Engineering Education

A Chemical Reaction Engineering Laboratory Experiment:

Isothermal Laminar-Flow Reactor

Nabeel S. AboGhander and Abdullah A. Shaikh∗

Chemical Engineering Department
King Fahd University of Petroleum & Minerals
Dhahran 31261, Saudi Arabia

Abstract

We report here an experiment for the chemical reaction engineering laboratory. The
reaction between sodium hydroxide and ethyl acetate is conducted isothermally in a
tubular reactor under isothermal, laminar flow, conditions. Steady-state reactor
performance is followed at different space times and Reynolds numbers. Analysis of
reactor performance is easily followed by direct sampling and titration. Comparisons
between theory and experiment are in reasonably good agreement. The experiment
demonstrates how concepts in transport phenomena, introduced early in the chemical
engineering curriculum, are related to reactor engineering.

Key words: reaction engineering laboratory; laminar flow reactor; kinetics.

∗ Author to whom correspondence should be addressed.

Tel: +966-3-860-2257; Fax: +966-3-860-4234; E-mail: [email protected]

1
Introduction introduce the student to the isothermal
The traditional approach in the laminar flow reactor (LFR). While
undergraduate chemical reaction plug flow is a good approximation for
engineering (CRE) laboratory, as most cases involving highly turbulent
practiced in our university (and flows in long tubes, it is not for cases
perhaps other universities) typically involving laminar flow. In the LFR,
consists of experiments involving radial gradients in velocity,
homogenous reactions in closed composition and temperature are
(batch) and open (continuous) tank and significant, whereas in the plug flow
tubular reactors. In our CRE reactor (PFR) the velocity profile is
laboratory, batch-reactor experiments assumed to be flat. Thus in the LFR
are used to collect concentration-time fluid properties vary in both axial and
data using ex-situ and in-situ radial directions.
techniques, with the objective of In doing this experiment, it is
determining reaction kinetics. also hoped to counter-balance the
Continuous tank- and tubular-reactor common impression, often perceived
experiments are used to study transient by chemical engineering graduates,
and steady-state reactor performance. that industrial tubular reactors are only
Experiments on the tubular operated as PFRs. This approach,
reactors, in particular, are normally while still based on homogenous
operated under turbulent-flow reactions, is also pedagogically
conditions such that plug flow is a advantageous, as it relates CRE to the
reasonable approximation of reactor first transport phenomena course that
flow conditions. This approach allows undergraduate students are introduced
students to compare experimentally to early in the chemical engineering
collected data with what is typically curriculum. Experiments on tubular
taught in the undergraduate CRE reactor operation, in the laminar flow
course. A typical CRE course mainly regime, were reported by Hovorka and
involves analysis of homogenous, Kendall (1960), and Kendall (1967).
single and multiple, reactions in However, the emphasis in those studies
isothermal and non-isothermal, “ideal” was on effect of reactor geometry, and
reactors. In our course, time limitations comparisons with theoretical results
permit an elementary coverage of were limited.
heterogeneous, solid-catalyzed,
reactions and non-ideal homogeneous Experimental
reactors. Attempts to include The system of interest in this
experiments on non-ideal reactors or experiment is the reaction between
solid-catalytic reactions in the CRE aqueous solutions of sodium hydroxide
laboratory have been limited. and ethyl acetate:
Therefore we have been interested over
recent years in introducing one or two NaOH + CH3COOC2H5 → CH3COO−Na+ + H2O
new experiments that can widen the
scope of the laboratory and offer Kinetics of this well-known reaction
instructors and different groups of are studied in the first experiment in
students some practical options. the CRE laboratory using a batch
The objective of this reactor. Typical results of a kinetics
communication is to summarize our experiment are given in the next
experience with a variation of a classic section. The experimental set-up, used
experiment for the CRE laboratory. in the LFR experiment is manufactured
The experiment is intended to by Armfield [www.armfield.co.uk/cex

2
_datasheet.html]. The set-up is shown steady-state outlet concentration of
in Figure 1. sodium hydroxide was determined.
THERMOMETER

