3D Architecture and Replaceable Layers for Label-Free DNA Biochips Yuksel Temiz, Sandro Carrara*, Andrea Cavallini, Yusuf Leblebici, and Giovanni De Micheli Ecole Polytechnique Fédérale de Lausanne (EPFL) CH-1015 Lausanne (CH) * Corresponding author: sandro.carrara@epfl.ch Abstract - Recent advances in bio-sensing technologies have led to design of bio-sensor arrays for rapid identification and quantification of various biological agents such as drugs, gene expressions, proteins, cholesterol, fats, etc. Various dedicated sensing arrays are already available commercially to monitor some of these compounds in a sample. However, monitoring the simultaneous presence of multiple agents in a sample is still a challenging task. Multiple agents may often attach to the same probes on an array which makes it difficult to design a chip that can distinguish such agents (leading to low specificity). Thus, sophisticated algorithms for targets identification need to be implemented in biochip in order to maximize the number of distinguishable targets in the samples. The proposed algorithms are also required to introduce sophisticated signal processing and more intelligence on-chip. Dealing with these new processing and information technology demands constraints also require more innovative approaches towards hybrid integration technologies. To address such new demands, we discuss in this paper an innovative 3D-integrated bio-chips especially dedicated to label-free DNA detection. I. INTRODUCTION DNA microarrays are the most commonly used biosensors for gene expression measurements. They work on the principle of identifying mRNAs in a sample using complementary cDNA strands as probes. In addition to the quantification of gene expression, microarrays can also be used for identifying the presence or the absence of an organism in a sample. In such an application, a set of oligonucleotides (synthetic single strand DNA) can be used as probes that hybridize to the genome of the organism under investigation. The same principle can be extended to the identification of a certain set of viruses in a sample. In that case, the probes used are specific to the genome of each virus in the sample. This would work if the genomes of the viruses are very different from each other so that the designed probes can be unique to each virus. But very often biological applications, such as monitoring of water pollution, would require identification of viruses that are closely related or have similar genome. In such situations, it becomes difficult to find unique probes for each virus. To monitor multiple targets in a sample, a non-unique probe design method has been proposed in the past [1, 2], where the designed probes may hybridize to multiple targets (or viruses) such that it is still possible to differentiate between any two viruses in the sample. A possible solution of the generalized formulation for distinguishing the simultaneous presence of DNA-targets in a sample has been given in [3]. This fact demands for new, advanced and innovative systems able to address more intelligence on board. This can require large amounts of memory and computation, a serious drawback for portable and real-time applications. High level of programmability and flexibility with 3-D Chip hardware fabricated by using 0.7-μm CMOS technology has been demonstrated for individual classifiers [4]. Thus, 3-D chip architectures are suitable for implementing more intelligence in fully-electronic and portable DNA chip. Complex fully electronic DNA chip should also be reusable for economical reasons. Reusable chip concept can be addressed by developing a replaceable bio-layer on top of the 3D-stack chip structure. Aim of this paper is to present an innovative solution for 3D-DNA-Biochip with replaceable top layer and to show preliminary results in developing such a system. II. 3D CHIP WITH REPLACEABLE BIOLAYER For decades, the continuing demand for faster and smaller integrated circuits (ICs) has led to aggressive downscaling in transistor sizes. As a result, today’s state-of- the-art IC technology has offered fascinating levels of performance and functionality, while introducing new barriers and challenges for further downscaling. As conventional planar IC designs have almost reached their limits, the three-dimensional integrated circuit (3D-IC) technology offers new possibilities to continue the trend predicted by Moore’s Law [5]. This emerging technology enables the integration of multiple layers with vertical interconnections, providing potential performance improvements even in the absence of continued device scaling [6]. Today, there has already been an extensive research on 3D memories, FPGAs, and multiprocessors, where each individual chip in the stack has identical or similar properties. Besides this homogenous integration, the 3D-IC technology also offers novel opportunities for the realization of systems comprising heterogeneous layers, such as RF, analog, digital, MEMS, etc. We believe that this technology also holds great promise for biosensor arrays, which is totally a new application area for this technology and still conceptual. Figure 1 shows the illustration of a 3D-integrated biosensor system composed of heterogeneous chips [7]. In this system, it is aimed to improve overall performance by employing different fabrication technologies for each particular function, such as custom micro- fabrication for the bio-layer, specific technologies enabling low-noise operation for analog electronics, and high speed/density CMOS technologies for digital circuits and memories. 978-1-4244-4709-1/09/$25.00 ©2009 IEEE 35