JP2017090687A - Near-infrared absorbing glass wafer and semiconductor wafer laminate - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an optical filter which enables miniaturization and height-reduction of camera modules and offers superior uniformity of optical characteristics and high manufacturing productivity.SOLUTION: A near-infrared absorbing glass wafer 10 of the present invention comprises a glass wafer 12 and an absorption layer 11 formed on at least one principal surface of the glass wafer 12, the absorption layer 11 containing a transparent resin and an absorptive pigment and being configured to satisfy -15%≤Δt/t≤15%, where trepresent a mean thickness and Δt represents a thickness difference relative to the mean thickness t.SELECTED DRAWING: Figure 1A

Description

  Near-infrared absorption type glass that has an optical filter function for an imaging device for visible light images and is used integrally with a silicon semiconductor wafer (hereinafter also referred to as “Si wafer”) having a plurality of solid-state imaging devices including a photodetector array. The present invention relates to a semiconductor wafer laminate in which a wafer and a Si wafer and a near-infrared absorbing glass wafer are integrated.

  An imaging device (camera module) using a solid-state imaging element uses a sensitivity correction optical filter that transmits visible light and blocks near-infrared light.

  As the optical filter, near infrared volume absorption type glass (hereinafter also referred to as “volume absorption type glass”) in which CuO or the like is added to fluorophosphate glass or phosphate glass is used, and the addition concentration of CuO or the like is used. The transmittance of light having a wavelength of 700 to 1150 nm can be adjusted by adjusting the thickness. For example, the volume-absorbing glass exhibits an absorption maximum at a wavelength of 800 to 1000 nm. However, when the absorption capacity is increased, the transmittance of visible light may be reduced. On the other hand, if the decrease in visible light transmittance is suppressed, the near infrared light transmittance may increase and stray light may be generated. For this reason, an optical filter using volume absorption type glass sometimes uses a reflective layer made of a dielectric multilayer film in order to reduce the transmittance of near-infrared light. Specifically, an optical filter that has a reflective layer on a volume absorption glass, reflects near ultraviolet light, and absorbs and reflects near infrared light is disclosed (see Patent Document 1).

  Also, an optical filter comprising an absorption layer containing a near-infrared absorbing dye and a resin on at least one surface of a transparent glass that does not absorb visible light or near-infrared light, and a reflective layer that reflects near-infrared light (patent) Reference 2) is disclosed. Furthermore, an optical filter (see Patent Document 3) including an absorption layer containing a near-infrared absorbing pigment on at least one surface of the volume absorption glass and a reflection layer that reflects near-infrared light is also disclosed. These optical filters have a small incident angle dependency of the spectral transmittance curve, and can reduce stray light by absorption.

  In recent years, along with the downsizing of imaging devices, there has been a demand for miniaturization and low profile of camera modules, and there is a demand for thinner optical components and fewer parts in the camera modules. For this reason, studies have been made to integrate a solid-state imaging device and an optical component such as near-infrared absorbing glass including volume absorbing glass, for example.

The solid-state imaging device has a photodetector array of CMOS or CCD structure in which pixels of 1 to 4 μm square size are two-dimensionally arranged in hundreds of thousands to millions. Furthermore, the solid-state imaging device has an RGB mosaic color filter for each pixel on the incident side of the photodetector for generating a color image, and further collects incident light on the light receiving surface of the photodetector on each pixel. Has resin microlenses.
Here, when integrating the solid-state imaging device and the near-infrared absorbing glass, the productivity is low when the individual pieces of the near-infrared absorbing glass are individually integrated. Productivity can be improved by integrating near-infrared absorbing glass in a wafer state as part of a wafer process for forming a filter and a resin microlens for each solid-state imaging device.

  Here, by joining a wafer-like near-infrared absorbing glass (near-infrared absorbing glass wafer) to a semiconductor wafer, the camera module can be expected to be reduced in size, height and productivity. The size of the formed Si wafer is often 15 cm or more in diameter, and the use of a thin near-infrared absorbing glass is a prerequisite for reducing the height of the camera module. That is, for example, it has been desired to realize an optical filter that is a near infrared absorption type glass wafer having a diameter of 15 cm or more and a thickness of 0.4 mm or less and that is excellent in uniformity of optical characteristics within the effective surface of the wafer.

International Publication No. 14/034386 International Publication No. 12/169447 International Publication No. 14/030628

  The present invention is a near-infrared absorbing glass wafer having an absorption layer containing a near-infrared absorbing dye and a transparent resin on at least one surface of a glass wafer whose surface is an optical mirror surface. A near-infrared-absorbing glass wafer that provides near-infrared optical characteristics with excellent spectral uniformity by approximating the spectral sensitivity of a solid-state image sensor to visual sensitivity and blocking near-infrared light, and such near-infrared-absorbing glass It aims at providing the semiconductor wafer laminated body provided with the wafer.

The present invention includes a glass wafer and an absorption layer on at least one main surface of the glass wafer, the absorption layer contains a transparent resin and an absorption pigment, and the absorption layer has an average thickness of t _0. Provided is a near-infrared absorbing glass wafer that satisfies −15% ≦ Δt / t ₀ ≦ 15%, where Δt is the difference in film thickness with respect to the average value t ₀ .

  Moreover, this invention provides the semiconductor wafer laminated body which has a solid-state image sensor provided on Si wafer, and said near-infrared absorption type glass wafer.

The present invention has an absorption layer containing a near-infrared absorbing dye and a transparent resin on a glass wafer, the film thickness and the film thickness distribution are within a predetermined range within the effective surface of the glass wafer, and the spectral transmittance curve It is possible to provide a near-infrared absorbing glass wafer having a small distribution and excellent optical characteristics, and a camera module that is reduced in size and height.
Furthermore, the present invention realizes a reduction in size and height of the camera module and an improvement in productivity in manufacturing the camera module, as compared with the conventional case where an optical filter having a solid-state image sensor size is separated from the solid-state image sensor.

Sectional drawing which shows the example of a near-infrared absorption type glass wafer roughly. Sectional drawing which shows the other example of a near-infrared absorption type glass wafer roughly. Sectional drawing which shows the other example of a near-infrared absorption type glass wafer roughly. Sectional drawing which shows the optical system of a camera module roughly. Sectional drawing which shows roughly the optical path change of the condensing light beam of the center pixel light-receiving surface of a solid-state image sensor of a near-infrared absorption type glass wafer. The graph which shows the calculation result of beam diameter (phi) of the light-receiving pixel surface with respect to the thickness distribution of a near-infrared absorption type glass wafer about the imaging lens from which F number differs. In-plane distribution of the thickness of the near-infrared absorbing material in the near-infrared absorbing glass wafer (Δd / d ₀ of the volume-absorbing glass or Δt / t ₀ of the near-infrared absorbing layer) and the wavelength λ (T _50% ) The graph which shows the relationship of transmissivity difference (DELTA) T. In the near-infrared absorbing glass wafer, the in-plane distribution of the thickness of the near-infrared absorbing material (Δd / d _{0 of the} volume-absorbing glass) with respect to the ratio of the transmittance change ΔT to the wavelength change Δλ in the vicinity of λ (T _50% ). or graph showing calculation results of distribution Δλ of accompanying Delta] t / _{t 0)} of the near-infrared absorption layer λ _{(T 50%).} The perspective view which shows the example of a semiconductor wafer laminated body roughly. The expanded view of the cross section which shows the example of a semiconductor wafer laminated body roughly. The expanded view of the cross section which shows the other example of a semiconductor wafer laminated body roughly. The graph which shows the spectral transmittance of the reflection layer which consists of dielectric multilayers (incident angle: 0 degree). 5 is a graph showing the spectral transmittance of the near-infrared absorbing glass wafer of Example 1. 6 is a graph showing the spectral transmittance of the near-infrared absorbing glass wafer of Example 2. Sectional drawing which shows the principal part of the example of a camera module roughly.

The near-infrared absorption type glass wafer of this invention is demonstrated using FIG. 1A-FIG. 1C.
FIG. 1A shows a near-infrared absorbing glass wafer 10 having an absorption layer 11 that absorbs near-infrared light on one side of a glass wafer 12 having a parallel plane shape.

(Glass wafer)
The glass wafer 12 is made of a glass material that is transparent to visible light having a wavelength of at least 450 to 600 nm. For example, the planar shape may be a substantially circular shape having a diameter of 15 cm or more, but may be non-circular or polygonal. . Although the thickness is not limited, for example, a thickness of 0.1 to 0.4 mm can be used to reduce the size of the imaging apparatus. Further, the glass wafer 12 may be made of a glass material in which a local refractive index distribution such as striae that causes resolution deterioration of the solid-state imaging device and pixel defects, and bubbles do not remain. Furthermore, the surface of the glass wafer 12 only needs to have a surface flatness that can suppress generation of scattered light and transmission wavefront aberration that cause resolution degradation of the solid-state imaging device, and both sides are mirror-finished. May be.

