Decontamination of soil contaminated at the surface with Bacillus anthracis spores using dry thermal treatment

Abstract

In the event of a large, aerosol release of Bacillus anthracis spores in a major metropolitan area, soils and other outdoor materials may become contaminated with the biological agent. A study was conducted to assess the in-situ remediation of soil using a dry thermal treatment approach to inactivate a B. anthracis spore surrogate inoculated into soil samples. The study was conducted in two phases, using loam, clay and sand-based soils, as well as biological indicators and spore-inoculated stainless-steel coupons. Initial experiments were performed in an environmental test chamber with temperatures controlled between 80 and 110 °C, with and without added humidity, and with contact times ranging from 4 h to 7 weeks. Tests were then scaled up to assess the thermal inactivation of spores in small soil columns, in which a heating plate set to 141 °C was applied to the soil surface. These column tests were conducted to assess time requirements to inactivate spores as a function of soil depth and soil type. Results from the initial phase of testing showed that increasing the temperature and relative humidity reduced the time requirements to achieve samples in which no surrogate spores were detected. For the test at 80 °C with no added humidity, 49 days were required to achieve soil samples with no spores detected in clay and loam. At 110 °C, 24 h were required to achieve samples in which no spores were detected. In the column tests, no spores were detected at the 2.5 cm depth at four days and at the 5.1 cm depth at 21 days, for two of the three soils. The experiments described in the study demonstrate the feasibility of using dry thermal techniques to decontaminate soils that have been surficially contaminated with B. anthracis spores.

1. Introduction

Bacillus anthracis is a naturally-occurring bacterium in soils, causes anthrax disease in wildlife, livestock, and humans (Mullins et al., 2013); and may persist (in the spore form) for several years in soil environments (Hugh-Jones and Blackburn, 2009). Soils and other outdoor materials or environments will also become contaminated with B. anthracis spores in the event of an intentional outdoor release of the biological agent. Such an intentional aerosol release of a large quantity of B. anthracis spores in a major metropolitan area is one of 15 National Planning Scenarios (U.S. Department of Homeland Security, 2010). Regardless of whether the B. anthracis contamination of soil occurs from infected wildlife or livestock, or from an intentional or accidental release into the environment, the National Biodefense Strategy (U.S. Department of Defense et al., 2018) calls for the biodefense enterprise to prepare for such an event via the development and verification of decontamination strategies and techniques.

The efficacy of a decontaminant in inactivating B. anthracis spores is highly dependent on the material with which the spores are associated, and soil remains one of the most difficult materials to decontaminate. This is due to its relatively high organic content, other variable chemical constituents and physical properties such as density, particle sizes, and porosity (Wood and Adrion, 2019). Literature related to the investigation of treatments for soil materials contaminated with B. anthracis spores is sparse. Semi-quantitative field trials using liquid solutions of 5% formaldehyde in seawater were mostly effective in reducing the number of B. anthracis spores in soil on an island used for bioweapons tests during World War II (Manchee et al., 1994), although repeated applications of the formaldehyde solution were required in some areas. More recent controlled laboratory studies have shown that chemical treatments of soils using acidified chlorine bleach (Wood et al., 2011), peracetic acid (U.S. Environmental Protection Agency, 2010), or aqueous chlorine dioxide (U.S. Environmental Protection Agency, 2012) were ineffective in inactivating B. anthracis spores. Alternatively, chemistries such as activated sodium persulfate, gaseous chlorine dioxide, methyl bromide (U.S. Environmental Protection Agency, 2017), and metam sodium (U.S. Environmental Protection Agency, 2013) were generally effective in inactivating B. anthracis spores in soil, but the level of inactivation depended on test conditions and depth of spores in soil. Simulated sunlight produced in the laboratory was found to inactivate B. anthracis spores on hard nonporous materials such as glass but caused no reduction of B. anthracis spore numbers in a soil matrix (Wood et al., 2015). Studies investigating the addition of chemicals to soil to cause B. anthracis spores to germinate (e.g., L-alanine) and thus become less resistant to inactivation showed that this approach was effective in reducing spore numbers when followed by natural attenuation (Bishop, 2014) or the addition of peracetic acid (Celebi et al., 2016).

In addition to chemical treatments for soils, ex-situ treatment of B. anthracis-contaminated soil (e.g., excavation/removal of soil followed by off-site treatment such as incineration) is another remediation option, although this approach may aerosolize and disperse the B. anthracis spores during the excavation and transport processes and may cause further cross contamination. Thus, in-situ treatment of soil contaminated with B. anthracis soil, such as the use of thermal techniques, is another remediation option that would potentially avoid the aerosolization of spores. However, while the use of thermal treatment for chemical contamination (U.S. Environmental Protection Agency, 2014), disinfection (i.e., inactivation of vegetative bacteria and viruses) (Peruzzi et al., 2011) or disinfestation (Rainbolt et al., 2013) of soils is documented in the literature, the literature related to the use of thermal treatments to inactivate bacterial spores such as those of B. anthracis (i. e., to essentially sterilize the soil) is non-existent. Bacterial spores are one of the most resistant class of microorganisms to inactivate. If a treatment is effective in inactivating bacterial spores, it should also be effective in inactivating less-resistant microorganisms, such as vegetative bacteria and viruses (McDonnell, 2017; U.S. Department of Health and Human Services, 2009), as long as the microorganisms are in the same matrix such as soil. Further, whereas the sterilization of materials other than soil (e.g., foods, medical instruments) via thermal treatment is widely known, documented, and used (Joslyn, 2001), few data are available on the inactivation of bacterial spores at the relatively lower temperatures that may be more suitable for surficial soil decontamination (Brannen and Garst, 1972; Peeler et al., 1977a).

