cond-mat0411467/v3.tex
1: \documentstyle[multicol,aps,prl,epsf]{revtex}
2: 
3: \pagestyle{empty}
4: \begin{document}
5: 
6: \draft
7: 
8: \title{Electric field induced charge injection or exhaustion in organic 
9: thin film transistor}
10: 
11: \author{
12: Manabu Kiguchi$^1$, Manabu Nakayama$^1$, Toshihiro Shimada$^2$, Koichiro Saiki$^{1,2}$}
13: 
14: \address{$^1$Department of Complexity Science $\&$ Engineering, Graduate
15: School of Frontier Sciences, The University of Tokyo, 7-3-1 Hongo,
16: Bunkyo-ku,Tokyo 113-0033, Japan}
17: 
18: \address{$^2$Department of Chemistry, Graduate School of Science, The 
19: University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan}
20: 
21: \date{\today}
22: 
23: \maketitle
24: 
25: \begin{abstract}
26: The conductivity of organic semiconductors is measured 
27: {\it in-situ} and continuously with a 
28: bottom contact configuration, as a function of film thickness at various gate voltages. 
29: The depletion layer thickness can be directly determined 
30: as a shift of the threshold thickness at which electric 
31: current began to flow. The {\it in-situ} and continuous measurement can 
32: also determine qualitatively 
33: the accumulation layer thickness together with the distribution 
34: function of injected carriers. 
35: The accumulation layer thickness is a few mono layers, 
36: and it does not depend on gate voltages, rather depends on the chemical species. 
37: 
38: \end{abstract}
39: 
40: \medskip
41: 
42: \pacs{PACS numbers: 79.60.Jv, 61.14.Hg, 68.55.-a}
43: 
44: \begin{multicols}{2}
45: \narrowtext
46: 
47: \section{INTRODUCTION}
48: \label{sec1}
49: 
50: Charge injection or depletion by electric fields in semiconductors is 
51: the basis of field electric transistors (FET)\cite{1}. 
52: It is also the basis of the field effect control of physical properties, 
53: which is one of the hottest topics in the field of condensed matter science. 
54: In contrast with chemical doping, 
55: the charge injection by electric fields does not introduce 
56: any chemical or structural disorder in pristine materials\cite{2,3}, 
57: and it may be possible to inject carriers in host materials, to which charges 
58: can not be injected by chemical doping. In this context, determination of the region, 
59: where carriers are practically injected or depleted by electric fields, 
60: is an important and fundamental issue. 
61: However, there had been little investigation on evaluating experimentally 
62: the dimension of these 
63: regions\cite{4}.
64:  
65: Recently, Dodabalapur {\it et al.} reported a pioneering work 
66: relating to the accumulation layer 
67: thickness. They estimated that it was 1$\sim$2 ML by comparing FET characteristics of a 2 ML 
68: sexithiophene ($\alpha$-6T) film with that of bulk one\cite{5}. 
69: However, they did not quantitatively determine 
70: the thickness, and quantitative analysis has not been done yet. 
71: Recently, we proposed a simple 
72: method to determine quantitatively the thickness of the depletion and accumulation layers in 
73: organic semiconductor pentacene\cite{6}.
74:  The conductivity of pentacene thin film transistor (TFT) was 
75: measured {\it in-situ} and continuously with a bottom contact configuration, 
76: as a function of film thickness. 
77: However, we only discussed the thickness of these layers for one organic species at fixed 
78: gate voltage (0, $\pm$15 V). 
79: By comparing the result of pentacene TFT with that of another organic TFT, 
80: and that at various gate voltages, 
81: we could discuss the characteristics of the conductance in organic 
82: TFT in more detail. Furthermore, there are many points to discuss; 
83: way and validity of estimation, universality of the method, etc.
84:   
85: In this study, we have studied the {\it in-situ} 
86: and continuous measurement for pentacene and 6T. 
87: By measuring the gate voltage and chemical species dependence, 
88: we quantitatively discussed the 
89: thicknesses of the depletion and accumulation layers in detail, and characteristics of the 
90: conductance in organic materials. 
91: 
92: \section{EXPERIMENTAL}
93: \label{sec2}
94: 
95: Figure~\ref{fig1} shows the schematic sample layout. 
96: The substrate was a highly doped silicon wafer, acting as a gate electrode. 
