Classic coordination complexes: The kinetics and stereochemistry of acid hydrolysis of cis- and trans-[Co(NH3)4Cl2]+

Article abstract of DOI:10.1016/j.poly.2005.12.015

The kinetics and stereochemistry of acid hydrolysis of the classic complexes cis- and trans-[Co(NH₃)₄Cl₂]⁺ have been determined. They hydrolyse at virtually the same rate, 2.4 × 10^-3 s^-1 and 2.2 × 10^-3 s^-1 for the cis and trans ions, respectively, (0.02 M HClO₄, 25 °C). The 'immeasurably fast' hydrolysis of the cis isomer has been dismissed as a myth. The trans-[Co(NH₃)₄(OH₂)Cl]²⁺ isomer has been isolated for the first time, and the trans/cis isomerisation equilibration rate measured, again remarkably at virtually the same rate, 2.3 × 10^-3 s^-1. The cis isomer is the more stable at equilibrium, 88%. The steric course of substitution has been determined as 55% cis product for the cis-dichloro reactant and 35% cis for the trans reactant. The kinetic and equilibrium data are compared to that for the well-studied [Co(en)₂Cl₂]⁺ system, and the synthetic chemistry for the simpler tetraammine systems is elaborated.

Full text of DOI:10.1016/j.poly.2005.12.015

Polyhedron 25 (2006) 1955–1966
www.elsevier.com/locate/poly
Classic coordination complexes: The kinetics and stereochemistry
of acid hydrolysis of cis- and trans-[Co(NH3)4Cl2]+
*
W. Gregory Jackson
School of Physical, Environmental and Mathematical Sciences, Chemistry, The University of New South Wales,
Australian Defence Force Academy (UNSW@ADFA), Northcott Drive, Canberra ACT 2600, Australia
Received 25 October 2005; accepted 14 December 2005
Available online 28 February 2006
Abstract
The kinetics and stereochemistry of acid hydrolysis of the classic complexes cis- and trans-[Co(NH3)4Cl2]+ have been determined.
They hydrolyse at virtually the same rate, 2.4 · 10ꢀ3 sꢀ1 and 2.2 · 10ꢀ3 sꢀ1 for the cis and trans ions, respectively, (0.02 M HClO4,
25 ꢀC). The ‘immeasurably fast’ hydrolysis of the cis isomer has been dismissed as a myth. The trans-[Co(NH3)4(OH2)Cl]2+ isomer
has been isolated for the first time, and the trans/cis isomerisation equilibration rate measured, again remarkably at virtually the same
rate, 2.3 · 10ꢀ3 sꢀ1. The cis isomer is the more stable at equilibrium, 88%. The steric course of substitution has been determined as 55%
cis product for the cis-dichloro reactant and 35% cis for the trans reactant. The kinetic and equilibrium data are compared to that for the
well-studied [Co(en)2Cl2]+ system, and the synthetic chemistry for the simpler tetraammine systems is elaborated.
ꢁ 2006 Elsevier Ltd. All rights reserved.
Keywords: Werner complexes; Aquation kinetics; Stereochemistry; Cobalt(III); Inorganic synthesis
1. Introduction
mer has been quantified as 2.2 · 10ꢀ3 sꢀ1, in five indepen-
dent studies (25 ꢀC, I = 1 M) [8–12]. The stereochemistry
of substitution has not been directly determined for either
isomer, since trans-[Co(NH3)4(OH2)Cl]2+ was unknown.
Linck has estimated it for the trans-[Co(NH3)4Cl2]+ sub-
strate (45 10% trans product) by an indirect method
[10,13].
Herein we report the direct determination of the steric
course of hydrolysis of both cis- and trans-[Co(NH3)4Cl2]+.
Also determined or redetermined are the rates of hydroly-
sis, and the [Co(NH3)4(OH2)Cl]2+ isomerisation rates and
equilibrium position.
Alfred Werner reported the synthesis and characterisa-
tion of cis-[Co(NH3)4Cl2]+ in 1907 to complete the cis
and trans isomeric pair [1]. This historic achievement and
others to follow precipitated his award of the Nobel Prize
in chemistry in 1913, for reasons well-documented [2]. Wer-
ner also synthesised salts of a much wider range of the
structurally related cis- and trans-[Co(en)2AX]n+ ions,
described in a massive 272 page paper in 1912 [3], and it
is these complexes that have been the focus of detailed
investigations of the kinetics and stereochemistry of ligand
substitution over many years [4–7]. In contrast, the synthe-
sis and kinetics of substitution for the simpler tetraammine
complexes have been virtually ignored.
2. Experimental
The rate of hydrolysis for cis-[Co(NH3)4Cl2]+ has been
described as ‘very fast’ [1,3,8], while that for the trans iso-
All chemicals were AnalaR or an equivalent grade. Pro-
ton NMR spectra were recorded on a Varian Unity Plus
400 MHz instrument at 20 ꢀC. Solvents used were D2O
(pD ca. 3, DCl) with dioxane as the internal reference (d
3.75 relative to DSS), and DMF-d7 with the central peak
*
Tel.: +61 2 62688078; fax: +61 2 6268 8090.
E-mail address: g.jackson@adfa.edu.au.
0277-5387/$ - see front matter ꢁ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2005.12.015

