STAGEWISE CALCULATION FOR T H E SOLVENT EXTRACTION SYSTEM MONAZITE RARE EARTH N ITRATES-N ITRIC ACI D-TRI BUTYL
0S PHIA T E-W ‘ A TER BROOKS M . S H A R P ’ A N D
M O R T O N S M U T 2
Institute f o r Atomic Research and Department of Chemical Engineering, Iowa State University, Ames, Iowa
Using single-stage equilibrium data plus the assumption of mutual immiscibility of water and tributyl phosphate (TBP), a calculation method was developed to give the stagewise conditions in a cascade of equilibrium stages with four rare earth nitrates and nitric acid present as solutes in the flow streams. The method was checked by a series of simulated column experiments.
The agreement between predicted and experi-
mental stagewise conditions was considered reasonable, and the method is considered useful for engineering work. The calculation and correlation methods developed should b e applicable to systems with a larger number of rare earth solutes and possibly to some other multisolute systems containing a series of chemically similar solute!;.
THE project discussed is a n investigation of the application of solvent extraction to the separation of the monazite rare earths: specifically, the postulation and verification of a calculation method to give the stagewise conditions in a multistage cascade operating with the five-solute system La(K03)a-Pr(S03),-Nd ( N o s ) 3-Sml:SOs) 3-HNO3-TBP-H20. T h e basic problem, ,as in many projects of this type, was to develop a method of calculating the conditions in one equilibrium phase with the contacting phase completely specified. Because of the large number of solutes present, it is practically impossible to investiga1:e systematically the equilibrium of the system of interest over the range of composition and concentration encountered in a solvent extraction cascade. Therefore the basic approach \vas to use experimentally determined equilibrium data for contributing systems to predict the equilibrium of the more complex system of interest.
to calculate the total molality, ” 0 3 molality, and the rare earth molality of the unspecified phase. To use the separation factors between the rare earths obtained from data for the contributing three-solute systems in one of the equations
(4)
(5)
Method of Calculation
T h e data used in the calculation method are single-stage “shake-up” distributions for the systems I . L a ( x 0 3 ) 3-H?;03-TBP-H20 11. P r ( N 0 3 ) S-HSO3-TBP-HZO
VI. VII.
Nd ( S O3) 3-Pr (S O3) 3-HS03-TBP-H 2 0 S m ( SO3) 3-Pr (NO3) 3-HNO 3-TBP-H20
T h e data for systems I , I I ? V: VI, and VI1 were collected by the authors by the method described in detail by Sharp and Smutz (9). T h e bulk of the data for systems I11 and I V was supplemented to a small taken from the \vork of Schoenherr (a), degree by work of the authors. A complete compilation of the data used is given by Sharp and Smutz (70). Equilibrium Phase Calculation. T h e general procedure for calculating concentrations in one phase when all concentrations in the contacting phase are specified was T o obtain values of K I and K , by interpolation using data obtained from the contributing two-solute systems and then use the defining equations 1 Present address, McDonnell Aircraft Co., St. Louis, Ma.
to calculate the rare earth molalities of the unspecified phase. Equations 4 and 5 follow directly from the definition of separation factor and are applicable, of course, only to the rare earths present.
I n more detail, the d a t a from systems I, 11, 111, and I V were processed to give a series of plots of K l and K , us. the composition of a n equilibrium phase with the total molality of the same phase as parameter. Typical plots, shown in Figures 1 through 4, were prepared a t intervals over the ranges 3 5 .IdC 5 16 and 2 5 *VI 5 4 . 5 . A complete tabular compilation of the values from these plots is given by Sharp and Smutz (70). Abstracts from these tables for use in the following sample calculation are shown in Tables I and 11. T h e separation factors from systems \’, \-I, and VI1 were plotted against the total molality of the organic phase as shown in Figure 5. I n each case the rare earth separation factors are, to a good approximation, a function of the total nitrate molality of the equilibrium organic phase and relatively independent of the phase composition. Similar assumptions have been applied to organic systems by K a r r and Scheibel (2) and Scheibel (7). T h e curves of Figure 5 xvere fitted to straight lines by the least squares method to give VOL. 4
NO. 1
JANUARY
1965
49
Array of Values of K t for Systems RE(N03)3-HN03-TBP-H,0
Table 1.
