Decontamination of Bacillus Spores with Formaldehyde Vapor Under Varied Environmental Conditions

Introduction

The use of formaldehyde (H₂CO; CAS No. 50-00-0) vapor as an effective sterilant and decontaminant for indoor enclosures, especially against persistent biological select agents and toxins (BSAT), has been documented for decades. ^{1 -4} Testing has shown that gas and vapor decontamination technologies are more likely to be affected by variations in temperature and humidity ⁵ compared to the other forms (eg, liquids, foams). Although there is a voluminous amount of data with respect to decontamination efficacy from H₂CO vapor, ^6,7 little information is available with respect to the effect of environmental conditions on H₂CO air concentrations and corresponding efficacies. To conduct effective, full-scale response or remediation of malicious, accidental, or natural releases of BSAT, data are needed to better understand the effect of environmental conditions on H₂CO decontamination efficacy; remediation actions may need to be rendered under less-than-ideal conditions.

To achieve the project goals, reliable methods were required to determine H₂CO air concentrations. Some previous efforts have relied on theoretical H₂CO air concentrations based on the mass of sublimated paraformaldehyde (PF) or heated (evaporated) formalin. ⁸ A systematic study to examine H₂CO air concentrations, at specified time-points, will provide the data needed to develop H₂CO profiles to understand the fate of the vapor over a prolonged duration. Also, controlling as many of the environmental parameters of the test chamber enclosure as possible will contribute to the reliability of the data generated, as well as consistency from trial to trial. These factors combined with the corresponding agent recovery data will give insight into the conditions appropriate to achieve desired decontamination efficacy levels.

A previous effort by Ngabo et al ⁸ demonstrated that lower H₂CO air concentrations (752 parts per million [ppm] theoretical concentration) caused lower decontamination efficacies (less than 6-log reduction) when Bacillus atrophaeus spores inoculated onto nonporous stainless steel material coupons were exposed for 2 hours. When Ngabo et al extended the contact time to 6 hours in the same small test chamber (0.59 m³), a greater than 6-log reduction was achieved, but the confirmatory liquid broth assessments resulted in positive cultures, thereby showing that the B atrophaeus remained viable. Higher H₂CO air concentrations (ie, higher formalin volumes; greater than or equal to 1521 ppm theoretical concentration) resulted in greater than or equal to 6-log reduction of B atrophaeus with negative growth in the liquid broth at relatively short exposure times (30 minutes, 1 hour). Beswick et al ⁹ demonstrated log reductions between 2.27 and 5.55 of Geobacillus stearothermophilus spores inoculated onto stainless steel discs after a 3-hour exposure of 600 ppm H₂CO vapor (theoretically calculated; not actually measured) generated from formalin in a relatively large enclosure (35 m³). When H₂CO vapor was generated in a larger enclosure (150 m³), a much higher H₂CO concentration was used (1400 ppm theoretical concentration), and complete inactivation was achieved within the detectable limit (6-log reduction). ⁹ Taylor et al ¹⁰ described heating (ie, depolymerization) PF to generate a contaminant-free, dry H₂CO gas, whereas the H₂CO gas generated from formalin may have included methanol and formic acid. Rogers et al ⁷ described the effect when the vapor phase equilibrium of H₂CO was exceeded inside a test chamber. Specifically, less H₂CO is available in the air, so the concentration of H₂CO remained below the vapor phase equilibrium to prevent condensation onto the surfaces inside of the test chamber. Ackland et al ¹¹ determined that the H₂CO vapor phase equilibrium was 1628 ppm (20°C-21°C testing temperature). When the vapor phase equilibrium is exceeded, H₂CO condensation and subsequent deposition may occur on surfaces inside the test chamber, thereby reducing the H₂CO vapor in the air. Formaldehyde deposition onto surfaces within a fumigated space has an unknown effect on decontamination efficacy against contaminants residing on those surfaces.

