دانلود مقاله ISI انگلیسی شماره 9221
ترجمه فارسی عنوان مقاله

روش تجربی و مدل شبیه سازی FE برای کاوش پارامترهای ارزیابی ویفر

عنوان انگلیسی
Experimental method and FE simulation model for evaluation of wafer probing parameters
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
9221 2006 13 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Microelectronics Journal, Volume 37, Issue 9, September 2006, Pages 871–883

ترجمه کلمات کلیدی
آزمون کاوش ویفر - کاوشگر سوزن - آزمون نیروی کوچک - روش المان محدود
کلمات کلیدی انگلیسی
پیش نمایش مقاله
پیش نمایش مقاله  روش تجربی و مدل شبیه سازی FE برای کاوش پارامترهای ارزیابی ویفر

چکیده انگلیسی

Wafer probing technology is a critical testing technology that is used in the semiconductor manufacturing and packaging process. A well-designed probing system must enable low and stable contact resistance when each needle-like probe makes contact with the IC chip-bonding pad. Mechanical contact using excessive probe force causes over-sized scrub marks that may damage the die pad and sizably deform the probe tip. In this paper, an experimental setup of a single tungsten needle probe making contact with an Al pad was employed to investigate the relationships between the overdrive, contact force, and scrub mark length. A three-dimensional computational probing simulation model was developed for analyzing dynamic deformations of the contact phenomena during wafer testing. The mechanical tensile strength of the tungsten needle was tested with a micro-tester to examine the tensile stress–strain relationship. The elastoplastic behaviors of the probe and die were taken into account in the simulation model. The resultant scrub lengths from the simulation were verified against the experimental data. Additional critical data, such as data of the scrub mark sinking on the die surface and the maximum Von-Mises stress level location at the probe tips, can be predicted. The experimental and numerical methods presented here can be used as useful performance evaluation tools to support the choice of suitable probe geometry and wafer probe testing parameters.

مقدمه انگلیسی

Wafer probe cards can be divided into two major classifications: needle types [1], [2] and [3] and membrane types [4]. Needle types can be further divided, based on the shape of the pins, into vertical needle types and horizontal needle types. New types of probe cards, such as micro-spring probe cards [5] and [6] and MEMS probe cards [7], are also under development, and some of them have already been commercialized. Up to now, the cantilever needle card is still the most widely probe card used in IC testing. It usually consists of four major parts: needle-like probes, Epoxy card, isolation base, and PCB. Performing stable wafer testing requires consideration of the following critical parameters: (a) the shape of the probe, including tip type, tip diameter, tip length, and bend angle; (b) the probe material; (c) wafer contact speed; and (d) overdrive distance (OD), where the overdrive is the distance moved from the point of initial contact with a bonding pad. During probe testing, electric contacts between the probes and pads are achieved by applying a large force, with the help of scratching motions, to break the oxide film on the metal pads [8]. However, excessive probe contact force may damage the die pad or induce over-sized scrub marks on the die surface that adversely influence the wire bond or flip chip quality during packaging. In addition, sizable deformed probe tips cause more than the usual number of particles to accumulate at the top of the contact region and to increase contact resistance. Research has been carried out to investigate probe test parameters in order to improve probe card performance. Varnau [9] evaluated the flip chip reliability of the probe-before-bump process using different overdrives and different contact times. Tan et al. [10] estimated the package reliability of flip chips and used two possible probe-before-bump processes, probe-on-aluminum and probe-on-Ni/Au. These two processes were evaluated with different overdrives, temperatures, and probe cards taken into consideration. Hotchkiss et al. [11] investigated the relationship between scrub mark damages ranging from 10% to 45% of the total pad area and the package reliability of wire bonding.

نتیجه گیری انگلیسی

Both an experimental method and a 3D computational model for studying the relationship between needle geometry, test parameters, and sizes of scrub marks on die pads are presented in this paper. The single-needle probe test results indicate that the contact force and scrub length increased when the overdrive increased. For the same overdrive, however, the scrub length and contact force increased when the bend angle decreased. From the above results, it can be seen that when the contact angle between the probe tip and bond pad increases, the bond pad will be seriously damaged. The elastoplastic material behavior of tungsten needle was also tested using micro-force tensile tests. The test results indicate that the stress–strain relationship of the needle did not vary significantly with temperature in the elastic range. In addition, the method that resulted in the stress–strain curve was implemented into the computational model. The scrub marks from numerical simulation were compared with the profiles obtained from the single-needle probe tests. The discrepancy in the scrub length between the two methods is less than 9%. The results from the numerical simulation are as follows: