Journal of Fluorine Chemistry, 5 (1975) 93 - 98

0 Elsevier Sequoia S.A., Lausanne - Printed in the Netherlands 93

REACTIONS OF TITANIUM DIOXIDE WITH METAL FLUORIDES

B. LENG and J. H. MOSS

Department of Chemistry, Teesside Polytechnic, Borough Road, Middlesbrough, Teesside TSI 3BA (Great Britain)

(Received February 24, 1974)

Summary

TiOs reacts with several condensed-phase metal fluorides at elevated temperatures to form TiF, and the metal oxide. Known mixed titanium metal oxides are usually observed as intermediates. FeOF is an intermediate from the system FeFs + TiO,. With CaF, and BaFs, the reaction did not proceed beyond the intermediates CaTiO, and BaTiO,. Reactions usually start at approximately the temperatures predicted from thermochemical data.

Introduction

The reaction of TiOs with condensed-phase metal fluorides has been little investigated. Titanium has been volatilised by heating the oxide with fluorides at high temperatures for spectroscopic studies [l] . Potassium fluoride will react with Ti02 but there is no evidence for a volatile titanium species in these reactions [ 2 - 41.

Experimental

Reagents were purchased commercially (TiOz, CuO, CdFa, PbF2, CaF,, BaF,, CrF, and FeF,) or prepared by previously reported methods (TiOFs [5], FeFs [6], CuFa [7], AlFs [8], NiFa [9], CoF,[lO], ZnF, [lo] and MnF, [lo])

Thermal analyses (Table 1) were carried out in a 5 cm3 platinum crucible on a Stanton TRl or HT5 thermobalance using m mol quantities of TiOa under nitrogen with a gas flow rate of 400 cm3 min-I. X-Ray diffraction traces were recorded on intermediates and involatile products using a Siemens type-F diffractometer and CuKa radiation. Hygroscopic materials were hand- led in a dry box over Pz05.

Reactions (Table 2) in which TiF, was collected were carried out in uacuo in a Lawrence Smith platinum crucible (length 90 mm, internal diam- eter 21 mm). The exit gases were passed via PTFE connectors to a cold- finger condenser. The products collected on the cold finger were purified by sublimation.

TA

BL

E 1

Th

erm

al

anal

ysis

ex

peri

men

ts

RS

XtE

+ll

tSR

eact

ion

co

ndi

tion

sP

mdu

cLs

Flu

orid

eO

xide

Mol

e ra

tio

Hea

tin

g T

empe

ratu

re

Tem

pera

ture

s?&

Mas

s R

TI

in

solI

d p

rodu

ctsb

?F ,

n so

l,d

pro

du

cts”

So

l~ls

Inte

rmed

iate

s at

fir

stn

uor

ide

of m

inim

um

10S.S

oxid

ere

acti

on r

ate

Obs

erve

dm

mim

um

of

reac

tion

Cal

cula

ted

Obs

erve

dC

alcu

late

dra

te

.%?F

CaF

,B

aF2

MnF

2F

eF2

CoF

zC

oF2

NiF

2N

iFZ

CuF

2Z

llF

,Zn

Fz

CdF

ZP

bF2

AIF

CrF

3F

eF3

TiF

,

TiO

,5.

127

0RO

O95

0d69

021

50

0T

iOz

21

400

1280

1350

64

619

3,

19.6

TiO

z2:

140

013

0013

50d

3.R

9.8,

10

0T

iOz

2:l

400

890

940

9.7

1x 1

, 17

9T

!02

2:1

270

600

950

800

41.7

0.7,

0

5

TiO

z2.

1,4

0053

012

5077

0.

1020

,11

50‘4

5 5

66

, 6

0T

iOZ

2:3

400

530

950d

18.1

31 9

, 31

7

TiO

z2.

140

055

010

0074

04

4.9

0.2

, 0

.3T

iOZ

2.3

400

620

900

35 t

i:30

9.

31.0

TiO

z3

:l27

040

085

032

60

0TiO

l2.

14

00

50

01000

760,8

10

3x

x2

9.

‘9

TiO

s2:3

400

620

8t30

36.7

28.1

. 2x

2

TiO

g2

1,4

00800

850d

10 6

11 “,,

I

1

TiO

q2-l

27

0700

750d

5. 5

8.i.

