Title of Invention

A STACKED-LAYER TYPE PHOTOELECTRIC CONVERSION DEVICE

Abstract A slacked photoelectric convener comprising a plu- rahity of stacked photoelectric conversion units (3;5) each including one conductivity type layer (31,51), a photoelectric converting layer (32;52) of substantially intrinsic semiconductor, and a reverseconduc- tivity type layer (33.53) that are formed on a substrate (1) sequentially from the light incident side. At least one of the reverse conductivity lypc layer (33) in the front photoelectric conversion unit (3) arranged relatively on the light incident side anti the one conductivity type layer (51) in the rear photoelectric conversion unit (5 (arranged contiguously to the from photoelectric conversion unit (3) includes a silicon com- posite layer (4). The silicon composite layer (4) has a thickness of 20)30 nm and an oxygen concentration of 25-60 atm. and a sili- con-rich phase is included in an amorphous alloy phase of silicon and oxygen.
Full Text DESCRIPTION
A STACKED-LAYER TYPE,PHOTOELECTRIC CONVERSION^DEVICE
Technical Field
The present invention relates to improvement of conversion efficiency of a thin-
film photoelectric conversion device, and more particularly to improvement of
conversion efficiency of a stacked-layer type thin-film photoelectric conversion device
having a plurality of photoelectric conversion units stacked one another. Herein, the
terms "crystalline" and "microcrystalline" are used also for denoting a solid state
partially including amorphous parts, as generally used in the art.
Background Art
In recent years, to simultaneously achieve lower costs and higher efficiency of a
photoelectric conversion device, thin-film photoelectric conversion devices having
almost no problem from the standpoint of resources have attracted attention, and
development thereof has been tried vigorously. Application of the thin-film
photoelectric conversion devices to various uses such as solar batteries, optical sensors,
displays and others have been expected. An amorphous silicon photoelectric
conversion device as one type of the thin-film photoelectric conversion devices can be
formed on a low-temperature, large-area substrate such as glass substrate or stainless
steel substrate, with which cost reduction is expected.
A thin-film photoelectric conversion device generally includes a first electrode,
one or more semiconductor thin-film photoelectric conversion units, and a second
electrode, which are successively stacked on a substrate having an insulative surface.
Each thin-film photoelectric conversion unit includes an i-type layer sandwiched
between a p-type layer and an n-type layer.
The i-type layer, which is a substantially intrinsic semiconductor layer, occupies
the most part of thickness of the thin-film photoelectric conversion unit. Photoelectric
conversion occurs primarily in this i-type layer. Thus, a thicker i-type photoelectric
conversion layer is preferable from the standpoint of light absorption, though an
unnecessarily thick layer leads to increase of both costs and time for deposition thereof.
On the other hand, the p-type and n-type conductive layers serve to generate
diffusion potential in the photoelectric conversion unit The magnitude of the diffusion
potential influences the value of open-circuit voltage that is one of the critical
characteristics of the thin-film photoelectric conversion device. These conductive
layers, however, are inactive layers not contributing to photoelectric conversion. Light
absorbed by the impurities introduced into the conductive layers becomes loss without
contributing to generation of electric power. As such, it is preferable that the p-type
and n-type conductive layers are as thin as possible within a range ensuring generation
of sufficient diffusion potential.
A photoelectric conversion unit or a thin-film solar battery is called an
amorphous photoelectric conversion unit or an amorphous thin-film solar battery in the
case that it includes an amorphous i-type photoelectric conversion layer occupying its
main part whether its p-type and n-type conductive layers are amorphous or crystalline,
and is called a crystalline photoelectric conversion unit or a crystalline thin-film solar
battery in the case that it includes a crystalline i-type layer.
In general, in a semiconductor used for the photoelectric conversion layer, the
light absorption coefficient decreases as the light wavelength increases. In particular,
when the photoelectric conversion material is in a state of a thin film, sufficient light
absorption is not expected in the wavelength region of small absorption coefficient, and
thus the photoelectric conversion amount is restricted depending on the thickness of the
photoelectric conversion layer. Therefore, measures have been taken to generate a
large amount of photocurrent, by forming a light scattering structure for preventing light
having come into the photoelectric conversion device from easily escaping to the outside,
to thereby increase the substantial optical path length and cause sufficient absorption.
For example, when the light is incident on a transparent substrate side, a textured
transparent conductive film having fine unevenness is used as a light incident side
electrode.
Further, as a way of improving the conversion efficiency of the thin-film
photoelectric conversion device, it is known to form a stacked-layer type thin-film
photoelectric conversion device having at least two photoelectric conversion units
stacked one another. In such a way, a front photoelectric conversion unit including a
photoelectric conversion layer having a large band gap is arranged on a light incident
side of the photoelectric conversion device, and back photoelectric conversion units
each including a photoelectric conversion layer (of, e.g., Si-Ge alloy) having a smaller
band gap in turn are successively arranged on the back of the front unit so as to enable
photoelectric conversion over a wide wavelength range of the incident light, to thereby
improve the conversion efficiency of the entire device Of the stacked-layer type
photoelectric conversion devices, one including both the amorphous and crystalline
photoelectric conversion units is called a hybrid type photoelectric conversion device
In the hybrid type photoelectric conversion device, the wavelength of light
photoellectrically convertible with the amorphous silicon is about 800 nm on the longer
wavelength side, whereas the light of longer wavelength of up to about 1100 nm can be
photoellectrically converted by the crystalline silicon, so that effective photoelectric
conversion becomes possible over a wider wavelength range of the incident light.
In the stacked-layer type photoelectric conversion device, the photoelectric
conversion units are connected in series. The shorted-circuit current density (Jsc) of
the photoelectric conversion device is restricted by the smallest one of current values
generated by the photoelectric conversion units therein. Thus, it is preferable that the
current values of the photoelectric conversion units are as even as possible Further,
improvement of conversion efficiency is expected with the greater absolute value of the
current. In the stacked-layer type photoelectric conversion device, a conductive
intermediate reflective layer having both light transmitting and reflecting properties may
be interposed between the photoelectric conversion units. In this case, light having
reached the intermediate reflective layer is partially reflected, and thus it is possible to
increase the light absorption amount and then increase the current generated within the
front photoelectric conversion unit located closer to the light incident side than the
intermediate reflective layer. This means that the effective thickness of the front
photoelectric conversion unit is apparently increased
For example, in the case that an intermediate reflective layer is inserted in a
hybrid type photoelectric conversion device formed of a front amorphous silicon
photoelectric conversion unit and a back crystalline silicon photoelectric conversion unit,
the current generated in the front photoelectric conversion unit can be increased without
increasing the thickness of the amorphous silicon photoelectric conversion layer
Further, when the intermediate reflective layer is included, the thickness of the
amorphous silicon photoelectric conversion layer required to obtain the same current
value can be decreased as compared to the case not including the intermediate layer
As such, it is possible to restrict deterioration in properties of the amorphous silicon
photoelectric conversion unit due to the optical deterioration (Sraebler-Wronsky effect)
that becomes considerable in accordance with increase in thickness of the amorphous
silicon layer.
A conventional intermediate reflective layer is often formed of TCO (transparent
conductive oxide) such as polycrystalline ITO (indium tin oxide) or ZnO. particularly of
ZnO. The ZnO intermediate reflective layer, however, is formed by sputtering or
spraying, which requires a film deposition apparatus besides the plasma CVD (chemical
vapor deposition) apparatus generally used for formation of semiconductor films. This
increases the equipment costs and the production tact time. Further, when sputtering is
employed for formation of the ZnO layer, there is a possibility that the underlying
semiconductor film suffers degradation in its properties due to damages caused by
sputtering
Moreover, it is necessary to form a good ohmic contact at the interface between
the TCO intermediate reflective layer and the semiconductor layer so as to suppress an
adverse effect on the series resistance of the stacked-layer type photoelectric conversion
device. However, it is generally known that it is not easy to form an ohmic contact at
an interface between a ZnO layer and an amorphous silicon layer or a crystalline silicon
layer. More specifically, if the dark conductivity of the ZnO intermediate reflective
layer is lower than l.Ox 102 S/cm, the intermediate reflective layer cannot form a good
ohmic contact with the front photoelectric conversion unit or with the back
photoelectric conversion unit, leading to increase of the contact resistance, which will in
turn decrease the fill factor (FF) of the stacked-layer type photoelectric conversion
device On the contrary, if the dark conductivity of the ZnO layer is greater than l.0x
103 S/cm, the light transmittance will decrease, and then the shorted-circuit current
density (Jsc) of the stacked-layer type photoelectric conversion device will decrease
As such, it is necessary to set the dark conductivity of the TCO layer at a relatively high
level in a range from l.0x 102 S/cm to 1.0 x 103 S/cm by impurity doping or by
adjusting the degree of oxidation.
A large-area thin-film photoelectric conversion device is generally formed as an
integrated type thin-film photoelectric conversion module. The integrated type thin-
film photoelectric conversion module has a structure in which a plurality of photoelectric
conversion cells are separated from each other to have their respective small areas and
are electrically connected in series on a glass substrate Normally, each photoelectric
conversion cell is formed by successively depositing and patterning, on the glass
substrate, a transparent electrode layer, one or more thin-film semiconductor
photoelectric conversion unit layers, and a back electrode layer.
Fig. 30 is a schematic cross sectional view of an example of a conventional
integrated type thin-film photoelectric conversion module having a plurality of stacked-
layer photoelectric conversion cells connected in series, provided with no intermediate
reflective layer. Incidentally, throughout the drawings, the same reference characters
denote the same or corresponding portions. A photoelectric conversion module 101
has a structure in which a transparent electrode layer 103, a front amorphous silicon
photoelectric conversion unit layer 104a, a back crystalline silicon photoelectric
conversion unit layer 104b, and a back electrode layer 106 successively stacked on a
glass substrate 102.
Integrated type thin-film photoelectric conversion module 101 is provided with
first and second isolation grooves 121, 122 for electrically isolating photoelectric
conversion cells 110 from each other, and a connection groove 123 for electrically
connecting the cells in series. First and second isolation grooves 121, 122 and
connection groove 123 are parallel to each other, and extend in the direction
perpendicular to the paper plane of Fig. 30. That is, first isolation groove 121
separates transparent electrode layer 103 into a plurality of regions, corresponding to
the respective photoelectric conversion cells 110. Similarly, second isolation groove
122 separates front photoelectric conversion unit layer 104a, back photoelectric
conversion unit layer 104b, and back electrode layer 106 into a plurality of regions,
corresponding to the respective photoelectric conversion cells 110.
Connection groove 123 provided between first isolation groove 121 and second
isolation groove 122 penetrates through front and back photoelectric conversion unit
layers 104a and 104b. Connection groove 123 is filled with the same metal material as
that of back electrode layer 106, and electrically connects in series back electrode 106 of
one photoelectric conversion cell 110 to transparent electrode 103 of the neighboring
photoelectric conversion cell 110.
An integrated type thin-film photoelectric conversion module of Fig 31 differs
from the module of Fig. 30 only in that a TCO intermediate reflective layer 105 is
inserted between front and back photoelectric conversion unit layers 104a and 104b.
In the photoelectric conversion module of Fig 31. connection groove 123 penetrates
through front photoelectric conversion unit layer 104a, TCO intermediate reflective
layer 105, and back photoelectric conversion unit layer 104b, and is filled with the same
metal material as that of back electrode layer 106 That is, the metal material filled in
connection groove 123 comes into contact with TCO intermediate reflective layer 105
TCO intermediate reflective layer 105 has high dark conductivity in the range of
1.0 x 102 S/cm to 1.0 x 10' S/cm as described above, and thus electric current can easily
flow through TCO layer 105 in a direction parallel to substrate 102. Therefore, back
photoelectric conversion unit 104b is short-circuited due to the current path through
TCO intermediate reflective layer 105, connection groove 123 and back electrode layer
106, causing a large leakage current. As a result, in the photoelectric conversion
module of Fig. 31, the electric power generated at back photoelectric conversion unit
104b can hardly be taken out.
(Prior Art Example 1)
The above-described problem of leakage current can be solved by employing a
structure shown in Fig. 32 (see a first patent document: Japanese Patent Laying-Open
No. 2002-261308). More specifically, in the integrated type thin-film photoelectric
conversion module 101 of Fig. 32, in addition to first and second isolation grooves 121
and 122, a third isolation groove 124 is provided between first isolation groove 121 and
connection groove 123. Third isolation groove 124 penetrates through front
photoelectric conversion unit 104a and intermediate reflective layer 105. Although
third isolation groove 124 may be arranged such that first isolation groove 121 is
located between third isolation groove 124 and connection groove 123, provision of the
third isolation groove between first isolation groove 121 and connection groove 123 as
shown in Fig 32 is more advantageous in that the area effective for power generation
can readily be increased
In the photoelectric conversion module 101 of Fig. 32, provision of third
isolation groove 124 can prevent the current generated by front photoelectric conversion
unit 104a from leaking through TCO intermediate reflective layer 105 and connection
groove 123. However, photoelectric conversion module of Fig. 32 includes third
isolation groove 124 additionally as compared to the module of Fig. 30. The first
through third isolation grooves and the connection groove are usually formed by
patterning with YAG laser or the like. The number of patterning steps increases by one
for photoelectric conversion module of Fig. 32 compared to that of the module of Fig
31, leading to increase of the manufacturing costs and time.
Further, in forming the photoelectric conversion module of Fig. 30, it is possible
to successively form front and back photoelectric conversion units 104a and 104b in a
plasma CVD apparatus. By comparison, in forming the photoelectric conversion
module of Fig. 32, it is necessary to form front photoelectric conversion unit 104a by
plasma CVD and form TCO intermediate reflective layer 105 by sputtering, and then,
the substrate should be temporarily taken out of the vacuum chamber in order to carry
out patterning with YAG laser. Thereafter, with the substrate returned into the
vacuum chamber, it is necessary to form back photoelectric conversion unit 104b by
plasma CVD. As such, the manufacturing time and costs both increase in the case of
the photoelectric conversion module of Fig. 32, as compared to the module of Fig. 30.
Still further, since the substrate is taken out to the ambient air after formation of
TCO intermediate reflective layer 105, impurities in the ambient air may be adsorbed
into the interface between intermediate reflective layer 105 and back photoelectric
conversion unit 104b, possibly leading to degradation in properties of the photoelectric
conversion module, occurrence of peeling of the thin films, and then degradation of
reliability.
Furthermore, since provision of third isolation groove 124 increases loss of the
effective area of the thin-film photoelectric conversion cell, the effect of improving the
properties of the photoelectric conversion module by provision of TCO intermediate
reflective layer 105 may not fully be achieved
(Prior Art Example 2)
A second patent document of Japanese Patent Laying-Open No. 5-95126
discloses an example of using amorphous silicon oxide as a material of one conductivity-
type layer in a stacked-layer type photoelectric conversion device. In this stacked-layer
type photoelectric conversion device, a transparent electrode of SnO2 or the like, a first
p-type layer of amorphous silicon carbide, a first i-type layer of amorphous silicon, a
first n-type layer of amorphous silicon oxide, a second p-type layer of amorphous silicon
carbide, a second i-type layer of amorphous silicon, a second n-type layer of amorphous
silicon, and a metal electrode of Ag or the like, are stacked successively on a glass
substrate. Although it is common to use amorphous silicon or microcrystalline silicon
for the first n-type layer, it is reported in the second patent document that the use of
amorphous silicon oxide having a large band gap can reduce the light absorption loss
As a result, it becomes possible to increase the amount of the light transmitting the first
n-type layer within the front photoelectric conversion unit and reaching the second i-
type layer within the back photoelectric conversion unit, leading to improvement in the
shorted-circuit current density (Jsc) of the stacked-layer type photoelectric conversion
device.
The oxygen concentration of the amorphous silicon oxide layer can be adjusted
as desired. As the oxygen concentration increases, the energy band gap becomes wider,
leading to increase of the transmittance. In the amorphous silicon oxide layer, on the
other hand, conductivity decreases as the oxygen concentration increases It is
reported in the second patent document that the first n-type layer of amorphous silicon
oxide needs to have conductivity of greater than 1 x 10-6 S/cm when irradiated with
light. To this end, when the amorphous silicon oxide is represented by a general
expression of "a-Si)-xOx", the value of x should be less than 0.2.
The second patent document describes that the current generated in the back
photoelectric conversion unit increases because of the increase of the light reaching the
second i-type layer, and then Jsc of the stacked-layer type thin-film photoelectric
conversion device increases, leading to improvement of the conversion efficiency. The
second patent document, however, is silent about improvement of the current generated
in the front photoelectric conversion unit. In the second patent document, the oxygen
concentration of the amorphous silicon oxide film is restricted to less than 20%. Thus,
it can be said that the refractive index of the amorphous silicon oxide film regarding the
light of 600 nm wavelength is at least about 3, as shown in Fig 5. In such a case,
difference in refractive index between the amorphous silicon oxide layer and the
amorphous silicon layer is small, and therefore, increase of current in the front
photoelectric conversion unit by the effect of reflection at the interface cannot be
expected. To make the amorphous silicon oxide layer function as the intermediate
reflective layer, it is necessary to increase its oxygen concentration so as to decrease the
refractive index. In this case, however, the illuminated conductivity of the amorphous
silicon oxide layer decreases, leading to decrease of FF of the stacked-layer type
photoelectric conversion device and then decrease in the conversion efficiency. This is
why the amorphous silicon oxide layer is not used as the intermediate reflective layer in
the second patent document.
First patent document: Japanese Patent Laying-Open No. 2002-261308
Second patent document: Japanese Patent Laying-Open No. 5-95126
Disclosure of the Invention
Problems to be Solved by the Invention
As described above, when a TCO layer of ZnO or the like is used as the
intermediate reflective layer in the stacked-layer type photoelectric conversion device,
the method of forming the TCO layer differs from that for semiconductor layers, and
