Smith, R.M., Tyron, E.H., and Tyner, E.H., 1971, "SoilDevelopment on Min~ Spoil," Bulletin 6O4T, July. Agr. Exp.Sta.. West Virginia Univ.. Morgantown.

Spindler, D., 1979. "Moist Bulk Density Values," personalcorr~spond~nce. Lapd Recl. Div.. Dept. of Mines and Minerals,Illinois.

Fig. 4-Corn Root. Site 3

D. R. Nagaraj .and P. Somasundaran

(either as a chelate with dissolved copper or as a chelate detached from the surface),

b) the bulk chelate does not aid flotation of the mineraland,

c) the solubility of the mineral largely controls the extentof partitioning of the chelating agent and, hence, flotation.

Experimental

Materials

Abstract-Based on our finding that cemmercial copperchelating solvent extractants such as LIX6.5NR and LIX6JRareexceUent coUectors for copper ma'nerals, a detat'led study wasmade USJ'ng several water-soluble chelating agents belonging tothe class of aromatic hydroxyoximes and having a gradual struc-tural variation, A tJasic study of the system salicylaldoxime-tenorlte was also ca~"ed out, MIcro flotation tests as a functionof concentration of OXI~, pH, addition of copper, etc, , usingo.¥idi(' copper ma'nerals such as chrysocoila and tenorite, andb~ch laboratory jlot~ion tests using $)mIhetic mixtures of quar-tz and oxidic copper minerals were carried out.

The results have demon$trated that the hydroxyoximeistudied have comparable collector property, and irJ ail case$ theoxz'me was found to be partitioned between the mmeral surfaceand the copper m the bulk aqueou.sph4u in theform ofa bulkcopper chelate, The occurrence of the latter could tJe accountedfor by formation m the bulk phase of a chelate between oximeand copper releasedfrom the 11U'neral and/or b)' the detach-ment of the surface chelate. The role of surface and bulk"chelates in jlotatlorJ is discussed based OrJ the results of par-titioning of chelating agent into bulk arJd surface chelatel m thesalicylaldoxime-tenorite $)lItem, The effect of structural aspectssuch as isomerism and substitution in the molecule OrJ the coUec.tor property of the oximes is also discussed. SubstitutIon OrJoximic carbon by alkyl groups is found to promote coUector ac.tl"vit}', On.ly the anti-isomer has the ability to collect,

Introduction

Oximes: First, ~ survey of all the currently marketedhydroxyoxime solvent extractanU was made to find any Str~'tural variations between the reagents warranting a detailed in-vestigation of structure vs. collector property. Theft reagentswere used in the preliminary investigation after purifying thecommercial samples by chemical and liquid chromatographic(Prep LC, Waters Associates) methods and purity checked bythin layer chromatography (TLC) and infrared Jpectr~ (IR).The results of chrysocona flotation demonstrated that the com.mercial chelating extract ants were not suit~ble for the St~cture'colleCtor activity study because:

. The reagenu are watec.insoluble. Problems were encoun-tered in obtaining representative emulsions of theseoximes inwater.

. The benzophenone oximes are not monoisomeric, mutualtransformations between isomers being sensitive to a large num-ber of variables.

. The Cu chelate of these oximes is not a well defmed solid,unlike chelates formed by lower oximes such as salicylaldoxime.This becomes an important consideration in the separation ofbulk and surface chelates.

For these reasoqs, a series of sparingly water-solubleoximes-shown in Table I-was selected. Except for SALO andOHCHO, an the oximes were prepared from their respectivereagent-grade ketones essentiany by the methods of Blatt (1955)

D. R. Nagaraj, member SME, is with American Cyanamid Co.,Stamford, CT. P. Somasundaran is a professor at the HenryKrumb School of Mines, Columbia University, NY. SMEpreprint 79-76, SME-AIME Annual Meeting, New Orleans, LA,Feb. 1979. Manuscript Jan. 2, 1979. Discussion of this papermust be submitted, in duplicate, prior to Nov. 30, 1981.

Hydroxyoxime chelating agents such as the LIX series ofreagents are now widely used for commercial solvent extractionof copper from leach liquors. A new application for thesesolv~nt extractants-as fIotaids for copper mineralflotation - was discussed in earlier papers (Nagaraj and

Somasundaran, 1976 and 1977).In this paper, results of the investigation of structure vs.

collector property of water-soluble aromatic hydroxyoximes forchrysocolla flotation (both microflotation and batch laboratorytests) are discussed.

The mechanism of collector action of hydroxyoximes hasbeen discussed, based on the study of the tenorite-salicyladoxime system and in the light of findings that:

a) the chelating agent is partition~d between the mineralsurface (to form a surface chelate) and the bulk aqueous phase

MINING FNGINEERING

Table 1- The Structures and Water-Solubility ofHydroxy Oximes

WaterSoI..-

~

2Xl0-l

I StockSolution

M

10-2

~

SAI.O (Salicylaldoxi_) tit.!

(»W'O (O-Hydroxy

Acetophenone Oxi8e) 151.' :

!»tBuPO (O-Hydroxy Butyro-

phenone OXi8e),0110'1

179.i.

-4.SX10-4

62X10(1.2\acetone)

(llBePO CO-Hydroxy -=-

phenone Oxi_) 1.SXIO~

and Kohler and Bruce (1931) and recrystallized from benzene orpetroleum ether containing a small amount of benzene and thepurity checked by TLC (methylene chloride with 1 %isopropanol as the mobile phase.)

Salicylaldoxime from LaChat Chemicals was recrystallizedfrom petroleum ether containing benzene.

The standard methods of oximation failed to give the oximeof o-hydroxycyclohexanone (OHCHO). It was prepared essen-tially by the method of Nenz, et al. (1964) by reactingcyclohexene and nitrosylhydrogen sulfate in cyclohexane below-50 C. Crude OHCHO waS found by TLC to contain only theanti-isomer and was used as-is.

The water-solubility at 25-26°C of the oximes was determined(Nagaraj. 1979). Solubility values in terms of molar concen-tration of the saturated solutions are also given in Table 1.

For the investigation of the mechanism of collector action ofoximes, SALO was selected since it was water-soluble (solubility2 X 10 -1M). monoisomeric, and is one of the better studiedhydroxyoximes with regard to its chelating and other chemicalpropenies and the structure and stoichiometry of its Cu chelate.which is a well-defined crystalline solid.Minerals: Chrysocolla used in earlier work was also used for thestructure-collector property study, except that the size was- 300 + 150 11m (- ,!8 + 100 mesh)in the present investigation.Tenorite was used for the basic study using SALO. Its prepara-tion is described in detail elsewhere (Nagaraj, 1979). Briefly.natural cuprite (20%) from Kennecott was upgraded to about92% Cu20 by heavy liquid separations using tetrabromoethane(S.G. 2.95) and thallous format malonate (S.G. 4.3) and. afterthoroughly washing the mineral free of the above liquids, washeated in a fumace at 510°C for 100 hours to give -212 +106 11m (- 65 + 150 mesh) CuO analyzing 70.7% Cu, and6.7% insolubles. No CU20 was detected in the X-ray diffractionpattem and the infrared spectrum.

Method

2HSMeAPO (2-liydroxy-S.

Methyl Ace~

OXiM)

166,2 J~o-4

2HSMBAO (2-Hydroxy-S-

Methoxy Benuldoxi8e)

2lVlAO (2-liydroxy-l-

Napl\tbaldox~) 189.2 : l.m10-4

~::

~i"'CII.3HO -

~~...0I2a12C113III -

~irQJIII -

.a:"'al3¥;III -

~--r-"~~~,:"""

~

(a10I0 (O-Hydroxy Cy"lo-

hexanone Oxille) c~lSIlO

Structure vs. Collector Property: Microflotation Tests-Thesetup used for reagentizing is shown in Fig. 1. Chrysocolla (0.7g) was reagentized by agitating with aqueous collector solutioncontaining 1 % acetone (added to obtain smaller bubbles and amore uniform bubble size distribution) and at the required pHand ionic strength (1.5.); pH was continuously monitored byadding KOH or HNO3 from a Dosimat buret and recorded on astr:iP chan recorder. The reagentized mineral was floated in aHallimond tube for one minute at 20 mL/min nitrogen.

Batch Laboratory Cell Tests-Mixtures of chrysocolla withquanz were prepared by wet-grinding in a laboratory porcelainball mill at 60% solids by weight. and wet-screened to obtain- 212 + 44,no (- 65 + S~5 mesh) fraction assaying approx.imately 1.7 % Cu. Some ~50 g of this fraction was conditioned at1500 rpm for four minutes with 1 L water in a D-I Denver Lab-oratory cell, and pH was continuously monitored at 4.8. Col-lector solution was then added. pulp made up to I.~L. andreagentized for four minutes at 1800 rpm and pH 4.8. Frother(DOWFROTH 250) was added toward the end of the reagentiz-ing period. .The pulp was floated to completion. and the con.centrate was cleaned after conditioning it for one-half minutewith the supernatant from rougher flotation makeup water.More frother was added when necessary.

ylNC»! 129

HO ~-

Tenorite-salicylaldoxime System: Microflotation Tests-Fromthe initial tests it was observed that the pH of the aqueous SALOsolution decreased very sharply upon adding the mineral owingto chelate fonnation (by deprotonation) between SALO andcopper species on the mineral surface as well as those releasedfrom fhe mineral into the bulk aqueous phase:

[CuOJminl + [Cu-speaesJaq + xSALO-{Cu-SALOJlninl +[Cu-SALOJbulk + yH+

The pH decrease was m~t pronounced when the initial pH ofthe SALO solution was between 4.5-8.5.

