn from the comparison? 5.30. For reactions of the general type [Co(N}~3)5H20j3~
+

~ [Co(NH3)5Lj2~ 1120
=

demonstrate that log K0 is proportional to ~ Recail that AO AH TAS and that AG = RT In K 2.3RT log K. First determine the relationship between AH and the ML bond strength (DM_L) by analyzing what bonds are broken and formed in this reaction. Assume that AS is approximately constant for various reactions of this type. 5.31. The rate constants at 25C for the aquation of complexes of general formula [CoCI(LXen)2] are given ia the following table. Are these data consistent with an associative or dissociative mechanism? Briefiy rationalize your answer.
= =

cis.[Co(L)Cl(en)21

1120

~--

cis-[Co(H20)Cl(en)2]2~
logk

+ L

0i-1
C1

1.2 X i0_2 2.4 )< i0~

1.9 3.6

NI-13

x iC7

6.3

From lhe work of M. L Tobe et ai.; for exam pie, see Sei. Prog., 48:484 (1960). 5.32.

Compare and rationalize the rate constants for the aquation reactions of the following trans-cobalt(III) complexes. [Co(NH3)Br(en)2]2~ + 1120 [CoBr2(en)2] + H20
+

k=.12X106S~

[Co(NH3)(Hz0)(en)2]3~ + Br [Co(H20)Br(en)212~ + Br

k=1.4X

i04 C1

Barycent 122

COORDINATJON CHEMISTRY

5.33. In your own words, write a concise paragraph explaining the relationship bet-ween the rates of substitution reactions and changes in the crystal field stabilization energy. Assume that the dissociation of a ligand from an octahedral reactant to form a five-coordinate interniediate is the rate-determining step. 5.34. The splitting diagram for a square pyramidal crystal field is given below. Verify the change in crystal field stabilization energy given in Table 5.4 for a metal ion with a electronie configuration.

+O.914A

+O.O~6~ O.086A O.457,~

5.35. Using the square pyramidal crystal fleld splitting diagram given in Problem 5.34, verify the change in the CFSE given in Table 5.4 for a metal ion with a low-spin d6 electronic configuration. 5.36. The splitting diagram for a trigonal bipyramidal crystal field is given below. Calculate the change in CFSE for a d3 metal ion. Compare your result with that given in Table 5.4. Given your result, would you conclude that these ions are kinetically inert? d~2 O.707

E
-

Barycenter
d~2~~

d~~ O.082A

d~~ d~~ O.272A 5.37. Using the splitting diagram for a trigonal bipyramidal crystal fleld given in Problem 5.36, calculate the change in CFSE for an M2 d8 ion when the field changes from octahedral to trigonal bipyramidal. (Assume high-spin states.) (a) Would you conclude that these ions are kinetically inert? (b) Would you also conclude that a trigonal bipyramidal transition state is favored over one that is square pyramidal? Why or why not? 5.38. The Fe3~ and V3~ ions, while +3 charged like Cr34 and Co3~, are not classified as inert. Briefiy expIam why not. 5.39. Octahedral Ni2~ complex ions, while experiencing reductions in crystal field stabi lization energies upon losing ligands to form square pyramidal transition states, are not classified as inert. Briefly expiam why not. *5.40. Although octahedral Cu3 complexes would most likely be inert, not many have been observed. (a) ExpIam why such compounds would be expected to be inert. (b) ExpIam why not many such compounds are known. 5.41. Both [Fe(CN)6]4 and [IrCI6]2 exchange their Iigands rather slowly, yet the cross reaction between them (shown below) occurs veiy rapidly (k 4 x iO~ M r1). Propose a mechanism to account for this observation.
=

[Fe(CN)6]4

[JrCI6]2

[Fe(CN)6]3

[IrCI6j3

RATES

AND MECHANISMS OF REACrI0NS OF COORDINATION COMPOUNOS

123

5.42. Discuss the probable mechanism of the following reaction: [co(NH3)6]3~ + [Ru(NH3)612+

[Co(NH3)6j2~ + [Ru(NH3)6]3~

~..

