Pure & Appl.Chem., Vol.52, pp.2091—2103.Pergamon Press Ltd. 1980. Printed in Great Britain.

MOLECULAR SHAPE SELECTIVE CATALYSIS

Paul B. Weisz

Central Research Division, Mobil Research and DevelopmentCorporation, P.O. Box 1025, Princeton, New Jersey 08540, USA

Mstract — Molecular shape selective catalysis in intracrys—talline space of siliceous zeolites has opened up a spectrumof new opportunities in catalytic science and in industry.The mechanisms involved include: (a) selective, diffusionalmass transport dynamics, and (b) steric modification of theintrinsic reaction rates.

THE CHALLENGE

Almost any thermodynamically possible chemical reactions may occur to someextent at a sufficiently high temperature. It follows that the most uniquerole and purpose of a catalyst is to provide selectivity to direct the trans-formation along a very specific, desired path.

Manmade catalysts are usually only "class selective". For example, a hydro-genation catalyst will hydrogenate double bonds in any reactant molecule, andregardless of position in a molecule, with very few exceptions.

In contrast, nature's catalysts possess detailed structural selectivity, in-cluding position of the reactive group, structural shape, size, and symmetry.This presents a challenge to man's science and design of catalysts.

"ACCURATE" CATALYTIC SOLIDS, AND MOLECULAR SHAPE SELECTIVITY

Kinetic principles require existence of at least some 1018 reactive centersper cm3 of catalyst volume to obtain typically useful catalytic conversionrates (1) in a process requiring an activation energy of 30 kcal/mole, at500°C. For solid contact, this requires large specific surface areas, corre-sponding to subdivision in the range of hundreds of Angstroms or less. It wastraditionally assumed that only amorphous solids were candidates for suchcatalysis. In the 1950's, it was demonstrated (2) that a unique class ofcrystalline solids could provide catalytic activity at exceptionally high rateswithin intracrystalline space. These materials are certain siliceous zeolites,members of the tectosilicate family.

In these solids, the location of catalysis is the intracrystalline environment.As such, it is accurately describable by available physical chemical tech-niques. Controllable physical and chemical manipulations can be undertaken.

Intracrystalline molecular shape selectivityIn some of the available zeolites, the intracrystalline space accessible tomolecules has dimensions near those of the molecules themselves. Thus arosethe concept of providing suitable catalytic centers within such a crystallinestructure so that only certain molecular structures could move or could existwithin the structural environment of the intracrystalline catalytic apparatus.

Reactant selective catalysis was demonstrated (3) using the internal acidicsites of a Linde 5A zeolite. As shown in Fig. 1, it was possible to catalyzethe dehydration of n-butanol without reacting isobutanol on this catalyst. Onthe large pore CaX zeolite catalyst, such size discrimination was shown not tooccur. There, in fact, the isomeric alcohol is actually more actively decom-posed. Other demonstrations followed: the near complete discrimination incatalyzed combustion of linear aliphatic molecules without any noticeable

2091

z0Cl)

U>z00

TABLE 1. Highly selective combustion of hydrocarbon mixturesover Pt-containing CaA zeolite (see: Ref. 3b)

Reactant Mixture(about equal parts)

Conversion, %(combustion)

n-Butanei-Butane

97.0<0.1

Propylenen-Butene-2i-Butene

98.698.6<1.0

Molecular engineering became possible in the sense of deliberately "designing"a catalyst to accomplish a desired activity. A selective hydrogenation cata-lyst was devised (5) that hydrogenated only ethylene in admixture (1:1) withpropylene, by removing sodium from mordenite, then adsorbing platinum—contain-ing ions (that are otherwise too large to penetrate the sodium—mordenite) andthereafter replacing the sodium. The intracrystalline structure of a sodium—mordenite provides adequate mobility for ethylene, propylene, and ethane, butnot for propane (4). Therefore, linear olef ins and hydrogen could reach Ptsites, but after hydrogenation, only ethane could escape.

Product selectivity was operative in this case: the platinuni-sodium-mordeniteacted selectively by virtue of the difference in transport capabilities of theproduct alternatives, rather than by differentiation between reactant mole-cules.

