Impacts of pH and ammonia on the leaching of Cu(II) and Cd(II) from coal fly ash

Chemosphere 64 (2006) 1892–1898 www.elsevier.com/locate/chemosphere

Impacts of pH and ammonia on the leaching of Cu(II) and Cd(II) from coal fly ash Jianmin Wang a

a,*

, Heng Ban b, Xinjun Teng b, Hao Wang b, Ken Ladwig

c

Department of Civil, Architectural and Environmental Engineering, University of Missouri-Rolla, Rolla, MO 65409, United States b School of Engineering, University of Alabama at Birmingham, Birmingham, AL 35294, United States c Electric Power Research Institute (EPRI), 3412 Hillview Avenue, Palo Alto, CA 94304, United States Received 15 November 2005; received in revised form 17 January 2006; accepted 18 January 2006 Available online 28 February 2006

Abstract Many coal-fired power plants are implementing ammonia-based technologies to reduce NOx emissions. Excess ammonia in the flue gas often deposits on the coal fly ash. Ammonia can form complexes with many heavy metals and change the leaching characteristics of these metals. This research tends to develop a fundamental understanding of the ammonia impact on the leaching of some heavy metals, exemplified by Cu(II) and Cd(II), under different pH conditions. Batch results indicated that the adsorption is the main mechanism controlling Cu(II) and Cd(II) leaching, and high concentrations of ammonia (>5000 mg/l) can increase the release of Cu(II) and Cd(II) in the alkaline pH range. Based on the chemical reactions among fly ash, ammonia, and heavy metal ion, a mathematical model was developed to quantify effects of pH and ammonia on metal adsorption. The adsorption constants (log K) of Cu2+, Cu(OH)+, Cu(OH)2, and CuðNH3 Þ2þ m for the fly ash under investigation were respectively 6.0, 7.7, 9.6, and 2.9. For Cd(II), these constants were respectively 4.3, 6.9, 8.8, and 2.6. Metal speciation calculations indicated that the formation of less adsorbable metal–ammonia complexes decreased metal adsorption, therefore enhanced metal leaching.  2006 Elsevier Ltd. All rights reserved. Keywords: Fly ash; Heavy metal; Ammonia; Leaching; Adsorption

1. Introduction Coal fly ash is a combustion product from coal-fired power plants. According to American Coal Ash Association (ACAA), US electric utilities generated 122 million short tons of coal combustion products (CCPs) in 2004, and 58% of the CCPs were fly ash (ACAA, 2005). Depending on the source of coal and combustion conditions, the fly ash may contain various levels of trace metals such as antimony, arsenic, barium, boron, cadmium, chromium, cobalt, copper, lead, mercury, molybdenum, nickel, selenium, silver, and zinc (Kim and Cardone, 1997; Querol et al., 1999; Kim and Kazonich, 2001). These metals can potentially be

*

Corresponding author. Tel.: +1 573 341 7503; fax: +1 573 341 4729. E-mail address: [email protected] (J. Wang).

0045-6535/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2006.01.041

released to the soil, surface water, and groundwater by leaching processes. Previous studies indicated that pH is one of the most important factors affecting metal leaching from fly ash. Results showed that the leachability of cationic metals such as cadmium, chromium, zinc, lead, mercury, and silver increases with decreasing pH (Weng and Huang, 1994; Fleming et al., 1996; Kim and Cardone, 1997; Fytianos and Tsaniklidi, 1998; Brickett et al., 2000; Kim and Kazonich, 2001; Lin and Chang, 2001; Bayat, 2002; Rao et al., 2002). However, metal leaching under natural conditions from traditional fly ash was generally not an environmental concern (Kim and Kazonich, 2001). Recent EPA regulations on controlling regional transport of ground level ozone require 22 eastern states in US to adopt NOx emission controls on large stationary sources. To meet these regulations, ammonia-based NOx control technologies, such as selective catalytic reduction

