Systematic Quality Assurance/Quality Control Framework for a Volatile Organic Compound Real-Time Monitoring Instrument, Autonomous Rugged Optical Multigas Analyzer for VOCs

Abstract

This study presents a systematic quality assurance/quality control (QA/QC) framework for new technology adopters to validate environmental monitoring instruments when consistent, standardized validation and calibration procedures do not readily exist. This QA/QC framework is especially relevant for newly developed environmental monitoring instruments prior to deployment in field settings as part of environmental site investigations. Key QA/QC parameters used in the framework include repeatability, intermediate precision, reproducibility, linearity, limits of detection (LOD), limit of quantification (LOQ), trueness, and recovery. The QA/QC framework is demonstrated on the autonomous rugged optical multigas analyzer for VOCs (AROMA-VOC), a technology that employs cavity ring-down spectroscopy (CRDS) with preconcentration and chromatographic separation prior to CRDS detection. QA/QC protocols for nine VOCs, and results for AROMA-VOC were compared with laboratory-based gas chromatography/mass spectrometer (GC–MS). Results demonstrated high linearity (R ² > 0.95) and low LOQ values (0.0004 ± 0.0002–0.0455 ± 0.0192 μg/L). The instrument also exhibited ranges of LOD at least an order of magnitude lower than that of the referenced GC–MS. AROMA-VOC demonstrated acceptable precision [repeatability, intraday relative standard deviation (RSD %): 1.72–12.1%; intermediate, interday RSD %: 2.01–10.93%] and substantial recovery (104.8% to 212.4%). Interlaboratory testing also concluded that AROMA-VOC can produce consistent detections (R ² > 0.90). The z-score findings suggested that the performance of the instrument was satisfactory, with the exceptions of xylenes and 1,3-butadiene. Overall, using the systematic QA/QC framework presented in this study, AROMA-VOC is shown to be a robust method for volatile organic compound analysis.

1. Introduction

Volatile organic compounds (VOCs) are a significant environmental pollution concern and require environmental monitoring of soil, water, and air. VOCs are defined as a class of organic chemicals with a high tendency to vaporize into the air under normal atmospheric conditions of temperature and pressure (i.e., 25 °C, 1 atm). , VOCs are frequently found in outdoor and indoor media and frequently require environmental monitoring. Chlorinated VOCs have been widely used in industrial processes, such as degreasing operations and dry cleaning industries. , Additionally, petroleum hydrocarbon VOCs are commonly found in fuels, oils, and other materials. If releases occur, the VOCs may contaminate air, groundwater, and soil, potentially leading to harmful impacts on human health through a variety of exposure pathways. ,

Conventional VOC sampling and analytical techniques often involve collecting environmental field samples and sending them to off-site laboratories for analysis, which can be time-consuming and costly and delay data-informed decision making. , Conversely, real-time VOC monitoring provides continuous, instantaneous data or near real-time quantification of environmental conditions, facilitating the rapid identification of pollution sources and trends. Real-time monitoring facilitates a detailed understanding of temporal variations and spatial distributions of VOCs. Importantly, real-time monitoring instruments that are field-deployable provide opportunities for quicker decision-making and site-response actions at locations with environmental contamination. ,

To advance the practice and regulatory acceptance of sampling and analysis approaches, when early technology adopters begin deploying new real-time VOC monitoring instruments, validation is necessary. In addition, clear data quality objectives, standardized operating procedures, and quality-controlled protocols can decrease discrepancies in data accuracy and reliability, subsequently improving the integration and interpretation of real-time monitoring data across various platforms and settings. −

This study presents a systematic QA/QC framework that can be used by early technology adopters to validate environmental monitoring instruments. To demonstrate the application of this systematic QA/QC framework, herein we validate a new near real-time sensor, autonomous rugged optical multigas analyzer for VOCs (AROMA-VOC). AROMA-VOC, developed by Entanglement Technologies, Inc., uses cavity ring-down spectroscopy (CRDS) coupled with thermal desorption (TD-CRDS) to measure and separate VOCs in air (Table ). The nine VOCs in the standard AROMA-VOC library for this model include petroleum hydrocarbons (e.g., benzene, 1,3-butadiene, toluene, ethylbenzene, xylenes, isoprene, and styrene) and chlorinated solvents [e.g., trichloroethylene (TCE) and cis-1,2-dichloroethylene (cis-DCE)].

