JP2004342705A

JP2004342705A - Tantalum nitride thin-film resistor and its manufacturing method

Info

Publication number: JP2004342705A
Application number: JP2003135104A
Authority: JP
Inventors: Motoki Obata; 元樹小畑; Kiichi Kamimura; 喜一上村; Rinpei Hayashibe; 林平林部
Original assignee: Cimeo Precision Co Ltd
Current assignee: Cimeo Precision Co Ltd
Priority date: 2003-05-13
Filing date: 2003-05-13
Publication date: 2004-12-02

Abstract

<P>PROBLEM TO BE SOLVED: To provide a tantalum nitride thin-film resistor that is improved in accuracy and stabilized with time by a simple method by suppressing the TCR to a small value as much as possible over the whole region of used temperatures by also paying attention to the contact resistance of an electrode film. <P>SOLUTION: In a structure in which the electrode film 4 is formed on a thin tantalum nitride film 2 formed on an insulating substrate 1 through an intermediate film 3, the sum of a first temperature coefficient of resistance which is the combined temperature coefficient of resistance of the intermediate film 3 and electrode film 4, and a second temperature coefficient of resistance which is the temperature coefficient of resistance of the thin-film resistor, is adjusted to -10-0 ppm/°C. In addition, Ti, Au, and AlN are respectively used as the materials of the intermediate film 3, electrode film 4, and insulating substrate 1. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

