Metal-Compatible Stripper Electronic/EL Grade is a water-miscible organic stripper released under the EL-2000 designation for front-end and packaging photoresist removal where Cu, Al, TiW, TiN, and Ni/Au interfaces must be preserved. It is supplied as a clear, pale-yellow liquid in 1 L, 2.5 L, and 5 L high-density polyethylene containers with a nitrogen headspace after passage through a 0.22 µm membrane. The product removes positive-tone and negative-tone photoresists, polyimide residues, and post-etch debris in single-wafer spray processors, re-circulating immersion tanks, and ultrasonic-assisted wet benches. Working temperature is normally between 50 °C and 65 °C. The lower bound is set by the dissolution rate of novolak and acrylic resists, while the upper bound is constrained by retention of the Cu corrosion inhibitor. The stripper is not formulated for solvent exchange in dry etching, and should not be applied to unprotected silicon areas where alkaline attack could occur.

What Analytical Limits Separate Electronic/EL Grade Strippers from Technical-Grade Solvent Blends?

Electronic-grade classification is not defined solely by visual clarity. It requires lot-specific release limits for mobile-ion contamination and non-volatile residue, because stripper residues left after die attach, wire bonding, or underfill can create leakage paths and corrosion cells. Table 1 lists the release matrix. Each trace metal cation is measured by inductively coupled plasma mass spectrometry after acid evaporation of a 100 mL sample; this technique provides detection limits below 1 ppb for the monitored elements. Anion control is critical because chloride residues accelerate pad corrosion; chloride is measured by ion chromatography using a sodium carbonate/bicarbonate eluent and suppressed conductivity detection. Particle control is equally important because particles greater than 0.5 µm can block spin-rinse dryers and contaminate exposed bond pads. Water content is measured by coulometric Karl Fischer titration rather than volumetric titration to improve precision in the 0.05–0.3 wt% range. Viscosity and density are controlled to ensure consistent dispense behaviour in spray tools; batch-to-batch variation beyond the specified window alters droplet size and can change clean time.

When Copper Pillars and Aluminium Bond Pads Are Simultaneously Exposed During Resist Removal

Metal compatibility is qualified on blanket Cu and Al films before the stripper is released for use on bumping lines. A 30 min immersion at 25 °C in undiluted EL Grade must produce no more than 2 Å/min Cu thickness loss and no more than 1 Å/min Al thickness loss. The Cu measurement is made by four-point probe sheet-resistance shift, converted to thickness using bulk copper resistivity, and cross-checked by X-ray fluorescence on 200 mm or 300 mm blanket wafers. For Al, stylus profilometry on a lithographically defined step is used to resolve sub-10 Å changes. The specification is tighter than conventional hydroxylamine-based strippers because the EL Grade contains no free fluoride and no free hydroxylamine; the corrosion inhibitor is a nitrogen-containing heterocyclic compound that adsorbs preferentially on Cu surfaces at the process pH. In single-wafer spray processing, a dispense temperature of 60 °C and total contact time of 5–15 min remove 2.5–10 µm positive-tone photoresist from wafer-level packages. The tool should use a fluoropolymer fluid path and a 0.1 µm point-of-use filter. After stripping, a deionised water rinse of 60–90 s at 800–1,200 rpm removes solubilised resist and prevents redeposition on exposed Cu pillars. If a negative-tone resist or polyimide residue is present, the contact time is extended to 15–20 min and the dispense temperature is raised to 65 °C, but the metal etch-rate specification still applies.

Re-circulating immersion baths are typically operated with a fluoropolymer or quartz vessel and a filtration loop that turns the bath over 4–6 times per minute through a 0.1 µm polytetrafluoroethylene membrane. Dissolved photoresist, not metal accumulation, is the principal bath-life variable. Replacement or replenishment is initiated when dissolved resist solids reach 10 g/L or when water content exceeds 5 wt%, whichever occurs first. Water uptake above 5 wt% is known to lower the effective pH at the metal surface and can shift Al etch rate toward the release limit. Moisture is controlled by sealing the bath lid and blanketing the headspace with dry nitrogen when the tool is idle. Ultrasonic-assisted stripping may be used only at frequencies above 40 kHz to avoid cavitation damage on wire bonds and thin passivation layers. Bath temperature should be ramped at no more than 2 °C/min during initial heat-up to avoid localised temperature overshoot and inhibitor degradation.

