JP3779352B2 - Infrared microspectroscopic analysis method and apparatus - Google Patents

Description

【００３４】
実施例のように、結晶をプローブとして用いるので、幾つかの条件を満たす必要がある。条件として、▲１▼中赤外域でのスペクトル測定に必要な波長域(約２〜１５μm)で透過率が高いこと、▲２▼結晶上に作製する開口の形状やサイズの確認を容易に行えるように可視光を透過すること、▲３▼水分を含む試料の測定、長時間の使用や、取り扱い易さを考慮して、潮解性がないこと、▲４▼結晶上での開口作製などにおいては結晶の加工等が必要になるため、加工、研磨が行いやすいこと、▲５▼加工時などの取り扱いを考慮して、可能な限り毒性が少ないこと、が挙げられる。
【００３５】
そこで、各結晶の性質を検討すると、ＺｎＳｅは、可視から赤外間での広い波長域で透過率が高く潮解性もない。ただし、医薬用外毒物に指定されているため、取り扱いに注意を要する。ＺｎＳは、ＺｎＳｅと同様の性質をもつが、透過波長域はＺｎＳｅに比べやや狭い。また医薬用外毒物に指定されているため、取り扱いに注意を要する。ＫＢｒをはじめＫＣｌ，ＮａＣｌなどのハロゲン化物は、可視領域から赤外領域の広い領域にわたって透過率が高い。潮解性がある。また、加工が困難である。ＭｇＦ₂をはじめＣａＦ₂，ＢａＦ₂などのフッ化物は、潮解性はほとんど無く、可視から赤外まで透過率はよい。ただし、１０μm以上の波長域では透過率が低くなり、この領域でのスペクトル測定は困難となる。Ｇｅ，Ｓｉの半導体は、潮解性がなく、赤外領域の長波長側まで透過率が高いが、可視光は透過しないため、結晶上に作製するスリットなどを可視光のもとで確認することができない。Ａｌ₂Ｏ₃，ＭｇＯなどの酸化物は、潮解性はなく、可視光に対する透過率も高い。ただし、長波長域での透過率は低く、また結晶が硬いため加工が困難である。
【００３６】
したがって、表１に掲げた赤外透過結晶にはどれも一長一短があり、上記の条件をすべて満足する結晶はない。ただし、▲５▼の毒性については、取り扱い時に十分注意を払えば問題はない。そこで、ＺｎＳｅかＺｎＳが最もよく条件を満たすと考え、さらに、ＺｎＳｅとＺｎＳとでは、ＺｎＳｅの方が透過波長域が広く、毒性もＺｎＳよりは低いので、実施例ではプローブ材料として最適のものとして、ＺｎＳｅを選択した。図７にＺｎＳｅ結晶の透過率スペクトルを示す。
ＺｎＳｅ結晶の円錐面の皮膜用コーティング金属には、金が用いられている。金の可視・赤外域における複素屈折率、および表皮深さは表２に示す通りである。表皮深さとは、光が金属中を進行するとき、その強度が１／ｅ(ｅは自然対数の
底)になる距離で定義されるものである。
【００３７】
【表２】

【００３８】
開口を作製する金属膜は、開口以外の場所では十分に光を遮断する必要がある。金属の膜厚を十分厚くすれば光は遮断されるが、その一方で、膜厚が大きいと、開口を通る光が開口内で散乱され、開口を透過する光の量が減少してしまうと考えられる。よって、金の膜厚は、光を十分に遮断できる厚さで、可能な限り薄いものが適当である。
【００３９】
表２から分かるように、可視域と赤外域での金の表皮深さは、若干可視域の方が深い程度である。したがって、コーティングした金の膜厚で赤外域の光を遮断できるかどうかは、可視域において、光が漏れ出してこないことを確認することで判断できる。金薄膜に可視光を照射し、透過光を確認したところ、金は約３００ｎｍ程度の膜厚で可視光を遮断することができることが分かった。そこで、結晶表面にスパッタ装置により金を約３００ｎｍコーティングした。金を結晶にコーティングするには、あらかじめ、結晶表面が十分に滑らかである必要がある。コーティングする膜厚が３００ｎｍ程度であるから、局所的にこれより大きな凹凸があると、そこから光が漏れ出ることがあるからである。したがって、金をコーティングする前に、結晶表面を研磨シートにより十分滑らかになるまで研磨した。研磨シートには、粒径０.３μmのラッピング・フィルムシートを用いた。
【００４０】
次は、スリット開口の作製である。例えば、スリットと同じ形状の遮蔽を結晶上に置いて、金属コーティングを行う方法が考えられるが、目的とするスリットは、スリット幅が約１〜２μmであるため、このような形状の遮蔽を作製したり、これを結晶上の先端部に配置したりすることは事実上不可能である。基板上に金属をコーティングし、スリット形状部分の金属をエッチングにより剥離する方法も考えられるが、ＺｎＳｅは酸に溶けるため、ＺｎＳｅ上でエッチングをおこなうことはできない。
【００４１】
したがって、スリットの作製には図８に示すような方法を用いた。すなわち、金属の針２６と、金属コーティングした結晶２０'とを衝突させるという方法である。金属の針２６の先端部は、拡大図示のように、開けるスリットと同一の幅をもつクサビ形状をしている。そして結晶２０'の位置の制御を、３軸ステージ２７およびピエゾ素子２８，２９により行い、固定した金属針２６に衝突させる。結晶の位置を制御するピエゾ素子は２方向Ｘ，Ｚに取り付けられている。つまり、結晶が針先端に近づく方向(図８のＺ軸方向)と、結晶と針とが横にずれる方向(図８のＸ軸方向)である。実際に金属針と結晶とを衝突させて開口を開けようとしたところ、Ｚ軸方向にのみ結晶を動かして針と衝突させただけでは、結晶上の金属に傷はつくものの、光を透過するような開口を作製することは困難であった。そこで、結晶と金属針２６とが接触した時点で結晶をＺ軸方向に動かしながら、ピエゾ素子２８により同時にＸ軸方向にも動かしたところ、スリット状の開口を作製することができた。
金属針と結晶との接触の様子、および、微小スリットの結晶表面上における位置を調節するため、金属針の先端と結晶表面の先端部分の接触箇所を、図８のＸ軸方向とＹ軸方向の２方向から、顕微鏡を用いて観測しながら、スリット作製を行った。また、金属針と結晶とが接触してからの、結晶のＺ軸方向への移動量を調節するには、金属針と結晶とが接触する位置を正確に把握する必要がある。しかし、２方向からの顕微鏡観察のみでは接触した瞬間を正確に見極めることが難しい。したがって、金属針と結晶表面の金属との間に電流を流し、電流計で、電流の流れ始める位置を確認することで、金属針と結晶との接触した厳密な位置を確認した。
【００４２】
なお、針２６の材質としては、結晶２０'との衝突で針先の形状が崩れることのないよう硬度の高い、ステンレスを用いた。硬度の高い材質として、チタンも用いたが、もろいため、加工が困難であった。
直径約１ｍｍのステンレスの金属棒の先端を、研磨シートを用いて研磨し、目的とするスリットと同一の幅をもつ、図８のようなクサビ形の形状に加工した。用いた研磨シートは粒径０.３μmのラッピング・フィルムシートである。針の研磨は、針の先端を顕微鏡により観察しながら行った。
【００４３】
以上の方法で作製したスリット状微小開口２１は、プローブ２０の球面側から円錐面の裏側を照明し、円錐面の表面上を顕微鏡により観測して、形状、長さを確認することができる。一例では、スリット幅２μm、スリット長約７０μmと確認された。また、数例を作製し、後述の具体的な実施例で適用した。
【００４４】
【より具体的な実施例】
図５の装置システムにより、図９に示すような、基盤５１上にコーティングした金属膜５２を、スリット状開口２１に直交するように面内で１次元走査し、金属膜５２のエッジ部分での赤外透過スペクトルの変化を測定した。
試料５０の被検体としての金属膜５２は、膜厚が約１５ｎｍの金(Ａｕ)であり、基盤５１は赤外光に対し透過率の高いＢａＦ₂を用い厚さは約１ｍｍである。スリット状開口２１のスリット幅は約２μm、スリット長さは約５０μmである。そして、この実施例の場合は、各測定位置で、微小開口プローブ２０のスリット状開口２１と試料５０とを接触させて測定を行っている。
【００４５】
図１０は、試料５０を矢印の方向に走査したときに得られた、金属膜５２のエッジ付近での透過率スペクトルの変化を示したものである。図１０の(a)、(b)、(c)、(d)は、それぞれ２.５μm離れた４点で測定されたスペクトルである。なお、各スペクトル測定の積算回数は１００回であった。図１０より、(a)−(b)、(c)−(d)間では測定波数全域にわたって、透過率の低下は約１０％であるが、(b)−(c)間では約５０％も低下しているのが分かる。つまり、(b)−(c)間で金属膜５２のエッジを検出していると考えられる。なお、各スペクトルの２３５０ｃｍ^-1付近で透過率が大きく変動しているのはＣＯ₂の吸収によるものである。
【００４６】
次に、(１)４０００ｃｍ^-1(２.５μm)、(２)２５００ｃｍ^-1(４μm)、(３)１６６６ｃｍ^-1(６μm)、(４)１２５０ｃｍ^-1(８μm)、(５)１０００ｃｍ^-1(１０μm)の各波数(波長)における、透過率の変化を図１１、図１２に示す。両図の(ｎ)−ａはプローブ２０を用いて測定したとき、(ｎ)−ｂはこれを用いなかったときの結果である。ｂの測定時の視野絞りは約８μm×８μmであった。ａ、ｂそれぞれの測定における、各波数ごとの分解能を検討するために、測定された透過率のうちの最大値と最小値との差Ｄに対して、０.７５Ｄから０.２５Ｄへ変化するのに要する距離を、この例における空間分解能Ｒと規定する。図１１、図１２それぞれの０.７５Ｄと０.２５Ｄとを表す直線を合わせ示している。各グラフから読み取ったＲの値を表３にまとめて示す。
【００４７】
【表３】

【００４８】
表３から、プローブを用いて測定したときの分解能Ｒは、各波長に対して約２〜３μm前後であるのに対して、プローブなしで測定したときのＲは、各波長によって約７μmから約１５μｍまで大きく異なっている。プローブがない場合の空間分解能は、視野絞りの大きさに加え、波長とカセグレン対物鏡の開口数ＮＡによって決まる像の拡がりによって決定されるので、波長が長くなればなるほどＲもそれに応じて大きくなる。一方、スリット・プローブを用いたニアフィールド測定では、(１)から(５)の各波長でＲがほぼ一定の値をとり、空間分解能が波長に依存せず、スリット幅によって決まることが分かる。
【００４９】
この実施例で用いたカセグレン対物鏡の開口数ＮＡは０.６２である。したがって、レーリーの式(Δｒ＝０.６１λ／ＮＡ)より得られる、回折限界は使用する波長とほぼ同程度である。表３の結果を見ると、スリットを用いた場合は、いずれの波長においても回折限界を超えた空間分解能を実現していることが分かる。
【００５０】
もう一つの実施例は、高分子フィルムの赤外吸収スペクトルの測定である。図１３に示すように、厚さ約１ｍｍのＢａＦ₂基板５３上に密着固定したポリイミド系耐熱高分子フィルム５４をスリット状微小開口プローブ２０のスリット２１のスリット幅の方向に１次元走査して、赤外吸収スペクトルを求めたものである。走査方向は図中の矢印方向であり、各スペクトルの測定の積算回数は１００回である。なお、適用した微小開口プローブ２０のスリット状開口２１のスリット幅は約２μmで、スリット長は約３０μmである。また、ポリイミド系耐熱高分子フィルム５４の厚さは約５０μmである。この試料５４の、通常の赤外分光測定で得られる透過率スペクトルと吸光度スペクトルをそれぞれ図１４(a)，(b)に示す。