  Since the near-infrared absorption type glass wafer 10 of the present invention joins a glass wafer and a Si wafer using an optical adhesive, the glass wafer material has a lower alkali component content, leading to deterioration of semiconductor operation. However, it is often preferable that high adhesion and reliability can be obtained. Among these, borosilicate glass is easy to process and can suppress the occurrence of scratches and foreign matters on the optical surface.

The material of the glass wafer that is transparent from visible light to near-infrared light is, for example, AF33 manufactured by Schott, Tempax (trademark), D263, B270, SW-3, SW-Y, SW manufactured by Asahi Glass. -YY, AN100, EN-A1, FP1, FP01eco and the like (product names) are preferable.
In addition, when using the glass wafer containing an alkali component, it is good to provide the passivation film which interrupts | blocks the movement of an alkali element at least on the Si wafer joint surface side. Examples of the passivation film include dielectric films with little visible light absorption such as SiO ₂ , SiO _x N _y , and Si ₃ N _4. Sputtering, CVD, ion-assisted deposition, sol-gel, etc. effective for forming a dense film The film can be formed with a predetermined film thickness by the film forming method.

  Furthermore, near infrared light (wavelength 700-1100 nm) is absorbed as a glass material as described above, for example, CuO-containing fluorophosphate glass or CuO-containing phosphate glass (these are referred to as “CuO-containing glass”), etc. Near infrared volume absorption type glass may be used.

  The CuO-containing glass has an absorption maximum at a wavelength of 750 to 1000 nm in an absorption spectrum at a wavelength of 400 to 1100 nm, and the transmittance can be adjusted by the CuO content and thickness. FIG. 1B shows a near-infrared absorbing glass wafer 20 having an absorbing layer 11 on a glass wafer 13 made of near-infrared volume absorbing glass. The glass wafer 13 exhibits high transmittance with visible light and can absorb near infrared light that cannot be sufficiently blocked by the absorption layer 11 alone.

Typical examples of the CuO-containing glass include the following compositions. The “phosphate glass” includes silicic acid phosphate glass in which a part of the glass skeleton is composed of SiO ₂ .

(1) represented by _{mass%, P 2 O 5 46~70%,} AlF 3 0.2~20%, LiF + NaF + KF0~25%, MgF 2 + CaF 2 + SrF 2 + BaF 2 + PbF 2 1~50%, however, F 0 The glass which contains CuO: 0.5-7 mass part by the outer split with respect to 100 mass parts of basic glass containing 0.5-32% and O 26-54%.

(2) represented by _{mass%, P 2 O 5 25~60%,} Al 2 OF 3 1~13%, MgO 1~10%, CaO 1~16%, BaO 1~26%, SrO 0~16%, _{ZnO 0~16%, Li 2 O 0~13} %, Na 2 O 0~10%, K 2 O 0~11%, CuO 1~7%, ΣRO (R = Mg, Ca, Sr, Ba) 15~ _{40%, ΣR '2 O (} R' = Li, Na, K) 3~18% ( however, ^{O 2-ions} to 39% molar amount F ^- ions is substituted with) glass consisting of.

(3) By mass%, P ₂ O ₅ 5 to 45%, AlF _{3 1 to} 35%, RF (R is Li, Na, K) 0 to 40%, R′F ₂ (R ′ is Mg, Ca , Sr, Ba, Pb, Zn) 10 to 75%, R ″ F _m (R ″ is La, Y, Cd, Si, B, Zr, Ta, m is a number corresponding to the valence of R ″) 0 Glass containing 15% (but up to 70% of the total fluoride weight can be replaced by oxide), and 0.2-15% CuO.

(4) In terms of cation%, P ⁵⁺ 11 to 43%, Al ³⁺ 1 to 29%, R cation (total amount of Mg, Ca, Sr, Ba, Pb, Zn ion) 14 to 50%, R ′ cation ( Total amount of Li, Na, K ions) 0 to 43%, R ″ cation (Total amount of La, Y, Gd, Si, B, Zr, Ta ions) 0 to 8%, and Cu ²⁺ 0.5 to 13 %, And further contains F ^- 17 to 80% in terms of anion%.

(5) In terms of cation%, P ⁵⁺ 23 to 41%, Al ³⁺ 4 to 16%, Li ⁺ 11 to 40%, Na ⁺ 3 to 13%, R ²⁺ (Mg ²⁺ , Ca ²⁺ , Sr ²⁺ , Ba ²⁺ , Zn ²⁺ ) 12 to 53% and Cu ²⁺ 2.6 to 4.7%, and further anion% F ⁻ 25 to 48% and O ² 52 to 75%.

(6) represented by _{mass%, P 2 O 5 70~85%,} Al 2 O 3 8~17%, B 2 O 3 1~10%, Li 2 O 0~3%, Na 2 O 0~5% , K ₂ O 0-5%, provided that 0.1% of CuO is 0.1% by weight based on 100 parts by mass of the basic glass composed of Li ₂ O + Na ₂ O + K ₂ O 0.1-5% and SiO ₂ 0-3%. Glass containing 5 parts by mass.

  Examples of commercially available products include, for example, (1) glass, NF-50E, NF-50EX, NF-50T, NF-50TX (manufactured by Asahi Glass Co., Ltd., trade name), etc., and (2) glass, BG Examples of the glass of (5) such as -60, BG-61 (manufactured by Shot Corporation, trade name), and the like include CD5000 (made by HOYA, trade name).

The above-described CuO-containing glass may further contain a metal oxide. When the metal oxide contains, for example, one or more of Fe ₂ O ₃ , MoO ₃ , WO ₃ , CeO ₂ , Sb ₂ O ₃ , V ₂ O _5, etc., the CuO-containing glass has ultraviolet absorption characteristics. The content of these metal oxides, relative to the CuO-containing glass 100 parts by _weight, the _{Fe 2 O 3, MoO 3,} WO 3 and at least one selected from the group consisting of CeO _{2, Fe} 2 _{O 3} 0.6-5 parts by mass, MoO ₃ 0.5-5 parts by mass, WO ₃ 1-6 parts by mass, CeO ₂ 2.5-6 parts by mass, or Fe ₂ O ₃ and Sb ₂ O ₃ Fe ₂ O ₃ 0.6 to 5 parts by ₊ Sb ₂ O 3 0.1 to 5 parts by weight, or _V 2 _{O 5} two and CeO _{2 V} 2 _O 5 0.01 to 0.5 parts by + CeO ₂ It is good to set it as 1-6 mass parts.

  In addition, the near-infrared absorption type glass wafer of the present invention can be expected to reduce the size and thickness of the imaging device if it also includes a cover glass function for protecting the solid-state imaging device. If the glass wafer contains an α-ray emitting element (radioisotope) as an impurity, the α-ray emitting element may be emitted to cause a soft error in the solid-state imaging device. It is recommended to use a glass material with a small amount of high purity. In the glass raw material, the content of U and Th among the α-ray emitting elements is preferably 20 ppb or less, and more preferably 5 ppb or less. Moreover, the near-infrared absorbing glass wafer may be provided with a film that shields α rays on one side close to the solid-state imaging device.

  The glass material can be molded by being drawn into a glass plate after being melted and cooled, and a float method, a downdraw method, an overflow fusion method, or the like can be used. Depending on the production method, a glass wafer having a thickness of 0.1 to 0.4 mm with an optical mirror surface can be obtained. However, in order to obtain a glass wafer having a thickness distribution described later, the glass substrate is processed, for example, with a diameter of 15 cm or more. Then, it may be processed so as to obtain an optical mirror surface having a thickness of 0.1 to 0.4 mm by double-side polishing.

  Next, consider an optical system in which a visible light image is focused on a solid-state imaging device through the near-infrared absorbing glass wafer of the present invention. FIG. 2 is a schematic diagram showing an optical system of the camera module, and FIG. 3 is a schematic diagram showing an optical path focused on the central pixel light receiving surface 16 of the solid-state imaging device on the optical axis of the imaging lens 31. . Here, with reference to FIG. 2 and FIG. 3, the influence on the resolution from the amount of blur on the imaging plane caused by the thickness distribution of the glass wafer and the pixel size of the solid-state imaging device will be described.

