• 検索結果がありません。

ε -Caprolactam Study on Environmentally Benign Catalytic Processes for the Production of

N/A
N/A
Protected

Academic year: 2022

シェア "ε -Caprolactam Study on Environmentally Benign Catalytic Processes for the Production of"

Copied!
68
0
0

読み込み中.... (全文を見る)

全文

(1)

Study on Environmentally Benign Catalytic Processes for the Production of ε-Caprolactam

ε‑カプロラクタム製造のための環境調和型接触プロセスの研究

July 2003

Hiroshi ICHIHASHI

市橋 宏

(2)

CONTENTS

INTRODUCTION

Chapter 1: CURRENT ε-CAPROLACTAM MANUFACTURING PROCESSES 1.1. Overview of current ε-caprolactam manufacturing processes

1.2. Cyclohexanone manufacturing

1.3. Hydroxylamine and cyclohexanone oxime manufacturing 1.4. Caprolactam manufacturing

1.5. Ammonium sulfate generation in caprolactam processes References

Chapter 2: NEW ENVIRONMENTALLY BENIGN PROCESSES FOR THE PRODUCTION OF ε-CAPROLACTAM

2.1. New environmentally benign processes

2.2. Comparison between the current processes and the proposed new processes 2.3. A review of ammoximation processes

2.3.1. A history and overview of ammoximation reaction 2.3.2. Ammoximation with hydrogen peroxide

2.3.3. By-products through the ammoximation process using hydrogen peroxide References

1

3 3 4 4 6 7 7

8 8 10 13 13 14 16 17

(3)

Chapter 3: A STUDY AND DEVELOPMENT OF THE CATALYSTS FOR THE SELECTIVE HYDROGENATION OF BENZENE

3.1. Background of the study

3.2. Outline of a cyclohexanol manufacturing process via cyclohexene from benzene 3.3. Working hypotheses for searching a fine selective hydrogenation catalyst 3.4. Experimental procedure

3.5. Experimental results and discussion 3.5.1. The search for carrier materials

3.5.2. Effects of the addition of some metal salts aqueous solution into the reaction system

3.5.3. Reaction scheme

3.5.4. The results of catalyst search 3.6. Conclusions

References

Chapter 4: THE DEVELOPMENT OF A CATALYST FOR THE VAPOR PHASE BECKMANN REARRANGEMENT PROCESS

4.1. A history and overview of the study on the vapor phase Beckmann rearrangement 4.2. The outline of Sumitomo's vapor phase Beckmann rearrangement process

4.3. Experimental procedure

4.4. The experimental results and discussion

19 19 21 22 23 23 23

25 27 28 31 31

33 33 34 35 36

(4)

4.4.1. Acidity of the catalyst

4.4.2. Effect of methanol addition into the reaction system 4.4.3. Effect of ammonia addition into the reaction system

4.4.4. Reaction temperature suitable for the vapor phase Beckmann rearrangement 4.4.5. Regeneration of the catalyst

References

Chapter 5: THE ACTIVE SITES OF A HIGH SILICA MFI ZEOLITE CATALYST AND THE REACTION MECHANISM OF VAPOR PHASE BECKMANN REARRANGEMENT

5.1. The active site of the catalyst 5.2. The reaction mechanism

5.3. Vapor Phase Beckmann rearrangement of other oximes References

Chapter 6: CONCLUDING REMARKS

ACKNOWLEDGEMENTS

LITERATURE RELATED TO THIS THESIS

36

38

41

43

45

45

47

47

53

54

56

58

60

61

(5)

INTRODUCTION

ε-Caprolactam (hereafter caprolactam) is an important intermediate used in the production of Nylon 6 fibers and resins, the worldwide production of which is estimated at about 4 million tons annually. All caprolactam is produced through the Beckmann rearrangement of cyclohexanone oxime, except a small amount that is produced through depolymerization of nylon 6 polymers.

The current caprolactam manufacturing processes produce large amounts of ammonium sulfate as a byproduct (at least 1.6 times of caprolactam by weight), because oleum or sulfuric acid is used for the reaction promoter in the Beckmann rearrangement process. As the profitability of caprolactam production strongly depends upon the amounts of ammonium sulfate, and a chemical process to manufacture only desired products is required, new processes to avoid its generation have been sought in many laboratories in industry and academia for a long time. The problems in current processes are discussed in Chapter 1.

In this thesis, the author proposes a new catalytic route for the production of caprolactam without producing any ammonium sulfate as a byproduct. The route consists of the following five reaction processes; selective hydrogenation of benzene to cyclohexene, hydration of cyclohexene to cyclohexanol, dehydrogenation of cyclohexanol to cyclohexanone, ammoximation of cyclohexanone to cyclohexanone oxime and vapor phase Beckmann rearrangement of cyclohexanone oxime to caprolactam. Cyclohexane and water are produced as byproducts through the processes.

However, cyclohexane is a valuable chemical intermediate and it is manufactured by hydrogenation of benzene commercially, hence its production does not cause drawbacks to chemical industries. Thus, the proposed catalytic route for the production of caprolactam can be considered environmentally benign. The author discusses the route in Chapter 2.

Sumitomo Chemical Co., Ltd. has industrialized the combined process of ammoximation and vapor phase Beckmann rearrangement for the first time in the world in 2003. The author reviews the characteristics of the ammoximation process from an industrial point of view in Chapter 2. A brief history of the research, the reaction mechanism, and the by-products that can be produced during the reaction are discussed.

(6)

The issues concerning the development of a selective hydrogenation catalyst for the production of cyclohexene from benzene are discussed in Chapter 3. When we started the development of the cyclohexene manufacturing process, no effective methods had been proposed in an industrial sense. We obtained a good catalyst and could prove that cyclohexene was produced by the selective hydrogenation of benzene with the catalyst in a good yield. The author describes the results and hypothesis for searching good catalysts.

The results from the development of the vapor phase Beckmann rearrangement are discussed in Chapter 4. The active site of the catalyst and reaction mechanism are discussed in Chapter 5. The Beckmann rearrangement reaction is one of the typical acid catalyzed reactions, but the catalyst that we have developed is mainly composed of a high silica MFI zeolite (Silicalite-1) and does not posses acidity that can be detected by ammonia TPD measurements.

Methanol fed into the reactor with cyclohexanone oxime improves the yield of caprolactam. Methanol reacts with terminal silanols on the zeolite surface, and converts them to methoxyl groups. The modification of the catalyst by methanol plays an important role in the Beckmann rearrangement reaction.

The author and his colleagues reveal that nest silanols close to the pore mouth of an MFI zeolite are the active sites of the catalyst. The coordination between the NOH group of cyclohexanone oxime molecule and the nest silanols is responsible for the reaction. The reaction mechanism of the Beckmann rearrangement under vapor phase conditions is also discussed and the author concludes that it is the same as in the liquid phase, namely the alkyl group in anti position against the hydroxyl group of the oxime migrates to the N atom position.

