1275155 九、發明說明： 【發明所屬之技術領域】 本發明是有關於一種檢測半導體摻雜過程的方法，特 別是指一種以高頻技術檢測半導體摻雜過程的方法。 【先前技術】 在半導體材料中，藉由額外摻雜其他金屬元素便可控 制半導體中的電性、磁性及光學性質，是半導體工業發展 传以如此豐富的主要原因之一。然而，推雜元素能否有效 取代半導體中的原子，是控制半導體性質上一項重要的關 鍵。 由於70件尺度的日益减小。成功推雜與否變得越來越 難檢測，但摻雜不完全所形成的金屬奈米團鎮對材料性質 的影響卻越來越大，也影響該半導體材料之品質優劣。、 現有技術在檢測摻雜過程中是否有金屬奈米團鎮形成 在半導體間最常被使用的技術便是穿透式電子顯微鏡（ tra_lssion electr〇n micr〇sc〇py ’簡稱 tem)，雖缺穿透 電子顯微鏡可以直接檢測出奈米團蔟的微結構，作是此: 需要相當繁複的樣品準備過程，而日所兹, 疋此法 a 而且所觀測的影像也只是 局#的結構並非是呈現整個半導體材料的品f。因此，* 透式電子顯微鏡相當費時且所觀察到的結果並不= 於整個半導體材料的品質。 ° 近來也有其他研究提出—1其 園铭A > α ， 乃對牛導體内奈米 團联進仃檢測，例如同步輻射 Μ . 次收光邊技術、超導量子千 〜儀低溫磁化量測，或低溫電 电得V里測等方法等，上述之 1275155 方法有些需要經過高強度的同步輻射光源技術或液氦的低 溫系統才能檢測出金屬奈米團簇是否存在於半導體中，不 僅程序複雜，檢測之成本也較高。因此仍然不能夠提供一 個簡便、快速又可靠的檢測方法。 【發明内容】 W牡扠供一種用於檢測半導 内是否有金屬奈米團簇析出的方法，α簡便、快速又可；1275155 IX. Description of the Invention: [Technical Field] The present invention relates to a method of detecting a doping process of a semiconductor, and more particularly to a method of detecting a doping process of a semiconductor by a high frequency technique. [Prior Art] In semiconductor materials, the electrical, magnetic and optical properties of semiconductors can be controlled by additionally doping other metal elements, which is one of the main reasons for the development of the semiconductor industry. However, whether the doping element can effectively replace the atoms in the semiconductor is an important key to controlling the nature of the semiconductor. Due to the decreasing size of 70 pieces. It is becoming more and more difficult to detect whether the hybridization is difficult or not, but the influence of the metal nano-town formed by incomplete doping on the material properties is increasing, which also affects the quality of the semiconductor material. The prior art technique for detecting the presence or absence of a metal nanoparticle in the doping process is the most commonly used technique among semiconductors. It is a transmission electron microscope (tra_lssion electr〇n micr〇sc〇py 'referred to as tem), although it is lacking. Penetrating electron microscopy can directly detect the microstructure of nano-cluster, which is the result: a fairly complicated sample preparation process is required, and the method is abundance, and the observed image is only the structure of the bureau#. The whole semiconductor material f. Therefore, *transmissive electron microscopy is quite time consuming and the observed results are not = the quality of the entire semiconductor material. ° Recently, other studies have proposed -1, its gardening A > α, is the detection of the inner conductor of the bovine conductor, such as synchrotron radiation. Sub-lighting edge technology, superconducting quantum kilometer low temperature magnetization measurement , or low-temperature electric power V-measuring method, etc., some of the above-mentioned 1275155 methods need to pass through high-intensity synchrotron radiation source technology or liquid helium cryogenic system to detect whether metal nanoclusters exist in the semiconductor, not only complicated procedures The cost of testing is also higher. Therefore, it is still not possible to provide a simple, fast and reliable detection method. SUMMARY OF THE INVENTION W-frying fork is a method for detecting the presence or absence of precipitation of metal nano-clusters in a semi-conductor, and α is simple, rapid, and can be used;