The purpose of the current study was to understand H₂CO vapor profiles under different environmental conditions and their effect on decontamination efficacy against dried, surface-bound Bacillus spores. The results of this effort may be used to reduce the cost and time required to recover after biological contamination incidents by enhancing our understanding of method performance outside ideal conditions and potentially expanding the options for effective decontamination. Many studies have been conducted to assess the effectiveness of H₂CO vapor as a decontaminant, but some questions remain about the influence of the environment on H₂CO vapor and the consequences of H₂CO vapor adsorption or absorption (ie, sorption) on the available surfaces that undergo decontamination. A better understanding of this fumigant is necessary for prescription of real-world decontamination approaches.

Materials and Methods

Sample Processing and Data Collection

Testing that involved only the B.t. and B.g. (ie, surrogates) was done under Biosafety Level (BSL)–2 conditions. All aspects of B.a. testing were performed under BSL-3 conditions, including whenever the virulent and surrogate strains of each organism were tested side-by-side. Coupons were laid flat in a Class II Biological Safety Cabinet (BSC) and inoculated with approximately 1 × 10⁸ CFU per coupon. A 100-µL aliquot of the test organism (approximately 1 × 10⁹ CFU/mL) was dispensed as 10 droplets (10 µL per droplet) across the surface of the test coupons. For each material type, 5 coupons were used at each of the 5 time-points (48 hours, 32 hours, 24 hours, 16 hours, and 8 hours H₂CO vapor exposure) for Tests 1 to 4 to assess efficacy of the decontamination of the organism at each combination of environmental condition and H₂CO vapor concentrations. (For Tests 5 to 8, three coupons were used at each of the same time-points.) Five additional coupons were used as controls (inoculated, exposed to ambient environmental conditions). All material coupons were dried for a minimum of 1 hour in the BSC under ambient conditions (approximately 22°C and 40% relative humidity [RH]) before testing. In addition, 3 coupons of each material were used as blanks (not inoculated) and were included for each time-point tested. The blank coupons served as controls for potential cross-contamination during testing, as well as the sterility state of the coupons.

Inoculated coupons were exposed to either ambient (approximately 22°C) or low temperature (approximately 10°C) and 70% to 75% RH for up to 48 hours as outlined in Table 1. At each of the 5 time-points tested, H₂CO decontamination efficacy under different environmental conditions was determined. After inoculation and initial drying period, the coupons were placed into an airtight container (Lock and Lock, HPL838P). The container was placed into a BSC. Elevated RH conditions (70%-75%) were achieved either by adding a container of humidity fixed-point salt solution (ie, saturated salt solution) located at the bottom (ie, floor) of the test chamber (ASTM International, 2012) or by the vaporization of water on the same hot-plate used to generate H₂CO vapor. A data-logger (Onset HOBO MX1101) was also placed on the floor of the test chamber and collected both temperature and RH readings every minute for the duration of a specific trial. At the end of the H₂CO vapor exposure duration, samples were collected and extracted by placing each coupon in its corresponding conical tube that contained 10 mL PBST. All tubes were agitated on their sides at room temperature on an orbital shaker for 15 minutes at 200 revolutions per minute.

OPEN IN VIEWER

Table 1.

Test Matrix.^a

Test No.	Test Organism Spores	Targeted PF Mass Sublimated Into Chamber Air, g	Temp, °C	RH, %	Time-points
1	B globigii, B thuringiensis	1	21-22	70-75b	8, 16, 24, 32, 48
2	B globigii, B thuringiensis	5.2	10	70-75b	8, 16, 24, 32, 48
3	B anthracis	1	21-22	70-75c	8, 16, 24, 32, 48
4	B anthracis	5.2	10	70-75c	8, 16, 24, 32, 48
5	B globigii, B thuringiensis, B anthracis	Formalin equivalent to 1	21-22	70-75c	8, 16, 24, 32, 48
6	B globigii, B thuringiensis, B anthracis	Formalin equivalent to 5.2	10	70-75c	8, 16, 24, 32, 48
7	B globigii, B thuringiensis, B anthracis	1	10	70-75c	8, 16, 24, 32, 48
8	B globigii, B thuringiensis, B anthracis	Formalin equivalent to 1	10	70-75c	8, 16, 24, 32, 48

^a PF, paraformaldehyde; RH, relative humidity.

^b RH generated with saturated salt in containers.

