Structural, spectroscopic, and theoretical characterization of bis(μ-oxo)dicopper complexes, novel intermediates in copper-mediated dioxygen activation

J. Am. Chem. Soc. 1996, 118, 11555-11574

11555

Structural, Spectroscopic, and Theoretical Characterization of

Bis(µ-oxo)dicopper Complexes, Novel Intermediates in

Copper-Mediated Dioxygen Activation

Samiran Mahapatra, Jason A. Halfen, Elizabeth C. Wilkinson, Gaofeng Pan,

Xuedong Wang, Victor G. Young, Jr., Christopher J. Cramer,*

Lawrence Que, Jr.,* and William B. Tolman*

Contribution from the Department of Chemistry, Center for Metals in Biocatalysis, and

Supercomputer Institute, UniVersity of Minnesota, 207 Pleasant St. SE,

Minneapolis, Minnesota 55455

ReceiVed July 5, 1996^X

Abstract: A description of the structure and bonding of novel bis(µ-oxo)dicopper complexes and their bis(µ-hydroxo)-

dicopper decomposition products was derived from combined X-ray crystallographic, spectroscopic, and ab initio

theoretical studies. The compounds [(LCu)₂(µ-O)₂]X₂were generated from the reaction of solutions of [LCu-

(CH₃CN)]X with O₂at -80 °C (L ) 1,4,7-tribenzyl-1,4,7-triazacyclononane, L^Bn; 1,4,7-triisopropyl-1,4,7-

3

triazacyclononane, L^iPr; or 1-benzyl-4,7-diisopropyl-1,4,7-triazacyclononane, L^{iPr Bn}; X ) variety of anions). The

3

2

geometry of the [Cu₂(µ-O)₂]²⁺core was defined by X-ray crystallography for [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂and by

3

EXAFS spectroscopy for the complexes capped by L^Bnand L^iPr; notable dimensions include short Cu-O (∼1.80

Å) and Cu‚‚‚Cu (∼2.80 Å) distances like those reported for analogous M₂(µ-O)₂(M ) Fe or Mn) rhombs. The core

geometry is contracted compared to those of the bis(µ-hydroxo)dicopper(II) compounds that result from decomposition

3

of the bis(µ-oxo) complexes upon warming. X-ray structures of the decomposition products [(L^BnCu)(L^{Bn H}Cu)-

3

2

(µ-OH)₂](O₃SCF₃)₂‚2CH₃CO, [(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚2THF, and [(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂showed that

they arise from N-dealkylation of the original capping macrocycles. Manometric, electrospray mass spectrometric,

and UV-vis, EPR, NMR, and resonance Raman spectroscopic data for the bis(µ-oxo)dicopper complexes in solution

revealed important topological and electronic structural features of the intact [Cu₂(µ-O)₂]²⁺core. The bis(µ-oxo)-

dicopper unit is diamagnetic, undergoes a rapid fluxional process involving interchange of equatorial and axial N-donor

ligand environments, and exhibits a diagnostic ∼600 cm^-^{1 18}O-sensitive feature in Raman spectra. Ab initio

calculations on a model system, {[(NH₃)₃Cu]₂(µ-O)₂}²⁺, predicted a closed-shell singlet ground-state structure that

agrees well with the bis(µ-oxo)dicopper geometry determined by experiment and helps to rationalize many of its

physicochemical properties. On the basis of an analysis of the theoretical and experimental results (including a

2

bond valence sum analysis), a formal oxidation level assignment for the core is suggested to be [Cu^III₂(µ-O^2-)₂]²⁺

,

although a more complete molecular orbital description indicates that the oxygen and copper fragment orbitals are

significantly mixed (i.e., there is a high degree of covalency).

The binding and activation of dioxygen by copper ions is

important in numerous and diverse biological and catalytic

processes.¹Biomolecules having one or more copper ions in

their active sites utilize the oxidizing power of O₂for respiration

and for the functionalization of diverse substrates during

metabolically significant biochemical pathways. Copper-

dioxygen interactions also play a central role in synthetically

useful stoichiometric and catalytic oxidative conversions of

organic molecules. Thus, an important research objective that

has attracted the attention of an interdisciplinary community of

scientists is to elucidate the nature and reactivity of copper-

dioxygen adducts.^1,2

been structurally characterized by X-ray crystallography are

shown in Figure 1;^3,4alternative binding modes have been

suggested for additional complexes based on spectroscopic and

kinetic data.⁵A representative demonstration of the successful

interplay of direct biophysical studies and synthetic modeling

work is the development of the now coherent picture of the

reversible binding of O₂to hemocyanin (Hc), the dioxygen

carrier protein found in arthropods and mollusks. In Hc, two

Cu(I) ions, each ligated by three histidyl imidazoles, form a

(2) (a) Solomon, E. I.; Baldwin, M. J.; Lowery, M. D. Chem. ReV. 1992,

92, 521-542. (b) Solomon, E. I.; Tuczek, F.; Root, D. E.; Brown, C. A.

Chem. ReV. 1994, 94, 827-856.

(3) (a) Tyekla´r, Z.; Jacobson, R. R.; Wei, N.; Murthy, N. N.; Zubieta,

J.; Karlin, K. D. J. Am. Chem. Soc. 1993, 115, 2677-2689. (b) Kitajima,

N.; Fujisawa, K.; Fujimoto, C.; Moro-oka, Y.; Hashimoto, S.; Kitagawa,

T.; Toriumi, K.; Tatsumi, K.; Nakamura, A. J. Am. Chem. Soc. 1992, 114,

1277-1291. (c) Fujisawa, K.; Tanaka, M.; Moro-oka, Y.; Kitajima, N. J.

Am. Chem. Soc. 1994, 116, 12079-12080. (d) Reim, J.; Krebs, B. Angew.

Chem., Int. Ed. Engl. 1994, 33, 1969-1971.

The few such Cu_n-O₂units in synthetic compounds that have

* To whom correspondence should be addressed. E-mail: cramer@

umn.edu, que@chem.umn.edu, and tolman@chem.umn.edu. FAX: 612-

624-7029.

^XAbstract published in AdVance ACS Abstracts, November 1, 1996.

(1) For reviews, see: (a) Karlin, K. D.; Gultneh, Y. Prog. Inorg. Chem.

1987, 35, 219-328. (b) Sorrell, T. N. Tetrahedron 1989, 40, 3-68. (c)

Karlin, K. D.; Tyekla´r, Z.; Zuberbu¨hler, A. D. In Bioinorganic Catalysis;

Reedjik, J., Ed.; Marcel Dekker, Inc.: New York, 1993; pp 261-315. (d)

Kitajima, N.; Moro-oka, Y. Chem. ReV. 1994, 94, 737. (e) Karlin, K. D.;

Tyekla´r, Z. AdV. Inorg. Biochem. 1994, 9, 123-172. (f) Fox, S.; Karlin, K.

D. In ActiVe Oxygen in Biochemistry; Valentine, J. S., Foote, C. S.,

Greenberg, A., Liebman, J. F., Eds.; Blackie Academic & Professional,

Chapman & Hall: Glasgow, Scotland, 1995; pp 188-231.

(4) An X-ray crystal structure of a complex comprised of an (η¹-

superoxo)copper(II) unit stabilized by hydrogen bonding interactions with

ligand amide groups has been reported,^4abut the X-ray data were incorrectly

interpreted. A subsequent reanalysis of the data showed the molecule to

contain a terminal hydroxide rather than a superoxide ligand.^4b(a) Harata,

M.; Jitsukawa, K.; Masuda, H.; Einaga, H. J. Am. Chem. Soc. 1994, 116,

10817-10818. (b) Berreau, L.; Mahapatra, S.; Halfen, J. A.; Young, V.

G., Jr.; Tolman, W. B. Inorg. Chem. 1996, 35, 6339-6342.

11556 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

for other metal-promoted oxidations, such as those involving

heme-iron sites,¹¹understanding of these issues has been

inhibited by a dearth of definitively identified intermediates

beyond the Cu_n-O₂(n ) 1 or 2) adduct stage. In synthetic

chemistry, oxygenation of Cu(I) complexes commonly occurs

with a 4:1 Cu:O₂reaction stoichiometry to yield oxo- or

hydroxocopper(II) products.¹²These products generally are

believed to arise via formation of a (peroxo)dicopper(II)

intermediate that is induced to undergo O-O bond cleavage

by the additional Cu(I) ions, which act as sources of electrons

(eq 1). Although examples of structurally characterized bis-

2Cu^I

2

2Cu^I+ O₂f Cu^II-(O₂)-Cu^II

8 2Cu^II-O(H)-Cu^II(1)

(µ-hydroxo)dicopper(II) are numerous,¹³only a few high

nuclearity oxocopper(II) clusters¹⁴and discrete oxodicopper-

(II) species have been identified,^12c,15,16and crystallographically

characterized examples exist only for species comprised of more

than two copper ions.

Increasingly oxidized species resulting from O-O bond

cleavage without further injection of electrons from additional

Cu(I) ions have been even more elusive, yet they are key

intermediates in proposed mechanisms of oxidative transforma-

tions carried out by copper-containing enzymes or synthetic

catalysts.^17,18For example, Klinman and co-workers have

suggested that in dopamine â-monooxygenase a tyrosine residue

and a nearby monocopper hydroperoxide [Cu(II)sOOH] react

Figure 1. Chem3D representations of the existing X-ray structures of

copper-dioxygen adducts.

dioxygen adduct (oxyHc) having a planar (µ-η²:η²-peroxo)-

dicopper(II) core. The geometric features of this [Cu₂(µ-η²:

η²-O₂)]²⁺unit have been defined by X-ray crystallography in

oxyHc⁶and, in greater detail, in the synthetic model complex

iPr₂

[(Tp^iPrCu)₂(O₂)] [Tp

) tris(3,5-diisopropyl-1-pyrazolyl)-

2

hydroborate] (Figure 1).^3bDiagnostic spectroscopic features of

this core include EPR silence due to strong antiferromagnetic

coupling between the Cu(II) ions, a low O-O stretching

frequency in the resonance Raman spectrum [∼750 cm^-1

;

(11) Cytochrome P-450: Structure, Mechanism, and Biochemistry, 2nd

ed.; Ortiz de Montellano, P. R., Ed.; Plenum Press: New York, 1995.

(12) (a) Davies, G.; El-Sayed, M. A. Comments Inorg. Chem. 1985, 4,

151-162. (b) Davies, G.; El-Sayed, M. A. Inorg. Chem. 1983, 22, 1257-

1266. (c) Caulton, K. G.; Davies, G.; Holt, E. M. Polyhedron 1990, 9, 2319-

2351 and references cited therein. (d) Sanyal, I.; Mahroof-Tahir, M.; Nasir,

M. S.; Ghosh, P.; Cohen, B. I.; Gultneh, Y.; Cruse, R. W.; Farooq, A.;

Karlin, K. D.; Liu, S.; Zubieta, J. Inorg. Chem. 1992, 31, 4322-4332. (e)

Lee, D.-H.; Murthy, N. N.; Karlin, K. D. Inorg. Chem. 1996, 35, 804-805

and references cited therein.

(13) (a) Lee, S. C.; Holm, R. H. Inorg. Chem. 1993, 32, 4745-4753

and references cited therein. (b) Hodgson, D. J. Prog. Inorg. Chem. 1975,

19, 173-242. (c) Kitajima, N.; Hikichi, S.; Tanaka, M.; Moro-oka, Y. J.

Am. Chem. Soc. 1993, 115, 5496-5508.

(14) (a) Ross, P. F.; Murmann, R. K.; Schlemper, E. O. Acta Crystallogr.

1974, B30, 1120-1123. (b) Butcher, R. J.; O’Connor, C. J.; Sinn, E. Inorg.

Chem. 1981, 20, 537-545. (c) Erdonmez, A.; van Diemen, J. H.; de Graaff,

R. A. G.; Reedjik, J. Acta Crystallogr. 1990, C46, 402-404. (d) McKee,

V.; Tandon, S. S. J. Chem. Soc., Dalton Trans. 1991, 221-230. (e) Agnus,

Y.; Louis, R.; Metz, B.; Boudon, C.; Gisselbrecht, J. P.; Gross, M. Inorg.

Chem. 1991, 30, 3155-3161 and references cited therein. (f) Davies, G.;

El-Shazly, M. F.; Rupich, M. W.; Churchill, M. R.; Rotella, R. J. J. Chem.

Soc., Chem. Commun. 1978, 1045-1046. (g) Hosokawa, T.; Takano, M.;

Murahashi, S.-I. J. Am. Chem. Soc. 1996, 118, 3990-3991.

∆ν(¹⁸O₂) ) 42 cm^-1], and peroxo f Cu(II) charge transfer

bands in the optical absorption spectrum [λ_max) 360 (ꢀ 20 000

M^-1cm^-1), 510 (ꢀ 1000) nm].^2,7Since its characterization in

[(Tp^iPrCu)₂(O₂)] and oxyHc, the [Cu₂(µ-η :η -O₂)]²⁺fragment

has been indentified in a number of other complexes, all with

facially coordinating, tridentate N-donor capping ligands.^8,9

Moreover, reversible O₂binding behavior that was not observed

2

for the Tp^iPrcomplex has been demonstrated in these instances.

2

The physicochemical characterization data for the copper-

dioxygen adducts shown in Figure 1 in conjunction with

comprehensive kinetic and thermodynamic data for the forma-

tion of these and other complexes^5,10has provided considerable

insight into the first steps involved in dioxygen binding and

activation. Nonetheless, many important issues remain unre-

solved, in particular the nature of the subsequent O-O bond

cleavage that invariably occurs in oxidation reactions. Whether

O-O bond disruption occurs prior to or concomitant with attack

on substrate is another key question. In contrast to the situation

(15) (a) Lapinte, C.; Riviere, H.; Roselli, A. J. Chem. Soc., Chem.

Commun. 1981, 1109-1110. (b) Kitajima, N.; Koda, T.; Moro-oka, Y.

Chem. Lett. 1988, 347-350.

(16) Kitajima, N.; Koda, T.; Iwata, Y.; Moro-oka, Y. J. Am. Chem. Soc.

1990, 112, 8833-8839.

(5) Some selected examples: (a) Karlin, K. D.; Tyekla´r, Z.; Farooq, A.;

Haka, M. S.; Ghosh, P.; Cruse, R. W.; Gultneh, Y.; Hayes, J. C.; Toscano,

P. J.; Zubieta, J. Inorg. Chem. 1992, 31, 1436-1451. (b) Sanyal, I.; Strange,

R. W.; Blackburn, N. J.; Karlin, K. D. J. Am. Chem. Soc. 1991, 113, 4692-

4693. (c) Karlin, K. D.; Wei, N.; Jung, B.; Kaderli, S.; Niklaus, P.;

Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1993, 115, 9506-9514. (d) Jung,

B.; Karlin, K. D.; Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1996, 118, 3763-

3764.

(6) (a) Magnus, K. A.; Hazes, B.; Ton-That, H.; Bonaventura, C.;

Bonaventura, J.; Hol, W. G. J. Proteins: Struct., Funct., Genet. 1994, 19,

302-309. (b) Magnus, K. A.; Ton-That, H.; Carpenter, J. E. Chem. ReV.

1994, 94, 727-735.

(7) Baldwin, M. J.; Root, D. E.; Pate, J. E.; Fujisawa, K.; Kitajima, N.;

Solomon, E. I. J. Am. Chem. Soc. 1992, 114, 10421-10431.

(8) (a) Lynch, W. E.; Kurtz, D. M., Jr.; Wang, S.; Scott, R. A. J. Am.

Chem. Soc. 1994, 116, 11030-11038. (b) Sorrell, T. N.; Allen, W. E.;

White, P. S. Inorg. Chem. 1995, 34, 952-960.

(9) Mahapatra, S.; Halfen, J. A.; Wilkinson, E. C.; Que, L., Jr.; Tolman,

W. B. J. Am. Chem. Soc. 1994, 116, 9785-9786.

(17) Some examples where the involvement of Cu(III)dO T Cu(II)sO^•

or related high oxidation state species have been suggested: (a) Villafranca,

J. J. In Copper Proteins; Spiro, T. G., Ed.; Wiley: New York, 1981; pp

264-289. (b) Capdevielle, P.; Baranne-Lafont, J.; Sparfel, D.; Cuong, N.

K.; Maumy, M. J. Mol. Catal. 1988, 47, 59-66. (c) Re´glier, M.; Amade¨ı,

E.; Tadayoni, R.; Waegell, B. J. Chem. Soc., Chem. Commun. 1989, 447-

450. (d) Capdevielle, P.; Sparfel, D.; Baranne-Lafont, J.; Cuong, N. K.;

Maumy, M. J. Chem. Soc., Chem. Commun. 1990, 565-566. (e) Reinaud,

O.; Capdevielle, P.; Maumy, M. J. Chem. Soc., Chem. Commun. 1990, 566-

568. (f) Capdevielle, P.; Maumy, M. Tetrahedron Lett. 1991, 32, 3831-

3834. (g) Reddy, K. V.; Jin, S.-J.; Arora, P. K.; Sfeir, D. S.; Maloney, S.

C. F.; Urbach, F. L.; Sayre, L. M. J. Am. Chem. Soc. 1990, 112, 2332-

2340 and references cited therein. (h) Itoh, S.; Kondo, T.; Komatsu, M.;

Ohshiro, Y.; Li, C.; Kanehisa, N.; Kai, Y.; Fukuzumi, S. J. Am. Chem.

Soc. 1995, 117, 4714-4715. (i) Chan, S. I.; Nguyen, H.-H. T.; Shiemke,

A. K.; Lidstrom, M. E. In Bioinorganic Chemistry of Copper; Karlin, K.

D., Tyekla´r, Z., Eds.; Chapman & Hall: New York, 1993; pp 184-195.

(j) Ogura, T.; Hirota, S.; Proshlyakov, D. A.; Shinzawa-Itoh, K.; Yoshikawa,

S.; Kitagawa, T. J. Am. Chem. Soc. 1996, 118, 5443-5449.

(10) (a) Wei, N.; Lee, D.-H.; Murthy, N. N.; Tyekla´r, Z.; Karlin, K. D.;

Kaderli, S.; Jung, B.; Zuberbu¨hler, A. D. Inorg. Chem. 1994, 33, 4625-

4626. (b) Cruse, R. W.; Kaderli, S.; Karlin, K. D.; Zuberbuhler, A. D. J.

Am. Chem. Soc. 1988, 110, 6882-6883. (c) Karlin, K. D.; Wei, N.; Jung,

B.; Kaderli, S.; Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1991, 113, 5868-

5870.

(18) Tian, G.; Berry, J. A.; Klinman, J. P. Biochemistry 1994, 33, 226-

234.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11557

to generate a tyrosyl radical and a mono(oxocopper) unit [often

written as Cu(III)dO T Cu(II)sO^•, although the former is

untenable]¹⁹that abstracts a benzylic hydrogen atom from

substrate during the catalytic mechanism.¹⁸On the basis of

convert, and that they both decompose to bis(µ-hydroxo)-

dicopper(II) species by pathways involving rate-determining

cleavage of ligand substituent C-H bonds.^27,28

Stimulated by its novelty and its relevance to biological

oxidations and catalysis, we have endeavored to fully character-

ize the novel [Cu₂(µ-O)₂]²⁺core by structural, spectroscopic,

and theoretical methods. Herein we provide a full description

of the results of these studies, which have utilized X-ray

crystallography, EXAFS, UV-vis absorption, NMR, and reso-

nance Raman spectroscopies, electrospray mass spectrometry,

bond valence sum analysis, and ab initio calculations to

understand the structure and bonding of the bis(µ-oxo)dicopper

complexes. For comparative purposes, we also describe the

nature of various bis(µ-hydroxo)dicopper(II) complexes having

N-dealkylated ligands that are produced upon decomposition

of the title compounds. Studies of the mechanism by which

the bis(µ-oxo)dicopper complexes undergo ligand N-dealkyla-

tion via C-H bond activation are presented separately.³⁵

kinetic investigations, analogous Tp^MeCuO fragments were

2

postulated to form upon cleavage of [(Tp^MeCu)₂(µ-η :η -O₂)]

in the rate-determining step of its decomposition,¹⁶but structural

or spectroscopic support for such a mono(oxocopper) species

in this or other synthetic or biological systems is notably absent.

