201107145 六、發明說明 【發明所屬之技術領域】 本發明關於微機電系統（MEMS )裝置的領域，特別 是關於噴墨列印頭。硏發該等噴墨列印頭主要用於改善熱 彎曲致動器在微機電系統製造期間和作業期間的堅固耐用 性。 【先前技術】 本案申請人先前已描述使用熱彎曲致動之微機電系統 噴墨噴嘴的噴墨過多之問題。熱彎曲致動通常意指由電流 通過一種材料，然後該材料的熱膨脹相對於另一材料所產 生的彎曲運動。結果的彎曲運動可用於從噴嘴的開口噴射 墨水，選擇性地經由輪葉或葉片的運動來噴射墨水，因爲 該運動會在噴嘴腔室內產生壓力波。 本案申請人的第US64 1 6 1 67號美國專利（其內容倂 入本案做參考）描述的噴墨噴嘴，具有設於噴嘴腔室內的 輪葉和設於噴嘴腔室外部的熱彎曲致動器。致動器採取傳 導材料之下主動樑（例如氮化鈦）融合至非傳導材料之上 被動樑（例如二氧化矽）的形式。致動器經由一臂部而連 接至輪葉，該臂部容置並穿過在噴嘴腔室之壁中的槽。當 電流通過下主動樑時，致動器向下彎曲，結果輪葉朝向噴 嘴開口運動，該噴嘴開口界定在噴嘴腔室的頂部，藉此噴 射墨水液滴。此設計的優點是其結構的簡單性。此設計的 缺點在於：輪葉的兩面須對抗噴嘴腔室內側相對黏性的墨 -5- 201107145 水而工作。 申請人之第US626095 3號美國專利（茲將其內容倂 入本案作參考）描述一種噴墨噴嘴，其中的致動器形成噴 嘴腔室的運動頂部。致動器採用聚合體材料包覆傳導材料 之螺旋形芯部的形式。當致動時，致動器朝向噴嘴腔室的 底部彎曲，增加腔室內的壓力，並迫使墨水液滴流出噴嘴 開口，該噴嘴開口界定在腔室的頂部中。該噴嘴開口界定 在該腔室頂部之非運動部分中。此設計的優點是運動頂部 的一面必須對抗噴嘴腔室內側相對黏性的墨水工作。此設 計的缺點在於：聚合體材料包覆螺旋形傳導性元件之致動 器構造，難以達成微機電系統的製程。 申請人的第US6 62 3101號美國專利（茲將其內容倂 入本案作參考）描述一種噴墨噴嘴，其包含具有可運動頂 部的噴嘴腔室，該頂部具有界定在其內的噴嘴開口。可運 動頂部經由一臂部連接至設置在噴嘴腔室外部的熱彎曲致 動器。致動器採用上主動樑和下被動樑間隔開的形式。藉 由將主動樑和被動樑間隔開，熱彎曲效率會最大化，因爲 被動樑不能做爲主動樑的散熱器。當電流一通過主動樑 時，具有噴嘴開口界定在其內的可運動頂部，被朝向噴嘴 腔室的底部轉動，藉此噴射穿過噴嘴開口。因爲噴嘴開口 隨著頂部而運動，所以藉由適當修飾噴嘴邊緣的形狀，可 控制液滴飛行的方向。此設計的優點是運動頂部只有一個 面必須對抗噴嘴腔室內側相對黏性的墨水工作。另一個優 點是藉由將主動樑構件和被動樑構件間隔開來，以使熱損 -6- 201107145 失最小化°此設計的缺點在於：喪失間隔開來之主動樑構 件和被動樑構件的構造剛性。 申請人的第US2008/0 1 29795號美國專利公開申請案 (茲將其內容倂入本案作參考）描述一種噴墨噴嘴，其包 含具有可運動頂部的噴嘴腔室，該頂部具有界定在其內的 噴嘴開□。該可運動頂部包含熱彎曲致動器，用於朝向腔 室的底部運動可運動頂部。用於改善致動器之效率的各種 裝置已被描述了，包括使用多孔性的二氧化矽於致動器的 被動層。 需要改善熱彎曲噴墨噴嘴的設計，以獲得更有效率的 液滴噴射和改善的機構堅固耐用性。從噴墨噴嘴的作業特 性和其製造兩個觀點而言，機構堅固耐用性是重要的因 素。製造時需要微機電系統製造步驟的順序，以便以高全 程產量來提供列印頭積體電路。 【發明內容】 在第一方面，提供一種熱彎曲致動器，包含：主動樑 和被動樑：主動樑用於連接至驅動電路；被動樑和該主動 樑機械式地耦合，使得當電流通過該主動樑時，該主動樑 相對於該被動樑膨脹，導致該致動器彎曲；其中該被動樑 包括第一層和第二層；該第一層包含氮化矽；該第二層包 含二氧化矽，且位於該第一層和該主動樑之間。 本發明的熱彎曲致動器的優點包括堅固且抗龜裂，同 時維持優良的熱效率。第一層的氮化矽提供抗龜裂，同時 201107145 第二層的二氧化矽提供熱絕緣，其維持整體 於主動樑和被動樑內不可避免的應力，所以 致動器內的問題，特別是被動樑。通常由二 動樑，二氧化矽具有良好的熱絕緣性質。藉 描述的雙層被動樑，本發明解決龜裂問題。 選擇性地，該第一層比該第二層還厚。 矽可爲第二層的二氧化矽厚2至20倍，選 20倍。 選擇性地，該第一層比該第二層至少厚 地至少厚四倍，或選擇性地至少厚八倍。 選擇性地，該第二層的厚度在〇.〇1和 圍內，選擇性地在〇.〇2和0.3微米的範圍 在0.05和0.2微米的範圍內，或選擇性地爲 選擇性地，該第一層的厚度在0.0 5和 圍內，選擇性地在1·〇和2.0微米的範圍內 爲約1.4微米。 選擇性地，該主動樑的厚度在0.05和 圍內，選擇性地在1·〇和3.0微米的範圍內 1.5和2.0微米的範圍內，或選擇性地爲約1 選擇性地，該主動樑經由一對電性接點 電路，該對接點位在該致動器的一端。 選擇性地，藉由沉積製程，該主動樑被 樑。 選擇性地，該主動樑包含傳導性熱彈性 的高效率。由 龜裂是熱彎曲 氧化矽形成被 由使用本文所 第一層的氮化 澤性地厚8至 兩倍，選擇性 0.5微米的範 內，選擇性地 約0.1微米。 5.0微米的範 ，或選擇性地 5 . 〇微米的範 ，選擇性地在 .7微米。 連接至該驅動 融合至該被動 材料，該材料 -8 - 201107145 選擇性地選自一群組，該群組由氮化鈦、氮化鈦鋁、和鋁 合金組成。 選擇性地，該主動樑包含釩鋁合金。 在第二方面，提供一種噴墨噴嘴組合體，包含：噴嘴 腔室，具有噴嘴開口和墨水入口；和熱彎曲致動器，用於 噴射墨水經過該噴嘴開口。該致動器包括：主動樑和被動 樑；主動樑用於連接至驅動電路；被動樑和該主動樑機械 式地耦合，使得當電流通過該主動樑時，該主動樑相對於 該被動樑膨脹，導致該致動器彎曲；其中該被動樑包括第 一層和第二層；該第一層包含氮化矽：該第二層包含二氧 化矽，且位於該第一層和該主動樑之間。 除了上文關於第一方面所討論的優點以外，第二方面 之噴墨噴嘴組合體的其他優點爲：第二層的氮化矽示對噴 嘴腔室內所含之液體的滲漏阻礙物。因此含水離子部能穿 過被動樑而溶出，且不能污染主動樑。該污染會導致噴嘴 故障。從熱墨水溶出含水離子，已被本申請案證明是熱彎 曲致動器的故障機制，該致動器具有只由二氧化矽製成之 被動樑。 選擇性地，該噴嘴腔室包括底部和頂部，該頂部具有 運動部分，藉此，該致動器的致動將該運動部分朝向該底 部運動。 選擇性地，該運動部分包括該致動器。 選擇性地，相對於該噴嘴腔室的該底部，該主動樑設 在該被動樑的上表面上。 -9- 201107145 選擇性地’該噴嘴開口被界定在該運動部分中，使得 該噴嘴開口可相對於該底部運動。 選擇性地，該致動器可相對於該噴嘴開口運動。 選擇性地，以聚合材料塗覆該頂部，該聚合材料例如 本文構詳細描述之聚合矽氧烷。 在第三方面，提供一種噴墨列印頭，包含複數噴嘴組 合體，每一噴嘴處合體包括：噴嘴腔室，具有噴嘴開口和 墨水入口；和熱彎曲致動器，用於噴射墨水經過該噴嘴開 口。該致動器包含：主動樑和被動樑；主動樑用於連接至 驅動電路；被動樑和該主動樑機械式地耦合，使得當電流 通過該主動樑時，該主動樑相對於該被動樑膨脹，導致該 致動器彎曲。其中該被動樑包括第一層和第二層；該第一 層包含氮化矽；該第二層包含二氧化矽，且位於該第一層 和該主動樑之間。 在第四方面，提供一種微機電系統裝置，包含一或更 多熱彎曲致動器，每一熱彎曲致動器包括：主動樑和被動 樑。主動樑連接至驅動電路；被動樑和該主動樑機械式地 耦合，使得當電流通過該主動樑時，該主動樑相對於該被 動樑膨脹，導致該致動器彎曲。其中該被動樑包括第一層 和第二層：該第一層包含氮化矽；該第二層包含二氧化 矽，且位於該第一層和該主動樑之間。 此等微機電系統裝置的例子包括晶片實驗室（LOC ) 閥和晶片實驗室泵（如同申請人之第1 2/142779號美國申 請案中所述）、感測器、開關等。熟悉技藝人士充分瞭解 -10- 201107145 包含有熱彎曲致動器之微機電系統裝置的液體過多之問 題。 在第五方面，提供製造熱彎曲致動器的方法，該方法 的步驟包括：（a)沉積第一層至犧牲支架上，該第一層 包括氮化矽：（b)沉積第二層至該第一層上，該第二層 包括二氧化矽；（c)沉積主動樑層至該第二層上；（d) 蝕刻該主動樑層、第一層、和第二層，以界定該熱彎曲致 動器，該熱彎曲致動器包括主動樑和被動樑，該被動樑包 括該第一層和第二層；和（e)藉由移除該犧牲支架，而 釋出該熱彎曲致動器。 選擇性地，該犧牲支架包括光阻劑或聚醯亞胺。 選擇性地，藉由氧化的電漿移除犧牲支架，在該技藝 中稱爲「灰化（ashing )」。使用氧（02 )電漿、氧/氮 (〇2/N2 )電漿、或任何其他氧化的電漿，可達成灰化。 選擇性地，在釋出熱彎曲致動器以後，被動樑中的殘 留應力主要存在於第一層內。 選擇性地，該方法形成用於噴墨噴嘴組合體之微機電 系統製造製程的至少一部分。 選擇性地，該第一和第二層界定噴嘴腔室的頂部。 選擇性地，該頂部包括運動部分，該運動部分包括熱 彎曲致動器。 選擇性地，在釋出該熱彎曲致動器以前，先在頂部內 界定噴嘴開口。 選擇性地，噴嘴開口界定在頂部的運動部分內。 -11 - 201107145 ^擇性地’在釋出該熱彎曲致動器以前，先以聚合材 料塗覆頂部。 選擇性地’在釋出該熱彎曲致動器以前，先以金屬層 保護聚合材料。 選擇性地’藉由旋轉製程將聚合材料塗覆在頂部上。 選擇性地’聚合材料爲聚合矽氧烷，例如聚二甲基倍 半砂氧烷、聚甲基倍半矽氧烷、或聚苯基倍半矽氧烷。 當然應瞭解’結合第一方面之熱彎曲致動器所描述的 選擇性方面，可等同地應用至第二、第三'第四、和第五 方面。 【實施方式】 應瞭解本發明可和具有主動樑融合至被動樑之任何熱 彎曲致動器一起使用。發現此等熱彎曲致動器使用在許多 微機電系統裝置中，包括噴墨噴嘴、開關、感測器、泵、 閥等。例如如同在第12/142779號美國案中所描述者，申 請人已展示在晶片實驗室（lab-on-a-chip )裝置中使用熱 彎曲致動器，該案的內容倂入本文做參考。如同在本文所 註明之交互參考專利案和專利申請案中所描述者，申請人 也展示噴墨噴嘴之液體過多的問題。雖然微機電系統熱彎 曲致動器可有許多不同的用途，但是本發明在本文將參考 申請人之噴墨噴嘴組合體其中之一做描述。當然應瞭解， 本發明不限於此特定的裝置。 圖1至13顯示申請人較早之第2008/0 309728號美國 -12- 201107145 專利案中所描述之用於噴墨噴嘴組合體1 00的微機電系統 之製造步驟的順序，該案的內容倂入本文做參考。圖 12、13所示之已完成的噴墨噴嘴組合體100使用熱彎曲 致動器，藉此，頂部的運動部分朝向基板彎曲，導致噴射 墨水。 製造微機電系統的開始點是標準的c Μ Ο S晶圓，該 C Μ Ο S晶圓具有形成在矽晶圓上部的C Μ Ο S驅動電路。在 微機電系統製造製程的末端，將該晶圓切割成個別的列印 頭積體電路，且每一積體電路包含驅動電路和複數噴嘴組 合體。 如圖1和2所示，基板101具有形成在其上部的電極 102。電極102是一對相鄰電極（正極和接地）其中之 ―，用於供給電力至噴墨噴嘴100的致動器。電極接受來 自 CMOS驅動電路（未示）的電力，該CMOS驅動電路 在基板101的上層。 圖1和2所示之另一電極103是用於供給電力至相鄰 的噴墨噴嘴。圖式大致顯示用於噴嘴組合體的微機電系統 製造步驟，該噴嘴組合體是一陣列噴嘴組合體其中之一。 下列的描述聚焦在這些噴嘴組合體其中之一噴嘴組合體的 製造步驟。但是當然應瞭解：對應的步驟可同時實施於形 成在晶圓上的各噴嘴組合體。