Precision Evaluation of Inorganic Cadmium Determination in PM2.5: Analytical Methods, Statistical Me

Precision evaluation of inorganic cadmium determination in PM2.5 is a critical quality assurance process that verifies the repeatability and reproducibility of heavy metal measurements in fine particulate matter. By systematically assessing relative standard deviation (RSD), spike recovery, and replicate consistency, laboratories can ensure that cadmium concentration data meet regulatory requirements for ambient air monitoring, health risk assessment, and source apportionment studies.

Technical Overview

Inorganic cadmium is one of the most closely monitored toxic trace metals in fine particulate matter (PM2.5) due to its confirmed carcinogenicity, long biological half-life, and severe respiratory and renal toxicity following inhalation exposure. Because ambient cadmium concentrations in PM2.5 typically fall in the low ng/m³ range, the analytical result is highly sensitive to minor variations in sample preparation, instrument response, and operator technique. Precision evaluation is therefore an indispensable component of the overall quality control system for cadmium determination.

Precision evaluation refers to the statistical assessment of the closeness of agreement between independent test results obtained under stipulated conditions. For inorganic cadmium in PM2.5, precision is evaluated at three levels: repeatability (same operator, same instrument, same day), intermediate precision (different days, operators, or calibration sequences), and reproducibility (different laboratories or methods). The primary statistical metric is the relative standard deviation (RSD), calculated as the ratio of the standard deviation to the mean concentration, expressed as a percentage.

The dominant analytical platform for this evaluation is inductively coupled plasma mass spectrometry (ICP-MS), which offers detection limits at the sub-ng/L level in digested filter extracts. Graphite furnace atomic absorption spectrometry (GFAAS) and energy-dispersive X-ray fluorescence (ED-XRF) serve as complementary or confirmatory techniques. Reference procedures include HJ 657 (determination of metals in ambient particulate matter by ICP-MS), EPA Compendium Method IO-3.5, and EN 14902 for lead and cadmium analysis on suspended particulate matter.

Testing Samples

Membrane Filter Samples

PM2.5 particles are collected on 47 mm quartz fiber or polytetrafluoroethylene (PTFE/Teflon) membranes using medium- or high-volume samplers equipped with PM2.5 impactor heads at a flow rate of typically 100 L/min for 23–24 hours. The loaded filters are conditioned in a constant-temperature, constant-humidity chamber (typically 15–30 °C and 45–55% relative humidity) before gravimetric determination and are then sectioned for chemical digestion. Quartz membranes are preferred for digestion-based methods because of their low cadmium blank, whereas Teflon membranes are standard when non-destructive XRF verification is also required.

Digested Extract Solutions

Filter subsamples or portions (typically 1/4 to 1/2 of the membrane) are digested using a mixed acid system of nitric acid, hydrofluoric acid, and hydrogen peroxide (HNO₃–HF–H₂O₂) in a closed-vessel microwave digestion system. The resulting clear solution is diluted to a fixed volume (25 or 50 mL) and constitutes the analytical sample for instrumental cadmium quantification. Hydrofluoric acid is essential to dissolve aluminosilicate crustal particles that may carry bound cadmium.

Certified Reference Materials and Quality Control Samples

Filter-based certified reference materials (such as NIST SRM 2783, Air Particulate on Filter Media) and externally validated QC samples are incorporated into each analytical batch. These materials provide an independent check on both the trueness and precision of the complete measurement chain, from digestion through instrumental detection.

Fortified and Blank Samples

Laboratory reagent blanks, digestion blanks, duplicate field samples, and matrix-spiked samples (fortified with known cadmium standards) are prepared alongside routine specimens. Spike recoveries between 85% and 115% and blank values below one-tenth of the method detection limit are generally required to validate batch precision and freedom from contamination.

Testing Items

The precision evaluation program for inorganic cadmium in PM2.5 encompasses the following test items:

• Total inorganic cadmium concentration in PM2.5, expressed in ng/m³ (atmospheric basis) and μg/L or ng/L (extract basis)
• Repeatability testing: at least six independent replicate determinations of the same homogenized digest within a single analytical sequence, reported as intra-batch RSD
• Intermediate precision testing: replicate determinations across five or more days, involving different calibration curves, operators, and instrument tuning conditions, reported as inter-day RSD
• Parallel filter subsample precision: digestion and analysis of multiple sections cut from the same loaded membrane to assess filter homogeneity and sub-sampling variance
• Spike recovery evaluation: fortification of representative filter digests at 0.5×, 1×, and 2× of the native cadmium level
• Method detection limit (MDL) determination: seven replicates of low-level spiked blanks, with MDL calculated as t(n−1, 0.99) × standard deviation
• Calibration verification: initial calibration verification (ICV) and continuing calibration verification (CCV) checks, with acceptance typically within ±10% of the certified value
• Internal standard recovery monitoring: indium (¹¹⁵In) or rhodium (¹⁰³Rh) response drift within ±20% of the initial calibration value

Testing Methods

Sample Preparation Protocol

A representative filter section is placed in a pre-cleaned PTFE microwave digestion vessel with 6 mL of concentrated nitric acid (65%, guaranteed reagent grade), 2 mL of hydrofluoric acid, and 2 mL of hydrogen peroxide. The vessel is subjected to a programmed microwave ramp (e.g., to 180–200 °C over 15 minutes, held for 20–30 minutes). After cooling, the digest is evaporated gently to near dryness to remove excess HF, reconstituted in 2% nitric acid, transferred quantitatively to a volumetric flask, and diluted to volume with ultrapure water (resistivity ≥18.2 MΩ·cm). A full procedural blank accompanies every digestion batch of no more than 20 samples.

