JP2011199027A - Solid state image pickup element and method for manufacturing the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an occurrence of a smear due to an unnecessary charge in an element isolation part in a solid state image pickup element including the element isolation part provided between a photoelectric conversion part and a charge transfer part.SOLUTION: A CCD image sensor that is a solid state image pickup element is equipped with: n-type layers 20 constituting each PD as a photoelectric conversion part; and n-type layers 21 constituting each VCCD as a charge transfer part. A p+ layer 22 constituting an element isolation part is formed between each of the n-type layers 20 and the n-type layers 21. The p+ layer 22 has a concentration gradient in which impurity concentration on a VCCD side is higher than that on a PD side. Thereby, a charge generated in the element isolation part flows into the PD side according to the concentration gradient. Accordingly, an occurrence of a smear due to the flowing of the charge generated in the element isolation part into the VCCD side is prevented.

Description

  The present invention relates to a solid-state imaging device in which an element separation unit is provided between a photoelectric conversion unit and a charge transfer unit, and a manufacturing method thereof.

  A CCD image sensor is known as one of solid-state imaging devices. The CCD image sensor has a so-called smear problem that when a bright subject is imaged, white lines appear in the vertical direction. Smear significantly impairs the appearance of the acquired image. For this reason, various measures have been conventionally taken in CCD image sensors to prevent the occurrence of smear.

  For example, Patent Document 1 describes that the potential barrier of the element isolation portion provided between the photodiode and the vertical transfer path is made higher than the element isolation portion provided between adjacent photodiodes. ing. By so doing, it is possible to prevent the occurrence of smear caused by unnecessary charges generated in the element separation portion between the photodiodes flowing into the vertical transfer path.

  In Patent Document 2, the horizontal concentration distribution of the element isolation portion provided between the photodiodes adjacent in the vertical direction is set so that the portion corresponding to both ends of the photodiode is higher than the portion corresponding to the photodiode center. The concentration is described. In this way, it is possible to prevent smear caused by the signal charge generated in the photodiode flowing into the vertical transfer path via the element isolation portion.

Japanese Patent Laid-Open No. 10-242447 JP 2006-73681 A

  As described in Patent Document 1, an element isolation portion which is a potential barrier for preventing charge movement is provided between a photodiode and a vertical transfer path for an adjacent photodiode. The vertical transfer path and the element isolation portion are covered with a light shielding film to prevent generation of unnecessary charges due to incidence of light. However, light incident obliquely on the photodiode may enter the element isolation portion through the opening end of the light shielding film. Then, when the incident light generates unnecessary charges in the element separating portion and the charges flow into the vertical transfer path, smear occurs.

  In recent years, with the miniaturization of a CCD image sensor, the width of the element isolation portion has been narrowed. When the width of the element isolation portion is narrowed, unnecessary charges generated in the element isolation portion easily flow into the vertical transfer path. Therefore, when the CCD image sensor is miniaturized, the occurrence of smear due to unnecessary charges at the element isolation portion provided between the photodiode and the vertical transfer path becomes significant.

  However, in each of the above patent documents, no consideration is given to the occurrence of such smear. For this reason, in the CCD image sensor, it is desired to appropriately prevent the occurrence of smear due to unnecessary charges in the element separation portion provided between the photodiode and the vertical transfer path.

  The present invention has been made in view of the above problems, and a solid-state imaging device in which an element separation unit is provided between a photoelectric conversion unit such as a photodiode and a charge transfer unit such as a vertical transfer path like a CCD image sensor. It is an object of the present invention to appropriately prevent the occurrence of smear due to unnecessary charges in the element isolation portion.

  In order to achieve the above object, the present invention provides a plurality of photoelectric conversion units arranged in a matrix on a semiconductor substrate and photoelectrically converting incident light to generate signal charges, and the photoelectric conversion unit generates the photoelectric conversion unit. A plurality of charge transfer units for transferring signal charges in a predetermined direction, and disposed between the photoelectric conversion unit and the charge transfer unit of the adjacent photoelectric conversion unit, and having a conductivity type opposite to the photoelectric conversion unit In the solid-state imaging device having a plurality of element separation units that prevent the signal charge from flowing into the charge transfer unit of the adjacent photoelectric conversion unit by increasing the concentration of the impurities, the charge transfer unit side is more A concentration gradient in which the impurity concentration is higher than that on the photoelectric conversion portion side is provided in each element isolation portion.