圖式中顯示相鄰噴嘴組合體 的一部分，此是爲了本發明的目的而忽略另一部分。因 此，本文不詳細描述相鄰噴嘴組合體的電極103和全部特 徵。事實上，爲了清楚起見，一些微機電系統製造步驟未 -13- 201107145 顯示在相鄰噴嘴組合體上。 在圖1、2所示之步驟順序中，首先在基板1 〇 1上沉 積8微米的二氧化矽層。二氧化矽的厚度界定噴墨噴嘴之 噴嘴腔室105的深度。在沉積二氧化矽（Si 02 )層以後， 蝕刻該層以界定壁1〇4’該等壁104將成爲噴嘴腔室105 的側壁。 如圖3、4所示，然後以光阻劑或聚醯亞胺1 06充滿 噴嘴腔室105，該光阻劑於後續的沉積步驟當作犧牲支 架。使用標準技術將聚醯亞胺106旋轉塗覆（spin)至晶 圓上，紫外線硬化和/或烤硬（hardbaked )，然後經歷化 學機械平面化停止在二氧化矽壁104的上表面。 在圖4、5中，形成高傳導性的連接器柱1 08和噴嘴 腔室105的頂部構件107，該等連接器柱108向下延伸至 電極102。如圖12、13所示，部分的頂部構件107被用 於界定被動樑116，該被動樑116在完成的噴墨噴嘴組合 體中用於熱彎曲致動器115。在申請人先前的噴墨噴嘴設 計中，頂部構件1〇7(和藉此之熱彎曲致動器的被動樑） 是由二氧化矽製成。二氧化矽的熱傳導性不佳，此性質使 得在致動期間傳輸離開熱彎曲致動器之主動樑的熱量最小 化。藉由使用具有不好熱傳導性的被動樑，改善了裝置的 整個效率。但是在微機電系統製造期間和已完成之噴墨噴 嘴組合體的操作期間，二氧化矽容易受影響而龜裂。二氧 化矽的另一缺點是其具有某些程度的含水離子（例如氯化 物離子）滲透率，經由從噴嘴腔室中的熱墨水溶出 -14- 201107145 (leach )含水離子，導致隨著時間的經過而污染主動樑 層。此污染的機制會導致主動樑和熱彎曲致動器故障。非 常不希望出現此故障。 相較於二氧化矽，氮化矽較不易受影響而龜裂，且允 許較大範圍的殘留應力——壓應力和張應力兩者。氮化矽 也完全不具滲透率，此性質能將經由從噴嘴腔室中的墨水 溶出離子導致噴嘴故障最小化。但是氮化矽比二氧化矽的 熱傳導性高很多，導致熱彎曲致動器的效率較差。因此儘 管氮化矽比二氧化矽具有較佳的機械性質，但是通常不用 氮化矽當作被動樑。 在本發明中，頂部構件1 0 7界定致動器完成品的被動 樑。頂部構件1 07包含相對厚的氮化矽層1 3 1 (約1 .4微 米）和相對薄的二氧化矽層130 (約0·1微米）。暫時參 考圖12，在完成的致動器115中，二氧化矽層130位於 主動樑110和氮化矽層131之間。此配置改善了微機電系 統的製造，因爲當藉由移除犧牲聚醯亞胺或光阻劑而 「釋出（release)」致動器時，頂部構件107較不易受影 響而龜裂，特別是頂部構件107界定熱彎曲致動器之被動 樑的部分較不受影響。也改善了列印頭成品中之被動樑 1 1 6和連續頂部構件1 〇7所界定之列印頭的噴嘴板的堅固 耐用性，而不必明顯地折衷熱效率。再者’頂部構件107 不會允許從熱墨水溶出任何的含水離子至熱彎曲致動器的 主動樑。因此可瞭解：雙層被動樑改善了致動器的操作和 致動器的製造。 -15- 201107145 現在回到圖5、6，在沉積雙層的頂部構件 後，使用標準非等向性深活性離子蝕刻（DRIE )在 內形成一對通孔，且向下至電極102。此蝕刻經由 孔而暴露該對電極102。其次，使用無電式電鍍以 的高傳導性金屬充滿通孔。已沉積的銅柱1 〇 8遭受 械平面化（CMP )，停止在雙層頂部構件107上， 平坦構造。可看得到，在無電式銅電鍍期間所形成 接器柱1 08碰到各電極1 02，以提供線性傳導路徑 頂部構件107。 在圖7、8中，藉由初始地沉積0.3微米的鋁 層頂部構件107和連接器柱108上，而形成金屬室 可使用任何的高傳導性金屬（例如鋁、鈦等），且 應沉積約〇 · 5微米或以下的厚度，以免太嚴重影響 合體的整體平面度。金屬墊109位在連接器柱108 在頂部構件1 0 7的上面，且在熱彈性主動樑構件 「彎曲區域」中。 在圖9、1 0中，熱彈性主動樑構件1 1 0形成在 部107上方。藉由將部分的頂部構件107融合至主 件1 1 〇，該部分的頂部構件1 當作機械熱彎曲致 下被動樑構件116之用，該致動器是由主動樑110 樑1 1 6所界定。熱彈性主動樑構件11 0可由任何適 彈性材料製成，例如氮化鈦、氮化鈦鋁、和鋁合金 在申請人較早之第2008/0 1 29793號美國案（茲將 倂入做參考）所解釋者，釩鋁合金是較佳的材料， 107以 壁 104 個別通 例如銅 化學機 以提供 的銅連 向上至 層在雙 109。 該金屬 噴嘴組 上方且 之預定 雙層頂 動樑構 動器的 和被動 合的熱 。如同 其內容 因爲釩 -16- 201107145 鋁合金組合高熱膨脹和低密度及高楊氏模數的有利性質。 爲了形成主動樑構件110，藉由標準的電漿輔助化學 氣相沉積（PEC VD )初始地沉積i . 5微米傳導性熱彈性主 動材料層。然後使用標準金屬鈾刻法蝕刻樑的材料，以界 定主動樑構件1 1 〇。在完成如圖9、1 0所示的金屬蝕刻以 後，主動樑構件1 1 0包含部分的噴嘴開口 1 1 1和樑元件 1 12，該樑元件1 12經由連接器柱108電性地連接在正電 極和接地電極102的每一端。平坦的樑元件112從第一 (正）連接器柱108的頂部延伸，且彎曲180度回到第二 (接地）連接器柱的頂部。 仍然參考圖9、1 0，設置金屬墊1 09的位置，以利在 潛在性較高阻抗區域中的電流流動。一個金屬墊109位在 樑元件1 1 2的彎曲區域，且位在主動樑構件1 1 0和被動樑 構件1 16之間。其他的金屬墊109位在連接器柱108的頂 部和樑元件1 1 2的端部之間。 參考圖1 1，將疏水性聚合物層80沉積至晶圓上，並 以保護性金屬層90 (例如1 00奈米的鋁）覆蓋疏水性聚 合物層80。在適當的遮蔽以後，金屬層90、聚合物層 80、和雙層的頂部構件1 07被蝕刻，以界定頂部之完整的 噴嘴開口 1 1 3和運動部分1 1 4。 運動部分114包含熱彎曲致動器115，熱彎曲致動器 1 1 5本身包括主動樑構件1 1 〇和下面的被動樑構件1 1 6。 噴嘴開口 1 1 3被界定在頂部的運動部分1 1 4內，所以在致 動期間，噴嘴開口隨著致動器運動。如同第200 8/0 1 29793 201107145 號美國案所述，噴嘴開口 Π3相對於運動部分114呈靜止 的結構也有可能，且也在本發明的範圍內。 圍繞頂部之運動部分114的周圍區域117將運動部分 和頂部的靜止部分118分離。當致動器115致動時，此周 圍區域117允許運動部分114彎曲進入噴嘴腔室105內和 朝向基板101。疏水性聚合物層80塡充周圍區域117，以 提供頂部1 07之運動部分1 1 4和靜止部分1 1 8之間的機械 性密封。聚合物具有充分低的楊氏模數，以允許致動器朝 向基板1 0 1彎曲，同時防止墨水在致動期間從間隙1 1 7滲 漏。 聚合物層80通常由聚合的矽氧烷構成，可使用旋轉 (spin· on)製程將聚合的矽氧烷沉積成薄層（例如〇.5至 2.0微米）並烤硬（hardbaked )。合適之聚合材料的例子 爲：聚（烷基倍半矽氧烷），例如聚（甲基倍半矽氧 烷）；聚（芳基倍半矽氧烷），例如聚（苯基倍半矽氧 烷）；聚（二烷基矽氧烷），例如聚二甲基矽氧烷。聚合 材料可倂入奈米顆粒，以改善其耐久性、抗摩耗性、抗疲 勞性等。 在最終的微機電處理步驟中，且如圖12、13所示， 蝕刻形成墨水供給通道1 2 0，從基板1 〇 1的背側貫穿至噴 嘴腔室105。雖然圖12、13中所示的墨水供給通道120 對齊噴嘴開口 1 1 3，但是墨水供給通道1 2〇也可形成在偏 離噴嘴開口的位置。 在蝕刻墨水供給通道以後，藉由在氧化的電漿中灰化 -18- 201107145 (ashing )，而移除已塡注在噴嘴腔！ 106，且藉由氟化氬（HF)或過氧化氫 除金屬膜90 ’以提供噴嘴組合體1〇〇。 此領域的一般工作者可瞭解’如 例，可對本發明做很多種改變修飾’而 描述之本發明精神和範圍。因此應從各 爲例示性而非限制性。 【圖式簡單說明】 上文藉由只做爲例子並參考附圖的 的選擇性實施例。附圖爲： 圖1是在各步驟中的弟一順序以後 噴墨噴嘴組合體的側剖視圖，在該第一 腔室側壁； 圖2是圖1所示之局部製造噴墨 圖， 圖3是在各步驟中的第二順序以後 噴嘴組合體的側剖視圖，在該第二順序 滿噴嘴腔室； 圖4是圖3所示之局部製造噴墨 Γο! · 圖， 圖5是在各步驟中的第三順序以後 噴嘴組合體的側剖視圖，在該第三順序 上達腔室頂部； 105的聚醯亞胺 (Η2ο2)清洗而移 司所示的特定實施 不會脫離已寬廣地 方面考慮本實施例 方式，描述本發明 ，局部製造之替代 順序中，形成噴嘴 噴嘴組合體的立體 '局部製造之噴墨 中，以聚醯亞胺注 噴嘴組合體的立體 ，局部製造之噴墨 中，形成連接器柱 -19- 201107145 圖6是圖5所示之局部製造噴墨噴嘴組合體的立體 圖； 圖7是在各步驟中的第四順序以後，局部製造之噴墨 噴嘴組合體的側剖視圖，在該第四順序中，形成傳導性金 屬板； 圖8是圖7所示之局部製造噴墨噴嘴組合體的立體 I C3.I . 圖， 圖9是在各步驟中的第五順序以後，局部製造之噴墨 噴嘴組合體的側剖視圖，在該第五順序中，形成熱彎曲致 動器的主動樑構件； 圖10是圖9所示之局部製造噴墨噴嘴組合體的立體 圖； 圖11是在各步驟中的第六順序以後、在以聚合物層 塗覆以後、在以金屬層保護以後、和在蝕刻噴嘴開口以 後，局部製造之噴墨噴嘴組合體的側剖視圖； 圖1 2是在後側微機電系統處理和移除光阻劑以後， 已完成之噴墨噴嘴組合體的側剖視圖；和 圖13是圖12所示之噴墨噴嘴組合體的切除立體圖。 【主要元件符號說明】 8〇 :(疏水性）聚合物層 9〇 :(保護性）金屬層 100 :噴墨噴嘴（組合體） ιοί :基板 -20- 201107145 1 02 :電極 1 03 :電極 104 :壁 105 :噴嘴腔室 106 :光阻劑（聚醯亞胺） 107 :頂部（構件） 1 〇 8 :連接器柱 I 09 :金屬墊 II 〇 :主動樑（構件） 111:(部分的）噴嘴開口 1 1 2 :樑元件 113：(完整的）噴嘴開口 1 1 4 :運動部分 115:(熱彎曲）致動器 1 1 6 :被動樑（構件） 1 1 7 :周邊區域（間隙） 1 18 :靜止部分 120 :墨水供給通道 1 3 0 :二氧化矽層 1 3 1 :氮化矽層 -21 -201107145 VI. Description of the Invention [Technical Field] The present invention relates to the field of microelectromechanical systems (MEMS) devices, and more particularly to ink jet print heads. These ink jet print heads are primarily used to improve the robustness of the thermal bending actuator during and during manufacture of the MEMS. [Prior Art] The applicant of the present application has previously described the problem of excessive ink ejection using a thermal bending actuated MEMS inkjet