converted to dehydrosemiascorbate during turnover.7 In vivo, DM is mostly located within chromaffin granules, which are neurosecretory vesicles of the adrenal gland.6 Both membrane- associated and soluble forms of DM are found in these granules; most spectroscopic and mechan- istic data for DM have come from the soluble bovine enzyme. Both these forms have a quaternary structure derived from proteindimers, inwhich monomers ofMr75 kDa are covalently linked by a disulfide bridge.8These dimers inturnpartially associate further into tetramers insolution, ina pH-dependent equilibrium; the tetrameric form is likely to be dominantin vivo.9,10Interestingly, the dimeric and tetrameric forms have different kinetic properties, for reasons that are unclear.10 Each proteinmonomer of DM contains 2 mol equivalents of copper.11
Peptidylglycine-amidating enzyme (PAM, EC 1.14.17.3) is also found in the nervous systems of higher organisms.4,12 The enzyme catalyzes the cleavage of the terminal acetate group from C-terminus glycyl peptides, through a two-step process that is accomplished at two different active sites withinthe same protein. First, a peptidylglycine-hydroxylating monooxygenase (PHM, EC 1.4.17.3) domainwithinthe proteinmonooxygenates the methylene carbonatom of the acetate group (Equation(4)).4,12 Then, the resultant carbinolamide is cleaved by a different, peptidylamidoglycolate lyase (PAL, EC 4.3.2.5) domainto give the final product (Equation(6)).13
CO2H N
H R
O OH
CO2H O NH2
R O
+ ð6ị
As for DM, the electrons required to accomplish the PHM reaction (Equation(4)) are provided by two equivalents of ascorbate. Multiple forms of PAM are produced in vivo, with Mrof between40 kDa and 100 kDa.14The soluble form of the enzyme, which hasMrẳ75 kDa, has beenthe most studied. The catalytically active PHM and PAL domains of PAM canbe produced separately from this protein, either by cleaving intact PAM or by using clones that produce only the relevant half of the polypeptide.15,16 The PHM polypeptide contains two mole equivalents of copper;17 however, PAL is a zinc enzyme,13and so is not discussed further here.
The PHM domaininpeptidylglycine-amidating enzyme shows 28% sequence identity to part of the amino acid sequence in DM.18As described below, the spectroscopic and redox properties of the copper ions in DM and PHM are also highly similar. Hence, although crystallographic data are only available for PHM, the active site structures and catalytic mechanisms of DM an d PHM are assumed to be similar. For this reason, they will be discussed together.
A crystal structure of the ‘‘catalytic core’’ (i.e., the part of the amino acid sequence that contains the two copper centers and the substrate binding site) of oxidized rat PHM at 2.3 A˚
resolutionhas beenpublished.19,20 This catalytic core gives indistinguishable EPR and copper EXAFS data to the complete PAM protein,15so that the crystallographic structures of the copper sites inthe truncated proteinshould be the same as inthe complete enzyme. The peptide is divided into two nine-stranded-sandwich domains of about 150 residues each. These domains are separated by a hydrophobic cleft, which has anaverage width of 8 A˚. The two copper ions lie oneither side of this cleft, and are separated by 11 A˚ (Figure 1). One of these, termed ‘‘CuH’’ or
‘‘CuA’’ inthe literature, is aT-shaped complex with three histidine ligands. The other copper ion, referred to as ‘‘CuM’’ or ‘‘CuB,’’ has a trigonal pyramidal structure with two histidine and one water ligands in its basal plane, and a weak interaction of 2.68 A˚ to an apical methionine side- chain. Interestingly, the protein conformation in the crystal, and coordination geometries of CuH
and CuM, are barely perturbed following reduction of the copper ions with ascorbate.20The only notable structural changes between reduced and oxidized crystals are: contraction of one of the CuN bonds at CuH upon reduction, so that the reduced center has a slightly more planar geometry; and a small movement of the water molecule coordinated to CuM. Finally, a 2.1 A˚
structure was obtained from a crystal of oxidized PHM catalytic core incubated with the substrate analogueN--acetyl-3,5-diiodotyrosylglycine.20The substrate molecule is bound within the inter- domain cleft, positioned so that the reactive methylene carbon atom is 4.3 A˚ from CuM, an d oriented towards the CuMwater ligand (Figure 2). This is a suitable distance to allow the reactive CH bond to interact with a peroxo or hydroperoxo ligand coordinated to CuM.
