JP4774598B2 - Polysilicon evaluation apparatus and thin film transistor manufacturing system - Google Patents

Description

【００８８】
この自己相関関数Ｒ（τ）は、ある関数ｆ（ｘ）をτだけｘ方向に平行移動させたときの相関を示す関数である。
【００８９】
このポリシリコン膜評価装置２０では、以下のようなウィンナーヒンチンの定理を用いて、ポリシリコン膜の表面画像の自己相関関数を求めている。なお、ここでは、具体的に取り込んだ画像情報を“ｉ”としている。
１ 取り込み画像“ｉ”の２次元フーリエ変換する。
：ｆ=ｆｏｕｒｉｅｒ（ｉ）
２ フーリエ級数“ｆ”を二乗してパワースペクトル“ｐｓ”を生成する。
：ｐｓ＝｜ｆ｜^２
３ パワースペクトル“ｐｓ”を逆フーリエ変換して２次元の自己相関関数“ａｃ”を生成する。
：ａｃ＝ｉｎｖｅｒｓｆｏｕｒｉｅｒ（ｐｓ）
４ 自己相関関数“ａｃ”の絶対値をとり、自己相関関数の実数“ａｃａ”を求める。
：ａｃａ＝｜ａｃ｜
【００９０】
このように生成された自己相関関数“ａｃａ”を表示すると、図１１及び図１２に示すような関数となる。図１１は、自己相関が高い画像、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が良いものの自己相関関数である。それに対して、図１２は、自己相関が低い画像、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が悪いものの自己相関関数である。
【００９１】
ポリシリコン膜評価装置２０は、このようなウィンナーヒンチンの定理を用いて計算した自己相関画像から、さらに、整列方向（即ち、直線性を有する方向）と垂直で、画面上の座標（０，０）を含む面を切り出して、その切り出したときに得られる関数を求める。ここで、画面上の座標（０，０）を含む面を切り出すのは、照明光量やＣＣＤゲイン等の実験パラメータによって変化する自己相関関数からの値を規格化するために行っている。
【００９２】
このように切り出したときに得られる関数が、上述した整列方向と垂直な方向の自己相関関数Ｒ（τ）に対応する関数となる。
【００９３】
また、ここで、上述したステップＳ１〜Ｓ３は、以下の図１３のステップＳ１１〜Ｓ１４に示すように行ってもよい。
【００９４】
また、このような評価の手順に代えて以下のような評価を行ってもよい。
【００９５】
この評価の処理手順は、図１３のフローチャートに示すように、まず、ポリシリコン膜の表面の画像取り込み処理を行う（ステップＳ１１）。続いて、レーザビームの進行方向（直線性がある方向：ｘ方向）と垂直な方向（周期性がある方向：ｙ方向）の取り込み画像の１ライン分を切り出す（ステップＳ１２）。続いて、この１ラインに関して自己相関関数の計算を行う（ステップＳ１３）。続いて、必要に応じて、これらの作業を数回繰り返し、各ラインの平均化を行う（ステップＳ１４）。
【００９６】
この場合における自己相関関数は、ウィンナーヒンチンの定理を用いて、以下のように求められる。なお、ここでは、具体的に取り込んだ１ライン分の画像情報を“ｌ”としている。
１ 取り込み画像の１ライン“ｌ”に関してのフーリエ変換をする。
：ｆｌ=ｆｏｕｒｉｅｒ（ｌ）
２ フーリエ級数“ｆｌ”を二乗してパワースペクトル“ｐｓｌ”を生成する。
：ｐｓｌ＝｜ｆｌ｜^２
３ パワースペクトル“ｐｓｌ”を逆フーリエ変換して２次元の自己相関関数“ａｃｌ”を生成する。
：ａｃｌ＝ｉｎｖｅｒｓｆｏｕｒｉｅｒ（ｐｓｌ）
４ 自己相関関数“ａｃｌ”の絶対値をとり、自己相関関数の実数“ａｃａｌ”を求める。
：ａｃａｌ＝｜ａｃｌ｜
【００９７】
このように生成された自己相関関数ａｃａｌをグラフ上に表すと、図１４及び図１５に示すような関数となる。図１４は、自己相関が高い関数、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が良いものの自己相関関数である。
それに対して、図１５は、自己相関が低い関数、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が悪いものの自己相関関数である。
【００９８】
これら１ラインの自己相関関数を取り込み画像の全てのラインに関して行い、各自己相関関数の平均化を施す。これが上述した整列方向（すなわち、直線性を有する方向）と垂直な方向の自己相関関数Ｒ（τ）に対応する関数となる。
【００９９】
ポリシリコン膜評価装置２０は、続いて、この得られた関数から、極大ピーク値と、サイドピーク値とを求める。そして、サイドピーク値に対する極大ピーク値の比を求め、この値をＡＣ値とする。
【０１００】
したがって、ＡＣ値は、自己相関が高い画像、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が良い場合は、極大ピーク値とサイドピーク値との差が大きくなり、その値が大きくなる。それに対して、自己相関が低い画像、即ち、ポリシリコン膜の表面空間構造の周期性及び直線性が悪い場合は、極大ピーク値とサイドピーク値との差が小さくなり、その値が小さくなる。
【０１０１】
以上のように、ボトムゲート型ＴＦＴ１では、ポリシリコン膜の表面画像を撮像して、その撮像画像の自己相関関数を求め、ポリシリコン膜の表面空間構造の直線性及び周期性を数値化している。
【０１０２】
具体的に、その撮像画像の一例に対するＡＣ値を示すと、図１６に示すようになる。
【０１０３】
以上のようにエキシマレーザアニールがされたポリシリコン膜は、図５に示したように、膜の表面に複数の直線状の凸部が形成され、さらに、この直線状の凸部が周期的に配列されることから、ＡＣ値によりその状態を評価することができる。すなわち、ポリシリコン膜の表面の空間構造が、反射型のグレーティング状となっているため、ＡＣ値で評価することができる。
【０１０４】
（３） つぎに、ＡＣ値を求める際の演算範囲について説明をする。
【０１０５】
上述したようにＡＣ値は、ポリシリコン膜の表面画像を取り込んで、その取り込んだ画像範囲に対して、２次元フーリエ変換を行い算出される。
【０１０６】
通常、その演算範囲（演算する表面画像の範囲で２次元的なもの）が大きければ大きいほど、ポリシリコン膜の表面空間構造の直線性及び周期性を正確に示すこととなる。しかしながら、ＡＣ値は、上述したように２次元フーリエ変換を行って求められるため、その演算量が膨大となるため、より高速に処理行うためには、演算範囲をより小さくすることが望ましい。
【０１０７】
一方、その演算範囲が小さすぎると、表面全体のＡＣ値との誤差が大きくなり、正確にポリシリコン膜の直線性及び周期性を評価することができない。
【０１０８】
このように、ＡＣ値を求める演算範囲は、精度と処理速度とに影響する。従って、より効率的に再現性よくＡＣ値を求めるために、演算範囲を最適に設定しなければならない。
【０１０９】
以下、表面の空間構造の周期性が悪いポリシリコン膜の表面画像、表面の空間構造の周期性が中程度のポリシリコン膜の表面画像、表面の空間構造の周期性が良いポリシリコン膜の表面画像をそれぞれ示し、これらの特徴について考察しながら、ＡＣ値の最適な演算範囲について説明をする。なお、以下、この最適な演算範囲の説明するにあたり、エキシマレーザアニール時におけるレーザの走査方向に平行な方向をＸ方向、レーザの走査方向に直交する方向をＹ方向というものとする。
【０１１０】
図１７（Ａ）に、表面の空間構造の周期性が全体として悪いポリシリコン膜の表面画像を示し、図１７（Ｂ）にこの画像を模式的に示したものを示す。表面の空間構造の周期性が悪い場合、本来ＡＣ値は小さくなるはずである。しかしながら、このように全体として周期性が悪い場合であったとしても、周期性が良くなっている部分が、Ｘ方向に対して発生する特徴がある。この周期性が良くなっている部分は、Ｘ方向にある程度の長さに亘り形成されている。そのため、もし、Ｘ方向の演算範囲を、この周期性が良くなっている部分のより狭い範囲に設定してしまうと、この周期性が良い部分の画像のみを切り出してＡＣ値の演算を行ってしまう可能性がある。すなわち、このような狭い範囲を設定して演算をしてしまうと、本来ＡＣ値が小さくなるべき画像であるにもかかわらず、算出したＡＣ値が大きくなってしまい、正確な値を得ることができない可能性がある。従って、Ｘ方向に対する演算範囲は、周期性が悪いポリシリコン膜の表面画像に対して演算する場合に、この周期性が良くなっている部分が演算範囲に含まれたとしても、充分ＡＣ値が低くなるように定めなければならない。例えば、Ｘ方向に対するＡＣ値の演算範囲を、周期性が良くなっている部分の長さ（図２中Ｌ１の長さ）の２倍以上とし、この周期性が良くなっている部分の影響を半減させ、正確な値を得るようにするのがよい。
【０１１１】
続いて、図１８（Ａ）に、表面の空間構造の周期性が全体として中程度のポリシリコン膜の表面画像を示し、図１８（Ｂ）にこの画像を模式的に示したものを示す。表面の空間構造の周期性が中程度の場合、本来ＡＣ値は中程度の値となるはずである。しかしながら、このように全体として周期性が中程度のポリシリコン膜の表面空間構造は、周期性の良い部分と、周期性の悪い部分とが交互に繰り返されて発生する特徴がある。この周期性が良くなっている部分及び周期性が悪くなっている部分は、Ｘ方向にある程度の長さに亘り形成されている。そのため、もし、Ｘ方向の演算範囲をこの周期性が良くなっている部分より狭い範囲（或いは周期性が悪くなっている部分より狭い範囲）に設定してしまうと、この周期性が良い部分の画像のみ（或いは周期性が悪い部分の画像のみ）を切り出して演算を行ってしまう可能性がある。すなわち、このような狭い範囲を設定して演算をしてしまうと、本来ＡＣ値が中程度になるべき画像であるにもかかわらず、算出したＡＣ値が極端に大きくなってしまったり或いは極端に小さくなってしまったりし、正確な値を得ることができない可能性がある。従って、Ｘ方向に対する演算範囲は、Ｘ方向に対して周期性の良い部分と周期性の悪い部分の繰り返しの１周期分の長さ（図３中Ｌ２の長さ）以上とし、この周期性が良くなっている部分及び悪くなっている部分の繰り返しの影響を半減させ、正確な値を得るようにするのがよい。
【０１１２】
図１９（Ａ）に、表面の空間構造の周期性が全体として良いポリシリコン膜の表面画像を示し、図１９（Ｂ）にこの画像を模式的に示したものを示す。表面の空間構造の周期性が良い場合、本来ＡＣ値は大きくなるはずである。しかしながら、このように全体として周期性が良い場合であったとしても、周期性が悪くなっている部分が、Ｘ方向に対して発生する特徴がある。この周期性が悪くなっている部分は、Ｘ方向にある程度の長さに亘り形成されている。そのため、もし、Ｘ方向の演算範囲をこの周期性が悪くなっている部分より狭い範囲に設定してしまうと、この周期性が悪い部分の画像のみを切り出して演算を行ってしまう可能性がある。すなわち、このような狭い範囲を設定して演算をしてしまうと、本来ＡＣ値が大きくなるべき画像であるにもかかわらず、算出したＡＣ値が小さくなってしまい、正確な値を得ることができない可能性がある。従って、Ｘ方向に対する演算範囲は、周期性が良いポリシリコン膜の表面画像に対して演算する場合に、この周期性が悪くなっている部分が演算範囲に含まれたとしても、充分ＡＣ値が低くなるように定めなければならない。例えば、Ｘ方向に対するＡＣ値の演算範囲を、周期性が悪くなっている部分の長さ（図４中Ｌ３の長さ）の２倍以上とし、この周期性が悪くなっている部分の影響を半減させ、正確な値を得るようにするのがよい。
【０１１３】
Ｘ方向に対する演算範囲は、以上の３つの条件を満足するように設定すると、正確なＡＣ値を得ることができる。すなわち、表面の空間構造の周期性が全体として悪いポリシリコン膜における周期性の良い部分の２倍の範囲以上であること、表面の空間構造の周期性が全体として中程度のポリシリコン膜における周期性の良い部分と周期性の悪い部分の繰り返しの１周期分以上であること、表面の空間構造の周期性が全体として良いポリシリコン膜における周期性の悪い部分の２倍の範囲以上であること、を満足するように設定すると正確なＡＣ値を得ることができる。
【０１１４】
一方、Ｙ方向に対する演算範囲は、最適なエネルギー密度でレーザアニールをした際に形成される複数の直線状の凸部の配列周期に対して、周期性を算出しているわけであるので、自己相関を求めるためには、図１９に示しているように、その２周期分以上の範囲が必要となる。
【０１１５】
なお、精度を高めるために演算範囲を広げる場合、Ｙ方向に演算範囲を広げるよりも、Ｘ方向に演算範囲を広げる方が、精度が高くなる可能性が高い。これは、Ｘ方向に演算範囲を広げれば、周期性が悪いポリシリコン膜に発生している周期性の良い部分の影響、周期性が中程度のポリシリコン膜の周期性が良い部分と悪い部分のＸ方向の繰り返しの影響、周期性が良いポリシリコン膜に発生している周期性の悪い部分の影響が、除去されるためである。従って、演算範囲を設定する場合には、Ｘ方向の長さが、Ｙ方向の長さ以上となっている長方形の範囲とするのがよい。
【０１１６】
また、上述したようにレーザアニールをした際に生じる複数の直線状の凸部の配列の周期は、０．３μｍなので、Ｙ方向の演算範囲は０．６μｍ以上とするのが望ましい。また、経験上、全体として周期性が中程度のポリシリコン膜の場合、Ｘ方向に対する周期性の良い部分と周期性の悪い部分との繰り返し周期は、５μｍであったので、Ｘ方向の演算範囲は５μｍ以上とするのが望ましい。
【０１１７】
（４） つぎに、再現性よくＡＣ値を算出するためのポリシリコン膜評価装置２０の紫外光の光量の制御について説明をする。
【０１１８】
ポリシリコン膜評価装置２０では、紫外光をポリシリコン膜の表面に照射し、その反射光をＣＣＤにより撮像し、その撮像画像からＡＣ値を算出している。
【０１１９】
ここで、ＡＣ値は、ＣＣＤにより撮像された撮像画像に基づき算出されるので、その撮像画像の明るさによって、ＡＣ値の値が変動してしまう場合がある。例えば、ＣＣＤへの露光量不足の場合や露光量がオーバーした場合などは、その撮像画像自体のコントラストが少なくなり、凹凸の状態を認識できなくなるためである。
【０１２０】
図２０に、ＣＣＤへの紫外光の露光量に対するＡＣ値の変動を示したグラフを示す。露光量（μＪ／ｃｍ^２）は、ＣＣＤのシャッタが開いている時間（ｓ）×照射光量（ｍＷ／ｃｍ^２）で表される。ここで、この図２０に示すように、ＡＣ値は、露光量がある一定の範囲（図中Ｔ_１で示した範囲）にある場合、その値が一定となっている。すなわち、露光量をこの一定の範囲内とするように制御することによって、ＡＣ値が再現性よく得られるように、ＣＣＤへの入射光量を制御することができる。
【０１２１】
そのため本実施の形態のポリシリコン膜評価装置２０では、対物レンズからの紫外光の出射光量を例えばフォトディテクタ等で検出して、ポリシリコン膜への照射光量を検出する。そして、この照射光量とＣＣＤのシャッタスピードから露光量を算出し、ＣＣＤへの入射光量をモニタする。そして、この入射光量をモニタリングしながら、例えば、紫外光レーザのパーワー、シャッタスピード等をフィードバック制御して、露光量が上記一定の範囲に入るようにコントロールする。
【０１２２】
また、図２１に、撮像画像の画面の明るさに対するＡＣ値の変動を示したグラフを示す。撮像画像の画面の明るさは、ここでは２５６階調で表している。ここで、この図２１に示すように、ＡＣ値は、撮像画像の明るさがある一定の範囲（図中Ｔ_２で示した範囲）にある場合、その値が一定となっている。すなわち、撮像画像の明るさをこの一定の範囲内とするように制御することによって、ＡＣ値が再現性よく得られるように、ＣＣＤへの入射光量を制御することができる。このように上述したように露光量を検出するのではなく、撮像画像の明るさを検出しても、ＣＣＤへの入射光量を制御することが可能である。
