difference ApKdE* of uncharged acids is of the same order of magnitude in protophobic A N and protophylic aprotic solvents D M F and DMSO, and quite different from that in water or methanol. No constant differences in pKdgn+ in the three aprotic solvents are found for cationic acids. For example, for aniline ( A P K ~ ~ ~ + ) * ~=D6.2, M F for n-butylamine, 9.4. Also for n-butylamine (ApKdBH+)DMFnM~o = f2.0, for ammonia, - 1.0. The conclusion drawn by Paul et al. (24) from condutance data that amines are present as dimers in D M F is incorrect. The conductivities reported by them are due to impurities, but the free bases are too weak to yield a measurable conductance.
Explanations for simularities and differences in pKdaoid values in different solvents will be discussed in a subsequent paper in connection with the distribution coefficients of the components of the acid-base systems. RECE~VED for review July 10, 1970. Accepted August 28, 1970. This work was supported by the Directorate of Chemical Sciences, Air Force Office of Scientific Research, under Grant AF-AFOSR-1223-67. One of US (H. S.) expresses his gratitude to the Kosciuszko Foundation, New York, N. Y.,for a grant which allowed him to work for a year at the University of Minnesota.
An Infrared Method for Rapid Analysis of the Sulfate Content of Reacted Lime and Limestone Materials Edwin F. Rissmann General Technologies Corporation, A Subsidiary of Cities Service Company, 1821 Michael Faraday Drive, Reston, Va. 22070 Robert L. Larkin National Air Pollution Control Administration, Department of Health, Education, and Welfare, 3914 Virginia Avenue, Cincinnati, Ohio 45227 A thin film infrared cell technique has been developed employing an aqueous system for the rapid analysis of the sulfate content of calcined limestones which have been reacted with SO, in a flue gas. Accurate analyses can be reproducibly conducted by dissolving the reacted limestone material in tetrasodium EDTA (ethylenediaminetetraacetic acid) saturated aqueous solutions and obtaining spectra of those in a 0.003 m m infrared liquid cell. For the infrared determinations, an identical cell, containing a saturated EDTA aqueous solution is placed in the reference beam of the spectrometer.. Results obtained are in good agreement with determinations made independently by different techniques. It has been demonstrated that similar analysis can also be conducted for the carbonate and hydroxide contents of limestone with the use of EDTA saturated D,O as the solvent. The general approach should be useful for many solids analyses
POLLUTION OF THE ATMOSPHERE by the sulfur dioxide formed during the combustion of fossil fuels has caused a rapid growth in research and development programs to devise effective control technology. One of many approaches being investigated for removing SO, from stack gases in the gas phase reaction with solid absorbents which can either be regenerated and recycled or removed from the system. Substantial mechanistic studies of the reactions occurring with a majority of the solid absorbents reveal alkaline earth sulfates as the primary product of reaction with SOZ. Evaluation and interpretation of the efficiency attained by a particular process depends very heavily upon an accurate and convenient method for determining the amount of SO, uptake by the solid absorbent. In addition to the low concentrations involved and the interferences from metal ions, analysis of these reaction products is further complicated by low solubilities and very slow dissolution rates resulting from exposure of the materials to the high temperature combustion conditions in the boiler. 1628
For the dry limestone injection processes as well as wet limestone control technology, a clear need exists for the capability to rapidly analyze reacted limestone materials for their sulfate content and on many occasions, carbonate, hydroxide, and silicate are of equal significance. The chemical methods of analysis (Le., turbidometric, colorimetric, and titrimetric methods) are too time consuming when analyses are to be made for several constituents. Thus, other analytical methods, such as those based on infrared technology, must be developed which have an inherent capability for simultaneous analyses for more than one component. One such method is to dissolve the limestone in a solvent and determine the sulfate content of the resulting solution by infrared quantitative analysis. After a considerable search, in which a number of organic solutions and solvent mixtures were investigated, tetrasodium EDTA (ethylenediaminetetraacetic acid) saturated aqueous solution was found to be the only solvent to satisfactorily dissolve all of the limestone constituents of interest without decomposition. Water, in cells of less than 0.01 mm in thickness, is known to transmit over most of the infrared region, with the exception of two areas (Le., 3500-3000 cm-1 and 1700-1500-1) ( I ) where intense absorption occurs. The dissolved EDTA also possesses a number of absorption regions, but these bands are less intense than those of water. The main problem encountered in the use of aqueous solutions is, thus, the requirement that very thin cells be used. To date, water has been employed as a solvent in several infrared investigations, most of which, because of difficulties with thin cells, have been of essentially a qualitative nature
“Handbook of Industrial Infrared Analysis,” Plenum Press, New York, N. Y., 1964, p 107.
