Canadian.Minsralogtst
Vol. 16, pp.397-404(1978)
BORNITE(GU,FESN):
STABITITVAND CRYSTALSTRUCTURE
OF THE II{TERMEDIATE
FORM
YASUO KANAZAWA*, KICHIRO KOTO AND NOBUO MORIMOTO
Institute of Scientific and Indwtrial Research,Qsaka University, YamadakatniSulta,
Osaka 565, tapan
ABSIRACT
The intermediate form of bornite (CurFeSJ has
been determined to be stable in tho range from
170oC to 235"C by differential scnnning calorimetry and high-temperature X-ray methods. The
changes in the cell dimensions on rapid cooling
indicate the appearance of a transitional meta$able
form after the hieh form, different from the intermediate form.
The crystal structure of Xhe intermediate form
has been determined at 185"C. The space group
is Fm3m, and c 10.981(1)A. Sulfur atoms form
the ideal cubic clos€st packing, and metal atoms are
distributed statistically in the tetrahedral interstices
among sulfur atoms. Tllre structure consists of two
different kinds sf cubes with tle antifluorite-ty1n
structure; one has half a metal atom in each tetrahedron and represents disorder of Cu atoms and
vacancies, whereas the otler has one metal atom
in each tetrahedron and represents disorder of Cu
and Fe atoms. The internrediate forrr represents a
stage between the low and high fonns.
SoIvrUenr
Ir champ de stabilit6 de la forme interm6diaire
de bornite (Cu'FeSJ s'6tend de 170' e 235'C,
ainsi qu'on l'a d6termin6 par calorim6trie diff6rentielle i balayage et par diffraction-X e haute t€mp6rature. Un refroidissement xapide de la forme de
haute temp6rature modifie les dimensions de la
maille, signalant ainsi l'apparition d'une forme de
transition m6tastable, diff6rente de la forme interm6diaire, La structure cristalline de la forme interm6diaire a 6t6 d6termin6e i 185"C: groupe spatial
Fm3m, a 10.981(1)A. Irs atomes de soufre constituent un empilement compact cubique, et les atomes m6talliques occupent, de fagon al6atoire, les
trois quafis des interstices t6ta6driques. La structure contient des mailles de deux sortes du type
antifluorine, les t6trabdres 6tant occup6s, dans les
unes, d raison d'un demi-atome de cuiwe par t€traddre et, dans les autres, par un atom€ m6tallique. Dans les premidres, le d6sordre se produit
entre atomes de cuiwe et lacunes; dans les secondes, ce sont les atom€s Cu et Fe qui sont d6sor*Present address: Geological Survey of Japau,
8 Kawada-cho, Shinjuku-ku, Tokyo 162, Japan.
donn6s.La forme interm6diaire repr6senteun stade
structural Sansitionnel entre les formes de haute
et de bassetemp6rature.
(Iraduit par la R6daction)
INrnopuc"rtoN
In spite of extensive studies of natural and
synthetic bornite (CuiFeSe), the polymorphic
relations and the crystal structures of the polymorphs remained unsolved. Frueh (1950) studied the polymorphic relations of bornite with
X-ray, thermal and electrical methods and found
two polymorphs: low- and high-temperature
forrrs (hereafter low and high forms, respectively). Morimoto & Kullerud (1961) described two
stable forms (ow and high forms) and one
transitional metastable form found through uss
of high-temperature single-crystal X-ray methods. They found the high form to be stable
above 228"C, and to gradually change at room
temperature to the low form through a transitional stage. However, DTA curves (Sugaki &
Shima 1965), the magnetic susceptibility (Allais
1965) and electrical resistance and magnetic
susceptibility of bornite (fakeno et al. 1968)
indicated the existence of another stable intermediate-temperature forq (hereafter intermediate form) between the low and high forms.
However, the relations between the metastable
and intermediate forms have not yet been clarified.
The structure of the high form was determined to be of antifluorite type with a 5.50A
and space group Frn3rn (Morimoto L964). The
metal atoms are statistically distributed nmotrg
24 equivalent sites in the sulfur tetrahedron. The
crystal structure of the metastable form was
studied by Morimoto (1964) and Allais (1968).
