201203352 六、發明說明： 【發明所屬之技術領域】 本發明係關於一種姓刻 籬早飩熱^ ㈣係_-種反應性 離子關的方法，以及用於實施祕刻方法的裝置。 【先前技術】 至7為止，例如專利文獻丨 + 為對由半導體材料或 料蚊獻2 ^己載，作 刻之裝置的H、已ft相形成之基板施行乾絲 ,,,^ 〜、已头有平行平板型之蝕刻裝置。平行 刻；f通常係於導入峨體之真空槽内，具 之’α電極、與和平台電極相對向而配置之上 ί電極。而平台電極係透過用以整合組抗之匹配箱而連接 著用以對基板施加偏壓之高頻電源。 ,此種平行平板型之蝕刻裴置中，當實施蝕刻處理時， 首先係，過設r上部電極之氣體“σ，將㈣氣體導 入。接著，由咼頻電源對基板施加高頻電壓，藉此由真空 槽内之蝕刻氣體生成電漿。然後，朝著相對於電漿偏壓為 負的平台電極’將電漿中的正離子引入，藉此化學性或物 理性地#刻基板。如此，於平行平板型之钮刻裝置中，藉 由所謂反應性蝕刻，於基板之厚度方向形成呈現預定形狀 之凹部。 [專利文獻1 ]日本專利公開公報特開平6-53191號 [專利文獻2]日本專利公開公報特開2004-128236號 【發明内容】 3 201203352 (發明所欲解決之課題） 惟，在如上述般朝著基板引入正離子之過程中，首先 係於基板之表面上形成具有大致一樣厚度之覆層。其後， 進入覆層内之正離子係藉由形成於覆層之電場而加速，且 在具有對應於電場加速之能量的狀態下，正離子係衝突於 基板之表面。 此時，若進入上述覆層内之正離子係全部沿著基板表 面之法線方向行進，不管覆層之厚度多大，該正離子均會 被引入電場方向之法線方向。因此，僅於基板之厚度方向 進行钱刻。 然而，進如上述覆層内之粒子中，除了行進於法線方 向之正離子，亦包含行進於與法線方向不同的方向之正離 子或自由基。此外，行進於與法線方向不同的方向之正離 子或自由基，當慢慢形成之凹部的底面與覆層上端部之距 離越大，則越難到達該凹部之底面。其結果，由於覆層之 厚度變大或者凹部之深度變大，粒子之行進距離變大，當 行進於法線方向之正離子在蝕刻凹部底面之期間，行進於 與法線方向不同的方向之正離子或自由基係僅在凹部開口 附近持續姓刻。然後，凹部之側壁的餘刻量變大，結杲導 致所謂CD (Critical Dimensions，臨界尺寸）損失變大。 尤其在對基板厚度方向之姓刻量大的加工（例如對石夕基板 形成貫通孔之加工），此種問題變得明顯。 另外，此種問題係可藉由提高電漿密度並減少覆層厚 度而減輕，但即便提高電漿密度，仍有其限制。又，CD損 失之問題不僅在上述平行平板型之蝕刻裝置發生，只要是 201203352 蝕刻對象物之基板被載置於電極上方，且對基板施加偏壓 之高頻電源係連接於該電極，具有此種構成之蝕刻裝置大 致上均會發生。 本發明係有鑑於上述之習知實情而完成者，其目的在 於提供可提高從基板表面於其厚度方向所形成之蝕刻形狀 的異向性之蝕刻方法以及使用於實施該蝕刻方法之裝置。 (解決課題之技術手段） 以下，記載用以解決上述課題的技術手段及其作用效 果。 本發明之第一態樣係一種蝕刻方法，其係將載置於真 空槽内之平台電極之基板於其厚度方向予以蝕刻之蝕刻方 法。該蝕刻方法具備：對上述真空槽内供給蝕刻氣體；以 60MHz以上且150MHz以下之頻率，對上述平台電極供給 高頻電力；以及以射入至上述基板之離子的能量分佈具有 如下之一對第一波峰與第二波峰的壓力而實施蝕刻，該一 對第一波峰係因由上述高頻電力之頻率所決定的相異之電 場加速，而於上述能量分佈之高能量侧一端所發生，該第 二波峰係於較上述一對第一波峰更低之能量區域所發生， 具有較上述一對第一波峰更高之強度。 本發明之第二態樣係將基板於其厚度方向予以蝕刻之 蝕刻裝置，具備：真空槽；平台電極，係配置於上述真空 槽内並載置上述基板；蝕刻氣體供給部，係對上述真空槽 内供給蝕刻氣體；高頻電源，係對上述平台電極供給高頻 電力；排氣部，係將上述真空槽内予以排氣；以及控制部， 係控制上述真空槽内之壓力，俾成為以射入至上述基板之 201203352 離子的能量分佈具有如下之一對第一波峰盥结 ^ 十丹弟二波峰的壓 力，該一對第一波峰係因由60MHZ以上且15〇MHz以 上述高頻電力之頻率所決定的相異之電場加速，而於上 能量分佈之高能量側-端所發生，該第二波峰係於較上^ 一對第一波峰更低之能量區域所發生，具有較上述一對 一波峰更高之強度。 當離子在基板表面上以預定厚度形成之覆層内不與其 他粒子衝突地到達基板之情況，在此離子遵循之6〇ΜΗζ' ^ 上且150MHz以下之頻率帶域中，經由該離子到達基板時 之高頻電力的相位，會生成受到大的電場加速之離子與未 受大的電場加速之離子。藉此，到達基板之離子之間會產 生能量差。其結果’由上述高頻電力之頻率所決定之互異 的電場加速所產生之一對波峰（所謂的雙峰波峰(bim〇dal peak))係可於離子的能量分佈内之高能量侧的一端觀察 到。 ^ 又’有在上述覆層内生成新的正離子之情況。例如， 仏有進入覆層内之正離子與中性粒子的電荷交換而生成正 離子之情況。此情況’新生成之正離子係由其所生成之位 置經由電場而被加速。因此，當覆層内新生成正離子之情 況’在較上述雙峰波峰更低能量之區域，係可觀察到與該 雙峰波峰不同的波峰。 根據本發明之第—及第二態樣，係在上述雙岭波峰與 較該雙峰波峰更低能量區域的波峰可在離子能量之分佈中 觀察到之壓力區域中，實行似彳。在此種壓力區域中，行 進於覆層内而被引人基板之離子（亦即對#刻有幫助之離 201203352 子），係由下述[第一離子]〜[第四離子]之4種類所構成。 [第一離子]沿著基板表面之法線方向而於覆層内行進 之離子。 [第二離子]沿著與基板表面之法線方向不同的方向 而於覆層内行進之離子。 [第三離子]生成於覆層内並於基板表面之法線方向行 進之離子。 [第四離子]生成於覆層内並於與基板表面之法線方 向不同的方向行進之離子。 在此，歸屬於第一波峰之雙峰波峰的離子中，從與基 板之法線方向不同的方向到達基板之上述[第二離子]，係 當蝕刻對象區域之深度越大，則越難到達該對象區域之底 面。另一方面，上述[第三離子]及上述[第四離子]，相較於 從覆層之上端部朝基板行進之上述[第一離子]及[第二離 子]，若到達該基板為止之行進距離越短，則越容易到達蝕 刻對象區域之底面。因此，若為使用上述[第一離子]〜[第 四離子]之蝕刻方法，當越生成上述[第三離子]及[第四離 子]，則越可能提高形成於基板厚度方向之蝕刻形狀的異向 性。尤其若為對基板厚度方向之蝕刻量增大的加工，其效 果更為顯著。 此外，根據上述蝕刻方法，由於亦存在具有相對高的 能量之上述[第一離子]及[第二離子]，亦可抑制壓力過度變 高反而使蝕刻形狀之異向性喪失的情況。 本發明之第三態樣之要旨係於上述第一態樣之蝕刻方 法中，上述壓力為50Pa以上且150Pa以下，且電漿之密度 201203352 為 lxl〇1()/cm3 以上且 5xl〇12/Cm3 以下。 本發明之第四態樣之要旨係於上述第二態樣之蚀刻裝 置：，上述控制部控制供給至上述真空槽内之触刻氣體的 >瓜里，俾使上述壓力為50Pa以上且ΐ5〇ρ“χ下，且電聚之 密度為ixio1(Vcm3以上且5xl〇!2/cm3以下。 基板上之覆層的厚度，係由電聚中所含之離子 之電聚密度所界定。在此，-般而言，刪 形成越薄的覆層，電漿密度越低，則形成越厚 掛第二波蜷此’不僅第一波峰與第二波峰之大小關係，當 之情況/，對於第二波♦強度的比率等要求再現性 於：點，二係定為如此的電漿密度範圍與壓力範圍。關 係設定為〜上且二之第二及第四態樣’由於壓力 lxl01(Vcm3以上 12 '以下，且電漿密度係設定為為 波峰強度相對μ 下’故可獲得上述之第一 另外，作t 峰度的㈣等之再現性。 方法，係可將生成f黎之空_壓力調整為預定值之 排氣量。關於此:整之流量，與調整触刻氣體之 調整姓刻氣體之:體之流量而調整，故相較於 i产。因此 排軋H的方式，可提高基板表面之流體的 籍^刻心若為提高生成電漿之空間的壓力之情況，可 制的區域巾^進行、’在控他刻速度之所謂反應速度控 ^^ ^耳轭上述蝕刻。因此可抑制蝕刻速度之下降， 同時刻形狀之異向性。 本發明$ & 义第五態樣之要旨係於上述第一或第三態樣之 8 201203352 蝕刻方法中，上述基板為矽基板，且 含氟氣體與具有氣、溴及碘之至少〜迹蝕刻氣體為含有 合氣體。 者的_化氫氣體之混 石夕基板之飯刻一般係使用與石夕级 合物之I化石夕的含氣氣體。另一方、、。:形成高揮發性化 土、漠及埃與石夕鍵結而成的函化石夕，位，以外之-元素的 性，而氫與發鍵結而成之氫化矽之=述氟化石夕更低 更n。 之揮發性係較氟化矽201203352 VI. Description of the Invention: [Technical Field to Be Invented by the Invention] The present invention relates to a method of surname heat, a system of reactive ionization, and a device for carrying out a secret engraving method. [Prior Art] Up to 7, for example, the patent document 丨+ is a dry wire that is formed on a substrate formed of a semiconductor material or a mosquito, which has been engraved, and has been formed by the ft phase, and has been dried. The head has a parallel plate type etching device. Parallel engraving; f is usually in the vacuum chamber into which the crucible is introduced, and has an 'α electrode, and an upper electrode disposed opposite to the platform electrode. The platform electrode is connected to a high frequency power source for biasing the substrate through a matching box for integrating the group resistance. In the etching apparatus of the parallel flat type, when the etching process is performed, first, the gas "σ" is introduced through the gas of the upper electrode of the r, and then the (four) gas is introduced. Then, the high frequency voltage is applied to the substrate by the frequency power source. This generates a plasma from the etching gas in the vacuum chamber. Then, positive ions in the plasma are introduced toward the platform electrode 'negatively negative with respect to the plasma bias, thereby chemically or physically engraving the substrate. In the stencil-type squeezing device, a recessed portion having a predetermined shape is formed in the thickness direction of the substrate by a so-called reactive etching. [Patent Document 1] Japanese Laid-Open Patent Publication No. Hei 6-53191 [Patent Document 2] Japanese Patent Laid-Open Publication No. 2004-128236 [Summary of the Invention] 3 201203352 (Problem to be Solved by the Invention) However, in the process of introducing positive ions toward a substrate as described above, firstly, a surface is formed on the surface of the substrate. a coating of the same thickness. Thereafter, the positive ions entering the cladding are accelerated by an electric field formed in the cladding, and in a state having an energy corresponding to the acceleration of the electric field, the positive ion system is rushed. Projecting on the surface of the substrate. At this time, if all the positive ions entering the coating travel along the normal direction of the surface of the substrate, regardless of the thickness of the coating, the positive ions are introduced into the normal direction of the electric field. Therefore, only in the thickness direction of the substrate, the particles in the coating layer include positive ions that travel in a direction different from the normal direction, in addition to the positive ions traveling in the normal direction. In addition, the positive ions or radicals traveling in a direction different from the normal direction, the greater the distance between the bottom surface of the recessed portion and the upper end portion of the coating, the harder it is to reach the bottom surface of the concave portion. Since the thickness of the coating becomes large or the depth of the concave portion becomes large, the traveling distance of the particles becomes large, and positive ions that travel in a direction different from the normal direction while positive ions traveling in the normal direction are etched under the bottom surface of the concave portion Or the radicals only continue to be surnamed near the opening of the recess. Then, the amount of the remaining side wall of the recess becomes large, and the crusting causes a so-called CD (Critical Dimensions) loss. This problem is particularly