- Email: [email protected]
Extraction of samarium and gadolinium from aqueous nitrate solution with D2EHPA in a pulsed disc and doughnut column
Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Extraction of samarium and gadolinium from aqueous nitrate solution with D2EHPA in a pulsed disc and doughnut column R. Torkaman a,b,*, J. Safdari b, M. Torab-Mostaedi b, M.A. Moosavian a, M. Asadollahzadeh b a b
Oil and Gas Centre of Excellence, School of Chemical Engineering, University College of Engineering, University of Tehran, Tehran, Iran Nuclear Fuel Cycle Research School, Nuclear Science & Technology Research Institute, PO Box 14155-1339, Tehran, Iran
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 May 2014 Received in revised form 2 October 2014 Accepted 12 October 2014 Available online 11 February 2015
The extraction of samarium and gadolinium from aqueous nitrate solution with D2EHPA using a pulsed disc and doughnut column was investigated. It was found from batch experiments for separation of Sm(III) from Gd(III) that the initial aqueous pH, concentration of D2EHPA and concentration of nitric acid as a stripping agent were optimized at 1.5, 0.12 M, and 0.1 M, respectively. In continuous experiments, the effects of variables such as pulsation intensity, dispersed and continuous phase velocity on holdup, mean drop sizes, separation factors and stripping efficiencies were studied. Empirical correlations for prediction of the dispersed phase holdup and mean drop sizes in terms of the operating variables and the physical properties were compared with the experimental results. The results of the continuous experiments demonstrated the feasibility of operating extraction process in the pulsed disc and doughnut column, with good efficiency for separation of samarium from gadolinium in the extraction and stripping stages. ß 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Keywords: Rare earth elements Dispersed phase holdup Pulsed disc and doughnut column Sauter mean drop diameter
1. Introduction Liquid–liquid extraction is an important separation process that is used in a wide range of industries such as petroleum refining, food industry, nuclear fuel processing, pharmaceuticals, biochemistry, metal extraction, waste management and other areas [1]. The purification and separation of rare earth elements (REEs) by the solvent extraction method have more attention worldwide because of their wide applications in many main technological areas [2]. For example, Gadolinium is used in the medical field as a contrasting agent in images created by magnetic resonance, as well as in the nuclear area, as a thermal neutron absorber. As little as 1% of gadolinium in chromium, iron and related alloys improves the resistance to high oxidation and high temperatures. Samarium; another element of rare earths, has an interesting application in the ceramic industry for coloring glass. The development of the samarium-cobalt permanent magnet, with flux densities extremely higher than those of similar current products has turned samarium into an important industrial material [3].
* Corresponding author at: University College of Engineering, University of Tehran, Enghelab Ave., Tehran, Iran. Tel.: +98 21 66409774; fax: +98 21 66480290. E-mail address: [email protected] (R. Torkaman).
With the development of rare earth applications, commercial products develop gradually from primary products to single highpurity and high value-added products. The preparation of highpurity rare earth products for the metallurgical industry is of great interest. Mixer-settlers are virtually the unique extractors used for rare earth purification and separation in the industry. Being stable in operation, easy to startup and less sensitive to changes in feed and environmental conditions can be listed as the main advantages of mixer-settlers. However, large space requirement, long residence time, large solvent inventory and poor sealing of the system are the main problems with the mixer-settlers [1]. Compared with mixer-settlers, columns are found to be highly efficient and economical in respect of stage numbers, settler area, solvent inventory, site area and maintenance. They are suitable for process control, which requires a quick response to changes in operating and environmental conditions [4]. Among the various types of columns, the pulsed extraction columns are recommended for processing the corrosive or radioactive solutions because the pulsing unit can be isolated from the column. Although the rare earth elements are not radioactive, they always coexist with radioactive elements, such as uranium and thorium. Therefore, the pulsed columns are suitable for extraction and separation of rare earth elements. Benedetto and co-workers investigated the separation of samarium and gadolinium by mixer-settlers with 16 stages. A
http://dx.doi.org/10.1016/j.jtice.2014.10.016 1876-1070/ß 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
comparative study on batch and continuous scale, has been carried out with three extractants [5]. Abdeltawab and co-workers studied the separation of La and Ce with PC88A from nitrate media by using a multistage counter-current mixer-settler extraction column in seven stages [6]. A process took place for the recovery of lanthanum with mixersettlers in 22 stages: 8 for extraction, 8 for scrubbing and 6 for stripping by using D2EHPA and HEH(EHP) from chloride medium solution [7]. The recovery of Dy with 98% purity and Y with 93% purity from rare earth chloride solution with mixer-settlers in 20 stages has been reported [8]. Ansari and co-workers employed the mixer–settler unit in 16 stages for extraction of uranium and lanthanides from simulated high-level waste [9]. Takahashi and co-workers investigated the multistage extraction of samarium and gadolinium by means of a mixer–settler extraction column. Then the stage efficiency based on hydrodynamics and mass transfer within the column was studied [10]. Separation factors are typically 1.5–2.5 for neighboring members of rare earth series. The preparation of high-purity products requires 30–60 stages of mixer-settlers. As Megon developed a process for producing high- purity yttrium oxide starting from the xenotime concentrates. The solvent extraction circuit consisted of a selective extraction by D2EHPA followed by four extraction, twelve scrubbing and four stripping unit. Yttrium nitrate solution with other rare earths was fed to the second circuit by using 26, 6, 8 stage for extraction, scrubbing and stripping unit, respectively [11]. Liao and co-workers showed that the reciprocating extraction column is highly promising as an alternative of mixer-settlers in the extraction and separation of rare earth elements (Nd group/Sm group) with D2EHPA extractant [12]. Among the various types of extraction columns, the pulsed disc and doughnut extraction column is one type of extractor whose application has rarely been referred to the literature [13–16]. An installation of the pulsed disc and doughnut columns to treat the uranium solutions from Western Mining Corporations in South Australia has shown that these columns have a clear advantage over the mixer-settlers. The estimated capital expenditure associated with pulsed column is about 20% lower than that of similar size plants using mixer-settlers [13,17]. The success of pulsed columns in the uranium industry is leading to more interest in pulsed columns for other metals- cobalt, zinc, copper and nickel. The use of pulsed columns for the GORO Nickel project in New Caledonia will mark the entrance of pulsed columns into the base metals industry and will lead to applications for other metals, as well [18]. The applications of the pulsed disc and doughnut extraction columns were not observed in the literature for extraction and separation of rare earth elements. Furthermore, the knowledge concerning the design and performance of these columns is still far from satisfactory. The reason relates mainly to the complex behaviors of the hydrodynamics and mass transfer performance. In order to develop the appropriate design procedure, the knowledge of average drop size and dispersed phase holdup in terms of the operating variables, column geometry and liquid physical properties is essential. The empirical correlations have been reported for the dispersed phase holdup and Sauter mean drop diameter in terms of physical properties and operating variables for different operating regimes [19–21]. The purpose of this study was to investigate the feasibility of the pulsed disc and doughnut extraction columns for extraction and separation of samarium and gadolinium from aqueous nitrate solution with D2EHPA extractant and to obtain information on drop size and holdup. The effects of operational variables such as pulsation intensity, dispersed and continuous phase velocities on holdup, drop size and separation factor in extraction and stripping stages were investigated.
