JP2004189584A

JP2004189584A - Method of measuring point defect distribution in silicon single crystal ingot

Info

Publication number: JP2004189584A
Application number: JP2003207860A
Authority: JP
Inventors: Kazunari Kurita; 一成栗田
Original assignee: Sumitomo Mitsubishi Silicon Corp
Current assignee: Sumco Corp
Priority date: 2002-10-18
Filing date: 2003-08-19
Publication date: 2004-07-08
Anticipated expiration: 2023-08-19
Also published as: JP4200845B2

Abstract

<P>PROBLEM TO BE SOLVED: To identify a region [Pv], a region [Pi] and the boundary of these with high accuracy in a short time without depending on the oxygen concentration dissolved in an ingot. <P>SOLUTION: A single crystal ingot is cut along the axial direction including the center axis of the ingot to obtain a sample for measurement including a region [V], a region [Pv], a region [Pi] and a region [I], and the sample is cut into halves symmetric around the center axis to prepare first and second samples. The surface of the first sample is contaminated with a first transition metal, while the surface of the second sample is contaminated with a second transition metal different from the first transition metal. The first and second samples contaminated with metals are heat treated to diffuse the first and second transition metals into the respective samples. The recombination life time in each of the whole first and second samples is measured, and the obtained measurement results in the vertical direction are superposed, from which the boundary between the region [Pi] and the region [I], and the boundary between the region [V] and the [Pv] are determined. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

【０００１】
【発明の属する技術分野】
本発明は、チョクラルスキー法（以下、ＣＺ法という。）により引上げられたシリコン単結晶インゴット（以下、インゴットという。）内に形成される真性点欠陥の欠陥分布を測定する方法に関する。更に詳しくは、無欠陥領域シリコンウェーハの製造を容易にするシリコンウェーハの真性点欠陥の欠陥分布評価方法に関するものである。
【０００２】
【従来の技術】
近年の半導体集積回路の超微細化にともないデバイスの歩留まりを低下させる要因として、シリコン単結晶インゴットの結晶成長中に形成される結晶成長時導入欠陥である結晶に起因したパーティクル（ＣｒｙｓｔａｌＯｒｉｇｉｎａｔｅｄＰａｒｔｉｃｌｅ、以下、ＣＯＰという。）や、酸化誘起積層欠陥（ＯｘｉｄａｔｉｏｎＩｎｄｕｃｅｄＳｔａｃｋｉｎｇＦａｕｌｔ、以下、ＯＩＳＦという。）の核となる酸素析出物の微小欠陥や、或いは侵入型転位（Ｉｎｔｅｒｓｔｉｔｉａｌ−ｔｙｐｅＬａｒｇｅＤｉｓｌｏｃａｔｉｏｎ、以下、Ｌ／Ｄという。）の存在が挙げられている。
ＣＯＰは、鏡面研磨されたシリコンウェーハをアンモニアと過酸化水素の混合液でＳＣ−１洗浄すると、ウェーハ表面に出現する結晶起因のピットである。このウェーハをパーティクルカウンタで測定すると、このピットがパーティクル（ＬｉｇｈｔＰｏｉｎｔＤｅｆｅｃｔ、ＬＰＤ）として検出される。ＣＯＰは電気的特性、例えば酸化膜の経時絶縁破壊特性（ＴｉｍｅＤｅｐｅｎｄｅｎｔｄｉｅｌｅｃｔｒｉｃＢｒｅａｋｄｏｗｎ、ＴＤＤＢ）、酸化膜耐圧特性（ＴｉｍｅＺｅｒｏＤｉｅｌｅｃｔｒｉｃＢｒｅａｋｄｏｗｎ、ＴＺＤＢ）等を劣化させる原因となる。またＣＯＰがウェーハ表面に存在するとデバイスの配線工程において段差を生じ、断線の原因となり得る。そして素子分離部分においてもリーク等の原因となり、製品の歩留りを低くする。
【０００３】
ＯＩＳＦは、結晶成長時に形成される微小な酸素析出が核となっていると考えられ、半導体デバイスを製造する際の熱酸化工程等で顕在化する積層欠陥である。このＯＩＳＦは、デバイスのリーク電流を増加させる等の不良原因になる。Ｌ／Ｄは、転位クラスタとも呼ばれたり、或いはこの欠陥を生じたシリコンウェーハをフッ酸を主成分とする選択エッチング液に浸漬すると方位を持ったエッチングピットを生じることから転位ピットとも呼ばれる。このＬ／Ｄも、電気的特性、例えばリーク特性、アイソレーション特性等を劣化させる原因となる。
以上のことから、半導体集積回路を製造するために用いられるシリコンウェーハからＣＯＰ、ＯＩＳＦ及びＬ／Ｄを減少させることが必要となっている。
【０００４】
このＣＯＰ、ＯＩＳＦ及びＬ／Ｄを有しない無欠陥のインゴット及びこのインゴットからスライスされたシリコンウェーハが開示されている（例えば、特許文献１及び２参照。）。この無欠陥のインゴットは、インゴット内での空孔型点欠陥の凝集体及び格子間シリコン型点欠陥の凝集体がそれぞれ検出されないパーフェクト領域を［Ｐ］とするとき、パーフェクト領域［Ｐ］からなるインゴットである。パーフェクト領域［Ｐ］は、インゴット内で空孔型点欠陥が優勢であって過飽和な空孔が凝集した欠陥を有する領域［Ｖ］と、格子間シリコン型点欠陥が優勢であって過飽和な格子間シリコンが凝集した欠陥を有する領域［Ｉ］との間に介在する。
【０００５】
また、点欠陥が凝集した欠陥を有しないパーフェクト領域［Ｐ］が空孔型点欠陥が優勢である領域［Ｐｖ］と、格子間シリコン型点欠陥が優勢である領域［Ｐｉ］とに分類されることが示される（例えば、特許文献３参照。）。領域［Ｐｖ］は領域［Ｖ］に隣接し、かつＯＩＳＦ核を形成し得る最低の空孔型点欠陥濃度未満の空孔型点欠陥濃度を有する領域である。領域［Ｐｉ］は領域［Ｉ］に隣接し、かつ侵入型転位を形成し得る最低の格子間シリコン型点欠陥濃度未満の格子間シリコン型点欠陥濃度を有する領域である。
パーフェクト領域［Ｐ］からなるインゴットは、インゴットの引上げ速度をＶ（ｍｍ／分）とし、シリコン融液とシリコンインゴットの固液界面近傍におけるインゴット鉛直方向の温度勾配をＧ（℃／ｍｍ）とするとき、熱酸化処理をした際にリング状に発生するＯＩＳＦ（Ｐバンド）がウェーハ中心部で消滅し、かつＬ／Ｄ（Ｂバンド）を発生しない領域のＶ／Ｇ（ｍｍ^２／分・℃）の範囲内で作られる。
【０００６】
無欠陥のシリコンウェーハを製造するためには、軸方向及び径方向の点欠陥濃度分布を制御することが重要であることから、従来、インゴット内の熱処理によって生じた２次欠陥、即ち凝集した欠陥分布を測定するためには次の方法が採られていた。先ず軸方向及び径方向の点欠陥濃度分布評価には無欠陥領域が形成される結晶引上げ条件でシリコン単結晶インゴットを作製する。次いでインゴットを軸方向にスライスしてサンプルを作製する。次にこのサンプルをミラーエッチングした後、窒素又は酸化性雰囲気下で８００℃で４時間熱処理し、更に続いて１０００℃で１６時間熱処理する。この熱処理したサンプルを銅デコレーション（ｃｏｐｐｅｒｄｅｃｏｒａｔｉｏｎ）、セコエッチング（ｓｅｃｃｏ−ｅｔｃｈｉｎｇ）、Ｘ線トポグラフ像（Ｘ−ＲａｙＴｏｐｏｇｒａｐｈｙ）分析、ライフタイム（ｌｉｆｅｔｉｍｅ）測定等の方法により測定する。図１３及び図１４に示すように、上記熱処理により酸素析出物がインゴット内に出現するため、この酸素析出物から各領域及び各境界を識別、判別していた。ここで図１３は、無欠陥のシリコン単結晶インゴットを作製可能なホットゾーンを有する結晶引上げ装置で引上げ速度を高速から低速に変化させて作製された領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含むサンプルに酸素析出熱処理した場合の再結合ライフタイムの面内分布を示す図であり、図１４は図１３のＡ−Ａ線断面における再結合ライフタイムを示す図である。
【０００７】
【特許文献１】
米国特許第６，０４５，６１０号明細書
【特許文献２】
特開平１１−１３９３号公報
【特許文献３】
特開２００１−１０２３８５号公報
【０００８】
【発明が解決しようとする課題】
しかしながら、上記測定方法の場合、再結合ライフタイムの測定値がサンプル内の酸素濃度、酸素析出熱処理条件に著しく依存することが知られている。例えば、インゴット内に固溶する酸素濃度が低濃度である場合、酸素析出熱処理によって形成される酸素析出物の密度は、酸素濃度が高濃度に固溶するサンプルと比較して低密度であることから、再結合ライフタイムの測定値の差分値が小さくなる。
このように低酸素濃度、高酸素濃度において、領域［Ｐｖ］と領域［Ｐｉ］との境界領域を明確にすることができない問題があった。
