Properties of nitric acid palladium solutions with a high metal concentration

Article abstract of DOI:10.1134/S1070427207050023

Nitric acid solutions (c_Pd up to 3.2 M) with variable HNO ₃ concentration were studied by electronic spectroscopy, ¹⁴N and ¹⁷O NMR, acid-base titration, gravimetry, and other methods. Solid phases that precipitated on storing these solutions were studied by X-ray phase analysis, thermogravimetry, and IR spectroscopy. The conditions for stability of particular Pd(II) species were determined, and specific features of aging of such highly concentrated palladium solutions were revealed. A procedure for palladium isolation as [Pd(NH₃) ₄](NO₃)₂ from nitric acid solutions in 98% yield was developed.

Full text of DOI:10.1134/S1070427207050023

ISSN 1070-4272, Russian Journal of Applied Chemistry, 2007, Vol. 80, No. 5, pp. 695 704. Pleiades Publishing, Ltd., 2007.
Original Russian Text A.B. Venediktov, S.V. Korenev, S.P. Khranenko, S.V. Tkachev, P.E. Plyusnin, S.N. Mamonov, L.V. Ivanova, V.A. Vostrikov, 2007,
published in Zhurnal Prikladnoi Khimii, 2007, Vol. 80, No. 5, pp. 716 725.
INORGANIC SYNTHESIS
AND INDUSTRIAL INORGANIC CHEMISTRY
Properties of Nitric Acid Palladium Solutions
with a High Metal Concentration
A. B. Venediktov, S. V. Korenev, S. P. Khranenko, S. V. Tkachev, P. E. Plyusnin,
S. N. Mamonov, L. V. Ivanova, and V. A. Vostrikov
Nikolaev Institute of Inorganic Chemistry, Siberian Division, Russian Academy of Sciences,
Novosibirsk, Russia
Gulidov Krasnoyarsk Factory of Nonferrous Metals, Open Joint-Stock Company, Krasnoyarsk, Russia
Received May 4, 2006; in final form, September 2006
Abstract Nitric acid solutions (cPd up to 3.2 M) with variable HNO3 concentration were studied by
electronic spectroscopy, 14N and 17O NMR, acid-base titration, gravimetry, and other methods. Solid phases
that precipitated on storing these solutions were studied by X-ray phase analysis, thermogravimetry, and
IR spectroscopy. The conditions for stability of particular Pd(II) species were determined, and specific features
of aging of such highly concentrated palladium solutions were revealed. A procedure for palladium isolation
as [Pd(NH3)4](NO3)2 from nitric acid solutions in 98% yield was developed.
DOI: 10.1134/S1070427207050023
The main body of the data on speciation of plati-
num group metals (PGMs) in media unconventional
for their refining, in particular, in nitrate, sulfate, and
phosphate solutions, were obtained by the middle of
the 1970s. The results of these studies are summarized
in reviews cited in [1]. According to the data ob-
tained, PGMs in these acid solutions exist in the form
of polynuclear or oligomeric species, and chemical
processes occurring in the systems are extremely
complicated and variable. It should be noted that
the data obtained concerned solutions with low PGM
concentrations (of the order of mM). In the sub-
sequent decades, studies in this direction were not
systematic.
in the macrocomponents of the systems,
and to evaluate the stability of solutions in their stor-
age under various experimental conditions.
EXPERIMENTAL
The electronic absorption spectra (EAS) of all solu-
tions without dilution were recorded on a Hewlett
Packard Vectra instrument in cells of 0.0019 cm
thickness, and the NMR spectra, on a Bruker CXP-
300 spectrometer with a working frequency of
14
17
21.865 MHz for N and 40.700 MHz for O, at
natural abundance of the isotopes in solutions. The
chemical shifts, which are positive in a weak field, are
given in the scale and measured relative to conven-
tional standards. The error in determining chemical
shifts did not exceed 0.05 ppm. Thermogravimetric
studies were carried out on a Q-1000 device modified
for operating in various gaseous media with numerali-
zation of an analog signal. The other data were ob-
tained using standard modern devices and equipment.
At present, interest in such studies, mainly con-
cerning the speciation and stability of PGM com-
plexes in solutions other than hydrochloric acid, in-
creased. It is caused by the need for expanding the
assortment of commercially produced PGM com-
pounds and catalytic systems based on them. Special
attention is given to highly concentrated or saturated
PGM solutions. The progress in this direction is
limited by the lack of experimental data on the PGM
behavior in such media.
The starting nitric acid palladium solutions
(NAPSs) were prepared as follows. A powder of met-
allic palladium containing no less than 99.97% Pd
was dissolved at 90 95 C within 1 h in 40 ml of
The aim of this study was to examine the processes
occurring in solutions maximally concentrated with
3
HNO (ultrapure grade, density d = 1.419 g cm )
3
with addition of 0.02 ml HCl (chemically pure grade,
d = 1.165 g cm ) per 10 g of Pd. Then the solution
respect to platinum metal, using the system Pd HNO
as an example, to reveal the state of Pd and changes
3
3
695

696
m, %
VENEDIKTOV et al.
endothermic effects are clearly detected in the DTA
curves (in air) at 160 and 817 C. The first of them is
caused by the reaction
[Pd(NO3)2(H2O)2]
PdO + H2O + 2HNO3. (1)
The theoretical weight loss for this reaction is
54.06, and the observed loss, 53.8%. The second ef-
fect is caused by PdO decomposition to the metal, as
follows from the correspondence between the theoret-
ical and experimental weight losses.
The data on the crystal structure of [Pd(NO )
3 2
(H O) ] obtained from the powder X-ray patterns by
ab initio method were published in [3]. The starting
substance was synthesized by dissolution of metallic
2
2
T, C
Fig. 1. Curve of [Pd(NO ) (H O) ] weight loss in air.
