Anodizing Bath Control: Free Acid, Dissolved Aluminum and Temperature
Table of Contents
Controlling a sulfuric acid anodizing bath means holding free sulfuric acid, dissolved aluminum and temperature inside the ranges your process control document (PCD) sets. Run each load at a set current density (or voltage ramp) and time, and titrate the acid and aluminum on a fixed schedule so that drift is found and corrected before it shows on parts.
Practitioner guidance for Type II is 15 to 20% sulfuric acid by weight at 68 to 80 °F (20 to 27 °C). The same guidance gives 5 to 15 g/L dissolved aluminum as the range generally used in the US for both Type II and Type III. For Type III hard anodizing, a general range cited in the hard anodizing literature is −10 to +5 °C with 165 to 225 g/L free acid, though shops run outside it. None of these figures is a requirement: the contract, the governing specification and the PCD set the limits. For US military work, MIL-PRF-8625F requires the PCD to state them.
📊 What Does Each Bath Variable Do to the Oxide?
Free acid, temperature, current density and time each change the oxide’s thickness or hardness; dissolved aluminum acts mainly through the bath, on burning risk, conductivity and deposits. Acid and temperature act through two opposing effects, oxide dissolution and electrolyte resistance. Their net effect changes direction with alloy, acid level and current density, so hold the PCD ranges rather than extrapolating from any one study (see What the Research Shows).
Anodizing Bath Control Reference
Ranges are labeled by source; practitioner guidance is from Larry Chesterfield's Products Finishing columns. Current density is in amperes per square foot (asf) or per square decimeter (A/dm²). None is a requirement; the contract, the governing specification and the PCD set the limits.
| Variable | What it controls | Reference range (source) | How to measure or verify | What to investigate when it moves |
|---|---|---|---|---|
| Free sulfuric acid | Conductivity; how fast the oxide redissolves | Type II: 15–20% by weight, about 165–228 g/L at 20 °C (practitioner guidance). Type III: 165–225 g/L (general literature range; shops run outside it) | NaOH titration (potassium fluoride, methyl orange or pH endpoint methods), not a pH meter reading | Drag-out or dilution; rising dissolved aluminum; additions; a suspect titration (repeat it) |
| Dissolved aluminum | Indirect: too little raises the risk of burning; high levels lower conductivity and, above about 25–30 g/L, deposit aluminum sulfate | 5–15 g/L generally used in the US for Type II and III, 8–12 g/L for closer control; some specifications call for less than 5 g/L (practitioner guidance) | Two endpoint NaOH titration; aluminum by difference | Upward trend against the PCD limit; decant or ion exchange as the PCD specifies |
| Temperature | Oxide dissolution and electrolyte resistance; hard anodizing runs cold for a hard oxide | Type II: 68–80 °F (20–27 °C) (practitioner guidance). Type III: −10 to +5 °C (general literature range) | The tank instrument the PCD names, with its reading location and calibration interval | Cooling capacity against rectifier heat load; agitation; probe location relative to the work; calibration |
| Current density (or voltage ramp) | Build rate; the startup ramp sets uniformity | Type II: 10–24 asf (about 1.1–2.6 A/dm²) (practitioner guidance). Type III: set in the PCD by alloy and part geometry | Rectifier current ÷ load surface area; ramp time; per-load record | Burning, thin or uneven coating; contact size; load area estimate |
| Time | Thickness at a given current density | Type II: usually no more than 60 min (practitioner guidance); otherwise from the Rule of 720 and required thickness | Coating weight or thickness on lot samples and process control specimens, by the specification’s method | Thin coating: alloy mix in the load, current density, area estimate |
Dissolved Aluminum
A little dissolved aluminum buffers the electrolyte against burning but does not prevent it (Chesterfield, Controlling Dissolved Aluminum). More lowers conductivity, though in his experience the change is virtually unnoticeable from 5 to 20 g/L. High levels also deposit aluminum sulfate (see Deposits, below), and an instrument maker notes that too much aluminum can harm the final surface appearance and raise electrical resistance (Metrohm AN-PAN-1018). Aluminum dissolving from the work binds sulfuric acid as aluminum sulfate: 3 H2SO4 per 2 Al, or about 5.5 g of acid per gram of aluminum (294 ÷ 54 g). That is a sulfate balance relationship, not a guaranteed change in the next free acid titer. Drag-out, water makeup and acid additions also move the result, and titration methods differ in how they treat aluminum. If aluminum rose from 5 to 15 g/L, the theoretical sulfuric acid equivalent associated with that sulfate balance is about 55 g/L of H2SO4; it is not a guaranteed change in the measured free acid titer. Restore free acid from a validated titration, not from that stoichiometric estimate, and do not add a second charge on top of the titer. Seeding a new bath with aluminum sulfate brings the sulfate with the salt and does not consume 5.5 g of free acid per gram of aluminum.