Similarly, in the cracking of hydrocarbons on protonic sites contained ingmelinite (3b), isoparaff in products are virtually absent. Chen, Lucki andMower (6) demonstrated this extensively for cracking in erionite of n-paraff insas large as from C10 to C36.

2092 PAUL B. WEISZ

00

50

i-C4OH

Ca X

n - C0H

n-C4OH

EF

I ...-i-C4OH 4

0ooC 300°C

Fig. 1. Conversion of isobutanol vs. n-butanol on non-shape selective CaX and on shape selective CaA catalyst(data from Ref. 2).

reaction of their branched species (3b) as shown in Table 1; selective hydro-genation of the linear diolef in piperylene without conversion of isoprene(3b); and selective hydrogenation of linear olef ins (3a,b).

Molecular shape selective catalysis 2093

EARLY CHOICES AND NEW ZEOLITES

These early developments were greatly limited by two circumstances: first,a limited choice of zeolites--the only narrow pore structure readily availablewas the Linde A zeolite structure; second, high concentration of Al sites in asilica zeolite structure prevents the ready conversion to and maintenance ofprotonic sites.

In a search for zeolites with lower Al site density (e.g., with high Si/Alratio) , we had to resort to the use of mineral collectors' samples, such aschabazite or gmelinite (3b) . Awareness of large scale deposits of sedimen-tary zeolites developed later. But most of these are either dense (e.g.,phillipsite), or penetrable by only linear molecules (e.g., chabazite anderionite) . However, access to erionite in sizeable quantity led to the firstindustrial shape selective process (Selectoforming (7)).

There was a need for synthetic zeolites with high Si/Al ratio, and for struc-tures with intracrystalline apertures larger than the 8 oxygen ring structuresin zeolite A, chabazite, erionite, etc. For example, shape selective crackingof linear paraffins was potentially useful for removing waxes from petroleumoils, but these waxes turned out to contain appreciable numbers of long chainparaffins with single methyl groups. They would not penetrate these zeolites.

Mordenite was available, with some potential for shape selectivity, by inser-tion of suitably bulky ions (for a review, see (8)). Mordenite has a linearchannel structure with no intersecting pathways. This presented special prob-lems: single file diffusion (9) tends to crèate4a tendency for rapid deacti-vation due to channel obstruction by relatively few adsorbed molecular en-tities, or decomposition products.

The discovery of methods for zeolite synthesis that ised large organic cations(10, 11) as templates, in place of the traditional all-inorganic ionic species,opened the way to the synthesis of many new zeolites (11).

One of the new zeolites was ZSM-5 (12). Chen and Garwood observed that it wasable to accept and crack mono—methyl in addition to normal paraff ins. It wastherefore a candidate for "dewaxing" heavy oils. It also allowed selectivereaction involving some mono-cyclic hydrocarbons (see: Ref. 14). Its crystalstructure (13) has a two dimensional network of intersecting channels, withdimensions of about 5.4 x 5.6 Angstroms (see Note a) created by a 10 oxygenatom ring. This allows a large variety of molecular discriminations to occur(14)

In a relatively short time, a number of novel commercial processes were de-veloped which made use of ZSM-5-type catalysts.

CONVERSION PROCESSES

Table 2 summarizes the major shape selective conversion processes that havebeen developed.

"Selectoforming" was developed and used to crack n-paraff ins from a gasoline,to mostly propane (7). This simultaneously removes from the gasoline thecomponents having the lowest octane number (ON) contribution (thus raising theON), and produces LPG as a valuable product.

In M-forming (15), the unique and graduated selectivities of ZSM-5 for crack-ing various paraffin structures are utilized. These are illustrated by thedata of Chen and Garwood (14) in Fig. 2. The cracking ability of the catalystfor these paraffin structures correlates inversely with their blending octanenumbers, Fig. 3, making the ZSM-5 conversion catalyst an "intelligent" octanenumber improver. In the same process, alkylation of aromatics with crackedfragments also occurs, thereby increasing the liquid yield. Contrary to con-ventional relative alkylation activities of benzene and toluene, benzene isreacted preferentially (14). This results in lowering benzene/toluene ratiosin the product.

Note a. Based on X-ray structure determination using an oxygen radius of 1.35Angstroms. Zeolite A, on this basis, has a circular aperture diameter of 4.1Angstroms.

