Mechanism of CO2 and H+ reduction by Ni(cyclam)+ in aqueous solution. A pulse and continuous radiolysis study

Article abstract of DOI:10.1021/ic980902p

The reduction of CO₂ and H⁺ by Ni(cyclam)⁺ in aqueous solution was found to proceed by the reduction of Ni(cyclam)(CO₂)⁺ or Ni(cyclam)(H)²⁺ by Ni(cyclam)⁺. At ambient temperature the rate constants for the bimolecular reactions are 1.6 × 10⁸ and 7.2 × 10⁷ M^-1 s^-1, respectively. Continuous γ-radiolysis studies have demonstrated nearly quantitative formation of CO from two Ni(cyclam)⁺ at pH 5.2 under 1 atm of CO₂ and in the presence of formate. Under the same conditions, but at low CO₂ concentration, the reduction of protons was found to compete with CO₂ reduction.

Full text of DOI:10.1021/ic980902p

Inorg. Chem. 1999, 38, 1579-1584
1579
Mechanism of CO2 and H+ Reduction by Ni(cyclam)+ in Aqueous Solution. A Pulse and
Continuous Radiolysis Study
Craig A. Kelly,*,†,1 Elliott L. Blinn,‡,1 Nadia Camaioni,2 Mila D’Angelantonio,2 and
Quinto G. Mulazzani*,2
Istituto di Fotochimica e Radiazioni d’Alta Energia del CNR, Via P. Gobetti 101, 40129 Bologna, Italy,
and Center for Photochemical Sciences, Department of Chemistry, Bowling Green State University,
Bowling Green, Ohio 43403
ReceiVed July 28, 1998
The reduction of CO2 and H+ by Ni(cyclam)+ in aqueous solution was found to proceed by the reduction of
Ni(cyclam)(CO2)+ or Ni(cyclam)(H)2+ by Ni(cyclam)+. At ambient temperature the rate constants for the
bimolecular reactions are 1.6 × 108 and 7.2 × 107 M-1 s-1, respectively. Continuous γ-radiolysis studies have
demonstrated nearly quantitative formation of CO from two Ni(cyclam)+ at pH 5.2 under 1 atm of CO2 and in
the presence of formate. Under the same conditions, but at low CO2 concentration, the reduction of protons was
found to compete with CO2 reduction.
Introduction
CO2 to the +1 oxidation-state metal centers, little knowledge
exists on the mechanism of H2 and CO evolution from these
Since the work of Brown et al.3 and that of Fisher and
Eisenberg4 almost 20 years ago, a remarkable effort has been
put forward to study the photo- and electrochemical reduction
of protons and/or CO2 by cobalt(II) and nickel(II) macrocycle
complexes. As a result, considerable knowledge now exists on
the chemistry and physical properties of many Co+ and Ni+
macrocycle complexes in the solid state,5 in solution,5-9 and at
an electrode.10-18 However, despite the accumulation of kinetic
and thermodynamic information on the binding of protons and
intermediate species. A notable exception is the work of Fujita
et al. who studied the slow reduction of CO2 by CoL+, where
L is 5,7,7,12,14,14-hexamethyl-1,4,8,11-tetraazacyclotetradeca-
4,11-diene, in acetonitrile and found it to proceed by a second-
order process involving two CoL+ centers.11a
Until recently,18 the most efficient and selective catalyst for
the reduction of CO2 to CO in acidic (pH 4) aqueous solution
was the complex Ni(cyclam)2+, where cyclam ) 1,4,8,11-
tetraazacyclotetradecane.12a,b Hydrogen evolution in this system
was observed below pH 4.12b While equally high current
efficiencies were observed earlier,4 the selectivity displayed
toward CO2 in the presence of protons is a remarkable aspect
of Ni(cyclam)+ chemistry. Introducing Ni(cyclam)+ into the
photochemical system of Brown et al.3 has been demonstrated
to reduce CO2 at pH 5 although with low yields for CO
formation.8b
† Present address: Department of Chemistry, Johns Hopkins University,
3400 North Charles St., Baltimore, MD 21218.
‡Deceased, May 7, 1997.
(1) Bowling Green State University.
(2) Istituto FRAE-CNR.
(3) Brown, G. M.; Brunschwig, B. S.; Creutz, C.; Endicott, J. F.; Sutin,
N. J. Am. Chem. Soc. 1979, 101, 1298-1300.
(4) Fisher, B.; Eisenberg, R. J. Am. Chem. Soc. 1980, 102, 7361-7363.
(5) (a) Szalda, D. J.; Fujita, S.; Creutz, C. Inorg. Chem. 1989, 28, 1446-
1450. (b) Fujita, E.; Creutz, C.; Sutin, N.; Szalda, D. J. J. Am. Chem.
Soc. 1991, 113, 343-353. (c) Szalda, D. J.; Fujita, E.; Sanzenbacher,
R.; Paulus, H.; Elias, H. Inorg. Chem. 1994, 33, 5855-5863.