noticeable in the case of processing a large amount of the surname in the thickness direction of the substrate (for example, the formation of a through hole in the Shih-hs substrate). In addition, this problem can be achieved by increasing the plasma density and reducing Although the thickness of the coating is reduced, there is still a limit to the increase in the plasma density. Moreover, the problem of CD loss occurs not only in the above-described parallel plate type etching apparatus, but also as the substrate of 201203352 etching object is placed above the electrode. Further, a high-frequency power source that biases a substrate is connected to the electrode, and an etching apparatus having such a configuration substantially occurs. The present invention has been made in view of the above-described conventional circumstances, and an object thereof is to provide an improvement An etching method of an anisotropic shape of an etching shape formed on a surface of a substrate in a thickness direction thereof and an apparatus used for carrying out the etching method. (Technical means for solving the problem) Hereinafter, technical means for solving the above problems and effects thereof will be described. The first aspect of the present invention is an etching method which is an etching method in which a substrate of a stage electrode placed in a vacuum chamber is etched in a thickness direction thereof. The etching method includes: supplying an etching gas into the vacuum chamber; supplying high frequency power to the terrace electrode at a frequency of 60 MHz or more and 150 MHz or less; and having an energy distribution of ions incident on the substrate as follows: Etching is performed by a peak and a second peak, and the pair of first peaks are accelerated by a different electric field determined by the frequency of the high frequency power, and are generated at one end of the energy distribution at the high energy side. The two-peak system occurs in an energy region lower than the pair of first peaks, and has a higher intensity than the pair of first peaks. A second aspect of the present invention is an etching apparatus for etching a substrate in a thickness direction thereof, comprising: a vacuum chamber; a stage electrode disposed in the vacuum chamber to mount the substrate; and an etching gas supply unit for the vacuum An etching gas is supplied to the tank; a high-frequency power source supplies high-frequency power to the platform electrode; an exhaust unit evacuates the vacuum chamber; and a control unit controls the pressure in the vacuum chamber to The energy distribution of the 201203352 ion incident on the substrate has a pressure of one of the first peaks and the first peak of the tenth peak, and the pair of first peaks is caused by the above-mentioned high frequency power of 60 MHz or more and 15 〇 MHz. The different electric field determined by the frequency is accelerated, and occurs at the high energy side-end of the upper energy distribution, and the second peak occurs in an energy region lower than the upper first pair of peaks, having the above one Higher intensity for a peak. When the ions reach the substrate in a coating layer having a predetermined thickness on the surface of the substrate without colliding with other particles, the ions reach the substrate via the ions in the frequency band of 6 〇ΜΗζ' ^ and 150 MHz or less. The phase of the high-frequency power at that time generates ions that are accelerated by a large electric field and ions that are not accelerated by a large electric field. Thereby, an energy difference is generated between ions reaching the substrate. As a result, one of the pair of peaks (so-called bim peaks) generated by the acceleration of the mutually different electric fields determined by the frequency of the high-frequency power can be on the high energy side of the energy distribution of the ions. Observed at one end. ^ Also, there is a case where a new positive ion is generated in the above coating. For example, 仏 has a charge exchange between positive ions and neutral particles entering the cladding to generate positive ions. In this case, the newly generated positive ion is accelerated by the electric field from the position generated by it. Therefore, when a positive ion is newly formed in the cladding layer, a peak different from the peak of the bimodal peak can be observed in a region lower than the peak of the bimodal peak. According to the first and second aspects of the present invention, the peaks in the lower energy region of the double ridge peak and the peak of the bimodal peak can be similarly observed in the pressure region observed in the distribution of the ion energy. In such a pressure region, the ions that travel in the cladding and are attracted to the substrate (that is, the help of 201203352) are caused by the following [first ion] to [fourth ion] The type consists of. [First ion] ions traveling in the cladding along the normal direction of the surface of the substrate. [Second ion] ions traveling in the cladding in a direction different from the normal direction of the surface of the substrate. [Third ion] An ion generated in the cladding and traveling in the normal direction of the surface of the substrate. [Fourth ion] An ion generated in the cladding and traveling in a direction different from the normal direction of the surface of the substrate. Here, in the ions belonging to the bimodal peak of the first peak, the above-mentioned [second ion] reaching the substrate from a direction different from the normal direction of the substrate is more difficult to reach when the depth of the etching target region is larger. The bottom surface of the object area. On the other hand, the above [third ion] and the above [fourth ion] are compared with the above-mentioned [first ion] and [second ion] which travel toward the substrate from the upper end portion of the cladding layer, and reach the substrate. The shorter the travel distance, the easier it is to reach the bottom surface of the etching target region. Therefore, in the etching method using the above [first ion] to [fourth ion], the more the above-mentioned [third ion] and [fourth ion] are formed, the more likely the etching shape formed in the thickness direction of the substrate is increased. Anisotropy. In particular, if the etching amount in the thickness direction of the substrate is increased, the effect is more remarkable. Further, according to the above etching method, since the above-mentioned [first ion] and [second ion] having relatively high energy are also present, it is possible to suppress the excessive increase in pressure and the loss of the anisotropy of the etching shape. According to a third aspect of the present invention, in the etching method of the first aspect, the pressure is 50 Pa or more and 150 Pa or less, and the plasma density 201203352 is lxl 〇 1 ()/cm 3 or more and 5 x 1 〇 12 / Below Cm3. According to a fourth aspect of the present invention, in the etching apparatus of the second aspect, the control unit controls the melon gas supplied to the gas in the vacuum chamber, and the pressure is 50 Pa or more and ΐ5 〇ρ"χ, and the density of electropolymerization is ixio1 (Vcm3 or more and 5xl〇!2/cm3 or less. The thickness of the coating on the substrate