Hydrometallurgy 2003 – Fifth International Conference in Honor of Professor Ian Ritchie –

Volume 1: Leaching and Solution Purification

Edited by C.A. Young, A.M. Alfantazi, C.G. Anderson, D.B. Dreisinger, B. Harris and A. James
TMS (The Minerals, Metals & Materials Society), 2003

EQUILIBRIUM MODIFIERS IN COPPER SOLVENT

EXTRACTION REAGENTS - FRIEND OR FOE?
Gary Kordosky and Michael Virnig1
Cognis Corporation
5051 Estecreek Drive
Cincinnati, OH 45232-1446

Abstract

Equilibrium modifiers, first introduced into copper solvent extraction reagents in the late
1960’s, are widely used. The properties of various equilibrium modifiers with respect to copper
extraction, selectivity, reagent stability and entrainment are discussed and some conclusions are
drawn relating the structure of the modifier to the properties of the modifier. A new highly
selective, low viscosity modifier is also introduced.

Introduction
As used in solvent extraction modifiers are chemical compounds added to the organic phase in a
solvent extraction process that change one or more properties of the organic phase. For
example, the use of isodecanol or tridecanol as a modifier in uranium solvent extraction plants
using tertiary amine extractants was a common and well known practice at the time copper
solvent extraction was introduced. The modifier is needed in the amine system to increase the
solubility of the amine uranium complex in the hydrocarbon diluent or else a third phase,
consisting of the amine uranium complex and protonated amine, forms. This third phase
gathers in the settler between the diluent phase and the aqueous phase and soon there is no
extractant in the circulating organic phase and solvent extraction ceases.
A modifier of this type is not needed in copper solvent extraction because General Mills (now
Cognis), in developing the first copper solvent extractants, built the active extractant so that the
extractant and the extractant copper complex would be highly soluble in hydrocarbon diluents
similar to ordinary kerosene. The fact that General Mills designed a copper extractant to be
highly soluble in a hydrocarbon diluent is not a surprise since General Mills, as the developer of
tertiary amines for uranium extraction, recognized this as an important extractant property.
The commercial practice of leach-solvent extraction-electrowinning (L-SX-EW) in the copper
industry began in March 1968 when the Bluebird Mine in Miami, Arizona, commenced their L-
SX-EW operations [1]. This was three years after General Mills, Inc. introduced LIX 64, the
first viable copper solvent extraction reagent, to the market.
Initially acceptance of this new technology was relatively slow for several reasons including the
fact that the single viable extraction reagent, LIX 64N, was only moderately strong as a copper
extractant, had relatively slow metal transfer kinetics and only moderately good Cu/Fe
selectivity. This limited the leach solutions that could be treated by solvent extraction to those
having moderate amounts of copper and iron at a pH above ~ 1.5.
In the late 1960’s Ashland Chemical tried to overcome the shortcomings of LIX 64N with the
introduction of the reagent Kelex® 100, a copper extraction reagent based on the molecule
alkylated 8-hydroxyquinoline [2]. This molecule is a very strong copper extractant that does
not strip well and later Ashland also introduced the idea of using varying amounts of modifier,
either nonylphenol or isodecanol, and a proper diluent to facilitate the stripping of copper with
905
normal tankhouse electrolytes [3]. The fact that Ashland Chemical thought to use isodecanol as
a modifier is not a surprise since Ashland Chemical also supplied tertiary amines for uranium
extraction. However, the fact that Ashland Chemical discovered that modifiers used to improve
the phase separation and entrainment characteristics of the amine extractants could also alter the
copper extraction œ stripping equilibrium is the first report of the use of equilibrium modifiers
in copper solvent extraction and represents a significant advance in copper solvent extraction
reagent technology. For a variety of reasons Kelex reagents were not viable as copper solvent
extraction reagents and the idea of using nonylphenol or isodecanol to alter the extraction œ
stripping equilibrium reaction for copper solvent extraction was lost for a time.
In 1974 Birch reported that P-1 from Acorga Ltd. was a very strong copper extractant with
rapid copper transfer kinetics and excellent Cu/Fe selectivity [4]. Unfortunately, the reagent
was such a strong copper extractant that efficient stripping required some 250 g/l acid, an acid
content not compatible with electrowinning, and for this reason P-1 initially had little impact on
copper solvent extraction technology.
Requirements of Copper Solvent Extraction Reagents
It is generally accepted that a copper solvent extraction reagent has to satisfy the following
criteria: copper extraction from solutions generated by the copper leaching process, copper
stripping under conditions consistent with electrowinning high quality copper, high selectivity
for copper, fast copper transfer kinetics, fast phase separation, low tendency to form crud,
stability under circuit conditions, high solubility in a low cost hydrocarbon diluent, low aqueous
solubility and a relatively low cost. Reagent properties are related in one way or another to the
structure of the extractant, the method by which the extractant is produced and the components
of the organic phase in which the extractant is dissolved. For example, in the opinion of the
author and others crud formation is influenced by uncomplexed extractant and the reagent (14).
It is not a surprise that a change in one property of the extraction system will often result in a
change in another property. Finally, successful copper solvent extraction reagents tend to have
a good balance of all the properties listed above.
The chemical equation below describes in a simplified way the extraction and stripping of
copper with a chelating type copper extractant.
2RH(org) + Cu2+(aq) R2Cu(org) + 2H+(aq) (1)
If the reagent is a weak extractant the above equilibrium will lie to the left, that is, the copper
extractant will not have enough extraction power to give the desired copper recovery. If the
reagent is too strong an extractant the above equilibrium will lie far to the right, that is, the
copper extractant will hold the copper so tightly that copper will not strip well when the reagent
copper complex is contacted with a typical copper tankhouse electrolyte.
For the reagents SME 529 and LIX 64N the equilibrium above is shifted to the left about as far
as it could be while still having enough reagent strength to be a viable copper extractant. For
the extraction molecule in P-1 the equilibrium above is shifted so far to the right that P-1 holds
copper too tightly to be an economically viable copper extractant. Thus in the mid 1970’s,
there still existed a need for a copper solvent extraction reagent that could more efficiently treat
leach solutions with a higher copper concentration and/or a lower pH than could LIX 64N or
SME 529. A reagent with faster kinetics and better selectivity would provide additional benefit.
In the mid 1970’s all of the then viable (LIX 64N and SME 529) or to be viable (P-1) copper
solvent extractant molecules were based on the general structure shown in Figure 1. Note that
the only difference between the structure for the extractant in SME 529 and the structure for the
extractant in P-1 is the A group in SME 529 is CH3 while the A group in P-1 is H.
906
 REAGENT R A TYPE
OH NOH
LIX 64 C12H25 C6H5 Ketoxime
A LIX 64N C9H19 C6H5 Ketoxime
SME 529, LIX 84-I C9H19 CH3 Ketoxime
LIX 860-I C12H25 H Aldoxime
R
P1, LIX 860N-I C9H19 H Aldoxime

