201233973 六、發明說明： 【發明所屬之技術領域】 本發明是關於一種檢測方法及檢測裝置。 【先前技術】 在提馬半導體裝置的封裝密度的目的下，將形成有 電子電路的複數個基板層疊而成的層疊型半導體裝置 受矚目。有一種基板貼合裝置，將複數個基板層疊的狀 況下，將基板間的位置對準而貼合（參照例如專利文獻 1 )。 【專利文獻1】特開2009-231671號公報 ,將複數個基板層疊的狀況下，於定位有各基板的外 形做為基準。在這種狀況是外形被透過型光學系統檢201233973 VI. Description of the Invention: [Technical Field of the Invention] The present invention relates to a detecting method and a detecting device. [Prior Art] A laminated semiconductor device in which a plurality of substrates on which an electronic circuit is formed is laminated for the purpose of packing density of a Tamar semiconductor device has been attracting attention. There is a substrate bonding apparatus in which a plurality of substrates are stacked, and the positions between the substrates are aligned and bonded (see, for example, Patent Document 1). [Patent Document 1] JP-A-2009-231671, in which a plurality of substrates are stacked, the external shape in which each substrate is positioned is used as a reference. In this situation, the shape is checked by the optical system.
Hi ’檢測對象的基板是已經將複數個基板層疊而 ^的層i基板’特別是在上段基板比下段基板外形小的 狀況，正確地檢測該上段基板的外形是困難的。 【發明内容】 ^在本發明的第-形態是提供—種檢測方法，檢測層 登的複數個基板中的特定基板的位置，具備：將照明照 射至包含特定基板邊緣-部分的區域；取得攝影區域得 到的圖像；以及根據顯示於圖像的段差部的位置，確認 特定基板的邊緣位置；其中將照明照射的步驟中將照明 照射至區_料’以及，取得圖像的步驟+將區域攝 § 3 201233973 影的方向的至少一者，對於基板面方向是傾斜的。 ^ ^本發明的第二形態是提供一種檢測裝置，檢測層 甓的複數個基板中的特定基板的位置，具備：照明部， 將照明照射至包含特定基板邊緣一部分的區域；圖像取 得部，取得從對於基板的面方向傾斜的方向來攝影區域 得到的圖像；以及位置確認部，根據顯示於圖像的段差 部的位置，確認特定基板的邊緣位置；其中在照明^中 將照明照射至區域的方向，以及，取得圖像部中攝^區 域的方向的至少一者，對於基板的面方向是傾斜的= 又，上述發明概要，並非將本發明所有必要特徵 舉者。又，這些特徵群的次結合也能成為發明。玉1 【實施方式】 以下經由發明實施形態來說明本發明，但以下的實 施形態並非限定關於申請專利範圍的發明。又’在實二 形態中說明的特徵的所有組合，並不限於發明解決手段 所必需。 、又 第一圖是一實施形態的檢測裝置100的結構的示 意斜視圖。檢測裝置100檢測層疊的下部基板1〇2以及 上部基板104中上部基板1〇4的位置。檢測裝置1〇〇具 備：平台101、照明部108、圖像取得部no以及位置 確認部120。 下部基板102以及上部基板1〇4是被基板貼合裝置 等在厚度方向重合。上部基板丨〇4的外形比下部基板 1〇2的外形小。因此，於上部基板1〇4的邊緣，上部基 4 201233973 板104 $上面與下部基才反1〇2 #上面之間形成有段差。 平台101載置檢測對象的下部基板102及上部基板 104。平纟101相對於χ車由、γ軸及z轴平行移動。該 平台101也可以是再貼合其它基板於上部基板104等的 裝置的平台。這種狀況的平台101也可以在X軸、Y軸 及z軸周®旋轉。平台1G1的上面設有基準標記1〇3。 包含第-圖的斜視圖中，平台1〇1的上面内的左右方向 做為X軸’岫後方向做為γ軸。垂直於X軸、Y軸而 向上的是Z軸。 二，準標記103用於照明部1〇8及圖像取得部11〇的 5周整等。❹’採用做為使光學系統聚焦的基準，藉以 在基板位置的檢測作業前，將缝隙像114照射至基準標 s己103’清晰的基準標記1〇3的像可成像於攝影部 105 ’。再者，該基準標記1〇3，係藉由該基準標記1〇3， 做為比對平台101上的位置與攝影於攝影部1〇5的圖像 上的位置的基準。 照明部108提供基板位置檢測用的縫隙像114。照 明部108依序具有：光源119、透鏡118、縫隙116以 及透鏡115。 光源119發射攝影部1〇5可檢測波長的光，例如在 攝影部105可攝影可見光的狀況，發射可見光。透鏡118 將來自光源119的光聚焦。縫隙116決定檢測上部基板 104的位置時的照明範圍。透鏡115聚焦通過縫隙 後的光，將縫隙像114成像於下部基板1〇2及上部基板 104的上面。 201233973 照明部108係從對於下部基板i〇2及上部基板i〇4 的面方向傾斜，在第一圖中從左上，照射下部基板1〇2 及上部基板104。因照明部1〇8產生下部基板1〇2及上 部基板104上的縫隙像114,具有往圓盤狀下部基板1〇2 及上部基板104的徑方向延伸的細長形狀，其照射範圍 包含上部基板104的邊緣的一部分。照明部ι〇8，在層 疊基板正確載置於平台101的特定位置的狀況，預先記 憶邊緣應過來的位置，將照明照射至該位置。該邊緣在 下σ卩基板1〇2專為圓盤狀的狀況是圓周。邊緣也可以包 含槽口（notch)等特徵處。 圖像取得部110包含攝影部1〇5與透鏡112。圖像 取得部110將包含上部基板104邊緣一部分的區域，從 對於上部基板104等的面方向為傾斜的方向，在第一圖 中從右上攝影。這種狀況也是’圖像取得部110，在層 疊基板被正確載置於平台1〇1的特定位置的狀況，預先 記憶邊緣應過來的位置，攝影包含該位置的區域。 透鏡112將從下部基板1〇2及上部基板1〇4的上面 反射的光成像於攝影部1〇5。攝影部1〇5的一例’是圖 像被二維排列的CCD、CMOS等。攝影部105藉由將 成像於攝影面的像的光訊號各像素地變換為電性訊 號’來取得圖像106。位置確認部120分析圖像106， 根據呈現於圖像106的段差部的位置，確認上部基板 104的邊緣位置。 第一圖所示結構，並非限定照明部108及圖像取得 部110的光學系統。例如，透鏡118、115、112是概略 6 201233973 可以是偏向It是非偏向（™)光學系統，也 傾斜，來對主光…=光㈣統精由將攝影部10 5 1〇2的^ 線傾斜社部敍1G4及下部基板 102的表面，可以廣範圍地聚焦。 土牙 =下’用第-圖所示的檢測裝置100來說 郤土反104的位置的檢測方法。本檢測方法具備杧 像的步驟與確認位置的步驟。取得圖像的步驟了以 照明部108於包含上部基板1〇4邊緣一部分的/區. 左上將照明照射來形成縫隙像114的步驟；以及== 部基板102與上部基板1〇4的上面反射的縫隙: 從對於下部基板102與上部基板104傾斜的方 影部105來攝影而取得圖像1〇6的步驟。 第二圖是圖像取得部11()已取得的基板邊緣部的圖 像106的概念圖。於圖像106的上部基板反射像η】是 缝隙像114中在上部基板1 〇4反射後的部分的圖像。另 一方面，下部基板反射像134是縫隙像Π4中在下部基 板102反射部分的圖像。 確認位置的步驟具有：將圖像1〇6從攝影部1〇5傳 送至位置碟認部12〇的步驟；以及以位置確認部120的 圖像分析，根據呈現於上部基板反射像132與下部基板 反射像134之間的段差部E的位置，確認上部基板1〇4 的邊緣位置。 圖像106上的段差部E對應於上部基板丨〇4的邊緣 位置。在第一圖中’當縫隙像114照射區域内的上部基The substrate on which the Hi' is to be detected is a layer i substrate in which a plurality of substrates have been stacked, and in particular, the upper substrate has a smaller outer shape than the lower substrate, and it is difficult to accurately detect the outer shape of the upper substrate. SUMMARY OF THE INVENTION In the first aspect of the present invention, there is provided a method for detecting a position of a specific substrate among a plurality of substrates of a layer, comprising: illuminating an area including an edge portion of a specific substrate; and obtaining a photograph An image obtained by the region; and confirming an edge position of the specific substrate according to the position of the step portion displayed on the image; wherein the step of illuminating the illumination illuminates the region to the region and the step of acquiring the image + the region Photograph § 3 201233973 At least one of the directions of the shadow is inclined for the direction of the substrate surface. A second aspect of the present invention provides a detecting device that detects a position of a specific substrate among a plurality of substrates of a layer, and includes: an illuminating unit that illuminates an area including a part of a specific substrate edge; and an image obtaining unit; Acquiring an image obtained from the image capturing