Synthesis, characterization, and structures of copper(II)-thiosulfate complexes incorporating tripodal tetraamine ligands

Article abstract of DOI:10.1021/ic0492800

The reaction of [Cu(L)(H₂O)]₂₊ with an excess of thiosulfate in aqueous solution produces a blue to green color change indicative of thiosulfate coordination to Cu(II) [L = tren, Bz₃tren, Me ₆tren, and Me₃tren; tren = tris(2-aminoethyl)amine, Bz₃tren = tris(2-benzylaminoethyl)amine, Me₆tren = tris(2,2-dimethylaminoethyl)amine, and Me₃tren = tris(2- methylaminoethyl)amine]. In excess thiosulfate, only [Cu(Me₆tren) (H₂O)]²⁺ promotes the oxidation of thiosulfate to polythionates. Products suitable for single-crystal X-ray diffraction analyses were obtained for three thiosulfate complexes, namely, [Cu(tren)(S ₂O₃)]·H₂O, [Cu(Bz₃tren) (S₂O₃)]·MeOH, and (H₃Me ₃tren)[Cu(Me₃tren)(S₂O₃)] ₂-(CIO₄)₃. Isolation of [Cu(Me ₆tren)(S₂O₃)] was prevented by its reactivity. In each complex, the copper(II) center is found in a trigonal bipyramidal (TBP) geometry consisting of four amine nitrogen atoms, with the bridgehead nitrogen in an axial position and an S-bound thiosulfate in the other axial site. Each structure exhibits H bonding (involving the amine ligand, thiosulfate, and solvent molecule, if present), forming either 2D sheets or 1D chains. The structure of [Cu(Me₃tren)(MeCN)](CIO₄)₂ was also determined for comparison since no structures of mononuclear Cu(II)-Me ₃tren complexes have been reported. The thiosulfate binding constant was determined spectrophotometrically for each Cu(II)-amine complex. Three complexes yielded the highest values reported to date [K_f = (1.82 ± 0.09) × 10³ M^-1 for tren, (4.30 ± 0.21) × 10⁴ M^-1 for Bz₃tren, and (2.13 ± 0.05) × 10³ M^-1 for Me₃tren], while for Me₆tren, the binding constant was much smaller (40 ± 10 M^-1).

Full text of DOI:10.1021/ic0492800

Inorg. Chem. 2004, 43, 6568−6578
Synthesis, Characterization, and Structures of Copper(II)
Complexes Incorporating Tripodal Tetraamine Ligands
−Thiosulfate
Adam J. Fischmann, Andrew C. Warden, Jay Black, and Leone Spiccia*
School of Chemistry, Monash UniVersity, Victoria 3800, Australia
Received June 2, 2004
+
The reaction of [Cu(L)(H2O)]2 with an excess of thiosulfate in aqueous solution produces a blue to green color
change indicative of thiosulfate coordination to Cu(II) [L
aminoethyl)amine, Bz3tren tris(2-benzylaminoethyl)amine, Me6tren
tren
tris(2-methylaminoethyl)amine]. In excess thiosulfate, only [Cu(Me6tren)(H2O)]2 promotes the oxidation of
thiosulfate to polythionates. Products suitable for single-crystal X-ray diffraction analyses were obtained for three
thiosulfate complexes, namely, [Cu(tren)(S2O3)] H2O, [Cu(Bz3tren)(S2O3)] MeOH, and (H3Me3tren)[Cu(Me3tren)(S2O3)]2-
)
tren, Bz3tren, Me6tren, and Me3tren; tren
) tris(2-
)
)
tris(2,2-dimethylaminoethyl)amine, and Me3-
+
)
‚
‚
(ClO4)3. Isolation of [Cu(Me6tren)(S2O3)] was prevented by its reactivity. In each complex, the copper(II) center is
found in a trigonal bipyramidal (TBP) geometry consisting of four amine nitrogen atoms, with the bridgehead nitrogen
in an axial position and an S-bound thiosulfate in the other axial site. Each structure exhibits H bonding (involving
the amine ligand, thiosulfate, and solvent molecule, if present), forming either 2D sheets or 1D chains. The structure
of [Cu(Me3tren)(MeCN)](ClO4)2 was also determined for comparison since no structures of mononuclear Cu(II)
−
Me3tren complexes have been reported. The thiosulfate binding constant was determined spectrophotometrically
for each Cu(II)
×
Me6tren, the binding constant was much smaller (40
−
amine complex. Three complexes yielded the highest values reported to date [Kf
)
(1.82
± 0.09)
1
1
1
103 M- for tren, (4.30
±
0.21)
×
104 M- for Bz3tren, and (2.13
±
0.05)
×
103 M- for Me3tren], while for
±
10 M-1).
Au0(s) + 2S2O32- / [Au(S2O3)2]3- + e-
[Cu(NH3)4]2+ + 3S2O32- + e- / [Cu(S2O3)3]5- + 4NH3
(1)
Introduction
The hydrometallurgical processing of gold requires an
oxidant to convert metallic gold to either Au(I) or Au(III)
ions and a ligand to complex the oxidized gold and stabilize
it in aqueous solution. In the traditional cyanidation process,
oxygen is used as the oxidant and cyanide as the complexing
agent.1 Growing concerns about the potential environmental
hazards presented by the production, transport, and use of
cyanide are stimulating interest in alternative gold-processing
methods.2-4 One promising, more benign alternative utilizes
a copper(II) tetraamine catalyst {[Cu(NH3)4]2+} to oxidize
native gold to Au(I) and thiosulfate (S2O32-) as the com-
plexing ligand, forming [Au(S2O3)2]3-.3
(2)
The Cu(II) complex oxidizes gold at a much faster rate
than oxygen, increasing the efficiency of the process.3 In
the reaction, the [Cu(NH3)4]2+ catalyst is converted into the
Cu(I) species, and for the process to continue, the catalyst
needs to be regenerated, usually by air oxidation. An
impediment to the industrial use of the thiosulfate leaching
process is that [Cu(NH3)4]2+ also oxidizes thiosulfate, initially
to tetrathionate (S4O62-) (eq 3).3 This leads to unacceptable
consumption of thiosulfate and has an adverse impact on
economic viability.
* Author to whom correspondence should be addressed. E-mail:
leone.spiccia@sci.monash.edu.au. Fax: +61-3-9905-4597.
(1) Sparrow, G. J.; Woodcock, J. T. Miner. Process. Extr. Metall. ReV.
1995, 14, 193-247.
2[Cu(NH3)4]2+ + 8S2O3
/
2-
2[Cu(S2O3)3]5- + S4O62- + 8NH3 (3)
(2) Grosse, A. C.; Dicinoski, G. W.; Shaw, M. J.; Haddad, P. R.
Hydrometallurgy 2003, 69, 1-21.
We have embarked on a search for alternative Cu(II)
complexes that effectively and rapidly oxidize gold but do
(3) Aylmore, M. G.; Muir, D. M. Miner. Eng. 2001, 14, 135-174.
(4) Molleman, E.; Dreisinger, D. Hydrometallurgy 2002, 66, 1-21.
6568 Inorganic Chemistry, Vol. 43, No. 21, 2004
10.1021/ic0492800 CCC: $27.50
© 2004 American Chemical Society
Published on Web 09/14/2004

Cu(II)-Thiosulfate Complexes with Tetraamine Ligands
not oxidize thiosulfate.5 A feature to emerge from these
studies is that color changes, presumably signifying Cu(II)-
thiosulfate complexation, occur immediately on the addition
of thiosulfate to many Cu(II) complexes. As indicated in our
previous publication, subsequent reactions, leading to thio-
sulfate oxidation, are sufficiently fast to prevent isolation
and allow full analysis of the product.5
In fact, few constants (Kf) for the coordination of thio-
sulfate to Cu(II) are available,6-9 and at the present time,
only one crystal structure of a Cu(II)-thiosulfate complex
has been reported in which thiosulfate occupies the axial
position in a square pyramidal geometry.10 In the course of
our studies, the Cu(II) complexes of tren [tris(2-aminoethyl)-
amine], Bz3tren [tris(2-benzylaminoethyl)amine], Me3tren
[tris(2-methylaminoethyl)amine], and Me6tren [tris(2,2-di-
methylaminoethyl)amine] were found to react with thiosulfate
in aqueous solution to form complexes which, with the
exception of Me6tren, are stable for long periods under certain
conditions. This has enabled the thiosulfate binding constant
to be measured in aqueous solution and the isolation,
characterization, and X-ray structure analysis of [Cu(tren)-
(S2O3)]‚H2O, [Cu(Bz3tren)(S2O3)]‚MeOH, and (H3Me3tren)-
[Cu(Me3tren)(S2O3)]2(ClO4)3 reported herein. Our study was
aided by the ability of the tetradentate ligands to bind strongly
to four of five coordination sites on a TBP Cu(II) center11
[tren,12 log K ) 19.58; Me3tren,13 log K ) 19.11; Me6tren,12
log K ) 15.65; and Bz3tren, log K ) 15.10 (vide infra)],
leaving only one site available for the coordination of other
ligands.