  In the concentration gradient, it is preferable that a concentration difference between the highest density portion adjacent to the charge transfer portion and the lowest concentration portion adjacent to the photoelectric conversion portion is 10% or more. The width of the element isolation part is preferably 0.3 μm or less.

  According to another aspect of the present invention, a plurality of photoelectric conversion units arranged in a matrix on a semiconductor substrate and photoelectrically converting incident light to generate signal charges, and the signal charges generated by the photoelectric conversion units are predetermined. A plurality of charge transfer units for transferring in the direction, and between the photoelectric conversion unit and the charge transfer unit of the adjacent photoelectric conversion unit, the concentration of impurities of the conductivity type opposite to the photoelectric conversion unit In the method of manufacturing a solid-state imaging device having a plurality of element separation units that prevent the signal charge from flowing into the charge transfer unit of the adjacent photoelectric conversion unit by increasing the charge transfer unit side, A step of providing each element isolation portion with a concentration gradient in which the impurity concentration is higher than that of the conversion portion side is provided.

  The step of providing the concentration gradient is preferably a step of forming the concentration gradient by performing the implantation of the impurity a plurality of times while changing the angle with respect to the semiconductor substrate. Further, the step of forming the concentration gradient may be performed by implanting the impurity a plurality of times using a plurality of photoresists having different patterns. Further, the step of forming the concentration gradient may be performed by implanting the impurity a plurality of times while shifting the photoresist of the same pattern. Further, the step of forming the concentration gradient by exposing the photoresist a plurality of times using a plurality of photomasks having different patterns and forming an inclined portion or a step portion on the photoresist may be performed. The step of forming the concentration gradient may be performed by exposing the photoresist a plurality of times while shifting the same photomask, and forming an inclined portion or a step portion on the photoresist. Furthermore, the step of forming the concentration gradient may be performed by exposing the photoresist using a halftone mask and forming an inclined portion or a step portion on the photoresist.

  In the present invention, the element separation part provided between the photoelectric conversion part and the charge transfer part is provided with a concentration gradient in which the impurity concentration is higher on the charge transfer part side than on the photoelectric conversion part side. In this way, unnecessary charges generated in the element separation unit flow into the photoelectric conversion unit side according to the concentration gradient, so that it is possible to appropriately prevent smear due to the charge flowing into the charge transfer unit side. . In addition, since the charges generated in the element separation unit can be appropriately collected in the photoelectric conversion unit, the sensitivity of the photoelectric conversion unit can be improved.

It is explanatory drawing which shows the structure of a CCD image sensor roughly. It is explanatory drawing which shows the impurity concentration of PD, VCCD, and an element isolation part. It is explanatory drawing which shows the outline of the manufacturing procedure of a CCD image sensor. It is explanatory drawing which shows the example using a some photoresist. It is explanatory drawing which shows the example which uses the same photoresist shifted. It is explanatory drawing which shows the example using a some photomask. It is explanatory drawing which shows the example which formed the level | step-difference part. It is explanatory drawing which shows the example which uses the same photomask shifting. It is explanatory drawing which shows the example using a halftone mask.

[First Embodiment]
As shown in FIG. 1, the CCD image sensor 10 has an n-type semiconductor substrate 11. The n-type semiconductor substrate 11 is provided with a photodiode (PD) 12 that is a photoelectric conversion unit, a vertical transfer path (VCCD) 13 that is a charge transfer unit, and an element isolation unit 14. A plurality of PDs 12 are provided on the n-type semiconductor substrate 11 and are arranged in a matrix. Each of these PDs 12 photoelectrically converts the incident light and generates a signal charge corresponding to the amount of the light.

  A plurality of VCCDs 13 and element separation units 14 are provided for each PD 12 column. The VCCD 13 is connected to the PD 12 via the read gate 15. The signal charge generated by the PD 12 is read out to the VCCD 13 through the read gate 15. The VCCD 13 transfers the signal charge read from the PD 12 via the read gate 15 in the vertical direction (direction perpendicular to the paper surface) toward a horizontal transfer path (not shown).

