Mechanism and Application of Imidazolium Derivatives as a Bifunctional Additive for Void-free Blind via Copper Filling

Abstract

In conventional acidic copper electroplating, the accelerator and leveler are typically provided by different molecules, resulting in complex additive formulations and difficult-to-control intermolecular interactions. This study proposes an “intramolecular functional group division” strategy and designs a series of imidazolium salts bearing different substituents as bifunctional additives, integrating the leveling function of the imidazolium cation and the accelerating function of the pendant groups within a single molecule. Electrochemical measurements, electroplating of silicon-based micro blind vias, and multi-scale characterization demonstrate that a simplified binary system consisting solely of this additive and PEG8000 can achieve void-free superconformal filling. Systematic comparison of four imidazole derivatives containing sulfonate, cyano, and carboxyl substituents establishes a clear structure–activity relationship: the sulfonate group is the key moiety that confers the accelerating effect, Cl^– bridging is a prerequisite for strong adsorption, and the structural basis for acceleration is that the sulfonate group remains exposed on the outer surface without participating in copper-surface adsorption. DFT calculations further elucidate this mechanism from the perspectives of electronic structure and adsorption configuration. This work elevates the integration of dual functions in copper electroplating additives from intermolecular synergy to the single-molecule level, providing a new paradigm for the design of multifunctional electroplating additives.

1. Introduction

The advancement of 5G technology and artificial intelligence (AI) has driven increasingly stringent demands on semiconductor packaging technology, where copper electroplating filling of silicon blind vias serves as a core process in semiconductor chip manufacturing and advanced packaging. , With the continuous scaling of chip technology nodes, the shrinking feature size of blind via structures imposes ever-harsher requirements on the process efficiency of copper electroplating filling. , In additive-free electrolytes, non-uniform current distribution typically results in the “dog-bone effect”, leading to incomplete filling and defect formation. The incorporation of functional additives can effectively mitigate these issues and significantly improve the quality of copper deposition. Consequently, the development of highly efficient and broadly applicable additives has become a topic of considerable scientific and technological importance.

Commonly employed composite additive systems consist of inhibitors, accelerators, and levelers. Inhibitors, such as polyethylene glycol (PEG), adsorb onto the copper surface to form a polymeric film that inhibits the deposition of copper ions. Accelerators, such as bis­(3-sulfopropyl) disulfide (SPS) or 3-mercapto-1-propanesulfonate (MPS), compete with inhibitors for surface adsorption sites through the synergistic interaction between thiol and sulfonate groups, thereby enhancing the deposition rate of copper ions. Levelers, which are predominantly composed of quaternary ammonium salts, preferentially adsorb onto regions of high electrical potential on the copper surface, thereby improving surface flatness. − The additives exhibit synergistic effects, and by tailoring their concentration ratios, super-filling of blind vias with varying feature dimensions can be achieved.

To improve the efficiency and quality of copper electroplating, various additives have been developed. Examples of levelers include pyridine-containing compounds, certain dyes (e.g., rhodamine B), quaternary ammonium salts, phosphonium chlorides, and imidazole derivatives. , Regarding accelerators, examples include thioether-containing sulfonate compounds, disulfide-based sulfonates, and dithiocarbamoyl-functionalized sulfonates. These additives are often required to be used in combination with other inhibitor/accelerator or inhibitor/leveler systems. Compared with single-function additives, multifunctional additives offer advantages such as simplified bath maintenance and reduced operational cost. However, studies on such additives remain relatively limited. Available reports indicate that sodium thiazolinyl-dithiopropanesulfonate (SH110) and 2-mercapto-1-methylimidazole (MMI) can be employed as multifunctional additives. Therefore, the development of novel multifunctional additives, together with a deeper understanding of their underlying mechanisms, remains of significant scientific and practical interest.

Previous studies have confirmed the application potential of imidazole-based compounds in this field. , However, imidazole-based compounds have been predominantly utilized as leveler molecules; only MMI has been demonstrated to function as a multifunctional additive. Systematic investigations into the effects of substituent groups and electronic properties of imidazole compounds on their adsorption behavior, electronic structure, and the regulatory mechanisms governing the acceleration/leveling balance during copper electroplating have yet to be reported.

