U.S. patent application number 17/399207 was filed with the patent office on 2021-12-02 for molecular sensor based on virtual buried nanowire.
This patent application is currently assigned to Ramot at Tel-Aviv University Ltd.. The applicant listed for this patent is Ramot at Tel-Aviv University Ltd.. Invention is credited to Yossi ROSENWAKS, Gil SHALEV.
Application Number | 20210372965 17/399207 |
Document ID | / |
Family ID | 1000005770018 |
Filed Date | 2021-12-02 |
United States Patent
Application |
20210372965 |
Kind Code |
A1 |
SHALEV; Gil ; et
al. |
December 2, 2021 |
MOLECULAR SENSOR BASED ON VIRTUAL BURIED NANOWIRE
Abstract
The present invention provides a method and a system based on a
multi-gate field effect transistor for sensing molecules in a gas
or liquid sample. The said FET transistor comprises dual gate
lateral electrodes (and optionally a back gate electrode) located
on the two sides of an active region, and a sensing surface on top
of the said active region. Appling voltages to the lateral gate
electrodes, creates a conductive channel in the active region,
wherein the width and the lateral position of the said channel can
be controlled. Enhanced sensing sensitivity is achieved by
measuring the channels conductivity at a plurality of positions in
the lateral direction. The use of an array of the said FTE for
electronic nose is also disclosed.
Inventors: |
SHALEV; Gil;
(Ramat-HaSharon, IL) ; ROSENWAKS; Yossi;
(Hod-HaSharon, IL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Ramot at Tel-Aviv University Ltd. |
Tel-Aviv |
|
IL |
|
|
Assignee: |
Ramot at Tel-Aviv University
Ltd.
Tel-Aviv
IL
|
Family ID: |
1000005770018 |
Appl. No.: |
17/399207 |
Filed: |
August 11, 2021 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
16027403 |
Jul 5, 2018 |
11112379 |
|
|
17399207 |
|
|
|
|
14380732 |
Aug 25, 2014 |
10054562 |
|
|
PCT/IL2013/050182 |
Feb 28, 2013 |
|
|
|
16027403 |
|
|
|
|
61604041 |
Feb 28, 2012 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01N 27/4146 20130101;
H01L 21/324 20130101; G01N 27/30 20130101; H01L 21/22 20130101;
G01N 33/551 20130101; G01N 27/12 20130101; G01N 33/54373 20130101;
G01N 27/4145 20130101 |
International
Class: |
G01N 27/414 20060101
G01N027/414; G01N 33/543 20060101 G01N033/543; G01N 27/12 20060101
G01N027/12; G01N 27/30 20060101 G01N027/30; H01L 21/22 20060101
H01L021/22; H01L 21/324 20060101 H01L021/324 |
Claims
1. A system for sensing at least one type of molecules in a gas or
liquid sample, comprising: a) at least one multi-gate field effect
transistor, comprising: 1) a piece of semiconductor with a first
region extending between a source region and a drain region, and
left and right lateral regions extending along the first region on
different sides; 2) left and right lateral gate electrodes that
respectively produce an electric field in the left and right
lateral regions, creating a conducting channel in the first region
when appropriate voltages are applied to them, a position of the
conducting channel depending on the applied voltages; 3) a sensing
surface adjacent to the first region, that molecules of the at
least one type adhere to when the sensing surface is exposed to the
molecules, the conductivity of the conducting channel being
measurably affected by a local concentration of the adhering
molecules near the position of the conducting channel; and b) a
controller that controls power supplies to successively apply
different voltages to the lateral gate electrodes of the
transistor, moving the conducting channel to a plurality of
different positions in a lateral direction, and at each position
uses a circuit to measure the conductivity of the conducting
channel by measuring a source to drain current at a source to drain
voltage, and uses the measured conductivity to calculate a local
concentration of the adhering molecules at that position.
2. A system according to claim 1, wherein the sensing surface is
coated with a ligand that binds specifically to the molecules that
are being sensed.
3. A system according to claim 1, wherein the source region and
drain region are doped with dopants of a same sign, and the left
and right lateral regions are doped with dopants of an opposite
sign to the source and drain regions.
4. A system according to claim 3, wherein the first region is doped
with a dopant of the same sign as the source and drain regions.
5. A system according to claim 4, wherein the concentration of
dopants of the lateral regions extends into the first region,
falling off gradually over a scale length greater than the width of
the conducting channel.
6. A method of modifying the field effect transistor in the system
of claim 4, comprising heat treating the transistor under
conditions such that some of the dopants from the left and right
lateral regions diffuse into the first region, reducing an
effective width of the first region by at least 30% at its
narrowest point, but not reducing the effective width to zero at
any point.
7. A system according to claim 1, wherein the first region is
narrower than 1 micrometer between the left and right lateral
regions.
8. A system according to claim 1, wherein the field effect
transistor also comprises a back gate electrode, located in a
direction away from the sensing surface and separated from the
first region at least by an insulator layer.
9. A system according to claim 1, which, for at least one choice of
gate electrode voltages, would have a width of the conducting
channel and a distance of the conducting channel from the sensing
surface such that an equilibrium concentration of the adhering
molecules could be determined when a concentration of the molecules
in air that the sensing surface is exposed to is only 100 parts per
million.
10. A system according to claim 1, wherein the at least one field
effect transistor comprises a plurality of field effect
transistors, with their sensing surfaces chemically modified in
substantially a same way for binding to molecules in a gas or
liquid sample, and the controller controls power supplies to change
the lateral gate voltages to change the position of the conductive
channel in a lateral direction in each transistor, and uses a
circuit to measure the conductivity of the conductive channel at a
plurality of different positions in each transistor, after exposing
the sensing surfaces of the transistors to the sample, and
calculates a greatest concentration of said molecules adhering near
any of the positions, for each transistor, from the measured
conductivities, and calculates an average over the transistors of
the greatest concentrations of the adhering molecules.
11. A system according to claim 1 for use as an electronic nose for
sensing a plurality of different types of molecules, wherein the at
least one field effect transistor comprises a plurality of field
effect transistors with sensing surfaces having different chemical
properties, causing them to have different relative tendencies for
the different molecules to adhere to them, and the controller
controls power supplies to change lateral gate voltages to change
the position of the conducting channel in a lateral direction, and
calculates a concentration of any adhering molecules near each of
the positions of the conducting channels after the sensing surfaces
are exposed to a gas or liquid sample, from the conductivity
measured at each of the positions, for each transistor, and finds
the type of molecules present in the sample by comparing a pattern
of the concentrations of molecules adhering to each field effect
transistor, to an expected pattern of concentrations of adhering
molecules for each of the types of molecules.
12. A system according to claim 1, wherein the field effect
transistor also comprises a dielectric layer situated over the
first region, wherein the sensing surface comprises a surface of
the dielectric layer.
13. A system according to claim 1, wherein the sensing surface of
the transistor comprises an exposed surface of the first
region.
14. A system according to claim 13, wherein the semiconductor
comprises silicon, and the exposed surface of the active region
comprises methyl-terminated silicon.
15. A system according to claim 1, wherein the sensing surface
lacks a reservoir for holding a liquid sample.
16. A system according to claim 1, also comprising a reservoir that
includes the sensing surface, exposing the sensing surface to
liquid samples contained in the reservoir.
17. A system according to claim 1, wherein the controller measures
the source to drain current at a source to drain voltage by one or
more of keeping the voltage at a constant value at the different
positions and measuring changes in the current, keeping the current
at a constant value at the different positions and measuring
changes in the voltage, and keeping a function of the current and
the voltage at a constant value at the different positions and
measuring changes in a different function of the current and the
voltage.
Description
RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 16/027,403 filed on Jul. 5, 2018, which is a
continuation of U.S. patent application Ser. No. 14/380,732 filed
on Aug. 25, 2014, now U.S. Pat. No. 10,054,562, which is a National
Phase of PCT Patent Application No. PCT/IL2013/050182 having
International Filing Date of Feb. 28, 2013, which claims the
benefit of priority under 35 U.S.C. .sctn. 119(e) from U.S.
Provisional Patent Application No. 61/604,041 filed on Feb. 28,
2012. The contents of the above applications are all incorporated
by reference as if fully set forth herein in their entirety.
FIELD AND BACKGROUND OF THE INVENTION
[0002] The present invention, in some embodiments thereof, relates
to a semiconductor chemical sensor and, more particularly, but not
exclusively, to a gas sensor based on a field effect
transistor.
[0003] Commercially available gas sensors include IR sensors, Toxic
sensors, and Pellistors, all sold by City Technology, Ltd., and
metal oxide gas detectors, sold by Figaro USA, Inc. (FIS, Inc.).
