Beruflich Dokumente
Kultur Dokumente
PII: S0263-8762(16)30299-4
DOI: http://dx.doi.org/doi:10.1016/j.cherd.2016.09.014
Reference: CHERD 2400
To appear in:
Please cite this article as: Reddy, S.Ramakrishna, Pandey, Niranjan K., Mallika,
C., Mudali, U.Kamachi, Equilibrium and Kinetics of Adsorption of Ruthenium on
Activated Charcoal from Nitric Acid Solutions.Chemical Engineering Research and
Design http://dx.doi.org/10.1016/j.cherd.2016.09.014
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Equilibrium and Kinetics of Adsorption of Ruthenium on
Activated Charcoal from Nitric Acid Solutions
S. Ramakrishna Reddy, Niranjan K. Pandey*, C. Mallika, U. Kamachi Mudali
1
kf = 1.6x10-6 cm/s
0.9
0.8
kf =1.6x10-5 cm/s
C/C0
0.7
0.6
kf = 1.6x10-4 cm/s
0.5
0.4
0 2000 4000 6000 8000 10000 12000
Time [s]
Effect of external mass transfer coefficient on the transport of Ru on to the adsorbent surface
Highlights
____________________________
*Corresponding Author: N.K. Pandey, Head, PE&MS, RR&DD, MPEDG, RpG, IGCAR, Kalpakkam, Tamilnadu,
India 603102
Email: nkpandey@igcar.gov.in; Ph.: 0091-44-27480126; Fax: 0091-44-27480126
1. Introduction
In the aqueous reprocessing of spent nuclear fuels by PUREX process, uranium and
plutonium are separated by solvent extraction using 30% tri-n-butyl phosphate (TBP) in
paraffin hydrocarbon diluent. The high level liquid waste (HLLW) solution generated after
the removal of U and Pu contains the radioactive fission products in nitric acid. The HLLW is
subjected to various waste management processes such as concentration, de-nitration and
vitrification in order to reduce the volume of the waste and to convert it into a solid vitrified
body for safe disposal. Ruthenium is one of the most troublesome fission products because of
103 106
its oxidation to highly volatile and chemically unstable radiotoxic RuO4 and RuO4
(McKibben, 1984). Ruthenium tetroxide decomposes to a non-volatile black deposit of
ruthenium dioxide (RuO2) on contact with the cooler parts of the surface of stainless steel
equipment during waste treatment and thus, the radiation dose of the plant increases.
Deposition of ruthenium was observed to plug the off-gas piping of the equipment owing to
appreciable volatilization of RuO4 during waste treatment. During vitrification of the waste,
platinum group fission product metals (Ru, Rh and Pd) form precipitates as alloy and/or
oxides in molten glass and Ru forms needle like electrically conducting RuO2 crystals
(Krause and Luckscheiter, 1991). It is reported that the vitrification of HLLW was
temporarily stopped in a nuclear reprocessing plant operated by Japan Nuclear Fuel Ltd
(JNPL) due to problems arising from the sedimentation of platinum group metals in the glass
melter, which had led to electric energy loss, local overheating, high viscous glass formation
and unexpected low production rate of vitrified glass (Krause and Luckscheiter, 1991). Thus,
separation of ruthenium species from the HLLW is desirable, prior to nuclear waste treatment
for safe disposal.
