Detector for detecting electromagnetic radiation
The invention refers to an detector for detecting electromagnetic radiation according to the preamble of claim 1.
Detecting electromagnetic radiation generally is based on the registration of a dc response (photo-voltage or photo-resistance) of a solid state device, in particular a semiconductor heterostructure, to electromagnetic radiation. Such detectors are able to register changes of a small current through the device upon irradiation with electromagnetic radiation.
Highly sensitive detectors capable of detecting electromagnetic radiation in the microwave, terahertz and infrared frequency range can be used in many different areas, including security, medicine, chemical and biological agent detection, material characterization, THz imaging, telecommunication, astronomy, and other fields.
In the prior art, semiconductor devices are known which are comprising a two-dimensional (2D) charge carrier gas, in particular an electron gas, which is confined in a conducting layer arranged between dielectric layers, wherein the conducting layer is electrically connected at their longitudinal ends to a source contact and a drain contact, respectively and on top of the upper dielectric layer a gate contact in the form of a split gate is arranged. By electrically biasing the gate contact with respect to the conducting layer, a depletion zone is developing in the conducting layer in the regions under the split gate contact. The split gate contact is thus defining a quantum point contact in the conducting layer through which the charge carriers (electrons) can tunnel from one side of the depletion zone to the other side, if the gate voltage applied to the split gate contact is high enough (as disclosed for example in van Wees et.al, Phys. Rev. Lett. 60, 848 (1988) and in Wharam et.al, Solid Stade Phys. 21, L209 (1988)). Such split gate structures can be used for building semiconductor detector devices for electromagnetic radiation, for example as a GaAs/AlGaAs heterostructure. Dorozhkin et.al, Appl. Phys. Lett 87, 092107 (2005) describe for example such a system for detection of microwave radiation.
The sensitivity of the known devices to electromagnetic radiation, however, is often poor. It is therefore the object of the invention to provide a detector for electromagnetic radiation with an improved sensitivity.
This object is achieved with a detector comprising the features of claim 1. Preferred embodiments of the detector according to the invention are shown in the dependent claims. The device of the invention comprises a solid state structure which is confining a two- dimensional (2D) charge carrier gas, in particular an electron gas, and has at least a source contact, a drain contact and a gate contact, with the gate contact operating as a quantum point contact. A substantially enhanced sensitivity of the device to electromagnetic radiation is achieved through the use of a special shape of the gate contact which is formed as a bridged gate.
The detector according to the invention is comprising a substrate, a conducting layer which is confining a two-dimensional charge carrier gas, a dielectric layer lying on the conducting layer, at least two electric contacts (namely a source contact and a drain contact) being electrically connected with the conducting layer and at least one top gate contact covering a part of the dielectric layer, wherein the top gate contact is comprising a first section and a second section which are connected with each other by a bridge section having a minimum width (d) smaller than the width (D) of the first and second sections. The bridge section thereby is forming a constricted portion of the top gate contact, which is defining a quantum point contact (QPC) in the conducting layer when the top gate contact is electrically biased with respect to the conducting layer.
The detector according to the invention preferably can be manufactured as a solid-state device, in particular a heterostructure, with the conducting layer being a semiconductor layer which is confining a two-dimensional charge carrier gas, in particular an electron gas. The conducting layer also can be a transition metal dichalcogenide monolayer, e.g. materials of the type MX2, with M a transition metal (Molybdenum, Tungsten) and X a chalcogen atom (Sulphur, Selenium, or Tellurium), which also can confine a two-dimensional hole gas. In an operation mode of the detector, a (positive or negative) bias is applied to the top gate contact with respect to the conducting layer (depending on whether the charge carriers are holes or electrons, respectively). In operation of the detector, the electrical resistance or conductance and/or a photovoltage between the contacts developing upon irradiation of the detector with an electromagnetic radiation is measured, in order to measure the intensity of the electromagnetic radiation.
In a preferred embodiment of the invention, the first and second section of the top gate contact at least substantially have the form of a stripe having a predetermined width (D) and the bridge section is having a minimum width (d) smaller than the width (D) of the stripe. The bridge section, which is defining the constricted portion of the top gate contact, can be realized by known etching or lithography techniques in a semiconductor structure and may be formed by at least one substantially U- or V-shaped recess in the top gate contact.
In a preferred embodiment of the invention, the conducting layer is for example extending in a longitudinal direction (x) and is having a first longitudinal end which is electrically connected with one of the electric contacts (source contact) and a second longitudinal end which is connected with the other one of the electric contacts (drain contact). The top gate contact preferably and substantially has the form of a stripe which is extending in a transversal direction (y), wherein the transversal direction (y) is (at least substantially) perpendicular to the longitudinal direction (x). In an alternative embodiment of the detector, the first section and the second section of the top gate contact are inclining an angle of lower than 180° with each other, for example an angle between 10° and 90°.
