T H E GROWTH OF SINGLE CRYSTALS OF BARIUM TITANATE FROM MELTS W I T H BARIUM C H L O R I D E S~ANISLAV ~AFRATA, VENDULKA BEDNifCOVi, JOSEF BENE~ Institute o] Physics, Czeehosl. Acad. Sci., Prague, Department o/ Atomistics and Solid State Physics, Charles University, Prague

On the basis o/ experiments carried out on the growing o] barium titanate single crystals (BaTiO3) ~t is shown that the data given by B. T. Matthias [1] are incomplete and insu]]ivient thus leading in certain cases to negative results. This paper gives the conditions /or growing crystals which are suitable/or ]undamental physical research. 1. I N T R O D U C T I O N

A number of methods for the growing of single crystals of barium titanate '(BaTiO3) are found in the literature. The crystals are either grown from their own melts [2, 3, 4] or from solution [1, 5, 6, 7, 8]. Great, differences are sometimes found in these methods. Inspire of this the underlying reasons are not given. This paper investigates Matthias' data on the growing of single crystals of barium titanate [1] in greater detail. 2. E X P E R I M E N T A L V E R I F I C A T I O N OF MATTHIAS' DATA

In an earlier paper [9] it was pointed out that certain deficiencies exist in Matthias'paper oa the growing of single crystals of barium titanate. According to the results of our previous experiments the melting procedure is not sufficiently defined by MAT~H~AS. Facts on the course of the temperature in the neighbourhood of maximum temperature, where the evaporation of the components of the melt m a y have an important effect on the composition, are not given. The shape of the vessel in which growth takes place is likewise omitted which introduces a further uncertainty into t h e s u m total of the melting conditions. During our experiments on the growing of single crystals in open boats the molar ratio between the components of the melt published b y Matthias was shown to be unsuitable. We found another molar ratio (BaCI~ : BaCO3 : :TiO2 -- 2 : 2 : 1 ) which for the conditions described in the paper led to positive results. Crystals were obtained which are suitable for further physical research. However the method described by us has a fundamental deficiency in the same way as Matthias' method. I t is not possible to determine exactly the composition of the melt from which the crystals of barium titanate are produced. During the evaporation of barium chloride other components also escape from the melt thus changing the composition of the melt at different places and in different phases of crystal growth. I n order to guarantee more clearly defined ratios of the components in the melt and their exact determination it was considered necessary to use an experimental arrangement which would ensure to a sufficient extent t h a t the composition of the melt during evaporation would change as little as possible. The experiments were carried out as follows. The mixture of barium chloride (for analysis), barium carbonate (purissimum) and titanium oxide (purum) was melted in a conventional alundum crucible made of the purest A1203. The internal diameter of the base of the crucible was 20 mm, the diameter of the Czechosl. Journ. Phys. 6 (195() 2