Interestingly, EXAFS measurements on DM and PHM show some differences from these crystallographic structures. The oxidized proteins show 2–3 histidine plus 1–2 other O/N ligands per copper center, all at a distance of 1.97 A˚; no S scatterer was detected in this state.15,16,21–25
EXAFS data onascorbate-reduced PHM showed one fewer O/N scatterers per dicopper protein
compared to oxidized PHM, together with 0.5 of a S/Cl scatterer at 2.27 A˚.16,22,24,25This implies a substantial shortening of the CuMS bond following reduction, which is not observed in the crystal. An identical result was obtained from the EXAFS of reduced half-apo DM and PHM, inwhich the copper ionhas beenselectively removed from the CuHsite.26,27It has beensuggested that these different crystallographic and EXAFS structures may reflect the different pHs at which the two series of measurements were carried out.20
Oxidized DM and PHM afford only one EPR signal, suggesting that the EPR properties of CuHand CuMare very similar.28This ‘‘averaged’’ spectrum is typical of a magnetically isolated type 2 copper(II) biosite but has varied slightly betweenproteins, with (gj jẳ2.25–2.29, g?ẳ2.05, Aj j{63,65Cu}ẳ142–162 G).15,16,28–33 ESEEM spectra of oxidized DM allowed the detectionof histidine ligation to copper,34consistent with the crystallographic data.19,20Treatment with ascor- bate causes reductionof both copper ions;29,30,35redox titrations have shown that CuHand CuMof DM have very similar midpoint potentials, of þ35040 mV vs. nHe.29,30Both cyanide and azide can bind to both copper centers in oxidized DM, although one copper site has a much greater affinity for these ions than the other.33,36However, inreduced DM and PHM only CuM26,27can bind carbon monoxide16,37–39 or isocyanides.40,41 Both these adducts show IR spectra that are typical of terminal metal CO or RNC complexes,26,27,37–41 while EXAFS spectra show that the methionine ligand is not displaced from the copper ion upon binding of these molecules.26,27,38–40
The increased affinity of CuMfor small molecules compared to CuHis consistent with the crystal- lographic conclusion, that CuMis the site of O2binding during catalysis.20However, the situation is complicated by the observation that, upon incubation of PHM with a substrate analog, CuH
also becomes accessible to CO.39 This is discussed inmore detail below.
Reductionof oxidized CuHand CuM by ascorbate is rapid, and yields catalytically competent enzyme that can bind O2and organic substrate.42,43For PHM, anequilibrium ordered mechanism
Figure 2 Binding ofN--acetyl-3,5-diiodotyrosylglycine to the active site of PHM, as revealed by protein X-ray crystallography.20
Figure 1 Structures of the copper centers in the catalytic core of PHM, as determined by protein X-ray crystallography.19,20
was demonstrated, in which substrate binding to prereduced enzyme is followed by binding of O2.44 This contrasts with DM, where binding of substrate and O2to the reduced enzyme can occur ina random order.45 The rate-determining step of DM catalysis is hydroxylationof the substrate CH bon d.45 For PHM, substrate hydroxylationor product dissociationcanbe rate- limiting, depending on whether the PAL lyase domain is present in the protein.43,44 DM and PHM yield very similar intrinsic primary2H isotope effects of 10.6–10.9,43,46and-secondary
2H isotope effects of 1.2,44,47 for substrate hydroxylation. For PHM, these two parameters are only slightly temperature dependent,48 which is characteristic of non-classical kinetic behavior.
These data imply that CH cleavage occurs by a hydrogenatom transfer step, which has a substantial quantum mechanical tunneling component. A linear free-energy relationship with ẳ–1.5 was derived from hydroxylation of substituted phenylethylamine substrates by DM, which is again typical of a radical-generating mechanism.47 A free-energy analysis of individual steps of these reactions implied that the final step of the catalytic cycle is product dissociation from an inner-sphere complex to copper.45,47Finally, several mechanism-based inhibition studies have also indicated a free radical CH oxygenation mechanism.49–58
The CH hydroxylationstep inthe DM catalytic cycle involves uptake of one proton per turnover.45This was used to infer the intermediacy of a copper hydroperoxide complex species in the hydroxylation reaction, generated by protonation of a bound dioxygen species. An intrinsic
18O kinetic isotope effect of 1.025 has been measured for monooxygenation of dopamine by DM;59importantly, the18O isotope effect for hydroxylation of a series of phenylethylamines by DM increases as the reactivity of the substrates decreases.59These data suggest that OO bond cleavage occursbeforeoxygen atom insertion into the substrate. This OO cleavage event is not promoted by further reduction of the enzyme: peroxo complex by exogenous electron donors.7To accommodate these results, it was suggested that a tyrosine residue might donate an electron to a copper(II) hydroperoxo species, thus effecting OO cleavage (Figure 3).59No tyrosyl radical has ever beendirectly detected inDM or PHM under any conditions. However, treatment of oxidized DM with hydrogen peroxide yields a protein-bound quinone residue, which could be circumstantial evidence for generation of a transient tyrosyl radical under these conditions.60 Other mechanisms for DM or PHM catalysis have also beenproposed that avoid the need for a tyrosyl radical.20,39,58But these do not take account of the need to cleave the OO bond prior to substrate hydroxylation.
