Chapter 2 Theory of ISFET
3.3 Measurement system
3.3.2 Setup of measurement
In the beginning of the measurement, the reference electrode is suspended on the air over the container. Then, pH-solution is filled in the container. It is noticed that the pH-solution must touch the sensing layer entirely because of the small opening.
In the setup of HP-4156, substrate voltage is ground and the reference electrode is sweeping to different voltage. In the measurement of sensitivity, the response of the
pH-ISFET is the function of time. According to P. Woias [10], the first equilibrium achieves in a minute. Therefore, we measured the I-V curves after 4 minutes, which the pH-ISFET had been covered by the pH-solution. Because of the drift, we measured again after two minutes later to make sure that the response was saturated.
The pH-solution in the container is about several milliliters. In order to control the accuracy of the result, the container has to be washed by the next pH-solution after measuring previous pH-solution.
In the measuring of the drift, we apply a constant gate voltage for a period of time. The drift changes fast at beginning but changes slowly several hours later.
According to this phenomenon, we can observe the reaction on the surface and in the barrier. Total measurement time is 43100 seconds and Vg depends on the Id current we choose.
3.4 References
[1] U. Guth,
“
Investigation of corrosion phenomena on chemical microsensors”, Electrochimica Acta 47 pp. 201–210 , 2001.[2] George T. Yu, “Hydrogen ion diffusion coefficient of silicon nitride thin films”, Applied Surface Science 202 pp.68–72, 2002.
[3] Y.Vlasov, “Investigation of pH-sensitivity ISFETs with oxide and nitride membranes using colloid chemistry method”, Sensors and Actuators B, 1 pp.357–360 1990.
[4] P. Hein, “Drift behavior on ISFET with nitride gate insulator”, Sensors and Actuators B, 13-14 pp.655–656 1993.
[5] R.M. Cohen,A study of insulator materials used on ISFET gates, Thin Solid Film, 53 pp.169-173 1978.
[6] K. Najafi, “A high yield IC-compatible multichannel recording array, IEEE Trans.
Electron. Devices 32, pp.1206-1211, 1985
[8] Hung-Kwei Liao, “Multi-structure ion sensitive field effect transistor with a metal light shield”, Sensors and Actuators B61 pp.1–5 1999
[7] I-Yu Huang, “Fabrication and characterization of a new planar solid-state reference electrode for ISFET sensors”, Thin Solid Films 406 pp.255–261 2002 [9] P. Bergveld, “How electrical and chemical requirements for REFETs may coincide”, Sensors and Actuators 18 pp.309-327, 1999
[10] P. Woias, “Slow pH response effects of silicon nitride ISFET sensors”, Sensors and Actuators B 48, pp.501-504, 1998.
Chapter 4
Results and discussions
4-1 The drift and sensitivity characteristic for TaN
According to I-V curves (Fig4-1), we find some characteristics in our pH-ISFET.
When the TaN barrier is under the PE-oxide, the sensitivity is higher and more linear.
In the standard pH-ISFET with sensing layer PE-oxide, the sensitivity is very unstable and the range is usually from 30-50 mV/pH [1]. However, the sensitivity in our experiment approaches to 61.2 mV/pH (Fig.4-2) after inserting TaN barrier.
On the other hand, we observe the first and the second measurement at 4 and 6 minutes. Sensitivities are different because of the fast response has not completed [2].
The sensitivity is higher after a long period of time (Fig. 4-3).
In the discussion of the drift, the standard pH-ISFET with sensing layer PE-oxide always unstable even after several hours [3]. After inserting a TaN barrier, the drift becomes stable (Fig.4-4). It is linear after 10000 seconds although the drift approaches to 18mV/hour. By observing the curve, we find the Vth is graduating lower by the time. We explain the situation for the H+ ions diffusing in the sensing layer and barriers. In the standard pH-ISFET, the drift is unstable because the PE-oxide cannot block ions [4]. After inserting TaN barrier, the ion will be blocked by it.
4-2 The drift and sensitivity characteristics for W
Although TaN can improve the stability of ISFET, the drift is still not good enough. W and WN are barriers we choose to improve the characteristics. For W, the
drift can approach to 3.6mV/hour (Fig. 4-5). Moreover, WN barrier can reduce the drift to 1.08 mV/hour (Fig. 4-6). The results show that different barriers can change the drift characteristic. WN barrier is the best barrier to show the lowest drift in our experiment
W and WN barrier show lower drift. However, there are different in the response of sensitivities. I-V curves for the W and WN are shown in Fig.4-7 and Fig.4-8.