The sensors sold by City Technology are described, for example, at
worldwideweb(dot)citytech(dot)com, and the sensors sold by Figaro
are described, for example, at
worldwideweb(dot)figarosensor(dot)com. Gas sensors that could be
manufactured more cheaply, and/or have greater sensitivity and/or
greater specificity, would be useful.
[0004] Gas sensors based on nanowire of various materials, for
example Si, ZnO, SnO, and other materials, can exhibit
exceptionally high resolution and sensitivity. However, the
manufacture of commercial gas sensors based on such nanowires may
not be feasible at the present time, since the fabrication of these
structures, for example with the VLS method, cannot accommodate
high volume manufacturing (HVM). Alternatively, high volume CMOS
manufacturing of nanowires could be realized in the future but with
a substantial increase in cost, even several orders of
magnitude.
[0005] Additional background art includes U.S. Pat. No. 6,173,602
to Moseley, "Transition metal oxide gas sensor;" WO 2005/004204 to
Heath, "An electrochemical method and resulting structures for
attaching molecular and biomolecular structures to semiconductor
micro and nanostructures;" WO 2008/030395 to Amori, "Apparatus and
method for quantitative determination of target molecules;" WO
2009/013754 to Haick, "Chemically sensitive field effect
transistors and uses thereof in electronic noise devices;" U.S.
Pat. No. 7,628,959 to Penner, "Hydrogen gas sensor;" U.S. Pat. No.
7,631,540 to Chueh, "Gas sensors with zinc oxide or indium/zinc
mixed oxides and method of detecting NOX gas;" U.S. Pat. No.
7,662,652 to Zhou, "Chemical sensor using semiconducting metal
oxide nanowires;" US 2010/0198521 to Haick, "Chemically sensitive
field effect transistors and uses thereof in electronic noise
devices;" U.S. Pat. No. 7,963,148 to Liu, "Gas sensor made of field
effect transistor made of ZnO nanowires;" Zhou et al, "Silicon
Nanowires as Chemical Sensors," Chem. Phys. Lett. 369 p. 220
(2003); Eliol et al, "Integrated Nanoscale Silicon Sensors Using
Top-Down Fabrication," Appl. Phys. Lett. 83 p. 4613 (2003); Sysoev
et al, "Toward the nanoscopic `electronic nose`: hydrogen vs.
carbon monoxide discrimination with an array of individual metal
oxide nano- and mesowire sensors," Nano Lett. 6(8):1584-8 (2006);
McAlpine et al, "Highly ordered nanowire arrays on plastic
substrates for ultrasensitive flexible chemical sensors," Nature
mater. 6(5) 379-384 (2007); Sysoev et al, "A Gradient Microarray
Electronic Nose Based on Percolating SnO2 Nanowire Sensing
Elements," NANO LETTERS, Vol. 7, No. 10, 3182-3188; McAlpine et al,
"Peptide-Nanowire Hybrid Materials for Selective Sensing of Small
Molecules," Peptide-Nanowire Hybrid Materials for Selective Sensing
of Small Molecules (2008); Engel et al, "Supersensitive Detection
of Explosives by Silicon Nanowire Arrays," Angew. Chem. Int. Ed.,
49, 6830-6835 (2010); U.S. Pat. No. 8,010,591 to Mojarradi et al,
"Four-Gate Transistor Analog Multiplier Circuit;" and Haick et al,
"Electrical Characteristics and Chemical Stability of Non-Oxidized
Methyl-Terminated Silicon Nanowires," J. Am. Chem. Soc. 128,
8990-8991 (2006).
SUMMARY OF THE INVENTION
[0006] An aspect of some embodiments of the invention concerns a
multi-gate field effect transistor (FET) with a conducting channel
that acts like a virtual buried nanowire, whose conductivity is
sensitive to a local concentration of molecules from a gas or
liquid sample adhering to a surface of the FET, and whose
transverse position is controllable by the gates, allowing the FET
to function as a molecular sensor with improved sensitivity.
[0007] There is thus provided, in accordance with an exemplary
embodiment of the invention, a system for sensing molecules in a
gas or liquid sample, comprising: [0008] a) at least one multi-gate
field effect transistor, comprising: [0009] 1) a piece of
semiconductor with an active region extending between a source
region and a drain region, and left and right lateral regions
extending along the active region on different sides; [0010] 2)
left and right lateral gate electrodes that respectively produce an
electric field in the left and right lateral regions, creating a
conducting channel in the active region when appropriate voltages
are applied to them, a position of the conducting channel depending
on the applied voltages; [0011] 3) a sensing surface adjacent to
the active region, that the molecules adhere to, a local
concentration of the adhering molecules near the position of the
conducting channel affecting its conductivity; and [0012] b) a
controller adapted to successively apply different voltages to the
lateral gate electrodes of the transistor, and move the conducting
channel to a plurality of different positions, and at each position
to measure its conductivity.
[0013] Optionally, the sensing surface is coated with a ligand that
binds specifically to the molecules that are being sensed.
[0014] Optionally, the source region and drain region are doped
with dopants of a same sign, and the left and right lateral regions
are doped with dopants of an opposite sign to the source and drain
regions.
[0015] Optionally, the active region is doped with a dopant of the
same sign as the source and drain regions.
[0016] Optionally, the concentration of dopants of the lateral
regions extends into the active region, falling off gradually over
a scale length greater than the width of the conducting
channel.
[0017] There is further provided, according to an exemplary
embodiment of the invention, a method of manufacturing the field
effect transistor in the system according to an embodiment of the
invention, comprising heat treating the transistor under conditions
such that some of the dopants from the left and right lateral
regions diffuse into the active region, reducing an effective width
of the active region by at least 30% at its narrowest point, but
not reducing the effective width to zero at any point.
[0018] Optionally, the active region is narrower than 1 micrometer
between the left and right lateral regions.
[0019] Optionally, the field effect transistor also comprises a
back gate electrode, located in a direction away from the sensing
surface and separated from the active region at least by an
insulator layer, a voltage of the back gate electrode affecting one
or both of an average distance and a range of distance of the
conducting channel from the sensing surface.
[0020] Optionally, the controller is adapted to determine a
concentration of adhering molecules adjacent to each of the
positions of the conducting channels, from the conductivity
measured at each of the positions.
[0021] Optionally, for at least one choice of gate electrode
voltages, the system has a width of the conducting channel and a
distance of the conducting channel from the sensing surface such
that an equilibrium concentration of the adhering molecules can be
determined when a concentration of the molecules in air that the
sensing surface is exposed to is only 100 parts per million.
[0022] In an embodiment of the invention, the at least one field
effect transistor comprises a plurality of field effect
transistors, and the controller is adapted to change the position
of the conductive channel in each transistor and to measure its
conductivity at a plurality of different positions, to find a
greatest concentration of adhering molecules near any of the
positions, for each transistor, and to find an average over the
transistors of the greatest concentrations of adhering gas
molecules.
[0023] Optionally, the system is for use as an electronic nose for
sensing a plurality of different types of molecules, and the at
least one field effect transistor comprises a plurality of field
effect transistors with sensing surfaces having different chemical
properties, causing them to have different relative tendencies for
the different molecules to adhere to them, and the controller is
adapted to change the position of the conducting channel and
determine a concentration of adhering molecules near each of the
positions of the conducting channels, from the conductivity
measured at each of the positions, for each transistor, and to find
the type of molecules present by comparing a pattern of the
concentrations of molecules adhering to each field effect
transistor, to an expected pattern of concentrations of adhering
molecules for each of the types of molecules.
[0024] Optionally, the field effect transistor also comprises a
dielectric layer situated over the active region, and the sensing
surface comprises a surface of the dielectric layer.
[0025] Optionally, the sensing surface of the transistor comprises
an exposed surface of the active region.
[0026] Optionally, the semiconductor comprises silicon, and the
exposed surface of the active region comprises methyl-terminated
silicon.
[0027] Optionally, the sensing surface is adapted for exposure to a
gas sample.
[0028] Alternatively, the system also comprises a reservoir adapted
for holding a liquid sample and for exposing the sensing surface to
the liquid sample.
[0029] There is further provided, in accordance with an exemplary
embodiment of the invention, a method of sensing molecules in a gas
or liquid sample with a multi-gate field effect transistor having a
conductive channel connecting a source region to a drain region, a
position of the conductive channel in a lateral direction
controllable by changing two lateral gate voltages, and the
conductivity of the conducting channel affected by the molecules
adhering to a sensing surface of the transistor at a position near
the conducting channel, the method comprising: [0030] a) exposing
the sensing surface to the gas or liquid sample; [0031] b) changing
the position of the conducting channel in the lateral direction,
and measuring a conductivity of the channel at a plurality of
positions of the channel; and [0032] c) detecting the molecules by
observing a change in conductivity of the conducting channel when
it is in a position such that it passes close to one of the
adhering molecules, or close to a fluctuation in a concentration of
the adhering molecules on the sensing surface.
