Bis(μ-oxo)dicopper(III) complexes of a homologous series of simple peralkylated 1,2-diamines: Steric modulation of structure, stability, and reactivity

Inorg. Chem. 2005, 44, 7345−7364

Bis(µ-oxo)dicopper(III) Complexes of a Homologous Series of Simple

Peralkylated 1,2-Diamines: Steric Modulation of Structure, Stability, and

Reactivity

Adam P. Cole, Viswanath Mahadevan, Liviu M. Mirica, Xavier Ottenwaelder, and T. Daniel P. Stack*

Contribution from the Department of Chemistry, Stanford UniVersity, Stanford, California 94305

Received March 3, 2005

We have synthesized and characterized bis(

peralkylated trans-(1R,2R)-cyclohexanediamine (CD) ligands, self-assembled from the corresponding

[LCu(MeCN)]CF₃SO₃species 1a 4a and O₂at 193 K in aprotic media; additional Os based on peralkylated

ethylenediamine and tridentate polyazacyclononane ligands were synthesized analogously for comparative purposes

µ-oxo)dicopper(III) dimers 1b−4b (Os) based on a core family of

−

(5b−7b and 8b−

9b, respectively). Trigonal-planar [LCu(MeCN)]¹⁺species are proposed as the active O precursors.

The 3-coordinate Cu(I) complexes [(L^TE)Cu(MeCN)]CF₃SO₃(4a) and [(L^TB)Cu(MeCN)]CF₃SO₃(10a) were structurally

−

characterized; the apparent O₂inertness of 10a correlates with the steric demands of its four benzyl substituents.

+

The rate of O formation, a multistep process that likely proceeds via associative formation of a 1:1 [LCu(O₂)]¹

intermediate, exhibits significant dependence upon ligand sterics and solvent: oxygenation of 4a

reacting O precursor of the CD series is first-order with respect to [4a] and proceeds at least 300 times faster in

tetrahydrofuran than in CH₂Cl₂. The EPR, UV vis, and resonance Raman spectra of 1b 9b are all characteristic

of the diamagnetic bis( -oxo)dicopper(III) core. The intense ligand-to-metal charge transfer absorption maxima of

sthe slowest-

s

−

µ

CD-based Os are red-shifted proportionally with increasing peripheral ligand bulk, an effect ascribed to a slight

distortion of the [Cu₂O₂] rhomb. The well-ordered crystal structure of [(L^ME)₂Cu₂(

µ

-O)₂](CF₃SO₃)₂

‚

4CH₂Cl₂([3b

‚

+

4CH₂Cl₂]) features the most metrically compact [Cu₂O₂]²core among structurally characterized Os (av Cu

−O

1.802(7) Å; Cu‚‚‚Cu 2.744(1) Å) and exemplifies the minimal square-planar ligation environment necessary for

stabilization of Cu(III). The reported Os are mild oxidants with moderate reactivity toward coordinating substrates,

readily oxidizing thiols, certain activated alkoxides, and electron-rich phenols in a net 2e^-, 2H⁺process. In the

absence of substrates, 1b

−

9b undergo thermally induced autolysis with concomitant degradation of the polyamine

D

ligands. Ligand product distribution and primary kinetic isotope effects (k_obs^H/k_obs

≈

8, 1b/d₂₄-1b, 293 K) support

a unimolecular mechanism involving rate-determining C

−

H bond cleavage at accessible ligand N-alkyl substituents.

Decomposition half-lives span almost 3 orders of magnitude at 293 K, ranging from

∼

2 s for 4b to almost 30 min

for d₂₄-1b, the most thermally robust dicationic O yet reported. Dealkylation is highly selective where ligand rigidity

constrains accessibility; in 3b, the ethyl groups are attacked preferentially. The observed relative thermal stabilities

and dealkylation selectivities of 1b

ligand backbone rigidity, and ligand denticity/axial donor strength. Among the peralkylated amines surveyed, bidentate

ligands with oxidatively robust NC^R

H bonds provide optimal stabilization for Os. Fortuitously, the least sterically

− −H bond dissociation energies, statistical factors,

9b are correlated with NC^R

−

demanding N-alkyl substituent (methyl) gives rise to the most thermally stable and most physically accessible O

core, retaining the potential for exogenous substrate reactivity.

Introduction

relevance to metallobiology and synthetic oxidative catalysis

that has attracted strong interest over the past two decades.^1-12

Recent work has probed the relationship between the

structure and function of natural Cu-O₂species, as well as

synthetic model systems, further contributing to a wealth of

The bis(µ-oxo)dicopper(III) dimer is one of several recur-

ring structural motifs in Cu-O₂chemistry, an area of

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

stanford.edu.

10.1021/ic050331i CCC: $30.25

Published on Web 09/21/2005

Inorganic Chemistry, Vol. 44, No. 21, 2005 7345

Cole et al.

structural, spectroscopic, and mechanistic data concerning

their formation and reactivity.^1,2Synthetic binuclear Cu-

O₂systems have frequently employed tridentate aromatic or

aliphatic nitrogen ligands in emulation of natural biochemical

environments such as that of hemocyanin (Hc), in which each

Cu is ligated by three histidine residues in both the

oxygenated and the reduced forms.¹³Yet, in oxyHc and all

other structurally characterized Cu-O₂complexes that

comprise such tridentate amine ligands, the third, axially

disposed N donor is but weakly bound to Cu.^13-20A marked

preference for square-planar coordination dominates the

chemistry of trivalent Cu and forms the basis for an emerging

trend in which sterically demanding bidentate nitrogen

ligands are used to stabilize bis(µ-oxo)dicopper(III) (O)^21-33

and µ-η²:η²-peroxodicopper(II) (P)³⁴dimers, as well as

interconverting mixtures thereof.³⁵

Simple aliphatic diamines constitute a vast and versatile

pool of inexpensive, easily modified bidentate ligands. Yet,

despite their well-known empirical use in conjunction with

Cu salts and atmospheric O₂for various “one-pot” oxidative

transformations involving unknown active species,^36,37they

have been largely ignored in the context of designing ligands

to support Cu-O₂intermediates. In 1984, Thompson did

report a putative Cu(II)-peroxo species, obtained by reacting

[(L^TEED)Cu]¹⁺(L^TEED) N,N,N′,N′-tetraethyl-1,2-ethanedi-

amine) with O₂at 193 K in wet methanol (MeOH).^38,39The

adduct is stable in the solid state at room temperature but

readily decomposes in solution via hydroxylation of the

ligand. Encouraged by this result and by the general synthetic

flexibility afforded by N-peralkylated diamine ligands (PDLs),

we systematically explored the low-temperature O₂reactivity

of a series of Cu(I)-PDL complexes in aprotic solvents with

weakly coordinating counteranions.⁴⁰In several instances, a

new Cu-O₂species with spectroscopic features distinct from

those of any Cu-O₂complex then known was observed.^41,42

One set of such features is now diagnostic for Os, and

numerous spectroscopically and/or structurally characterized

examples have since been reported.^{8,16,17,21-33,35,43-53}

(1) Mirica, L. M.; Ottenwaelder, X.; Stack, T. D. P. Chem. ReV. 2004,

104, 1013-1045.

(2) Lewis, E. A.; Tolman, W. B. Chem. ReV. 2004, 104, 1047-1076.

(3) Karlin, K. D.; Tyekla´r, Z., Eds. Bioinorganic Chemistry of Copper;

Chapman and Hall: New York, 1993.

(4) Barton, D. H. R.; Martell, A. E.; Sawyer, D. T. The ActiVation of

Dioxygen and Homogeneous Catalysis; Plenum Press: New York,

1993.

(5) Kitajima, N.; Moro-Oka, Y. Chem. ReV. 1994, 94, 737-757.

(6) Fox, S.; Karlin, K. D. In ActiVe Oxygen in Biochemistry; Valentine,

J. S., Foote, C. S., Greenberg, A., Liebman, J. F., Eds.; Blackie

Academic and Professional: London, 1995; pp 188-231.

(7) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem. ReV. 1996,

96, 2563-2606.

(8) Tolman, W. B. Acc. Chem. Res. 1997, 30, 227-237.

(9) Blackman, A. G.; Tolman, W. B. Struct. Bond. 2000, 97, 179-211.

(10) Schindler, S. Eur. J. Inorg. Chem. 2000, 2311-2326.

(11) Mahadevan, V.; Klein Gebbink, R. J. M.; Stack, T. D. P. Curr. Opin.

Chem. Biol. 2000, 4, 228-234.

(30) Taki, M.; Teramae, S.; Nagatomo, S.; Tachi, Y.; Kitagawa, T.; Itoh,

S.; Fukuzumi, S. J. Am. Chem. Soc. 2002, 124, 6367-6377.

(31) Spencer, D. J. E.; Aboelella, N. W.; Reynolds, A. M.; Holland, P. L.;

Tolman, W. B. J. Am. Chem. Soc. 2002, 124, 2108-2109.

(32) Aboelella, N. W.; Lewis, E. A.; Reynolds, A. M.; Brennessel, W. W.;

Cramer, C. J.; Tolman, W. B. J. Am. Chem. Soc. 2002, 124, 10660-

10661.

(12) Solomon, E. I.; Chen, P.; Metz, M.; Lee, S.-K.; Palmer, A. E. Angew.

Chem., Int. Ed. 2001, 40, 4570-4590.

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

tura, J.; Hol, W. G. J. Proteins: Struct., Funct., Genet. 1994, 19, 302-

309.

(33) Spencer, D. J. E.; Reynolds, A. M.; Holland, P. L.; Jazdzewski, B.

A.; Duboc-Toia, C.; Le Pape, L.; Yokota, S.; Tachi, Y.; Itoh, S.;

Tolman, W. B. Inorg. Chem. 2002, 41, 6307-6321.

(14) Kitajima, N.; Fujisawa, K.; Moro-Oka, Y.; Toriumi, K. J. Am. Chem.

Soc. 1989, 111, 8975-8976.

(34) Mirica, L. M.; Vance, M.; Rudd Jackson, D.; Hedman, B.; Hodgson,

K. O.; Solomon, E. I.; Stack, T. D. P. J. Am. Chem. Soc. 2002, 124,

9332-9333.

(15) Fujisawa, K.; Tanaka, M.; Moro-Oka, Y.; Kitajima, N. J. Am. Chem.

Soc. 1994, 116, 12079-12080.

(16) Mahapatra, S.; Halfen, J. A.; Wilkinson, E. C.; Pan, G. F.; Wang, X.

D.; Young, V. G.; Cramer, C. J.; Que, L., Jr.; Tolman, W. B. J. Am.

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

(35) Mahadevan, V.; Henson, M. J.; Solomon, E. I.; Stack, T. D. P. J. Am.

Chem. Soc. 2000, 122, 10249-10250.

(36) Aycock, D.; Abolins, V.; White, D. M. Encyclopedia of Polymer

Science and Engineering, 2nd ed.; John Wiley and Sons: New York,

1986; Vol. 13.

(17) Mahapatra, S.; Young, V. G.; Kaderli, S.; Zuberbu¨hler, A. D.; Tolman,

W. B. Angew. Chem., Int. Ed. Engl. 1997, 36, 130-133.

(18) Pidcock, E.; Debeer, S.; Obias, H. V.; Hedman, B.; Hodgson, K. O.;

Karlin, K. D.; Solomon, E. I. J. Am. Chem. Soc. 1999, 121, 1870-

1878.

(37) Oyaizu, K.; Saito, K.; Tsuchida, E. Chem. Lett. 2000, 1318-1319.

(38) Thompson, J. S. J. Am. Chem. Soc. 1984, 106, 8308-8309.

(39) Thompson, J. S. In Biological and Inorganic Copper Chemistry; Karlin,

K. D., Zubieta, J., Eds.; Adenine Press: Guilderland, NY, 1986.

(40) Stack, T. D. P. Dalton Trans. 2003, 10, 1881-1889.

(41) Mahapatra, S.; Halfen, J.; Wilkinson, E.; Pan, G.; Cramer, C. J.; Que,

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

(42) Cole, A. P.; Root, D. E.; Mukherjee, P.; Solomon, E. I.; Stack, T. D.

P. Science 1996, 273, 1848-1850.

(43) Mahapatra, S.; Kadereli, S.; Llobet, A.; Neuhold, Y. M.; Palanche,

T.; Halfen, J. A.; Young, V. G.; Kaden, T. A.; Que, L.; Zuberbu¨hler,

A. D.; Tolman, W. B. Inorg. Chem. 1997, 36, 6343-6356.

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

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

271, 1397-1400.

(45) Mahapatra, S.; Halfen, J. A.; Tolman, W. B. J. Am. Chem. Soc. 1996,

118, 11575-11586.

(46) Lam, B. M. T.; Halfen, J. A.; Young, V. G.; Hagadorn, J. R.; Holland,

P. L.; Lledo´s, A.; Cucurull-Sa´nchez, L.; Novoa, J. J.; Alvarez, S.;

Tolman, W. B. Inorg. Chem. 2000, 39, 4059-4072.

(47) Liang, H.-C.; Zhang, C. X.; Henson, M. J.; Sommer, R. D.; Hatwell,

K. R.; Kaderli, S.; Zuberbu¨hler, A. D.; Rheingold, A. L.; Solomon,

E. I.; Karlin, K. D. J. Am. Chem. Soc. 2002, 124, 4170-4171.

(48) Uozumi, K.; Hayashi, Y.; Suzuki, M.; Uehara, A. Chem. Lett. 1993,

963-966.

(49) Hayashi, H.; Fujinami, S.; Nagatomo, S.; Ogo, S.; Suzuki, M.; Uehara,

A.; Watanabe, Y.; Kitagawa, T. J. Am. Chem. Soc. 2000, 122, 2124-

2125.

(19) Kodera, M.; Katayama, K.; Tachi, Y.; Kano, K.; Hirota, S.; Fujinami,

S.; Suzuki, M. J. Am. Chem. Soc. 1999, 121, 11006-11007.

(20) Kodera, M.; Kajita, Y.; Tachi, Y.; Katayama, K.; Kano, K.; Hirota,

S.; Fujinami, S.; Suzuki, M. Angew. Chem., Int. Ed. 2004, 43, 334-

337.

(21) Mahadevan, V.; Hou, Z. G.; Cole, A. P.; Root, D. E.; Lal, T. K.;

Solomon, E. I.; Stack, T. D. P. J. Am. Chem. Soc. 1997, 119, 11996-

11997.

(22) Mahadevan, V.; Dubois, J. L.; Hedman, B.; Hodgson, K. O.; Stack,

T. D. P. J. Am. Chem. Soc. 1999, 121, 5583-5584.

(23) Holland, P. L.; Rodgers, K. R.; Tolman, W. B. Angew. Chem., Int.

Ed. 1999, 38, 1139-1142.

(24) Itoh, S.; Taki, M.; Nakao, H.; Holland, P. L.; Tolman, W. B.; Que,

L., Jr.; Fukuzumi, S. Angew. Chem., Int. Ed. 2000, 39, 398-400.

(25) Funahashi, Y.; Nakaya, K.; Hirota, S.; Yamauchi, O. Chem. Lett. 2000,

1172-1173.

(26) Holland, P. L.; Cramer, C. J.; Wilkinson, E. C.; Mahapatra, S.; Rodgers,

K. R.; Itoh, S.; Taki, M.; Fukuzumi, S.; Que, L.; Tolman, W. B. J.

Am. Chem. Soc. 2000, 122, 792-802.

(27) Straub, B. F.; Rominger, F.; Hofmann, P. Chem. Commun. 2000,

1611-1612.

(28) Taki, M.; Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 6203-

6204.

(29) Taki, M.; Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2002, 124, 998-

1002.

7346 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

Table 1. Nomenclature of Ligands and Cu Complexes

Scheme 1. Synthesis of a Family of Peralkylated Enantiopure _RRCD

Ligands

ligand

type

CD

ED

R₁

R₂

R₃

R₄Cu(I)^a,bO^b,c

L^TM

L^EtMe

L^ME

L^TE

Me Me Me Me

1a

2a

3a

4a

5a

6a

7a

8a

9a

10a

1b

2b

3b

4b

5b

6b

7b

8b

9b

-

3

Et

Me Me Me

Me Et

Et

Me Et

Me

Et

Me

Et

-

Et

L^TEED

L^{(MeEt) ED}

2

L^{Me Et ED}

Me Me Et

2

L^{Me TACN}(X ) N)

[9]ane Me Me Me

Me Me

CD

3

L^{Me ODACN}(X ) O) [9]ane

-

2

L^TB

^aCu(I) ) [LCu(MeCN)]¹⁺

Bz Bz Bz Bz

-

.

^bCF₃SO₃counteranion assumed unless

stated otherwise. ^cO ) [L₂Cu₂(µ-O)₂]²⁺

.

Extensive study of almost 20 such Os generated from a

broad range of PDLs has helped define the general structural,

spectroscopic, and reactive character of these complexes. We

take particular interest in ligand structural attributes that

determine the products and mechanism of thermal decay for

each O, as these provide an intrinsic failure analysis for

ligand design. Developing improved ligands that support

more thermally stable Os will require a firm understanding

of these factors. Such stability, often viewed as an end in

itself, is also a prerequisite to designing a useful reagent

capable of directing its oxidizing capacity beyond its own

ligands to exogenous substrates. A more thermally robust

oxidant will provide access to substrate oxidation processes

with much larger activation barriers.

This article confines itself to the discussion and compari-

son of seven Os generated from a series of related PDLs

based on the trans-(1R,2R)-cyclohexanediamine (CD) and

1,2-ethanediamine (ED) skeletons, as well as two tridentate

Os based on polyazacyclononane ([9]ane) ligands. Although

much of the data presented in depth is derived from

compounds representative of the first group, the general

arguments herein apply to all Os studied unless otherwise

indicated. Selected Os are also synoptically compared with

other reported [Cu₂(µ-O)₂]²⁺species.

ligands L^{Me TACN}and L^{Me ODACN}were prepared according

3

2

to literature methods.^54-56The bidentate ligands L^TM

,

L

rac

,

TM

TE

L^ME, L^EtMe, L , and L^TBwere synthesized by peralkylation

3

of enantiopure CD or its racemate _racCD (Scheme 1); L^TEED

and the substitution isomers L^{Me Et ED}(nonsymmetric) and

2

L^{(MeEt) ED}(symmetric) were either purchased or obtained

2

through analogous manipulation of ED. Only the nonsym-

metrically substituted L^EtMepresented a potential synthetic

3

challenge. The desired substituent pattern was introduced via

a modified procedure for N-acylation popularized by Na-

gao^57,58in which a 3-acyl-1,3-thiazolidine-2-thione ester is

substituted for the traditional acyl chloride. This milder

reagent affords better selectivity and greater yield of the

N-monoacetylated derivative, which is conveniently isolated

and purified through sublimation. Subsequent reduction and

permethylation of this key intermediate were accomplished

in a routine manner.

