201143049 六、發明說明： 【發明所屬之技術領域】 所揭示之標的物大體上係關於用於製造固態影像感 測器之結構以及方法。 本申請案主張2009年11月8日申請之美國臨時專利 申請案第61/259，180號之優先權。 【先前技術】 攝影設備（諸如，數位相機以及數位攝錄影機）可包 含電子影像感測器’該等電子影像感測器擷取光以供處理 成靜止影像或視訊影像。電子影像感測器通常包含上百萬 個光電轉換單元，諸如光電二極體。 、固態影像感測器可為電荷耦合裝置（CCD)類型或互 補金屬氧化物半導體（CM〇s)類型中之任一者。在任一 類型之影像感測器中’光電轉換單元形成於一基底中且配 ，成二維陣列。影像感測器通常包含各自包括一光電轉換 早之上百萬個像素以提供—高解析度影像。為改良光掏 取之效率，某些影像感測器具有光導（或波導），以引導光 朝向該等光電轉換單元。料光導可包括透光材料（例如， =Si3N4之氮切透紐料之折射率高於周圍絕緣材 ^例如，氧切）之折射率，使得在料之㈣處存在 =反射⑽止光外泄。或者’該f光導可在側壁上呈有 ，屬塗層以提供反射，且填充有透明材料⑷如，氧化石^ 或有機樹脂或旋塗式玻璃⑼G))。_像素可包括一個以 201143049 亡光導，該等光導中一者堆疊於另一者上方以形成級聯式 光導。在距基底任一高度處之光導通常彼此相距一給定間 距，且在距基底任-給定高度處共用—制水平橫截面輪 廓。保持一恆定間距提供沿著從左至右以及從上至下方向 (平打於像素陣狀平面）郷於影像感卿之面上的影像 之均-取樣’由此更好地匹配像素在諸如電腦顯示器之顯 示器及印表機上排列的方式。 【發明内容】 根據本發明之第-方面，一種影像感測器包括一像素 陣列，該像素_包括概個像素，其中每—像素包括:（a) 一光電轉換單元，該光電轉換單元在—絕緣層下方且在亦 埋置於該職層巾之複數條t線下方；以及⑻一光導， =光導埋置於該絕緣層中且在該複數條電線之間以將光透 轉換單元，其中在一方向上並排地排列且在該 =個像素當中的五個或五個以上像素當中的每一對連續 光導之間的一水平間隔跨該五個或五個以上像 =3=1該光導可包含染料或彩色顏料。該彩色 厂、".·，、有機顏料，或無機顏料，或有機金屬顏料。 在第方面巾’軸魏五個或五個 ::連續像素之該等光導之間的一水平間距 該==:二=水平間距跨 望該水平間距跨該五個或五個以上;素變化= 6 201143049 大 在第一方面中，該像素可更包括一彩色濾光片，該彩 色濾光片經耦接以經由該光導將光透射呈該光電轉換單 元’在該五個或五個以上像素當中的每—對連續像素之該 等彩色濾光片之間存在一間隙，該間隙具有一跨該五個或 五個以上像素非單調地變化的寬度。係期望該間隙寬度跨 在一方向上並排地排列之十六個像素變化〇 1 或更 大。更期望該間隙寬度跨該十六個像素變化0.2 μιη或更 大。 在第-方面中，該像素可更包括一彩色慮光片，該彩 色濾光^經耦接以經由該光導將光透射至該光電轉換單 元，在该五個或五個以上像素當中的每一對連續像素之該 等彩色遽光片之間存在-間隙，且—間隙間距跨該五個或 五個以上像素非單調地變化，該間關距為該等間隙之每 :對連續中心線之_-水平距離（在平行於光電轉換單 7L之平面的平財）。係期望該間隙間距跨在—方向上並 地排列之十六轉素㈣0] μιη蚊Α 距跨該十六個像素變化0.2 μιη或更大。m棚隙間 該間隙可包含空氣或一氣體。或者， 液體或固體材料，該液體或固體材料具二日二二二 片之-折射率低至少20%的折射率 ;== 彩色滤光片之間具有-不大於〇45帅之^間隙在鄰近 期望該間隙用凸頂板作為頂部。更進—I度。進一步係 光片之底部至凸頂板之頂部存在至少^ _望從彩色遽 201143049 根據本發月之第二方面，一種影像感測器包括一像素 陣列’該像素陣列包括複數個像素，其中每—像素包括:u 一光電轉換單元’該光電轉換單元在-絕緣層下方且在亦 埋置於該絕緣層中之複數條電線下方；以及⑴一光導， 該光導埋置於該絕緣層巾且在該複數條·之間以將光透 射至該光電轉換單元’其中該光導具有_跨五個或五個以 上像素非單調地變化之寬度。進-步侧魏光導之該寬 度針對在該五麵五個以上像素#中的經組態以僅债測較 短波長之光的—像素小於針對在該五個或五個以上像素當 中的經組態以侧較長波長之光的另—像素。亦進一步係 期望該光導之該寬度針對—藍色像素小於針對—紅色像 素:亦進-步係期望該光導之該寬度冊—藍色像素小於 針對-綠色像素。亦進―步係鮮該光導之該寬度針對一 綠色像素小於針對一紅色像素。 ^在上文中，係期望該等彩色濾光片包括一著色劑。該 者色劑可為染料或彩色綱。貞料可為有機顏料、無機 顏料，或有機金屬顏料。 ”' 根據本發明之第三方面，提供一種用 ，偵測-影像之方法，該方法包括：（a)在—絕 且在埋置於魏緣層巾之概條電線下方提供複數個光 ;以及⑴提供埋置於該絕緣層中且在該複數 條電線之_複數個光導㈣光透射至該光電轉換單元， ，中在y方向上並排地排狀在該複數個像素當中的五個 ’五個以上像素當中的每—對連續像素之該等光導之間的 8 201143049 水平間隔跨該五個或五個以上像素非單調地變化。 在第三方面中，係期望該影像感測器具有來自第—方 面之期望特徵中之任一者。 _根據本發明之第四方面，—種影像感顧包括一像素 一忠’該像素陣列包括複數個像素，其中每一像素包括：（a) ㈣電轉換單元，該光電轉換單元在—絕緣層下方且在亦 9#於°亥絕緣層中之複數條電線下方；以及（b) 一光導， j導埋置於賴緣層巾且在該複數條電線之間以將光透 、u°亥光電轉換單兀’其中在—方向上並排地排列且在該 =固像素當中的五個或五個以上像素當中的每一對連續 =之該等光導之間的—水平間隔跨該五個或五個以上像 I t與減小之間交替。該光導可包含染料或彩色顏 結社4色顏料可為有機顏料，或無機顏料，或有機金屬 顧料。 色減面中’該像素可更包括一彩色縣片，該彩 一、、.坐耦接以經由該光導將光透射至該光電轉換單 US個或五個以上像素當中的每一對連續像素之該 光片之間存在—間隙’該間隙具有—跨該五個或 素在增大與減小之間交替的寬度。_望該間 ^又5在一方向上並排地排列之十六個像素變化OJ μιη :大。更期望該間隙寬度跨該十六個像素變化0 吏大0 〆 在第四方面中’該像素可更包括—彩色濾光片 k光片經耦接以經由該光導將光透射至該光電轉換單 9 201143049 在该五個或五個以上像素當 等彩色據光片之間存在一間隙，且— ： 像素之該 五個以上像素在增大與減小間^間距跨該五個或 間隙之每一對連續中心線之‘ ί 2 f隙間距為該等 該間隙間距跨針六個像錢化0.2 μπΓ=^Α。更期望 根據本發明之第七方面，— 〔 陣列，該像素陣列包括複數個像素，ί中每像素 該絕緣層中之複數條電線下:;= 埋置於該絕緣層巾且在該複 = 光電轉換單元，其中該光導具有一跨五丄= 象素在增大與減奴間㈣之寬 _ s=:nr上像素當中== 進像^中的經組態以偵測較長波長之光的另-像素。亦 紅色=期ΐ該光導之該寬度針對—藍色像素小於針對一 夸色像素。亦進一步係期望該光導之該寬度針對-藍色像 針對綠色像素。亦進—步係期望該光導之該寬度 ^ 、、彔色像素小於針對一紅色像素。 測ϋί據本發明之第八方面，提供—種用於使用—影像感 方'則—影像之方法，該方法包括：（a)在一絕緣層下 且在埋置於該絕緣層中之複數條電線下方提供複數個光 201143049 元；以及⑻提供埋置於該絕緣層中且在該 複數個光導以將光透射域光電轉換單元， =在了方向上並排地排列且在該複數個像素當中的 5 hi以上像素當中的每一對連續像素之該等光導 ^水平_跨該五個或五似上像素在增续減小之^交 在第八方面中，係期望該影像感測器具有來自 面之期望特徵中之任一者。 方 根據本發明之第九方面，一種影像感測器包括 陣列’該像素陣列包括複數個像素，其中每一像素包括a、 一光電轉換單元，該光電轉換單元在—絕緣層下方且在= 埋置於該絕緣層中之複數條電線下方；以及一光 该光導埋置於該絕緣層中且在該複數條電線之間以將 射至該光電轉換單元’其中在-方向上並排地排列且在节 複數個像素當中的五個或五個以上像素當中的每一對連= 像素之該等光導之_-水平卩板跨該五個或五個以 素具有-較寬水平_，該較寬水平_緊跟在一較 隔之後，且該較寬水平間隔自身由另—較窄間隔緊跟。該 光導可包含祕或彩色顏料。該彩色顏料可為有機顏料了 或無機顏料，或有機金屬顏料。 【實施方式】 揭示一種影像感測器，該影像感測器具有一像素陣 列，該像素陣列包括各自包括—光電轉換單元之複數個像 201143049 素忒等像素中之每一者包括一光導，該光導埋置於一絕 緣層中且在亦埋置於魏緣體層中之多條電線之間 ，以使 $透射至縣電賴單元。齡_绩長之光之像素的光 導可在其底部處具有—寬度（「底部寬度」），該寬度小於僅 偵測車又長波長之光的另—像素之光導。該光導之—底部處 的-垂直中心_—橫向側上之—鄰近像素的光導之一莖 直中心線的—距離可大於距—相對橫向側上之-鄰近像素 的光導之-垂直中錢的—轉。該光導之—底部與一橫 向側上之-鄰近像素的光導之—底部之間的-間隔(「底部 間隔」)可大於該光導之該底部與—相對橫向側上之一鄰近 像素的光導之-底部之_ —間隔。該像素可包括一彩色 ;慮光片’知色濾光片包括彩色材料。在該複數個像素當 中的每-對並排像素之彩色據光片之間可存在—間隙。在 -對並排像素與另-對並排像素之間，間隙之寬度（「間隙 寬度」）可不同。跨在-方向上並排地排狀三個或三個以 上像素的間隙之間距（「間隙間距」）可變化15%或更少。 在-影像感測H巾且詳言之-彩色影像感測器中具有以上 技術特徵中之-或多者准許在絕緣層下以及絕緣層内之積 體電路特徵（例如’閘極電極、聚觸點（PGly⑺ntaet)、電 ，以及擴散觸點）之較高佈局密度。一基底支樓光電轉換 單兀’且可為輕雜成第—導電類型（較佳地P型）且進 二步較佳地具有在5el4W肖5el5/em3之間的播雜濃度的 半導體基底。基底106可為摻雜濃度超過lel9/em3之重推 雜P基底上的P磊晶層（P-epi layer)。舉例而言，基底1〇2 12 201143049 可具有以5el4/cm3與5el5/c 3 諸如，P+基底上之習知層的切’ 麥看圖式，更特定言之，參看參考 器H)，該影像感測器1Q包括像素14之=— ==r22之-群組連接至列 綠光讀二器電 ΐί待對輸出錢18之㈣樣本執行減法與放°=8 = tADC24將該（該等）類比信號轉換成ΑΙΧ：^ 匯=排6上之數位影像資料。若影像感測_ ι〇為彩 像感測器，則像素陣列12包括—彩色濾、光片陣列，該^ 1片陣列包括以使得每-像素14有―個彩色據光片^ 方式以二維排列之多個彩色濾光片。 圖8A說明-彩色濾、光片陣列之實例，該彩色滤光片 陣列可女置於像素陣列12之上且作為像素陣列12之部 分。圖8A展示一拜耳原色圖案（Bayer ―町⑶心 pattern)’該圖案包括彩色濾光片之2χ2區塊（虛線内）之 重複二維陣列，該等彩色濾光片各自具有綠色（G)、紅色 (R)以及藍色（B)中之一者。—對綠色滤光片沿著該2χ2 區塊之一對角線安置。一紅色濾光片以及一藍色濾光片所 組成之一對沿著另一對角線安置。在彩色濾光片陣列之此 實施例中，該等彩色濾光片在頁面上從左至右且在頁面上 從上至下並排地排列。 201143049 D8B說明圖8A中所展示之彩色遽光片陣列之一替代 實施例。在此變化中’彩色渡光片排列之方向相對於從左 至右以及從上至下方向，以及相對於影像掃描之從下至上 方向（以向上指之箭頭所展示）旋轉45度。該等像素14 在使用此彩色濾光片陣列之影像感測器1〇之實施例中同 樣地排列。 圖9展示像素陣列12之像素14之一實施例的示意 圖。像素14包括一光電轉換單元1〇2。舉例而言，該光電 轉換單元102可為光電二極體。該光電轉換單元1〇2可經 由傳送閘117連接至一重設開關112。該光電轉換單元1〇2 亦可經由一輸出（亦即，源極隨耦器）電晶體116耦接至 一選擇開關114。電晶體112、114、116、117可Λ場力女雷 晶_)。傳送問m之閉極連接至 重5叉開關112之閘極連接至—RST(n)線118。重設開關112 之汲極節點可連接至一取線12〇。選擇開關114之閘極連 接至一SEL線122。選擇開關114之源極節點連接至—〇υτ 線124。RST⑻線118、SEL⑻線122以及TF⑻線126可 為像素陣列12中之一整列像素所共用。同樣地，1]^線12〇 以及OUT線124可為像素陣列12中之一整行像素所共 用。RST⑻線118、SEL⑻線122以及TF⑻線121連接至 列解碼器20且為控制線22之部分。〇UT(m)線124連接至 光讀取器16且為垂直信號線18之部分。 圖1A展示基底1〇6上之彩色影像感測器1()之像素陣 列12中的四個鄰近像素14之一實施例，該四個鄰近像素 201143049 14在一方向上並排地排列。所展示之每一像素14包括一 光電轉換單元102a或102b，光電轉換單元102a或102b 將光子能量轉換成電荷。關於使用4T像素架構（諸如圖9 中所展示）之像素陣列12或其變體（諸如，在多對光電二 極體102與傳送開關117當中共用重設開關112、選擇開 關114以及輸出電晶體116)，閘極電極i〇4a、1〇4b可各 自為用以傳送電荷的不同傳送開關117之閘極電極。或 者，關於使用3T像素架構（諸如圖1〇中所展示）之像素 陣列12，閘極電極104a、i〇4b可分別為用以重設光電轉 換單元102a、l〇2b之不同重設開關112之閘極電極。閘極 電極104c可為像素陣列12内伺服不同功能之電晶體（例 如，重没開關112，或選擇開關114，或輸出電晶體116) 的閘極電極。閘極電極1 〇4a、1 〇4b、104c以及轉換單元 102a 102b形成於基底1〇6上或基底1〇6中。閘^電極 104a、lG4b、lG4e，光電轉換單元1Q2a、丨咖，以及基底 106可被覆蓋在保護層230 τ，保護層23〇包括氣化石夕且 具有200埃至1〇〇〇埃之間的厚度。保護層23〇使基底1⑽ 與金屬離子以及濕氣絕緣。絕緣層11〇之層覆蓋基底1〇6。 多條電線108埋置於絕緣層m中且在閘極電極馳、 104b、104C上方。電線⑽彳為包括銘或銅之導電互連電 線。其他互連電線（未圖示）可形成於其他佈線平面中， 母=平面包括多條互連電線（該多敍連電線可為金屬的） 且ί ^電極難、娜、1⑽上方處於與電線刚不同 之南度處。電線108可經由導電（諸如，金屬）介層孔連 15 201143049 接，導電介層孔橫跨至在不同高度處之鄰近佈線平面上 互連電線。 ' ' 光電轉換單元102a、102b可分別與下光導316a、316b 配對，下光導316a、316b埋置於絕緣層11〇中且在多條電 線刪之間。下光導她、鳩可包括折射率高於絕緣層 110 (為外部材料，例如，氧化矽）的透射材料（例如，嗜 如S^N4之氮化矽），且使用該光導與該外部材料之間的全 内反射來幫助防止光外泄出該等下光導。或者，下光導可 填充有諸如旋塗式玻璃（S0G)之透明材料或有機材料之 可甚至包括彩色材料(諸如，有機或無機顏 料或有機金屬顏料），且在其側壁上具有反射金屬塗 射金屬塗層類型）以向内反射光從而幫助防止光外泄曰出下 光導。 上光導130可位於下光導31如、鳩上方且可包 =下光導316a、316b相同之（多種）材料 同材料。上光導130以及下光導316a、316b可均為全^ = 反射金屬塗層類型，或其中-者可為-種 類型。上光導130之頂端寬於底端， 在底k處上先導13G會合下料施、。201143049 VI. DESCRIPTION OF THE INVENTION: TECHNICAL FIELD The disclosed subject matter is generally related to structures and methods for fabricating solid state image sensors. The present application claims priority to U.S. Provisional Patent Application Serial No. 61/259,180, filed on November 8, 2009. [Prior Art] A photographing device (such as a digital camera and a digital video camera) may include an electronic image sensor'. The electronic image sensor picks up light for processing into a still image or a video image. Electronic image sensors typically contain millions of photoelectric conversion units, such as photodiodes. The solid state image sensor can be any of a charge coupled device (CCD) type or a complementary metal oxide semiconductor (CM〇s) type. In any type of image sensor, the photoelectric conversion units are formed in a substrate and arranged in a two-dimensional array. Image sensors typically contain a million pixels that are each included in a photoelectric conversion to provide a high resolution image. To improve the efficiency of the pupil, some image sensors have a light guide (or waveguide) to direct light toward the photoelectric conversion units. The material light guide may comprise a light transmissive material (for example, the refractive index of the nitrogen cut through the =Si3N4 is higher than the refractive index of the surrounding insulating material, for example, oxygen cut), so that there is a reflection at the (4) of the material (reflection) . Alternatively, the f-light guide may be present on the side wall to provide a reflection and to be filled with a transparent material (4) such as an oxidized stone or an organic resin or a spin-on glass (9) G). The _pixel may include a light guide with 201143049, one of which is stacked over the other to form a cascaded light guide. The light guides at any height from the substrate are typically spaced apart from each other by a given spacing and are shared at any given height from the substrate to produce a horizontal cross-sectional profile. Maintaining a constant spacing provides a uniform-sampling of the image along the left-to-right and top-to-bottom (flat-to-pixel plane) image on the image-sensing surface, thereby better matching pixels in, for example, a computer The display of the display and the way it is arranged on the printer. SUMMARY OF THE INVENTION According to a first aspect of the present invention, an image sensor includes an array of pixels, wherein the pixel includes a substantially pixel, wherein each pixel includes: (a) a photoelectric conversion unit, wherein the photoelectric conversion unit is in- Below the insulating layer and below a plurality of t-lines that are also embedded in the blanket; and (8) a light guide, the light guide is embedded in the insulating layer and between the plurality of wires to convert the light through the unit, wherein A horizontal interval between each pair of continuous light guides arranged side by side in a side direction and between five or more pixels among the = pixels across the five or more images = 3 = 1 Contains dyes or color pigments. The color factory, ".