2094 PAUL B. WEISZ

TABLE 2. Industrial processes based on shape selective zeolites

MAJOR CHEMICAL/PROCESSCHARACTERISTICS

Selective n-paraffincracking

Cracking depending ondegree of branching;aromatics alkylation bycracked fragments

Cracking of highmolecular weightn- and mono-methylparaff ins

High throughput,long cycle life;suppression of sidereactions

Synthesis of hydrocarbonsonly, restricted to gaso-line range (C4 to C10)including aromatics

cc-c-c-c c-c-c-c

I I S

CC C

0.09 0.09

C

c-c-c-c-c c-c-cc-c c-c-c-c-cI S I $CC CC C0.09 0.05 0.17

Fig. 2. Relative cracking rate constants at 340°C in HZSM-5catalyst (after Ref. 14).

0

U00

0

Molecular shape selective catalysis 2095

9

8

7

• C7• C6

A C5

S

4

3

2

"0 I I I20 40 60 80

Research Octane Number (R+ 0)

Fig. 3. Selective cracking vs. octane number of paraff ins(after Ref. 18)

The dewaxing processes (16-19) have made it possible to make use of a largerfraction of heavy portions of petroleum raw material as diesel and heatingoils, and to provide high grade lubricatinq oils with low energy expendi-ture. The catalyst removes selectively those paraffinic portions of a heavyoil responsible for its deficient flow properties at low ambient temperature(high "pour point"). The catalyst also has the potential of converting waxycrude oils into upgraded, transportable fluids.

The new xylene isomerization and ethyl benzene processes (16, 21, 35), basedon ZSM-5 type catalysts, create the desired products at unprecedented yieldsby reducing all important side reactions. The xylene processes are now usedin some 90% of U.S. and 60% of existing western world capacity. Some of themechanistic aspects are noted below.

Toluene disproportionation catalysis (16, 18, 35) provides high product selec-tivity in the manufacture of xylenes and benzene from toluene. Again, sideproduct suppression is achieved by the catalyst. Potential improvements thatwould selectively create high concentrations of para-xylene produced directlyhave been demonstrated, as noted below.

Development of a methanol-to-gasoline (MTG) process (16, 22, 23) followed thefinding that the ZSM-5 catalyst is capable of converting oxygen compounds suchas alcohols, aldehydes, ketones, carboxylic acids, and esters to hydrocarbonsin a limited range of structural sizes, as demonstrated by Chang and Silvestri(20), and Kaeding and Butter (24).

The MTG process, developed to the stage for commercialization, converts theone carbon compound methanol into a mixture of hydrocarbons, in the gasolinerange, with high antiknock quality. The process is the final link in systemsto convert coal or natural gas into motor fuel. Fixed bed and fluidized bedreactors have been demonstrated in pilot units. Construction of a pilot plantfor scale-up of the fluidized bed reaction is planned with cooperation of theUnited States and the Federal Republic of Germany. The New Zealand governmentplans to produce gasoline from natural gas in a commercial plant based on thefixed bed scheme.

PRINCIPLES

It is convenient to consider a pair of molecular species, P and A. P shall bethe "proselective" species, and A the "antiselective" species in the sense thatshape selective discrimination will favor the reaction (or creation) of P.The degree of shape selectivity S is the ratio of the appropriate observed

P.A.A.C. 52/9—D

2096 PAUL B. WEISZ

effective rates or rate constants, as compared to their usual ratio in theabsence of structural constraints on molecular motion:

S(k'p/k'a)/(kp/ka).

Or, if the intrinsic rates in.the absence of spatial constraInts are equal,then:

S = k'p/k'.Shape selectivity via mass transport discriminationShape selectivity (S > 1) will occur, not only when A is entirely unable tomove through intracrystalline space, but also when Da<< D , i.e., when thediffusivity of the antiselective species A is 'sufficientrv smaller than thatof the proselective species P. Shape selectivity wifl result from the wellknown diffusion/reaction dynamics.

The effective diffusivity. In shape selective catalysis, the motion of mole-cules is restricted when the structural dimensions of the catalyst approachthose of the molecules. We have termed this the regime of "configUrational"diffusion (1). Even subtle changes in the dimensions of molecules can resultin large changes in diffusivity.

