Environ. Sci. Technol.

1999, 33, 2607-2610

Role of Temperature and pH in versus dissolution experiments. If differing solid age is

dominant in controlling cupric hydroxide solubility, the
Cu(OH)2 Solubility shorter term precipitation experiments would produce solids
with higher solubility (1, 18). Likewise, if a kinetic limitation
prevented equilibrium from being achieved with a particular
LOAY HIDMI AND MARC EDWARDS* solid phase, an erroneously high solubility product would be
Department of Civil Engineering, Virginia Polytechnic estimated for the case of precipitation and a lower solubility
Institute, 407 NEB, Blacksburg, Virginia 24061-0246 would be expected for dissolution. Interestingly, only few
researchers (12) have directly addressed the approach to
equilibrium through both dissolution and precipitation
reactions. Finally, since the type of solid that forms depends
The wide variation in reported cupric hydroxide solubility on the anion present, and chemical order of addition has
constants is attributed to the age of the solid, kinetic limitations been noted to influence results in the copper hydroxide
in approaching equilibrium, and transient formation of system (1), unsuspected formation of basic copper nitrate
basic copper nitrate solids. In situations similar to those solids could influence solubility under some circumstances.
where cupric hydroxide solubility constants were determined This work examines each of these possibilities by ex-
via precipitation in dilute nitrate solutions, copper nitrate panding the experimental scope and resulting interpretations
solids (gerhadtite) form first followed by a rapid transition to of previous researchers (1). The pH range from 7 to 9, not
directly examined in that work but of obvious importance
cupric hydroxide and then a slower transition to tenorite.
in engineered systems, is a special focus of this investigation.
In solutions containing hydroxide and nitrate as the Moreover, the practical impacts of temperature on Cu(OH)2
sole anionic constituents, fresh cupric hydroxide solids solubility and aging are established.
control soluble copper concentrations for minutes to days
depending on pH and temperature. Thus, although a log Materials and Methods
K of 9.36 ( 0.02 was most consistent with the experimental All chemicals used in these experiments were analytical
data for Cu(OH)2 solubility, decreasing log K values from reagent grade. Solutions were prepared from deionized type
9.36 [fresh Cu(OH)2] to 7.6 [CuO] can be expected with aging III reagent grade water (Milli-Q water reagent system) filtered
as controlled by temperature, pH, and other factors. through a 0.2 µm pore size filter before use.
Batch Tests. After preliminary experiments at pH 9
demonstrated identical results using either Teflon or glass
Introduction beakers, 2 L glass beakers were used in all experiments. For
Understanding freshly precipitated copper hydroxide solu- precipitation experiments, 2 L of solution containing 0.5 mM
bility has long been of key importance in the treatment of of Cu(NO3)2 and 1 mM NaNO3 was titrated with 50 µL
copper-containing wastewater (1, 2). Recently, copper hy- increments of 1 M freshly prepared NaOH dosed at 1 min
droxide solubility has been implicated in controlling cor- intervals. Throughout the experiment, the system was mixed
rosion byproduct release to drinking water from relatively with a 5 cm magnetic stirrer at 200 rpm and monitored for
new copper plumbing (3); thus, the solubility product is of time periods of up to 1 month. The final ionic strength was
direct relevance to compliance with copper regulations in 2 mM as NaNO3 after all copper was precipitated. The pH
drinking water and in assessing human exposure. Corrosion was monitored using a Beckman Φ10 pH meter and a Corning
of plumbing is also a key contributor to copper loading of combination electrode. During experiments over the course
wastewater treatment plants and the environment (4). of a single day, the recalibration was every 2 h, whereas the
Even under well-defined conditions, reported solubility meter was recalibrated immediately before measurements
products (5-21) for freshly precipitated Cu(OH)2 vary by at longer time periods. The pH was maintained to within
nearly 3 orders of magnitude (Table 1). Of these constants, (0.15 pH units of the target value through addition of 1 N
that determined by Schindler et al. (1965) is thought to be NaOH throughout the entire experiment. Temperature was
“the most reliable of those reported but it is classified as measured using a digital thermometer and maintained to
tentative because of the instability of the Cu(OH)2-alkali within (0.3 °C of the target value.
