Sulphur dioxide sorption ash mixtures John
B. Green*
and Stanley
by humic acid-fly
E. Manahan
Department of Chemistry, 123 Chemistry 65211 USA (Received 15 October 1980)
Bldg., University
of Missouri,
Columbia,
MO
The sulphur dioxide absorption capabilities of aqueous humic acid-fly ash mixtures were investigated. Humic acid (HA) promotes the dissolution of basic material in fly ash, resulting in the formation of humate salts. These salts absorb SO2 effectively via an acid-base reaction resulting in formation of metal bisulphites and HA as products. The extent of fly ash dissolution in HA is a function of the HA/fly ash ratio, with greater dissolution occurring as the ratio increases, until an excess of HA is present. The utilization of base from fly ash for neutralization of SO2 is a function of its dissolution, and hence is also dependent on the HA/fly ash ratio. Besides neutralization to bisulphite, HA-fly ash mixtures can also However, the importance of this reaction is absorb SO, through a specific HA-S02C,qJ interaction. relatively small. On the basis of this work, it can safely be predicted that HA-fly ash mixtures would greatly outperform water-fly ash slurries in stack gas scrubbing applications.
Fly ash (pulverized fuel ash) derived from coal and especially from lignite frequently contains appreciable base, in the form of alkali metal and alkali earth oxides, which could potentially be used to scrub sulphur oxides from stack gases produced during coal combustion. However, the success of previous attempts at sulphur oxide removal with aqueous fly ash suspensions has been limited by the slight solubility of the ash in water’,‘. The authors have attempted to circumvent this problem via prior dissolution of fly ash in solutions of humic acid (HA). The effectiveness of HA in dissolving fly ash has been shown3x4. Briefly, HA treatment of fly ash causes release of base via dissolution of slightly-soluble alkali metal and alkali earth containing glasses and, at high HA/fly ash ratios, release of basic material entrapped in iron and aluminum matrices. Humate salts thus formed are shown here to be sufficiently basic to be effective in sulphur dioxide removal. The overall process is summarized in equations (l)(3): M,O,
+ ~xHAF-?~[M(~~‘“)+ -(HA),,,]
basic oxide
humic
in
proton source
fly ash
+xH,O
(1)
metal humate
acid, a
ri[M(2X’“)+ -(HA),,,] nM$;‘“)+
+2xS0,,,,,+2xH20ti
+ 2xHSO,,,,+
2xHA
(2)
Net reaction:
M,O, + 2xSO,,,,,+
xH,O~+2xHSO;~,,+
nM{:;‘@+ (3)
Thus, as shown in equation (3) the net reaction neutralization of sulphur dioxide to bisulphite * Present address: Bartlesville Energy Technology of Energy, Bartlesville, OK 74003 USA 0016-2361/81/04033Ck05$2.00 @ 1981 IPC Business Press
330
FUEL,
1981,
Vol 60, April
Center,
is simply ion. The U.S. Dept.
HA ‘catalyses’ the reaction by converting base contained in fly ash to a (humate) form which reacts rapidly with so,. Similar studies with sodium humates’ have shown that a ‘complex’ is formed between HA and aqueous sulphur dioxide. The effect of the di- and trivalent metal ions extracted from fly ash on this interaction was investigated. Finally, the per cent utilization of available base in fly ash was studied as a function the HA/fly ash ratio. Because, as previously mentioned, high HA/fly ash ratios favour enhanced release of basic constituents, greater utilizations of fly ash base would be expected under those conditions. EXPERIMENTAL Materials
The fly ash was obtained from cyclone dust collectors in the Basin Electric Coop power plant, Stanton, ND, October 17, 1973 (US Bur. Mines sample GF 73-1921). Portions of the sample were oven dried at 150°C overnight prior to use. Elemental analysis of the sample6,’ showed it to contain relatively high concentrations of alkali metal and alkali earth oxides typical of lignitederived fly ashes. Solubility studies have indicated that the majority of these basic oxides are insoluble in water, but dissolve readily when small amounts of dilute acid are added to neutralize excess hydroxide ion6,‘. The balance ( =S35%) of the basic material is entrapped in Fe, Al, and Si matrices. Treatment with excess dilute acid sufficient for dissolution of Al and non-magnetic Fe matrices results in the release of ~93% of the total basic material. Release of the remaining base in the refractory Fe,O, and SiO, residue requires more rigorous conditions. The total available base (excluding that in Fe,O, or SiO,) was determined to be 11.7f 0.1 meq OH-/g. The HA was obtained from Illinois No. 6 bituminous coal via nitric acid oxidation followed by extraction with sodium hydroxide using a procedure described pre-
SO, sorption Tab/e 1 Composition
of humic acid-fly
tNa+l
meq H+/ HA
Ash
0.025
no.