^c RH generated with water vapor (boiled).

To assess recovery from coupons inoculated with spores, resulting liquid extracts were removed and a series of 10-fold dilutions was prepared in sterile, cell culture–grade water (Sigma Cat. No. W3500). An aliquot (0.1 mL) of the selected dilutions and, when necessary, the undiluted extracts was plated onto tryptic soy agar (TSA; Hardy Diagnostics Cat. No. G60) in triplicate. The agar plates were incubated at 37°C ± 2°C for 16 to 24 hours. Colonies were observed for colony morphology to ensure physical appearance consistent with inoculated spores and enumerated to determine H₂CO vapor decontamination efficacy after exposure.

Efficacy Calculations

Results from the decontamination tests are presented as total recovery from both control and decontaminated coupons and as decontamination efficacy. Recovery of viable test organism is expressed as the base-10 logarithm of total CFU recovered from the 5 control or test coupons for each material type and time-point tested. Efficacy of the decontamination process was determined by subtracting the base-10 logarithm of the number of viable test organisms extracted from the 5 test coupons from the mean of the base-10 logarithm of the control coupons and was expressed as log reduction (LR) using Equation 1:

LR D e c o n = ( log 10 C F U c ) ¯ − ( log 10 C F U t ) ¯

1

where CFUc is the number of viable test organisms extracted from the control samples and CFUt is the number of viable test organisms extracted from the test samples. For decontaminated test samples where no viable organisms were detected, the efficacy was calculated as the log of the mean number of viable organisms recovered from the control coupons. Using the calculated LR for each test coupon, the mean (±standard deviation [SD]) log reduction was calculated.

Each of the LR results is reported with an associated 95% confidence interval (CI), calculated using Equation 2:

95% CI = efficacy ± ( 1.96 × SE )

2

The term SE is the pooled standard error and was calculated using Equation 3:

SE = S pc 2 5 + S t 2 5

3

where S is the standard deviation of the LR results for either the 5 positive controls (S _pc) or the 5 test coupons (S _t) for each combination of H₂CO vapor concentration, environmental conditions, coupon, and spore tested.

Test Chamber

A 498-L static glovebox (Plas-Labs Model No. 830-ABB) was modified with high-efficiency particulate air (HEPA) filters to allow for the rapid evacuation of H₂CO gas from the test chamber via ducts to the laboratory air-scrubbing system (Figure 1). Due to the ability to rapidly evacuate H₂CO from the test chamber interior, neutralization of the H₂CO with ammonium carbonate (Luftman, 2005) was not needed to reduce H₂CO air concentration at the terminus of experimental trials. ²⁵ Therefore, removal of the resulting hexamethylenetetramine residue on all surfaces, a byproduct of the ammonium carbonate neutralization reaction with H₂CO, was not required. A heat exchanger-fan unit (Alphacool Model No. 14215) was mounted and operated inside the test chamber for efficient air mixing and heat exchange. A circulating water-bath (Thermo Scientific Model No. NESLAB RTE7), filled with automotive anti-freeze solution, was used to achieve low air temperature conditions when prescribed. The test chamber was leak-tested before each trial. A data logger (Onset Cat. No. HOBO MX1101) was placed in the sealed test chamber and collected both temperature and RH measurements every minute for the 48-hour duration of testing.

OPEN IN VIEWER

Figure 1.

Test chamber configuration. (1) HEPA filters for rapid evacuation of H₂CO, (2) heat exchanger-fan unit to circulate air and control test chamber temperature, (3) circulating water-bath to cool test chamber temperature, and (4) HOBO data-logger.

Results

Inoculation levels and recovery from positive controls. Over the course of the study, an average of 7.21 × 10⁷ CFU of B.g. spores, 7.01 × 10⁷ CFU of B.t. spores, and 6.74 × 10⁷ of B.a. spores was applied to the coupons. Positive control coupons (inoculated and processed in the same manner as those exposed to the H₂CO vapor but stored in a separate test chamber) resulted in overall average recoveries of 2.55 × 10⁷ CFU for B.g. (35% recovery), 4.11 × 10⁷ CFU for B.t. (57% recovery), and 2.18 × 10⁷ CFU for B.a. (32% recovery), as shown in Table 2.

