New recovery process for rhodium using metal vapor

Article abstract of DOI:10.1016/S0925-8388(03)00666-2

Rhodium (Rh) is an essential element as a catalyst in automotive catalytic converters, and the recovery of Rh from scrap is an important issue. However, the chemical inertness of Rh makes it difficult to recover it from scrap. This study focused on a new process for recovering platinum group metals (PGM) from scrap with the purpose of improving Rh dissolution in acid. Reactive metal vapors such as magnesium (Mg) and calcium (Ca) were reacted with powdered Rh in a closed stainless steel reaction vessel at a constant temperature ranging from 873 to 1173 K. Mg and Ca deposited on Rh at temperatures above 973 and 1073 K, respectively. Dissolution experiments of the obtained compounds were conducted at 323-333 K for 1 h using an aqueous HCl solution or aqua regia. After reactive metal treatment 99% of Rh was dissolved by aqua regia leaching. Under the same leaching condition only 8% of the untreated pure Rh powder was dissolved. Rh was recovered from the solution obtained by leaching Mg reacted Rh compound using precipitation or cementation methods, demonstrating the feasibility of such recovery. These results show the possibility of developing a new Rh recovery process utilizing reactive metal vapor treatment.

Full text of DOI:10.1016/S0925-8388(03)00666-2

Journal of Alloys and Compounds 365 (2004) 211–220
New recovery process for rhodium using metal vapor
Yoshihiro Kayanuma∗, Toru H. Okabe, Yoshitaka Mitsuda, Masafumi Maeda
Institute of Industrial Science, The University of Tokyo, 4-6-1, Komaba, Meguro-ku, Tokyo 153-8505, Japan
Received 27 June 2003; accepted 1 July 2003
Abstract
Rhodium (Rh) is an essential element as a catalyst in automotive catalytic converters, and the recovery of Rh from scrap is an important
issue. However, the chemical inertness of Rh makes it difficult to recover it from scrap. This study focused on a new process for recovering
platinum group metals (PGM) from scrap with the purpose of improving Rh dissolution in acid. Reactive metal vapors such as magnesium
(Mg) and calcium (Ca) were reacted with powdered Rh in a closed stainless steel reaction vessel at a constant temperature ranging from 873 to
1173 K. Mg and Ca deposited on Rh at temperatures above 973 and 1073 K, respectively. Dissolution experiments of the obtained compounds
were conducted at 323–333 K for 1 h using an aqueous HCl solution or aqua regia. After reactive metal treatment 99% of Rh was dissolved
by aqua regia leaching. Under the same leaching condition only 8% of the untreated pure Rh powder was dissolved. Rh was recovered from
the solution obtained by leaching Mg reacted Rh compound using precipitation or cementation methods, demonstrating the feasibility of such
recovery. These results show the possibility of developing a new Rh recovery process utilizing reactive metal vapor treatment.
© 2003 Elsevier B.V. All rights reserved.
Keywords: Precious metals; Scrap recovery; Vapor deposition
1. Introduction
involves the extraction and concentration of PGMs from
scrap, often including the process of pyrometallurgical con-
centration or hydrometallurgical leaching. In the past, a col-
lector metal such as copper, tin or lead was used to extract
and concentrate the PGMs [2–7]. The second stage involves
the separation and purification of PGMs, including the pro-
cesses of precipitation, solvent extraction and ion exchange.
Among these processes, the dissolution of PGMs is an im-
portant procedure, because PGMs are usually separated and
purified in solution during the second stage. However, these
metals are chemically stable and hardly dissolve in most
acids. In bulk form, Rh dissolves very slowly even in aqua
regia, and therefore pre-treatment is necessary to obtain a
large surface area to enhance the dissolution process. The re-
covery of PGMs thus requires a large amount of energy and
strong acids. Another problem is that a considerable amount
of wastewater is also generated which is difficult to treat; this
is therefore a serious environmental issue. For these reasons,
it is important to develop an efficient dissolution process
for PGMs.
Precious metals have long been used to manufacture
jewelry and ornaments. Platinum group metals (PGMs) is
the generic term used to refer to the six precious metal
elements—platinum, palladium, rhodium, iridium, ruthe-
nium, and osmium. As the physical and chemical properties
of these metals were revealed, their application expanded
to many other fields. PGMs have unique properties of heat
resistance, corrosion resistance and catalytic activity, and
are used in many fields such as electronics, and also as den-
tal alloys and automotive catalysts [1]. Among the PGMs,
rhodium (Rh) has a special catalytic activity to reduce ni-
trogen oxides (NOx) to N2, and is currently an essential
element as an automotive catalyst. Rhodium is the most
stable element, namely, it is the most difficult material to
dissolve in aqueous solution.