【０１２３】
そのため、本実施の形態のポリシリコン膜評価装置２０では、撮像画像の画面の明るさを、ＣＰＵなどで演算して、入射光量をモニタする。そして、この入射光量をモニタリングしながら、例えば、紫外光レーザのパーワー、シャッタスピード等をフィードバック制御して、露光量が上記一定の範囲に入るようにコントロールする。
【０１２４】
（５） 次に、上述したように演算した結果得られるＡＣ値と、ポリシリコン膜のグレーンサイズ及びポリシリコン膜に与えられるエネルギーとの関係を説明する。
【０１２５】
ＡＣ値は、図２２に示すように、エキシマレーザアニールによりポリシリコン膜に与えられるエネルギーが、あるエネルギーＥ_５となったときからその値が比例的に上昇し、あるエネルギーＥ_Ｔでその値が最大となる。そして、ＡＣ値は、この最大となるエネルギーＥ_Ｔでピーク値を迎え、その後その値が比例的に減少し、あるエネルギーＥ_Ｂ２でその減少が終了し、その値が最小値となる。このようにＡＣ値は、与えられるエネルギーに対してピーク特性を有している。
【０１２６】
このようなＡＣ値のピーク特性を、図３で示したポリシリコン膜のグレーンサイズの変化の特性に重ね合わせてみると、図２３に示すようになる。この図２３に示すように、ＡＣ値のピーク特性を示すグラフの最大値が、ポリシリコン膜のグレーンサイズが適正となるエネルギー範囲内に入ることが分かる。さらに、ＡＣ値が比例的に上昇を開始するエネルギーＥ_５が、ポリシリコン膜に与えてグレーンサイズが適正となる許容最低エネルギーＬよりも低くなる。また、ＡＣ値の比例的な減少が停止して最低値となったときのエネルギーＥ_Ｂ２が、ポリシリコン膜の結晶粒径が微結晶化する閾値のエネルギーである許容最大エネルギーＨよりも高くなる。
【０１２７】
したがって、このようなピーク特性を有するＡＣ値からポリシリコン膜のグレーンサイズが良好なものであるかどうかを評価する場合には、ＡＣ値が図２３中太線で示した範囲の値に入っているかどうかを判断すればよいこととなる。
【０１２８】
（６） このような特性を有するＡＣ値を評価して、ポリシリコン膜が良品であるか否かの検査を行う場合には、例えば、検査対象となる基板のＡＣ値が、許容最低エネルギーＬ或いは許容最大エネルギーＨを与えたときに求められるＡＣ_Ｌのいずれか高い方の値を閾値として、この閾値よりも大きければ良品であると判断することにより検査が可能である。
【０１２９】
また、このような特性を有するＡＣ値を評価して、エキシマレーザアニール装置から出射されるレーザのエネルギ密度を最適に設定する場合には、例えば、エキシマレーザのエネルギ密度を変動させながら、複数個の基板に対してレーザアニール処理を行う。そして、各エネルギ密度に対応させたＡＣ値の特性図を描き、具体的には、図２２に示すような特性図を描き、この特性図から最適なエネルギ密度を求めればよい。
【０１３０】
（７） ところで、上述したようにボトムゲート型ＴＦＴでは、ゲート電極３がポリシリコン膜６の下層に位置しているので、レーザアニールを行った場合におけるエネルギーの拡散性が、ガラス基板２上（ソース／ドレイン領域上）のポリシリコン膜６よりも、ゲート電極３上のポリシリコン膜６の方が高くなる。そのため、エキシマレーザアニール装置から与えられるエネルギ密度が同一であっても、ゲート電極３上のポリシリコン膜６と、ガラス基板２上（ソース／ドレイン領域上）のポリシリコン膜６とで、与えられるエネルギーが異なることとなり、その影響からグレーンサイズが双方で異なってしまう。
【０１３１】
一般に、エキシマレーザアニール装置によりレーザアニールを行った場合、ゲート電極上に位置するポリシリコン膜と、ガラス基板上（ソース／ドレイン領域上）に位置するポリシリコン膜とで、エネルギ密度を変化させるような制御をすることはできず、同一のエネルギ密度の設定で、一律にエキシマレーザアニールを行うこととなる。
【０１３２】
そのため、ボトムゲート型ＴＦＴでは、エキシマレーザのエネルギ密度に対するＡＣ値の特性は、図２４に示すようになり、ガラス基板上（ソース／ドレイン領域上）と、ゲート電極上とで、そのピーク値が異なる位置となってしまう。具体的には、ガラス基板上（ソース／ドレイン領域上）に位置するポリシリコン膜のＡＣ値の方が、ゲート電極上に位置するポリシリコン膜よりも低いエネルギ密度でピーク値を迎えることとなる。
【０１３３】
したがって、ＡＣ値を評価して、ポリシリコン膜が良品であるか否かの検査を行う場合、並びに、ＡＣ値を評価してエキシマレーザアニール装置から出射されるエネルギ密度を最適に設定する場合には、この両者（ガラス基板上及びゲート電極上）のポリシリコン膜が良好となるような値とする必要がある。
【０１３４】
続いて、ボトムゲート型ＴＦＴのポリシリコン膜について、エキシマレーザのエネルギ密度に対するＡＣ値の具体的な実験データの一例を図２５に示す。この図２５に示すように、ＡＣ値は、ゲート電極上とガラス基板上とで異なるピーク値となる特性となる。例えば、この図２５で示した特性図上では、エキシマレーザアニールでのエネルギ密度を３８０ｍＪで設定することが最適であることが分かる。
【０１３５】
（８） 以上のように、ボトムゲート型ＴＦＴに形成されたポリシリコン膜を評価する場合に、ポリシリコン膜表面の空間構造の直線性及び／又は周期性を評価することにより、非破壊で容易にポリシリコンの検査をすることができ、検査工程を製造工程に組み込むことが可能となる。また、この直線性及び／又は周期性を数値化するので、目視検査等によらず数値演算が可能となる。さらに、数値化を行って評価を行うので、そのため自動検査が可能となり、また、高い精度で客観的な検査を行うことができる。また、検査結果をアニール処理工程にフィードバックして、製造する薄膜トランジスタの歩留まりを高くすることができる。
【０１３６】
なお、以上ポリシリコン膜を撮像する装置として、波長２６６ｎｍの紫外光レーザを用いた顕微鏡装置を適用していたが、ポリシリコン膜の表面空間構造の直線性及び／又は周期性を評価するための元画像を撮像する装置は、このような装置に限られない。例えば、ＳＥＭにより観察した画像に基づいて、ポリシリコン膜の表面空間構造の直線性及び／周期性を評価しても良い。例えば、図２６に示すように、紫外光レーザを用いた顕微鏡装置（ＤＵＶ）で撮像した画像に基づきＡＣ値を求めた場合の特性と、ＳＥＭで撮像した画像に基づきＡＣ値を求めた場合の特性とを比較すると、ＳＥＭの方がより精彩な画像となるため相対的にＡＣ値が低くなるもの、その特性を示す曲線はほぼ同一となることが分かる。
【０１３７】
また、直線性及び／周期性を数値化する手法として自己相関関数を用いた例を詳細に説明したが、数値化の手法もこの自己相関関数を用いた例に限られない。
【０１３８】
ボトムゲート型ＴＦＴの製造プロセスにおける具体的な適用例
つぎに、ボトムゲート型ＴＦＴの製造プロセスに上記ポリシリコン膜評価装置２０を適用した具体的な適用例について説明する。
【０１３９】
まず、図２７に示すような、ボトムゲート型ＴＦＴのポリシリコン膜の撮像画像から得られたＡＣ値を評価して、その評価結果をエキシマレーザアニール装置にフィードバックし、このエキシマレーザアニール装置３０から出射されるレーザパワーを最適に設定する適用例（ＥＱＣ：Equipment Quality Control）について説明する。
【０１４０】
エキシマレーザアニール装置は、上述したようにレーザパワーの設定値に対して、実際のレーザパワーの出力値の変動が比較的に大きい。出力されるレーザパワーは、ガウス分布的な特性を示しばらつきが生じ、所定のパワー設定値に対してある程度のばらつきが生じる。これに対して、ボトムゲート型ＴＦＴの場合、ポリシリコン膜に与えられるエネルギーの製造マージン（この範囲より外のエネルギーが与えられた場合には不良品となるエネルギー範囲）は、そのばらつきに対して相対的に大きな値となる。
【０１４１】
したがって、図２８に示すように、ポリシリコン膜の製造マージンの中心位置がレーザパワーの設定値の最適値となり、この最適値にレーザパワーが設定してあれば、レーザパワーが変動したとしてもポリシリコン膜に与えられるエネルギーは製造マージン内に入ることとなり、高い歩留まりを得ることができる。しかしながら、図２９に示すように、レーザパワーの設定値が、製造マージンの最適値に設定されていない場合には、レーザパワーが変動すると、ポリシリコン膜に与えられるエネルギーが製造マージンから外れる場合が多く、歩留まりが低くなってしまう。
【０１４２】
そのため、本適用例は、ボトムゲート型ＴＦＴのＡＣ値のピーク特性を利用して、以下のように、エキシマレーザアニール装置のレーザパワーを最適値に設定する。
【０１４３】
まず、本適用例では、ポリシリコン膜を形成した複数枚の基板を製造する。このとき、各基板毎にエキシマレーザアニール装置のレーザパワーの設定を変化させ、それぞれの基板についてゲート電極上及びガラス基板上のＡＣ値を求める。
【０１４４】
すると、図３０や図３１に示すような、ＡＣ値のピーク曲線をグラフ上に描くことができる。
【０１４５】
このようなＡＣ値のピーク曲線を描くと、ゲート電極上とガラス基板上ともに良好なグレーンサイズを得られるレーザパワーの許容範囲（ポリシリコン膜の製造マージン）を求めることができる。具体的には、製造マージンの下位限界のレーザパワーは、ゲート電極上のポリシリコン膜に与えられるエネルギーの最低許容エネルギー（Ｌ）に対応したレーザパワー、具体的には、図３０及び図３１に示したゲート電極上のＡＣ値の太線で描いた部分の左端のレーザパワー（ＭＯ（Ｌ））となる。また、製造マージンの上位限界のレーザパワーは、ガラス基板上のポリシリコン膜に与えられるエネルギーの最高許容エネルギー（Ｈ）に対応したレーザパワー、具体的には、図３０及び図３１に示したガラス基板上のＡＣ値の太線で描いた部分の右端のレーザパワー（Ｇ（Ｈ））となる。
【０１４６】
そして、このように求めた製造マージンの中間値を求め、この中間値におけるレーザパワーを最適値として設定する。
【０１４７】
以上のようにＡＣ値を求めて、製造マージンを求め、この製造マージンを最適値として設定することによって、ボトムゲート型ＴＦＴの歩留まりを高くすることができる。
【０１４８】
【発明の効果】
本発明にかかるポリシリコンの評価方法並びに薄膜トランジスタ製造システム及び方法では、レーザアニール処理により形成されたポリシリコン膜の膜表面を撮像して、その撮像画像の自己相関を算出し、空間構造の直線性及び／又は周期性を評価する。その際に、レーザアニール処理でのレーザの走査方向に直交した方向に対する自己相関の演算範囲を、少なくともポリシリコン膜の表面空間構造の周期性が現れた際における２周期以上とする。
【０１４９】
また、本発明では、レーザアニール処理でのレーザの走査方向に並行した方向に対する自己相関の演算範囲を、上記直交した方向より長くする。
【０１５０】
このことにより本発明では、低温多結晶化プロセスで形成されたポリシリコン膜の表面の空間構造を自己相関関数を用いて数値化して膜の状態を評価する際に、その演算範囲を最適に設定し、再現性がよく且つ短時間で膜の状態を評価することができる。
【図面の簡単な説明】
【図１】ボトムゲート型ＴＦＴの模式的な断面構成を説明する図である。
【図２】ポリシリコン膜を形成したのちのボトムゲート型ＴＦＴの断面構造を説明するための図である。
【図３】ポリシリコン膜のグレーンサイズと、エキシマレーザアニールで与えられるエネルギーとの関係を説明するための図である。
【図４】出力するレーザのエネルギ密度を最適値としてエキシマレーザアニールを行ったときのポリシリコン膜の膜表面の画像と、最適値よりも少ないパワーとしたときのポリシリコン膜の膜表面の画像と、最適値よりも大きいパワーとしたときのポリシリコン膜の膜表面の画像とを説明するための図である。
【図５】（Ａ）は、エキシマレーザアニールをする前のＴＦＴ基板の断面構造を説明するための図である。（Ｂ）は、良好なエネルギ密度でエキシマレーザアニールを行った場合のＴＦＴ基板の断面構造を説明するための図である。（Ｃ）は、良好ではないエネルギ密度でエキシマレーザアニールを行った場合のＴＦＴ基板の断面構造を説明するための図である。
【図６】エキシマレーザアニールをする前の工程に、表面酸化工程を設けたＴＦＴの製造プロセスを説明するための図である。
【図７】直線性および周期性があるポリシリコン膜の撮像画像を模式的に表した図である。
【図８】直線性および周期性がないポリシリコン膜の撮像画像を模式的に表した図である。
【図９】本発明の実施の形態のポリシリコン膜の評価装置の構成図である。
【図１０】ポリシリコン膜の評価手順を説明するためのフローチャートである。
【図１１】周期性が高い場合の自己相関関数を説明するための図である。
【図１２】周期性が低い場合の自己相関関数を説明するための図である。
【図１３】ポリシリコン膜の他の評価手順を説明するためのフローチャートである。
【図１４】上記他の評価手順で評価した場合の周期性が高い場合の自己相関画像を説明するための図である。
【図１５】上記他の評価手順で評価した場合の周期性が低い場合の自己相関画像を説明するための図である。
【図１６】具体的な撮像画像に対する求められたＡＣ値の特性を説明する図である。
【図１７】表面の空間構造の周期性が悪いポリシリコン膜の表面画像を示す図、及びその模式的な構造を示す図である。
【図１８】表面の空間構造の周期性が中程度のポリシリコン膜の表面画像を示す図、及び、その模式的な構成を示す図である。
【図１９】表面の空間構造の周期性が良いポリシリコン膜の表面画像を示す図、及びその模式的な構造を示す図である。
【図２０】ＣＣＤへの紫外光の露光量に対するＡＣ値の変動を示したグラフを示す。
【図２１】撮像画像の画面の明るさに対するＡＣ値の変動を示したグラフを示す。
【図２２】ポリシリコン膜に与えられるエネルギーに対する自己相関値の特性を説明するための図である。
【図２３】ポリシリコン膜に与えられるエネルギーに対するＡＣ値及びグレーンサイズの特性を説明するための図である。
【図２４】ボトムゲート型ＴＦＴにおけるエキシマレーザのエネルギ密度に対するＡＣ値の特性を説明するための図である。
【図２５】ボトムゲート型ＴＦＴのポリシリコン膜について、エキシマレーザのエネルギ密度に対するＡＣ値の具体的な実験データの一例を説明するための図である。
【図２６】紫外光レーザを用いた顕微鏡装置（ＤＵＶ）で撮像した画像に基づきＡＣ値を求めた場合の特性と、ＳＥＭで撮像した画像に基づきＡＣ値を求めた場合の特性とを比較して説明するための図である。
【図２７】ボトムゲート型ＴＦＴの製造プロセスにポリシリコン膜評価装置を適用した具体的な適用例（ＥＱＣ）の構成を説明するための図である。
【図２８】ポリシリコン膜に与えられるエネルギーの製造マージンと、エキシマレーザのエネルギ密度の変動との関係を説明するための図である（最適にエネルギ密度が設定されている場合）。
【図２９】ポリシリコン膜に与えられるエネルギーの製造マージンと、エキシマレーザのエネルギ密度の変動との関係を説明するための図である（最適にレーザのエネルギ密度が設定されていない場合）。
【図３０】ボトムゲート型ＴＦＴの製造マージンと、エキシマレーザのエネルギ密度との関係の一例を示し、この一例からエネルギ密度の最適値を求めるための方法を説明するための図である。
【図３１】ボトムゲート型ＴＦＴの製造マージンと、エキシマレーザのエネルギ密度との関係の他の一例を示し、この他の一例からエネルギ密度の最適値を求めるための方法を説明するための図である。
【符号の説明】
１ ボトムゲート型ＴＦＴ、２ ガラス基板、３ ゲート電極、４ 第１のゲート絶縁膜、５ 第２のゲート絶縁膜、６ ポリシリコン膜、２０ ポリシリコン膜評価装置、３０ エキシマレーザアニール装置[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an apparatus and method for evaluating polysilicon formed by a low temperature polycrystallization process, and a thin film transistor manufacturing system and method.