system and because there is no other work to substantiate CO2 was initially stripped from the water by bubbling the
it (7)”. The wide variation in reported constants has been solution with N2, and during the remainder of the experiment,
attributed to surface area effects (6, 18) and/or aging (1). beakers were covered to the extent possible with a double
Our review indicates that reported solubility products can layer of Parafilm. The total initial carbonate was <0.02 mM,
also be demarcated into groups of low (<8.64) and high (>9.6) whereas the final carbonate after 1 month of aging was 0.04
solubility, corresponding to experiments in which equilibrium mM at pH 7.0, 0.05 mM at pH 8.0, and 0.057 mM at pH 9.0,
was approached through dissolution or precipitation reac- as determined by titration and established equilibrium
tions, respectively (Table 1). This differentiation may also be constants. Thus, traces of carbonate were present in the
important since equilibration times and experimental pro- systems tested. However, replication of experiments at pH
cedures are roughly similar within each type of experiment. 7.0 and 8.0 at a temperature of 23 °C produced undetectable
That is, equilibration times were apparently always greater differences in the approach described above versus a
than 2.5 days when dissolution experiments were conducted, continuous N2 atmosphere, so effects of trace CO2 were
whereas a few hours or less is commonplace in precipitation deemed insignificant for solubility data reported in this work.
experiments. Determining Total Soluble Copper, Free Copper, and
At least three phenomena could produce the wide Solubility Constants. The determination of soluble copper
variation in copper solubility reported for the precipitation required a solid/liquid separation using either centrifugation
or filtration. Different filter sizes and materials were tested
* Corresponding author; fax: 540 231-7916; e-mail: edwardsm@ (0.45 µm Nylon, 0.02 µm cellulose acetate, and Amicon
vt.edu. ultrafilters YM30 of size 30K Da) and compared to soluble
10.1021/es981121q CCC: $18.00  1999 American Chemical Society VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2607
Published on Web 06/26/1999
TABLE 1. Reported log K [K ) {Cu2+}/{H+}2] for Cupric Hydroxide
ref log K equilibration time experimental procedure
Dissolution Reactions
5 8.17 not reported cupric hydroxide prepared by adding NaOH to cupric sulfate.
19 8.2 8 days sufficient hydrated copper oxide was added to known strength HNO3 and shaken for 8 days.
18 8.64 not reported, cupric hydroxide was prepared by treating Cu(NO3)2 - 3Cu(OH)2 with NaOH. Dissolution in
but 2.5 days 0.2 NaClO4 M until the pH equilibrates.
when replica-
ted in our lab
Precipitation Reactions
11 10.7 90 min NaOH was added to 0.7 NaClO4 containing between 6.0 × 10-5 to 2.4 × 10-4 m total Cu.
Free copper measured after 90 min.
10 10.3 not reported 10-3 of copper nitrate present in 0.1 KNO3 M precipitated by adding 1 M KOH to the medium.
The pH was changed between 2 and 12.
21 9.6 10 min copper standard added to 40% carbonate free NaOH, 10-3 M formaldehyde and 5 × 10-2 M.
pH was adjusted by adding known amounts of diluted HClO4 to lower the pH from 9.7 to
6.0. with 10 min equilibrating period.
17 10.3 not reported stability constants for copper hydroxide complexes were measured in solutions of CuCl2 in
NaClO4 at pH 4.0 and used to calculate the solubility product.
9 10.3-10.0 3-120 min copper concentrations > 5 × 10-6 M in 0.1 M of NaNO3 precipitated by adding dilute NaOH.
The value of log K of 10.3 was obtained after equilibration for 3-5 min. However, when
that period was changed to 120 min the obtained value for log K was 10.0.

