Chapter 2 Experiment
2.4 Characteristic measurement of devices
We use HP 4284A precision LCR meter parameter to analyze Capacitance-Voltage (C-V) characteristic diagrams at 1MHz and the characteristic curves of Current-Voltage (I-V) are measured with semiconductor parameter analyzer by HP 4156. We measure all measurements are at room temperature in an air atmosphere.
Figure 2-1 Schematic of a corona discharge.
Figure 2-2 Schematic of a silent discharge (1) metallic electrodes and (2) dielectric barrier coating.
Figure 2-4 Schematic of the atmospheric-pressure plasma jet for the deposition of silica films.
Figure 2-5. Numerical simulation of the concentrations of species in the effluent of the plasma jet as a function of the distance from the nozzle in a 1.0%
O2/He plasma at 760 torr.
Source
Breakdown Plasma density Density(cm-3) voltages(KV) (cm-3) O+,O2+,O- O O3
Corona plasma 10~50 109~1013 1010 1010 1018
Dielectric barrier discharge 5~25 1012~1015 1010 1012 1018
plasma torch 10~50 1016~1019 1015 1018 <1010
plasma jet 0.05~0.2 1011~1012 1012 1016 1016 Table 2-1 Densities of oxygen species in the various plasma discharges.
Figure 2-7 Schematic of the atmospheric-pressure plasma jet for the deposition of silicon oxide.
Experimental parameters
Main gas CDA, N2, O2
Speed (mm/sec) 30
Gap distance (cm) 2.2
Substrate temperature (℃) 150
Ar flow rate (sccm) 100
Scanning times 60 or 15 (N2, O2)
Table 2-2 shows the detail experimental parameters of silicon oxide deposited with different kinds of main gases.
Experimental parameters
Table 2-3 shows the detail experimental parameters of silicon oxide deposited with different kinds of gap distances.
Table 2-4 shows the detail experimental parameters of silicon oxide deposited with different kinds of Ar flow rates.
Figure 2-9 shows the schematic of organic thin film transistor structure with top contact.
Chapter 3
Results and Discussion
Low temperature processes and good quality gate insulator are urgent for organic thin film transistor . We utilized the atmospheric pressure plasma jet to deposit silicon oxide as the gate insulator of organic thin film transistor due to the cold chemically active species could reduce the thermal damage on the substrate. In addition, APPJ dose not require vacuum system which could overcome size limit and reduce cost of equipments.The gate dielectric deposited by APPJ shows low leakage current density at the range of 1.59E-8~3.85E-8 A/cm2 at 0.5 MV/cm, and it is comparable to some gate dielectrics deposited in a vacuum chamber such as sputter and electron-beam evaporation .
3.1 Result of different conditions
3.1.1 The influence of different main gases
We utilized oxygen, nitrogen, and CDA as the main gas of APPJ, to decompose tetraethoxysilane (TEOS), Si(OC2H5)4. Oxygen, nitrogen, and CDA were used for investigating the influence of different main gases of APPJ on deposition rate and quality of silicon oxide. Leakage current density versus electric field of silicon oxide deposited with different kinds of main gases was shown in Figure 3-1, and The deposition rate and thickness show in Table 3-1. The leakage current density (A/cm2) at 0.5 and 1 MV/cm with different main gases show in Table 3-2. We could find that the
O atom, metastable oxygen, and ozone [22] Show the densities of oxygen species in the plasma discharges show in Table3-3 [21]. The deposition rate of silicon oxide deposited by oxygen main gas is the fastest because oxygen has better ability to decompose TEOS vapor. However, the faster deposition rate may decrease the denseness of silicon oxide and increase the surface roughness of silicon oxide generate higher leakage current density. The silicon oxide deposited with varied main gases (CDA, Nitrogen and Oxygen respectively) and, the SEM images Figure 3-2 had the same tendency as the AFM images Figure 3-3 .We could find that the uniformity of silicon oxide fabricated with CDA main gas is better than oxygen main gas.The schematic illustration of deopsited rate versus RMS are shown in Figure 3-4 and we found that the increase of deposition rate will increases RMS of surface.The structure of silicon oxide fabricated with oxygen gas is loose, and this may cause higher leakage current density.
Silicon oxide deposited with CDA main gas has the best quality but the deposition rate is the slowest. The oxygen percentage of CDA is about 20 % which is much higher than nitrogen, but the deposition rate of silicon oxide deposited with CDA is the slowest.