地檢測半導體摻雜製程之品質優劣。 於是，本發明半導體摻雜製程的檢測方法，適用於檢 測一半導體於摻雜過程中是 法包含以下步驟： 有金屬團族析出，該檢測方 抗頻以—高頻複數阻抗分析儀量測出所述半導體之阻 ⑻取阻抗頻譜之實部的對數為橫軸，虛部 縱軸緣圖並求取特徵斜率。 ί數為 (C)如該特徵斜率為〇 5， 種介電遲豫行為產生，亦即半 ι體中僅有一 的析出。 導體摻雜過程中無金屬團簇 【實施方式】 有關本發明之前述及其 C2 Τ斯人会土㈤ 技術内谷、特點盘功柃 , =配合參考圖式之—個較佳實施例明力心在 清楚的呈現。 干、、田°兒明中，將可 參閱圖1，本發明半導體摻 施例適用於檢測—半導 〜、^叫測方法之較佳實 牛^體1貫 疋古有金屬團簇 1275155 析出，所述半導體1上具有 有複數電極11，該半導體摻雜製 私的檢測方法包含以下步驟·· ⑷以-高頻複數阻抗分析儀2量測出半導體】之阻 =員：。本實施例中該高頻複數阻抗分析儀2是型號 P4294A之阻抗量測計，作實 # 1一耳她上不以此為限。該高頻複 數阻抗分析儀2具有二量測線 ㈣1測線21分別電連 妾μ半V體之電極11，以量測所述丰 W述牛導體1之阻抗頻譜，The quality of the semiconductor doping process is tested. Therefore, the method for detecting a semiconductor doping process of the present invention is suitable for detecting a semiconductor in a doping process. The method comprises the following steps:: metal group precipitation, the detection frequency is measured by a high frequency complex impedance analyzer The resistance of the semiconductor (8) takes the logarithm of the real part of the impedance spectrum as the horizontal axis, the imaginary part of the vertical axis, and obtains the characteristic slope. The ί number is (C) If the characteristic slope is 〇 5, the dielectric hesitation behavior occurs, that is, only one of the half bodies is precipitated. Metal-free clusters in the doping process of conductors [Embodiment] The foregoing and related C2 of the present invention (5) technology valley, characteristic disk work, = a preferred embodiment of the reference pattern The heart is clearly presented. In the dry, the field, the children's Ming, will refer to Figure 1, the semiconductor admixture of the present invention is suitable for the detection - semi-conducting ~, ^ called the measurement method of the best real bovine body 1 through the ancient metal cluster 1275155 precipitation The semiconductor 1 has a plurality of electrodes 11 thereon. The semiconductor doping method includes the following steps: (4) measuring the semiconductor by the high-frequency complex impedance analyzer 2; In this embodiment, the high-frequency complex impedance analyzer 2 is an impedance measuring instrument of the model P4294A, and it is not limited to this. The high-frequency complex impedance analyzer 2 has two measuring lines (4), and the measuring line 21 is electrically connected to the electrode 11 of the 半μ half-V body, respectively, to measure the impedance spectrum of the ferro-conductor 1
例中疋以具有電極11之半導體1為例作說明，但 實際實施時’只要量測線21可與所述半導體i電連接即可 ’實施範圍不以半導體！上是否具有電㈣為限。 ⑻將量測出之阻抗頻譜的阻抗值實部為橫軸，阻抗 ^之虛部為縱歸^Gle_C()le圖。便可發現其賴概略是 由一個半圓弧線所組成的。 (C)取阻抗頻譜之實部的對數為橫軸，虛部之對 縱軸繪製圖並求取特徵斜率。 …In the example, the semiconductor 1 having the electrode 11 is taken as an example, but in actual practice, as long as the measuring line 21 can be electrically connected to the semiconductor i, the implementation range is not semiconductor! Whether there is electricity (4) on the limit. (8) The real part of the impedance value of the impedance spectrum measured is the horizontal axis, and the imaginary part of the impedance ^ is the vertical ^Gle_C()le diagram. It can be found that the outline is composed of a semi-circular arc. (C) Take the logarithm of the real part of the impedance spectrum as the horizontal axis, and plot the vertical axis of the imaginary part and find the characteristic slope. ...
