Decontamination options for indoor surfaces contaminated with realistic fentanyl preparations

Abstract

The significant increase in illegal use of the synthetic opioid fentanyl is leading to unintentional overdose fatalities. Spills of fentanyl where it is abused or prepared for illegal distribution can result in persistent contamination of areas. Remediation can be attempted through physical removal but may benefit greatly from application of decontamination solutions that provide in-situ degradation of fentanyl. This work investigates the efficacy of decontamination technologies for degradation of fentanyl-HCl on indoor surfaces. Decontamination studies were conducted to evaluate the oxidative degradation of fentanyl based on percarbonate, hydrogen peroxide, peracetic acid, and chlorine (bleach) chemistries. This study utilized an experimental design relevant to field operations to provide direct information to first or hazardous materials responders and providers of environmental fentanyl remediation services, who may otherwise rely on unverified approaches. Across a range of nonporous indoor surfaces, results suggest that water (with or without detergent) spraying alone can physically remove 70–90% of fentanyl (with all fentanyl recovered in runoff). In nearly all cases, the spray application of peracetic acid or acetified bleach oxidants resulted in statistically significant degradation of fentanyl (>95% reduction), with noticeably lower efficacy for other oxidants (e.g., pH neutral bleach and OxiClean^™). The decontamination efficacy was significantly reduced upon the addition of cutting agents that competed for oxidant demand.

1. Introduction

Fentanyl, a commonly abused substance in the class of drugs knowns as opioids, is a synthetic, short-acting opioid analgesic that is 80–100 times more potent than morphine (See https://www.dea.gov/s, 2020). The horrific personal and societal impacts of opioid abuse are the topic of much discussion, including how other public health crises such as the COVID-19 pandemic (Volkow, 2020) can amplify it. From a hazardous material perspective, it is important to understand how to clean up fentanyl spills (entanyl and Fac, 2018; Fentanyl Remediation: Gui, 2020), which can impact laboratory personnel analyzing fentanyl samples, first responders, and even the public who may unknowingly encounter these spills. The need for cleanup approaches is driven by occupational concerns, although the same principles can help avoid extending opioid exposure to the unsuspecting public.

Fentanyl cleanup efforts are reported at scientific conferences (Doerflein, 2019), by public health services (See https://www.albertahe, 2020), and in the media (https://globalnews.ca/new, 2020) and are connected to spills that resulted from a failure to contain the fentanyl, either from incorrect operation or design. Once the controls are non-functional, the release can be totally uncontrolled. Practices for cleanup of fentanyl have benefitted from the availability of guidance for other drugs like methamphetamine (Voluntary Guidelines for, 2013), but there are significant differences and challenges resulting from both the chemical properties of fentanyl and its extreme toxicity.

A recent review (Bazley et al., 2020) highlights some of the technical challenges of fentanyl cleanup activities, the paucity of scientific literature on basic aspects of this topic, and the greater lack of scientific studies geared toward the development of guidance for people tasked to clean up fentanyl spills. Operational cleanup challenges include the need to always consider the safety of the responders; the recognition that a fentanyl spill may have spread throughout a property as a fine dust or powder; and the need to clean every surface. These challenges lead to cleanup activities that can take days to weeks to complete, depending on size and level of initial contamination (Fentanyl Lab Cleanup and, 2017). Some cleanup applications, like forensic laboratories (Froelich et al., 2018; Sisco et al., 2019), are focused on specific needs. Such laboratories are designed for regular cleaning of highly toxic substances; most fentanyl spills are not in these areas but rather are inside vehicles, residences, commercial buildings, and at likely significantly higher amounts.

Because cost is often a main driver in cleanup, techniques usually involve a combination of physical removal (e.g., HEPA vacuuming, detergent/water washing, etc.) and chemical degradation (e.g., the application of common oxidants like chlorine bleach). However, the recent review (Bazley et al., 2020) noted that relative contributions of physical removal versus chemical degradation during cleanup is infrequently studied directly, and these relative contributions can be a confounding factor in interpreting field cleanup data. Further, even basic issues such as the effect on cleanup approaches of other substances combined with fentanyl within illegal products (e.g., co-contaminants like cutting agents and adulterants) have not been reported in scientific literature.

The purpose of this paper is to report results of controlled laboratory studies relevant to fentanyl cleanup from frequently encountered surface types inside buildings. Fentanyl contaminated building materials were decontaminated through application of various commercially available decontamination solutions by spray under realistic field conditions. After a fixed contact time, the building materials were extracted for quantification of residual fentanyl. Runoff from the materials was also analyzed for residual fentanyl. Decontamination efficacies were derived from fentanyl amounts recovered from materials against fentanyl amounts recovered from materials that were not exposed to the decontamination solution. The design of these experiments is intended to result in information that can directly be applied to help address cleanup of spills, including co-contaminants of fentanyl.

