A method for electroplating a high aspect ratio via in a multilayer circuit board

Abstract

This application provides an electroplating method for high aspect ratio vias on multilayer circuit boards, relating to the field of printed circuit board manufacturing technology. The method employs a three-stage gradient increasing pulse current waveform, dividing the electroplating process into a low current density penetration stage, a medium current density filling stage, and a high current density sealing stage, executed sequentially. The peak current density and duty cycle of each stage increase progressively. During stage switching, a precision metering pump adjusts the concentration ratio of leveling agent to accelerator in the electroplating solution in real time, causing the ratio to decrease progressively from a high value in the first stage to a low value in the third stage. During pulse shutdown, oscillating jet stirring forces liquid phase renewal in the vias. An auxiliary reference electrode array monitors the local overpotential at different depths of the vias in real time to adaptively determine the stage switching timing. This method solves the technical problem of copper layer uniformity deterioration in traditional electroplating processes under aspect ratios ≥10:1.

Description

Technical Field

[0001] This application relates to the field of printed circuit board manufacturing technology, and more specifically, to an electroplating method for high aspect ratio through holes in multilayer circuit boards. Background Technology

[0002] In the manufacturing process of multilayer printed circuit boards (PCBs), through-hole plating is a key step in achieving interlayer electrical interconnection. As electronic products become increasingly integrated with higher density, the number of layers and the thickness of multilayer PCBs continue to increase, while the diameter of through-holes is constantly shrinking due to wiring density constraints, leading to a continuous rise in the thickness-to-diameter ratio (the ratio of board thickness to hole diameter). When the thickness-to-diameter ratio reaches 10:1 or higher, traditional DC plating or simple forward and reverse pulse plating processes face two major coupling difficulties: limited mass transfer of the plating solution within the hole and uneven current density distribution. This results in excessive copper growth at the hole opening while insufficient copper layer is present in the hole center, leading to a sharp deterioration in copper layer uniformity and making it difficult to achieve the deep plating capability required for mass production. Summary of the Invention

[0003] This application provides an electroplating method for high aspect ratio through-holes on a multilayer circuit board. The method includes: preparing an acidic copper sulfate electroplating solution containing a leveling agent, an accelerator, and an inhibitor to a first-stage target concentration; loading the multilayer circuit board to be plated into an electroplating tank; performing the first-stage electroplating with a pulse current output at a first peak current density and a first duty cycle; performing jet stirring on the through-hole during the pulse off-peak period; simultaneously collecting local overpotentials at different depths of the through-hole using an auxiliary reference electrode array; determining the completion of the first-stage electroplating when the overpotential difference between the hole opening and the hole center continuously meets a first threshold condition; and adjusting the concentration of the leveling agent and the accelerator... The concentration ratio is adjusted from the first ratio to the second ratio, and the second stage of electroplating is performed with the second peak current density and the second duty cycle. When the maximum value of the overpotential difference between adjacent reference electrodes continuously meets the second threshold condition and the cumulative deposition thickness reaches a preset proportion of the target thickness, the second stage of electroplating is determined to be completed. The concentration ratio is then adjusted from the second ratio to the third ratio, and the third stage of electroplating is performed with the third peak current density and the third duty cycle. Electroplating is terminated when the cumulative deposition thickness reaches the target thickness. The peak current density and duty cycle increase step by step from the first to the third stage, and the concentration ratio decreases step by step from the first to the third stage.

[0004] Optionally, the first peak current density is 0.5 to 1.0 A / dm², the first duty cycle is 15% to 25%, the second peak current density is 1.5 to 2.5 A / dm², the second duty cycle is 35% to 45%, the third peak current density is 3.0 to 5.0 A / dm², and the third duty cycle is 55% to 65%.

[0005] Optionally, the first ratio is 1.5 to 3.0, the second ratio is 0.5 to 1.0, and the third ratio is 0.2 to 0.5.

[0006] Optionally, the leveling agent is a polyethyleneimine compound with a molecular weight of 10,000 to 30,000, the accelerator is sodium polydisulfide dipropane sulfonate, and the inhibitor is polyethylene glycol with a molecular weight of 4,000 to 8,000, wherein the inhibitor is maintained at a constant concentration throughout all three stages.

[0007] Optionally, in the step of adjusting the concentration ratio of leveling agent to accelerator in the electroplating solution from a first ratio to a second ratio: the actual concentration of the leveling agent is detected by a cyclic voltammetric stripping online analyzer; when the actual concentration is higher than the target concentration of the leveling agent in the second stage, the leveling agent supply pump is paused; when the actual concentration is lower than 90% of the target concentration of the leveling agent, the leveling agent supply pump is resumed; and the accelerator concentration is adjusted to the target concentration of the accelerator in the second stage according to the accelerator supply formula.

[0008] Optionally, according to the method of claim 1, the oscillating jet stirring is driven by a servo motor to deflect the jet nozzle array in a sinusoidal manner with a deflection amplitude of ±15°, the pressure and oscillation frequency of the jet stirring increase with each stage, and the jet stirring is triggered synchronously with the off-period of the pulse current.

[0009] Optionally, the first threshold condition is that the overpotential difference between the orifice and the center of the orifice does not exceed 15 mV for 30 consecutive pulse cycles, and the second threshold condition is that the maximum overpotential difference between adjacent reference electrodes does not exceed 10 mV for 50 consecutive pulse cycles.

[0010] Optionally, the cumulative deposition thickness is obtained by piecewise summation using Faraday's law.

[0011] Optionally, the first stage electroplating and the second stage electroplating are respectively provided with timeout protection. When the duration of the first stage electroplating exceeds the first maximum duration, the process is forcibly switched to the second stage electroplating. When the duration of the second stage electroplating exceeds the second maximum duration, the process is forcibly switched to the third stage electroplating.

[0012] Optionally, the method further includes post-processing and quality inspection steps: the multilayer circuit board that has been electroplated is removed after being rinsed with three-stage countercurrent water; metallographic cross-section is performed on the test through holes to measure the copper layer thickness at the hole opening, the copper layer thickness at the hole center, and the copper layer thickness at the hole bottom; the deep plating capability and uniformity deviation are calculated based on the copper layer thickness at the hole opening, the copper layer thickness at the hole center, and the copper layer thickness at the hole bottom; when the deep plating capability is not less than 80%, the uniformity deviation does not exceed 15%, and the minimum copper thickness is not less than the target thickness, it is judged as qualified.

[0013] Optionally, the auxiliary reference electrode array includes a plurality of miniature silver-silver chloride reference electrodes arranged at equal intervals along the thickness direction of the multilayer circuit board to be plated, wherein the effective sensing area of ​​each reference electrode does not exceed 1 square millimeter.

[0014] Optionally, the electroplating solution is formulated as follows: copper sulfate pentahydrate concentration of 200 to 220 g / L, sulfuric acid concentration of 50 to 60 g / L, chloride ion concentration of 40 to 60 ppm, and temperature stable at 22 to 24°C.

[0015] Optionally, the step of loading the multilayer circuit board to be plated into the electroplating tank further includes: starting the temperature control circulation system to stabilize the temperature of the electroplating solution within a set range, performing open-circuit potential calibration on the auxiliary reference electrode array, and aligning the jet nozzle array with the effective electroplating area of ​​the multilayer circuit board to be plated in a surface coverage mode.

[0016] The beneficial effects of this application are as follows: the method decouples the electric field distribution control, the timing of additive action, and the deep hole mass transfer supply in the three dimensions of time domain, chemical domain, and fluid domain by using a pulse current waveform with a three-order gradient increase, a phased dynamic control of the leveling agent and accelerator concentration ratio, and the synergistic cooperation of oscillating jet stirring. This results in improvements in three indicators: applicable thickness-to-diameter ratio, deep plating capability, and uniformity deviation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall process of the high aspect ratio through-hole electroplating method for multilayer circuit boards provided in the embodiments of this application.

[0018] Figure 2 This is a flowchart illustrating the system initialization and calibration phases provided in an embodiment of this application.

[0019] Figure 3 This is a schematic diagram of the first stage of the low current density penetration period provided in the embodiments of this application.

[0020] Figure 4 This is a flowchart illustrating the current density filling period in the second stage, as provided in an embodiment of this application.

[0021] Figure 5 This is a schematic diagram of the third-stage high current density sealing period provided in the embodiments of this application.

[0022] Figure 6 This is a flowchart illustrating the post-processing and quality inspection stages provided in an embodiment of this application.

[0023] Figure 7 This is a schematic diagram of the module architecture of the electroplating system provided in the embodiments of this application.

[0024] Figure 8 This is a time-series comparison diagram of the third-order gradient pulse waveform and the additive concentration ratio provided in the embodiments of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0027] This embodiment provides an electroplating method for high aspect ratio through-holes in multilayer circuit boards. In a specific implementation, this method employs a three-order gradient increasing pulse current waveform combined with staged dynamic control of the leveling agent and accelerator concentration ratio, and oscillating jet stirring. This allows for the independent implementation of electric field distribution control, the time-sequential selective inhibition / promotion of additives, and deep-hole mass transfer replenishment in three orthogonally controllable dimensions: the time domain, the chemical domain, and the fluid domain. This enables electroplating filling with a depth plating capability ≥80% and a copper layer uniformity deviation ≤15% in micro-through-holes with an aspect ratio ≥10:1. This method solves the technical problem of copper layer uniformity deterioration caused by the two major coupling difficulties of uneven current density distribution and limited mass transfer within the hole under high aspect ratio conditions in traditional DC electroplating or simple forward and reverse pulse electroplating. It achieves the beneficial effect of high-uniformity, high aspect ratio through-hole electroplating with low equipment modification costs while remaining compatible with existing mass-produced vertical continuous electroplating production lines.