Figure 1. UV-visible spectra of aqueous solutions of [Cu(L)(H2O)]2+
and S2O32-, L ) tren (s), Me3tren (s s), Bz3tren* (- ‚ -), and Me6tren
(- ‚‚ -). [Cu]total ) 0.25 mM; [S2O32-] ) 25 mM. For Bz3tren*, [Cu]total
) 0.5 mM, [S2O32-] ) 2.5 mM, and absorbance is scaled by a factor of
one-half.
composition reaction which involved several processes.5 For
the conditions of [Cu]total ) 2.5 mM and [S2O32-] ) 0.25
mM, the rate constant of the final (slowest) process was (1.7
( 0.1) × 10-4 s-1, and at the end of the reaction, the
spectrum matched that of [Cu(tren)(H2O)]2+. The Cu(I)
species that formed did not absorb between 300 and 500
nm, the wavelength range monitored.5 The formation constant
for [Cu(Me6tren)(S2O3)] is the smallest in the series (vide
infra). Moreover, for this system, bleaching of the solution
that is indicative of thiosulfate oxidation was observed for
all thiosulfate/Cu(II) ratios examined.
Synthesis and Characterization. Crystals of [Cu(tren)-
(S2O3)]‚H2O were deposited on slow evaporation of a
concentrated aqueous solution of [Cu(tren)(H2O)](ClO4)2 and
Na2S2O3‚5H2O (1:1 molar ratio). Figure 2 shows that the
four absorption maxima found in the UV-visible diffuse
Results and Discussion
Exploratory Studies. The addition of thiosulfate to
solutions of [Cu(L)(H2O)]2+ was found to cause a blue to
green color change reflecting coordination of thiosulfate (eq
4), which results in a thiosulfate-S to Cu(II) ligand-to-metal
charge transfer (LMCT) transition between 300 and 450 nm
(Figure 1).
[Cu(L)(H2O)]2+ + S2O32- / [Cu(L)(S2O3)] + H2O (4)
reflectance spectrum of the solid match those in the spectrum
2-
of an aqueous solution of [Cu(tren)(H2O)]2+ and S2O3
,
The intensity of the LMCT bands varies with the thiosul-
fate concentration, but there was no change in the spectrum
over many days when excess thiosulfate was present (the
Me6tren complex was the only exception). In the case of
tren, when the total concentration of Cu(II) was in excess
over thiosulfate, the thiosulfate complex underwent a de-
confirming that the solution and solid-state chromophores
must be structurally similar.
The IR spectrum exhibits absorptions indicating the
presence of water, tren, and thiosulfate and the absence of
perchlorate. The position of the asymmetric stretch of the
SO3 group, νas(SO3), normally found in the 1100-1200 cm-1
region, is diagnostic of the thiosulfate binding mode;14
S-bridging (>1175 cm-1) can be distinguished from S-
coordination (1130-1175 cm-1) and O-coordination (<1130
cm-1). In ionic thiosulfates, νas(SO3) is found at ∼1130 cm-1.
Thus, the strong band at 1154 cm-1 supports S-coordination.
The symmetric SO3 stretch near 1000 cm-1 (1008 cm-1 for
this compound) is less diagnostic, but the position can be
correlated to S-coordination (>1000 cm-1) or O-coordination
(<1000 cm-1)15 and again supports the former.
(5) Brown, T. A.; Fischmann, A. J.; Spiccia, L.; McPhail, D. C.
Hydrometallurgy 2003. Fifth International Conference in Honor of
Professor Ian Ritchie, Vancouver, BC, Canada, Aug 24-27, 2003; 1,
213-226.
(6) Ra´bai, G.; Epstein, I. R. Inorg. Chem. 1992, 31, 3239-3242.
(7) Hemmes, P.; Petrucci, S. J. Phys. Chem. 1968, 72, 3986-3992.
(8) Hemmes, P.; Petrucci, S. J. Phys. Chem. 1970, 74, 467-468.
(9) Matheson, R. J. Phys. Chem. 1967, 71, 1302-1308.
(10) Podberezskaya, N. V.; Borisov, S. V.; Bakakin, V. V. J. Struct. Chem.
1971, 12, E770-E774 (English translation).
(11) Golub, G.; Lashaz, A.; Cohen, H.; Paoletti, P.; Bencini, A.; Valtancoli,
B.; Meyerstein, D. Inorg. Chim. Acta 1997, 255, 111-115.
(12) Anderegg, G.; Gramlich, V. HelV. Chim. Acta 1994, 77, 685-690.
(13) Thaler, F.; Hubbard, C. D.; Heinemann, F. W.; van Eldik, R.; Schindler,
S.; Fa´bia´n, I.; Dittler-Klingemann, A. M.; Hahn, F. E.; Orvig, C. Inorg.
Chem. 1998, 37, 4022-4029.
(14) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coor-
dination Compounds, 4th ed.; Wiley: New York, 1986.
Inorganic Chemistry, Vol. 43, No. 21, 2004 6569

Fischmann et al.
[Cu(Me3tren)(MeCN)](ClO4)2. Since the only previously
published crystal structure of a Cu(II)-Me3tren complex is
that of a cyanide-bridged binuclear species, [Cu2(Me3tren)2-
(CN)](ClO4)3‚2MeCN,13 the structure of [Cu(Me3tren)-
(MeCN)](ClO4)2 was determined for comparison with the
2-
Cu(II)-Me3tren-S2O3 structure. The Cu(II) geometry of
[Cu(Me3tren)(MeCN)](ClO4)2 is almost an ideal TBP struc-
ture (τ ) 94%), and acetonitrile coordinates to the free axial
position via the nitrile nitrogen (Figure 3). The methyl groups
point down toward the acetonitrile and radiate out sym-
metrically. As has been found for other Cu(II) complexes
of tren-type ligands,13 the Cu(II) center projects away from
the bridgehead nitrogen, in this case, by 0.194(2) Å out of
the N(2), N(3), and N(4) plane. Moderately strong H bonding
between some perchlorate oxygens and the amine protons
(Table 2 and Figure 4) arranges the structure into 1D chains
that align at an angle to the c axis. One perchlorate [Cl(2)]
bridges adjacent [Cu(Me3tren)(MeCN)]2+ cations via O(6)
and O(7), whereas the other projects out from the chain and
only bonds with H through O(4).
Thiosulfate Complexes. In the structure of the only
compound to feature the coordination of thiosulfate to Cu-
(II), [Cu(en)2(S2O3)], thiosulfate is weakly coordinated via
the terminal sulfur (Cu-S bond distance of 2.71 Å) to a
copper(II) center in a square pyramidal geometry.17 The lack
of crystallographic data arises because Cu(II) complexes
often promote the oxidation of thiosulfate to polythionates,6
and thiosulfate complexes are commonly too unstable to be
crystallized. Crystal structures featuring the monodentate S
binding of thiosulfate to other divalent transition-metal ions
have been reported for Ni(II),10,18 Co(II),18 Zn(II),19,20 and
Pd(II)21 (Table 4); however, none of these have TBP
geometry. As observed previously,18 the mean S-O bond
distance of the Cu(II) complexes is similar to that in ionic
thiosulfates (e.g., the Na2S2O3 described below). The S-S
and M-S bonds show greater variation, and for three of the
complexes {[Cu(tren)(S2O3)]‚H2O, complex b of [Cu(Me3-
tren)(S2O3)], and [Cu(Bz3tren)(S2O3)]‚MeOH}, there appears
to be an inverse relationship between the S-S and M-S
distances. In terms of the Pearson HSAB scheme,22 the Pd-
(II) ion should favor S binding; accordingly, the Pd(II)
complex has the shortest M-S and the longest S-S bond
distances. When compared to those of other 3d metal ions,
the M-S distances in the [Cu(L)(S2O3)] complexes are
shorter than those of the Co(II) and Ni(II) complexes and
closest to that in the Zn(II) complex.
Figure 2. Comparison of the solid-state diffuse reflectance (top) and
aqueous solution (bottom) electronic spectra of [Cu(tren)(S2O3)] (s). For
solution spectrum, [Cu(tren)(H2O)2+] ) 0.25 mM and [S2O32-] ) 2.5 mM
(inset: 2.5, 25 mM, respectively). The diffuse reflectance spectrum of [Cu-
(Bz3tren)(S2O3)]‚MeOH is also shown (- - -).
The IR spectrum of the green crystals of [Cu(Bz3tren)-
(S2O3)]‚MeOH indicated S-bound thiosulfate14 [νas(SO3) )
1157 cm-1], and the (S2O32-)-Cu(II) LMCT transition in
the diffuse reflectance of the UV-visible spectrum (Figure
2) matched that from the aqueous solution (367 nm). The
two d-d transitions (868 and 630 nm, Figure 2) are
characteristic of TBP Cu(II) geometry, as would be expected
with a tripodal ligand.