19
2. Experimental 2.1. Reagents The commercial extractant, 2-ethylhexyl phosphoric acid (D2EHPA) was purchased from Aldrich. This extractant was dissolved in kerosene to achieve the required concentration. The aqueous working solutions were prepared by diluting gadolinium and samarium nitrate hexahydrate (Sm(NO3)36H2O, Gd(NO3)36H2O, Middle East Ferro Alloy Company, 99.9% purity) in deionized water. The initial concentrations of two metals were maintained at 500 ppm. 2.2. Batch experiments The experiments were performed by contacting equal volumes of aqueous phase and organic phase in a shaker for 30 min at room temperature (initial experiments showed that equilibrium was obtained during 10 min). After extraction, the phases were separated by means of a separation funnel. Metal concentrations in the aqueous phase before and after extraction were determined by using a Perkin-Elmer model 5500 inductively coupled plasmaatomic emission spectroscopy (ICP-AES). The distribution coefficient (D), extraction efficiency (E), separation factor (b) and stripping efficiency (S) are defined as follows: D¼
½M t ½M a ½M a
%E ¼
b¼
(1)
D 100 D þ V aq =V org
(2)
D1 D2
%S ¼
(3) ½M aq;a V aq
½Morg;o V org þ ½M aq;a V aq
100
(4)
where [M]t and [M]a express the initial and final concentrations of metal ions in the aqueous phase, Vaq and Vorg are the volumes of the aqueous and organic phases, [M]aq,a is the equilibrium concentration of metal ion in the stripping acid and [M]org,o is the equilibrium concentration of metal ion in the loaded organic phase, respectively. 2.3. Pulsed disc and doughnut column extraction experiments The extraction experiments were performed at room temperature in a pilot plant using a pulsed disc and doughnut column extraction. The continuous and dispersed phases flowed countercurrently, the aqueous phase in descending direction and the organic phase in ascending direction. The schematic arrangement of column and auxiliary equipment is shown diagrammatically in Fig. 1. In extraction experiments, the nitrate solution of samarium and gadolinium with 500 ppm concentration was prepared for continuous phase. The dispersed phase organic solution contained 0.12 M D2EHPA extractant diluted with kerosene. In stripping experiments, the nitrate solution with 0.1 M concentration was prepared for continuous phase. The loaded organic phase of samarium and gadolinium with 500 ppm concentration was used for the dispersed phase organic solution. The main column section comprised a 76 mm internal diameter glass tube and the effective height of the column was 74 cm. The column divided by 30 pairs of disc and doughnut and was made of stainless steel with a thickness of 2 mm.
20
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25 Table 1 Physical properties of the phases used in the system, at 25 8C.
Aqueous phase Dispersed phase
r (kg/m3)
m (103 kg/m/s)
s (103 N/m)
1005 789
0.9216 1.2587
22.1 –
from each photo. In the case of non-spherical droplets, the major and axes, d1 and d2 were measured and the equivalent diameter, de, was calculated from the following equation: 1=3 de ¼ d21 d2
Fig. 1. Schematic flow diagram of pulsed disc and doughnut column.
(5)
The Sauter mean drop diameter (d32) is an average of droplet size that it is defined as the diameter of a sphere that has the same volume/surface area ratio as a droplet of interest. The drops were classified into 0.1 mm size intervals and the Sauter mean drop diameter was then calculated from the following equation: Pn ni d3i (6) d32 ¼ Pi¼0 n 2 i¼0 ni di where ni is the number of droplets of mean diameter di within a narrow size range i.
The discs and doughnuts were arranged alternately and spaced 10 mm apart, resulting in a 20 mm compartment height; the structure was held in place by means of three tie rods (3.2 mm O.D.) with SS spacer sleeves. The discs and doughnut apertures were 67 and 36 mm in diameter, respectively. The open free area based on these arrangements was 23.5%. Two settlers with 112 mm diameter at each end of the column permitted the liquids to coalesce and decant separately. Pulsation was applied using compressed air at the required amplitude and frequency into the pulse leg. The frequency of the pulses was controlled by means of two solenoid valves while the air pressure was controlled by a regulator to provide pulses of the required amplitude. The flow rates of the organic and liquid phases were controlled by two rotameters. The inlet and outlet of the column were connected to four tanks, each of 80 L capacity. The liquid–liquid interface was maintained approximately 250 mm above the top compartment employing an optical sensor. A solenoid valve was installed in outlet streamline of the heavy phase. When the interface location was going to vary, the optical sensor sent a signal to the solenoid valve and the aqueous phase was allowed to leave the column by opening the diaphragm of the solenoid valve. Consequently, the organic phase was allowed to leave the column with overflow. The amount of gadolinium and samarium transferred to the organic phase in the outlet column was estimated after stripping with 3 M HNO3; so as to analyze the gadolinium and samarium concentrations in the inlet and outlet phases by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The densities of the aqueous and the organic phases were determined by the pycnometer method. The interfacial tension measurements were obtained utilizing a Kru¨ss tensiometer. The viscosity of both phases was measured with DVI-Prime viscometer. Table 1 lists the physical properties of the two phases. 2.3.1. Droplet size measurements Drop sizes were determined by taking a digital photo of the inside column using a Nikon D5000 digital camera. The dimensions of the drops were then determined using AutoCAD software. The sizes of drops were converted to absolute dimensions by comparing the measured value with the disc and doughnut thickness and spacing. Approximately 300 drops were analyzed
2.3.2. Holdup measurements The fractional volumetric holdup is described as the volume fraction of the main section of the column that is occupied by the dispersed phase: xd ¼
vd ve
(7)
where vd and ve represent the volume of the dispersed phase and the total volume of the two phases for the effective length of the column, respectively. The holdup of the dispersed phase was measured by the shutdown method, the pulsation was turned off and the inlet and outlet flows were stopped simultaneously after reaching the steady state. The dispersion was then coalesced at the interface and the holdup was then measured by determining the change of interfacial height. 3. Result and discussion 3.1. Extraction and separation of samarium and gadolinium with D2EHPA in batch experiments The effects of operating parameters such as hydrogen ion, nitrate ion, extractant concentration and stripping acid were investigated to determine maximum separation factor for samarium and gadolinium. Based on the experimental results reported by Torkaman and co-workers [22,23], the extraction reaction of Sm(III) and Gd(III) by D2EHPA can be represented by the following equation: SmðIIIÞðaÞ or GdðIIIÞðaÞ K eq
þ 3H2 A2ðoÞ , SmH3 A6ðoÞ or GdH3 A6ðoÞ þ 3Hþ ðaÞ
(8)
where H2A2 represents the dimeric forms of D2EHPA. Fig. 2 shows the effect of pH on the Sm(III) and Gd(III) extraction. The similar curves of the change in the initial aqueous pH are obtained because two metals have similar properties. The extraction efficiency increases with increasing the initial aqueous pH, the low extraction efficiency is achieved in the lower pH region (pH > 1) because the reaction is shifted preferentially in the back extraction reaction according to Eq. (8). It is evident from Fig. 2 that the maximum
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
21
studied, the results of which are given in Table 2. Although high stripping percentages of both Gd(III) and Sm(III) were observed when using 2 M HNO3 as stripping agent, 0.1 M HNO3 is preferred for stripping as it gave 59.03% for Sm(III) and 32.41% for Gd(III) in a one stage stripping under the experimental conditions used. After two stages, the stripping percent reached 77.21% and 21.5% for Sm(III) and Gd(III), respectively, which may lead to good separation of the two metals. Therefore, the initial aqueous pH, concentration of D2EHPA and nitric acid agent in stripping stage were optimized at 1.5, 0.12 M, and 0.1 M, respectively. 3.2. Extraction and separation of samarium and gadolinium with pulsed disc and doughnut column
Fig. 2. Effect of initial aqueous pH on the percentage extraction of samarium and gadolinium and separation factor ([D2EHPA] = 0.12 M, [Sm(III)] = 500 ppm, [Gd(III)] = 500 ppm, O/A ratio = 1, T = 298.15 K).