Other aspects of the catalytic cycles of these enzymes are less clear-cut. One issue is how electron-transfer from CuHto CuMcantake place, since crystallographically the shortest feasible electron-transfer pathway between the two copper sites is prohibitively long, at 24 amino acid residues.20Intriguingly, the crystal structure of the PHM catalytic core:substrate complex implies a potential pathway for electron transfer between the copper centers mediated by bound sub- strate. This led to the proposal that electrontransfer betweenCuHand CuMis gated by substrate binding.20 Alternatively, the fact that binding of carbon monoxide to CuH is promoted by substrate binding in PHM has led to another suggestion, that O2is reduced intwo one-electron steps.39 Under this scheme, O2 first binds to, and is reduced by, CuH; thenthe resultant super- oxide anion migrates to CuM through an as-yet-unidentified channel in the protein, where it is reduced further to the peroxide level.39 As yet, there are no data available that would distinguish betweenthese proposals. Another problem is the structure of the putative CuMhydroperoxo and/
or CuMoxyl intermediates, which have never been directly observed and for which no spectro- scopic or structural data are available.
8.16.2.2 Structural Models for the CuH Center: Three-coordinate Copper Complexes with All-nitrogen Ligation
The only known examples of synthetic three-coordinate copper(II) complexes are (1)–(7) (Scheme 1), which have near-planar Y-shaped geometries inthe crystal with the Cu ionbeing displaced from the N2X (XẳCl, O, S) donor plane by up to 0.22 A˚ (Figure 4).61–63 These complexes are characterized by axial or slightly rhombic EPR spectra with a ‘‘normal’’ gj j > g? > ge pattern and a small hyperfine coupling to Cu;gj jẳ2.17–2.22,g?ẳ2.04–2.05,Aj j{63,65Cu}ẳ90–129 G.61–64 It is uncertain how these data relate to the EPR spectrum of oxidized CuH, since accurategand Avalues for this site have not been deconvoluted from those of CuM(see above). No compounds related to (1)–(7) containing an N3donor set that would more closely replicate the CuHsite inits þ2 oxidationstate have yet beendescribed.
Dopamine CuHII CuMII O.
CuHI CuMI2e–
CuHII CuMII OH
CuHI CuMI O– O O
NH2HCuHII CuMIIO
O– CuHII CuMIIO.
O– O
H O
NH2
HCuHII CuMIIO OH O– O.
H O
NH2
H
H O
NH2
H OH O O
NH2
H
H2O
O2 .
Norepinephrine Figure3ProposedmechanismfordopaminehydroxylationbyDM,involvingone-electroncleavageofacopper(II)hydroperoxidespeciesbyatyrosineresidue.
N Cu
N Cl
N Cu
N S R
R = Ph (2) R = CH2OCH3(3)
X = H, R = Me, R' = H (4) X = H, R = H, R' = OMe (5) X = H, R = H, R' = t Bu (6) X = Cl, R = H, R' = t Bu (7)
N Cu
N O X
R R
R' (1)
Scheme 1
However, there is a large number of trigonal monocopper(I) complexes with atris-N-donor set, which could provide insight into the chemistry of the reduced CuHcomplex. Examples are known containing three monodentate ligands,65–72 usually alkylated pyridines65,66 or imidazoles.69–72 Although crystal structures of these compounds always show approximately planar copper centers, their N–Cu–N angles can show substantial deviations from 120.65–68 Alternatively, a three-coordinate copper(I) ion may be bound by one bidentate and one monodentate ligand.73–78 A planar Y-shaped geometry is most commonfor these compounds, owing to the limited bite angles of the chelating ligands; there is one T-shaped example, however.76 Finally, the three N donors may be derived from one tridentate ligand. The coordination geometry of the copper ion here is dictated by the disposition of donors within the tridentate ligand. Most examples contain meridional tris-N-donor ligands, which form planar T-shaped copper(I) complexes (8)–(31) (Scheme 2). These most resemble the T-shaped geometry of the CuH site, and are discussed in more detail below. There are also isolated examples containing tridentate cyclic ligands, which afford roughly C3v-symmetric pyramidal (32)79 or planar (33)80 stereochemistries at copper (see below) (Scheme 3).
