ISSN 1063455X, Journal of Water Chemistry and Technology, 2014, Vol. 36, No. 3, pp. 125–133. © Allerton Press, Inc., 2014. Original Russian Text © K. Nadafi, M. Vosoughi, A. Asadi, M. Omidvar Borna, M. Shirmardi, 2014, published in Khimiya i Tekhnologiya Vody, 2014, Vol. 36, No. 3, pp. 229– 244.
PHYSICAL CHEMISTRY OF WATER TREATMENT PROCESSES
Reactive Red 120 Dye Removal From Aqueous Solution by Adsorption on NanoAlumina K. Nadafia, M. Vosoughib, *, A. Asadia, M. Omidvar Bornaa, and M. Shirmardic a
University of Medical Sciences, Tehran; Shahid Beheshti University of Medical Sciences, Tehran; c Jondishapor University of Medical Sciences, Ahvaz, Iran *e-mail:
[email protected]
b
Received April 23, 2012
Abstract—The adsorption of Reactive Red 120 dye from aqueous solutions by using nanoalumina has been investigated. The batch adsorption studies were carried out to determine the impact of pH, contact time, concentration of dye, and the adsorbent dose on adsorption process. The maximum adsorption effi ciency was observed at pH 3. However with an increase of the adsorbent dose, the dye removal efficiency increased, while the amount of dye adsorbed per unit mass (mg/g) decreased. A pseudosecondorder model best described the adsorption kinetics of the specified dye onto nanoalumina. In this case the Langmuir isotherm model appeared to be most suitable. Findings of the present study reveal that nano alumina can be an effective adsorbent for the removal of Reactive Red 120 from aqueous solutions. DOI: 10.3103/S1063455X14030059 Keywords: adsorption, adsorption isotherms, kinetics, reactive Red 120, nanoalumina.
INTRODUCTION Many industrial processes use synthetic dyes for producing textile, plastic, leather, food, paper, etc. [1–3]. According to the Color Index (CI), currently more than 10000 various types of dye are synthesized and avail able at the world market. Although no recent data is available on worldwide dye production, the annual pro duction of over 700000 tons has often been reported in literature. It is estimated that from 10 to 15% of dyes are wasted in the effluent during the dying process [4, 5]. Azo reactive dyes are most common compounds comprising about 60–70% of the total amount of dyes used at enterprises of different industries [6]. The discharge of colored wastewater to receiving waters poses serious environmental, aesthetical, and san itary problems [7]. The presence of dye in water not only causes aesthetical complains but also hinders the sun light penetration that disturbs the normal existence of hydrobionts (aquatic flora and fauna); in addition, syn thetic dyes are toxic for aquatic microorganisms and animals [8, 9]. Moreover some dyes and their byproducts are carcinogens [9, 10]. Various methods have been used for the removal of dye. They include chemical precipitation, adsorption, reverse osmosis, etc. [11, 12]. Among them the most common method is adsorption [4, 9, 12, 13]. Many kinds of activated carbon have been studied during the study of dye adsorption [12–18]. Disadvan tages of activated carbon are its high production and treatment costs [19–21]. Therefore it is necessary to opti mize the adsorption process and develop novel alternative and lowcost adsorbents. In recent years, the nanotechnology methods have been used for the treatment of ordinary water and wastewater [4, 20]. One of the benefits of using nanomaterials can be such properties as their selfassembly, large surface area, and increased reactivity, so they can be potentially useful for water and wastewater remedi ation [3, 20]. The purpose of this paper is to assess the adsorption capacity of nanoalumina in relation to the reactive Red 120 dye ( RR120) removal from aqueous solution. EXPERIMENTAL TECHNIQUE RR120 dye (C44H24C12N14O20S6Na6) produced by Alvan Sabet Co. Ltd. was used without additional purification. The chemical structure of Reactive Red 120 is presented in Fig. 1. A stock solution (1000 mg/dm3) required for subsequent experiments was prepared by dissolving RR120 in distilled water. A Perkin Elmer UV/Vis spectrophotometer of model Lambda 25 was used to determine the maximum absor 125
126
NADAFI et al.
bance (optical density) of the dye observed at the wavelength (λmax = 537 nm) with the absorption spectrum of 200–800 nm. Other chemicals used in this study were provided by Merck Company, Germany. SO3Na
NaO3 S
NaO3S
N=N
N=N
OH N
N N
N
HN
NH
N
N Cl
NaO3 S
HO
NH
HN
SO3Na
Cl
SO3Na
Fig. 1. The structure of Reactive Red 120 Dye.
