Scaling Oxygen Scavengers in Hermetic Bags for Improved Grain Storage

Abstract

The phasing out of most chemicals has created a demand for alternative methods to preserve grain quality and market value. Hermetic storage offers a chemical-free solution for pest control by creating an airtight environment that naturally leads to insect death. Adding oxygen scavengers can further enhance hermetic storage by accelerating oxygen depletion. However, no study has examined scaling hand warmers in hermetic storage bags used by large grain handlers and farmers. We evaluated the effects of 1, 2, or 3 hand warmers in 25-kg PICS bags and 2, 4, or 6 hand warmers in 50-kg PICS bags on oxygen consumption and grain quality. We hypothesized that doubling the number of hand warmers used in 25-kg to 50-kg PICS bags would maintain the same rate of oxygen reduction. Oxygen levels decreased as the number of hand warmers increased. Additionally, oxygen concentrations in 25-kg PICS bags with 1, 2, or 3 hand warmers closely mirrored those of 2, 4, or 6 hand warmers in 50-kg PICS bags, respectively. Using 2 or 3 hand warmers in 25-kg PICS bags and 4 or 6 hand warmers in 50-kg PICS bags reduced oxygen concentrations below the 5% threshold for pest suppression within 12 h and maintained it for at least 8 days. While a slight rise in relative humidity was observed with more hand warmers, this did not negatively affect seed moisture content or germination rates. Doubling hand warmers along with the bag size from 25 to 50 kg produced similar oxygen depletion rates. These findings are helpful for large grain handlers and farmers who use 50-kg hermetic bags to store seeds or specialty crops to maintain quality. Hermetic bags combined with hand warmers promote sustainability by reducing chemical usage and minimizing food and nutrient losses.

1. Introduction

Postharvest storage poses significant challenges from biotic and abiotic factors, which can lead to quality degradation. Hermetic storage effectively mitigates losses caused by these threats [1,2,3,4,5,6]. Insects in hermetic containers consume oxygen through feeding and respiration, and below 5%, their metabolism slows, ultimately causing mortality [7,8,9,10,11]. However, in cases of low infestation, high residual oxygen, or low temperature, oxygen depletion may take longer, potentially resulting in grain damage [12,13,14] or quality (nutrient) loss [15,16].

Due to regulatory and health concerns, many countries maintain a low tolerance for insect presence in grains. With the phasing out of several pesticides and insecticide resistance, achieving this standard has become challenging, highlighting the urgent need for alternative pest control methods. Although hermetic storage limits pest infestation, insects can sometimes survive for months due to intergranular residual oxygen. This excess residual oxygen requires additional mechanisms for removal. Iron-based oxygen scavengers, typically composed of ferrous iron powder, salt, and activated carbon, are used to remove this excess oxygen in food packaging [17,18,19]. Combining hermetic storage with oxygen scavengers accelerates oxygen depletion, providing an effective solution.

Hand warmers, though not originally designed as oxygen scavengers, share the same iron-based composition as commercial alternatives such as OxySorb [20]. While ingredient variations exist, oxygen depletion consistently occurs via oxidation (Equation (1)). When exposed to air, ferrous iron oxide absorbs moisture, triggering its oxidation into hydrated iron [17].

$$4\mathrm{F}\mathrm{e}+3{\mathrm{O}}_2+6{\mathrm{H}}_2\mathrm{O}\to 4\mathrm{F}\mathrm{e}({\mathrm{O}\mathrm{H})}_3$$

(1)

Hand warmers have proven to be more cost-effective for oxygen depletion in hermetic storage than OxySorb [21]. They are particularly appealing due to their lower cost and ability to deplete twice as much oxygen. Within 24 h, they effectively reduce oxygen levels below 5%, leading to insect mortality, decreased egg laying, and suppressed progeny development [22]. Moreover, they do not significantly affect temperature, relative humidity, grain moisture content, or seed germination.

Hand warmers have been tested as oxygen scavengers in small containers such as 15.2-L jars and 25-kg hermetic bags [21,22]. However, these are smaller than containers typically used by farmers and grain handlers. Scaling their use to 50-kg hermetic bags is valuable, as these are commonly used for grain and seed storage. The 50-kg PICS bags are recommended for their ease of handling [23]. Testing oxygen scavengers in these bags offers a promising chemical-free alternative for preserving stored products in large facilities. In addition, it provides a solution for preserving oxygen-sensitive products such as provitamin A maize or cassava, which degrade with prolonged exposure to oxygen [24,25,26,27,28,29].