  The element isolation unit 14 is disposed between the PD 12 and the VCCD 13 of the adjacent PD 12 to prevent the signal charge generated by the PD 12 from flowing into the VCCD 13 of the adjacent PD 12. The width D1 (see FIG. 2) of the element isolation portion 14 is 0.3 μm or less, and more preferably 0.2 μm or less.

  A p-well layer 16 is formed on the surface of the n-type semiconductor substrate 11. On the surface layer of the p-well layer 16, an n-type layer 20 constituting the PD 12, an n-type layer 21 constituting the VCCD 13, a p + layer 22 serving as the element isolation portion 14, and a p-type layer 23 serving as the read gate 15 are formed. ing. In addition, a p-type layer 24 for suppressing white scratches is formed on the surface of the n-type layer 20. The CCD image sensor 10 is manufactured by forming each of the above parts on the n-type semiconductor substrate 11 using a known technique such as vapor deposition, doping, photolithography, and etching.

  The PD 12 is configured by a junction of the p-well layer 16 and the n-type layer 20. The PD 12 generates electron-hole pairs in response to light incident on these junctions, and accumulates the electrons in the n-type layer 20. The n-type layer 20 and the n-type layer 21 are arranged at a predetermined interval, and are separated from each other by a p-type layer 23 positioned therebetween. The n-type layer 20 is separated from the adjacent n-type layer 21 for the PD 12 by the p + layer 22. This prevents the signal charge accumulated in the n-type layer 20 from unintentionally moving to another region.

  The VCCD 13 includes an n-type layer 21 and a transfer electrode 25 provided on the n-type layer 21. The read gate 15 includes a p-type layer 23 that separates the n-type layer 20 and the n-type layer 21, and a transfer electrode 26 provided on the p-type layer 23. For example, polysilicon is used for the transfer electrodes 25 and 26.

  The signal charges accumulated in the n-type layer 20 are transferred to the n-type layer 21 by applying a voltage to the transfer electrode 26 and changing the potential of the p-type layer 23 between the n-type layers 20 and 21. The The signal charge transferred to the n-type layer 21 is transferred in the vertical direction according to the voltage applied to the transfer electrode 25 at a predetermined timing. Thereby, the signal charge generated by the PD 12 is transferred by the VCCD 13 toward the horizontal transfer path.

  The element isolation unit 14 is configured by a p + layer 22 provided between the n-type layers 21 and 22. Thus, the element isolation part 14 is a potential barrier in which the concentration of impurities of the conductivity type opposite to that of the n-type layer 20 constituting the PD 12 and the n-type layer 21 constituting the VCCD 13 is increased. Accordingly, the element separation unit 14 prevents the signal charge from flowing into the VCCD 13 for the adjacent PD 12.

  A light shielding film 27, a planarizing layer 28, a color filter layer 29, and a microlens 30 are further provided on the p-well layer 16 on which the transfer electrodes 25 and 26 are formed. The light shielding film 27 is formed so as to cover the entire surface of the p well layer 16. In addition, the light shielding film 27 is provided with a plurality of openings 27 a that expose portions corresponding to the respective PDs 12. Thereby, the light shielding film 27 prevents extra light from entering a portion other than the PD 12.

  The planarization layer 28 fills the unevenness on the substrate caused by the transfer electrodes 25 and 26 and forms a plane for forming the color filter layer 29 and the microlens 30. For the planarizing layer 28, a light-transmitting material such as BPSG is used. A plurality of color filter layers 29 are provided corresponding to each PD 12. Each of these color filter layers 29 is, as is well known, composed of three types: one that transmits only red light, one that transmits only green light, and one that transmits only blue light, each having a predetermined pattern. Are arranged in In the CCD image sensor 10, a color image can be obtained by making each color light transmitted through each color filter layer 29 enter the corresponding PD 12. A plurality of microlenses 30 are provided corresponding to each PD 12. Each microlens 30 is formed in a substantially hemispherical shape, and condenses light on the corresponding n-type layer 20 of the PD 12.

