Classifications

• • G—PHYSICS
  • G02—OPTICS
  • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
  • G02B5/00—Optical elements other than lenses
  • G02B5/20—Filters
  • G02B5/201—Filters in the form of arrays
• • G—PHYSICS
  • G02—OPTICS
  • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
  • G02B5/00—Optical elements other than lenses
  • G02B5/20—Filters
  • G02B5/28—Interference filters
  • G02B5/285—Interference filters comprising deposited thin solid films
  • G02B5/288—Interference filters comprising deposited thin solid films comprising at least one thin film resonant cavity, e.g. in bandpass filters
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
  • H01L27/14—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components sensitive to infrared radiation, light, electromagnetic radiation of shorter wavelength or corpuscular radiation and specially adapted either for the conversion of the energy of such radiation into electrical energy or for the control of electrical energy by such radiation
  • H01L27/144—Devices controlled by radiation
  • H01L27/146—Imager structures
  • H01L27/14601—Structural or functional details thereof
  • H01L27/1462—Coatings
  • H01L27/14621—Colour filter arrangements
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
  • H01L27/14—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components sensitive to infrared radiation, light, electromagnetic radiation of shorter wavelength or corpuscular radiation and specially adapted either for the conversion of the energy of such radiation into electrical energy or for the control of electrical energy by such radiation
  • H01L27/144—Devices controlled by radiation
  • H01L27/146—Imager structures
  • H01L27/14643—Photodiode arrays; MOS imagers
  • H01L27/14645—Colour imagers

