**2. Potentiometric sensor array**

The detection of electric potential change based on a field-effect transistor (FET) [1] has shown excellent sensitivity such as for ion concentration and specific DNA sequences in‐ cluding single-nucleotide polymorphisms (SNPs). There are two detection principles.

One principle is the detection of electronic charge around an electrode and there is no elec‐ tron transfer to the electrode. The gate potential is determined by Poisson's equation. First, a probe layer is formed on an FET. Then, target molecules are supplied. Specific molecules are selectively taken into the probe layer on the FET channel, which detects the molecular charge in the probe layer. In the case of DNA detection, the probe is single-stranded (ss) DNA with a known sequence, immobilized on the substrate. When the target ssDNA is sup‐ plied, specific hybridization occurs if the target DNA is complementary to the probe DNA. Occurrence or nonoccurrence of specific hybridization can be detected by the difference in charge since a nucleotide has a negative charge on the phosphate group.

The other principle is the detection of chemical equilibrium potential, i.e., redox potential, accomplished by electron exchange between the electrolyte/molecule and the electrode. Fer‐ rocenyl-alkanethiol immobilized gold electrode is used to detect an enzyme reaction through a redox reaction. In this case, the gate potential is determined by the Nernst equation.

#### **2.1. CMOS Source-Drain Follower**

For the integrated sensor array, the structure must be compatible with CMOS integrated circuits. Employment of extended-gate electrodes is one solution, as shown in Figure 1(b). Molecules and/or membrane are formed on the extended-gate electrodes. Our goal is the realization of a million-sensor array on a single chip. One sensor must occupy a small area, and consume low power. Since the detection signal is in the order of 1 mV, high accuracy is essential. To meet these targets, we proposed a new integrated sensor circuit, a CMOS sourcedrain follower, where both the gate-source and gate-drain voltages of the sensor transistor are maintained at constant values [2, 3]. The source-drain follower has the merit of not influenc‐ ing the sensing system since the input impedance is infinite for both DC and AC signals.

The basic circuitry of the CMOS source-drain follower is shown in Figure 2(a). The sensor transistor N detects the extended-gate electrode voltage *VIN*. This circuit works as a voltage follower (*VOUT* = *VIN*) with high input and low output impedances. A benefit of the voltage follower is that the output voltage is independent of device parameters such as threshold voltage and environmental conditions such as temperature. This circuit also works as a source-drain follower for sensor transistor N when current *I* is kept constant.

**Figure 2.** (a) Basic CMOS source-drain follower. *VDD* and *VSS* are high and low power supply voltages, respectively. (b) 16×16 integrated sensor array with CMOS source-drain followers and peripheral circuits. A heater and thermometer are also integrated on the chip.

#### **2.2. pH Detection**

probes can be formed for parallel detection. In addition, the same kind of probe can be used to observe the time evolution of the spatial distribution of biomolecular interactions as well as to improve the detection accuracy since biomolecular interactions are a stochastic process. In this paper, several biosensor arrays are described based on the detection of electric poten‐

The detection of electric potential change based on a field-effect transistor (FET) [1] has shown excellent sensitivity such as for ion concentration and specific DNA sequences in‐

One principle is the detection of electronic charge around an electrode and there is no elec‐ tron transfer to the electrode. The gate potential is determined by Poisson's equation. First, a probe layer is formed on an FET. Then, target molecules are supplied. Specific molecules are selectively taken into the probe layer on the FET channel, which detects the molecular charge in the probe layer. In the case of DNA detection, the probe is single-stranded (ss) DNA with a known sequence, immobilized on the substrate. When the target ssDNA is sup‐ plied, specific hybridization occurs if the target DNA is complementary to the probe DNA. Occurrence or nonoccurrence of specific hybridization can be detected by the difference in

The other principle is the detection of chemical equilibrium potential, i.e., redox potential, accomplished by electron exchange between the electrolyte/molecule and the electrode. Fer‐ rocenyl-alkanethiol immobilized gold electrode is used to detect an enzyme reaction through a redox reaction. In this case, the gate potential is determined by the Nernst equation.