97: The gate dielectric layer was a 700 nm thermally grown silicon dioxide. 
98: On top of the surface, 30 nm thick gold source(S) and drain(D) electrodes 
99: were deposited through a shadow mask. 
100: The channel length and width were 100 $\mu$m and 5.4 mm. 
101: Pentacene (Aldrich) and $\alpha$-6T (Syncom BV) were deposited 
102: by means of vacuum deposition.
103: The film thickness was monitored by a quartz crystal oscillator, 
104: and it was calibrated by Auger Electron Spectroscopy.
105: The deposition rate was 0.1 nm/min, 
106: and the substrate temperature was kept at 310 K during the growth. 
107: Under high-vacuum condition, the S-D current of TFT was measured 
108: as a function of film thickness with a bottom-contact configuration. 
109: During the growth of the organic films, electric fields were not 
110: applied in order to exclude the influence of the electric current on the film growth. 
111: As a function of the film thickness, 
112: the gradual change in the conductivity and FET characteristics were measured 
113: {\it in-situ} under the same condition, avoiding the problem of specimen dependence.
114: 
115: \section{RESULTS}
116: \label{sec3}
117: \subsection{Depletion layer}
118: \label{sec3a}
119:  Figure~\ref{fig2} shows the S-D current ($I_{SD}$) of pentacene 
120: and 6T as a function of film thickness at various 
121: gate voltages ($V_{G}$); $V_{G}$ =-30, -15, 0, and +15 V. 
122: The S-D voltage ($V_{SD}$) was kept at 1 V. Since both 
123: pentacene and 6T are p-type semiconductors, 
124: large $I_{SD}$ is observed at negative gate voltages. Here 
125: we find the presence of threshold thickness ($d_{th}$) 
126: at which electric current begins to flow. 
127: There are two points to note about $d_{th}$. First, $d_{th}$ is $\sim$1 nm at $V_{G}$=0 V, 
128: and finite electric current is observed even for the 1.0 nm thick film. 
129: The present result indicates that the S, D electrodes (width 100 $\mu$m) 
130: could be electrically connected by only 1 nm thick organic semiconductors. 
131: Furthermore, typical p-type FET characteristics were observed for the this film \cite{7}. 
132: Second, $d_{th}$ shows clear $V_{G}$ dependence, 
133: and it shifts up to 4.4($\pm$0.5) nm for pentacene and $d_{th}$=5.8($\pm$0.6) nm 
134: for 6T at $V_{G}$=15 V. The value of $d_{th}$ was well reproduced for several experiments.
135: When positive $V_{G}$ is applied to p-type semiconductors, 
136: the depletion layer with a low conductance is 
137: formed at the semiconductor/insulator interface. 
138: Then, the whole film can be depleted on the thin 
139: limit, and $I_{SD}$ does not flow at positive $V_{G}$. 
140: Consequently, it is suggested that $d_{th}$ corresponds to the depletion layer thickness. 
141: 
142: In order to discuss the validity of the above discussion, 
143: we estimated the concentration of major carriers by two different ways 
144: with the conventional model which has been already established for 
145: inorganic semiconductors\cite{1,8}. 
146: When a depletion layer is formed at the semiconductor/insulator interface, 
147: carriers are exhausted and ionized acceptors are left in the depletion layer. 
148: The induced charge amounts to be $Q=CV_{G}$, 
149: if a flat band condition is satisfied at $V_{G}$=0 V. 
150: Here, $C$ is the capacitance of the gate dielectric, 
151: and it is represented as $C=\epsilon _{OX} / t_{OX}$, 
152: where $\epsilon _{OX}$: dielectric constant of 
153: SiO$_{2}$, $t_{OX}$: thickness of SiO$_{2}$. 
154: In the depletion approximation, only ionized acceptors are left in the 
155: depletion layer and distributed uniformly in the depletion layer. 
156: The concentration of the acceptors ($N$) (i.e. exhausted major carriers) is, 
157: thus, obtained by dividing induced charge ($Q$) by the depletion 
158: layer thickness ($t_{s}$), and it is represented as $N=Q/et_{s}$ . 
159: Because of the charge neutrality at $V_{G}$=0 V, 
160: concentration of the major carriers is equal to that of acceptors. 