2. Materials and methods

2.1. Indoor surface coupons

Decontamination efficacy testing was conducted using the following types of indoor materials: stainless steel, laminate, acrylic, and painted drywall. Relevant properties of these materials can be found in the Supplementary Material, Table SM-1. Material coupons were cut from large sections or panels to a uniform length (4.0 cm) and width (2.5 cm), equal to a 10-cm² surface area. Additional material coupons of medium size (30 cm by 10 cm) were cut from laminate and heavy-duty vinyl upholstery. Use of these larger coupons allowed for wipe sampling of the surface to determine residual fentanyl instead of extraction of the material coupon.

2.2. Application of fentanyl and additives to coupons

Fentanyl-HCl powder (CAS 1443-54-5, Cayman Chemical, Ann Arbor, MI, USA, >99% purity) was utilized as received. Particle size distribution analyses were performed on the purchased chemical using an aerodynamic particle sizer, as described in Supplementary Material, Fig. SM-1. The fentanyl-HCl salt was characterized as having a mean (±standard deviation; n = 3) aerodynamic diameter of 7.3 ± 2.1 μm by number and 15 ± 1.4 μm by mass with a bimodal size distribution with aerodynamic diameter peaks at 2.5 and 10 μm. Fentanyl-HCl was applied to surfaces as a solid salt consisting of fine particles using a 50-μL Drummond Series 500 Digital Microdispenser (3-000-550, Drummond, Broomall, PA, USA). To avoid compaction, the powder was stirred regularly with the pipette tip. The fentanyl mass application target was 1 mg. Following application of fentanyl onto the surface of a coupon, the solid was spread using an anti-static spatula (14-245-99, Fisher Scientific, Pittsburgh, PA). Fentanyl-HCl was spread as evenly as possible (visually) across at least 50% of the coupon surface area, It was then allowed to stay on the surface for 60 min before the decontamination solution was applied.

Toxicologically benign substances, which are used as cutting agents (hereafter referred to as additives) in some illicit drug preparations (Fentanyl Signature Profil, 2019; Methods for Impurity Prof, 2005), consisted of lactose (anhydrous, PHR1025–1G, Millipore Sigma, St. Louis, MO), mannitol (PHR1007–1G, Millipore Sigma, St. Louis, MO), and ascorbic acid (PHR1008–2G, Millipore Sigma, St. Louis, MO). Additives were applied to the material surface so that the fentanyl concentration equaled 5% of total mass of solids applied, which was the average fentanyl purity of 100–999 g size fentanyl seizures in 2019 (Fentanyl Signature Profil, 2019). Here, this corresponded to a 19-mg addition of an additive to the 1 mg of fentanyl-HCl on the material coupon.

2.3. Decontamination solutions

Decontamination efficacy studies were conducted using commercially available products with occasional modifications to adjust pH. These products covered percarbonate, chlorine, hydrogen peroxide, and peracetic acid chemistries and were expected to be suitable oxidants for fentanyl (Qi et al., 2011). Table 1 provides information on the decontaminants, active ingredients, measured pH, and active ingredient concentrations. All products were prepared immediately prior to use and per manufacturer instructions. Specific information on the preparation of all decontamination solutions can be found in the Supplementary Material. Laboratory tap water (West Jefferson, OH) was included in the study as a reference solution. OxiClean^™ was included based on reported degradation of fentanyl and fentanyl analogs (Froelich et al., 2018; Sisco et al., 2019). Chlorine-based oxidants are frequently used as decontamination solutions for surfaces of highly toxic substances like chemical warfare agents (econtamination, 2010; Love et al., 2011; Stone et al., 2015) and mis- or overused pesticides (Oudejans et al., 2020). The active species varies with pH and includes hypochlorite ion and hypochlorous acid. Based on the pH dependence of the solubility of fentanyl, which has a reported pKa of 8.99 (Roy and Flynn, 1989), bleach was adjusted to pH 7 (neutral), as well as a further acidified bleach (pH = 5) that is similar in pH to the hypochlorite solution tested by Qi et al. (2011) in their stirred reactor fentanyl degradation study. For the pH 5 acidified bleach, droplets were visually observed to bead on some of the material surfaces, leading to the inclusion of a third chlorine bleach solution that contained surfactants, acidified to pH 5 and having a free available chlorine concentration equal to that of the other pH 5 bleach solution. Both the DF200 and Dahlgren Decon^™ products are being publicized for their effective degradation of fentanyl (Bazley et al., 2020). However, independent verification of their claims under realistic field conditions, especially for decontamination of surfaces, is lacking. Because these products are being considered for purchase by responders, they were investigated as prepared as specified by the manufacturers, i.e., the individual components were not investigated separately, a meritorious subject beyond the scope of this paper.

