The pH Detect Technique - 離子感測場效電晶體對於溫度及閘極電壓應力調變之可靠度研究

Chapter 1 Introduction

1.4 The pH Detect Technique

In tradition, the measuring methods for pH values fall roughly into four categories:

indictor reagents, pH test strips, metal electrode methods and glass electrode methods. The glass electrode is most widely used for pH measurement due to ideal Nernstian response independent of redox interferences, short balancing time of electrical potential, high reproducibility and long lifetime. However, glass electrode has several drawbacks for many industrial applications. Firstly, they are unstable in alkaline or HF solutions or at temperatures higher 100℃. Also, they exhibit a sluggish response and are difficult to miniaturize. Moreover, they cannot be used in food or in in vivo applications due to their brittle nature. There is an increasing need to for alternative pH electrodes.

The new techniques for pH detection which include: (1) Optical-fiber-based pH sensors.

(2) Mass-sensitive pH sensor. (3) Metal oxide pH sensors. (4) Conducting polymer pH sensors. (5) Nano-constructed cantilever-based pH sensors. (6) ISFET-based pH sensors. (7) pH-image sensor.

The problems in practical applications about reliability of ISFETs are continuing to be investigated in this study.

Chapter 2 Theories for the Investigation of ISFET’s Temperature and Drift Characteristics

2.1 Basic Principles of ISFET

An anisotropic ion accumulation exists at the contact interface between an electrochemically active surface and a liquid electrolyte (Figure 2-1). Due to their different size and charge, the ions from a well-confined electric double layer close to the surface and outer charges exists between the Helmholtz planes and the neutral bulk of the solution.

Since the first report of the ISFET by Bergveld, research on new material and fabrication process to improve the sensitivity and stability has been continuously proposed [4-6]. At the same time, the mechanism of the pH response of pH ISFET has also been studied extensively [5-10].The followings are the theoretical foundations which are mostly adopted to characterize the ISFET.

2.1.1 From MOSFET to ISFET

The operation of an ISFET can best be described by comparing it with its purely electronic analogue. Figure 2-2 illustrates the similarities and differences between these two devices: The metal gate of the MOSFET of Figure 2-2(a) is replaced by the metal of a reference bare gate insulator ISFET in Figure 2-2(b). Mounting of the chips is of course different: a MOSFET can be completely encapsulated, whereas for an ISFET source and drain leads as well as chip edges have to be encapsulated carefully, meanwhile leaving the

gate area open for the contact with liquid.

For both devices the following equation is valid for the non-saturated region:

(

_GS _T

)

_DS _DS

with COX is the oxide capacitance per unit area, μ the electron mobility in the channel , W and L the width and the length of the channel.

In addition, the fabrication process for MOSFET devices are so well under control that VT is also a constant, which manifests itself only as a certain threshold voltage, hence its name, it was initially debated whether the observed ion sensitivity should be described as an additional input variable in terms of a modification of Vgs or a modification of VT. Therefore, the second important MOSFET equation is that of the threshold voltage:

where the first term reflects the difference in work function between the gate meta (ΦM) and silicon (Φsi), the second term is due to accumulated charge in the oxide (QOX), at the oxide-silicon interface (QSS) and the depletion charge in the silicon(QB), whereas the last term determines the onset of inversion depending on the doping level of the silicon. All terms are purely physical in nature.

In case of the ISFET, the same fabrication process is used, resulting in the same constant physical part of the threshold voltage (Eq. (2-2)). However, in addition to this, two more contributions manifest themselves: the constant potential of the reference electrode, Eref , and the interfacial potential Eref + χ^sol at the solution/oxide interface of which Ψ is the chemical input parameter, shown to be a function of the solution pH and χ^sol is the surface dipole potential of the solvent and thus having a constant value. The term Ψ0 representing the surface potential at the oxide-electrolyte interface is the key element that makes ISFET pH-sensitive.

The resulting equation for the threshold voltage of an ISFET is thus given by:

2.1.2 The Oxide-Electrolyte Interface of pH Response

The surface of any metal oxide always contains hydroxyl groups, in the case of silicon dioxide SiOH groups. These group may donate or accept a proton from solution, leaving a negatively charged or a positively charged surface group respectively. In the case of silicon dioxide SiOH groups. It is indicated that equilibrium reactions can occur between protons in the solution and the hydroxyl at the SiO2 solution interface. The oxide surface charge can be described by the site-binding model, as schematically represented by Figure 2-3 which describes the equilibrium between the so-called amphoteric SiOH surface sited and the H⁺ -ions in the solution. These reactions can be expressed by

The generally expression for the pH sensitivity of an ISFET, this response is given by:

q α

the parameter α is a dimensionless sensitivity parameter what varies between 0 and 1, depending on the intrinsic buffer capacity βint, of the oxide surface and the differential double-layer capacitance C^diff. If α=1, the theoretical maximum sensitivity of -59.2mV/pH at room temperature can be obtained.