8, = YRE 15 20 25 30
RE 0 0 0 0
0 0 0 0
Table II.
= La 430 413 399 389
RE 0 0 0 0
= Pr 495 484 475 469
RE 0 0 0 0
RE = La 0.756 0.831 0.910 0.991
YRE
0.15 0.20 0 25 0.30
.Vd 508 499 492 487
RE 0 0 0 0
S6 543 541 541 544
=
RE 0 0 0 0
La 354 345 337 329
RE 0 0 0 0
=
Array of Values of K, for Systems RE(NOd)3-HN03-TBP-H20
.et =
R E = .Vd 0 589 0,612 0.641 0.680
RE = Sm 0.571 0.584 0.600 0.619
.e,
I
[
0.4
0.6
OB
(
,
RE 0 0 0 0
442 435 424 415
I
Sm 458 460 461 462 =
and Y K E
mt
3.5 R E = ‘Vd 0.520 0.557 0.597 0.637
=
RE = Pr 0.563 0 607 0,653 0,701
I
l
RE = Sm 0,467 0.482 0,500 0,523
l
Mi ~ 5 . 0
I.9
02
0 0 0 0
as a Function of
RE = La 0,701 0.797 0.895 0,995
2.0
0.0
= Pr 426 420 413 406
3.0
R E = Pr 0.628 0.661 0.700 0,748
35 RE = .l-d
= =
atand YRE
as a Function of
30
0.5
1.0
0.0 0.1 0.2 0.3 0.4 05 06 0.7 0.8 0.9 1.0
YRE
X~~
Figure 1 . Nitric acid distribution coefficient for RE(NO&-HNO~-TBP-HZO systems as a function of composition of organic phase at M , = 3.0
-
Figure 3. Nitric acid distribution coeffisystems cient for RE( NO&-HNOZ-TBP-HzO as a function of composition of aqueous phase at M , = 5.0
0.7
0.6
0.71
0.5
I
I
I
I
I
I
M+ = 5.0 RE?;Lo 0 RE=Nd R E = Pr A R E = Sm 0
I
I
I
I I
n
0.3 0.0
0.2
a4
0.6
0.8
LO
Y~~
Figure 2. Total distribution coefficient for RE(N03)3-HN03-TBP-H20 systems as a f u n z tion of composition of organic phase at Mt = 50
3.0
I & E C PROCESS D E S I G N A N D DEVELOPMENT
XIIE
Figure 4. Total distribution coefficient for RE(NOa)3-HN03-TBP-H20 systems as a function of composition of aqueous phase at M , =
5.0
n
q
i1
$30
/
3
3
Figure 6. I
I
Cascade A
2
r
3
Figure 7.
n
Cascade B
PLa-Pr 0.0 0.0
I.o
2.0
3.0
4.0
5.0-
6D
7.0
TOTAL MOLALITY OF ORGANIC PHASE, M t
Figure
5.
Separation
P r ( N 0 3 ) 3 for
factor
between
RE(NO& and
RE(N03)3--Pr(N03)3-HN03-TBP-H20 systems
as
a function of concentrution of organic phase
K t = 0.5201 K , = 0.6116 T h e values of M I , M u , and M R E were then calculated from
0.1106
JT,
(6)
= 1.0448 f 0.09874
JIt
(7)
Equations 1, 2, and 3 to be
(8)
M , = Ji,/Ki M , = -U,/K, M R E M , - A4,
a ~ ~ -= p 0.8187 ~ /3sd-pr
-
X (0.0595) = 0.0240. T h e sum of the four contributions wai then the value of K , used. Exactly the same procedure: using Table II! was then used to get a value of K, for the case in question. T h e values so obtained were
/3sm-pr = -0.3795
+ 0.9214 aVt
These correlations were assumed to hold with J i t greater than 1.75. At this point the gross assumption was made that Equations 6, 7 , and 8 Lvould hold also in the five-solute system La(SO3)3(SO3) 3-HSO 3-TBP-H 2 0 . The Pr (SO3) 3-Nd (SO3 ) limitations of such a s\r.eeping assumption may only be ascertained empiricalll-, and the usefulness of the approximation must be judged by the icomparison \vith experimental results in the "Verification of Model" section. Sample Calculation. ,4numerical example illustrates the equilibrium phase calculation. I t was assumed that a n equilibrium organic phase has the following concentrations ;
J i t =3.0261
Jipr
Ji, -
-IIsd
= 2.3590
= 0.0460
0,0706
.iism
- 1 f ~=~ 0.0397
= 0.5108 *URE = 0.6671
From these data were calculated
Using
*Vt =
=
3.0261/@.5201
=
5,8183
= 2.3590/0.6116 = 3.8571
= 1.3612 3.0261 the separation factors were calculated
as
/~L*-P~= Psd-pr
0.4840
= 1.3436
Psm-pr = 2.4087 By definition, PPr-Pr is equal to 1.O. The rare earth molalities of the aqueous phase were then calculated by Equation 5 as follo\vs. T ( - U ~ / P ~ - P ~=) -IfLn!~La--Pr
t = 1
+
JiPr/~Pr--Pr
+
-Ifsd//3sd-pr
-
*1fsm,' T (.vt/Pt-pr)
= 0.0397/0.4840
+ 0.0460
1 = 1
0.0706/1.3436
1,0000
+
B5rn-p
+
+ 0.5108,'2.4087
T (.Vt,'Pi-pr)
5
0.3926
I = 1
Jia/-vt
Y,
= - -. = 2 3590:3.0261 = 0.7796 = -IIRE/.III = 0 6671 '3.0261 = 0,2204 - J L ~ = J f L h t J f R E = 0.0397j0.6671 = 0.0595
YR,
0.0460/0.6671 = 0.0690 @.07@6;0.6671 = 0.1058 ~= R 0.5108~0.6671 E = 0.7657
ipr
= - v p r j . u ~ l= ~
?sa
= av~ci
jsm
=
RE =
- ~ S ~ / -