Table 1.

Decontamination Solutions, Active Ingredients, Concentrations, and pH.

Decontaminant	Part No./manufacturer	Active ingredient(s)	Concentration of chemical	PH
Tap water	N/A	None	N/A	6.9
OxiClean™ Versatile stain remover	Church & Dwight	Percarbonate and hydrogen peroxide	0.37% H2O2	10.9
Bleach – pH 7	SP5101-4, K–O–K Products, Inc.	Chlorine	0.55–0.59%, FAC	6.7–6.8
Bleach – pH 5	SP5101-4, K–O–K Products, Inc.	Chlorine	0.51% FAC	5.1
Bleach – pH 5 with surfactants	Clorox™ ProResults® garage and driveway cleaner	Chlorine	0.53% FAC	4.9
EasyDECON® DF200 [DF200]	Intelagard, Inc.	Hydrogen peroxide and peroxygen	3.8–4.3% H2O2	9.4
Dahlgren Decon™	First Line Technology	Peracetic acid	1.4–1.7% PAA	7.3

FAC: Free Available Chlorine, PAA: peracetic acid, H₂O₂: hydrogen peroxide.

2.4. Experimental setup

The distribution of the solid fentanyl-HCl salt across the material surface required a decontamination solution application approach that covered the complete surface, as opposed to a more common approach of applying of the decontamination solution as a single droplet covering the localized contaminant. Here, decontamination solutions were applied via a moderately low flow spray using a typical pump pressurization-style sprayer (12U469, Grainger, Lake Forest, IL). Integration of the sprayer into a glovebox-style test chamber involved removing the sprayer’s shut-off valve and nozzle assembly from the extension wand. The valve and nozzle were attached to each end of a chemical-resistant Versilon^™ polyvinyl chloride (PVC) tubing lined with fluorinated ethylene propylene (FEP, 6519T14, McMaster-Carr, Aurora, OH) that was run into the test chamber through a port on the side wall. As illustrated in Fig. 1, the nozzle assembly was mounted to a rail installed at the top of the test chamber that allowed the spray delivered from the nozzle to be swept from side to side (left to right in Fig. 1). A motorized pulley system was used to move the nozzle across the rail at a uniform and constant rate. The speed with which the nozzle moved was adjusted to deliver 60 μL per cm² reproducibly to the coupon surfaces at a constant 20 psi sprayer pressure. This amount was implemented as it was the practical decontaminant application rate from backpack sprayers used by responders in a full-scale field study in which 60 L of decontamination solution were used to spray a 100 m² surface area (io-Response Op, 2013).

Fig. 1.

Sprayer and test chamber setup.

Test and procedural blank coupons were placed into separate acrylic boxes (4.4 cm square by 2.5 cm height; part no. 3790-CL, G&G Distributors, Saddle Brook, NJ) on top of a small polypropylene plastic (PP) mesh disk (3.5 cm diameter, 1.3 mm thickness; cut from larger material, part no. 9265T47, McMaster-Carr, Aurora, OH). The placement of coupons in individual boxes allowed for the quantification of fentanyl remaining on the coupon surface and in the decontaminant runoff from the coupon surface. The PP mesh disk was included to prevent contact between the underside of the material and decontamination solution that was collected in the box following spray application. The acrylic boxes holding individual coupons were arranged on a tray in two rows of eight boxes. The distance from the nozzle outlet to the top surface of the coupons/panels placed underneath was approximately 26 cm and was set such that the spray fan/cone delivered from the nozzle extended past the outer edges of the plastic boxes placed on the tray below.