Blue crystals of [Cu(Me3tren)(MeCN)](ClO4)2 and green
crystals of (H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3 showed
absorptions in their IR spectra confirming the presence of
perchlorate, Me3tren, MeCN (for the former), and thiosulfate
(for the latter).
Crystallography. The molecular structures of the com-
plexes reveal that the tripodal tetradentate ligands enforce a
trigonal bipyramidal (TBP) Cu(II) geometry, which is typical
of Cu(II) complexes of tren and derivatives thereof.11 In each
case, the bridgehead amine nitrogen occupies an axial
position and the three pendant amines (either primary for
tren or secondary for Me3tren and Bz3tren) form the
equatorial plane. The remaining axial position is occupied
by S-coordinated thiosulfate or acetonitrile. In the discussion
to follow, the τ parameter of Addison et al.16 is used to
quantify the degree of distortion of these five-coordinate
complexes from ideal trigonal bipyramidal geometry (τ )
100%) toward ideal square pyramidal geometry (τ ) 0%).
[Cu(tren)(S2O3)]‚H2O (Figure 5) and [Cu(Bz3tren)(S2O3)]‚
MeOH (Figure 6) have simple structures consisting of a
(17) Podberenzskaya, N. V.; Borisov, S. V.; Bahakin, V. V. J. Struct. Chem.
1971, 12, E770-E774 (English translation).
(18) Carter, A.; Drew, M. G. B. Polyhedron 1999, 18, 1445-1453.
(19) Baggio, R.; Baggio, S.; Pardo, M. I.; Garland, M. T. Acta Crystallogr.,
Sect. C: Cryst. Struct. Commun. 1996, 52, 820-823.
(20) Andreetti, G. D.; Cavalca, L.; Domiano, P.; Musatti, A. Ric. Sci. 1968,
38, 1100-1101.
(21) Baggio, S.; Amzel, L. M.; Becka, L. N. Acta Crystallogr., Sect. B:
Struct. Sci. 1970, 26, 1698-1705.
(22) Pearson, R. G. In Hard and Soft Acids and Bases; Pearson, R. G.,
Ed.; Dowden, Hutchinson & Ross: Stroudsburg, PA, 1973; pp 67-
107.
(15) Freedman, A. N.; Straughan, B. P. Spectrochim. Acta, Part A 1971,
27, 1455-1465.
(16) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rign, J.; Verschoor, G.
C. J. Chem. Soc., Dalton Trans. 1984, 1349-1356.
6570 Inorganic Chemistry, Vol. 43, No. 21, 2004

Cu(II)-Thiosulfate Complexes with Tetraamine Ligands
Table 1. Crystal Data for [Cu(Me3tren)(MeCN)](ClO4)2 (I), (H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3 (II), [Cu(tren)(S2O3)]‚H2O (III), and
[Cu(Bz3tren)(S2O3)]‚MeOH (IV)
I
II
III
IV
empirical formula
C11H27Cl2CuN5O8
491.82
123(2)
C27H69Cl3Cu2N12O18S4
1211.61
123(2)
C6H20CuN4O4S2
339.92
123(2)
C28H40CuN4O4S2
624.3
123(2)
fw (g mol-1
)
T (K)
crystal system
space group
a (Å)
orthorhombic
Pna2(1)
19.2487(3)
8.49330(10)
12.35080(10)
90
orthorhombic
P2(1)2(1)2(1)
12.0421(3)
12.8448(3)
33.9332(9)
90
monoclinic
P2(1)/c
10.894(2)
10.053(2)
12.238(2)
102.36
orthorhombic
Pbca
16.1557(4)
14.1477(4)
26.5325(9)
90
b (Å)
c (Å)
â (deg)
V (Å3)
2019.17(4)
4
5248.7(2)
4
1309.2(5)
4
6064.4(3)
8
Z
F
calc (g cm-3
)
1.618
1.533
1.725
1.368
µ Mo KR (mm-1
)
1.394
1.196
1.996
0.897
reflections collected
27524
36419
14532
46210
independent reflections
4699
1.05
10602
3202
6941
GOF on F2
1.088
1.045
1.024
R1,a wR2b [I > 2σ(I)]
maximum difference peak, hole (e Å-3
0.0320, 0.0710
0.744, -0.398
0.0940, 0.2045
1.814, -0.897
0.0304, 0.0663
0.396, -0.537
0.0696, 0.0951
0.517, -0.478
)
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) [∑w(Fo - Fc )2/∑w(Fo )2]1/2
.
2
2
2
Table 3. Selected Bond Distances (Å) and Bond Angles (deg) in
[Cu(Me3tren)(MeCN)](ClO4)2
Cu(1)-N(5)
Cu(1)-N(1)
Cu(1)-N(2)
Cu(1)-N(4)
Cu(1)-N(3)
N(5)-C(10)
C(10)-C(11)
1.982(3)
2.030(2)
2.080(2)
2.093(2)
2.111(3)
1.135(4)
1.449(5)
N(5)-Cu(1)-N(1)
N(5)-Cu(1)-N(2)
N(1)-Cu(1)-N(2)
N(5)-Cu(1)-N(4)
N(1)-Cu(1)-N(4)
N(2)-Cu(1)-N(4)
N(5)-Cu(1)-N(3)
N(1)-Cu(1)-N(3)
N(2)-Cu(1)-N(3)
N(4)-Cu(1)-N(3)
C(10)-N(5)-Cu(1)
N(5)-C(10)-C(11)
179.3(1)
95.1(1)
85.3(1)
95.6(1)
84.72(9)
123.2(1)
95.2(1)
84.0(1)
Figure 3. ORTEP view of the complex in [Cu(Me3tren)(MeCN)](ClO4)2
(hydrogens omitted for clarity).
124.1(1)
110.2(1)
175.2(2)
178.8(3)
Table 2. Hydrogen Bonding in [Cu(Me3tren)(MeCN)](ClO4)2
(distances in Å and angles in deg)a
[Cu(Me3tren)(S2O3)] moieties (complex a in Figure 7) has a
highly distorted geometry intermediate between TBP and
square pyramidal (τ ) 57%), whereas the other (τ ) 86%,
complex b of Figure 7) is similar to the tren and Bz3tren
complexes (τ ) 81 and 82%, respectively). The high degree
of distortion of complex a is reflected in a tight Cu-S-S
angle (<100°), which helps to accommodate the hydrogen
bonding of O(2) to the (H3Me3tren)3+ cation (Figure 8). The
N(2)-Cu(1)-N(3) angle increases from 120 (ideal) to 139.2-
(4)° to permit the thiosulfate to adopt this orientation.
D-H
H‚‚‚A
D‚‚‚A
<(DHA)
N(2)-H(2)‚‚‚O(7)1
N(4)-H(4)‚‚‚O(4)2
N(3)-H(3)‚‚‚O(6)
0.93(4)
0.86(4)
0.93(5)
2.50(3)
2.37(4)
2.29(5)
3.263(3)
3.199(4)
3.183(4)
140(3)
163(4)
162(4)
a Symmetry transformations used to generate equivalent atoms: (1) -x
+
1/2, y + 1/2, z + 1/2; (2) x - 1/2, -y - 1/2, z. D ) donor atom; A )
acceptor atom.
With the coordination to Cu(II), the S-S bond is signifi-
cantly lengthened from the ionic distance24 of 2.0054(4) Å,
yet the S-O distances are indistinguishable from those in
ionic compounds [1.468(4) Å].25 As is typical for Cu(II)
complexes of the tren family of ligands,13 the axial Cu-N
distance in each complex is shorter than the average
equatorial distance (Table 5), and the angles between the
axial nitrogen, Cu(II), and the equatorial nitrogen atoms are
Figure 4. Hydrogen bonding in [Cu(Me3tren)(MeCN)](ClO4)2 (non-amine
hydrogens omitted).
(23) Freire, E.; Baggio, S.; Baggio, R.; Suescun, L. Acta Crystallogr., Sect.
C: Cryst. Struct. Commun. 1999, 55, 1780-1784.
neutral, monomeric coordination complex and a solvent
molecule. The Me3tren complex, however, contains two
inequivalent [Cu(Me3tren)(S2O3)] moieties (Figure 7), a triply
protonated Me3tren and three perchlorate anions. One of the
(24) Teng, S. T.; Fuess, H.; Bats, J. W. Acta Crystallogr., Sect. C: Cryst.
Struct. Commun. 1984, 40, 1787-1789.
(25) Cotton, A. F.; Wilkinson, G. AdVanced Inorganic Chemistry, 5th ed.;
Wiley-Interscience: New York, 1988.
Inorganic Chemistry, Vol. 43, No. 21, 2004 6571

Fischmann et al.