1956
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
of the higher field CD2H quintet as the internal reference
(d 2.74 ppm, relative to SiMe4). Time dependent visible
(340–800 nm) and UV (220–320 nm) spectra and absorbance-
time data generated from such data were recorded on a
HP8453A diode array Vis/UV spectrophotometer thermo-
stated to 25.0 0.1 ꢀC with use of a Lauda RM6 cir-
culating water bath. Reported Vis/UV spectra for all
compounds have been extrapolated to zero reaction time.
Kinetic data were analysed by weighted non-linear regres-
sion on a Macintosh with the usual packages, or by import-
ing the HP .kd files into Specfitꢂ and carrying out Global
Spectral Analysis (SVD data reduction) and subsequent
fitting to either a single or double exponential integrated
rate expression [14]. Carbon dioxide free Milli-Qꢂ water
was used for all physical measurements.
5 min was filtered, washed with ethanol and ether, and air
dried. Two recrystallisations from a saturated aqueous
solution by adding a third volume of conc. HCl and cool-
ing yielded the pure chloride salt. Anal. Calc. for CoN4-
H14Cl3O: H, 5.61; N, 22.29; Cl, 42.30. Found: H, 5.39;
N, 22.04; Cl, 42.05%. Vis/UV spectrum (0.02 M HClO4):
1
531 nm, e(max) 57.5; 372 nm, e(max) 49.0 Mꢀ1 cmꢀ1. H
NMR spectrum (D2O, DCl): d 4.30 (6H, cis-Cl, OH2); d
3.63 (3H, trans-OH2); d 3.20 ppm (3H, trans-Cl).
cis-[Co(NH3)4(OH2)Cl]S2O6 Æ H2O was obtained by
adding aqueous Na2S2O6 to a solution of the complex chlo-
ride. Crystals which formed on cooling overnight were
washed with ice water, methanol and then ether and air
dried. Anal. Calc. for CoN4H16ClS2O8: H, 4.50; N, 15.62;
Cl, 9.88; S, 17.88. Found: H, 4.24; N, 15.63; Cl, 9.91; S,
17.50%.
2.1. Synthesis and characterisation
2.1.3. trans-[Co(NH3)5(OH2)Cl]Cl2
[Co(NH3)4(OCO2)]NO3 Æ 0.5H2O
and
[Co(NH3)4-
trans-[Co(NH3)5Cl2]Cl2 Æ H2O [18,21] (12.7 g, 50.5
mmol) was partly suspended/dissolved in a mixture of ice
(150 g) and water (150 mL) acidified with a few drops of
neat CF3SO3H or CF3CO2H while well-stirred in an ice
bath. To this was added dropwise over 4–5 min an ice-cold
suspension (yellow) of HgO (red; 10.9 g, 50.3 mmol,
1.0 equiv.) which had been well-stirred and cooled in ice
(25 g)/water (10 mL) containing neat CF3SO3H (16 g,
107 mmol). The green salt gradually dissolved to yield a
violet solution. Some unreacted green complex remained.
The addition of a mixture of ice (25 g) and HClO4 (70%;
25 mL) was added to ensure complete precipitation of the
unreacted green substrate. The mixture was then filtered,
and the residual green solid (<0.2 g) was washed with one
bed volume of acetone (50 mL). The filtrate and washing
was treated with ice (30 g)/conc. H2SO4 (20 mL), and ace-
tone (75 mL) was added to the point of cloudiness. After
further stirring on ice (ca. 3 min), the mixture was filtered
to remove the precipitate of cis-[Co(NH3)4(OH2)Cl]SO4
(ca. 6 g) and washed with acetone (50 mL). The now laven-
der-grey filtrate was carefully diluted with acetone (ca.
125 mL) just to the point of permanent cloudiness. After
5 min stirring on ice and a then further addition of acetone
(100 mL), the crystallisation of trans-[Co(NH3)4(OH2)-
Cl]SO4 was complete. It was removed by filtration, washed
with acetone and ether, and air dried (5.1 g). A second crop
(1.4 g) was obtained from the filtrate by diluting further
with acetone (to 1 L total volume) and allowing to crystal-
lise for 5 min.
(OCO2)]Cl were prepared by two standard methods, the
faster H2O2 oxidation procedure [15,16] being preferred
over the aerial oxidation route [15,17]. [Co(NH3)4-
(OCO2)]2SO4 Æ 3H2O was made by the published procedure
[10,18], as were cis-[Co(NH3)4(OH2)Cl]SO4 and cis-[Co-
(NH3)4(OH2)2]2(SO4)3 Æ 3H2O derived from it [15,18,19].
The carbonato sulfate was converted to the chloride
by metathesis using the stoichiometric amount of BaCl2 Æ
2H2O to remove the sulfate and then crystallised by precip-
itation with methanol. cis-Diaqua sulfate was generated
in situ from the carbonato sulfate with the required amount
of H2SO4, and then treated with pyridine and Na2S2O6 Æ
2H2O to afford cis-[Co(NH3)4(OH2)OH] S2O6 [20]. Much
less pyridine (30 g) is required than that reported (85 g)
for the synthesis on the same scale [20]. Dehydration of
the latter product (130 ꢀC, 3 h) yielded the cis-diol quanti-
tatively [20]; the observed weight loss on dehydration was
precisely that calculated. The resultant cis–cis-[(NH3)4Co-
(OH)2Co(NH3)4](S2O6)2 was converted to the chloride salt
by trituration of a thin aqueous slurry with excess NH4Cl,
analogous to that described for the Cr(III) analogue [20].
The product was recrystallised from a saturated aqueous
by addition of a fifth volume of saturated NH4Cl and care-
ful dilution with methanol. The product was collected, and
washed with methanol and ether, and air dried.
2.1.1. cis-[Co(NH3)4(OH2)Cl]SO4
The crude material was recrystallised by adding suffi-
cient conc. H2SO4 to a thick aqueous suspension of the
crude sulfate salt to dissolve much of it, filtering rapidly,
and then collecting the resultant crystals on cooling. The
process was repeated to afford the pure sulfate. Anal. Calc.
for CoN4H14ClSO5: H, 5.30; N, 19.66; Cl, 12.27; S, 11.22.
Found: H, 5.05; N, 19.99; Cl, 12.63; S, 11.29%.
The crude trans isomer was stirred in aqueous HCl
(5 M) whence it transformed from a lavender grey to a fine
grey-green Clꢀ salt. It was filtered to remove the cis isomer
which had selectively extracted. The extraction was
repeated until the washings were virtually colorless and
the residual solid had become a dull green. The purified
trans-[Co(NH3)4(OH2)Cl]Cl2 was finally recrystallised from
a minimum volume of ice water by filtering and quickly
adding cold 5 M HCl. Shiny green-grey needles resulted,
which were quickly collected, washed with methanol and
2.1.2. cis-[Co(NH3)4(OH2)Cl]Cl2
cis-[Co(NH3)4(OH2)Cl]SO4 (10 g) was treated with conc.
HCl (50 mL) at ambient temperature, and after stirring for