Because of their scarcity and high price, the recovery of
PGMs from scrap is an important issue. The PGMs recov-
ery process generally consists of two stages. The first stage
The authors have been studying a new process to recover
PGMs, especially platinum (Pt), from scrap. The extraction
mediums used in this process are reactive metal (R), such as
alkaline earth elements like magnesium (Mg) and calcium
∗
Corresponding author. Tel.: +81-3-5452-6298;
fax: +81-3-5452-6299.
E-mail address: kayanuma@iis.u-tokyo.ac.jp (Y. Kayanuma).
0925-8388/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0925-8388(03)00666-2

212
Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
Fig. 2. Experimental procedures of rhodium recovery using reactive metals
as an extraction medium.
2. Experiment
Fig. 1. Vapor pressure of reactive metals and rhodium as a function of
temperature (aM = 1).
2.1. Experimental procedures
Fig. 2 illustrates the experimental procedures of the Rh
recovery process using reactive metals as the extraction
medium. First, reactive metal vapor was supplied to Rh at
an elevated temperature ranging from 873 to 1173 K, and an
R–Rh (reactive metal–Rh) compound was formed. In some
experiments, the obtained R–Rh compounds were oxidized
in air at 1173 K for 1 h. Experiments on dissolution of the
R–Rh compound or the oxidized R–Rh compound were car-
ried out using aqueous HCl solution or aqua regia. The so-
lution thus obtained was then subjected to the Rh recovery
process, Rh was recovered from the solution using the pre-
cipitation or cementation method.
(Ca) supplied in vapor form. In a previous study, it was
demonstrated that compounds of those metals and platinum
(R–Pt compounds) or oxidized samples of these compounds
(RPtxOy) dissolved easily in aqua regia [8,9]. Based on these
results, the authors used this reactive metals vapor treatment
for the Rh recovery process. The advantage of using reac-
tive metals like Mg and Ca is that Rh has a strong affinity
for these metals, leading to the formation of a number of
intermetallic compounds including Mg44Rh7 and CaRh2
[10]. Furthermore, complex oxides of the R–Rh–O ternary
system such as MgRh2O4 [11] and CaRh2O4 [12], also
exist. Using the reactive metal treatment to form these com-
pounds, it is possible to enhance the dissolution ratio of Rh
in acid. Another advantage of using reactive metals is that
they can be supplied in vapor form to the Rh present in the
scrap. Fig. 1 illustrates the vapor pressure of selected reac-
tive metals and Rh as a function of temperature—calculated
from reference data [13]. The vapor pressure of reactive
metals is sufficiently higher than Rh at temperature ranging
from 873 to 1173 K, and it can supply the reactive metals
vapor to the solid form of Rh that exists in the complicated
structure of scraps. By controlling temperature and reaction
time, it is possible to regulate the amount of vapor supplied.
In this paper, we have undertaken a fundamental study on
the reaction between reactive metal vapor and Rh, the dis-
solution of the obtained compounds or oxidized samples in
acid, and the recovery of Rh from the solution.
2.2. R–Rh compound formation
The materials used in this study are listed in Table 1. Rh
powder (from 10 to 100 ␮m in size) was prepared from
a small Rh block by sawing and removing small Fe parti-
cles by magnet. Mg lumps (∼2 g, 10 × 10 × 10 mm) were
cut from Mg ingots, and Ca chips (∼0.5 g) were cut from
bulk form of Ca. Fig. 3 illustrates the experimental appara-
tus for reactive metal vapor deposition on Rh powder. The
Rh powder (0.5 g) was placed on a stainless steel sample
holder. The sample holder and reactive metal were covered
with a Ti sheet (0.2 mm thick) and a stainless steel sheet
(0.1 mm thick). The holder was placed in the middle of a
stainless steel container, and the reactive metal was placed
at the bottom of the container, so that its vapors could be

Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
213
Table 1
Materials used in this study
Material
Form
Purity or Supplier
conc. (%)
Rh
Mg
Ca
Ti
Ingot
Ingot
Chips
Sponge
Powder
Powder
Powder
99.9 up
99.0 up
98.0
99.0
99.0
98.5 up
99.0 up
99.5 up
99.0 up
Tanaka Kikinzoku Kogyo K. K.