(1) R. G. White,
ANALYTICAL CHEMISTRY, VOL. 42, NO. 13, NOVEMBER 1970
~
(
I
1
” I I .003 mm Film of Silver Deposited onto Outer Portions Only of Bottom AgCl Plate
I
////A
I I
I ,
Top AgCl Plate
/ Bottom A Plate
I
s1
I
Table I. Spectra of Insoluble Limestone Residues Band position, Sample (cm-1) Strength Assignment number 3650 very weak OH109 1400 very weak CO3 1240 very weak ? 1035 strong Si-0 970 strong Si-0 880 strong Si-0 48C-570 strong Si-0 1400 very weak CO3 107 1025 strong Si-0 1250 weak ? 970 strong Si-0 880 strong Si-0 490 medium Si-0 1420 very weak CO3’105 1250 very weak ? 1080 strong Si-0 1025 strong Si-0 935 strong Si-0 800 strong Si-0 475 strong Si-0 1420 very weak cos 110 1080 weak Si-0 lo00 weak Si-0 880 weak Si-0 480 weak Si-0 1440 weak co3 104 1240 very weak ? 1030 strong Si-0 950 weak Si-0 880 medium Si-0 480 medium Si-0 no residue-sample entirely dissolved 108 no residue-sample entirely dissolved 106 1410 very weak Coal103 1255 very weak ? 995 strong Si-0 995 weak Si-0 900 strong Si-0 840 weak Si-0 770 weak ? 500 weak Si-0 1420 very weak cos’102 io50 strong Si-0 940 strong Si-0 870 strong Si-0 830 weak Si-0 765 weak ? 500 strong Si-0 no residue-sample entirely dissolved 101
’-
Figure 1. Infrared cell for use with limestone solvent (2-5). However, recent advances in cell technologies have made it possible to design cells of a few microns in thickness which can be used for quantitative studies (6, 7). This paper presents an analytical technique for sulfate determination using thin (0.003 mm) cells and EDTA saturated aqueous solutions. EXPERIMENTAL
Quantitative determinations for sulfate have been conducted by the infrared method using cells such as those shown in Figure 1. The cells consisted of two optically flat silver chloride plates. On the outer portions of the bottom plate was deposited a 0.003-mm thick silver film which served as a cell spacer. The other plate had two small holes drilled through it for admission of sample. This assembly was of the proper size that it could be used in conjunction with commercial cell holders. For this analytical study, a pair of such cells were used; one containing a saturated aqueous solution of tetrasodium EDTA which was placed in the spectrometer reference beam, and the other containing a solution of calcium sulfate or dissolved limestone in this solvent system which was used in the sample beam. All spectra were run on a Beckman IR-10 infrared spectrometer. The saturated tetrasodium EDTA solutions used were prepared by dissolving this material in distilled water to saturation and storing the resulting solutions over a n excess of the EDTA salt. Solutions of either calcium sulfate or limestone were prepared by dissolving a known amount of the material of interest in a specified volume of solvent. In the case of some of the limestone samples, difficulties were encountered in dissolving all of the materials. In some cases, a n insoluble residue remained. This is in accord with the results of Hill and Goebel (8) who found that limestone constituents such as silica could not be dissolved by EDTA. The residues were collected by filtration, washed free of EDTA solution with distilled water, dried, and then infrared spectra were examined by the KBr pellet technique. The results, as shown in Table 1, revealed that they contained n o sulfate but large amounts of silicates. Calibration curves for sulfate absorbance were obtained from the data on a number of sulfate solutions using the dual cell method described above. In the data analysis, the absorbances a t the sulfate band maximum (1 110 cm-l) were measured assuming the background over the sulfate band region was linear. Studies using the dual cells, with only saturated EDTA present in each cell have shown this assumption to be valid. For the calibration studies, a plot was then made of absorbance us,
’-
’’-
(2) R. N. Jones and C. Sandorfy, “Infrared and Raman Spectra Applications in Chemical Applications of Spectroscopy,” W. West, Ed., Interscience, New York, N. Y., 1959, p 246. (3) F. C. Nachod and C. M. Martin, Appl. Spectrosc., 13,45 (1959). (4) H. Sternglanz, ibid., 10, 77 (1956). (5) S. D. Kulbom and H. F. Smith, ANAL.CHEM., 35, 912 (1963). (6) W. K. Thompson, Trans. Faraday Soc., 61, 2635 (1965). (7) W. A. Senior and R. E. Vernall, J. Phys. Chem., 73, 4242
concentration for the CaS04 samples studied. This is shown in Figure 2. As the same cell thicknesses were employed each time, no corrections had to be made to convert the absorbance data to absorbances per unit thickness. The analytical results (Table II), are expressed as per cent SO,. These numbers were arrived at assuming all of the sulfate in the reacted limestone samples was present as CaS04. Such a n assumption agrees well with the findings of Borgwardt (9) in studies of the reactions of calcined limestone with flue gases. Wet chemical analysis of the limestones studied was conducted t o corroborate the infrared results. Specifically, the limestone samples were ground to a fine powder after drying
(1969). (8) W. E. Hill and E. D. Goebel, Bull. 165, Part 7, State Geological Survey of Kansas, 1963.
