Effects of hydroxyl and trifluoromethyl substituents on the corrosion inhibition of benzotriazole derivatives in copper surface planarization

Abstract

Benzotriazole derivatives are well-known corrosion inhibitors in chemical mechanical planarization (CMP) of integrated circuits, but the effect of their substituents on steric hindrance and hydrophilicity has been almost ignored. In this work, the effect of hydroxyl and trifluoromethyl substituents on the electronic structure, steric hindrance and hydrophilicity of benzotriazole derivatives is studied to evaluate their corrosion inhibition performance in the CMP of copper. The electron-inductive effect of the substituents (–OH and –CF₃) can alter the electron density of the adsorption site, thus affecting the adsorption affinity of corrosion inhibitors. Meanwhile, the steric hindrance and hydrophilicity of –OH might contribute to a looser passivation film, thus facilitating the corrosion processes. Although the steric hindrance of –CF₃ is greater than that of –OH, its hydrophobicity might provide a barrier for film passivation to limit corrosion processes. This work may inspire the development of next-generation corrosion inhibitors for CMP of integrated circuits.

---

Introduction

Chemical mechanical planarization (CMP), which involves complex interfacial physicochemical processes, is a key technology process for global surface flattening of ultra-large-scale integrated circuits.^1,2 So far, copper is still the dominant interconnect metal material in integrated circuits, so it is crucial to develop CMP slurries for copper.^3,4 Corrosion inhibitors in the CMP slurries act as a balance between chemical corrosion and mechanical wear, thus achieving sub-nm surface roughness.^5,6 Recently, significant progress has been made in the development of corrosion inhibitors for copper CMP, driven by the growing demand for high-precision surface planarization in integrated circuits.^7–15 Benzotriazole (BTA) and its derivatives remain the most widely studied inhibitors due to their ability to form stable passivation films, effectively suppressing chemical corrosion during CMP processes.^16,17 Advanced formulations now integrate hybrid inhibitors combining multiple functional groups (e.g., –NH₂, –COOH, –CH₃, –OH) to enhance adsorption efficiency.^11,18,19 For example, X. Niu et al. reported the corrosion inhibition performance of 5-methyl-benzotriazole (5CH₃-BTA) on copper film in the CMP of a large-scale integrated circuit.²⁰ J. Cheng et al. compared the corrosion inhibition mechanism of 5CH₃-BTA with BTA and 5-carboxybenzotriazole (5COOH-BTA) in the CMP of cobalt by combining theoretical calculations and experiments.⁶

Unfortunately, the vast majority of existing studies have focused only on the performance enhancement or the elaboration of corrosion inhibition mechanisms, especially the effect of substituent groups on the electronic structure of the active sites in corrosion inhibitors. The factors of steric hindrance and hydrophilicity of substituent groups, which have a crucial influence on the densification and permeability of passivation films in CMP, have been neglected. Considering these points, this study aims to investigate the effect of substituent groups (–OH and –CF₃) on the electron density, steric hindrance and hydrophilicity of benzotriazole derivatives to evaluate their corrosion inhibition performance in CMP.

Materials and methods

As shown in Fig. 1a, BTA, 1-hydroxybenzotriazole (n-OH-BTA), and 1-hydroxy-6-trifluoromethylbenzotriazole (n-OH-BTA-4CF3) are used as corrosion inhibitors. Density functional theory (DFT) computations are conducted using Gaussian 16 and VASP 6.3.0. Electrochemical tests are recorded on an electrochemical workstation (CHI660e). CMP experiments are performed on polishing apparatus (UNIPOL-1202). The detailed parameters can be found in the SI.

Fig. 1 (a) Molecular structures and (b) ESP of BTA, n-OH-BTA, and n-OH-BTA-4CF₃. (c) Schematic diagram of adsorption affinity differences of BTA, n-OH-BTA, and n-OH-BTA-4CF₃ on the copper surface.

Results and discussion

Generally, the corrosion inhibition capacity of BTA derivatives on copper mainly originates from the chemical bonding between the N atoms at their adsorption sites (N-SITE) and copper, especially through the formation of coordination bonds (σ bonds) by donating electrons.²¹ Therefore, the effect of substituents on the electronic structure of the derivatives is first analyzed. From Fig. 1b, the ESP shows that the negative charges of three corrosion inhibitors are concentrated on the same N atoms, which will absolutely serve as N-SITE. But both –OH and –CF₃ lead to a decrease in electronegativity of N-SITE for n-OH-BTA (−1.56 eV) and n-OH-BTA-4CF₃ (−1.34 eV), compared to BTA (−1.64 eV). As shown in Fig. 1c, –OH can decrease the electron density of N-SITE (N1/N2) in n-OH-BTA through the electron-withdrawing inductive effect, giving rise to a decay in adsorption affinity. The electron density of N-SITE in n-OH-BTA-4CF₃ will be further reduced because of the combined effect of –OH and –CF₃, and lead to a remarkable decrease in adsorption affinity.