Table 4. Comparison of Geometric Parameters for Selected Thiosulfate Complexesa
S-S bond
mean S-O bond
distance (Å)
M-S bond
distance (Å)
complex
geometry
distance (Å)
ref
[Cu(tren)(S2O3)]‚H2O
TBP
TBP
2.051(1)
2.042(3),
2.014(3)
2.061(2)
2.05
1.464
1.459
2.316(1)
2.330(3),
2.280(3)
2.301(1)
2.71
this study
this study
[Cu(Me3tren)(S2O3)]
[Cu(Bz3tren)(S2O3)]‚MeOH
[Cu(en)2S2O3]
TBP
SP
1.460
1.47
this study
17
[Ni(phen)(OH2)3(S2O3)]‚H2O
O
O
T
2.003(1)
2.011†
2.048(3),
2.030(2)
2.061(6),
2.072(6)
1.4731
1.460†
1.449
2.449(1)
2.488†
2.252(2),
2.275(8)
2.282(6),
2.312(6)
23
18
19
[N(CH3)4]2[Co(H2O)4(S2O3)2]
[Zn(phen)(H2O)2][Zn(phen)(S2O3)2]‚H2O
[Pd(en)2][Pd(en)(S2O3)2]
P
1.457(7)
21
Na2S2O3
2.0054(4)
1.4765
2.9638#
24
a Average (#); identical for both thiosulfates (†). O ) octahedral; T ) tetrahedral; SP ) square pyramidal; P ) square planar; and TBP ) trigonal
bipyramidal.
Figure 7. ORTEP view of the two inequivalent complexes (a and b) in
(H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3 (hydrogens omitted for clarity).
N(2′) and C(26) in complex b were refined isotropically and are shown
without the octant slice.
Figure 5. ORTEP view of the complex in [Cu(tren)(S2O3)]‚H2O (hydro-
gens omitted for clarity).
Figure 8. H-bonding interactions involving (H3Me3tren)3+ (most hydro-
gens omitted for clarity). The major conformation (66%) of (H3Me3tren)3+
is shown.
Figure 6. ORTEP view of [Cu(Bz3tren)(S2O3)]‚MeOH (hydrogens omitted
for clarity).
less than 90°, which position each Cu(II) atom out of the
plane defined by the equatorial nitrogen atoms toward the
thiosulfate (Table 6). Compared to those of the tren complex,
the axial Cu-N bonds in the Me3tren and Bz3tren complexes
are shorter, yet the equatorial Cu-N bonds are generally
longer. Such a trend is consistent with that observed by
Komiyama et al.26 (who studied the Cu(II)-chloride com-
plexes of Bz3tren and Me3Bz3tren), in which the average Cu-
Neq distance increased in the order tmpa < tren < Bz3tren
< Me3Bz3tren ≈ Me6tren; Viz., the distance increased as the
steric bulk of the terminal nitrogen donors increased from
heterocyclic to primary, secondary, and tertiary amines.
Me3tren can be added to this trend as being approximately
equal to Bz3tren as a result of this work, as would be
expected for a terminal donor set consisting of secondary
amine nitrogens.
In each [Cu(Me3tren)(S2O3)] unit of (H3Me3tren)[Cu(Me3-
tren)(S2O3)]2(ClO4)3 and [Cu(Bz3tren)(S2O3)]‚MeOH, the
alkyl groups on the same side of the complex as the
thiosulfate-SO3 group are oriented away from the thiosul-
fate, whereas in [Cu(Me3tren)(MeCN)](ClO4)2 and [Cu(Bz3-
tren)Cl]Cl,27 where the axial ligand is symmetrical, the alkyl
groups are arranged symmetrically around the Cu(II) center
(Figure 9).
(26) Komiyama, K.; Furutachi, H.; Nagatomo, S.; Hashimoto, A.; Hayashi,
H.; Fujinami, S.; Suzuki, M.; Kitagawa, T. Bull. Chem. Soc. Jpn. 2004,
77, 59-72.
6572 Inorganic Chemistry, Vol. 43, No. 21, 2004

Cu(II)-Thiosulfate Complexes with Tetraamine Ligands
Table 5. Selected Bond Lengths (Å) for Cu(II)-L-S2O3 Complexes
bonda
Cu-Nax Cu-N(1) 2.075(2)
tren
Bz3tren Me3tren (a) Me3tren (b)
2.064(3)
2.052(9)
2.075(8)
Cu-Neq Cu-N(2)
Cu-N(3)
2.047(2)
2.062(2)
2.112(2)
2.098(4)
2.100(4)
2.168(4)
2.041(8)
2.103(8)
2.194(7)
2.08(2)
2.09(1)
2.10(1)
Cu-N(4)
Cu-S
S-S
Cu-S(1)
2.3158(8) 2.301(1)
2.329(3)
2.044(3)
2.281(3)
2.014(3)
S(2)-S(1) 2.0513(8) 2.061(2)
S-O
S(2)-O(1) 1.464(2)
S(2)-O(2) 1.461(2)
S(2)-O(3) 1.465(2)
1.465(3)
1.461(3)
1.455(3)
1.496(7)
1.460(6)
1.440(7)
1.434(7)
1.464(7)
1.458(7)
a Nax ) axial nitrogen; Neq ) equatorial nitrogen.
Figure 9. Orientation of methyl groups around the Cu(II) center in Me3-
tren complexes: (a) [Cu(Me3tren)(MeCN)](ClO4)2; (b) (H3Me3tren)[Cu-
(Me3tren)(S2O3)]2(ClO4)3.
Figure 10. (a) H-bond network and (b) crystal packing showing H bonding
between sheets in [Cu(tren)(S2O3)]‚H2O.
H-Bonding Interactions in Thiosulfate Complexes. In
each complex, intramolecular H bonding exists between one
or more of the N-H groups on the amine ligand and one or
more of the oxygen atoms on the coordinated thiosulfate. In
[Cu(tren)(S2O3)]‚H2O, further H bonding between adjacent
[Cu(tren)(S2O3)] units forms two-dimensional sheets that
pack along the bc plane and are separated from each other
by H-bonded layers of water (Figure 10 and Table 7).
Similarly, (H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3 ex-
hibits intramolecular thiosulfate-amine H bonds in addition
to the H-bonding interaction between H3Me3tren3+ and O(2)
described above. In this complex, the perchlorate anions are
found to be H bonded to amine protons on adjacent
complexes.
Potentiometric Titrations. The acid dissociation constants
for protonated Bz3tren, (H4Bz3tren)4+, and Cu(II)-L complex
formation constants were determined by potentiometric pH
titrations, as described in the Experimental Section. HY-
PERQUAD28 was used to analyze the data, and the results
are shown in Table 8.
In [Cu(Bz3tren)(S2O3)]‚MeOH, the MeOH molecule bridges
O(2) and H(3) through H bonding, and another H bond
between H(2) and O(1) completes a 10-member ring (Figure
11a). Furthermore, [Cu(Bz3tren)(S2O3)] units H bond to each
other from H(3) to O(3), forming one-dimensional chains
along the b axis. Figure 11b shows that although there is no
direct interaction between the chains, they pack into pseu-
dosheets along the ab plane. The hydrophobic groups
(benzene rings and methyl group of the methanol) project
out in the c direction. Figure 11c shows that the benzyl
groups on adjacent pseudosheets are weakly intercalated,
namely, the slight interdigitation of the aromatic ring
hydrogens.
The substitution of benzyl groups for methyl groups
dramatically lowers the pKa values of the secondary nitro-
gens, and this lower affinity for protons is also reflected in
the lower stability constant of the Cu(II) complex. The pKa
of the coordinated water molecule is very low compared to
those of the other complexes. This is possibly due to the
proximity of the hydrophobic benzyl groups which in Zn-
Table 6. Selected Bond Angles (deg) for Cu(II)-L-S2O3 Complexes
anglea
tren
Bz3tren
Me3tren (a)
Me3tren (b)
Neq-Cu-Nax
Neq-Cu-Neq
Neq-Cu-S
N(1)-Cu(1)-N(4)
N(3)-Cu(1)-N(1)
N(2)-Cu(1)-N(1)
N(3)-Cu(1)-N(4)
N(2)-Cu(1)-N(4)
N(2)-Cu(1)-N(3)
N(2)-Cu(1)-S(1)
N(4)-Cu(1)-S(1)
N(3)-Cu(1)-S(1)
N(1)-Cu(1)-S(1)
S(2)-S(1)-Cu(1)
83.67(8)
83.76(7)
84.76(7)
112.97(8)
117.01(8)
126.85(8)
91.22(6)
96.85(6)
99.90(6)
175.71(5)
103.28(3)
83.9(1)
83.8(1)
84.1(1)
114.0(2)
113.8(2)
129.0(2)
94.3(1)
97.1(1)
97.0(1)
178.3(1)
102.52(6)
83.8(3)
84.3(4)
85.3(4)
106.9(3)
111.1(3)
139.2(4)
90.6(3)
102.8(2)
95.3(2)
173.1(3)
97.1(1)
81.7(4)
83.1(4)
84.9(5)
112.9(4)
120.1(5)
123.0(5)
89.7(4)
101.6(3)
99.4(3)
Nax-Cu-S
Cu-S-S
174.5(3)
109.5(1)
a Nax ) axial nitrogen; Neq ) equatorial nitrogen.
