Synthetic models of the inactive Copper(II)-tyrosinate and active Copper(II)-tyrosyl radical forms of galactose and glyoxal oxidases

Article abstract of DOI:10.1021/ja9700663

A series of Cu(II) and Zn(II) complexes with new ligands having either one or two substituted phenolates appended to the 1,4,7-triazacyclononane frame were prepared and characterized by optical absorption, EPR, NMR, and/or resonance Raman spectroscopy, cy

Full text of DOI:10.1021/ja9700663

J. Am. Chem. Soc. 1997, 119, 8217-8227
8217
Synthetic Models of the Inactive Copper(II)-Tyrosinate and
Active Copper(II)-Tyrosyl Radical Forms of Galactose and
Glyoxal Oxidases
Jason A. Halfen, Brian A. Jazdzewski, Samiran Mahapatra, Lisa M. Berreau,
Elizabeth C. Wilkinson, Lawrence Que, Jr., and William B. Tolman*
Contribution from the Department of Chemistry and Center for Metals in Biocatalysis,
UniVersity of Minnesota, 207 Pleasant St. SE, Minneapolis, Minnesota 55455
ReceiVed January 8, 1997X
Abstract: A series of CuII and ZnII complexes with new ligands having either one or two substituted phenolates
appended to the 1,4,7-triazacyclononane frame were prepared and characterized by optical absorption, EPR, NMR,
and/or resonance Raman spectroscopy, cyclic voltammetry, and, in eight cases, X-ray crystallography. Features of
the active site geometries of the CuII-tyrosinate forms of galactose and glyoxal oxidases (GAO and GLO) were
modeled by these complexes, including the binding of a redox-active phenolate and an exogenous ligand (Cl-,
CH3CO2-, or CH3CN) in a cis-equatorial position of a square pyramidal metal ion. The role of the unique ortho
S-C covalent bond between a cysteine (C228) and the equatorial tyrosinate (Y272) in the proteins was probed
through an examination of the optical absorption and electrochemical properties of sets of similar complexes comprised
of phenolate ligands with differing ortho substituents, including thioether groups. The o-alkylthio unit influences
the PhO- f CuII LMCT transition and the MII-phenolate/MII-phenoxyl radical redox potential, but to a relatively
small degree. Electrochemical and chemical one-electron oxidations of the CuII and ZnII complexes of ligands having
tert-butyl protecting groups on the phenolates yielded new species that were identified as novel MII-phenoxyl radical
compounds analogous to the active CuII-tyrosyl radical forms of GAO and GLO. The MII-phenoxyl radical species
were characterized by optical absorption, EPR, and resonance Raman spectroscopy, as well as by their stoichiometry
of formation and chemical reduction. Notable features of the CuII-phenoxyl radical compounds that are similar to
their protein counterparts include EPR silence indicative of magnetic coupling between the CuII ion and the bound
radical, a band with λmax ≈ 410 nm (ꢀ ≈ 3900 M-1 cm-1) in UV-vis spectra diagnostic for the phenoxyl radical,
and a feature attributable to the phenoxyl radical C-O vibration (ν7a) in resonance Raman spectra. Similar Raman
spectra and electrochemical behavior for the ZnII analogs, as well as an isotropic signal at g ) 2.00 in their X-band
EPR spectra, further corroborate the formulations of the MII-phenoxyl radical species.
Introduction
The copper-containing metalloenzymes galactose oxidase
(GAO)1 and glyoxal oxidase (GLO)2 from fungi perform the
two-electron oxidations of primary alcohols or aldehydes,
respectively, coupled with the reduction of O2 to H2O2 (eqs 1
and 2). These enzymes are unusual because, in contrast to most
GAO: RCH2OH + O2 f RCHO + H2O2
(1)
(2)
GLO: RCHO + H2O + O2 f RCO2H + H2O2
copper proteins that effect multielectron redox reactions by using
multinuclear active sites (i.e., 1 e- per Cu, shuttling between
CuI and CuII),3 GAO and GLO use an isolated, monocopper
center to carry out two-electron redox chemistry. Details of
the geometry of this center in GAO have been revealed by X-ray
crystallography (Figure 1).4 The copper ion is square pyramidal,
with a tyrosinate ligand (or protonated form) weakly bound in
the axial position and two histidine imidazoles, a second
Figure 1. Drawing of the active site of galactose oxidase at pH 4.5 as
determined by X-ray crystallography (ref 4).
tyrosinate, and either H2O (pH 7.0) or acetate (from buffer, pH
4.5) in the equatorial sites. Interestingly, the equatorial tyro-
sinate is linked to a nearby cysteine via an ortho C-S bond to
afford a thioether-modified phenolate ligand (Y272-C228),
which is within π-stacking distance of a tryptophan residue (not
shown). Spectroscopic studies of GAO and GLO indicate that
the isolated inactive enzyme contains a CuII-tyrosinate fragment
(inactiVe, Figure 2) which upon treatment with oxidants yields
the functionally competent form (actiVe) consisting of a novel
X Abstract published in AdVance ACS Abstracts, August 1, 1997.
(1) (a) Whittaker, J. W. In Metalloenzymes InVolVing Amino Acid Residue
and Related Radicals; Sigel, H., Sigel, A., Eds.; Marcel Dekker: New York,
1994; Vol. 30, pp 315-360. (b) Knowles, P. F.; Ito, N. PerspectiVes in
Bio-inorganic Chemistry; Jai Press Ltd.: London, 1994; Vol. 2, pp 207-
244. (c) Whittaker, J. W. In Bioinorganic Chemistry of Copper; Karlin, K.
D., Tyekla´r, Z., Eds.; Chapman & Hall, Inc.: New York, 1993; pp 447-
458.
(2) Whittaker, M. M.; Kersten, P. J.; Nakamura, N.; Sanders-Loehr, J.;
Schweizer, E. S.; Whittaker, J. W. J. Biol. Chem. 1996, 271, 681-687.
(3) (a) Solomon, E. I.; Baldwin, M. J.; Lowery, M. D. Chem. ReV. 1992,
92, 521-542. (b) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem.
ReV. 1996, 96, 2563-2605.
(4) (a) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson,
M. J.; Keen, J. N.; Yadav, K. D. S.; Knowles, P. F. Nature 1991, 350,
87-90. (b) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson,
M. J.; Keen, J. N.; Yadav, K. D. S.; Knowles, P. R. Faraday Discuss. 1992,
93, 75-84. (c) Ito, N.; Phillips, S. E. V.; Yadav, K. D. S.; Knowles, P. F.
J. Mol. Biol. 1994, 238, 794-814.
S0002-7863(97)00066-8 CCC: $14.00 © 1997 American Chemical Society

8218 J. Am. Chem. Soc., Vol. 119, No. 35, 1997
Halfen et al.
metal complexes with proximate or directly bonded phenoxyl
radicals are rare;11 notable examples are Lippard’s monozinc-
(II) and diiron(III) compounds with pendant, uncoordinated
radicals12 and Wieghardt’s phenoxyl radicals coordinated to tri-
and tetravalent metals (FeIII, GaIII, MnIII, MnIV, and CrIII;
structurally characterized for CrIII) derived from the oxidation
of complexes of 1,4,7-tris(phenolato)-1,4,7-triazacyclononanes.13
We report herein our progress toward the synthesis and
structural and spectroscopic characterization of models of GAO
and GLO active site species. We have elucidated the properties
of new mono- and bis(phenolato)copper(II) and -zinc(II)
complexes relevant to the inactive form of the enzymes, with a
particular view toward evaluating the spectroscopic and redox
influences of phenolate ring substituents, including alkylthio
groups that mimic the enzyme cysteine modification. In
addition, we have prepared via electrochemical and chemical
means several oxidized versions which, on the basis of
spectroscopic and other evidence, we assign as novel metal-
phenoxyl radical species analogous to the active enzyme form
(Figure 2).14 Portions of this work have been communicated
previously.15
Figure 2. Active site forms and catalytic mechanism postulated for
galactose oxidase (refs 1, 4, and 5-7).
CuII-radical pair in the active site, with the radical localized
on the equatorial, covalently modified tyrosinate ligand.1,2,5,6
Key supporting evidence for this hypothesized formulation
includes EPR silence attributed to antiferromagnetic coupling
between the two S ) 1/2 centers, X-ray absorption edge data
indicative of the presence of CuII, resonance Raman data
showing diagnostic phenoxyl radical vibrations, and intense
optical absorption bands, including one with λmax ) 444 nm (ꢀ
) 5194 M-1 cm-1, GAO) that is postulated to be due in part to
a π f π* transition of the phenoxyl radical. The ability of the
mononuclear active site to carry out two-electron redox chem-
istry has been rationalized by invoking cycling between this
unique CuII-tyrosyl radical unit (actiVe, Figure 2) and a CuI
state (reduced), with the CuII-tyrosyl radical form being directly
responsible for hydrogen atom abstraction from the substrate
bound in the equatorial position adjacent to Y272 during
turnover.1,4,6,7
The unusual nature of the cysteine-modified CuII-tyrosyl
radical species, its critical functional role in enzyme catalysis,
and the provocative mechanisms proposed to explain how it
oxidizes substrates provide ample impetus for undertaking in-
depth studies of the properties of synthetic analogs of the GAO
and GLO active sites. Further rationale for study derives from
recognition of the CuII-tyrosyl radical species in these enzymes
as interesting members of a growing class of functionally
important metal-radical arrays in proteins that delocalize
multiple oxidizing equivalents in a manner complementary to
those involving polynuclear metal-containing active sites.8
Previous synthetic modeling efforts primarily have focused on
CuII-phenolate complexes akin to the inactive state of the
enzyme.9 Ligands with (methylthio)phenolate donors also have
been used in order to more closely mimic the cysteine-modified
tyrosinate in the proteins, but accurate mononuclear model
complexes and derived metal-phenoxyl radical species sup-
ported by these ligands have been elusive.10 Well-characterized
Results and Discussion
A. CuII- and ZnII-Phenolate Complexes. Synthesis. In
order to replicate key features of the active site environment of
GAO and GLO, we designed a series of chelates that would
favor monomeric square pyramidal CuII geometries, provide one
phenolate donor in an equatorial position for eventual oxidation
to a phenoxyl radical, and be amenable to adjustment of steric
and electronic influences in order to tune the spectroscopic and
redox properties of derived complexes and enhance the stability
of metal-phenoxyl radical species. The chosen ligands HLRR′
(Scheme 1, where R and R′ refer to the substituents para and
ortho to the phenolic oxygen atom, respectively)16 were isolated
as crystalline solids in good to excellent yields from the reaction
of the Mannich base l-(hydroxymethyl)-4,7-diisopropyl-1,4,7-
triazacyclononane (produced in situ from the reaction of 1,4-
diisopropyl-1,4,7-triazacyclononane17 and aqueous formalde-
hyde) with 2,4-disubstituted phenols in refluxing methanol.
(10) (a) Whittaker, M.; Chuang, Y.; Whittaker, J. J. Am. Chem. Soc.
1993, 115, 10029-10035. (b) Whittaker, M. M.; Duncan, W. R.; Whittaker,
J. W. Inorg. Chem. 1996, 35, 382-386.
(11) Goldberg, D. P.; Lippard, S. J. In Mechanistic Bioinorganic
Chemistry; Thorp, H. H., Pecoraro, V. L., Eds.; American Chemical
Society: Washington, DC, 1995; Adv. Chem. Ser. Vol. 246, pp 61-81.
(12) (a) Goldberg, D. P.; Watton, S. P.; Masschelein, A.; Wimmer, L.;
Lippard, S. J. J. Am. Chem. Soc. 1993, 115, 5346-5347. (b) Goldberg, D.
