Patina Uniformity on Copper Alloys with Sodium Selenite

Uniformity of selenite-derived patina on copper alloys is governed by the spatial balance between anodic copper dissolution and cathodic reduction of dissolved Se⁴⁺ species. Sodium selenite, Na₂SeO₃, is freely soluble in water and, when acidified with sulfuric acid to a pH between 2.0 and 3.5, yields a mixture of HSeO₃⁻ and H₂SeO₃. The cathodic half-reaction H₂SeO₃ + 4H⁺ + 4e⁻ → Se + 3H₂O has a standard potential of approximately +0.74 V versus the standard hydrogen electrode, while the anodic half-reaction Cu → Cu²⁺ + 2e⁻ has a standard potential of +0.34 V. The resulting mixed potential therefore drives spontaneous film formation without an external rectifier; however, kinetic factors—mass transport, nucleation density, local pH shift, and substrate microstructure—rather than thermodynamic feasibility determine whether the final selenium-containing film is optically uniform. The deposit is typically a mixed copper selenide containing Cu₂Se and CuSe, with the proportion of elemental selenium increasing as pH decreases. On production-scale lines using polypropylene dip tanks of 500–1500 L capacity, the principal uniformity defects are edge build-up, part-to-part shade variation, gas-pit speckling, and light patches at drainage points. Edge build-up occurs because protruding edges and corners experience enhanced convective transport of fresh selenite solution relative to recessed surfaces; the local current density at those sites becomes disproportionately high. Gas-pit speckling occurs when hydrogen bubbles or air bubbles nucleate, adhere to the surface, and mask the area from solution contact. These failure modes are managed through solution agitation, surfactant addition, controlled rack geometry, and pH buffering. Sodium selenite baths are immersion processes, not electroplating processes, so the anode-to-cathode rack ratios used in rectifier-driven operations do not directly apply; instead, bath loading is expressed as total exposed metal area per unit volume, commonly 0.5–1.5 m² per 100 L. The consumption of selenite is progressive and proportional to the square metres processed; copper dissolution products accumulate and can shift the mixed potential, so replenishment must be based on analytical selenium and copper concentrations rather than elapsed time. Operational boundaries include the exclusion of chloride contamination because chloride complexes copper and alters the anodic Tafel slope; even 50 mg/L of chloride can produce local pitting and non-uniform film morphology. The bath is also incompatible with strong reducing agents because acidified selenite solutions can be reduced to volatile hydrogen selenide under appropriate redox conditions. Ventilation across the tank face must be maintained at 0.5–1.0 m/s capture velocity when the bath is heated above 35°C.

How does solution pH shift the selenite reduction pathway on wrought copper?

Below pH 2.0, the selenite bath becomes aggressive toward wrought copper, and the deposit is dark but coarse with a higher elemental selenium content. On C11000 electrolytic tough pitch copper, uniform black coverage can be obtained in 1–2 min at pH 1.8–2.0, but the film has a powdery surface and adhesion may be compromised because rapid copper dissolution undercuts the forming selenide layer. Between pH 2.4 and 2.8 at 25–35°C with 10–20 g/L Na₂SeO₃, the deposition rate moderates to a brown-to-black film of 200–500 nm thickness within 2–5 min. This is the preferred window for C11000 because the HSeO₃⁻/H₂SeO₃ ratio provides a balance between nucleation density and continuous film coalescence. Above pH 3.5, the reaction rate slows sharply; films become translucent reddish-brown and non-uniform because coverage is incomplete, and the bath may precipitate insoluble copper selenite or basic copper sulfate if copper concentration exceeds 15 g/L. pH must therefore be controlled to within ±0.2 of the setpoint. Acid consumption is inherent because each mole of selenite reduced to copper selenide consumes protons; on a 750 L tank processing 25 m² of copper parts per shift, pH drift of +0.5 unit can occur within 4 h unless dilute sulfuric acid or selenious acid is metered by peristaltic pump. Continuous pH measurement with an inline flat-surface electrode is preferred to immersion probes because suspended selenide precipitates coat the glass membrane. When automatic dosing is not used, a manual adjustment schedule based on a calibrated benchtop pH meter per ASTM E70 is the minimum control. Temperature also interacts with pH: a 10°C increase raises the reduction rate by a factor of approximately 2–3, so a bath operated at 35°C requires a lower pH setpoint than one operated at 25°C to achieve equivalent coverage. The uniformity of the film is generally better at lower temperature and longer time because nucleation density remains high relative to vertical film growth. In practice, HVAC heat losses from uninsulated tank walls produce cold spots and localized slower deposition, so tank insulation and calibrated immersion heaters are necessary.