copper obtained after temperature-controlled centrifugation

at 4000 rpm for 20 min. For the comparison, a 2 L solution
with an initial copper concentration of 1 mM Cu(NO3)2 was
adjusted to pH 6.0, 7.0., 8.0, and 9.0. Free copper was also
determined in each solution using an Orion cupric ion specific
electrode and a single junction reference electrode (1). The
electrode was calibrated immediately before use with stan-
dards in the concentration range under consideration. Using
this information, soluble copper could then be calculated
with the electrode based on the measured free copper
concentration, soluble pH, and reported cupric hydroxide
complexation constants (20). Soluble copper in filtered or
centrifuged samples was determined directly using a Varian
Liberty axial inductive coupled plasma emission spectroscopy FIGURE 1. Effect of time on the soluble copper concentration as
(ICP-ES) instrument according to standard method 3120 with measured through dissolution and precipitation of Cu(OH)2 solids.
a detection limit of (1 µg/L. Experimental conditions: pH 7.0, 23.4 °C, I ) 2 mM NaNO3 and 0.5
At a pH of 8 or less, the difference in calculating free mM Cu2+total.
copper using the ISE and the ICP approach was always less Kinetics of Equilibration: Precipitation vs Dissolution.
than 9%, regardless of whether centrifugation or filtration If the kinetic approach to solid/liquid equilibrium is insig-
was used for solid/liquid separation. For routine experiments, nificant when compared to the experimental timespan, the
0.45 µm filters were used for solid/liquid separation and the solubility product determined through precipitation or
ICP-ES was used to quantify soluble copper. The exception dissolution experiments should not differ. For precipitation
was at pH 8.0-9.5 and at total soluble copper concentrations of cupric ion at pH 7.0, the soluble copper concentration
of e20 µg/L, in which case more than 20% of the copper was decreased to half its initial value in just 3 min, and then
unavoidably lost to the filter. To obtain an accurate estimate steadily decreased to a stable level of 6.2 ( 0.5 mg/L Cu
of soluble copper at this higher pH range, a combination of between 2 and 23 h (Figure 1). During the period of relative
centrifugation (providing an upper bound to actual solubility) stability, it was not necessary to add base to maintain a
and filtration (providing a lower bound to actual solubility) constant pH. After about 24 h, the soluble copper concen-
measurements was necessary. tration began to decrease, corresponding to a time period in
All solubility constants [K ) [Cu2+]/[H+]2] reported in this which base was once again required to maintain constant
work are corrected to 298.1 K using reported enthalpy values pH at 7.0. Between 24 and 720 h, the color of the solid
(20) and to zero ionic strength using the extended Debye- gradually changed from blue to dark brown.
Huckel approximation (22). For the dissolution experiment, a subsample of the solid
Solid Characterization. Solids collected on the 0.45 µm formed during the precipitation experiment was collected
pore size filter were rinsed with MQ reagent grade water by filtration. This solid was then immediately resuspended
three times and then dried at room temperature in a in a solution of identical pH and ionic strength but without
desiccator. These solids were then inspected by scanning copper. Soluble copper concentrations increased steadily
electron microscopy (ISI-SX 30 KV), X-ray diffraction, and during the first 6 h before plateauing at the same 6.2 ( 0.2
wet chemical analysis after sulfuric acid digestion. mg/L Cu level observed during the concurrent precipitation
experiment (Figure 1). Thereafter, trends in solubility for the
Results and Discussion precipitation and dissolution experiments were nearly iden-
Initial experiments focused on the kinetics of equilibration tical, although the concentration of copper in the precipita-
for dissolution versus precipitation reactions. Thereafter, the tion experiment was slightly lower from that in the dissolution
effect of pH and temperature on solubility was examined experiment between 30 and 92 h. After 720 h, soluble copper
followed by a determination of solids composition and decreased to a relatively stable value of 0.054 mg/L in the
morphology. precipitation experiment and 0.064 mg/L in the dissolution

2608 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999
TABLE 2. Particle Characterization for Solids Aged for 30 Minutes and 96 Hours
aging period SEM XRD
pH 7.0 30 min no visible crystals gerhardtite (Cu2NO3(OH)3)
5h tapelike and oval shaped solids. gerhardtite (Cu2NO3(OH)3), likasite (Cu3NO3(OH)5‚2H2O), cupric
hydroxide (Cu(OH)2) nitrate solids < 20% of copper solids
96 h oval and slab shaped solids likasite and cupric hydroxide nitrate solids < 15% of copper solids
720 h large oval shape tenorite (CuO)
pH 8.0 30 min small oval shape gerhardtite and tenorite (CuO)
96 h large oval shape tenorite (CuO)
720 h large oval shape tenorite (CuO)
pH 9.0 30 min small oval shape gerhardtite, likasite, and tenorite
96 h large oval shape tenorite (CuO)
96 h large oval shape tenorite (CuO)

the same at 6.3 ( 0.2 mg/L. This suggests that the period of
metastability is dependent on very slight differences in
chemical addition rate and other subtle factors not precisely
controlled in experiments, a phenomena consistent with the
instability of the system under consideration. Trends at pH
7.5 were similar to those at pH 7.0 (Figure 1), but the
metastable copper level of 6.11 ( 0.2 mg/L persisted for just
4 h instead of the 24 h observed at pH 7.0.