In an outdoor release of B. anthracis spores, it is expected that the spores would initially remain near the soil surface, although migration of bacterial spores downward through the soil (due to rainfall or other factors) column may occur and remains a research gap needing further investigation. The use of a thermal blanket placed on top of the soil, an in-situ approach for remediation of surficial soil (Iben et al., 1996) was successfully demonstrated at field-scale for polychlorinated biphenyl contamination, and we envision a similar approach could be used for B. anthracis contamination. Hence a similar approach was evaluated in the present study, albeit at a smaller scale.

In the present study, we conducted tests with soil samples contaminated with a B. anthracis spore surrogate. The focus of the study was on the use of dry heat, which is characterized as having a relative humidity (RH) less than 100% (Joslyn, 2001). The tests were initially performed in an environmental chamber with temperatures between 80 and 110 °C, with and without added humidity. (Tests have shown that adding humidity to hot air reduced times required to inactivate spores (Buhr et al., 2015; Buhr et al., 2016)). This temperature range was determined to be low enough that it would be feasible (less heat input required) to implement in the field, but effective in inactivating bacterial spores within a reasonable time. Once the time requirements to inactivate spores in various soils as a function of temperature were determined, tests were scaled up to assess the thermal inactivation of spores in small soil columns, in which a heating plate was applied to the soil surface (similar to the Iben study). These column tests were conducted to assess time requirements to inactivate spores as a function of soil depth and soil type. In summary, the research presented here attempted to prove the concept of using dry heat in-situ to remediate soil contaminated with B. anthracis spores, by determining time and temperature requirements to inactivate bacterial spores as a function of soil type and depth.

2. Materials and methods

2.1. Overview

The in-situ thermal remediation of soil contaminated with Bacillus spores was evaluated in two phases. In Phase 1, we examined the efficacy of heat treatment in inactivating Bacillus spores inoculated on the surface of soil samples placed in Petri dishes. The soil samples were exposed at different test chamber air temperatures, contact times, and RH levels. Three soil types were used in the study: a loam, sand, and clay soil. In addition to the spore-inoculated soil samples, the Phase 1 tests also included commercial biological indicators (BIs) as well as spore-inoculated stainless-steel (SS) coupons.

In Phase 2, we examined the efficacy of inactivating Bacillus spores placed at different depths in a soil column with the same three soil types, using a heating plate or “blanket” applied to the surface of the soil column for 21 consecutive days. This time period was selected based in part on the results from the Phase 1 tests and was also determined to be a reasonable period to allow for the remediation to be successful.

2.2. Microorganism

Bacillus atrophaeus var. globigii (Bg), a surrogate for the spore-forming bacterial agent B. anthracis, was used as the test organism for this study. Bg (also known as Bacillus atrophaeus) was inoculated onto our soil samples and was also the organism in the commercial BIs we used. Like B. anthracis, Bg is a soil dwelling, Gram-positive, spore forming, aerobic microorganism, but unlike B. anthracis, it is non-pathogenic. Bg forms an orange pigment when grown on nutrient agar, a desirable characteristic when there is a need to distinguish viable Bg spores in complex environmental samples. Bg has a long history of use in the biodefense community as a simulant for B. anthracis-associated biowarfare and bioterrorism events (Gibbons et al., 2011) and is the recommended indicator organism for dry heat sterilization processes (Murphey, 2007). Bg has also been shown to possess similar or higher resistance to inactivation compared to B. anthracis, for many decontamination techniques (Richter et al., 2018; Wood and Adrion, 2019). The source of the Bg varied according to the method of spore inoculation, discussed in detail below.

2.3. Soil materials

Three soil materials were used in this study and are defined by the USDA textural class using the hydrometer method (U.S. Department of Agriculture, 2014). The soils, from Eastern North Dakota, were collected and analyzed by Agvise Laboratories (Northwood, ND; USA). The three soils were as follows: a clay loam (Agvise sample M-CL-PF; referred to as “clay”) that contained a large portion of high-density clay that tended to become compact and hold large amounts of water; a loamy sand (Agvise sample GT-S-PF; referred to as “sand”) susceptible to drying out due to poor water and mineral retention; and a loam (Agvise sample HCB-SI-PF). The physical and chemical characteristics of the soils as received are listed in Table S1 of the Supplementary Material (SM) and are based on tests (U.S. Department of Agriculture, 2014) completed in compliance with Good Laboratory Practice (40 CFR Part 160). The soil materials were used in the decontamination experiments as is, i.e., they were not pre-sterilized prior to testing.

2.4. Phase 1 chamber tests

2.4.1. Inoculation of samples

Each Phase 1 soil sample was inoculated with a nominal value of 107 Bg spores, using a modified round Aerosol Deposition Apparatus (ADA) (Calfee et al., 2013; Lee et al., 2011) fitted with an actuator and a metered dose inhaler (MDI), as illustrated in Figure S1 of SM. The MDI containing spore suspension was prepared using dry Bg spores received from Dugway Proving Ground (Tooele County, UT; USA), resuspended in 100% ethanol, then combined into each MDI canister with 1,1,1,2-tetrafluoroethane (HFA-134a), a hydrofluorocarbon propellant with less ozone depletion capacity than traditionally used chlorofluorocarbons. The MDI was situated inside the actuator so that each time the actuator was depressed a repeatable number of spores was deposited on the soil sample.

