About Chemical Innovation - Subscription Information
March 2001
Vol. 31, No. 3, pp 15–21.
Enabling Science

Table of Contents

Aspi K. Kolah
Qi Zhiwen
Sanjay M. Mahajani

Dimerized isobutene:
An alternative to MTBE

Methyl tert-butyl ether (MTBE) has been banned as a gasoline additive in California as of 2002 (1) and New York is likely to follow suit (2). Groundwater contamination associated with MTBE has forced refiners all over the world to find a viable alternative. The current demand for MTBE and the forecast for the next 5 years (Figure 1) give a clear idea of this impact.

Figure 1. Worldwide MTBE production and projected future demand.
Figure 1. Worldwide MTBE production and projected future demand. Source: Reference 1. Used with permission.
Worldwide production of biologically derived ethanol is not likely to fulfill the demand created by the MTBE phaseout. Although the other fuel additive ethers such as ethyl tert-butyl ether (ETBE) and tert-amyl methyl ether (TAME) are alternatives to MTBE, they are not likely to be favored by refineries, lest they “go the MTBE way”. MTBE has been a major consumer of isobutene from the C4 hydrocarbon stock, and its phaseout will cause a major decline in the downstream consumption of isobutene from fluid catalytic cracking (FCC) and steam cracking product streams. Therefore, an additive that comes from isobutene is likely to have an edge over the non-C4 candidates such as TAME, ethanol, and isomerates (highly branched alkanes derived from pentanes and hexanes).

Alkylates are branched paraffins with carbon numbers in the range C8–C12 (isooctane is one example). Alkylates from C4 streams possess the high octane numbers required for gasoline additives and are expected to find a prominent place in the gasoline additives market. The average octane number for alkylates is in the range of 93–96. (This number is obtained by taking an arithmetic average of the research octane number and the motor octane number.)

Although alkylates have the potential to match most of MTBE’s properties, their boiling range (70–150 °C) is substantially higher than the boiling point of MTBE (52 °C). An appropriate blend of C5–C6 isomerates with alkylates would match the volatility of gasoline and be a perfect replacement for MTBE. However, the mandatory minimum 2% oxygen requirement in gasoline, as stipulated by the 1990 Clean Air Act (3), would have to be relaxed for this blend to be used in the United States.

Two processes are available for producing alkylates from C4 streams. The first, which has been practiced industrially for several years, is direct alkylation of isobutane with 1-butene in the presence of hydrofluoric acid, sulfuric acid, or solid super acid catalysts. The various aspects of this reaction have been reviewed recently (4, 5).

The second route, indirect alkylation, is carried out in two steps: selective dimerization of isobutene (from C4 streams) to form diisobutene (DIB), followed by hydrogenation to form the saturated product isooctane. Selectivity problems and catalyst deactivation hinder the isobutene dimerization reaction. Because this reaction decides the quality and properties of the alkylates formed, it is a crucial step in this process. The second step, olefin hydrogenation, is a well-established process. It uses either noble metal or non-noble metal catalysts, depending on the feed quality. Non-noble metal catalysts are less prone to deactivation and fouling, but require high hydrogen partial pressures, and hence, higher operating and capital costs.

The indirect alkylation process can be commercialized easily by revamping existing MTBE units. The other merits of indirect alkylation are described in the next section. Although companies such as UOP LLC, Institut Français du Pétrole (IFP), Snamprogetti (Italy), Fortum Oil (Finland), and Gas Oy (Finland) have licensed commercial technologies for synthesizing alkylates, there is not much information available in the open literature. In this article, we review the information on dimerization available in patents and the literature to set a basis for future research in this challenging area.

