JP3667515B2

JP3667515B2 - Method for producing quartz glass crucible

Info

Publication number: JP3667515B2
Application number: JP33580397A
Authority: JP
Inventors: 優新保
Original assignee: 東芝セラミックス株式会社
Priority date: 1997-12-05
Filing date: 1997-12-05
Publication date: 2005-07-06
Anticipated expiration: 2017-12-05
Also published as: JPH11171571A

Description

【０００１】
【発明の属する技術分野】
この発明は、シリコン単結晶引き上げ装置に使用されるシリコン溶融用の石英ガラスルツボの製造方法にかかり、とくに高純度の単結晶を長時間にわたって引き上げ可能なルツボ内の気泡を含む層の構造とその製法の工夫に関する。
【０００２】
【従来の技術】
シリコン融液中に浸けた種結晶を引き上げて単結晶を製造するシリコン単結晶引き上げ装置では、シリコン融液を形成・保持する石英ガラスルツボが使用されている。このルツボの壁内には多数の気泡を含有する層（以下「保温層」と言う）が含まれ、この保温層がルツボ内のシリコン融液の保温およびその温度分布の均質化を図る役割を担う。この場合の保温能力は気泡の数やその大きさ等の気泡状態で決まるため、この状態をルツボ成形時にシリカ粉末の粒度分布や溶融時の雰囲気調整用の溶融条件等に基づいて調整することで保温能力が間接的に制御される。
【０００３】
【発明が解決しようとする課題】
上述した従来の石英ガラスルツボでは、単結晶引き上げ時でのルツボの温度圧力状態、たとえば約１５００℃および減圧下の条件により保温層中の気泡が膨張して径が増加するため、熱線散乱の度合いが時間と共に増加し、ヒータからルツボを介してシリコン融液に加えられる熱量が低下していく。その結果、微妙な温度制御が困難となって単結晶引き上げの諸条件の精度が乱されると共に、熱効率の低下により消費電力が著しく増加するといった問題があった。
【０００４】
一方、保温層の保温能力を制御する方法は目標とする泡の残留量を経験的に求めて間接的におこなうものであるため、精度を高めるには製造条件の設定に極めて精妙かつ高度な技術が要求されるといった問題もあった。
【０００５】
特に保温層における諸問題は、近年要求されている大口径のシリコン単結晶を引き上げる場合により顕著となる。この場合には、大口径の石英ガラスルツボが使用され、その引き上げ工程も長時間にわたるため、ルツボ壁面の熱負荷が増大したり、シリコン融液の重量が増加する等によりルツボの変形や変質が生じて単結晶化率が低下する等の固有の問題もあり、これが上述の諸問題とともに相乗効果として増幅されるためである。
【０００６】
この発明は、このような従来の問題を改善するものであり、ルツボ使用時における保温層中の気泡の膨張を抑えることができる石英ガラスルツボの製造方法提供することを目的とする。この発明は同時に気泡の含有量を簡単にかつ直接的に制御することを別の目的とする。
【０００７】
【課題を解決するための手段】
上記目的を達成するため、本発明者は石英ガラスルツボ使用時の条件でも保温層の泡の膨れが実質的に生じない状態となるように様々な実験・検討をおこなった。その結果、保温層中に所定の結晶化促進剤を含めた場合には、石英ルツボの開口部内に入れたシリコン融液用の多結晶シリコンが高温度で溶解している間、石英ルツボの壁体内で石英ガラスが結晶化促進剤を核とする周囲で速やかに結晶成長し、そこで固体化した結晶により気泡の内圧が増加してもその膨張を阻止できるといった知見を得た。
【０００８】
この発明は、このような知見に基づいて完成させたもので、シリコン単結晶引き上げ用開口を有するルツボ壁を備え、このルツボ壁を石英ガラスで構成した石英ガラスルツボの製造方法であって、前記開口内を保温するための多数の気泡を有し且つ前記ルツボ壁内に層状に分布する保温層を備え、この保温層は前記石英ガラス用の結晶化促進剤を含むことを特徴とする。
【０００９】
結晶化促進剤は、この発明の範囲内での一例として各種の金属イオン、あるいはクリストバライトやトリジマイト等に代表される結晶質シリカの微粉末等を例示できる。その他、石英ガラスの溶融状態からの結晶化を促進させる添加材であれば適用可能である。とくに本発明者が行なった実験によれば、結晶化促進剤としてＣａ、Ｍｇ、Ｓｒ、Ｂａ、Ａｌ、Ｙ等の金属イオンを使用した場合により良好な結果を得ることがわかった。これらの金属イオンを使用すれば、石英ガラス中の拡散が小さく、また高温に晒されても変質しにくいためと考えられる。
【００１０】
