現今的積體電路隨著半導體製程技術發展,電子元件的數量與密度急遽增加,伴隨而來的是單位面積高發熱量的問題,這將會讓晶片操作於較高的環境溫度,結果導致電子元件速度與穩定性大幅降低;以2004年Intel發表的CPU而言,單位面積發熱率可達7.80W/cm2,總發熱量高達109.6W[2];未來如何能有效將熱能導出並且提高散熱效能,是積體電路發展刻不容緩的課題。 目前電腦中最常使用的散熱方法為均熱片與風扇結合的裝置,惟此種方法提供之散熱能力已經面臨臨界點,2004年Intel的CPU發展進度即出現因散熱效能不足而受阻。熱管為另一種常見的散熱裝置,熱管利用相變化提供優異的散熱能力,小型的熱管已經應用在筆記型電腦中。在1984年Cotter 提出微熱管 (Micro Heat Pipe) 可做為電子元件散熱的方式後[17],關於微熱管的理論分析與實驗量測在二十年來如雨後春筍般被提出,惟在微尺度之流體熱傳行為仍存在諸多尚待探討的議題,尤其對於重要的流道內部的溫度場分佈數據,一直要到最近兩、三年才有Zohar [10, 26]從事微熱管現地溫度量測與數據分析。 本研究藉著微機電製程之體型及表面微加工技術研製微熱管晶片,其流道寬度為150μm、長度為15mm;同時首度將白金熱阻式微溫度感測器整合於微熱管管壁上,藉以量測微熱管內部的現地溫度分布,並藉以估算所能提供的散熱能力,希望能對於未來電子元件散熱技術之進展有所貢獻。 本實驗著重的課題有以下四點:一、試著將微加熱器、微熱管與微溫度感測器整合在同一測試晶片上。二、研究陽極接合及環氧樹脂膠黏技術應用於結合4吋(100)矽晶圓與Pyrex7740玻璃的可能性。三、使用簡單且價格低廉的方法進行真空注水。四、量測微熱管在不同施加功率下的溫度分佈並分析熱傳性能。 本實驗使用微加熱器模擬微熱點長時間發熱的情形,用微熱管作為微型散熱元件,測試不同填充率的晶片軸向溫度的分佈;沒有填充工作流體以作為參考用的Chip_0在功率為800mW時即燒毀,有微熱管作動的晶片在功率為1W的情況下仍能運作;接著用重力輔助工作流體、在冷凝端加上熱沉兩種方式進行實驗,發覺兩種方式都對於微熱管作動都有所幫助;雖然加上微熱管的實驗晶片最大熱傳係數分別只有8.333、4.545 W/mK,但是比起未填充工作流體的晶片卻提升了3至五倍,未來希望能以改變接合方式來改善微熱管的性能。
Due to the improvement of semiconductor manufacturing techniques, the quantities and densities of electric components on a single chip increase quickly. The heat dissipated of miniaturized make chips to be operated in higher temperature which will decrease not only the speed of electric components but also the stability of the chips. Taking the CPU, “Intel Pentium 4 Processor Extreme Edition on 0.13 Micron Process in the 775-land Package” published in 2004 for example, its thermal design power can reach about 109.6 Watt. To remove heat generated inside an operating chip becomes more and more important from now on. The most popular electric cooling devices are fins and fans now, but those devices face the critical point of the ability to transfer heat from local hot spot. In 2004 Intel slows down their new products of CPU, and can not find a suitable device to provide cooling is one of the reasons. Heat pipe is one of the well known cooling devices, and some of them have been integrated with Notebook cooling module. Since Cotter suggest that Micro Heat Pipe could a new electric cooling device in the future, experiments and researches are proposed at the past 20 years. However, the research about how the working fluid inside micro heat pipes still have a lot of issues to be worked out. In this experiment I fabricate micro heat pipe chip using surface machining and bulk machining of MEMS manufacturing. The the channel of micro heat pipe is 150μm in width, and 15mm in length. This is the first time to fabricate a micro heat pipe chip integrated with micro heater and In-situ micro temperature sensor array. I hope that this research can promote the development of the electric cooling industry. There are four subjects in this experiment: First, integrated micro heater and micro heat pipe and micro temperature sensors in one single chip; Second, combining the glass chip and silicon chip using anodic bonding and epoxy pasting; Third, filling working fluid in a convenient way; Fourth, analyzing the behavior of the micro heat pipe under different applied works. I use micro heater to simulate hot spot, and use micro heat pipe as a cooling device to find out the performance of this design. Chip_0 which no working fluid is filled burned out when 800mW is applied on the heater; however, Chip_1 and Chip_2 which is filled several working fluids can work when 1W is applied on the heater. Then, gravity and heat sink are use to improve the performance of the micro heat pipe, and both of them are workable. Although the maximum thermal conductivity of Chip_1 and Chip_2 are about 8.333 and 4.545 W/mK, but they improve the maximum thermal conductivity about 3 to 5times than Chip_0.