[0033] Optionally, the multi-gate field effect transistor is a
field effect transistor comprising a back gate electrode that
affects one or both of an average distance and range of distance of
the conducting channel from the sensing surface, the method also
comprising adjusting a voltage of the back gate electrode to
improve a sensitivity of the conductivity of the conducting channel
to the adhering molecules.
[0034] Optionally, changing the two lateral gate voltages affects a
cross-sectional area of the conducting channel, a cross-sectional
shape of the conducting channel, or both, at least partly
independently of the position of the conducting channel in the
lateral direction, as well as affecting the position of the
conducting channel in the lateral direction.
[0035] Optionally, changing a position of the conducting channel in
the lateral direction comprises keeping the two lateral gate
voltages at values such that the conducting channel has a width in
the lateral direction no greater than 50% of a full range of the
positions that the conducting channel can move to in the lateral
direction.
[0036] Optionally, changing a position of the conducting channel in
the lateral direction comprises keeping the two lateral gate
voltages at values such that the conducting channel has a width in
the lateral direction no greater than 200 nanometers.
[0037] Optionally, the sample comprises a gas sample.
[0038] Alternatively, the sample comprises a liquid sample, and
exposing the sensing surface to the liquid sample comprises holding
the liquid sample in a reservoir.
[0039] There is further provided, according to an exemplary
embodiment of the invention, a method of moving a conducting
channel in a multi-gate field effect transistor with an active
region between a source region and a drain region, and at least
lateral gate electrodes that create a depletion region in part of
the active region, the method comprising: [0040] a) setting
voltages of the gate electrodes to create a non-depleted conducting
channel, narrower than the active region, connecting the source and
drain regions through the active region; and [0041] b) changing
voltages of the lateral gate electrodes to move the conducting
channel to different positions in a direction transverse to its
length.
[0042] Unless otherwise defined, all technical and/or scientific
terms used herein have the same meaning as commonly understood by
one of ordinary skill in the art to which the invention pertains.
Although methods and materials similar or equivalent to those
described herein can be used in the practice or testing of
embodiments of the invention, exemplary methods and/or materials
are described below. In case of conflict, the patent specification,
including definitions, will control. In addition, the materials,
methods, and examples are illustrative only and are not intended to
be necessarily limiting.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0043] Some embodiments of the invention are herein described, by
way of example only, with reference to the accompanying drawings.
With specific reference now to the drawings in detail, it is
stressed that the particulars shown are by way of example and for
purposes of illustrative discussion of embodiments of the
invention. In this regard, the description taken with the drawings
makes apparent to those skilled in the art how embodiments of the
invention may be practiced.
[0044] In the drawings:
[0045] FIG. 1 schematically shows a perspective view of a virtual
buried nanowire gas sensor, according to an exemplary embodiment of
the invention;
[0046] FIG. 2 schematically shows a cross section of the sensor in
FIG. 1, perpendicular to the direction of the conductive channel in
the middle of the channel, according to an exemplary embodiment of
the invention;
[0047] FIGS. 3A-3C schematically shows perspective views of a
cross-section of the sensor in FIGS. 1 and 2, not drawn to scale,
showing the conducting channel moved to different lateral positions
by changing the lateral gate voltages, and the response of channel
cross-section when the channel passes close to an adhering
molecule;
[0048] FIGS. 3D-3F schematically show a cross-section of the sensor
in FIGS. 1 and 2, seen from above, not drawn to scale, showing the
conducting channel moved to different lateral positions by changing
the lateral gate voltages;
[0049] FIG. 3G shows a flowchart of a procedure for using the
sensor shown in FIGS. 1 and 2, according to an exemplary embodiment
of the invention;
[0050] FIG. 4A shows a plot of simulation results for effective
channel width, and test results for a shift in source to drain
threshold voltage .DELTA.V.sub.Tf, as a function of lateral gate
voltage, for a virtual buried nanowire gas sensor similar to that
shown in FIGS. 1 and 2, used for specific anti-troponin detection,
in aqueous conditions and with a reference electrode, according to
an exemplary embodiment of the invention;
[0051] FIG. 4B shows contour plots of the carrier density in the
active region, for different values of gate voltage, from a
simulation of the sensor used for FIG. 4A;
[0052] FIG. 5 is a plot of simulation results for conducting
channel width W.sub.eff, and shift in average potential of the
upper surface of the active region, due to a given charge placed on
top of the gate dielectric above the center of the conducting
channel, as a function of lateral gate voltage V.sub.Gj, for a
virtual buried nanowire gas sensor similar to that shown in FIGS. 1
and 2, according to an exemplary embodiment of the invention;
[0053] FIG. 6 is a plot of simulation results for the carrier
density as a function of x, near the surface of the active region,
for a carrier channel centered at five different lateral positions
by changing the left and right lateral gate voltages, according to
an exemplary embodiment of the invention;
[0054] FIGS. 7A and 7B are simulation results showing contour plots
of the carrier density in a cross-section of the active region, for
two different values of the lateral gate voltages, showing how the
channel width can be adjusted, for a virtual buried nanowire gas
sensor similar to that shown in FIGS. 1 and 2, according to an
exemplary embodiment of the invention; and
[0055] FIGS. 8A and 8B are simulation results of contour plots
similar to those in FIGS. 7A and 7B, but for a virtual buried
nanowire gas sensor in which heat treatment has been used to cause
dopants from the lateral gate regions to move into the active
region, making the active region narrower, and allowing the
conductive channel to be narrower, according to an exemplary
embodiment of the invention.
DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION
[0056] The present invention, in some embodiments thereof, relates
to a semiconductor chemical sensor and, more particularly, but not
exclusively, to a gas sensor based on a field effect
transistor.
[0057] An aspect of some exemplary embodiments of the invention
concerns a multi-gate field effect transistor (FET) used for
sensing molecules in a gas or liquid sample. The molecules adhere
to an exposed surface of the FET, and affect the conductivity of a
conducting channel going through the active region, which acts like
a virtual buried nanowire, connecting a source region to a drain
region. Lateral gate electrodes are optionally used to control a
position of the conducting channel in a direction transverse to the
length of the conducting channel. By measuring the conductivity of
the conducting channel as its position is varied in the transverse
direction, fluctuations in the concentration of adhering molecules
can be detected, due for example to the small number of molecules,
potentially making the sensor much more sensitive than a FET with a
conducting channel that does not change its position, or to a
molecular sensor using a real nanowire made of a different material
and buried in the silicon at a fixed position. For example, in some
embodiments of the invention, the sensor produces a response signal
that depends on the conductivity of the conducting channel at a
position of the conducting channel where the concentration of
adhering molecules is greatest. Such a multi-gate FET molecular
sensor using a virtual buried nanowire is also potentially much
cheaper to mass produce than a conventional nanowire molecular
sensor using a real buried nanowire. For example, it could be
produced with conventional high volume, low cost CMOS manufacturing
methods, since, optionally, no low-dimensional design rules are
needed.
[0058] Other potential advantages of a virtual buried nanowire
molecular sensor, over conventional nanowire molecular sensors,
include increased SNR, enhanced gain, enhanced resolution, and
faster device characterization and development. For conventional
buried nanowire based sensors the dimensions of the nanowire need
to be optimized in accordance with the organic system to be
detected. This implies a lengthy characterization and development
phase where nanowires of various compositions and dimensions need
to be tested. In the virtual buried nanowire approach, the device
is optionally fabricated only once. The optimization of the device
for use in detecting a specific analyte is optionally accomplished
by adjusting the gate voltages to produce virtual nanowires of
different cross-sectional areas and shapes, and testing them.
[0059] The virtual buried nanowire molecular sensor is optimized,
in different embodiments of the invention, to sense different
analytes, for example, for medical diagnostic applications, for
environmental applications, for military applications, or for other
applications.
[0060] An aspect of some embodiments of the invention concerns a
multi-gate FET with a virtual buried nanowire, in which the
conducting channel is made narrower by increasing a dopant
concentration in the active region, while using a heat treatment to
cause dopants of the opposite sign to diffuse from the lateral gate
regions part way into the active region from the sides. This makes
the active region effectively narrower, while avoiding breakdown at
the PN junctions between the lateral regions and the active region,
and potentially with little or no reduction in the carrier density
in the conducting channel. If the FET is used as a molecular
sensor, with the conducting channel scanned laterally across the
active region, the narrower channel potentially gives the sensor
increased sensitivity, resolution, and/or SNR.
