ZrO 2 gated pH-ISFETs
4.4 Summ
A proposed scheme and compensation method to improve the pH-dependent drift for ZrO2 pHISFETs was successfully demonstrated. By the sequential measurement and the post processing of the signals for both n-channel and p-channel ISFETs, the results show that the 75–100% pH-dependent drifts are significantly improved and the nearly constant drift rate versus pH value is obtained. Meanwhile, the pH measurement sensitivities are maintained and a practical integrated scheme adopting this method is also illustrated. With the integrated scheme, the calibration accuracy will be greatly improved and no complicated compensation circuit design is required.
References
[4.1] Bergveld, P. Development of an ion-sensitive solid-state device for neuro-physiological measurements. IEEE Trans. Biomed Eng. BME 1970, 17, 70-71.
[4.2] Bousse, L.; Bergveld, P. The role of buried OH sites in the response mechanism of inorganic-gate pH-sensitive ISFETs. Sens. Actuat. 1984, 6, 65-78.
[4.3] Jamasb, S.; Collins, S. D.; Smith, R.L. A physical model for threshold voltage instability in Si3N4-gate H+-sensitive FET’s (pH-ISFET’s). IEEE Trans. Electron Devices 1998, 45, 1239-1245.
[4.4] Abe, H.; Esashi, M.; Matsuo, T. ISFET’s using inorganic gate thin films. IEEE Trans. Electron Devices 1979, 26, 1939-1944.
[4.5] Topkar, A.; Lal, R. Effect of electrolyte exposure on silicon dioxide in electrolyte-oxide-semiconductor structures. Thin Solid Films 1993, 232, 265-270.
[4.6] Buck. R. P. Kinetics and drift of gate voltages for electrolyte-bathed chemically sensitive semiconductor devices. IEEE Trans. Electron Devices 1982, 29, 108-115.
[4.7] Dun, Y.; Wei, Y. D.; Wang, G. H. Time-dependent response characteristics of pH-sensitive ISFET. Sens. Actuat. B 1991, 3, 279-285.
[4.8] Esashi, M; Matsuo, T. Integrated micro multi ion sensor using field effect of semiconductor. IEEE Trans. Biomed Eng. BME 1978, 25, 84-92.
[4.9] Tsai, C. N.; Chou, J. C.; Sun, T. P.; Hsiung, S. K. Study on the time-dependent slow response of the tin oxide pH electrode. IEEE sensors journal 2006, 6, 1243-1249.
[4.10] Chou, J. C.; Wang, Y. F. Preparation and study on the drift and hysteresis properties of the tin oxide gate ISFET by the sol–gel method. Sens. Actuat. B 2002, 86, 58-62.
[4.11] Chou, J. C.; Hsiao, C N. The hysteresis and drift effect of hydrogenated amorphous silicon for ISFET. Sens. Actuat. B 2000, 66, 181-183.
[4.12] Chiang, J. L.; Jan, S. S.; Chou, J. C.; Chen, Y. C. Study on the temperature effect, hysteresis and drift of pH-ISFET devices based on amorphous tungsten oxide. Sens.
Actuat. B 2001, 76, 624-628.
[4.13] Chiang, J. L.; Chou, J. C., Chen, Y.C.; Liau, G. S.; Cheng, C. C. Drift and hysteresis effects on AlN/SiO2 gate pH ion-sensitive field-effect transistor. Jpn. J.
Appl. Phys. 2003, 42, 4973–4977.
[4.14] Bousse, L.; Van Den Vlekkert, H.H.; De Rooij, N. F. Hysteresis in Al2O3-gate ISFETs. Sens. Actuat. B 1990, 2, 103-110.
[4.15] Hendrikse, J.; Olthuis, W.; Bergveld, P. A drift free nernstian irdium oxide pH sensor. Proceedings Transducers 1997, 2, 1367-1370.
[4.16] Hein, P.; Egger, P. Drift behaviour of ISFETs with Si3N4-SiO2 gate insulator. Sens.
Actuat. B 1993, 13, 655-656.
