- Email: [email protected]
Removal of crystal violet from aqueous solutions using coal
Journal of Colloid and Interface Science 422 (2014) 1–8
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Removal of crystal violet from aqueous solutions using coal Martin A. Schoonen a,⇑, Jan M.T. Schoonen a,b a b
Department of Geosciences, Stony Brook University, Stony Brook, NY 11794-2100, United States Department of Chemical Engineering and Material Science, Stevens Institute of Technology, Hoboken, NJ 07030, United States
a r t i c l e
i n f o
Article history: Received 23 September 2013 Accepted 5 February 2014 Available online 13 February 2014 Keywords: Crystal violet Wastewater Coal Sorption Pyrite
a b s t r a c t The interaction of crystal violet (CV) and six standard reference coals with varying amounts of pyrite was studied using batch sorption experiments. The experiments were designed to test the hypothesis that pyrite-containing coal removes CV through a combination of sorption and a Fenton-like degradation reaction involving pyrite. While pure pyrite does degrade CV slowly through a Fenton-like mechanism, bituminous coals containing pyrite showed far less CV removal than subbituminous coals without pyrite. Hence, the presence of pyrite in coal does not lead to an enhanced removal of CV from solution. Instead, the surface charge of coal appears to exert a primary role on the uptake of CV. The subbituminous coals tested in this study have a negative surface charge between pH 3 and 8, which facilitates the uptake of the cationic dye. Sorption of cationic CV onto subbituminous coal leads to a charge reversal. Modeling of the sorption kinetics suggest that CV diffuses into pore space within the coal after sorbing onto the surface, which is consistent with the fact that CV is not released after uptake by the coal. The results of this study indicate that subbituminous coal might be a useful sorbent for CV contained in waste streams generated in dye processes. Coal is a cheap bulk commodity, CV does not desorb easily, and the resulting CV-containing coal could be burned to incinerate the contaminant while producing energy. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Crystal violet (CV), a triarylmethane dye, is a common component in ink used in pens and printers [1]. CV and several closely related compounds are also widely used to color textile and leather. Over 100,000 different dyes are used and more than 0.7 million tons of dyes are produced annually [2]. It has been estimated that somewhere between 10% and 20% of the triarylmethane dyes are released to the environment, typically by discharge into surface waters [3,4]. Given that CV and other triarylmethane dyes are suspected carcinogens, the release of synthetic dyes represents a serious environmental problem and possible public health hazard in countries with thriving textile industries, such as India and China [1]. The coloring of surface waters with discharged dyes is also aesthetically undesirable [5]. The environmental and public health concerns have led to the development of an array of remediation strategies [2]. Most strategies rely on sorption of the dye from the process water. Several natural organic waste products, such as grapefruit peels [6], ginger ⇑ Corresponding author. Current address: Environmental Sciences Department, Brookhaven National Laboratory, Upton, NY 11973-5000, United States. Fax: +1 6316328246. E-mail addresses: [email protected], [email protected] (M.A. Schoonen), [email protected] (J.M.T. Schoonen). http://dx.doi.org/10.1016/j.jcis.2014.02.008 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
waste [7], pine bark [8], de-oiled soyabean waste [9], and palm kernel fiber and chitosan [5] have been evaluated as sorbents. The use of a wide range of agricultural solid waste to remove cationic and anionic dyes has recently been reviewed [1]. Besides agricultural solid waste, fly ash, lignite and charcoal have been evaluated as sorbents [9–11]. Other strategies rely on the degradation of the dye in the process water by microorganisms [3] or by (photo) chemical reactions. Chemical processes that have been evaluated include electrochemical degradation, photochemical degradation, reduction reactions using zero-valent iron, as well as Fenton-like, radical-based degradation [12–16]. As summarized in the recent review by Salleh and co-workers, chemical, photochemical, or microbial degradation technique have significant disadvantages and/or are costly to implement. Therefore, removal strategies based on sorption, particularly if the sorbent is a cheap bulk or waste material are of interest [9]. Chemical pretreatment of agricultural waste products to produce suitable dye sorption characteristics or the use of activated carbon adds cost. Hence, strategies based on sorption using a bulk sorbent that does not require chemical pretreatment have the lowest economic barrier to implementation. Here we evaluate the removal of dissolved CV by coal containing various amounts of pyrite. The rationale for studying coal with different pyrite content is based on the fact that pyrite-containing
2
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
coal spontaneously generates hydroxyl radicals (OH) when dispersed in water [17] as pyrite is oxidized by dissolved molecular oxygen [18]. The formation of OH radicals in pyrite slurries is the result of step-wise reduction of dissolved molecular oxygen at the pyrite surface [18] and has been conclusively demonstrated using Electron Spin Resonance spectroscopy [19]. We have shown that OH radicals formed in aerated pyrite slurries degrade RNA [20], DNA [19], adenine [21], and phenylanaline [22]. Independent of our studies, the same mechanism has been shown to degrade trichloroethylene [23] and lactate [24] in pyrite slurries. Because of the similarities to the classical Fenton reaction between hydrogen peroxide and dissolved ferrous iron, the formation of OH radicals in pyrite slurries has been referred to as a Fenton-like reaction [23]. Given that it is well established that CV degrades upon reaction with OH radicals [12,25–28], our hypothesis is that pyritecontaining coal removes CV from solution through a combination of sorption and degradation via a Fenton-like mechanism. Hence, the expectation is that coals containing pyrite would be superior materials to remove CV from solution compared to pyrite-free coals. Given that coal is a cheap bulk commodity it might be possible to process waters containing CV at relatively low cost. After drying, the coal powder could be burned once it is no longer effective in removing dye. Burning would effectively incinerate any sorbed CV. Earlier work has shown that a lignite coal sorbs CV [11], but there has been no study conducted with coal reference materials of different coal rank, nor has there been a study to evaluate the use of pyrite or pyrite-containing coal to address this pollution problem. Coal rank describes the level of coalification of organic matter. It ranges from the lowest degree of coalification found in lignite, to intermediate levels found in subbituminous to bituminous coal, to the highest level found in anthracite coal [29].
2. Materials and methods Five coal reference materials were obtained from the National Institute of Science and Technology (NIST) and one was obtained from the US Geological Survey (USGS). Full analyses of the reference materials are available on-line from NIST [30] and the USGS [31]. Table 1 provides several relevant characteristics of the coal reference materials. Included in Table 1 is the amount of OH radical formed in each of the coal slurries upon dispersion in water. These measurements were obtained in earlier work [17]. The reference materials are provided as powders and were used without any further pretreatment. The specific surface areas for the six coal reference materials were determined in earlier work using a five-point BET analysis on a Quantachrome NOVA 2000 instrument with nitrogen as adsorbate. As per instructions by the manufacturer, only measurements on at least 1 m2 of material are reported. Hence, by using sufficient amounts of material it is possible to attain accurate specific surface area measurements on materials with less than 1 m2/g (pers. Comm. Matthias Thommes, Quantachrome). The standard error in the BET measurement is estimated to be 2.5%. In addition to the coal reference materials, pure pyrite was used to study its interaction with CV in solution. The natural pyrite, originating from Huanzala, Peru, was obtained from Wards (Rochester, NY, USA). The material was crushed and sieved to a 63–90 micron size fraction. The specific surface area of the pyrite used in this study was 0.428 m2/g. A 2.46 mM primary CV stock solution was prepared by dissolving 1 g of CV in 1L deionized, ultra-filtered EasyPure™ water (hereafter abbreviated as DI). A secondary stock solution was prepared by 80-fold dilution of the primary stock solution with DI. This secondary solution (31 lM) and further dilutions of this solution were used as starting solutions in all experiments. Measurements of the
Table 1 Characteristics of coal reference materials used in study#. Number
Location
Coal type
Pyritic sulfur content (wt.%)
OH (24 h) (lM)
NIST #1635
Erie, Colorado Gillette, Wyoming Holden, West Virginia Marion, Illinios Captina, West Virginia Castleman, Maryland
Subbituminous
0.00
0
Subbituminous
0.01
0
Bituminous
0.49
0.3
Bituminous
0.52
1.0
Bituminous
1.15
1.5
Bituminous
0.67
0.4
NIST #2682b NIST #2692b
NIST #2684b NIST #2685b
USGS CLB-1
# Table adapted from Cohn et al. [17]. The amount of OH generated within a period of 24 h was measured in that study using a modified 30 -(p-Aminophenyl) fluorescein assay on a coal slurry with a surface loading of 40 m2/L [17].