[0061] U.S. Pat. No. 8,007,727, to Shalev et al, "Virtual
semiconductor nanowire, and method of using same," describes a
multiple-gate field-effect transistor that includes a fluid in a
top gate, two lateral gates, and a bottom gate. The multiple-gate
field-effect transistor also includes a patterned depletion zone
and a virtual depletion zone that has a lesser width than the
patterned depletion zone. The virtual depletion zone width creates
a virtual semiconductor nanowire that is lesser in width than the
patterned depletion zone. This patent has a common inventor with
the present application, but a different assignee.
[0062] Before explaining at least one embodiment of the invention
in detail, it is to be understood that the invention is not
necessarily limited in its application to the details of
construction and the arrangement of the components and/or methods
set forth in the following description and/or illustrated in the
drawings and/or the Examples. The invention is capable of other
embodiments or of being practiced or carried out in various
ways.
[0063] Referring now to the drawings, FIG. 1 shows an exemplary
multi-gate FET 100, comprising a semiconductor layer built on top
of an insulator layer 102, for example a buried oxide (BOX) layer
of silicon oxide, optionally on top of a substrate 104, optionally
made of the same material as the semiconductor layer, for example
silicon. The semiconductor layer over the insulator layer is
sometimes referred to herein as an SOI (silicon on insulator)
layer, although other semiconductor materials are used instead of
silicon in some embodiments of the invention, and materials other
than silicon oxide are optionally used for the insulator layer. It
should be understood that terms such as "on top of," "above," and
"over," as used herein, refer to a direction that is shown as
vertical in the drawings, but need not be literally vertical with
respect to gravity; generally the device may be oriented in any
direction with respect to gravity, without affecting its
operation.
[0064] The semiconductor layer comprises a source region 106 at one
end, and a drain region 108 at the other end, both doped with an
implant of the same charge, for example an N implant. A right
lateral gate region 110 and a left lateral gate region 112 are both
doped with an implant of an opposite charge to the implant of the
source and drain regions, for example a P implant. Alternatively,
the source and drain regions are doped with a P implant and the
gate regions are doped with an N implant. The rest of the
semiconductor layer comprises a portion 114 adjacent to the source
region, a portion 116 adjacent to the drain region, and a narrower
active region 118 connecting the source region to the drain region.
Portions 114 and 116, and active region 118 are optionally doped
with an implant of the same sign charge as the implant of the
source and drain regions, but are less strongly doped than the
source and drain region. A source electrode 120 is connected to
source region 106, a drain electrode 122 is connected to drain
region 108, a right lateral gate electrode 124 is connected to
right gate region 110, and a left lateral gate electrode 126 is
connected to left gate region 112. Connectors 128 allow the
electrodes to be connected to an external circuit which can control
the voltage on each of the electrodes, and can measure the current
between the source and drain electrodes.
[0065] Optionally, there is a back gate electrode, not shown in
FIG. 1, attached to the bottom of substrate 104, or to the bottom
of insulator layer 102 if there is no substrate 104 beneath the
insulator layer. The presence of insulator layer 102 between the
back gate electrode and the other electrodes makes it possible for
the back gate electrode to affect the electric field and hence the
carrier distribution in the active region, without drawing any
current. Substrate layer 104 may be present as a result of the
method of manufacture, in some methods of manufacturing FET
100.
[0066] Optionally, there is a gate dielectric layer, not shown in
FIG. 1, above active region 118. The gate dielectric is optionally
made of silicon oxide. Alternatively, other materials are used for
the gate dielectric, including for example any of HfO.sub.2,
Si.sub.3N.sub.4, Al.sub.2O.sub.3, and Ta.sub.2O.sub.5.
[0067] It should be understood that the FET need not have the
rectilinear geometry shown in FIG. 1, with the active region
oriented along the y-direction, the lateral gate regions
surrounding it in the x-direction, and the different layers
arranged in the z-direction. Instead, the FET may be curved or
twisted in any way, for example with the active region C-shaped, or
S-shaped, or with the layers having surface curvature, as long as
certain features are present, for example a path through the active
region connects the source and drain regions, and the lateral gate
regions are adjacent to the active region on its sides. However, a
rectilinear geometry potentially makes the FET easier to
manufacture by conventional manufacturing methods for semiconductor
devices. The gate electrodes, which generally do not have
substantial current running through them in normal operation, need
not be in physical contact with the semiconductor layer or
insulator layer, but could be separated from them by an air gap,
although for reasons of mechanical strength it is potentially
advantageous to have any electrodes in direct contact with
semiconductor or insulator.
[0068] FIG. 2 shows a cross-section 200 of FET 100, perpendicular
to the direction from the source to the sink, and half way between
the source and the sink. A dielectric layer 202 optionally covers
active region 118, and a back gate electrode 204 is optionally
attached to the bottom of substrate 104. A conducting channel 206,
connecting the source and drain regions through active region 118,
is created by applying appropriate voltages to the gate electrodes,
a voltage V.sub.Gj1 to the left gate electrode, a voltage V.sub.Gj2
to the right gate electrode, and optionally a voltage V.sub.Gb to
the back electrode. These voltages are measured, for example,
relative to ground, and typically the source electrode is grounded.
The voltage on the lateral gate electrodes creates an electric
field in the semiconductor, which creates a depletion region,
without charge carriers, at the interface of the lateral regions
with the active region, extending into the active region, while the
voltage on the back electrode creates a depletion region at the
interface between the insulator layer and the active region,
extending into the active region. For appropriate values of the
gate voltages, the depletion region covers much of the active
region, leaving only the relatively narrow undepleted conducting
channel 206. When a voltage V.sub.SD is then applied between the
source and the drain electrodes, a current flows between them which
depends on the cross-sectional area of the conducting channel.
[0069] The FET functions as a gas sensor because the
cross-sectional area of the conducting channel is sensitive to the
charge of gas molecules adhering to the surface of the dielectric
layer, sometimes referred to herein as a sensing surface. It should
be understood that, although we describe herein embodiments of the
invention that are used as gas sensors, other embodiments of the
invention are used to detect molecules in a liquid sample, for
example by adding a reservoir for holding a liquid sample to the
top of the sensing surface. The charge of these adhering gas
molecules changes the potential of the surface of dielectric layer
202, as if there were another gate electrode there, and this
changes the cross-sectional area of the conducting channel. For
this reason dielectric layer 202 is sometimes referred to herein as
a gate dielectric, although in the embodiments described, there is
optionally no physical gate electrode on top of the active region.
Typically, for a given set of gate voltages, the current between
the source and the drain is essentially zero up to a threshold
voltage between the source and the drain, and increases rapidly
above the threshold voltage. Optionally, the voltage between the
source and the drain is set just below the threshold voltage in the
absence of adhering gas molecules on the gate dielectric, so that
even a small decrease in the threshold voltage, caused by a small
number of adhering gas molecules, can greatly increase the current
between the source and the drain, making the FET potentially a very
sensitive detector of gas molecules. Typically, the threshold
voltage between the source and the drain is between 10 mV and 100
mV, and the source to drain voltage is optionally kept at such low
levels, much less than the lateral and back gate electrode
voltages, which are typically a few volts.
Changing Lateral Position of Conducting Channel
[0070] FIGS. 3A-3C illustrate how applying a different voltage on
the left and right lateral gate electrodes optionally controls the
lateral position of the conducting channel, and how changing the
lateral position of the conducting channel is optionally used to
increase the sensitivity of FET 100 as a gas sensor. In FIGS.
3A-3C, there is a relatively low concentration of gas above gate
dielectric 202, and only a few gas molecules adhere to the surface.
The left and right gate electrodes are optionally kept at voltages
such that the conducting channel is very narrow, for example
narrower than the width of the active region by a factor of 5, 10,
20, 50, or a lower, higher or intermediate value, and at a given
lateral position of the conducting channel, there is less than one
molecule, on average, adhering directly above any part of the
conducting channel. In absolute dimensions, the width of the
conducting channel is, for example, 500 nm, 200 nm, 100 nm, 50 nm,
20 nm, 10 nm, 5 nm, or a lower, higher, or intermediate value. As
used herein, the width of the conducting channel means the
full-width, in the y direction, at half-maximum of the carrier
density.
[0071] As the lateral position of the conducting channel is scanned
from left to right, by changing the voltage of the left lateral
electrode relative to the voltage of the right lateral electrode,
the cross-sectional area of the channel increases whenever it
passes close to an adsorbed molecule. For example, in FIG. 3A, the
channel passes under gas molecule 302, and its cross-sectional area
is relatively large. In FIG. 3B, the channel does not pass close to
any gas molecules, and its cross-sectional area is smaller. In FIG.