尚、この実施例では、スリット状開口２１を試料５４には接触させることなく、走査している。試料５４とスリット状微小開口２１間の距離は、図５に示される、ＣＣＤカメラ１０２を備える観察手段１００によって観察したところ、約１.５μmであった。
【００５１】
図１５の(a)、(b)、(c)は、試料５４のエッジ付近において、２μm間隔で測定した透過率スペクトルを示している。(a)から(c)の４μm間で、試料５４の吸収ピークが消えるのが確認できる。
【００５２】
図１４(b)に示した試料５４の吸収ピークのうち、矢印で示した、(１)２９６４ｃｍ^-1(３.３７μm)、(２)１７７８ｃｍ^-1(５.６２μm)、(３)１７２２ｃｍ^-1(５.８１μm)、(４)１６００ｃｍ^-1(６.２５μm)、(５)１５０２ｃｍ^-1(６.６６μm)、(６)１３８２ｃｍ^-1(７.２４μm)の６つの波数(波長)における透過率の変化を図１６、図１７に示す。図中の(ｎ)−(a)はプローブ２０を用いて測定した場合、(ｎ)−(b)はプローブを用いずに測定した場合である。なお、(b)においては、約５μm×５μmの視野絞りを用いて測定されている。
【００５３】
図１６、図１７の各グラフにおける縦軸は、各波数における透過率を、試料５４の吸収のない２１６０ｃｍ^-1の透過率で割った値である。割り算を行った理由は次の通りである。すなわち、測定中、試料のエッジ付近では信号全体の強度が減少する現象が見られた。これは、スリットの直前に試料が存在するときには試料からの透過光と試料表面からの散乱光がスリットに入射していたのに対して、試料のエッジ部分がスリットの真下を通過し、試料がスリット直前に存在していないときは、スリットから約５０μm(試料の厚さ)離れた基板を透過した光だけがスリットに入射し、基板表面からの散乱光はスリットに到達しないためであると考えられ、このため、試料による吸収がない波数での透過率を参照(割り算)することにより、スペクトル変化のうち試料の吸収のみの変化をみることができると考えられるからである。
【００５４】
図１６、図１７により明らかなように、プローブ２０を用いて測定した場合、各波数の前記の規定によるＲの値はおよそ３〜４μmで一定であることが分かる。また、(b)の場合は、Ｒは大きくなる傾向が現れている。なお、(b)では、Ｓ／Ｎ比が低く、Ｒの誤差も大きいと理解される。この実施例においても、先に示したように、プローブなしの計測では、波長が長くなるほど、空間分解能が低下するのに対して、スリット・プローブを用いた計測では、空間分解能は波長によらず、スリット幅で決まることが明白に確証されている。
【００５５】
なお、プローブを用いた(a)の例で、スリット幅が２μmで、Ｒが３〜４μm程度であるのに対して、先の実施例では、同じ幅のスリットを使っているにもかかわらず、Ｒが２.５μm程度戸、値が異なるのは、前者の場合よりも後者の場合の方が、プローブと試料との距離大きかったためであると考えられる。また、後者の実施例で用いた試料の厚さが約５０μmであるため、前者の実施例に比べ、エッジ部分の形状が鈍っているものと考えられる。
【００５６】
上記の実施例により、測定に用いる光の波長よりも短い幅を有するスリット状開口のプローブを用いることによって回折限界を超えた空間分解能で赤外分光スペクトルを測定できることが分かる。また、通常の赤外顕微分光測定では、空間分解能は回折限界で制限され、その制限は波長に依存するため、測定波数が小さいほど空間分解能が低下するという現象が現れるが、スリット状開口プローブを用いたニアフィールド顕微分光測定では空間分解能が波長によって制限されるということがなく、基本的には、スリット状開口のスリット幅の大きさによってのみ決定されるというものである。実施例によれば、波長域２.５〜１０μmにわたって、約２.５μmの空間分解能が実現されている。
【００５７】
したがって、実施例のような赤外ニアフィールド顕微分光システムを用いれば、有機多層膜試料など、試料の１次元方向の物質分布や微細構造を、回折限界を超える空間分解能で分析できる。
【００５８】
実施例のプローブのスリット幅は約２μmであったが、このスリット幅をより小さくすることで、さらに高い空間分解能が実現可能であり、サブミクロンのオーダも事実上、実現が可能である。尤も、空間分解能を向上させるためにスリット幅をより狭くすると、信号光は微弱化するが、▲１▼スリット長を長くする、▲２▼光源強度や検出器感度を向上させる、▲３▼スペクトル測定時の測定・積算回数を増やす、▲４▼さらに高い開口数ＮＡのカセグレン対物鏡を用いる、▲５▼ノイズ光を抑圧する特段の工夫をする、及び▲６▼スリット状開口の直線性を厳格にするとともに開口エッジの凹凸をなくす、等の単独あるいは組合わせによる工夫により、解決が可能である。
【００５９】
また、スリット状微小開口をプローブとして用いたとき、回折限界を超える空間分解能が得られるのは、スリット幅の方向であり、スリットの長さ方向に高い分解能は得られない。したがって、多層フィルムの断面などの試料を測定する場合はスリット状微小開口を幅方向に１次元走査することで試料の分布(組成等に基づく境界を超えて、Ａ，Ｂ，Ｃ，Ｄ……と識別できる)を測定できるが、たとえば、試料内に異なる種類の物質が２次元的に点在しているいるような場合は、スリット状微小開口を１次元方向に走査するだけでは、その分布を観測することができない。
【００６０】
スリット状微小開口を用いながら、試料の２次元分布を計測する方法として、スリット状微小開口を一方向にだけでなく、試料面内で様々な方向に走査することが考えられる。
図１８に図解されるように、試料中に吸収等の２次元分布をもった物体５５が存在する試料を測定するような場合を考えると、スリット２１を１次元方向にのみ走査しただけでは、スリットの長さ方向における試料の情報は得られない。そこで、スリット２１を回転させて同様にスリット幅の方向に走査を行う。この操作により、回転した後のスリット幅の方向における試料の情報が得られる。スリットの回転角φを一定で細かく割り出し、φ₁，φ₂，φ₃，φ₄，……の多方向においてスリットの走査を行えば、それだけ試料の２次元の情報を多く得ることができる。ある波長λ₀における２次元の透過率の分布は、各回転角φ_nにおける１次元の透過率分布を再構成することにより求めることができる。つまり、２次元ＣＴの原理を用いるもので、１次元投影データを測定し、そこからもとの物体の２次元断面像を再構成する方法(ラドン変換とラドン逆変換と呼ばれる手法)である。このような、試料への多方向からの観測結果から試料の構造を再構成する計測法として、Ｘ線ＣＴの技術が知られている。Ｘ線ＣＴで用いられているバックプロジェクションの手法をそのまま、スリット状微小開口による赤外透過率分布測定に適用することが可能である。
【００６１】
まず、図１８に示すように、▲１▼スリットを角度φだけ回転させて、Ｘ_φ方向に関する吸収強度分布を得る。これは、投影とも、プロジェクションともいわれ、数学的にはラドン変換の操作である。次に、▲２▼φを変化させて、多数の方向(φ_n：ｎ＝１，２，３，４，……)におけるプロジェクションを得る。▲３▼各プロジェクションにおける、Ｘ_φ方向の吸収分布を、Ｙ_φ方向に沿って並べる。これは、バックプロジェクションとよばれる操作である。そして、これを各φについて行い、測定領域全体においてその総和をとる。なお、バックプロジェクションから得られた再構成像(例えば、吸収分布像)は、試料の構造のうち、空間周波数の低い構造が強調されたものとなるため、一般には、上記の▲２▼の操作の後、各プロジェクションに対して、低周波を抑圧するためのフィルタリング処理を施してから、▲３▼の操作を行う。これをフィルタード・バックプロジェクションと称して、Ｘ線ＣＴでも同様に行われている。
【００６２】
尚、上述の実施例においては、赤外分光器９０はフーリエ変換型の分光器としたが、もちろん分散型の赤外分光器を適用することができる。また、試料ないし被検体５０を載置するステージ５０は、これに駆動機構が付帯しステージ５０が走査される構成であるが、ステージは固定でスリット状微小開口２１がステージ上方を走査する構成、すなわち微小開口プローブ２０を駆動し走査する構成とすることも可能である。同様に、ステージ５０は回転可能で所定角度で割出し制御されるようにも構成されているが、ステージは固定で微小開口プローブ２０を所定角度だけ割出しできるようにこの微小開口プローブを回転可能に支持するような構成としてもよい。
【００６３】
【発明の効果】
以上のように、本発明によれば、被検体の表面近傍にエバネッセント波の場を形成し、このエバネッセント波の場の中に幅が微小なスリット状開口を臨ませ、このスリット状開口を経由する光を検出するようにしたので、信号光の強度が増大してＳＮ比が格段に改善され微小な円形開口では不可能であった有意な信号が得られて赤外スペクトル分布を求めることが可能となるとともに、スリット状開口の幅方向に幅の大きさで決まる回折限界を超えた空間分解能を実現することができる。
【図面の簡単な説明】
【図１】 (a)，(b)のそれぞれは本発明の原理の説明図である。
【図２】 (a)，(b)のそれぞれはまた別の態様における本発明の原理の説明図である。
【図３】 (a)は微小円形開口を示し、(b)はスリット状開口を示す本発明に係る具体的な説明図である。
【図４】本発明の一実施例の要部の断面説明図である。
【図５】本発明の一実施例の全体を示す構成図である。
【図６】微小開口プローブの具体的な形状・構造を示す正面図である。
【図７】ＺｎＳｅの透過率スペクトルを示すグラフである。
【図８】スリット状微小開口を作製する装置を示す説明図である。
【図９】具体的な実施例における試料の走査を説明する図である。
【図１０】 (a)，(b)，(c)，(d)は透過率スペクトルの変化を示すグラフである。
【図１１】 (１)−ａ，(１)−ｂ，(２)−ａ，(２)−ｂ，(３)−ａ，(３)−ｂは各波数における透過率変化を示すグラフである。
【図１２】 (４)−ａ，(４)−ｂ，(５)−ａ，(５)−ｂはは各波数における透過率変化を示すグラフである。
【図１３】別の具体的な実施例における試料の走査を説明する図である。
【図１４】 (a)はポリイミド系耐熱高分子フィルムの透過率スペクトル、(b)は同吸収度スペクトルを示すグラフである。
【図１５】 (a)，(b)，(c)は透過率変化を示すグラフである。
【図１６】 (１)−(ａ)，(１)−(ｂ)，(２)−(ａ)，(２)−(ｂ)，(３)−(ａ)，(３)−(ｂ)は各波数における透過率変化を示すグラフである。
