Journal of Molecular Liquids 194 (2014) 68–76

Contents lists available at ScienceDirect

Journal of Molecular Liquids

journal homepage: www.elsevier.com/locate/molliq

Can pure gold be dissolved in seawater mixed with aqueous nitric acid?
Masashi Hojo a,⁎, Yuki Uji-yie a, Shohei Tsubota a, Mika Tamura a, Masahiko Yamamoto a,
Kei Okamura b, Kenji Isshiki c
a
Department of Chemistry, Faculty of Science, Kochi University, Akebono-cho, Kochi 780-8520, Japan
b
Center for Advanced Marine Core Research, Kochi University, B200 Monobe, Nankoku, Kochi 783-8502, Japan
c
Department of Environmental Science, University of Kochi, Eikokuji-cho, Kochi 780-8515, Japan

a r t i c l e i n f o a b s t r a c t

Article history: In order to demonstrate further the oxidation ability of diluted aqueous nitric acid containing abundant
Received 30 September 2013 amounts of salts, the dissolution of precious metals (Au, Pt, and Pd), especially, of gold has been examined in
Received in revised form 28 December 2013 0.1–2 mol dm−3 HNO3 accompanied by alkali metal, alkaline earth metal, and aluminum chlorides. The complete
Accepted 4 January 2014
dissolution time of pure gold plate (20 ± 2 mg, 0.1 mm thickness) in 2.0 mol dm−3 HNO3 accompanied by
Available online 24 January 2014
1.0 mol dm−3 AlCl3 was shortened remarkably with a temperature increase from 15 to 80 °C. The dissolution
Keywords:
rate constants, log (k/s−1), of a piece of gold wire (19.7 ± 0.5 mg) in 20 mL of 2.0 mol dm−3 HNO3 accompanied
Dilute aqua regia by the metal chlorides, in general, increase with increasing salt concentrations at 40 and 60 °C. For instance, the
Redox potential log (k/s−1) values linearly increase from −4.15 via −3.77, −3.45 to −3.14 by the addition of 1.0., 2.0, 3.0, and
Tetrachloroaurate 4.0 mol dm−3 LiCl in a 2.0 mol dm−3 HNO3 solution at 60 °C. The gold can be dissolved in the solution of
Concentrated salt 1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl, i.e. a “dilute aqua regia.” We have achieved a total dissolution of
Au or Pt electrode natural dissolution five pieces (0.10 g) of the gold wire in 100 mL of the 1:1 mixture between seawater and 2.0 mol dm−3 HNO3
Raman spectrum at ca. 100 °C, with a complete dissolution time of within 17 h and log (k/s−1) = −4.52. At 60 °C, the complete
dissolution experiments of the gold wire in much lower HNO3 concentrations (down to 0.1 mol dm−3) with
the chloride salts have been also successfully performed. The distortion of bulk water and alternation of water
properties are discussed based on the Raman spectra with increasing HCl concentrations.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction in the 19th century, but its use diminished following the introduction
of the cyanide process in 1889. Thiosulfate, thiourea, ammonia, alkaline
It has been widely recognized that every natural element present in sulfide, and other halides (eg., bromide, iodide) solutions have been in-
the earth is dissolved in seawater [1]. Concentrations of elements in vestigated extensively in the laboratory. Putnam [9] has reported that
seawater may depend on the amounts in the earth's crust, the solubility, the addition of chlorides to a chlorine solution will hasten the dissolu-
the mean residence time, and so on. Sonstadt [2] reported the gold tion of gold particles.
concentration in seawater to be 65 μg L−1. After World War I, Haber [3] The redox potentials [10a] for gold species are as follows:
tried to recover gold from seawater in vain. So far, it is believed that the
þ ‐ o 3þ ‐ o
gold content in seawater amounts to 145 ± 45 fmol dm−3 (i.e. 29 ± Au þ e ⇆Au E ¼ 1:83 V and Au þ 3e ⇆Au E ¼ 1:52 V;
9 pg L−1) in open ocean pacific waters [4], or 50–150 fmol dm−3 in the
‐ ‐ ‐ o
Atlantic and Northeast Pacific ocean [5]. AuCl2 þ e ⇆Au þ 2 Cl E ¼ 1:154 V; ð1Þ
In an editorial article [6] it is written that “many people have the im-
‐ ‐ ‐ o
pression that gold occurs as nuggets in streambeds and being a noble AuCl4 þ 3e ⇆Au þ 4 Cl E ¼ 1:002 V: ð2Þ
metal is only dissolved by aqua regia, mixture of concentrated hydro-
chloric and nitric acids.” Beckham et al. [7] described that Geber, an Ara- A comparison between Eqs. (1) and (2) indicates that the Au(III)
bian chemist, mentioned aqua regia in the eighth century. In the complex is more stable than the Au(I) species by 0.15 V. Oxidation
metallurgy of gold, many leaching (dissolution in liquids) methods will occur only above approximately, 1.2 V, and therefore a strong oxi-
were known [8]. Chlorine–chloride leaching was applied commercially dant, such as Cl2 or O3, is required to dissolve gold at a reasonable rate
[11]. The redox potential of Cl2 (g) to Cl− (aq) is reported to be 1.358
V [10b] and that of O3 to O2 to be 2.075 V or 1.246 V in acidic or basic
solution, respectively [10c].
Nitric acid has two different functions: it is a strong acid and a strong
⁎ Corresponding author. Tel.: +81 88 844 8306; fax: +81 88 844 8359. oxidizing agent in dilute and concentrated aqueous solutions, respec-
E-mail address: mhojo@kochi-u.ac.jp (M. Hojo). tively. Cotton and Wilkinson [12] have described that 0.1 mol dm− 3