This form gives apparent cubic symmetry with
space group Fd3m and a cell dimension of c
LO94A. fwice that of the hieh form. Reflections
in the X-ray diffraction pattern show extra systematic absences in that only reflections with
h, k, I all odd, all 4n or all 4n!2 appear. Moiimoto (1964) explained these extra systematic
absencesby a twin hypothesis based on a structure consisting of incoherent domains with
397
THB cANADIAN MINEMLOGIST
398
rhombohedral symmetry, a structure first proposed by Donnay et al. (1958) for synthetic
digenite. However, Alais (1968) considered the
size of the rhombohedral domains to be of the
order of coherency of the X-rays and determined a statistical structure of cubic symmetry.
In both structures, sulfur atoms form tle ideal
cubic closest packing. The metal atoms are
distributed statistically in the tetrahedral interstices in tle sulfur-atom array.
A variety of cell dimensionsand spacegloups
has been reported for the low form flunell &
Adams 1949, Fruelr 1950, Morimoto & Kullerud 1961). Koto & Morimoto (1972, L975)
have shown that the low form is orthorhombic
(pseudotetragonal)with a 10.950(1),b 21.862Q),
c 10.950(1)A, space groap Pbca. If the high and
the transitional forms are representedby aXaxa
and 2aX2aXZa cells, respectively,the low form
is representedby a 2ax4aY2a supercell. Koto
& Morimoto (1975) determined the structure of
the low form by a modified partial Patterson
function. As will be described later in more
detail, tle structure is characterized by a
mosaic pattern arrangement of two structural
units: antifluorite-type and sphalerite-type cubes.
Putnis & Grace (1976 have observed the
phase transformations of bornite by transmission electron microscopy. They found that two
types of zaxZaxZe supercells are produced
from the high form by different cooling rates.
The 2ax2ax2a supercell with the systematically absent supercell reflections is obtained by
rapid cooling; this corresponds to the metastible form. Slow cooling leads to another
2aX2aX2a supercell without systematic absences.This may correspond to the intermediate
form describedin this paPer.
In this investigation, the stability range of
the intermediate form of bornite has been determined by differential scanning calorimetry
(DSC) and high-temperature X-ray single-crystal methods. The crystal strusture of the intermediate form also has been determined and is
compared with those of other polymorphs of
bornite.
Sresu,ITv RELATIoNSmS
Materials
Bornite from Redruth, Cornwall, England
and synthetic bornite (CusFeSa)were used for
the thermal analysis and the high-temperature
X-ray study. Bornite was obtained by dry synthesis from a starting mixture of copper, iron
and sulfur in the ratio 5:1:4. The purity of the
metals was 99.99% and that of the sulfur
99.999Vo.
The mixture was heated at 70OoC for a day
and was quenched in cold water. The reaction
product was ground in acetone; heating and
grinding were repeated two more times. After
heating at 500oC for twenty days, the product
was cooled very slowly. The final product was
identified, by X-ray powder diffraction and reflection microscopy, as the low form of bornite.
Exo
I
.h.-\
II "'lll"'"nl''''
V
Bornite
(Redruth)
V
roo
2ooT/\
3oo
Frc. 1. DSC curves for synthetic bornite and bornite from Cornwall,
England.
INTERIVIEDIATB
399
FORIVI OF BORNITE
Therrnal analysis by DSC
Thermal analysis of natural bornite and synthetic CueFe& was obtained with DSC. The
DSC curves wele measured from room temperature to 450"C and vice versa in nitrogen
at one atmosphere pressure, using an aluminum
sample holder. The rates of increase and decreaseof'temperature were 10 6d JoQ,/min.,
respectively. Temperature was measured by a
Pt-Platinel thermocouple and controlled automatically.
The DSC curves of synthetic and natural
bornite are shown in Figure 1. Two endothermic
peals appear as temperature is increased. One
peak begins at about 170"C and is complete at
2OO"C for botL synthetic and natural bornite.
The other peak is broad and weak and extends
over a range from 20O'C to Z6O"C. The DSC
curves are similar to the DTA curves reported
by Sugaki & Shima (1965), though the peaks
are shifted toward the higher temperature side
by 5 to 1O'C in the latter. On cooling, the DSC
curve of synthetic bornite shows two exottrermic
peaks, whereas that of natural bornite shows
only one peak displaced toward lower temperature. The DSC surves are reproducible on repeated runs, although the exothermic peak on
the high+emperature side becomes insignificant
for natural specimens on cooling.
