,\

.r \

Il .

l'

. ,

(

1

t

__ "_. ____ .... H~

MAITE/SLAG BEHAVIOUR UNDER HIGH PS02 CONDITI?NS

by

Francisco Javier Tavera Miranda

.\

"

A Thesis Submi tted to th.e ' Facul ty of Graduate Stuaie!; and Research

"

in Partial Fulfi11ment of the Requirements for the Degrée of Maste~ of Engineering in Meta11urgical Engineering.

, .

MeGill University Ment reat • Canada Dece}Dber 1977

. -----~Francisco Javier Tav.era Miranda

<C'

1977 . ,

--/ -<-

• *'

( ~ -..

"

~

-.. '.

f

/'

,D'

l' ... ,

1

"\ 1$

t

"

Ô

/

.. TO IRMA AND LUZ ELISA f

,. '-

~

\' ., /'

J

•

( ..

-,

( . \

..

(

)

(

A~STRAcr

An experimental programme in which Cu-Fe-S mattes, silica

,saturated Fe-O-Si02

slags and controlled pS02 atmospheres were equili­

bratedrJhas been caTrled out with the principal objective of examining /

~pp~ concentrations in

atmosphere) conditions.

the sla~s under oxidizing (pS02 - O.~ to l , ('

i.

The experiments '~howed that when slags and mattes are equili- , ,

brated, the copper-in-s~àg concentrations 'remain low as long as the

mattes cont.ain less than 60% Cu, under al! temperature and pS02 condi-

tiOIl6. Copper-ip-slag concentrations increase catastrophically under

oxidizing conditions when the mattes con~ore than 65% Cu. The

eXperiments also showed that whereas FeS has a high solubility for

oxygen and is itself soluble in slag under oxidizing conditions, Cu2S

and slag are virtually completely immiscible. Cu-Fe-S mattes behave

in an intermediate manner. 1

In a theoretical Pïrt of the work, the cdnditions under which

a pure copper phase will be at equilibrium with Fe-O:SiOZ slags and ,

Cu-Fe-S'mattes have been predicted on the basis of previous experim~ntal

apeO and p02 data tor sl~gS, and acuzs

data for mattes. It is shown

that under SiOZ

.saturation conditions m~tilllic copper will occur in

equilibrium with: Ci) 45-50% Cu mattes at iron saturation, with (ii)

79% Cu mattes at magnetite s~turation and with (iii) intermediate matte

compositions in contact with totally liquid slags.

\ il' ~;JV

\

1

..

/ ./

f

(

\ .,

iL - .'

1" 1 ..

The èverall conclusion of the investigation i~ that industrial

prodûction of mattes containing 60% Cu or below should not cause e~ces-• .,. 1). ... (

if , -:sive copper-in':slag losses under- any circumstances. Copper losses may ""

become.exceslive above this matte grade especially under high~ ~ " ("

oxidizing sonditions. "' ~ ~

1 (

r " .

" "

1

' .

/ , ,

...

"

~ .......... --------------------------------------------~.----.

,)

(

, '

RESUME

~

Un ,programme expé~imentàl a été lXécuté dan~ le but , , , 1

d'étudier la cencentration du cuivre dans lès scories dans des

conditions oxydantes (PSO - 0.1 a 1 a Çe programme de

recherche comprenait 2 " ... ;

des ~xpérientes d'éq i1ibrage éntre des matfes~

.du système Cu-fe-S, des' scories satur&es e silice 4u système Fe-O-SiO

et des atmosphères d'une teneur , 1.

contrôlée en 502

,

Les expérie~ces ont prôuvé qu', en léquilibre entre les " mattes ,

e't les scories, la concentration du cuivr e dans les scories n'est pas

élevée pour autant que les ma~tes contienneft moins que 60% Cu, et ce

p6ur toute la gamme de t~mpératures et PSOz\ étudiées. La teneur des

\ scories en cuivre augmente de façon catastrofhique sous des conditions

2

oxydantes quand les mattes contiennent plus ue 65% Cu. Les expériences , ,

ont aussi ~~ouvé que l'oxygène est· très s~lub e'dans les FeS et ce der-

nier est soluble dans les scories sous des co d~tions oxydpntes, tandis

que le Cu 5 et la scoriè sont complètement imm scibles. 2 \

Les mattes du

système Cu-Fe-S ont un comportement intermédia re.

Les conditions sous lesquelles une p de cuivre pur sera

en équilibre avec des scories F~-0~Si~2 et les attes Cu-Fe-5 ont été î

'théoriquement calculées sur la base de données Xpérimentales déjà

existantes sur aFeO et p pour des scorie , O2

pour des

mattes. Il est démontré que le cuivre méta1li ue, sous conditions de

saturation en 5i02, sera en équilibre avec (i) des mattes contenant

!

11i.

"

.'

" . ,

iv • . , "

'.

( 45-50% Cu sous saturatiqn en fer, avec

19% Cu sous saturation en magnétite, ',et a~ (Ut) des mattês de

ri' 'compo,sition intermédiaire;en contact avec, des scoties complètement

liquides.

'La conclusion globale de cette étude est que la prodùction r ,,,,"

industri~lleode mattes contenant 60% Cu ~u'moins, ne devrait, en ' .. 1

àucune' circonstance, "avoir comme résultat des pertes excessives de

cuivre dans la' scorie~ - Les pertes en cuivre pourraient dèvenir ... excessives quand' ,la teneur en cuivre de,t; mattes 'dépasse 60%. et

surtout sous des conditions oxydantes.

•

>,

1 l ,

,,'

(

,1 1

"

(

ACKNOWLEDGEMENTS, \

Sincere appreciation is extanded to Professor. W. G. D~ven~rt \

~ . " for supetvision and assistance given during the course of this work. . , ,

Deepest thanks~are extended for his helpful criticism, patience ~nd

guidance.

v.

Thanks are due ta Professor W.M: Williams, Chairman, Departme~t

df Metallurgical Engineerïng, McGill University, for his encouragement· / '

and for making avâilable the facilities o~ the Department . \ .

The author wishes to extend his thânks ta Mr. M. Knoepfel for

- ' his technical assistance and to Mr. M. Tanaka for his assistance in

, l ,

tr~slating JOU~~lS from Jt;.ne,e into English.

Sincerést ,appreci~tion is extended to~he author!s fellow

students, in particular, H. Salomon de Friedberg, J.A. Fallavoliita and

A.. Hafeez aluable assistance and for creating a most

pl~asant l ,

The author is indebted ta the CONACYT (Mexico) and Mç'Gill

University for their financlal assistance in support of this work~

and the Metallurgical Research Institute (University of Michoacan, /

Mexico) for encouraging him in his Canadian research.

f 1

i J

-(

...

(

/

TABLE OF tONTENTS

, ABSTRACT

RESUME

ACKNOWLEDGEMENTS

PART I. INTRODUCTION

1.1 A Brief Review of Copper Matte Smelting 1. 2 Smelting. Methôds ' 1.3 Chemistry of Matte Smelting.-·lts 1ndustrial

Importance

PART II. L1TERATURE SURVEY

11.1 II. 2 II. 3 II. 4

Mattes Slags" , .1.-

Smelting Atmosph~s' Previous Experimental Work on (Metal and/or Matte)/ Slag/Gas Systems

1I.4.A

1I.4.B 11.4.C

Cu (Metal)/Slag!Oxygen Studies (Cu-Fe-O-Si0 2 System) Matte/Slag/Gas (0 , S , SO ) System Cù (Metal)/Matte/Slag7Gas t02' 82' S02) System

II.5 9Ver~ll Status of Cu-Fe-O-S-Si02 Studies

PART III. Exp,OENTAL t

l' PART IV.

III.1 Apparatus III. 2 Materia1s III.3 Procedure 111.4 Chemica1 Analysis !ILS Results

DISCUSSION

IV.I Thermodynamic Interpretation of Results

IV.l.A Estimation of pOZ IV .1. B Ca1culation of pS2 IV .1. Cr#1 Estimation of aCu2S IV.l.D Ca1culation of aCu IV.l.E Calculation of aCu20

J'

vi.

• 1 1

Hi

v , -",

1

Z 3

7

10

10 13 15

,.. 15 t

2Z 27 ~

33

40

42

42 47' 47 49 51

54

54

56 56 58 58 60

l-

. . ,

'1>' •

I . " . ~,

(

\

•

(

J

IV.2 Behaviour of Copper in Matte and Slag ()

1

IV.2.A Thermodynamic Interpretation of Copper-in-

IV.3 IV.4 IV.5

'Slag Concent.rations ~J IV.2.B Effect of pS02 on Copper jhaViOUr

°B~haviour of Sulphur in Slag -~! Behayiour of OXygen in Matte Structural Representation of Results ;

,

IV.S.A Cat~strophic Copper-in-S1ag Loss' for 60 ... ~ Cu Mattes '

IV.S.B 'Relative Solubilities of and in CU2S and FeSo. _

IV.6 Prediction of Cdrdi tions under whicp a Separat,e Copper Phase will Occur

IV.7 lndustrial Impli~ations df the Results 1

PART V. FUTURE WORK

PART VI. CONCLUSIONS

AP·PENDIX 1.1 Details of ExpeTiments •

REFERENCES ""'­

\

(

..

1

, .

•

"

1

L, vii.

~

Page

, . "

60

66 66

67 69 71

72

72

74 81

83

84

86

87

(

,

\

".

-- i'

1

1 1 , ,

1 k

1

(

, -,

(

. ,.

"

1· 1. INTRODUCTION

Flash smelting and other~utogenous or semi-autogenous

processes are gradually.being adopted for the smelting of copper, 1

1 1

prirnarily as a consequence of th;ir relatively low overa1l energy

requirements(l). These processes operate at 'somewhat higher S02 pres­

sures than traditional fuel-fired processes (in which '!the 5°2 is diluted

by hydrocarbon combustio~ products) and, hence their chemistry of opera­

tion is somewhat different ~han that'of the oider ~rocesses. ·In

addition, flash sme1ting and the other new p!ocesses produce mattes

of~gh copper grade (> 50% Cu) as comparedfto the 30 to 50% Cu mattes

produced by the traditional r~v~rberatory furnace.

This thesis descr~bes an experimentai programme in which

matte!slag behaviour under conditions similar tcr those encountered in

flash smelting furnaces, i.e.

k_

ItSOz temperature

,~ matte grade

slag tr

0.1 to 1 atmosphere

1423 to 1573 K

30 tq 79% Cu

SiOZ saturated

""

wa~,studied; the abject being to determine the nature and co~ositions ( \ , .

of the phases which are like1y to predominate in flash furnace' (and

other novel process) units.

A1though the study was primarily scientific in nature, specifie

attention was paid to the copper contents of the experimental slags and

t

,.

_.

(

, .#11

( ..- .,

" / .

" *

l~-" ..

\ u

to the factors which control them. ,Copper-in-slag concentration is an

important economic vâriable because it ~terminës wh ether a smelting . ,.# •

slag ca~ be discarded directly or whether i t requi'res subsequent copper

j

, recovery t~eatme~t befot~ disea~d. • Il -tt

Other variables rec~~ting specifie <"- \

a~tention werero6Xy~en corent in ma~te and sulphur content in slag.

1 1.1 A BRIEF REVIEW OF COPPER MATTE SMELTING

f

Matte ~elting, ~s applied to copper, consists essenti~lfr ... ~ ~ / .

of melting and pa;rtially oxidizing,lcopper sulphide concentrates (with . / 1

suitable fluxes) to form: .

a)' a liquid matte phase con~aining 30 to 60% copper; and' .... -. "

h) a liquid slag (oxide) phase as devoid as possiblè of copper. }

The matte is more dense (p ~ 4.6 tonnes,m-3) than the slag (1) ).

'CP ~ 3.5 tonnes m- 3) and,hedCe the two phases are readilY separated

'~;"s' of tap-holes at different depths in the, furnace .. ,~ ,

After separation from slag, matte goes on ~o be ma~( ipto r, . ~'I 1

erude b~ister copper. This is performed by oxidizin~ iron and sulphur

fram the matte with air (converting) in a separate co~verting unit~ i.e." - 1

by the r~actions: . , F~ 02 + Feo'~ sci~ , .

i~, '1 ,. ,

(air) (slagged. with 5i02)

C~2S + 02 + 2Cu + S02

(plister copper) ,

(1 • o •

..

{,

)

l,

..

1 ,,'

/~

, . 1

1

(

------_. .."..

The ,blister CoppeT is subs~uently fire- and electrorefined

to its final e1ectrical gtade form.

Analysis of the mattes and slags of industrial matte sme1ting

~ operations are summarized in Table 1.1.'

~

1.2 §MELTING METHODS j

Fig. ,1.1 summarizes, in, f,lowsheet form .... curre~t (1977) industrial .... _-

'matte smelting me~hods •• Far a~ ~way the most ~mport~t smelting units

are the fuel-fired reverberatory'furnace and the flash furnace. The

latter is gradually bei~ adopted as the principal copper smelting method

largely as a consequence of its low hYdrocarbon ~~el requirement(l). Its , .

major dVferepcç, f'rom reverberatory furnace smelting is that the thermal

requirement of the proce~s is largely provided by combusting the sulphide

concentrate in place of hydrocarbon fuel. ' , '-)

In recent years. several novel proce5ses which are also designed ,

to combust the sulphide concentrates for most of their heat requirement . \

h~ve be.n developed. The most important of these are Ci) the Mitsubishi ~

Process. in which concentrates and pure oxygen are blown through ,tances , "

into a circular hearth furnace; and (ii) the ~oranda-ProcesSJ in which

.?xygen-enr:iched air is

Î' _

blown through tuferes into a matte phase while

concentrates ar~ continuously added to the bath. ,

lh the latter case the -,

sulphides are melted to for~ matte by the heat generated in the matte

bath. Several installations of ~ach type are currently in production or Î \:

are being constructed: Fuel .requirements of~opper matte smelting

processes are summarized in Table I.2. ~

, () . ."

., .....

••

f , v

l \

(

•

(

'\

/,

t'

(

. ,

; .

----_.\_.-

.. .. ; !

3 .. . . .. ., . .. • .. ,. .

:z . . .... ...

..

., .9 ~ l'

"

.. .

. ::

•• ; ..

• :

, ..

' .... ==.== .. . . .. .. • .. •

. ... :: = 1: ::

,~ :: .. .. ;;

-.

:c J

= ::

a = = = ~ ; x ~= "= = è fi

.. ; ...

;: :z

-... . ..

:: =: ::

• .. .

= tI = -.. .. ..

.. , ..

:: .. ..

J B •• = = ~ • = = ::... ~ = = •

il • .! i

• j 1= ~ i • 1 i, a il .. .. i J

~ o 1/1

~ P 1/1 Q) 1/1,

. 1-1

'-­ \

4.

.~ 1 1

f , 1

- ------ - ~ -~

.,.

< L -, -!....--- >;"""f ~",. ~

~ 5. I-

f ! ./

'" --J ( .'

<::. Sllt .... d. Of .. l\:r t. 2.,. Cil J ( J

1 , J

»2}\f c.",,,, ... l,,oft

1 . OMo FI.tot.on , Conc.-ftf,ottl 120 to JO "10 Cllt --------- - -1 -- 1

~ Orro ( . - 1

Or)'" Gr"" & ""~""O .. (ontlrn"."

tJ. qp~. ,lP .'

~ AOOltl/IO

Flalh lu,noct • t

f 1

~(ont,nuou. > ~ [tlrctue eloat '",noCt , IItoCn, ••

A .... Hbt'ofo'y J 1

" fvrnoct • rnOtt ""'-- ' .... Mottlr/30 to !!!!·t.CJI).... ...... ."..

, t:-'" ~ .,

.".. ."

conftr .. n9~[j~ ." lt -."

1 1" /' Il a ..... ' topper (98 ~t% Cu, r:

.' AIIOdt ,.'""no ~O ~ftCI een .... ' _ ~

:/ r' . AlIOdu (99 ~ -.4 (u'

[IKt",,'.",nollllllfl:~UIlIIII 0' Q

1 Cothodu (9999+% Cu,

···,~n -l . , ~

. ' A OP." mould ~ r (ont'""01l, ,ëa.ton, . (0'''''9 ~ •

'fJ> f

J:obt,(bt,OII ..... .". " '1

/-

1 , .

'f < _1 , "

'If- ? :-

," FIG. 1.1 Pri.ncipal proce~ses for ext'racting copper if from sulphide ores. Parallel Unes indi-, , cate alterna~ive processes (-~-- Rare{l)

( -.-.-. Unde~ Development) ,'Davenport

1 L

~ l , 1

l' . , '...-; .. .. t ,

_ ~_.'w~" _~'''''''''_'''''''~~''' ...-,~·I."."~ __ .~.~ .... , ... ,~ .. ~.~ .. ".~ ... ,..,........~I ...oL>.4>'r _ ,_, ....... L ..... _.L.""'''''''''~ .... ~.M.Io"~_ • .-~''',1'oI!r..,,.<...IIJ~~.~'''t.!'',_.:. .... ~

1 ~

t r

·,1 , , J .~

~ 1 t

1 1

t

" '­,r

~

~.-._~-.......

,1

l'

l-

~ Smelting Process

INCO flash

}lèctric

Noranda single step (30\ 02 in tuyeres)

.Outokumpu flash

Mitsubishi continuous

i

~

cal~ine charged reverberatory

Green charged reverberatory

,':-,

" '. TABLE 1.2 ..

, i ",'

r

'.