更に、酸素析出熱処理には長時間の熱処理が必要なだけでなく、熱処理条件によってインゴット内の酸素析出に影響を与えることから点欠陥領域の境界領域が領域［Ｐｖ］又は領域［Ｐｉ］のいずれかの領域側にシフトすることになる。従って、本来の点欠陥領域を識別、判別することが困難となっていた。
【０００９】
本発明の目的は、インゴットに固溶する酸素濃度に依存することなくインゴット内の領域［Ｐｖ］及び領域［Ｐｉ］並びにこれらの境界を高い精度でかつ短時間に識別する、シリコン単結晶インゴットの点欠陥分布を測定する方法を提供することにある。
【００１０】
【課題を解決するための手段】
請求項１に係る発明は、図１に示すように、（ａ）シリコン融液から引上げ速度を変えて引上げられたシリコン単結晶インゴットを軸方向に単結晶インゴット中心軸を含むように切断して、領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含む測定用サンプルを作製する工程と、（ｂ）測定用サンプルをインゴット中心軸に対して対称となるように２分割して第１及び第２サンプルを作製する工程と、（ｃ）第１遷移金属が１〜１０００ｐｐｍの濃度で溶解している第１遷移金属溶液を第１サンプルの表面に塗布して金属汚染する工程と、（ｄ）第１遷移金属と異なる第２遷移金属が１〜１０００ｐｐｍの濃度で溶解している第２遷移金属溶液を第２サンプルの表面に塗布して金属汚染する工程と、（ｅ）金属汚染された第１及び第２サンプルをアルゴン、窒素、酸素、水素、又はそれらの混合ガス雰囲気下、６００℃〜１１５０℃で０．５時間〜３０時間熱処理して第１及び第２サンプル表面にそれぞれ塗布した第１及び第２遷移金属をサンプル内部に拡散させる拡散熱処理工程と、（ｆ）熱処理された第１及び第２サンプル全体における再結合ライフタイムをそれぞれ測定する工程と、（ｇ）上記（ｆ）工程の測定結果の鉛直方向の測定値を重ね合わせる工程と、（ｈ）上記（ｇ）工程の測定結果から、領域［Ｐｉ］と領域［Ｉ］の境界及び領域［Ｖ］と領域［Ｐｖ］の境界をそれぞれ規定する工程とを含むシリコン単結晶インゴットの点欠陥分布を測定する方法である。
但し、領域［Ｖ］は空孔型点欠陥が優勢であって過剰な空孔が凝集した欠陥を有する領域、領域［Ｐｖ］は空孔型点欠陥が優勢であって空孔が凝集した欠陥を有しない領域、領域［Ｐｉ］は格子間シリコン型点欠陥が優勢であって格子間シリコンが凝集した欠陥を有しない領域及び領域［Ｉ］は格子間シリコン型点欠陥が優勢であって格子間シリコンが凝集した欠陥を有する領域である。
【００１１】
請求項１に係る発明では、高い精度で領域［Ｖ］と領域［Ｐｖ］の境界を規定する第１遷移金属の再結合ライフタイム分布と、高い精度で領域［Ｐｉ］と領域［Ｉ］の境界を規定する第２遷移金属の再結合ライフタイム分布を組合わせることで、インゴット内に固溶する酸素濃度に依存することなく、高い精度でかつ短時間に領域［Ｐｖ］及び領域［Ｐｉ］並びにこれらの境界を規定する。
【００１２】
請求項２に係る発明は、図７に示すように、（ａ）シリコン融液から引上げ速度を変えて引上げられたシリコン単結晶インゴットを軸方向に単結晶インゴット中心軸を含むように切断して、領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含む測定用サンプルを作製する工程と、（ｂ）測定用サンプルをインゴット中心軸に対して対称となるように２分割して第１及び第２サンプルを作製する工程と、（ｃ）第１遷移金属が１〜１０００ｐｐｍの濃度で溶解している第１遷移金属溶液を第１サンプルの表面に塗布して金属汚染する工程と、（ｄ）第１遷移金属と異なる第２遷移金属が１〜１０００ｐｐｍの濃度で溶解している第２遷移金属溶液を第２サンプルの表面に塗布して金属汚染する工程と、（ｅ）金属汚染された第１及び第２サンプルをアルゴン、窒素、酸素、水素、又はそれらの混合ガス雰囲気下、６００℃〜１１５０℃で０．５時間〜３０時間熱処理して第１及び第２サンプル表面にそれぞれ塗布した第１及び第２遷移金属をサンプル内部に拡散させる拡散熱処理工程と、（ｆ）熱処理された第１及び第２サンプル全体における再結合ライフタイムをそれぞれ測定する工程と、（ｇ）上記（ｆ）工程の測定結果の鉛直方向の測定値を重ね合わせる工程と、（ｉ）熱処理された第１サンプルをＴＩＤ法により第１遷移金属の鉛直方向の濃度を求める工程と、（ｊ）熱処理された第２サンプルをＤＬＴＳ法により第２遷移金属の鉛直方向の濃度を求める工程と、（ｋ）上記（ｆ）工程の第１サンプルにおける鉛直方向の測定結果と上記（ｉ）工程の測定結果から、第１遷移金属濃度と再結合ライフタイムの相関直線を作成する工程と、（ｌ）上記（ｆ）工程の第２サンプルにおける鉛直方向の測定結果と上記（ｊ）工程の測定結果から、第２遷移金属濃度と再結合ライフタイムの相関直線を作成する工程と、（ｍ）上記（ｇ）工程の測定結果、上記（ｋ）工程の相関直線及び上記（ｌ）工程の相関直線から、領域［Ｖ］及び領域［Ｐｖ］の境界、領域［Ｐｉ］及び領域［Ｉ］の境界をそれぞれ規定する工程とを含むシリコン単結晶インゴットの点欠陥分布を測定する方法である。
但し、領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］は請求項１にそれぞれ記載した意味と同じであり、ＴＩＤ法とはシリコン単結晶中に固溶するＣｕ濃度を金属半導体接合ダイオードの過渡容量特性の解析から定量化する方法であり、ＤＬＴＳ法とは、接合（或いは界面）に逆方向電界をかけた状態で正方向のパルス電圧を加え、空乏層内の準位にキャリアを捕獲させる方法である。
【００１３】
請求項２に係る発明では、第１遷移金属の再結合ライフタイム分布と第２遷移金属の再結合ライフタイム分布を組合わせ、更に、第１遷移金属の金属濃度と第１遷移金属の再結合ライフタイム分布から相関直線を、同様に、第２遷移金属の金属濃度と第２遷移金属の再結合ライフタイム分布から相関直線をそれぞれ作成し、組合わせた再結合ライフタイム分布と２つの相関直線から、インゴット内に固溶する酸素濃度に依存することなく、高い精度でかつ短時間に領域［Ｐｖ］及び領域［Ｐｉ］並びにこれらの境界を規定する。
【００１４】
請求項３に係る発明は、請求項１に係る発明であって、第１遷移金属がＣｕ又はＦｅであって、第２遷移金属がＮｉ又はＣｏである測定方法である。
請求項３に係る発明では、上記第１及び第２遷移金属のうち、第１遷移金属をＣｕ、第２遷移金属をＮｉの組合わせにすると、領域［Ｐｖ］及び領域［Ｐｉ］の再結合ライフタイムが顕著に異なるため、領域［Ｐｖ］及び領域［Ｐｉ］の境界をより明確に判別できる。
請求項４に係る発明は、請求項２に係る発明であって、第１遷移金属がＣｕであって、第２遷移金属がＦｅである測定方法である。
【００１５】
請求項５に係る発明は、請求項１又は２に係る発明であって、（ｅ）工程における第１遷移金属及び第２遷移金属の拡散熱処理を６００℃〜１１５０℃で０．５時間〜２４時間熱処理する測定方法である。
請求項６に係る発明は、請求項１又は２に係る発明であって、（ｆ）工程における再結合ライフタイム測定をＬＭ−ＰＣＤ（レーザ／マイクロ波光導電率減衰法）を用いて測定する測定方法である。
【００１６】
請求項７に係る発明は、請求項１又は３に係る発明であって、第２サンプルに金属汚染させる第２遷移金属がＮｉであるとき、（ｅ）工程で熱処理された第２サンプル表面を選択エッチングする工程を更に含む測定方法である。
第２遷移金属がＮｉの場合、第２サンプルに拡散熱処理を施すとサンプル表面にはＮｉシリサイドが形成される。このＮｉシリサイドが形成されると、次に続く再結合ライフタイム測定を施す工程において、表面再結合速度の増加を招き、正確な数値が測定できない不具合を生じる。そのため、サンプル表面を選択エッチングしてＮｉシリサイドを除去する。
【００１７】
【発明の実施の形態】
（１）本発明の測定方法の対象となるインゴット
本発明の測定方法の対象となるインゴットは、シリコン融液から領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含むようにＶ／Ｇ比を制御して引上げられる。
またこのインゴットはその酸素濃度が８．０×１０^１７ａｔｏｍｓ／ｃｍ^３〜１．０×１０^１８ａｔｏｍｓ／ｃｍ^３（旧ＡＳＴＭ、以下同じ）の範囲にあるか、又はこのインゴットをスライスされたサンプルを窒素雰囲気下で８００℃で４時間熱処理し、更に続いて１０００℃で１６時間熱処理した後に再結合ライフタイムを測定したときにこのサンプルにおける、領域［Ｐｖ］と領域［Ｐｉ］の境界を識別不能であるインゴットである。
【００１８】
こうしたインゴットは、ＣＺ法又は磁場印加ＣＺ法によりホットゾーン炉内のシリコン融液からボロンコフ（Ｖｏｒｏｎｋｏｖ）の理論に基づいて引上げられる。一般的に、ＣＺ法又は磁場印加ＣＺ法によりホットゾーン炉内のシリコン融液からシリコン単結晶のインゴットを引上げたときには、シリコン単結晶における欠陥として、点欠陥（ｐｏｉｎｔｄｅｆｅｃｔ）と点欠陥の凝集体（ａｇｇｌｏｍｅｒａｔｅｓ：三次元欠陥）が発生する。点欠陥は空孔型点欠陥と格子間シリコン型点欠陥という二つの一般的な形態がある。空孔型点欠陥は一つのシリコン原子がシリコン結晶格子で正常的な位置の一つから離脱したものである。このような空孔が凝集し空孔型欠陥を形成する。一方、格子シリコンが凝集し格子間シリコン型欠陥を形成する。
【００１９】
点欠陥は一般的にシリコン融液（溶融シリコン）とインゴット（固状シリコン）の間の接触面で導入される。しかし、インゴットを継続的に引上げることによって接触面であった部分は引上げとともに冷却していく。冷却の間、空孔又は格子間シリコンは拡散や対消滅反応をする。約１１００℃まで冷却した時点での過剰な点欠陥が空孔型欠陥の凝集体（ｖａｃａｎｃｙａｇｇｌｏｍｅｒａｔｅｓ）又は格子間シリコン型欠陥の凝集体（ｉｎｔｅｒｓｔｉｔｉａｌａｇｇｌｏｍｅｒａｔｅｓ）を形成する。言い換えれば、過剰な点欠陥が凝集体を形成して発生する三次元構造である。
空孔型欠陥の凝集体は前述したＣＯＰの他に、ＬＳＴＤ（ＬａｓｅｒＳｃａｔｔｅｒｉｎｇＴｏｍｏｇｒａｐｈＤｅｆｅｃｔｓ）又はＦＰＤ（ＦｌｏｗＰａｔｔｅｒｎＤｅｆｅｃｔｓ）と呼ばれる欠陥を含み、格子間シリコン型欠陥の凝集体は前述したＬ／Ｄと呼ばれる欠陥を含む。ＦＰＤとは、インゴットをスライスして作製されたシリコンウェーハを３０分間セコエッチング（Ｓｅｃｃｏｅｔｃｈｉｎｇ、ＨＦ：Ｋ_２Ｃｒ_２Ｏ_７（０．１５ｍｏｌ／ｌ）＝２：１の混合液によるエッチング）したときに現れる特異なフローパターンを呈する痕跡の源であり、ＬＳＴＤとは、シリコン単結晶内に赤外線を照射したときにシリコンとは異なる屈折率を有し散乱光を発生する源である。
【００２０】