[0028] like Figure 1 As shown, this method includes the following steps: S000: System initialization and calibration.

[0029] like Figure 2 As shown, this stage completes all preparatory steps, including electroplating solution preparation, additive addition, temperature control startup, auxiliary electrode calibration, and loading of the circuit board to be plated.

[0030] S010: Prepare an acidic copper sulfate electroplating solution containing copper sulfate, sulfuric acid, and a chloride ion source.

[0031] Using deionized water as a solvent, copper sulfate pentahydrate, concentrated sulfuric acid, and hydrochloric acid are mixed according to a preset concentration formula to prepare an acidic copper sulfate electroplating solution. In the preset concentration formula, the concentration of copper sulfate pentahydrate is 200 to 220 g / L, corresponding to a copper ion concentration of approximately 50 to 55 g / L; the concentration of sulfuric acid is 50 to 60 g / L; and the concentration of chloride ions is 40 to 60 ppm. After preparation, the electroplating solution is injected into a vertical continuous electroplating (VCP) tank, with the liquid level 50 mm above the upper edge of the anode. The anode is a phosphor bronze plate symmetrically arranged on both sides of the circuit board to be plated.

[0032] In a preferred embodiment, for example, the concentration of copper sulfate pentahydrate is 210 g / L, corresponding to a copper ion concentration of approximately 52.5 g / L; the sulfuric acid concentration is 55 g / L; and the chloride ion concentration is 50 ppm. The tank volume is 200 L (800 mm long × 500 mm wide × 500 mm high). After filling, the liquid level is 500 mm, the upper edge of the anode phosphor bronze plate is 450 mm high, and the liquid level exceeds the upper edge of the anode by 50 mm, ensuring that the anode plate is always completely submerged during the electroplating process. Copper sulfate provides copper ions for electrodeposition, sulfuric acid acts as a conductive salt to increase the conductivity κ of the electroplating solution (typically approximately 0.5 S / cm), and chloride ions provided by hydrochloric acid participate in the kinetics of copper deposition as an essential co-adsorbed component of the inhibitor (PEG / chloride ion synergistic system).

[0033] S020: Add leveling agent, accelerator and inhibitor to the electroplating solution to the target concentration for the first stage.

[0034] Three organic additives were sequentially added to the prepared base electroplating solution. The leveling agent was a nitrogen-containing polymer of polyethyleneimine (PEI) with a molecular weight (MW) of 10,000 to 30,000, and a target concentration C_L1 of 8 to 12 mg / L in the first stage. The accelerator was sodium polydisulfide dipropane sulfonate (SPS), with a target concentration C_A1 of 4 to 6 mg / L in the first stage. The inhibitor was polyethylene glycol (PEG) with a molecular weight of 4,000 to 8,000 and a concentration C_S of 200 to 400 mg / L, which was kept constant throughout all three electroplating stages. After addition, the concentration ratio of leveling agent to accelerator in the first stage was... It falls within the range of 1.5 to 3.0.

[0035] For example, with C_L1 = 10 mg / L (PEI, MW = 20000), C_A1 = 5 mg / L (SPS), and C_S = 300 mg / L (PEG, MW = 6000), the concentration ratio in the first stage is... The physical significance of this high ratio is as follows: the leveling agent (PEI) is a high molecular weight nitrogen-containing compound that preferentially adsorbs in the high current density region at the wellhead to form an inhibition layer, thereby reducing the copper deposition rate in that region; the accelerator (SPS) is a small molecule sulfur-containing compound that adsorbs in the low current density region at the bottom of the well to promote copper reduction kinetics. This ensures that the first stage exerts the highest level of inhibition on the orifice, consistent with the goal of maximizing uniformity in this stage. The inhibitor PEG and chloride ions synergistically adsorb onto the entire copper surface to form a substrate inhibition layer, providing a background for the differentiated adsorption of the leveling agent and accelerator.

[0036] S030: Start the temperature control circulation system to stabilize the temperature of the electroplating solution within the set range.

[0037] The temperature control circulation system (hereinafter referred to as the electroplating solution circulation and temperature control module M400 in the device embodiment) is activated. This module includes a magnetically driven circulation pump (flow rate 200 to 300 L / h, acid-resistant PVDF material), which performs heating or cooling operations on the electroplating solution through a plate titanium heat exchanger (heating power 6 kW / cooling power 4 kW) to stably control the temperature of the electroplating solution at a certain level. The temperature is set within the °C range. Temperature acquisition uses a Pt100 platinum resistance temperature probe with an accuracy of ±0.1°C. The temperature control system employs PID closed-loop control (exemplarily, P = 2.0, I = 0.5, D = 0.1) to drive the cold / hot water solenoid valves of the plate heat exchanger, with a temperature control response time of less than 30 seconds.

[0038] Simultaneously, the filtration system is activated. The activated carbon filter assembly with a filtration accuracy of 1 μm operates in a dual-cylinder switching configuration (the first cylinder removes solid particles, and the second cylinder adsorbs organic decomposition products), with a circulation flow rate of 4 to 6 times the tank volume per hour. For example, for a 200 L tank, the filtration circulation flow rate is set to 1000 L / h (i.e., 5 times / hour), and a replacement alarm is triggered when the pressure difference across the filter element exceeds 0.5 bar.

[0039] In addition, this module also includes an online copper ion concentration monitoring subunit and a pH monitoring subunit. The copper ion concentration monitoring is based on UV-Vis colorimetry (measuring the absorbance of the dd transition of copper ions at a wavelength of 810 nm), automatically sampling and analyzing every 30 minutes, with a detection range of 40 to 70 g / L and an accuracy of ±2 g / L; when the copper ion concentration falls below a set lower limit (exemplarily 48 g / L), an alarm for adding copper sulfate is triggered. The pH monitoring uses an online glass pH electrode (range 0 to 4), collecting data once per minute, maintaining the pH within the range of 0.3 to 0.8.

[0040] S040: Perform open-circuit potential calibration on the auxiliary reference electrode array.

[0041] The auxiliary reference electrode array consists of N miniature silver-silver chloride (Ag / AgCl) reference electrodes, evenly spaced on the anode frame along the thickness direction of the multilayer circuit board to be plated. For example, N = 5, and for a circuit board with a thickness H = 3.2 mm, the spacing between adjacent electrodes is... mm. The effective sensing area of ​​each reference electrode does not exceed 1 square millimeter, and it is led out to the high-impedance differential amplifier module outside the slot through a flexible wire.

[0042] During calibration, the open-circuit potential (OCP) of each reference electrode is recorded as a baseline under conditions without external current. The calibration criterion is that the OCP reading deviation of each electrode does not exceed ±2 mV. Electrodes that exceed the tolerance must be replaced and recalibrated. After calibration, the OCP baseline data is stored in the central controller in the data structure {electrode_id, OCP_mV, calibration_timestamp}.

[0043] For example, the OCP baselines of the five electrodes are OCP1 = -198.3 mV, OCP2 = -197.8 mV, OCP3 = -198.1 mV, OCP4 = -197.5 mV, and OCP5 = -198.0 mV, respectively. The maximum deviation is |(-198.3) - (-197.5)| = 0.8 mV < 2 mV, which meets the calibration requirements. Each reference electrode is equipped with an instrumentation amplifier to amplify the millivolt-level potential signal to an output range of 0 to 5 V with a common-mode rejection ratio greater than 100 dB. The amplified signal is synchronously acquired by a 16-bit precision, 100 Hz sampling rate multi-channel analog-to-digital converter (ADC) module and transmitted to the central controller via the SPI bus.

[0044] S050: Load the multilayer circuit board to be plated, which has already undergone chemical copper plating, into the electroplating tank, and align the jet nozzle array with the effective electroplating area of ​​the multilayer circuit board in a surface coverage mode.

[0045] A multilayer printed circuit board (PCB) that has undergone drilling, deburring, and PTH chemical copper plating (seed layer thickness approximately 0.3 to 0.5 μm) is mounted vertically on a VCP fixture, ensuring the via axes are horizontal. A jet nozzle array is installed at the bottom of the plating tank, with nozzles arranged closely at equal intervals not exceeding 2 mm, forming a uniform high-pressure jet field perpendicular to the PCB surface, covering the entire effective plating area of ​​the PCB. This surface coverage pattern does not require individual alignment with each via. The parallelism deviation between the PCB surface and the nozzle array plate does not exceed ±1°, and the distance between the PCB surface and the nozzle array plate is 15 to 25 mm.

[0046] For example, the circuit board to be plated is a 12-layer multilayer board with a thickness of [missing information]. mm, through hole diameter mm, aspect ratio The PTH chemical copper plating seed layer is 0.4 μm thick, providing the conductive substrate required for plating initiation. The nozzle array board measures 460 mm × 610 mm, with nozzle orifice diameters of 0.8 mm, arranged in a close-packed configuration at 1.5 mm intervals, comprising approximately [number missing] nozzles. There are one nozzle orifice. The distance between the PCB board and the nozzle array board is set to 20 mm, and the measured parallelism deviation is ±0.5°. The jet nozzle array board is made of 316L stainless steel, and its acid resistance meets the requirements for long-term immersion in acidic copper sulfate electroplating solution.