(7)

Chapter 1

CURRENT ε-CAPROLACTAM MANUFACTURING PROCESSES

1.1. Overview of current ε-caprolactam manufacturing processes

Caprolactam has been manufactured currently through several reaction processes as shown in figure 1.1. Caprolactam is mostly produced from benzene through three intermediates; cyclohexane, cyclohexanone and cyclohexanone oxime.

Cyclohexanone is mainly produced by the oxidation of cyclohexane with air, but a small part of it is obtained by the hydrogenation of phenol. It can also be produced through the selective hydrogenation of benzene to cyclohexene, subsequent hydration of cyclohexene, and dehydrogenation of cyclohexanol. The route via cyclohexene has been commercialized by Asahi Chemical Company in Japan for adipic acid manufacturing, but the process has not yet been applied for caprolactam production.

Hydroxylamine is also very important as a raw material for the production of cyclohexanone oxime.

Toray in Japan produces cyclohexanone oxime directly from cyclohexane by the PNC process.

O

OH

NOH

NH O

+ O2 + NH2OH + H2SO4

+ NH3

+ H2 + H2O

OH

- H2

+ H2

+ H2 + NOCl

PNC (Toray)

Oximation

Beckmann rearrangement Neutralization Oxidation

Hydrogenation Selective

hydrogenation

Hydration Hydrogenation

NH2OH Production Raschig process

Catalytic NO reduction (BASF process) Catalytic NO3-reduction (HPO process) Figure 1.1. Current caprolactam manufacturing process

(8)

1.2. Cyclohexanone manufacturing

Cyclohexanone is mostly manufactured by the cobalt-catalyzed air oxidation of cyclohexane. Typical reaction conditions in commercial processes are as follows:

temperature=150-160 ℃ , pressure=0.81-1.013MPa (8-10atm) with catalyst concentrations of 0.3-3ppm. Cyclohexane conversion is usually held at 4-6mol% and selectivity to cyclohexanol and cyclohexanone is around 80% with a ratio (alcohol/ketone) of about 2:1[1.1].

Cyclohexanol is converted to cyclohexanone by catalytic vapor phase dehydrogenation. The reaction generally runs at 200-400℃ and atmospheric pressure.

Lower temperature reactions (200-300℃) give conversions of 40-70% and selectivity near 100%. Zn, Cu, Mg, and Ni compounds are widely used as catalysts, and are often promoted by other metals [1.2]. Hydrogen produced through the reaction can be used as a raw material in chemical industries. This oxidation process has the following drawbacks.

(1) Since the conversion of cyclohexane per pass through the oxidation reactor is very low, substantial quantities of cyclohexane must be removed by distillation from the oxidizer effluent for recycle to the reactor feed. The energy consumption for recycling cyclohexane is large.

(2) About 20% of undesired substances are produced through the process.

1.3. Hydroxylamine and cyclohexanone oxime manufacturing

There are three methods for the production of hydroxylamine: Raschig process (reduction of ammonium nitrite with sulfur dioxide), BASF process (catalytic hydrogenation of NO), and DSM process (catalytic hydrogenation of NO3- anion).

The Raschig process produces large amounts of ammonium sulfate, but some manufacturers still use this method. Hydroxylamine sulfate is obtained via hydroxylamine disulfonate by reducing ammonium nitrite with sulfur dioxide, and hydrolyzing in water (equation 1.1).

NH4NO2 + NH3 + 2SO2 + H2O       HON(SO3NH4)2

HON(SO3NH4)2 + 2H2O NH2OH・H2SO4 + (NH4)2SO4 (1.1)

(9)

For the benefit of reducing ammonium sulfate formation, two processes are commercially applied; one is the catalytic hydrogenation of NO developed by BASF, and the other is the catalytic hydrogenation of nitrate anion developed by DSM.

The former process produces hydroxylamine hydrogensulfate by hydrogenation of NO in dilute sulfuric acid aqueous media with the help of Pt on carbon catalyst as indicated in equation 1.2.

DSM in the Netherlands developed a hydroxylamine manufacturing process without producing any ammonium sulfate as in equation 1.3, where a nitrate anion is hydrogenated in a buffer solution composed of phosphoric acid and ammonia aqueous solution with the aid of Pd catalyst. The reaction products containing hydroxylamine are sent to the oximation process. Hydroxylamine reacts with cyclohexanone, and then cyclohexanone oxime is formed (equation 1.4).

N H4N O3 + 2 H3PO4 + 3 H2

[N H3O H]+[H2P O4]- + N H4H2PO4 + 2 H2O P d C a ta lys t

(1 .3 )

[NH3OH]+[H2PO4]- + NH4H2PO4 + 2 H2O + O

NOH + H3PO4 + NH4H2PO4 + 3 H2O

(1 .4 )

The mixture of the products is extracted with toluene, and cyclohexanone oxime is obtained. The aqueous solution of phosphoric acid and ammonium dihydrogen phosphate is recovered and recycled to the hydroxylamine-forming process with nitric acid.

NO  +  3/2 H2 + H2SO4           NH2OH・H2SO4

Pt/C Catalyst

(1.2)

(10)

When cyclohexanone oxime is produced by the reaction between cyclohexanone and hydroxylamine obtained by Raschig process, ammonium sulfate is produced in amounts about 2.8 times the weight of caprolactam. Hydroxylamine produced by BASF process generates about 1.0 times of caprolactam, because the oximation reaction is conducted with aqueous hydroxylamine sulfate solution containing ammonium sulfate [1.3].

Toray produces cyclohexanone oxime by the reaction named PNC (Photonitrosation of cyclohexanone oxime), where cyclohexane is reacted with nitrosyl chloride (NOCl) under the radiation of UV light (equation 1.5).

Hydrogen chloride eliminated in the Beckmann rearrangement process is recovered and recycled into the NOCl production. All cyclohexanone oxime, with the exception of the PNC process, is manufactured by the reaction between cyclohexanone and hydroxylamine.

1.4. Caprolactam manufacturing

Currently all caprolactam is produced by the Beckmann rearrangement process, where large amounts of sulfuric acid or oleum are required as a reaction promoter.

Caprolactam is produced as its sulfate form; therefore it is necessary to neutralize the reaction products with ammonia for its isolation. Ammonium sulfate is produced in amounts at least 1.6 times the weight of caprolactam unavoidably (equation 1.6).

+ NOCl + HCl NOH

.

2 HCl

NH

O N

OH

NOH Oleum .1/2 H2.SO4 NH3 + 1/2 (NH4)2SO4 (1.1)

(1.5)

(1.6)

(11)

1.5. Ammonium sulfate generation in caprolactam processes

The produced amounts of ammonium sulfate are summarized in Table 1.1 [1.3].

Currently 1.6〜4.4 kg ammonium sulfate are generated per kg of caprolactam. The Sumitomo's new process commercialized in 2003 does not produce any ammonium sulfate. The main theme of this thesis concerns the development of the process and research of the catalysis.

Table 1.1. Ammonium sulfate generation in caprolactam processes

References

[1.1] Kirk-Othmer Encyclopedia of Chemical Technology, vol 1, p514-515.