The only oxocopper(III) species to be structurally characterized

appear in the solid state chemical literature;²⁰these include

MCuO₂(M ) group IA alkali metal),²¹SrLaCuO₄,²²and Na₃-

Cu₂O₃.^21,23Electrochemical oxidation of oxotricopper(II) spe-

cies has been reported to yield mixed-valent (µ-oxo)tricopper-

(II,II,III) complexes,²⁴but these have not been definitively

characterized by X-ray structural methods nor have they been

generated from O₂.

2

We recently discovered that solutions of [L^iPrCu(CH₃CN)]-

3

O₃SCF₃(L^iPr) 1,4,7-triisopropyl-1,4,7-triazacyclononane)²⁵in

3

CH₂Cl₂at -80 °C bind O₂reversibly to form a (µ-η²:η²-

peroxo)dicopper(II) species analogous to oxyHc.⁹By perform-

ing the oxygenation reaction in THF or by starting with a Cu(I)

3

Results and Discussion

Syntheses. The Cu(I) starting materials [LCu(CH₃CN)]X [L

) L^Bnor 1-benzyl-4,7-diisopropyl-1,4,7-triazacyclononane

3

complex of 1,4,7-tribenzyl-1,4,7-triazaclononane (L^Bn

)

²⁶instead

-

(L^{iPr Bn}), X ) ClO₄^-, O₃SCF₃^-, or SbF₆^-; L ) L^iPr, X ) ClO₄

,

2

3

of L^iPrin a variety of solvents, new species with spectroscopic

properties different from those associated with the [Cu₂(µ-η²:

η²-O₂)]²⁺unit were obtained.^27,28We have identified these new

species as having previously unobserved [Cu₂(µ-O)₂]²⁺cores

[formally bis(µ-oxo)dicopper(III)], akin to analogous rhombs

containing iron(III and/or IV)²⁹or manganese(III and/or IV)³⁰

ions. Such high oxidation state “diamond-like” cores are

becoming recognized as important entities in a number of

processes carried out by multinuclear metalloprotein active

sites,³¹including those that feature dioxygen activation (methane

monooxygenase,³²ribonucleotide reductase)³³and generation

(oxygen evolving complex in photosystem II).³⁴Indeed, the

potential involvement of such cores in both types of processes

has been placed on firmer ground by our discovery that the

[Cu₂(µ-O)₂]²⁺and the [Cu₂(µ-η²:η²-O₂)]²⁺species can inter-

3

PF₆^-, BPh₄^-, SbF₆^-, O₃SCF₃^-, or B(Ar_f)₄

]

were prepared

- 36

in straightforward fashion, either by mixing the macrocycle L

with [Cu(CH₃CN)₄]X according to a published procedure³⁷or

by a subsequent anion exchange to further vary X. The ligand

L^{iPr Bn}is new; it was prepared by benzylation of 1,4-diisopropyl-

2

1,4,7-triazacyclononane,³⁸which was constructed in turn from

1-tosyl-1,4,7-triazacyclononane³⁹through alkylation with iso-

propyl bromide followed by a final detosylation (see the

Experimental Section). All Cu(I) complexes were isolated as

air-sensitive colorless or pale yellow solids and were character-

ized by NMR and IR spectroscopy and elemental analysis. The

iPr₂Bn

anticipated trend in electron donating abilities L^iPr> L

>

3

L^Bnwas confirmed by comparison of the ν(CO) of derived

3

[LCuCO]⁺adducts prepared by treating representative com-

plexes with 1 atm CO; ν(CO) ) 2067, 2069, and 2084 cm^-1

,

(19) Mayer, J. M. Comments Inorg. Chem. 1988, 8, 125-135.

(20) Levason, W.; Spicer, M. D. Coord. Chem. ReV. 1987, 76, 45-120.

(21) Hestermann, K.; Hoppe, R. Z. Anorg. Allg. Chem. 1969, 367, 249-

280.

(22) Goodenough, J. B.; Demazeau, G.; Pouchard, M.; Hagenmuller, P.

J. Solid State Chem. 1973, 8, 325-330.

(23) A tricopper(II,II,III) cluster capped by O atoms derived from O₂

and bonded to the Cu(III) ion at short (∼1.83 Å) distances was reported

recently: (a) Cole, A. P.; Root, D. E.; Solomon, E. I.; Stack, T. D. P.

Abstracts of Papers, 211th National Meeting of the American Chemical

Society, New Orleans, LA; American Chemical Society: Washington, DC,

1996; INORG 060. (b) Cole, A. P.; Root, D. E.; Solomon, E. I.; Stack, T.

D. P. Science 1996, 273, 1848-1850.

(24) (a) Datta, D.; Mascharak, P. K.; Chakravorty, A. Inorg. Chem. 1981,

20, 1673-1679. (b) Datta, D.; Chakravorty, A. Inorg. Chem. 1982, 21,

363-368. (c) Datta, D.; Chakravorty, A. Inorg. Chem. 1983, 22, 1611-

1613.

(25) Haselhorst, G.; Stoetzel, S.; Strassburger, A.; Walz, W.; Wieghardt,

K.; Nuber, B. J. Chem. Soc., Dalton Trans. 1993, 83-90.

(26) Beissel, T.; Della Vedova, B. S. P. C.; Wieghardt, K.; Boese, R.

Inorg. Chem. 1990, 29, 1736-1741.

(27) Mahapatra, S.; Halfen, J. A.; Wilkinson, E. C.; Pan, G.; Cramer, C.

J.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1995, 117, 8865-8866.

(28) Halfen, J. A.; Mahapatra, S.; Wilkinson, E. C.; Kaderli, S.; Young,

V. G., Jr.; Que, L., Jr.; Zuberbu¨hler, A. D.; Tolman, W. B. Science 1996,

271, 1397-1400.

(29) (a) Zang, Y.; Dong, Y.; Que, L., Jr.; Kauffmann, K.; Mu¨nck, E. J.

Am. Chem. Soc. 1995, 117, 1169-1170. (b) Dong, Y.; Fujii, H.; Hendrich,

M. P.; Leising, R. A.; Pan, G.; Randall, C. R.; Wilkinson, E. C.; Zang, Y.;

Que, L., Jr.; Fox, B. G.; Kauffmann, K.; Mu¨nck, E. J. Am. Chem. Soc.

1995, 117, 2778-2792.

respectively. Although distinguishable by IR spectroscopy, the

differences in electron donating characteristics of the three

ligands are not large, as is shown by the only slightly differing

E_1/2values in their cyclic voltammograms [E_1/2(∆E_p) ) 360

(110), 415 (170), and 420 (310) mV vs SCE in CH₃CN with

0.2 M Bu₄NPF₆at a scan rate of 100 mV s^-1, respectively].

Bubbling O₂through colorless to pale yellow solutions of

-

[L^BnCu(CH₃CN)]X or [L^{iPr Bn}Cu(CH₃CN)]X (X ) ClO₄

,

3

2

O₃SCF₃^-, or SbF₆^-) in CH₂Cl₂, acetone, or THF at -80 °C

induced a color change to deep orange-brown that we attribute

to the formation of the respective bis(µ-oxo)dicopper complexes

[(LCu)₂(µ-O)₂](X)₂(L ) L^Bnor L^{iPr Bn}, Scheme 1). The optical

3

2

absorbance spectra of these deeply colored solutions contain

(33) (a) Stubbe, J. AdV. Enzymol. 1990, 63, 349-419. (b) Fontecave,

M.; Nordlund, P.; Eklund, H.; Reichard, P. AdV. Enzymol. 1992, 65, 147-

183. (c) Sturgeon, B. E.; Burdi, D.; Chen, S.; Huynh, B.-H.; Edmondson,

D. E.; Stubbe, J.; Hoffman, B. M. J. Am. Chem. Soc. 1996, 118, 7551-

7557.

(34) Sauer, K.; Yachandra, V. K.; Britt, R. D.; Klein, M. P. In Manganese

Redox Enzymes; Pecoraro, V. L., Ed.; VCH: New York, 1992; pp 141-

175.

(35) Mahapatra, S.; Halfen, J. A.; Tolman, W. B. J. Am. Chem. Soc.

1996, 118, 11575-11586.

(36) B(Ar_f)₄^-) B[3,5-(CF₃)₂C₆H₃]₄^-; Brookhart, M.; Grant, B.; Volpe,

A. F., Jr. Organometallics 1992, 11, 3920-3922.

(37) 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.

(30) (a) Wieghardt, K. Angew. Chem., Int. Ed. Engl. 1989, 28, 1153-

1172. (b) Manchanda, R.; Brudvig, G. W.; Crabtree, R. H. Coord. Chem.

ReV. 1995, 144, 1-38.

(31) Que, L., Jr.; Dong, Y. Acc. Chem. Res. 1996, 29, 190-196.

(32) (a) Lipscomb, J. D. Annu. ReV. Microbiol. 1994, 48, 371-399. (b)

Wallar, B. J.; Lipscomb, J. D. Chem. ReV., in press.

(38) Houser, R. P.; Halfen, J. A.; Young, V. G., Jr.; Blackburn, N. J.;

Tolman, W. B. J. Am. Chem. Soc. 1995, 117, 10745-10746.

(39) Sessler, J. L.; Sibert, J. W.; Lynch, V. Inorg. Chem. 1990, 29, 4143-

4146.

11558 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

Scheme 1

Table 1. Spectroscopic Properties of Bis(µ-oxo)dicopper Complexes

manometry

Cu:O₂

UV-vis^aλ_max

,

resonance Raman^b

_{Cu O}, cm^-1

complex

nm (ꢀ, M^-1cm^-1

)

ν

NMR^cδ, ppm

2

[(L^BnCu)₂(µ-O)₂](ClO₄)₂

2.2(2):1

430 (14 000), 318 (12 000)

602, 612 (583)

¹H (500 MHz): 7.37 (s, 10H), 7.29 (s, 20H),

4.59 (s, 12H), 3.48 (s, 12H), 2.55 (s, 12H).

¹³C{¹H} (125 MHz): 133.0, 132.3, 128.9,

59.5, 51.6 (br)

3

[(L^iPrCu)₂(µ-O)₂](ClO₄)₂

2.0(2):1

448 (13 000), 324 (11 000)

436 (16 000), 322 (12 000)

600 (580)

594 (572)

3

[(L^{iPr Bn}Cu)₂(µ-O)₂](ClO₄)₂

2

^aMeasured in either CH₂Cl₂(L^Bnand L^{iPr Bn}) or THF (L^iPr) at -80 °C. ^bThe resonance Raman spectra were obtained at 77 K as frozen CH₂Cl₂

or THF solutions using 457 nm laser excitation; only ¹⁸O sensitive vibrations are quoted with specific ¹⁸O shifts in parentheses. ^cRecorded in 1:2

CD₂Cl₂/(CD₃)₂CO solvent mixture at -75 °C.

3

2

3

shown by UV-vis and/or resonance Raman (vide infra)

spectroscopic monitoring. For example, in CH₂Cl₂for all

counterions (X) the predominant (>90%) product is [(L^iPrCu)₂-

3

-

(µ-η²:η²-O₂)](X)₂, in THF or acetone with X ) BPh₄^-, SbF₆

,

O₃SCF₃^-, and/or B(Ar_f)₄both species are present [∼4:1

-

peroxo:bis(µ-oxo) ratio], and in THF with X ) ClO₄^-or PF₆

only [(L^iPrCu)₂(µ-O)₂](X)₂forms. A detailed analysis of this

3

evidently complex influence of solvent and counterion on the

nature and interconversion of these oxygenation products will

be the subject of a future report. Importantly in the current

context, solutions resulting from oxygenation of [L^iPrCu(CH₃-

3

CN)]X (X ) ClO₄^-or PF₆^-) in THF exhibit UV-vis absorption

features nearly identical to those observed when L ) L^Bnor

3

L^{iPr Bn}(Table 1), demonstrating the clean generation of analo-

2

gous bis(µ-oxo)dicopper cores and no µ-η²:η²-peroxo isomers

in each case.

Dinuclear formulations for the complexes in solution were

supported by manometric measurements of O₂uptake and, in

one instance, electrospray mass spectrometry.⁴⁰Manometry

experiments (3-5 replicate runs) for the reaction of O₂with

Figure 2. UV-vis spectrum of [(L^BnCu)₂(µ-O)₂](SbF₆)₂in CH₂Cl₂

3

at -80 °C (-) and of the solution resulting upon its decomposition

(- - -).

two intense, presumably charge transfer bands with λ_max= 320

nm (ꢀ 12 000 M^-1cm^-1per dicopper complex) and 430 nm (ꢀ

14 000 M^-1cm^-1) (Table 1 and Figure 2). These features vary

little among the species generated with the two macrocyclic

[LCu(CH₃CN)]ClO₄(L ) L^Bnor L^{iPr Bn}) in CH₂Cl₂and with

3

2

[L^iPrCu(CH₃CN)]ClO₄in THF at -80 °C indicated Cu:O₂

3

stoichiometries of 2:1 (Table 1). Electrospray mass spectra were

coligands L^Bnor L^{iPr Bn}or the various counterions or solvents.

They are distinct from the absorption bands characteristic for

the planar (µ-η²:η²-peroxo)dicopper(II) core.^3b,8,9The oxygen-

3

2

-

obtained for solutions of [(d₂₁-L^BnCu)₂(µ-O)₂](X)₂(X ) SbF₆

3

or ClO₄^-), perdeuterated at the ligand alkyl substituents in order

to enhance stability (vide infra), in CH₂Cl₂at ca. -40 °C. We

observed parent ion envelopes with isotope patterns consistent

-

ation reaction of [L^iPrCu(CH₃CN)]X [X ) ClO₄^-, PF₆^-, BPh₄

,

3

SbF₆^-, O₃SCF₃^-, or B(Ar_f)₄^-] is more complicated, since either

bis(µ-oxo)- or (µ-η²:η²-peroxo)dicopper species or both are

generated depending on the solvent and/or counterion used as

(40) Kim, J.; Dong, Y.; Larka, E.; Que, L., Jr. Inorg. Chem. 1996, 35,

2369-2372.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11559

shown in Scheme 2, either the symmetric species [(L^BnCu)₂-

3

(µ-OH)₂]²⁺or the unsymmetric compound [(L^BnCu)(L^{Bn H}Cu)-

(µ-OH)₂]²⁺having a singly N-dealkylated capping ligand was

crystallized from the solution resulting from decomposition of

3

2

[(L^BnCu)₂(µ-O)₂]²⁺, the identity of the isolated product being

3

dependent on the counterion used (ClO₄^-or CF₃SO₃^-, respec-

tively) and the crystallization conditions. The products isolated

from the decompositions of [(LCu)₂(µ-O)₂](ClO₄)₂(L ) L^iPr

3

or L^{iPr Bn}) were [(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂, in which only

2

N-dealkylated caps are present, or [(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃-

2

SCF₃)₂, with unperturbed substituents, respectively (the different

counterions were introduced via metathesis after completion of

the decomposition). The isolated yields of the various products

were variable and in no instance were we able to conclusively

identify all copper-containing species after the decomposition

reactions. We suspect that, for each case, the green decomposed

solutions contain rapidly interconverting statistical mixtures of

bis(hydroxo)dicopper(II) complexes with two intact, one intact

and one dealkylated, or two dealkylated ligands and that the

specific counterion and crystallization conditions dictate which

dinuclear species crystallizes.

Figure 3. Experimental (top) and calculated (bottom) peak envelopes

in the positive ion (A) and negative ion (B) electrospray mass spectra

of [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂in CH₂Cl₂.

3

with the indicated formulations; exemplary positive and negative

ion mass spectra and simulations for the respective parent ions

In order to substantiate the hypothesis that the bis(hydroxo)

compounds equilibrate in solution, we performed a simple cross-

over experiment. Electrospray mass spectrometric analysis of

3

+

Bn₃

-

{[(d₂₁-L^BnCu)₂(µ-O)₂]SbF₆} and {[(d₂₁-L Cu)₂(µ-O)₂](SbF₆)₃}

are shown in Figure 3. Electrospray data for hydroxo-bridged

decomposition products contain parent ion envelopes of similar

shape, but, as expected, are shifted 2 m/z units higher.

3

the solution obtained by mixing equimolar solutions of [(L^Bn

-

Cu)₂(µ-OH)₂](SbF₆)₂and [(d₂₁-L^BnCu)₂(µ-OH)₂](SbF₆)₂in CH₂-

3

Cl₂at room temperature for 30 min indicated the presence of a

The bis(µ-oxo)dicopper complexes are unstable in solution

and in the solid state, as evidenced by a change in color from

orange-brown to green that is accompanied by a loss of the

intense CT absorption bands associated with the [Cu₂(µ-O)₂]²⁺

core (Figure 1). The decomposition rates were significantly

slowed by using ligands with deuterated substituents on the

triazacyclononane ligands; a detailed discussion of these kinetic

isotope effects and the mechanism of decomposition is presented

elsewhere.³⁵Various bis(µ-hydroxo)dicopper(II) complexes

with differing capping ligands were isolated from the green

decomposed solutions (Scheme 2), three of which were structur-

ally characterized by X-ray crystallography (vide infra). As

+

1:2:1 ratio of {[(L^BnCu)₂(µ-OH)₂](SbF₆)} (M⁺envelope at

3

m/z ∼1197), {[(L^BnCu)(d₂₁-L Cu)(µ-OH)₂](SbF₆)} (M⁺at

Bn₃

+

3

m/z ∼1218), and {[(d₂₁-L^BnCu)₂(µ-OH)₂](SbF₆)} (M at m/z

∼1239). The observation of this statistical mixture supports

equilibration among the various hydroxo-bridged complexes in

solution, presumably via a bridge cleavage/recombination

process(es) that would be expected to be more facile than

alternatives involving dissociation/reassociation of the capping

macrocyclic ligands.

+

3

X-ray Crystallography. Crystallographic data and selected

bond distances and angles for the X-ray structures of [(d₂₁-

Scheme 2

11560 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Table 2. Summary of X-ray Crystallographic Data

Mahapatra et al.

[(d₂₁-L^BnCu)₂-

[(L^BnCu)(L^{Bn H}Cu)-

(µ-OH)₂](CF₃SO₃)₂‚

2(CH₃)₂CO

[(L^{iPr H}Cu)₂-

3

2

(µ-O)₂](SbF₆)₂‚

7(CH₃)₂CO‚2CH₃CN

(µ-OH)₂](BPh₄)₂‚

[(L^{iPr Bn}Cu)₂-

2THF

(µ-OH)₂](CF₃SO₃)₂

emp form

fw

C₇₉H₇₂D₄₂F₁₂Cu₂N₈O₉Sb₂

1960.70

C₅₅H₇₄F₆Cu₂N₆O₁₀S₂

1284.4

C₈₀H₁₁₂B₂Cu₂N₆O₄

1370.46

C₄₀H₆₈F₆Cu₂N₆O₈S₂

1066.21

cryst syst

space group

a (Å)

b (Å)

c (Å)

monoclinic

C2/c

monoclinic

P2₁/n

monoclinic

Cc

monoclinic

P2₁/a (no. 14)

21.326(7)

10.887(2)

23.316(8)

113.36(1)

4970(5)

26.1916(6)

15.3560(4)

22.2705(6)

96.493(1)

8899.7(4)

8

15.1315(2)

26.0344(4)

15.3733(3)

92.788(1)

6049.0(2)

4

23.7906(4)

16.7455(4)

21.7924(2)

120.462(1)

7483.4(2)

4

â (deg)

V (Å³)

Z

4

density(calcd) (mg/cm³)

temp (K)

1.462

123(2)

1.410

173(2)

1.216

173(2)

1.425

297(2)

crystal size (mm)

diffractometer

radiation

0.50 × 0.38 × 0.30

Siemens SMART

Mo KR (0.710 73 Å)

1.153 mm^-1

50.32

22487

7865

6599

523

0.1012

0.35 × 0.30 × 0.15

Siemens SMART

Mo KR (0.710 73 Å)

0.850 mm^-1

48.22

24233

9409

6822

801

0.077

0.40 × 0.24 × 0.04

Siemens SMART

Mo KR (0.710 73 Å)

0.621 mm^-1

48.14

15400

10282

8393

856

0.0824

0.36 × 0.24 × 0.15

Enraf-Nonius CAD-4

Mo KR (0.710 73 Å)

10.11 cm^-1

50.1

9499

9237

3528

577

0.069

0.058

absn coeff

2θ max (deg)

no. of reflns colld

no. of ind reflns

no. of obsd reflns [I > 2σ(I)]

params

R1^a[I > 2σ(I)]

Rw^a

wR2^b

0.2532

0.1585

0.1797

goodness-of-fit

1.051

1.091

0.947, -0.661

1.142

0.568, -0.557

1.41

0.56, -0.66

largest diff peak and hole (e^-Å^-3

)

2.941, -2.38^c

2

^aR1 ) ∑||F_o| - |F_c||/∑|F_o|; Rw ) [(∑w(|F_o| - |F_c|)²/∑wF_o)]^1/2, where w ) 4F_o/σ²(F_o), σ²(F_o) ) [S²(C + R²B) + (pF_o)²]/Lp², S ) scan rate,

C ) total integrated peak count, R ) ratio of scan time to background counting time, B ) total background count, Lp ) Lorentz-polarization factor,

and p ) p-factor. ^bwR2 ) [∑[w(F_o- F_c²)²]/∑[w(F_o)²]]^1/2, where w ) 1/σ²(F_o) + (aP)²+ bP. ^cLocated near disordered SbF₆anion.

