INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY
1734
TABLE IX. CL~SSIFICATIOX OF P L ~ S T I C I ZBCCORDI\G ERS TO TEMPERATURE R E T R k C r I O V A U D C O X P R E S b I O S RET DATi
P1astic:zer G R O U PI Diiso-octyl succinate Tri(2-ethylhexyl) phosphate-CaliRux G P Thiol ester G a o v I1 ~ Tri(2-ethyIhexyl) phosphate Dibutyl phthalate Di-n-ootyl phthalate Dibutoxyethyl adipate G R O L P111 Diisobutvl aaelate
1 4 t o 20
67 to 69
2 2 t o 28
22
23
75 to 79
33 to 45
28 t o 30
81 to 86
50 t o 5 5
10
great'est tendency to crystallize and the others mere intcrmediate. The compression set data \yere in general agreement with the temperature retraction data, in that masterbatches exhibiting high compression set also had relat,ively large differences between the TR-10 and TR-70 values, although the masterbatches could not, be lined up strictly in the same order of excellence in all the tests. The vulcanizat,e with diiso-octyl succinate had the lon-ePt set a t both -35' and -65 * F.and thatwithphosphate-oilmixture gave the nest best, performance. The masterbatches can be divided into t'hree groups (Table I S ) according to t,he temperature retraction and compression set teste. I n only t'no cases are there conflicts bctm-een the ratings arrived at by the three test procedures. The temperature retraction test' ~ o u l drank the thiol ester in Group I, ~ C e r e a saccording to the compreesion set data it should be in Group 11. The diisobutyl azelate should
Vol. 46, No. 8
be in Croup I11 according to cornpression set data a t -65" F. and the temperature retraction test data, but, in Croup I1 xcording t o compression set data a t 4 5 ' F. A comparison of the lo^ temperat,ure performance x i t h the calculiit,ed plast'icizer content as shon-n in Table V failed to shorn any correlation between the t v o . Tile variations betn-wn the low temperature performance of the masterbatches are greatel, t,han and often in the opposite direction t o \That, n-ould kit. expected from the variation in the plast No one plast,icizer out, of the twelve tested made a compound superior to the others in t,ensile strength, low temperature fiexibility, and freedom from Crystallization tendencies at low tenipcratures. The masterbatches having the best balance of low t,emperature properties appear to be those cont,aining diieooctyl succinate, tri(2-ethylhexylj phosphate, or dibutoxgethyl adipate. The maoterbat,ches offering the best compromise between good tensile properties at 212' F. and good low temperature performance are probably t,hoae containing dibutyl pht,hslate, dibutoxyethyl adipate, or T5 :25 tri(2-ethylhexylj phosphate-Califlux GP. LITERATURE CITED
(1) Am. SOC. Testing Materials, Philadelphia, ASTX D 1053-5:T. ( 2 ) Ibid., ASTRI D 1229-52T. ( 3 ) G e h m a n , S. D., Woodford, D. E., a n d Wilkinson, C . S., IXD. EXG.CHEM.,39, 1108 (1947). (4) ,Ju\*e. R.D., and Marsh. J. IT., Ihid., 41,2535 (1949). ( 5 ) RIorris, R. E., and H o l l i d e r . J. TI7.. Rubber Age (-V.l',), 70, 195 (1F)RlI. ~(6) Morris, R. E., Hollister, J . JV., a n d Shew, F. L., ISD. CXG.CHEIII,, 43, 2497 (1961). ( 7 ) Pollack, 11.A , , Irzdia Rltbber T o d d . 127, 497 (1953). ( 8 ) Smith. 0 . H.. et a l . , Anal. C'hem., 23, 322 (1951). \
I
RECE:VED for review January 2 5 , 1951. ACCEPTEUM a y 1, 1954. 1Yor;c peiforrned as a part of the research project sponsored by the Reconstruction Finance Corp., Ofiice of Synthetic Rubber, i n connection x i t h t h e government synthetic rubber program.
System Calcium Peroxide-Calcium Oxide-Oxygen SIR: The peroxidation of lime to calcium peroxide T o d d be of interest as the basis for a method for the commercial manufacture of hydrogen peroxide if it could be caused t o o c c u ~readily and in reasonable yield. The purpose of this communication is to analyze briefly the previous data concerning this reaction and to present some nev experimental studies. The reaction is represent,ed by Equation 1. 2Ca0 i- 02
--f
Cn03
(1)
The system has been regarded as univariant, probably by analogy t o barium peroxide-barium oxide and strontium peroxidestrontium oxide, in which systems the equilibrium pressure has been found t o be independent of the gross composition of the solid present and dependent only on temperature. Two contradictory reports of the equilibyium of Reaction 1 appear in t,he literat,ure, one by Bergius ( 1 ) and the other by Blument,hal ( d ) . However, both sets of data have appeared highly questionable since calcium peroxide has been found t,o deconipoee under conditions t'hat, these data indicate should be stable, P R E V I O U S INT'ESTIGATIONS. Three kinds of information are available which give clues to the equilibrium relationships.
