Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02204124 1997-04-30
Trichromatic Sensor
The invention relates to a component based on amorphous
silicon and its alloys, comprising two, in respect to each
other antiserially arranged, p-i-n or n-i-p or Schottky-
contact structures, in which the active layers are in each
case arranged in the normal way in the direction of light
incidence, whereby in the area of the first structure in
the direction of light incidence, the charge carriers
generated by blue light are collected for a first (V1)
voltage; and in the area of the second structure, in the
area of light incidence, the charge carriers generated by
red or green light are collected for a second (V2) or a
third (V3) voltage, and whereby at least one of the two
intrinsically conducting layers is constructed from two
partial layers.
A component of this type is known from the essay "New type
of thin film colour image sensor", Q. ZHU, H. Stiebig, P.
Rieve, J. Giehl, M. Sommer, M. Bohm; Conference Europto -
Sensors and Control for Advanced Automation II, Frankfurt /
Main, 20-24 June 1994.
Compared with components of crystalline silicon, photo
sensitive electronic components based on amorphous silicon
(a-Si:H) have the advantage of significantly increased
absorption of visible light. Basically such a photo
sensitive electronic component consists of two PIN diodes
connected antiserially in respect to each other, whereby
the alternatives NIPIN or PINIP are known; or of two metal-
semiconductor junctions (Schottky contacts) connected to
each other antiserially.
Technologically, such a component is produced by separating
a multitude of a-Si:H layers at low temperature (typically
250~C) by means of the PECVD (Plasma Enhanced Chemical
Vapour Deposition) process. Deposition takes place at first
for example on an insulating substrate, usually glass, by
CA 02204124 1997-04-30
connecting a translucent conductive oxide layer (TCO) which
later establishes a contact for an electrical voltage to be
applied externally to the component. By applying an
alternating field, a plasma is generated at the PECVD
reactor by the gas SiH4(silane) being decomposed into
silicon radicals and hydrogen. In this process, silicon
condenses on the substrate as an amorphous film containing
hydrogen. Starting from this, by adding phosphine, an n-
doped layer can be produced; or by adding boroethane, a p-
doped layer can be produced. In addition, it is well known
to increase the band gap of the amorphous silicon by adding
methane (CH4) to the silane, or to decrease the band gap by
adding germane (Ge H4).
The multilayer component produced in this way, having the
desired layer sequence, is now penetrated by radiation of
visible light in such a way that the direction of light
incidence is perpendicular to the plane of the layers.
Because the absorption coefficient of the sensor material
depends on the wave length of the incoming light and on the
band gap of the sensor material, different penetration
depths of light into the semiconductor material results.
This leads to blue light (wavelength approx. 450 nm) having
a significantly lesser penetration depth (absorption
length) than green or red light. By selecting the
respective size and polarity of the exterior direct voltage
applied to the component, and therefore the interior
electrical field, a spectral sensitivity of the component
can be attained. For example by applying respective
voltages to the element, a sensitivity for RGB light (red,
green, blue) in the multilayer structure can be attained.
In this the principal collecting region of the light-
generated charge carriers along the length of the component
and consequently its spectral sensitivity, is shifted,
depending on the exterior voltage applied.
. CA 02204124 1997-04-30
In order to optimise a NIPIN structure in respect of a
trichromatic sensor, it is known from the above-mentioned
report, on both sides of the p-doped intermediate layer, to
additionally provide intrinsically conducting defect layers
in whose thickness region the band gap is increased in
comparison to the remaining intrinsically conducting layers
(e.g. from 1.74 eV to 1.9 eV). This leads to an improved
red/green separation in the NIP structure in front in the
direction of light incidence, or to an improved red/green
separation in the subsequent PIN structure in the direction
of light incidence, (so-called band-gap engineering).
From US patent 53 11 047 a photo sensitive electronic
component based on amorphous silicon with NIPIN structure
is known.
From Applied Physics Letters 52(4) 1988, 275-277, a
heterojunction component (phototransistor) of the type
NIPIN is known which comprises two additionally-inserted
intrinsically conducting layers. From this it is known that
in the first junction preferably blue light, and in the
second junction preferably green or red light, are
absorbed. The second intrinsically conducting layer serves
to increase blue absorption.
It is the object of the invention to describe an electronic
component of the type mentioned in the introduction, as
well as a method for its manufacture which ensures that
under economic fabrication conditions a component sensitive
to RGB colours is created which features a high spectral
separation for RGB colours with negligible
infrared/ultraviolet contributions to the output signal.