·, organic pigments, or inorganic pigments, or organometallic pigments. In the first aspect of the 'Axis Wei five or five:: consecutive pixels of a horizontal spacing between the light guides ==: two = horizontal spacing across the horizontal distance across the five or more; = 6 201143049 In a first aspect, the pixel may further comprise a color filter coupled to transmit light through the light guide to the photoelectric conversion unit 'in the five or more Each of the pixels has a gap between the color filters of successive pixels, the gap having a width that varies non-monotonically across the five or more pixels. It is desirable that the gap width varies by sixteen pixels arranged side by side in one direction, 〇 1 or more. It is more desirable that the gap width varies by 0.2 μm or more across the sixteen pixels. In a first aspect, the pixel may further include a color filter that is coupled to transmit light to the photoelectric conversion unit via the light guide, each of the five or more pixels There is a gap between the color pupils of a pair of consecutive pixels, and the gap spacing varies non-monotonically across the five or more pixels, the gap being each of the gaps: a continuous centerline _-horizontal distance (in the plane parallel to the plane of the photoelectric conversion sheet 7L). It is desirable that the gap spacing spans in the - direction and is arranged in a sixteen-transliter (four) 0] μιη mosquito 变化 distance of 0.2 μηη or more across the sixteen pixels. m Between the gaps The gap may contain air or a gas. Or a liquid or solid material having a refractive index of at least 20% lower than the refractive index of the second day of the second or second piece; == having a gap between the color filters of - not more than 〇45 A convex top plate is used as the top adjacent to the gap. More progress - I degree. Further, there is at least a portion from the bottom of the light sheet to the top of the convex top plate. From the second aspect of the present month, an image sensor includes a pixel array 'the pixel array includes a plurality of pixels, each of which The pixel includes: u a photoelectric conversion unit 'below the insulating layer and under a plurality of wires buried in the insulating layer; and (1) a light guide buried in the insulating layer and The plurality of strips are transmitted to the photoelectric conversion unit 'where the light guide has a width that varies non-monotonically across five or more pixels. The width of the advance-step side light guide is for pixels in the five-sided five or more pixels # configured to only measure light of shorter wavelengths - the pixels smaller than for the five or more pixels Configure another pixel with light of longer wavelengths on the side. It is further desirable that the width of the light guide is such that the blue pixel is smaller than the target for the red pixel: the step size is that the blue pixel is smaller than the target green pixel. The width of the light guide is also smaller for a green pixel than for a red pixel. ^ In the above, it is desirable that the color filters include a colorant. The toner can be a dye or a color scheme. The dip can be an organic pigment, an inorganic pigment, or an organometallic pigment. According to a third aspect of the present invention, there is provided a method of detecting, imaging, the method comprising: (a) providing a plurality of lights underneath a wire that is embedded in a Wei edge layer; And (1) providing a plurality of light guides (four) buried in the insulating layer and transmitting light to the photoelectric conversion unit in the plurality of wires, wherein five of the plurality of pixels are arranged side by side in the y direction. Each of the five or more pixels-to-continuous pixels of the 8 201143049 horizontal spacing between the light guides varies non-monotonically across the five or more pixels. In a third aspect, it is desirable for the image sensor to have Any one of the desired features of the first aspect. According to a fourth aspect of the invention, the image sensing comprises a pixel-loyalty. The pixel array comprises a plurality of pixels, wherein each pixel comprises: (a) (4) an electrical conversion unit under the insulating layer and under a plurality of wires in the insulating layer of the insulating layer; and (b) a light guide, the guiding light is embedded in the rim blanket and is Multiple wires between the wires , u ° hai photoelectric conversion unit 其中 where each of the five or more pixels among the = solid pixels arranged side by side in the - direction = horizontal interval between the light guides The five or more images may alternate between It and decrease. The light guide may comprise a dye or a color pigment. The 4-color pigment may be an organic pigment, or an inorganic pigment, or an organometallic compound. The pixel may further include a color county piece, the color is coupled to transmit light through the light guide to the optical film of each pair of consecutive pixels of the photoelectric conversion unit US or more than five pixels There is a gap - the gap has - a width that alternates between the increase and decrease across the five or tens. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ It is more desirable that the gap width varies from the sixteen pixels by 0 吏 to 0. In the fourth aspect, the pixel may further include a color filter k-ray sheet coupled to transmit light to the photoelectric conversion via the light guide. Single 9 201143049 in the five or more pixels when There is a gap between the color light films, and - : the five or more pixels of the pixel are between the increasing and decreasing intervals, spanning the ' ί 2 f gap spacing of each of the five or gaps of the continuous center line For the spacing of the gaps, six pixels are pulverized by 0.2 μπΓ=^. More preferably, according to the seventh aspect of the invention, the array has a plurality of pixels, and each pixel is in the insulating layer. Under a plurality of wires:; = buried in the insulating layer and in the complex = photoelectric conversion unit, wherein the light guide has a span of five 丄 = pixels in the width of the increase and subtraction (four) _ s =: nr Among the pixels, == in the image is configured to detect another pixel of longer wavelength light. Also red = period 该 The width of the light guide is for - the blue pixel is smaller than for a quart pixel. It is further desirable that the width of the light guide is directed to a blue image for a green pixel. Further, the step is expected to have the width of the light guide ^, and the color pixel is smaller than for a red pixel. According to an eighth aspect of the present invention, there is provided a method for using an image-sensing image, the method comprising: (a) a plurality of layers under an insulating layer and embedded in the insulating layer a plurality of light 201143049 elements are provided under the wires; and (8) provided in the insulating layer and in the plurality of light guides to illuminate the light transmissive domain photoelectric conversion units, = side by side in the direction and among the plurality of pixels The light guides of each pair of consecutive pixels of 5 hi