3.5 Analysis Instruments
4.1.4 TCAD SILVACO ATLAS Simulation
After analyzing the experimental gas sensing response, TCAD SILVACO ATLAS
44
software is used to simulate the potential distribution in the vertical channel and the ideal transfer characteristics of SCLT. The simulation includes self-consistence of electron distributions and uniform doping distribution in the SCLT.
NH3 Sensing Simulation
The influences of electron doping (e-doping) on SCLT are investigated to reflect the gas sensing response. Material parameters of TCAD simulation were defined in Ref.4.9.
Figures 4-13(a) and 4-13(b) show 2-dimensional potential profiles of the vertical SCLT channel for e-doping concentrations of 1015 and 1016, respectively. SCLT is biased in off state with VCE = -1.2 V and VBE = 1.5 V. It is observed that increasing e-doping concentration in P3HT results in the increase of the potential barrier, particularly in the central region of the vertical channel. The potential distributions along the central vertical channel of bottom and top injection structures with various e-doping concentrations are plotted in Fig. 4-14. For bottom injection, the potential distribution is from top Al (C) to bottom ITO (E); for top injection, it is from top Al/MoO3 (E) to bottom ITO (C). From Fig. 4-14, the potential of both top and bottom injection structures increases with the increase of e-doping concentration. Although there are slight differences in the increase of potential with e-doping concentration as 1016 cm-3, the reason why bottom injection structure exhibits better sensitivity to NH3 is still unclear. cThe corresponding ideal JCE - VBE curves are shown in Fig. 4-15. With a fixed VCE as -1.2 V, JCE-VBE curves shifts to the left when e-doping concentration increases,
45
indicating an increase of the potential barrier in the vertical channel. This simulation results agrees well with the sensing response for JCE-VBE described in Fig. 4-7. The current density variation ratios (△J/J0) as a function of VBE for various doping concentrations are plotted shown in Fig. 4-16. The current density variation ratio (△J/J0), representing the response sensitivity, is defined as mentioned in Eq. 4.1-1, while JCE0 here is the collector current density at a electron doping concentration of 1015. It was found that △J/J0 is strongly dependent on VBE and the maximum △J/J0 is obtained in the switching region (e.g. 0.5 V < VBE < 2 V in this case). As anticipated, the simulated results agree well with the experimental results in Fig. 4-8. As aforementioned, when SCLT is biased in the switching region, a slight change of base potential causes a significant current variation. The maximum △J/J0 in the switching region indicates that porous SCLT biased in the switching region exhibits largest response to e-doping as well as the gas molecule interaction.
NO Sensing Simulation
In addition to simulate the sensor response to e-doping agents like NH3, the response to e-dedoping agents is also investigated. Material parameters of TCAD simulation were also defined in Ref.49. Figures 4-17(a) and 4-17(b) show 2-dimensional potential profiles of the vertical SCLT channel for e-dedoping concentrations of 1015 and 1016, respectively. SCLT is biased in off state with VCE = -1.2 V and VBE = 0 V. It is observed that increasing e-dedoping concentration in P3HT results in the decrease of the potential barrier, particularly in the central
46
region of the vertical channel. The potential distributions along the central vertical channel from top Al (C) to bottom ITO (E) with various e-dedoping concentrations are plotted in Fig.
4-18. The corresponding ideal JCE - VBE curves are shown in Fig. 4-19. With a fixed VCE as -1.2 V, JCE-VBE curves shifts to the right and JCE continuously increases with e-dedoping concentration increases. Eventually, the JCE becomes so high that the transistor cannot be turned off. This simulation results agrees well with the sensing response.
4.2 Amorphous IGZO TFT Hybrid Sensor
In this section, a concept of applying a-IGZO TFT on gas sensing by capping an inorganic sensing layer of metal-oxide semiconductor material (W18O49) is demonstrated. The test gases here are nitric oxide (NO) and ammonia (NH3). The lowest detectable ammonia concentration is 50 ppb for tungsten oxide (W18O49) capped a-IGZO TFT hybrid sensor.
Besides, 3-10 ppm nitric oxide gas sensing were also demonstrated.
Fig. 4-20 shows the transfer characteristics (IDS -VGS) of standard (STD) a-IGZO TFT and W18O49-capped a-IGZO TFT where VDS is biased at 20 V. The channel length and width are 200 μm and 1000 μm respectively Parameters of a-IGZO TFT are shown in the inset of Fig. 4-20. It is observed that, with W18O49 capping, IDS of a-IGZO TFT increases significantly and the transistor cannot be turned off. To solve this problem, we decrease the drain voltage so that the transistor can be turned off.
47
4.2.1 Ammonia Sensing Results
In our previous work, for STD a-IGZO TFT, no significant response can be found when ammonia concentration is 0.8 or 10 ppm [50]. The response of W18O49-capped a-IGZO TFT to ammonia is investigated. Drain current as a function of time is plotted in Fig. 4-21. The ammonia concentration here is 50-1000 ppm and the device is biased at VDS = VGS = 3V.
Besides, the W18O49 is capped by spin coating. There is an obvious current drop when the device is exposed in ammonia. The sensing mechanism is supposed that ammonia molecules absorbed onto W18O49 act like reducing agents to inject electrons into W18O49 [51]. The negative potential on W18O49 layer, as the second gate, shifts the threshold voltage to be more positive and produces a current drop in a-IGZO TFT.