nozzle. Thermal bending actuation generally refers to the bending motion of a material that is passed through a material and then thermally expanded relative to the other material. The resulting bending motion can be used to eject ink from the opening of the nozzle, selectively ejecting ink via the movement of the vanes or vanes, as this motion creates a pressure wave within the nozzle chamber. The inkjet nozzle described in the U.S. Patent No. 6,416,167, the disclosure of which is incorporated herein by reference in its entirety in the entire entire entire entire entire entire entire entire entire entire entire entire entire portion . The actuator takes the form of a passive beam (e.g., cerium oxide) fused to an active beam (e.g., titanium nitride) beneath the non-conductive material. The actuator is coupled to the vane via an arm that receives and passes through a slot in the wall of the nozzle chamber. When current is passed through the lower active beam, the actuator is bent downward, with the result that the vane moves toward the nozzle opening, which is defined at the top of the nozzle chamber, thereby ejecting ink droplets. The advantage of this design is the simplicity of its structure. The disadvantage of this design is that the two sides of the vane must work against the relatively viscous ink on the inside of the nozzle chamber -5 - 201107145. U.S. Pat. The actuator is in the form of a spiral core encasing the conductive material with a polymeric material. When actuated, the actuator bends toward the bottom of the nozzle chamber, increasing the pressure within the chamber and forcing ink droplets out of the nozzle opening, which is defined in the top of the chamber. The nozzle opening is defined in a non-moving portion of the top of the chamber. The advantage of this design is that the side of the top of the motion must work against the relatively viscous ink inside the nozzle chamber. A disadvantage of this design is that the polymer material is coated with the actuator structure of the spiral conductive element, making it difficult to achieve the MEMS process. The applicant's U.S. Patent No. 6,62, 311, the disclosure of which is incorporated herein by reference in its entirety in its entire entire entire entire entire entire entire entire entire entire entire entire entire entire entire entire entire entire entire portion The movable top is connected via an arm to a thermal bending actuator disposed outside the nozzle chamber. The actuator is in the form of an upper active beam and a lower passive beam spaced apart. By spacing the active and passive beams apart, the thermal bending efficiency is maximized because the passive beam cannot be used as a heat sink for the active beam. As soon as the current passes through the active beam, the movable top having the nozzle opening defined therein is rotated toward the bottom of the nozzle chamber, thereby being ejected through the nozzle opening. Since the nozzle opening moves with the top, the direction in which the droplets fly can be controlled by appropriately modifying the shape of the nozzle edge. The advantage of this design is that only one side of the top of the motion must work against the relatively viscous ink inside the nozzle chamber. Another advantage is that the active beam member and the passive beam member are spaced apart to minimize heat loss -6-201107145. This design has the disadvantage of losing the spacing of the active beam member and the passive beam member. rigidity. An inkjet nozzle comprising a nozzle chamber having a movable top having a defined therein is disclosed in the applicant's US Patent Application Publication No. US-A-2008/0 1, 295, the disclosure of which is incorporated herein by reference. The nozzle is open. The moveable top includes a thermal bending actuator for moving the movable top toward the bottom of the chamber. Various devices for improving the efficiency of the actuator have been described, including the use of porous ceria in the passive layer of the actuator. There is a need to improve the design of thermally curved inkjet nozzles for more efficient droplet ejection and improved mechanical robustness. From the standpoint of the operational characteristics of the inkjet nozzle and its manufacture, the robustness of the mechanism is an important factor. The order of fabrication steps of the MEMS is required at the time of manufacture to provide the print head integrated circuit at a high full throughput. SUMMARY OF THE INVENTION In a first aspect, a thermal bending actuator is provided, comprising: an active beam and a passive beam: an active beam is coupled to the drive circuit; the passive beam and the active beam are mechanically coupled such that when current flows through the The active beam expands relative to the passive beam, causing the actuator to bend; wherein the passive beam includes a first layer and a second layer; the first