or more are horizontally spanned across the five or five like pixels. In the eighth aspect, it is desirable that the image sensor has From any of the desired features of the face. According to a ninth aspect of the present invention, an image sensor includes an array of 'the pixel array including a plurality of pixels, wherein each pixel includes a, a photoelectric conversion unit, and the photoelectric conversion unit is under the insulating layer and buried at = Placed under a plurality of wires in the insulating layer; and a light guide is embedded in the insulating layer and arranged between the plurality of wires to be aligned side by side in the direction of the photoelectric conversion unit Each of the five or more pixels among the plurality of pixels of the plurality of pixels = the horizontal 卩 of the light guides of the pixels has a wider level _ across the five or five elements, The wide level _ follows a relatively long interval, and the wider horizontal interval itself follows the other narrower spacing. The light guide can comprise a secret or colored pigment. The color pigment may be an organic pigment or an inorganic pigment, or an organometallic pigment. [Embodiment] An image sensor is disclosed. The image sensor includes a pixel array including a plurality of pixels each including a photoelectric conversion unit, such as 201143049. The light guide includes a light guide. It is buried in an insulating layer and is also embedded between the plurality of wires in the layer of the Wei margin so that the $ is transmitted to the county. The light guide of the pixel of the age-length light can have a width ("bottom width") at the bottom that is smaller than the light guide of another pixel that only detects the light of the vehicle and the long wavelength. The light guide—the vertical center at the bottom—the lateral side—the distance from the straight centerline of one of the light guides of the adjacent pixels—can be greater than the distance—the light guide of the adjacent pixel on the opposite lateral side—the vertical -turn. The spacing between the bottom of the light guide and the bottom of the light guide of a neighboring pixel on a lateral side ("bottom spacing") may be greater than the light guide of the bottom of the light guide and the adjacent pixel on the opposite lateral side - _ at the bottom - interval. The pixel may comprise a color; the light-sensitive sheet' color filter comprises a color material. There may be a gap between the color data sheets of each of the pair of pixels in the plurality of pixels. The width of the gap ("gap width") may be different between the - side-by-side pixel and the other-side side-by-side pixel. The gap between the gaps ("gap spacing") of three or more pixels arranged side by side in the - direction may vary by 15% or less. In the image sensing H-mask and in detail - the color image sensor has one or more of the above technical features permitting integrated circuit features under the insulating layer and within the insulating layer (eg 'gate electrode, poly Higher placement density of contacts (PGly (7) ntaet), electricity, and diffusion contacts). A substrate support photoelectric conversion unit 兀 and may be of a light-to-first conductivity type (preferably P-type) and preferably has a semiconductor substrate having a doping concentration between 5el4W and 5el5/em3. Substrate 106 can be a P-epi layer on a heavily doped P substrate having a doping concentration in excess of le9/em3. For example, the substrate 1〇2 12 201143049 may have a cut-grain pattern of 5el4/cm3 and 5el5/c3 such as a conventional layer on a P+ substrate, more specifically, see referencer H), The image sensor 1Q includes the pixel 14 = - = = r22 - the group is connected to the column green light reading device ί to the output money 18 (four) sample is subjected to subtraction and release ° = 8 = tADC24 will (these The analog signal is converted into ΑΙΧ: ^ 汇 = digital image data on row 6. If the image sensing _ 〇 is a color image sensor, the pixel array 12 includes a color filter, a light film array, and the image is included such that each pixel 14 has a color light film ^ Dimensional arrangement of multiple color filters. Figure 8A illustrates an example of a color filter, array of light filters that can be placed on top of pixel array 12 and as part of pixel array 12. 8A shows a Bayer primordial pattern (Bayer-cho (3) heart pattern)'. The pattern includes a repeating two-dimensional array of 2 χ 2 blocks (within a dotted line) of color filters, each of which has a green (G), One of red (R) and blue (B). - Place the green filter diagonally along one of the 2χ2 blocks. One pair of a red filter and a blue filter are placed along the other diagonal. In this embodiment of the color filter array, the color filters are arranged side by side on the page from left to right and on the page from top to bottom. 201143049 D8B illustrates an alternative embodiment of the color enamel array shown in Figure 8A. In this variation, the direction of the color illuminator is rotated by 45 degrees with respect to the left-to-right and top-to-bottom directions, and from the bottom-up direction (shown by the upward pointing arrow) relative to the image scanning. The pixels 14 are similarly arranged in an embodiment of the image sensor 1 using the color filter array. FIG. 9 shows a schematic diagram of one embodiment of a pixel 14 of a pixel array 12. The pixel 14 includes a photoelectric conversion unit 1〇2. For example, the photoelectric conversion unit 102 can be a photodiode. The photoelectric conversion unit 1〇2 is connectable to a reset switch 112 via a transfer gate 117. The photoelectric conversion unit 1〇2 can also be coupled to a selection switch 114 via an output (i.e., source follower) transistor 116. The transistors 112, 114, 116, and 117 can be used in the field. The gate of the transfer switch m is connected to the gate of the heavy 5-switch 112 to the -RST(n) line 118. The drain node of the reset switch 112 can be connected to a take-up line 12A. The gate of select switch 114 is coupled to a SEL line 122. The source node of select switch 114 is coupled to the -ττ line 124. RST (8) line 118, SEL (8) line 122, and TF (8) line 126 may be shared by one of the columns of pixels in pixel array 12. Similarly, the 1]^ line 12〇 and the OUT line 124 can be used in common for one row of pixels in the pixel array 12. RST (8) line 118, SEL (8) line 122, and TF (8) line 121 are coupled to column decoder 20 and are part of control line 22. The 〇 UT (m) line 124 is connected to the optical reader 16 and is part of the vertical signal line 18. 1A shows an embodiment of four adjacent pixels 14 in a pixel array 12 of color image sensor 1() on substrate 1〇6, which are arranged side by side in one direction. Each of the pixels 14 shown includes a photoelectric conversion unit 102a or 102b that converts photon energy into electric charges. Regarding the pixel array 12 using a 4T pixel architecture (such as shown in FIG. 9) or a variant thereof (such as sharing the reset switch 112, the selection switch 114, and the output transistor among the plurality of pairs of photodiodes 102 and the transfer switch 117) 116), the gate electrodes i〇4a, 1〇4b may each be a gate electrode of a different transfer switch 117 for transferring charges. Alternatively, with respect to the pixel array 12 using a 3T pixel architecture (such as that shown in FIG. 1A), the gate electrodes 104a, i〇4b may be different reset switches 112 for resetting the photoelectric conversion units 102a, 102b, respectively. The gate electrode. The gate electrode 104c can be a gate electrode of a transistor (e.g., the reset switch 112, or the select switch 114, or the output transistor 116) that serves different functions in the pixel array 12. The gate electrodes 1 〇 4a, 1 〇 4b, 104c and the conversion unit 102a 102b are formed on the substrate 1〇6 or in the substrate 1〇6. The gate electrodes 104a, lG4b, lG4e, the photoelectric conversion unit 1Q2a, the enamel, and the substrate 106 may be covered on the protective layer 230 τ, and the protective layer 23 气 includes gas fossils and has between 200 angstroms and 1 angstrom thickness. The protective layer 23 is used to insulate the substrate 1 (10) from metal ions and moisture. The layer of insulating layer 11 覆盖 covers the substrate 1〇6. A plurality of wires 108 are buried in the insulating layer m and above the gate electrodes, 104b, 