Mononuclear Cu(I) Precursor Complexes. Mononuclear

Cu(I) complexes 1a-10a having the general formula [LCu-

(MeCN)]CF₃SO₃(Table 1) were synthesized by equimolar

admixture of [Cu(MeCN)₄]CF₃SO₃with ligand in purified,

anhydrous solvent (CH₂Cl₂, unless otherwise stated) in a N₂

drybox. With the exception of 2a and 7a, which contain

nonsymmetric ligands, the ¹H NMR spectra of the CD- and

ED-based 1:1 Cu(I) complexes reflect overall C₂molecular

symmetry. When 1a-10a are prepared in situ (i.e., in the

presence of 3 equiv of excess MeCN), their ¹H NMR spectra

contain only a single resonance due to MeCN at δ ) 2.05-

Results

Ligand Syntheses. All ligand syntheses involved routine

direct N-peralkylation or N-acylation followed by reduction

and further alkylation; ligand nomenclature is given in Table

1. CD-based PDLs (Scheme 1) were synthesized as the pure

1R,2R enantiomers (_RRL^X, abbreviated L^X) unless otherwise

specified (_SSL^Xor _racL^X). Tridentate facial-capping [9]ane

(54) Wieghardt, K.; Chaudhuri, P.; Nuber, B.; Weiss, J. Inorg. Chem. 1982,

21, 3086-3090.

(55) Reinen, D.; Ozarowski, A.; Jakob, B.; Pebler, J.; Stratemeier, H.;

Wieghardt, K.; Tolksdorf, I. Inorg. Chem. 1987, 26, 4010-4017.

(56) Benson, S.; Cai, P.; Colon, M.; Haiza, M.; Tokles, M.; Snyder, J. J.

Org. Chem. 1988, 53, 5335-5341.

(57) Nagao, Y.; Miyasaka, T.; Seno, K.; Yagi, M.; Fujita, E. Chem. Lett.

1981, 4, 463-466.

(58) Nagao, Y.; Seno, K.; Miyasaka, T.; Fujita, E. Chem. Lett. 1980, 159-

162.

(50) Hayashi, H.; Uozumi, K.; Fujinami, S.; Nagatomo, S.; Shiren, K.;

Furutachi, H.; Suzuki, M.; Uehara, A.; Kitagawa, T. Chem. Lett. 2002,

416-417.

(51) Mizuno, M.; Hayashi, H.; Fujinami, S.; Furutachi, H.; Nagatomo, S.;

Otake, S.; Uozumi, K.; Suzuki, M.; Kitagawa, T. Inorg. Chem. 2003,

42, 8534-8544.

(52) Berreau, L. M.; Halfen, J. A.; Young, V. G.; Tolman, W. B. Inorg.

Chim. Acta 2000, 297, 115-128.

(53) Enomoto, M.; Aida, T. J. Am. Chem. Soc. 1999, 121, 874-875.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7347

Cole et al.

Table 2. Crystal Data and Experimental Parameters for Structurally Characterized Compounds

1c 1d _RR,RR1e _RR,SS1e

[3b‚4CH₂Cl₂]

4a

[10a‚2CH₂Cl₂]

formula

CuSF₃O₃N₄C₂₁H₄₄CuC₂₉H₃₇N₂PSO₃F₃Cu₂N₄O₈F₆S₂C₂₂H₄₆Cu₂N₄O₈F₆S₂C₂₂H₄₆Cu₂S₂Cl₈F₆O₈N₄C₃₀H₆₀CuC₁₇H₃₃F₃N₃O₃S₃CuC₃₈H₄₃Cl₂F₃N₃O₃S

fw (g mol^-1

)

553.21

SMART CCD

176(3)

orthorhombic P

P2₁2₁2₁(no. 19)

14.5032(3)

25.298(1)

90

5321.3(2)

8

645.20

CAD4

203(1)

799.83

CAD4

203(1)

monoclinic P

P2₁(no. 4)

13.770(2)

8.508(1)

15.258(1)

90

108.212(8)

90

1697.9(3)

2

799.83

CAD4

203(1)

orthorhombic P

Pbca (no. 61)

8.35(1)

15.53(1)

26.16(2)

90

3392(5)

4

1193.65

SMART CCD

143(1)

monoclinic P

P2₁(no. 4)

12.1336(2)

11.9387(3)

18.4717(3)

90

108.387(1)

90

2539.19(8)

2

480.07

CAD4

203(1)

monoclinic P

P2₁(no. 4)

10.127(1)

8.290(1)

13.980(1)

90

105.52(1)

90

1130.9(2)

2

813.28

SMART CCD

159(1)

monoclinic P

P2₁(no. 4)

9.3741(2)

36.4647(7)

11.3652(3)

90

93.314(1)

90

3878.3(2)

4

diffractometer

T (K)

lattice type

space group

a (Å)

b (Å)

c (Å)

R (°)

â (°)

triclinic P

P1 (no. 1)

8.141(1)

9.017(1)

12.239(1)

68.490(7)

74.858(5)

67.016(7)

762.1(2)

1

γ (°)

V (Å³)

Z

µ

_calc(mm^-1

)

0.86

0.89

1.46

1.41

1.10

0.809

F₀₀₀

2352

1.381 [1.39]

336

1.41 [1.40]

828

1.56 [1.56]

1656

1.57 [1.55]

1224

504

1.41 [1.41]

1688

1.39 [1.40]

a

d

_calc[d_obs

]

1.56 [not measured]

(g cm^-3

)

size (mm³)

0.35 × 0.32 × 0.25 0.50 × 0.30 × 0.30 0.40 × 0.30 × 0.20 0.40 × 0.20 × 0.20 0.20 × 0.10 × 0.10

0.50 × 0.20 × 0.20 0.42 × 0.35 × 0.20

scan width ω (°)

scan rate

0.3

0.57 + 0.56 tan θ

5.5° min^-1in ω

50.0

22

2862

2811

0.035

0.66 + 0.84 tan θ

5.4° min^-1in ω

44.9

22

2495

2388

1974

0.027

0.72 + 0.75 tan θ

0.3

0.57 + 0.56 tan θ

5.5° min^-1in ω

50.0

0.3

10s exposure

52.1

9999

25,668

9340

7354

8.2° min^-1in ω

10s exposure

47.3

6569

10,340

6446

5722

20s exposure

51.6

5512

18 537

10 975

9240

max 2θ (°)

indexing reflns

total reflns

unique reflns

obsd. reflns^b

R_int

39.2

9

1756

1485

925

n/a

25

2222

2128

1834

0.059

0.047

0.051

0.030

LS parameters

612

15.26

0.052 [0.112]

0.98

363

7.89

0.028 [0.076]

1.07

0.0542 [0.2208]

0.43 [-0.47]

397

6.02

0.053 [0.150]

1.07

0.0856 [4.050]

0.89 [-0.49]

203

7.32

0.061 [0.188]

1.04

0.1046 [10.25]

0.53 [-0.36]

550

11.72

0.055 [0.129]

1.08

0.0137 [10.5184]

0.88 [-0.71]

258

8.25

0.040 [0.109]

1.04

0.0644 [0.5900]

0.43 [-0.60]

919

11.94

0.036 [0.071]

1.00

0.0203 [0.00]

0.24 [-0.29]

c

N_o/N_v

R₁[wR₂]^d

GOF, S^e

LS weights m [n]^f0.0504 [0.00]

F

_max[F_min

]

0.93 [-0.66]

(e^-Å^-3

)

^aDetermined by the neutral buoyancy method in CCl₄/hexane. ^bF_o> 4σ(F_o). ^cReflection-to-parameter ratio. ^dR₁) Σ ||F_o| - |F_c||/Σ |F_o| based on

2

observations with F_o> 4σ(F_o); wR₂) [(Σ w(F_o- F_c)²/Σ wF_o)]^1/2based on all observations. ^eS ) standard deviation of an observation of unit weight,

or [Σ w(F_o²- F_c)²/(N_o- N_v)]^1/2, where N_o) number of observations and N_v) number of variables. ^fThe least-squares function minimized is Σ w(F_o

-

2

F_c)², where w ) [σ²(F_o) + (mP)²+ nP]^-1and P ) [Max{F_o, 0} + 2 F_c]/3.

2

2.20, even at 193 K; no distinct resonance corresponding to

coordinated MeCN is observed. When samples of 4a and

10a are prepared from isolated solid, however, their spectra

exhibit a sharp singlet resonance at δ ≈ 2.15 corresponding

to one molecule of MeCN. We conclude that the auxiliary

MeCN ligand in each of these in situ-prepared Cu(I)

complexes undergoes facile exchange on the NMR time

scale. In notable contrast to the other CD- and ED-based

1

Cu(I) complexes, the H NMR spectrum of the ED-based

5a exhibits only an averaged singlet resonance corresponding

to the backbone methylenes, even at 193 K. This observation

reflects the flexibility of the ED ligand backbone in conjunc-

tion with the flat conformational potential energy profile due

to the four identical N-ethyl substituents.

Figure 1. ORTEP representation (50% probability) of the monocation of

4a, excluding hydrogen atoms. Selected interatomic distances (Å) and angles

(deg): Cu(1)-N(1), 1.864(4); Cu(1)-N(2), 2.082(4); Cu(1)-N(3),

2.047(4); N(1)-C(15), 1.130(6); N(1)Cu(1)N(2), 130.2(2); N(1)Cu(1)N-

(3), 141.7(2); N(2)Cu(1)N(3), 88.1(2); Cu(1)N(1)C(15), 174.7(4).

Crystal Structures of 1:1 Cu(I) Species. The 1:1

L/Cu(I) complexes 1a-10a were generally prepared in

situ and used in further experiments without isolation

or purification; however, suitable single crystals of

[(L^TE)Cu(MeCN)]CF₃SO₃(4a) and [(L^TB)Cu(MeCN)]CF₃-

SO₃(10a) were isolated and structurally characterized. In

the solid-state structure of 4a (Figure 1, Table 2), the Cu

atom is 3-coordinate and is ligated in an approximately

trigonal-planar fashion, the third coordination site being

occupied by the MeCN molecule. The N-Cu-N bond

angles deviate from the ideal trigonal-planar value of 120°

because of the constrained 88° bite angle of the L^TEligand.

The asymmetric unit of the [10a‚2CH₂Cl₂] crystal structure

(Figure 2 and Table 2) comprises two crystallographically

independent monocations that are isostructural to within

0.346 Å rms as a whole and to within 0.078 Å rms excluding

phenyl ring carbons. As in 4a, the approximately trigonal-

planar geometry about each 3-coordinate Cu atom is distorted

from the ideal as a result of the ∼89° bite angle of the L^TB

ligand. The relative orientations of the N-benzyl substituents

within each cation are indicative of pairwise edge-to-face

π-stacking interactions: the ortho protons of one monocation

(on C43 and C47, Figure 2) and those of the other (on C32

and C48)⁵⁹reside at distances of 2.59, 2.67, 2.59, and 2.50

Å, respectively, from the corresponding least-squares planes

(59) See Supporting Information.

7348 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

Figure 4. ORTEP representation (50% probability) of one monocation

of 1c, excluding hydrogen atoms. Selected interatomic distances (Å) and

angles (deg): Cu(1)-N(1), 2.143(6); Cu(1)-N(2), 2.146(6); Cu(1)-N(3),

2.161(6); Cu(1)-N(4), 2.126(7); N(4)Cu(1)N(1), 122.5(3); N(4)Cu(1)-

N(2), 84.8(3); N(1)Cu(1)N(2), 124.1(3); N(4)Cu(1)N(3), 122.8(3); N(1)-

Cu(1)N(3), 84.7(3); N(2)Cu(1)N(3), 122.9(3). Note: atoms belonging to

the depicted monocation are denoted in the tables⁵⁹by the suffix 1.

Figure 2. ORTEP representation (50% probability) of one monocation

of the crystal structure of [10a‚2CH₂Cl₂], excluding hydrogen atoms.

Selected interatomic distances (Å) and angles (deg): Cu(1)-N(1),

2.094(3); Cu(1)-N(3), 2.082(3); Cu(1)-N(5), 1.870(4); N(5)-C(69),

1.114(5); N(1)Cu(1)N(3), 88.7(1); N(1)Cu(1)N(5), 132.3(1); N(3)Cu(1)-

N(5), 139.0(1); Cu(1)N(5)C(69), 173.1(4).

[(_SSL^TM)₂Cu]CF₃SO₃(1c), is isolated that presents an inter-

esting case of ligand self-sorting (vide infra). Single crystals

of the racemic⁶⁰1c were characterized by X-ray diffraction

(Figure 4, Table 2). Despite its tetragonal Laue symmetry,

the crystal lattice in 1c is best represented as two ortho-

rhombic P2₁2₁2₁unit cells twinned at a ∼45:55% ratio by

interchange of the a and b axes. The asymmetric unit of each

twin component contains two enantiomeric yet crystallo-

graphically unique monocations; if one of the pair is inverted,

their coordinates are then isostructural to within 0.13 Å rms.⁶¹

Relative to the 1:1 L/Cu(I) complexes reported above, the

Cu-N_aminedistances in 1c (av 2.15(2) Å) are lengthened by

0.05-0.08 Å and the L^TMchelate bite angles are reduced

by 3-4°, reflecting the electrostatic consequences of in-

creased coordination number at a closed-shell d¹⁰metal

center.

Low-Temperature Oxygenation of 1a-9a and Char-

acterization of Os. At appropriate concentrations,⁶²the

Cu(I) precursor complexes 1a-9a react with dry O₂at 193

K in CH₂Cl₂, THF, or acetone to afford the yellow- to red-

brown, thermally unstable Os 1b-9b, respectively. A 2:1

Cu/O₂uptake stoichiometry was documented for several of

these reactions by manometry (3a, 4a, 5a, 7a; [Cu] ) 30

mM, 1 atm O₂, CH₂Cl₂). The incorporated O₂could not be

reversibly displaced from any of the resulting Os by cycles

of evacuation and purging with N₂, or by exposure to excess

CO_(g)(1 atm) or PPh₃(∼100 equiv). EPR spectra at 77 K

were essentially featureless for all Os; in no case did

mononuclear Cu(II) signals account for more than 5% of

Figure 3. ORTEP representation (50% probability) of the monocation of

1d, excluding hydrogen atoms. Selected interatomic distances (Å) and angles

(deg): Cu(1)-N(1), 2.041(3); Cu(1)-N(2), 2.065(3); Cu(1)-P(1),

2.150(1); N(1)Cu(1)N(2), 88.2(1); N(1)Cu(1)P(1), 140.3(1); N(2)Cu(1)-

P(1), 131.3(1).

defined in each case by the six carbon atoms of the nearest

neighboring phenyl ring. Solutions of 10a appeared to be

air-stable upon standing for several days at 298 K; however,

a CO adduct, [(L^TB)Cu(CO)]CF₃SO₃, having an IR stretch

of ν_C≡O) 2113 cm^-1formed readily upon exposure of a

CDCl₃solution of 10a to CO_(g)(1.7 atm, 15 min).

The auxiliary MeCN ligands in 1a-9a are readily

displaced by PPh₃; X-ray-quality crystals of a representative

air-stable adduct, [(L^TM)Cu(PPh₃)]CF₃SO₃, 1d, were readily

isolated following addition of a stoichiometric quantity of

(60) A similar air-stable 2:1 L/Cu compound was also derived from

TM

enantiopure

obtained.

L

, but crystals suitable for diffraction could not be

RR

1

PPh₃to an anaerobic CH₂Cl₂solution of 1a. The H NMR

(61) Considered separately, the monocations of 1c obey the higher

tetragonal space symmetry implied by the lattice dimensions: ignoring

counterions, the model may be adequately represented in the space

group I-42d, with each monocation residing on a crystallographic

2-fold axis and being related to the others by the principal 4-fold

improper rotation axis.

(62) Oxygenation of 1a, 2a, or 6a also gives rise to significant amounts of

a 3:1 Cu/O₂species when the Cu/O₂concentration ratio is sufficiently

greater than unity. We have previously reported structural⁴²and

spectroscopic^63,64characterizations of the first of these, a trinuclear

[L₃Cu₃(µ₃-O)₂]³⁺complex; Itoh has also recently reported a similar

3:1 Cu/O₂species.³⁰

spectrum of 1d in the L^TMligand region is analogous to that

of 1a, which was not isolated in solid form. The geometry

about the Cu(I) center in 1d is roughly trigonal-planar but

is again distorted by the L^TMligand bite angle of ∼88°

(Figure 3); the Cu-N and Cu-P distances are unexceptional.

TM

2:1

L

/Cu(I) Complex. When a 2:1

L

rac

/Cu stoi-

rac

chiometry is employed in place of an equimolar mixture, a

colorless, air-stable compound, [(_RRL^TM)₂Cu]CF₃SO₃‚

Inorganic Chemistry, Vol. 44, No. 21, 2005 7349

Cole et al.

