3846
Vol. 76
The Phototropy of the Alkaline Earth Titanates] and calcium oxide were each ground in a ball mill with equivalent quantities of titanium dioxide, and BY 1 5 7 11 ~MACSEWS ~ ~ ~AXD ~P. R. ~ OGLE* ~ ~ ignited a t 1200' for five hours. Barium titanate RECEIVED ~ I A R C22, H 1964 was a white powder. Calcium titanate was This note reports obseryations on the phenome- slightly violet on the surface when first examined in non of phototropy in alkaline earth titanates. daylight. Impurities of salts of FelI1, Znrl, SblI1, ChaUrley3reviewed the general subject of phototropy Sbv, Vv, Ag', CUI', SnIV, Zr*V and CoI1 ranging in 1929. Tanaka4has reported the development of from 0.08 to 0.5% of metal were dissolved in water a red color on exposure of calcium titanate to light or dilute hydrochloric acid. Slurries prepared and stated that both oxygen and moisture were with solutions of each impurity and samples of necessary. He also found that temperature affected the synthesized barium and calcium titanates were the speed of reversal. Weyl and Forlands concluded evaporated to dryness, ground and ignited a t 1000" from a study of titanium dioxide that when light for three hours. Impure and pure samples of the strikes an impurity ion, for example, FeTII in a ti- synthesized titanates were then exposed to ultratanium dioxide lattice, an electron of the for- violet light. Among the pure titanates only caleign ion is excited and the electron moves either cium showed a visible trace of violet color. Several into an oxygen vacancy of the defective rutile of the impure titanates of barium and calcium esstructure thus producing an FeTVion or the elec- Hibited phototropic change. Table I summarizes tron attaches itself to a TiIV ion to give colored the observations. I n general the effect is more inTi"'. LVeyl and Johnson6 reached similar con- tense with calcium than with barium titanate. The clusions. Preliminary experiments showed that effects do tend to parallel in the two titanates. random samples of barium, calcium and strontium Iron (FelI1), zinc (ZnII), antimony (SbV)and vanatitanates exhibited the phenomenon strongly, dium (Vv) gave the most pronounced effects whereas while it did not appear a t all in magnesium titanate. silver (Ag'), copper (Cu"), antimony (Sb"'), tin zirconium ( Z P ) and cobalt (Co'Ij produced I t was of interest therefore to consider possible (SnTV), no detectable color in barium titanate and no detectreasons for this difference in response to light. A spectrographic analysis of the barium, calcium TABLE I and strontium titanates showed the presence of imINDUCTION O F P H O T O T R O P Y I N BARIUM AND C A L C I U M purities of iron, silver, antimony and copper. In TITANATES BY I M P U R I T I E S magnesium titanate, iron was the only impurity C a T i O1 I m p u r i t y , yo BaTiOa detected. An attempt was made to incorporate None Sone None added impurities of silver, antimony and copper into magnesium titanate by the method of William~on,~ None Sone Fe"' 0.02 which consists of evaporating a slurry of magneSone .06 Violet sium titanate containing soluble salts of the desired Dark violet T'iolet .08 impurities equivalent to 0.2570 of silver, antimony Dark violet Violet .16 and copper. The dry mixtures were ground and S o n e Sone .02 Ag' ignited a t 1000' for three hours. Upon exposure None Sone 1.0 to ultraviolet light, no color development could be Sone detected. Magnesium titanate also was prepared 0.02 None CUI1 by the method of Tanakaa in the presence of each of None None .22 these metals. No phototropy could be detected. Sone None .02 Sb"' These results indicate the probable basic dependNone Sone .26 ence of the phenomenon upon the different specific Sone None Sn Iv I02 crystal structures of barium, calcium and stronSone Sone .30 tium titanates (octahedral) and the crystal structure of magnesium titanate (close-packed cubic) Sone Violet .02 Zn" rather than primarily upon the presence of impurity Dark violet .20 Gray metals. Gray Dark violet .30 The necessity for impurities for the phenomenon Sone .10 Mg" was further investigated by synthesizing barium h-one .20 and calcium titanates with the incorporation of impurities during preparation. Barium carbonate Violet .10 Gray Sb' (1) T h i s research was supported i n p a r t by a contract between t h e Charles F . Kettering Foundation of D a y t o n , Ohio, and The Ohio S t a t e University Research Foundation. (2) Research Fellow, 1951-1952. (3) L. Chalkley, Chem. Reus., 6, 217 (1929). (4) Y. T a n a k a , Bull. Chem. Sac. J a p a n , 16, 458 (1941). ( 5 ) W . A. Weyl and T. Forland, I n d . Bng. Chnn., 4'2, 263 (1880). (6) W . A . Weyl a n d G. Johnson, J. A m . Cemm. Sot., 82,898 (1849). (7) W. 0. Williecnson, Notrave, M S , 279 (1088). ( 8 ) Y . T a n a k a , Bull. Chcm. Soc. J a p a n , 16, 428 (1841).