2

-

Figure 5. Representation of the X-ray crystal structure of [(L^BnCu)-

3

Figure 4. Representation of the dication in the X-ray crystal structure

(L^{Bn H}Cu)(µ-OH)₂](O₃SCF₃)₂‚2(CH₃)₂CO (35% ellipsoids, acetone sol-

vate and all hydrogen atoms except those bonded to N or O omitted

for clarity, dashed lines indicate postulated hydrogen bonds).

2

of [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂‚7(CH₃)₂CO‚2CH₃CN (35% ellipsoids,

3

hydrogen and deuterium atoms omitted for clarity).

L^BnCu)₂(µ-O)₂](SbF₆)₂‚7(CH₃)₂CO‚2CH₃CN and the bis(µ-

3

hydroxo)dicopper(II) complexes [(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂]-

3

2

core geometries typical for this molecular motif,¹³but with

different capping ligands and hydrogen-bonding patterns. The

hydroxo bridges occupy equatorial coordination sites in all three

structures, with average parameters Cu-O ) 1.94 Å [range

1.921(5)-1.975(6) Å], Cu-O-Cu ) 101.5° [range 100.5(2)-

102.9(3)°], and Cu‚‚‚Cu ) 3.01 Å [range 2.9769(9)-3.037(3)

Å]. The copper ions in each complex have slightly distorted

square pyramidal coordination geometries; the axial N-donors

are disposed anti with respect to the [Cu₂(µ-OH)₂]²⁺core.

While the original macrocyclic ligand is retained intact in

(O₃SCF₃)₂‚2CH₃CO, [(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚2THF, and

2

[(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂are summarized in Tables 2 and

2

3, with drawings shown in Figures 4-8 {only one of the two

independent but chemically similar molecules in the asymmetric

unit of [(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂is shown}. Complete

2

listings of positional parameters, temperature factors, and bond

distances and angles for all of the compounds are provided as

supporting information. We discuss the structures of the bis-

(µ-hydroxo)dicopper(II) complexes first in order to more readily

draw comparisons of their cores with that of the unique bis(µ-

oxo) compound.

Bis(µ-hydroxo)dicopper(II) Complexes. Examination of

the molecular structures of the three bis(µ-hydroxo)di-

coppper(II) complexes shown in Figures 5-7 reveals similar

[(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂(Figure 7), the occurrence of

2

N-dealkylation chemistry upon decomposition of the respective

bis(µ-oxo)dicopper precursors supported by L^Bnor L^iPris

evident from the structures shown in Figures 5 and 6, which

contain either one or two N-H instead of N-R groups (R )

3

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11561

Table 3. Selected Bond Lengths (Å) and Angles (deg) for

Complexes Characterized by X-ray Crystallography^a

[(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂‚7(CH₃)₂CO‚2CH₃CN

3

Cu(1)-O(1)

1.803(5)

1.987(6)

1.986(6)

2.794(2)

102.3(2)

Cu(1)-O(1A)

Cu(1)-N(4)

1.808(5)

2.298(6)

Cu(1)-N(1)

Cu(1)-N(7)

O(1)-Cu(1)-O(1A) 78.6(2)

Cu(1)‚‚‚Cu(1A)

O(1)-Cu(1)-N(4)

O(1)-Cu(1)-N(1)

O(1)-Cu(1)-N(7)

172.3(2)

96.2(2)

O(1A)-Cu(1)-N(1) 95.6(2)

O(1A)-Cu(1)-N(7) 171.1(2)

O(1A)-Cu(1)-N(4) 103.8(2)

Cu(1)-O(1)-Cu(1A) 101.4(2)

[(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](CF₃SO₃)₂‚2(CH₃)₂CO

3

2

Cu(1)-O(1)

1.949(4)

2.087(5)

2.289(5)

1.940(4)

2.085(5)

2.9769(9) O(1)-Cu(1)-O(1′)

176.2(2)

99.1(2)

Cu(1)-O(1′)

Cu(1)-N(4)

Cu(1′)-O(1)

Cu(1′)-N(1′)

Cu(1′)-N(7′)

1.924(5)

2.054(4)

1.921(5)

2.266(5)

2.029(5)

78.6(2)

97.7(2)

98.3(2)

106.1(2)

Cu(1)-N(1)

Cu(1)-N(7)

Cu(1′)-O(1′)

Cu(1′)-N(4′)

Cu(1)‚‚‚Cu(1′)

O(1)-Cu(1)-N(1)

O(1)-Cu(1)-N(7)

O(1)-Cu(1)-N(4)

O(1′)-Cu(1)-N(1)

O(1′)-Cu(1)-N(7)

O(1′)-Cu(1′)-N(1′) 112.9(2)

O(1′)-Cu(1′)-N(7′) 94.9(2)

Cu(1)-O(1′)-Cu(1′) 100.8(2)

Figure 6. Representation of the dication and the THF solvate molecules

O(1′)-Cu(1)-N(4) 170.4(2)

O(1′)-Cu(1′)-O(1) 78.9(2)

O(1′)-Cu(1′)-N(4′) 175.1(2)

Cu(1)-O(1)-Cu(1′) 100.5(2)

in the X-ray crystal structure of [(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚2THF (35%

2

ellipsoids, all hydrogen atoms except those bonded to N or O omitted

for clarity, dashed lines indicate postulated hydrogen bonds).

[(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚2THF

2

Cu(1)-O(1)

Cu(1)-N(11)

Cu(1)-N(17)

Cu(2)-O(2)

Cu(2)-N(24)

Cu(1)‚‚‚Cu(2)

1.922(6)

2.017(8)

Cu(1)-O(2)

Cu(1)-N(14)

1.951(6)

2.303(7)

1.930(6)

2.008(9)

2.048(8)

78.3(3)

2.086(8)

1.951(6)

2.335(9)

Cu(2)-O(1)

Cu(2)-N(21)

Cu(2)-N(27)

3.0088(13) O(1)-Cu(1)-O(2)

O(1)-Cu(1)-N(14) 115.8(3)

O(1)-Cu(2)-N(21) 93.7(3)

O(2)-Cu(1)-N(11) 96.5(3)

O(2)-Cu(2)-N(17) 175.8(3)

O(2)-Cu(2)-N(24) 107.6(3)

O(1)-Cu(1)-Cu(2) 38.7(2)

O(1)-Cu(2)-Cu(1) 38.6(2)

O(2)-Cu(2)-Cu(1) 39.6(2)

Cu(2)-O(2)-Cu(1) 100.9(3)

O(1)-Cu(1)-N(11)

O(1)-Cu(1)-N(17)

O(1)-Cu(2)-N(27)

O(2)-Cu(1)-N(14)

O(2)-Cu(2)-N(21)

O(2)-Cu(2)-N(27)

O(1)-Cu(1)-O(2)

O(2)-Cu(1)-Cu(2)

Cu(1)-O(1)-Cu(2)

161.2(3)

100.1(3)

168.2(3)

99.3(3)

169.1(3)

101.7(3)

78.1(2)

39.5(2)

102.7(3)

[(L^{iPr Bn}Cu)₂(µ-OH)₂](CF₃SO₃)₂

2

Molecule 1

Cu(1)-O(1)

Cu(1)-N(1)

Cu(1)-N(7)

1.950(6)

2.113(8)

2.260(8)

Cu(1)-O(1′)

1.933(6)

2.065(8)

3.037(3)

159.4(3)

117.1(3)

173.6(3)

Cu(1)-N(4)

Cu(1)‚‚‚Cu(1′)

Figure 7. Representation of the dication in the X-ray crystal structure

O(1)-Cu(1)-O(1′) 77.1(3)

O(1)-Cu(1)-N(1)

O(1)-Cu(1)-N(7)

O(1′)-Cu(1)-N(4)

of [(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂(35% ellipsoids, all hydrogen atoms

2

O(1)-Cu(1)-N(4)

98.6(3)

except those bonded to O omitted for clarity).

O(1′)-Cu(1)-N(1) 98.0(3)

O(1′)-Cu(1)-N(7) 101.1(3)

Cu(1)-O(1)-Cu(1′) 102.9(3)

macrocyclic ligands during decomposition of the bis(µ-oxo)-

dicopper precursors.

Cu(2)-O(2)

Cu(2)-N(11)

Cu(2)-N(17)

1.947(6)

2.053(8)

3.033(3)

162.7(3)

113.2(3)

Bis(µ-oxo)dicopper Complex [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂‚

3

7(CH₃)₂CO‚2CH₃CN. Crystals of this compound suitable for

X-ray crystallographic characterization were grown by allowing

an oxygenated solution of [d₂₁-L^BnCu(CH₃CN)]SbF₆in 1:1 CH₂-

3

Cl₂/(CH₃)₂CO to stand at -80 °C for 2-3 days, the use of the

deuterated ligand d₂₁-L^Bnbeing critical for preventing decom-

3

position of the complex over the time period required for

crystallization. As described in detail in the Experimental

Section, significant disorder of the SbF₆^-counterion, the solvent

molecules, and phenyl groups necessitated imposition of bond

distance and anisotropic displacement parameter constraints to

force these units to conform to their average respective

substructures and to better fit rigid body motion. These aspects

of the structure undoubtedly contribute to the relatively high R

value (10.1%), which nonetheless compares favorably to those

reported for other copper-dioxygen complex structures.³

A crystallographic inversion center relates the two N₃CuO

halves of the novel bis(µ-oxo)dicopper complex (Figure 4). Like

the aforementioned bis(µ-hydroxo)dicopper compounds, each

copper ion adopts a slightly distorted square pyramidal geometry

with the long axial bonds in an anti relationship, skewed in

this instance by 17° from the axis normal to the equatorial plane.

However, the geometrical parameters of the planar [Cu₂(µ-

isopropyl or benzyl). In [(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](O₃-

SCF₃)₂‚2(CH₃)₂CO (Figure 5), the single N-H group and the

hydroxide bridges are hydrogen bonded to triflate counterions

in the crystal as shown by the dashed lines; relevant distances

are O1‚‚‚O102 ) 2.839(5) Å, O1′‚‚‚O203 ) 3.067(5) Å, N7′‚‚‚

O203 ) 3.098(5) Å, and N7′‚‚‚O201 ) 3.202(5) Å. The THF

3

2

solvate molecules in the crystal of [(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚

2

2THF (Figure 6) are hydrogen bonded to the µ-hydroxo groups

[O1‚‚‚O1S ) 3.016(8) Å and O2‚‚‚O2S ) 2.930(8) Å], but

N-H‚‚‚X interactions are not evident in this instance. Finally,

there are no close contacts between the µ-hydroxo groups and

other potential hydrogen bond acceptors in the crystal structure

of [(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂(Figure 7). Overall, the three

2

structures exemplify typical [Cu₂(µ-OH)₂]²⁺core geometries¹³

and provide unequivocal evidence for N-dealkylation of the

11562 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

oxidation state greater than +2 to the copper ions.²⁰In support

of this notion, metal-ligand distances in the few structurally

characterized examples of discrete Cu(III) compounds (Table

4, entries 8-13) include a short Cu-O separation of 1.826(3)

Å^23,45and Cu-N distances of 1.80-1.89 Å,^45,46the latter

depending on the nature of the N-donor (amide vs amine). More

relevant, perhaps, are the similarities between the core of the

bis(oxo) complex and the analogous Cu₂(µ-O)₂rhombs present

in the oxocopper(III) extended solid MCuO₂(M ) group IA

alkali metal ion);²¹the Cu-O and Cu‚‚‚Cu distance and Cu-

O-Cu angle parameters of 1.84 Å, 2.71 Å, and 95°, respec-

tively, for KCuO₂compare favorably to the like parameters 1.81

Å, 2.79 Å, and 101° for the bis(oxo) complex.⁴⁷Indeed, the

[Cu₂(µ-O)₂]²⁺core geometries in the bis(oxo) complex and

MCuO₂closely resemble those of other planar bis(µ-oxo)dimetal

units containing iron and manganese in their +3 and/or +4

oxidation states (Table 5),^29,30thus indicating a general structural

uniformity among the different metal-containing cores.

The above qualitative relationship between structural param-

eters and metal oxidation state can be more quantitatively

described through a bond valence sum (BVS) analysis.⁴⁸The

BVS for a metal ion is the sum of bond valences s_ijfor each

metal-ligand bond calculated according to eq 2, where r₀is a

Figure 8. Representation of the copper coordination spheres (50%

thermal ellipsoids) in the X-ray crystal structures of (A) [(d₂₁-L^BnCu)₂-

3

(µ-O)₂](SbF₆)₂‚7(CH₃)₂CO‚2CH₃CN and (B) [(L^BnCu)(L^{Bn H}Cu)(µ-

OH)₂](O₃SCF₃)₂‚2(CH₃)₂CO.

3

2

BVS ) s ) exp[(r - r )/0.37]

(2)

∑ _ij∑

0

ij

O)₂]²⁺core differ significantly from those of the isomeric [Cu₂-

parameter characteristic of that bond and r_ijis the experimentally

determined bond distance. The empirical r₀values for many

metal-ligand bonds for various metal oxidation states have been

determined by averaging r₀values for sets of like complexes,

where each r₀is calculated by solving eq 2 for r₀using

experimentally determined r_ijnumbers in complexes of known

oxidation state.⁴⁸Using r₀) 1.679 or 1.719 for Cu(II)-O or

Cu(II)-N bonds, respectively,^48cwe calculated BVS values for

the series of complexes listed in Table 4. Good agreement

((0.25) between the BVS and the copper oxidation state is

evident for a range of bona fide Cu(II) compounds, including

those ligated by 1,4,7-triazacyclononanes (entries 1-4) and

those containing relatively short copper-oxo or -hydroxo bonds

(entries 5-8). In contrast, BVS values significantly greater than

2.0 are obtained when the Cu(II) r₀values are used for the few

crystallographically characterized Cu(III) complexes (entries

9-13) and the bis(µ-oxo)dicopper compounds structurally

3b

(µ-η²:η²-O₂)]²⁺unit in [(Tp^iPrCu)₂(O₂)] and those typically

2

observed in [Cu₂(OH)₂]²⁺cores, as illustrated by the compa-

rision with that of [(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](O₃SCF₃)₂·2CH₃-

CO in Figure 8. Note that the 2.794(2) Å Cu‚‚‚Cu and 1.81 Å

(av) Cu-O distances in the bis(µ-oxo) compound are signifi-

cantly shorter than those in the bis(µ-hydroxo) complex [2.9769-

(9) Å and 1.93 Å (av), respectively]. Interestingly, the Cu-

O(H)-Cu angles are essentially identical in the two cores,

suggesting that the contracted metal-metal separation in the

bis(µ-oxo) complex arises entirely from the shortened Cu-O

distances. The Cu‚‚‚Cu and O‚‚‚O distances of 2.794(2) and

3

2

2.287(5) Å, respectively, for [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂rule

3

out a possible formulation of the compound as a µ-η²:η²-peroxo

complex, since the corresponding distances in this latter core

are significantly longer and shorter, respectively (3.56 and 1.41

Å).^3b

defined by X-ray crystallography (L ) L^Bn; entry 14) or EXAFS

Examples of short Cu-O distances in copper(II) coordination

chemistry occur in some mononuclear Cu(II)-OH species

{Cu-O ) 1.875(2) Å in [Cu(Me₆tren)(OH)]ClO₄,⁴¹1.878(2)

Å in [(tppa)Cu(OH)]ClO₄},⁴²a trans-1,2-peroxodicopper(II)

complex {Cu-O ) 1.852(5) Å in {[(TMPA)Cu]₂(O₂)}(PF₆)₂},^3a

an oxocopper(II) cluster {Cu-O ) 1.85 Å in [L′₃Cu(O)-

(ClO₄)]₂},⁴³and in Cu(II)-O-Fe(III) assemblies [1.829(4) Å⁴¹

and 1.856(5) Å⁴⁴]. In these complexes, most of the other Cu-L

(L ) N or O donor ligand) bond lengths are typical for Cu(II)

compounds [∼2.0 Å], thus corroborating the +2 formal oxida-

tion levels assigned to the copper ions in each case. In [(d₂₁-

3

(L ) L^iPr, vide infra; entry 15). Moreover, by using new r₀

3

values for Cu(III)-N and -O bonds of 1.738 and 1.753,

respectively, that were calculated from the data available for

the Cu(III) compounds in entries 9, 10, and 12, we obtained

BVS ) 3.0 ( 0.25 for all the Cu(III) species (entries 9-13)

and the bis(µ-oxo)dicopper complexes (entries 14 and 15). Thus,

BVS analysis supports assignment of a +3 oxidation state to

the copper ions in the bis(µ-oxo) compounds, notwithstanding

the approximations inherent to using a small reference database

for the derivation of the Cu(III) parameters.

L^BnCu)₂(µ-O)₂](SbF₆)₂, however, the Cu-O distances are

3

An additional noteworthy feature of the X-ray crystal structure

of [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂is the presence of an array of

shorter than those noted above and the equatorial Cu-N

distances [1.99 Å] also are slightly shorter than those typically

found in bona fide Cu(II) complexes of triazacyclononane

ligands [2.0-2.1 Å]. The observation of this contracted

coordination sphere is consistent with assignment of a formal

3

(45) Diaddario, L. L.; Robinson, W. R.; Margerum, D. W. Inorg. Chem.

1983, 22, 1021-1025.

(46) (a) Clark, G. R.; Skelton, B. W.; Waters, T. N. J. Chem. Soc., Dalton

Trans. 1976, 1528-1536. (b) Birker, P. J. M. W. L. Inorg. Chem. 1977,

16, 2478-2482.

(47) A Cu₂O cluster examined in the gas phase using photoelectron

spectroscopy and density functional calculations has been assigned geometric

(Cu-O ) 1.78 Å) and vibrational (ν_rhomb) 630 cm^-1) properties similar to

those described here: Wang, L.-S.; Wu, H.; Desai, S. R.; Lou, L. Phys.

ReV. B 1996, 53, 8028-8031.

(48) (a) Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 240-

244, 244-247. (b) Thorp, H. H. Inorg. Chem. 1992, 31, 1585-1588. (c)

Hati, S.; Datta, D. J. Chem. Soc., Dalton Trans. 1995, 1177-1182.

(41) Lee, S. C.; Holm, R. H. J. Am. Chem. Soc. 1993, 115, 11789-

11798.

(42) tppa ) tris[(6-pivaloylamino)-2-pyridyl)methyl]amine; see ref 4.