1. Physical and Chemical Properties of Chemical Compounds.

volatile compound	molecular weight (g/mol)	boiling point (°C)	density (g/cm3)	vapor pressure (mmHg)	conversion factors
1,3-butadiene	54.09	–4.5	0.6149 @ 25 °C	2.11 × 103 @ 25 °C	1 ppm = 2.21 mg/m3
benzene	78.11	80.1	0.8787 @ 15 °C	75 @ 20 °C	1 ppm = 3.26 mg/m3
cis-1,2-dichloroethene	96.94	59.6 @ 745 mmHg	1.2837 @ 20 °C	180 @ 20 °C	1 ppm = 3.96 mg/m3
ethylbenzene		136.2	0.8671 @ 25 °C	7 @ 20 °C	1 ppm = 4.34 mg/m3
isoprene	68.12				1 ppm = 2.79 mg/m3
styrene	104.15	145.2	0.9059 @ 20 °C	5 @ 20 °C	1 ppm = 4.33 mg/m3
toluene	92.14	110.6	0.8631 @ 20 °C	28.4 @ 25 °C	1 ppm = 3.76 mg/m3
trichloroethylene	131.38	87.2	1.4642 @ 20 °C	69 @ 25 °C	1 ppm = 5.46 mg/m3
xylene	106.16	137–140	0.864 @ 20 °C	6.72 @ 21 °C	1 ppm = 4.34 mg/m3

^aInformation obtained from refs and .

There are several other instruments that perform monitoring and/or sampling of VOCs that are comparable in one or more ways to AROMA-VOC. For example, HAPSITE, a mobile gas chromatography/mass spectrometer (GC–MS), which was originally designed to detect toxic industrial chemicals and chemical warfare agents, has been adapted for environmental monitoring. − Other real-time instruments that can detect VOCs in environmental samples include the 8610C (GC), Griffin G510 (GC–MS), and FROG-5000 [GC-PID (photoionization detector)]. Additionally, the United States Environmental Protection Agency (EPA) has invested in the development of a trace atmospheric gas analyzer, which is a mobile laboratory that contains standard operated GC–MS instruments inside a recreational vehicle. TAGAs are capable of identifying and quantifying VOCs or other organic chemicals at concentrations of parts-per-billion (ppbV) or lower. Direct injection mass spectrometry has also been utilized in certain instances to monitor VOCs in real time. For example, proton transfer reaction mass spectrometry and selected ion flow tube mass spectrometry have been widely adopted to facilitate rapid, sensitive, and continuous measurements of VOCs directly from the air or other matrices. − These technologies utilize soft ionization methods, where reagent ions (e.g., H₃O⁺, NO⁺, O₂ ⁺, OH^–, NO₂ ^–, and NO₃ ^–) interact with target VOCs to facilitate compound identification and quantification. All of these real-time instruments may have challenges with sensor sensitivity and precision, which can lead to false positives or negatives as a result of environmental interferences and varying detection limits. , Calibration and maintenance for these technologies require frequent adjustments and specialized knowledge to operate the instrument. Additionally, conditions and factors, such as temperature, humidity, battery life, and costs of deployment, further complicate the analysis process.

Compared with other available real-time instruments, AROMA-VOC is a newer technology that has been gaining popularity for environmental monitoring. The purpose of this research is to present and evaluate analytical performance quality assurance parameters for AROMA-VOC, which uses TD/GC-CRDS, to lab-based GC–MS. GC-based VOC analytical methods, in particular GC–MS, are considered the “gold standard” in environmental analysis of VOCs, due to regulatory acceptance. ,

As real-time instruments are implemented by early adopters and deployed for site investigations, a critical gap remains in practice exists when there is a lack of availability of consistent and standardized validation and calibration procedures (e.g., Narayana et al. 2022). Several researchers contribute to this gap in research. Zhu et al. conducted a validation study on metal oxide sensors, comparing their results with thermal desorption-GC-mass spectrometry (TD-GC-MS). The study addressed certain validation elements, such as detection limits, calibration, and retention time. Additionally, You et al., Radica et al., Hong et al., and Xu et al. introduced innovative and unique VOC analysis technologies at various settings and partial validation components. ,−

This study builds on these past efforts to present a systematic quality assurance/quality control (QA/QC) framework to validate a real-time instrument and ensure data reliability. The systematic QA/QC framework presented herein can be applied for other emerging on-site real-time instruments to ensure they meet rigorous QA/QC criteria prior to these instruments being deployed for site investigations.

