Rates of Ammonia Absorption and Release in Calcium Chloride

Minnesota, 421 Washington Avenue SE, Minneapolis,. Minnesota 55455, United States. ‡ Chemical Engineering, Texas Tech University, 807 Canton Ave., ...
1 downloads 0 Views 4MB Size
Subscriber access provided by University of Groningen

Article

Rates of Ammonia Absorption and Release in Calcium Chloride Collin Smith, Mahdi Malmali, Chen-Yu Liu, Alon V. McCormick, and Edward L. Cussler ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.8b02108 • Publication Date (Web): 01 Aug 2018 Downloaded from http://pubs.acs.org on August 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Rates of Ammonia Absorption and Release in Calcium Chloride Collin Smith†, Mahdi Malmali ‡, Chen-Yu Liu†, Alon V. McCormick†, and E. L. Cussler*† † Department of Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Avenue SE, Minneapolis, Minnesota 55455, United States ‡ Chemical Engineering, Texas Tech University, 807 Canton Ave., Lubbock, TX 79409, United States *Email: [email protected]

Abstract After synthesis, ammonia can be selectively absorbed by calcium chloride; nitrogen and hydrogen are not absorbed. The kinetics of release seem to be diffusion controlled. The kinetics of absorption are consistent with a first order reaction after an initial period of a higher order reaction, which may indicate a nucleation event. At 225 oC, both absorption and release show half-lives of around ten minutes if the ammonia partial pressure is swung from two to one bar, which allows design of an absorber for the periodic separation of ammonia. When this absorber replaces the ammonia condenser in a conventional ammonia synthesis, the ammonia production at low pressure can have the same rate as the conventional process operating at higher pressure.

Keywords: Absorption, Metal Halide, Kinetics, Ammonia, Pressure Swing Absorption

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction Ammonia is a key chemical in the world’s economy. It is crucial to food production; without ammonia over half the population on earth are expected to starve.[1] Ammonia is also a potential fuel, a way to store and transport hydrogen which has an energy density close to that in hydrocarbons.[2] However, ammonia presently is made from fossil fuels, like coal and natural gas; its manufacture produces about one percent of all carbon dioxide emissions.[3] Making ammonia without making carbon dioxide is thus a major world-wide goal. Ammonia can be made sustainably, from wind, air, and water.[4] The wind supplies the energy for electricity; the electricity is used to separate nitrogen from air, and to make hydrogen from water by electrolysis. The nitrogen and hydrogen are then reacted over a metal oxide catalyst to make ammonia. This reaction is carried out at high temperature, around 400 °C, to break the strong nitrogen triple bond at high rates. Unfortunately, such high temperature also reduces the single-pass conversion of reactants. To circumvent this limitation, the reaction is also carried out at high pressure – above 104 kPa - which does increase the conversion. The ammonia produced is then partly separated by cooling the gas mixture to liquefy the ammonia, and the unreacted hydrogen and nitrogen are recycled. This process, optimized by over a century of effort, is the key to commercial ammonia manufacture.[5-8] We are studying a variety of methods of making ammonia from wind.[4,9-10] Some of these closely parallel the current Haber-Bosch process but are at small scale, suitable for example, for an agricultural co-op. However, this small scale has not surprisingly meant that the capital investment per mass of ammonia produced is too high to compete with the existing fossil fuel based processes. As a result, we are exploring small scale, low-pressure processes which would have potentially lower cost. This cost is lower because the capital expense of the process itself is lower, and because the operating costs of compression of the hydrogen and nitrogen are reduced. The best developed of these low-pressure processes replaces ammonia liquefaction with ammonia absorption, that is, selectively taking up the ammonia using solid crystals of alkaline earth metal halides.[9] These salts, called ammines, can absorb large quantities of ammonia in their ammine form: for example, at room temperature, a tablespoon of magnesium chloride will absorb over 50 liters of ammonia at standard temperature and pressure. This large capacity under ambient conditions has motivated studies of its use for ammonia storage.[11] While the amount of

ACS Paragon Plus Environment

Page 2 of 36

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

absorption is reduced at higher temperatures, it is still significant even at temperatures approaching those of the ammonia synthesis reaction. In this paper, we report further studies of these solid absorbents and their absorption/release kinetics. We have already shown which salts absorb more ammonia.[10] We have found that in their pure form, these absorbents are not always stable, but that this stability is dramatically improved by supporting the salt on microporous, inert solids.[12] We have not shown how fast their absorption and release can be. That is the goal of this paper. In the following sections, we report measurements of absorption and release rates. We explore the different rate-controlling steps involved. Finally, we show how these measurements can guide the development of pilotscale processes which will show if this modified ammonia synthesis can be an alternative route to making this important chemical. Before beginning our report, we review the terminology and methods of measuring absorption and adsorption. By “absorption,” we mean here that a target species is taken into a solid by reaction, in our case a crystal. This reactive absorption is selective and can occur at temperatures higher than ambient. By “adsorption,” we consider uptake onto the surface of a solid, often made microporous to increase the surface area available. Usually, adsorption is less selective than absorption, and adsorption may drop more swiftly with increased temperature. In addition, we define “release” to mean the desorption of ammonia after absorption to distinguish it from desorption after adsorption. There are three main groups of methods for measuring kinetics of absorption. The first group uses gas flow through a packed column of solid absorbent, with a detector to measure the outlet concentration in the gas. While the mathematical analysis to calculate kinetic parameters can be complex,[13] it can be simplified significantly when the packed column is so short that absorbent is effectively exposed to a uniform concentration of gas.[14,15] The second group of methods suddenly changes the pressure in a static system of gas and absorbent, and measures either the change in pressure[16] or the amount injected to keep the pressure constant. The third group of methods, which is that used in this paper, again begins with a static system of gas and absorbent, but alters absorption by a change in temperature and measures absorption in a similar manner as the second group of methods. The results obtained from this method are complicated because the initial change in temperature must be fast relative to the time over which pressure change is measured. So far, we have found changing the temperature and measuring the pressure to be the