[0002]
[Prior art]
In recent years, thin film transistors using a polysilicon film as a channel layer have been put into practical use. When polysilicon is used for the channel layer, the electric field mobility of the thin film transistor becomes very high. For example, when it is used as a driving circuit for a liquid crystal display, etc., high resolution, high speed, miniaturization, etc. are realized. Will be able to.
[0003]
In recent years, a so-called low-temperature polycrystallization process for forming a polysilicon film by heat-treating amorphous silicon using an excimer laser annealing apparatus has been developed. By applying such a low-temperature polycrystalline process to a thin film transistor manufacturing process, thermal damage to the glass substrate is reduced, and a heat-resistant large-area and inexpensive glass substrate can be used.
[0004]
By the way, since the output power of the excimer laser annealing apparatus used in the low-temperature polycrystal process is unstable, the grain size of the formed polysilicon largely fluctuates. For this reason, the polysilicon film formed using the excimer laser annealing apparatus does not always have a good grain size. For example, the polysilicon film becomes a so-called linear defect in which the silicon crystal is microcrystallized, or has a sufficiently large grain size. There is a problem in that it becomes a so-called write failure that cannot be obtained.
[0005]
Therefore, in general, when performing an annealing process using such an excimer laser annealing apparatus, the energy information given to the polysilicon film is fed back to the excimer laser annealing apparatus to set the optimum energy density of the laser. Processing is performed.
[0006]
However, in order to evaluate a polysilicon film, there is conventionally only a sensory method in which a surface image is taken using a spectroscopic ellipso, a scanning electron microscope or the like, and the state of the crystal is judged by visualizing the surface image. It was impossible to judge objectively without contact.
[0007]
Therefore, the present applicant objectively and automatically evaluates the state of the polysilicon film formed by using the low-temperature polycrystallization process, and optimizes the energy density of the laser emitted from the laser annealing apparatus based on the information. Japanese Patent Application Nos. 2000-005994, 2000-005995, and 2000-005996 propose a thin film transistor manufacturing system that achieves the above.
[0008]
Applicants have found that when an amorphous silicon film is subjected to excimer laser annealing to form a polysilicon film, linearity and periodicity appear in the spatial structure of the film surface, depending on the energy density given to the amorphous silicon. We found that the structure of sex and periodicity changes. In each of the above applications, using such characteristics, a surface image of the polysilicon film is captured with ultraviolet light, and the periodicity of the spatial structure of the surface of the polysilicon film is numerically calculated from the captured image using an autocorrelation function. A thin film transistor manufacturing system has been proposed in which the state of the polysilicon film formed is evaluated based on this numerical value, and the evaluation result is fed back to the setting of the energy density of the excimer laser.
[0009]
[Problems to be solved by the invention]
By the way, when the periodicity of the spatial structure of the surface of the polysilicon film is quantified using the autocorrelation function, a two-dimensional Fourier operation must be performed on the captured image.
[0010]
In general, the larger the calculation range of the two-dimensional Fourier calculation, the more accurately the periodicity of the surface space structure of the polysilicon film. However, since the AC value is obtained by performing the two-dimensional Fourier transform as described above, the amount of calculation becomes enormous. Therefore, in order to perform processing at a higher speed, it is desirable to make the calculation range smaller.
[0011]
On the other hand, if the calculation range is too small, an error from the AC value of the entire surface increases, and the periodicity of the polysilicon film cannot be accurately evaluated.
[0012]
The present invention has been made in view of such circumstances, and evaluates the state of the film by quantifying the spatial structure of the surface of the polysilicon film formed by the low-temperature polycrystallization process using an autocorrelation function. At the same time, an object of the present invention is to provide a polysilicon evaluation apparatus and method, and a thin film transistor manufacturing system and method, which can optimally set the calculation range and evaluate the film state with high reproducibility in a short time.
[0013]
[Means for Solving the Problems]
  The polysilicon evaluation apparatus according to the present invention forms an amorphous silicon film on a substrate and applies the formed amorphous silicon film to the amorphous silicon film.ExcimerA polysilicon evaluation apparatus for evaluating a polysilicon film formed by laser annealing, wherein the ultraviolet light is focused on the polysilicon film.laserAn optical system that detects the reflected light and images the reflected light detected by the optical systemCamera with high sensitivity to ultraviolet lightAnd aboveCamera with high sensitivity to ultraviolet lightImaged by,The captured image of the polysilicon film,AutocorrelationRatio of maximum peak value to side peak value of function (AC value)To calculateTheEvaluation of periodicity of surface space structure of polysilicon filmThe crystalline state of the polysilicon film is evaluated based on this periodicity.And an evaluation unit that performs the evaluation,ExcimerAutocorrelation for the direction perpendicular to the laser scanning direction in laser annealing.When calculating the ratio (AC value) of the maximum peak value to the side peak value of the functionAt least the periodicity of the surface space structure of the polysilicon film appeared in the calculation range ofportionIt is characterized by having two or more periods.
[0014]
In this polysilicon evaluation apparatus, the surface of a polysilicon film formed by laser annealing is imaged, the autocorrelation of the captured image is calculated, and the linearity and / or periodicity of the spatial structure is evaluated. At this time, the calculation range of the autocorrelation with respect to the direction orthogonal to the laser scanning direction in the laser annealing process is set to at least two periods when the periodicity of the surface space structure of the polysilicon film appears.
