GAS SENSOR
[0001] The invention relates to a gas sensor with at least one gas-sensitive layer,
having at least one surface area, in which the electron affinity depends on the
concentration of a target gas coming in contact with the surface area, and with at
least one potential sensor capacitively coupled to the surface area via an air gap.
[0002] Such a gas sensor for the measurement of hydrogen gas concentrations is
disclosed in DE 43 33 875 C2. The gas sensor has a silicon substrate in which a field
effect transistor is integrated. The field effect transistor has a gate electrode that is
conductively connected to a sensor electrode, over which a gas-sensitive layer is
arranged, separated from the sensor electrode by an air gap and capacitively
coupled to the sensor electeode via the air gap. A cover electrode is applied on the
reverse side of the gas-sensitive layer opposite the sensor electrode. A surface area
of the gas-sensitive layer opposite the sensor electrode comes in contact with the
target gas, which is adsorbed upon contact with the surface area. When there is a
change in the concentration of the target gas, the electron affinity in the surface area
of the gas-sensitive layer changes. Because the sensor electrode is capacitively
coupled to the surface area, the electrical potential at the gate electrode also
changes. Depending on the change in potential, the current is guided between the
drain and source terminal of the field effect transistor.
[0003] In normal, ambient air, a thin layer of atmospheric oxygen is dissociatively
adsorbed on the surface of the gas-sensitive layer, i.e. as oxygen atoms, not oxygen
molecules as they occur in the air. If the target gas enters the vicinity of the gas-
sensitive layer, the target gas is first adsorbed on the surface, whereby the particular
gas partially displaces the atmospheric oxygen previously adsorbed on the surface
and assumes its adsorption positions. Both effects, the adsorption of the target gas
and reduction of oxygen, additionally contribute to the change in surface electron
affinity. Simultaneously, however, a reaction between hydrogen and oxygen takes
place on the surface, aided by the catalytic effect of the gas-sensitive layer, creating
water. At low temperatures below approx. 60 °C, this causes only the surface
hydrogen layer to gradually diminish; at higher temperatures, above approx. 60 °C,
this reaction occurs so quickly that it additionally causes the concentration of
hydrogen in the immediate vicinity of the gas-sensitive layer to diminish as well. This
allows the oxygen layer on the surface to increase again. All three effects distort the

electron affinity in the opposing direction. This reaction can take place over hours or
even seconds, depending on the temperature of the gas-sensitive layer, which can
drastically disturb the measurement signal. Furthermore, the gas sensor's sensitivity
towards the target gas decreases logarithmically when the target gas concentration
increases. Analysis of the gas sensor measurement signal is therefore especially
difficult with higher target gas concentrations.
[0004] Object of the invention is to create a gas sensor of the type described above
that makes it possible to measure a predetermined target gas concentration with
great precision.
[0005] This object is achived for the invention where in-the surface area of the gas-
sensitive layer is covered by an electric insulating layer;that is inert to the target gas
and is bonded ito the gas sensitive layer and is designed so as to be permeable to
the target gas and a differient non-target gas that can adsorb on the surface area so
that the coating for the target gas and the non targeted gas exhibit differient diffusion
constants and that the diffusion constants are coordinated in such a way so that the
measurement sensitivity of the gas sensor increases for the target gas when the
target gas concentration, in the presence of the non-target gas, crosses a
predetermined concentration threshold. The term "measurement sensitivity" refers to
the amount of change-in the sensor signal of the potential sensor divided by the
amount of change in the gas concentration after the abatement of possible transient
signal particles contained in the measurement signal.
[0006] Surprisingly, it ha;s been determined that with the gas sensor according to the
invention, the sensitivity to the target gas increases drastically when, in the presence
of a nearly constant concentration of the non-target gas, it crosses the concentration
threshold. In a corresponding manner, the sensitivity to the target gas decreases
drastically when the target gas, in the presence of a nearly constant concentration of
the non-target gas, falls below the concentration threshold. Furthermore, coating
disturbances in the measurement signal of the gas sensor that can be traced back to
the interplay between target gas and non-target gas can be largely avoided.
Advantageously, this enables the very precise detection of whether the target gas
concentration lies above or below the concentration threshold. The gas sensor can
particularly be used as a gas leak sensor.
[0007] In a preferred embodiment of the invention, the coating should contain at
least one polymer. With this type of coating, a gas sensor using a gas-sensitive layer
of platinum in a 1-4% concentration range for the detection of the target gas
hydrogen can attain a high level of sensitivity. The polymer can be a polyimide, for
example.

