Preliminary Study of Molten Ammonium Salt Hydrates as Carbon Dioxide Absorbents for Environmental Control and Life Support System Applications
 

Paul A. Flowers

Department of Chemistry and Physics, University of North Carolina at Pembroke, Pembroke, North Carolina 28372-1510


Aspects of carbon dioxide absorption by molten tetramethylammonium fluoride tetrahydrate (TMAF) were examined in a preliminary study of the salt's suitability for use in environmental control and life support systems.  An apparatus was constructed that permitted control of the composition and flow rate of gaseous mixtures bubbled through molten TMAF and allowed for continuous monitoring of the sparge gas carbon dioxide concentration via an infrared flow cell.  Calibration of the apparatus showed a linear response up to 3 %(v/v) CO2, with negative deviations from linearity observed at higher pressures.  Results from initial experiments employing a 3 % CO2 sparge gas and a 1 mL/min flow rate indicated a near 100% removal efficiency under these conditions.


Introduction

    Environmental control and life support systems are required elements of any habitats or vehicles used for manned exploration of sea or space.  Among the functions of these systems is the removal of carbon dioxide produced by respiration of human inhabitants from the habitat's atmosphere in order to prevent accumulation of this metabolic waste gas to harmful levels.  Carbon dioxide removal is typically achieved by sorption methods, either chemical (chemisorption) or physical (physisorption), and a variety of strategies have been developed to date.  For example, lithium hydroxide has long been used and remains the most commonly employed chemisorption reagent, removing carbon dioxide according to the equation

2LiOH.H2O(s) + CO2(g)  =  Li2CO3(s) + 3H2O(l)                                                          (1)


 Despite the historical and continued usage of this system, the sorption reaction is difficult to reverse and LiOH scrubbers are thus essentially nonregenerable, prohibiting their use on prolonged missions.  One example of a regenerable system involves chemisorption of carbon dioxide by the amine functional groups of a polymeric resins, e.g.,

R3N(s) + H2O(s) + CO2(g)  =  R3NH+(s) + HCO3-(s)                                                          (2)


 The amine resin may be regenerated by application of heat and/or vacuum to desorb the carbon dioxide via reverse of equation (2) above.  Finally, an example of one promising physisorption strategy involves the use of zeolites ("molecular sieves") as adsorptive traps for carbon dioxide.  These systems exploit the high surface area and adsorptive properties of the zeolites, and are regenerated by a combination of heat and vacuum.  A general summary of carbon dioxide removal strategies and a survey of related primary literature may be found in references [ 1 ]and [ 2 ] of this report.
    Previous reports by Quinn et al. [ 3 ] have shown that certain molten quaternary ammonium salt hydrates possess interesting CO2 absorption properties, including large absorption capacities and rapid desorption upon either solidification or application of vacuum.  Recent work in the author's laboratory has extended Quinn's investigations to examine the suitability of these molten salt hydrates as solvents for the electrochemical reduction of carbon dioxide[ 4 , 5 ].  Results from these studies indicate carbon dioxide is initially absorbed by these media via a chemical reaction with the salt's basic anion, e.g., for the case of tetramethylammonium fluoride tetrahydrate (TMAF),

2F- .nH2O(l) + CO2(g)  =  HCO3-(l)  +  HF2- .(2n-1)H2O(l)                                              (3)


Upon completion of this chemical reaction, carbon dioxide continues to be absorbed via simple, physical dissolution with Henry's law constant on the order of 0.1 M/atm.  These results have motivated the preliminary study of the suitability of TMAF as a carbon dioxide sorbent for environmental control and life support system applications described herein.

Experimental

    Reagents.  Tetramethylammonium fluoride tetrahydrate (98%, Fluka), nitrogen (standard grade, Dunn Machine & Welding Co.) and carbon dioxide (SFC-grade, Air Products) were used as received from the vendors.

    Apparatus.  A block diagram describing the experimental setup is shown in Figure 1.  A compressed gas cylinder of carbon dioxide was connected to the inlet of an Isco 260D syringe pump whose outlet was fed to a Rheodyne XXXX loop injector.  The sample loop ports of the injector were connected to a small scrubber assembly (see detail in Figure 2) containing the TMAF absorbent.  The scrubber assembly was fabricated from a 7 mm fritted tube connector (porosity C, Ace Glass, Inc.) and included 7 mm rubber septa (Agilent) to accept gas inlet and outlet tubes and an additional septum with a short segment of tubing placed near the outlet to serve as a foaming baffle. All gas connections between the carbon dioxide tank and the loop injector were made using either 1/8" or 1/16" stainless steel tubing.  The loop injector's outlet port was connected to an infrared flow cell (BioRad Model GC/C 32) using 1/16" steel tubing (exiting the injector) sealed with epoxy (Torr Seal) to 0.3 mm flexible glass capillary tubing (Agilent) to accommodate the flow cell's inlet port.  The flow cell was interfaced to a BioRad FTS40 FTIR spectrometer.


Figure 1.  Block diagram of the experimental apparatus used for carbon dioxide absorption experiments.  Solid lines represent steel or glass tubing connections.