A somewhat different trend was observed for pHs higher
than 8.0 in that an initial metastable period was not observed.
Instead, within the first hour, the color of the solid im-
mediately changed from blue to dark brown. As discussed
before, samples of this solution were centrifuged and filtered
to determine an upper and lower bound to the actual soluble
FIGURE 2. Effect of pH on cupric ion precipitation in 2 mM NaNO3 copper concentration. Initially, the difference in the measured
solutions at 23 °C. copper concentration between the two methods was ap-
proximately 0.4 mg/L for all tested pHs; however, that
experiment. After considering the significant Cu(OH)+, Cu- difference decreased with time. For example, at pH 8.0 and
(OH)20, and Cu2(OH)2-2 complexes (20), the log K of the after 24 h of aging, soluble copper was 0.12 mg/L and 0.043
metastable equilibrium period between 6 and 23 h was 9.36 mg/L as determined by filtration and centrifugation, re-
( 0.02, whereas after 1 month of aging, log K was 7.35 ( 0.05. spectively. Thus, the solubility product log K was between
The calculated values of log K are highly dependent on the 7.9 and 7.4. After 1 month, however, soluble copper levels
selected value of B2 for Cu(OH)2 formation, for which there were 0.043 ( 0.005 mg/L for both approaches.
is significant variation in the literature. If the lower range of Solid Phases. Precipitated solids were collected after 30
B2 values (i.e., log B2 ) -16.24) is deemed more accurate (1), min and 96 h of reaction time at pHs 7.0, 8.0, and 9.0. Another
the above values for log K would increase to 9.69 and 7.75 sample of solids aged at pH 7.0 for 5 h was collected to better
for data at 6-23 h and 1 month of aging, respectively. characterize the solids present during the metastable equi-
Since the soluble copper concentrations obtained through librium. In all cases, nitrate-containing copper solids (likasite
precipitation and dissolution reactions were the same only or gerhardtite) were identified by XRD after 30 min reaction
after 6 h, a significant portion of the previously cited 2 order time (Table 2). After 5-96 h, crystalline cupric hydroxides or
of magnitude differences in log K for cupric hydroxide solids tenorite were also present, either exclusively or in addition
may be attributed to kinetics of precipitation and dissolution. to the nitrates.
For example, if solubility had been determined after only 25 To better define the concentration of copper nitrate solids
min in the precipitation and dissolution experiments for the present, solids were dissolved in sulfuric acid and nitrate
exact same solid, log K’s of 9.6 and 8.45 would have been was determined by ion chromatography. After 5 h, copper
determined, respectively. nitrate solids always accounted for less than 20% of the total
Consistent with the observations of previous researchers copper solids present, indicating that copper hydroxides were
(1, 6, 18), solid aging and/or surface area effects also seem dominant. The relative concentration of copper nitrates
dominant within the two types of experiments. That is, a tended to decrease with time, although they were always
metastable equilibrium condition is obtained for solids aged present through 96 h of aging. After 1 month of aging and
a few hours, which roughly corresponds to the log K reported at all pHs tenorite was the dominant solid and there were
for precipitation reactions, whereas another metastable no traces of copper nitrate solids. Thus, nitrate-containing
equilibrium condition is achieved after days or weeks of aging solids seem to be of low significance after 5 h aging time. On
which corresponds to solubility products that have been the other hand, at shorter time periods, the results strongly
reported for dissolution reactions (Table 1). At intermediate support the presence of basic cupric nitrate solids. Though
timespans, log K’s between these extremes were valid. the XRD data on likasite or gerhardtite might result from
Solubility and pH. The precipitation experiment was reaction of sorbed NO3- on amorphous Cu(OH)2 as the solids
repeated at a range of pHs (Figure 2). At pH 7.0, soluble dried, this seems highly unlikely given the dilute solutions
copper concentrations at equilibrium were 6.41 mg/L, a level employed, the solid-rinsing procedure used, and the com-
which was stable for 24 h. Thereafter, the soluble copper plete absence of these phases at later experimental times.