TABLEI. CELLCONSTA]\TS
OF BORNITE
AT ELEVATED
TEMPERATURES
?.4
50
75
100
125
'150
a(A)
'r0.9s8(2)
710
220
230
240
zou
300
l0.es9(2)
10.964(2)
ro.e68(2)
10,972(2)
1o.e72(2)
1 0 . e 8 3 )( 1
I 0 . 9 7 6(67 )
r 0 . 9 8 0 (68 )
l 0 . 9 8 3 (56 )
l 0.98e0(4)
r 0 . e e l2( 6)
l 0 . 9 9 5 (24 )
5 . 5 0 7 (52 )
5 .s ] 0 6( 4 )
5 . 5 16 6( 6 )
250
240
5.5106(6)
5.5106(4)
lot
'175
t6t
'loE
,aA
220
210
200
190
r' r8700
160
120
80
28
b(A)
21.858(s)
2t.87s(3)
21.881(2\
zt.egs(z)
21.e]3(4)
2l.el7(3)
21.e33(3)
"(l)
'ro.
esr(2)
(3)
r 0.957
(l )
r 0.965
r0.e6e(l)
(2)
l 0.e73
(r )
r 0.977
2640.
I 0 . 9 7 9)0 '2644,
t322.
1324.
1325.
1327.
1327.8
I 329.3
167
1 6 7. 3 4
167
16 7 . 3 4 ( )6
r67.34(3)
' t13637s..2(934 ( 2 )
)
I 3 3 4 . 5( 3 )
q Enoq/rl
l l.0134(8)
ll.0096(9)
I 1 . 0 0 2(65 )
10.9962(6)
r0.9e15(8)
r0.9882(4)
1 0 . 9 8 5 ()l
2t.e28(2)
1
0.980(l)
21.s04(2)
'10.971(2)
21.882(2)
i0.e59(l)
21.862(2)
/(1'3)
2623.
2627.
2630.7
r33r.e(2)
1329.6(2)
1327.e(3)
1326.7(1)
2r.e58(2)
21.e38(2)
21.920(3)
21.e02(2)
temperature before measurement and about two
hours were necessaryto obtain the 2d values of
the necessary reflections at constant temperatures. The cell dimensions were determined by
the least-squaresrefinement of from 5 to 13
independent refl ections.
The cell dimensions of bornite are listed for
X-ray study
different temperatures in Table 1. In Figure 2,
Natural bornite was used for a high-tempera- the cell dimensions are shown as a function of
ture single-crystal X-ray study. Precession and temperature in terms of tlose of the respective
cone-axis photograpbs at room temperature, subcells: a/2, b/4 and c/2 for the low form
195'C, arLd265"C show diffrastion patterns of
and a/2 for the intermediate form. Changes.of
the low, intermediate and high forms, respec- the cell dimensions take place discontinuously
tively. The diffraction pattern of the low form
at about t70"C and 235oC when temperature
shows orthorhombic symmetry as reported by is increased. These changes correspond to the
Koto & Morimoto (iglS). The intermediate phase transitions from the low to intermediate
form has weak reflections such as 200 and 600 form and from the intermediate to high form,
that were not observed in the metastable form
respectively, as detected by the DSC curves for
reported by Morimoto (1964) and Allais (1968). natural bornite (Fig. 1).
The lack of systematic absences in the interWhen temperature is decreased, however,
mediate form indicates that its possible space the change of the cell dimension at the transigroup is Fm3rn, F-43m or F432.
tion of 23O"C is continuous and the lines repreA single crystal of 0.10X0.08X0.08 mm was senting the change of the cell dimensions show
used for determination of the cell dimensions slightly different inclinations above and below
at high temperatures. The crystal was enclosed the transition temperature. This continuous
in an evacuated silica capillary tube 0.1 mm change of the cell dimension at the transition
thick, and was heated with a small furnace ar- explains well the fact that the transition was
tached to a four-circle diffractometer. Tem- not detected in the DSC cooling curve for naperature was measure{ by a thermocouple at- tural bornite. The other phase transition takes
tached near the crystdl and controlled witlin
place at about 165oC, which is lower by 5'C
-1"C. The 20 values were measuredfrom toom than in the case of heating. In this transition,
temperature to 300oC at intervals of 10-20oC the single crystal changed into twinned crystals
using MoKo radiation 0,.-0.70926A). The of orthorhombic symmetry with a 10.959(1),
specimen was kept for 3o'minutes at constant b 21.862(2), c 21.9O2(2)A at 28"C.