Various methods reported in the literature for the removal of ruthenium from HLLW
are volatilization (Holgye and Krivanek, 1978; Holgye, 1987), precipitation (Moore, 1960;
Beamish, 1966; Gandon et al., 1993), solvent extraction (Hyman and Leader, 1959; Kiba et
al., 1963; Fieberg and Edwards, 1978; Fitoussi et al., 1981; Lingen et al., 1989; Dhami et al.,
2012), ion-exchange chromatography (Lee and Chung, 2003; Liu et al., 1993), electro-
oxidation/reduction (Lietzke and Griess, 1953; Kobayashi et al., 1965; Motojima, 1990;
Yoneya et al., 1995; Mousset et al., 2004; Pravati Swain et al., 2014, 2015) and sorption
(Berak et al., 1975; Akatsu et al., 1979; Siddhanta and Das, 1985; Motoki et al., 1986;
Tikhomirova et al., 1992; Pankaj et al., 2007; Qadeer, 2007). Among these methods,
adsorption is preferred over other processes because of its cost effectiveness and the high-
quality treated effluents it produces. Sorption method is suitable for the separation and pre-
concentration of ultra-trace amounts of noble metals from solutions containing large excess
of non-noble metals such as Cu, Fe, Ni, Co, Al, Mg and Ca. Inorganic (activated carbon,
zeolites, silica gel, silica-alumina gel etc.) and organic complex forming sorbents (Amberlite
IRC-50, polystyrene-divinylbenzene (8%) resin with thiosemicarbazide, poly-4-vinyl
pyridine, cross-linked by divinyl benzene etc.) are often considered for the separation of
ruthenium.
Activated carbon is the widely used adsorbent for several applications (Ping and
Lynne, 1997; Ho and McKay, 2000; Sharma et al., 2004; Qadeer and Khalid, 2005) because
of the chemical properties of its surface. It is a non-hazardous, processed carbonaceous
product having an intricate porous structure and a large surface area that is contained
predominantly within the micro-pores. The adsorbent properties of the activated charcoal are
essentially attributed to its large surface area, high degree of surface reactivity, and
favourable pore size distribution.
An understanding of the adsorption kinetics of ruthenium within activated carbon
particles will ensure a better prediction of the transport behavior of ruthenium in fixed bed
columns. As there is lack of information in the literature pertaining to the kinetics of
adsorption of ruthenium within activated carbon particles, it is desirable to carry out a
systematic investigation on the subject. In the present study, fundamental aspects of
adsorption equilibrium and kinetics of ruthenium on activated carbon has been investigated to
establish the mechanistic aspect of the process. Sorption data generated were correlated with
the Langmuir and Freundlich isotherm models. Pore diffusion model was used to interpret the
experimental adsorption kinetic data. Based on the results, pore diffusion coefficient and
intra-particle effective diffusivity of ruthenium in activated charcoal were determined.
2. Experimental
The chemicals used in this study were ruthenium nitrosyl nitrate and activated
charcoal (obtained from pine tree) procured from M/s. Arora Matthey, Kolkata,
hydroxylamine chloride, sodium hydroxide and nitric acid (AR Grade; Merck make),
analytical reagent grade 1,10- phenanthroline monohydrate (chromogenic agent) supplied by
M/s. Loba Chemie and AR grade ethanol. The equipment employed were Thermo shaker
(IKA KS 4000), REMI centrifuge, JULABO circulating water bath (JULABO F-34), Double
walled jacketed vessel Vertex mixer (REMI RQ 20 PLUS), Metrohm 744 pH meter and UV-
Visible spectrophotometer (CHEMITO Spectra scan UV 2600). The characteristics of
activated charcoal used are reported in Table 1.
The influence of contact time on the transport of ruthenium onto the surface of
adsorbent was studied in a separate batch experiment by measuring the change in
concentration of ruthenium in solution with time. In this experiment, a known mass of
activated charcoal and 1000 mL ruthenium solution of known concentration were taken in a
double jacketed vessel, which was maintained at constant temperature (300K) using a
circulating water bath (JULABO-SW23). The content of the vessel was mixed at 200 rpm
using a vertex mixer (REMI RQ 20 PLUS). The liquid samples were withdrawn at regular
intervals and analysed for ruthenium concentration until it attained the equilibrium value.
Samples were analysed more frequently in the initial phase of the experiment than towards
the end, to achieve a suitable distribution of measured concentration.
2.4. Check for the Reproducibility of Experimental Data
All the experiments were repeated twice. The average value for each set of experiments
is presented in this study. One such data set is shown in Fig. 1. To verify the precision of
experimental results, paired t-test was carried out using two experimental data from the same
set. The calculated t value was 1.62. The t value from the 95% confidence table (Vogel 1989)
was 2.23. As tcal < ttab, it is confirmed with 95% confidence that there is no significant difference
between the two experiments of the same set. Hence, the average values have been selected for
the results of subsequent experiments.