Direct vs. indirect alkylation
The direct alkylation process uses a mixture of isobutane and n-butenes, but not isobutene, as a feedstock. The process is limited because it cannot use the C4 stream coming from steam reforming, which contains only 4–8% isobutane (Table 1).
Table 1
Composition of various C4 product streams

Fluid catalytic cracking, % Steam cracking, % Dehydrogenation, % Concentrated isobutene, % a
Paraffins (isobutane and n-butane) 30–60 4–8 45–55 <10
n-Butene 25–50 35–60 0 <10
Isobutene 10–25 30–46 45–55 >90
aReferences 10 and 20.
In contrast, indirect alkylation has wider feedstock flexibility because the main feedstock is isobutene from C4. Adding dehydrogenation and isomerization units into the indirect alkylation complex can be beneficial, making it possible to use n-butane and isobutane present in the C4 stream (4).

The direct alkylation process mainly uses homogeneous strong acids such as HF or H2SO4. Although the process has been used for several years, concern is growing over a possible catastrophic release of HF from these units. These acids are environmentally hazardous and pose waste disposal and corrosion problems. The recent trend is to explore the use of solid acids for this reaction (5). One advantage of dimerization by indirect alkylation is that it can be performed conveniently under relatively mild conditions in the presence of heterogeneous catalysts such as ion exchange resins (IER) and solid phosphoric acid (SPA), which are safer than their homogeneous counterparts.

The alkylate produced by direct alkylation has an aver age octane number between 90 and 93, which is inferior to the octane number (>95) of the alkylate produced by indirect alkylation.

Process for indirect alkylation
A generalized flow diagram for the conventional commercial process is shown in Figure 2. The C4 stream, consisting mainly of isobutane, n-butane, isobutene, and n-butenes, is fed to the dimerization reactor, where isobutene is dimerized selectively in the presence of an acid catalyst.
Figure 2. Conventional process for producing alkylates by indirect alkylation.
Figure 2. Conventional process for producing alkylates by indirect alkylation.
Sometimes this indirect alkylation process is designed to dimerize n-butenes to some extent with a proper choice of catalyst. This is necessary to match the balance of isobutane and n-butenes in the outgoing C4 stream. The isobutane in the C4 stream is normally alkylated with the n-butenes using direct alkylation; the proper ratio of isobutane to n-butenes is essential for complete conversion and utilization of the C4 stream.

The reaction is exothermic, and heat must be removed to avoid temperature rises that can lead to the formation of undesired oligomers. These oligomers have relatively high molecular weights and boiling points and are not suitable as gasoline blends; they also rapidly deactivate the catalyst. Depending on the catalyst, an appropriate solvent may be needed to increase the selectivity toward the dimers. The product stream from the reactor is fed to a distillation column, where dimerized and heavy products are separated from the unreacted C4 components and solvent (if any). The dimer is then saturated in a separate reactor to form alkylates in the presence of a hydrogenation catalyst.

Figure 3
Figure 3. Reaction scheme for the oligomerization of isobutene.
Reaction scheme and mechanism
The reaction scheme for dimerization of isobutene is shown in Figure 3, and the reaction mechanism is shown in Figure 4. The reaction is irreversible and highly exothermic (δH = –19.8 kcal/mol). As in all acid-catalyzed reactions, tertiary carbocations are formed as a result of the reaction of isobutene with catalytic protons. The carbocation then reacts with another molecule of isobutene to form an unstable complex, which on release of a proton may form two isomers: 2,4,4-trimethyl-1-pentene or 2,4,4-trimethyl-2-pentene. The mixture of these two isomers, commonly called DIB, can further oligomerize to C12 (trimer), C16 (tetramer), or higher oligomers in a similar fashion. The experimental evidence, however, suggests that the trimers and tetramers can form directly from isobutene through parallel reactions. The tertiary carbocation may also react with n-butene, which leads to the formation of other isomers of DIB; however, this reaction needs relatively harsh conditions. The typical concentration profiles of all the components in a single batch reaction (Figure 5), exhibit the dynamic progress of the reacting system (6).

Figure 4
Figure 4. Reaction mechanism for the oligomerization of isobutene.

Figure 5
Figure 5. Typical concentration profiles for the oligomerization of isobutene.
Factors influencing dimerization kinetics
Much of the work on dimerization was carried out either before 1980 (6, 7) or in the late 1990s (8–13). The gap of almost two decades may be attributed to the dominance of MTBE as a gasoline additive during this time. Table 2, a survey of important published studies on isobutene oligomerization, gives a fair idea of the influence of different parameters on the reaction performance.