(6) (a) Tait, A. M.; Hoffman, M. Z.; Hayon, E. J. Am. Chem. Soc. 1976,
98, 86-93. (b) Tait, A. M.; Hoffman, M. Z.; Hayon, E. Inorg. Chem.
1976, 15, 934-939.
(7) (a) Jubran, N.; Ginzburg, G.; Cohen, H.; Meyerstein, D. J. Chem. Soc.,
Chem. Commun. 1982, 517-519. (b) Jubran, N.; Cohen, H.; Meyer-
stein, D. Isr. J. Chem. 1985, 25, 118-121. (c) Jubran, N.; Ginzburg,
G.; Cohen, H.; Koresh, Y.; Meyerstein, D. Inorg. Chem. 1985, 24,
251-258. (d) Jubran, N.; Meyerstein, D.; Cohen, H. Inorg. Chim. Acta
1986, 117, 129-132.
(8) (a) Grant, J. L.; Goswami, K.; Spreer, L. O.; Otvos, J. W.; Calvin, M.
J. Chem. Soc., Dalton Trans. 1987, 2105-2109. (b) Craig, C. A.;
Spreer, L. O.; Otvos, J. W.; Calvin, M. J. Phys. Chem. 1990, 94, 7957-
7960.
(9) (a) Creutz, C.; Schwarz, H. A.; Wishart, J. F.; Fujita, E.; Sutin, N. J.
Am. Chem. Soc. 1989, 111, 1153-1154. (b) Creutz, C.; Schwarz, H.
A.; Wishart, J. F.; Fujita, E.; Sutin, N. J. Am. Chem. Soc. 1991, 113,
3361-3371.
(10) Schmidt, M. H.; Miskelly, G. M.; Lewis, N. S. J. Am. Chem. Soc.
1990, 112, 3420-3426.
(11) (a) Fujita, E.; Szalda, D. J.; Creutz, C.; Sutin, N. J. Am. Chem. Soc.
1988, 110, 4870-4871. (b) Fujita, E.; Creutz, C.; Sutin, N.; Brun-
schwig, B. S. Inorg. Chem. 1993, 32, 2657-2662. (c) Fujita, E.;
Creutz, C. Inorg. Chem. 1994, 33, 1729-1730.
Electrochemical investigations of the reduction of CO2 by
Ni(cyclam)+ have demonstrated that the active catalyst is
absorbed on the surface of the electrode and have suggested
(12) (a) Beley, M.; Collin, J.-P.; Ruppert, R.; Sauvage, J.-P. J. Chem. Soc.,
Chem. Commun. 1984, 1315-1316. (b) Beley, M.; Collin, J.-P.;
Ruppert, R.; Sauvage, J.-P. J. Am. Chem. Soc. 1986, 108, 7461-7467.
(c) Collin, J.-P.; Jouaiti, A.; Sauvage, J.-P. Inorg. Chem. 1988, 27,
1986-1990.
(13) Tinnemans, A. H. A.; Koster, T. P. M.; Thewissen, D. H. M. W.;
Mackor, A. Recl. TraV. Chim. Pays-Bas 1984, 103, 288-295.
(14) Pearce, D. J.; Pletcher, D. J. Electroanal. Chem. 1986, 197, 317-
330.
(15) (a) Fujihira, M.; Hirata, Y.; Suga, K. J. Electroanal. Chem. 1990, 292,
199-215. (b) Hirata, Y.; Suga, K.; Fujihira, M. Chem. Lett. 1990,
1155-1158. (c) Fujihira, M.; Nakamura, Y.; Hirata, Y.; Akiba, U.;
Suga, K. Denki Kagaku 1991, 59, 532-539.
(16) (a) Balazs, G. B.; Anson, F. C. J. Electroanal. Chem. 1992, 322, 325-
345. (b) Balazs, G. B.; Anson, F. C. J. Electroanal. Chem. 1993, 361,
149-157.
(17) (a) Smith, C. I.; Crayston, J. A.; Hay, R. W. J. Chem. Soc., Dalton
Trans. 1993, 3267-3269. (b) Hay, R. W.; Kinsman, B.; Smith, C. I.
Polyhedron 1995, 14, 1245-1249.
(18) Fujita, E.; Haff, J.; Sanzenbacher, R.; Elias, H. Inorg. Chem. 1994,
33, 4627-4628.
10.1021/ic980902p CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/16/1999

1580 Inorganic Chemistry, Vol. 38, No. 7, 1999
Kelly et al.
Table 1. Product G Values Following γ Irradiationa
vacuum-degassed soln CO2-satd soln
that homogeneous Ni(cyclam)+ is inactive as a catalyst.16a The
low CO yields observed during the photochemical reduction of
CO2 by Ni(cyclam)+,8b coupled with the importance of the
absorbed complex in electrochemical studies,16a raise questions
concerning the ability of Ni(cyclam)+ to function as a homo-
geneous catalyst. This is an unfortunate conclusion since the
G(H2)
G(CO2)
G(CO)
G(H2)
G(CO)
obsdb
calcdc
0.24
0.10
0.26
0.34
0.10
0.00
0.10
0.10
0.24
0.00
a Conditions: 0.5 mM Ni(cyclam)(ClO4)2, 0.1 M HCO2-, pH 5.2,
strongly reducing nature of Ni(cyclam)+, E°(Ni(cyclam)2+/+
)
- 1.33 V,12b and the low pKa of Ni(cyclam)(H)2+, pKa ) 1.8,19
suggest that Ni(cyclam)+ is an ideal aqueous reducing agent
for CO2.
