200918330 九、發明說明 【發明所屬之技術領域】 本發明係有關於使用微電子機械系統（MEMS)技術來 製造噴墨列印頭的領域。 【先前技術】 許多不同種類的列印已被發明出來，其中的大多數現 在仍在使用中。列印的已知形式使用各種方法來用相關的 標記媒介來標記列印媒介。一般使用的列印形式包括偏位 列印，雷射列印及影印裝置，點矩陣式撞擊印表機，熱紙 印表機，薄膜記錄器，熱繼印表機，顏料次昇華印表機及 噴墨印表機兩者都是隨選液滴（drop on demand)及連續流 式印表機。當考量結構與操作的成本，速度，品質，可靠 性，簡單性時，每一種印表機都有其優點及缺點。 最近幾年，噴墨列印領域（每一獨立的墨水畫素都是 從一或多個墨水噴嘴得到的）已變得愈來愈流行，只要是 因爲它便宜及揮發的本質。 在噴墨列印上的許多不同技術已被發明出來。對於此 領域的檢視，可參考由 J Moore在 Output Hard Copy Device, Editors R Dube ck and S Sherr, pages 207-220(1988)所發表的一篇文章”Non-Impact Printing: Introduction and Historical Perspective” 。 噴墨印表機本身即有許多不同的種類。在噴墨列印中 使用連續的墨水流的技術可回溯至1 929年之授予Hansell 200918330 的美國專利第1,941，001號，且揭露了連續流靜電噴墨列 印的簡單形式。 美國專利第3,596,275號（Sweet等人）亦揭露了 一種連 續噴墨列印的處理其包括用高頻靜電場來調製該噴墨流用 以造成液滴分離的步驟。此技術仍爲數個製造商所使用， 包括Elmjet及Scitex(參見美國專利第3,373,437號（Sweet 等人））。 壓電式噴墨印表機亦是一種常被使用的噴墨列印技裝 置。壓電式系統是被 Kyser等人揭露在美國專利第 3,946,3 98號中其利用隔膜模式的操作，Zolten揭露在美 國專利第3,6 83,2 1 2號中其揭露一種壓電石英的壓擠操作 模式，Stemme在美國專利第3,747,120號中揭露壓電操作 的彎折模式，Howkins在美國專利第4,459,60 1號中揭露 一種噴墨流之壓電推出模式的致動及Fischbeck在美國專 利第4,5 8 4，5 9 0號中揭露一剪力模式的壓電換能器元件。 更現代地，熱噴墨列印已成爲一極爲流行的噴墨列印 形式。噴墨列印技術包括Endo等人在英國專利第 GB 2007162號及Vaught等人在美國專利第4,490,728號中所 揭露的技術。上述的這兩個專利文獻揭露的噴墨列印技術 依賴一電熱致動器的作動，其會造成在一被限制的空間中 (譬如一噴嘴中，產生一氣泡藉以造成墨水從一連接至該 被限制的空間的孔射出到一相關的列印摩介上。利用此電 熱致動器的列印裝置是由Canon及Hewlett Packard等製 造商所製造的。 -6- 200918330 從以上的描述可知，有許多不同的列印種類了供使用 。確實’一種列印技術應具有多種有育的屬性。這些屬性 包括建造及操作很便宜，高速操作，安全及長期的連續操 作等等。每一種技術在成本，速度，品質，可靠度，功率 使用，建造作業的簡單性，耐用性及可消費性等領域中有 其優點及缺點。 許多噴墨列印頭都是利用微電機系統（MEMS)技術來 建造。因此，它們都依賴沉積平面層於一矽晶圓上並將平 面層的某些部分蝕刻掉之標準的積體電路建造/製造技術 。在矽電路製造技術中，某些技術是較爲人所知道的。例 如’與其它用於奇特的電路製造（如，鐵電材質，砷化鎵 材質）的技術比較起來，與CMOS電路的製造相關聯的技 術是較常被使用的。因此，在Μ E M S結構中較佳地係使用 已被充分證明的半導體製技術其並不需要任何”奇特的，’處 理或材質。當然，將會有一定程度的犧牲，因爲如果使用 奇特的材質的優點遠大於其缺點的話，則無論如何都將偏 向於使用奇特的材質。然而，如果使用一般的材質可達到 相同或近似的特性的話，則奇特材質的問題就可被避免掉 〇 任何噴墨印表機的一項重要的議題爲列印頭保養。列 印頭保養可延長一列印頭的使用壽命並讓該列印頭在閒置 期之後可被使用。列印頭保養之典型的目標是要將微粒從 列印頭上去除掉，去除掉氾濫至該列印頭表面上的墨水， 及疏通被墨水或微粒堵塞住的噴嘴。迄今，已有各種技術 -7- 200918330 被用於列印頭保養上，譬如抽吸蓋或擠壓式刮除器。 然而’列印頭保養的一般問題在申請人的頁寬列印頭 上被惡化’頁寬列印頭上有用MEMS技術建造在一矽晶圓 上之高密度噴嘴。雖然這些列印頭在製造上很便宜，但與 其它列印頭比較起來它們較不耐用，因此到目前爲止，其 在列印頭保養枋面需要特別的考量。因此，本案申請人已 經提出許多用於列印頭保養的新穎技術，包括非接觸式保 養技術。這些保養技術的一部分已被體現在美國專利申請 案第 1 1 /246,6 8 8 號（2005 年 10月 11日提申）；第 11 /2 46，707 號（2005 年 10 月 11 日提申）；第 11 /2 46,693 號 (2005 年 10 月 11 日提申）；第 1 1 /246,6 8 8 號（2005 年 10 月11日提申）；第11 /4 82,958號（2006年7月10日提申） :及第11 /2 46,688號（2006年7月31曰提申），這些專利 申請案的每一件的內容都藉遺次參照被倂於本文中。 提供一種可經得起多種列印頭保養技術（包括接觸式 保養技術）的考驗之MEMS頁寬列印頭是所想要的。提供 一種具有絕佳的機械強健性之MEMS列印頭更是所想要的 。提供一種會捕捉住最少量的微粒且有利於列印頭保養之 MEMS列印頭更是所想要的。 【發明內容】 在一第一態樣中，一種噴墨列印頭被提供，其包含一 橫跨多個噴嘴之強化的雙層式噴嘴板結構° 選擇上地，每一噴嘴都包含一噴嘴室其具有一室頂’ -8 - 200918330 每一室頂都是由該噴嘴板結構的一部分所界定的。 選擇上地，噴嘴室係被形成在一基材上。 選擇上地，每一噴嘴室都包含該室頂其與該基材分隔 開來，及側壁其延伸於該室頂與該基材之間。 選擇上地，每一室都具有一噴嘴孔界定於其內。 選擇上地，該噴嘴板結構包含： 一第一噴嘴板，其橫跨多個噴嘴，該第一噴嘴板具有 多個穴室界定於其內； 塡充該等穴室的光阻劑；及 一第二噴嘴板其覆蓋該第一噴嘴板及該光阻劑。 選擇上地，該第二噴嘴板界定該列印頭的一平面的外 表面。 選擇上地，該第一及第二噴嘴板是用相同的或不同的 材質製成。 選擇上地，該等材質爲可用PECVD沉積的陶瓷材質 〇 選擇上地，該等材質係被獨立地選自於包含：氮化矽 ，氧化矽及氮氧化矽的組群中。 選擇上地，每一噴嘴都包含一形成在一基材上的噴嘴 室，該噴嘴室包含一頂其與該基材間隔開來及側壁其延伸 於該室頂與該基材之間，其中該第一噴嘴板與側壁是用相 同材質組成的。 在第二態樣中，一種噴墨列印頭積體電路被提供，其 包含： -9- 200918330 一基材，其具有多個形成於其上的噴嘴； 驅動電路，其被電連接至與噴嘴相關聯的致動器；及 一強化的雙層式噴嘴板結構，其橫跨該等多個噴嘴。 在第三態樣中，一種製造一具有一平的噴嘴板的噴墨 列印頭的方法被提供，該方法包含的步驟爲： (a) 提供部分製造好的一列印頭，其具有一由第一材 質組成之橫跨多個噴嘴的第一噴嘴板，該第一噴嘴板具有 多個穴室； (b) 用一塡料來塡充該等穴室，使得該第一噴嘴板的 上表面與該塡料的上表面一起界定一連續的平的表面；及 (c) 沉積一第二材質於該平的平面上用以形成一第二 噴嘴板其具有一平的外表面。 選擇上地，該第二材質是用PECVD來沉積的。 選擇上地，該第一材質係用PECVD而被沉積到一非 平面的犧牲性台架上用以形成該第一噴嘴板。 選擇上地，該第一及第二材質是彼此相同的或是不同 的。 選擇上地，該第一及第二材質係被獨立地選自於包含 :氮化矽，氧化矽及氮氧化矽的組群中。 選擇上地，該塡料是光阻劑。 選擇上地，步驟（b)是用下面的子步驟來實施： (b)(i)沉積一光阻劑層於該第一噴嘴板上用以塡充該 等穴室：及 (b)(ii)去除掉該光阻劑的一部分使得該第一噴嘴板的 -10- 200918330 一上表面及塡充該等穴室的該光阻劑的一上表面一起界定 一連續的平的表面。 選擇上地’該方法進一步包含的步驟爲： 將該光阻劑熱平流（reflowing)用以促進該等穴室之完 全塡充。 選擇上地，步驟（b)(ii)是用化學機械平坦化或用光阻 劑蝕刻來實施。 選擇上地，該方法進一步包含的步驟爲： (d)界定穿過該第一及第二噴嘴板的噴嘴孔。 選擇上地，每一噴嘴都包含一形成在一基材上的噴嘴 室，該噴嘴室包含一室頂其與該基材間隔開來及側壁其延 伸於該室頂與該基材之間，其中該第一噴嘴板與側壁是用 相同材質組成的。 依據本發明的列印頭包含多個噴嘴，及典型地對應於 每一噴嘴的一室與一致動器（如，加熱器元件）。該列印頭 的最小重復單元具有一供墨入口，其給送墨水至一或多個 室。一整個噴嘴陣列是藉由重復這些獨立的單元來形成的 。此一獨立的單元在本文中被統稱爲爲一，’單元細胞”。一 列印頭可由多個列印頭積體電路組成，每一列印頭積體電 路都包含多個噴嘴。 本文中所用之”墨水（ink)”一詞意謂著任何可噴出的液 體，且並不侷限於含有著色染料的傳統墨水。非著色（11〇11-c ο 1 〇 r e d)墨水包括定色劑’紅外線吸收劑墨水，功能化的 化學物，黏劑，生物流體’藥劑，水及其它溶劑，等等。 -11 - 200918330 該墨水或可噴出的液體亦不一定是一液體，且可包含物體 顆粒的懸浮液。 【實施方式】 一開始參照圖36，其顯示一 MEMS列印頭積體電路 的切開立體圖，如描述於吾人之美國專利申請案第 11/246,684號（2005年10月11曰）中者，該專利申請案內 容藉由此參照被倂於本文中。如圖3 6所示，每一列噴嘴 都具有一各自的供墨通道2 7沿著其長度延伸並供應墨水 至每一列中之多個墨水入口 1 5。該等墨水入口然後將墨水 供應至用於每一列的墨水導管23，其中每一噴嘴室接受來 自沿著每一列縱長向地延伸的共同墨水導管的墨水。具有 各自的噴嘴邊25之噴嘴孔26被界定於一噴嘴板1〇1上， 該噴嘴板橫跨噴嘴列及噴嘴行。如將於下文中詳細說明的 ，該噴嘴板1 01係將一陶瓷物質（如’化矽）PECVD於一光 阻劑台架上來形成的。由於此沉積處理的關係，該噴嘴板 101具有多個界定於其上的穴室102。穴室102被設置在 一列噴嘴中相鄰的噴嘴之間。這些穴室1 02典型地有數微 米深（如，1至5微米深）且造成該噴嘴板的不連續整體效 果爲一噴嘴板，其由於這些穴室1 02的關係而實質上是非 平面的。 依照特定的噴嘴設計與製造處理，穴室102可以實質 上比圖36所示的更大（更長’更寬或更深）。它們可延伸於 噴嘴列之間或噴行之間。 -12- 200918330 基於述項理由，穴室1 02所引起之該噴嘴板101的不 連續性與非平面性是不利的。首先，穴室1 02是噴嘴板 1 0 1上的脆弱點且降低噴嘴板的整體機械強健度，特別是 關於施加在整個噴嘴板上的剪力。這是特別重要的，因爲 掃過該噴嘴板101的整個表面的擦拭動作（在某些列印頭 保養期間會實施的動作）會造成相當大的剪力。其次，該 等穴室102能夠輕易地捕捉住墨水及/或顆粒，然後這些 墨水及/或顆粒就很難去除掉。穴室102靠近噴嘴孔26是 特別不想要，因爲任何被捕捉住的顆粒都很可能會阻塞噴 嘴並影響到列印品質。 爲了要對本發明有完整的瞭解，下文所描述的是圖36 中之列印頭積體電路是如何用一 MEMS製程來製造。此外 ，下文中還提供依據本發明的另一製程，其中該噴嘴板 1 0 1的平坦性被顯著地改善。 MEMS製程 該MEMS製程在完成CMOS處理之後建造噴嘴結構於 一矽晶圓上。圖2爲一噴嘴單元細胞1〇〇在完成CMOS處 理之後及在MEMS處理之前的一切開立體圖。 在晶圓的CMOS處理期間，四層金屬層被沉積在一矽 晶圓2上，其中這些金屬層散置在層監介電（ILD)層之間 。至四層金屬層被稱爲Ml，M2，M3及M4層且在CMOS 處理期間被依序地建構在晶圓上。這些CMOS層提供用來 操作該列印頭之所有驅動電路及邏輯電路。 -13- 200918330 在完成的列印頭中，每一加熱器元件致 定在最外層的M4層上的一對電極被連接至 此，該M4 COMS層爲該晶圓之後續MEMS 該M4層亦界定沿著每一列印頭積體電路的 合墊。這些黏合墊（未示出）讓該CMOS能夠 延伸出的電線黏合件而被連接至一微處理器 圖1及2顯示鋁M4層3，其具有一被 其上。（只有M4層的MEMS特徵結構被示 ;M4層的主要CMOS特徵結構被放置在該 外。）該M4層3具有1微米的厚度且本身被 厚的CVD氧化物層5上。如圖1及2所示 有一墨水入口開口 6及一坑開口 7。這些開 於MEMS處理中被形成的墨水入口及凹坑的 在該單元細胞11的MEMS處理開始之 列印頭積體電路的縱長邊緣的黏合墊藉由蝕 層4而被界定。此餓刻露出在黏合墊位置處 該噴嘴單元細鑣被用於此步驟的光阻完全地 會受到蝕刻的影響。 翻到圖3至5，該MEMS處理的第一階 坑8穿透該被動層4與該CVD氧化物層5。 一光阻劑層（未示出）來實施，該光阻劑層用另 暗色調凹坑光罩來加以曝光。該凹坑8具有 度，其是從該M4層3的頂端量起的深度。 8的同時，電極9藉由讓M4層3穿過被動 動器都經由界 該 CMOS。因 處理的基礎。 縱長邊緣的黏 經由從黏合墊 〇 動層4形成於 於這兩個圖中 噴嘴單元細胞 沉積在2微米 5，M4層3具 口界定出後續 位置。 前，沿著每一 刻穿透該被動 的M4層3。 遮蓋，因此不 段係蝕刻一凹 此蝕刻係使用 民於圖3中之 一 2微米的深 在蝕刻該凹坑 層4被部分地 -14- 200918330 露出而被形成在該凹坑的兩側。在完成的噴嘴中’ 一加熱 器元件被懸跨於凹坑8兩側的電極9之間。 在下一個步驟中（圖6至8)，該凹坑8被塡以光阻劑 10的第一犧牲層（“SAC 1”）。一 2微米厚的高黏性光阻劑首 先被旋施於該晶圓上，然後使用示於圖6中之暗色調凹坑 光罩來加以曝光。該SAC 1光阻劑1 0形成一台架，供後 續的加熱器物質沉積橫跨於該凹坑8兩側的電極9之間。 因此，該SAC1光阻劑10具有一與電極9的上表面齊平 之平的上表面是很重要的。