【０００１】
【発明の属する技術分野】
本発明は、サブマウント技術等に用いられる高精度高信頼性の窒化タンタル薄膜抵抗体及びその製造方法に関する。
【０００２】
【従来の技術】
近年、情報処理分野を始めとする集積回路等の半導体素子を用いた各種電子機器は、全て直流電源を必要としており、大規模化、高性能化に伴ってより大容量で高安定な直流電源が求められている。また、直流電源の高精度化には電流制御が不可欠であり、低抵抗値の高精度抵抗器が必要とされている。電流制御用とした場合には、一般的に知られた抵抗材料であるニクロムやマンガン等では十分な性能とは言えず、特に抵抗温度係数（ＴＣＲ）が著しく大きくなる事が指摘されている。
【０００３】
一方、窒化タンタル（ＴａＮ）は物理的、化学的に強固な特性を持つことに加えて、窒素の組成比を制御することにより、ＴＣＲを低くすることが可能であることから、高信頼性薄膜抵抗材料として着目されている。最近では、抵抗材料としての用途以外に、集積回路の配線技術において、銅配線の拡散バリア層としても注目されている。
【０００４】
ＴａＮ薄膜の堆積についてはすでに多くの報告があるが（例えば、非特許文献１参照。）、抵抗薄膜自体の抵抗値に着目した検討が大部分であり、電極との接触抵抗に関する報告は見あたらない。薄膜抵抗を回路要素として用いる場合、電極材料としてＡｕが使用される場合が多いが、基板との付着力が弱いため、界面にＣｒやＴｉ等の層を用いて基板との付着強度を高める方法が用いられる。
【０００５】
また、薄膜抵抗精度の温度依存性を向上させる為にＴＣＲを極力低くする事が必要であるが、この方法として、一つには窒化タンタル薄膜抵抗を形成した後一定温度の窒素雰囲気で熱処理をし、さらに酸化させる方法（例えば、特許文献１や特許文献２参照。）や、窒化タンタル薄膜抵抗形成中の窒素流量を調整して、ＴａとＴａ_２Ｎの正と負の温度係数を持つ抵抗膜の積層構造とする事（例えば、特許文献３や特許文献４参照。）が提唱されている。
【０００６】
【特許文献１】
特開昭５８−３３８０２号公報
【特許文献２】
特開昭５４−７６００号公報
【特許文献３】
特開昭５６−６４４０５号公報
【特許文献４】
特開昭５３−１０３１９４号公報
【非特許文献１】
「明石照久、竹盛英明、友部哲哉、小泉俊晃、ＡｌＮサブマウントに形成した窒化タンタル薄膜抵抗の特性評価、精密工学会誌Ｖｏｌ．６６、Ｎｏ８、ｐｐ．１０５２−１０５６、２００２年」
【０００７】
【発明が解決しようとする課題】
しかしながら、窒化タンタル薄膜抵抗を形成した後一定温度の窒素雰囲気で熱処理をする方法においては、熱処理工程が必要となり、さらに薄膜抵抗生成時の窒素流量によっては熱処理を行っても安定したＴＣＲを実現出来ないという問題があった。また、窒化タンタル薄膜抵抗生成中の窒素流量を変化させる事による正と負の薄膜抵抗の積層構造においても、使用温度領域全てに渡ってＴＣＲを低く維持する事は困難であり、安定したＴＣＲの窒化タンタル薄膜抵抗を形成出来ないという問題があった。また、この様な精度を制御する要因は電極とその接触面に起因すると考えられるが、明確な測定結果は従来報告されておらず、電極部分の接触抵抗は半導体素子の特性を制御する要因とされ、特に公称抵抗値１００ｍΩ以下の抵抗器では電極金属膜の抵抗値も無視出来ないという問題があった。
【０００８】
本発明の目的は、上記課題を解決して、使用温度領域全域に渡ってＴＣＲがほぼ０ｐｐｍ／℃を維持出来、高精度で安定した窒化タンタル薄膜抵抗を容易な方法で提供することにある。
【０００９】
【課題を解決するための手段】
上記目的を達成するため、本発明の請求項１及び請求項２記載の窒化タンタル薄膜抵抗体は、絶縁基板上に形成した窒化タンタル薄膜の上面に中間膜を介して電極膜を形成する構造において、中間膜と電極膜との合成抵抗温度係数を第１の抵抗温度係数とし、薄膜抵抗体の抵抗温度係数を第２の抵抗体温度係数とした時、第１と第２の温度係数の和を−１０ｐｐｍ／℃から０ｐｐｍ／℃とした。さらに、この時の中間膜の材質をＴｉ、電極膜の材質をＡｕとし、絶縁基板の材質をＡｌＮとしたことを特徴とする。
【００１０】
また、本発明の請求項３、請求項４及び請求項５記載の窒化タンタル薄膜抵抗体の製造方法は、絶縁基板上に形成した窒化タンタル薄膜の上面に中間膜を介して電極膜を形成する工程において、あらかじめ電送線路モデルを使用して前記中間膜と電極膜の合成抵抗温度係数である第１の抵抗温度係数を求める工程と、絶縁基板の片面全面に窒素雰囲気でのＲＦ反応性スパッタ法により窒化タンタル薄膜を形成し、この時の窒素分圧比を１０〜２０％、好ましくは１３〜１７％とする工程と、フォトレジスト法とＤＣスパッタ法の組合せにより所望の中間膜パターン及び電極膜パターンとを形成する工程を有し、さらに、この時の中間膜の材質をＴｉ、電極膜の材質をＡｕとし、絶縁基板の材質をＡｌＮとした事を特徴とする。
【００１１】
【発明の実施の形態】
以下、図面を用いて本発明の窒化タンタル薄膜抵抗体及びその製造方法の実施の形態を詳述する。反応性スパッタ法によりＴａＮ薄膜を堆積させ、スパッタガスとしてはＡｒを用い、基板にはＴａ板を使用した。基板裏面に磁石を置き、マグネトロンスパッタとした。スパッタガスに窒素を混合して、スパッタされたＴａ原子と反応させることによりＴａ薄膜を堆積させた。
【００１２】
図１は本発明の窒化タンタル薄膜抵抗体の基本構造断面図である。また、図２は本発明の窒化タンタル薄膜抵抗体をサブマウト基板に搭載した場合の上面図である。絶縁基板１の上面に形成された窒化タンタル薄膜２の上面側部及び絶縁基板上面には中間膜３を介して電極膜４が形成されている。図３は本発明の窒化タンタル薄膜抵抗の電送線路モデル試料の上面及び断面図であり、ガラス基板上に堆積させた薄膜について電気的特性評価を行い、比較のため窒素を流さずに堆積させた試料も作製した。スパッタガスであるＡｒの流量を２０ｓｃｃｍとし、Ｎ_２の流量を０、３、５ｓｃｃｍと変化させて堆積したが、この時の雰囲気の窒素分圧比はそれぞれ０、１５、２５％であった。基板温度はおよそ３００℃とし、スパッタ時間により膜圧を制御した。作製したＴａＮ薄膜はＸ線回折による結晶学的評価、ＳＥＭによる表面形状の観察、ＸＰＳによる組成分析、オージェ電子分光法による元素分析等を行った。
【００１３】
ＴａＮ薄膜の幅は１ｍｍとし、この上に電極としてＡｕ堆積した。基板とＡｕの付着強度を補うために、Ａｕ堆積前に中間膜としてＣｒ、Ｔａ、Ｔｉのいずれかを堆積させた。中間膜の厚さはおよそ６０ｎｍ、Ａｕ層の厚さはおよそ１４０ｎｍとした。電極の間隔を１ｍｍから８ｍｍまで変化させることにより、電極部分の抵抗とＴａＮ薄膜自体の抵抗を分離して評価した。この試料について、測定温度を−２０℃〜＋９０℃まで変化させて抵抗を測定し、ＴＣＲを電極部分と抵抗体自体をそれぞれ独立に評価した。
【００１４】
図３に示したガラス基板上に堆積したＴａＮ薄膜のＸ線回折パターンを図４に示した。Ｘ線光源は通常のＣｕＫαである。薄膜は基板温度３００℃で堆積した。膜厚はおよそ１５０ｎｍである。ＴａＮは高融点材料であり、基板温度の３００℃という値は、結晶化のために十分な値ではないため、結晶粒径が小さく十分な回折強度が得られなかったと考えられる。
【００１５】
窒素を流さないで堆積した場合は、純Ｔａの薄膜が形成された。窒素分圧比１５％で堆積した薄膜では、Ｔａ_２Ｎの（１０１）面からの回折位置付近にピークが現れた。窒素分圧比２５％に増すと、このピークに加えてＴａＮの（１１１）面からの回折と考えられる位置にもピークが現れている。これらの結果から、窒素流量を増すに従って薄膜中の窒素濃度が増加することが推察される。
【００１６】
全体的に薄膜からの回折ピークは半値幅が大きく、正確な結晶構造の決定はできなかったが、薄膜が極めて小さい結晶粒の多結晶で構成されていることを示している。ＸＰＳによる組成分析では、表面付近から多量の酸素が検出された。薄膜をスパッタエッチングしながら測定すると、この酸素は急激に減少した。ＸＰＳで検出された酸素原子は主に表面付近に吸着されているものと考えられる。酸化物と窒化物は近い位置に回折ピークが現れる可能性もあるので、この結果から結晶構造を特定することは難しい。
【００１７】
また、絶縁基板としてシリコン（Ｓｉ）基板を用いた場合のオージェ電子分光法により膜の構成物質を調べた結果を図６、図７に示す。両図では膜構成の深さ方向におけるプロファイルを示している。横軸はスパッタ時間（エッチング時間）を表していおり、これは膜表面からの深さに相当する。この結果より、元素構成は膜の深さ方向で一定であることが分かる。小さな不規則なふれはシステムの検出限界でありノイズと考えられる。膜表面に酸素が検出されたが１〜２ｍｉｎエッチングすると検出されなくなった。酸素やシリコンが検出されているが、膜中では非常に小さく検出限界である。膜中の窒素濃度は、窒素分圧比を１５％から２５％に増すことで、２０％から３０％に増加した。
【００１８】
窒素分圧比１５％で堆積した場合の膜の窒素濃度２０％は、Ｔａ_２Ｎ構造で予想される値より小さいので、ＴａとＴａ_２Ｎ構造が交じり合っていると考えられる。また、窒素分圧比２５％で堆積した場合の膜の窒素濃度はＴａＮ構造から予想される値よりも小さい。この場合は、ＴａとＴａ_２ＮとＴａＮが入り混じった構造になっていると考えられる。ここで、ピークの幅が広いことが、他の構造が混ざっていることを示唆している。また窒素濃度がＴａＮｘ膜とシリコンウエハ間で少し増加している。逆に、Ｔａの濃度はこの領域で減少している。これは、堆積前あるいは、堆積の初期の段階で窒化シリコンがシリコンウエハ表面に形成されたと考えられる。
【００１９】
次に、電極抵抗と薄膜の抵抗を分離するため、図３に示したように、幅１ｍｍのＴａＮ薄膜に２×２ｍｍの電極を、間隔を変えて設置して測定した。電極間距離と抵抗値の関係を図５に示す。抵抗値は、電極間隔の増加に対して直線的に増加している。しかしながら、電極部分の抵抗値は長さに無関係であるためこの値がｙ切片に現れている。薄膜の面抵抗をＲｓ、電極間隔をｄ、電極の抵抗をＲｃとすれば、全体の抵抗ＲはＲ＝Ｒｓ×ｄ＋２Ｒｃとなる。
【００２０】
図５において、ＴａＮ堆積時の窒素分圧比がパラメータにとってあり、窒素分圧比が増すに従って傾きすなわち薄膜の面抵抗Ｒｓが増加している。また、ｄ＝０における抵抗値は電極の抵抗を表しているが、この値も窒素分圧比に影響を受けていることがわかる。最小２乗法により、電極間距離と抵抗の関係を表す直線を求め、傾きとｄ＝０における抵抗値から薄膜の面抵抗Ｒｓと電極部分の抵抗Ｒｅを求めた結果を表１に示す。
【００２１】
【表１】