The weight of the starting sample is taken as 100%.
3 2
2
2
palladium in concentrated HNO with the subsequent
3
(m) Sample weight and (T) temperature.
evaporation of the solution. It is a light brown powder,
relatively stable in air depending on atmospheric
conditions. Dinitratodiaquapalladium(II) has an ortho-
rhombic lattice, space group Pbca, Z = 4. Unit cell
parameters: a = 5.0036(3), b = 10.6073(7), and c =
11.7223(8) . The coordination surrounding of Pd(II)
is square planar, and nitrato groups are monodentate.
It is interesting that in the previous studies concerning
the synthesis and properties of [Pd(NO ) (H O) ] its
3
d, g cm
2 2
2
2
geometric configuration was not discussed. Indirect
data about its trans structure are known [4], and in [3]
its trans configuration was clearly determined.
cPd, M
We have obtained EAS of [Pd(NO ) (H O) ] solu-
3 2
2
2
tions in 2 2.5 M HNO . The concentration of Pd(II)
3
Fig. 2. Dependence of the density d of Pd(II) nitric acid
was varied from 0.00123 to 0.184 M. The spec-
trum of this compound has a very broad absorption
band; therefore, accurate determination of its maxi-
mum is difficult: its position varies from 397 to
404 nm, and the extinction coefficient, from 138 to
solutions on the Pd(II) concentration c . Concentration of
free nitric acid in solutions varied from 1.5 to 2.6 M.
Pd
was evaporated at the same temperature and diluted
with water to c = 2 3 M; the concentration of free
Pd
1
1
HNO was varied from 3.8 to 5.6 M.
145 l mol cm .
3
In certain experiments the starting substance for
obtaining NAPS was crystalline [Pd(NO ) (H O) ].
Determination of acidity. To determine c and
Pd
acidity of NAPS, aliquots of solutions containing
from 100 to 200 mg of Pd(II) were weighed. To ob-
tain correct data, it was necessary to measure the
densities of these solutions. The results of such meas-
urements are shown in Fig. 2.
3 2
2
2
It was prepared from NAPS by concentrating in air to
wet crystals. Then the dish with the wet precipitate
was placed for 1 2 days in a desiccator over KOH
to complete removal of liquid phase traces, after
which it was kept in a desiccator without a desiccant.
The compound was identified by elemental analysis,
IR spectroscopy, X-ray phase analysis, and ther-
mogravimetry.
It is seen from Fig. 2 that the dependence of NAPS
density on Pd(II) concentration is linear and is ex-
pressed by the equation d = 1.03 0.18c . These data
Pd
make it possible to determine accurately the palladium
concentration. The measurement of NAPS density is
also the simplest and the most effective method of
estimating the state of solutions in their storage.
The curve of the weight loss of the synthesized
[Pd(NO ) (H O) ] on heating is shown in Fig. 1. It
3 2
2
2
agrees well with the data [2]. Furthermore, the Pd(II)
content in the crystalline product (39.9 wt %) deter-
mined from this curve corresponds to its content in
[Pd(NO ) (H O) ] (calculated 39.94 wt % Pd). Two
The relatively low reproducibility of the data at c
Pd
of about 3 M will be discussed below. It should be
noted that the insignificant scatter of the experimental
3 2
2
2
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

PROPERTIES OF NITRIC ACID PALLADIUM SOLUTIONS
697
data is caused by two reasons. The first of them is
tion. The main difference was that the excess of the
precipitant did not exceed 10% of the Pd(II) : DMG =
1 : 2 stoichiometry. The mother liquor and the whole
amount of the washing liquid were transferred to a
200-ml volumetric flask, and 50-ml aliquots were
titrated with a standard NaOH solution, using a mix-
ture of Phenol Red and Bromocresol Green indicators
[9]. According to the data of [9], its color changes
at pH 7.2 0.2. It was found experimentally that the
use of indicators with a higher pT results in regular
overestimation of results. When calculating the con-
centration of free HNO , 2c was subtracted from the
the fact that the concentration of free HNO in solu-
3
tions varied from 1.5 to 2.6 M. However, it is possible
to estimate these deviations. In particular, it follows
from the reference data on the HNO density that its
3
change is no more than 3.3 rel. % in this interval of
concentrations, which is the maximal error of the
measurements. The second reason consists in the fact
that the NAPS density was measured at room tem-
perature, which varied within the range 20 28 C. It
was experimentally found that the density variation
under these conditions was about 1.5%.
3
Pd
+
found total acidity, as two H ions are released into
solution upon Pd(II) precipitation with DMG.
Several techniques were tested for the acidity deter-
mination. The first (method no. 1) consisted in that
a NAPS aliquot diluted by a factor of 12.5 was treated
Finally, a direct pH-metric titration of freshly pre-
pared NAPS solutions resulted in the formation of
dark brown precipitates. In this case, it was necessary
to wait for no less than 1 h for reaching a stable po-
tential; therefore, this technique is unacceptable.
with a tenfold amount of solid NaNO relative to
2
Pd(II) for quantitative binding of Pd(II) in the tetra-
nitro complex, with the subsequent pH-metric titration
to pH 6.8 7.0. Unreliability of this technique con-
sisted, firstly, in the formation of small amounts of
nitrogen oxides. Secondly, the titration curves of free
HNO and of the same amount of HNO with an addi-
To check the correctness of techniques for the acid-
ity determination, we have carried out introduced
found experiments using model solutions. They were
prepared by the dissolving accurately weighed por-
tions of crystalline [Pd(NO ) (H O) ] in a fixed
3
3
tion of Na [Pd(NO ) ] and of the required excess of
2
2 4
NaNO differed essentially.