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

most complete method to explore mechanisms for metal halide salts because it offers the possibility of isolating individual “steps” in the salt capacity. While we limit our scope to the transition from pure salt to one molar equivalent of ammonia, salts such as CaCl2 will also transition from one to two, two to four, and four to eight molar equivalents under the right conditions. Because metal halide absorbents form discrete coordination numbers with ammonia,[11] isotherms should form discrete “steps” in different levels of saturation. Experimental isotherms have largely supported this hypothesis.[17] Steps in isotherms have also been observed in other systems that involve a gas species reacting with a solid to form another solid.[18] Curvature in experimental isotherms could result from kinetic limitations in reaching equilibrium between the salt and the surrounding environment. How much these steps and curves are reflected in experiments is illustrated by the results below.

Experimental Materials & Apparatus. Anhydrous CaCl2 in powder >97% purity (No. C4901 CAS 1004352-4, Lot No. SLBQ3073V) with inhomogeneous particle size was purchased from SigmaAldrich (St. Louis, MO). Anhydrous ammonia (>99.99%) was purchased from Matheson TriGas (Eagan, MN). The experimental apparatus, as shown in Figure 1, was built using Swagelok 316 stainless steel tubing and fittings (Chaska, MN). The sample chamber, 2.5 inches of half inch tubing containing 54.2 mg of CaCl2 physically mixed in 2 grams of glass beads (200 micron) was retained on either side by stainless steel wool. Even though previous studies indicate that supported MgCl2 or MgBr2 are the best sorbents for the process, unsupported CaCl2 was used here both because CaCl2 releases ammonia at lower temperatures, which makes it easier to systematically study, and because determining the surface area of CaCl2 is more straightforward. An Omega K-type thermocouple (Model DRG-SC-TC) in the center of the sample chamber was utilized to measure the temperature. The chamber temperature was controlled using a PID controller, which was tuned such that the temperature in the chamber deviated less than 0.5°C once the setpoint temperature was reached. The chamber was connected on each side to long sections of empty tubing at ambient temperature. Each section of tubing was separated by a Swagelok needle valve to split the system into three sections. The volumes of each section were 26, 15, and 37 cm3. The middle chamber was used in isolation to measure the

ACS Paragon Plus Environment

Page 4 of 36

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

equilibrium condition at any given temperature by restricting the absorbent from reaching saturation, while all three chambers were used in conjunction when measuring the kinetics of absorption. An ammonia mass flow controller (Alicat Scientific; Tuscon, AZ) was used to control the gas feed into the absorber. The PID controller and a pressure transducer (Dwyer; Michigan City, IN) were connected to a National Instruments CompactRio real time controller (Austin, TX) routed to a computer for data collection and apparatus control. General Procedure: When ammonia is absorbed into the salt, one of two conditions are generally true: either there is excess salt such that the pressure reaches the equilibrium pressure or there is excess ammonia such that the salt reaches saturation where there is one mole of ammonia per mole of calcium chloride. (While calcium chloride is capable of absorbing more ammonia, it does not occur under the experimental conditions considered.) The first case is utilized when measuring equilibrium at any given temperature and the second case is utilized when measuring kinetics at any given temperature and pressure. Saturation is not an equilibrium condition because if more salt is added the pressure continues to change even though the amount of salt in the system is not represented in the equilibrium constant and should be irrelevant at true equilibrium. The general process for implementing these two cases are shown on the theoretical equilibrium pressure curve in Figure 2. The gray arrows correspond to equilibrium measurements while the black arrows correspond to kinetic measurements. Equilibrium Measurements: To measure equilibrium pressures below atmospheric pressure, ammonia was first purged through the system at atmospheric pressure and 20 standard cubic centimeter per minute (SCCM). The temperature of the sample chamber was set at 250 °C because it is known that calcium chloride does not absorb ammonia at that temperature and pressure.[11] The mass flow controller was shut off and the valves surrounding the sample chamber, as well as the exhaust and inlet valve, were closed. So the starting point for these test was 1 bar of ammonia and 250 °C. Then we exposed the absorber chamber to a controlled decrease in temperature that resulted in ammonia being absorbed and the pressure dropping. As noted, we only used the middle section of the apparatus for this type of experiment for equilibrium measurements. We designed the absorber to have a volume so that at our starting ammonia pressure the sample would never able to absorb one molar equivalent of ammonia; so in this type of experiment, absorbent material reached the equilibrium pressure of formation of monoamminated structure at the given temperature, but it never achieved a fully saturated