thus an apparatus or a film deposition chamber for forming the TCO layer should be
additionally provided, which inevitably increases the apparatus costs. Specifically, the
semiconductor layers are formed by plasma CVD, while sputtering or spraying is used
for forming the TCO intermediate reflective layer
Further, in the case that the integrated type thin-film photoelectric conversion
module is formed including the TCO layer as the intermediate reflective layer, the
leakage current possibly occurs in the structure having the first and second isolation
grooves and the connection groove, leading to degradation in properties of the
photoelectric conversion module
Although the problem of the leakage current may be solved by providing the
third isolation groove 124 as shown in Fig. 32, this requires an additional patterning step,
leading to increase of the production costs and time. Further, the interface between the
intermediate reflective layer and the back photoelectric conversion unit may be
contaminated by exposure to the ambient air. Still further, provision of the third
isolation groove increases loss of the effective area of the effective photoelectric
conversion region.
In view of the foregoing problems in the prior art, an object of the present
invention is to provide a stacked-layer type thin-film photoelectric conversion device
and an integrated type thin-film photoelectric conversion module improved in
photoelectric conversion efficiency without substantially increasing costs and time
required for fabrication thereof.
Means to Solve the Problems
In a stacked-layer type photoelectric conversion device according to the present
invention, a plurality of photoelectric conversion units are stacked on a substrate, each
of which includes a one conductivity-type layer, a photoelectric conversion layer of
substantially intrinsic semiconductor, and an opposite conductivity-type layer in this
order from a light incident side. At least one of the opposite conductivity-type layer in
a front photoelectric conversion unit arranged relatively closer to the light incident side
and the one conductivity-type layer in a back photoelectric conversion unit arranged
adjacent to that front photoelectric conversion unit includes a silicon composite layer at
least in a part thereof. The silicon composite layer has a thickness of more than 20 nm
and less than 130 nm and an oxygen concentration of more than 25 atomic % and less
than 60 atomic %, and includes silicon-rich phase parts in an amorphous alloy phase of
silicon and oxygen. Herein, the term "silicon-rich" literally means that the silicon
concentration is high. Thus, the silicon-rich phase part refers to a phase part having a
high silicon concentration.
It is more preferable that the silicon composite layer has the oxygen
concentration of more than 40 atomic % and less than 55 atomic % to achieve low
refractive index. When the silicon composite layer has the thickness of more than 20
nm and less than 130 nm, it can bring about an effective reflection effect. It is more
preferable that the silicon composite layer has the thickness of more than 50 nm and less
than 100 nm to obtain an optimal reflection effect.
That is, when the oxygen concentration of the silicon composite layer is
increased to achieve low refractive index, the reflection effect at the interface with the
neighboring semiconductor layer is improved. The silicon composite layer can achieve
high dark conductivity, despite the high oxygen concentration, because it includes the
silicon-rich phase parts. As such, use of the silicon composite layer can ensure both the
high reflection effect and the high dark conductivity. Accordingly, the current
generated in the front photoelectric conversion unit can be increased, resulting in
improvement in the performance of the stacked-layer type photoelectric conversion
device.
It is desirable that the silicon-rich phase includes a silicon crystal phase. It is
considered that there exists a current path in the thickness direction of the silicon
composite layer through the silicon crystal phase, which is presumably one of the
reasons why a good ohmic contact can be formed. Alternatively, the silicon-rich phase
may include doped amorphous silicon. As well known, amorphous silicon of either n-
type or p-type sufficiently doped with impurities can make a film of sufficiently low
resistivity for forming the ohmic contact.
To achieve the sufficient reflection effect, the silicon composite layer has a
refractive index of preferably more than 1.7 and less than 2.5, and more preferably more
than 1.8 and less than 2.1, regarding light of 600 nm wavelength. When the substrate
of the stacked-layer type photoelectric conversion device is transparent, it is preferable
that reflected spectrum of light having transmitted the transparent substrate and reached
the silicon composite layer includes at least one maximal value and at least one minimal
value of reflectance in a wavelength range from 500 nm to 800 nm, and that the
difference between the maximal value and the minimal value is more than 1%.
It is preferable that the silicon composite layer has dark conductivity of more
than 10-8 S/cm and less than 10-1 S/cm. If the dark conductivity is too low, the fill
factor (FF) of the stacked-layer type photoelectric conversion device will decrease,
leading to degradation of the conversion efficiency. If the dark conductivity is too high,
the leakage current will occur in the thin-film photoelectric conversion module formed
of integrated stacked-layer type photoelectric conversion cells. To achieve the optimal
dark conductivity, in the silicon composite layer measured by Raman scattering, an
intensity ratio of the TO (optical transverse oscillation) mode peak of crystalline silicon
component to the TO mode peak of amorphous silicon component is preferably in a
range of more than 0.5 and less than 10. Further, the dopant atom concentration in the
silicon composite layer is preferably in a range from 3 x 1020 cm-3 to 1.8 x 1021 cm"3.
The silicon composite layer preferably has an optical energy gap of at least 2.2
eV to achieve a sufficient reflection effect. Further, in the silicon composite layer
measured by X-ray photoelectron spectroscopy, difference between the upper most
energy of the O ls photoelectrons having suffered interband excitation loss and the peak
energy of the Ols photoelectrons is preferably at least 2.2 eV.
In an integrated type photoelectric conversion module according to the present
invention, a first electrode layer, a plurality of photoelectric conversion unit layers and a
second electrode layer successively stacked on a substrate are separated by a plurality of
isolation grooves to form a plurality of photoelectric conversion cells, and the cells are
electrically connected in series with each other via a plurality of connection grooves
Each of the photoelectric conversion cells has a plurality of stacked photoelectric
conversion units, each including a one conductivity-type layer, a photoelectric
conversion layer of substantially intrinsic semiconductor and an opposite conductivity-
type layer in this order a the light incident side At least one of the opposite
conductivity-type layer in a front photoelectric conversion unit arranged relatively closer
to the light incident side and the one conductivity-type layer in the back photoelectric
conversion unit arranged adjacent to that front photoelectric conversion unit includes a
silicon composite layer at least in a part thereof. The silicon composite layer has a
thickness of more than 20 nm and less than 130 mn and an oxygen concentration of
more than 25 atomic % and less than 60 atomic %, and includes a silicon-rich phase
parts in an amorphous alloy phase of silicon and oxygen
The first electrode layer is separated into a plurality of regions corresponding to
the plurality of photoelectric conversion cells by a plurality of first isolation grooves, the
plurality of photoelectric conversion unit layers and the second electrode layer are
separated into a plurality of regions corresponding to the plurality of cells by a plurality
of second isolation grooves, and a connection groove is provided between the first
isolation groove and the second isolation groove to electrically connect the first
electrode of one of the cells with the second electrode of the next neighboring cell.
In the case of forming the stacked-layer type photoelectric conversion device,
when the silicon composite layer is deposited in a plasma CVD reaction chamber, it is
preferable that a mixing ratio of dopant source gas to silicon source gas is in a range
from 0.012 to 0.07. Further, in the case of forming the stacked-layer type
photoelectric conversion device, after deposition carried out up to a part of the total
thickness of the silicon composite layer on the substrate in a plasma CVD reaction
chamber, the substrate may be temporarily taken out to expose a surface of the silicon
composite layer to the ambient air, and then after the substrate is introduced again into
the plasma CVD reaction chamber, the remaining part of the total thickness of the
silicon composite layer may be deposited. In this case, it is preferable that the substrate
is taken out from the plasma CVD reaction chamber to the ambient air after at least 60%
of the total thickness of the silicon composite layer is deposited.
Effect of the Invention
In the stacked-layer type photoelectric conversion device according to the
present invention, the incident light is partially reflected at the interface between the
silicon composite layer and the semiconductor layer contacting the same. This can
increase the current generated by the front photoelectric conversion unit. Alternatively,
the photoelectric conversion layer in the front photoelectric conversion unit can be made
thinner to generate the same current as in the case that the silicon composite layer is not
provided.
Since the silicon composite layer can be formed by plasma CVD similarly as in
the case of the photoelectric conversion unit, the front photoelectric conversion unit, the
silicon composite layer and the back photoelectric conversion unit can all be formed in
the similar plasma CVD apparatus. This eliminates the need to provide a facility of
another system required for forming a conventional TCO intermediate reflective layer
Accordingly, it is possible to reduce the production costs and time of the stacked-layer
type photoelectric conversion device.
In the integrated type thin-film photoelectric conversion module according to the
present invention, the silicon composite layer has relatively high dark conductivity in a
direction parallel to the film surface. Accordingly, the patterning of the third isolation
groove for separation of the conventional TCO intermediate reflective layer is not
necessary, so that the production costs and time can be reduced. It is also possible to
avoid reduction in the photoelectric conversion efficiency due to the loss of the power
generating area because of provision of the third isolation groove
Brief Description of the Drawings
Fig. 1 is a graph showing the relation between the refractive index of the silicon
composite layer regarding light of 600 nm wavelength and the dark conductivity thereof.
Fig. 2 is a photograph showing an example of dark field image of the silicon
composite layer obtained by transmission electron microscopy
Fig 3 is a photograph showing an example of high-resolution image of the
silicon composite layer obtained by transmission electron microscopy
Fig. 4 is a graph showing an example of Raman scattering spectrum of the silicon
composite layer.
Fig. 5 is a graph showing the relation between the oxygen concentration of the
silicon composite layer and its refractive index regarding light of 600 nm wavelength
Fig. 6 is a graph showing the relation between the optical energy gap and the
absorption coefficient of the silicon composite layer.
Fig. 7 is a graph showing the relation between the refractive index regarding
light of 600 nm wavelength and the optical gap in the silicon composite layer.
Fig. 8 is a graph showing the photoelectron energy loss spectrum of O ls
measured by X-ray photoelectron spectroscopy in the silicon composite layer
Fig. 9 is a graph showing the energy difference between the upper most energy
of the photoelectron having suffered interband excitation loss of Ols and the peak
energy of the Ols photoelectron measured by X-ray photoelectron spectroscopy in the
silicon composite layer, in connection with the refractive index for light of 600 nm
wavelength.
Fig. 10 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to an embodiment of the present invention.
Fig. 11 is a graph showing the relation between the thickness of the silicon
composite layer and the reflectance at the interface taking account of interference
Fig. 12 is a photograph showing an example of bright field image by transmission
electron microscopy of a cross section in a stacked-layer type photoelectric conversion
device of the present invention
Fig. 13 is a photograph showing a dark field image corresponding to Fig. 12
Fig. 14 is a graph showing the reflectance of light incident on the glass substrate
side in the stacked-layer type photoelectric conversion device of the present invention
Fig. 15 is a schematic cross sectional view of a conventional stacked-layer type
photoelectric conversion device.
Fig. 16 is a schematic cross sectional view of another conventional stacked-layer
type photoelectric conversion device.
Fig. 17 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to another embodiment of the present invention.
Fig. 18 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to a further embodiment of the present invention.
Fig. 19 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to yet another embodiment of the present invention.
Fig. 20 is a graph showing the relative value of spectral sensitivity current in the
stacked-layer type photoelectric conversion device according to the present invention.
Fig. 21 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to yet another embodiment of the present invention.
Fig. 22 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to yet another embodiment of the present invention.
Fig. 23 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to yet another embodiment of the present invention.
Fig. 24 is a schematic cross sectional view of a stacked-layer type photoelectric
conversion device according to yet another embodiment of the present invention.
Fig. 25 is a schematic cross sectional view of a three-unit stacked-layer type
photoelectric conversion device according to yet another embodiment of the present
invention.
Fig. 26 is a schematic cross sectional view of a conventional three-unit stacked-
layer type photoelectric conversion device
Fig. 27 is a schematic cross sectional view of a three-unit stacked-layer type
photoelectric conversion device according to yet another embodiment of the present
invention.
Fig. 28 is a schematic cross sectional view of a three-unit stacked-layer type
photoelectric conversion device according to yet another embodiment of the present
invention
Fig. 29 is a schematic cross sectional view of an integrated type photoelectric
conversion module according to yet another embodiment of the present invention.
Fig. 30 is a schematic cross sectional view of an example of a conventional
integrated type photoelectric conversion module
Fig. 31 is a schematic cross sectional view of another example of a conventional
integrated type photoelectric conversion module
Fig. 32 is a schematic cross sectional view of a further example of a conventional
integrated type photoelectric conversion module
Fig. 33 is a graph showing distribution of phosphorus concentration and oxygen
concentration by SIMS for a stacked-layer type photoelectric conversion device
according to an embodiment of the present invention.
Fig. 34 is a graph showing distribution of oxygen concentration by XPS for a
stacked-layer type photoelectric conversion device according to another embodiment of
the present invention.
Fig. 35 is a graph showing other examples of Raman scattering spectrums of the
silicon composite layers.
Description of the Reference Characters
1: glass substrate, 2: transparent electrode layer; 3: front photoelectric
conversion unit; 3 a: first photoelectric conversion unit; 4: n-type silicon composite layer,
5: back photoelectric conversion unit; 5a: second photoelectric conversion unit; 6. back
electrode layer; 7: n-type silicon composite layer; 8 third photoelectric conversion unit,
31: p-type amorphous silicon carbide layer; 32, 32a: i-type amorphous silicon layer; 33
n-type microcrystalline silicon layer; 33a. n-type silicon composite layer; 33b, 33c. n-
type microcrystalline silicon layer, 34: n-type silicon composite layer; 35: n-type
microcrystalline silicon layer; 36: n-type microcrystalline silicon layer, 37: n-type silicon
composite layer; 38: n-type microcrystalline silicon layer, 39: n-type amorphous silicon
oxide layer, 51: p-type microcrystalline silicon layer, 51a: p-type silicon composite layer;
51b, 51c: p-type microcrystalline silicon layer; 52, 52a: i-type crystalline silicon layer,
53: n-type microcrystalline silicon layer; 53a: n-type silicon composite layer; 53b, 53c:
n-type microcrystalline silicon layer, 81: p-type microcrystalline silicon layer; 81 a: p-type
silicon composite layer, 82: i-type crystalline silicon layer, 83: n-type microcrystalline
silicon layer, 101: integrated type thin-film photoelectric conversion module, 102: glass
substrate; 103: transparent electrode layer; 104a front photoelectric conversion unit;
104b: back photoelectric conversion unit; 105: TCO intermediate reflective layer; 106:
back electrode layer; 107: silicon composite layer, 110: photoelectric conversion cell;
121: first isolation groove; 122: second isolation groove; 123 connection groove; and
124: third isolation groove.
Best Modes for Carrying Out the Invention
To find a material having both a low refractive index and a high conductivity, the
inventors concentrated on investigating a method of forming alloys of silicon and
oxygen by high-frequency plasma CVD. As a result, it was found that a layer including
silicon-rich phase parts in an amorphous alloy phase of silicon and oxygen (herein, called
a "silicon composite layer") can have both a low refractive index and a high conductivity
The graph of Fig. 1 shows the relation between the refractive index and the dark
conductivity measured for the silicon composite layers formed on glass substrates.
Here, spectral ellipsometry was used to measure the refractive index regarding light of
600 nm wavelength. The reason why the wavelength of 600 nm was selected is that in
a hybrid-type photoelectric conversion device as one type of stacked-layer type
photoelectric conversion device, the falling of the spectral sensitivity current of the
amorphous photoelectric conversion unit and the rising of the spectral sensitivity current
of the crystalline photoelectric conversion unit cross each other at the wavelength of
around 600 nm. It can be said that the film well reflecting light of wavelength of
around 600 nm, i.e., the film having a low refractive index regarding light of 600 nm
wavelength, is suitable for increasing current generated by the front amorphous
photoelectric conversion unit arranged closer to the light incident side.
With coplanar electrodes attached to the silicon composite layer on the substrate,
the dark conductivity was measured in connection with current flowing in a direction
parallel to the substrate. As a result of detailed investigation of the inventors, it was
found that the low refractive index of 1.7 to 2.5 and the relatively high dark conductivity
of 1CT-8 S/cm to 10-1 S/cm can be obtained simultaneously in the silicon composite layer,
as seen from Fig 1. Since the refractive index of amorphous silicon or crystalline
silicon regarding light of 600 nm wavelength is about 4, the difference in refractive index
between the silicon composite layer and the silicon layer is large, ensuring a sufficient
reflection effect.
As a preferred embodiment of the silicon composite layer, the silicon-rich phase
can include a silicon crystal phase. Fig. 2 shows a dark field image by transmission
electron microscopy (TEM) of a silicon composite layer taken out from a glass substrate,
viewed from a direction perpendicular to the film surface The dark field image is an
image obtained by electron beam diffracted at a particular crystallographic plane.
While diffraction does not occur in an amorphous portion, diffraction is caused by only
crystals including crystallographic planes at particular Bragg angles with respect to the
electron beam. Thus, the brightly imaged regions in the dark field image always
correspond to crystal phase parts. This means that Fig 2 indicates that the crystal
phase parts are included in the amorphous substance. Fig 3 is a high-resolution TEM
image of the same silicon composite layer as in Fig. 2, again viewed from the direction
perpendicular to the film surface. It is evident that the film contains the crystal phase
parts, since it can be confirmed that there exist partial regions in which crystal lattices
are arranged regularly.
The graph of Fig. 4 shows a Raman scattering spectrum of the silicon composite
layer of Fig. 2. In this graph, there is a steep TO mode peak of crystalline silicon in the
vicinity of 520 cm-1. This means that the silicon-rich phase parts in the silicon
composite layer include silicon crystals. At this time, the intensity ratio of the TO
mode peak of the crystalline silicon component is 2.5 with respect to the TO mode peak
of the amorphous silicon component in the vicinity of 480 cm'1.
It has been found from experiments that the silicon composite layer having both
a low refractive index and a high dark conductivity can be formed by plasma CVD,
using SiH4, CO2, H2 and PH3 (or B2H6) as reaction gases, under a large ratio of H2/SiH4