region in a direction inclined with respect to the surface direction of the substrate; and the position confirming portion confirming the edge position of the specific substrate based on the position of the step portion displayed on the image; wherein the illumination is irradiated to the illumination At least one of the direction of the region and the direction in which the image is captured in the image portion is inclined with respect to the plane direction of the substrate. Further, the above summary of the invention is not intended to be essential to the invention. Moreover, the sub-combination of these feature groups can also be an invention. Jade 1 [Embodiment] The present invention will be described below by way of embodiments of the invention, but the following embodiments are not intended to limit the invention. Further, all combinations of the features described in the second embodiment are not necessarily limited to the means for solving the invention. Further, Fig. 1 is a schematic perspective view showing the configuration of a detecting device 100 according to an embodiment. The detecting device 100 detects the positions of the stacked lower substrate 1〇2 and the upper substrate 1〇4 of the upper substrate 104. The detecting device 1 includes a platform 101, an illumination unit 108, an image acquisition unit no, and a position confirmation unit 120. The lower substrate 102 and the upper substrate 1〇4 are overlapped in the thickness direction by the substrate bonding apparatus or the like. The outer shape of the upper substrate 丨〇4 is smaller than the outer shape of the lower substrate 〇2. Therefore, at the edge of the upper substrate 1〇4, a step is formed between the upper base 4 and the lower surface of the lower layer of the 201233973 plate 104. The platform 101 mounts the lower substrate 102 and the upper substrate 104 to be detected. The flat jaw 101 moves in parallel with respect to the brake, the γ-axis, and the z-axis. The stage 101 may be a platform for reattaching other substrates to the upper substrate 104 or the like. The platform 101 in this state can also be rotated on the X-axis, the Y-axis, and the z-axis. A reference mark 1〇3 is provided on the upper surface of the platform 1G1. In the oblique view including the first figure, the left-right direction in the upper surface of the stage 1〇1 is taken as the X-axis and the rear direction is taken as the γ-axis. The Z axis is perpendicular to the X axis and the Y axis. Second, the quasi-mark 103 is used for the five-week alignment of the illumination unit 1〇8 and the image acquisition unit 11〇. ❹' is used as a reference for focusing the optical system, whereby an image in which the slit image 114 is irradiated to the fiducial mark 103' clear reference mark 1?3 before the substrate position detecting operation can be imaged on the photographing portion 105'. Further, the reference mark 1〇3 is used as a reference for aligning the position on the stage 101 with the position on the image captured by the photographing unit 1〇5 by the reference mark 1〇3. The illumination unit 108 provides a slit image 114 for detecting the substrate position. The illumination unit 108 has a light source 119, a lens 118, a slit 116, and a lens 115 in this order. The light source 119 emits light of a wavelength detectable by the photographing unit 1〇5, and emits visible light, for example, in a state where visible light can be captured by the photographing unit 105. Lens 118 focuses the light from source 119. The slit 116 determines the illumination range when the position of the upper substrate 104 is detected. The lens 115 focuses the light passing through the slit, and images the slit image 114 on the upper substrate 1〇2 and the upper substrate 104. 201233973 The illumination unit 108 is inclined from the surface direction of the lower substrate i〇2 and the upper substrate i〇4, and the lower substrate 1〇2 and the upper substrate 104 are irradiated from the upper left in the first drawing. The illumination unit 1〇8 generates the slit image 114 on the lower substrate 1〇2 and the upper substrate 104, and has an elongated shape extending in the radial direction of the disk-shaped lower substrate 1〇2 and the upper substrate 104, and the irradiation range includes the upper substrate. Part of the edge of 104. The illumination unit 〇8, in the case where the laminated substrate is correctly placed at a specific position of the stage 101, the position where the edge should come is recorded in advance, and the illumination is irradiated to the position. The edge is a circle in the case where the lower σ 卩 substrate 1 〇 2 is disk-shaped. Edges can also include features such as notches. The image acquisition unit 110 includes an imaging unit 1〇5 and a lens 112. The image acquisition unit 110 photographs a region including a part of the edge of the upper substrate 104 from the upper right in the first direction from the direction in which the surface of the upper substrate 