absorbance of the starting solution as well as dilutions of the starting solution (1/2, 1/4, 1/10, 1/25 dilutions) using a HACH™ DR4000 UV/Vis spectrometer showed that the absorbance decreased linearly with decreasing concentration. These measurements yielded a molar extinction coefficient of 86,024 M1cm1. 2.1. Fenton degradation experiments While there is considerable prior work to show that OH radicals generated through the Fenton reaction degrade CV [12,25–28], we conducted an experiment to determine how the visible spectra evolved over time. The experiment was conducted in a methacrylate cuvette with a 1-cm pathlength (Fischerbrand). The reaction was initiated by adding 150 lL of a 10 mM H2O2 solution to a mixture of 2 mL of a 10 lM CV solution and 150 lL of a 1 mM Fe2+ solution. After adding the hydrogen peroxide, the solution was mixed by inverting the cuvette once and then immediately place in a UV–vis spectrometer (HACH DR-4000). The solution was repeatedly scanned between 400 and 700 nm for a period of 35 min. The ferrous iron solution was prepared by dissolving ferrous ammonium sulfate salt in DI. The hydrogen peroxide solution was prepared by diluting concentrated hydrogen peroxide solution in DI. The molality of the 10 mM stock solution was checked by determining its absorbance at 240 nm. The Fenton-CV experiment was complemented with an experiment with pyrite to investigate changes in CV spectra over time. This complementary experiment was conducted by dispersing 0.5 g pyrite in 40 mL of a 10 lM CV solution. The solution was contained in a 50 mL conical tube (Falcon™). The tube was placed on an end-over-end shaker (Labquake, Thermo Science™) and kept in the dark. Periodically, the tube was removed from the shaker, centrifuged, and 2 mL of the supernatant were pipetted out of the tube and transferred in a methacrylate cuvette. The supernatant was returned to the tube after a spectrum between 400 and 700 nm was collected. A second tube with 40 mL CV solution without pyrite was treated in the same manner for comparison. The reaction was followed for a total of 19 h (1140 min). The initial CV concentration in the two Fenton experiments was chosen to be 10 lM after a series of preliminary experiments showed that at this initial concentration the reaction in the homogenous system proceeded at a rate that progress could be captured by consecutive scans across the 400–700 nm spectral range. The experiment with pyrite was conducted at that same initial CV concentration to be comparable to the homogeneous
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fenton experiment. For the sorption experiments, described below, the initial CV concentration was increased to 31 lM to make use of the full linear range of the spectrometer. 2.2. Batch sorption experiments Batch experiments were conducted in 50 mL plastic conical tubes (Falcon™). The tubes were loaded with 100 mg of a coal reference material or pyrite and 50 mL of the CV starting solution. The tubes were placed for 1 min on a vortex mixer to disperse the solids into the CV solution. After this initial agitation, the tubes were placed on a programmable shaker. Samples were withdrawn by taking the tubes off the shaker, opening the conical tubes, and withdrawing approximately 3 mL by plastic syringe. Some of the coal settled rapidly and supernatant was directly transferred into a 1 cm methacrylate cuvet. Samples with suspended coal particles were transferred into 2 mL conical centrifuge vials and spun down for 5 min at 5000 rpm. After centrifugation, the visible spectrum of the supernatant was measured from 400 to 700 nm. The absorbance at 590 nm was used to quantify the CV concentration after correcting for background absorbance. Background absorbance was determined by taking the absorbance values at 425 nm and 675 nm, assuming a linear relationship, and calculating a background value at 590 nm. This calculated value was subtracted from the observed absorbance at 590 nm. Samples were taken over a period of 5 days. After the completion of this set of batch sorption experiments an additional experiment with pure pyrite was conducted with 500 mg pyrite rather than 100 mg pyrite to further explore the interaction between pyrite and CV at the CV concentration levels of the sorption experiments. This experiment was conducted for a total of 24.5 h. On the basis of the results of the batch experiments, one coal reference material (NIST #2682b) was selected for additional sorption experiments. In this second phase of experiments, the particle loading and CV concentration were varied so that sorption isotherms could be obtained. To improve agitation compared to the exploratory experiment with all coals, the conical tubes were placed on an end-over-end shaker (Labquake, Thermo Science™) after the initial dispersion using the vortex mixer. A complementary desorption experiment was conducted on the two subbituminous coals (NIST #1635 and #2682b). First 500 mg of coal was dispersed in 50 mL CV starting solution contained in conical tubes. After a brief agitation on the vortex mixer, the tubes were placed on the end-over-end shaker for 24 h. At that point, a sample was withdrawn for CV analysis. The 50 mL plastic tubes with the remaining dispersion was centrifuged for 5 min at 5000 rpm, the solution was decanted, and the coal rinsed with DI water. After centrifuging this DI-coal dispersion, the water was decanted and replaced with a solution containing 1 M acetic acid and 1 M KCl. An experimental study with pine bark as sorbent showed that under these conditions CV was nearly completely desorbed within 12 h [8]. Color development in the slurry was monitored for a total of 6 days after the addition of the acetic acid/KCl solution. 2.3. Fourier Transform Infra Red (FTIR) Spectroscopy For each of the five coal reference materials a FTIR spectrum was collected. In addition, a spectrum of NIST coal #2682b recovered at the conclusion of a batch sorption experiment was collected as well as a spectrum of #2682b after it was exposed for a one day to a concentrated CV stock solution (200 mg coal in 4.3 mL 0.5 mM CV solution). A spectrum of pure CV powder was also collected for comparison. About 10 mg of the coal or CV powder was mixed with dry KBr (spectroscopy grade, Fisher Scientific). The spectrum was collected on the KBr sample mixture in Diffuse
3
Reflectance Infrared Fourier Transform spectroscopy mode (DRIFT) on a Nexus 670 Thermo Nicolet™ spectrometer. The spectrum was collected between 400 and 4000 wavenumbers (cm1) using a DGTS detector. The FTIR data was interpreted using Thermo Nicolet™ Omnic software. 2.4. Surface charge measurement The surface charge of the coal reference material was determined on a Brookhaven Instrument Corporation™ Zeta Potential instrument equipped with Phase Analysis Laser Spectroscopy [32,33]. The electrophoretic mobility was determined on a dispersion of 200 mg of the coal reference material in 100 mL deionized water. The pH was varied by adding either KOH or HCl to 20 mL subsamples of the coal dispersions. The pH in the subsample was measured using a gel-filled combination electrode (Sensorex™). The surface charge was measured at five to six pH values between pH 3 and 8.5. The pH electrode was calibrated using NIST-traceable pH buffers (4, 7, and 10). The estimated uncertainty in the pH measurements is 0.1 pH unit. In addition to the experiments on the coal-in-water dispersion, a complementary set of data was collected on dispersions of 200 mg of coal powder in CV solution (200 mg coal powder dispersed in a 31 lM CV starting solution). These experiments were conducted to determine if CV is a potential-determining sorbate. The zeta potential for each slurry was calculated from the measured electrophoretic mobility measurements using the Smulochowski approximation [34]. 3. Results and discussion Consistent with the earlier work by Fan and coworkers [28], CV is rapidly degraded by OH radical generated by the Fenton reaction. The work by Fan et al. [28] showed that the degradation creates a suite of intermediate reaction products that exhibit absorbance maxima below 590 nm, the absorbance maximum for CV. This effect is also seen in this study. Fig. 1a shows a shift in absorbance maximum and broadening of the absorbance peak as the reaction with OH radical proceeds. A shift to lower wavelength in the absorbance maximum of CV solutions is thus an indication that degradation via OH radical is taking place. This same shift is also seen when CV is exposed to pyrite (Fig. 1b). The similarity between the two spectra suggests therefore that pyrite can degrade CV via a Fenton-like process in aerated solutions. The results of the Fenton degradation experiments shown in Fig. 1 indicate that the position of the absorbance maximum over time can be used to interpret qualitatively whether radical-driven degradation plays a significant role in a CV-containing coal slurry or pyrite slurry. No shift to a lower wavelength is expected for systems in which sorption is the process that dominates CV removal. By contrast, systems in which CV is removed by a combination of sorption and radical-driven degradation are expected to exhibit a shift in absorbance maxima to lower wavelength. Fig. 2 shows the visible light spectra (400–700 nm) for selected batch experiments as function of time. All batch experiments with coal show a removal of CV from solution over time, but there is no evidence for a shift in the position of the absorbance maximum to lower wavelength. This is illustrated by the three sets of spectra presented in panels A through C in Fig. 2. These coals represent the coal with the highest amount of CV removal during the batch experiment (subbituminous coal sample NIST #2682b with 0.01 wt.% pyrite sulfur, Fig. 2a), bituminous coal NIST 2692b with 0.49 wt.% pyrite sulfur (Fig. 2b), and NIST #2685b with 1.15 wt.% pyrite sulfur (Fig. 2c). Hence, there is no evidence for OH-driven degradation of CV, even in experiments with coals that contain
4
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fig. 1. Homogeneous Fenton CV degradation experiment (A) and experiment with CV dissolved in aerated pyrite slurry (B). Note shift of absorbance maximum to lower wavelength as CV is degraded. See text for details.
significant amounts of pyrite. There is, however, a shift in the absorbance maximum in batch sorption experiments with pyrite as illustrated for the experiment with 500 mg pyrite (Fig. 2d). The experiment with 100 mg pyrite showed a shift of only 2 nm (not shown in Fig. 2).