After reactive ion etching of low-k dielectric or TiN hardmask, the residue contains organometallic species and fluoropolymers. The EL Grade is applied at 65 °C for 10–15 min in a single-wafer tool to swell the residue, followed by a short 0.2% dilute citric acid rinse to remove metal oxides. The stripper alone does not remove TiF₄ or NiF₂ residues; an intermediate dilute acid step is required. This is a critical process integration point: the acid rinse must be applied after the alkaline strip and before drying, because the acid can protonate the buffer and reduce the effectiveness of the corrosion inhibitor if mixed in the same bath. Rinse delay should not exceed 2 min to prevent residue curing under ambient moisture or ultraviolet exposure.

Resist Loading, Bath Contamination, and Corrosion Inhibitor Depletion

The operating window is defined by two competing failure modes: low temperatures slow resist removal, but high temperatures accelerate inhibitor desorption from Cu. At 50 °C, the dissolution rate of a heavily cross-linked positive resist is approximately 30–40% of the rate at 60 °C; at 65 °C, electrochemical linear polarization measurements on blanket Cu show a lower polarization resistance, indicating that the adsorbed inhibitor film begins to thin. The upper process limit is therefore set at 65 °C, not at the solvent boiling point. In immersion tools, inhibitor depletion occurs by adsorption onto suspended Cu particles and by reaction with oxidised resist byproducts. A production-scale bath with high Cu-particle loading may require more frequent replenishment than the dissolved-solids number alone would predict. Published data for this specific formulation under all production geometries are limited; bath-life validation should be conducted with the actual resist stack and metal area ratio. On a single-wafer platform, the use of fresh chemical at each wafer minimises inhibitor depletion and provides more stable metal etch-rate performance than immersion processing. The upper limit of dissolved resist solids is also relevant in spray tools because saturated droplets can dry on wafer edges and create post-rinse defects.

Post-strip rinse compatibility is controlled by the match between stripper pH and rinse pH. When the alkaline stripper contacts a heated wafer and begins to dry, polymerised residue can redeposit on metal surfaces. A two-stage rinse is recommended: a warm deionised water rinse at 50 °C for 60 s, followed by a room-temperature deionised water rinse for 60 s. The warm first rinse reduces viscosity of the stripper film and prevents thermal shock. Rinse water resistivity should be maintained above 18 MΩ·cm; cation loading from the rinse can otherwise exchange with the adsorbed inhibitor and increase Cu attack during the subsequent dry. On batch immersion tools, cascade overflow rinse tanks are preferred over static rinse tanks because they reduce the tank concentration of stripped residue from 10 g/L to <1 g/L within two tank turnovers.

Relative to earlier generation stripper formulations, the principal differences are the absence of fluoride, the controlled alkaline pH, and the electronic-grade trace-metal release. Fluoride-free chemistry reduces attack on SiO₂ and SiN passivation layers; a fluoride-containing stripper at pH above 10 can etch thermal oxide at several angstroms per minute, whereas the EL Grade exhibits no measurable oxide loss over 30 min at 60 °C by spectroscopic ellipsometry. The pH window is narrow enough to preserve Al bond pads but sufficiently alkaline to cleave novolak-based positive resists. Unlike solvent-only blends that lose stripping activity when moisture content rises, the EL Grade retains buffering capacity through moderate water uptake. Table 2 summarises the differences against two common product classes.

Operational boundaries apply to the EL Grade. The stripper should not be mixed with hydrogen peroxide, oxidising acids, or persulfate solutions because exothermic reactions may occur and the corrosion inhibitor may be destroyed. Mixing with amine-based strippers is not recommended because the pH shift and surfactant incompatibility can cause phase separation. The product is not intended for vapour degreasing; its high boiling point and inhibitor package are not suitable for distillation recovery. Store in the original sealed container at 5–30 °C; shelf life is 12 months from the date of manufacture. Before use, confirm that the process tool seals and elastomers are compatible with polar aprotic solvents and alkaline aqueous blends; nitrile and ethylene-propylene-diene monomer seals may swell or lose compression set over time. Rinse residues from the wafer within 2 min after stripping to avoid redeposition of stripped organic matter. When disposing of spent bath material, segregate it from strong acid waste streams to prevent heat generation and gas evolution.