~

/ . ~

Energy Consumption Per Tonne of Charge: Kcal

(Including Energy for 02 i' •. ; Manufacture)

Energy Consumption-Fer Tonne of Copper: Kcal

(Assuming 2S\ Cu in

-~---~~~~,-~~ ----_.' ----------

Charge)

IxlOS 4xl0S

4xlOS I6xl0S , -'-

4xl0S \ 16xlOS

~ ~ -

~ SxlOS 20xl0S

6xlOS 24x105 ~

8jtlO5 32xlOS

I3xIOS ~ S2xl05

iP .. ,

.~J For- Smeltin~ in Different es of Proces~(ll ;:;:

"

~

, '0-..,

.,..

'0

"

1

(

)

(

7.

1

An additional advantag~ of the processes which combust their

s~lphide concentrates for much of their thermal' requirement is that the . , effluent gases contain 502 at high concentration, i.e.; the 502 is not

heavily diluted by hydrocarbon combustion products. This permits

efficie~t recovery of S02 from the effluent,gases and relatively straight-. forward production of by-product sulphuric acid, liaeid ;02 or elemental

sulphur. 60). concentrations ~ the gases emanating from matte smelting ~ , -}

furnace5 are summarized·in Table 1.3.

1.3 \

CHEMISTRY OF MATTE SMELTING - ITS I~DUSTRIAL IMPORTANCE

The metallurgical pngineer of today)~st concentrate on three

, ftdafnental aspects, of the c~pper matte sme! ting process:

~ Ca) minimization of energy consumpt~on;

Cb) minimization of S02 (and other) pollution; and

Cc) . minimization of copper loss. - j 1

Thi~ thesis i5 principally concerned with the last of these

5ubjects, i.e., minimization of copper 1055 in slag. ,..

Matte ~melting is carried oUf under differing oxi~izing condi­

tions according to the type of process used. ,The electric furnaçe (with <

; . ' its carbon electrodes) operates under the most reducing cond~tion~ while

flash furnaceso

and ,the n~vel processes (wi~Fr small hydro'carbon

fuel use') are at the highly oxidizing end of the spectrum. The fuel-, " , 1 ..

burning reverberatory furnace (with its copious quantities of combustion

gases) is intermediate in oxidizing potential.

The shift in copper technology towards flash furnace processing

" has made smelting under highly oxidizing conditions (pSOZ - 0.1 to 1 . -

~ ..

.'

\

..

t

~ ~, li'

'f ~ ". b; ,.

- v

".

~I )

"

I~

r --~--_ ..

---.,.

" ,

Hearth roaste~

Fluid bed roaster

Reverberatory.furnace

Electric furnace

Flash furnace INCO Out Qkumpu

Continuous processes Mitsubishi

" , '

n Noranda Worcra

Converter Pierce-Smith Hoboken

,

~ ~

'-----,......

" \

Vol.% S02 in Flues

A~erage A...

Maximum ." Minimum

\ --= 5

10-15 C> 0.5-1.5 The-enahama smelter has by oxygen enrichment, draft control and rèduction of air infiltration upgraded the reverberatory off gas ,ta 2.2% S02' -

4-8

...

Electric furda~es can also be operated under non-oxidizing conditions, in,' which case the gases ~ontain in the order or 0~4\ S02'

80 '"12-15

8 5

o 4-5 8

~

10 • (higher wit5 02 enrichment)

-8-12 (reaction branch)

10 11

~

... \.

-',

1-2 (slag-~ettling branch)

o O~

---.,

l

TABLE I.3 Concentrations of SO~ in Gases Producedby Various Smelting and Convert}ng Processes(l)

~ .111111'11 t n ._jii'ûiP . ~ .. " .~~ ..... _- · .... 1. 5 r

."7

•

,

\

oc

t

(

',' " '"

... ~4 ..

(

\ ,

9.

at~'spheTes) much more common . .Jhe work presented in this thesis , ' ' J-

o

- recogni'zes :thi~ change in that it is concerned primarily with the matte/ , .

v' ,slag phases which should exist under flàsh smelting conditions, i.e.: '. pS02: 0_1 to 1 atmosphere, matte-grad'es' 30 to 79% Cu; and silica-. . s~turated slags. Thus, the investigation is intermediat,e 'between ·early

studies, under iron saturation conditions. (low oxidation'" potentials) and . .,. " recent copper metaI saturation studies which were directed towards single

step copper-making pré,cesses. ')

" .,y'

"' ~

1 , , . ..

•

. '

l,

,t

/ .

/'

, .".,

,-

, .

/' JI

a

,

./

,.

"

u

~ f,. "i,ti,\. ....... <\li ..... , .. !fI!

,:

10

j

] 1

(.

(

,~ .. ~

" . . ...

{'

;,

1

~

,,'

~~_........-""-'------' r----- . \

II. LITERATURE SURVEY

This section discusses the mattes and slags which are enc~untered

in copper smelting and it ~;E~nes in d~tai~ recent'metal/slag/gafo~ . ~

~ , ,~

~_::. matte/slag/gltS and metal/matt~!slag/gas studies in arder ,ta lay ther ~ #

groundwork for the experimental part of this thesis .

. Copper.smelting mattes may be thought of as homogeneous liquids

consisting essentially of Cu, "Fe and Sand containing a'significant

amount of dissolved oxygen.

such as Cp, ,Ni, Pb, ~. and .. Industrial mattes also contain i,mpUl'rt)e,s,

prect0us metals which originate in indus~ial,

ores.

"" ~ ~ Mattes are mor~se tha~ slags (= 4.6 tonnes m~3 as opposed

• 3"" 'Li_ to = 3.5 tonnes m- ) and, Hence, they tend to settle below a slag layer. - .. They have a relatively low viscosity (O! 10 centipoise). Their e,.fectrical

J •

conductivities (300 to 1000 ohm- l cm- l ) are much higher than those of

, '1 ... molten ionie salts (NaCl, 4 dhm- l cm-1) and slags «0.5 ohm-1 cm- l ), which .. .....

inqicates~at mattes are semiconductors rather than ionic conductors,

i. e., they are covai ently bonded. .. "'"

Mattes can be thopght, therefore, as being low viscosity, high .\..\ •

conductivity liquids.in which ir6n and copper atoms arè covilently bonded ... tb sulph,ur (and oxygen).

~tensive experimental~studies of the Cu-Fe-S system h\v~ b~ed; mad, and .tailed phase diag~ams have been prepared. The Most complete- '

work is that repor~ed ~igS. II.I.A, II~I.B)/bY Krivsky and Schuhmann(2)

'"

1 ,

" (

.\

)

. (

(

, 0

, "

,.

Fe

. ~ ~ ---r _____ ~ ____ ~~ ~ ,_

. Fe-Cu-S: 1150o'C~ Pt: , atm.

, ,

10

t \

Cu ,

Cu ---lIi~ \

+ GAS

30 "50 " ,- .

) -" ... ~...- ......... ~-~-- ; - --- .

, ..

FIG.II.l.A 1423 K i~othermal section of the Cu~Fe-S system (Kriv,sky and SChuhmann l2 )). .

. o .•

J '

"

1 01

Il.

1)

(

, .

(

. -\

. , ,

~

1"

".

1

Fe

"

J

j.

è. -"

. '---

--- ..... ----_.-_.-,-------~ .. _ ... - ~ ~

/'-

/ Ee-Cu-S; 1300°C:PT1~tm

'\ Cu'

1

1 , J

1 1

1 1

1 J matte·

1 • flJ 1 gas ::, ( PS2=1atm) ffll l' El 1

1 1

1 1 1

1

. -(.1 --FIG.II.l.S :l'S73 K isothermal section of the Cu-Fe-S

.' ~~~ystem çrepoite~ by Ellîott(3)). 1.\ ~ ••

. f"""'t ..... ok...IU~JR~'-....,.J~"""~~\\..~-.Ir". .. :,.,~~J~~ ........ ::.'f';'#w< .... ~ _t

J ,

12.

1""1

• 1

/'

• : ~-~, ..... ~"'1î\;:'f "1J! "'-

(

-

(

13 .•

who studied the system in an H2/H2S environment; i.e., under oxygen-free "-

conditions. Elliott(3) has combined these data with data on thAY Cu-Fe-O,

Cu-O-S and Fe-O-S'diagrams ta give an overall picture of matte systems.

Th1e 1iquidus range (1573 K, {p02 + pS2 ~ pS02·+ pS03} < 1 atmos.) is

illustrated in Fig. II.2.

II.2 SLAGS

Copper smelting slags are represented by the CU-Fe-~Si02 system with CaO and A1 203 as major impurities and otner oxides such as

MgO and ZnO as minor impurities. Copper smelting slags also contain up

to 5% sulphur by vittue of contact with matte and concentrate.

The compositions of smel'ting furnace slags are adjusted (by

adding appropriate amounts of flux) almost to the point of StP2-saturation.

li! As will be seen in subsequent discussions, this procedure tends to minimize

copper solubility in the slag.

In the high-silica compositTon range, slags are ionic in

structure and~hey are made up 1argely of metal catlons and silicate ..

a relati vely low

Th, eY'lare viscous ~SOO to 2000 centipoise) and they \ 1 -1 (1),

electrical conductivity (0.2 ta 2 ohm- cm) .

have polymer anions.

It is weIl established that silicate slags are ionic in nature.

It.is imperative. therefore, that ionie structure theory be taken into

account in any mechanistic analysis of copper and sulphur solubility in

slags. 1

\

i 1

t

--.,7

.. '-,

""

• ( ~

\

.......

-

\

;

'1

--~

-?

~

~

Cu :'.)./

..

S

~ 0

..

FIG.II.2 Phase 'volume of the liquid matte phase in the Cu~Fe-O-S tetrahedron (1573 K)(3). '

"

, t

14.

<.)

1 D

t

Fa

t "

r ('?) ,

• " 1

'.

(

~

)

(

(

15.

1.

), ,

II. 3 SMELTING ATMOSPHERES

The gases in smelting fur.naces have the fol1owing approximate ~

~

. ctmpos it ions :

Furnace

Reverberatory

E1ectric

INCO flash \)

Outokumpu flash

502

0.5 tp 1.5

0,4 to 5

90

1S to 20

Gas Composition (Vol O2

2

o

o

N2

85

7 ,

60 to 70

COZ CO

10 0

o o

S '<;'7,..0 to 10

%) H20

"

3

5 to 10

\ \

Comments

Measured 0.3 m above

slag surface. Oxygen

is due to air leaks.

Leaving furnace. S02

concentration depends

on type of operation.

'. No air leakage.

No air leak3:ge.

This table shows clear1y how the switcho~er to flash furnaces

.~esults in a sme1ting atmosphere which iS,much more concentrated in SO~

than the traditional reverberatory and e1ectric furices.

, f II.4 PREV10US EXPERIMENTAL WORK ON (METAL AND/OR MATTE)jSLAGjGAS SYSTEMS

Elliott (3) has made an e,xtensive review of existing information

on meta1jmatte/slag/gas systems. His review is ~st.summarized by

Figs. II.2 and II.3.A, II.3.B, the last being due primarily ta Yazawa

(4 5 6 7) and co-workers "J •

\

1

( -

'''1

"

.'

~ 1

FIG". II .). A

1

la

l' , ,

,

. ,

- FeS

Fea

~

Schemati~ diagram of. the CU2S-PeS-FeO-Si02 system 1~~73 K, iton-~aturatiory)(31{). 'mi' and 's.' des~ibe the mattes and slags wh~ch are af equilibriu~with each other under Sio2!~aturati6n conditions. '

"'"

, . , r ~, '

, -

16"

.. .

1 •

(

:

(

, ,.

• f

l, ,-

1

17.

,\ .

o

SiOz , "

Ft

FIG. II. 3. B IhField of ~tabH~ t; of the oxysulfide phase t. in the co~osit.~on tetrahedro~ of the

Fe-O-S-Si02 sys~em (1573 K, Elliott(3». 1

.,..

/ _P_"'~""""-~l4'~lr""""f'"

\ J - '

1 t.L..

)

(

'r

1

1 t l

(

'J --~ . ,,1

,""

r

1

1 /

- '" 0' . __ ~ ____ . ____ _

18.

'" ~-.I , Re~éntal work on copper smelting system~ çan ~

divided into three categories of st~dy: 1

(a) Cu (metal)/slag/oxygen system: iron, magnetfte and/or .

) silica satura:ted

system: saturated /

, 1

(c) Cu (metal)lmatte/slag/gas , iron, Wustite, ~gnetite o

ri

(02~ i2' 502) 'system: and/or silica saturated

These studies are summarized in Table II.l a~d· t~~ir results . are discussed in the following three sub-sections.

,

"

\

. , . ,

"

1 • b

o

"",,~"""l<o<I"'~"" ___ "''''- ---, . ' .

'''}

, .

"

l

I~ ,~

:!

, 1

>

- - - --"",~-...... ~ ~ ........ ' ~ '.>"-~ • .. -~ ...,.~ -- *"

~

, :.'

System Exp~rimenter ...

g'

Temperature Range

K

~

Oxygen Patential Range Atmos.

.1

."

1.

..".

"

""'"'

" Connnents

~ eu-Fe-O-Si02 ~ Rudd1e, 1573-1673 10-12 ta 10-6 Si1ica saturated. System ~aylo~'and ~

BatesC~4)

_JO

"

---------.

c.

"" . Toguri and Santander (15)

tman and ellogg(16)

Bai1ey et al. (8)

-"" --

-1573

1500-1560

1573

»

"'...-

,

10,:"10 to 10-7

H- . -6 10- to 10, .

10-12 ta 10-6

, Fayalite slags. with approximate1y 6.% A1 2.03 • -

Si02 saturated iron silicate slag.

_SiOi saturated iron silicate slags gave coppe~ solubilities simi1ar to above studies. An industria1 slag (38% Si02, 16% Fe. 19% CaO, 17% Al203' 4% (20) gave much lower Cu solu-bi1ities (11.4). ~

TABLE II.1 Summary of Recent.eu-Fe-0-S-Si02 Studies

~

~~ , ..,

e ,.

""'"

l-

.'

'-' t

1

1 1 \

l' !

I ~

10

,. '17 $ ,)

~ 1

------.. -)1 -,.,t urN' '%bir.t,M'iem'er ,,'ette Fil sr· 'p +

~. <l"I"! 1i"-7"' ... " ._~'" ~-~\~-~-~ .... ,.." ... -

i "

l ~

1 i r i

.....---,.

System

CU-Fe-O-S-Si02 (no copper phase)

~ ~. -----~-~

'\

"

Experimenter

Yazawa and Kameda(lO)

Spira and Themelis (1;2)

Bor and Tarassoff(ll)

, . Nagamori(9)

Temperature Range

K

1473-1573

1473

" 1413

1473

~

Oxygen Poten-tial Range Atll\Os.

Iron saturation

Upknown. more oxidizing than Yazawa's work

Iron and si1ica saturation

Iron and siliea saturation

; t-

" .....---,.

--J

'.-

CODDDents

% Cu in slag inereases~ \S in slag decreases. \ 0 in matte increases with inereasing matte grade (Figs. II.9. II.10. II.11).

\ Cu in slag and % 0 in matte higher than Yazawa's (F~g. II.9).'

--J \

\ 0 in matte only. Simi~ar results to Yazawa (Fig. II.12).

Higher ~Cu in sl~g and lower \ S in ,slag than Yazawa (Figs. II.9.-11.10).

"

" TABLE II.1 (cont'd.) Summary of Recent Cu-Fe-O-S-SiOZ Studies

C3

...

r 1

<' ..

l r t r

~

IV o

\

~];!if "lEm iiiilf7'U 1 t'à"'l1 •••• ' rt.' arillE' if 71 ft "m 1 n u _.- -~- -.. - .. ----.-------

l 1

....

..

!'"'

System '

Cu-Fe-0 ... S-Si02 (copper phase present)

" ~

...

Experlmenter

Johansen, Rosenqvist t\ and ' (19) Torgerson

(

Gevici and Rosenqvist(18)

-

Temperature RaJlge '. K

1400-1700

1523

.JI " ~~-..

... , ~~~

,"

O~gen Potential ""Range -Atmos.

pS02 -- l ,with or witfiout Fe304 saturation

Fe and/or FeO ~/or Fe304 Md/or Si02 sa tura tion '

"

~

Conunents

,Somewhat few experimental results.

Minimum \ Cu in 51ag found at Fe and Si02 double saturation (Fig • Il.13). Maximum at Fe304 and SiO~ double saturation. .

TABLE.II.l (cont'4.) Summary of Recent Cu-Fe-0-S-Si02 Studies

b

i

\ '!<-(' --fr

..

" 1)

'\

~

~. ~

• 1 r: : : 1 •• P.M 5 , ' 1 1 liI Ii 1 [17 1 ln 11 )'] 11 -

........

t,J f-'

•

..

1

~

1

( J

,.

(

22.

II.4.A CuÇMetal)/Slag/Oxygen Studies (Cu-Fe-O-Si02 System)

'Although' the Cu/slag/oxygen system is not of direct commercial ,

importance, it is a relatively straightforward system to study (under

CO/C02 atmospheres) and it forml one'of the quaternary bo~daries of

th~ CU-Fe-O!S-Si02 matte/sl~g quinary system. For these reasons it has -

received considerable attention as ~s summariz~d in Table II.1 and, (

"Fig. II .4.

'The experimental resul ts are unanimous in showing that the

-' concentr~tio? of copper in an oxide (slag) phase coexisting wi th coppeli

metal increa~es with:

data,

increasing p02'

increasing temperature . .