、⑼若將兩軸取對數，其半圓弧線取對數下的特徵斜 率為〇.5’代表所述半導體！中僅有—種介電遲豫行為產生 ，亦即所述半導體1摻雜過程中無金相蔡的形成。當斜率 偏離〇.5越多日夺可視為半導體中有二組或二組以上的遲豫 订為呈相對比例地增加。此額外的遲豫行為便是來自於取 代不完全的金屬團蔟與半導體^形成的介面所產生的社果 以下用-實驗例以說明本發明半導體摻雜過程之檢判 方法且以下所提及的參閱圖式皆須同時配合參圖丨。本實驗 1275155 例中是以一實驗組樣品及一 于…、、、且樣品進行檢測並予以比 較。實驗之前，先說明樣品製程與特性。樣品製程是利用 離子束麟系統在-塊氧化紹基板上成長5Gnm氧化辞換雜 5%-CoFe的半導體1為實驗έ 苴4 之樣品，以及Μ—塊氧化鋁 基板上成長50nm氧化鋅換雜1n()/ p p t 私拜I雜10%_CoFe的半導體1為對昭 組之樣品。(9) If the two axes are taken as a logarithm, the characteristic slope of the semi-circular line is logarithmically 〇.5' represents the semiconductor! There is only one kind of dielectric hesitation behavior, that is, there is no formation of metallographic Cai in the doping process of the semiconductor 1. When the slope deviates from 〇.5, the more the daytime is regarded as the delay of two or more groups in the semiconductor, the order is increased in a relative proportion. This additional hesitation behavior is derived from the substitution of the incomplete metal group and the interface formed by the semiconductor. The experimental example is used to illustrate the method for judging the semiconductor doping process of the present invention and is mentioned below. The reference drawings must be accompanied by a reference map. In the experiment 1275155, an experimental group sample and a sample were tested and compared. Before the experiment, the sample process and characteristics will be explained. The sample process is to use the ion beam lining system to grow 5Gnm oxidized 5%-CoFe semiconductor 1 as the experimental έ 之4 sample, and the 50 nm zinc oxide mixed on the Μ-block alumina substrate. 1n()/ppt A semiconductor 1 with a 10%_CoFe is a sample of the group.
、接者，利用-超導量子干涉儀量（圖未示）測樣品的 磁性質。由於在磁性半導體系統中，其磁性來源主要是來 自於過渡金屬取代半導體丨晶格中之原子形成磁性半導體工 化合物所引致的本質性的磁性，或者是由於取代不完全所 形成的磁性團簇所貢獻的外f磁性。兩者在作為自旋電子 元件的設計應用上完全不同，因此檢測材料中有無磁性團 簇的形成就可了解摻雜過程中是否有金屬團簇析出。圖2是 氧化辞摻雜5%_CoFe之實驗組樣品以及氧化鋅摻雜^ CoFe的對照組樣品於3〇〇κ所量測的磁滯曲線圖，兩樣。 在至溫皆觀測到鐵磁的磁滯曲線，顯現類似的室溫鐵磁性 行為，表示實驗組及對照阻之樣品都是磁性半導體玉。 之後，利用本發明半導體摻雜製程的檢測方法針對實 驗組及對照阻之樣品檢測其中是否有金屬團簇的析出、、 高頻複數阻抗分析儀刪麗利用兩點量測方式，以量二 實驗阻與對照組之樣品的的阻抗頻譜。其中量測的震盈電 壓為500mV ’外加額外直流偏壓〇〜i.5v，4 顆率範圍為 500Hz 至 110MHz 〇 再將兩樣品之頻譜的阻抗實部對阻抗虛部綠製出如圖3 1275*155 所=之Cole-Cole圖，便可發現實驗阻之頻譜弧線是概略呈 半圓的圓弧曲線’而對照組之圓弧曲線除高頻區之半圓 弧線外，於低頻時有—隨頻率增加而下垂之短弧線。 參閱圖4,若將頻譜之阻抗實部取對數為橫軸，對阻抗 虛部取對數為縱軸’並求取特徵斜率。實驗組樣品的半圓The receiver, the magnetic quantity of the sample is measured by the amount of the superconducting quantum interferometer (not shown). In magnetic semiconductor systems, the magnetic source is mainly derived from the essential magnetic properties caused by the transition metal substitution of atoms in the semiconductor germanium lattice to form a magnetic semiconductor compound, or the magnetic cluster formed by incomplete substitution. Contribution of external f magnetic. Both are completely different in the design and application as spintronic components. Therefore, it is possible to know whether or not metal clusters are precipitated during the doping process by detecting the presence or absence of magnetic clusters in the material. Fig. 2 is a hysteresis curve of the experimental group sample of the oxidized word doped 5%_CoFe and the control sample of the zinc oxide doped CoFe at 3 〇〇 κ, two. The hysteresis curve of ferromagnetic was observed at the temperature, showing similar room temperature ferromagnetic behavior, indicating that the samples of the experimental group and the control block were magnetic semiconductor jade. Then, using the detection method of the semiconductor doping process of the present invention, it is detected whether the metal clusters are precipitated in the experimental group and the control resisting sample, and the high-frequency complex impedance analyzer is used to delete the two-point measurement method. The impedance spectrum of the sample that is blocked from the control group. The measured jitter voltage is 500mV 'plus additional DC bias 〇~i.5v, 4 rate range is 500Hz to 110MHz 〇 and then the impedance of the two samples is real part of the impedance imaginary part green as shown in Figure 3. 