ボロンコフの理論は、欠陥の数が少ない高純度インゴットを成長させるために、インゴットの引上げ速度をＶ（ｍｍ／分）、シリコン融液とシリコンインゴットの固液界面近傍の温度勾配をＧ（℃／ｍｍ）とするときに、Ｖ／Ｇ（ｍｍ^２／分・℃）を制御することである。この理論では、図２に示すように、Ｖ／Ｇをよこ軸にとり、空孔型欠陥濃度と格子間シリコン型欠陥濃度を同一のたて軸にとって、Ｖ／Ｇと点欠陥濃度との関係を図式的に表現し、空孔領域と格子間シリコン領域の境界がＶ／Ｇによって決定されることを説明している。より詳しくは、Ｖ／Ｇ比が臨界点（Ｖ／Ｇ）_ｃ以上では空孔濃度が上昇したインゴットが形成される反面、Ｖ／Ｇ比が臨界点（Ｖ／Ｇ）_ｃ以下では格子間シリコン濃度が上昇したインゴットが形成される。図２において、［Ｉ］は格子間シリコンが優勢であって格子間シリコンが凝集した欠陥を有する領域（（Ｖ／Ｇ）_１以下）を示し、［Ｖ］は空孔が優勢であって空孔が凝集した欠陥を有する領域（（Ｖ／Ｇ）_２以上）を示し、［Ｐ］は空孔型欠陥の凝集体及び格子間シリコン型欠陥の凝集体が存在しないパーフェクト領域（（Ｖ／Ｇ）_１〜（Ｖ／Ｇ）_２）を示す。領域［Ｐ］に隣接する側の領域［Ｖ］との境界にはＯＩＳＦ核を形成するＰバンド領域（（Ｖ／Ｇ）_２〜（Ｖ／Ｇ）_３）が存在する。Ｐバンド領域は微小な板状析出物が存在し、酸化性雰囲気下での熱処理でＯＩＳＦ（積層欠陥）が形成される。また領域［Ｐ］に隣接する側の領域［Ｉ］の境界にはＢバンド領域（（Ｖ／Ｇ）_４〜（Ｖ／Ｇ）_１）が存在する。Ｂバンド領域とは、格子間シリコンの凝集体が核となり熱処理によって酸素析出が高濃度に発生している領域である。
【００２１】
このパーフェクト領域［Ｐ］は更に領域［Ｐｉ］と領域［Ｐｖ］に分類される。
［Ｐｉ］はＶ／Ｇ比が上記（Ｖ／Ｇ）_１から臨界点までの領域であり、格子間シリコンが優勢であって凝集した欠陥を有しない領域である。［Ｐｖ］はＶ／Ｇ比が臨界点から上記（Ｖ／Ｇ）_２までの領域であり、空孔が優勢であって凝集した欠陥を有しない領域である。即ち、［Ｐｉ］は領域［Ｉ］に隣接し、かつ侵入型転位を形成し得る最低の格子間シリコン濃度未満の格子間シリコン濃度を有する領域であり、［Ｐｖ］は領域［Ｖ］に隣接し、かつＯＩＳＦ核を形成し得る最低の空孔濃度未満の空孔濃度を有する領域である。
従って、本発明の測定方法の対象となるインゴットは、領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含むためには、Ｖ／Ｇ比が上記（Ｖ／Ｇ）_４以下から臨界点、（Ｖ／Ｇ）_３以上とそれぞれの領域を含むように制御されて引上げられることになる。
（２）本発明の第１測定方法
次に図１及び図３に基づいて本発明の第１測定方法を説明する。
図１に示すように、先ずテストインゴットから測定用サンプルを作製する（工程（ａ））。即ち、ＣＺ法又は磁場印加ＣＺ法に基づく引上げ装置の石英るつぼに貯留されたシリコン融液からインゴットを引上げる。このとき上述した領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］がインゴットの軸方向に含むように、インゴットの引上げ速度Ｖ（ｍｍ／分）を高速（トップ側）から低速（ボトム側）又は低速（ボトム側）から高速（トップ側）に速度を変えて引上げる。
【００２２】
図３に示すように、テストで得られたインゴットを軸方向にインゴット中心軸を含むように切断し、かつミラーエッチングすることにより、５００〜２０００μｍの厚さを有する、表面が鏡面化した測定用サンプルが作られる（図３（ａ）及び図３（ｂ））。この測定用サンプルには領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］がそれぞれ含まれる。図３（ａ）に示されるインゴットは、引上げ速度Ｖ（ｍｍ／分）を高速（トップ側）から低速（ボトム側）に速度を変えて引上げたインゴットである。
【００２３】
測定用サンプルに固溶する酸素濃度をＦＴ−ＩＲ（Ｆｏｕｒｉｅｒｔｒａｎｓｆｏｒｍｉｎｆｒａｒｅｄａｂｓｏｒｐｔｉｏｎｓｐｅｃｔｒｏｓｃｏｐｙ）法により測定し、酸素濃度が１．０×１０^１８ａｔｏｍｓ／ｃｍ^３以上である場合には、窒素雰囲気下で８００℃で４時間熱処理し、更に続いて１０００℃で１６時間熱処理する。この熱処理した測定用サンプルの再結合ライフタイムをＬＭ−ＰＣＤ法でサンプル全体にわたって測定する。図１３は測定結果の一例である。図１３に示すように、上記２段熱処理により酸素析出物が高密度にサンプル内に出現する。その濃度分布に応じて、サンプル全体が領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］が明瞭に識別される。領域［Ｖ］内にはＯＩＳＦ核を形成するＰバンド領域が存在し、領域［Ｉ］内にはＢバンド領域が存在する。
【００２４】
一方、測定用サンプルの酸素濃度が１．０×１０^１８ａｔｏｍｓ／ｃｍ^３未満である場合には、上記と同様に測定用サンプルを熱処理して測定用サンプルの再結合ライフタイムの測定しても、図５に示すように、この再結合ライフタイムの測定結果からでは、パーフェクト領域［Ｐ］が領域［Ｐｖ］であるか、又は領域［Ｐｉ］であるかを識別することができない。
【００２５】
このため、本発明では、測定用サンプルを図３（ｂ）の矢印に示す位置でインゴット中心軸に対して対称となるように２分割する（工程（ｂ））。上記工程（ｂ）により分割した測定用サンプルを、図３（ｃ）に示すように、第１サンプル及び第２サンプルとする。
図１に戻って、第１遷移金属が１〜１０００ｐｐｍの濃度で溶解している第１遷移金属溶液を第１サンプルの表面に塗布してこの第１サンプルを金属汚染する（工程（ｃ））。第１遷移金属としてはＣｕ、Ｆｅが好ましい。第１遷移金属溶液は第１遷移金属が１〜１０００ｐｐｍ、好ましくは１〜１００ｐｐｍの濃度で溶解している溶液である。特にＣｕ又はＦｅが１０〜１００ｐｐｍの濃度で溶解している原子吸光用標準溶液が入手が可能でかつ濃度精度に優れるため好ましい。
塗布方法としては、スピンコート法、ディッピング法が挙げられる。第１遷移金属溶液の濃度が下限値未満では、各領域の境界が十分に判別できず、再結合ライフタイムを高精度に測定することができない。また上限値を越えるとサンプルの表面及び裏面に遷移金属が拡散して表裏面の近傍で金属シリサイド（析出物）を形成することから、サンプル内部（結晶中）に深いエネルギ準位である再結合中心を形成することができない。
【００２６】
第１遷移金属と異なる第２遷移金属が１〜１０００ｐｐｍの濃度で溶解している第２遷移金属溶液を第２サンプルの表面に塗布してこの第２サンプルを金属汚染する（工程（ｄ））。第２遷移金属としてはＮｉ、Ｃｏが好ましい。第２遷移金属溶液は第２遷移金属が１〜１０００ｐｐｍ、好ましくは１〜１００ｐｐｍの濃度で溶解している溶液である。特にＮｉ又はＣｏが１０〜１００ｐｐｍの濃度で溶解している原子吸光用標準溶液が入手が可能でかつ濃度精度に優れるため好ましい。塗布方法は上記第１遷移金属の塗布方法と同様の方法を使用する。第２遷移金属溶液の濃度が下限値未満では、各領域の境界が十分に判別できず、再結合ライフタイムを高精度に測定することができない。また上限値を越えるとサンプルの表面及び裏面に遷移金属が拡散して表裏面の近傍で金属シリサイド（析出物）を形成することから、サンプル内部（結晶中）に深いエネルギ準位である再結合中心を形成することができない。
【００２７】
次に、第１及び第２サンプルをアルゴン、窒素、酸素、水素又はそれらの混合ガス雰囲気下、６００℃〜１１５０℃で０．５時間〜３０時間熱処理して、第１及び第２サンプル表面にそれぞれ塗布した第１及び第２遷移金属をサンプル全体に拡散させる（工程（ｅ））。熱処理の方法として、０．５〜１０℃／分の速度で昇温して６００〜１１５０℃の温度で０．５〜３０時間、好ましくは０．５時間〜２４時間熱処理する。５〜１０℃／分の速度で昇温して９００〜１０００℃で１〜２時間熱処理することがより好ましい。熱処理の時間及び温度が下限値未満では遷移金属が十分にサンプル内に拡散しない。また上限値を越えるとサンプルの表面及び裏面に遷移金属が拡散して表裏面の近傍で金属シリサイド（析出物）を形成することから、サンプル内部（結晶中）に深いエネルギ準位である再結合中心を形成することができない。
【００２８】
第２サンプルに汚染させた第２遷移金属がＮｉの場合、熱処理した第２サンプルの再結合ライフタイムを測定すると、図５に示すように、領域［Ｐｉ］と領域［Ｉ］の再結合ライフタイムが十分に増加せず、この領域間の境界を識別することができない。Ｎｉは第１遷移金属のＣｕやＦｅと比べて拡散係数が大きいため熱処理によりサンプル表面近傍に析出しているためである。そこでＮＨ_３とＨＦとＣＨ_３ＯＯＨを含むエッチング液でサンプル表面をエッチングする。このエッチングにより表面近傍に存在するＮｉは除去される。この状態で再度、第２サンプルの再結合ライフタイムを測定すると、化学エッチング後では、領域［Ｐｉ］と領域［Ｉ］での再結合ライフタイムの増加が顕著であり、一方、領域［Ｖ］は、化学エッチング前後での再結合ライフタイムに変化がないことが判る。
【００２９】
次に、熱処理された第１及び第２サンプルについて遷移金属の再結合ライフタイムの測定を行う（工程（ｆ））。測定にはＬＭ−ＰＣＤ（レーザ／マイクロ波光導電率減衰法）が好ましい。ＬＭ−ＰＣＤによりサンプル全体にわたって再結合ライフタイムを定量化する。第１及び第２サンプルの再結合ライフタイムを図４にそれぞれ示す。図４に示すように、第１サンプルは再結合ライフタイムが最も低い領域は、領域［Ｐｉ］であり、最も高いのは領域［Ｐｖ］であった。第２サンプルでは再結合ライフタイムが最も低い領域は、領域［Ｐｖ］であり、最も高いのは領域［Ｐｉ］であった。
【００３０】
第１及び第２サンプルの再結合ライフタイムを測定し、この測定結果を鉛直方向に分布を重ね合わせる（工程（ｇ））。第１遷移金属にＣｕを、第２遷移金属にＮｉを用い、第２サンプルに選択エッチングを施した第１及び第２サンプルの再結合ライフタイムを測定し、鉛直方向に分布の重ね合わせを行った結果を図６に示す。図６において、それぞれよこ軸はサンプルを切断する前のインゴットの軸方向に相当する。
【００３１】
Ｃｕが拡散した第１サンプルの場合は領域［Ｐｖ］の再結合ライフタイムがその他の点欠陥領域と比較して高い。一方、Ｎｉが拡散した第２サンプルの場合は、再結合ライフタイムが最も高い領域は領域［Ｐｉ］であることが判る。従って、遷移金属の種類と点欠陥領域の再結合ライフタイム依存性から領域［Ｐｖ］及び領域［Ｐｉ］を識別することができる。