2
3 2
2
2
The second technique (method no. 2) was based on
the results of [5 7] where it was proved that Pd(II) is
volume of HNO of precisely known concentration.
3
The final volume of such solutions was always higher
quantitatively extracted as compounds [Pd(NO ) L ]
3 2 2
than the initial HNO volume; therefore, to express
3
with sulfides according to Eq. (2), with the concentra-
concentrations in the M scale, we determined the
densities of the model solutions.
tion of free HNO remaining unchanged during the
3
process:
Method no. 3 gives the maximum discrepancy of
1.5% between the found and introduced amounts.
However, it is rather labor-consuming, and a DMG
excess larger than 10%, as well as wrong choice of an
indicator, result in the substantial overestimation of
the NAPS acidity. These drawbacks are compensated
by the possibility of the exact determination of both
{Pd2+ + 2NO3}(aq) + 2L(org)
[PdL2(NO3)2](org), (2)
where L designates a molecule of coordinated sulfide.
Then the raffinate containing only HNO was titrated
3
with a standard alkali solution. This technique of the
acidity determination could be close to ideal if it were
possible to determine precisely the volume of the raf-
finate after the separation of Pd(II) by extraction. It
appeared to be extremely difficult; therefore, for cal-
culations we used eight parameters and six equations,
which resulted in the accumulation of errors and low
reproducibility.
c
and the concentration of free HNO from a single
Pd
3
NAPS sample. The causes of regular errors in the
results obtained by method nos. 1 and 2 were con-
sidered above. For the other methods, the deviation of
acidity from that of model solutions did not exceed
5.6 (no. 2) and +9.4% (no. 1).
The third technique (method no. 3) allowed us to
NMR studies of NAPSs. The in situ determination
of the concentration of components in nearly saturated
solutions, like NAPSs under study, is a difficult prob-
lem. The NMR method seems to be the most informa-
tive in these cases. To elucidate its possibilities, we
have carried out several series of experiments in which
we varied the following parameters of solutions: con-
centrations of nitric acid and nitrate ions and the
determine simultaneously the content of free HNO
3
and also c during NAPS storage. Using the data
Pd
of Fig. 2, we took NAPS weighed portions, diluted
them with water approximately to 50 ml, and precipi-
tated Pd(II) with dimethylglyoxime (DMG) by the
standard procedure [8]. The difference from the tech-
nique [8] consisted in that the separated precipitate
was washed only with hot water without its acidifica-
+
2+
nature of a nitrate salt cation (Na or Zn ). Taking
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

698
VENEDIKTOV et al.
The content of Pd(II) in it is 20.07 wt % counting on
the metal (or 3.042 M), and the density of this solu-
3
tion at 20 C is 1.613 g cm .
14
In the N NMR spectrum of NAPS-1 taken in
the range from +460 to 560 ppm, only a symmetric
signal with a chemical shift of 8.93 ppm is detected.
This range essentially exceeds the range of 48
88 ppm characteristic of nitrite ins coordinated to
Pd(II) in various amounts [10]. This fact confirms our
earlier conclusion that Pd(II) nitrite complexes make a
negligible contribution to the material balance of the
NAPS system and allows us to rule out the presence
of noticeable amounts of nitrite ions or nitrous acid
(CS 200 ppm [10]) in Pd(II) nitric acid solutions.
, ppm
14
Fig. 3. Experimental
N NMR spectrum of NAPS-1
diluted by a factor of 2.5 with water and components of its
resolution by the Simplex method. ( ) Chemical shift.
into account the facts that the ionic radii of palladium
and zinc are close to each other and the Zn(II) com-
plex formation with nitrate ions is, presumably,
very weak, we chose zinc as a model simulating
aqueous solutions of Pd(II) nitrate. The following
regular trends were revealed. In 0.5 7 M solutions of
Na NO , the factor defining NO chemical shifts is
the solution acidity. Further, as the acidity of solu-
tions increases, the concentration dependence of the
Na NO chemical shift is displaced upfield and
becomes linear with a slope of 0.70. However, the
linear dependence of the chemical shift on the concen-
tration of Zn(NO ) has a slope of 1.15 when the
HNO concentration is varied. Hence, the dependences
The spectrum of the same solution diluted by a
factor of 2.5 with water, taken immediately after the
dilution, is shown in Fig. 3. After the dilution of the
starting NAPS-1 solution, a shoulder appears in the
high field in all the spectra, and it is the more pro-
nounced, the more diluted is the NAPS-1 solution
(by a factor of 10 at most). This is apparently due to
changes in the Pd(II) speciation. In this case, lines of
irregular and unsymmetric shapes were resolved into
a number of components specified by an operator.
Then, using the Simplex method, we carried out the
optimization of the following parameters of each
component: position in the spectrum (chemical shift),
amplitude, half-width, and Gaussian line contribution
to the shape of the signal resolved after Lorentz.
The optimization continued until the best fit of the
experimental spectrum with the total envelope of all
the components was attained. It should be noted that
the line of undiluted NAPS-1 was resolved neither
into two nor into three components, unlike the lines of
diluted solutions. The correctness of this approach is
illustrated by the fact that the experimental and op-
timized spectra of undiluted NAP-1 coincide to within
1%. This is a very good result for NMR spectroscopy.
14
14
3
3
14
3
3 2
3
of the chemical shift of free nitrate ions both on
their concentration and on the acidity are individual
for each specific cation. Thus, the determination of
the concentration of nitrate ions based on the analysis
14
of regular trends in variation of N chemical shifts
in nitrate solutions shows little promise.