For example, a greater than 200-fold reduction in the diffusion rate of butenein CaA zeolite results for the change from the trans to the cis form (5).Similarly, the measurements by Riekert (26) in T zeolite demonstrate a pro-found (three orders of magnitude) increase in mobility on "stiffening" onebond in the 4 carbon chain, as we compare n—butane with n—butene-2, Fig. 4.

Fig. 4. Diffusivity in HT zeolite. Effect of "stiffening"one bond in the 4 carbon hydrocarbon.

In molecular shape selectivity, not only the size, but also the dynamics ofthe structure of the molecules, need to be considered. For example, we havedemonstrated (27) that a triglyceride as large as C57H10406 will successfullyreact in ZSM-5 to substantially the same product spectrum of hydrocarbons asdoes methanol, CH3OH; see Fig. 5A. While the conventional "writing" of thestructural formula, Fig. 5B, suggests a structure too large to penetrate thecrystalline channel, molecular dynamics is such as to easily attain configur-ations that have an effective cross section no greater than that of trimethylbenzene, Fig. 5C.

DIFFUSIVITY IN H-T ZEOLITERIEKERT, A. I. CK E. J.l7 446, 971

Dcmec

Molecular shape selective catalysis 2097

Paratf Ins Ol.fins A'omatlcs NonaromaticsI

I.

Fig. 5. (A) Conversion of corn oil triglyceride C57H10406 com-pared to that of methanol, on ZSM-5; (B) "written"; (C) possibleconfiguration of C57 molecule.

In the configurational regime, 'motion occurs in a force field with replusivepotential rising rapidly with closer approach between molecular and structuralatoms. The effective cross section for successful motion will depend on trans-lational energy, resulting in a temperature dependency, i.e., an "activationenergy" for the diffusivity (28).

Just as has been demonstrated for rnonotonic gases in denser zeolites (see:e.g., Barrer (29)), this applies to organic molecules in shape selectivezeolites. For example, Gorring (30) has measured (see Fig. 6) rising activa-tion energies for diffusion range up to about 16 kcal/mole of ethane, n-butane,and n—octane in a T zeolite. These, and measurements by Gorring for some

Iol0 -

U

l2>.I.->(nDI.La. I4a

15.2

kcalMole

1516.001 .002

I/ TKFig. 6. Diffusion coefficients ofT zeolite.

Methanol30

20

10'

450°C, WHSV 0.67, 0% coke

CH3-OH

0 5 ib1

15

Corn oil C57HIn40a

20

c,&c

C:0

C

'0c'0 'c'0\ \Cl7 C17

0C''17cL-o—c' l7I

c.,_01.... 170

4OC, WHSV .4, 3% coke9

C1HC-0-CH2

9,C17HC-0-CH

olC17HC-0-CH2

I'0

00 5

L

lb IS

(A) (B) (C)

.003

hydrocarbons in potassium

2098 PAUL B. WEISZ

alkyl benzenes in Na-ZSM-5, Table 3, reflect the experience, common with otherfields (gases in dense zeolites and glasses or carbons, dyes in fibers andpolymers), that the activation energy rises with closer approach between thevoid dimensions and those of the molecular configurations.

TABLE 3. Activation energies for diffusion in Na-ZSM-5, fromsorption studies at 250°C to 350°C

kcal/mole

o-Xylene 9

m-Xylene 14t-Butyl benzene 121,2,4—trimethyl benzene 141,3,5—trimethyl benzene 19

Diffusion/reaction interaction. The behavior that can be expected for shapeselective kinetics between the antiselective species A and the proselectiveP, due to mass transport inhibition, can be illustrated by using the usualdiffusion/reaction models for first order reactions (see Note b).

For each species, the reaction rate inhibition due to diffusion effects willbe determined by the effectiveness factor n and its well known relationship(Fig. 7, top) to the characteristic parameter 4) = k(R2/D). R is the particleradius of the zeolite crystallite, k the intrinsic reaction rate constant,and D the diffusivity. To test the discrimination between two species due toa large diffusivity difference, we assume k to be the same. Then, the shapeselectivity S can be read from the ratio of the values in Fig. 7 (top).