【００３２】
結晶の同一部位から作製されたサンプルにＮｉ及びＣｕをそれぞれのサンプルに表面汚染し拡散処理後の再結合ライフタイム分布を測定することにより遷移金属の種類によって各種点欠陥領域での再結合ライフタイムが顕著に異なる性質を利用して領域［Ｐｖ］及び領域［Ｐｉ］の識別を実現できる。
【００３３】
このように本発明の第１測定方法では、インゴット内に固溶する酸素を測定対象とした識別方法ではないため、インゴット内に固溶する酸素濃度に依存することがない。また酸素析出熱処理条件に左右されることがないため、高い精度で識別することができる。
（３）本発明の第２測定方法
次に図７に基づいて本発明の第２測定方法を説明する。
先ず、工程（ａ）及び工程（ｂ）は前述した第１測定方法と同様である。次いで、図７に示すように、第１遷移金属が１〜１０００ｐｐｍの濃度で溶解している第１遷移金属溶液を第１サンプルの表面に塗布してこの第１サンプルを金属汚染する（工程（ｃ））。ここでの第１遷移金属にはＣｕが選択される。次に、第１遷移金属と異なる第２遷移金属が１〜１０００ｐｐｍの濃度で溶解している第２遷移金属溶液を第２サンプルの表面に塗布してこの第２サンプルを金属汚染する（工程（ｄ））。ここでの第２遷移金属にはＦｅが選択される。工程（ｅ）では前述した第１測定方法と同様の条件で第１及び第２サンプルに拡散熱処理を施す。
【００３４】
次に、熱処理された第１及び第２サンプルについて遷移金属の再結合ライフタイムの測定を行う（工程（ｆ））。測定にはＬＭ−ＰＣＤが好ましい。ＬＭ−ＰＣＤによりサンプル全体にわたって再結合ライフタイムを定量化する。第１及び第２サンプルの再結合ライフタイムを図８及び図９にそれぞれ示す。図８に示すように、第１サンプルは再結合ライフタイムが最も高い領域は［Ｐｉ］であり、最も低いのは領域［Ｖ］であった。また図９に示すように、第２サンプルでは再結合ライフタイムが最も低い領域は、領域［Ｐｉ］であり、最も高いのは領域［Ｐｖ］であった。
【００３５】
この第１及び第２サンプルの再結合ライフタイム測定の結果を鉛直方向に分布を重ね合わせる（工程（ｇ））。第１及び第２サンプルの再結合ライフタイムを鉛直方向に分布の重ね合わせを行った結果を図１０に示す。図１０において、それぞれよこ軸はサンプルを切断する前のインゴットの軸方向に相当する。Ｃｕが拡散した第１サンプルの場合は領域［Ｐｖ］の再結合ライフタイムがその他の点欠陥領域と比較して高い。一方、Ｆｅが拡散した第２サンプルの場合は、再結合ライフタイムが最も低い領域は領域［Ｐｖ］であることが判る。従って、遷移金属の種類と点欠陥領域の再結合ライフタイム依存性から領域［Ｐｖ］及び領域［Ｐｉ］を識別することができる。
【００３６】
次に、熱処理された第１サンプルをＴＩＤ法により第１遷移金属の鉛直方向の濃度を求める（工程（ｉ））。ＴＩＤ法によりサンプル内に第１遷移金属であるＣｕによって形成される再結合中心の濃度を定量化する。ＴＩＤ（ＴｒａｎｓｉｅｎｔＩｏｎＤｒｉｆｔ）法とはシリコン単結晶中に固溶するＣｕ濃度を金属半導体接合ダイオードの過渡容量特性の解析から定量化する方法である。熱平衡状態においては空間電荷層領域に均一にＣｕｉイオンが分布していると仮定する。金属半導体接合ダイオード（Ｐ型Ｓｉ）に正電圧を印加すると表面近傍のＣｕｉイオンは空間電荷層の端にドリフトし、高濃度なＣｕｉイオンが蓄積される。蓄積状態においては空間電荷層の端では、Ｃｕｉイオンは熱拡散によりその一部が空間電荷層の外側へ拡散する。再び、ゼロバイアスの状態においてはＣｕｉイオンは表面近傍及び空間電荷層の端から熱拡散していくと考えられる。ＴＩＤ法はゼロバイアス状態からの空間電荷層領域に蓄積されたＣｕｉイオン濃度の濃度変化を接合容量の過渡応答減衰特性を測定し定量化を実現できる評価方法である。
【００３７】
続いて、熱処理された第２サンプルをＤＬＴＳ法により第２遷移金属の鉛直方向の濃度を求める（工程（ｊ））。ＤＬＴＳ法によりサンプル内に第２遷移金属であるＦｅによって形成される再結合中心の濃度を定量化する。ＤＬＴＳ（ＤｅｅｐＬｅｖｅｌＴｒａｎｓｉｅｎｔＳｐｅｃｔｒｏｓｃｏｐｙ；過渡容量分光法）法とは、接合（或いは界面）に逆方向電界をかけた状態で正方向のパルス電圧を加え、空乏層内の準位にキャリアを捕獲させる方法である。
【００３８】
再結合ライフタイムと再結合中心濃度の両者には相関があることが推察されることから、工程（ｆ）で得られた第１サンプルの再結合ライフタイム測定値の逆数と工程（ｉ）の測定値とをプロットして第１遷移金属濃度と再結合ライフタイムの相関直線を作成する（工程（ｋ））。図１１に工程（ｋ）で作成した相関直線を示す。更に、工程（ｆ）で得られた第２サンプルの再結合ライフタイム測定値の逆数と工程（ｊ）の測定値とをプロットして第２遷移金属濃度と再結合ライフタイムの相関直線を作成する（工程（ｌ））。図１２に工程（ｌ）で作成した相関直線を示す。図１１及び図１２の相関直線から各種点欠陥領域での再結合中心濃度が判り、遷移金属の種類と点欠陥種によって再結合中心濃度が顕著に異なることが判る。
【００３９】
最後に、工程（ｇ）の測定結果、工程（ｋ）の相関直線及び工程（ｌ）の相関直線から、領域［Ｖ］及び領域［Ｐｖ］の境界、領域［Ｐｉ］及び領域［Ｉ］の境界をそれぞれ規定する（工程（ｍ））。あらかじめ相関直線（検量線）を作成することにより、Ｃｕ及びＦｅ拡散試料の再結合ライフタイムをＬＭ−ＰＣＤで測定することにより、領域［Ｐｖ］及び領域［Ｐｉ］を識別することができる。
【００４０】
結晶の同一部位から作製されたサンプルにＣｕ及びＦｅをそれぞれのサンプルに表面汚染し拡散熱処理後の再結合ライフタイム分布を測定し、ＴＩＤ法及びＤＬＴＳ法によりＣｕ及びＦｅ濃度を測定し、更に相関直線（検量線）を作成することにより、遷移金属の種類によって各種点欠陥領域での再結合ライフタイムが顕著に異なる性質を利用して領域［Ｐｖ］及び領域［Ｐｉ］の識別を実現できる。
【００４１】
このように本発明の第２測定方法では、インゴット内に固溶する酸素を測定対象とした識別方法ではないため、インゴット内に固溶する酸素濃度に依存することがない。また酸素析出熱処理条件に左右されることがないため、高い精度で識別することができる。
【００４２】
【発明の効果】
以上述べたように、本発明のシリコン単結晶インゴットの点欠陥分布の第１及び第２測定方法は、前述した第１及び第２の測定方法における各工程を経ることにより、インゴット内の領域［Ｐｖ］及び領域［Ｐｉ］並びにこれらの境界を高い精度でかつ短時間に識別することができる。本発明の測定方法により領域［Ｐｖ］及び領域［Ｐｉ］を明確に判別できることから、従来無欠陥シリコン単結晶を製造する際に問題となっていたＶ／Ｇ制御を精密に制御することができ、無欠陥シリコン単結晶の安定した量産を実現できる。
【図面の簡単な説明】
【図１】本発明の第１測定方法における測定用サンプルに関するフローチャート。
【図２】ボロンコフの理論を基づいた、Ｖ／Ｇをよこ軸にとり、空孔型点欠陥濃度と格子間シリコン型点欠陥濃度を同一のたて軸にとって、Ｖ／Ｇと点欠陥濃度との関係を示す図。
【図３】本発明の第１及び第２測定方法の工程（ｃ）におけるインゴットから第１及び第２サンプルを作製する状況を示す図。
【図４】本発明の第１測定方法の工程（ｆ）における無欠陥のシリコン単結晶インゴットを作製可能なホットゾーンを有する結晶引上げ装置で作製された領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含む第１及び第２サンプルに拡散熱処理した場合の再結合ライフタイムの面内分布を示す図。
【図５】本発明の第１測定方法における低酸素濃度の第２サンプルをＬＭ−ＰＣＤ法により測定したときのエッチング前後におけるサンプル全体のＮｉの再結合ライフタイムを示す図。
【図６】本発明の第１測定方法の工程（ｇ）における低酸素濃度の第１及び第２サンプルをＬＭ−ＰＣＤ法により測定し、得られた再結合ライフタイムを鉛直方向に重ね合わせた再結合ライフタイム分布を示す図。
【図７】本発明の第２測定方法における測定用サンプルに関するフローチャート。
【図８】本発明の第２測定方法の工程（ｆ）における低酸素濃度の第１サンプルをＬＭ−ＰＣＤ法により測定したときのサンプル全体のＣｕの再結合ライフタイムの面内分布を示す図。
【図９】本発明の第２測定方法の工程（ｆ）における低酸素濃度の第２サンプルをＬＭ−ＰＣＤ法により測定したときのサンプル全体のＦｅの再結合ライフタイムの面内分布を示す図。
【図１０】本発明の第２測定方法の工程（ｇ）における低酸素濃度の第１及び第２サンプルをＬＭ−ＰＣＤ法により測定し、得られた再結合ライフタイムを鉛直方向に重ね合わせた再結合ライフタイム分布を示す図。
【図１１】本発明の第２測定方法の工程（ｋ）において、低酸素濃度の第１サンプルのＬＭ−ＰＣＤ法による測定値の逆数とＴＩＤ法による測定値とをプロットしたときの検量線を示す図。
【図１２】本発明の第２測定方法の工程（ｌ）において、低酸素濃度の第２サンプルのＬＭ−ＰＣＤ法による測定値の逆数とＤＬＴＳ法による測定値とをプロットしたときの検量線を示す図。
【図１３】無欠陥のシリコン単結晶インゴットを作製可能なホットゾーンを有する結晶引上げ装置で作製された領域［Ｖ］、領域［Ｐｖ］、領域［Ｐｉ］及び領域［Ｉ］を含むサンプルに析出熱処理した場合の再結合ライフタイムの面内分布を示す図。
【図１４】図１３のＡ−Ａ線断面における再結合ライフタイムを示す図。[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method for measuring a defect distribution of intrinsic point defects formed in a silicon single crystal ingot (hereinafter, referred to as an ingot) pulled by a Czochralski method (hereinafter, referred to as a CZ method). More specifically, the present invention relates to a method for evaluating the distribution of intrinsic point defects in a silicon wafer, which facilitates the manufacture of a silicon wafer having no defect.