P.; Koulougliotis, D.; Brudvig, G. W.; Lippard, S. J. J. Am. Chem. Soc.
1995, 117, 3134-3144.
(13) (a) Hockertz, J.; Steenken, S.; Wieghardt, K.; Hildebrandt, P. J.
Am. Chem. Soc. 1993, 115, 11222-11230. (b) Sokolowski, A.; Bothe, E.;
Bill, E.; Weyhermuller, T.; Wieghardt, K. J. Chem. Soc., Chem. Commun.
1996, 1671-1672. (c) Adam, B.; Bill, E.; Bothe, E.; Goerdt, B.; Haselhorst,
G.; Hildenbrand, K.; Sokolowski, A.; Steenken, S.; Weyhermu¨ller, T.;
Wieghardt, K. Chem. Eur. J. 1997, 3, 308-319.
(14) Since the initial submission of this manuscript, several reports of
the oxidation chemistry of CuII-phenolate complexes have appeared: (a)
Wang, Y.; Stack, T. D. P. J. Am. Chem. Soc. 1996, 118, 13097-13098. (b)
Zurita, D.; Gautier-Luneau, I.; Me´nage, S.; Pierre, J.-L.; Saint-Aman, E. J.
Biol. Inorg. Chem. 1997, 2, 46-55. (c) Zurita, D.; Scheer, C.; Pierre, J.-L.;
Saint-Aman, E. J. Chem. Soc., Dalton Trans. 1996, 4331-4336. (d) Itoh,
S.; Takayama, S.; Arakawa, R.; Furuta, A.; Komatsu, M.; Ishida, A.;
Takamuku, S.; Fukuzumi, S. Inorg. Chem. 1997, 36, 1407-1416.
(15) Halfen, J. A.; Young, V. G., Jr.; Tolman, W. B. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1687-1690.
(5) (a) Baron, A. J.; Stevens, C.; Wilmot, C.; Seneviratne, K. D.;
Blakeley, V.; Dooley, D. M.; Phillips, S. E.; Knowles, P. F.; McPherson,
M. J. J. Biol. Chem. 1994, 269, 25095-25105. (b) Knowles, P. F.; Brown,
R. D., III; Koenig, S. H.; Wang, S.; Scott, R. A.; McGuirl, M. A.; Brown,
D. E.; Dooley, D. M. Inorg. Chem. 1995, 34, 3895-3902.
(6) Clark, K.; Penner-Hahn, J. E.; Whittaker, M. M.; Whittaker, J. W. J.
Am. Chem. Soc. 1990, 112, 6433-6434.
(7) (a) Branchaud, B. P.; Montague-Smith, M. P.; Kosman, D. J.;
McLaren, F. R. J. Am. Chem. Soc. 1993, 1l5, 798-800. (b) Wachter, R.
M.; Branchaud, B. P. J. Am. Chem. Soc. 1996, 118, 2782-2789. (c)
Wachter, R. M.; Branchaud, B. P. Biochemistry 1996, 35, 14425-14435.
(8) (a) Metalloenzymes InVolVing Amino Acid Residue and Related
Radicals; Sigel, H., Sigel, A., Eds.; Marcel Dekker: New York, 1994; Vol.
30. (b) Stubbe, J. A. Annu. ReV. Biochem. 1989, 58, 257-285. (c) Pedersen,
J. Z.; Finazzi-Agro, A. FEBS Lett. 1993, 325, 53-58. (d) Fontecave, M.;
Pierre, J. L. Bull. Soc. Chim. Fr. 1996, 133, 653-660.
(9) (a) Kitajima, N.; Whang, K.; Moro-oka, Y.; Uchida, A.; Sasada, Y.
J. Chem. Soc., Chem. Commun. 1986, 1504-1505. (b) Rajendran, U.;
Viswanathan, R.; Palaniandavar, M.; Lakshminarayanan, M. J. Chem. Soc.,
Dalton Trans. 1992, 3563-3564. (c) Uma, R.; Viswanathan, R.; Palani-
andavar, M.; Lakshminarayanan, M. J. Chem. Soc., Dalton Trans. 1994,
1219-1226. (d) Adams, H.; Bailey, N. A.; Fenton, D. E.; He, Q. Inorg.
Chim. Acta 1994, 215, 1-3. (e) Adams, H.; Bailey, N. A.; de Barbarin, C.
O. R.; Fenton, D. E.; He, Q.-Y. J. Chem. Soc., Dalton Trans. 1995, 2323-
2331. (f) Shulz, D.; Weyhermuller, T.; Wieghardt, K.; Butzlaff, C.;
Trautwein, A. X. Inorg. Chim. Acta 1996, 246, 387-394. (g) Karlin, K.
D.; Cohen, B. I.; Hayes, J. C.; Farooq, A.; Zubieta, J. Inorg. Chem. 1987,
26, 147- 153.
(16) Related monophenolate 1,4,7-triazacyclononane ligands with ad-
ditional N-bound carboxylates have been examined: Stockheim, C.; Hoster,
L.; Weyhermu¨ller, T.; Wieghardt, K.; Nuber, B. J. Chem. Soc., Dalton Trans.
1996, 4409-4416.
(17) (a) Houser, R. P.; Halfen, J. A.; Young, V. G., Jr.; Blackburn, N.
J.; Tolman, W. B. J. Am. Chem. Soc. 1995, 117, 10745-10746. (b)
Mahapatra, S.; Halfen, J. A.; Wilkinson, E. C.; Pan, G.; Wang, X.; Young,
V. G., Jr.; Cramer, C. J.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc.
1996, 118, 11555-11574.

Synthetic Models of Galactose and Glyoxal Oxidases
J. Am. Chem. Soc., Vol. 119, No. 35, 1997 8219
Scheme 1
Scheme 3
previously)15 and selected bond distances and angles for six are
listed in Tables 1 and 2, respectively. A representative subset
of five structural drawings is shown in Figure 3. Complete data
for all structures are provided in the Supporting Information.
All of the structures verify binding of all available donor
atoms in the ligands to the metal ion to yield mononuclear
complexes, which for the monophenolate cases also contain a
fifth exogenous ligand (Cl-, CH3CO2-, or CH3CN). The CuII
compounds of the monophenolate chelates (typified by the
structures in Figure 3a,c,d) are best described as square
pyramidal, with the phenolate and exogenous groups occupying
Scheme 2
cis-equatorial positions; the maximum τ value23 is 0.27 [LMeOMe
-
Cu(O2CCH3); see the Supporting Information]. As reported
previously for other monomeric CuII-phenolate complexes,9,19a
the Cu-Ophenolate bond is relatively short [av 1.92 Å, range
1.904(5)-1.947(3) Å] and induces by a trans influence length-
Similar methods were used previously for the preparation of
1,4,7-tris(phenol)-1,4,7-triazacyclononane ligands,13b,c,18,19 as
well as mono(phenol)-appended bis(pyridylmethyl)amine tri-
pods.20 The ligands HLMeSMe and HLtBuSMe with alkylthio
appendages akin to the cysteine-modified tyrosinate ligand
Y272-C228 in the proteins were constructed from the ap-
propriate 4-alkyl-2-(alkylthio)phenols (alkyl ) methyl, tert-
butyl). These were prepared by the reaction of the parent
4-alkylphenols with chlorodimethylsulfonium bisulfate in H2-
SO4 followed by hydrolysis of the intermediate aryldimethyl-
sulfonium salt.21 Finally, in order to more closely model the
bis(tyrosinate) ligation in the proteins, a second class of ligands
with two phenolate arms was designed. We have prepared one
ening of the opposite Cu-N bond (e.g., for LMe CuCl, Cu-N4
2
is longer by 0.08 Å than the other Cu-Neq distance).
Comparison of the pairs of structures LMe MCl (M ) Cu or
2
Zn, Figure 3a,b) and LtBu M (M ) Cu or Zn, only shown for M
4
) Cu in Figure 3e) reveals the consequences of coordination
of the same ligand in complexes different only with respect to
the nature of the metal ion. Whereas LMe CuCl is square
2
pyramidal (τ ) 0.07), LMe ZnCl is distorted trigonal bipyramidal
2
(τ ) 0.61) with the phenolate and two amine donors (Nl and
N4) comprising the trigonal plane and the third amine N-atom
(Nl) and Cl- residing in pseudoaxial positions. On the other
hand, the structures of LtBu M (M ) Cu or Zn) are more similar,
example so far, H2LtBu (Scheme 2), by the Mannich base
4
4
with each best described as having metal ion geometries
intermediate between square pyramidal and trigonal bipyramidal
(Cu, τ ) 0.45; Zn, τ ) 0.60). Both contain a MeOH solvate
molecule that is hydrogen-bonded to the O-donor atom of one
procedure described above, but starting with the new synthon
1-isopropyl-1,4,7-triazacyclononane, which was constructed via
an alkylation/deprotection sequence from 1,4-ditosyl-1,4,7-
triazacyclononane.22
of the coordinated phenolates. The structure of LtBu Cu is quite
A series of crystalline CuII and ZnII complexes of the new
monophenolate ligands LRR′ was prepared by deprotonation of
the phenols HLRR′ with NaH in THF followed by treatment with
4
similar to that reported for a CuII complex of a tris(phenolato)-
1,4,7-triazacyclononane ligand, except for the presence of a
dangling, uncoordinated phenol in the latter.19a The bis-
(tyrosinate) ligation present in the GAO and GLO active sites
is modeled by these compounds insofar as they also have two
phenolates in symmetry inequivalent positions, although the
distinct long axial/short equatorial phenolate bonding pattern
for the square pyramidal copper center in the proteins is not
replicated in the synthetic complexes. We cannot provide a
definitive rationale for the significant difference between the
the appropriate anhydrous metal salts (Scheme 3); H2LtBu
4
similarly was converted to LtBu M (M ) Cu, Zn). Attempts to
4
prepare CuI complexes by treating the deprotonated ligands with
CuCl or [Cu(CH3CN)4]ClO4 generally failed due to dispropor-
tionation reactions, as indicated by the purple color of the
solutions (due to CuII-phenolate species) and the presence of
colloidal copper metal.
X-ray Crystal Structures. We have solved the X-ray crystal
structures of eight complexes of the new ligands; of these,
Cu/Zn pair of structures for the LMe case compared to the lesser
2
divergence seen for LtBu , although we surmise that the steric
crystallographic data for five (data for LMe CuCl were reported
4
2
hindrance due to the phenolate tert-butyl groups in the latter is
an important factor.
(18) Moore, D. A.; Fanwick, P. E.; Welch, M. J. Inorg. Chem. 1989,
28, 1504-1506.
The structure of the GAO active site crystallographically
characterized at pH 4.5 (Figure 1)4 is most closely modeled by
LMeSMeCu(O2CCH3) (Figure 3c), which contains phenolate and
acetate ligands bound in cis-equatorial positions to a distorted
square pyramidal CuII ion (τ ) 0.24). The acetate coordination
(19) (a) Auerbach, U.; Eckert, U.; Wieghardt, K.; Nuber, B.; Weiss, J.
Inorg. Chem. 1990, 29, 938-944. (b) Auerbach, U.; Weyhermuller, T.;
Wieghardt, K.; Nuber, B.; Brill, E.; Butzlaff, C.; Trautwein, A. X. lnorg.
Chem. 1993, 32, 508-519.
(20) Cox, D. D.; Que, L., Jr. J. Am. Chem. Soc. 1988, 110, 8085-8092.
(21) Pilgram, K. H.; Medema, D.; Soloway, S. B.; Gaertner, G. W. U.S.