Figure 1: General structure of oxime molecules used for copper recovery.

There is little that can be done to economically produce a molecule with the general structure
above having an A group between H and CH3. Therefore in order to make P-1 and other
aldoximes such as LIX 860N-I and LIX 860-I effective in copper solvent extraction reagent
formulations the extraction œ stripping equilibrium shown in Figure 1 needs to be altered by
something other than a change in the structure of the extractant.
The Properties of Equilibrium Modifiers in Copper Solvent Extraction
In 1977 the P-5000 Series of Reagents were described by Acorga Ltd. [5]. These reagents
combined the strong, difficult to strip extractant P-1 with increasing amounts of nonylphenol to
give a series of reagents, P5100, P5200 and P5300 respectively, where the copper extraction œ
stripping equilibrium was progressively shifted toward the stripping side. Figure 2 taken from
Dalton [6] is a graphic representation showing the decrease in the copper content of a stripped
organic at a fixed P-1 concentration as a function of the nonylphenol/P-1 ratio. Figure 3, also
taken from Dalton, shows extraction isotherms for the same organic phases. The stripping
isotherms were generated with an aqueous strip solution having 30 g/l Cu and 150 g/l sulfuric
acid. The extraction isotherms were generated with an aqueous solution containing 6.0 g/l Cu
at a pH of 2.0.

Figure 2. Strip isotherms for P-1 as a Figure 3. Extraction Isotherms for P-1 as a
function of nonylphenol/P-1 ratio. a function of the nonylphenol/P-1 ratio.