The removal of CV as measured by the absorbance at 590 nm is shown for all batch experiments with the six reference coals and pure pyrite at two loading in Fig. 3. The results indicate that the subbituminous coals (NIST #1635 and NIST #2682b) are the most effective sorbents (Fig. 3). After 121 h, these two coals had removed 67% and 96% of the initial CV, respectively (Table 2). Among the bituminous coals, NIST #2685b removed most CV, but even after more than 121 h only 38% had been removed. Pure pyrite removed or degraded 8% of the initial CV (Table 2). Table 2 also shows the experimental data after 121 h of contact on a surfacenormalized basis. The surface-normalized data show that coal #1635 removes more CV per unit surface area than #2682b. The results of this set of experiments do not show a systematic relationship between the decrease in CV concentration and pyrite concentration in the coal. Expressed on a surface-normalized basis the experiment with 0.1 g pyrite indicates that pyrite does remove CV on par with #2682b; however, to be practical the pyrite would have to have a higher surface area. The amount of CV removed or degraded increased from 8% in 7258 min to 15% in 1469 min by increasing the pyrite loading from 100 mg to 500 mg. This increase with particle loading suggests that the process is dependent on available surface loading but this was not further pursued in this study. While all coals sorb CV to some degree, none appear to reach a steady state concentration that might be interpreted as a partitioning equilibrium. The sorption experiments with the six coals (Fig. 3) all shows a continued loss of CV from solution over time. The rate of removal does, however, diminish over time. The follow-up experiments with a subbituminous coal, NIST #2682b (Fig. 4) shows for all but the highest particle loading a continued uptake of CV over time. In the experiment with NIST #2682b at the highest loading (200 mg/50 mL) essentially all CV is sorbed within the first 400 min (Fig. 4). On the basis of the results shown in Fig. 4, a set of experiments with NIST #2682b were conducted in which the initial CV concentration was varied and the contact time was sufficiently long (1140 min) to reach a stage in the process where the rate of loss of CV from solution is minor (Fig. 5). These experiments with varying initial CV concentration conform to a Langmuir sorption isotherm (inset Fig. 5). However, the results of
Fig. 2. Selected spectra of CV sorption experiments. Spectra in panels A–C collected after 30, 103, 243, 2023, and 7258 min. Spectra in panel D collected after 60, 159, 339, 549, 749, and 1469 min. Insets in panels B–D show location of absorbance maximum. Note that #2682b, a subbituminous coal with no pyrite shows much higher loss of CV than bituminous coal #2692b (0.49 wt.% pyrite sulfur), bituminous coal #2685b (1.15 wt.% pyrite sulfur), or pyrite. There is no shift in the position of the absorbance maximum in any of the coal experiments, while the absorbance maximum shifts to lower values in the experiment with pyrite.
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fig. 3. Batch experiments to evaluate efficacy of removal of CV by six standard coal reference materials and pure pyrite. Pyrite experiment with 0.5 g loading was conducted separately from other experiments for a shorter duration. See text for details.
desorption experiments with the two subbituminous coals (NIST #1635 and NIST #2682b) showed no release of any measurable amount of CV within six days. Hence, the loss of CV from solution in the presence of coal is irreversible on the time scale of days and does not represent an equilibrium partitioning of CV between solution and coal. The study of coals with different pyrite content was motivated by the notion that sorption onto coal combined with Fenton-like chemistry might lead to a more efficient removal and decomposition of CV from solution. Although pure pyrite appears to induce loss of CV via a Fenton-like mechanism, the results do not support the hypothesis that pyrite-containing coal has a higher efficacy for CV removal/degradation than coal without pyrite. The lack of a shift in the absorbance maximum in any of the batch experiments with coal (Fig. 2) indicate that sorption is the dominant process by which CV is removed regardless of the presence or absence of pyrite. It appears that coal rank controls the efficacy of CV removal. The results show that subbituminous coal is a far better sorbent than bituminous coal. The higher efficacy of subbituminous coal is likely related to the presence of functional groups that render the coals negatively charged, which facilitates the uptake of cationic CV. This notion is corroborated by the FTIR and electrophoresis measurements.
5
Fig. 4. Batch experiments with NIST #2682b with variable particle loading. Note that in experiment with highest loading nearly all CV is removed from solution within the first 400 min.
3400 cm1 region with coal rank, an increase of absorbance with coal rank in the 3000–3020 cm1 region, a decrease in absorbance in the 1675–1800 cm1 region with coal rank, and an increase of absorbance in the 700–900 cm1 spectral region with coal rank. The spectrum of NIST #2682b exposed to a 0.5 mM CV solution shows very little difference from the spectrum of unexposed coal (Fig. 6). Only a few sharp absorbance features in the 1200– 1600 cm1 region that are consistent with CV are barely visible in the spectrum of the exposed coal. These minor features are marked with a vertical dashed line in Fig. 6. Note that the concentration of CV used here (0.5 mM) is more than ten times the concentration used in the sorption experiments (31 lM). The major differences in the FTIR spectra with coal rank are consistent with a loss of hydroxyl functional groups (absorbance at 3400 cm1; RAOH stretch) and a loss of carboxylic groups (absorbance in 1675–1800 cm1 region; C@O stretch) paired with an increase in aromaticity (increase in absorbance at 3020 and 700–900 region; aromatic-CAH stretch and out-of-plane bend, respectively) [35]. These changes with coal rank have been reported in earlier work and show an overall loss of titratable OH functional groups [36,37] and condensation of aromatic organic matter [38,39].
3.1. FTIR spectroscopy The FTIR spectra show significant differences among the subbituminous and bituminous coals. The spectra for the coals are presented in Fig. 6 and raw spectra are available as Supplemental Material. The major differences in the spectra of the subbituminous coals and bituminous coals are the loss of absorbance in the
Table 2 Results of sorption experiment with coal reference materials and pyrite*.
*
Sample Surface area m2 g1
CV solution (lmol)
CV sorbed lmol
% Surface normalized Removed sorption (lmol m2)
#1635 #2682b #2692b #2684b #2685b CLB-1 FeS2
0.51 0.07 1.32 1.12 0.95 1.41 1.40
1.03 1.47 0.22 0.42 0.58 0.12 0.14
66.90 95.57 14.41 27.23 37.90 8.11 8.81
1.79 4.94 1.35 2.11 1.72 1.41 0.43
5.75 2.97 1.64 1.98 3.39 0.88 3.17
Sorption experiment conducted in 50 ml volume with 1.54 lmol of CV added to 0.1 g of sorbent. Data shown in table are after 121 h of contact.
Fig. 5. Result of batch experiments with NIST #2682b as a function of solution concentration. 50 mg of coal exposed to 50 mL of CV solution for 1140 min. Inset shows Langmuir plot for data. Note, however, that data shown in Figs. 3 and 4 clearly indicate equilibrium partitioning is not achieved within timeframe of experiment.
6
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Absorbance, arbitrary
0.2 abs
CV solid
#2682b + CV
#2682b #1635 #2692b #2684b #2685b #CLB-1
Wavenumber, cm-1 Fig. 6. FTIR spectra of standard reference coals, CV, and NIST #2682b exposed to concentrated CV solution. See text for details. Raw spectra available in Supplemental Materials.
magnitude of the positive charge below pH 7. The crossover point for NIST #2692b shifted from around 4 to 7.5 due to the fact that the entire charge-pH curve shifted upward by about 40 mV as a result of the addition of CV (see Supplemental Material). The change in the abundance of carboxylic acid functional groups as illustrated by the FTIR spectra is likely controlling the surface charge development on coals. Both organic acid-tritrable functional groups, such as carboxylic acids, as well as minerals contained in coal contribute to the overall surface charge or, more precisely, the net electrophoretic mobility of suspended coal particles [40]. The lower-rank coals are negatively charged over the entire pH range, similar to humic acids and soil organic carbon [41] and consistent with other work on subbituminous coal [42]. The data suggest that the loss of carboxylic acid functional groups with increasing coal rank limits their contribution and may lead to a relative larger contribution of other functional groups and mineral matter associated with the coal. Oxidation of pyrite and the presence of iron oxides would be consistent with an isoelectric point between pH 6 and 7 [43].
Fig. 7. Zeta potential measurements of a representative subbituminous coal (NIST #2682b) and a representative bituminous coal (NIST #2685b) with and without CV. Note that the addition of CV leads to a shift to more positive surface charge on both coals; however, the magnitude of the shift is much larger for the subbituminous coal. The shift in charge upon exposure to CV indicates that the cationic dye sorbs within the shear plane of the coal particles. See text for additional details. Supplemental Information contains zeta potential data for each coal reference material.
3.2. Surface charge on coal and effect of CV uptake The coals differed considerably in their surface charge (Fig. 7 and Supplemental Material). The two subbituminous coals (NIST #1635 and NIST #2682b) were strongly negatively charged over the entire pH range measured (pH 3–8), with zeta potential values between 24 and 63 mV (see Supplemental Material for detailed zeta potential data). Exposure to CV led to a positive surface charge, indicating strong interaction between the adsorbate (CV) and the coal. By contrast, bituminous coals dispersed in DI showed a moderate positive charge up to pH 6 or 7, with zeta potentials typically between +3 and +6 mV. Above pH 6–7, the bituminous coals attained a negative charge. Addition of CV did not significantly change the charge crossover point for most bituminous coals (#2692b is the one exception), but it did increase the
Fig. 8. Batch sorption data for NIST #2682b (Fig. 4) replotted versus square root of time (A). Note that the observed sorption kinetics are consistent with intra-particle uptake which is diffusion controlled, leading to a square root dependence on time. The data for the highest loading are ambiguous because there are not enough data points before all CV is removed from solution. Panel B shows the slope of the three experiments with the lowest particle loadings plot against the particle loading. The slope of the regression line through those three data points is an estimate of KID, see text.