(ml)
(g)
Ash
PH
1 2 3 4
0 1000 1000 1000
5.0484 4.1077 1.7128 1.2093
0 0.150 0.361 0.511
11.53 6.98 5.66 4.88
Table 2
cation concentrations
[ Na+l
a Normalization
[Ca2+I
[Mg2+1
[ Fe3+I
[A13+1
[so:-1
0.04 17.5 38.5 29.0
3.07 19.2 50.2 40.3
95 135 63 45
fppm)
digit, whichever
for comparison
tK+l
2.5 3.7 3.9 3.2
56 60 46 37
128 378 240 170
0.1 82.2 46.4 32.8
is higher
with Figure 2a
tca2+1
Solution no. 1 2 3 4
fK+l
g
are +l% or k 1 in the least significant
Normalized
ash mixtures: J. B. Green and S. E. Manahan
ash mixturesa
Solution
a Uncertainties
by HA/fly
tMg2+l
[ Fe3+1
[ AIs+
[so:-1
0.008 4.3 22.5 24.0
0.61 4.7 29.3 33.3
18.8 33.0 36.8 36.8
(ppm) 11.2 14.7 27.1 30.9 is based on a 1000
0.50 0.90 2.3 2.6
25.4 92 140 141
0.02 20.0 27.1 27.1
ppm ash suspension
viouslys. The crude product was purified by extensive treatment with ion exchange resins394. Characterization of this ‘deionized’ HA solution included analysis for soluble Na. K, Ca, Mg, Fe, and Al; and determination of the total acidity (0.0247 meq H+/ml) and gravimetric concentration (7.6OkO.03 mg m1-‘)6. IIse of previously studied HA and fly ash allowed preparation of HA-fly ash mixtures of known acid/base ratio, and also greatly facilitated interpretation of the results from sulphur dioxide absorption experiments. Procedures Prepuration and analysis of’ the stock HA-fly ush mixtures. Four fly ash mixtures were prepared by weighing the appropriate amount into 1 1 volumetric flasks and diluting with either stock deionized HA or deionized water. The amount of fly ash weighed into the three flasks containing HA was varied to obtain HA/fly ash ratios effecting different degrees of fly ash dissolution. The solutions were equilibrated for one month, with periodic stirring, to ensure equilibrium conditions. Then, undissolved ash was separated from aliquots from each flask, and each aliquot was digested in nitric acid and analysed for cations and sulphate. Procedures for ash separation, HNO, digestion, and atomic absorption spectroscopic analysis have been described e1sewhere3,4. Finally, the alkalinities of all solutions were determined by titration with standard acid using a procedure used for determining alkalinities of sodium humate solutions’. Sulphur dioxide absorption studies. The same basic procedures and apparatus described for SO, absorption studies with sodium humates5 were used to determine absorption capacities and efficiencies of the fly ash mixtures. However, experiments with sodium humates designed specifically to show reversibility of absorption and desorption of SO, or to confirm the existence of a non-acid-base HA-SO, interaction were not repeated with the fly ash preparations, for the conclusions from those experiments could reasonably be expected to hold for HA-fly ash mixtures also. All experiments with fly ash solutions were performed at 23°C using nitrogen carrier gas at 2 1 min-‘.
RESULTS Table I shows the amounts of fly ash and HA used in preparation, the pH, and the results from analysis for cations and sulphate for each of the four solutions. The cation data are corrected for cations originally present in the HA. Also shown in the Table are the HA/fly ash ratios expressed as the milliequivalents acidic groups in HA (meq Hf) per 25 mg fly ash. These ratios are given to facilitate comparison with a series of cation solubility curves derived in an earlier study of fly ash dissolution in HA’. On the basis of that study, the solutions containing HA (nos. 24) were prepared to obtain meq H+/0.025 g ash ratios corresponding to important transitions in fly ash solubility behaviour in HA. For example, it was shown that only the more soluble, basic constituents dissolved in HA solutions with a HA/fly ash ratio of 0.150 (solution 2). However, a significant portion of the iron and aluminum oxides in fly ash dissolved at meq H+/0.025 g ash ratios near 0.36 (solution 3). Finally, addition of HA to obtain a HA/ash ratio of 0.51 meq H+/0.025 g (solution 4) resulted in an excess of HA and only a slightly greater degree of fly ash dissolution in this work. Although the preceding statements may be deduced from Table I, they can be seen more clearly when the raw data are normalized as in Table 2. The normalized cation and SOiconcentrations shown in Table 2 may be interpreted to be the same concentrations as those obtained had exactly 1.0 g fly ash been added to exactly 1.0 1 water containing sufficient HA to provide the desired HA/ash ratio. Tuble 3 summarizes the important quantitative aspects of the sulphur dioxide absorption experiments. Although much of the Table is self-explanatory, some additional interpretation is necessary. Alkalinities of the stock solutions were determined by two different methods. The first method, based on results from cation and sulphate determinations shown in Tables I and 2, is explained in an earlier publication’. The method is based on a mass balance type of calculation assuming that all cations obtained from dissolution of the ash were originally present in the oxideor sulphate-form in the ash. The validity of that assum-
FUEL, 1981, Vol 60, April
331
SO, sorption
by HAlfly
Tab/e 3
Summary
Solution
no.