104C. The wire (10) is a conductive interconnect wire including Ming or copper. Other interconnecting wires (not shown) may be formed in other wiring planes, the mother=plane includes a plurality of interconnecting wires (the poly-serial wire may be metal) and the electrode is difficult, na, 1 (10) is above the wire Just different from the south. The wires 108 can be connected via conductive (such as metal) vias, which are interconnected across adjacent wiring planes at different heights. The 'optoelectronic conversion units 102a, 102b can be paired with the lower light guides 316a, 316b, respectively, and the lower light guides 316a, 316b are buried in the insulating layer 11A and between the plurality of wires. The lower light guide, her, may include a transmissive material having a higher refractive index than the insulating layer 110 (which is an external material such as yttrium oxide) (for example, tantalum nitride which is like S^N4), and the light guide and the external material are used. Total internal reflection to help prevent light from leaking out of the lower light guide. Alternatively, the lower light guide may be filled with a transparent material such as spin-on glass (S0G) or an organic material, and may even include a color material such as an organic or inorganic pigment or an organometallic pigment, and have a reflective metal coating on its side wall. The metal coating type) reflects light inwardly to help prevent light from leaking out of the lower light guide. The upper light guide 130 can be located on the lower light guide 31, such as above, and can include the same material(s) as the lower light guides 316a, 316b. The upper light guide 130 and the lower light guides 316a, 316b may all be of the type of reflective metal coating, or one of them may be of the type. The top end of the upper light guide 130 is wider than the bottom end, and at the bottom k, the pilot 13G is combined with the lower material.
彩色遽光片ma、U4b位於上光導⑽上 光片114a、114b可名k H 脂’其;該樹二有===== 201143049 或苯基之至少一有機基團之聚合物（一實例為聚矽氧）。或 者，彩色濾光片可包括透射無機材料（例如，氮化矽），透 ^無機材料中分散有彩色顏料（例如，諸如氧化鐵之無機 衫色顏料、鈷或錳或鋅或銅顏料，或有機金屬顏料，或複 合無機彩色顏料）之微粒。彩色濾光片114a、114b在白光 中展現不同顏色。較佳地，每一者在4〇〇11111至7〇〇nm之 波長（空氣中）之間分別具有為至少5〇%之最高透射率以 及至多10%之最低透射率。或者，最高透射率與最低透射 率之間的比率應大於4比1。 如圖1A中所展示，鄰近彩色濾光片U4a、U4b之間 具有多個間隙。彩色濾光片114a、114b之間的間隙422a、 422b之寬度為〇·45 或更小。間隙422a、422b可填充 有空氣或氣體。間隙422a、422b在鄰近彩色濾光片之間的 深度可為0.6 μηι或更大。具有上文所提出之尺寸限制之間 隙使間隙内的光轉向至鄰近彩色濾光片中且隨後藉由（多 個）光導導引至光電轉換單元102a或1〇2b。因此，歸因 於光穿過間隙422a、422b且穿透間隙422a、422b下方所 引起的照射於像素上之光之百分比損失（下文中「像素損 失」）實質上減小。 、 彩色濾光片114a(或114b)以及上光導π〇與下光導 316a (或316b)共同構成一「級聯式光導」，該級聯式光 導經由在與諸如絕緣體110之外部介質之界面處的全内反 射而將光導引至光電轉換單元l〇2a(或l〇2b)。（或者，上 光導以及下光導中之一者或兩者可具有用以向内反射光之 17 201143049 金屬側壁。）在圖1B之射線追蹤圖中，展示射線0、办、e 以及/經歷彩色濾光片及/或上光導以及下光導之侧壁處的 反射。落在第二像素以及第三像素之彩色濾光片之間的較 寬間隙中之射線C以及d分別轉向至第二像素以及第三像 素之彩色濾光片中，且到達各別光電轉換單元。 圖3展示影像感測器1〇之替代實施例，其中鄰近彩 色濾光片之間的間隙422a、422b覆蓋在透明薄膜500下， 且支樓薄膜134填充在鄰近上光導130之間。若上光導130 為全内反射類型，則支撐薄膜134應具有低於上光導130 之折射率。間隙422a、422b之頂板（ceiling) 510相對於 透明薄膜500可為凹入的（亦即，相對於間隙為凸起的）， 諸如呈圓頂形狀。間隙422a、422b可包含空氣或氣體。從 上方進入間隙，從而橫跨凸頂板之光朝向鄰近彩色濾光片 轉向。 圖4展示一替代實施例，其中圖3中之實施例被進一 步修改’使得彩色濾光片114a、114b具有向内傾斜之侧壁 且支樓薄膜134具有與彩色渡光片之側壁的界面i如在第 二實施例中，跨從左至右之四個像素，連續彩色濾光片之 間的間隙變得更寬，接著更窄，接著再次更窄。=管在間 隙内之不同高度處間隙具有不同寬度，但針對較寬間隙 422b’與較窄間隙422a’之間的比較，量測在割穿該等彩色 濾光片以及彩色濾光片之間的間隙之水平位準即广平 行於光電轉換單元之平面的水平位準）處之間隙寬度為足 夠的，如圖4中所展示。同樣地，針對一對連續間間 201143049 的一較寬間距與另一對連續間隙之間的一較窄間距之間的 比較，在相同水平位準處量測為足夠的。圖7為射線追蹤 圖，其展示進入兩個彩色渡光片114a、114b之間的間隙的 光射線之軌道。頂板之凸度用以使光射線朝向該等彩色濾 光片中之一者轉向。從彩色濾光片之底部至鄰近頂板之頂 部的高度應為0.6 μιη或更大。此提供用於使從該等頂板進 入間隙（頂板下方且橫向地鄰近於彩色濾光片）之光轉向 至鄰近彩色濾光片中的足夠深度。舉例而言，將從第一（從 左邊數）彩色濾光片114a之底部至第一彩色濾光片114a 與第二彩色濾光片114b之間的頂板5l〇a之頂部量測的高 度標註為Ha。同樣地，將從第二彩色濾光片U4b之底部 至第二彩色濾光片114b與第三彩色濾光片114&之間的頂 板510b之頂部量測的高度標註為Hb。 圖5展示一替代實施例，其中省去上光導13〇，且替 代地微透鏡318置放於下光導316a、316b上方，而在微透 鏡318與光導316a、316b之間有透明平坦化層320。微透 鏡318使光聚焦至光導316a、316b之上孔隙中，光導 316a、316b之上孔隙又使光向下透射至其各別光電轉換單 元102a、102b。關於彩色影像感測器，光導316a、316b 可包括著色劑’諸如染料’或有機或無機或有機金屬顏料， 以根據像素陣列12之著色圖案（例如，拜耳圖案）之多個 顏色向光導316a、316b給予不同顏色。 或者，圖1A、圖3、圖4以及圖5中所展示之實施例 中之間隙422a、422b可包含透明（液體或固體）介質，只 201143049 射率比該等彩色遽光片之折射率低至少 ’該透明介質可為折射率在丨.4與15之間 =二Γ等？色渡光片包括氮化石夕之微粒，氮化矽之 雜的讀經難以使得其折射輕成17或以上。 埋置於絕緣體層110中且在多條電後1〇8之間以僅透 2短波長之光的光導可在其底部處具有小於用以透射較 之光的另-光導的寬度（「底部寬度」），而不管該光 吏用㈣反射㈣止絲_或在其㈣上使用金屬 J二參看圖1A，彩色遽光片U4a可為藍色遽光片該 化色滤光片114a之透射率針對在空氣中在铜細至 5〇〇 nm之_波長高於針對其他波長（因此彩色滤光片 IHa之像素為藍色像素），且彩色濾光片⑽可為綠色滤 光片，該等彩色遽光片114b之透射率針對在空氣中在5〇〇 nm至600 nm之間的波長好於針對其他波長（因此彩色濾 光片114b之像素為綠色像素）。因此，透射藍光之下光導 316a (針對藍色像素）的底部318a之底部寬度小於僅 透射綠光之下光導316b (針對綠色像素）的底部31沾之 底部寬度wb (wa<wb)。-般而言，下光導316a窄於下 光導316b。或者，圖1A中所展示之彩色遽光片⑽可為 紅色濾光片，該等彩色濾光片114b之透射率針對在空氣中 在600 nm與700 rnn之間的波長高於針對其他波長（因此 彩色遽光片114b之像素為紅色像素）。或者，彩色滤光片 114a可為綠色濾光片，而彩色濾光片1141)為紅色濾光片。 圖1A亦展示底部寬度跨在一方向上並排地排列之多 20 201143049 個=素非單調地變化。在圖1A卜從左至右， =為wa、wb、Wa、Wb，mM_: 化序列·[增大、減小、择去\ , 9 }。在一岸代實施例中，底部寬 冋變化序列為可能的，同時為非單_，亦即，底 i^增大，繼之以減小，且進—步繼之以另—增大；… ^減小，繼之以增大，且進-步繼之以另-減小。詳言之， ==可為重複的非單調序列。舉例而言，底部寬度之 μ匕可遵循重複的型樣{增大、無改變、減小、增大、減小}。 在圖1Α中，彩色濾光片U4a可為藍色濾光片，且下 ^ 316a在其底部處具有底部寬度Wa。進一步至圖ία 中之右邊’彩色濾抑114b可為綠色滤光片，且下光導 ^在其底部處具有底部寬度Wb，其中底部寬度Wb大於 底4寬度wa(wa< wb)。關於埋置於絕緣層11G中用以透 射波長等於5GG nm之光的下光導，其底部寬度較佳地在 〇.2 μιη至0.35 μιη之間’更佳地為〇 27 μπι+/_ 1〇%。關於 用以透射波長等於600 nm之光的下光導，其底部寬度應 較，地在0.25 至0 4 μΓη之間，更佳地在〇 33 μιη +/_ 1〇/〇内。關於用以透射波長等於700 nm之光的下光導，其 底。卩寬度應較佳地在0.3 μιη至〇·5 之間，更佳地在0.4 μΐΠ +Λ 10%内。在底部處具有較小寬度准許絕緣體110下 之積體電路特徵（諸如，閘極電極104a、104b以及104c、 多晶石夕觸點以及擴散觸點）之較高裝填密度。 該等間隙422a、422b可使在並排像素之彩色濾光片 之間的間隙之寬度（「間隙寬度」）在該等間隙自身當中非 21 201143049 單調地變化。跨第—像素、第二像素以及第三像素，該等 像素在一方向上以此次序並排地排列，第一像素與第二像 素之間的間隙可寬於第二像素與第三像素之間的間隙。舉 例而a，圖1A展示在一特定方向上並排地排列的四個鄰 近像素之間的兩個不同間隙寬度％、Vb。 連續間隙422a、422b之中心線之間的距離p (「間隙 間距」）跨在一方向上並排地排列的五個或五個以上像素可 維持基本该定（亦即，在其最大值之5%内），而間隙窗 度非單調地變化。 间丨糸間距户可被允許非單調變化，該非單調變 跨在一方向上並排地排列的預定數目個像素具有其最大 =之至多2G%的最缝對最小值差。該預定數目可小於 ，更佳地不大於8。更特定言之，間隙間距尸跨在—方 ° J^排列的多個像素可交替地增大以及減小。針對1帅 像素間距，_間距P可變化差不多U卿，更佳 下=。使_間距p結合底部間隔（或更—般而言連 地置水平間隔）非單繼變化給打光導更佳 以此方p導自身的更多自由，同時保持良好的光彌取。 同時伟f ’下光導可移位至一_使此側上之間隔較窄， 觸對側上之間隔足夠寬以容納諸如閘極電極或擴散 :點之額外積體電路特徵’由此幫助支撐一更緻密像= 素在並排地排列的五個或五個以上像 直板截面中連續下光導之垂直中心線之間的距離 22 201143049 (「底部間距」）可具有非單調變化。舉例 左邊的三個下光導316a、316b之垂直中心展： 展”垂直虛線)之間的兩個不同的底部間距(u: 大底部間距xb在間隙422b之較大間隙 a b 乂 小底部間距xa衫—關422a讀小下方’ ^較 與較小底部間距Xa相比，較大底部間距又、下方。 110下容納更多積體電路特徵。 b幫助在絕緣層 如圖1A展示，第一像素（從左邊 間的底部_ Sa與第三像素與第四像門、=像素之 =同，且小於第二像素與第三像素之間===隔 苐一像素與第二像素之間的較大底部間隔s_ b 或基底10ό上方且鄰沂於其讫在絕緣體下 徵’此等特徵包括閘極電二諸如 間隔之習知 素盥當m 寬度#於~，即使在第一像 寬^隔以+之間u及第三像素與第四像素之間無需此較 尤間μ如此，從而導致不及最佳像素密度之問題。 在傻為從上方向下看像素_12之俯視®1，其展示 ’、車列12之第一、第二以及第三實施例之三個列以及 從:行中的十五個像素14，其中多個像素在從左至右以及 至下方向上並排地配置。間隙之中心線以每一對並排 ，之間的粗灰線繪製。區域B表示上光導130之頂部表 區域且區域C表示下光導316a、316b之底部表面區 23 201143049 域:區域A減區域b為彩色濾光片之間的間隙422a、422b 之區域’其中A表示像素區域。圖2A展示區域B與C之 四個不同對41與〇1、32與〇2、^與〇：3 ，以及B4與C4。 此等四個不同對以一規則圖案重複。 可將圖1A看作圖2A中之中間列的第一像素至第四 像素（從左邊數）之垂直剖面。在此上下文中，％ = t3、 vb = t4、Sa = s3 且 sb = S4。 如圖2A展示’並排像素之彩色濾光片之間的間隙具 有重複的寬度ti、t2、t3、t4、u〗、U2、U3以及U4。沿著圖 2A中之任何從左至右或從上至下切割，間隙寬度&至t4 或u〗至叫在其自身當中變化，而從第一間隙之中心線至 下一第一間隙之中心線的間距P保持恆定。間隙寬度跨在 垂直於間隙之方向上並排地排列之複數個像素而變化，該 方向在圖2A中針對在像素之左邊或右邊（如在圖2A中向 下看像素陣列之俯視圖中所展示）之間隙係從左至右（或 從右至左），或針對在像素之上方或下方（再次地，如在圖 2A中向下看像素陣列之俯視圖中所展示）之間隙係從上至 下（或從下至上）。該變化為非單調的，亦即，增大，繼之 以減小’又繼之以另一增大；及/或減小，繼之以增大，進 一步繼之以另一減小。舉例而言，在頂部列中，從左至右， 間隙寬度以序列t2、tl、t2、ti變化，其中t2 < tl，其展現 型樣{增大、減小、增大}之變化序列；在中間列中，間隙 寬度以序列13、1443、丨4變化，其中13<〖4，其展現型樣{增 大、減小、增大}之變化序列；且在底部列中，序列與頂部 24 201143049 歹J之序列相同。更特今 的並排像素在較寬間“度二巧^在-方向上排列 管在間隙寬度之此等兩個序=:„之間交替。儘 度之變化在增大與減小 二者:，展示間隙寬 詳言之，變化序列可非單調的。 隙寬度之變化可翁舌、“早°周序列。舉例而言，間 大二 重複的型樣{增大、無改變、減小、增 同樣地，圖2A亦展示底部間隔 2列之多個像素而變化。舉例而言，在中間列= f右’底部間隔之序列在…之間交替， 处而展現變化序列{增大、減小、 4 =隔跨在—方向上並獅的像 曰^之間交替。使用光導之習知影像感測器像素陣列以彼 均-地移位之方式置放光導，由此不能夠利用光導之間 、過小以致不能容納諸如閘極電極之積體電路特徵的小允 間。圖2A中所展示之實施例能夠使第二列之左邊的兩二 光導彼此更靠近，且亦使第二列之下兩個光導彼此更靠 近，以便在中間（亦即，第二光導與第三光導之間）產生 較寬空間，使得額外閘極電極l〇4c可裝配於其中。即使僅 在横向地鄰近之光導之某些對之間需要較寬間隔，在光導 <間具有均一間隔之習知像素陣列仍會必須使所有間隔車交 寬。在一替代實施例中，底部間隔之任何變化序列為可At 的，只要變化序列為非單調的，亦即，底部間隔増大，^ 25 201143049 =以減小、，且進—步繼之以另—增大；及/或減小，繼 坫大，且進一步繼之以另一減士、 重複的非單調序列。舉之文化序列可為 複的型樣ί增大，二部間=變化可遵猶重 9大減小、增大、減小}。較佳地， 部間隔在五個連續像素之群組内變化Ο.—或更大。- w 亦f示底部寬度在第二列中以底部寬度序列 3 b冒3'~1?、^^從左至右非單調地變化，其中~ Wb ’從而展現變化型樣{增大、減小、增大、減小}，发a 非單調的。詳言之’關於在頂部兩則以及最左邊之= 行中的四個像素，展示底部區域〇3之底部寬度最小， 區域Ci之底部寬度最A，且底部區域c2以及q之底部寬 度為中間的。在原色拜耳圖案用於像素陣列12之上之彩 遽光片的實施例t，具有底部區域C2以及C4之像封 綠色像素’具有底輕域C1讀封為紅色像素，且呈有 底部區域Q之像素可缝色像素。在交㈣素在兩種不同 顏色之間交替的方向上’下光導之寬度應類似地交替，使 ^下光導以及該等下光導之間隔具有用於導引光被偵測同 時亦最佳化積體電路特徵之空間的最佳大小。因此，在底 部區域〇：3對胁龍像素且底部_ (〕4職於綠色像素 的此實例中，在第二列中’在交替像素在從左至右方向上 在藍色像素與綠色像素之間交替之情況下，光導之寬度相 應地最佳化’從而在較窄寬度（針對藍色像素）與較k 度（針對綠色像素）之間交替’從而節約用於搁置諸如閘 極電極以及觸點之積體電路特徵的光導之間的空間。同樣 26 201143049 地，在第-列中，在交替像素在綠色像 之 與較寬寬度(針對紅色針對綠色像素) 更-般而言’出於比較埋置於絕緣層u 寬度的目的以及比較此等光導# +的連續光 間的水平間隔的目的，可在閑極電極刚a、i〇4b=十 準(亦即，平行於光電轉換單= 面的水平位準）處量測寬度以及水平間1在此高度範圍 ^水平間隔（以及因此之寬度）為有重大意義的，在於 其影，絕緣層110巾及/或下之積财路舰可裝填在一起 的緻密程度，此等積魏路特徵包括難電極、多晶石夕佈 線5線（未圖示）、多晶矽觸點（未圖示）、擴散觸點（未 圖不）以及金屬電線。