在此同時，該SAC1光阻劑10 必需完全塡滿該凹坑8用以避免延伸橫跨該凹坑的導電加 熱器物質的”縱樑（stringers)”及將電極9短路。 典型地，當用光阻劑塡充溝渠時，有必要讓光阻劑露 到溝渠的周邊外面用以確保光阻劑有塡滿到溝渠的側壁處 ，因而可避免掉在後續沉積步驟中的”縱樑”。然而，此技 術會造成光阻劑在溝渠的周邊的附近會產生壟起來的邊緣 。這是所不想要的，因爲在後續的沉積步驟中，物質會被 不均勻地沉積在該壟起的邊緣上，該壟起的邊緣的垂直或 有角度的表面接受到的沉積物質少於塡充該溝渠之光阻劑 的水平的平的表面接受的沉積物質。此結果是在物質被很 薄地沉積的區域中的”電阻熱點”。 如圖7所示，目前的處理使用圖6所示的光罩蓄意地 露出在凹坑8的周壁內部的SAC1光阻劑10(如，在0.5微 米內）。這可確保一平的SAC1光阻劑10的上表面且可避 免在該凹坑8的周邊邊緣附近的光阻劑壟起區域。 -15- 200918330 在SAC1光阻劑10的曝光之後，該光阻劑藉由加熱 而被平流。將光阻劑平流可讓光阻劑流至凹康8的側壁， 將凹坑8確實塡滿。圖9與10顯示出在平流之後的SAC1 光阻劑10。該光阻劑具有一平的上表面且與形成電極9之 該M4層3的上表面齊平。在平流之後，該SAC1光阻劑 10被U.V.硬化及/或被烘烤用以避免在後續之加熱器物質 的沉積步驟期間有任何的平流。 圖1 1及1 2顯示在0.5微米的加熱器物質1 1沉積到 該SAC1光阻劑10上之後的單元細胞。由於上述之平流 處理，該加熱器物質11被均勻地沉積成爲在電極9與 SAC1光阻劑10上的一平的層。該加熱器物質可以是用任 何適合的導電物質組成，譬如 TiAl，TiN，TiAIN， TiAlSiN等等。一種典型的加熱器物質沉積處理可包括 100A 之 TiAl 種子層，2500A 之 TiAIN 層，另一 100A 之200918330 IX. Description of the Invention [Technical Field of the Invention] The present invention relates to the field of manufacturing ink jet print heads using microelectromechanical systems (MEMS) technology. [Prior Art] Many different types of prints have been invented, and most of them are still in use today. The known form of printing uses various methods to mark the printing medium with the associated marking medium. Commonly used printing formats include offset printing, laser printing and photocopying devices, dot matrix impact printers, hot paper printers, film recorders, thermal relay printers, pigment sublimation printers And inkjet printers are both drop on demand and continuous flow printers. When considering the cost, speed, quality, reliability, and simplicity of structure and operation, each type of printer has its advantages and disadvantages. In recent years, the field of inkjet printing (each individual ink element is obtained from one or more ink nozzles) has become increasingly popular, as long as it is cheap and volatile. Many different techniques for ink jet printing have been invented. For an examination of this field, refer to an article "Non-Impact Printing: Introduction and Historical Perspective" by J Moore in Output Hard Copy Device, Editors R Dube ck and S Sherr, pages 207-220 (1988). There are many different types of inkjet printers themselves. The use of a continuous flow of ink in ink jet printing is back to U.S. Patent No. 1,941,001, issued to Hansell 2009, 183, the disclosure of which is incorporated herein by reference. U.S. Patent No. 3,596,275 (Sweet et al.) also discloses a process for continuous ink jet printing which comprises the step of modulating the ink jet stream with a high frequency electrostatic field to cause droplet separation. This technology is still used by several manufacturers, including Elmjet and Scitex (see U.S. Patent No. 3,373,437 (Sweet et al.)). Piezoelectric inkjet printers are also a commonly used inkjet printing technology. A piezoelectric system is disclosed in U.S. Patent No. 3,946, issued toK.S. Pat. No. 3,946, issued to U.S. Pat. In the squeezing mode of operation, Stemme discloses a bending mode of piezoelectric operation in U.S. Patent No. 3,747,120, the disclosure of which is incorporated herein by reference. A piezoelectric transducer element of a shear mode is disclosed in Patent No. 4,5,8,4,5,099. More modernly, thermal inkjet printing has become an extremely popular form of inkjet printing. Inkjet printing techniques include those disclosed in U.S. Patent No. 4,490,728 to Endo et al. The ink jet printing techniques disclosed in the above two patent documents rely on the actuation of an electrothermal actuator which causes a bubble to be generated in a restricted space (for example, a nozzle to cause ink to pass from a connection to the The holes of the restricted space are ejected onto an associated printing device. The printing device using the electrothermal actuator is manufactured by a manufacturer such as Canon and Hewlett Packard. -6- 200918330 There are many different types of prints to use. It is true that a printing technique should have a variety of fertile attributes. These properties include construction and operation are inexpensive, high-speed operation, safety and long-term continuous operation, etc. Each technology is Cost, speed, quality, reliability, power usage, simplicity of construction, durability and consumability have their advantages and disadvantages. Many inkjet printheads use microelectromechanical systems (MEMS) technology. Therefore, they all rely on standard integrated circuit construction/manufacturing techniques that deposit a planar layer on a wafer and etch away portions of the planar layer. Among the circuit manufacturing technologies, some technologies are known. For example, 'compared with other technologies for strange circuit manufacturing (eg, ferroelectric materials, gallium arsenide materials), related to the manufacture of CMOS circuits. The technology of the joint is more commonly used. Therefore, it is better to use the well-proven semiconductor technology in the Μ EMS structure, which does not require any "peculiar," processing or material. Of course, there will be certain The degree of sacrifice, because if the advantages of using strange materials far outweigh its disadvantages, then it will be biased to use strange materials anyway. However, if the same material is used to achieve the same or similar characteristics, then the odd material The problem can be avoided. An important issue in any inkjet printer is printhead maintenance. Printhead maintenance extends the life of a printhead and allows the printhead to be used after an idle period. The typical goal of printhead maintenance is to remove particles from the printhead, remove the ink that floods the surface of the printhead, and unblock the ink. Nozzles blocked by particles. To date, various techniques have been used for the maintenance of print heads, such as suction caps or squeeze scrapers. However, the general problem of printhead maintenance is at the applicant's The page width printhead is degraded. High-density nozzles built on a wafer with useful MEMS technology on the page width print head. Although these print