Hirano Seizaemon Co., Ltd.,
Mintech Japan K. K.
Toho Titanium Co., Ltd., 1.0–2.0 g
Kanto Kagaku, 1–5 ␮m in diameter
Kanto Kagaku
Kanto Kagaku
Kanto Kagaku
Kanto Kagaku
Kanto Kagaku
Zn
NaNO2
NH4Cl
2-Propanol Liquid
Acetone
HCl
Liquid
Aqueous 35.0a
Aqueous 60.0a
HNO3
Kanto Kagaku
a
Concentration of the solution.
Fig. 3. Experimental apparatus for supplying reactive metal vapor to
metallic rhodium samples.
efficiently supplied to the Rh powder. After preparation, the
sample containers were installed in the stainless steel reac-
tion vessel. In some experiments, an R–Rh compound was
synthesized in a tantalum crucible for reference, and the ob-
tained samples were subjected to oxidation and dissolution
experiments. Sponge Ti was also placed in the vessel to ab-
sorb the nitrogen and remove extraneous oxygen. The vessel
was sealed by tungsten inert gas (TIG) welding. The vessel
containing Rh and R (i.e. Mg or Ca) was then put in an elec-
tric resistance furnace maintained at a constant temperature,
and the pure Rh sample was made to react with the reactive
metal vapor.
from the furnace, quenched in water, mechanically opened
and the samples were recovered from the sample holders.
The obtained R–Rh compound samples were then crushed
in mortar; some of the samples were subsequently oxidized
in air. In the oxidation experiment, the weighed sample com-
pounds were put into an alumina crucible and heated in air
using an electric resistance furnace at 1173 K for 1 h. After
reaction, the alumina crucible was removed from the fur-
nace and cooled. The obtained samples were again weighed
and analyzed, and were then subjected to dissolution exper-
iments.
Experimental conditions are listed in Table 2. To main-
tain vapor pressure at least above 10−3 atm, the reaction
temperature ranged from 873 to 1173 K for Mg, and 1073
to 1173 K for Ca. The holding time was fixed at 2 or 3 h.
After the heat treatment, the reaction vessel was removed
2.3. Dissolution experiment
Aqueous HCl solution and aqua regia (a 3:1 mixture of
aqueous HCl solution and aqueous HNO3) were used for
Table 2
Experimental conditions of reactive metal (R)–Rh alloy formation and analytical results of the obtained compounds
Sample
Mass of
Rh, wRh
(g)
Reactive
metal, R
Contact
of R with
Rh
Mass of
R, wR (g)
Reaction
temperature,
T (K)
Reaction
time,
tꢀ (h)
Analytical results after reaction
Mass of
Composition
of compound
(mass %)a
obtained
compound,
wcomp. (g)
Mg, CMg
Ca, CCa
Rh, CRh
A
B
C
D
0.50
0.50
0.51
0.49
Mg
Mg
Mg
Mg
Vapor
Vapor
Vapor
Vapor
2.13
2.16
2.11
2.18
873
973
1073
1173
3
3
2
3
0.48
0.68
1.05
1.94
0.0
19.3
54.6
77.5
77.4b
59.2
72.7b
48.1
64.7b
100.0
80.7
45.4
22.5
22.6b
40.8
27.3b
51.9
35.3b
83.7
70.4
71.8
51.4
24.4b
E
F
0.51
0.40
Mg
Mg
Vapor
2.06
1.01
1173
1173
3
3
1.80
0.87
Liquid
G
H
I
0.41
0.55
0.50
0.40
Ca
Ca
Ca
Ca
Vapor
Vapor
Vapor
Liquid
2.13
2.12
2.19
0.80
1073
1173
1173
1173
3
3
3
3
0.43
0.65
0.62
0.98
16.3
29.6
28.2
48.6
J
a
Determined by EDS.
Determined by ICP.
b

214
Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
dissolution. The sample was put in a 100 ml beaker, and
15 ml of aqueous HCl solution was then added. When the
reaction between the sample and acid was expected to be
very fast and gas evolution to be violent, the sample was
first immersed in distilled water, and then acid was slowly
added to it. In the dissolution experiment using aqua regia, 5
ml of concentrated aqueous HNO3 was added to the sample
after placing it in an aqueous HCl solution. The solution
was then heated using a hot plate, and was maintained at
a temperature ranging from 323 to 333 K. The dissolution
time was fixed at 1 h. After the experiment, the solution and
residue were separated using a filter paper (JIS P 3801, No.