separation factor is obtained with pH value equal to 1.5 and further increase in initial aqueous pH results a decrease in separation factor values due to the increasing co-extraction of samarium. The effect of the change in nitrate ion concentration on the separation of Sm(III) and Gd(III) was studied. The nitrate ion concentration varied by means of different concentrations of sodium nitrate from 0.01 to 0.5 M while keeping the initial aqueous pH at 1.5. The results showed that the change in the investigated range had a negligible effect on the separation of samarium and gadolinium by D2EHPA. The variation of separation factor and extraction efficiency with D2EHPA concentration is shown in Fig. 3. The extraction efficiency increases with increasing D2EHPA concentration because the reaction is shifted in the right side of Eq. (1). Also, it can be seen that separation factor values increases from 1.5 to 3.82 with an increase in D2EHPA concentration. The maximum separation factor is obtained when the D2EHPA concentration is equal to 0.12 M and further increase in D2EHPA concentration results a decrease in separation factor values due to the increasing coextraction of samarium. The effect of nitric acid concentrations on the stripping of samarium and gadolinium from the loaded organic solutions was
The operating parameters such as pulsation intensity, dispersed phase and continuous phase velocity varied to investigate the effect of the parameters on the holdup, drop sizes, separation factors and stripping efficiencies in extraction and stripping stages. The disc and doughnut extraction column can operate in three different flow regimes such as mixer–settler, dispersion and emulsion, which depends on the pulse intensity (Af = amplitude frequency) [19]. Our preliminary tests in the pulsed disc and doughnut column were carried out by utilizing different pulse intensities. The different flow regimes were created on the column depending on using A and f. Thus, values for Af that leading to the dispersion regime were selected as the experimental column. 3.2.1. Effect of operating parameters on the holdup The experimental data for dispersed phase holdup under a variety of operating conditions in extraction and stripping stages are presented in Fig. 4. In dispersion regime, the holdup increases with an increase in pulsation intensity that the results of which are shown in Fig. 4a. The inertial and shear forces on droplets enhance drop breakage with an increase in pulsation intensity. It appears that the relative velocity between the two phases decreases with an increase in the number of drops in the column. Accordingly, the value of dispersed holdup will increase in dispersion regime. As can be seen in Fig. 4b, the dispersed phase holdup increases with an increase in dispersed phase velocity. An increase in dispersed phase velocity leads to the increase in holdup according to Eq. (7). Fig. 4c shows the change of continuous phase velocity on dispersed phase holdup. The drag forces between the continuous phase and dispersed drops increase with an increase in the continuous phase velocity. The drop movement will be limited and the residence time will increase. Therefore, the increase in the continuous phase velocity leads to an increase in the holdup. The number of drops in the column increase with an increase in the dispersed phase velocity. The effect of number of drops is larger than residence time on the hold up. Consequently, as can be seen in Fig. 4b and c, the effect of continuous phase velocity on holdup is smaller than that of dispersed phase velocity.
Table 2 Effect of nitric acid concentration on the stripping efficiency (%S). Nitric acid (M)
Fig. 3. Effect of D2EHPA concentration on the percentage extraction of samarium and gadolinium and separation factor (pH = 1.5, [Sm(III)] = 500 ppm, [Gd(III)] = 500 ppm, O/A ratio = 1, T = 298.15 K).
0.01 0.025 0.05 0.075 0.1 0.5 1 2
Stripping efficiency (%S) Gadolinium
Samarium
0.28 0.43 10.68 20.34 32.41 70.36 99.01 99.98
0.82 1.99 31.63 39.12 59.03 90.00 99.26 99.99
22
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
Fig. 4. Effect of pulsation intensity, dispersed and continuous phase velocity on dispersed phase holdup.
3.2.2. Effect of operating parameters on the Sauter mean drop diameter The change on the Sauter mean drop diameter with pulsation intensity is shown in Fig. 5a. The increases in pulsation intensity lead to intense drop breaking and consequently mean drop sizes will decrease. The values of Sauter mean diameter will change gradually at high pulsation intensity. The droplet coalescence is obtained at high pulsation intensity by the further possibility of droplet collision. The elevated rate of coalescence overcomes the great tendency for droplet breakage and the drop sizes are apparently stabilized. Fig. 5b shows the variation of mean drop sizes along the pulsed disc and doughnut column highlighting the effect of the dispersed phase velocity on the performance of the column. An increase in the dispersed phase flow rate tends to increase the Sauter mean drop sizes. The effect of dispersed phase velocity on the mean drop sizes is less significant in comparison to pulsation intensity. Moreover, the continuous phase velocity has no significant change in drop sizes according to the results shown in Fig. 5c. 3.2.3. Correlation for Sauter mean drop diameter and holdup Two correlations are presented for estimating the Sauter mean diameter in a pulsed disc and doughnut column. Delden and coworkers [21] used the proposed correlation in the pulsed sieve plate [24] which is defined as follow: pffiffiffiffiffiffiffiffi n3 n d32 r g n2 md g 1=4 s 4 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ¼ C 1 en1 hc 1=4 s 3=4 s s r s =Drg " ( )# Af c2 þ exp C 3 eð s g=r Þ1=4
(8)
The parameters e, hc, r* and s* represent the fractional free area, compartment height, the density and surface tension of water at 20 8C, respectively. The constant parameters C1, C2, C3, n1, n2, n3 and n4 in Eq. (8) are 2.84, 0.16, 2.59, 0.30, 0.18, 0.14, and 0.06, respectively. This correlation compared with the experimental values of the mean drop size and a relative deviation of 61.3% with Eq. (9) was observed in the experiments. AARE ¼
NDP 1 X jPredicted value Experimental valuej NDP i¼1 Experimental value
(9)
where NDP is the number values of data points. Torab-Mostaedi and co-workers [20] presented the other correlation for estimating the Sauter mean diameter in a pulsed disc and doughnut column. The correlation is represented by the following equation: !0:283 0:29 4 d32 A f rc da rc s pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ¼ 33:53 103 gs m2c s =Drg 0:13 rc s 4 Dr 2:86 md 0:085 hc 0:734 ð1 þ RÞ0:34 cmc mc da rc (10)
2
3 where c ¼ 2p 1 e2 = 3e2 Co2 hc ð A f Þ This correlation is applied for prediction of the mean drop sizes in the present work. The comparison between the experimental results and the predicted values is shown in Fig. 6, which reveals that this correlation is in good agreement with the experimental data. An average relative deviation (ARD) of 9.32% is obtained with Eq. (10).
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
23
Fig. 5. Effect of pulsation intensity, dispersed and continuous phase velocity on Sauter mean drop diameter.
This correlation is used for prediction of the holdup in the present work. The comparison between the experimental data and the predicted values is given in Fig. 7, which shows that this correlation is in good agreement with the experimental results and an average relative deviation (ARD) of 14.12% is obtained with using Eq. (12).
Torab-Mostaedi and co-workers [19] reported the proposed correlation for estimating holdup that is defined as follows: !0:095 !0:35 !0:06 4 Vd4 rc m4d g A f rc xd ¼ 2:57 gs gs rc s 3 0:88 0:91 V Dr 1þ d Vc rc mixer settler !0:2 !0:32 !0:08 4 Vd4 rc m4d g A f rc xd ¼ 12:31 gs gs rc s 3 (12) 0:92 0:60 Vd Dr 1þ transition and emulsion Vc rc
3.2.4. Separation factors and stripping efficiencies for separation of samarium from gadolinium The effect of operating conditions on the extraction and separation of Sm(III) from Gd(III) was studied. The results are shown in Fig. 8. It is observed from Fig. 8a that the separation factors and stripping percents increase with an increase in
Fig. 6. Comparison of the values calculated using Eq. (10) with experimental results.