Adhesion and visual uniformity of the selenide film are controlled by the cleanliness and micro-roughening of the copper alloy before immersion. A water-break-free surface after cleaning per ASTM B322-14 is the minimum acceptance criterion; any residual organic film, drawing lubricant, or oxide scale creates a non-uniform electrochemical potential across the surface. Alkaline soak cleaning is typically performed at 50–65°C with a silicated, non-etching cleaner, followed by an anodic electroclean at 4–6 A/dm² and a double cascade rinse with conductivity below 10 µS/cm. Acid activation for C11000 uses 5–15 vol% sulfuric acid at 20–30°C for 30–120 s, which removes residual oxide and establishes a uniform active surface. For brass alloys, nitric acid activators are not recommended because preferential zinc dissolution produces a smutty copper surface and dezincification zones that appear as red patches after selenite patination. Sulfamic acid or dilute sulfuric acid with a small addition of copper sulfate is used instead. Parts with brazed joints or solder must be masked or isolated because solder dissolution shifts the local mixed potential, producing a light halo around the joint. After activation, parts should be transferred to the selenite bath within 30 s to avoid drying stains and re-oxidation. If relative humidity exceeds 60%, water film condensation on cold parts may dilute the acid film and produce uneven initiation; forced air drying of the transfer area is recommended. The final rinse after patination must also be deionized water with conductivity below 10 µS/cm, followed by drying at 60–80°C for 10–20 min. In high-volume lines, a warm rinse at 40–50°C reduces drag-out viscosity and improves drainage from blind holes.

When C26000 and C27000 brass substrates are processed in acidic selenite solutions

Selective zinc dissolution is the dominant risk when C26000 cartridge brass (70 wt% Cu, 30 wt% Zn) or C27000 yellow brass (65 wt% Cu, 35 wt% Zn) is patinated in a sodium selenite bath. At pH values below 2.0, zinc-rich phases corrode preferentially, leaving behind a copper-enriched surface layer. That layer reacts with selenite at a different rate than the surrounding brass, producing a non-uniform brown-black film over a rosy copper background. Cross-sectional scanning electron microscopy of C26000 panels processed at pH 2.0 for 5 min commonly shows dezincification depths greater than 1 µm, whereas panels processed at pH 3.0–3.5 under the same conditions show no measurable dezincification layer below the film. The mixed potential shifts with zinc dissolution; copper ion accumulation in the diffusion layer accelerates selenite reduction on the copper-rich regions and starves the adjacent zinc-rich regions of selenite. To suppress this, process baths for brass are maintained at pH 3.0–3.5, and copper sulfate pentahydrate is added at 2–5 g/L to reduce the driving force for further copper dissolution. Total immersion time for brass should not exceed 5 min at 30°C; longer cycles amplify the differential between phases. Immediate rinsing within 10 s after withdrawal is necessary to stop acid attack. Parts are immersed with the final visible surface oriented upward to avoid droplets draining across the face during withdrawal; drainage patterns are a frequent source of longitudinal streaks. C27000 is more susceptible to dezincification than C26000 and should be processed in the upper half of the pH range with a 1–2 min shorter immersion than C26000. The use of small additions of benzotriazole at 0.1–0.5 g/L has been reported in some formulations to reduce zinc dissolution, but published data for this specific configuration is limited; if used, it must be validated for bath life because benzotriazole can accumulate and inhibit selenite reduction entirely at higher concentrations.

On C51000 and C52100 phosphor bronze, the tin content of 4–8 wt% and 1–2 wt% respectively introduces a tin oxide-rich surface that resists the initial acid activation step. Immersion in 10 vol% sulfuric acid at 25°C may not completely remove the passive tin oxide film, resulting in delayed initiation in the selenite bath and a patchy appearance after 2 min. A pre-dip in 10–15 vol% fluoroboric acid or a sulfuric acid–persulfate activator at 20–30°C is used to micro-roughen the tin-rich phase and produce uniform nucleation. Process time in the selenite bath must be extended by 30–90 s relative to C11000 to achieve equivalent darkening, and the resulting film is usually thinner and more brown-black. For C70600 and C71500 cupronickel alloys containing 10–30 wt% nickel, the passive nickel oxide layer requires aggressive activation; a fluoroboric acid dip or a Wood’s nickel strike of 1–3 µm may be necessary before the selenite patina will adhere. Without such activation, the sodium selenite bath produces a non-uniform grey film that can be wiped off after drying. In mixed-metal assemblies, galvanic coupling between the alloy and stainless steel racking or fasteners creates local cells that may deposit selenium preferentially on the more noble surface, leaving shade differences; insulated rack contacts and titanium or polymer-clad fixtures are required to prevent unintended galvanic effects. Visual evaluation under D65 illuminant per ASTM D1729-16 is used for rapid screening, but panel acceptance is tied to measured ΔE*ab rather than relying solely on visual inspection. Published data for the exact numeric shade response of each wrought alloy is limited, and process validation on each alloy lot should include reflectivity and color difference measurements rather than relying only on operator judgement.

