Polarimetric optical device with an insulating layer between detectors

The invention relates to an optical polarimetric detector making it possible to detect two polarizations and, at the same time, eliminating the background noise.

The object of the present invention is to describe the architecture making it possible to combine two specific functionalities already demonstrated by quantum well detectors; the possibility of producing bispectral devices (French patent No. 2 756 667) and the possibility of producing detectors integrating, in the active layer, the function of subtracting the continuous component, that is to say the dark current of thermal origin and the optical scene current (French patent No. 2 756 666). Each of these two functions requires a double stack of quantum wells and three stage connector technology. The invention makes it possible to keep this number of stacks and of connector technology stages, while making it possible to read in subtractive mode and according to three, vertical (V), horizontal (H) and mixed (V+H) polarimetric templates.

The invention therefore relates to a bifunctional optical detector comprising:

a first active detector element having a photoconductor with which a first diffraction grating is associated allowing incident light from a first specific polarization direction to be coupled in the first detector element and allowing the latter to detect the light having this first polarization direction;

a second active detector element having a photoconductor with which a second diffraction grating is associated allowing incident light from a second polarization direction to be coupled in the second detector element and allowing the latter to detect the light having this second polarization direction.

The invention also relates to a method of producing such a detector.

The various objects and features of the invention will become apparent on reading the description and the appended figures:

FIG. 1, an exemplary structure of a detector according to the invention;

FIG. 2, a top view of the structure of FIG. 1;

FIG. 3, a matrix of detectors according to the invention;

FIG. 4, a table of operation modes.

The proposed architecture is identical to that of a subtractive component. The spectral response of a quantum well detector is the convolution of the spectral absorption of the detector and of the spectral efficiency induced by the geometric resonance of the diffraction grating used for the optical coupling of the incident radiation. Thus, a one-dimensional diffraction grating resonating around λ and oriented in a specific direction (or in a perpendicular direction) coupled to the quantum structure described above will induce an optical response centered around λ and sensitive to the vertically (or horizontally) polarized incident light.

The principle of the invention is to separate the main pixel of size a×a into two subpixels of size a×a/2, each of these subpixels being a conventional subtractive structure, but having two different coupling arrays (FIGS. 1 and 2 ).

The bias of the lower stage Vref, common to the set of pixels remains common to both types of subpixels. On the other hand, the biased of the upper stage, which is also common to the set of pixels, must be divided in two, each subarray of subpixels being biased in a common manner. The connection technology of the upper stage is, in this case, provided by two bias lines V 1 and V 2 per pixel instead of a single line for a standard subtractive pixel, as will be seen in FIG. 4 .

The device may then be used in three subtractive modes corresponding to the three different polarization modes V, H and H+V, depending on the bias voltages which are applied to each of the upper electrodes (FIG. 4 ).

According to a preferred embodiment of the invention, an optical detector is produced on a transparent substrate through which the incident light reaches the detector. An optical detector of this type therefore comprises, on the substrate, a structure such as shown in FIG. 1 . This structure comprises, on one face of the substrate, a first ohmic contact layer 5 , this ohmic contact layer being transparent. A stack of layers 2 forming a stack of quantum wells capable of detecting at least one range of wavelengths is located on this ohmic contact layer. A second transparent ohmic contact layer 4 is on this stack of layers. Two identical or almost identical stacks of layers 1 , 1 ′, each one forming a stack of quantum wells capable of detecting at least one range of wavelengths λ, are on the second ohmic contact layer and finally, diffraction gratings 3 and 3 ′ having different structures are on each of these stacks of layers. Each of these gratings allow either the light polarized in one direction (for example vertical) or the light polarized perpendicularly (horizontally) to be coupled in the stack of layers ( 1 , 1 ′ on which the grating is produced). To do this, the two gratings are oriented perpendicularly (see FIG. 2 ).

As shown in FIG. 1, the assembly is encapsulated in an insulating layer 6 and connection means pass through this insulating layer in order to reach the diffraction gratings and produce connection means C 1 and C′ 1 enabling voltages +V 1 and +V 2 to be applied respectively to the diffraction gratings 3 and 3 ′. Also, at least one connection means passes through the insulating layer and reaches the second ohmic contact layer 4 so as to produce the common contact Cc allowing the photoconduction signal to be read. And finally, a connection means C 2 is produced on the first ohmic contact layer 5 so that the voltage −Vref can be applied to the assembly of the structure.

FIG. 2 shows, in top view, a pair of detector elements like that of FIG. 1 . In the exemplary embodiment of FIG. 1, the stack of quantum well layers 2 in fact forms two elementary detectors, each of them being associated with a detector element 1 or 1 ′ for the purpose of eliminating the background noise as was explained in French patent No. 2 756 666.

