標題: 有機摻雜半導體與非晶矽半導體之導納分析模型建立與電性研究
Electric Characteristics of Organic Doping Semiconductor and Amorphous Semiconductor Elucidated by Admittance spectroscopy
作者: 謝明達
Hsieh, Ming-Ta
陳振芳
Chen, Jenn-Fan
電子物理系所
關鍵字: 有機摻雜半導體;非晶矽;等效電路模型;導納頻譜;organic doping semiconductor;amorphous silicon;equivalent circuit model;admittance spectroscopy
公開日期: 2008
摘要: 本論文根據導納量測理論與有機材料高阻抗的特性提出了雙異質介面有機元件的導納等效電路模型,並且在世界上首次將其應用於有機摻雜與擴散的分析中。導納中的電容-頻率量測對於有機材料元件的幾何結構上可以準確的得到所對應的厚度與介電常數,這對於製程上的膜厚校驗提供一種很快速又精確的方法。而在導納量測中同時又可以得到有機材料的電阻-電容(RC)時間響應,並由變溫的實驗可以得到其相對應的活化能,接著再藉由半導體理論公式就可以進而得到材料的相關重要的物理參數。在不同厚度的N,N_-bis-_1-naphthyl_-N,N_ -diphenyl,1,1_-biphenyl-4 ,4_-diamine (NPB)與2-methyl-9,10-di_2-naphthyl_anthra- cene (MADN)實驗中,薄膜的品質與成長厚度是成正比的,隨著成長的厚度增加,NPB與MADN的活化能與電阻率皆變小並且會趨於飽和,這表示有機薄膜的特性隨著厚度的增加而變好並且趨於穩定。而在相同厚度且沒有摻雜下NPB的阻值或是活化能都來的比MADN小,這表示在沒有摻雜的狀態下NPB可能比MADN更適合當作有機發光元件中的電洞傳輸層。在實際的發光元件上可以發現,使用MADN的元件在驅動電壓上的表現確實比使用NPB來的高,但是元件在發光效率以及元件壽命的表現上,使用MADN的元件卻比NPB來的好,這是由於MADN提供元件較低的電洞遷移率與較高的電洞注入位障而導致元件注入發光層中的電子電洞數目有較佳的平衡,而較佳的平衡增加了電子電洞對在發光層中複合的機率同時也減少了因為多餘陽離子堆積所造成的消光與劣化現象。在第三章中我們利用導納頻譜與相對應的等效電路模型分析有機材料經摻雜後的薄膜特性。在p型摻雜研究中,NPB與MADN經由三氧化鎢(WO3)摻雜後都可以有效的改善電洞注入特性與薄膜的電洞阻抗,並且都在濃度為33%時有最好的電洞注入特性。而在對於摻雜物WO3的侷限能力上,MADN從電流-電壓量測可以看到濃度高於20%後元件在turn-on之後的電流傳導特性開始變差,這個現象是因為WO3由MADN擴散到tris(8-quinolinolato)aluminium (Alq3)所產生的缺陷而導致的。在p-i-i元件方面,使用MADN摻雜WO3為電洞傳輸層的發光元件在摻雜濃度為10%時可以達到有最佳的元件效率表現4.0 cd/A與 2.4 lm/W,這結果與沒有摻雜的元件效率上足足提高了43%,並且優於以傳統NPB摻雜WO3作為電洞傳輸層且最佳化下的結果。 在n型摻雜的研究方面,我們利用了導納與電容-電壓量測對於Bathophenanthroline (Bphen)摻雜dipotassium phthalate (PAK2)作薄膜上的物理分析,其結果可以知道將PAK2摻雜入Bphen中可以幫助電子的注入也可以降低Bphen的電子阻抗,將其應用在發光元件中,在發光效率與驅動電壓上都表現的比傳統使用LiF為電子注入層的元件提高約40%的效率,並且在元件壽命的表現也不會因為摻雜後而有所損失。在MADN經由Cs2CO3與CsF摻雜後的薄膜物理特性研究部分,從實驗數據可以發現MADN不單是可以經由WO3摻雜,變成良好的電洞傳輸與注入層,也可以經由Cs2CO3與CsF的摻雜而變成非常好的電子注入與傳輸層。由於製程上的限制在使用Cs2CO3摻雜最高只能達到15%的摻雜濃度,不過仍然可以將MADN的電子阻抗有效降低,並且MADN的活化能被改善到0.288 eV。而CsF可提供MADN較高的摻雜濃度,並在濃度為33%的時候能使MADN的活化能降低到約0.1eV。最後製作以MADN摻雜WO3為電洞傳輸層與MADN摻雜Cs2CO3為電子注入層的p-i-n元件,從實驗結果顯示其元件在20 mA/cm2電流密度下可以達到4.2V的驅動電壓、4.6 cd/A與3.0 lm/W的效率,這結果顯示,使用MADN為p-i-n元件中單一host材料是可行的。在第四章中,我們將第二章所提出的有機雙異質接面等效電路模型加以修正並且應用於雙異質接面元件的摻雜物擴散研究。在Cs2CO3摻雜入MADN中的擴散研究中發現,在摻雜濃度為10%的元件中MADN受擴散影響的區域範圍約為8.3 nm,而此範圍內的MADN因為部分的Cs2CO3擴散進入鄰近的Alq3中而使有效的摻雜濃度降低而導致阻抗增加。而在摻雜濃度為15%的元件可以發現MADN中受擴散影響的區域的阻抗仍可以與未受影響的MADN相似,所以能以等效的訊號被量測到。受限於製程與機台量測範圍的限制,在Cs2CO3摻雜入MADN的研究中在室溫下我們無法藉由導納量測觀察到Alq3受擴散影響區域的厚度。在濃度為5%的元件中,我們試著將溫度提高到420 K進行量測,從實驗數據可以得知經過420 K的高溫活化後MADN中的Cs2CO3被再擴散進入Alq3中並且擴散的深度約為4.3 nm。在CsF摻雜入MADN的研究部分,在摻雜濃度為10%與20%的元件中,在有限的導納量測範圍內仍無法清楚的觀察到其CsF擴散進Alq3的深度,但是從C-F量測中可以觀察到CsF摻雜濃度為10%的元件中MADN 受擴散效應所影響的厚度其值約為3.5 nm。而CsF摻雜濃度為33%的元件中,在室溫下就可以明顯的觀察到其Alq3中受擴散效應所影響區域,經由計算其厚度約為9.4 nm,而再經由420 K量測後會更近一步增加到14.8 nm。在第五章中我們提出非晶矽metal-insulator-amorphous silicon (MIAS)電容結構在不同操作區域下的等效電路模型,藉由此模型與導納量測可以分析出非晶矽薄膜的特性,並且可以解析出非晶矽薄膜與絕緣層間的介面缺陷能態密度,由實驗的結果驗證所提出的模型相當正確,而從MIAS電容結構所得到的實驗結果也可以直接反應在相同製程條件下薄膜電晶體 (TFTs)元件的輸出特性上。由以上結論可以說明在第五章中所提出的MIAS電容等效電路模型在對於非晶矽薄膜特性研究上提供了方便且正確的分析方法並且可以直接反應在TFTs元件特性上,相信此等效電路模型不僅可以應用在非晶矽薄膜元件,將來相信也可以應用於多晶矽與微晶矽元件上。