concentration started to decrease gradually until reaching a Although the copper nitrate solids represented a relatively
value of 0.056 mg/L after 1 month of aging. Interestingly, small fraction of the total copper solids present, the question
comparing results at pH 7.0 but repeated at different days remained as to whether these solids might be responsible
(Figure 1 vs Figure 2), the duration of pseudo-equilibrium for the initial metastable equilibrium period. To address this
changed slightly but the average level of soluble copper was question, a pure copper electrode was immersed in a solution

VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2609
 diffraction and stoichiometry (if nitrate concentrations were
above 0.01 M). The authors hypothesized that an amorphous
cupric hydroxide solid phase formed first, followed by a
gradual transition to basic cupric nitrate solids, although we
note that the data are also qualitatively consistent with soluble
Cu(OH)20 complex formation followed by precipitation of
basic cupric nitrate solid. In contrast, interpretation of
titration data by other investigators in 1 mM nitrate solutions
suggested that a transient basic cupric nitrate phase might
form first, followed by rapid conversion to cupric hydroxide
(1). The X-ray and wet chemical analysis reported in this
work are obviously more consistent with the latter hypothesis,
although subtle differences in rates of base addition and
FIGURE 3. Effect of temperature on the precipitation of cupric nitrate concentrations could also partly explain differences
hydroxide solids. I ) 2 mM NaNO3 and 0.5 mM Cu2+total. in the results obtained.
Acknowledgments
of deionized water open to atmospheric oxygen and galvani-
This work was supported by the National Science Foundation
cally oxidized at pH 5 according to procedures presented
under Grant BES-9729008 and the American Water Works
elsewhere (22). The stable reaction product was freshly
Association Research Foundation (AWWARF). The opinions,
precipitated cupric hydroxide, as predicted thermodynami-
findings, conclusions, or recommendations are those of the
cally and measured by XRD. When this freshly precipitated
authors and do not necessarily reflect the views of AWWARF
solid was collected and placed into a container of deionized
or NSF.
water maintained at pH 7.0 by NaOH addition, soluble copper
concentrations leveled off at 5.76 mg/L after 4 h. After Literature Cited
correcting for ionic strength effects, this translates to a log (1) Patterson, J. W.; Boice, R. E.; Marani, D. Environ. Sci. Technol.
K of 9.35 for cupric hydroxide compared to the log K of 9.36 1991, 25, 1780-1787.
( 0.02 determined for the mixed solids of age greater than (2) Marani, D.; Patterson, J. W.; Anderson, P. R. Water Res. 1995,
6 h in the presence of nitrate. Thus, at time periods greater 29 1317-1326.
(3) Edwards, M.; Schock, M. R.; Meyer, T. E. JAWWA 1996, 88 (3),
than 6 h in the original experiment, solubility was controlled 81-94.
by Cu(OH)2 and was not influenced by nitrate. (4) Isaac, R. A.; Gil, L.; Cooperman, A. N.; Hulme, K.; Eddy, B.; Ruiz,
Temperature Effects. The role of temperature was M.; Jacobson, K.; Larson, C.; Pancorbo, O. C. Environ. Sci.
examined in copper precipitation experiments at pH 7.0 using Technol. 1997, 31, 3198-3203.
filtration to determine soluble copper. The general trends (5) Akselrud, N. V.; Fialkov, Ya. A. Ukrain. Khim. Zh. 1950, 16, 283.