9 0

TiO

,4:3

27

0770

8SO

d9

.227.1

, 269

TiO

z5.

327

060

085

068

04

3 8

i 1

, I 3

TiO

,1:

327

045

0X

5063

04

9.8

02

, 0

.3

CU

O2.

1927

020

035

0

a T

he

tem

pwat

ure

ra

nge

s aw

th

ose

for

wh

ich

th

e ra

te o

f w

righ

t 1

0~s

was

grr

atrr

tha

n 0

001

g 11

1 ;I

m!n

.’

Th

e ca

lcu

late

d an

slya

~ f

igu

res

wer

e ob

tain

ed f

rom

th

r pc

rcrn

tagc

mas

s IO

SS b

y as

sum

ing

TiF

, w

as t

he o

nly

vo1

atJ~

~ m

,itcr

l,ii

’ L

ines

due

to

an u

nld

enti

fwd

mat

eria

l w

crc

pres

ent

here

.d

Rea

ctio

n h

eld

at t

he

max

imu

m t

empe

ratu

re f

or 2

h.

’ T

he

prea

rncp

UC

xon

(lll

) in

th

e so

lid

prod

uct

s is

pr

obab

ly

due

to t

he i

nab

ilit

y to

com

piri

rly

rxr

iud

c .~

tmo\

phrr

r o

x\gc

,n L

rom

the

th

erm

obal

ancr

f Int

erm

edaa

tea

at

cwon

d m

inim

um

of

rrar

tlon

rat

e. (

:o,T

10

, an

d (_

oF2,

an

d a

t th

>rd

min

imu

m o

f r~

act,<

m m

tv.

CO

O,

Co2

T,G

j a

nd (

‘oF

2g

Th

e C

uO w

as p

lace

d o

n t

op o

f th

r T

IF,

wh

xh s

ubl

imed

th

rou

gh i

t.

00

19 5

10.0

lr,

90

61

26

00

03

1.8 “2

2 1

30.9 0.0

14.0

2.9

,I .628

.1 s1.5

c, .6 23.6 / _

,1 i

,IO.3

GO

(“C

)

95

Titanium was analysed by the peroxide titanium yellow method [ 1 l] and fluoride by pyrohydrolysis and alkali titration [ 121.

Discussion

Several condensed-phase metal fluorides react with TiOa at elevated temperatures, only PbFa and CdFa of those investigated showing no evidence of reaction (Table 1). TiF, was isolated from the reactions of CuF2 and CrF, with TiO,, and the net weight loss in conjunction with analysis of the residues showed TiF, to be the material volatilising in most other cases. With MnF,, AlFs and to a lesser extent CoFa, extensive volatilisation of the reacting fluoride occurred in addition to TiF,.

The observed reactions of fluorides with TiO, may be subdivided into three classes: (a) direct formation of TiF, and of the metal oxide without the formation of observable intermediates; (b) reaction to form mixed oxides, with possible further reaction of the fluoride with the mixed oxides to form more TiF, and the metal oxide; and (c) reaction uia oxyfluoride intermediates.

(a) Direct formation of TiF, and the metal oxide: Only in the case of CuFa + TiOa has the simple one-stage reaction been observed.

2CuF2 + TiOz = TiF, + 2CuO

Only a small excess of CuF, is needed for complete volatilisation of the titanium. The reaction is partly reversible as CuO will react with TiF, at lower temperatures to form CuO + TiOF,. A further stage of the reverse reaction is the known pyrolysis of TiOF, to form TiF, + TiOa [5].

2TiOF, = TiF, + TiOa

(b) Reaction to form mixed oxides: Mixed oxide formation was observed in the reactions of TiOz with CaF,, BaF,, FeF,, CoF2, NiF, and ZnF,. With CaFa and BaF,, formation of the mixed oxides was incomplete after several hours at 1350 “C and further reaction was not observed. In all other cases further reaction to form the metal oxide occurred, although Co,Ti04 and ZnzTi04 were not completely converted to Co0 and ZnO, respectively. The mixed oxides were not usually obtained free from other reacting materials but by using a deficiency of CoF, it was also possible to prepare CoTiOs. In those cases in which several mixed oxides were formed, increasing the reaction temperature tended to increase the metal content of the mixed titanium oxide. In the case of cobalt, the various stages of the thermal anal- ysis could be reasonably assigned to the formation of particular mixed oxides. Usually all the known metal titanium oxides were observed as reaction inter- mediates but, in the case of nickel, NiaTiaOs and Ni,TiO, were not found.