vv
.20
Gray
Violet
.10
Gray Gray
Dark violet Dark violet
h-one None
None None
None None
None None
.20
ZrTV
.10 .20
co"
.10 .20
3847
July 20, 1954
NOTES
able increase in the slight coloration of the pure calcium titanate. These results indicate the necessity for impurities, for specific impurities and that the phototropic effect increases with the concentration of the impurity ion. Reflectance spectra were obtained with dried slurries of the impure barium and calcium titanates evaporated on glass plates in darkness. Duplicate preparations of each were made and placed in the diffuse reflectance of attachment of the DU spectrophotometer. One plate was covered while the other was exposed a t a distance of 4.5 inches from the ultraviolet radiation Source. Figure 1 shows the reflectance spectra a t 400 to 700 mp for barium and calcium titanates after 5-minute exposures to ultraviolet.
Samples of barium and calcium titanates in equilibrium with atmospheric moisture were also exposed simultaneously and a t various temperatures. Dry and moist samples responded in exactly the same way contrary to the report of Tanaka. Similar experiments in the presence and absence of oxygen showed no difference in color response due to oxygen. I
l
00
40 60 80 100 120 MINUTES OF ULTRAVIOLET LIGHT EXPOSURE.
20
I40
Fig. 2.-Calcium titanate extinction a t 580 mfi with varving time of ultraviolet light exposure.
0020
&/ I
400
L
I
500
600
J
700
W A V E LENGTH IN MILLIMICRONS.
Fig. 1.-Reflectance
spectra for barium and calcium titanates.
The effect of exposure time on the intensity of reflected light was determined by measurements of extinction a t 550 m p after different periods of exposure. Figure 2 shows that a t least a one-halfhour exposure is needed for the extinction for calcium titanate to become constant. Fading of the color produced on exposure to ultraviolet was also studied by measurement of the extinction curve for calcium titanate which had been exposed for 30 minutes and then stored in the dark a t 25’. The reversal temperature of 220’ for the color change reported by Tanaka for calcium titanate was found to be somewhat higher with calcium titanate used in this work. The violet color, developed in calcium titanate on exposure to ultraviolet, was retained until a temperature between 360 and 400’ was reached. With barium titanate the color disappeared between 220 and 270”. When the samples were heated during exposure to ultraviolet light, a temperature between 480 and 610’ was needed to prevent entirely the development of color in both barium and calcium titanates. These experiments on the development and discharge of color were repeated 25 times with no apparent change in the response of the samples. Samples of barium and calcium titanates were sealed in quartz tubes containing phosphorus pentoxide for 7 2 hours and then exposed to ultraviolet.
These results indicate that the occurrence of phototropy in barium and calcium titanate and not in magnesium titanate is basically dependent upon the existence of the proper crystal structure in barium and calcium titanates, which is different from that of magnesium titanate. They also indicate that an impurity must be present such as Fell’, ZnII, Vv or SbV, and that the phototropic effect increases with the amount of impurity. Calcium titanate has a structure in which CaI1 and 0” ions together form a close-packed lattice. The small TiIV ion is surrounded octahedrally by six 0 I 1 ions. However, in magnesium titanate, only the 0” ions form a close packed configuration and the MgI1 and Ti” ions exist in the interstices. Ferrous titanate is isomorphous with Fez03 and magnesium is known to replace iron.g Thus an iron impurity added to MgTiOa can replace either the MglI or TiIV ion without producing an unsymmetrical field. ZincII, similar in size to MgII and TiIV may replace one or both ions without strain on the close packed 0” structure. These results also indicate that effective impurities in inducing phototropy have ionic radii near that of TiIV. Ions larger than Ti1”, for example Ag‘ and CuII which cannot fit even a distorted lattice, do not induce phototropy. It is also apparent that in order for an impurity ion to induce phototropy, it must have a radius approximately that of TiIVbut not exactly the same, otherwise no distortion would occur. The impurity ion must also have a valence other than 4 in order that electron transfer be possible. The greater intensity of color development in calcium than in barium titanate may be due to slight reduction of TiIV to colored Ti111 during the preparation. Not only does the pure calcium titanate show a slight trace of color but upon exposure to ultraviolet, the color produced is more intense. This behavior suggests a “memory” factor which may be retained by the once reduced titanium ions (D) A. F. Wells, “Structural Inorganic Chemigtry,” Oxford tlniversitr Press, London, 1945, p. 932.
3848
NOTES
as a force field associated with this state. Similar phenomena have been observed by Feitknecht and Haberle l o
Vol. 76
potassium compound show it to he similar to sodium borohydride in many respects.