3C, the channel passes close to gas molecule 304, and its
cross-sectional area increases again. In other embodiments of the
invention, the current decreases, instead of increasing, when the
conducting channel passes near an adhering gas molecule. By
scanning the conducting channel across the active region, and
measuring the current between the source and the drain at each
lateral position of the channel, an accurate determination may be
made of the density of gas molecules adhering to the gate
dielectric. If the current between the source and drain were only
measured at a fixed lateral position of the conducting channel, it
would not be possible to do this, because the current would likely
either be very small, corresponding to no adhering gas molecules,
if the channel did not happen to pass close to any gas molecules,
or very large, corresponding to a much higher density of adhering
gas molecules than are actually present, if the channel happened to
pass close to one of the gas molecules.
[0072] FIGS. 3D, 3E and 3F respectively show a cross-section of the
semiconductor layer as seen from above, at the surface of the
semiconductor layer or at a depth below the surface where
conducting channel 206 is located, at three different positions of
the conducting channel. Active region 118 and conducting channel
206 are not necessarily drawn to scale, but are shown wider than
they typically are relative to the dimensions of the source, drain
and lateral gate regions, so that the change in position of the
conducting channel may be clearly seen. It should be noted that, as
long as the FET is operated with source to drain voltage much less
than the voltage between the source and the lateral and back gate
electrodes, for example less by a factor of 10 or 100, then the
conducting channel will generally be very uniform in cross-section
along the active region between the lateral regions, as shown in
FIGS. 3D-3F, although the conducting channel may fan out beyond the
ends of the lateral regions.
[0073] In order to scan the conducting channel laterally across the
active region, the lateral gate voltages, and optionally the back
gate voltage, are controlled by a controller 306, which controls a
power supply 308 that provides the left lateral gate voltage, a
power supply 310 that provides the right lateral gate voltage, and
optionally a power supply 312 that provides the back gate voltage,
all relative to ground which is, for example, connected to the
source electrode. Controller 306 optionally adjusts the gate
voltages to keep the depth and cross-sectional dimensions of the
conducting channel substantially constant as its position moves
laterally across the active region, for example by running software
that implements a control algorithm, or by using an electronic
circuit that produces the right relationship between the gate
voltages. The relationship between the different gate voltages that
will accomplish this is found, for example, by simulations, as
described below in FIG. 6, or by testing the FET in the absence of
adhering gas molecules. The differences in source to drain current,
at different lateral positions of the conducting channel, are then
due primarily to differences in the concentration of gas molecules
adhering above the conducting channel, and optionally the
sensitivity of the sensor is optimized at all positions of the
conducting channel.
[0074] Controller 306, or a different controller, optionally
calculates an average density of adhering gas molecules on the gate
dielectric surface, from the measured source to drain current as a
function of lateral position of the conducting channel. For
example, the average density is proportional to an average shift in
threshold voltage, for all positions of the conducting channel,
relative to the threshold voltage when there are no adhering gas
molecules, and the constant of proportionality is calibrated using
a sample with known concentration of the gas molecules.
Alternatively, the number of adhering gas molecules is counted, by
counting the number of times that the current in the conducting
channel has a significant rise and fall as the conducting channel
is scanned across the width of the active region, indicating that
the channel has passed by one adhering gas molecule, and the
density is found by dividing the number of adhering molecules by
the surface area of the active region. The average shift in
threshold voltage may produce a more accurate measure of the
density of adhering gas molecules when the density is relatively
high, or even when the density is relatively low if the change in
threshold voltage as a function of channel position always has
about the same width and height whenever the channel passes an
adhering molecule. Counting the number of adhering molecules may
produce more accurate results if there are relatively few adhering
molecules, so that the conducting channel will usually not pass
close to more than one adhering molecule at a time.
[0075] FIG. 3G shows a flowchart 320 for a procedure used to
measure the density of adhering gas molecules, according to an
exemplary embodiment of the invention. At 322, the source to drain
voltage is optionally set just below the threshold value in the
absence of adhering gas molecules, so that the presence of an
adhering gas molecule adjacent to the conducting channel will
increase the source to drain current, by decreasing the source to
drain voltage below the threshold voltage. In an embodiment where
an adhering gas molecule decreases the current at a given voltage,
the source to drain voltage is instead optionally set just above
the threshold voltage, so that an adhering gas molecule will
decrease the current. At 324 the lateral gate voltages, and
optionally the back gate voltage, are set to a value for which the
conducting channel will be at an initial position in the active
region, for example all the way to one side of the active region in
the lateral direction, or all the way at the beginning of a range
of positions over which the conducting channel is to be scanned. At
326, while or after the gate dielectric is exposed to a gas sample,
the source to drain current is measured and recorded. In another
embodiment of the invention, instead of setting the source to drain
voltage at a constant value at 322, and measuring changes in
current caused by adhering gas molecules at 326, the source to
drain current is kept at a constant value, for example at the
maximum slope of current as a function of voltage just above the
threshold voltage, and changes in voltage due to adhering gas
molecules are measured. In effect, this is similar to measuring
changes in the threshold voltage. Alternatively, a function of
current and voltage is kept constant, and changes in a different
function of current and voltage is measured.
[0076] At 328, if this scan is not done, then at 330, the lateral
gate voltages, and optionally the back gate voltage, are adjusted
to move the conducting channel to the next position. Optionally,
this is done in such a way that the width and depth of the channel
are not changed, or are changed very little, as described above.
The position of the conducting channel need not change
monotonically in time, but can jump around. However, it may be
simplest, in interpreting the data and in controlling the voltages
of the gate electrodes, to have the position of the conducting
channel go sequentially from one side of the active region to the
other side during a scan, making measurements at frequent
intervals. After the gate voltages have been set to the new values
at 330, moving the conducting channel to the new position, the
source to drain current is measured and recorded again at 326. This
loop is continued until the scan is done at 328, for example
because the position of the conducting channel is all the way on
the other side of the active region from what it was initially, or
is all the way on the other side of the range of positions over
which the conducting channel is being scanned.
[0077] When the scan is done, the number or density of adhering gas
molecules is found at 332, from the data recorded at 326, for
example source to drain current as a function of channel position
at constant voltage, or the threshold voltage as a function of
channel position, using any of the methods described above for
finding the number or density of adhering gas molecules. At 334, if
more scans are to be made, then the gate voltages are returned to
the values that will put the conducting channel at its initial
position, at 324, and a new scan is made. Optionally, scans are
made repeatedly, while the gate dielectric is exposed to gas
molecules, and the density of adhering gas molecules as a function
of time. When all scans are done, at 334, the density of adhering
gas molecules as a function of time is optionally supplied as
output to a user, at 336. Typically, the density of adhering gas
molecules will initially increase linearly with time, when the
sensor is first exposed to the gas molecules, and will then
saturate, as the gate dielectric becomes saturated with gas
molecules, or as the rate of adherence of gas molecules is balanced
by a rate of loss of adhering gas molecules from the surface. The
concentration of gas molecules in a sample may be inferred from the
initial rate of rise, and/or from the saturation level.
[0078] To estimate how much increase in sensitivity can be achieved
by varying the position of the conducting channel, note that in
general the sensitivity may become greater the narrower the
effective width of the conducting channel, and assume that the
noise level is low enough so that, for a conducting channel width
of W.sub.c, a single adhering gas molecule can be detected if it is
within W.sub.c/2 of the conducting channel. Then, if the width of
the active region is .DELTA.x, on average the gas molecules could
be detected, at a given position of the conducting channel, only if
at least .DELTA.x/W.sub.c gas molecules were adhering to the upper
surface of the FET. If the conducting channel were scanned across
the active region, and the greatest response at any position were
measured, then in principle even a single adhering gas molecule
could be detected, an increase in sensitivity of .DELTA.x/W.sub.c,
which could be, for example, a factor of 5, or 10, or 20, or 30, or
more. Although the greatest potential increase in sensitivity may
occur, due to scanning the conducting channel, if a single adhering
gas molecule could be detected directly over the conducting
channel, some increase is sensitivity will occur even if, for
example, a minimum of 2 or 3 adhering molecules are needed for
detection at a given position of the channel, since there will be
large fluctuations in the number of adhering molecules above the
conducting channel, due to Poisson statistics, if the average
number at a given position of the channel is a relatively small
number such as 2 or 3. Relatively less increase in sensitivity due
to scanning the conducting channel may occur, as the minimum number
of adhering molecules, needed for detection at a given position,
increases, and as .DELTA.x/W.sub.c decreases.