Figure 4 Single-crystal X-ray structure of (6), a three-coordinate copper(II) complex.63All hydrogenatoms have beenomitted for clarity.
N N N
N
N Cu
+
+
N N
N N H H N
Cu S
S
(30) (31)
N N
N N N
Cu
R' R'
R
+
N N
N N Cu
R' R
+
Cu N N
N R
+
R = H, R' = H (21) R = H, R' = Me (22) R = CH2Ph, R' = H (23) R = CH2Ph, R' = CH2Ph (24) R = C{O}Me, R' = Me (25)
R = H, R' = nBu (26) R = CH2Ph, R' = Me (27)
R = H (28) R = Me (29) Cu
N N
N R
+
N N
N N N
Cu
R' R'
R
+ +
HN N N
N
N Cu
R R
R R
R = Me (8) R = CH2Ph (9) R = C2H4Ph (10) R = Ph (11)
R = H, R' = H (12) R = H, R' = Me (13) R = H, R' = CH2Ph (14) R = CH2Ph, R' = H (15) R = CH2Ph, R' = Me (16) R = CH2Ph, R' = CH2Ph (17)
R = H (18) R = Me (19) R = t Bu (20)
Scheme 2
Figure 5 Single-crystal X-ray structures of [CuL]ỵ(Lẳ2,6-bis{[3-phenylpyrazol-1-yl]methyl}pyridine) (30) (A) and [CuL(OH)2)2]2þ (B),85 showing the increase in coordination number that typically occurs upon oxidation of a three-coordinate copper(I) center. All carbon-bound hydrogen atoms have been omitted
for clarity.
N
N N
Cu
+
+
S S
N N N
N Cu
N S N
(32) (33)
Scheme 3
The complexes (9),81 (11),82 (13),83 (28),84 (29),84 (30),85 and (31)86 have all beenstructurally characterized, and show a T-shaped geometry with a ‘‘transoid’’ N–Cu–N angle of 143.03(13)–168.72(8) (Figure 5A). In addition, although no crystal structure is available, (19) has beenshownto be three-coordinate inthe solid state by X-ray absorptionspectroscopy.90While they have not been structurally characterized, salts of (10),87(12),82(14)–(17),83,88(18)–(20),89–93(21)–
(25),88,93–95(26),83and (27)88containing weakly coordinating anions, are also assumed to be three- coordinate in the solid. In addition to these compounds, there are structurally characterized dicopper(I) complexes of ligands containing two independent bis(pyrid-2-ylethyl)amine96–101orbis (pyrazol-1-ylethyl)amine102domains linked by a hydrocarbon spacer. The individual Cu centers in these compounds are structurally essentially identical to those in (8)–(11) an d (18)–(20). It is noteworthy that all of (8)–(31) contain six- or seven-membered chelate rings at copper. There is to date only one crystallographically authenticated three-coordinate copper(I) center, with two five- membered chelate rings linking three meridional N-donors.103 Rather, meridional tridentate ligands containing five-membered chelate rings instead usually form copper(I) complexes that are either helical dimers,104,105or four-coordinate adducts containing an exogenous monodentate ligand.103,106–111
Despite being three-coordinate in the solid, (18)–(20),89,91 (30),85 (31)86 or other compounds closely related to (8)–(11),100,101,112–118
(19)102 and (30)85,119 will readily add a monodentate ligand such as Cl, CO, MeCN, or PPh3, or an endogenous alkene, imine, phenoxide, or thioether donor, to afford four-coordinate products. This affinity for a fourth donor mimics the ability of CO and isocyanides to bind to the CuHsite.26,27,37–41
However, giventhe spectroscopic inertness of Cu(I), this affinity for exogenous ligands means that (8)–(31) cannot be assumed to be three- coordinate in solution, particularly in donor solvents such as MeCN. Compounds (8)–(11)81,82,87 and copper(I) complexes of several other N-alkyl-bis(2-{pyrid-2-yl}ethyl)amines,120–123 (13),83,88 (14),83,88(28),84and (29)84react rapidly with O2insolutionto form 2:1 Cu:O2adducts, that have beencharacterized insolutionby manometry,81–83,87 resonance Raman84,87 and/or UV/vis spectroscopy82–84,87,88
(Chapter 8.15). Incontrast, (12),88 (15)–(17),88 (21)–(25),88 (30),85 and (31)86 are air-stable in solution. There are no known monocopper superoxo or peroxo species derived from (8)–(31), or any otherT-shaped copper(I) precursor, that might mimic the product of a putative reactionbetweenCuH and O2(seeSection8.16.4).