The rapid and large pH decrease (followed by an increase)occuring in the CuO-SALO system necessitated rapid and con-tinuous monitoring of pH during reagentizing. 0.8 g of CuOwas reagentized with 94 mL of SALO solution (containing1 % acetone) at an I.S. of 2 X 10 - 3M (KNO3) and constant

pH, and floated for one minute.Partitioning of SALO- The possibility of partitioning of

oximes into surface and bulk chelates was discussed in earlierpublications. In the present investigation these chelates havebeen separately determined quantitatively for the SALO-Tenorite system, using the scheme shown in Fig. 2, by extractingthe chelates with chlorofonn, which, among all the solventstested, had the highest solubility for chelates. Briefly. the testsinvolved the following steps:

a) The mineral was reagentized as described.b) The supernatant with the bulk chelate dispersed in it was

separated from the mineral bed by mixing, allowing the mineralto settle (the S.G. difference between mineral and chelate ismore than three units) and decanting the supernatant.

c) Step (b) was repeated using fresh 2 X 10 - 3M (KNO3)

until all the bulk chelate was separated from the mineral bed.d) TIle supernatant from steps (b) and (c) was filtered, bulk

chelate was washed with water and extracted with chlorofom1(to eliminate any mineral slimes reponing as bulk chelate); thesolvent was then evaporaed and the bulk chelate weighed.

MAGNE.TIC

STIRRER

Fig. 1-The experimental setup for mineral reagentizing.

la~ PoFPTFURFR 1981UINING &;:NGIN&;:&;:RING

PARTtTIOfiING OF SALO -. THE EXPERIMENTAL SCHEME

MINERA~

WASHINGSSUPERNATANT

I WASH MNERAL

BED OFF BULK

CHELATE

FILTRATEI FILTER SUPERNATANT

CONTAINING BULK

CHELATE

I CHCI, EXTRACTIONi OF CHELATE ONMINERAL

WASH CHELATE iON FILTER PAPER' RINSE ELECTRODE

AND MAGNETICBAR WIT.H CHCI3

CHCI3

CHC/3 EXTRACTION IN A

SEPARATORY FUNNELI (EVAPORATE

CHCI3) SAMPLE FOR pH.!Cu2+ AND SALO :ANALYSIS

(EVAPORATECHCI5)

WEIGHCHELATE

WEIGH!CHELATE

MEASUREDH

u.~ SPECTRO-PHOTOMETRICANALYSIS:OF SALO AT ipH 1.5 IN HClo.

I SURFACE

CHELATE

ADJUST pHTO 3.0

I ~TERMINE CUZ.jUSING COPPER

ELECTRODE

c) Substitution in the 5-position on the benzene ring, (Rl =CH~, R2 = CH~--2H5MeAPO and Rl = H, R2 = OCH~--2H5MBAO) drastically affects the collector efficiency; both2H5MeAPO and 2H5MBAO are inferior to SALO.

d) A fused aromatic ring (naphthalene), instead of the ben-zene, again lowers the collector efficiency, thus 2HNAO is in-ferior to SALO.

e) Substitution of a phenyl for Rl in (I) decreases both thecolleCtor efficiency and water-solubility; thus, OHBePO is in-ferior to even SALO.

f) OHCHO could be considered to belong to the class of (k-acyloin oximes, although the oxime is distinct because of thecyclic alkane. OHCHO was found to be highly soluble in water.Its chelate with copper was soluble in slightly acidicsolutions-no chelate probably formed below pH 4.0-andprecipitated in the basic pH range. OHCHO was a poor collec-tor for chrysocolla even at high concentrations (Table 2).

For the oximes OHBePO, 2H5MeAPO, and 2H5MBAO, itwas observed that bubbles during flotation were smaller than forthe other oximes at similar concentrations.

Compared with SALO, the collector efficiency for OHBuPOand OHAPO is higher at almost all the pH values, as can beseen from Fig. 4. Flotation of chrysocolla is maximum aroundpH 5.0 for these oximes and for 2HNAO. On the other hand,flotation with OHCHO increases with increase in pH from ~.O toabout 6.5 above which it remains constant.

Results of batch laboratory flotation tests presented in Fig. 5confirm the improved collector efficiency of OHAPO overSALO for chrysocolla flotation fTom its synthetic miXtures withquartz. The grade of chrysocolla concentrate using OHAPO isbetter than that using SALO, which suggests that OHAPO maybe more specific than SALO.

Salicylaldoxi~- Tenonte System

Microflotation Tests: Flotation behavior of tenorite as a funCtionof initial concentration of SALO at pH 4.8 and I.S. of 2 X10 - ~M (KNO~) and 10 minutes reagentizing is shown in Fig. 6.

As can be seen from Fig. 6, flotation of tenorite is complete atabout 5 X 10 - ~M SALO, and minimum at approximately ~ X10 - 4M SALO. It must be noted that in all tests Cu-SALO

chelate was found dispersed in the bulk phase, the amount in-creasing with concentration of SALO and reagentizing time.

The effect of pH on tenorite flotation at 7.5 X 10 - 4MSALO, I.S. = 2 X 10 - ~M and 10 minutes reagentizing is

shown in Fig. 7. Flotation of tenorite attains a maximum at pH2.5 and 9.0, and a minimum at pH ~.5.

Fig. 2- The experimental scheme for quantitative determina-tion of partitioning salicylaldoxime into bulk and surfacechelates.

e) The chelate on mineral was extracted by vigorously mixingthe mneral bed (after c) with chloroform and repeating this stepwith fresh chloroform. The solvent was evaporated and the sur-face chelate weighed.

Results

Chrysocolla-Oximes

InitialConceftot..tion, M

to-55 x 10-5

10-42 x to-4

10-3to-21,0-1

AeagentlzlngTime. Min. pH

10 48

%Fl081ed

15.015.520.012.09.55.59.0

Oxime

Chrysocolla flotation as a function of concentration of theoximes is shown in Fig. 3 and Table 2. The effect of pH onchrysocolla flotation is shown in Fig. 4 for SALO. OHAPO.and OHBuPO. The results of batch laboratory flotation aresummarized in Fig. 5 for SALO and OHAPO.

The following become evident from the results shown in Fig. 3and Table 2.

a) Substitution of .CH3 (to give OHAPO) in place of -H in Rl(seel) causes a large increase in the collector efficiency; concen-tration of OHAPO is about half that of SALO for 60%flotation. The water-solubility of the oxime is lowered by al~osttwo orders of magnitude upon -CH3 substitution.

OHCHO

10-5 4.82 x 10-4- 2.8

, 3.84.84.85.25.755.86.77.06.49.7

30 4.8By improved technique of solubilizing OHBePO.1.5 x 10-: 10 4.81 x 10-7.5 x 10-55 x 10-52.25 x 10-5

1010

14.529.233.818.824.845.350.480.843.436.519.118.557.4

(I) OHBePO

b) Increasing the chain length of RI by two more -CH2groups (to give OHBuPO) further improves the collector ef-ficiency. but not to the extent as in (a); the water-solubility ofOHBuPO is only an order of magnitude less than that ofOHAPO.

29.735.732.520.713.7

MINING ENGINEERING SEPTEMBeR 1981 1353

at>-a:I&J>0UI&Ja:

zQ

~~0-J

Fig. 3-Chrysocolla flotation as a function of oxime con.centratlon.

Fig. 4-Chrysocolla flotation as a function of pH at a fixedoxime concentration.

Fig, 5-Batch laboratory cell flotation of chrysocolla fromsynthetic mixtures with quartz using salicylaldoximeand o-hydroxy acteophenone oxime.

Partitioning of SALO: The total and individual amounts of bulkand surface chelates deteTUlined as a function of concentrationof SALO and pH, under the same conditions as those used formicroflotation tests, are shown in Figs. 8 and 9, respectively.

The amounts of bulk and surface chelates at 5 X 10 - 3M

SALO (Fig. 8) are only apparent values since tbe bulk chelateformed was very coarse and settled as fast as the mineral,making its complete separation from the mineral almost im.possible. The points at 5 X 10 - 3M are, therefore, joined by

dashed lines. The actual amount of bulk chelate should behigher, and that of surface chelate correspondingly lower, thanthe apparent value shown in Fig. 8. These apparent values werecorrected for using the same surface to bulk chelate ratio as ob.tained for 2 X 10- 3M SALO, and the corrected values are alsoplotted (asterisks) in Fig. 8. It is to be noted that the flotationbehavior (Fig. 6) is correlated very well by the results in Fig. 8,panicularly for surface chelate.