[4.17] Bousse, L.; De Rooji, N. F.; Bergveld, P. Operation of chemically sensitive field-effect sensors as a function of the insulator-electrolyte interfac. IEEE Trans.
Electron Devices 1983, 30, 1263-1270
[4.18] Chang, K. M.; Chao, K. Y.; Chou, T. W.; Chang, C. T. Characteristics of zirconium oxide gate ion-sensitive field-effect transistors. Jpn. J. Appl. Phys. 2007, 46, 4333-4337
Figure 4-1 Schematic diagram of ZrO2 gate ISFET, which was fabricated by the Metal Oxide Semiconductor Field Effect Transistor (MOSFET) technique.
Figure 4-2 The measurement setup and a HP4156A semiconductor parameter analyzer, which was used for measuring the IDS–VGS characteristics for the ZrO2 gate ISFETs.
Figure 4-3 The sensitivity of (a) n-channel and (b) p-channel ZrO2 gate ISFETs.
Figure 4-4 The drift of the n-channel ISFET over the first seven hours at (a) pH3 (b) pH5 (c) pH7 (d) pH9 (e) pH11. The drift over the first seven hours was (a) -58.55mV
(b) -51.54mV (c) -41.61mV (d) -34.64mV and (e) -32.52mV, respectively.
Figure 4-5 The drift of the p-channel ISFET over the first seven hours at (a) pH3 (b) pH5 (c) pH7 (d) pH9 (e) pH11. The drift over the first seven hours were (a) 13.33mV
(b) 6.04mV (c) -4.91mV (d) -25.92mV and (e) -30.82mV, respectively.
Figure 4-6 The drift rate of (a) n-channel ISFET and (b) p-channel ISFET for the first seven hours.
Figure 4-7 The schematic diagram of the proposed system and measurement method.
Figure 4-8 (a) The pH-dependent drift of n-channel, p-channel and corrective ISFETs measured at the seventh hour. (b) The pH sensitivities with the proposed compensation
method.
Figure 4-9 The variations of the drift measured at different pH values from the second hour to seventh hour for n-channel, p-channel ISFETs and the proposed compensation method.
Chapter 5
Development of FET-type reference electrodes for pH-ISFET applications
5.1 Introduction
Because the ion-sensitive field-effect transistors (ISFETs) were introduced by Bergveld [5.1] in 1970, they appear to be one of the potential sensors to be used for biological applications due to the advantages of fast response, high maturity, economical scale of semiconductor process, and feasibility of miniaturization, which is attractive for in situ measurements. However, a micro-ISFET requires a miniaturized on-chip reference system, which is challenging for fabrication processes. There are two categories of on-chip reference systems under development: The first type includes miniaturized Ag/AgCl reference electrodes with encapsulated liquid reference solutions [5.2-5.4]; however, the leakage of the reference solutions limits the device lifetime and affects the measurement accuracy. The other type is reference field effect transistor (REFET), which was first proposed by Matsuo [5.5] and Janata [5.6]. Matsuo first introduced the concept of a REFET, which is an identical ISFET whose sensing layer is modified to be chemically insensitive.
Janata incorporated the REFET into a system that consisted of an ISFET/REFET pair, a quasi-reference electrode (QRE), and differential measurement circuits. QRE is simply a noble metal that replaces a complicated reference electrode; however, the potential in the metal/liquid interface is unstable because the interface is polarized. Therefore, the interfacial potential is thermodynamically undefined, which becomes the common noise of the ISFET/REFET differential system. To eliminate this common noise and maintain
measuring sensitivity, the development of REFETs should match with ISFETs to satisfy the requirements of differential measurements.
