Cumene Production

C
Robert J. Schmidt
UOP LLC, Des Plaines, Illinois, U.S.A.

INTRODUCTION number of cumene units remain using the fixed-bed,

kieselguhr-supported solid phosphoric acid (SPA)
The history of the cumene market has been examined catalyst process developed by UOP and the homoge-
in great detail with much discussion regarding product nous AlCl3 and hydrogen chloride catalyst system
usage, emerging markets, and process economics over developed by Monsanto.
the past 10–20 yr.[1] With more than 90% of the world’s
phenol production technology currently based on the
Solid Phosphoric Acid Catalyst
cumene hydroperoxide route, it is the focus of this
entry to review the latest technology improvements in
Although SPA remains a viable catalyst for cumene
cumene processing made over the past 10 yr. Current
synthesis, it has several important limitations: 1)
state-of-the-art processes for the production of cumene
cumene yield is limited to about 95% because of the
as a feedstock for phenol involve zeolitic catalyst
oligomerization of propylene and the formation of
technology offerings from UOP, Badger Licensing
heavy alkylate by-products; 2) the process requires a
(formerly ExxonMobil and the Washington Group),
relatively high benzene=propylene (B=P) molar feed
and CDTech based on zeolitic catalysis. Much of
ratio on the order of 7=1 to maintain such a cumene
the improvements in these technologies relate to yield,
yield; and 3) the catalyst is not regenerable and must
stability, and operating costs.
be disposed of at the end of each short catalyst cycle.
Also, in recent years, producers have been given
CUMENE PRODUCTION increasing incentives for better cumene product quality
to improve the quality of the phenol, acetone, and
Cumene is produced commercially through the alkyla- especially a-methylstyrene (e.g., cumene requires a low
tion of benzene with propylene over an acid catalyst. butylbenzene content) produced from the downstream
Over the years, many different catalysts have been used phenol units.
for this alkylation reaction, including boron trifluoride, For the UOP SPA catalyst process, propylene feed,
hydrogen fluoride, aluminum chloride, and phosphoric fresh benzene feed, and recycle benzene are charged
acid. Cumene processes were originally developed upflow to a fixed-bed reactor, which operates at
between 1939 and 1945 to meet the demand for high- 3 4 MPa (400–600 psig) and at 200–260 C. The SPA
octane aviation gasoline during World War II.[2,3] By catalyst provides an essentially complete conversion
1989, about 95% of cumene demand was used as an of propylene on a one-pass basis. A typical reactor
intermediate for the production of phenol and acetone. effluent yield contains 94.8 wt% cumene and 3.1 wt%
Today, nearly all cumene is used for production of diisopropylbenzene (DIPB). The remaining 2.1% is
phenol and acetone with only a small percentage being primarily heavy aromatics. This high yield of cumene
used for the production of a-methylstyrene. The is achieved without transalkylation of DIPB and
demand for cumene has risen at an average rate of is a key advantage to the SPA catalyst process. The
2–4% per year from 1970 to 2003.[4,5] This trend is cumene product is 99.9 wt% pure. The heavy aro-
expected to continue through at least 2010. matics, which have a research octane number (RON)
Currently, over 80% of all cumene is produced by of about 109, can be either used as high-octane gaso-
using zeolite-based processes. Early processes using line-blending components or combined with additional
zeolite-based catalyst systems were developed in the benzene and sent to a transalkylation section of the
late 1980s and included Unocal’s technology based plant where DIPB is converted to cumene. The overall
on a conventional fixed-bed system and CR&L’s yield of cumene for this process based on benzene
catalytic distillation system based on an extension and propylene is typically 97–98 wt% if transalkylation
of the CR&L MTBE technology.[6–9] At present, the is included or 94–96 wt% without transalkylation.
Q-Max2 process offered by UOP and the Badger Generally, it has been difficult to justify the addition
Cumene Technology developed by ExxonMobil and of a transalkylation section to the SPA process based
offered by Badger Licensing represent the state-of- on the relatively low incremental yield improvement
the-art zeolite-based catalyst technologies. A limited that it provides.
Encyclopedia of Chemical Processing DOI: 10.1081/E-ECHP-120026490
Copyright # 2006 by Taylor & Francis. All rights reserved. 603
604 Cumene Production