1275*155 = Cole-Cole diagram, you can find that the spectral arc of the experimental resistance is a semi-circular arc curve. The curved curve of the control group except the half arc of the high-frequency region has a low frequency A short arc that increases in frequency and sag. Referring to Fig. 4, if the logarithm of the real part of the impedance of the spectrum is the horizontal axis, the logarithm of the imaginary part of the impedance is the vertical axis' and the characteristic slope is obtained. Semi-circle of experimental group samples
，弧曲線取對數下的特徵斜率為，表示實驗組之樣品遵 守單一晶界的遲豫行為，亦即實驗組之樣品是取代完全 (5 /〇 CoFe摻雜）的氧化辞。相反的，對照組樣品的圓弧曲線 取對數下的特徵斜率為〇.7，此結果表示了對照組樣品裡具 有複數η電遲豫行為產生。此額外的遲豫行為便是來自於 取代不完全的金屬團蔟與半導體丨所形成的介面所產生的結 果。亦即，對照組樣品是有金屬磁性團簇析出摻 雜）的氧化辞。 本發明半導體摻雜製程的檢測方法之主要原理是當斜 率為〇·5時，代表Cole-Cole圖中之圓弧曲線為單一的半圓 (簡略推導為R2+X2=Constant2同取數會得斜率為的方 耘式）’因此當斜率偏離0·5越多時，可視為有二組或二組 以上的遲豫行為產生，此額外的遲豫行為便是來自於取代 不完全的金屬團簇與半導體i所形成的介面產生的結果。且 該斜率並不會隨著不同材料之半導體丨或不同材料之摻雜 物質所影響，只會受到是否有第二雜相(如金屬團簇析出：) 而影響，因此可用於不同材料之半導體1上作為檢測之用 本發明半導體摻雜製程的檢測方法可以利用高頻複數 1275.155 阻抗技術很清楚的直觀的分辨出摻雜過程中有無金屬團箱 的^/成以辨別摻雜過程品質優劣。且檢測所得結果為所 述半導體1整體之特性，非局部特性。 以下利用兩種目前常用之檢測方法驗證實驗組及對照 組之樣品檢驗結果是否與本發明半導體摻雜製程的檢測方 法檢驗結果相同。 首先，利用台灣國家同步輻射中心的x光吸收光譜技 術鑑定材料的微結構，兩樣品經過同步輻射光源的吸收光 譜技術檢測後，其光譜圖是如圖5所示，首先於實驗組之 樣品的譜圖與ZnQ的譜圖很接近，於l U们A附近的兩 根波峰分別分別代表Co_〇與c〇_Zn的鍵結，表示摻雜的 Co沒有形成金屬的團簇，而是進行取代Zn原子的作用。 而於對照組樣品的結果在2.2A附近有明顯的波峰，此波峰 的位置與CoFe金屬峰的位置接近，表示對照組樣品中的 C〇是以金屬的鍵結方式存在，亦即對照組的樣品中有奈米 級金屬磁性團簇的形成。 參閱圖6及圖7，其次再利用等效電路法對實驗組及對 照組的樣品於不同外加偏壓下觀察其阻抗頻譜的結果與參 數，利用此方式經分析後可清楚的看到圖8之實驗組樣品 =需兩組RC就可作一完整的擬合，表示半導體中無金屬團 簇析出的阻抗頻譜圖，圖中R()gb、C()gb& R。g、c。g分別代表 晶界（oxide grain boundary)與晶粒（〇xide grain)的貢獻 。而圖9之對照組樣品須三組Rc才有辦法作一完整的擬合 其中二組 RC 分別為 Rogb、c〇gb、Rog、c〇g 及 Rm〇、Cm〇， 10 Ι275·155 、 m°、Cm。代表metal-oxide的介面，表示對照組 樣：中有金屬團簇析出。最後再由圖及圖U巾，於不同 偏壓下之參數變化’可發現金屬團蔟與半導體i形成之介面 項所貢獻的電阻與電容參數，皆隨著偏壓增加而減少，與 乳化物半導體1本身晶粒與晶界的行為*同，可更加確定證 明本發明之檢驗效果。 “、、星由上述之同步輻射光源的吸收光譜技術檢測及等效 電路法檢測實驗組與對照組之樣品結果，與本發明半導體 摻雜製程的檢測方法檢驗之結果相同，可證明本發明所提 供之檢測方法確實可正確檢驗半導體摻雜製程之優劣。上 述之同步輻射光源的吸收光譜技術檢測及等效電路法非本 發明之特徵技術，其詳細原理不再贅述。 μ綜上所述，本發明半導體摻雜製程的檢測方法利用量 測，便，成本較低之高頻複數阻抗分析儀2,制述半導體 1量測其阻抗頻譜’並經由取阻抗實部與虛部之對數緣出圖 形的特徵斜率，以判斷半㈣1是否有二組或以上的遲豫 行為IX U導體！摻雜過程中，是否有金屬團蔟的析出 ’:判別出摻雜製程的品質優劣，檢測程序簡便、快速又 可罪’所以確實可達到本發明之目的。 惟以上所述者，僅為本發明之較佳實施例而已，當不 能以此限定本發明實施之範圍，即大凡依本發明申請：利 範圍及發明說明内容所作之簡單料效變化與修飾，皆仍 屬本發明專利涵蓋之範圍内。 【圖式簡單說明】 11 1275155 圖1是本發明半導體摻雜製程的檢測方法之較佳實施 例的設備與一半導體的立體圖； 圖2是本發明半導體摻雜製程的檢測方法之實驗例的 一實驗組樣品與一對照組樣品於室溫下之磁滯曲線圖； 圖3是該實驗組與對照組之樣品將其高頻複數阻抗頻譜 之實部對虛部作圖； 囷4疋a亥實驗組與對照組之樣品將其高頻複數阻抗頻 譜之實部對虛部歸一化並取對數後所作之圖形； 圖5疋利用同步輕射X光研究該實驗例中實驗組及對 照組樣品結構的譜圖； 圖6是該實驗組樣品於不同外加偏壓下的阻抗頻譜之 實部對虛部作圖； 圖7是該對照組樣品於不同外加偏壓下的阻抗頻譜之 實部對虛部作圖； 圖8是類似於目4之示圖，以說明實驗組樣品之兩組 RC所代表的意義； 圖9疋類似於圖6之示圖，以說明對照組樣品之三組 RC所代表的意義； 、，圖10疋邊實驗組樣品於不同外加偏壓下氧化物晶界盥 晶粒個別貢獻之電阻與電容對偏壓之關係U 、 τ K 4對照組樣品於不同外加偏壓下氧化物金屬界 ,與θθ粒個別貢獻之電阻與電容對偏壓之關係圖 12 1275155 【主要元件符號說明】 1 半導體 2 高頻複數阻抗分析儀 11 電極 21 量測線The slope of the characteristic curve of the arc curve is that the sample of the experimental group obeys the single grain boundary, that is, the sample of the experimental group is replaced by the complete (5 / 〇 CoFe doping) oxidation word. On the contrary, the characteristic curve of the arc curve of the control sample was 〇.7, which indicates that the control sample had a complex η electrical hesitation behavior. This additional hesitation is the result of replacing the interface formed by the incomplete metal group and the semiconductor germanium. That is, the control sample is an oxidized word in which metal magnetic