Acrylonitrile-Butadiene Copolymer (NBR) Supplier Onboarding: NBR Factory Audit Checklist: 18 Points to Verify Before Onboarding a Supplier

Acrylonitrile-butadiene copolymer, designated NBR under ASTM D1418, is supplied as a raw elastomer for oil-resistant seals, gaskets, O-rings, fuel and oil hoses, diaphragms, and certain industrial gloves. Commercial grades contain 18–50 wt% bound acrylonitrile; standard Mooney viscosity values range from 25 MU to 105 MU when measured as ML 1+4 at 100°C according to ISO 289-1:2014 or ASTM D1646-19a. Bound acrylonitrile content increases resistance to aliphatic hydrocarbon swelling and raises the glass transition temperature from approximately -55°C at 18 wt% ACN to -5°C at 50 wt% ACN. Mooney viscosity influences green strength, extrusion die swell, and injection molding fill behavior. A supplier that can blend crumb or bales to a certificate of analysis without tight process control may still deliver a lot that meets release limits but fails downstream dispersion or dimensional stability requirements.

Supplier onboarding is therefore based on an 18-point audit. The checklist covers process control, bound monomer composition, molecular architecture, contamination, packaging, and data integrity. Selection of a supplier based on a single approved sample without examining 25 consecutive lot records is insufficient for a raw polymer in which ±2 wt% bound ACN drift can alter oil swell and low-temperature performance under ASTM D471-16a immersion testing.

Why Raw Polymer Consistency Precedes Every Downstream Compounding Decision

In a carbon black-reinforced NBR compound used for compression-set-resistant seals, the elastomer’s molecular weight, branching, and polar nitrile content determine mixing energy, dispersion quality, and crosslink distribution during vulcanization. A supplier lot at the low end of a ±5 MU Mooney window may mix faster but can reduce extrudate dimensional stability on a 20:1 cold-feed extruder; a lot at the high end can raise compound Mooney and require process oil adjustment. A drop in bound acrylonitrile content of 1.5 wt% below target may not be visible in durometer but can increase volume swell in aggressive test oils by 3–5% absolute after immersion. The precise magnitude depends on filler loading and plasticizer type, so the audit seeks raw polymer data rather than relying on downstream compound testing alone.

On a production-scale intermeshing twin-screw extruder with 44:1 L/D, gel particles in raw NBR can accumulate on screen packs and raise head pressure above a normal 12–18 MPa operating range. The audit therefore checks whether the supplier measures gel content on every lot or only when a customer complains. A gravimetric methyl ethyl ketone insolubles method should be calibrated against a reference NBR sample; values above 2 wt% for injection molding grades frequently produce surface roughness in thin-wall diaphragms and O-ring preforms.

For sealing applications, compression set after 22 h at 100°C under ISO 815-1:2014 is often used as a cured-compound acceptance criterion, but it is a lagging indicator. Raw polymer audits add leading controls for bound ACN, Mooney viscosity, gel content, and residual catalyst or coagulation salts.

The checklist below combines documentary controls with test results. An auditor should request the last 25 production lot certificates, internal release data, retained sample availability, and corrective action records. If a supplier’s laboratory cannot demonstrate correlation between its internal Mooney viscosity data and an independent ISO/IEC 17025 laboratory for at least 10 lots, the supplier’s release data should not be accepted without incoming lot testing at the receiving plant. For food-contact NBR articles, documentation under 21 CFR 177.2600 should be requested at the audit stage; this does not replace migration testing under specific end-use conditions.

How Should an Auditor Verify Drying, Metal Contamination, and SPC Records?

Volatile matter and water content are not the same measurement. ISO 248-1:2011 reports total volatiles by hot-roll or oven heating; water-specific values are obtained by Karl Fischer titration with an oven extraction cell operating at 160°C. Raw NBR with 0.4 wt% total volatile matter may still contain enough trapped moisture to create porosity in a 1.5 mm extruded profile during atmospheric hot-air vulcanization. The auditor should inspect daily dryer records for inlet air dew point, residence time, and exhaust temperature. A fluidized-bed dryer with an inlet air dew point above -20°C can reabsorb moisture into crumb during humid weather unless the packaging line is climate-controlled.

Metal contamination in raw NBR accelerates oxidative aging and causes pigmentation defects in light-colored or translucent compounds. Iron, copper, and manganese are the most catalytically active. Microwave digestion in nitric acid/hydrogen peroxide followed by ICP-OES gives the audit-relevant profile. The auditor should not accept a single “heavy metals < 10 mg/kg” statement without specifying the digestion method and the elements quantified. Where a supplier uses a calcium chloride coagulation system, residual calcium can range from 150 mg/kg to 400 mg/kg across campaigns; the absolute value is less important than tight lot-to-lot control because calcium residues can interact with acidic accelerators and alter scorch behavior.