1. Oxygen pressures measured above the system under what were believed to be equilibrium conditions 2. Conditions under which calcium peroxide decomposes 3. Conditions under which calcium peroxide has been formed from calcium oxide and oxygen.
The results of all previous and present st,udies are summarized in Figure 1. Bergius ( 1j dissolved calcium peroxide and lime in a eutectic mixture of sodium and pot,assium hydroxide and placed it in a bomb. Equilibrium oxygen pressures were reported over a considerable temperatmurerange for each charge studied, and the liquid phase was also analyzed for peroxide content. (Bergius' results have been erroneously quoted in the International Critical Tables. S'alues were actually reported by Bergius in atmospheres but were incorrectly assumed in the compilation of the Tables to be expressed in millimeters of mercury. Consequently, they were incorrectly divided by 760 to convert to atmospheres.) The validity of Bergius' results appears questionable from several viewpoints. Applicaticin of the gas law to the results of a typical run shows that the moles of gaseous oxygen present remained essentially constant as the temperat,ure and pressure r e r e varied. The pressure changes reported n-ere due primarily t,o expansion and contraction of the oxygen rather than to the
August 1954
INDUSTRIAL AND ENGINEERING CHEMISTRY
chemical reaction. Fischer and Plotze (5) also report that under the conditions of Bergius' experiment there is a 30% conversion of potassium hydroxide to potassium peroxide, which would indicate that the peroxide which Bergius reported to be formed and which he believed to be calcium peroxide was probably actually potassium peroxide. Blumenthal ( 2 ) placed samples of lime and calcium peroxide in a closed container, and the equilibrium oxygen pressure was measured a t each of several temperatures. The length of time that was allowed for equilibrium to be reached was not given, but it appears to be of the order of minutes or, a t the most, of a few hours. I n a preliminary experiment by the present authors a sample of calcium peroxide was placed in a leak-tight reactor a t 143' C., and the variation of pressure with time was observed. Even after 69 hours, the pressure was still slowly rising and the final pressure attained after 333 hours was still far below the equilibrium pressure indicated by later experiments. Thus it appears highly doubtful that equilibrium was actually attained in Blumenthal's work. Riesenfeld and Nottebohm (6) found varying degrees of calcium peroxide decomposition a t each of a number of combinations of temperature and pressure between 238' and 305' C. and 118 and 190 atmospheres pressure. Four attempts were made to form calcium peroxide from a calcium oxide produced by decomposition of calcium peroxide but without success. From the present results, the conditions a t the two higher temperatures studied appear to be unfavorable for peroxide formation. Their tests a t the two lower temperatures were for 15 and 20 hours, which may have been insufficient to allow significant peroxide formation, More recently, Hart and Smith (6)prepared samples of calcium oxide from various sources such as calcium carbonate, calcium peroxide, and calcium hydroxide and subjected them in a high pressure bomb to oxygen at various pressures and temperatures and for times ranging from 2 to 30 hours. I n no case was any calcium peroxide formed. I n each run, a sample of calcium peroxide was also subjected to the same experimental conditions and for each point shown, from 10 to 100% of the calcium peroxide decomposed. Except for one point, an extension of the Blumenthal data would predict these to be conditions under which calcium peroxide should he stable. The actual formation of calcium peroxide from calcium oxide and oxygen has only been reported once other than in the present work. Blumenthal stated that a sample of calcium oxide prepared by decomposition of calcium peroxide and retaining 0.12 to 0.13% active oxygen yielded 1.65% active oxygen after heating for 3 days a t 90' C. and atmospheric pressure. A sample of calcium oxide prepared by calcination of calcium nitrate gave negative results under similar conditions. A second sample yielded 1.53% calcium peroxide after being heated in 1atmosphere of oxvgen a t 75" and 40' C. The presence of peroxide was determined in these cases bv titration with permanganate. EXPERIMENTAL
The reaction rates are so slow at the temperatures involved here that the time required to reach equilibrium is long and there is considerable uncertainty as t o when, or if, equilibrium has been reached. For the present studies, a series of tests m r e made a t each of four different oxygen pressures, 1525, 1000, 115, and 14.7 pounds per square inch absolute. In each series, samples of calcium peroxide of known initial composition were subjected to the fixed oxygen pressure for a fixed length of time but a t different temperatures. Each sample was then analyzed and the composition determined. The dissociation temperature at the oxygen pressure studied was then taken to be the lowest temperature a t which some decomposition was observed. B,y making the time of each test sufficiently long, in principle this point can be determined with considerable accuracy. In practice, the time required may be so long that the method becomes unsuitable. The equipment con8isted of a high pressure reaction chamber, equipped with a pressure gage, to which oxygen was passed from a cylinder equipped with an oxygen pressure regulator. The
1735
reactor was held in a horizontal position in a low temperature oven, and samples of calcium peroxide were placed in the bomb in platinum boats. Temperatures were measured within 1' C. by a calibrated thermocouple on the outside of the bomb. Oxygen from a commercial cylinder, reported to be not less than 99.5% pure, was passed through soda lime and phosphorus pentoxide, in succession, to remove any carbon dioxide or water which might be present before introduction to the reactor. h boat containing phosphorus pentoxide v a s also placed in the reaction vessel.