FLOWMETERS

Results and Discussion

CONTROL
CONSOLE
REACTOR
LFRs are open systems
characterized by the continuous
SAMPLING exchange of mass between the reactor
POINT
boundaries and the environment. The
flow in these reactors is characterized
as laminar when the Reynolds number,
Fig. 1 Schematic of tubular reactor Re < 2100. Laminar flow therefore
experimental set-up. occurs in situations characterized by
low velocity, high viscosity, or small
Solutions of the 0.1 M of tube size.
caustic soda and 0.1 M of ethyl acetate In cylindrical tubes laminar
are placed upstream in two storage flow is further characterized by a
tanks containing at least 20 lit of each parabolic velocity profile (Wilkes,
solution. The tubular reactor, whose 1999); i.e. u=u(r). Due to the absence
volume is 0.4 dm3 and inside diameter of mixing in the axial (z) and the radial
is 4 mm, is immersed in a directions (r) in this reactor,
thermostatted water bath. To ensure a concentration (and temperature) will
homogeneous thermal field a vary in both directions, unlike the PFR
continuous stirrer is also mounted where properties are presumed to vary
within the water bath. In the only in the axial direction.
experiments reported below, the In the case of an isothermal,
temperature was maintained constant constant-density, irreversible reaction,
in the range 25-26 oC. Two flow whose rate is (-rA), the performance
meters were used to set the inflow of equation of the LFR is given by the
the two reagents. The reaction was following mole balance equation:
carried out at different volumetric flow
rates. The inlet flow rates of the two ∂C A
reactants were set and the total outlet u (r ) + ( − RA ) = 0 (1)
∂z
flow rate was measured using a
In term of the mean residence time τ,
graduated cylinder and a stop watch.
this equation becomes:
The sampling point is located at the
outlet of the reactor. Samples of 2.5
∂C A
dm3 from the reactor outlet were taken + ( − RA ) = 0 (2)
at different flow rates; quenched with ∂τ
an excess 0.05 M HCl solution; and
back-titrated using 0.05 M sodium These equations apply at a particular
hydroxide solution after adding two to radial position r; therefore one must
three drops of phenolphthalein. The integrate radially to obtain an average
volume of the sodium hydroxide concentration CA at any axial
solution needed to reach the end point position z ∈ (0, L] , where L is reactor
(i.e. turn the color of the solution into outlet. Note that CA(L) is of interest
stable pink) was recorded. Sampling at here as concentration is measured at
short time intervals was required to the outlet in this experiment. Solutions
ensure that steady state was reached of equation (2) in the case of
(usually in 4-5 minutes). Thus the Newtonian fluids were given by
Bosworth (1948), Cleland and

3
Wilhelm (1956), and Denbigh (1951), These results agree with those reported
for zero-, first- and second-order in the literature; the rate constant is
reactions, respectively. These solutions reported to be equal to 0.111 dm3 mol-1
are also reported in most standard CRE s-1 at 25 oC in Bamford and Tipper
textbooks. (1972).
For a LFR of inside diameter D The same reaction was carried
and length L, the average outlet out, at the same temperature, in the
concentration, in the case of second- tubular reactor to investigate its
order kinetics, where equimolar performance in the laminar-flow range.
amounts of the reactants are used as in The effect of Reynolds number on
the present experiment, is given by: conversion of sodium hydroxide is
shown in Figure 3.
⎡ ⎡ kC A 0τ 2 ⎤⎤
C A = C A 0 ⎢1 − kC A 0τ ⎢1 − A n (1 + )⎥ ⎥
⎣ ⎣ 2 kC A0τ ⎦⎦ 0.90

0.80
(3)
0.70
Therefore conversion of reactant A 0.60

(sodium hydroxide in our case) is

Conversion
0.50

given by (Missen et al., 1999): 0.40

⎡ k C A 0τ 2 ⎤ 0.30
xA = k C A 0τ ⎢ 1 − A n (1 + )⎥
⎣ 2 k C A 0τ ⎦ 0.20 Theoretical
Experimental

0.10
(4)
0.00
It is clear that use of Eq. (3) 200 300 400 500 600 700 800 900 1000

Reynolds number
requires kinetics data for the reaction. Fig. 3 Effect of Reynolds number in LFR
As alluded to earlier, such data were
collected in a batch reactor. An inverse relationship between Re
Concentrations of sodium hydroxide and conversion is observed: as Re
versus time were utilized to determine increases the conversion at the outlet of
the kinetics of the reaction under study. the reactor decreases. The maximum
By using the integral method, in which discrepancy between the theoretical
the reaction rate expression is assumed values of conversion, calculated using
and tested using experimental data, it Eq. (4), and the experimental data in
was confirmed that the reaction is Figure 3 is 15 %. In our opinion such a
second-order and the reaction rate discrepancy is acceptable in a
constant was calculated to be 0.102 demonstration experiment, considering
dm3 mol-1 s-1 at 25.9 oC. Typical data errors inherent in collecting samples
collected by the students are shown in and in the analytical technique.
Figure 2. The dependence of the outlet
450 conversion on space time (or mean
400 residence time) is the focus of analysis
350
in the CRE course, especially in the
300
case of homogenous CSTRs and PFRs.
, lit/mol