5C). The residue was washed with a 1 mol/l aqueous HCl
solution, dried over a hot plate, and weighed. The filtrate
was analyzed for its Rh content.
plasma-atomic emission spectrometry (ICP-AES, Seiko
Instruments, SPS 4000).
3. Results and discussion
3.1. Vapor deposition of reactive metals on Rh and
formation of R–Rh compound
The compositions of the compounds obtained at various
temperatures are summarized in the last three columns of
Table 2. Mg was not observed on the surface of Rh when
treated at 873 K (sample A). It was found that the vapor
pressure of Mg at this temperature (p◦Mg = 1.2 × 10−3 atm
at 873 K) was too low for compound formation (Fig. 1).
2.4. Rh recovery experiment
Rh was recovered from the solution by two different meth-
ods. One was by the precipitation of Rh as an ammonium
nitrite salt, and the other was by cementation using zinc (Zn)
powder [14]. In each process, the Mg–Rh compound was
dissolved in aqua regia, and the solution was then heated
and dried to remove nitric acid. The dried compound was
again dissolved in 1 mol/l aqueous HCl solution and then
subjected to Rh recovery.
In the process of Rh recovery by precipitation, sodium ni-
trite (NaNO2) was added to the solution until the generation
of nitrous oxide gas was no longer observed. The solution
was then aged at 343–353 K for 1 h, and rapidly cooled in ice
water. Saturated ammonium chloride (NH4Cl) solution was
added to the solution while cooling. The white precipitate
generated after cooling was recovered by filtration, rinsed
with saturated NH4Cl solution, then dried on a hot plate.
The chemical composition of the precipitate recovered from
the filter paper was analyzed.
In the process of Rh recovery by cementation, ∼7 g of
Zn powder (1–5 ␮m in diameter) was added to the solution
containing Rh (100 ml) at 323 K, until the solution color
became clear and a black precipitate was obtained. After
keeping the solution containing the precipitate at 323 K for
1 h, the precipitate was recovered by centrifugation, washed
by distilled water, 2-propanol and acetone, and then dried
in a vacuum dry oven at 473 K for 1 day, and Rh content in
the precipitate was analyzed.
2.5. Analysis
A surface analysis of the obtained solid sample was
performed using a scanning electron microscope (SEM,
JEOL, JSM-5600LV), and the distribution of elements
in the samples was analyzed using an energy-dispersive
X-ray spectrometer (EDS, JEOL, JED-2200). The phases
in the samples were identified using X-ray powder diffrac-
tion (XRD, Rigaku, RINT2000) analysis. The amount of
dissolved Rh was measured using inductively coupled
Fig. 4. SEM image (a) and corresponding EDS images (b, c) of Mg–Rh
compound obtained after Mg vapor deposition on Rh (sample D, R = Mg,
T = 1173 K, tꢀ = 3 h).

Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
215
At temperatures exceeding 973 K, the Mg content in the
sample increased as temperature increased. At 1173 K, the
Mg content in the Mg–Rh compound reached over 70%
Mg by mass (samples D and E). Sample F in Table 2 was
immersed in liquid Mg. The obtained compound was almost
the same composition as the vapor treated sample. In the
case of Ca treatment at 1073 K, 16.3 mass % of Ca was
observed on the surface of Rh (sample G, p◦Ca = 8.5 × 10−4
atm at 1073 K). At 1173 K, the Ca content in the sample
reached ∼30% Ca by mass (samples H and I). Liquid Ca
was used to synthesize Ca–Rh compounds (sample J). The
obtained sample showed a larger content of Ca than the vapor
treated samples. This implies that the synthesis of Ca–Rh
compounds may be controlled by the rate of mass transfer
of Ca or the interfacial chemical reaction.
Fig. 4 shows the SEM image and EDS results of the ob-
tained Rh powder after reacting Mg vapor at 1173 K for
3 h (sample D in Table 2), and Fig. 5 shows the obtained
Rh powder after reacting Ca vapor under the same experi-
mental condition (sample I in Table 2). The surfaces of the
compounds were homogeneous by SEM observation in the
present study. EDS analysis showed a wide distribution of
reactive metals on Rh. XRD analysis was conducted to de-
termine the independent phase, however, no apparent phase
of the compound was observed.