Several approaches to fabricate REFETs have been proposed. Matsuo and Nakajima [5.7] used an ion-blocking parylene as a chemically inert membrane. However, quasi-instability problems made parylene-gated REFET not practical in a differential ISFET/REFET measurement setup [5.8]. Another method to produce REFET was through a grafted membrane with a long alkyl chain silane [5.9]; nevertheless, the result shows that the gate surface sites cannot be captured sufficiently to obtain a low pH sensitivity. Van den Berg et al. reported that the total elimination of the pH-sensitive groups by chemical monolayer modification cannot be achieved [5.10]. A poly(vinyl chloride) membrane has also been used to form an ioninsensitive layer [5.11-5.17], and normally their operating range is in between pH 2 and 9. Through kinds of modification, such as silylating pretreatment [5.13], modifying a membrane by adding lipophilic cations [5.14], and using a buffered poly(2-hydroxyethl methacrylate) [5.15] or negative photoresist (PR) [5.16] layers, polymer-based low sensitivity ion-unblocking REFETs could be produced.
Nafion is a perfluorosulfonated material and is resistant to chemical attack. It has been widely used for ion exchange or immobilization applications. Nafion was used as an outer layer to immobilize the chloride ion or nanoparticles while developing electrodes of pH sensors [5.18]. It was also used for pH measurement with a decrease in oxidation–reduction potential error [5.19]. Anh et al. utilized the Nafion membrane to reduce ion interferences from the electrolyte [5.20]. Nafion is also popular at biosensor fields [5.21, 5.22], where enzymes were immobilized with Nafion. In this work, Nafion was introduced to combine with ioninsensitive polymers for REFET system fabrication. The sensitivity, electrical properties, and drift behavior were investigated.
5.2 Experiment
The Nafion solution (5 wt %) was obtained from DuPont, and pH buffer solutions were purchased from RDH (Frankfurt, Germany). Hexamethyldissilazene (HMDS) and PR (FH6400) were obtained from the Nano Facility Center, National Chiao Tung University.
The preparation of the NaCl solution was for the Na+ ion measurement. The NaCl salt was electronic grade, and solutions were prepared in deionized water with different mole concentrations, 10−3, 10−2, 10−1, and 1 M.
The fabrication of ZrO2-gated ISFETs was reported [5.23]. The ZrO2-gated ISFET fabricated by the metal-oxide-semiconductor field-effect transistor (MOSFET) technique was described in last chapters. To test the performance of REFETs, a QRE was also fabricated with Ti/Pt deposition by sputtering with a thickness of 150/350 Å.
Ion-insensitive polymer materials, PR (FH6400), and conductive polymer, poly(3-hexylthiophene) (P3HT), were incorporated to modify the ZrO2 film. Two types of membranes were prepared and tested: One type was a double-layer structure, as shown in Figure 5-1 (a); the Nafion and polymer-based materials, PR (FH6400) or P3HT, were deposited on the top of the original sensing layer (ZrO2) step by step. The other type was a single layer, as shown in Figurer 5-1 (b), with polymers entrapped in the Nafion solution and then casting on the top of the ZrO2 film. To improve the adhesion of PR and P3HT with ZrO2, thin HMDS was pretreated on the ZrO2 surface in advance. All of the above membranes were heated at 70°C for 1 min and then dried in air for over 12 hour. The epoxy resin film, which is an excellent electrical insulator, was deposited by casting on the top of
the ZrO2 layer to form the ion-blocking membrane for the electrical property comparison test.
The measurement setup for the REFET system evaluation is as described in last chapters. For the drift performance test, the same device was soaked in the buffer solution for the first and second tests, which lasted for 4 hour, respectively. The device was dried in air in between the two tests. To prevent from light influence, the measurements were entirely executed in a dark box.
5.3 Results and discussions
Figures 5-2 (a) and (b) show the pH and pNa sensitivities of the sole ZrO2 membrane and with Nafion coating, respectively. The pH and pNa sensitivities of the ZrO2 membrane with and without Nafion coating are the same. According to the study of Gorchkov [5.24], the pH response of an ISFET would be influenced by the natures of the polymeric membranes, as the consequence of the “equivalent membrane charge.” Therefore, the slightly higher sensitivity of the ZrO2 membrane with the Nafion coating could be explained as extra cations stored in the Nafion membrane that leads to a surface potential increase. However, the effect is limited. Meanwhile, Nafion is a perfluorinated polymer that can be divided into three parts: a hydrophobic fluorocarbon backbone C–F, an interfacial region of relatively large fractional void volume, and clustered regions where the majority of the ionic exchange sites, counter ions, and absorbed water exist [5.25, 5.26].
