Alkylation

Flammable and Combustible Liquids

Eric Stauffer , ... Reta Newman , in Fire Debris Analysis, 2008

C/ Alkylation

Whereas cracking reactions seek to decrease the molecular weight of large molecules in order to get them into the gasoline range, alkylation seeks to increase the molecular weight of smaller molecules, in hopes of increasing the quantity of products in the gasoline range. However, alkanes will not easily combine with one another; they lack the necessary reactivity. For alkylation to occur, one of the reactants must be of the more reactive olefin type. The carbon-to-carbon double bond (C=C) that exists in olefins is much more reactive than the C–C single bonds of alkanes, thus much more likely to undergo this type of reaction. Because of the highly reactive nature of the olefins, this reaction can actually occur without the aid of a catalyst. However, because the reaction would require high temperatures and pressures, it is economically more practical to use a catalyst. A typical alkylation reaction is the reaction of isobutane with isobutylene to form isooctane, as shown in Figure 7-5.

FIGURE 7-5. The most common reaction in the alkylation process is the combination of butane and isobutylene to produce isooctane (2,2,4-trimethylpentane).

Similar reactions using propylene, pentylene, or isopentane as reactants are also possible, but usually not as advantageous as the isobutane-isobutylene reaction. One reason is that isopentane is already a useful component of gasoline, having desirable octane rating and vapor pressure qualities, so it is not economically profitable to use it as a reactant [10]. In addition, the octane ratings of the products from the propylene or pentylene reactions are not as high as those of the products from the butylene reaction [3].

One interesting phenomenon of the alkylation process is the resulting reduction of volume. The volume and density of the product are different from those of the reactants and because two smaller molecules combine to form a larger one, there is also a reduction in the number of moles. The change of volume is of great importance in the refinery industry, because volume is the primary measurement used. Reductions in volume may be in the order of 20%: When one barrel goes in as reactant, 0.8 barrel comes out as product. The converse occurs in the cracking reaction; it is said to have an expansion of volume.

Alkylation and Organic Synthesis

It should be pointed out that the term "alkylation" used in the petroleum refining industry is different than the same term used in synthetic organic chemistry. In organic chemistry, alkylation is known as the addition or substitution of an alkyl group to an aromatic ring, such as via the well-known Friedel Crafts alkylation reaction. In terms of petroleum refining, alkylation refers to the combination of a small olefin with a small isoparaffin, typically isobutane, to create a gasoline-range isoparaffin. The alkylation process yields a product rich in high octane rating isoparaffins, and is therefore a valuable gasoline blending stock.

Several catalysts can be used for the alkylation reaction such as aluminum chloride (AlCl3), sulfuric acid (H2SO4), or anhydrous hydrofluoric acid (HF). When acid catalysts are used, there must be a tertiary carbon atom in the alkane that is to react with the olefin, such as in isobutane and isopentane. The mechanism goes through a carbonium ion, which needs to be in the tertiary form in order to "spread out" the charge. The use of these catalysts allows for the reaction to occur at relatively low temperatures. The alkylation reaction takes place with the presence of isoparaffins in significant excess relative to the olefins; the purpose is to minimize undesirable polymerization reactions. The olefins for the reaction, propylene and/or butylenes, generally come from the cat cracking process and combine with the isobutane in a chiller. Following the reaction, the acid is separated out and recycled, and the products of the reaction are treated with a caustic wash to neutralize any remaining acid. The desired end product is then separated from light gases by fractionation.

The end result from the alkylation process is a product called alkylate. Alkylate is a high octane rating blending stock with a relatively narrow boiling point range. In addition to having a high octane value, alkylate has no olefins, no sulfur, and no benzenes, however it exhibits a high vapor pressure [9]. Alkylate consists primarily of isoparaffins and is very desirable for blending into gasoline.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780126639711500117

Catalyst Deactivation 2001

A. Borgna , ... C. Apesteguía , in Studies in Surface Science and Catalysis, 2001

1 Introduction

The alkylation of toluene with methanol is readily catalyzed on synthetic zeolites. Previous work has shown that the aromatic-ring alkylation of toluene with methanol takes place over acid zeolites [ 1], while the side-chain alkylation occurs preferentially over basic zeolites [2,3]. The side-chain alkylation of toluene with methanol, for producing a mixture of styrene and ethylbenzene offers economical advantages compared with the conventional homogeneously catalyzed Friedel-Crafts process, which use ethylene and benzene as reactants [4].