T o obtain the value of I;, it was first assumed that all the rare earth nitrates present were La(NO3)s-that is: YR= = I'I.& = 0.2204. Then by straight-line interpolation between the values in 'Table I a value of K ,was obtained that ivould hold for the case of J i , = 3.0261 and Y,, = YLn = 0.2204. This \\'as I;: = 0.4039. Exactly the same procedure \vas then follolved to get values for I;(.assuming, in succession, that the rare earths present were Pr(SOa)3. Nd(SO3)a. and Sm(NO3)a. Then to get the K t used in the calculation. each I;,derived from the assumption of onl!; one rare earth solute was multiplied b!. the fraction of the rare earths present as that particular solute. I n other Lvords. the contribution from the La(N03)3 system to the final va1c.e of I;, \vas assumed to be equal to the value of I;, for the LaI:SOa) 8-HNOJBP-HyO system multi.-2rithmetically, the contribution \vas (0.4039) plied by >I.n.
and in exactly the same manner M p , = 0.2298 Msa = 0.2625 Ms, = 1.0594 If the reverse calculation had been desired, the same pro; cedure would have been followed: first using tables derived from plots of the type portrayed in Figures 3 and 4 to get K , and A',, calculating then -ITa.and RE\ and then, after calculating the separation factors, using Equation 4 to calculate the rare earth molalities. Application to Multistage Calculations. It was desired to apply the moclel to the calculation of the stagetvise conditions in an ideal multistage cascade. Referring to Figures 6 and 7 it is seen that if any point on the operating line is kno\vnthat is: if two streams passing each other are completely defined-the conditions for any number of stages may be calculated "backwards" through the column if the operating line equation is known. T h e direction such a calculation
-u13
VOL. 4
NO. 1
JANUARY 1 9 6 5
51
kvould follo\v is shown b!- the dashed arro\vs in Figures 6 and 7. The calculation \vould obviously be composed of alternating applications of the equilibrium model and the operating line. 1 he application of this method to a center-fed cascade in Jvhich only the feed stream is specified requires trial-anderror procedures. In the present work, mutual immiscibility of \vater and TBP \vas assumed. TZ'riting a general operating line for the cascade pictured in Figure 6 gives
CASCADE
F IC
CASCADE
II
+
+
[(.Udo5'01 [(-bfOnRnI = [ ( J d n - is,-11 [(Mt)1R11 ( 9 ) By the assumption of mutual immiscibility Sa = S-, and R , = R1 and so the subscripts may be deleted, and after rearrangement (Mi),= a(-Udn-l - c y ( J 7 d o I),!%'( (10) T h e assumption of mutual immiscibility of water and TBP \vas felt to be justified. T h e solubility of TBP in water is very lo\v>on the order of tenths of a gram per liter, and the solubility of Lvarer in TBP does not exceed approximately 67, by weight. Because the solvent is normally equilibrated \vith water before use. the assumption of constant water content does not introducc a large error. Using the proposed equilibrium model and operating lines of the form of Equation 10, several calculations of the stage\vise condirions in various cascades were carried out. The procedure \vas to choose the end conditions arbitrarily, either R 1 and So for the cascade in Figure 6, or Ro and SI for the cascade in Figure 7 . and then to proceed as explained previously. At this point an unexpected problem arose. In several of the calculations. the total molality wzas "pinched" at a high value to impress high separation factors on the system since, with the equilibrium model proposed. the separation factors increase n i t h increased total concentration. In each case in rvhich this \vas done the individual concentrations also quickly pinched. Tt-hen the end conditions and flow rate ratio were chosen so that the total concentration changed from stage to stage. no such pinch of the individual solutes \vas calculated. T h e study of this pinch? based on the limit of a sequence that can be r\-ritten giving the molality of a general component in successive streams of a cascade: is an interesting problem the authors have not been able to solve completely. Sharp and Smutz (70) discuss this approach and give an analytical solution for the special case of 7' = 2. . i s the stage\vise calculation is tedious and time-consuming, digital computer programs for the IBhf 7074 were \vritten for four cascades of interest. A complete discussion of these programs is given by Sharp and Smutz (70).