The following become evident from the results in Fig. 9:a) the amounts of surface chelate correlate very well with the

flotation behavior (Fig. 7),b) the total amount of chelate foTUled in the system reaches a

maximum at pH 3.5 and decreases sharply below pH 2.5 andabove 9.5, and

c) the curve for total chelate closely follows_that for bulkchelate.

1354 SEPTeMBER 1981 MINING ENGINEERING

-- -

CuO - SAlO SYSTEM cc~

pH 4B, 10MIN REAGENTIZING ;~

IS' 2.10-3 (KNO31 ~

:J:W~

f.

-H«-CHS<-CH~2CHS

20..

17.5C.O-SALO SYSTEM" .7S.10-4M SAlO "'

.,-. RE~EHTIZIHG ;I (089 C.O) 15,.!!I

1;

S,:

2.5

100

~~~

eo

QIII... 6049..

If

~~

20

~

,,- - . . .~~~ . 1 ... .10~ 10-4 10-3 ~-t

INITIAL CONCENTRATION OF SALO. ~

Fig. 6- Tenorlte flotation as a function of sallcylaldoximeconcentration.

Discussion

ChrysoooUa-oximes Systems

In formula (I) when R2 = Hand Rl = - H, - CHS or- CH2CH2CHS the apparent effect is the increase in the collec-tor efficiency in the order

~

This increase in collector efficiency with increase in chain lengthis in general agreement with resultS of Fuerstenau. Healy. andSomasundaran (1964) for quartz flotation using amine collectorsof increasing hydrocarbon chain length. As can be expected. in.troducing the first - CH2 in SALO results in a much largerchange in polarity of the molecule compared with funher ad.

dition of two - CH2 groups. This is reflected not only in the

water-solubility of the oximes-SALO, OHAPO and OH-BuPO - but also in the collector efficiency. Furthermore, theelectron donating tendency of - CH2 group may also favor

chelation reaction.Substitution of a phenyl group on Rl (R2 = H) will make the

molecule much less polar than SALO as reflected in the water-solubility. On the basis of this, OHBePO can be predicted tohave a higher collector efficiency than SALO. The actual fin-ding was, however, contrary to the prediction, and this can bedue to the possible steric hindrance to chelation offered by thesecond benzene ring and the slower kinetics of adsorption ofOHBePO. Recent adsorption tests (Nagaraj. 1979) have in\dicated that adsorption of OHBePO on chrysocolla is significan-tly higher than that of SALO at the same oxime concentration,thereby suggesting that steric hindrance or slow chelationkinetics may not be responsible for the lower collector efficiencyof OHBePO compared to that of SALO. Bubble size reductionobserved during flotation is, therefore, a major factor-

Substitution in R2 (see formula I) by - CH! (Rl = CH! or

H, re&pec;tively) not only decreased the water-solubility of theparent oXimes, but also decreased the collector efficiency. Thedecrease in water-solubility is to be expected from the decreasein polarity of the molecule as a result of substitution. Thedecrease in collector efficiency could be attributed to thedecrease in acid-strength (logPKH, the proton-ligand stabilityconstant, Irving &. Rossotti) of the molecule owing to the in.crease in electron density on the phenolic oxygen caused by thenucleophilic substituents (Patel and Patel, 1970). This can,however, increase the metal,ligand stability as is generally ob-served (Martell and Calvin, 1952). The effect of electron-releasing groups on the ring is, therefore, generally favorable forchelation. Again, the reduction in bubble size appears to be themajor factor responsible for the lower collector efficiency. Thisappears to hold for the low collector efficiency observed in thecase of OHNAO, although the solution concentrations studiedwere severly limited by its low solubility.

As could be expected from the high solubility of OHCHO and

.,.".." """"I, t""'!:;';'~./:

CoO - SALO SYSTEM

,H ~8, IOMIN R£AGENTIZING-3I-S-2.1O (KNO31

~

-, , r.1

~~

to r TOTAI.£/

CH£LATI ...*

~~

80

wa

~ 12.0-C

~U

0 10."-'c.".~

u

tM

~

J<.

~

~

0w

..

~~

of

iif

~~

If~

*.I~'0 ~..~~_.~\ /../

~

..4'Jy !

~~

~/

~

, "I , , ,y

Ut S . 5 6 7

~H DURING REAGENTIZING

0 . =--.{:r-'":~""I.' 10-5 10-4 10-3 10-2

INITIAL CONCENTRATION OF SALO. M

Fig. 8- The extent of partitioning of salicylaldoxime into bulkand surface chelates as a function of salicylaldoxlmeconcentration.

Fig. 7- Tenorite flotation as a function of pH at a fixedsalicylaldoxime concentration.

MINING ENGINEERING SEPTEMBER 1881 13&6

chelate. This indicates that the surface chelate is responsible forflotation. provided the bulk chelate does not contribute toflotation. Results of previous investigation showed that bulkCu - LIX65N chelate had no ability to collect. and the LIX65Nassociated with bulk chelate was essentially unavailable forflotation. These results were also confirmed for SALO-CuOSystem (Nagaraj. 1979).

In highly acidic solutions. chelation of Cu by SALO is notcomplete and even the chelate that is formed is highly soluble(Biefeld and Howe. 1959). As the pH is increased. say. from 1.0.Cu-SALO chelate starts precipitating and at pH"'!.6 theprecipitation is complae. The pH of quantitative precipitationis dependent on the nature and composition of the aqueousphase. In the present system. the total amount of Cu.SALOchelate increased with pH. reaching a maximum value atpH"'~.5. Between pH 1.5 and 5.5. therefore. lies theprecipitation edge for the chelate. A similar situation exists inthe basic pH region: a second precipitation edge exists betweenpH"'9.0 and up to. possibly. 10.5 (see total Cu-SALO chelate.Fig. 9). Along the precipitation edge. the kinetics of chelateprecipitation are expected to be slow unless a nucleating centersuch as the mineral surface exists to favor chelate formation(precipitation). This can explain the increasing amount ofchelate on mineral surface with increase in pH from 1.5. eventhough the mineral has very high solubility in this pH range(Nagaraj. 1979). Simultaneously. the amount of bulk chelateincreases. since conditions become more favorable in the bulkphase for precipitation of chelate as the pH is increased. Abovethe pH of quantitative precipitation. bulk chelationpredominates (because of the high concentration of copperspecies released from mineral) over surface chelation, whichreaches a maximum followed by a decrease. Bulk chelationshould also reach a maximum (and surface chelation aminimum) since the solubility of the mineral continuouslydecreases with increase in pH (Nagaraj. 1979). The maximumin flotation (pH"'!.5) corresponds to the maximum in theamount of surface chelate, and the minimum (pH ~.5)corresponds to both a minimum in amount of surface chelateand maximum in bullt chelate.

its copper chelate. it exhibiu poor collector efficiency. Thelower solubility of the copper chelate in the basic pH rangesuggests the possibility of an increase in flotation of chrysocollaby OHCHO with increase in pH. and this. in fact. was observed(Fig. 4). It would be interesting to test the effect of substituenuon the collector property of this clus of oximes. as there is no in-fonnation in the literature. For example. introduction of one ortwo CH2 groups on the ring might sufficiently decrease thepolarity as well as the solubility of both the oxime and its copperchelate in order for the oxime to be an efficient collector.

It is to be noted that except OHBePO. all the oximes weremonoisomeric; only the anti-isomer was present. Oximes such asOHBePO with an additional phenyl group have both the anti-and syn-iIOmen with comparable stability. and the mutual trans-formation being sensitive to temperature. pH. etc. For theoxime to function as a coUector. it must be in the anti fonn. Synisomer does not function as a collector. This finding wasdiscussed earlier for LIX65N (Nagaraj and Somasundaran.1977).

Salicylaldoxime- Tenorite System

The results of the earlier investigations of the systemsLIX65N-CU20 and LIX65N-chrysocolla gave a qualitativeindication that UX65N was partitioned: partly on the mineralsurface - aiding flotation - and partly in the form of copperchelate dispersed in the bulk. This has been confirmed for theSALO-CuO system in the present Study and, in addition, thesurface and bulk chelates have been separated and quan-titatively determined.

The results of Figs. 6 and 8 are plotted together in Fig. 10,and similarly those of Figs. 7 and 9 are plotted in Fig. II, in or-der to clearly show the correlation and to better underStand therole of partitioning of SALO.

It is clear from Figs. 10 and 11 that flotation follows th~ sametrend as the extent of SALO abstracted in the form of surface

1356 SEPTEMBER 1981 ~E~EERfiG