[0015]
In the polysilicon evaluation apparatus according to the present invention, the autocorrelation calculation range for the direction parallel to the laser scanning direction in the laser annealing process is made longer than the orthogonal direction.
[0016]
  In the polysilicon evaluation method according to the present invention, an amorphous silicon film is formed on a substrate, and the formed amorphous silicon film isExcimerA polysilicon evaluation method for evaluating a polysilicon film formed by laser annealing treatment, wherein the ultraviolet light is focused on the polysilicon film.laserThe reflected light is detected and the detected reflected light is detected.Cameras with high sensitivity to ultraviolet lightImageImaged with a camera with high sensitivity to the ultraviolet light,The captured image of the polysilicon film,AutocorrelationRatio of maximum peak value to side peak value of function (AC value)To calculateTheEvaluate the periodicity of the surface space structure of the polysilicon film,Based on this periodicity, the crystalline state of the polysilicon film is evaluated,When evaluating the periodicity of the surface space structure of the polysilicon film,ExcimerAutocorrelation for the direction perpendicular to the laser scanning direction in laser annealing.When calculating the ratio (AC value) of the maximum peak value to the side peak value of the functionAt least the periodicity of the surface space structure of the polysilicon film appeared in the calculation range ofportionIt is characterized by having two or more periods.
[0017]
In this polysilicon evaluation method, the film surface of a polysilicon film formed by laser annealing is imaged, the autocorrelation of the captured image is calculated, and the linearity and / or periodicity of the spatial structure is evaluated. At this time, the calculation range of the autocorrelation with respect to the direction orthogonal to the laser scanning direction in the laser annealing process is set to at least two periods when the periodicity of the surface space structure of the polysilicon film appears.
[0018]
Further, in the polysilicon evaluation method according to the present invention, the calculation range of the autocorrelation with respect to the direction parallel to the laser scanning direction in the laser annealing process is made longer than the orthogonal direction.
[0019]
  A thin film transistor manufacturing system according to the present invention includes a film forming apparatus for forming an amorphous silicon film on a substrate, and an amorphous silicon film.ExcimerA polysilicon film to be a channel layer is generated by laser annealing.ExcimerLaser annealing equipment and ultraviolet light focused on the polysilicon filmlaserAn optical system that detects the reflected light and images the reflected light detected by the optical systemCamera with high sensitivity to ultraviolet lightAnd aboveCamera with high sensitivity to ultraviolet lightImaged by,The captured image of the polysilicon film,AutocorrelationRatio of maximum peak value to side peak value of function (AC value)To calculateTheEvaluation of periodicity of surface space structure of polysilicon filmThe periodicity of the surface space structure of the polysilicon film is evaluated based on this periodicity.A polysilicon inspection apparatus, and the evaluation means of the polysilicon inspection apparatus comprises:ExcimerAutocorrelation for the direction perpendicular to the laser scanning direction in laser annealing.When calculating the ratio (AC value) of the maximum peak value to the side peak value of the functionAt least the periodicity of the surface space structure of the polysilicon film appeared in the calculation range ofportionIt is characterized by having two or more periods.
[0020]
In this thin film transistor manufacturing system, the surface of a polysilicon film formed by laser annealing is imaged, the autocorrelation of the captured image is calculated, and the linearity and / or periodicity of the spatial structure is evaluated. At this time, the calculation range of the autocorrelation with respect to the direction orthogonal to the laser scanning direction in the laser annealing process is set to at least two periods when the periodicity of the surface space structure of the polysilicon film appears.
[0021]
Further, in the thin film transistor manufacturing system according to the present invention, the calculation range of the autocorrelation with respect to the direction parallel to the laser scanning direction in the laser annealing process is made longer than the orthogonal direction.
[0022]
  In the thin film transistor manufacturing method, an amorphous silicon film is formed on a substrate.ExcimerUltraviolet light focused on the polysilicon film after forming a polysilicon film by laser annealinglaserThe reflected light is detected and the detected reflected light is detected.Camera with high sensitivity to ultraviolet lightTake an image withImaged with a camera with high sensitivity to the ultraviolet light,The captured image of the polysilicon film,AutocorrelationRatio of maximum peak value to side peak value of function (AC value)To evaluate the periodicity of the surface space structure of this polysilicon film,Based on this periodicity, the crystalline state of the polysilicon film is evaluated,When evaluating the periodicity of the surface space structure of the polysilicon film,ExcimerAutocorrelation for the direction perpendicular to the laser scanning direction in laser annealing.When calculating the ratio (AC value) of the maximum peak value to the side peak value of the functionAt least the periodicity of the surface space structure of the polysilicon film appeared in the calculation range ofportionIt is characterized by having two or more periods.
[0023]
In the thin film transistor manufacturing method, the surface of the polysilicon film formed by laser annealing is imaged, the autocorrelation of the captured image is calculated, and the linearity and / or periodicity of the spatial structure is evaluated. At this time, the calculation range of the autocorrelation with respect to the direction orthogonal to the laser scanning direction in the laser annealing process is set to at least two periods when the periodicity of the surface space structure of the polysilicon film appears.
[0024]
Further, in the thin film transistor manufacturing method according to the present invention, the calculation range of the autocorrelation with respect to the direction parallel to the laser scanning direction in the laser annealing process is made longer than the orthogonal direction.
[0025]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, as an embodiment of the present invention, a polysilicon film evaluation apparatus and an evaluation method to which the present invention is applied will be described.
[0026]
The polysilicon film evaluation apparatus described below as an embodiment of the present invention is used, for example, for inspection of a polysilicon film formed during a manufacturing process of a thin film transistor having a bottom gate structure (bottom gate type TFT). The bottom gate TFT is a thin film transistor in which, for example, a gate electrode, a gate insulating film, and a polysilicon film (channel layer) are sequentially stacked from a lower layer on a glass substrate. That is, the bottom gate TFT is a TFT having a configuration in which a gate electrode is formed between a polysilicon film serving as a channel layer and a glass substrate. Although an example of an evaluation apparatus for evaluating a bottom gate type TFT will be described here, the present invention is not limited to such a bottom gate type TFT, and a polysilicon film (channel layer) is formed on a glass substrate. Then, it can be applied to a so-called top gate type TFT in which a gate electrode is provided on the upper layer.
[0027]
Bottom gate TFT structure
First, a specific configuration example of the bottom gate TFT will be described with reference to FIG.
[0028]
As shown in FIG. 1, the bottom gate TFT 1 includes a gate electrode 3, a first gate insulating film 4, a second gate insulating film 5, a polysilicon film 6, a stopper on a glass substrate 2 having a thickness of 0.7 mm. 7, a first interlayer insulating film 8, a second interlayer insulating film 9, a wiring 10, a planarizing film 11, and a transparent conductive film 12 are laminated.
[0029]
The gate electrode 3 is formed by forming a film of molybdenum (Mo) of 100 to 200 nm on the glass substrate 2 and then patterning it by anisotropic etching.
[0030]
The first gate insulating film 4 is made of, for example, silicon nitride (SiN) having a thickness of 50 nm._xSilicon nitride (SiN)_xIs laminated on the glass substrate 2 on which the gate electrode 3 is formed.
[0031]
The second gate insulating film 5 is made of, for example, silicon dioxide (SiO 2 having a film thickness of 200 nm.₂This silicon dioxide (SiO2)₂) Are stacked on the first gate insulating film 5.
[0032]
The polysilicon film 6 is made of, for example, polysilicon (p-Si) having a film thickness of 30 to 80 nm. The polysilicon film 6 is formed by being laminated on the second gate insulating film 5. This polysilicon film 6 functions as a channel layer of the bottom gate type TFT 1. The polysilicon film 6 is formed, for example, by polycrystallizing an amorphous silicon (a-Si) film having a thickness of 30 to 80 nm by LPCVD or the like and then annealing the amorphous silicon. . In the polycrystallization process of the polysilicon film 6, laser annealing using an excimer laser that is an ultraviolet laser is used. In this excimer laser annealing process, a laser beam having a linear irradiation surface is emitted, and amorphous silicon is polycrystallized into polysilicon while moving the irradiation region of the pulse beam. The shape of the irradiated surface of the laser beam is, for example, 20 cm in length in the longitudinal direction, 400 μm in length in the short side direction, and a pulse frequency of 300 Hz. The scanning direction of the laser beam when performing the excimer laser annealing treatment is performed in a direction (that is, a short side direction) orthogonal to the longitudinal direction of the irradiation surface of the linear laser.
[0033]
The polysilicon film 6 is polycrystallized by excimer laser annealing, and then is doped with impurities to form source / drain regions. This ion doping is performed after the stopper 7 is formed at a position corresponding to the gate electrode 3 so that impurities are not implanted into the polysilicon film 6 on the gate electrode 3. This stopper 7 is made of, for example, silicon dioxide (SiO 2 having a film thickness of 200 nm.₂), And is formed using the mask used when the gate electrode 3 is formed.
[0034]
The first interlayer insulating film 8 is made of, for example, silicon nitride (SiN) having a film thickness of 300 nm._xThis silicon nitride (SiN)_x) Are stacked on the polysilicon film 6.
[0035]
The second interlayer insulating film 9 is made of, for example, silicon dioxide (SiO 2 having a film thickness of 150 nm.₂This silicon dioxide (SiO2)₂) Is laminated on the first interlayer insulating film 8.
[0036]
In the wiring 10, contact holes for connecting the source / drain regions of the polysilicon film 6 are opened at positions corresponding to the source / drain regions of the first interlayer insulating film 8 and the second interlayer insulating film 9. Thereafter, aluminum (Al) and titanium (Ti) are formed and patterned by etching. The wiring 10 connects the source / drain regions of each transistor formed on the polysilicon film 6 to form a predetermined circuit pattern on the substrate.
[0037]
The flattening film 11 is a film for flattening the surface of the bottom gate TFT 1 and is formed after the wiring 10 is formed, and the film thickness is set to 2 to 3 μm.
[0038]
The transparent conductive film 12 is made of, for example, a transparent conductive material made of ITO or the like, and is a conductive line for connecting the wiring 10 to an external element or external wiring existing outside the bottom gate TFT 1. The transparent conductive film 12 is formed on the planarizing film 11 after the contact hole is opened in the planarizing film 11.
[0039]
In the bottom gate type TFT 1 as described above, since the polysilicon is used for the channel layer, the electric field mobility of the channel layer becomes very high. Therefore, for example, when used as a drive circuit for a liquid crystal display or the like, it is possible to realize high definition, high speed, and small size of the display. In the bottom gate TFT 1 as described above, a so-called low-temperature polycrystallization process is used in which the polysilicon film 6 is formed by heat-treating amorphous silicon using excimer laser annealing. Therefore, thermal damage to the glass substrate 2 in the polycrystallization process is reduced, and it is possible to use an inexpensive glass substrate having a large area.
[0040]
Necessity of inspection of polysilicon film
By the way, it is said that an important factor determining the electric field mobility of the polysilicon film 6 is the grain size of polysilicon. The grain size greatly depends on the energy given to the polysilicon film 6 during the excimer laser annealing process. Therefore, the control and stabilization of the laser energy density during the excimer laser annealing process greatly affect the characteristics and yield of the completed bottom gate TFT 1.
[0041]
However, the excimer laser annealing apparatus used in the excimer laser annealing process has a relatively large output fluctuation of the energy density of the emitted laser. Therefore, when excimer laser annealing is performed using an excimer laser annealing apparatus, the fluctuation of energy given to the polysilicon film 6 with respect to the allowable energy range (manufacturing margin of the polysilicon film 6) in which a good grain size can be obtained. Becomes large, and it is difficult to stably manufacture the polysilicon film 6.
[0042]
Therefore, even when excimer laser annealing is performed under the same conditions, if the grain size of the polysilicon film 6 fluctuates greatly, for example, if the laser energy becomes too large, the silicon crystal is microcrystallized. When the energy of the laser becomes too small, a sufficiently large grain size cannot be obtained, so that sufficient electric field mobility cannot be obtained in any case, resulting in failure.
[0043]
However, as described above, the excimer laser annealing apparatus has a relatively large output fluctuation of the energy density of the emitted laser. Therefore, it is difficult to control the energy density of the laser so that the grain size of the polysilicon film 6 becomes a favorable size.
[0044]
Therefore, in general, when the annealing process is performed using such an excimer laser annealing apparatus, for example, it is formed on the outermost surface at the stage where the polycrystallizing process of the polysilicon film 6 as shown in FIG. 2 is completed. Whether the manufactured product is defective at this stage by inspecting all the crystal states of the polysilicon film 6 that has been manufactured or by randomly extracting products and inspecting the crystal state The energy density of the laser is set by feeding back the energy information given to the polysilicon film 6 to the excimer laser annealing apparatus.