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

monoamminated structure. Since it was generally thought that release would be faster than absorption, the temperature was first decreased to 50 °C. Then, the temperature was increased incrementally until the pressure was within 21 kPa of atmospheric pressure, at which point the salt was nearly dry of ammonia and the equilibrium pressure could not be accurately measured accurately. For each data point reported here, the pressure was monitored until it was changing by less than 0.4 kPa per hour. The final pressure achieved at each temperature was approximated to be the equilibrium pressure at that temperature. For equilibrium pressures above atmospheric pressure, all three chambers were charged with ammonia (valves between chambers open) to various pressures and 250-300 ºC. The temperature was decreased slowly such that only one molar equivalent of ammonia would absorb. After the pressure had dropped by approximately 15 kPa – indicating one mole of ammonia had been absorbed based on the total volume of the system and moles of salt – the valves surrounding the sample chamber were closed and the temperature was gradually increased in steps. Once the temperature increased enough for ammonia to be released, the isolated sample chamber exceeded the initial charge pressure and approached the equilibrium pressure at a given temperature. The pressure was recorded until it was increasing by only 1 kPa per hour and the exponential decay curve was used to approximate the final equilibrium pressure. This was repeated several times. Kinetic Measurements: To measure absorption rates, ammonia was flushed through the system and the exhaust valve was closed to charge the system to a desired pressure with the sample chamber at 250-300 ºC. Then, the mass flow controller was shut off and the inlet valve was closed. The temperature in the sample chamber was decreased to facilitate absorption of the first mole of ammonia. The pressure was recorded to measure the kinetics of this process. This was repeated at several temperatures and initial charge pressures. The target temperatures were chosen such that absorption would be fast enough to measure but not so fast that the sample would reach saturation before the target temperature was reached. The process to measure ammonia release was similar, except after absorption the temperature was increased and the return to the original pressure was measured.

ACS Paragon Plus Environment

Page 6 of 36

Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Results Before we made our measurements of absorption and release, we check the reproducibility of the calcium chloride absorbent. This is important because earlier studies show that these materials are not always stable with multiple cycles, especially when large amounts of pure salts are used without inert supports.[11] These instabilities often lead to capacities on the first cycle which are larger than those on, for example, the fifth cycle. This is apparently due to the salt fusing into a mass which causes much slower absorption. That this is not true here for the small amounts of pure salt used is shown by the data in Figure 3. The rates of absorption are the same on the first cycle as they are on the fortieth cycle. The small amount of calcium chloride used here shows reproducible absorption. The BET surface area of this stable absorbent, before and after use, is a constant 9.4 m2/gram. With this proof of stability, we next report the results under three headings. First, we report equilibria at different temperatures. These temperatures determine the equilibrium pressure (Pe), that is, the pressure at which the salt will begin absorbing ammonia and will reach saturation if an excess of gaseous ammonia is provided. Second, we report ammonia release kinetics, which seem to be diffusion-controlled. Last, we report absorption kinetics, which seem to be controlled by a first order process and may involve an initial, nucleation-like step. These results let us begin the design of the separation steps of a low-pressure ammonia synthesis, which is discussed later in the paper.

Equilibrium Pressure. The pressure of ammonia vs. time is recorded first for cases where the system is under a pressure and temperature where ammonia release to form dried calcium chloride begins, but where complete release is not possible because of the equilibrium pressure. The resulting profiles of pressure with time can be used to estimate the equilibrium pressure for the first mole of ammonia reacting with calcium chloride to make a solid mono-amminated structure. Since this reaction involves one gaseous species and two solid species – pure salt and amminated form - the equilibrium constant for the reaction is the inverse of the equilibrium pressure. Plotting the natural log of the equilibrium constant against the inverse of temperature gives a straight line like that shown in Figure 4. The slope on this plot is proportional to the enthalpy of reaction; and the y-intercept, that is, the equilibrium constant at infinite temperature,

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

is proportional to the entropy of reaction. Table 1 shows that the values of enthalpy and entropy found here are within a factor of two of those found for similar systems in the literature.

Table 1 | Comparison of ∆H and ∆S of the Equilibrium Pressure ∆H (kJ mol-1) ∆S (J mol-1 K-1) -51 -105 -69 -230 -43 -132 -69 --42 -134 -64 -97

Source This Work Neveu 1993[19] Sharanov 2006[20] Sorensen 2008[11] Carling 1981[21] Aoki 2015[22]

Material Calcium Chloride Calcium Chloride Calcium Chloride Calcium Chloride Calcium Chloride Magnesium Chloride

Release Kinetics We next consider the rates of ammonia release. To study this rate, the temperature was increased such that the release starts quickly, but complete release takes much longer. This means that the pressure shows a sharp initial increase but slows with time as shown by the raw data in Figure 5. To correlate these data more concisely, we next plot the pressure change (P-P0) divided by the square root of the difference between the equilibrium pressure and the initial pressure (Pe-P0) as is characteristic of a simplified shrinking core process as shown in the supplemental material. When this reduced pressure is plotted against the square root of time the data are approximately linear, as shown in Figure 6. Such a variation is a signature of diffusion-controlled processes. In comparison, the natural log of dimensionless pressure change, as is characteristic of a first order process, is plotted in Figure 7 against time to the first power. The nonlinearity of the plots suggests release is not a simple first-order process. If release is in fact diffusion controlled, the initial slope of the data in Figure 6 before normalizing the axes is a(2RTCsatD)0.5, where a is the absorbent area per volume of gas, D is the diffusion coefficient, Csat is the concentration of ammonia in saturated salt, and T is the temperature. The data in Figure 6 imply an average diffusion coefficient of approximately 5(±1) .