When electrically biased with respect to the conducting layer, the top gate contact is dividing the conducting layer into two regions (sides) separated by a depletion layer and the charge carriers confined in the conducting layer are enabled to quantum mechanically tunnel from one region (side) to the other region (side) through the quantum point contact (QPC).
The minimum width (d) of the constricted portion of the top gate contact is preferably in the order of the electronic wavelength (de Broglie wavelength) of the charge carriers confined in the conducting layer, or it is several times the charge carrier's de Broglie wavelength. In preferred embodiments, the minimum width (d) of the constricted portion of the top gate contact is smaller than 0,5 μιη and preferably is smaller than 0,1 μιη. The minimum width (d) of the constricted portion of the top gate contact is defining the tunneling resistance for the charge carriers tunneling from one region (side) of the conducting layer to the other region (side). The lower limit of the minimum width (d) of the constricted portion of the top gate contact is limited by experimental constraints, because the lower dimensions of structures in solid state heterostructures are limited by the resolution of available etching or lithography technologies.
The tunnelling resistance (or conductance) of the quantum point contact (QPC) is controllable by a gate voltage (Vgate) applied to the top gate contact which biasing the top gate contact with respect to the conducting layer. Since the tunnelling resistance or conductance of the quantum point contact (QPC) is sensitive to irradiation by electromagnetic radiation, the electrical resistance or conductance between the contacts is responsive to the intensity of the electromagnetic radiation irradiated on the detector. Therefore, by measuring the photovoltage induced between the contacts (source and drain contacts) by electromagnetic radiation irradiated on the detector, the intensity of the electromagnetic radiation can be determined.
In the detector according to the invention, irradiation of the detector by an electromagnetic radiation induces an ac local electric field near the edges of the top gate contact, wherein the (polarization) direction of the induced local electric field in the region of the constricted portion of the top gate contact is at least substantially parallel to the tunnelling direction of the charge carriers which are tunnelling through the quantum point contact (QPC). The induced local electric fields thereby can be much higher than the electric fields of the incident wave of the electromagnetic radiation. The locally induced electric fields are polarized normally to the edges of the gate contact. In contrast to known split gate contacts, in the bridge gate contact of the invention the polarization direction of the locally induced ac electric fields therefore is at least mainly parallel to the tunnelling direction of the charge carriers which are tunnelling through the quantum point contact (QPC). For this reason, the tunnelling charge carriers are much more influenced by the locally induced ac electric fields in the region of the quantum point contact (QPC) and therefore the detector according to the invention is more sensitive than known devices having a split gate contact.
In further embodiments of the invention, the detector is comprising a number of top gate contacts with a constricted portion arranged on top of the dielectric layer. These multi-gate structures can be used to implement frequency-sensitive detectors.
In such an embodiment of the invention, further gate contacts are arranged on top of the dielectric layer and between a first top gate contact and a second top gate contact, wherein both top gate contacts preferably are arranged in longitudinal direction (x) one behind the other and in a distance from each other and have the form of stripes which are running parallel
to each other. Thus, a double quantum point contact (QPC) can be realized. When biasing both top gate contacts and the further gate contacts with respect to the conducting layer, depletion layers are developing in the conducting layer under the biased contacts, which are surrounding an area in the conducting layer occupied by 2D charge carriers (like in a quantum dot structure). The size of the area and the charge carrier density are preferably selected to enable the charge carriers confined in this area to oscillate with their plasma frequency upon irradiation of an electromagnetic radiation having a frequency which is equivalent to the plasma frequency (cop). In this embodiment, the predetermined plasma frequency (cop) of the charge carriers confined in said area of the conducting layer can be controlled by external magnetic fields or electric fields, as for example by biasing a back gate electrode arranged on the backside of the substrate with which the charge carrier density in the conducting layer can be controlled. This embodiment of the detector therefore is sensitive to the frequency of the incident electromagnetic radiation.
These and further features and advantages of the invention will be disclosed in the following description showing preferred embodiments of the invention by reference to the
accompanying figures. The figures show:
Figure 1 : Basic geometry of a detector according to the invention in a side view (figure la) and a top view (Figure lb);
Figure 2: preferred embodiments of a detector according to the invention implemented as a GaAs/AlGaAs heterostructure for confining a 2D electron gas;
Figure 3: Basic geometry of a detector according to the invention implemented in a
graphene system in a side view (figure 3a) and a top view (figure 3b); Figure 4: Different embodiments for the shape of the top gate contact of a detector
according to the invention in a top view on the top gate contact;
Figure 5 : Schematic demonstration of the electric fields around the quantum point
contact (QPC) developed in the conducting layer of a detector according to the
invention having a bridged gate contact in a side view (Figure 5 a) and in a top view (Figure 5b);
Schematic demonstration of the electric fields around the quantum point contact (QPC) developed in the conducting layer of a prior art device with a split gate contact in a top view;
Comparison of the normalised photoconductan.ee measured in a detector according to the invention having a bridged-gate contact and measured i a detector according to the prior art having a split-gate contact as a function of the normalised intensity (power) of an incident electromagnetic radiation.
