New Method for Testing Catalysts

J. A. HINCKLEY, JR.1, AND HARRY R. SHEPPARD, JR. Mellon Institute, Pittsburgh, Pa. A new type of catalyst testing system employs a device for continuo...
10 downloads 0 Views 3MB Size
New Method for Testing Catalysts J . A. HINCKLEY, J R . ~ ,AND HARRY R. SHEPPARD, JR. Mellon Institute, Pittsburgh,, P a .

A new type of catalyst testing system employs a device for continuously separating the reaction products from a catalyst testing furnace. These fractions are more readily analyzed than the converter-make collected as a whole. This new system has resulted in improved accuracy in testing catalysts used for making butadiene from ethyl alcohol.

C

ATALYST testing in the laboratory usually consists of determining the change in composition of a feed stock that takes place under given conditions in the presence of the catalyst. The accuracy of the testing depends on the accuracy of the analysis of the reaction mixture for unreacted feed and main product. The main problems involved in the analysis usually ariqe from the presence of the inevitable profusion of by-products caused by the severe conditions imposed upon the feed stock, and from the difficulty of obtaining a proper sample of a mixture containing components ranging from gases to heavy tars. An improvement in accuracy obtained by modifying the usual I method of recovering the reaction products is described here. The modification consists essentially in Orifice Fig1ure 1. Feeder separating the products continuously and precisely, during the test, A . Reservoir B . Valve into two fractions containing conC. Union stituents boiling below and above presD. Nitrogen mure inlet room temperature. head E. Constant device The reaction under consideration F. Filter disk is that of a mixture of ethyl alcohol G . Constant temperature liquid and acetaldehyde to form 1,3-butaexit bore Constant H . I n addition, hydrogen, diene. capillary carbon monoxide and dioxide, light I . Constant temperature liquid hydrocarbons, aldehydes, alcohols, entrance esters, heavy hydrocarbons, oils, etc., are recovered. The reaction product is separated into a butadiene and lighter fraction and an acetaldehyde and heavier fraction. The light hydrocarbon fraction is collected in its entirety in a steel cylinder previously evacuated, from which a representative sample can be drawn for analysis by low-temperature distillation or other means. The acetaldehyde or heavier fraction is recovered in aqueous solutions, which may be analyzed for acetaldehyde and alcohol by distillation and chemical means. The butadiene is contained entirely in the gas fraction and the acetaldehyde and alcohol are entirely in the aqueous solution.

pressure, controlled by a diaphragm valve and mercury bubbler, pushes the feed through the orifice. The feeder is designed so that the weight of material fed in each test can be measured accurately. .1 reservoir, detachable from the metering orifice, holds the feed. The reservoir is weighed before and after each test to measure the amount of material fed into the system. The feeder is designed to prevent holdup of feed within the metering orifice. A constant-head device within the feeder assures constant feeding as the liquid level drops within the reservoir. Reactor. The catalyst is tested in a I-inch stainless steel reactor held at' constant temperature by an electric resistance furnace. The pressure over the catalyst bed is constantly atmospheric. Feed stock is vaporized over a steam coil before entering the reactor. The reaction products are separated immediately after leaving the catalyst. Product Separator. ;In estractive distillation column (Figure 2j, which separates the reaction products, consists of a scrubbing column integral with a fractionating column. The reaction products enter this separator through a glass tubing coil concentric with the scrubbing column. Ice water is circulated in a jacket about the scrubber and coil. Licluid reaction products are condensed and a fog form of entrainment is eliminated in the coil. Scrub Lyith water uncondensed materials in the packed scrubbing section to remove acetaldehyde from the gas. The scrubbing .section is cooled t o 0" C. by the ice water jacket, greatly lowering the vapor pressure of acetaldehyde and thereby increasing efficiency of the scrubber many fold. The scrubbing water picks up a small amount of butadiene in addition to acetaldehyde. This butadiene is conb$CE WATER IN tinuously fractionated out qf solution by means of an integral fractionating column located just below the scrubt+. bing section. An unpacked jacketed section between the fractionating column and the scrubbing column $i 4 serves as a reflux condenser for the fractionating column. The reflux temperature is controlled a t the boiling point of acetaldehyde, assuring complete removal of butadiene from the scrubbing water. The unreacted feed and liquid products are collected in the still pot n i t h the scrubbing Tvater. Scrubbed gas A o ~ sfroni the top of the product separator into the gas-collection system. Figure 2. Product Separator Gas-Collection System. The gas A . Product entrance flom from the product separator and to jacket reB . Entrainment is immediately and continuously tested moval coil for complete acetaldehyde removal. C. Product entrance