The soil samples in the Phase 1 soil thermal treatment testing were placed in 100-mm diameter Pyrex® Reusable Petri Dishes with clear lids (Corning Inc., Mfr. No. 3160–101, Corning, NY; USA) that withstand repeated sterilization (wet or dry). To inoculate the soil with spores, the round ADA was placed flush with the bottom of a sterile Petri dish (Figure S2-A, soil). Five grams of unsterilized soil were poured through the 3.8 cm diameter opening of the ADA prior to inoculation. An MDI with an actuator was then attached to the top of the ADA, the slide was opened, and the MDI activated. Following inoculation, the slide was closed, and the MDI actuator was removed. After the MDI was removed, the inoculum was allowed to settle/impact on the soil surface for a total of 1 min, after which the ADA was removed, and the soil/spores were allowed to spread into the bottom of the Petri dish as shown in Figure S2-B. The Petri dish was then covered before being used a day later.

Stainless steel coupons (7.6 cm by 7.6 cm), made from 16-gauge type 304 mill-finished steel sheet (McMaster-Carr, Elmhurst, IL; USA), were also used as test materials, as a quality control (QC) check to validate the microbiological measurements. The inoculation of the SS coupons was performed in the same manner as the soil inoculations with the ADA placed flush with the surface of the coupon. The Phase 1 tests also used commercially available BIs, with a nominal population of 10⁶ Bg spores on a 0.9 cm diameter stainless steel disc packaged in Tyvek® (#9372, Mesa Labs, Lakewood, CO; USA).

2.4.2. Test chamber

The Phase 1 tests were performed using a temperature/humidity-controlled environmental test chamber (ETC; model 1016H, TestEquity LLC, Moorpark, CA; USA). The listed working volume for this chamber is 442 L, with internal dimensions of 76 cm height × 76 cm width × 76 cm depth and includes a fan to improve heat distribution. This chamber was modified to include a series of trays to accommodate the large numbers of soil samples that were treated. One tray of samples (every type of sample) was removed from the chamber at each time-point. Tests were conducted in this chamber with the temperature programmed to remain constant for the test duration (standard deviations for the chamber temperatures were <0.5 °C), using temperatures of 80, 90, and 110 °C with no added humidity. With no added water vapor, the RH level was <1%. An additional chamber test was conducted at 80 °C, but with the RH controlled to 80% (standard deviation 1.6%).

For the Phase 1 chamber tests, four timepoints were used to assess inactivation kinetics, and were selected according to the temperature of the chamber. At each timepoint, a tray of the soil samples, BIs, and SS coupons was removed from the chamber, and the samples were assayed for viable spores quantified as Colony Forming Units (CFU, discussed below). For each sample type and timepoint, five replicates were used. One positive control (not heated; kept at laboratory ambient conditions at a temperature of approximately 21 °C) for each sample type was used for each time point and was assayed concurrent with the test samples.

2.5. Phase 2 column tests

2.5.1. Soil columns

The soil columns were assembled in a 35.6 cm width x 35.6 cm length x 15.3 cm depth configuration. The column walls and bottom were constructed with 0.64 cm thick Lexan (Part # 1NLK9 Grainger, Miami, FL; USA). Only the top surface of the column had heat applied. The soil columns were covered on the sides with 2.5 cm foil-backed fiberglass insulation (R Value = 4.3, McMaster-Carr, Cleveland, OH; USA) to limit heat loss.

Four custom-made soil pouches inoculated with Bg (discussed next), and the tip of one thermocouple (Super OMEGACLAD™ XL, Omega Engineering, Inc., Norwalk, CT; USA), were placed in the center of the column at 2.5 cm interval depths starting from the top surface of the soil to a depth of 10.2 cm. Five small sampling ports (corresponding to each layer location) were cut on each side of the column to enable the removal of each soil pouch at each sampling timepoint. A thread made from 100% mercerized cotton (Hobby Lobby, SKU, 191189) and tied to each pouch was pulled through the sampling port to extract each soil pouch from the column. A data acquisition system (personal DA/50 Series with USB data acquisition modules, Measurement Computing, Norton, MA; USA) was used to record the temperature of the soil at each depth throughout the experimental duration.

2.5.2. Custom-made soil pouches with Bg

To facilitate recovery of the Bg spores at various depths within the soil column during testing, custom-made soil pouches were used. Each pouch was loaded with 1 g of soil and then the soil was inoculated with ~10⁷ spores. This inoculation consisted of one 25 μL droplet of a Bg bacterial spore suspension (4.5 × 10⁸ CFU per mL, ATCC # 9372 in 29% ethanol, Yakibou Labs, Holly Springs, NC; USA), placed in the center of the soil mass. (No mixing was performed, but the liquid inoculum was visibly distributed throughout most of the soil.) The soil pouches were made from 10 μm pore-size filter material (F-58 Filter bags, Ankom Technology, Macedon, NY; USA) and heat-sealed after inoculation. Refer to Figure S3 in the SM.