907
Extrapolating from the graphs clearly shows the addition of nonylphenol as a modifier increases
the net copper transfer of the strong copper extractant P-1. For example, consider an SX circuit
having 2 extraction stages and 1 strip stage for the organic phases A, B, C and D, respectively.
Table I shows the copper content for the organic phases A, B, C and D taken from Figure 2
when a pregnant strip aqueous solution containing 45 g/l Cu is obtained in one strip stage.
Table I also shows the maximum copper loading for the same organic phases as taken from the
isotherms for these organic phases in Figure 3. Assuming that organic phases A, B, C and D
respectively, would run at 95%, 90%, 85% and 80% of the copper maximum loading in 2
extraction stages, reasonable assumptions based on the behavior of commercial copper SX
plants, a loaded organic for organic phases A, B, C and D is calculated and given in Table I.
The copper net transfer values shown in Table I is the loaded organic less the stripped organic.
Table I Increase in Copper Net Transfer of P-1With the Addition of Modifier

Organic Stripped Organic Cu Max Load Loaded Organic Net Copper Transfer

A 6.0 g/l Cu 8.0 g/l Cu 7.6 g/l Cu 1.6 g/l Cu

B 4.33 g/l Cu 7.82 g/l Cu 7.04 g/l Cu 2.71 g/l Cu

C 3.01 g/l Cu 7.84 g/l Cu 6.66 g/l Cu 3.65 g/l Cu

D 2.25 g/l Cu 7.75 g/l Cu 6.20 g/l Cu 3.95 g/l Cu

The increases in copper net transfer upon the addition of nonylphenol modifier are substantial,
for example, compare organic A which has no added modifier to organic C which has 140 g/l
nonylphenol added to 70 g/l P-1. The copper net transfer increases from ~1.6 g/l Cu to ~3.65
g/l Cu, an increase of ~128%. The gain in copper net transfer obtained in an operating SX plant
when modifier is added to strong copper extractants such as LIX 860N-I, LIX 860-I and P-1 is
dependent on the stripping conditions, the copper content of the leach solution, the number of
extraction and stripping stages and the copper recovery obtained. Gains in copper net transfer
may be greater or less than those shown in Table I, but, over a range of modifier addition these
gains are economically significant. In many cases the use of an equilibrium modifier allows a
lower extractant concentration to be used to recover the same amount of copper. In other cases
the organic flow can be reduced when using a modifier in combination with a strong copper
extractant rather than using a strong copper extractant without a modifier. Strong copper
extractants are almost always used in combination with either an equilibrium modifier or with a
moderately strong copper extractant such as LIX 84-I.
A close look at Figures 2 and 3 shows there is a gradual tailing off of the improvements to net
transfer as the nonylphenol/extractant ratio increases in the organic phase. At some modifier
concentration the cost for added modifier will be greater than the savings due to the increase in
copper net transfer.
The amount of modifier in P5100 (Organic A in Figures 2 & 3) was well chosen to allow P-1 to
be stripped more efficiently but still remain a strong copper extractant. This allowed leach
solutions with a high copper content and/or a low pH to be efficiently treated in 2 rather than 3
extraction stages. P5300 (Organic D in Figures 2 & 3), which had a nonylphenol/extractant
ratio 3 times higher than P5100, was formulated to compete with LIX 64N for leach solutions
with a relatively low copper content and a relatively high pH in existing plants [5, 14]. P5200
(Organic C in Figures 2 & 3) was formulated for leach solutions with a pH in the mid-range [5].
It was reported that this series of reagents could be used to effectively treat virtually any leach
solution [5], and with respect to copper extraction and stripping, this is correct. In addition the

908
rapid metal transfer kinetics of these reagents meant that at low temperatures these reagents
could be more efficient than LIX 64N.
However, as mentioned above, a successful copper SX reagent must have a good balance of
properties and often a change in one property of the reagent system will result in a change in
another property. For example, the Inspiration copper SX plant started up in October 1979
using P5300, but switched to P5100 as makeup reagent in June 1981 “in order to improve cold
weather phase separation and to handle higher copper and acid concentration feed solutions”[7].
The high concentration of nonylphenol in P5300 gave a viscous reagent and in the authors
opinion the high viscosity caused slow phase separation when the concentration of P5300 was
increased from about 8 volume % in June 1980 to about 20 volume % in May 1981 [7]. In
addition, nonylphenol increases the hydrolytic degradation of the P-1 [8], well controlled pilot
plant studies at an operating agitation leach plant showed that the addition of nonylphenol to a
copper solvent extraction reagent increased the amount of crud formed [9], and nonylphenol is
known to increase the swelling of nitrile rubber, for example, when nitrile rubber is used to line
mild steel pipes.

