U n iversa l O il P r o d u c ts C o m p a n y , R iv e rsid e , III.
T
H E synthetic cracking catalysts are capable of accelerating a variety of hydrocarbon reactions. T he ty p e of reaction depends on the hydrocarbon involved and th e therm odynam ic conditions applied to th e system containing th e hydrocarbon and the catalyst. I t has already been shown th a t th e synthetic cracking catalysts readily sever carbon-carbon bonds in olefins such as octene and hexadecene a t atm ospheric pressure and 350- 500° C. to produce olefins of lower m olecular weight (5). In one sense this cracking of olefins can be regarded as depolymerization. A classical catalyst—th a t is, a substance th a t hastens equilibrium—should be capable of accelerating th e reverse reac
tion, polymerization, if th e ap p ro p riate changes in therm ody
namic conditions are m ade. In th e work reported here, several different synthetic catalysts which h ad been shown to be active cracking catalysts were studied.
Gaseous olefins have already been polymerized by masses con
taining silica and alumina. T he n a tu ra l clay known as floridin or Florida earth has been studied extensively. A pparently Gurvich (8) was the first to observe th a t it would polymerize gaseous olefins. This clay is a hydrosilicate of alum ina containing 1 The four previous articles in this series appeared in J. Am. Chem. Soc.,
«1, 3571 (1939); 66, 1586, 1589, 1694 (1944).
b oth iron and magnesium compounds. T his and num erous simi
lar clays, either w ith or w ithout acid treatm en t, have received considerable scientific (2, 15, 16, S I, 22, 24, 26) and technical (7, 10, 12, 14, 19, 20, 25) study as polym erization catalysts. In general th e polymers resulting were dimers, trim ers, and te- tram ers, although small proportions of still higher polymers m ay have been formed. Isobutylene was especially easy to polymerize.
Floridin is capable of polymerizing isobutylene a t tem peratures as low as —100° C. {15, 22). T he m olecular weight of the iso
butylene polymers was found to increase as th e tem perature was lowered. T he higher-molecular-weight polymers were elastomers.
A pparently G ayer {6) was th e first to report work on a syn
thetic silica-alum ina catalyst for th e polymerization of gaseous olefins. His catalyst was prepared by the hydrolytic adsorption of alum ina on a silica xerogel. T he resulting catalyst contained about 1% alum ina. This catalyst produced m ainly dimers and trim ers under the conditions studied. The same type of catalyst was investigated in considerable detail b y Hoog, Sm ittenberg, and Visser (9) and K azanskil and R osengart {13). R elated synthetic silica-alum ina catalysts have been described for olefin polymerization {1, 4, 6, U , 18).
544 I N D U S T R I A L A N D E N G I N E E R I N G C H t M I S T R T 37, No. 6 polymerizing ability of synthetic silica-alum ina catalysts spe
cially prepared to have a high cracking activity. In addition, synthetic silica-zirconia, silica-alumina-zirconia, and silica- alum ina-thoria cracking catalysts and th e components silica, alum ina, and zirconia were investigated.
T he cause or seat of th e catalytic activity of either n atu ral or th e synthetic catalysts is n o t clear. T he synthetic catalysts are noncrystalline, while th e n a tu ra l clay catalysts are crystalline to x-rays. W hen properly prepared th e synthetic catalysts are more active th an th e n a tu ra l catalysts. T he synthetic catalysts
Condition Fresh Regen. Fresh Fresh Fresh Freeh
Weight, grams 28.7 28.7 38.0 25 .2 35.7 25
<* Prepared from ZrOCli.8HjO ju st as if a c atalyst were being m ade b u t in th e absence of the silica hvdrr><rf>l e This polymer yield was calculated from the C«H« disappearance; in m ost instances the polymer
somewhat smaller due to mechanical losses. y
/ When no polymer was formed a t 165® C. with sirconia, th e tem perature was gradually raised to 300° C
maintaining the charging stock flow. No polymer form ation was observed. wmie
I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 545
& The yield of polymer calculated from the inlet and outlet gases is somewhat higher.
The difference between the calculated polymer and the recovered polymer is presumed to be due to handling losses.
As th e cataly st aged, there was less tendency to form polym ers boiling outside th e gasoline range. T h a t is, there was a greater tendency for th e cataly st to pro
duce dimers ra th e r th an higher polymers.
T he antiknock quality of th e olefinic polymer com
pared favorably w ith th a t produced by other polym eri
zation catalysts.
Prelim inary tests were m ade to see if th e catalysts would act as polymerization catalysts and to seek usable reaction conditions.
These tests showed th a t polym erization took place a t 165° C., a t a pressure of 45 kg. per sq. cm., and a t a feed ra te corresponding to about 7 volumes of liquefied charge per hour per gross volume of catalyst.
These conditions were used for tests on a num ber of catalysts.
Table I is a sum m ary of th e results. From th e experience gained in these tests th e following observations have been m ade:
The silica-alumina, silica-zirconia, silica-alum ina-thoria, and silica-alumina-zirconia masses were polym erization catalysts.
Individually th e silica, alum ina, and zirconia did n ot act as
The method of catalyst p rep aratio n seems to be ju st as im portant as chemical composition in preparing active catalysts. B oth a suitable preparation m ethod and a suitable composition are nec
essary to produce an active polym erization catalyst.
As m ight have been expected from its known activity, iso
butylene polymerized m ore readily th an n-butylenes under these conditions.
P IL O T PLA NT R ESU L T S
Larger scale tests were m ade so th a t th e process and the prod
uct could be studied in m ore detail. T he conditions of tem pera
ture and pressure were nearly th e sam e as those used in the laboratory. The initial feed ra te was 8.7 volumes of liquid charge per hour per gross volume of catalyst, b u t as th e activ ity of the catalyst declined, the feed was reduced so th a t it was 4.5 a t the end of the test. Table I I is a sum m ary of th e results. Liquid product inspections are listed in T able I I I . From th e d a ta in these two tables the following observations have been m ade:
The polymerization occurred on th e p ilo t p la n t scale, very much as would have been expected from th e lab o rato ry results.
The catalyst lost activity w ith use. D uring th e first 24 hours the polymer production rate was ab o u t 2.5 tim es as fast as during the 48-71 hour period.
E xploratory tests were m ade on the pilot p lan t with a propan e-propylene fraction containing 18% propylene.
Raising th e tem perature to 250-300° C. a t 50 kg. per sq. cm. polymerizing propylene. A more detailed stu d y of th e propylene polym erization by these catalysts was n o t m ade because of (he rapid decline in catalyst activ ity owing to th e carbonaceous de
posits on the catalyst. I t was th ought th a t the rapid decline of th e cataly st activ ity under these conditions m ight prevent the catalyst from com peting w ith other polym erization catalysts that are known to be able to polymerize propylene while m aintaining activ ity over m uch longer periods of time.