(43) L′ ) 3-(phenylimimo)butanone 2-oxime; Butcher, R. J.; O’Connor,

C. J.; Sinn, E. Inorg. Chem. 1981, 20, 537-545.

(44) Karlin, K. D.; Nanthakumar, A.; Fox, S.; Murthy, N. N.; Ravi, N.;

Huynh, B. H.; Orosz, R. D.; Day, E. P. J. Am. Chem. Soc. 1994, 116, 4753-

4763.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11563

Table 4. Bond Valence Sums for Copper Ions in Selected Complexes

coord

sphere

BVS^a

BVS^b

entry

1

compound

Cu-N (Å)

Cu-O (Å)

[Cu(II) r₀] [Cu(III) r₀]

ref

[(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](O₃SCF₃)₂3N, 2O 2.054, 2.087, 2.289 1.924, 1.949

1.99

2.05

2.02

2.05

1.95

1.91

1.88

this work (X-ray)

3

2

2.029, 2.085, 2.266 1.921, 1.940

[(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂

3N, 2O 2.017, 2.086, 2.303 1.922, 1.951

2.008, 2.048, 2.335 1.930, 1.951

3N, 2O 2.113, 2.065, 2.260 1.950, 1.933

2.090, 2.053, 2.266 1.975, 1.967

3N, 2O 2.132, 2.088, 2.288 1.936, 1.957

4N, O 2.049, 2.189, 2.161, 1.875

2.163

this work (X-ray)

2

3

[(L^{iPr Bn}Cu)₂(µ-OH)₂](O₃SCF₃)₂

2

L^iPrCu(O₂CCH₃)₂

[Cu(Me₆tren)(OH)]ClO₄

h

41

3

4

5

c

6

7

8

[(OEP)Fe-O-Cu(Me₆tren)]ClO₄

4N, O 2.045, 2.120, 2.179, 1.830

2.01

2.22

41

44

43

2.157

d

[(F₈-TPP)Fe-O-Cu(TMPA)]ClO₄

4N, O 2.100, 2.009, 2.172, 1.856

1.984

2N, 2O 1.991, 1.98

1.988, 1.971

1.999, 1.980

4O

e

[L′₃Cu(O)(ClO₄)]₂

1.854, 1.956

1.860, 1.958

1.890, 1.999

1.85, 1.85, 1.85, 1.85

2.07

1.95

2.59

2.78

2.88

2.63

2.71

9

KCuO₂

3.04

3.21

2.98

2.99

21

46b

45

46a

22

10 (Bu₄N){Cu[o-C₆H₄(NCONHCONH)₂]} 4N

1.82, 1.85, 1.86, 1.89

11 [Cu(H_-2Aib)₃]^f

12 [L Cu(OH)]ClO₄

13 SrLaCuO₄

3N, O 1.804, 1.801, 1.898 1.826

4N, O 1.86, 1.87, 1.85, 1.88 2.74

5O

N

g

4

1.88 (eq), 1.88 (eq),

2.23 (ax)

14 [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂

3N, 2O 1.986, 1.987, 2.298 1.808, 1.803

2.60

2.96

2.94

this work (X-ray)

this work (EXAFS)

3

15 [(L^iPrCu)₂(µ-O)₂](ClO₄)₂

3N, 2O 1.89, 1.89, 2.3

1.89, 1.89

3

(assumed)

^aCalculated using r₀values Cu(II)-O ) 1.679 Å and Cu(II)-N ) 1.719 Å. Separate BVS values are reported for polynuclear complexes in

which the metal ions are inequivalent. ^bCalculated using r₀values Cu(III)-O ) 1.738 Å and Cu(III)-N ) 1.753 Å. These parameters, which are

necessarily approximate due to the limited available database, were obtained from r_ijdata for entries 9, 10, and 12. ^cCalculated for one of the

independent molecules (no. 2) in the unit cell of the CH₃CN solvate; OEP ) octaethylporphyrinate. ^dF₈-TPP ) tetrakis(2,6-difluorophenyl)porphyrinate.

^eL′H ) 3-(phenylimino)butanone 2-oxime. ^fAib ) tri-R-aminoisobutyric acid. ^gL^N) planar, tetradentate ligand derived from condensation of

4

oxalyldihydrazide, ammonia, and acetaldehyde. ^hHalfen, J. A.; Tolman, W. B. J. Am. Chem. Soc. 1994, 116, 5475-5476.

Table 5. Structural Parameters for Selected Complexes with Bis(µ-oxo/hydroxo)dimetal Cores

core structure

supporting ligand^a

M‚‚‚M (Å)

O‚‚‚O (Å)

M-O (Å)

M-O-M (deg)

ref

Mn^II,II₂(µ-OH)₂

Tp^iPr

3.314(1)

2.553(7)

2.094(4)

2.089(5)

1.806(5)

1.813(6)

1.808(6)

1.787(6)

1.805(5)

1.820(5)

1.805(3)

1.798(3)

1.841(4)

1.917(4)

1.906(8)

1.981(8)

1.77^e

104.8(2)

b

2

Mn^III,III₂(µ-O)₂

Tp^iPr

2.696(2)

2.369(8)

96.5(3)

97.0(3)

2

Mn^III,IV₂(µ-O)₂

Mn^IV,IV₂(µ-O)₂

Fe^III,III₂(µ-O)₂

phen

2.695(9)

2.748(2)

2.714(2)

2.95(1)

2.89^e

2.425(8)

2.326(4)

96.0

c

phen

99.5(2)

92.5(2)

98.7(4)

c

6-Me₃-TPA

5-Me₃-TPA

29a

d

Fe^III,III₂(µ-O)(µ-OH)

Fe^III,IV₂(µ-O)₂

2.53(1)

29b

1.94

Cu^III,III₂(µ-O)₂

L^Bn

2.794(2)

2.287(2)

2.454(2)

1.808(5)

1.803(5)

1.89^e

101.4(2)

this work (X-ray)

3

L^iPr

2.88^e

2.9769(9)

this work (EXAFS)

this work

3

Cu^II,II₂(µ-OH)₂

L^Bn, L

1.949(4)

1.924(5)

1.940(4)

1.921(5)

1.922(6)

1.951(6)

1.930(6)

1.951(6)

1.950(6)

1.933(6)

1.975(6)

1.947(6)

100.8(2)

100.5(2)

Bn

3

2

L^{iPr H}

3.0088(13)

2.445(5)

100.9(3)

102.7(3)

this work

2

L^{iPr Bn}

3.037(3)

3.033(3)

2.42(1)

2.49(1)

102.9(3)

101.3(3)

2

^aTp^iPr) hydrotris(3,5-diisopropyl-1-pyrazolyl)borate; phen ) 1,10-phenanthroline; 5-Me₃-TPA ) tris(5-methylpyridyl-2-methyl)amine. ^bKitajima,

N.; Singh, U. P.; Amagai, H.; Osawa, M.; Moro-oka, Y. J. Am. Chem. Soc. 1991, 112, 7757-7758. ^cStebler, M.; Ludi, A.; Bu¨rgi, H.-B. Inorg.

Chem. 1986, 25, 4743-4750. ^dZang, Y.; Pan, G.; Que, L., Jr.; Fox, B. G.; Mu¨nck, E. J. Am. Chem. Soc. 1994, 116, 3653-3654. ^eFrom EXAFS.

2

intramolecular C-D‚‚‚O hydrogen bonds involving the benzylic

C-D bonds and the bridging oxo group (Figure 9). Charac-

terization of C-H‚‚‚O interactions in organic crystals has been

discussed extensively,^49-51the recent analyses by Steiner and

11564 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

Figure 9. Chem3D representation of the hydrogen bonding interactions

in the core of [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂as determined by X-ray

3

crystallography, but using corrected C-D distances of 1.10 Å that

slightly affect other distances and angles involving the D atoms.

Saenger of carbohydrate data on the basis of metrical parameters

d_H‚‚‚O, d_C‚‚‚O, and R_CH‚‚‚O(the C-H-O angle) being particularly

pertinent to this discussion.^50,51In an examination of C-H‚‚‚O

interactions in crystal structures obtained by neutron diffraction,

they noted a prominent class of intramolecular hydrogen bonds

involving H and O atoms separated by four covalent bonds

(formally S5,⁵²where S denotes “self” and 5 refers to the number

of atoms in the ring):

Figure 10. EXAFS data (k ) 2-10.5 Å^-1) (- - -) and simulations (s)

for [(L^BnCu)₂(µ-O)₂](ClO₄)₂: (top) r′-space spectrum; (bottom) k-space

3

spectrum.

Mean geometric parameters for this type of interaction, which

do vary over a significant range, are d_H‚‚‚O) 2.60 Å, d_C‚‚‚O

)

2.92 Å, and R_CH‚‚‚O) 96°. Comparison of these parameters

with those for the bis(µ-oxo)dicopper complex shown in Figure

9 supports the presence of such intramolecular hydrogen bonds

in this molecule, as shown by the d_H‚‚‚O(corrected by extending

the 0.99 Å C-D distance assigned by X-ray crystallography to

1.10 Å) and the more accurately determined d_C‚‚‚Ovalues in

the complex that are shorter than the mean values for the

carbohydrates. We also note that the d_H‚‚‚O(either using the

X-ray C-H distance of 0.99 Å or the corrected distance of 1.10

Å) are shorter than the sum of the van der Waals radii of the O

and H atoms (∼2.7 Å),⁵³although it has been argued that the

importance of this observation in an analysis of C-H‚‚‚O

hydrogen bonding is questionable.⁵⁰In any case, it is clear that

C-H‚‚‚O bonding interactions exist in the solid state structure

of the bis(µ-oxo)dicopper compound. These ground state

interactions may be important in the mechanism by which these

molecules decompose via oxidative N-dealkylation, which

involves rate-determining cleavage of the hydrogen-bonded

C-H bonds,³⁵as well as in their vibrational spectroscopic

properties (see discussion of Raman data below).

X-ray Absorption Spectroscopy. Analysis of EXAFS data

collected on solid samples of [(LCu)₂(µ-O)₂](ClO₄)₂(L ) L^Bn

3

or L^iPr) corroborated the X-ray crystallographic results for the

3

(49) For example, see: (a) Green, R. D. Hydrogen Bonding by C-H

Groups; Wiley: New York, 1974. (b) Taylor, R.; Kennard, O. J. Am. Chem.

Soc. 1982, 104, 5063-5070.

(50) Steiner, T.; Saenger, W. J. Am. Chem. Soc. 1992, 114, 10146-

10154 and references cited therein.

Figure 11. EXAFS data (k ) 2-10.5 Å^-1) (- - -) and simulations (s)

for [(L^iPrCu)₂(µ-O)₂](ClO₄)₂: (top) r′-space spectrum; (bottom) k-space

3

spectrum.

(51) For additional, relevant analyses of C-H‚‚‚O hydrogen bonds

involving (a) water molecules see: Steiner, T.; Saenger, W. J. Am. Chem.

Soc. 1993, 115, 4540-4547, and (b) carbonyl ligands in organometallic

complexes see: Braga, D.; Grepioni, F.; Biradha, K.; Pedireddi, V. R.;

Desiraju, G. R. J. Am. Chem. Soc. 1995, 117, 3156-3166.

(52) Etter, M. C.; MacDonald, J. C.; Bernstein, J. Acta Crystallogr. 1990,

B46, 256-262.

complex capped by L^Bnand allowed definition of the core for

3

the analogous compound ligated by L^iPr. The r′-space spectra

3

for both complexes (top plots in Figures 10 and 11) are

dominated by intense features at r′ ≈ 1.5 Å and r′ ≈ 2.5 Å,

corresponding to scatterers in the first and second coordination

spheres, respectively, where the copper-scatterer distance r

(53) Bondi, A. J. Phys. Chem. 1964, 68, 441-451.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11565

iPr

9

Table 6. EXAFS Analysis of [(LCu)₂(µ-O)₂](ClO₄)₂(L ) L^Bnor L^iPr) and [(L Cu)₂(µ-OH)₂](ClO₄)₂

3

1st sphere

2nd sphere

r

σ²

r

σ²

ꢀ²(×10⁴)

a

b

[(L^BnCu)₂(µ-O)₂](ClO₄)₂

4 O/N at 1.90 Å

0.002

30.4

9.2

3

1 Cu at 2.84 Å

1 Cu at 2.86 Å

9 C at 2.79 Å

0.005

0.006

2.0

[(L^iPrCu)₂(µ-O)₂](ClO₄)₂

4 O/N at 1.89 Å

-0.001

0.000

34.0

8.7

3

1 Cu at 2.88 Å

1 Cu at 2.86 Å

9 C at 2.94 Å

0.003

0.005

0.004

1.3

c

[(L^iPrCu)₂(µ-OH)₂](ClO₄)₂

4 O/N at 1.99 Å

1 N at 2.30 Å

4 O/N at 1.98 Å

1 N at 2.29 Å

4 O/N at 1.98 Å

1 N at 2.30 Å

0.005

0.004

0.005

0.004

0.005

0.006

3

14.4

12.2

6.0

1 Cu at 3.07 Å

1 Cu at 3.08 Å

7 C at 2.91 Å

0.009

0.006

0.005

^aFourier transform range 2-10.5 Å^-1, back-transform range 1-3.1 Å. ^bFourier transform range 2-10.5 Å^-1, back-transform range 1-3.0 Å.

^cFourier transform range 2-14.0 Å^-1, back-transform range 1-3.3 Å.

approximately equals r′ + 0.4 Å. In terms of overall appear-

ance, the spectra closely resemble those reported previously for

complexes or proteins with bis(µ-oxo)diiron²⁹and -dimanganese

cores.^30,54In these iron and manganese EXAFS r′-space spectra,

the prominent second-sphere feature is attributed to a metal ion

scatterer at ∼2.7-3.0 Å. Similarly, we assign the second-sphere

features in the EXAFS spectra for the bis(µ-oxo)dicopper

complexes to a Cu scatterer, with additional contributions from

C atoms in the ligand backbone.

The best fits to the k-space spectra are summarized in Table

6 and shown in Figures 10 and 11. In both cases, the first

coordination sphere feature was best fit using a single shell of

four O/N scatterers at 1.89-1.90 Å, which corresponds to the

average of the Cu-µ-O (1.81 Å) and Cu-N_eq(1.99 Å) bond

lengths found in the crystal structure of [(d₂₁-L^BnCu)₂(µ-O)₂]-

3

(SbF₆)₂. Splitting of this shell into two subshells was not

possible due to the limited resolution of the data sets (k )

2-10.5 Å^-1, ∆r ) 0.19 Å).⁵⁵The ∼2.3 Å Cu-N_axdistance

Figure 12. 500 MHz ¹H NMR spectra of (top) [(d₂₁-L^BnCu)₂(µ-O)₂]-

3

observed in the crystal structure was not required to obtain a

(ClO₄)₂and (bottom) [(L^BnCu)₂(µ-O)₂](ClO₄)₂in 2:1 (CD₃)₂CO/CD₂-

Cl₂at -75 °C (s denotes peaks due to residual hydrogen atoms in the

solvents; * indicates THF impurity present in the Cu(I) precursor).

3

good fit.

The second coordination sphere was best fit with a copper

scatterer at 2.86 Å and additional carbon scatterers in both of

iPr₃

the L^Bn- and L -capped complexes. Attempted fits using only

between the Cu and C atoms makes it somewhat difficult to

3

extract accurate parameters for the Cu-only or C-only shells.

In any case, the correspondence between the EXAFS data and

Cu or only C to model the second-sphere feature were of poorer

quality. However, the Cu‚‚‚Cu distances determined by analysis

of the EXAFS data were 0.07 Å longer than that determined

iPr₃

fits for the bis(µ-oxo)dicopper complexes ligated by L^Bnor L

,

3

by X-ray crystallography {Cu‚‚‚Cu ) 2.79 Å in [(d₂₁-L^BnCu)₂-

in conjunction with the general agreement of the former with

the X-ray crystallographic data, indicates that the [Cu₂(µ-O)₂]²⁺

core structure is basically the same for both compounds.

NMR Spectroscopy. In Figure 12 are shown ¹H NMR

3

(µ-O)₂](SbF₆)₂}. A similar ∼0.07 Å shift in the Cu‚‚‚Cu

distance was observed in the EXAFS analysis of [(L^iPrCu)₂(µ-

3

OH)₂](ClO₄)₂compared to the crystal structure of its L^{iPr H}

-

2

Bn₃

spectra for [(L^BnCu)₂(µ-O)₂](ClO₄)₂(bottom) and [(d₂₁-L

-

3

capped analog. We attribute this apparently systematic dis-

crepancy to the Cu-shell parameters used in the EXAFS analysis,

which were derived from data acquired for [Cu(TMEN)(µ-

OH)]₂Br₂‚H₂O (TMEN ) tetramethylethylenediamine).⁵⁶There

are a number of carbon atoms at similar distance as the other

copper atom in this structure, and the resulting interference

Cu)₂(µ-O)₂](ClO₄)₂(top) in 2:1 (CD₃)₂CO/CD₂Cl₂at -75 °C.

Comparison of the two spectra allows assignment of the peaks

as shown. The chemical shifts differ from those of the Cu(I)

starting materials {7.3, 3.9, 2.7, 2.5, and 2.4 ppm for [L^BnCu-

3

(CH₃CN)]ClO₄at -75 °C}, which are clearly not present in

these solutions, but the differences are not as large as would be

expected if there were significant paramagnetic contributions.

Thus, the presence of sharp peaks with chemical shifts in the

0-10 ppm region and the absence of signals in EPR spectra

clearly indicate that the bis(µ-oxo)dicopper compounds are

diamagnetic. Sharp resonances in the normal regions also are

present in ¹³C{¹H} NMR spectra (Table 1). The simplicity of

the NMR spectra is not consistent with the structure of the

complex as determined in the solid state, which would be

expected to have many more inequivalent hydrogens due to the

asymmetry inherent to the square pyramidal copper coordination

(54) Recent lead references: (a) DeRose, V. J.; Mukerji, I.; Latimer, M.

J.; Yachandra, V. K.; Sauer, K.; Klein, M. P. J. Am. Chem. Soc. 1994, 116,

5239-5249. (b) Riggs-Gelasco, P. J.; Mei, R.; Yocum, C. F.; Penner-Hahn,

J. E. J. Am. Chem. Soc. 1996, 118, 2387-2399.

(55) The data collected for both complexes contained significant artifacts

in the region above k ) 10.5 Å^-1which could not be corrected for. These

artifacts may have arisen from imperfections in the monochromator crystal

used in the data collection. The use of data beyond k ) 10.5 Å^-1(as was

reported previously)²⁷engendered artifactual features in their Fourier-

transformed spectra that could only be fit with parameters requiring

unreasonably negative ∆σ²values.

(56) Mitchell, T. P.; Bernard, W. H.; Wasson, J. R. Acta Crystallogr.

1970, B26, 2096.

11566 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

Figure 13. Resonance Raman spectra obtained with 457 nm excitation

Figure 14. Resonance Raman spectra obtained with 457 nm excitation

on frozen (77 K) CH₂Cl₂solutions of (A) [(L^BnCu)₂(µ-O)₂](ClO₄)₂,

on frozen (77 K) solutions of (A) [(L^iPrCu)₂(µ-O)₂](ClO₄)₂in THF,

3

18

Bn

18

(B) [(L^BnCu)₂(µ- O)₂](ClO₄)₂, (C) [(d₂₁-L Cu)₂(µ-O)₂](ClO₄)₂, and

(B) [(L^iPrCu)₂(µ- O)₂](ClO₄)₂in THF, (C) [(L^{iPr Bn}Cu)₂(µ-O)₂](ClO₄)₂

3

2

18

(D) [(d₂₁-L^BnCu)₂(µ- O)₂](ClO₄)₂. Peaks most sensitive to oxygen

in acetone, and (D) [(L^{iPr Bn}Cu)₂(µ-¹⁸O)₂](ClO₄)₂in acetone. Peaks most

3

2

isotope substitution are labeled with their Raman shifts; s denotes bands

derived from CH₂Cl₂solvent.

sensitive to oxygen isotope substitution are labeled with their Raman

shifts; s denotes bands derived from acetone solvent; peaks due to THF

solvent in spectra A and B have been removed by subtraction.

environments. A rapid fluxional process must occur in solution

to interchange the inequivalent equatorial and axial coordination

positions and thus simplify the spectra, but the nature of this

process(es) is not known. We favor an intramolecular rear-

rangement (cf. turnstile rotation), primarily in view of the results

of cross-over experiments presented elsewhere³⁵that argue

against the generation of monomeric fragments or the inter-

change of capping triazacyclononane ligands in solutions of the

bis(µ-oxo)dicopper compounds. Theoretical calculations also

support a low energy barrier for an intramolecular half-turnstile

rotation process (vide infra).