is defined by the electropolymerization density of the ions contained in the electropolymer. Therefore, in general, the thinner the coating is formed, the lower the plasma density is, the thicker the second wave is formed. This is not only the relationship between the first peak and the second peak, but in the case of The second wave ♦ intensity ratio and the like require reproducibility at: point, the second system is set to such a plasma density range and pressure range. The relationship is set to ~ and the second and fourth aspects of the second 'because of the pressure lxl01 (Vcm3 Above 12' or less, and the plasma density is set to be the peak intensity relative to μ', so that the first one can be obtained, and the reproducibility of (t) of t kurtosis can be obtained. The pressure is adjusted to the predetermined amount of exhaust. About this: the whole flow, and adjustment The gas is adjusted to the gas of the surname: the volume of the body is adjusted, so it is compared with the production of the body. Therefore, the way of rolling the H can increase the pressure of the fluid on the surface of the substrate, so as to increase the pressure of the space for generating plasma. In the case where the area towel can be made, the so-called reaction speed control at the speed of the control is controlled by the above-mentioned etching, so that the etching speed can be suppressed and the anisotropy of the shape can be suppressed. The fifth aspect of the present invention is directed to the first or third aspect of the method of 201203352, wherein the substrate is a germanium substrate, and the fluorine-containing gas and at least the trace etching gas having gas, bromine and iodine are contained. The gas of the gas mixture of the _ hydrogen gas is generally used as the gas-containing gas of the I fossil of the stone grading compound. The other side, the formation of high volatility soil, infiltration The fossilized stone formed by the combination of Ai and Shi Xi, the bit, the nature of the element other than the element, and the hydrogenation of hydrogen and the bond of the hydrogenated ytterbium = lower and more n. Phlegm
户女=此*發明之第五態樣中，作A 在作為餘刻主體之含氣氣 乍為餘刻氣體，係使用 的鹵化氫氣體。 ’、 虱、〉臭及碘之至少一者 氫離子而被促進f ，基板厚度方向進行之餘刻係利用 齒化石夕之堆積物二，該厚度方向之凹部的周壁會因 得蝕刻形狀的異向性=被蝕刻。因此，可更容易地獲 【實施方式】 以下，參昭 於實施該飼刻^ 1至圖6說明本發明的飯刻方法及將用 圖1係夺-士之钱刻裝置予以具體化之—實施形態。 平板型或容量纟Γ人實施形態之蝕刻裝置，即一般稱為平行 之餘刻裝置係具:型之餘刻裝置的概略構成。平行平板型 配設的平台電極在真空槽1之内部透過絕緣間隔物2而 板S，且供給；3。平台電極3係載置有蝕刻對象物之基 之匹配箱5而S'員電力之高頻電源4係透過用以整合阻抗 平台電極3的^於平台電極3。高頻電源4係將供給給 、阿頻電力之頻率設定為6〇MHz以上且 201203352 150MHz 以下。 :外於平口電極3之附近係配置有用以分析射入至 :反t正離子的能量之離子能量分析器 收集朝平台電極3行進之粒子的—部分，將該粒 古Γΐ3之正離子依電子溫度（ev)進行分類，並測定具 有各電子溫度之正離子的數量。 於平台電極3之上方係配設著作為與其相對向之接地 能，且設有氣體導人孔之喷灑板7，該氣體導入孔 糸將從真空槽1之外部供給祕職體導人至真空槽！ =。喷灑板7係連接有氣體供給部8，其係供給基板§之 ，刻用的各種氣體。此外，真空们係連結有排氣部9, ;、係將，空槽丨内之*刻氣體或大氣等各種氣體予以排 氣本貫施形態中’係利用氣體供給部8調整從上述氣體 =給部8供給至真空槽i的每單位時間之氣體的流量，或 =用排氣部9調整排氣部9每單位時間的排氣量，藉此 使二槽1内之壓力調整為預定值（本實施形態中為箫a 以上且150Pa以下）。或者，藉由氣體供給部8及排氣部9 之〇作，使真空槽1内之壓力調整為上述預定值。 上述平行平板型之蝕刻裝置中實施蝕刻時，首先係將 作為則對象之基板S搬送至真空槽i内，載置於平台電 極^ ^。接著由排氣部9將真空槽1内之大氣等排出，並 二從氣體供給部8對真空槽1内供給預定流量之姓刻氣 此時，藉由從氣體供給部8所供給之蝕刻氣體的流量、 與上述排氣部9之排氣量，將真空槽i内調整為預定之壓 力於蝕刻氣體之供給之後，實施從高頻電源4向平台電 10 201203352 極3之高頻電極的供給，則因載置於平台電極3之基板S 與喷灑板7之間的放電，而生成蝕刻氣體之電漿。另外， 上述排氣部9所進行之排氣處理、氣體供給部8所進行之 氣體供給處理、高頻電源4所進行之電力供給處理等在實 施裝置中所實行之各種處理，係利用連接於蝕刻裝置之控 制部10驅動各部，藉此而實行。 於電漿之生成時，電漿中之電子在基板S及真空槽1 之表面衝突，藉此，該等表面相對於電漿係被偏壓為負， 且該表面係大致相同地形成具有預定厚度的覆層。此外， 若電漿中之正離子到達覆層與該覆層以外之電漿的邊界 (覆層之上端部），則該正離子被引入經偏壓為負的基板S 之表面。如此，引入至基板S表面之正離子係化學性或物 理性地蝕刻基板S，藉此，伴隨有預定蝕刻形狀之蝕刻係 自基板S之表面於該基板S之厚度方向進行。另外，形成 有從基板S之表面於該基板S之厚度方向延伸的凹部。 另外，本實施形態中，係使上述基板S例如為矽基板， 作為蝕刻氣體，係使用由六氟化琉（SF6)氣體、氧（02) 氣以及溴化氫（HBr)氣體所構成之混合氣體。惟，作為 蝕刻氣體，並不限於此種組合，只要為含氟氣體與具有氯、 溴及碘之至少一者的鹵化氫氣體混合而成之氣體，均可採 用作為該蝕刻氣體。更詳細而言，作為構成蝕刻氣體之含 氟氣體，例如可採用六氟化硫氣體、五氟化碘（IF3)氣體、 三氟化氣（C1F3)氣體、三氟化硼（BF3)氣體、亞硫醯氟 (SOF2)氣體、硫醯氟（S02F2)氣體、以及氟化羰基（COF2) 氣體中之至少一者。又，作為鹵化氫氣體，例如可採用氯 Ι；ϊ 11 201203352 化氫（HCl)氣體、填化 、 至少一者。 以及碘化氫（Hi)氣體中之 本實施形態中，作為蝕 刻時-般所用之含體，係使用含材基板姓 之溴化氫氣體的混合氣體。葬、氟化硫氣體、與_化氫氣體 行的蝕刻係利用形成揮發性二，對基板S之厚度方向進 子而被促進。又，於基板s 乂氟化矽更高的氫化矽之氫離 係因函切之堆積二變得之不厚延，凹部之周壁， 用其他氣體種類而進行蝕 破蝕刻。因此，相較於使 狀之異向性。 Λ情況，變得容易獲得蝕刻形 另外，本實施形態中，蝕 與鹵化氫氣體之外，進一牛人Χ ^除了上述含氟氣體 的凹部之周壁變得更不易^刻於基板8之厚度方向延伸 —、佳夕τm2上述反應性離子蝕刻時，於覆層内朝基板S 灯 子係不與其他粒子衝突地到達基板S之表面。 此情況’若高頻電力之頻率帶域為上述6GMHZ以上且 l5〇MHz以下’則正離子係遵循高週波電力所形成 其結果:當正離子到達基板s時，對應於高觀力之^ = 係生成又到大的電場加速之正離子與未受到大電場加速 正離子。&，該等正離子之間發生能量差。因此，由高 電力之頻率所決定的互異之電場加速所形成的1波略 (::胃雙峰波峰)，係於正離子之能量分佈内於高能量側之 —一 察到。 12 201203352 峰而峰波峰係以在正離子之能量分佈内之優越波 s 情況，則正離子之大部分係從形成於基板 =二=的上端部行進至基板s之表面。如此在覆 ==::子’除1行進於法線方向之正離子之外， 於與:線方向;=3:同的方向之正離子。如此行進 正離子，係若慢慢形成之凹部 二二:則越難到達該凹部之底面。其結果，在行進 之正離子增大凹部之深度的期間，行進於與法 、士 I雔《ιΙΓ方向之正離子係持續擴大凹部之開口。因 :觀=:係=離子之能量分佈内之優越波峰而 ’、 ^'凹0卩/衣度越大，則凹部側壁之蝕刻量越 大。其結果，所謂的CD損失變大。 曰因此’本實施形態巾U在正離子之能量分佈中能 里小的第—波峰較第_波峰之上述雙峰波峰成為更優越的 條件，尤其是壓力條件，來實施钱刻。 其次，參照圖2〜圖5，以下述蝕刻條件為一例，說明 咼頻電力之頻率及真空槽1内之壓力相對於引入至基板s 之正離子的能置分佈之依存性。圖2〜圖5係下述餘刻條 件之正離子的能量分佈，依序例示將高頻電力之頻率定為 40MHz、60MHz、150MHz、250MHz 時之能量分佈。 (蝕刻條件） 蝕刻氣體：由SF6氣體、〇2氣體、HBr氣體構成之混 合氣體。 平台電極3之高頻電力的輸出值：i〇w/cm2。 真空槽 1 内之壓力：0.2Pa、25Pa、50Pa、150Pa、250Pa。 Η 13 201203352 钮刻時之電漿密度：1 x 1 〇l 1。 另外，圖2〜圖5所示之離子能量分佈，係從設於上 祕刻裝置之離子能量分㈣6所輪4之輸出值所得。 又择各圖中’真空槽丨内之壓力為G.2l>a時之離子能量係 以實現表不，該壓力為25Pa時之離子能4細虛線表示。 又，真空槽1内之壓力為50Pa時的離子能量係以二點鍊接 線表不，祕力為丨娜時之離子料―-點鍊接線表 不，該壓力為250Pa時之離子能量係以教線表示。 如圖2所示，當高頻電力之頻率為4〇MHz:情況，在 壓力為〇.2Pa之條件與壓力為25Pa之條件中，係於高能量 侧之-端觀察到雙峰波峰BP，其由在約14_所觀察到的 第-向能量波峰？1與在約65eV所觀察叫第二高能量波 峰P2所構成。相對於此，當歷力為較高之條件（顺、 150Pa、250Pa)的情況，並未觀察到雙峰波峰Bp。尤其在 壓力為2施之條件中’幾乎未觀察到具有由雙峰波峰BP 所包夾之能量區域65〜140eV的能量之正離子。 如圖3所示，當高頻電力之頻率為6〇MHz2情況， 壓力為250Pa以外（〇.2Pa、25pa、挪、ι卿a)之條件， 係於高能量侧之一端觀察到雙峰波峰Bp，豆 125eV所觀察到的第一高能哥沽庵 八’、由在約 同此ά波峰pl與在約80eV所觀察 到的第二高能量波峰P2所構成。 、 另外’在壓力為50Pa之條件與壓力為15〇pa之 中於車乂該雙峰波峰Bp更低能量側，係可 雙 波峰BP具有更高強度的第二波峰之低能量波峰 顯不低能量波峰P3之正離子，係具有較由電場加速所賦予 201203352 之能量更低的能量。因此，具有低能量波峰p3之正離子係 可歸屬於在覆層内所生成並從覆層内之途巾開始受到電場 加速之正離子，例如因中性粒子與正離子之電荷交換而新 生成的正離子。另外，具有較雙峰波峰Bp之強度更高強 度的低能量波峰P3在該雙峰波峰抑之低能量側被觀察到 的上述傾向，係於5〇Pa以上且150Pa以下之全範圍中被觀 察到的傾向。 如圖4所示，當頻率為15〇MHz之情況，係與上述 60MHz時相同，在壓力為25〇Pa以外（〇 2ρ&、25ρ&、5〇pa、 150Pa)之條件中，係於尚能量側之一端觀察到雙峰波峰 BP，其係由在約n〇eV所觀察到的第一高能量波峰ρι盥 在約9〇eV所觀察到的第二高能量波峰p2所構成。又，^ 上述60MHz時_ ’在壓力為5〇Pa之條件與壓力為ΐ5〇ρ& 之條件巾，係於較該雙峰波锋BP更低能量顺察到具有 較雙峰波峰BP更高之強度的第二波峰之低能量波峰。 另外，於上述60MHz時，低能量波峰p3在5〇ρ&及ΐ5〇ρ& 之各壓力條件下僅觀察到一個，相對於此，於i5〇MHz， 係在50Pa及150Pa之各壓力條件中觀察到三個低能量波 峰。又，與上述60MHz時相同，具有較雙峰波峰δρ之強 度更高強度的低能量波峰Ρ3在該雙峰波峰Βρ之低能量侧 被觀察到的上述傾向，係於5〇Pa以上且15〇Pa以下之全 圍中被觀察到的傾向。 王巳 如圖5所示’當頻率為25〇MHz之情況，與上述 時相同，壓力為〇.2Pa之條件與壓力為“以之條件中，係 於高能量側之一端觀察到雙峰波峰Bp，其係由在約Wei 15 201203352 所觀察到的第-高能量波峰ρι 第二高能量波峰卩2&@ 、約〇5eV所觀察到的 如此，在射入至基=子=雙峰波峰BP。 上述雙較峰BP、與缺气量分佈巾，觀察到 強度高的低能量波峰pj::，更低能量之區域且 上且圓HZ以下，且真空之;頻率帶域為嶋z以 以下的範m卜本實_料，上且15〇Pa 施帅夺，頻率帶域係被定 二述钱刻裝置中實 下，且被定為滿足下述[第一條件上=顧沿以 圍。例如，在上述钮刻條件中J丄第一條件]之塵力範 以上且15〇_z以下，且壓力 _z 150Pa以下。 τ被疋為5〇Pa以上且 [第一條件]雙峰波峰3卩 &曰 [第二條件]觀察恥較雙峰被觀察到。 度較雙峰波峰BP更高之低能量波夸Μ更低能量側，且強 何電聚之空間的屋力調整為預定值之方法， 知可舉出调整韻刻氣體之⑽疋值之方法， 而從可調整基板S表面之产诚％氣體之排氣量， 利用朗氣體之流量來調整該塵力H點而言，較佳係 =而控制韻刻速度之所刻反應 二度，電漿中所 电聚在度而界定。—般而言，已知 201203352 =度越高，則形成越薄的覆 越低，則形 成越厚的覆層。因此，不僅 电 告對低炉旦、、&上述[苐〜條件]及[第二條件]， 二®…* 3之強度與雙峰波導BP之強度的比率等 a圍之=之情況’較佳係界定此種電榮密度範圍與塵力 3=。故，在本實施形態中，係將蝕刻氣體之流量 3 之輸出值設定為使壓力為50Pa以上且150Pa以 下，且電漿密度成為lxl〇1〇/cm3以上且5xi〇lw以下。 穷厭3 ’參照圖6、圖7,說明如上述般決定之頻率帶域 得之侧形狀。圖6係表示實施上述反應性 …刻時於基板S之表面附近的正離子之軌道。又，圖 7係示意地表示由反應性離子朗所形成之侧形狀。 如圖6所示，將基板3收容於上述截刻裝置後，對真 空槽1之餘刻氣體之供給與對平台電極3之高頻電流之供 給係依序實施，藉此，從由中性粒子Np所構成之钱刻氣 體’生成含有電子E與正離子IP之㈣Rp。接著，遵循 由而頻電流卿成之電場所生成的㈣速度較快之電子£ 係到達基板S之表面，藉此，基板s之表面係相對於電聚 被偏壓為負。藉此’於基板S之表面，電子E反斥而彈回 離開基板s表面之側，故在距基板s表面之預定硅離 係形成覆層Rs，其係正離子ip之數量較電子E之數量更 多之區域。然後，較覆層Rs更遠離基板s之區域係存在 正離子IP、電子E及中性粒子NP，且成為電性為中性之 電漿Pr之區域。電漿Rp中之正離子Ip若到達該電漿汉 與覆層Rs之邊界面BP，則正電位之正離子ip被引入經^ 壓為負的基板S表面，換言之，正離子IP係行進至基板$ 17 201203352 之表面。 