heads are inexpensive to manufacture, they are less expensive than other print heads. Durable, so far, it requires special considerations in the maintenance of the print head. Therefore, the applicant has proposed many novel techniques for print head maintenance, including non-contact maintenance technology. Part of these maintenance techniques It has been embodied in U.S. Patent Application No. 1 1 /246,6 8 8 (issued on October 11, 2005); 11/2/2 46,707 (issued on October 11, 2005); /2 46,693 (as amended on October 11, 2005); 1 1 /246, 6 8 8 (assigned on October 11, 2005); 11/4 82,958 (July 10, 2006) Shen): and 11/2/2 46,688 (July 31, 2006), these The contents of each of the patent applications are referred to in this article by reference. It is desirable to provide a MEMS pagewidth printhead that can withstand the challenges of a variety of printhead maintenance techniques, including contact maintenance techniques. It is even more desirable to provide a MEMS print head with excellent mechanical robustness. It is even more desirable to provide a MEMS printhead that captures the smallest amount of particles and facilitates printhead maintenance. SUMMARY OF THE INVENTION In a first aspect, an ink jet print head is provided that includes a reinforced two-layer nozzle plate structure that spans a plurality of nozzles. Each of the nozzles includes a nozzle. The chamber has a chamber top ' -8 - 200918330. Each chamber roof is defined by a portion of the nozzle plate structure. The upper chamber is selected and the nozzle chamber is formed on a substrate. Each of the nozzle chambers is selected to be spaced apart from the substrate, and the sidewalls extend between the top of the chamber and the substrate. Selecting the upper floor, each chamber has a nozzle hole defined therein. Selecting the upper plate, the nozzle plate structure comprises: a first nozzle plate spanning a plurality of nozzles, the first nozzle plate having a plurality of chambers defined therein; and a photoresist for filling the chambers; A second nozzle plate covers the first nozzle plate and the photoresist. The upper nozzle plate is selected to define an outer surface of a plane of the print head. The upper and second nozzle plates are made of the same or different materials. Selecting the upper layer, the materials are ceramic materials deposited by PECVD. 上 The upper materials are selected from the group consisting of: tantalum nitride, hafnium oxide and hafnium oxynitride. Optionally, each nozzle comprises a nozzle chamber formed on a substrate, the nozzle chamber including a top portion spaced apart from the substrate and a sidewall extending between the top of the chamber and the substrate, wherein The first nozzle plate and the side wall are made of the same material. In a second aspect, an ink jet printhead integrated circuit is provided comprising: -9-200918330 a substrate having a plurality of nozzles formed thereon; a drive circuit electrically coupled to a nozzle associated actuator; and a reinforced two-layer nozzle plate structure that spans the plurality of nozzles. In a third aspect, a method of making an ink jet printhead having a flat nozzle plate is provided, the method comprising the steps of: (a) providing a partially fabricated print head having a a material consisting of a first nozzle plate spanning a plurality of nozzles, the first nozzle plate having a plurality of pockets; (b) filling the chambers with a dip to the upper surface of the first nozzle plate Forming a continuous flat surface with the upper surface of the dip; and (c) depositing a second material on the flat surface to form a second nozzle plate having a flat outer surface. Selecting the upper layer, the second material is deposited by PECVD. In the upper layer, the first material is deposited by PECVD onto a non-planar sacrificial gantry to form the first nozzle plate. The upper and the second materials are the same or different from each other. Preferably, the first and second materials are independently selected from the group consisting of: tantalum nitride, hafnium oxide and hafnium oxynitride. Select the upper layer, the material is a photoresist. Selecting the upper layer, step (b) is carried out using the following sub-steps: (b) (i) depositing a photoresist layer on the first nozzle plate for filling the chambers: and (b) Ii) removing a portion of the photoresist such that an upper surface of the first nozzle plate -10-200918330 and an upper surface of the photoresist that fills the chambers together define a continuous flat surface. The method of selecting the upper layer further comprises the step of: reflowing the photoresist to promote complete filling of the chambers. The upper layer is selected, and the step (b) (ii) is carried out by chemical mechanical planarization or etching with a photoresist. In the above manner, the method further comprises the steps of: (d) defining nozzle holes through the first and second nozzle plates. Optionally, each nozzle comprises a nozzle chamber formed on a substrate, the nozzle chamber including a chamber top spaced apart from the substrate and a sidewall extending between the top of the chamber and the substrate, Wherein the first nozzle plate and the side wall are made of the same material. A printhead in accordance with the present invention includes a plurality of nozzles, and a chamber and an actuator (e.g., a heater element) that typically corresponds to each nozzle. The smallest repeating unit of the printhead has an ink supply inlet that feeds ink to one or more chambers. An entire array of nozzles is formed by repeating these individual cells. This separate unit is collectively referred to herein as a 'unit cell.' A row of print heads may be comprised of a plurality of print head integrated circuits, each of which includes a plurality of nozzles. The term "ink" means any liquid that can be ejected and is not limited to conventional inks containing a coloring dye. Non-colored (11〇11-c ο 1 〇red) inks include fixatives 'infrared absorption' Ink, functional