Processing math: 100%
J. Semicond. > 2015, Volume 36 > Issue 12 > 126002

SEMICONDUCTOR TECHNOLOGY

Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization

Shaohua Jia, Yuling Liu, Chenwei Wang and Chenqi Yan

+ Author Affiliations

 Corresponding author: Liu Yuling,Email:liuyl@jingling.com.cn

DOI: 10.1088/1674-4926/36/12/126002

PDF

Abstract: The article studied the electrochemical behavior of P2 alkaline polishing slurry. The main research is the changing discipline of Ecorr and Icorr in the Cu electrolyte at different concentrations of oxidant H2O2. It compares potentiodynamic polarization curves in different P2 slurries and analyzes the passivation function of H2O2 acting on controlling dishing. The result implies that the potential increases gradually and then levels off while the current density on the contrary decreases with the augment of H2O2 concentration. In addition, dishing declines with the increasing of H2O2 along with the optimization of planarization of the alkaline P2 slurry.

Key words: hydrogen peroxidepassivationdishingalkalineCu CMP

In the process of copper wiring CMP,on the basis of P1 (remove the copper film quickly) realizing the preliminary planarization,the main task of P2 (the elaborate polishing process of removing residual copper) is to remove the residual copper on the wafer surface completely and modify dishing to reduce the extension of it. P2 should guarantee that the polishing rate of the barrier material tends to be zero[1]. On one hand,a period of over polishing is essential in P2 in order to remove the residual copper completely[2]. Inescapably,dishing will extend in the process of over polishing. Therefore,dishing has become a worldwide problem that is difficult to solve as well as the key technology of the multi-layered copper routing. On the other hand,copper and the barrier layer have high selectivity,which means that the removal rate of copper is relatively higher than that of Ta/TaN. Distinction of the removal rate will also cause the dishing to extend. Dishing has many defects such as high resistance with growing RC delay and increasing losses of copper wire,which will lead to increasing thermal power consumption and serious electricity migration phenomenon. It can also affect the partial planarization of copper pattern wafer,which cannot meet the industrial requirements of CMP[3].

At present,almost all of the polishing slurry used on domestic devices of global planarization is imported. A new type of alkaline copper polishing slurry self-developed by the Microelectronics Technology and Material Research Institute in Hebei University of Technology is adopted. It has many advantages such as simple components and environmental protection. The oxidant of the slurry is H2O2. Being different from benzotriazole (BTA) in the acid slurry with side effects such as being different to clean and forming Cu-BTA film[4],H2O2 will not induce ionic soil as the most common oxidizing agent[5]. On one hand,H2O2 can make copper be oxidized into ionic to react with the FA/O chelating agent,which generate stable complex. The reaction can accelerate the removal of copper as well as the efficiency of planarization. On the other hand,oxidation film and hydroxide film generating in the H2O2 oxidation process can act as passivation film to protect the copper wiring in the recess area from being removed,which forms a characteristic self-passivation process. The passive function of H2O2 can reduce dishing in the process of P2 and optimize the effect of planarization[6]. This article uses the electrochemical method to investigate the passive character of H2O2 in the alkaline P2 slurry with different concentrations of H2O2. In addition,we apply the P2 slurry to the 12 inch copper pattern wafer and integrate its function of modifying dishing. The passivation effect of H2O2 affecting the planarization of 12 inch copper pattern wafer is examined.

Experimental materials: a new type of alkaline copper polishing slurry self-developed by the Microelectronics Technology and Material Research Institute in Hebei University of Technology; the FA/O chelating agent and nonionic surface active agent researched independently; oxidizing agent H2O2 whose mass fraction is 30%; add different concentrations of H2O2 to confect alkaline P2 slurries before using (PH of all P2 slurries range from 10.3 to 10.7); a 3 inch copper blanket applied to test DRR; a 12 inch copper blanket wafer applied to test MRR; and a 12 inch copper pattern wafer applied to research the planarization of P2 slurry.

Experimental equipment: the E460E polisher produced by France's Alpsitec company; X IC1000 (TM) A2 PAD DH2 24.7''ACAO; *G01 polishing pad; CHI600C electrochemical workstation; 300 step profiler of KLA Tencor/Amhios Technology company; XP-300 Surface Profiler analytical balance with 0.1 mg accuracy (Mettle Toledo AB204-N); and PHB-4 portable PH meter of the Shanghai INESA Scientific Instrument Company.

P2 slurry composition: abrasive SiO2 whose mass fraction is 40% and FA/O chelating agent and active agent compose P1 slurry. According to 1:6 diluted times dilute P1 slurry directly in order to unify the stock solution of P1 and P2. P2 slurry compound with adding different concentrations of H2O2 (5,10,15,20,25,30,35,40 ml/L).

