(19)
(11)EP 2 561 675 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
19.09.2018 Bulletin 2018/38

(21)Application number: 11816721.2

(22)Date of filing:  21.04.2011
(51)Int. Cl.: 
B23K 1/00  (2006.01)
B23K 1/20  (2006.01)
B23K 101/42  (2006.01)
B23K 1/19  (2006.01)
H01L 27/146  (2006.01)
B23K 103/00  (2006.01)
(86)International application number:
PCT/US2011/033496
(87)International publication number:
WO 2012/021182 (16.02.2012 Gazette  2012/07)

(54)

IMAGE SENSOR AND METHOD OF MAKING AN IMAGE SENSOR

BILDSENSOR UND VERFAHREN ZUR HERSTELLUNG EINES BILDSENSORS

CAPTEUR D'IMAGE ET METHODE DE FABRICATION D'UN CAPTEUR D'IMAGE


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 21.04.2010 US 764862

(43)Date of publication of application:
27.02.2013 Bulletin 2013/09

(73)Proprietor: Intevac, Inc.
Santa Clara, CA 95054 (US)

(72)Inventors:
  • COSTELLO, Kenneth, A.
    Union City CA 94587 (US)
  • RODERICK, Kevin, J.
    San Jose CA 95128 (US)
  • YIN, Edward
    Cupertino CA 95014 (US)
  • FOWLER, Douglas
    San Jose CA 95120 (US)

(74)Representative: Puschmann Borchert Bardehle Patentanwälte Partnerschaft mbB 
Postfach 10 12 31
80086 München
80086 München (DE)


(56)References cited: : 
US-A- 5 709 338
US-A1- 2005 258 212
US-A1- 2006 138 322
US-A1- 2009 045 478
US-A- 5 734 200
US-A1- 2006 043 438
US-A1- 2006 138 322
US-B2- 7 452 798
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description


    [0001] This invention relates to the field of bonding a collimator to an imager and, in particular, to using catch pads throughout the imaged area, and more particularly to an image sensor and a method of making an image sensor according to the preamble of claims 1 and 11 (see, for example, US2006/138322 A1).

    [0002] Focal plane imaging sensors are well known for the purpose of acquiring an electronic image from a sensor array implemented on a semiconductor die. Exemplary imaging sensors include CMOS imaging arrays (hereinafter, CMOS imagers) and charge coupled devices (hereinafter, CCD).

    [0003] Backthinning, removal of material from the backside of substrates, has been used to make the imaging sensors, particularly in the CCD applications. Performance advantages of backthinned sensors can include improved light sensitivity as a result of improved effective fill factor. When properly passivated by a method such as described in US 5 688 715, backthinned CMOS sensors can demonstrate high sensitivity to UV light and low energy (∼0,5 to 20 keV) electrons and low energy X-Rays (<∼10 keV). The sensitivity to low energy electrons of backthinned CMOS sensors makes them particularly suitable for use in vacuum tubes as a video based image intensifier. US 6 285 018 B1, details the use of a backthinned CMOS sensor in an electron-bombarded configuration, The backthinned CMOS sensor is mounted directly opposing a photocathode in a proximity-focused configuration.

    [0004] Image intensifiers experience a modulation transfer function (MTF) degradation of sensor image associated with elastic scattering of electrons as the electrons strike the anode of the tube. In a proximity-focused tube, the scattered (including backscattered) electrons are attracted to, and re-impact the anode within a circle of radius equal to ∼2× the tube gap. This effect, often referred to as "halo," is a particular problem when bright lights fall within the image intensifier field of view. There are a number of prior art approaches to minimize the impact of halo in image tubes incorporating a microchannel plate (MCP). US 6 483 231 attempts to minimize halo in the cathode to MCP gap, and US 5 495 141 attempts to minimize halo in the MCP to screen gap. However, in a tube without a microchannel plate, the image flux electrons lost in the collimator significantly reduce tube sensitivity. The glass draw technology described in US 5 495 141 is typically limited to open area ratios on the order of ∼ 80%. This estimate is roughly consistent with the statement that gain lost in the collimator can be regained by increasing the applied voltage from 6000 to 10000 V.

    [0005] Differences in the coefficient of expansion between the glass used to manufacture MCP-like structures and the silicon of CMOS die make it impossible to maintain pixel level alignments between a glass collimator and an electron bombarded active pixel imager over normal environmental temperature ranges. Modern dry etch technology is now capable of producing highly anisotropic etched structures in silicon. One method used to generate such structures is described in US 5 501 893, US 7 042 060 B2, describes collimator structures made using modern semiconductor techniques. Collimators made using anisotropic dry etching can exceed 90% open area ratio. Due to the exact coefficient of thermal expansion match between a silicon collimator and the silicon of a typical CMOS imager, silicon is typically used as the collimator material.

    [0006] One consequence of the anisotropic-etch approach, and the high associated open areas, is that the collimator lacks the structural rigidity found in an MCP. US 5 501 893 discloses placing the collimator in close proximity to the screen. The collimator is supported at the edge and spaced a few microns from the surface of the phosphor screen. US 7 042 060 B2 describes multiple approaches including the use of a monolithic collimator, an edge supported collimator and a collimator bonded over the full active area. Modern image intensifiers employ various means to control sensor gain over widely varying input light levels. Direct view sensors use a combination of duty cycle gating of the cathode to MCP voltage and MCP voltage control to achieve suitable output light levels. Consequently, the electrostatic environment between the output of the MCP and screen, seen by the collimator described in US 5 501, 93, is constant during normal operation. When collimators are used in proximity focused solid state imaging sensors that do not employ an MCP, gain control is primarily achieved via duty cycle gating of the cathode-anode voltage. Consequently, the collimator is exposed to the alternating electrostatic field associated with the gating voltage. The collimator is conductive in order to maintain a drift field in the channels of the collimator. Similarly, the collimator is maintained at the anode potential during duty cycle gating. The electric field between the collimator surface and the photocathode results in a physical force that attracts the two surfaces toward each other.