Experimental results of oxidation are summarized in
Table 3. From XRD analysis, MgO, Rh2O3 and MgRh2O4
were identified in the sample obtained after the oxidation of
the Mg–Rh compound, and CaO, Rh, Rh2O3 and CaRh2O4
were identified in the sample obtained after the oxidation
of the Ca–Rh compound. After oxidation, mass of the sam-
ple increased to ∼42% in the Mg–Rh compound and 15%
in the Ca–Rh compound. Based on the hypothesis that the
increase in mass of the obtained sample is due to the ox-
idation of reactive metals and rhodium, and assuming the
reaction, R–Rh compound+O2→RO+Rh2O3, the calcu-
lated mass of each sample is listed in parenthesis in the
fourth column of Table 3. Considering the loss during han-
dling of the compound powders, and the error of weighing,
the calculated mass is in agreement with the experimental
results. These results indicate that most of the R–Rh com-
Fig. 5. SEM image (a) and corresponding EDS images (b, c) of Ca–Rh
compound obtained after Ca vapor deposition on Rh (sample I, R = Ca,
T = 1173 K, tꢀ = 3 h).
Table 3
Results of R–Rh compound oxidation experiments at 1173 K for 1 h, and analytical results of the obtained compounds
Sample
Reactive
metal, R
Mass of feed
compounda,
wcomp. (g)
Mass of obtained
sampleb, woxi. (g)
Composition of samplec (mass %)
Mg,
Ca,
Rh,
O,
CO
CMg
CRh
K
Mg
Ca
0.40
0.56
(0.60)
16.3
(24.2)
(64.7 mass % Mg)
L
0.36
0.42
65.8
28.0
(6.2)
(75.6 mass % Ca)
(0.50)
a
R content of the R–Rh compound was determined by ICP analysis (Table 2).
Values in the lower line in parenthesis were calculated based on the mass and composition of the compound assuming the following reaction: R–Rh
b
compound+O2→RO+Rh2O3.
c
Determined by EDS. Values for oxygen analysis are listed for reference.

216
Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
pound was converted to RO, Rh2O3, and RRh2O4 complex
oxide mixtures after oxidation.
3.2. Rh dissolution in acid
Dissolution experiments of the obtained R–Rh com-
pounds or their oxidized samples using aqueous HCl solu-
tion and aqua regia were carried out at 323–333 K. For the
purpose of reference, pure Rh powder was also subjected
to dissolution. Following the violent reaction of the R–Rh
compounds with the acids, the temperature of the solution
increased rapidly and hydrogen gas was generated. Dur-
ing dissolution of the oxidized samples, the generation of
hydrogen gas was suppressed.
In the case of dissolution using aqueous HCl solution, the
color of the solution did not change after the experiment,
and a black residue was observed. Fig. 6 shows SEM im-
ages of the residues obtained after dissolution of the R–Rh
compounds by aqueous HCl solution. In both residues, small
holes were observed on the surface. The EDS analysis only
detected Rh in the residues. Results of the dissolution ex-
periments using aqueous HCl solution are listed in Table 4.
Pure Rh did not dissolve in aqueous HCl solution. In the
case of R–Rh compounds, the samples synthesized from
reactive metal and Rh (samples F and J in Table 2) were
used. The mass of Rh in feed samples (wRh) and the mass
of residue (wres) were almost the same, and the mass % of
dissolved Rh was very low. These results indicate that re-
active metals contained in R–Rh compounds dissolved in
aqueous HCl solution, but Rh did not dissolve in HCl and
remained as a residue. In the oxidized samples, wRh and wres
were different, and in the oxidized sample of Mg–Rh com-
Fig. 6. SEM images of Rh residue obtained after leaching in aqueous
HCl solution. Initial sample: (a) Mg–Rh, (b) Ca–Rh compound.