Patent 3772391, 1973.
(22) Sessler, J. L.; Sibert, J. W.; Lynch, V. Inorg. Chem. 1990, 29, 4143-
4146.
(23) Addison, A. W.; Rao, T. N.; Reedijk, J.; von Rijn, J.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans. 1984, 1349-1356.

8220 J. Am. Chem. Soc., Vol. 119, No. 35, 1997
Table 1. Summary of X-ray Crystallographic Dataa
Halfen et al.
[LMe ZnCl]‚
[LMeSMeCu(O2CCH3)]‚ [LtBu Cu(CH3CN)]‚
[LtBu Cu]‚
[LtBu Zn]‚
2
2
4
4
complex
empirical formula
formula weight
crystal system
space group
a (Å)
0.32H2O
CH2Cl2
CF3SO3‚C7H8
CH3OH
CH3OH
C21H36.64N3O1.32ClZn C24H41N3O3SCl2Cu
C37H59N4O4F3SCu
776.48
triclinic
C40H67N3O3Cu
701.51
monoclinic
P21/c
13.5373(5)
10.4219(4)
29.190(1)
C40H67N3O3Zn
703.34
monoclinic
P21/n
16.952(3)
11.435(2)
21.492(4)
452.51
monoclinic
P21/c
568.10
monoclinic
P21/n
P1h
12.746(3)
12.911(3)
13.648(3)
12.676(3)
13.228(3)
17.457(4)
11.798(2)
14.084(4)
15.065(2)
73.74(3)
67.52(3)
65.97(3)
2089.5(7)
2
b (Å)
c (Å)
R (deg)
â (deg)
90.36(3)
102.29(3)
102.396(1)
94.70(3)
γ (deg)
V (Å3)
2245.9(8)
4
2860(1)
4
4022.3(3)
4
1.158
173(2)
4152(1)
4
1.127
293(2)
Z
density(calcd) (g cm-3
temp (K)
)
1.338
293(2)
1.361
293(2)
1.234
293(2)
crystal size (mm)
diffractometer
0.50 × 0.35 × 0.25
0.45 × 0.40 × 0.30
0.50 × 0.45 × 0.10
0.50 × 0.15 × 0.08 0.45 × 0.37 × 0.35
Enraf-Nonius CAD-4 Enraf-Nonius CAD-4 Enraf-Nonius CAD-4 Siemens SMART Enraf-Nonius CAD-4
abs coeff (mm-1
)
1.230
50.00
4131
3942
1.053
49.98
5252
5009
0.626
50.04
7717
7325
0.581
50.06
0.628
50.00
2θ max (deg)
no. of reflections collected
no. of indep reflections
19943
7059
7289
4255
no. of reflections with I > 2σ(I) 2610
3013
307
0.0606/0.1410
1.021
0.542/-0.823
3670
416
0.0758/0.1648
1.042
0.476/-0.503
4750
504
0.0507/0.1159
1.001
0.709/-0.631
2689
454
0.0496/0.1017
1.046
0.192/-0.183
no. of variable parameters
258
b
R1/wR2
0.0460/0.1058
1.021
goodness of fit (F2)
largest diff features (e- Å-3
)
0.363/-0.362
2
2
a Radiation used: Mo KR (λ ) 0.071 073 Å). b R1 ) ∑||Fo| - |Fc||/∑|Fo|; wR2 ) [∑[w(Fo - Fc2)2]]1/2, where w ) 1/σ2(Fo ) + (aP)2 + bP.
borders on asymmetric bidentate,24 the oxygen atom O3 being
directly trans to the axial amine N-donor (N3; the N3-Cu1-
O3 angle is 173°) with an O3‚‚‚Cu1 distance of 2.72 Å. This
weak axial interaction of a CuII ion with a carbonyl oxygen
influences of the phenolate substituents, in particular the
alkylthio groups akin to the enzyme C228 appendage. The
diamagnetic ZnII complexes exhibit sharp lH NMR spectra with
increases in the number and multiplicity of the macrocycle
backbone methylene resonances compared to the free ligand
that are diagnostic for metal ion binding.25 Despite the
inequivalence of the phenolate dispositions in the X-ray structure
atom resembles that observed in [LiPr Cu(O2CCH3)2] (L
)
)
iPr3
3
1,4,7-triisopropyl-1,4,7-triazacyclononane; Naxial-Cu-Oacetate
156°; Cu‚‚‚Oacetate ) 3.15 Å)25 and some type 1 copper
proteins.26 The thiomethyl S-atom (Sl) does not coordinate to
the CuII ion [Cul‚‚‚Sl ) 4.606(2) Å] in contrast to the structures
of other CuII complexes containing o-thioether-substituted
phenolate ligands.10a An additional noteworthy structural feature
is the conformation of this thiomethyl substituent, which is
coplanar with the phenolate ring [the C14-C13-Sl-C17
torsion angle is -0.3(6)°]. This resembles the conformation
of the redox-active Y272-C228 pair in galactose oxidase in
which Sγ and Câ of Cys228 are coplanar with the phenolic ring
of Y272, an observation which led to the proposal that the
SCys-CTyr bond possesses partial double bond character.4b In
LMeSMeCu(O2CCH3) the Carene-S distance is 1.767(6) Å, which
is typical for such distances in thioether-substituted arenes
argued to have little double bond character.27 In addition, a
structure similar to that of LMeSMeCu(O2CCH3) was found for
the LMeOMe analog (see the Supporting Information). The
finding of identical conformations of the o-OMe and -SMe
groups suggests that intermolecular packing forces govern the
conformations of these side chains and that unique S-C multiple
bonding need not be invoked to explain the planarity of the
o-(methylthio)phenolate.
of LtBu Zn, only a single set of resonances appears in the NMR
4
spectrum, suggestive of the operation of a fluxional process in
solution that rapidly interchanges the phenolate environments.
All of the CuII complexes exhibit axial signals in their X-band
EPR spectra (frozen solutions at 77 K) that are consistent with
2
2
a dx -y ground state (g| > g > 2.0, A| ≈ 150-170 G) with
a minor rhombic perturbation (perhaps indicative of a distortion
toward a trigonal bipyramidal structure) that is most clearly
resolved in the acetate and azide complexes. The CuII com-
plexes are also deeply colored due to absorption features with
λmax ≈ 450-530 nm (ꢀ ≈ 1000-2000 M-1 cm-1) that we assign
as PhO- f CuII ligand-to-metal charge transfer (LMCT)
transitions by analogy to other known CuII-phenolate systems
(Table 3 and Figure 4).9 This assignment is supported by
Raman spectroscopic studies of [LtBu Cu(CH3CN)]CF3SO3 and
2
LtBu Cu, which show resonance-enhanced phenolate vibrational
4
features upon irradiation into the absorption band (Vide infra).
Also consistent with the LMCT attribution, comparison of the
λmax values for the isostructural series LRR′Cu(O2CCH3) and
LRR′CuCl (Table 3) shows that the energy of the transition within
each series increases (λmax decreases) as the energy of the filled
ligand-based orbitals decreases according to the order LMeOMe
Spectroscopic Properties. Spectroscopic characterization of
the CuII- and ZnII-phenolate complexes was undertaken in
order to correlate their solution structures with those determined
by X-ray crystallography and to understand the electronic
tBu2
(most electron rich) ≈ LMeSMe ≈ LtBuSMe > LMe ≈ L
(least
2
electron rich). The overall range of energies within each series
is not large, however, which by extrapolation implies a relatively
minor role for the thioether appendage in tuning the redox
properties of the equatorial tyrosinate in GAO and GLO. This
conclusion is further corroborated by the results of cyclic
voltammetry experiments (Vide infra). The LMCT feature in
(24) Rardin, R. L.; Tolman, W. B.; Lippard, S. J. New J. Chem. 1991,
15, 417-430.
(25) Halfen, J. A.; Mahapatra, S.; Wilkinson, E. C.; Gengenbach, A. J.;
Young, V. G., Jr.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1996,
118, 763-776.
(26) Adman, E. T. AdV. Protein Chem. 1991, 42, 145-197.
(27) (a) Peach, M. E.; Burschka, C. Can. J. Chem. 1982, 60, 2029-
2037. (b) Blackmore, W. R.; Abrahams, S. C. Acta Crystallogr. 1955, 8,
329-335. (c) Cox, E. G.; Gillot, R. J. J. H.; Jeffrey, G. A. Acta Crystallogr.
1949, 2, 356-363.
LtBu Cu (Figure 4) is shifted significantly to higher energy
4
compared to the monophenolate compounds, consistent with
greater electron density at the CuII ion and raised acceptor orbital
energies due to the second phenolate donor.

Synthetic Models of Galactose and Glyoxal Oxidases
J. Am. Chem. Soc., Vol. 119, No. 35, 1997 8221
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
Complexes Characterized by X-ray Crystallographya
LMe CuCl
2
Cu(1)-O(1)
Cu(1)-N(4)
Cu(1)-Cl(1)
1.916(5)
2.148(6)
2.286(2)
Cu(1)-N(1)
2.069(5)
2.291(6)
92.8(2)
109.3(2)
83.0(2)
171.4(2)
92.4(2)
Cu(1)-N(7)
O(1)-Cu(1)-N(1)
O(1)-Cu(1)-N(7)
N(1)-Cu(1)-N(4)
N(1)-Cu(1)-Cl(1)
N(4)-Cu(1)-Cl(1)
O(1)-Cu(1)-N(4) 167.4(2)
O(1)-Cu(1)-Cl(1) 90.3(2)
N(1)-Cu(1)-N(7) 83.9(3)
N(4)-Cu(1)-N(7) 82.2(2)
N(7)-Cu(1)-Cl(1) 102.7(2)
[LMe ZnCl]‚0.32H2O
2
Zn(1)-O(1)
Zn(1)-N(4)
Zn(1)-Cl(1)
1.931(3)
2.156(4)
Zn(1)-N(1)
Zn(1)-N(7)
2.316(3)
2.134(3)
86.86(11)
130.34(14)
79.25(12)
174.19(10)
95.31(10)
2.3410(13) O(1)-Zn(1)-N(1)
O(1)-Zn(1)-N(4) 137.63(13) O(1)-Zn(1)-N(7)
O(1)-Zn(1)-Cl(1) 95.94(9) N(1)-Zn(1)-N(4)
N(1)-Zn(1)-N(7) 80.03(12) N(1)-Zn(1)-Cl(1)
N(4)-Zn(1)-N(7) 86.51(14) N(4)-Zn(1)-Cl(1)
N(7)-Zn(1)-Cl(1) 101.82(10)
[LMeSMeCu(O2CCH3)]‚CH2Cl2
Cu(1)-O(1)
Cu(1)-N(1)
Cu(1)-N(7)
O(3)-C(25)
1.943(4)
2.046(4)
2.295(5)
1.236(7)
Cu(1)-O(2)
1.958(4)
2.170(5)
1.267(7)
91.6(2)
175.5(2)
84.1(2)
Cu(1)-N(4)
O(2)-C(25)
O(1)-Cu(1)-O(2)
O(1)-Cu(1)-N(4)
N(1)-Cu(1)-N(4)
N(1)-Cu(1)-O(2)
N(4)-Cu(1)-O(2)
O(1)-Cu(1)-N(1) 93.4(2)
O(1)-Cu(1)-N(7) 93.7(2)
N(1)-Cu(1)-N(7) 82.8(2)
N(4)-Cu(1)-N(7) 82.4(2)
N(7)-Cu(1)-O(2) 114.8(2)
161.3(2)
91.9(2)
Cu(1)-O(2)-C(25) 109.5(4)
2
Cu(1)-O(1)
Cu(1)-N(1)
Cu(1)-N(7)
Cu(1)-N(10)
Cu(1)-N(4)
2.011(6)
2.297(5)
106.5(2)
84.8(2)
[LtBu Cu]‚CH3OH
4
Cu(1)-O(1)
Cu(1)-N(1)
Cu(1)-N(3)
1.935(2)
Cu(1)-O(2)
1.916(2)
2.043(2)
89.45(9)
150.85(9)
85.26(9)
178.12(10)
94.17(9)
2.076(2)
2.363(2)
Cu(1)-N(2)
O(1)-Cu(1)-O(2)
O(1)-Cu(1)-N(2)
O(1)-Cu(1)-N(1) 94.17(9)
O(1)-Cu(1)-N(3) 126.49(8) N(1)-Cu(1)-N(2)
N(1)-Cu(1)-N(3) 80.29(9)
N(2)-Cu(1)-N(3) 82.22(9)
N(3)-Cu(1)-O(2) 101.42(9)
N(1)-Cu(1)-O(2)
N(2)-Cu(1)-O(2)
[LtBu Zn]‚CH3OH
4
Zn(1)-O(1)
Zn(1)-N(1)
Zn(1)-N(3)
1.946(3)
2.244(4)
2.177(5)
Zn(1)-O(2)
1.962(3)
2.110(5)
92.15(14)
136.5(2)
80.6(2)
172.7(2)
93.4(2)
Zn(1)-N(2)
O(1)-Zn(1)-O(2)
O(1)-Zn(1)-N(2)
N(1)-Zn(1)-N(2)
N(1)-Zn(1)-O(2)
N(2)-Zn(1)-O(2)
O(1)-Zn(1)-N(1) 89.5(2)
O(1)-Zn(1)-N(3) 135.1(2)
N(1)-Zn(1)-N(3) 80.3(2)
N(2)-Zn(1)-N(3) 85.0(2)
N(3)-Zn(1)-O(2) 103.4
a Estimated standard deviations indicated in parentheses.