Figure 7 (bottom) shows how the shape selectivity S depends on the ratio ofthe diffusivities Dp/Da., and on the location of the parameter k(R2/D) for theantiselective species.

Fig. 7. Top: the conventional dependence of utilization factor n.Bottom: selectivity S vs. k(R2/D) of the antiselective species.

Note b. Because of symmetry of the equations, the same considerations applyto reactant as well as to product selectivity, when these products are gener-ated internally in a first order reversible reaction.

I.0

0.1Ii

0.0I

100 -

io' 10 i02 } k(R2/D)

I0 S

Molecular shape selective catalysis 2099

The selectivity ratio can always vary from S = 1.0 (no shape selectivity) toS = (Dp/Da) for the highest shape induced selectivity. For any given in-trinsic activity level, we note that shape selectivity requires not only thefavorabl Dp/DaratiO but, in addition, demands operation with a minimum fac-tor of R /Da. This factor is proportional to the characteristic diffusion timeTa of the antiselective species in the crystallite, since Ta R2/Da.

Thus, a potential shape selective effect by virtue of a large Dp/Dai when notobserved, can be developed or increased by increasing the crystallite size. Asimilar shift in diffusive residence time would be achieved by inducing poreblockage in a fraction of the pore passages, thus increasing the random walktortuosity. This would keep constant Dp/Da, but would decrease the effective

Both of these methods have been used successfully to obtain selectivity forhighly selective generation of only the para—xylene isomer in methanol-tolueneaddition and in disproportionation of toluene, Fig. 8.

REL.D REL.D

>IOvWWr H20

ICC .—wwwv-.

>i W'Wv© +CH3OHII C C

I

> +I

I C-Jw-I—.AiWvW

I

LFig. 8. Diffusion/reaction kinetics in para-xylene generationfrom disproportionation or methanol alkylation of toluene.

Chen, Kaeding and Dwyer (31) have achieved para-xylene selectivities of 90% ormore of total xylenes by introducing inorganic matter into the pore structureof ZSM-5 catalyst. Also, increasing the crystallite size alone from <0.5pm to3pm increased para-xylene concentration from 23% (equilibrium)to 46%. Haagand Olson (33) have obtained this also by closure of a fraction of the exter-nal pore mouths, again effecting a tortuosity increase.

Haag and Olson (32) demonstrated data for para—xylene concentrations achievedin toluene disproportionation at standard conditions for twelve ZSM—5 cata—lysts modified in various ways--by crystal size variationA coke deposition, orinorganic fillers--all o which correlate uniquely with R/D, a diffusionresidence time for ortho-xylene (an antiselective species) determined indepen-dently by sorption measurements. Their data, shown in Fig. 9, are fully con-sistent with the shape selectivity resulting from a classical diffusion/reac-tion mechanism as illustrated in Fig. 7.

Para-selectivity has also been achieved in modified mordenites (34). For ex-ample, CeNa mordenite showed high selectivity for mono-alkylation, and forpara-positional generation of cymene.

Shape selectivity via local configurational constraintIn the configurational regime, molecular motion is subject to the energeticsof molecule/framework force fields. This constraint at the local level finds

2100 PAUL B. WEISZ

80

,700U

E50

4030

20I0

0

Fig. 9. Para-xylene selectivity vs. o-xylene sorption time invarious ZSN-5 catalysts. 0 = various crystallite sizes; t =increased tortuosity by inorganic salts; I = increased tortu-osity by coke plugging. (After Ref. 32)

expression in the sizable "activation energy" of diffusivity. If the cataly-tic sites are located within the same constraining structure, then the fre-quency of successful encounters in local kinetic events can be expected toalso be sensitively dependent on size, spatial molecular configuration. anddynamics, and on temperature; this could include successful moleóule/sitecontact, or successful creation or transition of a specific reaction complex.

Discrimination in reactions with bimolecular complexes. The accommodation ofsuccessful bimolecular encounters and reaction complexes on a catalytic sitein constrained space will be more difficult than a monomolecular event.

This is believed to be a major phenomenon in creating the high selectivity inthe ZSN-5 xylene isomerization process. Fig. 10 shows this in terms of theratio of the rate constants for the unwanted bimolecular disproportionationreaction to that of the desired monomolecular xylene isomerizatioñ, observedby Haag and Dwyer (35) for four zeolite catalysts of varying intracrystallinechannel dimensions. A similar discrimination exists in the ethyl benzene dis-proportionat ion.