With respect to copper extraction œ stripping properties, copper transfer kinetics and capital
cost there is no question that the equilibrium modifier nonylphenol is a friend. Nonylphenol
used in combination with P-1 expanded the copper leach solutions that could be effectively
treated by solvent extraction and allowed copper leach solutions to be treated under a wider
variety of conditions, for example, at colder temperatures or at lower pH or with less extraction
stages. At the same time it is the authors opinion that with respect to slow phase separation [7],
increased reagent degradation [8], the potential for increased crud generation [9], and the
incompatibility with some materials of construction nonylphenol can be a foe especially if the
nonylphenol is used at a high concentration.
Nonylphenol was the first equilibrium modifier used commercially in copper solvent extraction
but others have followed including tridecanol [TDA] [10], 2,2,4-trimethyl-1,3-pentanediol
diisobutyrate [TXIB] [11], a highly branched C18 alcohol thought to be isooctadecyl alcohol
[11], and a low viscosity ester [12]. The ability of a modifier to decrease the oxime extraction
strength is dependent on the particular modifier functionality with alcohols being the most
effective. Isotherms similar to those in Figures 2 & 3 can be obtained for all modifiers except
that the amount of modifier will be different.
The properties of modified reagents have been discussed extensively by reagent suppliers, but,
often the work cited in one paper does not agree with the work cited in another paper. For
example, a well controlled pilot plant study showed nonylphenol increased crud formation [9],
and a rational for this has been given [13]. It has also been reported that nonylphenol is
effective in controlling crud formation, however, the studies were not described [14]. It is the
experience of Cognis and others [14] that the conditions under which a study is done, differing
leach solutions and the age of a leach solution can result in different conclusions about the
effect a modifier may have on the properties of a copper solvent extraction reagent.
In addition, it has been the experience of Cognis that the way an equilibrium modifier will
behave with respect to entrainment, phase separation and crud generation in commercial
applications can be difficult to predict based on laboratory studies or even short term pilot plant
campaigns. The literature suggests the same has been found by others.
While there are many properties of modifier behavior on which the literature does not agree
there are features of modifiers that are not in dispute. Modifiers fall into two classes based on
structure, those that can only accept a hydrogen bond and those that can both accept and donate
a hydrogen bond (Table II).

909
 Table II Modifiers Based on Hydrogen Bonding Characteristics

Hydrogen Bond Acceptor Only

Common name General Structure Examples

Esters R(C=O)OR’ TXIB

Ethers ROR

Ketones RR’C=O

Hydrogen Bond Donor and Acceptor

Common Name General Structure Examples

Alkylphenols RC6H5OH Nonylphenol, Dodecylphenol

Alcohols RCH2OH, RR’CHOH Isodecanol, TDA, C18 fatty alcohol

These modifier classes have certain characteristics. In general modifiers that only accept a
hydrogen bond, such as esters, ketones and ethers, give higher Cu/Fe selectivity and increase
the stability of the oxime to degradation by normal hydrolysis. Modifiers that both donate and
accept a hydrogen bond, such as alcohols and phenols, typically show lower Cu/Fe selectivity
and decrease the stability of the oxime to degradation by hydrolysis. It is known that modifiers
interact with oximes via hydrogen bonding [14], but an explanation for the different behavior of
modifiers that only accept a hydrogen bond and modifiers that both donate and accept a
hydrogen bond has not been given. How a particular modifier or modifier class will affect
entrainment, phase separation and crud formation is still open to question and because each
leach liquor is different, plant designs may differ and the way a plant is run can be unique, a
given modifier may behave quite well in one application, but not in another [15].
A plant operational property not often discussed is the stability of the phase continuity in a
mixer. The ability of a mixer to run a single continuous phase results in a smoother operating
plant and usually lower entrainment. In addition unstable phase continuity can lead to double
phasing in a mixer and high entrainment or even phase flipping which can set off a crud run.
The ability to hold phase continuity also has implications in crud behavior. It has been
observed in a number of copper SX plants that running aqueous continuous in a mixer often
gives crud that extends up into the organic phase while running organic continuous often gives
crud that settles at the interface. Crud that extends into the organic phase is carried to the
stripping stages thereby carrying impurities to the electrolyte. There is some evidence that the
use of certain hydrogen bond acceptor only modifiers results in less stable mixer continuity,
especially in the presence of solids, but this behavior does not appear to be universal. The
reason for this behavior may be because of high interfacial tension. An organic phase
containing a copper extractant and a hydrogen bond acceptor only modifier will often show a
higher interfacial tension when in contact with an aqueous copper solution than an organic
phase that contains a copper extractant but not a hydrogen bond acceptor only modifier. At first
glance one might think that a high interfacial tension and the resulting fast phase separation
would be a good thing. However, if phase separation is so fast that appreciable phase
separation takes place in the mixer it seems reasonable this behavior might result in a less stable
continuous phase.
Until 1998 when the Lomas Bayas copper leach/SX/EW plant came on line in Chile the only
type of copper solvent extraction reagent degradation mentioned anywhere was acid catalyzed
hydrolysis. Lomas Bayas, which has treated leach liquor containing 5 to 7 g/l Cu, as high as 35
910
 g/l nitrate and up to 15 g/l chloride, started with the reagent LIX 662N, a strong copper
extraction reagent modified with TDA. Initially the SX plant ran well but in a month or two the
behavior of the SX plant rapidly deteriorated. Analysis of the plant organic showed the oxime
extractant was rapidly degrading to normal hydrolysis product along with several other
products. Cognis set up a series of stability tests where various organic phases were vigorously
stirred at 45ºC for a period of time against a synthetic aqueous phase having 0.5 g/l Cu, 2 g/l
Fe+3, 32 g/l NO3-1, 15 g/l Cl-1, 70 g/l total SO4-2, pH =1.0, emf =720 mV (Ag/AgCl). Under
these conditions LIX 622N degradation was devastating. A second set of stability tests showed
that TDA rapidly increases the degradation of the copper extractant in LIX 84-I, the reagent that
has been proven to be the most stable under Lomas Bayas conditions (Figure 4). Note the
~20% loss in copper loading in about 35 days when TDA is added to LIX 84-I compared to no
measurable loss in copper loading for LIX 84-I by itself.
The ester modifier TXIB also accelerates reagent degradation, but, less aggressively than TDA
when these modifiers are used with the extractant 5-nonylsalicylaldoxime compared to when 5-
nonylsalicyaldoxime is used in the reagent LIX 984N (Figure 5). In Figure 5, the hydrolysis
degradation product in the organic phase is plotted against time.