The sensitivity for W from pH1 to pH7 is 49.1 mV/pH (Fig.4-9) but from pH13 to pH7 is 69.1 mV/pH (Fig.4-10). The difference is close to 20 mV/pH. However, there is an opposite situation in the WN. The sensitivity from pH1 to pH7 is 75.8(Fig.4-11) mV/pH but from pH13 to pH7 is 36 mV/pH (Fig.4-12) in WN. Up to now, there is not a clear expression for this phenomenon. In standard theory [5], the surface reaction is the only factor for sensitivity. However, PE-oxide is not a good diffuse barrier so that the diffusion phenomenon has to be considered.
After inserting W and WN barrier under the pH-ISFET with sensing layer low-pressure nitride, the sensitivity is stable (Fig.4-13, Fig.4-14). The drift of WN barrier approaches to 1.44 mV/hour (Fig.4-15) and W barrier approaches to 2.52 mV/hour (Fig.4-16). The drift in the standard pH-ISFET with sensing layer LP-nitride is about 1-2 mV/hour. According to the result, we can conclude that the barrier is more useful to PE-nitride.
4-3 Conclusions
After comparing different barriers, we can give some conclusions:
1. For different barriers, the characteristic of pH-ISFET with sensing layer PE-oxide can be improved.
2. WN is the best barrier for the drift improving. The drift is 1.08 mV/hour when the
WN is under the PE-oxide.
3. WN and W barrier have different sensitivities in different pH range.
4-4 References
[1] Li-Te Yin, “Study of indium tin oxide thin film for separative extended gate ISFET”, Materials Chemistry and Physics 70, pp.12–16 2002.
[2] P. Bergveld, “Development of an ion sensitive solid-state device for neurophysiological measurements”, IEEE Trans. Biomed. Eng. 17 pp.70–71,1970 [3] George T. Yu, “Hydrogen ion diffusion coefficient of silicon nitride thin films”, Applied Surface Science 202 pp.68–72, 2002.
[4] H.C.G. Ligtenberg, “solutions for some basic problems of ISFET sensors”, Thesis, Twenty University of Technology, the Netherlands, 1987.
[5] L. Bousse., The role of buried site OH sites in the response mechanism of pH-ISFETs”, Sensors and Actuators B6, pp.65–78 1984.
Chapter 5 Future work
5-1 Future work
In our experiment, barriers can improve the characteristics of the pH-ISFET with sensing layer PE-oxide. Similarly, the other sensing layers with porous structures can be added barriers under the sensing layers. We suppose that is more stable than the original structure.
Although W, TaN and WN get better characteristics, we can attempt other barriers to identify their drifts and sensitivities. Perhaps there will be some barriers better than WN.
In the applications of the ISFET, it always companies with other systems. For example, FIA system contains the channels, ISFET array and circuit on a chip. In our group, we have had the techniques in the channel and circuits. The most important we will do in the future is to integrate all devices in a chip. Although there are still some problems, we believe it is an achievable target in the future.
Figure 1-1 An integral ISFET system
Figure 1-2 Basic diagram of a back-side contacted ISFET
Figure 1-3 Flow injection system
Figure 2-1 Potential profile and charge distribution at an
oxide electrolyte solution interface
Figure 2-2 Cross section of the standard ISFET
Figure 2-3 Gate layers of intermediate gate ISFET
Figure 3-1 Corresponding graph for process
Figure 3-1 Corresponding graph for process
Production method: CZ Type/ Dopant: P/BO
Crystal Axis: <100>
Resistivity (ohm-cm):0-100 Diameter (mm): 99.5-100.5 Thickness (μm) ：500-550
Figure3-2 Specifications of Wafer
parameters of W sputter parameters of WN sputter
power 150W power 150W
Ar:24sccm Ar: 24sccm N2=4.8sccm rate=0.3A/s rate=0.4A/s
pre sputter 10min 60W pre sputter 10min 60W pressure=7.6×10-3 pressure=7.6×10-3
parameters of Ta sputter parameters of TaN sputter