chemicals, adhesives, biological fluids, pharmaceuticals, water and other solvents, etc. -11 - 200918330 The ink or liquid that can be ejected is not necessarily a liquid and can contain particles of the object. [Embodiment] Referring initially to Figure 36, there is shown a cutaway perspective view of a MEMS printhead integrated circuit, as described in U.S. Patent Application Serial No. 11/246,684 (October 11, 2005). The contents of this patent application are hereby incorporated by reference herein to each of the entire disclosures of each of the entire entire entire entire entire entire entire entire entire entire entire entire portion Ink Ports 15. The ink inlets then supply ink to the ink conduits 23 for each column, wherein each nozzle chamber receives ink from a common ink conduit extending longitudinally along each column. A nozzle hole 26 of 25 is defined on a nozzle plate 1〇1 which spans the nozzle row and the nozzle row. As will be described in detail hereinafter, the nozzle plate 101 is a ceramic material (eg, PE) PECVD is formed on a photoresist gantry. Due to this deposition process, the nozzle plate 101 has a plurality of pockets 102 defined therein. The chambers 102 are disposed adjacent nozzles in a row of nozzles. Between these, the chambers 102 are typically several microns deep (e.g., 1 to 5 microns deep) and cause the discontinuous overall effect of the nozzle plate to be a nozzle plate that is substantially non-linear due to the relationship of the chambers 102. Depending on the particular nozzle design and manufacturing process, the chambers 102 can be substantially larger (longer 'wider or deeper') than shown in Figure 36. They can extend between nozzle rows or between rows of nozzles. -12- 200918330 Based on For reasons, the discontinuity and non-planarity of the nozzle plate 101 caused by the chamber 102 is disadvantageous. First, the chamber 102 is a weak point on the nozzle plate 110 and reduces the overall mechanical robustness of the nozzle plate. Especially with regard to the shear applied to the entire nozzle plate. This is particularly important because the wiping action (which is performed during maintenance of certain print heads) sweeping across the entire surface of the nozzle plate 101 can result in considerable Large shear forces. Secondly, the chambers 102 can easily capture ink and/or particles, and then the inks and/or particles are difficult to remove. The proximity of the chamber 102 to the nozzle holes 26 is particularly undesirable because of any The captured particles are likely to block the nozzle and affect the print quality. In order to have a complete understanding of the present invention, what is described below is how the printhead integrated circuit of Figure 36 is fabricated using a MEMS process. Further, another process according to the present invention is further provided below, in which the flatness of the nozzle plate 110 is remarkably improved. MEMS Process This MEMS process builds a nozzle structure on a wafer after CMOS processing. Figure 2 is a perspective view of a nozzle unit cell after completion of CMOS processing and prior to MEMS processing. During the CMOS processing of the wafer, four metal layers are deposited on a wafer 2, wherein the metal layers are interspersed between layers of interlayer dielectric (ILD) layers. The four metal layers are referred to as M1, M2, M3, and M4 layers and are sequentially constructed on the wafer during CMOS processing. These CMOS layers provide all of the driver and logic circuitry used to operate the printhead. -13- 200918330 In the completed print head, each heater element is assigned to a pair of electrodes on the outermost M4 layer, the M4 COMS layer is the subsequent MEMS of the wafer, and the M4 layer is also defined A pad of the integrated circuit along each column of the print head. These bonding pads (not shown) allow the CMOS to extend out of the wire bond to be connected to a microprocessor. Figures 1 and 2 show an aluminum M4 layer 3 having a layer thereon. (Only the MEMS features of the M4 layer are shown; the main CMOS features of the M4 layer are placed outside.) The M4 layer 3 has a thickness of 1 micron and is itself thick on the CVD oxide layer 5. As shown in Figures 1 and 2, there is an ink inlet opening 6 and a pit opening 7. These adhesive pads which are formed in the MEMS process and which are formed at the longitudinal edges of the print head integrated circuit at the beginning of the MEMS process of the cell 11 are defined by the etch layer 4. This hungry exposure at the position of the bonding pad is that the photoresist used in this step is completely affected by the etching. Turning to Figures 3 through 5, the MEMS processed first stage pit 8 penetrates the passive layer 4 and the CVD oxide layer 5. A photoresist layer (not shown) is used which is exposed by a dim mask. The pit 8 has a degree which is a depth measured from the top end of the M4 layer 3. At the same time as 8 , the electrode 9 passes through the CMOS by passing the M4 layer 3 through the passive actuator. Due to the basis of processing. The adhesion of the longitudinal edges is formed in the two figures by the adhesion layer 4 from the bonding pad. The nozzle unit cells are deposited at 2 μm 5 and the M4 layer 3 ports define the subsequent positions. Before, the passive M4 layer 3 is penetrated along each moment. The mask is covered, so that it is not etched by a recess. This etching is used in one of the depths of 2 micrometers in Fig. 3. The layer 4 is etched and partially exposed at -14-200918330 to be formed on both sides of the pit. In the completed nozzle, a heater element is suspended between the electrodes 9 on both sides of the pit 8. In the next step (Figs. 6 to 8), the pit 8 is etched with the first sacrificial layer ("SAC 1") of the photoresist 10. A 2 micron thick high viscosity photoresist is first applied to the wafer and then exposed using a dark tone pit mask as shown in FIG. The SAC 1 photoresist 10 forms a shelf for subsequent deposition of heater