layer comprises tantalum nitride; the second layer comprises dioxide矽 and located between the first layer and the active beam. Advantages of the thermal bending actuator of the present invention include robustness and resistance to cracking while maintaining excellent thermal efficiency. The first layer of tantalum nitride provides resistance to cracking, while the 201107145 second layer of niobium dioxide provides thermal insulation, which maintains the inevitable stresses in the active and passive beams as a whole, so problems within the actuator, especially Passive beam. Usually made of a moving beam, cerium oxide has good thermal insulation properties. The present invention solves the problem of cracking by the described double-layer passive beam. Optionally, the first layer is thicker than the second layer.矽 can be 2nd to 20 times thicker for the second layer of cerium oxide, 20 times. Optionally, the first layer is at least four times thicker than the second layer, or alternatively at least eight times thicker. Optionally, the second layer has a thickness in the range of 〇.〇1 and circumference, selectively in the range of 0.05 and 0.2 μm in the range of 〇.〇2 and 0.3 μm, or alternatively, selectively The first layer has a thickness of 0.05 and a circumference, and is selectively about 1.4 microns in the range of 1 Å and 2.0 μm. Optionally, the active beam has a thickness of 0.05 and circumference, selectively in the range of 1.5 and 2.0 microns in the range of 1·〇 and 3.0 microns, or alternatively about 1 selectively, the active beam The mating point is located at one end of the actuator via a pair of electrical contact circuits. Optionally, the active beam is beamed by a deposition process. Optionally, the active beam contains a high efficiency of conductive thermoelasticity. The formation of yttrium oxide by cracking is thermally etched by using the first layer of the present invention to a thickness of 8 to 2, and a selectivity of 0.5 μm, selectively about 0.1 μm. A range of 5.0 microns, or alternatively a range of 5 〇 microns, selectively at .7 microns. Connected to the drive fused to the passive material, the material -8 - 201107145 is selectively selected from the group consisting of titanium nitride, titanium aluminum nitride, and an aluminum alloy. Optionally, the active beam comprises a vanadium aluminum alloy. In a second aspect, an ink jet nozzle assembly is provided comprising: a nozzle chamber having a nozzle opening and an ink inlet; and a thermal bending actuator for ejecting ink through the nozzle opening. The actuator includes: an active beam and a passive beam; the active beam is for connecting to the drive circuit; the passive beam and the active beam are mechanically coupled such that when the current passes through the active beam, the active beam expands relative to the passive beam Causing the actuator to bend; wherein the passive beam comprises a first layer and a second layer; the first layer comprises tantalum nitride: the second layer comprises ceria and is located in the first layer and the active beam between. In addition to the advantages discussed above with respect to the first aspect, another advantage of the ink jet nozzle assembly of the second aspect is that the tantalum nitride of the second layer exhibits a barrier to leakage of liquid contained within the nozzle chamber. Therefore, the aqueous ion portion can be dissolved through the passive beam and cannot contaminate the active beam. This contamination can cause nozzle failure. The dissolution of aqueous ions from hot ink has been demonstrated by the present application as a failure mechanism for a hot bending actuator having a passive beam made only of cerium oxide. Optionally, the nozzle chamber includes a bottom portion and a top portion, the top portion having a moving portion whereby actuation of the actuator moves the moving portion toward the bottom portion. Optionally, the moving portion includes the actuator. Optionally, the active beam is disposed on the upper surface of the passive beam relative to the bottom of the nozzle chamber. -9- 201107145 Optionally the nozzle opening is defined in the moving portion such that the nozzle opening is movable relative to the bottom. Optionally, the actuator is moveable relative to the nozzle opening. Optionally, the top is coated with a polymeric material such as the polymeric siloxane described in detail herein. In a third aspect, an inkjet printhead is provided, comprising a plurality of nozzle assemblies, each nozzle assembly comprising: a nozzle chamber having a nozzle opening and an ink inlet; and a thermal bending actuator for ejecting ink therethrough Nozzle opening. The actuator includes: an active beam and a passive beam; the active beam is for connecting to the drive circuit; the passive beam and the active beam are mechanically coupled such that when the current passes through the active beam, the