Analytical control methods and surface metrology

For quantitative release testing of architectural components, color difference is measured with an integrating sphere spectrophotometer per ASTM D2244-21 using D65 illumination and the 10° standard observer. Uniformity is considered under control when the maximum ΔE*ab across five defined points on a 100 mm × 100 mm panel is less than or equal to 2.0; for architectural door hardware, supplier-specific limits may be tighter at 1.5. Gloss variation is measured with a 60° glossmeter per ISO 2813:2014, and variation greater than 5 gloss units across the panel indicates texture or thickness non-uniformity that may not be captured by color alone. Scratch adhesion screening is performed by cross-hatch tape pull per ASTM D3359-17, method B, with a 1 mm spacing cutter; classification 5B or 4B is typical for comparative screening, but cohesive failure within the patina film can occur and must be distinguished from adhesive failure at the substrate. For referee measurements of film thickness below 500 nm, cross-sectional scanning electron microscopy remains the most reliable method because X-ray fluorescence thickness calibration for Cu₂Se films is not universally available. Selenium-to-copper atomic ratios are determined by energy-dispersive X-ray spectroscopy at 20 kV accelerating voltage; ratios between 0.2 and 0.5 are typical for uniform black films on copper, but the exact ratio varies with bath pH and age. X-ray photoelectron spectroscopy is used to determine the selenium oxidation state when process troubleshooting requires distinction between Cu₂Se, CuSe, and elemental Se. The table below summarizes the analytical control matrix used for process qualification and routine batch release.

In production tanks, racking density, agitation, and bath ageing are interconnected variables that determine whether a technically acceptable ΔE*ab can be maintained over a full shift. Parts are racked with 5–10 cm minimum spacing along the vertical axis and 2–3 cm minimum spacing between adjacent parts; the assembly is tilted 15–20° from vertical so that trapped air can escape from recesses. For flat decorative panels, vertical orientation with the visible face upward reduces drainage streaking. Air sparging from a CPVC or PVDF manifold at the bottom of the tank delivers oil-free compressed air at 0.5–1.5 L/min per m² of bath surface; this creates bulk turnover without the high shear of eductors. If the sparge rate is too low, a stagnant boundary layer forms and the patina is thin in the center of large panels; if too high, air bubbles collect in concave details and create unpatinated specks. The use of non-ionic wetting agents at 0.1–0.5 mL/L lowers surface tension and improves bubble detachment, but excessive foaming may necessitate defoamer, which can interfere with film formation. Bath temperature is controlled to ±2°C using titanium immersion heaters and a process thermostat; quartz heaters are acceptable, but stainless steel immersion heaters are not used because selenium species can corrode stainless steel at low pH. The bath is filtered through 1 µm polypropylene wound cartridges at 2–3 turnovers/h to remove suspended particulates. Suspended selenide sludge will settle on upward-facing surfaces and produce rough black spots if not removed. Bath contamination by chloride should be kept below 50 mg/L; chloride is introduced through tap water or poorly rinsed acid tanks and causes pitting. When bath loading exceeds 1.5 m² per 100 L per shift, the depletion of selenite and accumulation of copper become too rapid for manual replenishment and uniformity drifts; either decrease loading or install automatic analytical dosing.