0167-7322/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molliq.2014.01.014
 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76
aqueous nitric acid is dissociated by about 93% and that the oxidizing
power of nitric acid is very dependent on concentration, and below
about 2 mol dm−3 the acid has virtually no oxidizing power. The disso-
ciation constants of HNO3 at different concentration ranges in aqueous
solution have been determined [13]. The OH stretch IR absorption of ni-
tric acid in CCl4 has provided definitive proof of molecular nitric acid
rather than NO− 3 [14].
We have proposed an innovative idea [15] that the properties of bulk
water originate from the intermolecular hydrogen-bonded network
among a huge number of H2O molecules (e.g., nw N ~ 107). Water can
loose its properties as bulk water to get that of a non-aqueous solvent,
such as an alcohol (R\O\H) or even an ether (R\O\R) if the highly
“self-assembled structure” of bulk water is disturbed in the following
causes: (a) the residual water (c(H2O) ~10−3 mol dm−3) in organic sol-
vents; (b) aqueous solutions or organic solvent–water mixtures con-
taining salts at high concentrations; (c) nanoscale water droplets in
nano-tubes or reversed micellar systems; (d) water on metal electrodes,
ion-exchange resins, proteins, organic solvents in the solvent extrac-
tions, and glass vessels; (e) water at higher temperatures or under
supercritical conditions. Such water can be “reduced” to authentic sin-
gular H2O molecules (also called “dihydrogen ether,” (H)\O\(H))
[16]. Reichardt et al. [17] have concisely interpreted “dihydrogen
ether” that, at high salt concentrations [c(salt) N 5 mol dm−3], region
C, according to the solvation model of Frank and Wen [18], can be
abolished and only regions A and B survive, resulting in an aqueous sol-
vent called “dihydrogen ether.”
Our two ideas are combined by chance: (1) the ion association can
take place between simple [15,19] or poly-anions [20] and alkali metal
or alkaline earth metal cations, (2) the bulk water may lose its proper-
ties completely by the addition of salts and/or organic solvents at high
concentrations. Then, for the first time, we have been able to explain
successfully the concentrated salt effects on solvolysis reactions of or-
ganic haloalkanes and related compounds without resorting to different
types of ion pairs [16]. The exponential increases in solvolysis rates of
SN1 substrates in the presence of concentrated alkali metal (M+) and al-
kaline earth metal (M2+) perchlorates are attributed to the favorable
formation of carbocations (R+) through association (direct “chemical”
interaction) between metal cations (M+ or M2 +) and the leaving
group anion (X−) of a substrate (R–X) in the modified solvent.
Previously [21], we have managed to demonstrate that the Br− ion
of the surfactant, CTAB, is oxidized to Br2 (or Br− 3 ) in the CHCl3/CTAB/
H2O reversed micelle system with diluted nitric acid (0.25–
2.5 mol dm−3 in the 1.0% (v/v) H2O phase) at 15–40 °C (CTAB stands
for cetyltrimethylammonium bromide). We have proposed that the
NO+ 2 ion should be the active species in this oxidation process. In a fol-
lowing study [22] the nitration of phenols has been performed by aque-
ous nitric acid (2.0 mol dm−3) in reversed micelle systems.
Our most important finding in the current work [22] is that nitric
acid of below 2 mol dm−3 in bulk water, provided it contains salts in
high concentrations, can oxidize the Cl− ion to Cl2 as well as Br− to
Br2 at 20–40 °C. Thus, we recognize that the nitronium ion (NO+ 2 ), the
active species for nitration or oxidation, can be generated not only in
the water phase of reversed micellar systems but also in bulk water con-
taining salts in higher concentrations. Eq. (3) is the common reaction
scheme to produce NO+ 2 in concentrate nitric acid and we have assumed
that the same scheme can be applied to even the diluted nitric acid in
reversed-micellar water droplets [21,22] as well as in bulk water con-
taining higher concentrations of salts [22].
þ −
2 HNO3 ⇆NO2 þ NO3 þ H2 O ð3Þ
With an intimate examination of Eq. (3), we may notice that en-
hanced formation of HNO3 in its molecular state and not dissociated
(H+ + NO− +
3 ) is essential for a favorite NO2 generation. In bulk aqueous
solution, however, many people may think that nitric acid diluted
enough is apt to dissociate completely. How could we keep the nitric
69
acid molecules from dissociating to protons and nitrate ions? Reducing
the solvent's permittivity (e.g. εr b 10) may be an excellent way for that
purpose. However, the permittivity of bulk aqueous solutions cannot be
reduced so well. Then, our idea, the alternation of H2O into “dihydrogen
ether” [16], could resolve some of the problems. Both the “strong” acid-
ity and basicity of water are lost at once when the water molecules in
the huge network are broken into isolated molecules by some of the
causes already mentioned, such as, the abundant addition of salts and
increased temperature. For solutions with very high salt concentrations,
we may not rely so much on the Debye–Hückel [23] and Pitzer [24] the-
ories, even though being highly developed, because such new or ex-
traordinary phenomena have not been predicted in advance by
calculations making use of these theories, at least at the moment. We
have already mentioned that the redox potential of NO+ 2 generated in
diluted nitric acid should be high enough to oxidize Cl− into Cl2 [22].
The present adventure, that is, dissolving gold in aqueous nitric acid,
has been initiated not by any coincidence but by our reasoned deduc-
tion. As described above, dilute nitric acid in bulk water, provided it
contains salts in high concentrations, can oxidize the Cl− ion to Cl2.
Therefore, it is an inevitable theory that the evolved Cl2 should oxidize
Au0 into Au3 +, which is then coordinated by the abundant Cl− ions,
when the aqueous nitric acid is accompanied by sufficient amounts of
chloride salts.
In the present paper, we examine the dissolution of precious metals
(Au, Pt, and Pd), and give detailed reports on gold dissolution in aque-
ous nitric acid (b 2 mol dm−3) containing alkali metal, alkaline earth
metal, and aluminum chlorides at 15–100 °C. The periods for complete
gold dissolution as well as the corresponding rate constants are mea-
sured, and the accelerating influence of cations on the rate constants
will be related to the crystal ion radii of the metal ions. Examined and
discussed are the dissolution ability of “dilute aqua regia” (e.g., an aque-
ous 1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl solution) and the accel-
eration and deceleration phenomena of the dissolution rate caused by
the addition of salts. At last, pieces of gold wire are dissolved into a mix-
ture between seawater with aqueous nitric acid. The distortion of the
bulk water structure is discussed based on IR Raman spectra measured
at increasing HCl concentrations.
2. Experimental
2.1. Materials
The samples of gold were all commercially available: gold plate (Al-
drich, 99.99%, 0.1 mm thick) and gold wire (Nilaco, 99.95%, 0.1, 0.25,
and 0.5 mm diameter). Both Pt and Pd were also obtained commercially,
Pt (Nilaco, 99.98%, 0.10 mm diameter) and Pd wire (Aldrich, 99.9%,
0.25 mm diameter). Powdered samples of Ir and Ru were complimenta-
rily given by Tanaka Precious Metals. Commercially available salts were
used as received: NaClO4, Mg(ClO4)2, and MgCl2 from Aldrich; LiClO4,
LiCl, NaCl, NaBr, KCl, CaCl2, AlCl3, LiNO3, NaNO3, Li2SO4, Na2SO4 (of all
GR grades) from Wako Pure Chemical Industries. NaAuCl4 2H2O was ob-
tained from Aldrich. Nitric acid of 60% of a GR grade (JIS S), HCl
(5.0 mol dm−3), and acetonitrile of a GR grade were purchased from
Wako. Pure water purified by a MilliQ System was used. The seawater
of “Muroto Deep Seawater” (sampled from a 350 m depth) off Muroto,
Kochi, Japan, was utilized as a standard seawater. The other seawater
sampled at a 1000 m depth in the Pacific Ocean off Hawaii, USA,
which was filtered in a common procedure, was also used.
2.2. Procedure and equipments
As the first stage, we measured the periods for the complete dissolu-
tion of gold plate at various temperatures. In a 20 mL volumetric flask, a
piece of gold plate (Aldrich, 99.99%) of 20 ± 2 mg (0.1 mm × 2 mm ×
5 mm) was placed into 2.0 mol dm−3 nitric acid, which is accompanied
by 1.0 mol dm−3 AlCl3. At the second stage, the gold plate was displaced
70 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76
with the gold wire (Nilaco 99.95%, 0.25 mm diameter), and the total dis-
solution periods as well as kinetics (monitoring the changes in the gold
mass and the absorbance at 375 nm of AuCl− 4 ) were measured in the
presence of various salts.
With such a procedure, the actual concentration of HNO3
(2.0 mol dm−3) should be slightly reduced by the presence of a large
amount of a salt, although we have a merit to be able to avoid the direct
contact between the salt crystals and concentrated or more than
2.0 mol dm−3 HNO3. Therefore, at the third stage, a 10.0 mol dm− 3
HNO3 solution was utilized to make solutions of precise HNO3 concen-
trations, such as 0.5, 1.0, or 2.0 mol dm−3, containing various salts.
A gold sample of 19.7 ± 0.5 mg (i.e. 5.0 × 10− 3 mol dm− 3) in a
20 mL volumetric flask was kept at a constant temperature of 15 to 80
(±0.1) °C in a Yamato Shaking Incubator model BT100, being shaken
60 times in a minute, equipped with a cooling dip (Yamato model BE-
200F). In another experiment, a gold sample of 0.10 g was placed with
100 mL nitric acid solution into a two-neck 300 mL flask, equipped
with a condenser, when the solution temperature approached its boil-
ing point (97–103 °C), which depends on the concentrations of nitric
acid and the salt.
UV–visible absorption spectra were measured by using a Shimadzu
spectrophotometer (model UV-2550) in a 0.1 cm (unless otherwise
noted) quartz cuvette at room temperature. The metal components dis-
solved in solution were determined by a PerkinElmer ICP emission
spectrometer (model Optima 4300DV). Raman spectra of H2O in the
presence of HCl were recorded with a Horiba-Jobin Yvon LabRam Spec-
trograph HR-800 at room temperature (~ 23 °C). Excitation for the
Raman spectroscopy was provided by a 514.5 nm argon laser.
3. Results and discussion
3.1. Total dissolution of pure gold in 2 mol dm−3 aqueous HNO3 solution
The gold dissolution has been examined with gold plate. Fig. 1 shows
the period (time in hour) needed for the complete dissolution of a piece
of gold plate (20 ± 2 mg, the thickness of 0.1 mm) in a 20 mL solution of
2.0 mol dm−3 HNO3, accompanied by 1.0 mol dm−3 AlCl3, at various
temperatures. It takes a long time, ca. 35 h at 15 °C, however, the disso-
lution period was shortened with increasing temperature, e.g. less than
30 min at 70 or 80 °C. The aqueous HNO3 solution acquires the in-
creased dissolution ability with the temperature increase.
Next, the gold dissolution into the aqueous HNO3 solution has been
examined with the gold wire instead of the gold plate. Fig. 2 shows the
time for complete dissolution of a piece of gold wire (19.7 ± 0.5 mg, the
diameter of 0.25 mm) in a 20 mL solution of 2.0 mol dm−3 HNO3, ac-
companied by various chloride salts, at 60 °C. With increasing salt con-
centrations (from 1.0 mol dm−3), generally speaking, the complete
dissolution time is shortened. However, an interesting reversal trend
is observed for more than 5.0 mol dm−3 LiCl: the gold dissolution in a
2 mol dm−3 HNO3 solution needs longer time in the presence of 7.0
Fig. 1. Time for the complete dissolution of gold plate (20 ± 2 mg) in 2.0 mol dm−3 HNO3,
accompanied by 1.0 mol dm−3 AlCl3 at various temperatures.
Fig. 2. Time for the complete dissolution of gold wire (19.7 ± 0.5 mg) in 2.0 mol dm−3
HNO3, accompanied by various chloride salts at 60 °C.
or 9.0 than 5.0 mol dm−3 LiCl. The most remarkable reversal effect is ex-
hibited by CaCl2. The dissolution time reaches its minimum at around
2.5 mol dm−3 CaCl2 and get longer again after reaching a concentration
of 3.0 mol of the salt.
As to dissolving gold, MgCl2 is more effective than CaCl2. The influ-
ence of metal chlorides (at 1.0 mol dm−3) for the rate of gold dissolu-
tion increases in the order KCl b NaCl b LiCl b CaCl2 b MgCl2 b AlCl3. At
a lower temperature, 40 °C, much longer time is needed for complete
dissolution of the gold wire in a 20 mL solution of 2.0 mol dm−3
HNO3 than at 60 °C. For instance, it takes 13 h for the solution with
3.0 mol dm−3 NaCl at 40 °C, but only 2 h at 60 °C. In any case, all the re-
sults obtained with the metal chlorides at 40 °C were very similar to
those observed at 60 °C. We would like to mention that the complete
dissolution experiments of the gold wire in the solutions of much
lower HNO3 (down to 0.1 mol dm−3) concentrations containing the
chloride salts have been also successfully performed at 60 °C.
At 80 °C, the dissolution periods have been examined for dilute salts
(0.l\1.0 mol dm−3) in 2.0 mol dm−3 HNO3 (Fig. 3). A piece of gold wire
is completely dissolved after a long time, ca. 150, 50, and 24 h, with
0.1 mol dm−3 NaCl, CaCl2, and AlCl3, respectively. No differences be-
tween LiCl and NaCl, or MgCl2 and CaCl2 have been observed for low
salt concentrations [c(salt) b 1.0 mol dm−3]. That is, the chloride salt in-
fluence depends upon just the number of the cation charges according
to MCl b MCl2 b MCl3. The dissolution rate constants, log (k/s−1), have
also demonstrated the consistency (for the same charges) as [− 4.15
and − 4.71], [− 3.72 and − 3.79] for 0.5 mol dm− 3 alkali metals [LiCl
and NaCl], and alkaline earth metal ions [MgCl2 and CaCl2], respectively.
Aluminum chloride at the same concentration causes a much faster rate
of −3.33. For higher salt concentrations [c(salt) N 1.0 mol dm−3], how-
ever, the differences between LiCl, NaCl, and KCl, or between MgCl2 and
CaCl2 have been observed.
The dissolution capacity for Au has been examined as follows:
a 2 mL solution of 2.0 mol dm− 3 HNO3 is poured to 0.57 g of MgCl2
Fig. 3. Logarithm of time for the complete dissolution of gold wire (19.7 ± 0.5 mg) in
2.0 mol dm−3 HNO3, accompanied by chloride salts of lower concentrations at 80 °C.
 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76 71