【００２２】
電極の抵抗２Ｒｓは、薄膜の面抵抗の１０分の１以上の値を示した事により、抵抗値を精密に設計する場合には電極の抵抗も考慮する必要があることが示された。また、ＴＣＲに対しても電極の抵抗が影響すると考えられる。
【００２３】
次に、ＴＣＲについて、抵抗器としての使用を考える場合、ＴＣＲが十分小さいことが求められる。電極部分と薄膜部分では異なる温度特性を持つことが考えられるので、相対的に電極部分の影響が表れやすい、電極間距離が短い試料について、窒素分圧比による窒化タンタル薄膜抵抗のＴＣＲの値を図８、図９、図１０に示す。電極は本来の電極として働くＡｕ薄膜とＴａＮとの間に、付着強度を増すための中間膜が挿入されている。中間膜は、接触抵抗の値に影響すると考えられるので、中間膜の材質を変えた場合について比較した。
【００２４】
窒素を流さないで堆積した場合には、Ｔａ薄膜が形成される。このため、ＴＣＲは金属的な正の値を示している。窒素を加えてスパッタすると、窒化タンタル薄膜が得られ、ＴＣＲは窒素分圧比が増すに従って負の大きな値となった。窒素分圧比１５％前後の領域でＴＣＲが正から負に変化しており、この近辺で最も低いＴＣＲが得られた。電極部分の抵抗値自体が小さいため、薄膜部分のＴＣＲが大きい場合には電極構造による差異は現れていない。しかしながら、窒素分圧比１５％で堆積した試料では、薄膜のＴＣＲが極めて低いため、中間膜の材質の違いがＴＣＲに現れている。
【００２５】
電極との間の単位面積あたりの接触抵抗が固有接触抵抗ρｃであり、接触面の電気的特性の評価基準の一つとして用いられている。電極金属の面抵抗が十分に低く無視できる場合、電極の抵抗Ｒｅは、数式１で表される。ここで、Ｗは試料の幅、Ｌは電極長である。電極長Ｌが数式２に比べて十分大きければ、数式３であり、固有接触抵抗ρｃは数式４となる。
【００２６】
【数１】

