1279836 17657twf.doc/g 九、發明說明： 【發明所屬之技術領域】 本發明是有關於-種影像感测器的製造方法，且特別 疋有關於-種在半&體基底上形成影像感測器的製造方 法。 【先前技術】 影像感顚（image sensGO是將光學資訊轉換為電信 號的裝置。影像的種類可大致分為顯像管與固定攝 像元件。映像管以電視為巾^ ’叙祕將影像處理技術 運用的測量、控制、識別等，發展為應用技術。. 目則，固定攝像兀件包括電荷耦合⑽argedC〇upled Device，CCD)型與互補式金屬氣化半導體(CM〇s)型兩種。 CMOS型影像劇器是_ CM〇s製造技術，使光學 影像轉換為電信號的元件。如像素數，採用—種以聰 ^ ά _關方式。CM()S感測器與 被正在廣泛、大^:用作現有影像感測器的CCD型影像感 測益比較，具有驅動方式簡便，可實現多種掃描方式；可 將信號處理電路製傾單-晶片，不僅可使產品.實現小體 積’而且具有相容於CMOS技術，因而節省製造成本、降 低電力損耗的優點。由於這種優點，所以近年來，CM〇s 型影像感測器比CCD型影像感測器更進一步得到大量應 用。 然而，CMOS電晶體影像感測器的聚光鏡叫㈣及彩 色濾光陣列(color filter array，CFA)均為光阻材料，其燃點 5 1279騰.doc/g 很低，約小於300。(：，因此以習知的製造方法無法在聚光 鏡上形成一保護層以避免微粒或其他污染源的傷害。此 外，由於聚光鏡由光阻材料所製作而成，因此其結構相當 脆弱)於後續對晶圓進行清洗製程時，常會使得聚光鏡的 結構受到破壞。 另一方面，由於聚光鏡之間有一間隙存在，因此容易 造成散射光線直接穿透間隙，甚至照射至相鄰之光感測 區，因而發生干擾(cross_talk)現象，使CMOS電晶體影像 感測為接收到的雜訊增加，降低感測度。 【發明内容】 I有鑑於此，本發明的目的就是在提供一種影像感測器 的製造方法，可避免聚光鏡的結構在清洗製程中受到破壞。 本發明的再一目的是提供另一種影像感測器的製造方 法，能有效縮小聚光鏡之間的間隙。 本發明提出一種影像感測器的製造方法，影像成 是在-半導體基底上製作，半導體基底具包含 列。此製造方法首先於半導體基底上形成一第一平坦層。 接著，於第—平坦層上形絲色濾光_，純濾光陣列 形成於相對應的感測器陣列區域上方。然後，於彩色濾光 陣列上形成-第二平坦層。接下來，於第三平坦層上形成 夕數個聚仏、，料鏡分卿成於相職的彩色濾光陣列 上方之後以四乙氧基矽院(她时出㈣⑽謝， ^應氣體f進行-個電漿增強學氣相沈積製程，於些 *光鏡及帛—平坦層上軸共形的賴層，賴層是在射 1279836 17657twf.doc/g 頻功率為250瓦(W)〜450瓦且四r，甘、 鐘150〜500毫克(mgm)的環境下开土石&的通1為每分 依照本發明的-較佳實施例所述，在 $ 器、的製造方法中，形成保護層的製程溫度為15〇。〇!= 則 依照本發明的一較佳實施例所、+、 ^ _ 〇UC 〇 器的製造方法中，形成保護層更包=、’返，影像感測 (He)。 匕括通入軋氣(〇2)及氦氣 依照本發明的-較佳實施例所述，在 中’氧氣及氦氣的通量包括每分鐘 器的製造= 二補象感測 體影像感測器。 匕栝互補式金虱+導體電晶 器的例所述，在上述之影像感測 依照本發光陣列。 器的製造方料π 例所逑，在上述之影像感測 列區域外形成接合=保護層之後，更包括於感測器陣 是在娜则製造方法，影像感測器 接著，於繁一丞 、牛泠胆基底上形成一弟一平坦層。 形成於相對座層上形成彩色滤光陣列，彩色濾、光陣列 陣=_戚卿列區域上方。然後，於細光 开/成-弟二平坦層。接下來，於第二平坦層上形成 7 1279碰-c/g 多數個聚光鏡，聚光鏡分卿成於相對應的彩色濾、光陣列 上方。之後，以四乙氧基矽烷為反應氣體源進行一個電漿 增強型化學氣相沈積製程，於些聚光鏡及第二平坦層上形 成共形的保護層，保護層是在射頻功率為25〇瓦〜45〇瓦、 製程壓力》2〜4托爾(Torr)且四乙氧基石夕烧的通量為每分 鐘150〜500亳克的環境下形成。1279836 17657twf.doc/g IX. Description of the Invention: [Technical Field of the Invention] The present invention relates to a method of manufacturing an image sensor, and particularly relates to the formation of a sense of image on a semi- & The manufacturing method of the detector. [Prior Art] Image sensGO is a device that converts optical information into electrical signals. The types of images can be roughly divided into picture tubes and fixed imaging elements. The image tube is made of TV as a towel. Measurement, control, identification, etc., developed into application technology. Objectives, fixed camera components include charge coupled (10) argedC〇upled Device, CCD) and complementary metal vaporized semiconductor (CM 〇s) type. The CMOS video player is a component of the CM 〇s manufacturing technology that converts optical images into electrical signals. For example, the number of pixels is adopted in the form of Cong ^ ά _ off. The CM()S sensor is compared with the CCD-type image sensing which is widely used and used as an existing image sensor. It has a simple driving method and can realize a variety of scanning methods; - The wafer not only enables the product to achieve a small volume ' but also has the advantages of being compatible with CMOS technology, thereby saving manufacturing costs and reducing power loss. Due to this advantage, in recent years, the CM〇s type image sensor has been further used in a large number of applications than the CCD type image sensor. However, the condensing mirror of the CMOS transistor image sensor is called (4) and the color filter array (CFA) is a photoresist material, and its ignition point is 5 1279. The doc/g is very low, about less than 300. (: Therefore, a protective layer cannot be formed on the concentrating mirror by a conventional manufacturing method to avoid damage from particles or other sources of pollution. In addition, since the condensing mirror is made of a photoresist material, its structure is quite fragile) When the circle is cleaned, the structure of the condenser is often destroyed. On the other hand, since there is a gap between the condensing mirrors, it is easy to cause the scattered light to directly penetrate the gap and even illuminate the adjacent light sensing region, thereby causing a cross_talk phenomenon, so that the CMOS transistor image is sensed as receiving. The noise that arrives increases, reducing the sensitivity. SUMMARY OF THE INVENTION In view of the above, it is an object of the present invention to provide a method of fabricating an image sensor that avoids damage to the structure of the concentrating mirror during the cleaning process. It is still another object of the present invention to provide a method of fabricating another image sensor which can effectively reduce the gap between the condensing mirrors. The present invention provides a method of fabricating an image sensor that is fabricated on a semiconductor substrate having a semiconductor substrate comprising columns. This fabrication method first forms a first planar layer on the semiconductor substrate. Next, a filament filter is formed on the first flat layer, and a pure filter array is formed over the corresponding sensor array region. Then, a second flat layer is formed on the color filter array. Next, on the third flat layer, a plurality of polyfluorenes are formed on the third flat layer, and the material mirror is formed on top of the color filter array of the position, and then the tetraethoxy enamel is used (she is out (four) (10), and the gas should be f Conducting a plasma enhanced vapor deposition process on the *-mirror and 帛-flat layer on the upper axis of the conformal layer, the layer is 1279836 17657twf.doc / g frequency power is 250 watts (W) ~ 450 watts and four r, gan, bell 150 to 500 milligrams (mgm) of the environment of the open earth stone & pass 1 is in accordance with the preferred embodiment of the present invention, in the manufacturing method of The process temperature for forming the protective layer is 15 〇. 