[0045]
For example, the polysilicon film evaluation apparatus evaluates the formed polysilicon film 6 at the stage where the polycrystallizing process of the polysilicon film 6 is completed, and whether the manufactured product is defective at this stage. This is used when determining whether or not to set the laser energy by feeding back information to the excimer laser annealing apparatus.
[0046]
Evaluation principle and method of polysilicon film
(1) First, the evaluation principle of the polysilicon film formed by the excimer laser annealing described above will be described.
[0047]
The mobility of a thin film transistor manufactured as described above is greatly influenced by the grain size of polysilicon. In order to obtain sufficient mobility, it is desirable that the grain size of the polysilicon is larger.
[0048]
The grain size of the polysilicon film greatly depends on the energy applied by excimer laser annealing. As shown in FIG. 3, the grain size of the polysilicon film increases as the applied energy increases. However, a certain predetermined energy (position L in the figure: the energy at this time is the allowable minimum energy L). .) When the above is reached, the grain size becomes sufficiently large, and then the change is reduced and stabilized. When the energy is further increased, the change in grain size increases from a certain position (the position of H in the figure. The energy at this time is the allowable maximum energy H), and the polysilicon becomes a boundary at a certain critical point. Becomes fine crystal grains.
[0049]
Therefore, in general, when excimer laser annealing is performed, irradiation is performed so that the grain size shown in FIG. 3 is within a range from the allowable minimum energy L at which the grain size is sufficiently large to the allowable maximum energy H before microcrystal grains. By controlling the energy density of the laser, a sufficiently large grain size is obtained. The mobility of the completed thin film transistor can be sufficiently increased by irradiating the amorphous silicon film with laser light that provides energy in such a range.
[0050]
(2) Next, an image of the surface of the polysilicon film when the excimer laser annealing is performed with the energy density of the laser being the optimum value, and the film surface of the polysilicon film when the energy density is less than the optimum value The image is compared with the image of the surface of the polysilicon film when the energy density is larger than the optimum value. FIG. 4 shows images in each case. 4A shows an image of the surface of the polysilicon film when the energy density is lower than the optimum value, and FIG. 4B shows the film of the polysilicon film when the energy density is the optimum value. It is a figure which shows the image of a surface, and is a figure which shows the image of the film | membrane surface of a polysilicon film when FIG.4 (C) is set as the energy density larger than an optimal value. Each image shown in FIG. 4 is an image captured by a microscope apparatus using ultraviolet light. Details of the microscope apparatus will be described later.
[0051]
In FIG. 4, the laser scanning direction of excimer laser annealing is the X direction in the figure. Note that, as described above, the amorphous silicon film is irradiated with the laser beam whose irradiation surface is linear, and the scanning direction is a direction orthogonal to the longitudinal direction of the irradiation surface shape of the laser beam.
[0052]
Here, when the energy density at the time of excimer laser annealing is set to the optimum value, the image of FIG. 4B is compared with the other images shown in FIG. 4A and FIG. Such a feature appears.
[0053]
First, the surface image of the polysilicon film when the energy density is set to the optimum value (FIG. 4B) is the surface image of the polysilicon film where the energy density is not optimized (FIGS. 4A and 4B). Compared with C)), the image shows linearity. Specifically, the image is linear with respect to the laser scanning direction (X direction in FIG. 4). That is, the surface of the polysilicon film when the energy density is set to the optimum value is characterized by a regular shape in which linearity appears in the spatial structure.
[0054]
Further, the surface image of the polysilicon film when the energy density is set to the optimum value (FIG. 4B) is the surface image of the polysilicon film where the energy density is not optimized (FIGS. 4A and 4B). Compared with C)), the image shows periodicity. Specifically, the image has a periodicity in a direction orthogonal to the laser scanning direction (Y direction in FIG. 4). That is, the surface of the polysilicon film when the energy density is set to the optimum value is characterized by a regular shape in which periodicity appears in the spatial structure.
[0055]
Therefore, in the polysilicon film evaluation apparatus according to the embodiment of the present invention, the state of the polysilicon film is inspected using the above characteristics. That is, in the polysilicon film evaluation apparatus according to the embodiment of the present invention, the surface image of the polysilicon film after excimer laser annealing is numerically analyzed, and linearity appears in the surface space structure of the polysilicon film. The state of the polysilicon film is inspected by evaluating whether periodicity appears in the surface space structure of the polysilicon film, or whether linearity and periodicity appear in the surface space structure of the polysilicon film. To do.
[0056]
Note that the image of the polysilicon film surface after excimer laser annealing with a good energy density is an image in which linearity and periodicity appear as shown in FIG. A natural oxide film (SiO2) is formed on the amorphous silicon film.₂). The film thickness of this natural oxide film is considered to be 3 to 4 nm. Further, it is considered that the natural oxide film stops oxidation until a certain film thickness and does not reach a certain film thickness.
[0057]
Then, when the amorphous silicon on which the natural oxide film is formed is annealed using an excimer laser, as shown in FIG. 5B, the natural oxide film (SiO₂) Is expected to rise. When the energy density of excimer laser annealing is good, a plurality of linear protrusions parallel to the excimer laser scanning direction are formed, and each straight line has periodicity at equal intervals. become. Therefore, when a film surface image of the polysilicon film is picked up by a microscope apparatus using ultraviolet light, a striped image as shown in FIG. 4 can be referred to due to the influence of reflected diffracted light by the linear convex portion. It is considered a thing. For example, the natural oxide film has a thickness of 3 to 4 nm, and the energy density of the XeCl excimer laser is 300 to 400 mJ / cm.²If so, the interval between the linear protrusions is about 0.3 μm. This interval is about the wavelength of the annealing laser light source.
[0058]
The natural oxide film is considered to change its shape after annealing as shown in FIG. 5C due to factors such as the energy density given by excimer laser annealing and the difference in film thickness of the amorphous silicon film. It is done. For example, when the energy density of excimer laser annealing is not good, it is considered that the linearity and periodicity of the shape of the ridges are destroyed. In addition, the linearity and periodicity may not be constant values at different locations on the same film surface due to various factors such as a difference in film thickness.
[0059]
(3) Here, normally, in the manufacturing process of the thin film transistor, an impurity element removing step using diluted hydrogen fluoride (dilute HF) is provided. Even in the manufacturing process of the bottom gate type TFT 1 described above, impurities are usually removed by dilute HF after the amorphous silicon film is formed and before laser annealing.
[0060]
However, when the substrate is washed with dilute HF after forming the amorphous silicon film, the natural oxide film formed on the amorphous silicon film is dissolved together with the impurities. Therefore, even if laser annealing is performed thereafter, the raised shape of the natural oxide film as shown in FIG. 5B is not formed on the surface of the polysilicon film. Therefore, linearity and periodicity do not appear in the image of the polysilicon film surface, and the state of the polysilicon film cannot be evaluated.
[0061]
Therefore, in order to make it possible to evaluate the state of the polysilicon film, as shown in FIG. 6, a surface oxidation process is performed as a pre-process for performing the laser annealing process. That is, a natural oxide film (SiO 2) is formed on the surface of the amorphous silicon film.₂) Is formed (state 1), and impurities are removed by dilute HF cleaning. When diluted with HF, the amorphous silicon film becomes a natural oxide film on the surface (SiO₂) Is dissolved (state 2). A surface oxidation step is provided as the next step. In this surface oxidation step, for example, the substrate may be left in the air for about one day to form a natural oxide film on the surface of the amorphous silicon film, or the substrate is immersed in ozone water of about 10 ppm, for example. The surface of the amorphous silicon film may be oxidized. Thus, by providing the surface oxidation step before the laser annealing treatment, a natural oxide film can be formed on the amorphous silicon film (state 3). As a result, the natural oxide film is raised by performing the laser annealing treatment. (State 4), and the polysilicon film can be evaluated.
[0062]
(4) Next, an example of a numerical method when the captured image of the polysilicon film has linearity, periodicity, linearity, and periodicity will be described.
[0063]
For example, when a captured image of a polysilicon film having linearity and periodicity is schematically represented, as shown in FIG. 7A, a large number of straight lines are arranged in parallel, and the intervals are constant. expressed. On the other hand, when a captured image of a polysilicon film having neither linearity nor periodicity is schematically represented, as shown in FIG. 8A, an irregular short straight line or the like appears irregularly. . When numerically evaluating the linearity and periodicity from these images, the image is shifted laterally in the direction perpendicular to the direction in which periodicity is expected, and the image when the lateral shift is performed. What is necessary is just to express the correlation of this numerically. For example, when an image having linearity and periodicity is shifted laterally, as shown in FIG. 7B, the image has a high degree of correlation with a large degree of image overlap for every certain period, that is, every certain amount of lateral displacement. An image appears. On the other hand, as shown in FIG. 8B, an image having neither linearity nor periodicity does not appear as a highly correlated image with many overlapping images even when shifted laterally. .
[0064]
By using the concept of quantifying the correlation of images when the image is laterally shifted as described above, the periodicity of the polysilicon film can be quantified and evaluated. Specifically, as one method for realizing such a method, there is a method of obtaining an autocorrelation function of an image, calculating a peak value and a side peak value of the autocorrelation function, and calculating a ratio between them. Here, the peak value is the second minimum value in the y direction from the origin value (used to reduce the defocus value. It may be the first or second value). It shall be the subtracted value. Further, the side peak value refers to a value obtained by subtracting the second minimum value in the y direction from the origin from the second maximum value in the y direction (not including the origin) from the origin.
[0065]
In the present invention, it is also possible to evaluate only one of linearity and periodicity and judge the state of the polysilicon film.
[0066]
In addition, as another example of the numerical method when the captured image of the polysilicon film has linearity, periodicity, linearity, and periodicity, for example, a standardized image is aligned in a direction with uniform linearity. There is a method of adding the values of all pixels and taking the degree of modulation. In addition, there is a method of taking the intensity of a certain frequency component by two-dimensional Fourier transform of a standardized image. Also, the coordinates of the extreme values (local minimum or local maximum) in the image (for example, an image that will have linearity in the y direction) are extracted, and within the vertically long range in the y direction (the center in the x direction is the extreme). There is a method of taking the variance in the x direction with respect to the coordinates of the value x the average value of the coordinates and the length in the x direction as the pitch of the array in the x direction. Also, the coordinates of the extreme values (local minimum or local maximum) in the image (for example, an image that will have linearity in the y direction) are extracted, and within the vertically long range in the y direction (the center in the x direction is the extreme). There is a method of taking an angle with a point near the top and bottom of each point with respect to a coordinate of (value × average value of coordinates and length in x direction as pitch of array in x direction).
[0067]
Specific configuration of polysilicon film evaluation system and processing details
(1) Next, a specific configuration example of the polysilicon film evaluation apparatus for evaluating the linearity and periodicity of the surface space structure of the polysilicon film as described above will be described.
[0068]
A polysilicon film evaluation apparatus is a bottom gate TFT manufacturing substrate (a substrate immediately after a polysilicon film is formed by excimer laser annealing on an amorphous silicon film) using a microscope device using an ultraviolet laser with a wavelength of 266 nm. ) And evaluates the state of the polysilicon film formed from the captured image.
[0069]
FIG. 9 shows a configuration diagram of the polysilicon film evaluation apparatus.
[0070]
A polysilicon film evaluation apparatus 20 shown in FIG. 9 includes a movable stage 21, an ultraviolet solid-state laser light source 22, a CCD camera 23, an optical fiber probe 24, a polarization beam splitter 25, an objective lens 26, and a quarter wavelength. A board 27, a control computer 28, and an image processing computer 29 are provided.
[0071]
The movable stage 21 is a stage for supporting the substrate 1 on which a polysilicon film to be inspected is formed. The movable stage 21 has a function of supporting the substrate 1 to be inspected and moving the substrate 1 to a predetermined inspection target position.
[0072]
Specifically, the movable stage 21 includes an X stage, a Y stage, a Z stage, a suction plate, and the like.
[0073]
The X stage and the Y stage are stages that move in the horizontal direction, and the substrate 1 to be inspected is moved in a direction orthogonal to each other between the X stage and the Y stage, and the substrate 1 to be inspected is subjected to predetermined inspection It tries to lead to the position. The Z stage is a stage that moves in the vertical direction, and is for adjusting the height of the stage. That is, the Z stage moves in the optical axis direction of the irradiated ultraviolet laser, in other words, in the direction perpendicular to the plane of the substrate 1. The moving direction of the Z stage is hereinafter referred to as the Z direction. The suction plate is for sucking and fixing the substrate 1 to be inspected.
[0074]
The ultraviolet solid state laser light source 22 is an ultraviolet light source having a wavelength of 266 nm, and for example, an Nd: YAG quadruple wave all solid state laser is used. As this ultraviolet laser light source, those having a wavelength of about 157 nm have been developed in recent years, and such a light source may be used as the light source.
[0075]
The CCD camera 23 is a camera that is highly sensitive to ultraviolet light, and includes a CCD image sensor as an image sensor inside, and images the surface of the substrate 1 with this CCD image sensor. The CCD camera 23 suppresses thermal noise, readout noise, circuit noise, and the like generated by a CCD image sensor or the like by cooling the main body.