Table 3. Spectroscopic, Structural, and Thermal Decomposition Data for Selected Os^a

rR shifts (cm^-1

decay rate

activation

parameters

ligand

products

% yield

)

T ) 263 K

r_{Cu‚‚‚Cu}(Å)

[r_Cu-O(Å)]

LMCT λ_max(nm)

ν¹⁶O₂[ν¹⁸O₂]

k_obs(s^-1

)

∆H^q(kJ mol^-1

)

ligand

O

[ꢀ (10³M^-1cm^-1)]

{∆ν (¹⁶O₂-¹⁸O₂)}

[t_1/2(s)]

[∆S^q(J K^-1mol^-1)]

56(1) [-102(5)]

[theor max]

L^TM

L^EtMe

L^ME

L^TE

1b

-

301^b[20]

399^b[25]

605 [581]^c

{23}

2.0 × 10^-4

30 [50]^d

-

[3500]

2b

3b

4b

5b

6b

7b

8b

9b

-

305 [19]

402 [27]

313^b[21]

408^b[28]

319^b[17]

413^b[23]

307^b[17]

407^b[24]

-

3.5 × 10^-4

-

3

[2000]

2.744,^e2.74^f,g

[1.802,^e1.81^f,g

610 [587]^c

{23}

616 [590]^c

{26}

603 [572]^c

{31}

4.0 × 10^-3

49(1) [-105(5)]

40 [50]^h

50 [50]

30 [50]

8,^k5^l[50]

15 [50]^h

22 [25]ⁿ

20 [25]ⁿ

-

]

[170]

-

3.3 × 10^-2

49(1)ⁱ[-84(4)]ⁱ

[21]

L^TEED

L^{(MeEt) ED}

2.74^f,g

7.0 × 10^-3

60(1)^i,j[-54(4)]^i,j

[1.80]^f,g

[100]

-

297 [23]

397 [25]

-

5.4 × 10^-4

-

2

[1280]

L^{Me Et ED}

-

297 [18]

397 [21]

307^b[16]

412^b[18]

-

1.1 × 10^-3

-

2

[630]

L^{Me TACN}

2.77^f,g

604 [581]^c

{23}

1.7 × 10^-2

52(1) [-78(4)]

3

[1.81]^f,g

[40]

L^{Me ODACN}

-

297 [16]

397 [15]

297^b[16]

397^b[24]

-

1.4 × 10^-3

61(2) [-66(4)]

2

[500]

L^TMPD

2.85^f

608 [581]^c

{27}

-

(Stack et al.)^m

[1.81]^f

L^{Bn TACN}

-

2.794^e

[1.81]^e

2.906^e

[1.86]^e

2.758^e

[1.80]^e

2.866^e

[1.83]^e

318 [12]

430 [14]

602/608^p[583]

{23}

9.6 × 10^-2

58(2) [-41(5)]

40 [50]

-

3

(Tolman et al.)^o

[7]

t

₂P(NSiMe₃)₂N

L^Bu

-

315 [shoulder]

444 [10]

-

(Hofmann et al.)^q

L^{Me tmpa}

-

300 [12]

378 [19]

590 [564]

{26}

-

2

(Suzuki et al.)^s

L^{Me etpy}

-

s

579 [551]

{28}

-

2

(Kitagawa, Suzuki et al.)^r

390 [shoulder]

HMe₂L^iPr/L

-

2.849^e

s

653 [625]

{28}

-

TMPD

2

(Tolman et al.)^t

[1.818]^e

398 [17]

L^{iPr dtne}

-

2.783^e

316 [14]

414 [13]

600 [582]

{18}

2.9 × 10^-2

56(2) [-59(5)]

-

4

(Tolman et al.)^u

[1.825]^e

[24]

^aAll measurements were performed in CH₂Cl₂solution except where otherwise noted. Listed concentrations and molar absorptivities are uncorrected for

the thermal contraction of the solvent, which can be as great as 15%. ^bλ_maxvalue includes a correction of +7 nm relative to previously reported values (refs

21, 66, and 69) to compensate for an instrument miscalibration. ^cSee refs 21 and 66. ^dAn additional formamide 4e^-oxidation product (L^X+O) is also

observed for [Cu] ) 2-5 mM (∼5% [20% theor], [Cu] ) 2.0 mM, 263 K). ^eDistance derived from crystal structure; Cu-O values are a molecular average.

^fDistance derived from solid-state EXAFS. ^gSee refs 64 and 69. ^hL^X-Etonly. ⁱEquivalent kinetic results observed for THF and CH₂Cl₂solutions. ^jEquivalent

kinetic results observed for [5b‚(SbF₆)₂] in THF. ^kL^{X-Et l}L^{X-Me m}L^TMPD) N,N,N′,N′-tetramethyl-1,3-propanediamine. See ref 22. ⁿ4e^-L^X+Oformamide

. .

product only, [Cu] ) 1.0 mM. ^oSee refs 2 and 16. ^pFermi doublet. ^qSee ref 27. ^rSee ref 49. ^sSee ref 51. ^tAssembled in a stepwise process from

[(HMe₂L^iPr)Cu(O₂)] and [(L^TMPD)Cu(MeCN)]⁺; see ref 32. ^uSee refs 2 and 17.

2

total Cu content. NMR susceptibility measurements⁶⁵

(190 K, CD₂Cl₂) did not indicate significant paramagnetism

in any of the reported Os. The electronic absorption spectra

of 1b-9b exhibit characteristically intense ligand-to-metal

charge transfer (LMCT) bands at λ_max≈ 300 (ꢀ ) 10 000-

20 000 M^-1cm^-1) and 400 nm (ꢀ ) 10 000-30 000 M^-1

cm^-1). The LMCT-enhanced (λ_excitation) 407 nm) resonance

Raman (rR) spectra of 1b, 3b, 4b, 5b, and 8b contain

similarly intense O₂isotope-sensitive peaks at ∼600 cm^-1.

summarized in Table 3 together with those of other pertinent

literature examples.

Crystal Structure of [(L^ME)₂Cu₂(µ-O)₂](CF₃SO₃)₂‚

4CH₂Cl₂. Thermally sensitive crystals of [3b‚4CH₂Cl₂] were

grown from a concentrated CH₂Cl₂solution by layering with

Et₂O at 193 K. These were characterized by X-ray diffraction

(Table 2); relevant metrical parameters are summarized in

Figure 5. The well-ordered asymmetric unit contains a

crystallographically unique binuclear [(L^ME)₂Cu₂(µ-O)₂]

cluster in which the dioxygen O-O bond has been cleaved

(O‚‚‚O, 2.334 Å). The most striking metrical features are

the extremely short Cu-O bonds at both Cu centers (av Cu-

O, 1.802(7) Å) and the short Cu‚‚‚Cu separation (2.744(1)

These spectroscopic features and rR O₂isotope shifts^21,66are

-

(63) Root, D. E.; Henson, M. J.; Machonkin, T.; Mukherjee, P.; Stack, T.

D. P.; Solomon, E. I. J. Am. Chem. Soc. 1998, 120, 4982-4990.

(64) Dubois, J. L.; Mukherjee, P.; Stack, T. D. P.; Hedman, B.; Solomon,

E. I.; Hodgson, K. O. J. Am. Chem. Soc. 2000, 122, 5775-5787.

(65) Evans, D. F. J. Chem. Soc. 1959, 2003-2005.

(66) Henson, M. J.; Mukherjee, P.; Root, D. E.; Stack, T. D. P.; Solomon,

E. I. J. Am. Chem. Soc. 1999, 121, 10332-10345.

-

Å). The two noncoordinating CF₃SO₃ions present in the

asymmetric unit imply a net 2+ charge on the cation. Each

Cu atom is ligated in an approximately square-planar [N₂O₂]

7350 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

spectroscopic runs gave k_obs) 2.3 × 10^-3s^-1for this mixed-

solvent system. An Eyring fit comprising k_obsvalues mea-

sured between 195 and 213 K yielded activation parameters

∆H^q) 25(4) kJ mol^-1/6(1) kcal mol^-1and ∆S^q

)

-210(40) J K^-1mol^-1/-50(10) cal K^-1mol^-1; the accessible

temperature range for kinetic experiments was constrained

by the freezing point of the solvent and the onset of rapid

thermal decomposition of 4b above 213 K.

Reactivity with Exogenous Substrates. O complexes

1b-7b were screened against several organic substrates in

order to determine the characteristic reactivity of the O core.

Typically, solutions of each O ([Cu] ) 1.0 mM) were treated

with 1-5 equiv of substrate at 193 K under a N₂atmosphere,

and the ensuing reactions were allowed to proceed until either

the characteristic UV-vis absorption features of the O

chromophore had been quenched or the maximum interval

of 6 h had elapsed. After workup, the organic products were

assayed by GC-MS and/or ¹H NMR spectroscopy (generally,

the Cu products were not characterized). Os 1b-7b reacted

rapidly in >95% yield with 2 equiv of 2,4-di-tert-butylphenol

(DBP), thiocresol, or tert-butylthiol to give 3,3′,5,5′-tetra-

tert-butyl-2,2′-bis(phenol), 4-methylphenyl disulfide, or tert-

butyl disulfide, respectively, but showed no apparent reac-

tivity (<5% yield) toward ethylene, cyclohexene, 1,4-

cyclohexadiene, 9,10-dihydroanthracene, xanthene, fluorene,

(CH₃)₂S, (CH₃)₂SO, or PhSCH₃(1-5 equiv). These results

are summarized in Scheme 2. Reaction of 1b-7b with the

lithium alkoxide salts of benzyl and cinnamyl alcohols (2

equiv) rapidly gave the corresponding aldehydes in >95%

yield, whereas the alcohols themselves displayed minimal

(<5%) reactivity even upon warming (233 K). A representa-

tive aliphatic primary alcohol, 1-octanol (2 equiv), was

similarly unreactive even when deprotonated. Spectropho-

tometric titrations of 1b-5b (193 K, N₂) with 2 equiv of

ferrocene (Fc) or decamethylferrocene (Me₁₀Fc) quenched

the characteristic LMCT optical bands, whereas 1.5 equiv

of acetylferrocene (AcFc) was unreactive through 6 h in all

cases.

Figure 5. ORTEP representation (50% probability) of the dication of

[3b‚4CH₂Cl₂], excluding hydrogen atoms. Selected interatomic distances

(Å) and angles (deg): Cu(1)-O(1), 1.809(6); Cu(1)-O(2), 1.808(6);

Cu(2)-O(1), 1.795(5); Cu(2)-O(2), 1.799(6); Cu(1)‚‚‚Cu(2), 2.744(1);

O(1)‚‚‚O(2), 2.334(1); O(2)Cu(1)O(1), 80.4(2); N(1)Cu(1)N(2), 89.9(3);

O(1)Cu(2)O(2), 81.0(2); N(4)Cu(2)N(3), 89.3(3); Cu(2)O(1)Cu(1),

99.2(3); Cu(2)O(2)Cu(1), 99.1(3).

environment; however, the [Cu₂O₂N₄] core as a whole is

slightly twisted, with a dihedral angle of 7.5(2)° between

the two [N₂CuO₂] planes. A highly ordered, pseudo-C₂-

symmetric alternation of N-ethyl and N-methyl substituents

within and between the ligands is observed. This arrangement

places one methylene proton on each of the four N-ethyl

substituents in proximity to the bridging oxide ligands (C8-

H8A‚‚‚O2, 2.51 Å; C11-H11A‚‚‚O1, 2.47 Å; C20-

H20B‚‚‚O1, 2.60 Å; C23-H23B‚‚‚O2, 2.60 Å). Restrained

refinement of the N-methyl hydrogens converges to the

expected staggered conformation in each case, thus placing

the four proximate methyl protons at somewhat greater

average distance from the oxide ligands (C7-H7A‚‚‚O2,

2.69 Å; C10-H10B‚‚‚O1, 2.63 Å; C19-H19B‚‚‚O1, 2.72

Å; C22-H22B‚‚‚O2, 2.58 Å).

Formation Kinetics. In general, the diffusion-limited

formation of 1b-9b from 1a-9a proceeds rapidly in O₂-

saturated CH₂Cl₂, acetone, or THF, with full O formation

occurring in ∼10-15 s (1 atm O₂, [Cu] ) 0.1-2.0 mM,

193 K).⁶⁷The sole exception to this norm, apart from the

wholly inert 10a (q.v.), is 4a, the oxygenation of which

proceeds with like alacrity in pure acetone or THF but

requires ∼2 h for completion in pure, O₂-saturated CH₂Cl₂.

A pseudo-first-order rate constant of k_obs) 4.4 × 10^-4s^-1

was obtained for the latter case by means of a multivariate

spectral analysis in which absorbance data (λ ) 280-450

nm) were fitted to a simple first-order A f B model. No

colored intermediate species were noted. The reaction was

also followed in O₂-saturated 1:1 (v/v) THF/CH₂Cl₂by

monitoring the optical absorbance of a characteristic LMCT

band due to 4b (λ_max) 413 nm) during the initial 5% (200

s) of the reaction coordinate (0.2 < [4a]₀< 2.0 mM, 1 atm

O₂, 193 K). A plot of log rate_initialvs log [4a]₀showed a

linear relationship corresponding to a rate law that is first-

order in [4a] with k_obs) 2.5 × 10^-3s^-1under the above

conditions;⁵⁹multivariate analyses (280-450 nm) of multiple

Bis(µ-hydroxo)dicopper(II) Dimers. The bis(µ-hydroxo)-

dicopper(II) dimer [(L^TM)₂Cu₂(µ-OH)₂](CF₃SO₃)₂(_RR,RR1e)

was synthesized in near-quantitative yield via reaction of 1b

with 2 equiv of a sacrificial 1e^-, 1H⁺donor such as DBP or

tert-butylthiol and was also isolated in lesser yield (60%)

from thermally decomposed solutions of 1b. The racemic

dimer [(_SSL^TM)(_RRL^TM)Cu₂(µ-OH)₂](CF₃SO₃)₂(_RR,SS1e) was

TM

obtained in an analogous manner using

L

. In contrast

rac

to 1b, _RR,RR1e and _RR,SS1e are stable indefinitely at 293 K in

aprotic solvents, and their electronic spectrum (λ_max) 269

nm; ꢀ ) 14 000 M^-1cm^-1) contain no low-energy (>300

nm) LMCT bands.

Both _RR,RR1e and _RR,SS1e were structurally characterized

(Table 2). The enantiomerically pure _RR,RR1e crystallizes in

the acentric group P2₁(Figure 6). Both Cu(II) atoms are

coordinated in an approximately square-planar geometry,

with no unusual bond distances and a substantially larger

Cu‚‚‚Cu separation (2.99 Å) relative to all structurally

characterized Os (Table 3). Each bridging hydroxo ligand

participates in a hydrogen-bonding interaction with a triflate

(67) The manner in which O₂is introduced into the reaction vessel has a

significant effect on formation rates. For example, full formation of

1b at 193 K is complete within 10 s if concentrated 1a is injected

into O₂-saturated CH₂Cl₂but requires several minutes if O₂is

introduced afterwards by bubbling.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7351

Cole et al.

Scheme 2. Reactivity of Selected Os toward Exogenous Substrates^a

b

^aConditions (unless otherwise stated): 193 K, CH₂Cl₂, N₂atmosphere, 2 equiv of substrate/O, max 6 h. 1 atm.

Figure 6. ORTEP representation (50% probability) of the dication of

_RR,RR1e, excluding hydrogen atoms. Selected interatomic distances (Å) and

angles (deg): Cu(1)-O(1), 1.89(1); Cu(1)-O(2), 1.917(8); Cu(2)-O(1),

1.916(8); Cu(2)-O(2), 1.92(1); Cu(1)‚‚‚Cu(2), 2.990(2); O(1)‚‚‚O(2), 2.37;

O(1)Cu(1)O(2), 77.1(4); N(2)Cu(1)N(1), 86.7(4); O(1)Cu(2)O(2), 76.5(4);

N(3)Cu(2)N(4), 87.4(4); Cu(1)O(1)Cu(2), 103.5(5); Cu(1)O(2)Cu(2),

102.5(5).

Figure 7. ORTEP representation (50% probability) of the dication of

_RR,SS1e, excluding hydrogen atoms. Asterisks denote inversion-related atoms.

Selected interatomic distances (Å) and angles (deg): Cu(1)-O(1),

1.913(8); Cu(1)-O(1*), 1.908(8); Cu(1)‚‚‚Cu(1*), 2.985(3); O(1)‚‚‚O(1*),

2.39; O(1*)Cu(1)O(1), 77.3(4); N(1)Cu(1)N(2), 87.1(4); Cu(1*)O(1)-

Cu(1), 102.7(4).

oxygen atom (O1‚‚‚O7, 2.89(2) Å; O2‚‚‚O5, 2.94(2) Å).

Although each half of the _RR,RR1e dimer is crystallographically

unique, the two parts are superimposable and isostructural

to within 0.1 Å rms. The least-squares planes defined by

each respective [N₂CuO₂] unit are twisted by 12.5° relative

to one another, a distortion mandated by the chirality of the

ligand.

By contrast, the racemic _RR,SS1e crystallizes in a cen-

trosymmetric space group and contains one ligand of RR type

and one of SS type, with each heterochiral cation resting on

a crystallographic inversion center (Figure 7). Thus, neither

intra- nor intermolecular spontaneous resolution to homo-

chiral enantiomers is observed in the solid state, and the Cu,

N, and O atoms of _RR,SS1e are rigorously coplanar as required

by crystallographic symmetry. Bond distances are again

typical of Cu(II) complexes, and weak hydroxo-triflate

hydrogen-bonding interactions similar to those of _RR,RR1e are

also evident (O1‚‚‚O7, 3.09(2) Å).

Thermal Decomposition. The thermal decomposition

kinetics of 1b-9b were studied by UV-vis spectroscopy

and multivariate spectral analysis. The time-dependent evolu-

tion of the absorbance profile (λ ) 280-450 nm) was

successfully fitted in each case to a simple first-order A f

B model involving two principal colored components.

Decomposition rate constants (k_obs) obtained in this manner

for complexes 1b, 3b, 4b, 5b, 8b, and 9b at temperatures

between 223 and 293 K⁶⁸were used to generate Eyring

plots⁵⁹from which the activation parameters ∆H^qand ∆S^q

were derived (Table 3). Primary kinetic isotope effect (KIE)

values for the thermal decay of 1b and 3b were obtained

through experiments conducted with the corresponding

(68) 1b: 293-263 K, [Cu] ) 1.0 mM, CH₂Cl₂; 3b: 280-248 K, [Cu] )

1.0 mM, CH₂Cl₂; d₈-3b: 291-253 K, [Cu] ) 1.0 mM, CH₂Cl₂; 4b:

263-223 K, [Cu] ) 1.0 mM, THF; 5b: 263-233 K, [Cu] ) 1.0

mM, THF; 8b: 263-233 K, [Cu] ) 0.2 mM, CH₂Cl₂; 9b: 278-243

K, [Cu] ) 1.0 mM, CH₂Cl₂.