2.2 Introduction the Temperature Phenomenon

Because an ISFET is a chemical sensor based on MOSFET, ISFET’s temperature characteristics are similar to a conventional MOSFET. According to the equation of the linear region in MOSFET, we concluded that only the mobility and the threshold voltage are temperature dependent. These two factors have a negative T.C. It implies which a zero T.C.

point exits in the ID-VG characteristics. We concluded a series of experiments on the temperature characteristics in different solutions, the operation currents of the zero T.C. are changed in different solutions. This results indicate that the operation conditions for zero T.C.

will change along with the change in solutions.

On the operating conditions for a zero T.C., a differential configuration was used to investigate the membrane/electrolyte interface temperature coefficient. The differential configuration used dual FET’s structures, one FET was a MOSFET, the other was an ISFET.

A typical example of the linear region of the dual FET’s structures in different temperatures is shown in Figure 2-4. This figure indicates that the operation currents for the MOSFET and the ISFET at different currents will yield a different T.C. Although the two FET’s have the same tendency, the T.C. of the membrane/electrolyte interface can be estimated from the following equation:

T.C.(ISFET) — T.C.(MOSFET)= T.C.⁽membrane/ electrolyte interface) (2-8)

Operated in the nonsaturation region in different temperatures of IDS-VGS curves for the MOSFET were obtained. Part of the ID-VG curves are shown in Figure 2-5. A deviation in the IDS-VGS characteristics occurred as the temperature changed and it is easily to see the isothermal point clearly. In our previous study, an ISFET with different sensing film was developed based on the MOSFET theory. Operated in the nonsaturation region in pH = 2, 4, 6, 7, 8, 10 and 12 in temperatures from 5 to 65℃ with a step of 10℃, a family of IDS-VGS

curves for the gate ISFET were obtained. Part of the I -V curves are shown in Figure 2-6. A

deviation in the IDS-VGS characteristics occurred as the temperature changed and it is easily to see the isothermal point clearly too. The ID-VG curve at a specific pH declined with an increase in temperature. Accordingly, and isothermal point, near zero temperature coefficient, it is indicates that a well-chosen operating point can eliminate the temperature influence in nonsaturation region because the threshold voltage is approach constant. In this study, the temperature effects on several sensing film ISFET characteristics operated in the nonsaturation and saturation regions were investigated. The threshold voltage( VT ) and drain-source current( IDS ) versus temperature characteristics will be discussed [11-12].

2.3 Introduction the Drift Phenomenon

Threshold voltage instability, commonly known as drift, has seriously limited the commercial viability of ISFET-based sensors by imposing special requirements for burn-in, packaging or compensation. Drift is typically characterized by a relatively slow, monotonic, temporal change in the threshold voltage of the ISFET, which is not caused by variations in the electrolyte composition. General explanations proposed for drift phenomenon include electric field enhanced ion migration within the gate insulator as well as electrochemical non-equilibrium conditions at the insulator -solution interface, injection of electrons from the electrolyte at strong anodic polarizations, creating negative space charge inside the insulator films, and slow surface effect. The drift is thought to be caused by the slow conversion of the surface to a hydrated during contact with the solution. The following models, which are classified according to the location where the mechanism of pH-sensitivity is presumed to occur, will help us to have a further understanding of the transport of mobile ions.

2.3.1 Physical Model for Drift

The model presented in this work quantitatively explains drift in terms of hydration [13-16]. In particular, the time dependence of drift is derived by considering the correlation between the rate of hydration and the hopping and /or trap-limited transport of water-related species. The gate voltage drift in pH ISFET’s is a relatively slow process which occurs over a period of several hours, it is reasonable to hypothesize that this phenomenon is associated with transport in the insulator. In particular, the motion within this gate insulator is expected to be characterized by relatively long transit time.