These inoculated soil pouches were placed in the center of the soil column, at 2.5 cm depths, starting on the top surface of the soil and continuing to a depth of 10.2 cm to determine the decontamination efficacy of the thermal process downward into the soil. The pouches were then removed at the designated times for each depth. Pouches placed nearer to the surface of the soil column and hence nearer to the heat source were typically removed sooner than pouches at lower depths. Four positive control soil pouches for each soil type were not exposed to heat treatment and were quantitated at the start of each experiment (time zero), and an additional positive control for each soil type was quantified at the same times as the test pouches.

2.5.3. Heat source

The surface of the soil column was heated using a heavy-duty silicon rubber explosion-resistant 30.5 cm × 30.5 cm heating plate (Part No. SEPHB-1212–360-120-T3, Omega™, Norwalk, CT; USA) with an adhesive backing used to affix it to a stainless-steel plate (16-gauge type 304 mill-finished steel sheet; McMaster-Carr, Elmhurst, IL; USA). The stainless-steel plate was in contact with the soil surface. A controller was used to regulate the temperature of the heating plate and set at 141 °C. The whole apparatus was covered with 2.5 cm thick foil-backed fiberglass insulation. A photograph of the column and heat plate is shown in Figure S4.

2.6. Soil moisture and pH measurement

The moisture of each type of soil was determined according to a standard method (ASTM International, 2010). In short, 150 g of soil was spread in shallow trays, and the mass of the containers was weighed with an analytical balance (Model # ML54/03, Mettler-Toledo, LLC, Columbus, OH; USA), and recorded to four decimal places. The containers were then heated to 110 ± 5 °C using a benchtop laboratory oven (VWR 1300 U, Sheldon Manufacturing Inc., Cornelius, OR; USA). The mass of the tray with the soil was checked daily removing it from the oven and covering it with aluminum foil to inhibit moisture intake from the environment. The soil was then cooled in a desiccator for at least 1 h, and then weighed. When soil mass was constant for 48 consecutive hours, the last check was performed three times, in 4-h long intervals (acceptance criteria ±0.5% mass change between triplicate measurements).

Measurement of soil pH was performed according to a standard method (ASTM International, 2013). This measurement was conducted in soil materials suspended in water and 0.01 M calcium chloride. The measurement of the pH of the soil materials measurement was performed using a calibrated pH meter (Oakton® Acorn™ pH 5, OAKTON Instruments, USA).

2.7. Test matrix

The test matrix for the study is summarized in Table 1.

Table 1.

Study test matrix.

	Test Materials	Heat Source	Target Decontamination Conditions	Timepoints (d)
Phase 1	5 g samples of clay, loam, or sand soil; commercial BIsa; SS coupons	Environmental Test Chamber	80 °C, no added humidity	7, 14, 21, 49
			80 °C, 80% RH	8, 15
			90 °C, no added humidity	2, 5, 7, 9
			90 °C, no added humidity	9, 15
			110 °C, no added humidity	0.17, 0.33, 0.5, 1
Phase 2	Columns comprised of either clay, loam, or sand soil; Bg inoculated soil pouches placed at surface and at 2.5 cm depths	Soil column with heat plate on top	Heating plate on column surface set at 141 °C	Depth (cm)	Timepoints (d)
				0	0.17, 0.25, 0.5, 1
				2.5	0.5, 1, 4, 7
				5.1	1, 4, 7, 21
				7.6	1, 4, 7, 21
				10.2	1, 4, 7, 21

^aThe BIs were inadvertently excluded from the first 90 °C test.

2.8. Recovery and analysis of samples

Each soil sample (both test samples and positive controls) was aseptically removed from the Pyrex® Petri dish or soil pouch using a sterile laboratory scoop and placed in a sterile 50-mL conical tube (Part No. 352098, Corning Life Sciences, Corning, NY; USA) containing 10 mL of sterile phosphate buffered saline with 0.05% Tween (PBST). Each tube was sonicated for 10 min and then vortexed continuously for 2 min. To filter out larger particles, the soil slurry was poured through a Falcon® 70 μm cell strainer (Part No. 352350, Corning Life Sciences, Corning, NY; USA) fitted onto a second 50-mL conical tube. Another 10 mL of sterile PBST was added to the original conical tube, vortexed to mix the remaining soil, and poured into the cell strainer, to collect a total extraction volume of ~20 mL per sample for analysis. This extraction volume was then heat treated at 80 °C for 20 min to reduce background organisms in the soil (Turnbull et al., 2007) (e.g., bacterial vegetative cells and fungal spores) that could potentially interfere with colony counting of the target organism.

The commercial BIs and sample wipes (used for sampling the SS coupon surface) were extracted the same way as the soil samples, but without the strainer filtering and heat shock steps, and with PBST volumes of 10 mL. The wipes (Sterile Kendall Curity™ Versalon all-purpose, rayon-polyester blend absorbent gauze sponges, 2'' x 2” sterile packed, #8042, four-ply; Covidien PLC, Dublin, Ireland, P/N 8042 wetted with sterile phosphate buffered saline with 0.005% TWEEN®-20) used for sampling the SS coupons were extracted by vortexing for 2-min in 10-s bursts with 20 mL of PBST.