In this thesis, the equivalent circuit models for studying the electrical characteristics and diffusion effects of a heterojunction OLED with doping layer were established. This is first time to observe the effect of dopant diffusion in organic doping device in the world. In a multi-layer OLED, each layer can be treated as a resistance-capacitance (RC) unit. Moreover, each layer in the OLED commonly shows an independent RC property, thus the geometric and electrical characteristics can be investigated separately by admittance spectroscopy (AS) due to the differences of RC responses. As a result, the electrical characteristics of each layer can be determined by AS with a suitable equivalent circuit model Based on the circuit model and capacitance-frequency measurements, the dielectric constant and/or the thickness of organic thin film can be obtained exactly, which provides a rapid and convenient method to monitor the accuracy of organic thin film thickness during the process. Moreover, the RC time constant of organic materials simultaneously can be obtained. In the study of 1,4-bis[N-(1-naphthyl)-N′-phenylamino]−4,4′ diamine (NPB) and 2-methyl-9,10-di(2-naphthyl)anthracene (MADN) with various thicknesses, the quality of organic thin film greatly depends on the thickness. As thickness increases, the resistivity and activation energy of organic material are improved and shows a saturate trend. Under the same thickness, the resistivity and activation energy of intrinsic MADN are larger than that of NPB. However, the admittance spectroscopy studies show that using MADN as hole transport material (HTM) can reduce the amount of hole via HTM layer resulting in a well-balanced carrier recombination. The green fluorescent 10-(2-benzothiazolyl)-1,1,7,7-tetramethyl-2,3,6,7-tetrahydro-1H,5- H,11H-benzo[l]-pyrano[6,7,8-ij]quinolizin-11-one doped device can achieve a current efficiency of 21.8 cd/A and a power efficiency of 10.4 lm/W at 20 mA/cm2 that are 65% higher than those of devices adopted NPB as HTM. The green-doped device also achieved a long half-decay lifetime of 22000 h at an initial brightness of 500 cd/m2. In the studies of p-type doping organic materials, the effect of tungstenoxide (WO3) incorporation into NPB layer is investigated in NPB-tris(8-hydroxyquinoline)aluminium (Alq3) heterojunction organic light-emitting diodes. The admittance spectroscopy studies show that increasing the WO3 volume percentage from 0 to 16 % can increase the hole concentration of the NBP layer from 1.97×1014 to 1.90×1017 cm-3 and decrease the activation energy of the resistance of the NPB layer from 0.354 to 0.176 eV. Thus, this incorporation reduces the ohmic loss and increases the band bending in the NBP layer near the interface, resulting in an improved hole injection via tunneling through a narrow depletion region. In addition, an efficient p-doped transport layer composed of an ambipolar material, MADN