OPEN IN VIEWER

Table 2.

Average Inoculum and Recovery per Coupon by Spore Type Across All Tests.^a

		Average Recovery, CFU/coupon		% Recovery
Strain	Average Inoculum, CFU/coupon	Wood	Painted Concrete	Wood	Painted Concrete
B globigii	7.21 × 107 ± 1.88 × 107	4.08 × 106 ± 3.14 × 106	4.43 × 107 ± 1.21 × 107	6	61
B thuringiensis	7.01 × 107 ± 1.11 × 107	1.08 × 107 ± 1.98 × 107	5.85 × 107 ± 9.42 × 107	15	84
B anthracis	6.74 × 107 ± 2.17 × 107	4.53 × 106 ± 2.35 × 106	4.00 × 107 ± 7.83 × 106	7	59

^a Data are means ± standard deviation.

Recoveries were lower for wood compared to the painted concrete across all the tests conducted and spore types used (Table 3). Bacillus globigii recovered from the positive control on wood (average 4.66 × 10⁶ CFU/coupon) were a log lower than spores recovered from painted concrete (average 4.63 × 10⁷ CFU/coupon). Similarly, the recoveries of B.t. from wood (average 1.28 × 10⁷ CFU/coupon) were slightly less than the recoveries of those extracted from painted concrete (average 6.95 × 10⁷ CFU/coupon). Bacillus anthracis recovered from wood (average 2.60 × 10⁶ CFU/coupon) were also a log lower than the spores from the painted concrete (average 3.14 × 10⁷ CFU/coupon). The relatively high standard deviations for B.t. may be attributed to clumping of the DiPel powder when introduced to PBST despite the preparation being thoroughly mixed.

OPEN IN VIEWER

Table 3.

Positive Control Average Recoveries by Material Coupon Type, Spore Type, and Test Conducted.^a

	Wood, CFU/coupon			Painted Concrete, CFU/coupon
Test No.	B globigii	B thuringiensis	B anthracis	B globigii	B thuringiensis	B anthracis
1	9.28 × 106	5.07 × 107		6.52 × 107	5.45 × 107
2	5.21 × 106	5.09 × 105		4.17 × 107	5.13 × 106
3			7.15 × 106			4.49 × 107
4			7.49 × 106			4.57 × 107
5	2.21 × 106	8.56 × 106	4.25 × 106	4.49 × 107	3.63 × 107	2.86 × 107
6	5.27 × 106	1.50 × 106	2.60 × 106	4.87 × 107	2.46 × 108	3.14 × 107
7	1.35 × 106	2.78 × 106	1.71 × 106	3.11 × 107	5.36 × 106	4.44 × 107
8	1.14 × 106	8.55 × 105	3.97 × 106	3.44 × 107	3.82 × 106	4.52 × 107
Averagea	4.08 × 106 ± 3.14 × 106	1.08 × 107 ± 1.98 × 107	4.53 × 106 ± 2.35 × 106	4.43 × 107 ± 1.21 × 107	5.85 × 107 ± 9.42 × 107	4.00 × 107 ± 7.83 × 106

^a Data are means ± standard deviation

Decontamination efficacy. Decontamination efficacy data, expressed as the means and 95% CIs for the log reduction, are presented by test in Figures 2 to 11 and in Supplementary Tables 1 to 8. The highest efficacies were generally achieved with 48 hours of H₂CO vapor exposure across both materials, with lesser efficacies on porous wood for all spore types.

OPEN IN VIEWER

Figure 2.

Test 1: efficacy determinations for vaporization of 1 g PF, room temperature, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration × time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration × time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Test conducted at room temperature (▴) with initial 70%-75% RH (×) environmental conditions. B globigii and B thuringiensis were concurrently tested but represented on separate graphs for clarity.