Resonance Raman spectra of frozen solutions of [LtBu Cu-
2
(CH3CN)]O3SCF3 and LtBu Cu in CD3CN obtained by excitation
4
into the LMCT bands are shown in Figure 5a,b. They contain
high-frequency features typical of metal-phenolate complexes.
For example, metal-tyrosinate proteins give rise to characteristic
vibrations at ca. 1170, 1270, 1500, and 1600 cm-l that derive
mainly from ring deformations ν9a, ν7a, ν19a, and ν9a, respec-
tively.28,29 Corresponding peaks are found at similar energies
in the spectra shown in Figure 5a,b, but the increased substitu-
Figure 3. Drawings of the X-ray crystal structures of (a) LMe CuCl,
2
(b) LMe ZnCl, (c) LMeSMeCu(O2CCH3), (d) the cationic fragment of
2
tBu
[LtBu Cu(CH3CN)]O3SCF3‚C7H8, and (e) L Cu‚MeOH. With the
exception of structure e, which shows the MeOH solvate molecule
involved in hydrogen bonding to the copper complex, solvate molecules
and hydrogen atoms are omitted for clarity. All ellipsoids are drawn at
the 50% probability level.
2
4
(28) Que, L., Jr. In Biological Applications of Raman Spectroscopy;
Spiro, T. G., Ed.; John Wiley & Sons: New York, 1988; Vol. 3, pp 491-
521. (b) Que, L., Jr. Coord. Chem. ReV. 1983, 50, 73.
(29) (a) Mukherjee, A.; McGlashen, M. L.; Spiro, T. G. J. Phys. Chem.
1995, 99, 4912-4917. (b) McGlashen, M. L.; Eads, D. D.; Spiro, T. G.;
Whittaker, J. W. J. Phys. Chem. 1995, 99, 4918-4922.

8222 J. Am. Chem. Soc., Vol. 119, No. 35, 1997
Halfen et al.
Table 3. UV-Vis Absorption Spectral Data for CuII-Phenolate
and MII-Phenoxyl Radical Species (M ) Cu or Zn)
a
λ
max, nm (ꢀ, M-1 cm-1
)
complex
CuII-Phenolates
LMe CuCl
340 (3100)
350 (3000)
520 (1500)
710 (sh, 440)
720 (sh, 400)
2
LtBu CuCl
526 (1200)
528 (1150)b
530 (950)b
474 (960)b
486 (900)b
484 (810)b
538 (1440)
448 (1800)
438 (1000)
559 (642)
2
LMeOMeCuCl
LtBuSMeCuCl
LtBu Cu(O2CCH3)
2
LMeOMeCu(O2CCH3)
LMeSMeCu(O2CCH3)
[LtBu Cu(CH3CN)]CF3SO3 342 (3600)
660 (sh, 700)c
706 (430)d
2
LtBu Cu‚MeOH
inactive GAO
300 (18000)
375 (1880)
370 (2670)
4
625 (1167)e,f
747 (500)e,f
678 (1365)e
728 (888)e
inactive GAO azide
inactive GLO
inactive GLO azide
451 (1875)
564 (1248)
CuII- and ZnII-Phenoxyl Radicals
2+
[LtBu Cu(CH3CN)]
410 (4000)
396 (2600)
398 (3900)
422 (sh, 3800) 672 (1000)g
2
[LtBuZnCl]2+
406 (2650)
568 (2200)
690 (250)d
646 (2200)h
722 (400)d
800 (3211)e,f
851 (4300)e
+
[LtBu Cu]
4
+
[LtBu Zn]
394 (sh, 2900) 408 (3200)
444 (5194)
4
active GAO
active GLO
448 (5700)
a Except where noted, spectra measured on CH2Cl2 solutions at room
temperature. b Additional intensity is present at longer wavelengths
(∼600-800 nm), but no bands are clearly resolved. c 10:1 CH2Cl2/
CH3CN. d CH2Cl2, -40 °C. e Reference 2. f Whittaker, M. M.; Whit-
+
Figure 5. Resonance Raman spectra of (a) [LtBu Cu(CH3CN)] , (b)
2
tBu
+
tBu
+
LtBu Cu, (c) [L Cu(CH3CN)]2+, (d) [LtBu Cu] , and (e) [L Zn] . All
spectra were acquired on frozen solutions at -196 °C in CD2Cl2
[spectrum b] or CD2Cl2/CD3CN mixtures [spectra a and c-e], with
excitation wavelengths of 514.5 nm [spectrum a] or 457.9 nm [spectra
b-e].
4
2
4
4
g
taker, J. W. J. Biol. Chem. 1988, 263, 6074-6080. ∼5:1 CH2Cl2/
CH3CN, -40 °C. h 1:0.1 CH2Cl2/CH3CN, -40 °C.
Table 4. Electrochemical Data from Cyclic Voltammetry
Experimentsa
E1/2
∆Ep
(mV)
E1/2
∆Ep
(mV)
complex
(V)b
complex
(V)b
LtBu CuCl
0.56
0.52
0.51
0.49
0.79
86
69
82
70
92
LtBu Cu
0.50
0.78
0.48
0.81
93
99
83
85
2
4
LtBu ZnCl
2
LtBuSMeCuCl
LtBuSMeZnCl
LtBu Zn
4
[LtBu Cu(CH3CN)]
+
2
a All experiments were performed using CH2Cl2 with 0.4 M
tetrabutylammonium hexafluorophosphate (TBAPF6), scan rate 100 mV
s-1, room temperature, ipa ≈ ipc. b Reported versus SCE.
modified tyrosinate at 1267 and 1596 cm-l, respectively, close
to their positions in the model compound spectra.2
B. CuII- and ZnII-Phenoxyl Radical Complexes. Elec-
trochemistry. Without exception, the LRR′CuX (X ) Cl, O2-
CCH3) complexes and their ZnII analogs supported by the
ligands LMe , LMeOMe, and LMeSMe exhibit irreversible oxidation
2
processes in their cyclic voltammograms (0.4 M TBAPF6 in
CH2Cl2 or CH3CN). We attribute this behavior to decomposi-
tion of the electrochemically generated species via C-C bond
coupling processes involving the benzylic positions para to the
phenolate (phenoxyl) oxygen atom, where significant spin
density in phenoxyl radicals is known to reside and to contribute
to enhanced reactivity.30
In contrast, the cyclic voltammograms (CH2Cl2, 0.4 M
TBAPF6) of complexes with tert-butyl groups in the positions
para to the phenolate oxygen atoms are characterized by
electrochemically (∆Ep < 100 mV) and chemically (ipc ≈ ipa)
reversible redox processes (Table 4). Controlled potential
coulometric experiments with the monophenolate compounds
revealed that these redox processes correspond to the transfer
Figure 4. Optical absorption spectra of (top) [LtBu Cu(CH3CN)]O3-
2
SCF3 in CH2Cl2 (s) and [LtBu Cu(CH3CN)]2+ resulting from oxidation
2
by 1 equiv of CeIV in 5:1 CH2Cl2/CH3CN at -40 °C (- - -) and (bottom)
tBu
+
LtBu Cu in CH2Cl2 (s) and [L Cu] resulting from oxidation by 1
4
4
equiv of CeIV in 1:0.1 CH2Cl2/CH3CN at -40 °C (- - -).
tBu4
of one electron. For example, low-temperature (-40 °C)
tion of the phenolates in LtBu and L relative to the tyrosinate
2
+
electrolysis of a solution of [LtBu Cu(CH3CN)] in CH2Cl2 (0.4
2
moiety very likely gives rise to the multiplicity of some of the
features.13a The ν7a and ν8a modes for the axial, non-redox-
active tyrosinate ligand in the inactive form of glyoxal oxidase
(pH 8.1) occur at 1260 and 1609 cm-l and for the cysteine-
(30) (a) Rigby, S. E. J.; Nugent, J. H. A. Biochemistry 1994, 33, 1734-
1742. (b) O’Malley, P. J.; MacFarlane, A. J.; Rigby, S. E. J.; Nugent, J. H.
A. Biochim. Biophys. Acta 1995, 1232, 175-179.

Synthetic Models of Galactose and Glyoxal Oxidases
J. Am. Chem. Soc., Vol. 119, No. 35, 1997 8223
M TBAPF6) at 1.0 V vs SCE resulted in the removal of 0.8 (
0.1 e- mol-1 of complex and a color change from purple to
deep green; similar color changes were observed upon coulo-
tBu2
metric oxidation of LtBu CuCl (purple to deep green) and L
-
2
ZnCl (colorless to light green). UV-vis spectroscopic (at -40
°C) analysis of the EPR-silent (at 77 K, X-band) green solution
obtained after the electrochemical oxidation of [LtBu Cu(CH3-
2
CN)]+ revealed the presence of new optical absorption features
with λmax ) 410 (ꢀ ) 3500 M-1 cm-1) and 648 (1000) nm. On
the basis of the resemblance of the 410 nm feature to similar
bands in activated GAO and GLO (Table 3) and hindered
phenoxyl radicals,31 as well as other chemical and spectroscopic
evidence (Vida infra), we postulate that one-electron oxidation
results in the generation of a CuII-phenoxyl radical complex.
The assignment of the oxidation process as ligand-based and
not as a CuII/CuIII couple is supported by the observation of
similar redox behavior for the ZnII-phenolate compounds (albeit
at a slightly different potential), in which the central metal ion
is not redox active.
Figure 6. Cyclic voltammograms of LtBu M (M ) Cu or Zn) in CH2-
4
Cl2 (0.4 M TBAPF6).