This study builds on previous research [21,22] to explore the scalability of hand warmers in different sizes of hermetic bags. Specifically, we aimed to (i) examine how the number of hand warmers and container size influence oxygen reduction and (ii) assess the effect of increased hand warmers on environmental conditions (e.g., temperature and relative humidity) and seed quality. We hypothesize that doubling the number of hand warmers with increasing container size will result in similar oxygen depletion patterns. This research fills key knowledge gaps, including (i) the relationship between hand warmer quantity, container size, and oxygen reduction, and (ii) the potential effect on grain quality. The findings will guide grain handlers in adopting sustainable storage practices, providing an eco-friendly alternative to chemical pesticides while enhancing food security through improved hermetic storage.

2. Materials and Methods

2.1. Experiment Set-Up

This study was conducted from fall 2023 to spring 2024 at the Postharvest Innovation for Crop Storage Laboratory in the Department of Entomology at Purdue University, West Lafayette, IN, USA. Untreated Pioneer^® hybrid maize seeds (variety 1222AM) from Corteva Agriscience (Johnston, IA, USA) were obtained from the Agronomy Center for Research and Education (ACRE) at Purdue University, West Lafayette, IN, USA. Seeds were cleaned using Seedboss^® seed-processing equipment (ALMACO, Nevada, IA, USA) at ACRE. HotHands^® 10-h hand warmers (2″ × 3.5″ pouches) from Kobayashi Healthcare International, Inc. (Dalton, GA, USA) were purchased online from Amazon.com, Inc.

Two sizes of the Purdue Improved Crop Storage (PICS) bags, 25 kg and 50 kg, were used in this experiment [23]. The 25-kg bags were filled with 20 kg of maize, while the 50-kg bags held 40 kg. After filling each bag with maize, hand warmers were placed on top of the maize before sealing the bags. Both polyethylene liners were tied together, while the woven bag was closed separately using zip ties. Treatments combined bag size and the number of hand warmers: 1, 2, or 3 hand warmers for 25-kg PICS bags and 2, 4, or 6 hand warmers for 50-kg PICS bags. Each treatment had 4 replicates, resulting in 24 PICS bags. Due to limited space in the laboratory, the 50-kg PICS bag experiment was conducted from 15 to 26 November 2023, followed by the 25-kg PICS bag experiment from 27 December 2023 to 6 January 2024.

2.2. Data Collection

2.2.1. Oxygen Concentration

To measure oxygen concentration, fluorescent yellow Oxydot sensors (Industrial Physics, Boston, MA, USA) were affixed to two Petri dishes and attached to the middle translucent lining of the PICS bags using PERMATEX^® Clear Adhesive Sealant Silicone RTV 80050 (Solon, OH, USA). Circular openings, matching the size of the Petri dishes, were cut in the inner liner and woven bag, while the middle liner remained intact to allow oxygen concentration monitoring using a sensor. Each Petri dish was positioned over the cut area with the Oxydot sensors facing the interior of the inner layer and secured using a hot glue gun. To assess the oxygen gradient, Oxydot sensors were placed one-quarter from the top and bottom of the grain in each PICS bag (Figure 1).

The oxygen level in PICS bags was monitored using an OxySense^® 525OI Oxygen Analyzer (Industrial Physics, Devens, MA, USA). The analyzer, equipped with a fiber optic oxygen reader, measured the oxygen concentration (O₂%) inside the PICS bags by detecting the changes in the intensity and fluorescence properties of the Oxydot sensors. Measurements were taken at set intervals: 0, 1, 3, 6, 8, 12, 18, and 24 h. Subsequently, oxygen levels were measured twice daily at 12-h intervals up to 240 h.

2.2.2. Temperature and Relative Humidity

Data loggers (EL-USB-2, Lascar Electronics Inc., Erie, PA, USA) were programmed to record the temperature and relative humidity at 30 min intervals. Two data loggers were placed inside each PICS bag near the Oxydot sensors one-quarter from the top and bottom to assess potential gradients (Figure 1). A control data logger was placed in the room to track the ambient temperature and relative humidity during the experiment.