J. Wang et al. / Chemosphere 64 (2006) 1892–1898

(SCR) and selective non-catalytic reduction (SNCR) processes, have become increasingly common in coal-fired power plants, and their use will continue to grow significantly over the next few years. It is estimated that over 55% of the coal-fired power generating capacity will install either SCRs or SNCRs (Chu et al., 2001). In these processes, excess ammonia that escapes in the flue gas, referred to as ammonia slip, can be adsorbed onto the fly ash. The concentration of ammonia can be as high as 2500–2700 ppmw (part per million by weight) in fly ash under high slip conditions, although much lower concentrations are expected in typical operations (Fisher and Blackstock, 1997; EPRI, 1999, 2001; Bittner et al., 2001; Cardone et al., 2005). In addition to SCR and SNCR, ammonia is sometimes used as an ash conditioner to control ash resistivity for better particulate control in electrostatic precipitators (ESPs). Therefore, power plants without SCR or SNCR may also generate coal fly ash that is contaminated by ammonia. The western US lignite and sub-bituminous coal typically produces fly ash that contains relatively high calcium content compared to the fly ash generated from the eastern bituminous and anthracite coal. Since the calcium in fly ash is mostly in the form of lime, high calcium content results in alkaline pH in the leachate from the western coal fly ash. The pH in the leachate from the eastern coal fly ash is normally neutral or slightly acidic. In 2004, more than 40% of coal consumed for electric production was western coal (EIA, 2005). Ammonia can form water-soluble complexes with many heavy metals (Stumm and Morgan, 1995), especially under alkaline pH conditions. Therefore, it may change the leaching characteristics of these metals from fly ash. The objectives of this research are to determine the impact of ammonia on the leaching of cationic metals from fly ash, and to develop the fundamental understanding of the leaching process. 2. Theoretical aspects 2.1. Chemical reactions in ammonia-contaminated fly ash Following reactions could occur in ammonia-contaminated fly ash: (1) Ash surface site dissociation: SOH = Hþ + SO; K H

ð1Þ

where SOH and SO are respectively protonated and free surface sites, and KH is the acidity constant (M). The free surface site, SO, is responsible for the adsorption of cationic metals. (2) Metal hydroxide formation: ð2nÞþ

M2þ þ nOH ¼ MðOHÞn

2n

;

bðOHÞn

ð2Þ

where M2+ and MðOHÞn are respectively free metal ion and metal hydroxide, bðOHÞn is the overall forma-

1893

tion constant for the metal hydroxide, and n is the number of hydroxide ions associated with each metal ion, n = 1, 2, . . . (3) Metal adsorption (i.e. surface complex formation): SO + M2þ = SO–Mþ ;

SO + M(OH)þ = SO–M(OH); 

SO þ MðOHÞ2 ¼

ð3Þ

KS K SðOHÞ1

SO–MðOHÞ 2;

K SðOHÞ2

ð4Þ ð5Þ

SO–MðOHÞ 2

+

where SO–M , SO–M(OH), and are surface complexes of free metal and metal hydroxides, and KS, K SðOHÞ1 , and K SðOHÞ2 are stability constants of these surface complexes (i.e., adsorption constants). (4) Ammonium deprotonation: þ NHþ 4 ¼ NH3 þ H ;

ð6Þ

K NH3

where K NH3 is the acidity constant of ammonium ðNHþ 4 Þ, and pK NH3 ¼ 9:25 (Lide, 2003). (5) Metal–ammonia complex formation: M2þ þ mNH3 ¼ MðNH3 Þ2þ m ;

ð7Þ

bðNH3 Þm

MðNH3 Þ2þ m

where is a metal–ammonia complex, bðNH3 Þm is the stability constant of the metal–ammonia complex, and m is the number of ammonia molecules associated with each metal ion. (6) Adsorption of metal–ammonia complexes: þ SO þ MðNH3 Þ2þ m ¼ SO–MðNH3 Þm ;

K SðNH3 Þm

ð8Þ

þ SO–MðNH3 Þm

where is the adsorbed metal–ammonia complexes, and K SðNH3 Þm is the adsorption constant of the metal–ammonia complex. In this study, we assume that the adsorption constants of all metal– ammonia complexes are same ðK SðNH3 Þ Þ. 3. Metal partitioning model During the adsorption experiment, we control the total metal concentration at a relatively low level compared to the total surface site concentration and the total ammonia concentration. This is often true for fly ash. Thus, the adsorptions of free metal ion, metal hydroxides, and metal–ammonia complexes are in the linear range of the Langmuir adsorption isotherm, while the complexation between free metal ion and ammonia is also in the linear range of the complexation equilibrium. Therefore, the concentrations of the adsorbed metal species are proportional to their respective concentrations in the solution, and the concentrations of the soluble metal–ammonia complexes are also proportional to the free metal ion concentration in the solution. The concentrations of metal hydroxide species can be expressed as ð2nÞþ

½MðOHÞn

n

 ¼ bðOHÞn ½OH  ½M2þ ;

n ¼ 1; 2; . . .