-

SPECIAL C4TALYST

$

Ej

Q

TESTING EQUIPMEKT

Feeders. Orifice-type feeders (Figure 1) are used to admit feed stock to the catalyst and water t o the product separator. They continuously supply a well-regulated feed by employing a constant-bore glass capillary as a metering orifice. The temperature of the capillary is held constant by a jacket throughwhich a liquid is circulated a t controlled temperature. Sitrogen gas 1

t o column

D . Product gas exit E . Scrubbing water entrance Fractionating column G . Thermometer H . Still pot 1. Scrubbing column

F.

Present address, Podbielniak, Inc., Chicago, Ill.

771

Acetaldehyde is detected by means of a concurrent scrubber in which the gas is mashed with a water solution of hydr6xylamine hydrochloride containing bromophenol blue indicator.

V O L U M E 19, N O . 10

112

ancc in tla: r:apillwy, however, the level of m ~ r c u r y in tha reservoir is just slight,ly below the entering point of the

!

, i)ressuEe drop in lhe capillary. 11. is readily seen t,hat. a vacuum could be

sm&.

Changcsofthisanrt arc allindi-

V

and elongates t,hc spring in the sealit from which it issuspended. The diametcr ofthcmercui.yi.Pservoir as calculated so that, in vicw of linear clonga, tionof t,hespring,tholcvelof mercuryin thc reservoir docs not chengc with respcct t o the tip d the capillary whek thrgasmtr~athrrcgulator.I n thisway t,hc devicr is cnmpletclv aut.omwt,ir. throughout, nn m t i r r run.

-

CATALYST TESTING IJNIT

Figure 3.

Catalyst 'lesti,ng U n i t

The hydroxylamine hydrochloridr sdut,ioll is placed irr t,he scrub. in the of her at a p~ of 4.6, so that gas will react with the hydroxylamine hydrachloride and cause the of the,solution to from blue-pink to If any acetaldehyde is detected, it can be dctermined directly by titrating it, with the buret, filled with 0.5 N sodium hydroxide attachod to the top of the scrubber. I n this way good operation of the seoaration unit is rtssured. for poor operation is detected and can be corrected immediately. This hydroxylamine scrubber is operated by an air-lift principlc which circulates the gas with the

.

-.

The foed is supplied a t a oonstant raie 1.0 tho unit from an orifice feeder, 4, operated by maintaining a constant different,ial Pressure BC?'Oss t,he feeder orifier hy moans of nitrogen pressure in cylindor B, which is controlled by mercury bubbler C and differential pressure manometer D. The feed is sdmittcd to a vertical rmctov t,ubp containing t,hr trst catalyst, which tube i~

~

a pressure regulator into an evseuat,ed storage cylinder.

Fonnected ta tho top of theindosing cylinder while the hottom of the cylinder is connected by means of a rubher hose to a cylindrical wooden pot suspended by chain from a spring scale. The wooden pat is filled with mercury and the pat-scale assembly is suspended by a linen cord from an adjusbing rod t h a t is used to raise and lower it. Before evacuating, the height of the mwcury

cylinder. so thrtt the dressure in the tank is equal to atmospheric

voir. The a c t u d pressure in the tank is measuredhy means of 8. closed-end manomkter, X . When the collection tank is evacuated and the system is ready for operation, the gas stream is directed into the mercury pressure regulator. The gas flaws in through the capillary tube and bubbles UD through the mercury. leaving the regulator and flowing

~

~ .~ ~ ~ ~ ~ ~~~.~~ .. ~ ~. . ~ ~ ~~~~

~

~~~~~

~

~

~

.

With such back p k u r e , the mercury reservoir must he l a k r e d so that the column of mercury above the entering point of the

~~

~~~~~~~~

mercury in the ~riiervoirj k t . opposit,< the tip-of the capillary where the gas flows int,o the mercury. Owing to gas flow resist-

.