In FIG. 3, the image of the subject is formed on the pixel light receiving surface 16 of the solid-state image sensor joined to the near-infrared absorbing glass wafer 10 (20, 30) by the imaging lens 31. In FIG. 3, the total thickness of the near-infrared absorbing glass wafer 10 (20, 30) composed of the glass wafer having the thickness d ₀ and the absorbing layer having the thickness t ₀ and the adhesive 21 having the thickness g is the surface of the glass wafer. Corresponds to the distance d from the pixel light receiving surface 16 of the solid-state imaging device. The subject image is incident from the glass wafer surface at a convergence angle θ of numerical aperture NA = 1 / (2F) = sin θ corresponding to the F number (= F) of the imaging lens, and is condensed on the pixel light receiving surface 16. The
The glass wafer thickness d ₀ is the average thickness of the glass wafer 12 (13), and the absorption layer thickness t ₀ is the average thickness of the absorption layer 11. The glass wafer thickness d ₀ is 10 times or more thicker than the total thickness of the absorption layer thickness t ₀ and the adhesive layer thickness g, and each thickness distribution has the same ratio, so the difference value of the thickness d ₀ of the distance d Δd can be approximated by a difference value of the glass wafer thickness d ₀ .

Here, when the difference value of the thickness d ₀ of the glass wafer is Δd (≠ 0), the distance that can be taken between the glass wafer surface and the pixel light receiving surface 16 of the solid-state imaging device is d ₀ + Δd, and FIG. As shown in FIG. 5, the condensing position in the optical axis direction is shifted by the distance Δ. As a result, the image formation point on the pixel light receiving surface is enlarged to the diameter φ. Here, light incident at an incident angle θ on a glass wafer having a refractive index n is
sin θ ′ = sin θ / n
The amount of shift Δ is as follows.
Δ = Δd × (tan θ−tan θ ′) / tan θ ′
The beam diameter (blur) of the enlarged image point is
φ = 2 × | Δ | × tan θ ′
It is calculated from.

Here, when an imaging lens having an F number of 2.0 and 1.4 is used and the difference value Δd of the thickness d ₀ of the glass wafer is given in a range of ± 50 μm, the beam diameter φ of the light receiving pixel surface is A calculation example is shown in FIG. If the beam diameter φ is larger than the pixel size of the solid-state image sensor, an image cannot be reproduced with a resolution that can be accepted by the solid-state image sensor. It is important to obtain a difference value Δd of the wafer thickness d ₀ .

  The minimum pixel size of the CMOS semiconductor solid-state imaging device is about 1 μm □, but the RGB color filter is formed with 4 pixels to form one color imaging pixel, so the beam diameter φ at the minimum pixel size must be 2 μm or less. It becomes. In addition, the actual camera module depends on the resolution (MTF) performance of the imaging lens in addition to the pixel size of the solid-state imaging device. Therefore, an imaging lens design that suppresses chromatic aberration with respect to visible light, resolution degradation in peripheral pixels of an off-axis solid-state imaging device, distortion of the imaging surface, and the like is required. As described above, since the imaging beam size on the pixel surface of the solid-state imaging device may increase due to manufacturing variations of the imaging lens, the thickness of the glass wafer 12 (13) and the near-infrared absorbing glass wafer 10 (20, 30). The lower the resolution degradation due to the distribution, the better.

From the calculation result (beam diameter φ) in FIG. 4, the difference value Δd of the (near infrared absorption type) glass wafer thickness d ₀ is within ± 30 μm, although it depends on the pixel size of the solid-state imaging device and the resolution of the imaging lens. It is sufficient that it is within ± 20 μm, and more preferably within ± 10 μm.

As described above, the difference value Δd of d ₀ has been described. However, since the imaging light beam on the outer peripheral portion of the solid-state imaging device is obliquely incident on the imaging lens 31, astigmatism and coma are compared with the light beam on the optical axis. Aberration tends to remain, and it is difficult to design a lens that ensures high resolution (MTF). As a result, the imaging light beam on the outer peripheral portion of the solid-state imaging device is more easily blurred than on the optical axis, and astigmatism and coma due to the thickness difference value Δd are likely to occur. .
In addition, resolution degradation caused by the shift amount Δ of the condensing position in the optical axis direction can be countered at a certain level by adjusting the position of the imaging lens 13 by assembling the camera module or by autofocusing. Correction of point aberration and coma is difficult, and in that sense, it is necessary to reduce the difference value Δd of d ₀ .

Next, the thickness distribution in the case where CuO-containing glass (also referred to as “CuO-containing glass wafer”) is used as the glass wafer will be described.
Here, using the near-infrared absorption characteristics of CuO-containing glass, the wavelength at which the transmittance for light with a wavelength of 620 to 690 nm is 50% is defined as λ (T _50% ), and the CuO-containing glass wafer at that time When the thickness is D ₀ , the absorption coefficient α at λ (T _50% ) of the CuO-containing glass wafer is used.
0.5 = exp (−α × D ₀ )
Related to. That is, even for the average thickness d ₀ of the glass wafer described above,
0.5 = exp (−α × d ₀ )
The wavelength λ (T _50% ) associated with is identified. Therefore, the transmittance T of each position of the CuO-containing glass wafer having a difference value Δd of _{d 0 (= d 0 × x} ) is represented by the following equation.
T = exp {−α × (d ₀ + Δd)}
= Exp {−α × d ₀ × (1 + x)} = 0.5 ^{(1 + x)}
Here, x represents the ratio (x = Δd / d ₀ ) of the difference value Δd with respect to the thickness d ₀ where λ (T _50% ) of the CuO-containing glass wafer is a set value.

FIG. 5 is a graph showing the calculation result of the transmittance difference ΔT at λ (T _50% ) when x varies in the range of ± 20% for the CuO-containing glass wafer. From FIG. 5, the CuO-containing glass wafer has an in-plane distribution of thickness difference Δd / d ₀ of −5.5% to −4.9% and a thickness distribution Δd / d ₀ of −15% to + 15%. When d ₀ is -5% to + 5%, an in-plane distribution of transmittance difference of + 1.8% to -1.7% occurs.

Moreover, the wavelength dependency (T (λ)) of the transmittance in the vicinity of λ (T _50% ) occurs in the CuO-containing glass wafer according to the near infrared absorption characteristics. That is, when a slope r (% / nm) indicating the ratio ΔT / Δλ (= differential coefficient) of the transmittance change ΔT with respect to the wavelength change Δλ in the vicinity of λ (T _50% ) is obtained, based on the value of r, It can be converted into a distribution Δλ (= ΔT / r) of λ (T _50% ) associated with the thickness distribution Δd / d ₀ .

FIG. 6 shows λ (T _50% ) with the thickness distribution Δd / d ₀ when the slope r (% / nm) is 0.4, 0.5, 0.6, 1.0, and 1.5. This is a calculation result of the distribution Δλ. In the case of a CuO-containing glass wafer, the transmittance near the absorption maximum wavelength and the slope r (% / nm) near λ (T _50% ) can be adjusted by the CuO content and thickness.

  For example, when adjusting so as to maintain high transmittance at a wavelength of 450 to 600 nm, r = 0.4 to 0.6 (% / nm). Then, the transmittance of the CuO-containing glass wafer can be linearly approximated from 80% to 40%, and the wavelength change Δλ when ΔT = 40% changes is 100 nm at r = 0.4 (% / nm), r = 0.5 (% / nm) corresponds to 80 nm, and r = 0.6 (% / nm) corresponds to 67 nm.

In addition, the larger the in-plane Δλ of the CuO-containing glass wafer, the lower the color reproducibility of the captured image and the possibility of causing color unevenness. Therefore, Δλ may be within ± 5 nm, preferably within ± 3 nm, more preferably within ± 2 nm, and even more preferably within ± 1 nm. Further, the CuO-containing glass wafer may be adjusted to have a concentration distribution of CuO content so that the in-plane Δλ is within a predetermined range, and the thickness distribution Δd / d ₀ may be within ± 15%, and ± 8 % Is preferable, within ± 5% is more preferable, and within ± 2% is more preferable.

From the calculation result of FIG. 6, the thickness distribution Δd / d _{0 in} which the wavelength change Δλ in the vicinity of λ (T _50% ) of the CuO-containing glass wafer is within ± 3 nm is r = 0.4 where the spectral transmittance change is moderate. Although it is within ± 3% at (% / nm), it is within ± 5% at a slightly steep r = 0.6 (% / nm). That is, when the average thickness d ₀ of the CuO-containing glass wafer is d ₀ = 0.4 mm, the thickness difference value Δd is allowed to be within a range of ± 12 μm to ± 20 μm, but when d ₀ = 0.1 mm, the thickness difference The difference value Δd needs to be in the range of ± 3 μm to ± 5 μm and highly accurate thickness control. Moreover, it is preferable that the specification is satisfied even in a glass wafer that does not contain CuO and is transparent in near infrared light.