この発明にかかる石英ガラスルツボの製造方法は、シリカ粉末をルツボ成形用の型内に入れて成形し、その成形体を加熱溶融し、その後に冷却して多数の気泡を有する保温層を含む石英ガラスのルツボ壁を形成する方法であって、前記シリカ粉末を成形する前にそのシリカ粉末に前記石英ガラス用の結晶化促進剤を添加して前記保温層用の原料を調整することを特徴とする。すなわちこの製造方法は、石英ガラスルツボ用の型の中にシリカ粉末の層を成形し、その成形体を回転させながらアーク等の熱源を用いて型の内側から溶解させるものであって、シリカ粉末の層のうちの溶解後に多数の泡を含む保温層となる部分にあらかじめ結晶化促進剤を添加するものである。
【００１１】
結晶化促進剤としては、金属イオン化合物の粉末あるいは溶液を使用し、これをシリカ粉末に混合すればよい。この場合に好ましくは、これらの金属イオンの添加材の一部、あるいは全てにたとえば炭酸塩や水酸化物、塩化物、炭化物、窒化物等のルツボ溶融時に分解し、ガス成分を発生させる化合物を使用するものとする。化合物の添加量を調整すれば、泡の含有量を自在に調整できるといった利点があるためである。
【００１２】
金属イオンの添加方法の一例として、これらのイオン化合物を含む水溶液をシリカ粉末に添加してスラリー状とし、ルツボ成形用型の中で成形し乾燥させると、成形時および乾燥後のシリカ体の凝集力が増すため、シリカ粉末の成形がきわめて容易になるといった利点があることがわかった。
【００１３】
結晶促進剤は、シリコン結晶引き上げ時に十分に結晶が成長して泡の膨れを実質的に防止可能な濃度が必要であるが、濃度が高すぎるとルツボ形成時に結晶が発達してしまい、強度の低下などを引き起こす。本発明者の行なった実験・検討によれば、泡の膨れ対策として実質的な効果を得るには、５ｐｐｍ以上の濃度が必要であるが、５００ｐｐｍを超えるとルツボ成型時に結晶化が起き、良好なルツボを得ることができないことがわかった。したがって、結晶化促進剤の濃度は５ｐｐｍ以上５００ｐｐｍ以下であることが望ましい。
【００１４】
シリコン単結晶引き上げ用ルツボでは、通常では泡を含有する保温層の内側に透明な石英ガラスからなる層をもつものが多い。本発明は、このような内面透明層をふくむルツボの製造方法にも適用できる。この場合には、例えば溶融前に保温層を構成する結晶化促進剤含有のシリカ層の内側に結晶化促進材を添加しないシリカ粉末の層を形成し、この促進材無添加の層を溶融する際に減圧雰囲気等によりして透明層を形成させる方法や、あるいは泡を含有する保温層を形成させた後、シリカ源を外部から供給しながら溶融し、ルツボ内面側に透明層を重ねる方法などが適用可能である。
【００１５】
【発明の実施の形態】
以下、この発明にかかる石英ガラスルツボの製造方法の実施形態を図面を参照して説明する。
【００１６】
（第１の実施形態）
実施例１
この実施例では、保温層に添加すべき結晶化促進剤としてＣａイオンを使用し、そのイオン濃度を５ｐｐｍに調整して石英ガラスルツボを製造し、このルツボに対して各種の試験を行った。その結果を表１に示す。
【００１７】
【表１】

【００１８】
表１に示す石英ガラスルツボの製造方法を説明する。まず結晶化促進剤としてＣａイオンの水溶液を調整する。この調整に際し、Ｃａイオンの水酸化物を準備し、その水酸化物をつかってイオン濃度で０．５％の水溶液をつくり、これを母液として所定の濃度に薄めた水溶液を作成する。この水溶液をシリカ粉末に所定の重量割合（水溶液：シリカ粉末＝２５：７５）で混ぜ合わせ、その混合物をルツボ成形用の型に入れる。この型は、１８インチの大きさのものを使用し、その外部が水冷されるようになっている。
【００１９】
この型を回転させることによりシリカ粉末の混合物をルツボ形状に成形し、その成形粉末を加熱・乾燥させる。この加熱・乾燥に際しては、型の回転をそのまま続けながら水冷部に温水を通すなどの方法を用いる。ここで乾燥させたシリカ粉は凝結していると共に、Ｘ線回折で調べたところ、Ｃａイオンの炭酸塩が生じていることが確認された。
【００２０】
次いで、このシリカ成形体上に通常のシリカ粉末、すなわち結晶化促進剤を添加していないシリカ粉末の層を成形させ、その成形体に対して型を回転させながら内部に設けたカーボン電極を用いてアーク溶融させることにより、石英ガラスルツボを得た。ここで促進剤無添加のシリカ粉末層の溶融期間中にＨｅ雰囲気を保ち、シリカ成形体を真空引きする等の方法を用いることにより、ルツボ内面側のガラス層を透明化させた。
【００２１】
得られたルツボに対して保温層内の泡の生成状況および結晶化の有無状況を調べた。その結果、表１に示すように、保温層内の気泡量はやや少ないもののその生成は概ね良好であり、また結晶も析出していないことが確認された。
【００２２】
このルツボを１５００℃で５時間２０Ｔｏｒｒの減圧下で熱処理した後の泡の膨れ状態を調べたところ、表１に示すように約１０％であることが確認された。比較のため、結晶化促進剤を添加していない従来のルツボを同じ条件で熱処理して調べたところ、泡は約２倍（２００％）に膨れていることが確認された。