[0047] In the electroplating solution preparation process of step S010, the proportions of each component are selected according to the following technical criteria. The concentration of copper sulfate pentahydrate is selected in the range of 200 to 220 g / L, corresponding to a copper ion concentration of approximately 50 to 55 g / L. The lower limit of this concentration range ensures that the copper ion concentration in the holes will not drop below the critical deposition concentration (approximately 15 g / L) due to excessive consumption under high current density conditions, while the upper limit avoids copper sulfate crystallization due to solution oversaturation. The sulfuric acid concentration is selected in the range of 50 to 60 g / L. Its mechanism is as follows: sulfuric acid acts as a supporting electrolyte, increasing the conductivity κ of the electroplating solution (from approximately 0.2 S / cm in pure copper sulfate solution to approximately 0.5 S / cm), reducing the ohmic resistance of the electroplating solution, thereby reducing the potential drop difference between the inlet and center of the via, and indirectly improving the uniformity of current distribution. Excessively high sulfuric acid concentrations (greater than 80 g / L) will inhibit the cathodic reduction kinetics of copper, reducing deposition efficiency; excessively low sulfuric acid concentrations (less than 30 g / L) will result in insufficient conductivity to support high current density electroplating. The chloride ion concentration was selected to be 40 to 60 ppm. Chloride ions and the inhibitor PEG synergistically formed a co-adsorption layer covering the copper surface. This co-adsorption layer provided the necessary surface chemical substrate for the differentiated competitive adsorption of leveling agents and accelerators. When the chloride ion concentration was below 30 ppm, the co-adsorption layer was incomplete, and when it was above 80 ppm, it promoted the formation of the anolyte passivation film, affecting the anolyte dissolution efficiency.

[0048] The anode is a phosphorus copper plate containing 0.04% to 0.06% phosphorus. The addition of phosphorus causes a black copper phosphide film to form on the copper surface during the anode dissolution process. This film can filter out the release of copper particles and anode sludge, reducing the interference of suspended particles in the electroplating solution on the copper deposition quality. The phosphorus copper anodes are installed symmetrically on both sides of the PCB in the electroplating tank to ensure that the electric field applied to the PCB surface is symmetrically distributed in the direction normal to the board surface.

[0049] During the additive addition process in step S020, the concentration of the inhibitor PEG (polyethylene glycol) remains constant throughout all three stages. The technical reason for this is that the substrate inhibition layer formed by PEG and chloride ions is a prerequisite for the normal operation of the entire additive system. The formation and maintenance of this substrate layer require a sufficient and stable concentration of PEG (200 to 400 mg / L) in the electroplating solution. If the PEG concentration is adjusted synchronously during stage switching, it will lead to uncontrollable changes in the thickness and density of the substrate inhibition layer, making the differentiated adsorption effects of the leveling agent and accelerator unpredictable. Therefore, the PEG concentration is fixed as an "invariant," and a staged control strategy is implemented only by adjusting the concentration ratio of the leveling agent and accelerator.

[0050] For example, PEG with a molecular weight of 6000 was selected at a concentration of 300 mg / L. The effect of PEG molecular weight on the inhibition effect is as follows: when the molecular weight is below 3000, the PEG molecules are too small to form an effective co-adsorption layer with chloride ions; when the molecular weight is above 10000, the diffusion of PEG in the micropores is more restricted, resulting in an excessive difference in the thickness of the substrate inhibition layer between the pore opening and the pore bottom. PEG with a molecular weight of 4000 to 8000 achieves a balance between these two constraints.

[0051] In the auxiliary electrode calibration of step S040, the technical purpose of open-circuit potential (OCP) calibration is to establish a benchmark for overpotential calculation. The overpotential... This is a physical quantity describing the degree to which the electrode surface deviates from thermodynamic equilibrium, and its value determines the local electrodeposition driving force at that location. Different reference electrodes may have different intrinsic potential offsets due to manufacturing differences. These offsets are quantified and recorded through OCP calibration, allowing subsequent overpotential calculations to exclude systematic errors inherent to the electrodes themselves. The ±2 mV calibration deviation tolerance is based on the following considerations: the first-stage judgment threshold is 15 mV. If the systematic deviation between electrodes reaches 5 mV, it will lead to a judgment error of approximately 33%, which is unacceptable; when controlled within 2 mV, the judgment error is less than 13%, which is within the acceptable range.

[0052] In step S050, the 15-25 mm spacing between the PCB board and the nozzle array board is selected based on fluid dynamics considerations. Too small a spacing (less than 10 mm) will result in excessive jet impact pressure, causing irregular turbulent vortices in the plating solution on the PCB surface, interfering with the electric field distribution; too large a spacing (greater than 30 mm) will lead to excessive jet attenuation, preventing effective penetration of the vias to renew the internal liquid phase. The 15-25 mm range ensures that the jet maintains sufficient momentum to penetrate the vias upon reaching the PCB surface, while maintaining acceptable spatial uniformity of the jet field. The requirement of a parallelism deviation not exceeding ±1° ensures that the jet pressure difference received by different areas of the PCB board does not exceed 5%, avoiding uneven mass transfer due to insufficient local jet flow.

[0053] In the inter-module communication architecture of the electroplating system, the central controller M600 acts as a multi-protocol communication hub, sequentially completing the following operations within a 100 ms master control cycle: First, it receives the latest potential snapshot data packet from the M500's SPI bus, taking approximately 5 ms; second, it executes a 10-point moving average filtering and stage determination algorithm, taking approximately 2 ms; third, if a stage switching condition is detected, it sends parameter commands to the M100 via Modbus RTU (taking approximately 10 ms), sends JSON commands to the M200 via TCP / IP (taking approximately 15 ms), and sends parameter commands to the M300 via Modbus RTU (taking approximately 10 ms); fourth, it receives the tank status register reported by the M400 (once every 60 seconds, taking approximately 5 ms); fifth, it updates the HMI display interface (taking approximately 10 ms). All operations are completed within a 100 ms cycle, ensuring the real-time performance of the control system. During normal operation (without stage switching), the master control cycle only executes the first, second, and fifth steps, with a total time of approximately 17 ms and a system load of less than 20%.

[0054] The M600 is also equipped with a data logging function. Of the three characteristic dimensions of the electroplating process state tensor, the overpotential η is directly derived from the measurements of the M500; the calculated current density J is obtained by a linear approximation of the Butler-Volmer equation. The exchange current density is derived from the overpotential. During the calibration phase, it is determined by Tafel curve fitting (exemplary). The copper ion concentration was calculated based on the diffusion overpotential component of the Nernst equation (A / cm²). This tensor was stored in HDF5 format on the solid-state drive of the industrial computer. The amount of data generated in a single electroplating process (95 minutes) was approximately [amount missing]. byte MB (10 Hz sampling rate, 5 electrode channels, 3 feature dimensions, 64-bit double-precision floating-point).

[0055] In this embodiment, the parameter design of the third-order gradient pulse waveform follows a complete mathematical modeling process from input to output. In the pulse electroplating, the duty cycle DC and peak current density... Together they determine the effective average current density A low duty cycle means a longer off-time, providing a more ample time window for the diffusion and replenishment of copper ions within the via; however, an excessively low duty cycle leads to low overall electroplating efficiency and insufficient copper deposition per unit time. The essence of the third-order gradient strategy lies in the gradual increase of the duty cycle and peak current density as the effective aspect ratio of the via gradually decreases due to copper deposition in the preceding stages, thus achieving a smooth transition from a uniformity-first stage to an efficiency-first stage.

[0056] The effective average current density for each stage is given by the following formula: Subscript Identify the electroplating stage number. The parameters for the three stages are designed according to the following gradient increasing relationship: , The incremental gradient between stages is approximately 2 to 2.5 times, with the duty cycle increasing by about 20 percentage points per stage. This ensures a gradual improvement rather than abrupt changes, avoiding the degradation of the copper layer structure quality caused by sudden changes in current distribution during the transition period.

[0057] For example, the complete parameter set for the three phases is shown below. Phase 1: A / dm², ms, ms, , A / dm². Second stage: A / dm², ms, ms, , A / dm². Third stage: A / dm², ms, ms, , A / dm². From Phase 1 to Phase 3. It gradually increased from 0.16 to 2.40 A / dm², a 15-fold increase.

[0058] In the design of the above three-stage parameters, the peak current density in the first stage... The upper limit is constrained by the Wagner number. Specifically, in terms of conductivity... S / cm of acidic copper sulfate plating solution and through-hole depth Under the condition of cm, in order to make (Engineering criterion for near-uniform current distribution). It must not exceed approximately 1.2 A / dm². A safety margin of 0.8 A / dm² is provided. Peak current density for stage three. The upper limit is constrained by the quality of copper crystals; in acidic copper sulfate systems, it should not exceed 6.0 A / dm² to avoid coarse dendritic crystal formation and hydrogen evolution at the cathode. 4.0 A / dm² is within the safe range.

[0059] As an alternative implementation, a linearly varied current density mode can be used to replace the third-order discrete stage—the current density changes from... The slope gradually changes linearly to a preset gradient. This alternative approach offers a smoother transition, but the additive concentration control is difficult to precisely synchronize with continuously changing current densities, resulting in higher engineering complexity. Another alternative is a two-stage scheme (omitting the low-current penetration period in the first stage). This simplified scheme may be sufficient for boards with an aspect ratio less than 8:1, but for boards with an aspect ratio of not less than 10:1, the lack of a low-current penetration period will lead to discontinuities in the initial copper layer at the hole center, making it impossible to guarantee the uniformity of filling in subsequent stages.

[0060] In the dynamic control of the additive concentration ratio, the concentration ratio of leveling agent to accelerator... In the three stages The concentration ratio decreases gradually. The value of the concentration ratio was determined through a Hull cell experiment: different concentrations were introduced into the Hull cell... Electroplating was performed, and the differences in gloss and thickness distribution between high-current-density and low-current-density areas were observed to determine the optimal conditions for each stage. Value. The preferred range is: , , .