在此’作為射入至基板S表面之正離子IP，係大致區 分為以下二者。 不與其他粒子衝突地到達基板S表面之第一正離子 IPa。 •於覆層Rs内’利用中性粒子NP與正離子Ip之電 荷交換，由中性粒子NP所新生成的第二正離子11>15。 上述第一正離子IPa係接受來自高頻電力之電場加 速，且至射入基板S為止均維持由該加速所獲得之能量。 亦第一正離子Ipa係歸屬於先前圖2〜圖5所示之= 一尚能量波峰P1及第二高能量波峰p2所構成之 BP的正離子。上述第-正離子IPa係如圖6所示，U峰 進於基板s表面之法線方向的第一正離子IPa與行進二: 該法線方向不同的方向之第一正離子IPa。另外，該等 於法線方向的第-正離子IPa與行進於與該法線方向= 的方向之第一正離子IPa，係不與其他正離子Ip或中同 子NP衝突地到達基板s之表面。惟’上述第一正離板 中沿著與法線方向不同的方向行進之第一正離子ipa, 形成於基板s之凹部的深度越大，則越難到達該凹部 另一方面，上述第二正離子IPb係由電場加速 能量較上述第一正離子IPa更小之正離子，亦即，係=之 於先前圖3、圖4所示之低能量波峰p3的正離子。^二屬 二正離子IPb係與上述第一正離子IPa同樣地，包含^第 於上述法線方向的第二正離子ipb與行進於與上述法綠= 201203352 + 篦二疋離子1押。如此生成於覆層以内之 正離子，於開始被電碭加速時向的分佈係與上述 第-正離子IPa相同，但開始被電場加速時之位置係較上 述第一正離子IPa更靠近基板亦即，第二正離子IPb之 行進距離係距上述電荷交換進行地點之距離Db，而由電荷 交換係於覆層Rs内發生之事實而言’距離Db係較相當於 覆層Rs厚度之上述距離Da更小。因此，由電荷交換所產 生之第二正離子IPb係因與上述第一正離子ipa相同的理 由，即便其行進方向偏離法線方向，到達基板s為只之行 進距離相較於上述第一正離子IPa為越短，則越容易到達 形成於基板S之凹部的底面。 利用上述第一正離子IPa及第二正離子lpb進行反應 性離子蝕刻之情況所形成的蝕刻形狀，係分別示意性地示 於圖7 ( a)及（b)。另外，如圖7所示，利用反應性離子 蝕刻於基板S形成預定的蝕刻形狀（例如形成凹部H)之 情況丄在反應性離子蝕刻之前，係於基板s之表面形成具 有預疋開口部之遮罩M。然後，藉由從此開口部對基板s 使上述正離子IP射入，基板s被蝕刻。 如圖7 (a)所示，當利用上述第一正離子IPa實施蝕 二之情況，係經由在與上述法線方向不同的方向行進之第 而正，子1Pa ’朝較遮罩M開口部Ma之法線方向更外側 進行蝕刻。藉此，蝕刻形狀之凹部H的最大直徑〇丨沾 係較開口部Ma之直獲此八更大。 lpb另:方面，如圖7 (b)所示，當利用上述第二正離子 實軛蝕刻之情況，由於第二正離子Ipb係從基板s之附 19 201203352 口部Ma之法線方向 的直彳生係大致等於開 近被電場加速，故容易沿著遮罩Μ門 進行蝕刻。因此，蝕刻形狀之凹部J 口部Ma之直徑DiaA。 因此’射入至基板s之正離子Ip 所佔之比例增大，亦即，如先前 ，第二正離子iPb 子能量分佈中，藉由以滿足上述3 2 4所不，於離 之條件來實施蝕刻，可提高凹部H 1 ]及[第二條件] 1 η之異向性。 另夕’如該圖3、圖4所示，歸屬於低能量波峰ρ ^-正離子IPb ’其離子能量係較歸屬於雙峰波峰即 =離子IPa低，故若以僅存在有第二正離子肌 = 實施餘刻’則有侧速度降低之虞1於此點，根據本= 施形態，係以歸屬於雙蜂波峰Bp之第一正離子ipa ^ t低能量料P3之^正料1料紅條她^ 故可經由第二正離子IPb而維持姓刻形狀之異向性，同時 經由第一正離子IPa而抑制蝕刻速度之降低。 [實施例1] 對厚度750μιη之8英吋矽基板塗佈具有直徑5〇μιη之 開口部的遮罩後，使用上述平行平板型之蝕刻裝置，以由 SF6氣體、〇2氣體、HBr氣體所構成之混合氣體作為蝕刻 氣體，實施蝕刻。此時，從高頻電源輸出之高頻電力之頻 率在疋為60MHz，輸出值係定為1 〇w/cm2。又，以電衆密 度為lxl01Q/cm3以上且5xl012/cm3以下，且蝕刻時之壓力 成為120Pa之方式，將混合氣體中所含之各種氣體以Sf6 氣體、〇2氣體、HBr氣體之順序分別以15〇sccni、15〇sccm、 30sccm之流量，供給至真空槽。 20 201203352 …後於上述條件下’實施3〇〇秒的姓刻，藉此獲得 實施例1之凹部H。使用掃猫型電子顯微鏡（SEM)拍攝 之實施例1的凹部Η之斷面影像係示於圖8 (a)。又，根 據圖8(a)之斷面影像所計測之凹部H的深度之最大值（最 大深度）及凹部Η的内徑之最大值（最大内徑）示於下。 •最大深度·· 154μπι •最大内徑：67μιη [比較例1] 以朗時之壓力成為25Pa之方式，變更混合氣體之轉 流買，並將其他條件定為與上述實施例丨相同，獲得比較 例1之凹部Η。使用掃目苗型電子顯微鏡（sem)拍攝 較例1的凹部Η之斷面影像係示於圖8⑴。又， 8⑴之斷面影像所制之凹部H的最大深度及 最大内徑示於下。 的 •最大深度：88μιη •最大内徑：68μιη [比較例2] ^㈣時之壓力成為25QPa之方式，變更混合氣體 C篁，並將其他條較為與上述實施例i相同， 較例2之凹部Η。使用掃㈤型電子顯微鏡（sem : 比較例2的凹部Η之斷面影像係示於圖8⑴。又，= ==像所計測之凹部Η的最大深度及‘ •最大深度·· 150μιη •最大内徑：71μιη 21 201203352 由此等結果觀察到，實施例1之最大深度係比較例1 之最大深度的2倍左右，且實施例丨之最大内徑與比較例 1之最大内經大致為同程度。由此最大深度之差異可知， 在可觀察到低能量波峰P3之壓力區域，相較於未觀察到其 之低壓區域’正離子更容易到達凹部Η之底面。又，根據 上述最大内徑之差異可知，在可觀察到低能量波峰Η之壓 力區域與未觀察到其之低壓區域中，到達凹部 離子係大致同程度。 U囟之止 六將比較例1之壓力改變為較50Pa更低之其他壓 行計測之結1相同，形狀進 深度之減少==伽5,低越多，則上述最大 例察到’實施例1之最大深度較比較 之最大内徑稍+ ’且實她例1之最大内徑較比較例2 觀察到低能量據此最大深度之差異可知，即便是可 失之高壓區域中，3之壓力區域，在雙峰波峰Βρ幾乎消 根據上述最大内正離子仍難以到達凹部Η之底面。又， 峰ρ3之壓力C可知，即便是可觀察到低能量波 中，此情死亦是雙峰波峰ΒΡ幾乎消失之高壓區域 另外，將比於達凹部Η側壁之正離子變多。 力，並將其他之壓力改變為高於l5GPa之其他壓 行計測之結果，=與比較例2相同，對以刻形狀進 大深度之減少一 高越多，則上述最 如以上所說明：二最大内徑之增加傾向越明顯。 據本實施形態之_方法及#刻裝 22 201203352 置’可獲得以下所列舉之效果。 (1)在尚頻電力之頻率為6〇MHz以上且150MHz以 下，且在離子能量之分佈中可觀察到雙峰波峰Bp、與強度 較其更向且具有低能量之低能量波峰P3的壓力區域中，實 施蝕刻。因此，上述第二正離子IPb在覆層Rs内新生成越 多，則形成於基板S之厚度方向的蝕刻形狀之異向性 提高。 (2) 又’由於亦進行具有相對高能量之上述第—正 子IPa之姓刻，可抑制因壓力過度提高反而使 異向性喪失的狀況。 欠之 (3) 由於以壓：hy· mu ίο f力在50Pa以上且15〇Pa以下，且曾將 密度為卜氣、上且5χ1〇1、3以下 = 刻，故可獲得雙夸波峰BP之強度相對於低能量波峰違订餘 強度的比率等之再現性。 $ (4) 作為蝕刻氣體，係使用含有含氟氣體之〜 之 P3 化硫氣體與含氯1及奴至少_者的祕氫氣體之 的六氟 漠化氫之浪合氣體1此，於基板s之厚度方向=:的 仙’係減料_進，絲成錄板s Μ部=之 係因鹵化矽之堆積物而不易被蝕刻。因此，可更容易二壁 得蝕刻形狀之異向性。 易地獲 另卜 述實知形態亦可如下述般適當改變而實 (5)餘刻氣體係除了六氟化硫氣體與漠化 外’亦含有氧氣。||此，當餘刻時，係形成相較於上从 化石夕、氟化石夕及*化石夕之任-者均為揮發性更c 石夕。因此，上述照射面之周壁更不易被㈣ 化 施 23 训2〇3352 •上述實施形態之電难 離子的 離子分佈中i足!範圍係只要為射入基板之 之範圍即可，當對低能量波H條件]及上述[第二條件] 強度的比率等不特 =Ρ3之強度與雙峰波峰ΒΡ之 種電漿密度之範圍。/再現性之情況，亦可捨去界定此 的離子分佈力範圍係只要為射人基板之離子 圍即可’根據敍刻氣二第種:條件]及上述[第二條件]之範 150pa以上。若為 種頬，例如亦可為50Pa以下或 類與蝕刻對象材料刻方法，則可擴大蝕刻氣體之種 3附近之離V能=之f子能4的分佈係由設於平台電極 可為使用單探6所取得。作為使用於其之方式， 等，只要為買里刀析之方式、使用發光分光之方式 並無特別限定7 、人至基u之正離子的能量之方式， :::二Ϊ用以取得離子能量分佈之裝置，以輪'i 裝置不同的裝置來双丈 /、傲幻 件]之>1力範圍的方式 _ —條件]及上述[第二條 施。二Ϊ :::法係使用平行平板型之银刻裝置而實 板載置亦可使用具有將作為姓刻對象物之基 頻電源的構成之其他侧裝置。 之间 【圖式簡單說明】 24 201203352 · 圖1係用於實施本發明一實施形態之蝕刻方法的蝕刻 裝置的概略構成圖。 圖2係表不而頻電力的頻數為40MHz時’到達基板的 離子的離子能量的分布圖。 圖3係表示高頻電力的頻數為60MHz時，到達基板的 離子的離子能量的分布圖。 圖4係表不尚頻電力的頻數為15 0MHz時’到達基板 的離子的離子能量的分布圖。 圖5係表示高頻電力的頻數為250MHz時，到達基板 的離子的離子能量的分布圖。 圖6係表示反應性離子蝕刻中正離子的執道的模式 圖。 圖7(a)(b)係表示藉由反應性離子而形成的蝕刻形狀的 模式圖。 圖8(a)〜⑷係表示以掃描型電子顯微鏡（SEM)拍攝 藉由蝕刻而於矽基板形成的凹部的影像。 【主要元件符號說明】 真空槽 2 ： 絕緣間隔物 平台電極 4 ： 兩頻電源 匹配箱 6 ： 離子能量分析器 喷灑板 8 ： 氣體供給部 排氣部 10 :控制部 S :基板 25In the fifth aspect of the invention, in the fifth aspect of the invention, the hydrogen-containing gas used in the gas containing the gas as the residual gas is used as the residual gas. At least one of ', 虱, > odor and iodine is promoted by hydrogen ions, and the thickness of the substrate is made by using the deposit of the tooth fossil eve, and the peripheral wall of the concave portion in the thickness direction is different in etching shape. Directional = etched. Therefore, it can be more easily obtained. [Embodiment] Hereinafter, the engraving method of the present invention will be described with reference to the engraving of the engravings 1 to 6 and the device of the engraving device of Fig. 1 will be embodied. Implementation form. An etching apparatus of a flat type or a capacity-applied embodiment is generally referred to as a parallel remnant apparatus: a schematic configuration of a type of remnant apparatus. The parallel plate-type platform electrode is passed through the insulating spacer 2 inside the vacuum chamber 1 to the plate S, and is supplied; The stage electrode 3 is provided with a matching box 5 on which the object to be etched is placed, and the high-frequency power source 4 of the S's power is transmitted through the stage electrode 3 for integrating the impedance platform electrode 3. The high-frequency power supply 4 sets the frequency of the supplied and A-frequency power to 6 〇 MHz or more and 201203352 150 MHz or less. The vicinity of the flat electrode 3 is configured to analyze a portion of the particle traveling toward the platform electrode 3 by an ion energy analyzer that analyzes the energy injected into the anti-t positive ion, and the positive ion of the grain 3 is electron-dependent. The temperature (ev) is classified, and the number of positive ions having respective electron temperatures is determined. Above the platform electrode 3, a spray plate 7 is provided which is opposite to the grounding energy and is provided with a gas guiding hole, and the gas introducing hole is supplied to the secret body from the outside of the vacuum tank 1 to Vacuum tank! =. The spray plate 7 is connected to a gas supply unit 8, which supplies various gases to be used for the substrate. In addition, the evacuation unit 9 is connected to the vacuum, and the various gases such as the gas or the atmosphere in the empty space are exhausted. In the present embodiment, the gas supply unit 8 is used to adjust the gas from the gas. The flow rate of the gas supplied to the vacuum tank i per unit time by the feeding portion 8 or the amount of exhaust gas per unit time of the exhaust portion 9 is adjusted by the exhaust portion 9, whereby the pressure in the two tanks 1 is adjusted to a predetermined value. (In the present embodiment, it is 箫a or more and 150 Pa or less). Alternatively, the pressure in the vacuum chamber 1 is adjusted to the predetermined value by the operation of the gas supply unit 8 and the exhaust unit 9. When etching is performed in the above-described