Conventional imidazole-based levelers rely on weak N–Cu coordination to achieve inhibition. Although MMI can serve as a single additive for void-free filling, its bifunctionality originates from concentration-dependent behavior and thiol-copper chemisorption, representing a rather specific mechanism. Quaternary ammonium levelers suppress copper deposition in high-current-density regions via electrostatic adsorption, but their function is limited to leveling. In SPS, the sulfonate group can capture Cu²⁺ and transfer it to Cl^– sites, enabling deposition acceleration through inner-sphere electron transfer; however, its accelerating function depends on thiol anchoring. To overcome these limitations, this work proposes a design strategy of “intramolecular functional group division”: the imidazolium cation serves as the anchoring group to provide leveling functionality, while a sulfonate group is introduced through side-chain modification as the accelerating unit, endowing a single molecule with both functions. Based on this strategy, we have constructed a simplified binary electroplating system consisting of PEG8000 and the bifunctional additive [SPMIm]­Cl, achieving defect-free filling of silicon blind vias without the need for conventional accelerators or levelers.

To validate this strategy and establish structure-activity relationships, this study selects four imidazole derivatives bearing different substituents (sulfonic acid, cyano, and carboxyl groups) as novel bifunctional additives and systematically investigates their performance in copper electroplating. Through electrochemical measurements, theoretical calculations, spectroscopic analyses, and silicon-based micro-blind via filling experiments, combined with multi-scale characterization (XPS/SEM/LSCM) and DFT calculations, we comprehensively examine the effects of substituents and counterions on adsorption behavior, leveling/acceleration performance, and filling outcomes. Theoretical calculations reveal that, under Cl^– bridging, the imidazolium ring of [SPMIm]Cl firmly adsorbs onto the copper surface to exert the leveling effect, while the sulfonate group remains exposed outward and facilitates Cu²⁺ reduction, thereby elucidating the structural origin of the dual functions at the molecular level. This strategy provides clear theoretical and experimental guidance for the design of next-generation high-performance, low-toxicity electroplating additives.

2. Methods

2.1. Chemicals

All chemicals including polyethylene glycol 8000 (PEG8000), SPS, Janus green B (JGB), 1-methyl-3-(3-sulfopropyl)-1H-imidazole-3-ium chloride ([SPMIm]­Cl), 1-carboxyethyl-3-methylimidazolium chloride ([CEMIm]­Cl), 1-(3-cyanopropyl)-3-methylimidazolium chloride ([CPMIm]­Cl), and 1-methyl-3-(3-sulfopropyl)-1H-imidazole-3-ium ([SPMIm]⁺) were used as received without further purification. PEG8000, SPS and JGB were purchased from Aladdin Reagent Co., Ltd. [SPMIm]­Cl, [CEMIm]­Cl, [CPMIm]Cl and [SPMIm]⁺ were obtained from Lanzhou GreenChem ILs. All copper sulfate electroplating solutions were prepared in-house by our research team.

2.2. Electrochemical Measurements

All electrochemical measurements were performed on a CHI 660E electrochemical workstation coupled with a standard three-electrode system at a constant temperature of 25 °C. A platinum rotating disk electrode (Pt-RDE, PINE Research, 5 mm in diameter) was employed as the working electrode, a platinum rod as the counter electrode, and a saturated silver–silver chloride electrode (Ag/AgCl) as the reference electrode. The electrochemical cell was loaded with 50 mL of electroplating solution. The virgin makeup solution (VMS) contained 156 g·L^–1 CuSO₄·5H₂O, 10 g·L^–1 H₂SO₄, and 20 ppm of Cl^– (HCl). When conducting experiments with [SPMIm]⁺, VMS did not contain HCl to investigate the effect of chloride ions on additives. Cyclic voltammetry (CV) was performed at a scan rate of 50 mV·s^–1 over a potential range of −0.6 V to 1.4 V. The Pt-RDE was rotated at 200 rpm during the measurement. Before each experiment, the Pt-RDE was soaked in dilute nitric acid for 5 min and thoroughly rinsed with deionized water to remove surface contaminants and ensure a consistent electrode surface condition.

The stripping charge density (Q) was calculated by Q = Q _total/A. Specifically, Q _total was determined by integrating the area under the anodic stripping peak using CHI 660E software after subtracting the capacitive charging contribution with a linear baseline connecting the two peak end points. The integration limits were consistently set from −0.4 to 1.6 V. The electrochemical active surface area (A) of the electrode was 0.0707 cm².