(M ) Fe or Mn) at 660-700 cm^{-1 29,57}

The slightly higher

.

energies of the latter features are likely due to a combination

of factors, including higher charge and smaller masses of the

metal centers and variations in the M-O-M angles. In support

of the vibrational assignment, the ¹⁸O isotope shifts for all of

the copper complexes are 20-24 cm^-1, approaching the

expected shift of 28 cm^-1for a simple Cu-O harmonic

oscillator at 600 cm^-1

. As described below, the observed

frequency and isotope shift agree well with those of the

theoretically calculated [Cu₂(µ-O)₂]²⁺rhomb [ν(¹⁶O) ) 630

cm^-1; ∆(¹⁸O) ) 29 cm^-1].⁴⁷Finally, the ca. 600 cm^-1feature

for [(d₂₁-L^BnCu)₂(µ-O)₂](ClO₄)₂exhibits a depolarization ratio

Resonance Raman Spectroscopy. Further structural insights

into the unique bis(µ-oxo)dicopper cores were obtained from

analysis of resonance Raman spectra collected on frozen

solutions at 77 K using 457.9 nm excitation (nearly coincident

with the longer wavelength CT absorption band of the com-

3

of 0.57 in CH₂Cl₂at -80 °C, signifying that the core vibration

is polarized and totally symmetric,⁵⁸consistent with our A₁

assignment.

We have also observed that the Cu₂(µ-O)₂core vibration is

sensitive to substitution of deuterium into the benzyl substituents

plexes). Spectra of solutions of [(L^BnCu)₂(µ-O)₂](ClO₄)₂and

3

isotopomers with perdeuterated benzyl substituents and/or ¹⁸O

in the core in CH₂Cl₂are shown in Figure 13. Data for ¹⁶O-

3

of L^Bn, downshifting by 7-10 cm^-1for the ¹⁶O compound and

4 cm^-1for the ¹⁸O analog; this effect is much less pronounced

and ¹⁸O-substituted complexes capped by L^iPr(in THF) and

for L ) L^iPr. This intriguing effect of benzylic deuteration

3

L^{iPr Bn}(in acetone) are presented in Figure 14, with oxygen

2

indicates coupling of ligand-based vibrations with that of the

Cu₂(µ-O)₂core. The coupling may be rationalized by consid-

isotope-sensitive features for all spectra listed in Table 1.

eration of the X-ray crystal structure of [(d₂₁-L^BnCu)₂(µ-O)₂]-

3

While a number of resonance enhanced Raman features are

evident in the spectra, only one set of features near ∼600 cm^-1

is common to all of the complexes and is sensitive to

(SbF₆)₂which shows the presence of interactions between the

oxo bridges and the benzylic deuterium atoms (vide supra); these

interactions may provide the conduit for the vibrational coupling.

Following this reasoning, the absence of a significant effect of

¹⁸O-substitution. For L ) L^iPr, a prominent peak is observed

3

at 600 cm^-1which shifts to 580 cm^-1upon ¹⁸O substitution;

deuteration on the 600 cm^-1feature of the L^iPrcomplexes

3

the corresponding features for L^{iPr Bn}are found at 594 and 572

2

suggests that such C-H‚‚‚O interactions may be less important

in the latter compounds, a hypothesis that needs corroboration.

Whatever the rationale for the deuterium effect, the near

constancy of the 600 cm^-1vibration across the variety of

triazacyclononane ligands studied argues that the assignment

of this mode to a Cu₂(µ-O)₂core remains a good approximation.

cm^-1, respectively. Interestingly for L ) L^Bn, there are two

3

peaks at 602 and 612 cm^-1, which shift to 583 cm^-1when the

complex is prepared with ¹⁸O₂. The conversion of the two

features in the complex generated with ¹⁶O₂to one when the

complex is generated with ¹⁸O₂suggests that the doublet results

from a Fermi resonance of the oxygen-sensitive vibration and

an overtone of a fundamental vibration near 300 cm^-1or a

combination band arising from the ligand. Similar behavior

has been noted for [Fe₂O₂(5-Me₃-TPA)₂]³⁺[5-Me₃-TPA ) tris-

(5-methylpyridyl-2-methyl)amine].^29bWe assign these ¹⁸O-

sensitive features at ∼600 cm^-1to an A₁vibration of the [Cu₂-

(µ-O)₂]²⁺core, by analogy to those of related [M₂(µ-O)₂]³⁺units

(57) (a) Czernuszewicz, R. S.; Dave, B.; Rankin, J. G. In Spectroscopy

of Biological Molecules; Hester, R. E., Girling, R. B., Eds.; Royal Society

of Chemistry: Cambridge, 1991; pp 285-288. (b) Gamelin, D. R.; Kirk,

M. L.; Stemmler, T. L.; Pal, S.; Armstrong, W. H.; Penner-Hahn, J. E.;

Solomon, E. I. J. Am. Chem. Soc. 1994, 116, 2392-2399.

(58) Nakamoto, K. Infrared and Raman Spectra of Inorganic and

Coordination Compounds, 4th ed.; John Wiley & Sons: New York, 1986;

pp 72-76.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11567

Table 8. Relative Energies (kcal/mol) of {[(NH₃)₃Cu]₂(µ-O)₂}²⁺

Structures^a

level of theory

anti^b

eclipsed

TS^q

RHF/STO-3G

XR/STO-3G

BLYP/STO-3G

CAS(2,2)/DZP

CASPT2N(2,2)/DZP

0.0

1.8

2.9

2.3

-1.7

8.7

2.7

2.4

1.9

^aAll results are at the RHF/STO-3G* geometries. ^bAbsolute energies

(au) for this column are -3723.975 99, -3718.784 80, -3729.233 71,

-3763.304 06, and -3765.175 94.

deviations are observed for the various bond angles O-Cu-N,

but this is only to be expected insofar as the ammonia nitrogen

atoms in the theoretical structure are unconstrained in contrast

to those found in the triazacyclononane ligands. In addition,

the Cu-O bonds are too short in the theoretical structure by

about 0.025 Å. This may be attributable to the electronic and

steric differences between the ammonia ligands and the triaza-

cyclononanes and/or it may reflect the overbinding associated

with the single-determinant RHF wave function (vide infra). In

any case, it appears possible that this level of theory will

continue to be useful for providing reasonably accurate structures

for future modeling studies.

The relative energies for the two local minima were deter-

mined at several different levels of theory, and the results are

provided in Table 8. At the RHF/STO-3G level, the two

stereoisomers differ by only 1.8 kcal/mol. However, this level

of theory should not, a priori, be expected to be especially

accurate. In addition to the small basis set, the use of a single

determinant (implicit in HF theory) may be an inadequate

approximation for this system. Simple [Cu₂(µ-η²:η²-O₂)]²⁺, for

instance, has been shown to be characterized by a significant

mixing of multiple determinants in its lowest closed-shell singlet

state.^60,61One approach to treat this configurational mixing is

to use density functional theory, where the theoretical method

optimizes the energy with respect to the total electron density

as opposed to the wave function. DFT has been shown for a

variety of systems to accurately account for configurational

mixing in closed-shell singlets (based on calculations of

geometries, multiplet splittings, etc.).⁶²Interestingly, the energy

difference between the anti and eclipsed conformers increases

only slightly when either XR or the gradient-corrected BLYP

levels of density functional theory are employed in place of

Hartree-Fock theory. Furthermore, there is no tendency for

the R and â Kohn-Sham molecular orbitals to break symmetry.

Indeed, symmetry broken starting densities smoothly converge

to fully symmetric solutions indistinguishable from restricted

calculations. This is in contrast to results obtained for [Cu₂-

(µ-η²:η²-O₂)]²⁺model systems,⁶³and may be taken as evidence

of the highly covalent nature of the bonding in the {[(NH₃)₃-

Cu]₂(µ-O)₂}²⁺system.⁶⁴

Figure 15. Structures for anti (A) and eclipsed (B) {[(NH₃)₃Cu]₂(µ-

O)₂}²⁺(optimized at the RHF/STO-3G level).

Table 7. Selected Bond Lengths (Å) and Angles (deg) for

Calculated and Experimental [Cu₂(µ-O)₂]²⁺Complexes

anti

eclipsed

{[(NH₃)₃Cu]₂- {[(NH₃)₃Cu]₂- [(d₂₁-L^BnCu)₂-

3

(µ-O)₂}²⁺

(µ-O)₂](SbF₆)₂

Cu-Cu

2.734

2.281

1.780

2.207

1.996

79.7

100.3

100.9

90.7

2.741

2.275

1.781

2.205

2.000

79.4

100.6

103.9

90.0

2.794

2.287

1.803, 1.808

2.298

1.986, 1.987

78.6

O-O

Cu-O

Cu-N_axial

Cu-N_equatorial

O-Cu-O

Cu-O-Cu

O-Cu-N_axial

O-Cu-N_eq,syn

O-Cu-N_eq,anti

101.4

102.3, 103.8

95.6, 96.2

171.1, 172.3

163.7

161.5

Nevertheless, it is clear that the Raman spectra of these Cu₂-

(µ-O)₂complexes provide a wealth of information that has yet

to be mined. For the L^iPrcase, low intensity peaks at 614 and

3

588 cm^-1that are sensitive to both ¹⁸O and deuterium

substitution have yet to be assigned. The remaining, non-¹⁸O-

sensitive peaks in the spectra of all the compounds shift upon

ligand deuteration to varying extents and are therefore attibutable

to ligand-based vibrations with at least partial substituent

C-CH_nstretching and/or bending character. A detailed Raman

study of these features is in progress and will be the subject of

a future report.⁵⁹

Theoretical Calculations. In order to obtain insights into

the stability and electronic structure of the novel bis(µ-oxo)-

dicopper core ab initio calculations at the restricted Hartree Fock

(RHF) level using the STO-3G basis set were performed on

the model system {[(NH₃)₃Cu]₂(µ-O)₂}²⁺; the experimental

X-ray crystallographic coordinates for corresponding atoms in

As a more rigorous test of the importance of multiple

configurations, we carried out 2-electron-in-2-orbital (2,2)

complete active space multiconfiguration self-consistent field

calculations (CAS). This calculation permits partial occupation

[(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂were used as a starting point. Two

3

different local minima, both of which are closed-shell singlet

states, were found for {[(NH₃)₃Cu]₂(µ-O)₂}²⁺at the RHF/STO-

3G level of theory; they are illustrated in Figure 15. The major

difference between the structures is the relative disposition of

the axial N-donor ligands, anti in structure A versus eclipsed

in structure B. The critical heavy atom-heavy atom bond

lengths and angles in these structures are provided in Table 7,

together with those determined from the X-ray crystal structure

(60) Maddaluno, J.; Geissner-Prettre, C. Inorg. Chem. 1991, 30, 3439.

(61) Bernardi, F.; Bottoni, A.; Casadio, R.; Fariselli, P.; Rigo, A. Int. J.

Quantum Chem. 1996, 58, 109-119.

(62) (a) Cramer, C. J.; Dulles, F. J.; Storer, J. W.; Worthington, S. E.

Chem. Phys. Lett. 1994, 218, 387. (b) Cramer, C. J.; Dulles, F. J.; Falvey,

D. E. J. Am. Chem. Soc. 1994, 116, 9787. (c) Cramer, C. J.; Worthington,

S. E. J. Phys. Chem. 1995, 99, 1462. (d) Lim, M. H.; Worthington, S. E.;

Dulles, F. J.; Cramer, C. J. In Density-Functional Methods in Chemistry;

Laird, B. B., Ross, R., Ziegler, T., Eds.; American Chemical Society:

Washington, DC, 1996; p 402.

of [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂, which has an anti arrangement

3

of axial N-ligands. Overall, the agreement between the theoreti-

cal and experimental anti structures is remarkable. The largest

(63) Ross, P. K.; Solomon, E. I. J. Am. Chem. Soc. 1991, 113, 3246-

3259.

(59) Wilkinson, E. C.; Rodgers, K.; Mahapatra, S.; Que, L., Jr.; Tolman,

W. B. Work in progress.

(64) Lovell, T.; McGrady, J. E.; Stranger, R.; Macgregor, S. A. Inorg.

Chem. 1996, 35, 3079-3080.

11568 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

Figure 17. Transition state structure that interconverts anti and eclipsed

{[(NH₃)₃Cu]₂(µ-O)₂}²⁺(optimized at the RHF/STO-3G* level).

conformer there is a 65:35 mixing of the HOMO²and LUMO²

states. Moreover, the nontrivial difference between the CAS-

(2,2) and CASPT2N(2,2) relative energies suggests that a larger

active space might be important to account for additional

important nondynamic correlation effects. However, insofar as

we are using these calculations simply to provide some insight

into the experimentally characterized systems, we did not

examine in detail any more costly calculations employing larger

active spaces here.

One noteworthy feature of the CAS calculations is that at

this level of theory both the anti and eclipsed conformers are

unstable to dissociation into fragments. This can be understood

in terms of the substantial contribution of the LUMO²config-

uration, which is antibonding between all rhomb atoms. This

feature, combined with the Coulomb repulsion between the two

cationic centers, renders the complexes unstable. It is possible

that this is an artifact that would disappear were geometry

optimization to proceed with further consideration of electron

correlation (although we did not find expansion of the active

space to (8,8) to be helpful), or it may be that {[(NH₃)₃Cu]₂-

(µ-O)₂}²⁺is indeed unstable to dissociation in the gas phase.

Of course, larger ligands, condensed-phase effects, and the

presence of counterions, all of which occur for the experimen-

tally characterized systems, would be expected to strongly

mitigate the Coulomb repulsion in the dication. In any case, it

appears that the RHF/STO-3G level serendipitously provides

qualitatively useful geometries and energies through cancelling

errors in the various theoretical approximations.

Other results from the theoretical calculations may be used

to further our understanding of the experimental behavior of

these systems. The observation of stable closed-shell singlet

states for the calculated structures corroborates the experimental

finding that the bis(µ-oxo) complexes are diamagnetic, as

revealed by EPR silence and sharp peaks in the diamagnetic

regions of NMR spectra. Moreover, the NMR results imply a

low barrier to the exchange of axial and equatorial N-donor

ligand positions (note that double-label experiments preclude

dissociation/reassociation as the operative mechanism). The

small calculated difference in energies for the anti and eclipsed

conformers at all levels of theory offers some support for this

hypothesis, and to further examine it we located a transition-

state structure that interconverts the two conformers via a half-

turnstile-like⁶⁶rotation of the nitrogen ligands at one copper

atom (Figure 17). The barrier associated with this rotation is

only 2.7 kcal/mol at the RHF/STO-3G level. At the density

functional levels, the transition state structure is calculated to

have a slightly lower energy than the eclipsed structure.

Presumably this is simply an artifact of not reoptimizing all of

the geometries at this level; the important conclusion to be drawn

from these calculations is that the barrier to ligand rotation is

Figure 16. From bottom to top, the HOMO, LUMO, and LUMO +

1 orbitals of anti {[(NH₃)₃Cu]₂(µ-O)₂}²⁺calculated at the RHF/STO-

3G* level. The plots are for the 0.032 bohr^-3isodensity surface.

of both the highest occupied and lowest unoccupied molecular

orbitals (HOMO and LUMO, respectively). For the anti isomer,

these orbitals are illustrated in Figure 16, along with the LUMO

+ 1 which will be discussed further below. The HOMO, 12a_u,

2

is formed from a combination of copper d_{x -y}orbitals and

oxygen p orbitals. The mixing is net bonding between copper

and oxygen and copper and copper, but net antibonding between

oxygen and oxygen (i.e., it may be thought of as a bonding

combination of a copper-copper π bond with an oxygen-

oxygen σ* bond). Note that this HOMO is spatially similar to

that previously reported for the (µ-η²:η²-peroxo)dicopper core,⁶³

but with greater overlap between the fragment orbitals due to

the different interatomic distances in the bis(µ-oxo) unit. The

2

LUMO, 12b_g, hybridizes copper d_{x -y}orbitals with oxygen p

orbitals in a different manner that is net antibonding between

all atoms in the rhomb. This pair of MOs correlates with the

HOMO (13b₁) and LUMO (10a₂) in the eclipsed conformer.

To increase the accuracy of these calculations, a considerably

more flexible polarized double-ú basis set⁶⁵was employed.

Although the CAS(2,2) calculations reverse the relative energies

of the two conformers, the net effect is only 3.5 kcal/mol. When

additional account of correlation is taken via multireference

second-order perturbation theory (CASPT2N), the anti isomer

is again found to be lower, and moreover by a larger margin,

namely 8.7 kcal/mol. From examination of the coefficients

associated with the two possible closed-shell singlet determi-

nants (i.e., for the anti conformer |...12a²_u12b_g⁰vs

|...12a⁰_u12b²_g), it is apparent that multiconfigurational wave

functions are a hallmark of the {[(NH₃)₃Cu]₂(µ-O)₂}²⁺system,

analogous to the (µ-η²:η²-peroxo)dicopper core. In each

(66) Ugi, I.; Marquarding, D.; Klusacek, H.; Gillespie, P.; Ramirez, F.

(65) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.

Acc. Chem. Res. 1971, 4, 288.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11569

Summary and Perspectives

very small. Although this barrier would certainly be expected

to increase in the actual experimental systems because of steric

congestion present in the eclipsed structures, it still appears quite

reasonable to postulate rapid rotation on the NMR time scale

at the experimental temperatures.

We have constructed a coherent picture of the structure and

bonding of a series of bis(µ-oxo)dicopper complexes, which

represent a new motif in copper-dioxygen chemistry, through

a combination of experimental and theoretical investigations.

A topology for the [Cu₂(µ-O)₂]²⁺rhomb characterized by short

Cu-O (∼1.80 Å) and Cu‚‚‚Cu (∼2.80 Å) distances was

revealed by X-ray crystallography, EXAFS spectroscopy, and

ab initio calculations. These interatomic distances are similar

to those previously identified in analogous cores comprised of

iron or manganese ions and are notably shorter than those typical

for bis(µ-hydroxo)dicopper(II) compounds (∼1.95 and ∼3.00

Å, respectively). Examples of the latter type of compound are

produced upon decomposition of the thermally unstable bis(µ-

oxo)dicopper complexes in reactions involving N-dealkylation

of the triazacyclononane ligands, as verified by X-ray crystal-

lographic studies. Spectroscopic and electrospray mass spec-

trometry experiments showed that the bis(µ-oxo)dicopper core

remains intact in solution at low temperature and revealed

specific geometric and electronic structural features. For

example, we identified a vibrational signature of the [Cu₂(µ-

O)₂]²⁺core that is coupled to macrocycle ligand vibrations in

resonance Raman spectra of isotopically labeled compounds.

In addition, the diamagnetism of the bis(µ-oxo)dicopper species,

as well as the existence of a fluxional process in solution that

interchanges equatorial and axial N-donor ligand positions, was

revealed by low-temperature NMR and EPR spectroscopy. Ab

initio theory for a model system predicts a closed-shell singlet

ground-state structure that agrees well with the experimentally

determined geometry and provides a basis for understanding

key properties of the bis(µ-oxo)dicopper core, such as its

diamagnetism, its resonance Raman signature, and the low

barrier to the fluxionality involving the N-donors. These

theoretical calculations, in conjunction with analysis of the

experimental data, are consistent with a formal portrayal of the

core as [Cu^III₂(µ-O^2-)₂]²⁺, but further suggest that a molecular

orbital description involving a high degree of mixing of oxygen

and copper fragment orbitals (i.e. covalency) is appropriate.