,!00
THB
CANADIAN
MINERALOGIST
168
-f
tt-'J
^'D'
-".-o'-o'-
roP-
?/i s.qg
5.46
150
T/"t
Fro. 2. Cell dimensions atrd volumes as a function of temperature for natural bornite for increasing
t€mperature (dotted lines) and for decreasing temperature (solid lines). The cell dimensions
and volumes of the subcells are plotted for the low and intermediate forms. Circles, triangles and
quadrangles of the low form are a, b and c axes, respectively,
Discussion
The thermal analysis and high-temperature
X-ray study have confirmed that bornite exists
in three stable polymorpbs: low, intermediate
and high forms. The low form changes to the
intermediate form at 170'C and the latter
changesto the high form at 235oC. The change
in cell dimensions as a function of temperature
@ig. D indicates the existenca of a transitional
state, because the cell dimensions observed at
the same temperatures differ and depend on
whether temperature is increasing or decreasing.
When temperature increases,the change in the
cell dimensions is discontinuous at the transition
temperature from the intermediate to the high
form, indicating a transition of the first order.
Ifowever, when temperature decreases, a different kin6 of transition takes place at almost
the same temperature, in which the cell dimension of a new trrhaseis continuous with that of
the high form. This transition is, therefore, apparently of a higher order. The cell dimensions
of this new phase are closer to those of the
high form than to the intermediate form. The
metastable form, characterized by the rule for
extra systematic absences in the diffraction
patterns (Morimoto 1964; Allais 1968; Putnis
& Grace L976), rs considered to be this new
phase supercooled to room temperature from
the high form.
A similar feature is observedin the low form
inverted from the meastable form: the c axis
of the low form maintains continuity with the
cell dimension of the metastable form at the
transition temperature. In this case, the low
form is considered to maintain to some extent
the disordered arrangementof atoms which does
not exist in natural bornite.
Cnvsrer- SrnucrunE, oF THE INtsnMsonrn
Fonu
Int ensity rnecrsuternent
Intensity data were collected at 185oC with
INTERMEDIATB
40t
FORM OF BORNITE
a four-circle diffractometer from the crystal
used for the determination of the cell dimensions. MoKa radiation monochromatized with
pyrolytic graphite was used throughout this investigation. The furnace and the temperaturecontrol system are the same as used for tle
measurement of the cell dimensions.
All reflections ft, & and I)0 were measured
within the range 2017Oo. The intensities were
sorrected for Lorentz and polarization factors.
The value of JaR is 1.0 for the crystal and no
absorption correction was made. The intensities
of equivalent reflections agree well for most
reflections, but they differ slightly for the weak
supercell reflections. Therefore, the intensities
of 187 independent reflections were obtained
by averaging the intensities of equivalent reflections. Of 187 measured intensities, 171 with
'observed relf'.1>2o(F"l) were regarded as
flections'.
TABLE2, FII{AL ATOMICPARAMETERS
ANDIIEIGHTSFORTHE TNTERMEDIATE
FORIIOF BORNITE
AT I85OC
Atom
l4(t-l)
Wyckoff
notatl on
32f
r',r(l-2) e6k
M(2)
96k
s(l)
24e
s(2)
8c
a
0 . 0 9 0 91( 0 )
0.1295(16)
0.3614(5)
0.2535(7)
1/4
y
s
c
1/2-a
o
1/4
z
a(12)
u
c
a . 6 ( 3 ) 0 . 3 1( 4 )
0 . r 0 6 e ( 2 3 ) r . 6 ( 4 ) 0 . 0 7 ( l)
o.ll33(8) 3.5(2) 0.33(1)
0
2.4(1)
1/4
2.6(2)
Starting from this model, the coordinates and
weights of metal atoms and the isotropic temperature factors were refined by the full-matrix
least-squares program of the UMCS system
(Sakurai 1967). The scattering factor of the
metal atom is assumedto be (51c"*f*)/6, \\e
R value obtained in this refinement was 0.27
for the observedreflections.
The metal atoms kept their initial weights
but were displaced from the c€ntres of the
tetrahedra, . and acquired large temperature
structure analysis
factors. At this stage of the analysis, a difference Fourier synthesis showed four positive
The cell dimension of the intermediate form
peaks around each detal atom, which were
possible
at l.85oC is 10.9806A Clable 1).
space displaced towards the faces of the sulfur tetragroups of the intermediate form are Fm3m,
hedra. The metal atoms were then divided into
F43m and F432, as determined from the diffour
sites and successivedifference Fourier synfraction aspect. The space group of the intertheseswere done.
mediate form was assumed to be centrosvmIn the final structure, the M(1) atom was
metric Fm3m, because only Fm3m incluOe.s
Pbca of. the low form as one of the subgroups. divided into four sites displaced from the cenTA8I.E3. OESENVEO
IIID CAICULAITO
STRUCTURA!