185

Stanislav ~a]rata, Vendulka Bednd~ovd, Jose] Beneg widest part at the top was 52 m m and the height of the crucible was 55 mm. I t was covered with a close-fitting alundum lid. The crucibles with the melt were placed in a vertical tubular furnace with a megapyre or platinum winding which enabled temperatures around 1200 ~ in air to be attained easily. This arrangement is shown schematically in Fig. 1. The temperature was measured with a P t - P t R h thermocouple touching the b o t t o m of the crucible in the furnace. The thermocouple gives the temperature in the crucible with an accuracy of a few tens of degrees. In the given arran/_L gement the temperature inside the melt in the crucible , is approximately the same throughout. The voltage ~-- ~ was led to the furnace from a variable transformer conI : ~~ nected to an a - - c stabilizer. The voltage on the furnace and thus also the temperature in the furnace were varied automatically with an auxiliary motor controlling " ~ 4 the variable transformer. I-In the first series of experiments the suitability of the ~ 4 molar ratio of barium carbonate and titanium oxide -II ,j~ ~ 2 : 1 -- was tested. For the various meltings this ratio [[( ~ was adjusted within the limits 2.1 : 1.0 to 2.04: 1.0, i. e. from an exact stoichiometric multiple to the ratio ~ ~= given b y Matthias. The content of barium chloride was varied in the various experiments so t h a t values in the limits 1.92 : 1 to 3.85 : 1, i. e. up to Matthias' 5 ratio, were chosen for the molar ratio of barium chloride and titanium oxide. The melt was heated for several hours to a m axi m um temperature of 1120 ~ to 1290 ~ from which for the various meltings it was cooled Fig. 1. Diagram show- to 800 ~ at a speed of 35 to 160 ~ The melt ing location of cruci- cooled down from 800 ~ to room temperature in about ble in furnace, l.asbes- 12 hours. The cooled melt was dissolved in water b y tos, 2. h e a t i n g winding, 3. crucible with the usual method. In none of these experiments were melt, 4. ceramic tube, barium t i t anat e crystals produced in the melt. 5. thermocouple in Parallel with this another series of experiments was protective sheath, carried out with lower molar ratios of barium carbonate to titanium oxide t han Matthias' ratio of 2.04 : 1. The molar ratio most often used was 1.5 : 1. This series of experiments showed the im por t ant role played by the period of hold-out at the m a x i m u m temperature. With a molar ratio of BaC12 : BaC03 : TiO2 -- 3 : 1.5 : 1, a maximum temperature of 1120 ~ and a cooling rate of 64 ~ no single crystals of barium t i t anat e are produced without a hold-out in the melt. With the given content of barium chloride the melt is not sufficiently saturated with barium titanate and crystals cannot therefore be produced. In order to judge of the efficiency of the melt let us introduce the so-called yield of the melt v by the relation v= 100%, ~th where m is the weight of the crystals of barium titanate larger than 0.1 mm and mt~ is the weight of the crystals which would theoretically be produced in the melt if all the titanium oxide reacted with aa equimolar amount of barium carbonate to produce barium titanate. In addition to the crystals larger t han 186

qexoc~. ~ 8 . mypa. 6 (1956) 2

The Growth o] Single Crystals o/Barium Titanate ]rom Melts with Barium Chloride

0.1 m m a certain a m o u n t of smaller crystals are p r o d u c e d in t h e melt which c a n n o t actually be used in physical research and which are t h e r e f o r e n o t included in the yield o f the melt. These small crystals are p r o d u c e d in the m e l t at a t e m p e r a t u r e n e a r to t h e i r p o i n t o f solidification. A v e r y low yield - - 0.2% was o b t a i n e d for the given melt. These crystals, which are fairly small, axe shown in Fig. 2. I n this figure, as well as in several others, one can see interference fringes which are p r o d u c e d between the surface of the crystal and the u n d e r l y i n g slide.

Fig. 2. Crystals of BaTiO s from a melt BaClz : BaCOa : TiO 2 - 2 : 1.5 : 1, without hold-out.

Fig. 3. Crystals of BaTiOa from a melt BaC12 :BaCOa : : TiO 2 - - 3 : 1.5 : 1, hold-out 4 hours.

The melted m i x t u r e is p r e p a r e d in such a w a y t h a t the initial c o m p o n e n t s are g r o u n d a n d m i x e d in a mortar. The grain size o f the reacting c o m p o n e n t s is of i m p o r t a n c e d u r i n g the f o r m a t i o n of b a r i u m t i t a n a t e in t h e melt. One of the possible mechanisms o f the production, for example, is such t h a t on the surface of a grain of t i t a n i u m oxide, b a r i u m t i t a n a t e is p r o d u c e d which dissolves in the melt a n d thus gradually uncovers f u r t h e r layers of t h e grain o f t i t a n i u m oxide. This process of course takes some t i m e during which the melt m u s t be k e p t at a high t e m p e r a t u r e so t h a t the desired reaction should t a k e place as completely as possible. If, for example, for a melt h a v i n g a composition BaCl~ : BaCon : Ti02 -- 3 : 1.5 : 1 a f o u r - h o u r hold-out at m a x i m u m t e m p e r a t u r e of 1170 ~ is used, single crystals of b a r i u m t i t a n a t e are o b t a i n e d for a cooling rate o f 18 ~ which can be seen in Fig. 3. The yield o f this melt was a b o u t 0.4%. F o r a composition BaCI 2 : BaCO 3 : Ti02 - - 2 : 1.5 : 1 a n d otherwise identical conditions a m u c h greater a m o u n t a n d larger crystals are o b t a i n e d (Fig. 4). The yield of this melt was a b o u t 6%. This yield is also m u c h larger t h a n in the case of melting w i t h o u t hold-out. W i t h a long hold-out of m a n y hours the result is b e t t e r still. W i t h a m e l t composition BaCl~ : BaCO s : TiO~ -- 3 : 1.5 : 1 a n d a 19-hour h o l d - o u t a t m a x i m u m t e m p e r a t u r e of 1200 ~ rate o f cooling 19 ~ crystals were grown a millimetre in size (Fig. 5). This melt also p r o d u c e d t h e greatest q u a n t i t y of crystals of all t h e cases described so far, the yield reaching 16%. W i t h a long hold-out a sufficient q u a n t i t y o f the initial c o m p o n e n t s react so t h a t s a t u r a t i o n occurs in the melt at a t e m p e r a t u r e high e n o u g h above t h e point o f solidification o f the melt. The crystals can t h e n grow for a sufficiently long period. I n this experiment, in addition to t h e n o r m a l m u l t i - d o m a i n Czechosl. Journ. Phys. 6 (1956) 2