2. Materials and Methods

2.1. Reagents and Chemicals

A VOC gas standard in ultrapure nitrogen (±5% uncertainty) was purchased from Apel-Riemer Environmental, Inc. (Miami, FL). The gas standard included 10 parts-per-billion (ppbV) of equivalent for 1,3-butadiene (0.0234 μg/L), benzene (0.0326 μg/L), cis-1,2-dichloroethene (0.0396 μg/L), ethylbenzene (0.0434 μg/L), isoprene (0.0279 μg/L), toluene (0.0376 μg/L), trichloroethylene (0.0546 μg/L), styrene (0.0433 μg/L), m-xylene (0.0434 μg/L), o-xylene (0.0434 μg/L), and p-xylene (0.0434 μg/L). Nitrogen gas (99.99999% laser grade) was used as a carrier gas in AROMA-VOC to perform analysis for the samples. The nitrogen gas was purchased from American Welding & Gas (AWG).

2.2. Samples

Samples for analysis were prepared in a Teflon bag (7 × 7″ Teflon Sample Bag from Analytical Specialties) installed with a DP4 fitting. Specific concentrations of samples are prepared by mixing a precise volume of clean air with a precise volume of gas standard to achieve the desired concentration. The experimental setup included two gas sources: one supplying clean air, connected to the fume hood via a hose, and the other supplying VOCs from the gas standard. Both gas sources are connected to a high accuracy flowmeter. Clean air, sourced from a building gas pump that extracts atmospheric air, was first analyzed with the instrument to confirm the baseline of the background VOC concentrations. By timing flow from both gas sources (5 ft³/h for air; 2 ft³/h for gas standard), an accurate volume of the two-gas mixture was supplied to a Teflon sample bag, ensuring the desired sample concentration for analysis. Refer to Table S3 for the lab setup.

2.3. Analysis by AROMA-VOC

The AROMA-VOC model used in this research was the ARVOC-181. AROMA-VOC combines sample trapping and preconcentration followed by a ramped thermal desorption to achieve chemical separation with a broadly tunable cavity ring-down spectrometer. In a cycle, a controlled volume of sample is absorbed onto a sample collection tube (Carbopack B) at 35 °C. Sample intake is controlled and measured by an internal mass flow controller. At the conclusion of sample collection, the sample collection tube is heated rapidly to 325 °C and returns adsorbed molecules to the vapor phase, which is carried into a focuser (Carbopack B). The temperature of the focus tube is ramped from 35 to 300 °C over 266 s, leading to sequential desorption of target analytes. The vapors are carried by a carrier gas stream [50 standard cubic centimeters (sccm)] into the CRDS instrument core. The CRDS core sequences between a discrete set of 6 wavelengths spanning 1623 nm to 1719 nm are a repeated sequence with a dwell time of 0.75 s per wavelength. At each wavelength, the CRDS collects approximately 825 ring-down events. Compound identification is performed using a combination of arrival time and recovered spectra, requiring both to match. The AROMA-VOC uses cavity-locked CRDS which provides high coupling efficiency and a high duty cycle to improve CRDS sensitivity.

The instrument offers two analysis modes: Rapidscan (continuous real-time) and Labscan (near real-time grab sampling). Rapidscan provides direct continuous analysis in less than 5 s, resulting in compound classification, rather than compound identification. The classification includes aromatic compounds, chlorinated compounds, dienes compounds, methane, and water. Likewise, the Labscan mode is a target compound batch analysis mode designed for measuring specific chemical concentrations that were listed above. Results are generated in 10 min for Labscan mode. In this mode, users can use the default action scripts or create custom action scripts by specifying parameters such as sample volume and collection time. The validation of the instrument in this study is performed in Labscan mode. The schematic of AROMA-VOC is shown in Figure .

1.

Schematic of AROMA-VOC.