  In addition, the near-infrared absorption glass wafer 10 (20) of the present invention having the absorption layer 11 on at least one surface of the glass wafer 12 (13) satisfying the above specifications also suppresses deterioration in imaging quality of the solid-state imaging device. An optical mirror surface is preferably obtained. Specifically, it can be evaluated by referring to the MIL Military Standard MIL-PRF-13830 scratch (scratch) and dig (butsu) reference examples, assuming a visual inspection that defines the surface quality of optical components. This is similarly required in the surface quality of the glass wafer before the absorption layer is formed.

  Although the glass wafer depends on the pixel size, it is sufficient that the glass wafer generally satisfies the following level in the optically effective surface of the solid-state imaging device. That is, the surface quality may be a precision grade 60-40 (scratch width 6 μm or less, dig diameter 40 μm or less) in the above standard, and a high precision grade 20-10 (scratch width 2 μm or less, dig diameter 10 μm or less). ) Is preferred. The flatness may be a precision grade flatness λ / 4, and a high precision grade flatness λ / 20 is preferable. Furthermore, the surface roughness of the air interface of the glass wafer may be a precision grade of 20 mm RMS, preferably a high precision grade of 5 mm RMS. In addition, when the absorption layer 11 is arrange | positioned at the joint surface side with Si wafer, if the refractive index difference of the adhesive agent and absorption layer to be used is 0.1 or less, the refractive index difference of an air interface and a glass wafer is about. Since the intensity of reflection and scattering is 1/25 or less compared to 0.5, the specification of the surface roughness of the absorbing layer surface can be relaxed.

  Further, the near-infrared absorbing glass wafer may further include a reflective layer made of a dielectric multilayer film. Near-ultraviolet light that cannot be sufficiently blocked by a glass wafer or absorbing layer made of near-infrared volume absorbing glass. Light and near-infrared light can be blocked by reflection. FIG. 1C is a cross-sectional view of a near-infrared absorbing glass wafer 30 provided with a reflective layer. The reflective layers 14a and 14b are provided on one side and / or both sides of the glass wafer 12 (13). A reflective layer 14c may be provided on the surface. The near-infrared absorbing glass wafer 30 may be provided with an antireflection film, and may be subjected to a surface treatment with a silane coupling agent for improving the adhesion and reliability of the absorbing layer 11, or a dielectric film. May be provided. Since one of 14a and 14c located on the surface of the near-infrared absorbing glass wafer 30 is bonded to the Si wafer by an adhesive, it may be designed in consideration of the refractive index of the adhesive.

  The near-infrared absorbing glass wafer 30 of the present invention is bonded to a Si wafer on which a solid-state imaging device is formed, and is disposed at a position close to the pixels. For this reason, if there are foreign matters or microdefects in the reflective layers 14a, 14b, and 14c, they can directly become pixel defects. Therefore, the allowable level of the size and the number of occurrences is stricter than the reflective layer in the non-junction type optical filter. There are many cases. Therefore, the near-infrared absorbing glass wafer 30 may include the reflective layers 14a, 14b, and 14c according to the quality level.

Next, the absorption layer 11 will be described below.
(Absorption layer)
The absorbing layer 11 is a layer containing an absorbing dye, particularly a near-infrared absorbing dye (A) (hereinafter also referred to as “dye (A)”) and a transparent resin (B), and is typically a transparent resin. This is a layer in which the dye (A) is uniformly dissolved or dispersed in (B). The absorption layer 11 may further contain a near-ultraviolet absorbing dye (U) (hereinafter also referred to as “dye (U)”).

  In addition, in the near-infrared absorption type | mold glass wafer of FIG. 1A-FIG. 1C, even when the absorption layer 11 contains a pigment | dye (U), although it illustrates in figure to be comprised by one layer, it is not restricted to this structure. . For example, when the absorption layer 11 contains the dye (A) and the transparent resin (B) and does not contain the dye (U), a configuration may be provided in which a near-ultraviolet absorption layer (not shown in FIGS. 1A to 1C) is separately provided. That is, the near-ultraviolet absorbing layer may contain a pigment (U) and a transparent resin and be provided as an independent layer.

  In this case, the near-ultraviolet absorption layer may be provided on the absorption layer 11 side of both the main surfaces of the glass wafer 12 (13), or may be provided on the side opposite to the absorption layer 11 side. There is no limit. However, even if it is the structure which provides a near-ultraviolet absorption layer separately, the optical characteristic same as the optical characteristic of the structure in which the near-infrared absorption type glass wafer of this invention contains the pigment | dye (U) further is obtained. Moreover, even when the absorption layer 11 contains the pigment (A), the transparent resin (B), and further the pigment (U), a near ultraviolet absorption layer containing the pigment (U) and the transparent resin (B) is separately provided. Also good. Hereinafter, the near-infrared absorption type glass wafer of this invention demonstrates as a structure by which the absorption layer 11 contains a pigment | dye (U), when a pigment | dye (U) is contained.

<Near-infrared absorbing dye (A)>
The near-infrared absorbing dye (A) is not particularly limited as long as it has the ability to transmit light in the visible region (wavelength 450 to 600 nm) and absorb light in the near-infrared region (wavelength 700 to 1150 nm). The dye in the present invention may be a pigment, that is, a state in which molecules are aggregated.

The dye (A) is preferably a material having an absorption maximum wavelength λ _max at a wavelength of 650 to 750 nm, and more preferably a material having λ _max at a wavelength of 680 to 720 nm. Moreover, the absorption layer 11 containing a pigment | dye (A) has a high freedom degree in selection of the kind and content of a material which can narrow an absorption wavelength band width compared with the CuO containing glass wafer which has absorption in a near infrared region. Therefore, in the near infrared absorption type glass wafer 20 using the CuO-containing glass wafer 13, the absorption layer 11 has a transmittance T (λ _max ) at the absorption maximum wavelength λ _max of the absorption maximum wavelength λ _Gmax of the CuO-containing glass wafer. By adjusting the transmittance to be lower than the transmittance T (λ _Gmax ) in the glass, a sharp light shielding is performed near the λ _max to the visible light side near the λ _max while suppressing a decrease in the transmittance due to the absorption remaining in the visible region of the CuO-containing glass. Can be realized.

Further, the reason why the spectral transmittance curve of the absorption layer 11 is “the absorption of the visible light is small and it is preferable that the absorption layer 11 has a steeper slope toward the visible light (short wavelength) side than λ _max ” is due to the visibility of the absorption layer 11 This is for realizing a spectral transmittance curve close to. That is, the absorption layer 11 maintains a high transmittance with respect to light with a wavelength of 550 to 600 nm having a high visibility, and a transmittance with respect to light with a wavelength of 600 to 650 nm at which the visibility is gradually lowered is approximately 40 to 60%. The transmittance with respect to light having a wavelength of 650 to 700 nm from a level where the visibility is low to a level which hardly occurs is reduced to 5% or less. Specifically, the dye (A) and the content thereof may be adjusted so that the transmittance T (λ _max ) at λ _max of the absorption layer 11 is 5% or less.

As a result of adjusting the dye (A) so that the transmittance with respect to light having a wavelength of 550 to 700 nm exhibits the above optical characteristics, the absorption layer 11 is preferably as wide as possible in the near infrared region of approximately 700 nm or more. Further, the absorption layer 11, the absorption wavelength band width transmittance is 20% or less at lambda _max vicinity may be any 30nm or more, more preferably if more than 40 nm. In addition, even if the absorption layer 11 has such an absorption wavelength band width, it is effective to use a reflection layer for near-infrared transmitted light that cannot be sufficiently blocked by absorption of the CuO-containing glass wafer and the absorption layer 11. Can be shaded. In other words, the reflection layer shifts the reflection band due to the incident angle dependence of the spectral transmittance curve for light having an incident angle of 0 ° to 30 °, but the wavelength with a transmittance of 50% located on the short wavelength side of the reflection band. Even if it shifts to the short wavelength side depending on the incident angle, it may be set so as to be within the change in the absorption wavelength band of the absorption layer. In this way, the near-infrared absorbing glass wafer based on the design can suppress the incident angle dependency of the reflective layer particularly in the near-infrared region.