ð9Þ

The concentrations of metal–ammonia complexes can be expressed as

1894

J. Wang et al. / Chemosphere 64 (2006) 1892–1898 2þ

m

½MðNH3 Þm  ¼ bðNH3 Þm ½NH3  ½M2þ ;

m ¼ 1; 2; . . .

ð10Þ

The concentrations of adsorbed metal species can be expressed as fSO–Mþ g = K S fSO g½M2þ  fSO–M(OH) = K SðOHÞ1 fSO g[M(OH)þ ] ¼ K SðOHÞ1 bðOHÞ1 fSO g[OH ] [M2þ ]  fSO–MðOHÞ 2 g ¼ K SðOHÞ2 fSO g½MðOHÞ2  2

¼ K SðOHÞ2 bðOHÞ2 fSO g½OH  ½M2þ 

X

þ

fSO–MðNH3 Þm g X ¼ K SðNH3 Þ fSO g ½MðNH3 Þ2þ m  X m  2þ ¼ K SðNH3 Þ fSO g½M  ðbðNH3 Þm ½NH3  Þ

ð11Þ

where {SO} is the concentration of the free surface site K S (M), fSO g ¼ ½HþHþKT H ; ST is the concentration of total metal binding site (M), which is a product of solids concentration (g/l) and the metal binding site density (mol/gsolids); [NH3] is the free ammonia concentration (M), K

N

3 T ½NH3  ¼ ½HþNH ; NT is the total ammonia concentration þK NH3 (M). As a result, the ratio of free metal ion to the total metal in the system can be expressed as X X ½M2þ   n m ¼ 1þ ðbðOHÞn ½OH  Þ þ ðbðNH3 Þm ½NH3  Þ MT þ K S fSO g þ K SðOHÞ1 bðOHÞ1 fSO g½OH 

2

þ K SðOHÞ2 bðOHÞ2 fSO g½OH  1 X m þ K SðNH3 Þ fSO g ðbðNH3 Þm ½NH3  Þ

ð12Þ

The ratio of metal adsorption, or metal partitioning R, can be expressed as Mads MT P fSO–Mþ g þ fSO–MðOHÞg þ fSO–MðOHÞ fSO–MðNH3 Þþ 2 gþ mg ¼ MT  ¼ K S fSO g þ K SðOHÞ1 bðOHÞ1 fSO g½OH 

R¼

þ K SðOHÞ2 bðOHÞ2 fSO g½OH 2  ½M2þ  X þ K SðNH3 Þ fSO g ðbðNH3 Þm ½NH3 m Þ MT

ð13Þ

Eq. (13) in association with Eq. (12) and expressions for {SO} and [NH3] quantifies the relationship of metal partitioning as functions of pH, the total ammonia concentration, and the total surface site concentration. 4. Materials and methods 4.1. Fly ash samples A coal fly ash generated from a full-scale power generating unit burning medium sulfur eastern bituminous coal was selected for this study. The ash sample was collected

