Fully Automatic Computer-Operated Reaction System in Laboratory (Part 1)

Design and Development of the System and Examples of Feed-back Optimization Experiments

Abstract

This report deals with a fully automatic computeroperated reaction system (FACORS) which was designed and developed for a wide variety of automatic experiments, with or without on-line analysis and feedback of data, in research on catalytic or on non-catalytic reactions carried on in a laboratory-scale continuous flow reaction apparatus.
Fig. 1 shows a schematic diagram of FACORS designed to automate the following seven processes which are very common in various experiments in the laboratory:
(1) controlling variables such as reactor, catalystbed temperatures; flow rates, compositions, and pressures of fluids; process times; the paths of flowing substances; etc., (2) sampling products and operations for analysis by such means as gas chromatograph and mass spectrometer, etc., (3) measuring and recording process variables, (4) logging analytical data, (5) summarizing data obtained and preparing tables and graphs, (6) performing analysis and comparison of the data, and (7) deciding whether or not to continue, change, or stop the experiments based on the feedback of the results from (6) and others, and deciding the experimental conditions for subsequent experiments.
The automation of these (1)-(7) processes by use of a single computer (see the legend of Fig. 1) made it possible to perform a variety of fully automatic, unattended feedback and/or feedforward experiments; such as activity and life tests of catalysts, determination of optimum reaction conditions, repetitive operations such as reaction/regeneration or adsorption/desorption cycles, various temperature programmed experiments for examination of the properties of catalysts.
Two examples of unattended feedback experiments, Ex-(A) and Ex-(B), were carried out by use of FACORS to reveal the advantages and characteristics of the system. The Ex-(A) is for the determination of optimum reaction conditions [reaction temperature(T) and flow rate of reactant(F)] to attain the maximum yield of ethylene in Eq. (1). The flow diagram for Ex-(A) is illustrated in Fig. 3 and the results are shown in Figs. 5 and 6, which were drawn by a GP (Fig. 1), and summarized in Table 1. The profiles of T and F (uncorrected values) recorded on a RC (Fig. 1) during the course of Ex-(A) are shown in Fig. 4.
The Ex-(B), which consists of alternate reaction/regeneration cycles of Eqs. (2) and (3), was carried out according to the flow diagram illustrated in Fig. 7. Variation in the concentration of methane produced (y(t)) and time-averaged yield (Y(t)) obtained by online integration of y(t) using Eq. (4) during the course of reaction are shown in Figs. 8 and 9, respectively. At about 5hr after the start of the optimization experiment (see Fig. 7) the computer found the optimum time of reaction ((t)_opt=27±2min) and that of regeneration ((t_o)_opt=12.5±0.2min) with the maximum timeaveraged yield ([Y(t)_m]_M=1.15+0.05μmol•min^-1).
In conclusion, the major advantages of the system developed here are as follows: (i) A wide variety of "unattended" feedback and feedforward experiments are possible. (ii) Experiments with previously unknown objective conditions could be carried out. (iii) Once the programs are designed, even a nonprofessional can readily perform complicated experiments. (iv) The present system gives us well-organized original data as well as analysed ones, which are recorded on a tape or a disk of a computer. They are convenient for storing, carrying, reproducing, re-analyzing, and transforming into tables and graphs. (v) The data obtained are of high accuracy, reliability, reproducibility, and objectivity.