Aerosol and Air Quality Research, 17: 936–941, 2017 Copyright © Taiwan Association for Aerosol Research ISSN: 1680-8584 print / 2071-1409 online doi: 10.4209/aaqr.2016.10.0457 Technical Note

Influence of Fluoride Ions Contamination in Front Opening Unified Pod (FOUP) Generating Defective Bonding Pad Soon Seok Kwon1, Sung Min Hwang1, Hyoung Ryeun Kim1, Hee Chang Jang1, Jeong Hoon Hong1, Gil Joo Song1, Hyun Yul Park1, Minsoo Kim2, Youhwan Shin2, Jin Young Kim3, Tae Yong Noh4, Seoung-Kyo Yoo4* 1 Contamination Quality Engineering, Manufacturing and Technology Division, SK Hynix Semiconductor Inc., Icheon-si, Gyeonggi-do 467-701, Korea 2 Center for Urban Energy Research, Korea Institute of Science and Technology, Seongbuk-gu, Seoul 02792, Korea 3 Fuel Cell Research Center, Korea Institute of Science and Technology, Seongbuk-gu, Seoul 02792, Korea 4 WITHTECH INC., Yuseong-gu, Daejeon 34036, Korea ABSTRACT

We analyzed defective bonding pad in various ways and determined the causes of defects that boosts oxidation of aluminium by fluoride residue on surface of pad with moisture. Additionally, we compared and evaluated methods to minimize pad defects in aspects such as etching and wafer storage environment. In case of wafers after pad open etching process using common CF4 stored in FOUP, the concentration of fluoride ions in FOUP was 230 ng L–1 and it decreased down to 170 ng L–1 when Ar sputtering step was added after using CF4. Also under the same condition, fluoride ion concentration in FOUP decreased down to 20 ng L–1 when nitrogen purge was introduced for 10 minutes to the FOUP where wafers were stored and the moisture also decreased from 40% before purge to 10% after purge. As a result of observation on pad surface after storing wafers in FOUP for 120 hours under each condition, negligible amount of defects were found when nitrogen was purged. Therefore, we conclude that defects on pad were generated by existing fluoride ions after etching process and moisture in the air. Keywords: Bonding pad; Fluoride ion; Etching process; Contamination; Front opening unified pod (FOUP). INTRODUCTION

According to miniaturization, speed acceleration and mass storage in semiconductor, not only manufacturing process to form chip on wafer but also packaging process to install the completed chip on circuit is continuously being developed (Mönch et al., 2011; Nakamura et al., 2013; Joyce et al., 2015). There is not a single part unimportant in packaging process, but interconnecting technology connecting chip and package which enables to interact electrically and structurally is considered as most the important. Among these techniques, the wire bonding technology is a technique which connects aluminium pad of chip with lead frame of package by gold wire and is already a very commonly used technology. But recently defects in this wire bonding are increasing (Tan et al., 2002; Goh et al., 2013). As well known, many kinds of * Corresponding author.

Tel.: 82-42-936-7117; Fax: 82-42-936-7228 E-mail address: skyoo@withtech.co.kr

chemicals are used and various kinds of particles form in the semiconductor manufacturing process and airborne molecular contaminations (AMCs) from these chemicals. Particles with moisture in the air not only contaminate wafer itself but also have effect on wire bonding which can generate product defects (Kikyuama et al., 1991; Den et al., 2006; Lin et al., 2010; Lee et al., 2015). To prevent this kind of contamination by exposure to the air, wafers on which chip is formed are stored or moved in the enclosed carrier called front opening unified pod (FOUP) and particles or AMCs remaining on wafers may diffuse into FOUP and contaminate it (Frickinger et al., 2000; Hu et al., 2005; Hu et al., 2009). Particles or AMCs adsorb onto inside of FOUP which can contaminate other wafers and FOUPs like a “contagious disease” and therefore various researches on FOUP cleaning and purge are in progress (Hu and Tsao, 2006; Hu et al., 2007; Yoo et al., 2012).