  As shown in FIG. 2, the p + layer 22 constituting the element isolation portion 14 has a concentration gradient in which the impurity concentration is higher on the VCCD 13 side than on the PD 12 side. This concentration gradient has a concentration difference of 10% or more between the highest density portion adjacent to the n-type layer 21 constituting the VCCD 13 and the lowest density portion adjacent to the n-type layer 20 constituting the PD 12. ing. Note that the density difference of the density gradient of the element isolation unit 14 is more preferably 20% or more.

  As described above, when a concentration gradient is provided in the element isolation portion 14, light obliquely incident on the PD 12 enters the element isolation portion 14 through the opening end of the light shielding film 27, as indicated by L 1 in FIG. When an incident charge is generated and an unnecessary charge is generated in the element separation unit 14, the charge flows into the PD 12 according to the concentration gradient. Therefore, even when an unnecessary charge is generated in the element separation unit 14 provided between the PD 12 and the VCCD 13, the charge is prevented from flowing into the VCCD 13, and smear caused by the unnecessary charge in the element separation unit 14 is prevented. Occurrence can be prevented appropriately. In addition, since the charges generated by the light incident on the element separation unit 14 can be appropriately collected in the PD 12, the sensitivity of the PD 12 can be improved.

  Next, a manufacturing procedure of the CCD image sensor 10 having the above configuration will be described with reference to FIG. When the CCD image sensor 10 is manufactured, first, as shown in FIG. 3A, a p-well layer 16, an n-type layer 20, and an n-type layer are formed on an n-type semiconductor substrate 11 by using a well-known photolithography technique or the like. After forming the p-type layer 23 and the p-type layer 24, a photoresist 40 having an opening 40 a in a portion corresponding to the p + layer 22 is formed on the n-type semiconductor substrate 11.

  After the photoresist 40 is formed, as shown in FIG. 3B, p-type impurities such as boron are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees. After the implantation at 0 degree, as shown in FIG. 3C, the implantation angle with respect to the n-type semiconductor substrate 11 is tilted by 20 degrees toward the n-type layer 21, and the p-type impurity is implanted again. When the implantation is performed at 20 degrees, as shown in FIG. 3D, the implantation angle with respect to the n-type semiconductor substrate 11 is inclined by 30 degrees toward the n-type layer 21 side, and further p-type impurities are implanted. . Thereby, the p + layer 22 is formed on the n-type semiconductor substrate 11.

  When the impurity is implanted while being inclined toward the n-type layer 21, the impurity is blocked by the corner of the opening 40 a of the photoresist 40, and the impurity does not enter the portion close to the n-type layer 20. The range in which impurities are not incident becomes wider as the implantation angle increases. Therefore, if the implantation angle is changed a plurality of times by changing the implantation angle with respect to the n-type semiconductor substrate 11 as described above, the impurity implantation amount increases and the concentration becomes higher in the portion closer to the n-type layer 21. The p + layer 22 having a concentration gradient in which the impurity concentration is higher on the side than on the PD 12 side can be formed.

  After the p + layer 22 is formed on the n-type semiconductor substrate 11, the photoresist 40 is removed. Thereafter, the transfer electrodes 25, 26, the light shielding film 27, the planarization layer 28, the color filter layer 29, and the microlens 30 are sequentially formed on the n-type semiconductor substrate 11 using a known method. The CCD image sensor 10 is manufactured by the above procedure.

  In this embodiment, the impurities are implanted in the order of 0 degrees, 20 degrees, and 30 degrees. However, the order of the implantation angles is not limited to this, and for example, 30 degrees and 20 degrees opposite to the above. In any order, such as 0 degree order. In this embodiment, the impurity implantation is performed three times at the implantation angles of 0 degrees, 20 degrees, and 30 degrees. However, the implantation angle and the number of implantations are not limited to this, and any angle and number of times are possible. It's okay. Furthermore, in this embodiment, the implantation angle is adjusted by changing the angle of the impurity to be implanted, but on the contrary, the implantation angle may be adjusted by changing the angle of the n-type semiconductor substrate 11.

[Second Embodiment]
Next, a second embodiment of the manufacturing procedure of the CCD image sensor 10 will be described with reference to FIG. The same functions and configurations as those of the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted. In the present embodiment, first, similarly to the first embodiment, the formation of the p-well layer 16, the n-type layer 20, the n-type layer 21, the p-type layer 23, and the p-type layer 24 on the n-type semiconductor substrate 11, and After the formation of the photoresist 40, p-type impurities are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees (see FIGS. 4A and 4B).