10-13 cm2/s, two orders of magnitude smaller than that inferred from earlier measurements.[23]

While we are unsure what causes this discrepancy, we suspect that the smaller values may be due to the use of MgCl2 in previous experiments rather than CaCl2 and an incorrect assessment of the equilibrium pressure.

ACS Paragon Plus Environment

Page 8 of 36

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Absorption Kinetics. The dynamics of pressure change during absorption are exemplified by the raw data in Figure 8. The absorption curves have an inflection at early times due to the time required to drop the temperature in the sample chamber. At longer times, the curves approach the same asymptote regardless of temperature, indicative of saturation – one molar equivalent of ammonia. To analyze these data more completely, we define time zero for each set of conditions as the time at which the temperature reaches the set point; and we take the final time as that when the sample reaches 75% absorption. In other words, we remove complications involved with having a transient temperature by only considering pressure measurements at or near the set point temperature. In trials where the set point temperature is low enough that absorption began before the set point is reached, time zero was defined as that when 25% absorption is reached, which, we found, is always less than 3 ºC from the set point. This only occurred in the fastest absorption cases, such as the data at 205 ℃ in Figure 8. In trials where the set point was overshot by the PID controller by less than 5 ºC, the temperature characterizing the trial is the average of the temperature over the points analyzed. Any pressure change is thus implicitly taken to be the result of ammonia absorption rather than a decrease in the temperature of the gas in the sample chamber. This seems reasonable, because the pressure only decreases by 0.2% per 10 ºC due to the properties of an ideal gas. To analyze these data, we again use a reduced pressure for diffusion limitations and a dimensionless pressure for a first-order process. We want to know if data like those in Figure 8 are controlled by diffusion, like that for release, or by some other rate controlling step. To see if they are controlled by diffusion, we plot the natural log reduced pressure vs. the square root of time as shown in Figure 9. While the data may approach linear behavior at larger time, they show non-zero intercepts at smaller times, inconsistent with diffusion control. When we plot the reduced pressure vs. time, the results are more nearly linear, as shown in Figure 10. These data contradict previous results on magnesium chloride, which concluded absorption was diffusion controlled.[23] While the data in Figure 10 imply a first order process, the scale of the y-axis of this figure indicates that the pressure never approaches the equilibrium pressure for calcium chloride. Because the necessary driving force for appreciable rate of absorption, observed to be approximately 69 kPa, is far from the equilibrium pressure at any given temperature, the driving force for the absorption does not significantly decrease as absorption proceeds. Once saturation

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

is reached, the equilibrium pressure will change to the current pressure so the driving force becomes zero. Therefore, Figure 10 does not remove the possibility of a higher order reaction, as has been observed in a number of other systems.[24] Because the absorption is highly exothermic as is shown in Table 1, transferring heat away from the particle is also a concern. But, we believe that heat transfer in the particle is fast relative to conductive heat transfer away from the particle. To support this belief, we note that a typical absorption in these experiments takes 30 minutes and releases 0.11W of energy if an upper estimate of 69 kJ per mole is assumed for the enthalpy of reaction. The calcium chloride absorbent has a particle size of approximately 200 µm, as measured under a light microscope. This corresponds to a surface area of 0.001 m2/g, which is 300 times less than the BET surface area. The heat transfer coefficient of ammonia gas is around 40 W m-2K-1, as estimated using a thermal conductivity of 0.04 W m-1K-1[25] and an overestimated distance of 1 mm between the absorbent particle and either a glass particle or the stainless-steel chamber. Using these estimates, the temperature rise in the particle is estimated to be only 3 ºC. In the case of the fastest reaction experiment, the temperature rise is estimated to be 18 ºC. The thermocouple in the sample chamber registered no such temperature rise. Therefore, while heat transfer limitations may be significant under fast reaction conditions, they are insignificant under slower reaction conditions. and we conclude that heat transfer limitations are not important here. Thus the results above suggest that the rate controlling step for release is diffusion, but the rate of adsorption depends more on detailed chemical steps. The implications of this result for the low-pressure synthesis of ammonia using absorption are discussed in greater detail in the next, final section of this paper.

Discussion The data above lead to two ideas which merit discussion. The first, more scientific idea concerns the mechanism of ammonia movement into and out of the solid phase. The second, more practical concern is the basis which these experiments provide for designing alternative ways to make ammonia. Each idea is discussed here. The data above show that ammonia release from calcium chloride is diffusion controlled, and that ammonia uptake is reaction controlled. However, these data do not identify the particular chemistry involved. To show this, we first consider release. If the calcium chloride is