Definitions

• the invention relates to a wavelength optical filtering structure and an image sensor which comprises an optical filtering structure according to the invention.
• the invention finds a particularly advantageous application for the production of small image sensors such as, for example, image sensors of miniature cameras of mobile phones.
• CCD Charge Coupled Device
• Some devices start by separating the three colors and then direct them to three image sensors. Others separate the colors directly to the surface of a single matrix of detectors: it is to this second type of sensor that the invention relates.
• FIG. 1 An example of a Bayer matrix, viewed from above, is shown in FIG. 1.
• the Bayer matrix shown in FIG. 1 is a 2x2 matrix (two rows x two columns). From left to right, the row 1 filters are Green and Red respectively, and the row 2 filters are Blue and Green respectively.
• the term "green filter” means a substantially transparent element for the green color and substantially opaque for all other colors of the visible spectrum.
• blue filter or red filter a substantially transparent element for the blue color or the color red and substantially opaque for all other colors of the visible spectrum.
• the APS CMOS sensor comprises a photosensitive semiconductor element 1, for example silicon, on the surface of which are formed photosensitive zones Zph and electronic circuits E1, a silica layer 2 in which are integrated electrical interconnections 3 which connect between they El electronic circuits, resin layers forming Blue B filters, Red R filters and Green V filters, a resin layer 4 and a set of microlenses MC.
• This sensor realization technique is currently well controlled.
• a disadvantage of this sensor is however its impossibility to eliminate the infrared. It is therefore necessary to add, after the fact, above the sensor, a glass sheet provided with a multilayer interference filter to eliminate the infrared.
• the resins are not very dense and it is currently necessary to put a resin thickness close to or greater than one micron to have a sufficient filtering effect.
• the pixel size of the recent image sensors is close to one micron (typically 2 ⁇ m). This dimension of the pixels then poses a problem when the rays arrive with a high angle of incidence on the surface of the sensor.
• Colored resins are also known to be easily inhomogeneous. The inhomogeneity of filtering is then all the more marked as the pixels are small. This also represents another disadvantage.
• Figures 3 and 4 show, respectively, a top view and a sectional view of a structure disclosed in the patent application FR 2 904 432. Filtering cells R, V, B are placed next to each other and select the respective colors Red, Green and Blue (see Figure 3).
• Figure 4 is a sectional view along the axis BB of Figure 3.
• the structure shown in Figure 4 comprises four dielectric layers D1-D4 and three metal layers M1-M3, a metal layer alternating with a dielectric layer.
• the layer D1 is in contact with a substrate S and the layer D4 is in contact with the incident medium which receives the light.
• Each of the two dielectric layers D1 and D4 has a constant thickness, as well as each of the three metal layers M1-M3.
• the two layers D2 and D3 each have a variable thickness depending on the filtering in which they participate.
• Three filter zones are then defined according to the thickness of the layers: a zone Zl for the Red, a zone Z2 for the Green and a zone Z3 for the Blue.
• a transition zone Za separates the zones Z1 and Z2 and a transition zone Zb separates the zones Z2 and Z3.
• Table 1 gives an example of numerical values for the thicknesses of the different dielectric and metallic layers of the structure of FIG. 4 as a function of the different zones.
• the notation e (Di) represents the thickness of the layer Di
• the dielectric layers are made of titanium oxide (TiO 2 ) and the metal layers are silver
• the number of intermediate dielectric layers of variable thicknesses influences the maximum transmission and the narrowness of the transmission window: in practice, a color filter of equal spectral width at 100nm (Blue, Green or Red) typically requires the presence of two intermediate dielectric layers surrounded by two metal layers and thus, according to the method described in the patent application FR 2 904 432, the formation of seven layers in total. It is then necessary to implement four separate lithographic printing processes, each lithographic process comprising a resin deposition, a lithography step, an exposure, a development, an etching and a residual resin removal. This represents a disadvantage.
• the two engravings are superimposed, which causes, on the one hand, a lateral shift related to the accuracy of the alignment between the levels and, d on the other hand, a step in thickness between the different filters.
• This recess is, in the example given above, of the order of 60 nm (40 to 80 nm) between the green and blue filters and of the order of 120 nm between the red and blue filters. From an optical point of view, the transitions between pixels then generate a diffraction artifact.
• FIGS. 5A and 5B illustrate the intensity I of an optical signal transmitted by a structure with two neighboring pixels as a function of the type of transition that exists between the two neighboring pixels.
• Figure 5A corresponds to an ideal transition between the neighboring pixels whereas
• Figure 5B corresponds to a real transition.
• the ideal transition is a frank transition between the pixels that leads to a step of intensity without disturbance while the real transition is a transition in thickness and misalignment that leads to an artifact of diffraction.
• This artifact reduces the usable area of each pixel and disrupts signal transmission. It is therefore difficult to make pixels of very small dimensions, for example of dimensions less than 1.5 .mu.m xl, 5 .mu.m.