The role of the acid-base properties of zeolite catalysts on the selectivity in the alkylation of aromatic compounds has been previously studied [5]. Particularly, it has been suggested that the side-chain alkylation of toluene with methanol requires a cooperative action of acid/base pairs for efficiently promoting the rate-limiting step in the reaction mechanism [6,7]. However, the exact requirements of acid-base site density and strength is still debated. Studies on the effect that cofeeding acid and basic compounds has on the catalyst activity and selectivity will therefore contribute to obtain more insight on the reaction mechanism pathways. In the present work, we report the effect of the addition of acid and basic compounds on the deactivation kinetics for the alkylation of toluene with methanol over ion-exchanged Y zeolites.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299101802007

Superacids

George A. Olah , G.K. Surya Prakash , in Encyclopedia of Physical Science and Technology (Third Edition), 2003

III.F Alkylation

Alkylation of aromatics is carried out industrially on a large scale; an example is the reaction of ethylene with benzene to produce ethylbenzene, which is then dehydrogenated to styrene, the monomer used in producing polystyrene. Traditionally, these alkylations have been carried out in solution with a Friedel–Crafts acid catalyst such as AlCl 3. However, these processes are quite energy consuming and form a complex mixture of products requiring large amounts of catalyst, most of which is tied up as complexes and can be difficult or impossible to recover. The use of solid superacidic catalyst permits clean, efficient heterogeneous alkylations with no concomitant complex formation.

Aliphatic alkylation is widely used to produce high-octane gasolines and other hydrocarbon products. Conventional paraffin (alkane)–olefin (alkene) alkylation is an acid-catalyzed reaction; it involves the addition of a tertiary alkyl cation, generated from an isoalkane (via hydride abstraction) to an olefin. An example of such a reaction is the isobutane–ethylene alkylation, yielding 2,3-dimethylbutane.

The great interest in strong-acid chemistry is further exemplified by the discovery that lower alkanes such as methane and ethane can be polycondensed in Magic Acid at 50   °C, yielding mainly C4 to C10 hydrocarbons of the gasoline range. The proposed mechanism (Fig. 4) necessitates the intermediacy of protonated alkanes (pentacoordinate carbonium ions), at least as high-lying intermediates or transition states. Hydrogen must be oxidatively removed (by either the excess superacid or added oxidants) to make the condensation of methane thermodynamically feasible.

FIGURE 4. Mechanism of oxidative methane oligocondensation.

Because of the high reactivity of primary and secondary ions under these conditions, the alkylation reaction is complicated by hydride transfer and related competing reactions. However, in this mechanism it is implicit that an energetic primary cation will react directly with methane or ethane. This opens the door to new chemistry through activation of these traditionally passive molecules.

A convenient way to prepare an energetic primary cation is to react ethylene with superacid. This has been used with HF–TaF5 catalyst to achieve ethylation of methane in a flow system at 50   °C. With a methane–ethylene mixture (85:14), propane is the major product.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B0122274105007456

Natural Gas Conversion V

V.M. Mysov , ... A.V. Toktarev , in Studies in Surface Science and Catalysis, 1998

3.2 Alkylation of aromatics with synthesis gas

Alkylation of toluene with methanol or ethene on modified by different elements ZSM-5 type zeolites to produce p-xylene or p-ethyltoluene is studied in detail previously [12-15]. In many papers it was shown that synthesis gas may be used as an alkylation agent instead of methanol. It was admitted [16,17] that the efficiency of alkylation by methanol being formed in the catalyst bed continuously, was significantly higher compared with that in the case of the conventional methylation with methanol. But, Namba et al. showed that although alkylation of toluene with methanol on modified with boron ZSM-5 zeolites proceeded with as high as 94% selectivity to p-xylene [16], the use of the same modified zeolites bound with Zn-Cr component in the bifunctional catalyst gave only 43% selectivity with toluene conversion of 6.5% only in alkylation of toluene with synthesis gas.