chelates has been quantitatively determined in order toelucidate the role of these chelates in flotation of tenorite andthe mechanism of adsorption of salicylaldoxime on tenorite.Results indicated that flotation correlates very well with the ex-tent of SALO abstracted in the form of a chelate on the mineralsurface. Solubility of the mineral has a large influence on the ex-tent of bulk chelation. The bulk chelate has no role to play incontributing to flotation since it has no ability to collect;moreover. it depletes the reagent. which otherwise would beavailable for adsorption on the mineral. The extent of SALOabstracted as surface chelate depends upon the solubility ofboth the mineral and the chelate itself. and upon the nature ofsurface copper species.

Acknowledgements

We acknowledge suppon for this work by the National ScienceFoundation, Interfacial Chemistry of Selected Fine ParticlesProcessing Systems; NSF-ENG-76-80139.

References

Biefeld, L. P., and Howe, D. E., 1959, "Separation and Deter-mination of Copper and Nickel by Salicylaldoxime," IndwtrialEngineering Chemistry, Anal. Ed., Vol. II, p. 251.

BI~tt, A.H., 1955,Joumal ofDTganic Chemistry, Vol.. 20, p. 591.Fuerstenau, D. W., Healy, T. W. and Somasundaran, P.,1964, "The Rore of the Hydrocarbon Chain of Alkyl Collectonin Flotation," Trans. SME, Vol. 229, pp. 521-5.Kohler, E. P., and Bruce, W. F., 1951, "The Oximes ofOnho-hydroxy benzophenone," Journal of the A merican ChemicalSocz.ety, Vol. 55, p. 1569.

Manell, A. E., and Calvin, M., .1952, Chemistry of the MetalChelate Compounds, Prentice Hall, New Jersey.

Nagaraj, D.R., 1979, Doctoral Dissertation, Columbia Univer-sity.Nagaraj, D. R., and Somasundaran, P., "Chelating Agents asFlotaids: LIX-Copper Minerals Systems," ACS Centennial An'nual Meeting, August 1976, San Francisco; published in"Recent Developments in Separation Science," Vol. 5, CRCPress, Florida, Chap. 7, pp. 81-94.Nagaraj, D.R., and Somasundaran, P., "CommercialChelating Extractants as Collectors: Flotation of CopperMinerals Using LIXR Reagents," AIME Annual Meeting,Atlanta, March 1977, published in Trans. SME/AIME, Vol.266, pp. 1892-97.Nenz, A., et al., 1964, "Method for Preparation of a-hydroxyand a-acyl°XY oximes," Belg. 655, 960.Patel, R. P., and Patel, R. D., 1970, "Proton-ligand stabilityconstants of some ortho-hydroxy phenones and their oxirDes,"Journal of Inorganic Nuclear Chemistry, Vol. ~2, pp. 2591-2600.

Detachment of surface chelate is also an important possibilityto be considered in evaluating the partitioning of SALO.However. in the present system. results have indicated thatdetachment of surface chelate does not have a significant in-fluence on the panitioning compared with that of solubility ofthe mineral and the chelate.

The concentration of copper species released from themineral decreases sharply with increase in pH up to about 5.0,and decreases slowly above pH 5.0. As a consequence, the rateand extent of bulk chelation also decrease above pH 5.0. Fur-thermore, surface chelation is influenced by the nature of thesurface copper species (depending upon pH) and the ease withwhich these species form a chelate with SALO. As seen fromfig. II, the amount of surface chelate increases with increase inpH from 4.0 and reaches a maximum at pHrv9.0. Flotationresults are again correlated by the amount of surface chelate inthis pH range. The continuous decrease in bulk chelate abovepH 3.5 follows the continuous decrease of solubility with pH.

Summary and Conclusion

The effect of the structure of hydroxyoxime chelating agentson their collector property and Some results of the basic study ofthe salicyialdoxime-tenorite system are discussed in this paper.

Microflotation of chrysocolla with a number of water-Solublearomatic (and one cycloalkane) hydroxyoxirnes indicated that:

a) Increasing chain length on Rl (R2 = H, see formula I) in-creases the collector efficiency.

b) Substitution in R2 (Rl = CHS or H) or substitution of aphenyl in Rl (R2 = H) decreases the collector efficiency tolevels inferior even to the parent oxime (R 1 = R2 = H).

c) A fused ring instead of the benzene ring does not improvethe collector efficiency.

d) Both ortho-hydroxy cyclohexanone oxime and its copperchelate have a high water Solubility compared with other oxirnesand, as expected. this oxime is a poor collector.

Panitioning of salicylaldoxime into bulk and surface copper

MINING ENGINEERING SEPTEM8ER 198f 1357

David H. Carlson and John II. Johnson

AIME Annual Meeting. An overview of monitoring technologyand the health effects index are presented in Johnson (1980), in-cluding the use of CO2 concentrations for pollutant control.

Mine Characteristics That Affect Air Monitoring

The unusual nature of underground mining requires specialconsiderations in designing monitoring systems. Some featurespeculiar to underground mines are not common in other en.closed spaces where monitoring must be performed:. Large spaces with few people occupying them.

. Most people work at a specific location for only part of ashift. This location will be unmanned during the remainder ofthe shift, and distances as great as 0.4 kIn between personnel ina single section at the face are common.

. Most work in underground mines is performed withmobile equipment. In conventional mining, driving-tunnelsusually involves a sequential cycle of drilling, blasting, loading.and hauling ore to conveyor feeders, and bolting for roof sup'pon. Usually one or two people operate a single piece of equip.ment. Mined out areas become inactive, and only cenain tun-nels are maintained through the old workings to the newworkings. Ventilation air and supplies are often brought con-siderable distances from main shafts or portals through oldworkings to those being mined on the perimeter.

. Underground lighting is usually limited to lunch areasand conveyor feed points. Electrical power suitable foroperation of electronic instruments is often unavailable. Powerused for lighting may have voltage variations that can damageinstruments or cause erroneous readings.

Tabl.1 .Ranges of Pollutant Concentrations and TotalAmounts Emitted by Diesel Engines at RatedPower and Speed (Marshall and Fleming, 1971).

Exhaust Pollutant ConcentrationRInge

200 - 3300 ppm

700-~pp",2-20%~/ty

24- 288 mg/1n3-8O-1140ppmC3-97 ppm

RMIgegl8HP-II2.3 - 10.06.1-19.84.0-12.90.1- 0.60.3- 3.10.1- 0.5

PollutantCarbon monoxideNitrogen oxidesSmoke

Unburned hydrocarbons (as CH2)Aldehydes

S, Specific Eml88lonaPollutant

CONOx"(asN~}

(uNO!ParticulatesHydrocarbons (as CHVAldehydes (as HCHQ)

. ThiS ,,,as not presented by Marshal/, but was calculeted from the smoke opacityusing the relationship presented by Vuk et. al. (SAE Trans., 1916} for a 102 mm

path length... Measured as NO by NDtR Instruments.

Abstract- The paper presents background information, adescription of mine characteristics that affect air momtoringtechniques, and a reVl'ew of portable measurement techniquesand instruments for gas and particulate matter samPling. Ageneral review of the laboratory and field research being con-ducted at Michigan Technological University in the area ofdiesel emissions is presented. The facilitl& l'Rvoltled with thisresearch are also descn'bed. These are the experimental mine,the diesel dilution tunnel test cell, the mine air monitoring lab(MAML), portable instruments, and the production mineswed.

The application of ~l powered equipment to qndergroqndmines is generally believed to be advantageous due toelimination of electric shock. fire hazards. methane ignition.and physical safety hazards associated with electric cables. andredqced costs resulting from greater mobility (Ogden. 1978).However, mine oper~tors and mine employee organizations areconcerned about the potential haz~rds from exh~ust emissions~nd the potential for incre~sed ventilation costs to reduceemission concentr~tions to levels not harmful to the health ofexposed personnel. These concemsform the basis for the researchwork described in this p~per. The ultimate goal is to optimizemethodology for air qu~lity control for the miner's health.A recent paper by Johnson. Reinbold ~nd Carlson. 1981discusses the engineering control of diesel pollut~nts in un-derground mines.

Development of suitable measurement techniques mustlogically precede ~doption of optimum methodology for control.The measurement of mine air quality has been prim~rily basedupon various collection devices and detector tubes that give asingle number representing the entire period over which themeasurement was made.