0

0.6 .8 1.0 1.2

nm

EFFECTiVE DIAMETER OF INTRACRYSTALLINE CAVITY

Fig. 10. Selectivity in xylene isomerization vs. zeolite pore diameter.

Sorption Time for o-xylene, t (mm)

50

40

30

10

Molecular shape selective catalysis 2101

Csicsery (8) has observed the suppression of bulky 1,3,5-trialkyl benzeneformation in the disproportionation of dialkyl benzenes in mordenites, and haspointed out the mechanisms of constraint or prohibition of the transitionstate (8).

The characteristic of these rea,ctions is the fact that the reaction complex islarger than either the reactants or the products. The question of a transportmechanism does not apply for the spatial inhibition mechanism operating on thecomplex.

Shape selective inhibition of reactant' reactiv4y. It is natural.to assumethat reactant selectivity in shape selective catalysts is due to mass transportinhibition of the antiselective species' in line with diffusion/reactioninteraction discussed previously. However, there is evidence that shape selec-tive constraints on the local site kinetics can be operative.

The classical diffusion/reaction mode requires the selectivity between speciesto increase with increasing temperature unless the activation energy of dif-fusion exceeds that of the intrinsic reactivity. The first observations madeon alcohol dehydration (3) showed selectivity decreasing with temperature.Then Chen, and Garwood (l4)established a strong decrease of n-hexane/3-methylpentane selectivity in cracking over HZSM-5 catalyst. Judging from the dif-fusivity measurements (Table 3) with much bulkier aromatics molecules, wecannot expect the paraffin diffusivities to have appreciable activation ener-gies, let alone exceed that of catalytic reactivity. Definitive proof forshape selectivity constraints on intrinsic kinetics, rather than via dif-fusive transport inhibition, developed from work by Frilette, Haag and Lago(36). Measurements of the cracking rates were made in a differential reactorfor catalyst with crystallite sizes differing by two orders of magnitude (orby four orders of magnitude for R2/D). Figure 11 shows virtual independenceof behavior on size .

z

1

103/T0K

Fig. 11. Cracking rates ofsizes, 0.02 to 0.5 rim (opensymbols).

ZSN-5 catalysts of two crystallitesymbols) and 2 to 4 nm (solid

It may be too early to speculate concerning the detailed mechanism of stericinhibition, whether it occurs due to the need for forming a large reactioncomplex between a molecule and a sizable site—anchored carbonium ion, orwhether the formation of the site complex'("sticking factor") requires suchspecial geometric approach as to be hindered by the surrounding crystalstructure.

1.4 1.5 1.6 1.7

2102 PAUL B. WEISZ

It is notable that Derouane, et al (37), in extensive recent work on ZSM—5selectivity, have been led to suggest behavior influenced by intermediate corn—plex structures being able to form only in special locations in ZSM-5; forexample, the formation of aromatics from cycloaddition of a carbonium ion andan olefin is a complex that may form only at a channel intersection, while thearomatic product is small enough to diffusethrough the channels. The systemwould be analogous to the cases described previously. The complex cannotmove, but the reactant and product molecules, being smaller, can diffuse ade—quately .

Shape selective inhibition in product formation. Since the early dernonstra—tions of product selectivity, many more different cases have been observed,many reviewed and reported by Csicsery (8) and Derouane (37) , including anti—Markownikoff additions of HBr to a-olefins by Fetterly and Koenitz (38) , acase of base catalyzed shape selectivity in crossed aldol condensations byShabtai (34) , and others. Shabtai also has demonstrated new potentials forshape selective catalysis with inorganic materials that are not siliceouszeolites (34).

It is not clear yet when product selectivities are produced by mass transport(diffusion) inhibition of the antiselective species, and where steric in-hibition in the formation of the relevant reaction complex or product is oper-ative. Most likely, there are cases where both mechanisms are operative. Thestudy of kinetic behavior of various crystallite sizes is likely to be themost important method in discriminating between the two effects in futurework.