This object is achieved according to the invention by the
two partial layers, in the presence of an electrical field,
having different collection lengths for the charge
carriers, in such a way that in the partial layer which is
CA 02204124 1997-04-30
in front in the direction of light incidence, a higher
product from charge carrier mobility and life, and in the
partial layer which is at the rear in the direction of
light incidence, a lower product from charge carrier
mobility and life is present, or that at least one of the
two intrinsically conductive layers is formed from two
partial layers which on the basis of respective doping, in
particular with SixGel-xl have such different mobilities
that the ~-tau products in the two partial layers differ
from each other by at least a factor of 10, whereby the
partial layer with the higher ~-tau product is in front in
the direction of light incidence.
The invention is characterized in that due to the splitting
up of at least one of the two intrinsically conducting
layers of the component, into two partial layers,
blue/green separation or green/red separation is improved.
In contrast to known processes (e.g. by influencing the
absorption or recombination behaviour through a partial
increase of the band gap "band-gap engineering") or by
incorporating additional defect layers, the variation of
the ~-tau product according to the invention has the
decisive advantage of making possible a significantly
improved spectral separation and the desired linear
behaviour.
As a result of this, the primary colours can be separated
with certainty in the case of both high and low light
intensity. The solution according to the invention results
in a component with a linear connection between the photo
voltage for the respective spectral peak and the photon
flow. This is th~ case over several orders of magnitude of
the phonon flow. Beyond this, there is no linear
interdependence of the spectral peaks. The linearity on the
one hand and the linear independence of the spectral peaks
on the other hand are the very reasons that make it
possible to be able to use such a component at all in
CA 02204124 1997-04-30
practice for colour recognition under various conditions of
illumination.
By incorporating an area with low, free mobility of the
charge carriers in the rear structure in the direction of
light incidence, charge carriers can only be extracted
during higher voltage and thus higher electrical field.
Holes which are generated in the front part by light of
short wavelength drift to the intermediate layer. The
electrons reach the barrier layer of the rear structure
because no recombination partners are generated in the rear
part. Longwave light (red) generates predominantly in this
area, so that the charge carriers recombine as a result of
the lower ~-tau product. Only with higher field strengths
is it possible for the holes to drift to the intermediate
layer. As a result, red/green separation occurs in the rear
structure.
When incorporating suitable materials and alloys (e.g. a-
Si:H/a-SixGe(lx):H) with different free mobilities in the
bands, in this partial layer, at lower voltages, charge
carriers can be collected less well due to the lower ~-tau
product. By contrast, these charge carriers can be
collected in preference in the case of higher voltages.
In the partial layer with the higher ~-tau product,
preferably green photons are collected, while in the rear
area in the direction of light incidence, preferably the
red charge carriers are collected. Due to the different
refractive indices, depending on the wavelength, of a-
Si:H/a-Si(C):H and a-SixGelx:H, for the frontal partial
layer of the subsequent intrinsically conducting layer an
area of a certain thickness can be determined in which more
green photons than red photons are absorbed. The charge
carriers generated can be extracted at lower negative
voltages and determine the maximum spectral sensitivity.
With higher negative voltages the maximum spectral
CA 02204124 1997-04-30
sensitivity is determined by the charge carriers which are
generated in the rear partial layer of this intrinsically
conducting layer.
The maximum spectral sensitivity for longwave light can be
set by the thickness of the layer of the SixGel-x material,
by the germanium content in this area and possibly by
gradation.
With the component according to the invention, by applying
three different voltages, three space charging zones
located at various depths in the component can be obtained
as a result, therefore yielding outstanding spectral
selectivity. A particular advantage consists of the
linearity extending over more than five orders of magnitude
of luminous intensity. In addition, the component according
to the invention has a low dark current and a high dynamic
range (> 120 dB, in respect of 1000 lux). A further
advantage consists of being able to preset the spectral
sensitivity in a customer-specific manner. Of particular
significance is the fact that selection of spectral
components is possible without the use of additional
optical filters.
Preferred embodiments are stated in the dependent claims.
A NIPIN or a PINIP structure can be considered as preferred
variants of photosensitive components which according to
the invention are modified by the partial layers with a
different ~-tau product. Technologically, the ~-tau
variation is preferably achieved by alloying germanium into
the amorphous silicon (a-SixGelx:H).