ALCL3 AND HYDROGEN CHLORIDE CATALYST

Historically, AlCl3 processes have been used more

extensively for the production of ethylbenzene (EB)
than for the production of cumene. In 1976, Monsanto
developed an improved cumene process that uses an
AlCl3 catalyst, and by the mid-1980s, the technology
had been successfully commercialized. The overall yield
of cumene for this process can be as high as 99 wt%
based on benzene and 98 wt% based on propylene.[10]
Detailed process flow information is widely published
in the literature for this technology.[11] Dry benzene,
both fresh and recycled, and propylene are mixed in Fig. 1 Beta zeolite. (View this art in color at www.dekker.-
an alkylation reaction zone with an AlCl3 and hydro- com.)
gen chloride catalyst at a temperature of less than
135 C and a pressure of less than 0.4 MPa (50 psig).[11]
The effluent from the alkylation zone is combined with acidic zeolites used at that time, such as mordenite
recycled polyisopropylbenzene and fed to a transalkyla- (MOR) or ZSM-5 (MFI).
tion zone also using AlCl3 catalyst, where polyiso- Fig. 2 shows a comparison of the main channels
propylbenzenes are transalkylated to cumene. The of these zeolites. Chevron discovered that the open
strongly acidic catalyst is separated from the organic 12-membered ring structure characteristic of beta zeo-
phase by washing the reactor effluent with water and lite coupled with its high acidity made it an excellent
caustic. The distillation section is designed to recover catalyst for aromatic alkylation. These properties were
a high-purity cumene product. The unconverted ben- key in the production of alkyl aromatics such as EB
zene and polyisopropylbenzenes are separated and and cumene in extremely high yields and with product
recycled to the reaction system. Propane in the propy- purities approaching 100%. Moreover, Chevron dis-
lene feed is recovered as liquid petroleum gas (LPG). covered that the combination of high activity and
porous structure imparted a high degree of tolerance
to many typical feed contaminants.
ZEOLITE CATALYSTS From a technical perspective, the process developed
by Chevron was a breakthrough technology in that the
In the past decade, beta zeolite (given the universal BEA) high cumene yields and purities were not attainable by
has rapidly become the catalyst of choice for commercial the other vapor phase or liquid phase processes of the
production of EB and cumene. Mobil invented the basic day. Nevertheless, the manufacturing cost of beta
beta zeolite composition of matter in 1967.[12] Since that zeolite was still too high for catalyst producers to make
time, catalysts utilizing beta have undergone a series of a commercially viable catalyst. UOP, however,
evolutionary steps leading to the development of the developed new manufacturing technology to make a
state-of-the-art catalysts such as QZ-20002 catalyst and beta zeolite based catalyst a commercial reality. In
QZ-20012 catalyst for cumene alkylation. 1991 a new cost-effective synthesis route was invented
Much of the effort between 1967 and the early 1980s by Cannan and Hinchey at UOP.[14] The new synthesis
involved characterization of the perplexing structure of route patented the substitution of diethanolamine, a
beta zeolite. It was quickly recognized that the BEA much less expensive templating agent, for a substantial
zeolite structure has a large-pore, three-dimensional fraction of tetraethylammonium hydroxide, which had
structure, and a high acidity capable of catalyzing been used in the synthesis previously. Moreover, the
many reactions. But it was not until early 1988 that route further enables the use of tetraethylammonium
scientists at Exxon finally determined the chiral nature bromide (instead of the hydroxide) as an additional
of the BEA structure, which is shown in Fig. 1. cost saving approach. Finally, the new synthesis route
While the structure of beta was being investigated, allows the practical synthesis of beta over a wider
new uses for this zeolite were being discovered. A range of silica to alumina ratios, a factor that has a
major breakthrough came in late 1988 when workers profound effect on the catalyst’s performance.
at Chevron invented a liquid phase alkylation process Subsequently, UOP sought to develop zeolitic
using beta zeolite catalyst. Chevron patented the catalysts that would overcome the limitations of SPA
process in 1990.[13] While Chevron had significant including a catalyst that is regenerable, produces higher
commercial experience with the use of Y (FAU) zeolite cumene yield, and decreases the cumene cost of
in liquid phase aromatic alkylation, Chevron quickly production. More than 100 different catalyst materials
recognized the benefits of beta over Y as well and other were screened, including mordenites, MFIs, Y-zeolites,
Cumene Production 605

Fig. 2 Comparison of various zeo-

lites. (View this art in color at
www.dekker.com.)