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
3.3. Uptake kinetics and mechanism The follow-up experiments with subbituminous coal #2682b show that the first 1140 min of the uptake of CV can be described by a Langmuir isotherm (inset Fig. 5). However, a partition equilibrium is not established. The coal continues to take up CV after this initial phase (Fig. 4). The continued uptake of CV suggests that the sorbent has multiple sorption sites with different affinities for CV or that the coal exhibits intra-particle sorption. Coals have porosities ranging from 4 to 23%, with higher porosity found in coal of higher rank [44]. The pore system in coal becomes more restrictive in diameter with increasing rank. Sub-bituminous coal contains pores with diameters in excess of 50 nm, while the pore diameter in bituminous coals is typically less than 50 nm [29,45]. Prior work on CV removal by lignite coal has yielded results that are consistent with uptake of CV within the coal pore system [11]. We will evaluate here whether the data collected in this study are consistent with uptake in the pore structure. The kinetics of intra-particle sorption is governed by diffusion of CV from surface sites into pores within the coal particle. Hence, the kinetics of the process is dependent on the concentration gradient between solution and the pores within the particle, which leads to a square root dependence of concentration on time [11]:
C sorbed ¼ K 0 t0:5
ð1Þ
K 0 ¼ K ID Smass
ð2Þ
where KID is the intra-particle diffusion coefficient and Smass the sorbent mass. As shown in Fig. 8a, the CV concentration change over the entire duration in the experiments with NIST #2682b conforms to the square-root dependence, which suggests that intra-particle uptake is responsible for the continued removal of CV from solution. The value of KID estimated on the basis of the sorption experiments equals 2.02 1010 (mol mg1 min0.5) (Fig. 8b), which is of the same order of magnitude as the value obtained in earlier work on CV absorption by lignite coal (7.9–13.3 1010 mol mg1 min0.5) [11]. Intra-particle uptake of CV is also corroborated by the FTIR spectrum of NIST coal #2682b after exposure to a highly concentrated CV solution. Most of the recorded FTIR signal is derived from the interaction of the IR beam with the surface of the powder. A much more pronounced contribution of CV to the spectrum would be expected if CV predominantly resided on the surface. 4. Conclusions The results of this study indicate that subbituminous coal provides an option for the irreversible removal of CV and possibly other cationic dyes from wastewater. As demonstrated in this study, development of a negative surface charge over a wide range of pH determines whether a coal will be a good sorbent of CV. Bituminous coals are far less suitable sorbents as they lack carboxylic surface functional groups to render the surface negatively charged in water. In this study we established that NIST coal #2682b is minimally capable of removing 6.25 mg CV per gram of coal. This is based on the removal of essentially all CV in the sorption experiment shown in Fig. 3. This lower limit is higher than that for coir pitch (2.56 mg g1) and Calotropis procera leaf (4.14 mg g1), but is substantially lower than the capacity of treated ginger waste (277.7 mg g1) or grapefruit peel (254.16 mg g1) [7]. However, subbituminous coal is widely available as a bulk commodity, it can be used without any chemical pretreatment, and it can be burned after saturation with CV to incinerate the pollutant while yielding energy. Pure pyrite does remove CV from solution, but sorption by subbituminous containing no pyrite is a far more effective removal
7
mechanism. At the pyrite concentration levels in coal (typically less than 5 wt.%), there is no advantage to selecting a subbituminous coal that contains pyrite as the contribution of Fenton-like CV degradation is negligible. In fact, the presence of pyrite may lead to acidification of the waste water and it would lead to SO2 should the coal be burned. Acknowledgments This paper benefitted significantly from comments by three thoughtful reviewers. The experimental work described in this contribution was conducted in the senior author’s laboratory at Stony Brook University and leveraged equipment and facilities acquired with prior support from Stony Brook University, NASA, USDOE, and NSF. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.02.008. References [1] M.A.M. Salleh, D.K. Mahmoud, W.A.W.A. Karim, A. Idris, Desalination 280 (2011) 1. [2] T. Robinson, G. McMullan, R. Marchant, P. Nigam, Bioresour. Technol. 77 (2001) 247. [3] C.J. Ogugbue, T. Sawidis, Biotechnol. Res. Int. 2011 (2011). [4] B. Noroozi, G.A. Sorial, J. Environ. Sci. 25 (2013) 419. [5] W.H. Cheung, Y.S. Szeto, G. McKay, Bioresour. Technol. 98 (2007) 2897. [6] A. Saeed, M. Sharif, M. Iqbal, J. Hazard. Mater. 179 (2010) 564. [7] R. Kumar, R. Ahmad, Desalination 265 (2011) 112. [8] R. Ahmad, J. Hazard. Mater. 171 (2009) 767. [9] A. Mittal, J. Mittal, A. Malviya, D. Kaur, V.K. Gupta, J. Colloid Interface Sci. 343 (2010) 463. [10] M.J. Iqbal, M.N. Ashiq, J. Hazard. Mater. B139 (2007) 57. [11] P. Janos, P. Michalek, L. Turek, Dyes Pigm. 74 (2007) 363. [12] B. Goodell, Y. Qian, J. Jellison, M. Richard, Water Environ. Res. 76 (2004) 2703. [13] K.P. Singh, S. Gupta, A.K. Singh, S. Sinha, J. Hazard. Mater. 186 (2011) 1462. [14] C. Tizaoui, N. Karodia, M. Aburowais, J. Chem. Technol. Biotechnol. 85 (2010) 234. [15] L. Wojnarovits, T. Palfi, E. Takacs, S.S. Emmi, Radiat. Phys. Chem. 74 (2005) 239. [16] Z. Chen, T. Wang, X. Jin, Z. Chen, M. Megharaj, R. Naidu, J. Colloid Interface Sci. 398 (2013) 59. [17] C. Cohn, R. Laffers, S. Simon, T. O’Riordan, M. Schoonen, Part. Fibre Toxicol. 3 (2006) 16. [18] M.A.A. Schoonen, A.D. Harrington, R. Laffers, D.R. Strongin, Geochim. Cosmochim. Acta 74 (2010) 4971. [19] C. Cohn, S. Mueller, E. Wimmer, N. Leifer, S. Greenbaum, D. Strongin, M. Schoonen, Geochem. Trans. 7 (2006). [20] C.A. Cohn, R. Laffers, M.A.A. Schoonen, Environ. Sci. Technol. 40 (2006) 2838. [21] C.A. Cohn, S.C. Fisher, B.J. Brownawell, M.A.A. Schoonen, Geochem. Trans. 11 (2010). [22] S.C. Fisher, M.A.A. Schoonen, B.J. Brownawell, Geochem. Trans. 13 (2012). [23] H. Pham, M. Kitsuneduka, J. Hara, K. Suto, C. Inoue, Environ. Sci. Technol. 42 (2008) 7470. [24] W. Wang, Y. Qu, B. Yang, X. Liu, W. Su, Chemosphere 86 (2012) 376. [25] F.A. Alshamsi, A.S. Albadwawi, M.M. Alnuaimi, M.A. Rauf, S.S. Ashraf, Dyes Pigm. 74 (2007) 283. [26] S. Jana, M.K. Purkait, K. Mohanty, Appl. Clay Sci. 50 (2010) 337. [27] H. Wu, M.M. Fan, C.F. Li, M. Peng, L.J. Sheng, Q. Pan, G.W. Song, Water Sci. Technol. 62 (2010) 1. [28] H.-J. Fan, S.-T. Huang, W.-H. Chung, J.-L. Jan, W.-Y. Lin, C.-C. Chen, J. Hazard. Mater. 171 (2009) 1032. [29] L. Thomas, Coal Geology, John Wiley & Sons, 2002. [30] National Institute for Standards and Technology, in: Standard reference materials, Washington, 2010,. [32] F. McNeil-Watson, W. Tscharnuter, J. Miller, Colloids Surf., A 140 (1998) 53. [33] W.W. Tscharnuter, Appl. Opt. 40 (2001) 3995. [34] R.J. Hunter, Zeta Potential in Colloid Science – Principles and Applications, Academic Press, New York, 1981. [35] B. Smith, Infrared Spectral Analysis: A Systematic Approach, CRC Press, Boca Raton, Fl, 1999. [36] P.C. Painter, R.W. Snyder, M. Starsinic, M.M. Coleman, D.W. Kuehn, A. Davis, ACS Symp. Ser. 205 (1982) 47. [37] M. Starsinic, Y. Otake, P.L. Walker, P.C. Painter, Fuel 63 (1984) 1002. [38] J.V. Ibarra, E. Munoz, R. Moliner, Org. Geochem. 24 (1996) 725.
8
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
[39] P.C. Painter, M. Starsinic, E. Squires, A.A. Davis, Fuel 62 (1983) 742. [40] D.W. Fuerstenau, J.M. Rosenbaum, Y.S. You, Energy Fuels 2 (1988) 241. [41] W. Stumm, J.J. Morgan, Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, third ed., Wiley-Interscience, New York, 1995. [42] M.V. Lopezramon, C. Morenocastilla, J. Riverautrilla, R. Hidalgoalvarez, Carbon 31 (1993) 815.