-3
x 10
5
4.8
[Ru]soln. [mol.cm-3]
4.6
4.4
4.2
3.8
0 0.5 1 1.5 2
Time [s] 4
x 10
Fig. 1 Check for reproducibility of experimental data
qe
Ce C0 V
(1)
m
q e K F C1 / n (2)
The Freundlich model of isotherm was found to fit the experimental data better than
Langmuir model, as could be observed from the comparison between the two models, in Fig.
2. Logarithmic linearization of Eq. (2) facilitated the determination of model constants KF
and n and the numerical values of these constants for the present case were 0.005 and 8.15
respectively. The model fits the experimental data reasonably well giving a standard
deviation of about 7.19%.
-3
x 10
1.8
1.6
1.4
1.2
1
qe [mol.g-1]
0.8
0.6
Exp.
0.4 Freeundlich model
Langmuir model
0.2
0
0 0.5 1 1.5 2
Ce [mol.cm-3] -5
x 10
1.5
1
[Ru]ad. (mol.g-1)
0.5
[Ru]aq,t=0 = 185ppm
[Ru]aq,t=0 = 270ppm
[Ru]aq,t=0 = 465ppm
0
0 2000 4000 6000 8000 10000 12000
Time (s)
Fig. 3 Amount of Ru adsorbed on activated charcoal from aqueous solution of different initial
concentrations against time
Co
C s (R p , t)
liquid filled
film pores in the solid
3.4. Mathematical Model for the Transport of Solute from the Solution into the Adsorbent
Phase
Development of model equations is based on the following assumptions: (a)
diffusivity is constant at constant temperature and pressure, (b) activated charcoal particles
are considered as spherical and porous, (c) the adsorption sites are uniformly distributed
throughout the grains and (d) the external mass transfer coefficient and diffusivity are
constant for the range of solute concentrations under consideration.
C 3M p k f
(C Cs ) (3)
t VR p
where C is the bulk concentration of solute in the solution phase, and Cs is the concentration
of solute at the outer surface of the spherical particle. R represents the radius of the particle,
p is the density of adsorbent powder, Mp is the total mass of the particle, V is the volume of
solution used in the batch reactor and kf is the external mass-transfer coefficient. Here, mass
flow rate of solute through external film is proportional to the concentration difference
between bulk phase Ru concentration (C) and concentration of Ru at the particle surface, Cs
(Rp, t)
The mass balance of Ru within a spherical particle at a distance r from the centre of
particle is given as (from Ficks 2nd law of diffusion in spherical co-ordinate)
C r q 1 C r
p p r De 2 r 2 (4)
t t r r r
where De is the effective diffusivity of solute within the particle (could be treated as a
combination of pore diffusion and surface diffusion), p is the porosity, Cr is the solution
concentration of Ru within the pores of charcoal particle and qr is the solid phase
concentration of Ru. The first and second terms on the left hand side of Eq. (4) give the
accumulation of Ru in the void space of the adsorbent particles and solid phase respectively.