Scheme 4
The lower redox potential of LtBu CuCl (0.56 V) compared
2
+
to the related complex [LtBu Cu(CH3CN)] (0.79 V) may be
2
attributed to the increased electron-donating ability of the
chloride ligand compared to the acetonitrile donor. More
interestingly, comparison of the electrochemical data for LtBuSMe
-
MCl (M ) Cu or Zn) with the data obtained for the analogous
complexes LtBu MCl (Table 4) reveals a modest lowering of
2
the E1/2 values in the thioether-appended systems (by 50 mV
for Cu and 30 mV for Zn). Thus, the electron-releasing property
of the o-SMe group (indicated, for example, by the lower energy
of the LMCT band in the CuII complex of LtBuSMe compared to
that of LtBu ; Table 3) facilitates formation of the metal-
2
phenoxyl radical complexes by lowering the phenolate/phenoxyl
radical redox potential, albeit not by a large amount. These
spectroscopic and electrochemical observations support the
notion that the o-cysteine modification of the redox active
tyrosinate ligand in the active sites of GAO and GLO serves in
part to facilitate the oxidation of the CuII-Y272 pair to CuII-
Y272• by O2 during the last step in the proposed enzyme
mechanism(s),1,4 although the effect appears to be moderate on
the basis of our observations with the synthetic compounds.
Other structural and/or electronic influences that are present in
the protein but are absent in the model complexes (e.g., the
π-stacking interaction with the nearby tryptophan residue) also
would be expected to contribute in nontrivial ways to the control
of the Y272/Y272• redox potential, which at E1/2 ) 0.23 V vs
SCE is significantly lower than those we have observed.32
The cyclic voltammograms of the diphenolate complexes
stable in solution at low temperature33 by using either of the
two one-electron oxidants (NH4)2[Ce(NO3)6] or [thianthryl]-
BF4,13a,34 both of which have the added advantage of existing
as spectroscopically (EPR, UV-vis, resonance Raman) innocent
species in their reduced forms [Ce(NO3)3 and thianthrene,
respectively]. For example, treatment of CH2Cl2 solutions of
+
[LtBu Cu(CH3CN)] with stoichiometric amounts of either
2
(NH4)2[Ce(NO3)6] in CH3CN at -40 °C or [thianthryl]BF4 in
CH2Cl2 at -80 °C resulted in a color change from purple to
deep green (Scheme 4). These green solutions are EPR silent
(X-band, 92 ( 5% by double integration of the residual signal
at 77 K), consistent with magnetic coupling between an S )
1/2 CuII ion and an S ) 1/2 phenoxyl radical. The type of
coupling is unclear at present; antiferromagnetic coupling
between a coordinated phenoxyl radical and a central metal ion
was observed by Wieghardt and co-workers in a FeIII-phenoxyl
LtBu M (M ) Cu or Zn) contain two reversible, one-electron
4
waves (Figure 6, Table 4). Again, the close correspondence of
the data for the CuII and ZnII compounds of LtBu argues for a
radical complex (J ) -88 cm-1 13a,c
and a structurally
)
4
ligand- rather than a metal-centered process to yield a MII-
phenoxyl radical species in the first oxidation; this is further
supported by spectroscopic data (Vide infra). Note that simul-
taneous phenolate and phenoxyl radical coordination is impli-
cated for these systems, just as in the active forms of GAO and
GLO. The second reversible oxidation also occurs at similar
potentials for the copper and zinc compounds, suggesting that
it involves another ligand-centered oxidation to MII-(phenoxyl)2
characterized CrIII-phenoxyl radical compound,13b but ferro-
magnetic interactions (S ) 1 ground state) have been identified
in equatorial coordinated CuII-semiquinato complexes that are
EPR silent at X-band frequency due to large zero field splitting.35
Extensive magnetochemical studies to address this issue are
planned. The UV-vis spectrum of the green product (Figure
4) is similar to that of the electrochemically oxidized solution,
the feature at ∼415 nm (ꢀ ≈ 3900 M-1 cm-1) similar to that
reported for the active forms of GAO and GLO1,2 being
indicative of a coordinated phenoxyl radical. Both EPR and
2+
species, [LtBu M]
. Experiments aimed at confirming this
4
formulation and unraveling the magnetochemical properties of
these compounds are underway and will be reported separately.
Chemical Synthesis and Spectroscopic Properties. In
addition to being accessible via electrochemical methods, we
were able to generate MII-phenoxyl radical species that are
(33) Repeated attempts to isolate the oxidized products as pure solids
have yet to be successful.
(34) Boduszek, B.; Shine, H. J. J. Org. Chem. 1988, 53, 5142-5143.
(35) For example, see: (a) Kahn, O.; Prins, R.; Reedijk, J.; Thompson,
J. S. Inorg. Chem. 1987, 26, 3557-3561. (b) Benelli, C.; Dei, A.; Gatteschi,
D.; Pardi, L. Inorg. Chem. 1990, 29, 3409-3415. (c) Dei, A.; Gatteschi,
D.; Pardi, L.; Barra, A. L.; Brunel, L. C. Chem. Phys. Lett. 1990, 175,
589-592.
(31) Altwicker, E. R. Chem. ReV. 1967, 67, 475-531.
(32) Dyrkacz, G. R.; Libby, R. D.; Hamilton, G. A. J. Am. Chem. Soc.
1976, 98, 626-628.

8224 J. Am. Chem. Soc., Vol. 119, No. 35, 1997
Halfen et al.
shifted to higher energy than those for the LtBu and LtBu
2
4
compounds, however. The lower energy of the bands for the
compounds we have prepared appears to be a consequence of
coordination of the phenoxyl radical to the metal ion. Finally,
while both CuII-phenoxyl radical compounds do not give rise
to an EPR signal, the ZnII complexes exhibit nearly isotropic
signals (∼19 G width, peak-to-trough) centered at g ) 2.00 in
their X-band EPR spectra at 77 K (Figure 7). Double integration
of these signals revealed the appropriate intensity for their
formulations as monooxidized ZnII-phenoxyl radicals.
Consistent with the reversible behavior apparent in cyclic
voltammograms of the MII- phenolate precursors, reduction of
the chemically synthesized MII-phenoxyl radical complexes
could be effected electrochemically or via stoichiometric
reaction with ferrocene (Fc) or N,N,N′,N′-tetramethyl-1,4-
phenylenediamine (TMPDA; cf. Scheme 4). For example, the
cyclic voltammograms (initial scan cathodic) of the oxidized
2+
+
complexes [LtBu ZnCl]
and [LtBu Cu] (CH2Cl2, 0.4 M
2
4
TBAPF6, -40 °C) are the exact complements of those of their
respective starting materials (initial scan anodic). Also, treat-
2+
ment of the green solution of [LtBu Cu(CH3CN)] in CH2Cl2
2
at -40 °C with Fc or TMPDA as monitored by UV-vis
spectroscopy revealed quenching of the CuII-phenoxyl radical
features and growth of those of the CuII-phenolate complex
and either Fc+ (λmax ) 620 nm) or the TMPDA radical cation
(λmax ) 570 nm, ꢀ ) 12500 M-1 cm-1).37 Titrations established
a 1:1 stoichiometry for these reactions.
Figure 7. X-band EPR (top, 1:1 CH2Cl2/toluene, -196 °C) and UV-
+
vis (bottom, CH2Cl2, 0.49 mM, -40 °C) spectra of [LtBu ZnCl] .
2
Further evidence in support of the formulation of the MII-
phenoxyl radical compounds was obtained from resonance
UV-vis titrations established the 1:1 stoichiometry of the
oxidation reaction, confirming the previously discussed coulo-
Raman spectroscopy of the solutions resulting from one-electron
metric results.
+
oxidation of [LtBu Cu(CH3CN)] and LtBu Cu (Figure 5). The
phenolate vibrations of the precursor complexes (Figure 5a,b)
are replaced by one prominent feature at 1495-1497 cm-1 upon
2
4
tBu4
Similar one-electron oxidations of LtBu ZnCl and L M (M
2
) Cu or Zn) were performed; the spectroscopic data for the
resulting MII-phenoxyl radical species are listed in Table 3 and
+
shown for [LtBu ZnCl] in Figure 7. The optical absorption data
oxidation (Figures 5c,d). This peak also is observed at 1508
2
cm-1 in [LtBu Zn] (Figure 5e) and is assigned to ν7a of the
+
4
are fairly similar for all of the compounds, although the intensity
of the longer wavelength features is greater for the CuII
complexes than for the ZnII cases, consistent with the expected
magnitude of additional contributions from ligand field transi-
tions in this region for the different ions. Moreover, the
phenoxyl radical ligand by analogy to similar features in other
species that contain such a radical, e.g., activated GAO (1487
cm-1),29b the R2 protein of ribonucleotide reductase (1498
cm-1),38 and a diiron(III) model complex (1504 cm-1).11,12 The
monoradical derived from LtBu Cu exhibits greater intensity in
ν7a vibration has a significant C-O bond stretching component,
4
and its shift to higher energy relative to that of the precursor
complex is consistent with the increased double bond character
expected for the oxidized species.29,39 In addition, there are
weaker features at ca. 1200 and near 1400 cm-1 that resemble
the positions and relative intensities of vibrations found in the
active forms of GAO and GLO which are attributed to the
coordinated tyrosyl radical.2,29b
the 500-700 nm region than shown by those prepared from
II
the LtBu Cu complexes (Figure 4), presumably due to the
2
presence of mixed phenolato-phenoxyl radical ligation. We
speculate that a low-energy PhO- f CuII LMCT (significantly
shifted to longer wavelength than in the reduced precursor LtBu
4
-
Cu, which has higher energy metal-based acceptor orbitals) and
a ligand field transition(s) are major contributors to this
absorption intensity in the monoradical.36 Additional intensity
Conclusions
increases in this region, as well as new features, appear upon
+
further oxidation of [LtBu M] (M ) Cu or Zn) with CeIV (data
4
By using new ligands comprised of one or two phenolate
donors appended to the strongly binding 1,4,7-triazacyclononane
frame, we have prepared and characterized a series of CuII
complexes that model the inactive forms of GAO and GLO.
The effect of the cysteine modification of the equatorial
tyrosinate ligand in the proteins was probed through the study
of isostructural series of complexes of ligands having varying
appendages, including biomimetic o-thioethers, on the phenolate
ligand. The alkylthio group was found to induce relatively small
shifts of the PhO- f CuII LMCT band and the CuII-phenolate/
CuII-phenoxyl radical redox potential consistent with increased
electron density on the ring relative to alkyl-substituted reference
not shown). These observations are consistent with the
electrochemical results described above that suggest that a
unique MII-(phenoxyl)2 species is generated upon two-electron
oxidation of LtBu M (under investigation). The UV-vis data
4
for the ZnII-phenoxyl radical compounds (cf. Figure 7) are
similar to those reported for [Zn(BIDPhE)Cl2],11,12 which
contains a pendant, noncoordinated 2,4,6-trisubstituted phenoxyl
radical and has absorption maxima at 378 nm (ꢀ ) 1500 M-1
cm-1), 394 nm (1700), and 638 nm (430). These features are
(36) A CuII-mediated phenolate-to-phenoxyl radical CT transition may
also be considered as contributing to the added intensity, although this type
of process ascribed to the proteins occurs at significantly longer wavelength
(λmax ) 800 nm).29b We anticipate that the disposition of the phenolate
and phenoxyl ligands would affect the energy and intensity of such a CT,
(37) Fujita, S.; Steenken, S. J. Am. Chem. Soc. 1981, 103, 2540-2545.