ACS Paragon Plus Environment

Page 10 of 36

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

homogeneous, then the probable mechanism involves a shell of CaCl2 surrounding a core of CaCl2.NH3. The release could involve the rapid decay of CaCl2.NH3 into pure CaCl2 and free NH3 ammonia; the free ammonia then slowly diffuses out from a core of complex to the surrounding gas through a shell of CaCl2. If the ammine salt was opaque and the salt clear, a cross section would show an opaque core and a clear shell. This would give the observed dependence of release on the square root of time. Alternatively, if the calcium chloride were not homogeneous but rather nanoporous, the ammonia could be quickly released within the solid, and slowly diffuse to the walls of the pores. Now, if the ammine were opaque, we would see gradations in the opaque of the solid. This would also explain the rate varying with the square root of time. However, it might imply that absorption should also be diffusion controlled, which is not supported by our data. Without more measurements of ammonia distribution within the absorbent, we cannot be sure of the detailed mechanism of release. In a similar sense, we cannot identify all details of the mechanism of absorption. The rate of absorption is slower than for release, consistent with the hypothesis that diffusion is not rate controlling. The data do show the characteristic behavior of a first order reaction, but require a large distance from equilibrium. This might be correlated by a higher order reaction, as high as seventh order, as will be discussed and shown in Figure 11. Alternatively, it could suggest some sort of nucleation event, perhaps a formation of an intermediate cluster of amminated CaCl2 nucleated on solid CaCl2. After this cluster formation, the addition of an additional NH3 causes the solid ammine crystal to form at a rate that is not limited by the transport of ammonia through the complex. While such behavior frequently occurs in catalytic systems, we have no direct evidence about the chemistry responsible here. For now, we only want to point to the complexity which must be involved. The second, more practical idea derived from the data above concerns the implications for small scale ammonia synthesis, which is the origin of this project. In particular, the results can guide speculation about the feasibility of an ammonia synthesis in which ammonia absorption by calcium chloride replaces ammonia condensation. We already know that such a process can give greater ammonia synthesis rates at least seven times lower pressure when compared to a conventional Haber-Bosch process pilot plant.[9] We know that the capital of the absorbent process will be lower because less metal is needed for the lower pressure reactor and separator, and because the lower pressure means that the compressor needs less energy. We also

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 36

know how to stabilize the ammonia selective absorbents[10,12], and which absorbent chemistries promise better performance. This paper lets us estimate two important aspects of absorbent operation. First, we now know the times required for absorption and release, which affect the size and expense of the absorber operation. Second, we can now estimate the specific conditions which will give the best absorber performance. These two aspects are explored further below. We first consider absorbent cycle times. In Table 2A, we list the half-lives at 207 kPa for the absorption of one molar equivalent of ammonia vs. the temperature and the distance from equilibrium, calculated using the equilibrium pressure. In Table 2B, we give the half-lives for release of one molar equivalent of ammonia at 207 kPa and various temperatures. The absorption half-lives indicate that absorption proceeds slowly at small distances from equilibrium, but increases rapidly before reaching a plateau at larger distances from equilibrium. This stands in stark contrast to the release rates, where only a distance of 20 kPa from equilibrium is required for the same rate as absorption at a distance of 110 kPa from equilibrium.

Table 2. Kinetics of Ammonia Interactions with Calcium Chloride A | Absorption at 207 kPa Temperature (ºC)

238

Distance from Equil. (kPa) 62 Half Life (min)

1700

235

232

225

220

215

210

205

195

76

90

110

124

138

145

152

165

670

300

21

15

8

6

4

3

B | Release at 207 kPa Temperature (ºC)

250

252

255

Distance from Equil. (kPa)

7

20

41

Half Life (min)

75

30

10

This general trend in the rate of absorption requiring large distances from equilibrium similarly appears when the half-life is plotted against the distance from equilibrium at various starting pressures, as shown in Figure 11.A, or at various temperatures, as shown in Figure 11.B. Temperatures are represented in a range because the absorption temperature was not the same for

ACS Paragon Plus Environment

Page 13 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

different starting pressures. The trend strongly indicates that the distance from equilibrium, rather than the temperature or starting pressure, is the independent variable in the rate of ammonia absorption. The model used to capture this trend, as shown in Figure 11, is a rate equation that is seventh order at small distances from equilibrium and first order at large distances from equilibrium. This is an interesting result; however, it is so far only applicable to the reaction of one molar equivalent of ammonia with calcium chloride and not further levels of absorption with two, four, or eight molar equivalents of ammonia. For the purpose of process sensitivity analysis, the sum of the half-life for absorption and the half-life for release is defined as the process time tp. The half-life for absorption is calculated from the model pictured in figure 11 and the half-life for release is calculated using a simple inverse relation with distance from equilibrium that reflects the data in Table 2.B. This process time tp is the inverse of the combined rate of absorption and release, which is to be optimized by changing the absorption and release conditions. Since the process will always become faster as the absorption and release pressure are independently increased and decreased, respectively, the temperature is the most useful sensitivity parameter. For simplicity, it is assumed that absorption and release occur at the same temperature. As a starting point, it is assumed the absorption pressure is 200 kPa (twice atmospheric pressure) and the release pressure is 100 kPa (atmospheric pressure). As shown by the solid line in Figure 12, there is a temperature at which the overall process time is at a minimum. The asymptote at higher temperature is due to the absorption becoming very slow and the asymptote at lower temperature is due to the release becoming very slow. With this as a base case, the effect of increasing the absorption pressure or decreasing the release pressure on the overall process time is examined. As shown in Figure 12, as the absorption pressure increases, the optimal temperature increases as the higher temperature asymptote shifts. Conversely, as release pressure decreases, the optimal temperature decreases as the lower temperature asymptote shifts. In either case, the overall rate of the process, and the individual rates of absorption and release, increase by the same amount whether the absorption pressure increases or the release pressure decreases, as shown in Table 3. This indicates that the absorption and release are fundamentally connected. For example, as the absorption pressure increases, the process temperature can be increased, which improves the rate of release.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 36