Fig. 7. Comparison of the values calculated using Eq. (12) with experimental results.
24
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
Fig. 8. Effect of pulsation intensity, dispersed and continuous phase velocity on separation factor and stripping efficiency.
pulsation intensity. At high pulsation intensity, extraction of gadolinium in the extraction stage and the extraction of samarium in the stripping stage increase and an increase in the separation of two metals occurs in extraction and stripping stages. The maximum separation factor was equal to 4.3 when the values of Af, Vc and Vd were 1.5 cm/s, 1.102 and 1.102 mm/s, respectively. The value of b was higher than the value reported by Benedetto and co-workers [5] in mixer-settlers. In the same conditions, the samarium and gadolinium reached the maximum stripping percent values of 95.5 and 52.7, respectively. The hydrodynamic conditions such as flow rates and pulsed intensity in disc and doughnut column leads to an increase in back extraction reaction rate for samarium in comparison with gadolinium. The Sm-D2EHPA complex was weaker than GdD2EHPA complex and its complex is easier to be broken than GdD2EHPA complex. Therefore the high stripping percent >95.5% for samarium is achieved in stripping stage with disc and doughnut column. Fig. 8b illustrates that the extraction and separation of Sm(III) from Gd(III) increase with an increase in the dispersed phase velocity. The increase in the separation factors and the decrease in stripping percents with an increase in the continuous phase velocity are observed from Fig. 8c. This behavior is related to direction of mass transfer and reaction rate. In extraction stage, the forward reaction rate was faster than backward reaction rate. The turbulence of continuous phase around droplets increases with continuous phase velocity and this effect helps to increase the rate of mass transfer and separation factors. In stripping stage, the slow backward reaction rate is more effective than the turbulence of continuous phase and it leads to the decrease in stripping percents for samarium and gadolinium.
4. Conclusions The performance of a pulsed disc and doughnut column for the extraction and separation of Sm(III) from Gd(III) was studied. Before testing in the pulsed column, bench scale studies on the solvent extraction were carried out for selecting the pH of the initial aqueous solution, composition of D2EHPA with kerosene and concentration of nitric acid in stripping stage to achieve the selective separation of samarium from gadolinium. These values were optimized at 1.5, 0.12 M, and 0.1 M, respectively. The tests performed in the continuous system enhanced a proper understanding of the hydrodynamics of the pulsed disc and doughnut column. The hydrodynamic experiments indicated operating variables such as pulsation intensity and dispersed and continuous phase velocity influences on holdup, mean drop sizes, separation factors and stripping efficiencies. The recovery of two metals in stripping stages was found to decrease with increasing continuous phase velocity and to increase with increasing pulsation intensity and dispersed phase velocity. The results of samarium and gadolinium extraction with D2EHPA in the pulsed disc and doughnut column, showed the feasibility of the column for rare earth separations, with a separation factor of 4.3 and stripping efficiencies of 95.5% and 52.7% for Sm(III) and Gd(III), respectively. References [1] Kislik VS. Solvent extraction: classical and novel approaches. Amsterdam: Elsevier; 2012. [2] Topp NE. The chemistry of the rare-earth elements. Amsterdam: Elsevier; 1965. [3] Gupta CK, Krishnamurthy N. Extractive metallurgy of rare-earths. Boca Raton, FL: CRC Press; 2005.
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25 [4] Thornton JD. Science and practice of liquid–liquid extraction. 1st ed. New York, NY: Oxford University Press; 1992. [5] Benedetto JS, Ciminelli VST, Neto JD. Comparison of extractants in the separation of samarium and gadolinium. Miner Eng 1993;6:597–605. [6] Abdeltawab AA, Nii S, Kawaizumi F, Takahashi K. Separation of La and Ce with PC-88A by counter-current mixer-settler extraction column. Sep Purif Technol 2002;26:265–72. [7] Morais CA, Ciminelli VST. Process development for the recovery of high-grade lanthanum by solvent extraction. Hydrometallurgy 2004;73:237–44. [8] Singh DK, Kotekar MK, Singh H. Development of a solvent extraction process for production of nuclear grade dysprosium oxide from a crude concentrate. Desalination 2008;232:49–58. [9] Ansari SA, Prabhu DR, Gujar RB, Kanekar AS, Rajeswari B, Kulkarni MJ, et al. Counter-current extraction of uranium and lanthanides from simulated highlevel waste using N,N,N0 ,N0 -tetraoctyl diglycolamide. Sep Purif Technol 2009;66:118–24. [10] Takahashi K, Abdeltawab AAS, Nii S, Yajima T, Kawaizumi F. Extraction of rare earth metals with a multistage mixer-settler extraction column. Chem Eng Sci 2002;57:469–78. [11] Xie F, Zhang T, Dreisinger D, Doyle F. A critical review on solvent extraction of rare earths from aqueous solutions. Miner Eng 2014;56:10–28. [12] Liao BQ, Wang J, Wan CR. Separation and recovery of rare earths in reciprocating extraction columns. Sep Sci Technol 2005;40:1685–700. [13] Jahya A, Stevens GW, Pratt HRC. Pulsed disc-and-doughnut column performance. Solvent Extr Ion Exch 2009;27:63–82. [14] Lobry E, Gourdon C, Xuereb C, Lasuye T. Liquid–liquid dispersion in co-current disc and doughnut pulsed column effect of the operating conditions, physical properties and materials parameters. Chem Eng J 2013;233:24–38.
25
[15] Torab-Mostaedi M, Ghaemi A, Asadollahzadeh M. Prediction of mass transfer coefficients in pulsed disc and doughnut extraction column. Can J Chem Eng 2012;90:1569–77. [16] Delden MLV, Vos GS, Kuipers NJM, Haan ABD. Extraction of caprolactam with an alternative benign solvent in a pulsed disc and doughnut column. Solvent Extr Ion Exch 2007;25:639–64. [17] Movsowitz RL, Kleinberger R, Buchalter EM. Application of Bateman pulse columns for uranium solvent extraction. Israel: Bateman Projects Ltd; 1997. [18] Vanca MF. Pulsed column and mixer-settler applications in solvent extraction. JOM 2003;55:43–5. [19] Torab-Mostaedi M, Jalilvand H, Outokesh M. Dispersed phase holdup in a pulsed disc and doughnut extraction column. Braz J Chem Eng 2011;28: 313–23. [20] Torab-Mostaedi M, Ghaemi A, Asadollahzadeh M. Flooding and drop size in a pulsed disc and doughnut extraction column. Chem Eng Res Des 2011;89: 2742–51. [21] Delden MLV, Vos GS, Kuipers NJM, Haan ABD. Extraction of caprolactam with toluene in a pulsed disc and doughnut column—Part II: Experimental evaluation of the hydraulic characteristics. Solvent Extr Ion Exch 2006;24:519–38. [22] Torkaman R, Moosavian MA, Torab-Mostaedi M, Safdari J. Solvent extraction of samarium from aqueous nitrate solution by Cyanex301 and D2EHPA. Hydrometallurgy 2013;137:101–7. [23] Torkaman R, Moosavian MA, Safdari J, Torab-Mostaedi M. Synergistic extraction of gadolinium from nitrate media by mixtures of bis(2,4,4-trimethylpentyl) dithiophosphinic acid and di-(2-ethylhexyl) phosphoric acid. Annal Nucl Energy 2013;62:284–90. [24] Kumar A, Hartland S. Unified correlations for the prediction of drop size in liquid–liquid extraction columns. Ind Eng Chem Res 1996;35:2682–95.