  After the impurity implantation, the photoresist 40 is removed, and a photoresist 42 having a pattern different from that of the photoresist 40 is formed on the n-type semiconductor substrate 11 as shown in FIG. The photoresist 42 is provided with an opening 42 a that is narrower than the p + layer 22 and is biased toward the n-type layer 21 at a portion corresponding to the p + layer 22.

  When the photoresist 42 is formed, as shown in FIG. 4D, p-type impurities are implanted again at an implantation angle of 0 degrees. In the photoresist 42 in which the opening 42a is formed as described above, impurities do not enter the portion close to the n-type layer 20. Accordingly, as in the first embodiment, since the impurity implantation amount in the portion close to the n-type layer 21 is increased and the concentration is increased, the VCCD 13 side has a higher impurity concentration concentration than the PD 12 side. A p + layer 22 having can be formed.

  Thus, the p + layer 22 having the above-described concentration gradient can be formed even if the p-type impurity implantation is performed a plurality of times using a plurality of photoresists 40 and 42 having different patterns. In this embodiment, the p-type impurity is implanted twice using the photoresists 40 and 42. However, more photoresists may be prepared and implanted three or more times. Further, in order to suppress channeling at the time of ion implantation, the implantation angle may be changed to 7 degrees or the like.

[Third Embodiment]
Next, a third embodiment of the manufacturing procedure of the CCD image sensor 10 will be described with reference to FIG. In the present embodiment, first, similarly to the above-described embodiments, formation of the p-well layer 16, the n-type layer 20, the n-type layer 21, the p-type layer 23, and the p-type layer 24 on the n-type semiconductor substrate 11, and the photoresist After forming 40, p-type impurities are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees (see FIGS. 5A and 5B).

  After the impurity implantation, the photoresist 40 is once removed, and as shown in FIG. 5C, the photoresist 40 is shifted again by a predetermined amount toward the n-type layer 21 within a range not exceeding the width of the p + layer 22. It is formed on the n-type semiconductor substrate 11. Thereafter, as shown in FIG. 5D, the p-type impurity is implanted again at an implantation angle of 0 degree. As described above, if the photoresist 40 is formed so as to be shifted to the n-type layer 21 side, impurities are not incident on a portion close to the n-type layer 20. Accordingly, as in the above embodiments, the impurity injection amount in the portion near the n-type layer 21 is increased and the concentration is increased, so that the VCCD 13 side has a higher concentration gradient of impurities than the PD 12 side. A p + layer 22 can be formed.

  As described above, the p + layer 22 having the above-described concentration gradient can be formed even if the p-type impurity implantation is performed a plurality of times while shifting the photoresist 40 having the same pattern to the n-type layer 21 side by a predetermined amount. In the second embodiment, a photomask corresponding to each pattern must be prepared for each of the photoresists 40 and 42. However, in this embodiment, only one photomask is sufficient. Therefore, in this embodiment, the manufacturing cost of the CCD image sensor 10 can be reduced as compared with the second embodiment. In the present embodiment, the formation of the photoresist 40 and the implantation of impurities are performed twice, but these times may be three or more times. Further, in order to suppress channeling at the time of ion implantation, the implantation angle may be changed to 7 degrees or the like.

[Fourth Embodiment]
Next, a fourth embodiment of the manufacturing procedure of the CCD image sensor 10 will be described with reference to FIG. In the present embodiment, first, as shown in FIG. 6A, a p-well layer 16, an n-type layer 20, an n-type layer 21, a p-type layer 23, and a p-type layer 24 are formed on an n-type semiconductor substrate 11. Thereafter, a positive photoresist 44 is applied on the n-type semiconductor substrate 11, and a photomask 46 on which a predetermined pattern is formed is set on the photoresist 44.

  The photomask 46 includes a transparent glass substrate 47 having translucency, and a light shielding film 48 provided on the glass substrate 47. In the light shielding film 48, an opening 48 a corresponding to the width of the p + layer 22 is formed in a portion corresponding to the p + layer 22. That is, each of the openings 48 a becomes a pattern of the photomask 46.