[1.2] SRI Process Economic Program, PEP'95 7C AEH, p4-15, SRI Consulting California.

[1.3] G. Petrini, G. Leofanti, M. A. Mantegazza and F. Pignataro, ACS Symp. Ser., 626 (Green Chemistry), 33-48 1996.

Process Ammonium Sulfate Generation kg/kg of Caprolactam

Oximation Rearrangement Total

Raschig 2.8 1.6 4.4

BASF 1 1.6 2.6

DSM 0 1.6 1.6

Sumitomo's New Process*) 0 0 0

*) Combined process between ammoxidation and vapor phase Beckmann rearrangement

(12)

Chapter 2

NEW ENVIRONMENTALLY BENIGN PROCESSES FOR THE PRODUCTION OF ε-CAPROLACTAM 2.1. New environmentally benign processes

The author proposes an environmentally benign route for the production of caprolactam that consists of the following five reaction processes as indicated in figure 2.1.

(1) Selective hydrogenation of benzene (2) Hydration of cyclohexene

(3) Dehydrogenation of cyclohexanol (4) Ammoximation

(5) Vapor phase Beckmann rearrangement.

Byproducts are cyclohexane and water through the processes. Cyclohexane produced in the first process can be used as a chemical intermediate or can be dehydrogenated into benzene and recycled to the hydrogenation reactor.

The molar selectivity of each process is shown in figure 2.1. As the byproduct cyclohexane is a useful chemical intermediate, the economical efficiency of the selective hydrogenation of benzene can be considered almost perfect. Thus, the yield of cyclohexanone can be considered more than 98% economically. Therefore the overall yield of caprolactam or efficiency of benzene can be considered 92%.

The selective hydrogenation of benzene is carried out in liquid phase with a noble heterogeneous catalyst. The development of the catalyst is discussed in Chapter 3.

The hydration of cyclohexene is conducted in liquid phase with an MFI zeolite as a catalyst at moderate temperature (100〜130℃). The method to produce cyclohexanol through the two processes has already been commercialized by Asahi Chemical Company [2.1].

The dehydrogenation of cyclohexanol has been operated by many caprolactam producers, because current processes need to employ this reaction in the systems.

Therefore the reaction process has been already well established as described in Chapter 1.

(13)

Cyclohexanone oxime is produced by ammoximation reaction, which EniChem developed. The author reviews the process and catalysis fully in Section 2.3.

Finally, caprolactam is manufactured with the vapor phase Beckmann rearrangement catalytic process that the author and his colleagues have developed. The development of the catalyst, the study of the catalysis, and the reaction mechanism are discussed in Chapters 4 and 5.

H

2

NOH

NH O

Selective Hydrogenation

Hydration

Dehydrogenation

Ammoximation

Beckmann Rearrangement H

2

H

2

O

NH

3

H

2

O

2

H

2

O

OH

O

(Selectivity %)

(75%)

(99%)

(99%)

(98%)

(>96%)

Figure 2.1. Environmentally benign caprolactam manufacturing route.

(14)

2.2. Comparison between the current processes and the proposed new processes A typical current caprolactam manufacturing route (Raschig or BASF process is employed for the production of hydroxylamine) is shown in figure 2.2.

H

2

NOH

NH O

Hydrogenation

Oxidation

Dehydrogenation

Oximation

Beckmann Rearrangement H

2

Air

(NH

4

)

2

SO

4

OH O

NH

2

OH+H

2

SO

4

NH

3 O

Byproducts

H

2

SO

4

NH

3

(NH

4

)

2

SO

4 (ca100%)

(80%) (20%)

(99%)

(ca100%)

(99%)

(Selectivity %)

Figure 2.2. A typical current caprolactam manufacturing route.

(15)

During the cyclohexane oxidation process, many kinds of byproducts (carboxylic acids; adipic acid, hydroxycapric acid, glutaric acid, succinic acid, tartaric acid, oxalic acid, acetic acid, formic acid, etc, esters; caprolactone, cyclohexyl adipate, etc, deeply oxidized products; carbon monoxide and carbon dioxide) are produced, and the total yield of byproducts is about 20 mol%. Thus the overall caprolactam yield of the typical current route is 78% when unreacted cyclohexane is completely recycled to the oxidation process. Therefore the efficiency of benzene converting to caprolactam is 78% in the current route and 92% in the newly proposed route.

The author would like to point out the large difference in the efficiency of benzene to cyclohexanone between the two; it is 98% in the new route and 79% in the current route. That means not only the benzene consumption of the new route is much improved upon that of the current ones, but also the waste-products treatment is simpler in the new processes than in the current ones.

The oxidation of cyclohexane with air is conducted in liquid phase and the conversion of cyclohexane is only 4-6%, therefore a large volume of cyclohexane must be treated in the process. The producers must take special care of safety, because the oxidation reaction produces some organic hydroperoxides that are unstable and auto-decomposing compounds. On the contrary, the hydration reaction between cyclohexene and water is much safer.

Figure 2.3 shows the comparison of cyclohexanone oxime manufacturing between the current and new processes. In a current process (BASF process), hydroxyl ammonium sulfate is manufactured by the hydrogenation of nitrogen oxide in dilute sulfuric acid media with the help of a special platinum catalyst. Nitrogen oxide is formed by the oxidation of ammonia with pure oxygen gas. Thus the process has disadvantages. Not only produces ammonium sulfate but also many manufacturing plants are required, such as cyclohexanone, ammonia, oxygen, nitrogen oxide, sulfuric acid, hydrogen, hydroxylamine manufacturing, and an ammonium sulfate recovery.

(16)

On the contrary, the new process ammoximation is very simple.

Cyclohexanone reacts with hydrogen peroxide and ammonia to produce cyclohexanone oxime in a good conversion (99.9%) and good yield (98.2%) [2.2]. Hydrogen peroxide can be manufactured by a well-established process (alkyl anthraquinone method), in which hydrogen is oxidized with air. Thus, the ammoximation process requires only cyclohexanone, hydrogen peroxide, hydrogen, and ammonia manufacturing plants. That means the investment cost via ammoximation process is much lower than that of the current process.

The key issues are summarized in the table 2.1. These data clearly indicate that the new route is much cleaner than the current one.

Current process Ammoximation process

Oximation

(NH3OH)2SO4

NO

(NH4)2SO4

NH3

H2 H2SO4aq

O2

NH3

O NOH

Oximation

H2O2

O

H2

Air NH3

NOH

Figure 2.3. Comparison of cyclohexanone oxime manufacturing between the current and the ammoximation processes.

(17)

Review of ammoximation

2.3. A review of ammoximation processes

2.3.1. A history and overview of ammoximation reaction

Lebedev et al. [2.3] reported an interesting experimental result to form cyclohexanone oxime by reaction of cyclohexanone, ammonia and hydrogen peroxide in liquid phase. Sodium tungstate was used as a catalyst.

The first patent concerning the method to obtain cyclohexanone oxime directly by the reaction (4.1) was published from Toa Gosei in 1967 [2.4].

12-Tungstophosphate was used as a catalyst. An improved catalyst named TS-1 (Titanium silicate having MFI framework) was invented for the reaction by Roffia et al.