(6) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; John Wiley
for changes in soluble copper with time were similar to those
& Sons: New York, 1976.
noticed at room temperature. However, differences in the (7) Dirkse, T. P.; Michalowiski, T.; Akaiwa, H.; Izumi, F. Solubility
soluble copper concentration and duration of the metastable Data Series; Pergamon Press: New York, 1984; Vol. 23.
equilibrium condition were noted (Figure 3) at 4.2, 12.5, and (8) Feitknecht, W.; Schindler, P. W. Pure Appl. Chem. 1963, 6, 130.
23.6 °C. At 4.2 °C, the precipitate was always a light-blue (9) Gulens, J.; Leeson, P. K.; Seguin, L. Anal. Chim. Acta 1984, 156,
color and a metastable copper concentration of 22.1 ( 0.4 19.
(10) Heijne, G. J. M.; van der Linden, W. E. Anal. Chim. Acta 1978,
mg/L was maintained for about 72 h. Thereafter, copper 96, 13.
solubility rapidly decreased to 1.51 ( 0.1 mg/L after 1 month. (11) Hittinger, R. C. Determination of Cupric-Hydroxide Ion Pair
At 12.5 °C, the initial duration of metastable equilibration Stability Constants at 0.7 Ionic Strength. Master Thesis,
was about the same as that at 4.2 °C, although the soluble University of Rhode Island, 1975.
copper concentration was 15 mg/L ( 0.15, eventually (12) Mcfaden, P.; Matijevic, E. Inorg. Nucl. Chem. 1973, 35, 1883.
declining to 1.11 ( 0.12 mg/L after 1 month. The solids (13) Näsänen, R. Ann. Acad. Sci. Fenn. 1942, A59, 7.
(14) Näsänen, R. Ann. Acad. Sci. Fenn. 1943, A59, 3.
remained blue during the entire experiment. At 23.6 °C, the (15) Näsänen, R.; Tamminen, V. J. Am. Chem. Soc. 1949, 71, 1994.
metastable equilibrium persisted for much less than 72 h as (16) Oka, Y. Nippon Kagaku Kaishi 1938, 59, 971.
was observed previously. (17) Paulson, A. J.; Kester, D. R. J. Solution Chem. 1980, 4, 269.
Assuming that the difference in copper levels during (18) Schindler, P.; Althaus, H.; Hofer, F.; Minder, W. Helv. Chim.
metastable equilibrium can be attributed to enthalpy con- Acta. 1965, 48, 1204.
(19) Sircar, S. C.; Prasad, B. J. Indian Chem. Soc. 1956, 33, 361.
siderations over the temperature range 5-35 °C using the
(20) Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum
Van’t Hoff relationship (23), ∆H for the Cu(OH)2 precipitation Press: New York, 1976; Vol. 4.
reaction was estimated as 14.5 kcal/mol using the data at (21) Stella, R.; Ganzerli-Valentini, M. T. Anal. Chem. 1979, 51, 2148.
12.5 and 23.6 °C. This value is close to the 13.19 kcal/mol (22) Edwards, M.; Ferguson, J. F. JAWWA. 1993, 10, 105.
reported elsewhere for Cu(OH)2 solids in the literature (20). (23) Stumm, W.; Morgan, J. Aquatic Chemistry, 3rd ed.; Wiley-
Final Comment on Sequence of Solid Formation. Recent Interscience: New York, 1996; p 997.
(24) Spark, K. M.; Johnson, B. B.; Wells, J. D. Aust. J. Chem. 1990, 43,
studies of cupric ion precipitation by base addition in dilute 749.
cupric nitrate solutions have led to divergent conclusions
regarding the likely sequence of solid formation. In one case
Received for review November 2, 1998. Revised manuscript
(24), initial titration data were consistent with 2 mol of OH-
received May 12, 1999. Accepted May 21, 1999.
consumption per mole of Cu2+ precipitated, followed by the
appearance of a Cu(OH)1.5(NO3)0.5 solid as confirmed by X-ray ES981121Q

2610 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999