(ca. 3 mol dm−3) in a small beaker, and gold wire of 0.10 g is put in the Table 1
solution. The beaker is placed on a hot-plate at ca. 80 °C, the gold wire The gold dissolution rate constant, log (k/s−1), measured in 2.0 mol dm−3 HNO3,
accompanied by various added chloride salts at 40 °C.
dissolved completely in 40 min: the capacity can be estimated to be
50 g Au/L of 2 mol dm−3 HNO3, accompanied by 3 mol dm−3 MgCl2. c(salt)/mol dm−3 LiCl NaCl MgCl2 CaCl2 AlCl3
On the other hand, it takes several days for a 0.040 g Pt wire (0.1 mm di- 0.5 −4.67a
ameter) to be dissolved in 10 mL of 2 mol dm−3 HNO3, accompanied by −4.63b
3 mol dm−3 MgCl2 at 80 °C, giving a Pt capacity of at least 4 g Pt/L. We 1.0 −4.49 −4.59 −4.00
−4.48 −4.57 −3.97
will also note that Pd can be dissolved in a 1:1 mixture between
1.5 −4.09 −3.61
2.0 mol dm−3 HNO3 and seawater (vide infra). −4.07 −3.60
For other precious metals, Ir and Ru, dissolution experiments have 2.0 −4.39 −4.62 −3.82 −4.09
been performed. However, these metals (in powder) are not dissolved −4.37 −4.69 −4.06
in 2 mol dm− 3 HNO3, accompanied by 3 mol dm− 3 MgCl2. It is well 3.0 −4.10 −4.29 −3.51 −3.99
−4.03 −4.27
known that both Ir and Ru are not dissolved even in proper aqua regia
3.5 −4.21 −3.54 −4.26
[25]. Our method for dissolving precious metals (such as Au, Pt, or Pd) −4.27
has a similar but weaker function than normal aqua regia. However, 4.0 −3.84
we would like to stress that our method has a higher efficiency of −3.84
5.0 −3.59
chemicals, without useless evolution of NOx and Cl2 during gold dissolu-
6.0 −3.62
tion in regular aqua regia.
a
The (k/s−1) values in the upper rows are evaluated from the loss of the gold mass, ln
(Mt/M0) vs. t/s where M0 and Mt are the gold masses at the initial and a certain time,
3.2. Dissolution rate constants and cation effects respectively.
b
The (k/s−1) values in the lower rows are evaluated from the absorbance values at
The UV–visible spectroscopy has been employed to identify the 375 nm, ln (A∞ − At) vs. t/s where A∞ and At are absorbance values after the complete
dissolved species formed from the gold wire. Fig. 4 shows the gold dissolution and at a certain time, respectively.
changes with time in the absorption spectra (in 0.1 cm path-
length) of a 2.0 mol dm − 3 HNO 3 solution containing a piece of of the chloride ion are quite high; in such cases, the solution turns al-
gold wire (ca. 19.7 mg) in the presence of 3.5 mol dm − 3 NaCl. A ready to a yellow color before placing gold. For its simplicity, we may
band appears and increases with time, which is accompanied by a take the rate constants obtained by the loss of the gold mass, except
peak at λ ~ 306 nm and a shoulder around 400 nm, while the initial for some cases in which the mass method could not be appropriately ap-
HNO3 gives a band at λ ~ 300 nm. The spectrum observed after 12 h in plied. In the present paper, the rate constants evaluated by the mass
a 0.05 cm cuvette is consistent to that of 5.0 × 10−3 mol dm−3 NaAuCl4 method are given, unless otherwise noted.
dissolved in a 2.0 mol dm−3 HNO3 solution. Therefore, we can safely Table 2 shows the effects of the mass and the diameter of gold wire
conclude that the species dissolved in the 2 mol dm− 3 HNO3 and on the dissolution rate constant in the solution of 1.0 mol dm−3 HNO3
3.5 mol dm− 3 NaCl solution is definitely the AuCl− 4 ion. Jones et al. and 2.0 mol dm−3 NaCl at 60 °C. The increasing mass or length of the
[26] have reported the single crystal X-ray structure of Mg[Au(OH)4]2 gold wire of 0.25 mm (diameter) causes just slightly the acceleration
and Ca[Au(OH)4]2, which have been obtained from reactions between in the log (k/s− 1) value, while those of 0.1 and 0.50 mm diameter
HAuCl4 and Ca(OH)2. remain constant. However, the increasing diameter (or decreasing
The dissolution reaction rate constants (the first-order reaction) surface area) of gold wire causes largely the deceleration of gold
have been evaluated with the changes in the solid gold mass as well dissolution. The total dissolution time prolongs as ca. 18, 24, and 30 h
as the absorbance at λ ~ 375 nm, for avoiding the rather large HNO3 ab- [log (k/s−1) = −4.12,−4.58, and −4.90] for 20 mg gold wire of 0.10,
sorbance (at λ ~ 300 nm). We have adopted the linear part of the plots 0.25, and 0.50 mm diameter, respectively. When the k/s−1 values are
of ln [S] vs. t, since the plots tend to deviate from the linearity down- divided by the A/V value, the log (k/s− 1) values can be converted to
ward (i.e., reaction accelerated) after the half-life period. It should be log [k/(s−1 m)] = − 8.72,− 8.79, and − 8.82, respectively, where A
mentioned that the rate constant is for the initial dissolution process and V are the surface area (m2) and the volume (m3) of a piece of
and that effects of the surface area on the reaction rate can be mini- gold wire. All the log [k/(s−1 m)] values in Table 2 are within −8.70 ~
mized by taking the values obtained with the initial part of each − 8.82. For avoiding the effect of difference in the initial surface area,
experiment. only the gold wire of 0.25 mm diameter is utilized in the following
In general, rate constants obtained by the different two ways are examination.
consistent to each other (cf. Table 1). However, the results from the ab- By the way, the yellow color before putting gold into the solution
sorbance method can be fluctuated, especially, when the concentrations comes from dilute NOCl formed through the reaction between HNO3
and the chloride ion. Beckham et al. [7] noted that, when nitric acid is
mixed with solid sodium chloride, there are formed sodium nitrate,