FACTORS
FORINIERI.IEDIAT8
BORIiITE
The structures of the low and intermediare
kl
Fo
fc
kI
Fo
lq
ktF6fc
kZ
Fo
fc
forms must be explained as superstructures of
n," I
3 7 116.5
130.7 tr t2 lt.4 2.6 l0 l4 9
'o.o
.8 8.5
that of the high form. The differences among
3
e
7.4
13.6
to 14 22.6 37.1 tz ti
9 ? --4!.9,!9.9
tz.g
0 4 603.1566,5 3 11 36.2 36.8 12 tZ 19.4 2,4 n S
the structures of these three polymorphs are
"
g q 3 q . 85 5 . 6 3 1 3 4 . 7 2 . e 4 . 3
S s o . s8 o . o
0 ,8 s26.452s.9 3 l5 0.5 14.9 3 3 211.2zo4.s s t oiA.z
az.t
due mainly to the arrangement of metal atoms
q lq -{.9 -9.1
3 t7 3.6 5.8
: s tr.a itz.r
6 s is'.Z 12.t
0 lz 15.5 t4.4
5 5 33.043.7 3 7 9o,7 g3.n s ri ia.a ib.e
and vacancies in the tetrahedra of sulfur atoms
9 1! 6.! 24.4 s 7 102.7104.7 : g ag.o :z.z
s rj
i:.0
q r q 9 9 . 1 q 9 . e 5 e 7 . 3 3 3 . 3 s n s : . s a s . a s i 5 r o.i
i.! :.r
which form tle face-centred cubic arrangez 2 3o.e 33.6 5 ll 5e.844.1 g.r: z.r o.g
I i i s'6.i
. 6 -t.t
ti.q
.15
1!.5
5
t3
19.1
z.t
3
l5.s 2.8
? ! !.9
t s
ment. The structure of the intermediate form
z 6 39.a 41,5 5 l5 7.8 lt.t
s u o.o o.o
i tl tz.2 s.o
tZ,? 17.7 7 7 33.3 3?.2
s
s
0.0
2o.i
?
-9
z
ii
a.6
i.q
with the space group Fm3m has two nonj is 6.i i.t
7 e 24.7 27.5
s t zs.z iz.q
? lg 8.q 8.4
?'!? 9J- 11.Q 7 11 t7.4 22.5 5 e 9.3 s.3
s c 16.8
' o . d -z3.z
q.z
equivalent tetrahedra of sulfur atoms in an
z t4 4.7 3.1
1
3
7
7.6 s.3
s tr ag.z ro.s
i ti
7 15 3.8 0.3
5 l3 0.0 14.0 s r: i.i o.s
? rq - ?.! ?.1
asymmetric unit. There are six metal atoms in
5 t5 14.5 3.8 il rr e.e z.r
! 1ry.9e,21.9 e 9 6.8 6.1
e lt 20.0 12.4 t 17 0.0 2,2
{ c - _ q . ?- ' ! ! . 9
6
4 ,8 t06.9108.9 9 13 19.0 ,!4.4 t t ts.s to.o h'o" 6 zle.t rsl.g
the subcell with /-5.540 and each sulfur tetrae15 0.0 0.8
7 s 3 5 . 22 a . s e e - s . i i 6 . s
1
1
9
?
.
9
1
9
.
6
3/q
hedron on the average contains
4 l2 l8r.z 169.3 ll lt 20.2 2.0
metal atoms.
z .tt1 3 tz.t ii.1
6 rri
ri.i ,ii.o
n t3 12.7 1.4
7
r3.4 s.e
{ l1 9.? l!.9
6 .ii ri,i ir.e
Therefore, two crystallographically different
4 16 7.4 22.6 h. 2
z rS rz.l rr.:
e i a q e . 06
- ai . to
q
e 7.0 18.8 s s ii.l
!q q a ? . 9 9 9 . 8 2 2 3 s 4 . 2 3 s 7 . 4 e
tetrahedra accommodate3/2 metal atoms in all,
11.? 34.4 2 4 0.0 15.7 g tt g.: t.t.z
s ro 6.5 e.z
6 l0 r'1.9 8.9
2 6 zso.A
2s4.s 9.13 ?.4 4.7
a
t
z
i
.
i
i.o
no matter what the detailed distribution of the
6 12 t3.6 20.1 2 8 0.6 6.1
s rs o.n c.z
s ti i.z i.t
q !1 r'!,! 9.4
z 1 01 1 9 . 9
1 2 1 . 4 l . t 1 1 1 8 . 7 l s . e r o i o r i . e 'zai . o
g
metal atoms is among the two tetrahedra.
