Reactor first stage

Oxidation Step. A review of mechanistic studies of partial oxidation of propylene has appeared (58). The oxidation process flow sheet (Fig. 2) shows equipment and typical operating conditions. The reactors are of the fixed-bed shell-and-tube type (about 3—5 mlong and 2.5 cm in diameter) with a molten salt coolant on the shell side. The tubes are packed with catalyst, a small amount of inert material at the top serving as a preheater section for the feed gases. Vaporized propylene is mixed with steam and ak and fed to the first-stage reactor. The feed composition is typically 5—7% propylene, 10—30%... 

If necessary, first-stage reactor effluent maybe further cooled to 200—250°C by an iaterstage cooler to prevent homogeneous and unselective oxidation of acroleia taking place in the pipes leading to the second-stage reactor (56,59). 

Several variations of the above process are practiced. In the Sumitomo-Nippon Shokubai process, the effluent from the first-stage reactor containing methacrolein and methacrylic acid is fed directiy to the second-stage oxidation without isolation or purification (125,126). In this process, overall yields are maximized by optimizing selectivity to methacrolein plus methacrylic acid in the first stage. Conversion of isobutjiene or tert-huty alcohol must be high because no recycling of material is possible. In another variation, Asahi Chemical has reported the oxidative esterification of methacrolein directiy to MMA in 80% yield without isolation of the intermediate MAA (127,128). 

A necessary feature of the alkylation reaction section is the use of two reactors the first-stage reactor completes the major part of the alkylation reaction, and in the second-stage reactor the last traces of unsaturated hydrocarbons react, and a sizable portion of the soluble polyaromatics is removed. Modem units with lower-diene-containing feeds employ a single alkylation reactor (79). 

In the second stage, a more active 2inc oxide—copper oxide catalyst is used. This higher catalytic activity permits operation at lower exit temperatures than the first-stage reactor, and the resulting product has as low as 0.2% carbon monoxide. For space velocities of 2000-4000 h , exit carbon monoxide... 

First-stage reactor conditions space velocity, vol/vol hr feed gas flow rate, lb/hr recycle flow rate, lb/hr recycle molecular weight reactor temperatures, °C... 

By maintaining the first-stage reactor just beyond the phase inversion point, the dispersed rubber phase is relatively rich in dissolved styrene. As polymerization subsequently proceeds in the LFR s, the dissolved styrene will react to form either a graft copolymer with the rubber or a homopolymer. The latter will remain within the rubber droplet as a separate occluded phase. Achieving the first-stage reactor conversion and temperature by recycling a portion of the hot second reactor effluent may permit simplification of the first reactor temperature control system. 

Diolefins saturation is carried out in a first stage reactor, in the presence of recycled hydrogen. The hydrocarbon stream product is mixed with fresh make-up hydrogen, at... 

After heating, the EB is mixed with superheated steam and fed to the first stage reactor. Both the first and second stage reactors are packed with a catalyst of metal oxide deposited on an activated charcoal or alumina pellets. Iron oxide, sometimes combined with chromium oxide or potassium carbonate, is commonly used. 

In continuous emulsion polymerization of styrene in a series of CSTR s, it was clarified that almost all the particles formed in the first reactor (.2/2) Since the rate of polymerization is, under normal reaction conditions, proportional to the number of polymer particles present, the number of succeeding reactors after the first can be decreased if the number of polymer particles produced in the first stage reactor is increased. This can be realized by increasing emulsifier and initiator concentrations in the feed stream and by lowering the temperature of the first reactor where particle formation is taking place (2) The former choice is not desirable because production cost and impurities which may be involved in the polymers will increase. The latter practice could be employed in parallel with the technique given in this paper. 

Our final goal in the present paper is to devise an optimal type of the first stage reactor and its operation method which will maximize the number of polymer particles produced in continuous emulsion polymerization. For this purpose, we need a mathematical reaction model which explains particle formation and other kinetic behavior of continuous emulsion polymerization of styrene. 

Since the depletion of emulsifier micelles occurs only because they break up and their molecules are adsorbed onto the surface of growing particles. The balance on the micelles in the first stage reactor is given by the following equation if the reaction is started so that the emulsifier concentration in the feed stream does not change with time ... 

Let us consider the steady state characteristics of continuous emulsion polymerization of styrene in the first stage reactor. The steady state value of the number of polymer particles formed in the first stage reactor can be calculated using the following equations. From Eqs. (1) and (2), we have ... 

This is the reason why the steady state value of the number of polymer particles coincide with each other, as shown in Figure 4, regardless of the form of e when the first stage reactor is operated at comparatively longer residence time. On the other hand, if Eq.(23) is used instead of Eq.(16) for calculating Ap value, we have ... 

Figure 5 represents a typical example of the variation of the number of polymer particles with mean residence time 0. The solid line shows the theoretical value predicted by the Nomura and Harada model with e= 1.28x 10 . The dotted line is that predicted by the Gershberg model(or the Nomura and Harada model with Case C for ), where Eq. (23) was used instead of Eq.(16) for Ap. The value of Nt produced at longer mean residence time differs, therefore, by a factor of T(5/3) between the solid and dotted lines in Figure 5. From the comparison between the experimental and theoretical results shown in Figure 5, it is confirmed that the steady state particle number can be maximized by operating the first stage reactor at a certain low value of mean residence time max which is considerably lower than that in the succeeding reactors. This is so-called "pre-reactor principle". It is, therefore, desirable to operate the first reactor at such mean residence time as producing something like a maximum number of polymer particles in order to increase the rate of polymerization in the succeeding reactors. This will result in a decrease in the number of necessary reactors and hence, in the capital cost.

It is apparent from Eq.(30) that the higher the temperature of the first stage reactor and the value of r, the smaller the value of %ax This may be the reason why Degraff and Poehlein could not find 8max in their experiments. Equation(31) suggests that one can estimate the value of max simply by determing tc by measuring surface tension with the use of, for example, a du-Nouy tensiometer. ... 

The maximum number of polymer particles produced in the first stage reactor being operated at max can be obtained by introducing Eq.(30) into Eq. (26). Thus,... 

This means that as long as a CSTR is used as the first stage reactor and all the recipe ingrediants are fed into the first stage reactor, one cannot have more than 57% of the number of particles produced in a batch reactor with the same recipe as in continuous operation. The validity of these expression is clear from the comparison between the experimental and theoretical values shown in Figure 5. From Figure 5, it is found that the optimum mean residence time of the first stage reactor is about 10 minutes under these reaction conditions. Equation(30) predicts 10.0 minutes, while experimental value is 10.4 minutes where the number of polymer particles is about 60% of that produced in a batch reactor. 

Another method to increase the number of polymer particles produced in the first stage reactor with initiator and emulsifier concentrations fixed is to employ a plug flow type reactor such as a tubular reactor for the first stage. The minimum residence time of a plug flow reactor 6 necessary to produce the same number of polymer particles as in E batch reactor is tc. Thus, from Eq.(31) We have ... 

MF = 0.01 g/cc-water because the residence times of both reactors are fixed at 0= 20 minutes. From this theoretical and experimental results, therefore, a plug flow type reactor with a divided monomer feed is recommended for the first stage reactor(pre-reactor), because the volume of the reactor can be decreased by decreasing monomer concentration in a feed stream. Nevertheless, the steady state particle number attained in this reactor can be increased. 

A small amount of water which just dissolves initiator and emulsifier is fed into the first stage reactor along with all the initiator and emulsifier, and the rest of water into the second stage reactor. 

The number of polymer particles formed in the first stage reactor can be generally expressed by the following form ... 

---