• optical filtering structures are also known from the prior art, such as, for example, the structures disclosed in US Pat. No. 6,031,653, in patent application EP 1,592,067, in patent application US 2007/0146888. or in the international patent application WO 2008/017490.
• US Pat. No. 6,031,653 discloses an optical filtering structure consisting of two superimposed Fabry-Perot type cavities that share the same semi-reflective surface. Optical filtering is achieved by varying the thickness of the dielectric layers placed between the semi-reflective surfaces.
• the patent application EP 1 592 067 discloses a multilayer optical filtering structure comprising two multilayer films ⁇ / 4 separated by an insulating layer whose thickness is different from ⁇ / 4, ⁇ being the transmission wavelength. The thickness of the insulating layer varies according to the wavelength to be transmitted.
• the structure of the invention does not have the disadvantages mentioned above.
• the invention relates to an optical filtering structure composed of at least two neighboring elementary optical filters, an elementary optical filter being centered on an optimal transmission frequency, characterized in that it comprises a stack of n metal layers and of n-1 dielectric layers, n being an odd integer greater than or equal to 3, each metal layer alternating with a dielectric layer so that the central layer of the stack is a metal layer, each of the layers of the stack having a thickness constant except for the central metal layer whose fixed thickness varies the optimal transmission frequency of an elementary filter.
• the dielectric and metal layers of the stack located on either side of the central metal layer and which have the same rank relative to the central metal layer have a substantially identical thickness.
• the dielectric and metal layers of the stack located on either side of the central metal layer and which have the same rank relative to the central metal layer are not the same. thickness, the thickest dielectric and metal layers being all located on the same side with respect to the central metal layer.
• at least one metal layer of the stack results from the superposition of two elementary metal layers formed in different metallic materials. A first of the two elementary metal layers, called "hooked layer", then allows a better adhesion of the second elemental metal layer at one stack.
• the metal layers of the stack which are not formed of two elementary metal layers are made of silver and the second elemental metal layer of the metal layers of the stack which result from the superposition of two elemental metallic layers is also silver.
• At least one dielectric layer of the stack results from the superposition of at least two elementary dielectric layers formed in different materials.
• a first elemental dielectric layer located between a metal layer and the second elementary dielectric layer is an interface layer which makes it possible to avoid a chemical reaction between the metal of the metal layer and the second elemental dielectric layer and / or constitutes a barrier. diffusion of the metal of the metal layer in the dielectric.
• the dielectric layers which do not result from the superposition of two elementary dielectric layers and the second elementary dielectric layer of the two elementary dielectric layers of each dielectric layer formed of two elementary dielectric layers are produced, for example, in one of the following materials: titanium dioxide (TiO 2 ), nitride Titanium doped aluminum (AlTiN), Zinc sulphide (ZnS), Zinc-Silicon sulphide alloy (ZnS (x) -SiO 2 (1-x), alumina (Al2O3), silicon nitride (Si x N y ) , magnesium fluoride (MgF 2).
• a first dielectric layer, optical adjustment is placed on the metal layer located at a first end of the stack and / or a second dielectric layer
• the optical adaptation layer is placed on the metal layer at a second end of the stack, opposite to the first end.
• the first dielectric layer of adaptation, placed at the base of the stack is then a dielectric layer adapted to optically adapting the index between the stack and a substrate on which the stack is placed, likewise the second matching dielectric layer, placed on the surface of the stack, is then a dielectric layer capable of optically adapting index between the stack and a superstrate that covers the stack.
• the superstrate can be, for example, air.
• the first adaptation dielectric layer and / or the second adaptation dielectric layer have a thickness that varies from at least a first elementary optical filter to at least a second elementary optical filter.
• the first and second adaptation dielectric layers are produced in a material identical to the material which constitutes the dielectric layers of the stack.
• the elementary optical filters are arranged in matrix form.
• the matrix is a Bayer matrix for filtering the three colors Red, Green and Blue.
• the invention also relates to an optical sensor comprising an optical filtering structure and a photosensitive semiconductor substrate on which the optical filtering structure is deposited, characterized in that the optical filtering structure is a structure according to the invention, a layer metallic end of the stack or the first dielectric layer of adaptation being fixed on a first face of the semiconductor substrate.
• the optical filtering structure is an asymmetrical structure such that the thickest dielectric and metallic layers are all situated on the same side with respect to the central metallic layer, it is the dielectric and metal layers of the least thick stack which are placed between the first face of the semiconductor substrate and the central metal layer.
• an optical filtering structure of the invention for example a Bayer matrix, may be such that all the optical filters matrix elements that make up the structure have a thickness smaller than the shortest useful wavelengths.
• FIG. 1, already described represents a top view of a Bayer matrix according to FIG. prior art