The advent of continuous battery operated portable in-struments and recording devices offers the important advan-tages of providing detailed data to evaluate control technology.The underground mine poses special problems for many pres-ently available instruments. and measurement techniques areunsuitable.

Michigan Technological University (MTU). under contractssupported by the Mine Safety and Health Administration(MSHA) and by the US Bureau of Mines. has put together acomprehensive research program in mine air monitoring. Theo~jective is the development of optimum methodology for con-trol of air quality. The work has been directed toward develop-ment of systems for monitoring mine ambient air. Studies havebeen carried out in three production mines and the MTU ex.perimental mine. The three production mines are the WhitePine mine (copper), St. Joe Minerals, Brushy Creek mine (lead).and the Colorado Westmoreland Orchard Valley mine (coal).

Ranges of concentrations of various pollutants in dieselexhaust (Marshall and Fleming. 1971) are listed in the upperpart of Table 1. Total amounts emitted are listed in the lowerpan of the same table. Each of the diesel exhaUst pollutants hasan effect on an exposed individual related to the degree of ex-posure.

Detailed discussions of the amounts and characteristics ofdiesel exhaUst pollutants and health effects and standards ap-plying to personal exposure limits are discussed in the refer-ences. A new health effects index that takes into considerationcombined hannful effects of CO. NO. respirable combustibleparticulate. S02. and N02. and sets a limit based on amathematical formula. is also discussed in Mogan (1978). Thispaper is based on a more detailed paper presented at the 1979

1358 SEPTEMBER 1981

DoH. Cartson, member SME, is a senior research engineer in-stitute of mineral research; J.H. Johnson is a presidentialprofessor in the department of mechanical engineering -engineering mechanics, both at Michigan TechnologicalUniversity, Houghton, MI. SME preprint 79-69, SME-AIME AnnualMeeting New Orleans, LA, Feb. 1979. Manuscript Feb. 1, 1979.Discussion of this paper must be submitted, in duplicate, priorto Nov. 30,1981.

MINING ENGINEERING

ABSTRACT

Organic ccxnplexing agents have received special auentioo over !he pastfew decades in the search for ~ents for improved flotatioo separationof minerals, especiaUy non-sulphides. The majority of such reagents thathave been pI'OI»sed over the past few decades have remained largely alaboratory curiosity, excep pedtaps alkyl hydroxamales, because theirseleaioo is seldom based 00 ccxnmercial viability. In this regard, thechoice of ccxnplexing agents such as oximes may be more realisticbecause of their widespread use in solvent extraction of ~r. Ourearlier studies have cmfirmed !hat !hey are also effea.ive as collectors forcopper oxide minerals, including chrysocoUa which is the most difficultcopper mineral to ~eficiate.

The flotation bel1aviour of chrysocolla for several structurally-relatedhydroxyoximes were studied as a functioo of pH and oxime coocentrationand the role of structural features examined. The structural changesstudied included substitutioos in b«h clIelaling and non-cltelating parts ofthe molecule. The results indicated !hat, cootrary to expectations, manyof the structural changes in !he basic molecule, salicylaldoxime, do nothave a positive effea. 00 collector efficiency. The reduction in collectorefficiency is ratiooalised on !he basis of possible Sleric factors affectingccxnplexation and packing of the collea.or species 00 !he mineral surfaceas well as effects due to changes in electron density brought about bysubstituents. Introduction of alkyl substituents on the oxirnic carbon ofsalicylaldoxirne yielded !he highest coUector efficiency.

INTRODUCTION

Over me past six decades or so, since me advent of solublesurf~tants as collectors in flotation, mere has been a relentlesssearch for new chemistries for improved separation of minerals,bom sulphides and non-sulphides. Whemer it is for flotation ordepression, organic-complexing agents or chelating agents. havereceived special attention (Barbery el ai, 1977; Drzymala andLaskowski, 1981; Fuerstenau and Pradip, 1984; Gutzeit, 1946;Holman, 1930; Marabini et ai, 1971, 1974, 1973, 1976; Nagaraj,1987; Somasundaran am Nagaraj, 1984; Taggart. 1930). This islargely driven by the success of these reagents in analyticalseparations. While in principle it is logical lo expect metalspecificity observed in analytical separations to translate intomineral specificity, in pc~tice this is far from a straight forwardtransfer. Nevertheless, an assumption is made in me majority ofstudies mat complexing agents adsorb on minerals by formingmetal complexes on mineral surfaces under conditions mostfavorable for metal complexation in solution as observed inanalytical separations, and that such metal complexes areresponsible for imparting hydrophobicity lo the solid. While avariety of complexing agents are alreooy in large scale use insulphides flotation, this is not the case in oxides flotation. Fattyacids and amines still dominate me industry though numerouscomplexing agents have been proposed. Perhaps the onlysignificant commercial utility in the non-sulphides area can beattributed to the alkyl hydroxamates (Fuerstenau and Pradip,1984) in the former Soviet Union at least until a few years ago.Current usage of mis is not known. Limited application can alsobe found for certain phosphonic acids (Nagaraj, 1987). Theprohibitively high cost and high dosages required of many of theproposed complexing agents are certainly major factors. Theunoorlying factor, however, may be the choice of a complexing

agent. Complexing agents have been chosen invariably on thebasis of success in analytical separation rather than on the basisof commercial viability. Thus many of the proposed structuresfall into the category of 'exotic' chemistry and remain largely alaboratory curiosity. h1 this regard the choice of complexingagents such as oximes may be more realistic. The commercialviability of oximes has been well established because of theirwidespread use in solvent extraction of copper.

One of the early references to the proposed use of oximes isthal of Vivian in 1927. Dimethylglyoxime (DMG) was proposedfor floating nickel oxide ores. Delitsina el al (1954, 1956)recommended the use of DMG for the flotation of chalcopyrite,bornite, malachite, pyrite and electrolytic copper and nickel.They concluded that the increase of hydrophobisalion of themineral was greater with DMG than with xanthate, which is anintriguing result, especially for sulphide minerals. Their fmdingscontradict those of Peterson el al (1965) who found lack offlotation of chrysocolla with DMG and of Drzymala andLaskowski (1981) who found lack of flotation of syntheticmillerite and other nickel minerals with DMG and otherdioximes. It should be noted here that DMG forms water-solublecopper chelate. Usoni el al (1971) showed thal a hydrocarbon oilwas necessary to obtain acceptable flotation of niccolite withDMG. Teoh el al (1982) used several homologous dioximes forthe flotation of nickel bearing minerals. Optimum flotation wasobtained with 2,3-nonanedione dioxime. Peterson el al (1965)also suggested the use of alpha-benzoin oxime for chrysocollaflotation.

De Win and Batchelder (1939) investigated the collectorfunction of salicylaldoxime (SALO), its meta- and para- isomers.and its monomethyl ether, for chalcocite. covellite, azurite,malachite and cuprite. Only SALO was found to function as acollector. The meta- and para- isomers could not function ascollectors since their structure does not permit chelate formation.It was also proposed that if the phenolic hydroxyl is replaced by amethoxy group. the resulting monomethyl ether of SALO wouldnot be a collector. Mukai and Wakamatsu (1976) also usedSALO as a promoter of xanthate adsorption for the flotation ofchrysocolla. Adsorption of xanthate increased with the additionof SALO, while a complete coating of SALO on chrysocollaooversely affected xanthate adsorption. No effon was made toinvestigate the collector action of SALO alone. It is not clearwhalthe mechanism of this coadsorption is and why a xanthate isrequired in addition to SALO.

Barbery el al (1977) and Cecile el al (1981) investigated thecollector function of oximes for malachite and chrysocolla. Theystudied adsorption of gALa on malachite with supponinginformation derived from IR analysis. Mullilayers of copperchelate were observed on the mineral surface.