104 or the like is inclined. In this case, the image acquisition unit 110 is in a position where the laminated substrate is correctly placed at a specific position of the stage 1〇1, and the position where the edge should come is stored in advance, and the area including the position is photographed. The lens 112 images the light reflected from the upper surface of the lower substrate 1A2 and the upper substrate 1A4 in the imaging unit 1〇5. An example of the photographing unit 1〇5 is a CCD, a CMOS, or the like in which images are two-dimensionally arranged. The photographing unit 105 obtains the image 106 by converting each of the optical signals of the image formed on the imaging surface into electrical signals ’. The position confirmation unit 120 analyzes the image 106, and confirms the edge position of the upper substrate 104 based on the position of the step portion presented in the image 106. The configuration shown in the first figure is not limited to the optical system of the illumination unit 108 and the image acquisition unit 110. For example, the lenses 118, 115, 112 are schematic 6 201233973 may be biased It is a non-biased (TM) optical system, also tilted, to the main light ... = light (four) is fine by tilting the line of the camera 10 5 1 〇 2 The surface of the 1G4 and lower substrate 102 can be widely focused. The method of detecting the position of the soil counter 104 by the detecting device 100 shown in Fig. This detection method has steps for anamorphism and steps for confirming the position. The step of obtaining an image is performed by the illumination unit 108 on a portion including the edge of the upper substrate 1〇4. The upper left side illuminates the illumination to form the slit image 114; and the == the upper surface of the substrate 102 and the upper substrate 1〇4 are reflected. Slot: A step of capturing an image 1 to 6 from the square portion 105 inclined to the lower substrate 102 and the upper substrate 104. The second diagram is a conceptual diagram of the image 106 of the edge portion of the substrate which has been acquired by the image acquisition unit 11(). The upper substrate reflection image η] of the image 106 is an image of a portion of the slit image 114 that is reflected by the upper substrate 1 〇4. On the other hand, the lower substrate reflection image 134 is an image of a portion of the slit image 在 4 reflected on the lower substrate 102. The step of confirming the position includes: a step of transmitting the image 1〇6 from the photographing unit 1〇5 to the position disc recognition unit 12〇; and an image analysis by the position confirming unit 120 according to the image reflected on the upper substrate 132 and the lower portion The position of the step portion E between the substrate reflection images 134 confirms the edge position of the upper substrate 1〇4. The step portion E on the image 106 corresponds to the edge position of the upper substrate 丨〇4. In the first figure, 'the upper base in the area where the slit image 114 is illuminated
S 7 201233973 板104的邊緣，對於紙面往内方向移動，則圖像1〇6的 段差部E的位置往左移動’當上部基板1〇4的邊緣對於 紙面往前方向移動，則圖像106的段差部E往右移動。· 因此，藉由分析段差部E的位置，可以確認上部基板 · 104的邊緣位置。 位置確認部120根據縫隙116的尺寸、照明部1〇8 以及圖像取得部110的光學倍率等，預先記憶上部基板 反射像132的縱寬D。位置確認部12〇根據缝隙116的 尺寸、照明部108以及圖像取得部110的光學倍率等， 預先記憶呈現於上部基板反射像132的橫寬L的最大值 Lmax° 分析圖像106時，首先以選擇窗136來選定分析圖 像範圍。在確認圖像106的上部基板反射像132的上下 位置的目的下，較佳為選擇窗136的縱寬b比上述寬D 還寬，選擇窗136的橫寬a比上述寬Lmax還窄。由於上 部基板反射像132的部分的亮度比周邊高，位置確認部 120藉由分析在選擇窗136選擇的圖像的縱方向亮度， 可以取得上部基板反射像132的上下位置以及寬D。 第三圖是確認段差部位置的說明圖。在確認段差部 E位置的目的下，較佳為選擇窗136的橫寬a比上述寬 Lmax還寬，選擇窗136的縱寬b比上述寬D還窄。位置 確認部120藉由分析在選擇窗136選擇的圖像的橫方向 亮度，可以確認段差部E的位置。從圖像106的段差部 E的位置，可以確認在缝隙像114照射區域的上部基板 1〇4的邊緣的，於平台1〇1的位置。 8 201233973 第四圖為呈現於圖像106的上部基板反射像132的 段差部E的亮度變化的示意曲線。橫軸表示第二圖等的 圖像106的橫方向座標，縱軸表示亮度。同圖是做為可 看到上部基板反射像132的亮度變化者。理想地，如折 線142’上部基板反射像132在段差部e表示清晰的亮 度變化，但現實地，以光學系統的像差如曲線144，在 段差部E具有上部基板反射像132的亮度和緩地變化的 區域。在此，亮度和缓地變化的區域的半值寬Sx稱為 散景（Bokeh )量。 因攝影面上的繞射產生散景量Sx，是/3又/NA的 位數(〇rder)。在此，万是光學系統成像倍率，λ是入射 光=波·^，ΝΑ是透鏡開口數。為了將段差部Ε正確地 確溪，較佳為包含三點以上的計測點於散景量的範圍 内例如丨在攝影部105用CCD的狀況，將CCD的像 =了，為u ’則包含三個以上的像素於&的範圍的 t j λ/ΝΑ) >3u。也就是說，該條件變成NA < I 0 人 /3u )。 例如，在， 你成 ΝΑ<η 、u==5//m、λ = 0.67㈣的狀況 货:成 να<0·045。該 να 的狀況的να的較佳上=件式表不用上W、u及' 、+、。 值。又，在用偏向光學系統€ '.L置換為偏向光學系統的橫倍率β一ϊ 可0 卜 的散ΐ i ^ m他條件的說明圖。除了第四圖所3 反射像134的縱方向^^板反射像132及下部基由 也存在散景量Sy。