active beam expands relative to the passive beam Causes the actuator to bend. Wherein the passive beam comprises a first layer and a second layer; the first layer comprises tantalum nitride; the second layer comprises hafnium oxide and is located between the first layer and the active beam. In a fourth aspect, a microelectromechanical system device is provided comprising one or more thermal bending actuators, each thermally curved actuator comprising: an active beam and a passive beam. The active beam is coupled to the drive circuit; the passive beam is mechanically coupled to the active beam such that when current is passed through the active beam, the active beam expands relative to the driven beam, causing the actuator to flex. Wherein the passive beam comprises a first layer and a second layer: the first layer comprises tantalum nitride; the second layer comprises hafnium oxide and is located between the first layer and the active beam. Examples of such MEMS devices include wafer lab (LOC) valves and wafer lab pumps (as described in the applicant's U.S. Patent Application Serial No. 1/142,779), sensors, switches, and the like. Familiar with the skilled person -10- 201107145 Excessive liquids in MEMS devices with hot bending actuators. In a fifth aspect, a method of making a thermal bending actuator is provided, the method comprising the steps of: (a) depositing a first layer onto a sacrificial support, the first layer comprising tantalum nitride: (b) depositing a second layer to On the first layer, the second layer includes cerium oxide; (c) depositing an active beam layer onto the second layer; (d) etching the active beam layer, the first layer, and the second layer to define the a thermal bending actuator comprising an active beam and a passive beam, the passive beam comprising the first layer and the second layer; and (e) releasing the thermal bending by removing the sacrificial bracket Actuator. Optionally, the sacrificial scaffold comprises a photoresist or polyimine. Optionally, the sacrificial scaffold is removed by oxidized plasma, referred to in the art as "ashing." Ashing can be achieved using oxygen (02) plasma, oxygen/nitrogen (〇2/N2) plasma, or any other oxidized plasma. Alternatively, after the thermal bending actuator is released, the residual stress in the passive beam is mainly present in the first layer. Optionally, the method forms at least a portion of a fabrication process for a microelectromechanical system for an inkjet nozzle assembly. Optionally, the first and second layers define a top of the nozzle chamber. Optionally, the top portion includes a moving portion that includes a thermal bending actuator. Optionally, the nozzle opening is defined within the top prior to the release of the thermal bending actuator. Optionally, the nozzle opening is defined within the moving portion of the top. -11 - 201107145 Selectively, the top is coated with a polymeric material prior to the release of the thermal bending actuator. The polymeric material is selectively protected with a metal layer prior to the release of the thermal bending actuator. The polymeric material is selectively applied to the top by a spin process. The 'polymeric material is a polymeric siloxane such as polydimethylsesquioxane, polymethylsesquioxanes, or polyphenylsesquioxanes. It should of course be understood that the selectivity aspects described in connection with the thermal bending actuator of the first aspect are equally applicable to the second, third 'fourth, and fifth aspects. [Embodiment] It will be appreciated that the present invention can be used with any thermal bending actuator having an active beam fused to a passive beam. These thermal bending actuators have been found to be used in many MEMS devices, including inkjet nozzles, switches, sensors, pumps, valves, and the like. For example, as described in U.S. Patent Application Serial No. 12/142,779, the Applicant has shown the use of a thermal-bending actuator in a lab-on-a-chip device, the contents of which are incorporated herein by reference. . Applicants have also shown problems with excessive liquids in ink jet nozzles, as described in the cross-reference patents and patent applications noted herein. While MEMS thermo-curvature actuators can have many different uses, the invention will be described herein with reference to one of applicant's inkjet nozzle assemblies. It should of course be understood that the invention is not limited to this particular device. 