CC BY
7
Научни трудове на Съюза на учените в България-Пловдив. Серия В. Техника и технологии, естествен ии хуманитарни науки, том XVI., Съюз на учените сесия "Международна конференция на младите учени" 13-15 юни 2013. Scientific research of the Union of Scientists in Bulgaria-Plovdiv, series C. Natural Sciences and Humanities, Vol. XVI, ISSN 1311-9192, Union of Scientists, International Conference of Young Scientists, 13 - 15 June 2013, Plovdiv.
THEORETICAL STUDIES OF SIMULTANEOUS ADSORPTION WITH DIFFUSION OF ACETIC ACID ON ICE BASED ON EXPERIMENTAL DATA FROM FLOW REACTORS
Atanas Terziyski1, Nikolay Kochev2, Vesselina Paskaleva3 i, 2, 3 University of Plovdiv, Department of Analytical Chemistry and
Computer Chemistry e-mail1: atanas@uni-plovdiv.net e-mail2: nick@uni-plovdiv.net e-mail3:vessy@uni-plovdiv.net
Abstract
In this work we present theoretical study of the physicochemical properties of gaseous acetic acid as an important atmospheric compound. The kinetic processes between the trace gases and the ice surface and the consequent ice bulk diffusion are studied by means of flow reactor experiments and corresponding theoretical simulations. The performed laboratory experiments are presented by a quadruple mass signal as function of the time. The kinetic and thermodynamic parameters for the studied processes are analyzed by fitting the raw experiment data with a kinetic model developed in our group for the simulation of flow tube reactor experiments. The studied lab experiments cover temperatures around 200K and gas phase concentration of acetic acid around 1012 cm-3. The implemented simulations give valuable theoretical estimations of the studied parameters as well as they reveal detailed information about the mechanism of the processes occurring along entire reactor.
Introduction
Acetic acid is one of the most abundant carboxylic acids in the atmosphere. It plays an important role in tropospheric radical interactions. Acetic acid with some other molecules produces free radicals that take part in the catalytic decay of ozone. We studied acetic acid interactions with ice surface at temperatures around 200K experimentally and theoretically. Coated wall flow tube (CWFT) reactors are commonly used to study both gas phase and heterogeneous reactions. In our research we approach raw experimental data measured at the University of Duisburg-Essen, Institute of physical and theoretical chemistry, Prof. Zellner's workgroup. The experiments were performed while exposing gaseous acetic acid injected from a movable injector on ice generated surfaces. The temperature range is around 200K and the gas phase concentration between 1011 and 1012 cm-3. The raw data outputted from the measurements represents the quadruple mass spectrometer signal which is plotted as a function of the laboratory time while injector is moving. In this work we present the results obtained from the application of the full theoretical model [1] published recently which describes thoroughly all processes along entire reactor.
Mass spectrometer signal characteristics
The typical measurement in a CWFT reactor (Figure 1) with sliding injector can be described by several stages as follow:
Stage 0. The nozzle of the movable injector is located in front of the quadruple mass spectrometer (QMS). The registered signal is normalized to its value, i.e. 1. On Figure 2 stage 0 is represented by the first 10 seconds;
Figure 1. Coated wall flow reactor tube. The Figure 2. Illustration of the five stages of typical picture is taken while the reactor working. measurements in CWFT reactor.
Stage 1. The movable injector is rapidly slid with a constant speed from the QMS to the end of the tube. During this stage we expose the constantly injected gas flow on the fresh ice surface and they immediately adsorb and penetrate into the bulk. This causes the rapid drop of the signal. The end of this stage is when the nozzle stops by reaching the programmed distance;
Stage 2. In this stage the nozzle of the injector is not moving and its position is the same as the one reached at the end of previous stage. The signal is constantly increasing until it reaches equilibrium that corresponds to a constant signal;
Stage 3. The nozzle is moved analogously as in stage 1 but towards QMS. The established equilibrium in the previous stage is destroyed due to the lowering of the gas phase concentration and thus follows large desorbtion/segregation from the whole tube. The resulted signal is a sum of injector flow and the desorbed molecules. Stage 3 is shown on Figure 2 between 65 and 80 seconds. It corresponds to the highest measured signal;
Stage 4. The nozzle position of stage 4 and stage 0 is the same. While in stage 0 we record only constant signal, normalized to 1, in stage 4 the previously adsorbed and consequently desorbing molecules are added to the signal. The decay of the signal slope in stage 4 is proportional to the desorption rate coefficient.