[43] J. Bebie, M.A.A. Schoonen, M. Fuhrmann, D.R. Strongin, Geochim. Cosmochim. Acta 62 (1998) 633. [44] H. Gan, P.L. Nandi, L. Walker, Fuel 51 (1972) 272. [45] C.F. Rodrigues, M.J. Lemos de Sousa, Int. J. Coal Geol. 48 (2002) 245.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Removal of crystal violet from aqueous solutions using coal Martin A. Schoonen a,⇑, Jan M.T. Schoonen a,b a b
Department of Geosciences, Stony Brook University, Stony Brook, NY 11794-2100, United States Department of Chemical Engineering and Material Science, Stevens Institute of Technology, Hoboken, NJ 07030, United States
a r t i c l e
i n f o
Article history: Received 23 September 2013 Accepted 5 February 2014 Available online 13 February 2014 Keywords: Crystal violet Wastewater Coal Sorption Pyrite
a b s t r a c t The interaction of crystal violet (CV) and six standard reference coals with varying amounts of pyrite was studied using batch sorption experiments. The experiments were designed to test the hypothesis that pyrite-containing coal removes CV through a combination of sorption and a Fenton-like degradation reaction involving pyrite. While pure pyrite does degrade CV slowly through a Fenton-like mechanism, bituminous coals containing pyrite showed far less CV removal than subbituminous coals without pyrite. Hence, the presence of pyrite in coal does not lead to an enhanced removal of CV from solution. Instead, the surface charge of coal appears to exert a primary role on the uptake of CV. The subbituminous coals tested in this study have a negative surface charge between pH 3 and 8, which facilitates the uptake of the cationic dye. Sorption of cationic CV onto subbituminous coal leads to a charge reversal. Modeling of the sorption kinetics suggest that CV diffuses into pore space within the coal after sorbing onto the surface, which is consistent with the fact that CV is not released after uptake by the coal. The results of this study indicate that subbituminous coal might be a useful sorbent for CV contained in waste streams generated in dye processes. Coal is a cheap bulk commodity, CV does not desorb easily, and the resulting CV-containing coal could be burned to incinerate the contaminant while producing energy. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Crystal violet (CV), a triarylmethane dye, is a common component in ink used in pens and printers [1]. CV and several closely related compounds are also widely used to color textile and leather. Over 100,000 different dyes are used and more than 0.7 million tons of dyes are produced annually [2]. It has been estimated that somewhere between 10% and 20% of the triarylmethane dyes are released to the environment, typically by discharge into surface waters [3,4]. Given that CV and other triarylmethane dyes are suspected carcinogens, the release of synthetic dyes represents a serious environmental problem and possible public health hazard in countries with thriving textile industries, such as India and China [1]. The coloring of surface waters with discharged dyes is also aesthetically undesirable [5]. The environmental and public health concerns have led to the development of an array of remediation strategies [2]. Most strategies rely on sorption of the dye from the process water. Several natural organic waste products, such as grapefruit peels [6], ginger ⇑ Corresponding author. Current address: Environmental Sciences Department, Brookhaven National Laboratory, Upton, NY 11973-5000, United States. Fax: +1 6316328246. E-mail addresses: [email protected], [email protected] (M.A. Schoonen), [email protected] (J.M.T. Schoonen). http://dx.doi.org/10.1016/j.jcis.2014.02.008 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
waste [7], pine bark [8], de-oiled soyabean waste [9], and palm kernel fiber and chitosan [5] have been evaluated as sorbents. The use of a wide range of agricultural solid waste to remove cationic and anionic dyes has recently been reviewed [1]. Besides agricultural solid waste, fly ash, lignite and charcoal have been evaluated as sorbents [9–11]. Other strategies rely on the degradation of the dye in the process water by microorganisms [3] or by (photo) chemical reactions. Chemical processes that have been evaluated include electrochemical degradation, photochemical degradation, reduction reactions using zero-valent iron, as well as Fenton-like, radical-based degradation [12–16]. As summarized in the recent review by Salleh and co-workers, chemical, photochemical, or microbial degradation technique have significant disadvantages and/or are costly to implement. Therefore, removal strategies based on sorption, particularly if the sorbent is a cheap bulk or waste material are of interest [9]. Chemical pretreatment of agricultural waste products to produce suitable dye sorption characteristics or the use of activated carbon adds cost. Hence, strategies based on sorption using a bulk sorbent that does not require chemical pretreatment have the lowest economic barrier to implementation. Here we evaluate the removal of dissolved CV by coal containing various amounts of pyrite. The rationale for studying coal with different pyrite content is based on the fact that pyrite-containing
2
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
coal spontaneously generates hydroxyl radicals (OH) when dispersed in water [17] as pyrite is oxidized by dissolved molecular oxygen [18]. The formation of OH radicals in pyrite slurries is the result of step-wise reduction of dissolved molecular oxygen at the pyrite surface [18] and has been conclusively demonstrated using Electron Spin Resonance spectroscopy [19]. We have shown that OH radicals formed in aerated pyrite slurries degrade RNA [20], DNA [19], adenine [21], and phenylanaline [22]. Independent of our studies, the same mechanism has been shown to degrade trichloroethylene [23] and lactate [24] in pyrite slurries. Because of the similarities to the classical Fenton reaction between hydrogen peroxide and dissolved ferrous iron, the formation of OH radicals in pyrite slurries has been referred to as a Fenton-like reaction [23]. Given that it is well established that CV degrades upon reaction with OH radicals [12,25–28], our hypothesis is that pyritecontaining coal removes CV from solution through a combination of sorption and degradation via a Fenton-like mechanism. Hence, the expectation is that coals containing pyrite would be superior materials to remove CV from solution compared to pyrite-free coals. Given that coal is a cheap bulk commodity it might be possible to process waters containing CV at relatively low cost. After drying, the coal powder could be burned once it is no longer effective in removing dye. Burning would effectively incinerate any sorbed CV. Earlier work has shown that a lignite coal sorbs CV [11], but there has been no study conducted with coal reference materials of different coal rank, nor has there been a study to evaluate the use of pyrite or pyrite-containing coal to address this pollution problem. Coal rank describes the level of coalification of organic matter. It ranges from the lowest degree of coalification found in lignite, to intermediate levels found in subbituminous to bituminous coal, to the highest level found in anthracite coal [29].
2. Materials and methods Five coal reference materials were obtained from the National Institute of Science and Technology (NIST) and one was obtained from the US Geological Survey (USGS). Full analyses of the reference materials are available on-line from NIST [30] and the USGS [31]. Table 1 provides several relevant characteristics of the coal reference materials. Included in Table 1 is the amount of OH radical formed in each of the coal slurries upon dispersion in water. These measurements were obtained in earlier work [17]. The reference materials are provided as powders and were used without any further pretreatment. The specific surface areas for the six coal reference materials were determined in earlier work using a five-point BET analysis on a Quantachrome NOVA 2000 instrument with nitrogen as adsorbate. As per instructions by the manufacturer, only measurements on at least 1 m2 of material are reported. Hence, by using sufficient amounts of material it is possible to attain accurate specific surface area measurements on materials with less than 1 m2/g (pers. Comm. Matthias Thommes, Quantachrome). The standard error in the BET measurement is estimated to be 2.5%. In addition to the coal reference materials, pure pyrite was used to study its interaction with CV in solution. The natural pyrite, originating from Huanzala, Peru, was obtained from Wards (Rochester, NY, USA). The material was crushed and sieved to a 63–90 micron size fraction. The specific surface area of the pyrite used in this study was 0.428 m2/g. A 2.46 mM primary CV stock solution was prepared by dissolving 1 g of CV in 1L deionized, ultra-filtered EasyPure™ water (hereafter abbreviated as DI). A secondary stock solution was prepared by 80-fold dilution of the primary stock solution with DI. This secondary solution (31 lM) and further dilutions of this solution were used as starting solutions in all experiments. Measurements of the
Table 1 Characteristics of coal reference materials used in study#. Number
Location
Coal type
Pyritic sulfur content (wt.%)
OH (24 h) (lM)
NIST #1635
Erie, Colorado Gillette, Wyoming Holden, West Virginia Marion, Illinios Captina, West Virginia Castleman, Maryland
Subbituminous
0.00
0
Subbituminous
0.01
0
Bituminous
0.49
0.3
Bituminous
0.52
1.0
Bituminous
1.15
1.5
Bituminous
0.67
0.4
NIST #2682b NIST #2692b
NIST #2684b NIST #2685b
USGS CLB-1
# Table adapted from Cohn et al. [17]. The amount of OH generated within a period of 24 h was measured in that study using a modified 30 -(p-Aminophenyl) fluorescein assay on a coal slurry with a surface loading of 40 m2/L [17].
