Thermal decomposition of some boracites

Thermochrmxa Acta, 56 (1982) 359-369 Elsevier Scienttfic Publishing Company, Amsterdam-Prmted

l’I-IERMAL DECOMPOSITION

m The Netherlands

359

OF SOME BOIPACITES

P. TISSOT, J. PAINOT l, J.P. RIVERA and H. SCHMID Department qf inorganic, Anaiytrcal and Apphed Chemistry, Geneva 4 (Swrtzerland)

Unrversrty of Geneva, CH-1211

(Received 27 January 1982)

ABSTRACT The thermal decompositron under argon of the boracites MsB,O,sX (M=divalent metal ion, X= halogen ion or OH-) has been studied by thennogravlmetry (TG) and differential thermal analysis (DTA) for the compositions MgCl. MgOH, CrCl, CrBr, CrI, MnCl, MnBr, MnI, FeCl, FeBr, FeI, CoCl, CoBr, CoI, NICI, NiBr, NiI, CuCl, CuBr, ZnCl. ZnBr, ZnI, CdCl, CdBr and CdI. The lithium boracites Li4B70t2X, with X=Cl and Br, have also been studied. The stability of the boracttes systemaucally decreases in the order MCl>MBr>MI. The most stable boracite is CrCl (1% weight loss at 1242OC) and the least stable one studied

here LSMgON (1% weight loss at 717°C).

INTRODUCTION

Boracites are compounds with the general formula M3B70,3X, hereafter designated by MX, where M is Mg, Cr, Mn, Fe, Co, ‘Ni, Cu, Zn or Cd, and X is OH, F, Cl, Br, I or NO, [l-3]. Chalcogenide boracites with X = S, Se or Te and charge compensating oxygen vacancies [4,5], lithium boracites [7] Li,B,0,2X [6] with X = Cl, Br, I, and lithium sulfur boracite Li,B,O,,S are also known. Many boracites are of interest because of their ferroelectric/ferroelastic phases and most of the 3d transition element boracites because of simultaneous ferroelectricity, ferroelasticity and weak ferromagnetism [l,S]. Lithium boracites are of interest because of their ionic conductivity [7]. With a view to growing single crystals for physical characterizations, the knowledge of the stability limit is very important. Therefore a systematic study of the thermal decomposition of a series of 26 halogen boracites and of MgOH with simultaneous TG and DTA analysis has been undertaken and is described in this paper. It has to be stressed that such decompositions necessarily take place under non-equilibrium conditions and finally cannot --

* Present address: Thermocompact SA, Cl-I-1202 Geneva, Suitzerland.

OO4O-6031/82/OOOO-0000/$02.75

0 1982 Elsevier Scientific Publishing Company

.,

- 1403 - 1433 -1403

1310 1322 1452 1223 1326 1249

Fe1 NrBr CuCl

FeCl CoCl COI ZnCl ZnI ZnI

1.8X IO-’ 4.0x lo+ 5.6X IO+ 6.7x 1O-5 1.3x lo-” ?

pressure (at@

TOtid

BOCI> F&l 2 > B,O, BOClSoCl 2 > B,O, B203> >CoCI, (I not measured) Zn>O,>ZnCI, >>BzO, Zn > 0, > r B,O, (I not measured) 20 I>ZnI

I>H,O>HBO, >Fc>B,O, >O, H,O>B,OJ(?)>NtBr>HBO,(‘CO,‘? “)>Br>O, >NI 0,>H,O>CuCl=CI~Cu~HBO,(‘CO,?“)~B,O,

Components of gas phase, relattve mtensny sequence at T(K)

-

_

_, .

1

_^

-

.

-

-

%

I

_.

.

.

.

_

’ Remark: according to Delfmo and Gentile [ 131HBO:, which has the same mass number as CO:, IS more probable. b J%In1=8X 10m8(units omttted) at room temperature; no indrcatron of decompositron temperature.

Temp (K)

MX

Ref.

Gaseous decomposrtton products of some boracitcs. detected as positively charged species in a mats spectromctcr (data compiled from rcfs [ 121and i 141)

TABLE 1

361

replace more detailed studies of the phase diagrams. Howejler, they prove to be vzry useful in solving crystal growth problems. Some boracite compositions have already been studied by previous authors with respect to their stability. For MnOH Muller [9] anti Levasseur [lo] respectively find that strong decomposition starts at about 700°C and comes to an end at about ‘15OOC. Kravchuk and Lazebnik [I 1] report that FeOH remains stable up tc 650°C under hydrothermal conditions. Gallagher [ 121 finds endothermic peaks-interpreted by melting-for C~Cl(l025~C), NiBr (1150°C) and Fe1 (965°C). Delfino and Gentile [ 131 report thermogravimetric curves measured in air for MgCl, CuCl, NiBr and FeI, and measured in vacuum for MgCl, NiBr and FeI. They contest the opinion [12] tha! CuCl, NiBr and Fe1 meit congruently. Vedenkina et al. [ 141 studied decomposition by mass spectrornetry for FeCl, FeBr, CoCI, CoI, ZnCl and ZnI and find the formation of MO Bz03(+ B,O,(,), MCI,,,, or MIBj + I,(,) in a first step and the final products to be B,O, and MO for Co and Fe boracites, and B,03 and free Zn fol zinc boracites. The complete list of gaseous species detected during decomposition of some boracites by means of mass spectrometry [ 12,141 has been compiled in Table 1. l