[0061] For example, ( mg / L, (mg / L) corresponds to the first stage of strong inhibition orifice mode: high concentration of leveling agent forms a dense adsorption layer at the orifice, reducing the orifice deposition rate by about 16% (calculated by the deposition rate correction model). ( mg / L, (mg / L) corresponds to the equilibrium mode of the second stage: the leveling agent inhibits the intensity of the depression, the accelerator promotes the effect, and the deposition rate at the bottom of the well is about 18% higher than that at the orifice. ( mg / L, (mg / L) corresponds to the efficiency-first mode of the third stage: the leveling agent significantly reduces the inhibition of the orifice, allowing the orifice to deposit at a rate close to that inside the orifice.

[0062] As an alternative implementation, a continuous flow mode can be used instead of the staged control, employing a peristaltic pump with a time function... Continuously reduce the leveling agent / accelerator ratio, where The decay rate constant is 1 / min. The control accuracy of this alternative method depends on the pump's flow linearity and mixing delay.

[0063] In the stage switching determination, the selection of the overpotential difference threshold follows the rules below. First stage threshold... mV (range 10 to 20 mV): A more lenient threshold to match the positioning of the first stage, "establishing the initial homogeneous layer". Second stage threshold. mV (range 8 to 15 mV): A stricter threshold ensures a high degree of consistency in the electrochemical state at all depths when the bulk filling phase is completed. The number of consecutive satisfactions in the two phases are 30 and 50, respectively, corresponding to continuous satisfaction windows of approximately 3 seconds and 5 seconds. The former provides a disturbance tolerance margin of 30 data points during the shorter pulse period (100 ms) of the first phase, while the latter ensures the reliability of switching decisions with a higher count requirement during the longer filling process of the second phase.

[0064] The complete execution logic of the stage determination algorithm is as follows: when the state machine is in STAGE1, calculate the overpotential difference between the orifice electrode and the orifice center electrode. ;when The `consecutive_count` is incremented sequentially; otherwise, it is reset to 0. The `consecutive_count` is incremented when `consecutive_count` is not less than `count_req1` (30 times) or when the duration is not less than [a certain value]. The Stage 1 to Stage 2 transition is triggered at 30 min. When the state machine is in Stage 2, the maximum overpotential difference between adjacent electrodes is calculated. ;when The `consecutive_count` is incremented sequentially; otherwise, it is reset to 0. Simultaneously, the Faraday cumulative subunit is invoked to calculate... When consecutive_count is not less than count_req2 (50 times) and or has lasted for no less than At 60 min, the Stage 2 → Stage 3 transition is triggered. When the state machine is in Stage 3, the Faraday cumulative deposition thickness is used only. Electroplating is terminated.

[0065] In practical applications of this embodiment, the method is particularly suitable for the following technical scenarios: through-hole electroplating of multilayer printed circuit boards with a board thickness of 3.0 to 5.0 mm, a via diameter of 0.15 to 0.35 mm, and an aspect ratio of 10:1 to 20:1. For cases where the board thickness is less than 2.0 mm and the via density is extremely high (hole spacing less than 0.5 mm), the spatial resolution of the reference electrode array may be insufficient to accurately characterize the independent electrochemical state of each via. In such cases, the number of reference electrodes can be reduced to... To compensate for resolution loss, the stage switching threshold can be lowered, or alternatively, a "time series statistical method" can be used—the state inside the well is estimated by monitoring the change in the time constant of the pulse response using a single reference electrode. Given the limitation of the CVS online analyzer in the limited number of analyses during the first stage (lasting only 15 to 20 minutes), feedforward control can be used instead of feedback control in the first stage—replenishing additives according to a preset schedule based on a pre-calibrated additive consumption rate curve, switching to CVS feedback closed-loop control in the second stage and thereafter.

[0066] S100: Perform the first stage of electroplating by outputting a pulse current with a first peak current density and a first duty cycle. During the turn-off period of the pulse current, perform oscillating jet stirring on the through hole through the jet nozzle array. At the same time, collect the local overpotential at different depth positions of the through hole through the auxiliary reference electrode array. When the overpotential difference between the hole opening and the hole center continuously meets the first threshold condition, it is determined that the first stage of electroplating is completed.

[0067] like Figure 3 As shown, this stage is the low current density penetration stage, during which a uniform initial copper deposition layer is established on the entire surface of the hole wall of the high aspect ratio via.

[0068] S110: The central controller sends the first-stage parameter command to the pulse power module.

[0069] The central controller is an industrial computer deployed in the control cabinet of the electroplating production line, running a Linux real-time operating system. Internally, the controller maintains a stage parameter mapping table, storing the complete parameter sets for the three stages. During stage switching, the controller reads the corresponding parameter set from the mapping table and distributes it to the pulse power supply module, additive control module, and jet stirring module.

[0070] The first-stage parameters sent by the controller to the pulse power supply module include: forward peak current density. Pulse conduction time Pulse off time The first duty cycle Effective average current density In the first stage, no reverse pulse is applied. Simultaneously, jet parameters (jet pressure) are sent to the jet stirring module. oscillation frequency .

[0071] For example, A / dm², ms, ms, , A / dm². Jet parameters are: bar Hz (the jet direction oscillates sinusoidally at 2 Hz with an amplitude of ±15°).

[0072] The selection is based on the Wagner number constraint. Wagner number It is a dimensionless criterion characterizing the uniformity of current distribution; the larger its value, the closer the current distribution is to a uniform primary current distribution. This is derived from the Butler-Volmer equation. At low overpotential ( Linear expansion under the condition of mV can yield the partial derivative of the overpotential with respect to the current density. ,in J / (mol·K) is the gas constant. The temperature of the electroplating solution is (K). , where is the charge transfer coefficient (dimensionless, typical value 0.5 for acid copper systems). The number of electrons transferred during the reduction of copper ions (dimensionless). C / mol is the Faraday constant. Substituting the partial derivative into the definition of the Wagner number... ,get: in The conductivity of the electroplating solution (S / cm) is given. The depth of the through hole is in cm.

[0073] For example, substitute S / cm, , K (23°C) , , , A / cm² (conversion from 0.8 A / dm²), cm (plate thickness 3.2 mm), then the molecule denominator , . It was confirmed that the current distribution in the first stage approximates the primary current distribution. S / cm, Under cm conditions, It needs to be below approximately 1.2 A / dm² to achieve this. (Engineering safety threshold), a safety margin of 0.8 A / dm² is provided. If... A / dm², then If the deposition rate drops below 3, the deposition rate at the pore opening will be much higher than that at the pore center, and the homogenization target of the first stage cannot be achieved.

[0074] The selection is based on the copper ion diffusion supply requirement during the off-period. This is estimated based on one-dimensional diffusion time. The diffusion coefficient of copper ions cm² / s, half-hole depth cm, s s. The time ms is much shorter than 2.6 s, but with jet stirring, the effective diffusion layer thickness is increased. The depth was reduced to less than 50 μm, so that an 80 ms off-time was sufficient to replenish copper ions in the diffusion layer.

[0075] The core hardware of the pulse power supply module is an IGBT power module (drive capacity ≥100 A / 12 V), and the power controller is based on a DSP (TI TMS320F28x series) to generate T with a time base accuracy of 1μs. on / T off Square wave signal. Current feedback uses a Hall effect current sensor (accuracy ±0.5%), performing real-time PI closed-loop control with a response time of less than 1 ms. Overcurrent / overvoltage protection circuit: when the current exceeds... If the voltage exceeds the maximum allowable value, the IGBT will be disconnected within less than 10μs. The module's internal maintenance phase parameter table is a fixed-size array, with each element containing the fields {J_peak_mA, T_on_us, T_off_us, ramp_rate_mA_per_ms}.

[0076] The controller sends stage switching commands to the pulse power module via an RS-485 bus (Modbus RTU protocol). The command format is {device_addr = 0x01, func_code = 0x06, register_addr = stage_reg, value = stage_id}. At the end of each pulse cycle, the pulse power module polls the Modbus register, and upon detecting a new stage_id, it takes effect in the next cycle.

[0077] S120: The pulse power module outputs a pulse current according to the parameters of the first stage, performs copper electrodeposition during the conduction period of each pulse cycle, and performs oscillating liquid phase renewal on the through hole by the jet stirring module during the off period.

[0078] During each pulse cycle, the conduction period An internal forward current is applied to the cathode (via the via wall), and copper ions are reduced and deposited as metallic copper at the cathode. Under low current density (0.8 A / dm²), the polarization potential is small, and the non-uniformity of the electric field distribution is weakened by the Wagner number effect, resulting in approximately uniform initial copper deposition across the entire via wall. During the shutdown period... Inside, the pulse power supply module stops outputting current. At the same time, the jet stirring module is synchronously triggered at the start of the shutdown period.

[0079] The core hardware of the jet mixing module consists of a high-pressure centrifugal pump (maximum flow rate 50 L / min, maximum pressure 3 bar), a jet nozzle array plate, and a servo motor-driven deflection mechanism. The high-pressure pump controls the jet pressure by adjusting its speed through a frequency converter, and a pressure sensor at the outlet performs PID closed-loop control, ensuring pressure fluctuations do not exceed ±0.1 bar. The deflection mechanism, driven by a servo motor via a crank-connecting rod mechanism, allows the nozzle array plate to perform sinusoidal reciprocating deflection with a deflection angle of ±15° and an adjustable frequency range of 1 to 5 Hz. The synchronous trigger signal is implemented via hardware wiring—the T signal from the pulse power module... off The synchronization pulse is directly connected to the start relay of the jet stirring module to ensure precise synchronization between the jet and the turn-off period. During this period, the jet is shut off or reduced to a minimum flow rate to avoid interfering with the electric field distribution.

[0080] The jet mixing module maintains a jet parameter register with a data structure of {P_target_bar, f_osc_Hz, amplitude_deg, sync_mode}, where the sync_mode value of ON_DURING_TOFF indicates that it only starts during shutdown. The controller sends parameters to the frequency converter and servo drive via Modbus RTU.