Characteristics of alumina nanopowder AL2O3 (Nano Pars Lima Co. Ltd.) are as follows: Bulk (apparent) density …………………………………… 0.9 g/cm3 Appearance …………………………………………………… White powder Percent purity ………………………………………………… 99% Special surface area …………………………………………. >160 m2/g Particle size ……………………………………………………. 20 nm Al2O3 ……………………………………………………………. 99% Ca ………………………………………………………………... < 25 ppm V.………………………………………………………………......< 25 ppm Cl ……………………………………………………………….... < 315 ppm Na ..………………………………………………………………. < 70 ppm Mn ………………………………………………………………... < 315 ppm Co ..………………………………………………………………...< 70 ppm As shown in Fig. 2a alumina nanoparticles in the solution have a spherical form with the mean diameter of 20 nm. These spherical particles have tendency to adhere together and form chain structures. The phase
Intensity, a. u.
Al2O3–Gamma
200 nm 20
40
60 2θ, deg (a) (b) Fig. 2. TEM image (a) and XRD pattern (b) of nanoalumina. JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
80
100
Vol. 36
No. 3
2014
REACTIVE RED 120 DYE REMOVAL
127
analysis of nanoalumina was performed by using XRD (Fig. 2b) that confirmed the presence of alumina phase in the sorbent. Adsorption Experiments 3
A stock solution (1000 mg/dm ) was prepared by dissolving the dye in distilled water. Desired concentra tions were obtained from the stock by diluting. The effects of several parameters such as contact time, initial concentration, pH and adsorbent dose were examined. The kinetic of adsorption was determined by analyzing adsorption of the dye at different time intervals. The dye solution (100 cm3) with different dye concentrations (25, 50, 75 and 100 mg/dm3) were put into 200 ml conical flasks containing 0.08 g of nanoalumina. The flasks were agitated in a shaker for mixing the contents during the predetermined time (at the speed of revolution of 150 rpm and pH 7). After centrifuging the dye solution at 4000 rpm for 10 minutes, residual concentrations were determined by spectrophotometer at wavelength λmax = 537 nm for different contact times. Dye adsorp tion experiments were performed to acquire isotherms by using an incubator shaker (model Innova 4340, USA) at the fixed dose of adsorbent (0.8 mg/dm3) during the constant time greater than the equilibrium time. Effects of different doses of adsorbent ranging from 0.2 to 1.2 g/dm3 were examined at various dye contrac tions. The influence of initial pH was studied in a wide range (2–11) at constant adsorbent dose and different initial dye concentrations. 0.1 M HCl and NaOH solutions prepared from analytic grade chemicals were used for pH adjustment. Blank samples were used as controls for each series of experiments. The percentage removal of dye (R, %) was calculated by using the following formula: C –C R = 0t × 100 , C0
(1)
where C0 and Ct are the initial and residual concentrations of dye (mg/dm3) at a given time t, respectively. Amounts of the adsorbed dye on nanoalumina were calculated by using the following formula: C –C q t = 0t V , m
(2)
where qt is the amount of adsorbed dye per unit of nanoalumina (mg/g); V is the volume of aqueous phase (l), and m is the amount of adsorbent (mg). RESULTS AND DISCUSSION Effect of Contact Time and Initial Concentration of Dye Experiments showed that the amount of adsorbed dye gradually increased with the rise of contact time. As follows from Fig. 3, the resultant equilibrium time amounts to 150 min. Also these results reveal that the removal of dye depends on its initial concentration. During the equilibrium time the amount of adsorbed dye (Ve) increases from 29 to 35.65 mg/g as the initial concentration of dye rises from 25 to 100 mg/dm3. For assessing the effect of contact time and initial concentration of dye, a fixed dose of nanoalumina and pH 7 were used for each concentration of the dye. Effect of pH As can be seen from Fig. 4, the dye removal efficiency increases with reduction of pH. In this case the pH value is optimal for the RR120 removal by using nanoalumina. The quantitative adsorption of RR120 was at its maximum at the initial pH of 3 and below. Low pH values lead to an increase in hydrogen ions concentration in the solution, and the surface of nano alumina obtains a positive charge by adsorbing H+ ions [12, 22]. As the nanoalumina surface is positively charged at low pH, a remarkably strong electrostatic attraction appears between the positively charged nano alumina surface and anionic RR120 molecule leading to the maximum removal of dye [23]. Moreover, low adsorption of RR120 dye at higher pH values is also due to the competition for adsorption sites on the part of excess OH ions and anionic RR120 molecules. Arcla has reported that maximum biosorption of RR120 onto native and modified fungus biomass preparations of Lentinus SajorCaju occurs at pH 3 [10]. However, the other study by Adak reported that the maximum removal of crystal violet dye with surfactantmodified alu JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
128
NADAFI et al.