Nagaraj and Somasundaran have conducted many systematicstudies on the adsorption of oximes on copper oxide minerals andtheir flowion (1979a, 1979b, 1981). The collector property ofcommercially used solvent extractants of the u:x type .wereevaluated for the flowion of chrysocolla and cuprite. Theseinvestigators also conducted a fundamental study of the systemtenorile-SALO in order to elucidate the mechanism of collector

. The terms 'complexing' and 'chelating' will be used

interchangeably in this paper since all of the discussion applie!equally to both, and since 'chelation' is but a special case of'complexation'.

XVllllntemationai Mineral Proc»ssing COf9'8SS Sydney,23. 28 May 1993 S77

TABLE 1The structures and water solubility of hydroxyoximes.

Structure NAmAMol.WI.

SALO

WaterSolu-bility, M

2.0 x 10-1(Salicytaldoxime) 137.1

OHAPO (o-Hydroxy Acet0-phenone Oxime) 4.5 x 10-3151.2

OHBuPO (O-Hydroxy Butyro-phenone Oxime) 5.0 x 10-4179.2

Q-: HC""II

OHtQt

~Y',c,.aiJ

OHNOH

Q-: af.afJO\C"'"IIOHtQtQ- ~ ~

~C~ IIOH~~

OHBePO (o-Hydroxy Benzo-phenone Oxime) .5 x 10-4213.2

2H5MeAPO

3.0 x 10-4(2-H)ldroxy 5-Methyl

Acetophenone Oxime) 166.2

2H5MBAO (2-Hydroxy 5-MethoxyBenzaldoxime) 5.0 x 10-3168.2

OHNAO (2-Hydroxy 1-Naphthal-doxlme) 1.0 x 10-4189.2

OHCHO (a-Hydroxy Cycio-hexanone Oxime) 129.0 -1.0

OMeAPO (o-Methoxy Aceto-phenone Oxime) 165.0

C Of.

II

OHAPOMeQ (o-Hydroxy Acet0-phenone o.methytOxime 165.0

Generalized Structure ..-'-,

.,..R,CII

QfNOtf

The structure of a complexing agent and, therefore, itsproperties, can be changed by incorporating substituents in eidierthe chelating part or die non-chelating part of the molecule. Theeffect of such changes on the collector efficiency of a number ofstructurally-related hydroxyoximes for chrysocolla as a functionof pH and reagent concentration is discussed in this paper.Information on structure-activity relationships is important fromthe point of view of not only understanding fundamentalmechanistic aspects, but also for the commercial viability of

~tion of hydroxyoximes. Aliaga and Somasundaran (1987)fouOO an interesting correlation betWeen dle UV spectra ofseveral strucwrally-related oximes, dteir copper chelates and dteircollector properties.

Kiersmicki et aJ (1981) studied the effect of alkyl chain lengdtin 5-alkyl SALO in the flotation of sphalerite, smidlSOnite anddolomite and suggested dlat a propyl group gave optimumresults.

578 Sydney, 23 - 28 May 1993 XVllllnlemationaJ Mineral PIO<»SSing Congress

OONOH

cor.

~:i~.OHNOH

~ ,~,~ ~...H00 NOH

Q.NOH00

Q- C ... 01,

cor.':o.r

Q- ~ ... 01,

001'«)CH,

~

>-Q::

~0(,)wQ::

Z0

~~a.;;..JlJ.

INITIAL CONCENTRATION, M

FKJ 1 . ~ of cluy~a as a func:lim of ~ of ~ oximeI: SAW; OHAPO; OHBuro

oximes in flotation. Chrysocolla would be a logical choiceamong die oxide-type CORJel minerals because it is one of d1emost difficult 10 float. Its flotation aspects were reviewedrecently (Laskowsk.i el ai, 1985).

EXPERIMENTAL

confinned by IR and TLC techniques. Water solubility of theoximes was measurrd by detennining d1e amount dissolved afta-shaking d1e oxime in 10 mI of biply distilled water for duee daysat 25 - 26°C. The values are given in Table 1. These solubilitiesIre expected 10 be close 10 the equilibrium v~.

Stock solutions of oximes in triply distilled water werelX'epared by adding one pa' cent acetone 10 ensure adequatesolubility. This amount of ~Io~ was also beneficial in flotationtests because it lX'oduced small, uniformly sized bubbles; aseparate frother was, therefore, not ~sary.

Mineral

Chrysocolla, CuSi03. obtained as high-grade lumps from Black.Hills Min~als. was crushed. hand-soned. and crushed again to-14 mesh. The -14 + 100 mesh fraction was passed through a drymagnetic separator, ground in an agate mortar to -48 mesh. The-48 + 100 mesh fraction was passed through. magndic separatoragain. deslirned and dried. It analysed 24 per cent Cu. X-raydiffraction indicated mioor amounts of quartz impurity.

Flotation

95 ml of the oxime solution at the desired co~ttation wasstirred in a 150 ml cy1iJKier. After- pH adjUSbnent 10 the desiredvalue, 0.7 g of the mineral was added and oonditioned for ther~uired period widl oontinuous conttol and recording of die pH.pH wu adjusted widl KOH and HNO) 10 within :t. 0.05 unit.After oonditioning. the mineral was floated in a modifiedHallirnoM tube for one minute widl 20 cc/min nitto&er1.

OximesThe oximes used are shown in Table 1. SAW obcained from LaChat Chemicals was purified by recrystallisation from petroleumedter-balzene mixwre. 2-hydroxy-l-cyclohexll¥me oxime wasprepared by die med¥>d described by N a1Z el aI (1964). All OIlIeroximes were p-epared from dieir respective ketones or aldehydesby the methods described by Blatt (1955) and Kohler aIKi Bruce(1931). All the oximes except OHAPOMeO w~ purified byrecrystallisation. OHAPOMeO was obtained as a liquid and itwas purified by ether exlractim. The IlnM:tures and purines w~e

R~ULTS

Flotation data for chrysocolla at pH 4.8 as a f\DlCtion ofconcentration of the duee h<Xnologous oximes - SAW. OHAro.and OHBuPO - are shown in Figure 1. pH 4.8 was sel~tai forthese tests ~ause preliminary work with UK type of reagentshad iOOicated thaI flotation was Optimum arotmd pH 4.8. The

579XVIII k1.ma~ MneraI PIOceIIkIg Congr8Is S~.23-28May1~

(ie 2H5MBAO, ~ Table 1) as well as methyl sumtitution in theS-posirlon of OHAPO (ie 2H5MeAPO) decreased the collectorefficimcy (more so with 2HSMeAPO), though both 2HSMBAOand 2H5MeAPO had lower water solubility compared with theirparent mol«ules.

Flotation of chrysocolla with 2HSMBAO and 2HSMeAPO aregiven in Figure 4 as a function of pH. Rotation behavior with2H5MBAO is very similar 10 that with SAW, while that with2H5MeAPO is quite diJIerent and unusual compared widt d1a1 ofOHAPO. Two flotation maxima are observed for 2HSMeAPO,one at pH 4 and the other at pH 8.

Flotation obtained with the remaining oximes - OHBero,OHNAO, OMeAPO, OHAPOMeO, and OHCHO at pH 4.8 areshown in Figure S as a function of concentration and in Figure 6as a function of pH. Flotation data with SAW (from Figure 1)are also plotted in Figures 5 and 6 for comparison. Most of theoximes tested in this group were characterised by very lowcolleclOr efficiency at pH 4.8 compared with that obtained withSAW. Thus subslibltion of. phenyl group for Rl (ie OHBePO,Table 1) decreased colleclOr efficimcy at pH 4.8. The pH foroplimum flotation with this oxime. oowever, was at pH 6, not at4.8. A fused aromatic ring (naphthalene) instead of the bmza1ering in SAW resulted in a very large lowering of collectorefficimcy at pH 4.8 (though it was optimum at awroximatelythis pH). Thus OHNAO was inferi<x 10 all the other closelyrelated oximes. Similarly, methylation of the phenolic OH inSAW almost completely destroyed the collector property. Onthe otha- hand. methylation of the oximic OH of SAW (ie

maximwn amoWlt of oxime dlat cook! be used in flotation waslimited by its solubility.

It is evident from the results given in Fig\U8 1 that m increasein die chain lengdi of RI (see generalised Slrucnlre. Table I)while keeping R2 = H results in m increase in die flotationefficiency in die order -H « -CH3 < -C3H7. AMing the fIrStcarbon in Rl resulted in die largest increase. Thus dieconcenlration of OHAPO required to obtain about 60 per centflotation is aoout half dlat of SAW and die water solubility isk>wa:ed by almost two orders of magnittxle. lncrease in diechain lengdi of Rl by tWo more carbons (to give OHBuPO)further im}X'oved die collector efficiency, but not to die extentobselved by introducing !he first cvoon. The water solubility ofOHBuro was only an orda: of magninlde less dtan diat ofOHAPO.

Flotation data for die three homok>gous oximes are given inFIgwe 2 as a CWlCtion of pH. ~-d1an-optimlDn concenttationswa:e chosen deliberately in order to observe d1e features offlotation behaviour vs pH md to discern differences betWeenoximes. These data, in general. su~ die results discussedabove. The oolleccor efficiencies of OHBuro and OHAPO werehigher dim diat of SAW at all pH values, diough theconcentrations used widi die forma: tWo oximes was lower diandtat used widt SAW. Flotation of chrysocolla was maximwn atpH 5.0. The effect of substination on die benzene ring on dieflotation of chrysocolla as a function of concentration of oximesat pH 4.8 is shown in Figure 3. Flotatim data for SAW andOHAPO (fr<Xn Figme I) ue replotted here for comparison. Itcan be seat dtat methoxy substination in the 5-JX>sition in SAW

100 I -I

~~

SAlOCHAPaOHBuPO

0 2.0 x 10-~M-4

6 1.2 x 10 M. -5- . ~ 0 9.0 x 10M

90

80~->-tr.4J>0(.)w~

z0

~

'0.;.J~

.~.070