CASCADE
III
+
Verification of Model
T h e simulated column technique was used to obtain experimental stagexvise concentrations for comparison Lvith the predicted values. Briefly, a simulated column experiment is a batch\vise approximation to a continuous countercurrent cascade. The error in the approximation may be made arbitrarily small by increasing the number of cycles the experiment is run. T h e mechanics and mathematical demonstrations involved have been discussed extensively by Scheibel (3-6),
Four of the prediction calculations discussed in the "Method of Calculation" section were chosen for simulation. Experiments lvere chosen Ivhich represented a \vide variation of flow rate ratio a . number of stages. and solute concentrations. The experiments Lvere carried out in laboratory glassware at room temperature. using loo-hp.laboratory stirrers for the mixing steps. .i minimum of 2 minutes' mixing and 3 minutes' settling time \vas maintained for each contact. 52
l & E C PROCESS D E S I G N A N D D E V E L O P M E N T
Figure 8. simulation
Table 111.
Expt.
I
I1 I11
IV
Continuous countercurrent cascades chosen for
Experimental Conditions for Simulated Column Experiments Volume hltrtc Rare Flou. per InjecAcid Earth Rate tion, Molal.+'itrate of Ratio. Stream MI. ity Molality Cycles cr Rii 74.6 1.86 6.89 11 0.5 SO 30.7 0.00 0.00 Ro 30.0 0.00 0.00 Si1 245.2 0.00 0.00 10 8.0 F 215.6 6.66 11.69 Re 40.5 7.14 4.49 10 2.0 so 61.3 0.00 0.00 Ro 54.2 4.97 7.15 10 1.0 s 6 40.9 0.00 0.00
Along with the phases from the final cycle, the leaving aqueous phase from each cycle was retained and analyzed as a check on the approach to steady state. In all cases the number of cycles was felt to be sufficient. All organic and aqueous phases from the final cycle were analyzed for total oxides, nitric acid content, and rare earth composition.
Materials. Commercial grade TBP, obtained from Commercial Solvents Corp., was pre-equilibrated with distilled water before use as the solvent. The TBP was stored in contact with a water phase to ensure equilibrium. Reagent grade nitric acid was used in all laboratory work. The oxides of La, Pr, Nd, and Sm, 99.9% pure, were obtained from the ion exchange group of the Ames Laboratory. Atomic Energy Commission. Stock nitrate solutions of each rare earth were prepared by dissolving the oxides in boiling, concentrated nitric acid. After complete reaction, the excess nitric acid was boiled off and distilled water added. All feed solutions for the simulated column experiments were prepared by mixing these stock solutions, reagent grade concentrated nitric acid, and distilled water. Analytical Methods. Weighed samples of each equilibrium phase were analyzed. T h e solutes in the organic samples were transferred to aqueous solution prior to analysis by bringing the sample a minimum of three times in contact Lvith distilled water. The aqueous phases were combined and analyzed in the same manner as the aqueous phase samples. For the determination of total rare earths the aqueous sample was diluted to approximately 200 ml. with distilled water and brought to a boil on a hot plate, and the rare earths were completely precipitated with a saturated solution of oxalic acid. The oxalates were filtered and converted to the oxides a t approximately 870' C. in a muffle furnace. The weight of the oxides was determined on a n analytical balance,
Table IV.