For the integrated sensor array, the structure must be compatible with CMOS integrated circuits. Employment of extended-gate electrodes is one solution, as shown in Figure 1(b). Molecules and/or membrane are formed on the extended-gate electrodes. Our goal is the realization of a million-sensor array on a single chip. One sensor must occupy a small area, and consume low power. Since the detection signal is in the order of 1 mV, high accuracy is essential. To meet these targets, we proposed a new integrated sensor circuit, a CMOS sourcedrain follower, where both the gate-source and gate-drain voltages of the sensor transistor are maintained at constant values [2, 3]. The source-drain follower has the merit of not influenc‐ ing the sensing system since the input impedance is infinite for both DC and AC signals.

The basic circuitry of the CMOS source-drain follower is shown in Figure 2(a). The sensor transistor N detects the extended-gate electrode voltage *VIN*. This circuit works as a voltage follower (*VOUT* = *VIN*) with high input and low output impedances. A benefit of the voltage follower is that the output voltage is independent of device parameters such as threshold voltage and environmental conditions such as temperature. This circuit also works as a

source-drain follower for sensor transistor N when current *I* is kept constant.

cluding single-nucleotide polymorphisms (SNPs). There are two detection principles.

charge since a nucleotide has a negative charge on the phosphate group.

tial, current, capacitance, and impedance.

**2. Potentiometric sensor array**

164 State of the Art in Biosensors - General Aspects

**2.1. CMOS Source-Drain Follower**

Many different biosensors have been developed based on pH sensors since various biomo‐ lecular interactions produce protons. Rothberg et al. recently demonstrated a genome se‐ quencing chip that contains 13 million pH sensors on a 17.5×17.5 mm2 die [4].

**Figure 3.** (a) Two-dimensional image of pH change. There are three faulty sensor units. (b) Cumulative probability of output voltage, and (c) median ±3σ plot as a function of pH from pH 5 to 9 (black) and pH 8 to 5 (white). The output voltage is the result after subtracting the initial values in order to eliminate the charge effect from the floating gate. (d) Cumulative probability of pH sensitivity of individual sensor cells.

The cumulative probability of pH sensitivity of 16×16 sensor cells with a 100-nm catalytic chemical vapor deposition (Cat-CVD) silicon nitride layer is plotted in Figure 3. Cat-CVD is a low-temperature (350°C) process and the deposited silicon nitride is of high quality, simi‐ lar to that obtained by low-pressure CVD [5]. The median pH sensitivity is −41 mV/pH, which is lower than the theoretical value of −57 mV/pH. The reason for this lower value may be explained by the oxygen-rich layer on the Si3N4 surface [5].

#### **2.3. DNA Detection**

Using the integrated potentiometric sensor array, preliminary experiments on DNA detec‐ tion were performed. Gold extended-gate electrodes were used to immobilize the probe DNA. Immobilization of a 5'-thiol-modified 20-mer oligonucleotide, and hybridization with the complementary oligonucleotide were detected in a 1 mM phosphate buffer (pH 7.0), as shown in Figure 4. Biomolecular interactions were observed as the time evolution of twodimensional distribution. Maximum voltage change was 80 mV for immobilization and 40 mV for hybridization. In this experiment, the uniformity of biomolecular interactions was not good. Long-term drift of the sensed voltage was observed as 30 mV/h.

**Figure 4.** Preliminary experiment on DNA detection using a 16×16 potentiometric sensor array. (a) Experimental set‐ up. (b) Output voltage change before/after immobilization and (c) before/after hybridization.

The drift was reduced to 2 mV/h when Cat-CVD silicon nitride was deposited on the ex‐ tended-gate electrode. DNA detection was also performed using the silane-coupling method for probe immobilization on Cat-CVD silicon nitride. The results show voltage changes of around 100 mV for probe immobilization, 12 mV for hybridization of complementary target DNA, and less than 1 mV for reverse-complementary target DNA.