Since the BTA backbone is a π-conjugated structure, the Hückel method can be used to quantify the intramolecular charge distribution. Fig. S1 shows that the N3 atoms have the highest negative charge, while the N1 atoms possess a positive charge, in line with the ESP results. Specifically, N3 atoms in BTA have more negative atomic charges than n-OH-BTA, followed by n-OH-BTA-4CF₃ (Fig. S2). As a result, N-SITE in BTA should have the best nucleophilicity to form coordination bonds with copper, followed by n-OH-BTA-4CF₃, and n-OH-BTA is the worst. Based on DFT, the adsorption energy and Cu–N bond lengths of different inhibitors on Cu₂O have been obtained (Fig. S3 and Fig. 1c). Notably, Cu₂O is utilized because it is extremely easy to form a Cu₂O film on the copper surface during CMP.²¹ In detail, the Cu–N bond length (1.905 Å) between n-OH-BTA and Cu₂O is larger than that of BTA (1.888 Å), demonstrating that the coordination bond between n-OH-BTA and Cu₂O is weaker than that of BTA. For n-OH-BTA-4CF₃, the Cu–N bond length further increases to 1.963 Å, indicating the weakest coordination bond with Cu₂O. It is worth noting that the adsorption energy of the corrosion inhibitor on Cu₂O does not seem to coincide with the Cu–N bond length results because the role of hydrogen bonds is taken into account (Fig. S3). Nevertheless, hydrogen bonds are difficult to form in the actual CMP due to the protonation of the acidic polishing slurry we used; therefore, the adsorption energy cannot be used to judge the performance of the corrosion inhibitor. Overall, the corrosion inhibition performance of the three corrosion inhibitors is in the order of BTA > n-OH-BTA > n-OH-BTA-4CF₃.

Then, the corrosion performance of different BTA-derivatives is verified by electrochemical tests. Electrochemical impedance spectroscopy (EIS) and Tafel plots are performed in CMP slurries that contain different corrosion inhibitors with the same concentration (0.001 moL L⁻¹). Prior to electrochemical tests, all copper wafers are treated with acetic acid for 5 min to remove the surface oxides. Fig. 2a shows the Tafel plots of copper in different acidic slurries, by which the corresponding corrosion current (I_corr) can be obtained (Fig. 2b and Table S1). Obviously, the I_corr value sharply decreases once corrosion inhibitors are employed, contributing to the formation of an inhibitor passivation film. Specifically, BTA exhibits the smallest I_corr value (2.073 × 10⁻⁶ A), followed by n-OH-BTA-4CF₃ (9.523 × 10⁻⁶ A) and n-OH-BTA (16.63 × 10⁻⁶ A). According to eqn. (S1) in the SI, the corresponding corrosion inhibition efficiency (η_Tafel) can be calculated and is illustrated in Fig. 2b and Table S1. Unsurprisingly, BTA exhibits the highest η_Tafel value (95.1%), due to its strong Cu–N bonding as mentioned above. Interestingly, the η_Tafel value (77.5%) of n-OH-BTA-4CF₃ is higher than that of n-OH-BTA (60.6%), which is inconsistent with the results of DFT. This will be discussed below.

Fig. 2 (a) Tafel plots of copper in different slurries. (b) The corresponding corrosion current and corrosion inhibition efficiency. W/O represents without corrosion inhibitor. (c) and (d) Nyquist plots of copper in different slurries. (e) Schematic diagram of the copper/slurries interface and equivalent circuit diagram. (f) The calculated R_pf and R_ct values.

Furthermore, Fig. 2c and d depict the Nyquist plots of copper in various slurries. The capacitive resistance arc in the high-frequency region corresponds to the passivation film resistance (R_pf), whereas the one in the low-frequency region is attributed to the charge transfer resistance (R_ct) at the passivation film/slurry interface. The equivalent circuit diagram for the Nyquist plots is illustrated (Fig. 2e), associated with the schematic diagram of copper/slurries, where R_s is the internal resistance, and C_dl and CPE are the interfacial bilayer capacitance and the constant phase angle element associated with the passivation film, respectively. Fitting by equivalent circuit, the R_pf and R_ct can be obtained (Table S1). Compared to the slurry without corrosion inhibitor, all other samples manifest an improvement in both R_pf and R_ct (Fig. 2f). Meanwhile, n-OH-BTA shows the smallest resistance (including R_pf and R_ct), compared to BTA and n-OH-BTA-4CF₃, implying that the n-OH-BTA passivation film is inferior to n-OH-BTA-4CF₃. According to eqn (S2) in the SI, the corrosion inhibition efficiency (η_EIS) of BTA, n-OH-BTA, and n-OH-BTA-4CF₃ is calculated to be 93.6%, 60.3%, and 68.6%, respectively (Fig. 3a and Table S1), which is consistent with η_Tafel, but still not in agreement with the prediction of DFT.