mina was observed at pH 8 [24]. The results show that pH plays an important role in the efficiency of dye removal by using nanoalumina. qt, mg/g 40
30
25 mg/L 20
50 mg/L 75 mg/L 100 mg/L
10
0
40
80
120
160
200
240
280
min
Fig. 3. The effect of contact time and initial concentration on the dye removal efficiency (pH 7, adsorbent dose is 0.8 g/dm3).
Dye removal , %
80
25 mg/L 50 mg/L 75 mg/L 100 mg/L
60
40
20
0
2
4
6
8
10
12 pH Fig. 4. The effect of pH on the dye removal efficiency (adsorbent dose is 0.8 g/dm3, t = 150 min).
Effect of Adsorbent Dose Figure 5 shows the effect of adsorbent dose on the dye removal at different initial concentrations of dye (25, 50, 75, and 100 mg/dm3). As the nanoalumina dose increases from 0.2 to 1.2 g/dm3, the efficiency of dye removal also increases. In this case the values of qe (mg/g) for different dye concentrations decrease with the rise of adsorbent dose. According to data in Fig. 5, the value of qe at the initial concentration of 25 mg/dm3 decreases from 26.69 to 18 mg/g, while at the initial concentration of 100 mg/dm3 the value of qe decreases JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
REACTIVE RED 120 DYE REMOVAL
129
from 51 to 45 mg/g as the adsorbent dose increases from 0.2 to 1.2 g/dm3. It is easily seen that the number of available adsorption sites increases with the increasing adsorbent dose; hence it results in the percentage rise of dye removal. However the experiments show that the amount of adsorbed dye per mass unit decreases with the increasing nanoalumina dose due to the presence of unsaturated sites of adsorbent [22, 25]. Dye removal, %
80
qe, mg/L 25 mg/L 50 mg/L 75 mg/L 100 mg/L
120
80
40 40
0
0.4
0
0.8
1.2 Adsorbent dose
Fig. 5. The effect of adsorbent dose on the dye removal efficiency: (adsorbent dose is 0.8 g/dm3, pH 3, and t = 150 min).
Adsorption Models. Adsorption Kinetics There are several kinds of kinetic models, but the pseudofirst and pseudosecond order models are most common in practical use. The Lagergren pseudofirst order model is one of the most widely used equations for the sorption of solute from a liquid solution [15]. The differential pseudofirst order equation has the form dq t = k1(qe – qt), dt
(3)
where qt and qe are the amounts of adsorbed dye (mg/g) at time t and at equilibrium time, respectively; and k1 is the adsorption rate constant (min). The integration of Eq. 3 with initial conditions qt = 0 at t = 0 leads to the following result: k2 log(qe – qt) = logqe – t . 2.303
(4)
The intercept and slope of the plot of log (qe – qt) versus t relationship were used to determine the values of equilibrium adsorption capacity qe and pseudofirst order rate constant k1, respectively [15]. Figure 6 shows the linear plot of the pseudofirst order kinetic equation at all concentrations of dye. The calculated rate constant, experimental and calculated values of qe and the values of corresponding cor relation coefficients are presented in Table 1, where it is shown that the rate constant k1 increases with a rise in concentration, while correlation coefficients (r2) are relatively small at all values of concentrations. It should be noted that the calculated values of qe for the pseudofirst order kinetic equation poorly agree with experimental values (see Table 1). These results show that the pseudofirst order model cannot be utilized in foretelling the kinetics of the RR120 adsorption onto nanoalumina. Ho's PseudoSecond Order Model [26] The corresponding differential equation can be expressed in the form: JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
130
NADAFI et al.
