• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.06a
• /
• pp.434-437
• /
• 2005

Chemical Mechanical Polishing (CMP) has become an essential step in the overall semiconductor wafer fabrication technology. In general, CMP is a surface planarization method in which a silicon wafer is rotated against a polishing pad in the presence of slurry under pressure. The polishing pad, generally a polyurethane-based material, consists of polymeric foam cell walls, which aid in removal of the reaction products at the wafer interface. It has been found that the material removal rate of any polishing pad decreases due to the so-called 'pad glazing' after several wafer lots have been processed. Therefore, the pad restoration and conditioning has become essential in CMP processes to keep the urethane polishing pad at the proper friction coefficient and to allow effective slurry transport to the wafer surface. Diamond pad conditioner employs a single layer of brazed bonded diamond crystals. Due to the corrosive nature of the polishing slurry required in low pH metal CMP such as copper, it is essential to minimize the possibility of chemical interaction between very low pH slurry (pH <2) and the bond alloy. In this paper, we report an exceptional protective coated conditioner for in-situ pad conditioning in low pH Cu CMP process. The protective Cr-coated conditioner has been tested in slurry with pH levels as low as 1.5 without bond degradation.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.06a
• /
• pp.86-89
• /
• 2003

Chemical Mechanical Planarization(CMP) has been accepted as the most effective processes for ultra large scale integrated (ULSI) chip manufacturing. However, as the polishing process continues, pad pores get to be glazed by polishing residues, which hinder the supply of new slurry. And pad surface is ununiformly deformed as real contact distance. These defects make material removal rate(MRR) decrease with a number of polishied wafer. Also the desired within-chip planarity, within wafer non-uniformity(WIWNU) and wafer to wafer non-uniformity(WTWNU) arc unable to be achieved. So, pad conditioning in CMP Process is essential to overcome these defects. The eletroplated or brazed diamond conditioner is used as the conventional conditioning. And. allumina long fiber, the jet power of high pressure deionized water, vacuum compression. ultrasonic conditioner aided by cavitation effect and ceramic plate conditioner are once used or under investigation. But. these methods arc not sufficient for ununiformly deformed pad surface and the limits of conditioning effect. So this paper focuses on the characteristics of diamond conditioner which reopens glazed pores and removes ununiformly deformed pad away.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2007.11a
• /
• pp.47-48
• /
• 2007

A CMP Process has many factors that affect result of a polished wafer. Dominant factors are velocity, pressure and temperature in process. A pad profile is also considered as affecting factor of CMP. Accoding to variation of a pad profile, the each pan of a wafer is differently pressured. It appears to affect the uniformity of a wafer. A pad profile varies as a swing arm conditioner which have been ordinarily used in industry. A swing arm conditioner has several sectors in its swing path. This study aims that a wafer get a good uniformity as swing arm conditioner's path on pad is analyzed and simulated. Through the simulation, tendency of pad profile after conditioning will be predicted and the result of simulation compared with the result of experiment. The optimized pad profile would be made by to vary swing arm's velocity on each sector. In order to maintain the optimized profile, conditioner design or swing arm's velocity should be changed and designed.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.21 no.11
• /
• pp.963-967
• /
• 2008

There are many factors to affect polishing performance normally in chemical mechanical polishing (CMP) process. One of the factors is a pad profile. A pad profile has not been considered as a significant factor. However, a pad profile is easily changed by conditioning process in CMP, and then changed pad profile affects polishing performance. Therefore, understanding how the pad profile is changed by conditioning process is very important. In this paper, through the simulation based on kinematic analysis, the variation of the pad profile was described in accordance with difference condition of conditioning process. A swing-arm type conditioner was applied in this simulation. A swing-arm type conditioner plays a role of generating asperities on pad surface. The conditions of conditioing process to get uniform removal were also investigated by comparing the simulation with the experiment.

• Tribology and Lubricants
• /
• v.35 no.6
• /
• pp.330-336
• /
• 2019

Conditioning is a process involving pad surface scraping by a moving metallic disk that is electrodeposited with diamond abrasives. It is an indispensable process in chemical-mechanical planarization, which regulates the pad roughness by removing the surface residues. Additionally, conditioning maintains the material removal rates and increases the pad lifetime. As the conditioning continues, the pad profile becomes unevenly to be deformed, which causes poor polishing quality. Simulation calculates the density at which the diamond abrasives on the conditioner scratch the unit area on the pad. It can predict the profile deformation through the control of conditioner dwell time. Previously, this effect of the diamond shape on conditioning has been investigated with regard to microscopic areas, such as surface roughness, rather than global pad-profile deformation. In this study, the effect of diamond shape on the pad profile is evaluated by comparing the simulated and experimental conditioning using two conditioners: a) random-shaped abrasive conditioner (RSC) and b) uniform-shaped abrasive conditioner (USC). Consequently, it is confirmed that the USC is incapable of controlling the pad profile, which is consistent with the simulation results.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2008.06a
• /
• pp.114-114
• /
• 2008