〇! = According to a preferred embodiment of the present invention, the manufacturing method of the +, ^ _ 〇 UC 〇 device forms a protective layer, and the image is sensed. Measured (He). Including the introduction of the gas (〇2) and helium in accordance with the preferred embodiment of the present invention, wherein the flux of oxygen and helium includes the manufacture of each minute = two complements The sensing body image sensor is described in the example of the 匕栝 complementary metal 虱 + conductor crystallizer, and the above image sensing is performed according to the illuminating array. After the formation of the bonding = protective layer outside the image sensing column area, the sensor array is included in the method of manufacturing, and the image sensor is then formed on the substrate of the scorpion and the scorpion a younger flat layer formed on the opposite seat layer to form a color filter array, color filter, light array array = _ 戚 列 column area above. Then, in the fine light open / into - two flat layer. Next, A plurality of concentrating mirrors are formed on the second flat layer, and a concentrating mirror is formed on the corresponding color filter and light array. Thereafter, a plasma enhanced type is prepared by using tetraethoxy decane as a reaction gas source. The chemical vapor deposition process forms a conformal protective layer on the concentrating mirror and the second flat layer, and the protective layer is at a radio frequency of 25 〜 to 45 〇, and the process pressure is 2 to 4 Torr and four The flux of ethoxylate is formed in an environment of 150 to 500 gram per minute.
。。依照本發明的-較佳實施例所述，在上述之影像感測 益的製造方法中，形成保護層的製程溫度為15(rc〜25(rc。 =本發明的-較佳實施例所述，在上述之影像感測 益的衣仏枝巾，形成保制更包括通人減及氣氣。 依照本發明的—較佳實施例所述 器的製造錢中，氧氣錢氣的通量包括每分鐘丨；^^ 公分。 施例所述，在上述之影像感測 為包括互補式金氧半導體電晶 依A?、本發明的一較佳實 器的製造方法中，影像感測 體影像感測器。 贫明的 器的製造方 … 土只她例所迷，在上述之影像感測 依昭本私明遽料列包括R/G/B彩色濾光陣列。 器的製造方i中實施例所述，在上述之影像感測 列區域外形成接合塾開ί保㈣之後，更包括於感測器陣 影像::ί發影像感測器的製造方法所製造的 或其他污_的傷^形财保藝，因此可避免微粒 8 /g 1279斷- 此外，由於聚光鏡上具有保護層，在後續對晶圓進行 清洗製程時，可以保護聚光鏡的結構不會受到破壞。 另外’保護層可以縮小聚光鏡之間的間隙，如此一來 可以增加聚光鏡吸收入射光的面積並減少入射光之間的干 擾(cross-talk)現象，可使CMOS電晶體影像感測器接收到 的雜訊降低，增加感測度。 另一方面，由於在本發明的製程壓力下所形成的保護 φ 層具有較小的應力值，因此可以避免聚光鏡發生破裂。 為讓本發明之上述和其他目的、特徵和優點能更明顯 易懂，下文特舉較佳實施例，並配合所附圖式，作詳細說 明如下。 【實施方式】 圖1A〜圖id所繪示為本發明一實施例之影像感測器 的製造流程剖面圖。 首先，凊參照圖1A，提供一半導體基底1〇〇，例如是 矽基底。半導體基底1〇〇中已形成有p型井區1〇2以及設 _ 置於P型井區102上的感測器陣列區域1〇4。感測器陣列 區域104例如是具有多數個感光二極體(未繪示)設置於p 型井區102上，以及多數個隔離結構1〇6設置於相鄰二個 感光二極體之間的半導體基底100中，且隔離結構106環 繞於感光二極體周圍。感光二極體例如是具有一 CMOS電 晶體(未繪示)設置於P型井區1〇2表面，以及一光感測區 108形成於p型井區1〇2表層並與CMOS電晶體電性連 接。隔離結構106例如是淺溝渠隔離結構。 9 1279磁 f.doc/g 接著，於半導體基底1〇〇上形成平坦層11〇，且平坦 層U0覆盍於感先二極體上。