clusters are precipitated and doped. The main principle of the detection method of the semiconductor doping process of the present invention is that when the slope is 〇·5, the arc curve representing the Cole-Cole diagram is a single semicircle (a simple derivation is that R2+X2=Constant2 has the same slope as the number of acquisitions). Therefore, when the slope deviates from 0·5, it can be considered that there are two or more groups of hesitation behaviors. This additional delay behavior comes from the replacement of incomplete metal clusters. The result produced by the interface formed by the semiconductor i. And the slope does not affect the semiconductor germanium of different materials or the doping substances of different materials, and is only affected by whether there is a second impurity phase (such as metal cluster precipitation:), so it can be used for semiconductors of different materials. The detection method of the semiconductor doping process of the present invention can be used to detect the presence or absence of a metal cluster in the doping process to distinguish the quality of the doping process by using the high frequency complex 1275.155 impedance technique. Further, the result of the detection is the characteristics of the entire semiconductor 1, and the non-local characteristics. The following two commonly used detection methods are used to verify whether the test results of the experimental group and the control group are the same as those of the semiconductor doping process of the present invention. First, the x-ray absorption spectroscopy technique of Taiwan National Synchrotron Radiation Center was used to identify the microstructure of the material. After the two samples were detected by the absorption spectroscopy technique of the synchrotron radiation source, the spectrum is shown in Figure 5, first in the experimental group. The spectrum is very close to the spectrum of ZnQ. The two peaks near L U are respectively the bonds of Co_〇 and c〇_Zn, indicating that the doped Co does not form a metal cluster, but rather Substituting the role of Zn atoms. The results of the control sample showed obvious peaks near 2.2A. The position of this peak was close to the position of the CoFe metal peak, indicating that the C〇 in the control sample was present by metal bonding, that is, the control group. There is a formation of nano-scale metal magnetic clusters in the sample. Referring to Fig. 6 and Fig. 7, the equivalent circuit method is used to observe the impedance spectrum results and parameters of the experimental group and the control group under different applied bias voltages. After this analysis, the figure 8 can be clearly seen. Experimental group sample = need two sets of RC to make a complete fit, indicating the impedance spectrum of the metal-free clusters in the semiconductor, R () gb, C () gb & R. g, c. g represents the contribution of the oxide grain boundary and the 〇xide grain, respectively. However, the control sample of Figure 9 requires three sets of Rc to have a complete fit. The two sets of RC are Rogb, c〇gb, Rog, c〇g and Rm〇, Cm〇, 10 Ι275·155, m. °, Cm. The interface representing metal-oxide indicates a control sample in which metal clusters are precipitated. Finally, from the graph and the U-sheet, the parameter changes under different biases can be found that the resistance and capacitance parameters contributed by the interface between the metal cluster and the semiconductor i are reduced with the increase of the bias voltage, and the emulsion The behavior of the semiconductor 1 itself and the grain boundary* can be more certain to prove the test effect of the present invention. ",, the star is