(Montedipe, now Enichem) [2.5]. They have been developing the process, and operated

NH3 2 H2O (1)

+ + H2O2 +

O NOH

(4.1)

Table 2.1. Comparison between a current and the new routes

Product Process

(Intermediate) Current New

cyclohexanone 1) Yield is 79% based on benzene.

2) Many kinds of byproducts are produced. Followings are examples.

 1) Significant amount of Cyclohexane is formed, however it can be used commercially.

2) The yield can be considered as 98%

based on benzene, because the by- product cyclohexane is a valuable chemical product.

3) The conditions of the hydration reaction are mild and the process is safer

3) Waste-products treatment is heavier  than that of the new process.

cyclohexanone oxime 1) Hydroxylammoiun sulfate is employed, therefore large amount of ammonium sulfate is produced.

1) Hydrogen peroxide is required.

2) The investment cost is much lower than that of the current process.

caprolactam  1) Oleum or sulfuric acid is required, hence large amount of ammonium sulfate is produced.

1) a fine catalyst is employed instead of sulfuric acid.

2) Non of ammonium sulfate is formed.

O O

OH OH

O OH OH

O

OH OH O

O

OH OH

O

O

OH OH O

O

OH O

HO O

O

O O

OH

O

O

OH O

HO

CO CO2

(18)

a demonstrative plant to verify the process and the product quality [2.6, 2.7].

A method in which oxygen is used as an oxidant instead of hydrogen peroxide has also been developed. According to a patent published by Allied Chemical [2.8], amorphous silica gel was used as a catalyst in vapor phase at 194℃, and cyclohexanone converted to oxime with 54% conversion and 51% selectivity. Ammoximation by oxygen has been investigated by Trifiro’s group. They recently reviewed the catalytic behavior of silica catalysts for the vapor phase ammoximation reaction [2.9].

Liquid phase ammoximation using molecular oxygen as an oxidant under mild conditions (90℃) has been also investigated. Mantegazza et al. [2.10] reviewed and compared three ammoximation methods; (i) in liquid phase with hydrogen peroxide as an oxidant (Enichem process), (ii) in liquid phase with molecular oxygen, and (iii) in vapor phase with molecular oxygen. They suggested that the reaction mechanisms with molecular oxygen in liquid phase and in gas phase were different, and hinted that in the liquid phase process, H2O2 could be formed in situ, and contribute to the ammoximation reaction in the presence of the TS-1 catalyst.

Fan et al. [2.11] reported an interesting route to make the oxime directly from cyclohexanone and ammonia without using any oxidant. They demonstrated the liquid phase reaction between cyclohexanone and ammonia with Na type zeolite catalysts such as Na-ZSM-5, NaY, and NaA, in an autoclave. They obtained cyclohexanone oxime with 20.3% conversion and 30% selectivity. The reaction was conducted at 300℃ in the presence of benzene. This reaction is probably different from ammoximation, because no oxidants are used.

2.3.2. Ammoximation with hydrogen peroxide

Titanium sites in the TS-1 are believed to be the active site and some reaction mechanisms have been proposed. Petrini et al. [2.12] studied the stability and the deactivation behavior of TS-1 during its use as a catalyst in cyclohexanone ammoximation, and revealed that there were three causes for the deactivation. (i)Silicon species dissolve in the ammonia alkaline aqueous solution accompanying Ti migration, and cause the accumulation of Ti on the external surface of the remaining solid. (ii) Accordingly, by the removal of Ti from the framework, the coordination number of Ti changes from four to six. (iii) The catalyst can irreversibly adsorb some by-products of

(19)

the reaction resulting in a blockage of active sites. Two reaction paths for the ammoximation have been proposed as shown in figure 2.4.

Cyclohexanone reacts firstly with ammonia, then converted into cyclohexylimine, and finally the imine is oxidized into oxime by hydrogen peroxide on TS-1 (Path 1 in figure 2.4) [2.13, 2.14]. In Path 2 (figure 2.4), hydrogen peroxide oxidizes ammonia to hydroxylamine on Ti sites of TS-1, then hydroxylamine reacts with cyclohexanone and finally the oxime is produced. The second path seems today to be the correct one [2.15, 2.16, 2.17].

Zecchina et al. [2.16] reviewed the catalysis of the reaction by means of physical methods (IR, Raman, UV-Vis and XAFS spectroscopies) and supported the Path 2. They concluded that the hydroxylamine produced on the Ti site can directly interact with cyclohexanone inside the channels to give the oxime, or can migrate towards the exterior of the microcrystal to react with ketones. They also proposed a mechanism for the formation of hydroxylamine (figure 2.5).

Mantegazza et al. [2.18] clearly showed that hydroxylamine was formed by the reaction between ammonia and hydrogen peroxide on TS-1. These results strongly support Path 2 in figure 2.4.

Figure 2.4. Reaction paths for the ammoximation of cyclohexanone.

O

NOH

O NH

Path 1

+ NH3 TS-1

+ H2O2

NH3 NH2OH

TS-1 + H2O2 Path 2

(20)

2.3.3. By-products through the ammoximation process using hydrogen peroxide The quality of caprolactam strongly depends upon the impurities present in the cyclohexanone oxime, and therefore the information about the impurities of the oxime is one of the very important matters from the industrial point of view. It is well known that some by-products can be produced during the ammoximation reaction of cyclohexanone. The organic impurities which are expected to be produced during the reaction are shown in table 2.2. All impurities listed in table 2.2 are supposed to be produced by non-catalytic homogeneous reactions, and titanium active sites in TS-1 are not involved in their formation [2.19]. This means that the catalytic reaction to form oxime and the homogeneous side reactions are competitive. Therefore, the catalysts are desired to be highly active for obtaining cyclohexanone oxime in high yield and preventing the formation of impurities. Some inorganic by-products such as nitrogen, nitrous oxide, ammonium nitrite, and ammonium nitrate are also produced (Cesana et al.

[2.20]).

T i H2O ( N H3) O

OH H2O (NH3) O

O

T i H2O ( N H3) O

O OOH O

H2O (NH3) T i

O H2O ( N H3) O O

O

- NH4+ H2O (NH3) O

NH3

NH2OH

H2O2

Figure 2.5. Reaction scheme for formation of hydroxyl amine.

(21)

Compounds Formulas Literature  

cyclohexenone oxime C6H9NO

[2.19]

[2.20]

 

nitrocyclohexane

C6H11NO2 [2.19]

[2.20]

[2.21]

 

cyclohexenylcyclohexanone

C12H18O [2.19]

[2.20]

 

cyclohexylcyclohexanone

C12H20O [2.14]

 

2-cyclohexylidenecyclohexane

C12H18O [2.14]

 

peroxydicyclohexylamine

C12H21NO2 [2.14]

cyclohexanone azine

C12H20N2 [2.19]

[2.20]

[2.21]

NOH

N O O

O

O

O

O O

N H

N N

References

[2.1] M. Kono, Y. Fukuoka, K.Kakagawa, H.Nagahara and H.Ishida, Kagaku to Kogyo, 46 (1993) 429.