W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1957
ether and air dried. cis Isomer is introduced if crystallisa-
tion is delayed. Anal. Calc. for CoN4H14Cl3O: H, 5.65;
N, 22.42; Cl, 42.55. Found: H, 5.64; N, 22.05; Cl,
41.68%. Vis/UV spectrum (0.02 M HClO4): 593 nm,
e(max) 30.7; 470 nm, e(max) 24.1; 367 nm, e(max) 50.0;
and quickly recrystallised from water/HClO4. Yield: 1.6 g
(46%, based on diol). Anal. Calc. for CoH12N4Cl3O4: H,
4.07; N, 18.84; Cl, 35.76. Found: H, 4.34; N, 18.51; Cl,
35.32%. Vis/UV spectrum (0.02 M HClO4): 594 nm, e(sh)
54.2; 563 nm, e(max) 57.1; 456 nm, e(min) 9.5; 390 nm,
530 nm, e 20.4 Mꢀ1 cmꢀ1
.
1H NMR spectrum (D2O,
e(sh) 61.0 Mꢀ1 cmꢀ1 1H NMR spectrum: DMF-d7: d
.
DCl): d 4.36 ppm (6H, cis-Cl, OH2).
4.25 (6H, cis-Cl); d 2.74 (6H, trans-Cl); D2O, DCl: d 4.10
(6H, cis-Cl); d 3.35 ppm (6H, trans-Cl).
An aqueous solution of the Clꢀ salt yielded crystals of
2ꢀ
2ꢀ
the SO4
or S2O6
using the appropriate precipitant
trans-[Co(NH3)4(NO2)2]Cl was synthesised as described
[23]. It was recrystallised from a saturated aqueous solution
by addition of a fifth volume of saturated NH4Cl and care-
ful dilution with methanol. The perchlorate salt was
obtained by metathesis in water using HClO4 as precipi-
tant. Vis/UV spectrum (H2O): k(max) 441 and 347 nm.
1H NMR spectrum (DMF-d7): d 3.82 (12H, cis-NO2).
cis-[Co(NH3)4(NO2)2]Cl was initially prepared as
reported by aerial oxidation of Co(II) in the presence of
(5 M H2SO4 or 5 M aqueous Li2S2O6).
An alternative synthesis of the trans-aquachloro com-
plex is via reaction of trans-[Co(NH3)4Cl2]Cl Æ H2O in neat
triflic acid, as described for the bis(ethylenediamine) ana-
logue [22]. The product mixture required fractionation of
the sulfate salts, as described above.
2.1.4. trans-[Co(NH3)4Cl2]Cl Æ H2O
ꢀ
This followed that described by Jørgensen, reaction of
cis-[Co(NH3)4(OH2)2]2(SO4)3 Æ 3H2O in conc. H2SO4, fol-
lowed by reaction with 37% HCl, all at ambient tempera-
ture [18,21]. By warming to 50–60 ꢀC, the time scale of
the synthesis can be reduced from 48 h to 30 min, and
reduced further by using [Co(NH3)4(OCO2)]2SO4 Æ 3H2O
directly rather than first converting it to the cis-diaqua
complex using aqueous H2SO4. The grass-green product
is reported to be the chloride salt, but it contains free sul-
fate and requires recrystallisation from H2O/HCl to yield
pure trans-[Co(NH3)4Cl2]Cl Æ H2O. The perchlorate salt
was obtained by metathesis in water using 5 M HClO4 as
precipitant. Anal. Calc. for CoN4H14Cl3O: H, 5.61; N,
22.29; Cl, 42.30. Found: H, 5.38; N, 22.11; Cl, 42.13%.
Vis/UV spectrum (0.02 M HClO4): 634 nm, e(max) 42.1;
538 nm, e(min) 4.7; 467 nm, e(max) 21.5; 377 nm, e(max)
excess NH3 and NO2 [18]; the yield of pure material is
poor. The high yield synthesis of cis-[Co(NH3)4-
(NO2)2]NO3 has been reported [23]. We have made the
Clꢀ salt by warming cis-[Co(NH3)4(OH2)Cl]Cl2 with excess
NaNO2 (5 equiv.) after the addition of HCl to first generate
the nitritochloro complex. The yellow product was re-
precipitated from water with HCl, and then freed of a little
trans isomer by extracting about 90% of the product with
water and crystallising using HCl. This process was
repeated to afford yellow needles of the pure cis salt. The
perchlorate was obtained as described for the trans isomer.
Vis/UV spectrum (H2O): k(max) 441 and 324 nm. 1H
NMR spectrum (DMF-d7): d 3.86 (6H, cis-NO2); d
3.60 ppm (6H, trans-NO2).
An alternative synthesis, and one that gives cis isomer
cleanly, is as follows. [Co(NH3)4(OCO2)]Cl in water was
treated with sufficient H2SO4 to generate the cis-[Co-
(NH3)4(OH2)2]3+ species, as described [20]. Addition of
excess NaNO2 (5 equiv.) was added, and the stirred mix-
ture was allowed to stand 48 h to allow the nitrito species
to isomerise to the nitro complex. Ethanol was added to
precipitate the product, and the filtered material was
recrystallised from hot water by addition of a fifth volume
of saturated NH4Cl and dilution with methanol. Fine yel-
low-orange crystals slowly deposited on cooling. Curiously,
this material was much less soluble than the cis-dinitro
chloride obtained as above, and proved to contain no chlo-
ride but rather anionic sulfate which had carried through
from the H2SO4 plus carbonato chloride initial step.
cis-[Co(NH3)4(NO2)Cl]Cl was obtained by allowing a
stirred, thin slurry of cis-[Co(NH3)4(NO2)2]Cl in conc.
HCl (1 g in 5 mL) to react for 18 h. The crude product
was collected by filtration, washed with ethanol and ether,
and air dried. This was extracted carefully with water to
yield a claret solution and a green residue of the less soluble
trans-[Co(NH3)4Cl2]Cl. Addition of HCl precipitated the
red-claret chloride salt which was recrystallised from water
using HCl as precipitant to afford the Clꢀ salt. Trituration
of the Clꢀ salt in a thin slurry with 70% HClO4 removed
Clꢀ as HCl, and the filtered product was recrystallised from
36.8; 370 nm, e(max) 46.4 Mꢀ1 cmꢀ1 1H NMR spectra:
.
DMF-d7: d 4.20 (12H, cis-Cl); D2O, DCl: d 4.06 ppm
(12H, cis-Cl).
2.2. cis-[Co(NH3)4Cl2]ClO4
The following is a modification of Werner’s procedure
[1]. Concentrated HCl which had been saturated with
HCl (g) at 0 ꢀC was allowed to warm to ca. 15 ꢀC and then
30 mL was added to cis–cis-[(NH3)4Co(OH)2Co(NH3)4]Cl4
(5.0 g) in a mortar and ground rapidly for ca. 40 s. The
crystals that had quickly changed consistency and colour
were filtered and the then chilled filtrate and ethanol wash-
ings (2 · 10 mL) were treated with cold HClO4 (5 M; 1:1
70% mixed with ice). On scratching blue crystals deposited
which were collected, washed with ethanol and ether and
air dried. The original solid was extracted with small por-
tions of ice water to give a red solution of the diaqua com-
plex which contained a little of the cis-dichloro complex;
this was again crystallised using 5 M HClO4. The solid res-
idue was crude blue cis-dichloro chloride which was
extracted with sufficient ice water to dissolve it all, filtered,
and treated with a fifth volume of cold HClO4 (5 M). The
three fractions of cis-[Co(NH3)4Cl2]ClO4 were combined