Table 3. Design Parameters from the Minimum Process Time (tp) in Figure 12 Absorption Pressure

Optimized

Rate of Absorption

Release Half

[Release Pressure] (kPa)

Temperature (°C)

10-5 mol min-1 g-1

Life (min)

200 [73]

222

15.7

27

228 [100]

231

15.4

27

214 [100]

229

9.5

38

200 [100]

227

5.0

57

If the case of 228 kPa absorption and 100 kPa release is taken as the design, the absorber temperature would be 231 ºC. In order to uptake the equivalent of 1 kg of ammonia per day, 260 g of calcium chloride is necessary with a rate of 15.4 .10-5 mole min-1g-1. The absorption cycle time would be 60 min and the release cycle time would be 60 min. The overall pressure would need to be at least 4560 kPa in order to reach a partial ammonia pressure of 228 kPa, assuming 5% ammonia in the reactor effluent. It is important to note that this analysis does not consider the decrease in absorption rate that would occur as the ammonia pressure would decrease with distance through the absorber. However, this is not crucial to the process, because as long as the catalyst is not poisoned by a significant recycle of ammonia these operating conditions will still be feasible. There are two general conclusions that are drawn from this optimization. The first is that increasing the absorption pressure or decreasing the release pressure will have an equal effect on improving the process, and the practical route should depend only on equipment considerations. For example, if a vacuum for release needs to be avoided, this can be mitigated by simply increasing the absorption pressure. Secondly, if a reactor and absorber at isothermal conditions is the goal, the overall pressure will need to be increased in order for the absorption to occur at a significant rate. The target pressure has been set as low as 2000 kPa so that compression costs are minimized. However, unless an absorbent that can absorb ammonia at an appreciable rate at a higher temperature is found, this target will need to be conceded for the sake of removing the equilibrium limitations of the ammonia synthesis reaction.

ACS Paragon Plus Environment

Page 15 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Acknowledgements This work was funded in part by the Advanced Research Projects Agency-Energy (ARPA-E), U.S. Department of Energy, under Award Number DE-AR0000804; in part by the Minnesota Environment and Natural Resources Trust Fund as recommended by the Legislative-Citizen Commission on Minnesota Resources (LCCMR,/ ML 2015, CH 76, SEC 2, SUBD 07A); and in part by the MnDRIVE initiative of the University of Minnesota (MNT11). The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. Giang Le is acknowledged for assistance. Supporting Information Description Brief mathematical derivation of gas pressure surrounding a particle as a function of time during ammonia release from a shrinking core where diffusion is limiting. References (1) Smil, V. Detonator of the Population Explosion, Nature 1999, 400, 415, DOI 10.4236/am.2014.513188. (2) Klerke, A.; Christensen, C. H.; Nørskov, J. K.; Vegge, T. Ammonia for hydrogen storage: challenges and opportunities. J. Mater. Chem. 2008, 18, 2304-2310, DOI 10.1039/B720020J. (3) Brown, Trevor. “Ammonia production causes 1% of total global GHG emissions.” Accessed June 15, 2018. https://ammoniaindustry.com/ammonia-production-causes-1percent-of-total-global-ghg-emissions. (4) Reese, M.; Marquart, C.; Malmali, M.; Wagner, K.; Buchanan, E.; McCormick, A.; and Cussler, E.L.; Performance of a small-scale Haber process, Industrial & Engineering Chemistry Research 2016, 55, 3742-3750, DOI 10.1021/acs.iecr.5b04909. (5) Dyson, D. C.; Simon, J. M. Kinetic expression with diffusion correction for ammonia synthesis on industrial catalyst. Ind. Eng. Chem. Fundam. 1968, 7, 605-610, DOI DOI: 10.1021/i160028a013. (6) Temkin, M.; Pyzhev, V. Kinetics of ammonia synthesis on promoted catalysts. Acta Physiochim. USSR 1940, 12, 327-356. (7) Annable, D. Application of the temkin kinetic equation to ammonia synthesis in largescale reactors. Chem. Eng. Sci. 1952, 1, 145-196, DOI 10.1016/0009-2509(52)87011-3.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(8) Horiuti, J.; Takezawa, N. The mechanism of catalyzed synthesis of ammonia in the presence of doubly promoted iron catalyst. J. Res. Inst. Catal., Hokkaido Univ. 1960, 170−187, DOI 10.1080/03602458108068066. (9) Malmali, M.; Wei, Y.; McCormick, A.; and Cussler, E.L.; Ammonia synthesis at reduced pressure via reactive separation, Industrial & Engineering Chemistry Research 2016 55, 8922-8932, DOI 10.3791/55691. (10) Malmali, M.; Le, G.; Hendrickson, J.; Prince, J.; McCormick, A.V.; and Cussler, E.L.; Better absorbents for ammonia separation. 2018 ACS Sustainable Chem. Eng, DOI 10.1021/acssuschemeng.7b04684. (11) Sørensen, R.; Hummelshøj, J.S.; Klerke, A.; Reves, J.B.; Vegge, T.; Nørskov, J.K.; and Christensen, C.H. Indirect, reversible high-density hydrogen storage in compact metal ammine salts. Journal of the American Chemical Society 2008, 130, 8660-8668, DOI 10.1021/ja076762c. (12) Wagner, K.; Malmali, M.; Smith, C.; McCormick, A.; Cussler, E. L.; Zhu, M.; Seaton, N. C. A. Column absorption for reproducible cyclic separation in small scale ammonia synthesis, AIChE Journal 2017 ,63, 3058-3068, DOI 10.1002/aic.15685. (13) Aribike, D.S. and Olafadehan, O.A. Modeling of fixed bed adsorption of phenols on activated carbon, Theor. Found Chem Eng 2008, 42, 257-263, DOI 10.1134/S0040579508030056. (14) Teixeira, A.; Chang, C.; Fan, W.; Dauenhauer, P.; Dominance of surface barriers in molecular transport through silicalite-1, Journal of Physical Chemistry C 2013, 117, 25545-25555, DOI 10.1021/jp4089595. (15) Marcussen, L. The kinetics of water adsorption on porous alumina,” Chemical Engineering Science 1970, 25, 1487-1499, DOI 10.1016/0009-2509(70)85070-9. (16) Aristov, Y.; Tokarev, M.; Freni, A.; Glaznev, I.; Restuccia, G.; Kinetics of water adsorption on silica Fuji Davison RD, Microporous and Mesoporous Materials 2006, 96, 65-71, DOI 10.1016/j.micromeso.2006.06.008. (17) Liu, C.; Aika, K.; Ammonia absorption on alkaline earth halides as ammonia separation and storage procedure. Bull. Chem. Soc. Jpn. 2004, 77, 123-131, DOI 10.1021/ie049874a.

