Introduction
A new design of optical biosensor platform is desired in order to increase detection limit and repeatability of biosensor chips used in the Sailor Lab. The need for the project is to maximize the signal to noise ratio (SNR) in optical measurement. The new design works in a similar mechanism as the old platform. However, it has distinguishable features such as: spectrometer mounted directly onto the flow cell, a straight flow channel, and an aluminum base with active temperature control. This new design fulfills the need by eliminating signal distortions caused by bubble, stabilizing flow cell temperature against ambient temperature fluctuation, minimizing effect of ambient light and maximizing flow chamber mixing efficiency.
In the Sailor Lab at UCSD's Biochemistry Department, porous silicon chips are used in a biosensor platform to detect, identify and quantify minute concentrations of proteins[1,2]. The detection mechanism relies in the local refractive index change as target proteins diffuse into the porous matrices, resulting in a shift in effective optical thickness (EOT), or 2nL (where n is refractive index and L is thickness)[3] of reflective spectrum. The current flow cell has an baseline error magnitude of 2nm in the EOT reading due to undetermined sources of noise, limiting the detection ability of the system to detect concentrations below 3ng/mL. The system must also be refined to produce accurate results in small timescales to enable the study the kinetics of reactions. The need in this project is to maximize the signal to noise ratio (SNR) in optical measurement, by first identifying significant noise sources. A new flow cell will then be designed to address and minimize these noise sources.
In the Sailor Lab at UCSD's Biochemistry Department, porous silicon chips are used in a biosensor platform to detect, identify and quantify minute concentrations of proteins[1,2]. The detection mechanism relies in the local refractive index change as target proteins diffuse into the porous matrices, resulting in a shift in effective optical thickness (EOT), or 2nL (where n is refractive index and L is thickness)[3] of reflective spectrum. The current flow cell has an baseline error magnitude of 2nm in the EOT reading due to undetermined sources of noise, limiting the detection ability of the system to detect concentrations below 3ng/mL. The system must also be refined to produce accurate results in small timescales to enable the study the kinetics of reactions. The need in this project is to maximize the signal to noise ratio (SNR) in optical measurement, by first identifying significant noise sources. A new flow cell will then be designed to address and minimize these noise sources.
Preliminary Testing & Identification of Noise Sources
In order to determine the noise sources to be targeted in the redesign, preliminary testing was necessary. Based on the main engineering principles involved in this system, optics and fluidics, the main theoretical sources of noise were light, temperature, and flow conditions. Using the existing design, each possible noise source was isolated for testing to define the magnitude of noise generated.
To test the effect of changes in ambient light, the flow cell was run under normal conditions to establish a baseline EOT reading and then covered from the ambient light in the room until changes in the data were stabilized. During that time period, the effective optical thickness, EOT, reading dropped a significant 16nm, indicating that fluctuating ambient light between the spectrometer and the biosensor platform (as shown in Figure 1) is a source of noise that should be considered in the redesign.
To test the effect of changes in ambient light, the flow cell was run under normal conditions to establish a baseline EOT reading and then covered from the ambient light in the room until changes in the data were stabilized. During that time period, the effective optical thickness, EOT, reading dropped a significant 16nm, indicating that fluctuating ambient light between the spectrometer and the biosensor platform (as shown in Figure 1) is a source of noise that should be considered in the redesign.
Testing was also performed on the effects of temperature on the
noise of the system. Considering the reflective reading from the flow of
water over a porous SiO2 layer (surface of the chip) and the EOT of
17000nm, a one degree temperature shift causes a 1nm drift in EOT
value. Based on this data for a typical setup with EOT value around
20,000nm, a 3C change in ambient temperature would result in a 7nm
change in the EOT value, a significant change regarding the baseline
noise level. For shorter experiments (<1hr), temperature should not
affect the readings given that ambient temperature stays relatively
constant during this time. However, for experiments that are run for
hours or days, fluctuating ambient temperature can cause considerable
changes in reading. Hence, a thermal control is needed in the redesign.
Lastly, the flow properties within the chamber were tested to quantify if the circular chamber shape (Figure 2) added to the noise. To do so, the time to equilibrium was tested for multiple flow rates, and at different locations on the chip. As the location of acquisition moved further away from the center of the chip, there were noticeable irregularities in the data, along with much slower time response compared. This showed that not only did the current chamber shape slow acquisition in some cases; it also was creating flow irregularities that were affecting data. The noise due to flow irregularities was extremely large in some cases (±15nm), showing that optimization of flow is a necessary step in minimizing overall noise.
Lastly, the flow properties within the chamber were tested to quantify if the circular chamber shape (Figure 2) added to the noise. To do so, the time to equilibrium was tested for multiple flow rates, and at different locations on the chip. As the location of acquisition moved further away from the center of the chip, there were noticeable irregularities in the data, along with much slower time response compared. This showed that not only did the current chamber shape slow acquisition in some cases; it also was creating flow irregularities that were affecting data. The noise due to flow irregularities was extremely large in some cases (±15nm), showing that optimization of flow is a necessary step in minimizing overall noise.
Design Problem Statement
The main sources of noise contributing to the effective optical thickness (EOT) readings in reflective spectrum are ambient light and temperature changes, and flow conditions. A new flow cell design needs to be created to minimize the effects of these sources of noise on optical bio-sensing with porous silicon chips. Reducing the effects of these sources of noise will ultimately increase the SNR, which is needed to detect lower concentrations and smaller particles.
-Minh Phan