All samples of calcium peroxide studied were previously dried in an oxygen atmosphere at room temperature over phosphorus pentoxide for a minimum duration of a month. Previous work has established that this will remove all water of hydration which may be present. One sample studied was a commercial product which had approximately the following analysis before drying; CaOz, 55 to 60%; CaO, 15 to 20%: CaCO,, 7 to 8%; Ca(OH)*, 12 to 13%; HzO, 1 to 1.5%. Other purer samples of calcium peroxide were prepared by mixing 90% aqueous hydrogen peroxide with a saturated solution of calcium chloride and adding ammonium chloride to precipitate calcium peroxide. The product was washed and then dried over phosphorus pentoxide a t room temperature for a minimum of one month. After drying, the solid contained 90 t o 95% calcium peroxide. The remainder was assumed to be calcium hydroxide. In each run the bomb was first flushed with oxygen and, after the run was completed, the bomb was allowed to cool before releasing the oxygen. The sample of calcium peroxide was analyzed immediately after removal from the reaction vessel by mixing with water, dissolving the slurry by addition of dilute hydrochloric acid, and titration with 0.1N potassium permanganate solution. Blanks were made with equal samples of lime. RESULTS
Studies were made a t oxygen pressures of 1525, 1000, and 115 pounds per square inch absolute with the commercial grade calcium peroxide and a t 1000, 115, and 14.7 pounds per square inch absolute with the higher purity calcium peroxide KO significant difference was found between the results with the two grades of material. The duration of the experiments was 24 hours a t 1525 and 1000 pounds per square inch absolute, 2 days a t 115 pounds per square inch absolute, and 2 Reeks a t 14.7 pounds per square inch absolute. At the two higher pressures studied, the results showed an abrupt initiation of substantial calcium peroxide decomposition as the temperature was raised in successive experiments over a few degree interval. For example, during a 24-hour interval at 1525 pounds per square inch absolute no decomposition occurred at 110' C. or below, 35% occurred a t 113' C. and 97% or more at 123' C. or higher. Likewise, a t 1000 pounds per square inch absolute, during a 24-hour period no decomposition occurred at 105.5' C., but 17% decomposed at 107' C. and 50% a t 109" C. If the system were truly univariant, these data would indicate that the equilibrium temperature a t 1525 pounds per square inch absolute is about 111' C., or slightly lower if the time allowed were insufficient for appreciable reaction t o occur. Correspondingly the equilibrium temperature a t 1000 pounds per square inch absolute is about 106" C. These points are shown on Figure 1. At the two lower pressures studied, no such abrupt decomposition was initiated over a short temperature range. Under these conditions the reaction was probably so slow that it is doubtful that the minimum temperature a t which decomposition was observed in each case truly represented an equilibrium temperature. The formation of calcium peroxide from calcium oxide was attempted under two different experimental conditions using the same equipment. A sample of reagent grade calcium oxide was found to contain 0.3% calcium peroxide, by solution in acid followed by titration with standardized potassium permanganate
INDUSTRIAL AND ENGINEERING CHEMISTRY
1736
t i
1
w 3 v)
w
L
Figure 1. Comparison of Previous a n d Present Studies on Apparent Equilibrium i n S y s t e m Calcium Peroxide-
Calcium Oxide-Oxygen
solution, after being held a t 115 pounds per square inch absolute and a temperature of $7" to 89" C. for a period of 28 days. Several titrations were made of the final product, as well as of the starting material, in order to establish definitely the presence of peroxide. Three other samples of calcium oxide were prepared by calcination of caicium carbonate, calcium acetate, and calcium peroxide, respectively. After 6 days a t 735 pounds per square inch absolute and 100" C., each of these samples contained from 0 14 to 0.17% calciuni peroxide. These points are shown on Figure 1. A check on the accuracy of equilibrium values may be provided by the van't Hoff equation