平坦層110例如是由一層富 矽氧化矽層（Silicon-rich oxide，SR0)及—層旋塗式玻璃 (spin-on glass)所組成。. . According to the preferred embodiment of the present invention, in the above method for manufacturing image sensing benefits, the process temperature for forming the protective layer is 15 (rc~25 (rc. = the preferred embodiment of the present invention). In the above-mentioned image sensing, the lychee towel is formed, and the formation of the protection includes the reduction of the air. In the manufacturing of the device according to the preferred embodiment of the present invention, the flux of oxygen and money includes每^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Sensors. The makers of the poorer ones... I only know how to use the R/G/B color filter array in the above-mentioned image sensing. In the embodiment, after the image sensing column area is formed outside the image opening column (4), it is further included in the sensor array image: manufacturing method of the image sensor or other materials. Injury ^ shape wealth protection, so you can avoid particles 8 / g 1279 off - in addition, due to the protective layer on the condenser In the subsequent cleaning process of the wafer, the structure of the concentrating mirror can be protected from damage. In addition, the 'protective layer can reduce the gap between the condensing mirrors, thus increasing the area of the condensing mirror that absorbs the incident light and reducing the interference between the incident light. The cross-talk phenomenon can reduce the noise received by the CMOS transistor image sensor and increase the sensitivity. On the other hand, since the protection φ layer formed under the process pressure of the present invention has a small stress The above and other objects, features, and advantages of the present invention will become more apparent and understood. 1A to 1D are cross-sectional views showing a manufacturing process of an image sensor according to an embodiment of the present invention. First, referring to FIG. 1A, a semiconductor substrate 1 is provided, such as a germanium substrate. A p-type well region 1〇2 has been formed in the crucible and a sensor array region 1〇4 disposed on the P-type well region 102. The sensor array region 104 is, for example, A plurality of photosensitive diodes (not shown) are disposed on the p-type well region 102, and a plurality of isolation structures 1〇6 are disposed in the semiconductor substrate 100 between the adjacent two photodiodes, and the isolation structure 106 Surrounding the photodiode, the photodiode has a CMOS transistor (not shown) disposed on the surface of the P-well region 1〇2, and a photo-sensing region 108 is formed in the p-well region. 2 The surface layer is electrically connected to the CMOS transistor. The isolation structure 106 is, for example, a shallow trench isolation structure. 9 1279 Magnetic f.doc/g Next, a flat layer 11〇 is formed on the semiconductor substrate 1〇〇, and the flat layer U0 is covered. The flat layer 110 is composed of, for example, a layer of a silicon-rich oxide layer (SR0) and a layer of spin-on glass.
然後，請參照圖1B，於平坦層110上形成多數個圖案 化金屬層112及114。圖案化金屬層112形成於感測器陣 列區域104上相對應的隔離結構1〇6上方，用來阻擋入射 光線的散射。圖案化金屬層114形成於感測器陣列區域1〇4 外相對應的隔離結構1〇6上方，用以作為接合墊金屬層。 接著，於平坦層110及圖案化金屬層112、114上^成 一平坦層116。平坦層116的材質例如是氧化石夕。平坦層 Π6的形成方法例如是以四乙氧基魏為反應氣體源 行一個電漿增強型化學氣相沈積製程而形成之。 此外，更可選擇性地於平坦層116上形成氮化石夕 繪示），可增加入射光的穩定性。氮化矽層的形成方法I 是電漿增強型化學氣相沈積法。 ϋThen, referring to FIG. 1B, a plurality of patterned metal layers 112 and 114 are formed on the planar layer 110. A patterned metal layer 112 is formed over the corresponding isolation structure 1〇6 on the sensor array region 104 for blocking scattering of incident light. The patterned metal layer 114 is formed over the corresponding isolation structure 1〇6 outside the sensor array region 1〇4 to serve as a bonding pad metal layer. Next, a flat layer 116 is formed on the planar layer 110 and the patterned metal layers 112, 114. The material of the flat layer 116 is, for example, oxidized stone. The formation method of the flat layer Π6 is formed, for example, by a plasma-enhanced chemical vapor deposition process using tetraethoxy Wei as a reaction gas source. In addition, it is more optional to form a nitride on the flat layer 116 to increase the stability of the incident light. The formation method of the tantalum nitride layer is a plasma enhanced chemical vapor deposition method. ϋ
之後，請參照圖1C，依序於平坦層116上形成串 光陣列118。彩色濾光陣列例如是紅色/綠色/藍色彩^^濾 陣列(R/G/BCFA)。彩色濾光陣列118的材質例如/思光 料，而彩色濾光陣列118的形成方法為與此技領材 常知識者所周知，而此不再贅述。 、^ Α有通 繼之，於彩色濾光陣列118上形成平坦層、， 層120的材質例如是光阻材料。 1 °平坦 隨後，平坦層120上形成多數個聚光鏡丨22，取 來光鏡 10 1279 碰 7twf.doc/g 二l2^別形成於相對應的彩色濾光陣列118上方，且二個 I光叙之間存在一間隙124。聚光鏡丨22的材質例如是 阻材料。 —接著，請參照圖1D，以四乙氧基矽烷為反應氣體源進 灯一個電漿增強型化學氣相沈積製程，於聚光鏡122及平 ，層120上形成共形的保護層126，可以縮小間隙124。保 邊層126的厚度例如是1500〜3000埃。保護層126是在射 φ 頻功率為250瓦〜450瓦且四乙氧基矽烷的通量為每分鐘 〇 %〇 4克的環境下形成。其中，射頻功率較佳的是 =5W’四乙氧基石夕烧的通量較佳的是每分鐘24()毫克。其 匕形成保護層126的製程條件如下所述，製程溫度例如是 在150°C〜250°C，較佳的是21(rc。於形成保護層126的 製程中更可以通入氧氣及氦氣，而氧氣及氦氣的通量例如 是每分鐘1000立方公分。 在另一較佳實施例中，保護層126是在射頻功率為250 瓦〜450瓦、製程壓力為2〜4托爾且四乙氧基石夕烧的通量為 鲁 每分鐘150〜500毫克的環境下形成。其中，射頻功率較佳 的是325W，製程壓力較佳的是2.5托爾，四乙氧基矽烷的 通量較佳的是每分鐘240毫克。其它形成保護層126的製 程條件如下所述，製程溫度例如是在150°C〜250°C，較佳 的是21〇°C。於形成保護層126的製程中更可以通入氧氣 及氦氣，而氧氣及氦氣的通量例如是每分鐘1〇〇〇立方公 分。 值得注意的是，依照本發明的製程條件所形成的保護 1279嫌啊 層126均勻地形成在聚光鏡122上，且與聚光鏡122間的 附者力大’因此保護層126不會有剝離的情況，也不會形 成具有缺陷的保護層126。此外，保護層126具有較低的 應力值、南均勻度、南透明度及低反射率index， RI)。保護層126的應力值例如是_2.5xl〇7dyne/cm2，均勻 度例如疋3 % ’ RI例如是1 · 6。 上述形成保護層126的各製程條件(射頻功率、TEOS . 的通量、製程壓力、氧氣及氦氣的通量等），例如是利用實 驗設計(design of experiment，DOE)求出各製程條件最佳的 參數範圍。 於形成保護層126之後，於感測器陣列區域1〇4外形 成接合墊開口 128。形成接合墊開口 128的方法例如是先 於保屢層126上形成圖案化光阻層（未綠示），再進行一個 蝕刻製程移除部份保護層126及平坦層116以曝露出圖案 化金屬層114。 I 、、本發明所提出之影像感測器的製造方法所製造的影像 感測器，由於在聚光鏡122上形成有保護層126，因此可 避免微粒或其他污染源的傷害。 此外，由於聚光鏡122上具有保護層126，在後續對 晶圓進行清洗製程時，保護層126可以保護聚光鏡122， 以確保聚光鏡122的結構不會受到破壞。 另外，保護層120可以縮小聚光鏡122之間的間隙 U4 ’甚至可以做到無間隙(卿㈣的情况，如此〆來可以 土曰加來光鏡122吸收人射光的面積並減少人射光之間的干 12 I279^34f.d〇c/g 擾現象，可使CMOS電晶體影像感測器接收到的雜訊降 低，增加感測度。 另一方面，由於在本發明的製程壓力下所形成的保護 層126具有較小的應力值，因此可以避免聚光鏡122產生 破裂的現象。 以下，為進行實際的實驗測試結果，用以說明使用本 發明所提出之光影像感測器的製造方法之優點。