from an electrostatic precipitator (ESP) that uses ammonia as a flue gas conditioner to enhance precipitator performance. This resulted in an ash with an ammonia concentration of approximately 150 ppmw. The ash has the BET surface area of 2.6 m2/g, carbon content (by CHN analyzer) of 3.6%, and loss on ignition (LOI) of 4.3% (EPRI, 2002). The raw ash was used to conduct batch leaching experiments under different ammonia additions. The washed ash, which was cleaned through a washing process before experiment, was used to conduct metal adsorption experiments. The objective of washing is to remove readily soluble materials and most heavy metals in fly ash so that their potential interference on adsorption can be reduced. 4.2. Ash washing The following procedure was followed for ash washing: (a) mix the ash with distilled and deionized water (DD water) at a solids/liquid (S/L) ratio of 1:5; (b) agitate the sample using air or a mechanical mixer for 10 h; (c) allow the mixture settle for approximately 1 h after the mixing is stopped; (d) decant the supernatant; (e) repeat the above procedure for another four times; (f) dry the sample completely at a temperature of 103–105 C; (g) sieve the dried ash sample using a #200 sieve (75 lm opening); and (h) mix and store the sieved sample in an air-tight container for later use. 4.3. Batch metal adsorption experiment A batch method was used to generate metal adsorption data under various pH and ammonia conditions. All experiments were conducted in single-metal systems. The experiments were carried out according to the following procedure: (a) A 10 g fly ash sample was added to each of the 125-ml bottles, except the blank. There were five groups of bottles corresponding to five different initial ammonia concentrations to be tested. Each group had 14 sample bottles for 14 pH conditions plus one blank. (b) One hundred milliliter of prepared water solution containing 0.01 M of NaNO3, a pre-selected concentration of heavy metal ion, and a preselected concentration of ammonia (use NH4NO3) were added to all bottles, including the blank. The same water solution was distributed to all bottles in the same group. (c) The pH was adjusted for all bottles in each group to a desired distribution in the pH range of 2–12 using tracemetal grade 1 M nitric acid (HNO3) or sodium hydroxide (NaOH) solutions. No acid or base was added to the blank sample. (d) All bottles were tightly sealed and placed in a mechanical shaker for 24 h at 130 revolutions per minute (rpm) to reach equilibrium (EPRI, 2005), and then allowed to settle for at least 30 min. (e) Twenty milliliter of the supernatant were filtered using a 0.45 lm syringe filter, and acidified. (f) Dissolved metal concentrations were measured on the acidified sample and final pH was measured on the mixture remaining in bottles.

J. Wang et al. / Chemosphere 64 (2006) 1892–1898

4.4. Batch leaching experiment The batch leaching procedure was similar to the above batch metal adsorption procedure, except that the raw ash sample was used, and the water solution added to the bottle contains only pre-selected concentrations of NH4NO3. 4.5. Chemical analyses An atomic absorption spectrometer (AAnalyst 800, Perkin–Elmer Corp., Norwalk, Connecticut) was used to determine the soluble Cu(II) and Cd(II) concentrations. The pH was measured using an Orion model 720A pH meter with Orion Ross model 81-02 pH electrode. An Orion model 95-12 ammonia electrode with ionic strength agent (ISA) was used to measure the ammonia concentration. 5. Results and discussion 5.1. Ammonia impact on Cu(II) and Cd(II) leaching from raw fly ash

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Unlike Cu(II), further increase of pH to above 10 does not decrease the Cd(II) leaching. The impact of ammonia on the leaching of Cd(II) is not significant when ammonia concentration is less than or equal to 1000 mg/l. Based on these batch leaching results, it can be concluded that high ammonia concentration can enhance the leaching of Cu(II) and Cd(II) under the alkaline pH condition. This is probably due to the formation of less adsorbable metal–ammonia complexes when ammonia presents in the system. 5.2. Ammonia impact on Cu(II) and Cd(II) adsorption by washed ash Five milligram per liter of Cu(II) or Cd(II) were added to the respective system containing washed ash. The equilibrium concentrations of Cu(II) and Cd(II) in supernatants were determined. Results indicated that the total soluble metal concentration under very low pH conditions were equal to or less than the added metal concentration. It suggests that the background metal concentration is significantly lower than the added metals concentration, and the washing protocol is appropriate to generate relatively clean ash for adsorption experiments. The metal adsorption ratio was calculated based on the soluble metal concentration and the total added metal concentration. Fig. 2 shows the adsorption results under different ammonia concentration conditions, plotted as ratio of Cu(II) and Cd(II) adsorption (R) as a function of pH. The Cu(II) partitioning data indicate that (a) Cu(II) adsorption increases with the increase of pH when pH is less than 7, and the Cu(II) adsorption is not affected by ammonia in neutral and low pH region; (b) when pH increases from 7 to 10, the Cu(II) adsorption under ammonia concentrations of 5000 and 10 000 mg/l decreases with the increase of pH, and the greater ammonia concentration results in the lesser Cu(II) adsorption; (c) when pH further increases to above 10, the Cu(II) adsorption increases with the increase of pH; (d) for lower ammonia concentrations (less than or equal to 1000 mg/l), the Cu(II) adsorption is not affected by ammonia. The Cu(II) adsorption trend is in good agreement with the Cu(II) leaching results shown in Fig. 1. Therefore, it can be concluded that adsorption is the major 0.12

Cu(II)

0 NH3 100 NH3 200 NH3 500 NH3 1000 NH3 5000 NH3 10 000 NH3

Cd Conc. (ppm)

Cu Conc. (ppm)