Table 2
Changes in the gold dissolution rate constant, log (k/s−1), with the mass (or length) and
the diameter (or surface area) of gold wire, measured in a 1.0 mol dm−3 HNO3
(prepared from 4.0 mol dm−3 HNO3) and 2.0 mol dm−3 NaCl solution (20 mL) at 60 °C.

Mass of gold wire/mg

Diameter/mm 10 20 40 60

0.10 log (k/s−1)a −4.10 −4.12

(Wire length) (7.0 cm) (14 cm)
0.25 log (k/s−1)a −4.60 −4.58 −4.54 −4.52
(Wire length) (1.1 cm) (2.2 cm) (4.4 cm) (6.6 cm)
0.50 log (k/s−1)a −4.90 −4.87 −4.88
(Wire length) (0.55 cm) (1.1 cm) (1.6 cm)

Fig. 4. Changes in the UV–vis absorption spectra with time of AuCl−

4 in 2.0 mol dm
−3
HNO3 a
The (k/s−1) values are evaluated from the loss of the gold mass, ln (Mt/M0) vs. t/s
containing 3.5 mol dm−3 NaCl at 40 °C. where M0 and Mt are the gold masses at the initial and a certain time, respectively.
72 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76
chlorine, and nitrosyl chloride: 3 NaCl + 4 HNO3 → 3NaNO3 + Cl2 +
NOCl + 2 H2O. They mentioned that “the system is similar to aqua
regia.” Concentrated NOCl in solution, such as aqua regia, should give a
more dark orange or reddish color. The UV–visible absorption cross-
sections of NOCl have been reported [27].
Fig. 5 shows the relationship between the dissolution rate constants
[log (k/s−1)] and salt concentrations in 2.0 mol dm−3 HNO3 at 60 °C.
The dissolution rate constants, log (k/s − 1 ), of a piece of Au wire
(19.7 ± 0.5 mg) in 20 mL of 2.0 mol dm − 3 HNO3 accompanied by
the metal chlorides, in general, increase with increasing salt concen-
trations. For instance, the log (k/s− 1) values linearly increase to give
− 4.15,− 3.77,− 3.45, and − 3.14 at 1.0. 2.0, 3.0, and 4.0 mol dm− 3
LiCl in 2.0 mol dm− 3 HNO3. However, the values begin to turn re-
versely to give − 3.13 and − 3.17 at 5.0 and 6.0 mol dm− 3.
Thus, we can state that the gold dissolution rate constant increases
with increasing concentration of each alkali metal, alkaline earth
metal, and aluminum chlorides. Very curiously, however, the rate con-
stant decreases after reaching its maximum value for LiCl, MaCl2, and
CaCl2. Similar maxima are observed somehow for the same chloride
salts at 40 °C, as listed in Table 1.
The role of metal cations on the gold dissolution in aqueous HNO3
has been considered. Fig. 6 shows the relationship between log
(k/s − 1 ) at 60 °C and the crystal ionic radius [28] per charge (r/z)
of the metal ion. The smaller the crystal ionic radius per charge (r/z)
of a metal ion, the faster the gold dissolution proceeds. The hydration
around the lithium ion should be stronger than that around the sodium
ion. The difference in the strength of hydration around metal ions may
cause the difference in the gold dissolution rates.
It is quite necessary to consider the cause of the decreases in log
(k/s− 1) values after the maxima. Based on the solution preparation
method, the actual concentration of 2.0 mol dm− 3 HNO3 should be
altered by increasing the salt concentrations. Indeed, we have
found that the actual concentration of 2.0 mol dm − 3 HNO 3 is de-
creased to be 1.88, 1.80, and 1.67 mol dm− 3, respectively, when ac-
companied by 2.0, 3.0, and 3.5 mol dm− 3 CaCl2. We wonder whether
the actual decrease in HNO3 concentration can account for the ap-
pearance of the maxima in the curves of log (k/s − 1 ) vs. c(salt), as
shown in Table 1 and Fig. 5 for 40 and 60 °C, respectively.
In order to clarify this problem, another procedure has been also
adopted for HNO3 preparation: a certain amount of a salt is dissolved
in an appropriate amount of pure water in a 20 mL volumetric flask,
to which a 4.0 mL solution of 10.0 mol HNO3 is added to prepare a
2.0 mol dm− 3 HNO 3 solution. Fig. 7 shows the comparison in log
(k/s− 1) values between two preparation methods of HNO3 solutions
containing CaCl 2 . Even the proper HNO 3 concentration (exactly
2.0 mol dm−3 HNO3) solutions have given the maxima in log (k/s−1)
at around 2.5 mol dm−3 CaCl2 at 40 °C as well as 60 °C. Therefore, the ap-
pearance of the maxima cannot be accounted for (only) by the actual de-
crease in the HNO3 concentration from 2.0 mol dm−3 with increasing
Fig. 5. Relationship between dissolution rate constants and salt concentrations in
2.0 mol dm−3 HNO3 at 60 °C.
Fig. 6. The relationship between the log (k/s−1) measured at 60 °C and the crystal ion radii
per charge (r/z) of metal ions.
amounts of salts. We will represent the reasonable comprehension for
these reversal phenomena in a later section.
Fig. 8 shows the Arrhenius plots of gold wire dissolution rate
constants in 2.0 mol dm− 3 HNO3 containing 2.0 mol dm− 3 NaCl or
1.0 mol dm−3 CaCl2. Both solutions give good linearity from 30 °C to
60 or 70 °C, and activation energies are given to be Ea = 86.4 and
89.2 kJ mo− 1 for NaCl and CaCl2, respectively. These large values
(just) suggest that the gold dissolution reaction in aqueous HNO3 de-
pends on the normal temperature effects.
At ca. 100 °C, the gold dissolution in an exact 1.0 mol dm−3 HNO3
solution containing various chloride salts gives the order of KCl b NaCl
b LiCl b CaCl2 b MgCl2 b AlCl3, with the log (k/s− 1) values of − 3.95,
−3.77,−3.71,−3.50, and −3.34 at 1.0 mol dm−3 salts.
3.3. Acceleration and declaration by salts added to “dilute aqua regia”
We can prepare “dilute aqua regia” (e.g., a 1.0 mol dm−3 HNO3 and
1.0 mol dm− 3 HCl solution) of 20 mL, making use of 10 mol dm−3
HNO3 (2.0 mL) and 5.0 mol dm−3 HCl (4.0 mL). The ability of the “dilute
aqua regia” has been examined for the gold wire dissolution at 60 °C. A
piece of gold wire (ca. 19.7 mg) is completely dissolved in the “dilute
aqua regia” within 35 h to give a log (k/s−1) value of −4.80.
Now, the influence of various chloride salts on the dissolution ability
of “dilute aqua regia” has been examined at 60 °C, and an acceleration
quantified by the log (k/s−1) values is observed, as shown in Fig. 9(a).
In principle, the chloride salt effects on the “dilute aqua regia” (of
1.0 mol dm−3 HNO3 and 1.0 mol dm− 3 HCl) increase in the order of
KCl b NaCl b LiCl b CaCl2 b MgCl2 b AlCl3 at higher salt concentrations
such as 1.0 mol dm−3, whereas the order is KCl ~ NaCl ~ LiCl b CaCl2 ~
MgCl2 b AlCl3 at lower salt concentrations, as shown typically at
Fig. 7. Relationship between the dissolution rate constants and CaCl2 concentrations in
2.0 mol dm−3 HNO3 of 20 mL solution at 40 and 60 °C; (●) CaCl2 in a (10 mol dm−3
HNO3 4 mL + H2O) solution at 40 °C; (▲) CaCl2 dissolved in 2.0 mol dm−3 HNO3 at
40 °C; (○) CaCl2 in a (10 mol dm−3 HNO3 4 mL + H2O) solution at 60 °C; (△) CaCl2 dis-
solved in 2.0 mol dm−3 HNO3 at 60 °C.
 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76 73