7352 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

deuterated ligands d₁₂-L^TM(perdeuterated methyl groups) and

d₄-L^ME(deuterated N-methylene only). For 1b/d₂₄-1b, a

calculations^70,82indicate that the Cu(III) d⁸system prefers a

rigorously square-planar coordination environment given a

ligand field of sufficient strength, which suggests that

bidentate strong σ-donor ligands with constrained bite angles

near 90° should provide optimal thermodynamic stabilization

of the d⁸Cu(III) sites in the O core. Comparison of Os based

on tridentate and bidentate amine ligands reveals extensive

spectroscopic and structural similarity (Table 3). Yet triden-

tate amine ligands give rise to 5-coordinate [N₃O₂] Cu(III)

sites with weakly bound axial nitrogen donors that restrict

physical access to the [Cu₂O₂]²⁺core relative to a more open

[N₂O₂] square-planar coordination environment. Such ac-

cessibility is essential to inner-sphere oxidative processes that

involve exogenous substrates.²²Its importance is well

documented in the case of Hc, which can acquire phenol

oxidase activity via removal of residues that block direct

substrate access.^83,84A similar concept may also underlie the

catalytic activity of tyrosinase: in its oxygenated form, the

tertiary structure of this Cu monooxygenase allows substrates

freer access to the P core than does its nonenzymatic,

spectroscopically analogous counterpart oxyHc.^85,86

D

primary KIE value of k_obs^H/k_obs≈ 8 was observed at 293

D

K, whereas for 3b/d₈-3b, a primary KIE value of k_obs^H/k_obs

≈ 3 was calculated for T ) 293 K by extrapolation from

linear Eyring equation fits. For each decomposed O, the

corresponding ligand products were recovered and analyzed.

Reproducible ratios of unchanged ligand (L^X) to mono-

dealkylated products indicative of 2e^-oxidation (L^X-R) and/

or oxygenated formamide products indicative of 4e^-oxida-

tion (L^X+O) were quantified by GC-MS (Table 3); the latter

were observed as minor products at higher concentrations

of 1b ([Cu] ) 2-5 mM) and were the major ligand products

recovered from the thermal decay of 8b and 9b.

Discussion

Ramifications of Ligand Design. A recently published

analysis⁷⁰of theoretical calculations^46,71,72that address the

relative energies of isomeric µ-η²:η²-peroxodicopper(II) (P)

and bis(µ-oxo)dicopper(III) (O) complexes underscores the

central role of ligand structure in determining the nature and

distribution of Cu-O₂species within a given coordination

environment. Ligand steric demands, denticity, chelation

geometry, and donor characteristics are critical factors. For

example, simple bidentate 1,2-PDLs differentiated solely on

the basis of N-alkyl substituent size and conformational

rigidity support a range of species encompassing 2:1 (P^34,35

and O^21,22), 3:1 (trinuclear),⁴²and 4:1^40,73Cu/O₂reaction

stoichiometries; a similar trend was also recently reported

for a series of 2-(2-pyridyl)ethylamines.³⁰Likewise, backbone

perturbations within a family of â-diketiminate ligands are

reported to facilitate selective formation of charge-neutral

[LCu(O₂)] (1:1) and O (2:1) species.^31-33,74Studies of

interconverting P/O systems demonstrate that the distribution

of species may be manipulated through the use of appropriate

solvents and counteranions.^35,44,75,76The variation of second-

ary electronic characteristics within a ligand family has also

demonstrated usefulness in fine-tuning such equilibria.⁷⁷

Chirality and Conformational Rigidity. Both ED and

CD provide convenient, modular synthetic access to a

homologous series of simple Cu(I)-complexing PDLs (Scheme

1) that possess the essential ligating characteristics necessary

to support generation of Os. These ligands span a steric

continuum within which the nature of the proximal N-alkyl

substituents may be incrementally varied, allowing systematic

study of the effects of ligand design on the formation,

stability, and reactivity of these complexes. In this context,

the CD backbone affords several refinements relative to ED.

In the most thermodynamically stable “chair” conformation,

the two trans amino groups are equatorially oriented, creating

a highly predisposed binding pocket for coordination of

transition metals. CD is also optically active, and its

enantiomers are conveniently resolved.⁸⁷The use of a

conformationally restricted enantiopure ligand in the spon-

taneous self-assembly of dimers significantly reduces the

likelihood of obtaining multiple diastereomeric products or

conformers of products,^88-90and it offers the additional

Although frequently observed in natural Cu proteins, axial

ligation is not a requisite condition for supporting Cu-O₂

intermediates. Electrochemical studies^78-81and theoretical

(79) Youngblood, M. P.; Margerum, D. W. Inorg. Chem. 1980, 19, 3068-

3072.

(80) Ruiz, R.; Surville-Barland, C.; Aukauloo, A.; Anxolabehere-Mallart,

E.; Journaux, Y.; Cano, J.; Mun˜oz, M. C. J. Chem. Soc., Dalton Trans.

1997, 745-751.

(81) Cervera, B.; Sanz, J. L.; Iba´n˜ez, M. J.; Vila, G.; Lloret, F.; Julve, M.;

Ruiz, R.; Ottenwaelder, X.; Aukauloo, A.; Poussereau, S.; Journaux,

Y.; Mun˜oz, M. C. J. Chem. Soc., Dalton Trans. 1998, 781-790.

(82) Liu, X. Y.; Palacios, A. A.; Novoa, J. J.; Alvarez, S. Inorg. Chem.

1998, 37, 1202-1212.

(83) Decker, H.; Dillinger, R.; Tuczek, F. Angew. Chem., Int. Ed. 2000,

39, 1591-1595.

(84) Decker, H.; Tuczek, F. Trends Biochem. Sci. 2000, 25, 392-397.

(85) Jaenicke, E.; Decker, H. ChemBioChem 2004, 5, 163-169.

(86) Decker, H.; Jaenicke, E. DeV. Comp. Immunol. 2004, 28, 673-687.

(87) Whitney, T. A. J. Org. Chem. 1980, 45, 4214-4216.

(88) Enemark, E. J.; Stack, T. D. P. Angew. Chem., Int. Ed. Engl. 1995,

34, 996-998.

(89) Enemark, E. J.; Stack, T. D. P. Angew. Chem., Int. Ed. 1998, 37, 932-

935.

(90) Masood, M. A.; Enemark, E. J.; Stack, T. D. P. Angew. Chem., Int.

Ed. 1998, 37, 928-932.

(69) Dubois, J. L.; Mukherjee, P.; Collier, A. M.; Mayer, J. M.; Solomon,

E. I.; Hedman, B.; Stack, T. D. P.; Hodgson, K. O. J. Am. Chem. Soc.

1997, 119, 8578-8579.

(70) Siegbahn, P. E. M. J. Biol. Inorg. Chem. 2003, 8, 577-585.

(71) Be´rces, A. Inorg. Chem. 1997, 36, 4831-4837.

(72) Flock, M.; Pierloot, K. J. Phys. Chem. A 1999, 103, 95-102.

(73) Mukherjee, P. Ph.D. Thesis, Stanford University, 2000.

(74) Aboelella, N. W.; Kryatov, S. V.; Gherman, B. F.; Brennessel, W.

W.; Young, V. G.; Sarangi, R.; Rybak-Akimova, E. V.; Hodgson, K.

O.; Hedman, B.; Solomon, E. I.; Cramer, C. J.; Tolman, W. B. J. Am.

Chem. Soc. 2004, 126, 16896-16911.

(75) Cahoy, J.; Holland, P. L.; Tolman, W. B. Inorg. Chem. 1999, 38,

2161-2168.

(76) Liang, H. C.; Henson, M. J.; Hatcher, L. Q.; Vance, M. A.; Zhang, C.

X.; Lahti, D.; Kaderli, S.; Sommer, R. D.; Rheingold, A. L.;

Zuberbuhler, A. D.; Solomon, E. I.; Karlin, K. D. Inorg. Chem. 2004,

43, 4115-4117.

(77) Henson, M. J.; Vance, M. A.; Zhang, C. X.; Liang, H. C.; Karlin, K.

D.; Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 5186-5192.

(78) Bossu, F. P.; Chellappa, K. L.; Margerum, D. W. J. Am. Chem. Soc.

1977, 99, 2195-2203.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7353

Cole et al.

prospect of asymmetric induction in reactions with exogenous

substrates.⁹¹

Crystal Structure of [3b‚4CH₂Cl₂]. Low-temperature

crystallography of Cu-O₂species is frequently hampered

by substantial disorder and other intrinsic problems that lead

to high residual values when refining against all intensity

data. In this context, the solid-state structure of [3b‚

4CH₂Cl₂] (Figure 5) is notable for its high data redundancy

(∼12 observations per refined parameter), low residual values

(R₁) 5.5%; wR₂) 12.9% on F²), the absence of significant

solvent or counteranion disorder, and the fact that its species

of interest occupies a general position (as opposed to a special

position, or symmetry element). The metrical parameters of

the O core in the crystal structure closely correspond to those

derived from solid-state EXAFS (vide infra).^64,69The refined

model provides insight into the constraints imposed by the

chiral L^MEligands on the self-assembly of 3b. The alternating

“up, down” geared intra- and interligand arrangement of the

N-ethyl and N-methyl substituents (Figure 5) affords steric

protection to the O core while minimizing energetically

unfavorable interligand interactions: the closest approach of

any pair of terminal substituent carbon atoms on alternate

As is customary, peripheral ligand alkyl groups are

employed to preclude the formation of a 2:1 L/Cu square-

planar Cu(II) complex via spontaneous disproportionation

of the Cu(I) precursor. With a sterically compact ligand such

as L^TM, formation of a 2:1 [L₂Cu]¹⁺Cu(I) complex is also

possible (1c, Figure 4), but the proximity of the ligand

N-alkyl substituents enforces a tetrahedral coordination

geometry that stabilizes the Cu(I) oxidation state. Accord-

ingly, 1c is inert to O₂in the solid state even at 298 K. Its

crystal structure provides an elegant example of intramo-

lecular chiral self-recognition, such as we have previously

reported⁹⁰for Cu(I) complexes of other _racCD-based

ligands: although synthesized from _racL^TM, 1c is composed

exclusively of homochiral units in the solid state.⁶⁰Com-

parison of the energies derived from hybrid DFT calculations

on homochiral (1c) and heterochiral (_RR,SS1c) formulations

suggests that the homochiral complex is more thermody-

namically stable than its hypothetical centrosymmetric

counterpart by ∆∆G° ≈ 0.9 kcal mol^-1at 298 K.⁵⁹The

influence of ligand chirality on the structures of self-

assembled complexes is further illustrated by comparison of

the crystal structures of the bis(µ-hydroxo)dicopper(II) dimers

_RR,RR1e and _RR,SS1e: the homochiral L^TMligands of the former

impart a 12.5° twisting distortion to its [N₄Cu₂O₂] basal

plane, whereas the inversion-related ligand pair in _RR,SS1e

gives rise to a planar [N₄Cu₂O₂] unit.

ligands is 4.57 Å (C10‚‚‚C21). The homochirality of the

ME

L

ligands imparts a slight ∼8° twist to the basal [N₂-

RR

CuO₂CuN₂] unit, similar to that observed in the homochiral

bis(µ-hydroxo)dicopper(II) dimer _RR,RR1e (vide supra). The

observed intramolecular C-H‚‚‚O interactions between the

alkyl substituent NC^Rprotons and the bridging oxide ligands

have been noted in reports of other structurally characterized

Os^16,51and are suggestive of the proposed thermal decom-

position mechanism, which involves selective C-H bond

cleavage at these positions (vide infra).

O₂Reactivity of 1:1 L/Cu(I) Complexes. The reaction

of each ligand with [Cu(MeCN)₄]CF₃SO₃at 1:1 stoichiom-

etry yields a discrete, monocationic Cu(I) complex with at

least one MeCN molecule bonded as a labile auxiliary. The

¹H NMR symmetry and chemical shifts of these complexes

are consistent with either a tetrahedral 4-coordinate [LCu-

(MeCN)₂]¹⁺formulation or a trigonal-planar 3-coordinate

[LCu(MeCN)]¹⁺configuration like that observed in the solid-

state structures of 4a and 10a. If the former species is present,

the rapid exchange of MeCN observed by NMR implies a

facile equilibrium with the latter. Given the presumed

associative nature of O₂binding (vide infra), we believe that

such 3-coordinate [LCu(MeCN)]¹⁺species are the active

precursors in the formation of Cu-O₂species with bidentate

ligands. A recent study of host-guest exchange interactions

between tridentate calix[6]arene-Cu(I) complexes and aux-

iliary nitrile proligands indicates that the large enthalpic cost

of Cu(I)-nitrile bond scission is significantly offset by the

favorable entropic contribution arising from the overall

increase in the number of particles upon dissociation of the

nitrile from the complex.⁹²The structurally characterized

Cu(I) species [{(-)-sparteine}Cu(MeCN)₂]CF₃SO₃, which

is 4-coordinate in the solid state, reacts readily with O₂at

low temperatures to form an O;²⁵again, a facile equilibrium

between 3- and 4-coordinate species is presumed to be

essential to the observed O₂reactivity.

Among structurally characterized Os, 3b is the most

metrically compact example yet documented. It belongs to

a group of several among which the Cu‚‚‚Cu distance varies

by only 0.05 Å (L ) L^ME, av Cu-O 1.802(7) Å, Cu‚‚‚Cu

2.744(1) Å; L^{Bn TACN}, av Cu-O 1.81 Å, Cu‚‚‚Cu 2.794 Å;¹⁶

3

L^{iPr dtne}, av Cu-O 1.825 Å, Cu‚‚‚Cu 2.783 Å;¹⁷L^{Me tmpa}

,

4

2

av Cu-O 1.80 Å, Cu‚‚‚Cu 2.76 Å;⁴⁹Table 3). For L )

L^{Bn TACN}, e.g., the [Cu₂O₂] fragment is isostructural with that

3

of [3b‚4CH₂Cl₂] to within 0.046 Å rms, and the [Cu₂N₄O₂]

fragment to within 0.094 Å rms. All of these Os are based

on ligands that form five-membered chelate rings with the

Cu ions. O complexes that contain six-membered chelate

rings, by contrast, exhibit longer Cu‚‚‚Cu separations

2

(2.866

Å,

L

)

L^{Me etpy 51}

;

2.849

2.85

Å,

Å

[(HMe₂L^iPr)Cu(µ-O)₂Cu(L^TMPD)]CF₃SO₃;³²

2

(EXAFS), L ) L^TMPD) N,N,N′,N′-tetramethyl-1,3-pro-

panediamine). This elongation of the Cu‚‚‚Cu distance

reflects a greater degree of steric interaction within the O

core, likely due to the larger bite angle of such ligands. The

thermally stable O reported by Hofmann and colleagues²⁷

also exhibits significantly divergent metrical features relative

to 3b (Cu-O 1.86 Å, Cu‚‚‚Cu 2.906 Å). This charge-neutral

complex is based on an anionic, strongly basic iminophos-

phanamide ligand that forms a four-membered chelate and

thus represents a quite different geometric and electronic

coordination profile.

(91) Yoon, T. P.; Jacobsen, E. N. Science 2003, 299, 1691-1693.

(92) Rondelez, Y.; Rager, M.-N.; Duprat, A.; Reinaud, O. J. Am. Chem.

Soc. 2002, 124, 1334-1340.

The observed thermal lability, short Cu‚‚‚Cu and Cu-O

distances, and absence of close contacts between the bridging

7354 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

Scheme 3. Resonance Raman-Active A_g-Symmetric Principal

Vibrational Modes of O Complexes⁶⁶

oxides and the counteranions clearly differentiate [3b‚

4CH₂Cl₂] from the superficially similar bis(µ-hydroxo)di-

copper(II) complexes _RR,RR1e and _RR,SS1e. Yet the +3 oxida-

tion state of the Cu centers in 3b and other Os cannot be

deduced from their crystal structures alone, since no sub-

stantially disparate Cu-O bond distances are present to allow

for an internal comparison^42,64and since similarly short

Cu-O bonds are observed, albeit rarely, among authentic

Cu(II) compounds.^15,93,94Moreover, in complexes that exhibit

substantial metal-ligand covalency, formally assigned oxi-

dation states are of dubious significance unless based upon

a direct probe of the electron density at the metal center.^64,95

In the case of 3b, a comparative Cu K-edge XAS study using

the structurally related _RR,RR1e as a reference unambiguously

indicated a formal +3 oxidation state for both Cu atoms.^64,69

The spectroscopically analogous species 1b-9b, accordingly,

are all best described as bis(µ-oxo)dicopper(III) complexes.

The XAS technique remains the sole experimental basis for

definitive assignment of valence in this class of compounds.

Spectroscopy. Together with the observed diamagnetism,

the intense LMCT and O₂isotope-sensitive rR features

exhibited by 1b-9b are diagnostic for Os,⁶⁶as the charac-

teristic maxima for other 2:1 Cu-O₂chromophores appear

at different energies (P: LMCT ∼360 and ∼540 nm, rR

ν_O-O∼720-760 cm^-1; trans-µ-1,2-peroxo: LMCT ∼520-

550 and ∼600-630 nm, rR ν_O-O∼830 cm^-1and ν_Cu-O

∼550 cm^-1).^1,96The intensity of the LMCT absorptions (ꢀ

) 10 000-30 000 M^-1cm^-1) indicates substantial covalency

in the Cu(III)-O^2-interaction. On the basis of electronic

structure calculations and rR studies, the ∼300 nm band has

been assigned to the oxygen in-plane π_σ* f Cu(III) d_xy

LMCT and the ∼400 nm band to the oxygen σ_u* f Cu(III)

d_xyLMCT.⁹⁷

The rR spectra of several Os reported herein were the

subject of recent detailed theoretical and experimental studies

in which most of the prominent features were assigned (Table

3).^26,66These include characteristically intense fundamental

bands at ∼120 and ∼600 cm^-1that are enhanced upon

excitation into the ∼400 nm LMCT transition. The isotopic

shift of the ∼600 cm^-1feature on substitution of ¹⁶O₂with

¹⁸O₂(∆ν(¹⁶O₂-¹⁸O₂) ≈ 26 cm^-1) is consistent with its

assignment as an A_g-symmetric [Cu₂O₂] synchronous “breath-

ing” mode involving substantial elongation along the Cu-O

bond vectors (Scheme 3). The O₂isotope insensitivity of

the ∼120 cm^-1feature comports with its assignment as a

symmetrical, pairwise antisynchronous “flexing” mode,

which primarily involves changes in the internal angles of

the [Cu₂O₂] rhomb (Scheme 3).⁶⁶Analogous rR features have

been observed for other metal bis(µ-oxo) M₂O₂cores (M )

Ni, Co, Fe, and Mn).²⁶

for examining spectroscopic trends because the corresponding

Os contain exclusively homochiral ligands locked into a

single backbone conformation. Systematic increases in

N-alkyl substituent steric demands within the series produce

subtle but significant corresponding bathochromic shifts in

both the higher- and lower-energy LMCT bands of the O

chromophore (1b: 301, 399 nm; 2b: 305, 402 nm; 3b: 313,

408 nm; 4b: 319, 413 nm). We attribute this general

phenomenon primarily to sterically induced perturbations of

the O core geometry: as peripheral ligand bulk increases,

the [Cu₂O₂] rhomb elongates, increasing the Cu‚‚‚Cu separa-

tion while decreasing the O‚‚‚O separation. Since the donor

orbitals for both LMCT transitions are antibonding with

respect to a potential O-O interaction, a shortening of the

O‚‚‚O distance should raise their energies relative to that of

the Cu(III) d_xyacceptor MO. This trend is borne out among

[9]ane-based Os (8b: Cu‚‚‚Cu 2.77 Å, LMCT 307 and 412

nm; [(L^{iPr dtne})Cu₂(µ-O)₂]²⁺: Cu‚‚‚Cu 2.78 Å, LMCT 316

4

and 414 nm;¹⁷[(L^{Bn TACN})₂Cu₂(µ-O)₂]²⁺: Cu‚‚‚Cu 2.79 Å,

3

LMCT 318 and 430 nm¹⁶).