(c) Formation of oxyfluoride intermediates: FeFs reacted with TiOa in two stages, FeOF being detected as the intermediate product at the end of the first stage. The reaction of CrFs with TiOa showed a similar two-stage reaction. X-Ray diffraction studies on the intermediate stage however only showed the

TA

BL

E

2

Prep

arat

ive

expe

rim

ents

Rea

ctan

ts

Rea

ctio

n

con

diti

ons

An

alys

is

of

TiF

, pr

odu

ceda

A

nal

ysis

of

so

lid

resi

dues

Flu

orid

e O

xide

M

ole

rati

o W

eigh

t of

T

emp.

(O

C)

Tim

e(m

in)

Mas

s g

%

Ti(

%)

F(%

) T

i(%

) F

(%)

flu

orid

e :

sam

ple

(9)

loss

(g)

nu

ide

Fou

nd

Cal

c.b

Fou

nd

Cal

eb

Cu

Fz

TiO

z 5:

2 6.

656

850

60

2.61

0 1.

390

53

38.6

, 38

.6

60.2

, 61

.0

0.0

0.0

8.8,

8.

7 8.

7

CrF

, T

iOz

4:3

0.67

2 85

0 60

0.

3052

0.

131

40

38.6

, 38

.5

60.9

, 60

.5

7.3,

7.

5 7.

5

a P

ure

TiF

, co

nta

ins

38.6

%

Ti

and

61.3

%

F.

b T

he

calc

ula

ted

anal

ysis

fi

gure

s w

ere

obta

ined

fr

om

the

mas

s lo

ss

by

assu

min

g T

iF,

was

th

e on

ly

vola

tile

m

ater

ial.

TA

BL

E

3

Initi

al

reac

tion

tem

pera

ture

s __

_-

Fluo

ride

re

actin

g w

ith

TiO

z A

gF

CaF

2 B

aFz

Mn

FZ

F

eFz

CoF

2 N

iF,

Cu

F2

Zn

F.2

C

dF,

PbF

2 A

IF,

CrF

, F

eF3

Obs

erve

d in

itia

l re

acti

on

tem

pera

ture

(“

C)”

60

0 12

80

1300

89

0 60

0 53

0 55

0 40

0 58

0 77

0 60

0 45

0

Cal

cula

ted

init

ial

reac

tion

te

mpe

ratu

re

(“C

)b

375

1970

22

50

550

700

630

640

418

570

810

860

790

280

350’

Cal

cula

ted

init

ial

reac

tion

te

mpe

ratu

re

for

form

atio

n

of

the

mix

ed

oxid

e M

TiO

s (“

C)b

12

70

1070

_

a T

he

obse

rved

in

itia

l re

acti

on

tem

pera

ture

w

as t

he

low

est

tem

pera

ture

at

wh

ich

th

e ra

te

of

TiF

, lo

ss

was

gre

ater

th

an

0.00

1 g

in 5

min

. b

Th

e ca

lcu

late

d in

itia

l re

acti

on

tem

pera

ture

w

as t

he

low

est

tem

pera

ture

at

wh

ich

th

e va

pou

r pr

rssu

re

of

TiF

, w

as c

alcu

late

d to

be

gr

eate

r th

an

1 /I

000

at

m.

’ C

alcu

lati

on

from

es

tim

ated

da

ta

give

n

in r

ef.

15.

97

presence of reactants and products. AgF also reacted with TiOa in a two-stage reaction. The final solid product was merely metallic silver, presumably formed from the decomposition of silver oxide which is unstable at the reac- tion temperature [ 131.

4AgF + TiOa = 4Ag + TiF, + 0,

In the first stage of the reaction no titanium was volatilised but there was a small weight loss. Metallic silver and an unidentified material were present in the solids obtained at the end of the first stage. The presence of silver suggests that the volatile material lost was oxygen. If this was so, one oxygen atom was lost for each TiOs molecule reacted. This suggests that the unidentified material was an AgF-TiOF, complex, possibly Ag,TiOFq analogous to the known KzTiOF4 [ 141.