for the condition for forming a microcrystalline layer, with a ratio of CCVSiH^ set in a
range from about 2 to about 10. At this time, as conditions for generating plasma, it is
possible to use capacitive coupling parallel-plate electrodes, a power supply frequency
of 10-100 MHz, a power density of 50-500 mW/cm2, a pressure of 50-1000 Pa, and a
substrate temperature of 150-250°C. With increase of the CCVSiRj ratio, the oxygen
concentration of the silicon composite layer increases monotonously. On the other
hand, it has been found through experiments that when the CO2/S1H4 ratio is changed in
a range of more than 0 and less than 4, the carbon concentration of the silicon composite
layer is less than 1 atomic %, indicating that carbon is hardly introduced into the film,
compared to oxygen.
The graph of Fig. 5 shows the relation between the oxygen concentration in the
silicon composite layer and the refractive index regarding light of 600 nm wavelength
As will be described later with reference to Fig 20, to obtain the refractive index less
than 2.5 of the silicon composite layer so as to increase the output current of the
stacked-layer type photoelectric conversion device by the reflection effect of the silicon
composite layer, the oxygen concentration may be set to exceed 25 atomic % as seen
from Fig. 5. To obtain the refractive index of less than 2 1 so as to increase the output
current by more than 10% by the reflection effect (see Fig. 20), the oxygen
concentration may be set to exceed 40 atomic % (see Fig. 5).
The dark conductivity of the silicon composite layer is determined depending on
the oxygen concentration, the doping impurity (P or B) concentration, and the
proportion of the silicon crystal phase parts in the layer. The oxygen concentration is
preferably 25-60 atomic % in order to adjust the dark conductivity of the silicon
composite layer to 10- 8 S/cm to 10-1 S/cm and the refractive index to 1.7 to 2.5. That
is, there is a preferable upper limit for the oxygen concentration of the silicon composite
layer, since increase of the oxygen concentration in the silicon composite layer causes
decrease of not only the refractive index but also the dark conductivity.
In the case of an n-type silicon composite layer, it is preferable to set the
concentration of P as the doping impurity to more than 5 x 1019 cm-3 and less than 2 x
1022 cm-3. In the case of a p-type silicon composite layer, the concentration of B as the
doping impurity is preferably set to more than 5 The dark conductivity of the silicon composite layer increases as the P or B
concentration increases. However, when the doping impurities become excessive, the
proportion of the crystal phase parts decreases, leading to undesirable decrease of the
dark conductivity of the silicon composite layer Thus, it is preferable to adjust the P
or B concentration within the above-described range.
Further, as an index of the proportion of the silicon crystal phase parts in the
silicon composite layer, the intensity ratio of the TO mode peak of the crystalline silicon
component with respect to the TO mode peak of the amorphous silicon component
measured by Raman scattering is preferably more than 0.5 and less than 10. Although
the dark conductivity of the silicon composite layer increases as the peak intensity ratio
increases, the proportion of the amorphous silicon oxide in the silicon composite layer
decreases and the refractive index increases when the peak intensity ratio becomes too
large. Thus, it is preferable to adjust the peak intensity ratio of the Raman scattering
within the above-described range.
In the graph of Fig 1, the dark conductivity varies with the same refractive index
of the silicon composite layer, because conditions of oxygen concentration, impurity
concentration and proportion of the silicon crystal phase parts are changed.
In the silicon composite layer of the present invention, the dark conductivity in
the film thickness direction can be maintained high even if the oxygen concentration is
increased to lower the refractive index to less than 2.5, since the silicon-rich phase parts
presumably serves as electron transport paths in the film thickness direction. As such,
provision of the silicon composite layer between the front and back, photoelectric
conversion units in the stacked-layer type photoelectric conversion device would only
slightly affect the series resistance of the photoelectric conversion device, and thus it is
possible to design a silicon composite layer having a thickness and a refractive index
optimal for confinement of light. Moreover, it is readily possible to control the
retractive index of the silicon composite layer by simply adjusting the oxygen
concentration by changing the CO2/SiH4 gas ratio. As such, it can be expected to
enhance the light confinement effect by more precise optical design, e.g., by periodically
changing the refractive index in the film thickness direction.
In order to suppress the adverse effect on the series resistance of the stacked-
layer type photoelectric conversion device, the conventional intermediate reflective layer
formed of TCO such as ZnO is required to have high dark conductivity of 102 S/cm to
103 S/cm. As is generally known, it is not easy to form an ohmic contact at the
interface between ZnO and amorphous or crystalline silicon. In particular, the ohmic
contact is unlikely to be formed at the interface between ZnO and p-type amorphous
silicon or p-type crystalline silicon. In contrast, it has been found through detailed
investigation that the silicon composite layer having dark conductivity of 10-8 S/cm to
10-1 S/cm can realize a favorable ohmic contact with an amorphous or crystalline silicon
photoelectric conversion unit. One conceivable reason thereof is that the silicon
composite layer is a semiconductor thin film including silicon as its main component,
similarly to the amorphous silicon and the crystalline silicon.
Further, one of the reasons why the favorable ohmic contact can be formed is
that current paths via the silicon crystal phase parts presumably exist in the thickness
direction of the silicon composite layer. While the dark conductivity in Fig. 1 was
measured through current flowing in a direction parallel to the film surface, current
flows mainly in the thickness direction of the silicon composite layer in the stacked-layer
type photoelectric conversion device. In the TEM dark field image of Fig. 2, crystal
phase parts appear as bright spots. It can be said that the silicon crystal phase parts
penetrating the overall thickness of the silicon composite layer are spread in two
dimensions. Thus, it is considered that even if the dark conductivity in the direction
parallel to the film surface is low in the silicon composite layer introduced in the
stacked-layer type photoelectric conversion device, the current flows in the thickness
direction primarily through the silicon crystal phase parts, and as a result it is possible to
suppress the increase of the series resistance of the stacked-layer type photoelectric
conversion device
The graph of Fig. 6 shows the relation between the optical energy (E) inversely
proportional to wavelength of light and the absorption coefficient (a) regarding the
silicon composite layer. In this graph, a plurality of curved lines a through g
correspond to a plurality of silicon composite layers having different optical energy gaps.
The optical gap reflects the forbidden bandwidth in the silicon composite layer. With a
V(aE) line plotted for optical energy (E), the optical gap is obtained as an energy at an
intersection between an extrapolated line of the linear portion in the plotted line and the
axis of a=0 (i.e., the optical gap is obtained from so-called Tauc's plot).
In the graph of Fig. 6, the curved lines shift to the right or downward with
increase of the optical gap of the silicon composite layer, and namely the absorption
coefficient exponentially decreases in comparison at the same light energy. More
specifically, when the silicon composite layer is used as the intermediate reflective layer,
the absorption loss can be decreased exponentially by increasing the optical gap.
Compared to the case of the optical gap of 2.05 eV corresponding to the optical gap
range that is disclosed as suitable in Prior Art Example 2 mentioned above, if the optical
gap is set to exceed 2.2 eV, the absorption coefficient can be decreased to less than 1/3
in the wide optical energy range. In other words, the absorption loss of the stacked-
layer type photoelectric conversion device can be reduced when the silicon composite
layer having an optical gap of more than 2.2 eV is employed as the intermediate
reflective layer.
The graph of Fig. 7 shows the relation between the refractive index regarding
light of 600 nm wavelength and the optical gap in connection with the silicon composite
layer. As seen from Fig. 7, the optical gap increases as the refractive index decreases,
and it considerably increases in the range of the refractive index less than 2.2. If the
optical gap is wide, the absorption loss due to the silicon composite layer decreases,
which is preferable for improvement in properties of the stacked-layer type photoelectric
conversion device. That is, when the refractive index of the silicon composite layer is
set to less than 2.2, not only the reflection effect but also the effect of reducing the
absorption loss because of the increase of the optical gap is enhanced, which leads to
improvement in properties of the stacked-layer type photoelectric conversion device.
When a silicon composite layer having a thickness of more than 300 nm is
deposited on a glass substrate, its optical gap can readily be determined from the
transmission spectrum, or from the transmission spectrum and the reflection spectrum.
However, it is difficult to measure the optical gap of a film having a very small thickness
or the optical gap of a layer included in a stacked-layer film Thus, measurement of X-
ray photoelectron spectroscopy (XPS) was carried to obtain another index, similar to
the optical gap, reflecting the forbidden bandwidth of the silicon composite layer. The
graph of Fig. 8 shows the photoelectron energy loss spectrum measured by the X-ray
photoelectron spectroscopy for the silicon composite layer.
In Fig. 8, an energy difference (hereinafter, referred to as "Exps") between the
upper most energy of the photoelectron having suffered interband excitation loss of Ols
and the peak energy of the Ols photoelectron was obtained as the index of the forbidden
bandwidth of the silicon composite layer. The X-ray photoelectron spectroscopy may
be abbreviated to XPS or ESCA as well. The inner-orbit photoelectrons exited by
irradiating a specimen with X-ray can provide some photoelectrons escaping into a
vacuum and detected without losing their excited energy, and the other photoelectrons
escaping into the vacuum and detected after at least partially loosing the excited energy
in the film. The energy loss in the film is primarily attributable to interband excitation
and plasmon excitation. In the case of the silicon composite layer, the plasmon
excitation is about 20 eV which is sufficiently greater than that of the interband
excitation, so that those signals can be separated from each other. With the linear
portion in the spectrum region including the interband excitation loss being extrapolated,
its crossing point with the base line was determined as the upper most energy, and then
the difference between the upper most energy and the Ols peak energy was obtained as
Exps. In the example of Fig. 8, Exps is 3 4 eV. XPS can be measured even for a film
thinner than 10 nm. Further, if the XPS measurement is carried out while removing the
film surface layer by ion sputtering, it is also possible to obtain the Exps profile in the
film depth direction. According to XPS, therefore, it is possible to measure Exps even
for a thin film of less than 10 nm thickness or for any layer included in the stacked-layer
type photoelectric conversion device.
The graph of Fig. 9 shows the relation between the refractive index regarding
light of 600 nm wavelength and Exps in connection with the silicon composite layer
As shown in this graph, Exps suddenly increases as the refractive index becomes less
than about 2.2. That is, it is shown that Exps may be set to more than 2.2 eV in order
to obtain the refractive index of less than 2 2 for the silicon composite layer in the
stacked-layer type photoelectric conversion device, so that the current is increased by
the reflection effect of the silicon composite layer. It can be said that the refractive
index of less than 2.2 is preferable, not only from the standpoint of the reflection effect,
but also for decreasing the absorption loss in the silicon composite layer.
As already described, in the case of employing a conventional TCO intermediate
reflective layer, it was necessary for the layer to have high dark conductivity of 102 S/cm
to 103 S/cm to suppress the adverse effect on the series resistance of the stacked-layer
type photoelectric conversion device. In contrast, it has been found through detailed
investigation that the silicon composite layer can realize a good ohmic contact with
either of the photoelectric conversion units of amorphous silicon and crystalline silicon,
despite its low dark conductivity in the direction parallel to the film surface.
Since the silicon composite layer can form a favorable ohmic contact even if its
dark conductivity is lower by several to ten digits than that of TCO, the structure of the
integrated type photoelectric conversion module can be simplified enabling improvement
in conversion efficiency of the module as well as cost reduction. That is, as will be
described later in detail, in the integrated type thin-film photoelectric conversion module,
even the structure unprovided with third isolation groove 124 as shown in Fig. 32 does
not suffer the problem of leakage current. Thus, the number of times of patterning is
decreased by 1 in the integrated type photoelectric conversion module, thereby leading
to reduction of the production costs and time Further, since the third isolation groove
124 is unnecessary, the area loss in the photoelectric conversion region can be decreased,
thereby leading to improvement in conversion efficiency of the integrated type
photoelectric conversion module.
As another preferred embodiment of the silicon composite layer, the silicon
crystal phase parts may be undetectable in the silicon-rich phase part. That is, there is a
case that the silicon-rich phase part includes only amorphous silicon.
Herein, the term "amorphous" refers to the state where the crystal phase is
undetectable. It depends on plasma CVD conditions as well whether the silicon
composite layer includes the silicon crystal phase parts or not. The inventors have
found that, even if the silicon crystal phase parts are not detected in the silicon
composite layer with a known analyzing technique, the silicon composite layer may
effectively serve as the intermediate reflective layer The following models (l)-(4)
show possible reasons why the silicon composite layer can effectively serve as the
intermediate reflective layer in the stacked-layer type photoelectric conversion device
whether or not it includes detectable crystal phase parts.
(1) The electrons may be transported in the film thickness direction through an
undetectable minute crystal phase parts or an undetectable low-density crystal phase
parts within the silicon composite layer.
(2) The electrons may be transported in the film thickness direction through a
silicon-rich region within the silicon composite layer. As is well known, with the
impurities sufficiently doped, the n-type amorphous silicon exhibits high dark
conductivity of 10-3 S/cm to 10-1 S/cm and the p-type amorphous silicon exhibits high
dark conductivity of 10-5 S/cm to 10-2 S/cm.
(3) In the state immediately before generation of the crystal phase parts, a
precursors of the crystals are generated in the silicon composite layer. Minute regions
of lower resistivity compared to the normal amorphous substance are locally created,
and accordingly, electrons can be transported in the film thickness direction It is well
known that in the case of a silicon-based thin film deposited by plasma CVD, generated
crystal phase parts each grow in a columnar shape in the film thickness direction. Thus,
it is considered that, even in the amorphous substance immediately before generation of
the crystal phase parts, the minute regions of low resistivity grow in the film thickness
direction, and electrons are likely to travel in the thickness direction.
(4) Even the silicon composite layer nol including the crystal phase parts can
form a good ohmic contact, probably because the silicon composite layer is a
semiconductor film including silicon as its main component, similarly to the amorphous
silicon and the crystalline silicon.
It is noted that presence of amorphous silicon as the silicon-rich phase can
readily be confirmed, since the TO mode peak of the amorphous silicon appears in the
vicinity of 480 cm"1 in Raman scattering measurement.
Fig. 10 schematically shows a cross sectional view of a stacked-layer type
photoelectric conversion device according to an embodiment of the present invention.
In this stacked-layer type photoelectric conversion device, a transparent electrode layer
2, an amorphous silicon photoelectric conversion unit as a first photoelectric conversion
unit 3, a silicon composite layer 4 of one conductivity type (p-type or n-type), a
crystalline silicon photoelectric conversion unit as a second photoelectric conversion
unit 5, and a back electrode layer 6 are successively stacked on a glass substrate 1 as a
transparent substrate. Although silicon composite layer 4 of the one conductivity type
is shown as a layer separate from each of first photoelectric conversion unit 3 and
second photoelectric conversion unit 5 in Fig. 10. it may be considered as a part of the
one conductivity type (p-type or n-type) layer included in either first photoelectric
conversion unit 3 or second photoelectric conversion unit 5.
For the transparent substrate, a transparent resin film or the like may be
employed besides glass, though a material as transparent as possible is preferred so as to
make a greater amount of sunlight transmitted to and absorbed in the photoelectric
conversion layer. For the similar reason, it is preferable to form a reflection-free
coating on the substrate surface from which sunlight enters, to thereby reduce the light
reflection loss.
For transparent electrode layer 2, transparent conductive oxide (TCO), such as
tin oxide (SnO2), indium tin oxide (ITO), zinc oxide (ZnO) or the like can be used,
among which SnO2 is particularly preferable. Further, the interface between
transparent electrode layer 2 and photoelectric conversion unit 3 preferably has
unevenness having an average pitch of 200-900 nm To this end, transparent electrode
layer 2 preferably has an average grain size of 200-900 nm.
Although two photoelectric conversion units are stacked in Fig. 10, it is of
course possible to stack three or more photoelectric conversion units one another, as
will be described later. In the case that three or more photoelectric conversion units
are stacked, only one silicon composite layer 4 may be inserted, or the layer may be
inserted between every possible pair of the photoelectric conversion units
The photoelectric conversion unit includes a one conductivity-type layer, an i-
type photoelectric conversion layer of substantially intrinsic semiconductor, and an
opposite conductivity-type layer The one conductivity-type layer may be a p-type
layer or an n-type layer, and correspondingly, the opposite conductivity-type layer
becomes an n-type layer or a p-type layer. Since a p-type layer is normally arranged on
the light incident side of the photoelectric conversion device, one conductivity-type
layers 31, 51 in the structure of Fig. 10 are normally p-type layers, and opposite
conductivity-type layers 33, 43 are n-type layers I-type layers 32, 52, designed to
absorb light and carry out photoelectric conversion, preferably have different band gaps
or are formed of materials different in light absorption wavelength range from each other
It is preferable that the stacked-layer type photoelectric conversion device as a whole is
capable of absorbing light in the main wavelength range (400-1200 nm) of sunlight. To
this end, it is preferable to select a combination of an i-type amorphous silicon layer and
an i-type amorphous silicon germanium layer, or a combination of an i-type amorphous
silicon layer and an i-type crystalline silicon layer.
In the case that an amorphous silicon thin-film photoelectric conversion unit is
formed as the front, first photoelectric conversion unit 3 in Fig. 10, although n-i-p-layers
may be stacked in this order by plasma CVD, it is preferable to stack p-i-n-layers in this
order from the standpoint of achieving better conversion efficiency. For example, a p-
type amorphous silicon carbide layer 31 doped with boron as the conductivity type
determining impurity at a dose of more than 0.01 atomic %, an i-type amorphous silicon
layer 32 to be the photoelectric conversion layer, and an n-type microcrystalline silicon
layer 33 doped with phosphorus as the conductivity type determining impurity at a dose
of more than 0.01 atomic %, may be stacked in this order, although the constituent
layers are not restricted to these examples. As the p-type layer, amorphous silicon,
microcrystalline silicon, or amorphous silicon nitride may also be employed instead. As
the n-type layer, amorphous silicon may also be employed instead. The one
conductivity-type (p-type or n-type) layer has a thickness of preferably more than 3 nm
and less than 100 nm, and more preferably more than 5 nm and less than 50 nm
Silicon composite layer 4, one of the most important features of the present
invention, serves to reflect a part of light having reached thereto into the front
