The water–iodine oxide system: a revised mechanism for hydration and dehydration

Abstract

Iodic acids are widely studied in atmospheric and biological applications but their inherent hydrophilic properties introduce complexities that affect their functionality and reactivity. We have shown that iodic acid (HIO₃) dehydrates directly into iodine pentoxide (I₂O₅) in contradiction to the generally accepted multi-step dehydration mechanism where HIO₃ dehydrates into HI₃O₈ first, then dehydrates into I₂O₅. The generally accepted mechanism is used to determine the concentration of iodic acid by TGA and is only valid for special conditions. The revised mechanism allows for the determination of concentrations of iodic acids under all conditions, and the more specific conditions where the accepted mechanism is valid are shown. The determination of concentration of iodic acid with the revised dehydration mechanism is dependent on assumptions of residual water and initial concentration of HI₃O₈. The validity of these assumptions is established by studying the absorption and hydration behavior of I₂O₅ from atmospheric water. These results will have an impact on the handling and use of iodine.

---

Introduction

Iodine(V) oxides have biocidal potential in neutralizing bacterial agents released in certain weapon systems. Handling iodine(V) oxides is one of the biggest challenges toward their effective utilization in this application. Specifically, iodine(V) oxides are hygroscopic to the point of being deliquescent. A pile of di-iodine pentoxide (I₂O₅) powder, for example, will absorb so much water from the surrounding atmosphere over time, that a puddle is produced. This property has limited iodine(V) oxides usefulness in energetic applications because an abundance of water can greatly consume liberated energy from a chemical reaction between iodine(V) oxide and a metal fuel, such as aluminum, thereby inhibiting combustion performance. Understanding, absorption, hydration, desorption and dehydration mechanisms associated with iodine(V) oxides may enable more tailored synthesis and control over this property for engineering future prompt-agent-defeat technologies (or bio-agent defeat systems).

Iodine(V) oxide water absorption behavior is studied extensively with inconsistent results that may stem from a limited understanding of the fundamental kinetics and dynamics of water absorption and reaction with iodine(V) oxides. The inconsistencies highlighted here involve iodine(V) oxides reaction with water and include: different humidity levels where absorption begins; variations in concentration of iodic acids (HIO₃, HI₃O₈); crystal structure; onset temperature for dehydration and decomposition; and, absorption rates. All of the inconsistencies discussed below are affected by the hydration and dehydration mechanism.

The accepted mechanism for hydration of I₂O₅ into iodic acids (i.e., HIO₃ and HI₃O₈) and dehydration of iodic acids into I₂O₅ (i.e., HIO₃ and HI₃O₈) was first proposed by Selte et al.¹ and shown in eqn (1a) and (1b), respectively.¹

Hydration:

3I₂O₅ + 3H₂O → 2HI₃O₈ + 2H₂O → 6HIO₃ (1a)

Dehydration:

6HIO₃ → 2HI₃O₈ + 2H₂O → 3I₂O₅ + 3H₂O (1b)

In dehydration, six moles of iodic acid in the form of HIO₃ form three moles of I₂O₅ and three moles of water with an intermediate step involving the formation of two moles of HI₃O₈. In this reaction, water comprises 5.11 wt% of the total mass. The mass loss from dehydration of water at each step is shown in eqn (2).

3HIO₃ → HI₃O₈ + H₂O: 3.41% water (2a)

2HI₃O₈ → 3I₂O₅ + H₂O: 1.76% water (2b)

Using mass loss from water calculated from eqn (2a) and (2b), concentrations of iodic acids can be determined from thermal equilibrium analysis using differential scanning calorimetry (DSC) and thermal gravimetric analysis (TGA).^1,2 The equations used to calculate concentration of HIO₃ (X_HIO3) and HI₃O₈ (X_HI3O8) are shown in eqn (3a) and (3b), respectively.

In eqn (3), ML% is measured mass loss from TGA. The mass loss percent is divided by the calculated maximum mass loss from eqn (2a) and (2b) for HIO₃ and HI₃O₈ respectively, to obtain the concentration of each iodic acid. The DSC/TGA data shows an endotherm and mass loss at 110 °C, where HIO₃ dehydrates into HI₃O₈, followed by an endotherm and mass loss at 210 °C, where HI₃O₈ dehydrates into I₂O₅. Dehydration of HI₃O₈ has always been observed when HIO₃ dehydration was observed, indicating an intermediate step where HIO₃ dehydrates into HI₃O₈.^1,2 In Selte et al.¹ and multiple references within, determination of iodic acid concentration by TGA is confirmed with XRD analysis. The temperatures where the onset of iodic acid dehydration begins is widely varied and often overlooked.

A review of the publications over the thermochemical properties of iodine oxides³ summarized data available for reported onset temperatures of each dehydration step and decomposition of I₂O₅. The onset temperatures reported for I₂O₅ decomposition are between 300 and 450 °C.³ Selte et al. report a lower onset temperature of 280 °C for amorphous I₂O₅, with onset temperatures between 300 and 420 °C for crystalline I₂O₅.¹ Iodic acids such as HI₃O₈ onset temperatures are reported between 195 and 220 °C (ref. 1) and between 100 and 130 °C for HIO₃.¹ The reason for the variation in temperature is not known. Formation of hydrated species of iodine oxides is a complex, dynamic process that can depend on variables such as relative humidity (RH), particle size, hydration and desorption rates, available water, and crystal structure. These variables can be dependent on each other and could affect properties of the final material including dehydration temperature.