Fig. 3 (a) The corrosion inhibition efficiency (η_EIS). The proposed mechanism model of the formed passivation film for (b) BTA, (c) n-OH-BTA, and (d) n-OH-BTA-4CF₃.

Therefore, besides the electronic structure and bonding strength of N-SITE, some other factors should be taken into account. In fact, the bonding ability of a single corrosion inhibitor molecule with a metal is difficult to truly reflect its actual performance in CMP. The corrosion inhibition capability is determined by the passivation film composed of corrosion inhibitor molecules. In the passivation film, in addition to the affinity of the corrosion inhibitor to the substrate, the arrangement and density of the corrosion inhibitor directly affects its ability to resist external corrosion. At the same time, it is also necessary to consider the influence of the external environment, such as the acidity and alkalinity of the CMP polishing slurries. Thus, the steric hindrance and hydrophilicity should be considered, both of which can affect the arrangement, density, and thicknesses of passivation films and thereby alter the interfacial corrosion kinetics. It is well-known that –OH is hydrophilic due to its strong hydrogen bonding ability, whereas –CF₃ is highly hydrophobic due to its low polarity, low polarizability, and weak hydrogen bonding ability. The effect of –OH and –CF₃ groups on the hydrophilicity of derivatives can be demonstrated by the lipid–water partition coefficient (Table S2). Meanwhile, the van der Waals volume of –OH (approx. 12–15 ų) is significantly smaller than that of –CF₃ (approx. 40–50 ų), and thus –CF₃ will introduce a non-negligible steric hindrance within the passivation film. The proposed mechanism model of the formed passivation film is illustrated in Fig. 3b and c. For BTA, due to its inherent high hydrophobicity, a dense passivation film might be formed (Fig. 3b) in the water-based CMP slurry, leading to the highest R_ct and R_pf. For n-OH-BTA, the steric hindrance and hydrophilicity of –OH could result in a looser passivation film with protonation (Fig. 3c), facilitating mass transfer and corrosion processes. Although the steric hindrance of n-OH-BTA-4CF₃ is greater than that of n-OH-BTA, the hydrophobic –CF₃ might provide a barrier for film passivation to limit mass transfer and corrosion processes (Fig. 3d). This is why n-OH-BTA-4CF₃ shows better corrosion inhibition efficiency than n-OH-BTA in electrochemical tests and is different from the prediction of DFT.

Next, the ability of the different corrosion inhibitors is further verified in an actual CMP. Generally, smaller surface roughness (S_a) can be achieved with superior-performance corrosion inhibitors. Fig. 4a–c shows the surface morphology of copper after CMP using slurries containing different BTA derivatives. Obvious local corrosion and pitting can be observed for n-OH-BTA and n-OH-BTA-4CF₃, but not BTA, attributed to the excellent passivation film of BTA. The S_a of copper treated with BTA-based slurry reaches 0.985 nm, followed by n-OH-BTA-4CF₃ (1.102 nm) and n-OH-BTA (1.371 nm). The passivation film formed by corrosion inhibitors on the copper surface can not only reduce copper's erosion, but also block the mechanical force wear in CMP. According to eqn (S1) in the SI, the material removal rate (MRR) can be obtained. Consequently, by virtue of the hydrophobic –CF₃, the MRR of n-OH-BTA-4CF₃ is 245 nm min⁻¹, which is higher than that of BTA (168 nm min⁻¹) due to steric hindrance, but significantly lower than that of n-OH-BTA (327 nm min⁻¹). Overall, the corrosion inhibition order in CMP is: BTA > n-OH-BTA-4CF₃ > n-OH-BTA, which is the result of the combined effect of substituent groups on the electron density, steric hindrance and hydrophilicity.

Fig. 4 Surface morphology of copper after CMP using slurries containing (a) BTA, (b) n-OH-BTA, and (c) n-OH-BTA-4CF₃. (d) Corresponding material removal rate of copper after CMP.

Conclusions

In this work, the effect of substituent groups on the electron density, steric hindrance and hydrophilicity of the BTA-derivatives is comprehensively considered to evaluate their corrosion inhibition performance. The corrosion inhibition ability of BTA on copper mainly comes from the passivation film formed by BTA adsorption, and the adsorption site is the N atom (N-SITE) on triazole. The electron-inductive effect of the substituent groups (–OH and –CF₃) can alter the electron density of N-SITE, thus affecting the adsorption affinity of corrosion inhibitors. Meanwhile, the steric hindrance and hydrophilicity of –OH result in a looser passivation film for n-OH-BTA, facilitating mass transfer and corrosion processes. Although the steric hindrance of n-OH-BTA-4CF₃ is greater than that of n-OH-BTA, the hydrophobic –CF₃ provides a barrier for the passivation film to limit mass transfer and corrosion processes. Eventually, the S_a of copper treated with n-OH-BTA-4CF₃-based slurry reaches 1.102 nm, which is worse than BTA (0.985 nm) due to steric hindrance, but better than n-OH-BTA (1.371 nm) thanks to the hydrophobic –CF₃. This work may inspire the design and development of high-performance corrosion inhibitors in the field of anti-corrosion.