the dissolution rate of pure gold even in the “dilute aqua regia” (of
1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl) at 60 °C. The log (k/s−1)
value just slightly changes from−4.80 to−4.71 and then to −4.79 by
the addition of 0.5 and 1.0 mol dm−3 Li2SO4, respectively, to the “dilute
aqua regia”. However, sodium sulfate causes a distinct deceleration in
the log (k/s−1) value: −5.03 in the presence of 0.5 mol dm−3 Na2SO4.
Further addition of Na2SO4, e.g., of 0.7 mol dm−3, seems to cause further
obstruction against gold-dissolving with the time progress, therefore,
the precise rate constant cannot be measured.

3.4. Anion effects on limited acceleration

However, it is an easy task for us to clarify the reason why the log
Fig. 8. Arrhenius plots for gold wire dissolution rate constants in 2.0 mol dm−3 HNO3 con-
taining 2.0 mol dm−3 NaCl (○) and 1.0 mol dm−3 CaCl2 (●). (k/s− 1) values are diminished by the addition of Na2SO4 because we
have experienced similar depression phenomena in the previous work
[22]. We had observed the generation of Br2 from a 2.0 mol dm−3
0.2 mol dm−3. The similar chloride salt effects depend upon the number
HNO3 solution, accompanied by 3.5 mol dm−3 lithium or sodium salts
of the cation charges according to MCl b MCl2 b MCl3 as has been ob-
(and 0.010 mol dm−3 LiBr) at 20–35 °C. In the course of these experi-
served in the gold wire dissolution in 2 mol dm−3 HNO3 containing rel-
ments, the anion effects of the added sulfate and sulfonate salts are in
atively low chloride salt concentrations at 80 °C (cf. Fig. 3). − − − −
the order: SO2−4 ≪ CH3SO3 (b ClO4 ) b CF3SO3 b HSO4 with the counter
Fig. 9(b) shows the influence of various non-chloride salts on the
cations Li+ or Na+. Actually, lithium and sodium sulfates up to satura-
ability of “dilute aqua regia” at 60 °C. Magnesium perchlorate has a dis-
tion (bca. 1.5 mol dm− 3) in 2.0 mol dm− 3 HNO3 solutions have
tinct acceleration effect, the log (k/s−1) value linearly increases to give
completely suppressed the Br2 generation. The weaker positive effects
−4.80,−4.42, and−4.02 for 0.0, 0.5, and 1.0 mol dm−3 Mg(ClO4)2, re-
with CH3SO− 2−
3 as well as the complete obstruction with SO4 for the ox-
spectively. Compared to Mg(ClO4)2, alkali metal perchlorates have
idation ability of the nitric acid solution can be attributed to the (rela-
weaker acceleration effects: log (k/s− 1) = − 4.40 and − 4.42 for
tively “large”) basicities of these anions. It is reported that pKa =
1.0 mol dm−3 LiClO4 and NaClO4, respectively. Nitrate salts cause stron-
− 3.91 ± 0.50 [29] 1.75 ± 0.50 [29], and 1.92 [30] for CF3SO3H,
ger effects than perchlorate salts. That is, the acceleration effects of the
CH3SO3H, and HSO− 4 , respectively, at ambient aqueous conditions.
anions seem to increase in the order of ClO− − −
4 b NO3 b Cl .
The sulfate ions are able to pull out protons from HNO3 in solution:
Needless to say, chloride ions can assist strongly the formation of the
HNO3 + SO2− ⇆ NO− 3 + HSO4 .
−
chloro-complex AuCl− 4 to cause the formal redox potential of Au
3+
/Au 4
Therefore, in the present study, we can safely conclude that the pro-
to be lower (Eo = 1.002 V) [10a]. However, the perchlorate ion may
ton capture from HNO3 should cause the depression effects expressed
not exhibit any specific chemical reaction in the “dilute aqua regia” be-
by Li2SO4 [no apparent increase in log (k/s− 1) with increasing salt
sides increasing the strength and/or further distortion of the bulk
amounts] and Na2SO4 (distinct negative changes) in Fig. 9(b). In the
water structure. On the other hand, the presence of nitrate salts in a (di-
course of the previous work [22], the salt of CF3COO−(Li+) had been ef-
luted) HNO3 and HCl solution may produce more HNO3 species (which
fective for the Br2 generation but never CH3COO−(Na+) had been. Re-
is the source of NO+ 2 ), through the anion exchange between MNO3 and
gardless of the cations, we can attribute the difference in these
HCl without a loss of the total amount of chloride anions.
phenomena to the difference in the basicities of the conjugate anions
In the previous paper [22] we have reported an opposite anion
from acids: pKa = 0.3 [31] and 4.75 [32] for CF3COOH and CH3COOH.
effects of NO− −
3 b ClO4 for the oxidizing ability of 2.0 mol dm
−3
HNO3,
accompanied by 3.5 mol dm−3 LiCl, that is, the nitrate gave a weaker ef-
fect than the perchlorate. This lesser effect of NO−
3 should have been ex- 3.5. Discussions on the reversal phenomena in log (k/s−1)
−
plained by Le Chatelier's principle in Eq. (3): 2 HNO3 ⇆ NO+ 2 + NO3 +
H2O. With the addition of alkali metal nitrates, the unfavorable forma- Now, we will consider the reversal phenomena of rate constants
tion of NO+ 2 may be caused. Contrastingly, in the present system, the ini- with higher concentrations of chloride salts at 60 °C, as shown in
tial presence of HCl (as the “dilute aqua regia”) can cause a favorable Fig. 5. The rate constant decreases after reaching its maximum value
formation of additional HNO3 by the addition of LiNO3 or NaNO3. for LiCl, MgCl2, and CaCl2. As described above, we have already demon-
The most interesting feature in Fig. 9(b) is the influence of added strated at 40 and 60 °C, (cf. Fig. 7) that suppression after maxima is not
SO24 − salts. The addition of sulfate salts causes a deceleration in caused only by the actual decreases in HNO3 concentrations (from