【００２３】
次にこのルツボを使って実際にシリコン単結晶の引き上げを試みた。その結果、いずれも時間の経過にともなうルツボの保温層の変質が認められず、熱効率についても引き上げ終了まで良好に推移し、消費電力が従来とくらべて約５０％削減できることが確認された。使用後のルツボ状態を調べたところ、保温層が結晶化し、クリストバライトが検出された。
【００２４】
実施例２
この実施例では、Ｃａのイオン濃度を５０ｐｐｍとし、その他は実施例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡の状態が良好で初期の結晶が析出しておらず、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げをおこなったところ、実施例１と同様の良好な結果が得られた。
【００２５】
実施例３
この実施例では、Ｃａのイオン濃度を１００ｐｐｍとし、その他は実施例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡の状態が良好で初期の結晶が析出しておらず、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げをおこなったところ、実施例１と同様の良好な結果が得られた。
【００２６】
実施例４
この実施例では、結晶化促進イオンとしてＳｒを採用し、そのイオン濃度を２５ｐｐｍとし、その他は実施例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡の状態が良好で初期の結晶が析出しておらず、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げをおこなったところ、実施例１と同様の良好な結果が得られた。
【００２７】
実施例５
この実施例では、Ｓｒのイオン濃度を２５ｐｐｍとし、その他は実施例４と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡の状態が良好で初期の結晶が析出しておらず、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げをおこなったところ、実施例１と同様の良好な結果が得られた。
【００２８】
比較例１
この比較例では、結晶化促進イオンとしてＢａを採用し、そのイオン濃度を３ｐｐｍとし、その他は実施例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡がやや少なく、初期の結晶は析出しておらず、熱処理後の泡の膨れも１５０％とやや高めであることが確認された。これは、Ｂａのイオン濃度がやや低く、泡の膨れ防止用の結晶化が十分ではなかったためと考えられる。
【００２９】
実施例７
この実施例では、Ｂａのイオン濃度を８０ｐｐｍとし、その他は比較例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡の状態が良好で初期の結晶も析出しておらず、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げを行なったところ、実施例１と同様の良好な結果が得られた。
【００３０】
実施例８
この実施例では、Ｂａのイオン濃度を４５０ｐｐｍとし、その他は比較例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。得られたルツボは、表１に示すように保温層内の泡がやや多く、初期の結晶がやや析出し、熱処理後の泡の膨れが０％であることが確認された。またこのルツボを使ってシリコン単結晶の引き上げを行なったところ、実施例１と同様の良好な結果が得られた。
【００３１】
比較例２
この比較例では、Ｂａのイオン濃度を８００ｐｐｍとし、その他は比較例１と同様の条件で石英ガラスルツボを製造し、同様の試験を行った。この結果を表１に示す。この場合には、表１に示すようにＢａのイオン濃度が高いために保温層の結晶化が発達してしまい、ルツボの成形が困難であることが確認された。
【００３２】
（第２の実施形態）
この実施形態では、結晶化促進剤として窒化アルミニウム粉末を使用し、これをシリカ粉末にアルミニウムイオン換算で１５０ｐｐｍとなる状態で混合し、さらに１０％相当量の水を加えて良く攪拌し、このイオン添加した珪石粉を実施例１と同様の条件で成形・溶融した。さらに溶融時にルツボ内部に珪石粉末を連続的に供給する方法で内面透明層も形成した。
【００３３】
得られたルツボを用いてシリコン単結晶の引き上げ試験を行った。その結果、約３０時間の実験過程ではルツボの変質に原因する熱効率の変化は見られず、実験終了後のルツボの肉厚の増加量は約１０％であり、従来ルツボで見られる６０％以上の肉厚変化量と比べて大幅に抑制されることが確認された。
【００３４】
（第３の実施形態）
実施例１０
この実施例では、硝酸マグネシウムのアルコール溶液とシリカ粉末とを混合、乾燥させ、９００℃で焼成し、そのＭｇイオンが３０ｐｐｍ含まれるシリカ粉末を作成した。このシリカ粉末にＭｇイオンの塩化物をさらに添加し、塩化物濃度で２０ｐｐｍに調整した。このシリカ粉末を用いて第２の実施形態と同様の条件でルツボを作成し、各種試験をおこなった。その結果を表２に示す。
【００３５】
【表２】

【００３６】
得られたルツボは良好な外観を示し、結晶の析出は観察できず、熱線の透過率を測定したところ、表２に示すように９２％であった。
【００３７】
実施例１１
この実施例では、Ｍｇイオンの塩化物濃度を１００ｐｐｍとし、その他は実施例１０と同様の条件でルツボを作成し、各種試験をおこなった。その結果を表２に示す。得られたルツボは良好な外観を示し、結晶の析出は観察できず、熱線の透過率を測定したところ、表２に示すように８７％であった。
【００３８】
実施例１２
この実施例では、Ｍｇイオンの塩化物濃度を４００ｐｐｍとし、その他は実施例１０と同様の条件でルツボを作成し、各種試験をおこなった。その結果を表２に示す。得られたルツボは良好な外観を示し、結晶の析出は観察できず、熱線の透過率を測定したところ、表２に示すように７４％であった。
【００３９】
実施例１３
この実施例では、硝酸イットリウムのアルコール溶液とシリカ粉末とを混合、乾燥させ、９００℃で焼成し、そのＹイオンが３０ｐｐｍ含まれるシリカ粉末を作成した。このシリカ粉末にＹイオンの塩化物をさらに添加し、塩化物濃度で５０ｐｐｍに調整した。このシリカ粉末を用いてと同様の条件でルツボを作成し、各種試験をおこなった。その結果を表２に示す。得られたルツボは良好な外観を示し、結晶の析出は観察できず、熱線の透過率を測定したところ、表２に示すように９０％であった。
【００４０】
実施例１４
この実施例では、Ｙイオンの塩化物濃度を３００ｐｐｍとし、その他は実施例１０と同様の条件でルツボを作成し、各種試験をおこなった。その結果を表２に示す。得られたルツボは良好な外観を示し、結晶の析出は観察できず、熱線の透過率を測定したところ、表２に示すように８２％であった。
【００４１】
以上の実施例１０〜１４では、イオン種の違いに関係なく、塩化物の添加量が増すにつれて初期の透過率が増加する傾向を示し、その添加量の大小により保温層の保温能力を制御できることが確認された。
【００４２】
そこで、実施例１０〜１４で得られた各ルツボを１５００℃で１０時間、２０Ｔｏｒｒで熱処理したところ、すべてのサンプルで泡の膨れが２０％以内であることが確認された。同様に熱線の透過率を測定し、熱処理前と比べたところ、すべてのサンプルで実質的に変化していないことが判明した。
【００４３】
比較のため、従来のルツボを評価したところ、熱処理前の熱線透過率は９４％とやや高く、上記と同様の熱処理を行うと熱線透過率が５５％まで減少することが確認された。正常な形状のルツボに成形するためには透過率の変化は３％以内の範囲でしか調整できないこともわかった。
【００４４】
【発明の効果】
以上説明したように、この発明によれば、ルツボ内部の多数の気泡を含む保温層に結晶化促進剤を含有させる構造を構築したため、ルツボ内に入れた多結晶シリコンを高温溶解させている間に結晶化促進剤を核として速やかに結晶が成長し、泡の周囲のガラスが固体化して泡の内圧増加による膨張を抑制できる。その結果、従来のように保温層中の泡の膨れにより、単結晶引き上げの温度制御条件が乱れたり、熱効率が低下したり、消費電力が増加したりする事態を大幅に解消できる。
【００４５】
結晶化促進剤として、金属イオン化合物でその一部あるいはすべてにルツボ溶融時に分解してガス成分を発生させる化合物、たとえば炭酸塩、水酸化物、塩化物、窒化物などの化合物を使用すれば、その添加量を調整することにより気泡の含有量を簡単にかつ直接的に制御できる利点がある。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for producing a quartz glass crucible for melting silicon used in a silicon single crystal pulling apparatus, and in particular, a structure of a layer containing bubbles in a crucible capable of pulling a high purity single crystal over a long period of time and its structure Concerning the method of manufacturing.