Author contributions

Anguo Zhang and Xinyu Li: investigation, data curation, formal analysis, and writing – original draft. Chun Cao: supervision, conceptualization, funding acquisition, resources, writing – original draft, and writing – review & editing. Yuhan Chen: investigation, data curation, and formal analysis. Jianting Liu and Chunjing Shi: supervision and resources.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included in the paper and the supplementary information (SI) of this article. Supplementary information is available. See DOI: https://doi.org/10.1039/d5nj04126k.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22105180) and Start-up Research Fund of Hangzhou Dianzi University (KYS015624245 and KYS015625245).

Notes and references

1. J. Liu, X. Niu, Y. Jia, N. Zhan, Y. Zou, Y. Shi and J. Zhou, Tribol. Int., 2024, 198, 109832.
2. Q. Sun, D. Yang, T. Liu, J. Liu, S. Wang, S. Hu, S. Liu and Y. Song, Microsyst. Nanoeng., 2023, 9, 50.
3. C. He, J. Zhou, R. Zhou, C. Chen, S. Jing, K. Mu, Y.-T. Huang, C.-C. Chung, S.-J. Cherng, Y. Lu, K.-N. Tu and S.-P. Feng, Nat. Commun., 2024, 15, 7095.
4. M. Krishnan, J. W. Nalaskowski and L. M. Cook, Chem. Rev., 2010, 110, 178–204.
5. J. Liu, X. Niu, Y. Jia, N. Zhan, Y. Zou, Y. Shi and J. Zhou, Appl. Surf. Sci., 2024, 654, 159469.
6. J. Cheng, Y. Lv, F. Zhang, P. Han, Q. Miao and Z. Huang, Appl. Surf. Sci., 2025, 682, 161684.
7. J. Zhou, X. Niu, Y. Cui, Z. Wang, J. Wang and R. Wang, Appl. Surf. Sci., 2020, 505, 144507.
8. J. Zhou, X. Niu, Z. Wang, Y. Cui, J. Wang, C. Yang, Z. Huo and R. Wang, Colloids Surf., A, 2020, 586, 124293.
9. G. Yang, H. Wang, N. Wang, R. Sun and C.-P. Wong, J. Alloys Compd., 2019, 770, 175–182.
10. H. Yan, X. Niu, M. Qu, F. Luo, N. Zhan, J. Liu and Y. Zou, Int. J. Adv. Des. Manuf. Technol., 2023, 125, 47–71.
11. X. Wang, W. Li, B. Tan, F. Wang, H. Du, R. Liu, X. Han and S. Zhang, J. Mol. Liq., 2023, 390, 122985.
12. T. T. H. Tran, J. Gichovi, J. Commane, E. J. Podlaha and J. Seo, J. Environ. Chem. Eng., 2024, 12, 113669.
13. T. T. H. Tran, E. J. Podlaha and J. Seo, J. Environ. Chem. Eng., 2025, 13, 115134.
14. J. Seo, S. Ryu, J. Kwon and K. Lee, Appl. Surf. Sci., 2024, 663, 160149.
15. R. Liu, X. Han, B. Tan, W. Li, F. Wang, X. Wang, J. Zhao and X. Zhao, Colloids Surf., A, 2024, 698, 134624.
16. T. Ma, B. Tan, Y. Xu, D. Yin, G. Liu, N. Zeng, G. Song, Z. Kao and Y. Liu, Colloids Surf., A, 2020, 599, 124872.
17. A. Kokalj, S. Peljhan, M. Finšgar and I. Milošev, J. Am. Chem. Soc., 2010, 132, 16657–16668.
18. D. Liu, Z. Zhang, H. Zhou, X. Deng, C. Shi, F. Meng, Z. Yu and J. Feng, Appl. Surf. Sci., 2023, 640, 158382.
19. Z. Huang, P. Chang, Y. Chen, N. Lv, M. Li and T. Hang, Mater. Lett., 2024, 366, 136530.
20. H. Yan, X. Niu, F. Luo, M. Qu, N. Zhan, J. Liu and Y. Zou, ECS J. Solid State Sci. Technol., 2023, 12, 044007.
21. B.-J. Cho, S. Shima, S. Hamada and J.-G. Park, Appl. Surf. Sci., 2016, 384, 505–510.

---