    [0007] In proximity focused solid state imaging sensors, optimum image resolution is achieved by minimizing photoelectron time of flight, Time of flight is in turn minimized by maintaining tight spacing between the cathode and anode. Spacing is typically limited by the required operational voltage of the sensor (typically between 500 and 8000V). The net result of close spacing and a relatively high, alternating electric field present at the surface of the collimator is that significant movement can occur in an edge supported collimator. Experience has shown that in the geometries used in commercial EBAPS sensors, this movement can exceed 100 microns, Increasing sensor gap by 100 microns is sufficient to measurably degrade sensor performance. Placing an edge supported collimator in close proximity to the anode in a gated electric field can result in physical damage to both the collimator and the anode. Similarly, momentum transfer from the moving collimator is transferred to the outside of the sensor vacuum envelope resulting in a measurable acoustic signature that changes with sensor varying high voltage gating conditions.

    [0008] US 7 042 060 B2 discloses two approaches that can address both the physical damage and acoustic signature issues. The first approach generates a monolithic collimator via the use of area selective backside thinning. In practice, it is a costly process to develop and a difficult process to implement at high yield. The cost can be minimized by fabricating discrete collimators via established semiconductor methods and subsequently bonding known good collimators to known good backside thinned imagers.

    [0009] The second approach requires the collimator to be bonded to the underlying solid state imager throughout the active imaging area of the sensor via the use of a bonding medium. In the case of proximity focused solid state imagers that use GaAs, InGaAs, InP or other semiconductor photocathodes, ultra high vacuum (UHV) compatibility constraints apply to an acceptable bonding medium. The bonding media must both exclusively be composed of low vapor pressure materials and be compatible with the high temperature bake-out profiles necessary to achieve UHV pressures. The bond is made between the collimator and the passivated surface (also referred to as passivated layer) of the back-side thinned semiconductor imager, and this passivation layer is required to achieve good collection of electrons generated near the back surface of the imager. In order to achieve the best possible performance, the passivation layer is quite thin (typically <∼500 Angstroms). Consequently, any interaction or contamination of the surface of the backside thinned imager resulting from the presence of the bonding media during UHV thermal processing may result in performance degradation of the sensor. Compatibility of the bonding media with the passivation layer represents an additional constraint on the bonding media.

    [0010] US 2006/138322 A1 discloses backthinning in an area selective manner is applied to CMOS imaging sensors for use in electron bombarded active pixel array devices. A further arrangement results in an array of collimators aligned with pixels or groups of pixels of an active pixel array providing improved image contrast of such image sensor. Provision of a thin P-doped layer on the illuminated rear surface provides both a diffusion barrier resulting in improved resolution and a functional shield for reference pixels. A gradient in concentration of P-doped layer optimizes electron collection at the pixel array.

    [0011] According to US 5 709 338 A a soldering is performed in such a manner that two solder portions having different melting points are put between a metallization on a substrate for joining and a part. The chip portion having a higher melting point is made thicker than a solder foil portion having a lower melting point. Initially, the part is held to the metallization by the solder chip having the higher melting point. The solder foil having the lower melting point is first melted in such a state and, after that, the solder chip having the higher melting point is melted. By such a soldering method, a reliable soldering can be performed, without causing a void on the undersurface of a part joint section, by one soldering operation, so that a presoldering operation becomes unnecessary.

    [0012] According to US 7 452 798 B2 a chip mounting substrate for bonding a semiconductor chip to a substrate, comprises a solder layer on the substrate, the solder layer being connectable to a semiconductor chip, wherein the solder layer comprises a layer including delta-phase crystal grains of an Au-Sn alloy at a surface of the solder layer. The solder layer comprising a layer including delta-phase crystal grains of an Au-Sn alloy is formed at a surface of the solder layer. On mounting a semiconductor chip on the substrate, the substrate and the solder layer are heated and an image of the solder layer is shot to perform an image evaluation to detect the timing of mounting the semiconductor chip on the solder layer of the substrate and a position of the chip.

    [0013] In view of the above, an image sensor and a method for making the same are required to bond the collimator to the solid state imager that do not suffer from the above drawbacks and maintain pixel level alignments between a glass collimator and an electron bombarded active pixel imager over normal environmental temperature ranges.

    [0014] The following summary of the invention is included in order to provide a basic understanding of some aspects and features of the invention. This summary is not an extensive overview of the invention and as such it is not intended to particularly identify key or critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concepts of the invention in a simplified form as a prelude to the more detailed description that is presented below.

    [0015] According to an aspect of the invention, an image sensor according to the invention comprises the features of claim 1.

    [0016] In a preferred embodiment, the image sensor includes a solid state semiconductor imager comprising a backside thinned pixel array sensor and having a passivated silicon surface wherein a portion of the passivated silicon surface includes a pixel array region; a silicon collimator; and a metal bond inside the pixel array region joining the passivated silicon surface of the solid state semiconductor imager to the silicon collimator.

    [0017] The metal bond may be a thermo-compression bond between a metallized catch pad and metal layer.

    [0018] The metal bond may be a brazed metallized catch pad and metal layer.

    [0019] The metallized catch pad may include a top metal layer, an adhesion layer and a barrier layer.

    [0020] The metallized catch pad may include a top metal layer and a barrier layer. The metallized catch pad may include a gold layer, a titanium layer and a SiOxNy layer.

    [0021] The collimator may further include an adhesion layer between a collimator body of the collimator and the metallized layer.