The dye (A) has an absorption maximum wavelength at a wavelength of 650 to 750 nm in an absorption spectrum of a wavelength of 400 to 850 nm measured using a resin layer obtained by dispersing the dye (A) in the transparent resin (B). It is good to have. The dye (A) having the absorption characteristic is referred to as the dye (A1), and the absorption maximum wavelength in the absorption spectrum is referred to as λ _{max of} the dye (A1). The absorption spectrum of the dye (A1) has an absorption peak having an absorption peak at the wavelength λ _max (hereinafter referred to as “λ _max absorption peak”). The absorption spectrum of the dye (A1) also has little absorption of visible light, and it is preferable that the absorption peak of λ _max has a steep inclination on the visible light side. Further, the slope should be gentle on the long wavelength side of the absorption peak at λ _max .

  Examples of the dye (A1) include cyanine compounds, phthalocyanine compounds, naphthalocyanine compounds, dithiol metal complex compounds, diimonium compounds, polymethine compounds, phthalide compounds, naphthoquinone compounds, anthraquinone compounds, indophenol compounds, Examples include squarylium compounds.

Among these, squarylium compounds, cyanine compounds and phthalocyanine compounds are more preferred, and squarylium compounds are particularly preferred. The dye (A1) composed of a squarylium compound has little absorption of visible light in the above absorption spectrum, the absorption peak of λ _max has a steep slope on the visible light side, and has high storage stability and light stability. Therefore, it is preferable.

  For the squarylium-based compound, for example, WO2012 / 169447 can be referred to, and this document shows a dye having an absorption maximum wavelength of 695 to 747 nm. Moreover, the absorption layer which contains a squarylium type compound in various transparent resins can refer to, for example, WO2014 / 088063, and examples of the absorption maximum wavelength 691 to 722 nm are shown in this document.

The dye (A1) composed of a cyanine compound is preferable because the absorption spectrum has little visible light absorption and high light absorption on the long wavelength side in the wavelength region near _λmax . In addition, the cyanine compound is low in cost, and long-term stability can be secured by forming a salt. A dye (A1) made of a phthalocyanine compound is preferable because of excellent heat resistance and weather resistance. WO 2014/030628 can also be referred to, and this document describes a specific example of an absorption layer containing, in a transparent resin, a cyanine compound having absorption maximum wavelengths of 694, 740, and 747 nm and a phthalocyanine compound having absorption maximum wavelength of 681 nm. An example is shown.

<Near UV absorbing dye (U)>
The near-ultraviolet absorbing dye (U) absorbs light having a wavelength of 430 nm or less. As the dye (U), a compound satisfying the following requirements (iv-1) and (iv-2) (hereinafter referred to as dye (U1)) is preferable.

(Iv-1) In a light absorption spectrum having a wavelength of 350 to 800 nm measured by dissolving in dichloromethane, the light absorption spectrum has at least one absorption maximum wavelength in the region of wavelength 415 nm or less, and the absorption maximum in the region of wavelength 415 nm or less The absorption maximum wavelength λ _max (UV) on the longest wavelength side is at a wavelength of 360 to 415 nm.
In (iv-2) spectral transmittance curve measured was dissolved in dichloromethane, when the transmittance at the maximum absorption wavelength lambda _{max (UV)} is 10% transmissive at wavelengths longer than the absorption maximum wavelength lambda _{max (UV)} The difference λ _L90 -λ _L50 between the wavelength λ _{L90 at} which the rate is 90% and the wavelength λ _L50 at which the transmittance is 50% longer than the absorption maximum wavelength λ _max (UV) is 13 nm or less.
In addition, the absorption maximum wavelength of the pigment | dye (U1) which satisfy | fills the requirements of (iv-1) and (iv-2) does not change a lot in transparent resin.

  Specific examples of the dye (U1) include oxazole, merocyanine, cyanine, naphthalimide, oxadiazole, oxazine, oxazolidine, naphthalic acid, styryl, anthracene, cyclic carbonyl, and triazole. Etc.

  Commercially available products include, for example, Uvitex (trademark) OB (trade name, manufactured by Ciba), Hakcol (trademark) RF-K (trade name, manufactured by Showa Chemical Industry Co., Ltd.), Nikkafluor EFS, and Nikkafluor SB-conc. (All are trade names manufactured by Nippon Chemical Industry Co., Ltd.). Examples of the merocyanine series include S0511 (trade name, manufactured by Few Chemicals). Examples of cyanine include SMP370 and SMP416 (all are trade names manufactured by Hayashibara Co., Ltd.). Examples of naphthalimide include Lumogen (trademark) F violet 570 (trade name, manufactured by BASF).

As mentioned above, the absorption layer 11 contains the pigment | dye (A) and transparent resin (B), and it is good to contain a pigment | dye (U) further. It is preferable that the absorption layer 11 has the following optical properties (a1) and (a2) by containing the dye (A).
(A1) The absorption spectrum has an absorption maximum wavelength (λ _max ) at a wavelength of 650 to 750 nm.
(A2) The transmittance of light having a wavelength of 450 to 550 nm is 80% or more.

Further, the absorption layer 11 on the glass wafer 12 (13) is a near infrared absorption type glass wafer 30 (optical filter) obtained in combination with the reflection layer, and the following optical components (i-1) and (i-2) are used. It may be configured to have characteristics.
(I-1) The average transmittance of light having an incident angle of 0 ° at a wavelength of 450 to 550 nm is 80% or more.
(I-2) The average transmittance of light having an incident angle of 0 ° at a wavelength of 650 to 720 nm is 15% or less.
Here, since the absorption layer 11 satisfies the above (a1) and (a2), the optical characteristics of (i-1) and (i-2) can be easily obtained as a near-infrared absorbing glass wafer. ,preferable.

  In addition, content of the pigment | dye (A) in the absorption layer 11 is taken as the quantity which the absorption layer 11 satisfies the said optical characteristics (a1) and (a2). Further, the content of the dye (A) in the absorption layer 11 is 50% of the light transmittance in a region longer than the wavelength 600 nm of the spectral transmittance curve at an incident angle of 0 ° of the optical filter, preferably in the wavelength of 620 to 660 nm. It is preferable to adjust so as to have a wavelength. Specifically, the pigment (A) is preferably 0.1 to 30 parts by mass, more preferably 0.5 to 25 parts by mass with respect to 100 parts by mass of the transparent resin (B) in the absorption layer 11. 1-20 mass parts is especially preferable.

Moreover, when the absorption layer 11 further contains a pigment (U), the following (ii) is used as a near-infrared absorption type glass wafer obtained by combining the absorption layer 11, the glass wafer 12 (13), and the reflection layer. It may be configured to have the optical characteristics of -1) and (ii-2).
(Ii-1) At a wavelength of 430 to 450 nm, the average value of the transmittance of light having an incident angle of 0 ° is 70% or more.
(Ii-2) At a wavelength of 350 to 390 nm, the average transmittance of light with an incident angle of 0 ° is 5% or less.

  Furthermore, the content of the dye (U) in the absorption layer 11 is a region shorter than a wavelength of 450 nm in the spectral transmittance curve at an incident angle of 0 ° of the near-infrared absorbing glass wafer of the present invention, preferably a wavelength of 400 to 425 nm. It is preferable to determine the wavelength so that the transmittance is 50%. In addition, it is preferable that 0.01-30 mass parts of pigment | dye (U) is contained with respect to 100 mass parts of transparent resin (B) in the absorption layer 11, and 0.05-25 mass parts is more preferable. 0.1 to 20 parts by mass is even more preferable.

  The absorbing layer 11 is not limited to the pigment (A), the transparent resin (B), and the optional component pigment (U), and does not inhibit the effects of the present invention. Agents, leveling agents, antistatic agents, heat stabilizers, light stabilizers, antioxidants, dispersants, flame retardants, lubricants, plasticizers, and the like. Moreover, the component added to the coating liquid used when forming the absorption layer 11 mentioned later, for example, the component derived from a silane coupling agent, a heat | fever or photoinitiator, a polymerization catalyst, etc. are mentioned. The content of these other optional components in the absorbent layer is preferably 15 parts by mass or less for 100 parts by mass of the transparent resin (B).

  As for the film thickness of the absorption layer 11, 0.1-10 micrometers is preferable. If the film thickness is less than 0.1 μm, the near-infrared absorbing ability may not be sufficiently exhibited. On the other hand, if the film thickness exceeds 10 μm, the flatness of the film is lowered, and there is a possibility that the absorption rate varies. The film thickness is more preferably 1 to 10 μm. If it exists in this range, sufficient near-infrared absorptivity and flatness of a film thickness can be compatible. Even when a near ultraviolet absorbing layer is separately provided, the film thickness of the near ultraviolet absorbing layer only needs to satisfy the above range.