Inorganic Chemistry, Vol. 43, No. 21, 2004 6573

Fischmann et al.
Table 7. Hydrogen Bonding in [Cu(tren)(S2O3)]‚H2O (I),
[Cu(Bz3tren)(S2O3)]‚MeOH (II), and (H3Me3tren)[Cu(Me3tren)-
(S2O3)]2(ClO4)3 (III) (bond distances in Å and angles in deg)
complex
bonda
D-H H‚‚‚A D‚‚‚A <(DHA)
I
N(4)-H(5)‚‚‚O(3)1
N(2)-H(6)‚‚‚O(3)2
N(4)-H(8)‚‚‚O(2)
N(3)-H(9)‚‚‚S(1)2
N(2)-H(14)‚‚‚O(2)1
N(3)-H(18)‚‚‚S(1)3
0.83(3) 2.18(3) 2.991(3) 164(2)
0.82(3) 2.17(3) 2.986(3) 173(2)
0.87(3) 2.34(3) 3.077(3) 143(2)
0.80(3) 2.75(3) 3.439(2) 145(2)
0.80(3) 2.17(3) 2.939(3) 159(2)
0.83(2) 2.81(2) 3.564(2) 152(2)
O(1W)-H(1W)‚‚‚O(1)4 0.81(4) 2.18(4) 2.990(3) 176(4)
O(1W)-H(2W)‚‚‚O(1) 0.82(3) 2.08(3) 2.901(3) 174(3)
II
N(4)-H(4)‚‚‚O(3)5
N(3)-H(3)‚‚‚O(5)
N(2)-H(2)‚‚‚O(1)
O(5)-H(5)‚‚‚O(2)
0.80(4) 2.35(4) 3.062(5) 149(4)
0.80(3) 2.41(4) 3.155(5) 155(3)
0.83(4) 2.06(4) 2.866(5) 163(3)
0.88(6) 1.94(6) 2.795(5) 162(5)
III
N(12)-H(12C)‚‚‚O(2) 0.92
N(10)-H(10D)‚‚‚O(11) 0.92
2.11
2.18
2.34
2.88
2.29
2.26
2.15
1.97
2.13
2.67
2.22
2.11
2.77(1)
3.01(1)
3.09(1)
3.113(8)
2.95(1)
3.18(1)
2.96(1)
2.87(1)
2.98(1)
3.55(1)
3.11(2)
2.77(1)
127.1
149.1
137.2
95.8
127.9
174.6
145.8
166
154.1
157
160.3
127.1
N(3′)-H(3′)‚‚‚O(1′)
N(2)-H(2)‚‚‚S(1)
N(3)-H(3)‚‚‚O(1)
0.93
0.93
0.93
N(12)-H(12D)‚‚‚O(1′)6 0.92
N(11)-H(11C)‚‚‚O(1′)6 0.92
N(11)-H(11D)‚‚‚O(2′)7 0.92
N(10)-H(10C)‚‚‚O(2′)7 0.92
N(4′)-H(4′)‚‚‚S(1)8
N(4)-H(4)‚‚‚O(20)9
0.93
0.93
N(12)-H(12C)‚‚‚O(2) 0.92
a Symmetry transformations used to generate equivalent atoms. (1) -x
1
3
1
+ 1, y + /2, -z + /2; (2) -x + 1, -y, -z + 1; (3) -x + 1, y - /2, -z
3
1
1
+ /2; (4) -x, -y, -z + 1; (5) -x + /2, y - /2, z; (6) x + 1, y, z; (7) x
+ /2, -y - /2, -z; (8) x - 1, y, z; (9) -x + 1, y - /2, -z + /2. D )
donor atom; A ) acceptor atom.
1
1
1
1
(II) complexes and enzymes is responsible for increasing the
acidity of coordinated water.29
Thiosulfate Binding Constants. The striking spectral
changes in the 300-450 nm region were used to determine
the thiosulfate binding constant to [Cu(L)(H2O)]2+ (Figure
12). This was preferred to the minor changes in the visible
spectrum due to the d-d bands.
Matrix decomposition was applied to the absorbance data
to determine the spectrum and concentration of each absorb-
ing complex. The EQBRM program30 was used through the
Matlab interface to calculate speciation from a chemical
model consisting of mass-balance (including charge balance)
-
and mass-action equations for Na+, ClO4 , NaClO4, MOPS,
MOPS-, NaMOPS, S2O32-, [Cu(L)(H2O)]2+, and [Cu(L)-
(S2O3)], where appropriate. The procedure has been recently
described in detail by Brugger et al.31 Examples of the fit to
the data are given for [Cu(tren)(H2O)]2+ and [Cu(tren)(S2O3)]
(Figure 12), and the calculated molar absorbance spectrum
of each thiosulfate complex is shown in Figure 13. The
calculated position of the S-Cu(II) LMCT maximum was
in agreement with the diffuse reflectance of the UV-visible
spectrum of the corresponding solid compound {note that
Figure 11. (a) Moderately strong H bonding forming the 10-member ring
(most hydrogens omitted for clarity), (b) packing of chains along the ab
plane into a pseudosheet (most hydrogens omitted for clarity), and (c) weak
intercalation of aromatic rings between pseudosheets in [Cu(Bz3tren)(S2O3)]‚
MeOH.
(27) Schatz, M.; Becker, M.; Walter, O.; Liehr, G.; Schindler, S. Inorg.
Chim. Acta 2001, 324, 173-179.
no comparison is possible for [Cu(Me6tren)(S2O3)], for which
no solid product could be obtained}.
The formation constants for [Cu(tren)(S2O3)] [Kf ) (1.82
( 0.09) × 103 M-1], [Cu(Me3tren)(S2O3)] [(2.13 ( 0.05) ×
103 M-1], and [Cu(Bz3tren)(S2O3)] [(4.30 ( 0.21) × 104
M-1] are higher than those for any other divalent transition-
(28) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739-1753.
(29) Mareque-Rivas, J. C.; Prabaharan, R.; Parsons, S. J. Chem. Soc., Dalton
Trans. 2004, 1648-1655.
(30) Anderson, G. M.; Crerar, D. A. Thermodynamics in Geochemistry:
The Equilibrium Model; Oxford University Press: New York, 1993.
(31) Brugger, J.; McPhail, D. C.; Black, J.; Spiccia, L. Geochim. Cosmo-
chim. Acta 2001, 65, 2691-2708.
6574 Inorganic Chemistry, Vol. 43, No. 21, 2004

Cu(II)-Thiosulfate Complexes with Tetraamine Ligands
Table 8. pKa Values of tren-Based Ligands and Formation Constants for [CuL(H2O)]2+ and [CuLH(H2O)]3+ and pKa Values for the Ligated Water in
[CuL(H2O)]2+ [25.0 ( 0.1 °C, I ) 1 M (NaClO4)]a
ligand
tren
Me3tren
Bz3trenb
Me6tren
pKa1
pKa2
pKa3
pKa4
<1
<1
1.2(1)
<1
log KCuL
log KCuLH
pKa(CuL)
ref
10.42(1)
10.93(1)
9.16(1)
9.88(1)
10.24(1)
8.56(2)
9.32(1)
8.915(5)
9.17(1)
7.17(3)
8.17(1)
19.58(3)
19.11(2)
15.10(4)
15.65(3)
13.22(4)
9.4
12
13
9.08(2)
7.10(9)
8.1
3.77(7)
9.53(4)
this study
12
10.13(1)
a Numbers in parentheses are the standard deviations in the last digit. b I ) 0.1 M (NaNO3); KCuL ) [CuL2+]/([Cu2+][L]); KCuLH ) [Cu(LH)3+]/([Cu2+][LH+]).
Figure 13. Calculated spectra of [Cu(L)(S2O3)] complexes; L ) tren (s)
λmax 356 nm, Me3tren (- -) λmax 364 nm, Bz3tren (- ‚ -) λmax 367 nm,
and Me6tren (- ‚‚ -) λmax 410 nm.
10 M-1), and that it oxidizes thiosulfate in the presence of
an excess of thiosulfate. Golub et al.11 concluded from their
study of anions binding to [Cu(tren)(H2O)]2+ and [Cu(Me6-
tren)(H2O)]2+ that N-methylation of tren caused the copper-
(II) center to behave as a stronger and harder Lewis acid.