Table 4
Results of the experiments on dissolution of Rh, R–Rh compounds and their oxidized samples in aqueous HCl solutiona
Experiment
Leaching
sample
Mass of
leaching
sample,
wi (g)
Mass of Rh
in sampleb,
wRh (g)
Residue,
wres (g)
Mass of
Mass of
dissolved
Rhc,
Mass % of
dissolved
Rꢀhd,
dissolved
material,
wsol (g)
wd (g)
C Rh
( = wi−wres
)
1
2
Rh
0.050
0.099
0.050
0.035
0.048
0.031
0.002
0.002
4.0
Mg–Rh
0.068
0.002
7.0
(0.004)
(11.3)
3
4
Oxidized
Mg–Rh
0.099
0.099
0.099
0.016e
0.024
0.028e
0.021
0.022
0.022
0.078
0.077
0.077
0.003
(−0.005)
0.001
(0.002)
16.5
–
Ca–Rh
6.0
(9.3)
5
Oxidized
Ca–Rh
0.001
(0.006)
3.4
(20.6)
a
Experimental conditions: solution volume: 20 ml; reaction temperature: 323–333 K; reaction time: 1 h.
Calculated from ICP analysis.
Values in upper line were determined by ICP analysis, and values in lower line in parenthesis were calculated from the following equation:
b
c
wd = wRh−wres
.
d
Values in upper line were calculated from ICP results, and values in lower line in parenthesis were calculated from the following equation:
CꢀRh = (wd/wRh) × 100.
e
Calculated from EDS analysis.

Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
217
K for 1 h. Rh, R–Rh compound and the oxidized sample
of R–Rh compound hardly dissolved in the aqueous HCl
solution.
In the dissolution using aqua regia, the color of the solu-
tion changed from clear yellow to rosy red, indicating that
Rh dissolved in the solution. The experimental results of
dissolution in aqua regia at temperatures ranging from 323
to 333 K are listed in Table 5. In the R–Rh compound dis-
solution, both the samples obtained from vapor deposition
(experiments 2, 3, 6 and 7) and those synthesized from re-
active metal and Rh (experiments 4 and 8) were used, and it
is obvious that Rh dissolution was improved by the reactive
metals treatment. This result shows that the reactive metals
treatment is effective in enhancing the dissolution of Rh in
aqua regia. In the case of the oxidized sample, the mass of
dissolved Rh increased in comparison to pure Rh powder,
but it was not as much as the mass of dissolved Rh for the
R–Rh compound. Fig. 8 shows the results of the mass % of
the dissolved Rh in the dissolution experiments using aqua
regia at 323–333 K for 1 h. Pure Rh powder dissolved only
8 mass % in aqua regia. In the Mg–Rh compound, the mass
% of the dissolved Rh dramatically increased and reached
a maximum of 99% (experiments 2, 3 and 4 in Table 5).
Fig. 7. Percentage of dissolved Rh after leaching Rh, R–Rh compound
and R–Rh oxide samples with aqueous HCl solution (leaching conditions:
mass of sample: 0.05–0.1 g, T = 323–333 K, tꢀ = 1 h).
pound, wd ( = wRh−wres) showed a negative value (experi-
ment 3 in Table 4). This is believed to be due to the mass
of the remaining reactive metal compound. Fig. 7 shows
the results of the mass % of the dissolved Rh in the disso-
lution experiments using aqueous HCl solution at 323–333
Table 5
Results of the experiments on dissolution of Rh, R–Rh compounds and their oxidized samples in aqua regiaa
Experiment
Leaching
sample
Mass of
leaching
sample,
wi (g)
Mass of Rh
in sampleb,
wRh (g)
Residue,
wres (g)
Mass of
Mass of
dissolved
Rhc,
Mass % of
dissolved
Rꢀhd,
dissolved
material,
wsol (g)
wd (g)
C Rh
( = wi−wres
)
1
2
Rh
0.051
0.049
0.051
0.011
0.047
0.000
0.004
0.004
7.8
Mg–Rh
0.049
0.011
98.7
(0.011)
(100)
3
4
5
6
7
8
0.110
0.102
0.101
0.049
0.100
0.100
0.101
0.030
0.036
0.016e
0.034e
0.072e
0.024
0.028e
0.000
0.000
0.028
0.020
0.045
0.000
0.011
0.110
0.102
0.073
0.029
0.055
0.100
0.090
0.029
(0.030)
98.0
(100)
0.033
(0.036)
90.7
(100)
Oxidized
Mg–Rh
0.002
(−0.012)
0.008
(0.020)
13.3
–
Ca–Rh
24.4
(42.0)
0.022
(0.027)
30.4
(37.3)
0.021
(0.024)
87.7
(100)
9
Oxidized
Ca–Rh
0.006
(0.017)
22.9
(61.1)
a
Experimental conditions: solution volume: 20 ml (15 ml of conc. HCl aq. and 5 ml of conc. HNO3 aq. were mixed); reaction temperature: 323–333
K; reaction time: 1 h.
b
Calculated from ICP analysis.
c
Values in upper line were determined by ICP analysis, and values in lower line in parenthesis were calculated from the following equation:
wd = wRh−wres
.
d
Values in upper line were calculated from ICP results, and values in lower line in parenthesis were calculated from the following equation:
CꢀRh = (wd/wRh) × 100.
e
Calculated from EDS analysis.