2.2.3. Seed Moisture Content and Germination

Seed samples were collected before and after the experiment to assess changes in seed quality, namely moisture content and seed germination. Initial samples were randomly collected from the top, middle, and bottom of the grain in each PICS bag using a probe and mixed to form 250 g for testing. Three mesh bags, each containing 100 g of grain, were placed at the top, middle, and bottom of the PICS bags for post-experiment quality assessment. The top mesh bag represented the grain closest to the hand warmers, while the middle and bottom bags represented the grain further away from the hand warmers.

Moisture content was measured using the oven-drying method [30]. Three 15 g maize samples from each mesh bag were dried in tin cups at 103 ± 1 °C for 72 h. Moisture content was calculated based on the weight difference before and after drying. Seed germination testing followed the International Rules for Seed Testing protocol [31]. Four 25 seeds (100 seeds total) were taken from both initial and post-experiment sample bags. The seeds were evenly placed in a Petri dish lined with damp filter paper and kept moist to stimulate germination. Germination rates were recorded daily over seven days to evaluate seed quality.

2.3. Data Analysis

All statistical analyses were conducted using Prism-GraphPad version 10.0.3 (GraphPad Software, San Diego, CA, USA). Tukey’s multiple comparison test was performed to compare the means of oxygen concentration, temperature, relative humidity, and moisture content, with a 95% confidence level for mean separation. When no statistical differences were observed between the top and bottom for oxygen concentration, temperature, and relative humidity or between the top, middle, and bottom for moisture content and germination, the data were pulled together for further analysis. Simple and multiple regression models were applied to examine the relationship between the oxygen concentration, number of hand warmers in each PICS bag size, and exposure time.

3. Results

3.1. Oxygen Concentration

There was no difference in the average oxygen concentration (%) between the Oxydot sensors placed toward the top and bottom of the grain in the PICS bags at any time point, indicating no oxygen gradient. The data were pulled together for further analysis. The F-statistic measures the ratio of explained to unexplained variance, indicating the significance of the regression model. Oxygen depletion was significantly influenced by the number of hand warmers (F_{(5, 18)} = 37.52, p < 0.0001) and exposure time (F_{(9, 162)} = 5543, p < 0.0001). A significant interaction was observed between treatment and exposure time (F_{(45, 162)} = 23.53, p < 0.0001). For most treatments, except those with 1 hand warmer in 25-kg PICS bags and 2 hand warmers in 50-kg PICS bags, oxygen levels dropped below the critical 5% threshold for pest suppression within the first 12 h (Figure 2). These exceptions reached the 5% threshold after 48 h. Additionally, the oxygen depletion trend for a treatment in 25-kg PICS bags closely mirrored that of its corresponding double in the 50-kg PICS bags.

During the first 48 h, the average oxygen concentration (%) in PICS bags was similar across comparable hand warmer treatments, except for the treatment consisting of 1 hand warmer in a 25-kgPICS bag and 2 hand warmers in a 50-kg bag at 36 and 48 h (

Table 1. Average oxygen concentration (%) inside grain-filled 25-kg and 50-kg PICS bags over 48 h.

		Oxygen Concentration (%, Mean ± SEM) *
			Time
Treatment **	0 h	6 h	12 h	24 h	36 h	48 h
25-kg PICS bag—1 hand warmer	20.9 ± 0.20 aA	14.2 ± 0.65 aB	9.1 ± 1.07 aC	5.9 ± 0.94 aD	5.5 ± 0.73 aE	5.2 ± 0.52 aE
25-kg PICS bag—2 hand warmers	20.9 ± 0.15 aA	9.7 ± 0.78 bB	4.7 ± 0.53 bC	2.1 ± 0.30 bD	1.7 ± 0.19 cDE	1.3 ± 0.12 cE
25-kg PICS bag—3 hand warmers	20.9 ± 0.06 aA	6.9 ± 0.12 cB	2.5 ± 0.09 cC	0.9 ± 0.07 bD	0.8 ± 0.04 cD	0.7 ± 0.01 cD
50-kg PICS bag—2 hand warmers	20.4 ± 0.11 aA	13.7 ± 0.11 aB	8.5 ± 0.27 aC	4.8 ± 0.30 aD	3.6 ± 0.28 bD	3.0 ± 0.24 bD
50-kg PICS bag—4 hand warmers	20.4 ± 0.04 aA	10.0 ± 0.35 bB	4.5 ± 0.25 bC	2.0 ± 0.34 bD	1.2 ± 0.15 cD	0.9 ± 0.18 cD
50-kg PICS bag—6 hand warmers	20.4 ± 0.08 aA	7.3 ± 0.32 cB	2.4 ± 0.13 cC	0.8 ± 0.09 bD	0.7 ± 0.08 cD	0.6 ± 0.08 cD