In this work, SPS was used as the accelerator, PEG8000 as the inhibitor, JGB as the leveler. The four imidazolium-based compounds, namely [SPMIm]­Cl, [CEMIm]­Cl, [CPMIm]­Cl, [SPMIm]⁺, were evaluated as bifunctional additives. The concentrations of additives in the plating bath were as follows: SPS at 1 ppm, JGB at 5 ppm, PEG8000 at 300 ppm, and imidazole additive at 10 ppm.

2.3. Electroplating

The silicon blind vias used for electroplating had a diameter of 10 μm and a depth of 30 μm. Electroplating experiments were performed using VMS mixed with different types of additives. For [SPMIm]⁺, VMS did not contain HCl. During the electroplating process, the electroplating sample served as the cathode, and a copper plate was used as the anode. Prior to the experiments, the surface of the sample was cleaned sequentially with ethanol and ultrapure water, followed by vacuum treatment. Direct current (DC) electroplating was conducted at a constant current density of 0.5 A·dm^–2 for 1 h.

2.4. Characterization

To evaluate the throwing power of the additives, the cross sections of the silicon blind vias were observed using an optical microscope (Carl Zeiss Microscopy GmbH). The surface morphology of the copper deposits was characterized by scanning electron microscopy (SEM, Ciqtek-SEM5000X). The crystal structure of the as-deposited copper layer was analyzed via X-ray diffraction (XRD, Bruker-D8 ADVANCE). The chemical states of the elements on the coating surface were identified using X-ray photoelectron spectroscopy (XPS, Thermo Fisher-K-Alpha). The surface roughness of the copper deposits was measured by laser scanning confocal microscopy (LSCM, Olympus-LEXT OLS5100).

2.5. Theoretical Calculations

A series of theoretical calculations was conducted on the four compounds using the Gaussian 16 W program with the B3LYP method at the 6-31G basis set level, based on DFT. Subsequent calculations at the B3LYP/6-31G level yielded the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energies for each compound. The energy gap, designated as ΔE, was defined as the difference between the LUMO and HOMO energies (i.e., ΔE = E _LUMO – E _HOMO). The molecular orbitals and the electrostatic potential (ESP) were calculated, analyzed, and visualized by using Multiwfn and the VMD program package. ,

Density functional theory (DFT) as implemented in the Vienna Ab-initio Simulation Package (VASP) with the Perdew–Burke–Ernzerhof (PBE) functional and the projected augmented wave (PAW) method. − A plane-wave cutoff of 450 eV was employed. Convergence criteria were set to 10^–5 eV for electronic energy and 0.04 eV/Å for forces. van der Waals interactions were treated with the DFT + D3 correction. The Cu (200) surface was modeled as a 5 × 5 supercell (108 Cu atoms, a = b = 18.07 Å) with a 20 Å vacuum layer. The Brillouin zone was sampled using a 2 × 2 × 1 Monkhorst–Pack k-point mesh. Adsorption energies were calculated as

$$E_{\mathrm{a}\mathrm{d}\mathrm{s}}=E_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}-E_{\mathrm{C}\mathrm{u}}-E_{\mathrm{a}\mathrm{d}\mathrm{d}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}$$

3. Results and Discussion

3.1. Analysis of Imidazole Molecular Properties

The four imidazolium-based compounds investigated in this study share a common 1-methylimidazole backbone but differ significantly in their terminal substituents and counterion characteristics, leading to distinct physicochemical properties and electrochemical behaviors during copper electrodeposition. Specifically, [SPMIm]Cl and [SPMIm]⁺ contain a sulfonate functional group, [CEMIm]Cl incorporates a carboxyl group, while [CPMIm]Cl features a cyano group (Scheme ). These substituents play a decisive role in modulating molecular polarity, electron distribution, and interfacial adsorption capability. Notably, the key distinction between [SPMIm]Cl and [SPMIm]⁺ lies in the presence or absence of the counter anion (Cl^–), which may further affect electrostatic interactions and adsorption characteristics at the electrode/electrolyte interface. A systematic comparison of these compounds can elucidate the effects of substituent types, molecular polarity, coordination ability, and counterion effects on their properties in silicon-based micro blind hole copper plating processes.