The extraction liquid from the soil samples was spread plated with various aliquots and dilutions to obtain three plates within the countable range of 30–300 CFU/plate. The commercial BI and wipe samples were plated in triplicate using a spiral plater (Autoplate 5000, Advanced Instruments Inc., Norwood, MA; USA). The automated spiral plater deposits a sample extract in exponentially decreasing amounts across a rotating agar plate in concentric lines to achieve three, 10-fold serial dilutions on each plate. Plates were incubated at 35 ± 2 °C for 18–20 h. During incubation, the colonies develop along the lines where the sample is deposited. Colonies on each tryptic soy agar (TSA) plate were enumerated either manually (soil spread plates) or with a QCount® colony counter (Advanced Instruments Inc., Norwood, MA; USA). The QCount® colony counter was used to automatically calculate the CFU/ mL in a sample based on the dilution plated and the number of colonies detected on the plate. The QCount® data were recorded in a Microsoft Excel spreadsheet.

Soil samples with plate results below the 30-CFU threshold were replated with a more concentrated sample aliquot or spread-plated in triplicate with a maximum volume of 0.25 mL. No soil samples were filter-plated due to the high debris loads. If the BI or SS coupon samples fell below the 30-CFU threshold, they were filter-plated with the remaining extraction volume using a 100-mL capacity Micro-Funnel™ unit with 0.45 μm GN-6 Metricel membranes, and a vacuum manifold (Pall Corporation, Port Washington, NY; USA). The filters were placed onto TSA plates and incubated at 35 ± 2 °C for 20–24 h before manual enumeration. Levels of detection for the samples were determined by using a value of 1 CFU divided by the largest volume plated and multiplied by the total sample volume (Mikelonis et al., 2020) (i.e., the volume of PBST used to extract spores from the soil material, BI disc, or wipe). The plate and sample volumes analyzed for soil samples were approximately 0.25 and 20 mL, respectively, resulting in soil sample detection levels typically between 70 and 90 CFU. For the BIs and SS coupons, the filter plate and sample volumes analyzed were approximately 8 and 10 mL, respectively, resulting in sample detection levels typically less than 2 CFU.

2.9. Statistical analysis

For the Phase 1 chamber test results, the recovery of Bg spores (in terms of log CFU) from each soil or sample type was plotted as a function of time, for each chamber temperature. These thermal resistance plots enabled the quantification of results in terms of a D-value, which is the time required to obtain a 1 log₁₀ reduction (e.g., 90% reduction) for a given test condition. The D-values were determined from the linear fit of the CFU recovery over time with the CFU recovery results plotted on a log-scale. OriginPro 2019 software (OriginLab Corp., Northampton, MA) was used to determine the linear fit of the data points, the slope, and the R-square values, with the D-value calculated as the negative reciprocal of the slope. When replicate samples or analyses were used, data are reported as the mean ± standard deviation.

3. Results

The results from the Phase 1 (determining inactivation kinetics as a function of soil type and temperature, using a temperature-controlled environmental chamber) and Phase 2 (determining inactivation efficacy as a function of soil type and depth, using a heat plate applied to the soil column surface) tests are presented in this section.

3.1. Environmental chamber tests

3.1.1. Decontamination results

The results for the Phase 1 environmental chamber tests are shown in Fig. 1 and in the SM Figure S5. These results are reported as the log CFU of Bg recovered over time, for the three soils as well as the SS coupons and the BIs. Two experiments were conducted at 80 °C (the first with no added humidity, and the second one with RH controlled to 80%), two experiments at 90 °C, and one at 110 °C. (The second 90 °C test was conducted with a longer contact time.) Detection limits for the BIs and SS controls were approximately 0.1–0.4 log CFU, while for the soils, the detection limits were in the range of approximately 1.8–1.9 CFU per sample.

Fig. 1.

Recovery of Bg log colony forming units (mean ± SD) from soils and other samples. A = test chamber 80 °C, low RH; B = 80 °C, high RH; C = 90 °C; D = 110 °C. Black square = SS, red circle = clay, upward pointing blue triangle = loam, downward pointing green triangle = sand, purple diamond = BI. Dashed horizontal line approximate detection limit for SS and BI; horizontal line detection limit for soil samples.

For the test at 80 °C with no added humidity (Fig. 1A), the soil samples with no spores detected (clay and loam) did not occur until the 49-d timepoint. At this timepoint, an average of >3 log CFU Bg were recovered from the loam samples. At the 28-d timepoint, the BIs were completely inactivated, but Bg spores were detected in all other sample types. In contrast, at 80 °C and with the RH elevated to 80% (Fig. 1B), no spores were detected in any sample at the 15-day mark, which was the longest timepoint evaluated for this test condition. In addition, at the 8-day timepoint, no spores were detected in the sand, BIs, and SS coupons.

In the first 90 °C test (Fig. 1C), all samples had detectable spores at the longest timepoint tested of 9 days. For this reason, another 90 °C test was conducted to 15 days (Figure S5), which resulted in having no spores detected at the 15-day mark from any samples. In addition, at the 9-day timepoint for the second 90 °C test, the BIs, SS coupons, and sand samples had no spores detected from any samples. In the 110 °C environmental chamber test (Fig. 1D), at 24 h, the longest time tested, no spores were detected in any of the samples except for the sand. At this timepoint and temperature, an average of 3.3 log CFU of Bg were recovered from the sand.