Representation of the electron distribution function in the conducting layer of a detector according to the invention in the absence of electromagnetic radiation (Figure 8a) and under the action of electromagnetic radiation linearly polarized in the x-direction (Figure 8b);
Top view of a basic geometry of a frequency- sensitive embodiment of a detector according to the invention having a double bridge-gate geometry with a first and a second top gate contact;
Top view of a frequency-sensitive detector according to the embodiment of
Figure 9, implemented in a GaAs/AlGaAs heterostructure with the central part of one of the top gate contacts shown i an enhanced representation in the insert (bottom left part of the picture).
As shown in the basic geometry of Figure 1, the detector according to the invention is comprising a substrate 3, a (semi)conducting layer 1, a dielectric layer 4 lying on the (semi)conducting layer 1, at least two electric contacts, namely a source contact 2a and a drain contact 2b, and at least one top gate contact 5 covering a part of the dielectric layer 4. The (semi)conducting layer 1 is made of a conducting or a semiconducting material. When in the following a reference is made to the conducting layer 1, this means that this layer also can be a semiconductor layer. The conducting layer 1 is confining a two-dimensional charge carrier gas, which can be an electron gas or a hole gas, depending on the type of carriers present in
the conducting layer 1. The conducting layer 1 is lying in a horizontal plane (x-y-plane) and is extending in longitudinal direction x between a first end la and a second end lb. The first end la of the conducting layer 1 is electrically connected with the source contact 2a and the second end lb is electrically connected with the drain contact 2b.
According to the invention, the top gate contact 5 is formed as a "bridged-gate" and is comprising a first section 5 a and a second section 5b which are connected with each other by a bridge section 5c. The top gate contact 5 in general is made of an electrically conducting material, in particular a metal or a semimetal, and the sections 5 a, 5b and the bridge section 5 c are preferably all made of the same material. In the embodiment shown in Figure lb, the first and the second section 5a, 5b both have substantially the form of a stripe with a (medium) width D and the bridge section 5 c is having a minimum width d smaller than the general width D of the sections 5a, 5b. The bridge section 5c is thus forming a constricted portion of the top gate contact 5. When the top gate contact 5 is electrically biased with respect to the conducting layer 1 by applying a gate voltage Vgate to the top gate contact 5, a depletion zone is developed in the conducting layer 1 in the areas under the top gate contact 5 having a lower carrier density than in the other areas of the conducting layer 1. The depletion layers are thus defining a quantum point contact (QPC) in the conducting layer 1 in the region of the conducting layer 1 under the constricted portion (in the bridge section 5c) of the top gate contact 5.
In the operation mode of the detector, a negative or positive bias is applied to the top gate contact 5 with respect to the 2D conducting layer 1, depending on the type of carriers in the conducting layer 1 (a negative bias is applied when a 2D electron gas is confined in the conducting layer and a positive bias is applied in the case of a 2D hole gas). The bias of the top gate contact 5 leads to a reduction of the charge carrier density under the top gate contact 5 and thus a depletion layer is formed in the conducting layer in the area under the top gate contact 5. The depletion layer is hence separating the conducting layer 1 in two regions (sides a and b, shown in Figure lb), so that the charge carriers have to quantum-mechanically tunnel from the left region (left side a) to the right region (right side b). In the tunneling regime the resistance of the QPC can be made very high (Mega-Ohms scale). The tunnelling resistance of the QPC can be controlled by the gate voltage Vgate applied to the top gate contact 5 for biasing it against the conducting layer 1. For detecting whether the detector is irradiated with an electromagnetic radiation, the resistance (conductance) or the induced photovoltage between the two contacts 2 (source contact 2a, drain contact 2b) is measured.
Due to the constricted portion of the top gate contact 5 having a minimum width d, as shown in Figure lb, the charge carriers preferably tunnel through this narrow region. The constricted portion of the top gate contact 5 is thus forming a quantum point contact (QPC). The detector according to the invention now comprises a bridged gate as the top gate contact 5 (instead of of a split gate known from the prior art). Due to physical effects which will be explained in the following, the QPC of the detector according to the invention is substantially more sensitive to electromagnetic radiation. In the tunnel regime, the QPC devices are sensitive to irradiation of an electromagnetic radiation and can be used for detecting it. Since the top gate contact 5 is conducting, the electromagnetic wave of the incident electromagnetic radiation with its electric field 6 induces near the edges of the top gate contact 5 strong local ac electric fields 7 which can be much larger than the field of the incident wave itself (known as the lightning-rod effect: the field strongly increases near sharp edges of metallic (conducting) objects). The charge carriers of the 2D conducting layer 1, which should tunnel from one region (left side a) of the conducting layer to the other region (right side b), are influenced by this strong induced and oscillating electric fields. Their energy and the effective local Fermi energy increases, which leads to a strongly enhanced tunnel current.