1 •

On the basis of these results and other measured and estimated

Elliott(3) has constructed compbsition surfaces for the copper-

saturated iron-silica oxide phase as is shown in Figs. II.5. II.6 and ".

J

II.1). In essenc~, these surfaces indicate for.a given temperature.

p02 ~d Fe/Si02 ratïo, the concentration of copper which will be prese~t

in an oxide phase in equilibrium with metallic copper. The-effect of

increasing oxygen potential is clearly shown.

The effects of other slag components (Al203, CaO, MgO, ZnO)

been reviewed by BaiIeyeS). He was unable 1

have ..-

èonclusions. to~ any defi~tiV.

• •

••

* '

"

'. ·1 <

~

..

i 1

, , t

~ 1 f t .

\

#

,.

f ~

'-

t:..s. , 10·0

lo- BAlLE Y AND GARNER (SILlCA SATURATED ) (8)

8.0 1

" 01 o 6·0 -(/) C

:l 4·0 U"

~ 0 2.J· 1

Q.Q

Ô , RUDOLE, TAYLOR AND BATES (SllICA SA.::rURATEnJ (14)

0 TOGURI AND SANTANDER fFAYALITE SLAG, 6% AtZ031.<15l

/.

'-.......

- Cl

6 0. ~

og 00 o

-Il -lO -9 -8 -7'

Log P 02 FIG.II.4 Measured èopper-in-slag concentra~ions. Cu~e-O-Si02 system, ~der

, cbppeT and SiO] saturation conditions. The increase in copper -s~lubility with increasing poZ is notabl •.

• •• F ._~=-=-_.!-..<"":"::::,:;:;;;;;;_,,,,,,-:--'-"== ______ H --- - -,~_. m II 1 5 'OElrME n Il ,

..... -.......

• '1 ,.

-6

{

'1..

l'V <.>1

.,

t

_\.

(

ï

y

~,

~ •

..

/

( -

1

/

(\

(

..

.~

"\

\-

. ~ . FIG.II.5 Extension of the regions of stability of

the liquid ferrous-silicate phase (slag, of 1,1) and copper-silicate (t'2) phase ~nto the quaternary volume (1473 K, Elliott l3)).

l'-

24.

(

•

, r

,

~

, .

.,IJ " \J

~ .

\ . (

"

• i,

..

«"

L ... __ .... _

. ,

o ,1

Liquid 0 xide· Ph'ose in Cu-Fe-Q-Si02

·"System". 130QoC - . ~--~~----~------------~o

0..0-

/

-12 ""7--""';..I;::---:l::-----::-L-----1----=:::::J

\c~ 80

FIG.~I.6 Region of stability of the liquid oxy-sili~atér \.. 'phase in the c.v-Fe-O-SiOZ system (1573 K, " Elliott:,.(3)).

• 1

..

25 •

..

1.

/

. ~ >

\ ' , r' l t ,

(

Cu

( ..;, .

-~--.----.----------

•

. r'

%Fe-

,.

\ %0

80

1

FIG. II. 7 Position of the liqui~ oxidé phase in Cu-Fe-O-Si02 comEosition tetrahedrori (1573 K, El1iottl3~).

\ ,li .' 1

''1 ~

, . ,

,r<~ .... _~_ ~ ...... J .... -l ~ \~ • t. .. " ". v, . ~.

26 •

" ,/

Fe

~

~ l,

/II.

. . the

t ....

(

(

, . 27.

II.4.8 Matte/Slag/Gas (02~2~2) System ~

Nagamori(9), Yazawa and Kameda(l~), Bor and Tarassoff Cll ), and

Spira and Themelis(12) have carried out significant studies on this

system - the first three under iron and silica double saturation condi-

tions and the last without iron saturation (Table II.1). In aIl cases ,

the experiments '"were carried out under a neutral atmosphere into which J

OZ, S2 and SOZ could slowly escape. The partial pressures of the gases

were not, therefore, truly fi)Ced and the systems cannot be said to have

of 02,'S2 (

been in true equilibrium. The partial pressures and SO, a,re,

however, very low under iron saturation conditions and ~hus continued

flushing of them from the system should not, therefore, appreciably •

affect th~ compositioft0 of the coe:trstent condensed phases (Yazawa)'C13).

The results of these investigators are presented,in Figs. II.8, ..

II.9, II.10, II.11 and II.12 from which it can, be concluded (under iron

and silica saturation conditions, at least) that with a matte of

increasing copper content:

Ca) % Cu in the coexistent slag phase increases~ .fS . Cb) % S in the sléj.g phase decreases; and

Cc) % ° in the matte decreases.

Spira and Themelis(12) obtained similar results ~r non-iron saturation

conditions.

" Elliott(3) has developed a more general view of the matte

slag ~ystem based largely on the experimental work ~f Yazawa and co­

workers(4,S,6,7) (Figs. II.2, II.4, II.8). The overa11 matte/slag system , is described in terms of a CuzS, FeS, FeO, SiOZ quaternary ,diagram (Fig. II.3.A)

while the copper-free part of the system is described in ternary and

quaternar1 form (Fig~. II.3.B, II.8).

, / -

1

, , r

1

. "

\ (

/

...

... t

(

• ( ,

. t/'

-----,_. ---...----- -- - -

..

CI

, 5 +m

0

'so ·FeO"~s) +m

. ,

\:

60 80 Fe$ FeS-'

• 't

" ~)

PIG •. II.8 The Féo-FeS-Si02 system at iron-saturation conditions (1573 K, Elliott(3)) •

28.

-. ,

•

, ... ' .......... -.<. ..... _-""'--~-~~ ••• - ~ •

)

':"""

4

~/

~

1 ~ 0

0-8

Ô'O-6 C -(J)

c: -- 0·4 '!3

Ü

~ o

\' , "

'"

0-2

.... , '

o

/'

,;

t-

o SPIRA AND TH~MELIS (II °/0 CaO) (12 ) _ ~ - ~

-YAZAWA AND KAMEDA (Fe-SATURATION; 10-3 % Ca 0) (la)' A-- NAGAMQRI (~e-.~io2-SATURATlb~; 0,°/0 .CaO)(9)'

---

'-

,.

/~-

-,,'" ,/

'" - 6/ , ,~

,P' .H~

1 ... o

o

o '6 o \..,..........: ,

Il 0 fo~OOO '" o

20 40 60 0/0 Cu in 'matte ' ..

FIG.II.9 Measured copper-in-slag concentrations, Cu-Fe-O-S-Si02 system, as a functton ~'matte grade. The experimental conditions are shown in the Figure. ..

-_w._ 11.-: 'e' • 1"~ ~ "t' '1/11. ' --lti;iU *111 'nup 'n mi §' t- 'i 11 1 r •• q ne m 'f' 2 •••

'" /

..

"

;J

/ ,.........,

o

""-

,

80

, N 10

., Iq

.-~~ ... , ........",."._~ ___ ~ __ • ____ ....... ........,!- t)i $~ U2",,~I;.4'l~Spi ;.* f4'Ji\î!iiZP .. ~~*" .. ~-.o;.r~~..,.~~~~_ .. , .... ~.':. ......

r . 1" ~ ~

-,

~

-... ...

-i'

.' ).

-. ! ....

'f' " --~ • Il"

..

'j f J

50 ., 1 > <

. 1 NAGAMORI O{f:e ... SiO-a':SATURAT~ON)f9), \ 5 i ~

' ..

, ,t 1

• , ,1

",

l 1

'1

J.

\

f' "

4·0

" 1· - ,

0- 3·0' ~ c -oU)

c: --en 2·0

~ o

'·0

~

-- YAZAWA AND KAMEDA lFe-SATURATIO~i 10·3 % Ca 0) ( tO) --- ;) '\

...

'\ "

t( ~

'"

~

. ..

t

, .... / , , .:- " ,-r---

•

è

, .

\

~

l

L ,."

,... ," ~

~--~'.

~

'"" _ 1- _~ !~

1 5 7 3 K . ~ .. ~ ......... - ~:)~-

"

..

.... .'

..

-:=.1

.il

i 0 '1 80 20 ,40 60'

1 ..,

, ~

, ., C' LI cin matte 0/0 ~

, ,~ ,~ \)

FIG.II.IO Measured sulphur contents of slags, Cu-Fe-O-S-Si02 system Ciron saturated), as a function of matte grade.

t

'l'

î 1 ,

1 .

'1

----/"

.,.

"

J

q,

~ 0 .

. ---- ...... =,( .... - .•••••• rertnnU .,t •. '5n M'" '" 11'111 21', ,- • l' S'.,---~--1 •

, t . ~

f

'1 TI

)

t

J ! i - i 1

J '"

~--

....

'\-.

,.--.."

et

~ vi"

4

t

j>

10.0, _ , 6~

BOR 4NO TARASSOFF ( Il) \

_K~A AND YAZAWA (NO Ca 0 OR ~1203) (10)

YAZAWA (11 % CaO J (10) -

~ S·O.

~ • CI) ...... ~

,

.i ~ 6·0

•

c .--o 4·0

~ o

2-0

o

0.'.,

0 VAZ f-WA (7_ % A '2 <!..a) ( 10) 1

J --- YAZAWA AND K~MED~ tu % CaO) PRESENTED SY SPfRA

AND TtiEMELIS (12) "\

"'l

,.,

"" ~ v

'~ "'Y,",,- _ S

--- ..... -- -- -- "'" ---- ~--- --. - -- --f -- -1 1

20 40 60 °/0 Cu • matte ln ,.

, i

~IG. II. 11 Measured oxygen contents of mattes, C~~Fe-O-S-SiO system Ciron saturated), as a function Qf matte gr~pe. The Bot and Tarassoff(11) data a~e detailed in Figure II.12. ~

~

m; nt SPEmy- 1 - t - '" ~ ...... - la 1. st ~" lil 1 1 r . $ - Y 'F P't'n ' fi. 1i.8nllfrezœ, ct " U l' nn., • T

'!t

\~. \

,i

-...

{

- \

....

80

,

~ ~

i

1

1

1 i

\

. .. ._~

•

- t

: i l '" t + ; * .. , l. 1 ,-

. , , , ! "

t .. l-

""'

r

_ ~ : . .-'_;:'A"""h:ry~~"'''''-f%î42*;;SP~''''-~~<~-' .... ! _.'_"'< ...

-----.

, \

~ o

&

-----. Ë

~

''1 \ . ~

10·0 l ' ..... -- B OR AND TARASSOFF " (-6 IRON CRUC\BlE; 7·5 % CaO IN SlAG)

( a SlllCA CRUCBLEi 7·5 % CaO IN SLA<G) $-

8·0 .. Q) ..... ....

.- Ac

( 0 F\REClAY CRUCIBlE; 6 °10 CaO, 10 °/0 AI20 . 3

IN SLAG) ( Il ) , 1

" "!-

e E 6-0 , c --

--

°4·0 ~ , •

~ "

0

z·o -wt

.. -,.t

.. ~ A

o 20 40 60 \ 80 %_ Cu in ,matte

,

'> O~ygen solubility in mat~s under various experime~tal c~ndi)io~S as ~asured by Bor and Tarassoff(ll) (1473 ~. - \

- FIG.IL12

, Fi Je ._- .' Ur anm n_nATt1Er IliIaiJirliRlI",n If 1:":: nIf unli' 1" n rn.pm ln Il 7 ..

~

I.#l N'

" !

..

1

1 ("

J, , .. r ~

i, , ~'

f ~

'C ..

~ l"

- \~

___ ._._. -..1\~-

33.

i As would be expected from the dis tussions abovè, each repre-

sentation i~dicate5 the pr~ce of large C~mp05ition ~egions where

matte (oxysulphide) and .il~t. (.lag) Pha1e. coex\.to , The general

conclusions from the diagrams are that: \ /

(a) the separation of a sulphide fhase from

is most efficient when the sy~tem is silica saturated;

i.e., under these conditions'~he oxide contains a minimum

concentration of sulphur and ~~e sulphide pffase contains

a minimum :on~entra~on of oxy~en; ~

the concentration'o; copper in\the oxide (slag) phase

i5 minimtzed by silica saturàt~on (u~der iron saturation \

conditions, at least, Fig. II.3.A).

. (3). h 10tt pOlnts out t at an import4nt missing piece of \

information s the behaviour of the CU2S-FeS-FeO-SiO~ _ ~\ystem under non-

Iron satura~~ conditions.

II.4.C Cu (Metal)/Matte/S1ag/Gas (02~~2) System

In additi~n to the matte/slag studies described immediately ,

above, several studies of the copper metal/matte/slag syste~ have been'

made . ,( The most complete of these studies i5 the work of Rosenqvist •

d k (18.19) h . 'f' lt i Id . an ~wor ers w ose maré S1gn1 1cant resu sare swmp.ar ze 1D /. 1 F1gS. II.13, II.14. II.15, II.16). In essence. these investigators

prepared mixes of copper metal. iron,powder, copper sulphide and ' aster'

slag

give

"---...

(50 ta 55% FeO. 12% Fe203• 37% Si02) in proportions calculat~ ta

copper metal and a j1desired matte and slag composi tian. !hese mixes

1 . ,

1.

" '\

(

-,

"

( .

,

-1-· M

34. ~ c 8·0 / -en c -- 1-

ct • ::s 4·0 en '(\.1 U 0 .- >

~ en , 0

,0 cu

0·0 LL

,,-

z· z 0

·~2·0 0 J '-I-

+- ct C 0::

E ::J 1-e(

C en . /

-- 1·0 '(\.1 0

0 ü)

1 0

O'b ·0

0/ z -0

0'2.0 c -en , c:

l·O --en CIJ

~ , cu

0 U-

O:€) 60 70

% Cu in matte FIG.II.13 Copper-in-slag. sulphur-in-s1ag and oxygen­

in-matte concentrations, Cu-Fe-O-S-Si02

1 sys~em (copper and Si02 saturation condi- (18) tions), as measurcd by Geveci and Rosenqvist (1523 ,K).

l-ct 0::: :::> t-e( (/)

z 0 1-ct 0::: ::J 1-e( (/)

1

1

/

"

1

)

, )

;:

"

(

. ,

(

~,.. .... t Pfft A;

-35.

C' 8·0··......,..,.-------------r-----t C ---

cD ..... +-o E .-

'0 ~ o

4·0

10.0 ...... ~---..;;.---------.,.._-__tif

6·0

2-0

z o 1-~ Cl: ::) 1-~ en 1

,...0 cu u. 1 cu

L&-

z o 1-ct Cl: :;:)

ti en IN o en

-e~~p-__ ...d ~ li...

t' 5.0.--.....------------....,----,

-en c: --

zr-..... ~ o 1-ct Cl:

3·0 ~ ct Cf)

J o cu

L&-

z o 1-ct Cl: :;:) 1-<t (f)

1 ON

I CI) 1 \·0 fi)

1

~ o 4)

u..

30 40 50 0/0 Cu in matte

FIG.II.14 Coppe~-in-slag, sulphur-in-slag and oxygen-in-matte concentrations, Cu-Fe-O­S-Si02 system (copper and iron saturation

'conditions), as measured by Geveci and Roscnqvist (18) (1523 K).

•

•

Q) .J .... 0·8 .....

0 E C

0·4 .-0

/"

.'~

o 0.0

~0.8 0 -fi)

c: 0.4 .-•

en

\ ~ 0 0·0

• ( \ 'FIG. II .15

i t 1 j 1

~ . t

\

,..; < en Iv

0 r()

al u.. 1

0 Q)

LI..

. .... e:( (/)

, f

OV

t' "'" 1

o tf

77 78 Cu in

79' matte '

1 '

z o ~ ft: :l, .... <t en J N

o ü5 IV

01") cu

11.

o 80

Copper-in-slag, sulphur-in-slag and oxygen-in-matte concéntrations, Cu-Fe-O-S-Si02 system (copperland Fe t04-saturation conditions), as measured by Geveci and Rosenqvist(18 (1523 K).

, ...... , , .. ' l __

P 1

37. \ ,

1

~ r 1

f

,(

/'

-'

(

38.

,l

were then melted and held at 1526 K for 5 to 17 hours in a flawing •

nitrogen atmosphere. ! The 502 c?ntent of the effluent nitrogen was also

'determined for sorne experiments.

This 4-phase j '5-component (Cu j Fe, 0, St Si) system hast by

the phase rule, only two degr~reedom in addition to temperature.

As soon, therefore t as two additional parameters are specified, e.g., ,

(a) the activity or concentration of Cu25 in the matte; and Cb) the

activity or concentration of Fe304 in the slag; the system iF fully

de,.fined.

Most of Rosenqvist's(18) work was. in fact, carrieà out in the

p:~sence of a fifth separate phase~ Le., ïn irft, Magnetite, or silica

crUcib1es sa that aft~r temperature (1523 Kl had been specified, on1y

one degree of freedom remained to fully define the system. Matte grade \

(t Cu in the matte phase) was usually chosen for this final specification.

Based on this discussion and the experimenta'I resu1ts of Gevici

and Rosenqvist(18), the followi~g observations can be drawn.