Table 2.2. Impurities expected to be formed during the cyclohexanone ammoximation

(22)

[2.2] G. Petrini, G. Leofanti, M. A. Mantegazza and F. Pignataro, ACS Symp. Ser., 626 (Green Chemistry), 33-48 1996.

[2.3] O. L. Lebedev and S. N. Kazarnovskii, Zhur. Obshchei Khim. 30 (1960) 1631.

[2.4] E. Yasui, T. Kawaguchi, T. Matsubara and H. Kato, Ger. Patent 1 245 371 (1967) to Toa Gosei Chem. Ind.

[2.5] P. Roffia, M. Padovan, E. Moretti, and G. De Alberti (Montedipe), Eur. Patent 208 311 (1987) to Montedipe S.p.A.

[2.6] European Chemical News, 23 (6-12 February 1995).

[2.7] V. Alessi, R. Penzo, S. Pattaro, R. Tessari and M. J. Slater, Value Adding Solvent Extr. [Pap.ISEC’96], 2 (1996) 1673.

[2.8] J. N. Armor, US Patent 4 163 756 (1979) to Allied Chem. Corp.

[2.9] G. Fornasari, and F. Trifiro, Catal. Today, 41 (1998) 443.

[2.10] M. A. Mantegazza, G. Paparatto, G. Petrini, G, Fornasari and F. Trifiro, Chem.

Ind.(Dekker), 62 (1995) 353.

[2.11] L. Fan, J. Liu and K. Fujimoto, Sekiyu Gakkaishi, 42 (1999) 201.

[2.12] G. Petrini, A. Cesana, G. De Alberti, F. Genoni, G. Leofanti, M. Padovan, G.

Paparatto and P. Roffia, Stud. Surf. Sci. Catal., 68 (1991) 761.

[2.13] Z. Tvaruzkova, K. Habersberger, N. Zilkova and P. Jiru, Appl. Catal. A, 79 (1991) 105.

[2.14] A. Thangaraj, S. Sivasanker and P. Ratnasamy, J. Catal., 131 (1991) 394.

[2.15] A. Zecchina, G. Spoto, S. Bordiga, F. Geobaldo, G. Petrini, G. Leofanti, M.

Padovan, M. Mantegazza and P. Roffia, Stud. Surf. Sci. Catal., 75 (1993) 719.

[2.16] A. Zecchina, S. Bordiga, C. Lamberti, G. Ricchiardi, C. Lamberti, G. Ricchiardi, Scarano, G. Petrini, G. Leofanti and M. Mantegazza, Catal. Today, 32 (1996) 97.

[2.17] J. Le Bars, J. Dakka, and R. A. Sheldon, Appl. Catal. A, 136 (1886) 69.

[2.18] M. A. Mantegazza, G. Leofanti, G. Petrini, M. Padovan, A. Zecchina and S.

Bordiga, Stud. Surf. Sci. Catal., 82 (1994) 541.

[2.19] M. A. Mantegazza, A. Cesana and M. Pastori, Topics Catal., 3 (1996) 327.

[2.20] A. Cesana, M. A. Mantegazza and M. Pastori, J. Mol. Catal. A, 117 (1997) 367.

[2.21] J. P. schimann, F. weiss, Tetrahedron Lett. 7 (1972) 635.

(23)

Chapter 3

A STUDY AND DEVELOPMENT OF THE CATALYSTS FOR THE SELECTIVE HYDROGENATION OF BENZENE 3.1. Background of the study

It is well known that cyclohexene can be converted into cyclohexanol by hydration reaction with the aid of an acid catalyst. Thus, if we can obtain cyclohexene economically, the process to produce cyclohexanol via cyclohexene will be very interesting for caprolactam manufacturers.

When the author started the development of the cyclohexene manufacturing process in 1983, however, no effective method to produce cyclohexene had been proposed in an industrial sense.

Bond mentioned in his book "Catalysis by Metals" [3.1] that it was unanimously found that the only isolatable product from the hydrogenation of benzene was cyclohexane. In the book he described the reason as follows. From the heat of complete hydrogenation of benzene and possible intermediates given in table 3.1, it appears that the formation of 1,3-cyclohexadiene is endothermic by 5.6 kcal/mol at 82 ℃ : Now the heat of hydrogenation of the hypothetical non-resonating cyclohexatriene may be estimated as three times the value for cis-2-butene, viz. 85.7 kcal/mol, and the formation of cyclohexadiene from benzene is endothermic because the heat of hydrogenation of benzene is less than the "expected" value by the resonance energy of 36 kcal/mol. Since the formation of cyclohexene from benzene is exothermic, its absence from the products of the hydrogenation must either be due to its greater reactivity or be a consequence of the mechanism.

(24)

Values of ΔG for the formation of products from benzene and hydrogen at 25℃ are as follows.

1,3-cyclohexadiene; +13.2 kcal/mol cyclohexene; -4.7 kcal/mol

cyclohexane; -23.4 kcal/mol

These data suggest that the formation of cyclohexane is most favorable in the hydrogenation of benzene.

The difficulty to manufacture cyclohexene by a catalytic partial hydrogenation of benzene was generally accepted, because benzene is stabilized by the resonance energy, and therefore it cannot be hydrogenated easily. In contrast with this, the C=C double bond of cyclohexene is easily hydrogenated. Thus, it can be concluded that it is very difficult to obtain cyclohexene in good yield by the hydrogenation of benzene.

The author reviewed the studies on benzene hydrogenation and found several interesting results. Hartog et al. [3.2] reported that they obtained cyclohexene in 0.18%

yield by the hydrogenation of benzene with a Ru black catalyst in 1963. An interesting patent was issued from DuPont in 1973, in which they reported that cyclohexene could be obtained from benzene in good yield (Conversion=50%, Selectivity=65%, Yield=32.5%) when the hydrogenation was conducted with water [3.3]. Since then, some patents and papers appeared.

The author reviewed papers and patents, and discussed the reaction scheme with his colleagues. We finally made our own hypothesis on how to develop a fine catalyst for selective hydrogenation of benzene to cyclohexene. Our catalysts search was based on the hypothesis described in Section 3.3.

Table 3.1. Heats of complete hydrogenation of benzene and possible intermediates

Compound -ΔHh (kcal/mol)

Benzene 49.80

1,3-Cyclohexadiene 55.37

Cyclohexene 28.59

(25)

3.2. Outline of a cyclohexanol manufacturing process via cyclohexene from benzene

Cyclohexene produced is sent to the separation system, where benzene, cyclohexene, and cyclohexane are separated. The boiling points of the three components are very close; benzene=80.13, cyclohexene=83.3, cyclohexane=80.8℃. Therefore, it is impossible to separate them with an ordinary distillation technique; an azeotropic distillation should be applied. The scheme of the distillation is shown in figure 3.1. The mixture of the three components is fed into the first column with a solvent, such as trimethylphosphte, dipropyleneglycol, etc [3.4]. Cyclohexane is obtained from the top of the first column. In the second column, cyclohexene is obtained, and finally benzene and the solvent are recovered.