Fig. 1 shows the batch leaching results for Cu(II) and Cd(II) from the raw fly ash under different pH and ammonia conditions. The Cu(II) leaching data indicate that (a) when pH is less than 7, the Cu(II) leaching decreases with the increase of pH, and the ammonia does not impact the Cu(II) leaching; (b) when pH increases from 7 to 10, the Cu(II) leaching under ammonia concentrations of 5000 and 10 000 mg/l increases with the increase of pH, and the greater ammonia concentration results in the greater Cu(II) leaching; (c) when pH further increases to above 10, the leaching of Cu(II) decreases with the increase of pH; (d) for lower ammonia concentrations (less than or equal to 1000 mg/l), the Cu(II) leaching is not affected by ammonia. Fig. 1 also shows that the leaching of Cd(II) only decreases slightly with the increase of pH under neutral and low pH conditions. However, when pH is greater than 8, the leaching of Cd(II) under ammonia concentrations of 5000 and 10 000 mg/l increases with the increase of pH.

1895

Cd(II)

0.1

0 NH3 100 NH3

0.8

200 NH3

0.6

500 NH3

0.4

1000 NH3 5000 NH3

0.2

10 000 NH3

0 0

2

4

6

8 pH

10

12

14

0

2

4

6

8

10

12

14

pH

Fig. 1. Ammonia impact on the leaching of Cu(II) and Cd(II) from raw ash: soluble metal concentration as a function of pH under different ammonia addition conditions. Experimental conditions: temperature = 20–25 C; equilibration time = 24 h.

1896

J. Wang et al. / Chemosphere 64 (2006) 1892–1898

1.0

1.0

Cd(II)

0.8

Adsorption Ratio

Adsorption Ratio

Cu(II) 0.6 0 ppm NH3 200 ppm NH3

0.4

1000 ppm NH3

0.2

5000 ppm NH3

0.8 0.6 0 ppm NH3

0.4

200 ppm NH3 1000 ppm NH3

0.2

5000 ppm NH3

10 000 ppm NH3

0.0

10 000 ppm NH3

0.0 2

4

6

8 pH

10

12

14

2

4

6

8 pH

10

12

14

Fig. 2. Cu(II) and Cd(II) partitioning profiles in systems containing 100 g/l fly ash under different ammonia concentrations. The adsorption ratio is defined in Eq. (13). Experimental conditions: metal concentration = 5 mg/l; ion strength = 0.01 M (NaNO3); temperature = 20–25 C; and equilibration time = 24 h.

mechanism controlling Cu(II) leaching from fly ash. Fig. 2 also indicates that the adsorption behaviors of Cu(II) and Cd(II) are similar. In most cases, the Cd(II) adsorption results support the leaching data for raw ash. However, when pH is greater than 9.5, the adsorption of Cd(II) increases with the increase of pH, which does not agree with its leaching behavior for the same pH region. This is could be caused by the dissolution of other Cd(II) complexation ligands under high pH conditions. This will be further investigated in future. Since the ammonium ion has the acidity constant (pKa) of 9.25, the relative concentration of free ammonia increases significantly with the increase of pH in the alkaline pH region. Once the pH is greater than 9.25, free ammonia is the dominant species. Since free ammonia can form complexes with Cu(II) and Cd(II), it can be hypothesized that the decrease of Cu(II) and Cd(II) adsorption with the increase of pH under high ammonia concentrations is caused by the formation of less adsorbable metal–ammonia complexes.

metal binding site concentration (ST) used in this research was 1.8 · 103 M. Metal hydroxide species can be formed at alkaline pH. For Cu(II), the overall formation constants (log b(OH)) 2 for Cu(OH)+, Cu(OH)2, and CuðOHÞ4 are, respectively, 6.3, 11.8, and 16.4 (Stumm and Morgan, 1995). If a system contains ammonia, the free metal ion can react with free ammonia to form metal–ammonia complexes. For Cu(II), the overall formation constants ðlog bðNH3 Þ Þ for CuðNH3 Þ , CuðNH3 Þ2 , CuðNH3 Þ3 , and CuðNH3 Þ4 are, respectively, 4.0, 7.5, 10.3, and 11.8 (Stumm and Morgan, 1995). Based on these constants, the Cu(II) speciation as a function of pH was calculated. Fig. 3 shows the Cu(II) speciation profile in a system containing a total ammonia concentration (NT) of 0.6 M (10 000 mg/l). It shows that in the pH range from 6 to 12, most copper is in the form of Cu(II)–ammonia complexes. It also shows that when 2 pH is greater than 12, the negatively charged CuðOHÞ4 is the dominant species. The Cu(II) speciation determine its adsorption behavior.