The detector according to the invention with the QPC developed under and by the biased and bridged-gate top gate contact 5 now is substantially more sensitive to the irradiation with electromagnetic radiation. The reason is that, independent of the electric field polarization of the incident electromagnetic wave, the electric field 7 locally induced in the region of the edges of the top gate contact 5 turns out to be polarized normally with respect to the edge of the top gate contact 5, as shown in Figure 5. In the traditional split-gate geometry, shown for comparison in Fig. 6, the direction of the ac electric field 7 turns out to be mainly perpendicular to the tunneling direction of the charge carriers (shown by the large horizontal arrow in Figures 5 and 6). In contrast, in the bridged-gate geometry according to the invention, the direction of the ac electric field 7 is parallel to the tunneling direction of the charge carriers. It is possible to theoretically demonstrate that the time-averaged non- equilibrium distribution function f(px, py) of 2D electrons in the momentum (p = (px, py)) space is strongly elongated in the direction of the applied ac electric field, as shown in Figure 8. The left panel in Fig. 8 (Fig. 8a) shows the electron distribution function in the absence of
radiation (the conventional Fermi-Dirac distribution). The right panel (Fig. 8b) shows the stationary (averaged over time) electron distribution function in the presence of electromagnetic radiation linearly polarized in the x-direction. The function f(px, py) is found by solving the kinetic Boltzmann equation in the momentum-relaxation-time approximation. One sees that the average energy of electrons for the motion along the external ac electric field 7 is strongly increased remaining the same in the direction perpendicular to the external field polarization. This leads to the much stronger influence of electromagnetic radiation on the tunnel current in the bridged-gate QPC according to the invention, as compared to the traditional split-gate QPC.
This result is shown in Figure 7, where a comparison of the normalised photoconductance measured in a detector according to the invention having a bridged-gate contact (top gate contact 5) is made with the photoconductance measured in a detector according to the prior art having a split-gate contact. In the measurement, a source-drain- voltage was applied between the source and drain contacts (contacts 2a and 2b), so that a weak current flows, which current is changing under irradiation and the current change was measured in order to detect the intensity of the incident electromagnetic radiation. Likewise, also the photo voltage appearing between the source and drain contacts 2a, 2b under irradiation could be measured for detection.
Figure 7 shows the measured and normalised photoconductance as a function of the normalised intensity (power) of an incident electromagnetic radiation. It can be seen that the detector according to the invention reveals a much better sensitivity than the prior art device. One observes a more than one-order-of-magnitude difference of the photo-conductance.
In Figure 4, different embodiments for the shape and geometry of the top gate contact 5 are shown.
Figure 4a shows a top gate contact 5 in the form a stripe having two recesses approximately in the middle of the stripe length at opposite sides. Between these two recesses a bridge section 5c is arranged which is connecting (bridging) a first stripe section 5a with a second stripe section 5b of the top gate contact 5. The recesses substantially can have a V- shape (as shown in Figure 4a) or a U- shape and can be realized by etching the stripe of the top gate contact 5 and/or by electron beam or photolithography. The smallest width d of the bridge section is
mostly limited by technological restrictions of the used etching or photolithography techniques and is preferably smaller than 0,5 μιη and more preferred smaller than 0,1 μιη, in order to be small enough to enable the charge carriers to tunnel through the QPC. The width D of the stripe sections 5a, 5b is not crucial and can be for example in the range of one up to a few μιη.
Figure 4b shows an alternative embodiment for the shape of the top gate contact 5 with only one U-shaped recess in the generally stripe-formed contact, building a bridge section 5c lying at the outer edge of the contact 5 and which again is connecting the two stripe sections 5 a and 5b of the top gate contact 5.
In Figure 4c, a modification of the embodiment of Figure 4b is shown. Therein, an additional element in the form of a conducting stripe 9 is placed on top of the dielectric layer 4 on one side (here: left side a) of the structure. The conducting stripe 9 is directed towards the bridge section 5c of the top gate contact 5 and preferably is of metal. The conducting stripe 9 is not biased (no dc voltage is applied between the stripe 9 and the underlying 2D charge carrier gas in the conducting layer 1). As described above, a dc voltage is applied between the top gate contact 5 and the underlying 2D charge carrier gas in the conducting layer 1. As a result, the density of charge carriers under the top gate contact 5 is close to zero, thereby forming a depletion layer in the conducting layer 1 under the top gate contact 5 which is separating the conducting layer 1 into two sides, a and b. The charge carriers have to tunnel from one side (left side a) of the structure to the other side (right side b). Under the stripe 9 the density of charge carriers is the same as in the whole left side a. The role of the stripe 9 is to serve as an antenna which concentrates the electric field of the incident electromagnetic radiation in the quantum-point-contact (QPC) of the device, which is arranged in the conducting layer 1 under the bridge section 5c of the top gate contact 5. This further can improve the sensitivity of the detector, since the tunnelling resistance (conductance) of the charge carriers tunnelling from one side (a) to the other side (b) is now even more influenced by the electric field of the incident electromagnetic radiation.