2.4. Method Validation Framework

Verifying adequate quality assurance and quality control (QA/QC) of analytical instruments is common and necessary for environmental monitoring. Thus, this section incorporates a methodological framework recommended by literature and regulatory agencies. − Table presents quality assurance parameters that were included in the framework along with comprehensive procedures, definitions, and regulatory standards. These parameters were selected because they are commonly used for QA/QC, and they were included in the QA/QC validation of HAPSITE instruments by the U.S. EPA. ,

2. Definitions and Methods for all Quality Assurance (QA) and Quality Control (QC) Parameters .

parameter	definition	expressions	methods	regulatory standards
repeatability ,,	condition of measurement that is held fixed while performing two or more measurements over a short period of time	RSD = 100 × SD/ $\mathrm{x̅}$ ; RSD: relative standard deviation SD: standard deviation $\mathrm{x̅}$ : mean. Horwitz function = 2 × exp(1–0.5 × log C); C: certified concentration	measure 6 replicate samples of 10 ppbV VOCs certified materials on the same day	USEPA: RSD ≤ 20% TO-15: RSD ≤ 30% Horwitz function: RSD ≤ 32%
intermediate ,,	estimate of the variation in results when measurements are made in a single laboratory but under conditions that are more variable than repeatability		measure single or multiple samples of 10 ppbV VOCs certified materials on multiple extended periods of days
reproducibility ,,	condition of measurement, out of a set of conditions that includes different locations, operators, measuring systems, and replicate measurements on the same or similar objects		measure at least 5 samples of 10 ppbV VOCs certified materials using different instruments in a different laboratory
trueness ,,,	closeness of agreement between the average of measured values obtained by replicate measurements and a reference value	$Z=\frac{x_{\mathrm{f}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{d}}-x_{\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{i}\mathrm{g}\mathrm{n}\mathrm{e}\mathrm{d}}}{\mathrm{S}\mathrm{D}\mathrm{P}\mathrm{A}}$ ; xfound: measurement mean, xassigned: certified value (provider estimate value), SDPA: standard deviation for proficiency assessment	measure single or multiple samples of 10 ppbV VOCs certified materials on multiple extended periods of days	Z-score ≤ 2, satisfactory, 2 ≤ Z-score ≤ 3, warning signal, Z-score > 3, unsatisfactory
recovery ,,,	the fraction of analyte added to the test sample (fortified or spiked) prior to analysis, which is measured by the method	recovery (%) = $\frac{{\mathrm{x}}_{\mathrm{f}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{d}}}{{\mathrm{x}}_{\mathrm{c}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{e}\mathrm{d}}}\times 100\%$ ; x found: mean detections from instrument x certified: certified values for the standard provided by vendor	measure at least 5 samples of certified materials	USEPA: recovery = 70–120%, CODEX: recovery = 40–120%
limit of blank (LOB)	the highest apparent analyte concentration expected to be found when replicates of a blank sample containing no analyte are tested	LOB = x bl + 1.645 × SDbl; x bl: mean blank samples, SDbl: standard deviation of blank samples	measure at least 5 blank samples	N/A
limit of detection (LOD) ,	the lowest concentration or amount of analyte, that can reliably be distinguished from zero or that can be detected with reasonable statistical certainty	xL = x bl + 3SDbl; $\mathrm{L}\mathrm{O}\mathrm{D}=\frac{x_{\mathrm{L}}}{S}$ S: slope of calibration curve	measure at least 5 certified materials at the known concentration	N/A
limit of quantification (LOQ) ,	the lowest concentration or amount of analyte that can be determined quantitatively with an acceptable level of repeatability precision and trueness	xL = x bl + 10SDbl; $\mathrm{L}\mathrm{O}\mathrm{Q}=\frac{x_{\mathrm{L}}}{S}$	measure at least 5 certified materials at known concentration	N/A
linearity ,−	the ability of the method to obtain test results proportional to the concentration of analyte (within a given range)	linearity demonstrated using calibration curves	construct calibration curve	N/A

^aNote: Information, regulatory standards, and guidelines were referred to credited organizations for practical assessment.