[0002]
[Prior art]
In a silicon single crystal pulling apparatus for pulling up a seed crystal immersed in a silicon melt to produce a single crystal, a quartz glass crucible for forming and holding a silicon melt is used. The crucible wall includes a layer containing a large number of bubbles (hereinafter referred to as “heat-retaining layer”), and this heat-retaining layer plays a role of maintaining the temperature of the silicon melt in the crucible and homogenizing its temperature distribution. Bear. Since the heat retention capacity in this case is determined by the bubble state such as the number of bubbles and their size, this state can be adjusted based on the particle size distribution of the silica powder at the time of crucible molding and the melting conditions for adjusting the atmosphere at the time of melting. Insulation capacity is indirectly controlled.
[0003]
[Problems to be solved by the invention]
In the conventional quartz glass crucible described above, the temperature and pressure state of the crucible at the time of pulling the single crystal, for example, about 1500 ° C. and the condition under reduced pressure, the bubbles in the heat insulating layer expand and the diameter increases, so the degree of heat ray scattering Increases with time, and the amount of heat applied from the heater to the silicon melt through the crucible decreases. As a result, subtle temperature control becomes difficult, the accuracy of various conditions for pulling the single crystal is disturbed, and power consumption is significantly increased due to a decrease in thermal efficiency.
[0004]
On the other hand, the method of controlling the heat retaining capacity of the heat retaining layer is indirectly performed by empirically determining the target residual amount of foam, so in order to improve accuracy, extremely sophisticated and advanced technology is required for setting the manufacturing conditions. There was also a problem that was required.
[0005]
In particular, various problems in the heat insulating layer become more prominent when pulling up a large-diameter silicon single crystal that has been required in recent years. In this case, a quartz glass crucible with a large diameter is used, and the pulling process takes a long time, so that the heat load on the crucible wall increases, the weight of the silicon melt increases, and the crucible is deformed or altered. This is because there are inherent problems such as a reduction in the single crystallization rate, which is amplified as a synergistic effect together with the above-mentioned problems.
[0006]
An object of the present invention is to improve such a conventional problem and to provide a method for producing a quartz glass crucible that can suppress expansion of bubbles in a heat insulating layer when the crucible is used. Another object of the present invention is to easily and directly control the bubble content.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the present inventor has conducted various experiments and examinations so that the foam of the heat retaining layer does not substantially swell even under the condition of using the quartz glass crucible. As a result, when a predetermined crystallization accelerator is included in the heat retaining layer, the wall of the quartz crucible is melted while the polycrystalline silicon for silicon melt placed in the opening of the quartz crucible is dissolved at a high temperature. We have found that quartz glass grows rapidly around the body with a crystallization accelerator as a nucleus in the body, and the expansion of the bubbles can be prevented even if the internal pressure of bubbles increases due to the solidified crystals.
[0008]
The present invention has been completed based on such knowledge, and is a method for producing a quartz glass crucible comprising a crucible wall having a silicon single crystal pulling opening, and the crucible wall is made of quartz glass. A heat insulating layer having a large number of bubbles for maintaining heat in the opening and distributed in layers in the crucible wall is provided, and the heat insulating layer includes a crystallization accelerator for the quartz glass.
[0009]
As an example of the crystallization accelerator within the scope of the present invention, various metal ions, fine powder of crystalline silica represented by cristobalite, tridymite, and the like can be exemplified. In addition, any additive material that promotes crystallization from the molten state of quartz glass is applicable. In particular , experiments conducted by the present inventors have shown that better results are obtained when metal ions such as Ca, Mg, Sr, Ba, Al, and Y are used as crystallization accelerators. If these metal ions are used, it is considered that the diffusion in the quartz glass is small, and it is difficult to be altered even when exposed to high temperatures.
[0010]
In the method for producing a quartz glass crucible according to the present invention, a silica powder is placed in a crucible molding mold, molded, the molded body is heated and melted, and then cooled to include a heat insulating layer having a large number of bubbles. A method for forming a crucible wall of glass, characterized in that, before forming the silica powder, a crystallization accelerator for the quartz glass is added to the silica powder to adjust a raw material for the heat retaining layer. To do. That is, in this manufacturing method, a silica powder layer is formed in a mold for a silica glass crucible, and melted from the inside of the mold using a heat source such as an arc while rotating the formed body. In this layer, a crystallization accelerator is added in advance to a portion that becomes a heat retaining layer containing a large number of bubbles after dissolution.