~~~ ~60

0 ~

~'~50

ILl "

40 ~~r30

20Condo timeriot. time

10 min

1 min.0

to-3

0.S - 2 x 10 -.. KNOJ '

6 8 .2 144

CONDITIONING pHFIG 2 - H«aIim <i dtrylOCX)lia as a fwx:&jm <i pH wiIb ~ OI8Des: SALO; OHAPO-, OHBuPO

Sydney, 23.28.-v 1- XVIII Inlema1k)nal Mn8r8J ~ C«IGr8IS-

~

>-~w>0()4J~

Z0

~~0-.JLL

INITIAL CONCENTRATION, MFKi 3 . Acxatioo IX chrysoc:oUa u a ~ of ~ IX 0---: SALO; O~ 2H.SMBAO: 2HSMr.APO.

OHAPOMeO) only reduced the collector efficiency withoutdestroying it completely. pH fcx optimum flowion with thisoxime was at approximately 8.

The case of OHCHO is a special one because it belongs to theclass of alpha-acyloin oximes. though the oxime is distirx;(~ of the cyclic alkane. OHCHO was found to have anawreciable solubility in water (the highest of the oximes leSled).Its copper chelate was soluble in mildly acidic pH range andprecipitated in the basic pH range. In accord with this OHCHOexhibited extremely low collector efficiency even at very highcoocenlrations.

DISCUSSION

The flotation behavior for chrysocolla with the majority of theoximes snJdied was die same; optimum notation occ\med atapproximately pH 5. For UX65N (2-hydroxy-5-nonyl~ne oxime) two maxima in flotation occur, one at pH 5u¥i the; oth~ at pH 10. Chrysocolla flotation with octylhydroxamate has a maximum at pH 6 (Peterson el ai, 1965).Altl¥>ugh the optimum in notation of ctu'ysocolla around pH 5can be explained on the basis of CuOH+ species on the surf.:e(Palm~ et ai, 1915), an alremative explanation would involve thepartitioning of the oxime between the mineral surf~e aM thebulk Iq~US phase in relation to mineral solubility, as in the caseof Ia1Orite-SALO system (Nagaraj and Somasumaran, 1919).pKa of the oxime would also make a significant contribution tothe location of the; notation optimum.

The increase in colloctor efficiency observed with increase inchain length (ie SALO « OHAPO < OHBuPO. Figures 1 and 2)in the homologous series is in general agreement withobserv ations made in other systems. For example, F\lerstcnau,Healy and Somasundaran (1964) documented this for quartzflotation widt homologous amines. As can be expected.inlrodocing the first -CH2 in SAW causes a moch larger changein polarity of dte molecule compared with further Idditi<X1 of twoCH2 groups. This is reflected in the water solubility of theoximes - SAW » OHAP<> > OHBuPO - and also in thecollector efficiency. Furthermore, the irx:rease in collectorefficiency of the higher homologues may be related to theelectron-releasing (inductive) elfoct of CH3, relative to -H. whichwould inaease the electton density on the nitrogen. Watersolubility of homologous compounds should provide anapproximate indication of the hydrophobicity that they wouldimpart to dte mineral surface assuming that their mode ofadsorption is similar. This has been observed in dte case of fany~ids and amines.

Substitution of a phenyl group in RI (OHBePO. see Table 1)decreased the water solubility awreciably. which is to beexpected from the less polar benzene ring. On the basis of lowersolubility of OHBePO relative to SAW one might haveJX"edicled a higher collector efficiency than with SAW. Theresults. however. were quite contrary to this at pH 4.8 (Figure 5).An imponant reason for this is the shift of the pH of o.-imumflotation for OHBePO from pH 5 with SAW to -6 withOHRePO. This shift is possibly related to a change in PKa of theoxime in going from SAW to OHBePO. which can be expected

581XVIII kI.malkln8 MneraI p~~ C<M1gress SYckleY. 23 - 28 May 1 ~

~.

>-a:::

lIJ

>

0

u

w

a::

z0

i=

~

9L&..

:5

CONDITIONING pHfKJ 4 - flowioo of duyscx:UJa as a fl8laioo of pH wiIh osDnes: SAI.D; OHAPO; 2HSMBAO; 2H5MeAPO.

medtoxy group has reduced the hydrophobicity slighdy becauseof dte orientation effects.

The low collector efficiency of OHNAO (naphdtalene group,see Table I), u in dte case of OHBePO, may also be related tolow h)-drophobicity imparted to the s\D'f~ aftu ~on Didhindered packing on dte s\D'f~ because of the naphdtalene rings.The very low waler sojubility reslricted die maximumCOtK:altration studied to -10 M.

The drastic reduction in dte collector efficiency associated widtmethyl substitution for phenolic hydrogen (ie OMeAPO, Figures5 and 6) is expected because this molecule is unable to form achelate widt copper. This result is in agreement with dtatobserved by De Witt 81d Batcl1e~ (1939) for o-methylbenzalooxime and sUPlX>rts the hypothesis that chelate formationis a lX'erequisite for flotation. In dtis context. the absence ofcollector property observed by Nagaraj and Somasundaran(1979) for me syn-isomer of LIX65N is notewordty; dtesyn-isomer is also incapable of chelate formation. It can beargued, howevQ', d1a1 OMeAPO could stiliidsorb on chrysocoUaby fonning a complex (d1Ough not a chelate) widt cower via theoximic nitrogen aJ¥i function as a collector because simplealdoximes (such u R-NOH) are known to complex wid1 copper(Chakravarty, 1974). This ooes not appear to be me case from meflotation data obtained here. probably because dte methylsubstitution on phenolic OH may impose steric hmdrance to sucha complex formation widt oximic nitrogen on the mineral surf~e.

Medtylation of dte oxime group (as in OHAPOMeO, Table 1)

b«ause the benzene ring can alter the el~tron density on theoxime group significantly. Significant differences betweenorienlaboos of adsorbed SAW and OHBePO can also beexpected based on their structures. which then b'anslate intodiffer~ in hydrophobicity. Water con~t angles. for example.on surf.:es comprising -CH2 or -CH3 are higher than thosecomprising aromatic groups (WU. 1982). Also in the case ofOHBePO. there is a possibility dtat close packing of surfacearomatic groups is hindered. This would further contribute to thelowering of contact angles.

The lower collector efficiency of oximes widt substitution inthe 5-position on d1e ring, va 2H5MBAO and 2H5MeAPO.relative to SAW is rather surprising (Figures 3 and 4) in view ofdie fKt d1at dtey boch fonn stable, insoluble copper chelates withcopper ioos. The lower water solubility would suggest that theywould be less p>lar. Separate adsorption measurements (notshown ht;re) indicated d1at both substituted compotD1ds adsorbedas effectively as SAW. In the case of 2H5MeAPO. d1e pH ofoptimum flotation is shifted (in f~t there are two maxima. seeFigure 4, at pH 4 and 8). This may explain the k»w fk>tation atpH 4.8. The reason for the tWo flotation maxima is unknown andit cannot be rationalised readily b«ause all other oximes showed~ flotation maximum. In the case of 2H5MBAO, pH ofoptimum flotation was 8PlX'oximately the same as that obtainedwith SAW, though the methoxy group on the ring can beexpected to alter the pK. because of the elecu-oo releasingteJ¥leI¥:y. The only explanation that can be offered at this pointfor the k»wer coll~tor efficiency of 2H5MBAO is that the

Sy«18Y. 23 - 28 May 1993 XVllllntema1Xln81 MneraI PIOceIaiIg CongreII582

is not expected to prevent chelate fonnation with copper and.therefore. its collector p-operty should be similar to that ofOHAPO. but the results obtained here are quite contrary to thisexpectation (see Figures 5 and 6). OHAPOMeO exhibited onlyweak collector property even at high concentrations. This can berationalised at this stage only on the basis of possible sterichindrance to complex fonnation on the mineral surface. This isnot unreasonable because steric f~tors are much more importantfor surface chelation than for chelation in bulk *Iueous solutions.

The very poor collector efficiency of OHCHO (see Figures 5and 6) is also rather sutprising since after adsorption thiscompound should expose the cyclohexane group which should besufficiently hydrophobic. OHCHO was not only Vel-y soluble inwater (the highest among the oximes studied). but it fonnedsoluble cower chelates in mildly acid and neutral pH range. Theresults given in Figure 6 suggest its coll~tor activity to increasewith increase in pH. but a very high concentration may berequired to ~hieve high flotation of chrysocolla.

strucwral changes, however, made a negative contribution to diecollector efficiency. Correlation existed between reduction inwater solubility of the oxime and increase in its collectorefficiency only for the homologous oximes.

Substitution of alkyl group on die oximic carbon ofsalicylaldoxime had the largest positive contribution to thecollector efficiency which followed the order. o-hydroxybutyrophenone oxime> o-hydroxy acetophenone oxime »salicylaldoxime. Substiwtion of methoxy or methyl in the5-position on the benzene ring in SALO and o-hydroxy~tophenone oxime, respectively, decreased the collectorefficiency. This finding is contrary to expectations based onwater solubility and chelation of copper in solutions, and can berationalised at diis stage on the basis of steric hindrance tocomplexation and packing on chrysocolla. The methoxy group.in addition. may Iw to a reduction in hydrophobicity because ofits more polar character. A fused aromatic ring (naphthalene forexample) instead of the benzene ring in SALO lowered diecollector efficiency drastically. Substiwtion of a phenyl group onthe oximic carbon in SALO lowered die collector efficiency atpH 4.8, but it also shifted the pH of optimum flotation to 12Also, methylation of phenolic group in SALO destroyed diecollector property as expected. Mediylation of die oxime group,on die oilier hand. merely lowered the collector efficiency. Thisis attributed to the negative contribution from steric factorsaffecting complexation and plK:king on mineral surface. Ahydroxyoxime group on a cyclohexane ring offered noadvantage. In fact this oxime had extremely poor collectorefficiency even at very high concentrations. This is attributed tothe very high solubility of the copper chelate.