Expierimental and Predicted Compositions and Concentrations for Simulated Column Experiments
Weieht Per Cent in Total Oxides .Z’d203 Exptl. Pred. ~
SmOs Exptl. Pred.
Gram/Gram Solution Total Oxides H,VO a Exptl. Pred. Exptl. Pred.
SIUULATED COLEMN.EXPERIMENT I 32.3 30.0 28.0 28 8 28.4 29.9 29 29.1 29.2 30.1 29.0 12.1 8 .O 9.4 9,3 9.4 10.0 10.1 7 5 10.2 8.2
32.3 29.5 29.2 29.3 29.3 29.3 29 3 29.3 29.3 29 3 29 3 13 1 10.0 9 6 9,5 9.6 9.6 9.6 9.6 9.6 9.6
25 26 25 24 26 25 24 24 24 24 24 22 19 18 19 19 18 19 20 19 19
2 8 4 7 4 2 8 4 8 9 9 4 9 8 1 0 7 0 5 0 8
25 24 24 24 24 24 24 24 24 24 24 21 19 18 18 18 18 18 18 18 18
3 8 5 4 4 4 4 4 4 4 4 6 1 6 5 5 5 5
5 5 5
25 6 25 0 25 9 25 7 25 4 25 0 25 4 28 8 26 5 26 0 26 5 28 4 28 8 28 1 27.5 27.1 27 0 27 2 28 3 26 0 27.5
25 0 25 6 25 4 25 3 25 2 25 2 25 2 25 2 25 2 25 2 25 2 28.8 27 4 26 7 26.6 26.6 26 6 26 6 26 6 26 6 26.6
17 0 18 2 20 7 20 8 19 8 19 9 20 2 17 8 19 5 18 9 19 6 37 1 43 3 43.8 44.1 44 5 44 4 43 6 43 7 44.8 44.5
17 3 20 1 21 0 21 1 21 0 21 0 21 0 21 0 21 0 21 0 21 0 36 6 43 5 45.2 45.4 45.4 45 3 45 3 45 3 45.3 45.3
0.191 0.208 0.206 0.204 0 204 0 203 0 202 0.202 0.202 0.205 0 204 0 088 0 085 0 082 0.081 0.080 0 081 0 080 0 081 0.081 0,081
0.195 0.211 0.208 0 207 0 207 0 207 0 207 0.207 0.207 0.207 0 207 0 088 0 083 0 080 0.078 0.078 0 078 0 078 0 078 0.078 0.078
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
039 055 060 062 062 062 062 062 062 062 062 054 071 076 078 078 078 078 078 077 077
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
037 053 060 062 062 063 063 063 063 063 063 054 074 080 082 083 083 083 083 083 083
SIUCLATED COLUMN.EXPERIMENT I1 7.3 14.6 26.4 37.4 50.6 62.8 72.9 82.3 81.3 86.4 83.0 0.0 3.4 6.5 10.5 18.6 25.5 36.3 45,4 51.2 50.7
11.2 19.9 31 . O 43.4 55.4 65.8 74.0 80.0 84.4 87.8 81.2 3.8 5.6 9.0 14.1 20.7 28.3 36.1 43.4 49.8 55.5
10 12 15 14 12 9 7 3 6 5 2 5 7 8 8 9 9 7 5 5 5
4 9 0 7 0 8 5 4 1 8 9 7 1 2 9 2 3 4 7 4 0
9 11 12 11 9 7 6 4 3 2 3 5 6 7 8 8 7 6 6 5 4
5 7 3 5 8 9 1 6 6 8 0 9 8 8 4 3 8 9 0 1 4
11 4 11.9 10 5 8.6 7.0 4 2 2 6 1 2 2.5 1.9 0 8
10 8 11 9 12 5 12.4 11.2 10.5 9.9 8.4 7.5 5.7
12.1 13.2 12 6 10.9 8.7 6 7 5 1 3 9 3 1 2.4 3 0 9 7 10 3 10 9 10.8 10.2 9.3 8.2 7.1 6.2 5.5
70 9 60 6 48 2 39 3 30 4 232 1’0 13.2 10.1 5.8 13 2 835 77 6 72 8 68.3 61.0 54.6 46.5 40.5 35.9 38.6
67.3 55 2 44 1 34 2 26 0 196 149 11.4 8.9 7.0 12 8 806 77 3 72 3 66.7 60.7 54.6 48.8 43.5 38.8 34.6
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
052 072 087 103 122 140 163 184 204 220 235 023 029 032 034 037 040 044 047 052 057
058 076 092 108 126 145 166 186 207 228 238 024 031 035 037 040 044 047 052 056 062
0.139 0.168 0 172 0 165 0 152 0 136 0 117 0 102 0 088 0 076 0 153 0 108 0 124 0 127 0 129 0 128 0 126 0 124 0.121 0.117 0 115
0.136 0.167 0 171 0 164 0 151 0 136 0 119 0 104 0 090 0 077 0 156 0 107 0 122 0 126 0 127 0 126 0 125 0 122 0.119 0.116 0.114
0,060 0.061 0,046 0 035 0 030 0 127 0 183 0 174 0 151 0 136 0 128