[0081] In the first stage, the jet pressure is 1.5 bar and the oscillation frequency is 2 Hz. The jet sweeps across all the vias on the PCB board, physically pushing out the copper-depleted plating solution inside the vias due to electrodeposition, and replenishing it with fresh plating solution to eliminate the concentration gradient. At the same time, the leveling agent (PEI) migrates and adsorbs into the high current density area at the via opening through convection diffusion, forming an inhibition layer in that area.

[0082] S130: The auxiliary reference electrode array collects the cathode overpotential at each depth position at the end of each pulse conduction period.

[0083] Two milliseconds before the end of each pulse conduction period, the auxiliary reference electrode array acquires the cathode potential value of each reference electrode point at a sampling rate of 100 Hz. ( The controller calculates the local overpotential. The controller packages the data into potential snapshot data packets, in the format {timestamp_ms,channel_id[1..N],potential_mV[1..N],bath_temp_C, pH_value}, generating one data packet every 10ms.

[0084] The data acquisition and preprocessing subunit inside the controller performs a 10-point moving average filter on the received potential snapshot to remove high-frequency noise and generate a smoothed overpotential sequence. This sequence serves as the input to the stage determination logic. The controller also maintains an electroplating process state tensor with dimension [missing information]. ,in The number of time sampling points, The number of electrodes is used, and the three characteristic dimensions are overpotential η, estimated current density J, and estimated copper ion concentration, which are used for post-process analysis and parameter optimization.

[0085] For example, at the start of electroplating At time min, the overpotentials collected by the five electrodes were as follows: mV (orifice) mV, mV (hole center) mV, The absolute value of the overpotential at the orifice is much greater than that at the orifice center, reflecting the non-uniform distribution of the orifice current density in the initial state, which is much higher than that at the orifice center.

[0086] S140: The controller performs the first stage completion judgment. When the overpotential difference between the orifice and the center of the orifice continuously meets the first threshold condition, it outputs a stage switching signal.

[0087] The stage determination logic subunit inside the controller maintains a state machine {INIT → STAGE1 → STAGE2 → STAGE3 → COMPLETE} and executes the following determination algorithm: calculates the orifice electrode ( ) and the center electrode of the hole ( overpotential difference .when Satisfying within 30 consecutive pulse cycles When mV is reached, the first stage is considered complete, triggering the switch from Stage 1 to Stage 2.

[0088] The first threshold The mV value was determined based on the linear approximation of the Butler-Volmer equation under conditions of 50 g / L copper ion concentration and 23°C. It is estimated that a 15 mV overpotential difference corresponds to approximately 8% to 12% local current density deviation, meeting the uniformity target of the first stage. Its value range is 10 to 20 mV; below 10 mV is too stringent, resulting in excessively long first-stage time, while above 20 mV, uniformity is insufficient. The requirement is approximately 30 consecutive counts (approximately...). ms The continuous satisfaction of condition s is used to filter out false triggers caused by short-term fluctuations. If the counting is interrupted due to the failure to meet the condition, the counter is reset to 0 and starts accumulating again.

[0089] Timeout protection mechanism: If the duration of the first phase exceeds the first maximum duration. If the condition is still not met, the controller will forcibly switch to the second stage and record an alarm.

[0090] For example, in min time, mV, mV, mV mV, condition not met. min time, mV, mV, mV The voltage level has reached mV, and this has been met for more than 30 consecutive pulse cycles, triggering the Stage1→Stage2 switching signal. The actual duration of the first stage is approximately 18 minutes (including the switching delay).

[0091] During the pulse electroplating process in step S120, the copper electrodeposition reaction It occurs on the surface of the cathode (through-hole wall). The kinetics of the reaction are governed by the Butler-Volmer equation at a given overpotential. Under these conditions, cathode current density and The relationship between them is exponential. In the low overpotential linear region ( mV). and Approximately proportional, at this point the current distribution is mainly determined by the ohmic resistance (i.e., conductivity) of the electroplating solution. The Wagner number is determined by the geometry of the through-hole and the bore. It is under this linear approximation condition that the dimensionless criterion characterizing the uniformity of current distribution is derived.

[0092] During the pulse conduction period Within the via, copper ions are consumed from the diffusion layer near the hole wall, resulting in a copper ion concentration in the diffusion layer that is lower than the bulk concentration. At depths of the via, since natural convection of the plating solution is almost impossible, copper ion replenishment relies entirely on molecular diffusion, and the replenishment rate is constrained by Fick's diffusion law. The thickness of the diffusion layer... Under static conditions, the diameter can reach hundreds of micrometers, while under the forced convection conditions of oscillating jet stirring, it is compressed to below 50 μm, which greatly shortens the mass transfer path of copper ions from the bulk phase to the electrode surface.

[0093] During the shutdown period Inside, the current is zero, and no electrodeposition reaction occurs. At this time, the copper ion concentration inside the orifice begins to recover due to the cessation of consumption; jet stirring accelerates this recovery process through forced convection. Another important function of the turn-off period is that leveling agent molecules continuously migrate and adsorb towards the high current density region at the orifice via convective diffusion during the turn-off period, preparing for differentiated deposition in the next conduction period. The leveling agent PEI is a high molecular weight nitrogen-containing compound (molecular weight 10,000 to 30,000), and its diffusion coefficient in solution is approximately [missing information]. cm² / s, much smaller than the diffusion coefficient of copper ions ( (cm² / s), therefore, jet agitation during the shut-off period is particularly critical for the transport of leveling agent to the orifice.

[0094] In step S230, the technical differences between the second-stage pulse electroplating and the first-stage process are mainly reflected in three aspects. First, the current density increases from 0.8 A / dm² to 2.0 A / dm², resulting in a corresponding increase in deposition rate. However, since a uniform initial copper layer of approximately 0.6 μm thickness has already been established on the hole wall in the first stage, the conductivity of the via is improved, the potential drop within the hole decreases, and the non-uniformity of the current distribution is improved compared to the initial bare hole state. Second, the duty cycle increases from 20% to 40%, increasing the on-time and shortening the off-time. This requires a simultaneous increase in the pressure and frequency of jet stirring to ensure sufficient copper ion replenishment within a shorter off-time. Third, the additive concentration ratio decreases from 2.0 to 0.75, reducing the inhibitory strength of the leveling agent while increasing the promoting strength of the accelerator, causing the deposition rate distribution to shift from a "strong homogenization" mode to a "moderate homogenization + high efficiency" mode.

[0095] In step S320, the success of the high-current-density pulse electroplating in the third stage relies on the effective modification of the via geometry in the first two stages. After approximately 70 minutes of electroplating in the first and second stages, a copper layer of approximately 15.4 μm was deposited on the via wall (approximately 0.6 μm in the first stage and approximately 14.8 μm in the second stage), and the effective via diameter was reduced from the initial 0.25 mm to approximately [missing value]. mm (in reality, due to slightly more deposition at the orifice than at the center, the effective diameter at the orifice may shrink to approximately 0.20 mm). At this point, the effective thickness-to-diameter ratio of the through-hole decreases from the initial 12.8:1 to approximately... (At the opening) to (At the center of the hole), because the copper layer itself has a much higher conductivity than the chemically deposited copper seed layer (the conductivity of copper is approximately...), The actual effective conductivity path characteristics (S / cm) have been significantly improved. The uniformity of current distribution is no longer dominated by the initial geometric aspect ratio, but is guaranteed by the uniform conductivity of the copper layer itself. This is the physical basis for the third stage to operate at higher current densities without sacrificing the final uniformity.

[0096] S200: Adjust the concentration ratio of leveling agent to accelerator in the electroplating solution from a first ratio to a second ratio, and perform the second stage of electroplating by outputting pulse current with a second peak current density and a second duty cycle. When the maximum value of the overpotential difference between adjacent reference electrodes continuously meets the second threshold condition and the cumulative deposition thickness reaches a preset ratio of the target thickness, the second stage of electroplating is determined to be completed.

[0097] like Figure 4 As shown, this stage is the medium current density filling stage. Based on the uniform initial copper layer established in the first stage, copper deposition is accelerated at a medium current density to achieve the main filling of the via.

[0098] S210: The controller sends the second-stage parameter instructions to the pulse power module, the additive control module, and the jet stirring module.

[0099] Upon receiving the Stage1→Stage2 switching signal, the controller reads the second stage parameter set from the stage parameter mapping table and distributes it. The parameters sent to the pulse power supply module include: forward peak current density. Pulse conduction time Pulse off time The second duty cycle Effective average current density Send an additive concentration adjustment command to the additive control module, adjusting the leveling agent concentration from... Down to accelerator concentration from Rise to , so that the concentration ratio is from Adjust to Send updated jet parameters to the jet mixing module: jet pressure. oscillation frequency .

[0100] For example, A / dm², ms, ms, , A / dm². The additive concentration was adjusted to... mg / L, mg / L, The jet parameters have been updated to... bar Hz.

[0101] Compared to the first stage, Increased from 0.8 to 2.0 A / dm² (2.5 times). Increased from 20% to 40% (an increase of 20 percentage points). The efficiency was increased from 0.16 to 0.80 A / dm² (5 times). This incremental gradient design allowed for a gradual increase in deposition rate, avoiding abrupt changes in current distribution caused by abrupt jumps in parameters from low to high. Verification was performed using the Wagner number formula: It is still much greater than 1, but it is lower than that of the first stage (622) - since the first stage has established an initial copper layer of about 0.6 μm on the hole wall, the effective aperture of the via has been slightly reduced (about 2 to 5 μm on each side), and the geometric non-uniformity of the electric field distribution has been improved.