此水平間隔應跨在—方向上並排地 排列之五個或五個以上像素非單調地變化，如圖1Α中所 展示’以准3f-光導移位至_側以佔據過小以致不能容納 了積，電路特徵之小空間，使得更多空間聚集於另一側上 以變得足夠大從而容納一積體電路特徵。較佳地，此變化 跨在一方向上並排地排列之八個或少於八個之像素展現自 身。進一步較佳地，此水平間隔跨八個或少於八個之像素 (更佳地在五個連續像素之群組内）在多個連續光導之間展 =最寬水平間隔與最窄水平間隔之間的〇1 μιη或更大之 變化。更進一步較佳地，該變化為0.2 μιη或更大。使光導 之此寬度儘可能小（而不妨礙光之透射）有助於使水平間 隔加寬’由此有益於絕緣層110中以及下之積體電路特徵 27 201143049 的更緻密裝填以及像素陣列12中之像素的更緻密裝填。較 佳地，如圖1A中所展示，此寬度跨在一方向上並排地排 列之五個或五個以上像素非單調地變化以利用如下事實’ 即此寬度針對僅透射較短波長之光的光導可小於針對透射 較長波長之光的光導。關於埋置於絕緣層110中用以透射 在空氣中之波長等於500 nm之光的光導，此寬度較佳地 在 0.2 μιη 至 0.35 μιη 之間’更佳地為 0.27 μιη +/- 10%。關 於用以透射在空氣中之波長等於6〇〇 nm之光的光導，此 寬度應較佳地在0.25 μιη至0.4 μιη之間，更佳地在0.33 μιη +/_ ίο%内。關於用以透射在空氣中之波長等於7〇()11111之 光的光導，此寬度應較佳地在〇 3 μιη至〇 5 μιη之間，更 佳地在 0.4 μιη +/- 10% 内。 非單調地變化在一電線108與閘極電極1〇4a之間的 水平位準（亦即，平行於光電轉換單元之平面&水平位準） 處連續光導之間的水平間隔（或詳言之底部間隔）及/或一 電線108與閘極電極lG4a (包括在内）之間的高度處光導 之間的水平間距（或詳言之底部間距）准許—光導移位至 一侧以佔據過小以致不能容納一積體電路特徵之小空間， 使得更多空間輕於H以變得足夠大從而容納一積 體電路特徵。此在圖1A中可見，在圖u作若所有下光 導相等地間隔麟可能的情況相比，在第二:ρ光導與第三 下光導（從左邊數）之_在多條電線1G8與閘極 馳、_、胸（包括在内）之間的任何水平位準（亦 即，平行於光㈣換單元之平_水平㈣）處連續光導 28 201143049 之間的增大之水平間隔（或詳言之底部間隔sb)容納一額 外閘極電極104c。 非單調地變化多個彩色濾光片之間的間隙寬度及/或 間隙間距准許該等彩色濾光片以如下方式與各別下1導一 起移位··使得各別下光導保持耦接以接收來自各別彩色滤 光片之光。 在間隙填充有空氣或（多種）氣體，或折射率低於彩 色濾光片之液體或固體介質的情況下，或在凸頂板以在頂 板上方之折射率高於頂板下之折射率的狀況處於間隙之上 的情況下，甚至在間隙寬度改變時，光仍能夠從間隙轉向 至彩色濾光片中。因此’在連續的第一、第二、第三以及 第四彩色濾光片之序列中，第二彩色濾光片可移位成較靠 近弟一彩色;慮光片，且苐二衫色遽光片可移位成較靠近第 四彩色濾光片，而所有四者在收集光方面仍為有效的。如 圖1B中所展示，落在第二像素與第三像素之彩色濾光片 之間的較寬間隙中之射線c以及射線d仍由各別彩色濾光 片擷取且透射至各別光電轉換單元。在彩色濾光片之間具 有間隙且能夠使一些間隙之間隙寬度減小同時使變得較窄 之彼等間隙之間的間隙加寬准許該等彩色濾光片下方之下 光導在較窄間隙下方具有類似地減小的間隔且在加寬之間 隙下方具有加寬的間隔，同時保持所涉及之像素之良好的 光擷取。 圖2B展示圖2A中之十五個像素之相同俯視圖。可將 圖1A看作沿著圖2B中之切割線zz，的從左邊數之第一像 29 201143049 素至第四像素之垂直剖面。底部區域内且切割線22 粗「+」標記表示在垂直於基底1〇6且包含切 ^ 面中此等像素之下光導的垂直令心線（在圖1A = 虛線）。將沿著_線22，之_像素之_底部間距=為 Xb展示為非單調地改變。詳言之，沿著切割線z 3、 :}距之序列{Xa、xb、Xa、Xb}展現型樣{增大減小:增 上文之描述已展示積體電路特徵之裝填密度可 在一方向上排列之像素之連續光導之間的水平間隔^古 ^，底部間隔）之非單職化增強的方式。水平間隔 吕之，底部間隔）之非單調變化由以下各者中之任何一 ^者之非單調變化幫助：在一方向上排列之像素之連續光 導之間的水平間距（詳言之’底部間距）、間隙寬度以及間 隙間距。舉例而言’可允許間關距變化高達。較佳地， 變化之非單調性發生於在一方向上並排地排列之小數目個 像素之範圍内，例如32個像素，或例如16個像素，或更 佳地在8個像素内。在此範圍内，該（該等）變化展現增 大，繼之以減小，進一步繼之以另一增大；或減小，繼之 以增大’繼之以另一減小。 上文之描述亦展示裝填密度可藉由關於光導所透射 之不同顏色的光最佳化埋置於絕緣層中之光導之寬度來增 強以便佔用可能之最小空間的方式。僅透射較短波長之光 的光導應窄於透射較長波長之光的光導。 儘管已在附圖中描述並展示某些例示性實施例，但應 30 201143049 理解，此等實施例僅說明而非限制廣義之本發明，且本發 明不限於所展示並描述之特定構造以及配置，此係因為一 般熟習此項技術者可想出各種其他修改。 舉例而言，像素陣列中之像素可在與從左至右以及從 上至下方向（平行於光電轉換單元之平面）成45度之方向 上並排地配置。 舉例而言，上/下光導之頂部/底部開口可採取與矩形 不同之形狀，諸如八邊形或具有圓形轉角之矩形。 【圖式簡單說明】 圖1A為一說明，其展示本發明之一實施例之四個影 像感測器像素的橫截面； ^ 圖1B為射線追蹤圖，其用於圖1A之四個影像感測器 像素的相同橫截面； 圖2A為一說明，其展示一陣列内之十五個像素之俯 > 圖2B展示十五個像素之相同俯視圖； 器像=截:明，其展示一替代實施例之四個影像感測 模式的一替代實施例之 圖4為一說明，其展示為最佳 四個影像感測器像素之橫截面； 替代實施例之四個影像感測 圖5為一說明，其展示一 器像素之橫截面； 圖6為一影像感测器之示意圖 31 201143049 圖7為射線追蹤圖，其用於圖4中所展示之第三實施 例。 圖 8A 為一原色拜耳圖案（primary color Bayer pattern ) 之說明； 圖8B為旋轉45度之原色拜耳圖案之說明； 圖9為4T像素之示意圖； 圖10為3T像素之示意圖。 【主要元件符號說明】 10 :影像感測器 12 :像素陣列 14 :像素 16 :光讀取器電路/光讀取器 18 :輸出信號/垂直信號線 20 :列解碼器 22 :控制信號/控制線 24 :類比至數位轉換器（ADC) 66 : ADC輸出匯流排 102 :光電轉換單元/光電二極體 102a :光電轉換單元 102b :光電轉換單元 104a :閘極電極 104b :閘極電極 104c :閘極電極 32 201143049 106 :基底 108 :電線 110 :絕緣層/絕緣體 112 :重設開關/電晶體 114 :選擇開關/電晶體 114a :彩色濾光片 114b :彩色濾光片 116 :輸出電晶體 117 :傳送閘/電晶體/傳送開關 118 : RST(n)線 120 : IN 線 121 : TF(n)線 122 : SEL 線/SEL(n)線 124 : OUT 線/OUT(m)線 126 : TF(n)線 130 :上光導 134 :支撐薄膜 230 :保護層 316a :下光導 316b :下光導 318 :微透鏡 318a :下光導的底部 318b :下光導的底部 320 :透明平坦化層 33 201143049 422a :間隙 422a':較窄間隙 422b :間隙 422b':較寬間隙 500 :透明薄膜 510 :頂板 510a :頂板 510b :頂板 A : 區域 B ： 區域 B! 區域 b2 區域 b3 區域 b4 區域 c ： 區域 Cl 區域 C2 區域 C3 區域 C4 區域 34The color grading sheets ma, U4b are located on the upper light guide (10). The light sheets 114a, 114b may be named as H H ''; the tree 2 has ===== 201143049 or a polymer of at least one organic group of the phenyl group (an example) For polyoxyl). Alternatively, the color filter may include a transmissive inorganic material (for example, tantalum nitride) in which a color pigment (for example, an inorganic shirt pigment such as iron oxide, cobalt or manganese or zinc or copper pigment, or Particles of organometallic pigments, or composite inorganic color pigments. The color filters 114a, 114b exhibit different colors in white light. Preferably, each has a maximum transmission of at least 5% by weight and a minimum transmission of at most 10% between wavelengths of 4 〇〇 11111 to 7 〇〇 nm (in air). Alternatively, the ratio between the highest transmission and the lowest transmission should be greater than 4 to 1. As shown in Fig. 1A, there are a plurality of gaps between adjacent color filters U4a, U4b. The widths of the gaps 422a, 422b between the color filters 114a, 114b are 〇·45 or less. The gaps 422a, 422b may be filled with air or gas. The depth of the gaps 422a, 422b between adjacent color filters may be 0.6 μηι or more. The size limiting gap proposed above causes the light within the gap to be diverted into adjacent color filters and then directed to the photoelectric conversion unit 102a or 1〇2b by the (s) light guides. Therefore, the percentage loss of light (hereinafter "pixel loss") due to light passing through the gaps 422a, 422b and penetrating the gaps below the gaps 422a, 422b (hereinafter "pixel loss") is substantially reduced. The color filter 114a (or 114b) and the upper light guide π 〇 and the lower light guide 316a (or 316b) together form a "cascade light guide" that passes through an interface with an external medium such as the insulator 110. The total internal reflection reflects light to the photoelectric conversion unit 10a 2a (or l〇2b). (Alternatively, one or both of the upper and lower light guides may have 17 201143049 metal sidewalls for internally reflecting light.) In the ray trace diagram of Figure 1B, the ray 0, office, e, and/or color are displayed. Reflection of the filter and/or the upper light guide and the sidewalls of the lower light guide. The rays C and d falling in the wider gap between the color filters of the second pixel and the third pixel are respectively turned into the color filters of the second pixel and the third pixel, and reach the respective photoelectric conversion units . 3 shows an alternate embodiment of an image sensor 1 in which gaps 422a, 422b between adjacent color filters are covered under transparent film 500 and a via film 134 is filled between adjacent upper light guides 130. If the upper light guide 130 is of the total internal reflection type, the support film 134 should have a lower refractive index than the upper light guide 130. The ceiling 510 of the gaps 422a, 422b may be concave (i.e., convex relative to the gap) relative to the transparent film 500, such as in the shape of a dome. The gaps 422a, 422b may contain air or gas. The gap is entered from above so that the light across the convex top plate is turned toward the adjacent color filter. 4 shows an alternative embodiment in which the embodiment of FIG. 3 is further modified 'such that the color filters 114a, 114b have sidewalls that are inclined inwardly and the floor film 134 has an interface with the sidewalls of the color light guides. As in the second embodiment, across the four pixels from left to right, the gap between successive color filters becomes wider, then narrower, and then narrower again. = the tubes have different widths at different heights within the gap, but for comparison between the wider gap 422b' and the narrower gap 422a', the measurement is taken between the color filters and the color filters The gap width at the horizontal level of the gap, that is, the horizontal level parallel to the plane of the photoelectric conversion unit, is sufficient, as shown in FIG. Similarly, a comparison between a wider spacing between a pair of consecutive rooms 201143049 and a narrower spacing between another pair of consecutive gaps is sufficient at the same level. Figure 7 is a ray tracing diagram showing the trajectory of the light rays entering the gap between the two color fins 114a, 114b. The crown of the top plate is used to steer the light rays toward one of the color filters. The height from the bottom of the color filter to the top of the adjacent top plate should be 0.6 μm or more. This provides for sufficient depth to divert light from the top plates into the gap (below the top plate and laterally adjacent to the color filter) into adjacent color filters. For example, the height measured from the bottom of the first (from the left) color filter 114a to the top of the top plate 51a between the first color filter 114a and the second color filter 114b For Ha. Similarly, the height measured from the bottom of the second color filter U4b to the top of the top plate 510b between the second color filter 114b and