  As the near-infrared absorber, inorganic fine particles can be preferably used as those that do not impair the effects of the optical properties of the dye (A), preferably the dye (A1). Specific examples include fine particles such as ITO (Indium Tin Oxide), ATO (Antimony-doped Tin Oxide), cesium tungstate, and lanthanum boride. Among these, ITO fine particles and cesium tungstate fine particles have a high visible light transmittance and a wide range of light absorption including an infrared region exceeding 1200 nm. Therefore, particularly when infrared light shielding is required. preferable.

  The absorption layer 11 is, for example, a coating liquid prepared by dispersing and dissolving a dye (A) and a transparent resin (B) or a raw material component of the transparent resin (B), and optionally a dye (U) in a solvent. It can be manufactured by coating on a glass wafer 12 (13), drying, and further curing as necessary. By forming the absorption layer 11 by such a method, it is possible to produce the absorption layer 11 uniformly with a desired film thickness. When the absorption layer 11 contains the said arbitrary component, a coating liquid contains this arbitrary component.

  As the solvent, as long as it is a dispersion medium that can stably disperse the dye (A), the raw material component of the transparent resin (B) or the transparent resin (B), and further optionally the dye (U) or a solvent that can be dissolved, There is no particular limitation. 10-5000 mass parts is preferable with respect to 100 mass parts of transparent resin (B), and, as for the quantity of a solvent, 30-2000 mass parts is especially preferable. In addition, 2-50 mass% is preferable with respect to the coating liquid whole quantity, and, as for content of the non-volatile component (solid content) in a coating liquid, 5-40 mass% is especially preferable.

  For the preparation of the coating liquid, a stirring device such as a magnetic stirrer, a rotation / revolution mixer, a bead mill, a planetary mill, or an ultrasonic homogenizer can be used. For coating of coating liquid, dip coating method, cast coating method, spray coating method, spinner coating method, bead coating method, wire bar coating method, blade coating method, roller coating method, slit die coater method, etc. The law can be used. In addition, a bar coater method, a screen printing method, a flexographic printing method, etc. can also be used.

  After the coating liquid is applied onto the glass wafer 12 (13), the absorbent layer 11 is formed on the glass wafer 12 (13) by drying. When the coating solution contains the raw material component of the transparent resin (B), a curing treatment is further performed. When the reaction is thermosetting, drying and curing can be performed simultaneously. However, in the case of photocuring, a curing process is provided separately from the drying.

  In the near-infrared absorbing glass wafer of the present invention, the spectral transmittance curve is determined by the near-infrared absorption characteristics of the absorption layer 11 formed on the glass wafer 12 (13). Since the near-infrared absorbing glass wafer is bonded to the Si wafer on which a plurality of solid-state imaging elements are formed, the absorption layer 11 is within the effective surface of the near-infrared absorbing glass wafer facing the solid-state imaging element. It is important that the in-plane distribution of the spectral transmittance curve has uniformity.

Specifically, the absorption layer 11 is a coating solution prepared by dispersing and dissolving the pigment (A), the transparent resin (B) or the raw material component of the transparent resin (B), and optionally the pigment (U). Is coated on the glass wafer 12 (13), dried, and further cured as necessary. It is important that the absorption layer 11 formed in this way maintain uniformity so as to be within a predetermined Δt (= t ₀ × x) that is a difference value between a predetermined film thickness value t ₀ and t ₀ within the glass wafer surface. It becomes.

Here, with respect to the absorption layer 11, assuming that the film thickness at the wavelength λ (T _50% ) at which the transmittance with respect to light having a wavelength of 620 to 690 nm is 50% is T ₀ , the absorption layer 11 at λ (T _50% ) Using the absorption coefficient α,
0.5 = exp (−α × T ₀ )
Related to. That is, even for the average thickness t ₀ of the absorption layer 11,
0.5 = exp (−α × t ₀ )
The wavelength λ (T _50% ) associated with is identified. Therefore, the transmittance T at each position of the absorption layer 11 having the difference value Δt (= t ₀ × x) of t _{0 in} the plane is expressed by the following equation.
T = exp {−α × (t ₀ + Δt)}
= Exp {−α × t ₀ × (1 + x)} = 0.5 ^{(1 + x)}
Here, x represents a ratio (film thickness distribution) x = Δt / t ₀ of the difference value Δt with respect to the thickness t ₀ at which λ (T _50% ) of the absorption layer 11 is a set value.

FIG. 5 is a graph that also shows the calculation result of the transmittance difference ΔT at λ (T _50% ) when x varies within a range of ± 20% for the absorption layer. Although the figure shows the relationship between the thickness distribution Δd / d ₀ and the transmittance difference ΔT, as described above, the same relationship with ΔT holds for the film thickness distribution Δt / t _{0 of the} absorption layer 11. From FIG. 5, the film thickness distribution Δt / t ₀ is −15% to + 15%, the in-plane distribution of the transmittance difference of + 5.5% to −4.9, and the film thickness distribution Δt / t ₀ is −5% to At + 5%, an in-plane distribution of + 1.8% to -1.7% transmittance difference occurs.

Further, the absorption layer 11 has wavelength dependency (T (λ)) of transmittance in the vicinity of λ (T _50% ) according to the near-infrared absorption characteristics. That is, when a slope r (% / nm) indicating the ratio ΔT / Δλ (= differential coefficient) of the transmittance change ΔT with respect to the wavelength change Δλ in the vicinity of λ (T _50% ) is obtained, based on the value of r, It can be converted into a distribution Δλ (= ΔT / r) of λ (T _50% ) accompanying the film thickness distribution Δt / t ₀ .

FIG. 6 shows λ (T _50% ) with the film thickness distribution Δt / t ₀ when the gradient r (% / nm) is 0.4, 0.5, 0.6, 1.0, and 1.5. The calculation result of the distribution Δλ is also shown. For example, the transmittance of the absorption layer 11 can be linearly approximated from 90% to 10%, and the wavelength change Δλ when ΔT = 80% changes is 160 nm when r = 0.5 (% / nm), and r = 1. 0.0 (% / nm) corresponds to 80 nm, and r = 1.5 (% / nm) corresponds to 40 nm.

From FIG. 6, in order to suppress Δλ of the absorption layer 11 within ± 3 nm for the purpose of suppressing the decrease in color reproducibility and color unevenness of the captured image, r = ΔT / Δλ is 1.5 (% /%). nm), the film thickness distribution Δt / t ₀ may be within ± 13%. On the other hand, when r is 0.5 (% / nm), the film thickness distribution Δt / t ₀ may be within ± 5%.

Further, the near-infrared absorbing glass wafer may have a wavelength distribution Δλ within ± 5 nm, preferably within ± 3 nm, more preferably within ± 2 nm, and further preferably within ± 1 nm. Therefore, the absorption layer 11 adjusts the dye (A) concentration distribution in the absorption layer 11 so that the in-plane wavelength distribution Δλ is within this range, and the film thickness distribution Δt / t ₀ is within ± 15%. It is preferable if it is within ± 8%, more preferably within ± 5%, and even more preferably within ± 2%.

  Therefore, for example, in order to obtain the absorption layer 11 having a small film thickness distribution on the surface of a glass wafer having a diameter of 15 cm or more, the pigment (A) and the raw material components of the transparent resin (B) or the transparent resin (B), and further optional It is assumed that the coating liquid prepared by dispersing and dissolving the dye (U) to be added to the solvent is applied to a uniform film thickness. In order to realize this, in the above-described coating liquid coating method, a coating method capable of ensuring the coating liquid film thickness distribution may be used.

  Specifically, when using a spin coat method used for coating an organic dye recording layer or a protective film in some optical disks such as a photoresist film coating on a semiconductor wafer, DVD, Blu-ray (registered trademark), The film thickness distribution can be controlled within ± 1% with respect to a resin layer having a thickness of several μm spin-coated on a wafer size of 8 to 12 inches in diameter. Moreover, you may use the slit die coater method at the time of forming the RGB color filter with a film thickness of several micrometers on the large sized glass substrate of a liquid crystal television. For example, when the slit die coater method is used, a color resist having a film thickness of 2 μm is applied to a glass substrate surface of 800 mm □ or more within a film thickness distribution of ± 3%, and a film having a thickness of about 1% excluding the peripheral portion of the glass substrate. A thickness distribution is obtained.