1. Structure of Imidazole Additives.

3.2. Electrochemical Analysis

CV was employed to evaluate the influence of different additives on the electrochemical behavior of the plating bath. Figure systematically illustrates the electrochemical effects of imidazolium-based additives on copper electrodeposition from multiple perspectives. , A mixed solution of VMS and PEG8000 (300 ppm) was used as the blank sample, and different additives were separately introduced for comparative experiments. As shown in Figure a, the electrolyte containing [SPMIm]Cl exhibits the most positive shift in cathodic peak potential, accompanied by a significant increase in cathodic current density and integrated peak area. In contrast, [CPMIm]­Cl, [CEMIm]­Cl, and [SPMIm]⁺ show relatively weaker enhancements (For [SPMIm]⁺, VMS does not contain HCl), implying limited ability to facilitate charge transfer and deposition kinetics. Meanwhile, slight differences can also be observed in the charge density, with detailed results presented in Figure b. The highest charge density obtained with [SPMIm]­Cl confirms its superior ability to promote copper deposition. These phenomena may be attributed to the interaction of the sulfonate groups in [SPMIm]Cl with copper ions in the solution, while the imidazole ring adsorbs onto the copper surface, thereby facilitating the reduction of Cu²⁺. Moreover, chloride ions are essential for this acceleration process, as a comparison between [SPMIm]Cl and SPMIm⁺ indicates that Cl^– can interact synergistically with both the additive and the copper surface to enhance charge transfer kinetics. Overall, combined analysis of CV curves and integrated charge densities demonstrates that [SPMIm]­Cl most effectively promotes copper electrodeposition, highlighting its potential as a high-performance accelerator for electroplating systems.

1.

(a) CV curves at a scan rate of 50 mV·s^–1 with different additives; (b) calculated stripping charge density under different additive conditions; (c) CV curves at a scan rate of 50 mV·s^–1 with different concentrations of [SPMIm]­Cl; (d) calculated stripping charge density under different concentrations of [SPMIm]­Cl; (e) CV curves at a scan rate of 50 mV·s^–1 with different rotational speed; (f) calculated stripping charge density under different rotational speed.

Figure c illustrates the CV curves of the electrolyte containing PEG8000 (300 ppm) with varying concentrations of [SPMIm]­Cl. With increasing [SPMIm]Cl concentration from 1 to 10 ppm, the cathodic peak current gradually increases and the corresponding peak area becomes larger, confirming the enhanced catalytic activity and deposition efficiency. However, when the concentration is further increased to 20 ppm, the cathodic peak current and peak area show a slight decrease compared with those at 10 ppm, indicating a decline in the acceleration effect (Figure d). This indicates that [SPMIm]Cl exhibits an optimal concentration window (around 10 ppm) for maximizing its accelerating effect, beyond which its performance deteriorates due to adsorption saturation. Therefore, the concentration of 10 ppm represents an optimal balance between acceleration and surface coverage, thereby achieving the most effective copper electrodeposition. Figure e presents the CV curves obtained at different rotation rates in the electrolyte containing PEG8000 (300 ppm) and [SPMIm]Cl (10 ppm). As the rotation rate increases from 200 to 1500 rpm, the cathodic peak current gradually rises. This behavior can be attributed to improved mass transport under higher rotation speeds. To quantitatively evaluate this effect, the integrated cathodic charge densities are summarized in Figure f. In summary, these results demonstrate that [SPMIm]Cl functions as an effective accelerator under appropriate conditions.

3.3. Copper Electroplating Experiments and Characterization

The four imidazolium-based molecules were evaluated in copper electroplating trials, and the corresponding cross-sectional morphologies are presented in Figure . In the absence of any additives, the deposited copper exhibits severe defects, including a large central void and poor bottom-up filling (Figure a). The voids were slightly alleviated after adding 300 ppm of the inhibitor PEG8000 (Figure b), this improvement can be attributed to the adsorption of PEG8000 on the copper surface, forming a suppressive polymer layer that inhibits excessive deposition at high-current-density regions.

2.