The presence of nitrite ions in NAPSs is of a prin-
ciple importance in their analysis. The identification
of nitrite ions in concentrated nitric acid Pd(II) solu-
tions can be rather simple when using the NMR meth-
od. It was found experimentally that, when the
14
4
number of scans of the N resonance is about 10
The presence of a single symmetric line in the
NMR spectrum of undiluted NAPS-1, which is not
resolved into any number of signals, can be, in prin-
ciple, associated with the closeness of the chemical
shifts of Pd(II) species present in solution and of nitric
(experiment duration about 6000 s), a signal of nitrite
ions can be readily seen at their content of 0.01 M.
Such a detection limit of nitro groups is sufficient for
determination of their content in the systems under
consideration. In any case, when the content of nitro
groups in NAPSs with a Pd(II) content of about 3.0 M
acid, or with a high rate of exchange of coordinated
14
and free nitrate ions, or with the fact that the
N
14
is monitored by N NMR, the contribution Pd(II)
nitro complexes to the material balance is negligible.
NMR signals of NAPSs are considerably broader than
those of nitrate ions. In our case, the line width is
95.8 Hz. To compare, the line width of the signal of
3 M nitric acid is 9 Hz, and that of a nitric acid solu-
tion with a total concentration of nitrate ions of 7.0 M
is 13.9 Hz. The presence of Pd(II) in a concentration
of about 3 M in NAPS can affect the viscosity and
To elucidate the nature of Pd(II) complex species
and to determine the concentrations of certain NAP
macrocomponents, we also used NMR. The over-
whelming majority of such experiments were carried
out with a sample hereinafter designated as NAPS-1.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

PROPERTIES OF NITRIC ACID PALLADIUM SOLUTIONS
699
other parameters of solutions, but the comparison of
the above-given data shows that it would hardly lead
to an increase in the line width in the NAPS-1 spec-
trum by an order of magnitude.
determined from the extraction data sharply differ
from the results of [11, 12]. This question is discussed
in [13]. The approach to the estimation of the con-
stants is based on the examination of the dependences
of the Pd(II) distribution coefficient D on the con-
Pd
This fact, together with the asymmetry of lines of
NAP solutions diluted with water, was the reason why
we attempted to resolve the NMR spectra of dilute
solutions into components.
centration of HNO in the aqueous phase in the case
3
of the extraction from nitric acid media with neutral
phosphoryl-containing compounds, nitrates of amines,
and salts of quarternary ammonium bases. In so doing,
it was postulated that the nonideality of the aqueous
phase is caused solely by the formation of Pd(II)
The change in the symmetry of lines in the spectra
of solutions diluted with water and the presence of
three resolved components of these spectra adequately
describing the experimental spectra suggest either
the appearance of hydrolyzed Pd(II) species upon
lowering the acidity or the dissociation of complex
Pd(II) nitrates resulting in an increase in the concen-
nitrato complexes. Such approach gives the
value
1
lying in the range 14 20. Moreover, the constants
of formation of di- and trinitrato complexes were es-
timated at 2 and 6, respectively. However, Shmidt
et al. [13] consider these estimates of
pirical parameters in the dependences of D
only as em-
i
tration of NO ions. These reasons are supported
on
3
Pd
by the data on the stability Pd(II) nitratoaqua com-
HNO concentration in the aqueous phase, as there are
3
2+
plexes and by the constants of [Pd(H O) ] acid dis-
sociation.
no quantitative approaches to the determination of
the activity coefficients in a wide range of conditions.
2
4
Stability constants of Pd(II) nitratoaqua com-
plexes. There are only three publications [11 13] in
which the constants of formation of Pd(II) nitratoaqua
complexes are given. Shoittsov [11] reported that a
bathochromic shift in EAS of Pd(II) aqua complexes
The first paper on the hydrolysis of Pd(II) aqua
ions was published in 1967 [14]. In that paper, solu-
tions of palladium aqua ions (c = 0.001 0.006 M)
Pd
were prepared by dissolving in perchloric acid the pal-
ladium hydroxide precipitated from nitric acid solu-
tions. Hydrolysis constants were determined both
from the data of pH-metric titration and by spectro-
photometry, and then extrapolated to zero ionic
strength. Reaching constant pH values took about
5 h in each point of the titration curve. The con-
with
concentrations of nitrate ions and Pd(II) at c
0.008 M. In that study, the formation of Pd(II) nitrato
complexes was examined spectrophotometrically up to
= 390 nm is observed at 2.1 : 1 ratio of
max
=
Pd
3.6 M NO concentration at a total ionic strength of
3
4.0 M. It was found that only the mononitrato complex
+
2+
stants
= [PdOH (aq)]/[Pd (aq)][OH ] and
=
+
1
2
[Pd(H O) (NO )] with a stability constant
= 1.6
2+
2
2
3
3
1
[Pd(OH) (aq)]/[Pd (aq)][OH ]
were determined.
at 25 C averaged by
2
0.3 is formed in the examined concentration range.