5.43.

5.44.
5.45. 5.46.

5.47.

(ft1~9 5.48.

5.49.

5.50.

5.51.

Also speculate why this reaction is rather slow (k = 1 x 10_2 M1 s~1). (Hint: Examine the d-orbital occupancy during the reaction.) The self-exchange electron-transfer reaction between [Co(en)3]3~ and [Co(en)3J2~ is rather slow (k i0~). Expiam this observation on the basis of d-brbital occupa tions. Are the units of the rato constants in Table 5.5 consistent with an overail second-order reaction? Demonstrate your answer cleariy. Given the second-order nature of outer-sphere electron-transfer reactions, speculate on a rate-determining step for these reactions. The ML distances in tris(ethylenediamine) complexos of Ru(II) and Ru(III) are 2.12 and 2.10 A, respectively. Speculate on the mechanism and rato of the self exchange electron-transfer reaction between these two complexos. As part of your answer write an equation representing the reaction. In addition to the energy involved in readjusting the M L bond lengths in the coordinated metal ions that serve as reactants, ther~ are several other contributions to the energy of activation of an outer-sphere redox reaction. Speculate on what these contributions might involve. For each of the foliowing olectron-transfer reactions, speculate whether the mecha nism is outer-sphere or inner-sphere: (a) [IrCl6]2+[W(CN)3]4. [IrCl6]3+[W(CN)8]3 (b) [Co(NH3)5CN]2~+ [Cr(U2O)6]2~ [Cr(H2O)5NC]2~ + [Co(NH3)5(H2O)J2~ (c) [* Cr(1-12O)6]2~ + [Cr(H2O)5F]2~ [* Cr(H20)5F]2+ [Cr(H2O)6]2~ Write out the mechanism for the inner-sphero electron-transfer reaction between [Co(NH3)5SCN]2~ and [Cr(H20)6]2 that produces the pentaaquo-N-thiocyanato chromium(III) cation as one of its products. While the N-thiocyanato form of the [Cr(H20)5NCS]2 is the principal product (approximately 70 percent) of the above inner-sphere electron transfer (seo Problem 5.49), the S-thiocyanato linkage isonier also forms. Write a mechanism that shows how the S-linked isomer might be formed. For reactions of the general type [* Cr(H20)6] 2+ + [CrI-120)5X]
2+,

[* Cr(H2O)5Xj2~ + [Cr(H20)6] 2-1-

rate constants increase in the order X= F, cr, Br. Write out the mechanism for the above reaction and discuss the trend in the rato constant data in the light of your proposed mechanism. .4 5.52. Givon that the substitution of squaro planar complexos occurs by an associativo mechanism, speculate (a) on the order of the reaction with respect to the various reactants and (b) on the dependence of the rato of these reactions on the steric bulk of the nonreacting Jigands of the complex, the bulk and charge of the entering / ligands, and the overali charge carried by the reactant complex. 5.53. Using the kinotic trans effect, show how the three possible geometric isomers of amminebromochloro(pyridine)platinum(II) can be prepared. 5.54. Cispiatin, cis-diamminedichloroplatinum(II), is an extremely potent antitumor agent. Show how it can be prepared in high purity to the exclusion of the trans geometric isomer by employing the kinetic trans effect. 5.55. Given the polarization theory for explaining the lcinetic trans effect, where would you speculate that the hydroxide ion, 0H, might show up in the trans-directing series? Justify your answer. 5.56. Given the polarization theory for explaining the kinetic trans effect, where would you speculate that the organic sulfides, R2S, might show up in the trans-directing series? Justify your answer.

--

124

COORDINATION CREMISTRY

5.57. Predict the products of the following reactions: (a) [Pt(CO)C13h+ py

c ci ~Pt( + Nfl3 H3N NO2 (c) [PtCI3SCN]2+ H20

-

(b)

(d) [PtCI3CN]2+ NH3