2.5. Decontaminant spraying and sample collection procedure

For the small-size coupons, the 2 × 8 layout of materials allowed for a single spray application of decontaminant to four materials (all in triplicate, plus a single procedural blank per material). For the medium-size coupons, each material was sprayed individually. Decontaminant was sprayed and, after a 60-min decontaminant contact period, residual fentanyl was recovered by either extractive sampling of the small (10-cm² surface area) coupons or wipe sampling for the medium-size coupons (for which extractive sampling was experimentally impractical). Small coupons were transferred into 60-mL glass jars (05-719-257, Fisher Scientific, Pittsburgh, PA, USA) containing 10 mL of isopropyl alcohol (IPA) and 5 mL of 3 M sodium thiosulfate (STS) solution. Decontaminant that ran off the test or procedural blank coupon surfaces following spray delivery, as well as the overspray, was collected in the acrylic boxes holding the coupons and extracted in the same manner as the coupons, except that the extraction volumes were 20 mL of IPA and 5 mL of 3 M STS solution. Extraction jars were swirled for 10 s and sonicated for 10 min. This amount of IPA led to layers, which were allowed to separate. Aliquots from the IPA layer of approximately 0.5 mL from each extraction jar were transferred to individual vials and sealed (21140 [vial], 24,670 [vial cap], Fisher Scientific [Restek Corp.], Hanover Park, IL). Extracts were stored at −20 ± 10 °C temperatures if not analyzed on the same day.

In separate experiments (results not shown), the described sampling procedures were verified for the presence of residual active ingredients to avoid continued degradation. This would otherwise critically bias the recovered fentanyl mass, leading to artificially higher decontamination efficacy values than those associated with the degradation of fentanyl on the surface in the presence of the decontaminant. The procedure also enabled three-days storage at −20 °C (longer times were not studied) without impacting quantitation when immediate analysis was not possible.

Medium-size coupons were sampled via surface wiping using lint-free 5 cm by 5 cm four-ply rayon/polyester blend (gauze) sponges (22-037-921, Fisher Scientific, Pittsburgh, PA) with a sampling method demonstrated for pesticide sampling (Willison et al., 2019). Extraction efficiencies, defined as the percent ratio between recovered fentanyl mass from a coupon material and the ratio from a spike control into the same extraction solvent, were determined for all materials prior to the decontamination studies. All recoveries (see Table SM-2, Supplementary Material) using the extraction method exceeded the initial quality control recovery target values (better than 70%, but not higher than 120%, recovered mass) with a low coefficient of variation among test coupons (less than 30%). For medium-size coupons, the first wipe was a dry wipe to absorb any remaining decontamination solution followed by two wipes semi-saturated with isopropyl alcohol (chosen for consistency with the coupon extraction approach). Total fentanyl recoveries after three wipes were 93 and 94% with respect to the spike control recoveries after wiping laminate and vinyl upholstery, respectively.

2.6. Fentanyl and reaction byproduct analysis

An AB Sciex 5500 triple quadrupole mass spectrometer (SCIEX, Framingham, MA) coupled to a Shimadzu 20 XR series LC (Shimadzu, Columbia, MD) was used for liquid chromatography/tandem mass spectrometry (LC-MS/MS) sample analyses. Fentanyl was quantitated via isotope dilution in sample extracts using a reversed-phase high performance liquid chromatography (HPLC) method and multiple reaction monitoring (MRM). Fentanyl-d₅ (F-001–1ML, Sigma-Aldrich, St. Louis, MO) was added as the internal standard (IS) to calibration standards, controls, and test samples just prior to LC-MS/MS analysis (following dilution in water, as needed). The nominal IS concentration in samples after addition was 0.45 ng/mL. The lower limit of quantitation (LLOQ) for fentanyl free base was set at 0.010 ng/mL, equal to the concentration of the lowest calibration standard (LCS) used to generate the calibration curve. The signal-to-noise ratio for the LCS was approximately 30:1. This LLOQ translates to an extraction sampling method quantitation limit for fentanyl-HCl of 1.1 ng/coupon or 0.11 ng/cm². Samples that calculated to be below the LCS concentration, or displayed area counts below that of the LCS, were reported as less than the LLOQ; i.e., <0.01 ng/mL). Samples that quantitated above the highest calibration standard were diluted with IPA and reanalyzed. Details on the LC-MS/MS method are provided in the Supplementary Material. Samples (e.g., aliquots of decontaminant runoff) were matrix-matched to the calibration standards by addition of IPA to a final concentration of approximately 10%. Alternative dilution factors were used for samples of high analyte concentration or to reduce sample matrix effects (such as observed in the presence of residual Dahlgren Decon^™ decontaminant).

Additional analyses of selected samples were performed to identify byproducts produced during decontamination of fentanyl by two of the test decontaminants. Qualitative byproduct analyses were performed by LC-MS/MS, LC- quadrupole time of flight (QTOF), and gas chromatography-MS (GC/MS) using test, control, and blank samples produced during decontamination of laminate coupons by Dahlgren Decon^™ and pH 5 bleach. A targeted, semi-quantitative analysis for norfentanyl was performed via LC-MS/MS with a 0.1 ng/mL norfentanyl standard (N-031–1ML, Cerilliant, Round Rock, TX). Information on the LC/QTOF and GC/MS methods can be found in the Supplementary Material.

