Elsevier

Forensic Chemistry

Volume 22, March 2021, 100301
Forensic Chemistry

Characterization of a headspace sampling method with a five-component diesel fuel surrogate

https://doi.org/10.1016/j.forc.2020.100301Get rights and content

Highlights

  • A dynamic headspace concentration method was characterized with diesel surrogate.

  • Stable vaporization of diesel surrogate was achieved at or below 1.5 scc/min.

  • Recovery of low-volatility octadecane was associated with breakthrough.

  • Wood debris shifted carbon number and class distribution of diesel headspace.

Abstract

PLOT-cryoadsorption concentrates headspace vapors by sweeping them through porous layer open tubular (PLOT) capillaries chilled to 0 °C to promote adsorption onto an alumina layer. Compared to passive headspace concentration employing activated charcoal strips (ACSs) as adsorbents, PLOT-cryoadsorption has several potential advantages for extracting ignitable liquid (IL) residue from fire debris evidence. Capillary vapor traps can be monitored for breakthrough to avoid problematic displacement that occurs when an adsorbent becomes saturated. Capillaries can also be eluted with acetone to avoid highly toxic carbon disulfide. Future adoption of PLOT-cryoadsorption for fire debris analysis requires investigation of sampling parameters including collection volume, flow rate, and temperature. Here we employ a simple five-component surrogate for diesel fuel (that is, a simulated or artificial diesel fuel) to study the effect of flow rate and sample temperature on the composition and spatial distribution of the collected headspace. Flow rates at or below 1.5 scc (standard cubic centimeters)/min yielded the most repeatable results. Across the range 60 °C to 120 °C, we found that high sample temperatures allowed for the collection of more total sample and shifted its composition towards lower-volatility components. Consistent with chromatographic theory, higher-volatility components traveled furthest along the length of the capillary and were more prone to breakthrough. We conclude that the surrogate mixture allowed for effective, quantitative comparisons between sampling conditions that could be translated to real diesel fuel.

Introduction

Arson remains a challenging crime to investigate because most of the physical evidence is destroyed by the fire and by fire-fighting efforts. National arson rates are unknown due to irregular reporting, but 45% of reported arson offenses involved structures (2017 data) [1]. For structural fires, fire investigations often involve the collection of solid debris for subsequent laboratory analysis to extract and identify ignitable liquid (IL), which can indicate arson. The Ignitable Liquids Reference Collection Database provides chromatographic data and classification according to ASTM Standard Test Method E1618 for neat, weathered, and biologically degraded ILs [2], [3]. There have also been efforts to infer the specific source of neat and weathered gasolines [4].

Analyzing fire debris for IL residue requires extracting and concentrating IL compounds from debris into a form suitable for analytical instruments. ASTM Standard Practice E1412 describes the use of activated charcoal strips (ACSs) for passive headspace concentration and was originally approved in 1991; the method is well established in current forensic practice and is commonly used by forensic laboratories in the United States. The typical procedure is to sample headspace vapors by suspending an ACS above the fire debris in a steel can for 16 h to 24 h at 60 °C to 80 °C. Analytes are then desorbed from the ACS with solvent and analyzed by gas chromatography with mass spectrometry (GC–MS). The effects of time, temperature, and strip size on the recovery of IL compounds have been investigated over more than 20 years [5], [6]. If the results suggest displacement has occurred, the fire debris headspace may be re-sampled under different conditions, for example, for a shorter time and/or at a lower temperature. Displacement refers to the substitution of high-volatility compounds with low-volatility compounds at adsorbent sites. This process is driven by equilibration of each compound between the adsorbent and the headspace and occurs when the ACS becomes saturated due to high IL concentration (not easily controlled, depends on the debris) and/or small adsorbent capacity (controlled, depends on the analyst) [7].