【００２７】
【数２】

【００２８】
【数３】

【００２９】
【数４】

【００３０】
電極部分の抵抗Ｒｅと薄膜の面抵抗Ｒｅは、図５に示した電極間距離と抵抗値の関係から求めることができる。図１１、図１２、図１３はこのようにして計算した固有接触抵抗の値を、測定温度の関数として示したものである。電極間隔と抵抗値の関係を最小２乗法で直線近似し、その傾きから薄膜の面抵抗Ｒｓを求め、ｄ＝０に外挿した点の抵抗値から電極抵抗Ｒｅを求め、数式４により固有接触抵抗率ρｃを計算した。固有接触抵抗率の値は、おおむね１０^−８Ω／ｍ^２のオーダーとなり、中間膜の種類や薄膜の組成により異る値が得られ、この結果から明らかなように、温度に対する依存性は小さいと言える。図１１、図１２、図１３の結果から、ＴａＮ薄膜に対しては、Ｔｉが接触抵抗の低い中間膜として有効であることがわかった。また、窒素を流さずに堆積した場合は、中間膜にＴａを用いた場合が最も低い固有接触抵抗率を示した。これは、膜自体がＴａであることによると考えられる。
【００３１】
図１４は窒化タンタル薄膜抵抗のＴＣＲを窒素流量の関数として表した場合のグラフである。電気的特性は堆積対象となる基板の面質に大きく影響されるため、表面形状が無視できるガラス基板に堆積させたＴａＮ薄膜について評価を行った。これより、窒素流量が変化してもＴＣＲ特性がほとんど変化しないプラトー領域が存在していることがわかる。本発明では、スパッタガスであるＡｒの流量を２０ｓｃｃｍとした時の窒素流量が２ｓｃｃｍから３ｓｃｃｍ、つまり窒素分圧１０％から２０％付近がプラトー領域であり、Ｔａ_２Ｎの結晶構造付近であり、さらに窒素分圧が１３％から１７％がより好ましい事が確認されている。また固有抵抗率もこの付近で一定の値を示している。
【００３２】
図１５は本発明の薄膜抵抗体の製造方法を示す。高精度薄膜抵抗を作製する場合、抵抗体自体はパターン精度で制御できるが、電極を搭載した場合の信頼性までは制御することが難しい。そこで、予め抵抗体と電極のＴＣＲをそれぞれ独立して評価し、それらを組み合わせた場合のトータル的なＴＣＲ設計を行うことで信頼性の高い抵抗器を作製する。そのサブ的なプロセスを以下に簡単に説明する。
▲１▼：フォトリソグラフィー（フォトレジスト法）とドライエッチング法及びリフトオフ法にてＴａＮｘ薄膜のＴＣＲ評価試料を作製し、ＴａＮｘ薄膜のＴＣＲ特性を算出する。
▲２▼：電送線路モデル試料を▲１▼と同様な方法で作製し、電極のＴＣＲ特性を算出する。
▲３▼：▲１▼及び▲２▼で算出したＴＣＲから抵抗器とてのＴＣＲが最も小さくなる窒素分圧比を設計する。
▲４▼：▲３▼で得られた条件下で、実際の薄膜抵抗体を作製し、サブマウントに搭載する。
【００３３】
前述の図２に示したサブマウト基板は以下の工程を経て作成される。
（１）両面研磨したＡｌＮ基板の片面全面にＲＦ反応性スパッタ法にてＴａ_２Ｎ薄膜を成膜する。このスパッタにおいて、前述したＴＣＲ設計の条件を考慮したスパッタを行う。次にこのＴａ_２Ｎ薄膜の抵抗パターンを形成するために、ポジ型レジストをスピン塗布し、露光現像してレジストパターンを形成する。
（２）このレジストパターンをマスクとしてイオンミリング装置を用いてＴａ_２Ｎ薄膜抵抗パターンを形成する。
（３）同じポジ型レジストをスピン塗布し、Ｔａ_２Ｎ薄膜抵抗体パターンの一部を保護する。
（４）電極パターンを形成するために、ロードロック式ＤＣスパッタ装置を用いてＴｉ／Ｐｔ／Ａｕの３層を連続成膜する。
（５）次にネガ型のドライフィルムレジストをラミネートし、電極用のレジストパターンを形成する。
（６）このレジストパターンをマスクとして、イオンミリングにより電極パターンを形成する。このとき、保護用のポジ型レジストを挟んでＴａ_２Ｎ薄膜上に形成されたＴｉ／Ｐｔ／Ａｕ膜もイオンミリングされる。次に、残った保護用レジストとパターン形成に用いたフィルムレジストを剥離・除去する。
（７）絶縁パターンであるＳｉＯ_２薄膜がのるＡｕ部にバッファ層を挿入するため、ポジ型レジストを塗布し、レジストパターンを形成する。同時に半田流れ防止パターンも形成する。次に、この上にＤＣスパッタ装置にてＴｉを成膜、リフトオフ法にてＴｉパターンを形成する。
（８）ポジ型レジストを塗布し、ＳｉＯ_２用のレジストパターンを形成する。次に、ＲＦスパッタ装置にて、ＳｉＯ_２を断続成膜する。
（９）ポジ型レジストを塗布し、ＳｉＯ_２パターンと正反対の逆パターンをレジストで形成する。次に、イオンミリングによりＳｉＯ_２パターン以外のＳｉＯ_２膜をエッチング除去し、残ったレジストを剥離し、ＳｉＯ_２パターンを形成する。
（１０）ＬＤを搭載する部分にＡｕ−Ｓｎ共晶半田（Ａｕ：７５ｗｔ％、Ｓｎ：２５ｗｔ％）を電子ビーム＆抵抗加熱の２元同時蒸着可能な真空蒸着装置にて成膜する。
（１１）リフトオフ法にてＡｕ−Ｓｎ半田パターンを形成し、最後にダイシングしてチップとする。
【００３４】
本発明の窒化タンタル抵抗薄膜体について、信頼性試験を行った結果を、図１６に通電加熱試験、図１７に高温高湿試験として示す。各試験は１０００ｈ経過するまで実施し、各試験の途中で雰囲気状態を保ったまま、サンプルの抵抗値をモニターし、窒化タンタル薄膜抵抗体の抵抗値変動を調査した。通電加熱試験は高温状態に置かれた素子に通常の２０倍の負荷を１０００ｈ加え、抵抗値の経時変化を調査したが、試験前後で変化は見られなかった。高温高湿試験では、８５℃、８５％、１０００ｈの条件のもと抵抗値の経時変化を調査し、ＡｌＮ基板での変化率１．５１％であった。よって、サブマウントとして形成したＴａ_２Ｎ薄膜抵抗体の信頼性は高いと言える。
【００３５】
【発明の効果】
本発明によれば、電極間隔を変えた薄膜抵抗について、抵抗値と電極間距離の関係ならびにその温度特性を測定し、中間層の材料に着目して比較し、電送線路モデルにより、ＴａＮ薄膜と電極との固有接触抵抗とその温度係数を求め、堆積条件との関係を求めた。電極部分の抵抗は、抵抗体部分の面抵抗の１０分の１程度の値を持ち、精密な値の薄膜抵抗を作製する場合には考慮する必要がある。また、固有接触抵抗は、薄膜の窒素濃度に依存し、およそ１０^−８Ω／ｍ^２のオーダーであった。これにより、基板材料により、金電極と基板間に介在させる中間膜の材質を選ぶ必要があるが、その場合には同時に薄膜との固有接触抵抗率も考慮する必要がある。固有接触抵抗率の温度係数は正であり、ＴａＮ薄膜のＴＣＲをうち消す方向に働いている。ＴＣＲが低い領域のＴａＮ薄膜に対してはＴｉが中間層として優れていること分かった。
【００３６】
以上により、電極膜の接触抵抗をも考慮し、使用温度領域全域に渡ってＴＣＲを極力小さく押さえ、高精度で経時的にも安定した窒化タンタル薄膜抵抗体を、容易な方法により供給する事が可能となった。
【図面の簡単な説明】
【図１】本発明の窒化タンタル薄膜抵抗体の基本構造断面図。
【図２】本発明の窒化タンタル薄膜抵抗体を搭載したサブマウント基板の上面図。
【図３】本発明の窒化タンタル薄膜抵抗の電送線路モデル試料の上面及び断面図。
【図４】Ｓｉ基板上に堆積した窒化タンタル薄膜のＸ線回折結果を示すグラフ。
【図５】窒化タンタル薄膜抵抗の電極間距離と抵抗値との関係を示すグラフ。
【図６】窒素分圧比１５％の場合のオージェ電子分光法を用いた元素分析結果のグラフ。
【図７】窒素分圧比２５％の場合のオージェ電子分光法を用いた元素分析結果のグラフ。
【図８】窒素分圧比０％の場合の窒化タンタル薄膜抵抗のＴＣＲを示すグラフ。
【図９】窒素分圧比１５％の場合の窒化タンタル薄膜抵抗のＴＣＲを示すグラフ。
【図１０】窒素分圧比２５％の場合の窒化タンタル薄膜抵抗のＴＣＲを示すグラフ。
【図１１】窒素分圧比０％の場合の窒化タンタル薄膜抵抗に対する固有接触抵抗率の抵抗温度特性を示すグラフ。
【図１２】窒素分圧比１５％の場合の窒化タンタル薄膜抵抗に対する固有接触抵抗率の抵抗温度特性を示すグラフ。
【図１３】窒素分圧比２５％の場合の窒化タンタル薄膜抵抗に対する固有接触抵抗率の抵抗温度特性を示すグラフ。
【図１４】窒化タンタル薄膜抵抗のＴＣＲを窒素流量の関数として表した場合のグラフ。
【図１５】本発明の窒化タンタル薄膜抵抗体の製造方法を示す工程フロー図。
【図１６】本発明の窒化タンタル薄膜抵抗体の通電加熱試験の結果を示すグラフ。
【図１７】本発明の窒化タンタル薄膜抵抗体の高温高湿試験の結果を示すグラフ。
【符号の説明】
１絶縁基板
２窒化タンタル薄膜抵抗
３中間膜
４電極膜[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a high-accuracy and high-reliability tantalum nitride thin-film resistor used for a submount technique and the like, and a method for manufacturing the same.
[0002]
[Prior art]