material across the electrodes 9 on either side of the pit 8. Therefore, it is important that the SAC1 photoresist 10 has a flat upper surface that is flush with the upper surface of the electrode 9. At the same time, the SAC1 photoresist 10 must completely fill the pits 8 to avoid "stringers" of conductive heater material extending across the pits and to short the electrodes 9. Typically, when a trench is filled with a photoresist, it is necessary to expose the photoresist to the outside of the trench to ensure that the photoresist is filled to the sidewall of the trench, thereby avoiding the subsequent deposition step. "Stringer". However, this technique causes the photoresist to have a ridged edge near the perimeter of the trench. This is undesirable because in the subsequent deposition step, the material will be deposited unevenly on the edge of the ridge, and the vertical or angled surface of the ridged edge will receive less deposited material than 塡A flat surface that receives the photoresist of the ditch receives the deposited material. This result is a "resistance hot spot" in the area where the material is deposited very thinly. As shown in Fig. 7, the current process deliberately exposes the SAC1 photoresist 10 (e.g., within 0.5 micrometer) inside the peripheral wall of the pit 8 using the photomask shown in Fig. 6. This ensures the upper surface of a flat SAC1 photoresist 10 and avoids the photoresist ridges in the vicinity of the peripheral edge of the pit 8. -15- 200918330 After exposure of the SAC1 photoresist 10, the photoresist is advected by heating. Advection of the photoresist allows the photoresist to flow to the sidewalls of the recess 8 to completely fill the pits 8. Figures 9 and 10 show the SAC1 photoresist 10 after advection. The photoresist has a flat upper surface and is flush with the upper surface of the M4 layer 3 forming the electrode 9. After advection, the SAC1 photoresist 10 is hardened and/or baked by U.V. to avoid any advection during the subsequent deposition step of the heater material. Figures 11 and 12 show unit cells after a 0.5 micron heater substance 11 is deposited onto the SAC1 photoresist 10. Due to the above-described advection treatment, the heater substance 11 is uniformly deposited as a flat layer on the electrode 9 and the SAC1 photoresist 10. The heater material may be composed of any suitable conductive material such as TiAl, TiN, TiAIN, TiAlSiN or the like. A typical heater material deposition process can include a 100A TiAl seed layer, a 2500A TiAIN layer, and another 100A
TiAl種子層與最終之另一 2500A之TiAIN層的依續沉積 〇 參照圖13至15，在下面的步驟中，該加熱器物質層 11被蝕刻用以界定該熱致動器12。每一致動器12都具有 接點28其對該SAC 1光阻劑1 〇兩側上的各別電極9建立 一電連接。一加熱器元件29橫跨在相應的接點28之間。 此蝕刻係由一使用圖13所示的暗色調光罩曝光之光 阻劑層（未示出）來界定。如圖1 5所示，該加熱器元件！ 2 爲一直線樑其橫跨於該對電極9之間。然而，該加熱器元 件12可以採用其它的結構，譬如描述於本案申請人所擁 -16- 200918330 有之美國專利第6,75 5,5 09號中所描述的結構，該專利內 容藉由此參照被倂於本文中。 在接下來的步驟程序中，一用於噴嘴的墨水入口被蝕 刻穿透該被動層4，該氧化物層5及該矽晶圓2。在 CMOS處理期間，每一金屬層都具有一在墨水入口蝕刻處 理中被鈾刻穿透金屬層本身之墨水入口開口（參見圖1中 M4層3上的開口 6)。這些金屬層與散置的ILD層一起形 成用於該墨水入口的密封環，用來防止墨水滲出到C Μ Ο S 層中。 參照圖1 6至1 8，一相對厚的光阻層1 3被旋施於該晶 圓上且使用圖1 6所示之暗色調光罩加以曝光。該光阻劑 1 3的厚度將視用來蝕刻該墨水入口之深反應離子蝕刻 (DRIE)的選擇性而定。在一墨水入口開口 14被形成在該 光阻劑1 3上之後’該晶圓即已準備好進行後續的蝕刻步 驟。 在第一蝕刻步驟中（圖19及20)，介電層（被動層4與 氧化物層5 )被蝕刻穿透到達底下的矽晶圓。任何標準的氧 化物蝕刻（如’ 〇2/C4F8電漿）都可被使用。 在第二蝕刻步驟中（圖21及22)，—墨水入口 15藉由 使用同一光阻光罩1 3而被蝕刻至該矽晶圓2中達到2 5微 米的深度。任何標準的非等方向性DRIE都可被用來實施 此餓刻’譬如Bosch蝕刻（參見美國專利第6,5 0 1,893號及 第 6,284，148 號）。 在下一個步驟中，該墨水入口 15被塞入光阻劑及一 -17- 200918330 光阻劑16的第二犧牲層（“SAC2”）被近構於該SAC1光阻 劑10與該被動層14的上方。該SAC2光阻劑16將作爲 室頂物質之後續沉積的台架’其形成每一噴嘴室的頂及側 壁。參照圖2 3至2 5，一約6微米的高黏性光阻劑層被旋 施於該晶圓上且使用圖23所示的暗色調光罩加以曝光。 如圖23及25所示，該光罩露出在該SAC2光阻劑16 上對應於室側壁及墨水導管側壁的位置的側壁開口 1 7。此 外，開口 18及19被露出且分別與被塡塞的入口 15及噴 嘴室入口相鄰。這些入口 18及19在後續的室頂沉積步驟 中將被塡入室頂物質並在此噴嘴設計中提供獨特的優點。 詳言之，被塡入室頂物質之開口 18具有一啓始的特徵結 構般的功能，用以幫助將墨水從入口 1 5吸到每一噴嘴室 內。被塡入室頂物質的開口 1 9如瀘波器結構及流體串音 (cross talk)阻擋器般作用。這些都有助於防止氣泡進入到 噴嘴室中及防止熱致動器12產生擴散壓脈衝。 參照圖26及27，下一步驟用PECVD沉積3微米的室 頂物質20於該SAC2光阻劑16上。室頂物質20塡滿在 該SAC2光阻劑16上的開口 17，18及19用以形成具有 室頂2 1及側壁22的噴嘴室24。一用來供應墨水至每一噴 嘴室的墨水導管亦在該室頂物質20的沉積期間被形成。 此外，任何基礎特徵及濾波器結構（未示於圖26及27中） 都在同一時間被形成。室頂2 1 (每一室頂都對應到一各的 噴嘴室24)橫跨在同一列上之相鄰的噴嘴室用以形成一噴 嘴板。該室頂物質20可用任何適當的物質來組成，譬如 -18- 200918330 氮化矽’氧化矽，氮氧化矽，氮化鋁等等。如上文中討論 過的’噴嘴板101在介於噴嘴之間的區域處具有穴室102( 示於圖3 6中）。 參照圖2 8至3 0，下一階段藉由將2微米的室頂物質 20蝕刻掉來界定在室頂21上的橢圓形噴嘴邊緣25。此蝕 刻係使用一層以示於圖28中之暗色調的光罩曝光過的光 阻層（未示出）來界定的。該橢圓形邊緣25包含兩個同軸的 邊唇25a及25b，其位於它們各自的熱致動器12上。 參照圖3 1至3 3，下一階段藉由整個蝕刻穿透剩下的 室頂物質20來界定該橢圓形噴嘴孔26於該室頂21上， 該噴嘴孔係被邊圓25圍起來。此蝕刻係使用一層以示於 圖31中之暗色調的光罩曝光過的光阻層（未示出）來界定的 。該橢圓形噴嘴孔26如圖33所示係被設置在該熱致動器 12上方。 所有的MEMS噴嘴特徵現已完全被形成，後續的階段 爲以被側DRIE界定供墨通道27，用02電漿灰化（ashing) 來去除掉所有犧牲性光阻劑（包括SAC 1與SAC2光阻劑層 1 0及1 6在內），及用背側蝕刻將晶圓薄化約1 3 5微米。圖 34及35顯示完成的單元細胞，而圖36則以該完成的列印 頭積體電路的切開立體圖形式顯示三個相鄰的噴嘴列。 另一提供平的噴嘴板之MEMS製程 上文所述之MEMS製程的優點之一爲噴嘴板101以 PECVD力口以沉積。此表示該噴嘴丰反的製造可被加人到一使 -19- 200918330 用標準的COMS沉積/蝕刻技術的MEMS製程中。因此， 該列印頭的整體製造成本可被保持的很低。相反地，許多 前技列印頭具有層疊的噴嘴板，其不只有層與層分離的疑 慮，其還需要有一分離的層疊步驟來實施標準CMOS處理 。而這最終將提高此類列印頭的成本。 然而，噴嘴板的PECVD沉積亦具有它本身的挑戰。 沉積一具有足夠厚度的室頂物質（如，氮化矽）使得噴嘴板 不會太過脆弱是很重要的。當沉積在平的結構上時，任積 並不是問題，然而，從圖24-27可清楚地瞭解到，室頂物 質20的沉積亦必需形成噴嘴室24的側壁22。SAC2光阻 劑16可具有斜的壁（未示於圖24中）用以幫助室頂物物質 沉積至側壁區域17中。然而，爲了要確保室側壁22接受 到足夠室頂物質2 0的覆蓋，在相鄰的噴嘴之間必需要有 至少一些間隔。雖然從室頂沉積的角度來看此噴嘴間的間 隔是有利的’但是所得到的室頂21 (及噴嘴板1〇1)無可避 免地在噴嘴之間包含多個穴室1 〇 2。如已經討論過的’這 些穴室1 02會像是陷阱一般地捕捉住微粒及溢流的墨水， 並因而防礙列印頭的保養。 現參照圖37至40，其顯示另一種MEMS製程’該製 程將上述的一些問題減至最小。在圖26及27所示的列印 頭製造階段中，該室頂21(其形成該噴嘴板1〇1)先被平坦 化，而不是在其前面先實施噴嘴邊緣及噴嘴孔的實施。平 坦化是階5由沉積一額外的光阻劑層（約1 〇微米厚）於該室 頂2 1上，其會塡滿所有的穴室1 02。典型地’此光阻劑然 -20- 200918330 後被熱平流用以確保該等穴室102被完全地塡滿。該光阻 劑層然後被去除到該室頂2 1的高度，使得該室頂2 1的上 表面與沉積在穴室102內的光阻劑1〇3的上表面共同形成 一連續的平的表面。光阻劑的去除可藉由任何適當的技術 來實施，譬如化學機械硏磨（C Μ P)或控制下的光阻劑蝕刻（ 如’ 〇2電漿）。如圖3 7所示，所得到的單元細胞具有完全 塡滿穴室102的光阻劑103。 下一個階段係藉由PECVD將額外的室頂物質（如，一 微米厚的層）沉積到圖3 7所示之該平的結構上。如圖3 8 及39所示，該所得到的單元細胞具有一第一室頂21Α及 一第二室頂21Β。很重要的是，在外面的該第二室頂21Β 由於它沉積在一平的結構上所以它是完全平的。再者，該 第二室頂21Β被塡充至該第一室頂21Α的穴室1〇2內之底 下的光阻劑103強化。 此經過強化的雙層式室頂結構與圖2 7所示的單一室 頂結構相比自機械強度上要強健很多。此增加的厚度與噴 嘴間的強化可改善該室頂結構的一般強健度。又，外面的 第二室頂2 1 Β的平坦性可提供相對於該室頂之剪力的強健 度。 該第一及第二室頂21Α及21Β可用相同或不同的物質 來形成。典型地，該第一及第二室頂是用獨立地選自於包 含：氮化矽，氧化矽及氮氧化矽的組群中的物質組成的。 在一實施例中’該第一室頂21Α是由氮化砍組成的及k桌 二室頂2 1 B是由氧化矽組成的。 -21 - 200918330 從圖3 8及3 9所示的單元細胞接下來的是’後續的 MEMS處理可類似地進行配合圖28至36描述的相應步驟 上。因此，噴嘴邊緣及噴嘴孔蝕刻被實施，之後接著的是 背側DRIE用以界定供墨通道27，晶圓薄化與光阻劑去除 。