CONCLUS IONS

Shape selective catalysis has created opportunities for new chemistry and newindustrial processes. We appear to be in the typical early stage of enteringnew territory: discoveries and utilization of new features have been out—pacing our ability to grasp all quantitative detail of the mechanism. It nowappears that selective diffusion inhibition, as well as selective stericinhibition in site kinetics, or a combination of both phenomena, are oper-ative in shape selective catalysis. The selectivities achievable move ourabilities a step further toward the precise structure specific selectivitiesof the biochemical processes.

ACKNOWLEDGMENT

Many past and present colleagues made it possible to report on science andapplications of shape selective catalysis. Thanks go to W. 0. Haag, N. Y.Chen, and R. L. Gorring for the use of material and assistance in preparingthis paper.

REFERENCES

1. p. B. Weisz, Chemtech 3, 498 (1973).2. V. J. Frilette, P. B. Weisz, R.. L. Golden, J. Cat. 1, 301 (1962).3. (a) P. B. Weisz, V. J. Frilette, R. W. Maatman, E. B. Mower, J.Cat. 1,

307 (1962); (b) P. B. Weisz, Erdol und Kohle 18, 527 (1965).4. R. N. Barrer, Trans. Faraday Soc. 40, 555 (1944).5. N. Y. Chen, P. B. Weisz, Eng. Prog. Symp. Ser. 63 (73) 86 (1967).6. N. Y. Chen, S. J. Lucki, E. B. Mower, J. Cat. 13, 329 (1969).7. N. Y. Chen, J. Maziuk, A. B. Schwartz, P. B. Weisz, Oil & Gas J. 66 (47)

154 (1968); from Selectoforming, Hydrocarbon Proc. 49 (9) 192 (1970).8. S. M. Csicsery, chapter on Shape Selective Catalysis, from Zeolite Chem-

istry and Catalysis, ACS Monograph 171, J. A. Rabo, Ed., MiericanChemical Society, Washington, D.C. (1976).

9. A. C. Hodgkin, R. D. Keynes, 3. Physiol. (London) 128, 61 (1955);L. Riekert, chapter on Sorption, Diffusion and Catalytic Reaction inZeolites, from Advances in Catalysis, Volume 21, Academic Press,New York (1970).

10. R. N. Barrer, P. J. Denny, J. Chem. Soc. 971 (1961).11. G. T. Kerr, Science 140, 1412 (1963); G. T. Kerr, J. Inorg. Chern. 5,

1537 (1966)12. R. J. Argauer, G. R. Landolt, U.S. Patent 3,702,886.

(8961) zoi '•Pui •uIattD J0 •3OS L96T 1T(dV 1UOpUOnI1JU03 98A8TS inoeiow 'z;UeO>j • •)I 'c-t;ta;;ei i 8E

(ssaxd tIT) Nt0A MOM '998td OTO 'S 'uosqoo .r •Y •s •w 'Ic;t:sTuLet.ID UOTTOteUI uxo;tj: '9eYT09Z

UT STSTP3 OATD9T9S edTtS pU_SUOTTU1T[ TUOT9flTU 'UflOtQ •E •: •(•:3 •Q o: pauiqns) xepul 4uTSUOD

aq: )q se4TToez ZTS etOd 8TP88UI 0 — _OUTUmTV euTTs!x3 1q STSJTP3 '°5i U 'BH 0 M 'eTT 'f A

6L6T srL6nV 's;es —flT4OSSJ 'UOSO ''I5u Uj EUI U13 SUI 'IIIV 'SSAT3 ;TIOz

9ZTS 3t0d a Tpal]UaUI q3TM BUTssaoOEd soT;urotv 'aa 'fn O M SE (6L61 'ssetd UT) a POU1q3 I (LL61) c1Pi

'U1TW 's;ueEei S St9UIA{Od U0 U03 '0td 'TqqS r 'tS'L6O't' ' fl 'UOS[J 'H 'ci '5WH O M 'E '9O'L11't U9d S fl 'Uos[o 'H 'ci 'EH '0 'M '?