The thiosulfate binding constants are in agreement, and the
harder Cu(II) ion in [Cu(Me6tren)(H2O)]2+ has the weakest
binding interaction with the soft thiosulfate-S. However, the
Bz3tren binding constant seems anomalous because [Cu(Bz3-
tren)(H2O)]2+ is more acidic than [Cu(Me6tren)(H2O)]2+ (7.1
vs 8.3), but in this case, the aromatic groups may be
influencing the acidity of the coordinated water, as seen in
the Zn(II) complexes mentioned above,29 and the interaction
of thiosulfate with the copper(II) center.
Figure 12. Changes in UV spectrum with thiosulfate concentration:
measured data (s) and fit (+) for (a) [Cu(tren)(H2O)]2+ and (b) [Cu(Bz3-
tren)(H2O)]2+
.
The mechanism of the electron-transfer process that gives
rise to oxidation of thiosulfate and concurrent reduction of
Cu(II) is unclear as there are literature examples of both
inner-sphere and outer-sphere reactions. The fact that the
LMCT for the Me6tren complex is at lower energy indicates
that electron transfer within this system is more facile, while
the lower binding constant means that the concentrations of
free [Cu(Me6tren)(H2O)]2+ and S2O32- in solution would be
higher for this system than for the others; therefore, the rate
of an outer-sphere electron-transfer process involving the
reaction between [Cu(Me6tren)(H2O)]2+ and either the [Cu-
(Me6tren)(S2O3)] complex or free S2O32- would be enhanced.
A detailed kinetic investigation of this reaction is in progress
to attempt to elucidate the mechanism of these complex
reactions.
metal ion (Table 9) and, in particular, higher than that for
[Cu(en)2(S2O3)] (the most closely related complex shown).
The only other reported Cu(II)-thiosulfate formation con-
stant is the â2 value for [Cu(S2O3)2]2-, which cannot be
compared directly with Kf for [Cu(L)(S2O3)]. The thiosulfate
binding constants for other divalent transition-metal ions
(Ni2+, Zn2+, Co2+, and Mn2+) are 1-2 orders of magnitude
smaller than those for the above Cu(II) complexes.
There is no clear correlation between the measured stability
constants and crystallographic measurements of bonding
strength (i.e., S-S and Cu-S distances) and the location of
the Cu(II) center relative to the tripodal plane. Crystal
packing effects may be masking any relationship that may
be present.
[Cu(Me6tren)(S2O3)] differs from the other three com-
plexes in that its formation constant is much lower (40 (
The Me3tren and the Bz3tren complexes not only exhibit
the highest thiosulfate binding constants of the four com-
Inorganic Chemistry, Vol. 43, No. 21, 2004 6575

Fischmann et al.
Table 9. Comparison of Constants for the Binding of Thiosulfate to Divalent Metal Ions at 25 °C
reaction
methoda
ionic strength
log Kf
4.63
3.31
3.26
1.60
ref
[Cu(Bz3tren)(H2O)]2+ + S2O32- / [Cu(Bz3tren)(S2O3)] + H2O
[Cu(Me3tren)(H2O)]2+ + S2O32- / [Cu(Me3tren)(S2O3)] + H2O
[Cu(tren)(H2O)]2+ + S2O32- / [Cu(tren)(S2O3)] + H2O
spec
spec
spec
spec
0.03-0.04 M
0.15 M
0.25 M
this study
this study
this study
this study
[Cu(Me6tren)(H2O)]2+ + S2O32- / [Cu(Me6tren)(S2O3)] + H2O
0.15 M
Cu2+ + 2S2O32- / [Cu(S2O3)2]2-
spec
0.2 M
4.56b
6
[Cu(en)2(H2O)2]2+ + S2O32- / [Cu(en)2(S2O3)] + 2H2O
spec
spec
0.0 Mc
0.0 Mc
2.28, 2.34
2.2-2.5
7, 8
9
Ni2+ + S2O32- / [Ni(S2O3)]
Zn2+ + S2O32- / [Zn(S2O3)]
cal
0.50 M
0.78
32
cal
ext
pot
pot
0.5 M
1.00 M
2.0 M
3.0 M
1.12
0.62
1.10
0.96
32
33
34
35
Co2+ + S2O32- / [Co(S2O3)]
Mn2+ + S2O32- / [Mn(S2O3)]
cal
cal
0.50 M
0.5 M
0.77
0.67
32
32
a Spec ) spectrophotometric, cal ) calorimetric, ext ) solvent extraction, and pot ) potentiometric. b Value is log10 â2, not log10 Kf. c Results of experiments
at different ionic strengths extrapolated to I ) 0.
from commercial suppliers and were of analytical grade unless
otherwise indicated. Water was freshly distilled prior to use, and
CO2-free water was prepared by boiling distilled water under
nitrogen for 2 h.
Instrumentation. UV-vis-NIR spectra were recorded on a
Varian Cary 5G spectrophotometer fitted with a water-jacketed cell
holder. An externally circulating water bath (Varian) maintained
the temperature to a precision of (0.1 °C. Quartz cells (1 cm) were
first rinsed with distilled water and then with the sample prior to
use.
Solution pH was measured on either a Metrohm electrode, fitted
to a Metrohm pH meter, or a Eutech electrode, fitted to an EcoScan
pH 5-6 meter. A Metrohm 736 GP Titrino was used to dispense
the small volumes of NaOH solution required to fix the pH of the
samples for [Cu(tren)(S2O3)]. All pH measurements were made at
25 °C (temperature maintained by a Haake W 19 water bath). The
electrodes were calibrated using standard phthalate, phosphate, and
borate buffer solutions.
plexes measured but also are the least reactive in terms of
thiosulfate oxidation. This observation suggests that the
oxidation reaction may not be proceeding via an intramo-
lecular electron-transfer process; otherwise, an increase in
reactivity would be expected. Moreover, the intermolecular
or outer-sphere pathway is blocked because the strong
binding of thiosulfate leaves very little free [Cu(L)(H2O)]2+
in solution to react with either free thiosulfate or the
thiosulfate complex. Due to the low solubility of [Cu(Bz3-
tren)(H2O)]2+, whether thiosulfate oxidation is possible when
an excess of the Cu(II) complex is present, this theory cannot
be tested as it has been for tren.5
Conclusions
The results presented herein demonstrate that relatively
minor modifications of tris(2-aminoethyl)amine, achieved
through N-alkylation, result in copper(II) complexes with
significantly different thiosulfate binding characteristics and
abilities to promote the oxidation of thiosulfate. The much
higher reactivity, lower stability, and lower-energy LMCT
transition of [Cu(Me6tren)(S2O3)], when compared with those
of the other three complexes, have yet to be fully rationalized.
The origin of this behavior will be further explored through
studies of the copper(II) complexes of other tren derivatives
with all tertiary amine donors. This work demonstrates that
large differences in reactivity between Cu(II) and thiosulfate
can be brought about through functionalization of amine
ligands. Such a result may be promising for gold processing
applications; if a Cu(II)-L complex is found that is satisfac-
tory in terms of gold oxidation properties and the thiosulfate
oxidation behavior inhibited through the choice of spectator
ligand, then the potential exists for developing improved gold
leaching agents.
IR spectra were recorded on a Perkin-Elmer 1600 spectrometer
as KBr disks or Nujol mulls. Elemental analyses were performed
by Campbell Microanalytical Service, Otago University, Otago,
New Zealand. 1H NMR spectra were recorded in CDCl3 on either
a Bruker AC200 or Bruker DPX300 spectrometer. The residual
solvent peak was used as an internal standard. Mass spectra were
recorded on a Biomass Platform II mass spectrometer using an
electrospray ionization source.
1
Me6tren. Yield: 20%. H NMR δ (ppm): 2.23 [-N(CH3)2, s,
18H], 2.35, 2.38, 2.40, 2.42, 2.59, 2.61, 2.63, 2.66 (-CH2CH2-,
m, 12H). ESI-MS m/z (ion, %): MH+ 231.1 (100), MNa+ 253.2
(15), MK+ 269.1 (8).
Bz3tren. Yield: 66%. 1H NMR δ (ppm): 1.64 (NH, br s), 2.55-
2.71 (m, 12H, CH2CH2), 3.74 (6H, CH2Ph), 7.21-7.33 (m, C6H5),
(32) Aruga, R. J. Inorg. Nucl. Chem. 1974, 36, 3779-3782.
(33) Moriya, H.; Sekine, T. Bull. Chem. Soc. Jpn. 1974, 47, 747-748.
(34) Nazarova, L.; Efremova, T.; Orenshtein, S. Russ. J. Inorg. Chem. 1972,
17, 186-188.
(35) Persson, H. Acta Chem. Scand. 1970, 24, 3739-3750.
(36) Giumanini, A. G.; Chiavari, G.; Scarponi, F. L. Anal. Chem. 1976,
48, 484-489.