0.044 0.098 0 137 0 156 0 163 0.025 0 081 0 148 0 199 0 223 0 232
0.031 0 083 0 128 0 154 0 166 0.016 0 055 0 122 0 182 0 216 0 232
0 0 0 0 0
0 142 0 142 0 138 0 127 0 090 0.149 0.149 0.147 0 140 0.117 0.071
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
SIMULATED COLUMN.EXPERIMENT I11 65.8 43.7 29.1 26.9 17.0 79.6 77.8 64 5 61.8 62.4 63,9
66.8 47.4 28.5 19.6 17.8 79.3 73.7 64.3 58.7 58.8 60.4
10 12 10 7 6 8 9 10 8 7 7
5 8 7 1 7 2 3 2 7 8 1
11 12 11 8 6 8 9 10 9 8 7
4 8 1 2 7 3 7 4 4 3 8
11.5 16.7 17.0 13.5 10.2 7.4 5 2 11 3 9 9 8 6 7 5
12.1 16.7 17.4 14.1 11.3 7.2 9 4 11 6 113 9 5 8 4
12.3 26.8 43.3 52.5 66 0 4 9 7 7 13 9 19 5 21 2 21.4
9.7 23.1 43.1 58.1 64 3 5 3 7 3 13 6 20 6 23 4 23.4
0.049 0.050 0.039 0 032 0 030 0 118 0 163 0 158 0 139 0 130 0 126
SIMULATED COLUMN.EXPERIMENT IV 7.0 6.9 7.1 7 7 14.3 33 4 33.2 33.5 34.8 36.4
9.1 9.1 9.1 10.1 16.0 34.0 34.0 33.9 33.6 34.3 38.9
17 17 18 19 24
9 9 2 7 5
23 24 25 25 24
9 2 9 8 6
19 19 19 20 24 25 25 25 25 26 27
0 0 2 7 7 8 8 8 9 6 1
21.8 22.3 22.7 25.1 27.6 20.7 21 7 21 9 22 0 21 8
21.8 21.9 22.2 23.6 25.2 20.4 20.4 20 5 20 7 21 1 20 1
53.4 52.9 51.9 47.5 33.6 22 21 18 17 17
0 0 8 4 2
50.0 50.1 49.5 45.7 34.1 19 8 19 9 19 9 19 7 18 0 13 9
0 0 0 0 0
052 053 054 059 067
0 0 0 0 0
193 194 197 204 185
VOL. 4
0 0 0 0 0 0 0 0 0 0 0
050 050 052 058 067 191 191 192 196 204 188
NO. 1
136 135 132 122 088
0.147 0 145 0 138 0.118 0.072
JANUARY
1965
53
The composition of the final rare earth oxide sample was determined by a flame photometric method developed by members of the analytical chemistry group headed by V. A. Fassel, Ames Laboratory, Atomic Energy Commission (7). T o determine the nitric acid content, the samples were diluted to approximately 200 ml. and the rare earths present precipitated by addition of saturated potassium ferrocyanide. The nitric acid content \vas then determined by titration with standardized sodium hydroxide. An automatic titrater was used and the end point was taken at a p H of 8.7. The nitric acid content of aqueous solutions of nitric acid and potassium ferrocyanide diminished with time. This was thought to be due to oxidation by the acid. To obtain satisfactory analyses for the nitric acid it was necessary to titrate the aqueous solutions of nitric acid. potassium ferrocyanide, and precipitated rare earth ferrocyanides within about 15 minutes after addition of the potassium ferrocyanide solution. Simulated Column Experiments. Four cascades for which prediction calculations had been carried out were chosen for simulation. Figure 8 gives the continuous cascade being simulated, and Table I11 contains a compilation of pertinent experimental conditions. Table I\’ compares experimental and predicted concentrations and compositions. Conclusions