It may due to that the higher oxygen percentage in nitrogen plasma has strong electronegativity which may decrease the electron density of plasma [23] and cause the decrease of the excited species. Although some researches [22] indicated that the deposition rate of silicon oxide would be proportion to oxygen partial pressure but the percentage of oxygen to total gas was below 2% in these studies. The use of pure nitrogen main gas can generate a lot of ozone because the high density nitrogen plasma generated by APPJ could excite air to create ozone and excited oxygen atoms.
X-ray photoelectron spectroscopy was used to analyze the composition of the films.
Show in Table 3-4 .The silicon oxide deposited by CDA main gas were composed of 33.12 at.% silicon, 62.45 at.% oxygen, 4.1 at.% carbon. Oxygen to silicon ratio (O/Si ratio) was satisfied nearly 2. The presence of carbon in the films was convinced to be
due to incomplete decomposition of TEOS precursor and carbon contamination from open air. In present deposition system, carbon content did not exceed 10 at.%, though films were deposited in open air. From the XPS analysis, we confirmed that nearly inorganic SiO2 layer was deposited of the atmospheric pressure plasma jet with rare carbon contaminants.
3.1.2 The influence of different gap distances
Leakage current density versus electric field of silicon oxide deposited with different kinds of gap distances was shown in Fig. 3-5, and The deposition rate and thickness show in Table 3-5. The leakage current density (A/cm2) at 0.5 and 1 MV/cm with different gap distances show in Table 3-6.
CDA was used as main gas for analyzing the influence of gap distance because the deposition rate of silicon oxide was more stable than the others. The deposition rate of silicon oxide increased with the decrease of gap distance which may due to the concentration of species and substrate temperature increased. In atmospheric pressure, the mean free path of reactive species would decrease so the concentration of reactive species near the substrate would decrease when the gap distance increases. In our case, when the gap distance increased over 2.5cm at 30 mm/s scan rate the deposition rate was almost zero. On the other hand, the temperature in the nozzle may over 150 so ℃ the substrate temperature would increase with the decrease of gap distance. The schematic illustration of deopsited rate versus RMS are shown in Figure 3-6 The experimental data indicated that the quality of silicon oxide was degraded with the increase of deposition rate.
3.1.3 The influence of different Ar flow rates
Leakage current density versus electric field of silicon oxide deposited with different kinds of Ar flow rates was shown in Figure 3-7, The deposition rate and thickness show in Table 3-7.Leakage curren denstity (A/cm2) at 0.5 and 1 MV/cm in different Ar flow rates show in Table 3-8.
Due to the Ar gas delivers precursors into the plasma reactor. After plasma dissociate it, the active sites such like Si-O bond will diffuse onto the substrate with spray flow and be deposited as silicon dioxide thin films.Therefore the Ar flow rate has relationship with amount of active sites, which can increase deposition rate of film. But the higher deposition rate doesn’t mean that the higher quality.
The SEM image of silicon oxide deposited with different Ar flow rates deposited at differents 60sccm, 100sccm, 140sccm, 200sccm and 3000sccm is shown in Figure 3-8.
The morphology of surface structure changed from smooth to rough. The AFM images Figure 3-9 had the same tendency as the SEM images. The root mean square of the surface changed from 2.84nm to 12.67 nm when the smooth surface was being changed to a rough surface. The morphology of surface roughness is also increasing with Ar flow rate. A large number of active sites accumulate on substrate quickly when Ar flow rate increase.The active sites have become deposited film before migrate to suitable position. If the active sites cannot fill in the vacancy on surface, hence that will lead into porous film and leakage current will increase.
The schematic illustration of deopsited rate versus RMS are shown in Figure 3-10 and we found that the increase of deposition rate will increases RMS of surface. The experimental data indicated that the quality of silicon oxide was degraded with the increase of deposition rate.
To summarize this section, we would choose 100 sccm as our optimal parameter.
The leakage current density of 100sccm is minimum and the morphology of surface is best between the six deposition flow rates of APPJ
.
3.2 OTFT electric characteristics discussion
According to the previous discussion, we consider the concentration of ionization species may dominate the deposition rate of silicon oxide deposited with APPJ and the quality of silicon oxide would degrade with higher deposition rate.