Catalysts. As mentioned earlier, the reaction is acid catalyzed, and homogeneous or heterogeneous catalysts can be used. The heterogeneous catalysts offer several engineering benefits and are preferred. Catalysts, such as strong acid IERs, SPA, and aluminosilicates, have been studied in the past. Some patents also claim the possibility of using heteropolyacids, sulfonated polysilanes, and similar acids (14). The commercial processes use either IER or SPA.

The choice of catalyst is a crucial aspect of production because undesired impurities in the feedstock and the formation of side products can foul the catalyst. Typically, SPA catalyst has a 12–16-month life range. A pretreatment to remove impurities such as sulfur and moisture from the feed stock is desirable to maintain reasonably high catalytic activity.

Most existing MTBE units use IER catalysts, and any strategy for revamping existing MTBE units must consider the characteristics of these catalysts. In general, using IER catalyst leads to the formation of impurities, and this catalyst is prone to a higher degree of fouling than SPA. Moreover, IER cannot codimerize butene because of either low strength or thermal limitations. IER catalysts generally need a diluent to obtain high selectivity toward the dimer. The flexibility of a plant to produce MTBE or alkylate, depending on the market forces, may give IER an edge. An excellent comparison of the processes with these two catalysts has been presented in Reference 4. The economics of refinery MTBE conversion to indirect alkylation suggest that SPA is superior to IER because of the low capital cost, high flexibility due to n-butene conversion, and relatively longer catalyst life.

The vapor-phase oligomerization of isobutene (and other C3–C4 olefins) has also been studied extensively using zeolites such as ZSM-5 and [H]-mordenite as catalysts (15–18). Zeolite catalysts exhibit shape selectivity, producing the dimer with the required structure; however, these extremely active catalysts undergo very fast deactivation. This is the main reason zeolites may lag behind other solid acids as suitable catalysts for indirect alkylation.

Temperature. The operating temperature, which must be chosen carefully, is determined mainly by the kind of catalyst used. A high temperature under otherwise similar conditions increases the rate of the reaction, but also increases the formation of undesired higher oligomers. If the reaction is performed in a gas–liquid mode at relatively low pressure, by dissolving isobutene in a proper solvent, the increase in temperature adversely affects the reaction because the solubility of isobutene in the solvent drops. IER-catalyzed reactions are performed at relatively lower temperatures than those used for SPA-catalyzed dimerizations (Table 2).

Solvent. The presence of solvent in the reaction mixture is very important for this reaction. The selectivity toward dimer can be influenced significantly with a proper choice and proportion of a solvent. The solvent molecules cover the active catalytic sites and favorably influence the relative reaction rates of main and side reactions. As can be seen from Table 2, inert solvents (cyclohexane, isooctane, other paraffins) and solvents that are reactive under the operating conditions (H2O–tert-butyl alcohol and MeOH–MTBE) have been investigated, and the results are encouraging with regard to the selectivity toward the dimer. The choice and necessity of the solvent depend on the type of catalyst. The SPA catalyst process works well even in the absence of solvent.

The studies by Cunill and co-workers (14, 19) and by the group from Snamprogetti research laboratory (10, 20) deserve a special mention here. These groups have investigated the presence and influence of MTBE and MeOH, and these studies are important for retrofitting existing MTBE units. The rate of reaction decreases in the presence of MeOH–MTBE (19). At low MeOH/isobutene ratios, oligomer formation is favored over MTBE formation. The optimum ratio of ~0.6 is recommended because an extremely low ratio results in formation of undesired more highly polymerized oligomers, leading to poor selectivity (10).

Choice of reactor. Plug-flow type reactors are preferred for commercial production of the dimers because consecutive series reactions that may form higher oligomers have to be suppressed. In view of this, a fixed-bed reactor with proper temperature control can be a good choice for dimerization. The reactor should be designed to control the conversion of isobutene and avoid the formation of higher oligomers that may result from the increased concentration of dimer in the reactor. The conventional process separates the unreacted isobutene in a downstream distillation column and recycles it back to the reactor.