The verification of the anti and eclipsed structures as local

minima at the RHF/STO-3G level was accomplished by analytic

frequency calculations. These calculations were then used to

determine the frequency shift associated with ¹⁸O isotopic

substitution for the Raman active rhomb breathing mode

discussed above. For the anti conformer, we calculate ν(¹⁶O)

) 633 cm^-1and ν(¹⁸O) ) 604 cm^-1, i.e., a frequency shift of

29 cm^-1. These data compare favorably with the experimentally

observed absorptions and isotopic shifts in the complexes

described above and further support the experimental assign-

ments of this fundamental. For the eclipsed conformer, we

calculate ν(¹⁶O) ) 630 cm^-1and ν(¹⁸O) ) 601 cm^-1. Although

it is possible that the experimental observation of two peaks in

the Raman spectrum for this fundamental corresponds to the

presence of both anti and eclipsed triazacyclononane conformers

(with an accidental isochronicity in the ¹⁸O case), this is probably

less likely than the above proposed Fermi resonance for the

¹⁶O case. The steric demands of the triazacyclononane ligands

are considerably larger than those of the theoretical ammonia

ligands, so that the equilibrium population of the eclipsed

conformer might be expected to be below detection limits. In

the solid state, on the other hand, it may be that crystal packing

effects (involving additionally counterions and solvent mol-

ecules) can counter the higher steric energy associated with the

eclipsed conformation, and it is intriguing to speculate that this

mode of complexation may be observable under suitable

conditions and/or in different synthetic or biological systems.

The LUMO + 1 orbital (Figure 16) provides some insight

into the apparent electrophilic character exhibited by the rhomb

oxygen atoms (as judged by their proclivity to indulge in

hydrogen atom abstractions, which is discussed more fully in a

separate paper).³⁵This 13a_uorbital is formed from an all

2

antibonding combination of copper d_{x -y}orbitals and oxygen

p orbitals, i.e., the same fragment orbitals used to construct the

12a_uHOMO but taken with opposite phase. In the 12a_uHOMO,

the copper d orbital coefficients are larger than the oxygen p

orbital coefficients; in the 13a_uLUMO + 1, it is the oxygen p

orbital coefficients that dominate. This orbital would be

expected to be quite high in energy in a peroxo species, since

it is dominated by σ*_OOcharacter, but the large separation

between the oxygen atoms in the rhomb for the bis-oxo motif

stabilizes this virtual orbital so that it is only very slightly higher

in energy than the 12b_gLUMO. The 13a_uorbital, a low-energy

acceptor orbital dominated by contributions from in-plane

oxygen p fragments, thus rationalizes the observed electrophi-

licity of the rhomb oxygen atoms in the experimental systems.

The reaction of the LCu(I) precursors with O₂in a 2:1 ratio

to generate bis(µ-oxo)dicopper(III) species stands in stark

contrast to the usual process involving a 4:1 stoichiometry and

production of µ-oxo- or µ-hydroxo-bridged Cu(II) compounds

(eq 1). As in the typical 4:1 reaction, a (µ-peroxo)dicopper(II)

complex is a likely intermediate along the pathway to the bis-

(µ-oxo)dicopper(III) product. Instead of being reduced by

additional Cu(I) ions, however, O-O bond cleavage occurs

entirely intramolecularly; indeed, we have observed this conver-

28

sion and its reverse directly for the case where L ) L^iPr

.

In

3

a formal sense, the bis(µ-oxo)dicopper(III) core can be viewed

as being derived from homolysis of the (µ-peroxo) O-O bond

concomitant with one electron oxidation of each of the

copper(II) ions.⁶⁷Thus, the (µ-peroxo)dicopper(II) and bis(µ-

oxo)dicopper(III) cores are isomers residing at the same overall

oxidation level and, therefore, each can be considered equally

competent as potential oxidants in copper/dioxygen chemistry

and biochemistry.

Finally, a comparison of the HOMO in the bis(µ-oxo) core

to that reported for the analogous µ-η²:η²-peroxo system^2,63

provides some insight into the formal oxidation levels of the

atoms in the core. Because of the large distance between the

oxygen atoms in the [Cu₂(µ-O)₂]²⁺rhomb, there is much less

energetic cost associated with mixing a substantial fraction of

oxygen σ* into the HOMO than in the peroxo case where the

O-O bond is present (and, as noted above, this mixing increases

bonding with the copper fragments in the bis(µ-oxo) core). Such

hybridization moves electronic charge off of the copper atoms

and onto the oxygen atoms relative to the peroxo system. Since

the latter are well described as Cu^IIspecies, this orbital

modification may be used to suggest a formal [Cu^III₂(µ-O^2-)₂]²⁺

oxidation state for the bis(µ-oxo)dicopper unit.

The [Cu₂(µ-O)₂]²⁺core also may be envisioned to be a close

counterpart of the [M^IV₂(µ-O^2-)₂]⁴⁺(M ) Fe or Mn) unit, which

for M ) Mn has been structurally defined but which for M )

Fe has yet to be identified in a synthetic complex, notwithstand-

ing the fact that it has captured the imagination of those

investigating how the diiron active site of methane monooxy-

(67) Detailed theoretical and experimental studies of the actual peroxo/

bis(µ-oxo) isomerization process are in progress. See: Cramer, C. J.; Smith,

B. A.; Tolman, W. B. J. Am. Chem. Soc., in press.

11570 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

genase may effect the conversion CH₄f CH₃OH.^31,32,68,69The

analogy between the M₂O₂“diamond-like” cores of Cu, Mn,

and Fe complexes examined to date (including those at lower

oxidation levels) is underscored by their topological similarities

(Table 5). On the other hand, the apparently diverse roles of

the various cores in dioxygen activation and generation pro-

cesses implicate significant reactivity differences among them.

Understanding the geometric and electronic structural bases for

these differences remains an important research objective in

bioinorganic chemistry.

introduced. The instrument was scanned in 0.2-dalton steps, and signal

averaging was used to enhance the signal-to-noise ratio. Routine low

resolution electron impact mass spectra (LREIMS) were obtained on a

Hewlett-Packard GCD series G1800A GC/MS equipped with a 30 m

HP-5 (crosslinked 5% PhMe-silicone) column. Fast atom bombardment

(FAB) mass spectra were determined using a VG 7070E-HF mass

spectrometer with xenon gas ionization in a matrix of m-nitrobenzyl

alcohol (MNBA). Electrochemical experiments were perfomed as

previously described³⁷using an EG&G Princeton Applied Research

(PAR) VeraStat potentiostat driven by EG&G PAR Model 250 software.

X-ray absorption spectra (XAS) were collected at the Cu K-edge

(∼9.0 KeV) at the Cornell High Energy Synchrotron Source (CHESS,

station C2) and the National Synchrotron Light Source (Brookhaven

National Laboratory, station X9). The monochromator was calibrated

using the 8993 eV peak in the spectrum of Cu foil. Data were obtained

in transmission mode [A_exp) log(I₀/I₁)] on samples comprised of

microcrystalline solids dispersed in boron nitride (see below for sample

preparation details). Data analysis was performed as previously

described,²⁹via a variation of the FABM (“fine adjustment based on

models”) procedure that uses theoretical amplitude and phase functions

calculated according to a curved-wave formalism. The crystallographi-

cally characterized complexes [Cu(O₂CCH₃)₂‚H₂O]₂,⁷³[Cu(CH₃CN)₄]-

ClO₄,⁷⁴and [(TMEN)Cu(OH)]₂(Br)₂‚H₂O⁵⁶were used to determine the

empirical amplitude reduction factors (A) and shell-specific edge shifts

(∆E) employed to partially compensate for imperfections in the

theoretical amplitude and phase functions. The viability of these

empirical parameters was tested by using them to analyze EXAFS data

Experimental Section

General Procedures. All reagents and solvents were obtained from

commercial sources and used as received unless noted otherwise.

Solvents were dried according to published procedures⁷⁰and distilled

under N₂immediately prior to use. Dioxygen gas was dried by passing

through two short columns of supported P₄O₁₀and Drierite. Labeled

dioxygen (¹⁸O₂, 99%) was obtained from Cambridge Isotopes, Inc.,

and used without further purification. All air-sensitive reactions were

performed either in a Vacuum Atmospheres inert-atmosophere glovebox

under a N₂atmosphere or by using standard Schlenk and vacuum line

techniques. The salts [Cu(CH₃CN)₄]X (X ) ClO₄^-, SbF₆^-, or CF₃SO₃^-

)

were prepared as described in the literature.⁷¹The ligands 1,4,7-

triisopropyl-1,4,7-triazacyclononane (L^iPr

)

²⁵and 1,4,7-tribenzyl-1,4,7-

3

triazacyclononane (L^Bn

)

²⁶were synthesized using published procedures.

3

for [(L^iPrCu)₂(OH)₂](X)₂(Table 6).

d₇-2-Bromopropane [(CD₃)₃CDBr, 98%] was purchased from Aldrich

and used as received; d₇-benzyl chloride (C₆D₅CD₂Cl) was prepared

by refluxing d₈-toluene (99%; Aldrich) with SO₂Cl₂for ∼30 min in

the presence of a trace amount of benzoyl peroxide.⁷²Deuterium

incorporation was determined by integration of the residual proton

3

Manometry. Low temperature gas uptake experiments were

performed by using an apparatus as described earlier.⁷⁵A standard 25

mL Schlenk tube, used as the reaction vessel, was connected via its

side arm and a short piece of rubber tubing through a glass “T” to a

second Teflon stopcock, then to a three-way glass stopcock which was

linked to both vacuum and dioxygen inlet lines. The powdered sample

[Cu(I) starting material or calibration standard, Vaska’s complex] was

stored in a small flask connected to the Schlenk flask with a ground

glass joint, initially isolated from the system with a Teflon stopcock.

The system, including the Schlenk flask containing the solvent (CH₂-

Cl₂), was cooled to -75 °C using an isopropyl alcohol/dry ice bath

and evacuated and refilled with dioxygen three times to ensure a pure

dioxygen atmosphere. The system was then shut off from the gas/

vacuum inlet with the Teflon stopcock, and the initial pressure was

recorded after allowing equilibration at -75 °C for 30 min. The

powdered sample was then opened and allowed to fall into the solvent

in the Schlenk flask whereupon it started reacting immediately. The

solution was kept at -75 °C and vigorously stirred with a magnetic

stir bar for 4 h. The pressure change in the system was recorded until

it stabilized. The moles of dioxygen gas absorbed were then calculated

using the ideal gas law.

1

signals in H NMR spectra. Elemental analyses were performed by

Atlantic Microlabs of Norcross, GA.

Safety Note. Caution! Perchlorate salts of metal complexes with

organic ligands are potentially explosiVe. Only small amounts of

material should be prepared, and these should be handled with great

care.

Physical Methods. IR spectra were recorded on a Mattson Polaris

FTIR spectrometer, as KBr pellets or neat films between NaCl plates.

Low-temperature UV-vis spectra were recorded on a Hewlett-Packard

HP8452A diode array spectrophotometer (190-820 nm scan range)

by using a custom manufactured vacuum dewar equipped with quartz

windows. Low temperatures were achieved by adding liquid nitrogen

to a MeOH bath. The bath temperature was monitored by using a

resistance thermocouple probe (Fluke 51 K/J Thermometer). Resonance

Raman spectra were collected on a Spex 1403 spectrometer interfaced

with a DM3000 data collection system using a Spectra-Physics Model

2030-15 argon ion laser. The spectra were obtained at 77 K using a

backscattering geometry; samples were frozen onto a gold-plated copper

cold finger in thermal contact with a dewar containing liquid nitrogen.

Raman frequencies were referenced to the intense features at 710 or

806 cm^-1from frozen CH₂Cl₂or THF, respectively. ¹H and ¹³C{¹H}

NMR spectra were recorded on Varian VXR-300 or VXR-500

spectrometers. Chemical shifts (in ppm) are referenced to the residual

solvent peak(s). EPR spectra were obtained by using a Bruker ESP300

spectrometer fitted with a liquid nitrogen finger dewar (77 K, ∼9.4

GHz). Electrospray mass spectral data were collected using a Sciex

API III mass spectrometer (Sciex, Thornhill, Ontario, Canada). Cold

CH₂Cl₂(<-40 °C; syringe cooled externally with dry ice) was

introduced by a syringe infusion pump (Model 22, Harvard Apparatus,

South Natick, MA) through a 76 mm internal diameter fused silica

capillary line at flow rates of ∼10 mL/min for about 5 min, and then

the cold sample solution, which was prepared at -80 °C, was

1,4,7-Triisopropyl(d₇)-1,4,7-triazacyclononane (d₂₁-L^iPr). This

3

ligand was prepared as described in the literature²⁵using d₇-2-

bromopropane and was isolated as clear, colorless oil. Deuterium

content: 99% (-CD(CD₃)₂), 96% (-CD(CD₃)₂). ¹H NMR (300 MHz,

CDCl₃): δ 2.59 (s, 12H) ppm. ¹³C{¹H}NMR (75 MHz, CDCl₃): δ

53.6 (t, ¹J_CD) 19.4 Hz), 52.8, 17.4 (br) ppm. LREIMS: m/z (rel intens)

276 [M⁺, 1], 108 (100).

1,4,7-Tribenzyl(d₇)-1,4,7-triazacyclononane (d₂₁-L^Bn). This ligand

3

26

was synthesized in a similar manner as described for L^Bn

,

except

3

d₇-benzyl chloride was reacted with 1,4,7-triazacyclononane. The

ligand was isolated as thick yellow oil which solidified upon standing.

Deuterium content: 98% (-CD₂C₆D₅). ¹H NMR (CDCl₃, 300 MHz):

δ 2.83 (s, 12H) ppm. ¹³C{¹H} NMR (CDCl₃, 75 MHz): δ 55.3 ppm.

FTIR (KBr, cm^-1): 2277 (C-D), 2165 (C-D), 2045 (C-D); HREIMS

calcd for C₂₇H₁₂D₂₁N₃420.3993, found 420.4007.

1,4-Diisopropyl-7-(p-tolylsulfonyl)-1,4,7-triazacyclononane. 1-(p-

Tolylsulfonyl)-1,4,7-triazacyclononane (5.08 g, 0.018 mol) and 2-bro-

(68) Shteinman, A. A. FEBS Lett. 1995, 362, 5-9.

(69) (a) Feig, A. L.; Lippard, S. J. Chem. ReV. 1994, 94, 759-805. (b)

Liu, K. E.; Johnson, C. C.; Newcomb, M.; Lippard, S. J. J. Am. Chem.

Soc. 1993, 115, 939-947.

(73) Brown, G. M.; Chidambaram, R. Acta Crystallogr. 1973, B29,

(70) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory

Chemicals; Pergamon Press: New York, 1988.

2393-2403.

(74) Cso¨regh, I.; Kierkegaard, P.; Norrestam, R. Acta Crystallogr. 1975,

B31, 314-317.

(71) Kubas, G. J. Inorg. Synth. 1979, 19, 90-92; 1990, 28, 68-70.

(72) Furness, B. S.; Hannaford, A. J.; Smith, P. W. G.; Tatchell, A. R.

Vogel’s Textbook of Practical Organic Chemistry; Longman Scientific &

Technical: Essex, U.K., 1989; p 864.

(75) Ruggiero, C. E.; Carrier, S. M.; Antholine, W. E.; Whittaker, J.

W.; Cramer, C. J.; Tolman, W. B. J. Am. Chem. Soc. 1993, 115, 11285-

11298.

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11571

mopropane (8.84 g, 0.072 mol) were combined in degassed CH₃CN

(20 mL). Na₂CO₃(8.04 g) was added as a solid to the above colorless

solution, and the resultant mixture was heated at reflux under N₂for

18 h. The then yellow mixture was filtered from excess Na₂CO₃, and

the solid washed with fresh CH₃CN. The combined filtrates were

evaporated under reduced pressure to yield a yellow oil with a small

amount of a white solid. The residue was redissolved in CHCl₃(20

mL) and then washed with 1 M NaOH. The organic layer was removed,

and the aqueous phase was extracted with CHCl₃(3 × 10 mL). The

combined CHCl₃layers were dried (MgSO₄), and the solvent removed

under reduced pressure to yield 5.94 g (90%) of a tan solid. Mp: 58-

59 °C. ¹H NMR (CDCl₃, 500 MHz): δ 7.66 (d, J ) 8.0 Hz, 2H), 7.26

(d, J ) 8.0 Hz, 2H), 3.28-3.26 (m, 4H), 2.85-2.83 (m, 4H), 2.75

(septet, J ) 6.5 Hz, 2H), 2.43 (s, 4H), 2.39 (s, 3H), 0.91 (d, J ) 6.5

Hz, 12H) ppm. ¹³C{¹H} NMR (CDCl₃, 125 MHz): δ 142.54, 136.29,

129.33, 126.94, 53.53, 53.51, 52.11, 50.17, 21.28, 18.08 ppm. LRE-

IMS: m/z 366 [M⁺, 1], 141 (100). Anal. Calcd for C₁₉H₃₃N₃O₂S: C,

62.09; H, 9.05; N, 11.43. Found: C, 61.95; H, 9.05; N, 11.43.

(KBr, cm^-1): 2267 (CtN), 657 (SbF₆^-). Anal. Calcd for C₁₇H₃₆-

N₄CuSbF₆: C, 34.34; H, 6.11; N, 9.43. Found: C, 34.43; H, 6.10; N,

9.43.

[L^iPrCu(CH₃CN)]BPh₄. A solution of CuCl (0.20 g, 2.00 mmol)

3

in CH₃CN (5 mL) was treated with a solution of L^iPr(0.52 g, 2.04

3

mmol) in THF (3 mL). After stirring the colorless solution for 30 min,

a solution of NaBPh₄(0.68 g, 1.99 mmol) in CH₃CN (5 mL) was added,

causing the precipitation of NaCl. After stirring for an additional 30

min the mixture was filtered through a pad of Celite and the filtrate

treated with Et₂O (20 mL). Storage at -20 °C caused the deposition

of colorless plate crystals of the product (1.09 g, 80%). ¹H NMR (500

MHz, CD₂Cl₂): δ 7.34-7.32 (m, 8H), 7.04 (t, J ) 6.5 Hz, 8H), 6.89

(t, J ) 7.5 Hz, 4H), 3.05 (septet, J ) 6.5 Hz, 3H), 2.79-2.72 (m, 6H),

2.48-2.42 (m, 6H), 1.94 (s, 3H), 1.20 (d, J ) 6.5 Hz, 18H) ppm.

¹³C{¹H} NMR (125 MHz, CD₂Cl₂): δ 164.6, 135.4, 126.2, 126.1, 122.2,

58.6, 51.2, 20.0, 3.2 ppm. FTIR (KBr, cm^-1): 2265 (CtN), 734

(BPh₄^-), 704 (BPh₄^-). Anal. Calcd for C₄₁H₅₆N₄BCu: C, 72.50; H,

8.31; N, 8.25. Found: C, 72.13; H, 8.33; N, 8.33.

1,4-Diisopropyl-1,4,7-triazacyclononane (L^{iPr H}). A mixture of 1,4-

3

[L^iPrCu(CH₃CN)][B(C₆H₃(CF₃)₂)₄]. Na[B(C₆H₃(CF₃)₂)₄]³⁶(0.35 g,

2

diisopropyl-7-(p-tolylsulfonyl)-1,4,7-triazacyclononane (5.94 g, 0.016

mol) and 18 M H₂SO₄(20 mL) was heated at 120 °C for 18 h under

N₂. After cooling to room temperature, the dark mixture was poured

into crushed ice (100 g), and the resultant dark mixture brought to pH

> 11 with aqueous NaOH. The heterogenous mixture was extracted

with CHCl₃until the organic extracts were colorless (ca. 500 mL total).