Cross sectional image of micropores filled with copper electroplating solutions containing different additives at a current of 0.5 A·dm^–2 for 1 h: (a) VMS; (b) PEG8000 (300 ppm); (c) PEG8000 (300 ppm) + [SPMIm]Cl (10 ppm); (d) PEG8000 (300 ppm) + [CEMIm]­Cl (10 ppm); (e) PEG8000 (300 ppm) + [CPMIm]Cl (10 ppm); (f) PEG8000 (300 ppm) + [SPMIm]⁺ (10 ppm, without HCl); (g) PEG8000 (300 ppm) + SPS (1 ppm) + JGB (5 ppm); (h) VMS + [SPMIm]Cl (10 ppm). Cross-sectional images of microvias filled with Cu electroplating solution with [SPMIm]Cl for different concentrations: (i) [SPMIm]­Cl (5 ppm); (j) [SPMIm]Cl (20 ppm). Cross-sectional images of microvias filled with Cu electroplating solution with [SPMIm]Cl for different current densities: (k) 0.1 A·dm^–2; (l) 1.0 A·dm^–2.

The four imidazole-based compounds were separately introduced into a VMS containing 300 ppm PEG (without HCl added in the case of [SPMIm]⁺). The corresponding cross-sectional images are presented in Figure c–f. A notable improvement in copper electroplating performance was observed in all cases. Specifically, the copper deposit obtained with [CEMIm]­Cl exhibited a relatively rough surface (Figure d); when [CPMIm]Cl was used, the filling of blind vias is achieved, but defects still exist at the opening of the blind vias (Figure e); when using [SPMIm]⁺, the formation of voids was observed inside the blind vias (Figure f). These observations indicate that the leveling capability of [CEMIm]Cl is relatively poor, while the acceleration effects of [CPMIm]Cl and [SPMIm]⁺ are inadequate, which may be attributed to the absence of chloride ions and the weak interaction between the cyano group and copper ions. Consequently, [SPMIm]Cl demonstrated the optimal filling performance (Figure c), which is consistent with the conclusions drawn from the electrochemical measurements. Figure g shows the cross-sectional morphology of the micro-blind vias obtained using SPS as the accelerator and JGB as the leveler in the presence of PEG8000. As expected, this well-established additive system enables typical bottom-up super-filling behavior. Importantly, the filling morphology obtained with the [SPMIm]­Cl-containing system is comparable to that achieved with this additive system. However, when only 10 ppm [SPMIm]Cl was added without PEG8000 (Figure h), severe voids and non-uniform deposition reappeared. This indicates that the synergistic interaction between additives and inhibitors is crucial for achieving super-filling.

The concentration of [SPMIm]Cl also affects the morphology of the cross-section. At a reduced concentration of 5 ppm (Figure i), the via filling is incomplete, and small voids are observed within the interior. This behavior suggests that the concentration of [SPMIm]Cl is too low to provide an effective acceleration effect at the via bottom, leading to limited deposition rate and the formation of defects. In contrast, when the concentration is increased to 20 ppm (Figure j), the small-scale voids appeared again in the blind via. This phenomenon can be attributed to excessive adsorption of [SPMIm]­Cl molecules on the copper surface, forming a dense and relatively stable adsorption layer that inhibits Cu²⁺ reduction. Therefore, an intermediate concentration of 10 ppm (Figure c) is required to achieve an optimal balance between sufficient acceleration at the via bottom and moderate surface coverage, thereby enabling efficient bottom-up filling without defects. When the current density decreases to 0.1 A·dm^–2 (Figure k), the vias are not fully filled after 1 h of electroplating, and a noticeable unfilled region remains in the via interior, which indicates that the rate of electrochemical reduction is limited. In contrast, when the current density is increased to 1.0 A·dm^–2 (Figure l), the filling quality deteriorates, and distinct voids reappear inside the vias, accompanied by a rough and uneven surface morphology. This behavior can be attributed to intensified concentration polarization and accelerated consumption of Cu²⁺ ions at the electrode surface under high current density. As a result, copper deposition becomes preferentially localized near the via opening, while the interior region suffers from insufficient ion supply, ultimately leading to defect formation. A moderate current density (0.5 A·dm^–2, Figure c) provides an optimal balance between deposition rate and mass transport, enabling uniform bottom-up filling of micro-blind vias.