absorbance of the starting solution as well as dilutions of the starting solution (1/2, 1/4, 1/10, 1/25 dilutions) using a HACH™ DR4000 UV/Vis spectrometer showed that the absorbance decreased linearly with decreasing concentration. These measurements yielded a molar extinction coefficient of 86,024 M1cm1. 2.1. Fenton degradation experiments While there is considerable prior work to show that OH radicals generated through the Fenton reaction degrade CV [12,25–28], we conducted an experiment to determine how the visible spectra evolved over time. The experiment was conducted in a methacrylate cuvette with a 1-cm pathlength (Fischerbrand). The reaction was initiated by adding 150 lL of a 10 mM H2O2 solution to a mixture of 2 mL of a 10 lM CV solution and 150 lL of a 1 mM Fe2+ solution. After adding the hydrogen peroxide, the solution was mixed by inverting the cuvette once and then immediately place in a UV–vis spectrometer (HACH DR-4000). The solution was repeatedly scanned between 400 and 700 nm for a period of 35 min. The ferrous iron solution was prepared by dissolving ferrous ammonium sulfate salt in DI. The hydrogen peroxide solution was prepared by diluting concentrated hydrogen peroxide solution in DI. The molality of the 10 mM stock solution was checked by determining its absorbance at 240 nm. The Fenton-CV experiment was complemented with an experiment with pyrite to investigate changes in CV spectra over time. This complementary experiment was conducted by dispersing 0.5 g pyrite in 40 mL of a 10 lM CV solution. The solution was contained in a 50 mL conical tube (Falcon™). The tube was placed on an end-over-end shaker (Labquake, Thermo Science™) and kept in the dark. Periodically, the tube was removed from the shaker, centrifuged, and 2 mL of the supernatant were pipetted out of the tube and transferred in a methacrylate cuvette. The supernatant was returned to the tube after a spectrum between 400 and 700 nm was collected. A second tube with 40 mL CV solution without pyrite was treated in the same manner for comparison. The reaction was followed for a total of 19 h (1140 min). The initial CV concentration in the two Fenton experiments was chosen to be 10 lM after a series of preliminary experiments showed that at this initial concentration the reaction in the homogenous system proceeded at a rate that progress could be captured by consecutive scans across the 400–700 nm spectral range. The experiment with pyrite was conducted at that same initial CV concentration to be comparable to the homogeneous
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fenton experiment. For the sorption experiments, described below, the initial CV concentration was increased to 31 lM to make use of the full linear range of the spectrometer. 2.2. Batch sorption experiments Batch experiments were conducted in 50 mL plastic conical tubes (Falcon™). The tubes were loaded with 100 mg of a coal reference material or pyrite and 50 mL of the CV starting solution. The tubes were placed for 1 min on a vortex mixer to disperse the solids into the CV solution. After this initial agitation, the tubes were placed on a programmable shaker. Samples were withdrawn by taking the tubes off the shaker, opening the conical tubes, and withdrawing approximately 3 mL by plastic syringe. Some of the coal settled rapidly and supernatant was directly transferred into a 1 cm methacrylate cuvet. Samples with suspended coal particles were transferred into 2 mL conical centrifuge vials and spun down for 5 min at 5000 rpm. After centrifugation, the visible spectrum of the supernatant was measured from 400 to 700 nm. The absorbance at 590 nm was used to quantify the CV concentration after correcting for background absorbance. Background absorbance was determined by taking the absorbance values at 425 nm and 675 nm, assuming a linear relationship, and calculating a background value at 590 nm. This calculated value was subtracted from the observed absorbance at 590 nm. Samples were taken over a period of 5 days. After the completion of this set of batch sorption experiments an additional experiment with pure pyrite was conducted with 500 mg pyrite rather than 100 mg pyrite to further explore the interaction between pyrite and CV at the CV concentration levels of the sorption experiments. This experiment was conducted for a total of 24.5 h. On the basis of the results of the batch experiments, one coal reference material (NIST #2682b) was selected for additional sorption experiments. In this second phase of experiments, the particle loading and CV concentration were varied so that sorption isotherms could be obtained. To improve agitation compared to the exploratory experiment with all coals, the conical tubes were placed on an end-over-end shaker (Labquake, Thermo Science™) after the initial dispersion using the vortex mixer. A complementary desorption experiment was conducted on the two subbituminous coals (NIST #1635 and #2682b). First 500 mg of coal was dispersed in 50 mL CV starting solution contained in conical tubes. After a brief agitation on the vortex mixer, the tubes were placed on the end-over-end shaker for 24 h. At that point, a sample was withdrawn for CV analysis. The 50 mL plastic tubes with the remaining dispersion was centrifuged for 5 min at 5000 rpm, the solution was decanted, and the coal rinsed with DI water. After centrifuging this DI-coal dispersion, the water was decanted and replaced with a solution containing 1 M acetic acid and 1 M KCl. An experimental study with pine bark as sorbent showed that under these conditions CV was nearly completely desorbed within 12 h [8]. Color development in the slurry was monitored for a total of 6 days after the addition of the acetic acid/KCl solution. 2.3. Fourier Transform Infra Red (FTIR) Spectroscopy For each of the five coal reference materials a FTIR spectrum was collected. In addition, a spectrum of NIST coal #2682b recovered at the conclusion of a batch sorption experiment was collected as well as a spectrum of #2682b after it was exposed for a one day to a concentrated CV stock solution (200 mg coal in 4.3 mL 0.5 mM CV solution). A spectrum of pure CV powder was also collected for comparison. About 10 mg of the coal or CV powder was mixed with dry KBr (spectroscopy grade, Fisher Scientific). The spectrum was collected on the KBr sample mixture in Diffuse
3
Reflectance Infrared Fourier Transform spectroscopy mode (DRIFT) on a Nexus 670 Thermo Nicolet™ spectrometer. The spectrum was collected between 400 and 4000 wavenumbers (cm1) using a DGTS detector. The FTIR data was interpreted using Thermo Nicolet™ Omnic software. 2.4. Surface charge measurement The surface charge of the coal reference material was determined on a Brookhaven Instrument Corporation™ Zeta Potential instrument equipped with Phase Analysis Laser Spectroscopy [32,33]. The electrophoretic mobility was determined on a dispersion of 200 mg of the coal reference material in 100 mL deionized water. The pH was varied by adding either KOH or HCl to 20 mL subsamples of the coal dispersions. The pH in the subsample was measured using a gel-filled combination electrode (Sensorex™). The surface charge was measured at five to six pH values between pH 3 and 8.5. The pH electrode was calibrated using NIST-traceable pH buffers (4, 7, and 10). The estimated uncertainty in the pH measurements is 0.1 pH unit. In addition to the experiments on the coal-in-water dispersion, a complementary set of data was collected on dispersions of 200 mg of coal powder in CV solution (200 mg coal powder dispersed in a 31 lM CV starting solution). These experiments were conducted to determine if CV is a potential-determining sorbate. The zeta potential for each slurry was calculated from the measured electrophoretic mobility measurements using the Smulochowski approximation [34]. 3. Results and discussion Consistent with the earlier work by Fan and coworkers [28], CV is rapidly degraded by OH radical generated by the Fenton reaction. The work by Fan et al. [28] showed that the degradation creates a suite of intermediate reaction products that exhibit absorbance maxima below 590 nm, the absorbance maximum for CV. This effect is also seen in this study. Fig. 1a shows a shift in absorbance maximum and broadening of the absorbance peak as the reaction with OH radical proceeds. A shift to lower wavelength in the absorbance maximum of CV solutions is thus an indication that degradation via OH radical is taking place. This same shift is also seen when CV is exposed to pyrite (Fig. 1b). The similarity between the two spectra suggests therefore that pyrite can degrade CV via a Fenton-like process in aerated solutions. The results of the Fenton degradation experiments shown in Fig. 1 indicate that the position of the absorbance maximum over time can be used to interpret qualitatively whether radical-driven degradation plays a significant role in a CV-containing coal slurry or pyrite slurry. No shift to a lower wavelength is expected for systems in which sorption is the process that dominates CV removal. By contrast, systems in which CV is removed by a combination of sorption and radical-driven degradation are expected to exhibit a shift in absorbance maxima to lower wavelength. Fig. 2 shows the visible light spectra (400–700 nm) for selected batch experiments as function of time. All batch experiments with coal show a removal of CV from solution over time, but there is no evidence for a shift in the position of the absorbance maximum to lower wavelength. This is illustrated by the three sets of spectra presented in panels A through C in Fig. 2. These coals represent the coal with the highest amount of CV removal during the batch experiment (subbituminous coal sample NIST #2682b with 0.01 wt.% pyrite sulfur, Fig. 2a), bituminous coal NIST 2692b with 0.49 wt.% pyrite sulfur (Fig. 2b), and NIST #2685b with 1.15 wt.% pyrite sulfur (Fig. 2c). Hence, there is no evidence for OH-driven degradation of CV, even in experiments with coals that contain
4
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fig. 1. Homogeneous Fenton CV degradation experiment (A) and experiment with CV dissolved in aerated pyrite slurry (B). Note shift of absorbance maximum to lower wavelength as CV is degraded. See text for details.
significant amounts of pyrite. There is, however, a shift in the absorbance maximum in batch sorption experiments with pyrite as illustrated for the experiment with 500 mg pyrite (Fig. 2d). The experiment with 100 mg pyrite showed a shift of only 2 nm (not shown in Fig. 2).