The calculated EAS parameters of this species are as
= 415 nm and
The formation constant of the mononitrato complex
The values of log
and log
1
2
both methods are 12.7 0.6 and 26.2 0.7, respective-
1
1
follows:
= 155 l mol cm .
max
max
ly. The logarithm of the constant K of the equilibri-
s
um [Pd(OH) (s)] = [Pd(OH) (s)] is 2.7. Combining
2
2
= 1.2 0.4 was also determined spectrophotomet-
1
1
this value with log
, we obtain the solubility prod-
2
rically in [12]. Note that the
values from [11, 12]
1
28.9
uct Pd(OH) (s) equal to 10
. The constant
completely agree with each other within the limits of
the errors given in these papers. The parameters of
2
1
found for the reaction
+
[Pd(H O) NO ] EAS in these two papers also
2
3
3
[Pd(H2O)4]2+ + OH = [Pd(H2O)3OH]+ + H2O (3)
practically coincide. It was shown in [12] that the
position of the maximum and its intensity in EAS of
can be readily converted to the constant K of
+
d
[Pd(H O) NO ] noticeably decrease both on replace-
2+
2
3
3
[Pd(H O) ] dissociation by the equation
2
4
ment of HNO by NaNO and with a decrease in
HNO concentration. The procedures for preparing
3
3
[Pd(H2O)4]2+ = [Pd(H2O)3OH]+ + H+.
In this case, K = K . The values of pK calcu-
(4)
3
the nitrate solutions in [11, 12] are the same and are
based on the dissolution of Pd powder in concentrated
d
1
w
d
HNO with the subsequent boiling to eliminate pos-
3
lated in such a way from the data of [11, 14] are 1.2
0.1, which shows that the acid dissociation constants
calculated using data from two independent sources
actually coincide. It remains unclear why such an
important quantity as the dissociation constant of
sible sources of NO generation. The range of work-
2
ing Pd(II) concentrations in these studies was 0.001
0.01 M.
The stability constants of Pd(II) nitrato complexes
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

700
VENEDIKTOV et al.
Pd(II) aqua ion has not been calculated previously. It
follows from the value cited above that, first, the
Pd(II) aqua ion is in fact a strong acid, and, second,
in solutions without complexing ligands (e.g., per-
The idea of the first series consisted in the following.
In the NAPS material balance, only hydroxide and
nitrate ions can be anions. To check it, we have taken
the N NMR spectrum of one of NAPSs with an
internal reference, nitromethane (0.124 M) in NAPS
14
+
chlorate) at [H] = 1 M approximately 10% of Pd(II)
+
with the initial concentration c
= 21.05 wt %,
will be in the monohydroxo form, and at [H] = 10 M
Pd
diluted by a factor of 3. As a result, the difference
between the calculated concentration of nitrate ions
and the concentration found from the integral intensi-
ties of the total nitrate lines and the internal reference
line appeared to be 3.5%, which is a very accurate
result for the method in use. The second series of
experiments consisted in measuring pH of freshly
prepared solutions of the same NAPS with a dilution
by a factor from 12.5 to 193. An important result is
the fact that the measured pH values for the first solu-
tion were 0.56 0.04, which agrees well with the
concentration of hydrogen ions calculated on the basis
of the acid dissociation constant of Pd(II) aqua ions.
this fraction will be about 1%. In the specified acidity
range, the fraction of hydrolyzed Pd(II) will always be
several percents.
The dissociation processes are very fast; hence, the
subsequent slow changes in properties of solutions,
mainly precipitation, are probably associated with
slow polymerization of the hydrolyzed species. The
presence of ligands capable of forming complexes
with Pd(II) will lead to a decrease in the strength of
acids such as [Pd(H O) L]; therefore, the degree of
2
3
hydrolysis decreases. Hence, the contribution of poly-
merization processes to the common patztern of the
occurring reactions also decreases.
17
The O NMR spectra of both the starting NAPS-1
solution and solution after its dilution (at most ten-
fold) contain only two lines with chemical shifts of
412 1 ppm and from 121 to 128 (variation with
increasing dilution). The first signal exactly coincides
with the spectrum of 3.12 M nitric acid. The com-
parison of the second signal with the data of [15]
shows that it corresponds to the Pd(II) aqua ion with
a good accuracy. It confirms the fact that the exchange
of free nitrate ions with the nitrate groups coordinated
to Pd(II) is very fast, and the difference between these
ions cannot be detected by NMR.
Speciation of Pd(II) in nitric acid solutions. In
view of the considered data on the stability of nitrate
complexes and acid properties of Pd(II) aqua ions, us-
ing the NMR data for solutions, we can obtain the in-
formation on the nature of species present in NAPS.
14
The regular trends in variation of the N NMR spec-
tra of NAPS-1 solutions on their 2.5-, 5-, and 10-fold
dilution with water are as follows. For the species
giving the lowest-field signal, the shift changes from
2.4 to 0.8 upon dilution, with the fraction of the
species increasing by a factor of almost 2 with dilu-
tion. We can readily assign this signal to free NO
Thus, the above-discussed results allow the follow-
3
ions on the basis of the previously obtained data on
the resonance of concentrated NaNO + HNO solu-
ing conclusion. The undiluted NAPS solution has a
14
symmetric signal in the
N NMR spectrum (
3
3
tions. Hence, on dilution nitrate ions are accumulated
in solutions, which can be caused only by decomposi-
tion of the Pd(II) nitrate complex. The chemical shift
of the next signal varies from 4.2 to 1.6 ppm, with
the fraction of this species, 0.31 0.05, remaining
approximately constant. This signal presumably cor-
responds to the Pd(II) nitratotriaqua complex, which
should be the most stable complex species at these
concentrations. The highest-field signal has a stable
chemical shift ( 7.6 0.1 ppm), with the fraction of
the species decreasing from 0.24 to 0.12. It should be
noted that the ratio of the concentration of coordinated
8.9 ppm, half-width about 96 Hz), contains the
dominating form [Pd(NO ) (H O) ], and approxi-
3 2
2
2
+
mately equal low concentrations of [Pd(NO )(H O) ]
3
2
3
and [Pd(NO ) (H O)] providing the line symmetry.