* All data are means ± standard error of means (SEM). Under the same variable, within the same time or among treatments (lowercase) or within each treatment or across times (uppercase), means followed by the same letter are not significantly different (p < 0.05). ** Treatments consisted of 1, 2, or 3 10-h hand warmers in 25-kg PICS bags or 2, 4, or 6 10-h hand warmers in 50-kg bags.

). In the first 12 h, each treatment of a hand warmer in 25-kg PICS bags and its corresponding double in 50-kg bags showed significant differences. Beyond this point, no differences were observed between the 2 or 3 hand warmers in 25-kg bags and their corresponding double (4 or 6 hand warmers) in 50-kg bags. After 24 h, no significant differences were observed within each treatment, except for the 1 and 2 hand warmers in 25-kg PICS bags.

Table 2. Multiple regression of oxygen concentration (%) as a function of the number of hand warmer units and time (log_hour) with estimated marginal slopes, standard errors, and t- and p-values.

Treatment *	Estimated Marginal Slopes	Standard Error	t-Value	p-Value
Intercept	18.58	0.4481	41.46	<0.0001
25-kg PICS bag—1 hand warmer	0	0	.
25-kg PICS bag—2 hand warmers	−3.342	0.5419	6.166	<0.0001
25-kg PICS bag—3 hand warmers	−4.734	0.5419	8.736	<0.0001
Time (log_hour)	−8.234	0.2865	28.75	<0.0001
Intercept	17.06	0.3474	49.12	<0.0001
50-kg PICS bag—2 hand warmers	0	0	.
50-kg PICS bag—4 hand warmers	−2.596	0.4201	6.18	<0.0001
50-kg PICS bag—6 hand warmers	−4.144	0.4201	9.864	<0.0001
Time (log_hour)	−8.068	0.2221	36.33	<0.0001

* Treatments consisted of 1, 2, or 3 10-h hand warmers in 25-kg PICS bags or 2, 4, or 6 10-h hand warmers in 50-kg bags for over 48 h.

presents the multiple regression analysis of the effects of hand warmer number, PICS bag size, and time (log_hour) on oxygen concentration over 48 h. The data include all treatment replications (36 total). Increasing the number of hand warmers in each PICS bag size reduced the oxygen concentration, as indicated by the negative slope estimates (Table 2). The multiple regression models for both 25-kg PICS bags (F_{(3, 116)} = 302.3, p < 0.0001) and the 50-kg PICS bags (F_{(3, 116)} = 473.0, p < 0.0001) were highly significant. The R² values for these models were 0.8866 for the 25-kg bags and 0.9244 for the 50-kg bags, indicating a strong fit.

3.2. Temperature and Relative Humidity

No difference was observed in the temperature or relative humidity between the data loggers at the top and bottom of the grain across all treatments, indicating that no temperature or humidity gradients existed. Data were combined for further analysis. The combined data analysis revealed significant differences among treatments for temperature (F_{(7, 40)} = 54.19, p < 0.001) and relative humidity (F_{(7, 40)} = 47.07, p < 0.001). Temperatures in 25-kg PICS bags were significantly higher than in the 50-kg PICS bag. Throughout the experiment, temperatures for the 25-kg PICS bags ranged from 20.09 ± 0.0 °C to 20.53 ± 0.0 °C in the room and from 21.41 ± 0.25 °C to 22.97 ± 0.65 °C across treatments. In comparison, for the 50-kg PICS bags, room temperatures ranged from 19.68 ± 0.0 °C to 20.57 ± 0.0°C, while treatment temperatures ranged from 19.04 ± 0.11 °C to 20.44 ± 0.19 °C.