(9) R. H. Borgwardt, Enuiron. Sei. Technol., 4, 59 (1970).
ANALYTICAL CHEMISTRY, VOL. 42, NO. 13, NOVEMBER 1970
1629
Table 11. Analysis of Known Reacted Limestone Sample for Sulfate Weight
SO3
Sample number
by infrared
Average
by wet chemical
101
17.63
17.52
16.5
1.06
17.40 17.11
17.78
16.5
1.08
41.03
39.7
1 035
38.94
37.5
1.04
45.88
44.64 44.42 27.04 27.22 42.75 42.82 30.34 30.52 10.50 10.40 28,46 28.30 16.41 16.69 16.48
1.03
SOp
(Solution I*) 101 101
(Solution IIb)
0.4
0
0.8
1.2
1.6
2.0
Concentration (g C a S O ~ t O O c c )
Figure 2. Calibration curve with 0.003-mm cell for sulfate ion a t 1110 cm-' 1
I
I; 0.1
I!! 2.0
1 1750
I
I
1600
1400
1200
IC00
800
Ratio infrared/ wet chemical
101 102 102 103 103 104 104 105 105 106 106 107 107 108 108 109 109 110 110 110
18.45 40.30 41.77 37.76 40.12 47.26 44.50 28.08 27.61 47.12 46.29 27.88 29.20 11.56 11.50 30.45 29.86 18.37 17.61 17.39
27.84 46.7 28.54 11.53 30.15 17.79
I
1.03
1.09 .95 1.10 1.06 1.08
Calculations to arrive at this number assume all sulfate present as CaS04. Solutions I and JI of sample 101 were different concentrations of the same limestone material in saturated tetrasodium EDTA, The second solution, which exhibited the higher pH, was the more concentrated. drogen ions. This releases the sulfate to solutions as H2S04. Other slightly soluble sulfates are made soluble in the same manner. Simultaneously, the resin renders the solution virtually free of cations which might otherwise interfere with the final titrimetric determination. The solution was filtered through a wad of glass wool into a volumetric flask with washings from the glass wool. A suitable aliquot was taken, made t o 80% with isopropyl alcohol and titrated with 0.005N Ba(C104)susing Thorin indicator (11). As can be seen, good agreement exists for all samples. Duplicate determinations for each of the samples further show that method reproducibility is quite satisfactory. Some exploratory work was also conducted for analysis of the carbonate and dissolvable silicate contents of limestone using the same techniques and brief studies were also made, using D20 based solvents, of the feasibility of determining hydroxides in limestone.
Wavenumber cm-1
Figure 3. Spectra of EDTA solution alone and EDTA solution containing dissolved reacted limestone in a n oven for 1 hour to remove physically absorbed water. An accurately weighed sample of 0.2-0.3 g was heated in a mixture of distilled water and cation exchange resin, Rexyn 101 hydrogen form action exchange resin, (-20 g) for 1 hour. It may be noted that no sulfuric acid was introduced into the solutions by this procedure either from resin decomposition or dissolution. The ion-exchange resin (10) serves two purposes in this procedure. The first is that the equilibrum of Cas04 is shifted toward the right by continually removing calcium ions from the solution and replacing them with hy(10) H. N. S. Schafer, ANAL.CHEM., 35, 53 (1963).
1630
RESULTS
Sulfate Analyses. Calibrations for sulfate were made using solutions of calcium sulfate dissolved in the saturated aqueous tetrasodium EDTA solvent. In these experiments, the test solutions were placed in the cell in the spectrometer sample beam. In the reference beam was placed a duplicate cell containing the saturated EDTA solution, which acted to virtually eliminate most of the spectrum due to water and EDTA. This technique was completely successful for the sulfate calibrations as can be seen in Figure 2, which shows little scatter and good reproducibility using the 1110 cm-l sulfate band. (11) J. S. Fritz and S. S. Yamamura, ANAL.CHEM.,27, 1461 (1955).
ANALYTICAL CHEMISTRY, VOL. 42, NO. 13, NOVEMBER 1970
\
0
+0.4 '
0
.-2-
-.2
