AVS2011 Session BI-MoA: Sensors and Fluidics for Biomedical Applications

We have demonstrated the improvement of sensitivity by biasing the AlGaN/GaN HFET biosensors into the subthreshold regime. To bias the channel into the subthreshold regime, V_g=~-4 V was. The gate voltage causes ions in the physiological buffer solution to move and result in measurement noise. To avoid side effects from a high gate voltage, we adopted the gate recess process to shift the subthreshold gate voltage to zero/near zero volt and retain high sensitivity.

The AlGaN/GaN heterostructure used in this study has a 23 nm thick undoped AlGaN barrier. The recession of AlGaN barrier was conducted with an Oxford Plasmalab 100 system. A two-step recession process was used. The first step uses BCl₃ to etch the AlGaN barrier and the second step uses a combination of Cl₂/N₂/O₂ to passivate the etched surface. To achieve zero/near-zero subthreshold gate voltage, we recessed devices with the first step duration to be 50, 55, 60, 65, and 70 sec respectively. We also fabricated diodes with the Schottky area recessed with the same conditions to check the material property changes. The measured C-V characteristics show that the threshold voltage shifts along the positive direction with the increase of etching time. With 70 sec etching time, the threshold voltage of Ni diodes achieve subthreshold regime at Va=0 V. Extracted from C-V curves, the etched depths are 12.0, 11.6, 11.2, 9.7, and 8.8 nm respectively. AFM measurements of both recessed and original AlGaN surfaces show that the surface smoothness is improved after recession processes.

The fabricated AlGaN/GaN HFET biosensors have a recessed gate of 100 um (L) × 2 mm (W). To detect streptavidin, the AlGaN surface was modified by silanization and biotinylation. The measured Id-Vg curves of AlGaN/GaN HFETs with (a) recession time of 65 sec, and without any recession. There are two major differences: 1) the threshold voltage of recessed device shifted to around 0 V; and 2) the off-state current of recessed device is 1-2 orders lower. The threshold voltage promises high sensitivity when the gate is floating or biased at a very low voltage. The decrease of the off-state current indicates that signal-to-noise ratio is improved. The difference between the threshold voltage of diodes and HFETs mainly because of 1) the diode Schottky metal is Ni and the HFET gate is PBS and 2) the HFET is treated with oxygen plasma for surface modifications. The detection of 25 pg/ml (473 fM) SA solution in 1X PBS shows that the drain current is decreased by 22.7% because SA carry negative charges in PBS solution. The sensitivity is increased by one order compared to our previously published results with non-recessed devices.

AlGaN/GaN heterojunction features two-dimensional electron gas which is highly sensitive to proximal charges and excellent stability against chemicals. These unique advantages imply the potential of AlGaN/GaN based field effect transistor (FET) in facilitating various biological and chemical studies. We have previously reported the detection of hybridization between probe DNAs and fully complementary target DNAs with AlGaN/GaN HFET biosensors. To further improve the sensitivity, comprehensive understanding of the working principle is necessary. In this work, we focus on a systematical investigation on the effects of the distance between the charged layer and the sensing surface, i.e., the distance from target DNA biomocules to the AlGaN surface. We immobilized a monolayer of single-strand probe DNAs modified with thiol groups on the Au-coated active gate surface of an AlGaN/GaN HFET. Different target ssDNA molecules were designed so that distance between the hybridization sites and the AlGaN surface varied. Experimental results show an explicit relationship between the detection sensitivity and the charge layer distance.

Six types of synthesized 25-mer oligonucletides (Allele) solutions were prepared, including a probe DNA, a fully complementary target DNA, a 21-bp complementary DNA, a 13-bp complementary DNA, a 9-bp complementary DNA and a mismatched DNA. The amount of charges is assumed identical for DNAs with a constant length. The distance of the charge layer can be well controlled by varying the hybridization sites of target DNAs to the probe DNAs. For example, the 21-bp complementary DNA is 4-bp further away from the surface than the fully complementary DNA.

We compared the I_DS-V_DS characteristics of devices before and after hybridization. A decrease of drain current was noticed, revealing the binding of negatively charged DNAs. More importantly, the decreases in I_DS-V_DS demonstrated a clear dependence of charge layer distance by an observation of the current change ratio, defined by ΔI_DS / I_DS, (gate bias: -1 V): 16.1% for fully complementary target DNA (0.1 µ mol), 14.9% for 21-bp complementary DNA, 10.9% for 13-bp complementary DNA, 6.89 % for 9-bp complementary DNA, and 0.75% for mismatch DNA, which was comparably insignificant. The current change decreases with the distance between the target DNA and sensing surface, although the amount of charges for the target DNA is constant. These results demonstrate that the detection sensitivity is significantly dependent on the charge layer distance. Theoretical analysis for varying charge layer distances of target DNAs will also be studied and compared with the experimental results.