parallel flat type etching apparatus, first, the substrate S to be the target is transferred into the vacuum chamber i, and placed on the stage electrode. Then, the atmosphere in the vacuum chamber 1 is discharged by the exhaust unit 9, and the gas is supplied from the gas supply unit 8 to the inside of the vacuum chamber 1 at a predetermined flow rate. At this time, the etching gas supplied from the gas supply unit 8 is supplied. The flow rate and the amount of exhaust gas from the exhaust unit 9 are adjusted to a predetermined pressure in the vacuum chamber i after the supply of the etching gas, and the supply of the high-frequency electrode from the high-frequency power source 4 to the platform power 10 201203352 pole 3 is performed. Then, due to the discharge between the substrate S placed on the platform electrode 3 and the spray plate 7, a plasma of an etching gas is generated. In addition, the exhaust gas treatment by the exhaust unit 9, the gas supply process by the gas supply unit 8, and the power supply process by the high-frequency power source 4, etc., are performed by various processes performed by the device. The control unit 10 of the etching apparatus drives each part and performs this. At the time of plasma formation, electrons in the plasma collide on the surface of the substrate S and the vacuum chamber 1, whereby the surfaces are biased negative with respect to the plasma system, and the surface is formed substantially identically with a predetermined Thickness of the coating. Further, if the positive ions in the plasma reach the boundary between the cladding and the plasma other than the cladding (the upper end of the cladding), the positive ions are introduced into the surface of the substrate S which is biased negative. Thus, the positive ions introduced to the surface of the substrate S chemically or physically etch the substrate S, whereby the etching accompanying the predetermined etching shape is performed from the surface of the substrate S in the thickness direction of the substrate S. Further, a concave portion extending from the surface of the substrate S in the thickness direction of the substrate S is formed. Further, in the present embodiment, the substrate S is, for example, a tantalum substrate, and a mixture of hexafluoride (SF6) gas, oxygen (02) gas, and hydrogen bromide (HBr) gas is used as the etching gas. gas. However, the etching gas is not limited to such a combination, and any gas obtained by mixing a fluorine-containing gas and a hydrogen halide gas having at least one of chlorine, bromine and iodine may be used as the etching gas. More specifically, as the fluorine-containing gas constituting the etching gas, for example, sulfur hexafluoride gas, iodine pentafluoride (IF3) gas, trifluorinated gas (C1F3) gas, boron trifluoride (BF3) gas, or the like may be used. At least one of a sulphur sulphur (SOF2) gas, a sulphur fluorinated fluorine (S02F2) gas, and a fluorinated carbonyl (COF2) gas. Further, as the hydrogen halide gas, for example, ruthenium chloride; ϊ 11 201203352 hydrogenation (HCl) gas, and at least one of them may be used. In the hydrogen iodide (Hi) gas, in the present embodiment, as the inclusion body used for the etching, a mixed gas of hydrogen bromide gas of the substrate substrate is used. The etching of the burial, sulfur fluoride gas, and _hydrogen gas is promoted by forming the volatility in the thickness direction of the substrate S. Further, the hydrogen hydride of the ruthenium fluoride which is higher in the substrate s yttrium fluoride is not thickened by the deposition of the cleavage, and the peripheral wall of the concave portion is etched and etched by other gas species. Therefore, it is compared to the anisotropy of the shape. In the case of the crucible, it is easy to obtain an etched shape. In addition, in the present embodiment, in addition to the etch and hydrogen halide gas, the peripheral wall of the concave portion other than the fluorine-containing gas becomes less likely to be elongated in the thickness direction of the substrate 8. -, 佳夕τm2 In the above reactive ion etching, the lamp S system reaches the surface of the substrate S without colliding with other particles in the coating. In this case, if the frequency band of the high-frequency power is above 6GMHZ and below 15 〇MHz, the positive ion system follows the high-frequency power. As a result, when the positive ions reach the substrate s, the ^= generation corresponding to the high-power is generated. The positive ions accelerated by a large electric field and the positive ions are not accelerated by a large electric field. &, the energy difference between these positive ions. Therefore, the 1 wave (:: gastric bimodal peak) formed by the acceleration of the mutually different electric field determined by the frequency of the high power is found in the energy distribution of the positive ion on the high energy side. 12 201203352 The peak and peak peaks are in the case of a superior wave s in the energy distribution of positive ions, and most of the positive ions travel from the upper end portion formed on the substrate = two = to the surface of the substrate s. Thus, in addition to the positive ions that travel in the normal direction, the positive ions in the same direction as the line direction; = 3: The positive ions are thus traveled, and if the concave portion is formed slowly, the more difficult it is to reach the bottom surface of the concave portion. As a result, while the traveling positive ions increase the depth of the concave portion, they proceed to the opening of the concave portion of the positive ion system in the direction of the ΙΓ 士 士. Because: View =: the superior peak in the energy distribution of the ion = ', ^' concave 0卩 / the greater the degree of clothing, the greater the amount of etching of the sidewall of the recess. As a result, the so-called CD loss becomes large. Therefore, the first peak of the energy of the positive ion in the energy distribution of the present embodiment is more excellent than the above-mentioned bimodal peak of the first peak, and particularly the pressure condition is used to carry out the engraving. Next, the dependence of the frequency of the 咼 frequency power and the pressure in the vacuum chamber 1 with respect to the energy distribution of the positive ions introduced into the substrate s will be described with reference to Figs. 2 to 5 as an example. Fig. 2 to Fig. 5 are energy distributions of positive ions in the following residual conditions, and sequentially exemplify the energy distribution when the frequency of the high-frequency power is set to 40 MHz, 60 MHz, 150 MHz, and 250 MHz. (etching conditions) Etching gas: a mixed gas composed of SF6 gas, helium 2 gas, and HBr gas. The output value of the high-frequency power of the stage electrode 3 is i〇w/cm2. The pressure in the vacuum chamber 1 is 0.2 Pa, 25 Pa, 50 Pa, 150 Pa, 250 Pa. Η 13 201203352 Plasma density when the button is engraved: 1 x 1 〇l 1. Further, the ion energy distribution shown in Figs. 2 to 5 is obtained from the output value of the wheel 4 of the ion energy component (4) provided in the upper secret device. Further, in the respective figures, the ion energy in the vacuum chamber is G.2l>a, and the ion energy is shown by the thin dotted line at 25 Pa. In addition, when the pressure in the vacuum chamber 1 is 50 Pa, the ion energy is expressed by a two-point link line, and the secret force is the ion material of the enamel--point link line, and the ion energy at the pressure of 250 Pa is The teaching line said. As shown in Fig. 2, when the frequency of the high-frequency power is 4 〇 MHz: in the case where the pressure is 〇. 2 Pa and the pressure is 25 Pa, the bimodal peak BP is observed at the end of the high energy side. Is it caused by the first-direction energy peak observed at about 14_? 