The CHCl₃solution of the product was dried (MgSO₄), and the solvent

removed under reduced pressure to yield 3.25 g (94%) of an amber

oil. While suitable for further synthetic use, the crude product may be

purified by Kugelrohr distillation (80-82 °C, 0.01 Torr) to yield a

colorless oil, 2.60 g (75%). ¹H NMR (CDCl₃, 300 MHz): δ 3.36 (br

s, 1H), 2.74 (septet, J ) 6.6 Hz, 2H), 2.57-2.53 (m, 4H), 2.47-2.43

(m, 4H), 2.36 (s, 4H), 0.88 (d, J ) 6.6 Hz, 12H) ppm. ¹³C{¹H} NMR

(CDCl₃, 75 MHz) δ 52.72, 48.90, 47.86, 47.03, 18.49 ppm. FTIR

(neat): 3370 (br, ν_NH), 2966, 2931, 2867, 1680, 1490, 1487, 1462, 1458,

1437, 1382, 1360, 1319, 1303, 1264, 1253, 1190, 1169, 1149, 1132,

0.40 mmol) was dissolved in Et₂O (2 mL), and then diluted with CH₂-

Cl₂(3 mL). A solution of [L^iPrCu(CH₃CN)]SbF₆(0.24 g, 0.40 mmol)

3

in CH₂Cl₂(3 mL) was added dropwise with stirring to the solution of

Na[B(C₆H₃(CF₃)₂)₄] causing the immediate deposition of a colorless

precipitate (NaSbF₆). The mixture was stirred for 30 min and then

filtered through a pad of Celite. The solvent was removed from the

nearly colorless filtrate and the resultant residue was extracted with

Et₂O (5 mL). Evaporation of the solvent under reduced pressure yielded

the product as a light tan solid (0.24 g, 50%). ¹H NMR (300 MHz,

CD₃CN): δ 7.72-7.70 (br m, 12H), 3.06 (septet, J ) 6.6 Hz, 3H),

2.84-2.74 (m, 6H), 2.59-2.48 (m, 6H), 1.19 (d, J ) 6.6 Hz, 12H)

ppm. FTIR (KBr, cm^-1): 2277 (CtN). Anal. Calcd for C₄₉H₄₈N₄-

BCuF_24:C, 48.11; H, 3.96; N, 4.58. Found: C, 47.70; H, 4.07; N,

4.64.

[d₂₁-L^iPrCu(CH₃CN)]ClO₄. This complex was prepared in an

3

iPr

analogous fashion to [L^iPrCu(CH₃CN)]ClO₄except using d₂₁-L on

0.733 mmol scale (0.31 g, 90%). ¹H NMR (300 MHz, CD₂Cl₂): δ

2.74-2.84 (m, 6H), 2.50-2.59 (m, 6H), 2.22 (s, 3H) ppm. ¹³C{¹H}

NMR (75 MHz, CD₂Cl₂): δ 57.7 (t, J_CD) 21 Hz), 51.1, 18.9 (br), 3.4

ppm. FTIR (KBr, cm^-1): 2267 (CtN), 2227 (CsD), 2136 (CsD),

2070 (CsD), 1093 (ClO₄^-), 623 (ClO₄^-).

3

1116, 1108, 1093, 1040, 751, 714, 708, 704, 701 cm^-1

. HRCIMS

(NH₃): calcd for C₁₂H₂₈N₃(M + H) 214.2283, found 214.2277. Anal.

Calcd for C₁₂H₂₇N₃: C, 67.55; H, 12.75; N, 19.69. Found: C, 66.79;

H, 12.65; N, 19.69.

1-Benzyl-4,7-diisopropyl-1,4,7-triazacyclononane (L^{iPr Bn}). To a

2

[L^iPrCu(CO)]SbF₆. Solid [L^iPrCu(CH₃CN)]SbF₆(0.20 g, 0.33

mmol) was dissolved in THF (10 mL) in a Schlenk flask. Carbon

monoxide was then bubbled through the solution for 10 min at room

temperature. Addition of Et₂O (10 mL) to the colorless solution yielded

a white precipitate, which was recrystallized from CH₂Cl₂/Et₂O (0.18

g, 92%). ¹H NMR (300 MHz, CDCl₃): δ 3.23 (septet, J ) 6.6 Hz,

3H), 2.98-2.87 (m, 6H), 2.80-2.71 (m, 6H), 1.27 (d, J ) 6.6 Hz,

18H) ppm. ¹³C{¹H} NMR (75 MHz, CDCl₃): δ 206.8, 59.1, 51.4,

20.1 ppm. FTIR (KBr, cm^-1): 2067 (CtO), 658 (SbF₆^-). Anal. Calcd

for C₁₆H₃₃N₃OCuSbF₆: C, 32.98; H, 5.71; N, 7.21. Found: C, 33.00;

H, 5.71; N, 7.14.

solution of 1,4-diisopropyl-1,4,7-triazacyclononane (0.22 g, 1.03 mmol)

in CH₃CN (10 mL) was added benzyl chloride (0.13 g, 1.03 mmol).

After addition of Na₂CO₃(∼0.5 g) and tetrabutylammonium bromide

(10 mg), the mixture was heated at reflux under N₂for 18 h. After

cooling to room temperature, the mixture was filtered, the filter cake

washed with CHCl₃(10 mL), and the combined filtrates dried (MgSO₄).

Removal of solvent under reduced pressure afforded the crude product

(0.22 g, 70%) which was suitable for metal complex synthesis; further

purification was effected by a vacuum distillation (bp 116 °C, 0.01

Torr). ¹H NMR (CDCl₃, 300 MHz): δ 7.34-7.27 (m, 5H), 3.64 (s,

2H), 2.90-2.83 (m, 4H), 2.88 (septet, J ) 6.6 Hz, 2H), 2.59-2.62

(m, 4H), 2.60 (s, 4H), 0.94 (d, J ) 6.6 Hz, 12H) ppm. ¹³C{¹H} NMR

(CDCl₃, 75 MHz): δ 140.74, 128.9, 128.0, 126.5, 62.2, 55.3, 54.8,

52.7, 52.3, 18.3 ppm. FTIR (neat): 3084, 3064, 3027, 2964, 2932,

2795, 1945, 1873, 1813, 1681, 1607, 1496, 1456, 1381, 1361, 1307,

1262, 1173, 1156, 1116, 1096, 1030, 1001, 975, 910, 780, 731, 699

cm^-1; HREIMS: calcd for C₁₉H₃₃N₃303.2674, found 303.2683. Anal.

Calcd for C₁₉H₃₃N₃: C, 75.19; H, 10.96; N, 13.85. Found: C, 75.02;

H, 10.94; N, 13.82.

3

[L^BnCu(CH₃CN)]X (X ) ClO₄^-, SbF₆^-, and CF₃SO₃^-). To a

3

stirred solution of 0.20 g (0.50 mmol) of L^Bnin THF (2 mL) was

3

added [Cu(CH₃CN)₄]X (X ) ClO₄^-, SbF₆^-and CF₃SO₃^-) (0.50 mmol).

After stirring the mixture for 10 min, Et₂O (∼5 mL) was added to the

light yellow solution with vigorous stirring. The off-white precipitate

that formed was collected by filtration, washed with Et₂O and dried

under vacuum or under a stream of nitrogen. Recrystallization of this

solid from THF/ether or CH₂Cl₂/ether yielded the products as off-white

microcrystals. (X ) ClO₄^-). Yield: 0.28 g, 93%. ¹H NMR (CD₂-

Cl₂, 300 MHz): δ 7.36-7.32 (m, 15H), 3.85 (s, 6H), 2.75-2.65 (m,

6H), 2.59-2.49 (m, 6H), 2.36 (s, 3H) ppm. ¹³C{¹H} NMR (CD₂Cl₂,

75 MHz): δ 135.1, 131.2, 128.8, 64.3, 52.4 ppm. FTIR (KBr, cm^-1):

2270 (CtN), 1098 (ClO₄^-), 623 (ClO₄^-). FAB-MS (MNBA): m/z

[L^iPrCu(CH₃CN)]X (X ) ClO₄^-, CF₃SO₃^-, SbF₆^-). These

3

complexes were prepared as described in the literature for [L^iPrCu-

3

(CH₃CN)]PF₆,³⁷substituting the appropriate [Cu(CH₃CN)₄]X salt, and

were isolated as colorless solids in 85-95% yield. (X ) ClO₄^-). H

1

NMR (300 MHz, CD₂Cl₂): δ 3.08 (septet, J ) 6.6 Hz, 3H), 2.85-

2.77 (m, 6H), 2.59-2.51 (m, 6H), 2.24 (s, 3H), 1.21 (d, J ) 6.6 Hz,

18H) ppm. ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ 58.5, 51.2, 19.9 ppm.

FTIR (KBr, cm^-1): 2267 (CtN), 1090 (ClO₄^-), 623 (ClO₄^-). Anal.

Calcd for C₁₇H₃₆N₄CuClO₄: C, 44.44; H, 7.90; N, 12.19. Found: C,

44.29; H, 7.77; N, 11.94. (X ) CF₃SO₃^-). FTIR (KBr, cm^-1): 2267

(CtN), 1225 (CF₃SO₃^-), 1150 (CF₃SO₃^-), 1029 (CF₃SO₃^-), 637

(CF₃SO₃^-). Anal. Calcd for C₁₈H₃₆N₄CuF₃O₃S: C, 42.47; H, 7.13;

N, 11.00. Found: C, 42.86; H, 7.43; N, 10.18. (X ) SbF₆^-). FTIR

(rel intens) 561 {[L^BnCu(ClO₄) - H] , 5} 462 {[L Cu] , 100}. Anal.

Calcd for C₂₉H₃₆N₄CuClO₄: C, 57.79; H, 6.03; N, 9.30. Found: C,

57.51; H, 5.94; N, 9.28. (X ) SbF₆^-). Yield: 0.33 g, 90%. FTIR

(KBr, cm^-1): 2267 (CtN), 657 (SbF₆^-), 520. FAB-MS (MNBA) m/z

(rel intens) 699 {[L^BnCu(SbF₆) - H] , 1}, 462 {[L Cu] , 100}. Anal.

Calcd for C₂₉H₃₆N₄CuSbF₆: C, 47.15; H, 4.92; N, 7.59. Found: C,

47.56; H, 5.10; N, 7.57. (X ) CF₃SO₃^-). Yield: 0.28 g, 85%. FTIR

(KBr, cm^-1): 1252 (CF₃SO₃^-), 1161 (CF₃SO₃^-), 1030 (CF₃SO₃^-), 638

+

Bn

+

3

+

Bn

+

3

11572 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

(CF₃SO₃^-). Anal. Calcd for C₃₀H₃₆N₄CuSO₃F₃: C, 55.20; H, 5.56;

N, 8.59. Found: C, 55.51; H, 5.54; N, 8.64.

CD₂Cl₂(2:1), 75 MHz): δ 133.0, 132.3, 128.9, 59.5, 51.0 (br) ppm.

Samples of [(d₂₁-L^BnCu)₂(µ-O)₂](X)₂(X ) ClO₄^-, SbF₆^-, and CF₃SO₃^-

)

3

-

[d₂₁-L^BnCu(CH₃CN)]X (X ) ClO₄and SbF₆^-). These com-

were prepared similarly. NMR characterization was performed for the

3

-

case X ) ClO₄

.

¹H NMR (-75 °C, CD₂Cl₂, 500 MHz): δ 3.54 (br,

pounds were prepared in a manner analogous to that described for [L^Bn

3

-

s, 12H), 2.48 (br s, 12H) ppm. ¹H NMR [-75 °C, d₆-acetone/CD₂Cl₂

(2:1), 500 MHz]: δ 3.70 (br s,12H), 2.73 (br s, 12H) ppm. electrospray-

MS (CH₂Cl₂, ∼-40 °C): m/z (rel intens) (X ) ClO₄^-) 1099 {[(d₂₁-

Cu(CH₃CN)]X (X ) ClO₄and SbF₆^-) except d₂₁-L^Bn(0.2 g, 0.475

mmol) was used. (X ) ClO₄^-). Yield 0.27 g, 91%. ¹H NMR (CD₂-

Cl₂, 300 MHz): δ 2.74-2.64 (m, 6H), 2.56-2.46 (m, 6H), 2.36 (s,

3H) ppm. ¹³C{¹H} NMR (CD₂Cl₂, 75 MHz): δ 52.2 ppm. FTIR (KBr,

cm^-1): 2273 (CtN), 2247 (CsD), 2209 (CsD), 2149 (CsD), 2102

-

3

L^BnCu)₂(µ-O)₂(ClO₄)] , 55}, 525 {[d₂₁-L Cu(CH₃CN)] , 100}; (X

+

Bn

+

3

) SbF₆^-) positive ion 1235 {[(d₂₁-L^BnCu)₂(µ-O)₂(SbF₆)] , 20}, 525

+

3

{[d₂₁-L^BnCu(CH₃CN)] , 100}; negative ion 1707 {[(d₂₁-L Cu)₂(µ-

+

Bn

3

(CsD), 2067 (CsD), 2021 (CsD), 1098 (ClO₄^-), 622 (ClO₄^-). FAB-

3

O)₂(SbF₆)₃]^-, 100}.

MS (MNBA): m/z (rel intens) 582 {[d₂₁-L^BnCu(ClO₄) - H] , 5}, 483

+

3

{[d₂₁-L^BnCu] , 100}. Anal. Calcd for C₂₉H₁₅D₂₁N₄CuClO₄: C, 55.83;

+

3

Preparation of [(LCu)₂(µ-O)₂](X)₂Samples for EXAFS Experi-

ments. A solution of [L^iPrCu(CH₃CN)]ClO₄(0.10 g) in THF (10 mL)

3

H/D, 5.81; N, 8.99. Found: C, 55.51; H/D, 5.94; N, 9.08. (X )

SbF₆^-). Yield: 0.32 g, 88%. FTIR (KBr, cm^-1): 2278 (CtN), 2204

was cooled to -80 °C and oxygenated by purging with dry O₂for 1 h,

causing a color change from colorless to deep orange-brown and the

precipitation of an orange-brown solid. While stirring vigorously, cold

(-80 °C) Et₂O (30 mL) was added to complete the precipitation of the

complex. After the solid was allowed to settle, the supernatant was

removed with a frozen pipet. The dispersant boron nitride (precooled

to -80 °C) was then added to the moist solid and the mixture

homogenized with a frozen spatula. An additional amount of cold

diethyl ether (2 mL) was then added to facilitate transfer of the mixture

to the sample holder, which was frozen in a liquid N₂bath. The Et₂O

suspension of the sample mixture was taken up in a frozen glass pipette

and transferred to the frozen sample holder where the sample im-

mediately froze. After the sample holder was loaded, the sample was

covered with a window of mylar film and stored below 100 K until

subjected to X-ray analysis. A similar method was used to prepare

(CsD), 2155 (CsD), 2100 (CsD), 2064 (CsD), 655 (SbF₆^-). FAB-

+

MS (MNBA): m/z (rel intens) 720 {[d₂₁-L^BnCu(SbF₆) - H] , 3}, 483

3

{[d₂₁-L^BnCu] , 100}. Anal. Calcd for C₂₉H₁₅D₂₁N₄CuSbF₆: C, 45.83;

+

3

H/D, 4.78; N, 7.38. Found: C, 46.06; H, 4.95; N, 7.22.

[L^BnCu(CO)]ClO₄. Solid [L^BnCu(CH₃CN)]ClO₄(0.10 g, 0.17

mmol) was dissolved in CH₂Cl₂(5 mL) in a Schlenk flask. Carbon

monoxide was then bubbled through the solution for 10 min at room

temperature. A white precipitate formed while bubbling the CO; Et₂O

(10 mL) was then added to complete the precipitation. The white

microcrystalline product was collected, washed with CH₂Cl₂/Et₂O, and

then dried under vacuum (0.83 g, 85%). ¹H NMR (500 MHz, CD₂-

Cl₂): δ 7.37-7.34 (m, 15H), 3.94 (s, 6H), 2.92-2.86 (m, 6H), 2.84-

2.79 (m, 6H) ppm. FTIR (KBr, cm^-1): 2084 (CtO), 1089 (ClO₄^-),

622 (ClO₄^-). Anal. Calcd for C₂₈H₃₃N₃OCuClO₄: C, 57.03; H, 5.65;

N, 7.13. Found: C, 56.28; H, 5.56; N, 7.01.

3

samples of [(L^BnCu)₂(µ-O)₂](ClO₄)₂, which was initially prepared in

3

[L^{iPr Bn}Cu(CH₃CN)]ClO₄. This compound was prepared in a

cold acetone and precipitated with cold Et₂O.

2

[(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂. A solution of [L^iPrCu(CH₃CN)]ClO₄

(0.13 g, 0.28 mmol) in THF (10 mL) was cooled to -80 °C and

oxygenated by purging with dry O₂for 1 h; excess O₂was then removed

by purging with N₂for 30 min. The heterogenous mixture was then

allowed to warm to room temperature, causing a color change from

orange-brown to blue-green and the precipitation of a green solid. The

solid was redissolved in CH₃CN and the solution treated with excess

NaBPh₄(0.50 g). The resultant mixture was stored at -20 °C

overnight, causing the deposition of blue-green microcrystals (0.10 g,

30%). Recrystallization of this solid from THF/CH₃CN at -20 °C

manner analogous to that described for [L^BnCu(CH₃CN)]ClO₄^-except

2

3

L^{iPr Bn}(0.2 g, 0.66 mmol) was used instead of L^Bn. Yield: 0.31 g,

2

3

1

91%. H NMR (acetone-d₆, 300 MHz): δ 7.64-7.61 (m, 2H), 7.45-

7.35 (m, 3H), 4.01 (s, 2H), 3.06 (septet, J ) 6.6 Hz, 2H); 3.05-2.95

(m, 2H), 2.90-2.50 (overlapping multiplets, 10H), 2.43 (br s, 3H), 1.20

(d, J ) 6.6 Hz, 6H), 1.17 (d, J ) 6.6 Hz, 6H) ppm. ¹³C{¹H} NMR

(acetone-d₆, 75 MHz): δ 132.2, 129.0, 64.8, 58.5, 54.6, 51.8, 49.8,

20.3, 19.3 ppm. FTIR (KBr, cm^-1): 2263 (CtN), 1090 (ClO₄^-), 622

(ClO₄^-). Anal. Calcd for C₂₁H₃₆N₄CuClO₄: C, 49.70; H, 7.15; N,

11.04. Found: C, 50.05; H, 7.18; N, 10.93.

produced a small number of blue crystals of [(L^{iPr H}Cu)₂(µ-OH)₂]-

2

[L^{iPr Bn}Cu(CO)]ClO₄. This compound was prepared in a manner

2

analogous to that described for [L^iPrCu(CO)]ClO₄and recrystallized

(BPh₄)₂‚2THF suitable for crystallographic analysis. EPR (1:1 CH₃-

CN/toluene, 9.4 GHz, 77 K): silent. FTIR (KBr, cm^-1): 3416 (br,

O-H), 3275 (N-H), 710 (BPh₄^-). Electrospray-MS (CH₃CN) m/z 908

3

from THF/Et₂O. Yield: 0.088 g, 90%. ¹H NMR (500 MHz, CDCl₃)

δ 7.40-7.37 (m, 4H), 4.00 (s, 2H), 3.10 (septet, J ) 6.5 Hz, 2H),

3.05-2.98 (m, 4H), 2.89-2.79 (m, 4H), 2.76-2.67 (m, 4H), 1.16-

1.13 (two overlapping doublets, J ) 6.6 Hz, 12H) ppm. ¹³C{¹H} NMR

(125 MHz, CDCl₃): δ 134.8, 131.4, 129.1, 128.8, 65.2, 59.0, 54.7,

51.8, 50.0, 20.5, 19.8 ppm. FTIR (KBr, cm^-1): 2069 (CtO), 1091

(ClO₄^-), 622 (ClO₄^-). Anal. Calcd for C₂₂H₃₆N₃OCuClO₄: C, 48.67;

H, 6.74; N, 8.52. Found: C, 48.13; H, 6.62; N, 8.30.

{[(L^{iPr H}Cu)₂(OH)₂(BPh₄)]⁺, 100}.

2

Bn

[(L^BnCu)₂(µ-OH)₂](ClO₄)₂‚DMF‚H₂O. A solution of [(L Cu)₂-

(µ-O)₂](ClO₄)₂in CH₂Cl₂at -80 °C was prepared as described above

[0.10 g, 0.17 mmol scale of Cu(I) precursor]. After removing excess

oxygen by purging nitrogen at -80 °C for 15 min, the solution was

warmed to room temperature. The orange-brown solution turned blue-

green at room temperature, whereupon a blue-green solid precipitated,

leaving a green filtrate. The blue-green precipitate was then recrystal-

lized by diffusing methanol into DMF solution at room temperature.