【図１７】 (４)−(ａ)，(４)−(ｂ)，(５)−(ａ)，(５)−(ｂ)，(６)−(ａ)，(６)−(ｂ)は各波数における透過率変化を示すグラフである。
【図１８】スリット状微小開口を用いる２次元分布の計測方法の説明図である。
【符号の説明】
１、２１ スリット状微小開口
３ 透過型試料
８ 反射型試料
Ｅ エバネッセント波ないしエバネッセント波の場
２０ 赤外透過プリズム、微小開口プローブ
２４ 金属コーティング膜
３０、６０ カセグレン対物鏡
４０ ステージ
５０ 試料ないし被検体
７０ ＭＣＴ検出器
８０ コンピュータ
９０ 赤外フーリエ変換分光器[0001]
[Industrial application fields]
The present invention relates to an infrared microspectrophotometric analysis method and apparatus, and more particularly, an infrared microspectrophotometric analysis method and apparatus capable of detecting an evanescent wave using a probe in the near field (near field) and realizing a spatial resolution exceeding the diffraction limit. About.
[0002]
[Background technology]
In recent years, fine processing technology has been developed in many industrial fields. Integrated circuit technology is just a typical example, and sub-micron processing is performed. In addition, the significant reduction in the existing product size found in micromachines and the increase in the density of information storage such as compact discs and video discs are demanding further progress in miniaturization and high technology. With the progress of such microfabrication technology, from the viewpoint of quality control, it is indispensable to measure these microfabricated products in a micro area. The measurement in the minute region is, for example, observation of the surface shape after fine processing of the object, local composition analysis on the surface, inspection of impurities, or the like. As the microstructure produced by processing technology shrinks and increases in density, the resolution required for measurement technology inevitably increases.
Infrared microspectroscopic analysis, which is one of the measurement methods in such a minute region, is widely used in the inspection process of products and analysis of materials. For example, analysis of deposits and impurities on the surface of magnetic disks and semiconductors, confirmation of products in the field of synthetic chemistry, composition analysis and structure analysis of polymer materials, and fine structure analysis of biological samples It is applied in.
However, since this infrared microspectroscopic analysis method uses light in the wavelength range of about 2 to 15 μm, its spatial resolution is about 10 μm at most due to the diffraction limit. Therefore, it is not possible to meet the demand for measuring the distribution of materials with a spatial resolution of micron order or submicron order.
[0003]
Recently, near-field scanning optical microscopes that can realize high spatial resolution without being limited by the wavelength of light have been studied. In this microscope, an evanescent wave that is localized on the sample surface and cannot propagate in the space is detected using a probe. As the probe, an optical fiber, a micropipette, a microscopic aperture, a metal needle, a microsphere, or the like with a pointed tip is used. By making the size of these probes (fiber tip diameter, aperture diameter, etc.) smaller than the wavelength, it is possible to observe a structure finer than the wavelength that could not be observed with light until now.
[0004]
[Prior art]
As a prior art, a technique “Near Field Scanning Optical Microscope” disclosed in Japanese Patent Laid-Open No. Hei 4-307510 is known. In this technology, probe light is supplied to a sample through a pinhole formed at the tip of an optical probe. The probe light is pulsed and the photodetector is operated in synchronism with the background light. The resolution is improved by reducing the influence and detecting the probe light with good S / N.
[0005]
On the other hand, an apparatus capable of realizing a spatial resolution exceeding the diffraction limit using an evanescent wave in infrared microscopic measurement is known from JP-A-6-94613. The technique disclosed here places an infrared transparent and high refractive index hemispherical prism in close contact with the upper surface of the test sample, narrows the focused spot diameter using the high refractive index of the prism, By using reflection, an evanescent wave is generated in a low refractive index medium (test sample), and light that has interacted with a substance is collected to obtain a spectral distribution.
[0006]
[Problems of the prior art]
Near-field scanning optical microscopes are said to be able to achieve high resolution without being limited by wavelength, but the probe aperture diameter and tip diameter must be smaller than the wavelength, and the sample can be probed with probe light. In any case where the light from the sample is picked up by the probe or the probe picks up the light, the detection light to be a signal is very weak and has a big problem that the SN ratio is very small.