[0002]
[Prior art]
As a factor of lowering the device yield with the recent miniaturization of a semiconductor integrated circuit, a particle (Crystal Originated Particle, hereinafter) caused by a crystal which is a defect introduced during crystal growth formed during crystal growth of a silicon single crystal ingot. , COP), oxygen-induced stacking faults (hereinafter, referred to as OISF), micro defects of oxygen precipitates serving as nuclei, or interstitial-type large dislocation (hereinafter, L / L). D.).
COPs are pits caused by crystals that appear on the wafer surface when a mirror-polished silicon wafer is SC-1 washed with a mixed solution of ammonia and hydrogen peroxide. When this wafer is measured by a particle counter, the pits are detected as particles (Light Point Defect, LPD). The COP causes deterioration of electrical characteristics, for example, a time-dependent dielectric breakdown characteristic (Time Dependent Breakdown, TDDB) of an oxide film, and a dielectric breakdown voltage characteristic (Time Zero Dielectric Breakdown, TZDB) of an oxide film. Also, if COP exists on the wafer surface, a step is generated in a device wiring process, which may cause disconnection. This also causes a leak or the like in the element isolation portion, and lowers the product yield.
[0003]
OISF is considered to be a nucleus formed by minute oxygen precipitates formed during crystal growth, and is a stacking fault that becomes apparent in a thermal oxidation step or the like when manufacturing a semiconductor device. The OISF causes a defect such as an increase in the leak current of the device. L / D is also called a dislocation cluster because a silicon wafer having this defect is immersed in a selective etching solution containing hydrofluoric acid as a main component to generate etching pits having an orientation. This L / D also causes deterioration of electrical characteristics such as leak characteristics and isolation characteristics.
From the above, it is necessary to reduce COP, OISF and L / D from a silicon wafer used for manufacturing a semiconductor integrated circuit.
[0004]
A defect-free ingot having no COP, OISF and L / D and a silicon wafer sliced from the ingot are disclosed (for example, see Patent Documents 1 and 2). This defect-free ingot is composed of a perfect region [P], where a perfect region in which an aggregate of vacancy type point defects and an aggregate of interstitial silicon type point defects in the ingot are not detected is [P]. Ingot. The perfect region [P] includes a region [V] having a defect in which vacancy-type point defects are dominant and supersaturated vacancies are aggregated in the ingot, and a region [V] in which an interstitial silicon-type point defect is predominant in the ingot. The silicon intervenes between the region [I] having the aggregated defects.
[0005]
Further, the perfect region [P] having no defect in which point defects are aggregated is classified into a region [Pv] in which vacancy type point defects are dominant and a region [Pi] in which interstitial silicon type point defects are dominant. (See, for example, Patent Document 3). The region [Pv] is a region adjacent to the region [V] and having a vacancy point defect concentration lower than the lowest vacancy point defect concentration capable of forming an OISF nucleus. The region [Pi] is a region adjacent to the region [I] and having an interstitial silicon-type point defect concentration lower than the lowest interstitial silicon-type point defect concentration capable of forming an interstitial dislocation.
In the ingot composed of the perfect region [P], the pulling speed of the ingot is V (mm / min), and the temperature gradient in the vertical direction of the ingot near the solid-liquid interface between the silicon melt and the silicon ingot is G (° C./mm). At this time, the OISF (P band) generated in a ring shape when the thermal oxidation treatment is performed disappears at the center of the wafer, and the V / G (mm) in a region where L / D (B band) is not generated. ² / Min · ° C).
[0006]
In order to manufacture a defect-free silicon wafer, it is important to control the point defect concentration distribution in the axial direction and the radial direction. Therefore, conventionally, secondary defects caused by heat treatment in an ingot, that is, aggregated defects The following method was used to measure the distribution. First, for the evaluation of the point defect concentration distribution in the axial direction and the radial direction, a silicon single crystal ingot is manufactured under crystal pulling conditions in which a defect-free region is formed. Next, a sample is prepared by slicing the ingot in the axial direction. Next, this sample is mirror-etched, and then heat-treated at 800 ° C. for 4 hours in a nitrogen or oxidizing atmosphere, and subsequently heat-treated at 1000 ° C. for 16 hours. The heat-treated sample is measured by a method such as copper decoration (copper decoration), secco-etching, X-ray topographic image analysis (X-Ray Topography), and lifetime measurement (lifetime). As shown in FIGS. 13 and 14, oxygen precipitates appear in the ingot due to the heat treatment. Therefore, each region and each boundary were identified and distinguished from the oxygen precipitates. Here, FIG. 13 shows a region [V], a region [Pv], and a region [V] produced by changing a pulling speed from a high speed to a low speed using a crystal pulling apparatus having a hot zone capable of producing a defect-free silicon single crystal ingot. FIG. 14 is a diagram showing an in-plane distribution of recombination lifetime when oxygen precipitation heat treatment is performed on a sample including Pi] and region [I], and FIG. 14 is a diagram showing a recombination lifetime in a cross section taken along line AA in FIG. 13. It is.
[0007]
[Patent Document 1]
US Patent No. 6,045,610
[Patent Document 2]
JP-A-11-1393
[Patent Document 3]
JP 2001-102385 A
[0008]
[Problems to be solved by the invention]
However, in the case of the above measurement method, it is known that the measured value of the recombination lifetime remarkably depends on the oxygen concentration in the sample and the oxygen precipitation heat treatment conditions. For example, when the concentration of oxygen dissolved in the ingot is low, the density of the oxygen precipitate formed by the oxygen precipitation heat treatment is lower than that of the sample in which the oxygen concentration is dissolved in a high concentration. Therefore, the difference value of the measured value of the recombination lifetime becomes smaller.
As described above, there is a problem that the boundary region between the region [Pv] and the region [Pi] cannot be clarified at the low oxygen concentration and the high oxygen concentration.
Further, the oxygen precipitation heat treatment not only requires a long heat treatment, but also affects the oxygen precipitation in the ingot depending on the heat treatment conditions. Therefore, the boundary region of the point defect region is either the region [Pv] or the region [Pi]. This shifts to the area side. Therefore, it has been difficult to identify and determine the original point defect area.
[0009]
An object of the present invention is to identify a region [Pv] and a region [Pi] in an ingot and their boundaries with high accuracy and in a short time without depending on the concentration of oxygen dissolved in the ingot. An object of the present invention is to provide a method for measuring a point defect distribution.
[0010]
[Means for Solving the Problems]
The invention according to claim 1 is, as shown in FIG. 1, (a) cutting a silicon single crystal ingot pulled from a silicon melt by changing the pulling speed so as to include the central axis of the single crystal ingot in the axial direction. Forming a measurement sample including the region [V], the region [Pv], the region [Pi], and the region [I]; and (b) arranging the measurement sample so as to be symmetric with respect to the center axis of the ingot. (C) applying a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm on the surface of the first sample to cause metal contamination; (D) applying a second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the second sample to contaminate the metal; e) Metal-contaminated first The first and second samples were heat-treated at 600 ° C. to 1150 ° C. for 0.5 to 30 hours in an atmosphere of argon, nitrogen, oxygen, hydrogen, or a mixed gas thereof, and applied to the surfaces of the first and second samples, respectively. A diffusion heat treatment step for diffusing the first and second transition metals into the inside of the sample; (f) a step of measuring the recombination lifetime of the entire first and second heat-treated samples; and (g) a step of: A step of superimposing the measured values in the vertical direction of the measurement result; and (h) a boundary between the area [Pi] and the area [I] and a boundary between the area [V] and the area [Pv] based on the measurement result in the step (g). And measuring the point defect distribution of the silicon single crystal ingot.
However, the region [V] is a region having a defect in which vacancy-type point defects are dominant and has excess vacancies aggregated, and the region [Pv] is a defect in which vacancy-type point defects are dominant and vacancies are aggregated. And the region [Pi] where the interstitial silicon type point defect is dominant and the region where no interstitial silicon aggregated defect is present, and the region [I] where the interstitial silicon type point defect is dominant and the lattice This is a region having a defect in which silicon is aggregated.
