Catalysis of Organic..

370 MIBK Synthesis via CDreboiler product reaching as high as 78 wt%. This is particularly noteworthy sincethe syntheses of DAA and mesityl oxide are strongly equilibrium limited reactions.The selectivity to mesityl oxide remained within a narrow range from 85 to 90%.Evidently, conditions that resulted in a higher conversion of DAA also resulted in anincreased rate of production of undesirable higher molecular weight productsincluding mesitlyene, phorone and isophorone. Similarly, the conditions for whichthese undesirable consecutive reactions were averted resulted in a greater amount ofunreacted DAA remaining in the system. A major finding was that the undesirableconsecutive reactions from mesityl oxide could be mitigated while simultaneouslyincreasing mesityl oxide productivity by increasing the liquid reflux flow rate in theCD column. Note the effect of increasing the reboiler duty from 300 to 350 W at114°C. This shows that CD technology allows improved selectivity to a desiredintermediate by the rapid removal of the desired product from the reaction zone.In the second CD experiment, MIBK was successfully produced from acetone ina single stage. However, the MIBK yield was relatively low in this experiment. Forexample, when hydrogen was introduced into the CD reactor at 60 L/hr (STP) with areboiler duty of 350 W and a reaction temperature of 119°C, the MIBK productivitywas 0.10 [g MIBK /(hr*g cat )]. The MIBK productivity was calculated from an averageMIBK concentration in the reboiler product of 3.98 ± 0.031 wt% based on 11measurements over a 10 hour period of steady state operation. The mesityl oxideconversion was 15.1% and the hydrogen utilization was less than 2%. Thehydrogenation of acetone to produce 2-propanol was the only significant competinghydrogenation reaction and the selectivity of the hydrogenation was 84.4% for theconditions described above. It is evident that the hydrogenation of mesityl oxide toproduce MIBK is currently the limiting step of this CD process. It should also benoted that the locations and catalyst amounts for the reaction zones were notoptimized for this preliminary experiment.Although the MIBK productivity was comparable to the data reported byLawson and Nkosi (6), there was evidence of significant hydrogenation catalystdeactivation. At the end of the experiment, the hydrogenation catalyst was recoveredfrom the reactor under a nitrogen blanket and protected in solvent. Its activity formesityl oxide hydrogenation was subsequently tested in an autoclave and was foundto have an activity of 0.302 relative to the fresh catalyst. Spectroscopic datapresented in the next section suggests that the strong adsorption of MIBK may havehad a detrimental effect on the long-term performance of the catalyst in the CDreactor. The operating pressure for this experiment was constrained to less than 0.6MPa due to the poor thermal stability of Amberlyst 15, which has a maximumoperating temperature of 120°C. Consequently, a low MIBK yield was expected dueto the relatively low hydrogen partial pressure in this experiment.Catalyst Characterization via DRIFT SpectroscopyThe commercial Pd/Al 2 O 3 catalyst used in this pilot scale study was characterized viaDRIFT spectroscopy. In situ DRIFT spectra of carbon monoxide, MIBK, acetone