Relative humidity (RH) has a significant influence on the oxidation state and kinetics of iodine(V) oxides. Selte et al.¹ report that crystalline HI₃O₈ powder can be stored in open vessels for an indefinite period independent of atmospheric humidity but fine-powdered samples completely convert to HIO₃ in moist air. Kumar et al.⁴ studied the deliquescence and efflorescence behavior of I₂O₅ and HIO₃. They describe deliquescence as the transformation from a solid crystal phase to an aqueous solution and efflorescence as the reverse, when an aqueous solution crystallizes to a solid phase; and, define these differences by changes in the optical properties of the sample. Deliquescence and efflorescence are seen at specific threshold RH levels. Also, I₂O₅ and HIO₃ crystals with an average size of 10–50 μm were studied at different temperatures. Deliquescence was reported at 85% RH and room temperature for HIO₃, and at 80% RH and room temperature for I₂O₅.⁴ Efflorescence was not seen in iodine(V) oxide samples with RH as low as 1.0%. Little et al.² observed hydration according to the steps in the hydration mechanism in eqn (1a). They explain different hydrated species of iodine(V) oxide can precipitate out when exposed to water and recrystallize depending on impurities, rate of evaporation, and crystal structure.² This finding is in contrast to Selte et al.,¹ where all samples were exclusively HIO₃. Little et al.² note that multiple iterations of this method produced varying concentrations of iodic acid. Also, a threshold value of 60% RH was found by Little et al.² to hydrate commercially available samples without dissolving the solid in water and recrystallization. When commercially available samples are exposed to air above 60% RH threshold, the product is HIO₃ and when exposed to air below 60% RH, no hydration is observed. Because the commercial sample was entirely HI₃O₈, a threshold of 60% RH was assumed to be the RH that HI₃O₈ started absorbing water.² To obtain I₂O₅, a sample initially HI₃O₈ was ball milled, and produced a sample 75 wt% I₂O₅ and 25 wt% HI₃O₈. For hydration from I₂O₅ to HI₃O₈, Little et al.² exposed the ball milled samples to 40% RH over a period of several days. The TG analysis indicated increased mass loss at 210 °C such that samples with higher concentrations of I₂O₅ were hydrated at 40% RH. It is interesting to note that the observation of an endotherm with an onset temperature of 170 °C remained constant with increased hydration. An endotherm with an onset temperature of 170 °C is outside of the range of generally accepted dehydration temperatures for hydrated species of iodine oxides. Over time with exposure to 40% RH, the endothermic peak indicating HI₃O₈ increased and formed a double peak. Commercially available I₂O₅ was used to study the hydration step from HI₃O₈ to HIO₃ (eqn (1a)). No hydration was seen below 50% RH over a period of weeks, however, hydration was seen when RH increased to 70% RH. Hydration rates were measured at 70% RH for HI₃O₈ to HIO₃ (eqn (1a)) and at 40% RH for I₂O₅ to HI₃O₈ (eqn (1a)). At 40% RH, the hydration rate went to 0 before 80 wt% of the sample was HI₃O₈, but at 70% RH hydration continued until the sample was solely HIO₃. In summary, a discrepancy from previous research is that water absorption thresholds for any iodic acid widely range from 0 to 85% RH. The reason for the dramatic variation may be due to limited understanding of the fundamental mechanism of hydration and dehydration. A better understanding of the fundamental hydration and dehydration kinetics is needed to reconcile these discrepancies.

Our goal is to improve the fundamental mechanism of hydration and dehydration of I₂O₅ by studying hydration and dehydration from atmospheric water (i.e. relative humidity). The objectives are to understand the fundamental absorption mechanism of I₂O₅; understand how I₂O₅ reacts with water to form iodic acids; and, identify how the reactions between I₂O₅ and water affect the kinetics of hydration and dehydration. It is noted that hydration and absorption are not interchangeable terms. For the purpose of this study, absorption refers to water that is weakly bonded and not yet reacted to hydrate I₂O₅ into HIO₃. Hydration refers to absorbed water that has reacted to form an IO₃⁻ solution and will form an HIO₃ crystal structure upon desorption. Desorption is when mass loss occurs without heating and dehydration is mass loss as a result of HIO₃ forming I₂O₅ from heating.

Experimental

I₂O₅

Two synthesis procedures are used to produce I₂O₅ powders. One method is based on the thermal dehydration of iodic acid to form I₂O₅.^5,6 These samples were supplied by Sigma Aldrich (St. Louis, MO) and come as crystals that require the sample to be crushed using a mortar and pestle. This sample is referred to as 17 μm I₂O₅ because it has an average particle diameter of 16.9 μm.⁷ The 17 μm I₂O₅ is the same as commercial I₂O₅ used in Smith et al.⁷ where characterization of these samples was performed. The second method is a “dry” method that combines elemental iodine and oxygen using a non-thermal plasma energy source to form spherical I₂O₅ particles with an average diameter of 4.7 μm.⁷ These samples were labeled nano C I₂O₅ when used in Smith et al.⁷ because they were initially measured erroneously to have an average particle diameter of 400 nm.⁷ This sample is now referred to as 5 μm I₂O₅ here. The two distinct synthesis methods enable production of particles with discretely different properties inherent from the synthesis methods and tailored for experimental designs that focus on deriving a more general hydration/dehydration mechanism.