and WO3 has been developed. The admittance spectroscopy studies show that the incorporation of WO3 into MADN can greatly improve the hole injection and the conductivity of the device. Moreover, when this p-doped layer was incorporated in the tris(8-quinolinolato)aluminium based device, it achieved a current efficiency of 4.0 cd/A and a power efficiency of 2.4 lm/W at 20 mA/cm2. This work paves the way to simplify the fabrication of future p-i-n OLED with a single common ambipolar MADN material. In the n-type doping study, the electrical characterization of bathophenanthroline (BPhen) doped with PAK2 is investigated by current-voltage (I-V) and admittance spectroscopy measurements. The investigations show that the incorporation of PAK2 into BPhen is found to raise the Fermi level from 1.7 eV to only around 0.5 eV below BPhen’s lowest unoccupied molecular orbital, which further enhances the efficiency of electron injection from Al cathode. When this n-doped layer is adopted in OLED device, the green fluorescent 10-(2-benzothiazolyl)-1,1,7,7-tetramethyl-2,3,6,7-tetrahydro-1H,5H,11H-benzo[l]-pyrano[6,7,8-ij]quinolizin-11-one doped device can achieve a current efficiency of 16 cd/A and a power efficiency of 10.9 lm/W at 1000 cd/m2. In another n-type doping work, the admittance spectroscopy studies show that doping cesium fluoride (CsF) into MADN can greatly decrease the resistance of MADN and raises the Fermi level from deep level to only 0.1 eV below the lowest unoccupied molecular orbital, resulting in enhancing the electron injection. In addition, the diffusion width of CsF from doped MADN layer into tris(8-quinolinolato)aluminium is clearly observed by capacitance-frequency measurement and is about 9.4 nm. Moreover, the diffusion width is significant to be affected by external thermal. Detailed admittance spectroscopy measurements were made on a metal-silicon nitride-hydrogenated amorphous silicon (MIAS) structure. Based on the properties of hydrogenated amorphous silicon (a-Si:H), three simplified equivalent circuit models under various operating conditions (accumulation, depletion and full depletion) are presented along with an alternative direct measurement method at room temperature. Admittance spectroscopy shows that the interface states density between silicon nitride (SiNx) and a-Si:H can be determined from the depletion equivalent circuit model. The resisivity and activation energy of a-Si:H also can be obtained using accumulation and depletion equivalent circuit models. These models can be employed easily to monitor the fabrication parameters of thin-films transistors (TFTs) and to obtain accurately and directly the capacitance model parameters of TFTs .
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT009421802
http://hdl.handle.net/11536/81286
Appears in Collections:Thesis


Files in This Item:

  1. 180201.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.