    [0022] The solid state semiconductor imager may be a backside thinned pixel array sensor having a passivated surface.

    [0023] The metal bond may be formed by a thermo-compression process. The metal bond may be formed by a brazing process. The metal bond may be formed by a soldering process.

    [0024] According to a further aspect of the invention, a method of making an image sensor of the invention comprises the features of claim 11. Joining the collimator with the solid state semiconductor imager to form the metal bond may include performing a thermo-compression process.

    [0025] Joining the collimator with the solid state semiconductor imager to form the metal bond may include performing a brazing process.

    [0026] Joining the collimator with the solid state semiconductor imager to form the metal bond may include performing a soldering process.

    [0027] The method may also include depositing a barrier layer on the solid state imaging sensor; depositing an adhesion layer on the barrier layer; and depositing a top metal layer on the adhesion layer.

    [0028] The method may also include depositing an adhesion layer on the collimator; and depositing a surface metal layer on the adhesion layer.

    [0029] The solid state semiconductor imager may include a metallic catch pad and the collimator may include a metallized layer, and joining the collimator with the solid state semiconductor imager to form the metal bond may include joining the metallic catch pad and the metallized layer.

    [0030] Preferred embodiments of the image sensor and the method according to the invention are characterized in the sub-claims.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0031] The accompanying drawings, which are incorporated in and constitute a part of this specification, exemplify the embodiments of the present invention and, together with the description, serve to explain and illustrate principles of the invention. The drawings are intended to illustrate major features of the exemplary embodiments in a diagrammatic manner. The drawings are not intended to depict every feature of actual embodiments nor relative dimensions of the depicted elements, and are not drawn to scale.

    Figure 1 is a schematic diagram showing an image sensor according to one embodiment of the invention;

    Figure 2 is a perspective cross-sectional diagram showing the operation of an aligned collimator according to one embodiment of the invention;

    Figure 3 is a top view of an aligned collimator assembly showing the location of bonds joining the collimator to the underlying solid state semiconductor imager according to one embodiment of the invention;

    Figure 4 is a detailed diagram showing the metal catch pad formed on the solid state semiconductor imager according to one embodiment of the invention; and

    Figure 5 is a schematic cross-sectional view of a collimator and imager assembly showing the metal bond and the collimator metallization according to one embodiment of the invention.


    DETAILED DESCRIPTION



    [0032] According to embodiments of the invention, a method is provided for bonding a collimator to an underlying solid state semiconductor imager via a metal bond. The metal bond is formed between a metalized layer on the surface of the collimator that faces the sensor anode and a metalized catch pad on the surface of the solid state semiconductor imager. The catch pads are interspersed throughout the imaging array.

    [0033] Metal bonds are advantageous because they avoid hydroscopic issues associated with some bonding materials such as Boron Oxide containing glasses (i.e., storage and subsequent UHV processing of the bonding anode assemblies is problematic because of the hydroscopic nature of the glass).

    [0034] Degradation of sensor optical performance can be minimized by limiting the area of the catch pads to that required to reliably mechanically stabilize the collimator. Bond coverage of about 0.2% of the image plane area is sufficient to achieve more than ten times the adhesion strength required to resist the electrostatic forces present in the sensor during operation. It will be appreciated that bond coverage of more or less than 0.2% of the image plane area may be used. Catch pad coverage within the image array may range between about 0.002% and about 20% image plane coverage.

    [0035] In addition, the spacing between the collimator and the solid state semiconductor imager should exceed the anticipated deflection under the electrostatic load. In one example, if the vertical deflection of the collimator in the span between isolated catch pad bonds is less than about 0.1 micron, a metal bond thickness of more than about 0.1 micron limits the residual deflection of the collimator so it does not result in an impact between the collimator and the underlying anode. A solid border may also be incorporated into the assembly to stiffen the sides of the collimator to limit vertical deflection on cantilevered collimator edges.

    [0036] An exemplary electron bombarded active pixel array component is shown in Figure 1. In Figure 1, the active pixel array component includes a vacuum enclosure 3 that houses a photocathode 12 in proximity focus with a specialized active pixel array sensor chip 13. The sensor chip 13 forms the anode for receiving proximity focused photoelectrons from the photocathode 12. The active pixel array sensor 13 is mounted with its backside facing the photocathode 12. In one embodiment, the backside of the substrate of the active pixel array sensor 13 is thinned. Thinning the backside of the substrate provides for a greater collection efficiency of the photoelectron excited carriers and reduces the crosstalk between pixels as a result of diffusion. The active pixel sensor array 13 may also include a passivated surface or layer. A collimator assembly 10 is mounted on the backside of the pixel array sensor chip 13. The collimator assembly 10 spatially filters scattered electrons from the sensor chip 13 thereby minimizing image artifacts and improving the modulation transfer function (MTF) of the photocathode based electron bombarded active pixel sensor.

    [0037] The base of the vacuum device 3 is a transparent faceplate 21. The photocathode 12 is positioned on the faceplate 21. Vacuum enclosure sidewalls 22 extend between the transparent faceplate 21 and the header assembly 23. The active pixel sensor chip 13 is positioned on the header assembly 23. The header assembly 23 also allows for electrical feedthroughs for applying and reading back control voltages 17, bias voltage 18 and signal output lines 20 associated with the active pixel array sensor chip 13. Control electronics (not shown) are used to read out and operate the pixel array.

    [0038] The photocathode 12 can be a GaAs, an InP/InGaAs, a transferred electron photocathode, a multi-alkali photocathode, or any other photocathode known to those skilled in the art. An electron accelerating voltage 37 is applied to the photocathode 12. Photoelectrons 15 are emitted from the photocathode 12 in response to incident light in the form of an optical image 9. Optical components (not shown) are typically employed to form the optical image on the photocathode 12. A projection system may be used to place the image on the photocathode 12.