EXPERIMENTAL

All boracite compositions have been synthesized by chemical vapour transport [ 15-171 with the exception of MgOH which has been obtained by hydrothermal synthesis [93 and which was characterized by small amounts of manganese as impurity (i.e. MnOH served as seed for MgOH). Thermal analysis was carried out using a Mettler TAl device, equipped for simultEneous TG and DTA. The samples weighed 15-20 mg (small crystals, c 0.1 mm). Platinum crucibles were used. A heating rate of 4 deg min- ’ and an argon flow of 5 1 h-l were chosen.

RESULTS

AND

DISCUSSION

Fig. 1 shows the TG and DTA curves obtained for all boracites studied. The weight loss and the peak height are normalized for unit weight, so that all the curves are directly comparable. Most of the compounds present a complex decomposition, usually with two endothermic peaks on heating and sometimes a multiple step weight loss. The DTA of the chromium boracites has also been performed in sealed crucibles, in order to prevent decomposition (results not presented here); the peaks observed are the same as the former ones obtained with an open crucible and were reversible on cooling. The nature of the double peaks is not yet understood. Table 2 reports the temperature of the ‘beginning of decomposition’ (i.e. at

864 89716)

948 990[131 1002[13]

1242

1071

MI3

Cr

Mll

TG

Cl

X

Ll

M

“(2

1200 1050

983

1070

933

I 100

830

823

1196

TG

DTA

TG

-

I

Br

1250

870

DTA

-._

977

1150

DTA

Temperatures of ‘beynnmg of decomposition’ (I % weight loss, TG) and of DTA peak; comparison

TABLE 2

._

._--..--...-.“^__

700[9]n 700[lo] b

717

TG

OH

with hteraturc

715

DTA

in

Argon ? ?

Argon

Argon Vacuum AU

Argon 1-

Atmosphere

h data (temperatures

1130

1040[14]a 1005

929

Ni

cu

960

516[14]a 958

Cd

951

889 1032[131

923

1075[131 1033[131

993

1049

999

940

1125 1150(12]

955

955

852

712[14]”

852

1013 643[14]=

1035

925

920 908[13] 9691131

1003

a Does not correspond to 1%weight loss. b 700°C corresponds to a loss of 0.3 OH mole-‘; strong weight loss above 7OOOC.

920

929

Zn

9781131

1075

985[141a 1073

co

993 1025[121

985

1014

Pe

Argon Air Vacuum Argon

915

Argon Au Vacuum

Argon Au Vacuum

Argon Vacuum

Argon Air Vacuum

900

975

1070

980 965112)

364

(a)

r 700

’

I 800

’

I 900

I

I

1000

A

rlbo

’

liO0

‘T oc

FJ~ I(a)-(c). TG and DTA curves of the boracltes studlzd. Heating rate, 4 deg min-‘, atmosphere, argon (5 1 h-‘); sample weight. 15-20 mg

1% weight loss) and that of the first DTA Feak. Generally these two temperatures are close to each other; however in the case of the boracites MgOH, M&I, CrI, NiCl, NiBr, FeI, LMnBr, MnI, NiI, CuCI, ZnBr, ZnI, CdI and Cal, the decomposition starts at a lower temperature than the first DTA peak. It is interestin, * to note that all MI boracites behave in this manner, probably because of the smaller bond strength caused by the large size of iodine. Table3 reports the weight loss, expressed in percent, measured between room temperature and the beginning of the steady state. (The boracites having an unsharp end of weight loss are not included in this Table,) The:

365

,

I

I

TABLE

I

900

800

,

1000

1100

I 1200

Weight

CUCI

CuBr ZnCl ZnBr Znl CdCl CdBr Cd1

T “C

3

Weight loss (W) of some boracltes (characterized room temperature and the steady state

Fe1 NiCl NiBr NiI

I

loss %

Exptl.

Calcd.

17.5

26.8

12.3 15.2 20.8 15.8 19.7 7.0 14.6 17.9

13.1 20.1

14.3

14.0

17.1 18.7

19.4 24.5

26.6 13.2

20.2 13.2 20.0 26.3

[eqn. (l)]

by sharp decomposltlon

points)

between

366

I900

‘

900

results are hypothetical 2 M3B70,3X

1

I

1000

1100

1

I

1200

compared with the weight decomposition reaction 4 residue + MXzf,,

TOC

loss

calculated

according

to the (1)

This comparison shows that the decomposition is generally incomplete, with the exception of CuCl, CuBr and CdCl. Examination of the residue shows that it is covered by B203 on the bottom of the Pt crucible. This oxide may hinder the diffusion of gaseous MX, and prevent total decomposition of the boraci le. Another reason for the difference between the experimental and the calculated weight loss may be the decomposition of MX, at elevated we have not detected metals in the residue, but we often temperature;

.I

#I

ylyl,

,,,.