~ I

SUMMARY

Flotation behaviour of chrysocolla as a function of pH andconcenb'ation of several, sparingly soluble, sttucturally-relatedhydroxyoximes is discussed. Substitution in oodi die chelatingand non-chelating parts of die basic hydroxyoxinte molecule, diesalicylaldoxinte (SAW), was considered.

Flotation of chrysocolla was oplintum at pH-5 widi most ofdie oxintes. The major flotation effects observed in this study arereadily explained by examining die changes in s!ructural featuresof die oxintes. Conb'ary to what might be expected, many

100 I ta:.At'1~

9<>.

SALO .

OHBePO .

OHNAO -. -OMeAPO ~

OHAPOMeO OHCHO -.-

90

~..8Q

~->-''Q::

~.f~'""tiJ

~

~~<~,~-~

',$ 'Condo time = 10 minFlot. time = 1 min.Condo pH = 4.8~

-3$~Z 10 M KNO

50

.0,«

,

~~.<:>, g$'"

. ~.~ -v- "\7~ . """"""---4

30

2Q

}O~--' -

Q0-6 ~$e~' c 0-4 1i;\;t'3~

.,

Q1Z0 -1

,

INITIAL CONCENTRATION, MFIG S - F1«atioo of chrysocolla as a functioo of ooncentntioo of oximea: SALO; OHBePO; OHNAO: OMeAPO; OHAPOMeO; OHCHO.

XVIII International Mineral PIO~ssing Congress Sydney, 23 - 28 May 1993 583

~

>-Ck:lLJ

~UlLJ~

Z0~:<t-:0

~

CONDITIONING pHPKJ 6 - FklCalioo ci dlrysocoUa as a fImdion of pH widI OK8DeS: SALO; 0HBeP0; OHNAO; OMeAPO; OHAPOMeO: OHCHO.

ACKNOWLEDGEMENT

The authors ~knowledge me financial SUPlX>n provided by theNational Science Foundation and the Engelhard Corporation forcarrying out mis work.

Fuentenau, D W, Healy, T W md SanasWldaran, P, 1964. The Role fXH~ Q\ain at Alkyl CoUedms in FlOIaIi<Xt, TraIlS SME,229:321-325.

Fuentenau, D W and Pradip, 1984. in Reagelll.r ill Mincrals llldllltry(Eds: Jooes, M J, and 0bIan, R), pp 161-168 (The InJIiIUIioo atMiniJl& and Ma.lJurgy:~).

GUtzeil, G, 1946. O1ellte-FIXmin& Organic Can~ds as RotationReagenu, Trani AIME, 169:272.

Holman, B W, 1930. F1ocatioo Reagaa., BlIllllUt Mill Met, 315:53.Kienznicki, T, Majewski and M~k:, J, 1981. lIII J Miller ProcU$,

7:311-318.Kobler, E P and BNCZ, W F, 1931. The Oximes at QrdK)-Hyd~y

Beo~.J Am CM./It Soc, S3:1569-1514.Laskowski, J, Fuenlenau, D W, GooDiez. G and Urbina, R H 1985.

Studies 00 die f1awioo at O1ry1Ga)11a in Mwr PrOCU$ Eztr M.tallRn,2:135-155.

Ludt, R W and De Wilt, C C, 1949. The R«.alioo of C~ Silicate fnxnSilica. TraIUAlME,I84:49-SI.

Marabini, A M and ~ G, 1973. R~CXI at Pitmtiatde wilh aO1elaling Agmt and Fuel Oil, Tr4ll.f lllSt Min M.tall, 82:C225-228.

Mukai, S and Wakamauu, T, 1976. Sane Protieml at Copper OxideMineral Aocalion ~ at SME-AIME Fall M~. Deaver,Q), USA-

Nagaraj, D R and S<mallDldaran, P, 1979L in R.cent O.v.lopiMlII.f inS.pamtioll Sc;,.nce (Ed: N, U), Vol V, CRC Press, Boca Raton, A.017.

Nagaraj, DR, and SCMnaSlXldann, P,I979b. Trall.r SME, 266: 1892-1897.Nagaraj, D R and SanaslDldaran, P, 1981. Mining Eng, September, pp

13S1-13S7.

REFERENC~Aliag.. w UK! ~, P, 1987. M~ar 0rbita1 MOOeling UK!

UV~c lnveItiaatim oC Adsorption oC Oxime SurfaaanuinLA",ffUlir,3:1103.

Barbery, G, CeciJe, J L. FfiP81, J and Pliclla1., V, 1977. in XI/Illl.cnlGliOflDl Minloral Proccssiftl CMgrus Spccia/ PubliCtJlioll. Vol1/,S80Paulo,pp 19-34.

81an, A H, 1955. J Org CMm, 20: 591.Cccik., J L, CiU7., M I. Barbcry, G and Fripiai. J J, 1981. InCra~ SIx:CUa1

Study oC Species Formed m M&1ad1iIC Surf.:e by AdSorption CranA~s Salicylalooxime ~~m in J Coli 1111." Sci, 80:589-597.

OIakravMy, A, 1974. Struaura1 OIemillry oC Transilioo MetalCcxnpiexcs oC Oximel in ComdCMmRev, 13:1-46.

Dcliuina, G B and Yuyukevich, S M, 1954. A~ltim oC die PrincipalMinerals oC ~r .lid InxI wiIb !he Use oC Collecaon that FonnCanpicx ~ in Sb Nal£lI Tr MOIk llISl, T.rvcI M.1al1 iZolOIa. 24: 40.

Dc1iuina, G B, 1956. in Sb Tr Krworozhrk GtNMr~ IIISI, 5: 126.DeWiU, C C and vm Baldlelder. F, 1939. Old- CanPO'81ds as

A~m Reaa~ inJ Am CMm Soc, 61:1247.Dnymala, J UK! Laskowski, J, 1981. OIeIatina CanpoundS as Acxltim

Reaa~ in Pilysicocll.mical Pro~1IIS 0/ Mineral Processi""13:39-64.

584 s"..,. 23 - 28 May 1993 XVlllln18malionai ~ Pro<*l"wlg Cor9HS

Nagaraj. D R, 1987. The Otemiltl}' and AwlicaUoo of 0Iel81in& (X'Complexina Agents in Minerals Separatims. in RealMls ill MiMralTecllllOlogy (Eds: P. Somaslmdaran and B M, Moudgi1). PI> 251-334(Marcel Dekker).

Nenz, A 61 G/, 1964. -Hydroxy and -Acyioxy Oximes in Belg Patelll635.960.

Palmer. B R. Gulien'ez, B. G, Fuenlenau, M C and Aplan, F F, 1915.Mechmisms Involved in the F1«81ioo of Oxides 8I¥i Silicates withAniooic Coneaon . Parts I and n in TraM SME.AJME. 258:257.

PfScnoo, H D, Fuenta\8I, M C. Rickard, R S and Miller. J D, 1965.0uyI0c0I1a ~ by the FomIatioo or InsoIImie Surf~OIelaes, TraM AlME. 232:388-392-

Rinelli, G and Mara~i, AM, 1914. Flotation of Zinc and wd Oxide,Sulfide Ores with 0Iel81ing Agents in Proc~edings of X lMPC (TheJnstiIutioo of Minina and MeuUuray:f.A)ndoo).

Rinr.Ui, G. Marabini. A M md Ales~, V, 1976. Aow¥xl of Cassiteritewith Salicylalddtyde as a CdIusor in FiOIaiiOll - A M GaldillMemorial Vol- / (Ed: M C, ~), p 549 (A-ncan Inaiwteof Mining, MecaUuIJY and PeuoIaIm Engineen).

Somasundaran, P and Nagaraj, DR, 1984. O1cmisuy md Awlicalions ofOIe1a1ing AgCl1u ill ~Iioo md AoccuIaIioo, in R"leJIU ill theMiMral /ltdlUlry (Ed.: M I, Jones aDd R. ObIau), R' ~-220 (TheInstitulioo of Millin, and MetalluIJY: ~).

Taggart. A F, Taylor, T C and Ince, R, 1930. Experimmts with A~liooReagmts, TraIlS A/ME, 87:285.

Teob. E C, LawSCX1, F 81M! Haft, K N,1982, TraM IMM, 91:CI42-C152.U~ 1., RiIIdli. G and Marabini, A M, 1971. ~ Agent aIxi Fud

Oil: A New Way 10 FIocaIion ill Pnprilll' 71-8-10. AlME MUlill"New YOfk, FcblUary 1971.

Vivian. A C, 1927. A«a1im of Tm Ole. ill Millin, Ma,azw, 36:348.Wu, S, 1982. PolYrfWr IN..jacU alwl AdAuiDfI (Man:c1 Dckkc.-).

585SyGIey. 23 . 28 May 1993XVIII kUemalional Mneral p~ Congress

~ Sydney. 23 - 28 May 1993 XVllllntemational ~eraI Proc»ssing Congress