the third color filter 114& is denoted as Hb. Figure 5 shows an alternative embodiment in which the upper light guide 13A is omitted and instead the microlens 318 is placed over the lower light guides 316a, 316b with a transparent planarization layer 320 between the microlens 318 and the light guides 316a, 316b. . The microlens 318 focuses the light into the apertures above the light guides 316a, 316b, which in turn transmit the light downwardly to its respective photoelectric conversion units 102a, 102b. With regard to color image sensors, the light guides 316a, 316b can include a colorant 'such as a dye' or an organic or inorganic or organometallic pigment to direct the light guide 316a according to a plurality of colors of the colored pattern of the pixel array 12 (eg, a Bayer pattern), 316b gives different colors. Alternatively, the gaps 422a, 422b in the embodiment shown in Figures 1A, 3, 4, and 5 may comprise a transparent (liquid or solid) medium, and only the 201143049 rate is lower than the refractive index of the color patches. At least 'the transparent medium can have a refractive index between 丨.4 and 15 = two Γ, etc.? The color light-receiving sheet includes the particles of the cerium nitride, and the miscellaneous reading of the tantalum nitride is difficult to make the refractive light of 17 or more. A light guide embedded in the insulator layer 110 and having a light transmission of only 2 short wavelengths between the plurality of electrical electrodes 1 〇 8 may have a width at the bottom thereof that is smaller than the other light guide for transmitting light ("bottom Width"), regardless of the diaphragm (4) reflection (four) wire _ or use of metal (J) on it (see Figure 1A), color enamel U4a can be blue ray film transmission of the color filter 114a The rate is such that the wavelength of copper in the air is as fine as 5 〇〇 nm is higher than that for other wavelengths (so the pixels of the color filter IHa are blue pixels), and the color filter (10) can be a green filter, The transmittance of the equal-color calender sheet 114b is better for wavelengths between 5 〇〇 nm and 600 nm in air than for other wavelengths (so the pixels of the color filter 114b are green pixels). Thus, the bottom width of the bottom 318a of the light guide 316a (for the blue pixel) transmitted under the blue light is smaller than the bottom width wb (wa<wb) of the bottom 31 of the light guide 316b (for the green pixel) only under the green light. In general, the lower light guide 316a is narrower than the lower light guide 316b. Alternatively, the color light-emitting sheet (10) shown in FIG. 1A may be a red color filter having a transmittance higher than that for other wavelengths in the air between 600 nm and 700 rnn ( Therefore, the pixels of the color calender sheet 114b are red pixels). Alternatively, the color filter 114a may be a green filter and the color filter 1141) may be a red filter. Figure 1A also shows that the bottom width is arranged side by side across one direction. 20 201143049 = non-monotonically varying. In Fig. 1A, from left to right, = is wa, wb, Wa, Wb, mM_: sequence · [increase, decrease, select \, 9 }. In the one-shore embodiment, the bottom width variation sequence is possible, and at the same time, it is not a single _, that is, the bottom i^ is increased, followed by the decrease, and the further step is followed by another increase; ... ^ decreases, followed by an increase, and the further step is followed by another - decrease. In particular, == can be a repeating non-monotonic sequence. For example, the μ宽度 of the bottom width can follow the repeated pattern {increase, no change, decrease, increase, decrease}. In Fig. 1A, the color filter U4a may be a blue filter, and the lower ^ 316a has a bottom width Wa at the bottom thereof. Further to the right of the image ία, the color filter 114b may be a green filter, and the lower light guide has a bottom width Wb at the bottom thereof, wherein the bottom width Wb is greater than the bottom 4 width wa (wa< wb). Regarding the lower light guide buried in the insulating layer 11G for transmitting light having a wavelength of 5 GG nm, the bottom width thereof is preferably between 〇.2 μιη to 0.35 μηη, more preferably 〇27 μπι+/_1〇 %. Regarding the lower light guide for transmitting light having a wavelength equal to 600 nm, the bottom width should be between 0.25 and 0 4 μΓη, more preferably 〇 33 μιη +/_ 1〇/〇. Regarding the lower light guide for transmitting light of a wavelength equal to 700 nm, the bottom. The width of the crucible should preferably be between 0.3 μηη and 〇·5, more preferably within 0.4 μΐΠ + Λ 10%. Having a smaller width at the bottom permits higher packing densities of integrated circuit features under insulator 110, such as gate electrodes 104a, 104b, and 104c, polycrystalline whisk contacts, and diffused contacts. The gaps 422a, 422b can vary the width of the gap between the color filters of the side-by-side pixels ("gap width") monotonically within the gaps themselves, 21 201143049. Across the first pixel, the second pixel, and the third pixel, the pixels are arranged side by side in this order in one direction, and a gap between the first pixel and the second pixel may be wider than between the second pixel and the third pixel gap. By way of example, a, Figure 1A shows two different gap widths %, Vb between four adjacent pixels arranged side by side in a particular direction. The distance p ("gap gap") between the centerlines of the continuous gaps 422a, 422b can be maintained substantially equal to five or more pixels arranged side by side in one direction (i.e., at 5% of its maximum value) Inside), while the gap window does not change monotonically. The inter-pitch spacing may be allowed to be non-monotonic, the non-monotonic variation of a predetermined number of pixels arranged side by side across a direction having a maximum seam-to-minimum difference of at most 2G%. The predetermined number may be less than, more preferably not more than 8. More specifically, the plurality of pixels arranged in the gap spacing can be alternately increased and decreased. For a handsome pixel pitch, the _pitch P can vary by almost U, better. Having the _pitch p combined with the bottom spacing (or more generally, the horizontal spacing) is not a single change to the light guide. This way, the p-guide itself is more free, while maintaining good light absorption. At the same time, the lower light guide can be shifted to a _ such that the spacing on the side is narrower, and the spacing on the opposite side is wide enough to accommodate additional integrated circuit features such as gate electrodes or diffusion: points to help support A more dense image = the distance 22 between the vertical centerlines of successive lower light guides in five or more like straight sections arranged side by side. 201143049 ("bottom spacing") may have a non-monotonic change. For example, the vertical center of the three lower light guides 316a, 316b on the left: two different bottom spacings between the "vertical dashed lines" (u: the large bottom spacing xb in the gap 422b, the larger gap ab 乂 the small bottom spacing xa shirt - Off 422a read small lower '^ Compared with the smaller bottom spacing Xa, the larger bottom spacing is again below. 