detected by the absorption spectrum technique of the above-mentioned synchrotron radiation source and the equivalent circuit method detects the sample results of the experimental group and the control group, and the result of the test method of the semiconductor doping process of the present invention is the same, and the invention can be proved. The detection method provided can correctly verify the advantages and disadvantages of the semiconductor doping process. The absorption spectrum technique detection and equivalent circuit method of the above-mentioned synchrotron radiation source are not the characteristic technology of the present invention, and the detailed principle thereof will not be described again. The detection method of the semiconductor doping process of the present invention utilizes the measurement, and the low-frequency high-frequency complex impedance analyzer 2, and the semiconductor 1 measures the impedance spectrum 'and obtains the logarithm of the real part and the imaginary part of the impedance The characteristic slope of the graph to determine whether the half (four) 1 has two or more sets of delayed behavior IX U conductors! Whether there is precipitation of metal clusters during the doping process': discriminate the quality of the doping process, the detection procedure is simple, It is fast and guilty, so it is indeed possible to achieve the object of the present invention. However, the above is only a preferred embodiment of the present invention, when it is not possible The scope of the present invention is defined by the scope of the present invention, and the simple changes and modifications made by the present invention are still within the scope of the present invention. [Simplified Schematic] 11 1275155 1 is a perspective view of a device and a semiconductor of a preferred embodiment of the method for detecting a semiconductor doping process of the present invention; FIG. 2 is an experimental group sample and a control sample of an experimental example of the method for detecting a semiconductor doping process of the present invention; The hysteresis curve at room temperature; Figure 3 is a plot of the real part of the high-frequency complex impedance spectrum of the experimental and control samples against the imaginary part; the sample of the 囷4疋ahai experimental group and the control group The real part of the high-frequency complex impedance spectrum is normalized to the imaginary part and the logarithm is taken; Figure 5疋 The spectrum of the experimental and control samples in the experimental example is studied by using synchronous light-emitting X-ray; The experimental group samples are plotted against the imaginary part of the impedance spectrum under different applied bias voltages. Figure 7 is a plot of the real part of the impedance spectrum of the control sample under different applied biases to the imaginary part; similar Figure 4 is a diagram to illustrate the significance of the two sets of RC of the experimental group samples; Figure 9 is similar to the diagram of Figure 6 to illustrate the significance of the three sets of RC of the control sample; Figure 10疋The relationship between the resistance and the capacitance of the experimental group sample under different applied bias voltages and the capacitance to the bias voltage U, τ K 4 control sample under different applied bias voltage oxide metal boundary, and θ θ particles The relationship between the resistance and the capacitance of the individual contribution to the bias voltage. Figure 12 1275155 [Explanation of the main components] 1 Semiconductor 2 High-frequency complex impedance analyzer 11 Electrode 21 Measurement line
1313