An initial test that involved the sublimation of 1 g PF at room temperature with a starting RH of 70% to 75% (Test 1, Table 1) resulted in LRs of 3.69 ± 0.10 from wood and 7.80 ± 0.09 from painted concrete for B.g. and 7.70 ± 0.07 from wood and 7.66 ± 0.30 from painted concrete for B.t. after the maximum 48 hour H₂CO vapor exposure (H₂CO concentration × time = 1826 ppm-hours). Time-points 8, 16, 24, and 32 hours resulted in < 5 log reduction (Figure 2a).

In Test 2, 5.2 g of PF was sublimated at approximately 10°C with a starting RH of 70% to 75% (Test 2, Table 1), and these conditions resulted in efficacies of 2.59 ± 0.73 from wood and 3.56 ± 0.49 from painted concrete for B globigii and 7.13 ± 0.95 from wood and 6.70 ± 0.11 from painted concrete for B thuringiensis after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 1815 ppm-hours). High decontamination efficacies were observed for painted concrete at the 24- and 32-hour time-points (5.79 and 6.36 log reductions, respectively) for B globigii, as well as 16- and 24-hour time-points (6.70 and 5.73 log reductions, respectively) for B thuringiensis. The resulting H₂CO concentration × time values for 16-, 24-, and 32-hour time-points were 711, 714, and 735 ppm-hours, respectively. Other time-points resulted in < 5 log reduction (Figure 3).

OPEN IN VIEWER

Figure 3.

Test 2: efficacy determinations for vaporization of 5.2 g PF, approximately 10°C, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Low temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration. B globigii and B thuringiensis were concurrently tested but represented on separate graphs for clarity.

A subset of tests was conducted with B.a. to confirm the decontamination efficacy data obtained with B globigii and B thuringiensis. In Test 3 (compared to Test 1), the sublimation of 1 g PF at room temperature with a starting RH of 70% to 75% (Table 1) resulted in B. anthracis decontamination efficacies of 6.84 ± 0.12 from wood and 7.65 ± 0.02 from painted concrete after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 14,526 ppm-hours) (Figure 4). High decontamination efficacies were observed for painted concrete at the 24- and 32-hour time-points (5.85 and 6.99 log reductions, respectively). The resulting H₂CO concentration × time values for the 24- and 32-hour time-points were 2269 and 3235 ppm-hours, respectively. Other time-points resulted in <4 log reduction (Figure 4).

OPEN IN VIEWER

Figure 4.

Test 3: efficacy determinations for vaporization of 1 g PF, room temperature, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Room temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration.

In Test 4 (compared to Test 2), the sublimation of 5.2 g PF at approximately 10°C with a starting RH of 70% to 75% (Table 1) resulted in B anthracis efficacies of 2.86 ± 0.72 from wood and 7.65 ± 0.07 from painted concrete after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 5499 ppm-hours). High decontamination efficacies were observed for painted concrete at the 16-, 24-, and 32-hour time-points (6.34, 7.65, and 7.65 log reductions, respectively). The resulting H₂CO concentration × time values for the 16- and 32-hour time-points were 2182 and 2290 ppm-hours, respectively (the H₂CO reading for time-point 24 hours was not collected). Other time-points resulted in < 3 log reduction (Figure 5).

OPEN IN VIEWER

Figure 5.

Test 4: efficacy determinations for vaporization of 5.2 g PF, approximately 10°C, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Low temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration.

During Test 5, the PF-equivalent mass of formalin for 1 g PF was vaporized at room temperature with a starting RH of 70% to 75% (Test 5, Table 1). These conditions resulted in efficacies of 6.26 ± 0.38 from wood and 7.64 ± 0.16 from painted concrete for B.g., 6.65 ± 0.68 from wood and 7.47 ± 0.44 from painted concrete for B.t., and 6.62 ± 0.09 from wood and 7.43 ± 0.21 from painted concrete for B.a. after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 7222 ppm-hours). High decontamination efficacies were observed for painted concrete at the 32-hour time-point (5.04 log reductions) for B thuringiensis, as well as 24- and 32-hour time-points (6.61 and 6.92 log reductions, respectively) for B anthracis. The resulting H₂CO concentration × time values for 24- and 32-hour time-points were 1115 and 1622 ppm-hours, respectively. Other time-points resulted in < 5 log reduction (Figure 6).