2.7. Decontamination test matrix, calculations, and statistical analysis

Table 2 summarizes the overall decontamination test matrix. For two oxidants, additional decontamination tests were conducted that separately included the presence of one of the three different additives. Decontamination testing for the medium-size coupons was limited to two of the better performing decontaminants for laminate and vinyl upholstery, and it also included one test with an extended 4-h contact time between the decontaminant and the laminate surface.

Table 2.

Test matrix.

Decontaminant	Acrylic	Laminate	Painted drywall	Stainless steel	Laminate with additives	Vinyl upholstery
Water	X	X	X	X
OxiClean™	X	X	X	X
pH 7 Bleach	X	X	X	X
pH 5 Bleach	X	X	X	X
pH 5 Bleach with surfactant	X	X, Y	X	X	X
DF200	X	X	X	X
Dahlgren Decon™	X	X, Y	X	X	X	Y

X: Included in small coupon test; Y: Included in medium coupon size test.

Total sample mass was calculated using the masses recovered from extraction of the coupon and extraction of the associated runoff sample, according to Equation (1):

$$Mas{s}_{Tot}=Mas{s}_{Rec\left(\mathit{\text{coupon}}\right)}+Mas{s}_{Rec\left(\mathit{\text{runoff}}\right)}$$ (1)

where: MassTot = Total fentanyl mass recovered (μg), MassRec (coupon) = Fentanyl mass recovered from the coupon (μg), and MassRec (runoff) = Fentanyl mass recovered from the runoff (μg). For decontamination tests using surface-wiped, medium-size coupons, the recovered fentanyl mass is based on the total amount recovered from the three extracted wipes.

Percent efficacy of decontamination from each individual test coupon or percent total efficacy for each coupon/runoff combination was calculated according to Equation (2):

$$\mathit{\text{Efficacy}}=\left(\frac{Mas{s}_{Rec\left(pos\right)}-Mas{s}_{Tot}}{Mas{s}_{Rec\left(pos\right)}}\right)\times 100\%$$ (2)

where: MassRec (pos) = Fentanyl mass recovered from the associated positive control (μg).

The standard deviation in the mean mass recovered from test and positive control coupons was used to derive the standard deviation in decontamination efficacy through propagation of error.

A 95% confidence level was utilized for all statistical tests. A one-way ANOVA model was fitted separately to each material (acrylic, laminate, painted drywall, or stainless steel) for all positive control and test coupon sample sets, with an effect for decontaminant (Table 1) to investigate if there were significant performance differences among the different decontaminants. A second one-way ANOVA model with an effect for surface material was fitted separately to each decontaminant for all positive control and test coupon sample sets to investigate if there were significant performance differences among the different materials for each decontaminant. A Tukey’s multiple comparisons procedure was performed for the 21 possible decontaminant pairwise comparisons within each material and six possible material pairwise comparisons within each decontaminant group to investigate which pairs of decontaminants or materials had mean total mass recoveries that were significantly different from each other.

3. Results and discussion

Interaction of the fentanyl salt with the sprayed decontaminants occasionally led to visual observation of small clumps of agglomerated fentanyl, which might have reduced mass transfer and the ability of the decontaminant to degrade the fentanyl. The amount and nature of agglomerates is expected to be related to factors such as the material surface, preparation method and purity of the specific sample source. Therefore, the results and discussion below are specific to the fentanyl sample investigated but are illustrative of underlying phenomena surrounding decontamination of fentanyl preparations found in the field.

3.1. Water and OxiClean results

Fig. 2 shows the recovered mean fentanyl amounts and standard deviation following the spray application of tap water and the OxiClean^™ solution on the four coupon materials after a 1-h dwell time of the solution on the surface. For the tap water spray, a significant amount of the applied fentanyl was found remaining on the coupon surface (5–38% across all materials in respect to positive control recovery), as well as in the overspray and runoff (33–80%). While the spray pattern can be categorized as a light spray, some of the distributed fentanyl salt would run off the surface due to the physical impact of the spray with the solid fentanyl, which is only loosely associated with these smooth surfaces. Further, some of the materials are hydrophobic, so water will run off the surface more easily, taking fentanyl with it. For the water spray test, the sum of the two recovered fentanyl mass amounts was found to be not statistically significantly different from the positive control recoveries for three of the four materials (Tukey adjusted p-value>0.05), indicative, as expected, of no degradation of fentanyl in water within the limitations of this study (1-h contact time). The exception was the painted drywall material, perhaps an artifact resulting from the higher precision in the fentanyl mass recovery for the corresponding positive controls.

Fig. 2.