3. Results and Discussions

3.1. Method Validation

3.1.1. Linearity

Linearity represents the relationship between responses obtained by an analytical method to the known analyte concentration in a sample that can be visualized as a straight line in a 2-dimensional Euclidean space. ,− To validate quantitative analytical methods, the range over which linearity is evaluated includes the response values over which the method is intended to be. To establish a calibration curve, 5-point calibration was conducted on AROMA-VOC at concentrations of 0, 3, 5, 7, and 10 ppbV. Each sample was prepared in triplicate to ensure accuracy and reproducibility. The instrument’s response was measured, and the resulting data points were plotted to generate the curve. Table presents linear regression equations and coefficient of determination R ² representing linearity, which is equal to the square of the linear correlation coefficient. The right side of Table is calculated from AROMA-VOC calibration measurements, and the left side includes literature-reported GC–MS calibration results. For GC–MS, the literature values were reported for a calibration range approximately an order of magnitude greater than AROMA-VOC’s calibration range.

3. Calibration Curve Comparison with GC–MS .

	GC–MS		AROMA-VOC
VOCs	calibration curve equation	linearity (R 2)	calibration curve equation	linearity (R 2)
1,3-butadiene	y = 16,910x + 2436.6	0.9994	y = 2.0598x – 0.4741	0.9783
benzene	y = 75,303x + 9609.7	0.9997	y = 1.0766x + 0.4261	0.9794
cis-1,2-dichloroethene			y = 1.0735x – 0.2149	0.9818
ethylbenzene	y = 121,781x + 6841.1	0.9999	y = 1.2711x – 0.0925	0.9757
isoprene		0.9890	y = 1.0285x + 0.7142	0.9794
styrene	y = 57766x – 10,087	0.9999	y = 1.1498x + 0.351	0.9699
toluene	y = 97,609x + 97,944	0.9996	y = 0.9979x + 1.786	0.9750
trichloroethylene	y = 24,918x – 6931.5	0.9990	y = 1.167x – 0.1766	0.9819
xylene	y = 93,040x – 2895.5	0.9998	y = 2.1199x – 1.436	0.9570

^aNote: Information obtained from refs and .

^bThe calibration range is approximately an order of magnitude greater than AROMA-VOC’s calibration.

The R ² calculated for the AROMA-VOC method was at least 0.95 and above, indicating very good linearity across the tested range and highlighting the consistency of the method. Additionally, the R ² values for AROMA-VOC compared with GC–MS values reveals that AROMA-VOC performance is similar to GC–MS (“gold-standard” technique for VOC environmental sampling and analysis). , The consistency between both sets of values highlights the reliability and accuracy of the AROMA-VOC method in quantifying VOCs and supports its use as a viable alternative to traditional GC techniques.

3.1.2. Precision (Also Referred to as Comparison)

Precision is a parameter that compares measurements from the instrument between each sample, either within a short interval (same day), long period (months), or multiple instruments. This parameter examines whether the instrument can produce a consistent signal when identical samples are fed into the system, even under different conditions. In this study, three types of precisionrepeatability, intermediate, and reproducibilitywere determined for AROMA-VOC and the results were compared with previous studies on GC–MS. ,

Repeatability was determined in working sessions of 6 replicates (intraday analyses). Intermediate precision (interday analyses) of each volatile was determined with 11 replicates (interday analyses) in 6 nonconsecutive working sessions (at least once a week). Reproducibility precision was conducted using different AROMA-VOC instruments (Instruments A and Instrument B). This comparison was done to evaluate “off-the-shelf” results from two different instruments. As discussed below, one of the instruments was not calibrated for all of the compounds.

Figure presents the results for precision as the relative standard deviation (RSD %). Values for repeatability are within the good range (<20% RSD) ,as described by EPA as the recommended threshold value. For all except 2 compounds (xylenes and styrene), % RSD values are less than 5%. Additionally, % RSD of less than 30% was recommended in the compendium method TO-15 and the values that were obtained in this study fell within the range. In comparison with GC–MS analysis, AROMA-VOC demonstrated better repeatability (lower RSD %) for all compounds, except styrene, xylene, and ethylbenzene, which have at least 2 methyl functional groups. Table S1 includes precision values and detection limits for AROMA-VOC and GC–MS.

2.

Precision values for AROMA-VOC.

The results from intermediate precision for AROMA-VOC are also <20% RSD, suggesting that AROMA-VOC has better or equal intermediate precision performance compared to GC–MS for measuring VOCs at respective concentrations. It is worth noting that GC–MS can analyze higher concentrations (>2000 ppbV), whereas AROMA-VOC analyzes concentrations up to approximately 2000 ppbV (compound-dependent). For concentrations >2000 ppbV, a sample loop is required to mitigate the finite dead-volume internal to the analyzer. The use of a sample loop was not investigated in this study.