1958
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1
water/HClO4. H NMR spectrum (DMF-d7): d 4.54 (6H,
cis-NO2, cis-Cl); d 4.04 (3H, cis-Cl, trans-NO2); d
3.48 ppm (3H, trans-Cl).
cleavage reaction! Werner envisaged the reaction as a sim-
ple cis-Co–O cleavage reaction, with such bond rupture
occurring on the same Co centre. However, acid cleavage
of the diol might have proceeded either way, as shown in
Scheme 1.
trans-[Co(NH3)4(NO2)Cl]Cl was synthesised from trans-
[Co(NH3)4(NO2)2]Cl as described for the cis isomer above
but allowing a 27 h reaction time. The trans-dinitro complex
is not very soluble in HCl but cleanly transforms to the
trans-nitrochloro product quantitatively, whereas the cis
isomꢀer partly dissolves in HCl and some loss of the second
However, we can confirm that no observable cis-[Co-
(NH3)4(OH2)Cl]Cl2 is produced in the diol cleavage reac-
tion. Similarly, the acid cleavage reactions of the carbonato
and dinitro complexes, each reported to yield the cis-
[Co(NH3)4Cl2]Cl complex, might have expected to yield
aqua complexes, given the known modes of bond cleavage
revealed through tracer studies [4,26], Scheme 2.
In our hands, no cis-dichloro complex was formed.
We have found that neither cis- nor trans-[Co-
(NH3)4(OH2)Cl]2+ yield cis-[Co(NH3)4Cl2]+ on reaction
with HCl under a variety of conditions, nor does evapo-
ration of an aqueous solution of the trans-dichloro chlo-
ride, a strategy successful in the bis(ethylenediamine)
system.
Finally, the anation of either cis-[Co(NH3)4(OSMe2)2]-
(ClO4)3 using Et4Cl in either DMSO or DMF failed to
yield appreciable cis-dichloro isomer, yet in the corre-
sponding bis(ethylenediamine) systems [27] the cis-dichloro
complex is formed and is strongly stabilised by ion-pairing
in such dipolar aprotic media.
NO2 ligand occurs. The salmon pink trans-nitrochloro
product was obtained as Clꢀ and ClO4 salts. H NMR
1
ꢀ
spectrum (DMF-d7): d 3.70 ppm (12H, cis-NO2, cis-Cl).
2.3. Kinetics of acid hydrolysis
Reactions were followed spectrophotometrically using a
8-cell compartment housed in an HP8453A diode array
Vis/UV spectrophotometer and the in situ method
described previously [24]. The solvent was 1.0, 0.1 or more
usually 0.02 M HClO4.
3. Results and discussion
3.1. Syntheses
cis-[Co(NH3)4(OH2)Cl]SO4 and cis-[Co(NH3)4(OH2)Cl]-
S2O6 Æ H2O were too insoluble to be generally useful, while
the more soluble cis-[Co(NH3)4(OH2)Cl]Cl2 as prepared by
various published routes was not pure (1H NMR). Conver-
sion to the sulfate and back to the chloride removed these
impurities. The chloride and sulfate salts were anhydrous;
hydrates have been reported [13].
trans-[Co(NH3)4Cl2]Cl Æ H2O was obtained by a modifi-
cation of a published method [21]. It is much less soluble in
water than the bis(ethylenediamine) analogue, and this is
also true of methanol as solvent in which it is virtually
insoluble. For the perchlorate salts, the water solubilities
are reversed. The perchlorate counter ion does not precip-
itate the chloroaqua ions, and thus both cis- and trans-
[Co(NH3)4Cl2]ClO4 are obtained free of the chloroaqua
species which otherwise readily crystallise when Clꢀ,
2ꢀ
2ꢀ
SO4 or S2O6 are used as counterions.
Cl
H2O
Cl
+
Werner’s classic cis-[Co(NH3)4Cl2]Cl complex was first
synthesised by HCl cleavage of the cis-diol [(NH3)4Co-
(l-OH)2Co(NH3)4]Cl4 [1]. A 1:1 mixture of cis-[Co(NH3)4-
Cl2]Cl and cis-[Co(NH3)4(OH2)2]Cl3 results. The HCl
needs to be above the standard concentrated grade
(11.7 M) or the yield of the dichloro complex from the
bridge cleavage process is reduced. Treatment of cis-[Co(N-
H3)4(NO2)2]Cl with concentrated HCl has been reported
[18,25] as (rapidly) producing cis-[Co(NH3)4Cl2]Cl, but in
our hands it yielded (slowly) cis-[Co(NH3)4(NO2)Cl]Cl,
and on prolonged standing, trans-[Co(NH3)4Cl2]Cl. Simi-
larly, the trans-dinitro complex yielded trans-[Co(NH3)4-
(NO2)Cl]Cl; these syntheses parallel the bis(ethylenedia-
mine) chemistry [3]. Werner’s alternative syntheses [3] of
cis-[Co(NH3)4Cl2]Cl via reaction of [Co(NH3)4-(OCO2)]Cl
in aqueous or ethanolic HCl yielded only cis-[Co(NH3)4-
(OH2)Cl]Cl2 in our hands. Also, we found that the latter
gives trans-[Co(NH3)4Cl2]Cl on prolonged reaction in
HCl.
Co
Co
OH2
HCl
or
H
O
H
O
HCl
Co
Co
Co
Co
O
H
OH2Cl
HCl
OH2
OH2
Cl
Cl
+
Co
Co
In hindsight, it is surprising that any of these syntheses
for the cis-dichloro complex work, even the successful diol
Scheme 1. Alternative modes of acid catalysed OH– bridge cleavage in the
diol.