We showed that the selectivity of the process to p-xylene was 52% with toluene conversion more than 40% when bifunctional catalysts were not modified [10]. Moreover, the composite catalysts containing ZSM-5 zeolite modified with Mg and Si oxides exhibited selectivity to p- xylene and p-ethyltoluene nearly 70-91% and 76-97%, respectively. Toluene conversion was 20-30% but syngas conversion was over 30%.

We studied the bifimctional catalysts containing ZSM-5, ZSM-12 and β zeolites for reaction of 2-MN methylation with synthesis gas. The main alkylation products were dimethylnaphthalenes (65-93% in total naphthalene products), but naphthalene, trimethylnaphthalenes and syngas conversion products were formed too. It was shown that selectivity of 2-MN alkylation process depends on temperature, contact time, catalyst composition, dimensions of zeolite channels and acidity of zeolite. The selectivity to (2,6   +   2,7)-DMN for ZSM-5 and the one to 1,2-DMN for β zeolites changed with process time. Effect of zeolite channels dimensions on the process selectivity is shown in Table 2.

Table 2. Alkylation of 2-methylnaphthalene with synthesis gas

Type of zeolite ZSM-5 ZSM-12 β zeolite
Dimensions of the channels, Ã… 5.3   ×   5.6 5.5   ×   6.2 6.4   ×   7.6
Hours on stream 2 12 2 12 2 12
DMN distribution, wt %
2,6   +   2,7 48.2 91.5 18.1 16.6 25.6 4.1
1,7 26.8 1.1 15.7 15.8 8.0 15.7
1,3   + 1,6 18.5 3.4 21.1 21.4 35.5 14.7
2,3   +   1,4 2.6 0.8 9.9 10.1 16.8 7.7
1,5 trace - trace - trace trace
1,2 3.9 3.2 35.2 36.1 14.1 57.8

Reaction temperature - 440   °C, pressure - 80   atm.

From Table 2 one can see that increased selectivity to (2,6   +   2,7)-DMN is achieved on ZSM-5 and increases more with process time. On wide pore zeolites the distribution of formed DMN is in accordance either with thermodynamic equilibrium (β zeolite, first 2   hours), or with an increased amount of 1,2-isomer.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299198804862

Scientific Bases for the Preparation of Heterogeneous Catalysts

Norikazu Nishiyama , ... Korekazu Ueyama , in Studies in Surface Science and Catalysis, 2006

3.3 Alkylation of toluene

Alkylation of toluene with methanol was carried out over the uncoated H-ZSM-5 and silicalite/H-ZSM-5 ( x = 0.12) catalysts. The evolution of toluene conversion and para-selectivity over the uncoated H-ZSM-5 catalysts with reaction time are shown in Figure 3. Reaction temperature was 673   K and the space time (W/F) was 0.06   kg-catalyst h mol−1. The toluene conversion and para-selectivity at the early period were more than 40% and 65%, respectively. The para-selectivity over the uncoated ZSM-5 catalyst was higher than that of the thermodynamic equilibrium value (23%). The toluene conversion decreased to 18% after 3   h because of coking, but no change in the para-selectivity was observed for 3   h.

Figure 3. Catalytic activity of uncoated H-ZSM-5 for the alkylation of toluene with methanol: circles, toluene conversion; squares, para-selectivity. Toluene/methanol = 1.0. Reaction temperature = 673   K. W/F = 0.06   kg-catalyst h mol−1.

Figure 4 shows the evolution of the toluene conversion and para-selectivity over the silicalite/H-ZSM-5 catalyst for the alkylation of toluene with reaction time. Reaction temperature was 673   K and the space time (W/F) was 0.06   kg-catalyst h mol−1. The toluene conversion and para-selectivity over the silicalite/H-ZSM-5 catalyst after the reaction time of 2   h were higher than 35   % and 98   %, respectively. However, the toluene conversion decreased with time similarly to the results of the uncoated H-ZSM-5 catalysts. The toluene conversion went down to lower than 10   % after 6   h. The para-selectivity also slightly decreased to approximately 88   % after the reaction time of 8   h. A possible reason is that the rate of isomerization on only a few acid sites near the external surface did not change with reaction time although the rate of p-xylene formation was decreased with reaction time by coking inside the pores of H-ZSM-5.