    [0039] An anode voltage 18 is applied to the sensor 13. The photoelectrons 15 are accelerated by virtue of the difference between the accelerating voltage 37 and the anode voltage 18. In one embodiment, the accelerating voltage 37 is negative with respect to the anode voltage 18. This permits biasing the sensor 13 to near (or at) ground potential to simplify interfacing with other components. Control signals and bias voltages 17 are applied to the active pixel sensor 3 and a video output signal 20 may be taken off the sensor 13.

    [0040] The electron bombarded active pixel array sensor 13 is usefully employed in various systems, such as cameras, night vision imagers and the like. To that end, optical lenses or the like form an image of real objects on a focal plane which is ordinarily disposed to coincide with the photocathode 12. Under conditions of a sufficiently bright image, the focal plane is formed on the backside thinned surface of the pixel array and the bias may be adjusted to preclude the photoelectron flux 15 and operate in a conventional optical image regime. See, for example, US Patent No. 4,687,922, the entirety of which is hereby incorporated by reference. Taken together, the electron bombarded active pixel array component with optics, power supply and control electronics may be regarded as a camera system.

    [0041] In one embodiment, the image sensor 13 includes an array of nominally 10 micron square pixels with center to center displacement of approximately 10 microns and a 50 micron (height) collimator 10. It will be appreciated that the dimensions of the image sensor 13 and the collimator 10 may be less than or greater than these exemplary dimensions.

    [0042] Figure 2 illustrates the operation of backside thinned sensor 13 and the collimator assembly 10 in further detail. As shown in Figure 2, an elemental collimator 40 (i.e., an element of the collimator assembly 10) is aligned with the pixel array region 42, a four pixel sub-array. The four pixel sub-array 42 is disposed within and proximate the front side of the sensor chip 13. The collimator is aligned with the pixel array when the normal to the center of a pixel (or center of a group of pixels) is collinear with the long axis of the corresponding collimator.

    [0043] An exemplary backscattered electron trajectory 46 is shown terminating on the interior walls 43 of the elemental collimator 40. But for the collimator 10, such backscattered electrons would excite pixels other than that pixel directly on the initial trajectory 44 of a proximity focused electron, thus degrading resolution and contributing artifact such as the halo effect. Thus, the collimator 10 functions as a baffle in respect to scattered electrons. The backscattered electrons are a relatively small effect and the halo effect, as such, becomes significant for very bright image features; however, the suppression of backscattered electrons is desirable because it is a source of background noise and veiling glare in the image sensor, resulting in reduced MTF.

    [0044] The thermal budget of the solid state semiconductor anode imposes a constraint on the bond used to join the solid state imager 13 to the silicon collimator assembly 10. Standard CMOS imagers typically show signs of damage if peak processing temperatures exceed about 450°C. This 450°C temperature limit makes the use of many active braze materials difficult. Thermo-compression bonds, brazes and solders typically do not bond well directly to a silicon surface. Accordingly, in embodiments of the present invention, a catch pad is provided on the solid state semiconductor imager to form a high reliability bond between the imager 13 and the collimator 10.

    [0045] Figure 3 is a detailed view of the exposed backside thinned, passivated semiconductor image array surface 45 illustrating an embodiment in which catch pads are used to form the high reliability bond. The catch pads 104 are deposited onto the backside, passivated surface 44 of the semiconductor image array. The catch pads 104, shown by dotted circles, are provided inside and outside of the pixel array region 42. The catch pads 104 are interspersed across the array surface 45.

    [0046] In the illustrated embodiment, one catch pad is used for every four pixels. Consequently, a one megapixel imaging sensor employs on the order of 250,000 catch pads. The catch pads illustrated in Figure 3 cover about 0.2% of the focal plane area. It will be appreciated that these numbers are exemplary and that the imaging sensor may include less than or more than 250,000 catch pads, and the catch pads may cover less than or more than 0.2% of the focal plane area. In one embodiment, the catch pads do not exceed about 50% pixel coverage for a single pixel.

    [0047] Exemplary dimensions for the features of the image array surface are 10 micron pixel pitch, 40 micron collimator pitch, 1 micron collimator wall thickness and 4 micron diameter catch pads. It will be appreciated that these dimensions are only exemplary and that the dimensions may be less than or greater than the exemplary dimensions. The catch pad dimensions may be limited by the limitations of the existing lithography techniques. For example, currently available lithography techniques limit dimensions of the catch pads to about 0.1 microns. It will be appreciated that tradeoffs can be made between catch pad geometry and number. In one embodiment, the catch pads can take the form of a line segment that exceeds the collimator pitch so that there is a collimator wall intersection even if the collimator is not aligned to the underlying pixel area. For example, an alternate catch pad geometry is a 1 micron by 28.3 micron line segment aligned at a 45 degree angle to the pixel array.

    [0048] The catch pads 104 may be deposited using known deposition techniques, such as, for example, atomic layer deposition (ALD), electro-plating, chemical vapor deposition (CVD) and physical vapor deposition (PVD) processes, and the like, and combinations thereof.

    [0049] The collimator assembly 10, which includes a metallized surface, is then aligned and pressed into the catch pads 104. Thermal and pressure profiles are generated to generate a reliable metal bond between the collimator 10 and the image sensor 13. The pressure and temperature profile required to generate a reliable bond depends on the details of the metallization approach. In one example, gold to gold thermo-compression bond(s) may be used to join the collimator assembly 10 and the image sensor 3. In the case of gold to gold thermo-compression bonds, a temperature profile of 300°C for 10 minutes at a pressure sufficient to yield approximately a 30% compression of the overall gold thickness resulted in reliable bonds.