ACSs have a high affinity for hydrocarbons, which makes them effective adsorbents but demands a strong elution solvent. Carbon disulfide is usually the solvent employed, because its extreme affinity for carbon allows it to effectively displace IL compounds. The significant drawback of carbon disulfide is its high toxicity and flammability. Small exposures to carbon disulfide can cause cardiovascular, neurological, and reproductive toxicity. Accelerated solvent extraction with acetone is a proposed alternative to carbon disulfide for volatile organic compounds captured during air quality measurements in occupational environments; however, this approach has not been applied to less volatile target compounds found in ILs such as gasoline or diesel fuel [8]. Other solvents have been evaluated for their ability to desorb target compounds found in gasoline and diesel fuel towards the goal of improving laboratory safety [9]. ASTM Standard Practice E1412 permits diethyl ether and pentane as alternatives to carbon disulfide; however, both solvents are less effective at desorbing gasoline and diesel fuel compounds than carbon disulfide and are not commonly employed by forensic laboratories [5], [10], [11]. Dichloromethane has been proposed as a safer alternative to carbon disulfide, despite both its carcinogenicity and inferior desorption of target compounds from gasoline and diesel fuel [11].

There have been efforts to develop passive headspace concentration methods that avoid carbon disulfide in favor of thermal desorption [12]. Solid-phase microextraction (SPME) utilizes fibers coated with an adsorbent phase, often polydimethylsiloxane, that are thermally desorbed directly into a gas chromatograph for analysis [13]. ASTM Standard Practice E2154 describes SPME passive headspace concentration as suitable for screening fire debris samples and was originally approved in 2001 [14]. SPME fibers are more prone to displacement than ACSs due to the lower number of adsorption sites, and SPME fibers may preferentially collect aliphatic or aromatic compounds depending on the adsorbent phase and temperature [15], [16] because of competitive adsorption. The relative concentration of a compound equilibrated between two phases is described by its partition coefficient, but even if this value can be predicted at the sample temperature [17], relating adsorbent concentration to vapor concentration is impractical for complex mixtures. Headspace sorptive extraction (HSSE) is similar in concept but utilizes polydimethylsiloxane stir bars to provide approximately 100x the adsorbent volume [18]. The method appears promising and should be less prone to displacement, but it requires additional validation. Common among these passive concentration methods is the issue of competitive adsorption or displacement. Capillary microextraction of volatiles (CMV), utilizing glass fiber filters modified with a sol–gel adsorbent phase [19], was recently employed to extract IL compounds from simulated fire debris [20]. CMV fibers are like SPME fibers in that both can be thermally desorbed; however, CMV filters are robust and can be used for dynamic headspace concentration.

Displacement is less of a concern for dynamic headspace concentration methods because they can be monitored for breakthrough. PLOT (porous layer open tubular)-cryoadsorption (PLOT-cryo) was developed at NIST in 2008 to concentrate vapors from low-volatility explosives and identify compounds in the headspace of these materials [21], [22]. This headspace concentration method sweeps headspace vapor through a capillary vapor “trap” coated with a porous alumina adsorbent layer that is cooled to promote adsorption. Breakthrough may be a sign that the adsorbent is becoming saturated (leading to displacement), but it can also indicate that the flow rate is too high to permit adsorption of all compounds onto the porous alumina. In either scenario, breakthrough can be detected by bubbling the vapor exiting the capillary through a vial of solvent and subsequently analyzing this solvent for the presence of analytes. Capillaries are easily eluted with solvent and the eluate can be analyzed by GC–MS or other analytical techniques. Over the last decade, PLOT-cryo has been applied to detection of grave soil, poultry spoilage, and natural gas contaminants, among other applications [23], [24], [25]. The portable PLOT-cryo instrument developed in 2014 was first applied to detect diesel fuel spiked on glass beads (with a detection limit below 1 ppm) and more recently used to detect explosives- and decomposition-related compounds inside a simulated shipping container [26], [27], [28].