Humidity chamber

A humidity chamber was designed and built to control relative humidity, RH. The chamber was designed from a 60 × 30 × 30 cm acrylic air tight box. A 1.27 cm hole was drilled in the wall of the chamber and a variable speed fan was placed outside the chamber so the current of air from the fan passed over a hole to draw air out of the inside the chamber. A heating pad was placed inside the chamber and a Pyrex™ pan filled with water positioned on top of the heating pad. Humidity was controlled by adjusting the temperature on the heating pad and the speed of the air from the fan. Relative humidity was measured with an AMPROBE TH-3 relative humidity temperature meter and was held above 80% but below the dew point at a temperature of 23 °C so drops of water did not condense inside the chamber. The 80% RH threshold was selected because it is the highest RH without having droplets form and thus a reasonable upper bound for this study.

Thermal equilibrium analysis

Simultaneous thermal analysis (STA) was performed using a Netzsch model STA 449 differential scanning calorimeter (DSC) with thermo-gravimetric (TG) analyzer. All powder samples measured approximately 10 mg and were placed inside alumina crucibles designed for the STA 449 DSC-TGA. To ensure the samples were initially comprised of only I₂O₅ and not another form of iodic acid, all samples were heated in a Neytech QEX oven to 250 °C, above the dehydration temperature of HIO₃ (110 °C) and HI₃O₈ (210 °C) but below the dissociation temperature of I₂O₅ (350 °C), and held for 5 minutes. The heat flow and mass loss showing that samples are pure I₂O₅ after heat treatment are shown in ESI Fig. S1.† The sample covers were removed and the samples were placed inside the humidity controlled chamber for varying intervals. The samples were weighed before and after they were placed inside the humidity controlled chamber and the mass change is reported as a percentage in Tables 1 and 2. It is important to note that the starting material was purposefully heated such that the material is initially solely I₂O₅ as shown in Fig. S1.† Most other studies reporting iodine(V) compounds may not start with pure I₂O₅ as the precursor because a heat treatment procedure as described above is not mentioned.^2,4,8,9 Each sample was covered and loaded into the STA manually to minimize exposure time to the ambient held at an average 20% RH. Samples were heated to 600 °C at 10 °C min⁻¹ in an argon atmosphere. The argon atmosphere was created by a constant flow of argon into the STA at 100 ml min⁻¹. The control software for the STA requires a 1 min equilibration period between sample loading and heating that ensures the atmosphere inside the STA is argon. This equilibration period is included in the time interval between when the samples were removed from 80% RH and measurements.

Table 1 Mass gain from absorption and hydration at 80% RH (23 °C) and mass loss from TG analysis for 17 μm I₂O₅. Mass gain is measured by weighing samples before and after exposure to 80% RH and shows absorption. Mass loss is obtained by TG measurements in the iodic acid dehydration temperature range and indicates hydration. The ratio HIO₃/HI₃O₈ indicates percent mass loss by HIO₃ to percent mass loss by HI₃O₈

Exposure time (min)	Mass gain, %	HIO3 mass loss, %	HI3O8 mass loss, %	Total mass loss	HIO3/HI3O8
1	1.0%	0.0%	0.02%	0.02%	0.00
5	1.5%	0.0%	0.08%	0.08%	0.00
10	3.2%	0.0%	0.06%	0.06%	0.00
15	9.0%	1.63%	0.23%	1.86%	7.09
30	13.7%	2.70%	0.10%	2.80%	27.00
180	34.8%	5.36%	0.22%	5.58%	24.36
900	32.7%	7.41%	0.17%	7.58%	43.59
2680	23.1%	8.49%	0.29%	8.78%	29.28

---

Table 2 Mass gain from absorption and hydration at 80% RH (23 °C) and mass loss from TG analysis for 5 μm I₂O₅. Mass gain is by weighing samples before and after exposure to 80% RH and shows absorption. Mass loss is obtained by TG measurements in the iodic acid dehydration temperature range and indicates hydration. The ratio HIO₃/HI₃O₈ indicates percent mass loss by HIO₃ to percent mass loss by HI₃O₈

Exposure time (min)	Mass gain, %	HIO3 mass loss, %	HI3O8 mass loss, %	Total mass loss	HIO3/HI3O8
1	0.7%	1.12%	0.42%	1.54%	2.67
5	4.7%	2.01%	0.57%	2.58%	3.53
10	6.9%	4.63%	0.29%	4.92%	15.97
30	9.1%	4.74%	0.32%	5.06%	14.81
85	6.1%	4.55%	0.50%	5.05%	9.10
220	4.7%	4.81%	0.49%	5.30%	9.82
2880	4.8%	4.87%	0.42%	5.29%	11.60