amorphous silica–aluminas, and beta zeolite. The most ratio. The low B=P feed ratio (2) represents the lowest
promising materials were modified to improve their in the industry and affords cumene producers the
selectivity and then subjected to more rigorous testing. option to expand capacity and=or revamp existing
On the process side, Unocal developed an early fractionation equipment with significant cost savings.
liquid phase fixed-bed reactor system based on a The Q-Max process flow scheme based on a liquid
Y-type zeolite catalyst in the 1980s.[15] The higher phase process is shown in Fig. 3. The alkylation reactor
yields associated with the liquid phase based process is divided into four catalyst beds contained in a single
were quickly recognized and adapted by the industry reactor vessel. Fresh benzene feed is routed through
with the selectivity to cumene generally falling between the upper-mid section of the depropanizer column to
70 and 90 wt% based on converted benzene and propy- remove excess water that may be present in the fresh
lene dependent on operating conditions. Major side benzene feed. Relatively dry benzene is withdrawn
products in the process are primarily polyisopropyl- from the depropanizer for routing to the alkylation
benzenes, which are transalkylated to cumene in a reactor. Recycle benzene to the alkylation and trans-
separate reaction zone to give an overall yield of alkylation reactors is recovered as a sidedraw from
cumene of about 99 wt%. The distillation requirements the benzene column. A mixture of fresh and recycle
involve the separation of propane for LPG use, the benzene is charged downflow into the alkylation
recycle of excess benzene to the reaction zones, the reactor. Propylene feed is divided into portions and
separation of polyisopropylbenzene for transalkylation injected into the alkylation reactor between the catalyst
to cumene, and the production of a purified cumene beds and each portion is essentially completely
product. consumed in each bed. An excess of benzene is used
By 1992, UOP had selected the most promising in the alkylation reactor to avoid polyalkylation and
catalyst, based on beta zeolite, for cumene production to help minimize olefin oligomerization. The alkylation
and then began to optimize a liquid phase based reaction is highly exothermic and the temperature rise
process design around this new catalyst. The result of in the reactor is controlled by recycling a portion of the
this work led to the commercialization of the UOP alkylation reactor effluent to the reactor inlet to act as
Q-Max process and the QZ-2000 catalyst in 1996. a heat sink. In addition, the inlet temperature of each
More recently in 2001, UOP commercialized a new downstream bed is reduced to the same temperature
alkylation catalyst, QZ-2001, which offers improved as the first bed inlet by injecting a portion of cooled
stability and operation as low as 2 B=P molar feed reactor effluent between beds.
606 Cumene Production

Fig. 3 Q-Max process. (View this art in color at www.dekker.com.)

Effluent from the alkylation reactor flows to the reactor. The expected catalyst cycle length is 2–4 yr,
depropanizer column, which removes any propane that and the catalyst should not need replacement for at least
may have entered with the propylene feed along with three cycles with proper care. At the end of each cycle,
excess water that may have entered with the fresh the catalyst is typically regenerated ex situ via a simple
benzene feed. The bottom stream of the depropanizer carbon burn by a certified regeneration contractor.
column goes to the benzene column where excess However, the unit can also be designed for in situ regen-
benzene is collected overhead and recycled. Any trace eration. Mild operating conditions and a corrosion-free
nonaromatics that may have been in the fresh benzene process environment permit the use of carbon–steel
feed can be purged from the benzene column to avoid construction and conventional process equipment.
an unacceptable accumulation of nonaromatic species An alternative zeolite process was developed by
in the benzene recycle stream. The benzene column CR&L and licensed by CDTech and is based on the
bottom stream goes to the cumene column where the concept of catalytic distillation, which is a combination
cumene product is recovered overhead. The cumene of catalytic reaction and distillation in a single col-
column bottoms stream, which contains predomi- umn.[6–9] Catalytic distillation uses the heat of reaction
nantly DIPB, goes to the DIPB column. If the propy- directly to supply heat for distillation of the reaction
lene feed contains an excessive amount of butylenes, mixture. This concept has been applied commercially
or if the benzene feed contains an excessive amount for producing not only cumene but also EB and methyl
of toluene, higher levels of butylbenzenes and=or tert-butyl ether (MTBE). The use of a single column
cymenes can be formed in the alkylation reactor. These that performs both the reaction and the distillation
compounds are distilled out and purged from the functions has the potential of realizing substantial
overhead section of the DIPB column. The DIPB savings in capital cost by essentially eliminating the
stream leaves the DIPB column by a sidedraw and is need for a separate reactor section. Unfortunately,
passed to the transalkylation reactor. The DIPB many available zeolite catalysts that are ordinarily very
column bottom stream consists of heavy aromatic effective in promoting alkylation in a fixed-bed
by-products, which are normally blended into fuel environment are much less effective when used in the
oil. Steam or hot oil typically provides the heat require- environment of the catalytic distillation column. Also,
ments of the product fractionation section. a separate fixed-bed finishing reactor may be required
The sidedraw from the DIPB column containing to ensure that complete conversion (100%) of the olefin
mainly DIPB combines with a portion of the recycle occurs to avoid yield losses of propylene to the LPG
benzene and is charged downflow into the transalkyla- product stream. As such, the amount of catalyst and
tion reactor. In the transalkylation reactor, DIPB and physical size of the distillation column may be substan-
benzene are converted to additional cumene. The tially larger than the benzene column used in the
effluent from the transalkylation reactor is then sent conventional fixed-bed process. Thus, the savings
to the benzene column. realized by the elimination of the reactor section may
QZ-2000 or the newer QZ-2001 catalyst can be uti- be more than offset by the increased catalytic
lized in the alkylation reactor while QZ-2000 catalyst distillation column size, catalyst cost, and addition
remains the catalyst of choice for the transalkylation of a finishing reactor. Depending on the relative values
Cumene Production 607