meq H+/0.025
ash mixtures.. J. B. Green and S. E. Manahan
of alkalinity
and SO2 sorption
dataa 1
g ash
2
3
4
0
0.150
0.361
0.511
0.0072
0.0255
0.0167
0.0120
0.0072 0.0077
0.0195 0.0228
0.0115 0.0146
0.0667 0.0092
la) to pH 3.00
0.0076
0.0202
0.0083
(b) to pH 2.50
-
-
0.0115 f 0.0002 0.0146 f 0.0003
Alkalinities (1) (2)
Calc. from cation -SO:-anal. Det. by HCI titn. (a) to pH 3.00 (b) to pH 2.50
fmeq ml-t)
(meq ml-1l (meq ml-t)
Mmole SO2 Absorbed/ml
SO, absorption
efficiencies
1%)
(a) pH 3.00 (b) pH 2.56 Total base availableb % Base utilized
99.7 from ash (meq ml-l)
for absorption
0.0591
lmeq ml-‘1
42 -
-
a Relative standard deviation = il% b Available base = 11.7 meq g-l
10
20
ml stock humtc
0.0247
-
103
30 ocld -fly
40 ash/ 250
50
60
ml
ption was shown for the sample of fly ash used here’. The second method, titration with standard acid to a specified pH, is described in the Experimental section. The alkalinity given by method 1 is the total alkalinity for each solution. Thus, comparison of the alkalinities
FUEL,
1981,
99.7 96.9
i 0.2 i 0.3
0.0200
99.7
i 0.3
0.0141
58 73 0.0247 68
59 0.0247 49
unless noted otherwise
Figure 7 Sulphur dioxide absorption versus humic acid-flv ash concentration for solutions: A, 3; 6,2; C, 3; D, 4; E, 1. l, pH, 3.00; A, pH, 2.50
332
f 0.3
0.0481
13 -
% HA acidic groups reacting with fly ash
0
99.3
of SO2
(a) to pH 3.00 (b) to pH 2.50 HA acidity
f 0.2
Vol 60, April
from method 2 with that from 1 shows the extent of H’ exchange at pH 3.00 and pH 2.50 of metal ions existing as humate salts. Because no HA was present in solution 1, the results from the two methods agreed. (The slightly higher value obtained with method 2 at pH 2.50 was due to reaction of HCl with solid ash particles incompletely removed prior to titration. This reaction was sufftciently slow at pH 3.00 so at to cause no error in that value.) Data for solutions containing HA show incomplete exchange even at pH 2.50, thus indicating the tendency of HA to bind significant amounts of metal ions even under quite acidic conditionsg. Finally, it should be noted that solution 1 (no HA), which contained the largest amount of fly ash (Table I), also had the lowest alkalinity. The sulphur dioxide absorption capacities generally agreed, within experimental error, with the alkalinities from method 2, in accordance with equations (lH3). The large discrepancy between results for solution 4 will be discussed later. Figure 1 shows the linearity of absorption capacity uersus concentrations of stock solution (ml/250 ml total volume) for all solutions. Absorption capacities were obtained from slopes of plots similar to those in Figure 1. The absorption efficiencies obtained, essentially the same for all solutions, were largely a function of the apparatus used and probably do not accurately reflect obtained under stack gas scrubbing efficiencies conditions. The total available base data were calculated from the amount of fly ash added to each solution, and the base content (see Experimental) of the ash, and when divided into the SO, absorption capacities yielded the base utilization results shown. The superiority in base utilization exhibited by solutions containing HA is clear. Also, base utilization data for solution 3 indicate enhanced utilization whenever concentrations of SO, in the stack gas are high enough to lower the pH of the scrubbing medium below pH 3. A previous calculation indicated that pH 2.50 can be attained with SO, con-
SO, sorption 11
9
8-
7-
\ 6-
DyY&I&_
2I
I
I
1
I
I
I
I
I
Relative SOa absorption Figure2 Variation of the pH of humic acid-fly ash mixtures l-4 with absorption of sulphur dioxide. (Curves offset progressively to the right for clarity; [HA-FAI = 50 ml stock/250 ml; the horizontal bars indicate relative absorption capacity). Solutions: A, 1; 6, 2; C.3.D.4