Experimental Section
(37) Ciampolini, M.; Nardi, N. Inorg. Chem. 1966, 5, 41-44.
(38) Naiini, A. A.; Menge, W. M. P. B.; Verkade, J. G. Inorg. Chem. 1991,
30, 5009-5012.
Materials and Reagents. Me6tren,36,37 Bz3tren,38 [Cu(tren)(H2O)]-
40
(ClO4)2,39 and [Cu(Me6tren)(H2O)](ClO4)2 were synthesized by
(39) Fry, F. Polynuclear Metal Complexes of Macrocyclic Ligands and
their Ability to Hydrolyse Phosphate Esters. Ph.D. Thesis, Monash
University, Melbourne, AU, 2002.
published procedures with minor modifications. The synthesis of
[Cu(Me3tren)(H2O)](ClO4)2 was based on the method used for [Cu-
(Me6tren)(H2O)](ClO4)2. All other chemicals were used as received
(40) Lee, S. C.; Holm, R. H. J. Am. Chem. Soc. 1993, 115, 11789-11798.
6576 Inorganic Chemistry, Vol. 43, No. 21, 2004

Cu(II)-Thiosulfate Complexes with Tetraamine Ligands
7.11-7.36 (m). ESI-MS m/z (ion, %): 417 MH+ (>98%); minor
peak at 327 indicates Bz2tren H+ (<2%).
10.8. Calcd for C12H32Cl2CuN4O9 (%): C, 28.3; H, 6.3; N, 11.0.
Selected IR bands [KBr, ν (cm-1)]: 3449 (br), 2981 (m), 2898
(m), 2850 (m), 1474 (s), 1148 (s), 1094 (s), 1005 (m), 624 (s).
Diffuse reflectance UV-vis-NIR spectrum [λmax, nm (reflectance,
%)]: 871 (8), 700 sh (16), 330 (5), 280 sh (7). Solution UV-vis-
NIR spectrum (H2O) [λmax, nm (molar absorbance, M-1 cm-1)]:
290 (4270), 710 sh (160), 875 (450).
[Cu(Me3tren)(H2O)](ClO4)2. Synthesis was based on the method
published for [Cu(Me6tren)(H2O)](ClO4)2.40 Typically, Me3tren
(Aldrich; 0.943 g, 5 mmol) was reacted with a hot solution of Cu-
(ClO4)2‚6H2O (1.857 g, 5 mmol) in 20 mL of EtOH. Yield: 1.526
g (65%). Characterization microanalyses. Found (%): C, 23.5; H,
5.7; N, 12.1. Calcd for C9H26Cl2CuN4O9 (%): C, 23.1; H, 5.6; N,
12.0. Selected IR bands [KBr, ν (cm-1)]: 3500 (br), 3273 (m),
2882 (m), 1473 (m), 1091 (s), 626 (s). Diffuse reflectance UV-
vis-NIR spectrum [λmax, nm (reflectance, %)]: 871 (9), 669 (9),
[Cu(tren)(S2O3)]‚H2O. Copper perchlorate hexahydrate (0.579
g, 1.56 mmol) and tris(2-aminoethyl)amine (0.240 g, 1.64 mmol)
were dissolved in a minimum of distilled water and formed a deep-
blue solution. Upon addition of sodium thiosulfate pentahydrate
(0.386 g, 1.56 mmol), the solution became dark green. Some solid
remained in the bottom of the vial. A sample of the supernatant
solution (∼1.5 mL) was taken and left to evaporate in air over a
period of 3 weeks. The blue and white solid residue was then
dissolved in a minimum of distilled water, and EtOH was added to
give a green solution. This was sealed and placed in the freezer.
After 10 days, small blue and green crystals had formed. One blue
and one green crystal were analyzed by single-crystal XRD, and
both were found to be [Cu(tren)(S2O3)]‚H2O.
A second batch of the complex was made by dissolving [Cu-
(tren)(H2O)](ClO4)2 (0.306 g, 0.7 mmol) in a minimum of distilled
water and by adding Na2S2O3‚5H2O (0.175 g, 0.7 mmol), dissolved
in a minimum volume of distilled water. EtOH was added until
precipitation of a fine green solid occurred, and the suspension was
placed in the freezer. Flaky green crystals formed overnight. These
were washed with Et2O (80 mL) and dried in a desiccator. Yield
) 0.162 g (67%). At this point, the product was analyzed by IR
spectroscopy (results are the same as for the sample obtained above).
The flaky green crystals were redissolved in a minimum quantity
of distilled water and placed in an evaporating dish. Diamond-
shaped dark-green crystals, with the same unit-cell dimensions as
those above, formed overnight on evaporation of the water.
Characterization microanalyses. Found (%): C, 21.4; H, 5.9; N,
16.4. Calcd for C6H20N4O4S2Cu (%): C, 21.2; H, 5.9; N, 16.5.
Selected IR bands [KBr, ν (cm-1)]: 3548 (s), 3478 (s), 3269 (s),
3159 (s), 2880 (s), 1612 (m), 1473 (m), 1154 (s), 1066 (m), 1008
(s), 902 (m), 638 (s), 459 (m). Diffuse reflectance UV-vis-NIR
spectrum [λmax, nm (reflectance, %)]: 280 (8.5), 360 (4.5), 655
(15), 860 (19.5). Solution UV-vis-NIR spectrum (H2O) [λmax, nm
(molar absorbance, M-1 cm-1)]: 257 (2930), 356 (4910), 686 (110),
853 (130). ESI-MS m/z (ion, %) H2O\CH3OH: 209 (100), 211 (43)
[Cu(tren)]+; 344 (94), 346 (51) [Cu(tren)(S2O3)]‚Na+; 322 (4), 344
(3) [Cu(tren)(S2O3)]‚H+.
[Cu(Bz3tren)(S2O3)]‚MeOH. Bz3tren (0.205 g, 0.5 mmol) was
dissolved in 10 mL of a 1:1 H2O/MeOH mixture. To this solution
were added Cu(ClO4)2‚6H2O (0.185 g, 0.5 mmol) and Na2S2O3‚
5H2O (1.239 g, 5 mmol). Immediately when the thiosulfate was
added, a dark-green solid formed and was separated by vacuum
filtration. The dark-green filtrate was left to stand (covered) for 2
days, whereupon fine dark-green needles (unsuitable for crystal-
lography) and plates (used in X-ray structure determination) formed.
Characterization (needles and plates) microanalyses. Found (%):
C, 51.7; H, 5.9; N, 8.8. Calcd for C27H40CuN4O5S2 (i.e., replacement
of MeOH with 2 × H2O): C, 51.7; H, 6.4; N, 8.9. Selected IR
bands [KBr, ν (cm-1)]: 3423 (br), 3230 (m), 2926 (w), 2897 (w),
1619 (m), 1457 (m), 1157 (s), 1000 (s), 743 (m), 703 (m), 640 (s).
Diffuse reflectance UV-vis-NIR spectrum [λmax, nm (reflectance,
%)]: 868 (9), 630 (7), 403 (5), 302 (7). Solution UV-vis-NIR
spectrum (MeCN) [λmax, nm (molar absorbance, M-1 cm-1)]: 268
(4860), 275 (4900), 287 (4930), 391 (5480), 637 (330), 871 (330).
[Cu(Me6tren)(H2O)](ClO4)2. Synthesis followed a published
method40 with minor modifications to the workup. Typically, to a
hot solution of Cu(ClO4)2‚6H2O (2.785 g, 7.5 mmol) in 30 mL of
EtOH (pale-blue solution) was added a solution of Me6tren (1.743
g, 7.6 mmol) in 18 mL of EtOH, resulting in a dark-blue solution,
which gave the product on workup. Yield: 1.117 g (29%).
Characterization microanalyses. Found (%): C, 28.2; H, 6.5; N,
320 (5), 263 (8). Solution UV-vis-NIR spectrum (MeCN) [λmax
,
nm (molar absorbance, M-1 cm-1)]: 274 (3770), 690 sh (100), 854
(230).
[Cu(Me3tren)(MeCN)](ClO4)2. Slow diffusion of Et2O into 6
mL of a MeCN solution of [Cu(Me3tren)(H2O)](ClO4)2 (0.05 g,
0.1 mmol) yielded dark-blue crystals after 3 weeks. Yield: 0.019
g (38%). Characterization. Selected IR bands [Nujol, ν (cm-1)]:
3259 (s), 2723 (m), 2319 (s), 2291 (s), 2022 (s), 1715 (w), 1660
(w), 1294 (s), 1220 (s), 1086 (br), 621 (s). Diffuse reflectance UV-
vis-NIR spectrum [λmax, nm (reflectance, %)]: 323 (10), 664 (14),
847 (15). Solution UV-vis-NIR spectrum (MeCN) [λmax, nm
(molar absorbance, M-1 cm-1)]: 284 (7480), 646 sh (90), 820 (270).