As characterized by SEM, electroplating performed with a bath containing only the VMS and the inhibitor PEG8000 resulted in a relatively rough surface of the as-obtained copper layer (Figure a). Upon separate addition of the additive molecules [SPMIm]­Cl, [CEMIm]­Cl, [CPMIm]Cl and [SPMIm]⁺ into the plating bath, the surface flatness of the copper layers was improved in all experimental groups (Figure b–e). This observation indicates that the imidazolium moiety in these additive molecules exerts a certain level of leveling effect. Among all the additives tested, [SPMIm]Cl afforded the optimal copper surface quality (Figure b). The surface morphology achieved with the [SPMIm]­Cl-containing plating bath was comparable to that obtained from the plating system formulated with the SPS and the JGB (Figure f). This result demonstrates that [SPMIm]­Cl can not only accelerate the copper deposition process inside blind vias, but also effectively suppress the “tip effect”, thereby enabling the fabrication of electroplated copper layers with excellent surface morphology.

3.

SEM images of electroplating solution containing different additives: (a) PEG8000 (300 ppm); (b) PEG8000 (300 ppm) + [SPMIm]Cl (10 ppm); (c) PEG8000 (300 ppm) + [CEMIm]Cl (10 ppm); (d) PEG8000 (300 ppm) + [CPMIm]Cl (10 ppm); (e) PEG8000 (300 ppm) + [SPMIm]⁺ (10 ppm, without HCl); (f) PEG8000 (300 ppm) + SPS (1 ppm) + JGB (5 ppm).

To further characterize the roughness of the surface copper, LSCM was employed (Figure ). The surface roughness results are summarized in Table . For the sample obtained from the VMS (Figure a), the arithmetic mean height (Sa) was 0.205 μm, the root mean square height (Sq) was 0.262 μm, the maximum peak height (Sp) was 1.505 μm, the maximum pit depth (Sv) was 1.593 μm, and the developed interfacial area ratio (Sdr) was 22.985%. After the addition of the inhibitor PEG8000 and the additive [SPMIm]Cl (Figure b), the surface flatness was significantly improved (Sa: 0.044 vs 0.205; Sq: 0.059 vs 0.262; Sp: 0.427 vs 1.505; Sv: 0.892 vs 1.593; Sdr: 0.711 vs 22.985). This demonstrates that the [SPMIm]­Cl additive exerts a leveling effect on the copper surface and reduces the surface roughness. This pronounced improvement can be attributed to the synergistic effect between PEG8000 and [SPMIm]­Cl. PEG8000 acts as an inhibitor that preferentially suppresses copper deposition, while [SPMIm]Cl further enhances this selective adsorption behavior through its imidazolium structure and functional groups. As a result, localized overgrowth at surface asperities is effectively suppressed, while deposition in recessed regions is relatively promoted, leading to a more uniform surface copper layer.

4.

Characterization of LSCM: (a) VMS; (b) PEG8000 (300 ppm) + [SPMIm]­Cl (10 ppm).

1. Summary of Roughness Results.

Entry	Sa (μm)	Sq (μm)	Sp (μm)	Sv (μm)	Sdr (%)
a	0.205	0.262	1.505	1.593	22.985
b	0.044	0.059	0.427	0.892	0.711

3.4. Structural Analysis

To reveal the influence of [SPMIm]Cl on the preferred crystallographic orientation of electrodeposited copper, XRD characterization was carried out to identify the dominant crystal planes, as displayed in Figure a. For the VMS, the (220) and (200) planes served as the main preferred orientations. In the plating bath consisting of PEG8000, SPS, and JGB, the (200) plane was dominant. When [SPMIm]Cl was used as a bifunctional additive, the (200) plane consistently exhibited a preferred orientation (texture coefficients are provided in Table S8 in the Supporting Information). Although crystallographic texture can be influenced by multiple factorsincluding additive type, current density, and film thickness − the present study was designed so that only the additive type was varied, while all other parameters were kept strictly constant. Under these controlled conditions, the texture shifted from a mixed (220)/(200) pattern to a pronounced (200) preferred orientation, indicating that the additive molecules selectively adsorb on different crystal planes, thereby regulating interfacial growth kinetics and promoting (200)-oriented deposition. However, it should be noted that the present DFT calculations were performed only on the Cu (200) surface; therefore, a direct assignment of inhibition effects on other planes, such as the (220) facet, would require further computational and experimental investigation. Nevertheless, previous studies have shown that a strong (220) reflection is generally associated with rapid localized growth and increased surface roughness; hence, the observed texture change indirectly supports the surface planarization function of [SPMIm]­Cl.