Current Density
Run to a current density rather than a voltage when a set thickness in a set time is the goal, with voltage as a useful gauge (Anodizing Fundamentals). A PCD may instead specify voltage ramps and hold times, which MIL-PRF-8625F allows (4.3.1). In Type II practice the relationship is broadly linear: 24 asf (2.6 A/dm²) builds 1 mil (25 μm) in 30 minutes, and 12 asf (1.3 A/dm²) in 60. Ramp the current slowly and evenly at the start, because the initial buildup sets the coating’s uniformity: generally 30 seconds to 5 minutes for Type II, and often a standard 5-minute ramp for hard, thick coatings.
Time
Anodizing time follows the Rule of 720 (Hardcoat Anodize Thickness Problem):
minutes = mils of coating × 720 ÷ current density (asf)
minutes ≈ μm × 3.12 ÷ A/dm² (Rule of 312)
Converting 720 exactly gives about 3.05, within about 2% of 3.12. The rule needs the load’s total surface area and required thickness, and it is an approximation that alloy and bath conditions shift. It works very straightforwardly for 6063, 5005 and 5052; 6061 may take a few minutes longer, depending on current density, and 2000 series alloys up to 50% longer in some cases. In a mixed load, the time must let the slower anodizing alloys reach a minimum thickness, which in some cases may give improper coatings on the faster ones (MIL-PRF-8625F, note 6.5; the specification’s section 6 notes are guidance, not requirements).
Chesterfield advises against stopping a load part way through to check thickness, because each stop and restart forms a separate layer that may lack the integrity of a single one. For Type III (hardcoat) thickness, MIL-PRF-8625F note 6.10.4 lists measuring the coated part while still wet and returning it to the bath for a fixed period as one control technique. If a line uses it, the PCD should define it.
⚙️ Type II and Type III Conditions, Hardcoat Current Density and Cooling
Type III differs from Type II mainly in temperature and current density; the acid ranges overlap. A general industrial range cited in the hard anodizing literature is −10 to +5 °C with 165 to 225 g/L free acid (Scampone et al., 2022; the acid span is cited there from Poyner, 1991), but shops run outside it. One shop hard anodizing 6061-T651 ran 220 to 240 g/L at −3 to −5 °C, at 2.5 A/dm² (23 asf) for 80 minutes, including a 15-minute ramp. Chesterfield had no issue with those bath conditions but judged the current density below optimum for 6061. At 3.9 A/dm² (36 asf), 2 mils should take about 45 to 50 minutes, including a 5-minute ramp. That current density might burn thin gauge parts with sharp edges; for those, he advises finding the highest current density that anodizes without burning when the conditions are set up on the first load. The PCD then fixes Type III current density and time for each alloy and part geometry.
Hard anodizing often runs at 0 °C or below, forming a hard porous oxide of large Vickers hardness (Kikuchi et al., 2015), so temperature control is largely a cooling problem. A chemical supplier advises enough cooling to keep the tank within a couple of degrees during anodizing (Couden, 2011). It sizes the cooling against the rectifier’s heat load of amps × volts × 3.412 Btu/h, with a water/glycol coolant in the colder hardcoat baths. The same supplier treats agitation as part of temperature control. A typical air sparger effectively agitates a 6 to 9 inch (about 15 to 23 cm) zone, pumps are sized for 4 to 6 tank turnovers per hour, and eductors can improve flow at the workpiece. A 5086 laboratory study stirred its bath against local temperature differences, citing earlier work linking a 5 °C local difference to oxide cracks at the interface (Raid et al., 2017). Check that the temperature reading location represents the work zone.
🧪 Titrating Free Acid and Dissolved Aluminum
A pH meter reading alone cannot determine the acid concentration (Hanna Instruments). Titrate free acid and aluminum with sodium hydroxide (NaOH). Three published shop procedures, not ASTM or ISO methods, are given by Chesterfield, Titrating Sulfuric Acid Anodizing Baths. Each has an acid endpoint (volume A here) and a total endpoint for acid plus aluminum (volume T), and aluminum follows from the difference. Published versions letter the volumes differently, so check which endpoint each letter means before using a formula. Use only a procedure your laboratory has approved and checked against a sample of known acid and aluminum.
Potassium fluoride method. Reagents: standard 1.0 N NaOH, and 50 g KF·2H2O in 100 mL of water, neutralized to phenolphthalein and stored in plastic because KF etches glass. Potassium fluoride is toxic if swallowed, in contact with skin or if inhaled, and fluoride added to the still acidic sample can form some hydrogen fluoride. Run this method only under your laboratory’s approved procedure, with the SDS, PPE, ventilation and fluoride waste controls it specifies.
- Dilute two 10 mL samples to about 150 mL each with DI or distilled water and add phenolphthalein.
- Titrate one to the phenolphthalein endpoint: volume T (acid plus aluminum).
- Titrate the other slowly until the alumina formed by each drop only slowly re-dissolves. Add 20 mL of the KF solution (or about 5 g of KF powder) to precipitate the aluminum, and continue to the endpoint: volume A (the acid).