Fentanyl Recoveries Following Application of Water and OxiClean^™ Solution. S Steel is Stainless Steel; Drywall is Painted Drywall Paper.

Recovered fentanyl amounts following spray application of OxiClean^™ solution were found to be similar to those following the water spray application (22–50% remaining on the surface and 32–66% recovered in the runoff across all materials) and the associated positive control recoveries, indicating no appreciable fentanyl degradation outside experimental error. This finding may appear to contradict Froelich et al. (2018) and Sisco et al. (2019) for fentanyl and fentanyl analogs, in which minimal amounts of fentanyl or fentanyl analogs remained on surfaces following application of OxiClean^™. However, in contrast to the present study, which examines decontamination through physical removal via water run-off and chemical oxidation, the previous two studies employed physical removal via wipes following application of OxiClean^™. The deployed wipe would likely have retained unreacted fentanyl, as mentioned by the authors in one study (Froelich et al., 2018). Sisco et al. (2019) reported more than 80% reduction within 1 min in measured amounts for cyclopropyl fentanyl and carfentanil on wipe material in the presence of OxiClean^™. No changes in the recovered amounts were reported over the next 15 min while a 30 min contact time resulted in complete carfentanil reduction but no change for cyclopropyl fentanyl. These results may arise from a combination of the poor recovery of these fentanyl analogs from the wipes and/or differences in oxidation reaction mechanisms for different fentanyl analogs. Neither of these possibilities has been investigated and merit further study.

The OxiClean^™ results also illustrate other important aspects of surface decontamination studies. The active ingredient concentration was approximately equal to the hydrogen peroxide concentration used in the stirred reactor study by Qi et al. (2011) which reported a 50% degradation of fentanyl after 1 h. In a stirred reactor experiment, oxidation occurs in the solution phase, rather than the mechanistically more complex case for surface decontamination in which fentanyl needs to dissolve into the aqueous solution before it can be oxidized. Considering the measured pH 11 of the OxiClean^™ solution, fentanyl dissolution is likely to be a slow process and limited by the low solubility of fentanyl in water at high pH (Roy and Flynn, 1989). Indeed, the recent review article summarizing most of the fentanyl decontamination literature (Bazley et al., 2020), highlights the importance of pH in decontamination studies. In the present OxiClean^™ study, some solid particulate was visually observed on the coupon surface prior to coupon extraction. It is plausible that this was unreacted fentanyl did not dissolve in the OxiClean^™ solution (during the 1-h contact time). Accordingly, in contrast to Qi’s solution phase results (Qi et al., 2011), the OxiClean^™ solution resulted in measured concentrations of fentanyl on the surfaces that did not differ significantly from that of water, even including that unreacted fentanyl dissolved in the extraction solvent during both the water and OxiClean^™ experiments.

3.2. Chlorine-containing solution results

Fentanyl recoveries following spray application of the chlorine-containing decontamination solutions are shown in Fig. 3. Total mean fentanyl recovery (coupon extract plus runoff and overspray) following pH 7 adjusted bleach spray decontamination of all materials was significantly lower than recoveries of positive controls (Tukey-adjusted p-value <0.05). Total recoveries ranged from 223 to 392 μg across all materials versus 821–920 μg for the positive controls across the same materials. This indicates that degradation of fentanyl occurred.

Fig. 3.

Fentanyl Recoveries Following Application of Hypochlorite-based Products. S Steel is Stainless Steel; Drywall is Painted Drywall Paper.

A significant reduction in recovered fentanyl was found for the acidified bleach solution at pH 5. Fentanyl recoveries were 42–102 μg or less than 10% of the positive control recoveries across all materials. Recovered fentanyl masses were lower for the pH 5 bleach product that contained a surfactant (15–73 μg fentanyl recovered across all four materials). Both pH 5 bleach products were prepared to have similar FAC (Table 1). The surfactant containing bleach product yielded (results not shown) a more visually even distribution of the solution across the surface as opposed to the distinct droplets of pH 5 bleach on the surface. While this is suggestive of modification of surface properties and hence mechanisms, Tukey adjusted p-values for a direct comparison of pH 5 bleach and pH 5 bleach with surfactant solutions were always>0.05 for all four materials. In all, chlorine containing bleach solutions buffered to low pH (to increase solubility) show a strong capability to degrade fentanyl on these surfaces. Additional experiments are warranted to elucidate the role of surfactants in degradation of fentanyl on the surface.

