TECHVIEW. Effects of Diamond Shape and Size on Polyurethane Pad Conditioning. abrasive technology

Transcription

1 abrasive technology TECHVIEW Effects of Diamond Shape and Size on Polyurethane Pad Conditioning Mark Bubnick, Ph.D., Sohail Qamar, Stephen McGregor, Ph.D., Thomas Namola, and Timothy White Abrasive Technology, 84 Green Meadows Dr., Lewis Center, OH 4335

2 Polyurethane Pad Conditioning Effects of Diamond Shape and Size on Polyurethane Pad Conditioning Mark Bubnick, Ph.D., Sohail Qamar, Stephen McGregor, Ph.D., Thomas Namola, Timothy White Abrasive Technology, 84 Green Meadows Dr., Lewis Center, OH 4335 Tel: , Fax: , ABSTRACT The effects of tightly controlled diamond size and diamond shape on the performance of pad conditioning disks were investigated. Pad removal rate performance of conditioning disks was demonstrated on Rodel IC1 pads and Thomas West polyurethane pads. Results indicate that by manipulating the shape of diamonds used on pad conditioning disks 1, the polishing pad removal rate can be varied by more than 4 on a Rodel IC1 pad. The impact of tightly controlling diamond size distribution on polishing pad removal rate varies with mean diamond size. By correctly controlling the diamond shape and size distribution, the conditioning disk 2 and polishing pad life can be increased. Data indicates that a similar relationship between diamond shape and pad removal rate exists on Thomas West polishing pads. Manipulating diamond shape on the conditioning disk allows the polishing pad removal rate to be controlled. INTRODUCTION Several diamond variables were investigated to determine their effects on PRR (polishing pad removal rate), RAS (Relative Abrasive Sharpness 3 ), and conditioning disk life. The variables investigated were diamond concentration, diamond size distribution, diamond shape, and diamond mesh size. The effects of the diamond variables are demonstrated on both Rodel IC1 and Thomas West 813-C239G-24 polishing pads. Selected disks were subjected to extended PRR tests in oxide slurry to demonstrate the effects of tight sized, tight shaped abrasive on diamond conditioning disk life. EXPERIMENTS The diamond used in the experiments was characterized by size distribution and shape. Diamond sizing was performed according to ANSI B using precision sieves. Diamond shaping was performed on a vibratory shaping table. The PRR tests were performed on a lab-scale conditioning machine. See Table 1 for test parameters. The conditioning disk platforms were standard Mirra universal mounts. All conditioning pads were 8.25 in diameter. RAS measurements were made using proprietary Abrasive Technology test equipment.

3 Table 1: Test parameters for polishing pad removal rate (PRR) tests Loading force lbs (27 N) Conditioner disk rotation 1 rpm Pad rotation 9 rpm Disk sweep 37 mm at 5 mm/s Slurry DI 12 ml/min Diamond and Diamond Disk Construction The effects of diamond concentration (Fig. 1) and diamond size distribution (Fig. 2) on PRR and RAS were investigated initially. Diamond Concentration 3X 2X 1X 3X 2X 1X Tight Sized Non-Tight Sized Particle Size Distribution Fig. 1: Test disks for abrasive concentration and size distribution Percent Diamond Size Fig. 2: Tight sized and non-tight sized 1 mesh size distribution The next tests compared four grades of 1 mesh and four grades of 22 mesh diamond for a fixed diamond concentration. The diamond size distributions and shape characterizations are shown in Figs. 3 & 4. The 22 mesh diamond size distributions and shape characterizations are shown in Figs. 5 & Mesh Diamond Size Distribution Non-Tight sized diamond Tight sized diamond Mesh Diamond Size Distribution Abrasive Shape A Abrasive Shape B Abrasive Shape C Abrasive Shape D Diamond Size Fig. 3: Size distribution for tight sized 1 mesh diamond Mesh Diamond Size Distribution Abrasive Shape A Abrasive Shape B Abrasive Shape C Abrasive Shape D Diamond Size Fig. 5: Size distribution for tight sized 22 mesh diamond Shape A Shape B Shape C Shape D Blocky 1 Mesh Shaping Characterization Non Blocky Fig. 4: Shape characterization for tight sized 1 mesh diamond 22 Mesh Diamond Shaping Characterization Blocky Shape A Shape B Shape C Shape D Non Blocky Fig. : Shape characterization for tight shaped 22 mesh diamond