[0011]
As a crystallization accelerator, a metal ion compound powder or solution may be used and mixed with silica powder. In this case, it is preferable that some or all of these metal ion additives be decomposed when the crucible melts, for example, carbonate, hydroxide, chloride, carbide, nitride, etc., to generate a gas component. Shall be used. This is because adjusting the amount of the compound added has the advantage that the foam content can be freely adjusted.
[0012]
As an example of a method for adding metal ions, an aqueous solution containing these ionic compounds is added to silica powder to form a slurry, which is then molded and dried in a crucible mold, and the silica body aggregates during and after molding. It has been found that there is an advantage that the molding of the silica powder becomes very easy due to the increased force.
[0013]
The crystal accelerator needs to have a concentration that can sufficiently prevent the bubble from expanding by sufficiently growing the crystal when the silicon crystal is pulled, but if the concentration is too high, the crystal will develop during crucible formation, and the strength of the crystal accelerator will increase. Causes a decrease. According to the experiments and examinations conducted by the present inventors, a concentration of 5 ppm or more is necessary to obtain a substantial effect as a countermeasure against bubble swelling, but if it exceeds 500 ppm, crystallization occurs during crucible molding, which is good. I found that I could not get a crucible. Therefore, the concentration of the crystallization accelerator is preferably 5 ppm or more and 500 ppm or less.
[0014]
Many crucibles for pulling a silicon single crystal usually have a layer made of transparent quartz glass inside a heat retaining layer containing bubbles. The present invention can also be applied to a method of manufacturing a crucible including such an inner transparent layer. In this case, for example, a layer of silica powder to which no crystallization accelerator is added is formed inside the silica layer containing the crystallization accelerator that constitutes the heat retaining layer before melting, and the layer without the addition of the accelerator is melted. A method of forming a transparent layer in a reduced pressure atmosphere or the like, or a method of forming a heat-retaining layer containing bubbles and then melting while supplying a silica source from the outside, and overlaying the transparent layer on the crucible inner surface side, etc. Is applicable.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of a method for producing a quartz glass crucible according to the present invention will be described below with reference to the drawings.
[0016]
(First embodiment)
Example 1
In this example, Ca ions were used as a crystallization accelerator to be added to the heat retaining layer, the ion concentration was adjusted to 5 ppm, a quartz glass crucible was produced, and various tests were performed on this crucible. The results are shown in Table 1.
[0017]
[Table 1]

[0018]
A method for producing the quartz glass crucible shown in Table 1 will be described. First, an aqueous Ca ion solution is prepared as a crystallization accelerator. In this adjustment, a Ca ion hydroxide is prepared, an aqueous solution having an ion concentration of 0.5% is prepared using the hydroxide, and an aqueous solution diluted to a predetermined concentration is prepared using this as a mother liquor. This aqueous solution is mixed with silica powder at a predetermined weight ratio (aqueous solution: silica powder = 25: 75), and the mixture is put into a crucible mold. This type uses a 18-inch size, and the outside is water-cooled.
[0019]
By rotating this mold, a mixture of silica powder is formed into a crucible shape, and the molded powder is heated and dried. For this heating / drying, a method such as passing warm water through the water-cooling part while continuing the rotation of the mold is used. The dried silica powder was condensed and examined by X-ray diffraction. As a result, it was confirmed that Ca ion carbonate was formed.
[0020]
Next, a layer of ordinary silica powder, that is, a silica powder to which no crystallization accelerator is added is formed on the silica molded body, and a carbon electrode provided inside is used while rotating the mold with respect to the molded body. A quartz glass crucible was obtained by arc melting. Here, the glass layer on the inner surface side of the crucible was made transparent by using a method such as maintaining the He atmosphere during the melting period of the silica powder layer to which no accelerator was added and evacuating the silica compact.
[0021]
With respect to the obtained crucible, the formation state of bubbles in the heat insulating layer and the presence or absence of crystallization were investigated. As a result, as shown in Table 1, it was confirmed that although the amount of bubbles in the heat retaining layer was slightly small, the generation was generally good and crystals were not precipitated.