The removal of CV as measured by the absorbance at 590 nm is shown for all batch experiments with the six reference coals and pure pyrite at two loading in Fig. 3. The results indicate that the subbituminous coals (NIST #1635 and NIST #2682b) are the most effective sorbents (Fig. 3). After 121 h, these two coals had removed 67% and 96% of the initial CV, respectively (Table 2). Among the bituminous coals, NIST #2685b removed most CV, but even after more than 121 h only 38% had been removed. Pure pyrite removed or degraded 8% of the initial CV (Table 2). Table 2 also shows the experimental data after 121 h of contact on a surfacenormalized basis. The surface-normalized data show that coal #1635 removes more CV per unit surface area than #2682b. The results of this set of experiments do not show a systematic relationship between the decrease in CV concentration and pyrite concentration in the coal. Expressed on a surface-normalized basis the experiment with 0.1 g pyrite indicates that pyrite does remove CV on par with #2682b; however, to be practical the pyrite would have to have a higher surface area. The amount of CV removed or degraded increased from 8% in 7258 min to 15% in 1469 min by increasing the pyrite loading from 100 mg to 500 mg. This increase with particle loading suggests that the process is dependent on available surface loading but this was not further pursued in this study. While all coals sorb CV to some degree, none appear to reach a steady state concentration that might be interpreted as a partitioning equilibrium. The sorption experiments with the six coals (Fig. 3) all shows a continued loss of CV from solution over time. The rate of removal does, however, diminish over time. The follow-up experiments with a subbituminous coal, NIST #2682b (Fig. 4) shows for all but the highest particle loading a continued uptake of CV over time. In the experiment with NIST #2682b at the highest loading (200 mg/50 mL) essentially all CV is sorbed within the first 400 min (Fig. 4). On the basis of the results shown in Fig. 4, a set of experiments with NIST #2682b were conducted in which the initial CV concentration was varied and the contact time was sufficiently long (1140 min) to reach a stage in the process where the rate of loss of CV from solution is minor (Fig. 5). These experiments with varying initial CV concentration conform to a Langmuir sorption isotherm (inset Fig. 5). However, the results of
Fig. 2. Selected spectra of CV sorption experiments. Spectra in panels A–C collected after 30, 103, 243, 2023, and 7258 min. Spectra in panel D collected after 60, 159, 339, 549, 749, and 1469 min. Insets in panels B–D show location of absorbance maximum. Note that #2682b, a subbituminous coal with no pyrite shows much higher loss of CV than bituminous coal #2692b (0.49 wt.% pyrite sulfur), bituminous coal #2685b (1.15 wt.% pyrite sulfur), or pyrite. There is no shift in the position of the absorbance maximum in any of the coal experiments, while the absorbance maximum shifts to lower values in the experiment with pyrite.
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Fig. 3. Batch experiments to evaluate efficacy of removal of CV by six standard coal reference materials and pure pyrite. Pyrite experiment with 0.5 g loading was conducted separately from other experiments for a shorter duration. See text for details.
desorption experiments with the two subbituminous coals (NIST #1635 and NIST #2682b) showed no release of any measurable amount of CV within six days. Hence, the loss of CV from solution in the presence of coal is irreversible on the time scale of days and does not represent an equilibrium partitioning of CV between solution and coal. The study of coals with different pyrite content was motivated by the notion that sorption onto coal combined with Fenton-like chemistry might lead to a more efficient removal and decomposition of CV from solution. Although pure pyrite appears to induce loss of CV via a Fenton-like mechanism, the results do not support the hypothesis that pyrite-containing coal has a higher efficacy for CV removal/degradation than coal without pyrite. The lack of a shift in the absorbance maximum in any of the batch experiments with coal (Fig. 2) indicate that sorption is the dominant process by which CV is removed regardless of the presence or absence of pyrite. It appears that coal rank controls the efficacy of CV removal. The results show that subbituminous coal is a far better sorbent than bituminous coal. The higher efficacy of subbituminous coal is likely related to the presence of functional groups that render the coals negatively charged, which facilitates the uptake of cationic CV. This notion is corroborated by the FTIR and electrophoresis measurements.
5
Fig. 4. Batch experiments with NIST #2682b with variable particle loading. Note that in experiment with highest loading nearly all CV is removed from solution within the first 400 min.
3400 cm1 region with coal rank, an increase of absorbance with coal rank in the 3000–3020 cm1 region, a decrease in absorbance in the 1675–1800 cm1 region with coal rank, and an increase of absorbance in the 700–900 cm1 spectral region with coal rank. The spectrum of NIST #2682b exposed to a 0.5 mM CV solution shows very little difference from the spectrum of unexposed coal (Fig. 6). Only a few sharp absorbance features in the 1200– 1600 cm1 region that are consistent with CV are barely visible in the spectrum of the exposed coal. These minor features are marked with a vertical dashed line in Fig. 6. Note that the concentration of CV used here (0.5 mM) is more than ten times the concentration used in the sorption experiments (31 lM). The major differences in the FTIR spectra with coal rank are consistent with a loss of hydroxyl functional groups (absorbance at 3400 cm1; RAOH stretch) and a loss of carboxylic groups (absorbance in 1675–1800 cm1 region; C@O stretch) paired with an increase in aromaticity (increase in absorbance at 3020 and 700–900 region; aromatic-CAH stretch and out-of-plane bend, respectively) [35]. These changes with coal rank have been reported in earlier work and show an overall loss of titratable OH functional groups [36,37] and condensation of aromatic organic matter [38,39].
3.1. FTIR spectroscopy The FTIR spectra show significant differences among the subbituminous and bituminous coals. The spectra for the coals are presented in Fig. 6 and raw spectra are available as Supplemental Material. The major differences in the spectra of the subbituminous coals and bituminous coals are the loss of absorbance in the
Table 2 Results of sorption experiment with coal reference materials and pyrite*.
*
Sample Surface area m2 g1
CV solution (lmol)
CV sorbed lmol
% Surface normalized Removed sorption (lmol m2)
#1635 #2682b #2692b #2684b #2685b CLB-1 FeS2
0.51 0.07 1.32 1.12 0.95 1.41 1.40
1.03 1.47 0.22 0.42 0.58 0.12 0.14
66.90 95.57 14.41 27.23 37.90 8.11 8.81
1.79 4.94 1.35 2.11 1.72 1.41 0.43
5.75 2.97 1.64 1.98 3.39 0.88 3.17
Sorption experiment conducted in 50 ml volume with 1.54 lmol of CV added to 0.1 g of sorbent. Data shown in table are after 121 h of contact.
Fig. 5. Result of batch experiments with NIST #2682b as a function of solution concentration. 50 mg of coal exposed to 50 mL of CV solution for 1140 min. Inset shows Langmuir plot for data. Note, however, that data shown in Figs. 3 and 4 clearly indicate equilibrium partitioning is not achieved within timeframe of experiment.
6
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
Absorbance, arbitrary
0.2 abs
CV solid
#2682b + CV
#2682b #1635 #2692b #2684b #2685b #CLB-1
Wavenumber, cm-1 Fig. 6. FTIR spectra of standard reference coals, CV, and NIST #2682b exposed to concentrated CV solution. See text for details. Raw spectra available in Supplemental Materials.
magnitude of the positive charge below pH 7. The crossover point for NIST #2692b shifted from around 4 to 7.5 due to the fact that the entire charge-pH curve shifted upward by about 40 mV as a result of the addition of CV (see Supplemental Material). The change in the abundance of carboxylic acid functional groups as illustrated by the FTIR spectra is likely controlling the surface charge development on coals. Both organic acid-tritrable functional groups, such as carboxylic acids, as well as minerals contained in coal contribute to the overall surface charge or, more precisely, the net electrophoretic mobility of suspended coal particles [40]. The lower-rank coals are negatively charged over the entire pH range, similar to humic acids and soil organic carbon [41] and consistent with other work on subbituminous coal [42]. The data suggest that the loss of carboxylic acid functional groups with increasing coal rank limits their contribution and may lead to a relative larger contribution of other functional groups and mineral matter associated with the coal. Oxidation of pyrite and the presence of iron oxides would be consistent with an isoelectric point between pH 6 and 7 [43].
Fig. 7. Zeta potential measurements of a representative subbituminous coal (NIST #2682b) and a representative bituminous coal (NIST #2685b) with and without CV. Note that the addition of CV leads to a shift to more positive surface charge on both coals; however, the magnitude of the shift is much larger for the subbituminous coal. The shift in charge upon exposure to CV indicates that the cationic dye sorbs within the shear plane of the coal particles. See text for additional details. Supplemental Information contains zeta potential data for each coal reference material.
3.2. Surface charge on coal and effect of CV uptake The coals differed considerably in their surface charge (Fig. 7 and Supplemental Material). The two subbituminous coals (NIST #1635 and NIST #2682b) were strongly negatively charged over the entire pH range measured (pH 3–8), with zeta potential values between 24 and 63 mV (see Supplemental Material for detailed zeta potential data). Exposure to CV led to a positive surface charge, indicating strong interaction between the adsorbate (CV) and the coal. By contrast, bituminous coals dispersed in DI showed a moderate positive charge up to pH 6 or 7, with zeta potentials typically between +3 and +6 mV. Above pH 6–7, the bituminous coals attained a negative charge. Addition of CV did not significantly change the charge crossover point for most bituminous coals (#2692b is the one exception), but it did increase the
Fig. 8. Batch sorption data for NIST #2682b (Fig. 4) replotted versus square root of time (A). Note that the observed sorption kinetics are consistent with intra-particle uptake which is diffusion controlled, leading to a square root dependence on time. The data for the highest loading are ambiguous because there are not enough data points before all CV is removed from solution. Panel B shows the slope of the three experiments with the lowest particle loadings plot against the particle loading. The slope of the regression line through those three data points is an estimate of KID, see text.