- For these 10 mL samples, free sulfuric acid (g/L) = A × N × 4.9 and dissolved aluminum (g/L) = (T − A) × N × 0.9, N being the NaOH normality.
Methyl orange method. Chesterfield’s published procedure uses one 5.0 mL sample: dilute it to about 50 mL with DI water, titrate with 1.0 N NaOH from orange to a yellow methyl orange endpoint (A), then add phenolphthalein and continue on the same sample to a slight pink that persists (T). Free acid (g/L) = A × 9.8 and aluminum (g/L) = (T − A) × 1.8; these factors carry no normality term and assume that method’s 1.0 N NaOH and 5.0 mL sample. A Henkel additive technical data sheet uses the same factors on two separately prepared 5 mL samples (methyl orange on one, phenolphthalein on the other). Those are alternative procedures; do not mix their steps.
pH electrode method. Chesterfield’s published procedure: pipette a 5 mL sample into a beaker, dilute to about 200 mL, and titrate with 1.0 N NaOH to pH 3.6 (A), then continue on the same sample to pH 10.0 (T). Free acid (g/L) = A × N × 9.8 and aluminum (g/L) = (T − A) × N × 1.35. The 1.35 factor is the published conversion for that pH 10 endpoint; it treats each aluminum ion as consuming four hydroxide ions (aluminate). A separate Metrohm potentiometric bulletin (AB-089) uses a 2.0 mL sample, locates two inflection points rather than fixed pH values, and calculates aluminum on three equivalents (8.994 mg Al per mL of 1 mol/L NaOH). Do not carry endpoints or factors between those methods. The 0.9 and 1.8 factors in the other shop methods assume three equivalents.
The factors are equivalent weights (g/eq) divided by sample volume in millilitres. For sulfuric acid, 49.04 g/eq ÷ 10 mL gives about 4.9 and ÷ 5 mL gives about 9.8 (both protons). For aluminum, 26.98 g/mol ÷ 3 eq/mol = 8.99 g/eq, which ÷ 10 mL is about 0.9 and ÷ 5 mL is about 1.8; ÷ 4 eq/mol gives 6.745 g/eq, which ÷ 5 mL is about 1.35. Aluminum hydrolysis can begin near the methyl orange or pH 3.6 free acid endpoint, so those acid titers can run slightly high.
Published Shop Titration Factors (Not ASTM or ISO)
A is the free acid endpoint volume and T is the total endpoint, as defined in this article. Chesterfield letters the same volumes differently in print. Do not carry an aluminum factor between methods.
| Method | Sample | Free acid, g/L | Aluminum, g/L | Al equivalents assumed |
|---|---|---|---|---|
| Potassium fluoride (Chesterfield) | Two 10 mL aliquots | A × N × 4.9 | (T − A) × N × 0.9 | 3 |
| Methyl orange (Chesterfield, sequential) | One 5.0 mL aliquot; 1.0 N NaOH | A × 9.8 | (T − A) × 1.8 | 3 |
| pH 3.6 then 10.0 (Chesterfield) | One 5 mL aliquot | A × N × 9.8 | (T − A) × N × 1.35 | 4 |
| Potentiometric inflections (Metrohm AB-089) | 2.0 mL; two equivalence points | 49.04 mg H2SO4 per mL of 1 mol/L NaOH | 8.994 mg Al per mL of 1 mol/L NaOH | 3 |
Implementation Tip: Check any method against a sample of known acid and aluminum, including aluminum, before using it for bath control.
Automated titration. An automated potentiometric titrator can run the acid, aluminum and chloride titrations, avoiding inconsistent endpoint judgment between operators and trouble reading color changes in heavily colored samples (Hanna); check its repeatability with a gage R&R study for titration. Potentiometric methods also cover chloride, oxalic acid and sulfate (Metrohm AB-089), and online analyzers can track dissolved aluminum and free acid continuously for near real time adjustment (Metrohm AN-PAN-1018).
Chloride and makeup water. Contaminants, particularly chlorides, or improper racking may cause severe attack by the electrolyte (MIL-PRF-8625F, note 6.4). The PCD sets makeup water quality and any chloride limit. As practitioner guidance, not a specification limit, Chesterfield recommends deionized water of at least 200,000 ohm-cm resistivity (5 μS/cm) for making up most anodizing line baths, and a chloride limit of 50 ppm in the sulfuric anodizing bath itself. Resistivity does not measure chloride: cation resin is often regenerated with hydrochloric acid, so chloride can appear in water that still meets that resistivity. He advises testing the makeup water periodically with silver nitrate (cloudiness is an alert) and analyzing the bath for chloride rather than inferring it from resistivity.
⚠️ What Does Drift Look Like on Parts and in the Data?
On parts, drift shows as burning, soft or thin coatings, deposits and poor dye uptake; in the data, as conductivity and voltage changes. Most signs have more than one cause, so each calls for an investigation rather than a single adjustment.