187

Stanislav ~a]rata, Vendullca Bednd~'ovd, Joss/Bensa

crystals (Fig. 6), there was also a considerable quantity of crystals formed of a few domains (Fig. 7). These crystals are particularly suitable for physical research. Approximately the same results are obtained by a melt with an initial composition of BaCl= : BaC03 : Ti02 -- 2 : 1.5 : 1, other conditions being identical to those described last. With melts having a composition according to Matthias, we also wished to eliminate insufficient reaction between the components since, as was shown, this considerably effects the results. The melt was therefore kept at a maximum temperature of 1200 ~ for 18 hours. Three cases of the composition BaCI 2 : : B a C O a : T i 0 2 -- 4 : 2 : 1 , 3 : 2 : 1 and 2 : 2 : I were investigated. Ir~ none

Fig. 4. Crystals of BaTiOa from a melt BaCI 2 : BaCO 3 : : TiO 2 -- 2 : 1.5 : I, hold-out 4 hours.

Fig. 5. Crystals of BaTiO 3 from a melt BaCI~ : BaCO :TiO~ -- 3: 1.5: I, hold-out 19 hours.

3 :

of these melts were single crystals of barium titanate produced for a cooling rate of 19 ~ When Matthias' ratio of barium carbonate and titanium oxide was used bluish crystals having the shape of hexagonal prisms and being several millimetres in length sometimes appeared. These crystals were slightly soluble in water. According to X-ray analysis of these crystals carried out by M. STRAXOVA from the Institute of Physics of the Czechoslovak Academy of Sciences, it was found t h a t they were barium orthotitanate (Ba~TiO4). This verified the comparison of the debyegrams of the grown crystals both with an analogous photograph in the literature [I0] and with a debyegram of crystals of barium orthotitanate prepared in our laboratory by another m~thod [ll]. Fig. 8a gives the debyegram of the crystals from a melt having Matthias' composition, while Fig. 8b shows the debyegram of crystals of barium orthotitanate. On varying the ratio of barium carbonate and titanium oxide to a lower content of barium carbonate it was found t h a t with a composition of BaCOn : : TiOa -- 1.7 : 1 no single crystals of barium titanate were produced in the melt.

188

~Iexocat. 0~a. ~ICypH. 6 (1956) 2

The Growth o/Single Crystals o] Barium Titanate ]rom Melts with Barium Chloride 3. CONCLUSIONS

It is n o t p o s s i b l e to grow single crystals o f b a r i u m t i t a n a t e in closed cruc bles from melts h a v i n g the same c o m p o s i t i o n as t h a t used b y M a t t h i a s for growing t h e s e crystals in o p e n crucibles. W i t h i n the limits o f t h e m o l a r ratios BaC12 : B a C Q : TiO~ - - 1.92 : 1.7 : 1 to 3.85 : 2.04 : l, predominantly crystals

of barium orthotitanate are produced. Matthias' method can only be used

Fig. 6. Domain structure of crystals of BaTiO a. Microphotograph in polarized light, crossed Nicols.