ACS Paragon Plus Environment

Page 16 of 36

Page 17 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

(18) Sandler, S.I. Chemical, Biochemical, and Engineering Thermodynamics. John Wiley & Sons: Hoboken, NJ, 2006. 738-739. (19) Neveu, P.; Castaing, J.; Solid-gas chemical heat pumps: Field of application and performance of the internal heat of reaction recovery process. Heat Recovery Systems and CHP 1993, 13, 233-251, DOI 10.1016/0890-4332(93)90014-M. (20) Sharonov, V.; Veselovskaya, J.; Aristov, Y. Ammonia sorption on composites ‘CaCl 2 in inorganic host matrix’: isosteric chart and its performance. International Journal of Low-carbon Technologies 2006, 1, 191-200, DOI 10.1093/ijlct/1.3.191. (21) Carling, R.; Dissociation pressures and enthalpies of reaction in MgCl2 · nH2O and CaCl2 · nNH3. J. Chem. Thermodynamics 1981, 13, 503–512, DOI 10.1016/00219614(81)90105-1. (22) Aoki, T.; Miyaoka, H.; Inokawa, H.; Ichikawa, T.; Kojima, Y. Activation on ammonia absorbing reaction for magnesium chloride. The Journal of Physical Chemistry C 2015. 119. 26296−26302, DOI 10.1021/acs.jpcc.5b07965. (23) Huberty, M. S., Wagner, A. L., McCormick, A. and Cussler, E. Ammonia absorption at haber process conditions. AIChE J. 2012, 58: 3526–3532, DOI 10.1002/aic.13744. (24) Aris, R.; Astarita, G.; Continuous lumping of nonlinear chemical kinetics, Chemical Engineering and Processing: Process Intensification 1989, 26, 63-69, DOI 10.1016/0255-2701(89)87007-2. (25) Afshar, R.; Murad, S.; and Saxena, S.C. Thermal conductivity of gaseous ammonia in the temperature range 358-925 K, Chemical Engineering Communications 1981, 10, 111, DOI 10.1080/00986448108910921.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1 Schematic of apparatus. An absorbent is heated in a small chamber and surrounded by a large volume at ambient temperature. Pressure change indicates absorption.

Figure 2 Schematic of equilibrium (gray) and kinetic (black) experiments. The goal in measuring equilibrium is to approach the equilibrium pressure gradually, while the goal in measuring kinetics approach saturation in an reasonable amount of time. The equilibrium pressure is approached during the release rather than the absorption of ammonia because generally release is much faster.

ACS Paragon Plus Environment

Page 18 of 36

Page 19 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 3 Stability of absorbent over many absorption desorption cycles. The slower rate before 0.5 hours is due to a lag in temperature.

Figure 4 Inverse of equilibrium pressure vs. inverse temperature. The equilibrium constant for the absorption of ammonia is simply the inverse of the ammonia pressure because other species are solids. The slope is proportional to enthalpy of reaction and the intercept is proportional to entropy of reaction. Insert: raw data of equilibrium pressure as a function of temperature.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5 Ammonia release at 207 kPa starting pressure. The profiles indicate a fast initial rate and a slower rate at longer times, which suggests a diffusion limitation. Complete release of ammonia corresponds to14 kPa of pressure change.

Figure 6 Ammonia release kinetics with a diffusion limitation. The linearity with the square-root of time is indicative of a diffusion controlled process.

ACS Paragon Plus Environment

Page 20 of 36

Page 21 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 7 Ammonia release kinetics with only a first-order reaction. In comparison to Figure 6, a lack of first-order kinetics is shown by nonlinearity with time.