In order to improve blue/red separation as well as for
improved quantum efficiency, the intrinsically conducting
layer of the front structure in the direction of light
incidence, may consist of a carbonised layer (a-Si(C):H),
CA 02204124 1997-04-30
so that in addition to the ~-tau optimisation, an
adaptation of the band gap also takes place.
The particularly preferred embodiment of the invention
provides that the component is part of a colour sensor in
which the sandwich structure comprising the component is
arranged on the surface of an integrated circuit. Combining
a crystalline component (for example an ASIC) with the
multilayer component produced according to low-temperature
PECVD technology, results in a simple combination which can
be produced economically and which achieves high resolution
as a picture-yielding colour sensor. In this, each surface
element of the photosensitive component, according to the
micro structure of the integrated circuit or the ASIC, acts
as a single pixel element. Colour selectivity is set by the
voltage, presettable by the circuit applied to the region
of the pixel area. The result is a so-called thin film on
ASIC (TFA) sensor system which combines the advantages of
traditional ASICs made of crystalline silicon with an
optical sensor based on amorphous silicon.
Below, the invention is explained in more detail by means
of a drawing, as follows:
~ig. 1 is a schematic diagram to illustrate the layer
construction of a photosensitive electronic component
according to prior art
~ig. 2 is a schematic diagram to illustrate the operational
mode of the element according to Fig. 1, whereby
~ig. 2a shows the spatial arrangement of the individual
layers of the multilayer component
CA 02204124 1997-04-30
~ig. 2b shows the local flow of the electrical field strength
at U~0, and
~ig. 2c shows the local flow of the electrical field strength
at U<0
~ig. 3 is a schematic diagram to illustrate the layer
construction of a photosensitive electronic component
according to the invention
~ig. 4 is a schematic diagram to illustrate the operational
mode of the element according to Fig. 3, whereby
~ig. 4a shows the spatial arrangement of the individual
layers of the multilayer component
~ig. 4b shows the local flow of the electrical field strength
at U>0
~ig. 4c shows the local flow of the electrical field strength
at U<0, as well as the local flow of the ~-tau
product, and
~ig. 5 is a schematic diagram to illustrate the layer
construction of the combination of a photosensitive
electronic component according to the invention with
an integrated circuit.
Fig. 1 shows a cross section through a NIPIN layer system
in which the NIPIN layer sequence is deposited on a carrier
(glass). The glass substrate is subsequently coated with a
TCO layer comprising a light-transmitting, conductive
oxide. On top of this, subsequently, the individual
amorphous silicon layers are deposited in the sequence as
shown in Fig. 1.
- CA 02204124 1997-04-30
The deposition process is by means of the known PECVD
technology in which amorphous silicon is deposited in the
desired layer thickness at relatively low temperature
(approx. 250~C).
Back contact is established by an aluminium electrode to
which an exterior electrical voltage U is applied which
drives a current I that flows into the element, whereby the
TCO layer forms the reference potential. In this way the
arrangement acts like an antiserially connected combination
of two PIN diodes.
As is also shown in Fig. 1, light incidence is through a
glass substrate into the NIPIN layer system, perpendicular
to the surfaces of the layers.
A schematic diagram of the semiconductor structure is shown
in Fig. 2a. In the n-doped regions forming the barrier
layers of the NIPIN component, there is strong doping. In
these areas no charge carrier collection takes place
because there is a high recombination probability between
electrons and holes, corresponding to the high defect
densities. In that location, because of the doping
concentration, assumed to be constant, as is shown in Figs.
2b and 2c respectively, a linear increase in the electrical
field strength takes place. In the intrinsically conducting
regions assumed to be free of space charges, there is a
spatially approximately constant distribution of the
electrical field strength, whereby the contribution to the
space charge by mobile charge carriers, defects and
impurities is negligible. By means of the electrical field,
photo-generated charge carriers are collected as the
primary photo current. Thermally generated charge carriers
contribute to the dark current. Because the NIPIN structure
can be regarded as consisting of two antiserially connected
PIN diodes, the major drop in voltage occurs in the region
of the diode polarised in the direction of blocking. By
CA 02204124 1997-04-30
contrast, the electrical field in the diode poled in the
direction of transmission is negligible. As can be seen
from Figs 2b and 2c, provided the thermally-generated
charge carriers are negligible, the field strength existing
in the region of the intermediate p-layer is a function of
the applied voltage U, whereby in the centre of the p-layer
at a certain location, the field strength has a zero
passage.
Depending on the voltage applied externally to the element,
the flow of the electrical field strength is either
according to Fig. 2b (U>O) or according to Fig. 2c (UcO).
If now, according to Fig. 2b a positive voltage, for
example +2 volt, is applied to the element shown in Fig. 1,
then the front diode is poled in the direction of blocking,
so that inside it a strong electrical field forms in which
charge carriers can be separated. Because this region is
located in the front in the direction of light incidence,
the spectral components of the light with lower absorption
length, i.e. blue light, are absorbed there.
By contrast, if a negative voltage is applied to the
component, the space charge zone forms in the region of the
rear PIN diode, so that there the light is absorbed in the
green or red spectral range which has a greater penetration
depth.
Starting from this, Fig. 3 shows an embodiment of a photo-
sensitive electronic component according to the invention.
Supplementary to the component shown in connection with
Fig. 1, according to Fig. 3 the intrinsically conducting
rear layer in the direction of light incidence, is
subdivided into two partial layers I, II. Subdivision takes
place because in partial layer I there is a larger ~-tau
product than in partial layer II. This connection is
evident from Fig 4c, by the ~-tau characterlstic drawn in
- CA 02204124 1997-04-30
bold from which the gradation in the areas I, II can be
recognlsed .
With otherwise identical basic functions of the component,
this leads to the following mode of operation:
Corresponding to Fig. 4b, charge carriers generated by blue
light are again collected in preference in those
circumstances where a positive exterior voltage is applied
to the component.
By contrast, Fig. 4c shows the case where a negative
voltage U is applied to the component. Let us first examine
the case where the amount of negative voltage is relatively
small, for example where a voltage of -0.5 volt is applied
to the component. By incorporating materials with low free
mobilities in the rear area of the second intrinsically
conducting layer (partial layer II) and the lower ~-tau
product resulting therefrom, collection of charge carriers
which are generated by light with high penetration depth
(corresponding to the red light) is worsened. Therefore, in
the case of such a voltage, in preference, charge carriers
are collected which are generated by green photons. Because
partial layer I does not contain germanium, the band gap is
so large there that red photons cannot be absorbed or can
only be absorbed insufficiently.
By contrast, if the amount of negative voltage rises, then
in preference red charge carriers can be collected. Due to
the fact that additionally, the band gap in partial layer
II is narrow as a result of the layer containing germanium,
red photons are especially well absorbed there.
Thus gradating the ~-tau product in the area of the rear
intrinsically conducting layer, results in a significant
improvement in the red/green separation and therefore, in
combination with the blue absorption occurring in the
CA 02204124 1997-04-30
region of the frontal intrinsically conducting layer, pro-
vides the possibility of a trichromatic sensor with very
high spectral selectivity.
Starting with the basic functions, as described above, of
the component according to the invention, preferably the
following improvements are also provided:
The intrinsically conducting layer of the NIP diode located
in front in the direction of light incidence, can consist
of a carbonised layer so that improved blue/red separation
results. To reduce the dark current, an area with a
narrower band gap can be provided within this intrinsically
conducting layer.
In order to improve the maximum quantum efficiency, the n-
layer of the front structure in the direction of light
incidence may be carbonised or may be made as a micro
crystalline layer whereby the middle p-layer, too, may be
carbonised.
In addition, the sensitivity of the first structure as
regards blue absorption may be moved towards shorter
wavelengths by a slight doping in the transitional area
between the intrinsically conducting layer and the middle
p-layer.
In order to optimise spectral selectivity, a gradated a-
SiXGElx:H layer or further undoped layers may be placed
behind or in front of the a-SixGElx:H layer.
Starting from the component described in the context of
Fig. 3, a chromatic picture sensor, shown in Fig. 5, can be
created, in which the multilayer structure as described is
placed as a substrate, in the shape of an ASIC, on top of
an integrated circuit. The production process takes place
by interconnecting an insulation layer and a metal layer.
CA 02204124 1997-04-30
In this, the layer sequence during deposition is the other
way round than that described in connection with Fig. 1 or
3 respectively.
According to Fig. 5, the light then enters the structure
from above. Depending on the selection of the optical cell
by the crystalline ASIC structure, for each pixel element,
depending on the applied voltage, different spectral
behaviour results. In this way the irradiated light can be
analysed, pixel by pixel, in regard to its spectral
components, and the light signal converted in this way may
be further processed electrically.