---

Results

Fig. 1a and b are the heat flow from DSC and mass change from TG measurements as a function of exposure time to 80% RH from 1 to 3060 minutes for 17 μm I₂O₅, respectively. Fig. 1a shows endothermic reactions starting at an exposure time of 10 minutes. At an exposure time of 15 minutes, there are two distinct endotherms below 160 °C. The second endotherm, with onset at 110 °C, increases in magnitude with time while the first endotherm reduces in magnitude until 900 minutes exposure time where it is no longer detectable. To the authors' knowledge, two distinct peaks in the range of HIO₃ dehydration have not been reported previously. However, similar endothermic activity reported in Smith et al.¹⁰ observed a single endotherm with a double peak before 110 °C. They proposed the multiple endothermic activity in this temperature range was latent water or IO₃⁻ solution formed when iodine oxides react with water, and that theory is supported by the results shown here. The first endotherm in samples that show two endotherms is assumed to be evaporation of latent water and only mass loss from the second endotherm is used for calculation of mass loss from dehydration of HIO₃. Additionally, most samples showed slight endothermic activity at 210 °C with corresponding mass loss (Fig. 1b). The beginning and ending points for mass loss in Fig. 1b are taken from endotherm onset and ending temperatures in Fig. 1a. All mass loss for HIO₃ and HI₃O₈ dehydration corresponding to the endotherms in Fig. 1a are shown in Fig. 1b and tabulated along with the mass increase from absorption and mass loss from hydration when exposed to 80% RH in Table 1.

Fig. 1 (a) DSC measurements of heat flow for 17 μm I₂O₅ at different exposure time intervals to 80% RH as indicated on the curves; and, (b) TG measurements of mass change for 17 μm I₂O₅ at different exposure time intervals to 80% RH as indicated on the curves. Onset and end temperature for measurements of mass loss are determined by endotherm onset and end temperatures.

Fig. 2a and b are the heat flow from DSC and mass change from TG measurements as a function of exposure time to 80% RH from 1 to 2880 minutes for 5 μm I₂O₅. Fig. 2a shows two distinct endotherms with onset temperature below 110 °C at an exposure time of 1 minute. With an exposure time of 5 minutes, there is only one distinct endotherm with an onset temperature of 90 °C. These endotherms increase in magnitude and onset temperature through the entire range tested. Fig. 2a shows most 5 μm I₂O₅ samples produce an endotherm at 210 °C with an associated mass loss (Fig. 2b). The beginning and ending temperatures for mass loss in Fig. 2b are taken from endotherm onset and end temperatures in Fig. 2a. Table 2 summarizes the data in Fig. 2b along with the mass increase from absorption.

Fig. 2 (a) DSC measurements of heat flow for 5 μm I₂O₅ at different exposure time intervals to 80% RH as indicated on the curves; and, (b) TG measurements of mass change for 5 μm I₂O₅ after different exposure time intervals to 80% RH as indicated on the curves. Onset and end temperature for measurements of mass loss are determined by endotherm onset and end temperatures.

Fig. 3a and b show the mass gain for 17 μm I₂O₅ and 5 μm I₂O₅ respectively, determined by weighing the samples before and after exposure to 80% RH at 23 °C for the indicated exposure times. The mass gain for 17 μm I₂O₅ increases to a maximum of 34.8% at an exposure time of 180 minutes and then decreases to 7.0% at an exposure time of 3060 minutes. The mass gain for 5 μm I₂O₅ increases to a maximum of 9.1% at an exposure time of 30 minutes then decreases to 4.8% at an exposure time of 2880 minutes. An opposite trend is seen in the mass gain rate in Fig. 3a and b after the mass gain rate starts to decrease, the decrease in the 5 μm I₂O₅ mass gain after an exposure time of 30 min is initially fast and slows with increased exposure time, while 17 μm I₂O₅ is initially slow and increases with exposure time.

Fig. 3 All graphs provide information on the relative rates of absorption, hydration or dehydration as a function of particle size. Samples shown are exposed to 80% RH (23 °C) throughout the entire duration. (a) Mass gain percentage as a function of time for 17 μm I₂O₅. (b) Mass gain percentage as a function of time for 5 μm I₂O₅ up to 220 minutes (i.e., 2880 minute data excluded from graph for clarity of presentation but all data shown in Table 2 and Fig S2†). (c) Mass gain for both 5 μm I₂O₅ (dashed line) and 17 μm I₂O₅ (solid line) for the first 30 minutes of exposure to 80% RH.

Fig. 3c shows the mass gain for both 5 μm I₂O₅ and 17 μm I₂O₅ for exposure time periods less than 30 minutes. The mass gain rate for the 5 μm I₂O₅ is linear until an exposure time of 5 minutes, then starts plateauing at an exposure time of 5 minutes. In contrast, the mass gain for 17 μm I₂O₅ is negligible until an exposure time of 5 minutes, then starts to increase.

Discussion

Mechanism

The mechanism in eqn (1a) and (1b) assumes that one mole of water and two moles of I₂O₅ initially react to form HI₃O₈, and then two moles of water and two moles of HI₃O₈ react to form HIO₃. This intermediate step (i.e., formation of HI₃O₈ before HIO₃) in the formation of HIO₃ from I₂O₅ is assumed because in almost all reported cases, there is an endothermic peak at 210 °C, the dehydration temperature of HI₃O₈, and an associated mass loss that corresponds to the mechanism in eqn (1a) and (1b).^1,2 Eqn (1) and TG mass loss (Tables 1 and 2 along with Fig. 1b and 2b) can be used to calculate the initial concentration of HIO₃. Eqn (2a) and (2b) show the dehydration steps of both iodic acids along with the mass percent of water in each step. If the mechanism accepted and described in eqn (1a) and (1b) is correct, the mass loss in the HIO₃ dehydration temperature range cannot be greater than 3.41% if there is no residual water. Mass loss from HIO₃ greater than 3.41% will give a concentration of HIO₃ greater than 100% when calculated using eqn (3a).