Accordingly, the Nafion membrane was a low impedance film, which unblocks the hydrogen and sodium ions; those ions from the electrolyte can pass through the Nafion membrane with a low resist and an established potential on the ZrO surface. The pNa
sensitivity was much lower than the pH sensitivity. The result reveals that it was mainly dominated by the selectivity property of the ZrO2 film, though the fact that the mobility of H+ was about 5 times that of Na+ in the Nafion membrane [5.27]. The selectivity coefficient can be determined by the methods based on the Nicolsky–Eisenman equation [5.28]. Meanwhile, the linearity of the pNa response was not as good as the pH response.
The difference might be that the Nafion membrane has a slightly higher affinity for Na+ than for H+ [5.27], which affects the interface potential between the Nafion membrane and the ZrO2 film. As shown in Figure 5-2 (c), the IDS-VGS curves were similar for both membranes except the threshold voltage shift; it revealed that ZrO2 films with or without Nafion coating have identical electrical characteristics. The REFETs produced by additional membrane coating on the top of the ISFETs should maintain the original electrical properties, such as transconductance, for differential measurements.
Table 5-1 and Figure 5-3 show the measurement results of pH and pNa sensitivities for the tested conditions. Sensitivities to the hydrogen ions for all tested polymer-based REFETs were below 10 mV/pH except P3HT without HMDS pretreatment. However, pH sensitivities of all the REFETs were reduced. The reasons for the dramatic pH sensitivity decrease can be explained by employing a site-binding model [5.29, 5.30]. The pH sensitivity of an ISFET is expressed in Equations 4-1 and 4-2. According to the equations, large values of Cdif /βint would result in a lower pH sensitivity. This means that surfaces with small buffer capacity (small value of surface reactive sites and of surface charges) show low pH sensitivity. For the double-layer polymer/Nafion as shown in Figure 5-1 (a), the sensing layers were PR and P3HT, whose surface site densities are smaller than ZrO2; therefore, the pH sensitivities were low. For the single-layer PR/Nafion composite, as shown in Figure 5-1 (b), it was long chain hydrocarbon molecules which have the characteristic of a small buffer capacity and result in a dramatic decrease in the pH
sensitivities. For the single-layer P3HT/Nafion composite, because P3HT is the conductive polymer that provides extra paths for electrons to pass through the membrane, the pH sensitivity was higher than that of the PR/Nafion composite.
HMDS is an adhesion promoter for hydrophobic polymers that attempt to attach on hydrophilic inorganic films. By comparing the pH sensitivity of the tested composite films 2 and 3, 4 and 5, and 6 and 7, all films with the HMDS layer have less sensitivity variation than those without the HMDS layer. It shows that the additional deposition of HMDS improved the stability of test films. Polymerbased HMDS also reduced the pH sensitivity.
Figure 5-6 shows the Ids-Vgs curves of the HMDS coated ZrO2 membrane. The sensitivity was decreased from 57.89 to 37.22 mV/pH, and the coating of HMDS caused more than a 35% reduction in sensitivity. Accordingly, the insensitivity character of REFET was not only dominated by the entrapped hydrophobic polymers but also contributed by HMDS. It was evident by comparing films 2 and 3, 4 and 5, and 6 and 7 that the films with the HMDS layer obtained lower pH sensitivity than those without the HMDS layer.
Among all the test films, the combination of Nafion-PR composite/HMDS has the lowest pH sensitivity, and the similar pNa response as the ZrO2 film, as shown in Figure 5-5, demonstrated the capability to serve as a REFET. Comparing with Figure 5-2 (b), the linearity became worse for the pH response but became better for pNa. The reasons for both cases are different. For pH, it is because the measured potential range was highly reduced that the measurement error becomes more significant. For pNa, a possible reason is that the property of the Nafion film was altered by the mixed composition, which restricted the transport of Na+ and the potential variation and, therefore, became more stable.