On the other hand, in the above infrared microscopic measurement apparatus (Japanese Patent Laid-Open No. 6-94613), since the evanescent wave is generated using the total reflection of the high refractive index prism, the spatial resolution that can be realized is the high refractive index. There is a problem that it is limited by the refractive index of the prism, and a problem that a special device is required for the illumination optical system in order to obtain an incident light beam having a critical angle or more.
The present invention aims to solve the above problems.
[0007]
[Means for solving problems]
In the present invention, The width is less than the shortest wavelength of the infrared spectroscopic light used Using the slit-shaped aperture, the subject and the slit-shaped aperture are relatively scanned in the width direction of the slit-shaped aperture, and the subject and the slit-shaped aperture are relatively rotated by a predetermined angle to perform a plurality of scans. The two-dimensional spectrum distribution of the subject is obtained from a plurality of information thus obtained.
[0008]
According to one aspect of the present invention, an evanescent wave field is formed in the vicinity of the surface of the subject using spectroscopic light in the infrared region, an opening is made to face in the evanescent wave field, and the propagation light from the opening Alternatively, in the infrared microspectroscopy analysis method of collecting the propagation light from the minute region of the subject facing the aperture, obtaining a signal by photoelectrically converting the collected light, and obtaining a spectrum distribution based on this signal,
The opening is The width is less than the shortest wavelength of the infrared spectroscopic light used A plurality of scans, each having a slit-like opening, relatively scanning the subject and the slit-like opening in the width direction of the slit-like opening, and relatively rotating the subject and the slit-like opening by a predetermined angle. An infrared microspectroscopy analysis method is provided, characterized in that the two-dimensional spectral distribution of the subject is obtained from a plurality of obtained information.
[0009]
According to another aspect of the present invention, a stage on which the subject is placed, means for condensing and projecting light from the infrared spectrometer toward the subject placed on the stage, An infrared transmitting prism having an opening on the side facing the surface; means for condensing light that has passed through the opening of the infrared transmitting prism; means for photoelectrically converting the collected light into a signal; and Infrared microspectroscopic analyzer comprising means for calculating a spectral distribution based on
The opening is The width is less than the shortest wavelength of the infrared spectroscopic light used A slit-like opening, means for relatively scanning the stage and the infrared transmitting prism in the width direction of the slit-like opening, and means for relatively rotating the stage and the infrared transmitting prism by a predetermined angle. There is provided an infrared microspectroscopic analyzer characterized by being arranged.
[0010]
Furthermore, according to the present invention, as an infrared transmission prism that can be suitably used in the above infrared microspectroscopy analysis method and apparatus, the upper part is hemispherical, the lower part is conical, and the lower conical surface is A metal coating thin film is applied, and the lower end is disposed by forming a notch in the metal coating thin film. The width is less than the shortest wavelength of the infrared spectroscopic light used There is provided an infrared transmitting prism characterized in that there is a slit-like opening.
[0011]
[Principle Background of the Invention]
The present invention utilizes evanescent waves. First, a method for generating an evanescent wave will be described. There are (1) a method using a high refractive index crystal, (2) a method using a diffraction grating, and (3) a method using a minute aperture.
In (1) above, when light is incident on the low refractive index medium from the high refractive index medium at an incident angle greater than the critical angle, total reflection occurs, and at this time, an evanescent wave is generated in the low refractive index medium. In the method (2) using the diffraction grating, light only needs to be incident on the diffraction grating. An evanescent wave is generated regardless of the incident angle of light and the lattice spacing. The wavelength of diffracted light (including evanescent waves) generated by the diffraction grating ranges from the same size as the wavelength of propagating light to an infinitely small one. Further, if the grating interval is reduced, the proportion of the diffracted light that becomes an evanescent wave in the diffracted light of all orders increases. When a diffraction grating is used, when an evanescent wave is incident on the diffraction grating instead of propagating light, the diffracted light includes the evanescent wave together with the propagating light.
[0012]
The method (3) using the minute aperture can be considered by reducing it to a diffraction grating. That is, the structure of the one-dimensional aperture is considered to be a transmittance distribution represented by a rect function, and when this rect function is Fourier-transformed, it has a spatial frequency component from 0 to infinity. That is, the structure of the one-dimensional aperture can be considered as a superposition of diffraction gratings having various grating intervals. Therefore, the diffracted wave from the aperture can be considered in the same way as the diffracted wave from the diffraction grating. When the wavelength of incident light is λ, a spatial frequency component smaller than 1 / λ represents a diffraction wave component of a diffraction grating having a grating interval longer than the wavelength, and a spatial frequency component larger than 1 / λ is a grating interval shorter than the wavelength. Represents a diffracted wave component from a diffraction grating having Therefore, if the size of the aperture is made sufficiently smaller than the wavelength, the ratio of the diffracted wave component from the grating shorter than the wavelength to the entire diffracted wave increases, and as a result, many evanescent wave components exist. In other words, most of the diffracted waves are evanescent waves that give a resolution exceeding the wavelength, so if a diffracted wave from an aperture sufficiently smaller than the wavelength is used, a spatial resolution exceeding the diffraction limit can be realized. However, even when the size of the aperture is smaller than the wavelength, the diffracted wave from the aperture contains a propagating light component. As is well known, since the evanescent wave is attenuated exponentially, the evanescent wave is present only immediately after the opening (near field).
[0013]
As described above, there are propagating light and evanescent wave which is non-propagating light in the diffracted wave by the minute aperture. An object having a certain structure such as an uneven shape or transmittance distribution as well as a minute aperture can be described as a superposition of several diffraction gratings when the structure is expanded by Fourier series. Since the diffracted wave from the diffraction grating always includes an evanescent wave component, the diffracted wave from any structure has an evanescent wave. Of the spatial frequency components of the structure, the higher the proportion of high-frequency components (the greater the number of grating components having an interval shorter than the wavelength when Fourier series expansion is performed), the greater the proportion of evanescent waves in the diffracted waves.
The near-field optical microscope applies a probe to the field of the evanescent wave generated in the near-field region by any of the methods (1), (2), and (3) above, and the sample placed in the near-field region is It is a device that observes with light.
[0014]
In a near-field microscope using a microscopic aperture as a probe, it is the size of the aperture that determines the spatial resolution, and in principle there is no limit to the spatial resolution that can be realized. The action of the microscopic aperture associated with the sample of the microscope optical system using the microscopic aperture according to the present invention will be described with reference to FIGS.
[0015]
FIG. 1A shows an optical system in which incident light 2 is incident on an optical minute aperture 1, a sample 3 having an absorption distribution, for example, is placed immediately after the aperture 1, and transmitted light 4 from the sample 2 is detected. It is. At this time, the aperture 1 serves as a minute light source having a wavelength equal to or smaller than the wavelength. Of the diffracted light from the aperture 1, the evanescent wave E exists immediately after the aperture 1 in a region having the same size as the aperture size (for example, aperture diameter). Since the sample 3 having a structure such as an absorption distribution is placed in this region, the evanescent wave E is diffracted by the structure of the sample 3, and the propagating light component of the diffracted light is emitted as the transmitted light 4. This is detected. By making the size of the opening 1 sufficiently small, spatial resolution exceeding the wavelength can be realized.
[0016]
On the other hand, FIG. 1B shows an optical system in which incident light 5 is incident on the sample 3 and diffracted light due to the structure of the sample 3 is detected by the minute aperture 1. The light diffracted by the structure of the sample 3 also includes a propagating light component and an evanescent wave component, as in the case of the minute aperture. Among these, the evanescent wave component E is diffracted by the spatial frequency component below the wavelength in the structure of the sample 3. This Therefore, a spatial resolution exceeding the diffraction limit can be obtained by detecting this light. By placing the minute aperture 1 immediately after the sample 3, the evanescent wave E from the sample 3 is diffracted by the minute aperture 1, and the propagating light component 6 of the diffracted light is detected. FIG. 1 shows a transmission type optical system for a transparent sample 3, but cannot be applied to measurement of an opaque sample. A reflection type optical system as shown in FIG. 2 is used for an opaque sample.
[0017]
In FIG. 2A, incident light 7 is incident on the minute aperture 1, and the sample 8 is illuminated by diffracted light (including evanescent waves) from the aperture 1, so that the diffracted light (including evanescent waves) from the sample 8 is illuminated. ) Is detected as reflected light 9 as propagating light by the same minute aperture 1. Since the illumination of the sample 8 and the detection of the reflected light 9 can be performed by the single opening 1, the optical system is simplified. However, it is inevitable that the light wave will pass through the aperture 1 smaller than the wavelength twice, resulting in a very poor S / N ratio. In order to improve this, as shown in FIG. light Reference numeral 10 does not pass through the opening 1 but directly enters the sample 8. In this case, the incident light 10 is diffracted by the structure of the sample 8, and the evanescent wave component of the diffracted light (reflected light) is diffracted by the minute aperture 1 and detected as reflected light 11 as propagating light.