The Cu(II/I) couples shownby T-shaped copper(I) compounds of meridional tris-N-donor ligands are often chemically reversible, and show cyclic voltammograms consistent with fast electrontransfer.82,85,86,93,124Since the copper(II) products generated by oxidation will expand their coordination numbers to four or five, through addition of solvent-derived ligands (Figure 5B),85,86,93,120–124
, this reversibility implies that the kinetics of exogenous ligand coordination to Cu(II), and decoordination from Cu(I), are rapid on the voltammetric timescale. The Cu(II/I) half-potential in these compounds can be sensitive to small changes in the donor properties of the tridentate ligand,82,85,93which makes it impossible to use these data to predict the effects of three-coordinationonthe oxidationpotential of the reduced CuHbiosite.
Compound (33) is the only crystallographically characterized tris-imidazole-ligated cop- per(I) complex to adopt a near-regular trigonal planar geometry.80 This complex is inert to ligand addition, does not react with O2, and undergoes only a semi-reversible electrochemical oxidationto copper(II).80 It is unclear how these properties would change following distor- tion of its coordination geometry from trigonal planar to the T-shaped structure in CuH. A T-shaped structure has beendetermined by EXAFS for [Cu(Me2Im)3]ỵ (Me2Imẳ1,2- dimethylimidazole);72 interestingly, this complex reacts rapidly with O2 insolution, whereas [Cu(Me2Im)2]þ is air-stable.71,72 There are few other three-coordinate copper(I) tris-imidazole complexes.69,70
8.16.2.3 Structural Models for the CuMCenter
8.16.2.3.1 Copper complexes with mixed nitrogen/thioether ligation
There are no synthetic copper(II) compounds with biological donors that adopt a coordination geometry analogous to that shown by CuM. The CuM architecture of a trigonal pyramidal structure bearing a basal exogenous solvent ligand, and a very long apical CuS{thioether}
distance, represents a formidable synthetic target in copper chemistry because of the strong
tendency of the copper(II) ion to form tetragonal or C2-distorted tetrahedral complexes when four-coordinate. Hence, while there is a large number of known copper(II) complexes containing mixed nitrogen/thioether ligation, these give us little information about the chemistry to be expected of the CuMcenter. Hence, rather than provide a complete survey of a large and disparate series of complexes, this section concentrates on a smaller number of compounds with mixed N/S ligation that can yield conclusions about structural changes that might accompany redox at the CuM site.
The compound that most closely approximates the coordination geometry of CuM is (34) (Scheme 4),62 which was originally targeted as a model of the type 1 cupredoxin active site (Chapter 8.4). This compound is closely related to complexes (1)–(7) already cited as models for CuH. The crystal structure of (34) shows a copper(II) ionwith a flattened tetrahedral stereo- chemistry (Figure 6A). The Cu atom is displaced 0.484(2) A˚ above the N2S{thiolate} donor plane towards the thioether S donor, which results in an ‘‘apical’’ CuS{thioether} distance of 2.403(1) A˚. This is at the short end of the range typically observed for apical or axial Cu(II)S bonds. Unfortunately, the EPR spectrum of (34) (g1ẳ2.15, g2ẳ2.06, g3ẳ2.01, A1{63,65Cu}ẳ98 G)62 is very different from the ‘‘combined’’ EPR signal from CuHỵCuM
(gj jẳ2.252.29, g?ẳ2.05, Aj j{63,65Cu}ẳ142162 G)15,16,28–33
so that despite the structural simi- larities this is a poor model for the CuMbiosite.
Figure 6 Single-crystal X-ray structures of (34) (A)62and (35) (B),127two copper(II) complexes with apical thioether donors whose molecular structures contain some features of the copper biosite. All hydrogen atoms
have beenomitted for clarity.