Fig. 7. Crystals of BaTiO~ with small n u m b e r of domains. Microphotograpt~ in polarized light, crossed Nicots. Magnification identical to that in Fig. 6.

a)

b)

Fig. 8. Debyegrams: a) crystals from melts composed according to MATTI~AS [1], b) crystals of b a r i u m orthotitanate. Czechosl. Journ. Phys. 6 (1956) 2

189

Stanislav ~a/rata, VenduIlca Bednd~ovd, Jose/Beneg

therefore for growing crystals in open crucibles where noticeable evaporatiort of the components of the melt occurs thus c~using, at least locally, considerable change in this composition. The m e t h o d of B~A~T~EI~, KA~ZlO and MEI~Z w i t h a ratio of BaCl~ : BaCOa : TiO~ - - 3.3 : 1.4 : 1 or o f ELIZABETIt WOOD with ~ ratio of 2.95 : 1.62 : 1 can therefore be used with greater hope o f a positive result w h e n growing in covered crucibles. According to our experiments crystals suitable for physical research were produced from a melt with ~ composition of 2 : 1.5 : 1 to 3 : 1.5 : 1 for a hold-out of 19 hours at a m a x i m u m temperature of 1200 ~ and subsequent cooling at a rate of 19 ~ The growing process used provides a reliable m e t h o d b y w h i c h reproducible results can easily be obtained. l ~ e c e i v e d 12. 5. 1955. BLIPAIIIHBAHHE

MOHOHPHCTAJI~OB

THTAHATA

BAPHH

( C o A e p m a r m e n p e g L t ~ y m e ~ CTaTbI~I) STANISLAV ~AFRATA, VENDULKA BEDN-~f:tOV/I~, JOSEF BENE~