In recent years, various electronic devices using semiconductor elements such as integrated circuits, including information processing fields, all require a DC power supply. Is required. Further, current control is indispensable for improving the accuracy of a DC power supply, and a high-precision resistor having a low resistance value is required. When used for current control, it is pointed out that generally known resistance materials, such as nichrome and manganese, cannot be said to have sufficient performance, and in particular, the temperature coefficient of resistance (TCR) is significantly increased.
[0003]
On the other hand, tantalum nitride (TaN) has physically and chemically strong characteristics, and it is possible to lower the TCR by controlling the composition ratio of nitrogen. It has received attention as a resistance material. Recently, in addition to use as a resistance material, in a wiring technology of an integrated circuit, a diffusion barrier layer of a copper wiring has been attracting attention.
[0004]
Although there are many reports on the deposition of a TaN thin film (for example, see Non-Patent Document 1), most studies focus on the resistance value of the resistive thin film itself, and no reports on the contact resistance with the electrode are found. . When a thin film resistor is used as a circuit element, Au is often used as an electrode material. However, since the adhesion to the substrate is weak, a method of increasing the adhesion to the substrate by using a layer of Cr or Ti at the interface is used. Is used.
[0005]
It is necessary to lower the TCR as much as possible in order to improve the temperature dependence of the thin film resistance accuracy. One of the methods is to form a tantalum nitride thin film resistor and then perform heat treatment in a nitrogen atmosphere at a certain temperature. Then, a resistor having a positive and negative temperature coefficient of Ta and Ta ₂ N is adjusted by further oxidizing (see, for example, Patent Documents 1 and 2) and adjusting the flow rate of nitrogen during the formation of the tantalum nitride thin film resistor. It has been proposed to adopt a laminated structure of films (for example, see Patent Literature 3 and Patent Literature 4).
[0006]
[Patent Document 1]
JP-A-58-33802 [Patent Document 2]
JP-A-54-7600 [Patent Document 3]
JP-A-56-64405 [Patent Document 4]
JP-A-53-103194 [Non-Patent Document 1]
"Teruhisa Akashi, Hideaki Takemori, Tetsuya Tomobe, Toshiaki Koizumi, Characteristic Evaluation of Tantalum Nitride Thin Film Resistor Formed on AlN Submount, Journal of the Japan Society of Precision Engineering, Vol. 66, No. 8, pp. 1052-1056, 2002"
[0007]
[Problems to be solved by the invention]
However, in the method of performing heat treatment in a nitrogen atmosphere at a constant temperature after forming a tantalum nitride thin film resistor, a heat treatment step is required, and a stable TCR can be realized even with the heat treatment depending on the nitrogen flow rate at the time of thin film resistor generation. There was no problem. Also, in a laminated structure of positive and negative thin film resistors by changing the flow rate of nitrogen during the formation of the tantalum nitride thin film resistor, it is difficult to keep the TCR low over the entire operating temperature range. There is a problem that a tantalum nitride thin film resistor cannot be formed. In addition, it is thought that the factors controlling such accuracy are due to the electrodes and their contact surfaces, but no clear measurement results have been reported so far, and the contact resistance of the electrode portions is a factor controlling the characteristics of the semiconductor element. In particular, a resistor having a nominal resistance value of 100 mΩ or less has a problem that the resistance value of the electrode metal film cannot be ignored.
[0008]
SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to provide a highly accurate and stable tantalum nitride thin film resistor that can maintain a TCR of almost 0 ppm / ° C over the entire operating temperature range by an easy method.
[0009]
[Means for Solving the Problems]
In order to achieve the above object, a tantalum nitride thin film resistor according to