OPEN IN VIEWER

Figure 6.

Test 5: efficacy determinations for vaporization of 1 g PF-equivalent mass of formalin, room temperature, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (D) Room temperature (▴) with starting 70-%75% RH (×) environmental conditions over a 48-hour exposure duration. B globigii, B thuringiensis, and B anthracis were concurrently tested but represented on separate graphs for clarity.

For Test 6, the PF-equivalent mass of formalin for 5.2 g PF was vaporized at approximately 10°C with a starting RH of 70% to 75% (Test 6, Table 1). These conditions resulted in efficacies of 6.70 ± 0.20 from wood and 7.68 ± 0.08 from painted concrete for B.g., 6.13 ± 0.29 from wood and 7.44 ± 1.41 from painted concrete for B.t., and 6.37 ± 0.31 from wood and 7.49 ± 0.08 from painted concrete for B.a. after the maximum 48 hour H₂CO vapor exposure (H₂CO concentration × time = 6414 ppm-hours). High decontamination efficacies were observed for wood at the 24- and 32-hour time-points (6.13 and 6.13 log reductions, respectively) for B thuringiensis as well as 16-, 24-, and 32-hour time-points (6.37, 6.37, and 5.86 log reductions, respectively) for B anthracis. The resulting H₂CO concentration × time values for the 16-, 24-, and 32-hour time-points were 1602, 1723, and 1901 ppm-hours, respectively. Other time-points resulted in <5 log reduction (Figure 7).

OPEN IN VIEWER

Figure 7.

Test 6: efficacy determinations for vaporization of 5.2 g PF-equivalent mass of formalin, approximately 10°C, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (D) Low temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration. B globigii, B thuringiensis, and B anthracis were concurrently tested but represented on separate graphs for clarity.

In Test 7, the sublimation of 1 g PF at approximately 10°C with a starting RH of 70% to 75% (Table 1) resulted in efficacies of 6.49 ± 1.95 from painted concrete for B.g., 6.21 ± 0.68 from wood and 6.72 ± 0.087 from painted concrete for B.t., and 6.84 ± 1.59 from painted concrete for B.a. after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 6614 ppm-hours). High decontamination efficacies were observed for painted concrete at the 32-hour time-point (6.22 log reduction) for B thuringiensis. At the 48-hour time-point for B.a. on wood, the highest log reduction was 5.05 ± 1.19. The resulting H₂CO concentration × time values for the 32- and 48-hour time-points were 2419 and 6614 ppm-hours, respectively. Other time-points resulted in <4 log reduction (Figure 8).

OPEN IN VIEWER

Figure 8.

Test 7: decontamination efficacy of 1 g PF, approximately 10°C, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (D) Low temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration. B globigii, B thuringiensis, and B anthracis were concurrently tested but represented on separate graphs for clarity.

During Test 8, the PF-equivalent mass of formalin for 1 g PF was vaporized at approximately 10°C with a starting RH of 70% to 75% (Test 8, Table 1). These conditions resulted in lower efficacies than previous tests. (This test was the first test that involved the cold temperature with less than saturation H₂CO vapor present in the test chamber.) At the 48-hour time-point, low efficacies of 1.70 ± 0.31 from wood and 2.62 ± 1.048 from painted concrete for B.g., higher efficacies of 5.05 ± 1.62 from wood and 4.43 ± 2.19 from painted concrete for B.t., and 3.82 ± 0.74 from wood and 5.48 ± 2.15 from painted concrete for B.a. after the maximum 48-hour H₂CO vapor exposure (H₂CO concentration × time = 1654 ppm-hours) were observed, as shown in Figure 9.

OPEN IN VIEWER

Figure 9.

Test 8: decontamination efficacy of 1 g PF-equivalent mass of formalin, approximately 10°C, initial 70%-75% RH, 48-hour exposure duration. (A) Decontamination efficacies determined at 5 time-points for B globigii on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (B) Decontamination efficacies determined at 5 time-points for B thuringiensis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (C) Decontamination efficacies determined at 5 time-points for B anthracis on wood (•) and painted concrete (□), and concentration x time (ppm-hours) of H₂CO vapor calculated at each time-point (- - -). (D) Low temperature (▴) with starting 70%-75% RH (×) environmental conditions over a 48-hour exposure duration. B globigii, B thuringiensis, and B anthracis were concurrently tested but represented on separate graphs for clarity.