Cyclohexene is converted to cyclohexanol by a hydration reaction with the help of an MFI zeolite catalyst. Many acid catalysts are known for hydration of olefins.

Homogeneous catalysts such as sulfuric acid or aryl sulfonic acid produce cyclohexanol in good yield. The separation of cyclohexanol, however, is somewhat difficult.

The author's colleagues found very interesting phenomena when the hydration was carried out with a solid acid catalyst; the addition of a third component such as cresol to the reaction mixture enhances the conversion of cyclohexene and cyclohexanol was obtained with good yield. For example, the conversion is higher than 30% and the

solvent solvent

solvent + + Solvent + Solvent

solvent solvent

solvent + + Solvent + Solvent

Figure 3.1. Separation of products.

(26)

selectivity is almost 100% [3.5].

The mixture produced is sent to a decanter and separated into oil and water layers. The catalyst is placed in the latter. Water and the catalyst are returned into the reactor and the oil layer is distilled. Thus, cyclohexanol is produced.

3.3. Working hypotheses for searching a fine selective hydrogenation catalyst Judging from the literature, the author and his colleagues made hypotheses for searching a fine catalyst. The following is the digest.

(1) Water addition in the reaction system might improve the catalyst performance.

(2) The mechanism to extract the product smoothly from the surface of the catalyst should be employed for preventing further hydrogenation of cyclohexene. Maybe, when water is added in the reaction system, and a hydrophilic catalyst is used, the product moves from the catalyst surface to the oil layer through the water layer. Thus the consecutive hydrogenation will be hindered.

(3) Group VIII precious metals (Ru, Rh, Pd, Ir, and Pt) can be candidates; especially Ru might be anticipated as a fine catalyst.

(4) The modification of catalysts by addition of some metal components will improve the catalyst performance.

H2O

OH

CH3 OH

OH

H2O + Catalyst

+ H2O

H2O

OH

CH3 OH OH

CH3 OH

OH

H2O + Catalyst

+ H2O

Figure 3.2. Hydration and separation system.

(27)

3.4. Experimental procedure

Catalysts are prepared as in the methods described in the patents [3.6]. A typical example to make 2%Ru/BaSO4 is as follows: 0.190 g of RuCl3.3H2O is added to a 500 ml flask, and 200 ml of water is added to dissolve it. After adding 3.6 g of BaSO4, the flask is mounted on a rotary evaporator. The BaSO4 is impregnated with the aqueous solution with stirring at room temperature for 1 hour, and then at 60 for 1 hour.

Thereafter, water is evaporated at 80 under reduced pressure.

The dry-up product obtained is filled in a Pyrex glass tube of 5 mm inside diameter, heated to 200 while passing hydrogen gas at a rate of 100 ml/min and kept at this temperature for 4 hours to activate the catalyst. The composition of the catalyst obtained is 2%Ru/BaSO4.

Selective hydrogenation reaction is carried out with an autoclave as described in the patents [3.6]. Fifty ml of water are added to a 100 ml stainless steel autoclave, in which the air is sufficiently replaced by argon in advance, and 0.5 g of a catalyst that is previously reduced by hydrogen gas at 200 and 15 ml of benzene are added in this order. Hydrogen gas is introduced in the autoclave under a pressure of about 4 MPa.

The reaction temperature is about 140-180. The reactants are stirred vigorously. The products are analyzed with vapor phase chromatography.

3.5. Experimental results and discussion 3.5.1. The search for carrier materials

Ru metal was selected for the main component and carrier materials were searched. Two weight percent of Ru on a carrier such as MgO, CaO, SrO, BaO, Al2O3, SiO2, SnO2, ZnO, CdO, TiO2, ZrO2, V2O5, Nb2O5, Ta2O5, La2O3, and Ce2O3 were tested.

The experimental results are listed in figure 3.3.

Where the electronegativity χi is adopted from the following equation [3.7].

χi = (1 + 2z)/χ0

z: Charge of metal cation

χ0: Electronegativity of neutral metal atom

The authors expected that some correlation would be observed between the selectivity and the electronegativity of the metallic cations of carrier materials. However, no special relationship could be obtained as indicated in figure 3.3.

(28)

Two kinds of MgO, one was received from a chemicals supplier and the other was prepared by the hydrolysis of magnesium t-buthoxide. These materials were used as a carrier and compared their catalytic activities. The results are also indicated in figure 3.3. Ru on silica catalysts that were prepared from silica gel or silica made from tetraethyl orthosilicate were tested and the results are also shown in figure 3.3. Large differences are observed in these catalysts. The data clearly imply that the catalytic activity or selectivity strongly depends on the preparation methods. Among the carriers tested, Ru on ZnO catalyst gave a good selectivity to cyclohexene. We were very interested in the catalyst and investigated the role of the ZnO carrier. The results are described in the next section.

0 10 20 30 40 50

0 5 10 15 20

BaO CaO

SrO MgO MgO (Mg(OtBu)2

ZnO

Al2O3

CeO2 ZrO2

Ta2O5 TiO2

SiO (Si(OEt)4)

SiO2 Nb2O5 V2O5 SnO2

Selectivity (%)

MnO2 La2O3 CdO

Figure 3.3. Selectivity to cyclohexene at benzene conversion=50%.

Reaction conditions: catalyst=0.5g, water=50ml, benzene=15ml temperature=180, H2 pressure=4.0MPa,

Electro negativity χi

(29)

3.5.2. Effects of the addition of some metal salts aqueous solution into the reaction system

The author and his colleagues analyzed the water layer of the product when Ru/ZnO catalyst was employed, and found that small amounts of ZnCl2 existed in the layer. The Cl- anion came from the raw material (RuCl3・xH2O) used for the catalyst preparation. When the catalyst precursor was reduced by hydrogen at an elevated temperature, HCl was eliminated from RuCl3 in the precursor and HCl reacted with ZnO. Thus, ZnCl2 was contained in the catalyst. We supposed that metal salts such as ZnCl2 promoted the selectivity to cyclohexene, so we tested several metal salts as additives. We found that not only ZnCl2 but also other metal salts such as CaSO4 could promote the selectivity [3.6]. The results when aqueous solution of CoSO4 is used with 2%Ru/Al2O3 catalyst are shown in figure 3.4 as compared with water and without water.

0 20 40 60 80

0 20 40 60 80 100

Conversion (%)

Selectivity (%)

CoSO4 aq Water

Without H2O

Figure 3.4. The effect of CoSO4 aqueous solution addition in the reaction system.

Catalyst: 2%Ru/Al2O3 0.5g Benzene: 15ml H2O or 3.5%CoSO4 aq: 15ml Reaction temperature: 180℃  Hydrogen pressure: 4.0MPa

(30)

For the purpose of better understanding the effect of CoSO4 aq solution, the hydrogenation reactions of benzene and cyclohexene were carried out with a 2%Ru-1%Fe/BaSO4 catalyst as a function of reaction time with the three following cases; water is added, not added, and CoSO4 aq is added in the reaction system. The dependence between the reaction time and the amounts of reacted benzene or cyclohexene are shown in figure 3.5.