218
Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
Fig. 8. Percentage of dissolved Rh after leaching Rh, R–Rh compound
and R–Rh oxide samples with aqua regia (leaching conditions: mass of
sample: 0.05–0.1 g, T = 323–333 K, tꢀ = 1 h).
Dissolution results of Ca–Rh compound also showed an
increment in the dissolution ratio of Rh, and it reached a
maximum of 88% (experiments 6, 7 and 8 in Table 5). The
reason for such improvements in dissolution ratio is not
fully understood but is probably due to increment of the
surface area of the dissolving samples. This surface modi-
fication enhanced the reactivity of Rh with aqua regia, and
the Rh dissolution was accelerated. In the case of the oxi-
Fig. 9. SEM image (a) and XRD pattern (b) of the precipitate obtained
from Mg–Rh compound by precipitation (experiment 2 in Table 6).
Table 6
Experimental results of Rh recovery from the Mg–Rh compound by precipitation
Experiment
Starting sample
Mass of
Mass of Rh in samples
Overall
compound,
wcomp. (g)
Initial,
winit. (g)
Rh in
solutiona,
wA (g)
Recovered
(NH4)2Na[Rh(NO2)6],
wB (g)
Recovered
Rhb,
wRh (g)
yieldc,
Yall (%)
1
Mg–Rh
(sample D in Table 2)
0.102
0.102
0.023
0.028
0.023
0.075
0.018
76.4
86.0
2
Mg–Rh
0.027
0.102
0.024
(sample E in Table 2)
a
Samples were dissolved in aqua regia.
b
c
wRh = wB × NRh/NA (NRh: atomic weight of Rh, 102.9, NA: molecular weight of (NH4)2Na[Rh(NO2)6], 437.9).
Yall = 100 × wRh/winit.
.
Table 7
Experimental results of Rh recovery from the Mg–Rh compound by cementation
Experiment
Starting sample
Mass of
Mass of Rh in samples
Overall
compound,
wcomp. (g)
Initial,
winit. (g)
Rh in
solutiona,
wA (g)
Sediment,
wB (g)
Rh content
in sedimentb,
CRh (mass %)
Recovered
Rhc,
wRh (g)
yieldd,
Yall (%)
1
Mg–Rh
(sample D in Table 2)
0.421
0.407
0.095
0.111
0.096
0.118
0.141
81.9
0.096
101
2
Mg–Rh
0.111
78.0
0.110
98.9
(sample E in Table 2)
a
Samples were dissolved in aqua regia.
Determined by EDS.
b
c
d
wRh = wB × CRh/100.
Yall = 100 × wRh/winit.
.

Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
219
dized sample, the mass % of dissolved Rh reached 10–20%
(experiments 5 and 9 in Table 5). Rh in the oxidized sample
was in the form of oxide or complex oxide, and unlike the
R–Rh compounds, neither of these dissolved easily in aqua
regia.
shows the SEM image of the compound recovered by pre-
cipitation, and Fig. 9(b) shows its XRD pattern. The recov-
ered precipitate was identified as a mixture of ammonium
sodium rhodium nitrite, (NH4)2Na[Rh(NO2)6], and NH4Cl.
In this Rh recovery experiment from Mg–Rh compound, the
precipitate was found to be free of Mg. Results of Rh recov-
ery and the yields of Rh by precipitation are summarized
in Table 6. The overall yield of Rh was calculated from the
mass that was recovered and the initial content. In this study,
the yield of recovered Rh as (NH4)2Na[Rh(NO2)6] reached
a maximum of 86%. The loss of Rh occurred mainly during
the filtration procedure.
3.3. Rh recovery from aqueous solution
The solution obtained after dissolution of the Mg–Rh
compound in aqua regia was used for Rh recovery. Fig. 9(a)
Fig. 10(a) shows an SEM image of the recovered
precipitate obtained from the solution containing Rh by
cementation using Zn powder. The recovered precipitate
was observed as spherical particles 2–3 ␮m in diameter.