110 is to accommodate more integrated circuit features. b Help in the insulating layer as shown in Figure 1A, the first pixel ( From the bottom between the left side _ Sa and the third pixel and the fourth image gate, = pixel = the same, and less than between the second pixel and the third pixel == = the larger between the pixel and the second pixel The bottom spacing s_b or the top of the substrate 10ό and adjacent to the 讫 讫 under the insulator' such features include the gate electrode 2 such as the interval of the conventional 盥 m m m m width # 在 ~, even at the first image width ^ + between u and the third pixel and the fourth pixel do not need this more than the μ, which leads to the problem of less than the optimal pixel density. In the silly view from the top down pixel_12 overlooking the ®1, its display ', the first, second and third embodiments of the train 12 are three columns and from: Fifteen pixels 14 in which a plurality of pixels are arranged side by side from left to right and down to the bottom. The center line of the gap is drawn with a thick gray line between each pair side by side. Area B represents the upper light guide 130 The top table area and the area C represent the bottom surface area 23 of the lower light guides 316a, 316b. 201143049 Field: Area A minus area b is the area of the gaps 422a, 422b between the color filters 'where A represents the pixel area. Figure 2A shows the area Four different pairs of B and C are 41 and 〇 1, 32 and 〇 2, ^ and 〇: 3, and B4 and C4. These four different pairs are repeated in a regular pattern. Figure 1A can be seen as Figure 2A. A vertical section of the first to fourth pixels (counting from the left) of the middle column. In this context, % = t3, vb = t4, Sa = s3, and sb = S4. Figure 2A shows the color of side-by-side pixels. The gap between the filters has repeated widths ti, t2, t3, t4, u, U2, U3, and U4. Cut from left to right or top to bottom along any of Figure 2A, gap width & To t4 or u to change in its own, and from the centerline of the first gap to the next first The pitch P of the center line of the gap remains constant. The gap width varies across a plurality of pixels arranged side by side in a direction perpendicular to the gap, which is directed to the left or right of the pixel in Figure 2A (as in Figure 2A) The gaps shown in the top view of the pixel array are from left to right (or right to left), or above or below the pixel (again, as seen in the top view of the pixel array in Figure 2A) The gap shown is from top to bottom (or from bottom to top). The change is non-monotonic, that is, increased, followed by decreasing 'and then another increase; and/or decreasing , followed by an increase, further followed by another decrease. For example, in the top column, from left to right, the gap width varies in the sequence t2, tl, t2, ti, where t2 < tl, which exhibits a sequence of changes in the pattern {increase, decrease, increase} In the middle column, the gap width varies in the sequence 13, 1443, 丨4, where 13<4, which exhibits a sequence of variations of the pattern {increase, decrease, increase}; and in the bottom column, the sequence and Top 24 201143049 The sequence of 歹J is the same. More recently, the side-by-side pixels are arranged in a wide range of degrees. The two rows in the gap width are alternated between: The change in the degree of progress is increasing and decreasing. The display gap is wider. In detail, the sequence of changes can be non-monotonic. The change in the width of the gap can be followed by an "early-period sequence. For example, the pattern of the second-largest repeat {increased, no change, decrease, increase the same, Figure 2A also shows the bottom interval of two columns. For example, the sequence of the middle column = f right 'bottom interval alternates between..., showing the sequence of changes {increase, decrease, 4 = the span in the direction of the lion and the image of the lion Between the two. The conventional image sensor pixel arrays using light guides are placed in a manner that they are uniformly displaced, thereby making it impossible to utilize an integrated circuit such as a gate electrode that is too small to be accommodated between the light guides. The small embodiment of the feature. The embodiment shown in Figure 2A enables the two light guides to the left of the second column to be closer to each other and also to bring the two light guides below the second column closer together so as to be in the middle (i. a wider space between the second light guide and the third light guide such that the additional gate electrode 10c4 can be fitted therein. Even if only a relatively wide spacing is required between certain pairs of laterally adjacent light guides, the light guide <conventional pixel array with uniform spacing between still All of the bays must be widened. In an alternative embodiment, any sequence of variations of the bottom spacing is At, as long as the sequence of changes is non-monotonic, that is, the bottom spacing is large, ^ 25 201143049 = to decrease, And the step-by-step is followed by another increase; and/or decrease, followed by a large, and further followed by another non-monotonic sequence of repetitions and repetitions. The cultural sequence can be a complex type. Large, between the two parts = change can be reduced, increased, decreased by 9 times. Preferably, the interval varies within a group of five consecutive pixels —.- or greater. - w also f The width of the bottom is shown in the second column as the bottom width sequence 3 b as 3'~1?, ^^ varies non-monotonously from left to right, where ~ Wb ' thus exhibits a variation pattern {increase, decrease, increase , reduce}, send a non-monotonous. In detail, for the four pixels in the top two and the leftmost = row, the bottom area 〇3 has the smallest bottom width, and the area Ci has the bottom width A, And the bottom widths of the bottom regions c2 and q are intermediate. The color of the primary color Bayer pattern is used for the pixel array 12 In the embodiment t, the image green pixel having the bottom regions C2 and C4 has a bottom light domain C1 read as a red pixel, and a pixel separable pixel having a bottom region Q. The intersection is in two different colors. The widths of the lower light guides in the alternating directions should be similarly alternated so that the spacing of the lower light guides and the lower light guides is optimal for guiding the light to be detected while also optimizing the space of the integrated circuit features. Size. Therefore, in the bottom area 〇: 3 pairs of tyrannosaurus pixels and bottom _ () 4 in the green pixel of this instance, in the second column 'in alternating pixels in the left to right direction in the blue pixel with In the case of alternating green pixels, the width of the light guide is correspondingly optimized 'to alternate between narrower widths (for blue pixels) and more k degrees (for green pixels) to save for holding such as gates The space between the electrodes and the light guides of the integrated circuit features of the contacts. Similarly 26 201143049, in the first column, in the alternating pixel in the green image with a wider width (for red for green pixels) more generally - for the purpose of comparing the buried u width of the insulation layer and compare this The purpose of the horizontal interval between the continuous light of the light guide # + can be measured at the idle electrode just a, i 〇 4b = ten (that is, parallel to the horizontal level of the photoelectric conversion single = surface) and the horizontal The extent of the interval 1 in this height range ^ and thus the width (and hence the width) is significant, in terms of the degree of density of the insulating layer 110 and/or the underlying wealth road can be loaded together. The road features include a hard electrode, a polycrystalline wire wiring 5 line (not shown), a polysilicon contact (not shown), a diffusion contact (not shown), and a metal wire. This horizontal interval should be non-monotonically varied across five or more pixels arranged side by side in the - direction, as shown in Figure 1A, with the quasi-3f-lightguide shifted to the _ side to occupy too small to accommodate the product. The small space of the circuit features allows more space to be concentrated on the other side to become large enough to accommodate an integrated circuit feature. Preferably, the variation manifests itself in eight or fewer than eight pixels arranged side by side in one direction. Further preferably, the horizontal interval spans eight or fewer pixels (more preferably within a group of five consecutive pixels) between a plurality of consecutive light guides = the widest horizontal interval and the narrowest horizontal interval The change between 〇1 μηη or greater. Still more preferably, the change is 0.2 μηη or more. Making this width of the light guide as small as possible (without obstructing the transmission of light) helps to widen the horizontal spacing ' thereby benefiting the denser packing of the integrated circuit features 27 201143049 in and below the insulating layer 110 and the pixel array 12 More dense filling of the pixels in the middle. Preferably, as shown in FIG. 1A, the width varies non-monotonically across five or more pixels arranged side by side in one direction to take advantage of the fact that the width is for a light guide that transmits only light of shorter wavelengths. It can be smaller than the light guide for transmitting light of longer wavelengths. Regarding the light guide buried in the insulating layer 110 for transmitting