After ∼48 h blue crystals were separated, washed with methanol and

air dried (0.06 g, 60%). FTIR (KBr, cm^-1): 3574 (OH), 3444 (br,

H₂O), 1662 (CdO, DMF), 1100 (ClO₄^-), 623 (ClO₄^-). UV-vis

(DMF): (λ_max, nm (ꢀ, M^-1cm^-1)): 622 (190), 292 (3800). EPR (DMF,

3

[(LCu)₂(µ-O)₂](X)₂(L ) L^Bn, L , and L^{iPr Bn}; X ) ClO₄^-, SbF₆

,

iPr

-

3

2

and CF₃SO₃^-). Solid [LCu(CH₃CN)]X (∼0.20 g) was placed in a

Schlenk flask and cooled to -80 °C. Dry CH₂Cl₂(10 mL) (L^Bnor

3

L^{iPr Bn}) or THF (L^iPr) was then injected into the Schlenk flask and dry

dioxygen was then bubbled through the cold solution for 15 min. The

initially colorless solution became deep orange-brown. Spectroscopic

characterization of these compounds were performed at a low temper-

2

3

1

9.4 GHz, 77 K): silent. electrospray-MS (DMF): m/z positive ion

ature (∼-80 °C) immediately after generation. Low-temperature H

+

1059 {[(L^BnCu)₂(OH)₂(ClO₄)] , 28}, 521 (100); negative ion 1257

and ¹³C NMR spectra were obtained at -75 °C using the Varian VXR-

3

{[(L^BnCu)₂(OH)₂(ClO₄)₃] , 38}, 721 (100). Anal. Calcd for

-

500 spectrometer as follows. In the glovebox, ∼ 0.03 g of [L^BnCu-

3

C₅₇H₇₇N₇Cu₂Cl₂O₁₁: C, 54.84; H, 6.22; N, 7.86. Found: C, 54.68; H,

6.06; N, 7.66.

(CH₃CN)]ClO₄was dissolved in ∼0.5 mL CD₂Cl₂or in 0.6 mL of

d₆-acetone/CD₂Cl₂(2:1) solvent mixture in a NMR tube capped with

a septum. After removal from the glovebox, the tube was immersed

into a cold bath maintained at -80 °C and O₂was then bubbled in

through a needle for 15 min. After oxygenation, the dark brown mixture

was purged with N₂to remove excess O₂. The NMR tube was then

transferred to the precooled (-75 °C) probe of the spectrometer. EPR

[(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](CF₃SO₃)₂‚2(CH₃)₂CO. Following a

3

2

similar methodology as described above, when an orange-brown

solution of [(L^BnCu)₂(µ-O)₂](CF₃SO₃)₂warmed slowly from -80 °C

3

to room temperature under a positive dinitrogen flow, a blue-green

solution resulted. A blue-green solid was isolated from this mixture

by the addition of excess Et₂O. Recrystallization of this solid from

acetone/Et₂O at room temperature yielded a small number of blue-

green crystals suitable for X-ray crystallography. FTIR (KBr, cm^-1):

3573 (O-H), 3292 (N-H), 1253 (CF₃SO₃^-), 1160 (CF₃SO₃^-), 1030

(CF₃SO₃^-), 639 (CF₃SO₃^-). EPR (1:1 CH₃CN/toluene, 9.4 GHz, 77

K): silent.

(CH₂Cl₂, 9.4 GHz, 77 K): silent. (L ) L^Bn). ¹H NMR (-75 °C, CD₂-

3

Cl₂, 500 MHz) δ 7.44-7.36 (m, 30H), 4.57 (s, 12H), 3.54 (br, s,12H),

2.50 (br s, 12H) ppm. ¹³C{¹H} NMR (-75 °C, CD₂Cl₂, 75 MHz): δ

134.6, 132.1, 131.0, 128.9, 59.7 ppm. ¹H NMR [-75 °C, d₆-acetone/

CD₂Cl₂(2:1), 500 MHz]: δ 7.54-7.45 (m, 30H), 4.87 (s, 12H), 3.71

(br s,12H), 2.73 (br s, 12H) ppm. ¹³C{¹H} NMR (-75 °C, d₆-acetone/

Characterization of Bis(µ-oxo)dicopper Complexes

J. Am. Chem. Soc., Vol. 118, No. 46, 1996 11573

[(L^{iPr Bn}Cu)₂(µ-OH)₂](CF₃SO₃)₂. Following a similar methodology

collection. Data collection was performed as described above. Im-

portant crystallographic information is given in Table 2. The space

group P2₁/n was determined on the basis of systematic absences and

intensity statistics. A successful direct-methods solution was calculated

which provided most non-hydrogen atoms from the E-map. Several

full-matrix least squares/difference Fourier cycles were performed which

located the remainder of the non-hydrogen atoms. All non-hydrogen

atoms were refined with anisotropic displacement parameters and all

hydrogen atoms were placed in ideal positions and were refined as

riding atoms with individual isotropic displacement parameters.

2

as described above, when an orange-brown solution of [(L^{iPr Bn}Cu)₂-

2

(µ-O)₂](CF₃SO₃)₂warmed slowly from -80 °C to room temperature

under a positive dinitrogen flow, a blue-green solution with a blue-

green precipitate resulted. The blue-green precipitate was collected

and recrystallized by diffusing Et₂O into a DMF solution at room

temperature (0.043 g, 45%). FTIR (KBr, cm^-1): 3606 (OH), 1265

(CF₃SO₃^-), 1160 (CF₃SO₃^-), 1031 (CF₃SO₃^-), 636 (CF₃SO₃^-). EPR

(DMF, 9.4 GHz, 77 K): silent. UV-vis (DMF) (λ_max, nm (ꢀ, M^-1

cm^-1)): 360 (sh, 2200), 650 (200). Anal. Calcd for C₄₀H₆₈N₆-

Cu₂F₆S₂O₈: C, 45.10; H, 6.44; N, 7.89. Found: C, 44.29; H, 6.33; N,

7.99.

-

One of the two CF₃SO₃counterions and one of the (CH₃)₂CO

molecules exhibit significant translational and torsional rigid-body

motion. Modeling the former as a disordered group was considered,

but not attempted. Although a TLS rigid body refinement of CF₃SO₃

would yield more realistic results than a disordered refinement, it is

doubtful that a significantly better residual would result. The isotropic

displacement parameters of the hydrogens on the bridging hydroxyls

and on N7′ were tied to 1.2 times that of their host. Drawings of the

cation appear in Figures 5 and 8, and selected bond lengths and angles

are presented in Table 3. Full details of the structure determination,

including tables of bond lengths and angles, atomic positional

parameters, and final thermal parameters for non-hydrogen atoms, are

given in the supporting information.

Crossover Experiment. Equimolar amounts of the bis(µ-hydroxo)-

Bn

dicopper(II) compounds [(L^BnCu)₂(µ-OH)₂](SbF₆)₂and [(d₂₁-L Cu)₂-

(µ-OD)₂](SbF₆)₂in CH₂Cl₂were mixed at room temperature. After

stirring for 30 min this mixture was injected to an electrospray-MS

3

detector at room temperature. Electrospray MS (CH₂Cl₂): m/z (rel

+

intens) positive ion 1197 {[(L^BnCu)₂(µ-OH)₂(SbF₆)] , 30}, 1218

3

Bn

+

Bn

3

{[(d₂₁-L^BnCuOD)(L CuOH)(SbF₆)] , 60}, 1237 {[(d₂₁-L Cu)₂(µ-

3

-

OD)₂(SbF₆)]⁺, 35}; negative ion 1667 {[(L^BnCu)₂(µ-OH)₂(SbF₆)₃] ,

3

Bn

-

52}, 1689 {[(d₂₁-L^BnCuOD)(L CuOH)(SbF₆)₃] , 100}, 1709

3

-

{[(d₂₁-L^BnCu)₂(µ-OD)₂(SbF₆)₃] , 55}.

3

X-ray Crystallography. [(d₂₁-L^BnCu)₂(µ-O)₂](SbF₆)₂‚7(CH₃)₂CO‚

3

[(L^{iPr H}Cu)₂(µ-OH)₂](BPh₄)₂‚2THF. A blue plate crystal of the

2

2CH₃CN. Crystals were grown by storing a solution of the complex

prepared by oxygenation of [d₂₁-L^BnCu(CH₃CN)]ClO₄in 1:1 CH₂Cl₂/

3

compound was attached to a glass fiber with heavy-weight oil and

mounted on the Siemens SMART system for data collection at 173(2)

K. Data collection was performed as described above. Important

crystallographic information is given in Table 2. The space group Cc

was determined based on systematic absences and intensity statistics,

which suggested a noncentrosymmetric space group. A successful

direct-methods solution was calculated which provided all non-hydrogen

atoms from the E-map. All non-hydrogen atoms were refined with

anisotropic displacement parameters. All hydrogen atoms were placed

in idealized positions and refined as riding atoms with individual

isotropic displacement parameters.

(CH₃)₂CO at -80 °C. A dark brown rectangular-shaped crystal was

mounted to the side of a glass capillary at low temperature (<-40 °C)

using a locally designed cold stage and then transferred to the

goniometer of a Siemens SMART system, with the cryogenic stream

set to 123(2) K for data collection. An initial set of cell constants was

calculated from 354 reflections harvested from three sets of 30 frames.

Final cell constants were calculated from a set of 6155 strong reflections

from the data collection. A hemisphere collection involves the survey

of a randomly oriented region of reciprocal space to the extent of 1.3

hemispheres to a resolution of 0.84 Å. Three major swaths of frames

were collected with 0.30° steps in ω. A semiempirical absorption

correction afforded minimum and maximum transmission factors of

0.624 and 0.871, respectively. See Table 2 for additional crystal and

refinement information.

The space groups Cc or C2/c were determined based on systematic

absences. Intensity statistics favored neither one. A successful direct-

methods solution was calculated in Cc which provided most non-

hydrogen atoms from the E-map. The structure was transformed into

the centrosymmetric space group C2/c after checking with MISSYM.⁷⁶

Several full-matrix least squares/difference Fourier cycles were per-

formed which located the remainder of the non-hydrogen atoms. All

non-hydrogen atoms were refined with anisotropic displacement

parameters. All hydrogen or deuterium atoms were placed in idealized

positions and were refined as riding atoms with relative (1.2 times host

for methylene, 1.5 times host for methyl) isotropic displacement

parameters. All calculations were performed using SGI INDY R4400-

SC or Pentium computers with the SHELXTL V5.0 program suite.⁷⁷

The asymmetric unit contains one half of the dimeric cation as well

as CH₃CN and 3.5 (CH₃)₂CO solvent molecules and a rotationally

The unit cell is composed of four dicationic dimers, eight THF

solvent molecules, and eight tetraphenylborate anions. The structure

solved easily in Cc, which presented a model that appeared to be

centrosymmetric. The dication, THF molecules, and tetraphenylborate

are labeled as though pairs of related atoms pass through a point

equidistant from both copper atoms. Attempts to transform one-half

of the contents to the centrosymmetric equivalent in space group C2/c

failed. Use of the MISSYM⁷⁶also failed to find any additional

symmetry. Therefore, we believe the model presented here to be

correct. This specimen desolvated very rapidly to yield crystals with

surface fractures. Additionally, the model was refined as a racemic

twin in a 66:34 ratio. A thermal ellipsoid drawing of the cation appears

in Figure 6, and selected bond lengths and angles are presented in Table

3. Full details of the structure determination, including tables of bond

lengths and angles, atomic positional parameters, and final thermal

parameters for non-hydrogen atoms, are given in the Supporting

Information.

[(L^{iPr Bn}Cu)₂(µ-OH)₂](CF₃SO₃)₂. A blue prism crystal of the

2

compound was mounted on a glass fiber and transferred to an Enraf-

Nonius CAD-4 diffractometer with graphite monochromated Mo KR

radiation (λ ) 0.710 73 Å). Relevant crystallographic information is

given in Table 2. Cell constants and an orientation matrix for data

collection were obtained from a least squares refinement of the setting

angles for 25 carefully centered reflections in the range 18.12° < 2θ

<32.05°. The data were collected at a temperature of 24 ( 1 °C using

the ω-2θ scan technique to a maximum 2θ value of 50.1°. An empirical

absorption correction was applied (using ψ-scans of three reflections)

which resulted in transmission factors ranging from 0.90 to 1.00. The

data were corrected for Lorentz and polarization effects; no decay

correction was applied.

The structure was solved by direct methods using the TEXSAN

software package.⁷⁸Non-hydrogen atoms were refined anisotropically,

and the hydrogen atoms were placed at calculated positions. The

maximum and minimum peaks on the final difference Fourier map

corresponded to 0.56 and -0.66 e^-/Å³, respectively. An ORTEP

drawing of the cation appears in Figure 7, and selected bond lengths

-

disordered SbF₆counterion (disordered 7:1). Copious distance

restraints were placed on the solvent molecules, the SbF₆^-counterions,

and the phenyl groups of the cation to force these units to conform

their respective average substructures. Additionally, the anisotropic

displacement parameters for these were restrained to better fit rigid

body motion. The residual electron density in the final difference

Fourier map was located close to the disordered SbF₆^-. Drawings of

the cation appear in Figures 4 and 8, and selected bond lengths and

angles are presented in Table 3. Full details of the structure determi-

nation, including tables of bond lengths and angles, atomic positional

parameters, and final thermal parameters for non-hydrogen atoms, are

given in the Supporting Information.

[(L^BnCu)(L^{Bn H}Cu)(µ-OH)₂](CF₃SO₃)₂‚2(CH₃)₂CO. A blue-green

plate-like crystal of the compound was attached to a glass fiber with

heavy-weight oil and transferred to the goniometer of the Siemens

SMART system, with cryogenic stream set to 173(2) K for data

3

2

(76) PLATON: Spek, A. L. Acta Crystallogr. 1990, A46, C34.

(77) SHELXTL V5.0, Siemens Energy & Automation, Inc., Madison,

WI 53719-1173.

(78) TEXSAN: Single Crystal Structure Analysis Software, Version 5.0

(1989), Molecular Structure Corporation, The Woodlands, TX 77381.

11574 J. Am. Chem. Soc., Vol. 118, No. 46, 1996

Mahapatra et al.

and angles are presented in Table 3. Full details of the structure

determination, including tables of bond lengths and angles, atomic

positional parameters, and final thermal parameters for non-hydrogen

atoms, are given in the supporting information.

structures compare well with experimentally derived structures, we

consider it a reasonable approximation to employ these structures for

higher level calculations.

The HF and DFT calculations were carried out using the Gaussian94

suite of programs;⁸⁷the MCSCF and CASPT2N calculations were

carried out using the MOLCAS suite of programs.⁸⁸

Computational Methods. Anti (C_2h) and eclipsed (C_2V) conformers

of {[(NH₃)₃Cu]₂(µ-O)₂}²⁺were fully optimized at the restricted

Hartree-Fock (RHF) level of theory using the STO-3G basis set,^79,80

as was a transition state structure that interconnects them via a half-

turnstile-like rotation. Analytic frequency calculations verified the

nature of these stationary points; frequencies when reported have been

scaled by a factor of 0.89.⁸¹Single-point energies for these three

structures were calculated at a variety of levels of theory. Using the

STO-3G* basis set, density functional (DFT) calculations were carried

out using the local XR functional⁸²and also the gradient corrected

exchange and correlation functionals of Becke⁸³and Lee-Yang-Parr,⁸⁴

respectively (BLYP). To test the stability of the restricted XR

calculation on the eclipsed isomer, a broken-symmetry initial density

was constructed by zeroing out all atomic orbital (AO) coefficients

associated with one copper atom in the R Kohn-Sham (KS) highest

occupied molecular orbital (HOMO), and all AO coefficients associated

with the other copper atom in the â KS HOMO. This calculation

smoothly converged to the restricted solution.

Additional single-point calculations with the RHF/STO-3G geom-

etries were carried out at the multiconfiguration self-consistent-field

(MCSCF) level using the double-ú polarized basis set of Schafer et

al.⁸⁵Nondynamic correlation effects were partially accounted for with

a 2-electron/2-orbital (2,2) active space, and additional correlation

effects were accounted for using multireference second-order perturba-

tion theory (CASPT2N).⁸⁶For the anti conformer, the active space is

formed from molecular orbitals 12a_uand 12b_g; for the eclipsed

conformer, the active space is formed from molecular orbitals 13b₁

and 10a₂. Geometry optimization at the MCSCF level of theory led to

dissociation, suggesting that in the gas phase, the dications are either

intrinsically unstable or the calculations require the inclusion of

additional correlation to stabilize the (µ-O)₂structures. We investigated

active spaces up to (8,8) in size, but found dissociation to proceed

without barrier in every case; additional electron correlation effects

calculated at the CASPT2N level strongly stabilize the (µ-O)₂structures

relative to the peroxo, but analytic derivatives for this method are not

available, and that precludes efficiently accounting for those effects in

geometry optimization. Apparently the RHF/STO-3G level serendipi-

tously overstabilizes the complex. Insofar as the RHF/STO-3G

Acknowledgment. We thank the National Institutes of

Health (Grant GM47365 to W.B.T., Grant GM33162 to L.Q.,

and a predoctoral traineeship, Grant GM07323, to E.C.W.), the

Searle Scholars Program/Chicago Community Trust (W.B.T.),

the National Science Foundation (NYI award to W.B.T., Grant

CHE-9525819 to C.J.C., 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

training grants to J.A.H. and S.M. through the Center for Metals

in Biocatalysis), and the Alfred P. Sloan and Camille & Henry

Dreyfus Foundations (fellowships to W.B.T) for providing

funding in support of this research. We are grateful for high-

performance vector and parallel computing resources made

available by the Minnesota Supercomputer Institute and the

University of Minnesota-IBM Shared Research Project, respec-

tively. EXAFS data were collected at the National Synchrotron

Light Source (Brookhaven National Laboratory, Beamline

X-9A) and the Cornell High Energy Synchrotron Source

(CHESS, station C-2). We also thank Professor Kenton Rodgers

(North Dakota State University) for assistance with the Raman

polarization experiments, Bradley Smith for technical assistance

with orbital visualization, Professor Bernt Krebs (Mu¨nster, FRG)

for supplying the coordinates for the X-ray structure described

in ref 3d, and Professor Holden H. Thorp (University of North

Carolina) for helpful discussions.

Supporting Information Available: 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 (66 pages). See any current masthead

page for ordering and Internet access instructions.

JA962305C

(79) Hehre, W. H.; Stewart, R. F.; Pople, J. A. J. Chem. Phys. 1969, 51,

2657.

(80) Pietro, W. J.; Hehre, W. J. J. Comput. Chem. 1983, 4, 241.

(81) Pople, J. A.; Scott, A. P.; Wong, M. W.; Radom, L. Isr. J. Chem.

1993, 33, 345.

(82) Slater, J. C. Phys. ReV. 1951, 81, 385.

(83) Becke, A. D. Phys. ReV. A 1988, 38, 3098.

(84) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.

(85) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.

(86) (a) Andersson, K.; Malmqvist, P.-A° .; Roos, B. O.; Sadlej, A. J.;

Wolinski, K. J. Phys. Chem. 1990, 94, 5483. (b) Andersson, K.; Malmqvist,

P.-A° .; Roos, B. O. J. Chem. Phys. 1992, 96, 1218. (c) Andersson, K.; Roos,

B. O. Int. J. Quant. Chem. 1993, 45, 591.

(87) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;

Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.

A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,

V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;

Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;

Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;

Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-

Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94 ReVC.1; Gaussian

Inc.: Pittsburgh, PA, 1995.

(88) Andersson, K.; Blomberg, M. R. A.; Fu¨lscher, M. P.; Karlstro¨m,

G.; Kello¨, V.; Lindh, R.; Malmqvist, P.-A° .; Noga, J.; Olsen, J.; Roos, B.

O.; Sadlej, A. J.; Siegbahn, P. E. M.; Urban, M.; Widmark, P.-O. MOLCAS-

3; University of Lund: Sweden, 1994.

Article Doi

Structural, spectroscopic, and theoretical characterization of bis(μ-oxo)dicopper complexes, novel intermediates in copper-mediated dioxygen activation

DOI: 10.1021/ja962305c

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja962305c

Full text of DOI:10.1021/ja962305c

Products guided by the article

R&D Labs maybe for 184219-48-5

Relevant to this article

Hot Product