Twisting of the ligand N-Cu-N chelate planes relative

to the [Cu₂O₂] plane could also influence the LMCT

transition energies, since geometric distortions that reduce

the overlap of the amine nitrogen σ-donor orbitals with the

primarily Cu(III) d_xyLUMO should lower the energy of the

latter relative to the oxygen-based donors. Although it is

difficult to confirm whether the ∼8° dihedral core twist

observed in the crystal structure of [3b‚4CH₂Cl₂] persists in

solution, any such distortion arising from the chirality of the

CD backbone is likely operative to a similar degree in all of

1b-4b. Notably, samples of 4b exhibit the same LMCT

absorption maxima whether they are prepared from racemic

(_racL^TE) or enantiomerically pure (_RRL^TE) ligand, implying

that spontaneous chiral self-sorting takes place, that both

ligands induce similar twists, or that whatever degree of

N-Cu-N twisting distortion does persist in solution has

negligible effect on the LMCT energy. The lupin alkaloid

(-)-sparteine, a rigid bicyclic amine that enforces a pro-

nounced tetrahedral distortion of the Cu centers in its bis-

(µ-hydroxo)dicopper(II) dimer,²⁵forms an O that exhibits

relatively low-energy LMCT transitions (330, 427 nm). It is

not known whether this relative bathochromic shift is due

to an analogous twisting distortion, an increased Cu‚‚‚Cu

separation mandated by similar steric concerns, or the larger

chelate bite angle (at parity of Cu-N distances) that

differentiates (-)-sparteine from the CD, ED, and [9]ane

ligand families employed in this work.

The homologous series of enantiopure CD ligands em-

ployed in this work (Table 1) provides an excellent platform

(93) Jacobson, R. R.; Tyekla´r, Z.; Farooq, A.; Karlin, K. D.; Liu, S.; Zubieta,

J. J. Am. Chem. Soc. 1988, 110, 3690-3692.

(94) Lee, S. C.; Holm, R. H. J. Am. Chem. Soc. 1993, 115, 11789-11798.

(95) Kau, L. S.; Spira-Solomon, D. J.; Penner-Hahn, J. E.; Hodgson, K.

O.; Solomon, E. I. J. Am. Chem. Soc. 1987, 109, 6433-6442.

(96) Solomon, E. I.; Tuczek, F.; Root, D. E.; Brown, C. A. Chem. ReV.

1994, 94, 827-856.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7355

Cole et al.

Scheme 4. Presumed General Mechanism for the Formation of Os

Scheme 5. Proposed Mechanism of Formation for 4b

Although ligand-modulated changes in solvation and ion-

pairing that create an altered dielectric environment at the

O core cannot be entirely ruled out as an additional influence

on the observed LMCT energies, inspection of a representa-

tive sampling of 1b-7b synthesized in various solvents

(THF, CH₂Cl₂, acetone) and with various weakly coordinat-

stable species^74,111-113in other Cu-O₂systems that employ

sufficiently bulky ligands. Two structurally characterized

examples of charge-neutral [LCu(O₂)] complexes also ex-

ist;^15,32,74one of these reacts anaerobically with sterically

compact Cu(I) and Ni(I) species to generate nonsymmetric

mixed-ligand Os³²and a heterobimetallic bis(µ-oxo)cop-

per(III)nickel(III) complex,¹¹⁴respectively.

The substantial disparity between the formation rates of

4b (t_1/2≈ 1600 s) and 1b-3b (t_1/2e 5 s) in CH₂Cl₂is

intriguing. We suggest that the large N-alkyl substituent steric

effect on the rate of O formation (k_1b-3b/k_4bg 300; CH₂Cl₂,

193 K) arises from an increase in coordination number upon

addition of O₂during the RDS, from 3-coordinate Cu(I) to

4-coordinate or possibly 5-coordinate Cu(II) in the transition

state (Scheme 5), consistent with the strongly negative value

of ∆S^q) -210(40) J K^-1mol^-1/-50(10) cal K^-1mol^-1

measured upon oxygenation of 4a in 1:1 (v/v) THF/

CH₂Cl₂. Given the poor donor capability of the O₂molecule

prior to efficient electron transfer¹¹⁵and the comparatively

strong Cu-MeCN interaction evidenced by short (∼1.86 Å)

Cu-N_nitrilebonds in the crystal structures of 4a and 10a (vide

supra), this associative model offers a more energetically

reasonable description of the O₂binding step than does

-

ing counteranions (CF₃SO₃, ClO₄, SbF₆) suggests that

such factors are subordinate to ligand identity in this regard.⁴⁰

Formation Mechanism. Among the Os studied, only the

formation of 4b in O₂-saturated CH₂Cl₂or mixtures thereof

occurs slowly enough to facilitate monitoring by conventional

mixing techniques under the conditions employed (1 atm O₂,

193 K). The reaction is first-order in [4a] (i.e., rate )

k_obs[4a]) and is presumed to be second-order overall (i.e.,

k_obs) k[O₂]). This finding is consistent with the general

mechanism originally proposed for self-assembled Cu-O₂

species by Zuberbu¨hler,^98,99which involves initial formation

of a mononuclear [LCu(O₂)]¹⁺species (S, Scheme 4) in the

rate-determining step (RDS), followed by rapid reaction with

a second equivalent of the Cu(I) precursor to generate a

binuclear “peroxide-level” intermediate (X, Scheme 4).¹⁰⁰

Similar first-order formation kinetics have been reported for

Os based on the mononucleating L^{iPr TACN}and dinucleating

3

L^{iPr dtne}ligands.^17,44Neither S nor X is observable spectro-

4

(106) Comba, P.; Kerscher, M.; Merz, M.; Mu¨ller, V.; Pritzkow, H.;

Remenyi, R.; Schiek, W.; Xiong, Y. Chem. Eur. J. 2002, 8, 5750-

5760.

(107) Weitzer, M.; Schatz, M.; Hampel, F.; Heinemann, F. W.; Schindler,

S. J. Chem. Soc., Dalton Trans. 2002, 5, 686-694.

(108) Schatz, M.; Leibold, M.; Foxon, S. P.; Weitzer, M.; Heinemann, F.

W.; Hampel, F.; Walter, O.; Schindler, S. Dalton Trans. 2003, 1480-

1487.

scopically in these cases because fast subsequent steps

preclude their accumulation. However, such 1:1 inter-

mediatesswhich may be formally described as either su-

peroxocopper(II) or peroxocopper(III) speciesshave been

observed spectroscopically as both transient^{10,98,101-110}and

(109) Zhang, C.; Kaderli, S.; Costas, M.; Kim, E.; Neuhold, Y.; Karlin,

K.; Zuberbu¨hler, A. D. Inorg. Chem. 2003, 42, 1807-1824.

(110) Weitzer, M.; Schindler, S.; Brehm, G.; Schneider, S.; Hormann, E.;

Jung, B.; Kaderli, S.; Zuberbu¨hler, A. D. Inorg. Chem. 2003, 42,

1800-1806.

(97) Orbital symmetries and bonding/antibonding designations are given

relative to a hypothetical O-O bond vector.⁶⁶

(98) Karlin, K. D.; Wei, N.; Jung, B.; Kaderli, S.; Zuberbu¨hler, A. D. J.

Am. Chem. Soc. 1991, 113, 5868-5870.

(99) Zuberbu¨hler, A. D. In Bioinorganic Chemistry of Copper; Karlin, K.

D., Tyekla´r, Z., Eds.; Chapman and Hall: New York, 1993.

(100) It is possible that intermediate X (Scheme 3) represents a multistep

sequence involving different peroxide-level Cu(II) intermediates, not

all of which are bona fide P species.

(111) Chen, P.; Root, D. E.; Campochiaro, C.; Fujisawa, K.; Solomon, E.

I. J. Am. Chem. Soc. 2003, 125, 466-474.

(112) Schatz, M.; Raab, V.; Foxon, S. P.; Brehm, G.; Schneider, S.; Reiher,

M.; Holthausen, M. C.; Sundermeyer, J.; Schindler, S. Angew. Chem.,

Int. Ed. 2004, 43, 4360-4363.

(101) Karlin, K. D.; Wei, N.; Jung, B.; Kaderli, S.; Niklaus, P.; Zuberbu¨hler,

A. D. J. Am. Chem. Soc. 1993, 115, 9506-9514.

(113) Komiyama, K.; Furutachi, H.; Nagatomo, S.; Hashimoto, A.; Hayashi,

H.; Fujinami, S.; Suzuki, M.; Kitagawa, T. Bull. Chem. Soc. Jpn.

2004, 77, 59-72.

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

Kaderli, S.; Jung, B.; Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1995,

117, 12498-12513.

(114) Aboelella, N. W.; York, J. T.; Reynolds, A. M.; Fujita, K.; Kinsinger,

C. R.; Cramer, C. J.; Riordan, C. G.; Tolman, W. B. Chem. Commun.

2004, 1716-1717.

(103) Karlin, K. D.; Lee, D. H.; Kaderli, S.; Zuberbu¨hler, A. D. Chem.

Commun. 1997, 475-476.

(115) Given that cyclic voltammograms of the Cu(I) complexes 1a-4a

and 10a in anaerobic CH₂Cl₂each exhibit a quasi-reversible oxidation

wave with E ) 0.9-1.2 V vs SCE, we contend that outer-sphere

electron transfer from Cu(I) to O₂is not a practicable initial step in

the O formation mechanism.

(104) Becker, M.; Heinemann, F. W.; Schindler, S. Chem. Eur. J. 1999, 5,

3124-3129.

(105) Schatz, M.; Becker, M.; Walter, O.; Liehr, G.; Schindler, S. Inorg.

Chim. Acta 2001, 324, 173-179.

7356 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

dissociative MeCN substitution, which would require a

nonlinear, thermodynamically unfavorable, 2-coordinate Cu-

(I) intermediate. Bulkier N-alkyl substituents should inhibit

associative expansion of the coordination sphere, resulting

in slower oxygenation rates.

AcFc (750 mV) within 6 h at 193 K. All of the Os surveyed

were capable of oxidizing alkyl and aryl thiols to the

corresponding disulfides and DBP to a coupled biaryl in near-

quantitative yields at 193 K under N₂sa net 2e^-, 2H⁺

process. Noncoordinating hydrocarbon substrates with weak

C-H bonds are not appreciably oxidized by any of 1b-7b

at 193 K under the conditions employed in this study,

although such reactivity has recently been documented for

certain Os²⁸and O-containing isomeric mixtures,^18,119as well

as for other anionically stabilized Cu(III) complexes.¹²⁰

Consistent with this mechanism, the most sterically

demanding CD-based Cu(I) complex examined, 10a, does

not form an optically detectable Cu-O₂species at 193 K,

and it appears to be air-stable even at ambient temperature.

Its ostensible O₂inertness cannot be a kinetic phenomenon,

since the auxiliary MeCN ligand undergoes rapid exchange

on the NMR time scale and is readily displaced by CO_(g).

Space-filling molecular modeling based on the crystal

structure of [10a‚2CH₂Cl₂] indicates sufficient space for

associative substitution of the MeCN ligand by a dioxygen

molecule, even in a “side-on” η²orientation. The failure to

observe spectroscopic features attributable to a 1:1 Cu/O₂

adduct suggests that formation of such an intermediate is

thermodynamically unfavorable relative to 10a and O₂. It is

important to note, however, that even significant equilibrium

concentrations of a transient Cu(II)-superoxide species could

remain undetectable via optical spectroscopy: a recent

Os 1b-7b are limited in their reactivity toward alcohols.

The oxidatively susceptible benzyl and cinnamyl alcohols

are converted stoichiometrically to the corresponding con-

jugated aldehydes only when first activated by deprotonation,

and saturated primary alcohols are unreactive even when

converted to lithium alkoxide salts. In our report of [(L^TMPD)₂-

Cu₂(µ-O)₂](CF₃SO₃)₂, a similar O that does oxidize primary

alcohols directly at 233 K, we attributed the enhanced

reactivity to structural and electronic differences: the weak

anionic axial association indicated by EXAFS connotes more

facile substrate access to the [Cu₂O₂]²⁺core relative to the

Os reported in this study.²²Altered geometry due to increased

chelate ring size (i.e., a six-membered chelate for L^TMPDvs

a five-membered chelate for CD-, ED-, and [9]ane-based Os)

may also be a contributing factor: in electrochemical

studies,^81,121,122distortions from ideal square-planar coordina-

tion geometry have been shown to increase the potential of

the Cu(III/II) couple.

spectroscopic study of [Cu(O₂){L^{HB(3-Ad-5-iPrpz)}}] (Ad )

3

1-adamantyl) revealed no molar absorptivities greater than

400 M^-1cm^-1over the entirety of the UV-vis and NIR

regions.¹¹¹If a 1:1 intermediate does form upon reaction of

10a with O₂, it may be prevented from further aggregation

by interligand steric interference due to interlocking edge-

to-face π-stacking interactions between the four forward-

projecting N-benzyl substituents (Figure 2). The strong

upfield shift of the doublet resonance (δ ) 6.55) correspond-

ing to the relevant ortho aryl protons in the ¹H NMR

spectrum of 10a suggests that this π-stacked conformer

persists in solution.

The observed formation rate of 4b varies dramatically with

solvent choice. At 193 K and 1 atm O₂pressure, oxygenation

of 4a proceeds at least 300 times faster in THF or acetone

than in CH₂Cl₂sthat is, at a rate indistinguishable from the

lower bound documented for 1b-3b via conventional mixing

techniques. The greater relative solubility of O₂in THF or

acetone,^43,116while certainly a contributing factor, is insuf-

ficient in magnitude to account for such a large discrepancy

in rates. As an inner-sphere process,¹¹⁵O₂activation at Cu

is greatly dependent upon the accessibility of the metal ion.

The observed solvent effect may thus be attributed reasonably

to subtle differences in ligand conformation, solvent associa-

tion, or ion-pairing proclivity that modulate the accessibility

of the metal.

Kinetic Aspects of Thermal Decomposition. Although

stable at low temperatures, Os 1b-9b decompose upon

warming through a mechanism primarily involving oxygen-

ation of the ligand, as previously described for 1b, 3b, 4b,

and 8b,²¹as well as other Os.²The thermal decomposition

of each O (Scheme 6) is fitted readily to a first-order kinetic

model and entails 2e^-sand sometimes 4e^-soxidation of an

N-alkyl pendant group. Our interest in decomposition studies

was motivated by broad disparities in thermal sensitivity and

ligand product distribution among these structurally homolo-

gous Os, which are distinguished solely by the nature of their

N-alkyl substituents. For example, at 263 K, 1b is ∼20 times

more thermally stable than 3b and ∼170 times more

thermally stable than 4b. Notably, the thermal decay process

appears comparatively uninfluenced by solvent and coun-

teranion effects, as is evidenced by the equivalent values of

∆H^qand ∆S^qobtained from Eyring fits for the decay of [5b‚

(CF₃SO₃)₂] in THF and CH₂Cl₂and for [5b‚(CF₃SO₃)₂] and

[5b‚(SbF₆)₂] in THF.⁵⁹

Reactivity. Toward exogenous substrates the O core

exhibits a pattern of reactivity (Scheme 2) best characterized

as that of a mild 2e^-oxidant, often acting via sequential

1e^-hydrogen atom abstractions. The characteristic intense

LMCT bands of 1b-5b are fully quenched by spectropho-

tometric titration with 2 equiv of either Me₁₀Fc (E_1/2) -100

mV vs SCE)¹¹⁷or Fc (480 mV)^117,118but are unaffected by

(118) Chang, D.; Malinski, T.; Ulman, A.; Kadish, K. Inorg. Chem. 1984,

23, 817-824.

(119) The report by Karlin et al. of such reactivity at -80 °C (J. Am. Chem.

Soc. 1998, 120, 12960-12961) has been reevaluated. See Shearer,

J.; Xin Zhang, C.; Zakharov, L. N.; Rheingold, A. L.; Karlin, K. D.

J. Am. Chem. Soc. 2005, 127, 5469-5483.

(120) Lockwood, M. A.; Blubaugh, T. J.; Collier, A. M.; Lovell, S.; Mayer,

J. M. Angew. Chem., Int. Ed. 1999, 38, 225-227.

(121) Hanss, J.; Beckmann, A.; Kru¨ger, H.-J. Eur. J. Inorg. Chem. 1999,

163-172.

(116) Jung, B.; Karlin, K. D.; Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1996,

118, 3763-3764.

(117) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877-910.

(122) Ribas, X.; Jackson, D. A.; Donnadieu, B.; Mahia, J.; Parella, T.; Xifra,

R.; Hedman, B.; Hodgson, K. O.; Llobet, A.; Stack, T. D. P. Angew.

Chem., Int. Ed. 2002, 41, 2991-2994.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7357

Cole et al.