Figure 4. Catalytic activity of silicalite/H-ZSM-5 for the alkylation of toluene with methanol: circles, toluene conversion; squares, para-selectivity. Toluene/methanol = 1.0. Reaction temperature = 673   K. W/F = 0.06   kg-catalyst h mol−1.

The toluene conversion and para-selectivity after the reaction time of 60   min as a function of space time, W/F, are shown in Figure 5. The para-selectivity on the uncoated H-ZSM-5 catalyst decreased with increasing the toluene conversion and finally the para-selectivity went down to approximately 30   % at high W/F. On the other hand, the silicalite/H-ZSM-5 catalyst showed excellent para-selectivity. The para-selectivity was not largely decreased with increasing space time. The toluene conversion was constant at high space time, because the diffusion resistance of reactants through the silicalite-1 layer is dominant.

Figure 5. The toluene conversion and para-selectivity as a function of space time for the alkylation of toluene with methanol: circles, uncoated H-ZSM-5; square, silicalite/H-ZSM-5; open symbols, toluene conversion; closed symbols, para-selectivity. Toluene/methanol = 1.0. Reaction temperature = 673   K.

The alkylation of toluene with methanol was performed with hydrogen gas in the feed to inhibit the catalyst coking. The results of catalytic testing with reaction time are shown in Figure 6. Compared to the results without hydrogen gas, hydrogen gas in feed prevented not only toluene conversion but also para-selectivity from decreasing with reaction time. This result indicates that catalytic deactivation is related to the para-selectivity. However, to clarify the details of a deactivation mechanism, further investigation on the relation between coking and para-selectivity is required.

Figure 6. Catalytic activity of silicalite/H-ZSM-5 for the alkylation of toluene with methanol with 40   ml/min of hydrogen in feed, circles, toluene conversion; squares, para-selectivity. Toluene/methanol= 1.0. Reaction temperature = 673   K.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299106809171

Nanoporous Materials III

Young-Woong Suh , Hyun-Ku Rhee , in Studies in Surface Science and Catalysis, 2002

2.6 Catalytic experiment

Alkylation of isobutane with 1-butene was carried out in a fixed-bed flow reactor at atmospheric pressure. The reaction in the gas phase condition can lead to a low concentration of the reactants in the reaction system so that the deactivation rate of catalysts is slow. Isobutane was first charged into the reactor at room temperature to avoid fast polymerization of olefins. Once the desired reaction temperature was stabilized, the reaction was started by feeding 1-butene into the reactor. An isobutane/1-butene volumetric flow rate ratio of 10 was used and the olefin space velocity, WHSV, was kept at 3   h−   1 in these experiments. The reactants and the products were quantitatively analyzed by gas chromatography using an FID and a 50   m HP-1 capillary column. Data at different times on stream were obtained from the samples extracted directly from the reactor. The sulfated zirconia catalysts were activated in the reactor by calcining it in air at 550   °C for 3   h. Air flow was then replaced by nitrogen and the catalyst temperature was lowered to room temperature.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299102805547

Introduction to Zeolite Science and Practice

Herman van Bekkum , Herman W. Kouwenhoven , in Studies in Surface Science and Catalysis, 2007

7.1.1 Alkylation-cyclization

Aromatic alkylation followed by alkylation of the side chain and cyclization can lead to interesting new compounds. In the following example [ 115] zeolite Beta appears to be a unique catalyst (Figure 47).

Figure 47.

Conditions: autoclave charged with naphthalene (10mmol), isopropyl alcohol 20mmol, 0.5g zeolite H-Beta (Si/Al = 12.5), cyclohexane (100 ml), undecane (10mmol) as internal standard, 200°C, 2MPa N2. After 1h the selectivities to 2-isopropylnaphthalene and to the cyclic compound I are 50 and 46%, respectively, at 19% naphthalene conversion.

The mechanism presumably involves an intermediate cation (Figure 48) which reacts with propene (iPrOH) to another cation (Figure 49) which undergoes fast cyclization.

Figure 48.

Figure 49.