The mechanism in eqn (1) also has a condition that HIO₃ dehydrates directly into HI₃O₈. This condition is represented by the ratio of mass loss of HIO₃ to HI₃O₈. The maximum ratio of mass loss percentage at 110 °C for HIO₃ to mass loss percentage at 210 °C for HI₃O₈ should be 3.41 : 1.76 (i.e., a ratio of 1.94). A ratio higher than 1.94 indicates that a portion of HIO₃ is dehydrating directly into I₂O₅ (not HI₃O₈) and is seen for multiple samples (highlighted in Tables 1 and 2). These results indicate a new mechanism is needed and presented in eqn (4a) and (4b) for hydration and dehydration of I₂O₅ that does not have an intermediate step resulting in the formation of HI₃O₈.

Hydration:

I₂O₅ + H₂O → 2HIO₃ (4a)

Dehydration:

2HIO₃ → I₂O₅ + H₂O (4b)

The mechanism in eqn (4a) and (4b) is proposed because the accepted mechanism (eqn (1)) is valid only under the condition that all HIO₃ forms HI₃O₈ (eqn (2a)) during heating. The initial concentrations of HIO₃ and I₂O₅ can be calculated from TG analysis of mass loss as tabulated in Tables 1 and 2 and eqn (4). The equations to calculate initial concentration of HIO₃ and I₂O₅ from eqn (4) are shown in eqn (5a) and (5b).

X_I2O5 = 1 − X_HIO3 (5b)

In eqn (5), X is the concentration of HIO₃ or I₂O₅ as indicated, ML%_total is the measured mass loss for the entire iodic acid dehydration temperature range, and X_{HI3O8,initial} is the initial concentration of HI₃O₈. The total measured mass loss is divided by the calculated mass loss percent of total water (5.11%), not including residual water, calculated from eqn (4b).

Eqn (1)–(3) represent a model for the previously accepted hydration/dehydration mechanism that enables determination of the concentration of iodine species by TGA mass change measurements. A new model is presented in eqn (4)–(7). The major difference between the models shown is seen by comparing eqn (5) to (3) and is that HIO₃ dehydrates into I₂O₅ instead of HI₃O₈, and that HI₃O₈ is formed during heating. The amount of HI₃O₈ that is formed during heating is accounted for by using the total mass loss in the iodic acid (HIO₃ and HI₃O₈) dehydration temperature range instead of the individual mass loss from HIO₃ and HI₃O₈ dehydration respectively (e.g., used in eqn (3a) and (3b)). The model in eqn (5) includes a term for initial concentration of HI₃O₈. The DSC cannot differentiate between initial HI₃O₈ and HI₃O₈ that is formed during heating. To accurately determine the concentration of HIO₃ and I₂O₅ using eqn (5), the initial concentration of HI₃O₈ is needed. In this study, the initial concentration of HI₃O₈ is assumed to be 0.0% because the samples are initially dehydrated before hydration in 80% RH. The validity of this assumption is discussed below. The amount of HIO₃ that dehydrates into I₂O₅ (eqn (6b)) and the amount of HIO₃ that forms HI₃O₈ during heating (eqn (6a)) is shown in eqn (6).

X_HIO3→I2O5 = X_HIO3 − X_HI3O8 (6b)

In eqn (6), ML%_HI3O8 is the measured mass loss from TG analysis for HI₃O₈, X_HIO3→I2O5 is the concentration of HIO₃ that dehydrates directly into I₂O₅, X_HI3O8 is the concentration of HI₃O₈ used in eqn (3b) and is also the concentration of HIO₃ that forms HI₃O₈ during heating (X_HIO3→HI3O8) using eqn (4) assuming the initial concentration of HI₃O₈ is 0%. The mechanism in eqn (4) is proposed because the accepted mechanism (eqn (1)) is a special condition but eqn (7) shows the conditions where the mechanism in eqn (4) is equivalent to the mechanism in eqn (1).

X_HIO3→HI3O8 = X_HIO3 (7a)

X_HIO3→I2O5 = 0 (7b)

The concentrations of HIO₃ and HI₃O₈ calculated using the mechanism in eqn (1) and (4) are shown in ESI Tables S1 and S2.† The samples for which eqn (1) is not valid, because the concentration of HIO₃ is greater than 100% (X_HIO3 > 100%), are shaded in ESI Tables S1 and S2.† The mechanism in eqn (4a) and (4b) may not have been observed by others because the methods previously used to study hydration of I₂O₅ are similar and thus produce similar results^1,4,8,11,12 that follow eqn (1) and (7) where HIO₃ dehydrates directly into HI₃O₈. The variables that determine whether the conditions in eqn (7) are met are not known. Multiple different experimental conditions (DSC heating rate, water impurities, hydration method (e.g., atmospheric vs. dissolution in water), particle size, I₂O₅ synthesis method (e.g., dry or dehydration of HIO₃)) have been tested in an effort to determine the condition when eqn (1) is valid; however, a direct relation has not been found.