photoelectric conversion unit 3 located on the light incident side, and transmit the
remaining part of the light to the back photoelectric conversion unit 5. When a silicon-
based material is used for the photoelectric conversion layer, the refractive index of the
photoelectric conversion layer regarding light of 600 nm wavelength is about 4. Thus,
the refractive index of silicon composite layer 4 is preferably in the range of more than
1.7 and less than 2.5. The dark conductivity of silicon composite layer 4 is preferably
more than 10-8 S/cm and less than 10-1 S/cm, since output current flows therethrough.
The graph of Fig. 11 shows the relation between the thickness of silicon
composite layer 4 and the reflectance of light of 600 nm wavelength, taking account of
interference between its both sides. At this time, the silicon composite layer has the
refractive index of 2 regarding light of 600 nm wavelength. It is found from this graph
that silicon composite layer 4 preferably has a thickness of more than 20 nm and less
than 130 nm to cause more than 10% of light to be reflected to the front photoelectric
conversion unit 3 side. It is also found that silicon composite layer 4 preferably has a
thickness of more than 50 nm and less than 100 nm such that more than 30% of light is
reflected to the front photoelectric conversion unit 3 side.
Front photoelectric conversion unit 3, silicon composite layer 4, and back
photoelectric conversion unit 5 are preferably formed continuously, without being taken
out to the ambient air. Here, "not taken out to the ambient air" means that they are
kept in the environment free from surface contamination. To this end, any of various
known methods may be employed.
The structure and thickness of silicon composite layer 4 thus adapted to the
stacked-layer type photoelectric conversion device can be analyzed with transmission
electron microscopy (TEM). Fig. 12 is a bright-field TEM image of the stacked-layer
type photoelectric conversion device having a structure of glass substrate / transparent
electrode layer / amorphous photoelectric conversion unit / silicon composite layer /
crystalline photoelectric conversion unit / back electrode layer, showing a cross section
taken in the thickness direction in the vicinity of the silicon composite layer. In the
bright field image of Fig 12, the silicon composite layer (mc-SiO) is shown whitish
because of its lower density as compared with the amorphous silicon layer (a-Si) and the
crystalline silicon layer (poly-Si) on both sides thereof. Fig. 13 is a dark field image
showing the same region as that of the bright field image of Fig. 12 In this dark field
image, there are minute bright parts spread through the silicon composite layer,
indicating existence of minute crystal phase parts in the silicon composite layer.
The oxygen concentration and the P or B concentration of the silicon composite
layer in the stacked-layer type photoelectric conversion device can be detected by using
any of known analyzing methods. For example, SIMS (secondary ion mass
spectrometry), ESCA (electron spectroscopy for chemical analysis), EPMA (electron
probe micro analysis), or Auger electron spectroscopy can be employed to analyze the
composition, after removing the back electrode layer of the stacked-layer type
photoelectric conversion device by an acid such as HC1, while removing the surface
layer by wet etching, plasma etching, ion sputtering or the like to change the depth of
the region to be measured
Further, the refractive index of the silicon composite layer in the stacked-layer
type photoelectric conversion device can be detected with ellipsometry, by exposing the
silicon composite layer. Presence/absence of the silicon composite layer can readily be
determined based on a difference in reflectance of light having entered from the glass
substrate. The graph of Fig. 14 shows the reflection spectra of light having entered
from the glass substrate side, in the stacked-layer photoelectric conversion devices
including a 60 nm thick silicon composite layer, a 30 nm thick silicon composite layers,
and no silicon composite layer, respectively. With provision of the silicon composite
layer, light is reflected back and forth within the amorphous photoelectric conversion
unit thereby causing interference, and then, a maximal reflectance value and a minimal
reflectance value having a difference more than 1% therebetween appear in the
wavelength range of 500-800 nm. In contrast, with no provision of the silicon
composite layer, well-defined maximal and minimal reflectance values do not appear in
this wavelength range.
In the case that a crystalline silicon photoelectric conversion unit, for example, is
formed as second photoelectric conversion unit 5 on silicon composite layer 4 in the
stacked-layer type photoelectric conversion device of Fig. 10, it is preferable that p-i-n-
layers are stacked in this order by plasma CVD at a substrate temperature of lower than
400°C. That is, by forming the same at a low temperature, the crystalline silicon-based
photoelectric conversion layer as i-type photoelectric conversion layer 52 preferably
includes a great number of hydrogen atoms that terminate and inactivate defects in
crystal grain boundaries and inside grains. Specifically, the hydrogen content of i-type
photoelectric conversion layer 52 is preferably in a range of 1-15 atomic %. Further, it
is preferable that the i-type layer is formed as a film of substantially intrinsic
semiconductor having a density of conductivity type determining impurity atoms of less
than 1 x 1018 cm"3.
Further, it is preferable that many of crystal grains included in i-type crystalline
silicon layer 52 each grow and extend in a columnar shape in a direction approximately
perpendicular to the film surface, and cause a preferential orientation plane (110) parallel
to the film surface. In the case of the crystalline silicon thin film having such crystal
orientation, even if the upper surface of transparent electrode 2 is substantially flat, the
upper surface of photoelectric conversion unit 5 deposited thereon exhibits a textured
surface structure with fine unevenness. If the upper surface of transparent electrode 2
also has a textured surface structure with unevenness, a textured structure with fine
unevenness superposed on the unevenness of transparent electrode 2 is caused on the
upper surface of photoelectric conversion unit 5, and thus it is possible to obtain a
textured structure having a favorable light-confining effect suitable for reflecting light of
a wide wavelength range. The i-type crystalline silicon layer preferably has a thickness
of more than 0.1 mm and less than 10 mm. It is noted that the i-type crystalline silicon
layer may be replaced with a layer of amorphous silicon germanium (containing
germanium of, e.g., 30 atomic %) or crystalline silicon germanium as an alloyed material,
since it is preferred that the thin-film photoelectric conversion unit is capable of
absorbing light in the main wavelength range (400-1200 nm) of sunlight.
Preferably, p-type crystalline silicon layer 51 in crystalline silicon photoelectric
conversion unit 5 has a thickness in a range of 3 nm to 25 nm. If p-type crystalline
silicon layer 51 is thinner than 3 nm, it is not possible to generate an internal electric
field sufficient for externally extracting carriers generated by light irradiation within i-
type crystalline silicon photoelectric conversion layer 52. If p-type layer 5 lis thicker
than 25 nm, the light absorption loss is increased in p-type layer 51 itself. The
thickness of n-type crystalline silicon layer 53 is preferably in a range from 3 nm to 20
nm, similarly as in the case of the p-type crystalline silicon layer.
As back electrode layer 6, at least one metal layer containing at least one of Al,
Ag, Au, Cu, Pt and Cr is preferably formed by sputtering or evaporation. A TCO layer
(not shown) of ITO, SnO2, ZnO or the like may be formed between photoelectric
conversion unit 5 and metal electrode 6.
For Example, back electrode 6 is preferably a multi-layered film having a 10 nm
to 150 nm thick ZnO film and a 30 nm to 500 nm thick silver film formed in this order.
If the ZnO film is thinner than 10 nm, adhesion between crystalline silicon photoelectric
conversion unit 5 and silver film 6 will be poor On the other hand, if it is thicker than
150 nm, light absorption in the ZnO film itself will increase, leading to degradation in
properties of the photoelectric conversion device. Silver film 6 serves to reflect light of
the longer wavelength side that can not easily be absorbed by crystalline silicon
photoelectric conversion unit 5, to make the light reenter the photoelectric conversion
unit 5. If silver film 6 is thinner than 30 nm, the effect as the reflective layer will
considerably decrease. If it is thicker than 500 nm, the production costs will increase.
Although transparent substrate 1 has been employed in the example of Fig. 10,
the stacked-layer type photoelectric conversion device may be formed of the back
electrode layer, back photoelectric conversion unit, silicon composite layer, front
photoelectric conversion unit and transparent electrode layer successively stacked on a
non-transparent substrate, in which case also the current generated in the front
photoelectric conversion unit can be increased to improve the conversion efficiency. In
this case, however, each of the back photoelectric conversion unit and the front
photoelectric conversion unit preferably has layers of n-i-p-types stacked in this order
from the substrate side.
Hereinafter, comparative example of the conventional art and examples of the
present invention will be described. Throughout the drawings, the same reference
characters denote the same or corresponding portions, and descriptions thereof will not
be repeated.
Firstly, comparative examples 1 and 2 according to the conventional art and
examples 1-4 according to the present invention are described for a two-unit stacked-
layer type photoelectric conversion device. Table 1 shows properties of the stacked-
layer type photoelectric conversion devices of Comparative Examples 1, 2 and Examples
1-4. Each photoelectric conversion device has an effective area of 1 cm2, and its
output properties were measured at 25°C with irradiation of AM 1.5 light at 100
mW/cm2. In Table 1, the open-circuit voltage (Voc), shorted-circuit current density
(Jsc), fill factor (FF) and conversion efficiency (Eff) are shown for each photoelectric
conversion device. Also shown in Table 1 are the output current of the front
photoelectric conversion unit, the output current of the back photoelectric conversion
unit, and the total output current obtained by spectral sensitivity measurement, which
are shown as relative values normalized with the values of Comparative Example 1.
(Comparative Example 1)
A stacked-layer type photoelectric conversion device as shown in Fig. 15 was
fabricated as Comparative Example 1. Firstly, on a 1.1 mm thick glass substrate 1 of
127 mm square, a SnCh film having surface unevenness with small pyramids and an
average thickness of 800 nm was formed as a transparent electrode layer 2 by thermal
CVD. The obtained transparent electrode layer 2 had a sheet resistivity of about 9 O/ .
Its Haze ratio measured with a C light source was 12%, and the average level difference
d of the surface unevenness was about 100 nm A front photoelectric conversion unit 3
including a 15 nm thick p-type amorphous silicon carbide layer 31, a 0.3 u.m thick i-type
amorphous silicon layer 32, and a 30 nm thick n-type microcrystalline silicon layer 33
was formed on transparent electrode layer 2 by plasma CVD, followed by formation of a
back photoelectric conversion unit 5 including a 15 nm thick p-type microcrystalline
silicon layer 51, a 2.5 u.m thick i-type crystalline silicon layer 52, and a 15 nm thick n-
type microcrystalline silicon layer Thereafter, a 90 nm thick A1-doped ZnO film and a
300 nm thick Ag film were successively formed as a back electrode layer 6 by sputtering.
(Comparative Example 2)
A stacked-layer type photoelectric conversion device as shown in Fig. 16 was
fabricated as Comparative Example 2. The device of Comparative Example 2 of Fig
16 is identical to the device of Comparative Example 1 except that n-type
microcrystalline silicon layer 33 in Fig. 15 is replaced with a 30 nm thick n-type
amorphous silicon oxide layer 39. The device of Comparative Example 2 has a
structure similar to that of Prior Art Example 2 described above, except that its back
photoelectric conversion unit is a crystalline photoelectric conversion unit. N-type
amorphous silicon oxide layer 39 was formed under conditions of a gas flow rate ratio
of SiH4/CO2/PH3/H2=5/2.5/0.1/100 sccm, a power supply frequency of 13.56 MHz, a
power density of 20 mW/cm2, a pressure of 100 Pa, and a substrate temperature of
200°C. The obtained amorphous silicon oxide layer 39 had an oxygen concentration of
18 atomic %, a refractive index of 3.0 regarding light of 600 nm wavelength, and a dark
conductivity of 1.2 x 10-6 S/cm. Then, its intensity ratio of the TO mode peak of
crystalline silicon component to the TO mode peak of amorphous silicon component
measured by Raman scattering was 0 indicating that there is no crystal phase.
As seen from Table 1, Comparative Example 2 has properties similar to those of
Comparative Example 1, exhibiting no increase of Jsc and no significant change in the
spectral sensitivity current of the front photoelectric conversion unit. It thus can be
said that n-type amorphous silicon oxide layer 39 having the oxygen concentration of 18
atomic % does not have the effect of reflecting light to the front photoelectric
conversion unit 3 side. In other words, it can be said that it is hardly possible to obtain
the reflection effect with this n-type amorphous silicon oxide layer 39, since its refractive
index is as high as 3.0, only slightly different from those of the amorphous silicon layer
and the crystalline silicon layer.
(Example 1)
A stacked-layer type photoelectric conversion device as shown in Fig. 10 was
fabricated as Example 1. The device of Example 1 differs from the device of
Comparative Example 1 of Fig. 15 only in that a 30 nm thick n-type silicon composite
layer 4 is formed between front and back photoelectric conversion units 3 and 5. This
n-type silicon composite layer 4 may be considered as a part of the n-type layer included
in front photoelectric conversion unit 3.
N-type silicon composite layer 4 was formed under conditions of a gas flow rate
ratio of SiH4/CO2/PH3/H2=5/10/0.1/1000 sccm, a power supply frequency of 13.56
MHz, a power density of 100 mW/cm2, a pressure of 100 Pa, and a substrate
temperature of 200°C. The obtained n-type silicon composition layer 4 had an oxygen
concentration of 42 atomic %, an optical gap of 2.37 eV, Exps of 3.5 eV measured by
XPS, and a refractive index of 2.0 regarding light of 600 nm wavelength. Then, its
intensity ratio of the TO mode peak of crystalline silicon component to the TO mode
peak of amorphous silicon component measured by Raman scattering was 2.0, and its
dark conductivity was 5 x 10-6 S/cm.
As seen from Table 1, although FF is slightly decreased in Example 1 compared
to Comparative Example 1, Jsc is increased by more than 1 mA/cm2 and Eff is improved.
The spectral sensitivity current of the front photoelectric conversion unit is increased by
9%, indicating that silicon composite layer 4 effectively reflects the incident light to the
front photoelectric conversion unit side. Further, the spectral sensitivity current of the
back photoelectric conversion unit is also increased by 6%. It is considered that the
light scattering caused by silicon composite layer 4 also results in the extended optical
path length within the back photoelectric conversion unit.
(Example 2)
A stacked-layer type photoelectric conversion device as shown in Fig. 17 was
fabricated as Example 2. The device of Example 2 differs from the device of Example
1 of Fig. 10 only in that a 30 ran thick n-type silicon composite layer 4 is employed as
the n-type layer of front photoelectric conversion unit 3 to serve both as the
intermediate reflective layer and the n-type layer.
As seen from Table 1, in Example 2 compared to Example 1, Jsc is further
increased and Eff is further improved, and the spectral sensitivity currents of both the
front and back photoelectric conversion units are increased, presumably for the
following reason. Since silicon composite layer 4 also serves as the n-type layer of
front photoelectric conversion unit 3, light reflected to the front photoelectric
conversion unit 3 side and light transmitted to the back photoelectric conversion unit 5
side are both unnecessary to transmit the 30 nm thick n-type microcrystalline silicon
layer, resulting in decrease of the absorption loss. However, it is noted that FF is
decreased in Example 2 compared to Comparative Example 1 and Example 1, and it is
considered that the contact resistance is increased at the interface between i-type
amorphous silicon layer 32 and silicon composite layer 4.
A graph of Fig. 33 shows distribution of phosphorus concentration and oxygen
concentration measured by SIMS in the thickness direction of the photoelectric
conversion device having the stacked-layer structure shown in Fig. 17. A horizontal
axis in this graph represents the distance (mm) in the thickness direction, a left vertical
axis represents the atomic concentration (cm-3), and a right vertical axis represents the
number of counts per second (c/s) of SIMS. Back surface electrode 6 was removed in
advance by wet etching with hydrochloric acid, and the phosphorus concentration and
the oxygen concentration were measured in the depth (thickness) direction by SIMS,
while conducting ion sputtering toward the substrate 1. As shown in the graph of Fig.
33, phosphorus and oxygen have clear peaks at the same location in the depth direction,
which means that there exists an n-type silicon composite layer.
A graph of Fig. 34 shows distribution of oxygen concentration measured by XPS
in the thickness direction of the photoelectric conversion device having the stacked-layer
structure of Fig. 17. A horizontal axis in this graph represents the sputtering time
(min) in the thickness direction, and a vertical axis represents the atomic %. The
silicon composite layer included in the stacked-layer type photoelectric conversion
device measured by XPS had a thickness of 50 nm and refractive index of 2.18. Back
surface electrode 6 was removed in advance by wet etching with hydrochloric acid, and
the oxygen concentration in the depth (thickness) direction was measured by XPS, while
conducting ion sputtering toward the substrate 1. As shown in the graph of Fig. 34,
the oxygen concentration has a clear peak, proving the existence of the silicon
composite layer. The peak value of the oxygen concentration in this graph is about 15
atomic %, and its half width is about 110 nm. Taking the actual thickness of 50 nm of
the silicon composite layer into consideration, it is found that the oxygen concentration
of the silicon composite layer measured by XPS is 33 atomic %, because (15 atomic %)
x (110nm/50nm) = 33.
(Example 3)
A stacked-layer type photoelectric conversion device as shown in Fig. 18 was
fabricated as Example 3. The device of Example 3 differs from the device of Example
2 of Fig. 17 only in that the n-type layer of front photoelectric conversion unit 3 is
formed by stacking a 30 nm thick n-type silicon composite layer 34 as a first n-type layer
and a 5 nm thick n-type microcrystalline silicon layer 35 as a second n-type layer.
Needless to say, the n-type silicon composite layer 34 of Example 3 is formed under
plasma CVD conditions similar to those for n-type silicon composite layer 4 in Examples
1 and 2.
As seen from Table 1, although Jsc in Example 3 is slightly decreased compared
to Example 2, FF and Eff are improved. Further, although the spectral sensitivity
current of the front photoelectric conversion unit in Example 3 is slightly lower than that
of Example 2, it is higher than those of Comparative Example 1 and Example 1. This
is presumably because insertion of n-type microcrystalline silicon layer 35 between n-
type silicon composite layer 34 and p-type microcrystalline silicon layer 51 has
decreased the contact resistance at the n/p interface, resulting in improvement of FF.
Incidentally, the decrease of Jsc in Example 3 compared to Example 2 is not large,
because n-type microcrystalline silicon layer 35 inserted in Example 3 is as thin as 5 nm.
(Example 4)
A stacked-layer type photoelectric conversion device as shown in Fig. 19 was
fabricated as Example 4. The device of Example 4 differs from the device of Example
2 of Fig. 17 only in that the n-type layer of front photoelectric conversion unit 3 is
formed by stacking a 10 nm thick n-type microcrystalline silicon layer 36 as a first n-type
layer, a 60 nm thick n-type silicon composite layer 37 as a second n-type layer and a 5
nm thick n-type microcrystalline silicon layer 38 as a third n-type layer.
As seen from Table 1, although Jsc in Example 4 is slightly decreased compared
to Example 3, FF and Eff are improved. This is presumably because n-type
microcrystalline silicon layer 36 inserted between i-type amorphous silicon layer 32 and
n-type silicon composite layer 37 has decreased the contact resistance at the i/n interface,
and thus FF is further improved compared to Example 3. In Example 4, n-type
microcrystalline silicon layer 36 and n-type microcrystalline silicon layer 38, excluding n-
type silicon composite layer 37, have a total thickness of only 15 nm, which is thinner
than 30 nm thick n-type microcrystalline silicon layer 33 of Example 1. This decreases
the absorption loss, and Jsc is increased even compared to Example 1.
(Example 5)
As Example 5, shown in a graph of Fig. 20 are relative values of the spectral