Scheme 6. Proposed Unimolecular (2e^-) and Bimolecular (4e^-) Decomposition Pathways (CD-Based Os Shown)

Scheme 7. Selectivity during Thermal Decay of 3b

KIE studies were undertaken in order to investigate the

thermal decay process further. The comparison of 1b with

d₂₄-1b presents a straightforward and informative picture of

the thermal decay process, as both Os give rise to analogous

ligand products that imply a common decomposition path-

ethylated (d₄-L^ME-Et) and de-methylated (d₄-L^ME-Me) prod-

ucts in a ∼1:1 ratio (Scheme 7). The NC^R-D bond

dissociation energy (BDE) in d₈-3b is sufficiently elevated

relative to h₈-3b to activate a competitive decay pathway

involving rate-limiting oxidation of a N-methyl group.

D

way. The observed primary KIE of k_obs^H/k_obs≈ 8 for this

Rate-determining NC^R-H bond activation is the norm

among other reported Os, some of which exhibit much larger,

nonclassical KIEs attributed to quantum-mechanical proton

tunneling.^24,45C-H bond cleavage could theoretically pro-

ceed via hydride transfer; however, on the basis of Hammett

studies of Os derived from tridentate ligands incorporating

para-substituted N-benzyl groups, the hydride model has

generally been rejected in favor of either a hydrogen

abstraction or a nonsynchronous concerted mechanism

(Scheme 6, paths (a) and (b), respectively).^2,24,45,133In the

case of 1b and 3b, the ∼2-fold decrease of the thermal

decomposition rate in the presence of O₂(1 atm) suggests

involvement of radical species.

reaction at 293 K¹²³is in accord with the classically predicted

value of ∼6.5.¹²⁴This observation provides evidence for

cleavage of a methyl C-H bond during the RDS (Scheme

6). By contrast, selective deuteration of the methylene units

of the N-ethyl substituents in 3b yields an extrapolated

D

primary KIE of k_obs^H/k_obs≈ 3 at 293 K. Ligand product

analyses (vide infra) offer a plausible explanation for the

apparent attenuation of the KIE relative to 1b: whereas

h₈-3b generates only the de-ethylated (L^ME-Et) product upon

thermal decomposition and workup, d₈-3b yields both de-

(123) The primary KIE for the thermal decay of 1b/d₂₄-1b might be even

larger, since only a 40% yield (i.e., 80% theoretical) of the L^TM-Me

product is observed.

(124) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic

Chemistry, 3rd ed.; HarperCollins: New York, 1987.

The nature of the N-alkyl substituents dramatically affects

the thermal stability of Os: at ambient temperature, decom-

7358 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

position half-lives (t_1/2) for the reported compounds span

almost 3 orders of magnitude (d₂₄-1b, t_1/2) 1600 s; 4b,

extrapolated t_1/2≈ 2 s; 293 K). The observed 20-fold

difference in decay rates between 1b and 3b (Table 3) is

predominantly attributable to the disparity in their activation

mechanism outlined in Scheme 6. In several instances,

however, the observed yields of such products are insufficient

to account for all oxidizing equivalents on a 2e^-/O basis.

For example, upon decomposition and workup, 1b and 3b

generate the L^TM-Meand L^ME-Etproducts in 60% and 80%

of theoretical yields, respectively, whereas 4b produces the

L^TE-Etproduct in near-quantitative (>95%) yield. Yields of

L^X-Rproducts for the ED-based Os 5b-7b are likewise

substantially substoichiometric despite excellent (>90%)

ligand mass recovery during analysis. These reproducible

observations raise the prospect of a competing exogenous

reductant, although comparative thermal decomposition

studies conducted in deuterated solvents have revealed no

significant solvent KIE, and no other decay products were

enthalpies: ∆∆H^q

) 7 kJ mol^-1/1.8 kcal mol^-1;

1b-3b

∆∆S^q

) -3 J K^-1mol^-1/-0.8 cal K^-1mol^-1. This

1b-3b

corresponds well with the 1.9 kcal mol^-1difference in the

reported molar NC^R-H BDEs of trimethylamine (92.6 ( 1

kcal mol^-1) and triethylamine (90.7 ( 0.4 kcal mol^-1).¹²⁵

By contrast, the 8-fold difference between the decay rates

of 3b and 4b arises almost exclusively from the entropic

terms: ∆∆H^q

) -0.5 kJ mol^-1/-0.1 kcal mol^-1;

3b-4b

∆∆S^q

) - 21 J K^-1mol^-1/-5 cal K^-1mol^-1. The

3b-4b

negligible ∆∆H^qis consistent with a mechanism in which

both 3b and 4b decompose solely via NC^R-H bond cleavage

of N-ethyl substituents having equivalent BDEs; as already

noted, ligand degradation product analysis confirms that the

N-ethyl substituent is indeed selectively oxidized in the case

of 3b. The significant disparity in activation entropy terms

may reflect the 2-fold relative increase in the number of

accessible N-ethyl groups, as well as greater preorganization

of these substituents owing to increased steric congestion

about the O core in 4b.

detected by GC-MS or H NMR analysis. The fate of the

remaining oxidizing equivalents is currently under investiga-

tion.

1

The appearance of a 4e^-oxidized formamide L^TM+O

product in up to 20% theoretical yield at higher concentra-

tions of 1b ([Cu] ) 2-5 mM, 263 K, THF) strongly suggests

the involvement of an aminol intermediate (Scheme 6),

subsequent oxidation of which leads to the formamide

product.¹²⁶Analogous 4e^--oxidized species are the major

ligand products in the thermal decomposition of the [9]ane-

based Os 8b and 9b, and similar products have been reported

for other [9]ane systems.⁵²As proposed by Itoh and Tol-

man,^24,45the rate-limiting oxygen atom insertion into the

NC^R-H bond generates an aminol intermediate. This un-

stable species, the major ligand product in the chiefly

unimolecular decay of CD- and ED-based Os, converts to

the dealkylated L^X-Rend product upon workup, the detached

N-alkyl substituent having been oxidized to the corresponding

aldehyde. The formamide product is proposed to arise from

intermolecular overoxidation of the aminol intermediate by

a second equivalent of O in a bimolecular process that

becomes significant at higher concentrations of 1b and

predominates generally in the cases of 8b and 9b. Evidence

for such a processsformally, a net oxygen atom transfer from

the Cu₂O₂core to the ligandshas been given previously via

¹⁸O₂labeling studies in which the label was transferred from

Comparison of Eyring plots⁵⁹for the thermal decomposi-

tion of Os based on the permethylated facial-capping

tridentate ligands L^{Me TACN}and L^{Me ODACN}(8b and 9b,

respectively) gives insight into the effect of ligand denticitys

more specifically, the presence or absence of axial donorss

on thermal stability. The superior thermal robustness of 9b

derives mainly from enthalpic considerations that may reflect

the substantially weaker donor strength of the axially bonded

3

2

ether oxygen atom of L^{Me ODACN}(∆∆H^q_9b-8b) 9.0 kJ mol^-1/

2

2.2 kcal mol^-1; ∆∆S^q

) -12 J K^-1mol^-1/-2 cal K^-1

9b-8b

mol^-1). The [N₂O] donor set in L^{Me ODACN}generates an O

2

that is more thermally stable than its [N₃] L^{Me TACN}analogue

3

by an order of magnitude, suggesting that the stronger axial

σ donation in the latter destabilizes its O core. Indeed, 1b,

which lacks an integral axial ligand donor altogether, enjoys

a 100-fold advantage in thermal stability over 8b, at parity

of N-alkyl substituents (viz. methyl). Coordination of axial

ligands is known to increase the reduction potential of

Cu(III) complexes generally,^79,120although the particular

comparison of 1b with [9]ane-based Os is complicated by

the fluxional axial-equatorial ligand exchange documented

for some members of the latter family.¹⁶That the disparity

[(L^{Bn TACN})₂Cu₂(µ-¹⁸O)₂]²⁺to the carbonyl product.⁴⁵

3

C-H Bond Cleavage. The selectivity of intraligand C-H

bond cleavage in Os depends on the nature of the N-alkyl

ligand substituents and the degree to which they are

physically accessible to the [Cu₂O₂]²⁺core. A comparison

of the ligand products from the decay of 3b and 6b

underscores the importance of N-alkyl group positioning as

well as NC^R-H BDE in steering the thermal decomposition

process: whereas 3b generates exclusively L^ME-Et, 6b yields

in activation entropies (∆∆S^q

) -24 J K^-1mol^-1/-6

1b-8b

cal K^-1mol^-1) is counterstatistical with regard to the number

of available methyl hydrogens tends to support the relevance

of other factors. The large quantum tunneling-induced

primary KIEs observed in some [9]ane-based systems⁴⁵also

suggest a fundamentally different juxtaposition of N-alkyl

substituents to the [Cu₂O₂]²⁺core.

both L^{(MeEt) ED-Et}and L^{(MeEt) ED-Me}products in comparable

quantities but at a much lower combined yield (25% of

theoretical vs 80% for 3b). The absence of selectivity in the

latter case is attributed to the increased conformational

freedom of the ED ligand backbone, which affords the

N-methyl groups more competitive access to the O core. The

marked difference in decay rates between ED- and CD-based

2

Ligand Degradation Product Analysis. The observed

distribution of N-dealkylated ligand products within the CD

and ED series (Table 3) is consistent with the general

(125) Dombrowski, G. W. D., J. P.; Farid, S.; Goodman, J. L.; Gould, I.

(126) No corresponding 4e^--oxidized product was detected for the thermal

R. J. Org. Chem. 1999, 64, 427-431.

decomposition of 3b at Cu concentrations up to 50 mM.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7359

Cole et al.

Os is likewise consistent with this analysis: at 263 K, 6b

decays 8 times more slowly than the more rigid 3b because

the greater backbone flexibility of the former allows the

N-alkyl substituents to occupy a wider range of positions

less conducive to NC^R-H bond cleavage. Similarly, 4b

decays ∼5 times faster than 5b despite a larger entropic

barrier attributable to its relatively more constrained transition

state [∆∆H^q_4b-5b) 11.0 kJ mol^-1/-2.7 kcal mol^-1; ∆∆S^q_4b-5b

) -30 J K^-1mol^-1/-7 cal K^-1mol^-1]. At parity of N-alkyl

substituents, the comparatively low overall yields of N-

dealkylated ligand products among the ED-based Os are also

consistent with the slower observed decay rates, assuming

that such substoichiometric yields are due in general to a

competing process involving an exogenous reductant (vide

supra).

the [Cu₂O₂] least-squares plane in the solid state, and they

possess a much greater degree of rotational freedom that

should further reduce their exposure to the core.

Although the crystal structure of [3b‚4CH₂Cl₂] certainly

provides insight into the types of ligand-core interactions

that may influence the NC^R-H bond cleavage mechanism,

the limitations of such parallels between solid-state structural

data and the behavior of solution species become evident in

considering the thermal decay of d₈-3b. In this case, attack

of the L^MEN-methyl substituents becomes competitive with

de-ethylation merely upon selective deuteration of the

N-methylene units, implying that relative differences in

C-H(D) BDE exert the deciding influence in this system.

More generally, this surprising result suggests that it is in

fact a complex interplay of factors including relative NC^R-H

BDEs, thermally induced vibrational distortions, sterically

imposed constraints on ligand substituent conformations, and

statistical factors that governs the ligand product selectivity

observed in each particular instance.

The importance of core accessibility at parity of backbone

structure is further exemplified by comparison of the isomeric

ED-based Os 6b and 7b. Whereas the former yields a

mixture of ligand products (vide supra), the latter decays 20

times faster and generates the L^{Me Et ED-Et}product exclu-

2

sively. The L^{(MeEt) ED}ligand in 6b can achieve greater

2

Conclusions

conformational flexibility by virtue of the symmetric distri-

bution of its bulk throughout the complex; hence, both

N-ethyl and N-methyl groups are susceptible to oxidation.

By contrast, the conformational freedom of the nonsymmetric

We have synthesized and characterized a family of simple

vicinal PDL-Cu(I) precursors and their corresponding

thermally labile Os, as well as related tridentate Os based

on the [9]ane skeleton. The essential spectroscopic features

of these Os are consistent over a large number of peralkylated

di- and triamine ligands, and include intense LMCT bands

at ∼300 and ∼400 nm, an intense O₂isotope-sensitive band

in the rR spectrum at ∼600 cm^-1(∆ν(¹⁶O₂-¹⁸O₂) ≈ 25

cm^-1), characteristically short Cu-O (∼1.80 Å) and

Cu‚‚‚Cu (∼2.75 Å) distances, and an XAS edge profile

typical of trivalent Cu. The well-ordered crystal structure of

[3b‚4CH₂Cl₂] provides one of the most metrically compact

examples of an O core yet reportedsa result we ascribe to

the minimization of intramolecular steric interactions. The

square-planar geometry and bridging dianionic oxide ligands

in the [Cu₂(µ-O)₂]²⁺core provide a highly stabilizing

coordination environment for Cu(III). The mechanism of

formation of these Os is best described as rate-determining

generation of a Cu(II)-superoxide species followed by

reaction with another molecule of the Cu(I) precursor; the

rate of formation may be influenced by ligand steric factors

as well as solvent effects.

L^{Me Et ED}ligand in 7b is limited by steric congestion of its

geminal N-ethyl substituents, which may be thus constrained

to positions more accessible to the [Cu₂O₂]²⁺core. These

steric considerations, in conjunction with the 1.9 kcal mol^-1

disparity in BDE between N-methyl and N-methylene C-H

bonds, are sufficient to produce the observed selectivity.

Upon inspection of the crystal structure of [3b‚4CH₂Cl₂],

it is possible to make similar inferences regarding the

importance of core accessibility in the thermal decay of 3b.

In explaining its dealkylation selectivity, one might assign

equal relevance to both the disparity in NC^R-H BDEs and

the steric constraints imposed on the N-ethyl substituents by

the rigid L^MEligand backbone. In the solid state, at least,

the four inward-directed methylene hydrogens of these

N-ethyl groups participate in C-H‚‚‚O hydrogen bonding

interactions that place them as little as 2.45 Å away from

the bridging oxides and as close as 0.62 Å to the [Cu₂O₂]

least-squares plane (C-H bond lengths fixed to 1.1 Å). Since

the LUMO + 1 orbital that is presumed to be the main

acceptor MO for the NC^R-H bond cleavage process is a

predominantly metal-based combination of the in-plane Cu

d_xyand oxygen σ* orbitals,^45,66,127C-H bonds that approach

coplanarity with the [Cu₂O₂] rhomb should achieve better

overlap with this orbital, and thus exhibit greater susceptibil-

ity to NC^R-H bond cleavage. Moreover, space-filling

molecular models of the [3b‚4CH₂Cl₂] crystal structure

suggest that the geared configuration of the N-alkyl substit-

uents restricts free rotation of the N-ethyl groups about the

C-N bond, effectively locking them into an orientation that

is conducive to NC^R-H bond cleavage. The N-methyl

hydrogens, by comparison, are at least 2.29 Å distant from

2

The chemical reactivity profile of Os is that of a mild

oxidant that reacts with suitable exogenous substrates, or with

its own ligands through a NC^R-H bond cleavage mechanism.

Careful analysis of the dealkylated ligand decay products

provides further support for the proposed mechanism and

suggests that the relative thermal stabilities of the reported

Os are dictated not solely by the strength of ligand substituent

NC^R-H bonds but also by their position and degree of

conformational freedom relative to the [Cu₂O₂]²⁺core. This

latter attribute is, in turn, very much dependent on ligand

geometry and conformational rigidity. The same factors also

have a marked effect on the internal selectivity of ligand

oxidation during thermal decay, as ligand substituents that

have weak NC^R-H bonds constrained in a comparatively

(127) Pidcock, E.; Obias, H. V.; Zhang, C. X.; Karlin, K. D.; Solomon, E.

I. J. Am. Chem. Soc. 1998, 120, 7841-7847.

7360 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

with a variable optical path length (0.01, 0.1, and 1.0 cm) and

custom air-free sample cells (ChemGlass). X-ray crystal data were

collected on Enraf-Nonius CAD4 and Bruker-Siemens SMART

diffractometers and were refined using teXsan 1.04 and Crystal-

Structure 2.0 from Molecular Structure Corp., as well as the

SHELX-97 routine of George Sheldrick. The refined models were

checked and evaluated using PLATON for Windows;¹²⁹ORTEP

figures were generated via Ortep-3 for Windows¹³⁰and POV-ray

3.5 for Windows. ¹H and ¹³C NMR spectra (including low-

temperature magnetic susceptibility measurements)⁶⁵were per-

formed on a Varian XL 400 MHz instrument. The derived molar

susceptibilities were corrected for the diamagnetic contributions of

the ligands and solvent, which were estimated using Pascal’s

constants.¹³¹Gas chromatography/mass spectrometry (GC-MS) was

performed on an HP 5970 gas chromatograph with an in-line MS

detector, using an AT-5 (Alltech) column. Ligand purity was

assessed on an HP 5890 gas chromatograph with an HP-DB1

column and an FID detector. Multivariate analysis of UV-vis

kinetic data was performed using SPECFIT.¹³²Cyclic voltammo-

grams of Cu(I) species were recorded at a scan rate of 100 mV s^-1

on a CV-50W voltammetric analyzer from Bioanalytical Systems

using a Pt wire counter electrode, a Pt button working electrode,

and a Ag/AgCl_(s)reference electrode; conditions were as follows:

1.0 mM analyte concentration, N₂atmosphere, CH₂Cl₂solvent, 0.1

M tetra-n-butylammonium perchlorate supporting electrolyte. Solu-

tion FTIR spectra were collected on an ATI Mattson Infinity Series

spectrophotometer using an air-free CaF₂liquid flow cell. Elemental

analyses were performed by Desert Analytics (Tucson, AZ);

samples were dried under vacuum overnight prior to analysis and

handled anaerobically.

Syntheses of Ligands. trans-(1R,2R)-Cyclohexanediamine

(CD) and (()-trans-1,2-Cyclohexanediamine (_racCD). Enantiopure

CD tartrate (25 g, 99 mmol) or an equivalent amount of _racCD

oxalate was dissolved in the minimum amount of 10 M KOH. KOH

pellets (20 g) and Et₂O (500 mL) were added, and the mixture was

stirred vigorously for 24 h. The resulting solution was decanted,

and the solid residue was pulverized and washed with 50 mL

portions of Et₂O. To minimize the formation of insoluble carbonate

salts, exposure to atmospheric moisture and CO₂was limited by

appropriate use of N₂blanketing. The solvent was removed from

the combined organic phases in vacuo at 293 K, and the residue

was distilled in vacuo from CaH_2(s)to yield 8.9 g of the free base

(80%). ¹H NMR (CDCl₃, 400 MHz): δ 1.05-1.2 (m, 2H); 1.25-

1.35 (br m, 6H); 1.7 (m, 2H); 1.85 (d, 2H); 2.25 (m, 2H).