Best fit parameters:
Ea = 47000 J/mol (activation energy)
a0 = 10-14 cm2 (area of an active surface site)
A = 5x10" s-1 (frequency factor)
Y = 1.6x10-4 (uptake coefficient)
c = 1.2x1015 cm-2 (maximal surface coverage)
s,max v ° '
ksol = 0.2 s-1 (solution coefficient)
kseg = 10-20 cm3s-1 (segregation coefficient)
D = 2x10-11 cm2s-1 (diffusion coefficient)
Lab time [s]
Reactor experiment conditions and geometry:
T = 190K, c =3.64x1011cm-3, P , =9.57x10-4Pa
' gas ' partial
radius=1.2cm, length = 15cm Figure 3. Acetic acid on ice, measured in CWFT reactor (dots) and the solid line - kinetic
23mulation of adsorbtion with ice bulk diffusion.
Modeling procedures
The experimental signal G depending on the gas phase concentration is often scattered. The noise level is estimated to be asound 0.2 a.u. Signal G is fitted with a simulated signal G(a) regarded as a function of the studied parameters p = (y, Ea, o0, A, csmax, ksol, k ). Tln^ fit is performed by means of the numerical procedures [1] implemented within AdDeSSa software system [2]. The rheoretical eatimation for the parameters is found eor the minimum root mean squire error between Gea and G(a). Typical fitting result is shown in figure 3._
C /Cs,max 100%
without diffusion
■■ kr
mm
time [s]
Cs/Cs,max 100%%
with diffusion
time [s]
Figure 4. Plot of the relative surface concentration (c /c ) as a function of the laboratory
° v s s,max) J
time anc[ the re actor position (segment). Top graphics represents the reacSo r stftes without diffusion
Results and Discussion
ADESSA software simulates the resulting QMS signal (see Figure 3) as well as it can trace the states of each reactor segment as function of the time. The latter results are visualized as 3D surface plots. Figure 4 shows in detail the relative surface concentration for all reactor segments. Typically the surface concentration for each segment sharply rises up reaching a plateau (this part of the profile corresponds to the adsorption processes reaching the equilibrium, see stages 0,1 and 2) and then with a unique profile for each segment, the concentration reaches down to zero (the desorption processes in stages 3 and 4). The comparison of the top and bottom plots from figure 4 shows a significant difference due to the bulk diffusion. The profiles of the back segments (the segments with higher numbers) is drastically changed thus they do not have the plateau. Also in the bottom plot (where diffusion is included) the system does not reach equilibrium within the studied time interval.
Table 1. Maximal relative surface concentration for the entire reactor as function of y
Cs/Cs,max-diffe re nce
Y max(c /c ) v s s,max' without diffusion max(c /c ) v s s,max' with diffusion
1.03 x 10-4 4.2% 3.8%
5.24 x 10-4 18.4% 16.3%
8.08 x 10-4 25.5% 23.0%
14.50 x 10-4 37.7% 34.6%
20.0 x 10-4 46.6% 43.0%
Figure 5. The difference between maximal relative surface concentrations with and without diffusion expressed as function of y.
¡U 1.5 a>
■a
0.5 -0
0.9973
0.00E+00 5.00E-04 1.00E-03 1.50E-03 Gamma
2.00E-03 2.50E-03
2.5
2 --
Table 1 shows the maximal surface concentrations within entire reactor for different values of the uptake coefficient y. As it can be seen the maximum is lower for the case with included diffusion. Currently in similar researches, the diffusion is not taken into account directly but indirectly by correcting y coefficient. Figure 5 shows strong logarithmic dependence of the difference between maximum surface concentration calculated respectively with and without diffusion.
Summary
The combination of CWFT reactor experiment with an application of ADDESSA model can reveal valuable kinetic and thermodynamic constants as well as important details for the studied processes. Currently we are analyzing large set of experimental data taken at different conditions. The kinetic and thermodynamic values will be reported in the forthcoming publications.
Acknowledgments
This work is supported by the Bulgarian National Fund for Scientific Research NFNI (project MU02/12).
References:
1. N. Kochev, A. Terziyskil and M. Milev, Numerical Modeling of Three-Phase Mass Transition with an Application in Atmospheric Chemistry, Applied Mathematics, Vol.4 No.8A, 2013
2. A. Terziyski and N. Kochev, "Distributed Software Sys-tem for Data Evaluation and Numerical Simulations of Atmospheric Processes," LNCS (Title: Numerical Methods and Applications), Vol. 6046, 2011, pp. 182-189. doi: 10.1007/978-3-642-18466-6_21