4.2.2 Nitric Oxide Sensing Results
To further verify the aforementioned mechanism, a typical oxidizing gas, NO, is used to stimulate the W18O49-capped device. In our previous work, when STD device is exposed to NO gas, a slow current drop is observed and the current drop is not able to be recovered even when NO is removed. Drain current as a function of time is plotted in Fig. 4-22, and W18O49 sensing layer here is formed by drop. The NO concentration is 3 to 10 ppm, the operating temperature is 80℃, and the device is biased at VDS = VGS = 0.5V. It is observed that the drain current increases when the device is exposed to NO and a recovery behavior is also obtained when NO is removed. The W18O49 capping blocks the reaction between NO and a-IGZO film.
48
The sensing behavior is dominated by the reaction between NO and W18O49 in which NO molecules act as oxidizing agents to withdraw electrons from W18O49 [51]. The positive potential on W18O49 film helps to turn on the transistor and thus increase the current of a-IGZO TFT.
49
Figures of Chapter 4
Fig. 4-1 The diode characteristics of porous SCLT, including EC, EB and BC diode. (a) EC diode characteristics between emitter and collector where emitter is biased and collector is grounded. (b) EB diode characteristics between emitter and base where emitter is biased and base is grounded. (c) BC diode characteristics between base and collector where base is biased and collector is grounded.
50
Fig.4-2 Output characteristics, the collector current density (JCE) as a function of the collector voltage (VCE), of the porous SCLT. With collector bias (VCE) as -2.4 V and -1.2 V, porous SCLT exhibits an on/off current ratio as 4750 and 890, and a switching swing as 140 mV/dec.
and 122 mV/dec., respectively.
Fig.4-3 Transfer characteristics, the collector current density (JCE) as a function of the base voltage (VBE), of the porous SCLT.
51
Fig. 4-4 A plot of sensitivity-NH3 concentration with different VCE bias conditions. The NH3 concentrations were ranged from 30 ppb to 1000 ppb; the VCE bias conditions are -1.2 V, -1.8 V and -2.4 V.
Fig. 4-5 A plot of sensitivity as a function of NH3 concentration with different injection of holes.
52
Fig. 4-6 Porous SCLT with different injection of holes. (a) Bottom injection of holes where ITO acts as emitter and Al acts as collector. (b) Top injection of holes where MoO3/Al acts as emitter and ITO acts collector.
53
Fig. 4-7(a) A plot of JCE-VBE, representing the porous SCLT’s sensing response to NH3. VCE was fixed as -1.2 V and the NH3 concentrations were ranged from 30 ppb to 1000 ppb; (b) the response of the switching region (VBE = -0.4 V to 0 V) of JCE-VBE.
54
Fig. 4-8 The sensing sensitivities △J/J0 as a function of VBE for various NH3 concentrations;
(b) Maximum sensitivity as a function of NH3 concentration of SCLT sensor and diode sensor.
55
Fig. 4-9(a) Effects of bias stress and sensing response on JEC-VBE curve. Porous SCLT is bias as VBE = -0.2 V and VCE = -1.2 V and NH3 concentration is 100ppb for 200 seconds. (b) Maximum sensitivity as function of the time after finishing of bias stress and NH3
concentration is 100 ppb.
56
Fig. 4-10 The switching function of the porous SCLT under NH3 sensing.
Fig. 4-11 The real-time NH3 recovering response at VBE = -0.9 V, 0 V and 0.9 V.
57
Fig. 4-12 The real-time sensing response of JCE to nitric oxide (NO).
Fig. 4-13 2-dimensional potential profiles of the vertical SCLT channel for e-doping concentrations. (a) e-doping concentrations of 1015. (b) e-doping concentrations of 1016. SCLT is biased in off state with VCE = -1.2 V and VBE = 1.5 V.
58
Fig. 4-14 The potential distributions along the central vertical channel with various e-doping concentrations. (a) Top injection: from top Al (C) to bottom ITO (E); (b) bottom injection:
from top Al/MoO3 (E) to bottom ITO (C).
Fig.4-15 The corresponding ideal JCE - VBE curves.
59
Fig. 4-16 The current density variation ratios (△J/J0) as a function of VBE for various doping concentrations.
Fig. 4-17 2-dimensional potential profiles of the vertical SCLT channel for e-dedoping concentrations. (a) e-dedoping concentrations of 1015. (b) e-dedoping concentrations of 1016. SCLT is biased in off state with VCE = -1.2 V and VBE = 0 V.
60
Fig. 4-18 The potential distributions along the central vertical channel from top Al (C) to bottom ITO (E) with various e-dedoping concentrations.
Fig.4-19 The ideal JCE - VBE curves.
61
Fig.4-20 Transfer characteristic of STD a-IGZO TFT and W18O49-capped a-IGZO TFT. The channel length and width are 200 μm and 1000 μm respectively. Parameters of a-IGZO TFT are shown in the inset of Fig. 4-20.
Fig. 4-21 Drain current as a function of time. The ammonia concentration here is 50-1000 ppm and the device is biased at VDS = VGS = 3V. Besides, the W18O49 is capped by spin coating.
62
Fig. 4-22 Drain current as a function of time is plotted in and W18O49 sensing layer here is formed by drop. The NO concentration is 3 to 10 ppm, the operating temperature is 80℃, and the device is biased at VDS = VGS = 0.5 V.
63