Statistical process control review is the least intrusive test of whether an NBR supplier can hold raw properties. The auditor should request individual moving-range charts for Mooney viscosity and bound ACN over the last 25 lots. A supplier that reports only batch averages and not within-lot standard deviation is likely blending crumb or using selective sampling. For lot release, Cpk ≥ 1.33 for Mooney viscosity and bound ACN are practical minimums; Ppk ≥ 1.67 is preferable when the supplier has collected 100 or more lots.

Acrylonitrile Content, Mooney Viscosity, and Bound Antioxidant Control

Bound acrylonitrile content is a stoichiometric calculation from nitrogen. Acrylonitrile contains 26.4 wt% nitrogen; the conversion factor from nitrogen to bound acrylonitrile is 3.79. Thus a nitrogen value of 8.0 wt% corresponds to 30.3 wt% bound acrylonitrile. The ISO 24698-1:2018 method uses Kjeldahl digestion, but some plants use calibrated FTIR for speed. The audit verifies that the FTIR model covers the full ACN range and is not extrapolated from a 20–35 wt% calibration when releasing a 45 wt% grade.

Mooney viscosity is an empirical torque measurement. A change in specimen mass, mill roll temperature during sample preparation, or viscometer rotor temperature can shift the result by 3–5 MU. The supplier should run a certified reference rubber at the beginning and end of each shift and maintain the viscometer according to ISO 289-1:2014 annex requirements. Auditors should also verify that sample sheets are not stored in direct sunlight or near mill roll heat; premature crosslinking from poor sample handling can elevate apparent viscosity and mask true raw polymer processability.

Bound antioxidants in NBR are not inert; they are consumed during drying, storage, and end-use heat aging. Extraction with acetone/THF and HPLC-UV at 280 nm can identify hindered phenols and some phosphite degradation products. The audit should compare antioxidant peak area per gram against the supplier’s master formula. If the supplier uses polymerized trimethyl dihydroquinoline, the UV response may be broad; in that case a nitrogen-specific detector or LC-MS is more appropriate. A lot with low antioxidant can pass standard Mooney and ACN tests but fail long-term heat aging at 100°C for 168 h under ISO 188.

Batch records for cold-polymerized NBR should show monomer feed ratio, initiator injection, shortstop addition, residual monomer stripping, coagulation aid, crumb washing, and dryer residence time. Drift in coagulation pH or salt concentration changes crumb size and residual soap content; larger crumb can carry more calcium into the dryer and generate localized gel particles. On a 200 L internal mixer, a calcium residue shift from 180 mg/kg to 620 mg/kg has been observed in production audits to reduce scorch time by 8–15% in a sulfur donor compound. The interaction is formulation-specific, and the auditor should request a designed experiment generated on the supplier’s pilot line rather than relying on a certificate of analysis.

When a Supplier’s Batch Records Diverge from Certificate of Analysis Data

The audit’s most severe finding is not a single out-of-specification lot; it is a systemic divergence between laboratory data and the certificate of analysis. An auditor should compare the raw chromatogram or viscometer chart timestamp against the certificate lot number. If the supplier’s database permits manual editing of Mooney viscosity or bound ACN results without a traceable reason code, the supplier cannot demonstrate data integrity. This is particularly important for NBR because a 5 MU Mooney shift can be hidden by blending a high-viscosity lot with a low-viscosity lot; the blended product may meet the release range but has broader molecular weight distribution, different shear thinning behavior, and increased die swell.

A practical audit sequence includes selecting 5 retained samples at random, resubmitting them blind to the supplier’s laboratory, and comparing the results to the original certificate. For Mooney viscosity, the repeatability stated in ISO 289-1:2014 should be reviewed; a difference of more than 3 MU between two laboratories warrants a full measurement system analysis. For bound ACN, the ISO 24698-1:2018 repeatability limit should be confirmed by proficiency testing. The audit should also verify that balance and viscometer calibration records are traceable to ISO/IEC 17025 and that the laboratory participates in interlaboratory rubber reference programs.

Retained samples from the first three production lots should be sealed in aluminum-lined polyethylene bags, stored at 23±2°C and 50±5% RH, and retained for 24 months. Incoming inspection at the receiving plant should include bound ACN, Mooney viscosity, ash, volatile matter, and gel content on every container if the lot count is below 10; above 10 lots, the sampling plan may be reduced to every second lot only after 15 consecutive acceptances. The audit closes when all 18 points have documented evidence and the first three incoming lots pass the receiving plant’s release panel.