若要段差部：S 7 201233973 The edge of the board 104 moves in the inward direction of the paper, and the position of the step portion E of the image 1〇6 moves to the left. When the edge of the upper substrate 1〇4 moves in the forward direction of the paper, the image 106 The step portion E moves to the right. Therefore, by analyzing the position of the step portion E, the edge position of the upper substrate 104 can be confirmed. The position confirmation unit 120 stores the vertical width D of the upper substrate reflection image 132 in advance based on the size of the slit 116, the illumination unit 1〇8, and the optical magnification of the image acquisition unit 110. The position confirmation unit 12 预先 preliminarily stores the maximum value Lmax of the lateral width L of the upper substrate reflection image 132 based on the size of the slit 116, the optical magnification of the illumination unit 108 and the image acquisition unit 110, and the like. The analysis image range is selected by selection window 136. For the purpose of confirming the upper and lower positions of the upper substrate reflection image 132 of the image 106, it is preferable that the vertical width b of the selection window 136 is wider than the width D, and the lateral width a of the selection window 136 is narrower than the width Lmax. Since the luminance of the portion of the upper substrate reflection image 132 is higher than the periphery, the position confirmation unit 120 can obtain the vertical position and the width D of the upper substrate reflection image 132 by analyzing the longitudinal luminance of the image selected in the selection window 136. The third figure is an explanatory diagram for confirming the position of the step portion. For the purpose of confirming the position of the step portion E, it is preferable that the width a of the selection window 136 is wider than the width Lmax, and the width b of the selection window 136 is narrower than the width D. The position confirming unit 120 can confirm the position of the step portion E by analyzing the horizontal brightness of the image selected in the selection window 136. From the position of the step portion E of the image 106, the position of the stage 1〇1 at the edge of the upper substrate 1〇4 of the region where the slit image 114 is irradiated can be confirmed. 8 201233973 The fourth figure is a schematic curve showing the change in luminance of the step portion E of the upper substrate reflection image 132 of the image 106. The horizontal axis represents the horizontal coordinate of the image 106 of the second figure or the like, and the vertical axis represents the luminance. The same figure is used as a change in brightness of the upper substrate reflection image 132. Ideally, the upper substrate reflection image 132 such as the fold line 142' indicates a clear change in luminance at the step portion e, but realistically, the aberration of the optical system such as the curve 144 has the brightness and gentleness of the upper substrate reflection image 132 at the step portion E. Changing area. Here, the half value width Sx of the luminance and the gently changing region is referred to as the bokeh (Bokeh) amount. The amount of bokeh Sx due to diffraction on the photographic surface is /3 and the number of bits of NA (〇rder). Here, 10,000 is the imaging magnification of the optical system, λ is the incident light = wave · ^, and ΝΑ is the number of lens openings. In order to correct the ridges correctly, it is preferable to include the measurement points of three or more points in the range of the bokeh amount, for example, the CCD image is used in the imaging unit 105, and the image of the CCD is =, and u 'includes More than three pixels in the range of & tj λ/ΝΑ) > 3u. That is, the condition becomes NA < I 0 person /3u ). For example, if you become a condition of <η , u==5//m, λ = 0.67 (four), the goods are: να<0·045. The να of the condition of να is preferably the upper part table without W, u and ', +, . value. Further, in the case where the deflection optical system is replaced by the deflection optical system, the horizontal magnification β of the deflection optical system is 0 ϊ. The bokeh amount Sy is also present in the longitudinal direction of the reflection image 132 and the lower base of the reflection image 134 in the fourth figure. To the section:
S 9 201233973 f確認，較佳為段差部E的高度Η大於（Sy+muhm 疋用來確遇段差部E的像素數，所以是—以上的整數。 當考慮散景;g Sy也是0 λ/ΝΑ的位數，則較佳為滿足 接下來的上述式。 Η> ( β λ/ΝΑ) + mu ( 1 ) 段差部Ε的高度Η，是對應下部基板1〇2的上面與 上部基板104的上面的間隔h，即對應上部基板1〇4的 厚度，Η的大小是以下式來決定。 H=2hySsin6li (2) 其中，h是下部基板1〇2的上面與上部基板1〇4的 上面的間隔，~是人射光的人射角。在用偏向光學系 統的狀況.，將冷置換為偏向光學系統的橫倍率召-。 在0|為90°的狀況，sin^i表示最大值1。因此， Η的最大值可以用下式來表現。S 9 201233973 f confirms that it is preferable that the height Η of the step portion E is larger than (Sy+muhm 疋 is used to confirm the number of pixels of the step portion E, so it is an integer above. When considering bokeh; g Sy is also 0 λ/ The number of digits of ΝΑ is preferably such that the following equation is satisfied. Η > ( β λ / ΝΑ ) + mu ( 1 ) The height 段 of the step portion 对应 corresponds to the upper surface of the lower substrate 1 〇 2 and the upper substrate 104 The upper interval h corresponds to the thickness of the upper substrate 1〇4, and the size of the crucible is determined by the following equation: H=2hySsin6li (2) where h is the upper surface of the lower substrate 1〇2 and the upper surface of the upper substrate 1〇4 Interval, ~ is the angle of the person who emits light. In the case of using the bias optical system, the cold is replaced by the horizontal magnification of the biasing optical system. In the case where 0| is 90°, sin^i represents the maximum value of 1. Therefore, the maximum value of Η can be expressed by the following formula.
