Another material tested as semiconducting photocatode was the nanoporous
black silicon together with a FeFe-hydrogenase [56]. This precise preparation of
silicon brings advantages such as absorbing more light than bright silicon and
suppressing the attachment of H 2 bubbles due to its lower surface tension.
A nanoporous structure was implemented to the Si by a several step process
including the formation of Ag nanoparticles and its further etching in acidic solution, yielding pores ranging from 30 to 100 nm in diameter. The nanoporous
structure allowed a high loading of the hydrogenase. The hydrogenase-modified
electrode showed an onset potential of −0.22 versus Ag/AgCl at pH 6.8, much more
positive than the onset potential of the bare Si electrode, starting at −0.50 and equal
to that of platinum nanoparticles. The photobioelectrode delivered hydrogen production under low irradiation power, starting at 10 mW Â cm
−2 , although an
increase on the light intensity yielded a photocurrent increase. The photocurrents
measured were correlated to the hydrogen production, yielding a TOF of 1,300 s
−1
for a photocurrent density of 3 mA Â cm
−2 at −0.5 V versus Ag/AgCl, pH 6.8 and
50 mW Â cm
−2 irradiation.
Silicon can also be covered with a protective layer of TiO 2 and still work as
support for hydrogenases for the H 2 evolution [57]. Such TiO 2 layer prevents
silicon from oxidation and facilitates the excited electron transport through its
conductive band to the immobilized hydrogenase, which reduces the protons to H 2 .
The initial control tested if a FTO electrode modified with amorphous TiO 2 and
hydrogenase was suitable, instead of crystalline TiO 2 . This configuration was able
to reduce protons to H 2 when a potential of −0.35 V versus SHE was applied at pH
6. After 1 h, it produced 90 nmol of H 2 with a Faradaic yield of 96%, demonstrating that amorphous TiO 2 with hydrogenase was suitable. The next experiment
transferred the construction to the surface of a p–Si semiconductor. The p–Si–TiO 2 -
hydrogenase electrode was set to 1 h of white light illumination, producing
25 nmol of H 2 with a Faradaic yield of 95%. Control experiments demonstrated
that lacking any of the components yielded negligible amounts of H 2 . This photobioelectrochemical approach allowed avoiding the use of sacrificial electron
donors. After determining the p–Si–TiO 2 as a successful combination of materials
to develop a suitable photoelectrode to host hydrogenase, the next step was its
nanostructuring. A hierarchical structure of inverse opal TiO 2 was built on the top
of the silicon. The structure showed a high pore control and cavities, where the
NiFeSe-hydrogenase from D. baculatum was immobilized. This construction
showed an onset potential of +0.35 V versus RHE. The experiments of controlled
potential photoelectrolysis (0 V vs. RHE) without enzyme yielded 0.5 lmol H 2
cm
−2 with a Faradaic efficiency of 45% during 5 h, whereas the addition of
hydrogenase yielded 17 lmol H 2 Ácm
−2 with a Faradaic efficiency of 86% during
5 h. The electrode was coupled to a BiVO 4 photoanode to assemble a whole
2-compartment photoelectrochemical cell including a Nafion membrane to separate
them. As a result of 5 h illumination, 0.46 lmol H 2 and 0.20 lmol of O 2 were
synthesized in absence of an external bias; the Faradaic efficiencies were 98 and
84%, respectively. Reisner’s group has very recently established the benchmark for
photoelectroenzymatic H 2 production by integrating the TiO 2 -hydrogenase
Biological Production of Hydrogen
257
black silicon together with a FeFe-hydrogenase [56]. This precise preparation of
silicon brings advantages such as absorbing more light than bright silicon and
suppressing the attachment of H 2 bubbles due to its lower surface tension.
A nanoporous structure was implemented to the Si by a several step process
including the formation of Ag nanoparticles and its further etching in acidic solution, yielding pores ranging from 30 to 100 nm in diameter. The nanoporous
structure allowed a high loading of the hydrogenase. The hydrogenase-modified
electrode showed an onset potential of −0.22 versus Ag/AgCl at pH 6.8, much more
positive than the onset potential of the bare Si electrode, starting at −0.50 and equal
to that of platinum nanoparticles. The photobioelectrode delivered hydrogen production under low irradiation power, starting at 10 mW Â cm
−2 , although an
increase on the light intensity yielded a photocurrent increase. The photocurrents
measured were correlated to the hydrogen production, yielding a TOF of 1,300 s
−1
for a photocurrent density of 3 mA Â cm
−2 at −0.5 V versus Ag/AgCl, pH 6.8 and
50 mW Â cm
−2 irradiation.
Silicon can also be covered with a protective layer of TiO 2 and still work as
support for hydrogenases for the H 2 evolution [57]. Such TiO 2 layer prevents
silicon from oxidation and facilitates the excited electron transport through its
conductive band to the immobilized hydrogenase, which reduces the protons to H 2 .
The initial control tested if a FTO electrode modified with amorphous TiO 2 and
hydrogenase was suitable, instead of crystalline TiO 2 . This configuration was able
to reduce protons to H 2 when a potential of −0.35 V versus SHE was applied at pH
6. After 1 h, it produced 90 nmol of H 2 with a Faradaic yield of 96%, demonstrating that amorphous TiO 2 with hydrogenase was suitable. The next experiment
transferred the construction to the surface of a p–Si semiconductor. The p–Si–TiO 2 -
hydrogenase electrode was set to 1 h of white light illumination, producing
25 nmol of H 2 with a Faradaic yield of 95%. Control experiments demonstrated
that lacking any of the components yielded negligible amounts of H 2 . This photobioelectrochemical approach allowed avoiding the use of sacrificial electron
donors. After determining the p–Si–TiO 2 as a successful combination of materials
to develop a suitable photoelectrode to host hydrogenase, the next step was its
nanostructuring. A hierarchical structure of inverse opal TiO 2 was built on the top
of the silicon. The structure showed a high pore control and cavities, where the
NiFeSe-hydrogenase from D. baculatum was immobilized. This construction
showed an onset potential of +0.35 V versus RHE. The experiments of controlled
potential photoelectrolysis (0 V vs. RHE) without enzyme yielded 0.5 lmol H 2
cm
−2 with a Faradaic efficiency of 45% during 5 h, whereas the addition of
hydrogenase yielded 17 lmol H 2 Ácm
−2 with a Faradaic efficiency of 86% during
5 h. The electrode was coupled to a BiVO 4 photoanode to assemble a whole
2-compartment photoelectrochemical cell including a Nafion membrane to separate
them. As a result of 5 h illumination, 0.46 lmol H 2 and 0.20 lmol of O 2 were
synthesized in absence of an external bias; the Faradaic efficiencies were 98 and
84%, respectively. Reisner’s group has very recently established the benchmark for
photoelectroenzymatic H 2 production by integrating the TiO 2 -hydrogenase
Biological Production of Hydrogen
257