Reproducibility precision (% RSD) was evaluated by comparing results from Instrument A and Instrument B. Instrument-A was used for all measurements in this study and is primarily used for laboratory analysis of the VOCs included in this study. This instrument is routinely calibrated for all of the VOCs included in this study. Instrument-B is operated by GSI Environmental Inc. in Austin, TX and is primarily used for field screening to measure TCE and cis-DCE. A hardware component of Instrument B had recently been replaced, and the analytical method was not updated for all target analytes prior to this study. Instrument-B was also not fully recalibrated prior to this study.

The results for reproducibility precision (% RSD) are presented in Figure and are roughly 10 times higher than the repeatability precision values for each analyte calculated for Instrument A. Similarly, GC–MS reproducibility precision is also approximately 10 times higher than the repeatability precision for each analyte (Table S1).

While reproducibility precision was included for evaluation purposes, caution should be exercised when interpreting the data because Instrument B was not fully calibrated, as noted above. To understand the “off-the-shelf” operating condition of Instrument B in detail, a 5-point calibration curve was constructed by plotting Instrument B’s response (concentration in μg) against sample concentration (μg). Again, R ² represents the strength of the linear relationship between the calibration points and measured values. The R ² for Instrument B were mostly >0.90, except for 1,3-butadiene and styrene (see Table ). For most analytes tested, Instrument B produced a linear response over the range of tested concentrations; however, because calibration was not conducted for all VOCs measured, the data include some measurement uncertainty. Table S2 includes the mean detections of the three measurements collected using Instrument B of each sample.

4. Calibration Linear Equations and Correlations for AROMA-VOC at GSI Environmental Inc.

	Instrument B (11/21/23)
volatile compound	calibration curve equation	linearity (R 2)
1,3-butadiene	y = 0.2366x	0.1694
benzene	y = 0.5374x + 0.0009	0.9422
cis-1,2-dichloroethene	y = 0.514x + 0.0008	0.9320
ethylbenzene	y = 0.3804x + 0.0005	0.967
isoprene	y = 0.8323x + 0.0005	0.9299
styrene	y = 0.0006x + 0.0007	0
toluene	y = 0.5736x + 0.0009	0.9542
trichloroethylene	y = 0.4417x + 0.0011	0.9457
xylene	y = 1.7627x + 0.0009	0.9944

3.1.3. Limit of Detection (LOD), LOQ, and Limit of Blank

The limits of detection (LODs) are the lowest concentrations of an analyte that can be detected or distinguished from limit of blank (LOB) by the method with a specific level of confidence. ,, LOB is the highest measurement when no analyte is present. LOB acts as the baseline noise level of the analytical method, above which any signal can be considered to originate from the analyte. The definitions used in this study were adopted based on the methodologies employed; however, there are other practical definitions and formula that are commonly used in other applicable settings. The LOD addresses the qualitative detection of the analyte but does not indicate whether the analyte can be reliably quantified. , The LOQ is the lowest concentration or amount of analyte that can be determined quantitatively with an acceptable level of repeatability precision and trueness. , The LOD can be calculated with standard deviation of the lowest concentration sample or can be derived with empirical formula. , In this study, LOD and LOQ were calculated using eqs and (adapted from ref )­

$$\mathrm{L}\mathrm{O}\mathrm{D}=\frac{x_{\mathrm{b}\mathrm{l}}+3{\mathrm{S}\mathrm{D}}_{\mathrm{b}\mathrm{l}}}{S}$$ 1

And

$$\mathrm{L}\mathrm{O}\mathrm{Q}=\frac{x_{\mathrm{b}\mathrm{l}}+10{\mathrm{S}\mathrm{D}}_{\mathrm{b}\mathrm{l}}}{S}$$ 2

where x _bl is mean blank samples; SD_bl is the standard deviation of blank samples; and S is the slope of the calibration curve. Calculations for LOQ are typically similar to those for LOD, except the standard deviation SD_bl is multiplied by a factor of 10 instead of 3. This value is often represented as a coefficient variation, where LOQ is calculated to be no higher than % RSD of 10%. Sometimes, this value is also referred to as a signal-to-noise ratio of 10:1. The mean of blank samples was calculated from 9 measurements in 3 nonconsecutive days. To establish a baseline for background noise, calculations and values for LOB were included and are expressed as eq (retrieved from ref )­

$$\mathrm{L}\mathrm{O}\mathrm{B}=x_{\mathrm{b}\mathrm{l}}+1.645\times {\mathrm{S}\mathrm{D}}_{\mathrm{b}\mathrm{l}}$$ 3

where x _bl is mean blank samples; SD_bl is the standard deviation of blank samples.