Gas Molecules Adhering Directly to Semiconductor with No Gate
Dielectric
[0079] In some embodiments of the invention, the sensing surface
that the gas molecules adhere to is, at least in part, an upper
surface of the active region itself, and there need not be any
dielectric layer over the active region. Optionally, in those
embodiments, the semiconductor comprises silicon, and the upper
surface of the active region, that the gas molecules adhere to,
comprises methyl-terminated silicon, as described, for conventional
silicon nanowires, by Haick et al, J. Am. Chem. Soc. 128, 8990-8991
(2006), cited above. Alternatively, the silicon is coated with a
polar monolayer of organic molecules, as described by Paska and
Haick, "Controlling properties of field effect transistors by
intermolecular cross-linking of molecular dipoles," Appl. Phys.
Lett. 95, 233103 (2009), and "Controlling surface energetics of
silicon by intermolecular interactions between parallel
self-assembled molecular dipoles," J. Chem. Phys. C 113, 1993-1997
(2009), by the same authors. Alternatively, the silicon is coated
with dense hydrophobic organic hexyltrichlorosilane monolayers,
that are especially suitable for nonpolar molecules, as described
by Paska et al, "Enhanced sensing of nonpolar volatile organic
compounds by silicon nanowire field effect transistors," ACS Nano
5, 5620-5626 (2011). Paska et al also describe other suitable
coatings for this purpose.
[0080] Having the gas molecules adhere directly to the upper
surface of the active region has the potential advantage of
improving the sensitivity of the gas sensor. Using a dielectric
layer, in particular a silicon dioxide dielectric layer, above the
active region, has the potential advantage that the design is
closer to the design of a conventional FET, and it may be possible
to use more conventional manufacturing methods. Also, the
technology of chemically modifying dielectric surfaces, for binding
to specific gas molecules, may be more advanced than the technology
of chemically modifying semiconductor surfaces, potentially
allowing more flexibility in choosing which gas molecules are to be
detected, for a sensor using a dielectric layer. But it should be
understood that any of the devices and methods shown in the
drawings could also be implemented without a dielectric layer above
the active region, and having the gas molecules adhere directly to
an upper surface of the active region.
Controlling Conducting Channel Dimensions
[0081] Optionally, the lateral gate electrodes also control a width
of the channel in the transverse direction, at least partly
independently of the position of the channel. This can be done if
the voltage of the left and right lateral gate electrodes is
controlled independently. Optionally the voltage of the back gate
electrode, possibly together with the voltages of the lateral gate
electrodes, controls a distance of the conducting channel from the
sensing surface where the gas molecules adhere, and/or a range of
distances of the conducting channel from the sensing surface. For
the geometry shown in FIG. 2, this means controlling a vertical
position and/or a vertical width of the conducting channel. This
has the potential advantage that the conducting channel can be at a
vertical position that is reasonably close to the top dielectric
layer, for enhanced sensitivity, but not so close to the top
dielectric layer that the sensor suffers from noise generated by
noise centers at the interface between the semiconductor and the
top dielectric layer. Optionally, the transverse and vertical width
of the conducting channel, and/or the vertical position of the
conducting channel, are set at values that give the gas sensor
better sensitivity than it would have for other values. For
example, the width of the channel, transversely and/or vertically,
at its narrowest point, or on average over the length of the active
region, is greater than 200 nanometers, or between 200 and 100
nanometers, or between 100 and 30 nanometers, or between 30 and 10
nanometers, or between 10 and 3 nanometers, or less than 3
nanometers, or more than 50% of the width of the active region in
the transverse direction, or between 50% and 30% of the width, or
between 20% and 10% of the width, or between 10% and 5% of the
width, or less than 5% of the width. Optionally, the width of the
active region in the transverse direction, at its narrowest point
or on average over its length, is greater than 1 micrometer, or
between 1 micrometer and 500 nanometers, or between 500 nanometers
and 200 nanometers, or between 200 and 100 nanometers, or less than
100 nanometers. Optionally, the conducting channel is located
vertically close to the top of the active region, i.e. close to the
dielectric layer in the case where there is a dielectric layer, for
example a distance of 200 nanometers, 100 nanometers, 30
nanometers, 10 nanometers, or 3 nanometers from the top of the
active layer, or a greater, smaller, or intermediate distance, or a
distance of 50%, 30%, 20%, 10% or 5% of the vertical thickness of
the active region, from the top of the active region, or a greater,
smaller, or intermediate distance.
[0082] It should be understood that if the active region is too
wide, then the voltage that is applied to the lateral gate
electrodes, in order to create a conducting channel of a given
width, may be greater than the breakdown voltage of the PN
junctions between the lateral gate regions and the active region.
Making the conducting channel wider may make it less sensitive to
adhering gas molecules. Making the conducting channel too narrow
may result in it not having any carriers in it, on average, at a
given time. The number of carriers can be increased by increasing
the dopant concentration in the active region, but if the dopant
concentration is too high, then the breakdown voltage of the PN
junctions may decrease, and breakdown may occur at the voltage
applied to the lateral gate electrodes. Making the active region
longer may make the conducting channel more sensitive to adhering
molecules, since a molecule adhering anywhere along the length of
the conducting channel may significantly affect the conductivity.
But increasing the length of the active region may also increase
the threshold voltage between the source and drain, and if the
voltage between the source and drain is not small enough relative
to the lateral gate voltage, then the conducting channel may not be
uniform in width along its length, and it may be less sensitive to
adhering molecules. The dimensions and dopant concentrations given
in the "Examples" section below represent a set of parameters that
has been found to work well, both experimentally and according to
simulations.
Embodiment with Narrower Active Region and Conducting Channel
[0083] In some embodiments of the invention, the FET is heat
treated, causing dopants from the left and right lateral gate
regions to diffuse part way into the active region from the sides,
giving a portion of the active region, adjacent to the lateral gate
regions, a net dopant concentration of the same sign as the lateral
gate regions, and opposite to the rest of the active region. This
diffusion of dopants in effect causes the lateral gate regions to
extend part way into the active region, making the active region
narrower in the lateral direction, and allowing the conducting
channel to be narrower, as will be described below in the
"Examples" section. The effective width of the active region is
defined herein as the width of the part of the original active
region where the net dopant concentration (the P dopant
concentration minus the N dopant concentration) has the same sign
as the net dopant concentration had originally in the active
region, before the heat treatment, which is opposite to the sign of
the net dopant concentration in the lateral gate regions.
Optionally, the heat treatment reduces the effective width of the
active region at its narrowest point by at least 20%, or at least
30%, or at least 40%, or at least 50%, or at least 60%, or at least
70%, or a greater than 70%. Optionally, the heat treatment is done
at a temperature for which the diffusion rate of the lateral gate
region dopant in the semiconductor, for example boron as a dopant
in silicon, is the square of a desired diffusion distance divided
by a desired time of the heat treatment. The desired time is, for
example, less than 15 seconds, or between 15 and 30 seconds, or
between 30 and 60 seconds, or between 60 and 90 seconds, or between
90 and 150 seconds, or between 150 and 300 seconds, or more than
300 seconds. The desired diffusion distance is, for example, less
than 5% of the width of the active region, or between 5% and 10% of
the width of the active region, or between 10% and 20%, or between
20% and 30%, or between 30% and 40%, or between 40% and 50%, or
more than 50% of the width of the active region.
[0084] A potential advantage of reducing the width of the active
region by using such a heat treatment, rather than making the
active region narrower to begin with, even if that is possible with
the lithography used, is that the net concentration of dopants
changes more gradually between the lateral gate regions and the
active region, making breakdown less likely at the PN junctions
between the lateral gate regions and the active region. In some
embodiments of the invention, the active region is doped at a
higher concentration originally, so that, after the heat treatment,
it still has a net dopant concentration at its center that is at a
high enough level to produce a desired carrier density in the
conducting channel, for example high enough to produce a carrier
density greater than 10.sup.18 cm.sup.-3, or greater than
5.times.10.sup.17 or 3.times.10.sup.17 or 2.times.10.sup.17 or
1.times.10.sup.17 cm.sup.-3, even when the conducting channel is
very narrow, for example narrower than 50 nm or 30 nm or 20 nm or
10 nm.
[0085] In an exemplary embodiment of the invention, the SOI layer
is silicon, the lateral gate regions are doped with boron, the
width of the active region is 400 nm, and the heat treatment is a
temperature of 1050.degree. C., and lasts for 75 seconds, causing
boron from the lateral gate regions to diffuse far enough into the
active region to reduce the effective width of the active region to
only 90 nm at its narrowest point, and 130 nm at the top, adjacent
to the gate dielectric. Optionally, the heat treatment is at a
temperature below 900.degree. C., or between 900.degree. and
1000.degree. C., or between 1000.degree. C. and 1100.degree. C., or
between 1100.degree. and 1200.degree. C., or above 1200.degree. C.