ICP-MS Determination

Cadmium is quantified by ICP-MS using isotopes ¹¹¹Cd (primary) and ¹¹⁴Cd (confirmatory) to monitor and correct for potential isobaric interferences from molybdenum oxides (⁹⁸MoO⁺ on ¹¹⁴Cd) and tin (¹¹²Sn, ¹¹⁴Sn). Indium (¹¹⁵In) is commonly used as the internal standard to compensate for matrix suppression and instrument drift. A multi-point external calibration curve spanning the expected concentration range (typically 0.05–10 μg/L for Cd) with a correlation coefficient r ≥ 0.999 is established, and the calibration blank must be below the MDL. Collision/reaction cell conditions are optimized if oxide interferences exceed 2% (as indicated by the CeO⁺/Ce⁺ ratio).

Precision Evaluation Procedure and Statistical Criteria

For repeatability assessment, a homogeneous digest is analyzed six consecutive times (n = 6); the mean, standard deviation (s), and RSD = (s / mean) × 100% are computed. For inter-day precision, the same stable sample is analyzed on at least five separate days under recalibrated conditions. Consensus acceptance criteria applied in routine practice are summarized below:

• Intra-batch (repeatability) RSD: ≤ 5% for extract concentrations above 10 times the MDL; ≤ 10% for low-level samples near the quantification limit
• Inter-day (intermediate precision) RSD: ≤ 10%
• Reproducibility RSD (between laboratories): ≤ 20% for ambient-level cadmium
• Relative deviation of parallel filter subsamples: ≤ 15% when concentrations exceed 10 times the MDL
• Spike recovery: 85–115%
• Internal standard recovery: 80–120% of the calibration-time signal

Results failing these criteria trigger investigation of digestion completeness, calibration integrity, interference correction, and blank contamination, followed by full re-analysis of the affected batch. Precision data are trended on Shewhart control charts (mean ± 2s warning limits, ± 3s control limits) to demonstrate ongoing statistical control as required for laboratory accreditation.

Confirmatory Techniques

For verification, a subset of membranes may be analyzed non-destructively by energy-dispersive X-ray fluorescence (ED-XRF) following EPA Method IO-3.3, and selected digests may be re-measured by graphite furnace atomic absorption spectrophotometry with a cadmium hollow cathode lamp. Agreement between methods within ±20% supports the robustness of the reported precision. Standard additions may be applied to confirm the absence of matrix effects in difficult samples.

Applications

Precision evaluation of inorganic cadmium determination in PM2.5 supports a wide range of scientific and regulatory applications:

• National and regional ambient air quality monitoring networks, where long-term trend analysis requires measurement uncertainty small enough to detect year-on-year concentration changes
• Compliance assessment against ambient cadmium reference values, such as the EU limit of 5 ng/m³ (annual mean) and targets set under WHO air quality guidelines
• Source apportionment studies using receptor models (PMF, CMB), in which input uncertainty matrices are derived directly from the precision (RSD) of each elemental measurement
• Human exposure and health risk assessment, where inhalation dose estimates for carcinogenic metals depend on defensible concentration data
• Atmospheric research on heavy metal transport, transformation, and deposition, including haze episode characterization and long-range transport tracing
• Workplace and industrial emission monitoring around smelters, battery manufacturing, and waste incineration facilities
• Laboratory accreditation and method validation under ISO/IEC 17025, CMA, and CNAS frameworks, where documented precision studies are mandatory technical evidence
• Development and verification of standard reference methods and proficiency testing schemes for inter-laboratory comparison programs

FAQ

What is an acceptable RSD for cadmium determination in PM2.5? For routine ICP-MS analysis of digested filter samples, an intra-batch RSD of ≤ 5% is expected at concentrations well above the detection limit, while ≤ 10% is acceptable for low-level ambient samples. Inter-day RSD should remain within 10%, and between-laboratory reproducibility within 20% is considered satisfactory at ng/m³ concentration levels.

Why must precision be evaluated separately from accuracy? Precision characterizes random error—the scatter among replicate results—whereas accuracy reflects the closeness of the measured value to the true concentration. A method can be precise yet biased (for example, due to incomplete digestion or calibration error), so precision evaluation is combined with spike recovery and certified reference material analysis to demonstrate total method validity.

Which cadmium isotope should be selected for ICP-MS measurement? Isotope ¹¹¹Cd is generally preferred as the primary quantitative mass because it suffers fewer isobaric interferences; ¹¹⁴Cd is monitored concurrently for confirmation, with mathematical correction applied for ⁹⁸MoO⁺ and ¹¹⁴Sn overlap where necessary.

What are the main sources of imprecision in this measurement? Dominant contributors include non-homogeneous particle loading across the filter membrane, incomplete acid digestion, cadmium contamination from reagents or labware (a significant risk given the low ng-level target), instrumental drift and matrix suppression, and isobaric spectral interferences. Rigorous clean-room handling, closed-vessel microwave digestion, internal standardization, and full procedural blanks control these sources.

How is the method detection limit related to precision evaluation? The MDL is itself a precision-based statistic, calculated from the standard deviation of seven replicate low-level spiked blanks multiplied by the Student's t-value (3.143 for n = 7 at 99% confidence). Reporting limits are commonly set at 3–5 times the MDL, below which precision degrades unacceptably and RSD criteria no longer apply.

How frequently should precision evaluation be performed? A full precision study (six replicates, multi-day design) is required during initial method validation. Thereafter, ongoing control is demonstrated through duplicate analyses in every batch (minimum 10% of samples), continuing calibration verification every 10 samples, quarterly MDL verification, and annual participation in inter-laboratory proficiency testing.