4 The last tests compared a standard AT Infinity disk and conditioning disk made with tight sized, tight shaped abrasive; the new AT Infinity v.1 disk. Life testing was run in Rodel Klebosol 3 N 5 slurry and the PRR and RAS decay for the two disks were compared. RESULTS Effects of diamond size distribution and concentration on conditioning disks: There was no discernable trend resulting from diamond concentration or size distribution on polishing pad removal rate and RAS over the ranges tested. (Fig. 7). Pad Removal Rate (mm/hr) Polishing Pad Removal Rate and RAS for diamond size distribution and diamond concentration Tight sized abrasive Non tight sized abrasive Tight sized abrasive (RAS) Non tight sized abrasive (RAS) X 2X 3X Diamond Concentration Fig 7: Effects of diamond size distribution and concentration on PRR and RAS Relative Abrasive Sharpness Effects of diamond shape on polishing pad removal rate and RAS: The PRR and RAS results for the 1 mesh diamond disks on the Rodel IC1 pad are shown in Fig 8. The PRR and RAS results for the 22 mesh disks on the Rodel IC 1 and Thomas West 813 pads are shown in Fig. 9. Pad Removal Rate (mm/hr) Pad Removal Rate and RAS as function of 1 mesh Abrasive Shape Rodel IC 1.1. Pad Removal Rate.14 Calibrated RAS Shape A Shape B Shape C Shape D Fig. 8: Effects of diamond shape on PRR and RAS for 1 mesh disks (Rodel) Relative Abrasive Sharpness Pad Removal Rate and RAS as function of 22 mesh Abrasive Shape Rodel IC 1 and Thomas West 813-C239G Pad Removal Rate (Rodel).5.4 Pad Removal Rate (Thomas West).35 Calibrated RAS Shape A Shape B Shape C Shape D Fig. 9: Effects of diamond shape on PRR and RAS for 22 mesh disks Relative Abrasive Sharpness As the diamond became increasingly non-blocky the PRR and RAS increased. This is demonstrated for both the 1 mesh and 22 mesh size diamond. The 1 mesh diamond showed an increase of more than from the most nonaggressive to most aggressive disk while the 22 mesh showed an increase of more than 4. The 22 mesh conditioners had PPR approximately twice that of the 1 mesh conditioners for a similar grade diamond. Effects of tight size diamond and concentration on disk life: The first tests demonstrated that size distribution and diamond concentration did not have noticeable effects on initial PRR or RAS. However, tight sizing and tight shaping significantly effect useful life of a conditioning disk. Two disks were tested, one standard AT Infinity disk and a tight sized, tight shaped AT Infinity v.1 disk. The disks were subjected to long term conditioning in Klebosol 3 N 5 slurry to compare PRR and RAS decay (Figs. 1 & 11).

5 1.2 Pad Removal Rate Decay as a Function of Time Klebosol 3 N 5 Oxide Slurry, Rodel IC-1 Pad RAS Decay as a Function of Time Klebosol 3 N 5 Oxide Slurry, Rodel IC-1 Pad Normalized Pad Cut-rate AT Infinity AT Infinity v Time (hours) Standardized RAS AT Infinity AT Infinity v Time (Hours) Fig. 1: PRR decay vs. time for the AT Infinity disk and the AT Infinity v.1 disk. Fig. 11: Standardized RAS vs. time for the AT Infinity disk and the AT Infinity v.1 disk. The tight sized, tight shaped abrasive disk showed significantly improved PRR and RAS characteristics. The useful life of the disk increased by more than 5. CONCLUSIONS The mesh size of the diamond crystal greatly effected conditioning disk performance. The 22 mesh diamond was twice as aggressive as a 1 mesh diamond of similar grade. By controlling diamond shape, the PRR by the conditioning disks can be greatly increased or decreased (up to ). This is useful for optimizing pad life and removal rate and developing processes on alternate pads. Diamond particle size distribution and diamond concentration did not have an effect on diamond disk initial PRR or RAS. However, tight sizing and tight shaping the diamond did greatly increase the useful conditioning life of a disk. 1. Mark Bubnick, Sohail Qamar, Thomas Namola, and Dave McClew, Impact of Diamond Conditioning Disk Characteristics on Removal Rates of Polyurethane Polishing Pads, CMP-MIC Proceedings, P.119 (22) 2. Mark Bubnick, Sohail Qamar, Optimizing Diamond Conditioning Disks for the Tungsten CMP Process, CMP-MIC Proceedings, P.43 (23) 3. G. Prabhu, S. Kumaraswamy, D. Flynn, S. Qamar, T. Namola, Pad Live Optimization by Characterization of a Fundamental Pad-Disk Interaction Property, CMP-MIC Proceedings, P.293 (2)

6 World headquarters - Lewis Center, Ohio USA abrasive technology WORLD HEADQUARTERS 84 Green Meadows Dr. P.O. Box 545 Lewis Center, Ohio 4335 USA Ph: Fax: Abrasive Technology EUROPEAN HEADQUARTERS Roxby Place, Fulham London SW 1RT United Kingdom Ph: Fax: ASIAN HEADQUARTERS Blk 193 Lower Delta Rd. 5-11/13 Tiong Bahru Industrial Estate Singapore 1924 Ph: Fax: TAIWAN OFFICE 27 Alley 4, Lane 43 San Ming Rd, Yang Mei Taoyuan, Taiwan Ph: Fax: /5