[0011]
In the invention according to claim 1, the recombination lifetime distribution of the first transition metal that defines the boundary between the region [V] and the region [Pv] with high accuracy, and the distribution of the region [Pi] and the region [I] with high accuracy By combining the recombination lifetime distribution of the second transition metal that defines the boundary, the region [Pv] and the region [Pi] can be formed with high accuracy and in a short time without depending on the concentration of oxygen dissolved in the ingot. And define these boundaries.
[0012]
As shown in FIG. 7, the invention according to claim 2 is as follows: (a) cutting a silicon single crystal ingot pulled from a silicon melt by changing the pulling speed so as to include the central axis of the single crystal ingot in the axial direction; Forming a measurement sample including the region [V], the region [Pv], the region [Pi], and the region [I]; and (b) arranging the measurement sample so as to be symmetric with respect to the center axis of the ingot. (C) applying a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm on the surface of the first sample to cause metal contamination; (D) applying a second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the second sample to contaminate the metal; e) Metal-contaminated first The first and second samples were heat-treated at 600 ° C. to 1150 ° C. for 0.5 to 30 hours in an atmosphere of argon, nitrogen, oxygen, hydrogen, or a mixed gas thereof, and applied to the surfaces of the first and second samples, respectively. A diffusion heat treatment step for diffusing the first and second transition metals into the inside of the sample; (f) a step of measuring the recombination lifetime of the entire first and second heat-treated samples; and (g) a step of: Superimposing the measured values in the vertical direction of the measurement results; (i) obtaining the vertical concentration of the first transition metal in the heat-treated first sample by a TID method; and (j) heat-treating the second sample. Calculating the vertical concentration of the second transition metal by the DLTS method, and (k) measuring the vertical concentration of the first sample in the step (f) with the measurement result in the step (i). A step of creating a correlation line between the first transition metal concentration and the recombination lifetime from (1) the vertical measurement result of the second sample in the above (f) step and the measurement result of the above (j) step, Forming a correlation line between the second transition metal concentration and the recombination lifetime; (m) from the measurement result of the step (g), the correlation line of the step (k), and the correlation line of the step (l), A method of measuring a point defect distribution of a silicon single crystal ingot, including a step of defining a boundary between a region [V] and a region [Pv] and a boundary between a region [Pi] and a region [I], respectively.
However, the region [V], the region [Pv], the region [Pi], and the region [I] have the same meaning as described in claim 1, respectively. This is a method of quantification by analyzing the transient capacitance characteristics of a metal-semiconductor junction diode. The DLTS method applies a positive-direction pulse voltage in a state where a reverse electric field is applied to a junction (or an interface) and applies a quasi-voltage in a depletion layer. This is a method of capturing carriers on the order.
[0013]
In the invention according to claim 2, the recombination lifetime distribution of the first transition metal and the recombination lifetime distribution of the second transition metal are combined, and the metal concentration of the first transition metal and the recombination of the first transition metal are further combined. A correlation line is created from the lifetime distribution, and similarly, a correlation line is created from the metal concentration of the second transition metal and the recombination lifetime distribution of the second transition metal, and the combined recombination lifetime distribution and two correlation lines are created. Thus, the region [Pv] and the region [Pi] and their boundaries are defined with high accuracy and in a short time without depending on the concentration of oxygen dissolved in the ingot.
[0014]
The invention according to claim 3 is the invention according to claim 1, wherein the first transition metal is Cu or Fe and the second transition metal is Ni or Co.
In the invention according to claim 3, when the first transition metal is a combination of Cu and the second transition metal is a combination of Ni, the recombination of the region [Pv] and the region [Pi] is performed. Since the lifetimes are significantly different, the boundary between the region [Pv] and the region [Pi] can be more clearly determined.
The invention according to claim 4 is the measurement method according to claim 2, wherein the first transition metal is Cu and the second transition metal is Fe.
[0015]
The invention according to claim 5 is the invention according to claim 1 or 2, wherein the diffusion heat treatment of the first transition metal and the second transition metal in the step (e) is performed at 600 ° C. to 1150 ° C. for 0.5 hour to 24 hours. This is a measuring method in which heat treatment is performed for a long time.
The invention according to claim 6 is the invention according to claim 1 or 2, wherein the recombination lifetime measurement in the step (f) is measured using LM-PCD (laser / microwave photoconductivity decay method). Is the way.
[0016]
The invention according to claim 7 is the invention according to claim 1 or 3, wherein when the second transition metal that causes metal contamination to the second sample is Ni, the surface of the second sample heat-treated in the step (e) is removed. This measurement method further includes a step of performing selective etching.
When the second transition metal is Ni, Ni silicide is formed on the surface of the sample when the diffusion heat treatment is performed on the second sample. The formation of this Ni silicide causes an increase in the surface recombination rate in the subsequent step of measuring the recombination lifetime, causing a problem that an accurate numerical value cannot be measured. Therefore, Ni silicide is removed by selectively etching the sample surface.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
(1) Ingots subject to the measurement method of the present invention
The ingot to be subjected to the measurement method of the present invention is pulled from the silicon melt by controlling the V / G ratio so as to include the region [V], the region [Pv], the region [Pi], and the region [I].
This ingot has an oxygen concentration of 8.0 × 10 ¹⁷ atoms / cm ³ ~ 1.0 × 10 ¹⁸ atoms / cm ³ (Old ASTM, the same applies hereinafter) or a sample obtained by slicing this ingot in a nitrogen atmosphere at 800 ° C. for 4 hours, followed by heat treatment at 1000 ° C. for 16 hours, and then the recombination lifetime is reduced. This is an ingot in which the boundary between the region [Pv] and the region [Pi] in the sample cannot be identified when measured.
[0018]
Such an ingot is pulled up from a silicon melt in a hot zone furnace based on the Voronkov theory by the CZ method or the CZ method applying a magnetic field. Generally, when a silicon single crystal ingot is pulled up from a silicon melt in a hot zone furnace by a CZ method or a magnetic field application CZ method, a point defect and an aggregate of the point defects are regarded as defects in the silicon single crystal. (Agglomerates: three-dimensional defect). Point defects have two general forms: vacancy type point defects and interstitial silicon type point defects. A vacancy-type point defect is one in which one silicon atom has separated from one of the normal positions in the silicon crystal lattice. Such vacancies aggregate to form vacancy-type defects. On the other hand, lattice silicon aggregates to form interstitial silicon type defects.
[0019]
Point defects are generally introduced at the interface between the silicon melt (molten silicon) and the ingot (solid silicon). However, by continuously pulling the ingot, the portion that was the contact surface cools with the pulling. During cooling, the vacancies or interstitial silicon undergo a diffusion or annihilation reaction. Excessive point defects upon cooling to about 1100 ° C. form vacancy agglomerates or interstitial silicon agglomerates. In other words, it is a three-dimensional structure in which excessive point defects are generated by forming aggregates.
Aggregates of vacancy-type defects include defects called LSTDs (Laser Scattering Tomograph Defects) or FPDs (Flow Pattern Defects), in addition to the above-mentioned COPs. Aggregates of interstitial silicon-type defects have the L / D described above. Including defects called. FPD means that a silicon wafer produced by slicing an ingot is secco-etched for 30 minutes (Secco etching, HF: K ₂ Cr ₂ O ₇ (0.15 mol / l) = a source of a trace exhibiting a unique flow pattern that appears when the silicon single crystal is irradiated with infrared rays. Are sources having different refractive indices and generating scattered light.
[0020]
Boronkov's theory states that, in order to grow a high-purity ingot with a small number of defects, the pulling speed of the ingot is V (mm / min), and the temperature gradient near the solid-liquid interface between the silicon melt and the silicon ingot is G (° C. / mm), V / G (mm ² / Min · ° C). In this theory, as shown in FIG. 2, the relationship between V / G and point defect concentration is defined by taking V / G as the horizontal axis and taking the vacancy type defect concentration and the interstitial silicon type defect concentration as the same vertical axis. Schematically, it illustrates that the boundary between the vacancy region and the interstitial silicon region is determined by V / G. More specifically, the V / G ratio is a critical point (V / G) _c In the above, an ingot having an increased vacancy concentration is formed, but the V / G ratio is changed to the critical point (V / G). _c In the following, an ingot having an increased interstitial silicon concentration is formed. In FIG. 2, [I] is a region having a defect where interstitial silicon is dominant and interstitial silicon is aggregated ((V / G) ₁ Below), and [V] indicates a region having a defect in which vacancies are dominant and vacancies are aggregated ((V / G) ₂ Above, and [P] indicates a perfect region ((V / G) in which no aggregates of vacancy type defects and aggregates of interstitial silicon type defects exist. ₁ ~ (V / G) ₂ ). The P-band region ((V / G) forming the OISF nucleus is located at the boundary with the region [V] on the side adjacent to the region [P]. ₂ ~ (V / G) ₃ ) Exists. In the P band region, fine plate-like precipitates are present, and OISF (stacking fault) is formed by heat treatment in an oxidizing atmosphere. A B-band region ((V / G)) is located at the boundary of the region [I] on the side adjacent to the region [P]. ₄ ~ (V / G) ₁ ) Exists. The B band region is a region in which aggregates of interstitial silicon are used as nuclei and oxygen precipitation occurs at a high concentration by heat treatment.
[0021]
This perfect area [P] is further classified into an area [Pi] and an area [Pv].
[Pi] is such that the V / G ratio is (V / G) ₁ To the critical point, where the interstitial silicon is dominant and has no aggregated defects. [Pv] is the above (V / G) when the V / G ratio is from the critical point. ₂ This is a region in which vacancies are dominant and have no aggregated defects. That is, [Pi] is a region adjacent to the region [I] and has an interstitial silicon concentration lower than the lowest interstitial silicon concentration capable of forming an interstitial dislocation, and [Pv] is adjacent to the region [V]. And a region having a vacancy concentration lower than the lowest vacancy concentration capable of forming an OISF nucleus.
Therefore, in order for the ingot to be subjected to the measurement method of the present invention to include the region [V], the region [Pv], the region [Pi], and the region [I], the V / G ratio is set to the above (V / G). ₄ Critical point from below, (V / G) ₃ As described above, it is controlled so as to include each area and pulled up.
(2) First measurement method of the present invention
Next, a first measuring method of the present invention will be described with reference to FIGS.
As shown in FIG. 1, first, a measurement sample is prepared from a test ingot (step (a)). That is, the ingot is pulled from the silicon melt stored in the quartz crucible of the pulling device based on the CZ method or the magnetic field application CZ method. At this time, the pulling speed V (mm / min) of the ingot is increased from the high speed (top side) so that the above-mentioned region [V], region [Pv], region [Pi] and region [I] are included in the axial direction of the ingot. The speed is changed from low speed (bottom side) or low speed (bottom side) to high speed (top side) to pull up.
[0022]
As shown in FIG. 3, the ingot obtained in the test is cut in the axial direction so as to include the center axis of the ingot, and is mirror-etched to have a thickness of 500 to 2000 μm. A sample is made (FIGS. 3 (a) and 3 (b)). The measurement sample includes a region [V], a region [Pv], a region [Pi], and a region [I]. The ingot shown in FIG. 3A is an ingot obtained by changing the pulling speed V (mm / min) from a high speed (top side) to a low speed (bottom side).
[0023]
The concentration of oxygen dissolved in the sample for measurement was measured by FT-IR (Fourier transform insulation absorption spectroscopy), and the oxygen concentration was 1.0 × 10 4 ¹⁸ atoms / cm ³ In this case, heat treatment is performed at 800 ° C. for 4 hours in a nitrogen atmosphere, followed by heat treatment at 1000 ° C. for 16 hours. The recombination lifetime of the heat-treated measurement sample is measured over the entire sample by the LM-PCD method. FIG. 13 shows an example of the measurement result. As shown in FIG. 13, oxygen precipitates appear at high density in the sample by the two-step heat treatment. According to the concentration distribution, the region [V], the region [Pv], the region [Pi], and the region [I] are clearly identified in the entire sample. In the region [V], there is a P band region that forms the OISF nucleus, and in the region [I], there is a B band region.
[0024]
On the other hand, the oxygen concentration of the measurement sample was 1.0 × 10 ¹⁸ atoms / cm ³ If less than, even if the measurement sample is heat-treated similarly to the above and the recombination lifetime of the measurement sample is measured, as shown in FIG. 5, from the measurement result of the recombination lifetime, It cannot be determined whether the perfect area [P] is the area [Pv] or the area [Pi].
[0025]
For this reason, in the present invention, the measurement sample is divided into two so as to be symmetric with respect to the center axis of the ingot at the position indicated by the arrow in FIG. 3B (step (b)). The measurement sample divided in the step (b) is referred to as a first sample and a second sample as shown in FIG.
Returning to FIG. 1, a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm is applied to the surface of the first sample to contaminate the first sample with metal (step (c)). . Cu and Fe are preferable as the first transition metal. The first transition metal solution is a solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm, preferably 1 to 100 ppm. In particular, a standard solution for atomic absorption in which Cu or Fe is dissolved at a concentration of 10 to 100 ppm is preferable because it is available and has excellent concentration accuracy.
Examples of the coating method include a spin coating method and a dipping method. If the concentration of the first transition metal solution is less than the lower limit, the boundaries between the regions cannot be sufficiently determined, and the recombination lifetime cannot be measured with high accuracy. If the upper limit is exceeded, the transition metal diffuses into the front and back surfaces of the sample and forms metal silicide (precipitate) near the front and back surfaces, so that recombination, which is a deep energy level inside the sample (in the crystal), occurs. The center cannot be formed.
[0026]
A second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm is applied to the surface of the second sample to contaminate the second sample with metal (step (d)). . Ni and Co are preferable as the second transition metal. The second transition metal solution is a solution in which the second transition metal is dissolved at a concentration of 1 to 1000 ppm, preferably 1 to 100 ppm. In particular, a standard solution for atomic absorption in which Ni or Co is dissolved at a concentration of 10 to 100 ppm is preferable because it is available and has excellent concentration accuracy. The method of application is the same as the method of applying the first transition metal. If the concentration of the second transition metal solution is less than the lower limit, the boundary of each region cannot be determined sufficiently, and the recombination lifetime cannot be measured with high accuracy. If the upper limit is exceeded, the transition metal diffuses into the front and back surfaces of the sample to form metal silicide (precipitate) near the front and back surfaces, so that recombination, which is a deep energy level inside the sample (in the crystal), occurs. The center cannot be formed.
[0027]
Next, the first and second samples are heat-treated in an atmosphere of argon, nitrogen, oxygen, hydrogen or a mixed gas thereof at 600 ° C. to 1150 ° C. for 0.5 hours to 30 hours, so that the first and second samples have a surface. The applied first and second transition metals are diffused throughout the sample (step (e)). As a heat treatment method, the temperature is raised at a rate of 0.5 to 10 ° C./min, and the heat treatment is performed at a temperature of 600 to 1150 ° C. for 0.5 to 30 hours, preferably 0.5 to 24 hours. More preferably, the temperature is raised at a rate of 5 to 10 ° C./min and heat treatment is performed at 900 to 1000 ° C. for 1 to 2 hours. If the time and temperature of the heat treatment are less than the lower limit, the transition metal does not sufficiently diffuse into the sample. If the upper limit is exceeded, the transition metal diffuses into the front and back surfaces of the sample to form metal silicide (precipitate) near the front and back surfaces, so that recombination, which is a deep energy level inside the sample (in the crystal), occurs. The center cannot be formed.
[0028]
When the second transition metal contaminated by the second sample is Ni, the recombination lifetime of the heat-treated second sample is measured. As shown in FIG. 5, the recombination life of the region [Pi] and the region [I] is obtained. The time has not increased sufficiently and the boundaries between the regions cannot be identified. This is because Ni has a larger diffusion coefficient than the first transition metal, Cu or Fe, and is thus deposited near the sample surface by heat treatment. So NH ₃ And HF and CH ₃ The sample surface is etched with an etchant containing OOH. By this etching, Ni existing near the surface is removed. When the recombination lifetime of the second sample is measured again in this state, the increase in the recombination lifetime in the region [Pi] and the region [I] is remarkable after the chemical etching, while the region [V] Shows that there is no change in the recombination lifetime before and after chemical etching.
[0029]
Next, the recombination lifetime of the transition metal is measured for the heat-treated first and second samples (step (f)). For measurement, LM-PCD (laser / microwave photoconductivity decay method) is preferable. Quantify the recombination lifetime across the sample by LM-PCD. FIG. 4 shows the recombination lifetimes of the first and second samples, respectively. As shown in FIG. 4, in the first sample, the region with the lowest recombination lifetime was the region [Pi], and the region with the highest recombination lifetime was the region [Pv]. In the second sample, the region having the lowest recombination lifetime was the region [Pv], and the region having the highest recombination lifetime was the region [Pi].
[0030]
The recombination lifetimes of the first and second samples are measured, and the measurement results are superimposed on the distribution in the vertical direction (step (g)). Using Cu as the first transition metal and Ni as the second transition metal, the recombination lifetimes of the first and second samples obtained by selectively etching the second sample are measured, and the distribution is superimposed in the vertical direction. FIG. 6 shows the results. In FIG. 6, each horizontal axis corresponds to the axial direction of the ingot before cutting the sample.
[0031]
In the case of the first sample in which Cu is diffused, the recombination lifetime of the region [Pv] is longer than that of the other point defect regions. On the other hand, in the case of the second sample in which Ni is diffused, it can be seen that the region having the highest recombination lifetime is the region [Pi]. Therefore, the region [Pv] and the region [Pi] can be identified from the type of the transition metal and the recombination lifetime dependence of the point defect region.
[0032]
The samples prepared from the same part of the crystal are subjected to surface contamination with Ni and Cu on each sample, and the recombination lifetime distribution after diffusion treatment is measured. Can distinguish the region [Pv] and the region [Pi] by utilizing the remarkably different properties.
[0033]
As described above, since the first measurement method of the present invention is not an identification method in which oxygen dissolved in the ingot is a measurement target, the first measurement method does not depend on the concentration of oxygen dissolved in the ingot. In addition, since it is not affected by the oxygen precipitation heat treatment conditions, the identification can be performed with high accuracy.
(3) Second measuring method of the present invention
Next, a second measurement method of the present invention will be described with reference to FIG.
First, the steps (a) and (b) are the same as in the first measuring method described above. Next, as shown in FIG. 7, a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm is applied to the surface of the first sample to contaminate the first sample with a metal (step ( c)). Here, Cu is selected as the first transition metal. Next, a second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm is applied to the surface of the second sample to contaminate the second sample with metal (step ( d)). Here, Fe is selected as the second transition metal. In step (e), the first and second samples are subjected to a diffusion heat treatment under the same conditions as in the above-described first measurement method.
[0034]
Next, the recombination lifetime of the transition metal is measured for the heat-treated first and second samples (step (f)). LM-PCD is preferred for measurement. Quantify the recombination lifetime across the sample by LM-PCD. The recombination lifetimes of the first and second samples are shown in FIGS. 8 and 9, respectively. As shown in FIG. 8, in the first sample, the region with the highest recombination lifetime was [Pi], and the region with the lowest recombination lifetime was [V]. As shown in FIG. 9, in the second sample, the region with the lowest recombination lifetime was the region [Pi], and the region with the highest recombination lifetime was the region [Pv].
[0035]
The distribution of the results of the recombination lifetime measurement of the first and second samples is overlapped in the vertical direction (step (g)). FIG. 10 shows the result of superimposing the distributions of the recombination lifetimes of the first and second samples in the vertical direction. In FIG. 10, each horizontal axis corresponds to the axial direction of the ingot before cutting the sample. In the case of the first sample in which Cu is diffused, the recombination lifetime of the region [Pv] is longer than that of the other point defect regions. On the other hand, in the case of the second sample in which Fe is diffused, it can be seen that the region having the lowest recombination lifetime is the region [Pv]. Therefore, the region [Pv] and the region [Pi] can be identified from the type of the transition metal and the recombination lifetime dependence of the point defect region.
[0036]
Next, the concentration of the first transition metal in the vertical direction is determined from the heat-treated first sample by the TID method (step (i)). The concentration of the recombination center formed by Cu as the first transition metal in the sample is quantified by the TID method. The TID (Transient Ion Drift) method is a method for quantifying the concentration of Cu dissolved in a silicon single crystal by analyzing the transient capacitance characteristics of a metal-semiconductor junction diode. It is assumed that Cui ions are uniformly distributed in the space charge layer region in the thermal equilibrium state. When a positive voltage is applied to the metal-semiconductor junction diode (P-type Si), Cui ions near the surface drift to the end of the space charge layer, and high-concentration Cui ions are accumulated. In the accumulation state, at the edge of the space charge layer, part of the Cui ions diffuses outside the space charge layer by thermal diffusion. Again, it is considered that the Cui ions thermally diffuse from the vicinity of the surface and from the end of the space charge layer in the state of zero bias. The TID method is an evaluation method capable of quantifying a change in the concentration of Cui ions accumulated in a space charge layer region from a zero bias state by measuring a transient response attenuation characteristic of a junction capacitance.