light having a wavelength of 500 nm in air, the width is preferably between 0.2 μm and 0.35 μm, more preferably 0.27 μηη +/- 10%. Regarding the light guide for transmitting light having a wavelength in the air equal to 6 〇〇 nm, the width should preferably be between 0.25 μηη and 0.4 μηη, more preferably within 0.33 μηη +/_ ίο%. With respect to a light guide for transmitting light having a wavelength in air equal to 7 〇 (11111), the width should preferably be between 〇 3 μηη and 〇 5 μιη, more preferably within 0.4 μηη +/- 10%. Non-monotonically varying the horizontal level between successive light guides at a horizontal level between a wire 108 and a gate electrode 1〇4a (ie, parallel to the plane & horizontal level of the photoelectric conversion unit) (or in detail) The horizontal spacing (or in detail the bottom spacing) between the light guides at the height between the wire 108 and the gate electrode 1G4a (inclusive) permits the light guide to shift to one side to occupy too small The small space that does not accommodate an integrated circuit feature is such that more space is lighter than H to become large enough to accommodate an integrated circuit feature. This can be seen in Figure 1A, in Figure u, if all of the lower light guides are equally spaced, in the second: ρ light guide and the third lower light guide (from the left) _ in the multiple wires 1G8 and gate The horizontal interval between the continuous light guide 28 201143049 at any level between the extreme, _, and chest (inclusive) (ie, parallel to the flat (level) of the light (four) change unit) (or The bottom spacer sb) accommodates an additional gate electrode 104c. Non-monotonically varying the gap width and/or gap spacing between the plurality of color filters permits the color filters to be displaced with the respective lower ones in such a manner that the respective lower light guides remain coupled Receive light from individual color filters. In the case where the gap is filled with air or a gas, or a liquid or solid medium having a lower refractive index than the color filter, or in the case where the convex top plate has a higher refractive index above the top plate than the lower plate In the case of a gap above, light can be diverted from the gap into the color filter even when the gap width is changed. Therefore, in the sequence of consecutive first, second, third, and fourth color filters, the second color filter can be shifted to be closer to the color of the younger; the light-sensitive sheet, and the second color The light sheet can be displaced closer to the fourth color filter, and all four are still effective in collecting light. As shown in FIG. 1B, the ray c and the ray d falling in the wider gap between the color filters of the second pixel and the third pixel are still extracted by the respective color filters and transmitted to the respective photoelectrics. Conversion unit. Having a gap between the color filters and enabling the gap width of some of the gaps to be reduced while widening the gap between the gaps that become narrower allows the light guides below the color filters to be in a narrower gap There is a similarly reduced spacing below and a widening interval below the widened gap while maintaining good light extraction of the pixels involved. Figure 2B shows the same top view of the fifteen pixels in Figure 2A. Fig. 1A can be regarded as a vertical cross section of the first image 29 201143049 from the left to the fourth pixel along the cutting line zz in Fig. 2B. The thick "+" mark in the bottom region and the cut line 22 indicates a vertical dangling line (in Figure 1A = dashed line) that is perpendicular to the substrate 1〇6 and includes the light guides below the pixels in the tangent plane. The _pixel bottom spacing = Xb along _ line 22 is shown to be non-monotonically changed. In detail, the pattern {Xa, xb, Xa, Xb} along the cutting line z 3, :} shows the pattern {increase and decrease: increasing the above description has shown that the packing density of the integrated circuit features can be The horizontal spacing between successive light guides of pixels arranged in one direction is a non-single-intensity enhancement. The non-monotonic variation of the horizontal interval, the bottom spacing, is aided by a non-monotonic change in any of the following: the horizontal spacing between successive light guides of pixels arranged in one direction (detailed 'bottom spacing') , gap width and gap spacing. For example, the allowable change in the clearance can be as high as possible. Preferably, the non-monotonicity of the variation occurs within a small number of pixels arranged side by side in a direction, such as 32 pixels, or for example 16 pixels, or more preferably within 8 pixels. Within this range, the (the) change exhibits an increase, followed by a decrease, followed by another increase; or a decrease, followed by an increase followed by another decrease. The above description also shows that the packing density can be enhanced by optimizing the width of the light guide embedded in the insulating layer with respect to the light of the different colors transmitted by the light guide to take up the smallest possible space. Light guides that transmit only light of shorter wavelengths should be narrower than light guides that transmit light of longer wavelengths. Although the present invention has been described and illustrated in the drawings, it is understood that the embodiments are only illustrative and not limiting, and the invention is not limited to the specific construction and configuration shown and described. This is because other people who are familiar with the technology can come up with various other modifications. For example, the pixels in the pixel array can be arranged side by side in a direction from left to right and from top to bottom (parallel to the plane of the photoelectric conversion unit) at 45 degrees. For example, the top/bottom opening of the upper/lower light guide may take a different shape than the rectangle, such as an octagon or a rectangle having a rounded corner. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1A is a diagram showing a cross section of four image sensor pixels according to an embodiment of the present invention; FIG. 1B is a ray tracing diagram for four image senses of FIG. 1A. Figure 2A is an illustration showing an image of fifteen pixels in an array. Figure 2B shows the same top view of fifteen pixels. Image = cut: explicit, which shows an alternative FIG. 4, which is an alternative embodiment of the four image sensing modes of the embodiment, is an illustration showing the cross section of the best four image sensor pixels; the four image sensing patterns of the alternative embodiment are one. Illustrated, it shows a cross section of a pixel; Figure 6 is a schematic diagram of an image sensor 31 201143049 Figure 7 is a ray tracing diagram for the third embodiment shown in Figure 4. 8A is a description of a primary color Bayer pattern; FIG. 8B is a description of a primary color Bayer pattern rotated by 45 degrees; FIG. 9 is a schematic view of a 4T pixel; and FIG. 10 is a schematic view of a 3T pixel. [Main Component Symbol Description] 10: Image Sensor 12: Pixel Array 14: Pixel 16: Optical Reader Circuit/Optical Reader 18: Output Signal/Vertical Signal Line 20: Column Decoder 22: Control Signal/Control Line 24: analog to digital converter (ADC) 66: ADC output bus 102: photoelectric conversion unit / photodiode 102a: photoelectric conversion unit 102b: photoelectric conversion unit 104a: gate electrode 104b: gate electrode 104c: gate Electrode electrode 32 201143049 106 : Substrate 108 : Wire 110 : Insulation layer / insulator 112 : Reset switch / transistor 114 : Select switch / transistor 114a : Color filter 114b : Color filter 116 : Output transistor 117 : Transfer gate/transistor/transfer switch 118: RST(n) line 120: IN line 121: TF(n) line 122: SEL line/SEL(n) line 124: OUT line/OUT(m) line 126: TF( n) Line 130: upper light guide 134: support film 230: protective layer 316a: lower light guide 316b: lower light guide 318: microlens 318a: bottom 318b of lower light guide: bottom 320 of lower light guide: transparent planarization layer 33 201143049 422a: gap 422a': narrower gap 422b: gap 422b': wider gap 500: transparent film 510 : Top plate 510a : Top plate 510b : Top plate A : Area B : Area B! Area b2 Area b3 Area b4 Area c : Area Cl Area C2 Area C3 Area C4 Area 34