[0008] It is especially advantageous if the polymer is polymethylmethacrylate
(PMMA). An almost abrupt increase in the sensor signal of the potential sensor can
be achieved when the concentration threshold is crossed with a gas sensor with this
coating. Furthermore, it has been determined that the gas sensor with this coating
has long-term stability at high temperatures up to approx. 180 °C and/or in damp
conditions.
[0009] It is advantageous if the coating is a lacquer. The coating can then be
inexpensively applied to the gas-sensitive layer during manufacture of the gas
sensor.
[0010] In a preferred embodiment of the invention, the thickness of the coating is
between 0.3 urn and 4 urn, particularly between 0.5 urn and 2.5 urn. This layer
thickness enables a favorable measurement signal characteristic for the gas sensor
of the target gas.
[0011] In an advantageous embodiment of the invention, the level of the
concentration threshold is dependent upon the temperature of the gas-sensitive layer
and coating and the gas sensor has a temperature control unit, preferably a heater,
for setting the temperature of the gas-sensitive layer and coating. Through heating
and/or cooling of the gas-sensitive layer, the concentration threshold can be set to a
desired level.
[0012] In a further modification of the invention, the gas sensor has a control device
with an actual value entry connected to the potential sensor, a set point entry
connected to a set point device, and a control signal output connected to the
temperature control unit, whereby, in order to customize the concentration threshold
to the target gas concentration, a set point can be assigned to the set point entry with
the help of the set point device, whose value corresponds to the sensor signal of the
potential sensor at the concentration threshold and/or has a target gas concentration
above the concentration threshold. Different target gas concentrations can thus be
measured with great precision, the concentration threshold being set so that each
threshold agrees with or is slightly smaller than the target gas concentration.
[0013] In a preferred embodiment of the invention, the gas-sensitive layer is made of
platinum or palladium. The gas sensor then has a high level of detection sensitivity
towards the target gas hydrogen.
[0014] The potential sensor is preferably a field effect transistor with a substrate to
which a drain and a source are attached, whereby a channel is formed between drain
and source and whereby the channel is capacitively coupled to the surface area of
the gas-sensitive layer, either directly via the air gap or indirectly via a gate electrode
acting with the channel and with a sensor electrode conductively connected to the

gate electrode. The field effect transistor can thus be a SGFET or a CCFET.
Thereby, the gas sensor can have a compact design and can furthermore be easily
integrated in a semiconductor chip. Thus, an evaluation device can also be
integrated into the semiconductor chip to process the measurement signals of the
gas sensor.
[0015] In another advantageous embodiment of the invention, the gas sensor is
designed as a Kelvin probe, in which the potential sensor is capacitively coupled to
the surface area of the gas-sensitive layer via an electrode that can move towards
and away from the gas-sensitive layer and is separated from the surface area of the
gas-sensitive layer by the air gap. Thereby, the electrode can, for example, be
positioned relative to the gas-sensitive layer and brought to oscillate by means of a
Piezo actuator. The potential sensor is assigned an evaluation and piloting device,
that feeds a counter voltage to the electrode chosen so that the potential measured
by the potential sensor averages zero. The counter voltage is a gauge for the
concentration of the target gas in contact with the surface area of the gas-sensitive
layer.
[0016] Illustrative embodiments are explained in more detail in the following, with
reference to the drawings, wherein:
[0017] Fig. 1 a cross-section of a gas sensor that has a SGFET, whose channel is
capacitively coupled to a gas-sensitive layer with a passive coating via an air gap,
[0018] Fig. 2 a cross-section of a gas sensor that has a CCFET, whose sensor
electrode is capacitively coupled to a gas-sensitive layer with a passive coating via
an air gap,
[0019] Fig. 3 a cross-section of a gas sensor designed as a Kelvin probe, where the
gas-sensitive layer has a passive coating,
[0020] Fig. 4 a graph illustration of a sensor signal (top curve) and the target gas
concentration (bottom curve) in an illustrative embodiment of the gas sensor,
whereby the x-axis is time t, the left y-axis is the amplitude S of a potential sensor's
sensor signal and the right y-axis is the target gas concentration k,
[0021] Fig. 5 a similar illustration to Fig. 4 whereby, however, the temperature of the
gas sensor is lower than in Fig. 4,
[0022] Fig. 6 a similar illustration to Fig. 5 whereby, however, the temperature of the
gas sensor is lower than in Fig. 5,
[0023] Fig. 7 a graph illustration of a threshold for the target gas concentration of
the gas sensor, whereby the x-axis is the temperature and the y-axis is the threshold,
and
[0024] Fig. 8 a schematic illustration of a control device.