Burning. Forcing the higher current a large part needs through contact areas sized for smaller parts may cause burning (MIL-PRF-8625F, note 6.14). In practitioner guidance (Chesterfield), a bath with no dissolved aluminum, or a high Type III current density on thin gauge parts with sharp edges, also raises the risk. Check agitation and solution flow at the work: the oxide surface heats locally during anodizing, less so in a colder bath (Scampone et al.).
Soft or powdery coatings. Before adjusting anything, check the bath analysis, temperature record, agitation and the load’s current schedule. Powdery areas and loose films fail the workmanship requirement of MIL-PRF-8625F (3.13), checked visually on lot samples (4.4). Studies link a soft or loose oxide to faster dissolution at its surface, but the direction depends on alloy and bath.
Thin coating or low coating weight. Check the load’s alloy mix, current density and area estimate, and test coating weight or thickness against the specification. MIL-PRF-8625F sets Type III thickness as a tolerance band around the nominal, not a minimum, so overshooting is nonconforming too (see Records and Process Control Tests). Chesterfield advises setting the conditions on the first load and adjusting later loads if needed. Keep adjustments within the PCD’s current density and time ranges, and check the first load’s thickness before release.
Deposits and poor dye uptake. Above roughly 25 to 30 g/L dissolved aluminum, aluminum sulfate tends to precipitate on the work, tanks, bus bars, ventilation hoods, pumps and heat exchangers and can impair dyeing (Chesterfield, Controlling Dissolved Aluminum). Other dyeing factors lie outside the bath: cast alloys may show dye bleed-out or lack of color from their porosity (MIL-PRF-8625F, 3.6), and coatings must not dry before dyeing (3.12).
Rising voltage. Trend the voltage at the same point of the cycle for comparable loads (same alloy family, current density and time). Under constant current in acidic solutions, the voltage climbs as the barrier oxide grows, dips slightly as the porous layer forms, then levels off once oxide formation and dissolution balance (Kikuchi et al.). The rise within one load, at constant current, mainly follows that sequence; it is not by itself a bath chemistry signal, and a temperature excursion or a contact change during the load can still move the voltage. Four factors are linked to the voltage:
- Temperature: warmer electrolyte lowered anodizing voltage in sulfate solutions in earlier work (Ling, Chen, Wang and Li, 2008, as cited in a 2021 foil study); cold raises the electrolyte’s resistance (Scampone et al.).
- Acid: stronger acid (0.4 to 2.0 M) gave a lower voltage in the same foil study, at constant current and 280 K.
- Dissolved aluminum: far above the usual range, it lowers conductivity (see Controlling Dissolved Aluminum).
- Cathodes: properly installed aluminum cathodes gave 1 to 2 V less cell voltage than a standard lead cathode system (Aluminum Anodizers Council Technical Bulletin 1-00), a comparison of setups, not of a bath over time.
A rising anodizing voltage does not identify a single cause. Compare its trend with bath temperature, acid concentration, dissolved aluminum, electrical contacts and cathode condition before selecting an adjustment. Any voltage action level comes from the PCD; none has a universal value. For slow drift, see why control limits do not detect gradual process drift.
🧰 Controlling Dissolved Aluminum and Making Up the Bath
Dissolved aluminum is held in range by periodic decanting or by ion exchange of the electrolyte. There is no universal decant point, decant fraction or replacement point: the PCD sets the aluminum limit, and the trend of successive analyses shows when the bath is approaching it. The theoretical dilution fraction to remove, assuming aluminum-free replacement liquid, return to the same final volume and ideal mixing, is
1 − (target ÷ measured aluminum)
As an illustration only, if the PCD target is 10 g/L and the bath measures 18 g/L, that fraction is about 44% (1 − 10 ÷ 18 ≈ 0.44). It is not an approved production procedure. The same fraction of free acid leaves with the decant. Handle the decanted electrolyte under the site’s waste procedures and wastewater permit; where they allow, it can serve for pH adjustment in waste treatment.
Refill with water of the quality the PCD specifies, leaving room for the acid addition. Size the acid addition from a fresh titration of the volume actually in the tank when the sample is taken, not by multiplying a heel titer by the finished volume. Add the acid under the makeup precautions below. Then bring the bath to level, mix and re-titrate, and check that the temperature is back in the PCD range before running work.
Chesterfield’s aluminum figures, from lowest to highest:
- About 2 g/L (startup): his advice for a new Type II or Type III bath.
- 2–5 g/L up to 20 g/L (observed): the normal operating ranges he has seen.
- 5–15 g/L (routine): generally used in the US for both types, 8–12 g/L for closer control; some specifications call for less than 5 g/L, and he advises following the one in use.
- 15 g/L (one shop’s maximum): recommended for one shop’s 6061-T651 hardcoat bath at 220 to 240 g/L acid and −3 to −5 °C, instead of the 30 g/L at the top of its range (Hardcoat Anodize Thickness Problem).
- Roughly 25–30 g/L and above (precipitation): aluminum sulfate deposits.
- Upwards of 80 g/L (extreme case): conductivity had fallen so far that one anodizer could get only half the rectifier’s output at full capacity (Controlling Dissolved Aluminum).