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Extraction of samarium and gadolinium from aqueous nitrate solution with D2EHPA in a pulsed disc and doughnut column R. Torkaman a,b,*, J. Safdari b, M. Torab-Mostaedi b, M.A. Moosavian a, M. Asadollahzadeh b a b
Oil and Gas Centre of Excellence, School of Chemical Engineering, University College of Engineering, University of Tehran, Tehran, Iran Nuclear Fuel Cycle Research School, Nuclear Science & Technology Research Institute, PO Box 14155-1339, Tehran, Iran
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 May 2014 Received in revised form 2 October 2014 Accepted 12 October 2014 Available online 11 February 2015
The extraction of samarium and gadolinium from aqueous nitrate solution with D2EHPA using a pulsed disc and doughnut column was investigated. It was found from batch experiments for separation of Sm(III) from Gd(III) that the initial aqueous pH, concentration of D2EHPA and concentration of nitric acid as a stripping agent were optimized at 1.5, 0.12 M, and 0.1 M, respectively. In continuous experiments, the effects of variables such as pulsation intensity, dispersed and continuous phase velocity on holdup, mean drop sizes, separation factors and stripping efficiencies were studied. Empirical correlations for prediction of the dispersed phase holdup and mean drop sizes in terms of the operating variables and the physical properties were compared with the experimental results. The results of the continuous experiments demonstrated the feasibility of operating extraction process in the pulsed disc and doughnut column, with good efficiency for separation of samarium from gadolinium in the extraction and stripping stages. ß 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Keywords: Rare earth elements Dispersed phase holdup Pulsed disc and doughnut column Sauter mean drop diameter
1. Introduction Liquid–liquid extraction is an important separation process that is used in a wide range of industries such as petroleum refining, food industry, nuclear fuel processing, pharmaceuticals, biochemistry, metal extraction, waste management and other areas [1]. The purification and separation of rare earth elements (REEs) by the solvent extraction method have more attention worldwide because of their wide applications in many main technological areas [2]. For example, Gadolinium is used in the medical field as a contrasting agent in images created by magnetic resonance, as well as in the nuclear area, as a thermal neutron absorber. As little as 1% of gadolinium in chromium, iron and related alloys improves the resistance to high oxidation and high temperatures. Samarium; another element of rare earths, has an interesting application in the ceramic industry for coloring glass. The development of the samarium-cobalt permanent magnet, with flux densities extremely higher than those of similar current products has turned samarium into an important industrial material [3].
* Corresponding author at: University College of Engineering, University of Tehran, Enghelab Ave., Tehran, Iran. Tel.: +98 21 66409774; fax: +98 21 66480290. E-mail address: [email protected] (R. Torkaman).
With the development of rare earth applications, commercial products develop gradually from primary products to single highpurity and high value-added products. The preparation of highpurity rare earth products for the metallurgical industry is of great interest. Mixer-settlers are virtually the unique extractors used for rare earth purification and separation in the industry. Being stable in operation, easy to startup and less sensitive to changes in feed and environmental conditions can be listed as the main advantages of mixer-settlers. However, large space requirement, long residence time, large solvent inventory and poor sealing of the system are the main problems with the mixer-settlers [1]. Compared with mixer-settlers, columns are found to be highly efficient and economical in respect of stage numbers, settler area, solvent inventory, site area and maintenance. They are suitable for process control, which requires a quick response to changes in operating and environmental conditions [4]. Among the various types of columns, the pulsed extraction columns are recommended for processing the corrosive or radioactive solutions because the pulsing unit can be isolated from the column. Although the rare earth elements are not radioactive, they always coexist with radioactive elements, such as uranium and thorium. Therefore, the pulsed columns are suitable for extraction and separation of rare earth elements. Benedetto and co-workers investigated the separation of samarium and gadolinium by mixer-settlers with 16 stages. A
http://dx.doi.org/10.1016/j.jtice.2014.10.016 1876-1070/ß 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
comparative study on batch and continuous scale, has been carried out with three extractants [5]. Abdeltawab and co-workers studied the separation of La and Ce with PC88A from nitrate media by using a multistage counter-current mixer-settler extraction column in seven stages [6]. A process took place for the recovery of lanthanum with mixersettlers in 22 stages: 8 for extraction, 8 for scrubbing and 6 for stripping by using D2EHPA and HEH(EHP) from chloride medium solution [7]. The recovery of Dy with 98% purity and Y with 93% purity from rare earth chloride solution with mixer-settlers in 20 stages has been reported [8]. Ansari and co-workers employed the mixer–settler unit in 16 stages for extraction of uranium and lanthanides from simulated high-level waste [9]. Takahashi and co-workers investigated the multistage extraction of samarium and gadolinium by means of a mixer–settler extraction column. Then the stage efficiency based on hydrodynamics and mass transfer within the column was studied [10]. Separation factors are typically 1.5–2.5 for neighboring members of rare earth series. The preparation of high-purity products requires 30–60 stages of mixer-settlers. As Megon developed a process for producing high- purity yttrium oxide starting from the xenotime concentrates. The solvent extraction circuit consisted of a selective extraction by D2EHPA followed by four extraction, twelve scrubbing and four stripping unit. Yttrium nitrate solution with other rare earths was fed to the second circuit by using 26, 6, 8 stage for extraction, scrubbing and stripping unit, respectively [11]. Liao and co-workers showed that the reciprocating extraction column is highly promising as an alternative of mixer-settlers in the extraction and separation of rare earth elements (Nd group/Sm group) with D2EHPA extractant [12]. Among the various types of extraction columns, the pulsed disc and doughnut extraction column is one type of extractor whose application has rarely been referred to the literature [13–16]. An installation of the pulsed disc and doughnut columns to treat the uranium solutions from Western Mining Corporations in South Australia has shown that these columns have a clear advantage over the mixer-settlers. The estimated capital expenditure associated with pulsed column is about 20% lower than that of similar size plants using mixer-settlers [13,17]. The success of pulsed columns in the uranium industry is leading to more interest in pulsed columns for other metals- cobalt, zinc, copper and nickel. The use of pulsed columns for the GORO Nickel project in New Caledonia will mark the entrance of pulsed columns into the base metals industry and will lead to applications for other metals, as well [18]. The applications of the pulsed disc and doughnut extraction columns were not observed in the literature for extraction and separation of rare earth elements. Furthermore, the knowledge concerning the design and performance of these columns is still far from satisfactory. The reason relates mainly to the complex behaviors of the hydrodynamics and mass transfer performance. In order to develop the appropriate design procedure, the knowledge of average drop size and dispersed phase holdup in terms of the operating variables, column geometry and liquid physical properties is essential. The empirical correlations have been reported for the dispersed phase holdup and Sauter mean drop diameter in terms of physical properties and operating variables for different operating regimes [19–21]. The purpose of this study was to investigate the feasibility of the pulsed disc and doughnut extraction columns for extraction and separation of samarium and gadolinium from aqueous nitrate solution with D2EHPA extractant and to obtain information on drop size and holdup. The effects of operational variables such as pulsation intensity, dispersed and continuous phase velocities on holdup, drop size and separation factor in extraction and stripping stages were investigated.