The Freundlich isotherm can be used for the correlation of Cr and qr. Then Eq. (4)
becomes
C r De 1 2 C r
1n / n 2 r (5)
t p ( p K F / n )C r r r
t = 0: C = C0, Cr = 0 (6)
C
r 0: r 0 (7)
r r 0
C
r R : k f ( C CS ) D e r (8)
r r R
X 3M p
Bi (X X S ) (9)
V P
X r 1 2 Xr 2Xr
(10)
p ( p K f / n )C 1n / n x x x
2
1 2 Xr
ar x
x 2 x x
and the initial and boundary conditions in dimensionless form can be written as,
0 : X 1, X r 0 (11)
X
x 0 : r 0 (12)
x x 0
X
x 1 : r Bi(X X S ) (13)
x x 1
C Cr r tDe kf R
X , Xr , x , , and Bi
C0 C0 R R2 De
For solving the set of partial differential Eqs. (9) and (10), they are changed to
coupled ordinary differential equations by finite difference methods as follows:
dX i X 2X i X i 1 X i 1 X i 1
a (i) i 1 (14)
d (x ) 2 x i x
Applying the boundary conditions (12) and (13) then, Eq. (10) becomes
The total N+1 coupled ordinary differential equations for X, X1, X2, ---- XN are solved
numerically by the fourth order Runge-Kutta method using ODE-45 solver of MATLAB
software, in order to obtain concentration profiles. The external mass transfer (kf) and the
effective diffusivity (De) are the two unknown parameters, which are estimated by
superimposing the experimental data on to the diagrams of numerical solution of model
equations using a nonlinear regression technique. The best values were obtained by
minimizing the following objective function:
1
Exp. (C 0 ~ 185ppm)
Calc. (C 0 ~ 185ppm)
0.7
0.6
0.5
0.4
0 2000 4000 6000 8000 10000 12000
Time [s]
Fig. 5 Comparison between experimental and model predicted concentrations against time
Sensitivity analysis was performed for the changes in kf and De. Figure 6 represents
the effect of kf on model predicted results. If kf is 10-100 times lower than the estimated
value of 2.6 x 10-4 cm/s, then the removal rate of ruthenium is slow. Nevertheless, when kf
value exceeds the estimated value, the removal rate remained nearly unchanged, because the
external mass transfer coefficient is a function of intensity of mixing. In a stirred vessel, the
thickness of the stagnant interfacial film surrounding the particle is thicker at low impeller
speed and diffusion limits the rate. With increase in impeller speed, stationary diffusion film
becomes thinner and the external diffusional resistance becomes smaller. At some value of
stirring speed, the external mass transfer coefficient has no influence on the entire removal
process.
Figure 7 shows the effect of diffusivity on model predicted results. If the effective
diffusivity (De) value is 10 to 100 times higher than the specified value of 2.3 x 10-7 cm2/s,
then the removal rate becomes faster. However, when De is lower than the estimated value,
the removal rate is almost similar to that of estimated value. Thus, from simulation studies it
is predicted that the effect of De on the removal rate of Ru is less sensitive in comparison to
the effect of external mass transfer coefficient.
1
kf = 1.6x10-6 cm/s
0.9
0.8
kf =1.6x10-5 cm/s
C/C0
0.7
0.6
kf = 1.6x10-4 cm/s
0.5
0.4
0 2000 4000 6000 8000 10000 12000
Time [s]
Fig. 6 Effect of external mass transfer coefficient on the transport of Ru from solution to the
adsorbent surface
1
0.9
0.8
D =2.3x10-9 cm2/s
C/C0
e
0.7
D = 2.3x10-8 cm2/s
0.6 e
-7 2
D =2.3x10 cm /s
e
0.5
0.4
0 2000 4000 6000 8000 10000 12000
Time [s]
Fig. 7 Effect of diffusivity on the transport of Ru from solution to the adsorbent surface
14
12
qe [mol.g-1]
10
2
0 2 4 6 8 10 12
pH of the solution
Fig. 8 Effect of pH on the transport of Ru from solution to the adsorbent surface (initial
concentration of Ru in the solution: 1g/L
4. Conclusions
a. The adsorption isotherm could be correlated using Freundlich model of adsorption. The
Freundlich isotherm fitted well with experimental data.
b. The mechanism of adsorption involved an initial rapid rate for the removal of ruthenium
due to surface diffusion followed by intrapaticle diffusion. At later stages, rate of
adsorption of ruthenium decreased with time, possibly due to the decreased concentration
gradient.
c. The diffusion model was numerically assessed and there was conformity between
theoretical prediction and batch experimental data. Theoretical values of De and kf were
estimated from diffusion model.
d. Kinetic modelling revealed that pore diffusion model to be appropriate for the description
of transport of ruthenium within activated charcoal particles.
Nomenclature
n Freundlich constant
t Time (s)