Repeated immersion cycles thicken the film but do not linearly improve uniformity

After the first immersion cycle, a continuous selenide film of 200–500 nm on C11000 has typically reached full optical coverage. A second immersion in freshly replenished solution can increase total thickness to 800–1200 nm, but macro color variation measured as ΔE*ab may not improve proportionally. Thicker films reduce total reflectance, particularly in the 400–700 nm range, but edge build-up is amplified because the existing conductive selenide film provides a ready cathodic surface for further selenite reduction; edges therefore thicken faster than recessed areas. For cosmetic architectural parts, a single immersion per side is often preferred to avoid this non-uniformity. If a second cycle is required to darken the film, the part must be rinsed thoroughly in deionized water with conductivity below 10 µS/cm and inspected for drainage stains before re-immersion. Each return to the bath introduces fresh solution into porous film regions and can dissolve loosely adhering selenide particles; these particles then re-deposit as sludge on the surface and become embedded in the growing film. On brass, repeated cycles increase cumulative dezincification risk and should not exceed a total immersion time of 5 min at 30°C unless cross-sectional evaluation verifies that the dezincification layer remains below 0.5 µm. On C51000, a second short cycle of 60–120 s may improve colour depth without significant adhesion loss. Oven drying at 60–80°C after the final rinse stabilizes the film and prevents water spotting; drying above 100°C can oxidize the selenide film and shift the hue toward brown or grey. The final decision to use repeated cycles should be based on measured ΔE*ab, gloss, and SEM thickness, not on visual darkness alone.

Selenium compound waste streams and airborne emissions require engineering controls because sodium selenite is toxic by ingestion and inhalation and is classified under Regulation (EC) No 1272/2008 as Acute Tox. 3 H301 and H331, with chronic target-organ effects under STOT RE 1 H372. The 0.2 mg/m³ 8-hour time-weighted average for selenium compounds as selenium under 29 CFR 1910.1000 Table Z-1 and the equivalent NIOSH REL is the maximum exposure concentration for production staff; local exhaust ventilation must provide 0.5–1.0 m/s capture velocity across the tank face and at the drying station. Acid mist from the bath is controlled by a packed scrubber or mist eliminator because selenious acid droplets can be inhaled. Wastewater from rinsing operations contains selenite and must not be discharged to storm sewers. Selenite is not efficiently removed by simple hydroxide precipitation at pH 9–10; ferric sulfate co-precipitation at pH 5–6 or zero-valent iron media can reduce selenium to below 0.05 mg/L, the US EPA National Primary Drinking Water Regulation maximum contaminant level under 40 CFR 141.62. Spills of the sodium selenite salt or concentrated solution are collected with a dedicated wet vacuum or moistened absorbent and disposed as hazardous waste; dry sweeping is prohibited because it creates inhalable dust. The material must be stored separately from strong reducing agents, metal powders, and concentrated acids to prevent uncontrolled reduction and possible hydrogen selenide generation. In the event of fire, dry chemical or carbon dioxide extinguishers are used; water spray may spread soluble selenite. Process operators must use nitrile gloves, chemical goggles, and acid-resistant aprons; respiratory protection with NIOSH-approved cartridges for acid gas and particulates is required if engineering controls cannot maintain exposure below the OEL. Rinse water and spent baths must be collected in dedicated holding tanks and treated or shipped to a licensed facility. Discharge permits typically require daily composite sampling by EPA Method 200.7 or equivalent; local publicly owned treatment works may impose more stringent selenium limits than the federal standard, and the permit limit governs.

For blind-hole or recessed geometries, such as architectural escutcheon plates or optical baffles with threaded holes, air bubble entrapment is a dominant cause of unpatinated spots. Holes smaller than 3 mm diameter retain air because surface tension exceeds buoyancy; pre-wetting with a surfactant solution or plugging with removable polymer plugs is used. Parts are immersed slowly at 10–20 cm/s vertical speed to allow air to escape. After withdrawal, tilting and rotation during drainage prevents solution from pooling in recesses and creating edge-weep stains. In optical applications using C11000 where low reflectance below 5% at 550 nm is required, the bath is operated at the lower end of the concentration range, 10–12 g/L, and the film is polished with a soft felt wheel after patination to remove loosely adherent particles; reflectance is verified with a spectrophotometer per ASTM E903-20. For soldered or brazed assemblies, the galvanic coupling between solder and copper alters the local deposition rate; a copper electroplated strike over the solder of 3–5 µm is applied before patination to homogenize the surface potential. For architectural bronze door hardware, the sealed patina is tested for corrosion resistance under neutral salt spray ISO 9227:2022 for 96 h; acceptance criteria are typically defined by the end-user specification rather than by the patination process standard, and published data for this specific configuration is limited. Across all configurations, process capability studies should collect at least 30 consecutive lots of ΔE*ab, gloss, and thickness data to establish upper and lower control limits per ISO 7870-2. The process is considered stable when no more than one data point in each lot exceeds the specified ΔE*ab limit and no systematic upward trend in copper concentration is observed.