and trade-off of these considerations including utility in the propylene feed, there is always the potential also
values, catalytic distillation may still be an appro- for butylbenzene to form through the oligomerization
priate option for producers in certain circumstances, of propylene to nonene, followed by cracking and C
although market interest in this process option for alkylation to produce butylbenzenes and amylbenzenes.
cumene production has been low in recent years. As a result of having relatively high activity and oper-
ating at relatively low temperature, beta zeolite catalyst
systems tend to eliminate oligomerization. This results
Zeolitic Alkylation Chemistry in essentially no butylbenzene formation other than
that formed from the butylenes in the propylene feed.
The production of cumene proceeds by a modified The cumene product from a beta zeolite based process
Friedel–Crafts alkylation of benzene with propylene. such as the Q-Max process unit fed with a butylene-free
This reaction can be promoted with varying degrees propylene feedstock typically contains less than
of effectiveness using many different acid catalysts. 15 wt ppm butylbenzenes.
The basic alkylation chemistry and reaction mechanism The Q-Max process typically produces near-
are shown in Fig. 4. The olefin forms a carbonium ion equilibrium levels of cumene (between 85 and 95 mol%)
intermediate, which attacks the benzene ring in an elec- and DIPB (between 5 and 15 mol%). The DIPB is
trophilic substitution. The addition of propylene to the fractionated from the cumene and reacted with recycle
benzene ring is at the middle carbon of C3 olefin double benzene at transalkylation conditions to produce addi-
bond, in accordance with Markovnikov’s rule. The tional cumene. The transalkylation reaction is believed
presence of the isopropyl group on the benzene ring to occur by the acid catalyzed transfer of one isopropyl
weakly activates the ring toward further alkylation, group from DIPB to a benzene molecule to form two
producing DIPB and heavier polyalkylate by-products. molecules of cumene, as shown in Fig. 5
Because new high-activity beta zeolite catalysts such Beta zeolite catalyst is also an extremely effective
as QZ-2000 catalyst are such strong acids, they can be catalyst for the transalkylation of DIPB to produce
used at lower reaction temperatures than SPA catalyst cumene. Because of the high activity of beta zeolite,
or other relatively lower-activity zeolites such as transalkylation promoted by beta zeolite can take
MCM-22 catalyst.[16,17] The lower reaction tempera- place at very low temperature to achieve high conver-
ture in turn reduces the olefin oligomerization reaction sion and minimum side products such as heavy aro-
rate, which is relatively high for SPA catalyst. The matics and additional n-propylbenzene as highlighted
result is that beta zeolite catalysts tend to have higher in Fig. 6. Virtually no tri-isopropyl benzene is pro-
selectivity to cumene and lower selectivity to both duced in the beta system owing to the shape selectivity
nonaromatics that distill with cumene (such as olefins, of the three-dimensional beta zeolite structure, which
which are analyzed as Bromine Index, and saturates) inhibits compounds heavier than DIPB from forming.
and heavy by-products. For example, although butyl- As a result of the high activity and selectivity
benzene is typically produced from traces of butylene properties of beta zeolite, a beta zeolite based catalyst

Fig. 4 Alkylation chemistry. (View this art in color at www.dekker.com.)

608 Cumene Production

Fig. 5 Transalkylation chemistry. (View this art in color at www.dekker.com.)

(e.g., QZ-2001 catalyst or QZ-2000 catalyst) is specified the Q-Max process can be at least 99.7 wt% or higher.
for both the alkylation and the transalkylation sections Because the Q-Max process uses small, fixed-bed reac-
of the Q-Max process. tors and carbon–steel construction, the erected cost is
With both alkylation and transalkylation reactors relatively low. Also, because the QZ-2001=QZ-2000
working together to take full advantage of the catalyst system is more tolerant of feedstock impurities
QZ-2001=QZ-2000 catalyst system, the overall yield (such as water, p-dioxane, sulfur, etc.) compared to
of cumene based on benzene and propylene feed in other catalysts available, the Q-Max process requires

Fig. 6 Possible alkylation side reactions. (View this art in color at www.dekker.com.)
Cumene Production 609