19
2. Experimental 2.1. Reagents The commercial extractant, 2-ethylhexyl phosphoric acid (D2EHPA) was purchased from Aldrich. This extractant was dissolved in kerosene to achieve the required concentration. The aqueous working solutions were prepared by diluting gadolinium and samarium nitrate hexahydrate (Sm(NO3)36H2O, Gd(NO3)36H2O, Middle East Ferro Alloy Company, 99.9% purity) in deionized water. The initial concentrations of two metals were maintained at 500 ppm. 2.2. Batch experiments The experiments were performed by contacting equal volumes of aqueous phase and organic phase in a shaker for 30 min at room temperature (initial experiments showed that equilibrium was obtained during 10 min). After extraction, the phases were separated by means of a separation funnel. Metal concentrations in the aqueous phase before and after extraction were determined by using a Perkin-Elmer model 5500 inductively coupled plasmaatomic emission spectroscopy (ICP-AES). The distribution coefficient (D), extraction efficiency (E), separation factor (b) and stripping efficiency (S) are defined as follows: D¼
½M t ½M a ½M a
%E ¼
b¼
(1)
D 100 D þ V aq =V org
(2)
D1 D2
%S ¼
(3) ½M aq;a V aq
½Morg;o V org þ ½M aq;a V aq
100
(4)
where [M]t and [M]a express the initial and final concentrations of metal ions in the aqueous phase, Vaq and Vorg are the volumes of the aqueous and organic phases, [M]aq,a is the equilibrium concentration of metal ion in the stripping acid and [M]org,o is the equilibrium concentration of metal ion in the loaded organic phase, respectively. 2.3. Pulsed disc and doughnut column extraction experiments The extraction experiments were performed at room temperature in a pilot plant using a pulsed disc and doughnut column extraction. The continuous and dispersed phases flowed countercurrently, the aqueous phase in descending direction and the organic phase in ascending direction. The schematic arrangement of column and auxiliary equipment is shown diagrammatically in Fig. 1. In extraction experiments, the nitrate solution of samarium and gadolinium with 500 ppm concentration was prepared for continuous phase. The dispersed phase organic solution contained 0.12 M D2EHPA extractant diluted with kerosene. In stripping experiments, the nitrate solution with 0.1 M concentration was prepared for continuous phase. The loaded organic phase of samarium and gadolinium with 500 ppm concentration was used for the dispersed phase organic solution. The main column section comprised a 76 mm internal diameter glass tube and the effective height of the column was 74 cm. The column divided by 30 pairs of disc and doughnut and was made of stainless steel with a thickness of 2 mm.
20
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25 Table 1 Physical properties of the phases used in the system, at 25 8C.
Aqueous phase Dispersed phase
r (kg/m3)
m (103 kg/m/s)
s (103 N/m)
1005 789
0.9216 1.2587
22.1 –
from each photo. In the case of non-spherical droplets, the major and axes, d1 and d2 were measured and the equivalent diameter, de, was calculated from the following equation: 1=3 de ¼ d21 d2
Fig. 1. Schematic flow diagram of pulsed disc and doughnut column.
(5)
The Sauter mean drop diameter (d32) is an average of droplet size that it is defined as the diameter of a sphere that has the same volume/surface area ratio as a droplet of interest. The drops were classified into 0.1 mm size intervals and the Sauter mean drop diameter was then calculated from the following equation: Pn ni d3i (6) d32 ¼ Pi¼0 n 2 i¼0 ni di where ni is the number of droplets of mean diameter di within a narrow size range i.
The discs and doughnuts were arranged alternately and spaced 10 mm apart, resulting in a 20 mm compartment height; the structure was held in place by means of three tie rods (3.2 mm O.D.) with SS spacer sleeves. The discs and doughnut apertures were 67 and 36 mm in diameter, respectively. The open free area based on these arrangements was 23.5%. Two settlers with 112 mm diameter at each end of the column permitted the liquids to coalesce and decant separately. Pulsation was applied using compressed air at the required amplitude and frequency into the pulse leg. The frequency of the pulses was controlled by means of two solenoid valves while the air pressure was controlled by a regulator to provide pulses of the required amplitude. The flow rates of the organic and liquid phases were controlled by two rotameters. The inlet and outlet of the column were connected to four tanks, each of 80 L capacity. The liquid–liquid interface was maintained approximately 250 mm above the top compartment employing an optical sensor. A solenoid valve was installed in outlet streamline of the heavy phase. When the interface location was going to vary, the optical sensor sent a signal to the solenoid valve and the aqueous phase was allowed to leave the column by opening the diaphragm of the solenoid valve. Consequently, the organic phase was allowed to leave the column with overflow. The amount of gadolinium and samarium transferred to the organic phase in the outlet column was estimated after stripping with 3 M HNO3; so as to analyze the gadolinium and samarium concentrations in the inlet and outlet phases by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The densities of the aqueous and the organic phases were determined by the pycnometer method. The interfacial tension measurements were obtained utilizing a Kru¨ss tensiometer. The viscosity of both phases was measured with DVI-Prime viscometer. Table 1 lists the physical properties of the two phases. 2.3.1. Droplet size measurements Drop sizes were determined by taking a digital photo of the inside column using a Nikon D5000 digital camera. The dimensions of the drops were then determined using AutoCAD software. The sizes of drops were converted to absolute dimensions by comparing the measured value with the disc and doughnut thickness and spacing. Approximately 300 drops were analyzed
2.3.2. Holdup measurements The fractional volumetric holdup is described as the volume fraction of the main section of the column that is occupied by the dispersed phase: xd ¼
vd ve
(7)
where vd and ve represent the volume of the dispersed phase and the total volume of the two phases for the effective length of the column, respectively. The holdup of the dispersed phase was measured by the shutdown method, the pulsation was turned off and the inlet and outlet flows were stopped simultaneously after reaching the steady state. The dispersion was then coalesced at the interface and the holdup was then measured by determining the change of interfacial height. 3. Result and discussion 3.1. Extraction and separation of samarium and gadolinium with D2EHPA in batch experiments The effects of operating parameters such as hydrogen ion, nitrate ion, extractant concentration and stripping acid were investigated to determine maximum separation factor for samarium and gadolinium. Based on the experimental results reported by Torkaman and co-workers [22,23], the extraction reaction of Sm(III) and Gd(III) by D2EHPA can be represented by the following equation: SmðIIIÞðaÞ or GdðIIIÞðaÞ K eq
þ 3H2 A2ðoÞ , SmH3 A6ðoÞ or GdH3 A6ðoÞ þ 3Hþ ðaÞ
(8)
where H2A2 represents the dimeric forms of D2EHPA. Fig. 2 shows the effect of pH on the Sm(III) and Gd(III) extraction. The similar curves of the change in the initial aqueous pH are obtained because two metals have similar properties. The extraction efficiency increases with increasing the initial aqueous pH, the low extraction efficiency is achieved in the lower pH region (pH > 1) because the reaction is shifted preferentially in the back extraction reaction according to Eq. (8). It is evident from Fig. 2 that the maximum
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
21
studied, the results of which are given in Table 2. Although high stripping percentages of both Gd(III) and Sm(III) were observed when using 2 M HNO3 as stripping agent, 0.1 M HNO3 is preferred for stripping as it gave 59.03% for Sm(III) and 32.41% for Gd(III) in a one stage stripping under the experimental conditions used. After two stages, the stripping percent reached 77.21% and 21.5% for Sm(III) and Gd(III), respectively, which may lead to good separation of the two metals. Therefore, the initial aqueous pH, concentration of D2EHPA and nitric acid agent in stripping stage were optimized at 1.5, 0.12 M, and 0.1 M, respectively. 3.2. Extraction and separation of samarium and gadolinium with pulsed disc and doughnut column
Fig. 2. Effect of initial aqueous pH on the percentage extraction of samarium and gadolinium and separation factor ([D2EHPA] = 0.12 M, [Sm(III)] = 500 ppm, [Gd(III)] = 500 ppm, O/A ratio = 1, T = 298.15 K).