power 500W power 500W
gas:Ar gas Ar:N2=47:2.6 rate=2.07A/s rate=2.146A/s
pre sputter 5min 300W pre sputter 5min 300W pressure=7.6×10-3 pressure=7.6×10-3
Figure 3-3 Parameters of barriers deposition
parameters of LP-nitride deposition
NH3= 17sccm SiH2Cl2=85sccm 18min =1018A
temperature=850C pressure=180mT
n=2.321
parameters of PE-oxide deposition
temperature=300C TEOS=10sccm
2000A=4 ' 44''
Figure 3-4 Parameters of sensing layers deposition
HDP-RIE
process pressure=10m torr flow rate of CHF3=40sccm
flow rate of Ar=40sccm ICP power=600
Bias power=150 etch time(min)
TaN 3'
Ta 2'
WN 3'
W 2'
LP 3'
PE 2'
Figure 3-5 Recipe of HDP-RIE etch rate
Figure 3-6 System of measurement
-1 0 1 2 3 4
Figure 4-1-1 I-V curves of PE-oxide/TaN from pH1-pH13 after 4 minutes dipping
-1 0 1 2 3 4
Figure 4-1-2 I-V curves of PE-oxide/TaN from pH1-pH13 after 6 minutes dipping
0 2 4 6 8 10 12 14 Figure 4-2 Sensitivity of PE-oxide/TaN after 4 minutes dipping
0 2 4 6 8 10 12 14
Figure 4-3 Sensitivity of PE-oxide/TaN after 6 minute dipping
0 10000 20000 30000 40000 0.2
0.4 0.6 0.8 1.0 1.2 1.4
Vg(V)
time(s)
drift=18mV/hour after 10000s
Figure 4-4 Drift of PE-oxide/TaN at pH7, Vg at Id=300µA
0 10000 20000 30000 40000 50000
0.2 0.4 0.6 0.8 1.0 1.2 1.4
drift=3.6mV/hour after 10000s
Vg(V)
time(s)
Figure 4-5 Drift of PE-oxide/W at pH7, Vg at Id=300µA
0 10000 20000 30000 40000 50000
drift=1.08 mV/hour after 10000s
Vg(V)
time(s)
Figure 4-6 Drift of PE-oxide/WN at pH7, Vg at Id=300µA
-2 -1 0 1 2 3
Figure 4-7-1 I-V curves from pH1-pH7 for PE-oxide/W barrier after 4 minutes dipping
-2 -1 0 1 2 3
Figure 4-7-2 I-V curves from pH1-pH7 for PE-oxide/W barrier after 6 minutes dipping
-2 -1 0 1 2 3
Figure 4-7-3 I-V curves from pH7-pH13 for PE-oxide/W barrier after 4 minutes dipping
-2 -1 0 1 2 3
Figure 4-7-4 I-V curves from pH7-pH13 for PE-oxide/W barrier after 6 minutes dipping
-1 0 1 2 3
Figure 4-8-1 I-V curves for PE-oxide/WN from pH1-pH7 after 4 minutes dipping
-1 0 1 2 3
Figure 4-8-2 I-V curves for PE-oxide/WN from pH1-pH7 after 6 minutes dipping
-1 0 1 2 3
Figure 4-8-3 I-V curves for PE-oxide/WN from pH7-pH13 after 4 minutes dipping
-1 0 1 2 3
Figure 4-8-4 I-V curves for PE-oxide/WN from pH7-pH13 after 6 minutes dipping
1 2 3 4 5 6 7
Figure 4-9-1 Sensitivity of PE-oxide/W from pH1-pH7 after 4 minutes dipping.
1 2 3 4 5 6 7
Figure 4-9-2 Sensitivity of PE-oxide/W from pH1-pH7 after 6 minutes dipping.
7 8 9 10 11 12 13
Figure 4-10-1 Sensitivity of PE-oxide/W from pH7-pH13 after 4 minutes dipping.
7 8 9 10 11 12 13
Figure 4-10-2 Sensitivity of PE-oxide/W from pH7-pH13 after 6 minutes dipping.
1 2 3 4 5 6 7
Figure 4-11-1 Sensitivity of PE-oxide/WN from pH1-pH7 after 4 minutes dipping.
1 2 3 4 5 6 7
Figure 4-11-2 Sensitivity of PE-oxide/WN from pH1-pH7 after 6 minutes dipping.
Figure 4-12-1 Sensitivity of PE-oxide/WN from pH7-pH13 after 4 minutes dipping.
7 8 9 10 11 12 13
Figure 4-12-2 Sensitivity of PE-oxide/WN from pH7-pH13 after 6 minutes dipping,
Figure 4-13-1 I-V curves of LP-nitride/W from pH1 to pH7 after 4 minutes dipping
-2 -1 0 1 2 3
Figure 4-13-2 I-V curves of LP-nitride/W from pH1 to pH7 after 6 minutes dipping
Figure 4-13-3 I-V curves of LP-nitride/W from pH7 to pH13 after 4 minutes dipping
-2 -1 0 1 2 3
Figure 4-13-4 I-V curves of LP-nitride/W from pH7 to pH13 after 6 minutes dipping
Figure 4-14-1 Sensitivity of LP nitride/W from pH1-pH13 after 4 minutes dipping
0 2 4 6 8 10 12 14
Figure 4-14-2 Sensitivity of LP nitride/W from pH1-pH13 after 6 minutes dipping,
0 10000 20000 30000 40000 50000
0.2
Figure 4-15 Drift of LP-nitride/WN at pH 7, Vg at Id=300µ
0 10000 20000 30000 40000 50000 0.2
0.4 0.6 0.8 1.0 1.2 1.4
time(s)
drift=2.52 mV/hour
Vg(V)
Figure 4-15 Drift of LP-nitride/W at pH 7, Vg at Id=300µ