Discussion

Except for prions, bacterial spores exhibit higher resistance to chemical and physical inactivation than all other microorganisms. ¹⁶ For this reason, the foremost interest was with B.a. since any effective decontamination approach against the B.a. spores will likely be efficacious against non–spore-forming bacteria. Also, Bacillus spores were tested in this study due to the historical data available in the decontamination arena. ⁴

A possible leak from the test chamber could explain a disparity of H₂CO vapor concentrations between Tests 1 and 3 (Table 1). Both tests involved similar parameters (coupon types, environmental conditions, and time-points), with the only difference being spore type. Test 1 was one of the first tests conducted in this study, and there was a clear difference between individual H₂CO vapor concentrations per time-point, which translated into disproportional concentration × time ppm-hour values between the 2 trials. Practices to prevent H₂CO vapor leakage as well as safety modifications and a different exhausting technique were employed as described by Coldiron et al ¹⁷ upon reviewing the data from Tests 1 and 3.

Munro et al ⁶ showed that the optimal formaldehyde gas decontamination parameters for B.a. spores were 66% RH, a minimum temperature of 28°C, and 10.5 g PF/m³. Since the test chamber used for this effort was 0.5 m³, the maximum mass of PF used was 5.25 g (10.5 g PF/m³; Trial Nos. 2, 4, and 6, Table 1). In other trials, 1 g of PF (Trial Nos. 1, 3, 7, and 8, Table 1) was used to achieve less-than-saturation H₂CO conditions in the test chamber. Rogers et al⁷ showed consistent H₂CO air measurements out to 10 hours after vapor generation. The decrease in H₂CO concentration, due to sorption to all test chamber interior components, may have occurred between the 10-hour and 16-hour time-points since the generation methods between the work described in Rogers et al⁷ and this effort were similar (ie, both generation methods used PF heated in a container). An indicator that H₂CO sorption readily occurred inside the test chamber was that it took several days (up to 6 days in some cases) to clear the test chamber of H₂CO (as measured with a low-level sensor, Draeger Gas Detection Tubes, Cole-Parmer Cat. No. EW-86514-34) for safe reentry after H₂CO evacuation (0.75 ppm permissible exposure limit measured as an 8-hour time-weighted average, 29 CFR 1910.1048) in preparation for the next trial. ¹⁸

There was an appreciable amount of surface area inside the test chamber due to the heat exchanger-fan unit (ie, presence of heat-conductive aluminum tubing throughout the unit) being present. The circulating water-bath that was connected to the heat exchanger-fan unit was set at approximately 3°C to attain a test chamber air temperature of 10°C ± 2°C for the low-temperature tests. Since the dew point with the low-temperature test chamber configuration was calculated at 4.8°C, any temperature less than 4.8°C would have produced a condensate on the heat exchanger-fan’s aluminum tubing due to the elevated RH, thereby removing moisture from the air. Air measurements taken immediately after sublimation of 5.2 g PF resulted in concentrations between 128 and 532 ppm, much lower than those measured at room temperature. H₂CO readily solubilizes in water. ¹⁹ Therefore, the condensate formed on the aluminum tubing may have scavenged H₂CO from the air, thereby reducing the air concentrations. Abraham et al ²⁰ described having to pay close attention to RH levels during the winter months when decontaminating filter canisters mounted in laboratories. The cold supply-air (introduced from the outside) was heated once inside the laboratory facility (from 10°C to 21°C), thereby decreasing the RH in the air. In the study Abraham et al, which involved 604 separate, large-scale tests with the introduction of more water vapor to increase RH to the prescribed levels during winter, did not result in noticeably higher spore strip failure. Also, Spiner and Hoffmann ² clearly showed the relationship between RH, H₂CO vapor concentration, nature of the material surface, and even lot-to-lot variability with microorganism preparations can greatly affect decontamination efficacy rates. In addition, high RH environmental conditions (≥65%) were responsible for the highest H₂CO vapor decontamination efficacies against Bacillus spores as described in comprehensive testing by Abraham et al, ²⁰ as well as in a review of available building decontamination data by the EPA. ²¹