) SiOl Saturation Conditions (Fig. II.13 and II.16)

When 'th~ system ts altered from ir9n saturation towards magnetite

saturation, the coexistent phases adjust in the f~llowing manner: . t •

Ma~te phase: % Cu increases

Slag phase:

Gas phase:

% Fe decreases

, \ Q decreases

J Cu -increases

\ S decreases

Fe++ JFe+++ increases (not shown on graphs)·

pSOz increases 0

1 "

,. "

1 K. l !

(

\

/

,-

('

./' Iron Saturation Conditions (Figs. II.14, II.16)

1 Alteration of the system from FeO to 5102 saturation affects

~he phases in the following ways:

Matte phase: % Cu increases

e% Fe decreases

% ·0 decreases

&lag phase: % Cu decrease.s·

% S decreases

++ +++ Fe IFe decreases (not shown)

1 Gas phase: PSOz decreases (not shown)

fe~ Saturation Conditions ,(Fiss. II.1S, II.16)

,Changing the system fpom FeO to Si02 sa~r~tion conditions

alters the p\ases as follow~:

Matte phase: % Cu increases

% Fe decreases

% 0 increases

51ag phase: % Cu increases

% S decreases

Fj+ IFe+"'+ increases (not shown)

Gas phase: pS02 increases

Geveci and Rosenqvist(18) conc1ude that the copper-in-slag

content is:

(a) at a minimum (O.S% Cu) upder Fe and Si02 sàturation

co~ditions (50% Cu in matte); and

f'

! l' ,

~'''.'''''~'-..t 1!4II~~"ifi .. _""'~'" - ~., .. ..,~, .......... .,..

(

(b) at a maximu~ (7.6% Cu) under Fe304 and Si02 saturation

conditions (79.5% Cu in matte).

Su1phur in slag is, on the ~ther hand: .

(a) at a minimum (0.01% S) at.FQ304 and S102 satpr~n

, J (79.4% Cu in matte); and \ .

Cb) at a maximum (4.4% S) at FeO and Fe saturation ,(32% Cu

in matte);

whi1e oxygen in matte is:

(a) at a minimum (0.05% 0) under Fe''304 a.t1d Si\02"oSaturation

c~nditions (79.5% Cu in matte); and

(b) at a maximum (5.5% 0) at Fe and FeO saturation (32% Cu

in matte).

)

11.5 OVERALL' STATUS OF Cu-Fe-,O-S-Si02 STUDIES

The above review indicates tha,t the Cu-Fe-S (matte) ternary - 1

system i5 weIl knOwn. Likewise, sla~ struètures and properties have .. , .i

been studied exte~sively.

The Cu-Fe-O-Si02 system ha5 been c10sely examined for oxygen

40.

-12 -6 pressures between 10 and 10 a tmospheres and,. further experiments in

this range would seem t~ merely repeat previous studies. The 5ys~em i5 ,

-6 0 only partially ~derstood at oxygen pressures between 10 and 10

atmospheres and further study in this pressure region is weIl warrante~

Matte/s1ag'studies have been much fewer in number than CU-O-Si~2

studies ~ecause of their greater experimenta1 comple~ity. The most

complete study jltthat of Gr'iCi and ~OSenqVist~18) for the ~tte/slag

system in the presence of ~eparate copper phase. trere i5 some scatter "

.. ...

,<1

1

(

(

. ---_ .. ------,----,...-.,.., _., -----_.--

41.

1

in these results and the experiments are restricted to one.temperature

50 that this work should be repeated and expanded.

Most other matte/sl~g studies have been carried out under

iron saturation conditions, i.e., under conditions which are much more

reducing'than conditiohs in flash smelting f~aces. These studies

should be extended to more highly oxidizing conditions~

The work in this thesis represents .such an attempt to exPand 1

ma1!te/slag knowledge away from iron saturation conditions towards

industrial smelting (high pS02) conditions.

I~

1/

,

/'

/

\ '.

\

.,(,

• j

. \

o

. , .' J

f

"

1

. ' /'

1

III. EXPERIMENTAL ,

The experiments of this study consisted essentia~ly of c

equilibrating Cu2S-FeS mattes and silica-saturated Fe-0-Si02 slags

under controlled pressûres of 502' The principal measured parameters

were: " 1 1 ~ .. tndependent yariables J pS02 (0.1 to 1.0 atmos) .

.,,

'1

, Dependent variables '" •

(in coexistent phases)

III.1 . APPARATUS

Temperature (1423 to 1573 K).

Initial matte composition.

Initial slag compositi~n.

Silica saturation conditions

(si1ica crucible).

% Cu ïÏn slag.

++ 'JI +++ ~ (lf,e ra,~io in 5 llLg.

" a.... 1

%. S ~n slag,.

% a in matte. 1

The' experiments were a11 carried out in quartz crucibles

(23 mm 0.0., 20 mm 1:0., 35 to,45 mm long, LaSalle G1àss Blowing, Co ••

~ . Mon~real) placed ~n agas tight refractory tube assembly (Fig. 111.11,

42.

aIl heated in a 3.7 Kw Kanthal-wound resistance tube furnace (Pigs.,III.2.

\. ... Furnace temperature was controlled'by meansl o~ a Pt/Pt-13% Rh

thermocouple s~nsor placed against the bottom of the outer refractory

tube and a Lindber& Typ~ 59545 temperature controiler (Tech-Met Canada Ltd.,

'.

l f

/

! . ~

,10

.. '

_ _ .. _. • .",,_4 ___ _ /'

Fig. II!.l

)

..

- ,

/

• p,

••

'\

" -~ 1 7

E ERIMENTAL REACTION SYSTEM SHOWING UPPER ~AOO) L(MIfR P RTS OF THE REACTION TUBE ;

1. Glass èover. 1

2 Glass cover brass O-ring adapter. 3. 'Silica lance brass O-ring adapter. !'> 4. Thermocouple insulator brass O-ring adapt~r. 5. Gas outlet. 6. 7.

-Copper water cooling jacket. Reaction tube (600 mm long x 70 mm 0.0., 5 mm thickness,) •

8. Ga lance (sflica) (6 mm 0.0., 4 mm 1.0.). 9. ,Alu na thermocoupHt insulator.

10. ,Quart reaction crucible (23 mm 0.0., 20 mm 1.0., ~ 35, to 4 long) • .

Il. 'Alumina halder. \ 1

, l ' fil

..

/'

•

~

fi t '\l

. ( ,)

~. -• .,. .,.

~

. ,.

/

)' ~

# () .-

r(,

, ,

.. (

,l

-'

1 'f

f'

;

.,

/

"

,/

.j

/

. )

. ~

.. ~~-,;. ~~~IM.~ ____ ~*2 __ tW_' __

/ (

5

'" \

43.

~I

/

, ' ,

. ,

'1 J _ l

, , 1 1 1

: ,;

. t l t

,1 i 1

,'1 " . : . ' • .i

d :'-

, , i

~_ 1

:-

\1 " :, ....

,1 cl q ., ,f

r

~

j .. '\

~

~ q

-ma 1 -- /.

.fi'

'.,

2

2

li.

, __ 1 N 2 4i:

...

J

"..

.. "

,)

.. - A

r.:"===ai""""ï.==.i

3

4 1\ II~ ~~' , 8

~' ~ 'i"'Î, F \.

.. Il 1111 ~ ( ~ 19

'\

3

, )

-

5 4

'"" ..... t'~

7 . - ,........ """'. . ...............

!l /'

1 l

FIG.I1I.2 Schematie representntipn of the experi­'mental apparat\ls: 1) gas tank (S02 or N2); 2) microvalve; 3) H2S04 bubbling flask; 4) capillary flow­meter; 5) U2S04 manorneter; 6) reaction tube; 7) vertical~ube furnace; 8) heat exchangcr;l 9) 502 absorption tube.

r

4 )

~/ ~

~ _""

l "

~

CI

b

/

;

'"

.,

\ ...

"'" "'"

•

l

'i

i ~

..... _.:......_-------,~~---------_ .... _---_ .. _--------------------•

,.,)

-"

,.

'1

'"

'\

•

'\

~ -~.

....

FIG.III.3 Experimental apparatus assembly showing vertical tube furnaces, gas train and temperature control1er. ~

, .

~

.JtI

-.

) '.

.,.. VI

( J

.........

1 46.

~ Toronto). The temperature of the reactïon crucible ~easured with a

Pt/Pt-l3% Rh thermocouple plàced immediat1ly adjacent to the crucible. ' l ' _ l '

This temperature was held within ± l K du~ing the course of each experiment. {' ,

The ~uter refractory tube (quarA, mullite or alumina, McDanel l ' 1 1

Industrial Ceramics, ~eaver Falls, Pennsylvania) was closed at,one end. 1 1

It was rendered gas-tight at the open end bi means of a fitted-glass " ..

cover, a brass adaptor ring and '0' ripgs as IlshOwn in F~g. IILL A

second brass plate and '0' ring asse,mbly at the top of the apparatus, t'

permitted entry of a gas lance (for S02 and S02-N2 mixtures) and~ thermo~

couple into thè system. Watercooling of the top 200 mm of the refractory

tube permitted the use of glass, brass and rubber fittings at the top of

the apparatus.

Gas Train

The sulphur dioxide pressure of the gas imm~diately above the , ~.

reacti~ crucible was controlled by passing S02 or a specified S02-N2 #

mixture through a quartz lance onto the surface of the crucible contents" l,

The icomposition of ~he gas was controlled by fixing the rates of 802 and

N2 flow into the lance using microflowmeters (Figs. 111.2, 111.3).

The S02 and N2 were controlled to ± 2% of ~heir concentration ~; checked

by measurements of flows through a water-displacement burette. The totàl

gas flow into reaction c~cible (i.e., onto the liquid surfaces) was / • 1 l 'l

., maintained at 5 cm3 (NTP) per minute. Gas was removed 'from the appatatus

. "

through a hbrizontal outl~ in the glass cover (Figs. IIL1. y2. III. 3) .

f 1 1

. ' , : '

i \

, 1

(

~-"-------/-::O---------'

•

47.

III.2, MATERIALS

gases: Sulphur dioxide, pure anhydrous, ,

% 99.98 S02 (Welding Products Ltd.,

Nitr~gen, ultra high purity,

matte componénts:

slag components:

Canadian Liquid Air, Montr~al). 'Il

1de powd~r (Glidden-Durbee

D~vision of S.C.M., Cleveland).

Perrous sulph;de lump (Anachemia

Chemicals Limited, ~ontreal).

Silica powder (Anachemia Chemicals

Limited, Montreal).

Ferric ~xide (Fisher Scientific Co.

~ Limited, Montreal).

Master Synthetic'Slag

The slag phase in each experiment was formed from powdered

master synthetic slag, thefeby ensuring a common starting point for

41/ allf experiments. e master slag was pre~ared by melting Fe203 and Si02 ,

duction furnace under argon. The resulting slag was cast

and o a-ISO micrometer powder. Its composition,after casting

27.9% FeO, 19.9% Fe203' 52.2% Si02 (Fe/SiOZ - 0.68) . •

111.3 PROCEDURE

Each experiment was be~ by charging cu2S, FeS, synthe tic slag

/'

and silicà powders to"the silica crucible in proportiQns which would

produce the matte and slag compositions desired forthat specific experi-

ment ... The crucible was then placed in an alumina h'older near the bottom

""~~""~~~~I_~4n ... " ..... ",~~~ .. _ ..... :.:-

/ 't

(

(

1

.. 1 _ ...... _________ ._A. ___ _

48.

of the refractory tube (Fig. 111'.1); the system was closed; the lance

and thermbcouPle were fixed- in place and a flow of nitrogen was directed

through the lance to displace air from the system.

After approxi~tely 1 hour of nitrogen purging the nitrogen

flow was switched to the prescribed S02-N2 mixture for the experiment,

the tube ~urnace w:: raised up ovar the refractory tube and heating was il<

begun. Heating was carried out very slow1y with 7 to 8 hours being ~

allowed for attainment of the experimental temperature. The flow througfi'-,

the silica lan~e into the reaction crucible was ~intained at 5 cm3 min- l

during heating and at 5 ± 0.1 cm3 min- l once,th~ experimental' temperatu~e

was reached. This low flowrate ensured that the surface of t~e slag was

blanketed by S02 of known partial pressure but that turb~lence in the

liquids was minimal.

Each experiment was/ conduct~d for 20 to 30 hours at temperature,

to ensure a close approach to equil~brium. This point i5 discussed in ,-J~ '-

detail i~the experimental re5ults section. At the end of the experi-

ment, \the furnace power was shut off, witl1- the gas ,flow be.i;ng maintained. . ,

It is estimated that under these conditions freezing of the liquid5 wa~

complet~ in about 10 minutes, i.e., t~e temperature of the crucible fell ....

ta 1073 K in about this time. 1 ~

After the system had cooled ta '500 K, the furnace was dropped

below the refractory assembly and the reaction crucible was femoved /

from the apparatus. The crucible was gently Broken open and the slag

and matte layers were separated for chemical analysis. In general, the

layers separated very easily due ta a lack of mixing between-the ph~ses'

-'l and ta their difference in' freezing temperatures.

1 J

(

( , .'

• b

- (

III.4 OlEMlCAL ANALYSIS

-i'The samples were ground (separately)"to -ISO ~m and analysed

quantitatively for: .-li) matte Cu

,- total Fe \

0

, 5i0

2 , -/ slag Cu

\ Fe++

Fe+++

5

5i02

49.,

Attempts were made- to analyse F:+ and F:++ in matte but

copper at high concentration interfered with the1 analyses and prevented

aquisition of any meaningful results.

The fol1owing analytical methods were used:

1. COPEer: 'long iodide ' methodfdupÎicated br r.tomic

absorption (after dissolving in HCl and HN03).

2. Iron (total): dissolution of sample in HCl and HF, , .

reduction of a1l iron to Fl+ with SnC12; and back titra-

ti~n of the Fët+ against potass~um dichromate.

3. F ++. 1 e ln s ag: dissolution of the s~le into boiling

H2S0d and HF with .the crucible (platinum) cover,ed to-avoid ++ Il' -' .~.

oxidation,of Pe . The resulting solution was then diluted

and titrated with standard potassium dichromate solution ....

(sodium diphenylamine indi'cator).

î

/

(

, .

ft 1

/

(

4.

s.

\ ---_ .... ,--_._.-~---- ~ - --.- .-. -

50.

1"" Fe+++ in slag: dissolution into boiling l:~, HCI fOltowed

by hot titration (to clearness) with a standard stannous

chloride solution.

Sulphur in slag: 'barium sulphate' metho~ (dissolution IP

of sample in HN03 and HCl with quantitative precipitation

of sulphur as barium ~ulphate) duplicated by analysis in.

a LECO Model 762-100 sulphur ,analyser (LECO Instruments,

Ltd., Mississauga, Ontario). The latter analysis was 'madé

by mixing a small slag sample with pure iron and by treating

the m:lxture like a normal iron or steel s'ample (combustion

to 502' infrared S02 concentration measurement).

6. Silica in matte and slag: dissolution of sample in boiling

HCl and HN03 and oxidation with perchloric acid. Fo11owing

evaporation, the soluble salts were dissolved away from the ~. 1/

Si02 component 'and the dried Si02 residue was weighed.

7. Oxygen in matte (deteTnrlned by LECO CModel 775-100) oxygen 1

analysis): In this case a 0.1 gram sample<of matte was . ~

wrapp~d in aluminum foil of known oxygen content, and the comb~nation was 'melted inductively in a stream of hydrogen.

The LECO instl"Ument quantified the amount of oxygen by

infrare9 detection of the H2b formed from exposing the . \

mel t~d sample to the hydrogen stream. \ .

The analyses were normally done in tri~licate wi~h a repro-

ducibility generally wi th in ± 2 or 3% of the resul t.

,

~ (

(

,A

)

-----*_._.-------------------

51.

~RESULTS The principal experimental results are tabulated in Table III.l

showing pS02' temperat~~and chemical analyses of the coexistent ph~ses. \,

Mà.t~e/Slag Equilibration Tim~ "" The experiments were aIl carried out at temperature for 20 to

2S hours. For comparison, preyious investigators have carried out (under

somewhat different condit~ons) their experiments efor the fo11owing

periods of time.

Equilibration JI Workers System rime ~ TemPerature

Cu-Fe-0-S-Si02 Geviéi and Rosenqvist (18) (copper phase present) 5 to 17 hrs. 1523 K

Johansen, Rosen3vist and Torgersen (1 )

Cu-Fe-0-~-Si02 ," (co~per phase present) 28 hrs.

1400 ta 1700 K

Nagamori (9) Cu-Fe-0-S-Si02 3 to 5 hrs. 1473 K

Bor and Tarassoff(ll) Cu-Fe-O-S-Si02 3 hrs. 1473 K

Toguti and Santander(15) Cu-Fe-O-SiOZ 48 hrs. 1573 K

It can be seen that the equi1ibration times of this'work are

in the same range as those used prev~ously i~ similar studies. The

suitability of these eXperimental times was tested by carrying out a

group of experiments in which pure Cu2S was equilibrated with master,

" slag for various lengths of time. The parameter examined for invari-

ability with time was-Wt% Fe in the matte phase. The results ar~

summarized in Table III.2 from which H i~ apparent that steady condi-'

tions are achieved within 20.to 2S hours of contact.

1

•

y.

•

r ~~ -~ ... _~~*" .--- ,.b iii:

\- '" ..,,52.

, ' ....