3 3
2
The following processes occur on dilution of the
solutions with water:
[(PdNO3)3(H2O)] + H2O
[Pd(NO3)2(H2O)2] + H2O
[Pd(NO3)(H2O)3]+ + H2O
[Pd(NO3)3(H2O)2] + NO3,
(5)
[Pd(NO3)(H2O)3]+ + NO3,
(6)
nitrate ions to c , calculated from the integral intensi-
Pd
[Pd(H2O)4]2+ + NO3, (7)
[Pd(OH)(H2O)3]+ + H+.
(8)
ties of lines, decreases on dilution from 1.9 to 1.4.
[Pd(H2O)4]2+
To confirm the absence of hydrolytic processes and
the prevalence of dissociation reactions of complex
Pd(II) nitrates, as indicated by all experimental
data, we have carried our two series of experiments.
Reactions (5) (7) result in increasing concentration
of nitrate ions in solutions, which accounts for the
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

PROPERTIES OF NITRIC ACID PALLADIUM SOLUTIONS
701
essence of the occurring processes. The occurrence of
process (8) is proved without controversy by the fact
that, in all NAPSs diluted with water, dark brown fine
precipitates are formed within 5 7 days after the dilu-
tion. The data of IR spectroscopy, X-ray diffraction,
and elemental analysis suggest that the precipitate is
hydrated palladium oxide Pd(OH) xH O, where
The reaction of 5 ml of Pd(NO ) 2H O aqueous
3 2
2
solution ( 1 M) with an excess of concentrated aque-
ous solution of ammonia at room temperature yielded
a light brown solution containing a small amount of
a voluminous brown precipitate. The reaction mixture
was heated with excess ammonia, the precipitate was
filtered off, and the resulting solution was evaporated
to dryness on a water bath. Yield 98%.
2
2
x varies within the range 1.20 1.45. The greater the
dilution, the higher x. The maximal dilution corre-
The reaction of 10 ml of an aqueous solution con-
taining 0.8 M Pd(NO ) and 1.6 M free HNO with
sponded to the 25-fold decrease in c . Dilute acid
Pd
3 2
3
HCl did not dissolve these precipitates, and concen-
trated HCl, only on boiling. This agrees completely
with the data of [16].
an excess of concentrated aqueous solution of am-
monia at room temperature proceeds as follows. At
the first moment after the addition of the ammonia
solution, a light yellow voluminous flocculent precipi-
tate is formed, which completely dissolves on further
adding ammonia. The resulting solution has a light
brown color. No clear precipitate is observed. Then
the solution is evaporated up to the formation of
a film of crystals and is cooled. The precipitated crys-
tals are separated by filtration. To remove an NH NO
Isolation of Pd(II) from NAPS in the form of
[Pd(NH ) ](NO ) . It should be noted that we were
3 4
3 2
able to reach a nearly quantitative yield of [Pd(NO )
3 2
(H O) ] crystals. The [Pd(NO ) (H O) ] obtained can
2
2
3 2
2
2
be used as the starting compound in the synthesis of a
very important product, [Pd(NH ) ](NO ) . Owing to
3 4
3 2
its high solubility, the compound can find use in
the production of supported palladium catalysts or
precursors of polymetallic powders. Furthermore, the
use of [Pd(NH ) ](NO ) makes it possible to carry
4
3
impurity, the product is recrystallized from aqueous
solution. Yield 70%. The reaction at room tempera-
ture with a solution containing 2.5 g of NH HCO in
3 4
3 2
4
3
out syntheses in the presence of small amounts of
HCl, in contrast to [Pd(NH ) ]Cl whose application
concentrated ammonia proceeds similary.
3 4
2
Our experiments show that small amounts of a pre-
cipitate contaminating the target product are formed
inevitably leads to the formation of trans-[Pd(NH )
3 2
Cl ]. The complex [Pd(NH ) ](NO ) was obtained
2
3 4
3 2
during the synthesis. In the synthesis from Pd(NO )
3 2
from [Pd(NO ) (H O) ] in the presence of an ex-
3 2
2
2
2H O solutions containing free HNO , such precipi-
2
3
cess of concentrated ammonia, after which the result-
ing reaction mixture was evaporated. As a result,
[Pd(NH ) ](NO ) was obtained in a high yield:
tate is not formed. We failed to determine the com-
position of the precipitate. However, data of IR spec-
troscopy and of X-ray phase and thermal analysis
suggest that the precipitate composition can be ex-
pressed by the general formula PdO(NH ) (H O) .
3 4
3 2
[Pd(NO3)2(H2O)2] + 4NH3
+ 2NO3 + 2H2O.
[Pd(NH3)4]2+
3 x
2
y
(9)
In particular, this assumption is supported by the fact
that the dried precipitates of this phase readily dis-
solve in dilute HCl without heating, which is quite
uncharacteristic of dried PdO(H O) .
To optimize the synthesis procedure, a series of
experiments with various starting conditions have
2
x
been carried out. The reaction of solid [Pd(NO )
3 2
Variation of NAPS properties in time. We have
obtained the following experimental data on the NAPS
stability in time. First of all, it should be noted that
freshly prepared NAPSs can contain solutions over-
saturated with respect to Pd(II), which are fairly stable
for a certain time. Depending on the Pd(II) concentra-
tion in solution, they remain homogeneous for a
period from 2 to 24 h, but after that a crystalline phase
of [Pd(NO ) (H O) ] precipitates. We have obtained
(H O) ] ( 0.01 mol) with an excess of concentrated
2
2
aqueous solution of ammonia ( 0.5 mol) at room
temperature or at 50 60 C yields a light brown solu-
tion containing a small amount of a voluminous
brown precipitate. On further heating of the reaction
mixture with the added ammonia, the color becomes
lighter within several hours, and the solid practically
completely precipitates on the bottom. Then the reac-
tion mixture is filtered through a porous glass filter,
and the mother liquor is evaporated to dryness. The
target product [Pd(NH ) ](NO ) represents crystals
3 2
2
2
NAPS containing up to 26.22% Pd(II). If the precipi-
tate is separated from it and the solution is stored
further, crystalline [Pd(NO ) (H O) ] precipitates
3 4
3 2
in the form of colorless or slightly yellowish plates.