minimal pretreatment of the feeds, which further  Significant reduction in undesirable polymerization
minimizes the capital costs. and by-product side reactions.
This is in distinct contrast to other technologies  High tolerance to feedstock impurities and poisons. C
based on zeolites other than beta where extensive feed
contaminant guard beds are required to protect the The resistance of new zeolitic catalysts to temporary
catalyst from rapid and precipitous deactivation and and permanent catalyst poisons is essential to the eco-
loss of conversion when exposed to trace amounts of nomic and commercial success of a zeolitic based
sulfur, water, oxygenates, and nitrogen. cumene process. The following commercial data
obtained using beta zeolite as a catalyst illustrates the
Cumene Product Impurities outstanding ability of beta zeolite to cope with a wide
range of feedstock contaminants:
Beta zeolite catalyst can be optimized to nearly elimi-
nate all undesirable side reactions in the production
Cyclopropane
of cumene. The improvement in beta zeolite catalyst
quality has occurred to the point that any significant
n-Propylbenzene (nPB) is formed by alkylation of
impurities in the cumene product are governed largely
benzene with cyclopropane or n-propanol, and by
by trace impurities in the feeds. The selectivity of the
anti-Markovnikov alkylation of benzene with propy-
catalyst typically reduces by-products to a level result-
lene. Cyclopropane is a common impurity in propylene
ing in production of ultrahigh cumene product purities
feed and approximately half of this species is converted
up to 99.97 wt%. At this level, the only significant by-
to nPB in the alkylation reactor. Essentially, all alkyla-
product is n-propylbenzene with the catalyst producing
tion catalysts produce some nPB by anti-Markovnikov
essentially no EB, butylbenzene, or cymene beyond
alkylation of propylene. The tendency to form nPB
precursors in the feed. Fig. 7 shows the reactions of
rather than cumene decreases as the reaction tempera-
some common feedstock impurities that produce these
ture is lowered, making it possible to compensate for
cumene impurities.
cyclopropane in the feed to some extent. As the operat-
ing temperature of zeolitic alkylation catalyst is
Beta Catalyst Resistance to decreased, the deactivation rate increases. However,
Feed Contaminants because of the exceptional stability of the beta zeolite
catalyst system, a unit operating with beta zeolite
Beta zeolite is a large pore zeolite with optimized acid catalyst can be operated for extended cycle lengths
site densities that exhibits: and still maintain an acceptable level of nPB in the
cumene product. For example, with FCC-grade propy-
 Good mass transfer properties. lene feed containing typical amounts of cyclopropane,

Fig. 7 Reactions of feed impurities. (View

this art in color at www.dekker.com.)
610 Cumene Production

a beta zeolite based process can produce an overall by-products that boil within the boiling range of cumene,
cumene product containing 250–300 wt ppm nPB while the sulfur content of the cumene product depends on the
achieving an acceptable catalyst cycle length. sulfur content of the propylene and especially benzene
feeds to a certain extent. Sulfur at the levels usually
Water present in propylene and benzene feeds considered for
cumene production will normally result in cumene
Water can act in this environment as a Bronsted base product sulfur content that is within specifications.
to neutralize some of the weaker zeolite acid sites. This
effect is not harmful to any appreciable extent to the
Unsaturates
beta zeolite catalyst at typical feed stock moisture
levels and under normal alkylation and transalkylation
Use of beta zeolite catalyst does not require the
conditions. This includes processing of feedstocks up
benzene feed to be clay treated prior to use in alkyla-
to the normal water saturation condition (typically tion service. Some of the unsaturated material in the
500–1000 ppm) resulting in 10–150 ppm water in the
benzene can lead to the formation in the alkylation
feed to the alkylation reactor dependent on feed
reactors of polycyclic-aromatic material which will
and=or recycle stream fractionation efficiency.
get preferentially trapped in some zeolites having rela-
tively small-sized pores. This can lead to increased
Oxygenates deactivation rates in such small-pore zeolites. Beta
zeolite’s large pore structure makes it possible to more
Small quantities of methanol and ethanol are some- easily handle this polycyclic-aromatic material and as a
times added to the C3s in pipelines to protect against result does not require further treatment of the benzene
freezing because of hydrate formation. Although the feed to remove unsaturated material. In addition,
beta zeolite catalyst is tolerant of these alcohols, alpha-methylstyrene (AMS) is produced by alkylation
removing them from the feed by a water wash may still of benzene with methylacetylene or propadiene. Some
be desirable to achieve the lowest possible levels of EB of the AMS alkylates with benzene, forming diphenyl-
or cymene in the cumene product. Cymene is formed propane, a heavy aromatic that leaves the unit with the
by the alkylation of toluene with propylene. The DIPB column bottoms.
toluene may already be present as an impurity in the
benzene feed, or it may be formed in the alkylation
reactor from methanol and benzene. Ethylbenzene is Nitrogen, Metal Cations, and Arsine
primarily formed from ethylene impurities in the
propylene feed. However, similar to cymene, EB can The presence of trace amounts of organic nitrogen
also be formed from ethanol. compounds and metal cations in the benzene feed or
p-Dioxane is sometimes present in benzene from arsine in the olefin feed has been known to neutralize
extraction units that use ethylene glycol based solvents. the acid sites of any zeolite catalyst. Good feedstock
It is reported to cause deactivation in some zeolitic alky- treating practice or proven guard-bed technology
lation catalysts even at very low ppm levels. However, easily handles these potential poisons. For example,
beta zeolite catalyst appears to be tolerant to p-dioxane basic nitrogen, which can sometimes be present in the
at levels typically found in benzene extraction processes benzene fresh feed, is easily removed using very low-
and does not require costly removal of this impurity. cost UOP guard-bed technology. To facilitate monitor-
ing of feeds for potential nitrogen based contaminants,
UOP has developed improved analytical techniques to
Sulfur help in the evaluation. These methods include UOP
971 (‘‘Trace Nitrogen in Light Aromatic Hydrocar-
Sulfur has no significant effect on beta zeolite catalyst bons’’ by Chemiluminescence) used to detect total
at the levels normally present in olefin and benzene nitrogen down to 30 ppb and UOP 974 (‘‘Nitrogen
feeds considered for cumene production. However, Compounds in Light Aromatic Hydrocarbons’’ by
even though the beta zeolite catalyst is sulfur tolerant, GC) used to detect individual nitrogen species down
trace sulfur that makes its way into the finished to about 100 ppb.
cumene unit product may be a feed quality concern
for downstream phenol processors where the typical
sulfur specification is <1 ppm. The majority of sulfur CURRENT STATE-OF-THE-ART
compounds associated with propylene (mercaptans) CUMENE TECHNOLOGY
and those associated with benzene (thiophenes) are
converted to by-products outside the boiling range Currently, the major processes for cumene production
of cumene. Because some sulfur compounds form are liquid phase technologies offered by UOP and
Cumene Production 611