( 1

\ ) J III s:: J-o

..-1 ~ . ----4_ - .... - ••••• .......... .. ........ .............. ... ~~-t., ..... '"':-:~~-:~ . ........ .-

1 ... .: . ......... ... ..:.: ...... .:..;. . ..; .... .:. ......... . ..... . ...... . ... .:.: '1j s:: 0 u ... ......... _ ...... - ~"4!~"':J ................. ~"': ...... --:":~., ................. ;1 ! , i=:i== ====:it ;;:I::.~ .lià~à:i::ir:== == =:t~1Ï =:;=. :i:i==::i:i ,~'=;hi l, 1

i 1 1 *t ~~~~~ ..... -... ~":~"":to: ~"!"-;~~~-:--:~~ ":~ -!~~., "';~~~ ........... ....... ......... ...... ..;.;. ---- ... _ ............... ~- ... .. ....... . ....... ..;.;.;...;.; ... .;..;"' .... f,z

s:: 0 ,~

.j.I

«1 J..t =' .j.I

«1 U)

«1 U .~

o-t .~

a . ~~--:-: .......... -- ........ ..................... •• _ .... ...... .. .......... . ...... ... ........ ....... ............ .... ..ti . .. .... .. ..... .. ..... 1 ......... U)

J..t ~

'1j

• J ...... , ........... ....... ................ .. .. ....... -..... . ..... .. ........ .. ..... ..: .. . . ..: .. ..;. ...... -..~ .... ..: .. .:~ •• ..... ~ ... ....:.:..;",,, .. .: ... • s:: .~

=' III " .j.I

.r 1( ••••• ....... . ..... .~~"!-:-:~~-:~ ..- ~~~-: .. ~ ....

~-:~-:-:~ . ..... "

.... ..: • ..i •••• . . .: .. ............ -.. •• ..... -" .. . ....... . . .: .. r-!

=' III ~ ~ ,

j ........ _.-... . ..... - ~~~.~~":~~~ .. -:-:~~ "'!~~~ ~":":-'!~~ ......... ... ;:==== =~=~== ;;:::.:= ....... _ .. ""' ....... " .. .., ...... • "N", _ .... "' .... :i~::;l 1 ..................... .. .. .. ... "" ..... " -.......... • .......... ... ..... .. ..... ... ... 4 .......... ~~

_ ..... .. ... " ...... .. "!!:== ....... ~ - ~~~"':~~ ........... ~-.. === ....::=:=~.=' ..:=== .. ;:" • .... i., J

.... _ ....... ... ::-=;:; •

a . .-....... -- .......... ..... ,.. -- ... -- ........... . _ .... ..- .......... .......... .. i:::i== i=;;;; ::i::::o ==i;;::::=ii == i==d '::::J:6 iii;:i '''' ' , , J ;: ... s;=

5555& a5SSES ..... " ..... :::::n:;:::::~::::= == =::::! =~::!= .............. =~=r::: le ::::: M ............. ................ ~ :: :::: :::: = .... _ .... ===== .. _ ... ,.-- .. _--- ... ... -...... _ .... -- ---_ ...