separation factor is obtained with pH value equal to 1.5 and further increase in initial aqueous pH results a decrease in separation factor values due to the increasing co-extraction of samarium. The effect of the change in nitrate ion concentration on the separation of Sm(III) and Gd(III) was studied. The nitrate ion concentration varied by means of different concentrations of sodium nitrate from 0.01 to 0.5 M while keeping the initial aqueous pH at 1.5. The results showed that the change in the investigated range had a negligible effect on the separation of samarium and gadolinium by D2EHPA. The variation of separation factor and extraction efficiency with D2EHPA concentration is shown in Fig. 3. The extraction efficiency increases with increasing D2EHPA concentration because the reaction is shifted in the right side of Eq. (1). Also, it can be seen that separation factor values increases from 1.5 to 3.82 with an increase in D2EHPA concentration. The maximum separation factor is obtained when the D2EHPA concentration is equal to 0.12 M and further increase in D2EHPA concentration results a decrease in separation factor values due to the increasing coextraction of samarium. The effect of nitric acid concentrations on the stripping of samarium and gadolinium from the loaded organic solutions was
The operating parameters such as pulsation intensity, dispersed phase and continuous phase velocity varied to investigate the effect of the parameters on the holdup, drop sizes, separation factors and stripping efficiencies in extraction and stripping stages. The disc and doughnut extraction column can operate in three different flow regimes such as mixer–settler, dispersion and emulsion, which depends on the pulse intensity (Af = amplitude frequency) [19]. Our preliminary tests in the pulsed disc and doughnut column were carried out by utilizing different pulse intensities. The different flow regimes were created on the column depending on using A and f. Thus, values for Af that leading to the dispersion regime were selected as the experimental column. 3.2.1. Effect of operating parameters on the holdup The experimental data for dispersed phase holdup under a variety of operating conditions in extraction and stripping stages are presented in Fig. 4. In dispersion regime, the holdup increases with an increase in pulsation intensity that the results of which are shown in Fig. 4a. The inertial and shear forces on droplets enhance drop breakage with an increase in pulsation intensity. It appears that the relative velocity between the two phases decreases with an increase in the number of drops in the column. Accordingly, the value of dispersed holdup will increase in dispersion regime. As can be seen in Fig. 4b, the dispersed phase holdup increases with an increase in dispersed phase velocity. An increase in dispersed phase velocity leads to the increase in holdup according to Eq. (7). Fig. 4c shows the change of continuous phase velocity on dispersed phase holdup. The drag forces between the continuous phase and dispersed drops increase with an increase in the continuous phase velocity. The drop movement will be limited and the residence time will increase. Therefore, the increase in the continuous phase velocity leads to an increase in the holdup. The number of drops in the column increase with an increase in the dispersed phase velocity. The effect of number of drops is larger than residence time on the hold up. Consequently, as can be seen in Fig. 4b and c, the effect of continuous phase velocity on holdup is smaller than that of dispersed phase velocity.
Table 2 Effect of nitric acid concentration on the stripping efficiency (%S). Nitric acid (M)
Fig. 3. Effect of D2EHPA concentration on the percentage extraction of samarium and gadolinium and separation factor (pH = 1.5, [Sm(III)] = 500 ppm, [Gd(III)] = 500 ppm, O/A ratio = 1, T = 298.15 K).
0.01 0.025 0.05 0.075 0.1 0.5 1 2
Stripping efficiency (%S) Gadolinium
Samarium
0.28 0.43 10.68 20.34 32.41 70.36 99.01 99.98
0.82 1.99 31.63 39.12 59.03 90.00 99.26 99.99
22
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
Fig. 4. Effect of pulsation intensity, dispersed and continuous phase velocity on dispersed phase holdup.
3.2.2. Effect of operating parameters on the Sauter mean drop diameter The change on the Sauter mean drop diameter with pulsation intensity is shown in Fig. 5a. The increases in pulsation intensity lead to intense drop breaking and consequently mean drop sizes will decrease. The values of Sauter mean diameter will change gradually at high pulsation intensity. The droplet coalescence is obtained at high pulsation intensity by the further possibility of droplet collision. The elevated rate of coalescence overcomes the great tendency for droplet breakage and the drop sizes are apparently stabilized. Fig. 5b shows the variation of mean drop sizes along the pulsed disc and doughnut column highlighting the effect of the dispersed phase velocity on the performance of the column. An increase in the dispersed phase flow rate tends to increase the Sauter mean drop sizes. The effect of dispersed phase velocity on the mean drop sizes is less significant in comparison to pulsation intensity. Moreover, the continuous phase velocity has no significant change in drop sizes according to the results shown in Fig. 5c. 3.2.3. Correlation for Sauter mean drop diameter and holdup Two correlations are presented for estimating the Sauter mean diameter in a pulsed disc and doughnut column. Delden and coworkers [21] used the proposed correlation in the pulsed sieve plate [24] which is defined as follow: pffiffiffiffiffiffiffiffi n3 n d32 r g n2 md g 1=4 s 4 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ¼ C 1 en1 hc 1=4 s 3=4 s s r s =Drg " ( )# Af c2 þ exp C 3 eð s g=r Þ1=4
(8)
The parameters e, hc, r* and s* represent the fractional free area, compartment height, the density and surface tension of water at 20 8C, respectively. The constant parameters C1, C2, C3, n1, n2, n3 and n4 in Eq. (8) are 2.84, 0.16, 2.59, 0.30, 0.18, 0.14, and 0.06, respectively. This correlation compared with the experimental values of the mean drop size and a relative deviation of 61.3% with Eq. (9) was observed in the experiments. AARE ¼
NDP 1 X jPredicted value Experimental valuej NDP i¼1 Experimental value
(9)
where NDP is the number values of data points. Torab-Mostaedi and co-workers [20] presented the other correlation for estimating the Sauter mean diameter in a pulsed disc and doughnut column. The correlation is represented by the following equation: !0:283 0:29 4 d32 A f rc da rc s pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ¼ 33:53 103 gs m2c s =Drg 0:13 rc s 4 Dr 2:86 md 0:085 hc 0:734 ð1 þ RÞ0:34 cmc mc da rc (10)
2
3 where c ¼ 2p 1 e2 = 3e2 Co2 hc ð A f Þ This correlation is applied for prediction of the mean drop sizes in the present work. The comparison between the experimental results and the predicted values is shown in Fig. 6, which reveals that this correlation is in good agreement with the experimental data. An average relative deviation (ARD) of 9.32% is obtained with Eq. (10).
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
23
Fig. 5. Effect of pulsation intensity, dispersed and continuous phase velocity on Sauter mean drop diameter.
This correlation is used for prediction of the holdup in the present work. The comparison between the experimental data and the predicted values is given in Fig. 7, which shows that this correlation is in good agreement with the experimental results and an average relative deviation (ARD) of 14.12% is obtained with using Eq. (12).
Torab-Mostaedi and co-workers [19] reported the proposed correlation for estimating holdup that is defined as follows: !0:095 !0:35 !0:06 4 Vd4 rc m4d g A f rc xd ¼ 2:57 gs gs rc s 3 0:88 0:91 V Dr 1þ d Vc rc mixer settler !0:2 !0:32 !0:08 4 Vd4 rc m4d g A f rc xd ¼ 12:31 gs gs rc s 3 (12) 0:92 0:60 Vd Dr 1þ transition and emulsion Vc rc
3.2.4. Separation factors and stripping efficiencies for separation of samarium from gadolinium The effect of operating conditions on the extraction and separation of Sm(III) from Gd(III) was studied. The results are shown in Fig. 8. It is observed from Fig. 8a that the separation factors and stripping percents increase with an increase in
Fig. 6. Comparison of the values calculated using Eq. (10) with experimental results.