Hmax=2hy3 (3) 將式（3)帶入式（1)，則得到下式。 2hyS > ( β λ/ΝΑ) +mu (4) 例如，在召=1、u=5"m ' λ = 0.67^ m、i 的狀況，若要檢測30//m的間隔h，NA> 0.0012。該條 件式表示用上述0、u、λ及m的狀況的NA的較佳下 限值。 10 201233973 在更正確地檢測上部基板104的形狀的目的下，包 含入射光線與反射光線的入射平面，較佳為鄰接上部基 板104的邊緣。當入射平面從基板的切線方向偏離，則 在檢測結果產生誤差。在將該誤差抑制於容許範圍的目 的下，入射平面與上部基板104的切線方向所成角度， 較佳為調整在5°以内。 第六圖表示確認上部基板104的邊緣位置的其它 實施形態。該實施形態中，於上部基板104的邊緣的三 個區域’分別形成縫隙像114、172、174，取得在各區 域反射後的圖像，確認上部基板104的位置。在這種狀 況’具備分別對應縫隙像114、缝隙像172、缝隙像174 的檢測裴置100，各檢測裝置100以上述檢測方法，可 以確認對應的部位的邊緣位置。 •該實施形態中，上述基板104的形狀若為已知，則 藉由確認圖像106上的上部基板104的三處邊緣位置， 可以更正確地檢測平台101上的上部基板104的位置。 例如，若上部基板104為圓盤狀，則藉由確認上部基板 104的三處邊緣位置，可以確認上部基板1〇4的中心位 置及半徑。由此，可以正確檢測上部基板104的位置。 根據本實施形態，不只是檢測效率高，也可以防止在移 動基板檢測複數個地方的狀況所產生的誤差。 第七圖及第八圖表示確認上部基板104的邊緣位 置的又一實施形態。這些第七圖及第八圖表示繼續第六 圖的動作。在該實施形態中，將上部基板104等與被照 射至該上部基板104的照明及攝影區域相對地移動’並 201233973 取得複數個地方的圖.像，確認槽0等特徵處。 在這種狀況，如第七圖所示，Μ隙像L照射包含 ^部基板104的槽口的區域，鏠隙像172照射對槽口 9〇度旋轉的位置，縫隙像174照射對槽口刚度 的位置。縫隙像H4、縫隙像172、縫隙像174， ，置’形成往上部基板1G4的徑方向延伸的細長 裝置f取得各區__像，賴對應部位的邊 、、彖位置。如第七圖所示，藉由確認上部基板1〇 口，可以確認上部基板104的旋轉。 1曰 第八圖中，縫隙像114及縫隙 長邊方向疋Υ軸方向’縫隙像172的長邊方向是= 方向。這種狀況，，缝隙像114或縫隙像174的入 = 是垂直於Y軸的面。縫隙像172的入射方向 ：Hmax=2hy3 (3) By bringing the formula (3) into the formula (1), the following formula is obtained. 2hyS > ( β λ / ΝΑ ) + mu (4) For example, in the case of call = 1, u = 5 " m ' λ = 0.67 ^ m, i, to detect the interval h of 30 / / m, NA > 0.0012. This conditional expression indicates a preferred lower limit of the NA using the above-described conditions of 0, u, λ, and m. 10 201233973 The incident plane containing the incident ray and the reflected ray is preferably adjacent to the edge of the upper substrate 104 for the purpose of more accurately detecting the shape of the upper substrate 104. When the incident plane deviates from the tangential direction of the substrate, an error occurs in the detection result. In the case where the error is suppressed to the allowable range, the angle between the incident plane and the tangential direction of the upper substrate 104 is preferably adjusted to within 5°. Fig. 6 shows another embodiment for confirming the edge position of the upper substrate 104. In this embodiment, slit images 114, 172, and 174 are formed in the three regions ' at the edges of the upper substrate 104, and images reflected in the respective regions are obtained, and the position of the upper substrate 104 is confirmed. In this case, the detecting means 100 for the slit image 114, the slit image 172, and the slit image 174 are provided, and each detecting device 100 can confirm the edge position of the corresponding portion by the above-described detecting method. In the embodiment, if the shape of the substrate 104 is known, the position of the upper substrate 104 on the stage 101 can be more accurately detected by confirming the three edge positions of the upper substrate 104 on the image 106. For example, when the upper substrate 104 has a disk shape, the center position and radius of the upper substrate 1〇4 can be confirmed by confirming the three edge positions of the upper substrate 104. Thereby, the position of the upper substrate 104 can be correctly detected. According to this embodiment, it is possible to prevent an error caused by a situation in which a plurality of places are detected on the moving substrate, not only because the detection efficiency is high. The seventh and eighth figures show still another embodiment for confirming the edge position of the upper substrate 104. These seventh and eighth figures show the actions of continuing the sixth figure. In the above-described embodiment, the upper substrate 104 and the like are moved relative to the illumination and imaging regions that are irradiated onto the upper substrate 104, and the image of a plurality of places is acquired in 201233973, and features such as the groove 0 are confirmed. In this case, as shown in the seventh figure, the crevice image L illuminates the region including the notch of the substrate 104, the crevice image 172 illuminates the position rotated to the notch 9, and the slit image 174 illuminates the notch. The location of the stiffness. The slit image H4, the slit image 172, and the slit image 174 are formed so that the elongated device f extending in the radial direction of the upper substrate 1G4 obtains the image of each region, and the edge of the corresponding portion. As shown in the seventh figure, the rotation of the upper substrate 104 can be confirmed by confirming the opening of the upper substrate 1. 1A In the eighth diagram, the slit image 114 and the slit longitudinal direction are in the x-axis direction. The longitudinal direction of the slit image 172 is the = direction. In this case, the entrance of the slit image 114 or the slit image 174 is a plane perpendicular to the Y-axis. The incident direction of the slit image 172:
軸的面。 ；X 第七圖及第八圖中，藉由平台1〇1從第六圖所示 上部基板104等的位置往X方向，往圖中右方向移動 來移動上部基板104及下部基板102，檢測邊緣的複數 位置。這種狀況，在照明部108及圖像取得部u〇被固 定的狀態，平台101以一定速度移動，且圖像取得部攝 影複數圖像106。這種狀況，也可以是平台1〇1間歇土也 被移動，取得當平台101的停止時的圖像106。 苐九圖表示以從弟六圖到第八圖表示的實施形態、 獲得邊緣的複數位置資訊的一例。在第九圖中，表示^ 第六圖到第八圖中的縫隙像114所對應的圖像1〇6的段 差部位置Yl、Y2……，與沿著此時平台101的X細= 12 201233973 位置XI、X2......。由此，可以確認平台101上的槽口 的XY位置。 移動上部基板104等，確認邊緣的複數位置的方 法，不限於如第六圖到第八圖所示，將三個縫隙像照射 至層疊基板上的狀況，即，不限於用三個檢測裝置1〇〇， 確認層疊基板上的三點的狀況。在如第一圖所示，確認 一點的狀況，以及，在確認未滿三點、比三點多的複數 點的狀況也可以使用。即使是確認一點的狀況，藉由移 動上部基板104等，並確認複數位置的邊緣，可以檢測 上部基板104的位置及形狀。 在此，以平台101的移動所伴隨的振動等，變動移 動中平台101與攝影部105的位置關係，有產生位置的 特定誤差之虞。對此，在如第六圖到第八圖所示，確認 複數處的狀況，從對應縫隙像114及縫隙像174的檢測 裝置100，檢測因Y軸振動產生的移位。根據該Y軸的 振動移位值，修正從縫隙像172已確認位置資訊中Y軸 的值。同樣地，縫隙像172所對應的檢測裝置100，檢 測因X軸的振動導致的移位。根據該X軸的振動移位 值，修正從縫隙像114及縫隙像174已確認的位置資訊 中X軸的值。以該修正，可以更正確地檢測上部基板 104的形狀及位置。 第十圖及第十一圖是說明將照明掃瞄的實施形態 的正面圖。第十圖及第十一圖的實施形態中，替代移動 平台101，將照明的照射位置移動，並檢測邊緣的複數 位置。 13 201233973 如第十圖所示’照明部108於透鏡11 $的像侧，即 於與層疊基板之間’具有平行平板破璃182。在第十圖 所示狀態中，由於平行平板玻璃182的入射面垂直於主 光線地被配設，所以即使通過該平行平板玻璃182，照 明的照射位置也會變成從透鏡中心往主光線方向延長 的位置X,。 如第十一圖所示，對於來自透鏡的主光線，當平行 平板玻璃的角度傾斜，則可以保持往層疊基板的入射角 度’並將照射位置從乂]往X2移動。由此，可以變更平 行平板玻璃182的角度，並掃瞄照射位置，檢測邊緣的 複數位置。 第十二圖是說明將照明掃瞄的其它實施形態的正 面圖。第十二圖中，照明部108具有反射鏡184於瞳孔 位置。藉由變更反射鏡184的角度，可以移動縫隙像114 的照射位置。 根據第十圖到第十二圖的實施形態，可以不移動載 置層疊基板的平台1〇1。因此，可以更小巧地設計檢測 裝置100的整體。 在更正確地檢測上部基板104的形狀的目的下，具 有入射光線及反射光線的入射平面，較佳為鄰接上部基 板104的邊緣。若平台ι〇1或入射光線的移動範圍變 大’則在檢測範圍内與入射平面成大角度的邊緣也被包 含’這樣的部分的檢測精度會低落（參照第六圖至第八 圖）°因此’移動平台1〇1或入射光線並計測的狀況， 較佳為限定其移動範圍。例如，當用寬 0.065mm 的縫 201233973 隙116，檢測約300mm的上部基板l〇4的邊緣時，可以 預先調準成上部基板104的槽口對於基板中心向著γ方 向來配置於平台1〇1，平台1〇1或入射光線在5mm以 内的範圍移動，並正確地檢測上部基板1〇4的邊緣。 第十三圖表示確認四點的實施形態。除了第六圖中 的縫隙像114、縫隙像172及縫隙像174之外，更設有 縫隙像188，縫隙像188是照射相對於縫隙像172旋轉 180度的位置，藉此可以同時確認四個區域的邊緣位 置。這種狀況，四個縫隙像中，即使是一個縫隙像照射 上部基板104的槽口位置的狀況，以其他三個縫隙像， 也可以同時檢測上部基板104的中心位置。 第十四圖是在三片相異尺寸的基板被重疊的狀 況，以第一圖所示的檢測裝置1〇〇可取得的基板邊緣部 的圖像的概念圖。例如，將比下部基板102更大的基板 重合於第一圖的下部基板102的下方時’三片基板在圖 像106以從上開始的順序分別形成上部基板反射像 132、下部基板反射像134以及三層基板反射像I%。 即使是這種狀況，若對應最上方的上部基板1 〇4的邊 緣’於上部基板反射像132出現可識別程度的段差部 E，則以上述方法可以檢測該邊緣位置。 以上’根據本實施形態，使複數個基板貼合來製造 層疊半導體裝置的裝置中，可以正確地檢測應貼合的基 板外形、位置。由此’可以正確調整作為貼合對象的基 板彼此間的相對位置。 又，上述實施形態中有圖像取得部Π，圖像取得部 15 201233973 11取得因照明部108導致傾斜照明的正反射所產生的 圖像。但是，照明部1〇8及圖像取得部110的配置不限 於此。做為其他例，也可以是照明部108相對於基板面 方向為傾斜，圖像取得部110取得面的法線方向的圖 像。再做為其他例，也可以是照明部108從基板的面的 法線方向將照明照射，圖像取得部110從相對於面方向 傾斜來取得圖像。再做為其他例，也可以是照明部1〇8 及圖像取得部100之任一者，在相對於基板面為傾斜、 彼此偏離正反射的位置。 又，上述實施形態中，用縫隙像114做為照明。但 日 疋照明的例不限於此。做為其他例，也可以是縫隙像i i 4 與正負的關係，即，也可以是具有縫隙狀的影而周圍是 明冗的照明。這種狀況較佳為，於基板是圓形的狀況， 具有在徑方向延伸的圖案的照明。 以上，已用實施形態來說明本發明，但本發明的技 術，圍並不限於上述實施形態中記載的範圍。上述實施 幵中，可以施加各種變更或改良是所屬領域具有通常 =識者能明瞭的。經施加各種變更或改良的形態也能包 ί在本發明的技術範圍，從申請專利範圍的記載就能明 瞭0 么、，申請專利範圍、說明書及圖式中所示的裝置、系 方法中的動作、順序、步驟及階段等各處理 順序，只要未特別明示是「更之前」、「首先」等， 以二將前處理之輸*用在後處理的狀況，注意可 思順序來實現。關於申料利範圍、說明t及圖式 201233973 中的動作流程，即使是方便上用「首先」、「接下來」來 說明，也不代表必須以該順序來實施。 【圖式簡單說明】 第一圖：檢測裝置100結構的示意斜視圖。 第二圖：圖像取得部取得的基板邊緣部的圖像106的概 念圖。 第三圖：以位置確認部確認基板邊緣部位置的說明圖。 第四圖：於段差部E的亮度變化的示意曲線。 第五圖：檢測裝置的檢測條件的說明圖。 第六圖：從基板邊緣部的三處取得圖像的說明圖。 第七圖：移動基板並取得圖像的說明圖。 第八圖：移動基板並取得圖像的說明圖。 第九圖：以檢測出的基板邊緣位置來判斷基板外形及位 置的說明圖。 第十圖：說明掃瞄入射光線的實施形態的正面圖。 第十一圖：說明掃瞄入射光線的實施形態的正面圖。 第十二圖：說明掃瞄入射光線的其他實施形態的正面 圖。 第十三圖：從基板邊緣部的四處取得圖像的說明圖。 第十四圖：三層基板的基板邊緣部的圖像的概念圖。 【主要元件符號說明】 100 檢測裝置 101 平台 17 201233973 102 下部基板 103 基準標記 104 上部基板 105 攝影部 106 圖像 108 照明部 110 圖像取得部 112、 115、118 透鏡 114、 172、174 縫隙像 116 縫隙 119 光源 120 位置確認部 132 上部基板反射像 134 下部基板反射像 136 選擇窗 142 折線 144 曲線 182 平行平板玻璃 184 反射鏡 192 三層基板反射像 E 段差部 a ' b 、D、L 寬The face of the shaft. X, in the seventh and eighth figures, the upper substrate 104 and the lower substrate 102 are moved by moving the upper substrate 104 and the lower substrate 102 from the position of the upper substrate 104 and the like shown in the sixth figure to the X direction by the position of the upper substrate 104 and the like. The complex position of the edge. In this state, in a state where the illumination unit 108 and the image acquisition unit u are fixed, the stage 101 moves at a constant speed, and the image acquisition unit captures the complex image 106. In this case, the platform 1〇1 intermittent soil may also be moved to obtain an image 106 when the platform 101 is stopped. The figure IX shows an example of obtaining the complex position information of the edge in the embodiment shown from the sixth figure to the eighth figure. In the ninth diagram, the step position Y1, Y2, ... of the image 1 〇 6 corresponding to the slit image 114 in the sixth to eighth figures is shown, and X thin = 12 along the stage 101 at this time. 201233973 Location XI, X2.... Thereby, the XY position of the notch on the stage 101 can be confirmed. The method of moving the upper substrate 104 and the like to confirm the plural positions of the edges is not limited to the case where the three slit images are irradiated onto the laminated substrate as shown in FIGS. 6 to 8 , that is, not limited to three detecting devices 1 . 