5.3. Speciation of ash surface site and metal ions

5.4. Cu(II) partitioning modeling

In order to better understand the metal adsorption behavior under the influence of ammonia, the speciation of the metal binding sites on ash surface and the speciation of metal ions in solution must be considered. The surface site density and acidity constant of the fly ash used in this study were determined previously (Wang et al., 2004). The surface of this ash contains three types of acid sites, sites a, b, and c. Their densities are 2.1 · 104, 1.8 · 105, and 5.3 · 105 mol/g-fly ash, with acidity constants (pKH) of 2.7, 7.8, and 11.0, respectively. Previous study also indicated that the site a does not adsorb cationic metals ions (Wang et al., 2004). Since most cationic metal ions are in negatively charged hydroxide forms when pH approaches 11, they are unlikely to be adsorbed by free site c. Therefore, only the free surface site b is responsible for cationic metal adsorption. Since the total metal binding site density for this ash is 1.8 · 105 mol/g, and the total solids concentration is 100 g/l for the S/L = 1:10 ash sample, the total

Eq. (13) was used to fit the Cu(II) adsorption data in Fig. 2. Multiple variable non-linear regression programs

2þ

2þ

Ratio of different Cu(II) species

1

2þ

2þ

Cu(NH 3)42+

0.9

Cu(NH3)32+

0.8 Cu 2+

0.7

Cu(NH3)22+ Cu(NH3)2+

0.6

2-

Cu(OH) 4

0.5 0.4 0.3 0.2 0.1 0

2

4

6

8 pH

10

12

14

Fig. 3. Cu(II) speciation profile in a water solution containing 0.6 M (10 000 mg/l) of total ammonia.

J. Wang et al. / Chemosphere 64 (2006) 1892–1898

Ratio of different Cu(II) species

1 0.9

SO-Cu(OH)2-

0.8 0.7

SO-Cu + SO-Cu(OH)

Cu 2+

+

SO-Cu(NH 3)m

0.6

2+

0.4

Cu(NH3)2

0.3 2+

0.2 0.1 2

4

2þ

2þ

200 mg/l

1000 mg/l 5000 mg/l

0.2

10 000 mg/l

2þ

CdðNH3 Þ2 , CdðNH3 Þ3 , and CdðNH3 Þ4 are, respectively, 2.6, 4.6, 5.9, and 6.7 (Stumm and Morgan, 1995). These constants were directly applied to the model. Based on the Cd(II) adsorption results in Fig. 2, the adsorption

0 mg/l

0.3

14

For the validation purpose, Cd(II) adsorption data under the impact of ammonia was also modeled using Eq. (13). The overall formation constants (log b(OH)) for Cd(OH)+ and Cd(OH)2 are, respectively, 3.9 and 7.6, and the overall formation constants ðlog bðNH3 Þ Þ for Cd(NH3)2+,

0.8 0.7

0.4

12

5.5. Cd(II) adsorption in the presence of ammonia

0.8

200 mg/l

10

Fig. 5 shows the calculated Cu(II) speciation results in a system containing 100 g/l fly ash (S/L = 1:10) and 0.6 M (10 000 mg/l) ammonia. Results show that when pH is less than 8, the major soluble species is free Cu(II) ion. If pH is greater than 8, Cu(II)–ammonia complexes are the major soluble species. The concentrations of soluble Cu(II) hydroxide species are not significant in the entire experimental pH range, primarily due to their greater adsorption strength than free Cu(II). This calculation also indicates that the formation of less adsorbable Cu(II)–ammonia complexes results in the decrease of copper adsorption when pH is greater than 8.

1

0.5

8

Fig. 5. Cu(II) speciation profile in a system containing 100 g/l of fly ash and 0.6 M (10 000 mg/l) of total ammonia.

