Performance and Properties of a Solid Amine ... - ACS Publications

Przemyslaw RzepkaZoltán BacsikStef SmeetsThomas C. HansenNiklas HedinDariusz ...... James Murphy , Steven Petinakis , Katherine Zerdin , Andrew Scull...
0 downloads 0 Views 98KB Size
250

Energy & Fuels 2001, 15, 250-255

Performance and Properties of a Solid Amine Sorbent for Carbon Dioxide Removal in Space Life Support Applications Sunita Satyapal, Tom Filburn,*,‡ John Trela,† and Jeremy Strange‡ United Technologies Research Center, 411 Silver Lane, East Hartford, Connecticut, and Hamilton Sundstrand Division, One Hamilton Road, Windsor Locks, Connecticut Received October 25, 2000. Revised Manuscript Received January 15, 2001

NASA is currently using a solid amine sorbent known as HSC+ for regeneratively removing CO2 in space shuttle applications. This sorbent may also be of value for CO2 removal in various industrial processes such as greenhouse gas control, industrial syntheses, and natural gas purification. To design novel sorbents and to design a CO2 scrubber based on HSC+, physical and thermochemical property data are required. In this paper, we present a detailed experimental investigation of property data and long-term performance results using HSC+ as a CO2 sorbent. Differential scanning calorimetry was used to determine the heat capacity of the material. Cyclic and equilibrium capacities of the material for CO2 pickup were determined and long-term test data show excellent performance. In addition, we have determined the heat of adsorption associated with CO2 pickup by HSC+ and the effect of moisture, using isothermal flow calorimetry. We have also performed thermal gravimetric analyses on the materials to gain insight into the stability of the material and determine the temperatures at which CO2 and constituents of HSC+ leave the surface of the material.

Introduction The primary method used to remove CO2 in space life support systems until the early 1990s was based on lithium hydroxide (LiOH).1-7 This was used, for example, in the environmental control and life support system of space suits and the space shuttle to absorb CO2 from the air. Although the storage capacity of LiOH is high (∼30 wt %), the material cannot be practically regenerated. Therefore, the long-term occupation of a space station would require a CO2 sorbent that can be easily regenerated to reduce launch weight and storage volume. Hamilton Sunstrand Space Systems International (HSSSI) has developed a regenerable sorbent consisting of solid amine beads, known as HSC+. This proprietary material consists of a liquid amine, poly* Author to whom correspondence should be addressed. † United Technologies Research Center. ‡ Hamilton Sunstrand Division. (1) Johnson, S. R.; Garrard, G. Regenerative Trace Contaminant Control: New Test Method for Effects on Solid Amine. ICES paper 921349, July 1992. (2) Johnson, S. R., et al. Multifunction Air Revitalization Systems: Combined CO2-Trace Contaminant Removal Using Solid Amines. 30th Space Congress Proceedings, Cocoa Beach, Fl, April 1993. (3) Kazemi, A. R.; Mitchell, S. M. Advanced Testing and Modeling of a Modified Solid Amine Regenerative CO2 & H2O Removal System. ICES paper 932293, p 1624. (4) Kissenger, L. D. Experimental & Analytical Techniques for Multiple Gas-Adsorbent Equilibrium (on HSC). M.S. Thesis, Rice University, May 1977. (5) Yieh, D. T. Trace Contaminant Studies of HSC Adsorbent. M.S. Thesis, December, 1978. (6) Snowdon, D. Trace Contaminant Testing of HSC+. Hamilton Standard Memo, Anal. 95-100, July 20, 1995. (7) Saiyapal, S.; Filburn, T.; Michels, H.; Graf, J. A Unique Solid Amine Sorbent Useful for Capturing Low Concentrations of Carbon Dioxide. Proceedings of the 4th International Greenhouse Gas Control Conference; submitted August 1998.