Relative humidity was significantly higher in 50-kg PICS bags (Figure 3). The average relative humidity in 25-kg PICS bags remained stable over 240 h, measuring 40.4 ± 0.1% for 1 hand warmer, 42.3 ± 0.4% for 2 hand warmers, and 42.7 ± 0.3% for 3 hand warmers, with no significant differences between them (Figure 3). Similarly, in 50-kg PICS bags, relative humidity remained stable at 47.1 ± 0.2% for 2 hand warmers, 50.5 ± 0.4% for 4 hand warmers, and 48.3 ± 0.2% for 6 hand warmers, also showing no significant differences between them (Figure 3).

3.3. Seed Moisture Content and Seed Germination

Moisture content and germination rates showed no significant differences between samples taken from the top, middle, and bottom of the storage bags. Data were pooled together for further analysis. The moisture content of maize seeds stored in PICS bags with hand warmers was influenced by both the storage period (F_{(1, 132)} = 10.22, p < 0.0001) and the number of hand warmers (F_{(5, 132)} = 40.21, p < 0.0001). The interaction between these two factors was significant (F_{(5, 132)} = 6.01, p < 0.0001). Significant differences in seed moisture content between initial and post-experiment samples were only observed when using 1 hand warmer in 25-kg PICS bags and 2 hand warmers in 50-kg PICS bags (

Table 3. Average (± SEM, n = 4) maize moisture content and seed germination initially and after 240 h of storage.

	Moisture Content (%)		Seed Germination (%)
Treatment *	Initial	240 h	Initial	240 h
25-kg PIC bag—1 hand warmer	9.35 ± 0.00 aA **	9.12 ± 0.09 cB	100.00 ± 0.00 aA	99.00 ± 2.49 aA
25-kg PICS bag—2 hand warmers	9.35 ± 0.00 aA	9.27 ± 0.03 bcA	97.00 ± 3.83 aA	97.67 ± 3.17 aA
25-kg PICS bag—3 hand warmers	9.35 ± 0.00 aA	9.41 ± 0.06 bA	98.00 ± 4.00 aA	97.33 ± 3.11 aA
50-kg PICS bag—2 hand warmers	9.63 ± 0.00 bA	9.41 ± 0.04 bB	100.00 ± 0.00 aA	98.67 ± 2.61 aA
50-kg PICS bag—4 hand warmers	9.63 ± 0.00 bA	9.71 ± 0.04 aA	97.00 ± 3.83 aA	97.33 ± 2.61 aA
50-kg PICS bag—6 hand warmers	9.63 ± 0.00 bA	9.59 ± 0.04 aA	97.00 ± 3.83 aA	96.00 ± 2.95 aA

* Treatments were 1, 2, or 3 10-h hand warmers in 25-kg PICS bags or 2, 4, or 6 10-h hand warmers in 50-kg bags for over 240 h. ** All data are means ± standard error of means (SEM). Under the same variable, within the same time or among treatments (lowercase) or within each treatment or across times (uppercase), means followed by the same letter are not significantly different (p < 0.05).

).

Initially, maize stored in 50-kg PICS bags had a higher moisture content. After storage, the highest moisture levels were found in maize stored in 50-kg PICS bags with 4 or 6 hand warmers. Seed germination remained unchanged from the initial measurement to 240 h of storage (F_{(1, 18)} = 0.2889, p = 0.597) and did not vary significantly across treatments (F_{(5, 18)} = 2.78, p = 0.05). Additionally, there was no interaction between storage time and treatments (F_{(5, 18)} = 0.1263, p = 0.99). No variations were observed among treatments at either the start or the end of the experiment.

4. Discussion

Oxygen concentration rapidly decreased with an increased number of hand warmers. This is a significant improvement compared to relying solely on insect metabolism, which can take several days or even weeks, depending on initial infestation levels and environmental conditions [12,13,32,33]. The pest suppression threshold of 5% oxygen or lower was reached within 12 h using 2 or 3 hand warmers in 25-kg PICS bags and 4 or 6 hand warmers in 50-kg PICS bags. Our findings align with previous research showing that more oxygen scavengers, such as hand warmers or seeds, accelerate oxygen consumption [14,22,34]. After plateauing, oxygen levels began to rebound, a phenomenon observed in previous studies [22,35,36,37]. To address this challenge, using sufficient hand warmers is essential for sustaining low oxygen levels.