The best aluminum level also depends on acid strength: in Chesterfield’s experience, the lower the acid concentration, the higher the aluminum can be. An instrument maker lists 1 to 10 g/L as the aluminum range to expect when monitoring an anodizing bath (Metrohm AN-PAN-1018), not as a control limit.
To start a new bath at about 2 g/L, add aluminum sulfate. With the hydrated form normally supplied in North America, Al2(SO4)3·14.3H2O, 11 g per liter of bath raises dissolved aluminum by 1 g/L (Chesterfield, Adding Aluminum to a New Anodizing Bath). With the non-hydrated salt, Al2(SO4)3, the figure is 6.33 g per liter of brand new bath. Grades differ in hydration (13 or 14.3 waters, for example), so check the factor against the aluminum content on the supplier’s certificate and confirm the startup level by titration.
Making Up the Bath
Make up the bath by weight of acid, not by volume (Sulfuric Acid Anodizing Bath Makeup). Concentrated 93% (66° Baumé) acid is about 1.8 times as dense as water, so 15% by volume gives about 1.5 times the intended concentration. For example, 75 L of acid in 500 L of bath holds about 256 g/L of H2SO4 instead of 165 g/L.
Worked example, on assumed acid properties (93% w/w, density 1.834 kg/L): 500 L of finished bath (the final volume) at 165 g/L of pure H2SO4.
Pure H2SO4 (kg) = volume (L) × target (g/L) ÷ 1000
500 × 165 ÷ 1000 = 82.5 kg of pure acid
Commercial acid (kg) = pure H2SO4 ÷ assay
82.5 ÷ 0.93 ≈ 88.7 kg, or about 48 L (88.7 ÷ 1.834 ≈ 48.4 L)
A percent by weight target converts to g/L only through solution density. At 20 °C, 15 to 20% by weight is about 165 to 228 g/L (densities about 1,102 and 1,139 g/L: 0.15 × 1,102 ≈ 165; 0.20 × 1,139 ≈ 228). Chesterfield’s rounded factor of 11.1 g/L per percent (Adding Aluminum to a New Anodizing Bath) gives 167 to 222 g/L instead. Use the assay and density on your acid’s certificate of analysis, make up to the final volume, and confirm the finished bath by titration before use.
Concentrated sulfuric acid reacts violently with water, releasing heat.
Implementation Tip: Add the acid slowly to the water, never water to the acid (ILO International Chemical Safety Card 0362), with cooling, following your approved makeup SOP and the acid’s SDS.
In service, drag-out carries acid out at bath strength. When the lost volume is made up with water of the quality the PCD specifies (see Chloride and makeup water, above), or rinse water is dragged in, free acid falls. Dissolving aluminum also binds acid as sulfate (see Dissolved Aluminum). Calculate additions from the titrated free acid, not from a stoichiometric aluminum estimate and not from the original makeup (how to calculate chemical additions for plating baths). Water that only replaces evaporation does not dilute the bath.
✅ How Do You Run the Bath Control Loop?
Run seven steps for every analysis (measure, compare, contain, investigate, act, verify and record) so each decision can be traced later:
- Measure. Titrate free acid and dissolved aluminum at the frequency the PCD and contract set, and log temperature, current (or voltage ramp) and time for every load.
- Compare. Check each result against its PCD range and the trend of earlier results; a value can sit inside its range while drifting toward a limit.
- Contain. A result outside the PCD range is a nonconformance to the PCD unless the PCD defines a different response. Follow the reaction plan in the PCD or your nonconforming product procedure, which says whether to stop the line and which work processed since the last in range analysis to hold or trace for assessment. Restoring the bath does not dispose of that work. Treat a load run outside its PCD temperature, current density (or voltage ramp) or time range, for example after a chiller trip on a hardcoat line, the same way: segregate and assess it before release.
- Investigate. Before adjusting, repeat a suspect titration. Then check likely causes: drag-out, dilution and rising aluminum for falling free acid; work anodized since the last decant for rising aluminum; cooling capacity, agitation and instrument calibration for temperature; contacts, cathodes and load area for current and voltage changes.
- Act. Make the addition, decant or other correction the PCD approves, calculated from the titrated values.
- Verify and resume. Re-titrate after mixing and confirm that every affected PCD parameter (free acid, dissolved aluminum, temperature, and, if they moved, current density or voltage ramp and time) is back inside its PCD range. If the line was stopped, resume only after that verification, in compliance with the approved reaction plan, and with any authorization the plan requires. Restoring process conditions is not the same as evaluating or disposing of potentially affected product.
- Record. Enter the analysis, correction, verification and disposition of work processed while the bath was out of range in the bath history, and the disposition of any out of range load in its per-load record.