22 ( 2 °C and doses ) 0.5-1.0 kGy. G values are given in units of
µmol J-1 b Values measured in this work. c Calculation based on the
.
values given in the text, assuming no reaction between Ni(cyclam)+
and protons or CO2.
In previous work, we have characterized the rapid binding
of CO2 and protons by Ni(cyclam)+ using both pulse radiolysis
and laser flash photolysis.19 From that work it is now apparent
that the selectivity displayed by Ni(cyclam)+ toward CO2 over
protons at pH 4 is not the result of an intrinsic selectivity. In
fact, the reactivity of Ni(cyclam)+ toward CO2 and protons, and
the stabilities of the adducts, Ni(cyclam)(CO2)+ and Ni(cyclam)-
(H)2+, are quite similar.19 Rather, the apparent selectivity is
primarily due to concentration differences of the two species,
i.e., 36 mM CO2 (1 atm) and 0.1 mM protons (pH 4). In the
current work, we report the results of kinetic studies on the decay
of Ni(cyclam)+ as a function of proton and CO2 concentration.
Additionally, we have studied gaseous product yields from the
Continuous Radiolysis. Continuous radiolyses were carried
out at room temperature on 25 mL solutions using a 60Co-
Gammacell with a dose rate of ca. 60 Gy min-1. The exact
absorbed radiation dose was determined with the Fricke
chemical dosimeter assuming G(Fe3+) ) 1.61 µmol J-1 23
.
Solutions were degassed under vacuum or equilibrated with 1
atm of purified CO2 using standard vacuum-line techniques. The
gaseous products were removed from the irradiated sample by
means of an automatic Toepler pump, and their total volume
was measured in a gas buret. The gas was then analyzed by
gas chromatography.
continuous γ-radiolysis of aqueous solutions of Ni(cyclam)2+
.
Results
From these results, we propose a mechanism for the catalytic
Continuous Radiolysis. The gaseous product yields following
continuous γ-irradiation of solutions containing 0.5 mM Ni-
two-electron reduction of protons and CO2 by Ni(cyclam)+.
-
(cyclam)2+ and 0.1 M HCO2 at pH 5.2 in vacuum-degassed
Experimental Section
and CO2-saturated (1 atm) solutions are given in Table 1.
Materials. The compound Ni(cyclam)(ClO4)2 was prepared
according to published procedures20 and was purified by
repeated recrystallization from water and acetone. Sodium
formate was reagent grade and was used as received. Water
was obtained from a Millipore (Milli-Q) water purification
system. Solutions were purged with Ar or saturated with N2O,
CO2 or Ar/CO2 mixtures of known composition with 1 atm total
pressure. The solubility of CO2 in water at 1 atm partial pressure
was taken to be 36 ( 2 mM at 22 ( 2 °C,21 i.e., the temperature
at which the experiments were performed. In some cases,
solutions were degassed by stardard vacuum-line techniques.
Solution pH was adjusted with HClO4 (Merck, Suprapur).
Pulse Radiolysis. Pulse radiolysis was performed using the
12 MeV electron linear accelerator at the FRAE-CNR Institute
in Bologna. The irradiations were carried out at ambient
temperature, 22 ( 2 °C, on samples contained in Spectrosil cells
of 2 cm optical path length. Solutions were protected from the
analyzing light by means of a shutter and appropriate cutoff
filters. The monitoring light source was either a 450 W Xe arc
lamp or a 50 W tungsten-halogen lamp, the latter being used
when performing long time scale measurements. The bandwidth
used throughout the pulse radiolysis experiments was 5 nm. The
radiation dose per pulse was monitored by means of a charge
collector placed behind the irradiation cell and calibrated with
a N2O-saturated solution containing 0.1 M HCO2- and 0.5 mM
methyl viologen (1,1′-dimethyl-4,4′-bipyridinium dication; MV2+)
using Gꢀ ) 9.66 × 10-4 m2 J-1 at 602 nm.22 G(X) represents
the number of moles of species X formed or consumed per joule
of energy absorbed by the system.
-
Pulse Radiolysis. Pulse radiolysis was used to generate eaq
,
CO2•-, and H•, which we have used for the one-electron
reduction of Ni(cyclam)2+ 19
.