0.9

0 mg/l

6

pH

1

0.6

Cu(NH3)3

Cu(NH3)2+

0.9 0.7

2+

Cu(NH3)4

0.5

0

Adsorption ratio

Adsorption ratio

such as SigmaPlot were used for curve fitting. Since the formation constants for Cu(II) hydroxides ðbðOHÞn Þ and Cu(II)–ammonia complexes ðbðNH3 Þm Þ were found from references, the only unknown constants in Eq. (13) are KS, K SðOHÞ1 , K SðOHÞ2 , and K SðNH3 Þ . During the curve fitting, we arbitrarily apply all the constants found in literatures to the model without modification. Curve fitting results indicate that the logarithms of adsorption constants KS, K SðOHÞ1 , K SðOHÞ2 , and K SðNH3 Þ are 6.0, 7.7, 9.6, and 2.9, with respective standard errors of 0.04, 0.40, 0.14, and 0.10. Fig. 4 shows the curve fitting results in solid curves. It shows that the modeling results agree with the experimental data. The correlation coefficient (R2) for this curve fitting is 0.974. According to our previous study (Wang et al., 2004), the adsorption constant (log KS) of Cu2+ for the same fly ash determined in a system without ammonia is 6.1. The value of 6.0 obtained in the current study in a system containing ammonia agrees well with the previous result. It partially validated the metal adsorption model developed in this research. Fig. 4 shows that the predicted Cu(II) adsorption ratio under extremely high pH conditions (pH > 11) is greater than the experimental data. This may suggest that: (a) the actual formation constant for non-adsorbable CuðOHÞ2 4 is greater than those cited from the reference; or (b) the adsorption constants for Cu(II)–ammonia complexes decrease with the increase of the number of ammonia (m) associated with each Cu(II) in the complex. However, since the model can reasonably fit the experimental data in the pH range between 2 and 11, the metal adsorption model developed in this study is appropriate. Results from this study clearly show that the metal– ammonia complexes have much lower adsorption constants than free metal ion and metal hydroxides. Therefore, it is confirmed that the formation of metal–ammonia complexes reduces metal adsorption under high ammonia conditions. This same approach may be used to quantify the metal partitioning in systems that contain other metal ions and metal complexation agents.

1897

1000 mg/l 5000 mg/l

0.6

10 000 mg/l

0.5 0.4 0.3 0.2

0.1

0.1 0

0 2

4

6

8

10

12

14

pH Fig. 4. Modeling results (solid lines) for Cu(II) partitioning in systems containing 100 g/l of fly ash and different ammonia concentrations. The adsorption ratio is defined in Eq. (13).

2

4

6

8

10

12

14

pH Fig. 6. Modeling results (solid lines) for Cd(II) partitioning in systems containing 100 g/l fly ash and different ammonia concentrations. The adsorption ratio is defined in Eq. (13).

1898

J. Wang et al. / Chemosphere 64 (2006) 1892–1898

constants (log K) of Cd2+, Cd(OH)+, Cd(OH)2, and Cd(II)–ammonia complexes were determined. Their values are 4.3, 6.9, 8.8, and 2.6, respectively. The correlation coefficient (R2) for this curve fitting is 0.961. Fig. 6 shows curve fitting results in solid curves, which can reasonably reflect the trend of the experimental data. The log KS value for Cd2+ determined in this study is comparable to the previous value of 4.8 determined in a system without ammonia (Wang et al., 2004). It is expected that the curve fitting can be improved if the constants cited from the literatures are corrected based on the experimental conditions. It should be noted that many other elements such as mercury and nickel can form strong complexes with ammonia. It is possible that their leaching can be enhanced at a lower ammonia concentration. The approach used in this research could be used to investigate the leachablity of these elements and assess the potential environmental impact of ammonia contaminated fly ash. 6. Conclusion This research indicates that adsorption is the major mechanism controlling Cu(II) and Cd(II) leaching from fly ash, and high concentrations of ammonia (>5000 mg/l) enhance their leaching. The mathematical model developed in this study based on chemical reactions among heavy metal, fly ash surface site, and ammonia under different pH conditions can well quantify the metal adsorption behavior in fly ash that contains ammonia. By fitting metal partitioning data, the adsorption constants (log K) of Cu2+, Cu(OH)+, Cu(OH)2, and Cu(NH3)m for the tested fly ash are determined to be 6.0, 7.7, 9.6, and 2.9, respectively. For Cd(II), these constants are respectively 4.3, 6.9, 8.8, and 2.6. It is concluded that the formation of less adsorbable metal– ammonia complexes results in the decrease of metal adsorption under high ammonia concentration conditions. Acknowledgments This work was partially supported by the Electric Power Research Institute (EPRI) (EPRI PID043352). Authors wish to thank colleagues at University of Missouri-Rolla, Dr. Craig Addams, Dr. Glenn Morrison, Dr. Mohammad Qureshi for their invaluable comments. Conclusions and statements made in this paper are those of the authors and in no way reflect the endorsement of the funding agency. References American Coal Ash Association (ACAA), 2005. 2004 Coal Combustion Product (CCP) Production and Use Survey. Available from: .