As in the embodiments described above, the width of the narrowest area in the QPC (which is defined by the minimum width of the bridge section 5c of the top gate contact 5) can be a in the range of 100 nm or smaller, and up to some μιη. The optimal length of the stripe 9 in the
horizontal direction is depending on the wavelength λ of the incident radiation to be detected and is in the order of ~ λ/2, wherein the wavelength is i.e. λ ~ 0.15 mm for the frequency 1 THz and ~ 0.75 μιη for the telecommunication wavelength 1.55 μιη. The width of the stripe 9 indeed is not crucial and can be for example in the range of one up to a few μιη. In Figure 4d, a modification of the embodiment of Figure 4c is shown with a top gate contact 5 (which is biased against the conducting layer 1) and a stripe 9 (which again is not biased as in the embodiment of Figure 4c) serving as an antenna to direct the incident electromagnetic radiation towards the top gate contact 5. Compared to the embodiment of Figure 4c, the shape of the top gate contact 5 is different in the embodiment of Figure 4d. The top gate 5 has now the form as shown in Figure 4d with a first section 5a and a second section 5b, which are both formed here as conical stripes which are inclining an angle of for example between 10° and 90°. The conical form of the stripes, however, is not necessary or crucial for their function, any other shape of the sections 5a, 5b would work just as well. The inner ends of the first and second sections 5a, 5b which are directing towards the centre of the structure are connected (bridged) with each other by a bridge section 5c. Approximately in the middle of the bridge section 5c, it has a minimum width d (which again is in the order of 0,1 μιη or smaller). The minimum width d of the bridge section 5c is preferably smaller than the lowest width of the sections 5a and 5b of the top gate contact 5 (due to the conical form of the stripes defining the sections 5a and 5b, the lowest width of the sections 5a and 5b here is at their inner ends at which they are connected with the bridge section 5 c, as shown in Figure 4d). As in the embodiments described above, the first longitudinal end la of the conducting layer 1 is connected with a source contact 2a and its second longitudinal end lb is connected with a drain contact 2b, which now is arranged between the outer ends of the first and second section 5a, 5b of the top gate contact 5. Between the inner end of the stripe 9 and the bridge section 5c of the top gate contact 5, a gap area 7 is formed.
As described above, by biasing the top gate contact 5, the charge carrier gas under the gate 5 is depleted and the charge carriers have to tunnel through the QPC defined by the restricted portion of the top gate contact 5 in the bridge section 5 c, in order to move from one side (left side a) to the other side (right side b) of the 2D charge carrier gas confined in the conducting layer 1. Therein, the left side a is placed outside of the first and second sections 5a, 5b of the top gate contact 5 and the right side b is placed inside of them, as shown in Figure 4d. The shown shape of the top gate contact 5 now also serves as an antenna and is further directing the electric field of the incident radiation towards the region of the QPC formed in the conducting layer 1 under the bridge section 5c. As a result, both the top gate contact 5 and the
additional conducting stripe 9 serve as antennas focusing the electric field in the gap area 7 between the inner end of the stripe 9 and the bridge section 5 c of the top gate contact 5, which gap area 7 is quite in the vicinity of the QPC. This enhances the electric field induced in the region of the gap 7, as compared to the field of the incident wave of the electromagnetic radiation, and leads to an increase of the sensitivity of the detector.
Any other structures and shapes of antennas focusing the electric field in the near-QPC area (with the polarization of the field in the horizontal direction) are possible.
The invention has been technically implemented in semiconductor heterostructures. Figure 2 shows two different embodiments of suitable modulation doped GaAs/AlGaAs heterostructures which can be used for implementing the invention.
With reference to Fig. 2(a) there is shown a first embodiment of a GaAs/AlGaAs heterostructure suitable for realising a 2D charge carrier layer in the form of an electron gas (two dimensional electron gas, 2DEG) confined in a conducting layer.