Electrochemical electrolyte composition: FA/O chelating agent; different contents of H2O2 (5,10,15,20,25,30,35,40~ml/L).The process condition of CMP is shown in Table1.

Table  1.  The process conditions of CMP.
DownLoad: CSV  | Show Table

Using the CHI600E electrochemical workstation to observe and analyze the copper polarization curve,Ecorr and Icorr of P2 slurry with different concentrations of H2O2. Analyze the passivation effect of H2O2 on controlling the dishing after P2.

Amend the polishing pad for one minute before each P2 experiment. Use the same stock solution,according to 1:6 diluted times to dilute P1 slurry directly into P2 slurry. On the basis of realizing the preliminary planarization of P1 on the copper pattern wafer,use P2 slurry to polish the wafer and an XP 300 step profiler to test dishing.

The main reason that acid polishing slurry of copper CMP is used widely on the international is that copper will form the oxide and hydroxide,which is difficult to dissolve in the alkaline situation resulting in decrease of the copper removal rate. Besides these insoluble particles may cause scratches and defects[7]. However,acid polishing slurry has many disadvantages with the development of nodes to 45 nm and below. For example,it contains toxic increasing film BTA whose heavy mechanical action will destroy the multiple air brittle low k dielectric,which is influenced by technical parameters including polishing pressure,speed and pad selection[8, 9, 10, 11]. This article uses alkaline slurry developed independently and selects the self-developed complex agent to turn insoluble copper substance into soluble copper amine complexes which is extremely stable. Thus,the problem of insoluble copper formation in the alkaline circumstance is solved effectively[12, 13].

As an oxidant,H2O2 oxidizes the surface of copper into copper oxide and hydroxide. Then Cu2+ ionized out of Cu(OH)2 complex with the FA/O chelating agent quickly under the chemical mechanical impact and form a stable copper-amine complex,which improves the removal rate of copper[14]. Meanwhile,in order to solve the problem of dishing extending P2,the passivation effect of H2O2 on recession copper is more focused on. The oxide and hydroxide passivation film adhering to the surface of recession copper prevent recessed copper from dissolving and convex copper wiring has a higher kinetic energy and faster mass transfer. Cu2+ ionized out of the Cu(OH)2 complex can overcome the potential energy quickly and generate extremely stable resultant. Through this process,distinction of high and low removal rates becomes obvious and dishing decreases after P2,which optimizes the planarization effect. Meanwhile,in order to reduce the extension of dishing,P2 need that the removal rate of the barrier layer tend to zero,which lay a good foundation for barrier layer polishing.

Different concentrations of H2O2 alkaline electrolytes are made up. Analyze the corrosion state of copper (Figure2) by testing the potential polarization curve. The corrosion statue includes the corrosion potential (Ecorr) and corrosion current (Icorr). The changing rule of them in different concentrations of H2O2 is shown in Figure1.

Figure  Fig1.  Ecorr and Icorr of copper wafer in P2 electrolyte of different H2O2 concentration.
Figure  Fig2.  (Color online) Potential polarization curve in P2 electrolyte of different H2O2 concentrations.

We test the electrochemical behavior of the copper electrode in different P2 electrolytes respectively; the testing results of polarization curves in different P2 electrolytes are shown in Figure2. Electrochemical anode and cathode electrode reactions: \text{ Anode:}\ \text{Cu}=\text{C}{{\text{u}}^{2+}}+2{{\text{e}}^{-}}. \text{Cathode:}\ {\rm O}_{2}+{\rm H}_{2}{\rm O}_{2}+2{\rm H}_{2}{\rm O}+6{\rm e}^{-}=6{\rm OH}^{-}.

((1))

Through analyzing the change of Ecorr and Icorr and the potential polarization curve,the testing results show the following conclusions.

When the concentration of H2O2 is 5 ml/L,Ecorr leaps to a relatively high value. With the increased concentration of H2O2,Ecorr continues to increase slowly and Icorr augments persistently. According to the data in Table2,when the concentration of H2O2 is 35 ml/L,the Ecorr of copper reaches the peak value and Icorr continues to decrease. The high Ecorr shows that the corrosion of copper is slight and the passivation is obvious at this content of P2 electrolyte. Increasing the H2O2 content to 40 ml/L,Ecorr changes to decrease and Icorr tends to increase.