3.1.2. Positive control results

The spore recovery results for the positive controls used during the 80 °C test with no added humidity are shown in Table 2. The positive controls were not subject to any heat treatment and recovered concurrently with the test samples at the same timepoints. We present here the results for the positive controls for the 80 °C, low RH experiment, to illustrate the stability of the spores at ambient temperature for the longest test duration used in the study (49 days). As can be seen in Table 2, the controls were relatively stable for the clay and loam soils, with minimal loss in recovery over the 49-day period. The spore recovery from the SS coupons dropped nearly 1 log over the same time period. For the commercial BIs, the recovery of spores at time zero and day 28 averaged 5.39 log CFU and 5.20 log CFU, respectively; the recoveries for the control BIs at the other time points (8, 15, 49 days) were all > 6.30 log CFU. The results for the sand soil show a recovery of 8.56 log CFU at day 49, which was greater than a 1-log increase in recovery when compared to time zero (7.51 log CFU).

Table 2.

Positive control Bg spore recoveries for the 80 °C low RH test.

	Day 0a	Day 8	Day 15	Day 28	Day 49	Avg ±SD for Day 8 through Day 49
SS coupons	7.28 ± 0.31	7.32	6.89	7.36	6.43	7.00 ± 0.44
Clay soil	7.50 ± 0.26	7.38	7.34	7.63	7.59	7.49 ± 0.15
Loam soil	7.55 ± 0.20	7.33	7.23	7.61	7.52	7.42 ± 0.17
Sand soil	7.51 ± 0.13	7.39	7.49	7.92	8.56	7.84 ± 0.53
BIs	5.39 ± 0.09	6.65	6.99	5.20	6.33	6.29 ± 0.78

^an = 3 for Day 0; n = 1 for other time points; results reported as log colony forming units (CFU).

3.1.3. D-values

The D-values for the five Phase 1 environmental chamber tests are shown in Table 3, and provide the mean time required to reduce the spore recovery from each sample type by 90% or 1 log. The D-values are reported in units of days for all tests except the 110 °C test, which are reported in hours. As expected, the higher rate of inactivation and hence lower D-values were associated with the higher temperatures. The D-values were the lowest for the 110 °C test condition and were within a range of approximately 4–6 h for all five materials. The highest D-values occurred at the lowest temperature of 80 °C, with no humidity added; these D-values ranged from approximately 5–12 days. With added humidity at the 80 °C temperature, the D-values for the five sample types were reduced by approximately four-to seven-fold. The D-values for the first 90 °C test ranged from 2.0 to 2.8 days, whereas in the second 90 °C test, the D-values ranged from 1.0 to 1.8 days. The D-values were generally lower for the SS coupons and BIs, when compared to the soil materials. There were no apparent trends for the soil D-values, which on average fell within 27% of each other for each test.

Table 3.

D-values for Bg in soil by test condition and soil type.

	80 °C low RH		80 °C 80% RH		90 °C Test 1		90 °C Test 2		110 °C
	D-value (days)	R2	D-value (days)	R2	D-value (days)	R2	D-value (days)	R2	D-value (hours)	R2
SS	7.9 ± 0.5	.97	1.1 ± 0	1	2.0 ± 0.2	.98	1.3 ± 0	1	4.0 ± 0.4	0.95
Clay	9.0 ± 0.4	.99	2.7 ± 0.2	0.99	2.3 ± 0.3	.87	1.7 ± 0	1	4.4 ± 0.3	0.98
Loam	12.5 ± 1.4	.95	2.7 ± 0.1	0.99	2.6 ± 0.3	.93	1.8 ± 0	1	4.2 ± 0.2	0.99
Sand	8.9 ± 0.4	.99	1.4 ± 0	1	2.8 ± 0.4	.91	1.7 ± 0	1	6.4 ± 0.8	0.93
BI	5.4 ± 1.2	.78	1.3 ± 0	1	NT	NT	1.0 ± 0.1	.94	3.9 ± 0.3	0.97

D values reported as ± SD. NT = not tested.

The R square values, which provide an indication of the goodness of the linear fit of the log CFU data used to compute each D-value, were all above 0.9 except for two instances. The R square values equaling 1.0 signify that only two data points were used.

3.2. Soil column results

The efficacy results for the clay, loam, and sandy soil column tests are shown in Fig. 2, respectively. The results are presented in terms of the log CFU of Bg spores recovered over time as a function of depth in each soil column. A heat plate was placed on the surface of the soil column and set to 141 °C, for a contact time of 21 days. Four timepoints were assessed for each depth, with the shallower spore pouches recovered sooner. Also shown in the figures are the results for the positive controls, which were not exposed to heat treatment, and were analyzed concurrently with the test samples. The positive control recovery results are also summarized in Table 4 and indicate minimal loss in spore recovery over the 21-d time interval; the mean recovery of Bg spores from the positive control soil pouches ranged from 6.2 to 6.9 log CFU.

Fig. 2.

Recovery of Bg spores in the soil columns as a function of time and depth; log colony forming units. A = clay; B = loam; C = sand. Black square = positive control, red circle = surface, upward pointing blue triangle = 2.5 cm depth, downward pointing green triangle = 5.1 cm depth, purple diamond = 7.6 cm depth, left pointing gold triangle = 10.2 cm depth. Horizontal line shows detection limit for soil samples.

Table 4.

Recovery of Bg spores from the soil column positive controls.