[0025] A gas sensor designated in its entirety by 1 in Fig. 1 has a substrate 2, in
which a potential sensor 27 is integrated. The potential sensor 27 has a drain 3 and a
source 4 that are arranged in an n-doped transistor well. The drain 3 and the source
4 can be made of p-doped silicon, for example. The drain 3 is connected to a drain
terminal, which is not shown in any greater detail in this drawing, via electrical
conductors. In a corresponding manner, the source 4 is connected to a source
terminal. Between drain 3 and source 4 in the substrate 2, a channel 5 is formed, to
which an electrically insulating thin oxide layer is applied, which acts as a gate
dielectric.
[0026] Over the channel 5, a gas-sensitive layer 7 is applied to a substrate 6, which
can be made of, for example, a precious metal, particularly platinum or palladium,
and is separated from the channel 5 by an air gap 8. The surface area 9 of the gas
sensitive layer 7 that is opposite the channel 5 is capacitively coupled to the channel
5 via an air gap 8.
[0027] The substrate 6 is connected on both sides of the gas-sensitive layer 7 with
the substrate 2 via an electric insulating layer 10. It can be clearly discerned in Fig. 1
that the substrate 6 and the gas-sensitive layer 7 form a suspended gate.
[0028] The air gap 4 is connected to the atmosphere surrounding the gas sensor 1
via at least one of the openings, not shown in any greater detail in this drawing. Via
this opening, the surface area 9 of the gas-sensitive layer 7 can come in contact with
a detectable target gas, namely hydrogen, and a non-target gas, namely an
electronegative gas such as atmospheric oxygen. When in contact with the surface
area 9 the target gas and the non-target gas adsorb on the surface area 9. Upon the
adsorption of the target gas, the electron affinity of the surface area 9 changes which
leads to a change in the electrical potential in the channel 5.
[0029] In the illustrative embodiment in Fig. 1, the channel 5 is designed as open
(ISFET) and is capacitively coupled to the gas-sensitive layer 7 via the thin oxide
layer and the air gap 8. It can be clearly discerned that the channel 5 is arranged to
the side of the gas-sensitive layer 7 opposite the air gap 8.
[0030] In the illustrative embodiment in Fig. 2, the field effect transistor is configured
as a CCFET, in which the channel 5 is arranged laterally next to the gas-sensitive
layer 7 in the substrate 2 and is covered by a gate electrode 11. To capacitively
couple the channel 5 to the gas-sensitive layer 7, the gate electrode 11 is connected,
via an electric coupler 12, to a sensor electrode 13, which is arranged on the side of
the air gap 8 that is opposite the surface area 9 of the gas-sensitive layer 7 on an
insulating layer 10 found on the substrate 2. The insulating layer 10 can be a Si02
layer, for example. The construction of the SGFET's suspended gate corresponds to

Fig. 1.
[0031] In the illustrative embodiment in Fig. 3, the gas sensor 1 is designed as a
Kelvin probe. The gas-sensitive layer 7 is attached to an electrically conducting
substrate 14 and has a surface area 9 on the side opposite the substrate 14, on
which the target gas can adsorb. The surface area 9 is separated from an electrode
15 by an air gap 8 and forms an electrical capacitance with it.
[0032] The electrode 15 can be brought to oscillate with the help of an actuator that
is not shown in any greater detail in this drawing. Thereby, the electrode 15 moves
alternately towards and away from the gas-sensitive layer 7 in accordance with the
arrow Pf. The electrode 15 and the substrate 14, and respectively the gas-sensitive
layer 7, are connected to an evaluation and piloting device 17 with ports 16. This
evaluation and piloting device 17 has a potential sensor, not shown in any greater
detail, that is connected to the ports 16 for the measurement of electrical potential
between the gas-sensitive layer 7 and the electrode 15. The evaluation and piloting
device 17 also has a variable voltage supply source connected to the potential
sensor via a control connection that is applied via a counter voltage between the
potential sensor and the electrode 15 and/or the substrate 15. The counter voltage is
chosen so that the potential measured by the potential sensor will average zero.
[0033] In the previously described illustrative embodiments of the gas sensor 1, the
surface area 9 of the gas-sensitive layer 7 is in each case covered, by an electrically
insulating coating 18 that is inert to the target gas and that preferably consists of
polymethylmethacrylate (PMMA) or polyimide. The coating 18 is bonded to the gas-
sensitive layer 7. The coating 18 is configured as a thick layer with an almost
constant thickness, preferably between 0.5 urn and 2.5 urn.
[0034] The coating 18 is permeable for both the target gas as well as the non-target
gas. The coating 18, however, has a different diffusion constant for the target gas
than for the non-target gas. The diffusion constants, the target gas and the non-target
gas, are coordinated with each other in such a way that the sensitivity of the gas
sensor 1 to the target gas increases drastically when the concentration of the target
gas crosses a threshold 19 in the presence of the non-target gas. The level of the
threshold 19 depends on the temperature.
[0035] It is discernable in Figs. 4 through 6 that, at constant temperatures of target
gas concentrations that lay within an initial concentration range that is upwardly
limited by the concentration threshold 19 the sensor signal 20 of the potential sensor
27 initially increases logarithmically with the target gas concentration 21. The sensor
signal 20 of the potential sensor 27 lies in an initial modulation range 29 in the first
concentration area.