W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1959
O
HCl
HCl
17OH2
Cl
17O
17O
17O
C
Co
Co
Co
C
O
Cl
17O
HCl
HCl
17ON17
O
17OH2
N17O2
N17O2
Co
Co
Co
N17O2
N17O2
HCl
HCl
HCl
Cl
17OH2
17OH2
17OH2
Co
Co
Co
17OH2
17ON17
O
Scheme 2. Acid induced bond cleavage and rearrangement modes for carbonate and nitrite complexes of cobalt(III).
The trans-[Co(NH3)4(OH2)Cl]Cl2 complex is new. It
was synthesised by Hg(II) induced aquation of trans-
[Co(NH3)4Cl2]+, a method similar to that described for
the bis(ethylenediamine) analogue [28]. It was also
obtained [22] by reacting trans-[Co(NH3)4Cl2]Cl Æ H2O
with triflic acid to yield [Co(NH3)4(O3SCF3)Cl]CF3SO3,
which rapidly formed a cis-/trans-[Co(NH3)4(OH2)Cl]2+
mixture on hydrolysis, and this could be separated using
the sulfate counter-ion. cis/trans Mixtures of salts of
[Co(NH3)4(OH2)Cl]2+ proved virtually impossible to sep-
(340–700 nm) and UV (220–320, 270–300 nm) regions. In
the visible region, two sharp isosbestic points were
observed (589 nm, e 30.7 Mꢀ1 cmꢀ1; and 473 nm, e
23.8 Mꢀ1 cmꢀ1) and the absorbance changes were large
(Fig. 1). Absorbance changes in the UV were much smaller
(Figure 1S) because cis- and trans-[Co(NH3)4(OH2)Cl]2+
have very similar spectra throughout this region. Nonethe-
less, by working on the tail of the UV where the absor-
bance is low yet proportional absorbance changes are
appreciable, and through raising the [Co] to take advantage
of this, a sharp isosbestic point was observed at 284 nm
(Figure 2S), and the absorbance-time data accurately fitted
a single exponential (kobsd = 2.31 · 10ꢀ3 sꢀ1). The rate data
obtained from the three spectral regions were in good
agreement (Table 1), and kobsd is not appreciably medium
dependent (1.0, 0.1, 0.02 M HClO4).
The trans/cis equilibrium, while it favours the cis side
(88%), is not so far to that side that the rate could not be
measured starting with the cis isomer. The spectral changes
commencing with cis are small (Figure 3S) but sufficient to
define the rate well (Table 1), and even the two isosbestic
points (473 and 590 nm). The measured kobsd should be
the same since the sum of the forward and reverse rates
are measured commencing with either isomer, and indeed
this is born out by the results (Table 1).
arate completely as the Clꢀ, SO4 or S2O
salts, due
2ꢀ
2ꢀ
6
to the relatively rapid rate of trans to cis isomerisation
2ꢀ
and because, for SO4 and S2O62ꢀ, the trans isomer is
more soluble; the cis form then tends to co-crystallise.
As the Clꢀ salt, even small amounts of the cis form ren-
dered the crystals grey, although they appeared homoge-
neous under the microscope. It could, however, be
rendered completely free of it by successive extractions
with small amounts of 1 M HCl until the extracts were
a consistent pale violet colour. The residue was a dull
green salt.
The four nitro complexes, of which cis- and trans-
[Co(NH3)4(NO2)Cl]+ appear to be new, were reactants or
products of the synthetic reactions described, and which
did not proceed as reported. They are analogous to their
bis(ethylenediamine) counterparts, and were characterised
Equilibrium [Co(NH3)4(OH2)Cl]2+ data are given in
Table 2 and results are internally consistent for the four
different starting materials. This result, 88 1% cis, is
not too different to that estimated by Linck’s indirect
method [10,13] (90.5 1.0% cis).
1
by their Vis/UV and H NMR spectra; the characteristic
difference between cis and trans-nitro species in the UV is
noted.
3.2. Isomerisation of [Co(NH3)4(OH2)Cl]2+
The 1H NMR spectrum ( Fig. 2) show the single line spec-
trum for the trans isomer in D2O/DCl change into the 2:1:1
pattern expected for cis-[Co(NH3)4(OH2)Cl]2+, and confirm
its presence at equilibrium to the extent of ca. 10%.
The isomerisation of trans-[Co(NH3)4(OH2)Cl]2+ in
dilute HClO4 at 25 ꢀC was followed in both the visible

1960
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
Fig. 1. Visible spectral changes for the isomerization of trans-[Co(NH3)4-
(OH2)Cl]2+ in H2O/H+, 25 ꢀC.
Table 1
Rate data for the reversible isomerisation of trans-[Co(NH3)4(OH2)Cl]2+
in dilute HClO4 at 25 ꢀC
[HClO4] (M)
0.020
Regionc
103k2 (sꢀ1 a
)
vis
vis
UV
vis
vis
2.31 (10)
2.32 (3)b
2.17 (6)
2.36 (3)
2.62 (9)
0.10
1.0
a
k2 = ktc + kct; ktc/kct = 88/12 (Table 2). The number of determinations
is given in parenthesis.
b
Reactant was cis-[Co(NH3)4(OH2)Cl]Cl2.
Global analysis fits (refer text).
c
Table 2
Fig. 2. 1H NMR spectra (NH region) for the isomerization of trans-
(T2+) to cis- (C2+) [Co(NH3)4(OH2)Cl]2+ in D2O/D+ at ca. 21 ꢀC; times
increase from 1 to 4.
Equilibrium data for the reversible isomerisation of trans-[Co(NH3)4-
(OH2)Cl]2+ in 0.020 M HClO4 at 25 ꢀC
Reactant
e (530 nm) (Mꢀ1 cmꢀ1
)
%cis(equil)a
cis-[(OH2)Cl]SO4
cis-[(OH2)Cl]S2O6 Æ H2O
cis-[(OH2)Cl]Cl2
trans-[(OH2)Cl]Cl2
trans-[Cl2]Cl Æ H2O
cis-[Cl2]ClO4
53.0 (2)
52.8 (3)
53.7 (2)
53.5 (2)
53.0 (2)
52.6 (2)
88
87
90
89
88
87
3.3. Aquation of cis-[Co(NH3)4Cl2]+
The hydrolysis rate for the cis isomer has never been
measured. It has been described as ‘very fast’ by Pearson
et al. [8], and that the aquation was very rapid was implied
by Werner [1,3] in his experimental descriptions of poor
recovery through rapid recrystallisation from water. It is
a myth. It hydrolyses at about the same rate as the trans
Ave:
53.0 (13)
88 (13)
a
Number of determination in parenthesis; results calculated from e(530)
for the cis isomer, 57.5, and e(530) for the trans form, 20.4 Mꢀ1 cmꢀ1
.