Optionally, the heat treatment lasts for less than 15 seconds, or
between 15 and 30 seconds, or between 30 and 60 seconds, or between
60 and 90 seconds, or between 90 and 150 seconds, between 150 and
300 seconds, or more than 300 seconds. To achieve a given amount of
diffusion of dopants, less time may be needed if the temperature is
higher, and the temperature and time of the heat treatment may be
very different depending on the dopant used, since different
dopants may diffuse at very different rates at a given temperature.
Using a dopant that diffuses more easily has the potential
advantage that the heat treatment may be less expensive because it
takes less time and a lower temperature may be used. Using a dopant
that diffuses less easily has the potential advantage that it may
be easier to control the diffusion and to get repeatable results.
Using a longer heat treatment at a lower temperature may also make
the process more controllable and repeatable. But if the
temperature is too low, or if the dopant has too low a diffusion
rate, the time required to achieve a given degree of diffusion may
be impracticably long.
Sensitivity of Gas Sensor
[0086] Optionally, the conducting channel is controlled to be
narrow enough, and close enough to the dielectric layer, but not
too close, so that the gas sensor has a sensitivity to gas
molecules in air that the dielectric layer is exposed to,
sufficiently high so that the sensor can detect less than 100 parts
per million (ppm) of the gas, after exposure to the air for a long
enough time so that the concentration of adhering gas molecules
reaches an equilibrium, for example for at least several seconds,
or at least several tens of seconds. Optionally, the sensitivity is
sufficiently high to detect less than 30 ppm of the gas, or less
than 10 ppm, or less than 3 ppm, or less than 1 ppm, or less than
300 parts per billion (ppb), or less than 100 ppb, or less than 30
ppb, or less than 10 ppb, or less than 3 ppb, or less than 1 ppb,
or less than 0.3 ppb, or less than 0.1 ppb.
Chemical Treatment of Gate Dielectric
[0087] Optionally, the gate dielectric is chemically treated, for
example, a SiO.sub.2 gate dielectric is modified with APTMS, or
with AUTES, or in other ways. Optionally, the gate dielectric is
modified by coating it with a ligand, so that it binds specifically
to the gas molecules being sensed, in a "lock and key"
configuration. Alternatively, the gate dielectric is chemically
treated with a ligand that does not bind only to the gas molecules
being sensed. For example, the ligand also binds to one or more
other gas molecules that are potentially present in an environment
where the sensor is designed to be used. In some embodiments of the
invention, referred to sometimes as an electronic nose or "e-nose,"
an array of FETs is used, with the gate dielectrics of the
different FETs having different chemical treatments, and different
types of gas molecules have different relative tendencies to bind
to different FETs in the array, and/or different FETs in the array
have different sensitivities to one type of molecule, even if that
is the only type of molecule that the sensor is designed to detect.
The type or types of gas present is then optionally determined from
the signature of the response it produces from each of the FETs in
the array, for example using an algorithm.
Use of Other Materials
[0088] The FET may use any of a variety of semiconductors for the
active region, and any of a variety of dielectric materials for the
gate dielectric, and for the insulator layer between the active
region and back gate electrode if there is one. For convenience,
the semiconductor may be referred to herein as "silicon" and the
dielectric material may be referred to herein as "oxide," for
example "gate oxide" or "buried oxide," or "silicon oxide," but it
should be understood that other suitable materials may be used
instead.
Fabrication Method
[0089] An exemplary method of fabrication of FET 100 begins with a
silicon-on-insulator (SOI) wafer. A silicon island is optionally
shaped, with the silicon around the island etched completely away
until the buried oxide (BOX) is reached, as may be seen in FIGS. 1
and 2, where the silicon island is the semiconductor layer. In this
method of fabrication, optionally there is no silicon substrate
layer 104, but the insulator layer is the substrate. Alternatively,
the silicon island may be grown, for example as polysilicon, on the
insulator side of an SOI wafer, leaving the silicon layer as
substrate 104 beneath the insulator layer, as in FIGS. 1 and 2. The
silicon island has a length L along an axis shown as the y-axis in
FIG. 1, and a width W along a lateral axis shown as the x-axis in
FIG. 1. In an exemplary embodiment of the invention, a critical
dimension of the device is the distance between the two lateral
gate regions. This distance could is optionally defined with g-line
lithography, i-line lithography or a smaller wavelength. In an
exemplary embodiment of the invention, source and drain regions are
created via implants, for doping of the silicon, on each end of the
silicon island to allow for a conducting channel going between
them, parallel to the y-axis. The left and right lateral gate
regions are defined via implants on each side of the active region,
including the conducting channel, in the x-direction. Optionally,
the implants of the lateral gates are of opposite sign from that of
the source/drain implants, i.e. if one of them is P then the other
one is N. The active region is optionally implanted with the same
species as the source/drain implants, i.e. both of them are P or
both of them are N. In an exemplary embodiment of the invention,
metal contacts are then created for the source and drain and the
lateral gates, and optionally the back gate. Optionally,
inter-layer-dielectric (ILD) and passivation are incorporated in
order to isolate the contacts from the gas sample. A gate
dielectric layer is optionally added on top of the conducting
channel.
Depth of Conducting Channel
[0090] Nanowires inherently suffer from surface states. These
surface states may entail degradation in sensor performance in
terms of gain and SNR. In a virtual buried nanowire device, the
conducting channel is optionally removed from the noise centers at
the Si/SiO.sub.2 interface by using the gate voltages to adjust the
depth of the conducting channel, and adjusting the depth of the
implants in the semiconductor, which potentially achieves greater
gain and SNR for the sensor.
[0091] As used herein the term "about" refers to .+-.10%.
[0092] The terms "comprises", "comprising", "includes",
"including", "having" and their conjugates mean "including but not
limited to".
[0093] The term "consisting of" means "including and limited
to".
[0094] The term "consisting essentially of" means that the
composition, method or structure may include additional
ingredients, steps and/or parts, but only if the additional
ingredients, steps and/or parts do not materially alter the basic
and novel characteristics of the claimed composition, method or
structure.
[0095] As used herein, the singular form "a", "an" and "the"
include plural references unless the context clearly dictates
otherwise. For example, the term "a compound" or "at least one
compound" may include a plurality of compounds, including mixtures
thereof.
[0096] Throughout this application, various embodiments of this
invention may be presented in a range format. It should be
understood that the description in range format is merely for
convenience and brevity and should not be construed as an
inflexible limitation on the scope of the invention. Accordingly,
the description of a range should be considered to have
specifically disclosed all the possible subranges as well as
individual numerical values within that range. For example,
description of a range such as from 1 to 6 should be considered to
have specifically disclosed subranges such as from 1 to 3, from 1
to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as
well as individual numbers within that range, for example, 1, 2, 3,
4, 5, and 6. This applies regardless of the breadth of the
range.
[0097] Whenever a numerical range is indicated herein, it is meant
to include any cited numeral (fractional or integral) within the
indicated range. The phrases "ranging/ranges between" a first
indicate number and a second indicate number and "ranging/ranges
from" a first indicate number "to" a second indicate number are
used herein interchangeably and are meant to include the first and
second indicated numbers and all the fractional and integral
numerals therebetween.
[0098] It is appreciated that certain features of the invention,
which are, for clarity, described in the context of separate
embodiments, may also be provided in combination in a single
embodiment. Conversely, various features of the invention, which
are, for brevity, described in the context of a single embodiment,
may also be provided separately or in any suitable subcombination
or as suitable in any other described embodiment of the invention.
Certain features described in the context of various embodiments
are not to be considered essential features of those embodiments,
unless the embodiment is inoperative without those elements.
[0099] Various embodiments and aspects of the present invention as
delineated hereinabove and as claimed in the claims section below
find experimental and calculational support in the following
examples.
EXAMPLES
[0100] Reference is now made to the following examples, which
together with the above descriptions illustrate some embodiments of
the invention in a non limiting fashion.
[0101] The general silicon configuration of the virtual buried
nanowire gas sensor was demonstrated experimentally for biological
detection. In this experiment, the thickness of the SOI layer was
260 nm, with boron doping of 1.6.times.10.sup.14 cm.sup.3, giving
it a resistivity of 13 to 22 .OMEGA.cm. The thickness of the buried
oxide was 1 .mu.m. The thickness of the SiO.sub.2 gate dielectric
was 5 nm. The active region, including the wide portions adjacent
to the source and drain regions, was doped with arsenic in the
range of 1.6.times.10.sup.17 cm.sup.-3. The source and drain
regions were doped with arsenic in the range of 5.times.10.sup.19
cm.sup.-3, and the lateral gate regions were doped with boron in
the range of 5.times.10.sup.19 cm.sup.3. The distance from source
to drain region was 10 .mu.m, and the length of the active region,
defined as the length of the lateral gate regions, was 7 .mu.m. The
width of the active region was 400 nm.