1 is composed of a second high energy peak P2 observed at about 65 eV. On the other hand, in the case where the history is a high condition (cis, 150 Pa, 250 Pa), the bimodal peak Bp is not observed. Especially in the condition of a pressure of 2, almost no positive ions having energy of 65 to 140 eV surrounded by the bimodal peak BP were observed. As shown in Fig. 3, when the frequency of the high-frequency power is 6 〇 MHz 2 and the pressure is 250 Pa (〇. 2 Pa, 25 Pa, N, qing qing a), a bimodal peak is observed at one end of the high energy side. Bp, the first high energy peak observed by the bean 125eV, consists of the second high energy peak P2 observed at about the same peak pl and at about 80 eV. In addition, in the condition that the pressure is 50 Pa and the pressure is 15 〇pa on the lower energy side of the bimodal peak Bp of the rut, the low energy peak of the second peak with higher intensity BP double peak is not low. The positive ion of the energy peak P3 has a lower energy than the energy given to 201203352 by the electric field acceleration. Therefore, the positive ion system having the low energy peak p3 can be attributed to the positive ions generated in the coating and accelerated by the electric field from the towel in the coating, for example, due to the charge exchange between the neutral particles and the positive ions. Positive ion. Further, the above-described tendency that the low-energy peak P3 having a higher intensity than the bimodal peak Bp is observed on the low-energy side of the bimodal peak is observed in the entire range of 5 〇 Pa or more and 150 Pa or less. The tendency to get. As shown in Fig. 4, when the frequency is 15 〇 MHz, it is the same as the above 60 MHz, and the condition is other than the pressure of 25 〇Pa (〇2ρ&, 25ρ&, 5〇pa, 150Pa). A bimodal peak BP is observed at one end of the energy side, which is composed of a second high energy peak p2 observed at about 9 〇 eV observed at about n〇eV. Also, ^ at the above 60MHz _ 'the condition of the pressure of 5 〇 Pa and the condition of the pressure ΐ 5 〇 ρ & is lower than the bimodal wave front BP, the energy is higher than the bimodal peak BP The low energy peak of the second peak of the intensity. Further, at the above 60 MHz, the low energy peak p3 is observed only under the respective pressure conditions of 5 〇 ρ & and ΐ 5 〇 ρ & and, in contrast, at i5 〇 MHz, in the pressure conditions of 50 Pa and 150 Pa. Three low energy peaks were observed. Further, similarly to the above-described 60 MHz, the low energy peak Ρ3 having a higher intensity than the bimodal peak δρ is observed on the low energy side of the bimodal peak Βρ, and is inclined at 5 〇 Pa or more and 15 〇. The tendency observed in the whole circumference below Pa. Wang Wei, as shown in Fig. 5, when the frequency is 25 〇MHz, the same as the above, the condition and pressure of the pressure of 〇.2Pa are "in the condition that the bimodal peak Bp is observed at one end of the high energy side, This is observed from the first high energy peak ρι observed in Wei 15 201203352, the second high energy peak 卩2 & @, about 5eV, at the injection to base = sub = bimodal peak BP. The above-mentioned double peak BP and the gas-dissipation distribution towel were observed to have a high-energy low-energy peak pj::, a lower energy region and a circle below HZ, and vacuum; the frequency band is 嶋z to the following m Buben real _ material, on the 15 〇 Pa Shi Shuai, the frequency band system is determined in the second description of the money engraving device, and is determined to meet the following [first condition = = along the circumference. For example, in In the above-described buttoning condition, the dust force force of the first condition] is not more than 15 〇 _z, and the pressure _z is 150 Pa or less. τ is 〇5 〇Pa or more and [first condition] bimodal peak 3 卩 &曰[Second condition] Observed shame is observed as a double peak. The lower energy wave than the bimodal peak BP exaggerates the lower energy side, and The method of adjusting the room force of the space of the electricity collection to a predetermined value is known as a method of adjusting the (10) enthalpy of the rhythmic gas, and the gas volume of the gas produced by the surface of the substrate S can be adjusted, and the gas is used. The flow rate is adjusted to adjust the dust force H point, preferably = and control the rhythm of the moment of the reaction, and the electropolymerization in the plasma is defined by the degree. In general, it is known that 201203352 = degree If the height is higher, the thinner the coating is, the thicker the coating is formed. Therefore, not only the low furnace, but also the above [苐~ condition] and [second condition], two ®...* 3 The ratio of the intensity to the intensity of the bimodal waveguide BP, etc., a case of 'the circumference' is preferably defined as the range of the electric density and the dust force 3 =. Therefore, in the present embodiment, the flow rate of the etching gas is 3 The output value is set such that the pressure is 50 Pa or more and 150 Pa or less, and the plasma density is lxl 〇 1 〇 / cm 3 or more and 5 xi 〇 lw or less. Exhaustive 3 ' With reference to FIG. 6 and FIG. 7, the frequency band determined as described above will be described. The shape of the side of the domain. Fig. 6 shows the positive displacement near the surface of the substrate S when the above reactivity is carried out. Further, Fig. 7 is a schematic view showing a side shape formed by reactive ions. As shown in Fig. 6, after the substrate 3 is housed in the above-described cutting device, the supply of the gas to the vacuum chamber 1 is The supply of the high-frequency current to the platform electrode 3 is sequentially performed, whereby the (4) Rp containing the electron E and the positive ion IP is generated from the money engraved gas composed of the neutral particles Np. Then, the frequency current is followed. The (4) faster electrons generated by the electric field reach the surface of the substrate S, whereby the surface of the substrate s is biased negative relative to the electropolymer. Thus, on the surface of the substrate S, the electron E is inverted. Repulsively bounces off the side of the surface of the substrate s, so that a predetermined layer of silicon from the surface of the substrate s forms a coating Rs, which is a region in which the number of positive ions ip is larger than the number of electrons E. Then, in the region where the coating layer Rs is further away from the substrate s, there are positive ion IP, electron E, and neutral particle NP, and it becomes a region of the plasma Pr which is electrically neutral. If the positive ion Ip in the plasma Rp reaches the boundary surface BP of