[0022]
When this crucible was heat-treated at 1500 ° C. under a reduced pressure of 20 Torr for 5 hours, it was confirmed that it was about 10% as shown in Table 1. For comparison, when a conventional crucible to which no crystallization accelerator was added was subjected to heat treatment under the same conditions, it was confirmed that the bubbles expanded about twice (200%).
[0023]
Next, using this crucible, we tried to pull up the silicon single crystal. As a result, it was confirmed that none of the heat insulating layer of the crucible was changed over time, the thermal efficiency was excellent until the end of the pulling, and the power consumption could be reduced by about 50% compared to the conventional case. When the crucible state after use was examined, the heat retaining layer was crystallized and cristobalite was detected.
[0024]
Example 2
In this example, a quartz glass crucible was manufactured under the same conditions as in Example 1 except that the Ca ion concentration was 50 ppm, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a good state of bubbles in the heat retaining layer, no initial crystals were precipitated, and the bubble swelling after the heat treatment was 0%. When the silicon single crystal was pulled using this crucible, the same good results as in Example 1 were obtained.
[0025]
Example 3
In this example, a quartz glass crucible was manufactured under the same conditions as in Example 1 except that the Ca ion concentration was 100 ppm, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a good state of bubbles in the heat retaining layer, no initial crystals were precipitated, and the bubble swelling after the heat treatment was 0%. When the silicon single crystal was pulled using this crucible, the same good results as in Example 1 were obtained.
[0026]
Example 4
In this example, Sr was adopted as the crystallization promoting ion, the ion concentration was 25 ppm, and a quartz glass crucible was manufactured under the same conditions as in Example 1, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a good state of bubbles in the heat retaining layer, no initial crystals were precipitated, and the bubble swelling after the heat treatment was 0%. When the silicon single crystal was pulled using this crucible, the same good results as in Example 1 were obtained.
[0027]
Example 5
In this example, a quartz glass crucible was manufactured under the same conditions as in Example 4 except that the ion concentration of Sr was 25 ppm, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a good state of bubbles in the heat retaining layer, no initial crystals were precipitated, and the bubble swelling after the heat treatment was 0%. When the silicon single crystal was pulled using this crucible, the same good results as in Example 1 were obtained.
[0028]
Comparative Example 1
In this comparative example , Ba was adopted as the crystallization promoting ion, the ion concentration was 3 ppm, and a quartz glass crucible was manufactured under the same conditions as in Example 1, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a little amount of bubbles in the heat retaining layer, no initial crystals were precipitated, and the swelling of the bubbles after heat treatment was slightly high at 150%. . This is presumably because the ion concentration of Ba was slightly low and the crystallization for preventing bubble expansion was not sufficient.
[0029]
Example 7
In this example, a quartz glass crucible was manufactured under the same conditions as in Comparative Example 1 except that the ion concentration of Ba was 80 ppm, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a good state of bubbles in the heat retaining layer, no initial crystals were deposited, and the bubble swelling after the heat treatment was 0%. Also was subjected to a pulling of the silicon single crystal using the crucible, good results similar to those in Example 1 were obtained.
[0030]
Example 8
In this example, a quartz glass crucible was manufactured under the same conditions as in Comparative Example 1 except that the ion concentration of Ba was 450 ppm, and the same test was performed. The results are shown in Table 1. As shown in Table 1, it was confirmed that the obtained crucible had a little more bubbles in the heat retaining layer, some initial crystals were precipitated, and the bubble swelling after the heat treatment was 0%. Also was subjected to a pulling of the silicon single crystal using the crucible, good results similar to those in Example 1 were obtained.
[0031]
Comparative Example 2
In this comparative example , a quartz glass crucible was manufactured under the same conditions as in Comparative Example 1 except that the ion concentration of Ba was 800 ppm, and the same test was performed. The results are shown in Table 1. In this case, as shown in Table 1, since the Ba ion concentration was high, crystallization of the heat retaining layer developed, and it was confirmed that crucible molding was difficult.
[0032]
(Second Embodiment)
In this embodiment, aluminum nitride powder is used as a crystallization accelerator, and this is mixed with silica powder in a state of 150 ppm in terms of aluminum ions. Further, 10% equivalent amount of water is added and stirred well. The added silica powder was molded and melted under the same conditions as in Example 1. Furthermore, an inner transparent layer was also formed by continuously supplying silica powder into the crucible during melting.
[0033]
A pulling test of the silicon single crystal was performed using the obtained crucible. As a result, there was no change in thermal efficiency due to the crucible alteration in the experimental process for about 30 hours, and the increase in crucible wall thickness after the experiment was about 10%, which is more than 60% seen with conventional crucibles. It was confirmed that the thickness was significantly suppressed as compared with the wall thickness change amount.