2
12
LIX 84-I 1.8
10 1.6
TDA
% Aldehyde or Ketone

LIX 84-I + TDA 1.4

8
1.2
Cu ML

6 1
0.8
TXIB
4 0.6
0.4
2 LIX 984N
0.2
LIX 84-I
0 0
0 20 40 60 80 0 10 20 30 40 50
Days
Days

Figure 4. Degradation of LIX 84-I in the Figure 5. Degradation of 5-

nitrate system in the presence of TDA nonylsalicylaldoxime in the nitrate system
in the presence of modifiers
In these tests it can be seen that degradation levels off after a period of time. In the stability
tests shown above the respective organic phase was in constant contact with the same aliquot of
synthetic leach solution. Analysis showed that at the time reagent degradation leveled off the
nitrate content of the aqueous phase was greatly reduced. This implies that the nitrate has a
prominent role in the degradation of the oxime including degradation to the hydrolysis product.
A more detailed discussion of reagent stability in the presence of nitrate will be presented
elsewhere [16].
Through trial and error Cognis has learned that degradation tests must be run very carefully
under tightly controlled conditions in order to get consistent and repeatable results. For
example, something as simple as the way glassware is cleaned can dramatically change the
degradation rate of an oxime.
There is little doubt that both TDA and TXIB are foes when used in copper SX reagents to treat
leach solutions having a high nitrate concentration along with some chloride ion. These data
suggest that all modifiers may be foes in a nitrate chloride system, but, this has not been proven.

911
Cognis New Low Viscosity Modified Reagents
Even though it is generally accepted that reagent concentrations greater than 30 volume %
should be avoided because of viscosity there is very little information in the literature about the
viscosity of a reagent and its behavior in a copper solvent extraction plant. Intuitively it makes
sense that a highly viscous organic phase will have slower phase separation and higher
entrainment and, in retrospect, the fact that P5300 showed slow phase separation in cold
weather in the inspiration copper SX plant at about 20 volume % reagent might have been
because of its high viscosity.
Nonylphenol and TXIB are relatively viscous modifiers, TDA less so. For this reason Cognis
found it interesting to study low viscosity modifiers from the ketone and ether class with the
idea of using the less viscous modifiers in applications where a high reagent concentration is
needed. The reagents, LIX 973N, LIX 6422N-LV and M5640 were selected for a laboratory
study at 40 0C in a continuous mini-plant to treat a copper leach solution in 3 extraction and 2
stripping stages with 45 volume % reagent. LIX 973N, a 70/30 blend of the very strong
extractant LIX 860N-I with the moderately strong extractant LIX 84-I, has no added modifier.
LIX 6422N-LV contains the strong extractant LIX 860N-I in a mixture with a new low
viscosity modifier while M5640 contains the strong extractant P-1 in mixture with TXIB. The
viscosity for these reagents and 45 volume % solutions of these reagents are shown in Table III.
Table III Viscosity of Reagents as Received and at 45% Vol % in Phillips SX-7 Diluent.
Reagent Viscosity in cSt
45 Vol% at 40ºC 45 Vol% at 25ºC As Received at 25ºC
LIX 973N 3.59 5.36 28
LIX 6422N-LV 4.04 6.17 31
Avecia M5640 4.62 7.14 90