Muller [9] Levasseur [lo] This work

,,,

j,

,,,

Vedenkina et al. 1141

,y

_,

15-20

1.25 4

,

10 10

20 20

TG TG Knudsen cell+ mass spectrometry TG TG TG-DTA

Air (1 atm) Air Pure argon

Vacuum (8 x lo- ‘Otorr?, units omitted) at room temperature Air Vacuum (10S3 tort)

2

Evolved gas analysis by mass spectrometty (EGA)

83

Static

Static

Dry air

10

Delfino and Gentile [ 131

40

Dry an

10

4-8 (single crystal chips) 20-44 (single crystal chrps) 2-4

TG

Gallagher [ 121

DTA

Flow rate (cm3min-‘)

Atmosphere

Heating rate (deg mm-‘)

Weight of sample (mg)

Method

Author

Experimental conditrons of decompositron applied by drfferent authors

TABLE 4

observed an attack of the crucible which may have resulted from the formation of an alloy with platinum. This would be consistent with the fact that metal vapour was detected by mass spectrometry (Table 1) [ 12,141 during decomposition of certain boracites. The experimental conditions used by different authors to study the decomposition of boracites are indicated in Table 4.

CGNCLUSIQN

The results of the TG and DTA study of the decompcsition of boracites, in conJunction with the results obtained by other workers by means of different methods (EGA, X-ray analysis), allow the conditions of synthesis of these compounds to be chosen more Judiciously. No boracite was found to be stable in the molten stat,: in an open crucible, but the exact temperature corresponding to 1% weight loss (‘beginning of decomposition’) was found to be a very useful parameter for the upper temperature limit of crystal growth. Further detailed studies with a view to elucidating the nature of the double peaks are in progress.

ACKNOWLEDGEMENTS

The authors wish to thank Miss H. Lartigue for performing the thermoanalysis runs, Prof. J.C. Joubert (Grenoble) for providing the MgOH sample synthesized by 5. Muller. and the Fonds National Suisse de la Recherche Scientifique for financial support.

REFERENCES 1 Landolt-Bomstein, Numerical Data and Functional Relationshtps in Sctence and Technology, Vol. 16, Spnnger, Berlin, 1981, pp. 41-42, 261-284, 597-614. 2 R.J. Nelmes, J. Phys. Chem., Solid State Chem., 7 (1974) 3840. 3 T.A. Bither and H.S. Young, J. Solid State Chem.. 10 (1974) 302. 4 C. Fouassier, A. Levasseur and J.C. Joubert, 2. Anorg. Allg. Chem., 375 (1970) 202. 5 A. Levasseur, B. Rouby and C. Fouassier, C.R. Acad. Sci., Ser. C, 277 (1973) 421. 6 W. Jeltschlco, T.A. Bither and P.E. Bierstedt, Acta Crystallogr., Sect. B, 33 (1977) 2767. 7 J M. Rtau, A. Levasseur, G. Magniez, B. Cal&s, C. Fouassier and P. Hagenmuller, Mater. Res. Bull., 11 (1976) 1087. 8 H. S&mid, Int. J. Magn., 4 (1973) 337. 9 J. Muller, These, Universite de Grenoble, 1970, J.C. Joubert, J. Muller, M. Pemet and B. Ferrand, Bull. Sot. Fr. Mineral. Cristallogr., 95 (1972) 68. 10 A. Levasseur, Th&se, Universite de Bordeaux I, 1973. 11 T.A. Rravchuk and Y.D. Lazebnik, Zh. Neorg. Khim. 12 (1967) 44 [Russ. J. Inorg. Chem., 12 (1967) 211

369

12 PK. c&ag!vx Ttrermactiim. fwa, 23 (ZS?3) 565. 13 M. Dclfinc and P.S. Gentile, Therm~~hi~. Acta, 40 {198Q) 333. 14 L.G. Vedenlcina, A.V. Steblevskii, A.S. Alixanijan, V.I. Bugakov, V.P. Orlovskil and V.I. Gororalci, Izv. Akad. Nauk. S.S.S.R, Neorg. Mater., 16 ( 1980) 1301. 15 H. S&mid, J. Phys. Chem. Salids, 26 (1965) 973. 16 H. S&mid, and H. Tippmann, J. Cryst. Growth, 4 (1979) 723. 17 W. Depmeler, H. Schmid, B.I. Nolting and M.W. Richardson, J. Cryst. Growth, 46 (1979) 718.