    [0050] Figure 4 illustrates the catch pad layer structure in further detail. The catch pad 104 includes a top layer 105, an adhesion layer 107 and a barrier layer 109. The adhesion layer 107 is positioned between the top layer 105 and the barrier layer 109.

    [0051] The top layer 105 is used to form the metal bond with the metallized layer of the collimator 10. The catch pad assembly 104 shown in Figure 4 may be used with, for example, gold to gold thermo-compression bonding. In one embodiment, the top layer 105 is gold. It will be appreciated that other metals may be used as the top layer. For example, copper or silver can be cleaned to yield a reliable bond. It will be appreciated, however, that the malleable, relatively non-reactive nature of gold provides certain advantages over other metals. Exemplary dimensions of the top layer 105 are 2 microns in diameter and 2 microns thickness. It will be appreciated that the metal thickness, however, may be any thickness between about 0.2 to more than 5 microns. In one particular embodiment, the thickness is two microns, which is sufficient to overcome surface irregularities without imposing excessive difficulty for the use of lift-off lithography.

    [0052] The adhesion layer 107 is provided to improve adhesion between the top layer 105 and the barrier layer 109. In one embodiment, the adhesion layer 107 is titanium. It will be appreciated that the adhesion layer 107 may be other reactive metal materials. Other exemplary materials include but are not limited to Chromium, Tungsten, Niobium, Vanadium, and the like, and alloys thereof (e.g., Ti - W). In one embodiment, the adhesion layer 107 is about 500 Angstroms thick. It will be appreciated that the thickness of the adhesion layer 107 may be less than or greater than 500 Angstroms.

    [0053] The barrier layer 109 is provided to prevent the overlying metal layers 105, 107 from reacting with or excessively stressing the underlying passivated, backside thinned semiconductor image array 13. The barrier layer 109 protects the sensor surface throughout the thermal and pressure cycles required for bonding and any subsequent UHV processing. In one embodiment, the barrier layer 109 is approximately 1000 Angstroms and is made of SiO2. Other exemplary materials that may be used for the barrier layer 109 include but are not limited to SiO, SiOxNy, Al2O3 ,and the like, and refractory metals, such as Tungsten, Molybdenum, Tantalum, Niobium, Platinum, and the like. It will be appreciated that the thickness of the barrier layer 109 may be less than or greater than 1000 Angstroms.

    [0054] Typically, there is sufficient leakage to drain off the photocathode charge without the need for a discrete "grounding" contact; however, if the nature and thickness of the barrier layer 109 results in an electrically isolated bond, a path is provided to assure that the collimator remains at the potential of the underlying anode. One such path eliminates the barrier layer 109 for bonds that lie outside of the imaging area (but the barrier layer 109 remains for those bonds that lie within the imaging area).

    [0055] It will be appreciated that in embodiments in which the barrier layer 109 is made of a refractory metal, the adhesion layer 107 and barrier layer 109 may be combined (i.e., a single layer may be deposited under the top layer 105).

    [0056] It will be appreciated that the top layer 105, adhesion layer 107 and barrier 109 may be deposited using known techniques such as for example atomic layer deposition (ALD), electro-plating, chemical vapor deposition (CVD) and physical vapor deposition (PVD) processes, and the like, and combinations thereof.

    [0057] Figure 5 illustrates the metal bond between the collimator 40 and the image sensor 13 in further detail. Two metal layers are deposited on the silicon walls of the collimator 40 so the collimator 40 can be joined with the catch pads 104. The two metal layers include an adhesion layer 62 and a surface layer 65.

    [0058] The adhesion layer 62 improves adhesion between the surface layer 65 and the silicon collimator 40. In one embodiment, the adhesion layer 62 is Titanium and 500 Angstroms in thickness. It will be appreciated that the adhesion layer 62 may be made of any reactive metal. Other suitable materials include but are not limited to Chromium, Tungsten, Niobium, Vanadium, and the like, and alloys thereof (e.g., Ti - W). It will be appreciated that the thickness of the adhesion layer 62 may be less than or greater than 500 Angstroms.

    [0059] In one embodiment, the bottom layer 65 is gold. Exemplary dimensions of the bottom layer 65 are 1 micron wide (e.g., the width of a collimator wall) and 2 microns thick. It will be appreciated that these dimensions are merely exemplary and that the dimensions may be less than or greater than these exemplary dimensions. In one embodiment, the full surface of the collimator 40 may be coated by the surface layer 65.

    [0060] In Figure 5, a residual separation is shown between the top of the free surface of the collimator (defined by layer 65) and the free surface of the imager 13. In one embodiment, the separation between the two surfaces is approximately 1 micron. This separation is sufficient to insure that residual movement of the collimator under gated proximity focused sensor operation does not result in imager damage. In the case of thermo compression bonds, geometry and pressure are primary factors in controlling the residual collimator-to-sensor gap in the regions between catch pads.

    [0061] In the embodiment shown in Figure 5 a thermo-compression bond is used to form the metal bond between the image sensor 13 and the collimator 10. A temperature of 300°C for 10 minutes at a sufficient pressure can yield about a 10% to about a 50% compression of the overall gold thickness, which results in reliable bonds.