(H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3. [Cu(Me3tren)(H2O)]-
(ClO4)2 (0.091 g, 0.2 mmol) was dissolved in 6 mL of MeOH to
give a bright-blue solution. Na2S2O3‚5H2O (0.049 g, 0.2 mmol)
was added to [Cu(Me3tren)(H2O)](ClO4)2 (0.091 g, 0.2 mmol) in 6
mL of MeOH (bright blue), and the color changed to dark green.
Not all of the thiosulfate dissolved. Dark-green crystals formed upon
diffusion of Et2O into this solution. Yield: 0.049 g (40%).
Microanalyses. Found (%): C, 26.8; H, 5.9; N, 13.8. Calcd for
C27H75Cl3Cu2N12O18S4 (%): C, 26.6; H, 6.2; N, 13.8. Selected IR
bands [KBr, ν (cm-1)]: 3449 (br), 3263 (s), 2926 (m), 2888 (m),
1477 (m), 1457 (m), 1192 (s), 1146 (sh), 1087 (s), 990 (s), 859
(m), 830 (m), 649 (s), 623 (s). Diffuse reflectance UV-vis-NIR
spectrum [λmax, nm (reflectance, %)]: 290 (5), 379 (4), 684 (17),
894 (11). Solution UV-vis-NIR spectrum (MeCN) [λmax, nm
(molar absorbance, M-1 cm-1)]: 282 (8600), 383 (6190), 686 sh
(465), 850 (724).
X-ray Crystallography. Single-crystal X-ray data were collected
on an Enraf-Nonius CAD4 diffractometer with monochromated Mo
ΚR radiation (λ ) 0.71073 Å) at 123(2) K using æ and/or ω scans.
Data were corrected for Lorentz and polarization effects, and
absorption corrections were applied. The structure was solved by
the direct methods and refined using the full-matrix least-squares
method of the programs SHELXS-9741 and SHELXL-97,42 respec-
tively. The program X-Seed43 was used as an interface to the
SHELX programs and to prepare the figures. Most hydrogen atoms
were located on Fourier difference maps and refined isotropically;
otherwise, they were modeled in idealized geometries. All of the
non-hydrogen atoms were refined anisotropically, except O(22),
O(30)-(33), N(2′), and C(26) in [Cu(Me3tren)(S2O3)]2H3Me3tren-
(ClO4)3, which were modeled isotropically. C(30) and C(31), C(34)
(41) Sheldrick, G. M. SHELXS-97; University of Go¨ttingen: Go¨ttingen,
Germany, 1997.
(42) Sheldrick, G. M. SHELXL-97; University of Go¨ttingen: Go¨ttingen,
Germany, 1997.
(43) Barbour, L. J. J. Supramol. Chem. 1999, 1, 189-191.
Inorganic Chemistry, Vol. 43, No. 21, 2004 6577

Fischmann et al.
Table 10. Concentrations Used for Potentiometric Titrations
concentrations in the 0.5-50 mM range, which were stored in a
refrigerator until required. Samples for spectrophotometric measure-
ment were prepared by mixing 2 mL of each thiosulfate solution
with 2 mL of the Cu(II) complex stock solution. The UV-visible
spectrum of each sample was measured from 200 to 500 nm with
the appropriate background solution in the reference beam. The
spectra of the Cu(II)-free background solutions were measured and
subtracted from the measured spectra.
volume ratio (Cu/L)
0
1:2
1:1.3
1:1
[Cu(II)]total (mM)
[Bz3tren]total (mM)
0.0
5.0
1.7
3.3
2.1
2.9
2.5
2.5
and C(35), and C(38) and C(39) were refined as being disordered
across two positions in a 2:1 ratio, respectively.
Potentiometric Titrations. Solutions for pH tirations were
prepared using CO2-free water, and each titration was performed
in duplicate in a sealed vessel under a nitrogen atmosphere. NaNO3
and Cu(NO3)2‚2.5H2O were used due to the low solubility of [Cu-
(Bz3tren)(H2O)](ClO4)2 in the presence of the slightest excess of
ClO4-. The volume of the Bz3tren solution (5 mM) was constant
(20 mL) for each titration, whereas the volume of the Cu(II) solution
(5 mM) was varied (0, 10, 15, and 20 mL). NaOH solutions (∼0.1
M) were standardized against dried potassium hydrogen phthalate.
The number of moles of H+, Cu2+, and Bz3tren were refined in the
fit and did not differ significantly from the calculated values based
on the quantity of reagents used.
Spectrophotometric Studies. For the tren system, the pH was
fixed but not buffered. For all of the other systems, MOPS (3-
morpholinopropanesulfonic acid) was used. Solutions were left to
equilibrate for at least 10 min in the water-jacketed cell holder (T
) 25.0 ( 0.1 °C) prior to measurement.
[Cu(tren)(S2O3)]. Na2S2O3‚5H2O and [Cu(tren)(H2O)](ClO4)2
were used to volumetrically prepare stock solutions of thiosulfate
(75 mM) and the Cu(II) complex (0.75 mM), whereas perchlorate
solutions were prepared by neutralization of a standardized NaOH
solution (∼2 M) with perchloric acid. The thiosulfate solution was
kept in the dark, and if it was not used within 2 weeks, a fresh
solution was prepared. Solutions for UV-visible spectrophotometric
study were prepared by mixing 10 mL each of a [Cu(tren)(H2O)]-
[Cu(Me6tren)(S2O3)]. Both copper(II) and thiosulfate stock
solutions were prepared in the same way as for [Cu(Me3tren)(S2O3)].
The final concentrations of the solutions used in spectral analysis
were [Cu]total ) 0.25 mM, [S2O32-] ) 0-25 mM, and I ) 0.15 M
(NaClO4). Due to the reactivity of [Cu(Me6tren)(H2O)]2+ with
thiosulfate, a two-compartment mixing cell was used to allow the
two solutions to equilibrate to 25 ( 0.1 °C in the cell prior to
mixing. The UV-visible spectrum was then recorded from 300 to
500 nm within approximately 1 min of mixing, and the baseline
spectra were subtracted as above prior to analysis.
[Cu(Bz3tren)(S2O3)]. In this case, the Cu(II) complex was
prepared in situ. Bz3tren and MOPS were added to deoxygenated
water, the pH was adjusted to 7.50 using ∼1 M NaOH, and Cu-
(ClO4)2‚6H2O was added to generate a solution with [Cu]total ) 1
mM, [Bz3tren]total ) 1 mM, and MOPS (50 mM). It was not possible
to fix the ionic strength using NaClO4‚H2O because even the
slightest addition caused precipitation of a pale-blue solid, presum-
ably of [Cu(Bz3tren)(H2O)](ClO4)2. Solutions for spectral measure-
ment were prepared by mixing the Cu(II) solution with various
thiosulfate solutions such that the final concentrations were [Cu]tot
) 0.5 mM, [S2O32-] ) 0-2.5 mM, and I ) 0.03-0.04 M. Spectral
measurements were conducted as they were for [Cu(Me3tren)-
(S2O3)].
Acknowledgment. We are grateful to the ARC Center
for Green Chemistry, the Australian Postgraduate Award
Scheme (A.J.F. and J.B.), and the Australian Institute for
Nuclear Science and Engineering for financial support, to
Ms. A. Kutasi for kindly collecting the X-ray data for the
Bz3tren and Me3tren complexes, and to Dr. S. R. Batten for
assistance with modeling the disorder in the crystal structure
of (H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3.
(ClO4)2 solution, a sodium thiosulfate solution, and a sodium
2-
perchlorate solution in a sealable glass vial ([S2O3
]final ) 0.125-
25.0 mM, I ) 0.25 M). The pH of each sample was then fixed by
the addition of a small volume of NaOH from the autoburet such
that the volume change upon the addition of base was less than
2%.
[Cu(Me3tren)(S2O3)]. [Cu(Me3tren)(H2O)](ClO4)2, MOPS, and
NaClO4‚H2O were used to volumetrically prepare a Cu stock
solution in deoxygenated water, which had a composition of [Cu]total
) 0.5 mM, [MOPS]total ) 100 mM and I ) 0.15 M. The pH of the
stock was adjusted to 7.50 at 25 °C by titration with an ap-
proximately 1 M NaOH solution and then stored in a refrigerator.
Thiosulfate stock solutions (50, 20, and 0.5 mM; I ) 0.15 M) were
prepared with deoxygenated water using Na2S2O3‚5H2O and
NaClO4‚H2O and used to prepare a series of solutions with
Supporting Information Available: Crystallographic data for
[Cu(Me3tren)(MeCN)](ClO4)2, [Cu(tren)(S2O3)]‚H2O, [Cu(Bz3tren)-
(S2O3)]‚MeOH, and (H3Me3tren)[Cu(Me3tren)(S2O3)]2(ClO4)3 in
CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC0492800
6578 Inorganic Chemistry, Vol. 43, No. 21, 2004