LIX 973N has the lowest viscosity both as purchased and diluted to 45 volume % while M5640
has the highest viscosity both as purchased and diluted to 45 volume %. The small difference
in the viscosity of the reagents at 45 volume % might be dismissed as being important in the
performance of the reagents in a solvent extraction circuit, indeed this is understandable.
However, the results of the laboratory study suggest that viscosity may be important, especially
if low entrainment of aqueous in the loaded organic is needed (Table IV). The synthetic copper
solution treated in this study had 14.0 g/l Cu, 14 g/l sulfuric acid and 33 g/l total Fe (2 g/l
ferric), the barren strip aqueous had 33 g/l Cu and 200 g/l acid while the pregnant electrolyte
was built to about 46 g/l Cu. Mixer retention times were 2.5 to 3 minutes with settler residence
times of 5 to 5.5 minutes. The circuits were run and monitored until equilibrium had been
established (5 to 8 hours). Then the circuits were run for several more hours while entrainment
was determined and several profiles taken.
Entrainment of aqueous in the loaded organic was done be carefully collecting a 100 ml sample
of organic into a calibrated glass vessel as the organic stream was entering the loaded organic
surge tank, centrifuging the sample for 20 minutes and reading the aqueous entrainment in the
calibrated vessel. The reagent with the lowest viscosity showed the lowest entrainment of
aqueous in the loaded organic. While the difference in entrainment is not large in absolute
terms such a difference can be important if there are species in the aqueous copper feed solution
that can be detrimental to electrowinning of high quality copper and the concentration of these
species in the electrolyte needs to controlled with an electrolyte bleed.

912
 Table IV Typical Results at 40ºC for Continuous Mini-plant
Evaluations of 3 Copper SX Reagents at 45 Volume % Reagent

Reagent O/A Raffinate Recovery Ave. Aqueous Entrainment

Ext’n g/l Cu % in LO, ppm
LIX 973N 1.32 0.73 94.8 400
LIX 973N 1.19 0.75 94.6 400
LIX 973N 1.1 0.86 93.9

LIX 6422N-LV 1.28 0.73 94.8 450

LIX 6422N-LV 1.20 0.76 94.6

M5640 1.33 0.72 94.8 700

M5640 1.20 0.76 94.6 600
The aqueous entrainment in the loaded organic is quite low in all cases especially for a reagent
concentration of 45 volume %. The use of reagent concentrations of about 45 volume % has
been reported in two instances. Outokumpu has reported the use of 45 volume % reagent in a
specially designed mixer settler unit [17]. The copper SX plant in the CESL Process
demonstration leach-SX-EW plant uses a reagent concentration of ~45 vol % LIX 973N and
reports good success as long as the organic is kept clean by clay treatment and the temperature
across SX is maintained at about 40ºC [18]. This brief lab study and the results reported by
CESL and Outokumpu suggest that reagent concentrations as high as 45 volume % are viable
for use with leach solutions having a copper concentration above about 10 to 12 g/l Cu provided
copper SX will operate at about 40ºC or higher.
In this study the entrainment of organic in the raffinate and the pregnant electrolyte was at the
trace level in all continuous mini-plant runs. Thus the viscosity of the organic phase does not
appear to negatively affect the entrainment of organic in the aqueous phases. But with respect
to entrainment of aqueous in the loaded organic high viscosity modifiers may be a foe at high
reagent concentrations.
LIX 612N-LV which contains the same low viscosity modifier as LIX 6422N-LV, but a slightly
greater amount, has been added as makeup to a commercial plant since January 2002. Prior to
then the plant had been operating on LIX 984N for a number of years, but a change in operating
conditions made copper/iron selectivity much more important and the reagent change was made
in order to improve copper/iron selectivity. As expected Cu/Fe selectivity has improved with
the addition of LIX 612N-LV. It is difficult to judge how LIX 612N-LV is behaving with
respect to entrainment, crud generation and phase separation when compared to LIX 984N
because the circuit is currently running under much different conditions that it had been before
the reagent change. But it can be said that entrainment, crud generation and phase separation
for this new low viscosity modified reagent are within the expected range for a plant running
under the new conditions and the plant has not reported any unexpected operating problems.
Stability tests currently in progress in the Cognis laboratory show the new low viscosity
hydrogen bond acceptor only modifier in the reagents LIX 612N-LV and LIX 6422N-LV
increases the stability of the aldoxime to degradation by hydrolysis to the same extent, if not
more, than other hydrogen bond acceptor only modifiers such as TXIB.