For the case here the equilibrium constant of the reaction, in the form written in Equation 1, is simply the oxygen pressure. If the enthalpy change on reaction is constant over the temperature range studied, a plot of oxygen pressure x r s u s the reciprocal of the absolute temperature should be a straight line, the slope of which equals A H I R . Two values of the enthalpy of formation of calcium peroxide have been reported in the literature. T h a t of de Forcrand ( 4 ) is -10.86 kcal. for the reaction as written in Equation 1. Vedeneev, Kazarnovskaya, and Kazarnovskil ( 7 ) recently reported the enthalpy of formation of calcium peroxide to be -155.77 kcal. per mole, which combined with the accepted value of - 151.9 kcal. per mole for calcium oxide, gives -7.8 kcal. as the enthalpy change of Reaction 1, a reasonably fair check Kith the de Forcrand value. The equilibrium line calculated from the van't
Vol. 46, No. 8
Hoff equation using the more recent figure and drawn through the equilibrium points at 1000 and 1525 pounds per square inch absolut,e is s h o m in Figure 1. These conclusions from calcium peroxide decomposition studies must be reconciled wit>hthe fact. that small amounts of calcium peroxide have been formed from calcium oside and oxygen by both Blumenthal and in the present work under conditions in which calcium peroxide would be expected to be unstable if the system were strictly univariant. Any error in the present equilibrium points caused by not allowing sufficient time for equilibrium t o be att,ained would shift the curve t o the right and accentuate the anomaly. One possible interpretat,ion is that the small quantities of calcium peroxide formed might be attributed to an interfacial reaction of oxygen with surface molecules whose activity n-as far different from those comprisiug the bulk of the mass. Alternately, a solid solution may be formed when either calcium oxide or calcium peroxide is present in the solid phase in small concentration, although the system might still be univariant over most of the solid phase composition range. These possibilities are supported by the results of another test in which a sample of the high purity calcium peroxide containing initially 91.6% &On was held at 1000 pounds per square inch absolute and 103.5' C. for 7 days instead of the 24 hours used in other studies at this pressure. The equilibrium temperature has been t'alren as 106" C. for this pressure. At the end of this time, 1.77, of the peroxide initially present had decomposed. Another source of difficulty may he associated with the fact that some quantit.y of calcium hydroxide is always present in any calcium peroxide studied, whereas, none is present when it is attempted to approach the equilibrium from t,he other side by reaction of calcium oxide and oxygen. The presence of traces of water vapor can cause significant decomposition of calcium peroxide. For example, a sample of calcium peroxide supported in a desiccator over phosphorus pentoxide at, 40" C. lost, 5y0 of ita peroxygen content when oxygen from a commercial cylinder was slowly passed over it for 9 days. On repeating the experiment but with previous drying of the inlet oxygen with phosphorus pentoxide, no decomposition occurred. The equilibrium characterist,ics of the system are still in doubt, particularly since no substantial amounts of calcium peroxide have yet been formed from calcium oxide and oxygen. However, from the practical point of vievv, the results of t,he present st,udy definitely indicat'e that the equilibrium relationships in this system are much less favorable for the peroxidation of calciuni oxide than had been thought, previouely. ACKNOWLEDG.\IEST
Financial support was received from the C . S. Office of Kava1 Research under Contract, S o . N5ori-07819: 4R-092-008. The experimental results of Hart and Smith ITere o!,tained in work carried out a t a British =Idmiralty experimental establishment and are published by permission of .i. B. Hart and the .$dmiralty. LITERATURE CITED
(I)
(2) (3) (4) (6) (6)
Bergius, F.,"Sernst Festschrift," p. 68-85, Halle, Wilhelm Knapp, 1912. Blumenthal, hl., Ruczniki Chem., 12, 232 (1932). Fischer, F.,and Plotze. H., Z. unorg. Chem., 75, 30 (1912). Forcrand, R.de, Compt. rend., 130, 1388 (1900). Hart, A. B..and Smith, W. A , , private communication (1953). Riesenfeld, E. H., and h-ottebohm, mi., 2 . anorg. Chem., 90, 371 (1 914-15).
(7) Vedeneev. A. V., Kaaarnorskaya, L. I.. and Kazarnovski:, Zhur. Fie. Khim., 26, 1808 (1952).
CHARLES N. THEODORE
QATTERFIELD
w.STEIN
hlaEsachusetts Institute of Technology Cambridge, Mass. November 30, 1953
For further information, circle numbers 59 A-1, 59 A-2 on Readers' Service Card, page 125 A
-+-
