DOI QR코드

DOI QR Code

A Study on Pressure Distribution for Uniform Polishing of Sapphire Substrate

  • Received : 2016.01.07
  • Accepted : 2016.03.21
  • Published : 2016.04.30

Abstract

Total thickness variation (TTV), BOW, and surface roughness are essential characteristics for high quality sapphire substrates. Many researchers have attempted to increase removal rate by controlling the key process parameters like pressure and velocity owing to the high cost of consumables in sapphire chemical mechanical polishing (CMP). In case of the pressure approach, increased pressure owing to higher deviation of pressure over the wafer leads to significant degradation of the TTV. In this study, the authors focused on reducing TTV under the high-pressure conditions. When the production equipment polishes multiple wafers attached on a carrier, higher loads seem to be concentrated around the leading edge of the head; this occurs because of frictional force generated by the combination of table rotation and the height of the gimbal of the polishing head. We believe the skewed pressure distribution during polishing to be the main reason of within-wafer non-uniformity (WIWNU). The insertion of a hub ring between the polishing head and substrate carrier helped reduce the pressure deviation. Adjusting the location of the hub ring enables tuning of the pressure distribution. The results indicated that the position of the hub ring strongly affected the removal profile, which confirmed that the position of the hub ring changes the pressure distribution. Furthermore, we analyzed the deformation of the head via finite element method (FEM) to verify the pressure non-uniformity over the contact area Based on experiment and FEM results, we determined the optimal position of hub ring for achieving uniform polishing of the substrate.

Keywords

1. Introduction

Consumption of sapphire substrate is getting larger as growing of sapphire application market such as light emitting diode (LED) and sapphire cover-glass. Also, the price competition is intensified as that the number of substrate manufacturing companies are increased. As decreasing of the sapphire substrate cost and increasing of the consumables cost, efficiency of manufacturing process and the unit cost of production are more crucial issue than ever before. Many researchers have been attempting to increase removal rate for reducing the process time and cutting cost of consumable. In sapphire substrate manufacturing process, chemical mechanical polishing (CMP) is one of the most expensive and difficult processes which require long process time and consumption of consumables. Therefore, reducing process time and cutting the cost of the consumables are very attractive topic in the field of sapphire CMP. Traditionally, chemical as approach by formulating slurry has been developed to increase removal rate in CMP. However, chemical tuning method by itself is not enough to increase removal rate effectively due to its stable chemical properties [1-4]. Recently many researchers and companies are changing their approaches from chemical to mechanical method to improve removal rate. In many case, researchers are interested in the process pressure abling to easily predict results. Preston’s equation is representative empirical equation that predicts behavior of removal rate as a function of pressure. Short polishing time and cost down can be expected under the high pressure, but there are still many problems occurred by the high pressure. When sapphire substrate is polished at high pressure on mass-production equipment, removal amount is different according to wafer position.[5] Due to the gimbal structure of the equipment, the sapphire substrates located near leading edge and trail edge on carrier is usually overpolished [6]. The thickness variation as wafer position becomes larger when pressure is increased. In This paper, a hub ring was utilized at high pressure CMP in order to reduce total thickness variation (TTV) of the sapphire wafer.

 

2. Experimental Setup

2-1. Preliminary Experiment

The preliminary experiment was performed for analyzing removal rate and wafer shape as function of pressure in a mass-production equipment. Processing conditions of the experiment are shown in Table 1. Process variable is the pressure in this experiment. Fig. 1 shows that material removal rate increases with polishing pressure. Removal rate is 1.4 µm/hr at 400 g/cm2, and 7.1 µm/hr at 1200 g/cm2, respectively which make 5times difference. On the other hand, when pressure become over 1200 g/cm2, the test could not continue because substrates are slided from regular position because of melting of adhesive wax used for bonding the substrates due to high friciotnal heat. If high temperature is generated by friction, wax mount is melted and substrate are got out from its original regular position. Wax melting is started at about 80℃. It is hard to measuring surface between substrate and pad directly. Usually temperature is detected on pad next to carrier. When temperature is over 40℃ at insitue testing, substrates are dropped out of position.

Table 1.Experimental conditions for preliminary test

Fig. 1.Removal Rate as a function of pressure.