W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1961
Fig. 3. Visible spectral changes for the hydrolysis of cis-[Co(NH3)4Cl2]+ in H2O/H+, 25 ꢀC. The absorbance decreases with time at 600 nm.
isomer. The changes in visible spectra (340–700 nm; Fig. 3)
with time hide the fact that there are two early sharp isos-
bestic points, one of which changes quickly with time
(Fig. 4S. 440 nm, e 13.0; 514 nm, e 39.0 Mꢀ1 cmꢀ1). The
first phase is loss of one Clꢀ and the second is cis-/trans-
[Co(NH3)4(OH2)Cl]2+ isomerisation. The final spectrum
corresponds to a 88% cis/22% trans distribution (Table
2). The absorbance–time trace at 530 nm (Fig. 4) clearly
indicates the biphasic behaviour, and in particular it shows
that the second phase corresponds to trans–cis rather than
cis–trans isomerisation, i.e., the initial hydrolysis reaction
leads to not only both cis- and trans-[Co(NH3)4(OH2)Cl]2+
but an isomeric mixture which contains more trans isomer
than does the equilibrium distribution.
The changes in the UV absorption spectra accompany-
ing the hydrolysis of the cis-dichloro complex are relatively
small (Figure 5S), but nonetheless sufficient to determine a
rate constant for the hydrolysis. The data fitted a single
exponential because in the region examined (280–320 nm)
the absorbance changes for the secondary isomerisation
process are much smaller, and negligible. The results (see
below) bear this out – the visible and UV rate data agree.
The primary hydrolysis rate constant was determined by
fitting the 589 nm data to a single exponential. The fits were
good, since this wavelength is an isosbestic point for the
secondary trans/cis isomerisation reaction. The 530 nm
data (Fig. 4) were then fitted to two exponentials with
the first exponential (k1) fixed at the value determined from
the 589 nm data. This yielded k2 which is the aquachloro
Fig. 4. Absorbance–time trace at 530 nm for the hydrolysis of cis-
[Co(NH3)4Cl2]2+ in H2O/H+, 25 ꢀC, illustrating the biphasic nature of the
reaction.
isomerisation rate, and the check on the correctness of
the analysis is the reasonable agreement of this value and
that determined independently for this reaction (Table 3;
cf. Table 1).
It is evident that k1 and k2 are very similar in magnitude.
Because of this, the global analysis of the visible absorption
spectral data (340–700 nm) involves a shallow minimum
with considerable flexibility in the derived values for both
k1 and k2 and the spectrum for the intermediate (kinetic)

1962
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
Table 3
Rate data for the acid hydrolysis of cis-[Co(NH3)4Cl2]+ in 0.020 M HClO4
at 25 ꢀC
Region
103k1 (sꢀ1
)
103k1 (sꢀ1) (589 nm)a
103k2 (sꢀ1
)
(530 nm)b
UV (260–300 nm)
Vis
2.12 (3)
2.35 (9)b
2.39c
2.47 (9)
2.46c
2.33c
2.48c
2.37d
2.31d
a
Single exponential fit at one of the cis-/trans-[Co(NH3)4(OH2)Cl]2+
isomerisation isosbestic points.
b
Double exponential fit, with k1 fixed at the value determined at 589 nm
for the same run.
c
Global fit (340–700 nm) double exponential fit with no restriction on k1
or k2, for selected data sets (>95% reaction).
d
Double exponential fit with k2 fixed at the independently determined
value.
[Co(NH3)4(OH2)Cl]2+ isomer distribution. This is a general
problem when k1 ꢁ k2 [14]. However, fitting to a single
wavelength data set such as the fall-rise seen at 530 nm
gives much better defined parameters when the two rates
are similar, provided the data cover a time period in excess
of 95% reaction.
The steric course for the primary substitution reaction
was determined as described previously [29] by fitting the
e, t data to Eq. (1) which applies to Scheme 3 as shown:
eobsd ¼ e1 þ fðeA ꢀ eCÞ þ ðk1ðeB ꢀ eCÞ=ðk2 ꢀ k1ÞÞg
ꢂ expðꢀk1tÞ ꢀ fðk1ðeB ꢀ eCÞ=ðk2 ꢀ k1ÞÞg
Fig. 5. Computed intermediate (B) and final (C) cis-/trans-[Co(NH3)4-
(OH2)Cl]2+ visible spectra, for the hydrolysis of cis-[Co(NH3)4Cl2]+ (A) in
H2O/H+ at 25 ꢀC. The vertical scale is absorbance, for an arbitrary [Co].
ꢂ expðꢀk2tÞ.
ð1Þ
Here ‘B’ is the first formed cis-/trans-aquachloro mixture
and ‘C’ is the equilibrium mixture. The initial cis/trans ra-
tio is best defined from absorbance data at around 530 nm
where the component spectra differ the most. We also car-
ried out a multiwavelength version of this analysis by glob-
ally fitting (Specfitꢂ) the 340–700 nm data to a A ! B ! C
scheme fixing the values of k1 (determined at 589 nm, Table
3) and k2 (independently determined, Table 1), and deter-
mining the full spectrum of the intermediate isomer mix-
ture (Fig. 5). Note that even though we are dealing with
first order reactions, absolute times must be used to get
the correct spectrum of the intermediate B. The intermedi-
ate spectrum was fitted to the spectra of its two compo-
nents and yielded
distribution; similarly, the final spectrum corresponded to
88% cis, 12% trans.
a 55% cis, 45% trans kinetic
The 440 nm isosbestic point (e 13.0; e(cis) = 11.0,
e(trans) = 19.3 Mꢀ1 cmꢀ1) is not very sensitive to the
trans-/cis- product ratio but the 514 nm isosbestic is. The
initial product distribution calculated from this (e 39.0;
e(cis) = 53.5 and e(trans) = 19.7 Mꢀ1 cmꢀ1), while intrinsi-
cally less accurate, gives excellent agreement (57% cis). This
shifts quite quickly with time, and it was important to define
it using a number of scans collected within the first 40 s (10%
reaction). After 2 min the isosbestic, still reasonably well
defined, is at 516 nm (e 40.5 Mꢀ1 cmꢀ1) and this corre-
sponds to nearly 60% cis product. As the simulation shows
(Table 4), for an initial 55% and final 88% cis, the 55%
changes quickly in a system where k1 and k2 are very similar.
k1'
cis-[Co(NH3)4Cl2]+
cis-[Co(NH3)4(OH2)Cl]2+
kct
1
Fig. 6 shows the H NMR spectra for the cis-dichloro
complex hydrolysing in D2O (pD 2, DCl) at ca. 21 ꢀC.
The cis- and trans-aquachloro ions grow in parallel, before
ultimately the trans isomer decays back to the cis. The low-
est field peak is trans (12H) and the next lowest is the major
cis peak (6H), and from their areas the product distribution
at 2–3 min reaction time is estimated as 60% cis, 40% trans,
which is entirely in accord with the spectrophotometric
determination.
k1''
ktc
trans-[Co(NH3)4(OH2)Cl]2+
k1 = k1' + k1''
k2 = ktc+ kct
Scheme 3. The sequential hydrolysis reaction for cis-[Co(NH3)4Cl2]+.