當然，被該第一及第二室頂21A與21B所包圍起來的光 阻劑1 03不會曝露在任何灰化電漿下且在最後階段的光阻 劑去除期間保持未受損傷。 所得到之具有一平的雙層式強化噴嘴板之列印頭積體 電路被示於圖40中。該噴嘴板包含一第一噴嘴板l〇lA與 一外面的第二噴嘴板101B，它是完全平的，保留給噴嘴 邊緣及噴嘴孔。依據本發明的此列印頭積體電路便於列印 頭的保養作業。其改良的機械完整性意謂著可使用相對強 硬的清潔技術（如，擦拭）而不會傷到該列印頭。再者，在 該外面的第二噴嘴板102B上沒有穴室102可將微粒或墨 水被永久性地包陷在該列印頭上的風險降至最低。 當然，將可被瞭解的是，本發明已單純用舉例的方式 加以描述且在細節上的變化可在本發明的範圍內被達成， 本發明的範圍是由下面的申請專利範圍來加以界定的。 【圖式簡單說明】 本發明的實施例現將以舉例的方式參照附圖加以描述 ，其中： 圖1顯示依據本發明之在一列印頭上的MEMS噴嘴陣 列的部分製作好的單元細胞，該單元細胞係沿著圖3的 -22- 200918330 A - A線被剖開； 圖2顯示圖1中部分作好的單元細胞的立體圖； 圖3顯示與每一加熱器元件溝渠相關聯的光罩； 圖4爲該單元細胞在該溝槽蝕刻之後的剖面圖； 圖5爲示於圖4中之單元細胞的立體圖； 圖6爲與示於圖7中之犧牲性光阻的沉積相關聯的光 罩； 圖7顯示在該犧牲性光阻劑溝渠的沉積之後該單元細 胞’的遺中介於該犧牲性物質與該溝渠的側壁的邊緣之間 的間隙被放大； 圖8爲示於圖7中之單元細胞的立體圖； 圖9顯示用於閉合沿著溝渠的側壁之間隙的該犧牲性 光阻劑平流之後的單元細胞； 圖10爲示於圖9中之單元細胞的立體圖； 圖1 1爲一剖面圖其顯示該加熱器材料層的沉積； 圖12爲示於圖11中之單元細胞的立體圖； 圖13顯示圖14中之加熱器物質的金屬蝕刻相關聯的 光罩； 圖14爲一剖面圖其顯示用來昔塑加熱器致動器的金 屬蝕刻； 圖1 5爲示於圖1 4中之單元細胞的立體圖； 圖16顯示與示於圖17中之鈾刻相關聯的光罩； 圖17顯示該光阻劑層的沉積及後續到達該CMOS驅 動層的頂部之被動層之墨水入口的鈾刻； -23- 200918330 圖18爲示於圖17中之單元細胞的立體圖； 圖19顯示穿透該被動層與CMOS層到達底下的矽晶 圓之氧化物蝕刻； 圖20爲示於圖19中之單元細胞的立體圖； 圖21爲進入到該矽晶圓內之該墨水入口的深非等方 向性鈾刻； 圖22爲示於圖21中之單元細胞的立體圖； 圖2 3顯示與示於圖24中之光阻鈾刻相關聯的光罩； 圖24顯示該光阻劑蝕刻用以形成供室等與側壁用的 開口； 圖25爲示於圖24中之單元細胞的立體圖； 圖26顯示側壁及風險物質（risk material)的沉積； 圖27爲示於圖26中之單元細胞的立體圖； 圖2 8爲與圖2 9中之噴嘴緣蝕刻相關聯的光罩； 圖2 9顯示室頂層的鈾刻用以形成該噴嘴孔邊圓； 圖30爲示於圖29中之單元細胞的立體圖； 圖31爲與圖32中之噴嘴孔蝕刻相關聯的光罩； 圖3 2顯示室頂物質的触刻用以形成橢圓形噴嘴孔； 圖33爲示於圖32中之單元細胞的立體圖； 圖3 4顯示在背側蝕刻，電漿灰化及晶圓薄化之後的 單元細胞； 圖35爲示於圖34中之單元細胞的立體圖；及 圖3 6爲在一列印頭積體電路上的一噴嘴陣列的切開 立體圖； -24- 200918330 圖3 7爲圖27所示的單元細胞在穴室塡充之後的立體 圖； 圖3 8爲圖3 7所示的單元細胞在第二室頂沉積之後的 側視圖； 圖39爲示於圖38中之單元細胞的立體圖；及 圖40爲具有強化的雙層式噴嘴板之列印頭積體電路 的切開立體圖。 【主要元件符號說明】 27 :供墨通道 1 5 :入口孔 2 3 :墨水導管 2 6 :噴嘴孔 25 :噴嘴環 1 0 1 :噴嘴板 1 02 :穴室 2 :晶圓 100 :噴嘴單元細胞Subsequent deposition of the TiAl seed layer and finally another 2500A TiAIN layer 〇 Referring to Figures 13 through 15, in the next step, the heater material layer 11 is etched to define the thermal actuator 12. Each actuator 12 has a contact 28 that establishes an electrical connection to the respective electrodes 9 on either side of the SAC 1 photoresist 1 . A heater element 29 spans between the respective contacts 28. This etching is defined by a photoresist layer (not shown) exposed using a dark-tone mask as shown in FIG. As shown in Figure 15, the heater element! 2 is a straight beam which spans between the pair of electrodes 9. However, the heater element 12 can be constructed in other configurations, such as those described in U.S. Patent No. 6,75,5,09, the entire disclosure of which is incorporated by reference. References are cited in this article. In the next step of the procedure, an ink inlet for the nozzle is etched through the passive layer 4, the oxide layer 5 and the germanium wafer 2. During the CMOS process, each metal layer has an ink inlet opening that is etched through the metal layer itself by uranium during the ink inlet etching process (see opening 6 in layer M4 of Figure 1). These metal layers together with the interspersed ILD layer form a seal ring for the ink inlet to prevent ink from oozing out into the C Μ Ο S layer. Referring to Figures 16 to 18, a relatively thick photoresist layer 13 is spin-applied to the crystal and exposed using a dark-tone mask as shown in Figure 16. The thickness of the photoresist 13 will depend on the selectivity of the deep reactive ion etching (DRIE) used to etch the ink inlet. After an ink inlet opening 14 is formed on the photoresist 13, the wafer is ready for subsequent etching steps. In the first etching step (Figs. 19 and 20), the dielectric layer (passive layer 4 and oxide layer 5) is etched through to the underlying germanium wafer. Any standard oxide etch (such as '’2/C4F8 plasma) can be used. In the second etching step (Figs. 21 and 22), the ink inlet 15 is etched into the silicon wafer 2 to a depth of 25 μm by using the same photoresist mask 13. Any standard non-isotropic DRIE can be used to perform this hungry, such as Bosch etching (see U.S. Patent Nos. 6,500,893 and 6,284,148). In the next step, the ink inlet 15 is inserted into the photoresist and a second sacrificial layer ("SAC2") of a -17-200918330 photoresist 16 is proximally applied to the SAC1 photoresist 10 and the passive layer 14. Above. The SAC2 photoresist 16 will serve as a shelf for subsequent deposition of the roofing material 'which forms the top and side walls of each nozzle chamber. Referring to Figures 23 to 25, a layer of high viscosity photoresist of about 6 microns is applied to the wafer and exposed using a dark tone mask as shown in Figure 23. As shown in Figures 23 and 25, the mask exposes a sidewall opening 17 on the SAC2 photoresist 16 corresponding to the sidewalls of the chamber and the side walls of the ink conduit. In addition, the openings 18 and 19 are exposed and adjacent to the dammed inlet 15 and the nozzle chamber inlet, respectively. These inlets 18 and 19 will be forced into the roofing material during subsequent chamber top deposition steps and provide unique advantages in this nozzle design. In particular, the opening 18 that is inserted into the top material has an initial feature-like function to help draw ink from the inlet 15 into each nozzle chamber. The opening to the top of the chamber material 1 9 acts as a chopper structure and a fluid cross talk blocker. These all help to prevent air bubbles from entering the nozzle chamber and preventing the thermal actuator 12 from generating a diffusion pressure pulse. Referring to Figures 26 and 27, the next step is to deposit 3 microns of the topping material 20 onto the SAC2 photoresist 16 by PECVD. The openings 18, 19 and 19, which are filled with the top material 20 on the SAC2 photoresist 16, are used to form the nozzle chamber 24 having the chamber top 21 and the side walls 22. An ink conduit for supplying ink to each of the nozzle chambers is also formed during deposition of the chamber top material 20. In addition, any of the basic features and filter structures (not shown in Figures 26 and 27) are formed at the same time. The roofs 2 1 (each of which corresponds to a respective nozzle chamber 24) span adjacent nozzle chambers in the same