The aluminium pad of chip for wire bonding is a part that is exposed to the outside after wafer process, which can be continuously affected by the surrounding atmosphere until it is insulated from packaging. Especially, it can be contaminated by molecular contamination due to silicon

Kwon et al., Aerosol and Air Quality Research, 17: 936–941, 2017 937

dust from die sawing process which cuts off chips from silicon wafer, oxidization on pad surface by atmospheric moisture, and chemical contamination by silicon wafer protective layer or residual chemicals of etchant (Corum et al., 1993; Hsu, 2001). Contamination by etchant is highly likely due to the residual fluoric etchant solution or gas remaining after the etch process (Hua et al., 2014; Song et al., 2015). Contamination on pad surface can cause abnormal failure having various defects on the surface in black colour and Non-Stick-on-Pad (NSOP) failure having wire ball to be detached after wire bonding, and these failures can increase contact resistance with probe used in chip inspection which causes damage on probe pin or degrade electrical feature of the chip (Kim et al., 2003; Dan et al., 2015).

In this research, pads failed in actual manufacturing process were collected and inspected. The surfaces and composition were checked to figure out the causes of contamination. The correlation between pad contamination effect by open pad etch process condition and AMCs according to inner environment of FOUP is reported.

METHODS

Chips with defects, found in the inspection in actual

manufacturing process were collected to analyse the images of surfaces, cross sections, and substances according to the thickness of thin layers. The surfaces of normal and abnormal pad for visual inspection were observed by Microscope (CX-21, Olympus), and images of surfaces and cross sections were obtained by scanning electron microscopy with Energy Dispersive x-ray Spectroscopy (SEM-EDX, S-5200, Hitachi) and transmission electron microscopy (TEM, CM200, Philips), respectively. The substance analysis according to the thickness of thin layer was conducted by Auger Electron Spectroscopy (AES, PHI-670, Perkin Elmer).

The etching condition for pad open process was divided into as-is using carbon tetrafluoride (CF4) condition and introducing an argon sputtering step after using CF4 by pad etch facility (Producer, Applied Materials, Inc.) for investigating an etchant influence. To investigate the cause of defect on bonding pad surface, different etching process for pad opening was applied. For the first nine wafers, typical CF4 etching process was applied. Then three from these nine wafers were selected and each one was stored in three

different pre-cleaned FOUPs. The other rest six wafers had additional argon sputtering process and each one was stored in six different pre-cleaned FOUPs. And finally, nitrogen purge was applied to only the three FOUPs from the latter six FOUPs had both CF4 etching and argon sputtering. As for nitrogen purge, it remained for 10 minutes with 10 mL min–1 of flow rate and the measured moisture concentration inside of FOUP during purge was measured by a humidity sensor (Datalogger 177-H1, Testo). After storing each wafer in FOUPs for 2 hours under each condition, fluoride ion concentration inside of three FOUPs at each condition was measured and averaged by online AMCs monitoring system (proFast-2000, WITHTECH Inc.) and three pad surfaces at each condition were analysed by EDX for comparison. Also, defect occurrence on the pad surface stored in FOUP for 2 hours and the pad surface stored in FOUP for 120 hours under each condition was observed by microscopically.

RESULTS AND DISCUSSION

When an abnormally failed pad from pad open etching process was collected and inspected visually, the normal pad surface was bright but many defects with black colour were observed on the surface of abnormal pad as shown in Fig. 1. The results of observation on abnormal pad surfaces by SEM-EDX and TEM were shown in Fig. 2. From the result of SEM-EDX observation, defects on pad surfaces were in bump shape and Al, O and F elements were detected by EDX from these bumps with different contents. Also, as a result of diffraction pattern of defects part by TEM, it was found that defects have crystal structures, not amorphous. Fig. 3 shows the analytic result of substance according to the depth profile on the surface represented as sputtering time using AES, which shows the fact that F and O elements were located in deeper position on abnormal pad than normal pad. The depth profile of AES is calculated by sputter time and the thickness of Rutherford backscattering. From this method, 1 minute of sputter time is 100 Å of depth profile. Fig. 3(a) shows normal pad has fluoride ion from pad surface to 40 Å and Fig. 3(b) shows abnormal pad had fluoride ion to deeper area as 330 Å. When aluminium oxide layer (Al2O3) of each normal pad and abnormal pad were analyzed by SEM-EDX, it was found that the thicker the Al2O3 layer is, the higher the concentration of fluoride

(a) (b) (c)

Fig. 1. Microscopic images of pad surfaces of normal (a) and abnormal (b), and bonding pads (c).