    [0062] As described above, in other embodiments, the collimator 10 and image sensor 13 may be bonded using other bonding techniques such as, for example, brazing and soldering. In embodiments in which the collimator 10 and imaging sensor 13 are bonded using brazing, the layers 107 and 109 may be thickened to insure that a minimal residual collimator-to-sensor gap is maintained in the regions between the catch pads 104. In addition, the top layer 105 may instead be a braze alloy such as for example Au - Sn. In the example in which the top layer 105 is Au - Sn, the time and temperature of the bond may be set to allow inter-diffusion between the top layer 105 of the imaging sensor 13 and the bottom layer 65 of the collimator 40 to increase the melting point of the bond beyond that of the eutectic value. Braze metallizations that terminate in gold (Au) are advantageous because surface oxidation issues are avoided thereby improving braze reliability. Those skilled in the art of brazing can apply a wide variety of material systems to this bond while maintaining the disclosed geometry and meeting the constraints imposed by UHV processing. The small catch pad geometry (<∼5 microns wide) is important when high modulus higher melting point braze alloys are used. Other suitable solder / braze alloys include but are not limited to Gold - Silicon, Gold - Germanium and Gold - Tin. Transient liquid phase (TLP) braze-formulations / thermal profiles may also be used.

    [0063] Although all of the drawings depict catch pads that are aligned to the pixel array, the disclosed techniques and structures can also be used with an unaligned collimator structure. For example, a collimator structure that is hexagonal, and has a 17 micron pitch, 50 micron thick with a wall thickness of ∼1 micron can be used in conjunction with 20 micron long, 2 micron wide catch pads, and the image sensor can be a backside thinned electron bombarded active pixel sensor with a 10.8 micron pixel pitch.

    [0064] Pull testing confirmed the collimator adhesion strengths were within engineering limits. Elimination of anode damage due to collimator / anode impact and the reliability of the approach under repetitive gating load were demonstrated via extended gated high voltage reliability testing.

    [0065] Acoustic testing confirmed that switching from an edge bonded collimator to an area bonded collimator results in >10X reduction of the sound pressure level generated by the sensor during high voltage gated operation. Whereas the sound level of the edge-bonded collimator could be readily characterized, the area bonded collimator sensor employing metal bonds as described above did not result in a measurable increase in sound pressure level over the noise floor of the measurement system.

    [0066] Image testing of EBAPS sensors incorporating collimators bonded via metal bonds on catch pads showed the electron shadow of the collimator and catch pad are very stable over the life of the sensor. Catch pads and collimator wall thickness were designed such that individual pixel shadowing ranged from 0 to ∼40%. Limiting pixel shadowing to this range combined with the stable nature of the shadowing allows image correction algorithms to reliably correct for any reduced gain on the affected pixels. Although aligning collimator geometry to pixel geometry still results in the best possible image sensor performance, the stability of the fixed pattern noise combined with image correction algorithms allows non-aligned catch pad and collimator geometries to be used with minimal performance degradation, which, in turn, lowers the capital equipment cost required to produce the bonded collimator anode assemblies.

    [0067] In some embodiments of the invention, the bonding techniques disclosed herein can be used with front-surface electron sensitive read out integrated circuits (ROICs). In these embodiments, the underlying barrier layer may be omitted; the metal pads can rest directly on the SiOxNy barrier layer found in these sensors. Similarly, the catch pads on the front-side imagers can be built up from metal pads, providing direct electrical connection to the collimator.

    [0068] It should be understood that processes and techniques described herein are not inherently related to any particular apparatus and may be implemented by any suitable combination of components. Further, various types of general purpose devices may be used in accordance with the teachings described herein. The present invention has been described in relation to particular examples, which are intended in all respects to be illustrative rather than restrictive. Those skilled in the art will appreciate that many different combinations will be suitable for practicing the present invention.


    Claims

    1. An image sensor comprising:

    a solid state semiconductor imager comprising a sensor anode and a pixel array region; and

    a collimator (10) having a metallized layer that faces the sensor anode,

    characterized in that

    a plurality of metallized catch pads (104) are interspersed inside the pixel array region, and that

    the metallized layer is joined to the metallized catch pads (104) to form a metal bond.


     
    2. The image sensor of claim 1, characterized in that:

    the solid state semiconductor imager comprising a backside thinned pixel array sensor (13) having a passivated silicon surface (44) wherein a portion of the passivated silicon surface includes a pixel array region (45); and

    the metal bond inside the pixel array region (45) joins the passivated silicon surface (44) of the solid state semiconductor imager to the silicon collimator (10).


     
    3. The image sensor of claim 2, wherein the catch pads (104) comprise thermo-compressioned metallized catch pads (104) or brazed metallized catch pads (104) and the metal bond further comprises metal layers deposited on the silicon walls of the collimator (10).
     
    4. The image sensor of claim 2 or 3, wherein each of the catch pads (104) comprises a top metal layer (105), an adhesion layer (107) and a barrier layer (109) on the passivated silicon surface (44).
     
    5. The image sensor of claim 1 or 2, wherein each of the metallized catch pads (104) comprises a top metal layer (105) and a barrier layer (109) and, optionally, an adhesion layer (107).
     
    6. The image sensor of any of the preceding claims, wherein each of the metallized catch pads (104) comprises a gold layer, a titanium layer and a SiOxNy layer.
     
    7. The image sensor of claim 1, wherein the collimator (10) further comprises an adhesion layer between a collimator body of the collimator (10) and the metallized layer.
     
    8. The image sensor of any of the preceding claims, wherein the plurality of metallized catch pads (104) are positioned inside and outside of the pixel array region (45).
     
    9. The image sensor of any of the preceding claims, wherein the solid state semiconductor imager is a backside thinned pixel array sensor having a passivated surface.
     
    10. The image sensor of claim 1, wherein the metal bond is formed by a thermo-compression process or by a brazing process.
     
    11. A method of making an image sensor comprising a solid state semiconductor imager comprising a sensor anode an a pixel array region (45), wherein a metal layer that faces the sensor anode is deposited on a collimator (10),
    characterized by depositing a plurality of catch pads (104) onto a backside, passivated surface (44) of a semiconductor image array so as to be interspersed throughout the imaging array of a solid state semiconductor imager;
    collimator (10):

    aligning the collimator (10) with the solid state semiconductor imager; and

    joining the collimator (10) with the solid state semiconductor imager to form a metal bond between the collimator (10) and the solid state semiconductor imager, the metal bond positioned inside a pixel array region (45) of the solid state semiconductor imager.