However, 2-methyl-2-propanol was
not used as a hydroxyl radical scavenger due to the reaction of
Ni(cyclam)+ with the 2-methyl-2-propanyl radical previously
described by Jubran et al., eq 1.7c Although we did not study
Ni(cyclam)+ + •CH2C(CH3)2OH f Ni(cyclam)2+
+
CH2dC(CH3)2 + OH- (1)
this reaction in detail, we confirmed that Ni(cyclam)+ reacts
with 2-methyl-2-propanyl radical and have estimated the rate
constant to be 1 × 109 M-1 s-1
.
In pulse radiolysis studies using formate as the hydroxyl
radical scavenger in the absence of added CO2, 20-25% of
Ni(cyclam)+ formed by the reduction of Ni(cyclam)2+ with eaq
-
•-
and CO2 disappears according to a process which can be
approximated to first-order, Figure 1. The value of the observed
rate constant, kobs, was independent of [Ni(cyclam)2+] over the
range 0.1-0.5 mM, but increased linearly with dose, Figure 1
inset. The nature of this reaction will be discussed below.
The decay of the remaining Ni(cyclam)+ absorbance yielded
linear plots of 1/(∆A - ∆A∞) vs time, where ∆A and ∆A∞ are
the absorbance changes observed at time t and infinity after
the pulse, Figure S1. A representative fit of the absorbance
change data as a function of time is given in Figure 2. Kinetic
behavior of this type is consistent with a second-order process.
The observed rate constant, kobs, obtained from the absorbance
change data, is a function of the second-order rate constant, k,
the difference between the molar decadic absorption coefficients
of the reagents and products at a specified wavelength, ∆ꢀ, and
the optical path length of the sample, l. For an elementary
second-order reaction kobs ) k/(∆ꢀl). The dimensions of kobs, k,
(19) Kelly, C. A.; Mulazzani, Q. G.; Venturi, M.; Blinn, E. L.; Rodgers,
M. A. J. J. Am. Chem. Soc. 1995, 117, 4911-4919.
(20) (a) Barefield, E. K.; Wagner, F.; Herlinger, A. W.; Dahl, A. R. Inorg.
Synth. 1975, 16, 220-225. (b) Bosnich, B.; Tobe, M. L.; Webb, G.
A. Inorg. Chem. 1965, 4, 1109-112.
(21) Wilhelm, E.; Battino, R.; Wilcock, R. J. Chem. ReV. 1977, 77, 219-
262.
(22) Mulazzani, Q. G.; D’Angelantonio, M.; Venturi, M.; Hoffman, M.
Z.; Rodgers, M. A. J. J. Phys. Chem. 1986, 90, 5347-5352.
(23) Spinks, J. W. T.; Woods, R. J. An Introduction to Radiation Chemistry,
3rd ed.; J. Wiley: New York, 1990; p 100.

CO2 and H+ Reduction by Ni(cyclam)+ in Aqueous Solution
Inorganic Chemistry, Vol. 38, No. 7, 1999 1581
Figure 3. Plot of kobs vs [CO2] from the pulse irradiation of solutions
containing ca. 0.5 mM Ni(cyclam)2+ and 0.1 M HCO2 at pH 2 (1),
-
pH 3 ([), and pH 5.2 (b). Observations were made at 380 nm with an
optical path length of 2.0 cm and a dose per pulse of ca. 50 Gy. The
solid lines represent the best fits of eq 24 to the data.
Figure 1. Time dependence of the absorbance monitored at 380 nm
for the initial decay of Ni(cyclam)+ following the pulse irradiation of
-
Ar-purged solutions containing 0.5 mM Ni(cyclam)2+ and 0.1 M HCO2
The decay of Ni(cyclam)+ is accelerated by the presence of
CO2 but continues to be well represented by the second-order
model. The effect of CO2 concentration on the decay of Ni-
(cyclam)+ was investigated in the pH range 5.2-2, Figure 3.
Investigations of the CO2 reaction at higher pH was not
attempted due to the formation of a species with a visible
at pH ∼ 7. The dose per pulse was 88.6 Gy, and the optical path length
was 2.0 cm. The solid line represents the first-order kinetic fit to the
data. Inset: Plot of the observed first-order rate constant, kobs, vs the
irradiation dose, under the same conditions as those given above.
absorption spectrum similar to that of cis-Ni(cyclam)2+ 24
, which
we tentatively attribute to cis-[Ni(cyclam)(CO3)]. The chemistry
of the latter species was not investigated. At pH 5.2, the lifetime
of Ni(cyclam)+ was identical under 1 atm of H2 or 1 atm of
Ar.
Discussion
In the absence of CO2, the following reactions and associated
G-values (in parentheses, µmol J-1 25
) are anticipated assuming
no reaction between Ni(cyclam)+ and either protons or CO2.
Figure 2. Example of time dependence of absorbance at 380 nm from
the pulse irradiation of Ar-purged solutions containing 0.5 mM Ni-
H2O D> H• (0.06), •OH (0.28), eaq- (0.27),
(cyclam)2+ and 0.1 M HCO2 at pH 2; dose per pulse ) 48.6 Gy;
-
optical path ) 2.0 cm. The solid line represents the second-order kinetic
fit to the data. Inset: plot of log kobs (extrapolated to zero dose) vs pH.