In a preliminary investigation of its application to fire debris, PLOT-cryo was shown to extract representative IL compounds from gasoline and diesel fuel from laboratory-generated fire debris [29]. Several advantages stood out from that work. Unlike passive headspace concentration methods in which the distribution of compounds recovered may not represent the distribution of compounds in the headspace vapor due to competitive adsorption or displacement, headspace flow through a capillary vapor trap cooled to sub-ambient temperatures promoted adsorption of all IL compounds by the alumina adsorbent. Because alumina is a weaker adsorbent than carbon, capillaries could be completely eluted with acetone, which is a respiratory irritant, but still a much safer solvent than carbon disulfide. Solvent flow through the capillary also promotes complete desorption compared to equilibrium desorption with the same solvent. Alumina’s ability to be reused is another advantage, while carbon is considered disposable because of the difficulty of verifying complete sample removal. In our 2014 paper, PLOT-cryo was directly compared to the ACS method and found to collect a larger number of compounds on the low-volatility end of each fuel; however, we must note that due to safety concerns, the student researcher eluted ACSs with acetone instead of carbon disulfide. This choice surely had a negative impact on ACS performance, and additional head-to-head experiments are needed before any definitive conclusion can be made.

PLOT-cryo shares features with ASTM Standard Practices E3189-19 and E1413-19 for static and dynamic headspace concentration of IL residues onto adsorbent tubes [30], [31]. In the 2014 fire debris work, we used a flow rate of 1.3 mL/min, which is close to, but still less than, the recommendation for static headspace concentration: 2 mL/min to 80 mL/min. Since then, however, we have employed carrier gas flow rates of up to 10 mL/min in other work with pure compounds and rates up to 100 mL/min using the portable instrument [26], [43]. On the other hand, the collection volume may be greater than 10% of the headspace volume (guidance that separates static from dynamic headspace concentration), although this depends on the sample container. In the 2014 paper, we used a 2 mL sample vial and collected approximately 40 mL of headspace. This collection volume represents a small fraction of the volume of a quart (less than 5%) or gallon-sized steel can (~1%), which are commonly used to collect fire debris. However, this calculation does not account for the volume occupied by the debris, which typically occupies one-quarter to one-half of the can volume. PLOT-cryo therefore falls somewhere between the ASTM definitions for static and dynamic headspace concentration regarding flow rate and collection volume. Regarding temperature, samples in the 2014 paper were heated to 175 °C, which is higher than recommended in any ASTM standard for headspace concentration of IL residues. Although high sample temperatures generate rapid results, temperatures above 90 °C are unsafe for moist samples because the pressure exerted by water vapor can cause sample containers to violently fail. PLOT-cryo also utilizes a chilled adsorbent to enhance trapping, which is not recommended in any ASTM standard and is a key feature of the PLOT-cryo method.

Our primary goal here is to determine to what extent the PLOT-cryo method might be optimized with a surrogate mixture or “artificial” diesel fuel by examining the influence of flow rate and temperature on the composition and spatial distribution of the collected headspace compounds. Surrogate or artificial mixtures have been used in the past primarily for engine testing and have been designed to approximate kinetics, combustion, and certain thermophysical characteristics important in that application, such as volatility, viscosity, and energy content [32], [33]. In a recent application to forensic science, researchers used a seven-component artificial gasoline mixture to study the temperature dependence of IL weathering behavior [34], [35]. To investigate variables contributing to distortion of the collected headspace compounds related to ACS saturation, Williams et al. employed a simple hydrocarbon mixture [7]. The mixture was composed of five hydrocarbons selected for their relative volatilities and was an effective choice for the questions being investigated, but it was not intended to be a surrogate (i.e., it did not represent the thermophysical properties of any fuel). We propose that by selecting a thermodynamically representative surrogate mixture we will not only simplify the analysis of trends, but the method development will also be transferable to the more complex target fuel.

To investigate the fitness of surrogate mixtures for this purpose, we first selected a five-component diesel surrogate comprised of compounds with a range of volatilities (described in the next section). While varying the flow rate (0.5 scc/min – 2.5 scc/min) and sample temperature (60 °C – 120 °C), we examined PLOT-cryo collections from neat diesel surrogate and neat diesel fuel to compare the total quantity of collected headspace as well as trends in the volatility and hydrocarbon class. Our findings indicate that the surrogate was a useful tool for predicting the effects of variable sampling parameters on the diesel fuel collections, but it had limitations associated with its simplicity.