913
Ketoximes as Modifiers for Aldoximes?
The moderately strong ketoxime extractant LIX 84-I has been called a copper chelating
modifier when used in a blend with aldoximes such as LIX 860N-I [14]. While this may be
true to some extent, in that it appears the weaker extracting ketoxime lowers the strip point of
the stronger extracting aldoxime thereby increasing the copper net transfer to some extent [19],
with respect to other key modifier properties the aldoxime/ketoxime blends tend to be in a class
by themselves. For example, they are slightly slower kinetically than modified aldoximes and
they are less selective than aldoximes with hydrogen bonded acceptor only modifiers, but under
a variety of conditions they tend to give more stable mixer continuities, lower entrainment of
aqueous in the loaded organic and less crud than modified aldoximes [9]. For these reasons the
ketoximes should not be considered modifiers for aldoximes and the LIX 84-I blends with
aldoximes such as LIX 860N-I or LIX 860-I should be considered as a separate reagent class.
In addition, the aldoxime/ketoxime blends are the least viscous oxime copper SX reagents
available. There are cases where low viscosity can be very important. For example, consider
copper SX plants where a strong copper extractant is not needed, that is, with leach solutions
having a relatively high pH and/or where a very high copper recovery is not needed. If a
modified aldoxime were to be used the recommended aldoxime would contain a large amount
of modifier in order to maximize copper net transfer and if the modifier was viscous, for
example nonylphenol or TXIB, the resulting reagent would be viscous. Thus the high viscosity
of the reagent would to a certain extent limit the concentrations at which these viscous reagents
can be used. Inspiration’s experience with P5300 cited previously serves as an example.
If an aldoxime/ketoxime blend were to be used it would be a blend with a high proportion of
LIX 84-I since this blend would strip very well to maximize copper net transfer. Since these
blends are much less viscous they would have less limitations on the concentration at which
they can be used.
As shown in Figure 5, the aldoxime/ketoxime blend LIX 984N is much more stable to nitrate
containing leach solutions than is the aldoxime modified with TDA and TXIB. Furthermore
aldoxime/ketoxime blends such as LIX 984N show only about 6 % reagent degradation after
one year in a typical heap leach application so the ketoxime does not appear to decrease the
stability of the aldoxime to any measurable extent. Degradation tests run in the Cognis
laboratory shows that the aldoximes blended with ketoxime have stability about equal to the
stability of aldoximes modified with hydrogen bond acceptor only modifiers.
Conclusions
The title of this paper poses the question are equilibrium modifiers are a friend or foe in copper
solvent extraction reagents. Clearly they are both depending on the particular property of
interest. Modifiers are beneficial with respect to increasing the copper net transfer of an
aldoxime extractant and are always a friend from this point of view. The use of hydrogen bond
acceptor only modifiers results in superior copper over iron selectivity. In most cases this is
beneficial and these modifiers are a friend. In some cases selectivity can be too high and
ferrous sulfate needs to be added to the tankhouse in order to have enough iron in the electrolyte
to control the oxidation potential of the electrolyte in a proper range. In this case hydrogen
bond acceptor modifiers are foes, but a modifier like TDA is less selective and would be a
friend in these cases.
In the instances when modifiers result in a less stable mixer continuity, or more crud or higher
entrainment of aqueous in the loaded organic phase they are clearly foes while the LIX 84-I
blends with aldoximes appear to be more friendly in this area.