MIL-PRF-8625F attaches no failure response to an out of range solution analysis (4.3.2), so the reaction plan decides whether the line stops. A failed Table II process control test, by contrast, halts production at once. The reason is determined and corrected before production resumes, process control testing is performed at the start of production, and all traceable work (work whose location is known) from the time the failed specimens were anodized to the halt is rejected unless the contracting officer specifies otherwise (4.3.4; see Failed process control test in the records table below). A re-titration inside the PCD range is not that process control test.
Key Takeaway: Restoring the bath is not product disposition. Verify every affected PCD parameter before restart, and dispose of potentially affected work as the reaction plan directs.
Lab Wizard Cloud is built around this same loop: the analysis is entered, the result is compared with the limits, the addition is calculated, and the verification stays with the bath record.
📋 Records and Process Control Tests
For MIL-PRF-8625F work, keep a process control document and analyze every processing solution at least once every two weeks unless the procuring activity specifies otherwise. Keep each bath’s history of additions and analysis results for at least a year after the contract or purchase order is completed, and run the Table II process control tests at least monthly. The revision your contract or purchase order invokes governs. As of November 2026, ASSIST lists MIL-PRF-8625F Amendment 2 (23 November 2020) as the active revision (full text, third party copy). Operating conditions remain the supplier’s option unless the contract, purchase order or drawing specifies them (3.2). Contracts that still invoke MIL-A-8625F should note that Amendment 1 (2003) added Type III corrosion when sealed, and that Amendment 2 uses undyed sealed corrosion specimens, SAE AMS4037 panels, ASTM B244 or E376 for thickness, and Type III lot thickness examination.
MIL-PRF-8625F (Amendment 2, 2020): Records and Tests for the Anodizing Bath
Clause numbers refer to MIL-PRF-8625F with Amendment 2. The contract or the procuring activity can change several of these requirements, as the rows note.
| Record or test | Requirement (clause) |
|---|---|
| Process control document (PCD) | Developed, maintained and adhered to; at minimum it describes all processing steps, each step’s immersion time and temperature ranges, chemical constituents and allowable solution control ranges, and temperature, current density and anodizing time ranges (or voltage ramps and hold times) for individual alloys or alloy series (4.3.1) |
| Solution analysis | All processing solutions in the line, at least every two weeks unless the procuring activity specifies otherwise, against the PCD ranges (4.3.2) |
| Bath history record | Each bath’s added chemicals or treatment solutions and all analysis results, kept at least one year after the contract or purchase order is completed and, with test reports, available on the procuring activity’s request (4.3.2) |
| Process control tests | Table II specimens at least monthly, at intervals of no more than 35 days, and at the start of production after a month or longer without production to the specification (4.3.3.1) |
| Tests by type | Types I, IB, IC, II and IIB: coating weight, corrosion (undyed and sealed) and, for Class 2 when specified, light fastness (dyed and sealed); Type III: thickness (or weight, at the procuring activity’s option), abrasion and, only when specified as sealed, corrosion (Table II) |
| Failed process control test | Production halted immediately; the reason determined and corrected before production resumes; process control testing performed at the start of production; all traceable work (work whose location is known) from the time the failed specimens were anodized to the halt rejected unless the contracting officer specifies otherwise (4.3.4) |
| Coating weight | Unsealed, before dyeing or sealing: Type II at least 1,000 mg/ft² (3.7.1.1, Table I); Type III, where weight replaces thickness (procuring activity’s option), at least 4,320 mg/ft² for every 0.001 inch of coating (3.7.2.1.1) |
| Type III thickness | Nominal 0.002 inch (about 51 μm), varying by no more than ±20% for coatings up to 0.002 inch and no more than ±0.0004 inch (about 10 μm) for thicker coatings, unless the contract, purchase order or drawing specifies otherwise (3.7.2.1) |
| Type III thickness checks | Samples from every lot, after anodizing, sealing and dyeing if applicable, per MIL-STD-1916 Verification Level II; acceptance criteria as specified in the contract or order (4.4, 4.4.2). Monthly process control tests (Table II). ASTM B244 or E376, average of at least eight readings; metallographic cross section in a dispute (4.5.1) |
Table IV of the specification gives typical thicknesses of 0.00007 to 0.0010 inch (about 1.8 to 25 μm) for Type II and 0.0005 to 0.0045 inch (about 13 to 114 μm) for Type III. The drawing sets the thickness for a given part.
Beyond the bath history, keep each load’s record (alloy, surface area, current or ramp, time and temperature) with the lot, so a coating result can be traced to its conditions. For control limits on the analysis records, see how to set control limits for plating and metal finishing.
🔬 What the Research Shows About Acid and Temperature
Across these laboratory studies, a colder bath or stronger acid did not always give a thicker or harder layer; the direction depended on alloy, acid level and current density. Each result holds only for its own conditions, so none of them sets a production window.
Key Takeaway: Hold the PCD temperature and acid ranges rather than extrapolating a production window from any one laboratory study.