ethyleneimine (PEI), bonded to a high-surface-area, solid polymethyl methacrylate polymeric support. The material also consists of a second liquid phase coating, poly(ethylene glycol) (PEG), to enhance CO2 adsorption and desorption rates. NASA has used HSC+ in the Regenerable CO2 Removal System (RCRS) of the space shuttle, to remove CO2 from the crew compartment. In this application two equally sized beds of sorbent are used to maintain CO2 within tolerable limits (generally below 7.6 mm Hg). These beds operate in a semi-batch, Pressure Swing Adsorption (PSA) mode. One bed uses a fan to pass cabin air through the bed to remove CO2 and water vapor, while the second bed is exposed to space vacuum to regenerate the sorbent. This arrangement continues for a preset half-cycle time period of roughly 13 min at which point the bed functions alternate. This system has demonstrated its CO2 removal capability both in ground testing and while in orbit. Its capacity has been shown to exceed the space shuttle seven-person crew specification, which requires an average CO2 removal rate of 7 kg per day. As a pressure swing adsorption system it typically adsorbs CO2 at a partial pressure of 4.5 mmHg. Despite being exposed to space vacuum, restrictions in the outlet exhaust duct limit the final desorb pressure to only ∼1 mmHg. Therefore, the PSA system alternates between the 4.5 mmHg of the cabin pCO2 and 1 mmHg during the desorption mode. To design sorbents with higher capacities and favorable kinetics, a detailed study of the currently used sorbent, HSC+, is essential. In this paper, we report a

10.1021/ef0002391 CCC: $20.00 © 2001 American Chemical Society Published on Web 02/20/2001

HSC+ in CO2 Removal in Space Shuttle Applications

Energy & Fuels, Vol. 15, No. 2, 2001 251

Figure 1. RCRS simplified flight schematic. Table 1. Heat Capacity (J/g-°C) 25 °C 30 °C 35 °C 40 °C 45 °C 50 °C

HSC+

substrate

1.73 ( 12 1.74 ( 0.12 1.75 ( 0.13 1.78 ( 0.15 1.80 ( 0.15 1.81 ( 0.15

2.78 ( 10.14 2.76 ( 0.14 2.75 ( 0.14 2.75 ( 0.15 2.76 ( 0.15 2.76 ( 0.15

thorough experimental investigation of thermodynamic data for HSC+ such as heat capacities, heats of adsorption, adsorption capacity, and thermal gravimetric analyses. Due to the exothermic nature of the CO2 absorption reaction and the variability of CO2 capacity with temperature, physical property data is essential to optimize system parameters (power, weight, and volume). Our overall goal is to maximize the cyclic CO2 capacity of the system while adhering to system power, weight, and volume specifications. By reporting on fundamental property data, this sorbent may also be found applicable for other CO2 removal applications. Experimental Setup and Results Performance. Figure 1 shows a schematic of the RCRS, the system presently used for regenerative CO2 removal on-board the space shuttle. This system has been successfully used on 15 separate shuttle missions operating for over 188 days. This system occupies less than 0.32 m3 on board the orbiter and weighs less than 150 kg. While operating, it draws a meager 65 W of power. This combination of low weight, power, and volume makes it very attractive for this aerospace application. Heat Capacity Measurements. Differential Scanning Calorimetry (DSC; TA Instruments; model 2920) was used to determine the heat capacities of HSC+ and the substrate material over a range of temperatures. The heat capacities are measured by subjecting a sample of material to a linear temperature rise. Table 1 shows the heat capacity of HSC+ between 25 and 50 °C. Several sample sizes between roughly