  After applying the photoresist 44 and setting the photomask 46, the pattern of the photomask 46 is transferred to the photoresist 44 by irradiating light from above to expose the photoresist 44.

  When the photoresist 44 is exposed using the photomask 46, the photomask 46 is replaced with a photomask 50 as shown in FIG. Similar to the photomask 46, the photomask 50 includes a glass substrate 51 and a light shielding film 52. In the light shielding film 52, an opening 52 a is formed in a portion corresponding to the p + layer 22, which is narrower than the p + layer 22 and is biased toward the n-type layer 21.

  After the photomask 46 is replaced with the photomask 50, the pattern of the photomask 50 is transferred to the photoresist 44 by exposing the photoresist 44 again. As described above, when the photoresist 44 is exposed a plurality of times using a plurality of photomasks 46 and 50 having different patterns, a portion corresponding to the p + layer 22 of the photoresist 44 according to the width of each opening 48a and 52a. A difference in exposure occurs, and the amount of exposure in the portion close to the n-type layer 21 is greater than the portion close to the n-type layer 20.

  After the exposure of the photoresist 44 using the photomask 50, the photomask 50 is removed and the n-type semiconductor substrate 11 and the photoresist 44 are immersed in a developing solution. Remove from resist 44.

  When the exposed portion is removed from the photoresist 44, an opening 44a is formed in a portion corresponding to the p + layer 22 of the photoresist 44 as shown in FIG. At this time, an inclined portion 44b is formed in the opening 44a so as to overlap a portion corresponding to the p + layer 22 of the n-type semiconductor substrate 11 and a portion close to the n-type layer 20 due to a difference in exposure amount at the time of performing multiple exposures. It is formed.

  Then, after removing the exposed portion from the photoresist 44, as shown in FIG. 6D, p-type impurities are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees. When the impurity implantation is performed using the photoresist 44 having the inclined portion 44b, the impurity implantation amount gradually changes in the portion corresponding to the inclined portion 44b according to the inclination, that is, the change in the thickness of the photoresist 44. As a result, the closer to the n-type layer 20, the smaller the amount of implanted impurities, so that the p + layer 22 having a higher concentration gradient of impurities on the VCCD 13 side than on the PD 12 side is formed as in the above embodiments. can do.

  As described above, by exposing the photoresist 44 a plurality of times using a plurality of photomasks 46 and 50 having different patterns and forming the inclined portion 44b in the photoresist 44, the p + layer 22 having the above-described concentration gradient can be formed. Can be formed. In the present embodiment, the photoresist 44 is exposed twice using the photomasks 46 and 50. However, more photomasks may be prepared and exposed three times or more. In the present embodiment, the inclined portion 44b is formed in the photoresist 44 by exposing the photoresist 44 a plurality of times using a plurality of photomasks 46 and 50 having different patterns. However, the present invention is not limited to this. A step 44c may be formed in the photoresist 44 by exposing the photoresist 44 a plurality of times in the same procedure as described above, as shown in FIG. Further, in order to suppress channeling at the time of ion implantation, the implantation angle may be changed to 7 degrees or the like.

[Fifth Embodiment]
Next, a fifth embodiment of the manufacturing procedure of the CCD image sensor 10 will be described with reference to FIG. In the present embodiment, first, as in the fourth embodiment, formation of the p-well layer 16, the n-type layer 20, the n-type layer 21, the p-type layer 23, and the p-type layer 24 on the n-type semiconductor substrate 11 is performed. The resist 44 is applied and the photomask 46 is set, and the photoresist 44 is exposed using the photomask 46 (see FIG. 8A).

  When the photoresist 44 is exposed using the photomask 46, the photomask 46 is shifted to the n-type layer 21 side by a predetermined amount within a range not exceeding the width of the p + layer 22, as shown in FIG. In this state, the photoresist 44 is exposed again. As described above, even if the photomask 46 is shifted to the n-type layer 21 side and the photoresist 44 is exposed a plurality of times as in the case of the fourth embodiment, the photoresist 44 depends on the shift amount of the photomask 46. A difference in exposure amount occurs in the portion corresponding to the p + layer 22, and the portion near the n-type layer 21 has a larger exposure amount than the portion close to the n-type layer 20.