[0076]
The optical fiber probe 24 is a waveguide of ultraviolet laser light, and guides the ultraviolet laser emitted from the ultraviolet solid laser light source 22 to the polarization beam splitter 25.
[0077]
The polarization beam splitter 25 reflects the ultraviolet laser light from the ultraviolet solid laser light source 22 and irradiates the substrate 1 on the movable stage 21 through the objective lens 26, and also reflects the reflected light reflected from the substrate 1. The light is transmitted and irradiated to the high sensitivity low noise camera 3. That is, the polarization beam splitter 25 is a laser beam separator for separating the optical path of the outgoing light optical system such as the ultraviolet solid laser light source 22 and the optical path of the reflected light optical system to the CCD camera 23.
[0078]
The objective lens 26 is an optical element for enlarging and detecting the reflected light from the substrate 1. This objective lens 26 has, for example, NA of 0.9 and aberration correction at a wavelength of 266 nm. The objective lens 26 is disposed between the polarization beam splitter 25 and the movable stage 21.
[0079]
The quarter wavelength plate 27 extracts the reflected light component from the ultraviolet laser beam. The ultraviolet light that has been circularly polarized by the quarter-wave plate 27 is reflected by the substrate 1 and is transmitted through the quarter-wave plate 27 again, whereby the direction of 90 ° linearly polarized light is rotated. For this reason, the return light is transmitted through the polarization beam splitter 25.
[0080]
The control computer 28 performs lighting control of the laser light from the ultraviolet solid state laser light source 22, control of the moving position of the movable stage 21, switching control of the objective lens 26, and the like.
[0081]
The image processing computer 29 captures an image of the substrate 1 captured by the CCD image sensor provided in the CCD camera 23, analyzes the image, and evaluates the state of the polysilicon film formed on the substrate 1.
[0082]
In the evaluation apparatus 20 configured as described above, the ultraviolet light laser emitted from the ultraviolet solid laser light source 22 is transmitted through the optical fiber probe 24, the polarization beam splitter 25, the objective lens 26, and the quarter wavelength plate 27. 1 is irradiated. Light incident as linearly polarized light becomes circularly polarized light by the quarter-wave plate 27 and enters the substrate 1. The phase of the reflected return light changes by 90 °, and when passing through the quarter-wave plate 27 again, the direction of linearly polarized light is rotated by 90 °. Therefore, the reflected return light passes through the polarization beam splitter 25 and enters the CCD camera 23. Then, the CCD camera 23 images the incident reflected light with a CCD image sensor, and supplies the surface image information of the polysilicon film obtained by the imaging to the image processing computer 29.
[0083]
Then, as will be described below, the image processing computer 29 evaluates the state of the polysilicon film based on the information of the surface image of the taken polysilicon film. Then, based on the evaluation result, a set value of energy density at the time of excimer laser annealing for generating the polysilicon film is obtained, or the polysilicon film formed on the substrate 1 is a good product or not. Determine if it is a non-defective product.
[0084]
(2) Next, the procedure for evaluating the state of the polysilicon film of the image processing computer 29 will be described. The image processing computer 29 obtains a value (hereinafter referred to as an AC value) obtained by quantifying the periodicity from the surface image of the polysilicon film using autocorrelation, and linearity and period of the surface space structure of the polysilicon film. The state of the polysilicon film is evaluated by evaluating the properties.
[0085]
As shown in the flowchart of FIG. 10, in the evaluation processing procedure, first, an image capturing process of the surface of the polysilicon film is performed (step S1). Subsequently, an autocorrelation function is calculated from the captured image (step S2). Subsequently, a plane perpendicular to the alignment direction including (0, 0) on the image coordinates is cut out (step S3). Subsequently, the peak value and the side peak value of the autocorrelation function in the cut surface are calculated, and the AC value is obtained by taking the ratio between the peak value and the side peak value (step S4). Subsequently, the polysilicon film is evaluated based on the AC value (step S5).
[0086]
Here, the autocorrelation function is a function as shown in the following equation.
[0087]
[Expression 1]

[0088]
This autocorrelation function R (τ) is a function indicating the correlation when a certain function f (x) is translated in the x direction by τ.
[0089]
In this polysilicon film evaluation apparatus 20, the autocorrelation function of the surface image of the polysilicon film is obtained using the following Wiener Hinting's theorem. Note that here, the specifically captured image information is “i”.
1 Perform two-dimensional Fourier transform of the captured image “i”.
: F = fourier (i)
2 The power spectrum “ps” is generated by squaring the Fourier series “f”.
: Ps = | f |²
3 The power spectrum “ps” is inverse Fourier transformed to generate a two-dimensional autocorrelation function “ac”.
: Ac = inversefourier (ps)
4 Take the absolute value of the autocorrelation function “ac” to obtain the real number “aca” of the autocorrelation function.
: Aca = | ac |
[0090]
When the autocorrelation function “aca” generated in this way is displayed, a function as shown in FIGS. 11 and 12 is obtained. FIG. 11 shows an image having a high autocorrelation, that is, an autocorrelation function with good periodicity and linearity of the surface space structure of the polysilicon film. On the other hand, FIG. 12 shows an image with low autocorrelation, that is, an autocorrelation function with poor periodicity and linearity of the surface space structure of the polysilicon film.
[0091]
From the autocorrelation image calculated using the Wiener Hinting's theorem, the polysilicon film evaluation apparatus 20 is further perpendicular to the alignment direction (that is, the direction having linearity) and has coordinates (0, 0) is cut out, and a function obtained when cutting out is obtained. Here, the plane including the coordinates (0, 0) on the screen is cut out in order to normalize a value from an autocorrelation function that varies depending on experimental parameters such as the amount of illumination light and CCD gain.
[0092]
The function obtained by cutting out in this way is a function corresponding to the autocorrelation function R (τ) in the direction perpendicular to the alignment direction described above.
[0093]
Here, steps S1 to S3 described above may be performed as shown in steps S11 to S14 of FIG. 13 below.
[0094]
Further, instead of such an evaluation procedure, the following evaluation may be performed.
[0095]
As shown in the flowchart of FIG. 13, in this evaluation processing procedure, first, an image capturing process of the surface of the polysilicon film is performed (step S11). Subsequently, one line of the captured image is cut out in a direction perpendicular to the laser beam traveling direction (direction with linearity: x direction) (direction with periodicity: y direction) (step S12). Subsequently, an autocorrelation function is calculated for this one line (step S13). Subsequently, if necessary, these operations are repeated several times to average each line (step S14).
[0096]
The autocorrelation function in this case is obtained as follows using the Wiener Hinchin's theorem. Here, the image information for one line that is specifically captured is set to “l”.
1 Perform Fourier transform for one line “l” of the captured image.
: Fl = fourier (l)
2 The power spectrum “psl” is generated by squaring the Fourier series “fl”.
: Psl = | fl |²
3 The power spectrum “psl” is inverse Fourier transformed to generate a two-dimensional autocorrelation function “acl”.
: Acl = inversesfourier (psl)
4 Take the absolute value of the autocorrelation function “acl” to obtain the real number “acal” of the autocorrelation function.
: Acal = | acl |
[0097]
When the autocorrelation function acal generated in this way is represented on a graph, it becomes a function as shown in FIGS. FIG. 14 is a function having a high autocorrelation, that is, an autocorrelation function having good periodicity and linearity of the surface space structure of the polysilicon film.
On the other hand, FIG. 15 shows a function with low autocorrelation, that is, an autocorrelation function with poor periodicity and linearity of the surface space structure of the polysilicon film.
[0098]
These one-line autocorrelation functions are taken for all lines of the captured image, and each autocorrelation function is averaged. This is a function corresponding to the autocorrelation function R (τ) in the direction perpendicular to the alignment direction (that is, the direction having linearity) described above.
[0099]
Next, the polysilicon film evaluation apparatus 20 obtains a maximum peak value and a side peak value from the obtained function. Then, the ratio of the maximum peak value to the side peak value is obtained, and this value is set as the AC value.
[0100]
Therefore, when the AC value is an image with high autocorrelation, that is, when the periodicity and linearity of the surface space structure of the polysilicon film are good, the difference between the maximum peak value and the side peak value is large, and the value is large. Become. On the other hand, in the case of an image with low autocorrelation, that is, when the periodicity and linearity of the surface space structure of the polysilicon film are poor, the difference between the maximum peak value and the side peak value becomes small and the value becomes small.
[0101]
As described above, the bottom gate TFT 1 captures the surface image of the polysilicon film, obtains the autocorrelation function of the captured image, and quantifies the linearity and periodicity of the surface space structure of the polysilicon film. .
[0102]
Specifically, an AC value for an example of the captured image is shown in FIG.
[0103]
As shown in FIG. 5, the polysilicon film subjected to the excimer laser annealing as described above has a plurality of linear protrusions formed on the surface of the film, and the linear protrusions are periodically formed. Since it is arranged, its state can be evaluated by the AC value. That is, since the spatial structure of the surface of the polysilicon film has a reflective grating shape, it can be evaluated by an AC value.
[0104]
(3) Next, the calculation range for obtaining the AC value will be described.
[0105]
As described above, the AC value is calculated by capturing a surface image of the polysilicon film and performing two-dimensional Fourier transform on the captured image range.
[0106]
Usually, the larger the calculation range (two-dimensional range of the surface image to be calculated), the more accurately the linearity and periodicity of the surface space structure of the polysilicon film. However, since the AC value is obtained by performing the two-dimensional Fourier transform as described above, the amount of calculation becomes enormous. Therefore, in order to perform processing at a higher speed, it is desirable to make the calculation range smaller.
[0107]
On the other hand, if the calculation range is too small, an error from the AC value of the entire surface increases, and the linearity and periodicity of the polysilicon film cannot be accurately evaluated.
[0108]
Thus, the calculation range for obtaining the AC value affects the accuracy and the processing speed. Therefore, in order to obtain the AC value more efficiently and with good reproducibility, the calculation range must be set optimally.
[0109]
Below, surface image of polysilicon film with poor periodicity of surface spatial structure, surface image of polysilicon film with moderate periodicity of surface spatial structure, surface of polysilicon film with good periodicity of surface spatial structure Each image is shown and the optimum calculation range of the AC value is described while considering these features. In the following description of the optimum calculation range, the direction parallel to the laser scanning direction during excimer laser annealing is referred to as the X direction, and the direction orthogonal to the laser scanning direction is referred to as the Y direction.
[0110]
FIG. 17A shows a surface image of a polysilicon film having a poor periodicity of the surface spatial structure as a whole, and FIG. 17B schematically shows this image. If the periodicity of the surface spatial structure is poor, the AC value should be small. However, even if the periodicity is poor as a whole, there is a feature that a portion with improved periodicity occurs in the X direction. The portion with improved periodicity is formed over a certain length in the X direction. Therefore, if the calculation range in the X direction is set to a narrower range of the portion with good periodicity, only the image of the portion with good periodicity is cut out and the AC value is calculated. There is a possibility. In other words, if calculation is performed with such a narrow range set, the calculated AC value increases despite the fact that the AC value should originally be small, and an accurate value can be obtained. It may not be possible. Therefore, when the calculation range for the X direction is calculated for the surface image of the polysilicon film having poor periodicity, the AC value is sufficiently large even if the portion where the periodicity is improved is included in the calculation range. Must be set to be low. For example, the calculation range of the AC value with respect to the X direction is set to be twice or more the length of the portion with good periodicity (the length of L1 in FIG. 2), and the influence of this portion with good periodicity is affected. It is better to halve to get the correct value.
[0111]
Subsequently, FIG. 18A shows a surface image of a polysilicon film having a medium periodicity of the spatial structure of the surface as a whole, and FIG. 18B schematically shows this image. When the periodicity of the spatial structure of the surface is medium, the AC value should be a medium value. However, the surface space structure of the polysilicon film having a medium periodicity as described above is characterized in that portions having good periodicity and portions having poor periodicity are alternately repeated. The portion where the periodicity is good and the portion where the periodicity is bad are formed over a certain length in the X direction. Therefore, if the calculation range in the X direction is set to a range narrower than the portion where the periodicity is good (or a range narrower than the portion where the periodicity is bad), There is a possibility that the calculation is performed by cutting out only the image (or only the image of the portion with poor periodicity). That is, if the calculation is performed with such a narrow range, the calculated AC value may become extremely large or extremely large even though the image should originally have a medium AC value. There is a possibility that it becomes small and cannot obtain an accurate value. Accordingly, the calculation range for the X direction is set to be equal to or longer than the length of one cycle (the length of L2 in FIG. 3) of the portion having good periodicity and the portion having bad periodicity in the X direction. It is better to halve the effects of repeated good and bad parts and get accurate values.