N,N,N′,N′-Tetramethyl-trans-(1R,2R)-cyclohexanediamine

(L^TM).⁵⁶CD (3.0 g, 26.3 mmol) was added to 2.4 g of HCOOH

(52.6 mmol) and 4.2 g of HCHO (37% aqueous solution, 52.6

mmol) and the mixture refluxed for 24 h. The solution was cooled

to 298 K, made basic with 2 M aqueous NaOH, and extracted with

Et₂O (4 × 25 mL). The resulting solution was dried over K₂CO_3(s)

and the solvent removed by rotary evaporation. The resulting brown

oil was distilled in vacuo from CaH_2(s)to give 3.1 g of a colorless

oil (70%). _racL^TMwas prepared from _racCD in an identical manner.

¹H NMR (CDCl₃, 400 MHz): δ 1.0-1.2 (m, 4H); 1.7 (m, 2H);

1.77 (m, 2H); 2.23 (m, 12H); 2.55 (m, 2H). GC-MS: m/z 170,

M⁺.

favorable position relative to the O core are preferentially

attacked. Os based on bidentate ligands lack the destabilizing

influence of an axial σ donor for Cu(III), and so possess

greater thermal stability at parity of N-alkyl substituents than

those based on comparable tridentate ligands. With a half-

life of almost 30 min at 293 K, the perdeuterated d₂₄-1b is

the most thermally stable dicationic O species currently

known.

That a bidentate ligand with the most compact N-alkyl

substituents (methyl) provides the greatest thermal stability

is as fortunate as it is fortuitous, for it preserves a potential

for reactivity with exogenous organic substrates that is

determined primarily by accessibility to the O core. Although

the products are not yet known, the observation of sub-

stoichiometric yields of dealkylated ligands in the case of

more thermally stable Os such as 1b and 6b is encouraging

insofar as a portion of their reactivity is directed outward

toward substrates or solvent rather than inward toward the

ligands. Achieving such exogenous reactivity at biologically

relevant temperatures remains a primary goal in ligand

design.

Experimental Section

Materials. CH₂Cl₂obtained from Fisher Scientific was stirred

over concentrated H₂SO₄for several days, washed thoroughly with

several portions of 1 M KOH, and then washed with deionized

H₂O until clear. After fast drying with either K₂CO_3(s)or CaCl_2(s)

,

it was allowed to stand for several days over P₄O_10(s)in the absence

of light and O₂, and was subsequently distilled under like conditions

and stored in a N₂drybox. THF and diethyl ether (Et₂O) were

obtained from Fisher Scientific and distilled under N₂from the

corresponding solution of the Na-benzophenone ketyl radical.

Spectrophotometric-grade MeCN was obtained from Fisher Sci-

entific. Spectrophotometic-grade acetone was distilled from

K₂CO_3(s). NaH was obtained as a 60% mineral oil dispersion

(Aldrich), washed repeatedly with dry hexane in a N₂drybox, dried

in vacuo, and stored under N₂. Salts [Cu(MeCN)₄](X), where X^-

) CF₃SO₃^-, ClO₄^-, or SbF₆^-, were synthesized from Cu₂O

(Aldrich) and HX (Aldrich) by a variation of the literature

method.¹²⁸Prior to use, they were recrystallized twice from MeCN/

Et₂O and stored under dry N₂. Ferrocene, acetylferrocene, and

decamethylferrocene were obtained from Aldrich and either re-

crystallized or sublimed before use. trans-(1R,2R)-Cyclohexanedi-

amine (CD) was resolved as the tartrate salt from a commercially

available diamine mixture (DuPont DCH-99) according to the

method of Whitney.⁸⁷The racemic form, (()-trans-1,2-cyclohex-

anediamine (_racCD), was separated as the oxalate salt from the same

commercial mixture in an analogous procedure. All diamine ligands

except L^TBwere purified by flash distillation in vacuo from CaH_2(s)

or NaH_(s)and stored in a N₂drybox.

Methods. Preparation and manipulation of air-sensitive Cu

complexes were carried out using standard Schlenk techniques or

in a N₂drybox (M. Braun). Since many of the Cu(I) complexes

tended to decompose to yellowish, insoluble products on standing

in CH₂Cl₂solution, these compounds were generally prepared in

situ. Low-temperature UV-vis spectra were obtained using a

Polytec PI X-DAP-06 spectrophotometer in conjunction with a

custom-designed immersible quartz fiber-optic probe (Hellma, Inc.)

(129) Spek, A. J. Appl. Crystallogr. 2003, 36, 7-13.

(130) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.

(131) Drago, R. S. Physical Methods for Chemists, 2nd ed.; Saunders:

Orlando, FL, 1992.

(132) Binstead, R.; Zuberbu¨hler, A. Spectrum Software Associates, Marl-

borough, MA.

(128) Kubas, G. J.; Monzyk, B.; Crumbliss, A. L. Inorg. Synth. 1979, 19,

90-92.

(133) Spuhler, P.; Holthausen, M. C. Angew. Chem., Int. Ed. 2003, 42,

5961-5965.

Inorganic Chemistry, Vol. 44, No. 21, 2005 7361

Cole et al.

1

d₁₂-N,N,N′,N′-Tetramethyl-trans-(1R,2R)-cyclohexane-

diamine (d₁₂-L^TM). The perdeutero-N-methyl analogue of L^TMwas

prepared as above from CD (0.470 g, 4.12 mmol) using d₂-formic

acid (2.06 g, 41.2 mmol) and d₂-formaldehyde (20% aqueous

solution in D₂O, 1.32 g, 41.2 mmol). Yield 50%. ¹H NMR (CDCl₃,

400 MHz): δ 1.0-1.2 (m, 4H); 1.7 (m, 2H); 1.77 (m, 2H); 2.55

(m, 2H). GC-MS: m/z 182, M⁺.

crystallized (14.2 g, 50%). H NMR (CDCl₃, 400 MHz): δ 2.78

(s, 3H); 3.29 (t, 2H); 4.58 (t, 2H).

N-Acetyl-trans-(1R,2R)-cyclohexanediamine. A solution of

3-acetyl-1,3-thiazolidine-2-thione (7.2 g, 45 mmol) in 50 mL of

CH₂Cl₂was added dropwise over the course of 1 h to a solution of

CD (5.0 g, 44 mmol) in CH₂Cl₂(150 mL). The resulting bright

red solution was extracted with aqueous acetate buffer (1 M NaOAc,

2 M HOAc, 3 × 50 mL). The combined aqueous fractions were

washed with CH₂Cl₂(3 × 50 mL, washings discarded), made basic

(pH > 12) by addition of saturated aqueous KOH, and extracted

with CH₂Cl₂(5 × 50 mL). Drying the combined organic fractions

over K₂CO_3(s)and removing the solvent in vacuo yielded 4.8 g of

solid product, which was contaminated with significant amounts

of unreacted CD starting material, as well as the diacetylated

byproduct. The crude solid was then purified, first by heating to

373 K under a dynamic vacuum for several hours to remove the

more volatile CD (600 mg recovered) and thereafter by sublimation

in an open Pyrex tube at 473 K. The white, microcrystalline product

N,N′-Diacetyl-trans-(1R,2R)-cyclohexanediamine. CD (3.5 g,

31.3 mmol) was refluxed with a large excess of acetic anhydride

(100 mL) in dry toluene (PhCH₃, 100 mL) for 4 h with stirring.

The white crystals of the N,N′-diacetyl derivative that formed on

cooling to 273 K were filtered, washed with successive 10 mL

portions of cold PhCH₃and Et₂O, and recrystallized from 1:1

1

(v/v) PhCH₃/Et₂O (4.5 g, 75%). H NMR (CDCl₃, 400 MHz): δ

1.28 (br m, 4H); 1.75 (m, 2H); 1.95 (s, 6H); 2.05 (m, 2H); 3.65

(m, 2H); 6.1 (br s, 2H).

N,N′-Diethyl-trans-(1R,2R)-cyclohexanediamine. N,N′-diacetyl-

trans-(1R,2R)-cyclohexanediamine (4.5 g, 23.0 mmol) was sus-

pended in 250 mL of dry, freshly distilled, deoxygenated THF and

cooled to 273 K on an ice bath. To this mixture were added 120

mL of 1.0 M BH₃in THF via cannula. After the addition was

complete, the reaction mixture was warmed to 293 K, heated at

reflux 3 h with stirring under N₂, and finally acidified by cautious

addition of excess 4 M aqueous HCl (50 mL) and refluxed for an

additional 1 h. The resulting mixture was concentrated to a volume

of ∼50 mL by boiling, made basic (pH > 12) by cautious addition

of 10 M KOH at 273 K, and extracted with Et₂O (6 × 25 mL).

The combined organic fractions were dried over K₂CO_3(s)and

concentrated on a rotary evaporator, yielding an oily residue (3.5

g, 90%). The crude diamine product was subjected to methylation

directly.

d₄-N,N′-Diethyl-trans-(1R,2R)-cyclohexanediamine. A suspen-

sion of N,N′-diacetyl-trans-(1R,2R)-cyclohexanediamine (0.700 g,

3.54 mmol) in 200 mL of dry, deoxygenated THF was cooled to

273 K. LiAlD₄(Aldrich, 0.150 g, 36 mmol) was added cautiously

under N₂, and the mixture was stirred for 15 min with attendant

warming and vigorous effervescence. The reaction mixture was

heated at reflux with stirring under a N₂atmosphere for 12 h and

then cooled to 298 K, after which the excess LiAlD₄was quenched

via cautious addition of MeOH. The solvent was removed in vacuo

and the residue treated with excess 10 M KOH. The mixture was

then extracted with Et₂O (4 × 25 mL), and the combined organic

fractions were dried over K₂CO_3(s). The solvent was again removed

by rotary evaporation, leaving an oily residue of the crude diamine

(0.400 g, 70%), which was subjected to methylation directly.

N,N′-Diethyl-N,N′-dimethyl-trans-(1R,2R)-cyclohexanedi-

amine (L^ME) and d₄-N,N′-Diethyl-N,N′-dimethyl-trans-(1R,2R)-

cyclohexanediamine (d₄-L^ME). Methylation of crude N,N′-diethyl-

trans-(1R,2R)-cyclohexanediamine and d₄-N,N′-diethyl-trans-(1R,2R)-

cyclohexanediamine was performed as described for L^TM, with

1

was resublimed once in the same manner (3.7 g, 55%). H NMR

(CDCl₃, 400 MHz): δ 1.1-1.4 (br m, 4H); 1.75 (m, 4H); 1.95 (s,

3H); 2.02 (m, 3H); 3.63 (m, 1H); 5.89 (br s, 1H).

N-Ethyl-trans-(1R,2R)-cyclohexanediamine. N-Acetyl-trans-

(1R,2R)-cyclohexanediamine (3.7 g, 23.7 mmol) was suspended in

300 mL of dry, freshly distilled, deoxygenated THF and cooled to

273 K in an ice bath. To this mixture were added 100 mL of 1 M

BH₃in THF via cannula. After refluxing for 3 h with stirring under

N₂, the solution was acidified by cautious addition of 4 M aqueous

HCl (150 mL) and refluxed for an additional 1 h. The mixture was

then concentrated to a volume of 50 mL by boiling. The concentrate

was made basic (pH > 12) by cautious addition of KOH pellets,

diluted with sufficient H₂O to dissolve precipitated salts, and

extracted with Et₂O (6 × 25 mL). The combined organic fractions

were dried over K₂CO_3(s)and concentrated on a rotary evaporator,

yielding an oily, partially crystalline residue (4 g) which was

employed in further steps without characterization.

N-Ethyl-N,N′,N′-trimethyl-trans-(1R,2R)-cyclohexanedi-

amine (L^EtMe). Crude N-ethyl-trans-(1R,2R)-cyclohexanediamine

3

was permethylated and purified in the manner described for L^TM

(3.5 g colorless oil, 19 mmol, 43% overall yield). ¹H NMR (CDCl₃,

400 MHz): δ 1.07 (t, 3H); 1.02-1.21 (br m, 4H); 1.65-1.93 (br

m, 4H); 2.20 (s, 3H); 2.30 (m, 6H); 2.40-2.60 (br m, 2H); 2.52

(q, 2H). GC-MS: 353 K (1 min), 30 K min^-1to 473 K (10 min);

t_R) 9.7 min (m/z 184, M⁺).

N,N,N′,N′-Tetraethyl-trans-(1R,2R)-cyclohexanediamine (L^TE).

CD (3.10 g, 27.7 mmol) and NaH_(s)(10 g) were suspended in dry

PhCH₃(150 mL), and the resulting slurry was refluxed under N₂.

Excess diethyl sulfate (30 mL) was added slowly by syringe, and

the mixture was refluxed for 24 h. Residual hydride was quenched

by cautious addition of MeOH (50 mL), after which the mixture

was poured into 1.0 L of a 1 M aqueous NaOH solution and

extracted with Et₂O (4 × 50 mL). The combined organic layers

were dried over K₂CO_3(s)and the solvent removed by rotary

yields of ∼70% in each case. L^ME

:

¹H NMR (CDCl₃, 400 MHz):

δ 1.05 (t, 6H); 1.05-1.20 (br m, 4H); 1.65-1.80 (m, 4H); 2.23 (s,

6H); 2.55 (q + m, 6H). GC-MS: m/z 198, M⁺. d₄-L^ME¹H NMR

:

evaporation. The brown, oily residue was distilled in vacuo from

CaH_2(s)to give 3.5 g of a clear oil (60%).

TE

L

rac

was prepared in

(CDCl₃, 400 MHz): δ 1.05 (t, 6H); 1.05-1.20 (br m, 4H); 1.65-

1.80 (m, 4H); 2.23 (s, 6H); 2.55 (q, 2H). GC-MS: m/z 202, M⁺.

3-Acetyl-1,3-thiazolidine-2-thione. Triethylamine (30 g, 296

mmol) was added to a solution of 2-mercaptothiazoline (20.0 g,

168 mmol) in 400 mL of CH₂Cl₂. Acetyl chloride (15.8 g, 201

mmol) was added gradually with stirring, and the clear yellow

solution which resulted was then washed with successive 100-mL

portions of 10% aqueous HCl and 1 M aqueous NaOH, dried over

K₂CO_3(s), and concentrated to an oil by rotary evaporation. Upon

standing in a freezer for several days at 253 K, the oil slowly

an identical manner from _racCD. ¹H NMR (CDCl₃, 400 MHz): δ 1.01

(t, 12H); 1.05-1.20 (m, 4H); 1.64-1.70 (m, 2H); 1.76-1.84 (m, 2H);

2.40-2.50 (sextet, 4H); 2.56-2.68 (m, 6H). GC-MS: m/z 226, M⁺.

N,N′-Diethyl-N,N′-dimethyl-1,2-ethanediamine (L^{(MeEt) ED}). N,N′-

2

Diethyl-1,2-ethanediamine (Aldrich) was permethylated and purified

as described for L^TM(80%). ¹H NMR (CDCl₃, 400 MHz): δ 1.01

(t, 4H); 1.9 (s, 6H); 2.0-2.2 (m, 10H). GC-MS: m/z 144, M⁺.

N,N-Diethyl-N′,N′-dimethyl-1,2-ethanediamine (L^{Me Et ED}). N,N-

2

Diethyl-1,2-ethanediamine (Aldrich) was permethylated and purified

7362 Inorganic Chemistry, Vol. 44, No. 21, 2005

Bis(µ-oxo)dicopper(III) Complexes of 1,2-Diamines

1

as described for L^TM(80%). H NMR (d₆-acetone, 400 MHz):

was added _racL^TM(170 mg, 1.00 mmol). The resulting clear solution

was layered with Et₂O and allowed to stand for several days,

whereupon large crystals (colorless rectangular blocks) of 1c slowly

formed (195 mg, 35 mmol, 70%). Since 1c dissociates (with

appearance of free ligand resonances) in those NMR solvents in

which it dissolves (d₆-DMSO, CD₃CN), no ¹H NMR spectrum was

obtained.

δ 0.70 (t, 6H); 1.8 (s, 6H); 2.00-2.07 (m, 2H); 2.18-2.25 (q+m,

6H). GC-MS: m/z 144, M⁺.

N,N,N′,N′-Tetraethyl-1,2-ethanediamine (L^TEED). L^TEED(Al-

drich) was distilled from CaH_2(s)before use. ¹H NMR (CDCl₃, 400

MHz): δ 1.01 (t, 12H); 2.5 (d, 4H); 2.45-2.55 (m, 8H). GC-MS:

m/z 172, M⁺.

[(L^TM)Cu(PPh₃)]CF₃SO₃(1d). To

a

solution of

1,4,7-Trimethyl-1,4,7-triazacyclononane (L^{Me TACN}). L^{Me TACN}

was synthesized according to the literature method.^{54 1}H NMR

(CDCl₃, 400 MHz): δ 2.4 (s, 9H); 2.7-2.8 (m, 12H). GC-MS:

353 K (1 min), 40 K min^-1to 393 K (10 min); t_R) 9.3 min (m/z

171, M⁺).

3

[Cu(MeCN)₄]CF₃SO₃(188 mg, 0.50 mmol) in 2 mL of CH₂Cl₂

were added L^TM(85 mg, 0.50 mmol) and PPh₃(131 mg, 0.50

mmol). The resulting clear solution was layered with cyclohexane

and allowed to stand for several days, whereupon large crystals

(colorless rhombic blocks) of 1d slowly formed (270 mg, 0.43

mmol, 85%). ¹H NMR (CDCl₃): δ 1.10-1.20 (m, 2H); 1.20-1.30

(br m, 2H); 1.75-1.85 (d, 2H); 1.93-1.99 (d, 2H), 2.41 (s, 6H);

2.60 (s, 6H); 2.65-2.70 (br m, 2H).