Removal amount are shown in Fig. 2. That result shows about removal amount difference by position. And substrate locations and size are shown in Fig. 3. When substrate position is closed to 0 mm, the position is closed to center of carrier which a flat zone of substrate. In the other hand, When substrate position is closed to 100 mm, substrate is closed to edge of carrier. Removal amount at edge of carrier is larger than that at center. Table 2 shows uniformity of the sapphire substrate after CMP. Within-wafer non-uniformity (WIWNU) at 400 g/cm2 is bigger than that at 1000 g/cm2. It can be seen that the substrate is flatter on high pressure. The WIWNU should be reconsidered with removal amount. In this case, standard deviation (STDEV) should be measured. The standard deviation is getting worse, when pressure becomes higher, namely the substrate shape is worse at high pressure as shown at Fig. 2. Therefore, the high pressure CMP of the sapphire substrate requires more uniform removal pressurising mechanism.

Fig. 2.Removal amount along with substrate position.

Fig. 3.Ceramic plate and hub ring.

Table 2.Uniformity of sapphire substrate

2-2. FEM Analysis

For uniform polishing to overall substrate surface, pressure distribution is changed by using hub ring. Hub rings are attached between carrier and polishing head. Method of attachment of hub ring is shown in Fig. 3. The dimensions of hub ring are shown in Table 3. Polishing pressure is analyzed by Finite Elements Method (FEM). FEM approaches are widely used to explain polishing mechanism [7-9]. At first, when load is distributed by hub ring, it should be noted with respect to changes in pressure applied to sapphire substrate. Therefore structural analysis is proceeded. Fig. 4 shows schematic of finite element model.

Table 3.Dimensions of hub ring

Fig. 4.Schematic of finite element model.

In static state, load applied to substrate is changed to match the position pressure applied to carrier by hub ring. The result shows on Fig. 5. As case change from I to V, load to substrate is changed from flat zone to opposite.

Fig. 5.Pressure distribution at static state.

In dynamic state, load applied to substrate is changed by gimbal structure and pressure distribution. The result show on Fig. 6. Through pressure distribution to 6 substrates, removal trend can be found in dynamic state. 6 substrates adhere to one carrier. By tangential force, load to carrier is focused on opposite site of rotation direction. Pressure distribution is asymmetrically in 6 substrate.

Fig. 6.Pressure distribution at dynamic state.

2-3. Polishing Experiment

Polishing experiments were implemented for more understanding of removal mechanism and shaping trend in mass-production equipment using the hub ring developed. Generally, there are many variables affecting polishing results, but authors focused on the the effect of the hub ring. The experimental condition is same as preliminary condition, except of polishing pressure fixed to 1000 g/cm2. Removal amount can be easily measured at high pressure. And hub ring positions are same as FEM analysis. Fig. 7 shows removal amount trend according to hub ring position. According to the change of pressure distribution, the removal amount trend is followed. When the hub ring is positioned at center position of the carrier (Case I), comparatively more sapphire is removed near flat zone. On the other hand, when the hub ring is positioned at edge position of carrier (Case V), the opposite site from flat zone has higher removal amount. The case IV makes the substrate an ideal flat shape. In other word, uniform polishing is not occurred at case III that load is almost uniformly applied to substrate. Actually, it could be found that uniform polishing is occurred when load is applied with a little outside from center poisition of substrate. This result can be compared with FEM analysis.

Fig. 7.Removal amount as hub ring position.

Fig. 8 shows results removal rate and non-uniformity. Removal rate does not change in spite of changing the hub ring position. Furthermore rem oval rate is alsmost same in both cases with and without hub ring. However, non-unformity changes significantly according to the hub ring position. Finally case IV has the best non- unifomity of 7%.

Fig. 8.Removal rate and non-uniformity as hub ring position.

 

3. Discussion

Overpolishing can be occurred near edge of substrate during CMP because of unstable movement of gimbal structure and rotation of table and head. When table is rotated with load on the head, friction force is generated between substrate and pad surface in the horizontal direction of substrate. By this friction, head is pushed. Front part of the head is pressed and back part is floated in order for gimbal structure. Accordingly, concentration of load is generated to front part. This concentration of load makes removal amount imbalance. This effect is equally generated to mass-production equipment. Fig. 9 shows load changes on mass-production equipment. In preliminary experimental results, this load changed can be found . In the results, when pressure is high, removal rate is also large. But WIWNU of substrate is getting worse. The cause of this tendency is due to non-uniformity of load by gimbal structure.

Fig. 9.Schematics of the moment balance of the head.