Thereafter, referring to FIG. 1C, a string array 118 is formed on the planar layer 116 in sequence. The color filter array is, for example, a red/green/blue color filter array (R/G/BCFA). The material of the color filter array 118 is, for example, a light material, and the method of forming the color filter array 118 is well known to those skilled in the art, and will not be described again. A planar layer is formed on the color filter array 118, and the material of the layer 120 is, for example, a photoresist material. After 1 ° flattening, a plurality of concentrating mirrors 22 are formed on the flat layer 120, and the light mirror 10 1279 is touched by 7 twf.doc/g. The two mirrors are formed on the corresponding color filter array 118, and two I-lights are formed. There is a gap 124 between them. The material of the condensing mirror 22 is, for example, a resist material. - Next, referring to FIG. 1D, a plasma enhanced chemical vapor deposition process is carried out by using tetraethoxy decane as a reactive gas source, and a conformal protective layer 126 is formed on the concentrating mirror 122 and the flat layer 120, which can be reduced. Clearance 124. The thickness of the edge layer 126 is, for example, 1,500 to 3,000 angstroms. The protective layer 126 is formed in an environment where the φ frequency power is 250 watts to 450 watts and the flux of tetraethoxy decane is 〇 % 〇 4 gram per minute. Among them, the RF power is preferably that the flux of =5W'tetraethoxy zephyr is preferably 24 () mg per minute. The process conditions for forming the protective layer 126 are as follows. The process temperature is, for example, 150 ° C to 250 ° C, preferably 21 (rc. In the process of forming the protective layer 126, oxygen and helium may be introduced. The flux of oxygen and helium is, for example, 1000 cubic centimeters per minute. In another preferred embodiment, the protective layer 126 has a radio frequency power of 250 watts to 450 watts, a process pressure of 2 to 4 Torr, and four. The flux of ethoxylate is formed in an environment of 150 to 500 mg per minute. Among them, the RF power is preferably 325 W, the process pressure is preferably 2.5 Torr, and the flux of tetraethoxy decane is higher. Preferably, 240 mg per minute is used. Other process conditions for forming the protective layer 126 are as follows, and the process temperature is, for example, from 150 ° C to 250 ° C, preferably 21 ° C. In the process of forming the protective layer 126. It is also possible to pass oxygen and helium, and the flux of oxygen and helium is, for example, 1 cubic centimeter per minute. It is worth noting that the protection of the process conditions according to the present invention is 1279. Formed on the condensing mirror 122 and attached to the concentrating mirror 122 Therefore, the protective layer 126 does not have a peeling condition, nor does it form a defective protective layer 126. Further, the protective layer 126 has a lower stress value, a south uniformity, a south transparency, and a low reflectance index, RI). . The stress value of the protective layer 126 is, for example, _2.5xl 〇 7dyne/cm 2 , and the uniformity such as 疋 3 % ' RI is, for example, 1.6. The process conditions (radio frequency power, flux of TEOS, process pressure, flux of oxygen and helium, etc.) for forming the protective layer 126 described