A potentially important choice for such a reaction is the use of catalytic distillation (CD)—a novel multifunctional reactor design that combines reaction and distillation in a single unit (21). The potential of these reactors and the design and modeling aspects are detailed in References 22 and 23. Capital and recycle costs are reduced significantly, and the per-pass yield toward dimer is enhanced significantly. Because dimer in CD can be separated conveniently during the reaction as soon as it forms, its concentration can be maintained at a low level in the reaction zone, and consecutive reactions can be suppressed.

MTBE revamp
The outlook for MTBE in gasoline is extremely uncertain. In light of this, refineries in the United States and elsewhere are looking at cost-effective solutions for revamping MTBE manufacturing units, often in collaboration with process licensing companies. UOP LLC offers options that use indirect alkylation technology (InAlk process) with either SPA or resin catalysts. They claim a payback period of less than 2 years with their SPA catalyst. An additional oxygenate recovery unit is essential if resin catalyst is used.

The SP–Isoether process is commercially available from Snamprogetti. This technology converts their water-cooled tubular reactors using resin catalyst to produce either diisobutene or MTBE, depending on the market demand (24). This process offers a low-risk option for revamping MTBE plants. The Selectopol process by IFP is a retrofit option for existing MTBE units using SPA catalyst. The NExOCTANE process for isobutene dimerization is yet another commercially available technology from Fortum Oil and Gas Oy operating jointly with Kellogg Brown and Root. This process uses readily available standard commercial catalysts. These commercial technologies are easily located in the literature (23–27), and similar technologies from other MTBE licensors may exist.

What’s next?
The potential for a worldwide ban on MTBE has produced uncertainties in the international gasoline market, which should catalyze research on identifying and developing alternatives to MTBE in gasoline. Alkylates are a promising alternative, and alkylate process development will attract increasing attention from researchers and technologists in the years to come. The gray areas in this field are the research on reaction engineering and catalysis.

There is an increasing need to develop a kinetic rate equation that adequately describes the behavior of the reacting system, especially with regard to the formation of undesired higher oligomers. A proper solvent that can render higher selectivity and a catalyst that is less susceptible to fouling and deactivation must be identified. Catalytic distillation can bring compactness and cost effectiveness to the entire process. More exploration is necessary on the possibility of carrying out dimerization and hydrogenation in a single vessel with suitable location of the respective catalytic zones.