sensitivity currents in the case of changing the refractive index of silicon composite layer
4 in the stacked-layer type photoelectric conversion device having the structure shown
in Fig. 17. Silicon composite layer 4 in Example 5 was formed in a similar manner as in
Examples 1 and 2, except that the CO2/SiH4 ratio was changed within a range of 1 to 15
upon plasma CVD. In this graph, a horizontal axis represents the refractive index of
silicon composite layer 4 regarding light of 600 nm wavelength, and a vertical axis
represents the relative value of the spectral sensitivity current in the device of Example 5,
compared to that of Comparative Example 1 of Fig. 15 provided with no silicon
composite layer
As shown in Fig. 20, the spectral sensitivity current of front photoelectric
conversion unit 3 increases according to decrease of the refractive index of silicon
composite layer 4, and decreases when the refractive index becomes smaller than about
1.8, probably for the following reasons. With decrease of the refractive index of silicon
composite layer 4, light reflected to the front photoelectric conversion unit 3 increases,
leading to increase of the spectral sensitivity current. When the refractive index
becomes smaller than about 1.8, however, the dark conductivity of silicon composite
layer 4 decreases, and then the resistivity of silicon composite layer 4 as well as the
contact resistance at the interface considerably increases, resulting in decrease of the
current.
The spectral sensitivity current of back photoelectric conversion unit 5 also
increases according to decrease of the refractive index of silicon composite layer 4, and
decreases as the refractive index becomes smaller than about 2 Since the
transmittance of silicon composite layer 4 increases with decrease of the refractive index,
light reaching the back photoelectric conversion unit 5 increases, thereby increasing the
current. When the refractive index of silicon composite layer 4 becomes smaller than
about 2, light reflected to front photoelectric conversion unit 3 increases, and
correspondingly, light reaching back photoelectric conversion unit 5 decreases
considerably, which results in decrease of the current of back unit 5.
As seen from Fig. 20, the total spectral sensitivity current of front and back
photoelectric conversion units 3 and 5 has its maximal value dependent on the refractive
index. When the refractive index of silicon composite layer 4 is in a range from 1.7 to
2.5, the total spectral sensitivity current of the stacked-layer type photoelectric
conversion device of Example 5 increases as compared to Comparative Example 1. In
order to increase the total spectral sensitivity current of the stacked-layer type
photoelectric conversion device of Example 5 by more than 10% compared to
Comparative Example 1, the refractive index of silicon composite layer 4 should be in a
range from 1.8 to 2.1.
Hereinafter, explanation will be given for cases where the silicon composite layer
is used also as a part of one conductivity type layer (p-type layer) in the back
photoelectric conversion unit in the two-unit stacked-layer type photoelectric conversion
device, in conjunction with Comparative Example 1 of the prior art and Examples 6-9 of
the present invention. Relative values of properties of these stacked-layer type
photoelectric conversion devices are summarized and listed in Table 2.
(Example 6)
A stacked-layer type photoelectric conversion device as shown in Fig. 21 was
fabricated in Example 6. The device of Example 6 differs from the device of
Comparative Example shown in Fig. 15 only in that the p-type layer of back
photoelectric conversion unit 5 is replaced with a 30 nm thick p-type silicon composite
layer 51a.
P-type silicon composite layer 51a was formed under plasma CVD conditions of
a gas flow rate ratio of SiH4/CO2/B2H6/H2= 1/2/0.0025/500 sccm, a power supply
frequency of 13.56 MHz, a power density of 100 mW/cm2, a pressure of 800 Pa, and a
substrate temperature of 200°C. The obtained p-type silicon composition layer 51a
had an oxygen concentration of 29 atomic % and a refractive index of 2.45 regarding
light of 600 nm wavelength. Then, its intensity ratio of the TO mode peak of
crystalline silicon component to the TO mode peak of amorphous silicon component
measured by Raman scattering was 1.2 and its dark conductivity was 2 x 10-5 S/cm.
As shown in Table 2, Jsc and Eff in Example 6 also are improved as compared to
Comparative Example 1. It is considered that p-type silicon composite layer 51 a
functions as the intermediate reflective layer, effectively reflecting the incident light to
the front photoelectric conversion unit 3 side, and that since the p-type layer of back
photoelectric conversion unit 5 has been replaced with highly transparent p-type silicon
composite layer 51a, the light absorption loss is also decreased. However, it is noted
that FF in Example 6 is decreased as compared to Comparative Example 1, presumably
because the contact resistance is increased at the interface of n-type microcrystalline
silicon layer 33 / p-type silicon composite layer 51a or at the interface of p-type silicon
composite layer 51a / i-type crystalline silicon layer 52.
(Example 7)
A stacked-layer type photoelectric conversion device as shown in Fig. 22 was
fabricated in Example 7. The device of Example 7 differs from that of Comparative
Example 1 shown in Fig. 15 only in that the n-type layer of front photoelectric
conversion unit 3 is replaced with a 30 nm thick n-type silicon composite layer 33a, and
the p-type layer of back photoelectric conversion unit 5 is replaced with a 30 nm thick
p-type silicon composite layer 51a. N-type silicon composite layer 33a is formed under
plasma CVD conditions similar to those for n-type silicon composite layer 4 used in
Examples 1 and 2.
It is seen from Table 2 that in Example 7 Jsc is considerably increased and Eff is
improved as compared to Comparative Example 1, and the Jsc is increased and the Eff is
improved even as compared to Example 6, presumably for the following reasons. N-
type silicon composite layer 33a and p-type silicon composite layer 51a have a total
thickness of 60 nm, and thus their functions as the intermediate reflective layers are
further improved. Furthermore, since the n-type layer of front photoelectric
conversion unit 3 and the p-type layer of back photoelectric conversion unit 5 are
replaced with highly transparent silicon composite layers, the light absorption loss is
considerably reduced. However, it is noted that FF in Example 7 is decreased
compared to Comparative Example 1, presumably because the contact resistance at the
interface of i-type amorphous silicon layer 32 / silicon composite layer 33a or at the
interface of p-type silicon composite layer 51a / i-type crystalline silicon layer 52 is
increased.
(Example 8)
A stacked-layer type photoelectric conversion device as shown in Fig. 23 was
fabricated in Example 8. The device of Example 8 differs from that of Example 7
shown in Fig. 22 only in that a 10 nm thick n-type microcrystalline silicon layer 33b and
a 30 nm thick n-type silicon composite layer 33a are stacked in this order as n-type layer
33 of front photoelectric conversion unit 3, and a 30 nm thick p-type silicon composite
layer 51a and a 5 nm thick p-type microcrystalline silicon layer 51b are stacked in this
order as p-type layer 51 of back photoelectric conversion unit 5.
As shown in Table 2, FF in Example 8 is improved compared to Example 7,
presumably because the contact resistance at the interface of i-type amorphous silicon
layer 32 / n-type silicon composite layer 33a is reduced by insertion of thin n-type
microcrystalline silicon layer 33b, and the contact resistance at the interface of p-type
silicon composite layer 51a/ i-type crystalline silicon layer 52 is reduced by insertion of
thin p-type microcrystalline silicon layer 51b On the other hand, Jsc in Example 8 is
slightly decreased compared to Example 7, due to the light absorption loss at n-type
microcrystalline silicon layer 33b and p-type microcrystalline silicon layer 51b.
(Example 9)
A stacked-layer type photoelectric conversion device as shown in Fig. 24 was
fabricated in Example 9. The device of Example 9 differs from that of Example 8
shown in Fig. 23 only in that a 30 nm thick n-type silicon composite layer 33a and a 10
nm thick n-type microcrystalline silicon layer 33c are stacked in this order as n-type
layer 33 of front photoelectric conversion unit 3, and a 5 nm thick p-type
microcrystalline silicon layer 51c and a 30 nm thick p-type silicon composite layer 51a
are stacked in this order as p-type layer 51 of back photoelectric conversion unit 5.
As shown in Table 2, FF in Example 9 is further improved compared to Example
8, presumably because at the junction interface between front and back photoelectric
conversion units 3 and 5 where influence of the contact resistance would be most
considerable, a silicon composite layer is not disposed but the junction of n-type
microcrystalline silicon layer 33c / p-type microcrystalline silicon layer 51c is formed
similarly as in Comparative Example 1 of Fig. 15.
(Example 10)
A stacked-layer type photoelectric conversion device as shown in Fig. 17 was
fabricated in Example 10, similarly as in Example 2. The device of Example 10 differs
from that of Example 2 only in that the thickness of n-type silicon composite layer 4 of
front photoelectric conversion unit 3 is changed to 50 nm, and that the plasma CVD
conditions are changed. That is, the remaining conditions of forming the device of
Example 10 are similar to those of Comparative Example 1 and Example 2.
More specifically, n-type silicon composite layer 4 of Example 10 was formed
under the plasma CVD conditions of the gas flow rate ratio of
SiH4/CO2/PH3/H2=l/3/0.02/100, the power supply frequency of 13.56 MHz, the power
density of 100 mW/cm2, the pressure of 100 Pa. and the substrate temperature of 200°C.
In Example 10, unlike in the case of Comparative Example 2, the dilution of SiH4 with
H2 is increased to 100 times, which is the condition where the crystal phase readily
appears provided that CO2 is not contained. The n-type silicon composition layer 4
obtained under these conditions had an oxygen concentration of 44 atomic %, an optical
gap of 2.42 eV, Exps measured by XPS of 3.6 eV, and a dark conductivity of 1.2 x
10-12 S/cm. Then, its refractive index regarding light of 600 nm wavelength was as low
as 1.95, which is approximately comparable to that of the TCO layer. In this silicon
composite layer, a TO mode peak of amorphous silicon was detected by Raman
scattering measurement, proving that amorphous silicon is included locally, whereas a
TO mode peak of crystalline silicon was not detected. That is, the intensity ratio
(Ic/Ia) of the TO mode peak derived from the crystalline silicon phase to the TO mode
peak derived from the amorphous alloy phase was 0, and no silicon crystal phase was
detected. Even with X-ray diffraction and transmission electron microscopy, the
crystal phase was not detected in the silicon composite layer of Example 10.
Table 3 shows properties of the stacked-layer type photoelectric conversion
devices of Comparative Example 1 and Example 10. As seen from Table 3, even in
Example 10 in which the silicon composite layer does not include silicon crystal phase
parts, Jsc is increased and Eff is improved compared to Comparative Example 1. The
spectral sensitivity currents in both first and second photoelectric conversion units 3 and
5 were also increased in Example 10 compared to Comparative Example 1. This
shows that silicon composite layer 4 functions as the intermediate reflective layer,
effectively reflecting the incident light to the first photoelectric conversion unit 3 side.
Further, it is considered that the light scattering due to silicon composite layer 4
elongates the optical path length within second photoelectric conversion unit 5.
Further, the light absorption loss is reduced presumably because the n-type layer of first
photoelectric conversion unit 3 has been replaced with highly transparent silicon
composite layer 4. However, it is noted that FF in Example 10 is decreased compared
to Comparative Example 1, presumably because the contact resistance is increased at the
interface of i-type amorphous silicon layer 32 / silicon composite layer 4 or at the
interface of silicon composite layer 4 / p-type microcrystalline silicon layer 51.
The graph of Fig. 35 shows Raman scattering spectra measured while changing
the thickness of the silicon composite layer. In this graph, the silicon composite layers
of Example 11, Reference Example 1 and Reference Example 2 were formed under the
same plasma CVD conditions as those for the silicon composite layer of Example 10.
In Example 11, back electrode 6 of the stacked-layer type photoelectric conversion
device of Example 10 was removed, and the Raman scattering spectrum was measured
while exposing the surface of silicon composite layer 4 by ion sputtering. At this time,
silicon composite layer 4 had a thickness of 50 nm. In Reference Examples 1 and 2,
300 mm thick and 1 mm thick silicon composite layers were deposited on glass
substrates, respectively.
In each of Example 11 and Reference Example 1 having the silicon composite
layers of 50 nm and 300 nm, respectively, only a wide TO mode peak of amorphous
silicon component was observed in the vicinity of 480 cm-1, as shown in Fig. 35,
indicating that each composite layer does not include the crystal phase. In Reference
Example 2 having the 1 mm thick silicon composite layer, on the other hand, a shoulder
was observed in the vicinity of 520 cm"1 in the graph of Fig. 35, and the TO mode peak
by crystalline silicon component was detected
As shown in Fig. 35, the crystal phase is detected when the thickness of the
silicon composite layer is increased, and thus it is considered that the silicon composite
layer of Example 11 is in a state immediately before generation of the crystal phase.
That is, it can be said that under the same plasma CVD conditions, the silicon composite
layer in the stacked-layer type photoelectric conversion device is preferably deposited to
have a thickness of more than 1 mm to thereby include silicon crystal phase parts. It is
noted that in the silicon composite layers of Example 11 and Reference Example 1, the
crystal phase was not detected even with X-ray diffraction or transmission electron
microscopy.
In the case of depositing the silicon composite layer changing only the hydrogen
dilution ratio compared to Example 11, the crystal phase was detected only in the silicon
composite layer obtained with the hydrogen dilution ratio of more than 120. Further,
in the photoelectric conversion device having the stacked-layer structure of Example 1,
when only the hydrogen dilution ratio was changed in a rage of 50 to 120 at the time of
deposition of silicon composite layer 4, Jsc was increased by more than 5% compared to
Comparative Example 1 with the hydrogen dilution ratio of more than 70. Thus, when
the silicon composite layer is deposited for use as the intermediate reflective layer in the
stacked-layer type photoelectric conversion device, it can be said that it is one of
preferable conditions to set the hydrogen dilution ratio to more than 60% in which case
the crystal phase can be detected. Incidentally, when a silicon composite layer was
deposited with film deposition conditions other than the hydrogen dilution ratio being
shifted from the plasma CVD conditions suitable for generation of the crystal phase to
those suitable for generation of the amorphous phase, the formed silicon composite layer
was still effective for improving properties of the stacked-layer type photoelectric
conversion device.
(Example 12)
Fig. 25 schematically shows a three-unit stacked-layer type photoelectric
conversion device of Example 12 This photoelectric conversion device was fabricated
in a similar manner as in Example 1 of Fig. 10 up to completion of a second-level
photoelectric conversion unit 5a, except that thicknesses of the i-type layers were
changed. I-type amorphous silicon layer 32a of a first-level photoelectric conversion
unit 3a in Fig. 25 has a thickness of 100 nm, and i-type crystalline silicon layer 52a of
second-level photoelectric conversion unit 5a has a thickness of 1.2 mm. A 30 nm
thick second silicon composite layer 7 was formed on second-level photoelectric
conversion unit 5a, and then a 15 nm thick p-type microcrystalline silicon layer 81, a 2.0
urn thick i-type crystalline silicon layer 82 and a 15 nm thick n-type microcrystalline
silicon layer 83 were formed as a third-level photoelectric conversion unit 8
Thereafter, a 90 nm thick Al-doped ZnO layer and a 300 nm thick Ag layer were
successively formed as back electrode 6 by sputtering. First silicon composite layer 4
and second silicon composite layer 7 were formed under plasma CVD conditions similar
to those for silicon composite layer 4 of Example 1.
For the three-unit stacked-layer type photoelectric conversion device of Example
12 thus obtained, output properties were measured under conditions similar to those for
Table 1. As a result, Voc was 1.905 V, Jsc was 10.07 mA/cm2, FF was 0.745, and Eff
was 14.29%.
(Comparative Example 3)
A three-unit stacked-layer type photoelectric conversion device as shown in Fig.
26 was fabricated as Comparative Example 3. The device of Comparative Example 3
differs from that of Example 12 shown in Fig. 25 only in that it does not include first and
second silicon composite layers 4 and 7. As a result of measuring output properties of
this Comparative Example 3, Voc was 1.910 V, Jsc was 9.50 mA/cm2, FF was 0.749,
and Eff was 13.59%.
Comparing Example 12 and Comparative Example 3, it is found that in the
three-unit staked-layer type photoelectric conversion device as well, Jsc is increased by
the reflection effect of the silicon composite layer, and Eff is improved.
Hereinafter, in connection with the three-unit stacked-layer type photoelectric
conversion device, explanation will further given for Comparative Example 4 of the
prior art and Examples 13 and 14 of the present invention. Properties of these stacked-
layer type photoelectric conversion devices are summarized and listed with relative
values in Table 4.
(Comparative Example 4)
A three-unit stacked-layer type photoelectric conversion device of Comparative
Example 4 differs from that of Comparative Example 3 shown in Fig. 26 only in that the
thickness of i-type crystalline silicon layer 82 in third-level photoelectric conversion unit
8 is changed from 2.0 mm to 2.5 mm.
(Example 13)
A three-unit stacked-layer type photoelectric conversion device as shown in Fig
27 was fabricated in Example 13. The device of Fig. 27 differs from that of
Comparative Example 4 only in that the n-type layer of second photoelectric conversion
unit 5a is replaced with an n-type silicon composite layer 53a, and the p-type layer of
third photoelectric conversion unit 8 is replaced with a p-type silicon composite layer
81a. N-type silicon composite layer 53a was deposited under conditions similar to
those for the n-type silicon composite layer used in Example 1, and p-type silicon
composite layer 81a was deposited under conditions similar to those for the p-type
silicon composite layer used in Example 6.
As shown in Table 4, in the three-unit stacked-layer type photoelectric
conversion device of Example 13 also, Jsc is increased and Eff is improved compared to
Comparative Example 4, by virtue of the reflection effect of the intermediate silicon
composite layers.
(Example 14)
A three-unit stacked-layer type photoelectric conversion device as shown in Fig
28 was fabricated in Example 14,. The device of Example 14 differs from that of
Comparative Example 4 in that a 5 nm thick n-type microcrystalline silicon layer 33b, a
50 nm thick n-type silicon composite layer 33 a, and 5 nm thick n-type microcrystalline
silicon layer 33c are stacked in this order as n-type layer 33 of a first-level photoelectric
conversion unit 3a, and a 5 nm thick n-type microcrystalline silicon layer 53b, a 50 nm
thick n-type silicon composite layer 53a, and a 5 nm thick n-type microcrystalline silicon
layer 53c are stacked in this order as n-type layer 53 of a second-level photoelectric
conversion unit 5a. N-type silicon composite layer 33a and n-type silicon composite
layer 53a were deposited under conditions similar to those for n-type silicon composite
layer 4 of Example 10.
As shown in Table 4, in the three-unit stacked-layer type photoelectric
conversion device of Example 14 also, Jsc is increased and Eff is improved compared to
Comparative Example 4, by virtue of the reflection effect of the intermediate silicon
composite layers.
(Example 15)
Fig. 29 shows an integrated type thin-film photoelectric conversion module
according to Example 15 of the present invention. The module of Fig. 29 differs from
the conventional module of Fig. 31 only in that ZnO intermediate reflective layer 105 is
replaced with a silicon composite layer 107 Thicknesses and forming methods of the
layers included in the module of Example 15 are similar to those of Example 1. The
module of Fig. 29 has an area of 910 mm x 455 mm, and 100 photoelectric conversion
cells partitioned by patterning are connected in series with each other.
(Comparative Examples 5-7)
An integrated type thin-film photoelectric conversion module including no
intermediate reflective layer as shown in Fig 30 was fabricated as Comparative Example
5. A module as shown in Fig. 31, including a 30 nm thick ZnO intermediate reflective
layer 105 formed by sputtering, was fabricated as Comparative Example 6. A module
as shown in Fig. 32, including a ZnO intermediate reflective layer 105 and a third
isolation groove 124, was fabricated as Comparative Example 7.
Table 5 shows output properties of the integrated type thin-film photoelectric
conversion modules of Example 15 and Comparative Examples 5-7, measured under
conditions similar to those for Table 1.