Fig. 8 Effect of framework Al on beta

catalyst stability. (View this art in color
at www.dekker.com.)

Badger Licensing (ExxonMobil technology) based on toward oligomerization and were more prone toward
beta zeolite (such as QZ-2001 and QZ-2000 catalysts) coking.
and MCM-22 catalyst, respectively. Over the past This study parallels work performed at UOP, where,
decade, great progress has been made in improving through the use of nonconventional synthesis techni-
and optimizing catalyst formulations for use in applica- ques, samples have also been prepared with Si=Al2
tions to produce cumene from benzene alkylation. For ratios as low as 10. Through this work it has been
example, the ability to synthesize beta zeolite in a wide found that with a Si=Al2 ratio of 25, the catalyst
range of Si=Al2 ratios has given catalyst designers the maintains sufficient activity to achieve polyalkylate
ability to tailor the zeolite into a form that optimizes equilibrium (e.g., DIPB equilibrium) and, at the same
activity and selectivity. A parametric study on the time, minimizes the formation of heavier diphenyl com-
effects of Si=Al2 ratio on activity and selectivity was pounds (and hence maximizes yield) in cumene service.
published by Bellusi.[18] In this work, it was found that Perhaps the most critical understanding was devel-
as the silica to alumina ratio was increased from 28 to oped with regard to the need to minimize the Lewis
70, there was a decrease in both activity and selectivity acidity of the catalyst and at the same time maintain
toward isopropylbenzene-type compounds. Addition- high Brønsted acidity. Studies at UOP demonstrated
ally, the less active catalysts had a greater tendency that olefin oligomerization was directly related to the

Fig. 9 Stability of QZ-2001 catalyst vs. QZ-2000

catalyst. (View this art in color at www.dekker.-
com.)
612 Cumene Production

Fig. 10 JLM commercial data

on catalyst stability. (View this
art in color at www.dekker.com.)

Lewis acid function of the catalyst. Olefin oligomeriza- from other commercially available zeolite catalysts such
tion reactions can lead to the formation of heavy com- as MCM-22 catalyst, where significant activity and selec-
pounds (coke type precursors), which have a negative tivity can be lost upon regeneration.[19] The ability to
effect on catalyst stability. Thus, minimization of the regenerate catalyst is essential in a commercial environ-
Lewis character of the beta leads to a catalyst with high ment to provide additional flexibility to cope with a wide
stability. Generally, Lewis acidity in beta zeolite has range of feedstock sources, feedstock contaminants, and
been attributed to the existence of nonframework alu- potential operational upsets.
minum atoms. The most common mechanism for the The historical development of beta zeolite showed
formation of nonframework alumina is through steam that early versions of beta catalyst demonstrated less
dealumination during the catalyst calcination step of than optimum performance when compared to today’s
the manufacturing process. By careful control of the state-of-the-art formulation. Fig. 8 is a plot of the rela-
temperature, time, and steam levels during the manu- tive stability of beta zeolite as a function of the Si=Al2
facturing process, it is possible to produce a catalyst ratio of the beta zeolite structure in which the dominat-
that is extremely stable at typical alkylation conditions. ing influence of this parameter is evident. Uop has
From a commercial standpoint this knowledge has learned to stabilize the zeolite structure through careful
had the additional benefit of developing a regeneration process and chemical means. This has resulted in
protocol that is extremely robust. It has been demon- a catalyst system that is extremely robust, highly
strated in commercial in situ and ex situ procedures that regenerable, and tolerant of most common feedstock
the beta zeolite catalyst can be regenerated with excellent impurities.
results providing complete restoration of fresh catalyst Additional studies of beta zeolite have come to similar
performance. The feature of complete regenerability is conclusions. For example, Enichem has found that beta
another attribute that distinguishes beta zeolite catalysts zeolite is the most effective catalyst for cumene alkylation