(38) Bender, C; Sahlin, M; Babcock, G. T.; Chandrashekov, T. L.;
Sabowe, S. P.; Stubbe, J.; Lindstrom, B.; Petersson, L.; Ehrenberg, A.;
Sjoberg, B.-M. J. Am. Chem. Soc. 1989, 111, 8076-8083.
(39) Tripathi, G. N. R.; Schuler, R. H. J. Phys. Chem. 1988, 92, 5129-
5133.
however, and based on the crystallographic data for the LtBu M complexes
4
that show a difference between the relative geometries of these ligands
compared to the proteins a corresponding divergence in the CT bands would
not be surprising.

Synthetic Models of Galactose and Glyoxal Oxidases
J. Am. Chem. Soc., Vol. 119, No. 35, 1997 8225
4-Methyl-2-(methylthio)phenol.21 A 100 mL Schlenk flask fitted
with a gas adaptor and a pressure equalizing addition funnel was
charged with 6.60 g of KMnO4; this apparatus was connected via a
Teflon cannula to another 100 mL Schlenk flask which was charged
with H2SO4 (22 mL). The flask containing the H2SO4 was cooled to
0 °C, and dimethyl sulfide (DMS; 7.0 g, 113 mmol) was added via
syringe. After the entire apparatus was purged with N2 for 30 min,
HCl (40 mL) was placed in the dropping funnel, and the tip of the
Teflon cannula connecting the two halves of the reaction apparatus
placed below the surface of the H2SO4/DMS solution. The HCl was
then added dropwise to the solid KMnO4 with vigorous stirring, causing
the evolution of Cl2 (g), which was swept into the H2SO4/DMS mixture
with a gentle N2 flow. The addition of HCl took place over 45 min,
and the H2SO4/DMS mixture was maintained at 0 °C. Subsequently,
4-methylphenol (5.0 g, 46.2 mmol) was added to the chlorodimethyl-
sulfonium bisulfate mixture via syringe over a period of 1 h, causing
the evolution of HCl(g). After the addition of 4-methylphenol, the
mixture was removed from the ice bath and was stirred at room
temperature for 1 h. The yellow mixture was then poured into a solution
of NaCl (25 g) in H2O (160 mL), and the resultant solution heated at
reflux overnight. After the reflux period, the heterogenous mixture
was cooled to room temperature and the insoluble oil removed by
extraction with Et2O (150 mL). The Et2O phase was dried (Na2SO4)
and the solvent evaporated under reduced pressure to yield a yellow
oil. Vacuum distillation (bp 55-65 °C, 0.05 Torr) of the crude product
yielded a colorless oil, 2.45 g (34% based on 4-methylphenol): 1H
NMR (300 MHz, CDCl3) δ 7.27 (d, J ) 1.5 Hz, 1H), 7.03 (dd, J )
2.1, 8.4 Hz, lH), 6.86 (d, J ) 8.4 Hz, 1H), 6.45 (s, 1H), 2.30 (s, 3H),
2.25 (s, 3H) ppm; LREIMS m/z 154 [M+].
systems. The presence of the cysteine modification in the
proteins thus may be rationalized, at least in part, by invoking
analogous electronic influences that help to stabilize the CuII-
tyrosyl radical state. Low-temperature electrochemical and
chemical oxidation of the CuII and ZnII complexes having tert-
butyl protecting groups on the phenolate ligands yielded novel
MII-phenoxyl radical compounds akin to the active forms of
GAO and GLO. Identification of the oxidation as ligand-rather
than metal-centered is based on (i) the nearly identical electro-
chemical behavior of similar CuII and ZnII complexes and (ii)
the observation of UV-vis and resonance Raman signatures
of the phenoxyl radical, with additional corroboration from EPR
and UV-vis data for the paramagnetic ZnII cases. Although
stable for hours at temperatures below -40 °C, the metal-
phenoxyl radical complexes rapidly decompose to unidentified
products at room temperature, presumably via radical dispro-
portionation. This behavior stands in sharp contrast to the tri-
and tetravalent metal complexes of bis(phenolate)mono(phe-
noxyl radical) ligands reported by Wieghardt to be stable at
room temperature.13 We speculate that facile dissociation (or
solvent- or counterion-assisted displacement) of the lower
denticity phenoxyl radical ligand from the CuII or ZnII centers
that have an additional, potentially labile monodentate ligand
yields a “free” radical that is prone to decompose.40
Although progress toward understanding the nature of the
unusual GAO and GLO active sites has been attained through
the characterization of the CuII-phenoxyl radical complexes
we have prepared, shortcomings of the complexes as structural,
spectroscopic, and functional models need to be addressed. For
example, the intense, low-energy optical absorption (λmax ) 800
nm, ꢀ 3200 M-1 cm-1) attributed to charge transfer between
the axial tyrosinate and the equatorial tyrosyl radical in active
GAO and GLO has yet to be seen in a synthetic compound,
necessitating deeper exploration of the oxidation chemistry of
complexes with multiple phenolate donors. Oxidation of a
mixed CuII-alcoholato/phenolato complex had been shown
previously to yield aldehyde in a reaction modeling GAO
function,15 and in a recent report treatment of an exogenous
alcohol with a preformed CuII-phenoxyl radical to generate
aldehyde in a model “single-turnover” reaction was reported.14a,41
Mechanistic understanding of the reactions is lacking, however.
Current research continues to be inspired by these and other
issues that are pertinent not only to GAO and GLO structure
and function, but also within a broader context to the funda-
mental chemistry of metal-radical arrays in biological systems.8
4-tert-Butyl-2-(methylthio)phenol. This compound was prepared
in a fashion similar to that described above for 4-methyl-2-(methylthio)-
phenol, substituting 4-tert-butylphenol. Vacuum distillation afforded
the pure product as a colorless oil (bp 88-90 °C, 0.05 Torr): 1H NMR
(300 MHz, CDCl3) δ 7.50 (d, J ) 2.1 Hz, 1H), 7.29 (dd, J ) 2.1, 8.7
Hz, 1H), 6.93 (d, J ) 8.7 Hz, 1H), 6.51 (s, 1H), 2.34 (s, 3H), 1.30 (s,
9H) ppm; LREIMS m/z 196 [M+].
Synthesis of Ligands HLRR′
. The following procedure for the
preparation of 1-(2-hydroxy-3,5-dimethylbenzyl)-4,7-diisopropyl-1,4,7-
triazacyclononane (HLMe ) is typical for the class, with the only variation
2
for the different examples being the phenol used in the condensation
reaction (Scheme 1). To a solution of 1,4-diisopropyl-1,4,7-triazacy-
clononane17 (1.057 g, 4.95 mmol) in MeOH (15 mL) was added aqueous
formaldehyde (0.538 g, 6.64 mmol), and the resultant solution heated
at reflux for 1.5 h. A solution of 2,4-dimethylphenol (0.630 g, 5.16
mmol) in MeOH (5 mL) was added, and refluxing was continued
overnight. After cooling to room temperature, H2O (5 mL) was added
to the yellow solution, causing the deposition of a colorless oil. This
oil was collected by carefully decanting away the supernatant, and
subsequently redissolved in Et2O (20 mL) and washed with 3 M NaOH.
The Et2O phase was removed and dried with Na2SO4; evaporation of
the solvent under reduced pressure yielded the product as a colorless
oil which crystallized on standing (1.224 g, 71%): 1H NMR (500 MHz,
CDCl3) δ 6.81 (s, 1H), 6.60 (s, 1H), 3.73 (s, 2H), 2.98 (br m, 4H),
2.88 (heptet, J ) 6.5 Hz, 2H), 2.67 (t, J ) 4.5 Hz, 4H), 2.43 (s, 4H),
2.19 (s, 3H), 2.17 (s, 3H), 0.95 (d, J ) 6.5 Hz, 12H) ppm; HREIMS
calcd for C21H37N3O 347.2936, found 347.2936. Anal. Calcd for
C2lH37N3O: C, 72.58; H, 10.73; N, 12.09. Found: C, 71.95; H, 10.52;
N, 11.97.
Experimental Section
General Procedures. All reagents and solvents were obtained from
commercial sources and used as received unless otherwise noted.
Solvents were dried according to published procedures and distilled
under N2 immediately prior to use. Commercial NaH (60% dispersion
in oil) was washed with pentane to remove the dispersant, dried under
vacuum, and stored under a dry, inert atmosphere. All transition metal
complexes were synthesized in a Vacuum Atmospheres inert atmo-
sphere glovebox under a dry N2 atmosphere. Spectroscopic and
analytical data were collected as described previously.17b Quantitation
1-(3,5-Di-tert-butyl-2-hydroxybenzyl)-4,7-diisopropyl-1,4,7-triaza-
1
cyclononane (HLtBu ): Yield 90%; H NMR (300 MHz, CDCl3) δ 7.17
2
(d, J ) 2.4 Hz, 1H), 6.81 (d, J ) 2.4 Hz, 1H), 3.76 (s, 2H), 3.04-2.85
(overlapping m and heptet (J ) 6.6 Hz), 6H), 2.68 (t, J ) 4.8 Hz, 4H),
2.46 (s, 4H), 1.42 (s, 9H), 1.28 (s, 9H), 0.97 (d, J ) 6.6 Hz, 12H)
ppm; LREIMS m/z 431 [M+]. Anal. Calcd for C27H49N30: C, 75.12;
H, 11.44; N, 9.73. Found: C, 74.61; H, 11.22; N, 9.24.
of EPR signal intensity was accomplished by comparison to standard
25
solutions of [LiPr Cu(O3SCF3)(H2O)]CF3SO3 in 1:1 CH2Cl2/toluene
3
or CuS04 in glycerol/H2O (for copper complexes) or 2,2,6,6-tetramethyl-
1-piperidinyloxy (TEMPO) in toluene (for the zinc-phenoxyl radical
complexes).
1-(2-Hydroxy-3-methoxy-5-methylbenzyl)-4,7-diisopropyl-1,4,7-
triazacyclononane (HLMeOMe): yield 63%; 1H NMR (300 MHz,
CDCl3) δ 8.99 (br, 1H), 6.57 (s, 1H), 6.37 (s, 1H), 3.82 (s, 3H), 3.76
(s, 2H), 3.01 (m, 4H), 2.88 (heptet, J ) 6.6 Hz, 2H), 2.69 (t, J ) 4.8
Hz, 4H), 2.43 (s, 4H), 2.22 (s, 3H), 0.94 (d, J ) 6.6 Hz, 12H) ppm;
HREIMS calcd for C21H37N3O2 363.2886, found 363.2888. Anal. Calcd
for C21H37N3O2: C, 69.38; H, 10.26; N, 11.56. Found: C, 68.75; H,
10.22; N, 11.44.
(40) Presumably, this decomposition is facilitated by the presence of
unpaired spin density at the unprotected benzylic position (where the
phenolate is linked to the macrocycle). The design and synthesis of new
ligands protected at this position is underway (Jazdzewski, B. A., Tolman,
W. B. Unpublished results).
(41) We have explored the stoichiometic reactions of the CuII-phenoxyl
radical complexes described herein with alcohols, but have yet to observe
significant yields of aldehydes.

8226 J. Am. Chem. Soc., Vol. 119, No. 35, 1997
Halfen et al.
1-(2-Hydroxy-5-methyl-3-(methylthio)benzyl)-4,7-diisopropyl-
LtBu CuCl (86% yield): EPR (1:1 CH2Cl2/toluene, 77 K, 9.46 GHz)
g| ) 2.26, A| ) 161 × 10-4 cm-1, g ) 2.04. Anal. Calcd for
C27H48N3OCuCl: C, 61.22; H, 9.13; N, 7.93. Found: C, 60.85; H,
8.96; N, 7.90.