6 16 4.0 4.3
2 12 0.0 2.2 n t: z.e :.2
t o i i i .' o
t-.7i
9 l-q2 2 p . ) 2 1 2 . ? 2 t 4 6 3 . 86 s . 2 h - 4
q
The average structure of the low form based
216 e.5 3.2
4 4 2 4 0 . 1 2 7 0 . 3 7 3 8 . 33 1 . 1
l
l
.
1
?
s
.
?
q l? r!.9 !ft.l
q a s.o to.s
4 4 4.0 1.2
t g iz.i ii.s
on the unit cell of the intermediate form (Koto
4 6 lr.3 16.9 4 8 349.6
-s 14 7.r ll.0
3 6 5 . 2 t t i -zaa..il -i g
i ..gt
4 I 12.2 13.3 q ro s.r tz.a
M9 ,9.1 q.4
i
ii
q tz o.o tz.o
l0 t2 t0.4 ll.5
g g i.i
& Morimoto 1975) indicates the existence of
4 ro 6.6 3.2
z.a
l0 t4 4.r 8.8
4 12 0.0 6.7
a u o.o ra.s
i t i -zs6..oo r o . s
1 21 2 9 8 . 8 6 0 . 4 4 1 4 0 . 0 4 . 0
two kinds of tetrahedra, one occupied by half
4 t6 55.652.4 s i:
ii.c
h' -I . 1 - . ,
4 16 ll.7 1.6
6 6 27,821.3
B
I 3 6 . 9 4 8 . 6 6 6 1 0 7 . 5 1 2 9 . 2 6 A 6 . 9 2 ? . 2 hI . 8 . t 4 0 . 5-ti3. s6 . 4
a metal atom and the other by one metal atom.
I 3 zn.7 1s0.9 6 8 16.4 5.6
6 ln 3.1 .t.7
e t0 l.i
6 10 129.8123.8 6 tz 17.4 t3.t
Therefore, the structure model of the interI I -9!.9 ,91.1
s re a.j rs.o
l 7 147.5
1s8.6 6 12 8.2 9.7
6 14 r2.t 7.3 to ro i.e i.g
6 1 4 3 9 . 1 4 ? . 6 6 1 6 3 . 5 o . B r o r i r -i ."o o . c
mediate form was constructed on this distribu1 _9 lg.q !.4
| 11 77.6 74.3 6 16 l4.l 0.8
A g e o . t t O . O t' q" g
I 8 lt.6 3.1
tion of metal atoms. Sulfur atoms were first
a ro rr.r ri.i
I ll r9,q q.6
i -e8s..0o o . r
I l0 6,9 6.3
8 t2 lle.7 lol.n
M
9.q e.z
s ri
i.s
1 17 7.2 1.1
8 12 7.1 0.4
kept at the nodes of a cubic face-centred latI 14 6.4 5.7
3 3 1 9 5 . 8t 8 8 . 5 I 1 4 7 . 2 1 . 5 t o t o o . n z . :
3
5
7
3
.
3
8
8
.
0
l
0
t
0
5
6
.
7
5
4
.
8
t
z
t
o
s
,
z
tice and metal atoms at the centres of the
s.0
tetrahedra.
THE
MINERALOGIST
CANADIAN
tre towards the faces of tle tetrahedron, whereas the MQ) atom was divided into only tlree
sites slightly displaced around a three-fold axis.
1!s sseldinates and weights of all atoms were
finally obtained by fuIl-matrix least-squares re'
finement. Although the total number of metal
atoms was not constrained in this refinementn
tle total number converged to 48.3 per unit
cell. The R value is 0.175 for all reflections and
0.165 for observedreflections. The final atomic
paramete$ are given in Table 2, and the comparison of lF.l and lf"l in Table 3. All computations were done at the Computation Centre of
Osaka University.
1t4
Description of the structure
The projection along the a axis is shown
schematically in Figure 3. Sulfur atoms form
the cubic closest packing and metal atoms are
statistically distributed in the sulfur tetrahedra.