#### **2.4. Redox Potential Detection**

The direct charge detection method using FET had a number of serious problems, as ex‐ plained in Figure 5. First, the molecular charge is screened by ions in solution. Screening length is around 3 nm in the case of ion concentration of 10 mM. This can be extended if low ion concentration is used; however, in this case, a very high impedance environment is pro‐ duced, and the electric potential becomes unstable. Second, the charge distribution is influ‐ enced by the shape of the molecule. It is generally understood that single-stranded DNA takes a Gaussian shape, and double-stranded DNA takes a rod-like shape. It is unclear whether it is a change in charge or change in structure that is detected. Especially in a flow system, the molecular shape fluctuates, which leads to unstable electric potential. Third, the electrode enters a floating state. Embedded charge causes a large threshold voltage variation.

The cumulative probability of pH sensitivity of 16×16 sensor cells with a 100-nm catalytic chemical vapor deposition (Cat-CVD) silicon nitride layer is plotted in Figure 3. Cat-CVD is a low-temperature (350°C) process and the deposited silicon nitride is of high quality, simi‐ lar to that obtained by low-pressure CVD [5]. The median pH sensitivity is −41 mV/pH, which is lower than the theoretical value of −57 mV/pH. The reason for this lower value may

Using the integrated potentiometric sensor array, preliminary experiments on DNA detec‐ tion were performed. Gold extended-gate electrodes were used to immobilize the probe DNA. Immobilization of a 5'-thiol-modified 20-mer oligonucleotide, and hybridization with the complementary oligonucleotide were detected in a 1 mM phosphate buffer (pH 7.0), as shown in Figure 4. Biomolecular interactions were observed as the time evolution of twodimensional distribution. Maximum voltage change was 80 mV for immobilization and 40 mV for hybridization. In this experiment, the uniformity of biomolecular interactions was

**Figure 4.** Preliminary experiment on DNA detection using a 16×16 potentiometric sensor array. (a) Experimental set‐

The drift was reduced to 2 mV/h when Cat-CVD silicon nitride was deposited on the ex‐ tended-gate electrode. DNA detection was also performed using the silane-coupling method for probe immobilization on Cat-CVD silicon nitride. The results show voltage changes of around 100 mV for probe immobilization, 12 mV for hybridization of complementary target

The direct charge detection method using FET had a number of serious problems, as ex‐ plained in Figure 5. First, the molecular charge is screened by ions in solution. Screening length is around 3 nm in the case of ion concentration of 10 mM. This can be extended if low ion

be explained by the oxygen-rich layer on the Si3N4 surface [5].

not good. Long-term drift of the sensed voltage was observed as 30 mV/h.

up. (b) Output voltage change before/after immobilization and (c) before/after hybridization.

DNA, and less than 1 mV for reverse-complementary target DNA.

**2.4. Redox Potential Detection**

**2.3. DNA Detection**

166 State of the Art in Biosensors - General Aspects

**Figure 5.** Problems with direct charge detection method. (a) Molecular charge is screened by ions in solution, (b) charge distribution is influenced by the shape of the molecules, and (c) electrode enters a floating state.

Instead of using the direct charge detection method, a redox potential detection method was developed using ferrocenyl-alkanethiol modified gold electrode [6, 7]. This redox potential sensor detects the ratio of oxidizer to reducer concentration, as shown in Figure 6, and is not affected by the absolute concentration and pH.

**Figure 6.** a) Schematic cross section of redox potential sensor. (b) Potential versus ratio of oxidizer (ferricyanide) to reducer (ferrocyanide) concentration.