Section snippets

Materials

Table 1 shows the composition of the diesel surrogate. It contains n-octadecane (OD) to represent alkanes, 2,2,4,4,6,8,8-heptamethylnonane (HMN) to represent isoalkanes, 1,2,3,4-tetrahydronaphthalene or tetralin (THN) to represent naphthoaromatics, 1,2,4-trimethylbenzene (TMB) to represent alkylbenzenes, and 1-methyl naphthalene (MN) to represent polynuclear aromatics [33]. We chose this five-component surrogate because its distillation curve has the lowest average difference from the target

Influence of flow rate

In general, dynamic headspace concentration cannot be assumed to be an equilibrium process, because vapor–liquid equilibrium (VLE) cannot be established if the vapor is too rapidly removed from the headspace. However, it is possible under certain conditions. We have previously demonstrated that at flow rates up to 10 scc/min, vapor–liquid or vapor–solid equilibrium can be established for pure compounds, permitting the determination of vapor pressure as a function of temperature [22], [42], [43]

Conclusion

Surrogate mixtures designed to represent complex fluids with fewer components have long been used to simplify fuels research, but their potential contributions to forensic science have not yet been fully explored. We demonstrated here that a five-component diesel fuel surrogate was a valuable tool for developing PLOT-cryoadsorption for headspace concentration of ILs from fire debris. The thousands of components in diesel fuel lead to data analysis challenges, especially when looking for subtle

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

M. E. Harries was supported by a National Research Council (NRC) Postdoctoral Research Associateship and S. S. Wasserman was supported by a Summer Undergraduate Research Fellowship (SURF). This research was supported by funding from the NIST Special Programs Office and by funding from the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice through an interagency agreement (DJO-NIJ-19-RO-0007). The funders had no role in study design, data collection and

References (48)

  • M.E. Harries et al.

    Field demonstration of portable vapor sampling in a simulated cargo container

    Forensic Chem.

    (2019)
  • J.E. Nichols et al.

    Analysis of arson fire debris by low temperature dynamic headspace adsorption porous layer open tubular columns

    J. Chromatogr. A

    (2014)
  • H.L. Birks et al.

    The surprising effect of temperature on the weathering of gasoline

    Forensic Chem.

    (2017)
  • I.C. Willis et al.

    Weathering of ignitable liquids at elevated temperatures: a thermodynamic model, based on laws of ideal solutions, to predict weathering in structure fires

    Forensic Chem.

    (2020)
  • T.M. Lovestead et al.

    Determination of cannabinoid vapor pressures to aid in vapor phase detection of intoxication

    Forensic Chem.

    (2017)
  • U.S. Department of Justice Federal Bureau of Investigation, Uniform Crime Report: Crime in the United States 2017,...
  • National Center for Forensic Science, University of Central Florida, Ignitable Liquids Reference Collection Database,...
  • American Society for Testing and Materials, ASTM E1618-19: Standard test method for ignitable liquid residues in...
  • American Society for Testing and Materials, ASTM E1412-16: Standard practice for separation of ignitable liquid...
  • R.T. Newman et al.

    The use of activated charcoal strips for fire debris extractions by passive diffusion. Part 1: The effects of time, temperature, strip size, and sample concentration

    J. Forensic Sci.

    (1996)
  • M.R. Williams et al.

    Adsorption saturation and chromatographic distortion effects on passive headspace sampling with activated charcoal in fire debris analysis

    J. Forensic Sci.

    (2005)
  • G. Fabrizi et al.

    Occupational exposure to complex mixtures of volatile organic compounds in ambient air: desorption from activated charcoal using accelerated solvent extraction can replace carbon disulfide?

    Anal Bioanal Chem

    (2013)
  • J.J. Lentini et al.

    Comparison of the eluting efficiency of carbon disulfide with diethyl ether: the case for laboratory safety

    J. Forensic Sci.

    (1997)
  • R.T. Newman, J.A. Dolan, Solvent options for the desorption of activated carcoal in fire debris analysis. Proceedings...
  • Cited by (0)

    Contribution of the National Institute of Standards and Technology; not subject to copyright in the United States.

    View full text