On the more spacious zeolite HY the isopropylation of naphthalene is faster but the main products are mono-, di- and tri-isopropylnaphthalene.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299107808147

Reaction Kinetics and the Development of Catalytic Processes

Mohammed C. Al-Kinany , Soliman H. Al-Khowaiter , in Studies in Surface Science and Catalysis, 1999

1. Introduction

Alkylation of benzene with ethylene to produce ethylbenzene, an intermediate in the production of styrene, is well established process (1-5) . The major byproducts formed during the reaction are polyethylated benzenes such as o-, p- and m-diethylbenzenes, triethylbenzene isomers and tetraethylbenzene isomers. Higher ethylated products, including pentaethylbenzene and hexaethylbenzene are also produced. Considerable attention has been given to this chemistry because large amounts of ethylbenzene were required to meet the increasing demand for styrene in this highly competitive market. The importance of ethylbenzene as a precursor to styrene and other petrochemical derivatives is shown in scheme (1).

Scheme (1).

The following catalysts have been used.

1.

Metal halide catalysts (Lewis acids) : such as anhydrous aluminum chloride, boron trifluoride and titanium tetrachloride.

2.

Protonic acid catalysts (Bronsted acids) : such as sulphuric acid, hydrofluoric acid, and super acids.

3.

Solid acid catalysts: such as zeolites (6) , and solid phosphoric acid.

Transfer of an alkyl group from one to another molecules is the basis of the transalkylation reaction. This reaction is industrially of some interest as some of the low valued products like polyethylbenzenes can be converted to their monosubstituted homologues, which are in higher demand and of higher value.

The transalkylation reaction, first demonstrated in 1894 (8) , forms the basis of an important commercial synthetic process for the synthesis of ethylbenzene from fresh benzene and recycled higher ethylbenzenes in the presence of a metal halide (i.e. AlCl3, BF3), protonic acid, or solid acids.

Transalkylation is usually carried out on an industrial scale employing aluminium chloride (1) or trifluoroboron (5) at 100?C, and zeolite (2,3) catalysts at temperatures higher than 200?C. Supported phosphoric acid catalysts are ineffective (9) for this reaction. The diethylbenzenes, formed using this kind of catalyst, are not recycled for transalkylation, and to minimize their formation higher ratios of benzene to ethylene are often used. The transalkylation of diethylbenzene with benzene has also been studied over solid superacidic catalyst, such as Nafion-H (10) . The proposed reaction mechanism, for transalkylation of diethylebenzene with benzene in the presence of Lewis acid is presented in scheme (2).

Scheme (2).

The other reaction involving disproportionation of diethylbenzene to triethylbenzenes and ethylbenzene is shown in scheme (3).

Scheme (3).

Transalkylation of diethylbenzenes to obtain higher yield of ethylbenzene using the above and a wide range of catalysts has been intensively studied during the past decade. The major reactions include isomerization and transalkylation with their relative contributions controlled by the nature of catalyst and the reaction conditions. Considerable efforts have been aimed at the isomerization and disproportionation of diethylbenzene isomers, in the absence of benzene, to produce ethylbenzene by using trifluoromethanesulphonic acid catalyst (7) .

The present work was carried out to investigate the efficiency of trifluoromethanesulphonic acid as a catalyst for isomerization and transalkylation of diethylbenzene isomers in the presence of benzene at room temperature in order to synthesise higher yield of ethylbenzene.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167299199801694

Petroleum Refining

James H. Gary , in Encyclopedia of Physical Science and Technology (Third Edition), 2003

IV.D Alkylation

Alkylation, as used in the petroleum refining industry, refers to the process for reacting olefins with an isoparaffin to produce a higher molecular weight paraffin. Specifically, the process reacts propylenes and butylenes with isobutane to produce branched chain paraffins in the gasoline boiling range. The product is called alkylate and is a premium gasoline blending stock because it has high octane numbers and a low sensitivity.