(Pu3uuec~u(* u~cmumym ~tCAH, llpaea Ha96eOpa amoauo~t 96uzu~u u ~usu~u rneepOoeo mena I~ap~oea yuueepcumema, llpaea , B a a T e p a T y p e IIpHBO~I4TCa PAR OT~'IHqaloII4HXC~t ~ p y r OT ~ p y r a ClI0CO6OB m,tloami~BaHi~a IcpHcTazlaOB THTaHaTa 6ap~ta. C p e ~ ~ x BBI~e:I!HIOTC~t ClI0CO6LI, ~ c n o a L a y ~ o m ~ e p a c ~ a a B aeHHylO Maccy x a o p ~ i a 6 a p ~ a . THmI~n~IM ~tB~IaeTCa n p n e M , om~canHLii~ M a T T ~ I a C 0 M . ~TOT CIIOCO5 ~BHJICYI I4CXO~ttt)[M rIyHHTOM HalLII4X OrIt,ITOB, Ii3 t(OTOpl, IX CTaJIO BI4~HO, qTO ~aHHHe M a T T ~ a c a HelIO~IHHe, Tall qTO OH!~ MOryT B oiipe~eaeuH~iX c ~ I y q a u x np~iBeCTl~ ~ OTpI4I~aTe~II~H~IM pe~y~ILTaWaM. M a T T H a C , oqem,i~Ho, B CBOI4X OrI~ITaX loa6oTa~I c OTHp~TI~IMH MHCHaMI4, H3 HOTOpBIX X~IOp~l~ 5 a p ~ a MOF JIe!PH0 I4ClIapi:ITt)(~H. ~MH HaHIHX 0HcFIel0HMeHTOB Mt,I I4ClIOJII,3OBaJIH THFeJII~I I43 oqeHb qHCTOrO Cm~TepHopyH~(a, XOi0OmO I~pHHp],ITHe HphIIL~HaMH, Tall qT0 COCTaBJDIIOIIU~e p a c l i a a B a 3aMeTHO He Hcnap~IaI~Cl~. B IlepBOfi cepI~H 3HclIepHMeHTOB ~icco~e~osazlacb i4e~iecoo6loa3HocTt, MOJI~ipH0ro OTHOmeH~i~ 2 : 1 yrJIeHHcSiOF0 6 a p H a ~ 7~ByOt<~iCI~THTaHa. 1]p~ OT~eJH, HHX ~t~Iannax 0TO OTHOmeHHe MeH,qJIOCl~ OT 2 , 0 0 : 1,00 ~(0 2,04 : 1,00, T. e. OT TOqHOFO CTeXI4OMeT])14qecHoFo MHOH~I4TeJIH gO OTHOLUeHH~t, IIp~iBO~MOrO MaTT~IaC0M. Co;~epmaHIie x a o p H ~ a 6 a p ~ a MeH~iJ!OCb ilpI4 OT~e~IhH~IX -0~;ClIepHMeHTax TaI,;HM o6pa3oM: ~ t MO~I~pHOFO OTtiOIIIeHH~I x~Iop~l~a 5 a p ~ a i4 ~ByOHI4CI~ THTaHa nI,~5~paa~Ic~ a H a q e H n a B ilpe~e~iax OT 1,92 : 1 ;(0 3 , 8 5 : 1, T, e. ~o 0TH o m e H n a IIpHB0~I~iMOFO MaTTHaCOM. Pacn~Ian iiarpeBa~IcK B Te'~eHne HeCHOJIbHHX qaC0B ~O MaI~c~a~ISHOfi TeMrlepaTyp~i, B oSaacTl( OT 1210 ~ ~Io 1290 ~ I~ ~aTeM o x a a m ~ a ~ I c a ~ o TeMnepaTyp],i 8 0 0 ~ cgopocT5~o ]~ 0Tgeabn~,IX n~Iamcax 3 5 - - 1 6 0 ~ H , np~l O~(HOM Ha DTHK aHcIIepHMeHTOB MOHO.~{pHcTa~IZILIMeTaTI/ITaitaTa 6 a p ~ B p a c r l a a n e ~e o5I~apymeHia. Bo BTopoIi c e p m I OI~ClIepHMeHTOB ~iClIO~!~,30BaZIocb M0ZI~ipHoe OTHomeHne yraei~ICaOrO 6 a p ~ m ~ gByom~c~ TI4TaHa 1,5 : 1. IlloI4 TaI~0M OTHOUleHHH I43 pacn~IaBa 6Lia~l s~ipai~eH~r ~p~cTa313!~,I. /{afi~etto cyu~eCTBeHHoe B3Ig~HHe IIp0~oov~FHI~ITeJIlhHOCTH npeMeH~, B TeqeHm~ ~OWOpOrO p a c l I a a B Haxo~i~acg n p ~ Ma~cnMaa~HOA TeMnepaType. q e M ilpo~io~mgwe3i~Hee ttaxoAgwc~ p a c l i a a B n p ~ MaKcI4MaYIbHOITI T e ~ n e p a w y p e , weM 6oJIb~iIe ~ a y ~ m e (iip~ I l p o q g x p a m ~ i x yCZlOB~IaX) Bl,IpalI~eHH~Ie Hp~IcWa.g3ISI. CaM0e 6o3II~uioe HoaggecwBo HpI~CTaJI310B I I O a y u a a o c ~ B p a c n a a B e , c c o ~ e p m a H ~ e M BaCI~ : BaCO~ : TiO~ = 3 : 1,5 : 1; p a c n a a B r p e a c a 19 ~ a c o a Ilp~ TeMnepawype 1200 ~ ~ ox3Iam~azlca co C~r 19 ~ HpoMe O6])IHHOBeHHt,IX MHOFO~OMeHHIblX ]r IIp~I TagOi~ II~IaBge BCTpetta~OTCH TaHHce H O~I-IO~OMeHHa~Ie HplacTaa.~I HaI~ Hp~ICTao~IJIla, opaaoeaHi~,ie a~iml~ He60~blIIHM q~ICJIOM ~OMeH. TaI~Iae i~pgcTaaJI~,i Oc06eHHO llO~XO~(gT ~ I g (~I~3I~I~IecHHx Hcc~egOBaHHI]. B p a c l I a a B e (COCT&B iio MaTTHacy) ~ a m e IlOC~Ie MHOI'0qaCOBOFO Har!0eBaHHg Ilpg MaHCHMOJII~HOITI T e M n e p a T y p e Hp~IcTa3I.~Ib[ MeTaTI4TaHaTa 6apH~t He IIOHBHSIHCG. B p a c n J i a B a x c COCTa]~OM MaTw~4aca I1OHBJIYIJIHCt) C~HeBaT],Ie HIOBcTa~]JIbI , ~ I ~ g O ~ ~0 HeCHO~II~HI4X MHJIJIHMeTpoB, B mI~(e iiiecTHyroa~,H~iX !qpI43M. P e H T r e H o r p a ~ H q e c ~ i i ~ a H a z m a n o ~ a a a a , ~ o 0TO opToT~ITaHaT 6 a p u a (Ba TiOa). OTH I~plac~aaa~i ncvpe,taa~Ic~, T a R m e ~I B pac~i~iaBax c OTHOIIIeHrleM BaCO~ : TiO~ = 1 , 7 : 1 , r ~ e HpHCTa~IJII,I MeTaT~ITaHaTa

190

~Iexoc~. r

mypH. 6 (1956) 2