It is well known that the surface is slowly converted to a hydrated sensing layer as result of exposure to an aqueous electrolyte. The chemical modification of the insulator surface implies that the dielectric constant of the hydration layer will differ from that of the insulator bulk. Therefore, the overall insulator capacitance, which is determined by the series combination of the capacitance of the hydration layer, it will exhibit a relatively slow, temporal change as hydration proceeds. The rate of hydration which has recently been accurately modeled by a hopping transport mechanism, known as dispersive transport. In amorphous solid, dispersive transport arising from a hopping motion via localized states, result in a characteristic power-law time decay of diffusivity given by:

where D00 is a temperature-dependent diffusion coefficient which obeys an Arrhenius relationship, ω0 is the hopping attempt frequency, and β is the dispersion parameter satisfying 0<β<1. Physically, the time-dependent transport properties result from the dispersion in the separation distances between nearest-neighbor localized sites and/or the dispersion in trap energy levels. Dispersive transport leads to a decay in the density of sites/traps occupied by the species undergoing transport. This decay is described by the stretched-exponential time

dependence given by

where ΔNS/T(t) is the area density (units of cm^-2) of sites/traps occupied, τ is the time constant associated with structural relaxation, and β is the dispersion parameter characterizing dispersive transport.

2.3.2 Drift Expression

One means of operating an ISFET is in the feedback mode, where a constant drain current is maintained by applying a compensating feedback voltage to the solution side of the gate voltage ( e.g., a reference electrode). Therefore, the temporal change is the overall insulator capacitance resulting from hydration leads to a drift in compensating feedback voltage. In other words, when drift phenomenon occurs at the surface of an actively-biased ISFET, the gate voltage will simultaneously exhibit a change to keep a constant drain current.

The change in the gate voltage can be written as )

Since the voltage drop inside of the semiconductor is kept constant in the feedback mode, ΔVG(t) becomes

where VFB and Vins represent the flatband voltage and the voltage drop across the insulator, respectively. The flatband voltage is given by

OX The voltage drop across the insulator, Vins, is given by

where Qinv is the inversion charge. If the temperature, pH, and the ionic strength of the

solution are held constant, the variations in Eref, χ^sol, Ψ0, and ΦSi can be neglected, so the drift

The gate oxide of the fabricated ISFET was composed of two layers in this study, a lower layer of thermally-grown SiO2 of thickness, xL, and an upper layer of PECVD SiO2 of thickness, xU.. CI(0) is the effective insulator capacitance given by the series combination of the thermally-grown SiO2 capacitance, εL/xL, and the PECVD SiO2 capacitance, εU/xU. Ci(t) is analogous to CI(0), but an additional hydrated layer of capacitance, εHL/xHL, at the oxide-electrolyte interface must be took into consideration, and the PECVD SiO2 capacitance is now given by εU/[xU－xHL]. The series combinations of the capacitances are illustrated in Figure 2-7. The simplified expression for drift is, therefore, given by [17]

)

As is evident from From this equation (2-16), drift is directly proportional to the thickness of the modified surface layer. Therefore, the time dependence of drift is identical to that associated with the growth of this layer. By considering the time dependence of the diffusion coefficient associated with dispersive transport, an expression for xHL(t) is given by

[ ]

where AD represents the cross-sectional area, and Nhydr is the average density of the hydrating species per unit volume of hydration layer. The overall expression for the gate voltage drift is given by

[ ]

We can expect that if the time of gate oxide immersing in the test-solution is long enough (determined by the constant τ ), the gate voltage drift will approach a constant value which is greatly dependent on the hydration depth, xHL(∞).

Chapter 3 Procedures of the Experiment

To investigate the properties of membrane as the pH-sensing layers, the ISFET were fabricated. All processes were done in NDL (National Nano Device Laboratory)

and Nano Facility center. The schematic diagrams of ISFET is presented and corresponding graph is shown in Figure 3-1.

3.1 ISFET Fabrication Process Flow

(a) RCA clean

Wet-oxidation, 6000Å, temperature = 1050°C for 65 minutes (b) Defining of Source/Drain (mask 1)

BOE wet-etching of SiO2

Source/Drain annealing, 950°C, 60 minutes (d) PECVD SiO2 for passiveness, 1μm

(e) To define contact hole and gate region (mask 2) BOE wet-etching of SiO2

(f) Dry oxide thickness=100Å, temperature=850°C for 60 minutes (g) Sensing layer 1, 300Å

Defining of sensing region (mask 3) HF wet-etching of SiO2

(h) Sensing layer 2, 300Å (mask 4)

(i) Al evaporation, 5000Å (mask 5)

3.2 Experiment details

3.2.1 Gate Region Formation

RCA clean is usually performed at wafer starting to reduce the effect of diffusion ions, particles and native oxide. RCA clean will ensure the integrity of device electricity. In order to create a Source/Drain region, the next step 600Å thickness wet oxide is deposited as blocking layer for Source/Drain implant. The density and the energy of Source/Drain implant is 5E15 (1/cm²) and 25Kev with phosphorus dopant, respectively. In our experiment, p-type wafer is used. After Source/Drain implanting following a 950°C 30minutes N⁺ anneal performed to activate the dopants.