claims

1 and 2 of the present invention has a structure in which an electrode film is formed on an upper surface of a tantalum nitride thin film formed on an insulating substrate via an intermediate film. When the combined resistance temperature coefficient of the intermediate film and the electrode film is defined as the first resistance temperature coefficient, and the resistance temperature coefficient of the thin film resistor is defined as the second resistance temperature coefficient, the sum of the first and second temperature coefficients is obtained. From -10 ppm / ° C to 0 ppm / ° C. Further, the material of the intermediate film at this time is Ti, the material of the electrode film is Au, and the material of the insulating substrate is AlN.
[0010]
According to a third aspect of the present invention, there is provided a method for manufacturing a tantalum nitride thin film resistor, wherein an electrode film is formed on an upper surface of a tantalum nitride thin film formed on an insulating substrate via an intermediate film. A step of obtaining a first temperature coefficient of resistance, which is a combined resistance temperature coefficient of the intermediate film and the electrode film, using a transmission line model in advance, and an RF reactive sputtering method in a nitrogen atmosphere over the entire surface of one side of the insulating substrate. Forming a thin film of tantalum nitride, and setting the nitrogen partial pressure ratio at this time to 10 to 20%, preferably 13 to 17%, and a desired intermediate film pattern and electrode film pattern by a combination of a photoresist method and a DC sputtering method. In addition, the material of the intermediate film at this time is made of Ti, the material of the electrode film is made of Au, and the material of the insulating substrate is made of AlN.
[0011]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of a tantalum nitride thin film resistor and a method for manufacturing the same according to the present invention will be described in detail with reference to the drawings. A TaN thin film was deposited by a reactive sputtering method, Ar was used as a sputtering gas, and a Ta plate was used as a substrate. A magnet was placed on the back surface of the substrate to perform magnetron sputtering. A Ta thin film was deposited by mixing nitrogen with a sputter gas and reacting with sputtered Ta atoms.
[0012]
FIG. 1 is a sectional view showing the basic structure of a tantalum nitride thin film resistor according to the present invention. FIG. 2 is a top view when the tantalum nitride thin film resistor of the present invention is mounted on a submount substrate. An electrode film 4 is formed on an upper surface side of the tantalum nitride thin film 2 formed on the upper surface of the insulating substrate 1 and on the upper surface of the insulating substrate via an intermediate film 3. FIG. 3 is a top view and a cross-sectional view of a transmission line model sample of a tantalum nitride thin film resistor according to the present invention. The thin film deposited on a glass substrate was evaluated for electrical characteristics, and deposited without flowing nitrogen for comparison. Samples were also made. The deposition was performed by changing the flow rate of Ar as a sputtering gas to 20 sccm and the flow rate of N ₂ to 0, 3, and 5 sccm. At this time, the nitrogen partial pressure ratio in the atmosphere was 0, 15, and 25%, respectively. The substrate temperature was about 300 ° C., and the film pressure was controlled by the sputtering time. The produced TaN thin film was subjected to crystallographic evaluation by X-ray diffraction, observation of surface shape by SEM, composition analysis by XPS, elemental analysis by Auger electron spectroscopy, and the like.
[0013]
The width of the TaN thin film was 1 mm, and Au was deposited thereon as an electrode. In order to supplement the adhesion strength between the substrate and Au, any of Cr, Ta, and Ti was deposited as an intermediate film before depositing Au. The thickness of the intermediate film was about 60 nm, and the thickness of the Au layer was about 140 nm. By changing the distance between the electrodes from 1 mm to 8 mm, the resistance of the electrode portion and the resistance of the TaN thin film itself were separately evaluated. For this sample, the resistance was measured while changing the measurement temperature from -20 ° C to + 90 ° C, and the TCR was evaluated independently for the electrode portion and the resistor itself.
[0014]
FIG. 4 shows an X-ray diffraction pattern of the TaN thin film deposited on the glass substrate shown in FIG. The X-ray light source is ordinary CuKα. The thin film was deposited at a substrate temperature of 300 ° C. The thickness is about 150 nm. Since TaN is a high melting point material and the substrate temperature of 300 ° C. is not a value sufficient for crystallization, it is considered that the crystal grain size was small and sufficient diffraction intensity could not be obtained.
[0015]
When deposition was performed without flowing nitrogen, a thin film of pure Ta was formed. In the thin film deposited at a nitrogen partial pressure ratio of 15%, a peak appeared near the diffraction position from the (101) plane of Ta ₂ N. When the nitrogen partial pressure ratio is increased to 25%, a peak appears at a position considered to be diffraction from the (111) plane of TaN in addition to this peak. From these results, it is inferred that the nitrogen concentration in the thin film increases as the nitrogen flow rate increases.
[0016]
As a whole, the diffraction peak from the thin film has a large half-value width, and an accurate crystal structure could not be determined, but it indicates that the thin film is composed of polycrystals of extremely small crystal grains. In the composition analysis by XPS, a large amount of oxygen was detected near the surface. When measured while sputter etching the thin film, the oxygen decreased sharply. It is considered that oxygen atoms detected by XPS are mainly adsorbed near the surface. Since the oxide and nitride may have diffraction peaks at close positions, it is difficult to specify the crystal structure from the results.
[0017]
6 and 7 show the results of examining the constituent materials of the film by Auger electron spectroscopy when a silicon (Si) substrate is used as the insulating substrate. Both figures show profiles in the depth direction of the film configuration. The horizontal axis represents the sputtering time (etching time), which corresponds to the depth from the film surface. From this result, it is understood that the element constitution is constant in the depth direction of the film. Small irregular runout is the detection limit of the system and is considered noise. Oxygen was detected on the film surface, but was not detected after etching for 1 to 2 minutes. Oxygen and silicon have been detected, but are very small in the film and have a detection limit. The nitrogen concentration in the film was increased from 20% to 30% by increasing the nitrogen partial pressure ratio from 15% to 25%.
[0018]
Since the nitrogen concentration of the film deposited at a nitrogen partial pressure ratio of 15% is smaller than the value expected for the Ta ₂ N structure, it is considered that the Ta and Ta ₂ N structures are intermingled. The nitrogen concentration of the film deposited at a nitrogen partial pressure ratio of 25% is smaller than the value expected from the TaN structure. In this case, it is considered that the structure is a mixture of Ta, Ta ₂ N, and TaN. Here, a wide peak width suggests that other structures are mixed. Further, the nitrogen concentration is slightly increased between the TaNx film and the silicon wafer. Conversely, the concentration of Ta decreases in this region. This is presumably because silicon nitride was formed on the silicon wafer surface before the deposition or at an early stage of the deposition.
[0019]
Next, in order to separate the electrode resistance and the resistance of the thin film, as shown in FIG. 3, a 2 × 2 mm electrode was placed at different intervals on a 1 mm wide TaN thin film and measured. FIG. 5 shows the relationship between the distance between the electrodes and the resistance value. The resistance value increases linearly with an increase in the electrode spacing. However, since the resistance value of the electrode portion is independent of the length, this value appears in the y-intercept. Assuming that the sheet resistance of the thin film is Rs, the electrode interval is d, and the resistance of the electrode is Rc, the total resistance R is R = Rs × d + 2Rc.
[0020]
In FIG. 5, the nitrogen partial pressure ratio at the time of TaN deposition is a parameter, and the slope, that is, the sheet resistance Rs of the thin film increases as the nitrogen partial pressure ratio increases. Further, the resistance value at d = 0 represents the resistance of the electrode, and it can be seen that this value is also affected by the nitrogen partial pressure ratio. Table 1 shows the result of obtaining a straight line representing the relationship between the inter-electrode distance and the resistance by the least squares method, and obtaining the sheet resistance Rs of the thin film and the resistance Re of the electrode portion from the slope and the resistance value at d = 0.
[0021]
[Table 1]