column for forming a nozzle plate. The top material 20 can be composed of any suitable material, such as -18-200918330 tantalum nitride yttrium oxide, yttrium oxynitride, aluminum nitride, and the like. The nozzle plate 101, as discussed above, has a pocket 102 (shown in Figure 36) at the area between the nozzles. Referring to Figures 28 through 30, the next stage defines an elliptical nozzle edge 25 on the chamber roof 21 by etching away the 2 micron ceiling material 20. This etch is defined by a layer of photoresist (not shown) exposed by a visor shown in Fig. 28 in dark tones. The elliptical edge 25 includes two coaxial lips 25a and 25b that are located on their respective thermal actuators 12. Referring to Figures 31 to 3, the next stage defines the elliptical nozzle aperture 26 on the chamber top 21 by the entire etch through the remaining top material 20 which is surrounded by the rim 25. This etching is defined by a photoresist layer (not shown) exposed by a mask of the dark color shown in Fig. 31. The elliptical nozzle hole 26 is disposed above the thermal actuator 12 as shown in FIG. All MEMS nozzle features are now fully formed, with subsequent stages defining the ink supply channel 27 by side DRIE and removing all sacrificial photoresists (including SAC 1 and SAC2 light) with 02 plasma ashing. The resist layers 10 and 16 are included, and the wafer is thinned by a backside etch of about 135 microns. Figures 34 and 35 show the completed unit cells, and Figure 36 shows three adjacent nozzle rows in the form of a cutaway perspective view of the completed print head integrated circuit. Another MEMS process that provides a flat nozzle plate One of the advantages of the MEMS process described above is that the nozzle plate 101 is deposited with a PECVD force. This means that the fabrication of the nozzle can be added to a MEMS process using the standard COMS deposition/etching technique of -19-200918330. Therefore, the overall manufacturing cost of the print head can be kept low. Conversely, many front-end printheads have stacked nozzle plates that are not only layer-to-layer separation concerns, but also require a separate lamination step to implement standard CMOS processing. This will ultimately increase the cost of such print heads. However, PECVD deposition of nozzle plates also has its own challenges. It is important to deposit a ceiling material of sufficient thickness (e.g., tantalum nitride) so that the nozzle plate is not too weak. When deposited on a flat structure, any product is not an issue, however, as is clear from Figures 24-27, the deposition of the roofing material 20 must also form the sidewalls 22 of the nozzle chamber 24. The SAC2 photoresist 16 can have a sloped wall (not shown in Figure 24) to aid in the deposition of the topping material into the sidewall region 17. However, in order to ensure that the chamber sidewalls 22 receive sufficient coverage of the roofing material 20, at least some spacing between adjacent nozzles is required. Although the spacing between the nozzles is advantageous from the viewpoint of deposition at the top of the chamber, the resulting chamber top 21 (and the nozzle plate 1〇1) inevitably includes a plurality of chambers 1 〇 2 between the nozzles. As already discussed, these chambers 102 will trap particles and overflow ink like traps and thus prevent the maintenance of the print head. Referring now to Figures 37 through 40, another MEMS process is shown which minimizes some of the above problems. In the stage of the print head manufacturing shown in Figs. 26 and 27, the chamber top 21 (which forms the nozzle plate 1〇1) is first flattened instead of the nozzle edge and the nozzle hole being implemented in front of it. The flattening is that the order 5 consists of depositing an additional layer of photoresist (about 1 〇 microns thick) on the top 2 of the chamber which fills all of the chambers 102. Typically, this photoresist is tempered by -20-200918330 to ensure that the chambers 102 are completely filled. The photoresist layer is then removed to the height of the chamber top 21 such that the upper surface of the chamber top 21 forms a continuous flat with the upper surface of the photoresist 1〇3 deposited in the chamber 102. surface. The removal of the photoresist can be carried out by any suitable technique, such as chemical mechanical honing (C Μ P) or controlled photoresist etching (e.g., '〇2 plasma). As shown in Fig. 37, the resulting unit cells have a photoresist 103 which is completely filled with the chamber 102. The next stage is to deposit additional roofing material (e.g., a one micron thick layer) by PECVD onto the flat structure shown in Fig. 37. As shown in Figures 38 and 39, the resulting unit cells have a first chamber top 21Α and a second chamber top 21Β. It is important that the second chamber top 21 on the outside is completely flat due to its deposition on a flat structure. Further, the second chamber top 21 is reinforced by the photoresist 103 which is submerged under the chamber 1〇2 of the first chamber top 21Α. This reinforced double-layered roof structure is much more mechanically strong than the single roof structure shown in Figure 27. This increased thickness and reinforcement between the nozzles improves the general robustness of the roof structure. Moreover, the flatness of the outer second chamber top 2 1 可 provides robustness against the shear of the roof. The first and second chamber tops 21 and 21 can be formed of the same or different materials. Typically, the first and second chamber tops are comprised of materials independently selected from the group consisting of: tantalum nitride, cerium oxide, and cerium oxynitride. In one embodiment, the first chamber top 21 is composed of nitrided chopped and the k-table two-chamber top 2 1 B is composed of tantalum oxide. -21 - 200918330 From the unit cells shown in Figures 38 and 39, the subsequent MEMS processing can be similarly performed in the corresponding steps described in connection with Figures 28 through 36. Thus, nozzle