We do not need to deposit PE-oxide with thickness 1um in standard MOSFET, but it is necessary to do which protect the structure of a pH-ISFET, when the ISFET’s will operating in a long period, during this period, we need to avoid ion’s diffusion in the structure and affect the electrical characteristics [18]. In order to avoid this influence, a thick PE-oxide deposition can eliminate this effect. After PE-oxide depositing, 100Å thickness dry oxide was grown as gate oxide.

3.2.2 Sensing Layer Deposition

Methods of deposited sensing membrane as gate material are different which is the most important part in our experiment. The drift, hysteresis and sensitivity will improve by different layers [19]. For comparing these sensing layers, several deposition techniques were performed. Low- pressure nitride (LP- nitride) and PE-oxide are deposited as sensing

layers. We adopted LPCVD to obtain low stress nitride and high sensitivity, so it is good sensing layer. There are so many researches on it [20]. However, PE-oxide drift and sensitivity are unstable in different electrolytes. In CMOS process, tungsten and tantalum are popularly used. By using different barriers, drift lowing for a long period of tome and compatible with CMOS can be accomplished.

3.3 Measurement System

3.3.1 Preparation Before Measuring

To define the characteristics of the device, we use HP4156A to measure the I-V curve of ISFET. The measurement system is showed in Figure 3-2. Otherwise, light will produce serious influence on the ISFET, so that we measurement in the dark box to prevent light influence.

After device being made, we glue a container on the wafer. Entire sensing layer region must be included in the opening under the container. The material of the container is made by silica gel and the bottom has to be small enough to avoid touching other devices. However, the opening on the top has to be big enough for insert reference electrode.

The pH-solution that we use is made from Riedel-deHzen and pH-value is 1, 3, 5, 7 buffer solutions. The electric potential of the pH-solution is always floating. The disturbance from the environment would induce the electric potential variance of the solution. By eliminating this variance, a reference electrode is needed to put into the pH-solution.

3.3.2 Measurement Set-Up

In the beginning of the measurement, the reference electrode is suspended on the air over the container. The 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-4156A semiconductor parameter analyzer system, substrate is grounded 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 and at the first, we check the I^D-V^D curve to make sure the ISFET device is work as a MOSFET. On the other hand, we also should decide the drain bias from ID-VD to ensure the both IFET are operating in linear region while I^D-V^G measurement.

The ISFET held at VG= 1V, 2V, 3V, 4V, and 5V. The typical set of ID-VD curves for the ZrO2, Ta2O5, Thermal Oxide, and PE Oxide gate ISFET are shown in Figure 3-3, Figure 3-4, Figure 3-5, Figure 3-6.

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.

3.3.3 Temperature Measurement Set-Up

For characterizing the temperature influence of ISFETs, we measured I-V curves for etch film with changing the pH-solution in order of pH 1, 3, 5, 7 buffer solutions and controlling the different ambient temperature about 25°C, 35°C, 45°C, 55°C, 65°C, 75°C, and 85°C. For each temperature value, we wait the 15 minutes then measured the I^D-V^G curves which the pH-ISFET had been covered by the pH-solution. The measurement system is showed in Figure 3-7. Figure 3-8 illustrates the detection principle of pH. Firstly, we obtain

the pH1 transconductance, it is purpose to get maximum gain. The second step we decide the pH1’s VG then decide the IDS. At last, we can obtain the pH3, pH5, pH7’s VG. When we change the different pH buffer solution, we must use the pH7 buffer solution first, then pH5, pH3, and pH1, it is purpose to get better performance which let the pH buffer solution concentration from low to high.

Chapter 4 Results and Discussions

4.1 Temperature Sensitivities of Various Membranes

4.1.1 ZrO2 membrane gate ISFET

Figure 4-1~4-14 are the ID-VG curves and sensitivities of ZrO2 gate ISFET in pH=1, 3, 5, 7 buffer solutions at 25℃, 35℃, 45℃, 55℃, 65℃, 75℃, and 85℃. The measuring data are sorted in table 4-1. Figure 4-15 is temperature sensitivity correlation coefficient. The measuring data are sorted in table 4-3. Figure 4-16 is normalize the temperature sensitivities curve. The measuring data are sorted in table 4-2. According to the data about table 4-3, we can find that ZrO2 gat ISFET temperature sensitivity is increase progressively.

在文檔中離子感測場效電晶體對於溫度及閘極電壓應力調變之可靠度研究 (頁 17-0)