  After the photomask 46 is shifted and the exposure of the photoresist 44 is performed a plurality of times, the photomask 46 is removed and the exposed portion is removed from the photoresist 44 to form an opening 44 d in the photoresist 44. At this time, an inclined portion 44e is formed in the opening 44d so as to overlap with a portion close to the n-type layer 20 corresponding to the p + layer 22 of the n-type semiconductor substrate 11 due to an exposure amount difference when performing multiple exposures. In addition to being formed, an inclined portion 44 f is formed so as to overlap the n-type layer 21.

  Then, after removing the exposed portion from the photoresist 44, as shown in FIG. 8D, p-type impurities are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees. In this way, as in the fourth embodiment, since the amount of impurity implantation decreases as the inclined portion 44e approaches the n-type layer 20, the concentration gradient of impurities on the VCCD 13 side is higher than that on the PD 12 side. A p + layer 22 having can be formed.

  As described above, the p + layer 22 having the above-described concentration gradient can also be formed by shifting the same photomask 46 and exposing the photoresist 44 a plurality of times to form the inclined portion 44e in the photoresist 44. . In the fourth embodiment, a plurality of photomasks 46 and 50 must be prepared. In the present embodiment, only one photomask 46 is required. Therefore, in this embodiment, the manufacturing cost of the CCD image sensor 10 can be reduced as compared with the fourth embodiment. In the present embodiment, the photomask 46 is shifted and the photoresist 44 is exposed twice, but the number of exposures may be three or more. Further, in order to suppress channeling at the time of ion implantation, the implantation angle may be changed to 7 degrees or the like. In the present embodiment, the inclined portions 44e and 44f are formed in the photoresist 44, but may be stepped portions as shown in FIG.

[Sixth Embodiment]
Next, a sixth embodiment of the manufacturing procedure of the CCD image sensor 10 will be described with reference to FIG. In this embodiment, first, as shown in FIG. 9A, formation of a p-well layer 16, an n-type layer 20, an n-type layer 21, a p-type layer 23, and a p-type layer 24 on the n-type semiconductor substrate 11, and After applying the photoresist 44, a halftone mask 54 on which a predetermined pattern is formed is set on the photoresist 44.

  In the halftone mask 54, a gradation portion 54a having an optical density gradient in which the transmittance gradually decreases from the n-type layer 21 side toward the n-type layer 20 side is formed in a portion corresponding to the p + layer 22. ing. Further, the portion other than the gradation portion 54a of the halftone mask 54 is a light shielding portion 54b that prevents light from being transmitted. That is, the pattern of the halftone mask 54 is constituted by the gradation portion 54a.

  After the halftone mask 54 is set, the pattern of the halftone mask 54 is transferred to the photoresist 44 by irradiating light from above to expose the photoresist 44. As described above, when the photoresist 44 is exposed using the halftone mask 54, an exposure amount difference occurs in a portion corresponding to the p + layer 22 of the photoresist 44 in accordance with the optical density gradient of the gradation portion 54a. The portion closer to the n-type layer 21 has a larger exposure amount than the portion closer to the n-type layer 20.

  After the photoresist 44 is exposed using the halftone mask 54, the halftone mask 54 is removed and the exposed portion is removed from the photoresist 44 to form an opening 44g in the photoresist 44 as shown in FIG. 9B. To do. At this time, the opening 44g overlaps with the portion close to the n-type layer 20 corresponding to the p + layer 22 of the n-type semiconductor substrate 11 due to the exposure amount difference corresponding to the optical density gradient of the gradation portion 54a. An inclined portion 44h is formed.

  Then, after removing the exposed portion from the photoresist 44, as shown in FIG. 9C, p-type impurities are implanted into the n-type semiconductor substrate 11 at an implantation angle of 0 degrees. In this way, as in the above embodiments, the amount of impurity implantation decreases as the inclined portion 44h approaches the n-type layer 20, so that the VCCD 13 side has a higher concentration gradient of impurities than the PD 12 side. A p + layer 22 can be formed.

