Determination of Ten Volatile Sulfur Compounds Using Nutech 8910CF Cryogen Free Preconcentrator and GC FPD

Overview

This application note evaluates an electronically cooled Nutech 8910CF preconcentrator coupled to a gas chromatograph with a flame photometric detector (GC FPD) for ten volatile sulfur compounds in canister samples. In the reported laboratory study, the system produced log–log calibration correlation coefficients of 0.9979–0.9998, method detection limits of 0.01–0.03 nmol/mol (0.04–0.09 µg/m³), and relative standard deviations of 0.22–7.87% across three spike levels. The preconcentrator operated without liquid nitrogen.

Analytes and instrumentation

The analytes were hydrogen sulfide, carbonyl sulfide, methanethiol, ethanethiol, dimethyl sulfide, carbon disulfide, methylethyl sulfide, thiophene, diethyl sulfide, and dimethyl disulfide. A mixed sulfur compound gas standard was used. Nitrogen, hydrogen, and air each had a stated purity of at least 99.999%.

Nutech sampling canisters, the Nutech 2104 canister cleaner, Nutech 2208 precision diluter, Nutech 3608L autosampler, and Nutech 8910CF preconcentrator were used with a Shimadzu Nexis GC-2030 equipped with an FPD. The workflow covered canister cleaning, preparation of working standards, automated introduction, preconcentration, chromatographic separation, and sulfur selective detection.

Figure 1. Complete canister-based sulfur compound monitoring system integrating Nutech sample preparation instruments with GC–FPD.

Table 1 Operating conditions

Calibration and validation

Two working standard mixtures prepared with the Nutech 2208 were introduced at different volumes to establish six calibration levels. Concentrations in Table 2 are expressed as equivalents for a 400 mL sample volume: for a fixed analyte, the mass delivered depends on the working gas concentration multiplied by the actual injection volume. Calibration used the logarithm of molar fraction (nmol/mol) on the x axis and the logarithm of peak area on the y axis. This treatment accounts for the characteristic nonlinear sulfur response of a conventional FPD.

Table 2 Calibration levels in nmol/mol as 400 mL sample equivalents

For detection limits, seven replicate measurements at the lowest calibration level were used to calculate the standard deviation s. The reported method detection limit was MDL = 3.143 × s. Precision and recovery were assessed with six replicate measurements of spiked blank samples at low, medium, and high levels.

Results

All ten log–log calibration correlations were between 0.9979 and 0.9998. Table 3 provides analyte specific retention times and laboratory detection limits. Detection limits in nmol/mol are rounded as reported; the µg/m³ values are retained from the source dataset.

Table 3 Calibration and method detection limit results

RT = retention time; r = correlation coefficient for the log–log calibration; MDL = method detection limit under the stated laboratory procedure.

Precision and recovery

Across the low, medium, and high spiking levels, relative standard deviations were 0.84–3.60%, 1.14–5.76%, and 0.22–7.87%, respectively. Recoveries were 98.6–105.5%, 81.3–97.7%, and 91.6–108.7%, respectively. All individual results are retained in Table 4; reported mean concentrations are rounded independently of recovery percentages.

Table 4 Precision and recovery in spiked blank samples

Six replicates per concentration level. Spiked blank tests describe laboratory performance with this configuration and these test gases.