1 to 13 show the sequence of the manufacturing steps of the MEMS for the ink jet nozzle assembly 100 described in the applicant's earlier application No. 2008/0 309, 728, the disclosure of which is incorporated herein by reference. Refer to this article for reference. The completed ink jet nozzle assembly 100 shown in Figures 12, 13 uses a thermal bending actuator whereby the moving portion of the top is bent toward the substrate, causing the ink to be ejected. The starting point for fabricating a MEMS system is a standard c Μ Ο S wafer with a C Μ 驱动 S driver circuit formed on top of the 矽 wafer. At the end of the MEMS manufacturing process, the wafer is diced into individual print head integrated circuits, and each integrated circuit includes a drive circuit and a complex nozzle assembly. As shown in Figs. 1 and 2, the substrate 101 has an electrode 102 formed at an upper portion thereof. The electrode 102 is one of a pair of adjacent electrodes (positive electrode and ground) for supplying electric power to the actuator of the ink jet nozzle 100. The electrodes receive power from a CMOS drive circuit (not shown) which is on the upper layer of the substrate 101. The other electrode 103 shown in Figures 1 and 2 is for supplying electric power to an adjacent ink jet nozzle. The drawings generally show the MEMS manufacturing steps for a nozzle assembly that is one of an array of nozzle assemblies. The following description focuses on the manufacturing steps of one of the nozzle assemblies. However, it should of course be understood that the corresponding steps can be performed simultaneously on each nozzle assembly formed on the wafer. A portion of an adjacent nozzle assembly is shown in the drawings for the purpose of the present invention while ignoring the other portion. Therefore, the electrodes 103 and all features of adjacent nozzle assemblies are not described in detail herein. In fact, for the sake of clarity, some MEMS manufacturing steps are not shown on adjacent nozzle assemblies. In the sequence of steps shown in Figs. 1 and 2, an 8 μm layer of germanium dioxide is first deposited on the substrate 1 〇 1 . The thickness of the cerium oxide defines the depth of the nozzle chamber 105 of the ink jet nozzle. After depositing the cerium oxide (Si02) layer, the layer is etched to define walls 1〇4' which will become the sidewalls of the nozzle chamber 105. As shown in Figures 3 and 4, the nozzle chamber 105 is then filled with a photoresist or polyimide 106 which acts as a sacrificial support in subsequent deposition steps. Polyimine 106 is spin-coated onto the wafer using standard techniques, UV hardened and/or hardbaked, and then subjected to chemical mechanical planarization to stop on the upper surface of the ceria wall 104. In Figures 4, 5, a highly conductive connector post 108 and a top member 107 of the nozzle chamber 105 are formed, which extend down to the electrode 102. As shown in Figures 12 and 13, a portion of the top member 107 is used to define a passive beam 116 for use in the finished inkjet nozzle assembly for the thermal bending actuator 115. In the applicant's previous inkjet nozzle design, the top member 1〇7 (and the passive beam of the thermal bending actuator thereby) is made of cerium oxide. The thermal conductivity of cerium oxide is poor, and this property minimizes the heat transferred from the active beam of the thermal bending actuator during actuation. The overall efficiency of the device is improved by using a passive beam with poor thermal conductivity. However, during the manufacture of the MEMS system and during the operation of the completed ink jet nozzle assembly, cerium oxide is susceptible to cracking. Another disadvantage of cerium oxide is that it has some degree of ionic ion (eg chloride ion) permeability, which dissolves the aqueous ion through the hot ink from the nozzle chamber -14-201107145 (leach), resulting in time After the pollution, the active beam layer is polluted. This contamination mechanism can cause failure of the active beam and thermal bending actuators. This fault is highly undesirable. Compared to cerium oxide, tantalum nitride is less susceptible to cracking and allows a greater range of residual stresses—both compressive and tensile. Niobium nitride also has no permeability at all, which is capable of minimizing nozzle failure by dissolving ions from the ink in the nozzle chamber. However, tantalum nitride is much more thermally conductive than cerium oxide, resulting in less efficient thermal bending actuators. Therefore, although tantalum nitride has better mechanical properties than cerium oxide, tantalum nitride is generally not used as a passive beam. In the present invention, the top member 107 defines the passive beam of the actuator finish. The top member 107 comprises a relatively thick layer of tantalum nitride 1 31 (about 1.4 micrometers) and a relatively thin layer of tantalum dioxide 130 (about 0.1 micron). Referring briefly to Figure 12, in the completed actuator 115, the cerium oxide layer 130 is positioned between the active beam 110 and the tantalum nitride layer 131. This configuration improves the fabrication of the MEMS because the top member 107 is less susceptible to cracking when the actuator is "released" by removing the sacrificial polyimide or photoresist. It is the portion of the passive