When there is no water addition, the conversion of benzene is very low.

However, when some water or CoSO4 aq solution is added to the reaction system, the reaction velocity of benzene accelerated much more. It can be emphasized that water promotes the hydrogenation of benzene and CoSO4 does not affect the reaction velocity of benzene with water.

On the contrary, the velocity of cyclohexene hydrogenation is highly reduced when CoSO4 aq solution is employed. These results clearly describe why CoSO4 aq solution improves the selectivity to cyclohexene.

Figure 3.5. Hydrogenation of benzene (a) and cyclohexene (b) vs reaction time.

▲: no addition

●: 3.5% CoSO4 aqueous solution (volume ratio of the aq solution and benzene or cyclohexene = 1.0)

○: water (volume ratio of water and benzene or cyclohexene = 1.0)

Catalyst; 2%Ru1%Fe/BaSO4 200mg Benzene or Cyclohexene; 15ml Reaction temperature; 180℃  Hydrogen pressure; 4.0MPa

0 0.2 0.4 0.6 0.8

0 0.2 0.4 0.6

0 0.2 0.4 0.6 0.8

0 1 2 3 4 5

Amount of benzene reacted (mole/g-cat.)

Reaction time (h) Reaction time (h)

(a) (b)

Amount of cyclohexene reacted (mole/g-cat.)

(31)

3.5.3. Reaction scheme

The reaction scheme of selective benzene hydrogenation reaction can be regarded as a consecutive and parallel reaction as is shown in figure 3.6 [3.8].

In this scheme, when water is absent in the reaction system, the rate of benzene, namely k1 plus k2, is very low, but the reaction rate of cyclohexene k3 is high. Therefore cyclohexene cannot be obtained in good yield.

When water is present in the reactor, the rate of benzene hydrogenation is very high and the reaction rate of cyclohexene (k3) is almost the same as in the absence of water. As a result, the selectivity to cyclohexene is enhanced. Cobalt sulfate decreases the rate of cyclohexene consumption (k3) without affecting the reaction of benzene (k1 + K2). Therefore, the selectivity to cyclohexene increases particularly at high conversion of benzene.

k1

k2

k3

Figure 3.6. Reaction scheme of benzene hydrogenation.

(32)

Figure 3.7 is an image of the diffusion and reaction steps near and on the catalyst surface. The catalyst surface is surrounded by the water layer. Benzene is dissolved into water and adsorbs on the catalyst, then it is hydrogenated to cyclohexene.

Cyclohexene is extracted by an oil layer immediately. Thus, the chance for cyclohexene to be converted to cyclohexane is decreased. The solubility of benzene into water is about ten times larger than cyclohexene or cyclohexane. Niwa et al. [3.9] reported the importance of water addition into the reaction system, and the solubility data of the three components at 150℃ under 70 kg/cm2 pressure. The data can be read as follows from the figure in the paper.

Benzene; 0.2g/100gH2O Cyclohexene; 0.02g/100gH2O Cyclohexane; 0.02g/100gH2O

Fukuoka et al. [3.10] reported almost the same reaction scheme of selective hydrogenation of benzene in a journal. They studied the hydrogenation of benzene-d6, and proposed a new concept concerning the reaction mechanism and the role of the co-catalyst (a zinc salt).

3.5.4. The results of catalyst search

We understood that the hydrophilic property of the catalyst was very important for the selective hydrogenation of benzene as illustrated in figure 3.7, because the

H2gas phase

Oil layer

Water layer

Catalyst

Figure 3.7. Imaginative reaction scheme of selective hydrogenation of benzene.

(33)

surface of the catalyst had to be covered with water. Therefore we searched for several hydrophilic carrier materials and finally found BaSO4. BaSO4 is very hydrophilic and the density of the crystal is large, so this material exists in the water layer and is very easily separated from the reaction products by decantation. These properties are very favorable for the plant operation.

The Ru/BaSO4 catalyst gave cyclohexene in a good selectivity. We tried to improve the catalyst by means of adding some components on the catalyst. Finally, Ru-Co-Cu/BaSO4 catalyst system was obtained [3.6]. Typical reaction results are shown in comparison with a 2%Ru/Al2O3 catalyst in figure 3.8. Thus, it is proved that cyclohexene can be produced by the partial hydrogenation of benzene with more than 75% selectivity.

(34)

C onversion (%)

Selectuvity (%)

0 20 40 60 80 100

0 20 40 60 80 100

Figure 3.8. Selectivity to cyclohexene vs. conversion of benzene.

■:2%Ru-2%Co-0.2%Cu/BaSO4 (+3.5%CoSO4 aq solution) 

●:2%Ru/Al2O3 (+3.5%CoSO4 aq solution) Reaction conditions

Catalyst charged: 0.2〜0.5g, Benzene: 15ml, 3.5% CoSO4 aq solution: 50ml Reaction temperature: 160℃, Hydrogen pressure: 4.0MPa

(35)

3.6. Conclusions

(1) The catalyst composed of Ru, Co, Cu and BaSO4 gives cyclohexene in a good yield.

The relationship between the conversion and selectivity is for an example as follows:

Conversion=20%, Selectivity=80%

Conversion=35%, Selectivity=75%

Conversion=50%, Selectivity=70%

(2) Water promotes the conversion of benzene, and some metal salts such as ZnCl2 and CoSO4 suppress the hydrogenation of cyclohexene to cyclohexane. The improvement of the selectivity in the case of a CoSO4 aqueous solution added in the reaction system can be explained with this phenomenon.

(3) Our working hypothesis for searching a fine catalyst was useful to develop a good catalyst.

(4) Only Cyclohexane is produced as a byproduct, which is a useful material for the chemical industry. Therefore the process to manufacture cyclohexene is 100%

effective in the industrial sense of view.

References

[3.1] G. C. Bond, Catalysis by Metals, p314, Academic Press London and New York 1962.

[3.2] F. Hartog and Zwietering, J. Catalysis, 2 (1963) 79-81.

[3.3] W. C. Drinkard, USP3,767,720 (1973) to DuPont.

[3.4] Y. Ori and O. Moriya, Japanese Patent, 1671676, Japanese Unexamined Patent, S62-123136, S62-123135 to Sumitomo Chemical Co., Ltd.

[3.5] T. Shirafuji, K. Sakai and K. Hirose, US4,716,253 (1987) to Sumitomo Chemical Co., Ltd.

[3.6] H. Ichihashi and H. Yoshioka, US4,575,572 (1986), US4,665,274 (1987) to Sumitomo Chemical Co., Ltd.

[3.7] K. Tanaka and A. Ozaki, J. Catalysis, 8 (1967) 1-7.

[3.8] H. Ichihashi and H. Yoshioka, Joint Symposium between Chugoku-Shikoku and Kyushu branch unions of the Chemical Society of Japan, Tokushima, 1986, 1D12.