Bayat, B., 2002. Combined removal of zinc(II) and cadmium(II) for aqueous solutions by adsorption onto high-calcium Turkish fly ash. Water, Air and Soil Pollution 136, 69–92. Bittner, J., Gasiorowski, S., Hrach, F., 2001. Removing ammonia from fly ash. In: Proceedings, 2001 International Ash Utilization Symposium, Lexington, Kentucky. Brickett, L., Cardone, C., Dahlberg, M., Harrison, D., Kazonich, G., Kim, A., Knoer, J., Slivon, J., 2000. OST technical progress report – FY2000 results, by-product utilization term. Cardone, C., Kim, A., Schroeder, K., 2005. Release of ammonia from SCR/SNCR fly ashes. In: Proceedings, 2005 World of Coal Ash (WOCA), Lexington, Kentucky, USA. Chu, P., Goodman, N., Behrens, G., Roberson, R., 2001. Total and specified mercury emissions from US coal-fired power plants. In: Proceedings, 4th Electric Utilities Environmental Conference, Tucson, Arizona. Energy Information Administration (EIA), 2005. Annual Coal Report. DOE/EIA-0584 (2004). EPRI, 1999. Investigation of ammonia adsorption on fly ash and potential impacts of ammoniated ash. EPRI, Palo Alto, CA: Report TE-113777. EPRI, 2001. Characterization of ammonia leaching from coal fly ash. EPRI, Palo Alto, CA: Report 1005163. EPRI, 2002. Behavior of ammoniated fly ash: effects of ammonia on fly ash handling, disposal, and end-use. EPRI, Palo Alto, CA: Report 1003981. EPRI, 2005. Effects of ammonia on trace element leaching from coal fly ash. EPRI, Palo Alto, CA: Report 1010063. Fisher, B.C., Blackstock, T., 1997. Fly ash beneficiation using an ammonia stripping process. In: Proceedings, 12th International Symposium on Coal Combustion-by Products Management and Use, 1997, pp. 65-1–65-8. Fleming, L.N., Abinteh, H.N., Inyang, H.I., 1996. Leachant pH effects on the leachability of metals from fly ash. Journal of Soil Contamination 5, 53–59. Fytianos, K., Tsaniklidi, B., 1998. Leachability of heavy metals in Greek fly ash from coal combustion. Environmental International 24, 477– 486. Kim, A.G., Cardone, C., 1997. Preliminary statistical analysis of fly ash disposal in mined areasProc: 12th International Symposium on Coal Combustion By-Product Management and Use, vol. 1. American Coal Ash Association, pp. 11-1–11-13. Kim, A.G., Kazonich, G., 2001. Release of trace elements from CCB: maximum extractable fraction. In: Proceedings 14th International Symposium on Management and Use of Coal Combustion Products (CCPs), vol. 1, pp. 20-1–20-15. Lide, D.R., 2003. 2003–2004 Handbook of Chemistry and Physics, 84th ed. CRC Press. Lin, C.J., Chang, J., 2001. Effect of fly ash characteristics on the removal of Cu(II) from aqueous solution. Chemosphere 44, 1185–1192. Querol, X., Umana, J.C., Alastuey, A., Bertrana, C., Lopez-Soler, A., Plana, F., 1999. Physicochemical characterization of Spanish fly ashes. Energy Sources 21, 883–898. Rao, M., Parwate, A.V., Bhole, A.G., 2002. Removal of Cr6+, and Ni2+ from aqueous solution using bagasse and fly ash. Waste Management 22, 821–830. Stumm, W., Morgan, J., 1995. Aquatic Chemistry, 3rd ed. John Wiley & Sons. Wang, J., Teng, X., Wang, H., Ban, H., 2004. Characterizing the metal adsorption capability of a class F coal fly ash. ES&T 38, 6710–6715. Weng, C.H., Huang, C.P., 1994. Treatment of metal industrial wastewater by fly ash and cement fixation. Journal of Environmental Engineering 120, 1470–1487.