Table  2.  Ecorr and Icorr of copper wafer in different H2O2 concentration P2 electrolytes.
DownLoad: CSV  | Show Table

The polarization curve in Figure2 shows that anode current decreases with ascending H2O2 content until the H2O2 content is 35 ml/L. When the concentration of H2O2 is lower,H2O2 firstly oxides copper into cuprous ions,which is oxidized into cupric ion in the electrolyte. On the contrary,H2O2 oxides copper into cupric ions directly. The change of color (changing from green to light blue) in the electrolyte can prove the explanation given above. The passive film attaching to the anode copper electrode prevent copper from further dissolution which increases passivation and descends the anode corrosion current[15]. At the same time,due to the incrassation of the passive film electron ionized from the anode becoming less,which is unable to provide an electronic supplying the cathode reaction (2) continue to happen. Hence the current density of the cathode also shrinks. Synthesizing the changing rule of anode and cathode,this is thought that diminution of the corrosion current can also explain that the passive effect of H2O2 on the copper film surface goes up with the augment of H2O2 concentration.

Therefore,the electrochemical behavior implies that the electrolyte has the strongest inhibitory effect on copper corrosion as well as the best passivation effect at H2O2 concentration of 35 ml/L. The results can also correspond to the change of the removal rate of copper during the process of P2.

Figure3 shows changes in anodic current over the corrosion time in P2 slurry of 35 mL/L H2O2. It can be confirmed that the dynamic changes in anode current is beneficial for analyzing the dynamic time-phased reaction occurring in P2 slurry. In the first part of the curve (0-700 s),the anode current descends relatively faster. At this stage,the passivation film that blocks available reaction sites protects recession copper from further reaction. Thus the anode current drops sharply. However,in the rest of curve (700-1200 s),the current persistently eases down. It is suggested that the oxide formed during the early stage of the dissolution is porous,which means continuous dissolution could occur[16]. Therefore,the passivation character of the H2O2 changes over time because of the porous oxidization film,which increases the reaction area.

Figure  Fig3.  Changes in anode current over corrosion time

Apply the changing rule of the anode current at H2O2 content of 35 ml/L to the P2 process. In the convex area,H2O2 passivation needs to be weak (700-1200 s),which can realize removing residual copper quickly. On the contrary,H2O2 passivation should be enhanced (0-700 s) to block the recession copper from dissolution to guarantee dishing is not extended.

On the basis of the results of the electrochemical analysis experiment of copper in different H2O2 content in the electrolyte,prepare the P2 slurry of different contents of 5,15,25,35,40 mL/L respectively. Use these slurries to polish (P2) the 12 inch copper pattern wafer for 30 s.

Choose five points A,B,C,D,E on the 12 inch pattern wafer diameter uniformly as dishing test positions. Figure4 shows the effect of various H2O2 contents of P2 slurry on dishing. P1 can realize planarization preliminarily and P2 can control dishing to the needs of industry requirements on the basis of removing the residual copper remaining in P1 completely. As it can be seen,when the H2O2 content is 5 mL/L and 15~mL/L,dishing extended to some extent with respect to it after P1. Dishing was amended with the increase of H2O2 content and it reaches the minimum at a certain H2O2 concentration of 3 mL/L. At this concentration,dishing was controlled to the maximum extent.

Figure  Fig4.  Effect of various H2O2 content of P2 slurry on dishing.

Figure5 shows the effect of various H2O2 contents of P2 slurry on the removal rate of Cu/Ta as well as the static corrosion rate of Cu. With the increase,the copper removal rate of P2 reduced gradually. When the H2O2 content increased to 15~mL/L,the copper removal rate of P2 reduced continuously but not obviously. Meanwhile,dishing was controlled to the maximum extent at an H2O2 concentration of 35 mL/L. It can be explained that the P1 realized planarization preliminarily and Cu removal rate of P2 just need to meet the standard (about 200 nm/min). When the H2O2 concentration is 35 mL/L,combined with the minimum static corrosion rate of copper further showing the reinforcement of the passivation of H2O2,a strong passivation effect of H2O2 can protect recession copper from being removed and the removal amount of convex copper almost had no difference. At the same time P2 can meet the standard that the barrier layer removal rate is almost zero.

Figure  Fig5.  Influence of various H2O2 content P2 slurry on the removal rate of Cu/Ta and the static corrosion rate of Cu.