The past decade has witnessed the use of microcantilevers as mechanical transducers of molecular recognition and for the development of miniaturized and sensitive biochip platforms. Microcantilever based biosensors can be based on either mass adsorption or surface stress variation. The potential to operate a microcantilever sensor in liquid medium with extreme sensitivity makes it an ideal choice for the development of biological sensors. Selective detection is obtained by immobilizing receptor molecules on one side of the cantilever. Microcantilever based biological sensors predominantly operate in a liquid environment; this is done in order to retain the functionality of the biomolecules immobilized on the surface. Typically reference cantilevers serve to observe and analyze the effect of non specific interactions and fluid flow rate from specific biomolecular interactions. We have additionally observed that the interaction of analytes on the functionalized surface of the cantilever is influenced by the varying the flow rate of the solution used. The surface stress observed due to the adsorption of analyte molecules on to the receptors on cantilever surface in static condition is much higher compared to a dynamic condition where analyte molecules are allowed to flow across the cantilever surface with the help of a flow control system. Here we address the effect of flow rate on the biomolecular adsorption kinetics of the system and how it affects the sensitivity of Microcantilever based biological sensors.

Sensors for measuring pH are very important for understanding reactions of biological species such as proteins, enzymes and cells. While traditional sensors based upon such techniques as infrared spectroscopy, fluorescence and others have low sensitivity and slow response time, more recent nanowire field effect transistor sensors offer improved sensitivity and response time due to smaller size and increased surface areas. [1, 2]

We have fabricated nanowire FET sensors using conventional CMOS semiconductor processes. The nanowires were patterned by electron beam lithography and reactively ion etched into 30 nm thick silicon on insulator (SOI) substrates. The gate sensing surface is comprised of a hafnium oxide/ silicon dioxide stack covering the nanowire, while the source/drain regions are comprised of boron activated SOI with nickel-platinum silicide contacts. The resulting 16 nm wide nanowire devices show high sensitivity for pH measurements. The FET drain current increases by a factor of eight per unit change in pH, while the subthreshold slope is ~ 77 mV/decade. The sensors operate at a reduced sensing voltage of 0.5 V, making them promising candidates for low power, bio-medical applications.

[1] E. Stern, J. Klemic, et al., Nature, 445 (2007) 519

[2] S. Zafar, C. D’Emic, et al, to be published

The use of nanotechnology tools has opened new opportunities to explore analytical applications of the nano-engineered materials which attracted great attention due to their unique morphology, extraordinary physical and chemical properties towards development of biosensors to facilitate the improvement of the selectivity and sensitivity of the current methods. In this work, the importance of semiconducting La³⁺ doped TiO₂ a metal-oxide-based nanostructure platform is highlighted for biosensors platforms. La³⁺ doped TiO₂ nanostructure having nanometer-scale inner-core cavity which were exposed to the outer surface with different oxidation states having possibilities for redox-activity can make them attractive for sensing uses. Therefore, the use of La³⁺ doped TiO₂ for the development of electrochemical sensors will be discussed. These platforms has been characterized by XRD, XPS, FTIR, SEM, cyclic voltametry to determine structure, surface chemistry and electron transfer characteristics for biosensor applications. Cholesterol oxidase immobilized onto La³⁺ doped TiO₂-based nanostructured surfaces exhibited a pair of well-defined and quasireversible voltammetric peaks in CV measurements. We will also discuss the potential prospect of these surfaces as low cost stable platforms for biomedical diagnosis.

Antibody films for the use in biosensors and assays are usually deposited via wet-chemical attachment methods. The presented experiments demonstrate that pneumatic spray deposition of antibody thin films from aqueous solution yields films of similar sensitivity and durability without special surface treatments and attachment chemistries. The experiments were performed using a commercially available low flow nebulizer in combination with a syringe pump and a substrate rotation stage to homogenize the coating.

In the experiments E. coli O157:H7 antibody was deposited on cleaned microscopy glass slides without any other pre-treatment. Standard wet-chemically prepared silanized glass slides using the avidin-biotin attachment scheme were also prepared for direct comparison of sensitivity and longevity of the spray based substrates. After incubation with GFP-labeled E. coli O157:H7 cells (ATCC 35150) ranging from 10⁴-10⁶ CFU/ml the slides were rinsed and AF647-labeled detector antibody was added and incubated. After rinsing and drying the slides were interrogated with a 635 nm laser and visualized using a CCD camera. Slides were also visualized by epifluorescence microscopy to examine antibody patterns and determine E. coli capture efficiencies.

The results of the experiments demonstrate that there is little difference between spray and standard protocol wet-chemically prepared substrates. This indicates that antibody films can be prepared via physisorption without complex attachment chemistries, and that antibodies can directly attach to glass slides, while retaining their functionality.