A flow sheet, of the catalyst t,eding unit is prcscntea a5 rigurr 3.

Figure 4.

Catalyst Testin:: Unit

At top. furnaces and gas-eolleofion tanka on s t o d platform. Instrument board 1ooaf.d directly below gas-eolleofion tanks. Conatant-~rmpcmture machine nf left, pndeyesth -stairway. I'mduof-

OCTOBER 1947

773 t,he gases f l o through ~ ~t three-Lv This stopcock normally conne with the gas-product collecting tank, but' it can also he used for diverting the gas stream a t an?. instant into a gas-sampling bottle so that t h r catalyst performaric~ran tie checked at, any time. .ifter fluwing through thr thrrr-n-a?-~topeock. the gas is conducted to an automatic nierrur!. , allows tbe gas to flon. prrssurrregulator, t 7 that from the system at atmospheric pressure into an evacuated tank, \'. The gaaes, as they are CUILected in t h r tank, are sampled by means of a Toepler pump and analyzed. T h r tank is tht'ii brought up to atmospheric pressure by means of nitrogen from cylinder Yand is purged for sonic' time into a hood before rc~-evaciiatinghv i r e i n h

niaiutained a t t h r desired trriiperature by mea~isi i i furiiace E , The t,eniperature of the furnace i controlled b y means of a microswitch operated in conjunction \ t h a mercury relay. The back EXPERIXIENTAL RESULTS pressure caused by material flax across the unit is indicated by manonicter F. Directional f l o ~arrows on the line connecting l i t d t i frorn forty runs of 8 h(JUI'S' tluratioii u s t d t o st:rrldF with t,hecatalyst tuhtt are in both directions, since the line serves ardize the first catalyst testing unit aria shown in Table I. two purposes. First, tht, hac*kpressure from the unit is exerted I n the calculation of the ultimate efficiency prrccJritage, the through the line to F during operation, and secondly, the line varries nitrogen gas tliat it; uwd t o purge t h e unit before and molrs of rtAaetc.d acrtaldt.2iydr arcs divided by 0.92 because, in aftrr each test run. TtiP iiitrogrri is by-passed a c r o . manometer ~ plant oprration, rthanol is convrrtrti t o acetaldehyde by a sepaI ) through this dual-purposc linr to the reactor. ratr rcwcntion n-hivh is approximately 9 2 5 eficirrit. The ulThe converter-makc, flo\vs from the reactor in furriaw E into tiniatcs c.firit.nc*y is thercJforr rsprtAsseci 011 thr h I hr sPparation unit at point G. Here the converter-makr rnters R cooling coil immerstd in iri. \rater that ip circulatrd by a ('enrrifugal pump, H , from t h e icr watrr reservoir, I . -111 thr conskiisable materials i'riini t hr. converter are coridrrised iri this coil n u d flow into an unpackrtl sertion of the separation column at point J . :\t thi.: point t h t . liquid materials flon- down into the \vhiir thr uncondenscd gases flow up into iiwtionating column, h-, t l l t l lvater scrubher, L . l\'attir is admitted t o I, by mrans of an ,i,ific~ feeder, .If,oprrattd b y maintaining a caonatarit differential prtwurc across the foedt~i,orifice t)y means of riitrogrn pressure iri cylinder S,whirti is cwntrcilled hy niercurv bubbler 0 and dif:'rwntial pressure niaiio~nrtrrP. -1thrrniomcbter placrd a t the ve.5 to sho\v that thr butadiene fractionated out of ction. The hutadienr and 0th yaws f l o i ~fivm the t o p of the water icruhhrr into a hydroq-laminr hydwchloride iolrition scrubher, 0. From 0 thcs gas f l o w through drying :uhe R and then over a iiiei,c.ury-fillrd mariometcr, AS. which s h o w r t i e pr.(,ssurc in t h c *ystrm at that point. From thv riianomt~tc~r q

Quantitative Analysis with the X-Ray Spectrometer JOHA C. I