Figure presents LOD values and indicates that AROMA-VOC exhibits a lower LOD compared to GC–MS, with a difference of at least 10 orders of magnitude. The results demonstrate that AROMA-VOC typically provides lower LOD and LOQ values than the GC–MS values reported in the literature. Additionally, when compared to the TO-15 method, our detection limits fall well within the range suggested by the TO-15 compendium.

3.

LOD and LOQ for AROMA-VOC and literature obtained GC–MS values.

3.1.4. Trueness and Recovery

Recovery is an estimation of the true value measured with the instrument using a known concentration of analyte. Regulators and analytical practitioners often refer to an acceptable range of 70–120% or 40–120%, which is a standard established by the EPA and the Food and Agricultural Organization of the United Nations (FAO) respectively. , Recovery was calculated based on guideline listed in Table , expressed as eq (retrieved from ref )­

$$\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{y}(\%)=\frac{x_{\mathrm{f}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{d}}}{x_{\mathrm{c}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{e}\mathrm{d}}}\times 100\%$$ 4

where x _found is measurement mean; x _certified is certified samples mean.

Trueness was represented by a z-score calculated with eq (retrieved from refs , , and )­

$$Z=\frac{x_{\mathrm{f}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{d}}-x_{\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{i}\mathrm{g}\mathrm{n}\mathrm{e}\mathrm{d}}}{\mathrm{S}\mathrm{D}\mathrm{P}\mathrm{A}}$$ 5

where x _found is measurement mean; x _assigned is the referenced value for the standard gas; SPDA is the standard deviation for proficiency assessment.

The z-score was evaluated as a proficiency test to assess the validity of all laboratories verified analysis methodologies. Both measurements for trueness and recovery were obtained from 11 samples of standard concentration from 6 nonconsecutive working sessions.

Table shows recoveries for 7 compounds that meet EPA guidelines, except for 1,3-butadiene and xylene. The false detections on the compounds were, however, consistent across all testing with this calibration standard, which suggested that this error was a result of instrument calibration. The quality of calibration gas standards is an important factor ensuring the accuracy and precision of the AROMA-VOC measurements. In addition, trueness compares the difference between an instrument’s measurement and the true value of the analyte concentration. The difference between instrument measurement and the true value is accounted for in the expression for z-score. Z-score provides a representation of the accuracy of instrument’s detections and implies that data is normally distributed. Z-score below or equal to 2 (Z-score ≤ 2) is a guideline that practitioners often accept for performance. , Similar to recovery, AROMA-VOC exhibits satisfactory performance for 7 compounds, with the exception of xylenes and 1,3-butadiene.

5. Recoveries and Z-Scores of the Study .

	GC–MS				AROMA-VOC
volatile compound	certified value (μg/L)	measured value (μg/L)	recovery (%)	Z-score	certified value (μg/L)	measured value (μg/L)	recovery (%)	Z-score
1,3-butadiene	<4.3	0.06 ± 0.01	93.8	0.258	0.023	0.048	208.7	4.413
benzene	17.0	18.68 ± 0.33	109.9	0.729	0.035	0.040	114.3	0.268
	0.156		98.4
cis-1,2-dichloroethene					0.042	0.044	104.8	0.108
ethyl benzene	19.4	19.98 ± 0.31	103.0		0.044	0.058	131.8	1.355
isoprene	0.136		99.6		0.030	0.032	106.7	0.176
styrene	40.8	35.58 ± 0.32	87.9		0.045	0.053	117.8	0.645
toluene	11.0	11.04 ± 0.19	100.4	0.19	0.039	0.046	117.9	0.674
	0.184		90.3
trichloroethylene	75.2	80.59 ± 1.71	107.2	0.796	0.060	0.066	110.0	0.390
xylene	21.0	20.50 ± 0.19	97.6	0.745	0.137	0.291	212.4	4.638
	0.212		94.2