In previous studies,^1,2 concentration of iodic acid has been verified by XRD analysis. The mechanism in eqn (4a) and (4b) does not affect the validity of these results and the XRD analysis from previous studies show that TGA is an accurate technique for determination of iodic acids. The XRD analysis happens over 1–2 hours for accurate measurements and RH cannot be controlled during measurements. For this study, the samples exposure to 20% RH was limited to less than 5 minutes. The limitation on exposure time in 20% RH and the time needed for XRD analysis did not allow XRD data collection.

Assumptions

The model described in eqn (4)–(7) is based on assumptions that may not have been considered previously. We have shown that the accepted mechanism for dehydration of iodic acids is a special condition that is violated when the ratio of mass loss from HIO₃ to HI₃O₈ is greater than 1.9 or when the mass loss of water from HIO₃ is greater than 3.14%. These are general conditions and are only valid when the only mass loss is water from dehydration of HIO₃ (not residual water). When I₂O₅ is hydrated by exposure to 80% RH, this assumption does not always hold and is seen by a total mass loss greater than 5.11%. Samples that have a greater mass loss then 5.11% are assumed to have residual water and are underlined in ESI Tables S1 and S2.† For accurate calculation of initial HIO₃ and I₂O₅ concentration, it is assumed that if total mass loss is greater than 5.11%, the sample is initially pure HIO₃. The validity of the assumption that residual water in the sample indicates the sample is pure HIO₃ is discussed below. The calculated concentrations for each sample using eqn (3) and the assumption that the concentration of HIO₃ is 100% for samples with a greater total mass loss than 5.11% is shown in Tables 3 and 4.

Table 3 Composition of 17 μm I₂O₅ calculated using eqn (5) and (6). Values are calculated assuming X_{HI3O8,initial} = 0 and that mass loss over 5.11% indicates the sample is 100% HIO₃

Exposure time (min)	Residual water (g)	XHIO3	XHI3O8	XI2O5	XHIO3→I2O5
1	0	0.4%	1.1%	99.6%	−0.7%
5	0	1.6%	4.5%	98.4%	−3.0%
10	0	1.2%	3.4%	98.8%	−2.2%
15	0	36.4%	13.0%	63.6%	23.3%
30	0	54.8%	5.7%	45.2%	49.1%
180	0.47	100.0%	12.5%	0.0%	87.5%
900	0.24	100.0%	9.6%	0.0%	90.4%
2680	0.36	100.0%	16.4%	0.0%	83.6%

---

Table 4 Composition of 5 μm I₂O₅ calculated using eqn (5) and (6). Values are calculated assuming X_{HI3O8,initial} = 0 and that mass loss over 5.11% indicates the sample is 100% HIO₃

Exposure time (min)	Residual water (g)	XHIO3	XHI3O8	XI2O5	XHIO3→I2O5
1	0	30.1%	23.8%	69.9%	6.3%
5	0	50.5%	32.3%	49.5%	18.2%
10	0	96.2%	16.4%	3.8%	79.8%
30	0	99.0%	18.1%	1.0%	80.8%
85	0	98.8%	28.3%	1.2%	70.4%
220	0.019	100.0%	27.8%	0.0%	72.2%
2880	0.018	100.0%	23.8%	0.0%	76.2%

---

In Tables 3 and 4, the concentration of HIO₃ that dehydrates into I₂O₅ is negative for the 17 μm sample after an exposure time of 1, 5 and 10 minutes (underlined). The negative value for the concentration of HIO₃ that dehydrates into I₂O₅ is a result of the assumption that the initial concentration of HI₃O₈ is 0.0%. The new mechanism (i.e., eqn (4)) has an inherent assumption that during hydration, I₂O₅ hydrates directly into HIO₃. Any I₂O₅ that hydrates directly into HI₃O₈ will show up as initial HI₃O₈ in the model shown in eqn (4)–(7) because it is not possible to differentiate between HI₃O₈ that is formed during heating and HI₃O₈ that is formed during hydration. This leads to the small error (<1.6%) in HIO₃ concentration calculated using eqn (5a). Since HI₃O₈ is an adjunct of HIO₃ and I₂O₅,¹³ the small error in HIO₃ calculated using total mass in eqn (5a) is the amount of HIO₃ in HI₃O₈ from hydration. The model in eqn (4)–(7) can be improved by determining the amount of HI₃O₈ formed during hydration, but a method to determine the amount of HI₃O₈ formed during hydration has not been developed. Examining the absorption behavior of these samples can help verify the validity of the assumptions that initial concentration of HI₃O₈ is 0.0% and that mass loss over 5.11% indicates the samples in 100% HIO₃. By weighing samples before and after exposure to 80% RH, differences in absorption and hydration were measured.