W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1963
Table 4
Changing product distribution with time for a consecutive reaction scheme
for the cis-[Co(NH3)4Cl2]+ reactant and where k1 = 2.4 · 10ꢀ3 sꢀ1
,
k2 = 2.3 · 10ꢀ3 sꢀ1, and the initial and final cis-[Co(NH3)4(OH2)Cl]2+
product proportions are 55% and 88%
t
A
B
C
%cis
0
5
1.000
0.988
0.976
0.965
0.953
0.942
0.931
0.919
0.908
0.898
0.887
0.876
0.866
0.856
0.845
0.835
0.825
0.815
0.806
0.796
0.787
0.777
0.768
0.759
0.750
0.741
0.732
0.723
0.715
0.706
0.698
0.689
0.681
0.673
0.665
0.657
0.649
0.000
0.012
0.023
0.035
0.046
0.057
0.067
0.077
0.087
0.097
0.107
0.116
0.125
0.134
0.143
0.151
0.159
0.167
0.175
0.182
0.190
0.197
0.204
0.211
0.217
0.224
0.230
0.236
0.242
0.248
0.253
0.258
0.264
0.269
0.274
0.278
0.283
0.000
0.000
0.000
0.001
0.001
0.002
0.002
0.003
0.004
0.005
0.006
0.008
0.009
0.011
0.012
0.014
0.016
0.017
0.019
0.021
0.024
0.026
0.028
0.031
0.033
0.036
0.038
0.041
0.044
0.046
0.049
0.052
0.055
0.058
0.061
0.065
0.068
55.0
55.2
55.4
55.6
55.8
55.9
56.1
56.3
56.5
56.7
56.9
57.0
57.2
57.4
57.6
57.8
57.9
58.1
58.3
58.5
58.7
58.8
59.0
59.2
59.4
59.5
59.7
59.9
60.0
60.2
60.4
60.5
60.7
60.9
61.0
61.2
61.4
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
155
160
165
170
175
180
3.4. Aquation of trans-[Co(NH3)4Cl2]+
The UV absorption maximum at ca. 250 nm for the
trans-dichloro ion [10] is associated with Co–Cl charge
transfer absorption (e ca. 15,000 Mꢀ1 cmꢀ1). It is roughly
twice as intense and to lower energy than the correspond-
ing absorption for the trans- and cis-monochloro ions
(ca. 230 nm), and that, being associated with the Co–Cl
charge transfer, is insensitive to the cis or trans stereochem-
istry. Thus the hydrolysis of the dichloro to monochloro
ions involves a large absorption change in the UV region
(Figure 6S) whereas the trans-/cis-isomerisation involves
very little change. The latter change was sufficient, how-
ever, to easily determine the isomerisation rate spectropho-
tometrically (Table 1).
Fig. 6. 1H NMR spectra (NH region) for the hydrolysis of cis-
[Co(NH3)4Cl2]+ (C+) in D2O/D+ at ca. 21 ꢀC; C2+ and T2+ denote
the signals for cis- and trans-[Co(NH3)4(OH2)Cl]2+. Times increase, from
1 to 4.
trans-[Co(NH3)4(OH2)Cl]2+ are indeed very similar to each
other and to the trans-[Co(NH3)4Cl2]+ ion in this region,
and it confirms Linck’s surmise [10] on this point, i.e.,
the persistence of the sharp isosbestic point at 239 nm for
10 t1/2 of the primary reaction, despite the second step
having essentially the same rate as the first step, the
Although there is no isosbestic point for the isomerisa-
tion coincident with the isosbestic at 239 nm for the hydro-
lysis of trans-[Co(NH3)4Cl2]+ (the so-called triple isosbestic
point), we have determined that the spectra for cis- and

1964
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
primary aquation and secondary isomerisation reactions.
Indeed, the small absorbance changes in the UV for the
secondary reaction are the reason for the good single phase
kinetics for the primary reaction as measured in the UV
region at any wavelength. The rates obtained are all the
same, and in particular the same as those determined from
the UV data at 284 nm (Table 5), which is an isosbestic
point for the trans-/cis-[Co(NH3)4(OH2)Cl]2+ isomerisa-
tion reaction (Figure 2S).
The reaction studied in the visible spectral region
revealed four sharp isosbestic points: 353 nm, e 43.4;
397 nm, e 33.2; 467 nm, e 21.5; 597 nm, e 28.2 Mꢀ1 cmꢀ1
(Fig. 7). None were definitive of the steric course of substi-
tution because the spectra of cis- and trans-[Co-
(NH3)4(OH2)Cl]2+ are very similar at each of these wave-
lengths (see Figs. 1 and 8).
The data in the 340–700 nm visible range actually fit a
single exponential quite well, but this is fortuitous, as dis-
cussed elsewhere [30]. The data were refitted to a double
exponential, appropriate to a consecutive reaction scheme,
with k2 fixed at its known value 2.31 · 10ꢀ3 sꢀ1. The results
for k1 so obtained are presented in Table 5, and the value is
in agreement with that determined in a more straightfor-
ward fashion from the UV data.
Quite remarkably, the hydrolysis rate for the trans iso-
mer (k1, 2.2 · 10ꢀ3 sꢀ1) is almost identical to that deter-
mined for the cis isomer (2.4 · 10ꢀ3 sꢀ1), and also is
essentially that for the subsequent isomerisation reaction
(2.3 · 10ꢀ3 sꢀ1).
Table 5
Rate data for the acid hydrolysis of trans-[Co(NH3)4Cl2]+ in 0.020 M
The steric course of hydrolysis was determined as
described above for the cis isomer – globally fitting the vis-
ible spectral data (340–700 nm) to a consecutive reaction
scheme with the derived values for k1 and k2. The results
are shown in Fig. 8. The average 530 nm result is
eB = 32.7 Mꢀ1 cmꢀ1 (22 determinations), which corre-
sponds to 65% trans and 35% cis product. The error here
is likely 5%. Linck had estimated 45 10% trans product
by an indirect method [10,13].
HClO4 at 25 ꢀC
Region
103k1 (sꢀ1
)
UV
Vis
UV
a
2.22 (12)a
2.22 (22)b
2.23 (4)c
Global analysis (UV: 220–320 nm) single exponential fit.
b
Global analysis (Vis: 340–700 nm) double exponential fit, with k2 fixed
at 2.31 · 10ꢀ3 sꢀ1
.
c
Single exponential fit, for data collected at 284 nm, an isosbestic point
for [Co(NH3)4(OH2)Cl]2+ isomerisation.
Fig. 8. Computed intermediate (B) and final (C) cis-/trans-[Co(NH3)4-
(OH2)Cl]2+ visible spectra, for the hydrolysis of trans-[Co(NH3)4Cl2]+ (A)
in H2O/H+ at 25 ꢀC. The vertical scale is absorbance, for an arbitrary [Co].
Fig. 7. Visible spectral changes for the hydrolysis of trans-[Co(NH3)4Cl2]+
in H2O/H+, 25 ꢀC. The absorbance increases with time at 530 nm.