above are, for example, a design of experiment (DOE) to determine the most process conditions. Good range of parameters. After the protective layer 126 is formed, the pad array opening 128 is formed in the sensor array region 1〇4. The method of forming the bond pad opening 128 is, for example, forming a patterned photoresist layer (not shown in green) on the security layer 126, and performing an etching process to remove a portion of the protective layer 126 and the planar layer 116 to expose the patterned metal layer. 114. In the image sensor manufactured by the method for manufacturing an image sensor according to the present invention, since the protective layer 126 is formed on the condensing mirror 122, damage of particles or other sources of contamination can be avoided. In addition, since the concentrating mirror 122 has a protective layer 126, the protective layer 126 can protect the concentrating mirror 122 during the subsequent cleaning process of the wafer to ensure that the structure of the concentrating mirror 122 is not damaged. In addition, the protective layer 120 can reduce the gap U4 between the condensing mirrors 122' even without gaps (clear (four), so that the light can be added to the light mirror 122 to absorb the area of the human light and reduce the light between the human light. Dry 12 I279^34f.d〇c/g disturbing phenomenon can reduce the noise received by the CMOS transistor image sensor and increase the sensitivity. On the other hand, the protection formed under the process pressure of the present invention The layer 126 has a small stress value, so that the phenomenon that the condensing mirror 122 is broken can be avoided. Hereinafter, the actual experimental test results are used to illustrate the advantages of the manufacturing method of the optical image sensor proposed by the present invention.
【實驗例1】 本發明的實驗例1，是在製程壓力為2.5托爾，射頻 功率為325W，TEOS通量為24〇mgm，氧氣及氛氣流量為 lOOOsccm的製程條件下，於晝素尺寸318//m及2.4#m 的聚光鏡上形成保護層，聚光鏡的厚度約為61〇〇埃。下表 1所載為崎描式電子顯微鏡量卿成健層之後聚光鏡 最高點的厚度。[Experimental Example 1] Experimental Example 1 of the present invention was carried out under the conditions of a process pressure of 2.5 Torr, a radio frequency power of 325 W, a TEOS flux of 24 〇 mgm, and an oxygen and atmosphere flow rate of 1000 sccm. A protective layer is formed on the 318//m and 2.4#m concentrating mirrors, and the thickness of the condensing mirror is about 61 angstroms. Table 1 below shows the thickness of the highest point of the concentrator after the smear-type electron microscope.
砂(埃）Sand (A)
户又可以得知晝素尺寸3.18/Zm&2..4#m的聚光 後;沒有出現後度被壓縮的現象。亦即，保護== 壞t光鏡的結構。 13 Ι279·_ 圖 圖2為將聚光鏡切開後的掃描式電子顯微鏡的照片 請參照圖2 ’以晝素尺寸3.18//m位於晶圓中央位置 的聚光鏡為例，圖2中標示6050埃的部份為聚光鏡最高點 的厚度，而標示2500埃的部份為保護層的厚度。由圖2 中可清楚的看出聚光鏡的結構完整，由此可知ς發 護層並不會使聚光鏡產生剝離或是破裂的 χ ’、 【實驗例2】 ° 本發明的實驗例2，是在製程承六％ 分鱼兔、系曰达， 土刀為2·5托爾’射頻 功羊為325W，TEOS通置為240mgm，务 lOOOsccm的製程條件下，於晝素尺乳乳及乱軋▲里為 的聚光鏡上形成保護層。下表2所恭J8/Zm及2.4^m 鏡量測形成保護層_聚光鏡表掃描式電子顯微 矣〇瓦又止拉士 、、口構的尺寸變化。 保護層形 成前/後 畫素尺寸 (Um) 保護層形 3-18//m --1—------ 成之前 2·4 U m 保護層形 3·18 // m 成之後 2·4 “ mThe household can also know the concentration of the alizarin size 3.18/Zm&2..4#m; there is no post-compression phenomenon. That is, the protection == structure of the bad t-mirror. 