Fig. 9. (a). Changes in dissolution rate constants with added chloride salt concentrations in “dilute aqua regia” (of a 1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl solution) at 60 °C. (b).
Changes in dissolution rate constants and added non-chloride salt concentrations in “dilute aqua regia” (of a 1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl solution) at 60 °C.
74 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76

2.0 mol dm−3) with the addition of increasing amount of salts, which is
prepared in 20 mL volumetric flasks.
Therefore, another cause must be responsible for the appearance of
the maxima in a relationship between the log (k/s− 1) value and
c(salt), the chloride salt concentration. To tell the truth, this problem
has been puzzling us since the beginning of the present study. However,
we have thought of another cause for decreasing the effective HNO3
concentration, which should generate NO+ 2 unfavorably, now that
even the weak Brønsted bases, such as SO24 −, have been found to de-
crease the effective HNO3 concentration. The ion association between
Ca2 + and NO− 3 [33–35] at the higher ionic strength should decrease
the effective HNO3 concentrations. The ion association between Mg2+
and NO− 3 [36,37] under similar conditions may also give similar results.
Fig. 10. The increase of the dissolution rate constant of gold wire with increasing added
NaCl concentration in the 1.0 mol dm−3 HNO3 and seawater (a half-concentration) solu-
3.6. Gold dissolution into seawater and the seawater assisted by added tion at ca. 100 °C. The rate constants have been evaluated by means of spectroscopy.