3.3. DF200 and Dahlgren Decon results

Recovered mean fentanyl masses following spraying of the DF200 and Dahlgren Decon^™ products are shown in Fig. 4. Total recoveries ranged from 23 to 72 μg across all four materials following DF200 spray decontamination and ranged from 4.6 to 114 μg following Dahlgren Decon^™ spray decontamination and a 1-h dwell time. Although the fentanyl mass recovered following Dahlgren Decon^™ application was lower than after DF200 decontamination, the ANOVA test indicated that these differences were not statistically significantly different (Tukey adjusted p-value>0.05). In both studies, laminate appeared (Fig. 4) to be a more difficult to clean material. However, recovered fentanyl mass was not significantly different from that from other materials for both decontaminants (p>0.05). The wider spread in the recovered fentanyl amounts might be attributed to the observed agglomerates (of the fentanyl) in some of the runoff (DF200 decontamination) samples and on some of the laminate coupons (Dahlgren Decon decontamination).

Fig. 4.

Fentanyl Recoveries Following Application of DF200 and Dahlgren Decon^™. S Steel is Stainless Steel; Drywall is Painted Drywall Paper.

3.4. Impact of surface types on decontamination

ANOVA results for all positive controls indicated that there were no significant differences between any pairs of decontaminant sample sets for acrylic, painted drywall, and stainless steel with regard to the amount of fentanyl recovered. For laminate, there was one combination with a significant pairwise comparison: the positive control mean mass recovery for DF200 was significantly less than the positive control mean mass recovery for the pH 5 bleach decontamination. Considering that the random probability of measuring a significant difference when none truly exists is 0.05 at the 95% confidence level, this one significant difference out of 84 comparisons (twenty-one comparisons for each of the four materials) is arguably not sufficient to conclude there is a practical difference between the materials. Thus, within the boundaries of this study, it is reasonable to conclude that for materials of the type studied (hard, non-porous), decontamination should be similar regardless of surface type.

3.5. Impact of additives on decontamination

The addition of various additives to the fentanyl at a 19:1 ratio on laminate material resulted in noticeably higher fentanyl recoveries following decontamination with two of the better-performing decontamination solutions (pH 5 bleach with surfactant and Dahlgren Decon^™) as shown in Fig. 5. The impact of mannitol and lactose was relatively small while the presence of ascorbic acid resulted in significantly higher recovery of unreacted fentanyl (p < 0.05). Ascorbic acid (along with other cutting agents (Methods for Impurity Prof, 2005)) is well-known to be an effective reducing agent for both bleach and peracetic acid solutions, whereas sugars like mannitol and lactose are not. Here, the ascorbic acid creates a noticeable demand for the active ingredients resulting in less degradation of fentanyl. This demand might be overcome through reapplication of the same oxidant, which merits further investigation. Accordingly, in an actual fentanyl remediation case, any information on the chemical nature of additives will greatly facilitate a successful remediation.

Fig. 5.

Fentanyl Recoveries Following Application of Dahlgren Decon and pH 5 Bleach with surfactant in Presence of Additives.

3.6. Physical removal of residual fentanyl following oxidative degradation

After a 1-h contact time, most of the decontaminants were still present as a wet residue on the materials. In a response scenario, this would require removal of the residue. Here, the use of a dry surface wipe was assessed to remove this residue. The subsequent two IPA wetted wipes that followed the first dry wipe measured fentanyl remaining on the surface. Fig. 6 summarizes the mean fentanyl amounts recovered from the three extracted wipes from the laminate and vinyl upholstery coupons. Following the pH 5 bleach application and a 1-h contact time, a total of 227 μg fentanyl was recovered (8.8% of applied amount) with the majority on the first wipe (123 μg, 4.8%). Here, solids were observed on the laminate surface after the 1-h contact time, which may be attributed to measurement of unreacted fentanyl. Following decontamination of the medium-size coupons using Dahlgren Decon^™, the total residual fentanyl amounts were statistically significantly lower for laminate (1.2 μg or 0.13% of applied amount) and vinyl upholstery (1.0 μg or 0.04% of applied amount). The amounts recovered from the coupons were equal to 0.07% and 0.03% of the positive control recoveries for laminate and vinyl upholstery, respectively. Total fentanyl recoveries were 0.03 μg or lower after a 4-h contact time with amounts less than the LLOQ for the second and third wipes. The additional contact time likely led to a more complete dissolution and subsequent degradation of presumably agglomerated fentanyl. Results such as in Fig. 6 are highly informative but point to the need for better understanding of dissolution of spilled fentanyl preparations relative to decontamination approaches.

Fig. 6.

Relative Fentanyl Recoveries Following Application of pH 5 Bleach and Dahlgren Decon (DD) to Medium-size Coupons.