The semiconductor structure of the embodiment of Figure 2(a) comprises a GaAs substrate 10, typically with a thickness of 0.3 to 0.5 mm (however, this is not critical). A GaAs buffer layer 12 is grown on the GaAs substrate and is followed by a superlattice structure 14 comprising alternating layers of AlGaAs and GaAs, with the AlGaAs layers typically being 7 nm thick and GaAs layers typically being 3 nm thick. The total thickness of the buffer structure 12, 14 typically amounts to 1 μιη, however, this is again not critical. The idea is to prevent impurities from the GaAs substrate migrating into the active part of the device. The active part of the device comprises the layer 16 of GaAs (intrinsic material) which is followed by an intrinsic layer 18 of AlGaAs. After the intrinsic layer 18 has been grown it is followed by a silicon-doped AlGaAs layer 20 and a capping layer 22 of GaAs. On top of the capping layer 22 a metallic top gate contact is placed (not shown on the Figure), wherein this top gate contact is formed according to the invention and as described above.
By applying a dc gate voltage Vgate between the top gate contact and the 2DEG confined in the electric conducting layer 16, the density of 2D electrons under the top gate contact can be changed and controlled by the gate voltage Vgate. Alternatively, an additional conducting layer, separated by at least one barrier from the 2DEG, may be inserted in the layer structure of Fig. 2(a) below the 2D charge carrier layer (also not shown in the figures). It can serve as a
back-gate for changing the density of charge carriers (electrons) in the 2DEG: By applying a back gate voltage V back gate between this back-gate and the 2DEG, the density of 2D electrons in the 2DEG can be changed. The semiconductor heterostructure is, for example, grown by molecular beam epitaxy (MBE) or chemical vapour deposition (CVD), as are all other structures described here. The conduction band scheme for the structure of Fig. 2(a) is shown in the same figure in an insert right from the structure geometry. It can be seen that the interface between the undoped AlGaAs layer 18 and the undoped GaAs layer 16 forms a potential well 24. The potential well 24 has a quantised energy level 26 for electrons which are induced into the potential well by the silicon donor atoms provided in the layer 20. The electrons are thus sharply localised in the potential well 24 and this results in the two-dimensional electron gas illustrated as the layer 26 in the left part of Fig. 2 (a). The GaAs layer 16 therefore is equivalent to the conducting layer 1 of the basic geometry of a detector according to the invention as shown in Figure 1, since it is confining a two dimensional charge carrier (electron) gas.
As can be seen from the insert in Fig. 2(a), this two-dimensional electron gas is localised in the GaAs layer 16 having a thickness of approximately 5-10 nm. Although this is a very thin layer it is probably better described as a quasi two-dimensional electron gas rather than a strict two-dimensional electron gas which would have zero thickness.
It can also be seen from the insert in Fig. 2(a) that the layers 18, 20 and 22 provided on top of the GaAs layer 16 have a total thickness typically in the range from 35 nm to 250 nm, with the relative proportions being as shown in the drawing. For the detector operation these dimensions should be possibly small. The GaAs layer 16 typically has a thickness of approximately 1 μιη, but again this is not critical.
Turning now to Fig. 2(b) there can be seen an alternative embodiment for realising a two- dimensional electron gas. The structure of Fig. 2(b) is similar to that of Fig. 2(a) and the layers which are common to both structures have been given the same reference numerals.
It can be seen that the layer 16 of GaAs has been reduced in this structure to a thin layer having a thickness of ~10 nm and that it is sandwiched between two layers 18 of intrinsic
AlGaAs. These two layers 18 are in turn sandwiched between two layers 20 of AlGaAs doped with silicon.
The conduction band diagram for this structure in the vicinity of the thin intrinsic GaAs layer 16 is shown in the insert of Fig. 2(b), right. The thin layer 16 of intrinsic GaAs forms a quantum well 24 with a quantised energy level 26 in which electrons are localised. The electrons are induced by the presence of the silicon donors in the two layers 20. Again, as in the embodiment of Figure 2(a), the GaAs layer 16 is equivalent to the conducting layer 1 of the basic geometry of a detector according to the invention as shown in Figure 1, since it is confining a two dimensional charge carrier (electron) gas.
Theoretical calculations and the experimentally demonstrated heterostructures revealed the following key parameters of the detectors implemented by the above described semiconductor heterostructures:
The density n of the charge carriers confined in the conducting layer 1 (2D electrons) and their mobility μ typically is n ¾ (1 - 8) x 1011 cm2 and μ ¾ (1 - 3) x 106 cm2/Vs, corresponding to a mean free path of « 60 μιη which is substantially exceeding the dimensions of the QPC (defined by the minimum width (d) of the bridge section 5c).
The top gate contact 5 can be fabricated on the surface of the heterostructure (dielectric layer 4) using electron beam lithography. The distance between the top gate contact 5 and the 2DEG typically is 90 nm. The samples were irradiated by microwaves with the frequency in the range from 110 to 170 GHz and a power density ¾ (1 - 10) mW/cm2, corresponding to the microwave electric field of Εω ¾ (0.6 - 2) V/cm. The measurements were carried out in a VTI cryostat with a wave-guide delivering microwave radiation down to the sample and by using a conventional lock-in technique to measure the resistance R = l/σ. The sample was rotated with respect to the orientation of the waveguide but no polarization dependence has been observed. Nine devices were studied and similar results have been obtained for all of them.