The yield reaches 97%. The recrystallization of the
substance from water allows us to obtain a high-purity
product.
3 2
2
2
again from the solution. It is most likely that the crys-
talline precipitates are formed due to limited solubility
of [Pd(NO ) (H O) ] in nitric acid.
3 2
2
2
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

702
VENEDIKTOV et al.
Electronic absorption spectra of Pd(II) nitric acid solutions. Concentration of free HNO3 in solutions 3.7 5.7 M
of maxima, nm, at indicated keeping time, weeks
Solution
Solution*
cPd, M
no.
0
3
4
8
16
1
2
3
4
5
6
7
NAPS-1
NAPS-1
3.042
3.042
2.744
2.070
2.925
3.181
2.706
397(220)**
398(211)
396(286)
419(160)
419(178)
425(173)
428(194)
397(272)
398(244)
398(247)
393(244)
A
B
C
D***
E
398(294)
419(188)
421(172)
421(198)
428(206)
396(244)
421(170)
421(189)
422(180)
427(180)
396(283)
419(165)
419(172)
419(170)
426(181)
* Molar extinction coefficients.
** Solutions A, B, C, D, and E were prepared using the technique of acidity correction.
*** Prepared by dissoluving crystalline [Pd(NO ) (H O) ] in HNO .
3 2
2
2
3
From the experimental data we concluded that
NAPSs are the most stable if the Pd(II) content is 21
22%, or close to 3.0 M. Higher Pd(II) content results
in the precipitation of crystalline products of the
composition [Pd(NO ) (H O) ]. This fact is respons-
position of the maximum of NAPS absorption) by
100 pulses of 20 mJ each, which corresponds to ap-
proximately a week storage under usual laboratory il-
lumination. No difference was observed between the
spectra of the irradiated and initial solutions.
3 2
2
2
ible for the above-mentioned low reproducibility of
The characteristics of the absorption spectra of the
NAPSs under study are given in the table. All the
solutions were kept in the dark, except for solution
no. 2. The solutions designated by letters A, B, C, D,
and E were prepared using the acidity correction. It
should be noted that the contribution of nitric acid
to the absorption in its fairly concentrated solutions
does not give rise to an appreciable fixed error at
wavelengths longer than 350 nm, which is very favor-
able for characterizing EAS of Pd(II) nitric acid solu-
tions. Typical spectra of undiluted Pd(II) nitric acid
solutions are shown in Fig. 4.
the data on the NAPS density in the region of c 3 M
Pd
(Fig. 2). The most important conclusion from our
experiments is the fact that NAPSs are stable in time
if c does not exceed 21 wt % and the molar ratio of
Pd
free HNO and c is no less than 1 : 1.
3
Pd
This conclusion on the Pd(II) and HNO concentra-
3
tions ensuring the stability of NAPS solutions was
supplemented by a series of other measurements. We
have monitored the dynamics of variations of such
solution parameters as palladium(II) concentration,
acidity, and EAS taken in situ without dilution. These
observations were carried out during a period of 3
9 months. The weight fraction of Pd(II) in the solu-
tions studied was in the range 15.0 20.7 wt %, or
It is interesting that the EAS of the solutions are of
two types. The spectra of NAPS-1 and solution A have
a maximum at approximately 400 nm, which coin-
2.07 3.17 M. The content of free HNO was from
3
1
cides in position with the maximum of [Pd(NO )
3 2
172 to 316 g l . All the solutions were prepared ac-
(H O) ], but these spectra are much more intensive.
2
2
cording to the technique described above, but after the
evaporation to the necessary volume the acidity was
corrected by adding certain amounts of concentrated
At the same time, solutions B, C, D, and E are char-
acterized by a maximum at 425 4 nm, and EAS are
apparently superpositions of at least two spectra of
different Pd(II) species in NAPS. Furthermore, the
difference in the molar extinction coefficients of the
spectra under consideration is about 30%. The com-
parison of the spectra of solutions C and E is especial-
HNO to ensure the molar ratio of Pd(II) and free
3
HNO required for the NAPS stability.
3
It was found that the concentration of Pd(II) in
solutions prepared in such a way remains unchanged
in time within 0.1 0.6 rel. % and practically does not
depend on the fact whether these solutions were held
in the dark or in a scattered laboratory light. The latter
conclusion was confirmed by an experiment on laser
flash photolysis, carried out as follows. A solution
was irradiated at a wavelength of 308 nm (the average
ly characteristic, as they noticeably differ in c , but
Pd
have almost the same shape.
It should be noted that the characteristics of EAS
of Pd(II) nitric acid solutions depend, as follows from
the published data [11 13, 17], on the concentration
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

PROPERTIES OF NITRIC ACID PALLADIUM SOLUTIONS
703
HNO of 3.7 5.6 M during their storage both in light
3
and in the dark under various conditions was deter-
mined.
(2) At Pd(II) content higher than 21 wt %, the
crystalline phase [Pd(NO ) (H O) ] precipitates from
3 2
2
2
solutions. It is hygroscopic and unstable to the action
of light on storage. The technique of the preparation
of [Pd(NH ) ](NO ) from this substance with a yield
3 4
3 2
close to quantitative was suggested. Thanks to its
specific properties, the product can be used for prepar-
ing supported palladium catalysts.
