Table of Contents
Background and Theory ?????????????????????.
Experimental Procedure ?????????????????????.
Results and Discussion ??????????????????????
Conclusions and Recommendations ?????????????????
A. Data ??????????????????????????.
B. Equipment Diagram ????????????????????..
The hydrogenolysis of ethane on a nickel on kieselguhr catalyst has only been conducted using a cylindrical pellet shaped catalyst of unknown surface area. We used a powdered version of this catalyst and compared its performance with that of the pellet catalyst. This comparison was conducted using one gram of catalyst in each reactor. Holding all other variables constant, ethane conversion was tracked while varying total flowrate. We then varied the temperature while holding the flowrate, input composition, and all other variables constant.
The reaction that converts ethane to methane occurs on the surface of the catalyst. A determination of the surface area is an important factor in comparison of the performance of the pellet and powdered catalyst. We conducted the surface area determination using a nitrogen adsorption apparatus and analyzed the data in accordance with BET theory using the single point method.
Background and Theory
This experiment involves the kinetics of ethane cracking. Ethane cracking is where a molecule of ethane is hydrogenated and produces two molecules of methane, as seen in Equation 1. The catalyst is nickel on kieselguhr, a porous medium.
This reaction was conducted in a single tubular packed bed reactor (PBR). The catalyst was supported by glass wool and placed in the center section of the PBR. The composition of the product stream was measured using a gas chromatograph (GC). The GC detector signal was fed to an integrator. The integrated area under the signal peaks can be related to the conversion of ethane (XE) by the following equation:
where M and Eare the integrated areas under the corresponding peaks for methane, and ethane and CF is a correction factor based on a standard of known composition.
In addition, we investigated methods for the determination of both catalysts? surface area using a Quantasorb nitrogen adsorption apparatus. The adsorption apparatus consisted of a sample cell through which nitrogen and helium flowed. The effluent passed through a thermal conductivity detector that detected nitrogen. The detector signal was sent to an integrator where the area of the peak was displayed. The first step in the process involved affecting adsorption by immersion of the sample cell into liquid nitrogen. Once nitrogen adsorption was complete the sample cell was withdrawn from the liquid nitrogen bath. As the sample warmed to room temperature desorption of the nitrogen began and the integrated signal was recorded. A known volume of nitrogen calibration gas was then injected into the detector. The data was then analyzed in accordance with Brunauer, Teller and Emmet theory using the single point method. The following equation was used to determine the total surface area of the sample based upon volume of desorbed nitrogen:
where A/Ac is the ratio of the integrated areas of the desorbate to the calibration gas, Vc is the volume of calibration gas, Pa is the ambient pressure, N is Avogadro?s number, R is the gas constant, T is temperature, P/Po is the ratio of the partial pressure of the adsorbate to the saturated vapor pressure of the adsorbate and Acs is the cross-sectional area of the adsorbate molecule.
In this experiment we only used Reactor 1 from the diagram in Appendix B. We first collected data with 1 g of the pellet catalyst. The first set of trials involved varying the total flow rate while keeping the composition and temperature constant. We used a composition of ethane (53.3 vol. %), helium (26.7 vol. %), and hydrogen (20 vol. %), and maintained the reactor temperature at 240?C. The flow rates that we used were between 100 cc/min to 400 cc/min. Once this was accomplished we chose the flow rate of 200 cc/min and then varied the temperature for the second set of trials. The range of temperatures across which the experiment was conducted was 190?C ? 265?C. We then conducted the same series of varied flow rates and temperature trials upon 1 g of the powdered catalyst. In both cases we took several standard readings of the GC at the beginning of each lab period.
Matt, using glass wool as a support for the catalyst, packed the powdered reactor. The catalyst was distributed over a certain length as evenly as possible. A test was conducted to see if the high-pressure gasses would blow the catalyst out of the reactor, and the results indicated that the catalyst would stay in the glass wool inside the reactor.
A possible safety concern could have arisen if the powdered catalyst dramatically increased conversion. This reaction is exothermic, and by going to completion, it could have produced more heat that this experiment was designed for. We didn?t know the heat limitations, but we assumed that the powdered catalyst would only slightly increase the conversion, so it would not be a concern.