	4 h	6 h	12 h	24 h	4 d	7 d	21 d	Avg ±SD
Clay	6.78	6.81	6.79	6.46	6.87	6.35	6.35	6.68 ± 0.23
Loam	6.82	6.88	6.79	6.88	6.23	6.89	6.35	6.75 ± 0.28
Sand	6.40	6.41	6.47	6.53	6.39	6.74	6.57	6.49 ± 0.13

Recovery of Bg spores in log colony forming units.

Samples that were located on the surface were completely inactivated at the 12-h timepoint for all three soil types. At the 2.5-cm depth, minimal inactivation occurred at the 1-day timepoint for the sandy soil column, the longest timepoint assessed for that soil at that depth. For this reason, samples were recovered out to 1 week at the 2.5 cm depth for the other two soils. For the clay and loam soil columns, there was minimal inactivation of spores at the 2.5 cm depth at 1 day, but at 4 days, the samples were completely inactivated (no spores were detected). At the 5.1 cm depth, no spores were recovered at the 21-day timepoint for the loam and sand, but 4.4 log CFU Bg were recovered from the clay column. Minimal loss in recovery (compared to positive controls) of Bg spores was observed for the samples at the 7.6- and 10-cm depths for all timepoints, for all three soils.

The temperature data for the soil column tests are summarized in Table 5; the temperature profiles and soil moisture results for the soil columns for each experiment are provided in the SM as Figures S7 and S8, respectively. As shown in the figures, the soil temperature at the surface of each column approached the plate temperature of 141 °C (between 130 and 140 °C) within just a few hours. Average and maximum temperatures in the columns decreased as a function of depth. At the 2.5 cm depth, average temperatures for the three soil columns ranged from 128 to 134 °C, while average temperatures at the 10 cm depth ranged from 65 to 80 °C. Average temperatures for the three soils were similar (with ranges spanning between 1 °C and 6 °C) at the 0-, 2.5-, and 5.1-cm depths. At the 7.6 cm and 10 cm depths, the average temperatures for the sand column were higher than those for the loam and clay columns by approximately 12–15 °C.

Table 5.

Temperature data summary for the soil column tests.

		Soil Depth (cm)
		0	2.5	5.1	7.6	10.2
Loam	Average	136 ± 5	128 ± 9	105 ± 9	76 ± 8	71 ± 7
	Maximum	141	135	111	84	79
Clay	Average	133 ± 3	128 ± 7	104 ± 7	81 ± 6	65 ± 5
	Maximum	138	133	110	89	74
Sand	Average	136 ± 4	134 ± 6	103 ± 6	94 ± 7	80 ± 6
	Maximum	142	141	107	98	85

Average temperatures in °C ± SD.

4. Discussion

After an intentional release of B. anthracis spores in a major city, outdoor surfaces such as soil may become contaminated with the virulent biological agent. Soils may also become contaminated with B. anthracis spores from the animal carcasses resulting from natural anthrax disease outbreaks infecting livestock or wildlife (Griffin et al., 2020), and remain one of the most difficult materials to decontaminate. The present study examined some initial fundamental aspects of using thermal techniques for the in-situ decontamination of soil contaminated with bacterial spores. The present study focused on the use of dry heat, which is characterized as having RH less than 100%. (The use of wet heat, such as steam, to decontaminate soil contaminated with B. anthracis spores in-situ is the subject of an ongoing study by the authors.) The present study examined the inactivation of Bg spores, a surrogate for B. anthracis, in three types of soil, at temperatures that could be readily achievable in the field using an in-situ remediation approach.

The study was conducted in two phases, with Phase 1 examining the time requirements to achieve inactivation of a Bg spore population in small amounts of soil in a temperature-controlled environmental test chamber. In Phase 2 of the study, the scale of the testing was increased via the use of small soil columns, with heat applied to the surface, to examine inactivation of Bg spores over time as a function of depth of spores placed in the soil column. The application of heat at the soil surface via a thermal blanket or horizontal wells could be a potential in-situ decontamination tool for surficial contamination (Iben et al., 1996; Vidonish et al., 2016).

One primary benefit of conducting the soil remediation in-situ is the avoidance of virulent bacterial spores becoming aerosolized (and causing further contamination) during the excavation and transport activities that would occur with ex-situ treatment. The use of thermal treatment in particular has the added benefit of avoiding the use of chemicals for decontamination. Chemical decontaminants are typically hazardous themselves, and their use would create additional environmental, health, and safety issues, such as the need to remove excess decontaminant from the soil, occupational exposure to the decontaminants, and the disposal of hazardous waste. In addition, many chemical decontaminants used for inactivation of B. anthracis spores are not effective in soil. The primary disadvantages of using thermal treatment for soil remediation are the potential cost of producing the energy, and the associated environmental impacts thereof, needed to heat the soil.

4.1. Chamber tests

Our environmental chamber tests fill a gap in data needed to establish inactivation rates and times required to achieve inactivation of Bg spore populations in a dry heat environment at temperatures ranging from 80 to 110 °C. In particular, this study is the first to present data for the D-values and times required to achieve effective decontamination of soil materials contaminated with bacterial spores using heat treatment.