dq t = k2(qe – qt)2, dt
(5)
where k2 is the equilibrium rate constant calculated for the pseudosecond order adsorption model (g⋅mg–1⋅ min–1). Integrating Eq. 5 with the boundary conditions qt = 0 to qt and t = 0 to t we obtain: 1 1 = + k2t. qe qe – qt
(6)
log(qe – qt)
25 mg/ L 50 mg/ L 75 mg/L 100 mg/L
2
1
0
20
40
60
80
100
120
140
160 min
–1 Fig. 6. Pseudofirst order kinetic equation describing the RR120 dye adsorption onto nanoalumina (adsorbent dose is 0.8 g/dm3, pH 7).
Table 1. Pseudofirst and pseudosecond order adsorption rate constants and the calculated and experimental values of qe for the RR120 dye adsorption onto nanoalumina Sample, mg/dm3 25 50 75 100
Pseudofirst order equation qe(cal), k1, qe(exp), mg/dm3 mg/g min–1 29.8 20.3 1.899 × 10–1 31 24.3 1.911 × 10–3 33.7 18.2 1.711 × 10–3 9.6 2.828 × 10–3 36.6
r2 0.969 0.989 0.962 0.939
Pseudosecond order equation qe(cal), k2, mg/g g⋅mg–1⋅min–1 28.57 9.212 × 10–3 32.25 0.0161 36.03 0.02303 38.48 0.0299
r2 0.987 0.99 0.997 0.997
Equation 6 can be transformed to the following linear form: 1 1 1 = 2 + t . q qt e k2 qe
(7)
h = k2 q e2 ,
(8)
where h is the initial sorption rate determined from the point of interception. The value of qe is the maximum adsorption capacity (mg/g), and k2 is the equilibrium rate constant calculated for pseudosecond order adsorption model (g ⋅mg–1⋅min–1). The value of qe is determined from the slope of the plot of t/qt versus t rela tionship (Fig. 7), k2 can be calculated from the value of the initial sorption rate. The calculated values of qe are in very good agreement with experimental data (see Table 1). The correlation coefficients (r) for pseudosec ond order kinetic model are > 0.96 for all values of concentrations. This indicates that the adsorption of RR120 dye from solution onto nanoalumina obeys the pseudosecond order kinetic model. The same results were observed for nitrate adsorption onto nanoalumina [23].
JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
REACTIVE RED 120 DYE REMOVAL
131
t/qt 25 mg/L 50 mg/L
6
75 mg/L 100 mg/L 4
2
0
20
40
60
80
100
120
140
min
Fig. 7. Pseudosecond order kinetic equation for the RR120 dye adsorption onto nanoalumina: (adsorbent dose is 0.8 g/dm3, pH = 7).
