Benzene Contamination in Xylene: Detection Limits, Test Methods and Buyer Risk

Xylene is a C8 aromatic hydrocarbon fraction containing para-xylene, meta-xylene, ortho-xylene, and ethylbenzene in proportions determined by the production route. It enters commerce as mixed xylene for solvent and blending service, or as high-purity isomer streams for p-xylene oxidation, o-xylene oxidation, and m-xylene-derived isophthalic acid production. Catalytic reformate, pyrolysis gasoline, and toluene disproportionation streams can carry benzene into the xylene fraction when fractionation columns are operated at high throughput, when overhead recycle streams are not fully stabilised, or when upstream feed switches are not isolated. Benzene is an IARC Group 1 carcinogen with an OSHA 8-hour TWA of 1 ppm and a short-term exposure limit of 5 ppm under 29 CFR 1910.1028. In xylene purchasing, benzene is therefore controlled both as a regulatory exposure boundary and as a marker of incomplete aromatics fractionation. A product that meets total xylene purity may still be unacceptable if the benzene concentration exceeds the narrow limit set by downstream exposure or catalyst performance.

Industrial xylene grades are not interchangeable with respect to benzene. Mixed xylene for solvent or blending may be sold with a benzene maximum that reflects coating and degreasing exposure scenarios, while p-xylene feedstock for purified terephthalic acid and o-xylene feedstock for phthalic anhydride may carry tighter benzene clauses. The difference can be an order of magnitude or more, depending on the production route and contractual specification. A buyer that does not fix the test method code and the lower quantification limit on the purchase specification inherits analytical uncertainty that can exceed the specification itself. In practice, the benzene value in a certificate of analysis is a combination of the method, the calibration, the sampling, and the laboratory’s LOQ. Without all four, the reported concentration is not a reliable release parameter.

What Detection Limits Apply to Benzene in Mixed Xylene and Isomer-Grade Streams?

The detection limit for benzene in xylene is not a physical constant. It is a method-defined concentration produced by a calibrated gas chromatograph, an injection system, and a detector maintained within stated linearity. In xylene commerce, three standard methods account for most certificates of analysis: ASTM D7504, ASTM D2360, and UOP 931. ASTM D7504 uses capillary gas chromatography with flame ionisation detection and effective carbon number correction to determine trace impurities in monocyclic aromatic hydrocarbons, including benzene in mixed xylene and p-xylene feedstock. ASTM D2360 is an older capillary or packed-column GC-FID method for trace impurities in the same aromatic family. UOP 931 is a refinery gas chromatographic method configured specifically for mixed xylenes. Each method reports benzene as mass concentration in the xylene matrix, typically in mg/kg, but none fixes a universal detection limit without reference to injection mode, split ratio, column dimensions, and detector maintenance.

Because benzene is more volatile than xylene and has a distinct retention time on a nonpolar or polar capillary column, gas chromatographic separation is not the primary challenge. The primary challenge is quantitative accuracy at trace levels in a matrix containing higher-boiling methylated aromatics. A flame ionisation detector gives a mass-sensitive response that is linear over a wide range, but the response factor for benzene must be corrected for carbon number and detector conditions. ASTM D7504 formalises the effective carbon number correction so that benzene can be quantified against an internal standard without preparing an exhaustive curve for every impurity. The method covers monocyclic aromatic hydrocarbon streams and is suitable for benzene in mixed xylene, but the lower limit depends on the injection technique and the detector signal-to-noise ratio.

Method capability should be stated as a lower quantification level, not a detection limit. Detection limits based on a signal-to-noise ratio of 3:1 are screening values; a release test requires a validated LOQ at a signal-to-noise ratio of at least 10:1. A split injection with a 100:1 split ratio and a 1 µL liquid sample may not reach an LOQ below 5 mg/kg unless the detector signal is unusually high. A splitless injection of 1 µL to 2 µL on a 50 m to 60 m capillary column with 0.20 mm internal diameter and a 0.5 µm dimethylpolysiloxane or polyethylene glycol film can reduce the benzene LOQ into the low mg/kg range. Direct cool on-column injection can further reduce discrimination of the lower-boiling benzene but requires a retention gap and more frequent column maintenance.

Calibration should bracket the expected benzene concentration. A calibration curve prepared from 0.5 mg/kg to 25 mg/kg benzene in a xylene matrix is preferable to a single-point calibration because it reveals curvature at low levels. Laboratories may use a stock solution prepared gravimetrically from high-purity xylene and benzene, then dilute into the same mixed xylene matrix. The curve should be verified with a second independent standard, preferably from a different lot, and the relative standard deviation of replicate injections at the LOQ should be below 15 % for release testing. If the LOQ standard shows a relative standard deviation above 20 %, the method is not sufficiently stable at that concentration and the LOQ should be raised. Buyers can request the calibration curve and LOQ standard data as part of a method validation package.

When Benzene Limits Fall Below 1 mg/kg in High-Purity p-Xylene

High-purity p-xylene contracts may carry a benzene limit below 1 mg/kg because downstream oxidation catalysts and product purity requirements are sensitive to aromatic impurities. At this concentration, conventional split injection with flame ionisation detection can be inadequate unless the laboratory validates a splitless protocol, increases injection volume, and uses very clean inlet liners. Gas chromatography-mass spectrometry with selected ion monitoring at m/z 78 is used for confirmation in some laboratories, but co-introduced xylene isomers can raise ion source pressure and alter ion ratios if the method is not optimised. Published data for this specific configuration is limited, so laboratories should demonstrate an LOQ of no higher than half the specification limit before using GC-MS data for release.