centrations in the range 30~1000 ppm, depending on temperature and other factors5. The per cent of HA reacting with basic materials in fly ash was computed as the ratio of the total alkalinity (i.e., from method 1) and the HA acidity. Acidic groups not reacting with basic oxides either reacted with Al or Fe oxides (see Tables 1 and 2) or remained protonated. The data shown in Table 3 illustrate the inverse dependence of the reaction of HA groups with basic material on the HA fly ash ratio. Finally, Figure 2 summarizes, in a qualitative manner, the most significant outcome of this study. The curves shown are traces of recorder plots of pH versus SO, exposure time. The initial pH of the solution containing no HA (no. 1) is much higher than the ones with HA; yet, the sorption capacity (indicated by the horizontal bar) is relatively low. However, solution 4, with the lowest initial pH and highest HA/ash ratio, has substantial SO, capacity even’ though it contains slightly more than 25”:) of the amount of ash in solution 1. Thus, HA can significantly improve the performance of fly ash, or other slightly-soluble basic material, in applications requiring neutralization of acidic substances. DISCUSSION Ahscwption mechanisms
The major absorption mechanism has been shown to be acid-base5: Ka-HA(,,,
+ H,O + SOztg,*
absorption
HA+SO
\
3-
ash mixtures: J. B. Green and S. E. Manahan
with additional of so?:
21
10
I,
by HA/fly
for sodium
humates
YA( + Na,&, + HSO,,,, (4)
resulting
from ‘complexation’
2(ay)+HA~~~SOz
(5)
The close agreement between titration alkalinities (method 2) and sulphur dioxide absorption data in Table 3 indicates that HA-fly ash mixtures also absorb SO, mostly via the acid&base reaction shown previously (equation (2)). However, two data (solutions 2 and 4, pH 3.00) for absorption capacity indicate that SO, complexation also occurs in HA-fly ash mixtures. Work with sodium humates showed that complexation of SO, was negligible at pH 3.00 but caused significantly increased absorption (a lo’:/,) at pH 2.505. Thus, the observed enhanced (i.e., over that predicted from solution alkalinities) absorption at pH 3.00 for solutions 2 and 4 and zero enhancement for solution 3 at pH 2.50 appears contradictory to the previous results. However, there are two important differences between HA used in the two series of experiments which may be responsible for the variation in SO, complexation: (1) presence of di- and trivalent cations in HA-fly ash mixtures and (2) presence of significant amounts of sodium bicarbonate and sodium chloride in the crude sodium humate preparations. The importance ofexplanations (1) and (2) lies mostly in their probable effects on the structure and properties of the HA after exposure to SO,. As shown in equations (2) and (4), absorption of SO, leads to the protonation of acidic groups on HA resulting in formation of an HA precipitate. Although the flocculation mechanism of HA in acidic media is poorly understood, it is believed that hydrogen bonding plays an important role. When present, metal ions can also be incorporated into the flocculated HA and are believed to form intermolecular bridge bonds which contribute to the stability of the HA solid’. Furthermore, when present in sufficient concentration, metal ion bridges can replace hydrogen bonding as the dominant mechanism for holding the floe together. For example, a recent study with Cu(II) suggested that, at certain critical concentractions of Cu(II), the HA floe structure changed to accommodate greater amounts of cupric ion’, Briefly, the results presented in this paper suggest that changes in the HA brought about by the effects of cations derived from fly ash significantly affected the ability of the solid to bind aqueous sulphur dioxide. Although too little data exist for a detailed analysis of the problem, inspection of Tables IL3 suggests the following: (1) The presence of di- and especially trivalent cations inhibits the ability of HA to bind SO,. For example, HA in solution 4, with the least amount of cations (TUe I), was most active in complexation of SO,. HA in solution 3, containing the highest iron and aluminum concentrations, was totally inactive toward SO, binding. Finally, HA in solution 2, with low concentrations of Fe and Al but quite high calcium and magnesium concentrations, was only slightly active toward SO, complexation. (2) Complexation of SO, is inversely related to the ionic strength of the HA medium. The high ionic strength of the sodium humate preparations5 is the most probable cause for failure of HA to bind SO, at pH 3 in that series of experiments. As cation binding by humic substances is greatly affected by ionic strength lo 13, it is not unreasonable to assume the same for binding of SO,.