We fabricated a chip that integrates 32×32 redox potential sensors, as shown in Figure 7 [8]. The sensor chip was dipped in 500 µM 11-ferrocenyl-1-undecanethiol (11-FUT) in ethanol for 24 h. Hexacyanoferrate mixture totaling 10 mM was used for the oxidizer and reducer. Six orders of concentration ratio of oxidizer and reducer were detected by this sensor array, as shown in Figure 6(b). The sensitivity was 57.9 mV/decade, which is very close to the theo‐ retical value of 59 mV/decade at 25°C. Stability, i.e., long-term drift and fluctuation, of elec‐ tric potential was examined using a bare electrode and an 11-FUT modified electrode, and 10 mM PBS solution (pH 7.4) and redox PBS solution in which the 10 mM hexacyanoferrate was additionally added to 10 mM PBS solution, as shown in Figure 7. By using redox PBS solution, the drift of electric potential was reduced by nearly one order. Furthermore, 11- FUT modification reduced the drift to nearly one fourth. This experiment showed that the drift can be drastically reduced by the redox potential detection method compared to the di‐ rect charge detection method. Of all 32×32 sensor cells, each potential of 92% was within ±1mV from the median. For the 8% abnormal output sensor cells, microscopic observation showed that an Au electrode had peeled off.

**Figure 7.** Redox potential sensor array (32×32) and stability of electric potential.

This redox potential sensor array successfully detected the glucose level with an accuracy of 2 mg/dL, using the following enzyme-catalyzed redox reaction:

$$\begin{array}{rcl} \text{^{\text{H}K}} & \text{^{\text{H}K}} & \text{^{\text{C}}} \\ \text{^{\text{C}}} & \text{^{\text{C}}} & \text{^{\text{C}}} \\ \text{^{\text{C}}} \text{^{\text{C}}} & \text{^{\text{C}}} & \text{^{\text{C}}} \\ \text{^{\text{C}}} & \text{^{\text{C}}} & \text{^{\text{C}}} \\ \end{array}$$

where HK is hexokinase and G6PDH is glucose-6-phosphate dehydrogenase.

Continuous sample measurement was performed using a flow measurement system with a flow speed of 1 µl/s, as shown in Figure 8. We used two types of solutions: PBS solution (pH 7.4) and glucose sample solution (glucose, 9.9 mM potassium ferricyanide, 0.1 mM potassi‐ um ferrocyanide, 0.6 mM NAD, 2 mM ATP, 10 mM MgCl2). PBS solution was used to wash out the glucose sample. As shown in Figure 9(a), the gate voltage settled in the glucose sam‐ ple very rapidly. On the other hand, in the PBS solution, a long settling time was observed. Figure 9(b) shows the relationship between given and detected glucose concentrations, indi‐ cating fairly good linearity.

Potentiometric, Amperometric, and Impedimetric CMOS Biosensor Array http://dx.doi.org/10.5772/53319 169

**Figure 8.** Setup of measurement. The chip is controlled by a microcontroller unit (MCU).

tric potential was examined using a bare electrode and an 11-FUT modified electrode, and 10 mM PBS solution (pH 7.4) and redox PBS solution in which the 10 mM hexacyanoferrate was additionally added to 10 mM PBS solution, as shown in Figure 7. By using redox PBS solution, the drift of electric potential was reduced by nearly one order. Furthermore, 11- FUT modification reduced the drift to nearly one fourth. This experiment showed that the drift can be drastically reduced by the redox potential detection method compared to the di‐ rect charge detection method. Of all 32×32 sensor cells, each potential of 92% was within ±1mV from the median. For the 8% abnormal output sensor cells, microscopic observation

This redox potential sensor array successfully detected the glucose level with an accuracy of

*Gluconolactone* - 6 - *phosphate* + *NADH*

(1)

showed that an Au electrode had peeled off.

168 State of the Art in Biosensors - General Aspects

**Figure 7.** Redox potential sensor array (32×32) and stability of electric potential.

2 mg/dL, using the following enzyme-catalyzed redox reaction:

Diaphorate

*Glucose* - 6 - *Phosphate* + *ADP*

G6PDH

where HK is hexokinase and G6PDH is glucose-6-phosphate dehydrogenase.