The stacks of quantum well layers 1 , 1 ′, 2 are active or photoconducting at the same ranges of wavelengths λ. The stacks of layers 1 and 1 ′ are designed so as to be more absorbing at wavelengths λ than the stack of layer 2 . The stack of layer 2 is preferably designed so as to be virtually nonabsorbing. This may be achieved by different thicknesses of the stacks of layers or by greater doping of the quantum well layers of the more absorbing stacks.

Since the detector is illuminated by incident radiation on the ohmic contact layer 5 , the stack of layers 2 receives the radiation first. The latter therefore passes through the stack of layers 2 then the stacks of layers 1 and 1 ′ and they are coupled by the diffraction gratings 3 and 3 ′, depending on the polarization of the light in the stacks of layers 1 and 1 ′.

Although the diffraction grating 3 is provided in order to couple the vertically polarized light, the vertically polarized light will be diffracted toward the stack of layer 1 and will be absorbed or almost absorbed by this stack of layers. Similarly, although the diffraction grating 3 ′ is provided in order to diffract horizontally polarized light, the horizontally polarized light will be absorbed or almost absorbed by the stack of layer 1 ′.

FIG. 3 shows a matrix embodiment of a detector according to the invention. In this figure, four optical detectors as described above have been shown by way of example. Each detector therefore comprises two detector elements, each one with its diffraction grating. In this figure, the connection means making it possible to control, in a matrix manner, this matrix of detectors is essentially shown. It can therefore be seen in this FIG. 3 that the connection means C 1 are interconnected by their contact P 1 to the control potential V 1 . Similarly, the connection means C′ 1 are interconnected by their connection contact P′ 1 to the same control potential V 2 . The common contact layer, to which a contact P 2 is connected, is connected to the reference potential Vref. Finally, in the central part of each detector, a connection means Cc is located making it possible to connect an individual current measuring means for each detector.

In FIG. 3, a control circuit CU is also shown making it possible to deliver the reference potential Vref and the control potentials V 1 and V 2 .

A detection circuit DET receives information v 1 . 2 indicating the control mode delivered by the circuit CU (potentials V 1 and/or V 2 ) from the control circuit CU. It also receives read signals delivered by each pixel to its common contact means Cc.

FIG. 4 gives a control table for detectors produced in this way. This table thus illustrates the control of the detectors by the circuit CU. As can be seen in this FIG. 4, in operating mode, a reference potential Vref is applied to the various detectors. When a potential V 1 is applied to one detector and no potential V 2 is applied, this detector makes it possible to detect horizontally polarized light of wavelength λ. Conversely, if a potential V 2 is applied but no potential V 1 is applied to a detector, the latter detects the vertically polarized light (of wavelength λ). Finally, if both potentials V 1 and V 2 are applied to a detector, this makes it possible to detect both polarizations.

The normal operating mode may be considered as being the third mode, in order to acquire the maximum signal. The alternating use of modes 1 and 2 allows a polarimetry function to be added. Finally, in the event of optical countermeasure attack of one of the two polarizations, it is noted that the system can continue to operate and to acquire the signal in the other polarization.

An example of a method for producing a detector according to the invention will now be described. To simplify the description, the production of a detector comprising two detector elements such as the detector of FIG. 1 will be considered, but the application of this method could relate to producing a matrix of detectors.

This method comprises the following steps:

A first ohmic contact layer 5 , transparent to the wavelengths to be detected, is produced on a transparent substrate (not shown in FIG. 1 ).

A stack of quantum wells able to be photoconducting at a range of wavelengths λ is produced on this ohmic contact layer.

A second ohmic contact layer, transparent to the wavelengths to be detected, is then produced.

A second stack of quantum wells, able to be photoconducting at the range of wavelengths λ, is produced on this second ohmic contact layer.

Two diffraction gratings, oriented perpendicularly in order to diffract light to the second stack of quantum wells, each for different polarizations, are produced on this second stack.

The geometry of each detector is then defined by etching. To do this, the assembly of layers is etched to the first ohmic contact layer 5 . A groove is also produced between the two detector elements, at the boundary of the two gratings 3 and 3 ′, so that the stacks of layers 1 and 1 ′ of the two detector elements are also separated.

An insulating encapsulation layer 6 is deposited on the assembly thus obtained.

Finally, the connection means are produced. To do this, the following holes are made:

(a) the holes passing through the insulating layer to the diffraction gratings 3 and 3 ′ which are metalized in order to form the connection means C 1 and C′ 1 ;

(b) the common connection hole passing through the insulating layer and the second stack to the second ohmic contact layer in order to form the common connection means Cc allowing [lacuna] the read signal;

(c) a hole passing through the insulating layer and reaching the first ohmic contact layer 5 so as to produce the connection means C 2 .

It should be noted that, in the foregoing, the diffraction gratings 3 and 3 ′ may be metalized in order to apply the potentials V 1 and V 2 .

Thus, the method of the invention has made it possible to produce the detector of FIG. 1 . In order to produce a matrix of detectors, it would be appropriate to etch several detectors designed in this way on the first ohmic contact layer.