Fig. 11 JLM commercial data on sulfur in propylene

feed. (View this art in color at www.dekker.com.)
Cumene Production 613

Fig. 12 JLM commercial data on alkylation reactor

moisture content. (View this art in color at www.
dekker.com.)

among other zeolites tested including Y, mordenite, and Refinery grade, chemical grade, or polymer grade pro-
an isostructural synthesis of MCM-22 catalyst.[20] pylene feedstocks have been successfully used to make
The principles described above also led to the high-quality cumene product in the Q-Max process.
development of the new-generation QZ-2001 alkylation A good example of the ruggedness of the beta
catalyst.[21] In Fig. 9, results from accelerated stability zeolite catalyst can be found in the case of JLM’s Blue
testing of QZ-2000 and QZ-2001 catalysts demonstrate Island (Illinois) Q-Max operation. The operation
the superior stability of the latest catalyst system. QZ- started in August 1996 as the first Q-Max process
2001 catalyst exhibits as much as twice the stability when operation with UOP beta zeolite catalyst. Initial oper-
compared to QZ-2000 catalyst. The benefit afforded by ating results were reported in 1997.[22] The unit has
QZ-2001 catalyst can be utilized by cumene producers continued to operate with stable performance for more
in several ways. It can be taken directly through reduc- than 7 yr without catalyst regeneration in spite of the
tion of the catalyst loading for a specific run length, presence of significant levels of feed contaminants.
or alternatively as a convenient way to increase run Excellent monoalkylation selectivity has also been
length, or to increase throughput through the cumene observed over many years of service in the JLM opera-
unit, by allowing operation at a lower B=P feed ratio. tion as shown in Fig. 10. Under the normal operating
conditions of the unit, an equilibrium cumene selectivity
of about 91 mol% is predicted. Thus, results clearly show
Long-Term Stability of Beta Zeolite Catalyst that the beta zeolite catalyst is active enough to achieve
near-equilibrium selectivity. This is an important feature
Commercial operation with a wide variety of olefin of the catalyst as the amount of dialkylate that must be
feedstocks from different sources has demonstrated processed in the transalkylator and the subsequent cost
the flexibility of beta zeolite in the Q-Max process. of processing this material are minimized.

Fig. 13 JLM commercial data on benzene feed

p-dioxane content. (View this art in color at www.
dekker.com.)
614 Cumene Production

Fig. 14 JLM commercial data on cumene product purity. (View this art in color at www.dekker.com.)

Feed contaminants such as sulfur, water, and no impact on catalyst stability or performance, as seen
p-dioxane were monitored very closely during the first in Figs. 10 and 15. Note that the alkylation catalyst
2 yr onstream as shown in Figs. 11–13. selectivity and catalyst bed inlet temperature and
The impact of these impurities on cumene product weight average bed temperature remain virtually
quality was negligible during this period. The customer unchanged after years of operation.
was able to maintain extremely high-quality cumene
throughout the life of the beta catalyst, as shown in
Fig. 14. Beta Zeolite Catalyst Regeneration
Another important observation is that sulfur levels
from 3 to 7 ppm, moisture levels of 30–70 ppm, and As a result of beta’s high activity and robustness,
even p-dioxane excursions up to 70 ppm had virtually catalyst requirements are minimized. At the end of

Fig. 15 JLM commercial data. (View this art in color at www.dekker.com.)

Cumene Production 615

Fig. 16 Commercial Q-Max process data. Effect of in situ QZ-2000 catalyst regeneration on cumene selectivity. (View this art in
color at www.dekker.com.)

each cycle, the catalyst can be regenerated ex situ by a process unit can be designed to accommodate in situ
certified regeneration contractor. The regenerability of catalyst regeneration. Both options have been success-
beta zeolite catalyst provides an ultimate life of three fully demonstrated in commercial operation.
cycles or more if appropriate processing and regenera- Figs. 16 and 17 show an example of the perfor-
tion guidelines are followed. If desired, the Q-Max mance of QZ-2000 catalyst after multiple in situ