〇〇, confirm the status of the three points on the laminated substrate. As shown in the first figure, it is possible to confirm the situation at one point and to confirm the situation of a plurality of points that are less than three points and more than three points. Even if the situation is confirmed, the position and shape of the upper substrate 104 can be detected by moving the upper substrate 104 or the like and checking the edges of the plurality of positions. Here, the positional relationship between the stage 101 and the photographing unit 105 during the movement in the movement due to the vibration or the like accompanying the movement of the stage 101 may cause a specific error in the position. On the other hand, as shown in Figs. 6 to 8 , the state of the plurality of points is confirmed, and the displacement due to the Y-axis vibration is detected from the detecting device 100 corresponding to the slit image 114 and the slit image 174. Based on the vibration shift value of the Y-axis, the value of the Y-axis in the position information confirmed from the slit image 172 is corrected. Similarly, the detecting device 100 corresponding to the slit image 172 detects the displacement due to the vibration of the X-axis. Based on the vibration shift value of the X-axis, the value of the X-axis in the position information confirmed from the slit image 114 and the slit image 174 is corrected. With this correction, the shape and position of the upper substrate 104 can be detected more accurately. The tenth and eleventh drawings are front views for explaining an embodiment of scanning the illumination. In the embodiment of the tenth and eleventh embodiments, instead of the moving platform 101, the illumination position of the illumination is moved, and the plurality of positions of the edge are detected. 13 201233973 As shown in the tenth diagram, the illumination unit 108 has a parallel flat glass 182 on the image side of the lens 11 $, that is, between the laminated substrate. In the state shown in the tenth diagram, since the incident surface of the parallel plate glass 182 is disposed perpendicular to the chief ray, even if the parallel plate glass 182 is passed, the illumination position of the illumination becomes extended from the center of the lens toward the chief ray. The location of X,. As shown in Fig. 11, for the chief ray from the lens, when the angle of the parallel plate glass is inclined, the incident angle to the laminated substrate can be maintained and the irradiation position can be moved from 乂 to X2. Thereby, the angle of the parallel plate glass 182 can be changed, and the irradiation position can be scanned to detect the plural positions of the edges. Fig. 12 is a front elevational view showing another embodiment of scanning the illumination. In the twelfth figure, the illumination portion 108 has a mirror 184 at the pupil position. By changing the angle of the mirror 184, the irradiation position of the slit image 114 can be moved. According to the embodiment of the tenth to twelfth drawings, the stage 1〇1 on which the laminated substrate is placed can be prevented from moving. Therefore, the entirety of the detecting device 100 can be designed more compactly. The incident plane having the incident light and the reflected light is preferably adjacent to the edge of the upper substrate 104 for the purpose of more accurately detecting the shape of the upper substrate 104. If the plane ι〇1 or the range of the incident light becomes larger, the edge that is at a large angle to the incident plane within the detection range is also included. The detection accuracy of such a portion is low (refer to Fig. 6 to Fig. 8). Therefore, it is preferable to limit the movement range of the mobile platform 1〇1 or the incident light and the measured condition. For example, when the edge of the upper substrate 10b of about 300 mm is detected by the slit 201233973 slit 116 having a width of 0.065 mm, the notch of the upper substrate 104 can be pre-aligned to be disposed on the stage 1 对于 in the γ direction toward the center of the substrate. The platform 1〇1 or the incident light is moved within a range of 5 mm, and the edges of the upper substrate 1〇4 are correctly detected. The thirteenth figure shows an embodiment in which four points are confirmed. In addition to the slit image 114, the slit image 172, and the slit image 174 in the sixth figure, a slit image 188 is provided, and the slit image 188 is a position where the illumination is rotated by 180 degrees with respect to the slit image 172, whereby four can be confirmed at the same time. The edge position of the area. In this case, even if one of the four slit images is irradiated to the position of the notch of the upper substrate 104, the center position of the upper substrate 104 can be simultaneously detected by the other three slit images. Fig. 14 is a conceptual diagram of an image of the edge portion of the substrate which can be obtained by the detecting device 1 shown in Fig. 1 in a state in which three substrates of different sizes are overlapped. For example, when a substrate larger than the lower substrate 102 is overlapped under the lower substrate 102 of the first figure, the three substrates form the upper substrate reflection image 132 and the lower substrate reflection image 134 in the order