Comparison of the observed reaction temperatures with those predicted from thermochemical data is interesting (Table 3). If it is assumed that the first detectable mass loss in the thermal analysis occurred at ea. l/1000 atm pressure, the initial reaction temperature of the metathetical reactions

4MF, + nTi0, = 4MO,,, + nTiF,

may be estimated from thermochemical data to f 100 o C. The absence of observed reactions of CdFa and PbF, with TiOs is not surprising as both the fluorides are quite volatile at the predicted reaction temperatures. The small extent of reaction with AlF, may be similarly understood. For other fluorides, the observed reaction temperatures show a rough correspondence with those calculated. The largest deviations are for the reactions of CaF, and BaF, with TiOz where stable mixed oxides are formed. If the initial reaction temper-, ature calculations are repeated for the reactions

2MF, + 3Ti02 = 2MTiOs + TiF, (M = Ca, Ba)

the reaction temperatures fit much more closely with those observed. The failure to observe the oxides CaO and BaO as reaction products in these sys- tems is also accounted for. In other cases where mixed oxides are formed, a lowering of the observed reaction temperature is also observed. In the absence of thermochemical data, it is not possible to estimate the initial reaction temperatures. The relatively small lowering of the observed reaction temper- atures, coupled with the observation of further fluorination of the mixed metal titanium oxides, suggests that the heats of formation of these mixed oxides from their constituent metal oxides are small compared with those of CaTiOs and BaTiOa. Of the remaining significant discrepancies between observed and estimated reaction temperatures, that of the MnF2 + TiOa reac- tion may arise from the particularly large error in the reported heat of form- ation of MnFa [16] while that of the CrF, + TiOz reaction is possibly-assoc- iated with the slow reaction kinetics frequent with chromium(II1) compounds [ 171. As AgF initially reacts with evolution of oxygen, the calculation is not applicable to this system.

98

References

1 E. Schroll, U.S. Govt. Res. Rep., 39 (1964) 19; Chem. Abstr., 61 (1964) 15649d.2 A. A. Fotiev and F. P. Sheinkman, Zh. Neorg. Khim., 12 (1967) 135.3 V. G. Bamburov and N. V. Demenev, Tr. Inst. Khim. Akad. Nauk SSSR, Ural’ski.

Filial., 7 (1963) 19.4 V. G. Bamburov and A. A. Fotiev, Izv. Sibirsk. Otd. Akad. Nauk SSSR, Ser. Khim.

Nauk, 3 (1963) 42.5 J. H. Moss and A. Wright, J. Fluorine Chem., 5 (1974) 163.6 R. Colton and J. H. Canterford, Halides of the First Row Transition Metals, Inter-

science, New York, 1969, p. 272.7 C. Poulenc, Ann. Chim. Phys., 2 (1894) 58.8 W. Blitz and E. Rahlfs, Z. Anorg. Allg. Chem., 166 (1927) 370.9 0. Ruff and E. Ascher, Z. Anorg. Allg. Chem., 183 (1929) 193.

10 G. Bauer, Handbook of Preparative Inorganic Chemistry, Academic Press, London,1963, pp. 242 - 267.

11 F. 0. Snell and C. T. Snell, Calorimetric Methods of Analysis, Vol. IIA, Van Nostrand.New York, 1959, p. 325.

12 J. D. Rushmere and H. Mason, UKAEA (Industrial Group) Report, SCS-R-392, 1959.13 G. N. Lewis, 2. Phys. Chem., 52 (1905) 310.14 I. P. Sorokin, G. E. Dmitrevskii, Yu. I. Kol’tsov, B. S. Blinov and A. A. Belitskaya,

Russ. J. Inorg. Chem., 11 (1966) 1524.15 0. Kubaschewski, E. L. Evans and C. B. Alcock, Metallurgical Thermochemistry,

4th ed., Pergamon Press, Oxford, 1967.16 K. Jellinek and A. Rudat, Z. Anorg. Allg. Chem., 175 (1928) 281.17 F. Basolo and R. G. Pearson, Mechanisms of Organic Reactions, Wiley, New York,

1958, p. 121.