Adsorption Isotherms The analysis of experimental data was performed by using the models based on the Langmuir and Freun dlich isotherms. The linear form of the Langmuir equation can be written in the form: Ce 1 1 = + C e , qm kL qm qe
(9)
where Ce (mg/dm3) is the equilibrium concentration of RR120 in solution, qm is the maximum adsorption capacity corresponding to the monolayer coverage (mg/g), and kL (L/mg) is the Langmuir constant related to sorption energy. The ce/qe versus ce plot is presented by a straight line (Fig. 8), the slope and axisintercept of this line correspond to qm and kL, respectively [17, 27]. The calculated correlation coefficients (r) and the Langmuir constant for RR120 are presented in Table 2. The analysis of this table shows that the maximum sorption capacity of nanoalumina for this dye and the value of r amount to 53.63 mg/g and 0.99 at 25°C, respectively. The value of r from the Langmuir isotherm was greater than that from the Freundlich isotherm for the case of adsorption of the specified dye. This implies that the Langmuir model better describes the RR120 adsorption onto nanoalumina than the Freundlich model. Table 2. Isotherm parameters and correlation coefficients for the RR120 dye adsorption onto nanoalumina T, °C 25
Langmuir isotherm –1
–1
qm (mg⋅g )
kL (L⋅mg )
r2
65.23
0.1583
0.99
Freundlich isotherm RL
n
kF
0.059~0.38
2.873
11.91
r2 0.969
The essential attributes of dimensionless separation factor can be written as follows: R L = 1 , 1 + C0 kL
(10)
where C0 is the initial concentration of RR120 (mg/g). The values of RL in the interval from 0 to 1 indicate the effective adsorption. The adsorption process is irreversible when RL = 0 and linear when RL = 1, however it is unfavorable when RL > 1 [28]. The calculated values of RL for the examined adsorption system reveal that these values lie in the range from 0.059 to 0.38 for dye concentrations ranging from 25 to 100 mg/dm3 at 25°C. Accordingly, these values show that the adsorption process is favorable. Another possible model is the model for adsorption on amorphous surface based on the Freundlich iso therm. It assumes the heterogeneity of surface and the exponential distribution of active sites and their ener gies. The linear form of Freundlich isotherm can be presented as follows [27, 28]:
JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
132
NADAFI et al.
1 logqe = logkF + logCe, n
(11)
kF and n are the Freundlich constants (n is an indicator of how favorable the adsorption process is and kF is the adsorption capacity of adsorbent). These constants can be determined from the linear log qe versus log Ce plot (Fig. 9). kF is the adsorption or distribution coefficient representing the quantity of dye adsorbed onto nanoalumina at a unit equilibrium concentration. 1/qe
1. 2
0. 8
0.4
0
20
40
60
1/C e
Fig. 8. Langmuir adsorption isotherm for the RR120 dye adsorption onto nanoalumina (adsorbent dose is 0.8 g/dm3, pH 3, and T = 25°C).
logqe
1. 6
1. 2
0. 8
0
1
2 logCe
Fig. 9. Freundlich adsorption isotherm for the RR120 dye adsorption onto nanoalumina (adsorbent dose is 0.8 g/dm3, pH 3 and T = 25°C).
The slope (1/n) value ranging from 0 to 1 is a measure of the adsorption intensity or surface heterogeneity; in this case the degree of heterogeneity increases as the specified value gets closer to zero [29]. The magnitude of the exponent (1/n) may serve as an indicator of the favorable adsorption capacity. Hence, at n >1 the favorable adsorption conditions are created [30]. Values of kF and n are calculated from the axis–intercept and slope of the plot (see Table 2). The results suggest that RR120 dye is successfully adsorbed by nanoalumina. However, the values of the correlation coefficient r indicate that the Langmuir isotherm provides the best description of the RR120 adsorption on nanoalumina. JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
REACTIVE RED 120 DYE REMOVAL
133