HI₃O₈ formation from hydration

Tables 1 and 2 show the mass gain from absorption in 80% RH and the mass loss from HIO₃ dehydration. In both samples, the mass gained from absorption is higher than the total iodic acid mass loss from dehydration. After exposure to 80% RH in a humidity chamber, the samples were removed from the camber and exposed to 20% RH ambient environment for weighing and placement in the STA. Since only the RH of the environment is changed after the samples are removed from the humidity chamber, the difference between mass gain and mass loss is assumed to be caused by rapid desorption of residual water and/or IO₃⁻ solution. The time of exposure to low humidity was kept as constant as possible and was less than 30 seconds for weight measurements and less than 5 minutes for DSC loading. This limitation to the experimental setup inadvertently helps explain the differences between absorption and the hydration mechanism. The difference in mass gain from absorption and mass loss from dehydration of HIO₃ indicates that hydration of I₂O₅ into HIO₃ is a time dependent reaction. The time dependence of this reaction could influence the hydration mechanism and allow HI₃O₈ to form during hydration.

In Table 1 for 17 μm I₂O₅, there is no mass loss from dehydration of HIO₃ until 15 minutes exposure time, but a mass gain of 1.0%, 1.5%, and 3.2%, and a total mass loss of 0.02%, 0.08%, and 0.06% is seen at 1, 5, and 10 minutes exposure time, respectively. The results in Table 1 indicate the mass loss is due to dehydration of HI₃O₈, indicating that initially, HI₃O₈ is forming instead of HIO₃ for the 17 μm I₂O₅. Fig. 3c shows that mass gain in the first 30 minutes of exposure time for both 17 μm I₂O₅ and 5 μm I₂O₅. The 17 μm I₂O₅ initially has a fast rate of absorption with an inflection point at 10 minutes where the absorption rate decreases (Fig. 3c). After 10 minutes exposure time, mass loss in the HIO₃ dehydration temperature range is seen. The inflection point at 10 minutes and mass loss in the HIO₃ dehydration temperature range after 10 minutes exposure suggests that the increased rate of absorption in the 17 μm I₂O₅ is due to I₂O₅ forming directly into HI₃O₈, and after the rate of absorption slows, water that is absorbed starts to hydrate I₂O₅ into HIO₃. The 5 μm I₂O₅ does not have an inflection point where the rate of absorption slows and does not have exposure times where mass loss from HI₃O₈ and no mass loss from HIO₃ is seen. We speculate that the differences in absorption behavior and formation of HI₃O₈ are a result of the different particle size or processing methods. Decreased specific surface area of 17 μm I₂O₅ could slow absorption. If hydration of I₂O₅ into HIO₃ is dependent on absorption rate, the slower absorption rate of 17 μm I₂O₅ could slow the hydration rate of I₂O₅ into HIO₃. HI₃O₈ is an adjunct of HIO₃ and I₂O₅ and HIO₃ and I₂O₅ must both be present to form HI₃O₈. If the absorption rate of 17 μm I₂O₅ slows the hydration rate of I₂O₅ to the point where both I₂O₅ and HIO₃ are present, then formation of HI₃O₈ could be a result of particle size. Alternatively, the processing method between 17 μm I₂O₅ and 5 μm I₂O₅ could result in conditions where the formation of HI₃O₈ is favored over the formation of HIO₃ in 17 μm I₂O₅, and the difference in absorption is a result of formation of HI₃O₈. The 17 μm I₂O₅ were created by dehydration of commercial I₂O₅ that has been shown to be composed mostly of HI₃O₈.^2,7,10 The dehydration of HI₃O₈ during the heating process could leave discontinuities in the crystal lattice that could favor the formation of HI₃O₈ over HIO₃ during hydration. The initial formation of HI₃O₈ may not be seen in the 5 μm I₂O₅ because it is created using a “dry” method and may have more limited discontinuities in the crystal lattice formed from initial dehydration. HI₃O₈ that is formed during hydration cannot be distinguished from HI₃O₈ that is formed during heating and leads to the error shown in Tables 3 and 4. Because of this, further studies are need to determine the effects of particle size and processing methods on absorption and formation of HI₃O₈.

Residual water

The other major assumption used in the calculation of iodic acid concentration from TGA is that any mass loss over 5.11% indicates that the sample is entirely composed of HIO₃. In both 17 μm I₂O₅ and 5 μm I₂O₅, more water is absorbed than is needed to hydrate I₂O₅ into HIO₃. The maximum mass gained for 5 μm I₂O₅ is 9.1% at an exposure time of 30 minutes and the maximum mass gain for the 17 μm I₂O₅ is 34.8% at an exposure time of 180 minutes. Also, the first exposure time where the mass loss is over 5% for 5 μm I₂O₅ is at an exposure time of 30 minutes and the first exposure time where the mass loss is over 5% for 17 μm I₂O₅ is at an exposure time of 180 minutes. The exposure time where maximum mass gain is seen and mass loss first reaches 5% suggests that at these exposure times, all I₂O₅ has reacted with water to form HIO₃. Residual water is shown in Tables 3 and 4 and there is no residual water until after the exposure times where maximum mass gain is seen and mass loss first reaches 5%. This suggests that residual water is not formed until all I₂O₅ has reacted with water to form HIO₃ and the assumption that mass loss over 5.11% indicates the sample is pure HIO₃ is reasonable. Fig. 3a and b also show that the mass gained starts to decrease after all I₂O₅ has reacted to form HIO₃. We propose that maximum mass gain is related to the particle size and the extra water reacts with I₂O₅ to form an IO₃⁻ solution. When the samples are removed from 80% RH to 20% RH for weighing and loading into the DSC, rapid desorption of extra water occurs forming an HIO₃ crystal structure. The decrease in mass gain after 180 minutes exposure for 17 μm I₂O₅ and 30 minutes for 5 μm I₂O₅ indicates that an HIO₃ crystal structure is forming while still exposed to 80% RH. The HIO₃ crystal structure that is formed while the samples are still exposed to 80% RH acts as a diffusion barrier that slows the rapid desorption of water when the samples are taken from 80% RH to 20% RH. The diffusion barrier traps the extra water/IO₃⁻ solution in the sample. When the samples are heated in the DSC, discontinuities in the crystal structure are formed by dehydration of the HIO₃, reducing the effects of the diffusion barrier and allowing the residual water to be released in the dehydration temperature range of HIO₃. Further studies with a greater range of particle sizes and synthesis methods will allow us to determine the effects that these variables have on the formation of HI₃O₈ and residual water. A better understanding of these variables could eliminate the need for the assumptions that introduce error, but are necessary to calculate concentration of iodic acid by TG mass change and eqn (4)–(7).