FUEL,
1981,
Vol 60, April
333
SO, sorption
by HA/fly
ash mixtures:
J. B. Green and S. E. Manahan
Given the limited knowledge of HA structure and properties, prediction of the nature of the HA-SO, binding mechanism is not possible. However, the known capability of sulphur dioxide to act as a Lewis acid14 coupled with the apparent inhibition of interaction with HA by cations suggests some type of reaction with the same group(s) on HA responsible for cation binding. In summary, although the SO,-HA binding phenomenon is interesting, its effect on the practical use of HAfly ash mixtures for SO, removal would probably be negligible. HA solutions recycled through a scrubber several times should contain more than enough cations to completely inhibit the complexation interaction. Summary research
of
a potential
scrubbing
operation;
,further
In an actual scrubbing operation, HA-fly ash mixtures would remove both sulphur dioxide and fly ash from stack gases, because aqueous scrubbers are quite effective in removal of fly ash15. Ideally, the newly absorbed fly ash would then react with HA to reform basic humates. Much of the sulphur dioxide would eventually be removed as CaSO, and/or CaSO, utilizing Ca(I1) from the ash. However, a build-up of Mg(II), Na(I), and K(1) sulphites in the scrubbing medium would occur after several cycles through the scrubber. Possible methods for removal of these salts have been discussed6; however, research is needed in this area. Finally, the kinetics and overall reactivity of HA in spent scrubber solutions with fresh fly ash have not yet been investigated. Obviously, this is an important step in the scrubbing operation. In theory, less ash should be required for regeneration of HA than is needed for formation of the original metal humates, because not all of the original metal ions are exchanged for H + during the first SO, exposure (see Results). Thus, assuming that HA
334
FUEL, 1981, Vol 60, April
will retain a similar amount of residual cations after the second SO, exposure as it did after the first, the utilization of base from fly ash used to regenerate the HA will be higher than that from the original ash. Hence, the base utilization data reported in this paper are probably conservative estimates for a scrubber in continuous recycle operation. ACKNOWLEDGEMENT This research was supported by a grant from the US Department of the Interior Office of Water Research and Technology Matching Grant B-1 1.5MO. REFERENCES
4 5 6
7 8 9 10 11
12 13 14 15
Tufte, P. H. et al., in US Bur. Mines Inform. Circ. 8650, (Compilers G. H. Gronhovd and W. R. Kube), 1974, pp. 103-33 Annon. J. Air Pollut. Contr. Assoc. 1977, 21, 948 Green, J. B. and Manahan, S. E. in ‘Chem. of Wastewater Technol.‘, (Ed. A. J. Rubin), Ann Arbor Sci., Ann Arbor, MI Chap. 22, pp. 373402 Manahan, S. E. et al. US Dept. Energy Rept. Invest., Laramie Energy Res. Center, Laramie, WY, LETC/RI-78/5 Green, J. B. and Manahan, S. paper submitted to Fuel Green, J. B. ‘Potential of Humic Acids as Stack Gas Scrubbing Media for Removal of Sulfur Dioxide and Fly Ash,’ Ph.D. Diss., Un. MO-Columbia, 1977 Green, J. B. and Manahan, S. E. Analyst. Chem. 1978, 50, 1975 Green, J. B. and Manahan, S. E. J. Inory. Nucl. Chem. 1977, 39, 1023 Green, J. B. and Manahan, S. E. Gun. J. Chem. 1977, 55, 3248 Schnitzer, M. and Khan, S. U. ‘Humic Substances in the Environment,’ M. Dekker, MY, 1972 Stevenson, F. J. in ‘Environ. Biogeochem., Proc. Int. Symp., 2nd 1975,’ (Ed. J. 0. Nriagu) Vol. 2, Ann Arbor Sci., Ann Arbor, Mich., 1975, pp. 519-540 Stevenson, F. J. Soil Sci. 1977, 123, 10 Stevenson, F. J. Soil Sci. Sot. Am. J. 1976, 40, 682 Milanova, E. and Benoit, R. L. Can. J. Chem. 1977, 55, 2807 Perkins, H. C. ‘Air Pollution,‘Chap. 10, McGraw-Hill, NY, 1974