CONCLUSIONS The present study shows that nanoalumina is quite efficient for the removal of RR120 dye from aqueous solution. The effect of contact time, initial concentration, initial pH value and the adsorbent dose on the effi ciency of dye removal was examined. The optimum pH value for the RR120 adsorption onto nanoalumina was found to be no more than 3. The experiments showed that the kinetics of RR120 adsorption onto nano alumina could be well described by the pseudosecond order model. The simulation based on isotherm showed that the Langmuir isotherm equation could better describe the adsorption on nanoalumina as compared to the Freundlich equation. Nanoalumina can be produced in a simple and cost effective way. This research has been supported by the Tehran University of Medical Sciences, grant 90026113727. REFERENCES 1. Luo, P., Zhao, Y., Zhang, B., Liu, J., Yang, Y., and Liu, J., Water Res., 2010, vol. 44, no. 5, pp. 1489–1497. 2. Forgacs, E., CserhAti, T., and Oros, G., Environ. Int., 2004, vol. 30, no. 7, pp. 953–971. 3. Zhao, M., Tang, Z., and Liu, P., J. Hazard. Materials, 2008, vol. 158, no. 1, pp. 43–51. 4. Moussavi, G. and Mahmoud, M., ibid, 2009, vol. 168, no. 2/3, pp. 806–812. 5. Iram, M., Guo, C., Guan, Y., Ishfaq, A., and Liu, H., Ibid, 2010, vol. 181, no. 1/3, pp. 1039–1050. 6. Ozdemir, O., Turan, M., Turan, A.Z., Faki, A., and Engin, A.B., Ibid, 2009, vol. 166, no. 2/3, pp. 647–654. 7. Gupta, V.K. and Suhas, J., Environ. Management, 2009, vol. 90, no. 8, pp. 2313–2342. 8. Reddy, S.S. and Kotaiah, B., Iran. J. Environ. Health, Sci. and Eng., 2006, vol. 3, no. 4, pp. 239–246. 9. Ehrampoush, M., Ghanizadeh, G., and Ghaneian, M., ibid, 2011, vol. 8, no. 2, pp. 101–106. 10. Arcla, M.Y. and Bayramoglu, G., J. Hazar. Materials, 2007, vol. 149, no. 2, pp. 499–507. 11. Zendehdel, M., Barati, A., Alikhani, H., and Hekmat, A., Iran. J. Environ. Health Sci. and Eng., 2010, vol. 7, no. 5, pp. 431–436. 12. Malik, P.K., J. Hazard. Materials, 2004, vol. 113, no. 1/3, pp. 81–88. 13. Senthilkumaar, S., Kalaamani, P., Porkodi, K., Varadarajan, P.R., and Subburaam, C.V., Biores. Technol., 2006, vol. 97, no. 14, pp. 1618–1625. 14. Orfao, J.J.M., Silva, A.I.M., Pereira, J.C.V., Barata, S.A., Fonseca, I.M., Faria, P.C.C., and Pereira, M.F.R., J. Col loid and Interface Sci., 2006, vol. 296, no. 2, pp. 480–489. 15. Amin, N.K., Desalination, 2008, vol. 223, no. 1/3, pp. 152–161. 16. Namasivayam, C. and Kavitha, D., Dyes and Pigments, 2002, vol. 54, no. 1, pp. 47–58. 17. Santhy, K. and Selvapathy, P., Biores. Technol., 2006, vol. 97, no. 11, pp. 1329–1336. 18. Hameed, B.H., Ahmad, A.L., and Latiff, K.N.A., Dyes and Pigments, 2007, vol. 75, no. 1, pp. 143–149. 19. Aksu, Z., Proc. Biochem., 2005, vol. 40, no. 3/4, pp. 997–1026. 20. Absalan, G., Asadi, M., Kamran, S., Sheikhian, L., and Goltz, D.M., J. Hazard. Materials, 2011, vol. 192, no. 2, pp. 476–484. 21. Garg, V.K., Gupta, R., Bala Yadav, A., and Kumar, R., Biores. Technol., 2003, vol. 89, no. 2, pp. 121–124. 22. Arami, M., Limaee, N.Y., and Mahmoodi, N.M., Chem. Eng. J., 2008, vol. 139, no. 1, pp. 2– 10. 23. Bhatnagar, A., Kumar, E., and Sillanpaa, M., Ibid, 2010, vol. 163, no. 3, pp. 317–323. 24. Adak, A., Bandyopadhyay, M., and Pal, A., Separ. and Purif. Technol., 2005, vol. 44, no. 2, pp. 139–144. 25. Afkhami, A., SaberTehrani, M., and Bagheri, H., Desalination, 2010, vol. 263, no. 1/3, pp. 240–248. 26. Ho, Y.S. and McKay, G., Proc. Biochem., 1999, vol. 34, no. 5, pp. 451–465. 27. Tunali, S., Ozcan, A.S., Ozcan, A., and Gedikbey, T., J. Hazard. Materials, 2006, vol. 135, no. 1/3, pp. 141–148. 28. Tan, I.A.W., Hameed, B.H., and Ahmad, A.L., Chem. Eng. J., 2007, vol. 127, no. 1/3, pp. 111–119. 29. Haghseresht, F. and Lu, G.Q., Energy and Fuels, 1998, vol. 12, no. 6, pp. 1100–1107. 30. Fytianos, K., Voudrias, E., and Kokkalis, E., Chemosphere, 2000, vol. 40, no. 1, pp. 3–6. Translated by A. Zheldak
JOURNAL OF WATER CHEMISTRY AND TECHNOLOGY
Vol. 36
No. 3
2014