Conclusion

We have shown that the accepted dehydration mechanism is a special condition of a more general mechanism presented here. The condition where the accepted dehydration mechanism is equivalent to the new mechanism is when all HIO₃ dehydrates directly into HI₃O₈. The new dehydration mechanism for I₂O₅ was proven by detecting reduced concentrations of HI₃O₈ in two different I₂O₅ powder samples. When HIO₃ partially dehydrates directly into I₂O₅, the previously accepted mechanism breaks down because the concentration of HIO₃ calculated is greater than 100%. The model presented here more accurately predicts concentrations of iodic acids when assumptions about HI₃O₈ formation from hydration and residual water are made. The validity of the assumptions is shown by comparisons of mass gained during hydration and mass lost during heating. The data presented confirms the validity of the assumptions, but also show that these assumptions introduce small error in the iodic acid concentrations calculated. Further studies are needed that would allow us to expand the hydration/dehydration model that would allow the concentration of iodic acid to be calculated without assumptions that lead to error.

Acknowledgements

The authors Smith and Pantoya are grateful for partial support from DTRA under award HDTRA1-15-1-0029. The authors Parkey and Kesmez are grateful for support from DTRA under award HDTRA1-15-P-0037. We are all grateful for helpful discussion and encouragement from our program manager, Dr Douglas Allen Dalton.

References

1. K. Selte and A. Kjekshus, Iodine Oxides Part II on the System H₂O I₂O₅, Acta Chem. Scand., 1968, 22, 3309–3320.
2. B. K. Little, S. B. Emery, J. C. Nittinger, R. C. Fantasia and C. M. Lindsay, Physiochemical Characterization of Iodine(V) Oxide, Part 1: Hydration Rates, Propellants, Explos., Pyrotech., 2015, 40, 595–603.
3. M. W. Chase, NIST-JANAF Thermochemical Tables for the Iodine Oxides, J. Phys. Chem. Ref. Data, 1996, 25(5), 1297–1340.
4. R. Kumar, R. W. Saunders, a. S. Mahajan, J. M. C. Plane and B. J. Murray, Physical Properties of Iodate Solutions and the Deliquescence of Crystalline I₂O₅ and HIO₃, Atmos. Chem. Phys., 2010, 10(24), 12251–12260.
5. J. Feng, G. Jian, Q. Liu and M. R. Zachariah, Passivated Iodine Pentoxide Oxidizer for Potential Biocidal Nanoenergetic Applications, ACS Appl. Mater. Interfaces, 2013, 5(18), 8875–8880.
6. A. B. Lamb, W. C. Bray and W. J. Geldard, The Preparation of Iodic Acid and Its Anhydride, J. Am. Chem. Soc., 1920, 42(8), 1636–1648.
7. D. K. Smith, K. Hill, M. L. Pantoya, J. S. Parkey and M. Kesmez, Reactive Characterization of Anhydrous Iodine(V) Oxide (I₂O₅) with Aluminum: Amorphous versus Crystalline Microstructures, Thermochim. Acta, 2016, 641, 55–62.
8. K. Selte and A. Kjekshus, Iodine Oxides Part III. The Crystal Structure of I₂O₅, Acta Chem. Scand., 1970, 6(24), 1913–1924.
9. A. Wikjord, P. Taylor, D. Torgerson and L. Hachkowski, Thermal Behaviour of Corona-Precipitated Iodine Oxides, Thermochim. Acta, 1980, 36(3), 367–375.
10. D. K. Smith, J. McCollum and M. L. Pantoya, Effect of Environment of Iodine Oxidation State on Reactivity with Aluminum, Phys. Chem. Chem. Phys., 2016, 18, 11243–11250.
11. B. K. Little, E. J. Welle, S. B. Emery, M. B. Bogle, V. L. Ashley, A. M. Schrand and C. M. Lindsay, Chemical Dynamics of Nano-Aluminum/iodine(V) Oxide, J. Phys.: Conf. Ser., 2014, 500(5), 52025.
12. B. Little, S. Emery and C. Lindsay, Physiochemical Characterization of Iodine(V) Oxide Part II: Morphology and Crystal Structure of Particulate Films, Crystals, 2015, 5(4), 534–550.
13. A. Fischer, Redetermination of HI₃O₈, an Adduct of HIO₃ I₂O₅, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2005, 61, 278–279.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra27854j

---