Kwon et al., Aerosol and Air Quality Research, 17: 936–941, 2017 938

(a)

(b)

Fig. 2. SEM-EDX images and element mapping of abnormal pad surface (a) and TEM images with diffraction pattern of crystalline defect of abnormal pad (b).

(a) (b) Fig. 3. AES depth profile for component analysis on aluminium, fluorine and oxygen elements of normal (a) and abnormal (b) pads. ion on pad surface is as shown in Fig. 4. The defects on pad surface were not observed if the thickness of Al2O3 layer is under 80 Å and the fluoride ion concentration on the surface was 4%. On the other hand, defects were observed when the thickness of Al2O3 layer is over 180 Å with the higher the concentration of fluoride ion on pad surface. From this, it is regarded that Al2O3 layer abnormally becomes thicker due to presence of fluoride ion. Alberici et al. (2003) reports that the normal pad surface has Al2O3, fluorine used in pad etching process and aluminium from pad combines to form AlF3, and reaction with moisture in exposed environment to

outside develops crystalized compounds, AlFx.H2O, Al(OH)x, Al(OF)x, and AlF(OH)2. Chen et al. (2001) also reports that the humidity with fluoride ion was effective in acceleration of the growth of crystal defects. From these results and reports, it is considered that fluoride ion boosts oxidization of aluminium with moisture in the air and this growth forms a thick oxidization layer.

To investigate the influence of fluoride ion on this abnormal growth of Al2O3 layer, wafers after pad open etching process were stored in FOUP and fluoride ion substance inside of FOUP and on the pad surfaces was

Al O F

Kwon et al., Aerosol and Air Quality Research, 17: 936–941, 2017 939

Fig. 4. Correlation between fluoride ion concentration on pad surfaces and thickness of Al2O3 layers.

analysed. Also to split fluoride ion concentration, argon sputtering step (CF4 + Ar) other than common CF4 etching process (CF4) and nitrogen purge step (CF4 + Ar + N2) into FOUP were introduced. Table 1 shows result from each online AMCs monitoring system and EDX measured fluoride ion inside of FOUP and on pad surfaces under each condition. The fluoride ion concentration when using as-is CF4 was 230 ng L–1, decreased down to 170 ng L–1 upon introduction of argon sputtering step and to ng L–1 upon nitrogen purge. Fluoride ion concentration on pad surfaces also decreased down to each 23%, 11% and 4%. From this result, it was found that the fluoride ion concentration adsorbed onto wafer can be reduced by argon sputtering and also fluoride ion contamination diffused into FOUP from wafer can be reduced by nitrogen purge. This result supports the theoretical background of Fick’s 2nd law shown as Eq. (1), which insists that nitrogen purge can decrease fluoride ion concentration in FOUP and relatively higher fluoride ion concentration on wafer is diffused into FOUP to remove fluoride ion on wafer.

0 expa

desE dCR DkT dx

(1)

Rdes: Desorption rate [moles cm–2 s–1], D0: Rate constant, Ea: Desorption Enthalpy (Activation Energy), T: Temperature, dC/dx: Concentration gradient, (dC/dx): Concentration gradient between wafer and surround ↑, Removal rate ↑.

Furthermore, changes in moisture concentration were

observed during nitrogen purge to investigate the moisture effect when pad is exposed. As a result, moisture concentration was approximately 40% initially and decreased down to 10% level within 10 minutes after nitrogen purge, as shown in Fig. 5. Fig. 6 is image from observation of defect occurrence on pad surfaces after storing wafers in FOUP for each 2 hours and 120 hours under each condition. Under two conditions, a case used as-is CF4 and the other case added argon sputtering step, defects occurred on pad surfaces with slight differences even after storing 120 hours and in case nitrogen purge was introduced, there were barely any defects after 120 hours of storage. This can be interpreted that defect could occur due to existing fluoride ions on pad surfaces (4%) even after introducing nitrogen purge, but nitrogen purge could additionally remove moisture and restrain defect occurrence. From these results, it was found that the introduction of nitrogen purge can restrain defect occurrence on pad surfaces by removing over 90% of fluoride ion and it is necessary to remove not only fluoride ion but also moisture to prevent defect occurrence. CONCLUSIONS