The article researched many aspects such as the electrochemical performance of copper wafer in different H2O2 concentrations in alkaline P2 slurry,the effect of H2O2 passivation on dishing and also the removal rate of copper film. H2O2 not only can be used as an oxidant to accelerate the Cu removal rate it can also protect recession regions of copper wafer from being further dissolved due to dense oxide film formation. The changing rule of electrochemical corrosion potential reveals the rule of copper CMP in alkaline P2 slurry that corrosion potential ascends with the augment of the H2O2 content. The tendency revealed that the passivation effect of H2O2 is enhanced. Therefore,the passivation function of H2O2 plays an extremely critical role in P2 on controlling dishing from extending,which is also beneficial for global planarization.



[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Fig1.  Ecorr and Icorr of copper wafer in P2 electrolyte of different H2O2 concentration.

Fig2.  (Color online) Potential polarization curve in P2 electrolyte of different H2O2 concentrations.

Fig3.  Changes in anode current over corrosion time

Fig4.  Effect of various H2O2 content of P2 slurry on dishing.

Fig5.  Influence of various H2O2 content P2 slurry on the removal rate of Cu/Ta and the static corrosion rate of Cu.

Table 1.   The process conditions of CMP.

DownLoad: CSV

Table 2.   Ecorr and Icorr of copper wafer in different H2O2 concentration P2 electrolytes.

DownLoad: CSV
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
1

Effect of nitrogen incorporation and surface passivation on photoluminescence properties of InAs-based nanowires

Ratmir Ustimenko, Danila Karaulov, Maxim Vinnichenko, Ilya Norvatov, Andrey Kaveev, et al.

Journal of Semiconductors. doi: 10.1088/1674-4926/25030041

2

Enhanced efficiency of the Sb2Se3 thin-film solar cell by the anode passivation using an organic small molecular of TCTA

Yujie Hu, Zhixiang Chen, Yi Xiang, Chuanhui Cheng, Weifeng Liu, et al.

Journal of Semiconductors, 2023, 44(8): 082701. doi: 10.1088/1674-4926/44/8/082701

3

Investigation on the passivation, band alignment, gate charge, and mobility degradation of the Ge MOSFET with a GeOx/Al2O3 gate stack by ozone oxidation

Lixing Zhou, Jinjuan Xiang, Xiaolei Wang, Wenwu Wang

Journal of Semiconductors, 2022, 43(1): 013101. doi: 10.1088/1674-4926/43/1/013101

4

Surface passivation of perovskite film for efficient solar cells

Yang (Michael) Yang

Journal of Semiconductors, 2019, 40(4): 040204. doi: 10.1088/1674-4926/40/4/040204

5

Rational molecular passivation for high-performance perovskite light-emitting diodes

Jingbi You

Journal of Semiconductors, 2019, 40(4): 040203. doi: 10.1088/1674-4926/40/4/040203

6

The stability of a novel weakly alkaline slurry of copper interconnection CMP for GLSI

Caihong Yao, Chenwei Wang, Xinhuan Niu, Yan Wang, Shengjun Tian, et al.

Journal of Semiconductors, 2018, 39(2): 026002. doi: 10.1088/1674-4926/39/2/026002

7

Improved interfacial properties of GaAs MOS capacitor with NH3-plasma-treated ZnON as interfacial passivation layer

Jingkang Gong, Jingping Xu, Lu Liu, Hanhan Lu, Xiaoyu Liu, et al.

Journal of Semiconductors, 2017, 38(9): 094004. doi: 10.1088/1674-4926/38/9/094004

8

The effect of nitridation and sulfur passivation for In0.53Ga0.47As surfaces on their Al/Al2O3/InGaAs MOS capacitors properties

Zizeng Lin, Mingmin Cao, Shengkai Wang, Qi Li, Gongli Xiao, et al.

Journal of Semiconductors, 2016, 37(2): 026002. doi: 10.1088/1674-4926/37/2/026002

9

Characterization of the effects of nitrogen and hydrogen passivation on SiO2/4H-SiC interface by low temperature conductance measurements

Yiyu Wang, Zhaoyang Peng, Huajun Shen, Chengzhan Li, Jia Wu, et al.

Journal of Semiconductors, 2016, 37(2): 026001. doi: 10.1088/1674-4926/37/2/026001

10

Effect of FA/O complexing agents and H2O2 on chemical mechanical polishing of ruthenium in weakly alkaline slurry

Bo Duan, Weijing An, Jianwei Zhou, Shuai Wang

Journal of Semiconductors, 2015, 36(7): 076002. doi: 10.1088/1674-4926/36/7/076002

11

High performance AlGaN/GaN HEMTs with AlN/SiNx passivation

Xin Tan, Yuanjie Lü, Guodong Gu, Li Wang, Shaobo Dun, et al.