Experimental Materials.-Commercial sodium borohydride from Metal ?\ICPHCRSOS CHEMICAL LABORATORY Hydrides, Incorporated, was used for the preparation of poT H E OEilO STATE UVIVERSITY tassium borohydride. For the preparation of the rubidium COI.I-Nl3L-5, OHIO and cesium compounds, it was purified by the method of Davis, Mason and Stegeman.3 Rubidium and cesium metals were obtained from the Fairmount Chemical ComPotassium, Rubidium and Cesium Borohydrides pany. Organic compounds used in the reduction studies were obtained from the Eastman Kodak Company. P i 21 1 k i ( cx 4s BAYUS,ROBERT LV. BRACDOU A V D ALFRED Preparation of Potassium Borohydride.-A solution of A. HIYCKLEY 74.2 g. of 94.7% sodium borohydride in 130 cc. of distilled mater was added t o a stirred solution of 133.2 g. of potaqFEBRUARY 17, 1954 RECEIVED sium hydroxide (85% assay) in 110 cc. of water. The reThe relatively small ionic potentials (= charge/ sulting thick slurry was filtered through a medium-porosity, glass disc under nitrogen pressure. The precipirntllui) of the potassium, rubidium and cesium sintered tate was washed with three 100-cc. portion5 of ice-cold, ions and the consequent low coordinating power of 95% ethanol. The product was then dried in a vacuum the ions suggested th'dt the borohydrides of these oven a t 80' for three hours; 75.5 grams of potassium borometals would not form hydrates. Interest in these hydride was obtained, representing a yield of 75.5%. Anal. Calcd. for KBH4: K , 72.4X; B, 20.06; H, 7.47. compounds was prompted by this consideration, becsuse of the instability of the borohydride ion Found: K , 73.3; E , 19.8; H,7.25. Preparation of Rubidium and Cesium Borohydrides.toward water. Sodium borohydride forms a di- 10.8 grams of rubidium metal was dissolved in 750 Cc. Of hydrate under normal atmospheric conditions and absolute methanol t o give a clear, water-white solution of rubidium methoxide. This solution was concentrated by consequently has a poor shelf-life after exposure. Potassium borohydride has been prepared by distillation of the excess methanol to give a nearly saturated solution; 4.93 grams of 99.0% pure sodium borothe absorption of diborane by potassium tetra- methoxide hydride dissolved in a minimum amount of methanol was methouyborohydride' but none of its properties added t o the stirred methoxide solution. A white precipiwere reported. This is a difficult preparative tate formed immediately, which was isolated by filtration method because of the unusual nature of the chemi- on a medium-porosity, sintered glass disc, follow'ed by two washes with methanol. The product was dried in a stream cals and equipment necessary. This note de- of nitrogen, then in a vacuum oven a t 80' for two hours scribes a simple method for the preparation of po- The combined methanol washes were concentrated bp evaptassium borohydride, and in addition the hitherto oration to precipitate more product. This was filtered a\ unknown rubidium and cesium borohydrides by before, and washed with 10-20 cc. of methanol and dried. combined yield of 12.2 g., or 95.3'% of theory, was ohmetathetical reactions from sodium borohydride.2 A tained. In aqueous or alcoholic media the products are Anal. Calcd. for RbBH4: Rb, 85.20; B, 10.78; 11, easily obtained in a pure form, are non-hygroscopic, 4.02. Found: Rb,84.59; B, 10.74; H,4.00. and as such have good storage stability. InvestiCesium borohydride was prepared by a similar procedure gation of some of the properties and reactions of the in 85% yield. Anal. Calcd. for CsBH4: Cs, 89.95; B, 7.32; H, 2 7:3. 20 Found: c s , 89.44; B, 7.37; H, 2.71. Hygroscopicity of Potassium Borohydride.-A weighed sample of 99.0% potassium borohydride was placed in 3. c hygrostat over a saturated, aqueous solution of ammonium d chloride in contact with an excess of solid ammonium ride. In this manner a constant humidity of 79.5% a t 20 B (13.8 mm.) was maintained. Exposure for 24 hour.. re5c sulted in a weight gain of 0.85%. Further evidence of the % -. good storage stability of potassium compound was found . e when exposure of a weighed sample t o normal laboratory M 10 atmosphere for seven days resulted in no weight gain and no detectable change in purity. .*-r Solubilities of Potassium Borohydride.-Solubility measu urements were carried out by agitation of an excess of the I solid with the solvent in a warm water-bath, followed by ?c cooling of the mixture t o the desired temperature. A porx . tion of the solution was then removed with a pipet fitted with a glass wool plug t o prevent entry of solids. The solu-1 tion was weighed, the solvent vacuum evaporated, and the 3 weight of the residue determined. Purity was checked i n 7. each caSe by chemical analysis. It was observed that heat was absorbed during solution in water and alcohols. Data for the solubility in aqueous methanol solutions are presented in Fig. 1. Potassium borohydride was observed to CI-IjOH 20 40 60 80 Hz0 be very soluble in liquid ammonia, but no quantitative measurements were made. One hundred grams of 9.5% % He0 by volume. ethanol dissolved 0.25 g. of potassium borohydride a t 2"". Dig. 1.-Solubility of potassium borohydride in methanol, The compound is insoluble (