2
1
1,4,7-triazacyclononane (HLMeSMe): yield 70%; HNMR (300 MHz,
CDCl3) δ 11.16 (br, 1H), 6.86 (d, J ) 1.8 Hz, 1H), 6.58 (d, J ) 1.8
Hz, 1H), 3.74 (s, 2H), 3.02 (br m, 4H), 2.89 (heptet, J ) 6.6 Hz, 2H),
2.67 (t, J ) 4.8 Hz, 4H), 2.42 (s, 4H), 2.41 (s, 3H), 2.22 (s, 3H), 0.94
(d, J ) 6.6 Hz, 12H) ppm. Anal. Calcd for C21H37N3OS: C, 66.45;
H, 9.82; N, 11.07; S, 8.45. Found: C, 66.26; H, 9.81; N, 11.15; S,
8.54.
LMeOMeCuCl (87% yield): EPR (1:1 CH2Cl2/toluene, 77 K, 9.46
GHz) g| ) 2.26, A| ) 158 × 10-4 cm-1, g ) 2.02. Anal. Calcd for
C21H36N3O2CuCl‚0.5 CH2Cl2: C, 51.24; H, 7.40; N, 8.34. Found: C,
51.12; H, 7.50; N, 8.05.
1-(4-tert-Butyl-2-hydroxy-3-(methylthio)benzyl)-4,7-diisopropyl-
1,4,7-triazacyclononane (HLtBuSMe): yield 55%; 1H NMR (500 MHz,
CDC13) δ 7.11 (d, J ) 2.0 Hz, 1H), 6.78 (d, J ) 2.0 Hz, 1H), 3.78 (s,
2H), 3.03 (br m, 4H), 2.90 (heptet, J ) 6.5 Hz, 2H), 2.68 (t, J ) 4.8
Hz, 4H), 2.43 (br s, 7H), 1.26 (s, 9H), 0.95 (d, J ) 6.5 Hz, 12H) ppm.
Anal. Calcd for C24H43N3OS: C, 68.36; H, 10.28; N, 9.96. Found:
C, 67.79; H,10.00; N, 9.74.
LtBuSMeCuCl‚H2O (72% yield): EPR (1:1 CH2Cl2/toluene, 77 K,
9.46 GHz) g| ) 2.26, A| ) 157 × 10-4 cm-1, g ) 2.06. Anal. Calcd
for C24H42N3OSCuCl‚H2O: C, 53.71; H, 8.27; N, 7.83. Found: C,
54.12; H, 7.83; N, 7.74.
LMe ZnCl (60% yield): 1H NMR (300 MHz, CDCl3) δ 6.81 (d, J
2
) 1.5 Hz, 1H), 6.52 (d, J ) 1.5 Hz, 1H), 3.90-3.74 (overlapping singlet
and heptet, 4H), 3.16-3.05 (m, 2), 2.75-2.60 (m, 6H), 2.60-2.50 (m,
2H), 2.35-2.25 (m, 2H), 2.19 (s, 3H), 2.12 (s, 3H), 1.18 (d, J ) 6.6
Hz, 6H), 0.98 (d, J ) 6.6 Hz, 6H) ppm. Anal. Calcd for C21H36N3-
OClZn‚0.5 H2O: C, 55.27; H, 8.17; N, 9.21. Found: C, 55.01; H,
7.88; N, 9.24.
1-Isopropyl-1,4,7-triazacyclononane. To a solution of 1,4-ditosyl-
1,4,7 triazacyclononane22 (11.99 g, 0.027 mol) in distilled CH3CN (120
rnL) was added 2-bromopropane (7.40 g, 0.060 mol), Na2CO3 (6.49 g,
0.060 mol), and tetrabutylammonium bromide (20 mg). The resulting
solution was heated at reflux under N2 for 36 h. The reaction mixture
was then cooled to ambient temperature and filtered, and the remaining
solids were washed with CH2Cl2 (100 mL). The combined organic
filtrates were dried over MgSO4. Following filtration, the solvents were
removed under reduced pressure, yielding a thick, pale yellow oil. The
isolated oil was then dissolved in 18 M H2SO4 (60 mL), and the
resulting dark brown solution was heated at 120 °C under N2 for 40 h.
After cooling to room temperature, the dark mixture was poured into
crushed ice (100 g), and aqueous NaOH was added until the pH of the
solution exceeded 11. The brown solution was then extracted with
CHCl3 (4 × 300 mL). The combined CHCl3 fractions were dried
(MgSO4), and the solvent removed under reduced pressure to yield a
pale yellow oil. Purification by vacuum distillation (80-84 °C, 0.20
Torr) yielded the product as a clear colorless oil (1.97 g, 57%): 1H
NMR (CDCl3, 300 MHz) δ 2.85 (heptet, J ) 6.6 Hz, 1H), 2.73 (s,
4H), 2.67 (m, 4H), 2.52 (m, 4H), 2.10 (br, 2H), 0.99 (d, J ) 6.6 Hz,
6H) ppm; HREIMS calcd for C9H21N3 171.1735, found 171.1737. Anal.
Calcd for C9H21N3: C, 63.09; H, 12.36; N, 24.54. Found: C, 61.81;
H, 12.33; N, 24.05.
LtBu ZnCl (96% yield): 1H NMR (300 MHz, CD2Cl2) δ 7.20 (d, J
2
) 2.4 Hz, 1H), 6.80 (d, J ) 2.4 Hz, 1H), 3.89 (heptet, J ) 6.6 Hz,
2H), 3.81 (s, 2H), 3.25-3.23 (m, 2H), 2.98-2.93 (m, 2H), 2.71-2.64
(m, 6H), 2.48-2.41 (m, 2H), 1.86 (br., 2H, H2O), 1.49 (s, 9H), 1.29
(s, 9H), 1.18 (d, J ) 6.6 Hz, 6H), 1.12 (d, J ) 6.6 Hz, 6H) ppm. Anal.
Calcd for C27H48N3OClZn‚H2O: C, 59.01; H, 9.17; N, 7.65. Found:
C, 59.10; H, 9.17; N, 7.65.
LtBuSMeZnCl (93% yield): 1H NMR (500 MHz, CDCl3) δ 7.07 (d,
J ) 2.0 Hz, 1H), 6.71 (d, J ) 2.0 Hz, 1H), 3.88 (heptet, J ) 6.5 Hz,
2H), 3.82 (s, 2H), 3.18 (m, br, 2H), 2.74 (m, br, 4H), 2.72 (m, br, 2H),
2.59 (m, br, 2H), 2.43 (s, 3H), 2.35 (m, br, 2H), 1.26 (s, 9H), 1.24 (d,
J ) 6.5 Hz, 6H), 1.05 (d, J ) 6.5 Hz, 6H) ppm. Anal. Calcd for
C24H42N3OSZnCl: C, 55.47; H, 8.15; N, 8.09. Found: C, 55.29; H,
8.14; N, 8.12.
General Method for the Preparation of the Complexes LRR′Cu-
(O2CCH3). These complexes were prepared in a fashion similar to
that of LRR′CuCl, substituting anhydrous Cu(O2CCH3)2. Recrystalli-
zation by pentane diffusion into solutions of the complexes in CH2-
Cl2/toluene (1:2) yielded the LRR′Cu(O2CCH3) complexes as brown or
green crystalline solids in the yields noted.
1,4-bis(2-hydroxy-3,5-di-tert-butylbenzyl)-7-isopropyl-1,4,7-tri-
azacyclononane (H2LtBu ). To a solution of 1-isopropyl-1,4,7-triaza-
LtBu Cu(O2CCH3) (65% yield): EPR (1:1 CH2Cl2/toluene, 77 K,
4
2
9.46 GHz) g| ) 2.26, A| ) 175 × 10-4 cm-1, g ) 2.03; FAB-MS
cyclononane (0.300 g, 1.75 mmol) in MeOH (5 rnL) was added 37%
aqueous formaldehyde (0.355 g, 4.38 mmol). This solution was
refluxed under N2 for 3 h. A solution of 2,4-di-tert-butylphenol (0.753g,
3.65 mmol) in MeOH (2 mL) was then added. The resulting solution
was refluxed for an additional 14 h. Upon cooling the solution to room
temperature, a white precipitate began to form. Addition of 5 mL of
H2O to the light yellow solution induced the precipitation of an
additional amount of white solid. After the supernatant was decanted
away, the solid was dissolved in Et2O (20 mL) and washed with 3 M
NaOH (15 mL). The Et2O layer was removed and dried over MgSO4,
and the solvent was then evaporated under reduced pressure to yield
the product as a white solid (0.732 g, 69%): 1H NMR (300 MHz,
CDCl3) δ 10.44-11.31 (br, 2H), 7.19 (d, J ) 2.4 Hz, 2H), 6.77 (d, J
) 2.1 Hz, 2H), 3.73 (s, 4H), 3.02 (s, 4H), 2.93 (heptet, J ) 6.6 Hz,
1H), 2.71 (m, br, 4H), 2.59 (m, br, 4H), 1.41 (s, 18H), 1.25 (s, 18H),
0.96 (d, J ) 6.6 Hz, 6H) ppm; HRCIMS calcd for C39H65N3O2
607.5076, found 607.5063. Anal Calcd. for C39H65N3O2: C, 77.05;
H, 10.78; N, 6.91. Found: C, 76.09; H, 10.61; N, 6.65.
(MNBA) m/z 552 [M+].
LMeOMeCu(O2CCH3)‚CH2Cl2 (85% yield): EPR (1:1 CH2Cl2/
toluene, 77 K, 9.46 GHz) g| ) 2.27, A| ) 167 × 10-4 cm-1, g
)
2.04. Anal. Calcd for C24H41N3O4Cl2Cu: C, 50.57; H, 7.23; N, 7.37.
Found: C, 50.57; H, 7.12; N, 7.33.
LMeSMeCu(O2CCH3)‚CH2Cl2 (75% yield): EPR (1:1 CH2Cl2/
toluene, 77 K, 9.46 GHz) g| ) 2.27, A| ) 166 × 10-4 cm-1, g
)
2.02. Anal. Calcd for C24H41N3O3SCl2Cu: C, 49.18; H, 7.05; N, 7.17.
Found: C, 49.81; H, 6.84; N, 7.20.
tBu
[LtBu Cu(CH3CN)]CF3SO3. To a solution of Na[L ] in THF (5
2
2
mL), prepared by treatment of LtBu (0.319 g, 0.74 mmol) with excess
2
NaH followed by filtration, was added a solution of Cu(CF3SO3)2 (0.262
g, 0.72 mmol) dissolved in CH3CN (1 mL), causing a color change to
deep purple. After stirring for 30 min, the solvent was removed under
reduced pressure, and the resultant purple residue redissolved in CH2-
Cl2 (5 mL) and filtered through Celite, and the filtrate evaporated.
Recrystallization of the purple solid from a mixture of CH3CN/toluene/
pentane (1:5:10) at -20 °C for 5-7 days afforded the pure product as
purple crystals (0.45 g, 93%): EPR (1:1:0.1 CH2Cl2/toluene/CH3CN,
77 K, 9.46 GHz) g| ) 2.25, A| ) 160 × 10-4 cm-1, g ) 2.03; ISP-
MS (CH2Cl2) m/z 493 [M - CH3CN - CF3SO3+]. Anal. Calcd for
C30H51N404F3SCu: C, 52.65; H, 7.51; N, 8.19. Found: C, 52.21; H,
General Method for the Preparation of the Complexes LRR′MCl
(M ) Cu or Zn). A solution of LRR′ (typically 0.15 g) in THF (3
rnL) was treated with excess NaH, causing the evolution of H2; stirring
was continued until gas evolution ceased. The mixture was then
filtered, and the filtrate treated with solid anhydrous MCl2 (M ) Cu or
Zn, 1.0 equiv), in the case of M ) Cu causing the development of a
deep purple color. After stirring for 2 h, the mixture was filtered
through a pad of Celite and the filtrate evaporated to dryness.