4,7-Dimethyl-4,7-diaza-1-oxacyclononane (L^{Me ODACN}). 4,7-

2

Diaza-1-oxacyclononane (1.1 g, 8.5 mmol), synthesized by the

literature method,⁵⁵was permethylated and purified as described

for L^TM(0.67 g, 50%). H NMR (CDCl₃, 400 MHz): δ 2.40 (s,

1

[(L^TB)Cu(MeCN)]CF₃SO₃(10a). L^TB(50 mg, 0.110 mmol) and

[Cu(MeCN)₄]CF₃SO₃(40 mg, 0.11 mmol) were dissolved in 1 mL

of CH₂Cl₂. Dropwise addition of Et₂O caused the formation of fine

white needles of 10a, which were collected by suction filtration,

washed with Et₂O, and dried in vacuo (75 mg, 0.099 mmol, 90%).

¹H NMR (400 MHz, CDCl₃): δ 1.30-1.40 (m, 2H); 1.45-1.60

(m, 2H); 2.00-2.10 (m, 2H); 2.15 (s, 3H); 2.40-2.50 (m, 2H);

3.02 (d, 2H); 3.20-3.30 (m, 4H); 3.60 (d, 2H); 3.95 (d, 2H); 6.55

(d, 4H); 7.20-7.50 (m, 16H). ¹³C NMR (400 MHz, CDCl₃): δ

23.3, 25.0, 54.5, 56.1, 58.7, 128.6, 128.9, 129.3, 129.5, 131.2, 131.7,

6H); 2.68 (m, 8H); 3.70 (t, 4H). GC-MS: m/z 158, M⁺.

N,N,N′,N′-Tetrabenzyl-trans-(1R,2R)-cyclohexanediamine (L^TB).

Freshly distilled CD (2.0 g, 175 mmol) was dissolved in 50 mL of

dry, deoxygenated DMF. Finely powdered K₂CO_3(s)(12.1 g, 875

mmol) and PhCH₂Br (12.3 g, 700 mmol) were added, and the

resulting effervescent milky suspension was stirred 24 h in an oil

bath at 353 K under a N₂atmosphere. The mixture was treated

with 100 mL of CH₂Cl₂and washed with 1 M aqueous NaOH (8

× 100 mL). The organic layer was dried over K₂CO_3(s)and the

solvent removed in vacuo. The resulting pale oil was redissolved

in 20 mL of absolute EtOH and chilled over dry ice until

crystallization ensued. The resulting white solid was filtered, washed

with cold (193 K) EtOH (3 × 10 mL), and dried in vacuo (2.9 g,

134.5, 135.3. FTIR: ν_C≡N, 2292 cm^-1

.

[(L^TB)Cu(CO)]CF₃SO₃. A solution of [10a‚2CH₂Cl₂] (100 mg,

0.123 mmol) in CDCl₃(1 mL) was sparged briefly with CO_(g)and

allowed to stand for 15 min under a CO atmosphere (1.7 atm).

Dropwise addition of Et₂O caused the precipitation of a micro-

crystalline solid product, which was recovered by suction filtration

and dried in vacuo (74 mg, 0.117 mmol, 95%). ¹H NMR (400 MHz,

CDCl₃): δ 1.40-1.60 (m, 2H); 1.60-1.80 (m, 2H); 2.00-2.10 (m,

2H); 2.15 (s, 3H); 2.50-2.60 (m, 2H); 3.30 (d, 2H); 3.40-3.55

(m, 4H); 3.90 (d, 2H); 4.05 (d, 2H); 6.25 (d, 4H); 7.15-7.40 and

1

35%). H NMR (400 MHz, CDCl₃): δ 0.95-1.15 (m, 4H); 1.7

(m, 2H); 2.1 (m, 2H); 2.7 (m, 2H); 3.35 (d, 4H); 3.75 (d, 4H);

7.2-7.3 (m, 12H); 7.4-7.45 (m, 8H). ¹³C NMR (400 MHz,

CDCl₃): δ 24.9, 25.9, 53.2, 58.1, 126.5, 127.9, 128.9, 140.6. FAB-

MS: m/z 475.2, M⁺.

Syntheses and in situ Preparations of [LCu(MeCN)]CF₃SO₃

Complexes 1a-9a (General Procedure). To a solution of

[Cu(MeCN)₄]CF₃SO₃(38 mg, 0.10 mmol) in the appropriate amount

of CH₂Cl₂, THF, or acetone was added 0.10 mmol of the ligand L,

yielding solutions of 1a-9a. 1a: ¹H NMR (CD₂Cl₂, 400 MHz): δ

1.05-1.10 (m, 2H); 1.10-1.25 (br m, 2H); 1.70-1.80 (d, 2H);

1.90-1.98 (d, 2H); 2.19 (CH₃CN, s, 12H); 2.23 (s, 6H); 2.30-

2.40 (br m, 2H); 2.49 (s, 6H). 2a: ¹H NMR (CD₂Cl₂, 400 MHz):

δ 1.10-1.20 (br m, 2H); 1.20-1.35 (t + br m, 5H); 1.65-1.75

(m, 2H); 1.90-2.00 (br m, 2H); 2.09 (CH₃CN, s, 12H); 2.31 (s,

3H); 2.32 (s, 3H); 2.48-2.51 (br m, 2H); 2.60 (s, 3H); 2.55-2.65

(br m, 1H); 2.60-2.70 (br m, 1H). 3a: ¹H NMR (CD₂Cl₂, 400

MHz): δ 1.05-1.15 (m, 2H); 1.20-1.35 (m, 8H); 1.65-1.75 (m,

2H); 1.90-2.00 (m, 2H); 2.05 (CH₃CN, s, 12H); 2.27 (s, 6H); 2.55-

2.80 (br m, 6H). 4a: ¹H NMR (CD₂Cl₂, 400 MHz): δ 1.05-1.15

(br m, 2H); 1.20-1.35 (br m, 14H); 1.75-1.85 (br m, 2H); 1.95-

2.05 (br m, 2H); 2.07 (CH₃CN, s, 12H); 2.35 (sextet, 2H); 2.55-

2.70 (br m, 6H); 2.91 (sextet, 2H). 5a ¹H NMR (CD₂Cl₂, 300

MHz): δ 1.15-1.19 (t, 12H); 2.08 (CH₃CN, s, 12H); 2.62 (s, 4H);

2.68-2.75 (q, 8H). 6a ¹H NMR (CD₂Cl₂, 300 MHz): δ 1.18-1.13

(t, 6H); 2.08 (CH₃CN, s, 12H); 2.34 (br s, 6H); 2.3-2.9 (br m, 8H).

7a ¹H NMR (CD₂Cl₂, 200 MHz): δ 1.08-1.15 (t, 6H); 2.07 (CH₃-

CN, s, 12H); 2.44 (s, 6H); 2.45-2.65 (m, 4H); 2.66-2.77 (q, 4H).

7.50-7.70 (m, 16H). FTIR (CH₂Cl₂solution): ν_C≡O, 2113 cm^-1

Syntheses of Bis(µ-hydroxo)dicopper(II) Complexes

.

TM

[( L

RR

)₂Cu₂(µ-OH)₂](CF₃SO₃)₂(_RR,RR1e) and [(_SSL^TM)(_RRL^TM

)

Cu₂(µ-OH)₂](CF₃SO₃)₂_RR,SS1e). Method A (_RR,RR1e): A cold

(

solution of 1b in CH₂Cl₂(25 mL, [Cu] ) 2.0 mM) was injected

with 1.1 mL of a 50 mM CH₂Cl₂solution of DBP or tert-butylthiol

(2.2 equiv) at 193 K under a N₂atmosphere, whereupon the color

rapidly changed to violet. The solution was warmed to ambient

temperature, concentrated to ∼1 mL in vacuo, and layered with

Et₂O. Violet rhombic blocks of _RR,RR1e (19 mg, 90%) separated

upon standing. Method B (_RR,RR1e and _RR,SS1e): A CH₂Cl₂solution

TM

of 1a (10 mL, 30 mM; prepared with

L

or

L

rac

as

RR

appropriate) was exposed to the atmosphere in an open flask at

room temperature (i.e., in the presence of atmospheric moisture),

whereupon a rapid color change to deep blue-violet was observed.

The solution was layered with Et₂O and allowed to stand for several

days, whereupon violet rhombic blocks of the corresponding bis-

(µ-hydroxo)dicopper(II) dimer [L₂Cu₂(µ-OH)₂](CF₃SO₃)₂separated

(72 mg, 0.090 mmol, 60%).

General Preparation of Bis(µ-oxo)dicopper(III) Dimers (Os)

ME

[L₂Cu₂(µ-O)₂](CF₃SO₃)₂(L ) L^TM, 1b; L ) L^EtMe, 2b; L ) L

,

3

1

3b; L ) L^TE, 4b; L ) L^TEED, 5b; L ) L^{(MeEt) ED}, 6b; L )

2

8a H NMR (CD₂Cl₂, 500 MHz): δ 2.03 (CH₃CN, s, 12H); 2.68

(s, 4H); 2.69 (br m, 8H). Single diffraction-quality crystals (colorless

needles) of 4a were obtained upon layering a concentrated (50 mM,

L^{Me Et ED}, 7b; L ) L^{Me TACN}, 8b; L ) L^{Me ODACN}, 9b) and

Manometry. Air-free solutions of 1a-9a, prepared in situ in

concentrations ranging from 0.1 to 50 mM⁶²in purified CH₂Cl₂,

THF, or acetone, were chilled to 193 K in a dry ice/2-propanol

bath. Dry O₂was admitted by bubbling with a fine needle,

2

3

2

2.0 mL) CH₂Cl₂solution with Et₂O (22 mg, 0.046 mmol, 45%).

TM

[( L

RR

)₂Cu]CF₃SO₃‚[(_SSL^TM)₂Cu]CF₃SO₃(1c). To a solution

of [Cu(MeCN)₄]CF₃SO₃(188 mg, 0.50 mmol) in 2 mL of CH₂Cl₂

Inorganic Chemistry, Vol. 44, No. 21, 2005 7363

Cole et al.

whereupon the solutions turned yellow- to red-brown (1b-9b,

respectively). Manometric measurements were performed using a

calibrated Schlenk line with a dead volume of ∼100 mL, an open-

tube mercury manometer, and a 50 mL Teflon-valved flask. A

solution of the Cu(I) precursor (3a-5a, 7a: CH₂Cl₂, [Cu] ) 40

mM) was thoroughly degassed by freeze-pump-thaw cycles

(three) and allowed to react with O₂at 193 K (1 atm initial pressure)

for a minimum of 2 h or until gas uptake ceased (maximum 8 h,

4a). The cooling apparatus was then removed, and the system was

allowed to equilibrate at 293 K to account for solvent vapor pressure

effects. The decomposed solutions were immediately resubjected

to the same procedure to provide a reference measurement in each

case. Under the given conditions, a ∼6 cm difference between the

reaction and blank corresponded to a 2:1 (Cu/O₂) uptake ratio. The

calculated uptake ratios were as follows. 3a, 2.0(1):1 (average of

three trials); 4a, 2.2:1; 5a, 1.9(1):1 (average of four trials); 7a, 2.1-

(1):1 (average of three trials). Single crystals of [3b‚4CH₂Cl₂]

suitable for X-ray diffraction study (yellow-brown rhombic blocks)

were grown by direct diffusion of Et₂O into a concentrated ([Cu]

≈ 50 mM) CH₂Cl₂solution at 193 K. Selected crystals were

examined by light microscopy and mounted on a goniometer at

193 K using an improvised cold stage.

of neutral activated alumina (Brockmann I, ∼150 mesh, 58 Å)

followed by the aqueous layer. The column was rinsed with MeOH

(2 mL) and the combined metal-free eluents dried over K₂CO_3(s)

.

The solution was concentrated by rotary evaporation to ∼2 mL,

and a stoichiometric amount of cinnamaldehyde, acetophenone,

anthracene, or benzonitrile was added as an internal GC standard.

To avoid aerobic oxidation, atmospheric exposure was minimized

throughout the workup process via N₂blanketing and prompt

handling. The ligand products (total ligand mass recovery was

generally >90%) were subjected to GC/GC-MS analysis.⁵⁹

Spectrophotometric Titrations and Oxidation of Exogenous

Substrates (General Procedure). Deoxygenated (N₂, min. 15 min)

O solutions (1b-7b; [Cu] ) 1.0 mM; 10 mL CH₂Cl₂; 193 K) were

prepared and injected with the following substrates dissolved in

dry deoxygenated THF or CH₂Cl₂(0.5 mL, 2 equiv relative to O

unless otherwise specified): ethylene (1 atm); 9,10-dihydroan-

thracene (1-5 equiv); fluorene (1-5 equiv); xanthene (1-5 equiv);

cyclohexene (1-5 equiv); 1,4-cyclohexadiene (1-5 equiv); 1-

octanol (protonated and Li salt); PhCH₂OH (protonated and Li salt);

cinnamyl alcohol (protonated and Li salt); (CH₃)₂S (1-5 equiv);

(CH₃)₂SO (1-5 equiv); PhSCH₃(1-5 equiv); tert-butylthiol;

thiocresol; 2,4-di-tert-butylphenol. Lithium salts of the alcohols

were generated in situ by stoichiometric addition of n-butyllithium

to a THF solution. The reaction mixtures were maintained at the

desired temperature (generally 193 K) under N₂until no further

optical change was evident (max 6 h). The solutions were then

passed through a column of neutral activated alumina (Brockmann

I, ∼150 mesh, 58 Å) at 293 K followed by two column volumes

of MeOH. After drying over K₂CO_3(s), the combined copper-free

eluents were subjected to GC/GC-MS analysis. Mass recovery of

the products was generally >90% as deduced by addition of an

internal calibrant. Substrate oxidation products were identified by

their GC retention times (t_R) and molecular ion m/z (M⁺) peaks⁵⁹

Thermal Decomposition Kinetics. The thermal decomposition

reactions of 1b-9b, as well as deuterated complexes d₂₄-1b and

d₈-3b, were monitored in a custom-designed low-temperature cell

in CH₂Cl₂solution (except where otherwise noted) at the following

Cu concentrations: 1b/d₂₄-1b: 0.2, 0.5, 1.0, 2.0 mM; 2b: 1.0 mM;

3b/d₈-3b: 0.2, 1.0, 2.0 mM; 4b: 1.0, 2.0 mM (CH₂Cl₂and THF);

-

5b: 1.0, 2.0 mM (CH₂Cl₂and THF, also SbF₆counteranion in

THF); 6b: 1.0, 2.0 mM; 7b: 1.0, 2.0 mM; 8b: 0.2, 1.0 mM; 9b:

0.2, 2.0 mM. The Os were formed by injecting 1.0 mL of a

concentrated (10×) solution of the corresponding Cu(I) complex

(1a-9a) into 9.0 mL of cold, O₂-saturated solvent (193 K, 1 atm).

After complete O formation was verified by the stabilization of

optical absorbances at the expected values (Table 3), the excess

O₂was removed by vacuum/N₂purge cycles (four) and the complex

was allowed to decay at the desired temperature (213-293 K),

which was maintained by a thermostatic cooling bath (FTS Systems,

New York). Data collection for decay was started only after the

solution had attained the desired temperature as detected by an

OMEGA surface temperature probe; approximately 2-3 min were

normally required for thermal equilibration. A multiwavelength

(280-450 nm) component analysis of the data was performed using

SPECFIT to obtain rate constants. Data reduction was performed

by singular value decomposition analysis followed by Marquardt

minimization to the given model. The number of significant

eigenvectors obtained was generally 2 for all data sets, and a first-

order A f B reaction model provided a suitable fit to each.

Reproducibility of the results was checked in each case by

conducting multiple trials (minimum three). The decomposed O

solutions from kinetic runs were generally subjected to analysis

(vide infra) to verify ligand product distribution.

1

and were confirmed in some cases by comparison of their H and

¹³C NMR features with literature values.

Spectrophotometric titrations of 1b-5b (193 K, CH₂Cl₂, [Cu]

) 1.0 mM) with Fc, AcFc, and Me₁₀Fc were conducted in a similar

manner, with successive injections of 0.1 equiv of aliquots spaced

at intervals sufficient to allow equilibration. Additional reactions

of 1b-5b with AcFc were attempted under like conditions with

no change in optical absorbance noted after 6 h.

Acknowledgment. We thank Drs. Frederick J. Hollander

and Allen G. Oliver of the UC Berkeley College of

Chemistry X-ray Diffraction Facility (CHEXRAY) for col-

lection of X-ray diffraction data. Funding was provided by

NIH GM50730 (T.D.P.S). A.P.C. thanks the National Sci-

ence Foundation for a predoctoral fellowship, and L.M.M.

gratefully acknowledges a John Stauffer Graduate Fellow-

ship.

Supporting Information Available: Characterization of prod-

ucts from ligand decomposition and exogenous substrates oxidation;

initial rates plot for the oxygenation of 4a; Eyring plots and kinetic

data for thermal decomposition of 1b, 3b, 4b, 5b, 8b, and 9b;

experimental details, atomic positions, anisotropic thermal param-

eters, bond distances, bond angles, torsional angles, and CIF files

for the crystal structures of 1c, 1d, _RR,SS1e, _RR,RR1e, 3b, 4a, and

10a; input coordinates for hybrid DFT calculations on homochiral

and heterochiral formulations of 1c (PDF). This material is available

free of charge via the Internet at http://pubs.acs.org.

Ligand Decomposition Product Analysis (General Procedure

for 1b, 3b-9b). For ligand decomposition product analyses, 10-

mL CH₂Cl₂solutions of Os (1b, [Cu] ) 1.0-2.0 mM, also 2.0-

5.0 mM in THF; 3b-9b, [Cu] ) 0.2-2.0 mM) were prepared at

193 K, degassed by evacuation and backfilling with N₂(three

cycles), and allowed to decompose at a convenient temperature

between 213 and 293 K for several hours until the characteristic O

LMCT bands had completely disappeared. The vessel was removed

from the cooling bath, and aqueous NH₃(30%) was added dropwise

to the still-cold solution with vigorous shaking until the CH₂Cl₂

layer was colorless. The organic layer was passed through a column

IC050331I

7364 Inorganic Chemistry, Vol. 44, No. 21, 2005

Article Doi

Bis(μ-oxo)dicopper(III) complexes of a homologous series of simple peralkylated 1,2-diamines: Steric modulation of structure, stability, and reactivity

DOI: 10.1021/ic050331i

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic050331i

Full text of DOI:10.1021/ic050331i

Products guided by the article

R&D Labs maybe for 868832-83-1

Relevant to this article

Hot Product