[0018]
As described above, when illuminating a sample with an aperture, when light enters the aperture, the evanescent wave component having a shorter wavelength than the propagating light attenuates exponentially from the aperture directly, and the propagating light component that has passed through the aperture Since the intensity distribution spatially spreads as the distance from the aperture increases, the spatial resolution that can be realized decreases as the distance from the aperture to the sample increases. The same applies to the optical system that detects the diffracted light from the sample through the aperture. Therefore, in any optical system, measurement can be performed with higher spatial resolution as the aperture is closer to the sample.
[0019]
Further, in the transmission type system, light passing through the sample is scattered and detected by the structure of the sample surface, so that information on only the sample surface cannot be obtained. On the other hand, in the reflection type optical system, only the light scattered on the surface is detected by irradiating with light from the opening of the sample surface, so that information such as the inside of the sample is not mixed into the signal light. Only the information can be detected. Therefore, for such a purpose, measurement using a reflection type optical system is more suitable than transmission type.
Regarding the aperture shape of the probe, as the minute aperture 1, all the conventional apertures (in a broad sense) have been circular apertures.
In a near-field microscope using a minute aperture probe, only the light that has passed through the aperture is detected out of the light from the light source, so the light utilization efficiency is very poor. Further, in order to obtain a spatial resolution exceeding the diffraction limit, it is necessary to make the size of the aperture smaller than the wavelength. Therefore, the detected signal light is very weak and difficult to measure.
[0020]
Actually, a circular aperture having a diameter r of 2 μm as shown in FIG. 3A was prepared by using the method of an embodiment described later, and an infrared spectrum distribution was measured using this circular aperture. The intensity of light from the aperture was very weak and could not be detected by clearly separating the signal light. In addition, in infrared spectroscopic analysis, which is the subject of the present invention, rather than the signal light itself, the purpose is to obtain an infrared spectrum distribution, and noise is excessively superimposed on this infrared spectrum distribution. In the first place, it cannot be put to practical use. Therefore, the signal light needs to be obtained in a form separated from noise components to some extent, and it is essential to realize an S / N ratio of a certain level or more.
[0021]
Therefore, in order to solve the above problems, the inventors have devised a slit shape by paying attention to the opening shape of the probe. When the opening is formed into a slit shape, the measurement light intensity proportional to the slit area is obtained, the S / N is greatly improved, and the spatial resolution determined by the width is obtained in the slit width direction. However, high spatial resolution cannot be obtained in the slit length direction. That is, spatial resolution exceeding the wavelength can be realized only in the one-dimensional direction.
[0022]
FIG. 4 shows a main part of the embodiment, and FIG. 5 illustrates the whole embodiment in an explanatory manner. In FIG. 4, reference numeral 20 denotes a prism provided with a slit-like opening 21 having a very small width, and transmits infrared light. Reference numeral 30 denotes a Cassegrain objective mirror including a primary mirror and a secondary mirror, and 31 denotes a light beam of infrared light. The prism 20 has a spherical surface 22 on the upper surface and a conical surface 23 on the lower surface. A thin coating film 24 made of metal is formed on the conical surface 23, and the slit-shaped minute opening 21 is opened at the top of the conical surface 23 so as to cut out the coating film 24. The infrared transmission prism 20 having the slit-shaped minute aperture 21 constitutes a minute aperture probe. In the embodiment, the infrared transmission prism 20 is fixed to the Cassegrain objective mirror 30 with the optical axis aligned. Since the upper surface of the prism 20 is a spherical surface 22, the focused light from the Cassegrain objective mirror 30 can be efficiently focused on the microscopic aperture 21 or the diffracted light from the microscopic aperture 21 can be efficiently focused on the Cassegrain objective mirror 30. . Further, the conical surface 23 on the lower surface of the prism 20. Is In order to easily confirm the position of the opening in the sample surface of the sample located below the opening 21 and to facilitate confirmation of the distance between the opening 21 and the sample, a conical shape is used. Yes.
[0023]
In FIG. 5 showing the system configuration, the micro-aperture probe 20 fixed to the Cassegrain objective mirror 30 is positioned above a stage 40 having an opening at the center. Subject on stage 40 As The sample 50 is placed. The stage 40 can be finely moved up and down, and can be driven to scan in the X or Y direction of two axes, that is, a plane XY perpendicular to the microscope optical axis. The Cassegrain objective mirror 60 is applied for illuminating the sample 50 when the sample 50 is transparent (transmission mode), and focuses the incident light beam and projects it onto the sample 50. The mirror 61 is for mode selection. In the transmission mode, the mirror 61 is retracted from the optical path. When the sample 50 is an opaque reflecting object (reflection mode), Like It is located in the optical path and reflects incident light upward. The mirror 62 and the mirror 63 are both optical path conversion mirrors, and the mirror 64 is a semi-transmissive mirror. The MCT detector 70 as the photoelectric detection means is disposed above the semi-transparent mirror 64 so that the detection portion is positioned at the position where the slit-shaped micro aperture 21 of the micro aperture probe 20 is imaged by the Cassegrain objective mirror 30. ing. The MCT detector 70 is a detector made of a quantum compound semiconductor containing mercury, cadmium and tellurium, and can detect infrared light with high sensitivity in a wide wavelength range.
[0024]
The computer 80 is composed of, for example, a personal computer and is a compatible interface. - It is coupled to the MCT detector 70 via the face and to the infrared spectrometer 90. In this embodiment, the infrared spectrometer 90 is a Fourier transform infrared spectrometer composed of an infrared light source 91 and a Fourier transform interferometer 92, and the computer 80 scans a moving mirror in the interferometer 92. Control. In this case, the photoelectric signal obtained by the MCT detector 70 is an interferogram, and the computer 80 inputs the interferogram signal from the MCT detector 70 in synchronization with the operation control of the infrared spectrometer 90. To do. The computer 80 also has means (many programs are configured) for calculating the infrared spectrum distribution from the obtained interferogram, and the obtained infrared spectrum distribution can be visualized by a monitor or recording means attached. Can be output as information
[0025]
On the other hand, a means 100 for observation between the probe and the sample is attached to the microscope main body, and the observation means 100 includes a long-acting objective lens 101, a CCD camera 102, and a monitor 103. The long working objective 101 has a magnification of 20 times, NA of 0.28, and working distance of 30.5 mm. By this observation means, the distance between the slit-shaped minute opening 21 and the surface of the sample 50 can be enlarged and visually observed, and the distance between the two can be controlled and set minutely.
The Cassegrain objective mirrors 30 and 60 have a magnification of 25 times and an NA of 0.62. The apparatus system shown in FIG. 5 is an existing infrared microspectroscopic system, for example, manufactured by Horiba, Ltd., infrared microspectroscopic system, product number FT-520, etc., except for the microscopic aperture probe 20 and the observation means 100. Can be used.
[0026]
In the operation of the system of FIG. 5, as described above, in this system, the sample 50 has two modes: a transmission mode in which light is transmitted through the sample 50 and measurement is performed, and a reflection mode in which reflected light from the sample 50 is measured. It can be switched according to the measurement. The light from the infrared light source 91 is first guided to the interferometer 92. When the mirror 61 is retracted from the optical path, the spectroscopic light exiting from the interferometer 92 proceeds to the mirror 63 and is reflected there to enter the lower Cassegrain objective mirror 60. The Cassegrain objective mirror 60 transmits the incident light. The sample 50 is condensed. Diffracted light including evanescent waves is generated in the sample 50 irradiated with the spectroscopic light. The slit-shaped micro-aperture 21 of the micro-aperture probe 20 is positioned in the field of the evanescent wave formed on the surface of the sample 50. The evanescent wave is converted into propagating light by the slit-shaped micro-aperture 21, and the propagating light is converted into the micro-aperture probe. Proceed upward through 20 prisms. The upper Cassegrain objective mirror 30 focuses all of the propagating light coming from the slit-shaped minute aperture 21 and condenses it on the MCT detector 70.
.
[0027]
On the other hand, when the mirror 61 is placed in the optical path, it is a reflection type measurement. In this case, the spectroscopic light from the interferometer 92 is reflected by the mirrors 61, 62, 64 and enters the upper Cassegrain objective mirror 30. The Cassegrain objective mirror 30 focuses incident light and focuses it on the slit-shaped microscopic aperture 21 of the microscopic aperture probe 20. At this time, since the slit-like micro-aperture 21 of the micro-aperture probe 20 is located at the center of the probe spherical surface, the light travels straight without being diffracted by the spherical surface, and the micro-aperture prism 20 is made of a material having a relatively high refractive index. Therefore, even if aberration is taken into consideration, the degree of light collection is high. The spectroscopic light condensed on the slit-shaped minute aperture 21 generates diffracted light including an evanescent wave through the slit-shaped minute aperture 21 and irradiates the surface of the sample 50. The light reflected by the sample 50 forms an evanescent wave field and generates diffracted light. The light is propagated by interaction with the slit-shaped micro-aperture 21 and travels upward in the prism of the micro-aperture probe 21. The upper Cassegrain objective mirror 30 focuses all of the propagating light traveling upward through the slit-shaped minute aperture 21 and condenses it on the MCT detector 70 via the semi-transmissive mirror 64.