[0037]
Subsequently, the concentration of the second transition metal in the vertical direction is obtained from the heat-treated second sample by the DLTS method (step (j)). The concentration of the recombination center formed by Fe as the second transition metal in the sample is quantified by the DLTS method. The DLTS (Deep Level Transient Spectroscopy) method is a method in which a positive direction pulse voltage is applied while a reverse electric field is applied to a junction (or interface), and carriers are trapped at a level in a depletion layer. It is.
[0038]
Since it is presumed that there is a correlation between both the recombination lifetime and the recombination center concentration, the reciprocal of the measured recombination lifetime of the first sample obtained in the step (f) and the reciprocal of the step (i) The measured values are plotted to create a correlation line between the first transition metal concentration and the recombination lifetime (step (k)). FIG. 11 shows the correlation line created in step (k). Further, the reciprocal of the measured recombination lifetime of the second sample obtained in step (f) and the measured value of step (j) are plotted to create a correlation line between the second transition metal concentration and the recombination lifetime. (Step (l)). FIG. 12 shows the correlation line created in step (l). The recombination center concentrations in various point defect regions are found from the correlation lines in FIGS. 11 and 12, and it is found that the recombination center concentration is significantly different depending on the type of transition metal and the type of point defect.
[0039]
Finally, from the measurement result of the step (g), the correlation straight line of the step (k) and the correlation straight line of the step (l), the boundary of the area [V] and the area [Pv], the area [Pi] and the area [I] are obtained. The boundaries are defined (step (m)). The region [Pv] and the region [Pi] can be identified by preparing a correlation line (calibration curve) in advance and measuring the recombination lifetime of the Cu and Fe diffusion samples by LM-PCD.
[0040]
Samples made from the same part of the crystal were contaminated with Cu and Fe on the surface of each sample, the recombination lifetime distribution after diffusion heat treatment was measured, and the Cu and Fe concentrations were measured by the TID method and the DLTS method. By creating a straight line (calibration curve), it is possible to distinguish between the region [Pv] and the region [Pi] by utilizing the property that the recombination lifetime in various point defect regions is significantly different depending on the type of transition metal.
[0041]
As described above, since the second measurement method of the present invention is not an identification method in which oxygen dissolved in an ingot is a measurement target, it does not depend on the concentration of oxygen dissolved in the ingot. In addition, since it is not affected by the oxygen precipitation heat treatment conditions, the identification can be performed with high accuracy.
[0042]
【The invention's effect】
As described above, the first and second measurement methods of the point defect distribution of the silicon single crystal ingot of the present invention are performed by performing the respective steps in the first and second measurement methods described above, and thereby, the area [ Pv] and the region [Pi] and their boundaries can be identified with high accuracy and in a short time. Since the region [Pv] and the region [Pi] can be clearly distinguished by the measurement method of the present invention, the V / G control which has conventionally been a problem in producing a defect-free silicon single crystal can be precisely controlled. Thus, stable mass production of defect-free silicon single crystals can be realized.
[Brief description of the drawings]
FIG. 1 is a flowchart relating to a measurement sample in a first measurement method of the present invention.
FIG. 2 is a graph showing the relationship between V / G and point defect concentration, where V / G is taken on the horizontal axis based on Bornkov's theory and vacancy type point defect concentration and interstitial silicon type point defect concentration are taken on the same vertical axis. The figure which shows a relationship.
FIG. 3 is a diagram showing a situation in which first and second samples are produced from an ingot in step (c) of the first and second measurement methods of the present invention.
FIG. 4 shows a region [V], a region [Pv], and a region produced by a crystal pulling apparatus having a hot zone capable of producing a defect-free silicon single crystal ingot in step (f) of the first measurement method of the present invention. The figure which shows the in-plane distribution of the recombination lifetime when the diffusion heat treatment is carried out to the 1st and 2nd sample containing [Pi] and area | region [I].
FIG. 5 is a diagram showing the recombination lifetime of Ni in the entire sample before and after etching when a second sample having a low oxygen concentration is measured by the LM-PCD method in the first measurement method of the present invention.
FIG. 6: The first and second samples having a low oxygen concentration in the step (g) of the first measurement method of the present invention were measured by the LM-PCD method, and the obtained recombination lifetimes were overlapped in the vertical direction. The figure which shows the recombination lifetime distribution.
FIG. 7 is a flowchart relating to a measurement sample in the second measurement method of the present invention.
FIG. 8 is a view showing the in-plane distribution of the recombination lifetime of Cu of the entire sample when the first sample having a low oxygen concentration is measured by the LM-PCD method in step (f) of the second measurement method of the present invention. .
FIG. 9 is a diagram showing the in-plane distribution of the recombination lifetime of Fe in the entire sample when the second sample having a low oxygen concentration is measured by the LM-PCD method in step (f) of the second measurement method of the present invention. .
FIG. 10: The first and second samples having a low oxygen concentration in step (g) of the second measurement method of the present invention were measured by the LM-PCD method, and the obtained recombination lifetimes were overlapped in the vertical direction. The figure which shows the recombination lifetime distribution.
FIG. 11 shows a calibration curve obtained by plotting the reciprocal of the measured value of the first sample having a low oxygen concentration by the LM-PCD method and the measured value by the TID method in step (k) of the second measuring method of the present invention. FIG.
FIG. 12 shows a calibration curve obtained by plotting the reciprocal of the measured value of the second sample having a low oxygen concentration by the LM-PCD method and the measured value by the DLTS method in the step (l) of the second measuring method of the present invention. FIG.
FIG. 13 is deposited on a sample including a region [V], a region [Pv], a region [Pi], and a region [I] manufactured by a crystal pulling apparatus having a hot zone capable of manufacturing a defect-free silicon single crystal ingot. The figure which shows the in-plane distribution of the recombination lifetime at the time of heat processing.
FIG. 14 is a diagram showing a recombination lifetime in a cross section taken along line AA of FIG. 13;

Claims

(A) A silicon single crystal ingot pulled from a silicon melt at a different pulling speed is cut in the axial direction so as to include the central axis of the single crystal ingot, and a region [V], a region [Pv], and a region [Pi] are cut. And a step of preparing a measurement sample including the region [I];
(B) a step of preparing the first and second samples by dividing the measurement sample into two so as to be symmetric with respect to the center axis of the ingot;
(C) applying a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the first sample to contaminate the metal;
(D) applying a second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the second sample to contaminate the metal;
(E) heat-treating the first and second samples contaminated with metal at 600 ° C. to 1150 ° C. for 0.5 to 30 hours in an atmosphere of argon, nitrogen, oxygen, hydrogen or a mixed gas thereof. And a diffusion heat treatment step of diffusing the first and second transition metals respectively applied to the surface of the second sample into the sample,
(F) measuring recombination lifetimes of the entire first and second heat-treated samples;
(G) superimposing the measured values in the vertical direction of the measurement results of the step (f);
(H) a step of defining a boundary between the region [Pi] and the region [I] and a boundary between the region [V] and the region [Pv] based on the measurement result of the step (g). A method for measuring defect distribution.
However, the region [V] is a region having a defect in which vacancy-type point defects are dominant and has excess vacancies aggregated, and the region [Pv] is a defect in which vacancy-type point defects are dominant and vacancies are aggregated. And the region [Pi] where the interstitial silicon type point defect is dominant and the region where no interstitial silicon aggregated defect is present, and the region [I] where the interstitial silicon type point defect is dominant and the lattice This is a region having a defect in which silicon is aggregated.

(A) A silicon single crystal ingot pulled from a silicon melt at a different pulling speed is cut in the axial direction so as to include the central axis of the single crystal ingot, and a region [V], a region [Pv], and a region [Pi] are cut. And a step of preparing a measurement sample including the region [I];
(B) a step of preparing the first and second samples by dividing the measurement sample into two so as to be symmetric with respect to the center axis of the ingot;
(C) applying a first transition metal solution in which the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the first sample to contaminate the metal;
(D) applying a second transition metal solution in which a second transition metal different from the first transition metal is dissolved at a concentration of 1 to 1000 ppm to the surface of the second sample to contaminate the metal;
(E) heat-treating the first and second samples contaminated with metal at 600 ° C. to 1150 ° C. for 0.5 to 30 hours in an atmosphere of argon, nitrogen, oxygen, hydrogen or a mixed gas thereof. And a diffusion heat treatment step of diffusing the first and second transition metals respectively applied to the surface of the second sample into the sample,
(F) measuring recombination lifetimes of the entire first and second heat-treated samples;
(G) superimposing the measured values in the vertical direction of the measurement results of the step (f);
(I) determining the concentration of the first transition metal in the vertical direction from the heat-treated first sample by a TID method;
(J) determining the vertical concentration of the second transition metal in the heat-treated second sample by a DLTS method;
(K) forming a correlation line between the first transition metal concentration and the recombination lifetime from the measurement result in the vertical direction of the first sample in the step (f) and the measurement result in the step (i);
(L) forming a correlation straight line between the second transition metal concentration and the recombination lifetime from the measurement result in the vertical direction of the second sample in the step (f) and the measurement result in the step (j);
(M) From the measurement result of the step (g), the correlation line of the step (k) and the correlation line of the step (l), the boundary between the region [V] and the region [Pv], the region [Pi] and the region [ I) measuring the point defect distribution of the silicon single crystal ingot.
However, the region [V], the region [Pv], the region [Pi], and the region [I] have the same meaning as described in claim 1, respectively. This is a method of quantification by analyzing the transient capacitance characteristics of a metal-semiconductor junction diode. The DLTS method applies a positive-direction pulse voltage in a state where a reverse electric field is applied to a junction (or an interface) and applies a quasi-voltage in a depletion layer. This is a method of capturing carriers on the order.

The measurement method according to claim 1, wherein the first transition metal is Cu or Fe, and the second transition metal is Ni or Co.

3. The method according to claim 2, wherein the first transition metal is Cu and the second transition metal is Fe.

The method according to claim 1, wherein the heat treatment for diffusing the first transition metal and the second transition metal in the step (e) is performed at 600 ° C. to 1150 ° C. for 0.5 hours to 24 hours.

The measurement method according to claim 1 or 2, wherein the recombination lifetime measurement in the step (f) is performed using LM-PCD (laser / microwave photoconductivity decay method).

4. The measuring method according to claim 1, further comprising a step of selectively etching the surface of the second sample heat-treated in step (e) when the second transition metal that causes metal contamination to the second sample is Ni.