The solid line represents the linear fit to the data based on eq 21.
H2 (0.045), H2O2 (0.07) (2)
H• (0.06) + HCO2- f H2 (0.06) + CO2•- (0.06) (3)
∆ꢀ, and l are s-1, M-1 s-1, M-1 cm-1, and cm, respectively.
As will be discussed below, the observed second-order decay
is the result of a composite reaction mechanism and, as a result,
kobs * k/(∆ꢀl) in the present case. Evaluation of the elementary
rate constants will be discussed below.
•OH (0.28) + HCO2- f H2O + CO2•- (0.28)
eaq- (0.27) + Ni(cyclam)2+ f Ni(cyclam)+ (0.27) (5)
CO2•- (0.34) + Ni(cyclam)2+
(4)
f
CO2 (0.34) + Ni(cyclam)+ (0.34) (6)
The fit of the second-order model to the data required the
presence of a small, i.e., <10%, value for ∆A∞, indicative of
the formation of longer lived products. Unfortunately, we were
unable to obtain reliable kinetic or spectral information regarding
the nature of these products using the pulse radiolysis technique.
However, by irradiating a vacuum-degassed solution containing
The yield of Ni(cyclam)+ is therefore expected to be G ) 0.61,
i.e., the sum of eq 5 and 6. However, an initial loss of 20-
25% of Ni(cyclam)+ was observed that we attribute to the
oxidation of Ni(cyclam)+ by H2O2, eq 7. Hydrogen peroxide is
0.5 mM Ni(cyclam)2+ and 0.1 M HCO2 at pH 5.2 with 200
-
pulses of 50 ns duration and at a repetition rate of 5 Hz, with
each pulse delivering a dose of approximately 10 Gy, a
differential absorption spectrum could be acquired with a
conventional spectrometer ∼45 s after the irradiation. The long-
lived product exhibited a band at 354 nm, a shoulder at ∼400
nm, and a broad weak band at ∼620 nm.
2Ni(cyclam)+ + H2O2 f 2Ni(cyclam)2+ + 2OH- (7)
formed radiolytically with G ) 0.07,25 eq 2. On the basis of
the stoichiometry of eq 7, H2O2 produced during the irradiation
is capable of consuming ca. 23% of the Ni(cyclam)+ yield,
consistent with observation, Figure 1. The ∼8-fold excess of
In general, the observed rate constant, kobs, was found to be
a function of both dose (ca. 20-180 Gy) and pH, Figure S1,
inset. The values of kobs, extrapolated to zero dose, were found
to be 1.3 × 104, 9.4 × 102, 128, 7.3, and 0.73 s-1 at pH 2, 3,
4, 5.2, and ∼7, respectively, Figure 2, inset. At pH 2 and 3, the
values of kobs were independent of irradiation dose, within
experimental error.
(24) (a) Billo, E. J. Inorg. Chem. 1981, 20, 4019-4021. (b) Billo, E. J.
Inorg. Chem. 1984, 23, 2223-2227. (c) Barefield, E. K.; Bianchi,
A.; Billo, E. J.; Connolly, P. J.; Paoletti, P.; Summers, J. S.; Van
Derveer, D. G. Inorg. Chem. 1986, 25, 4197-4202.
(25) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys.
Chem. Ref. Data 1988, 17, 513-886.

1582 Inorganic Chemistry, Vol. 38, No. 7, 1999
Kelly et al.
Ni(cyclam)+over H2O2 results in a pseudo-first-order decay of
the reduced nickel complex. By converting the dose into [Ni-
(cyclam)+], assuming G(Ni(cyclam)+) ) 0.61, a bimolecular
rate constant was obtained from a plot of kobs vs dose, Figure 1
inset, yielding k7 ) 8 × 107 M-1 s-1. As a result of the hydrogen
peroxide oxidation, eq 7, the yield of Ni(cyclam)+ is reduced
to G ) 0.47.
The mechanism of hydrogen peroxide reduction is of some
interest. The multibodied nature of eq 7 implies a composite
reaction. However, the initiating reaction cannot be an outer-
sphere electron transfer, eq 8, since free hydroxyl radicals would
(G ) 0.26) CO2 yield. The yield of H2 (G ) 0.24) is in excess
of the anticipated yield by G ) 0.13. The sum of the CO and
excess hydrogen yields, G ) 0.23, is again equal to about half
the anticipated Ni(cyclam)+ yield. That is, H2 evolution
competes with CO production under low CO2 concentration.
In other words, at pH 5.2 and in the presence of formate, 97 (
5% of the reactivity of Ni(cyclam)+ remaining after the
hydrogen peroxide reaction, eq 7, is directed toward CO2 or
proton reduction.