"i ~~~~~ .......... -..... -- ~~~~~-:~~~~ ......... .._ .....

~~~4!41!, I!~II!~'! oS

i ----- ....... ••••• _ .... -....... --.. .... -i •• • ••• ------ .......... ' ..

k ........... , ...... :: ......... ===a;:;::= •• :::; lut"; :::;Ulll :;~a.;= ;;.a~ -_ ....... " ..

~

d

(

, J

.... _i\b'_ ... <.~s,,"WI. $.*'._0$. "'il~~~~~""""-"'~ .""--l'" \ té"

,-__ ...... 1 __________________ -, - -

53.

( " ')

Exp. If pS02(Atm) TeK) 'time (hrs.) '.Wt Fe (matte) , . >.

1 1.0 1423 10 0.3 2

, 1.0 1423 20 0.88

3 1.0 1423 , 2S 1.05 4 1.0 1423 5S r 0.92

S 1.0 1473 10 0.6 6 1.0 1473 .. 1S 0.95 7 1.0 1473 20 Il 0.87 8 1.0 1473 30 0.92

9 1.0 1523 10 0.33 10 1.0~ 1523 15 0.98 il 1.~ 1523 20 0.92 12 1. 1523 25 0.99 13 1.0 1523 ':;0 0.89

14 1.,0 1513 rO 0.83[ ~

lS 1.0 1573 1S 1.25 /

16 1.0 1573 20 1.35 17 1.0 1573 25 0.94 18 1.0 1573 30 ·1.10

[ l-t

, , TABLE III.2 Ex]2erimental Resu1 ts for Matte/Sla~

Eguilibration Time. EXEeriments \~ under SiO,2 Saturation Conditions.

0

1

"

.' i

(

..

~ .

1

(

7

(

---..... "._--

t, 54.

IV. DISCUSSION

This section discusses the impli~ations of the :~erimental

results under the headings:

IV'.l Thermodynamic interpretation of resul ts . r.

r

IV.2 Behaviour of copper in slag.

Iv.~ehaViour of sulphur in slag.

IV.4 ehaviour of oxygen in matte.

IV. ' tructural representation,of results.

IV.6 Prediction of conditions under which a

phase will occur. v

IV.7 Industrial implications of the results.

IV~ 1 TIlERMODYNAMIC rNl'ERPRETATI~N OF -lŒSULTS

..

separate copper

The principal objective of this thermodynamic interpretation

was to determine:

acu S 2

aCu 0 2

for each experimental equilibration, based on measured values of pS02'

matte composition and slag composition. Additional parameters cal-culated

along the way were p02 and pS2'

1

Before describing the calculations, it will be worthwhile to inter-

pret the present and previous matte/slag/gas experiments in terms of the

phase rule.

..

': 1

~

"

1

ri'

(

"1 ,

(

------------------ .--

/ 55. ,

1 1 ,

The experi;~ental ~ystem described in this thesis consists of , ' 4 phases (Si02, two lfquidS, gas) and 5 components (Cu, Fe, 0, S, Si0

2).

Br the phase rule (P A 'F - C + 2) ,. the system \~S three degrees of \ ,~ 1

freedom. Thus. once t~mperature, pS02 and matt~ grade are specified as , ,

/ ' 1 independent variables, ~he remainder of the SYitem is fully ~efined.

The experimental apprloa~ in this work was es~entia11y, therefore, to

fix temperature, '~o2\ a4d matte g,rade (t Cu) Jnd to let "the slag find \ :

its appropriatel compositaon (particularly with regard to its Fe++/F~+++ " ~ ~~

• \1 \ l , ratio and su1phur cont~nt). J, •

1\:' ~. . (12) This approaeh. ~s not di~imilar to prevI us work (Themelis • (10) . 1 (11) j -,

Yaz!wa ,Bor and Tara~soff ) in'which temperat re, iron-saturation

and matte grade were !ho$~n-a1 i~dependent variables and in which slag l.'I ' , ~

and gas compositions werd,\: allowed to find their appropriate values. The \ :

ess~ntial difference b~tw~~ these previo~ iron-saturation experiments

and those described here ~~ ~e fixed, relativelY,high pS02

work.

of the. pres.::~ ) ~ i\ ~~

Geveci and Rosén~vist (lB? also proceeded in, a somewhat simi1ar "t

manner but with metailie c per as a separate phase. Their system had 5

components, 5 phases (meta~

material which was Fe, Ee30

ic eQRpèr~;matte, slag, gas and~he container , 1

or",,:siOZ) and hence on1y two degrees of

freedom. Thus, these worke s fixed temperature (1523 K) and matte grade: -\ -

and allowed the gas and liq ~. d, phases to find their appropriate comPositions.

The present work o~ be considered as a further: ~xploration into

mà.tt'e/slag/gaJ systems uncier\\conditions ~onsider~bIY different (par:i~-1\ '

br~y with ~ect to pS02) ~:natL those of prèvi ... ous studies. .. " \\ .....

\\ • 1

,1 ".

.' t

(

./

, ,

)'

, 56.

IV.I.A E~timation of pOZ

, The first step in the t~ermodyna.mic ,interpretation of the resul ts

• was to estima~e p02 from Ci) th~lag cbBlposition of' each e~periment an~ii)

the ~e-O-Si02 oxygen pressure data ~f Muan (20) (Fig. IV .1). The slag~

/n th~. pre~ent wotk were aU very n~ar silica saturation (having be,~

equilibrated for 20+ hours in a quartz crucible) 50 that the oxygen s:-

pressure in each experimental sy~tem was,determined in accotdance wi~h

~ l.ts F~++ /Fe+++ ratio along 'tlie silica saturation Une in

Of cpurse the dat~ ~f Muan(20~ i5 for copper-

Pig. IV~l. "

and sulphur-free

slags 50 that ~ome error is int~oduced in this interpretation. However,

the total of copper plus sulphur in the sbg was usually J.t1ss ~an st • 1

so that the error is likely to be small. .' , ,

Thé results of t~~ calculations,are presented in Table IV.I. $

~V.1. B Calcula-çion of pS2 '-. . ,~

. .

Once p02,had been estima~ecl for each equilibrium. pS2 was

calculated from it and the appropriate experimental value of pS02 by mearts of

the equilibri,um e~S5'iO~S:

~/. ;

1/2 S2 + 02 ~. SOi

6GO -361000 + 12.4 T (joules) IV.l

'., ,

,

'fi'

~--- p'502 6Go (

and - RT ln '. PS21/2p02

IV.2 .

1. ...

, /

./

, .

.. l

~ :çq;; '_4t{ 1<4. "t"c -"'~ __ ."' _________ "_4 ___ """ __ ""'_

t ~

~ ----'., '~,) .,./ (-'.---

<'

-t' 1 ~

~ ~ ;0. '. ,

, ----f"" -;&--)

••

, ,

'\

~

.. , T -

.. .. "r-

"

t!>

_ 0 i(,fb

0\·

.. 7'

?-

r

.slOZ

_ 0\°

20 ~O # % Fe

2-0

3 ~fi

,:

"-

.. ...

'\

~,

,

.. -::?

;,

---" SiO~ -

~,

0 i(,fb

__ J~

0\°

, V Y V le X 'X ) .. Fe 0 1 1/ -.. v .ac y v y 't .... Fe 0 FeO , 20 40 60 Z 3 ,FeO " " An ~I'\ 2 3

% fe.Z

0 3 ) ,

FIG .. IV.l

-~--, Phase diagrams of the FeD-Fe20~-Si02 system showing equilibrium oxygèn pressurcs in the melt f~clds (~~n(20)) . ..,..., '-

.;4

"

CIl "-l .

Il 3 m 1 - ~ '7'3 II n III _fi "Nil. 1 n F A

,

(

o

4 ~ ,

) , 1 / '-' ' ,'"

, " , '~'-" , .

t -''\ î t l,

J 1 j l, > t .

\ . ( 1

1

1

or

'?

l'l'' ,

..

58.

\ As Table IV.l shows, sulphur pr~ssures in the experiments varied

-5 -Z ...... from 10 to 10 atmospheres.

IV .1. C Es tima tion of aCu s- .' 2 t

~ ~ivities of Cu2S i~ the equ~librated matie phases were esti-

mated from (i) each experimen(al matte co~osition and (ii) the CU2S-FeS

activity data of BaIe and Toguri(2l) and ,KriVskY)ind Schuhmann CZ)

(Fig. IV.rr Activities determined in this ,w}y are reported,in Table

IV.l-and tF~g. IV.3. Bal~ and Togu~i (21) suggest that the Cu2S-FeS pseudo

C" " 1 f' binary system is weIl described by temki~'s (2 ) mod~l 50 thatrno tempera-

/ . f,' "\.. ture adjustments were made in choosiny the aCu S v~lues.-

. 2 This determination suffers from the same problem as the

1 1 ~ 1

ddtermination of pOZ; i.e., the matte phas! contains a small quantity of 01.>

oxygen (and a trace of Si02) in addition to Cu, Fe and S. However',

components totalled less (xcept 'in a few cases- th,e o~gen and Si1\ca than st 50 that the error is once again not likely to be large, especially

D

with high matte grades.

IV.l.n Calculation of aCu-

The activity of copper in each expe~iment

fr~m the values of pS2,.-and aeuil:- in sections IV.!. B

·l·b· \ .~ ( 'eqUl l rlum expresslons:

il! " 1 .. 2Cu.t, + 1/2 S2 :t CU2SR, , g

\~Go _ -171000 + 58.6 T (joules)

.

( .

\ was readi1y ca1cuiated

and IV .1.C u\~ng ~e,

*,IV.3 •

This e:x:pression is in fJenera~ agreement 1.t1ith the reoeht data of BaZe and Toguri(2~) + 500 JouZes per moZe of Cu2S.

\ /

(

l '

...

(}8 /"

+ ., "

0

1 ~ ,

1

t

0.2

l t $

~' 0·0

, , ,

F~G.IV.2

1

1 ,( •. ,il

.(

(

•

59 •

... . '

, 1

• T06URI(21) - --KRIVSKY(a) - ·-TEMK1N'S MODEL

J .

(22)

... ,1473 K

~ \

0,4' 0·6 0·8 1·0 . X Fe S' ~ ~.

~/ Activity of CUlS in Cu-Fe-S mattes as determined by ~rivsky(2) anda.."etill(21). Activities predicted by Temkin's cat~onic ipnie lattice model(22) are also shawn. ,

1 •

1

1

; C 1

!. ,1

1

[

(

" 1 1

Î

~ .

(

- ---"--.-------------------------------, l

60.

( and the equilibrium constant:

( \ IV.4 1

The cal~ulated copper activity ~ues, are shown in Table IV.I and Fig. IV.3.

/rv-.1.E Calcplation of acu2n- ~ ~

As a final step in the thermod~amic interpretation the àctivity

" of Cu2

0 in each experimentàl system was ,calcu1ated.from the p02 and :eu. vaiues d e:,erm~,d in Sections IV. 1. A and IV. 1. D wi th the aid of the

equi1ibrium,exPressions:

and ,

-+ 2Cu~ + 1/2 02 + Cu2ü.t

g 6Co - -144000 + 58.6 T

K -E co )2C 0 )1/2 aCu . p 2

/

(joules)

/

. The calculated valu~s are presented in Table IV.I and

~ . ~ F~g.. IV. 3. They~are discussed in Section IV.2. -

IV. Z BEHAVIOUR OF COPPER IN MATIE AND· SLAC

IV.5

IV.6

Pe~haJs the mo~t important results of this work are presented

in Figs. IV.4 and IV.S, both of which plot wt% Cu i, slag vs. wt% Cu in , /

_ matte. "Fig. ~V.4 ,present~ aIl the, 'raw' data of the investigation while

1

1

I!IP

Fig. IV.5 cpmpares th~ 'hest fit' liue of the present' wqrk with the~es~ts

.. "of previous investigators. ~~ latteI' show~ clearly that the pre lent f' '

,,'

1 ____ ..--~ .. Itjf_~_""~ 'I"'_-r:.-r--........ _' ,,~ ...... ...-~ ____ ~ ... ~. - \ __ • -- , -..... !Qf.Z!IIIl!!:

\

,.-...,. ,....-.... .. -, ....

.. •

~/ -

~ ~

Exp • .., pS02(atmos.) T(K) \,Si02 Fe++ IFe+++- (weightj - p02 (atmos.) pS2(atmos.) Xeu S aeu S aCu a 2 2 Çu;Zo

.... -9 -3 ..., 1

9.3xI0-5 1 1.0 Il 1423 38.0 5 ... 2 5.0xl0 5.1xl0_3

0.94 0.93 0.09 2 ... J 1.0 1423 rt37 . 6 4.4

t. 6.4xl0-9 0.70 0~72 8.7xl0- 2 9.5X10-~ 3.5xlO

3 1.0 1423 34.7 5 ~1 ' S.5xlO-~' 4.2x10=~ 0.58 0.5Z- 7.1XI0:; 6.0XI0:S 4 1.0 1423 31.6 5.9 .r 2.5xl0=9 2.0xl0 0.54 0.49 4.6xl0 1.7xl0

5 1.0 1423 31.3 6.0 1.8x10 S.6xl0-2 .. 0.66 0.62 4.1xl0- 2 1.0xl0-S

.. , 1.0x10-9 3.2x10-2 0.92"- S.6xl0- 2 1.6xl0-~ 6 0.5 1423 38.5 6.4 0.93

7 0.5 '"" 1423 37~7 5.3 3.0xl0-9 f5 -3 0.91 0.90 9.7xl0=~ 3. xl0 2 8.4xl0-6 8 0.5 1423 34.3 6.7 8.OxlO-10 4. 9xl0-2- 0.68 0.65 . 4.2x10 8.3xlO- .

9 0.5 1423 34.0 7.1 6.0Xl0-10 8.8xl0- "0.49 0.41 2.9xlO-2 3.4xl0-6" 10 0.5 1423 28.1 4.6 7.0xlO-9 6.5XIO:: 0.40 -.. 0" 25 7.8xl0-2 8.3XI0=~ 11 0.5 lA23 27.4 4.6 6.8xlO-9 6.9x10 0.40 0.24 7.Sx10- 2 7.6x10 , , -

5.5xlO-9 -5 -3 12 0.1 1423 :n.8 5.0 4.2xlO_s 0.97 0.97 0.30 I.JxlO -3 13 0.1 1423 35.6 5.1 5.3xl0-9 4.5x10 0.94 ' 0.95 0.29 1. Oxl!) 14 0.1 14'23 33.0 5.0 5.8xl0-9 3.8xl0:~ 0.70 0.68 0.26 8.4XIO-: 15 0.1 1423 <- 30.v9 7.7 S.SdO- lO 3.8x10 3 0.49 0.40 6.4x10-2 1.6x~0-S 16 .. 0.1 H23 34.7 7.7 S.8x10-1O 3.8xl0- 0.52 0.44 6.7xl0-2 1.1x10-

l 17 1.0 1473 39.7 5.4 7.6xlO-9 1.6X10=~ - 0.69 0.68 7.2x10-2 -5 4.8x10_S 18 1.0 1473 39.5 5.2 4.6xl0-9 4.Sx10_

2 0.68 0.65 S • .fx10-2 2.2x10_

S 19 1.0 1473 - 30.2 5.4 7.6xl0-9 0.48 -2 1 1.6xl0 O.5~ 6.0xl0_2

3.4xl0_4 20 1.0 147-3 36.7 4.8 1.5XI0:~ 4.3Xl0:~ 0.61 0.56 9.1xl0_2 1. OxlO 5 21 LO 1473 39.0 5.1 5.9x10 2.7xl0_

2 0.62 0~58 5.8xlO 2.8xlO-

22 1.0 l47~ 38.4 5.6 4.5xlO=~ 4.8xlO 2 0.60 0~5S t(4.9XI0:~ 1.1xI0=~ 23 1.0 1473 33.0 8.1 4.7xlO 4.4xl0- 0.51 0.42 4.4xlO 1. 4xlO 24 ~1.0 1413 4.5 8.2x10-9 1.4xlO-2 0.43 0.30 4.9xlO-2 2.3xlO-S

.i. 4 -..-

25 )~ 14-r3 7.9 4.7XI0-~ 4.4110-~ -----i}.38 0-2 7.5X10=~ ,~ 0',22 3.2Xl_2 , 26 l.p l 3 ~.8 4.5xl0- 4.8x10- 0.33 0.12 2.3x10 3.8xl0

} . ... • .•.. (cont'd.)

- '" TABLE IV.l Thermodr!!.a~l~_Qùanti~tic:lJi Estimated and Calculated from Experim~n'tal Data ...... .

~

" -:-... m----.- ----:;.-, .... - ... , • If • 'm = nt'· tt '1 '1 tt et w ' r 'rr 7 F lM' 1.

'"

'. ..

i

-~ ....... ~ ~ ~~.- .. -....-~- .............. '> .. ..-<---~~~. "" ". ~,.~

~

TABLE IV.I (cont'd. )

Exp. , pSOZ(atmos.) T(K) " SiOZ

27 0.5 I4~3 38.5 28 0.5 147~ 37.7 29 0.5 1473 38.8

'30 0.5 1473 34.9 31 0.5 1473 34.3 32 0.5 1473 35.1

33 0.1 1473 39.7 .34 0.1 1473 ~7

3S 0.1 1473 32.5 36 0.1 1473 30.7

37 1.0 1523 ~8.9 38 1.0 1523 34.3 39 1.0 lS23 35.0

·40 .. 1.0 1523 38.0 41 1.0 1523 32.3 42 1.0 1523 31.2

43 1.0 1573 36.6 JO 44 . 1.0 1573 . 39.0

4S 1.0 1573 33.8 46 1.0 1573 34.6 47 1.0 1573 32.7

,

~ -~."'- ..... ~ ..-..... ...... ... - ...... - -

~ ...........

'" '\

\.

... ... , '\

Fl+ /Fe+++ (weight) pOz(atmos.) , pSz(atmos.) X a aeu aCu 0 Cur_ Cu2S 2

5.7 6 -9 7.9xlO- 3 0.97 0..96 0.1 7.9XJO-! 5. XIO_9 ... 4.9 -3 0.95 0.95 0.12 7.6xlO_9 4.2x10 2 1. 3xl0-

5.5 0.84 0.85 -2 6.2xIO=~ 4.9x10 l.Oxl0- 9.0xl0_2 5.2 . 9 4.4xI0=~ 0.61 0.57 , 7.4Xl0:9

9.0xl0 7.7xl0_S -7.0xl0-2 5.8 5.6xl0 7.8xl0 0.49 0.40- 3.5xl0 5

!L8 3.9XI0-9 1.6x10-2 0.43 0.26 4.0xl0-2 1. 3xl0-

5.7 5:.6X10=~ -4 0.97 0.96 0.24 -4 3.1xl0_3

4.4xl0_4 7.2 3.0xl0 1.1xl0 -2 0.79 0.78 . 0.15 1. 3xlO 5

8.8 9.3x10-10 l.lx10 0.48 0.40 6.0XIO:~ 1. 2xIO-5 5.8 3.0xl0-9 1.1xlO-3 0.39 0.24 8.0xlO 4.1xIO- -

4.5 2.9xlO-8 -3 0.74 0.73 -4 -8

-8.2x10_3 0.11 1.5xIO 4 4.3 6.0x~O_8 1.9xlO 0.67 0.64 0.15 4.lx10-5.2 2.2x10_ 8 I.4xID=; 0.77 0.76 0 .. 10 1.1xIO:! 4.9 la. 4.1xl0 -8 4:~xlO -3 0.57 0.50 0.11 "'1.7XIO_ 4 4.7 7.3xlO_

3 1. 3xl0 3 0.51 0.42 0.14 2 3.9xl0 5.0 4.1xl0 4.1xIO:" 0.37 0.19 6.0xI0- 6. Bxl0-5

3.8 -7 -4 0.86 0.85 0.43 -3 5.9x10_7 1. 2xl0 5 7.0x10_ 2 3.7 7.4xl0_ 7 7.5xlO:4 . 0.87 0.87 0.48 1.OxlO_3 3.9 ,6.2Xl0_7 1.lxl0 -4 0.83 0.82 0.43 7.4x10_4 4.8 - 2.3x10 7 7. 8dO_3 0.51 0.44 .().19_

2 9.lx10_S 5.4 1.lxlO- 3.2xlO 0.26 0.08 6.0x10 S.7xl0 ~ --- -- -- -~--- --- -- ---

..

) ~

...

0\ N

,. t: 1'1·': 7 tlll! .t:<mr.'IIIZHII ••• I?2 TE P : If \ 7 'a' q; 1 1. l

(

1 i

, Î

._------------- - -1

~ 63. '

0' '101 C CI)

c: 10 9 • .a..

Si 02 SATURATION • .. .... • • .JI ..

CONDITIONS ~ fo- 0 • • 0 !6 fi!- 10~1

QO Q Il o • Q • • ... tJi7_ .. ~o ... Il

-- J\. •. ati: .....

/'

:::s U

o

1 1

1

....... '

. 0 .. Q • Q QQ

~JIA ~.o~ .., • .0· 00 · (:) ••

.." Q~ ,

A Il~'" (:)

0·8 ..... otS;IiJ '\

~ ~

~, , 1-en

N :::s U 0·4

0, 1-

;çpI ~J -

fo- ......, • : .. 1

0·0 .. ,.,

--.,. \ FIG. IV.l

" 1

0-0

<:

. (,

. .. Q~ , • • • • , 1 , 1

0·2 0.4

X CU S 2

.. QQ

~ Q

~ O· 0

• • • • •• • i~ Il 0

l 1 i 1

0·6 0·8

in matte

00..

• 6Yf-.-

Il

1

-

10- 6

\·0

Experimental values of "'t'a Cu in shg, aeu' aCu Sand aCu,O",as fun,'O!15-of Xeu s in matte"silica saturated system. 2 The upward tTend" aIl variables with increàsing matte grade is notable. The experimenta / conditions were:O 0 0 1423 K. pS02 - 1.0, 0.5, 0.1 atmos; ••• 1473 K, pS02 -1.0,0.5,0.1 atmosjQ.pS02 -1.0 atmos, 1523, 1573 Je.

..

1- ..... ""~~~ ... - .....- fOI ....... "

"Y

64.

,; \

~

"-(

. ,

/'

7.0

Si O2 SATURATI?N

" 6·0t- CONDITIONS

r

5·0 - .. CI 0 -en

4·0 r- .,; A

C 1

--::; (.) 3-0 r-

• ~ ,

. '\ 'il 0 .'\1 • 1

2·0 , l'ô r-/ ;, • o.

~ . ~

'·0 • 1 0 ", ( - o .. f' 0

• • i QJ:;7 0

• '0 .6~. · .fJ ,1 ,i~~ • ·~oo. 1 1 ~,

30 40' 50 60 70 8'0-1

0/0 • VI

CU ln matte 1

. \. FIG.IV.4 Exp~rimental coppe~-in-slag results, silica-saturated system. The-

experimental conditions.were:OÔ01423 K, pS02 -1.0, O,~, 0.1 " a,tmos; ••• 147l, K. pS02 - 1.0. 0.5. 0.1 atmos;"9 pS02 - ].'0 atlIIOs. 1523, 1573 K. 1

( . '[ '1

~

'".

,,-...,

'",-

,

:i"

/

~

'" ~

~

- THIS WORK <SILICA-SATURATED SYSTEM)

5.àl 0 SPIRA AND THEMEUS (12)

-----rAZAWA AND KAMEOA liRON-SATURATION) (10) '" ~ • ~ 0 - - .

SEHNALEK AND IMRfS( IRON-SATURATION}( 23)

--0

1 a

4·0. Q NAGA~RI ( IRON-SILICA-SATURA TlON ) (9 )

4\,

l' GEVECI AND ROSENOVIST UftUCA-SATURATEO SYSTEM)( 18) - 1- • 0"

0 ." v -UL3·0 ..r c --, -:::J'

U 2.0 "

~ o .

':0 1

Q1 vV 9 V

Q1 V(p

0 20 40-"

60 0/0 Cu in matte

Ji- - -;

FIC.IV.S Experimental Cu-in-slag results of the present work compared ~ 1 with those of prcviou8 invcstigators.

'. .d-z::: . - - · 'Q. • ,':' H •

...........

Jo

"

•

A

B· &.

80 0-V1

~ -'-

~

"

i \ -, ------., .--X""-

( 7

- " .-......• ~----_ ... -_ .. _ ..... -. __ -:..ç -~- -- ....... -. --

66.

"-

results are in good agreemertt with those of previous. workers.*

A notable feature of the present work and that of Geveci and

Rosenqvist(18) 1s the sharp rise in copper-in-slag content,at matte

grades above about 60% Cu. Moreover ~t can be seen that under the

highly reducing conditions of iron saturation (Fig. IV.S) such a sharp