Recrystallization by diffusion of pentane into a solution of the complex
in CH2Cl2/toluene (1:2) yielded the LRR′MCl complexes as purple (M
) Cu) or colorless (M ) Zn) crystalline solids in the yields noted.
7.51; N, 8.31.
tBu
LtBu M‚MeOH. To a solution of H2L (0.300 g, 0.493 mmol) in
THF (5 mL) under nitrogen was added excess NaH (0.048 g, 2.00
mmol). The resultant mixture was stirred until H2 evolution ceased (2
h). The light yellow solution was filtered through a pad of Celite, and
the filtrate was treated with a THF suspension (4 mL) of anhydrous
MCl2 (M ) Cu or Zn, 1 equiv). This caused the formation of a murky
green (M ) Cu) or light yellow (M ) Zn) solution which was stirred
overnight at ambient temperature. The solution was then filtered
through a pad of Celite and the THF removed under reduced pressure.
4
4
LMe CuCl (70% yield): EPR (1:1 CH2Cl2/toluene, 77 K, 9.46 GHz)
2
g| ) 2.25, A| ) 160 × 10-4 cm-1, g ) 2.03; FAB-MS (MNBA) m/z
444 [M+]. Anal. Calcd for C21H36N3OCuCl‚0.5 H2O: C, 55.49; H,
8.20; N, 9.43. Found: C, 55.50; H, 7.87, N, 9.20.

Synthetic Models of Galactose and Glyoxal Oxidases
J. Am. Chem. Soc., Vol. 119, No. 35, 1997 8227
2
The green powder obtained was recrystallized by slow evaporation of
a MeOH/H2O (70:30) solution over the course of 3 days, yielding dark
green (M ) Cu) or light yellow (M ) Zn) crystals.
For [LtBu Cu(CH3CN)]CF3SO3‚C7H8, a 6.9% decrease in the intensities
of these reflections was noted, and the data were scaled to adjust for
this decay. The data were corrected for Lorentz and polarization effects.
Absorption corrections were determined from a series of azimuthal scans
of three intense reflections for each complex. Diffraction data for a
LtBu Cu‚MeOH (54% yield): EPR (1:1 CH2Cl2/toluene, 77 K, 9.46
4
GHz) g| ) 2.25, A| ) 169 × 10-4 cm-1, g ) 2.03; FAB-MS (MNBA)
m/e 669 (M+, 100). Anal. Calcd for C39H63N3O2Cu‚CH3OH: C, 68.48;
H, 9.63; N, 5.99. Found: C, 68.52; H, 9.67; N, 6.12.
green-brown plate crystal of LtBu Cu‚CH3OH were collected using a
4
Siemens SMART platform CCD diffractometer at 173(2) K. Initial
cell constants and an orientation matrix were determined from 60
reflections harvested from 20 frames. Final cell constants were
calculated from a set of 8192 strong reflections from the actual data
collection. An absorption correction was determined using the program
SADABS.42
LtBu Zn‚MeOH (56% yield): 1H NMR (500 MHz, CDCl3) δ 7.22
4
(d, J ) 2.5 Hz, 2H), 6.78 (d, J ) 2.5 Hz, 2H), 4.58 (d, J ) 11.5 Hz,
2H), 3.46 (s, 3H), 3.39 (d, J ) 11.5 Hz, 2H), 3.38 (heptet, J ) 6.5 Hz,
1H), 3.11 (m, br, 2H), 2.80 (m, br, 4H), 2.51 (m, br, 4H), 2.33 (m, br,
2H), 1.51 (s, 18H), 1.27 (s, 18H), 0.84 (d, J ) 6.5 Hz, 6H) ppm. Anal.
Calcd for C39H63N3O2Zn‚CH3OH: C, 68.31; H, 9.60; N, 5.97. Found:
Space group determinations were made on the basis of systematic
absences and intensity statistics. The structures were solved by direct
methods which provided the positions of most of the non-hydrogen
atoms. The remainder of the non-hydrogen atoms were located
following several full-matrix least squares/difference Fourier cycles.
All non-hydrogen atoms were refined with anisotropic displacement
parameters unless noted below. Hydrogen atoms were placed in
calculated positions and refined as riding atoms with fixed isotropic
displacement parameters [1.2 × host atom U(eq) for methylene and
methine, 1.5 × host atom U(eq) for methyl and hydroxylic protons].
The carbon atoms of the toluene molecule located in the asymmetric
C, 68.58; H, 9.55; N, 5.92.
tBu
Coulometric oxidations of [LtBu Cu(CH3CN)]CF3SO3 and L Cu.
2
4
A three-chamber electrochemical cell was charged with a solution of
-3
[LtBu Cu(CH3CN)]CF3SO3 (0.0045 g, 7.0 × 10 mmol) or LtBu Cu‚
MeOH (0.0062 g, 8.8 × 10-3 mmol) in CH2Cl2 (0.4 M TBAPF6), fitted
with a platinum gauze working electrode, a silver wire pseudoreference
electrode, and a platinum wire auxiliary electrode, and then cooled to
-40 °C under an atmosphere of dry N2. Coulometric oxidation was
2
4
then performed at 1.0 V (LtBu complex) or 0.71 V (LtBu complex),
2
4
resulting in the removal of 0.8 ( 0.1 e- mol-1 (LtBu ) or 1.0 ( 0.1 e
-
2
mol-1 (LtBu ) of the compound and a color change to deep green (L
unit of [LtBu Cu(CH3CN)]CF3SO3‚C7H8 were refined isotropically.
tBu
4
2
)
2
or dark blue (LtBu ). For the L case, an aliquot of the green solution
was removed by syringe and transferred to a precooled EPR tube, and
the X-band EPR spectrum subsequently recorded at 77 K. Analysis
revealed the presence of a residual axial Cu(II) signal superimposed
on a weak isotropic signal at g ) 2.00; double integration of the entire
signal revealed a 92 ( 5% decrease in the signal intensity compared
to that of the original sample. Concurrent UV-vis analysis of the green
solution (-40 °C) revealed the presence of new absorption features
Although the carbon atoms of the tert-butyl group para to the phenolic
oxygen atom exhibit large anisotropic displacement parameters, a
suitable split atom model could not be determined. A CH2Cl2 molecule
was located in the asymmetric unit of LMeSMeCu(O2CCH3)‚CH2Cl2, and
a partially occupied (0.32) water molecule was identified in the
tBu
4
2
asymmetric unit of LMe ZnCl‚0.32 H2O. Methanol molecules were
2
found to be hydrogen bonded to phenolate oxygens in the asymmetric
tBu
units of LtBu Cu‚CH3OH and L Zn‚CH3OH. In the latter complex,
one tert-butyl group was found to be rotationally disordered over two
positions in a 50:50 ratio. Bond length restraints were applied to the
carbon atoms of the disordered groups.
4
4
with λmax ) 410 (ꢀ ) 3500 M-1 cm-1) and 648 (ꢀ ) 1000 M-1 cm-1
)
nm.
+
tBu
2
Chemical Oxidation of [LtBu Cu(CH3CN)] , L ZnCl, and
2
LtBu M (M ) Cu or Zn). Solutions of the MII-phenoxyl radical
All calculations were performed using the SHELXTL-Plus V5.0 suite
of programs on an SGI INDY R4400-SC computer.43 Important
crystallographic details are summarized in Table 1, and relevant bond
lengths and angles are found in Table 2. Full details of the structure
determinations, including fractional atomic coordinates, and full tables
of bond lengths and angles, anisotropic displacement parameters, torsion
angles, and hydrogen atom coordinates are included in the Supporting
Information.
4
species derived from these precursors were prepared by treatment with
(NH4)2[Ce(NO3)6] in CH2Cl2/CH3CN or CH2Cl2/toluene/CH3CN mix-
tures at -40 °C or with [thianthryl]BF4 in CH2Cl2 at -80 °C (see Table
3 and Figures 4, 5, and 7 for spectroscopic data). For example, a
solution of [LtBu Cu(CH3CN)]CF3SO3 (0.013 g, 0.02 mmol) in CH2Cl2
2
(5.0 mL) was cooled to -40 °C under a N2 atmosphere and treated
with a solution of (NH4)2[Ce(NO3)6] (0.011 g, 0.02 mmol) dissolved
in CH3CN (1.5 mL), causing a color change from purple to deep green.
The sample for resonance Raman spectroscopy was produced by treating
Acknowledgment. Funding for this work was provided by
the National Institutes of Health (Grant GM47365 to W.B.T.,
Grant GM33162 to L.Q., a predoctoral traineeship to E.C.W.,
and a postdoctoral fellowship to L.M.B.), the National Science
Foundation (NYI award to W.B.T. and Grant CHE-9413114
for partial support for the purchase of the Siemens SMART
system), the University of Minnesota (dissertation fellowship
to J.A.H.), and the Alfred P. Sloan and Camille & Henry
Dreyfus Foundations (fellowships to W.B.T.).
a solution of [LtBu Cu(CH3CN)]CF3SO3 (0.026 g, 0.04 mmol) in CD3-
2
CN (2.0 mL) with a solution of (NH4)2[Ce(NO3)6] (0.022 g, 0.04 mmol)
in CD3CN (1.5 mL) at -40 °C, yielding a final concentration of
2+
+
[(LtBu )Cu(CH3CN)] of 11.5 mM. Solutions of [LtBu ZnCl] and
2
2
[LtBu M] (M ) Cu or Zn) were prepared similarly, except for [LtBu
-
+
4
2
ZnCl]+ a 1:4 CH2Cl2/toluene solution of LtBu ZnCl was used and for
2
[LtBu Zn] a 1:1 CH2Cl2/toluene solution of LtBu Zn was used. The
+
4
4
+
solution of [LtBu Cu] was dark blue, and those of the Zn compounds
4
were light green. EPR data for [LtBu ZnCl]+ and [LtBu Zn]+ (∼1:1:0.1
CH2Cl2/toluene/CH3CN, 77 K, 9.46 Ghz, field modulation amplitude
1.0 G, microwave power 20 µW): giso ) 2.00.
2
4
Supporting Information Available: FTIR and 13C{1H}
NMR spectral data and complete drawings and full details of
the X-ray structure determinations, including tables of bond
lengths and angles, atomic positional parameters, and final
thermal parameters (87 pages). See any current masthead page
for ordering and Internet access instructions.
X-ray Crystallography. A purple plate crystal of [LtBu Cu(CH3-
2
CN)]CF3SO3‚C7H8, a colorless prism of LMe ZnCl‚0.32H2O, a yellow
2
block crystal of LtBu Zn‚CH3OH, and a green brown prismatic crystal
4
of LMeSMeCu(O2CCH3)‚CH2Cl2 were fixed to thin glass fibers with epoxy
resin and mounted on an Enraf-Nonius CAD4 diffractometer for data
collection at 293(2) K. Cell constants and orientation matrices were
determined by least-squares refinement of the setting angles of 25
carefully centered reflections. The intensities of three standard reflec-
tions were monitored hourly in order to calculate corrections for decay.
JA9700663
(42) Blessing, R. H. Acta Crystallogr. 1995, A51, 33-38.
(43) SHELXTL V5.0, Siemens Energy & Automation, Inc., Madison,
WI 53719-1173.