In contrast to the reactivity of Ni(cyclam)+ with H2O2,
discussed above, or N2O, as reported by Jubran et al.,7c the
coordination of protons or carbon dioxide to Ni(cyclam)+ does
not lead to the spontaneous two-electron reduction of the bound
substrate with the concomitant formation of Ni(cyclam)3+. This
is due to the limited two-electron reduction potential of Ni-
(cyclam)3+, E(Ni(cyclam)3+/+) ) -0.18 V,19 which is marginal
at best for the two-electron reduction at pH 4 of either protons,
E4(H+/H2) ) -0.24 V, or CO2, E4(CO2/CO) ) -0.34 V, but
is more than sufficient for the two-electron reduction of H2O2,
E4(H2O2/H2O) ) 1.53 V, and N2O, E4(N2O/N2) ) 1.53 V.27
Ni(cyclam)+ + H2O2 f Ni(cyclam)2+ + •OH + OH- (8)
-
react rapidly, k ) 3.2 × 109 M-1 s-1 25
,
with HCO2 forming
CO2•-. The latter species is a strong reducing agent, capable of
regenerating Ni(cyclam)+, eq 6. The net result would be the
catalyzed reduction of H2O2 to OH-, with no net loss of Ni-
(cyclam)+. In contrast, an inner-sphere reaction provides a
mechanism for coordination of the hydroxyl radical to the metal
center, eqs 9-11, effectively preventing its release and reaction
with HCO2-. It is not clear if the hydrogen peroxide adduct,
The absorbance of Ni(cyclam)+, monitored at 380 nm, formed
by the pulse irradiation of solutions containing Ni(cyclam)2+
and formate under an argon atmosphere, decays with a rate
constant of 0.73 s-1 (extrapolated to zero dose), at pH ∼ 7. On
decreasing the pH (pH range 7-2) the observed rate constant
increases but remains second-order. In Figure 2 is shown, as
an example, the decay of Ni(cyclam)+, monitored at 380 nm,
observed after pulse irradiation (dose 48.6 Gy) of a solution
containing 0.5 mM Ni(cyclam)2+ and 0.1 M HCO2- at pH 2. It
was impractical to extend this study to lower pH as a
consequence of the depletion of Ni(cyclam)+ due to the
formation of the hydride, eq 15 (K15 ) 62 M-1), which does
Ni(cyclam)+ + H2O2 f [Ni(cyclam)(H2O2)]+
(9)
[Ni(cyclam)(H2O2)]+ f [Ni(cyclam)(OH)]2+ + OH- (10)
[Ni(cyclam)(OH)]2+ f Ni(cyclam)3+ + OH- (11)
represented here as [Ni(cyclam)(H2O2)]+, is a true intermediate
or if hydroxide dissociation, eq 10, is concerted with H2O2
coordination, eq 9.26 In the presence of excess Ni(cyclam)+ the
lifetime of Ni(cyclam)3+ formed by eq 11 would be limited by
the comproportionation reaction, eq 12. As will be discussed
not absorb significantly at an accessible wavelength range.19
A
Ni(cyclam)3+ + Ni(cyclam)+ f 2Ni(cyclam)2+ (12)
Ni(cyclam)+ + H+ h Ni(cyclam)(H)2+
(15)
below, the lifetime of the remaining Ni(cyclam)+ is ultimately
mechanism to account for the second-order decay of Ni-
(cyclam)+ induced by protons is given by eq 15 followed by
eq 16. From this mechanism, assuming that Ni(cyclam)+ is the
only absorbing species at 380 nm, the rate law given by eq 17
is derived.
•
limited by its reaction with CO2 produced by the H• and OH
scavenging reactions of formate, eqs 3 and 4.
Under CO2-saturated conditions, the hydrated electron is
rapidly scavenged according to eq 13. Therefore, as the
eaq- (0.27) + CO2 f CO2•- (0.27)
CO2•- (0.61) + Ni(cyclam)2+
(13)
Ni(cyclam)(H)2+ + Ni(cyclam)+ f P(H+)
(16)
(17)
f
dA380
dt
2K15k16[H+]
(1 + K15[H+])ꢀ380
2
CO2 (0.61) + Ni(cyclam)+ (0.61) (14)
)
(A380)
l
concentration of CO2 is increased, eq 5 becomes unimportant
relative to eq 13 and eq 14 replaces eq 5. The yield of Ni-
(cyclam)+, however, remains unchanged.
Integrating eq 17 yields eq 18,
2K15k16[H+]
(1 + K15[H+])ꢀ380
The product yields following continuous γ-irradiation, Table
1, reveal that in CO2-saturated solution the yield of H2 is close
to that expected on the basis of the radiolytic yield, eqs 2 and
3. The yield of CO (G ) 0.24), which cannot be accounted for
by radiolytic processes, is equal to half the yield of Ni(cyclam)+
after correcting for the H2O2 reaction, eq 7. This is consistent
with the fact that the two-electron reduction of CO2 to CO
requires two reducing equivalents, i.e., 2 Ni(cyclam)+.