2 Fe(CN)

6

Continuous sample measurement was performed using a flow measurement system with a flow speed of 1 µl/s, as shown in Figure 8. We used two types of solutions: PBS solution (pH 7.4) and glucose sample solution (glucose, 9.9 mM potassium ferricyanide, 0.1 mM potassi‐ um ferrocyanide, 0.6 mM NAD, 2 mM ATP, 10 mM MgCl2). PBS solution was used to wash out the glucose sample. As shown in Figure 9(a), the gate voltage settled in the glucose sam‐ ple very rapidly. On the other hand, in the PBS solution, a long settling time was observed. Figure 9(b) shows the relationship between given and detected glucose concentrations, indi‐

4- + *NAD*

*Glucose* + *ATP* →

6

cating fairly good linearity.

2 Fe(CN)

HK

*Glucose* - 6 - *Phosphate* + *NAD* →

3- + *NADH* →

**Figure 9.** a) Flow measurement of glucose. Blue areas indicate the flow of PBS solution, and yellow areas indicate the flow of glucose sample solution. (b) Detected glucose vs. given glucose.

This sensor array could be applied to genome sequencing by incorporating a primer exten‐ sion reaction, which produces pyrophosphate (PPi).

$$\begin{array}{rcl} \text{PPi} + \text{H}\_{2}\text{O} \rightarrow 2\text{Pi} \\\\ \text{P}i + \text{GAP} + \text{NAD}^{\cdot} \rightarrow \text{1,3}\text{BPG} + \text{NADH} \\\\ \text{2[Fe(CN)}\_{6}\text{]}^{3\text{-}} + \text{NADH} \rightarrow \text{2[Fe(CN)}\_{6}\text{]}^{4\text{-}} + \text{NAD}^{\cdot} \end{array} \tag{2}$$

where GAP is glyceraldehyde 3-phosphate and BPG is bisphosphoglycerate.

## **3. Amperometric sensor array**

Amperometric imaging offers great potential for multipoint rapid detection and the analysis of diffusion processes of target molecules. The microelectrode is one of the most versatile and powerful tools in amperometry. Although the current passing through a microelectrode is very small, it has the advantages of high mass transport density, small double-layer capaci‐ tance, and small ohmic drop. Moreover, a microelectrode has a steady-state current re‐ sponse in unstirred solutions. Such steady-state currents are easy to analyze and interpret. However, as it takes a few or tens of seconds before reaching a steady state, rapid multi‐ point measurement cannot be achieved with a simple switching scheme. Furthermore, when the inter-electrode distance is not sufficiently large, the diffusion layers begin to overlap and eventually merge to form a single planar diffusion layer. This overlapping of diffusion lay‐ ers is commonly referred to as "cross talk" or the "shielding" effect. When cross talk occurs, the microelectrode array loses its unique features and becomes similar to a large-area "macro" electrode, which makes local and quantitative analysis extremely difficult. We proposed a switching circuit that measures multiple microelectrode currents at high speed, and a micro‐ electrode structure to suppress diffusion layer expansion over the microelectrode array [9].

#### **3.1. Switching Circuit and Microelectrode Array Structure**

Figure 10(a) shows the proposed amperometric electrochemical sensor circuit. Multiple elec‐ trodes placed in an array are connected with one amperometric sensor circuit through the switches. Each electrode is connected to two switches. The electrode being measured is con‐ nected to the readout circuit via switch SWA, and on stand-by, the potential is fixed via switch SWB to maintain the steady-state current. When the reading electrode is switched, either switch of the two is kept closed. Therefore, the switching is carried out while the steady-state current is maintained. In this way, it is not necessary to wait for a steady-state current, thus realizing ultra-fast readout from each electrode.