     
    12. The method of claim 11, wherein joining the collimator (10) with the solid state semiconductor imager to form the metal bond comprises performing a thermo-compression process or a brazing process or a soldering process.
     
    13. The method of claim 11, wherein depositing a plurality of catch pads (104) comprising:
    depositing a barrier layer (109) on the solid state imaging sensor, and depositing a top metal layer (105) on the adhesion layer (107) and, optionally, depositing an adhesion layer (107) on the barrier layer (109).
     
    14. The method of claim 11, wherein the solid state semiconductor imager comprises a metallic catch pad and the collimator (10) comprises a metallized layer, and wherein joining the collimator (10) with the solid state semiconductor imager to form the metal bond comprises joining the metallic catch pads (104) and the metallized layer.
     


    Ansprüche

    1. Bildsensor, umfassend:

    einen Festkörper-Halbleiter-Bildaufnehmer mit einer Sensoranode und einem Pixelarray-Bereich; und

    einen Kollimator (10) mit einer metallisierten Schicht, die der Sensoranode zugewandt ist,

    dadurch gekennzeichnet,

    dass eine Vielzahl metallisierter Fangpads (104) innerhalb des Pixelarray-Bereichs verteilt angeordnet ist, und

    dass die metallisierte Schicht mit den metallisierten Fangpads (104) zu einer Metallverbindung verbunden wird.


     
    2. Bildsensor nach Anspruch 1, dadurch gekennzeichnet, dass der Festkörper-Halbleiter-Bildaufnehmer einen rückseitig verdünnten Pixelarraysensor (13) umfasst und eine passivierte Siliziumoberfläche (44) aufweist, wobei ein Teil der passivierten Siliziumoberfläche einen Pixelarray-Bereich (45) beinhaltet; und
    die Metallverbindung innerhalb des Pixelarray-Bereichs (45) die passivierte Siliziumoberfläche (44) des Festkörper-Halbleiter-Bildaufnehmers mit dem Siliziumkollimator (10) verbindet.
     
    3. Bildsensor nach Anspruch 2, bei dem die Fangpads (104) thermokomprimierte metallisierte Fangpads (104) oder hartgelötete metallisierte Fangpads (104) umfasst und die Metallverbindung des Weiteren Metallschichten umfasst, die auf den Siliziumwänden des Kollimators (10) aufgebracht sind.
     
    4. Bildsensor nach Anspruch 2 oder 3, bei dem jedes der Fangpads (104) eine oberste Metallschicht (105), eine Haftschicht (107) und eine Sperrschicht (109) auf der passivierten Siliziumoberfläche (44) umfasst.
     
    5. Bildsensor nach Anspruch 1 oder 2, bei dem jedes der metallisierten Fangpads (104) eine oberste Metallschicht (105) und eine Sperrschicht (109) sowie fakultativ eine Haftschicht (107) umfasst.
     
    6. Bildsensor nach einem der vorhergehenden Ansprüche, bei dem jedes der metallisierten Fangpads (104) eine Goldschicht, eine Titanschicht und eine SiOxNy-Schicht umfasst.
     
    7. Bildsensor nach Anspruch 1, bei dem der Kollimator (10) des Weiteren eine Haftschicht zwischen einem Kollimatorkörper des Kollimators (10) und der metallisierten Schicht umfasst.
     
    8. Bildsensor nach einem der vorhergehenden Ansprüche, bei dem sich die Vielzahl der metallisierten Fangpads (104) innerhalb und außerhalb des Pixelarray-Bereichs (45) befindet.
     
    9. Bildsensor nach einem der vorhergehenden Ansprüche, bei dem der Festkörper-Halbleiter-Bildaufnehmer ein rückseitig verdünnter Pixelarraysensor mit passivierter Oberfläche ist.
     
    10. Bildsensor nach Anspruch 1, bei den die Metallverbindung durch ein Thermokompressionsverfahren oder durch ein Hartlötverfahren gebildet wird.
     
    11. Verfahren zur Herstellung eines Bildsensors, der einen Festkörper-Halbleiter-Bildaufnehmer mit einer Sensoranode und einem Pixelarray-Bereich (45) umfasst, bei dem eine Metallschicht, die der Sensoranode zugewandt ist, auf einem Kollimator (10) aufgebracht wird,
    gekennzeichnet durch
    das Aufbringen einer Vielzahl von Fangpads (104) auf einer rückseitigen passivierten Oberfläche (44) eines Halbleiter-Bildarrays derart, dass sie in dem gesamten Bilderfassungsarray eines Festkörper-Halbleiter-Bildaufnehmers verteilt sind;
    einen Kollimator (10);
    das Ausrichten des Kollimators (10) mit dem Festkörper-Halbleiter-Bildaufnehmer; und
    das Verbinden des Kollimators (10) mit dem Festkörper-Halbleiter-Bildaufnehmer zur Bildung einer Metallverbindung zwischen dem Kollimator (10) und dem Festkörper-Halbleiter-Bildaufnehmer, wobei sich die Metallverbindung in einem Pixelarray-Bereich (45) des Festkörper-Halbleiter-Bildaufnehmers befindet.
     
    12. Verfahren nach Anspruch 11, bei dem das Verbinden des Kollimators (10) mit dem Festkörper-Halbteiter-Bildaufnehmer zur Bildung der Metallverbindung das Durchführen eines Thermokompressionsverfahrens oder eines Hartlötverfahrens oder eines Weichlötverfahrens umfasst.
     