(3) The ratio of concentrations of palladium(II)
and free nitric acid is important for the stability of Pd
nitric acid solutions in storage. Solutions with c
about 20 wt % can be stored for indefinitely long time
Pd
if the ratio of concentrations of free HNO and Pd(II)
3
is no less than 1 : 1 (i.e., no less than 3 M HNO , or
3
1
about 200 g l ).
(4) According to the data obtained, the Pd nitric
acid solutions contain a mixture of palladium(II) nitro-
aqua compounds in the presence of free nitric acid.
Nitrite ions or nitrous acid are present in negligible
amounts or absent at all. One of the main conditions
of the safe storage of nitric acid solutions containing
, nm
Fig. 4. EAS of Pd(II) nitric acid solutions. (D) Optical
density and ( ) wavelength. (A E) Solutions (see text).
of free nitric acid, which noticeably varied in different
NAPSs. Solutions A, B, and C have almost equal
acidity, but their spectra essentially differ in the
shape. It is probable that differences in the NAPS
spectra somehow depend on fine details of the solu-
tion preparation and acidity correction.
about 20% Pd and about 3 M free HNO is the exclu-
3
sion of their dilution by water. Otherwise Pd(II) hy-
droxides precipitate from the solutions within several
days.
The most important conclusion consists in the fact
that, when solutions are kept for three months, the
positions and intensities of maxima in EAS remain
constant within the limits of measurement errors,
which first of all depend on the thickness of the ab-
sorbing layer of the cell in use and on specific features
of its filling with solutions. This conclusion is also
proved by the measurements of the solution densities,
which, in spite of appreciable temperature variation
during monitoring, were nevertheless more accurate
and remained constant within 2%.
(5) The comparison of properties of solutions kept
in the light and in the dark shows that they are almost
inert photochemically.
(6) To diagnose properties of nitric acid solutions
in the course of their storage, it is suggested to use
a combination of visual observation and measurement
of solutions density. In this case, using calibration
dependences of the density of nitric acid solutions on
the Pd(II) concentration in them, it is possible to
determine palladium concentration accurate to 3%
without analytical measurements.
Along with recording EAS of solutions, we meas-
ured the acidity of solutions using NAPS-1 as an
example of the most aged solution. This monitoring
was carried out during a period of 10 months. In this
period, the acidity remained constant and was 198 3
ACKNOWLEDGMENTS
The authors are grateful to A.V. Belyaev for useful
discussion.
1
g l HNO (confidence level 95%).
3
REFERENCES
CONCLUSIONS
1. Khimiya platinovykh i tyazhelykh metallov (Chemistry
of Platinum and Heavy Metals), Ser.: Problemy koordi-
natsionnoi khimii (Problems of Coordination Chemis-
try), Shchelokov, R.N., Ed., Moscow: Nauka, 1975.
(1) The nature of solid phases precipitating
from Pd(II) nitric acid solutions with c of up to
Pd
20.67 wt % (or 3.181 M) and the concentration of free
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007

704
VENEDIKTOV et al.
2. Shmidt, V.S., Shorokhov, N.A., Vashman, A.A., and
Samsonov, V.E., Zh. Neorg. Khim, 1982, vol. 27, no. 5,
pp. 1254 1255.
proach), Reading, Massachusetts: Addison Wesley,
1964.
10. Belyaev, A.V., Fedotov, M.A., and Korenev, S.V.,
Koord. Khim., 1989, vol. 15, no. 11, pp. 1551 1555.
3. Laligant, Y., Ferey, G., and Le Ball A., Mater. Res.
Bull., 1991, vol. 26, no. 4, pp. 269 275.
11. Shoittsov Rudi, A Spectrophotometric Study of Pal-
ladium(2+) State in Aqueous Solutions, Cand. Sci. Dis-
sertation, Moscow, 1972.
12. Jorgensen, C.K. and Parthasarathy, V., Acta Chem.
Scand. A, 1978, vol. 32, no. 10, pp. 957 963.
13. Shmidt, V.S., Mezhov, E.A., Rubisov, V.N., et al.,
Radiokhim., 1986, vol. 28, no. 3, pp. 345 349.
14. Izatt, R.M., Eatough, D., and Christensen, J.J.,
4. Gatehouse, B.M., Livingstone, S.E., and Nyholm, R.S.,
J. Chem. Soc., 1957, no. 10, pp. 4222 4225.
5. Torgov, V.G., Tatarchuk, V.V., Druzhinina, I.A., et al.,
At. Energ., 1994, vol. 76, pp. 478 482.
6. Torgov, V.G., Tatarchuk, V.V., Druzhinina, I.A., et al.,
At. Energ., 1996, vol. 80, pp. 267 273.
7. Tatarchuk, V.V., Druzhinina, I.A., Korda, T.M., and
Torgov, V.G., Zh. Neorg. Khim., 2002, vol. 47, no. 12,
pp. 2082 2088.
J. Chem. Soc. A, 1967, pp. 1301 1304.
15. Fedotov, M.A., Troitskii, S.Yu., and Likholobov, V.A.,
Koord. Khim., 1990, vol. 16, no. 12, pp. 1675 1677.
8. Charlot, G., Les methods de la chimie analytique.
Analyse quantitative minerale, Paris: Masson, 1961,
4th ed.
16. Giemser, O. and Peuschell, G., Z. Anorg. Allg. Chem.,
1955, vol. 281, no. 1, pp. 44 47.
17. Shmidt, V.S. and Shorokhov, N.A., At. Energ., 1988,
9. Butler, J.N., Ionic Equilibrium (A Mathematical Ap-
vol. 64, no. 2, pp. 103 110.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 80 No. 5 2007