**Figure 10.** a) Amperometric electrochemical sensor circuit. Each electrode is connected to two switches. (b) Conven‐ tional and proposed electrode geometry.

Our working electrode (WE) structure shown in Figure 10(b2) is surrounded by a grid auxil‐ iary electrode (AE), and the redox reaction opposite the working electrode (WE) occurs in the AE. Therefore, the diffusion layer is confined around the WE, and the overlap is de‐ creased. The steady-state current is amplified by redox cycling, and the time to reach the steady-state is reduced.

#### **3.2. Fabricated Amperometric Sensor Array**

**3. Amperometric sensor array**

170 State of the Art in Biosensors - General Aspects

**3.1. Switching Circuit and Microelectrode Array Structure**

current, thus realizing ultra-fast readout from each electrode.

tional and proposed electrode geometry.

Amperometric imaging offers great potential for multipoint rapid detection and the analysis of diffusion processes of target molecules. The microelectrode is one of the most versatile and powerful tools in amperometry. Although the current passing through a microelectrode is very small, it has the advantages of high mass transport density, small double-layer capaci‐ tance, and small ohmic drop. Moreover, a microelectrode has a steady-state current re‐ sponse in unstirred solutions. Such steady-state currents are easy to analyze and interpret. However, as it takes a few or tens of seconds before reaching a steady state, rapid multi‐ point measurement cannot be achieved with a simple switching scheme. Furthermore, when the inter-electrode distance is not sufficiently large, the diffusion layers begin to overlap and eventually merge to form a single planar diffusion layer. This overlapping of diffusion lay‐ ers is commonly referred to as "cross talk" or the "shielding" effect. When cross talk occurs, the microelectrode array loses its unique features and becomes similar to a large-area "macro" electrode, which makes local and quantitative analysis extremely difficult. We proposed a switching circuit that measures multiple microelectrode currents at high speed, and a micro‐ electrode structure to suppress diffusion layer expansion over the microelectrode array [9].

Figure 10(a) shows the proposed amperometric electrochemical sensor circuit. Multiple elec‐ trodes placed in an array are connected with one amperometric sensor circuit through the switches. Each electrode is connected to two switches. The electrode being measured is con‐ nected to the readout circuit via switch SWA, and on stand-by, the potential is fixed via switch SWB to maintain the steady-state current. When the reading electrode is switched, either switch of the two is kept closed. Therefore, the switching is carried out while the steady-state current is maintained. In this way, it is not necessary to wait for a steady-state

**Figure 10.** a) Amperometric electrochemical sensor circuit. Each electrode is connected to two switches. (b) Conven‐

Our working electrode (WE) structure shown in Figure 10(b2) is surrounded by a grid auxil‐ iary electrode (AE), and the redox reaction opposite the working electrode (WE) occurs in A 16×16 amperometric sensor array was fabricated as shown in Figure 11. Ag/AgCl and Pt wire were used as reference and counter electrodes, respectively. The solution was com‐ posed of 100 mM sodium sulfate and 1 mM potassium ferrocyanide. The WE and the AE potential were fixed at 0.65 V and 0 V (vs Ag/AgCl), respectively. Figure 11 shows the cur‐ rent responses of 1 mM [Fe(CN) 6] 4− observed at the single microelectrode, conventional mi‐ croelectrode array, and proposed microelectrode array. This amperometric sensor array could be applied to genome sequencing by using allele-specific primers and electrochemical reaction [10].

**Figure 11.** Amperometric sensor array (16×16) and current responses of fabricated microelectrodes. The size of the working electrode is 25×25 μm2.

#### **4. Impedimetric sensor array**

#### **4.1. Capacitance Sensor Array**

We applied nonfaradaic impedimetric measurement by implementing charge-based capaci‐ tance measurement (CBCM) to realize a label-free, fully integrated capacitance biosensor. The proposed sensor exploits the capacitance changes of electrical double-layer properties as a result of biorecognition events at the sensing electrode/solution interface. Figure 12(a) shows a schematic of the proposed circuit [11]. To overcome the trade-offs between sensor area and performance, we employed a fully differential measurement circuit that would compensate for parasitic capacitances and reduce the effect of electronic noise, leading to improvement of the detection limit. A photomicrograph of the fabricated chip is shown in Figure 12(b).