[0022]
The resistance 2Rs of the electrode showed a value of 1/10 or more of the sheet resistance of the thin film, indicating that it is necessary to consider the resistance of the electrode when designing the resistance value precisely. It is also considered that the resistance of the electrode affects the TCR.
[0023]
Next, when using the TCR as a resistor, it is required that the TCR be sufficiently small. Since the electrode portion and the thin film portion may have different temperature characteristics, the TCR value of the tantalum nitride thin film resistor according to the nitrogen partial pressure ratio is shown for a sample where the effect of the electrode portion is relatively likely to appear and the distance between the electrodes is short. 8, FIG. 9 and FIG. In the electrode, an intermediate film for increasing the adhesive strength is inserted between the Au thin film serving as an original electrode and TaN. Since the intermediate film is considered to affect the value of the contact resistance, a comparison was made when the material of the intermediate film was changed.
[0024]
When the deposition is performed without flowing nitrogen, a Ta thin film is formed. Therefore, the TCR indicates a metallic positive value. When sputtering was performed by adding nitrogen, a tantalum nitride thin film was obtained, and the TCR became a large negative value as the nitrogen partial pressure ratio increased. The TCR changed from positive to negative in a region where the nitrogen partial pressure ratio was around 15%, and the lowest TCR was obtained in the vicinity of this. Since the resistance value itself of the electrode portion is small, no difference due to the electrode structure appears when the TCR of the thin film portion is large. However, in the sample deposited at a nitrogen partial pressure ratio of 15%, the TCR of the thin film is extremely low, so that a difference in the material of the intermediate film appears in the TCR.
[0025]
The contact resistance per unit area with the electrode is the specific contact resistance ρc, which is used as one of the evaluation criteria for the electrical characteristics of the contact surface. When the sheet resistance of the electrode metal is sufficiently low and can be ignored, the resistance Re of the electrode is expressed by the following equation (1). Here, W is the width of the sample, and L is the electrode length. If the electrode length L is sufficiently larger than Equation 2, Equation 3 is obtained, and the specific contact resistance ρc is Equation 4.
[0026]
(Equation 1)

[0027]
(Equation 2)

[0028]
[Equation 3]

[0029]
(Equation 4)

[0030]
The resistance Re of the electrode portion and the sheet resistance Re of the thin film can be obtained from the relationship between the distance between the electrodes and the resistance value shown in FIG. FIGS. 11, 12 and 13 show the values of the specific contact resistance calculated in this way as a function of the measured temperature. The relationship between the electrode spacing and the resistance value is linearly approximated by the least-squares method, the surface resistance Rs of the thin film is obtained from the slope, the electrode resistance Re is obtained from the resistance value at a point extrapolated to d = 0, The resistivity ρc was calculated. The value of the specific contact resistivity is on the order of 10 ⁻⁸ Ω / m ² , and different values are obtained depending on the type of the intermediate film and the composition of the thin film. As is clear from the result, the dependence on the temperature is small. It can be said. From the results of FIGS. 11, 12, and 13, it was found that Ti is effective as an intermediate film having a low contact resistance with respect to a TaN thin film. When the deposition was performed without flowing nitrogen, the lowest specific contact resistivity was obtained when Ta was used for the intermediate film. This is probably because the film itself is Ta.
[0031]
FIG. 14 is a graph showing the TCR of a tantalum nitride thin film resistor as a function of the nitrogen flow rate. Since the electrical characteristics are greatly affected by the surface quality of the substrate to be deposited, a TaN thin film deposited on a glass substrate having a negligible surface shape was evaluated. This indicates that there is a plateau region where the TCR characteristic hardly changes even when the nitrogen flow rate changes. In the present invention, when the flow rate of Ar as the sputtering gas is 20 sccm, the nitrogen flow rate is 2 sccm to 3 sccm, that is, the nitrogen partial pressure is around 10% to 20% is the plateau region, and the Ta ₂ N crystal structure is around, It has been confirmed that a nitrogen partial pressure of 13% to 17% is more preferable. The specific resistivity also shows a constant value in this vicinity.
[0032]
FIG. 15 shows a method of manufacturing a thin film resistor according to the present invention. When fabricating a high-precision thin-film resistor, the resistor itself can be controlled with pattern accuracy, but it is difficult to control the reliability when an electrode is mounted. Therefore, a highly reliable resistor is manufactured by independently evaluating the TCRs of the resistor and the electrode in advance and performing a total TCR design in the case of combining them. The sub-process is briefly described below.
{Circle around (1)} A TCR evaluation sample of a TaNx thin film is prepared by photolithography (photoresist method), dry etching method and lift-off method, and the TCR characteristics of the TaNx thin film are calculated.
(2): A transmission line model sample is prepared in the same manner as in (1), and the TCR characteristics of the electrode are calculated.
(3): Design a nitrogen partial pressure ratio that minimizes the TCR as a resistor from the TCR calculated in (1) and (2).
(4): Under the conditions obtained in (3), an actual thin film resistor is manufactured and mounted on a submount.
[0033]
The submount substrate shown in FIG. 2 is produced through the following steps.
(1) A Ta ₂ N thin film is formed on the entire surface of one side of an AlN substrate polished on both sides by RF reactive sputtering. In this sputtering, sputtering is performed in consideration of the TCR design conditions described above. Next, in order to form a resist pattern of the Ta ₂ N thin film, a positive resist is spin-coated, exposed and developed to form a resist pattern.
(2) Using the resist pattern as a mask, a Ta ₂ N thin film resistance pattern is formed using an ion milling device.
(3) The same positive type resist is spin-coated to protect a part of the Ta ₂ N thin film resistor pattern.
(4) In order to form an electrode pattern, three layers of Ti / Pt / Au are continuously formed using a load lock type DC sputtering apparatus.
(5) Next, a negative type dry film resist is laminated to form a resist pattern for an electrode.
(6) Using this resist pattern as a mask, an electrode pattern is formed by ion milling. At this time, the Ti / Pt / Au film formed on the Ta ₂ N thin film via the protective positive resist is also ion-milled. Next, the remaining protective resist and the film resist used for pattern formation are peeled and removed.
(7) A positive resist is applied to insert a buffer layer in the Au portion on which the SiO ₂ thin film as an insulating pattern is to be deposited, and a resist pattern is formed. At the same time, a solder flow prevention pattern is also formed. Next, a Ti film is formed thereon by a DC sputtering apparatus, and a Ti pattern is formed by a lift-off method.
(8) A positive resist is applied to form a resist pattern for SiO ₂ . Next, SiO ₂ is intermittently formed by an RF sputtering apparatus.
(9) A positive resist is applied, and a pattern opposite to the SiO ₂ pattern is formed with the resist. Next, the SiO ₂ film other than the SiO ₂ pattern is removed by etching by ion milling, and the remaining resist is stripped to form a SiO ₂ pattern.
(10) Au-Sn eutectic solder (Au: 75 wt%, Sn: 25 wt%) is formed on a portion on which the LD is to be mounted by a vacuum vapor deposition apparatus capable of simultaneous vapor deposition of electron beam and resistance heating.
(11) An Au-Sn solder pattern is formed by a lift-off method, and is finally diced to obtain a chip.
[0034]
The results of a reliability test performed on the tantalum nitride resistance thin film of the present invention are shown in FIG. 16 as an electrical heating test and FIG. 17 as a high-temperature and high-humidity test. Each test was performed until 1000 hours had elapsed, and the resistance value of the sample was monitored while maintaining the atmosphere state during each test, and the variation in the resistance value of the tantalum nitride thin film resistor was investigated. In the electrical heating test, a 20-fold normal load was applied to the element placed in a high temperature state for 1,000 hours, and the change with time in the resistance value was examined. No change was observed before and after the test. In the high-temperature and high-humidity test, the change with time in the resistance under the conditions of 85 ° C., 85%, and 1000 hours was examined, and the rate of change in the AlN substrate was 1.51%. Therefore, it can be said that the reliability of the Ta ₂ N thin film resistor formed as a submount is high.
[0035]
【The invention's effect】
According to the present invention, the relationship between the resistance value and the inter-electrode distance and the temperature characteristics of the thin-film resistor with the changed electrode spacing are measured and compared with a focus on the material of the intermediate layer. The specific contact resistance with the electrode and its temperature coefficient were determined, and the relationship with the deposition conditions was determined. The resistance of the electrode portion has a value of about 1/10 of the sheet resistance of the resistor portion, and it is necessary to consider it when producing a thin film resistor having a precise value. The specific contact resistance depends on the nitrogen concentration of the thin film, and was on the order of about 10 ⁻⁸ Ω / m ² . Accordingly, it is necessary to select the material of the intermediate film interposed between the gold electrode and the substrate depending on the substrate material. In this case, it is necessary to consider the specific contact resistivity with the thin film at the same time. The temperature coefficient of the specific contact resistivity is positive and works in a direction to eliminate the TCR of the TaN thin film. It was found that Ti was excellent as an intermediate layer for a TaN thin film in a region having a low TCR.
[0036]
As described above, in consideration of the contact resistance of the electrode film, the TCR can be suppressed as small as possible over the entire operating temperature range, and a tantalum nitride thin film resistor that is stable with time and highly accurate can be supplied by an easy method. It has become possible.
[Brief description of the drawings]
FIG. 1 is a sectional view of the basic structure of a tantalum nitride thin film resistor of the present invention.
FIG. 2 is a top view of a submount substrate on which the tantalum nitride thin film resistor of the present invention is mounted.
FIG. 3 is a top view and a cross-sectional view of a transmission line model sample of a tantalum nitride thin film resistor according to the present invention.
FIG. 4 is a graph showing an X-ray diffraction result of a tantalum nitride thin film deposited on a Si substrate.
FIG. 5 is a graph showing a relationship between a distance between electrodes and a resistance value of a tantalum nitride thin film resistor.
FIG. 6 is a graph showing the results of elemental analysis using Auger electron spectroscopy when the nitrogen partial pressure ratio is 15%.
FIG. 7 is a graph showing the results of elemental analysis using Auger electron spectroscopy when the nitrogen partial pressure ratio is 25%.
FIG. 8 is a graph showing the TCR of a tantalum nitride thin film resistor when the nitrogen partial pressure ratio is 0%.
FIG. 9 is a graph showing the TCR of a tantalum nitride thin film resistor when the nitrogen partial pressure ratio is 15%.
FIG. 10 is a graph showing the TCR of a tantalum nitride thin film resistor when the nitrogen partial pressure ratio is 25%.
FIG. 11 is a graph showing a resistance-temperature characteristic of a specific contact resistivity with respect to a tantalum nitride thin film resistor when a nitrogen partial pressure ratio is 0%.
FIG. 12 is a graph showing a resistance-temperature characteristic of a specific contact resistivity with respect to a tantalum nitride thin film resistor when a nitrogen partial pressure ratio is 15%.
FIG. 13 is a graph showing a resistance-temperature characteristic of a specific contact resistivity with respect to a tantalum nitride thin film resistor when a nitrogen partial pressure ratio is 25%.
FIG. 14 is a graph illustrating the TCR of a tantalum nitride thin film resistor as a function of nitrogen flow rate.
FIG. 15 is a process flow chart showing a method for manufacturing a tantalum nitride thin film resistor of the present invention.
FIG. 16 is a graph showing the results of a current heating test of the tantalum nitride thin film resistor of the present invention.
FIG. 17 is a graph showing the results of a high-temperature and high-humidity test of the tantalum nitride thin film resistor of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Insulating substrate 2 Tantalum nitride thin film resistor 3 Intermediate film 4 Electrode film