In the absence of CO2 there are excess yields of both H2 and
CO and a diminished yield of CO2, Table 1. The yield of CO
(G ) 0.10) is close to the difference in anticipated radiolytic
yield of CO2 from eqs 3, 4, and 6 (G ) 0.34), and the observed
1
1
2
)
)
(A380
)
(18)
(19)
A380 A0
l
or, on rearranging, eq 19,
1
1
)
) kobst
A380 A0
with the observed rate constant, kobs, being a nonlinear function
of the proton concentration, eq 20.
(26) We thank a reviewer for suggesting this point.
(27) Bratsch, S. G. J. Phys. Chem. Ref. Data 1989, 18, 1-21.

CO2 and H+ Reduction by Ni(cyclam)+ in Aqueous Solution
Inorganic Chemistry, Vol. 38, No. 7, 1999 1583
Table 2. Rate Constants Obtained for Reaction 23a
2K15k16[H+]
(1 + K15[H+])ꢀ380
pH
k23, M-1 s-1
kobs
)
(20)
l
5.2
3.0
2.0
(1.08 ( 0.06) × 108
(1.32 ( 0.22) × 108
(2.43 ( 0.07) × 108
At high pH, eq 20 reduces to the linear relation, eq 21.
a See text for details.
log kobs ) log(2K15k16/ꢀ380l) - pH
(21)
decomposition to yield the H2 or CO product and Ni(cyclam)2+
.
The extinction coefficient for Ni(cyclam)+ at 380 nm, ꢀ380
optical path length, l, and the equilibrium constant K15 have
values of 3650 M-1 cm-1, 2 cm, and 62 M-1, respectively.19
linear fit to the data, Figure 2 inset, yields k16 ) (7.2 ( 2.7) ×
,
The stoichiometry of the reaction is represented by eqs 26 and
27.30
A
Ni(cyclam)(H)+ + H+ f Ni(cyclam)2+ + H2 (26)
107 M-1 s-1 28
.
The CO2-Induced Decay of Ni(cyclam)+. The presence of
a second-order decay over a relatively wide CO2 concentration
range suggests a mechanism of the form observed for the proton
reaction, eqs 22 and 23.
Ni(cyclam)(CO2)0 + H+ f Ni(cyclam)2+ + CO + OH-
(27)
Alternatively, the second Ni(cyclam)+ may act as an inner-
sphere reductant, eq 28,
Ni(cyclam)+ + CO2 h Ni(cyclam)(CO2)+
(22)
Ni(cyclam)(CO2)+ + Ni(cyclam)+ f P(CO2) (23)
Ni(cyclam)+ + Ni(cyclam)(S)n+ f Ni2(cyclam)2(S)(n+1)+
(28)
However, in addition to the CO2 reactions, the competitive
proton reactions of eqs 15 and 16 also need to be considered.
Under these conditions, the observed second-order rate constant,
kobs, is given by eq 24.
in which the proton or CO2 bridges the two Ni(I) complexes.
In this case, the stoichiometry of the processes leading to the
final products would be represented by eqs 29 and 30.
2K15k16[H+] + 2K22k23[CO2]
(1 + K15[H+] + K22[CO2])ꢀ380
Ni2(cyclam)2(H)3+ + H+ f 2Ni(cyclam)2+ + H2 (29)
kobs
)
(24)
l
Ni2(cyclam)2(CO2)2+ + H+ f
When [CO2] ) 0, eq 24 reduces to eq 20, a situation that cannot
be obtained under our conditions since CO2 is always generated
by the reaction of the CO2•- radical with Ni(cyclam)2+ to give
Ni(cyclam)+, eq 6.29 Since the amount of CO2 will increase
with increasing irradiation dose, the rate of Ni(cyclam)+ decay
will follow the same trend, accounting for the dose dependence
of kobs, Figure S1, inset. The values of kobs reported in Figure
2, inset, were those extrapolated to zero dose. As we have
already discussed above, at pH 2 and 3 the values of kobs were
independent of irradiation dose. Under these conditions, the
amount of CO2 generated in the system becomes negligible in
comparison to the proton concentration.
2Ni(cyclam)2+ + CO + OH- (30)
Although we are unable to distinguish between these two
pathways, reaction 28 is appealing in light of the isolation by
Fujita et al. of the species [LCo(CO2H)CoL]3+ which contains
a formate bridging group.11a
The decay of Ni(cyclam)+ does not result in the clean
formation of the Ni(cyclam)2+ starting material. Instead, the
second-order decay, observed as a change of absorbance, does
not return to the baseline. The spectrum of the product, observed
ca. 45 s after irradiation, was identical to that obtained by
Sauvage and co-workers following continuous electrolysis of
The rate constants for reaction 23, obtained from the fit of
eq 24 to the data shown in Figure 3, are given in Table 2. The
value of K22 (16 M-1) has been reported previously.19 The
limited data set does not allow us to comment significantly on
the apparent increase in k23 with decreasing pH. It is likely that
the spread of rate constants is due to the accumulation of errors
in the parameters of eq 24. The average value of k23 is (1.6 (
Ni(cyclam)2+ 12b
by Furenlid et al. after reduction of Ni-
,
(cyclam)2+ by sodium amalgam in acetonitrile followed by
addition of CO,31 and by ourselves following pulse radiolysis
or laser flash photolysis of solutions containing CO.32 We have
previously assigned this spectrum to RSRS-Ni(cyclam)(CO)+.32
The formation of the latter species is presumably the result of
a competition between reaction 22 and reaction 31, for which
k31 ) (2.0 ( 0.2) × 109 M-1 s-1 has been reported.32 The
product of reaction 31 has been suggested to be RRSS-Ni-
(cyclam)(CO)+, which isomerizes to RSRS-Ni(cyclam)(CO)+
0.7) × 108 M-1 s-1
.