As seen from Table 5, in Comparative Example 6 including the ZnO intermediate
reflective layer, Voc and FF are noticeably lowered, and the maximum power (Pmax)
and the conversion efficiency (Eff) are considerably decreased, as compared to
Comparative Example 5 including no intermediate reflective layer. This is because
leakage current flows through the current path of ZnO intermediate reflective layer 105,
connection groove 123 and back electrode layer 106. In Comparative Example 7
including third isolation groove 124, on the other hand, such leakage current is
suppressed, and thus the shorted-circuit current (Jsc) increases and Pmax is improved by
about 3 W, as compared to Comparative Example 5.
In Example 15 including silicon composite layer 107, Jsc further increases
compared to Comparative Example 7, and Pmax is improved by about 10 W compared
to Comparative Example 5, presumably because the area loss due to third isolation
groove 124 is eliminated and then Jsc is improved. Further, in Example 15, front
photoelectric conversion unit layer 104a, silicon composite layer 107 and back
photoelectric conversion unit layer 104b can be formed continuously by plasma CVD.
This prevents atmospheric contamination of the interface between silicon composite
layer 107 and back photoelectric conversion unit 104b, leading to improvement of FF
Further, in Example 15, third isolation groove 124 is not needed. This
decreases the number of times of patterning, and thus the production costs and time can
be reduced compared to Comparative Example 7. Silicon composite layer 107 can be
formed by only adding a gas line for CO2 in the plasma CVD apparatus for forming the
photoelectric conversion unit. Other film formation equipment such as a sputtering
apparatus, otherwise required for forming the ZnO intermediate layer, becomes
unnecessary, so that the production costs can be reduced considerably. Still further,
the number of times of process steps of introducing the substrate into the vacuum
plasma CVD apparatus, heating the substrate therein, and taking the substrate out from
the plasma CVD apparatus are decreased by 1 each in Example 15 as compared to
Comparative Example 7, which further reduces the production costs and time.
(Example 16)
Two-unit stacked-layer type photoelectric devices each having such a stacked-
layer structure as shown in Fig. 17 were fabricated in Example 16, similarly as in
Example 2. In Example 16, however, photoelectric conversion devices as Samples 1A
and IB were fabricated through processes slightly different from each other.
For Sample 1 A, firstly, a transparent electrode layer including SnO2 as its main
component was formed on transparent glass substrate 1. The substrate was then
introduced into a first plasma CVD apparatus, and its temperature was raised.
Thereafter, p-type amorphous silicon carbide layer 31, i-type amorphous silicon
photoelectric conversion layer 32, and a first part of n-type silicon composite layer 4 of
amorphous silicon photoelectric conversion unit 3 were formed to thicknesses of 15 nm,
300 nm and 40 nm, respectively.
The first part of n-type silicon composite layer 4 was formed under conditions of
a gas flow rate ratio of SiH4/CO2/PH3/H2=l/3/0 025/200, a power supply frequency of
13.56 MHz, a power density of 120 mW/cm2, and a substrate temperature of 180°C.
The first part of n-type silicon composite layer 4 thus formed had an oxygen
concentration of 42 atomic %, and a refractive index of 2.0 regarding light of 600 nm
wavelength.
After deposition of the first part of n-type silicon composite layer 4, followed by
evacuation of the film deposition chamber, the substrate was immediately transferred to
an unload chamber of the first plasma CVD apparatus. Then, after chamber was
quickly filled with nitrogen gas, the substrate was taken out to the ambient air.
After the substrate was left in (or exposed to) the ambient air for about 40 hours,
it was introduced into a second plasma CVD apparatus and then its temperature was
raised. A second part of n-type silicon composite layer 4 was then formed to a
thickness of 10 nm. The second part of n-type silicon composite layer 4 had a
refractive index and an oxygen concentration similar to those of the first part of silicon
composite layer 4 formed in the first plasma CVD apparatus.
Subsequently, p-type microcrystalline silicon layer 51, non-doped i-type
crystalline silicon photoelectric conversion layer 52, and n-type microcrystalline silicon
layer 53 were formed in the same second plasma CVD apparatus to thicknesses of 15
nm, 2.5 mm and 15 nm, respectively. Thereafter, a 90 nm thick Al-doped ZnO layer
and a 300 nm thick Ag layer were successively deposited as back electrode 6 by
sputtering.
The stacked-layer type photoelectric conversion device of Sample 1B differs
from the device of Sample 1A only in that amorphous silicon photoelectric conversion
unit 3 through crystalline silicon photoelectric conversion unit 5 were formed
continuously in the same plasma CVD apparatus. That is, in the device of Sample 1B,
the entire thickness of n-type silicon composite layer 4 was deposited continuously in
the same plasma CVD apparatus, without taking the substrate out to the ambient air
during the deposition.
Photoelectric conversion properties of the photoelectric conversion devices of
Samples 1A and 1B were measured under conditions similar to those for Table 1. The
conversion efficiency of the photoelectric conversion device of Sample 1A was 1.01 in
relative value with respect to Sample 1B, which is high enough despite the exposure to
the ambient air
(Example 17)
Photoelectric conversion devices each having such a stacked-layer structure as
shown in Fig. 10 were fabricated in Example 17, similarly as in Example 1. In Example
17, photoelectric conversion devices as Samples 2A and 2B were formed through
processes slightly different from each other.
The fabricating process of the photoelectric conversion device of Sample 2 A is
similar to that of Sample 1 A. In the first plasma CVD apparatus, p-type amorphous
silicon carbide layer 31, i-type amorphous silicon photoelectric conversion layer 32, n-
type microcrystalline silicon layer 33, and a first part of n-type silicon composite layer 4
were formed to thicknesses of 15 nm, 300 nm. 10 nm, and 40 nm, respectively. That is,
Sample 2A differs from Sample 1A only in that formation of n-type silicon composite
layer 4 follows formation of n-type microcrystalline silicon layer 33 having high
conductivity and high crystallinity compared to the silicon composite layer. In Sample
2A, n-type microcrystalline silicon layer 33 formed in advance can reduce the contact
resistance at the interface between i-type amorphous silicon layer 32 and silicon
composite layer 4, and thus the photoelectric conversion properties are further improved
as compared to Sample 1A.
The stacked-layer type photoelectric conversion device of Sample 2B differs
from the device of Sample 2A only in that the entire thickness of n-type silicon
composite layer 4 is deposited continuously in a single plasma CVD apparatus, without
taking the substrate out to the ambient air during the deposition.
Photoelectric conversion properties of the photoelectric conversion devices of
Samples 2A and 2B were measured under conditions similar to those for Table 1. The
conversion efficiency of the photoelectric conversion device of Sample 2A compared to
Sample 2B was 1.01 in relative value, which is high enough despite the exposure to the
ambient air.
(Example 18)
In Example 18, additional samples were fabricated by variously changing the
refractive index regarding light of 600 nm wavelength and the oxygen concentration of
silicon composite layer 4 of Example 17. Table 6 shows photoelectric conversion
efficiencies of the samples exposed or unexposed to the ambient air after deposition of
up to the first part in thickness of silicon composite layer 4, in relative values with
respect to Sample 2B.
It is found from Table 6 that, as the oxygen concentration of silicon composite
layer 4 increases and the refractive index decreases, the effect as the intermediate
reflective layer increases and the photoelectric conversion efficiency is improved. In
comparison between the sample exposed to the ambient air and the other sample
unexposed to the ambient air both having the same oxygen concentration and the same
refractive index of silicon composite layer 4, Samples 2A, 3 A, 4A, 5A exposed to the
ambient air each having the oxygen concentration of more than 25% or the refractive
index of less than 2.5 exhibit their photoelectric conversion efficiencies equal to or
superior to those of Samples 2B, 3B, 4B, 5B unexposed to the ambient air. The silicon
composite layer originally contains oxygen and thus is less likely to suffer an adverse
effect of increase in resistance due to surface oxidation or the like. Therefore, higher
oxygen concentration and lower refractive index of the silicon composite layer can
reduce the influence of the exposure to the ambient air upon the photoelectric
conversion properties.
(Example 19)
Additional samples similar to Sample 2A of Example 17 were fabricated in
Example 19. Specifically, in Example 19, the ratio between the first and second parts
in the total thickness of silicon composite layer 4 was changed variously. Table 7
shows photoelectric conversion efficiencies of the samples exposed to the ambient air
after deposition of up to the first part in thickness of silicon composite layer 4, in
relative values with respect to Comparative Sample 02B unexposed to the ambient air.
As shown in Table 7, in Comparative Sample 03 A, the thickness of silicon
composite layer 4 before exposure to ambient air is 0, indicating that the sample is
exposed to the ambient air between formation of n-type microcrystalline silicon layer 33
and formation of n-type silicon composite layer 4 in Fig. 10. Under the influence of the
exposure to the ambient air of the surface of n-type microcrystalline silicon layer 33
originally containing almost no oxygen, the photoelectric conversion efficiency of
Comparative Sample 03 A is slightly decreased as compared to Comparative Sample 2B.
In Comparative Example 04A, deposition thickness of silicon composite layer 4 after
exposure to the ambient air is 0, indicating that the structure is exposed to the ambient
air between formation of n-type silicon composite layer 4 and p-type microcrystalline
silicon layer 51 in Fig. 10. Since the n-p junction interface or the tunnel junction
interface between the photoelectric conversion units is exposed to the ambient air, the
photoelectric conversion efficiency of Comparative Example 04 A is considerably
decreased as compared to Comparative Sample 02B.
In each of Samples 6A, 7A, 8A and 2A, on the other hand, parts in thickness of
the n-type silicon composite layer are formed immediately before and immediately after
the exposure to the ambient air, which makes it possible to obtain the photoelectric
conversion efficiency equal to or superior to that of Comparative Sample 02B
unexposed to the ambient air. Particularly, in each of Samples 8A and 2A, the first part
of 60% or more of the total thickness of n-type silicon composite layer 4 is formed
before the exposure to the ambient air, which makes it possible to obtain the
photoelectric conversion efficiency superior to that of Comparative Sample 02B. This
is presumably because excessive dopant atoms within silicon composite layer 4 and the
conductivity type layers are removed during evacuation of the film deposition chamber
and re-heating of the substrate which are carried out before film deposition in the second
plasma CVD apparatus into which the substrate has been introduced after the exposure
to the ambient air. Such an effect is more considerable when the first part of the silicon
composite layer formed before the exposure to the ambient air constitutes 60% or more
of the total thickness, resulting in improvement of the photoelectric conversion
efficiency.
(Example 20)
Photoelectric conversion devices each having such a stacked-layer structure as
shown in Fig. 10 were fabricated in Example 20, similarly as in Example 1. In Example
20, various samples of photoelectric conversion devices were fabricated by changing the
gas mixture ratios during plasma CVD of silicon composite layer 4.
In Sample 9, firstly, transparent electrode layer 2 including SnO2 as its main
component was formed on transparent glass substrate 1 Thereafter, the substrate was
introduced into a plasma CVD apparatus and its temperature was raised to deposit p-
type amorphous silicon carbide layer 31, i-type amorphous silicon photoelectric
conversion layer 32, and n-type microcrystalline silicon layer 33 of amorphous silicon
photoelectric conversion unit 3, and then n-type silicon composite layer 4 to thicknesses
of 15 nm, 300 nm, 10 nm and 50 nm, respectively
N-type silicon composite layer 4 was formed under plasma CVD conditions of a
gas flow rate ratio of SiH4CO2/PH3/H2 =1/5/0.025/360, a power supply frequency of
13.56 MHz, a power density of 150 mW/cm2, and a substrate temperature of 180°C.
An n-type silicon composite layer formed to a thickness of 300 nm on a glass substrate
under the same conditions had a refractive index of 1.83 regarding light of 600 nm
wavelength, a phosphorus concentration of 6.0 x 1020 cm-3 measured by secondary ion
mass spectrometry (SIMS), and an oxygen concentration of 52 atomic %. Then, it had
an intensity ratio (Ic/Ia) of 2.5 of the TO mode peak derived from the crystalline silicon
phase to the TO mode peak derived from the amorphous alloy phase in Raman
scattering spectrum.
Next, in the plasma CVD apparatus, p-type microcrystalline silicon layer 51,
non-doped i-type crystalline silicon photoelectric conversion layer 52, and n-type
microcrystalline silicon layer 53 of crystalline silicon photoelectric conversion unit 5
were formed to thicknesses of 15 nm, 2.5 mm and 15 nm, respectively. Thereafter, a
90 nm thick Al-doped ZnO layer and a 300 nm thick Ag layer were successively formed
as back electrode 5 by sputtering.
Samples 10-12 and Comparative Samples 05-07 differ from Sample 9 only in
that the mixing ratio of the doping source gas PH3 to the silicon source gas SiH4 in
plasma CVD is changed variously for n-type silicon composite layer 4 included therein.
Table 8 shows, for Samples 9-12 and Comparative Samples 05-07, the gas
mixture ratio for deposition of silicon composite layer 4, the refractive index regarding
light of 600 nm wavelength, the phosphorus concentration, the oxygen concentration,
and the peak intensity ratio (Ic/Ia) in Raman scattering spectrum, together with the
photoelectric conversion efficiency in relative value with respect to Sample 9.
As seen from Comparative Sample 05 in Table 8, when the silicon composite
layer has a low phosphorus concentration, the refractive index is low, while Ic/Ia is low,
causing a high resistivity of the silicon composite layer Then, the photoelectric
conversion efficiency is decreased due to increase of the series resistance in the stacked-
layer type photoelectric conversion device. On the other hand, it is considered that the
phosphorus atoms have the effect of promoting growth of the silicon crystal phase.
Thus, when the phosphorus concentration is increased as in Samples 9-12, conductivity
of the silicon composite layer is considerably improved with the synergic effect of
increase of Ic/Ia and increase of dopant concentration. If the mixture ratio of PH3 and
thus the phosphorus concentration are too high as in Comparative Samples 06-07,
however, although the conductivity is high because of the silicon crystal phase, the
excessive phosphorus atoms are undesirably introduced to the crystalline silicon
photoelectric conversion unit 5 side, resulting in degradation of the photoelectric
conversion properties. As such, it is found from the results of Samples 9-12 exhibiting
high photoelectric conversion efficiencies of 0.95 ore more in relative values, that the
preferable phosphorus concentration is in the range from 3 x 1020 cm"3 to 1.8 x 1021
cm"3, and the preferable PH3/SiH4 gas mixture ratio is in the range of 0.012 to 0.07.
Samples 13-15 and Comparative Samples 08-09 additionally fabricated differ
from Sample 9 only in that the mixing ratio of the oxygen source gas CO2 to the silicon
source gas SiH4 in plasma CVD is changed variously in a range from 2 to 7 for n-type
silicon composite layer 4 included therein.
Table 9 shows, for Samples 9, 13-15 and Comparative Samples 08-09, the gas
mixture ratio for deposition of silicon composite layer 4, the refractive index regarding
light of 600 nm wavelength, the phosphorus concentration, the oxygen concentration,
and the peak intensity ratio (Ic/Ia) in Raman scattering spectrum, together with the
photoelectric conversion efficiency in relative value with respect to Sample 9.
When the silicon composite layer has a low oxygen concentration and a high
refractive index as in Comparative Sample 08 in Table 9, its conductivity is high.
However, the reflection effect is weak and the absorption loss increases, so that it is not
possible to achieve a good conversion efficiency in the stacked-layer type photoelectric
conversion device. In Comparative Sample 09, the oxygen concentration in the silicon
composite layer is high and the refractive index is low, while Ic/Ia is low and thus the
proportion of the silicon crystal phase is extremely small, causing increase in resistivity
of the silicon composite layer. The photoelectric conversion efficiency is low due to
the increase in series resistance of the stacked-layer type photoelectric conversion device.
Accordingly, it is found from the results of Samples 9 and 13-15 exhibiting high
efficiencies of 0.95 or more in relative values, that the refractive index of the silicon
composite layer is preferably 2.0 or less, the oxygen concentration is preferably 40% or
more, and the peak intensity ratio Ic/Ia in Raman scattering spectrum is preferably 1.5 or
more
Industrial Applicability
As described above, according to the present invention, it is possible to improve
the photoelectric conversion efficiency and reduce the production costs of the stacked-
layer type thin-film photoelectric conversion device and the integrated type thin-film
photoelectric conversion module.