2 and 8 mg were tested due to the small sample size and potential inhomogeneities in the material. Each data point is the average of 6 trials. The procedure involved cooling the DSC sample holder enclosure to below room temperature (∼15 °C) and purging the sample with nitrogen. A ramp rate of 5 °C per minute was used until a final temperature of roughly 60 °C was obtained. A nitrogen purge of 50 mL/min was used throughout the runs. By measuring the rate of heat flow and temperature rise at each temperature, the specific heat of the material may be calculated. Determining the heat capacity of a sapphire standard validated the procedure. As shown in Table 1, the presence of the amine and coating on the substrate appears to have an effect on the heat capacity. The substrate consists of porous polymeric beads and the heat capacity is not expected to change significantly within the given temperature range. The heat capacity of HSC+ appears to be substantially different from the heat capacity of the substrate due to the presence of a thick amine and poly(ethylene glycol) coating. A heat transfer material with low specific heat would be beneficial in transferring heat away from the material. The RCRS on the space shuttle employs such a heat transfer material. The material is a reticulated aluminum foam that has a 95% open area and therefore adds negligible weight to the system. We have shown that the cyclic performance of a dual bed system is improved by 20% when using the heat transfer material. The foam allows heat released from the exothermicity of the adsorption process to be transferred to the desorbing bed, thereby aiding in desorption. Temperature Profiles and CO2 Concentration Breakthrough Data. The adsorption capacity for CO2 and temperature increase associated with adsorption was determined using a flow apparatus. The apparatus consisted of a Pyrex tube packed with HSC+. A thermocouple was placed at the center of the bed and the concentration of CO2 was measured at the inlet and exit of the tube using an infrared detector (Horiba, model PIR-2000). Figure 2 shows a typical breakthrough curve using an initial mass of 11.4 g of HSC+, packed in a 2.5 cm diameter tube. The flow rate of CO2 (2% in N2) was varied between ∼0.5 and 2 slpm. The maximum temperature determined in the center of the bed (with no aluminum

252

Energy & Fuels, Vol. 15, No. 2, 2001

Satyapal et al.

Figure 3. TGA data for polymeric substrate.

Figure 2. Typical breakthrough curve showing CO2 concentration and temperature profiles. foam packing material) was 52.9 °C, and the maximum capacity was found to be roughly 4% by mass. An improved version of the amine has shown a ∼8% capacity for CO2 absorption but has not yet been flight qualified for use in space applications7. Molecular sieve type 5a could be expected to have a 3.5 wt % capacity in a PSA system operating with a CO2 partial pressure between 7.6 and 1 mmHg. However, this same mole sieve system would need to dry the air stream before exposing the zeolite to the gas. Thermal Gravimetric Analysis (TGA). The solid amine sorbent, HSC+, is an improved version of an amine previously developed at HSSSI which was known as HSC. That material consisted of a solid polymeric support, coated with an amine. The HSC+ version uses a poly(ethylene glycol) coating along with the amine coating to improve the cyclic capacity of the sorbent material. To determine the thermal stability of HSC+, thermal gravimetric methods (TA Instruments, model 2950) were used. TGA (thermal gravimetric analysis) data allow us to determine the range of temperatures at which CO2, the coating material, and the amine leave the surface of HSC+ under controlled conditions. Figures 3-8 show TGA data of HSC+ and its constituents. All experiments were taken under consistent conditions and typical sample sizes varied between 20 and 50 mg. The runs were initiated at room temperature (approximately 25˚C), and the temperature was increased at a ramp rate of 1 0˚C per minute. Nitrogen was used as a purge gas at a flow rate of roughly 70 mL/min, throughout all experiments. TGA data for the three individual constituents of HSC+ (substrate, poly(ethylene glycol) coating, and amine) are shown in Figures 3-5. Figure 3 shows TGA data for the polymeric substrate and it is clear that significant loss of material does not occur until ∼300 °C. The slight loss in mass between room temperature and 300 °C may be due to the evaporation of adsorbed moisture. The steep slope seen between ∼300 and ∼330 °C, and again between ∼390 and ∼420 °C, is indicative of two different processes taking place and complete decomposition is seen by ∼500 °C. As seen in Figure 4, the poly(ethylene glycol) coating material by itself, begins to lose significant mass at temperatures as low as 150 °C. The most noteworthy point is that although the coating does not have a high boiling point by itself, it enhances the stability of the overall material by enhanced intermolecular interaction with the amine. In other words, the vapor pressure of the coating by itself is not indicative of the stabilizing effect it produces when combined with the other materials. Figure 5 shows TGA data for the liquid amine without any additional constituents. As seen in Figure 5, the amine has lost only ∼10% of its original mass even at a temperature of ∼250 °C. The slope of

Figure 4. TGA data for coating.