161: Therefore, the concentration of the 
162: major carriers of pentacene at $V_{G}$=0 V amounted to $N$=$1.1 \times 10 ^{18} cm^{-3}$ 
163: using $t_{s}$=4.4 nm at $V_{G}$=15 V, 
164: while it amounted to $N$=$8.0 \times 10 ^{17} cm^{-3}$ for 6T 
165: using $t_{s}$= 5.8 nm at $V_{G}$=15 V.
166: 
167: On the other hand, the concentration of the mobile major carriers can be estimated by the 
168: conductance of TFT at $V_{G}$=0 V. Here we assume zero contact potential between organic 
169: semiconductors and Au electrodes. This assumption is supported by ohmic behavior of the S, D 
170: contacts to the pentacene and 6T active layers observed in the present FET characteristics. 
171: The conductivity of the 4.4 nm thick pentacene film was 
172: $ 2.5 \times 10 ^{-2} \Omega ^{-1} cm ^{-1}$ using $I_{SD}$ ($ 6.0 \times 10 ^{-7}$ A) at 
173: $V_{G}$=0 V, and the conductivity of the 5.8 nm thick 6T film was 
174: $ 1.6 \times 10 ^{-6} \Omega ^{-1} cm ^{-1}$  
175: using $I_{SD}$ ($5.1 \times 10 ^{-11}$ A) at $V_{G}$=0 V. 
176: Since the conductivity is represented as $ne \mu$, 
177: where $n$, $e$, and $\mu$ are carrier density, unit charge, and mobility, 
178: the concentration of the mobile major carriers ($N_{m}$) at $V_{G}$=0 V is obtained 
179: by dividing the conductivity by the mobility of the organic film. 
180: Therefore, $N_{m}$ was estimated to be 
181: $7.6 \times 10^{17} cm^{-3}$ for pentacene and $N_{m}$=$8.2 \times 10^{16} cm^{-3}$ for 6T. 
182: 
183: The fact that $N$ determined by $d_{th}$, and $N_{m}$ determined 
184: by the conductance at $V_{G}$=0 V, have the same order, 
185: indicates that $d_{th}$ actually corresponds to the depletion layer thickness. 
186: In other words, we could directly determine the deletion layer thickness 
187: as a shift of $d_{th}$ by the present {\it in-situ} and continuous measurement. 
188: Furthermore, the measurement was revealed to be a rather general method 
189: to determine the depletion layer thickness, 
190: because a shift of $d_{th}$ was observed for both organic semiconductors.
191: 
192: Although the above discussion on $d_{th}$ is roughly justified, 
193: there is a difference between $N$ and $N_{m}$; $N_{m}$ is always lower than $N$. 
194: The difference could be caused by several assumptions stated above 
195: (i.e. flat band condition, zero contact potential and depletion approximation). 
196: On the other hand, the difference could include important information 
197: on the concentration of the major carriers. 
198: Some of carriers are trapped at trap sites, such as grain boundaries, 
199: defects, or semiconductor/insulator (metal) interfaces. 
200: Therefore, the difference of concentration between $N$ and $N_{m}$ can be considered 
201: to relate to the concentration of the trapped carriers. 
202: 
203: Here we comment on the effect of impurity on the determination of the depletion layer 
204: thickness by the present measurement. 
205: Finite conductivity at $V_{G}$=0 V means that the carrier concentration 
206: originating from the impurity is higher than the concentration of trap sites. 
207: This causes positive threshold gate voltage($V_{th}$) in the FET characteristics. 
208: In case of purified samples, however, the carrier concentration 
209: can be lower than the concentration of trap sites, and $d_{th}$ would 
210: not be observed at positive $V_{G}$. 
211: High concentration of impurity is, thus, essential for the present method. 
212: Since the depletion layer thickness is inversely proportional to 
213: the concentration of major carriers, 
214: high concentration of impurity help determination of the depletion layer thickness 
215: ($\sim$5 nm at $V_{G}$=15 V). 
216: 
217: 
218: \subsection{Accumulation layer}
219: \label{sec3b}
220: 
221: In the previous section, we discussed the depletion layer thickness. 
222: In this section, we discuss the thickness of the accumulation layer, 
223: and the distribution function of charge injected by electric fields. 
224: In the following analysis, we assume that the organic semiconductors grow 
225: in a layer-by-layer fashion\cite{afm}, and that the carrier concentration depends 
226: only on electric fields ($V_{G}$), and does not depend on the film thickness. 