Figure 2.  Detailed illustration of the scrubber assembly depicted in Figure 1.  Gas flow is from bottom-to-top of the scrubber as shown.

    Procedure.  Calibration of the apparatus was performed by first flushing with nitrogen gas with the loop injector in the "load" position (bypassing the scrubber assembly) and measuring a reference spectrum.  Next, the syringe pump was filled to capacity (ca. 260 mL) with a gaseous mixture of the desired composition by introducing appropriate volumes of nitrogen and carbon dioxide and waiting ca. 5 min to permit thorough diffusional mixing.  The apparatus was then flushed with this mixture for several minutes prior to measuring an absorbance spectrum versus the nitrogen reference spectrum.
    For CO2 sorption experiments, the scrubber assembly was loaded with a weighed quantity of TMAF and connected to the gas inlet and outlet tubes of the loop injector.  This operation was performed quickly in order to minimize absorption of water by the hygroscopic salt.  The salt was then gently heated with a heat gun until molten (melting point ca. 50 oC).  The insulating properties of the glass scrubber tube were adequate to keep the salt molten until it was exposed to CO2, whose dissolution depressed the salt's melting point to subambient values [ 3 ].  With the loop injector set to the "load" position (i.e., bypassing the scrubber assembly), the system was purged with nitrogen gas and a reference spectrum was acquired.  A CO2 / N2 mixture of desired composition was then prepared as described above and pumped through the system at the desired flow rate.  Acquisition of infrared absorbance spectra using the spectrometer data station's kinetic software was begun, and the loop injector was switched to the "inject" position (passing the gaseous mixture through the scrubber) after ca. 30 min.

Results and Discussion

    Calibration of the apparatus depicted in Figure 1 above yielded the plot shown in Figure 2 below.  A linear relation between peak absorbance at 2350 cm-1 and carbon dioxide partial pressure was observed up to ca. 0.03 atm (3 %CO2).  A negative deviation from linearity was observed at higher pressures that may be a result of pressure broadening of the spectral feature.  Regression analysis of the calibration data yielded a calibration equation of A = -3870 atm-2 P2  +  41.2 atm-1 P  +  0.005, where A is absorbance at 2350 cm-1 and P is the partial pressure of carbon dioxide.  This equation was used to convert absorbances measured during the CO2 sorption experiments described below.  Baseline noise levels in these spectra were on the order of 2 mau, corresponding to a detection limit (for S/N > 3) of ca. 0.0002 atm (0.02 %CO2).


Figure 3.  Calibration curve for carbon dioxide (absorbance at 2350 cm-1 versus partial pressure of CO2).
    Shown in Figure 4 are results from a sorption experiment employing a roughly 3 %CO2 gaseous mixture and a flow rate of 1 mL/min.  After monitoring the carbon dioxide pressure for ca. 30 min with the loop injector set to bypass the scrubber assembly, the injector valve was switched to redirect flow through the scrubber and a rapid decrease in CO2 pressure to nearly undetectable levels was observed.  Based on the CO2 flow rate, the mass of TMAF in the scrubber, and the reaction stoichiometry of equation (3) above, an estimated time of ca. 340 min would be required to chemically saturate the sorbent.  An unexpected foaming of the molten TMAF occurred approximately 2 h into this experimental run, resulting in blockage of the infrared flow cell's transfer lines and hence preventing further measurements.  The flow cell has been returned to the vendor for repairs.

Figure 4.  Variation of CO2 partial pressure with time.  Sparging at 1 mL/min begun at ca. 30 minutes.

Future Work

    Though quite limited in scope, the data obtained in this preliminary investigation are promising.  Continued pursuit of this project should involve the following:

a) redesign of the foaming baffle (see Figure 2) to prevent recurrence of the foam-over incident described above;
b) sorption studies at varying flow rates and sparge gas compositions (including controlled moisture levels);
c) similar studies of the desorption of saturated sorbent;
d) studies employing different salt hydrates; and,
e) studies employing different physical forms of salt hydrate (e.g., supported films).


Acknowledgments

    The author gratefully acknowledges financial support of this work by the North Carolina Space Grant Consortium.

Literature Cited

1.  "The Regenerative Life Support World Wide Web Site",  J.E. Atwater, UMPQUA Research Company, 1996:  http://ucs.orst.edu/~atwaterj/LifeSupport.html

2.  Literature survey results by the author available through this link.

3.  "Salt Hydrates:  New Reversible Absorbents for Carbon Dioxide", R. Quinn, J.B. Appleby, G.P. Pez, J. Am. Chem. Soc. 1995, 117, 329-335.

4.  "Dissolution and Electroreduction of Carbon Dioxide in Molten Ammonium Salt Hydrates", P.A. Flowers, 44th Annual Report on Research under Sponsorship of The Petroleum Research Fund, 2000, 328-329.

5.   "Dissolution and Electroreduction of Carbon Dioxide in Molten Ammonium Salt Hydrates", P.A. Flowers, 45th Annual Report on Research under Sponsorship of The Petroleum Research Fund, in press.