[0112]
FIG. 19A shows a surface image of a polysilicon film whose surface spatial structure has good periodicity as a whole, and FIG. 19B schematically shows this image. When the periodicity of the spatial structure of the surface is good, the AC value should naturally increase. However, even if the periodicity is good as a whole as described above, there is a feature that a portion where the periodicity is bad occurs in the X direction. The portion where the periodicity is deteriorated is formed over a certain length in the X direction. For this reason, if the calculation range in the X direction is set to a range narrower than the portion where the periodicity is poor, there is a possibility that the calculation is performed by cutting out only the image of the portion where the periodicity is bad. . In other words, if calculation is performed with such a narrow range set, the calculated AC value becomes small even though the image is supposed to have a large AC value, and an accurate value can be obtained. It may not be possible. Therefore, when the calculation range for the X direction is calculated for the surface image of the polysilicon film having good periodicity, the AC value is sufficient even if the calculation range includes a portion where the periodicity is deteriorated. Must be set to be low. For example, the calculation range of the AC value with respect to the X direction is set to be twice or more the length of the portion with the poor periodicity (the length of L3 in FIG. 4), and the influence of the portion with the poor periodicity is affected. It is better to halve to get the correct value.
[0113]
If the calculation range for the X direction is set so as to satisfy the above three conditions, an accurate AC value can be obtained. That is, the periodicity of the surface spatial structure is not less than twice the range of the good periodicity of the polysilicon film as a whole, and the periodicity of the surface of the polysilicon film is moderate as a whole. It must be at least one cycle of repeating the good part and the bad periodic part, and the periodicity of the surface spatial structure should be at least twice the range of the poor periodic part in the good polysilicon film as a whole. Therefore, an accurate AC value can be obtained.
[0114]
On the other hand, since the calculation range for the Y direction calculates periodicity with respect to the arrangement period of a plurality of linear protrusions formed when laser annealing is performed at an optimal energy density, In order to obtain the correlation, a range of two cycles or more is required as shown in FIG.
[0115]
Note that when the calculation range is expanded in order to increase accuracy, it is more likely that the calculation range is expanded in the X direction than the calculation range is expanded in the Y direction. This is because, if the calculation range is expanded in the X direction, the influence of the good periodicity generated in the poorly periodic polysilicon film, the good periodicity and the bad part of the moderately periodic polysilicon film This is because the influence of the repetition in the X direction and the influence of the portion with poor periodicity occurring in the polysilicon film with good periodicity are removed. Therefore, when the calculation range is set, it is preferable that the length in the X direction is a rectangular range that is equal to or longer than the length in the Y direction.
[0116]
Further, since the period of the arrangement of the plurality of linear protrusions generated when laser annealing is performed as described above is 0.3 μm, it is desirable that the calculation range in the Y direction be 0.6 μm or more. Further, from experience, in the case of a polysilicon film having a medium periodicity as a whole, the repetition period of the portion with good periodicity and the portion with poor periodicity with respect to the X direction was 5 μm. Is preferably 5 μm or more.
[0117]
(4) Next, control of the amount of ultraviolet light of the polysilicon film evaluation apparatus 20 for calculating the AC value with high reproducibility will be described.
[0118]
The polysilicon film evaluation apparatus 20 irradiates the surface of the polysilicon film with ultraviolet light, images the reflected light with a CCD, and calculates an AC value from the captured image.
[0119]
Here, since the AC value is calculated based on the captured image captured by the CCD, the value of the AC value may vary depending on the brightness of the captured image. For example, when the exposure amount to the CCD is insufficient or when the exposure amount is excessive, the contrast of the captured image itself decreases, and the uneven state cannot be recognized.
[0120]
FIG. 20 shows a graph showing the variation of the AC value with respect to the exposure amount of the ultraviolet light to the CCD. Exposure (μJ / cm²) Is the time when the shutter of the CCD is open (s) × the amount of irradiation light (mW / cm).²). Here, as shown in FIG. 20, the AC value is within a certain range (T in the figure).₁If the value is within the range indicated by (), the value is constant. That is, the amount of light incident on the CCD can be controlled so that the AC value can be obtained with good reproducibility by controlling the exposure amount to be within this certain range.
[0121]
Therefore, in the polysilicon film evaluation apparatus 20 according to the present embodiment, the amount of ultraviolet light emitted from the objective lens is detected by, for example, a photodetector, and the amount of light applied to the polysilicon film is detected. Then, the exposure amount is calculated from the irradiation light amount and the shutter speed of the CCD, and the incident light amount to the CCD is monitored. Then, while monitoring the amount of incident light, for example, the power of the ultraviolet laser, the shutter speed, and the like are feedback controlled to control the exposure amount to fall within the above-mentioned fixed range.
[0122]
FIG. 21 is a graph showing the variation of the AC value with respect to the brightness of the screen of the captured image. Here, the brightness of the screen of the captured image is represented by 256 gradations. Here, as shown in FIG. 21, the AC value is within a certain range (T₂If the value is within the range indicated by (), the value is constant. That is, by controlling the brightness of the captured image to be within this certain range, the amount of light incident on the CCD can be controlled so that the AC value can be obtained with good reproducibility. As described above, it is possible to control the amount of light incident on the CCD even if the brightness of the captured image is detected instead of detecting the exposure amount.
[0123]
Therefore, in the polysilicon film evaluation apparatus 20 of the present embodiment, the brightness of the screen of the captured image is calculated by a CPU or the like, and the incident light amount is monitored. Then, while monitoring the amount of incident light, for example, the power of the ultraviolet laser, the shutter speed, and the like are feedback controlled to control the exposure amount to fall within the above-mentioned fixed range.
[0124]
(5) Next, the relationship between the AC value obtained as a result of calculation as described above, the grain size of the polysilicon film, and the energy applied to the polysilicon film will be described.
[0125]
As shown in FIG. 22, the AC value is determined by the energy E given to the polysilicon film by excimer laser annealing.₅The value rises proportionally from when the_TThe value becomes the maximum. The AC value is the maximum energy E_TAt the peak value, then the value decreases proportionally, and some energy E_B2The reduction ends and the value becomes the minimum value. Thus, the AC value has a peak characteristic with respect to given energy.
[0126]
When such AC value peak characteristics are superimposed on the grain size change characteristics of the polysilicon film shown in FIG. 3, the result is as shown in FIG. As shown in FIG. 23, it can be seen that the maximum value of the graph showing the peak characteristic of the AC value falls within the energy range in which the grain size of the polysilicon film is appropriate. Furthermore, the energy E at which the AC value starts to rise proportionally₅However, it is lower than the allowable minimum energy L which is given to the polysilicon film and the grain size is appropriate. In addition, the energy E when the proportional decrease of the AC value stops and becomes the minimum value_B2However, the crystal grain size of the polysilicon film becomes higher than the allowable maximum energy H, which is the threshold energy for microcrystallization.
[0127]
Therefore, when evaluating whether the grain size of the polysilicon film is favorable from the AC value having such peak characteristics, is the AC value within the range indicated by the thick line in FIG. Judgment should be made.
[0128]
(6) When an AC value having such characteristics is evaluated to inspect whether the polysilicon film is a non-defective product, for example, the AC value of the substrate to be inspected is the allowable minimum energy L. Or AC required when the maximum allowable energy H is given_LThe higher one of the two values is used as a threshold value, and if it is larger than this threshold value, it can be inspected by determining that the product is non-defective.
[0129]
Further, when the AC value having such characteristics is evaluated and the energy density of the laser emitted from the excimer laser annealing apparatus is optimally set, for example, while changing the energy density of the excimer laser, a plurality of Laser annealing is performed on the substrate. Then, a characteristic diagram of the AC value corresponding to each energy density is drawn. Specifically, a characteristic chart as shown in FIG. 22 is drawn, and an optimum energy density may be obtained from this characteristic chart.
[0130]
(7) By the way, in the bottom gate type TFT as described above, since the gate electrode 3 is located in the lower layer of the polysilicon film 6, the energy diffusibility when the laser annealing is performed is on the glass substrate 2 ( The polysilicon film 6 on the gate electrode 3 is higher than the polysilicon film 6 on the source / drain region. Therefore, even if the energy density given from the excimer laser annealing apparatus is the same, it is given by the polysilicon film 6 on the gate electrode 3 and the polysilicon film 6 on the glass substrate 2 (on the source / drain region). The energy will be different and the grain size will be different on both sides.
[0131]
In general, when laser annealing is performed with an excimer laser annealing apparatus, the energy density is changed between a polysilicon film located on the gate electrode and a polysilicon film located on the glass substrate (on the source / drain region). Therefore, excimer laser annealing is uniformly performed with the same energy density setting.
[0132]
Therefore, in the bottom gate type TFT, the AC value characteristic with respect to the energy density of the excimer laser is as shown in FIG. 24, and the peak value is on the glass substrate (on the source / drain region) and on the gate electrode. It will be in a different position. Specifically, the AC value of the polysilicon film located on the glass substrate (on the source / drain region) reaches a peak value at a lower energy density than the polysilicon film located on the gate electrode. .
[0133]
Therefore, when the AC value is evaluated to inspect whether the polysilicon film is non-defective, and when the AC value is evaluated and the energy density emitted from the excimer laser annealing apparatus is optimally set. It is necessary to set such a value that the polysilicon films on both (on the glass substrate and on the gate electrode) are good.
[0134]
Next, FIG. 25 shows an example of specific experimental data of the AC value with respect to the energy density of the excimer laser for the polysilicon film of the bottom gate TFT. As shown in FIG. 25, the AC value has a different peak value between the gate electrode and the glass substrate. For example, in the characteristic diagram shown in FIG. 25, it can be seen that it is optimal to set the energy density in excimer laser annealing at 380 mJ.
[0135]
(8) As described above, when evaluating a polysilicon film formed on a bottom-gate TFT, it is easy and non-destructive by evaluating the linearity and / or periodicity of the spatial structure of the polysilicon film surface. In addition, polysilicon can be inspected, and the inspection process can be incorporated into the manufacturing process. In addition, since this linearity and / or periodicity is digitized, it is possible to perform numerical calculations without relying on visual inspections. Furthermore, since evaluation is performed by digitization, automatic inspection is possible, and objective inspection can be performed with high accuracy. In addition, the inspection result can be fed back to the annealing process to increase the yield of the thin film transistor to be manufactured.
[0136]
Although a microscope apparatus using an ultraviolet laser with a wavelength of 266 nm has been applied as an apparatus for imaging the polysilicon film, the linearity and / or periodicity of the surface space structure of the polysilicon film is evaluated. The device that captures the original image is not limited to such a device. For example, the linearity and / or periodicity of the surface space structure of the polysilicon film may be evaluated based on an image observed with an SEM. For example, as shown in FIG. 26, when the AC value is obtained based on an image captured by a microscope apparatus (DUV) using an ultraviolet laser and when the AC value is obtained based on an image captured by an SEM. Comparing the characteristics with each other, it can be seen that the SEM produces a clearer image, so that the AC value is relatively low, and the curves indicating the characteristics are substantially the same.
[0137]
Further, although an example using an autocorrelation function as a method for quantifying linearity and / or periodicity has been described in detail, the quantification method is not limited to an example using this autocorrelation function.
[0138]
Specific application examples in bottom gate TFT manufacturing process
Next, a specific application example in which the polysilicon film evaluation apparatus 20 is applied to a bottom gate type TFT manufacturing process will be described.
[0139]
First, the AC value obtained from the captured image of the polysilicon film of the bottom gate TFT as shown in FIG. 27 is evaluated, and the evaluation result is fed back to the excimer laser annealing apparatus. An application example (EQC: Equipment Quality Control) for optimally setting the emitted laser power will be described.
[0140]
As described above, the excimer laser annealing apparatus has a relatively large fluctuation in the actual output value of the laser power with respect to the set value of the laser power. The output laser power exhibits a Gaussian distribution characteristic and varies, and a certain amount of variation occurs with respect to a predetermined power setting value. On the other hand, in the case of a bottom gate type TFT, the manufacturing margin of energy given to the polysilicon film (energy range that becomes a defective product when energy outside this range is given) A relatively large value.
[0141]
Therefore, as shown in FIG. 28, the center position of the manufacturing margin of the polysilicon film is the optimum value of the laser power setting value. If the laser power is set to this optimum value, even if the laser power fluctuates, The energy given to the silicon film falls within the manufacturing margin, and a high yield can be obtained. However, as shown in FIG. 29, when the set value of the laser power is not set to the optimum value of the manufacturing margin, if the laser power fluctuates, the energy given to the polysilicon film may deviate from the manufacturing margin. Many yields are low.
[0142]
For this reason, in this application example, the laser power of the excimer laser annealing apparatus is set to the optimum value as follows using the peak characteristic of the AC value of the bottom gate TFT.
[0143]
First, in this application example, a plurality of substrates on which a polysilicon film is formed are manufactured. At this time, the laser power setting of the excimer laser annealing apparatus is changed for each substrate, and the AC value on the gate electrode and the glass substrate is obtained for each substrate.