The operation of detectors according to the invention could be demonstrated at microwave frequencies f ¾ 110 - 170 GHz and at low (4.2 K) temperatures, but the operating frequency range can be extended both to lower and higher (terahertz, IR) frequencies, and the working temperatures can be increased at least up to the liquid nitrogen temperature (experimentally verified) and probably up to room temperature.
The estimated internal responsivity of the detector should be about 106 V/W. This is several orders of magnitude larger than the responsivity of terahertz Schottky-diode detectors (which is between 100 and 1000 V/W). The detectors can operate in the resonant (frequency sensitive) regime.
The operation principle of the detectors according to the invention is based on electronic transitions. Its response time is therefore restricted by only very short electron relaxation times. This is substantially shorter than, for example, the response time of bolometers whose operation is based on the thermal response.
The effective area of a single working element of the detector is about 100 x 100 nm2, together with lead wires - smaller than 10 x 10 μιη2. The device can be realized in the form of a line or a matrix of detectors for sub-wavelength detection of microwave/terahertz radiation (sensitive microwave/terahertz/infrared camera).
The detectors implemented in the GaAs/AlGaAs heterostructure system comprising the above described bridged-gate structure (top gate contact 5 with a bridge section 5 c) revealed to be substantially more sensitive to the irradiation than traditional structures having a split-gate as the top gate contact. This is illustrated in Figure 7, where measurement results of detectors with a split-gate (according to the prior art) and a bridged-gate (according to the invention) are shown. In the measurements, the gate voltages Vgate are chosen so that the dark conductances (P = 0) are approximately the same in both devices. The maximum power Po was also approximately the same both in the split- and bridged-gate devices which was independently controlled by the amplitude of the microwave induced resistance oscillations measured on the Hall bar. The bridged-gate devices according to the invention demonstrate a more than one order of magnitude higher sensitivity (normalized photo-conductance σ/σ0) as compared to the split-gate detectors according to the prior art. Alternatively to the above described GaAs/AlGaAs heterostructures, the detectors also can be build using other materials, in particualar any other suitable semiconductor material system for confining a 2DEG in a (semi)conducting layer. Another option is to use recently discovered atomically thin two-dimensional crystals like, for example, transition metal dichalcogenides (MoS2, WSe2, etc) as the (semi) conducting layer 1 (according to the basic geometry of Fig. 1), boron nitride as the dielectric layer 4 and graphene for building the top
gate contact 5. Since the thickness of these layers (- 0.5 - few nm) is 10 - 100 times smaller than in GaAs/AlGaAs structures, the local electric fields 7 (Figure 5) near atomically sharp conducting edges are expected to be orders of magnitude larger. This should lead to an orders- of-magnitude larger responsivity of the detector.
Figure 3 shows an example for the geometry of a detector based on a Graphene Related Material (GRM) in a side view. A semiconducting layer 1 is arranged on top of a substrate 3 and is electrically connected with metallic contacts 2a and 2b. The substrate is a dielectric material, e.g. quartz, Si02 and should be sufficiently thick (in the range of -0.5 mm or thicker) to be mechanically stable. The structure can also lay, e.g. on top or side of a wave- guiding structure, for example on an optical wave guide in which the presence of the waves should be detected.
The semiconducting layer 1 is made out of, e.g., a mono-atomic layer of molybdenum disulfide, MoS2, with a thickness of -0.65 nm, and serves as a conducting layer in the sense of the basic structure of the detector according to the invention shown in Figure 1 (instead of a GaAs quantum well, as described above with reference to Figures 2(a) and 2(b)). Alternatively, other transition metal dichalcogenide (TMDC) monolayers which are atomically thin semiconductors of the type MX2, with M a transition metal atom (Mo, W, etc.) and X a chalcogen atom (S, Se, or Te) can be used, wherein one layer of M atoms is sandwiched between two layers of X atoms.