Fig. 10(b) and (c) are the corresponding EDS images of the
precipitate shown in Fig. 10(a). EDS analysis revealed that
the precipitate contained ∼80 mass % of Rh and about 20
mass % of Zn. An XRD pattern of the precipitate shown in
Fig. 10(d) indicates that the precipitate is mostly metallic
Rh, and the Zn phase was not identified by XRD analysis.
The existence of Zn was not clear, but it was expected to be
present due to a reaction of Zn with Rh in the cementation
process or adsorption of Zn ion on the surface of the re-
duced Rh powder. Results of Rh recovery and the yields of
Rh by cementation are summarized in Table 7. The overall
yield of Rh, calculated as stated, was almost 100%. This
shows that Rh was almost completely recovered from the
solution by cementation.
4. Conclusion
Reactive metal (Mg or Ca) vapor or liquid was reacted
with Rh powder at temperatures between 873 and 1173 K.
The Mg–Rh compound was formed at temperatures above
973 K, and the Ca–Rh compound at temperatures above
1073 K. The obtained compounds were homogeneous within
the sensitivity of the SEM-EDS analysis.
Dissolution experiments of the obtained compounds were
conducted using aqueous HCl solution and aqua regia at
323–333 K for 1 h. In the case of aqueous HCl solution, only
reactive metals in the compounds were dissolved, and Rh
residue, with a large surface area, remained. In aqua regia,
the dissolution ratio of the R–Rh compound was significantly
higher, and the mass % of the dissolved Rh reached 99% for
the Mg–Rh compound, and 88% for the Ca–Rh compound.
Rh recovery from the Mg–Rh compound was carried out
by precipitation and cementation. Using precipitation, Rh
was recovered as (NH4)2Na[Rh(NO2)6], and the yield was
86%. Using cementation, Rh was almost completely recov-
ered as a mixture with Zn.
The results demonstrated in this study show the possibility
of developing a new Rh dissolution process using reactive
metals.
Fig. 10. SEM image (a), corresponding EDS images (b, c) and XRD
pattern (d) of the precipitate obtained from Mg–Rh compound by cemen-
tation (experiment 2 in Table 7).

220
Y. Kayanuma et al. / Journal of Alloys and Compounds 365 (2004) 211–220
[3] F.E. Beamish, Talanta 5 (1960) 1.
[4] L.M. Banbury, F.E. Beamish, Fresenius’ Z. Anal. Chem. 211 (1965)
Acknowledgements
178.
The authors are grateful to Ms. Sachiko Yamamoto,
Messrs. Wataru Matsuno, Shusuke Iwata, and Itaru Mae-
bashi for their experimental assistance. They greatly ap-
preciate Dr. Masao Miyake for his suggestions on the
preparation of the manuscript.
[5] G.H. Faye, P.E. Moloughney, Talanta 19 (1972) 269.
[6] A. Diamantatos, Anal. Chim. Acta 94 (1977) 49.
[7] J.E. Hoffmann, J. Met. June (1988) 40.
[8] T.H. Okabe, Y. Kayanuma, S. Yamamoto, M. Maeda, Mat. Trans.
JIM 44 (2003) 1386.
[9] T.H. Okabe, S. Yamamoto, Y. Kayanuma, M. Maeda, J. Mater. Res.
18 (2003) 1960.
[10] H. Okamoto, Phase Diagrams for Binary Compounds, ASM Inter-
national, Materials Park, OH, 2000.
References
[11] J. Nell, H.St.C. O’Neill, Geochim. Cosmochim. Acta 61 (1997)
4159.
[12] K.T. Jacob, Y. Waseda, J. Solid State Chem. 150 (2000) 213.
[13] I. Barin, Thermochemical Data of Pure Substances, 3rd Edition,
VCH, Weinheim, 1995.
[1] T. Kendall (Ed.), Platinum 2002 Interim Review, Johnson Matthey,
London, 2002.
[2] R.K. Mishra, in: R.K. Mishra (Ed.), Proceedings of the 17th Inter-
national Precious Metals Conference, International Precious Metals
Institute, Allentown, PA, 1993, p. 449.
[14] A. Okuda, J. Shibata, Shigen-to-Sozai 118 (2002) 1.