We investigated the cause of defects on aluminium pad which often occurs in packaging process. Firstly, pad on which defects occurred was analyzed and found that the thickness of Al2O3 layer had abnormally increased, which is because of Al2O3 layer growing into an abnormal crystal structure due to fluoride ions existing on the surface and in the layer and moisture in the air. After pad open etching using common CF4, wafers are stored in FOUP and fluoride ions which is one of chemicals used in etching adsorbed onto wafer diffused into FOUP at 230 ng L–1. When argon

Table 1. Fluoride ion concentration in FOUP and pad surfaces depending on each process of the etch and purge.

Condition Pad Open Etch and N2 Purge Process CH4 CH4 + Ar CH4 + Ar + N2

F Conc. in FOUP (ng L–1) 230 170 20 F Conc. on Pad (Atomic %) 23 11 4

Kwon et al., Aerosol and Air Quality Research, 17: 936–941, 2017 940

Fig. 5. Changes of relative humidity in the air inside of FOUP depending on the time N2 purged into FOUP.

Fig. 6. Microscopic images of bonding pad surfaces splitted by storage time in the FOUP as 2 and 120 hours, depending on each process of CH4 etching (a), CH4 etching with Ar sputtering (b) and N2 purge after CH4 etching with Ar sputtering (c). sputtering step was added to as-is pad open etching process using only CF4, fluoride ion concentration in FOUP decreased down to 170 ng L–1 and that on the pad surface decreased from 25% to 10%. After storing wafers through each of two kinds of etching conditions for 120 hours in FOUP, plenty of defects were observed on pad surfaces under both conditions. When nitrogen purge into FOUP in which stores wafers after argon sputtering step adding to etching using CF4 was introduced for 10 minutes, fluoride ion concentration in FOUP decreased down to 20 ng L–1 and so did moisture down to 10% level each, and defects barely occurred on pad surfaces after storing for 120 hours. From this, it was understood that when wafers are stored in FOUP after pad open etching process, fluoride ions, the one of chemicals used in etching adsorbed onto wafer

diffuses into FOUP to exist at certain concentration, and this fluoride ion in FOUP causes defects on pad surfaces with moisture. It was discovered that, to prevent defect occurrence on aluminium pad surface which is only part exposed to the air until packaging process is completed after wafer process, it is necessary to manage fluoride ion and moisture content upon storing or moving wafers. ACKNOWLEDGMENTS

This research was financially supported by the Ministry of Trade, Industry and Energy (MOTIE) and Korea Institute for Advancement of Technology (KIAT) through the Promoting Regional specialized Industry Program under grants no. R0004804.

Kwon et al., Aerosol and Air Quality Research, 17: 936–941, 2017 941

REFERENCES Alberici, S., Coulon, D., Joubin, P., Mignot, Y., Oggioni,

L., Petruzza, P. and Piumi, D. (2003). Surface treatment of wire bonding metal pads. Microelectron. Eng. 70: 558–565.

Chen, J.S., Wei, L.K., Chang, Y.P. and Huang, C.C. (2001). Bond pad F-crystal defect control and monitoring. Semiconductor Manufacturing Symposium 2001, IEEE International, pp. 297-299.

Corum, D., Kim, S.Y. and Wills, K.S. (1993). Failures mechanism in integrated circuits. In Microelectronic Desk Reference (1st edition), pp. 278-300.

Den, W., Bai, H.L. and Kang, Y.H. (2006). Organic airborne molecular contamination in semiconductor fabrication clean rooms. J. Electrochem. Soc. 153: 149–159.

Frickinger, J., Bugler, J., Zielonka, G., Pfitzner, L., Ryssel, H., Hollemann, S. and Schneider, H. (2000). Reducing airborne molecular contamination by efficient purging of FOUPs for 300-mm wafers– The influence of materials properties. IEEE Trans. Semicond. Manuf. 13: 427–435.

Goh, C.S., Chong, W.L.E., Lee, T.K. and Breach, C. (2013). Corrosion study and Intermetallic formation in gold and copper wire bonding in microelectronics packaging. Crystals 3: 391-404.