[0036] In a second concentration range that shares a bottom border with the
concentration threshold and is considerably narrower than the first concentration
range, the sensor signal 20 increases drastically at a constant temperature. In the
second concentration range, the sensor signal 20 lies in a second modulation range
30, in which the measurement sensitivity of the gas sensor 1 is greater than in the
first modulation range 29. In a third concentration range that lies above and borders
the second concentration range, the sensor signal 20 of the potential sensor 27
remains essentially constant when the temperature is constant at a value that
borders the second concentration range.
[0037] It is discernable in Fig. 7 that the concentration threshold 19 depends on the
temperature of the layer sequence formed by the gas-sensitive layer 7 and the
coating 18 and continuously climbs with increasing temperature. The increase occurs
almost exponentially with temperature. If necessary, the exponential increase can be
linearly approximated in the range relevant for the concentration measurement.
[0038] The gas sensors depicted in Figs. 1 through 3 each have a temperature
control unit 22, that is only schematically depicted in Fig. 8, with which the
temperature of the gas-sensitive layer 7 and the coating 18 can be adjusted. A
control entry for the temperature control unit 22 is connected with a control signal
output 23 of a control device that serves to adjust the temperature of the gas-
sensitive layer 7 and the coating 18 so that the sensor signal 20 of the potential
sensor 27 can be essentially independent of the gas concentration and lies within the
second modulation range 30.
[0039] The control device has a comparator 24 that has an actual value entry
connected to the potential sensor 27 and a set point entry connected to a set point
device 25. An output of the comparator 24 is connected to the control signal output
23 via a controller 26. With the help of the set point device 25, a set point 28 is
assigned to the set point entry located in the second modulation range 30, thus
having a value where the sensor signal (20) of the potential sensor (27) has a target
gas concentration above the concentration threshold (19).
[0040] In one of the first operating modes for the gas sensor 1, the controller 26
controls the temperature control unit 22 in such a way so that in the event of a
discrepancy between the sensor signal 20 of the potential sensor 27 and the set
point 28, the temperature of the gas-sensitive layer 7 and the coating 18 is changed
in order to reduce the discrepancy. When the sensor signal 20 of the potential sensor
27 agrees with the set point 28, the temperature of the gas-sensitive layer 7 and the
coating 18 is a gauge for the target gas concentration.
[0041] In a second operating mode, with the help of the temperature control unit 22,

the temperature of the gas-sensitive layer 7 and the coating 18 are adjusted to a
constant value. Alternately, the temperature control unit 22 can also be turned off in
the second operating mode, so that the temperature of the gas sensor 1 then
approximately reflects the ambient air temperature. The second operating mode is
always then activated when the temperature detected by the controller 26 falls below
a predetermined minimum temperature. This temperature can lie between 60-80 °C,
for example.
[0042] In the second operating mode, the target gas concentration is established
based on the signal value of the sensor signal 20 of the potential sensor 27 and
based on parameters that could, for example, exist in the form of a characteristic
curve. In the second operating mode, the signal analysis essentially reflects that of a
conventional gas sensor As, soon as the minimum temperature is exceeded, the
mode switches over so as to determine the target gas concentration-based on the
preset temperature thus the first operating mode is utilized for high target-gas
concentrations and the second operating mode for lower target gas concentrations.
[0043] The first operating mode is preferably chosen when the target gas
concentration is between 1% and 4%. The appropriate concentration range can be
determined through experimentation. In this mode, a nearly exponential correlation
between temperature and: target gas concentration results. Thereby, the gas sensor
1 according to the invention enables a clearly improved solution in comparison with a
conventional gas sensor .-.'