[0102] The controller sends a JSON-formatted instruction to the PLC of the additive control module via Ethernet TCP / IP: {"stage": 2, "C_L_target": 6.0, "C_A_target": 8.0}. The additive control module reports the current concentration status to the controller every 60 seconds.

[0103] S220: The additive control module performs concentration ratio adjustment.

[0104] The additive control module is located in an independent control cabinet next to the circulation pipeline of the electroplating tank. It includes three precision peristaltic metering pumps (for quantitative injection of leveling agent, accelerator and inhibitor mother liquor respectively) and one cyclic voltammetric stripping (CVS) online analyzer.

[0105] Leveling agent concentration reduction: The CVS online analyzer automatically takes a 50 mL sample from the electroplating tank every 10 minutes and performs a cyclic voltammetric scan (scan range -0.3 V to +0.3 V vs. Ag / AgCl, scan rate 100 mV / s). The leveling agent concentration is quantified by the PEI reduction peak area, with an accuracy of ±5%. The PLC program performs a judgment based on the CVS analysis results: when the actual concentration is higher than... When the actual concentration is below a certain leveling agent leveling pump, stop the leveling agent supply pump; when the actual concentration is below a certain level... When the concentration reaches 5.4 mg / L, the replenishment pump is restarted. The time it takes for the leveling agent concentration to naturally decrease from 10 mg / L to the target value of 6 mg / L depends on the rate at which the leveling agent is consumed on the copper surface during the electroplating process.

[0106] Accelerator concentration adjustment: The PLC program calculates the required replenishment amount and drives the accelerator metering pump (accuracy ±0.5 mL / min). The accelerator replenishment amount formula is: in The volume of the tank liquid (L) The concentration of the accelerator mother liquor is (mg / L). This represents the current actual accelerator concentration (mg / L). For example, mg / L, mg / L, L, mg / L (1% stock solution), then L mL. The metering pump runs at a flow rate of 6 mL / min for 10 minutes to complete the injection. The injection point is located at the return end of the circulation pipeline, with a pipeline delay of approximately 30 to 60 seconds from the injection point to the electroplating tank. The PLC compensates for the advance of the pump start-up and shutdown sequence. The CVS online analyzer quantifies the accelerator concentration by measuring the SPS oxidation peak area, verifying the adjustment results.

[0107] The additive control module maintains an additive state vector: {C_L_measured, C_A_measured, C_S_measured, C_L_target, C_A_target, R_target, last_CVS_timestamp}. The `measured` field is updated after each CVS analysis, and the `target` field is updated after each controller command is received. The additive concentration in the electroplating solution transitions to the second-stage target value within 15 minutes.

[0108] Reduced concentration ratio The physical significance lies in: compared to the first stage The leveling agent's inhibitory effect on the wellhead is reduced, while the increased concentration of the accelerator deep within the well promotes bottom deposition. The net effect is: leveling agent adsorption still inhibits deposition at the wellhead (but with moderate intensity), the deposition rate inside the well is increased, and the difference in deposition rates between inside and outside the well is further reduced. This balance achieves both uniformity and efficiency.

[0109] The differentiated adsorption behavior of the additives was described by Langmuir adsorption isotherms. The leveling agent was located at... The surface coverage at this location is: The accelerator is in position The surface coverage at this location is: in and These are the adsorption equilibrium constants (L / mg) for the leveling agent and the accelerator, respectively. and For additives in Local concentration (mg / L) at the location. PEI (nitrogen-containing group) and SPS (sulfur-containing group) each chemically favor different active sites on the Cu surface. However, the total number of usable active sites on the Cu cathode surface is limited, and the adsorption layers of the two additive molecules spatially repel each other—the adsorption of large PEI molecules occupies a larger surface area, objectively reducing the surface area accessible to small SPS molecules. Therefore, a competitive Langmuir model is used as a macroscopic approximation. Typical parameter values ​​are: to L / mg, to L / mg, to (Leveling agent inhibition coefficient, dimensionless). to (Accelerator promoting coefficient, dimensionless).

[0110] The local deposition rate is corrected to: in When there are no additives The deposition rate at a given location is determined by the local current density. The correction factor... and All are dimensionless. At the orifice (high current density region). High (leveling agent preferentially diffuses to the orifice) makes Larger Suppressed; at the bottom of the hole (low current density region). Relative dominance Larger Promoted.

[0111] For example, take L / mg, L / mg, , In the second phase ( mg / L, mg / L), at the orifice ( Approximately equal to a tank solution concentration of 6 mg / L. (Approximately 5 mg / L, slightly lower than the tank solution concentration due to consumption) , , At the bottom of the hole ( Because diffusion is limited to approximately 2 mg / L, (Approximately 7 mg / L, slightly better due to diffusion of small molecules) , , The deposition rate at the bottom of the well is about 18% higher than that at the wellhead, which helps to homogenize the filling.

[0112] As an alternative implementation, an independent Langmuir model can also be used: , At this point, the adsorption of the two additives is completely independent. The two models are interchangeable within the engineering accuracy range (±10% coverage prediction error).

[0113] S230: The pulse power module outputs a pulse current of medium current density according to the parameters of the second stage to perform copper deposition.

[0114] A moderate current density (2.0 A / dm²) within ms drives a copper deposition rate approximately 5 times that of the first stage. The jet pressure was increased from 1.5 bar to 2.0 bar, and the oscillation frequency was increased from 2 Hz to 3 Hz to cope with the greater copper ion consumption rate under medium current density and to ensure that the copper ion concentration deep in the hole is adequately replenished.

[0115] S240: The controller performs the second phase completion determination.

[0116] The controller calculates the maximum overpotential difference between adjacent reference electrodes. .when Satisfying the condition within 50 consecutive pulse cycles At mV, the electrochemical state at each depth of the via tends to be consistent, and the bulk filling has reached uniformity. Unlike the first stage, the determination of the second stage also requires that the cumulative deposition thickness reaches a preset proportion of the target thickness.

[0117] The cumulative deposition thickness is calculated in real time using piecewise summation based on Faraday's law. The Faraday cumulative calculation subunit of the controller accumulates the thickness at the end of each pulse cycle based on the actual output current and running time obtained from feedback from the pulse power supply module. in g / mol is the molar mass of copper. The number of electrons transferred (dimensionless). C / mol is the Faraday constant. g / cm³ is the density of copper. Current efficiency (typically 95% to 98%, dimensionless). This represents the effective average current density at the current stage (calculated based on the orifice wall area, in A / cm²). (i.e., 70% of the target thickness) and satisfied for 50 consecutive times. When mV is reached, the Stage2→Stage3 switch is triggered.

[0118] The mV requirement is more stringent than the 15 mV in the first stage, as the vias are now partially filled, and higher uniformity is required to ensure the final deep plating capability meets the standard. Its value ranges from 8 to 15 mV. A continuous count requirement of 50 cycles (approximately...) ms (The continued satisfaction of s) provides a higher disturbance immunity margin.

[0119] Timeout protection: min.

[0120] For example, in min (5 minutes after entering the second stage), mV, mV, mV, condition not met. min time, mV mV. In min time, mV, and μm μm — here The 17.5 μm threshold has not yet been reached; electroplating continues. min time, μm If the value is μm and the condition is met consecutively more than 50 times, the Stage 2 → Stage 3 switch is triggered. The actual duration of the second stage is approximately 52 minutes.

[0121] S300: Adjust the concentration ratio from the second ratio to the third ratio, and perform the third stage of electroplating by outputting a pulse current with a third peak current density and a third duty cycle. The electroplating is terminated when the cumulative deposition thickness reaches the target thickness.

[0122] like Figure 5 As shown, this stage is the high current density sealing stage, where the final copper deposition and sealing is completed rapidly with a high current density after the via body is filled.

[0123] S310: The controller sends the third-stage parameter instructions to each module.

[0124] Upon receiving the Stage2→Stage3 switching signal, the controller sends the following parameters to the pulse power supply module: forward peak current density. Pulse conduction time Pulse off time The third duty cycle Effective average current density Send a command to the additive control module to adjust the leveling agent concentration to... The accelerator concentration was adjusted to Concentration ratio Send parameters to the jet mixing module: , .

[0125] For example, A / dm², ms, ms, , A / dm². mg / L, mg / L, . bar Hz.

[0126] The complete gradient for three-stage parameter increment is: From 0.8 → 2.0 → 4.0 A / dm² (approximately 2 to 2.5 times per stage). From 20% to 40% to 60% (20 percentage points per stage). The parameters decreased gradually from 2.0 to 0.75 to 0.3, the jet pressure from 1.5 to 2.0 to 2.5 bar, and the oscillation frequency from 2 to 3 to 4 Hz. This gradual increase avoided abrupt changes in the parameters.

[0127] Substituting Wagner numbers to verify the third stage: A / cm², . It is still greater than 1 but lower than the previous two stages. However, at this time, the effective aperture of the through hole has been reduced from the initial 0.25 mm to about 0.10 to 0.15 mm (the previous two stages deposited about 50 to 75 μm on each side), and the effective thickness-to-diameter ratio has been reduced from 12.8:1 to about 5:1 to 3:1. The uniformity problem of traditional electroplating no longer constitutes a bottleneck under this low thickness-to-diameter ratio condition.

[0128] This significantly reduces the inhibitory effect of the leveling agent on the orifice, allowing the orifice area to deposit at a rate close to that inside the orifice, thus accelerating sealing. At this point, the uniformity requirement has been ensured by the sufficient homogenization in the first two stages, and the third stage prioritizes efficiency.

[0129] like A value exceeding 6.0 A / dm² may cause "burning" of the copper layer (coarse dendritic crystals) and cathodic hydrogen evolution in an acidic copper sulfate system, leading to a decrease in the quality of the copper layer. A / dm² is within the safe range.