member of the top member 107 that defines the thermal bending actuator that is less affected. The robustness of the nozzle plate of the printhead defined by the passive beam 1 16 and the continuous top member 1 〇7 in the finished print head is also improved without having to significantly compromise thermal efficiency. Again, the top member 107 does not allow any aqueous ions to be eluted from the hot ink to the active beam of the thermal bending actuator. It is therefore understood that the double-layer passive beam improves the operation of the actuator and the manufacture of the actuator. -15- 201107145 Returning now to Figures 5 and 6, after depositing a double layer top member, a pair of through holes are formed therein using standard anisotropic deep reactive ion etching (DRIE) and down to electrode 102. This etching exposes the pair of electrodes 102 via the holes. Second, a high conductivity metal using electroless plating fills the through holes. The deposited copper pillars 1 遭受 8 are subjected to mechanical planarization (CMP) and are stopped on the double-layered top member 107 in a flat configuration. It can be seen that the connector post 108 hits each electrode 102 during electroless copper plating to provide a linear conductive path top member 107. In Figures 7 and 8, any of the highly conductive metals (e.g., aluminum, titanium, etc.) can be used to form the metal chamber by initially depositing a 0.3 micron aluminum layer top member 107 and the connector post 108, and should be deposited. A thickness of about 5 microns or less, so as not to seriously affect the overall flatness of the fit. The metal pad 109 is positioned over the connector post 108 above the top member 107 and in the "bending region" of the thermoelastic active beam member. In Figs. 9, 10, the thermoelastic active beam member 110 is formed above the portion 107. By splicing a portion of the top member 107 to the main member 1 1 〇, the portion of the top member 1 acts as a mechanical thermal bending to the lower passive beam member 116, which is made up of the active beam 110 beam 1 16 Defined. The thermoelastic active beam member 110 can be made of any suitable elastic material, such as titanium nitride, titanium aluminum nitride, and aluminum alloy, in the applicant's earlier US Patent No. 2008/0 1 29793 (for reference) As explained, vanadium-aluminum alloy is the preferred material, 107 is individually connected to the wall by 104, such as a copper chemical machine, to provide a copper connection up to the layer at 109. The passive heat of the predetermined double-layer top beam structure above the metal nozzle set. As its content, vanadium-16-201107145 aluminum alloy combination has the advantages of high thermal expansion and low density and high Young's modulus. To form the active beam member 110, a layer of i. 5 micron conductive thermoelastic material is initially deposited by standard plasma assisted chemical vapor deposition (PEC VD). The material of the beam is then etched using standard metal uranium engraving to define the active beam member 1 1 〇. After completing the metal etching as shown in FIGS. 9, 10, the active beam member 110 includes a portion of the nozzle opening 1 1 1 and the beam member 12, which is electrically connected via the connector post 108. Each end of the positive electrode and the ground electrode 102. A flat beam member 112 extends from the top of the first (positive) connector post 108 and is bent 180 degrees back to the top of the second (ground) connector post. Still referring to Figures 9, 10, the position of the metal pad 109 is set to facilitate current flow in potentially higher impedance regions. A metal pad 109 is located in the curved region of the beam member 112 and is positioned between the active beam member 110 and the passive beam member 116. Other metal pads 109 are located between the top of the connector post 108 and the end of the beam member 112. Referring to Figure 1, a hydrophobic polymer layer 80 is deposited onto the wafer and the hydrophobic polymer layer 80 is covered with a protective metal layer 90 (e.g., 100 nanometers of aluminum). After proper masking, metal layer 90, polymer layer 80, and double layer top member 107 are etched to define a complete nozzle opening 1 1 3 and moving portion 1 1 4 at the top. The moving portion 114 includes a thermal bending actuator 115 which itself includes the active beam member 1 1 〇 and the underlying passive beam member 1 16 . The nozzle opening 1 1 3 is defined within the top moving portion 1 1 4 so that the nozzle opening moves with the actuator during actuation. It is also possible that the nozzle opening Π3 is stationary with respect to the moving portion 114 as described in the U.S. Patent No. 2,800,0,,,,,,,,,,,. The surrounding portion 117 of the moving portion 114 around the top separates the moving portion from the stationary portion 118 at the top. This