^aNotes: GC–MS values were obtained from literature. ,

3.2. Importance of QA/QC

There are limited established QA/QC frameworks specifically designed for real-time environmental instruments. In 2000, the EPA developed a protocol for validating the HAPSITE on similar metrics, such as precision, reproducibility, quantification limits, etc. , Additionally, the protocols suggested similar evaluation criteria such as setting % RSD of less than 20%–30%, allowing for concentration deviation within 35%, and recovery of test results to be between 75 and 125%. However, a specific validation approach has not been widely adapted for other real-time instruments despite their increasing use in environmental monitoring.

The systematic QA/QC framework that we present as part of this research aims to set a comprehensive, technology-agnostic foundation, focusing on precision, accuracy, calibration, detection limits, and related quality control metrics. Unlike the EPA developed protocols, we apply the framework to a non-GC based tool, AROMA-VOC, which uses CRDS to analyze 9 compounds with potential risks to the environment. This framework aims to provide adaptable validation criteria to ensure data comparability with traditional lab-based analytical systems, such as GC–MS, while also being robust enough to handle varying field conditions and sensor platforms.

Implementing this framework allows us to establish a QA/QC standard that will serve as a basis for the future development and deployment of real-time monitoring instruments, ultimately enhancing data reliability in a variety of environmental contexts and technologies. This approach is consistent with the EPA’s objectives from the HAPSITE protocol but also extends its scope on real-time deployment to promote consistent QA/QC standards in site investigations.

4. Other Applications and Future Studies

Overall, AROMA-VOC demonstrated a satisfactory excellent performance. The overall precision % RSD is within the EPA designated range. The LOD for AROMA-VOC is 10 times lower than that for GC–MS. One significant limitation for AROMA-VOC is the suite of chemicals that can be analyzed. The current model included in this study is a beta version of AROMA-VOC, which was first manufactured in 2020. This model can analyze only up to nine chemical compounds, which is limited compared to GC–MS techniques. AROMA-VOC’s user-friendliness renders it an accessible option for a broader range of users and applications, potentially improving real-time monitoring capabilities and reducing the necessity for specialized knowledge and extensive knowledge.

Similar to other real-time instruments, the AROMA-VOC provides adaptability in a variety of applications. Water sample testing can be achieved by sparging gas into the samples, while the samples are gradually heated according to a standard procedure. This process enables precise detection and measurement of VOCs by volatilizing VOCs from liquid phase to the vapor phase. This application is useful in scenarios where rapid, real-time monitoring is required. Future research should focus on developing standard methods for water sampling using AROMA-VOC. Extensive field testing and development of robust standard procedures will be part of future studies to validate the system’s performance in various environmental conditions and operations to ensure its practical applicability.

Glossary

Nomenclature

AWG

American Welding & Gas

AROMA-VOC

autonomous rugged optical multigas analyzer for VOCs

cis-DCE

cis-1,2-dichloroethylene

CRDS

cavity ring-down spectroscopy

CWA

chemical warfare agent

DI-MS

direct injection-mass spectrometry

DQO

data quality objectives

EPA

United States Environmental Protection Agency

FAO

Food and Agricultural Organization of the United Nations

GC-MS

gas chromatography–mass spectrometer

GC-PID

gas chromatography-photoionization detector

LOB

limit of blank

LOD

limit of detection

LOQ

limit of quantification

MFC

mass flow controller

ppbV

parts-per-billion volume

PTR-MS

proton transfer reaction mass spectrometry

QA/QC

quality assurance/quality control

RSD

relative standard deviation

SIFT-MS

selected ion flow tube mass spectrometry

TAGA

trace atmospheric gas analyzer

TCE

trichloroethylene

TIC

toxic industrial chemicals

VOC

volatile organic compound

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c10634.

• Two separate tables and a figure, one presenting concentration data analyzed using a different model of AROMA-VOC instrument, owned by GSI Environmental, Inc., and another comparing the precision values and detection limits of the AROMA-VOC with GC–MS values reported in the literature, and an experimental design and setup (PDF)

The authors declare no competing financial interest.