[0130] S320: The pulse power module outputs a high current density pulse current according to the parameters of the third stage to quickly complete the sealing of the through hole.

[0131] The high duty cycle (60%) and high current density (4.0 A / dm²) significantly increased the deposition rate. A / dm² is 15 times that of the first stage ( This allows for the rapid completion of the final 20% to 30% copper deposition. The jet pressure is increased to 2.5 bar, and the oscillation frequency is 4 Hz, providing sufficient mass transfer to support the high current density copper ion consumption rate.

[0132] S330: The controller executes a termination decision, and terminates electroplating when the cumulative deposition thickness estimated by the piecewise summation of Faraday's law reaches the target thickness.

[0133] The Faraday cumulative calculation subunit of the controller continuously performs real-time cumulative calculation of the deposition thickness. The termination condition is: in Corresponding to three stages, The effective average current density for each stage (calculated based on the orifice wall area, unit: A / cm²). The duration (s) of each stage.

[0134] Area benchmark explanation: In the above formula This is the equivalent current density calculated based on the actual plated area of ​​the hole wall, not the macroscopic current density of the board surface at the power output of the equipment. For a single through hole ( mm, mm), hole wall area mm². If the equipment power supply is based on the board area. As a reference, output macroscopic current density The conversion relationship of the equivalent current density of the hole wall is as follows: ,in This is the sum of the wall areas of all through holes on the board surface. In the following exemplary calculation, the area used... The value has been converted as described above, and the borehole wall area is used as the benchmark directly.

[0135] For example, take (96%) μm: First-stage deposition thickness: μm.

[0136] Second-stage deposition thickness: μm.

[0137] Third-stage deposition thickness: μm.

[0138] total μm μm, meeting the target thickness requirement. The total electroplating time is approximately min. when When the electroplating is stopped, the controller outputs an electroplating termination signal, and the pulse power supply module stops outputting current.

[0139] Regarding the robustness of the third stage, a counterexample will be used to illustrate this below. During the second stage's execution... At time min, due to the startup of an adjacent electroplating tank causing a brief fluctuation (±5%) in the plant's power supply voltage, the actual output current of the pulse power module exhibited a transient deviation. The reference electrode array collected... mV (far exceeding) (mV), consequent_count was reset to 0. However, the voltage fluctuation recovered within 2 seconds. min time The voltage dropped back to 9 mV. Since the consecutive_count was reset, it needed to accumulate to 50 consecutive satisfactions again, delaying the phase switch by approximately 5 seconds. This transient disturbance was correctly identified as a non-steady-state event by the algorithm—the algorithm did not mistakenly trigger a phase switch or terminate electroplating due to a single fluctuation; the continuous counting requirement effectively filtered out such transient interferences.

[0140] S400: Post-processing and quality inspection.

[0141] like Figure 6 As shown, this stage involves cleaning the plated circuit boards after they have been removed from the slots and inspecting the cross-sectional quality of the through holes.

[0142] S410: The multilayer circuit board that has been electroplated is removed after being washed with three-stage countercurrent water.

[0143] Disconnect the pulse power supply module's current output and remove the electroplated multilayer circuit board from the VCP fixture. Perform a three-stage counter-current water rinse: the first stage uses deionized water (conductivity ≤1μS / cm) at 25°C for ≥30 seconds; the second and third stages progressively decrease the contamination concentration, with each stage lasting ≥30 seconds. The total rinsing time for the three stages is no less than 90 seconds, removing residual electroplating solution and additive decomposition products from the surface of the multilayer circuit board.

[0144] S420: Perform a metallographic cross-section slice on the test via hole, and measure the copper layer thickness at the hole opening, at the hole center, and at the hole bottom.

[0145] Remove the attached test board from the edge of the product board, and perform a metallographic cross-section slice along the axial direction of the via hole. Use an optical microscope (with a magnification of 200× to 500×) to measure the following key dimensions: (Copper layer thickness at the hole opening, 0.1 mm from the board surface), (Copper layer thickness at the hole center, at the midpoint of the board thickness), (Copper layer thickness at the hole bottom, 0.1 mm from the other board surface).

[0146] S430: Calculate the deep plating ability and the uniformity deviation based on the copper layer thickness at the hole opening, at the hole center, and at the hole bottom, and determine compliance.

[0147] The formula for calculating the deep plating ability is: The formula for calculating the uniformity deviation is: Where 、 and are the maximum value, the minimum value, and the average value among the three measurement points respectively.

[0148] The controller executes the compliance determination logic: When and and it is determined to be compliant. When or it is determined to be non-compliant, and the controller records the non-compliant parameters and triggers a suggestion for fine-tuning the process parameters.

[0149] Exemplarily, for a via hole with a thickness-to-diameter ratio of 12.8:1 (board thickness 3.2 mm, hole diameter 0.25 mm), the metallographic slice measurement results are: μm, μm, μm. , μm, μm, μm, . , , μm μm, it is determined to be compliant.

[0150] To verify the performance advantages of this method over existing technologies, a control group was set up for comparative testing. Control group A used traditional DC electroplating: a fixed current density of 2.0 A / dm² and a fixed additive concentration ( The control group (B) used standard forward and reverse pulse electroplating: forward pulse of 2.0 A / dm² for 50 ms and reverse pulse of 0.5 A / dm² for 10 ms, with fixed additives and standard jet agitation. Each group underwent repeated testing on at least 10 plates (each plate had 3 cross-sectional slices, for a total of at least 30 measurement points / group). The statistical significance is determined by a one-tailed t-test. A high aspect ratio through-hole in the process group ( )satisfy and And the same through holes in the control group or At that time, it was determined that this method had a performance advantage over the control group.

[0151] In addition to macroscopic indicators such as deep plating capability and uniformity deviation, the microscopic crystal quality of the electroplated copper layer is also evaluated. Scanning electron microscopy (SEM, magnification 5000× to 20000×) is used to observe the microstructure of metallographic sections, assessing the uniformity of copper grain size and the presence of abnormal crystal morphologies such as columnar or dendritic crystals. Focused ion beam (FIB) cross-sectional observation technology is used to measure the microporosity within the copper layer. Under the process parameters of this embodiment, the copper layer obtained by the three-order gradient pulse electroplating has a grain size of 0.5 to 2.0 μm, uniform grain distribution, no abnormal dendritic crystals, and a microporosity of less than 0.3%. The density of the copper layer meets the requirements for use in high-reliability electronic products.

[0152] like Figure 7 As shown, this application also provides an electroplating system for performing the above-described electroplating method. The electroplating system comprises six functional modules: a third-order gradient pulse power supply module M100, a dynamic additive control module M200, an oscillating jet stirring module M300, an electroplating solution circulation and temperature control module M400, a distributed potential monitoring module M500, and a central stage controller M600.

[0153] The third-order gradient pulse power supply module M100 is a programmable pulsed DC power supply installed in the power supply cabinet of the electroplating production line. The core hardware of this module is an IGBT power module with a drive capacity of no less than 100 A / 12 V, capable of supporting a peak current of 4.0 A / dm² under a maximum electroplating area of ​​approximately 25 dm². The power controller is based on a DSP digital signal processor. Its internal interrupt service routine generates T_on / T_off square wave signals with a 1 μs time base accuracy to drive the IGBT gate, supporting online parameter updates—after receiving a stage switching command from the central controller M600, it seamlessly switches to the new parameter set at the end of the current pulse cycle. The current feedback loop is based on a Hall effect current sensor (accuracy ±0.5%), tracking the target current in real time using a PI closed-loop method with a response time of less than 1 ms. The overcurrent / overvoltage protection circuit is composed of a hardware comparator, cutting off the IGBT within less than 10 μs. The module maintains a fixed-size array StageParams[3] as a stage parameter table, with each element containing four fields: {J_peak_mA, T_on_us, T_off_us, ramp_rate_mA_per_ms}. The M600 sends stage switching instructions via the RS-485 bus using the Modbus RTU protocol, and the M100 polls the Modbus register at the end of each pulse cycle and executes the new parameters starting in the next cycle.

[0154] The dynamic additive control module M200 is deployed in a separate control cabinet next to the circulation pipeline of the electroplating tank. This module includes three precision peristaltic metering pumps (accuracy ±0.5 mL / min) for the quantitative injection of leveling agent, accelerator, and inhibitor mother liquors, respectively; and one online CVS analyzer, which automatically samples 50 mL from the electroplating tank every 10 minutes to perform cyclic voltammetry scanning, quantifying the concentrations of accelerator and leveling agent using the SPS oxidation peak area and PEI reduction peak area, respectively, with an accuracy of ±5%. The PLC program calculates the replenishment amount and drives the metering pumps based on the CVS analysis results and the target concentration command issued by M600. A mixing delay compensation subunit performs advance compensation for the pipeline delay (approximately 30 to 60 seconds) from the injection point to the electroplating tank. The module maintains the additive state vector {C_L_measured, C_A_measured, C_S_measured, C_L_target, C_A_target, R_target, last_CVS_timestamp}. The M600 sends JSON-formatted instructions to the M200 via Ethernet TCP / IP, and the M200 reports the current concentration status every 60 seconds.