peripheral region 117 allows the moving portion 114 to bend into the nozzle chamber 105 and toward the substrate 101 when the actuator 115 is actuated. The hydrophobic polymer layer 80 fills the surrounding region 117 to provide a mechanical seal between the moving portion 1 14 and the stationary portion 1 18 of the top 107. The polymer has a sufficiently low Young's modulus to allow the actuator to bend toward the substrate 110 while preventing ink from escaping from the gap 117 during actuation. Polymer layer 80 is typically comprised of a polymerized decane, which can be deposited into a thin layer (e.g., 55 to 2.0 microns) and hardbaked using a spin on process. Examples of suitable polymeric materials are: poly(alkylsesquioxanes), such as poly(methylsesquioxanes); poly(arylsesquioxanes), such as poly(phenylsesquioxanes) Oxyalkylene); poly(dialkyloxane), such as polydimethyloxane. The polymeric material can be incorporated into the nanoparticle to improve its durability, abrasion resistance, fatigue resistance and the like. In the final microelectromechanical processing step, and as shown in Figs. 12 and 13, etching is performed to form the ink supply passage 120 from the back side of the substrate 1 〇 1 to the nozzle chamber 105. Although the ink supply passage 120 shown in Figs. 12, 13 is aligned with the nozzle opening 1 1 3, the ink supply passage 1 2 〇 may be formed at a position away from the nozzle opening. After etching the ink supply channel, the ash is -18-201107145 (ashing) in the oxidized plasma, and the removal has been injected into the nozzle cavity! 106, and the metal film 90' is removed by argon fluoride (HF) or hydrogen peroxide to provide the nozzle assembly 1A. The spirit and scope of the present invention will be described by those of ordinary skill in the art, as the invention may be modified. Therefore, each should be illustrative and not limiting. BRIEF DESCRIPTION OF THE DRAWINGS The above is an alternative embodiment by way of example only and with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a side cross-sectional view of the ink jet nozzle assembly after the sequence of the steps in each step, on the side wall of the first chamber; FIG. 2 is a partially fabricated ink jet view shown in FIG. A side cross-sectional view of the nozzle assembly after the second sequence in each step, in which the second sequence is full of the nozzle chamber; FIG. 4 is a partially fabricated ink jet 所示 ! · · · , , , , , , , , The third sequence is followed by a side cross-sectional view of the nozzle assembly, in the third sequence up to the top of the chamber; 105 polyimine (Η2ο2) cleaning and the specific implementation shown in the shift does not deviate from the broad considerations of this implementation Illustratively, the present invention is described in an alternative sequence of partial manufacturing in which a three-dimensional 'partially manufactured ink jet that forms a nozzle nozzle assembly is formed in a three-dimensional, partially fabricated ink jet of a polyimine injection nozzle assembly.柱柱-19- 201107145 Figure 6 is a perspective view of the partially fabricated inkjet nozzle assembly shown in Figure 5; Figure 7 is a side cross-sectional view of the partially fabricated inkjet nozzle assembly after the fourth sequence in each step, The fourth In the middle, a conductive metal plate is formed; FIG. 8 is a perspective view of the partially fabricated inkjet nozzle assembly shown in FIG. 7, and FIG. 9 is a partially fabricated inkjet after the fifth sequence in each step. A side cross-sectional view of the nozzle assembly in which the active beam member of the thermal bending actuator is formed; FIG. 10 is a perspective view of the partially fabricated inkjet nozzle assembly shown in FIG. 9; FIG. 11 is in each step After the sixth sequence, after coating with the polymer layer, after protection with the metal layer, and after etching the nozzle opening, a side cross-sectional view of the partially fabricated inkjet nozzle assembly; Figure 12 is the rear side microelectromechanical A side cross-sectional view of the completed ink jet nozzle assembly after system processing and removal of the photoresist; and FIG. 13 is an exploded perspective view of the ink jet nozzle assembly of FIG. [Explanation of main component symbols] 8〇: (hydrophobic) polymer layer 9〇: (protective) metal layer 100: inkjet nozzle (combination) ιοί: substrate-20-201107145 1 02: electrode 103: electrode 104 : Wall 105 : Nozzle chamber 106 : Photoresist (polyimide ) 107 : Top (member) 1 〇 8 : Connector column I 09 : Metal pad II 〇: Active beam (member) 111: (partial) Nozzle opening 1 1 2 : beam element 113: (complete) nozzle opening 1 1 4 : moving part 115: (thermal bending) actuator 1 1 6 : passive beam (member) 1 1 7 : peripheral area (gap) 1 18: still portion 120: ink supply passage 1 3 0 : ruthenium dioxide layer 1 3 1 : tantalum nitride layer-21 -