In this study, H₂CO vapor generated from formalin resulted in lower decontamination efficacies than H₂CO vapor generated from PF, at the low-temperature condition. Decontamination efficacies (log reduction) of 2.62, 4.43, and 5.48 were achieved with formalin, after 48-hour exposure at ∼10°C, while efficacies of 6.49, 6.72, and 6.84 were achieved with PF, for B.g., B.t., and B.a. spores on concrete coupons, respectively (Figure 9). Efficacy of H₂CO vapor generated from formalin was 1.70, 5.05, and 3.82 log reduction for B.g., B.t., and B.a. spores on pine wood coupons, respectively, after a 48-hour exposure at ∼10°C. Comparatively, efficacies of PF-generated H₂CO vapor were 3.49, 6.21, and 5.05, respectively, for B.g., B.t., and B.a. spores on pine wood.

Decontamination efficacies from PF were relatively consistent throughout the 5 tests, with low variability at each time-point (Figure 10), compared to higher variability in efficacies from similar conditions in tests using formalin (Figure 11). For both PF- and formalin-generated H₂CO vapor, however, decontamination efficacy varied directly with exposure duration (Figures 10 and 11).

OPEN IN VIEWER

Figure 10.

Regression analysis of exposure duration and decontamination efficacy for 3 spore types inoculated on wood from 5 separate tests that used PF to generate H₂CO vapor. (A) B globigii on wood, (B) B thuringiensis on wood, (C) B anthracis on wood.

OPEN IN VIEWER

Figure 11.

Regression analysis of exposure duration and decontamination efficacy for 3 spore types inoculated on wood from 3 separate tests that used formalin to generate H₂CO vapor. (A) B globigii on wood, (B) B thuringiensis on wood, (C) B anthracis on wood.

Another aspect that must be considered was that the Bacillus spores from suspension were dried onto the coupon surface before their exposure to H₂CO vapor for this study. The benefit of using a relatively small test chamber was that variables were closely controlled, including the mixture of the air with the heat exchanger and fan to distribute the H₂CO vapor throughout the interior. A release that involves dry, aerosolized spores is plausible, ^22,23 and the use of a fan to distribute a decontaminant such as H₂CO vapor may redistribute contaminants and increase the size of the area contaminated. It is unknown how aerosol-deposited dry spores would respond to H₂CO vapor treatments. Further work should be conducted to determine efficacy of H₂CO when used to treat surfaces contaminated by dry spore deposition. Each contamination incident may warrant a unique remediation approach.

Conclusions

In most tests, wood was more difficult to decontaminate (lower comparative efficacies) than painted concrete. The only exception occurred when a PF-equivalent mass of formalin for 5.2 g PF was vaporized at approximately 10°C (Test 6, Table 1). Trials conducted with approximately 10°C internal air temperatures (Test Nos. 2, 4, 6, 7, and 8, Table 1) resulted in a precipitous drop in H₂CO air concentrations. Regardless of the environmental conditions, a decrease in concentration consistently occurred from time-point 0 hour to 16 hours. At 16 to 48 hours, the concentrations remained relatively consistent for all environmental conditions when using 5.2 g PF, which suggests that an equilibrium was reached at some point before 16 hours for this test chamber configuration.

Variable and unpredictable environmental conditions are expected in the field. Responders who plan on deploying H₂CO vapor in a field setting will prescribe a process and predict results based on the data generated in a laboratory at the bench-scale. ²⁴ The bench-scale data presented here demonstrate that environmental conditions can affect H₂CO air concentrations and H₂CO decontamination efficacies. Careful planning of the decontamination operation, to eliminate potential cold surfaces and prevent condensation-based H₂CO loss, is warranted. The surface area-to-volume ratio of the decontamination space should also be minimized to the extent possible, to achieve proper mixing and reduce H₂CO loss to surfaces.

Supplemental Material