Figure 5. TGA data for amine. the mass-loss curve changes from a somewhat gradual slope to a steep slope at roughly 300 °C. There is another abrupt change in slope at ∼350 °C, indicative of yet another change in the mechanism of thermal loss. The amine is completely disintegrated by roughly 600 °C. It is evident that the amine has an extremely low vapor pressure, unlike commonly used amines such as MEA. This property makes it suitable for coating a substrate and for longevity under temperatures slightly above ambient conditions. Figures 6 through 8 show additional TGA data that characterize the sorbent’s thermal stability. Figure 6, for example, shows the effect of combining the amine and the poly(ethylene glycol) coating. If one compares the mass loss for pure coating at 200 °C (see Figure 4) with the mass loss for coating and amine at the same temperature (see Figure 6), it is clear that there is a difference of 60% loss (for pure coating) versus ∼15% loss (for amine + coating). It is also possible that the 15% loss includes trace

HSC+ in CO2 Removal in Space Shuttle Applications

Figure 6. TGA data for mixture of coating and amine.

Energy & Fuels, Vol. 15, No. 2, 2001 253

Figure 9. Heat of adsorption for HSC+. Table 2. Mass Loss Data as a Function of Temperature temperature °(C)

Figure 7. TGA data for HSC+.

Figure 8. TGA data for HSC+ pretreated with CO2. amounts of moisture and CO2. Therefore, the TGA data show that the presence of the amine increases the thermal stability of the coating material. However, as anticipated, combining low-boiling and high-boiling components results in a mixture that usually boils at an intermediate temperature (unless an azeotrope is formed). Finally, all components are combined to form the material known as HSC+, and the thermal stability of this combined material is shown in Figure 7. The initial peak in the mass loss derivative curve (at ∼60 °C) is due to removal of trace amounts of CO2. The second peak at ∼200 °C is due to removal of both amine and coating. There is a change in the mechanism beginning at roughly 230 °C, and complete decomposition occurs at ∼450 °C. The HSC+ sample was purged using CO2, just prior to obtaining TGA data, and the results are shown in Figure 8. The maximum rate of desorption of CO2 is seen to be between 60 and 70 °C. It is clear that loss of additional components (e.g., amine or coating)

wt %

amine

coating

amine + coating

substrate

HSC*

HSC + CO2

90 80 70 60 50 40 30 20 10

229.8 264.1 285.0 299.5 315.0 329.4 349.4 475.4 545.5

157.9 173.3 182.8 189.9 195.9 201.0 205.4 209.4 213.4

188.6 205.9 219.2 237.4 269.3 295.6 328.6 364.6 515.8

307.7 315.5 320.2 326.6 354.1 398.1 408.0 412.8 455.2

177.2 208.3 283.6 330.3 364.3 391.2 410.9 423.6 437.6

165.2 200.4 269.4 323.6 359.6 388.1 409.5 422.7 437.0

do not begin to take place until over 100 °C. The material therefore appears to be thermally stable under the operating conditions we employ for space life support systems. All TGA data are summarized in Table 2. This solid amine sorbent has been used on space shuttle missions over the last several years and no degradation of material has been shown. Heats of Reaction. To configure coupled adsorption and desorption beds, the heat of adsorption during the reaction of HSC+ and CO2 must be known. We have measured the heat of adsorption using isothermal flow microcalorimetery. This highly sensitive technique is valuable for thermochemical measurements in which equilibrium is attained in a relatively short time. The technique we have used is a conductometric method as opposed to an adiabatic method. Rather than maintaining adiabatic conditions (i.e., eliminating heat flow to or from the sample cell), we maintain isothermal conditions and measure heat flow to or from the cell. Integration of the heat flow over the time period of the adsorption process provides the heat of adsorption. The instrument (CSC, model 4400) is a differential (dual cell) unit and can measure heat flows as low as 0.1 µW (25 nanocalories/s). Operating temperatures range from 0 to 100 °C with an adsorbent bed volume of approximately 3 cm3. Figure 9 shows the heat generated during the adsorption of CO2 (2% in air) on HSC+. In this example, an approximately 0.6 g sample of HSC+ was exposed to a 2% mixture of CO2 in N2, at a flow rate of 30 mL/min. The heat of adsorption was calculated to be -94 ((8) kJ/mol CO2 which is consistent with results anticipated for amine + CO2 reactions. This value represents an average for 5 sample trials. The mass % of CO2 adsorbed was 3.7 ((0.4) %. For instance primary amines have heat of adsorption values around 84 kJ/mol, secondary amines have values of about 72 kJ/mol, and tertiary amines form a much weaker bond of about 48 kJ/mol.9 One of the major sources of error is believed to be incomplete degassing to (8) Goodridge, F. Kinetic Studies in Gas-Liquid Systems. Trans. Faraday Soc. 1955, 1703. (9) Kohl, A.; Nielsen, R.; Gas Purification, 5th ed.; Gulf Publishing Co.: Houston, TX, 1997.