227: The observed conductance of TFT is the sum of the conductance of each layer, 
228: which is parallel to the interface. The conductance of each layer can be, 
229: thus, estimated by the first derivative of the total conductance 
230: with respect to film thickness. Then, the carrier density of each layer 
231: is determined by dividing the conductance of each layer by $\mu$. 
232: Based on the assumption of layer growth, the mobility can be constant, 
233: because mobility depends on morphology of the TFT\cite{9}. 
234: The mobility is determined by the FET characteristics of the 20 nm 
235: thick film (0.23 $cm^{2}V^{-1}s^{-1}$ for pentacene, 
236: $2.5 \times 10 ^{-4}cm^{2}V^{-1}s^{-1}$ for 6T). 
237: 
238: Figures~\ref{fig3} and ~\ref{fig4} show the obtained distribution function 
239: ($n(x)$) of carriers as a function of distance $x$ from the interface 
240: for pentacene and 6T at various $V_{G}$. 
241: The large carrier density at $V_{G}$=0 V in small $x$ region is 
242: due to the charge transfer from the Au electrodes to the pentacene molecules \cite{10}. 
243: By subtracting the carrier density $n_{0}(x)$ at $V_{G}$=0 V from 
244: $n_{-15}(x)$ at $V_{G}$=-15 V, or $n_{-30}(x)$ at $V_{G}$=-30 V, 
245: the carrier density injected by electric fields ($n_{i}(x)$) is obtained. 
246: As seen in the figure, $n_{i}(x)$ decays steeply with increasing $x$, 
247: meaning that the injected carriers are localized at the interface. 
248: We could quantitatively determine the distribution function of carriers and the 
249: accumulation layer thickness for organic TFT by the {\it in-situ} and continuous measurement. 
250: It should be noticed that the above processes (differentiation, division by $\mu$) 
251: can be validated only for the continuous measurement results, 
252: in which the sample condition is kept unchanged. 
253: Otherwise, for an individual sample, the condition changes from one sample 
254: to another due to the difference in grain size, crystallinity, interface state, etc.
255: 
256: For quantitative estimation on the accumulation layer thickness, $n_{i}(x)$ is fitted with an 
257: exponential function ($f(x)=a \times exp(-x/b)$). 
258: The fitted value of $b$ was 2($\pm$1) nm for pentacene and 9($\pm$5) nm for 6T 
259: at both $V_{G}$=-15 V and -30 V \cite{11}. 
260: Again, the value of $b$ was well reproduced for several experiments.
261: Here, $b$ can be considered as an effective accumulation layer thickness. 
262: There are three points to note about the accumulation layer thickness ($d_{ac}$). 
263: First, most of injected carriers are localized in a few ML next to the interface.
264: (1ML thickness: 1.5 nm for pentacene, 2.2 nm for 6T)
265: We could quantitatively show the localization of injected carrier at the interface by 
266: experiments, while most previous works simply shows 
267: that carriers were localized at the interface\cite{5}. 
268: Second, $d_{ac}$ does not depend on gate voltages. 
269: Third, $d_{ac}$ of pentacene is smaller than that of 6T. 
270: In the following, we would discuss these gate voltage and chemical species dependence of the 
271: accumulation layer thickness.
272: 
273: First, we discuss the gate voltage dependence of the accumulation layer thickness. 
274: In an ideal semiconductor, $d_{ac}$ is inversely proportional to $V_{G}$\cite{12}. 
275: On the other hand, $d_{ac}$ of organic semiconductor pentacene and 
276: 6T did not show gate voltage dependence in the present study. 
277: The difference could be explained by the electron-electron repulsion, 
278: the localized states in organic semiconductors, and geometry-related factors. 
279: Electron-electron repulsion delocalizes the carriers, and leads to an increase in $d_{ac}$. 
280: Since the repulsion force increases with carrier concentration (i.e. $V_{G}$), 
281: a decrease of $d_{ac}$ could be compensated with an increasing $V_{G}$. 
282: This causes little $V_{G}$ dependence. 
283: On the other hand, $V_{G}$ dependence in an ideal semiconductor assumes 
284: the parabolic density of states for the conduction band. 
285: In organic TFT, however, there is no long range order in the specimens. 