increa5e does not apparently occur. These two observatïons suggest that \ "'

whereas copper-in-slag concentration i5 not appreciably affe~ted by

oxidizing potential (as represented by iron saturation conditions and 1

,high pS02 conditions as extremes) at matte grades below 60% coppet, the

effect becomes very noticeable for mattes above this grade.

IV.2.A Thermodynami><~nterpretation of Copper-in-Slaa:eoncentrati~ns 7'

Fig.~I~.3 shows weight% Cu in slag and the activities of Cu,

..

Cu2S and Cu20 cal~~lated in Section IV.l aIl plotted as ~ function of

mole fraction-of cu)s ~ the matte. All.three actLvities increase with

matte grade so it tomes alno surprise that ~pper-in-s;ag concentrations

s~id a1so i~creas,· with increasing matt~ grade. Interpretation of

these data is continued in Section IV.S.

'jil

IV.2.B Effect of pS02 on Copper Behaviour

At matte grades below 60% CU it appears (Fig: ~V.4) that pso2

has no appreciable effect on copper-in-slag concentrations sa tha't as

long as matte grades are kept bèlow that value there should be no """.

whelming copper 1055 problem with high pS02 processes such as the

flash furnace. 1

over-

:of INCi

*' The e:qJerimentaZ resuZts of Nagarnori(9J (Fig. IV.S) shobJ somewhat Zargez. Cu-in-aZag concentrations than previou8 lùOrker8. There is no apparent reason for this as he toOK. great actPe to en8UJge that thel'e ws no physiaaZ ent!'ainment of matte in his sZags.

, 1 .' J

,--, -, "-~~------ - ~_. . _. --._- -------- -

(-

..

-----_.,--------_ ..•.. -- '- '

67.

~ Above matte grades of 60% CU J the experimental resul ts become

confused, probably because a "smal1 error in ~e determination~f % tu in 1

matte terids to appreciably distor~ the p~cture. '4~ ,

It 15 clear. however, that unde~ the oxidi~ing conditions

encountered in smelting processes (i.e., non-tron saturation conditiqns),

production of mattes containing more than 60\ copper will always lead to " <f'

large copper-in-slag losses':

IV.3

1 ;'

BEHAVlorR OF SULPHUR IN SLA~ ,

Fig. IV.6 shows the experimental values of wt\ sulphur in slag

and the 'best'fit' line as a function of matte grade. The experimental.

values of prev~ous workers are aIso shown.

The clearest point e5tablished by ~helse results is that the

501ubility of pure CU2S in silica-saturated Fe-O-Si02

slags i5 very

limited, the solubility of pure FeS is~quite high (equivalent to 6(32)

or 7(10)\ S) and that the so1~bility of Cu2S-FeS mattes

mediate between these extremes. Somewhat surprising1y, .\~

seem to be a discernable correlation bet~een % S Jn slag

or pS02' J

inter-

does not

temperature

c Fig. 1V.6 sho~fthat for mattes containing less than a~out 60%

Cu, the sulphur-in-slag contents measured here are slightly higher than

those of recent investigators. However. there is considerable disagreement ....

between these workers themselves 50 that a clear-cut picture of sulphur "tr <

~olubHt ty a t low matte grades awai ts further e,xperimentation. ' On the

basis of this work, it 1s suggested that LECO ana1ysis (Section 11I.4)

for sulpllUr is a rapid and accurate analytical, technique for this purp05e.

"

\

"," __ . .1

..

li

f

~

~

_ ... ~~~ ...... _ ... ,..~ .. #1iI .~""",~!O!J '\!II ...... ~~"',...",.....-._ .. _~~ ...... _,. _~_ ... __ • '"' __ '" _ ...,.;;_ v __ ~_ ~

,.r--..

, ~

...... ~ .

D' o 8·0 ... -(/) t-c:: --

en 4·01-

0 è)'q t-

0·0

D' c,8·0 -tJ) '-

c: --en 4·0

~ 0 ~ ..

0 "

-'

.. " ..

'\

EXPERIMENTAL POINTS (THIS WORK)

~ --t

0 9

ç/l! • '\

0

• Àoe • ·0 A 0

DQcxPj.d~ 0111 J 1

, ,L 1

A MACLE AN ( 32 ) ~ TH1S WORK ( 1

GEVECI AND ROSENQVIST (18)

~ t 6 YAZAWA' AND KAMEDA no ) ~ 1 ..

NAGAMORI ( 9 )'-

~ ~

;r', \' -.

0 0-0,,8

" Ô. ô. )

101 20 30 40 50 60 70

PIG. j. Sulph= ;O)Ub~r.~,n f.u .. ~ ~un~o~ 1} !,~ grade.from thl. ~~~d previous work, si1i~ saturation con~itions, The negligib1e

s.olubility of CUlS and thfappreciable solubility of FeS in silica saturated slags are notable. T~ experimental conditi~n$ were:OÔ o 1423 K, pS02 - 1.0, 0.5, 0.1 atmosi ••• 1473 iC, pS02 - 1.0, 0.5, O~l atlIIOsiO. pSOZ"- 1.0 atmos, 1523, 1573 K. "

~

1

80

,.-...

,;

,

0\ CCI

i 1 _

~J

1

1

- 1

l 1

1", ..f' .. -------:--- -------~------~-w--.... ~ ...... ~ .. ~ .... .:.-.;~I.F .. '~5r~ .... ~------------~ ; 1 ; r l'tl-•• n ' 1 Pl, 7 r Il 1 ...... ?~.,-(-- -~-,.--,.

.~ -

(

I~

"

v-'

( ,

,1

t t 1

· t t ~

~t Q

____ 4_._ ... """ ....... _. __ ,.....'_._ .. r_~ __ _ 69.

~possible explanation for the rather high sulphur-in-slag

concentrations of the present work Ct Cu in matte < 60%), i$ that°totte,

slag was not ~ruly saturated with Si02 .

shown" this effect is very small unless

Howe,ver, as Yazawa(13) has Q J

the % Si02 in slag falls below

As a general conclusion about sulphur~olubility, then, the'

principal contrOUing factor is matte grade with aIl other factors being ,

secondary. The s~lphur-in-slag solubility of Cu2& is negligibl~ but the ~

solubility increases with decreasing matte grade. -t

IV.4 BEHAVIOUR OF OXYGEN IN MATIE \

The measuted concentrations ,of oxygen in the equilibrated

1 mattes of the present investigation are presented in Fig. IV.7/Which

, ,

also compares the 'best fit' line with the data of pr~~ious workers.

As with sulphur in slag, the Most noticeable feature of th'é ~ ,

resul~s i5 the remarkable difference between the behaviour of cu2s,and'

~ , FeS., Ther~ is virtually'no oxygen solu~lity in Cu2S but considerablé

(up to iQ) (5pira and Themelis(12)) in FeS, with~ntermediate values

for Cu-Fe-S mattes. Matte compo~ition appears to overwhelm aIl other

effects (i.e., temperaturé and p502). " '

Like sulphur solubility in slag (Section IV.3), the present ~

, ~,

experimental oxygen solubilities are, for matte grades below 60% Cu, '-\>

t

somewhat higher than those of previous workers. These-.~o sets of

resul~s aie therefore internally consistent at least. Yazawa(13) notes

that a lack of Si02 saturation will increase oxygen-in-matte solubility

but thiS'" effect is qui te small for 5i02

contents above 30%. ~

f

- -- -- -- - - ~ -\---- -- -- -' '~

. ,

_II1II"'_"""_ ... __ ....... ____ ...-'" ._ f~";Iii __ .. L111P kll!i .'."'f~~_ ........... ~..- .. ..,: ... ..".. ~ __ .. '" , .,,~.

. ,. --.

( \.

1

",G) 4-4-

• • • 8·0~ ~

0

E tif

• •

c: .- 4·0f â

0 , ~ 0 ,

'" -l1»- o -10

\

• ""'" '! FIG.-IV. 7

~

': / • 2:;-· -. ,

• • •

.... ~

• • ~~ 0

ft , • (:)

.!

.. ç

.t /'

'" -THIS WORK

o GEVECI AND ROSENQVIST ( 18) 1 •

'0 8~ AND'>T ARASSOFF (II V .,' SPIRA-' ~ND THEMELIS (12) .

• •

o • o é}t:P o

80 b

70 • 20 3~~.MO 50' % Cu (in ~ott e,

60

, \ o'

,.--.,.

;

Oxygen solubility in slag as a fbnction of matte grade from this and previous work, silica saturation conditions. The negligible solu­bility of oxygen in CuiS and the appreciable solubility in FeS ~e not"âble. .TIlI experimental conditions were: 0 601423 K, ,pS02 - 1.0, 0.5, 0.1 atmosj ••• 1413 K. pS02 _ 1.0, 0.5, 0.1 atmos;Qy pS02 -

~

1.0 at,os, 1523, 1513 K. 1 . • .

ni.' j : .' i li IJHnSU ; • 1 lm • J g 11Ft 1" DII ) • PH S' h ., • 1 riEl S ' . '

...., o

1 l,

"

1

, .

I­I

t

1 ! 1

. 1 1 f'

1 1 i

't 1

1 1 " ~,

'i

1 Il

:1 Il " 1 1

" j

}

(

. " ,

, l,

'"

(1

1 ' l ... _-_._-- --.....;.--\0(-------------------------

, '

~. 71. '

1 r "

~~v.4.i Si02 in ~tte

, As can be seen from the experimental 'data (Table 111.1), there-?

seems to be a Isma1l 5i02

solubility in tHe equi1ibrated mattes « 3%).

Bo; and Tarassoff(ll), in dis,cussing 5i02

sO,lubility in m.atte,

o suggest that ~n~er iron and 5i02

-of Si02

in matt/es (la to 40% Cu)

. , saturation conditions the solubility

i's le~s than 1%, decreJsing with

~ .

\ lt"l

ma, e grade. This trend towards low 5i02 solbbilitiês in , 1

v/ higq g e ~it.-es is confirmed by the present e~perimental resul ts (

J(Table III')'

,ample. f

" 1 1

*

1 1

. SQme of the highe~ values of the pr~sent,~ork may. in

}nclusion of traces of quartz crucible in the 'analrtical h 1

/

IV~5 STRUCTURAL REPRESENTATION OF RESULTS

f' Nagamo;i (9), T~guri (15) and Yazawa q3) have recently published 'JI , ,,~ J

.. )/ , ~ C

thor9ugh discussions las to how copper-in-slag behaviour might best be o " )

1 Q ,

e~ained ln si~ctura~ terms. 'Much of the discussion has ~entred around i.J'JJ ~ '.. 0

,n the distinction, between 1 oxide 1 and 1 sulphide 1 copper in the slags. How-l " L

; ever, this concept of s~rate categories of copper-i~-slag ddê~not fJt /, ,-\ 1 (,.. ,

~". 1,,1 weB witb the basic ,~oncep~ of, slags as .'s,oluti~s of anions and catioI\.s. ~

/ Thel e aré, ,rlevertheless, some basic observations which require 1/, 1 ~ '" exp1anation.

" la)

;'

Why does thecopper-in-s1ag concentration increase cata-~ , , , '

, ' <>

,strophically for matte grades above 60 or 65% Cu?

(b~ Why~ is FeS -more soluble int si'ags than Cu2S? ' )

',(c) Why is oxygen ~Ch ,m"re SOluble"in FeS than Cu2S? 1 _

These point~ are discussed in the next two subsections. " . .. ,

'~. . '-

ôl 1

• i ~( ~ .. l

"

;. . ( ,.

~ '1

-. . ' '\ \

j

1

~

(. ,

\

J r 1

1 i ! (>

t !

.. t l ' ,

"

"

\

/

72 •

• 1

.., IV.S.A 60 + % CU Mattes

r

that wt%

For SUlphur-free systems: Ta ri(lS,16) and Altman(17) 5uggest

CuJn~'Slag is rel~ted to CUZO- activity by, the relationships: '':''> . ~

1S00 K wt% Cu

1~23 K wt% Cu' "

JI

~,

'"!" 34 (a' ) 1/2 CuzO - 28(a )1/2 Cuzo •

"- Altman

Toguri 1

14e Toguri value i5 probabiy somewhat low because hi~-slags aiso ~ ,,' 111' ""

1 11.".

".ÇOllt8.ined some alumina dis sol ved from the expei-imental crucibles. ,1"'. ... fi) .< 1

Since the sulphur conterlt of the present eXperimental s'lags . -r f

was less than one per cènt'for matte ~ades fbove 65% C~, it would seern

that these relationsh'ips could be applied to the present results over , 1 , /

this experimental range.

There is, of course, considerable scatt~r in the present aCu 0 \' 2

results (F!i. IV.3) wh~ich makes pre~ise interpret~~on difficult but if

aCu 0 i5 said to be 'ënveloped fithin 10-2•5 and \0-4 at XCu S - .... 0.9, 2 . Z

(75% CUJ it can be seen that Alt~n predicts copper-in-slags of between . 0.3 and 2% which is reasonably close to the present experimental values

. t' .

(Fig. IV.4). The Ilower y~lues of w~% Cu in slag reported by Yazawa(10:

and Sehnalek(23) for iron saturation cond~tions (Fig. IV.5) are due, of ~ i \ . ,'I!t

course, to the low~alues of oxy~~n~artial pressure 4nd aeu 0 under • '> 2

such co~ditions.

IV.S.B Relativé Solubilities of and " 0

In examining'the reasons

t , \ . ,

4

)

\

,1 • ,J

1

1.

.-

·f 1

, r

if ( 1,

l'

.. --------, \

~

\ , "

,

'" (a) the negligible ~o1ubility of CuZS \

in s lag and. the

considerable solubi1ity of FeS in ~~ l' •

slag; "

(b) the ~egl~gible SQlubility, of oxygen in Cu2S and the

conside~able solubi~lty of oxygen ~n FeS; ~I , ..

73.

thf.p~~erties o~~these two sulphides (and FeO) can be tabulat~:

., .

PToperty

Me1ting point K(l)

Density at 1523 K (tonne~ m- 3J(24,2S)

Specifie e1ectri~al conductance at_II (24 26) 1523 K (g~l Cm ), '

,Cu:2§.

1130

5.8

100

,II

. '" "

FeS

1190

3.8

1500

• •

FeO T ~377 .?

5.7 (soUd) ,

200 (at melting,point)

.. The free energies of formation of t~se three compounds and

Cu2

0 aré:

"

•

2Cu.t + 1/2 52 .... Cu2SJ/. 'g

2,CU~" ,1/2 02\ .... "CU20-

•

'/ I~

o âG 1523 K - -84000

o âG 1523 K

, ,. / -73000

; .... 1

joules

joules

o " âG 1523 K - -163000 joul~s

1.

t l , 1

l'

It can be ~from the above data th1t liquid CU2S differs

"from FeS only with respect ta its DlUch greater ,densi~y* and its much (

smaller electrical conductivity. It appears to behave like a semi-

~

74.

, l, ~

1 conducter(26) '(stnaU·positive temperature effect on conductivity) whereas'

'the electrical behâviour 'of FeS resembles electronic conduction. f,

Thèse observations suggest that Cu2S is ~re strongly co - ~

bond~d than FeS and that as a consequence it resists being dissolve, ,and

broken up by the ionie structure of 5lags. , Of course, ~

must be regarded aSi speculative'and further structural 1 (

is required to fully/explain these phenpmena.

.,

i~.~~REDlcfION OF CONDITIONS UNDER·WHICH A SEPARATE COPPER PHASE , WILL)tCCUR

, )The~ecent extensive studies into single-step copper~king Diake it of great interest ,ta fredict the conditions under which â pure

• 'ù •

copper phase (aCu - 1) w111 occur. 1

There are (Section IV.I) only two degree~~f freedom in this ~ J

<#

system so that once temperature and one other parameter are specified (

the system is èompletely defined. The cilculations 6f t~is thesis 1 1 _

(summâriz~d again, below) can be used to make these predictions . .., The basi~ assumptions in the calculation are:

l,

(a) that the oxygen pressure and ~,eO of the system can be

cQ.lculated fram ,the slag co~osi tian (Fe++ /Fl++ ratio ~ ~ ~ ~ along . the Si02 saturation l~ne) ,as if Cu and S have no

* Intel'estingly Cu.,s~ lik$ watel'~ expands on fl'eezing bec~U8e it8 1

liquiçi densi'f7y a~ the !l'eezing point (5.9Jtonnes m- 3) excgeds its 80lid q,nsity (5.6 to 5.8 tonnes m-5). ,

l' ( . .....

~,

[

l'

__ -----:--_ . ____ --=-__ 1-

.' Il,

.. l't,

, 1

!

1 t

,r

! .

( ( 75.

) effect, L.a., that the data ~f ~an(20)J Darxen and

~r~(27,28), Sc?uhmann and Ensio(29) ~Mi1hal and

(30) (31:-) r, Schuhmann . and Oarken apply to the system;

(h) that the acti~ities of Cu2S and FeS in matte are closely

.~epresented by the Temkinl22) cationic lattice model as H (21)

noted by BaIe et al .

~ U The usual method of calculation was to specify the temperature

.' f

.....

"'-

,

J,? li

of the sYrtem a~ the Fe+~Fe+++ rat~o of the silica saturat~ slag,

which of cohrse fu11y defines the remainder of the system. The limiting

1 . f h F +T IF +++ . f ~ . , ' i < t d va ues 0 tee . e rat10 or trie slag were.at ~ron saturat on an

'1 Magnetite

, . lations:

<:.,. • 0

, .

..

saturation which mark the limits of the liquid slag field •. , ~

Ir' •

The predictions we~e made by the fOllOWt'ng sequence of calcu-. . ~

. (a) Having specified ~emperature, the F~++ IFt++ ratio along the

\

(h)

(c)

11"

(d)

pO~ and apeQ equivalent to the ~r~scri~ed'slag composi­

tion and temperature were determined from the data of - ,

Muan(20t, Oarken and Gurry(27,28) 1 Scbuhmann and Ensio(29), l , • 7

Michal and Schuhmann()O) and Darken(31) (as replotted in i" 1

Fig. IV.8) • . \1

1

aeu 0 /w~ calculated :5rom the p02 in (b) and aCu - 1 2 ' ('

(copper sa~ur~ted 'syste~) using equations IV.S aq9,IV.6.

The act~vity of Fe in the systèm was catrculated' from

'aFeO and p02 according ta the equations: .,

,. , , '1

. ~ 1

, 1

1 -~. ~ ... _ ..

/

_--~--F-

" e :1 -6

,) -8 N

0 0-

C' -10 0 ~

-'2 ,

.

0 coJ I.L.

0

FIG.IV.S

. aL • ---~ -- ~

'.

\ , 76.

-r t

JI'

>'0 MUAN (20) •

l

1573 K ,

,. . ~

" , "-0 1473r K

\...... O'5~ 6 DARKEN ( 31) • SCHUHMANN (29,30)

) \473 K • .... 0-4

~

........ _______ ooI--.a...-"---!-'--........ - ...... ----' O· 3 40 60' '. 80 20

..

-F ++'F ... ++ 1 e- 1 e 0 J

, . 1 Il ..;

A' ~ Esti~ted q~f~~ pressure values and acti vi ty valu~s of FeO for given Fe~/F ratio in slagrfrom available datak20,29,30,31). ~e aFeO lines, are4~ls9 basèd on aF~304 and p02odata(~~,21,28).

, '.

-""'------

.. /

-1 ! !

(

"

-/ ()

\

\. )

, \ 1

• 1

~I .----ts--.....;;..---.. ---t--

-+ + Fe0R.

âGO - -233000 + 46~0 T f

joules IV.7

IV.8 a (pO) 11.2

Fe 2 s

(e) The ratio aCu s/apeS of the system was calculateq-f~pm 2 . 1. ''. '

aCu, acu 0' ape and apeO yS1ng the equat1ons: .~ 2 \ t

•

AGO - 97000 - 103.0 T

-and

"." 2FeO + FeS :t 3Fe + S02

~

âGo - 220000 - 46.4 T

.which combine 'to give

, ,

joules

. joules \

1 ..

',. 0

• 1 IV.9

IV.IO

'( .\;

.1

IV.])

IV.I2'

0' ,

.,

#--

..

)

1 ..

( 1

..

/

-,~ .. .. -~

\

....

1

fl \. t:

!

;àoo -_ ... ",-~ .. --.,.

f 78.

(ape~3 2 KIl aeu 5 (aeu 0) E 2 4 -:-y- -x 2 x

(aeu) 6 KE 8 PeS (apeO) ,

or ~ /