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
3.3. Uptake kinetics and mechanism The follow-up experiments with subbituminous coal #2682b show that the first 1140 min of the uptake of CV can be described by a Langmuir isotherm (inset Fig. 5). However, a partition equilibrium is not established. The coal continues to take up CV after this initial phase (Fig. 4). The continued uptake of CV suggests that the sorbent has multiple sorption sites with different affinities for CV or that the coal exhibits intra-particle sorption. Coals have porosities ranging from 4 to 23%, with higher porosity found in coal of higher rank [44]. The pore system in coal becomes more restrictive in diameter with increasing rank. Sub-bituminous coal contains pores with diameters in excess of 50 nm, while the pore diameter in bituminous coals is typically less than 50 nm [29,45]. Prior work on CV removal by lignite coal has yielded results that are consistent with uptake of CV within the coal pore system [11]. We will evaluate here whether the data collected in this study are consistent with uptake in the pore structure. The kinetics of intra-particle sorption is governed by diffusion of CV from surface sites into pores within the coal particle. Hence, the kinetics of the process is dependent on the concentration gradient between solution and the pores within the particle, which leads to a square root dependence of concentration on time [11]:
C sorbed ¼ K 0 t0:5
ð1Þ
K 0 ¼ K ID Smass
ð2Þ
where KID is the intra-particle diffusion coefficient and Smass the sorbent mass. As shown in Fig. 8a, the CV concentration change over the entire duration in the experiments with NIST #2682b conforms to the square-root dependence, which suggests that intra-particle uptake is responsible for the continued removal of CV from solution. The value of KID estimated on the basis of the sorption experiments equals 2.02 1010 (mol mg1 min0.5) (Fig. 8b), which is of the same order of magnitude as the value obtained in earlier work on CV absorption by lignite coal (7.9–13.3 1010 mol mg1 min0.5) [11]. Intra-particle uptake of CV is also corroborated by the FTIR spectrum of NIST coal #2682b after exposure to a highly concentrated CV solution. Most of the recorded FTIR signal is derived from the interaction of the IR beam with the surface of the powder. A much more pronounced contribution of CV to the spectrum would be expected if CV predominantly resided on the surface. 4. Conclusions The results of this study indicate that subbituminous coal provides an option for the irreversible removal of CV and possibly other cationic dyes from wastewater. As demonstrated in this study, development of a negative surface charge over a wide range of pH determines whether a coal will be a good sorbent of CV. Bituminous coals are far less suitable sorbents as they lack carboxylic surface functional groups to render the surface negatively charged in water. In this study we established that NIST coal #2682b is minimally capable of removing 6.25 mg CV per gram of coal. This is based on the removal of essentially all CV in the sorption experiment shown in Fig. 3. This lower limit is higher than that for coir pitch (2.56 mg g1) and Calotropis procera leaf (4.14 mg g1), but is substantially lower than the capacity of treated ginger waste (277.7 mg g1) or grapefruit peel (254.16 mg g1) [7]. However, subbituminous coal is widely available as a bulk commodity, it can be used without any chemical pretreatment, and it can be burned after saturation with CV to incinerate the pollutant while yielding energy. Pure pyrite does remove CV from solution, but sorption by subbituminous containing no pyrite is a far more effective removal
7
mechanism. At the pyrite concentration levels in coal (typically less than 5 wt.%), there is no advantage to selecting a subbituminous coal that contains pyrite as the contribution of Fenton-like CV degradation is negligible. In fact, the presence of pyrite may lead to acidification of the waste water and it would lead to SO2 should the coal be burned. Acknowledgments This paper benefitted significantly from comments by three thoughtful reviewers. The experimental work described in this contribution was conducted in the senior author’s laboratory at Stony Brook University and leveraged equipment and facilities acquired with prior support from Stony Brook University, NASA, USDOE, and NSF. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.02.008. References [1] M.A.M. Salleh, D.K. Mahmoud, W.A.W.A. Karim, A. Idris, Desalination 280 (2011) 1. [2] T. Robinson, G. McMullan, R. Marchant, P. Nigam, Bioresour. Technol. 77 (2001) 247. [3] C.J. Ogugbue, T. Sawidis, Biotechnol. Res. Int. 2011 (2011). [4] B. Noroozi, G.A. Sorial, J. Environ. Sci. 25 (2013) 419. [5] W.H. Cheung, Y.S. Szeto, G. McKay, Bioresour. Technol. 98 (2007) 2897. [6] A. Saeed, M. Sharif, M. Iqbal, J. Hazard. Mater. 179 (2010) 564. [7] R. Kumar, R. Ahmad, Desalination 265 (2011) 112. [8] R. Ahmad, J. Hazard. Mater. 171 (2009) 767. [9] A. Mittal, J. Mittal, A. Malviya, D. Kaur, V.K. Gupta, J. Colloid Interface Sci. 343 (2010) 463. [10] M.J. Iqbal, M.N. Ashiq, J. Hazard. Mater. B139 (2007) 57. [11] P. Janos, P. Michalek, L. Turek, Dyes Pigm. 74 (2007) 363. [12] B. Goodell, Y. Qian, J. Jellison, M. Richard, Water Environ. Res. 76 (2004) 2703. [13] K.P. Singh, S. Gupta, A.K. Singh, S. Sinha, J. Hazard. Mater. 186 (2011) 1462. [14] C. Tizaoui, N. Karodia, M. Aburowais, J. Chem. Technol. Biotechnol. 85 (2010) 234. [15] L. Wojnarovits, T. Palfi, E. Takacs, S.S. Emmi, Radiat. Phys. Chem. 74 (2005) 239. [16] Z. Chen, T. Wang, X. Jin, Z. Chen, M. Megharaj, R. Naidu, J. Colloid Interface Sci. 398 (2013) 59. [17] C. Cohn, R. Laffers, S. Simon, T. O’Riordan, M. Schoonen, Part. Fibre Toxicol. 3 (2006) 16. [18] M.A.A. Schoonen, A.D. Harrington, R. Laffers, D.R. Strongin, Geochim. Cosmochim. Acta 74 (2010) 4971. [19] C. Cohn, S. Mueller, E. Wimmer, N. Leifer, S. Greenbaum, D. Strongin, M. Schoonen, Geochem. Trans. 7 (2006). [20] C.A. Cohn, R. Laffers, M.A.A. Schoonen, Environ. Sci. Technol. 40 (2006) 2838. [21] C.A. Cohn, S.C. Fisher, B.J. Brownawell, M.A.A. Schoonen, Geochem. Trans. 11 (2010). [22] S.C. Fisher, M.A.A. Schoonen, B.J. Brownawell, Geochem. Trans. 13 (2012). [23] H. Pham, M. Kitsuneduka, J. Hara, K. Suto, C. Inoue, Environ. Sci. Technol. 42 (2008) 7470. [24] W. Wang, Y. Qu, B. Yang, X. Liu, W. Su, Chemosphere 86 (2012) 376. [25] F.A. Alshamsi, A.S. Albadwawi, M.M. Alnuaimi, M.A. Rauf, S.S. Ashraf, Dyes Pigm. 74 (2007) 283. [26] S. Jana, M.K. Purkait, K. Mohanty, Appl. Clay Sci. 50 (2010) 337. [27] H. Wu, M.M. Fan, C.F. Li, M. Peng, L.J. Sheng, Q. Pan, G.W. Song, Water Sci. Technol. 62 (2010) 1. [28] H.-J. Fan, S.-T. Huang, W.-H. Chung, J.-L. Jan, W.-Y. Lin, C.-C. Chen, J. Hazard. Mater. 171 (2009) 1032. [29] L. Thomas, Coal Geology, John Wiley & Sons, 2002. [30] National Institute for Standards and Technology, in: Standard reference materials, Washington, 2010,
8
M.A. Schoonen, J.M.T. Schoonen / Journal of Colloid and Interface Science 422 (2014) 1–8
[39] P.C. Painter, M. Starsinic, E. Squires, A.A. Davis, Fuel 62 (1983) 742. [40] D.W. Fuerstenau, J.M. Rosenbaum, Y.S. You, Energy Fuels 2 (1988) 241. [41] W. Stumm, J.J. Morgan, Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, third ed., Wiley-Interscience, New York, 1995. [42] M.V. Lopezramon, C. Morenocastilla, J. Riverautrilla, R. Hidalgoalvarez, Carbon 31 (1993) 815.
[43] J. Bebie, M.A.A. Schoonen, M. Fuhrmann, D.R. Strongin, Geochim. Cosmochim. Acta 62 (1998) 633. [44] H. Gan, P.L. Nandi, L. Walker, Fuel 51 (1972) 272. [45] C.F. Rodrigues, M.J. Lemos de Sousa, Int. J. Coal Geol. 48 (2002) 245.