3.7. Decontamination by-product assessment

The medium size coupon wipe extracts were also investigated for the presence of potential decontamination by-products. A fentanyl degradation byproduct, norfentanyl, was detected in the wipe extracts of the test coupons (70–140 μg, semi-quantitatively) following pH 5 bleach decontamination. Norfentanyl, the main metabolite of fentanyl, has been identified as a byproduct following N-dealkylation at the piperi-dine ring during solution phase oxidation studies (Qi et al., 2011). Norfentanyl was also observed at a lower concentration in the positive control wipes (13 μg) which suggests it may also be a minor impurity in the fentanyl-HCl product (<0.02% by mass), perhaps an artifact of the synthetic route. Norfentanyl was not found in the surface wipes following Dahlgren Decon^™ decontamination, possibly indicating oxidation mechanisms in which norfentanyl formation is not occurring or, if formed, result in further oxidative degradation of norfentanyl. Another possible, unreported fentanyl oxidation byproduct was detected following pH 5 bleach decontamination at m/z 350.2, without further identification using available mass spectrometry libraries or structural elucidation via the available data. No evidence was found for the formation of chlorine containing byproducts following decontamination with the pH 5 bleach solution.

It should be noted that all byproduct analysis, in this study or elsewhere, is limited by the ability to prepare samples of analysis and by the analytical technique’s ability to detect them. Additional research on the formation of potential toxic degradation byproducts is warranted, especially when toxicity of byproducts with opiate receptor activity occurs at ultra-low concentrations.

3.8. Comparison of decontamination approaches

The use of the three successive surface wipes enables measurement of the residual fentanyl amounts on medium-size coupons. Their use can also be interpreted as representing the physical removal of residual fentanyl from a surface. This concept also holds true for fentanyl recoveries from the three wipes that were used to sample the positive control (no decontaminant applied). Fig. 7 illustrates for laminate the measured log₁₀ reduction in fentanyl mass for a physical removal approach only (defined here via repeated surface wiping) versus the decontamination spray and decontamination spray plus physical removal via wiping. The starting point is the approximately 1 mg/100 cm² contamination level for the medium-size material coupons. In Fig. 7, physical removal of fentanyl will reduce the fentanyl mass on the surface, however, fentanyl degradation via spray application of the identified oxidants is noticeably (and statistically) more efficient.

Fig. 7.

Measured log reduction in fentanyl mass from laminate.

Currently there are no published health-based levels that would establish whether a surface is clean. Therefore, a measure of effective remediation could be defined as when post-decontamination sampling results are reported below the limit of detection for the chosen sampling and analytical method. For instance, the final decontamination level following the remediation of a fentanyl contaminated home (Doerflein, 2019) was set to 0.1 μg/100 cm² (equal to the laboratory detection limit of the testing laboratory). This level is identified in Fig. 7 (dashed line), and it is 4 log₁₀ lower than the initial contamination level in this study. Chemical degradation via oxidation using an efficacious decontaminant enhances the ability to reach the intended cleanup level. Nevertheless, this approach can be limited in the likely case that other co-contaminants have oxidant demand. As there may be insufficient time to characterize the co-contaminants prior to performing decontamination, investigation of approaches to ensuring application of sufficient amounts of oxidants are warranted.

4. Conclusions

The research presented here focused on the decontamination of nonporous indoor materials. Oxidative chemistries as present in chlorine containing bleach solutions, and adjusted to lower pH, as well as in the DF200 and Dahlgren Decon^™ product are effective (>95%) in degrading fentanyl after a 1-h dwell time. Porous materials would be more difficult to clean and would therefore be preferably removed from a building prior to the remediation start (Doerflein, 2019).

Additional research is warranted to provide recommendations for responders who may come in contact with fentanyl that would lead to the need to clean their gear or PPE. Extrapolation of the results presented here should be considered with caution as the needs of gear or PPE decontamination deviate from the design of the present study in that the contact time with the decontaminant would be significantly shorter.

The use of HEPA dry-vacuuming, a suggested first step (entanyl and Fac, 2018; Fentanyl Remediation: Gui, 2020; Doerflein, 2019; https://globalnews.ca/new, 2020) in the remediation of a fentanyl-contaminated site, was beyond the focus of the present study. The efficiency of dry-vacuuming is likely dependent on the properties (particle size, agglomeration) of the fentanyl powder combined with its adhesion forces to different surfaces, as well as shear force created by the vacuum suction and the implementation of the technique by the operator. These factors would affect the statistical interpretation of results of such a study, perhaps even to a greater extent than statistical uncertainties that affected the present, highly-controlled study. A combination of HEPA vacuuming and the application of the efficacious decontamination techniques investigated here are tools in a toolbox of decontamination approaches to apply to a fentanyl contaminated site.