Figure 5-6 shows the differential result of the pH and pNa measurements for the ISFET/REFET arrangement; responses with more than 50 mV/pH sensitivity and less than 5 mV/pNa interference can be obtained in the range of pH 1–13. Polymers entrapped in
Nafion play the role of blocking the most paths of ion exchange to achieve a low ion sensitivity requirement for REFET. Nevertheless, the induced impedance change by the alteration of the overall membrane composition should be considered. Figure 5-7 describes the transconductance (gm) curves of the ISFET, the proposed REFET, and the REFET with epoxy coating membranes. The gm curve with epoxy-REFET shows different behaviors from the other two. On the other hand, the curves of the proposed REFET and ISFET show similar curvatures in the linear region (Vgs-Vth: 0 – 1.2 V). The gm curves represent the signal transformation characteristics of the device. An effective differential measurement requires similar gm curvatures for ISFET and REFET in pair. Otherwise, mismatch issues would occur. Because the epoxy is an ion-blocking membrane that causes a polarizable interface whose potential difference changes as a consequence of the potential difference across the whole system, it would alter the original electrical properties of an ISFET.
Consequently, the electrical mismatch of the ISFET/REFET differential pair would result in the common mode noises, which cannot be entirely eliminated. On the contrary, an ion-unblocking REFET as proposed causes a nonpolarizable interface, whose potential difference across it is virtually fixed. Meanwhile, it maintains the electrical properties of the original film and makes the ISFET/REFET differential pair match, which can easily eliminate the common mode noises with simple readout circuits. The gm of the ZrO2
membrane was slightly altered due to the addition of PR; however, the electrical characteristic in most linear regions was not altered accordingly, which provided an operating region for the ISFET/REFET pair in wide dynamic pH measurements.
In this experiment, it was found that the measurement could not be conducted without the Nafion protection because P3HT and PR alone would dissolve in the buffer solution within a very short time period. The drift performances of the Nafion-PR composite/HMDS film were 3.5 mV/h and less than 1 mV/h for the first and second 4 h tests, as shown in
Figure 5-8. It demonstrated that the incorporation of the Nafion enhanced the stability of the proposed REFET.
5.4 Summary
In this study, a method of fabricating reference ISFETs by combining Nafion and ion-insensitive polymers was proposed. With a differential arrangement, sensitivities of 52.1 mV/pH and 4.61 mV/pNa based on the ZrO2 sensing film were obtained in the range of pH 1–13; it demonstrated good performance for ISFET applications. Meanwhile, the drift performances were 3.5 mV/h and less than 1 mV/h for the first and second 4 h tests.
With the high pH sensitivity and the low interference of pNa performed by the proposed functional polymer combination, a chemically protective ion-unblocking membrane associated with the ion insensitivity polymers provides a promising way for REFET fabrications. Meanwhile, the transconductance match of the proposed REFET/ISFET pair would simplify the differential readout circuits.
References
[5.1] Bergveld, P. Development of an ion-sensitive solid-state device for neuro-physiological measurements. IEEE Trans. Biomed. Eng. BME 1970, 17, 70-71.
[5.2] Shimada, K.; Yano, M.; Shibatani, K.; Komoto, Y.; Esashi, M.; Matsuo, T.
Application of catheter-tip isfet for continuous in-vivo measurement. Med. Biol.
Eng. Comput. 1980, 18, 741-745.
[5.3] Smith, R. L.; Scott, D. C. A solid state miniature reference electrode. Proceedings of the IEEE/VSF Symposium on Biosensors, Los Angeles, CA, September 1984,
15-17, 61-62.
[5.4] Hiroaki, S.; Taishi, H.; Satoshi, S.; Isao, K. Micromachined liquid-junction Ag/AgCl reference electrode. Sens. Actuat. B 1998, 46, 146-154.
[5.5] Matsuo, T.; Esashi, M. 153rd Meet. Electrochem. Soc. Ext. Abstr. 1978, 78-1, 202-203.