  1. U.S. MTBE Restrictions Seem Likely. Oil and Gas Journal, Oct 9, 2000, p 52.
  2. Governor [George E.] Pataki signs legislation to ban MTBE in New York. Press release, May 24, 2000, www.state.ny.us/governor/press/year00/may24_00.htm (accessed Jan 2001).
  3. 42 USC 7401-7671q, P.L. 101-549, 104 Stat. 2399, www.epa.gov/oar/caa/contents.html (accessed Jan 2001).
  4. Meister, J. M.; Black, S. M.; Muldoon, B. S.; Wei, D. H.; Roeseler, C. M. Hydrocarb. Process. 2000, 79 (5), 63–75.
  5. Kundu, B.; Mukhopadhyay, S. Chem. Weekly (India) 2000, 46 (9), 139–146.
  6. Haag, W. O. Chem. Eng. Prog. Symp. Ser. 1967, 63, 140–147.
  7. Scharfe, G. Hydrocarb. Proc. 1973, 53 (4), 171–173.
  8. Bowman, W. G.; Stadig, W. P. U.S. Patent 4,100,220, 1978.
  9. Evans, T. I.; Karas, L. J.; Rameswaran, R. U.S. Patent 5,877,372, 1999.
  10. Girolamo, M. D.; Lami, M.; Marchionna, M.; Pescarollo, E.; Tagliabue, L.; Ancillotti, F. Ind. Eng. Chem. Res. 1997, 36, 4452–4458.
  11. Stine, L. O.; Muldoon, B. S.; Gimre, S. C.; Frame, R. R. U.S. Patent 5,895,830, 1999.
  12. Stine, L. O.; Muldoon, B. S.; Gimre, S. C.; Frame, R. R. U.S. Patent 6,080,903, 2000.
  13. LePage, J. F.; Cosyns, J.; Miquel, J.; Juguin, B. U.S. Patent 4,324,646, 1980.
  14. Izquierdo, J. F.; Vila, M.; Tejero, J.; Cunill, F.; Iborra, M. Appl. Catal., A 1993, 106, 155–165.
  15. Tabak, S. A.; Krambeck, F. J.; Garwood, W. E. AIChE J. 1986, 32, 1526– 1531.
  16. Ngandjui, L.M.T.; Thyrion, F. C. Ind. Eng. Chem. Res. 1996, 35, 1269– 1274.
  17. Kojima, M.; Rautenbach, M. W.; O’Connor, C. T. Ind. Eng. Chem. Res. 1988, 27, 248–252.
  18. Gnep, N. S.; Bouchet, F.; Guisnet, M. R. Prepr.—Am. Chem. Soc., Div. Petr. Chem. 1991, 620–626.
  19. Vila, M.; Cunill, F.; Izquierdo, J. F.; Gonzalez, J.; Hernandez, A. Appl. Catal., A 1994, 117, L99–L108.
  20. De Girolamo, M.; Tagliabue, L. U.S. Patent 6,011,191, 2000.
  21. Vora, B.; Hammershaimb, H. U. U.S. Patent 6,025,533, 2000.
  22. Taylor, R.; Krishna, R. Chem. Eng. Sci. 2000, 55, 5183–5229.
  23. Doherty, M. F.; Buzad, G. Chem. Eng. Res. Design. 1992, 70, 448–458.
  24. www.snamprogetti.it/inglese/isoether/isoether_articolo.htm (accessed Jan 2001).
  25. www.ifpna.com/clean_fuels.html (accessed Jan 2001).
  26. www.uop.com/home/refining/processes_and_products/inalk_intro.htm (accessed Jan 2001).
  27. www.nesteengineering.com/pages/products/ogc/nexoctane.htm (accessed Jan 2001).

Aspi K. Kolah manages R&D and application development at Thermax Ltd., a leading manufacturer and exporter of ion exchange resins and adsorbents (Thermax Ltd., Chemical Division, Pune 411026, India). He works on reactive distillation and specialty applications of ion exchange resins and adsorbents, which include catalysts for etherification, esterification, bisphenol-A synthesis, and downstream processing for the pharmaceutical and biotechnology industry. He received his Ph.D. from the University Department of Chemical Technology, Bombay, India. He then pursued a postdoctoral research assignment at the Max Planck Institut für Dynamik Komplexer Technischer Systeme in Magdeburg, Germany.

Qi Zhiwen has worked as a postdoctoral research fellow at the Max Planck Institute for the past two years (Max Planck Institut für Dynamik Komplexer Technischer Systeme, D-39210, Magdeburg, Germany). He is a specialist in simulating industrial reactive distillation systems. His research areas encompass the design and simulation of reactive distillation and other multifunctional reactors. He received his Ph.D. from East China University of Science and Technology, Shanghai.

Sanjay M. Mahajani is an assistant professor at the Indian Institute of Technology, Bombay (Department of Chemical Engineering, Mumbai 400 076, India; +91-22-5767246; fax: +91-22-5726895; sanjaym@che.iitb.ernet.in). His principal research emphasis is on reaction engineering, catalysis, and multifunctional reactors. He received his B.S. degree and his Ph.D. from the University Department of Chemical Technology, Bombay, India, and his M.S. degree in chemical engineering from the Indian Institute of Technology, Bombay. He worked as a postdoctoral research fellow at Monash University, Australia.

Way back when
These days we’re trying to work MTBE out of the motor fuel system. Back in the 1970s, MTBE was the great hope for replacing tetraethyllead as the primary antiknock additive in gasoline. Naturally, CHEMTECH, the forerunner to Chemical Innovation, was on top of this subject. In August 1979, we published an article by Brian Taniguchi of General Motors and Richard Johnson of the University of Missouri, Rolla, heralding the arrival of MTBE. The editors thought it would be interesting for the reader to compare excerpts from Taniguchi and Johnson with the discussion in the current article.