When we conducted surface area analysis the sample was out-gassed using the method of repetitive cycling . This involved conducting adsorption and desorption cycles until the adsorption/desorption readings stabilized. Ideally when this occurs its indicates that
It must be water, things most likely do not stick to it too well
It is possible that we should have conducted more cleaning cycles
Added graph 1 and caption
Added graph 2 and caption
Results and Discussion
Our results were very inconclusive for the conversion of ethane vs. the total flowrate using the pellet shaped catalyst. The purpose of this experiment was to see if running the input stream slower would result in higher conversion. This would mean the gasses would have more time to diffuse through the porous catalyst. However, if the gasses were already fully diffusing, then conversion wouldn?t change as a function of flowrate. Figure ******** below shows three different days of trials varying the flowrate.
Figure *****. The conversion of ethane related to total flowrate at constant temperature was very different on different days. We attributed this to residue left on the catalyst slowly bleeding off. The conclusion we reached was that the conversion was not dependent on the flowrate.
The first day seemed to show that the longer that the reaction had to take place on the catalyst, the higher conversion we would get. However, in the second day of trials, we saw that total flowrate had almost no effect on the conversion of ethane. The third day showed a slight increase in conversion with an increase in flowrate. We can attribute the third line as noise in the system, with the exception of the last data point, which is very hard to explain (human error?). Assuming that the second day?s results were correct, then we need to explain what happened on the first day of trials. It is possible that some standard was left running through the reactor, and therefore affected the amount of methane that was perceived by the GC. Most likely there was residue from the last group using the reactor, and there is a certain time that needs to elapse before all of this could bleed through. This is why, on the 10-5-00 line, after 4 trials (approximately 40 minutes), the conversion goes to zero and stays there. There is a possibility that there could be residue in the sample loop for the injection. The recommendations section addresses a future experiment to deal with this.
Results for increasing the temperature are what we expected. The higher we raised the temperature, the higher the conversion of ethane was. The results from two different trials of the pellet are shown below in Figure *****.
Figure *****. This graph shows that by increasing the temperature we increase conversion. The variation in rise time is possibly due to temperature equilibrium not being reached before the measurement was taken.
There is a large difference in conversion for the temperature of 250. First off, there could be differences in the measurement of the thermocouple, and the actual temperature of the catalyst. We left about 10 minutes in between each run for the temperature to rise and come to equilibrium. This might not have been enough time for steady state to be reached. Also, this reaction is exothermic, so it could be heating itself. The higher the temperature, the more heat would be produced in the reaction, which would bump up the temperature also. This would account for the variations in the graph. One of the runs was done at a total flowrate of 300 cc/min, and the other at 200. Given our conclusion that total flowrate doesn?t affect conversion, this shouldn?t have altered the results.
In general, conversion of ethane using a powdered catalyst was less than that of the pellet catalyst. The results from total flowrate variation are shown below in Figure ********. The conversion stayed low, and fluctuated around the different flow rates. This is probably due to noise in the system.
Insert graph from powdered catalyst ? total flowrate
The weight of powder used was twice that of the pellet, and since the conversion for the powdered catalyst was low, it reinforces the fact that the pellet seemed to work better. One day of trials with the powdered catalyst produced no conversion at all. This could have been due to the fact that the catalyst had ?blown out? of the reactor, or that it was quickly poisoned during another group?s trials.
Once again, we showed that increasing the temperature increased the conversion of ethane. We didn?t achieve the as high of a conversion as with the pellet catalyst.
Insert graph from powdered catalyst ? temp variation
Final comments on the comparison of the two. The pellet catalyst allows for higher conversion than the powdered catalyst. Both the total flowrate variation and the temperature variation showed this to be true.
Conclusions and Recommendations
? After doing the standard run, do a run with ethane on the bypass loop. Measure conversion to make sure there?s no residual methane in the sample loop.
? After doing the standard run, do a run with helium running through the reactor. Measure conversion to see if there?s residue left from previous trials.
We expected that the powdered catalyst would give greater conversion than the pellet catalyst under the same circumstances. On a superficial level it appears that the apparent effective surface area is much larger in the case of the powdered catalyst. Yet, due to the difference in packed bed void fraction, and the impact that this has upon convective boundary layer exchange of the reactants and products at the surface of the catalyst particle, this relationship may not be so simple. In addition, the method by which the catalyst is packed into the reactor may have a great impact upon diffusion. This is assuming that the contribution to overall ethane conversion of the internal reactive surface area of the pellet catalyst is negligible. If this is not so then perhaps operating the reactor at slower flow rates will exhibit a greater increase in pellet catalyst conversion than the powdered. In the end, these experiments are a small part of the overall problem-space and will provide insight for future experimental direction.