[0155] The M300 oscillating jet stirring module is installed at the bottom of the electroplating tank. The core hardware consists of a high-pressure centrifugal pump (maximum flow rate 50 L / min, maximum pressure 3 bar), a jet nozzle array plate (made of 316L stainless steel, nozzle orifice diameter 0.5 to 1.0 mm, arranged in close proximity at intervals not exceeding 2 mm to form a uniform, surface-covering jet field), and a servo motor-driven deflection mechanism. The high-pressure pump's speed is adjusted via frequency conversion, and a pressure sensor at the outlet performs PID closed-loop control, ensuring pressure fluctuations do not exceed ±0.1 bar. The deflection mechanism, driven by a servo motor through a crank-connecting rod mechanism, drives the nozzle array plate to perform sinusoidal reciprocating deflection with a deflection angle of ±15° and a frequency range of 1 to 5 Hz. Synchronization triggering uses hardware wiring—the T_off synchronization pulse of the M100 is directly connected to the start relay of the M300, ensuring the jet only operates during the pulse off-peak period. The module maintains the jet parameter register {P_target_bar, f_osc_Hz, amplitude_deg, sync_mode}. The M600 sends parameters to the frequency converter and servo drive via Modbus RTU.

[0156] The electroplating solution circulation and temperature control module M400 is deployed in the circulation pipeline system outside the electroplating tank. The core hardware includes a magnetically driven circulation pump (flow rate 200 to 300 L / h, acid-resistant PVDF material), a plate titanium heat exchanger (heating power 6 kW / cooling power 4 kW), a dual-cylinder switching 1 μm precision activated carbon filter group, and an online sensor group. The temperature control subunit uses a Pt100 probe (accuracy ±0.1°C) and a PID controller to stabilize the temperature at 23±1°C, with a temperature control response time of less than 30 seconds. The filtration circulation subunit operates in series, with a circulation flow rate of 4 to 6 times the tank volume per hour. A replacement alarm is triggered when the filter element pressure difference exceeds 0.5 bar. The Cu²⁺ concentration online monitoring subunit is based on UV-Vis colorimetry (wavelength 810 nm), sampling and analyzing every 30 minutes, with a detection range of 40 to 70 g / L and an accuracy of ±2 g / L. The pH monitoring subunit uses an online glass pH electrode, collecting data once per minute to maintain the pH within the range of 0.3 to 0.8. The module maintains the tank solution status register {T_current_C, T_setpoint_C, pH_current, Cu2_conc_gL, filter_dp_bar, pump_flow_Lh, last_update_timestamp}. The M400 reports its status to the M600 every 60 seconds via Modbus RTU, and the M600 can modify the temperature setpoint by writing to the register.

[0157] The distributed potential monitoring module M500 consists of an array of N (exemplarily N = 5) miniature Ag / AgCl reference electrodes, fixed to the anode frame via a flexible PCB carrier, and evenly spaced along the thickness direction of the circuit board to be plated. The effective sensing area of ​​each reference electrode does not exceed 1 square millimeter. Each channel is equipped with an instrumentation amplifier to amplify the millivolt-level potential signal to the range of 0 to 5 V. A 16-bit precision, 100 Hz sampling rate multi-channel ADC module synchronously acquires data from all channels and transmits it to the M600 via the SPI bus. Output potential snapshot data packets {timestamp_ms, channel_id[1..N], potential_mV[1..N], bath_temp_C, pH_value} are generated every 10 ms.

[0158] The central stage controller M600 is an industrial control computer deployed in the control cabinet of the electroplating production line, running a Linux real-time operating system and equipped with an Intel Core i5 processor and 8 GB of RAM. This controller is the communication hub and decision-making core of the entire system, containing five functional sub-units. The data acquisition and preprocessing sub-unit receives potential snapshots from M500 and performs a 10-point moving average filter to generate a smoothed overpotential sequence. The stage determination logic sub-unit executes the stage switching determination algorithms in S140 and S240, maintaining the state machine {INIT → STAGE1 → STAGE2 → STAGE3 → COMPLETE}. The parameter mapping sub-unit stores a three-stage parameter mapping table, reading the corresponding parameter set from the table and distributing it to M100 / M200 / M300 during stage switching. The Faraday cumulative calculation sub-unit calculates the cumulative deposition thickness in real time based on the actual output current fed back from M100. The HMI (Human-Machine Interface) sub-unit displays parameters such as the real-time potential distribution curve, current stage, cumulative thickness, and additive concentration via a touchscreen. This controller also maintains the electroplating process state tensor (dimensions). (This is used for post-process analysis.) The M600 communicates with the M100 / M300 via RS-485 Modbus RTU, with the M200 via Ethernet TCP / IP, and with the M500 via SPI bus. All communication is completed within a 100 ms master control cycle.

[0159] like Figure 8 As shown, in a complete electroplating process, the three-order gradient pulse waveform and the additive concentration ratio exhibit a time-series correlation: the first stage (low current density, high concentration ratio, low jet pressure) → the second stage (medium current density, medium concentration ratio, medium jet pressure) → the third stage (high current density, low concentration ratio, high jet pressure). The parameters in the three dimensions are independently adjustable and work together in the time domain to achieve a smooth transition from "uniformity priority" to "efficiency priority".

[0160] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the above embodiments, or equivalent substitutions can be made to some of the technical features; these modifications and substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0161] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware.

[0162] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for electroplating high aspect ratio through holes in a multilayer circuit board, characterized in that, include: The acidic copper sulfate electroplating solution containing leveling agent, accelerator and inhibitor is prepared to the target concentration of the first stage, and the multilayer circuit board to be plated is loaded into the electroplating tank. The first stage of electroplating is performed by outputting pulse current with the first peak current density and the first duty cycle. During the pulse turn-off period, jet stirring is performed on the through hole. At the same time, the local overpotential at different depths of the through hole is collected by the auxiliary reference electrode array. When the overpotential difference between the hole opening and the hole center continuously meets the first threshold condition, the first stage of electroplating is determined to be completed. The concentration ratio of leveling agent to accelerator is adjusted from the first ratio to the second ratio. The second stage of electroplating is performed with the second peak current density and the second duty cycle. When the maximum value of the overpotential difference between adjacent reference electrodes continuously meets the second threshold condition and the cumulative deposition thickness reaches the preset ratio of the target thickness, the second stage of electroplating is determined to be completed. The concentration ratio is adjusted from the second ratio to the third ratio, and the third stage of electroplating is performed with the third peak current density and the third duty cycle. Electroplating is terminated when the cumulative deposition thickness reaches the target thickness. The peak current density and duty cycle increase progressively from the first to the third stage, while the concentration ratio decreases progressively from the first to the third stage.

2. The method according to claim 1, characterized in that, The first peak current density is 0.5 to 1.0 A / dm², the first duty cycle is 15% to 25%, the second peak current density is 1.5 to 2.5 A / dm², the second duty cycle is 35% to 45%, the third peak current density is 3.0 to 5.0 A / dm², and the third duty cycle is 55% to 65%.

3. The method according to claim 1, characterized in that, The first ratio is 1.5 to 3.0, the second ratio is 0.5 to 1.0, and the third ratio is 0.2 to 0.

5.

4. The method according to claim 1, characterized in that, The leveling agent is a polyethyleneimine compound with a molecular weight of 10,000 to 30,000, the accelerator is sodium polydisulfide dipropane sulfonate, and the inhibitor is polyethylene glycol with a molecular weight of 4,000 to 8,000. The inhibitor is maintained at a constant concentration throughout all three stages.

5. The method according to claim 1, characterized in that, In the step of adjusting the concentration ratio of leveling agent to accelerator in the electroplating solution from a first ratio to a second ratio: The actual concentration of the leveling agent is detected by a cyclic voltammetric stripping online analyzer. When the actual concentration is higher than the target concentration of the leveling agent in the second stage, the leveling agent supply pump is stopped. When the actual concentration is lower than 90% of the target concentration of the leveling agent, the leveling agent supply pump is resumed. The accelerator concentration is adjusted to the target accelerator concentration for the second stage according to the accelerator replenishment formula. The accelerator replenishment formula is: the replenishment volume is equal to the difference between the target concentration and the current concentration multiplied by the tank liquid volume and then divided by the mother liquor concentration.

6. The method according to claim 1, characterized in that, The jet stirring is driven by a servo motor to deflect the jet nozzle array in a sinusoidal manner with a deflection amplitude of ±15°. The pressure and oscillation frequency of the jet stirring increase with each stage, and the jet stirring is triggered synchronously with the off-period of the pulse current.

7. The method according to claim 1, characterized in that, The first threshold condition is that the overpotential difference between the orifice and the center of the orifice does not exceed 15 mV for 30 consecutive pulse cycles. The second threshold condition is that the maximum overpotential difference between adjacent reference electrodes does not exceed 10 mV for 50 consecutive pulse cycles.

8. The method according to claim 1, characterized in that, The first stage electroplating and the second stage electroplating are respectively equipped with timeout protection. When the duration of the first stage electroplating exceeds the first maximum duration, it is forcibly switched to the second stage electroplating. When the duration of the second stage electroplating exceeds the second maximum duration, it is forcibly switched to the third stage electroplating.

9. The method according to claim 1, characterized in that, The method also includes post-processing and quality inspection steps: The multilayer circuit board that has been electroplated is removed after being rinsed with three stages of countercurrent water. Metallographic cross-sections were performed on the test through-holes to measure the copper layer thickness at the hole opening, the center of the hole, and the bottom of the hole. The deep plating capability and uniformity deviation are calculated based on the copper layer thickness at the orifice, the copper layer thickness at the center of the orifice, and the minimum copper thickness is not less than the target thickness when the deep plating capability is not less than 80%, the uniformity deviation is not more than 15%, and the minimum copper thickness is not less than the target thickness.

10. The method according to claim 1, characterized in that, The auxiliary reference electrode array includes multiple micro silver-silver chloride reference electrodes arranged at equal intervals along the thickness direction of the multilayer circuit board to be plated. The effective sensing area of ​​each reference electrode does not exceed 1 square millimeter. The collected overpotential data is amplified by high impedance differential and then synchronously sampled by a multi-channel analog-to-digital converter.

Images