edge and nozzle hole etching is performed, followed by backside DRIE to define ink supply channel 27, wafer thinning and photoresist removal. Of course, the photoresist 101 surrounded by the first and second chamber tops 21A and 21B is not exposed to any ashing plasma and remains unimpeded during the final stage of photoresist removal. The resulting print head integrated circuit having a flat double-layered reinforced nozzle plate is shown in FIG. The nozzle plate includes a first nozzle plate 101A and an outer second nozzle plate 101B which are completely flat and are reserved for the nozzle edge and the nozzle hole. This print head integrated circuit according to the present invention facilitates the maintenance work of the print head. Its improved mechanical integrity means that relatively strong cleaning techniques (eg, wiping) can be used without damaging the print head. Moreover, the absence of pockets 102 on the outer second nozzle plate 102B minimizes the risk of particulate or ink being permanently trapped on the printhead. Of course, it is to be understood that the invention has been described by way of example only and the details of the invention may be . BRIEF DESCRIPTION OF THE DRAWINGS [0007] Embodiments of the invention will now be described by way of example with reference to the accompanying drawings, in which: FIG. 1 shows a partially constituting unit cell of a MEMS nozzle array on a column of print heads in accordance with the present invention, the unit The cell line is sectioned along line -22-200918330 A-A of Figure 3; Figure 2 shows a perspective view of a portion of the cell cells of Figure 1; Figure 3 shows the mask associated with each heater element trench; 4 is a cross-sectional view of the unit cell after etching the trench; FIG. 5 is a perspective view of the unit cell shown in FIG. 4; and FIG. 6 is a light associated with deposition of the sacrificial photoresist shown in FIG. Shield; Figure 7 shows that after the deposition of the sacrificial photoresist trench, the gap between the sacrificial material and the edge of the sidewall of the trench is magnified; Figure 8 is shown in Figure 7. Figure 7 shows a unit cell after the advection of the sacrificial photoresist along the gap of the side wall of the trench; Figure 10 is a perspective view of the unit cell shown in Figure 9; a cross-sectional view showing the plus Figure 12 is a perspective view of the unit cell shown in Figure 11; Figure 13 is a photomask associated with the metal etching of the heater material of Figure 14; Figure 14 is a cross-sectional view showing the same Metal etch of the plastic heater actuator; Fig. 15 is a perspective view of the unit cell shown in Fig. 14; Fig. 16 shows the reticle associated with the uranium engraving shown in Fig. 17; Fig. 17 shows the photoresist Deposition of the agent layer and subsequent uranium engraving of the ink inlet to the passive layer at the top of the CMOS driver layer; -23- 200918330 Figure 18 is a perspective view of the cell shown in Figure 17; Figure 19 shows the penetration of the passive layer and The CMOS layer reaches the oxide etch of the underlying germanium wafer; FIG. 20 is a perspective view of the cell shown in FIG. 19; FIG. 21 is a deep non-isotropic uranium engraved into the ink inlet into the germanium wafer; Figure 22 is a perspective view of the unit cell shown in Figure 21; Figure 2 3 shows the photomask associated with the photoresist uranium engraving shown in Figure 24; Figure 24 shows the photoresist etching used to form a chamber, etc. Figure 25 is a perspective view of the unit cells shown in Figure 24; Figure 26 shows the deposition of sidewalls and risk material; Figure 27 is a perspective view of the unit cells shown in Figure 26; Figure 28 is a mask associated with the nozzle edge etching of Figure 29; Figure 2 9 The uranium engraved on the top of the display chamber is used to form the nozzle hole circle; Fig. 30 is a perspective view of the unit cell shown in Fig. 29; Fig. 31 is a reticle associated with the nozzle hole etching in Fig. 32; Fig. 3 2 shows The contact of the top material is used to form an elliptical nozzle hole; Fig. 33 is a perspective view of the unit cell shown in Fig. 32; Fig. 34 shows the unit cell after back side etching, plasma ashing and wafer thinning. Figure 35 is a perspective view of the cell shown in Figure 34; and Figure 36 is a cutaway perspective view of a nozzle array on a column of integrator circuits; -24- 200918330 Figure 37 is the cell shown in Figure 27. 3 is a side view of the cell shown in FIG. 37 after deposition on the top of the second chamber; FIG. 39 is a perspective view of the unit cell shown in FIG. 38; Cutaway perspective view of a printhead integrated circuit with a reinforced double-layer nozzle plate . [Main component symbol description] 27: Ink supply channel 1 5 : Inlet hole 2 3 : Ink duct 2 6 : Nozzle hole 25 : Nozzle ring 1 0 1 : Nozzle plate 1 02 : Chamber 2 : Wafer 100 : Nozzle unit cell
Ml :金屬層 M2 :金屬層 M3 :金屬層 M4 :金屬層 4 :被動層 6 :墨水入口開口 -25 - 200918330 7 :凹坑開口 8 :凹坑 3 : M3 層 9 :電極 10 ： SAC1 光阻劑 11 :加熱器金屬 1 2 :熱致動器 1 3 :光阻劑 2 8 :接點 29 :加熱器元件 1 4 :墨水入口孔 1 5 :墨水入口 1 6 : S A C 2光阻劑 1 7 :側壁開口 1 8 :開口 1 9 :開口 20 :室頂物質 2 1 :室頂 22 :側壁 2 3 :墨水導管 2 4 :噴嘴室 25a :邊緣唇 2 5 b :邊緣唇 2 5 :噴嘴邊緣 -26 200918330 2 6 :噴嘴孔 16 :台架 2 1 A :第一室頂 2 1 B :第二室頂 1 〇 3 :底下的光阻劑 27 :供墨通道 -27Ml : metal layer M2 : metal layer M3 : metal layer M4 : metal layer 4 : passive layer 6 : ink inlet opening - 25 - 200918330 7 : pit opening 8 : pit 3 : M3 layer 9 : electrode 10 : SAC1 photoresist Agent 11: heater metal 1 2 : thermal actuator 1 3 : photoresist 2 8 : contact 29 : heater element 1 4 : ink inlet hole 1 5 : ink inlet 1 6 : SAC 2 photoresist 1 7 : Side opening 1 8 : Opening 1 9 : Opening 20 : Roof material 2 1 : Room top 22 : Side wall 2 3 : Ink duct 2 4 : Nozzle chamber 25a : Edge lip 2 5 b : Edge lip 2 5 : Nozzle edge - 26 200918330 2 6 : Nozzle hole 16 : gantry 2 1 A : first chamber top 2 1 B : second chamber top 1 〇 3 : bottom photoresist 27 : ink supply channel -27