[0028]
The computer 80 captures the photoelectric conversion signal from the MCT detector 70 in synchronization with the operation control of the interferometer 92. Since the spectroscope 90 is an infrared Fourier transform spectroscope, in this case, the photoelectric conversion signal is an interferogram, and the computer 80 captures the interferogram and stores it in the storage means as data. When an Fourier transform is performed on an interferogram converted into data by starting a Fourier transform means (for example, constituted by a program) built in the computer 80, an infrared absorption spectrum distribution or an infrared reflection spectrum distribution can be obtained and visualized. And output.
[0029]
The sample 50 is placed on an XY biaxial stage 40. The stage 40 is scanned in the direction of the slit width of the slit-shaped minute opening 21. By this scanning, a one-dimensional infrared absorption or reflection spectrum distribution can be measured with a spatial resolution corresponding to the slit width. If the direction of the wavelength λ is also added to the dimension, it can be said to be a two-dimensional spectral distribution of (λ, X) or (λ, Y). In addition, the above apparatus and method can be applied to a sample having a structure only in a one-dimensional direction, such as an organic multilayer film (in the actual measurement, a cross section in the film thickness direction or a direction crossing the film thickness direction). Infrared microspectroscopy analysis can be performed with a spatial resolution exceeding the diffraction limit. Therefore, the sample corresponding to the inside of the slit-like opening of the slit-like minute opening 21 50 For example, when a part of a two-dimensional structure having an absorption distribution exists in the minute region, a correct spectral distribution cannot be obtained. However, for organic multilayer film samples, etc., the analysis of the material distribution and fine structure in the one-dimensional direction of the sample can be performed on the micron order or the submicron order by narrowing the slit width. Have.
[0030]
For example, when there is an absorption object having a two-dimensional distribution in the observation region of the sample, an absorption image of the two-dimensional distribution can be reconstructed by a projection method described later. This reconstructed absorption image has a spatial resolution on the order of microns. In order to perform the projection method, the stage 40 shown in FIG. 5 is also configured to be rotatable about the optical axis, and can be scanned in the slit width direction by indexing a predetermined angle φ. Has been. Of course, XY biaxial scanning control, angle φ index control, and the like can be performed under the control of the computer 80.
[0031]
Next, the minute aperture probe 20 which is an important element in the embodiment of the present invention will be described in more detail.
FIG. 6 shows the external structure. It is obtained by processing a hemisphere of an infrared transmission crystal having a radius R of 7.5 mm. The lower conical surface 23 is machined so as to form a 15.0 ° generatrix with respect to the bottom surface of the hemisphere. The metal thin film 24 as an optical shield applied to the conical surface 23 is a gold coating film, and is formed by sputtering. The slit-shaped minute opening 21 is formed by cutting out this gold coating film. The slit width is about 2 μm and the slit length is about 70 μm. The annular zone 25 between the spherical surface 22 and the conical surface 23 is a gripping part at the time of processing, and is also a fixing part when the probe 20 is assembled to the optical system. The diameter of the annular zone 25 is 13 mm.
[0032]
The infrared transmission crystal serving as a substrate for forming the slit opening is made of ZnSe. There are other applicable crystals, which are shown in Table 1 along with their characteristics.
[0033]
[Table 1]

[0034]
Since the crystal is used as a probe as in the embodiment, several conditions must be satisfied. As conditions, (1) the transmittance is high in the wavelength range (about 2 to 15 μm) necessary for spectrum measurement in the mid-infrared region, and (2) the shape and size of the opening formed on the crystal can be easily confirmed. (3) Measurement of moisture-containing samples, long-term use and ease of handling, no deliquescence, (4) Opening on crystals, etc. Since crystal processing is required, processing and polishing are easy to perform, and (5) considering the handling at the time of processing and the like, it has as little toxicity as possible.
[0035]
Therefore, when the properties of each crystal are examined, ZnSe has high transmittance and no deliquescence in a wide wavelength range from visible to infrared. However, since it is designated as a medicinal exogenous substance, it must be handled with care. ZnS has the same properties as ZnSe, but its transmission wavelength range is slightly narrower than that of ZnSe. In addition, since it is designated as a medicinal exotoxin, handling is required. Halides such as KBr, KCl, and NaCl have high transmittance over a wide region from the visible region to the infrared region. Deliquescent. Moreover, processing is difficult. MgF ₂ Including CaF ₂ , BaF ₂ Such fluorides have almost no deliquescence and good transmittance from visible to infrared. However, the transmittance is low in the wavelength region of 10 μm or more, and spectrum measurement in this region becomes difficult. Ge and Si semiconductors are not deliquescent and have high transmittance up to the long wavelength side of the infrared region, but do not transmit visible light, so check the slits and the like produced on the crystal under visible light. I can't. Al ₂ O _Three , MgO and other oxides have no deliquescence and high transmittance for visible light. However, the transmittance in the long wavelength region is low and the processing is difficult because the crystal is hard.
[0036]
Therefore, all of the infrared transmission crystals listed in Table 1 have advantages and disadvantages, and there are no crystals that satisfy all of the above conditions. However, (5) toxicity is not a problem if sufficient care is taken during handling. Therefore, ZnSe or ZnS is considered to satisfy the best conditions. Further, in ZnSe and ZnS, ZnSe has a wider transmission wavelength range and lower toxicity than ZnS. ZnSe was selected. FIG. 7 shows the transmittance spectrum of the ZnSe crystal.
Gold is used as the coating metal for the coating on the conical surface of the ZnSe crystal. The complex refractive index and skin depth in the visible / infrared region of gold are as shown in Table 2. Skin depth is the intensity of 1 / e (e is the natural logarithm) when light travels through a metal.
It is defined by the distance to be the bottom.
[0037]
[Table 2]

[0038]
The metal film for forming the opening needs to sufficiently block light at a place other than the opening. If the metal film is sufficiently thick, light is blocked. On the other hand, if the film thickness is large, the light passing through the opening is scattered in the opening and the amount of light passing through the opening is reduced. Conceivable. Therefore, the thickness of the gold film is a thickness that can sufficiently block light and is as thin as possible.
[0039]
As can be seen from Table 2, the skin depth of gold in the visible region and the infrared region is slightly deeper in the visible region. Therefore, whether or not the light in the infrared region can be blocked by the coated gold film thickness can be determined by confirming that no light leaks in the visible region. When the gold thin film was irradiated with visible light and the transmitted light was confirmed, it was found that gold can block visible light with a film thickness of about 300 nm. So, the crystal surface To About 300 nm of gold was coated with a putter apparatus. In order to coat gold with crystals, the crystal surface must be sufficiently smooth beforehand. This is because the film thickness to be coated is about 300 nm, and if there are unevenness locally larger than this, light may leak from there. Therefore, before coating with gold, the crystal surface was polished with an abrasive sheet until it was sufficiently smooth. A lapping film sheet having a particle size of 0.3 μm was used as the polishing sheet.
[0040]
Next is the production of the slit aperture. For example, it is possible to apply a metal coating by placing a shield having the same shape as the slit on the crystal. However, since the target slit has a slit width of about 1 to 2 μm, a shield having such a shape is produced. It is virtually impossible to place it at the tip of the crystal. Although a method of coating a metal on a substrate and peeling off the metal in the slit shape portion by etching is also conceivable, since ZnSe is soluble in an acid, etching cannot be performed on ZnSe.
[0041]
Therefore, the method as shown in FIG. That is, the metal needle 26 and the metal-coated crystal 20 ′ are collided. The tip of the metal needle 26 has a wedge shape having the same width as the slit to be opened as shown in the enlarged view. The position of the crystal 20 ′ is controlled by the triaxial stage 27 and the piezo elements 28 and 29, and is made to collide with the fixed metal needle 26. Piezo elements that control the position of the crystal are mounted in two directions X and Z. That is, the direction in which the crystal approaches the tip of the needle (Z-axis direction in FIG. 8) and the direction in which the crystal and the needle are laterally shifted (X-axis direction in FIG. 8). When we tried to open the opening by actually colliding the metal needle with the crystal, only moving the crystal in the Z-axis direction and colliding with the needle would damage the metal on the crystal but transmit light. It was difficult to make such an opening. Therefore, when the crystal and the metal needle 26 contacted each other, the crystal was moved in the Z-axis direction and simultaneously moved in the X-axis direction by the piezo element 28, whereby a slit-shaped opening could be produced.
In order to adjust the state of contact between the metal needle and the crystal and the position of the micro slit on the crystal surface, the contact location between the tip of the metal needle and the tip of the crystal surface is determined in the X-axis direction and Y-axis direction of FIG. From two directions, slits were produced while observing with a microscope. Further, in order to adjust the amount of movement of the crystal in the Z-axis direction after the metal needle and the crystal are in contact, it is necessary to accurately grasp the position where the metal needle and the crystal are in contact. However, it is difficult to accurately determine the moment of contact only by microscopic observation from two directions. Therefore, by passing an electric current between the metal needle and the metal on the crystal surface, and confirming the position where the current starts to flow with an ammeter, the exact position where the metal needle and the crystal contacted was confirmed.
[0042]
As the material of the needle 26, stainless steel having high hardness was used so that the shape of the needle tip does not collapse due to collision with the crystal 20 '. Titanium was also used as a material with high hardness, but it was fragile and difficult to process.
The tip of a stainless steel metal rod having a diameter of about 1 mm was polished with a polishing sheet and processed into a wedge shape as shown in FIG. 8 having the same width as the target slit. The abrasive sheet used was a lapping film sheet with a particle size of 0.3 μm. The polishing of the needle was performed while observing the tip of the needle with a microscope.
[0043]
The slit-shaped micro-opening 21 produced by the above method illuminates the back side of the conical surface from the spherical surface side of the probe 20, and the shape and length can be confirmed by observing the surface of the conical surface with a microscope. In one example, it was confirmed that the slit width was 2 μm and the slit length was about 70 μm. In addition, several examples were prepared and applied in specific examples described later.
[0044]
[More specific examples]
5, the metal film 52 coated on the substrate 51 as shown in FIG. 9 is one-dimensionally scanned in the plane so as to be orthogonal to the slit-shaped opening 21, and the metal film 52 is detected at the edge portion of the metal film 52. Changes in the infrared transmission spectrum were measured.
The metal film 52 as the specimen of the sample 50 is gold (Au) having a film thickness of about 15 nm, and the base 51 has a high transmittance with respect to infrared light. ₂ The thickness is about 1 mm. The slit-shaped opening 21 has a slit width of about 2 μm and a slit length of about 50 μm. In this embodiment, the measurement is performed by bringing the slit-shaped opening 21 of the micro-aperture probe 20 into contact with the sample 50 at each measurement position.
[0045]
FIG. 10 shows the change in the transmittance spectrum in the vicinity of the edge of the metal film 52 obtained when the sample 50 is scanned in the direction of the arrow. (A), (b), (c), and (d) in FIG. 10 are spectra measured at four points separated by 2.5 μm. In addition, the frequency | count of integration of each spectrum measurement was 100 times. From FIG. 10, (a)-(b), (c)-(d) while Then, it can be seen that the decrease in transmittance is about 10% over the entire measurement wave number, but it is about 50% between (b) and (c). That is, it is considered that the edge of the metal film 52 is detected between (b) and (c). In addition, 2350cm of each spectrum ^-1 In the vicinity, the transmittance fluctuates greatly. ₂ Is due to absorption of
[0046]
Next, (1) 4000cm ^-1 (2.5μm), (2) 2500cm ^-1 (4μm), (3) 1666cm ^-1 (6μm), (4) 1250cm ^-1 (8μm), (5) 1000cm ^-1 Changes in transmittance at each wave number (wavelength) of (10 μm) are shown in FIGS. In both figures, (n) -a is the result when the probe 20 is used for measurement, and (n) -b is the result when this is not used. The field stop at the time of measuring b was about 8 μm × 8 μm. In order to examine the resolution for each wave number in each measurement of a and b, the difference D between the maximum value and the minimum value of the measured transmittance changes from 0.75D to 0.25D. Is defined as the spatial resolution R in this example. The straight lines representing 0.75D and 0.25D in FIGS. 11 and 12 are also shown. Table 3 summarizes the values of R read from each graph.
[0047]
[Table 3]

[0048]
From Table 3, the resolution R when measured using a probe is for Whereas it is about 2 to 3 μm, R when measured without a probe varies greatly from about 7 μm to about 15 μm depending on each wavelength. The spatial resolution in the absence of the probe is determined not only by the size of the field stop, but also by the spread of the image determined by the wavelength and the numerical aperture NA of the Cassegrain objective. Therefore, the longer the wavelength, the larger the R accordingly. . On the other hand, in near-field measurement using a slit probe, it can be seen that R takes a substantially constant value at each wavelength (1) to (5), and the spatial resolution does not depend on the wavelength and is determined by the slit width.
[0049]
The numerical aperture NA of the Cassegrain objective used in this example is 0.62. Therefore, the diffraction limit obtained from the Rayleigh equation (Δr = 0.61λ / NA) is almost the same as the wavelength used. From the results shown in Table 3, it can be seen that when a slit is used, a spatial resolution exceeding the diffraction limit is realized at any wavelength.
[0050]
Another example is measurement of the infrared absorption spectrum of a polymer film. As shown in FIG. 13, BaF with a thickness of about 1 mm. ₂ An infrared absorption spectrum is obtained by one-dimensionally scanning the polyimide heat-resistant polymer film 54 tightly fixed on the substrate 53 in the slit width direction of the slit 21 of the slit-shaped microscopic aperture probe 20. The scanning direction is the direction of the arrow in the figure, and the number of measurements for each spectrum is 100 times. Note that the slit width of the slit-shaped opening 21 of the applied micro-aperture probe 20 is about 2 μm, and the slit length is about 30 μm. The polyimide heat-resistant polymer film 54 has a thickness of about 50 μm. FIGS. 14 (a) and 14 (b) show the transmittance spectrum and the absorbance spectrum of this sample 54 obtained by ordinary infrared spectroscopic measurement, respectively.
In this embodiment, the slit-shaped opening 21 is scanned without contacting the sample 54. The distance between the sample 54 and the slit-shaped minute aperture 21 was about 1.5 μm when observed by the observation means 100 including the CCD camera 102 shown in FIG.
[0051]
15A, 15B, and 15C show transmittance spectra measured at intervals of 2 μm near the edge of the sample 54. FIG. It can be confirmed that the absorption peak of the sample 54 disappears between 4 μm of (a) to (c).
[0052]
Among the absorption peaks of the sample 54 shown in FIG. 14B, (1) 2964 cm indicated by an arrow ^-1 (3.37 μm), (2) 1778 cm ^-1 (5.62 μm), (3) 1722 cm ^-1 (5.81 μm), (4) 1600 cm ^-1 (6.25 μm), (5) 1502 cm ^-1 (6.66 μm), (6) 1382 cm ^-1 Changes in transmittance at six wave numbers (wavelengths) of (7.24 μm) are shown in FIGS. (N)-(a) in the figure is the case where measurement is performed using the probe 20, and (n)-(b) is the case where measurement is performed without using the probe. In (b), the measurement is performed using a field stop of about 5 μm × 5 μm.
[0053]
The vertical axis in each graph of FIGS. 16 and 17 indicates the transmittance at each wave number of 2160 cm without absorption of the sample 54. ^-1 The value divided by the transmittance. The reason for the division is as follows. That is, during the measurement, a phenomenon was observed in which the intensity of the entire signal decreased near the edge of the sample. This is because when the sample exists just before the slit, the transmitted light from the sample and the scattered light from the sample surface were incident on the slit, whereas the edge portion of the sample passed directly under the slit, When it does not exist just before the slit, it is considered that only light that has passed through the substrate approximately 50 μm (sample thickness) away from the slit is incident on the slit, and scattered light from the substrate surface does not reach the slit. Therefore, by referring to (dividing) the transmittance at a wave number where there is no absorption by the sample, it is considered that only the absorption of the sample among the changes in spectrum can be seen.
[0054]
As is apparent from FIGS. 16 and 17, when measured using the probe 20, it can be seen that the value of R according to the above definition of each wave number is constant at about 3 to 4 μm. In the case of (b), R tends to increase. In (b), it is understood that the S / N ratio is low and the error of R is large. Also in this embodiment, as described above, in the measurement without the probe, the spatial resolution decreases as the wavelength becomes longer, whereas in the measurement using the slit probe, the spatial resolution does not depend on the wavelength. It is clearly confirmed that it is determined by the slit width.
[0055]
In the example of (a) using a probe, the slit width is 2 μm and R is about 3 to 4 μm, whereas the previous embodiment uses a slit having the same width. The reason why the value of R is about 2.5 μm is different because the distance between the probe and the sample is larger in the latter case than in the former case. Moreover, since the thickness of the sample used in the latter example is about 50 μm, it is considered that the shape of the edge portion is dull compared to the former example.
[0056]
From the above examples, it can be seen that an infrared spectrum can be measured with a spatial resolution exceeding the diffraction limit by using a probe having a slit-like opening having a width shorter than the wavelength of light used for measurement. In ordinary infrared microspectroscopy, the spatial resolution is limited by the diffraction limit, and the limitation depends on the wavelength. Therefore, the phenomenon that the spatial resolution decreases as the measured wave number decreases, In the near-field microspectroscopic measurement used, the spatial resolution is not limited by the wavelength, and is basically determined only by the size of the slit width of the slit-shaped opening. According to the embodiment, a spatial resolution of about 2.5 μm is realized over a wavelength range of 2.5 to 10 μm.
[0057]
Therefore, if an infrared near-field microspectroscopy system as in the embodiment is used, it is possible to analyze a material distribution and a fine structure in a one-dimensional direction of a sample such as an organic multilayer film sample with a spatial resolution exceeding the diffraction limit.