  Thus, the p + layer 22 having the above-mentioned concentration gradient can also be formed by exposing the photoresist 44 using the halftone mask 54 and forming the inclined portion 44 h in the photoresist 44. In the present embodiment, the inclined portion 44h is formed in the photoresist 44, but it may be a stepped portion as shown in FIG. Further, in order to suppress channeling at the time of ion implantation, the implantation angle may be changed to 7 degrees or the like.

  In each of the above embodiments, the n-type layer 20 constituting the PD 12 and the n-type layer 21 constituting the VCCD 13 are formed, and then the p + layer 22 constituting the element isolation portion 14 is formed. The formation order is not limited to this, and may be any order, for example, the p + layer 22 is formed first.

  In each of the above embodiments, the example in which the present invention is applied to the CCD image sensor 10 has been described. However, the present invention is not limited to this, and an element separation unit is provided between the photoelectric conversion unit and the charge transfer unit. The present invention may be applied to other types of solid-state imaging devices.

10 CCD image sensor (solid-state image sensor)
11 n-type semiconductor substrate 12 PD (photoelectric conversion part)
13 VCCD (charge transfer unit)
14 Element isolation part 40, 42, 44 Photoresist 46, 50 Photomask 54 Halftone mask

Claims (10)

A plurality of photoelectric conversion units arranged in a matrix on a semiconductor substrate and photoelectrically converting incident light to generate signal charges;
A plurality of charge transfer units for transferring the signal charges generated by the photoelectric conversion unit in a predetermined direction;
The signal charge is disposed between the photoelectric conversion unit and the charge transfer unit of the adjacent photoelectric conversion unit, and the signal charge is adjacent to the photoelectric conversion by increasing a concentration of an impurity of a conductivity type opposite to the photoelectric conversion unit. In a solid-state imaging device having a plurality of element separation portions to prevent flowing into the charge transfer portion of the part,
A solid-state imaging device, wherein a concentration gradient in which the impurity concentration is higher on the charge transfer unit side than on the photoelectric conversion unit side is provided in each of the element isolation units.   2. The density gradient according to claim 1, wherein a density difference between a highest density part adjacent to the charge transfer unit and a lowest density part adjacent to the photoelectric conversion unit is 10% or more. The solid-state imaging device described.   3. The solid-state imaging device according to claim 1, wherein a width of the element isolation portion is 0.3 μm or less. A plurality of photoelectric conversion units arranged in a matrix on a semiconductor substrate and photoelectrically converting incident light to generate signal charges;
A plurality of charge transfer units for transferring the signal charges generated by the photoelectric conversion unit in a predetermined direction;
The signal charge is disposed between the photoelectric conversion unit and the charge transfer unit of the adjacent photoelectric conversion unit, and the signal charge is adjacent to the photoelectric conversion by increasing a concentration of an impurity of a conductivity type opposite to the photoelectric conversion unit. In a method for manufacturing a solid-state imaging device having a plurality of element separation portions that prevent the flow from flowing into the charge transfer portion of a portion,
A method of manufacturing a solid-state imaging device, comprising: providing each element isolation unit with a concentration gradient in which the impurity concentration is higher on the charge transfer unit side than on the photoelectric conversion unit side.   5. The method of manufacturing a solid-state imaging device according to claim 4, wherein the concentration gradient is formed by performing the implantation of the impurity a plurality of times while changing the angle with respect to the semiconductor substrate.   5. The method of manufacturing a solid-state imaging device according to claim 4, wherein the concentration gradient is formed by performing the implantation of the impurities a plurality of times using a plurality of photoresists having different patterns.   5. The method of manufacturing a solid-state imaging device according to claim 4, wherein the concentration gradient is formed by performing the impurity implantation a plurality of times while shifting the photoresist having the same pattern.   5. The solid according to claim 4, wherein the concentration gradient is formed by exposing the photoresist a plurality of times using a plurality of photomasks having different patterns, and forming an inclined portion or a step portion on the photoresist. Manufacturing method of imaging device.   5. The solid-state imaging device according to claim 4, wherein the density gradient is formed by exposing the photoresist a plurality of times while shifting the same photomask, and forming an inclined portion or a step portion on the photoresist. Production method.   5. The method of manufacturing a solid-state imaging device according to claim 4, wherein the density gradient is formed by exposing the photoresist using a halftone mask and forming an inclined portion or a step portion on the photoresist.

Images