[0144]
Then, a peak curve of the AC value as shown in FIGS. 30 and 31 can be drawn on the graph.
[0145]
By drawing such a peak curve of AC value, it is possible to obtain an allowable range of laser power (polysilicon film manufacturing margin) at which a good grain size can be obtained on both the gate electrode and the glass substrate. Specifically, the laser power at the lower limit of the manufacturing margin is the laser power corresponding to the minimum allowable energy (L) of the energy given to the polysilicon film on the gate electrode, specifically, FIG. 30 and FIG. The laser power (MO (L)) at the left end of the portion drawn with a thick line of the AC value on the gate electrode shown. The upper limit laser power of the manufacturing margin is the laser power corresponding to the maximum allowable energy (H) of the energy given to the polysilicon film on the glass substrate, specifically, the glass shown in FIGS. This is the laser power (G (H)) at the right end of the portion drawn with a thick line of the AC value on the substrate.
[0146]
Then, an intermediate value of the manufacturing margin thus obtained is obtained, and the laser power at the intermediate value is set as an optimum value.
[0147]
As described above, by obtaining the AC value, obtaining the manufacturing margin, and setting this manufacturing margin as the optimum value, the yield of the bottom gate TFT can be increased.
[0148]
【The invention's effect】
In the polysilicon evaluation method and the thin film transistor manufacturing system and method according to the present invention, the surface of a polysilicon film formed by laser annealing is imaged, the autocorrelation of the captured image is calculated, and the linearity of the spatial structure And / or evaluate periodicity. At this time, the calculation range of the autocorrelation with respect to the direction orthogonal to the laser scanning direction in the laser annealing process is set to at least two periods when the periodicity of the surface space structure of the polysilicon film appears.
[0149]
In the present invention, the calculation range of the autocorrelation with respect to the direction parallel to the laser scanning direction in the laser annealing process is made longer than the orthogonal direction.
[0150]
As a result, in the present invention, when the state of the film is evaluated by quantifying the spatial structure of the surface of the polysilicon film formed by the low-temperature polycrystallization process using the autocorrelation function, the calculation range is optimally set. In addition, the state of the film can be evaluated with good reproducibility and in a short time.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a schematic cross-sectional configuration of a bottom gate TFT.
FIG. 2 is a diagram for explaining a cross-sectional structure of a bottom gate TFT after a polysilicon film is formed.
FIG. 3 is a diagram for explaining the relationship between the grain size of a polysilicon film and the energy given by excimer laser annealing.
FIG. 4 shows an image of the surface of a polysilicon film when excimer laser annealing is performed with the energy density of the output laser as an optimum value, and an image of the film surface of the polysilicon film when the power is less than the optimum value. FIG. 6 is a diagram for explaining an image of the surface of a polysilicon film when the power is greater than the optimum value.
FIG. 5A is a diagram for explaining a cross-sectional structure of a TFT substrate before excimer laser annealing. (B) is a figure for demonstrating the cross-section of a TFT substrate at the time of performing excimer laser annealing with a favorable energy density. (C) is a figure for demonstrating the cross-sectional structure of a TFT substrate at the time of performing excimer laser annealing with the energy density which is not favorable.
FIG. 6 is a diagram for explaining a manufacturing process of a TFT in which a surface oxidation step is provided in a step before excimer laser annealing.
FIG. 7 is a diagram schematically showing a captured image of a polysilicon film having linearity and periodicity.
FIG. 8 is a diagram schematically showing a captured image of a polysilicon film having no linearity and periodicity.
FIG. 9 is a configuration diagram of a polysilicon film evaluation apparatus according to an embodiment of the present invention.
FIG. 10 is a flowchart for explaining a procedure for evaluating a polysilicon film;
FIG. 11 is a diagram for explaining an autocorrelation function when periodicity is high.
FIG. 12 is a diagram for explaining an autocorrelation function when periodicity is low.
FIG. 13 is a flowchart for explaining another evaluation procedure of a polysilicon film.
FIG. 14 is a diagram for explaining an autocorrelation image when the periodicity when evaluated by the other evaluation procedure is high.
FIG. 15 is a diagram for explaining an autocorrelation image when periodicity is low when evaluation is performed by the other evaluation procedure.
FIG. 16 is a diagram for explaining the characteristics of the obtained AC value for a specific captured image.
FIG. 17 is a diagram showing a surface image of a polysilicon film with poor periodicity of the spatial structure of the surface, and a diagram showing a schematic structure thereof.
FIG. 18 is a diagram showing a surface image of a polysilicon film having a moderate periodicity of the surface spatial structure, and a diagram showing a schematic configuration thereof.
FIG. 19 is a diagram showing a surface image of a polysilicon film having a good periodicity of the spatial structure of the surface, and a diagram showing a schematic structure thereof.
FIG. 20 is a graph showing a change in AC value with respect to the exposure amount of the ultraviolet light to the CCD.
FIG. 21 is a graph showing a change in AC value with respect to screen brightness of a captured image;
FIG. 22 is a diagram for explaining a characteristic of an autocorrelation value with respect to energy applied to a polysilicon film.
FIG. 23 is a diagram for explaining the characteristics of the AC value and the grain size with respect to the energy given to the polysilicon film.
FIG. 24 is a diagram for explaining a characteristic of an AC value with respect to an energy density of an excimer laser in a bottom gate TFT.
FIG. 25 is a diagram for explaining an example of specific experimental data of the AC value with respect to the energy density of the excimer laser for the polysilicon film of the bottom gate TFT.
FIG. 26 compares characteristics when an AC value is obtained based on an image captured by a microscope apparatus (DUV) using an ultraviolet laser and characteristics obtained when an AC value is obtained based on an image captured by an SEM. FIG.
FIG. 27 is a diagram for explaining a configuration of a specific application example (EQC) in which a polysilicon film evaluation apparatus is applied to a manufacturing process of a bottom gate TFT.
FIG. 28 is a view for explaining the relationship between the manufacturing margin of energy given to a polysilicon film and the fluctuation of the energy density of the excimer laser (when the energy density is set optimally).
FIG. 29 is a diagram for explaining the relationship between the manufacturing margin of energy given to a polysilicon film and the fluctuation of the energy density of the excimer laser (when the laser energy density is not optimally set).
30 is a diagram illustrating an example of a relationship between a manufacturing margin of a bottom gate type TFT and an energy density of an excimer laser, and a diagram for explaining a method for obtaining an optimum value of the energy density from this example. FIG.
FIG. 31 shows another example of the relationship between the manufacturing margin of the bottom gate type TFT and the energy density of the excimer laser, and is a diagram for explaining a method for obtaining the optimum value of the energy density from the other example. is there.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Bottom gate type TFT, 2 Glass substrate, 3 Gate electrode, 1st gate insulating film, 5 2nd gate insulating film, 6 Polysilicon film, 20 Polysilicon film evaluation apparatus, 30 Excimer laser annealing apparatus

Claims (12)

In a polysilicon evaluation apparatus that evaluates a polysilicon film formed by forming an amorphous silicon film on a substrate and subjecting the formed amorphous silicon film to an excimer laser annealing process,
An optical system for irradiating the polysilicon film with the focused ultraviolet laser and detecting the reflected light;
A high-sensitivity camera for ultraviolet light that images the reflected light detected by the optical system;
Captured by the high sensitivity and camera with respect to the ultraviolet light, the captured image of the polysilicon film, calculates the ratio of the maximum peak value to the side peak value of the autocorrelation function (AC value), the poly An evaluation means for evaluating the periodicity of the surface space structure of the silicon film, and evaluating the crystalline state of the polysilicon film based on the periodicity ;
The evaluation means has a calculation range for obtaining a ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function with respect to the direction orthogonal to the laser scanning direction in the excimer laser annealing process, at least for the polysilicon film. A polysilicon evaluation apparatus characterized by having two or more periods in a portion where periodicity of the surface space structure appears. The evaluation means determines the calculation range when obtaining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction parallel to the laser scanning direction in the excimer laser annealing process from the orthogonal direction. The polysilicon evaluation apparatus according to claim 1, wherein the polysilicon evaluation apparatus is lengthened. In the excimer laser annealing process, the evaluation means sets the calculation range when determining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function in the direction parallel to the laser scanning direction to 5 μm or more. 2. The polysilicon evaluation apparatus according to claim 1, wherein: In a polysilicon evaluation method for evaluating a polysilicon film formed by forming an amorphous silicon film on a substrate and subjecting the formed amorphous silicon film to an excimer laser annealing process,
The polysilicon film is irradiated with a focused ultraviolet laser , the reflected light is detected, and the detected reflected light is imaged by a camera that is highly sensitive to ultraviolet light ,
Captured by the high sensitivity and camera with respect to the ultraviolet light, the captured image of the polysilicon film, calculates the ratio of the maximum peak value to the side peak value of the autocorrelation function (AC value), the poly Evaluate the periodicity of the surface space structure of the silicon film, evaluate the crystalline state of the polysilicon film based on this periodicity,
When evaluating the periodicity of the surface spatial structure of the polysilicon film, the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction orthogonal to the laser scanning direction in the excimer laser annealing treatment is calculated. The method for evaluating polysilicon is characterized in that a calculation range when obtaining is at least two periods in a portion where the periodicity of the surface space structure of the polysilicon film appears. In the excimer laser annealing process, the calculation range for obtaining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction parallel to the laser scanning direction is made longer than the orthogonal direction. The method for evaluating polysilicon according to claim 4. The calculation range for obtaining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction parallel to the laser scanning direction in the excimer laser annealing process is 5 μm or more. 5. The polysilicon evaluation method according to 4. In a thin film transistor manufacturing system for manufacturing a thin film transistor,
A film forming apparatus for forming an amorphous silicon film on a substrate;
An excimer laser annealing apparatus for generating a polysilicon film serving as a channel layer by performing an excimer laser annealing process on the amorphous silicon film;
The polysilicon film is irradiated with a focused ultraviolet laser and the reflected light is detected, and the sensitivity to the ultraviolet light for imaging the reflected light detected by the optical system is increased. a camera, captured by a high sensitivity and camera with respect to the ultraviolet light, the captured image of the polysilicon film, calculates the ratio of the maximum peak value to the side peak value of the autocorrelation function (AC value) , to evaluate the periodicity of the surface spatial structure of the polysilicon film, and a polysilicon inspection apparatus for evaluating the crystal state of the polysilicon film based on the periodicity,
The evaluation means of the polysilicon inspection apparatus has a calculation range when obtaining a ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function with respect to the direction orthogonal to the laser scanning direction in the excimer laser annealing process. A thin film transistor manufacturing system characterized in that the period is at least two in a portion where periodicity of the surface space structure of the polysilicon film appears. The evaluation means of the polysilicon inspection apparatus calculates a calculation range when obtaining a ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function with respect to a direction parallel to the laser scanning direction in the excimer laser annealing process. The thin film transistor manufacturing system according to claim 7, wherein the thin film transistor is longer than the orthogonal direction. The evaluation means of the polysilicon inspection apparatus calculates a calculation range when obtaining a ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function with respect to a direction parallel to the laser scanning direction in the excimer laser annealing process. The thin film transistor manufacturing system according to claim 7, wherein the thickness is 5 μm or more. In a thin film transistor manufacturing method for manufacturing a thin film transistor, the method includes a polysilicon film generation step of generating a polysilicon film to be a channel layer by performing an excimer laser annealing process on an amorphous silicon film with an excimer laser annealing apparatus.
An amorphous silicon film is formed on the substrate,
Excimer laser annealing is performed on the amorphous silicon film to form a polysilicon film,
The polysilicon film is irradiated with a focused ultraviolet laser , the reflected light is detected, and the detected reflected light is imaged with a camera made highly sensitive to ultraviolet light ,
Captured by the high sensitivity and camera with respect to the ultraviolet light, the captured image of the polysilicon film, calculates the ratio of the maximum peak value to the side peak value of the autocorrelation function (AC value), the poly Evaluate the periodicity of the surface space structure of the silicon film, evaluate the crystalline state of the polysilicon film based on this periodicity,
When evaluating the periodicity of the surface spatial structure of the polysilicon film, the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction orthogonal to the laser scanning direction in the excimer laser annealing treatment is calculated. A method for producing a thin film transistor, characterized in that a calculation range when obtaining is at least two periods in a portion where the periodicity of the surface space structure of the polysilicon film appears. In the excimer laser annealing process, the calculation range for obtaining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction parallel to the laser scanning direction is made longer than the orthogonal direction. The method for producing a thin film transistor according to claim 10. The calculation range for obtaining the ratio (AC value) of the maximum peak value to the side peak value of the autocorrelation function for the direction parallel to the laser scanning direction in the excimer laser annealing process is 5 μm or more. 10. The method for producing a thin film transistor according to 10.

Images