    13. Verfahren nach Anspruch 11, bei dem das Aufbringen einer Vielzahl von Fangpads (104) folgendes umfasst:
    Aufbringen einer Sperrschicht (109) auf dem Festkörper-Bildsensor, und Aufbringen einer obersten Metallschicht (105) auf der Haftschicht (107), und fakultativ Aufbringen einer Haftschicht (107) auf der Sperrschicht (109).
     
    14. Verfahren nach Anspruch 11, bei dem der Festkörper-Halbleiter-Bildaufnehmer ein metallisches Fangpad umfasst und der Kollimator (10) eine metallisierte Schicht umfasst, und bei dem das Verbinden des Kollimators (10) mit dem Festkörper-Halbleiter-Bildaufnehmer zur Bildung der Metallverbindung das Verbinden der metallischen Fangpads (104) mit der metallisierten Schicht umfasst.
     


    Revendications

    1. Capteur d'image comprenant :

    un imageur à semi-conducteur à l'état solide comprenant une anode de capteur et une région de matrice de pixels ; et

    un collimateur (10) présentant une couche métallisée qui est tournée vers l'anode de capteur,

    caractérisé en ce que

    une pluralité de zones de capture métallisées (104) sont intercalées à l'intérieur de la région de matrice de pixels, et en ce que la couche métallisée est raccordée aux zones de capture métallisées (104) pour former une liaison métallique.


     
    2. Capteur d'image selon la revendication 1, caractérisé en ce que :

    l'imageur à semi-conducteur à l'état solide comprend un capteur de matrice de pixels à face arrière amincie (13) et présente une surface passivée en silicium (44), dans lequel une partie de la surface passivée en silicium comporte une zone de matrice de pixels (45) ; et

    la liaison métallique à l'intérieur de la région de matrice de pixels (45) raccorde la surface passivée en silicium (44) de l'imageur à semi-conducteur à l'état solide au collimateur en silicium (10).


     
    3. Capteur d'image selon la revendication 2, dans lequel les zones de capture (104) comprennent des zones de capture (104) métallisées ayant subi une thermo-compression ou des zones de capture (104) métallisées brasées et la liaison métallique comprend en outre des couches métalliques déposées sur les parois en silicium du collimateur (10).
     
    4. Capteur d'image selon la revendication 2 ou 3, dans lequel chacune des zones de capture (104) comprend une couche métallique supérieure (105), une couche adhésive (107) et une couche barrière (109) sur la surface passivée en silicium (44).
     
    5. Capteur d'image selon la revendication 1 ou 2, dans lequel chacune des zones de capture métallisées (104) comprend une couche métallique supérieure (105) et une couche barrière (109) et, éventuellement, une couche adhésive (107).
     
    6. Capteur d'image selon l'une quelconque des revendications précédentes, dans lequel chacune des zones de capture métallisées (104) comprend une couche d'or, une couche de titane et une couche de SiOxNy.
     
    7. Capteur d'image selon la revendication 1, dans lequel le collimateur (10) comprend en outre une couche adhésive entre un corps de collimateur du collimateur (10) et la couche métallisée.
     
    8. Capteur d'image selon l'une quelconque des revendications précédentes, dans lequel la pluralité de zones de capture métallisées (104) sont positionnées à l'intérieur et à l'extérieur de la région de matrice de pixels (45).
     
    9. Capteur d'image selon l'une quelconque des revendications précédentes, dans lequel l'imageur à semi-conducteur à l'état solide est un capteur de matrice de pixels à face arrière amincie présentant une surface passivée.
     
    10. Capteur d'image selon la revendication 1, dans lequel la liaison métallique est formée par un processus de thermo-compression ou par un processus de brassage.
     
    11. Procédé de fabrication d'un capteur d'image comprenant un imageur à semi-conducteur à l'état solide comprenant une anode de capteur et une région de matrice de pixels (45), dans lequel une couche métallique qui est tournée vers l'anode de capteur est déposée sur un collimateur (10),
    caractérisé par
    le dépôt d'une pluralité de zones de capture (104) sur une surface passivée arrière (44) d'une matrice d'images à semi-conducteur de manière à ce qu'elles soient intercalées sur toute l'étendue de la matrice d'imagerie d'un imageur à semi-conducteur à l'état solide ;
    un collimateur (10) ;
    l'alignement du collimateur (10) sur l'imageur à semi-conducteur à l'état solide ; et
    le raccordement du collimateur (10) à l'imageur à semi-conducteur à l'état solide pour former une liaison métallique entre le collimateur (10) et l'imageur à semi-conducteur à l'état solide, la liaison métallique étant positionnée à l'intérieur de la région de matrice de pixels (45) de l'imageur à semi-conducteur à l'état solide.
     
    12. Procédé selon la revendication 11, dans lequel le raccordement du collimateur (10) à l'imageur à semi-conducteur à l'état solide pour former la liaison métallique comprend la mise en oeuvre d'un processus de thermo-compression ou d'un processus de brasage ou d'un processus de soudage.
     
    13. Procédé selon la revendication 11, dans lequel le dépôt d'une pluralité de zones de capture (104) comprend :
    le dépôt d'une couche barrière (109) sur le capteur d'imagerie à l'état solide, et le dépôt d'une couche supérieure métallique (105) sur la couche adhésive (107), et, éventuellement, le dépôt d'une couche adhésive (107) sur la couche barrière (109).
     
    14. Procédé selon la revendication 11, dans lequel l'imageur à semi-conducteur à l'état solide comprend une zone de capture métallique et le collimateur (10) comprend une couche métallisée, et dans lequel le raccordement du collimateur (10) à l'imageur à semi-conducteur à l'état solide pour former la liaison métallique comprend le raccordement des zones de capture métalliques (104) et de la couche métallisée.
     




    Drawing


















    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description