At 1 mg/kg benzene in p-xylene, the mass of benzene injected in a 1 µL splitless injection is approximately 1 ng; a modern flame ionisation detector can detect this, but matrix peaks can obscure integration. The benzene peak elutes before toluene and before the xylene isomers, so interference is usually limited to low-boiling nonaromatic hydrocarbons. If the column is a 50 m × 0.20 mm × 0.5 µm dimethylpolysiloxane capillary column with 1.0 mL/min helium carrier gas, benzene can be baseline-resolved from cyclohexane and low-boiling paraffins when the oven initial temperature is held at 35 °C to 40 °C for 5 min to 10 min before ramping. A polyethylene glycol column can provide different selectivity for polar impurities but may require higher initial temperature stability.

Sample preparation can be the limiting step. Benzene is present in laboratory air, in septum materials, and in poorly conditioned gas lines. A headspace sampler can reduce liquid matrix injection contamination, but the benzene headspace concentration is controlled by the partition coefficient between xylene and vapour; calibration must be performed with standard additions in the same xylene matrix rather than in a light naphtha diluent. Vials should be capped with PTFE-lined septa and stored under nitrogen headspace at 2 °C to 8 °C if analysis is delayed beyond 24 h. Repeated needle penetrations should be minimised because volatile benzene loss from a pierced septum can produce a negative bias that exceeds the method repeatability.

Standard addition is used when the xylene matrix contains unknown nonaromatic components that alter the benzene response factor. In standard addition, a known benzene spike is added to a split sample at two or three levels; the original benzene concentration is extrapolated. The technique corrects for matrix suppression or enhancement but is labour-intensive and not always used in routine purchase verification. Published data for standard addition in xylene with benzene below 1 mg/kg is limited; laboratories should validate the approach before applying it to a commercial dispute.

CoA Verification, Retained Sampling, and Common-Carrier Cross-Contamination

A certificate of analysis is not a substitute for a sampling protocol. Liquid xylene tanks can stratify by temperature and composition after transit; a single top sample may not represent the batch. Sampling should follow ASTM D4057 or an equivalent pipeline sampling practice, with flow-proportional composite sampling preferred for ship or barge transfers. The purchase order should specify the sampling point, the sample container, the purge procedure, and the analytical method code. A CoA that states benzene at 1.0 mg/kg but does not state the LOQ leaves the buyer unable to determine whether the result is a measurement or an upper bound.

Retained samples should be sealed in borosilicate glass with PTFE-lined closures and stored away from direct sunlight. Benzene loss from xylene can occur through closures that are not vapour-tight, and the loss rate is accelerated by repeated opening. If a retained sample is reanalysed after a dispute, the laboratory should compare the result with the method reproducibility, not with the original single value, because chromatographic integration, calibration standards, and injection technique all contribute scatter. Common-carrier tank trucks and ISO containers may carry benzene-containing gasoline or aromatics in previous loads; a heel of a few litres in a bottom hose or valve can contaminate an otherwise clean xylene load. Buyers should require a dedicated xylene compartment or a verified wash procedure with a solvent that does not leave a benzene residue.

Private specifications are often stricter than regulatory cut-offs. ASTM D5211 for xylene feedstock to p-xylene production includes composition limits for benzene and other hydrocarbons; the current edition should be referenced because limits are revised through consensus. REACH Annex XVII restricts benzene in substances and mixtures placed on the European market, but downstream polymer producers may impose lower contract values to protect food-contact packaging or pharmaceutical intermediates. When a buyer accepts material without a defined benzene method, the default supplier method may be optimised for total aromatics purity rather than trace benzene, and the test may not detect a specification-breaking concentration until it is already blended into a process stream.

Benzene in xylene is not only a product specification item. The liquid concentration influences the vapour concentration in storage and processing. A closed-loop loading system with vapour recovery can reduce worker exposure even if the benzene concentration is high, but open loading hatches and drainage points can exceed the short-term exposure limit if the liquid benzene is high. Under OSHA 29 CFR 1910.1028, the action level for benzene is 0.5 ppm as an 8-hour TWA; exposure monitoring is required when airborne benzene exceeds that level. In practice, a xylene with benzene at 10 mg/kg can still generate low airborne benzene in a poorly ventilated area, but the exposure depends on temperature, ventilation, and liquid surface area. The buyer should not assume that a low benzene CoA eliminates the need for workplace monitoring.

Benzene Content Shifts Downstream Yield, Catalyst Deactivation, and Occupational Exposure

In p-xylene oxidation to purified terephthalic acid, benzene enters the oxidation reactor as a lower-boiling aromatic that is not the intended methyl-substituted feedstock. The incremental benzene load can lower reactor selectivity if partial oxidation intermediates accumulate or if the off-gas treatment system is not sized for additional volatile organics. The effect is process-specific and should be evaluated through pilot-scale oxidation testing; published data for this specific configuration is limited. In o-xylene oxidation to phthalic anhydride, benzene can contribute to hot spots at the catalyst inlet and increase carbon dioxide formation relative to the desired anhydride yield. In mixed xylene solvent applications, benzene above the specification raises the exposure potential for open-surface coating, degreasing, and cleaning operations even when the total aromatic content remains unchanged.

Occupational exposure in downstream production is governed by the benzene concentration in the liquid raw material after accounting for vapour pressure and ventilation. A mixed xylene containing benzene at 10 mg/kg and an open-top cleaning tank does not create the same airborne concentration as a benzene-containing solvent with higher volatility in a confined space, but the benzene vapour component is not zero. Engineering controls, personal exposure monitoring, and substitution of low-benzene xylene grades are used where the task is prolonged. The buyer risk therefore extends beyond the certificate of analysis: it includes the quality of the supplier’s tank-cleaning records, the laboratory’s method validation package, and the ability to re-analyse retained samples under the same standard method.