We Claims -
1. Gas sensor (1) with at least one gas-sensitive layer (7) that has at least one
surface area (9) in which the electron affinity depends on the concentration of a
target gas brought in contact with the surface area (9) and with at least one
electrical potential sensor (27) that is capacitively coupled with the surface area
(9) via an air gap (8), characterized in that the surface area (9) of the gas-
sensitive layer (7) is covered by an electric insulating coating (18) that is inert to
the target gas and is bonded to the gas-sensitive layer (7) and designed so that it
is permeable on the surface layer (9) for the target gas and a different, non-target
gas, that the coating (18) has different diffusion constants for the target gas and
the non-target gas, and that the diffusion constants are coordinated with each
other in such a way that the measurement sensitivity of the gas sensor (1)
increases for the target gas when the target gas concentration exceeds a
predetermined concentration threshold in the presence of the non-target gas (19).
2. Gas sensor (1), as in claim 1, characterized in that the target gas is a reducing
gas, in particular hydrogen, and that the non-target gas is an electronegative gas,
in particular oxygen.
3. Gas sensor (1), as in claims 1 or 2, characterized in that the coating (18) contains
at least one polymer.
4. Gas sensor (1), as in claim 3, characterized in that the polymer is polyimide.
5. Gas sensor (1), as in claim 3, characterized in that the polymer is
polymethylmethacrylate.
6. Gas sensor (1), as in one of the claims 1 through 5, characterized in that the
coating (18) is a lacquer.
7. Gas sensor (1), as in one of the claims 1 through 6, characterized in that the
thickness of the coating (18) is between 0.3 urn and 4 urn and preferably between
0.5 urn and 2.5 urn.

8. Gas sensor (1), as in one of the claims 1 through 7, characterized in that the level
of the concentration threshold (19) depends on the temperature of the gas-
sensitive layer (7) and the coating (18) and that the gas sensor (1) has a
temperature control unit, preferably a heater, for setting the temperature of the
gas-sensitive layer (7) and the coating (18).
9. Gas sensor (1), as in one of the claims 1 through 8, characterized in that it has a
control device with an actual value entry connected to the potential sensor (27), a
set point entry connected to a set point device (25), and a control signal output
(23) connected to the temperature control unit (22), and that, to adjust the
concentration threshold(49) to the target gas concentration (21), with the help of
the set point device (25),ta set point (28) can be assigned to the set point entry,
which has a value where the sensor signal (20) of the potential sensor (27) lies at
the concentration (l9) and/or a target gas concentration that lies above
the concentration threshold (19).
10. Gas sensor (1), as in one of the claims 1 through 9, characterized in that the gas-
sensitive layer (7) is platinum or palladium.
11. Gas sensor (1), as in one of the claims 1 through 10, characterized in that the
potential sensor (27) is a field effect transistor with a substrate (2), on which a
drain (3) and a source (41) are arranged, that a channel (5) is formed between
drain (3) and source (4), and that the channel (5) is capacitively coupled to the
surface area (9) of the gas-sensitive layer (7) directly via the air gap (8) or
indirectly via a gate-electrode (11), acting with the channel (5), and a sensor
electrode (13) that is cohductively connected with the gate electrode (11).
12. Gas sensor (1), as in one of the claims 1 through 11, characterized in that it is
designed as a Kelvin probe, in which the potential sensor (27) is capacitively
coupled to the surface area (9) of the gas-sensitive layer (7) via an electrode (15)
that is separated from the surface area (9) of the gas-sensitive layer (7) by the air
gap (8) and that can be moved towards and away from the gas-sensitive layer
(7).

A gas sensor (1) has a gas-sensitive layer (7) with a surface area (9) where the
electron affinity depends on the concentration of a target gas brought in contact with
the surface area (9). An electrical potential sensor (27) is capacitively coupled to the
surface area (9) via an air gap (8). The surface area (9) of the gas-sensitive layer (7)
is covered by an electric insulating layer (18) that is inert to the target gas and is
bonded to the gas-sensitive layer (7). The coating (18) is designed in such a way that
it is permeable for the target gas and a different, non-target gas that can be adsorbed
on the surface area (9). The coating (18) has different diffusion constants for the
target gas and the non-target gas. The diffusion constants are coordinated with each
other in such a way that the sensitivity of the gas sensor (1) to the target gas
increases when the target gas concentration exceeds a predetermined concentration
threshold in the presence of the non-target gas.