Hsu, C.F. (2001). A study of the volatile defect formation and the polysilicon residue elimination. Proc. of SPIE 4405, Process and Equipment Control in Microelectronic Manufacturing II, pp. 131–141.

Hu, S.C. and Hsiao, T.R. (2005). Particle dynamics in a front-opening unified pod/load port unit minienvironment in the presence of a 300 mm wafer in various positions. Aerosol Sci. Technol. 39: 185–195.

Hu, S.C. and Tsao, J.M. (2006). Purging of front-opening unified pod with nitrogen for 300mm wafer manufacturing. Jpn. J. Appl. Phys. 45: 5269–5271.

Hu, S.C., Wub, T.M., Lina, H.C. and Hsu, K. (2007). Design and evaluation of a nitrogen purge system for the front opening unified pod (FOUP). Microelectron. Eng. 27: 1386–1393.

Hu, S.C., Ku, C.W., Shih, Y.C. and Hsu, K. (2009). Dynamic analysis on particle concentration induced by opening the door of a front opening unified pod (FOUP) that loaded with 25 pieces of 300 mm Wafer manufacturing processes. Aerosol Air Qual. Res. 9: 139–148.

Hua, Y., Xing, Z. and Li, X. (2014). Characterization studies of fluorine-induced corrosion crystal defects on microchip Al bondpads using X-ray photoelectron spectroscopy. 2014 IEEE 21st International Symposium on the Physical and Failure Analysis of Integrated

Circuits (IPFA), pp. 90–93. Joyce, R., Singh, K., Varghese, S. and Akhtar J. (2015).

Effective cleaning process and its influence on surface roughness in anodic bonding for semiconductor device packaging. Mater. Sci. Semicond. Process. 31: 84-93.

Kikyuama, H., Miki, N., Saka, K., Takano, J., Kawanabe, I., Miyashita, M. and Ohmi, T. (1991). Principles of wet chemical processing in ULSI microfabrication. IEEE Trans. Semicond. Manuf. 4: 26–35.

Kim, H.J., Lee, J.Y., Paik, K.W., Koh, K.W., Won, J.H., Choi, S.H., Lee, J., Moon, J.T. and Park, Y.J. (2003). Effects of Cu/AI intermetallic compound (IMC) on copper wire and aluminium pad bondability. IEEE Trans. Compon. Packag. Technol. 26: 367–374.

Lee, M. and Yook S.J. (2015). Investigation of particulate contamination of heated wafers contained in a closed environment. J. Aerosol Sci. 88: 148–158.

Lin, I.K., Bai, H. and Wu, B.J. (2010). Analysis of relationship between inorganic gases and fine particles in cleanroom environment. Aerosol Air Qual. Res. 10: 245–254.

Mönch, L., Fowler, J.W., Dauzère-Pérès, S., Mason, S.J. and Rose, O. (2011). A survey of problems, solution techniques, and future challenges in scheduling semiconductor manufacturing operations. J. Scheduling 14: 583–599.

Nakamura, Y. and Katogi, S. (2013). Technology trends and future history of semiconductor packaging substrate material. Hitachi Chemical Technology Report No. 55, pp. 24-29.

Posey, D.E. and Hossain, T.Z. (2015). Contaminant sources causing non-stick-on-pad during die bonding. Proc. SPIE 3215: 180–184.

Song, G.J., Hwang, S.M., Koo, S.J., Kim, H.R., Jang, H.C., Hong, J.H., Park, H.Y., Choi, E., Kim, J.Y., Noh, T.Y., Lee, E. and Yoo, S.K. (2015). Evaluation of hindrance to the growth of SiN passivation layer by contamination of fluoride ions in front opening unified pod (FOUP). Aerosol Air Qual. Res. 15: 2175–2183.

Tan, C.W., Daud, A.R. and Tarmo, M.A. (2002). Corrosion study at the Cu-Al interface in microelectronics packaging. Appl. Surf. Sci. 191: 67–73.

Yoo, S.K., Shiue, A., Lee, T.S., Hu, S.C., Lee, E. and Ju, Y. (2012). Monitoring and purge parameters of ammonia for the front opening unified pod (FOUP). Adv. Sci. Lett. 17: 280–284.

Received for review, October 24, 2016 Revised, December 26, 2016

Accepted, January 1, 2017