13 Ι279·_ Figure 2 is a photo of a scanning electron microscope after cutting the condenser. Refer to Figure 2 for an example of a concentrating mirror with a plaque size of 3.18/m at the center of the wafer. Figure 2 shows the 6050 angstrom. The part is the thickness of the highest point of the concentrating mirror, and the part marked with 2500 angstroms is the thickness of the protective layer. It can be clearly seen from Fig. 2 that the structure of the condensing mirror is complete, and thus it is known that the hair styling sheath does not cause peeling or cracking of the condensing mirror. [Experimental Example 2] Experimental Example 2 of the present invention is The process is divided into six parts, the fish rabbit, the system, the soil knife is 2. 5 Torr, the RF sheep is 325W, the TEOS is 240mgm, under the process of lOOOsccm, the 昼素尺乳乳 and the rolling ▲ A protective layer is formed on the concentrating mirror. Table 2 below shows the J8/Zm and 2.4^m mirror measurement to form a protective layer _ concentrating mirror table scanning electron microscope 矣〇 又 又 又 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 Protective layer formation front/rear pixel size (Um) Protective layer shape 3-18//m --1—------ Before formation 2·4 U m Protective layer shape 3·18 // m After 2 ·4 “ m
曲率半徑 (//m) 由表2可以得知’晝素7^^^^ 2.154 2.237 1.254 1.199 2.444 2.489 1.408 14 1279·' :wf.doc/g 光鏡不論是在晶圓巾央位置或是邊緣位置， 之後’聚域表面結構之間關隙 =面^ 的半徑變大。以畫素尺寸3.18_丄晶;== 光鏡為例，在形成保護層之前 ^央位置的聚 5 可以侍知在聚光鏡上形成保護層可以缩 光、兄之_間隙’且保護層能使 可 增加㈣鏡吸收人射光的面積。^ ™ 综上所述’本發少具有下顺點： 像二ti:提紐測， 害。 百保叹層，因此可避免微粒或其他污染源的傷 成的的製造方法中，所形 保護層，在後續對晶圓進行清洗製程 ”4 g可以保護聚光鏡的結構不會受到破壞。 像感細剛織所製造的影 達到無間可鳴聚光鏡之圆隙，甚至 面積並減少加聚光鏡吸收入射光的 影像n社 間的干擾現象’可降低CM0 s電晶體 〜像L接收到的雜訊，增加感測度。 5.在本發㈣製㈣力條件下卿成的保護層具有較 15Curvature radius (//m) can be seen from Table 2 '昼素7^^^^ 2.154 2.237 1.254 1.199 2.444 2.489 1.408 14 1279·' :wf.doc/g The light mirror is at the center of the wafer or The edge position, then the radius of the gap between the surface structure of the poly area = surface ^ becomes larger. Taking the pixel size 3.18_ twins; == light mirror as an example, before the formation of the protective layer, the poly 5 can be used to form a protective layer on the concentrating mirror, which can be dimmed, the gap of the brother and the protective layer can It can increase the area of the (4) mirror to absorb human light. ^ TM In summary, 'this hair has less to do with the next point: like two ti: mentioning New Zealand, harm. In the manufacturing method of avoiding the damage of particles or other pollution sources, the protective layer can be protected from the subsequent cleaning process of the wafer. 4 g can protect the structure of the condenser without being damaged. The shadow made by Gangzhi has reached the circular gap of the condensable condenser, and even the area and the interference phenomenon between the image and the image that absorbs the incident light by the condensing mirror can reduce the noise received by the CM0 s transistor~L, increasing the sense. Measure 5. Under the condition of (4) (4) force, the protective layer of Qingcheng has 15