We observed that seven weeks were required to achieve no detectable spores at the 80 °C temperature for two of the three soils (clay and loam); in contrast, thermal treatments for soil contaminated with hydrocarbon pollutants can take several years for full remediation (Vidonish et al., 2016). At the 80 °C temperature but with the air humidified to 80% RH, only 15 days were required to inactivate Bg in soil samples. The reduced contact time to achieve full efficacy, and the concomitant reduced D-values, resulting from humidifying the hot air was expected and is consistent with previous work demonstrating such phenomena (Buhr et al., 2015, 2016). At 110 °C, only a 24-h contact time was sufficient to completely inactivate the spore populations in the clay and loam samples. The use of higher temperatures would speed up the remediation process but would potentially result in a higher cost; an analysis of such costs and energy requirements is beyond the scope of this paper but is recommended for future research.

The inactivation rates, in terms of a D-value, for the Bg spore populations were determined for each of the five Phase 1 chamber tests, for each material. The relatively high R-square results for the D-value determinations attest to the good fit of the data using a log-linear scale. The D-values for the SS coupons and BIs were lower than the soil samples for each test, most likely due to the higher thermal conductivity for the SS materials (the commercial BIs were comprised of SS disks as well) compared to the soils. There were no other apparent trends among the D-values for the three soil materials, which is consistent with thermal conductivity varying little among soil types (Vidonish et al., 2016). For the two 90 °C tests, on average the D-values for the second test were approximately one third lower than the first one. Other than having a longer contact time and fewer samples within the test chamber (fewer time points), there were no differences in the procedures for the two tests to help explain the difference in results. Thus, the difference in results for the two 90 °C tests illustrates some inherent variability in using a biological system.

The determination of D-values for the present study allows us to compare our results with other similar dry-heat bacterial spore inactivation studies. Although literature on D-values for soil materials was non-existent, there are a few studies in the literature reporting D-values for Bg spores on metal materials in temperature ranges similar to ours. For example, in a study using Bg inoculated onto SS discs, Kempf et al. reported a D-value of 135 min for 115 °C (uncontrolled, unreported RH) (Kempf et al., 2005), a result comparable to the 3.9–4.0 h D-value we obtained at the somewhat lower temperature of 110 °C for the BIs and SS coupons. Brannen et al. (Brannen and Garst, 1972) reported a D-value range for Bg spores inoculated onto aluminum discs of 1.1–2.9 h (105 °C, RH = 0.003–1.7%). With Bg inoculated into SS cups and placed in a tin can, Peeler et al. (1977b) reported a D-value of ~0.7 h at 105 °C with RH < 0.001, and at 113 °C, with RH 5–100%, D-values ranged from approximately 1–20 h. However, in the same study but at 90 °C and RH < 0.001, Peeler et al. reported a D-value of approximately 3 h, which is considerably lower than the 1–2-day D-values we found at 90 °C for the SS coupons and BIs.

4.2. Column tests

The use of dry heat to decontaminate soil contaminated with B. anthracis spores, via heat applied to the surface using a thermal blanket or heating plate approach, was proven in concept using the small soil columns. With the temperature of the surface heat source set at 141 °C, we were able to achieve no detectable spores at the 2.5 cm depth at four days and at the 5.1 cm depth at 21 days, for two of the three soils. As the average temperatures for the soil columns decreased with depth, so too did the decontamination efficacy, as would be expected. Relatedly, the average temperatures for the three soil columns correlated to the 5.1 cm depth, consistent with low variation in thermal conductivity of different soil types (Vidonish et al., 2016).

Higher surface temperatures would allow for further heat penetration to lower depths and reduce times to achieve desired spore inactivation. Further engineering analysis and research of the tradeoffs between parameters such as surface temperature, energy requirements and cost, time requirements, and soil depths is recommended for a future study. A larger-scale field study would be required to conduct such an analysis, similar to the one that examined the use of a thermal blanket approach to desorb surficial polychlorinated biphenyl contamination (Iben et al., 1996). (In that field study, the investigators custom-made a thermal blanket that was placed on the surface of the soil. The blanket was comprised of resistive heating elements, then a cover of insulative material and a silicone rubber sheet was placed on top of that. We would envision a similar approach could potentially be used B. anthracis contamination of soils.) A field study could also address the aerosolization of B. anthracis spores.

We expected the controlled temperature Phase 1 environmental chamber tests to allow us to gauge times to expect for inactivating spores in the Phase 2 column tests. However, we found that the times to achieve spore inactivation in the soil columns at a given depth with an associated average temperature were considerably longer than expected based on similar temperatures from the Phase 1 test results. For example, at the 5.1 cm depth, temperatures averaged between 103 and 105 °C, which may be compared to the 110 °C chamber test. In that 110 °C test, we achieved nearly complete inactivation at 24 h, but in the column tests, > 5 log CFU were recovered from the pouch samples at 7 days. At the 7.6 cm depths for the loam and clay soils, their average temperatures (76 and 81 °C) are comparable to the 80 °C Phase 1 test. In that chamber test, we observed a reduction in spore recovery of approximately 2 log CFU at 21 d, compared to no inactivation of spores in the column for the same time period. It is likely that the lack of agreement between the two test phases in the present study is due to the additional time required to reach a plateau temperature in the column.

5. Conclusion

The experiments described in the study demonstrate the feasibility and characterize the parameters affecting the use of dry thermal techniques to decontaminate soils that have been surficially contaminated with B. anthracis spores. Further research is needed to determine the field applicability and optimization of this method. Conditions found to be effective in inactivating bacterial spores in soil should also be effective for less resistant microorganisms such as vegetative bacteria and viruses.