WE CLAIM:
1. A stacked-layer type photoelectric conversion device comprising a plurality
of photoelectric conversion units (3; 5) stacked on a substrate (1). each of which
includes a one conductivity-type layer (31; 51), a photoelectric conversion layer (32;
52) of substantially intrinsic semiconductor, and an opposite conductivity-type layer
(33; 53) in this order from a light incident side, wherein
at least one of said opposite conductivity-type layer (33) in a front photoelectric
conversion unit (3) arranged relatively closer to the light incident side and said one
conductivity-type layer (51) in a back photoelectric conversion unit (5) arranged
adjacent to said front photoelectric conversion unit (3) includes a silicon composite
layer (4) at least in a part thereof, and
said silicon composite layer (4) has a thickness of more than 20 ran and less
than 130 nm and an oxygen concentration of more than 25 atomic % and less than 60
atomic %, and includes silicon-rich phase parts dispersed in an amorphous alloy phase
matrix of silicon and oxygen.
2. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein said silicon-rich phase part includes a silicon crystal phase.
3. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein said silicon-rich phase part includes doped amorphous silicon.
4. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein refractive index of said silicon composite layer regarding light of 600 nm
wavelength is more than 1.7 and less than 2.5.
5. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein said substrate is transparent, and a reflection spectrum of light having passed
through the substrate and entered said stacked photoelectric conversion units has at
least one maximal value and at least one minimal value of reflectance in a wavelength
range of 500 nm to 800 nm, and a difference between said maximal value and said
minimal value is at least 1%.
6. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein dark conductivity of said silicon composite layer is more than 10-8 S/cm and
less than 10-1 S/cm.
7. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein in said silicon composite layer, an intensity ratio of a TO mode peak derived
from crystalline silicon phase parts to a TO mode peak derived from said amorphous
alloy phase, measured by Raman scattering, is more than 0.5 and less than 10.
8. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein an optical energy gap of said silicon composite layer is at least 2.2 eV.
9. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein in said silicon composite layer, an energy difference between upper most
energy of a photoelectron having suffered interband excitation loss of Ols and peak
energy of the Ols photoelectron, measured by X-ray photoelectron spectroscopy, is at
least 2.2 eV.
10. The stacked-layer type photoelectric conversion device as claimed in claim
1, wherein a dopant atom concentration in said silicon composite layer is in a range
from 3 x 1020 cm-3 to 1.8 x 1021 cm-3.
11. A method of forming the stacked-layer type photoelectric conversion
device as claimed in claim 1, wherein said substrate, having said silicon composite
layer deposited to a part of its total thickness in a plasma CVD reaction chamber, is
temporarily taken out to expose a surface of said silicon composite layer to the ambient
air, and then after said substrate is introduced again into a plasma CVD reaction
chamber, the remaining part of the total thickness of said silicon composite layer is
deposited.
12. The method as claimed in claim 11, wherein said substrate is taken out
from said plasma CVD reaction chamber to the ambient air after at least 60% of the
total thickness of said silicon composite layer is deposited.
13. A method for forming the stacked-layer type photoelectric conversion
device as claimed in claim 1, wherein a mixing ratio of doping source gas to silicon
source gas for deposition of said silicon composite layer in a plasma CVD reaction
chamber is in a range from 0.012 to 0.07.
14. An integrated type photoelectric conversion module, wherein
a first electrode layer (103), a plurality of photoelectric conversion unit layers
(104a; 104b) and a second electrode layer (106) successively stacked on a substrate
(102) are separated by a plurality of isolation grooves (121; 122) to form a plurality of
photoelectric conversion cells (110), and the cells are electrically connected in series
with each other via a plurality of connection grooves (123),
each of said photoelectric conversion cells has a plurality of stacked
photoelectric conversion units each including a one conductivity-type layer, a
photoelectric conversion layer of substantially intrinsic semiconductor and an opposite
conductivity-type layer in this order from a light-incident side,
at least one of said opposite conductivity-type layer in a front photoelectric
conversion unit (104a) arranged relatively closer to the light-incident side and said one
conductivity-type layer in a back photoelectric conversion unit (104b) arranged adjacent
to the front photoelectric conversion unit includes a silicon composite layer (107) at
least in a part thereof, and
said silicon composite layer (107) has a thickness of more than 20 nm and less
than 130 nm and an oxygen concentration of more than 25 atomic % and less than 60
atomic %, and includes silicon-rich phase parts dispersed in an amorphous alloy phase
matrix of silicon and oxygen.
15. The integrated type photoelectric conversion module as claimed in claim
14, wherein said first electrode layer (103) is separated into a plurality of regions
corresponding to said plurality of photoelectric conversion cells (110) by a plurality of
first isolation grooves (121), said plurality of photoelectric conversion unit layers
(104a; 104b) and said second electrode layer (106) are separated into a plurality of
regions corresponding to said plurality of cells by a plurality of second isolation
grooves (122), and a connection groove (123) is provided between said first isolation
groove (121) and said second isolation groove (122) to electrically connect said first
electrode of one of said cells with said second electrode of its neighboring cell.