5.

(a) XRD patterns of copper electroplating layer formed from a Cu electroplating solution containing different additives; (b) XPS spectra, (c) high resolution C 1s, (d) Cu 2p_3/2, (e) O 1s and (f) N 1s peaks of [SPMIm]Cl adsorbed on copper surface.

The copper surface was characterized by XPS. The interfacial interactions between additive molecules and the copper substrate were systematically investigated by analyzing the evolution of chemical valence states of characteristic elements (Figure b–f). Characteristic signals of C 1s, N 1s, O 1s, and Cu 2p_3/2 were clearly observed in the survey scan spectrum, which directly confirmed the effective adsorption of [SPMIm]Cl on the copper surface. , The high-resolution C 1s spectrum could be fitted into several characteristic peaks at 283.9 eV (Cu–C), 284.8 eV (C–H/C–C), 285.3 eV (C–N), 286.5 eV (C–N), and 288.6 eV (O–C–O), indicating the formation of a dense and well-structured organic adsorption layer of [SPMIm]Cl on the copper surface (Figure c). The Cu 2p_3/2 spectrum (Figure d) exhibits two main contributions at around 932.5 and 933.2 eV, which are assigned to Cu⁰/Cu⁺ and Cu²⁺ species, respectively, suggesting the presence of a thin surface oxide layer alongside metallic copper. The O 1s spectrum exhibited characteristic peaks at 530.5 eV (C–OH), 531.8 eV (O–Cu), and 533.1 eV (C–O–C), which were mainly derived from the adsorption of the inhibitor PEG8000 on the copper surface (Figure e). The N 1s spectrum could be fitted into two components at 399.3 eV (C–N) and 400 eV (Cu–N). The presence of the Cu–N characteristic peak directly verified the strong chemical interaction between [SPMIm]Cl and the copper surface (Figure f). XRD and XPS analyses revealed that [SPMIm]Cl effectively regulates the preferential growth of copper grains along the (200) crystal plane. Meanwhile, the additive molecules are stably adsorbed onto the copper surface via interfacial chemical interactions, thereby significantly improving the surface quality of the electroplated copper layer.

3.5. Theoretical Calculation Results and Discussion

DFT calculations were performed to elucidate the electronic properties of the four imidazolium-based additives. The calculated HOMO energies, LUMO energies, and HOMO–LUMO energy gaps (ΔE = E _LUMO – E _HOMO) are shown in Figure . The ΔE values follow the order: [SPMIm]Cl (5.864 eV) < [SPMIm]⁺ (5.870 eV) < [CPMIm]­Cl (5.999 eV) < [CEMIm]Cl (6.457 eV). A smaller ΔE value indicates stronger adsorption of the organic molecule onto the copper surface. The relatively small ΔE values observed for the sulfonate-containing compounds suggest high chemical reactivity and a pronounced tendency to engage in strong electronic interactions with the copper substrate, thereby facilitating robust adsorption. [CPMIm]Cl exhibits a moderate ΔE value, indicative of intermediate adsorption strength, whereas [CEMIm]­Cl displays the largest ΔE value, reflecting its relative inertness and weaker adsorption onto the copper surface. Previous studies have established a correlation between ΔE and the leveling capability of additives, with a smaller ΔE corresponding to enhanced leveling performance. , Consequently, the copper deposit obtained with [CEMIm]Cl exhibited a relatively rough surface.

6.

Distribution and orbital energy values of HOMO and LUMO for different imidazole molecules.

The ESP distributions reveal that the imidazole rings of all four additives are low electron density regions (Figure ), which are easily attracted to high current density sites on the copper surface by electrostatic interaction, thereby contributing to the leveling effect. Conversely, the high electron density regions are located on the sulfonate, carboxyl, and cyano groups. These electronegative moieties can interact with copper through coordination bonds (e.g., Cu–N or Cu–O), anchoring the additive molecules onto the electrode surface. , The combination of electrostatic attraction and coordination bonding enables a synergistic adsorption mechanism, which is essential for achieving gradient coverage and defect-free super-filling.

7.

ESP diagram of imidazolium compounds.