[0034]
(Third embodiment)
Example 10
In this example, an alcohol solution of magnesium nitrate and silica powder were mixed, dried, and fired at 900 ° C. to produce silica powder containing 30 ppm of Mg ions. Mg ion chloride was further added to the silica powder to adjust the chloride concentration to 20 ppm. Using this silica powder, a crucible was created under the same conditions as in the second embodiment, and various tests were performed. The results are shown in Table 2.
[0035]
[Table 2]

[0036]
The obtained crucible showed a good appearance, no crystal precipitation could be observed, and the heat ray transmittance was measured and found to be 92% as shown in Table 2.
[0037]
Example 11
In this example, the crucible was prepared under the same conditions as in Example 10 except that the chloride concentration of Mg ions was 100 ppm, and various tests were performed. The results are shown in Table 2. The obtained crucible showed a good appearance, no crystal precipitation could be observed, and the heat ray transmittance was measured and found to be 87% as shown in Table 2.
[0038]
Example 12
In this example, a crucible was prepared under the same conditions as in Example 10 except that the chloride concentration of Mg ions was 400 ppm, and various tests were performed. The results are shown in Table 2. The obtained crucible showed a good appearance, no crystal precipitation could be observed, and the heat ray transmittance was measured and found to be 74% as shown in Table 2.
[0039]
Example 13
In this example, an alcohol solution of yttrium nitrate and silica powder were mixed, dried, and baked at 900 ° C. to prepare silica powder containing 30 ppm of Y ions. Y ion chloride was further added to the silica powder to adjust the chloride concentration to 50 ppm. A crucible was prepared under the same conditions as those using this silica powder, and various tests were conducted. The results are shown in Table 2. The obtained crucible showed a good appearance, no crystal precipitation could be observed, and the heat ray transmittance was measured and found to be 90% as shown in Table 2.
[0040]
Example 14
In this example, a crucible was prepared under the same conditions as in Example 10 except that the Y ion chloride concentration was 300 ppm, and various tests were performed. The results are shown in Table 2. The obtained crucible showed a good appearance, no crystal precipitation could be observed, and the heat ray transmittance was measured and found to be 82% as shown in Table 2.
[0041]
In the above Examples 10 to 14, regardless of the difference in ion species, the initial transmittance tends to increase as the amount of added chloride increases, and the heat retaining ability of the heat insulating layer can be controlled by the amount of the added amount. Was confirmed.
[0042]
Thus, when each of the crucibles obtained in Examples 10 to 14 was heat-treated at 1500 ° C. for 10 hours at 20 Torr, it was confirmed that the foam expansion was within 20% in all the samples. Similarly, when the transmittance of the heat ray was measured and compared with that before the heat treatment, it was found that all the samples were not substantially changed.
[0043]
For comparison, when a conventional crucible was evaluated, it was confirmed that the heat ray transmittance before heat treatment was slightly high at 94%, and that the heat ray transmittance was reduced to 55% when heat treatment similar to the above was performed. It was also found that the change in transmittance can only be adjusted within 3% in order to form a crucible with a normal shape.
[0044]
【The invention's effect】
As described above, according to the present invention, since the structure in which the crystallization accelerator is contained in the heat insulating layer including a large number of bubbles inside the crucible is constructed, the polycrystalline silicon placed in the crucible is melted at a high temperature. In addition, the crystal grows quickly with the crystallization accelerator as a nucleus, and the glass around the foam is solidified to suppress expansion due to an increase in the internal pressure of the foam. As a result, it is possible to greatly eliminate the situation in which the temperature control conditions for pulling up the single crystal are disturbed, the thermal efficiency is lowered, or the power consumption is increased due to the expansion of bubbles in the heat retaining layer as in the past.
[0045]
As a crystallization accelerator, if a compound such as carbonate, hydroxide, chloride, nitride, etc., which is a metal ion compound that decomposes into a part or all of it at the time of crucible melting and generates a gas component, for example, By adjusting the addition amount, there is an advantage that the bubble content can be controlled easily and directly.

Claims

A method of forming a silica glass crucible wall including a heat insulating layer having a large number of bubbles by heating and melting the molded body by placing silica powder in a crucible molding mold, followed by cooling. Before forming the silica powder, a crystallization accelerator for the quartz glass is added to the silica powder to adjust the raw material for the heat retaining layer, and the concentration of the crystallization accelerator is 5 ppm or more and 500 ppm or less. A method for producing a quartz glass crucible, characterized by using a concentration.