**Figure 12.** a) Schematic of a fully differential capacitance sensor. CX1 is the capacitance due to molecules to be detect‐ ed. (b) Photomicrograph of a sensor chip with 4×4 µm2 planar electrodes.

When probe oligonucleotides were immobilized on the electrode surface, a self-assembled monolayer serving as an insulator was formed in conjunction with the electrical double lay‐ er. The resulting interfacial capacitance is a total of these series capacitances. When comple‐ mentary oligonucleotides were introduced to the probes, hybridization occurred and this interface property, i.e., double-layer thickness due to ion displacement, was altered, causing the corresponding capacitance to undergo further change. DNA detection is demonstrated by comparing the results of the capacitance measurements using bare, immobilized, and hy‐ bridized electrodes [11]. As observed in Figure 13, the immobilization gave rise to a maxi‐ mum of 50% capacitance reduction when 20-mer thiolated oligonucleotides were selfassembled at the gold electrode surface. A further 20% reduction in capacitance is also observed after hybridization, implying that the double layer has changed due to the hybridi‐ zation event.

**Figure 13.** Measured results of capacitance against frequency for bare electrode, after DNA immobilization, and after DNA hybridization.

#### **4.2. Impedance Sensor Array**

improvement of the detection limit. A photomicrograph of the fabricated chip is shown in

**Figure 12.** a) Schematic of a fully differential capacitance sensor. CX1 is the capacitance due to molecules to be detect‐

When probe oligonucleotides were immobilized on the electrode surface, a self-assembled monolayer serving as an insulator was formed in conjunction with the electrical double lay‐ er. The resulting interfacial capacitance is a total of these series capacitances. When comple‐ mentary oligonucleotides were introduced to the probes, hybridization occurred and this interface property, i.e., double-layer thickness due to ion displacement, was altered, causing the corresponding capacitance to undergo further change. DNA detection is demonstrated by comparing the results of the capacitance measurements using bare, immobilized, and hy‐ bridized electrodes [11]. As observed in Figure 13, the immobilization gave rise to a maxi‐ mum of 50% capacitance reduction when 20-mer thiolated oligonucleotides were selfassembled at the gold electrode surface. A further 20% reduction in capacitance is also observed after hybridization, implying that the double layer has changed due to the hybridi‐

**Figure 13.** Measured results of capacitance against frequency for bare electrode, after DNA immobilization, and after

ed. (b) Photomicrograph of a sensor chip with 4×4 µm2 planar electrodes.

Figure 12(b).

172 State of the Art in Biosensors - General Aspects

zation event.

DNA hybridization.

Electrochemical impedance was measured between two disc electrodes of 20 mm in diame‐ ter and with 1 mm of separation, as shown in Figure 14. The specific hybridization is charac‐ terized by the change in Curie– von Schweidler exponent *α* of the constant phase element, which implies the structural change of molecules [12]. This shows that the overall impe‐ dance characteristic is more important than the capacitance for detecting randomly distrib‐ uted molecules.

**Figure 14.** Electrochemical impedance spectroscopy using 20-mm-diameter Au disk electrodes, and Nyquist plot (Cole-Cole plot) of (a) bare electrode, (b) probe/mercaptohexanol immobilization, (c) electrode after noncomplemen‐ tary binding, and (d) target hybridization. Impedance of constant phase element is proportional to ( jω) −α, where *j* is the imaginary unit, ω is the angular frequency, and α is the Curie– von Schweidler exponent. The ratio of the imaginary part to the real part becomes a constant −tan(πα/2).