On top of the semiconducting layer 1 lies a dielectric layer 4, e.g., of hexagonal boron nitride (h-BN) which serves as a dielectric and should have a thickness of only a few atomic layers, e.g. in the range of 5-10 nm. The top gate contact 5 is arranged on top of the dielectric layer 4, which is preventing the tunnelling of electrons from the top gate contact 5 to the semiconducting layer 1. The top gate contact 5 can be made out of (a) graphene or (b) a metallic material. The top gate contact 5 should be sufficiently thick to be a good conductor and to serve as an antenna (e.g. a metallic layer - 50 - 100 nm thick). The above described detector devices are comprising a single QPC formed in the conducting layer 1 under the (bridged-gate) top gate contact 5. With such detector devices, it is possible to detect the intensity of electromagnetic radiation, but not its frequency. Using embodiments of the detector with two or more bridged-gate QPCs, it is possible to realize detectors which
are sensitive on the frequency of the incident radiation. The frequency-sensitivity of the devices can be electrically and/or magnetically controlled. In the following, such an embodiment of the invention is described in detail: A preferred embodiment of a detector comprising two-bridged-gate top gate contacts 5.1 and 5.2 is shown in Figure 9. The detector device shown in Figure 9 is comprising a first top gate contact 5.1 and a second top gate contact 5.2, in addition to the source and drain contacts 2a and 2b, wherein the first top gate contact 5.1 and the second top gate contact 5.2 both have a constricted portion 5c with a minimum width (d) smaller than the width (D) of the remaining portions of the respective top gate contact (as in the previous embodiments), wherein the constricted portion 5c of each top gate contact 5.1 , 5.2 is defining a quantum point contact (QPC) in the conducting layer 1, when the respective top gate contact 5.1 , 5.2 is electrically biased with respect to the conducting layer. The first top gate contact 5.1 and the second top gate contact 5.1 are arranged in longitudinal direction (x) one behind the other and in a distance from each other, wherein both top gate contacts 5.1 , 5.2 have the form of stripes which are running parallel to each other. Two further gate contacts 5.3, 5.4 are arranged on top of the dielectric layer and between the first top gate contact 5.1 and the second top gate contact 5.2. When biasing both top gate contacts 5.1 , 5.2 and the further gate contacts 5.3, 5.4 with respect to the conducting layer 1, depletion layers are developing in the conducting layer 1 under the biased contacts. The depletion layers are surrounding an area 9 in the conducting layer 1, which is occupied by 2D charge carriers (like in a quantum dot structure). The charge carriers in this well-defined area 9 of the conducting layer 1 are thus confined horizontally (x- y-plane) within the depletion layers surrounding the area and vertically (z-direction) within the conducting layer 1. Since the charge carriers confined in this area 9 of the conducting layer 1 have a predetermined charge carrier density (n) and a predetermined plasma frequency (cop), the size of the area 9 and the charge carrier density (n) can be selected to enable the charge carriers confined in the area to oscillate with their plasma frequency upon irradiation of an electromagnetic radiation having a frequency which is equivalent to the plasma frequency (cop). The predetermined plasma frequency (cop) of the charge carriers confined in the conducting layer can be controlled by external magnetic fields or electric fields, as for example by biasing a back gate electrode arranged on the backside of the substrate with which the charge carrier density in the conducting layer can be controlled. This embodiment of the detector therefore is also sensitive to the frequency of the incident electromagnetic radiation.
The voltage applied to the bridged-gate contacts 5.1 and 5.2 controls the current flow between the source and drain contacts 2a, 2b. A voltage applied to the further gates 5.3 and 5.4 creates depletion layers under these gates and can control the effective size of the area 9 occupied by 2D electrons. Inside the area 9 electrons can oscillate with the plasma frequency (cop), dependent on the electron density (n) and the size of the area 9. If the frequency ω of the incident electromagnetic radiation coincides with the plasma frequency (cop), the radiation resonantly excites the 2D plasmons and the source-drain current between the source contact 2a and the drain contact 2b resonantly depends on the frequency co. Figure 10 shows a photographic representation of a realized sample of a GaAs/AlGaAs heterostructure with a double top gate contact structure as schematically shown in Figure 9 and as described above, in a top view. The central part of one of the top gate contacts is show in an enhanced representation in the insert (bottom left part of the figure).
The resonance frequency (plasma frequency C0p) can be controlled by external magnetic fields: in the absence of a magnetic field B the plasma frequency C0p is proportional to -v/n/R and hence depends on the electron density n and the radius R of the well-defined area 9, which can be tuned by the gate voltages V5.3 and V5.4 applied to the further gates 5.3 and 5.4, respectively. If the external magnetic field B≠ 0, the plasma frequency C0p depends, in addition, on the magnetic field and particularly is proportional to n/BR (so called edge magnetoplasmons).
The resonance frequency (plasma frequency C0p) can also be controlled by external electric fields: Therefore a back gate electrode can be arranged on the substrate, By biasing the back gate electrode, an electric field is produced in vertical direction (z) which is controlling the charge carrier density (n) in the conducting layer 1 and hence the plasma frequency cop oc Vn/R . Technically it is possible with the invention (and even easier as compared to split gate structures) to realize a line or a matrix of a number of bridged-gate contacts on top of the dielectric layer 4. The detector devices according to the invention with a number of bridged- gate top gate contacts 5 (and a number of associated QPCs, one for each top gate contact) can
thus be used in a matrix to build a fast terahertz/infrared camera with a sub- wave length resolution and extremely high responsivity.