• Figure 2, already described shows a sectional view of a CMOS APS sensor according to the prior art
• FIG. 3, already described represents a view from above of an optical filtering matrix structure of the prior art
• Figure 4, already described shows a sectional view of the optical filter matrix structure shown in Figure 3
• FIGS. 5A and 5B already described, represent the intensity of an optical signal transmitted by a structure with two neighboring pixels as a function of the type of transition that exists between the two neighboring pixels
• FIG. 1 represents a top view of a Bayer matrix according to FIG. prior art
• Figure 2 already described shows a sectional view of a CMOS APS sensor according to the prior art
• FIG. 3, already described represents a view from above of an optical filtering matrix structure of the prior art
• Figure 4 already described shows a sectional view of the optical filter matrix structure shown in Figure 3
• FIGS. 5A and 5B already described, represent the intensity of an optical
• FIG. 6A represents a sectional view of a first exemplary optical filter matrix structure according to the invention
• Figs. 6B-6D illustrate the operation of an optical filtering structure of the invention according to Fig. 6A
• FIG. 7A represents a sectional view of a second exemplary optical filter matrix structure according to the invention
• FIGS. 7B-7D show, respectively, blue, green and blue filter transmission curves.
• FIG. 8 represents the theoretical spectral response curves of the "CIE 1931" colorimetric standard for the blue, green and red colors
• FIG. 9 represents a UV-blue peak attenuation spectral response of a structure according to the invention.
• FIG. 6A represents a sectional view of a first example of an optical filtering structure according to the invention.
• the structure of FIG. 6A comprises two dielectric layers d2-d3 and three metallic layers ml-m3 placed on a substrate S.
• the layers ml, d2, m2 constitute a first Fabry-Perot cavity and the layers m2, d3 and m3 constitute a second Fabry-Perot cavity.
• the dielectric layers d2 and d3 are the cavity resonators and the metal layers ml, m2, m3 are mirrors that partially let the light through.
• the central metallic layer m2 which is common to the two cavities of Fabry-Perot has a variable thickness, the two dielectric layers d2, d3 and the two metal layers ml, m3 surrounding the central layer m2 each having a constant thickness.
• the structure of FIG. 6A illustrates the embodiment of the invention in which the stack of the dielectric and metal layers has a symmetrical structure with respect to the central metallic layer, the dielectric and metal layers situated on either side of the metal layer. the central metal layer and which have the same rank relative to the central metal layer having a substantially identical thickness.
• the thickness variation of the central metallic layer leads to define three filter zones Z1, Z2, Z3, the zone Z1 being the filtering zone of the blue (i.e. the transparent zone for the
• the filtering structure of the invention has a thickness total advantageously lower than the total thickness of the structure disclosed in the patent application FR 2 904 432 mentioned above.
• the basic structure corresponds to the Blue filter, which has a thickness between 180 and 300 nm depending on the structures and dielectric materials used. This results in an advantageously ultra-thin structure.
• the structure of the invention retains all the advantages already mentioned for the structure described in the patent application FR 2 904 432, namely a large angular tolerance, the possibility of integrating a metal electrode in the filter and the possibility to filter the infrared (IR) and the far ultraviolet (far UV).
• FIGS. 6B-6D show the transmission coefficient T of a structure according to FIG. 6A as a function of the energy E of the photons, for different thicknesses of the central metallic layer.
• FIG. 6B corresponds to a case where the thickness of the central metallic layer is large (in this case 41.4 nm / zone Z1 in FIG. 6A), curve 6B to a case where the thickness is average (in FIG. 15nm / zone Z2 in FIG. 6A) and layer 6C in a case where the thickness is small (in this case 6.5nm / area Z3 in FIG. 6A).
• the central metallic layer besides being a mirror, also behaves like an optical tunnel barrier.
• the eigen modes of each of the two cavities coexist. This results in a single lobe in frequency, the central frequency of the lobe being the resonance frequency of the cavity (transmission of the color Blue).
• the thickness decreases (FIG. 6C)
• the coupling between the eigen modes of the cavity is accentuated and it follows a sliding and a doubling of the frequency lobe, one of the lobes corresponding to a transmission of the Green and the other than a transmission in the Ultraviolet.
• the thickness decreases further (FIG.
• the filtering structure of the invention transmits, in the visible, photons whose energy is substantially identical from one color to another (for example 2.7eV) and the waves of the Infrared range are always stopped (see the curves 6B-6D).
• FIG. 7A represents a sectional view of a second example of optical filtering structure according to the invention.
• FIG. 7A shows an optical filtering structure which comprises, in addition to the aforementioned layers ml, d2, m2, d3, m3, two dielectric optical matching layers d1 and d4.
• a first optical matching dielectric layer d1 is placed between the substrate S and the metal layer ml and the second optical adaptation layer d4 is placed on the metal layer m3.
• the dielectric matching layers have a constant thickness from one elementary optical filter to the other.
• the invention also relates to the case where the optical matching dielectric layer (s) have different thicknesses from one elementary filter to another.
• Table 2 below gives an example of numerical values for the thicknesses of the different dielectric and metallic layers of the structure of FIG. 7B as a function of the different filtering zones Z1, Z2 and Z3.