S Cu N N
O OH
+
S Cu N
N N
C N C N
C Cu C
N
∞
Cu S N
N N
NCMe 2+
R S N
N N
Cu
R R
S N N N Cu
R 2+
Cu S
N
N N N
+ N
Cu N S S
S Cu N N
N N
2+
Cu S N
N N
R'
R R +
Cu S N
NS L
n+
x +
Cu N N
NS Cl
Cu S N
NS x
+ (34 )
x = 1 (44) x = 2 (45) R = H (42)
R = Me (43)
(35) (36)
(37) (38) R = H, R' = Me (39)
R = Me, R' = iPr (40 )
(48) x = 1, L = SO42–, n = 0 (46)
x = 1, L = NO3–, n = 1 (47) (41)
( )
( )
Scheme 4
One series of ligands that controllably provide an N2S donor set with an apical thioether donor is small-ring diazathia macrocycles, of which 1,4-diaza-7-thiacyclononane ([9]aneN2S) is the prototype. The parent ligand affords octahedral [Cu([9]aneN2S)2]2þ with copper(II) salts of weakly coordinating anions.125,126However, several N-substituted derivatives of this macrocycle bearing donating side-arms controllably afford 1:1 complexes with copper(II), such as (35) (Figure 6B)127 and (36) (Scheme 4).128 Only one copper(I) complex of a [9]aneN2S derivative has been structurally characterized, namely the coordination polymer (37) which contains a distorted tetrahedral copper(I) center.129 The CuS bond lengths in (35), (36), and other square–pyramidal copper(II) complexes containing the [9]aneN2S ligand framework are 2.486(4)–2.648(2) A˚,127,128,130–133 while the Cu(I)S bond in (37) is 2.303(3) A˚.129 This shows that the weak apical Cu(II)S interaction in these compounds has shortened significantly upon reduction, as was concluded from EXAFS (but not crystallographic) studies of the CuMcenter.
There are a small number of other mixed nitrogen/thioether-donor copper systems for which structural data inthe þ1 an d þ2 oxidationstate are available (Scheme 4). Complex (38) was characterized crystallographically, and shows a square–pyramidal copper(II) ion with the S-donor ina basal, rather thanapical, position.134 EXAFS data for (38) an d (39) suggest that the CuS bond is shortened upon reduction, from 2.34 A˚ in(38) to 2.16 A˚ in(39); the CuN distances are also shorter inthe copper(I) complex.134 The closely related (41)135 has its S-donor in the apical positionof a square pyramid, with a very long Cu(II). . .S distance of 2.6035(3) A˚; while, (40),118 (42),136and (43)117,118show Cu(I)S bonds of 2.228(6)–2.273(1) A˚ by crystallography. Similarly, the Cu(I)S bonds in (44)137 and (45),138 at 2.230(5)–2.343(5) A˚, are up to 0.23 A˚ shorter thanfor the trigonal bipyramidal copper(II) congeners (46)137and (47).139Incontrast, the Cu(I)S bond in (48) is 0.17 A˚ longer than for copper(II) complexes containing this tridentate ligand system, at 2.469(9) A˚.140,141This is a consequence of the pseudo-linear geometry adopted by the copper(I) ion, resulting in a weaker Cu. . .S interaction.140
Although these data suggest that a CuS bond might be expected to shorten by up to 0.3 A˚
upon reduction from Cu(II) to Cu(I), these conclusions are complicated by the different coordin- ation numbers and geometries that are virtually always present in copper complex redox pairs.
Reflecting this complexity, a survey of all copper thioether complex crystal structures published in 1989 found that the mean Cu(I)S and Cu(II)S bond lengths were statistically indistinguish- able, at 2.31(5) and 2.36(5) A˚ respectively.142To conclude, although it seems reasonable to expect a shorter CuS bond in reduced compared to oxidized CuM, it is impossible to predict how significant this difference should be.
8.16.2.3.2 Copper(I) aqua complexes
Another noteworthy feature of the structure of the CuMcomplex is the presence of a water ligand inits reduced state. Crystallographically authenticated copper(I) aqua complexes are rare,143–148 and only two of the known examples are discrete monocopper compounds.143,144Each compound contains one water ligand per copper atom with a Cu(I)O bond of between 2.167(7) A˚
and 2.307(14) A˚, which is ca. 0.1–0.3 A˚ longer than is typically found in copper(II) complexes bearing a basal water ligand.142This change in bond length is consistent with the movement of the water ligand within the active site upon reduction of CuM.20