The structure is characterized by a three-dimensional alternating arrangement of two kinds of
cubes denoted by Ml and M2 (Fig. 4). Both
cubes have the unit-cell dimensions of the high
form and have tbe antifluorite-type structure,
though they contain different numbers of metal
atoms.
In the ML qubes, each sulfur tetrahedron
contains one M(1-1) and three equivalent
M(l-z) atoms (Fig. 4). The total number of
metal atoms is 0.52 in each tetrahedron. Therefore, each ML cube with eight tetrahedra con-
Frc. 4. The distribution of metal atoms in the sul'
fur tetrahedra in the intermediate form.
tains approximately four metal atoms. On the
other hand, each tetrahedron in the M2 qtbe
contains three equ.ivalent M(2) atoms (Fig. 4).
The total number of metal atoms in each tetrahedron is 0.99; therefore, each M2 cube contains about 8 metal atoms in all. The metalsulfur distanses are given in Table 4.
All the metal atoms are displaced towards
the centres of the triangles of the sulfur tetrahedra. The M(I-L) and three M(l-2) atoms in
the ML cube are displaced from the centre of
the tetrahedra toward the centres of the sulfur
triangles forming the tetrahedra. The value of
2.284 of.the M(1-1)-S distancesis equal to the
value of 2.28A found in a triangular coordrBation in the low form Koto & Morimoto 1975).
The average value of the three short M(1-2)-S
distances
--fiiangutaris 2344.
coordination of Cu atoms has
been reported in several sulfide minerals @vans
& Konnert L967), whercas Fe atoms have regular tetrahedral coordination in sulfide minerals'
Therefore, tjie ML cube is considered to be occupied onty by copper atoms.
in t}re liZ iuben three M(2) atoms in a sulfur
temahedron are slightly displaced from the sentre of the tetrahedron towards the sulfur triangles around the three-fold axis. The average of
2.lZA tor the M(2)-S distances is close to the
averase
-tn"- of. 2364 for the metal-sulfur distanqes
i"
Oirtorted tetrahedral coordination in the
low form.
TETRAHEDRON
TI THESULFUR
DISTAIICES(fr)
TABLE4. METAL.SULFUR
s ( )r
s(2)
M(j-r)
2.28(l)x3
3.03(r)
M(l-z)
2 . 2 e ( 2 )x 2
2.5s(2)
2.44(2\
?,40
N(2)
? . 3 q ( 1 )x 2
z . 6 0 (l )
2.2e(1)
2.37
fi..t|l
Frc. 3. The structure (from .e=0 to Ve of the unit
cell) of the intermediate form. The Ml arrd,M2
cubes are shown. Sulfur atoms are represented
by large circles. Dotted lines join the cenfes of
sulfur tetrahedra in the cubes.
rThe dlstance in the parcnthesis ls the mean of shortest
three one9.
INTERMBDIATE
S-cube A-cube
403
FORM OF BORNITE
M1-cube M2-cube
l.-
a -_--{
O sulfur
I
metal
E Yacancy
b
F-
a __{
(a)
(c)
FIG. 5. Three polymorphs of bornitel (a) low form, O) intermediate form and (c) high form. The
statistical distributions of metal atoms in the sulfur tetrahedra are shown by solid circlos representing the weights.
The large temperature factors of the metal
atoms are considered to be due to positional
disorder rather than to thermal vibrations at
high temperature, because large temperature
factors are obtained even for the metal atoms
of the low form (Koto & Morimoto 1975).
Splitting into three or four sites of metal atoms,
as found in this investigation, approximates the
actual statistical distribution of atoms as is discussed in more detail farther below.
Comparison of the stuctures of polymorphs
The structures of three polymorphs of bornite
(low, intermediate and high forms) are schematically illustrated in Figure 5. The structure
of the intermediate form is constructed from
two different kinds of cubes as in the low form
(Fig. 5b). These have the same size as the unit
cell of the high form and alternate three-d,imensionally, resulting in a mosaic arrangemenL
Both kinds of cubes of the intermediate form
have the antifluorite-type structure, but those
of the low form are of sphalerite-type (,S) and
antifluorite-type (l) @g. 5a).
From a structural viewpoint, the structure of
the intermediate form can be considered to
represent an intermediate stage between high
and low forms. For example, the ML and M2
cubes of the intermediate form contain numbers of metal atoms similar to the S and ./
cubes of the low form, respecnvely. Both M2
and .r4 cubes have eight atoms that are slightly
displaced towards the adjacent subes. On the
other hand, characteristic statistical distributions
of the metal atoms are observed in both intermediate and high forms.