The calculation method developed for the multistage multisolute system gave good agreement with experimental data from four simulated column experiments carried out under widely differing conditions. The agreement in all cases demonstrates that the calculation method can be used to select processing conditions for a desired separation with this system. T h e assumption that the separation factor between two rare earth nitrate solutes was a function only ‘of the total was greater than 1.75, molality of the organic phase, when gave surprisingly reliable results. The reason for the validity of this assumption is not knolvn, but it appears to be a good approximation. I t is considered reasonable to postulate that the correlation and calculation methods developed would be applicable to systems ivith a greater number of rare earth solutes, and possibly to some other multisolute systems containing a series of chemically similar solutes.
Avt
Nomenclature
All molalities are expressed in terms of the nitrate group associated with the solute in question-for example, a solution containing 1 kg. of water and one molecular weight (324.9 grams) of L a ( S 0 3 ) 3would be 3 molal L a ( N 0 J 3 .
54
l & E C PROCESS D E S I G N A N D D E V E L O P M E N T
M
Av = K
=
X
/3
= =
Y
=
y
=
S
=
R
=
T
= =
CY
kg. of solvent molality of an organic phase, moles of solute per kg. of solvent distribution coefficient, ratio of molality of a solute in an organic phase to molality of solute in equilibrium aqueous phase separation factor. ratio of two distribution coefficients ratio of molality of a solute in aqueous solution to total molality of solution ratio of molality of a solute in organic solution to total molality of solution ratio of molality of a rare earth solute in organic solution to total rare earth molality of solution flow rate of solvent in a n organic stream. kg. TBP:unit time flow rate of solvent in a n aqueous stream, kg. H 2 0 / / u n i t time total number of rare earth solutes flow rate ratio, SIR
= molality of a n aqueous phase, moles of solute per
SUBSCRIPTS
i
=
t RE La Pr Nd Sm
= = = =
arbitrary solute total solutes present, expressed as equivalents of nitrate total rare earth nitrate solutes present lanthanum nitrate praseodymium nitrate = neodymium nitrate = samarium nitrate a = nitric acid 0, 1, 2, . , , , n - 1: n = subscripted quantity refers to stream of multistage cascade Literature Cited
(1) D’Silva, A,, Fassel, V. A , , Kniseley, R. N., Anal. Chem. 36, 532 (1964). (2)’-Ka;r, A. E . , Scheibel, E. G., 2nd. Eng. Chem. 46, 1583 (1954). 3) Scheibel, E. G., Ibzd.,43, 242 (1951). id.,44, 2942 (1952). 14) 16’ (5) Ibid... 46,. 16 11954). (6) Ibid.,p. 43. (7) Scheibel, E. G., Petrol. Rejner 38, No. 9, 227 (1959). (8) Schoenherr, R. U., Ph.D. thesis in chemical engineering, Iowa State University, Ames, Iowa, 1959. (9) Sharp, B. M., Smutz, M., U. S. Atomic Energy Commission, Rept. IS-335 (1960). (10) Ibid., IS-780 (1963). ~
I
RECEIVED for review January 3: 1964 ACCEPTED June 25, 1964 Contribution 1422. M‘ork performed in the Ames Laboratory of the U. S. Atomic Energy Commission.