Fig. 7. Comparison of the values calculated using Eq. (12) with experimental results.
24
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25
Fig. 8. Effect of pulsation intensity, dispersed and continuous phase velocity on separation factor and stripping efficiency.
pulsation intensity. At high pulsation intensity, extraction of gadolinium in the extraction stage and the extraction of samarium in the stripping stage increase and an increase in the separation of two metals occurs in extraction and stripping stages. The maximum separation factor was equal to 4.3 when the values of Af, Vc and Vd were 1.5 cm/s, 1.102 and 1.102 mm/s, respectively. The value of b was higher than the value reported by Benedetto and co-workers [5] in mixer-settlers. In the same conditions, the samarium and gadolinium reached the maximum stripping percent values of 95.5 and 52.7, respectively. The hydrodynamic conditions such as flow rates and pulsed intensity in disc and doughnut column leads to an increase in back extraction reaction rate for samarium in comparison with gadolinium. The Sm-D2EHPA complex was weaker than GdD2EHPA complex and its complex is easier to be broken than GdD2EHPA complex. Therefore the high stripping percent >95.5% for samarium is achieved in stripping stage with disc and doughnut column. Fig. 8b illustrates that the extraction and separation of Sm(III) from Gd(III) increase with an increase in the dispersed phase velocity. The increase in the separation factors and the decrease in stripping percents with an increase in the continuous phase velocity are observed from Fig. 8c. This behavior is related to direction of mass transfer and reaction rate. In extraction stage, the forward reaction rate was faster than backward reaction rate. The turbulence of continuous phase around droplets increases with continuous phase velocity and this effect helps to increase the rate of mass transfer and separation factors. In stripping stage, the slow backward reaction rate is more effective than the turbulence of continuous phase and it leads to the decrease in stripping percents for samarium and gadolinium.
4. Conclusions The performance of a pulsed disc and doughnut column for the extraction and separation of Sm(III) from Gd(III) was studied. Before testing in the pulsed column, bench scale studies on the solvent extraction were carried out for selecting the pH of the initial aqueous solution, composition of D2EHPA with kerosene and concentration of nitric acid in stripping stage to achieve the selective separation of samarium from gadolinium. These values were optimized at 1.5, 0.12 M, and 0.1 M, respectively. The tests performed in the continuous system enhanced a proper understanding of the hydrodynamics of the pulsed disc and doughnut column. The hydrodynamic experiments indicated operating variables such as pulsation intensity and dispersed and continuous phase velocity influences on holdup, mean drop sizes, separation factors and stripping efficiencies. The recovery of two metals in stripping stages was found to decrease with increasing continuous phase velocity and to increase with increasing pulsation intensity and dispersed phase velocity. The results of samarium and gadolinium extraction with D2EHPA in the pulsed disc and doughnut column, showed the feasibility of the column for rare earth separations, with a separation factor of 4.3 and stripping efficiencies of 95.5% and 52.7% for Sm(III) and Gd(III), respectively. References [1] Kislik VS. Solvent extraction: classical and novel approaches. Amsterdam: Elsevier; 2012. [2] Topp NE. The chemistry of the rare-earth elements. Amsterdam: Elsevier; 1965. [3] Gupta CK, Krishnamurthy N. Extractive metallurgy of rare-earths. Boca Raton, FL: CRC Press; 2005.
R. Torkaman et al. / Journal of the Taiwan Institute of Chemical Engineers 48 (2015) 18–25 [4] Thornton JD. Science and practice of liquid–liquid extraction. 1st ed. New York, NY: Oxford University Press; 1992. [5] Benedetto JS, Ciminelli VST, Neto JD. Comparison of extractants in the separation of samarium and gadolinium. Miner Eng 1993;6:597–605. [6] Abdeltawab AA, Nii S, Kawaizumi F, Takahashi K. Separation of La and Ce with PC-88A by counter-current mixer-settler extraction column. Sep Purif Technol 2002;26:265–72. [7] Morais CA, Ciminelli VST. Process development for the recovery of high-grade lanthanum by solvent extraction. Hydrometallurgy 2004;73:237–44. [8] Singh DK, Kotekar MK, Singh H. Development of a solvent extraction process for production of nuclear grade dysprosium oxide from a crude concentrate. Desalination 2008;232:49–58. [9] Ansari SA, Prabhu DR, Gujar RB, Kanekar AS, Rajeswari B, Kulkarni MJ, et al. Counter-current extraction of uranium and lanthanides from simulated highlevel waste using N,N,N0 ,N0 -tetraoctyl diglycolamide. Sep Purif Technol 2009;66:118–24. [10] Takahashi K, Abdeltawab AAS, Nii S, Yajima T, Kawaizumi F. Extraction of rare earth metals with a multistage mixer-settler extraction column. Chem Eng Sci 2002;57:469–78. [11] Xie F, Zhang T, Dreisinger D, Doyle F. A critical review on solvent extraction of rare earths from aqueous solutions. Miner Eng 2014;56:10–28. [12] Liao BQ, Wang J, Wan CR. Separation and recovery of rare earths in reciprocating extraction columns. Sep Sci Technol 2005;40:1685–700. [13] Jahya A, Stevens GW, Pratt HRC. Pulsed disc-and-doughnut column performance. Solvent Extr Ion Exch 2009;27:63–82. [14] Lobry E, Gourdon C, Xuereb C, Lasuye T. Liquid–liquid dispersion in co-current disc and doughnut pulsed column effect of the operating conditions, physical properties and materials parameters. Chem Eng J 2013;233:24–38.
25
[15] Torab-Mostaedi M, Ghaemi A, Asadollahzadeh M. Prediction of mass transfer coefficients in pulsed disc and doughnut extraction column. Can J Chem Eng 2012;90:1569–77. [16] Delden MLV, Vos GS, Kuipers NJM, Haan ABD. Extraction of caprolactam with an alternative benign solvent in a pulsed disc and doughnut column. Solvent Extr Ion Exch 2007;25:639–64. [17] Movsowitz RL, Kleinberger R, Buchalter EM. Application of Bateman pulse columns for uranium solvent extraction. Israel: Bateman Projects Ltd; 1997. [18] Vanca MF. Pulsed column and mixer-settler applications in solvent extraction. JOM 2003;55:43–5. [19] Torab-Mostaedi M, Jalilvand H, Outokesh M. Dispersed phase holdup in a pulsed disc and doughnut extraction column. Braz J Chem Eng 2011;28: 313–23. [20] Torab-Mostaedi M, Ghaemi A, Asadollahzadeh M. Flooding and drop size in a pulsed disc and doughnut extraction column. Chem Eng Res Des 2011;89: 2742–51. [21] Delden MLV, Vos GS, Kuipers NJM, Haan ABD. Extraction of caprolactam with toluene in a pulsed disc and doughnut column—Part II: Experimental evaluation of the hydraulic characteristics. Solvent Extr Ion Exch 2006;24:519–38. [22] Torkaman R, Moosavian MA, Torab-Mostaedi M, Safdari J. Solvent extraction of samarium from aqueous nitrate solution by Cyanex301 and D2EHPA. Hydrometallurgy 2013;137:101–7. [23] Torkaman R, Moosavian MA, Safdari J, Torab-Mostaedi M. Synergistic extraction of gadolinium from nitrate media by mixtures of bis(2,4,4-trimethylpentyl) dithiophosphinic acid and di-(2-ethylhexyl) phosphoric acid. Annal Nucl Energy 2013;62:284–90. [24] Kumar A, Hartland S. Unified correlations for the prediction of drop size in liquid–liquid extraction columns. Ind Eng Chem Res 1996;35:2682–95.