variations of both the Cp, to make it more electron withdrawing and facilitate
catalysis [81], and the diphosphine. By analogy to the Ni family (see Sect. 3.3.3),
Bullock and coworkers have described a number of Co complexes with pendent
amines incorporated into the phosphine backbone to facilitate proton transfer [82–
84]. Initial complexes based on the P
R
2 N
R
0
2 (1,5-diaza-3,7-diphosphacyclooctane)
ligand had modest turnover frequencies, circa 100 s
À1 , at moderate overpotentials
of approximately 400 mV (Fig. 13b). However, these compounds are unstable in
acidic media. Thus complexes including tetradentate phosphines were created
(Fig. 13c). Although more stable compounds, the increased turnover frequency of
these catalysts comes at the price of high, more than 1 V, overpotential. This is
another reminder that decoupling turnover frequency and overpotential is
nontrivial.
Cobalt macrocyclic, or pseudo-macrocyclic, complexes have proven among the
most widespread cobalt-based proton reduction catalysts. For example, Fisher and
Eisenberg demonstrated in 1980 that some cobalt tetraazamacrocyclic complexes
are active in both CO 2 and H
+ reduction [75]. Similarly, porphyrins have been
extensively investigated. Nocera and coworkers showed that cobalt(II) hangman
porphyrins can catalyze proton reduction with less overpotential and weaker acids
than their standard porphyrin cousins (Fig. 13d) [85, 86]. Both features are thought
to be a result of the enhanced proton donation by the carboxylic acid of the
hangman substituent. Bren and coworkers showed that the biologically derived
cobalt-substituted microperoxidase-11 is stable with a turnover number of 25,000,
but the catalytic rate is relatively low at 6.7 s
À1 [87].
The pseudo-macrocyclic diimine and dioxime complexes of cobalt were also
already recognized in the mid-1980s as proton reduction catalysts [88], and the
groups of Artero, Fontecave, and Peters have extensively investigated the catalytic
properties of this family [89]. Using [Co(dmgBF 2 ) 2 (OH 2 ) 2 ] (for dmg¼dimethylglyoxime) (Fig. 13e) as a starting point, they have described variants for hydrogen
evolution from both aqueous and organic solutions, either electrocatalytically or
photocatalytically [90–92]. These complexes are both fast and efficient with some
reported to have turnover frequencies in excess of 1,000 s
À1 and most with
overpotentials of approximately 200 mV. Like the phosphine-coordinated complexes described above, enhanced stability has been achieved using tetradentate
ligands [93]. Importantly, systematic study of these compounds has shown that
although modifications of the equatorial ligand can be used to tune the reduction
potential, there is not a concomitant effect on overpotential for catalysis. The
problem is that reduction potential and nucleophilicity, i.e., ability to be protonated,
are tightly linked [91, 92, 94]. However, modifications of the axial ligand have been
used to tune the rate of catalysis without significant impact on overpotential
[90]. Finally, taking advantage of its ability to coordinate planar macrocyclic
complexes, apomyoglobin has been used as a scaffold to coordinate [Co
(dmgBF 2 ) 2 (OH 2 ) 2 ] and [Co(dmgH) 2 (OH 2 ) 2 ]. In this arrangement, the protein provides the axial ligand [95], and the reduction potential is 100 mV more negative
than the compound free in solution. Although catalytic activity was detected, it is
Biomimetic Complexes for Production of Dihydrogen and Reduction of CO 2
249
catalysis [81], and the diphosphine. By analogy to the Ni family (see Sect. 3.3.3),
Bullock and coworkers have described a number of Co complexes with pendent
amines incorporated into the phosphine backbone to facilitate proton transfer [82–
84]. Initial complexes based on the P
R
2 N
R
0
2 (1,5-diaza-3,7-diphosphacyclooctane)
ligand had modest turnover frequencies, circa 100 s
À1 , at moderate overpotentials
of approximately 400 mV (Fig. 13b). However, these compounds are unstable in
acidic media. Thus complexes including tetradentate phosphines were created
(Fig. 13c). Although more stable compounds, the increased turnover frequency of
these catalysts comes at the price of high, more than 1 V, overpotential. This is
another reminder that decoupling turnover frequency and overpotential is
nontrivial.
Cobalt macrocyclic, or pseudo-macrocyclic, complexes have proven among the
most widespread cobalt-based proton reduction catalysts. For example, Fisher and
Eisenberg demonstrated in 1980 that some cobalt tetraazamacrocyclic complexes
are active in both CO 2 and H
+ reduction [75]. Similarly, porphyrins have been
extensively investigated. Nocera and coworkers showed that cobalt(II) hangman
porphyrins can catalyze proton reduction with less overpotential and weaker acids
than their standard porphyrin cousins (Fig. 13d) [85, 86]. Both features are thought
to be a result of the enhanced proton donation by the carboxylic acid of the
hangman substituent. Bren and coworkers showed that the biologically derived
cobalt-substituted microperoxidase-11 is stable with a turnover number of 25,000,
but the catalytic rate is relatively low at 6.7 s
À1 [87].
The pseudo-macrocyclic diimine and dioxime complexes of cobalt were also
already recognized in the mid-1980s as proton reduction catalysts [88], and the
groups of Artero, Fontecave, and Peters have extensively investigated the catalytic
properties of this family [89]. Using [Co(dmgBF 2 ) 2 (OH 2 ) 2 ] (for dmg¼dimethylglyoxime) (Fig. 13e) as a starting point, they have described variants for hydrogen
evolution from both aqueous and organic solutions, either electrocatalytically or
photocatalytically [90–92]. These complexes are both fast and efficient with some
reported to have turnover frequencies in excess of 1,000 s
À1 and most with
overpotentials of approximately 200 mV. Like the phosphine-coordinated complexes described above, enhanced stability has been achieved using tetradentate
ligands [93]. Importantly, systematic study of these compounds has shown that
although modifications of the equatorial ligand can be used to tune the reduction
potential, there is not a concomitant effect on overpotential for catalysis. The
problem is that reduction potential and nucleophilicity, i.e., ability to be protonated,
are tightly linked [91, 92, 94]. However, modifications of the axial ligand have been
used to tune the rate of catalysis without significant impact on overpotential
[90]. Finally, taking advantage of its ability to coordinate planar macrocyclic
complexes, apomyoglobin has been used as a scaffold to coordinate [Co
(dmgBF 2 ) 2 (OH 2 ) 2 ] and [Co(dmgH) 2 (OH 2 ) 2 ]. In this arrangement, the protein provides the axial ligand [95], and the reduction potential is 100 mV more negative
than the compound free in solution. Although catalytic activity was detected, it is
Biomimetic Complexes for Production of Dihydrogen and Reduction of CO 2
249