Reduction of the proton or CO2 substrate requires an
additional reducing equivalent. In the current system, the second
electron is provided by a second Ni(cyclam)+ which can fill
one of two possible roles. First, it may function as an outer-
sphere reductant, eq 25,
with k ) 1.8 ( 0.2 s-1 32
Ni(cyclam)+ + CO f Ni(cyclam)(CO)+
Conclusions
.
(31)
Ni(cyclam)+ + Ni(cyclam)(S)n+
f
Ni(cyclam)2+ + Ni(cyclam)(S)n-1 (25)
The binding of protons or CO2 to Ni(cyclam)+ in aqueous
solution is insufficient to drive the two-electron reduction of
where S is either a proton or CO2 and n is 2 or 1, respectively,
with the product Ni(cyclam)(S)n-1 undergoing facile protonation/
(30) The species Ni(cyclam)(H)+ and Ni(cyclam)(CO2)0 are, formally, Ni-
2-
(II) species containing the H- and CO2 ligands, respectively.
(28) We are grateful to a reviewer for suggesting the use of eq 21 under
our experimental conditions rather than eq 20.
(29) At the lowest dose, ca. 25 Gy per pulse, the amount of CO2 evolved
as a result of reaction 6 is ∼8.5 µM.
(31) Furenlid, L. R.; Renner, M. W.; Szalda, D. J.; Fujita, E. J. Am. Chem.
Soc. 1991, 113, 883-892.
(32) Kelly, C. A.; Mulazzani, Q. G.; Blinn, E. L.; Rodgers, M. A. J. Inorg.
Chem. 1996, 35, 5122-5126.

1584 Inorganic Chemistry, Vol. 38, No. 7, 1999
Kelly et al.
the substrates with concomitant formation of Ni(cyclam)3+
.
cathode surface relative to the second-order mechanism that we
have discovered in homogeneous solution. The low yield
observed in the homogeneous photochemical system of Craig
et al.8b cannot be accounted for by the current work. While we
are not able to address specific points of the photochemical
system, the low yields cannot be due to inefficient reduction of
CO2 by Ni(cyclam)+.
Under the reaction conditions investigated, an additional
equivalent of Ni(cyclam)+ is required for the reduction of the
intermediate Ni(cyclam)(H)2+ or Ni(cyclam)(CO2)+ adducts.
The reactions of Ni(cyclam)+ with Ni(cyclam)(H)2+ and Ni-
(cyclam)(CO2)+, k ) 7.2 × 107 and 1.6 × 108 M-1 s-1
,
respectively, proceed at less than the diffusion controlled limit,
yet the rate constants differ by only a factor of ∼2, suggesting
similar driving forces for the two reactions. It is not clear
whether the second equivalent of Ni(cyclam)+ functions solely
as an outer-sphere reducing agent, or if a bridged dinuclear
intermediate, [(cyclam)Ni(H)Ni(cyclam)]3+ or [(cyclam)Ni-
(CO2)Ni(cyclam)]2+ (or a protonated form thereof), is of
importance. From the reduced mononuclear species Ni(cyclam)-
(H)+ and Ni(cyclam)(CO2)0 or from the dinuclear intermed-
iates [(cyclam)Ni(H)Ni(cyclam)]3+ and [(cyclam)Ni(CO2)Ni-
(cyclam)]2+, nearly quantitative yields of H2 and CO are
obtained.
Acknowledgment. The authors gratefully acknowledge Dr.
H. A. Schwarz for assistance in the kinetic analysis, Dr. A.
Martelli and A. Monti for assistance with the pulse radiolysis
experiments, and Dr. M. Venturi and Dr. M. A. J. Rodgers for
help during the early stages of the work. Funding was provided
in part by Consiglio Nazionale delle Ricerche of Italy (Progetto
Finalizzato Chimica Fine II and Progetto Strategico Tecnologie
Chimiche Innovative), Research Corporation (Grant C-2737),
and the Center for Photochemical Sciences at Bowling Green
State University.
Supporting Information Available: Plots of absorbance and
1/(∆A-∆A∞) vs time and plots of kobs vs dose. This material is available
free of charge via the Internet at http://pubs.acs.org.
Previous reports concerning the lack of reactivity of homo-
geneous Ni(cyclam)+ toward CO2 reduction are clearly over-
stated.16 The origin of the discrepancy is likely due to more
efficient reduction of Ni(cyclam)(CO2)+ while adsorbed to the
IC980902P