We selected CDA as main gas to deposit silicon oxide for the gate insulator of OTFTs. Figure 3-11 shows transfer characteristics (IDS −VG) of organic thin film transistor fabricated with low temperature silicon oxide as gate insulator deposited with atmospheric pressure plasma jet. The mobility was extracted in saturation region from the following equation: threshold voltage. Our device shows mobility was about 0.66 cm2/V‧s, threshold voltage was as low as -0.82 V, and the subthreshold swing was as low as 0.7 V/ decade.
Subthreshold swing was extracted from the following relationship:
G
Where Cd is the depletion-layer capacitance density, Cit is interface state
Output characteristics (IDS−VDS)of OTFTs are displayed in Figure 3-12, and the operation voltage was within -2~0 V which would reduce active power consumption (ID×VD) In order to reduce static power consumption, gate leakage current and source-drain leakage current must be suppressed. However, the source-drain leakage current would be a problem when the channel length scaling down for obtaining high operation speed. Moreover, the gate leakage current determined by gate insulator quality was not easily controlled at low temperature processes. Therefore, we made use of atmospheric pressure plasma jet to deposit silicon oxide at atmospheric pressure and low temperature which has low leakage current density about 1.59E-8 A/cm2..
Figure 3-1 Electric field versus leakage current density of silicon oxide with different main gases of atmospheric pressure plasma jet.
Main gas Scanning times Capacitance Thickness Deposition rate
(nF/cm2) (Å) ( Å /min‧cm2 )
CDA 60 97.93~94.74 88~91 18.2
Nitrogen 15 24.6~19.11 351~452 361.6
Oxygen 15 23.65~18.23 365~473 378.4
Main gas Leakage current density Leakage current density (A/cm2) at 0.5 MV/cm (A/cm2) at 1 MV/cm
CDA 1.59E-08 3.16E-08
N2 5.54E-06 1.63E-05
O2 4.41E-06 3.67E-05
Table 3-2 shows the leakage current density (A/cm2 ) at 0.5 and 1 MV/cm with different main gases.
Source
Density (cm-3) O+,O2+,O- O O3
Corona plasma 1010 1010 1018
Dielectric barrier discharge 1010 1012 1018
plasma torch 1015 1018 <1010
plasma jet 1012 1016 1016
Table 3-3 Densities of oxygen species in the plasma discharges.
(a)
(b)
(a)
(b)
(c)
Figure 3-3 AFM images of silicon oxide fabricated with different main gases of atmospheric pressure plasma jet. (a) CDA (b) N2 (c) O2
Figure 3-4 Deposited rate and RMS of silicon oxide deposited with different kinds of main gases of atmospheric pressure plasma jet.
Name (At. %) CDA N2 O2
Si 2p 33.12 32.8 31.13
O 1s 62.45 62.13 61.38
N 1s 0.33 0.32 0.63
C ls 4.1 4.75 6.86
Figure 3-5 Electric field versus leakage current density of silicon oxide with different gap distances of atmospheric pressure plasma jet.
Gap distance(cm) Capacitance Thickness Deposition rate (CDA main gas ) (nF/cm2) (Å) ( Å / min
‧
cm2 )1.8 161.2~159.3 67.2~71.4 14.28
2 57.17~54.94 151~157 31.4
2.2 97.93~94.74 88.3~91.5 18.2
2.5 201.43~199.84 42.9~43.1 8.6
Table 3-5 shows the deposition rate and thickness of silicon oxide deposited with different gap distances of atmospheric pressure plasma jet.
Gap distance Leakage current density Leakage current density (cm) (A/cm2) at 0.5 MV/cm (A/cm2) at 1 MV/cm
1.8 2
2.28631E-7 5.34395E-8
9.35127E-7 1.46051E-7
2.2 2.12E-8 3.41E-8
2.5 1.4615E-8 2.9388E-8
Table 3-6 shows the leakage current density (A/cm2 ) at 0.5 and 1 MV/cm with different gap distances.
Figure 3-7. Electric field versus leakage current density of silicon oxide with different Ar flow rates of atmospheric pressure plasma jet.