[0058]
The slit width of the probe of the example was about 2 μm, but by further reducing the slit width, a higher spatial resolution can be realized, and a submicron order can be practically realized. However, if the slit width is made narrower in order to improve the spatial resolution, the signal light is weakened, but (1) the slit length is lengthened, (2) the light source intensity and detector sensitivity are improved, (3) spectrum. Increase the number of measurements and integration during measurement, (4) Use a Cassegrain objective mirror with a higher numerical aperture NA, (5) Make special measures to suppress noise light, and (6) Increase the linearity of the slit aperture It can be solved by making it strict and eliminating the unevenness of the opening edge, either alone or in combination.
[0059]
In addition, when a slit-like minute aperture is used as a probe, a spatial resolution exceeding the diffraction limit is obtained in the slit width direction, and a high resolution cannot be obtained in the slit length direction. Therefore, when measuring a sample such as a cross section of a multilayer film, the distribution of the sample (beyond the boundary based on the composition etc., A, B, C, D... For example, when different types of substances are scattered two-dimensionally in the sample, the distribution is simply obtained by scanning the slit-shaped micro-apertures in a one-dimensional direction. Cannot be observed.
[0060]
As a method for measuring the two-dimensional distribution of the sample while using the slit-shaped minute aperture, it is conceivable to scan the slit-shaped minute aperture not only in one direction but also in various directions within the sample surface.
As illustrated in FIG. 18, considering a case where a sample in which an object 55 having a two-dimensional distribution such as absorption exists in the sample is measured, if the slit 21 is scanned only in the one-dimensional direction, Information on the sample in the length direction of the slit cannot be obtained. Therefore, the slit 21 is rotated to similarly scan in the direction of the slit width. By this operation, information on the sample in the direction of the slit width after rotation is obtained. Slit rotation angle φ is constant and finely determined. ₁ , Φ ₂ , Φ _Three , Φ _Four ...,... Can be obtained as much as two-dimensional information about the sample. A wavelength λ ₀ The two-dimensional transmittance distribution at _n Can be obtained by reconstructing the one-dimensional transmittance distribution at. That is, it uses the principle of two-dimensional CT, and is a method of measuring one-dimensional projection data and reconstructing a two-dimensional cross-sectional image of the original object therefrom (a method called radon transformation and radon inverse transformation). An X-ray CT technique is known as a measurement method for reconstructing the structure of a sample from such observation results from multiple directions on the sample. The back projection method used in X-ray CT can be applied to the infrared transmittance distribution measurement using a slit-shaped minute aperture as it is.
[0061]
First, as shown in FIG. 18, (1) the slit is rotated by an angle φ, _φ Obtain the absorption intensity distribution in the direction. This is also called projection or projection, and is mathematically a Radon transform operation. Next, (2) φ is changed and many directions (φ _n : Projection at n = 1, 2, 3, 4,. (3) X in each projection _φ The absorption distribution in the direction _φ Arrange along the direction. This is an operation called back projection. Then, this is performed for each φ, and the sum is obtained in the entire measurement region. The reconstructed image (for example, the absorption distribution image) obtained from the back projection is an emphasized structure having a low spatial frequency in the structure of the sample. After that, the filtering operation for suppressing the low frequency is performed on each projection, and then the operation (3) is performed. This is referred to as filtered back projection and is also performed in X-ray CT.
[0062]
In the above-described embodiment, the infrared spectrometer 90 is a Fourier transform type spectrometer, but of course, a dispersion type infrared spectrometer can be applied. In addition, the stage 50 on which the sample or the subject 50 is placed has a configuration in which a drive mechanism is attached to the stage 50 and the stage 50 is scanned. In other words, it is possible to drive and scan the minute aperture probe 20. Similarly, the stage 50 is configured to be rotatable and controlled to be indexed at a predetermined angle. However, the stage is fixed and the microscopic aperture probe 20 can be rotated so that the microscopic aperture probe 20 can be indexed at a predetermined angle. It is good also as a structure which supports.
[0063]
【The invention's effect】
As described above, according to the present invention, an evanescent wave field is formed in the vicinity of the surface of the subject, and a slit-shaped opening having a small width is exposed in the evanescent wave field. Since the signal light intensity is increased, the signal-to-noise ratio is remarkably improved, and a significant signal that is impossible with a small circular aperture can be obtained to obtain the infrared spectrum distribution. In addition, it is possible to realize a spatial resolution that exceeds the diffraction limit determined by the width in the width direction of the slit-shaped opening.
[Brief description of the drawings]
FIGS. 1A and 1B are explanatory diagrams of the principle of the present invention.
FIGS. 2A and 2B are explanatory views of the principle of the present invention in another embodiment. FIG.
FIGS. 3A and 3B are specific explanatory views according to the present invention showing a minute circular opening and FIG. 3B showing a slit-like opening.
FIG. 4 is an explanatory cross-sectional view of a main part of an embodiment of the present invention.
FIG. 5 is a block diagram showing the entirety of an embodiment of the present invention.
FIG. 6 is a front view showing a specific shape and structure of a minute aperture probe.
FIG. 7 is a graph showing a transmittance spectrum of ZnSe.
FIG. 8 is an explanatory view showing an apparatus for producing slit-shaped minute openings.
FIG. 9 is a diagram illustrating scanning of a sample in a specific example.
10 (a), (b), (c), and (d) are graphs showing changes in the transmittance spectrum.
FIGS. 11A and 11B are graphs showing a change in transmittance at each wave number; FIGS. 11A and 11B are (1) -a, (1) -b, (2) -a, (2) -b, (3) -a, and (3) -b. is there.
FIGS. 12A to 12C are graphs showing changes in transmittance at each wave number; FIGS.
FIG. 13 is a diagram for explaining scanning of a sample in another specific example.
14A is a graph showing the transmittance spectrum of a polyimide heat-resistant polymer film, and FIG. 14B is a graph showing the same absorption spectrum.
FIGS. 15A, 15B, and 15C are graphs showing transmittance changes. FIGS.
FIG. 16 (1)-(a), (1)-(b), (2)-(a), (2)-(b), (3)-(a), (3)-(b ) Is a graph showing a change in transmittance at each wave number.
FIG. 17 (4)-(a), (4)-(b), (5)-(a), (5)-(b), (6)-(a), (6)-(b ) Is a graph showing a change in transmittance at each wave number.
FIG. 18 is an explanatory diagram of a two-dimensional distribution measuring method using a slit-shaped minute aperture.
[Explanation of symbols]
1,21 Slit-shaped minute aperture
3. Transmission sample
8 Reflective samples
E Evanescent wave or evanescent wave field
20 Infrared transmitting prism, micro aperture probe
24 Metal coating film
30, 60 Cassegrain objective mirror
40 stages
50 samples or subjects
70 MCT detector
80 computers
90 Infrared Fourier Transform Spectrometer

Claims (6)

An evanescent wave field is formed in the vicinity of the surface of the subject using infrared spectral light, and an opening is made to face the evanescent wave field. In the infrared microspectroscopy analysis method for collecting the propagation light from the region, photoelectrically converting the collected light to obtain a signal, and obtaining the spectral distribution based on this signal,
The opening is a slit-like opening whose width is equal to or shorter than the shortest wavelength of infrared spectroscopic light used , the subject and the slit-like opening are relatively scanned in the width direction of the slit-like opening, and the subject and the slit An infrared microspectroscopic analysis method characterized in that a two-dimensional spectral distribution of a subject is obtained from a plurality of pieces of information obtained by rotating the scanning aperture a predetermined angle relative to each other and performing the scanning a plurality of times. A stage on which the subject is placed, means for condensing and projecting light from the infrared spectrometer toward the subject placed on the stage, and a red having an opening on the side facing the surface of the subject An outer transmission prism, means for condensing light passing through the aperture of the infrared transmission prism, means for photoelectrically converting the collected light into a signal, and means for calculating a spectral distribution based on the signal In the infrared microspectrometer equipped with,
The aperture is a slit-shaped aperture whose width is equal to or shorter than the shortest wavelength of infrared spectroscopic light used , and means for relatively scanning the stage and the infrared transmitting prism in the width direction of the slit-shaped aperture, the stage, and the stage An infrared microspectroscopic analyzer characterized in that means for rotating the infrared transmission prism relatively by a predetermined angle is provided. The infrared microspectroscopy apparatus according to claim 2, wherein the infrared transmission prism has a hemispherical upper part and a conical lower part facing the surface of the subject. 4. The infrared microspectroscopy apparatus according to claim 2, wherein the infrared transmission prism is made of ZnSe crystal or ZnS crystal. The infrared microspectroscopy apparatus according to any one of claims 2 to 4 , wherein the slit opening is formed by cutting out a metal coating thin film. The upper part is hemispherical and the lower part has a conical shape, and the lower conical surface has a metal coating thin film, and the lower end has a width provided by forming a notch in the metal coating thin film. An infrared transmitting prism characterized by having a slit-shaped opening having a wavelength shorter than the shortest wavelength of the infrared spectral light to be used .

Images