of the image 106 from the top. And the three-layer substrate reflection image I%. Even in such a case, if the edge portion E corresponding to the uppermost substrate 1 〇 4 has an identifiable step portion E on the upper substrate reflection image 132, the edge position can be detected by the above method. According to the present embodiment, in the apparatus for manufacturing a laminated semiconductor device by laminating a plurality of substrates, it is possible to accurately detect the outer shape and position of the substrate to be bonded. Thereby, the relative positions of the substrates as the bonding objects can be correctly adjusted. Further, in the above embodiment, the image acquisition unit Π, the image acquisition unit 15 201233973 11 acquires an image caused by the regular reflection of the oblique illumination by the illumination unit 108. However, the arrangement of the illumination unit 1〇8 and the image acquisition unit 110 is not limited to this. As another example, the illumination unit 108 may be inclined with respect to the substrate surface direction, and the image acquisition unit 110 may acquire an image of the normal direction of the surface. In another example, the illumination unit 108 may illuminate the illumination from the normal direction of the surface of the substrate, and the image acquisition unit 110 may obtain an image by tilting from the surface direction. As another example, any of the illumination unit 1 8 and the image acquisition unit 100 may be inclined at a position offset from the substrate surface by a regular reflection. Further, in the above embodiment, the slit image 114 is used as illumination. However, the example of the daylighting illumination is not limited to this. As another example, the slit image i i 4 may be in a positive or negative relationship, that is, it may have a slit-like shadow and a surrounding illumination. In such a case, it is preferable that the substrate has a circular pattern and has illumination of a pattern extending in the radial direction. The present invention has been described above by way of embodiments, but the technical scope of the present invention is not limited to the scope described in the above embodiments. In the above-mentioned embodiments, various changes or improvements can be applied in the field of the art. The various modifications and improvements are also possible in the technical scope of the present invention, and the description of the scope of the claims can be made in the device, the method of the invention, and the method and method shown in the drawings. The processing sequence, such as the operation, the sequence, the step, and the stage, is not described as "before" or "first", and the pre-processing is used in the post-processing. Regarding the application flow range, description t, and the operation flow in the drawing 201233973, even if it is convenient to use "first" or "next" to explain, it does not mean that it must be implemented in this order. BRIEF DESCRIPTION OF THE DRAWINGS First FIG.: A schematic oblique view of the structure of the detecting device 100. Second diagram: A conceptual diagram of an image 106 at the edge portion of the substrate obtained by the image acquisition unit. Third drawing: An explanatory view for confirming the position of the edge portion of the substrate by the position checking unit. Fourth figure: A schematic curve of the change in luminance of the step portion E. Fig. 5 is an explanatory diagram of detection conditions of the detecting device. Fig. 6 is an explanatory diagram of taking an image from three places on the edge portion of the substrate. Figure 7: An explanatory diagram of moving the substrate and taking an image. Figure 8: An explanatory diagram of moving the substrate and taking an image. Fig. 9 is an explanatory diagram for judging the shape and position of the substrate by detecting the position of the edge of the substrate. Fig. 10 is a front elevational view showing an embodiment of scanning incident light. Figure 11 is a front elevational view showing an embodiment of scanning incident light. Fig. 12 is a front elevational view showing another embodiment of scanning incident light. Thirteenth Diagram: An explanatory diagram of taking images from four places on the edge portion of the substrate. Fig. 14 is a conceptual diagram of an image of a peripheral portion of a substrate of a three-layer substrate. [Description of main components] 100 detection device 101 platform 17 201233973 102 lower substrate 103 reference mark 104 upper substrate 105 imaging unit 106 image 108 illumination unit 110 image acquisition unit 112, 115, 118 lens 114, 172, 174 slit image 116 Slit 119 Light source 120 Position confirmation unit 132 Upper substrate reflection image 134 Lower substrate reflection image 136 Selection window 142 Polyline 144 Curve 182 Parallel plate glass 184 Mirror 192 Three-layer substrate reflection image E Segment difference a ' b , D, L Width