____________________ (.6L61)

£8L9 '101 ''DOS 'U183U1 'f 'teJt ' I '6UTP9P)i 'M M 'Ue3 ' ' '- (EL61) £1 'IE '3 .f '5uPUtOE ' •OE (8961) L 9q1t4deS ';oi 'sbd ''put ' .req3 'tts w i '6Z

8L61 ')[tO Mej\I 'SSetd DTUIePe1 'SeAeTS tnzei°w pU sueqto5 s S1PIaUTH 1ce[3 PUP seT1OeZ 't1P5 'M 'l '8Z

'(6L61) LS '90Z eDUeTOS 'P1ePoH ) d 'hPPH 0 M 'ZSTeM 'El 'd 'LZ (1L61) 9 'LI '12 atl3IV 'ta[eTpj "1 9Z

(0861) ti Ainu '96—S6 sl5d '' 5f ITO 'SZ (0861) SSI '19 'P3 'f 't9fl5 'V 'S 'E,UTPe>I M 'M 'z

'0861 'uxpie;suIv ''03 'c(ttd DT;TUeTDS teTA8S1 'sUP1d 1eTuieq3 eE.xi atot; 'sseoota

eUç1OSPO 0 [oUPq;e qOM 'qaxn. s 'UPuD[eeM 'M 'A 'NnTzPW '12

'eei M (8L61) SS 'LI 'Aeci 'sea 'DOtd 'u1e143 'f)u[ 'pU] 'TIseAITS '12 'V 'TM '12 'f 'qoo 'w '5 fuPrI 'H 'M 'Qfl)I 'M '3 '12 'EUPq3 '(1 '3 '

_________ (9L61) 98 '9

qoeuIeT3 'ZSM 'El 'd 'H '3 'Ubno[InDDW d '12 'IeSTeW "-I 'S 'LL6I

4DPW 'UPTSTfl0I 'suPeIto MN ''6u; 'UV ELI ''DOS 'UI93 'UII 'TUfl euezue tehpPEl/lTgOw_ITeuuo3 StT 'xeAMQ '0 'I spterI '12 'ci

(9L61) 06 (I) ? '6u 'uiaq3 '.xepTeUqDs 'M ' 15M9nI '12 •d 'teAMQ 'I) 'd 'IZ (LL6I) 6tZ 'Lt 'P3 'f 'TseA1TS '1' 'V '5UPq3 'a '3 'oz ___________________

(6L61) 611 (s) 8S 'oo UOqJPDOPAH PzP n1 ';JPH 'S 'V 'TUTPeH 3 'pUPIaXI 'H H '61

'6L61 eqeAoN 'PTUt0TIP3 '0D5TDUaX, UPS ''5ui 'UV L 'J5UH 'W9ij3

;o 'sui 'ur2 'ss'P;3 e;Troez 9Z5 axo ePTpeuLIeUI .t9AO uoTsXeAUO3 eATDe1aS edq5 'ZtPt4qD5 'El 'V '5PPH '0 'M 'POOb1PO 'H M 'Ul3 'A 'N '81

______________ (LL6I) S91 (E)

'SL '12 SPf) ITO 'UT95 ' '31 'puP[axI ')J 'H 'fUT00 ' 'j 'U8T43 'A 'N 'LI 'LL6I Sfl&V

'STOUTIII 'O5PDTq3 ''bU '005 'uleq3 'ur L1 ''thtA pUeT oei 's;si —PP3 eToez eAO spDTuxeq3 pUP sien O UOTDflpO.Id eq; UT 5DUPLPV

U9D9H 'zsTeM 'El 'ci 'tI IPPI 'H '3 'If)flOI1fl3DW 'ci '12 'ITW .rI 'S '91 '0861 'sPxe31 'U045fl.OH 'bUi '005 'U18q3 'UL 'se;Tloez eAToeIeS

edqs eo ePuuoje ;o UTsseood 0Tk[PP3 30 1PTUeOd bUTSOOEl 9UP0Q 'POOt'XP0 'H 'M 'U3 'A 'N (LL6I) ES 'AeH 'P3 'UUPUX9UT8H 'H 'SI

'(8L61) ES7 'ZS 'P3 'f 'pOO!JItPD 'H 'M 'U943 'A 'N 'I — _______ ' (8L61) LEi 'ZLZ nPN 'leTew 'M 'UOSIO 'H 'ci IUOMprI •rI 'S 'OTTPONO)I '31 '1) 'El

sis aas edqs lfl0iON