• the dielectric layers are made of titanium oxide (TiO 2 ) and the metal layers are silver (Ag).
• the total thickness of the structure of the invention is very substantially less than the total thickness of a structure of the prior art having substantially identical filter performance.
• FIGS. 7B, 7C and 7D represent, respectively, the optical transmission coefficients T1, T2 and T3 of the respective filter pixels Red, Green and Blue of the structure represented in FIG. 6B (zones Z1, Z2 and Z3) as a function of the wavelength ⁇ .
• the optical transmission coefficient T1 of the red filter pixel is centered at a wavelength substantially equal to 605 nm and has a peak at substantially 380 nm.
• the optical transmission coefficient T2 of the green filter pixel is centered at a wavelength substantially equal to 540 nm and has a peak at substantially 406 nm.
• the optical transmission coefficient T3 of the blue filter pixel is centered at a wavelength substantially equal to 450 nm.
• the peaks at 605 nm, 540 nm and 450 nm of the respective transmission coefficients T1, T2, T3 correspond substantially to the maximum spectral responses of the "CIE 1931" standard, which are considered colorimetric references.
• the spectral responses R ( ⁇ ) of the "CIE 1931” standard are illustrated, for the record, in FIG. 8.
• the colorimetric response of an imager is all the better that the spectra of the pixels constituting the imager are closer to the spectra. of the "CIE 1931” standard. It is clear that the color filters of the invention are therefore very satisfactory from this point of view.
• the peaks at 380 nm and 406 nm of the respective transmission coefficients T1 and T2 are not useful. According to an improvement of the invention, means are provided for decreasing the amplitude of these peaks. There are three ways to perform this amplitude reduction.
• a first way is to achieve an asymmetrical stack with respect to the central metal layer.
• the dielectric and metallic layers that are between the substrate and the central metal layer are then of smaller thickness than the dielectric and metal layers that are between the central metal layer and the incident medium.
• Table 3 gives thickness values of dielectric and metal layers of a seven-layer asymmetrical structure (four dielectric layers, including two optical matching dielectric layers and three metal layers). :
• the thickness e (m3) of the layer m3 is substantially greater than the thickness e (ml) of the layer ml, and it would be the same between the metal layers of the same rank located on the side and another of the central metal layer of a structure having a greater number of metal layers, optimizing the stack to reduce the peak in the near UV systematically resulting in this characteristic.
• FIG. 9 represents the transmission coefficients T1, T2, T3 of the red, green and blue pixels of the asymmetrical structure of the invention whose table is given above (Table 3). We can observe a real decrease in unwanted peaks.
• a second way of reducing the amplitude of the undesirable peaks is the use, for producing the dielectric layers, of a material whose absorption cut-off wavelength is substantially around 400 nm, such as, for example, sulphide. of Zinc (ZnS).
• a third way is to simultaneously implement the first and second channels mentioned above.
• optical filtering structure of the invention is realized using microelectronic technological methods known per se.
• the dielectric and metal layers are deposited by vacuum cathode sputtering which is a "cold" process.
• vacuum cathode sputtering which is a "cold" process.
• Other techniques such as evaporation under vacuum are possible, however.
• the materials are, for example, silver (Ag) for producing the metal layers and titanium dioxide (TiO 2 ) for producing the dielectric layers.
• the control of the layer thicknesses can be done, for example, by the knowledge of the deposition rate.
• a first dielectric layer of constant thickness 20nm (layer d1)
• a first metal layer of constant thickness 15nm (layer ml)
• a second metallic layer of constant thickness 40 nm (layer intended to produce the layer m2).
• Two lithogravure steps follow the four deposition steps.
• a resin is used to protect the areas not to be etched.
• the etching is then done, for example reactive ion etching (for example CF4 + 02 gas).
• the stopping point of the etching is determined for example by optical interferometry.
• For the thickness values mentioned above see FIG. 6B and the associated table 3, namely: 36 nm for blue, 13 nm for green and 7 nm for red), after the deposition of the thick metal layer 36 nm, we do a first open masking on the "green" and "red” areas and is engraved on a thickness of 23nm. The first masking is then removed and a second one is made in which only the "red" areas are open and one is engraved again on a thickness of 6 nm. The metal layer of variable thickness is then formed. Then remove the resin

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Optics & Photonics (AREA)
• Engineering & Computer Science (AREA)
• Power Engineering (AREA)
• Electromagnetism (AREA)
• Condensed Matter Physics & Semiconductors (AREA)
• Computer Hardware Design (AREA)
• Microelectronics & Electronic Packaging (AREA)
• Solid State Image Pick-Up Elements (AREA)
• Optical Filters (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=40612818

Family Applications (1)

Country Status (4)

Families Citing this family (17)

Family Cites Families (8)

• 2008
  • 2008-10-22 FR FR0857157A patent/FR2937425B1/fr not_active Expired - Fee Related
• 2009
  • 2009-10-20 EP EP09736973A patent/EP2347290A1/de not_active Withdrawn
  • 2009-10-20 US US13/124,007 patent/US8675280B2/en not_active Expired - Fee Related
  • 2009-10-20 WO PCT/EP2009/063750 patent/WO2010046369A1/fr active Application Filing

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events