W.G. Jackson / Polyhedron 25 (2006) 1955–1966
1965
Fig. 9. 1H NMR spectra (NH region) for the hydrolysis of trans-(T+) [Co(NH3)4Cl2]+ in D2O/D+ at ca. 21 ꢀC; times increase from 1 to 6. C2+ and T2+
are cis- and trans-[Co(NH3)4(OH2)Cl]2+, respectively.
The 1H NMR spectra for trans-[Co(NH3)4Cl2]+ reacting
in D2O/DCl are shown in Fig. 9. Compared to the corre-
sponding changes for the cis isomer (Fig. 6), there is a little
more trans product formed initially, estimated as 60%
which is in good agreement with that determined
spectrophotometrically.
ler than the corresponding tetraammine rates; for the latter
the cis isomer is about 10-fold more reactive than the trans
isomer. The previous notion [8] that cis-[Co(NH3)4Cl2]+
hydrolyses extraordinarily rapidly is unfounded. Both iso-
mers yield cis- and trans-[Co(NH3)4(OH2)Cl]2+ products
directly, in different proportions, with the trans isomer
showing more rearrangement (35% cis product; 65% rear-
rangement) than the cis form (55% cis product, 45% rear-
rangement). This also parallels the bis(ethylenediamine)
chemistry, but there the extents of rearrangement are less
(the trans reactant gives 74% trans product, the cis reactant
gives 76% cis product; these numbers correspond to 26%
and 24% rearrangement, respectively) [7]. The aquachloro
4. Summary and conclusions
The cis- and trans-dichloro complexes hydrolyse at coin-
cidentally the same rate. For the bis(ethylenediamine) ana-
logues [7], the cis isomer is about sixfold more reactive than
the trans, and both rates are between 10- and 50-fold smal-

1966
W.G. Jackson / Polyhedron 25 (2006) 1955–1966
[6] W.G. Jackson, A.M. Sargeson, in: P. deMayo (Ed.), Rearrangement
in Coordination Complexes, vol. 2, Academic Press, New York, 1980,
p. 273.
[7] W.G. Jackson, in: I. Bernal (Ed.), The Stereochemistry of the Bailar
Inversion and Related Substitution Processes, vol. 1, Elsevier, 1986, p.
255.
[8] R.G. Pearson, R.C. Boston, F.J. Basolo, J. Phys. Chem. 59 (1955)
304.
[9] R. Tsuchida, Bull. Chem. Soc. Jpn. 11 (1936) 721.
[10] R.G. Linck, Inorg. Chem. 8 (1969) 1016.
equilibrium isomer distributions in the two systems are also
similar (en, 79% cis; NH3, 88% cis).
The new stereochemical data for the tetraammine sys-
tem again show that cis isomers are not universally reten-
tive. It was once believed [4] the norm for the
bis(ethylenediamine) complexes, but shown otherwise
[6,29]. It thus seems unlikely that a study of the kinetics
and stereochemistry of other [Co(NH3)4AX]n+ complexes
will uncover anything beyond what is described herein,
notwithstanding the fact that the synthetic chemistry is
not yet developed.
[11] L.L. Borer, H.W. Erdman, C. Norris, J. Williams, J. Worrell,
Inorganic Syntheses, vol. 31, Wiley, NY, 1997.
[12] G. Daffner, D.A. Palmer, H. Kelm, Inorg. Chim. Acta 61 (1982) 57.
[13] R.G. Linck, Inorg. Chem. 7 (1968) 2394.
[14] W.G. Jackson, A.F.M.M. Rahman, M.A. Wong, Inorg. Chim. Acta
357 (2004) 665.
Acknowledgement
[15] G.B. Kauffman, R.P. Pinnell, J.K. Crandall, W.L. Jolly, Inorganic
Syntheses, vol. 6, McGraw-Hill, NY, 1960, p. 177.
[16] R.J. Angelici, 2nd ed.Synthesis and Technique in Inorganic Chem-
istry, University Science Books, CA, 1986, p. 16.
[17] G. Schlessinger, J.W. Simmons, G. Jabs, M.M. Chamberlain,
Inorganic Syntheses, vol. 6, McGraw-Hill, 1960, p. 173.
[18] G. Brauer, in: Handbook of Preparative Inorganic Chemistry, vol. 2,
Academic, NY, 1965, p. 1537.
The author is grateful to Professor Hans Riesen for an
accurate translation of sections of Werner’s original Beri-
chte paper.
Appendix A. Supplementary data
[19] S.M. Jørgensen, Z. Anorg. Chem. 2 (1892) 296.
[20] J. Springborg, C.E. Schaffer, M.L. Wilson, J.M. Wharton, W.E.
Hatfield, Inorganic Syntheses, vol. 18, Wiley, NY, 1978, p. 75.
[21] S.M. Jørgensen, Z. Anorg. Chem. 14 (1897) 404.
[22] P. Comba, N.J. Curtis, W.G. Jackson, A.M. Sargeson, Aust. J.
Chem. 39 (1986) 1297.
Figures 1S–6S, visible/UV spectral data. Supplementary
data associated with this article can be found, in the online
version, at doi:10.1016/j.poly.2005.12.015.
[23] G.B. Kauffman, S.F. Abbott, S.E. Clark, J.M. Gibson, R.D. Meyers,
Inorganic Syntheses, vol. 18, Wiley, NY, 1978, p. 69.
[24] W.G. Jackson, Inorg. Chem. 41 (2002) 7077.
References
[1] A. Werner, Ber. Deut. Chem. Ges. 40 (1907) 4817.
[2] Werner Centennial, vol. 62, ACS, Washington, 1967.
[3] A. Werner, Ann 386 (1912) 1.
[4] F. Basolo, E.G. Pearson, Mechanisms of Inorganic Reactions, 2nd
ed., Wiley, NY, 1967.
[25] C. Duval, Comptes Rendus Hebd. Seances Acad. Sci. 182 (1926) 636.
[26] W.G. Jackson, Inorg. Chem. 26 (1987) 3857.
[27] W.R. Fitzgerald, A.J. Parker, D.W. Watts, J. Am. Chem. Soc. 90
(1968) 5744.
[28] W.G. Jackson, C.M. Begbie, Inorg. Synth. 21 (1982) 119.
[29] W.G. Jackson, A.M. Sargeson, Inorg. Chem. 17 (1978) 1348.
[30] W.G. Jackson, Int. J. Chem. Kinet. (submitted for publication).
[5] M.L. Tobe, in: Advances in Inorganic and Bioinorganic Mechanisms,
vol. 2, Academic Press, New York, 1983, p. 1.