懸臂樑式生物感測單一晶片之設計與實現

Abstract

由於生物與醫學的產業應用不斷增加，預估2015 年生物感測晶片產值將高達 1000~1500 億日圓。在各式的生物感測晶片設計中，利用微懸臂樑吸附特定物質後造成 懸臂樑共振頻率改變，藉由共振頻率的量測來推估生化物質的濃度是主要作法之一。 本計劃預計使用TSMC 2P4M 製程來設計製作生物感測晶片，晶片內部包含一電熱 驅動、壓阻感測的懸臂樑微振盪器（MEMS 元件），以及一個類比的鎖相迴路（IC 電路）。 所設計的微振盪器與鎖相迴路採「回授控制」的方式結合，即將微振盪器視為受控體， 鎖相迴路視為回授控制器，藉由鎖相迴路內部參數的設定，使微振盪器自動操作於共振 頻率，進而獲得待測物質的濃度。採用此特殊的MEMS 元件與IC 電路的整合方式，其 目的在於簡化IC 感測電路設計，進一步實現一不需外部儀器支援的生物感測單一晶片， 提高晶片量測精度並降低製作成本。缺點是參數設計不當可能導致系統不穩定。 在先前的研究中，本實驗室成功的利用TSMC 2P4M 製程製作出一微振盪器，並實 驗驗證其可行性。惟先前研究尚未針對懸臂樑結構進行最佳化設計，且未整合感測電 路，因此量測精度受限（5.6 pico-gram/Hz）。此外，該振盪器具有feedtrhough、三階系 統、輸入／輸出兩倍頻、等非理想性，使得該設計無法直接與鎖相迴路結合。因此本 計畫的工作重點在於：（1）最佳化微振盪器的設計；（2）修正訊號處理流程；（3）以 回授控制方式來結合鎖相迴路，完成此一生物感測晶片的設計與製作。

The market of the biosensor chip is growing rapidly because of the surge of the biomedical industry. The market is estimated in a size of 1~1.5 billion US dollar in the year 2015. Among various biosensor chips, one major approach is using the cantilever beam to measure the concentrations of bio-chemical materials. In that approach, the resonant frequency of the cantilever beam changes after absorbing the biochemical materials. Therefore, by measuring the change of the resonant frequency, one can obtain the concentration of the biochemical materials. This project aims to develop a biosensor chip using the foundry service TSMC 2P4M process. The chip contains two parts: one is the electrothermal-actuated, piezoresistive-sensored cantilever beam (Micro-resonator, MEMS device), and the other is an analog-type Phase Lock Loop (PLL, IC circuit). The micro-resonator and the PLL are integrated together to form a feedback control system. Meaning that, the micro-resonator is the controlled plant and the PLL is the controller. Through the parameter design of the PLL, the micro-resonator is regulated to operate at its resonant frequency, and thus to obtain the concentration of the measured biochemical material. The advantage of this approach is that the sensing circuits of the device can be simplified, which can achieve a standalone biosensor chip with better sensing accuracy and lower cost. However, the disadvantage is that the system can be unstable if the parameters are not properly designed. In our previous study, we have successfully fabricated a micro-resonator using TSMC 2P4M process. The feasibility of this device is experimentally verified. However, because the micro-resonator design is not optimized and no sensing circuit included, the sensing accuracy of the device is rather limited. Besides, our previous study also indicates that the fabricated device can not work with the PLL directly because of some device imperfections such as feedthrough, doubling frequency in input/ output channels, three-order resonator, etc. Therefore, the research focuses of this project are: (1) the optimal design of the micro-resonator; (2) the modification of the signal processing; (3) the integration of the micro-resonator and the PLL in a feedback control loop. Finally, we will completion the design and fabrication of the proposed biosensor chip.