(36)

[3.9] S. Niwa, F. Mizukami, M. Taba, T. Tsuchiya, K. Shimizu and J. Imamura, Shokubai, 31 (1989) 421-424.

[3.10] Y. Fukuoka, H. Nagahara and M. Konishi, Shokubai, 35 (1993) 34-39.

(37)

Chapter 4

THE DEVELOPMENT OF A CATALYST FOR THE VAPOR PHASE BECKMANN REARRANGEMENT PROCESS

4.1. A history and overview of the study on the vapor phase Beckmann rearrangement

The Beckmann rearrangement (equation 4.1) is a typical acid catalyzed reaction. Many attempts to replace sulfuric acid with a solid acid catalyst have been done.

Since Lazier and Rigby (Du Pont) [4.1] demonstrated in 1938 that caprolactam could be produced from cyclohexanone oxime with the aid of SiO2 gel catalyst in vapor phase, many catalysts have been searched. After Dawydoff (Leunawerke) [4.2] verified that the catalysts mainly composed of B2O3 gave an excellent selectivity to caprolactam with a very good conversion of cyclohexanone oxime, BASF [4.3], Bayer [4.4], and many other laboratories intensively investigated and developed vapor phase catalytic processes. According to the patents [4.3, 4.4], caprolactam was produced in more than 95% yield. The conversion and selectivity were very high, however, the vapor phase Beckmann rearrangement process based on the B2O3 catalysts has not yet been commercialized. B2O3 is easily volatile at an elevated temperature. Hence, the life of the catalyst cannot be expected long enough for its industrial use, and vaporized B2O3 is supposed to cause difficulties of purification and wastewater treatment.

Catalysts other than B2O3 have also been developed. In table 4.1 the author cites some examples from the patents applied since 1986. Many catalysts have been proposed, but the features of these catalysts are still not clear.

The author and his colleagues developed a fine catalyst mainly composed of a high silica MFI zeolite for the reaction and Sumitomo Chemical Co., Ltd. constructed a caprolactam manufacturing plant with the catalyst in 2003.

NOH

N O

H

(4.1) (2)

(38)

4.2. The outline of Sumitomo's vapor phase Beckmann rearrangement process The outline of Sumitomo's vapor phase Beckmann rearrangement process developed by us is shown in figure 4.1. Cyclohexanone oxime is vaporized and is fed into a reactor with methanol vapor, in which a catalyst mainly composed of high silica MFI zeolite is loaded. The reaction temperature is between 350 and 400, the pressure is around ambient. The gaseous products from the reactor are cooled, and methanol vapor is recovered and recycled into the reactor. Then the products are purified, and pure and high quality caprolactam is obtained. The addition of methanol vapor into the reactor is one of the main features of the process.

During the reaction, some amounts of carbon like material are deposited on the catalyst. Therefore it is necessary to remove the carbonaceous materials by aeration at an elevated temperature. A fluidized bed reaction system (figure 4.2) is introduced because it can conduct the reaction and regeneration of the catalyst continuously [4.5].

Cyclohexanone oxime is converted to caprolactam in good yield and high

Cyclohexanone

oxime Vaporization Beckmann rearrangement

Methanol

Recovery Purification ε-Caprolactam Cyclohexanone

oxime Vaporization Beckmann rearrangement

Methanol

Recovery Purification ε-Caprolactam

Figure 4.1. Sumitomo Chemical's Vapor Phase Beckmann Rearrangement Process.

Table 3 Som e catalysts for the vapor phase Beckm ann rearrangem ent 

Patent Patent Year W HS V C onversion Selectivity

assignee Num ber priority C atalyst Solvent 1/hr % %

UO P US4873325 1986 SAPO -11 acetonitrile 0.54 98 95

Sum itom o US4709024 1986 high silica M FI benzene 3 74 72

Sum itom o US4968793 1989 high silica M FI m ethanol 3.3 99 87

M obil US4927924 1989 ZSM -5 benzene 0.05 99 89

M itsubishi EP509493 1991 Ta2O5/SiO2 benzene 1 96 97

Sum itom o JP9-291074 1996 ALPO -5 ethanol 0.45 27 88

Enichem EP819675 1996 am orphous S iO2/Al2O3 m ethanol 2 99 78

Ube JP10-87612 1996 zeolite-L n-hexanol 0.83 99 97

Table 4.1 Some catalysts for the vapor phase Beckmann rearrangement

(39)

ε-Caprolactam

REACTOR

REGENERATOR

N2 Air

Cyclohexanone oxime Methanol

Off gas ε-Caprolactam

REACTOR

REGENERATOR

N2 Air

Cyclohexanone oxime Methanol

Off gas

Figure 4.2. Fluidization reaction system.

efficiency in the reactor. A part of the catalyst moves from the reactor to the regenerator and is treated by air. Then the regenerated catalyst returns from the regenerator to the reactor. The catalyst circulates continuously through the reactor and the regenerator thus the system is operated continuously.

4.3. Experimental procedure

The zeolite samples were synthesized in our laboratory as described in patents [4.6, 4.7, 4.8].

The vapor phase rearrangement reaction tests were carried out with fixed bed type micro-reactors. A mixture of cyclohexanone oxime and methanol and/or benzene was fed into the reactor (10 mm inner diameter, made of silica glass) with nitrogen as a carrier gas. The reaction product was collected under ice cooling and assayed by gas chromatography (HEWLETT PACKARD model HP 6890, and SHIMADZU model 9A).

The conversion of cyclohexanone oxime and the selectivity to caprolactam were calculated as follows: cyclohexanone oxime conversion (%) = {(moles of cyclohexanone oxime charged – moles of unaltered cyclohexanone oxime) / (moles of cyclohexanone oxime charged)} × 100; selectivity to caprolactam (%) = {(moles of caprolactam in product) / {(moles of cyclohexanone oxime charged – moles of unaltered cyclohexanone oxime) } × 100.

参照

関連したドキュメント

Answering a question of de la Harpe and Bridson in the Kourovka Notebook, we build the explicit embeddings of the additive group of rational numbers Q in a finitely generated group

Next, we prove bounds for the dimensions of p-adic MLV-spaces in Section 3, assuming results in Section 4, and make a conjecture about a special element in the motivic Galois group

Transirico, “Second order elliptic equations in weighted Sobolev spaces on unbounded domains,” Rendiconti della Accademia Nazionale delle Scienze detta dei XL.. Memorie di

In our previous paper [Ban1], we explicitly calculated the p-adic polylogarithm sheaf on the projective line minus three points, and calculated its specializa- tions to the d-th

Our method of proof can also be used to recover the rational homotopy of L K(2) S 0 as well as the chromatic splitting conjecture at primes p > 3 [16]; we only need to use the

Classical definitions of locally complete intersection (l.c.i.) homomor- phisms of commutative rings are limited to maps that are essentially of finite type, or flat.. The

Yin, “Global existence and blow-up phenomena for an integrable two-component Camassa-Holm shallow water system,” Journal of Differential Equations, vol.. Yin, “Global weak

This paper presents an investigation into the mechanics of this specific problem and develops an analytical approach that accounts for the effects of geometrical and material data on