Concentrated load should be removed for applying uniform load to overall surface of substrate. By changing pressure distribution, concentrated load can be removed. Hub ring is used for a method of changing distribution of pressure. Position of hub rings are indicated in Table 3 and Fig. 3. Through FEM analysis, it is assured to load distribution are changed by position of hub ring from Fig. 5. So it is predicted that load unformity is the best in case III during polishing from the result. But load distribution have other change at dynamic state as mentioned. Therefore load distribution to whole surface of 6 substrate is need to analyzing at dynamic state by FEM analysis. Result between Fig. 5 and Fig. 6 are different. Load is distributed by hub ring and gimbal sructure is the same time. Load is concentrated oneside of carrier by gimbal and specific part by hub ring from Fig. 6. In real experiment, flat zone part of substrate (carrier cneter) is removed more than edge at case I. This overpolishing phenomenon is caused by changing of load distribution according to hubring and gimbal structure. This is analyzed by FEM analysis result. Such as this result, removal trends of case II ~ V are matched to load distribtion by FEM analysis result. From removal experiment, the best flatness case is IV. During head rotation, the one substrate pass through the position of 6 substrates. So, by analyzing average of load generated in each of substrate from FEM analysis result, removal uniformity of substrate can be analyzed. When analyzed by this way, case IV is the most unifrom in FEM analysis result.

 

4. Conclusions

1. In high pressure CMP, removal rate is increased, but WIWNU is worse. This problem should be improved for epi-ready wafer and cover-glass.

2. When sapphire substrates are polished with mass-production equipment, non-unifrom polishing is generated. Gimbal structure makes concentration of load to edge. So overpolishing is occered on the edge. It makes non-uniformity of sapphire substrate.

3. Distribution of load is changed by using hub ring for removing non-unform polishing. Through experiment and FEM analysis, the best position of hub ring is found that having optimum uniformity. The best position of hubring is when hub ring is located a little outside from center of substrate.

References

  1. X. Hu, Z. Song, Z. Pan, W. Liu, L. Wu, “Planarization machining of sapphire wafers with boron carbide and colloidal silica as abrasives,” Appl. Surf. Sci., Vol. 225, 8230-8234, 2009.
  2. J. Murata, A. Kubota, K. Yagi, Y. Sano, H. Hara, K. Arima, T. Okamoto, H. Mimura, K. Yamauchi, “Chemical planarization of GaN using hydroxyl radicals generated on a catalyst plate in H2O2 solution”, J. Cryst. Growth, Vol. 310, 1637-1641, 2008. https://doi.org/10.1016/j.jcrysgro.2007.11.093
  3. J. H. An, G. S. Lee, W. J. Lee, B. C. Shin, J. D. Seo, K. R. Ku, H. D. Seo, H. D. Jeong, “Effect of process parameters on material removal rate in chemical mechanical polishing of 6H–SiC(0001)”, Mater. Sci. Forum, Vol. 600-603, 831-834, 2009. https://doi.org/10.4028/www.scientific.net/MSF.600-603.831
  4. C. J. Park, H. J. Kim, S. J. Lee H. D. Jeong, “The influence of abrasive size on high-pressure chemical mechanical polishing of sapphire wafer”, International journal of Precision Engineering and Manufacturing-Green Technology, Vol. 2, No. 2, 157-162, 2015. https://doi.org/10.1007/s40684-015-0020-0
  5. Darren DeNardisa, Yoshiyuki Seikeb, Mineo Takaokab, Keiji Miyachib, Ara Philipossiana, “Investigation of high-pressure micro jet technology as an alternative to diamond disc conditioning in ILD CMP”, Wear, Vol. 260, Issues 11-12, 30 June 2006, 1224-1231. https://doi.org/10.1016/j.wear.2005.08.002
  6. H. J. Kim, H. Y. Kim, H. D. Jeong, “Viscoelastic Behavior of Polishing Pad and Its Influence on Polishing Non-uniformity”, 5th CMP-MIC, pp. 275-282, 2000.
  7. H. Hocheng, Y. L Huang, L. J Chen, “Kinematic analysis and measurement of temperature rise on a pad in chemical mechanical planarization”, J. Electrochem. Soc., Vol. 146, No. 11, 4236-4239, 1999. https://doi.org/10.1149/1.1392620
  8. Jing Li, Yuhong Liu, Tongqing Wang, Xinchun Lu, Jianbin Luo, “Electrochemical investigation of copper passivation kinetics and its application to low-pressure CMP modeling,” Applied Surface Science, Vol. 265, 15 January 2013, 764-770. https://doi.org/10.1016/j.apsusc.2012.11.106
  9. G. Ahmadi, X. Xia, “A Model For Mechanical Wear and Abrasive Particle Adhesion during the Chemical Mechanical Polishing Process”, J. of The Electrochemical Society, Vol. 148, No. 3, G99-G109, 2001. https://doi.org/10.1149/1.1346614