The article began:

    The search for antiknock compounds has been long and continuing. Research dates back to the early 1900s, and a seemingly unending progression of materials has been evaluated. In 1921, tetraethyllead was discovered to be a powerful antiknock additive, and since that time, its use has been the most effective method of increasing the octane of gasoline.

    Then in the 1960s and early 1970s, concern about the effect of auto emissions led to legislative restrictions on automotive exhaust emissions. Noble metal exhaust catalyst systems, which are poisoned by lead, were proposed as one way to meet the legislated emissions standards. Resulting legislation requiring gradual removal of lead compounds from gasoline caused renewed interest in nonlead octane quality improvers. Among materials that received significant attention are organic oxygen compounds such as alcohols and ethers. Methanol and ethanol received particular attention because, in addition to being octane improvers, they represent avenues for converting coal and biomass into liquid fuels. Unfortunately, when these alcohols are blended with gasoline, changes occur in intake mixture stoichiometry and fuel properties that substantially alter the driveability, emissions, and reliability of existing automobiles.

    Methyl tert-butyl ether (MTBE), which is more compatible with gasoline, shows promise as a high octane number blending stock and a mechanism to put methanol into gasoline.

    To make MTBE, methanol and isobutene, from a refinery stream, are reacted over a catalyst at moderate temperature and pressure to yield a product containing at least 80% MTBE. This product is subsequently dried and purified to a purity of 95% MTBE or better. Approximately 36 mass % of the final product MTBE can be attributed to the methanol feedstock. Thus, if the methanol is produced from coal or biomass, a 15% blend of MTBE in gasoline will derive approximately 3% of its energy from nonpetroleum sources. This potential to add nonpetroleum energy to the gasoline pool, and the high octane quality of MTBE make it a fuel blending stock worthy of serious consideration.

    MTBE has been in commercial use in Italy and Germany and has now been introduced commercially in the United States.

    In this study we evaluated MTBE as a high-octane blending component, and determined its advantages and disadvantages. The evaluation was conducted in three phases:

    Fuel properties—MTBE–gasoline blends were compared with gasoline in terms of commonly measured properties such as octane number, Reid vapor pressure (RVP), etc.

    Single-cylinder engine evaluation—Differences between MTBE–gasoline blends and gasoline were determined. Variables measured included emissions, power output, and efficiency.

    Evaluation of MTBE–gasoline blends in a vehicle— Road octane ratings were correlated with research and motor octane ratings determined in the fuel properties program. Also, the effects of MTBE–gasoline blends on vehicle emissions and fuel economy were determined.

After carefully considering the chemical, engineering, performance, and economic consequences, the authors concluded:

    This study indicates that MTBE is a high-octane blending component that will improve road octane rating of unleaded gasolines. Although no substantial advantage or disad vantage of 15 wt % MTBE–gasoline blends was found in single-cylinder engine comparisons at the same equivalence ratio, the slightly altered fuel stoichiometry and energy content of blends can cause moderate changes in the behavior of vehicles calibrated for gasolines. Generally, the changes were those associated with leaner operation of the engine. A better understanding of the emissions, fuel economy, and road octane performance of MTBE–gasoline blends in vehicles will require a study with several vehicles. Furthermore, road driveability tests and road fuel economy tests should be added. Also, tests using vehicles equipped with Phase II emission control systems (three-way catalyst, closed-loop carburetor) should be conducted. In addition, the water–precipitate problem noted in this report should be further investigated.

Needless to say, the suggested additional studies, and more, were conducted and MTBE became one of the most used organic compounds in history. Please continue to read Chemical Innovation to keep abreast of future developments in fuel additives.

Reprinted in part from CHEMTECH, 1979, 9, 502–510. © 1979 American Chemical Society.


Cartoon of a man and a woman talking to each other

Return to Top || Table of Contents