Figure 8 Absorption profiles at 207 kPa starting pressure. The initial inflection point is due to a lag in achieving the temperature setpoint. All profiles reach an asymptote indicative of one molar equivalent of ammonia absorbed per mole of calcium chloride based on the system volume and amount of salt.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 9 Absorption kinetics with a diffusion limitation. From the data in Figure 8, time zero is defined as the time at which the set-point is reached. Even though the profiles are linear at long times, the nonlinearities at small times suggest a lack of diffusion limitation.

Figure 10 Absorption kinetics with only a first order reaction. From the data in Figure 8, time zero is defined as the time at which the set-point is reached. All profiles indicate strong linearity in comparison to Figure 9.

ACS Paragon Plus Environment

Page 22 of 36

Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 11 Summary of all absorption half-lives. The rate of absorption (half-life) is dependent only on the distance from equilibrium. The model shown is high order (seventh order here) at small distances from equilibrium, which suggests a nucleation event. At large distances from equilibrium, the model is first-order. Whether the data is organized by starting pressure (A) or temperature (B), the only trend is with the distance from equilibrium as measured at the start.

Figure 12 Process sensitivity with temperature at several absorption/release conditions. The overall process time (tp) is minimized at optimal conditions. The high temperature asymptote is due to absorption being very slow and the low temperature asymptote is due to release being very slow. The overall process time (tp) decreases by the same amount if the absorption pressure is increased or if the release pressure is decreased.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For Table of Contents Use Only

Synopsis Ammonia synthesis at low pressure using stranded renewable energy can be accelerated with calcium chloride as a selective absorbent.

ACS Paragon Plus Environment

Page 24 of 36

Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 1 Schematic of apparatus. An absorbent is heated in a small chamber and surrounded by a large volume at ambient temperature. Pressure change indicates absorption.

456x235mm (96 x 96 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2 Schematic of equilibrium (gray) and kinetic (black) experiments. The goal in measuring equilibrium is to approach the equilibrium pressure gradually, while the goal in measuring kinetics approach saturation in an reasonable amount of time. The equilibrium pressure is approached during the release rather than the absorption of ammonia because generally release is much faster.

354x242mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 26 of 36

Page 27 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 3 Stability of absorbent over many absorption desorption cycles. The slower rate before 0.5 hours is due to a lag in temperature. 289x199mm (149 x 149 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4 Inverse of equilibrium pressure vs. inverse temperature. The equilibrium constant for the absorption of ammonia is simply the inverse of the ammonia pressure because other species are solids. The slope is proportional to enthalpy of reaction and the intercept is proportional to entropy of reaction. Insert: raw data of equilibrium pressure as a function of temperature. 352x242mm (149 x 149 DPI)

ACS Paragon Plus Environment

Page 28 of 36

Page 29 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 5 Ammonia release at 207 kPa starting pressure. The profiles indicate a fast initial rate and a slower rate at longer times, which suggests a diffusion limitation. Complete release of ammonia corresponds to14 kPa of pressure change. 324x219mm (149 x 149 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6 Ammonia release kinetics with a diffusion limitation. The linearity with the square-root of time is indicative of a diffusion controlled process. 310x212mm (149 x 149 DPI)

ACS Paragon Plus Environment

Page 30 of 36

Page 31 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 7 Ammonia release kinetics with only a first-order reaction. In comparison to Figure 6, a lack of firstorder kinetics is shown by nonlinearity with time. 326x228mm (149 x 149 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 8 Absorption profiles at 207 kPa starting pressure. The initial inflection point is due to a lag in achieving the temperature setpoint. All profiles reach an asymptote indicative of one molar equivalent of ammonia absorbed per mole of calcium chloride based on the system volume and amount of salt. 347x214mm (149 x 149 DPI)

ACS Paragon Plus Environment

Page 32 of 36

Page 33 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 9 Absorption kinetics with a diffusion limitation. From the data in Figure 8, time zero is defined as the time at which the set-point is reached. Even though the profiles are linear at long times, the nonlinearities at small times suggest a lack of diffusion limitation. 388x234mm (149 x 149 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 10 Absorption kinetics with only a first order reaction. From the data in Figure 8, time zero is defined as the time at which the set-point is reached. All profiles indicate strong linearity in comparison to Figure 9. 356x219mm (149 x 149 DPI)

ACS Paragon Plus Environment

Page 34 of 36

Page 35 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Figure 11 Summary of all absorption data half-lives. The rate of absorption (half-life) is dependent only on the distance from equilibrium. The model shown is high order (seventh order here) at small distances from equilibrium, which suggests a nucleation event. At large distances from equilibrium, the model is first-order. Whether the data is organized by starting pressure (A) or temperature (B), the only trend is with the distance from equilibrium as measured at the start. 304x177mm (120 x 120 DPI)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 12 Process sensitivity with temperature at several absorption/release conditions. The overall process time (tp) is minimized at optimal conditions. The high temperature asymptote is due to absorption being very slow and the low temperature asymptote is due to release being very slow. The overall process time (tp) decreases by the same amount if the absorption pressure is increased or if the release pressure is decreased.

372x213mm (149 x 149 DPI)

ACS Paragon Plus Environment

Page 36 of 36