Fig. 17 Commercial Q-Max

process data. Effect of in situ
QZ-2000 catalyst regeneration
on cumene purity. (View this
art in color at www.dekker.com.)
616 Cumene Production

regenerations in a commercial cumene unit. The 9. Shoemaker, J.D.; Jones, E.M., Jr. Cumene by
cumene unit experienced premature deactivation catalytic distillation. Hydrocarbon Process 1987,
because of excessive basic nitrogen levels in the feed 66 (6), 57.
as well as unsatisfactory plant operation. 10. Canfield, R.C.; Unruh, T.L. Improving cumene
Taking advantage of the in situ regeneration yields via selective catalysis. Chem. Eng. 1983,
capability, the customer opted to regenerate the cata- 90 (6), 32.
lyst three times during this period. The results show 11. Canfield, R.C.; Cox, R.C.; McCarthy, D.M.
the remarkable resilience of the beta zeolite catalyst Monsanto=Lummus crest process produces low-
to the stresses of regeneration with virtually no loss est cost cumene. In AICHE 1988 Spring Meeting,
in monoalkylate selectivity or cumene product quality New Orleans, LA, Apr 6–10, 1986.
as a result of repeated regenerations. 12. Wadlinger, R.L.; Kerr, G.T. U.S. Patent
3,308,069, Mar 6, 1967. Mobil Oil Corporation.
13. Innes, R.A.; Zones, S.I.; Nacamuli, G.J. Liquid
CONCLUSIONS Phase Alkylation or Transalkylation Process
Using Zeolite Beta. U.S. Patent 4,891,458, Jan 2,
Fixed-bed zeolitic cumene technology is the process of 1990. Chevron Research and Technology
choice for the cumene=phenol industry. Most of the Company: San Francisco, CA.
previously installed cumene capacity based on older 14. Cannan, T.R.; Hinchey, R.J. Synthesis of Zeolite
AlCl3 and SPA technologies has now been replaced Beta. U.S. Patent 5,139,759, Aug 18, 1992. UOP:
with the newer zeolitic technology over the past 10 yr. Des Plaines, IL.
The result is greatly improved yields and reduced oper- 15. Inwood, T.V.; Wight, C.G.; Ward, J.W. Liquid-
ating and capital costs. UOP’s Q-Max process based on phase Alkylation and Transalkylation Process.
beta zeolite has emerged as the leader for cumene U.S. Patent 4,459,426, Jul 10, 1984, Union Oil.
technology. This is primarily due to the high activity, 16. Cheng, J.C., et al. A comparison of zeolites
robustness, and lower operating costs associated with MCM-22, beta, and USY for liquid phase alkyla-
operating a beta zeolite based catalyst system. tion of benzene with ethylene. Sci. Technol. Catal.
1998, 6, 52–60.
17. Cheng, J.C., Smith, C.M.; Venkat, C.R.; Walsh,
REFERENCES D.E. Continuous Process for Preparing Ethylben-
zene Using Liquid Phase Alkylation and Vapor
1. Bentham, M., et al. Process improvements for a Phase Transalkylation. U.S. Patent 5,600,048,
changing phenol market. In DeWitt Petro- Feb 4, 1997, Mobil Oil Corporation,
chemical Review Conference, Houston, TX, Mar 18. Bellussi, G., et al. Liquid phase alkylation of
19–21, 1991. benzene with light olefins catalyzed by beta
2. Stefanidakis, G.; Gwyn, J.E. In Encyclopedia of zeolites. J. Catal. 1995, 157, 227–234.
Chemical Processing and Design; McKetta, J.J., 19. Dandekar, A.B., et al. Regeneration of Aromatic
Cunningham, W.A., Eds.; Marcel Dekker: New Alkylation Catalysts Using Hydrocarbon Strip-
York, 1977; Vol. 2, 357. ping. WO 01=83408, Nov 2001. Mobil Oil Corp,
3. Keim, W.; Roper, M. In Ullmann’s Encyclopedia Baytown, TX.
of Industrial Chemistry; Gerhartz, W., Ed.; VCH 20. Perego, C., et al. Experimental and computational
Verlagsgesellschaft: Weinheim, 1985; Vol. A1, 185. study of beta, ZSM-12, Y, modernite and ERB-1
4. Sedaglat-Pour, Z. Cumene; CEH Data Summary; in cumene synthesis. Microporous Mater 1996, 6,
SRI International: Menlo Park, CA, Mar 1989, 395–404.
638.5000A. 21. Schmidt, R.; Zarchy, A.; Petersen, G. New
5. Chem. Mark. Rep. 1987, 232 (10), 54. developments in cumene and ethylbenzene alkyla-
6. Shoemaker, J.D.; Jones, E.M., Jr. Cumene by tion—Paper 124b. In 1st Annual Aromatic
catalytic distillation. In 1987 NPRA Annual Producers Conference—AIChE Spring Meeting,
Meeting, Mar 29–31, 1987. New Orleans, LA, Mar 10–14, 2002.
7. Jones, E.M., Jr.; Mawer, J. Cumene by catalytic 22. Jeanneret, J.; Greer, D.; Ho, P.; McGehee, J.;
distillation, In AICHE Meeting, New Orleans, Shakir, H. The UOP Q-Max process: setting the
LA, Apr 6–10, 1986. pace for cumene production, Presented at the
8. Stadig, W.P. Cumene. Chem. Process. 1987, 50 (2), 22nd Annual De Witt Petrochemical Review,
27. Houston, TX, Mar 18–20, 1997.