250
Figure 4 nicely illustrates the effect of
CNaOH

200

mean residence time on steady-state

−1

150

100
conversion in a LFR. As expected,
50 conversion increases as space time
0
0 500 1000 1500 2000 2500 3000 3500 4000
increases. This relationship can be
Time, s physically interpreted as the larger the
Fig. 2 Analysis of kinetic data obtained in space time, the smaller the volumetric
a batch reactor flow rate and the better the conversion.
The two curves in Figure 4 show a

4
maximum deviation around 20 % when such as the assumption of pure
the space time is 282 s. Again, we feel convection (Levenspiel, 1999) which is
such a discrepancy is not uncommon in implicit in the use of Eq. (1) as a model
laboratory demonstrations. of the reactor system. Indeed, there
may well be some degree of molecular
0.90
diffusion in the reactor. As this subject
0.80
is not traditionally covered in our
0.70

0.60
undergraduate reaction engineering
course, we do not expect the students
Conversion

0.50

0.40
to discuss this issue in the analysis of
0.30 the results.
0.20
Theoretical
At the end of our laboratory
0.10
Experimental
experiments on tubular reactors, we
0.00
120 180 240 300 360 420 480 540
hope the difference between laminar-
Space Time, s and plug-flow reactors is clearly
Fig. 4 Effect of space time on conversion demarcated.
in LFR

Concluding Remarks
A tubular reactor was operated
Acknowledgment:
isothermally in the laminar-flow
We are grateful to Mr. Ayman S.
regime to demonstrate basic transport
Ataya for his help in doing some of the
phenomena concepts to students in the
experiments.
context of the undergraduate CRE
laboratory. Students are able to test
theories presented in class lectures:
Nomenclature
reaction kinetics; steady-state reactor
CA concentration [mol .dm3]
performance; and space time versus
CA0 feed concentration [mol .dm3]
real time. Concepts such as mixing-cup
D inside diameter of tube [m]
average concentrations, Reynolds
k reaction rate constant
number, and effect of tube bends can
L reactor length [m]
also be part of the discussion in this
r radial position [m]
experiment.
(-RA) reaction rate [mol.dm-3.s-1]
The sodium hydroxide-ethyl
Re Reynolds number, ρu0D/µ
acetate reaction system was chosen
u velocity [m.s-1]
because of its well-known kinetics that
u0 superficial velocity [m.s-1]
our students confirm in batch-reactor
V reactor volume [m3]
experiments in the CRE laboratory. 3 -1
v0 volumetric flow rate [m .s ]
The relative ease of the analytical
xA conversion of reactant A
technique, which is based on direct
z axial position [m]
sampling and titration, is another
ρ density of mixture [kg.m-3]
reason for this choice of the reaction
system. Although the technique can be µ viscosity of reaction mixture [Pa]
tedious for students, careful planning is τ space time, V/v0 [s]
necessary to ensure the experiments
are completed within the duration of
the laboratory session. Literature Cited
In the comparisons of Bamford, C.H. and Tipper, C.F.H.
theoretical and experimental results we (eds.), “Comprehensive Chemical
see reasonable agreement. There might Kinetics”, Vol. 10, pp. 169, Elsevier,
be other reasons to explain deviations, Amsterdam, 1972.

5
Bosworth, R.C. L., “Distribution of
Reaction Times for Laminar Flow in
Cylindrical Reactors”, Phil. Mag., 39,
847-862 (1948).

Cleland, F.A. and Wilhelm, R.H.,

“Diffusion and Reaction in Viscous-
Flow Tubular Reactors”, AIChE J., 2,
489-497(1956).

Denbigh, K.G., “The Kinetics of

Continuous Reaction Processes:
Application to Polymerization”, J.
Appl. Chem., 1, 227 (1951).

Hovorka, R.B. and Kendall, H.B.,

“Tubular Reactor at Low Flow Rates”,
Chem. Eng. Prog., 56 (8), 58-62
(1960).

Kendall, H.B., “An Apparatus for

Undergraduate Experiments in Flow
System Reaction Kinetics”, Chem.
Eng. Prog. Symp. Series, 63 (70), 3
(1967).

Levenspiel, O., “Chemical Reaction

Engineering”, 3rd ed., pp 339, Wiley,
New York, 1999.

Missen, R.W., Mims, C.A., and

Saville, B.A., “Introduction to
Chemical Reaction Engineering and
Kinetics”, pp 400, Wiley, New York,
1999.

Wilkes, J.O., “Fluid Mechanics for

Chemical Engineers”, pp 117,
Prentice-Hall, Upper Saddle River,
1999.