Journal of Semiconductors, 2015, 36(7): 074008. doi: 10.1088/1674-4926/36/7/074008

12

Alkaline barrier slurry applied in TSV chemical mechanical planarization

Suohui Ma, Shengli Wang, Yuling Liu, Chenwei Wang, Yan Yang, et al.

Journal of Semiconductors, 2014, 35(2): 026002. doi: 10.1088/1674-4926/35/2/026002

13

Effect of guanidine hydrochloride on removal rate selectivity and wafer topography modification in barrier CMP

Hailong Li, Jin Kang, Yuling Liu, Chenwei Wang, Hong Liu, et al.

Journal of Semiconductors, 2014, 35(3): 036002. doi: 10.1088/1674-4926/35/3/036002

14

Influence of atomic layer deposition Al2O3 nano-layer on the surface passivation of silicon solar cells

Decheng Yang, Fang Lang, Zhuo Xu, Jinchao Shi, Gaofei Li, et al.

Journal of Semiconductors, 2014, 35(5): 052002. doi: 10.1088/1674-4926/35/5/052002

15

Effects of substrate characteristics on the passivation performance of ALD-Al2O3 thin film for high-efficiency solar cells

Zongcun Liang, Dianlei Wang, Yanbin Zhu

Journal of Semiconductors, 2014, 35(5): 054002. doi: 10.1088/1674-4926/35/5/054002

16

Effect of chelating agent concentration in alkaline Cu CMP process under the condition of different applied pressures

Haobo Yuan, Yuling Liu, Mengting Jiang, Weijuan Liu, Guodong Chen, et al.

Journal of Semiconductors, 2014, 35(11): 116005. doi: 10.1088/1674-4926/35/11/116005

17

An advanced alkaline slurry for barrier chemical mechanical planarization on patterned wafers

Wang Chenwei, Liu Yuling, Niu Xinhuan, Tian Jianying, Gao Baohong, et al.

Journal of Semiconductors, 2012, 33(4): 046001. doi: 10.1088/1674-4926/33/4/046001

18

Optimization of a Cu CMP process modeling parameters of nanometer integrated circuits

Ruan Wenbiao, Chen Lan, Ma Tianyu, Fang Jingjing, Zhang He, et al.

Journal of Semiconductors, 2012, 33(8): 086001. doi: 10.1088/1674-4926/33/8/086001

19

Effects of the reciprocating parameters of the carrier on material removal rate and non-uniformity in CMP

Wang Cailing, Kang Renke, Jin Zhuji, Guo Dongming

Journal of Semiconductors, 2010, 31(12): 126001. doi: 10.1088/1674-4926/31/12/126001

20

Performance Improvement of AlGaN/GaN HEMTs by Surface Treatment Prior to Si3N4 Passivation

Li Chengzhan, Liu Dan, Zheng Yingkui, Liu Xinyu, Liu Jian, et al.

Journal of Semiconductors, 2008, 29(2): 329-333.

  • Search

    Advanced Search >>

    GET CITATION

    Shaohua Jia, Yuling Liu, Chenwei Wang, Chenqi Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. Journal of Semiconductors, 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002
    S H Jia, Y L Liu, C W Wang, C Q Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. J. Semicond., 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002.
    shu

    Export: BibTex EndNote

    Article Metrics

    Article views: 2832 Times PDF downloads: 11 Times Cited by: 0 Times

    History

    Received: 23 March 2015 Revised: Online: Published: 01 December 2015

    Catalog

      Email This Article

      User name:
      Email:*请输入正确邮箱
      Code:*验证码错误
      Shaohua Jia, Yuling Liu, Chenwei Wang, Chenqi Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. Journal of Semiconductors, 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002 ****S H Jia, Y L Liu, C W Wang, C Q Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. J. Semicond., 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002.
      Citation:
      Shaohua Jia, Yuling Liu, Chenwei Wang, Chenqi Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. Journal of Semiconductors, 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002 ****
      S H Jia, Y L Liu, C W Wang, C Q Yan. Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization[J]. J. Semicond., 2015, 36(12): 126002. doi: 10.1088/1674-4926/36/12/126002.

      Influence of oxidant passivation on controlling dishing in alkaline chemical mechanical planarization

      DOI: 10.1088/1674-4926/36/12/126002
      Funds:

      Project supported by the Special Project Items No. 2 in National Long-Term Technology Development Plan, China(No. 2009ZX02308).

      More Information
      • Corresponding author: Liu Yuling,Email:liuyl@jingling.com.cn
      • Received Date: 2015-03-23
      • Accepted Date: 2015-05-28
      • Published Date: 2015-01-25

      Catalog

        /

        DownLoad:  Full-Size Img  PowerPoint
        Return
        Return