반도체 device의 성능을 향상시키기 위하여 패턴은 더욱 더 고 집적화 되고 배선 또한 다층배선 구조를 가지게 되었으며 요구되는 선폭 또한 더욱 미세화 되어 CMP 공정이 도입되게 되었다. 이러한 CMP 공정에 사용되는 소모품으로는 크게 세 가지의 중요한 부분으로 나눌 수 있다. 그것은 slurry와 pad, conditioner이다. 그중에 pad conditioning 공정은 CMP 공정시 pad의 마모에 따라 감소하는 removal rate(RR)값을 회복시키기 위한 공정으로 마모된 pad의 표면을 활성화 시켜주는 중요한 공정이다. 하지만 pad conditioning 공정을 장시간 진행하게 되면 conditioner 표면에 오염물이 발생하게 되며, 오염물로 인하여 wafer표면에 scratch 및 defect을 발생시키는 원인이 될 수 있다. 이러한 문제점을 보완하기 위하여 conditioner의 표면을 변화시켜 공정중의 오염이 발생하지 않도록 하는 것이 중요하다. 본 논문에서는 oxide CMP 실험을 통하여 conditioner표면에 오염물이 발생함을 확인하였으며 energy dispersive spectroscopy(EDS) 분석을 통하여 주오염물의 성분이 oxide slurry중 silica임을 확인하였다. Conditioner의 표면을 소수성으로 만들기 위하여 self assembled monolayer(SAM) 방법을 이용하여 표면에 코팅을 하였으며, 소수성 박막이 코팅된 conditioner와 코팅되지 않은 conditioner의 비교 실험을 통하여 오염 정도를 비교하였다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.10a
• /
• pp.362-365
• /
• 2005

Currently Chemical Mechanical Planarization (CMP) has become an essential step in the overall semiconductor wafer fabrication technology. Especially the CMP pad conditioner, one of the diamond tools, is required to have strong diamond retention. Strong cohesion between diamond grits and metal matrix prevents macro scratch on the wafer. If diamond retention is weak, the diamond will be pulled out of metal matrix. The pulled diamond grits are causative of macro scratch on wafer during CMP process. Firstly, some results will be reported of cohesion between diamond grits and metal matrix on the diamond tools prepared by three different manufacturing methods. A measuring instrument with sharp cemented carbide connected with a push-pull gauge was manufactured to measure the cohesion between diamond grits and metal matrix. The retention force of brazed diamond tool was stronger than the others. The retention force was also increased in proportion to the contact area of diamond grits and metal matrix. The brazed diamond tool has a strong chemical combination of the interlayer composed of chrome in metal matrix and carbon which enhance the interfacial cohesion strength between diamond grits and metal matrix. Secondly, we measured real-time data of the coefficient of friction and the pad wear rate by using CMP tester (CETR, CP-4). CMP pad conditioner samples were manufactured by brazed, electro-plated and sintered methods. The coefficient of friction and the pad wear rate were shown differently according to the arranged diamond patterns. Consequently, the coefficient of friction is increased according as the space between diamonds is increased or the concentration of diamonds is decreased. The pad wear rate is increased according as the degree of diamond protrusion is increased.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2007.05a
• /
• pp.27.2-27.2
• /
• 2007

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2006.05a
• /
• pp.19.2-19.2
• /
• 2006

• Tribology and Lubricants
• /
• v.40 no.2
• /
• pp.67-70
• /
• 2024

Chemical mechanical planarization (CMP) is an essential process for ensuring high integration when manufacturing semiconductor devices. CMP mainly requires the use of polyurethane-based polishing pads as an ultraprecise process to achieve mechanical material removal and the required chemical reactions. A diamond disk performs pad conditioning to remove processing residues on the pad surface and maintain sufficient surface roughness during CMP. However, the diamond grits attached to the disk cause uneven wear of the pad, leading to the poor uniformity of material removal during CMP. This study investigates the pad wear rate profile according to the swing motion of the conditioner during swing-arm-type CMP conditioning using deep learning. During conditioning, the motion of the swing arm is independently controlled in eight zones of the same pad radius. The experiment includes six swingmotion conditions to obtain actual data on the pad wear rate profile, and deep learning learns the pad wear rate profile obtained in the experiment. The absolute average error rate between the experimental values and learning results is 0.01%. This finding confirms that the experimental results can be well represented by learning. Pad wear rate profile prediction using the learning results reveals good agreement between the predicted and experimental values.