the plasma and the coating Rs, the positive ion ip of the positive potential is introduced into the surface of the substrate S which is negatively pressed, in other words, the positive ion IP system proceeds to The surface of the substrate $ 17 201203352. Here, the positive ions IP incident on the surface of the substrate S are roughly classified into the following two. The first positive ion IPa of the surface of the substrate S does not collide with other particles. • In the cladding Rs, the second positive ion 11 > 15 newly formed by the neutral particle NP is exchanged by the neutral particle NP and the positive ion Ip. The first positive ion IPa receives the electric field acceleration from the high-frequency power, and maintains the energy obtained by the acceleration until it enters the substrate S. The first positive ion Ipa is also a positive ion of BP which is formed by the previous energy peak P1 and the second high energy peak p2 shown in FIGS. 2 to 5 . As shown in Fig. 6, the first positive cation IPa is a first positive ion IPa in the normal direction of the surface of the substrate s and a first positive ion IPa in the direction different from the normal direction. Further, the first positive ion IPa equal to the normal direction and the first positive ion IPa traveling in the direction opposite to the normal direction = do not collide with the other positive ions Ip or the middle identical NP to reach the surface of the substrate s . However, the first positive ion ipa which travels in a direction different from the normal direction in the first off-board, the greater the depth formed in the concave portion of the substrate s, the harder it is to reach the concave portion, and the second The positive ion IPb is a positive ion whose electric field acceleration energy is smaller than the first positive ion IPa, that is, a positive ion which is lower than the low energy peak p3 shown in FIG. 3 and FIG. 4 previously. ^Two genus The di-negative IPb system, like the first positive ion IPa, includes a second positive ion ipb that is in the normal direction and travels with the above-mentioned method green = 201203352 + 篦 疋 ion. The positive ions generated in the cladding layer are the same as the first positive ions IPa when the electrons are accelerated by the electric enthalpy, but the position at which the electric field is accelerated is closer to the substrate than the first positive ions IPa. That is, the distance traveled by the second positive ion IPb is the distance Db from the charge exchange proceeding point, and the fact that the charge exchange system occurs in the cladding layer Rs is that the distance Db is more than the above-mentioned distance corresponding to the thickness of the cladding layer Rs. Da is smaller. Therefore, the second positive ion IPb generated by the charge exchange is the same as the first positive ion ipa, and even if the traveling direction deviates from the normal direction, the reaching distance of the substrate s is only the first positive distance. The shorter the ion IPa is, the easier it is to reach the bottom surface of the concave portion formed on the substrate S. The etching shapes formed by reactive ion etching using the first positive ions IPa and the second positive ions lpb are schematically shown in Figs. 7(a) and (b), respectively. Further, as shown in FIG. 7, a predetermined etching shape (for example, formation of the concave portion H) is formed on the substrate S by reactive ion etching, and a pre-turn opening portion is formed on the surface of the substrate s before the reactive ion etching. Mask M. Then, the positive ions IP are incident on the substrate s from the opening, and the substrate s is etched. As shown in FIG. 7(a), when the etched second is performed by the first positive ion IPa, the sub-portion 1Pa' is moved toward the opening of the mask M by the first direction traveling in a direction different from the normal direction. The normal direction of Ma is etched on the outside. Thereby, the maximum diameter 〇丨 of the concave portion H of the etched shape is larger than that of the opening portion Ma. In the aspect of lpb, as shown in Fig. 7 (b), when the second positive ion yoke is used, the second positive ion Ipb is straight from the substrate s attached 19 201203352 The twin system is roughly equal to the opening and is accelerated by the electric field, so it is easy to etch along the mask door. Therefore, the diameter DiaA of the concave portion J of the shape J is etched. Therefore, the proportion of the positive ions Ip incident on the substrate s increases, that is, as in the previous, the second positive ion iPb sub-energy distribution, by satisfying the above conditions, By performing etching, the anisotropy of the concave portion H 1 ] and the [second condition] 1 η can be improved. As shown in Fig. 3 and Fig. 4, the low energy peak ρ ^ - positive ion IPb ' is lower than the bimodal peak, ie, the ion IPa, so if there is only the second positive Ion muscle = implementation of the residual 'there is a side velocity reduction 虞 1 at this point, according to this = application form, is the first positive ion ipa ^ t low energy material P3 belonging to the double bee peak Bp The red strip can be used to maintain the anisotropy of the shape of the surname through the second positive ion IPb, while suppressing the decrease in the etching rate via the first positive ion IPa. [Example 1] After a mask having an opening of 5 μm in diameter was applied to an 8-inch substrate having a thickness of 750 μm, the parallel plate type etching apparatus was used, and SF6 gas, helium gas, and HBr gas were used. The mixed gas of the composition is etched as an etching gas. At this time, the frequency of the high-frequency power output from the high-frequency power source is 60 60 MHz, and the output value is set to 1 〇 w/cm 2 . In addition, in the order of the electric power density of lxl01Q/cm3 or more and 5xl012/cm3 or less, and the pressure at the time of etching becomes 120 Pa, the various gases contained in the mixed gas are respectively in the order of Sf6 gas, helium gas, and HBr gas. A flow rate of 15 〇sccni, 15 〇 sccm, and 30 sccm was supplied to the vacuum chamber. 20 201203352 ... After the above conditions were carried out, the last minute of 3 seconds was carried out, whereby the concave portion H of Example 1 was obtained. The cross-sectional image of the concave portion of Example 1 taken using a scanning cat type electron microscope (SEM) is shown in Fig. 8(a). Further, the maximum value (maximum depth) of the depth of the concave portion H measured based on the cross-sectional image of Fig. 8(a) and the maximum value (maximum inner diameter) of the inner diameter of the concave portion 示 are shown below. • Maximum depth·· 154 μm • Maximum inner diameter: 67 μm η [Comparative Example 1] The flow of the mixed gas was changed so that the pressure at Langshi became 25 Pa, and other conditions were determined to be the same as those in the above Example, and comparison was made. The concave portion of Example 1. The image of the cross section of the concave portion of Comparative Example 1 was taken using a scanning electron microscope (Sem) as shown in Fig. 8 (1). Further, the maximum depth and the maximum inner diameter of the concave portion H formed by the sectional image of 8 (1) are shown below. • Maximum depth: 88 μm η • Maximum inner diameter: 68 μm η [Comparative Example 2] ^ (4) When the pressure is 25 QPa, the mixed gas C 变更 is changed, and the other strips are the same as those of the above embodiment i, and the recess of the example 2 is Hey. A scanning electron microscope (sem: the cross-sectional image of the concave portion of Comparative Example 2 is shown in Fig. 8 (1). Further, = = = the maximum depth of the concave portion 计 as measured and ' • maximum depth · 150 μιη • maximum within Diameter: 71 μιη 21 201203352 It was observed from the results that the maximum depth of Example 1 was about twice the maximum depth of Comparative Example 1, and the maximum inner diameter of Example 大致 was approximately the same as the maximum internal diameter of Comparative Example 1. From this difference in maximum depth, it can be seen that in the pressure region where the low energy peak P3 can be observed, the positive ions are more likely to reach the bottom surface of the concave portion than the undetected low pressure region. Further, according to the above maximum inner diameter The difference is that, in the pressure region where the low energy peak 可 can be observed and the low pressure region where the low energy peak is not observed, the ion system reaching the concave portion is approximately the same degree. The pressure of the comparative example 1 is changed to be lower than 50 Pa. The other pressure measurement has the same knot 1 and the reduction of the shape depth == gamma 5, the more the lower, the above-mentioned maximum example sees that the maximum depth of the first embodiment is slightly smaller than the comparison of the maximum inner diameter of the case. The largest within Compared with Comparative Example 2, the difference between the maximum energy and the maximum depth is observed. Even in the high-pressure region that can be lost, the pressure region of 3 is hard to reach the bottom surface of the concave portion according to the maximum internal positive ion. Further, the pressure C of the peak ρ3 shows that even in the low-energy wave, the death is a high-pressure region in which the bimodal peak ΒΡ almost disappears, and the positive ions are more than the positive ions in the sidewall of the concave portion. And change the other pressures to other pressure measurement results higher than l5GPa, = same as in Comparative Example 2, the more the reduction in the depth of the engraved shape is higher, the above is most explained above: The tendency of the increase in the inner diameter is more pronounced. According to the method of the present embodiment and the method of #刻刻22 201203352, the following effects can be obtained. (1) The frequency of the frequency power is 6 〇 MHz or more and 150 MHz or less, and The etching is performed in a pressure region in which the bimodal peak Bp and the low energy peak P3 having a higher intensity and lower energy are observed in the distribution of the ion energy. Therefore, the second positive ion IPb is in the coating Rs. The more the new generation is, the more the anisotropy of the etching shape formed in the thickness direction of the substrate S is improved. (2) In addition, since the surname of the first positive sub-IPa having a relatively high energy is also performed, the excessive pressure increase can be suppressed. Conversely, the situation of the loss of anisotropy. (3) Because of the pressure: hy· mu ίο f force is above 50Pa and below 15〇Pa, and the density is b, above and 5χ1〇1, 3 or less = The reproducibility of the ratio of the intensity of the double-bend peak BP to the low-energy peak-to-order margin is obtained. $ (4) As the etching gas, the P3 sulfur gas containing the fluorine-containing gas is used and Chlorine 1 and the slave's hydrogen gas, the hexafluoro-hydrogenated hydrogen gas, the gas in the thickness direction of the substrate s =: the fairy 'system cut _ into, the silk into the board s Μ part = It is not easily etched due to the deposit of antimony halide. Therefore, it is easier to etch the anisotropy of the shape on both walls. It is also possible to change the actual form as described below. (5) The residual gas system contains oxygen in addition to sulfur hexafluoride gas and desertification. ||This, when the moment is left, is formed to be more volatile than the upper ones from the fossils, the fluorites, and the fossils. Therefore, the peripheral wall of the above-mentioned irradiation surface is less likely to be subjected to (4) the application of 23 〇 2 〇 3352. The ion distribution of the electro-difficulous ions of the above embodiment is in the range of the range of the substrate, and the low energy The ratio of the wave H condition] and the above [second condition] intensity is not particularly limited to the range of the intensity of the Ρ3 and the plasma density of the bimodal peak ΒΡ. In the case of reproducibility, it is also possible to define the range of ion distribution force as long as it is the ionization of the substrate, and the above-mentioned [second condition] is 150 Pa or more. . If it is a seed, for example, it may be 50 Pa or less or a method of etching the material to be etched, the distribution of the energy of the energy of the vicinity of the species 3 of the etching gas 3 can be expanded. Single exploration 6 was obtained. The method used for the method, etc., is not limited to the method of illuminating and splitting, and the method of using the illuminating spectroscopy is not particularly limited. The device for energy distribution, in the case of a different device of the wheel 'i device, the mode of the force range] __condition] and the above [second article. The second::: method uses a parallel plate type silver engraving device, and the solid plate mounting can also use other side devices having a configuration of a fundamental power source to be an object of surname. [Brief Description of the Drawings] 24 201203352 Fig. 1 is a schematic configuration diagram of an etching apparatus for carrying out an etching method according to an embodiment of the present invention. Fig. 2 is a diagram showing the distribution of ion energy of ions reaching the substrate when the frequency of the uninterruptible power is 40 MHz. Fig. 3 is a view showing the distribution of ion energy of ions reaching the substrate when the frequency of the high-frequency power is 60 MHz. Fig. 4 is a graph showing the distribution of ion energy of ions reaching the substrate when the frequency of the unused power is 150 MHz. Fig. 5 is a view showing the distribution of ion energy of ions reaching the substrate when the frequency of the high-frequency power is 250 MHz. Fig. 6 is a schematic view showing the behavior of positive ions in reactive ion etching. Fig. 7 (a) and (b) are schematic views showing an etching shape formed by reactive ions. Figs. 8(a) to (4) show images of a concave portion formed on a ruthenium substrate by etching using a scanning electron microscope (SEM). [Main component symbol description] Vacuum chamber 2 : Insulation spacer Platform electrode 4 : Two-frequency power supply Matching box 6 : Ion energy analyzer Spray plate 8 : Gas supply unit Exhaust unit 10 : Control unit S : Substrate 25