286: Because of absence of the long range order, 
287: the localized states with finite energy width are formed below the free 
288: electron like conduction band. Therefore, the equations assuming parabolic DOS could not be 
289: applicable to the organic TFT, and $d_{ac}$ was not inversely proportional to $V_{G}$.    
290: The geometry-related factors can also explain the experimental results. 
291: There are many grain boundaries in the growth film, 
292: and these grain boundaries and defects would affect the charge 
293: injection. Therefore, the equations assuming a perfect semiconductor crystal 
294: could not be applicable to the organic FET. 
295: Although we present some explanation of the little gate voltage dependence of 
296: the accumulation layer thickness, we can not present decisive one at present.
297: 
298: 
299: Second, the chemical species dependence of the accumulation thickness is discussed. 
300: For p-type organic semiconductors, holes are injected 
301: in the highest occupied molecular orbital (HOMO) of the semiconductor, 
302: and thus, $d_{ac}$ reflects the spatial distribution of HOMO. 
303: On the other hand, both pentacene and 6T are $\pi$-conjugated molecules, 
304: and their HOMOs spread in the whole molecules. 
305: Therefore, the molecular size should be taken into account to discuss $d_{ac}$. Since the 
306: molecular size of 6T is larger than that of pentacene, 
307: $d_{ac}$ of 6T was larger than that of pentacene. 
308: Besides the molecular size, another factors, such as difference in dielectric constant, 
309: intrinsic carrier density, could cause the difference in $d_{ac}$\cite{13}.
310: 
311: Here we estimate the amount of carriers injected by electric fields at the first layer, 
312: in order to obtain a guide to the electric fields control of physical properties. 
313: Since the distribution function of the injected carrier can be represented by 
314: $f(x)$=$a \times exp(-x/b)$, the amount of carriers in the first 
315: layer is obtained by integrating $f(x)$ from 0 to the thickness of 1 ML 
316: (1.5 nm for pentacene, 2.2 nm for 6T ). 
317: The amount of the injected carriers was $6.1 \times 10^{11}cm^{-2}$ for pentacene 
318: and $2.6 \times 10^{11}cm^{-2}$ for 6T at $V_{G}$=-30 V. 
319: One carrier was injected every 350 (810) pentacene (6T) molecules at $V_{G}$=-30 V.  
320: In most of previous studies, the amount of injected carriers per each molecule was estimated 
321: assuming that carrier were localized at the first layer. 
322: In the present work. however, distribution of injected carriers were estimated, 
323: showing that they were not necessarily localized at the first layer. 
324: The present result can be, thus, 
325: a guide to control of physical properties by electric fields. 
326: 
327: In the present study, the {\it in-situ} and continuous measurement was applied to organic 
328: semiconductor, in order to quantitatively determine the thickness 
329: of the depletion and accumulation layers. 
330: As discussed in the previous section, layer growth is essential to the present method. 
331: The reason why we studied organic semiconductors, 
332: is just that they do not need high substrate temperature to obtain flat films, 
333: in contrast with the inorganic semiconductor which needs high substrate temperature. 
334: Only if we can prepare flat films at low substrate temperature, the present 
335: method can be applied to inorganic semiconductor to estimate 
336: the thickness of the depletion and accumulation layers, 
337: and distribution function of the injected carriers.
338: 
339: Finally, we comment the control of physical properties by electric fields. 
340: Since carriers reside in a few ML from the semiconductor/insulator interface, 
341: the physical properties only at the interface could be controlled by the charge injection. 
342: This implies the importance of a well-ordered interface 
343: through which charges are efficiently injected. 
344: Furthermore, we propose that the semiconductor-metal or -superconductor transition 
345: is a promising target to study. When the 
346: semiconductors change into metal phase by field induced charge injection, 
347: metallic and semiconducting regions sit side by side 
348: with only an atomic distance apart in the organic film. 
349: Under this situation, free carriers can interact with exciton, 
350: associated with the semiconductor, at the interface. 
351: Ginzburg {\it et al.}\cite{14} proposed the possibility of superconductivity by exciton 
352: mechanism at metal/semiconductor interfaces. 
353: They discussed that the coexistence (in real space) of 
354: excitons and metallic carriers would make it possible interface superconductivity. 
355: Therefore, the situation realized at interface thus leads a possible ground 
356: for superconductivity by exciton mechanism.