[5.6] Janata, J.; Huber, R. J. Ion-sensitive field effect transistors. Ion-Sel. Electrode Rev.
1979, 1, 31-79.
[5.7] Matsuo, T.; Nakajima, H. Characteristics of reference electrodes using a polymer gate ISFET. Sens. Actuat. B 1984, 5, 293-305.
[5.8] Bergveld, P.; van den Berg, A.; van der Wal, P. D.; Skowronsk, P. M.; Sudho¨lter, E.J.R.; Reinhoudt, D.N. How electrical and chemical requirements for REFETs may coincide. Sens. Actuat. B 1989, 18, 309-327.
[5.9] Rocher, V.; Chovelon, J. M.; Jaffrezic-Renault, N.; Cros, Y.; Birot, D. An oxynitride ISFET modified for working in a differential mode for pH detection. J.
Electrochem. Soc. 1994, 141, 535-539.
[5.10] van den Berg, A.; Bergveld, P.; Reinhoudt, D. N. Sudho¨lter, E. J. R. Sensitivity control of ISFETs by chemical surface modification. Sens. Actuat. B 1985, 8, 129-148.
[5.11] van den Berg, A. Ion sensors based on ISFET’s with synthetic ionophores. PhD Thesis, Univ. of Twente, the Netherlands, 1988.
[5.12] van den Berg, A.; van der Wal, P. D.; Skowronsk, P. M.; Sudho¨lter, E. J. R.;
Reinhoudt, D. N.; Bergveld, P. Nature of anionic sites in plasticized poly(vinyl chloride) membranes. Anal. Chem. 1987, 59, 2827-2829.
[5.13] Errachid, A.; Baussells, J.; Jaffrezic-Renault, N. A simple REFET for pH detection in differential mode. Sens. Actuat. B 1999, 60, 43-48.
[5.14] Chudy, M.; Wróblewski, W.; Brzózka, Z. Towards REFET. Sens. Actuat. B 1999, 57, 47-50.
[5.15] Dawgul, M.; Pijanowska, D. G.; Krzyskow, A.; Kruk, J.; Torbicz, W. An influence of polyHEMA gate layer on properties of ChemFETs. Sensors 2003, 3, 146-159.
[5.16] Lee, Y. C.; Sohn, B. K. Development of an FET-type reference electrode for pH detection. J. Kor. Phys. Soc. 2002, 40, 601-604.
[5.17] Lai, C. S.; Lue, C. E.; Yang, C. M.; Dawgul, M.; Pijanowska, D. G. Optimization of a PVC membrane for reference field effect transistors. Sensors 2009, 9, 2076-2087.
[5.18] Kinlen, P. J.; Heider, J. E.; Hubbard, D. E. A solid-state pH sensor based on a Nafion-coated iridium oxide indicator electrode and a polymer-based silver chloride reference electrode. Sens. Actuat. B 1994, 22, 13-25.
[5.19] Luo, X. L.; Xu, J. J.; Zhao, W.; Chen, H. Y. Ascorbic acid sensor based on ion-sensitive field-effect transistor modified with MnO2 nanoparticles. Anal. Chi.
Acta. 2004, 512, 57-61.
[5.20] Anh, D. T. V.; Olthuis, W.; Bergveld, P. Hydrogen peroxide detection with improved selectivity and sensitivity using constant current potentiometry. Sens.
Actuat. B 2003, 91, 1-4.
[5.21] Soldatkin, A. P.; Volotovsky, V.; El’skaya, A. V.; Jaffrezic-Renault, N.; Martelet, C.
Improvement of urease based biosensor characteristics using additional layers of charged polymers. Anal. Chi. Acta. 2000, 403, 25-29.
[5.22] Volotovsky, V.; Young, J. N.; Kim, N. Urease-based biosensor for mercuric ions determination. Sens. Actuat. B 1997, 42, 233-237.
[5.22] Volotovsky, V.; Young, J. N.; Kim, N. Urease-based biosensor for mercuric ions determination. Sens. Actuat. B 1997, 42, 233-237.