Therefore, the coating thickness distribution Δt / t ₀ corresponding to the thickness distribution of the absorbing layer 11 is within ± 15%, preferably within ± 8%, more preferably within ± 5%, and even more preferably ± A coating method that is within 2% may be used. That is, when the (set value) film thickness t _{0 of} the absorption layer 11 is 1 to 10 μm, the film thickness distribution Δt of the absorption layer 11 in the glass wafer is within ± 20 nm in order to make Δt / t ₀ within ± 2%. It is good to control within ˜ ± 200 nm.

<Transparent resin (B)>
The transparent resin (B) is specifically acrylic resin, epoxy resin, ene thiol resin, polycarbonate resin, polyether resin, polyarylate resin, polysulfone resin, polyethersulfone resin, polyparaphenylene resin, polyarylene ether. Examples include phosphine oxide resins, polyimide resins, polyamideimide resins, polyolefin resins, cyclic olefin resins, and polyester resins. As transparent resin (B), 1 type may be used individually from these resin, and 2 or more types may be mixed and used for it.

  Among these, from the viewpoint of solubility of the dye (A) and the dye (U) in the transparent resin (B), the transparent resin is an acrylic resin, a polyester resin, a polycarbonate resin, an ene / thiol resin, an epoxy resin, and a cyclic olefin. One or more selected from resins are preferred. Further, the transparent resin is more preferably at least one selected from acrylic resin, polyester resin, polycarbonate resin, polyimide resin, polyamideimide resin, and cyclic olefin resin. As the polyester resin, polyethylene terephthalate resin, polyethylene naphthalate resin and the like are preferable.

(Semiconductor wafer laminate)
Furthermore, the present invention provides a semiconductor wafer laminate in which a semiconductor wafer on which a plurality of solid-state imaging elements for manufacturing a camera module are formed and a near-infrared absorbing glass wafer are joined. FIG. 7 schematically shows an example of a semiconductor wafer laminate 40 (50) in which a near-infrared absorbing glass wafer 10 (20, 30) of the present invention and a semiconductor wafer 15 on which a plurality of solid-state imaging devices 19 are formed are joined. It is a perspective view shown in FIG.

  8A and 8B are schematic cross-sectional views in which the periphery of the solid-state imaging device 19 of the semiconductor wafer laminate 40 (50) in which the near-infrared absorbing glass wafer 10 (20, 30) of the present invention is integrated with the solid-state imaging device 19 is enlarged. FIG. The solid-state imaging device 19 is formed by forming an Si semiconductor (CMOS, CCD) photodetector array 16 on one side of an Si wafer 15 and an RGB mosaic color filter 17 and a resin microlens 18 for each pixel. In the solid-state imaging device 19, the Si wafer 15 and the near-infrared absorbing glass wafer 10 (20, 30) are integrated via an adhesive 21 to form a semiconductor wafer laminated body 40 (50).

  The semiconductor wafer laminated body 40 of FIG. 8A is a structure integrated with the solid-state imaging device 19 via the adhesive 21 on the absorption layer 11 side of the near-infrared absorbing glass wafer 10 (20, 30). On the other hand, the semiconductor wafer laminated body 50 of FIG. 8B has a configuration in which the absorption layer 11 faces the air side and is integrated on the side facing the absorption layer 11 via the adhesive 21. The adhesive 21 may be any material that is transparent to visible light. In the semiconductor wafer laminate, the arrangement of the absorption layer 11 may be on the solid-state imaging device 19 side (FIG. 8A) or on the air side (FIG. 8B). Since the absorption layer is softer than a glass wafer, the surface is easily scratched. Therefore, when the absorption layer 11 is formed of a single layer, if it is disposed on the bonding surface side of the solid-state imaging device 19, the subsequent manufacturing process Scratches are unlikely to occur.

  The semiconductor wafer laminated body 50 is a structure obtained also in the process of forming the absorption layer 11 on the surface of the glass wafer 12 (13) after joining the glass wafer 12 (13) to the Si wafer 15. That is, the semiconductor wafer laminated body 50 can obtain the same configuration even if the order of the formation of the absorption layer 11 and the bonding of the glass wafer and the solid-state imaging device is not limited.

In the solid-state imaging device 19, the resin microlens 18 has a convex lens function that collects incident light on the light receiving surface of the photodetector array 16. Therefore, the refractive index n _{ML of the} transparent resin used for the resin microlens 18 and the refractive index n _G of the adhesive 21 satisfy n _ML > n _G , and the refractive index difference (n _ML −n _G ) is preferably as large as possible. Specifically, n _ML ≧ 1.8 is preferable, and n _ML ≧ 1.9 is more preferable. Further, n _G ≦ 1.5 is preferable, and n _G ≦ 1.45 is more preferable.

  The adhesive 21 may be either an ultraviolet curable type or a heat curable type, but an ultraviolet curable type is preferable in that an adhesive strength can be obtained in a short time. The ultraviolet curable adhesive can provide sufficient adhesive strength between the resin microlens 18 surface and the glass surface or absorption layer surface of the near-infrared absorbing glass wafer 10 (20, 30). Adhesive 21 has a polymerization shrinkage rate of 3% or less upon curing, is less susceptible to misalignment and lowering of adhesive strength due to ambient environmental conditions such as high temperature and high quality, and rapid temperature changes, and has a low halogen content. Those with less outgassing due to later unreacted components are preferred.

Bonding with the adhesive 21 is performed by applying an uncured adhesive to the near-infrared absorbing glass wafer 10 (20, 30), and uniformly between the near-infrared absorbing glass wafer 10 and the solid-state imaging device 19 with a thickness of 10 μm or less. It integrates so that it may become a film thickness, and the semiconductor wafer laminated body 40 (50) is obtained. In the case of using an ultraviolet curable adhesive, it is preferable to irradiate the adhesive 21 with ultraviolet rays from the near-infrared absorbing glass wafer 10 (20, 30) side for polymerization and curing. When using a thermosetting adhesive, the entire semiconductor wafer laminate 50 may be heated and polymerized to cure.
In the case where the absorption layer 11 does not transmit ultraviolet rays or is altered by heat treatment in the curing process of the adhesive 21, the glass wafer 12 (13) is bonded after the glass wafer 12 (13) and the solid-state imaging device 19 are bonded. The absorption layer 11 may be formed on the surface of

  Further, in the semiconductor wafer laminated body of FIG. 8A and FIG. 8B, electric wiring for voltage application and electric signal extraction of the solid-state imaging device 19 is omitted. Actually, in the case of a back-illuminated CMOS solid-state imaging device capable of suppressing a reduction in sensitivity due to pixel miniaturization, the electrical wiring is disposed on the side facing the photodetector array 16 of the semiconductor wafer 15, and the through electrode of the semiconductor wafer 15 is disposed. An example in which the electrode is drawn out to the back surface of the solid-state imaging device by a technique such as the above.

  The semiconductor wafer stacked body 40 (50) is cut into the size of the solid-state imaging device 19 using a dicing device or the like and mounted on the solid-state imaging device. FIG. 12 is a cross-sectional view schematically showing a main part of the solid-state imaging device 60, the solid-state imaging device 19 to which the near-infrared absorbing glass wafer 10 (20) is bonded, the reflective layer 14 on the front surface thereof, An imaging lens 31 and a housing 33 for fixing them are included. The imaging lens 31 is fixed by a lens unit 32 provided inside the housing 33. The reflective layer 14 has a dielectric multilayer film on one side or both sides of the transparent substrate, and is disposed in the optical path between the light incident side of the lens unit 32 and the solid-state imaging device 19. The solid-state imaging device 60 in FIG. 12 shows an example in which the reflective layer 14 is disposed between the lens unit 32 and the near-infrared absorbing glass wafer 10 (20). A configuration in which a dielectric multilayer film is formed on the surface of the imaging lens 31 may be employed.

Thus, since the semiconductor wafer laminate 40 (50) of the present invention can incorporate the optical filter function into the solid-state imaging device 19 at the wafer level, the productivity is improved and the characteristics are stabilized. Further, the conventional optical filter function is integrated in the solid-state image sensor 19 and the imaging lens 31, and the assembly adjustment of the camera module is simplified by reducing the number of optical filter components, and the solid-state imaging device can be miniaturized. Since the near infrared absorption glass wafer 10 (20, 30) has a difference value Δd of the glass wafer thickness d ₀ within ± 30 μm, the distance (focus) adjustment between the lens unit 32 and the solid-state imaging device 19 is adjusted. Can be relaxed.

Hereinafter, the present invention will be described in more detail by way of examples.
[Example 1]
A manufacturing example of the near-infrared absorbing glass wafer 10 shown in FIG. 1A will be described. The near-infrared absorbing glass wafer 10 includes an absorbing layer 11 on one side of a glass wafer 12 having a diameter of 15 cm and a thickness of 0.2 mm.