chloride salts
Seawater is an electrolyte solution. Its chemistry is dominated by the
presence of six ions (Na+, K+, Mg2+, Ca2+, Cl−, SO2− 4 ) which constitute
more than 99.5% of the dissolved constituents [38]. The concentrations
of the main components have been reported to be 0.468, 0.010, 0.053,
0.545, 0.028 mol kg−1, respectively, for a 35‰ seawater [39]. Taking
into account its density (d = 1.025 as an average) [40], we can regard
the seawater as a 0.55–0.56 mol dm−3 chloride solution with mainly so-
dium and partly magnesium cations.
All experimental results in the present work strongly suggest that
pure gold should dissolve in seawater if it is mixed with aqueous nitric
acid and then heated. We have utilized “Muroto Deep Seawater” as a
standard sample of seawater. The main components in the seawater
are well consistent with the typical 35‰ seawater. By mixing 50 mL sea-
water with 50 mL of 2.0 mol dm−3 HNO3, we have a 100 mL solution of
1.0 mol dm−3 HNO3 containing 0.278 mol dm−3 Cl−, 0.225 mol dm−3
Na+, 0.026 mol dm−3 Mg2+ (0.014 mol dm−3 SO2− 4 ) and other ions.
The mixed solution is heated up to ca. 100 °C in a flask, equipped with
a condenser. When boiling starts, then five pieces of gold wire (totally
0.10 g) are placed into the solution. Occasionally, samplings are carried
out from the boiling solution and the pieces of gold wire are dissolved
completely within 17 h. The absorbance values at λ = 375 nm are uti-
lized for evaluating the dissolution rate constant to give log (k/s−1) =
− 4.52. Another Pacific seawater off Hawaii (1000 m depth) gives a
value (− 4.54) similar to that of “Muroto Deep Seawater”. Although
ten pieces of gold wire (totally ca. 0.20 g) have appeared to dissolve
completely in the 100 mL (seawater and nitric acid mixture) solution,
after cooling down, some residue is noticed on the solution surface.
Therefore, we keep five pieces (0.10 g) of gold wire as the limit for the
seawater experiments. Incidentally, platinum is too tough to be dis-
solved in the seawater and 2.0 mol dm− 3 HNO3 (1:1) mixture even
after 10 days at ca. 100 °C. However, within 24 h, a piece of Pd wire
(0.04 g, 0.25 mm diameter, 99.9%) has been dissolved in a 50 mL
mixed (1:1) solution between the seawater and 2.0 mol dm−3 HNO3
at ca. 100 °C.
Fig. 10 shows the increasing dissolution rate constant with increasing
NaCl addition to the seawater. The solution preparation procedure is as
follows: The seawater is poured to 11.69 g NaCl crystals up to 50 mL in
a volumetric flask to prepare an additional 4.0 mol dm−3 NaCl seawater
solution. Mixing between 50 mL of 2.0 mol dm−3 HNO3 and 50 mL of
the additional 4.0 mol dm−3 NaCl seawater solution gives 100 mL of
1.0 mol dm−3 HNO3 containing additional 2.0 mol dm−3 NaCl seawater
solution (totally ca. 2.23 mol dm−3 Na+ and ca. 2.28 mol dm−3 Cl−).
The log (k/s−1) value increases remarkably from − 4.52 to − 3.71
with additional 0.5 mol dm−3 NaCl and it remains almost a constant
value up to 2.0 mol dm−3 NaCl. As for platinum, however, a 2.8 mg Pt
mass has been left without dissolving out of ca. 20 mg platinum
wire (0.1 mm diameter) in 100 mL of the 1.0 mol dm−3 HNO3 and
0.5 mol dm−3 NaCl added seawater after 6 days. Without the support
by additional salts, however, no Pt can be dissolved in the 1:1 mixture
between seawater and 2.0 mol dm−3 HNO3 (vide supra).
We just mention that the “excellent” dissolution ability of seawater
mixed with 2.0 mol dm−3 HNO3 for gold has been successfully applied
to collecting of gold from waste electronic devises, just mechanically
tipped. The ICP-atomic emission spectrometry has demonstrated the
dissolution of Au out of the waste electronic devices, as well as Cu, Ni,
Al, Si, Zn, and B. Recovery of gold from secondary sources has been
widely reviewed by S. Syed [41].
3.7. “Dilute aqua regia” of various HNO3 and HCl concentrations
In the CRC book [42] the preparation procedure of aqua regia is de-
scribed as follows: “Mix 1 part concentrated HNO3 with 3 parts of con-
centrated HCl. This formula should include one volume of water if the
aqua regia is to be stored for any length of time. Without water, objec-
tionable quantities of chlorine and other gases are evolved.” According
to this description, it has been well known that dilution with water is es-
sential for the long term storage of active aqua regia. Assuming
16 mol dm−3 HNO3 (70%) and 12 mol dm− 3 HCl, the original aqua
regia is composed of 4.0 mol dm−3 HNO3 and 9 mol dm−3 HCl, while
the water added solution is calculated to contain still 3.2 mol dm− 3
HNO3 and 7.2 mol dm−3 HCl.
The gold dissolution in “dilute aqua regia” has been examined with
different HNO3 and HCl combinations in various concentrations: 0.5,
1.0, and 2.0 mol dm−3 HNO3; 0.5, 1.0, (1.5), and 2.0 mol dm−3 HCl at
60 °C (cf. Fig. 11). As mentioned in a previous section, the [1.0, 1.0] “di-
lute aqua regia” (of 1.0 mol dm−3 HNO3 and 1.0 mol dm−3 HCl) gives a
Fig. 11. The dissolution rate constants of gold wire (19.7 ± 0.5 mg) in various concentra-
tions of “dilute aqua regia” at ca. 100 °C: (▽) 0.5 mol dm−3 HNO3 60 °C; (○) 1.0 mol dm−3
HNO3 60 °C; (●) 2.0 mol dm−3 HNO3 60 °C; (▼) 0.5 mol dm−3 HNO3 100 °C; (△)
1.0 mol dm−3 HNO3 100 °C. The rate constants at 60 and ca. 100 °C have been evaluated
by the mass and spectroscopy, respectively.
 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76
dissolution rate constant of log (k/s−1) = − 4.80 at 60 °C. With in-
creased HCl concentration, i.e., of 2.0 mol dm−3, the rate constant in-
creases: the [1.0, 2.0] “dilute aqua regia” gives − 3.74. The [2.0, 0.5],
[2.0, 1.0], and [2.0, 2.0] solutions give a linear relation between log
(k/s− 1) and c(HCl) values, as − 4.76,− 4.13, and − 3.33.
Comparing the [0.5, 2.0] with the [2.0, 0.5] solution, we have noticed
that the [0.5, 2.0] (−4.13) is faster than the [2.0, 0.5] (−4.76), probably
because the ratio between c(HNO3) and c(HCl) in the [0.5, 2.0] may be
more favorable than that of the [2.0, 0.5] as the “dilute aqua regia” me-
dium. Similarly, the [1.0, 2.0] (− 3.74) is faster than the [2.0, 1.0]
(−4.13) at 60 °C. At 100 °C, the rate constant of “dilute aqua regia” in-
creases remarkably, e.g. the [1.0, 1.0] (−3.14) at ca. 100 °C is 46 times
faster than the [1.0, 1.0] (− 4.80) at 60 °C. In a very “dilute aqua
regia,” it takes a very long time to dissolve gold wire (ca. 19.7 mg)
completely: 8 days in both the [0.5, 1.0] and the [1.0, 0.5] solutions at
60 °C.
3.8. Raman spectra of various HCl concentrations
In this final section, we would like to discuss the reason why the
“dilute aqua regia” as well as the proper aqua regia possesses the
gold dissolution ability. At first, we may mention that Raman
spectroscopy is a technique preferably used for studies of water
structure because that the proportions of intensities of main OH
stretching modes (around ν = 3200 and 3400 cm− 1 ) are thought
to reflect well a network of hydrogen bonds as well as its distur-
bance by the presence of some solute. Scherer et al. [43] have
reported Raman spectra of liquid H 2 O from − 10 to 90 °C and
proposed two classes of water, symmetrical, hydrogen-bonded
complex (dibonded: ν1 = 3225 cm−1, ν3 = 3440 cm−1), and an asym-
metrically hydrogen-bonded complex one strong (ν = 3422 cm−1) and
one weak bond (ν = 3615 cm− 1). Nevertheless, we have regarded
simply and successfully (cf., ref. [16e]) the OH stretching bands around
ν = 3250 cm−1 and that at 3400–3450 cm−1 as strongly hydrogen-
bonded and weakly hydrogen-bonded, respectively, disregarding the
band around 3600 cm−1 which is ideally approaching to the stretch
band (ν1 = 3657.1 cm−1 in infrared) of H2O in the gas phase [44].
Fig. 12 shows the Raman spectra of OH stretching bands of H2O con-
taining various HCl concentrations at room temperature. The band in-
tensity at ν = 3250 cm−1 decreases remarkably, compared with that
at 3400 cm− 1, with increasing HCl concentrations, and at the same
time, the band at 3400 cm−1 shifts toward higher wavenumbers up to
3460 cm− 1 at 5 mol dm− 3 HCl. The ratio of band intensities (I3250/
I3400) decreases from 0.88 (of no HCl) to 0.79, 0.73, and 0.60 for 1.0,
2.0, and 5.0 mol dm−3 HCl. Thus, the increased distortion of the bulk
Fig. 12. Raman spectra of H2O containing increasing concentrations of HCl at room tem-
perature. The intensities of the band around ν = 3400 cm−1 are standardized.
75
water structure with increasing HCl in H2O has been demonstrated by
the remarkable decrease of the band intensity at 3250 cm−1, compared
with that around 3400 cm−1. The temperature dependence in the H2O
Raman spectra by Scherer et al. [43] is quite correspondent to the HCl
concentration dependence, observed by us.
We may just take into account that the intensity ratios of the main
Raman components (around ν = 3250 and 3400 cm−1) of H2O have
been reported to alter with the different excitation wavelengths in
Raman spectroscopy [45]. It goes without saying that, in the present
work, a single wavelength (514.5 nm) has been utilized for excitation
in measuring the Raman spectra.
Levering et al. [46] have studied the effect of hydrogen-halide on the
bulk water structure with IR and Raman spectroscopy, and have con-
cluded that within the bulk liquid environment for the aqueous HCl,
HBr, and HI solutions, the hydrogen-bonding network of water becomes
increasingly disrupted by the solvation shells of the anion as the size,
polarizability, and concentration of the anions is increased. The Raman
spectra they have reported indicate remarkable increases in the intensi-
ty around ν = 3400 cm−1 at the expense of that at 3250 cm−3 for HCl
and HBr as well as NaCl and NaBr of a 0.1 molar fraction. The ratio of
I3250/I3400 has decreased remarkably in the order of HCl b HBr b NaCl
b NaBr.
Previously [22], we have discussed that the distortion (difference
from the intrinsic water structure) of bulk water should be caused by
not only NaCl but also LiCl, MgCl2, and CaCl2 of higher concentrations,
whereas the latter metal ions at lower concentrations have been
regarded to be of “structure-making” [18]. Now, we may propose that
even a strong acid, HCl, must not be of “structure-making” (in a tradi-
tional meaning) because the strong hydrogen-bonded structure con-
structed around the H+ ion is much different from that of the intrinsic
bulk water. All the properties of bulk water should be originated by
the natural bulk water structure, i.e., the highly “self-assembled struc-
ture.” The oxidation ability of HNO3 in the “dilute aqua regia” should
be acquired by the co-existence of HCl (0.5–2.0 mol dm−3) as well as
further distortion in higher temperature (60–100 °C). Even AlCl3 at
higher concentrations in aqueous solution may be also newly catego-
rized into “bulk-water-breakers”, now that it is found that the dilute ni-
tric acid (b2 mol dm−3) is vested with the strong oxidation ability by
AlCl3. We have to notice again that the aqueous nitric acid
(b2 mol dm−3) has been long regarded as no oxidizer [12]. Finally, we
would like to note that the following idea is becoming popular in solu-
tion chemistry: The term “structure-making” and “structure-breaking”
is not a very precise expression in that incorporation of small ions in
water invites a new structure (hydration structure) formation different
from the intrinsic water structure [47].
4. Conclusion
Although aqueous nitric acid (b2 mol dm−3) itself exhibits no oxida-
tion ability, it acquires the strong oxidation ability to oxidize Cl− to Cl2
in the presence of chloride salts. The chlorine (Cl2)–chloride (Cl−) sys-
tems provided by aqueous nitric acid containing enough amounts of
chloride salts, have been found to be excellent media for dissolving pre-
cious metals, especially, pure gold. In general, the increasing concentra-
tions of chloride salts cause the advanced ability for gold dissolution.
However, reversal rate decreases have been also observed by the excess
addition of chloride salts. The reversal rate decreases for CaCl2 at higher
concentrations are certainly accounted for by the association between
NO− 3 and Ca
2+
. At higher temperatures, such as 60 or 100 °C, gold can
easily be dissolved into the mixed solution between diluted aqueous
HNO3 and HCl, i.e. “dilute aqua regia”, which can be applied to prepare
a 1000 ppm gold stock solution easily. The salt addition of various
kinds of anions to “dilute aqua regia” causes both acceleration and
deceleration in the gold dissolution rate, and the obvious deceleration
by Na2SO4 is attributed to the Brønsted basicity of sulfate ions which
may reduce the effective HNO3 concentration. The “excellent” gold
76 M. Hojo et al. / Journal of Molecular Liquids 194 (2014) 68–76

