We report the first experimental observation of a semiconductor/liquid junction whose open circuit voltage Voc is controlled by bulk diffusion/recombination processes. Variation in temperature, minority‐carrier diffusion length, and/or in majority‐carrier concentration produces changes in the Voc of the n‐Si/CH3OH interface in accord with bulk recombination/diffusion theory. Under AM2 irradiation conditions, the extrapolated intercept at 0 K of Voc vs T plots yields activation energies for the dominant recombination process of 1.1–1.2 eV, in accord with the 1.12‐eV band gap of Si. A crucial factor in achieving optimum performance of the n‐Si/CH3OH interface is assigned to photoelectrochemical oxide formation, which passivates surface recombination sites at the n‐Si/CH3OH interface and minimizes deleterious effects of pinning of the Fermi level at the Si/CH3OH junction. Controlled Si oxide growth, combined with optimization of bulk crystal parameters in accord with diffusion theory, is found to yield improved photoelectrode output parameters, with 12.0±1.5% AM2 efficiencies and AM1 Voc values of 632–640 mV for 0.2‐Ω cm Si materials.

1.
C. M.
Gronet
,
N. S.
Lewis
,
G.
Cogan
, and
J.
Gibbons
,
Proc. Natl. Acad. Sci.
80
,
1152
(
1983
).
2.
S. M. Sze, Physics of Semiconductor Devices (Wiley, New York, 1981).
3.
A. J.
Bard
,
F.‐R. F.
Fan
,
A. J.
Gioda
,
G.
Nagasubramanian
, and
H. S.
White
,
Disc. Farad. Soc.
70
,
19
(
1980
);
A. J.
Bard
,
A. B.
Bocarsly
,
F.‐R. F.
Fan
,
E. G.
Walton
, and
M. S.
Wrighton
,
J. Am. Chem. Soc.
102
,
3671
(
1980
).
4.
G. W.
Cogan
,
C. M.
Gronet
,
J. F.
Gibbons
, and
N. S.
Lewis
,
Appl. Phys. Lett.
44
,
539
(
1984
).
5.
Measurement of Lp was performed by surface photovoltage techniques on the unmounted Si wafers, as well as by an analysis of the near‐infrared spectral response of n‐Si/CH3OH junctions. These two techniques gave results in agreement to ± 15%. We note that for the large Lp samples (where Lp is within a factor of 2 of the wafer thickness) that we have only determined “effective” Lp values.
6.
W.
Shockley
,
Bell. Syst. Tech. J.
28
,
435
(
1949
).
7.
A. L. Fahrenbruch and R. H. Bube, Fundamentals of Solar Cells (Academic, New York, 1983).
8.
For p‐n junctions, slow variations of the other terms in Eq. (2) with temperature typically produce intercepts which are 5%–10% higher than the true activation energy at room temperature. Although there have been few detailed studies of the effect of temperature on semiconductor/liquid junctions, similar effects are expected to prevail at semiconductor/liquid interfaces.
9.
A.
Heller
,
ACS Symp. Ser.
146
,
57
(
1981
);
A.
Heller
,
H. J.
Leamy
,
B.
Miller
, and
W. D.
Johnston
, Jr.
,
J. Phys. Chem.
87
,
3239
(
1983
).
10.
Our measurement techniques have been described in detail in Refs. 1 and 4. We note that our current‐voltage curves are obtained in a potentiostatic, three‐electrode configuration, and our efficiency data are thus photoelectrode parameters of the working electrode.
11.
R. B.
Godfrey
and
M. A.
Green
,
Appl. Phys. Lett.
34
,
790
(
1979
);
A. W.
Blakers
,
M. A.
Green
,
S.
Jiqun
,
E. M.
Keller
,
S. R.
Wenham
,
R. B.
Godfrey
,
T.
Szpitalak
, and
M. R.
Wilson
,
IEEE Electron Dev. Lett.
EDL‐5
,
12
(
1984
).
This content is only available via PDF.
You do not currently have access to this content.