254

Energy & Fuels, Vol. 15, No. 2, 2001

Satyapal et al.

Table 3. Heat of Adsorption Data HSC+ 2% Adsorption in N2

average std. dev.

∆H

initial mass (g)

wt % adsorbed

J/g(mi)

KJ/g(Am)

kJ/mol CO2

0.531 0.030

3.67 0.41

78.1 6.4

2.15 0.19

94.6 8.2

HSC+ Water Adsorption in N2 ∆H

average std. dev.

initial mass

wt % adsorbed

J/g(mi)

KJ/g(Am)

kJ/mol H2O

0.546 0.057

16.8 3.6

504 157

2.62 0.06

47.2 1.0

HSC+ Water Adsorption with 2% CO2 in N2

average std. dev.

initial mass

wt % adsorbed

J/g(mi)

∆H KJ/g(∆m)

0.535 0.027

27.3 2.2

688 57

2.52 0.05

remove CO2 and moisture. The sample cell is not directly connected to a vacuum pump and therefore adsorption of trace CO2 or moisture is anticipated during sample transfer. The flow rates were varied from 5 to 30 mL/min to verify that the instrument response time was sufficient for the extent of heat flow. We also measured the heat released as a stream of humid air passed through the sample cell containing HSC+. A water adsorption accessory is available for the instrument, which allows a controlled humidity stream to pass through the sample cell. Water vapor was continuously injected into a stream of nitrogen such that the relative humidity was stabilized at 80%, and the heat released as water adsorbed onto HSC+ was determined. We also injected water vapor into a stream of 2% CO2 in N2, and measured the overall heat released due to both reaction and adsorption. All results are tabulated in Table 3. The average values represent data from either 4 or 5 sample runs. Due to the long times required for stabilizing the calorimeter temperatures, each run was up to 50 h long.

Discussion and Conclusions Measuring heats of adsorption provides important information on the interaction between CO2 and the HSC+ sorbent. We also measured the heat of adsorption for water vapor on HSC+. Our results show a ∆H of 47.2 ((1.0) kJ/mol H2O and we found that nearly 17% of the initial mass of HSC+ is absorbed water. The material is therefore an efficient dehumidification sorbent. The heat of condensation of water vapor is -44.0 kJ/mol (at 25 °C) and the ∆H we observe is slightly larger. As anticipated, a physisorptive process typically results in a ∆H similar to the heat of condensation of the adsorbed gas. Due to the potential of additional chemisorption and the fact that the coating on HSC+ is hydrophilic, we expect the heat of adsorption to be larger than the heat of condensation. The difference, however, is minor and the predominant mechanism for moisture adsorption appears to be similar to capillary condensation. The heat released during adsorption of CO2 is significantly larger than the heat of condensation of water vapor, indicative of a strong interaction between the CO2 and the amine surface. These results indicate that even though the temperature required for desorb-