dissolution ability of the mixture of seawater with aqueous nitric acid (e) M. Hojo, T. Ueda, E. Ueno, T. Hamasaki, T. Nakano, Bull. Chem. Soc. Jpn. 83 (2010)
401;
can be applied to the recovery of gold from waste electronic devices. (f) M. Hojo, S. Aoki, Bull. Chem. Soc. Jpn. 85 (2012) 1023.
In electrochemical operation, the Au or Pt electrodes can be dissolved [17] C. Reichardt, D. Che, G. Heckenkemper, G. Schaefer, Eur. J. Org. Chem. (2001) 2343.
naturally in concentrated halides salt solutions. [18] H.S. Frank, W.-Y. Wen, Discuss. Faraday Soc. 24 (1957) 133.
[19] (a) M. Hojo, T. Ueda, H. Hamada, Z. Chen, S. Umetani, J. Mol. Liq. 145 (2009) 24;
(b) M. Hojo, T. Ueda, M. Ike, M. Kobayashi, H. Nakai, J. Mol. Liq. 145 (2009) 152;
Acknowledgment (c) M. Hojo, T. Ueda, A. Inoue, S. Tokita, J. Mol. Liq. 148 (2009) 109;
(d) M. Hojo, T. Ueda, T. Inoue, M. Ike, M. Kobayashi, H. Nakai, J. Phys. Chem. B 111
(2007) 1759.
We are grateful to Dr. Noriaki Hara of Tanaka Precious Metals for his [20] M. Hojo, S. Ohta, K. Ayabe, K. Okamura, K. Kobiro, Z. Chen, J. Mol. Liq. 177 (2013)
valuable technical suggestions and kindness in supplying us with not 145.
[21] M. Hojo, T. Ueda, C. Daike, F. Takezaki, Y. Furuya, K. Miyamoto, A. Narutaki, R. Kato,
only Ir and Ru powder samples but also the waste electronic devise
Bull. Chem. Soc. Jpn. 79 (2006) 1215.
sample. We also thank the Institute of Muroto Deep Seawater, Kochi [22] M. Hojo, R. Kato, A. Narutaki, T. Maeda, Y. Uji-yie, J. Mol. Liq. 163 (2011) 161.
Prefecture, for supplying us with a sample of seawater. The present [23] (a) P. Debye, E. Hückel, Phys. Z. 24 (185) (1923) 305;
(b) D. Fraenkel, J. Phys. Chem. B 115 (2011) 14634.
work is partially supported by a Grant-in-Aid for the commission pro-
[24] K.S. Pitzer, Activity Coefficients in Electrolyte Solutions, 2nd ed. CRC, Boca Raton,
ject of “Research Program on Climate Change Adaptation” from the Florida, 1991.
Ministry of Education, Culture, Sports, Science and Technology, Japan. [25] M. Nakahara, Dictionary of Inorganic Compounds and Complexes, Kodansha Scien-
tific, Tokyo, 1997. (in Japanese).
[26] P.G. Jones, R. Schelbach, E. Schwarzmann, Z. Naturforsch. B 42 (1987) 522.
References [27] C.M. Roehl, J.J. Orlando, J.G. Calvert, J. Photochem. Photobiol. A Chem. 69 (1992) 1.
[28] Y. Marcus, Ion Properties, Marcel Dekker, New York, 1997. 43.
[1] (a) M.S. Quinby-Hunt, K.K. Turekian, EOS Trans. AGU 64 (1983) 130; [29] SicFinder: Calculated using Advanced Chemistry Development (ACD/Labs) Software
(b) Y. Nozaki, EOS Trans. AGU 78 (1997) 221. V11.02 (© 1994–2013 ACD/Labs).
[2] E. Sonstadt, Chem. News 25 (1872) 196(231 and 241). [30] R.C. Weast, Handbook of Chemistry and Physics, 70th edn CRC, Boca Raton, Florida,
[3] E. Berl, J. Chem. Educ. 14 (1937) 203. 1989. D-165.
[4] M. Koide, V. Hodge, E.D. Goldberg, K. Bertine, Appl. Geochem. 3 (1988) 237. [31] The Merck Index 11th edn, 1989. 9595 Merck, Rahway, NJ.
[5] K.K. Falkner, J.M. Edmond, Earth Planet. Sci. Lett. 98 (1990) 208. [32] Ref. 30, D-163.
[6] P.H. Abelson, Science 233 (1986) 141. [33] S.K. Patil, H.D. Sharma, Can. J. Chem. 47 (1969) 3851.
[7] L.J. Beckham, W.A. Fessler, M.A. Kise, Chem. Rev. 48 (1951) 319. [34] (a) R.E. Hester, R.A. Plane, J. Chem. Phys. 40 (1964) 411;
[8] J.O. Marsden, C.I. House, The Chemistry of Gold Extraction, 2nd edn Society for (b) R.E. Hester, R.A. Plane, Inorg. Chem. 3 (1964) 769.
Mining, Metallurgy, and Exploration, Inc. (SME), Littleton, Colorado, 2006. 233. [35] D.E. Irish, G.E. Walrafen, J. Chem. Phys. 46 (1967) 378.
[9] G.L. Putnam, Eng. Min. J. 145 (1944) 70. [36] A.P. Kreshkov, K.A. Komarova, V.P. Gorbashev, Elektrokhimiya 10 (1974) 1082(E:
[10] (a) In: A.J. Bard, R. Parsons, J. Jordan (Eds.), Standard Potentials in Aqueous Solu- 1025).
tion, IUPAC, Marcel Dekker, New York, 1985, p. 318. [37] E. Hoegfeldt, E. Hoegfeldt, Part A: Inorganic Ligands, IUPAC Chemical Data Series,
(b) p. 71. No.21Pergamon, Oxford, UK, 1982.
(c) p. 58. [38] M. Whitfield, in: J.P. Riley, G. Skirrow (Eds.), 2nd ed., Chemical Oceanography, vol. 1,
[11] J.O. Marsden, C.I. House, The Chemistry of Gold Extraction, 2nd edn Society for Min- Academic Press, London, 1975, (Chapter 2).
ing, Metallurgy, and Exploration, Inc. (SME), Littleton, Colorado, 2006. 272. [39] H.D. Holland, The Chemistry of the Atmosphere and Ocean, Wiley, New York, 1978.
[12] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry: A Comprehensive Text, (Chapter 5).
4th ed. Wiley, New York, 1980. (ch. 6.). [40] http://en.wikipedia.org/wiki/Seawater.
[13] D.M. Petkovic, J. Chem. Soc. Dalton Trans. (1982) 2425. [41] S. Sayd, Hydrometallurgy 115–116 (2012) 30.
[14] M.H. Kuo, A. David, N. Kamelamela, M. White, M.J. Shultz, J. Phys. Chem. C 111 [42] Ref. 30, D-130.
(2007) 8827. [43] (a) J.R. Scherer, M.K. Go, S. Kint, J. Phys. Chem. 78 (1974) 1304;
[15] M. Hojo, Pure Appl. Chem. 80 (2008) 1539(and references cited therein). (b) D.E. Irish, M.H. Brooker, in: R.J.H. Clark, R.E. Hester (Eds.), Advances in Infrared
[16] (a) L.C. Manege, T. Ueda, M. Hojo, Bull. Chem. Soc. Jpn. 71 (1998) 589; and Raman Spectroscopy, vol. 2, Heyden, London, 1976, p. 239, (Chapter 6).
(b) L.C. Manege, T. Ueda, M. Hojo, M. Fujio, J. Chem. Soc., Perkin Trans. 2 (1998) 1961; [44] P.F. Bernath, Phys. Chem. Chem. Phys. 4 (2002) 1501.
(c) M. Hojo, T. Ueda, S. Inoue, Y. Kawahara, J. Chem. Soc., Perkin Trans. 2 (2000) [45] M. Pastorczak, M. Kozaneck, J. Ulanski, J. Phys. Chem. A 112 (2008) 10705.
1735; [46] L.M. Levering, M.R. Sierra-Hernandez, H.C. Allen, J. Phys. Chem. C 111 (2007) 8814.
(d) M. Hojo, T. Ueda, E. Ueno, T. Hamasaki, D. Fujimura, Bull. Chem. Soc. Jpn. 79 [47] K. Yonehama, Y. Yoshimura, T. Takekiyo, H. Kanno, Bull. Chem. Soc. Jpn. 82 (2009)
(2006) 751; 563.