Acid and Temperature Studies: Conditions and Results
Conditions as each study reports them.
| Study | Alloy | Bath and temperature | Current | Reported result |
|---|---|---|---|---|
| Die-cast study (Scampone et al., 2022) | AlSi11Cu2(Fe) (EN AC-46100) high pressure die castings, sandblasted; 11.5% Si measured (MIL-PRF-8625F bars Type III above 8.0% nominal silicon unless the procuring activity approves) | 168 or 217 g/L free acid at −4.5 or +5 °C | Galvanostatic: linear ramp from 0 to 0.88 A/dm² over 20 min, then 35 min at 0.88 A/dm² (an industrial schedule after Henley, 1982) | Mean layers, μm: 3.4 ± 1.2 (+5 °C, 168 g/L), 5.2 ± 2.1 (+5 °C, 217 g/L), 5.5 ± 1.6 (−4.5 °C, 168 g/L), 4.6 ± 1.5 (−4.5 °C, 217 g/L) |
| 5086 study (Raid et al., 2017) | 5086 (4.9 wt% Mg), 1 L laboratory cell, stirred | 20% by mass sulfuric acid; cooling set at −1 to 4 °C (low temperature) or 16 to 21 °C (room temperature) | Initial current 1.4 A; 30 to 300 min | Cold: layers up to 150 μm within a few hours, harder, with enhanced pore formation; room temperature: below 80 μm |
| Foil study (2021) | 99.99% aluminum foil, 10.5 μm | 0.4, 1.0 and 2.0 M sulfuric acid (about 39 to 196 g/L) at 280 K (about 7 °C) | Constant current, 0.5 to 10 mA/cm² | Stronger acid: no significant change in anodizing rate, a lower voltage and a falling volumetric growth factor (oxide thickness over the original aluminum thickness) |
| 6061 study (Coatings, 2022) | 6061 | 160 g/L sulfuric acid at 5, 15 and 25 °C | 15 V pulsed; current density varied to hold voltage. Steady-state current 0.4 A at 5 °C, 0.9 A at 15 °C, 2.4 A at 25 °C | Warmer: thicker films with a coral-like, loose porous surface (current also rose). Cooling from 25 to 5 °C: steady-state friction coefficient 0.46 to 0.39, wear rate down about 69% |
| Microhardness study (2015) | Aluminum | Sulfuric and oxalic acid electrolytes of several concentrations, −1.78 to 45.78 °C | 1, 3 and 5 A/dm² | Warming raised layer microhardness about 66% at 1 A/dm² (less with more sulfuric acid) and lowered it at 3 A/dm² for oxidation under 25 min; no significant temperature effect below 10.5 V |
On the die casting, the mean layer was thicker with a colder bath or stronger acid alone, but the reported spreads overlap, so those differences are not cleanly resolved. The authors interpret the colder bath gain as slower oxide dissolution. Together (−4.5 °C, 217 g/L), the mean was thinner than either change alone (4.6 μm, against 5.5 and 5.2 μm), and that combination also gave the lowest scratch resistance (critical load 2.5 ± 1 N, against about 5 ± 1 N otherwise). Those are the authors’ attributions: they credit the thinner layer to electrolyte resistance outweighing slower dissolution, and the low scratch resistance to dissolution of the outer layer. The die-cast layers were only 3.4 to 5.5 μm, so the study describes neither wrought alloys nor thick Type III coatings, and 0.88 A/dm² is not a general Type III current density. Why stronger acid lowers the foil study’s growth factor is unsettled: its authors propose faster oxide dissolution but judge chemical dissolution of the film surface unlikely at 7 °C.
❌ Common Mistakes to Avoid
- Do not make up the bath as percent by volume. 66° Baumé (93%) acid measured as 15% by volume gives about 1.5 times the intended concentration.
- Do not read free acid from a pH meter. Titrate with NaOH; a pH electrode is only a titration endpoint, not a concentration reading.
- Do not carry an aluminum factor between titration methods. The pH method’s factor assumes a different reaction.
- Consider seeding a new bath with dissolved aluminum. Unless the PCD says otherwise, Chesterfield advises starting at about 2 g/L, because a bath with none can burn parts more easily.
- Do not rush the ramp on long, thin walled extrusions. High current applied in only a few seconds may concentrate at the ends and leave more coating there than in the middle, a special concern for parts to be colored (Chesterfield, Anodizing Fundamentals).
🔍 How Lab Wizard Cloud Supports Anodizing Bath Control
Lab Wizard Cloud helps anodizing operations manage bath chemistry from routine analysis through corrective action and verification. It calculates free acid and dissolved aluminum concentrations from entered titration results, compares results against defined specification limits, and calculates chemical additions when adjustments are needed.
Analysis scheduling helps laboratories maintain consistent testing intervals, while historical trends and built-in SPC tools help operators and engineers recognize changes before they become larger process problems. Analysis results, chemical additions, and subsequent verification results remain available for review, supporting investigations and bath history documentation.
By keeping these activities connected, Lab Wizard Cloud reduces reliance on separate spreadsheets, handwritten records, and individual knowledge. The software supports consistent process decisions while the facility’s approved procedures continue to govern production holds, restarts, and product disposition.