The adsorption configurations of the four additives on the Cu (200) surface are shown in Figure . All four species exhibit chemisorption, yet with distinct binding modes. [SPMIm]Cl adsorbs via Cl^–-bridged coordination with additional van der Waals and Cl^–-π contributions. [CEMIm]Cl interacts solely through a Cl-bridged structure, whereas [CPMIm]Cl adopts a dual-site mode involving direct N-coordination and Cl-bridging. [SPMIm]⁺ binds directly through bidentate sulfonate-oxygen coordination. All adsorption energies are negative (E _ads < 0), indicating thermodynamically favorable spontaneous adsorption. The adsorption strength follows [SPMIm]Cl > [SPMIm]⁺ > [CPMIm]Cl > [CEMIm]­Cl, consistent with the trend of ΔE values. Based on the adsorption configurations, it can be inferred that the sulfonate group of the [SPMIm]Cl molecule does not participate in the adsorption process onto the copper surface. Consequently, this sulfonate group remains available to interact with copper ions in the solution, thereby enhancing the copper deposition rate and facilitating the acceleration process. In contrast, both the cyano group of [CPMIm]Cl and the sulfonate group of [SPMIm]⁺ interact with the copper surface, leading to comparatively poor acceleration capabilities. Furthermore, the relatively low adsorption energy of [CEMIm]Cl results in poor leveling performance. These findings are fully consistent with the results obtained from the electroplating experiments.

8.

Molecular adsorption configuration and adsorption energy of: (a) [SPMIm]­Cl; (b) [CEMIm]­Cl; (c) [CPMIm]­Cl; (d) [SPMIm]⁺ on the Cu (200) surface.

The DFT calculations described above adopt a vacuum single-molecule simplification, without explicitly accounting for solvent, ionic, or electric field effects. Nevertheless, the calculated relative order of adsorption energies is qualitatively consistent with the experimental electrochemical behavior and filling performance, confirming that the model adequately captures the key chemical trends.

Based on the above DFT calculation results described above, a possible dual-function mechanism of [SPMIm]Cl is illustrated in Figure . In high current density regions, the imidazole group exhibits strong adsorption on the copper surface through Cl^–-π interactions and van der Waals forces, thereby suppressing copper deposition and predominantly acting as a leveler. At the bottom of the blind via, where the current density is relatively low, the interaction between the intramolecular sulfonic acid group and copper ions in the solution becomes dominant, accelerating copper deposition. In this case, the adsorption of the imidazole group on the copper surface is weak and no longer dominates the interfacial behavior, so the molecule as a whole mainly serves as an accelerator. Therefore, the proposed mechanism is reasonable for a dual-function additive that integrates both leveling and accelerating effects.

9.

Possible mechanisms of dual function effects.

4. Conclusion

This work proposes an “intramolecular functional group division” strategy that integrates the leveling function of the imidazolium cation and the accelerating function of the sulfonate group within a single molecule. Four imidazolium derivatives (sulfonate, cyano, carboxyl) were evaluated, among which [SPMIm]Cl exhibited the optimal bifunctional performance: a simplified system consisting of [SPMIm]­Cl and PEG8000 achieved void-free superconformal filling of silicon-based micro blind vias, yielding a dense and smooth deposit. XRD revealed a preferred (200) crystallographic orientation, and XPS confirmed additive–copper surface interactions. Systematic comparison established that the sulfonate group is the key accelerating moiety, Cl^– bridging is a prerequisite for strong adsorption, and the structural basis for the accelerating effect lies in the sulfonate group remaining exposed on the outer surface without participating in copper surface adsorption. DFT calculations showed that [SPMIm]­Cl possesses the smallest ΔE value (5.864 eV) and the highest adsorption energy; ESP and VASP confirmed that the imidazole ring directly adsorbs on the copper surface while the pendant sulfonate group facilitates Cu²⁺ reduction. This study elevates bifunctional integration from intermolecular synergy to the single-molecule level, provides a multiscale mechanistic understanding from filling performance to electronic/adsorption structure, and opens new avenues for next-generation multifunctional additives.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c06002.

• This Supporting Information presents systematic comparisons with recent literature in conceptual approach, methodology, research scope and mechanism interpretation, summaries of plating solution formulas, charge density data, chloride effects on electrochemistry and deposition, filling quality metrics, LSCM and roughness results, texture coefficient calculations, chloride tolerance tests and detailed DFT computational methods (PDF)

The authors declare no competing financial interest.