A slacked photoelectric convener comprising a plu-
rahity of stacked photoelectric conversion units (3;5) each including
one conductivity type layer (31,51), a photoelectric converting layer
(32;52) of substantially intrinsic semiconductor, and a reverseconduc-
tivity type layer (33.53) that are formed on a substrate (1) sequentially
from the light incident side. At least one of the reverse conductivity
lypc layer (33) in the front photoelectric conversion unit (3) arranged
relatively on the light incident side anti the one conductivity type layer
(51) in the rear photoelectric conversion unit (5 (arranged contiguously
to the from photoelectric conversion unit (3) includes a silicon com-
posite layer (4). The silicon composite layer (4) has a thickness of
20)30 nm and an oxygen concentration of 25-60 atm. and a sili-
con-rich phase is included in an amorphous alloy phase of silicon and
oxygen.

Documents:

623-kolnp-2005-abstract.pdf

623-kolnp-2005-assignment.pdf

623-kolnp-2005-claims.pdf

623-KOLNP-2005-CORRESPONDENCE-1.1.pdf

623-kolnp-2005-correspondence.pdf

623-kolnp-2005-description (complete).pdf

623-kolnp-2005-drawings.pdf

623-kolnp-2005-examination report.pdf

623-kolnp-2005-form 1.pdf

623-kolnp-2005-form 18.pdf

623-KOLNP-2005-FORM 27.pdf

623-kolnp-2005-form 3.pdf

623-kolnp-2005-form 5.pdf

623-KOLNP-2005-FORM-27.pdf

623-kolnp-2005-gpa.pdf

623-kolnp-2005-granted-abstract.pdf

623-kolnp-2005-granted-assignment.pdf

623-kolnp-2005-granted-claims.pdf

623-kolnp-2005-granted-correspondence.pdf

623-kolnp-2005-granted-description (complete).pdf

623-kolnp-2005-granted-drawings.pdf

623-kolnp-2005-granted-examination report.pdf

623-kolnp-2005-granted-form 1.pdf

623-kolnp-2005-granted-form 18.pdf

623-kolnp-2005-granted-form 3.pdf

623-kolnp-2005-granted-form 5.pdf

623-kolnp-2005-granted-reply to examination report.pdf

623-kolnp-2005-granted-specification.pdf

623-kolnp-2005-reply to examination report.pdf

623-kolnp-2005-specification.pdf


Patent Number 235848
Indian Patent Application Number 623/KOLNP/2005
PG Journal Number 36/2009
Publication Date 04-Sep-2009
Grant Date 02-Sep-2009
Date of Filing 11-Apr-2005
Name of Patentee KANEKA CORPORATION
Applicant Address +2-4, NAKANOSHIMA 3-CHOME, KITA-KU, OSAKA-SHI, OSAKA
Inventors:
# Inventor's Name Inventor's Address
1 KOI YOHEI 24-8-304, KONOOKA-CHO, OTSU-SHI, SHIGA 520-0103
2 SASAKI TOSHIAKI 2-1-2-131, HIEITSUJI, OTSU, SHI, SHIGA 520-0104
3 YAMAMOTO KENJI 1-2-W1406, MIKATADI, NISHI-KU, KOBE-SHI, HYOGO 651-2277
4 YOSHIMI MASASHI 6-6-4, IBUKIDAINISHIMACHI, NISHI-KU, KOBE-SHI, HYOGO 651-2243
5 ICHIKAWA MITSURU 1-25-1, HIEITSUJI, OTSU-SHI, SHIGA 520-0104
6 YAMAMOTO KENJI 1-2-W1406, MIKATADI, NISHI-KU, KOBE-SHI, HYOGO 651-2277
7 YOSHIMI MASASHI 6-6-4, IBUKIDAINISHIMACHI, NISHI-KU, KOBE-SHI, HYOGO 651-2243
8 ICHIKAWA MITSURU 1-25-1, HIEITSUJI, OTSU-SHI, SHIGA 520-0104
9 SASAKI TOSHIAKI 2-1-2-131, HIEITSUJI, OTSU, SHI, SHIGA 520-0104
10 KOI YOHEI 24-8-304, KONOOKA-CHO, OTSU-SHI, SHIGA 520-0103
PCT International Classification Number H01L 31/075
PCT International Application Number PCT/JP2004/010115
PCT International Filing date 2004-07-15
PCT Conventions:
# PCT Application Number Date of Convention Priority Country
1 2003-367535 2003-10-28 Japan
2 2003-279493 2003-07-24 Japan
3 2003-367536 2003-10-28 Japan
4 2004-091897 2004-03-26 Japan