📌 Key Takeaways
- Control free acid, dissolved aluminum and temperature to your PCD and contract; the specification revision the contract invokes governs. As of November 2026, ASSIST lists MIL-PRF-8625F Amendment 2 (2020) as the active revision for US military work.
- Reference ranges are not limits. Practitioner guidance for Type II is 15 to 20% acid by weight, 68 to 80 °F and 10 to 24 asf. A general range cited in the literature for Type III is about −10 to +5 °C and 165 to 225 g/L, which shops run outside.
- Hold dissolved aluminum to the PCD limit by decanting or ion exchange; practitioner guidance gives 5 to 15 g/L (8 to 12 g/L for closer control), and some specifications call for less than 5 g/L. The dilution fraction 1 − (target ÷ measured) is a theoretical estimate, not a production procedure.
- Make up the bath by weight from pure acid and final volume, add acid to water under the approved SOP, and confirm by titration. Restore free acid from the titer, not from a stoichiometric aluminum estimate.
- Titrate with NaOH, not a pH meter reading, on the PCD schedule, with each method’s own endpoints and factors under an approved procedure, and act on the trend, not only on an out of range result.
- Run each load to the PCD’s current density (or voltage ramp) and time, and keep a per-load record.
- Treat an out of range analysis or load record as a nonconformance to the PCD unless the PCD defines a different response: hold or trace the affected work as the reaction plan directs, restore and verify all affected PCD parameters before restart, and record the disposition. Restoring the bath is not product disposition.
- Under MIL-PRF-8625F, a failed Table II process control test halts production, the cause is corrected, process control testing is performed at the start of production, and traceable work from the failed specimens to the halt is rejected unless the contracting officer specifies otherwise (4.3.4).
🔗 Related Resources
- How to Manage Plating Bath Chemistry Automatically. LIMS, SPC and alerts behind bath history records.
- Control pH, Conductivity, and Temperature in Plating. How pH, conductivity and temperature interact in wet process baths.
- Calculate Chemical Additions for Plating Baths. The general arithmetic behind acid replenishment.
- How to Set Control Limits for Plating & Metal Finishing. SPC control limits and how they differ from specification limits such as PCD ranges.
- Why Control Limits Don’t Detect Gradual Process Drift. Why a process can stay inside its control limits while drifting from target.
- MSA in Plating Labs: Gage R&R for Titration & pH. Gage R&R for titration, the check on the repeatability of your own bath analyses.
External Links
- DLA ASSIST QuickSearch: MIL-PRF-8625. Status and revision history.
- MIL-PRF-8625F with Amendment 2 (2020), third party copy of the ASSIST download. Full text with Amendment 2.
- Finishing & Coating: Rollout of MIL-PRF-8625 Rev F and Implications for Anodizers (2021). Westre and Munson (CHEMEON) on the revision; they advise reviewing standing contracts for supersession.
- MIL-A-8625F (1993, superseded) and its Amendment 1 (2003). For contracts that still invoke it; the amendment adds Type III to the corrosion test when sealed.
- Products Finishing columns by Larry Chesterfield: Anodizing Fundamentals (Type II conditions and ramping); Controlling Dissolved Aluminum in the Type II and Type III Anodizing Bath (aluminum ranges and decanting); Adding Aluminum to a New Anodizing Bath (aluminum sulfate additions); Sulfuric Acid Anodizing Bath Makeup (makeup by weight); Titrating Sulfuric Acid Anodizing Baths (three titration methods); Hardcoat Anodize Thickness Problem (Rule of 720 and a 6061 case); DI Water Quality (2004) (makeup water and chloride).
- Products Finishing: Aluminum Anodizing (Couden, 2011). A chemical supplier on cooling load and agitation.
- Aluminum Anodizers Council Technical Bulletin 1-00 (2013). Aluminum cathodes and cell voltage.
- PubChem: Potassium fluoride, GHS classification. Acute toxicity of the titration reagent.
- ILO International Chemical Safety Card 0362: Sulfuric acid, concentrated. Handling and dilution of concentrated acid.
- Hanna Instruments: Optimizing Your Anodizing Baths. An instrument maker on titration rather than pH alone.
- Metrohm AB-089. Potentiometric analysis of anodizing baths.
- Metrohm AN-PAN-1018. Online analysis of acids and aluminum in anodizing baths.
- Kikuchi et al. (2015). Review of porous oxide formation and the constant current voltage curve.
- Scampone et al. (2022). Electrolyte temperature and free acid in hard anodizing of die-cast Al-Si-Cu alloys.
- Raid et al. (2017). Layer growth and hardness on 5086, cold versus room temperature.
- Reduced-temperature sulfuric acid anodizing of foil (2021). Foil study with its own voltage results; cites Ling et al. (2008) for the temperature–voltage effect.
- Anodizing temperature and 6061 oxide films (Coatings, 2022). Electrolyte temperature, film structure and wear on 6061.
- Electrolyte temperature, current density and layer microhardness (2015). Microhardness across sulfuric and oxalic electrolytes.