[0102] In FIG. 4A, a plot 400 presents results for the specific
detection of anti-troponin. Horizontal axis 402 shows a voltage
V.sub.Gj, in volts, applied to both the left and right lateral gate
electrodes, relative to ground, with the source electrode grounded.
The back gate electrode was kept at a voltage of -7 volts. Curve
404 plots the effective width W.sub.eff of the conducting channel,
in nanometers, as a function of the lateral gate voltage, with the
values shown on vertical axis 406 on the right side of the plot, in
nanometers. Curve 408 plots the change in source to drain threshold
voltage associated with the presence of anti-troponin, as a
function of lateral gate voltage V.sub.Gj, with the values shown on
vertical axis 410 on the left side of the plot, in millivolts. Note
that the narrower the conducting channel is, the higher is the
shift in threshold voltage between the source and drain associated
with the presence of anti-troponin.
[0103] FIG. 4B shows contour plots of the carrier density, in this
case electron density, in a cross-section of the active region,
which is 250 nm high and 400 nm wide, perpendicular to the
direction of the conducting channel, half way between the source
and the drain regions, for different values of the lateral gate
voltage V.sub.Gj. Plots 412, 414, 416, 418, and 420 respectively
show the carrier density for V.sub.Gj equal to -2.0, -1.5, -1.0,
-0.5, and 0.0 volts. Contours 422, 424 and 426 respectively
correspond to carrier densities of 4.times.10.sup.19 cm.sup.-3,
2.times.10.sup.16 cm.sup.3, and 1.times.10.sup.13 cm.sup.-3.
[0104] Also, the principle of operation of the virtual buried
nanowire gas sensor was simulated using `Sentaurus` software, sold
by Synopsys, Inc. The parameters used in the simulation were the
same as the parameters described above for the experiment, except
that the length of the active region and the lateral gate regions
was only 3 .mu.m, in order to save on computation time. A SiO.sub.2
cubic of 10 nm side with fixed charge density of 10.sup.19
ecm.sup.-3 was placed at the center of the channel on top of a gate
dielectric, in this simulation, to represent a molecule adhering at
that location. The shift in the average potential of the SOI region
due to the presence of the charge was calculated for various
channel widths. The results are presented in FIG. 5, in a plot 500.
As in plot 400, horizontal axis 502 shows the voltage V.sub.G--, in
volts, applied to both the left and right lateral gate electrodes,
relative to the source electrode which is grounded, and curve 504,
which is in close agreement with curve 404 in plot 400, shows the
effective width W.sub.eff, in nanometers, of the conducting
channel, as a function of V.sub.G--, with the values shown on
vertical axis 506 on the right side of the plot. Curve 508 shows
the change in average potential of the active region, associated
with the presence of the 10 nm wide charged cube of SiO.sub.2
representing an adhering gas molecule in the simulation, with the
values, in millivolts, shown on vertical axis 510 on the left side
of the plot. The average is taken over the full 400 nm width and
260 nm depth of the active region, and extending over the full
length from source to drain regions. Note that the smaller the
width of the conducting channel, the greater the change in average
potential over the active region due to the presence of the
simulated adhering molecule. The change in potential averaged only
over the conducting channel, though not shown in FIG. 5, goes up
even more dramatically, with narrower channel width.
[0105] FIG. 6 shows a plot 600, illustrating the results of a
simulation of scanning the conducting channel across the active
region by varying the voltage V.sub.Gj1 on the left lateral gate
electrode, and the voltage V.sub.Gj2 on the right lateral gate
electrode. In this simulation, the parameters were the same as for
the experiment described above. Curve 602 shows the normalized
carrier (electron) density as function of lateral position x in the
active region, when V.sub.Gj1=0 volts and V.sub.Gj2=-5.16 volts.
Curve 604 shows the carrier density when V.sub.Gj1=-0.85 volts and
V.sub.Gj2=-3.43 volts. Curve 606 shows the carrier density when
V.sub.Gj1=-2.0 volts and V.sub.Gj2=-2.0 volts. Curve 608 shows the
carrier density when V.sub.Gj1=-3.43 volts and V.sub.Gj2=-0.85
volts. Curve 610 shows the carrier density when V.sub.Gj1=-5.16
volts and V.sub.Gj2=0 volts. By varying the left and right lateral
gate voltages in this way, the position of the conducting channel
moves from left to right, over a distance of 200 nm, while the
width of the conducting channel remains constant at 100 nm. The
active region extends from x=-200 nm to +200 nm.
[0106] FIG. 7A shows a contour plot 700, of the carrier density in
the active region as a function of position in a cross-section
perpendicular to the direction of the conducting channel, the
y-axis, when the lateral gate voltages are both zero, according to
a simulation. The values of x and z are given in nanometers. The
parameters are similar to those of the experiment described above,
except that the arsenic doping in the active region, still
averaging 1.6.times.10.sup.17 cm.sup.3, is not homogeneous, but is
higher closer to the gate dielectric and lower closer to the
insulator layer, which makes the conducting channel form adjacent
to the gate dielectric.
[0107] Curves 702 and 704 in FIG. 7A respectively show the
junctions of the active region with the left and right lateral gate
regions. Curve 706 is the boundary of the depletion zone in the
active region, where the carrier density goes to zero. Curves 708,
710, 712, 714, and 716 respectively show the contours for carrier
density of 0.5.times.10.sup.17, 1.0.times.10.sup.17,
1.5.times.10.sup.17, 2.0.times.10.sup.17, and 2.5.times.10.sup.17
cm.sup.-3. The conducting channel is about 150 nm in diameter, a
large fraction of the width of the active region, which is about
220 nm at its narrowest point.
[0108] FIG. 7B shows a similar contour plot 718, for the case where
the left and right lateral gate voltages are both -2.0 volts. Here
curve 720 is the boundary of the depletion zone in the active
region, where the carrier density goes to zero, and the two
contours inside curve 720 are the contours for carrier density of
0.5.times.10.sup.17 and 1.0.times.10.sup.17 cm.sup.-3. The
conducting channel is now about 40 nm in diameter.
[0109] FIGS. 8A and 8B show contour plots 800 and 802 of carrier
density in the active region, from a simulation, with parameters
similar to those in the plots in FIGS. 7A and 7B, but for a case in
which the FET has been heat treated, so that dopants from the
lateral gate regions have diffused somewhat into the active region,
and with a higher density of dopants implanted in the active
region, 10.sup.18 cm.sup.3. The heat treatment, at a temperature of
1050.degree. C., is applied for 75 seconds, causing boron to
diffuse from the lateral gate regions into the active region.
[0110] Curves 804 and 806 in FIGS. 8A and 8B show the left and
right boundaries of the effective active region, where the net
dopant density still has the same sign as before the heat
treatment. This effective active region has a width of only 90 nm
at its narrowest point, but about 130 nm at the depth of the center
of the conducting channel. FIG. 8A shows the case where the lateral
gate voltages are both zero. Curve 808 is the boundary of the
depletion zone in the active region, where the carrier density goes
to zero. Curves 810, 812, 814, and 816 respectively are the density
contours for 2.times.10.sup.18, 4.times.10.sup.18,
6.times.10.sup.18 and 8.times.10.sup.18 cm.sup.-3. The conductive
channel is about 90 nm wide. FIG. 8B shows the case where the
lateral gate voltages are both -2.0 volts. Curve 818 is boundary of
the depletion zone, and the two curves inside it are the density
contours for 2.times.10.sup.18 and 4.times.10.sup.18 cm.sup.-3. The
conductive channel is only about 25 nm, much narrower than in FIG.
7B, without heat treatment of the FET.
[0111] Although the invention has been described in conjunction
with specific embodiments thereof, it is evident that many
alternatives, modifications and variations will be apparent to
those skilled in the art. Accordingly, it is intended to embrace
all such alternatives, modifications and variations that fall
within the spirit and broad scope of the appended claims.
[0112] It is the intent of the Applicant(s) that all publications,
patents and patent applications referred to in this specification
are to be incorporated in their entirety by reference into the
specification, as if each individual publication, patent or patent
application was specifically and individually noted when referenced
that it is to be incorporated herein by reference. In addition,
citation or identification of any reference in this application
shall not be construed as an admission that such reference is
available as prior art to the present invention. To the extent that
section headings are used, they should not be construed as
necessarily limiting. In addition, any priority document(s) of this
application is/are hereby incorporated herein by reference in
its/their entirety.
* * * * *