914
The hydrogen bond acceptor only modifiers appear to be friends with respect to aldoxime
stability to acid catalyzed hydrolysis while the hydrogen bond acceptor and donor modifiers are
foes. However with respect to stability to nitrate containing leach solutions the modifiers TXIB
and TDA are foes while the LIX 84-I/aldoxime blends are much more friendly with the nitrate
containing leach solutions.
The low viscosity modifiers and the aldoxime/ketoxime blends may be friendlier than the high
viscosity modifiers in applications where a high reagent concentration is needed, for example,
with concentrate leach solutions or when the temperature is low.
All commercially successful copper solvent extraction reagents have a good balance of the
needed properties and no single reagent is best for all copper leach liquors or all copper solvent
extraction plants. However, determining which reagent is best for a given situation is not
always easy. McCabe-Thiele diagrams constructed on extraction and stripping isotherms will
predict with good accuracy the copper extraction and stripping behavior of reagents. Long term
pilot plant tests if well run with a truly representative leach solution should give comparative
data on some of the physical properties of reagents such as entrainment, phase separation and
crud generation. Stability tests if done carefully under well controlled conditions ought to give
comparative data on reagents.
However, the reality today is that few end users enter into a long testing program on copper
solvent extraction reagents relying instead on suppliers to provide needed information on
copper solvent extraction reagents. As long as the leach solution does not contain anything
unusual that will cause unanticipated problems this approach usually works well. If there is
something in the leach solution that causes unanticipated problems after the plant is operating
the operator does have the option to consider other reagents in an effort to solve or at least
minimize the problem.
The fact that the plant operator has a broad range of modified and nonmodified copper SX
reagents to consider and from which to choose, the fact that a wide variety of leach solutions
can now be successfully treated and the fact that modifiers spurred development of the
aldoxim/ketoxime blends leads to the final conclusion that overall modifiers are a friend.
References
1. K.L. Power, ”Operation of the First Commercial Liquid Ion Exchange and Electrowinning
Plant”, (Paper presented at the AIME Annual Meeting, Denver, February 1970).
2. Ashland Chemical Company, “Copper Solvent Extraction Reagents” (Preliminary Technical
Data Sheet, 1968).
3. J.A. Hartlage and A.D. Cronberg, "Chemical and Physical Factors to be Evaluated Upon
Designing a Kelex Extraction System", (Paper presented at CIM Conference of Metallurgists,
Quebec City, August 1973).

4. C.P. Birch, "The Evaluation of the New Copper Extractant P-1", Proceedings of the
International Solvent Extraction Conference, 1974, (ISEC '74), 2837-2871, SCI, London.

5. J.A. Tumilty, R.F. Dalton and J.P. Massam, "The Acorga P-5000 series: a novel range of
solvent-extraction reagents for copper" Advances in Extractive Metallurgy 1977, ed. Jones M.J.,
The Institution of Mining and Metallurgy, London 1977, pp 123 - 131..

6. R.F. Dalton (1977), “The Effect of Alkyl Phenols on the Copper Transfer Properties of the
Extractant Acorga P-1”, Proceedings of the International Solvent Extraction Conference, ISEC
’77, CIM Special Volume 21, 40-48.
915
7. J.O. Houseweart (1982),”Inspiration Solvent Extraction Plant”, Paper presented at the 1982
Spring Hydrometallurgical Session, Arizona Section of A.I.M.E., June 5, 1982.

8. “Acorga P-5100 and Acorga P5300: Long Term Laboratory Evaluation of Hydrolysis”,
(Acorga Mining Chemicals Technical Information MC.180).

9. G.A. Kordosky, J.M. Sierakoski and J.E. House, “The LIX 860 Series: Unmodified Copper
Extraction Reagents”, Proceedings of the International Solvent Extraction Conference,
ISEC’83, pp. 191-192, A.I.Chem.E.

10. Tridecanol was first commercialized in LIX 622 at the Pinto Valley SX plant (1981).

11. R.F. Dalton et al, U.S. Patent 4,978,788.(1990)

12. M.J. Virnig et al, U.S. Patent 6,177,055 (2001).

13. M.J. Virnig et al., “Crud Formation: Field Studies and Fundamental Studies”, Fourth
International Conference Copper 99-Cobre 99, Volume IV: Hydrometallurgy of Copper, ed.
S.K. Young, D.B. Dreisinger, R.P. Hackl and D.G. Dixon (Warrendale, PA: The Metallurgical
Society of AIME, 1999), 291-304.

14. R.F. Dalton, K.J. Severs and G. Stephens, “Advances in Solvent Extraction for Copper by
Optimized Use of Modifiers”, Paper presented at IMM’s Mining Latin America Conference,
November 1986, Chile, 67-75.

15. R.F. Dalton and G.W. Seward “The Role of Aryl Aldoxime Reagents in the Technical and
Economic Advancement of Copper Production by Solvent Extraction”, Proceedings of the
International Solvent Extraction Conference, ISEC ’86, Volume II, 11-18, Munich.

16. M.J. Virnig et al., Paper to be presented at Cobre / Copper 2003, Copper International
Conference, Santiago, Chile. November 2003
.
17. P. Pekkala, et al., “Flexible Outocompact Copper SX Track”, ALTA 2002 Copper -7,
ALTA Metallurgical Services, Perth Australia, May 2002.

18. D. Jones, Comments made at ALTA 2002 Copper-7, ALTA Metallurgical Services, Perth
Australia, May 2002.

19. G.A. Kordosky et al, US Patent 4,507,268 (1985).

916