The statistical distribution of the intermediate form represents approximately the superposition of the atomic distribution of the low
form displaced by b/2. Similarly, the statistical
distribution of the high form represents the superposition of the intermediate forms displaced
by a/2. The intermediate form can be interpreted as a partly disordered low form, because
the intermediate form has long-range order,ing
in alternation of the S and z4 cubeso but has
only short-range ordering in orientation of the
S cubes of the low form. Disorder in orientation
of the S cubes is possible by antiphase-domain
boundaries with displacement vectors of b/2
or (a*d/2 of the low form. Thus the intermediate form is considered to be an assem-
THB cANADIAN MINERALOGIST
4O4
blage of the antiphase domains of the S and A
cubes in the order of coherency for the X-rays.
On the other hand, the intermediate form can
be distinguished from the high form by longrange ordering of the Ml ar.d M2 cubes and
convergence of the statistical distribution of
metal atoms to three or four positions from
twenty-four possible positions in the sulfur
tetrahedron.
Allais (1968) studied at room temperature
the structure of bornite rapidly cooled from
high temperature. This structure bears a close
resemblance to the low form in that it forms
sphalerite-like and antifluorite-like cubes. Based
on this statistical structure, it is difficult to
explain the observed extra systematic absences
if cubic symmetry is assumed.
The metastable form seems to be closer to
the high form than to the intermediate form
judging from the cell-dimensions and the lack
of extra systematic absences in the intermediate form. The experiment by Putnis & Grace
(1976) also supports this. In the structure of
the metastable form, some of the metal atoms
retain the distribution of the high form, which
must explain the extra systematic absences.
AcKNowLBDGMENTS
We thank Dr. M, Watanabe of National Institute for Researchesin Inorganic Materials
for supplying a small furnace for a four-circle
diffractometer. Dr. M. Tokonami of tlis institute provided helpful suggestions.Part of the
expensesof this work was defrayed by a research grant from the Ministry of Education of
the JapaneseGovernment.
(1968): Etude de I'ordre dans les sulfures
de cuivre. Structure de la phasem6tastablede
la bornite. Soc' frang. Min€ral. Crist. Bull. 9L,
600-620.
DoNNAy,G., DoNNeY,J. D. H. & Kur-r-nnuo,G.
(1958): Crystal and twin structure of digenite,
Mineral. 49, 228'242,
Cl,sSs,Amer,
EvANs,H. T. JR. & KoNNnnl J. A. (1976): Crystal structure refinement of covellite. Amer, Iulineral. 6L, 996-1000.
FRUEH,A. J., tn. (1950): Disorder in the mineral
Amer. l. 9ci.248, 185-192,
bornite, Cu5FeSa.
Koro, K. & Momntoro, N. (1972): Superstructure
of bornite determinedby a modified partial Patterson method.Acta Cryst. A28' S70 (Abstr.).
(1975): Superstructure
investiga&tion of bornite, CurFeSo by the modified partial
Pattersonfunction. Acta Cryst. B,gL'2268'2273,
MoRrMoro, N. (1964): Structure of two polyActa Cryst. 17, 35Imorphic forms of CusFeS4.
360.
& Kur.r-rruo, G. (1961): Polymorphismin
bornite. Amer, Mineral. 46, 1270't282.
Purxrs, A. & GRAcB,I. (L916):The transformation
behaviour of bornite. Conff. Mineral. Petrology
55,311-315.
SAruner,T., ed. (1967): UniversityCrystallographic
Computation Program System.Cryst. Soc. Japan.
SuceKt,A. & Sruvre,H. (1965): Syntheticsulfide
minerals (II), Mem. Fac. Eng. Yamaguchi Univ.
L5,33-47.
TArrxo, S., Mesuvtoto, K. & Keurcarcru, T.
(1968): Electric and magnetic properties of
t. Sci. Hiroshima aniv., Ser.
bornite, CuoFeSa.
c. 6, 321-332.
TuNELL,G. & Aoeus, C. E. (1949): On the symmetry and crystal structure of bomite, Amer.
Mineral, 84,8A-829.
RBFERBNCES
Ar.r.lrs, G. (1965): Mesurede la susceptibilit6magn6tique de la bornite (CuslFeSJ.C.R. Acad,
Sci,Paris.260, 6583-6585.
Received.March 1978; revised manuscript accepted
lune 1978.