There are two basic processes, one using concentrated hydrofluoric acid (Fig. 15) as the catalyst and the other using concentrated sulfuric acid. Although the product yields and quality are essentially the same, there are significant differences in process operation because of differences in catalyst characteristics. The process using hydrofluoric acid as catalyst is performed at temperatures between 30 and 43   °C (85 and 110   °F) while the process using sulfuric acid operates from 4 to 13   °C (40 to 55   °F). The reactions are highly exothermic, and the heat created by the reactions must be removed from the reactor in order to keep the reactions under control and to achieve good yields and high alkylate quality. For the hydrofluoric acid process, the reaction temperature is sufficiently high that the temperature can be controlled by removing the heat of reaction with cooling water. For the sulfuric acid process, it is necessary to use refrigeration to remove the reaction heat and control the temperature. The need for refrigeration increases the operating cost significantly over that of the hydrofluoric acid process.

FIGURE 15. Hydrofluoric acid alkylation unit. (Courtesy of Phillips Petroleum.)

If a stream of mixed olefins (propylene and butylene) is reacted with isobutane, the product stream leaving the reactors consists of (1) C5 + alkylate, (2) propane, (3) n-butane, and (4) alkylate tar. The alkylate tar is a viscous brown oil consisting of polymerization products of high molecular weight. Only a small amount is made, usually less than 0.1% of the olefin feed.

For the mixed olefin feed the C5 + alkylate product has a research octane number (RON) of 94 and a motor octane number (MON) of 91. For each volume of olefin feed, 1.5 volumes of alkylate are produced. These numbers are approximate because the propylene/butylene ratio in the feed and operating conditions affect the product quality and quantity.

A simplified process flow diagram for the hydrofluoric acid process is shown in Fig. 16. The olefin and isobutane feed streams are dehydrated by passing them through a solid-bed desiccant unit. Good dehydration is necessary to prevent equipment corrosion and to prevent dilution of the acid catalyst. The dehydrated streams are mixed with each other and with the hydrofluoric acid catalyst at a pressure sufficiently high to prevent vaporization of the components. The mixture is passed into a separating vessel and separated into two layers. The top layer of reacted hydrocarbons and excess isobutane is sent through a series of distillation columns to separate the propane, isobutane, n-butane, and C5 + alkylate into streams. The propane and n-butane are sent to storage, the isobutane is recycled, and the alkylate is sent to product blending.

FIGURE 16. HF alkylation. A, Olefin feed drier; B, isobutane feed drier; C, acid settler; D, acid cooler; E, acid rerun column; F, acid accumulator; G, depropanizer; H, depropanizer accumulator; J, acid stripper; K, deisobutanizer; K, deisobutanizer accumulator; M, debutanizer; N, debutanizer accumulator; P, propane caustic treater; Q, butane caustic treater. [From Gary, J. H., and Handwerk, G. E. (1994). "Petroleum Refining, Technology and Economics," 3rd ed. Marcel Dekker, New York.]

The acid stream settling to the bottom of the separating vessel is recycled but a slip-stream is withdrawn to separate the alkylate tar from the hydrofluoric acid in a distillation column. The pure acid is put into the acid recycle stream and alkylate tar is sent to an accumulator.

Isobutane is recycled to give a high ratio of isobutane to olefins in the reactor. Typically, recycle ratios of isobutane to olefins in the feed are from 3 to 10.

The heat of reaction can be removed by passing the reactants through a water-cooled heat exchanger before entering the separating vessel or by passing the recycled acid through a cooler.

The process flow for a sulfuric acid system is similar to that of the hydrofluoric acid unit, with two major exceptions. First, sulfuric acid is more viscous than hydrofluoric acid, and it is necessary to use mixers to get good contact between the hydrocarbons and the acid catalyst. This introduces a mechanical heat of mixing that must be removed in addition to the heat of reaction in order to maintain the reaction temperature. Second, a refrigeration system with a compressor and condenser is required to remove the heat of reaction and mixing; it does so by vaporizing the propane and some of the isobutane in the system, compressing the gases produced, and condensing them for reintroduction into the system as a liquid. The distillation system to separate the components in the product stream is similar to that for the hydrofluoric acid process.

The capital costs for the two systems are similar because the acid recovery system and safety requirements for the hydrofluoric system balance the refrigeration system cost of the sulfuric acid unit. Both units are constructed of carbon steel, but the hydrofluoric acid unit requires that pump impellers, valve trim, and portions of the system where water–acid mixtures are a possibility be constructed of Monel metal or Monel clad steel.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B0122274105005561