ing CO2 is low, the mechanism of adsorption is not simple physisorption. To understand the effect of water vapor on the adsorption process, we also measured the heat released when a mixture of 2% CO2 in N2, at a relative humidity of 80%, is adsorbed onto HSC+. The important difference to note is the mass % adsorbed in the presence of water versus in the absence of water. In the case of pure CO2 (no water), there is only a 3.67 (( 0.41) % mass gain; in the case of pure water, there is a 16.8 (( 3.6)% mass gain; and in the case of CO2 + water, the mass gain is 27.3 (( 2.2)%. It is therefore clear that higher CO2 capabilities are achieved by coadsorption of water. The mechanism for CO2 removal using amines is known to be dependent on the presence of water.8-9 Without moisture present, the main reaction believed to account for CO2 removal is carbamate formation:

CO2 + 2R2NH ) R2NH2+ + R2NCOOThis shows that for every one mole of amine, only 1/2 a mole of CO2 is removed. However, when moisture is present, further reaction of the carbamate ion to form bicarbonate occurs:

R2NCOO- + 2H2O + CO2 ) R2NH2+ + 2HCO3Bicarbonate may also form directly from the amine + CO2 + water reaction:

CO2 + R2NH + H20 ) R2NH2+ + HCO3Therefore, in the presence of water, one mole of amine is effective in removing one mole of CO2. This mechanism has been discussed in the literature for several years.10 In our results, it appears as if roughly 3 times more CO2 is removed in the presence of water as compared to the absence of water. We did not differentiate between the mass adsorbed in terms of the molar fraction of water and CO2. Therefore, there may also be an enhanced adsorption of water in the experiments tabulated in the last section of Table 3. Another possibility is that reaction on the surface, and/or moisture adsorption, allows more of the active sites of the material to be available for CO2 removal. The fraction of CO2 removed on a molar basis as compared to the number of moles of active amine available on the surface has not been determined. We have theorized the presence of the poly(ethylene glycol) coating has two potential effects. The first may be to simply attract more water to the surface of the support due to the hygroscopic nature of the chemical. The second may be an enhancement due to the preponderance of OH- ions from the polyethlene glycol molecules. To design CO2 scrubbers for various applications, detailed performance studies of the sorbent are essential. We have performed a set of detailed experiments to determine the physical and thermochemical properties of the solid amine sorbent HSC+. The material has been used on the space shuttle for several years, and improved materials are currently under development. The material may also be of value for other CO2 removal (10) Otsubo, K., et al. International Symposium on Space Technology and Science, Proceedings, Vol. 2; 1992; pp 1431-1438, and references therein.

HSC+ in CO2 Removal in Space Shuttle Applications

applications and therefore property data such as heat capacities and heats of adsorption are necessary. The key advantage of this sorbent as opposed to membrane separations is that pressurization of the CO2rich stream is not required. The sorbent is capable of removing low concentrations of CO2 (∼1 Torr) under ambient temperatures and pressures, and is therefore less cost intensive than membrane separations. Liquid amines have been the most predominant method for CO2 removal but require substantial equipment for circulating liquids and must deal with the corrosive nature of the carbamates formed in the process. The use of solid amines provides ease of handling, applicability to microgravity environments, and regenerability in either pressure swing or temperature swing modes. As opposed to metal oxides for CO2 removal, we have shown that the solid amine may desorb CO2 at temperatures as low as ∼40 °C. We have measured the CO2 absorption capacity for HSC+ to be ∼4% at ambient pressures, and

Energy & Fuels, Vol. 15, No. 2, 2001 255

have shown that the sorbent may be regenerated using vacuum desorption at ∼1 mmHg. Thermal gravimetric analysis was used to show that the amine is strongly bonded to the substrate and the material does not begin to lose amine/coating components until over 100 °C. The material has been tested for hundreds of cycles with no loss in performance. These cycles have consisted of a period of CO2 adsorption followed by an equal time period of vacuum desorption. Vacuum levels generally range from 5 mmHg to 1 mmHg at the end of the desorption period. Acknowledgment. We thank Harvey Michels, Phil Birbara, and Joe Genovese for detailed discussions on this work. We also thank Larry Pryor for the setup of theTGA instrument. EF0002391