2 6 aeu 5 KIl (apeO) (aeu) .,. 2 E - :-lx 3 x 2 apeS KE (ape) (aeu 0)

2 "

(f) The composition of the matte which is equivalent to this , ~

acu2s/apes ratio was calculated using the Temkin cationic matte ~

model (assuming the Isystem to be oxygen-free) proposed by

(21) BaIe et al ,i.e.; nCu . 2

aeu S - Cn + n ) 2 Cu Fe

nFe , ns ncu + 'nFe • ns

""-"?

and the mole' fraction summationn

1

which lead' to

1 IV.13

IV.'14

IV.IS

rV.16

(g) The activity of CU2S was calculated from Xcu2s

·in Cf) ~

. "ùsing Equations IV.13~ and IV.IS.

(h) ~ pS02 was calculated fi-om aCuJ acu) and ac~ 0 by Equa-2 2

tians IV.9 and IV.IO. .l

\ ,-

...

i 1 1

(

i 1

t /

(,

?

": )

----------*--!-- ------79.

~ The results of the C~lculations are summari~ed ~n F?g. IV.9

which 'shows the,--ne composition range ovar which i t i~ possible te

The ~ produce meta1lic copp~r (Si02 saturation, liquid Fe-O-SJ02 slag).

1imit at the low matte grade en~s iron saturation i~ the system , _ . . (ap -,1) whi1e the upper 1imit is magnetite saturatio~ (a- 0 - 1).

e re3 4 .. 'The following pOilts can be noted.

(a)

!

(b)

Copper,can be at equilib~ium with matte/s1ag/silica/gas

in the range "of matte grades of

D .Ji ~

44% té 79% Cu '1573 K , 48% t,o ,79% Cu 1413 I<

equi1ib~ium pSO; is very low 'at •• rr The iron saturation

conditions flO- 6 or 10-7 atmos.) but it approaches

1 atm9sphere- at magnetite saturl~ion.·

t

1 .It 1s of interest to compare these predictions with the ,

~

me~urements of Geveci and Rosenqvist(18) who found, under' 5i02 s4tura­~

ti'on condi tion~ (1523 ~) J that:

(a) Cu cao be at equilibrium.with ~tte/slag/silicalgas for

matte grades-between s~ and 79.4% copper; ,. .

(b)- the equilibrium pS02 is very low under iron saturation

1

conditions apd increas~~ to ~.'~2 atmosp~eres at 79% Gu

in matte (the latter in comp~rison with'p~iction$. of

10-2 atmospheres (1473 K) and 4x10-1 atmospheres (1573 .~)).

..

;. , ,1

1 J ------ -~-.--. __ .~------ .~. ~"-="'"""''-'''-''--___ 1iIL ~

• 1 ,

( ,

...

' ...

, .

--'

( )

"

; .

80.

('

80 , 1573

+ + 60, + I.RON ~/

Q) ~ ')

SATURfON .. ,.

LL,. "-+ 40 ) .. I~ "-+ " Q) " LL '\ "-

"-20 ~

K~,\ 1473 ~, , ,

},

• • t ,. "

0 " . MAGNETITE· • SATURATION

t\I - 2 0 , en .. 0- -4

\. , 1573 K~

0' lRON

0 S

..J -6 ~

1473,K, ...

-.J --0

'40 ( 50'~ 60 70 8·0 ~

'" f \

,010 C u in motta, 4'

FIG.IV.9 Conditions, \lJlder which metallic copper (aCu - Il can be at ' eqûïlibrium ~ith.matte an1hslag as calculated in Section IV.6.

\.

(

\.

l

1

,

___ ..... _, _._. _Ml ". __ 111 _, _____ - ... -----.-__ .. ~ ~ -

81.

The predictions are, therefore, very close to the experimental

data reported by Géveci and Rosenqvist(~8).

\ with Cu,

The precise composition of~the matte 'which is in equilibrium , \

Fe3,0 4 and Si02 is predicted to be :

" 7fj,jJt Cu (1473 Kl: r

79.8% Cu (1573 K)

Geveci and Rosenqvist US) report 79.4% Cu under these candi- \z-, tians and Rosenqvist(19) 79.5% Cu and,Sehnalek (reported by Rosenqvist(19))

18.9% Cu so that thes~ictions a1so sèem to be rèasonably accurate.

It appears, therefore, th~Cu2S-FeS ac~ivities, p02 a~d aFeO

slag data, and~eta1/matte/slag/gas experiments are weIl tied-in together.\ --IV.7 INDUSTRIAL IMPLICATIONS OF THE RESULTS

~

The results of this wark, show that matte smelting presents very

few operating problems as long as the grade of the matte prodl;1ct is 'kept .\

below 60% Cu. Und~r these conditions copper losses in slag are at a

l6ler~ble level (Fig. IV.4) and they are not appreciably inf1uenced by

""-, temperature or ~02 pressure.

From wWat can be gathered from industrial proc~sses, it app:ars ~ .

that produttion of mattes containing more than about 60% Cu'also leads

'to difficult nfoblems with impurities in the .matte phase as well as high - <1

Cu-in-sIl, càncentrations. This problem was not studied in this work but

it ls ilDportant to note that such problems exist. ' , . , ... f

With rega~ to producing copper directly from concentrate it

i5 now clear that it is possible tOlproduce copp~r direct~y from rela-'Cl>

tively low grade mattes.'" ~onditions whiCh promote suçh métal production

are:

1

..

)-

,\

r

.. ;

~

~, 1

L

"'C 1

1 1

-, _._------_ .... _---82f.

(a) low oxidjzing potentials as r~presented by rela~ively • high Fe++/Fe+++ ratios in slagsj

(b) __ .Jow 502 pressures above the system.

Of course, it is not possible t~ ob tain these conditions while

o~gen or air are being passed through matte'to remove~sulphqr from the

system. However, the, might be ac?ieved by a~te~atelY.passing (i) air

or ,oxygen th~-(i~) reducing gas (CO or H2) th~gÜ tne system. Copper

would be produced dvrinf the redUCin?'gas stage from low 'grade mattes

thereby minimizing'the copper-in~slag S5 problem. ~ow other impurities . ... ,

might behave,puring such a process is not known. t •

• 1

(

f o

1

<;

'\ '\

•

( \.

~'

ç,

...... -

.:

. ,

~

,.,. .y"

..

! Î

. 1

1

~ i "

J..

\.

'-

~

,

----------~--_ ..

-'Î f~ ,

On the basis of this investigation it.is suggested that:

1. The present study shou1d be extènded to lower matte

2.

."..; .

i

grades in order to comp1ete the olferall picture of " 1

f the ~tte/slag/S02 'systrm;

The study should be ext~n~ed ta inclûde t~e CaO-Fe304 1

slags currently used by\the Mitsubishi process(33). 1

Theoretièal and experimcntal investigations of ~he

Cu-Fe~O-S-Si02 system under copper saturation and

controlled pS02 condi tio,ns "shOUld be u9taken ï ~ arder to confirm the calculations of Section IV.6.

83.

4.~ An investigation into th~ distribution of minor elements \ 1

(e.g. As) Bi, 'Sb) betweeh slag and matte (Dr slag , matte l , 1

1

and metal) under controlled pS02 conditions should be"

undertaken.

These studies will further -i rove our knowledge of copper

smel ting systems.

" ' .. "- ~~ ~l

'-~

" 1 N' '\

...

1

1

, ,

/

;'

..

."

,

"f

(

"

~

.. -

c>'

, 1 1

1:

\- \ , 1.

, t f.

t ~

f ~ ,

.)/', '-

t " ( t,

Il ,/

"

d,

1.

2.

. )

1": • , .... \~~,. ~.

,1,)

-__ ----.. -.-....... __ r...,~_ ... __ ~ __ _

" 0, ~

VI. CONCLUSIONS .

L

84.

An experimental programme. in which matte.~slag and SO' (0.1 to 1.0 , . 2

atmospheres) were equilibrated (1423-1573

with the pri~cipal objective Qf examining

tiQns underDoxidizing conditions.

K) has been carried out

C~Pfr-in-Slag concentra- .

~~--- .-

The experiments were carried out under S~02 saturation conditions

il .50 that the system consisted of 5 components (Cu. Fe, O. '"

A ' '

·S. Si) and 4 phases (quar~z container, matte, slag, gas), hençe 3

-, cÎegrees of freedom. ,Specifidtiort of pS02~ temperatùre and a

.particular matte, grade fully defined fhe syst'em. '"

3. The experiments showed that when slags and mattes ~}are equilibra~ed,

the copper-in-slag concentrations remain ~ow as long as the ~attes

contain less than 60% Cl,l, under all temper'ature ahd.psO/'conditions ..

4.

çopper-in-slag concentrations increase dramatically under oxidizing :li

conditi1ns wh~n the mattes contain more th~~ 65% Cu.

Cu2

S i~ virtual,ly insoluble in iron-silicate sl~gs .underall condi-,1' "1

,tions. FeS has considerable solubiUty (up to 7% S) and Cu2S-FeS

o mat~es exhibit int~rmediate behaviour. It is suggested that the

strong covalent nature of Cu2S prevents it from dissociàting in

ionic slags. ~ ~

.

S. Oxygen i~ virtually insoluble in CuZS but dissones up to IO~ i~~ ~,.-'

FeS, with mattes exhibiting intermediate b~~~viour. Oxygen will

apparenily not 'fit' into the covalent CuZS struc~ure.

.-, ,

, ._-.. . . , - - ---.--~ -----

1 \

\ r

.-i

! , • ,. . r

1

.' ~

, '" . , , " 1 •

! r ri , , , . , 1. 1 :

f ,

i ~ : ..

, f

1 : , ~

, , , f

1 'i

(

...

•

(

1

'. r

.\.

- ~"IIf'" -~ .. . . ,

"

7. /

(

: -'

, . ( 85 • ..

" 1 •

Conditions undir which a ~ure copper phase,will b~ at equilibrium "'" ....

" "" , ~ wi th Fe-0-Si02 slags and Cu-Fe-S maHes h~~e been pt'edicted -on the

/"'~ .. "~

basis of. previous experimental aF: O and P02 d,atS:- ;or slags: ~n~ acu2s d~ta for mattes. It is sho~thaû under ~i02 saturation co~ditions

mettÙlic. copper~ill ;~cur' ih equiÙbrium j~, Ci) .... 45-'50\ Cu mattes ~ l, .

. at iron saturation, wÜh (ii} 79'. Cu mattes at magneti te saturation .., • 1

and with (Hi) intermediate matte compositions within the liquid

slag field. These predictions confirm the experimeptal result~ oi

Gëveci and~Rosenqvist. )

. The experimenta t copper-in-slag '4 '\ 7'

results show that irtdustrial produc-

4 tiDn of mattes contai~ing 60% Cu or below shoul~ot cause excessive ....

copper-in-slag lqs~es. ,Copper los ses may become excessive above , a. f?

thi~~tte grade especially under highly oXi~Zini conditions.

'\

, >

1

)~ ...

d

: ./

, ~

a ( ,

~ .,. -7'

'. . ,,,,,, - ...

,

...

- ..... _-~~.,. t~r"""""~"'_1i ,l,lIla Ifi 14

..... ~ ~ ".'.'" --' , i '" '

~

'\

, "" •

.J

! 1 1 ! r

·1

.. , . ,,-"~"".P'" a

.. ~., -II ". '!"

' .. .,. 86. ( . <"t \ ~ .. <".

\

-{ • APPENDI~ 1.1 DetailS of ExPeriments ... \

J" A\. 0 . ,-.- . .,. -"" Exp. # W:t ."" CU2S gr. Wt. FeS gr. Wt • Si0

2 gr .. Wt. Slag gr. ,Total Weight

·v<9 t ,..". r \ 7

1 5.5496 3.3297 9.925 5.67 15.4743 2 6.4746 0.9249 '.- 4.7~8 12.1306 3 5.5496 1.8499

11 ~.4 4.7 8 12 .. S30~ 4 5.0000 5.0000 '1.33 6.3934 17.7234 S 4.5000 5.785-7 1.34 6.5761 18.2018 6 Il 6.4746 0.9249 4.7308 12.1303 7 5.5496 1.8499 0.4 4.7308 ~12.5303

8 5..5496 3.3297 0.925 5.6700 15.4743 9 5'.000Ô 5.30QO 1.335 ft- 6.~000 18.0350

10 5.0000 5.0000 r.33 6.3934 17.7234 v Il 4.5000 5.7857 1. 34 ~ 6.5761 18.2018 ,.

12 5.5496 1.8499 {)'.4 4.7308 12.5303 1.? 5.5496 3.3297 0.925 5.6700 15.4143 14. 6.4746 0.9249 4.1308 12.1303 15 5.0000 5.0000 1.33 6.3934 17.7234 16 4.5000 5.7857 1.34

~ 6.5761 18.2018

17 6.4746 4.1395 . 10.6141 18 6.0000 0.4746 4;1395 1P .6141 19 5.5496 3.3297 0.9250 5.6700 15.4743 20 5.5496 1.8499 4.7308 12.130,3 \.. 21 6.4746 0.9249 4.7308 12.1303 22 6.0120 1.3875 4.7308 12.1303 23 5.5496 2.5225 0.8Z31 5.2100 14.1552 24 5.5496 4.3164 1.0275 6.3000 ,17.1935 2S 5.0000 5.0000 1.33 6.3934 17.7234

; " 26 4.5000 5.7857 1.34 6.5761 18.2018

27 6.4746 0.9249 4.7308 12.1303 ,28 6.4746 . , 4.1395 10.6141 29 5.5496 1.8499 0.4 4.7308 12.5303 30 5.549,p • 3.3297 0:925 5.6700 15.4743

r '. 31 5.0000 5.0000 1.33 6.3934 17.'7234 ~t "32 4.5000 5.7857 1.34 6.5761 18.2018

33 6~4746 ;- 4.1395- 10.6141 ) 34 6.4746 0.9249 4.7308 12.1303 35 5.0000 5.0000 1.33 - 6.3934 17.7234 36 . 4.5000 5.7857 1.34 6.5761 18.2018 37 6.4746 0.9249 4.7308 12.1303

-j 38 5.5496 3.3297 0.925 5.6700 15.4743 39 5.5496 1.8499 0.4 4.7308 12.53Ô3 40 5.0000 5.0000 1.33 6.3934 17.7234

( 41 4.5000 5.7857 1.34 \ 6.5761 18.2018 . r , , 42 4.0000 6.2857 1.35 6.5761 18.2118 43 5.5496 1.8499 0.4 4.7308 12.5303

o( ; 44 6.4706 0.9249 4.7308 12.1303 4S 5.5496 .3.3297 0.925 5.6700 15.4743 46 4.5000 5.~7

, 1.34 6.5761 18.2018

l , , 47 5.0000 5.0000 1.33 6.3934 17.7234

1· ,

/ 1

1 lit

__ .. 60.J. ____ > ~ -. ~ "- .. ""~_.... , ~" , .... ~ ... .,. ... ;t. , .... J;' l

1 ..

("

(

..,::1'"

)'

\

1.

2.

87.

Biswas, A.K. and Davenpory, W.G. (1976), Extractive Meta11urgy of Copper, Pergamon International Library, Vo1.20, Pergamon Press

( Ltd." Oxford. . ~

, Krivsky) W.A. and Schuhmann, -R., Jr. (1957), Thermodynamics of the Cu-Fe-S System ~t Matte Smelting Temperatures, Trans. AIME, ~09, 981-988. ! ,),

3. E11iott, J.F. (1976), Phase Re1ationshi~ in thé Pyrometal1urgy . ôf Copper, Met. Tràns. B, The Metallurgièal Socïety of AIME, 7B,

5.

6.

17-33. • . ,

Yatawa, A. and Kameda, M. (1955),' Technol. Rep. Tohoku·Univ., 19, 239=-49.

,/

Kameda, M. and Yazawa, A. (1961), Phy. Chem. of Proc. Met., Part 2, G. St.Pierre, ed., 963-88'jTMS Conf. Proc.,"Interscience, New York.

Kameda, M. and Yatawa, A. (1970), Studies in Met., 159-66, Commemoration for Prof. Matoba, Tohoku Univ ..

7. Yazawa, A. (1956), Techn91. Rep. Tohoku Univ., 21, 31-50.

8. Bailey, J.B.W. and Garner, F.A. (1974). Composition on the Solubility of Copper SeL' Engng., 2, ~6-17.

The Effect of Slag in Smelting Slags, Mineral

~ 9. Nagamori, M:"l(1974), Metal "tess to Slag: Part I. Su1fiq.ie and

Oxidic Dissolution of Copper in Fayalite Slag fram Low Grade Matte, Met. Trans. Meta1lurgical Society of AIME, S, 531-38.

10. •

11.

12.

/

~zawaJ A. and Ka~edaJ M. (1953), Tech. Reports, Tohoku Univ., 16, 40 (as reported br Toguri, J.M., ThemelÎs, N.J. 'and Jennings, P:H. (1964), A Review of Recent Studies on Copper Sme1ting, Cano Met. Quart., 3, 197-220). .

1 .. Bor, F.Y. and Tarassoff, P. (1971), Solubi1ity of Oxygen in Copper Mattes, Cano Met. Quart., 10, 267-71.

, Spira, P. and Theme1is, N.J. (1969), The Solubility of Copper in Slags, J. of Metals~ 21, 35-42.

-13. Yazawa. A. (1974), The~dynamic Considerations of Copper Smelting,

14.

Cano Met. Quart., 13 (3),443-53. ' "

Ruddle, R.W., Taylor, B. and Bates,,~.P. (1966). The Solubility of Copper in Iron Silicate Slags, Trans. Instn. Min. Me~.~ 75" Cl-Il.

b

fg._rn_j._~.~.~.~~~_~~~~~_If'4('r""""'''''''''*'-'">tlf, .. *~, .

1

11

l ,

•

_ .... , ............... ~---

, 1 ) , 88. ...

15. ,Toguri, J.M. and Santander, N.H. (196~), The Solubility of Coppel' in Fayalite Slags at 1300oC, Cano Met. Quart., B (2), 167-71 . .

16.

17.

18.

19.

, 20.

Toguri, J.M. and Santander, N.H. (1972), Distribution of COppel' Between Cu-Au A110ys and~Si1ica,Saturated Fayalite Slags, Mét. Trans., Meta Il ur g'i cal Society of AIME. 3, 586-88-., ,.

, Altman, R. and Kellogg, H.H. (1972), ,Solubility of Coppel' in Silica­Saturated Iron Silicate Slag, Trans. I~stn. Min. Met., 81, ~163-7S.

! .

Geveci, A. and Rosenqvist, T. (1973), EquiIibrium Rel~tions Between Liquid Copper, Iron-Copper Matte and Iron Silicate SIag at-12S0oC,

• , 7

Trans. -Instn. Min. Met., 82, C193-201. . Johansen, E.B., Rosenqvist, T. and Torgersen, P.T. (1970); On the Thermodynamics of Continuous Cèpper Smew.ng J J. of MetaIs, 22, 39-47.

r ,\.-

Muan, A. (1955), Phase Equilibri(' in ·the System P·eO-Fe203-Si1S. Trans. AIME, 7, 965-76. N ' 2

# BaIe, C.W. and Toguri, J.M. (1976), Thermodynamics of the CU-S, Fe-S and Cu-Pe-S'Systems, Cano -Met. Quart., "15 (4), 30S-1S.

'" 22. Temkin, M. (1945), Acta Physiochimica URSS, 20, 411.

23.

24.

25.

SehnaIek, F. and Imris, I. (197t), Advances in Extractive Met. and Refining, M.J. Jones, ed., London. Instn. of Min. and Met., 39-62. 1

Dib, C;. (1974), li'he Thermodynamic Praperties 'of C6pper-S1ag Systems, INCRA Series on The Metallurgy of Copper, I~te~ational Copper Research Association Monograph III, New York. 1 '-, .... .., Bourgon, M., Derge, G. and Pound, G.M. (1958), 10ensity of Me1ten Cu-Fe Su1fide Mattes, Trans. AIME, 212, 338-40.

26. Pound, G.M., Dergè, G. and Osuch, G. (1955), E1ectrical Conduction in Malten Cu-Fe'Sulphide Mattes, Trans: AIME, 7, 481-~41

27. Darken, L.S. and Gurry, R.W. (1945), The System Iron-Oxygen, J. Amer. Chem. Soc., 67, 1398-412.

28. Darken, fL.S. and Gurry, R'.W. (1946), The System lton-Oxygen, J. ~er. Chem. Soc., 68, 798-816.

29. Schuhmann, R. Jr., and Ensio T J. (1951), Thermodynamics of Iron­Silicate Slag}: Slags Saturated with Gamma Iron, Trans. AIME, 191, 401-11.

-'·0 __ ( ~

. '

, .

....

0',

/,

« 1

..

r \

, .

, ,

" 1

• 30.

, .

31.

32.

, . 33.

,-----------------------

. Michal"E.J. imd Schuhmann, R.~ Jr. (195~), Thermodynamics of Iron~Silicate Slags: Slags Saturated vith Sol id Siliea, Trans. AIME, 194, 723-28. "

" , Darken, L.S. (1948), Melting Points of Iron Oxide on Silica: Phase Equilibria iii the Fej:'Si-O Syste~. Amer. Chem. Soc., 70, 2046-53. '

MacLean, ~.H. (1969), Liquid~s 'Phase Relations in the FeS-FeO­Fe304-S~02 System, and'The~r,Applieatio~in Geology, Economie Ge(dogy, 64 t 865 - 84 ~, '

89 .

Nagano, T. 'and Suzuki, T.·,(1976), fxtrâcti~e Meta.J.lurg)" of Copper, Vol.I, J.C. Yannopoulos Yannopoulos and J.C. Agarwal, ed~, The Mflt. Soc. of AIME, 439-5 7: '

"

,

j "

. 1 • \ ,

, '. , (

<'

/