357: 
358: \section{CONCLUSIONS}
359: \label{sec4}
360: We present the {\it in-situ} and continuous measurement of the conductivity 
361: of growing organic films, as a simple but powerful method to determine 
362: the distribution curve of injected carriers and the 
363: dimension of the accumulation and depletion layers. 
364: The depletion layer thickness could be directly 
365: determined as a shift of the threshold thickness at which electric current began to flow. 
366: The {\it in-situ} and continuous measurement 
367: could also determine qualitatively the accumulation layer thickness 
368: together with the distribution function of injected carriers. 
369: The accumulation layer thcikness was 2 nm for pentacene and 9 nm for 6T.
370: The accumulation layer thickness did not depend on gate voltages, 
371: rather depended on the chemical species.
372: 
373: \acknowledgments{This work was supported by the Grant-in-Aid 
374: from the Ministry of Education, Culture, Sports, 
375: Science and Technology of Japan (14GS0207).}
376: 
377: \begin{references}
378: 
379: 
380: \bibitem{1}S.M. Sze, Physics of Semiconductor Devices (John Wiley and Sons, New York, 1985)
381: \bibitem{2}H. Ohno, D. Chiba, F. Matsukura, T. Omiya, E. Abe, T. Dietl, 
382: Y. Ohno and K. Ohtani, Nature {\bf 408} (2000) 944.
383: \bibitem{3}C. H. Ahn, S. Gariglio, P. Paruch, T. Tybell, L. Antognazza, and J.-M. Triscone,
384:  Science {\bf 284} (1999) 1152.
385: \bibitem{4}The depletion layer thickness can be indirectly determined by the capacitance measurement. 
386: When the depletion layer is formed at the semiconductor/insulator interface, the capacitance of the system 
387: is $ C_{i}C_{d}/(C_{i}+C_{d})$, where $C_{i}$ and $C_{d}$ are the capacitance of the insulating and deletion layers. 
388: The capacitance of the depletion layer is determined using the known capacitance of the insulating layer. 
389: Therefore, the depletion layer thickness ($L_{d}$) can be determined by equation $L_{d}=\epsilon/C_{i}$, 
390: where $\epsilon$ is the dielectric constant of the semiconductor. 
391: On the other hand, the accumulation layer thickness is hard to be determined. 
392: In most of previous studies of organic FET, it is assumed that the whole charge (=$C_{i}V_{G}$) 
393: is injected only in the first layer at the interface. 
394: \bibitem{5}A. Dodabalapur, L. Torsi, and H.E. Katz, Science {\bf 268} (1995) 270.
395: \bibitem{6}M. Kiguchi, M. Nakayama, K. Fujiwara, K. Ueno, T. Shimada and K. Saiki, 
396: Jpn. J. Appl. Phys., 
397: {\bf 42} (2003) L1408.
398: \bibitem{7} We have measured the morphology of the film by atomic force microscopy (AFM). 
399: For the 0.7 nm thick film, the first monolayer pentacene molecules covered 
400: the SiO$_{2}$ substrate completely with their molecular long axes parallel to 
401: the surface ("lying"), and some molecules grew on the first layer with their molecular 
402: long axes normal to the surface ("standing"); M. Nakayama, M. Kiguchi, K. Saiki, 
403: unpublished results.
404: \bibitem{8}E.H. Rhoderick, Monographs in electriccal and electronic engineering. 
405: Metal-semiconductor contacts. (Clarendon press, Oxford, UK 1978)
406: \bibitem{afm} This assumption is supported by the AFM results. 
407: The film nearly covered the substrate completely, although there were some grain boundaries. 
408: Furthermore the deviation of height in the island was about 2 nm, 
409: which corresponded to 1 monolayer thickness. These experimental results supports 
410: that the nearly ideal layer growth occurred under the present growth condition.
411: \bibitem{9}T. Komoda, Y. Endo, K. Kyuno and A. Toriumi, 
412: Jpn. J. Appl. Phys. {\bf 41} (2002) 2767.
413: \bibitem{10}R. C. Haddon, A. S. Perel, R. C. Morris, T. T. M. Palstra, A. F. Hebard, 
414: and R. M. Fleming, 
415: Appl. Phys. Lett. {\bf 67} (1995) 121.