Doubling the number of hand warmers and the size of PICS bags yielded comparable oxygen depletion trends. This indicates that hand warmers can also be doubled when the bag size increases from 25 to 50 kg to achieve similar results. The multiple regression model further validates these findings. Both bag sizes benefit from increased hand warmers, though diminishing returns are observed beyond 2 hand warmers for 25-kg PICS bags and 4 for 50-kg PICS bags. The high R² values indicate strong model predictability, reinforcing our hypothesis about the feasibility of scaling hand warmers in larger storage containers. However, further research is needed to assess their effectiveness beyond 50-kg PICS bags, e.g., for 100-kg PICS bags, which are more cost-effective for smallholder farmers due to their lower cost per unit of stored volume [23].

Environmental conditions, such as temperature and relative humidity, are crucial for preserving grain quality. The slightly higher temperature (~2 °C) observed in 25-kg PICS bags compared to 50-kg PICS bags was due to the timing of the experiment rather than the treatments. Though hand warmers are designed to generate heat, they did not noticeably increase the temperature inside PICS bags. This corroborates findings from other studies [21,22]. Relative humidity is a major contributor to grain and seed storage losses [38,39,40]. Although relative humidity was significantly higher in 50-kg PICS bags than in 25-kg PICS bags, it remained stable, with no significant difference among hand warmers within each bag size. These results align with previous research showing minimal changes in relative humidity in hermetic storage [21,22]. Further research is needed to clarify the impact of doubling hand warmers alongside hermetic bag size on relative humidity, as no clear pattern emerged.

Despite variations in some treatments, moisture content remained consistent between the initial and final samples, suggesting that increasing the number of hand warmers had minimal impact on grain quality. In addition, seed germination remained unchanged, despite variations in relative humidity and moisture content, further confirming that hand warmers did not affect seed quality during 240 h of hermetic storage. These findings are consistent with previous research showing that oxygen scavengers, such as germinating seeds and hand warmers, have minimal impact on moisture content and seed germination [14,21,22].

Our findings indicate that doubling both hand warmers and PICS bag size maintained similar oxygen depletion trends, indicating scalability from 25- to 50-kg PICS bags. Using 2 or 3 hand warmers in 25-kg PICS bags or 4 or 6 hand warmers in 50-kg bags lowered oxygen levels below 5% within 12 h and maintained them for at least 8 days. Previous studies have shown that maintaining oxygen concentrations below 5% for 3 to 5 days significantly increases insect mortality, inhibits female oviposition, and suppresses progeny development [8,11,14,35,41,42]. Research is needed to assess the use of hand warmers in different types of grains and pests, as residual oxygen levels vary with seed size, and insect behavior differs accordingly. Further, studies should evaluate the effects of hand warmers on long-term storage (e.g., 3 months or more) of grain or seed quality. In addition, research should explore humidity-regulating materials for stabilizing moisture levels and preserving grain quality during storage.

5. Conclusions

Our study shows that hand warmers can be scaled to larger hermetic storage containers. Doubling the number of hand warmers with an increase in bag size from 25 to 50 kg yielded comparable results. However, diminishing returns were observed beyond specific thresholds. Using 2 or 3 hand warmers in 25-kg PICS bags and 4 or 6 hand warmers in 50-kg PICS bags reduced oxygen concentrations below the 5% threshold for pest suppression within 12 h and maintained this for several days. The findings also confirm the minimal impact on seed moisture content, relative humidity, and germination rates, highlighting their non-disruptive effect on grain quality. Future research should explore scaling hand warmers in 100-kg PICS bags, their effectiveness across grain types and pests, their long-term impact on grain quality, and humidity-regulating materials in storage.

The findings of this research present a valuable opportunity to address storage challenges, including providing alternatives to chemical treatments. Further, they offer a solution for preserving high-nutrient crops, such as provitamin A maize, which is sensitive to prolonged exposure to oxygen. This is crucial for institutions and initiatives aimed at reducing malnutrition among children, particularly in developing countries. Using hand warmers in hermetic storage could help reduce stored product losses, enhancing both food and nutrition security while promoting sustainable resource management.