We have designed an on-chip impedimetric sensor unit, which measures the amplitude of impedance at frequencies up to 10 MHz. The sensor unit and peripheral circuitry are shown in Figure 15. To eliminate the effect of turn-on resistance (~20 kΩ) of switch SWA and bit line capacitance (~400 fF), a current amplifier is included in each sensor unit as shown in Figure 15.

**Figure 15.** Circuitry of impedimetric sensor unit. AC current is amplified inside a sensor unit. *VBB* is DC bias voltage. A similar current amplifier was used in [13].

## **5. Multimodal Sensor Array**

In large-scale integration (LSI) circuit fabrication, the initial cost for making a set of photo‐ masks is quite high. On the other hand, the chip cost is extremely low if a large number of chips are produced. Table 1 shows the typical cost in several technologies. From this table, more than 10,000 chips are necessary to balance the initial cost. This means that standardiza‐ tion and general-purpose sensor chips are important. Our strategy is to realize a multimodal sensor array for synthetic analysis and standardization. The chip consists of amperometric, potentiometric, and impedimetric smart cells containing an amplifier in the sensor cell, ach‐ ieving noise reduction and not influencing the measurement system. The chip can be cus‐ tomized by patterning the insulator layer to cover the unused sensor cells, as shown in Figure 16.


**Table 1.** Typical cost of LSI fabrication

**Figure 16.** General-purpose sensor chip integrated with potentiometric, amperometric, and impedimetric sensor units. The chip can be customized by the post-CMOS process.

A multimodal sensor unit with potentiometric, amperometric, and impedimetric sensors is shown in Figure 17. The chip was fabricated by using a 1.2-µm 2P2M (2-polysilicon and 2 metal layers) CMOS process. Furthermore, a 16×16 multimodal sensor array with 0.24 mm pitch was designed using a 0.6-µm 2P3M mixed-signal general CMOS process.

**Figure 17.** A 4×4 1-mm-pitch multimodal sensor array integrated with potentiometric, amperometric, and impedi‐ metric sensor units.

## **6. Conclusion**

**5. Multimodal Sensor Array**

174 State of the Art in Biosensors - General Aspects

Figure 16.

**Table 1.** Typical cost of LSI fabrication

units. The chip can be customized by the post-CMOS process.

In large-scale integration (LSI) circuit fabrication, the initial cost for making a set of photo‐ masks is quite high. On the other hand, the chip cost is extremely low if a large number of chips are produced. Table 1 shows the typical cost in several technologies. From this table, more than 10,000 chips are necessary to balance the initial cost. This means that standardiza‐ tion and general-purpose sensor chips are important. Our strategy is to realize a multimodal sensor array for synthetic analysis and standardization. The chip consists of amperometric, potentiometric, and impedimetric smart cells containing an amplifier in the sensor cell, ach‐ ieving noise reduction and not influencing the measurement system. The chip can be cus‐ tomized by patterning the insulator layer to cover the unused sensor cells, as shown in

**Technology layers Cost of a set of photomasks Cost of 1-cm2 chip excluding**

**Figure 16.** General-purpose sensor chip integrated with potentiometric, amperometric, and impedimetric sensor

A multimodal sensor unit with potentiometric, amperometric, and impedimetric sensors is shown in Figure 17. The chip was fabricated by using a 1.2-µm 2P2M (2-polysilicon and 2

0.6 μm 2P3M \$ 30K \$ 5 0.25 μm 1P4M \$ 100K \$ 7 0.18 μm 1P6M \$ 240K \$ 9 0.13 μm 1P8M \$ 600K \$ 14

**photomasks**

Potentiometric, amperometric, and impedimetric sensor arrays using standard CMOS tech‐ nology are described. Biomolecular interactions were observed as the time evolution of twodimensional distribution. The multimodal sensor array with potentiometric, amperometric, and impedimetric sensor units will enable synthetic sensing and standardization of Bio‐ CMOS LSIs.