The glass wafer 12 is alkali-free glass (trade name: EN-A1 manufactured by Asahi Glass Co., Ltd.) having an alkali oxide (Li ₂ O, Na ₂ O, K ₂ O, etc.) content of 0.1% (mass% display) or less. ) Is used. The glass wafer 12 is processed by double-side polishing so that the in-plane thickness difference value Δd = 3 μm or less, that is, Δd / d ₀ = ± 1.5% or less. EN-A1 is a glass having a refractive index n _D = 1.52 and transparent to light having a wavelength of 350 to 1150 nm, and its thermal expansion coefficient is 3.3 ppm / K (50 to 200 ° C.).

Next, the dye (U) and the dye (A) were mixed in a 15% by mass cyclohexanone solution of a polyester resin (manufactured by Osaka Gas Chemical Co., Ltd., trade name: B-OKP2, refractive index: 1.64) The mixture is dissolved by stirring to prepare a coating solution. This coating solution is applied to one main surface of the glass wafer 12 by a spin coating method, and the solvent is heated and dried, and then an absorption layer 11 having an average thickness t ₀ = 2.7 μm in a φ15 cm plane is formed. To do.

Here, the dye (A) is a squarylium dye (A1) having an absorption maximum wavelength λ (T _min ) of 705 nm, and is mixed at an addition amount 3 (parts by mass relative to 100 parts by mass of the transparent resin (B)). In addition, the dye (U) is an oxazole-based (U1) Uvitex (trademark) OB having an absorption maximum wavelength λ (T _min ) of 396 nm, and the addition amount is 5 (parts by mass relative to 100 parts by mass of the transparent resin (B)). Mix. The difference value Δt of the thickness of the absorption layer 11 in the glass wafer surface is within ± 40 nm, and Δt / t ₀ is approximately within ± 1.5%.

FIG. 9 is a spectral transmittance curve (incident angle: 0 °) of a reflective layer made of a dielectric multilayer film used in combination with the near-infrared absorbing glass wafer 10. The reflection layer is formed by alternately stacking 40 layers of SiO ₂ films having a refractive index of 1.45 and TiO ₂ films having a refractive index of 2.41 as optical elements disposed in the camera module.

FIG. 10 shows a spectral transmittance curve of an optical part composed of a near-infrared absorbing glass wafer 10 and a reflective layer. From FIG. 10, the average transmittance in the near ultraviolet light having a wavelength of 350 to 400 nm is 0.3%, the average transmittance in the visible light having a wavelength of 430 to 600 nm is 92%, and the average transmittance in the near infrared light having a wavelength of 700 to 1150 nm. Is 0.9%, and the spectral transmittance change approximates the visibility at a wavelength of 600 to 700 nm. The average value in the near-infrared absorbing glass wafer surface of the wavelength λ (T _50% ) at which the transmittance of light with a wavelength of 600 to 700 nm is 50% is 645 nm, and the transmittance change ratio ΔT / Δλ is It shows approximately 1.08, and its in-plane distribution Δλ (T _50% ) is within ± 1 nm.

[Example 2]
Next, instead of EN-A1 used as the glass wafer 12 in Example 1, a CuO-containing glass wafer (trade name, manufactured by Asahi Glass Co., Ltd.) having a diameter of 15 cm, a thickness of 0.2 mm, and an absorption maximum wavelength λ (T _min ) of 850 nm. : NF-50T). The average value of the internal transmittance of the CuO-containing glass wafer in light having a wavelength of 850 nm is 10%, and the average value of the transmittance 50% wavelength λ (T _50% ) is 658 nm. Here, in the CuO-containing glass wafer thickness distribution, when Δd / d ₀ = ± 10% or less, that is, Δd = ± 20 μm or less, Δλ (T _50% ) is within ± 5.5 nm.

Next, in the same manner as in Example 1, the dye (U) and the dye (A) are mixed in the cyclohexanone solution of the polyester resin and sufficiently stirred to dissolve, thereby preparing a coating solution. This coating solution was applied to one main surface of the CuO-containing glass wafer 13 by a die coating method, and after the solvent was heated and dried, the average thickness t ₀ in the 15 cm diameter plane was 2.7 μm, An absorption layer 11 having a difference value Δt = ± 54 nm or less (thickness distribution Δt / t ₀ = ± 2% or less) is formed to obtain a near-infrared absorbing glass wafer 20.

  FIG. 11 is a spectral transmittance curve of an optical part (optical filter) composed of the near-infrared absorbing glass wafer 20 and the reflective layer of Example 1. From FIG. 11, the average transmittance in the near ultraviolet light having a wavelength of 350 to 400 nm is 0.3%, the average transmittance in the visible light having a wavelength of 430 to 600 nm is 90%, and the average transmittance in the near infrared light having a wavelength of 700 to 1150 nm. Is 0.1%, and the spectral transmittance change approximates the visibility at a wavelength of 600 to 700 nm. In Example 2, since the CuO-containing glass wafer 13 is used, the light shielding property in the near infrared region is improved as compared with Example 1. A wavelength of 600 to 700 nm shows a gradual change in transmittance.

Further, the average value in the glass wafer plane of the wavelength λ (T _50% ) at which the transmittance is 50% in the wavelength range of 600 to 700 nm is 621 nm, and the transmittance change ratio ΔT / Δλ is approximately 0.73. The in-plane distribution Δλ (T _50% ) is within ± 3.5 nm. That is, the thickness distribution Δd / d _{0 of} the CuO-containing glass wafer 13 is within ± 10%, and Δλ (T _50% ) is within ± 5.5 nm, but the absorption layer 11 has a thickness distribution Δt / t. By forming the film within ₀ = ± 2%, Δλ (T _50% ) of the entire near-infrared absorbing glass wafer 20 can be reduced.

  The near-infrared absorbing glass wafer and semiconductor wafer laminate of the present invention are useful for imaging devices such as digital still cameras and mobile phone cameras using solid-state imaging devices.

  DESCRIPTION OF SYMBOLS 10,20,30 ... Near-infrared absorption type glass wafer, 11 ... Absorption layer, 12 ... Glass wafer, 13 ... (Near-infrared volume absorption type) glass wafer, 14, 14a, 14b ... Reflection layer, 15 ... Si wafer, 16 DESCRIPTION OF SYMBOLS ... Photodetector array, 17 ... RGB mosaic color filter, 18 ... Resin microlens, 19 ... Solid-state image sensor, 21 ... Adhesive, 31 ... Imaging lens, 32 ... Lens unit, 33 ... Housing, 40, 50 ... Semiconductor Wafer stack, 60... Solid-state imaging device (camera module).

Claims (10)

A glass wafer;
An absorption layer is provided on at least one main surface of the glass wafer,
The absorption layer contains a transparent resin and an absorption pigment,
The absorption layer has an average value of the film thickness t ₀ , and a difference value of the film thickness with respect to the average value t ₀ is Δt.
−15% ≦ Δt / t ₀ ≦ 15%
Satisfying near-infrared absorbing glass wafer.   The near-infrared absorbing glass wafer according to claim 1, wherein the glass wafer is a CuO-containing glass wafer. 3. The near-infrared absorbing glass wafer according to claim 1, wherein the glass wafer has an average value of d ₀ and a difference value of thickness with respect to the average value d ₀ of Δd within ± 30 μm. The glass wafer is
−15% ≦ Δd / d ₀ ≦ 15%
The near-infrared absorption type glass wafer of Claim 3 satisfy | filling.   The near-infrared absorbing glass wafer according to any one of claims 1 to 4, wherein the glass wafer has a diameter of 15 cm or more.   The near-infrared absorbing glass wafer according to claim 1, wherein the absorbing dye includes a near-infrared absorbing dye. The absorption layer has an absorption maximum wavelength λ (T _min ) at a wavelength of 650 to 750 nm,
The transmittance T _min for the light having the absorption maximum wavelength λ (T _min ) is 5% or less,
Furthermore, the near-infrared absorption type | mold glass wafer of any one of Claims 1-6 which has wavelength (lambda) (T50 _% ) from which the transmittance | permeability becomes 50% in wavelength 620-690nm.   The near-infrared absorption type glass wafer of any one of Claims 1-7 provided with the reflection layer which has a dielectric multilayer on the at least one main surface of the said glass wafer.   The near-infrared absorbing glass wafer according to any one of claims 1 to 8, wherein the absorbing dye includes a near-ultraviolet absorbing dye.   The semiconductor wafer laminated body which has a solid-state image sensor provided on Si wafer, and the near-infrared absorption type glass wafer in any one of Claims 1-9.

Images