Claims

In a tantalum nitride thin film resistor in which an electrode film is formed on an upper surface of a tantalum nitride thin film formed on an insulating substrate via an intermediate film, a combined resistance temperature coefficient of the intermediate film and the electrode film is defined as a first resistance temperature coefficient, When the resistance temperature coefficient of the thin-film resistor is a second resistance temperature coefficient, the sum of the first and second temperature coefficients is from -10 ppm / ° C to 0 ppm / ° C. Resistor.

2. The tantalum nitride thin film resistor according to claim 1, wherein the material of the intermediate film is Ti, the material of the electrode film is Au, and the material of the insulating substrate is AlN.

In a method of manufacturing a tantalum nitride thin film resistor for forming an electrode film via an intermediate film on an upper surface of a tantalum nitride thin film formed on an insulating substrate,
Calculating a first temperature coefficient of resistance, which is a combined temperature coefficient of resistance of the intermediate film and the electrode film, using a transmission line model in advance;
Forming a tantalum nitride thin film on the entire surface of one surface of the insulating substrate by RF reactive sputtering in a nitrogen atmosphere, and setting a nitrogen partial pressure ratio at this time to 10 to 20%;
A method for manufacturing a tantalum nitride thin film resistor, comprising a step of forming a desired intermediate film pattern and an electrode film pattern by a combination of a photoresist method and a DC sputtering method.

4. The method of manufacturing a tantalum nitride thin film resistor according to claim 3, wherein said nitrogen partial pressure ratio is 13 to 17%.

5. The method according to claim 3, wherein the material of the intermediate film is Ti, the material of the electrode film is Au, and the material of the insulating substrate is AlN. .