416: \bibitem{11}The carrier density ($n(x)$) could be also obtained 
417: by dividing the local conductance by the mobility $\mu _{f}(x)$, 
418: which is determined from the FET characteristics at each film thickness. 
419: In this case, $b$ is 0.86 nm at -15 V, and 0.91 nm at -30 V for pentacene, 
420: $b$ is 3.4 nm at -15 V, and 3.9 nm at -30 V for 6T, 
421: which are smaller than $b$ in the text. In determining $\mu _{f}$, 
422: the carriers($Q=CV$) are assumed to be completely injected 
423: even for an ultra thin film. The injected carriers are, thus, 
424: overestimated, leading to underestimation of $\mu _{f}$. 
425: On the other hand, constant mobility ($\mu _{c}$) in the text overestimates 
426: the mobility especially in the thin film region. Therefore, the true mobility is 
427: between $\mu _{f}$ and $\mu _{c}$, and the true $b$ is between those estimated 
428: from $\mu _{f}$ and $\mu _{c}$. 
429: The main topic of the present work is based on the fact that $b$ does not depend on $V_{G}$, 
430: which was commonly observed in both cases. 
431: Therefore, the discussion in the text was not affected by choice of $\mu$.
432: \bibitem{12}G. Horowitz, M.E. Hajlaoui, and R. Hajlaoui, J. Appl. Phys. {\bf 87} (2000) 4456. 
433: \bibitem{13}The carrier density is represented as  $p=n_{i}exp((E_{i}-E_{F})/kT)$, 
434: where $n_{i}$, $E_{i}$, $E_{F}$ are intrinsic carrier density, 
435: intrinsic Fermi level, and Fermi level. Since the carrier density changes exponentially 
436: with the energy difference between intrinsic Fermi level and Fermi level, 
437: the small difference in the band bending would lead to large difference 
438: in distribution function of carriers. The band 
439: bending depends on dielectric constant of semiconductors.
440: \bibitem{14}Ginzburg, V.L. and Kirzhnits, D.A 
441: High-Temperature Superconductivity (Consultants Bureau, New York, 1982).
442: 
443: \end{references}
444: 
445: \begin{figure}
446: \begin{center}
447: \leavevmode\epsfysize=40mm \epsfbox{fig1.eps}
448: \caption{Schematic sample layout. The source (S)-drain (D) current was 
449: measured {\it in-situ}  at various gate voltages.}
450: \label{fig1}
451: \end{center}
452: \end{figure}
453: 
454: \begin{figure}
455: \begin{center}
456: \leavevmode\epsfysize=50mm \epsfbox{fig2.eps}
457: \caption{Source-drain current ($I_{SD}$) of pentacene and 6T measured 
458: as a function of film thickness at various gate voltages ($V_{G}$). 
459: Source-drain voltage was kept at 1 V. The dotted line is guideline to 
460: obtain the threshold thickness.}
461: \label{fig2}
462: \end{center}
463: \end{figure}
464: 
465: \begin{figure}
466: \begin{center}
467: \leavevmode\epsfysize=60mm \epsfbox{fig3.eps}
468: \caption{(a) Carrier density ($n(x)$) of pentacene as a function of the distance ($x$) 
469: from the interface at various gate voltages ($V_{G}$); 
470: $V_{G}$=-30 V(down-triangle), -15 V(upper-triangle), 0 V(circle). 
471: (b) The difference ($n_{i}(x)$) between charge carrier density at $V_{G}$=-15 V or -30 V 
472: and that at $V_{G}$=0 V.}
473: \label{fig3}
474: \end{center}
475: \end{figure}
476: 
477: \begin{figure}
478: \begin{center}
479: \leavevmode\epsfysize=60mm \epsfbox{fig4.eps}
480: \caption{ (a) Carrier density ($n(x)$) of 6T as a function of the distance ($x$) 
481: from the interface at various gate voltages ($V_{G}$); $V_{G}$=-30 V(down-triangle), 
482: -15 V(upper-triangle), 0 V(circle). (b) The difference 
483: ($n_{i}(x)$)  between charge carrier density at $V_{G}$=-15 V or -30 V 
484: and that at $V_{G}$=0 V.}
485: \label{fig4}
486: \end{center}
487: \end{figure}
488: 
489: 
490: \end{multicols}
491: \end{document} 
492: 
493: 
494: