RESEARCH AND IDEATE

Coming Up With an Idea

Through consulting Dr. Colin Dalton, a microfabrication expert, we quickly realized that this project’s scope required two years to complete. He drove home the need for accuracy, precision and reliability, which would require thoroughly vetting a sensor design. We looked at different types of biosensors to solve this problem. Mechanical cantilever-based sensors appeared attractive given their high sensitivity and versatility. However, we focused on impedance based biosensors due to their low cost, high sensitivity and in some cases high selectivity. Vitamin A has a binding protein, known as RBP. Retinol, outside of the liver in the blood, is found in a 1:1 complex with RBP (Sommer, 2001). An aptamer-based method appeared to be a solution by which we could extract retinol binding protein from the blood. Thomas Ljinse, a PhD student at the University of Calgary, provided guidance on how one can make an aptamer-based test that uses impedance to quantify a protein. By using an electrode coated with aptamers, one can quantify the amount of protein bound by the aptamers by measuring the change in impedance. With a two year timeline, we largely focused our efforts this season on developing a circuit capable of detecting changes in impedance.

Sensitivity of the Device

We needed to determine constraints on our design, and specifically, how sensitive our diagnostic device had to be. The WHO defines vitamin A deficiency (VAD) by serum retinol concentrations less than 0.70µmol/L (Whitehead et al., 2015). We found that serum retinol is the best established biochemical indicator of vitamin A status (De Pee & Dary, 2002). Our first motive was to develop a model to calculate how much impedance change RBP could cause, and thus indicate how sensitive our biosensor had to be. However, speaking with Dr. Colin Dalton, he informed us that this was a very complex phenomena to model. We were informed that many of these types of biosensor were empirically based. Without laboratory access, we relied upon literature. Using a similar method of impedance analysis of an electrochemical cell, a biosensor for RBP was able to detect quantities of RBP as low as 2.5 ag/mL (Şimşek et al., 2014). Using values of frequency and input waveforms from this paper, we were able to define constraints needed to achieve similar sensitivity.

Selectivity

Whole blood samples taken from a finger prick were an ideal candidate, but this raised questions about how we could separate the plasma from serum and quantify levels of RBP. Thus, we consulted Sultan Khetani, a PhD student with extensive experience in biosensor design. He informed us that depending on how the method used to select an aptamer, known as SELEX, was carried out, the aptamer could be specific enough to select RBP from whole blood samples. In theory, if an aptamer is selective enough, it could be used for selecting RBP from whole blood samples. Robert Mayall from FREDsense informed us that a common issue with using impedance based testing on blood, is fouling of the electrode. We were able to find an aptamer sequence in literature that was proven to be selective in human serum. Lab data and further aptamer research is needed to determine whether or not separation of plasma from serum will be necessary.

How it Works

Impedance is different from resistance in that it describes the sum of both the resistance and reactance. Systems, outside of conventional circuitry, can be described using combinations of impedances. The Randles Circuit is a common equivalent circuit used to model electrochemical cells, and is described in Figure 1.0.

Figure 1.0: Equivalent Randles Circuit of an Electrochemical Cell

In the diagram above, Zw is a warburg impedance. This models the slow diffusion process, something that is difficult to model with a simple differential equation. Zw can be modelled as:

Equation 1:

Equation 2:

The Cdl is known as the double layer capacitance, and models the effect of charges building up in the electrolyte on the electrode surface. Rct is the charge transfer resistance, accounting for voltage drop over the electrode-electrolyte interface due to a load. Lastly, Rs models the electrolyte resistance. Equation 3.0 is the impedance of a capacitor, and can thus be used to model the impedance of the double layer capacitance.

Equation 3:

The working principle of the Randle Cell Testing Device is the perturbation of the double layer capacitance in Figure 1.0 caused by the binding of RBP. This perturbation of Cdl has been known to cause significant changes in impedance. This change has been most noticeable at lower frequencies, less than 10kHz (Venkatraman et al., 2009). In addition, at low frequencies, the warburg impedance acts like a resistor (Lim, 2014). Notice that in Equation 2.0, the magnitude of the warburg impedance is directly proportional to the inverse of the square root of frequency. Therefore, as frequency approaches 0, the warburg impedance becomes infinite, analogous to an open circuit. For this reason, we simplified our model of our randle cell by assuming Zw to be an open circuit when operating at frequencies less than 10kHz.

As a result, we are left with a series circuit of Rs and Cdl, where we assume that Rs remains constant at lower frequencies.

Figure 2.0: Simplification of the Randles Circuit

In this simplified circuit, changes in the cells impedance can be largely attributed to the change in double layer capacitance, as long as Rs remain constant. This is because at low frequency (~10kHz), capacitance becomes large, seen by the inverse relation between ZC and frequency in Equation 3.0.

Selecting an Impedance Measurement Technique

We selected electrochemical impedance spectroscopy (EIS) , as it is a technique capable of measuring impedance across an electrochemical cell as a function of frequency. EIS is an AC technique, meaning it uses an alternating signal. This allows us to directly measure capacitance, by measuring impedance at a given frequency. Capacitance can be calculated by modifying Equation 3.0. We selected the AD5933 by Analog Devices due to its small size and high accuracy. This is a small integrated circuit that has the ability to run EIS.

The Electrodes

In an effort to make our design cheap and robust, we selected gold-plated interdigitated electrodes due to their ease of fabrication, high sensitivity and inexpensiveness (Mazlan, 2017). To ensure high sensitivity of the Randles Cell Circuit, the distance from the aptamer to the electrode surface needs to be as small as possible. We have talked to both Robert Mayall of FREDsense, as well as Sultan Khetani who have listed several options to immobilize aptamers to our electrode. The exact method is a question that will be addressed in future directions.

The Current Design

An electrochemical cell, featuring interdigitated electrodes. These electrodes are coated with aptamers found through the SELEX process, that have high specificity towards RBP (Lee et al., 2008). These electrodes are connected to the AD5933 embedded on the Pmod IA board through electrical leads. The Pmod IA board allows the AD9533 to interface with the Arduino Uno. All components were sourced commercially. Note that in Figure 3.0 and 4.0 we have not included the final electrode design.

Figure 3.0: Impedance Analysis Circuit Measuring a 1nF Capacitive Load Featuring the Arduino Uno, and AD9533 on the Pmod IA

Figure 4.0: The Randles Cell Testing Device measuring Impedance of an Aqueous Solution

Software

The AD5933 was calibrated and programmed using the popular Arduino IDE. The code was created following the AD5933 datasheet to ensure proper calibration and operation. Commands are written from the Arduino Uno, to the AD5933. Through the popularity of the Arduino IDE and our extensive documentation, we hope to ease the burden of datasheet synthesis and understanding of the AD5933 for future users.