Ar flow Scanning Capacitance Thickness Deposition rate (sccm) times (nF/cm2) (Å) ( Å / min
‧
cm2 )60 60 202.12~186.13 42.7~46.4 9.28
100 60 91.07~89.01 94.74~96.94 19.4
140 60 82.73~78.97 104.3~109.3 21.86
200 60 65.58~63.19 131.6~136.6 27.32
300 60 55.06~53.25 156.7~162.1 32.42
Table 3-7 shows the deposition rate and thickness of silicon oxide deposited with different Ar flow rates of atmospheric pressure plasma jet.
Ar flow (sccm)
Leakage current density Leakage current density (A/cm
2) at 0.5 MV/cm (A/cm
2) at 1 MV/cm
60 1.52E-08 3.16E-08
100 2.43E-08 3.85E-08
140 3.28E-06 7.65E-06
200 9.70E-06 2.05E-05
300 5.95E-05 2.34E-04
Table 3-8 shows the leakage current density (A/cm2 ) at 0.5 and 1 MV/cm with different Ar flow rates.
(a) (b)
(c) (d)
(e)
Figure 3-8. SEM images of silicon oxide fabricated with different Ar flow rates of atmospheric pressure plasma jet. (a) 60sccm (b) 100sccm (c) 140sccm (d) 200sccm (e) 300sccm
(a)
(b)
(d)
(e)
Figure 3-9 AFM images of silicon oxide fabricated with different Ar flow rates of atmospheric pressure plasma jet. (a) 60sccm (b) 100sccm (c) 140sccm (d) 200sccm (e) 300sccm
Figure 3-10 Deposited rate and RMS of silicon oxide deposited with different kinds of Ar flow rates of atmospheric pressure plasma jet.
Figure 3-12 Output characteristics of the penecene-based OTFTs with silicon oxide deposited with atmospheric pressure plasma jet.
Chapter 4
Conclusions and Future work
4.1 Conclusions
OTFTs have been fabricated with silicon oxide as gate insulator deposited with atmospheric pressure plasma jet at low temperature. Our OTFTs showed good electric characteristics including lower operation voltage, lower threshold voltage, lower subthreshold swing, and higher mobility. Silicon oxide quality and deposition rate are strongly dependent on selection of main gas and gap distance. The quality of silicon oxide with oxygen main gas is the worst because the fast deposition rate would cause loose thin film structure. Although oxygen in CDA would reduce decomposition rate of TEOS due to its higher electronegativity but the quality silicon oxide deposited with CDA is the best. Gap distance would influence the concentration of deposition species at substrate surface. The morphology of surface roughness is also increasing with Ar flow rate. A large number of active sites accumulate on substrate quickly when Ar flow rate increase.The active sites have become deposited film before migrate to suitable position. If the active sites cannot fill in the vacancy on surface, hence that will lead into porous film and leakage current will increase.Moreover, the substrate temperature would increase when the nozzle approaches the substrate. The increase of substrate temperature may increase the deposition rate of silicon oxide and degrade the electrical quality of silicon oxide. The leakage current density of silicon oxide with CDA main gas was about 1.59E-8 A/cm2.
4.2 Future work
• We may use a low temperature plasma treatment to improve gate insulator quality.
• We may operate the APPJ in the clean room to reduce the particles in air.
• Because Pentacene OTFT are sensitive to ambient conditions. Protection from the environment by encapsulation is critical to the stability of Pentacene OTFT.
Therefore, using a suitable material as passivation to protect Pentacene film from environmental effect is another important topic
Reference
[9] C.Bartic, H. Jansen, A. Campitelli, and S. Borghs, Organic electronics, 3, 65 (2002).
[14] Chrisos D. Dimitrakopoulos, D, J. Mascaro, IBM J. RES. & DEV. 45(1), 11,
[18] I. Kymissis, C. D. Dimitrakopoulos, and S. Purushothaman, IEEE Trans.
Electron Devices 48, 1060 (2001); C. D. Dimitrakopoulos and P. R. L.
[21] Andreas Sch¨utze, James Y. Jeong, Steven E. Babayan, Jaeyoung Park, Gary S.
Selwyn, and Robert F. Hicks,IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 26, NO. 6, DECEMBER (1998)
[22] S. E. Babayan, J. Y. Jeong, V. J. Tu, J. Park, G. S. Selwyn, and R. F. Hicks, Plasma Source Sci. Technol., 10, 573 (2001).
[23] B. R. Wang, W. C. Zhu, and Y. K. Pu, Plasma Science & Technology, 1.7, 3045 (2005).