Conversion of d-xylose to carboxylic acids in a capillary fluidized bed.

摘要

The demand for renewable biomass as a replacement for fossil fuels has never been greater. Many paths to convert carbohydrates into higher value products are under investigation. Few studies have reported data at high temperature and almost no experiments have examined high temperature catalytic partial oxidation of carbohydrates. We need generalized design criteria to exploit this technology commercially.;We present a new method to valorize C5 sugars and review state of the art C5 sugar reactions. We wanted to develop optimal process conditions to produce desirable acids and eventually to test new catalysts for other applications.;We studied the gas phase heterogeneous catalytic oxidation of xylose in a gas-solid catalytic capillary fluidized bed reactor at relatively high temperature and atmospheric pressure. We oxidized the sugar over three catalyst system (vanadyl pyrophosphate, molybdenum trioxide-cobalt oxide and iron molybdate) to form organic acids/anhydrides. We injected a water-sugar solution into a capillary fluidized bed reactor whose main component is a quartz tube (ID = 8 mm) in a furnace that operates at up to 1000 °C. We determine the range of possible operating conditions with a scouting study and then made an experimental design. The most important factors were temperature (200-550 °C), xylose concentration (3 %wt, 7 %wt, 10 % wt in water), residence time (0.1 s, 0.2 s), oxygen partial pressure (0 %vol, 3 %vol, 10 %vol, 21 %vol in nitrogen), and catalyst (VPP, MoO3/CoO, FeMoO). Co-feeding nitrogen improved atomization. Parameters that affected atomization are gas-to-liquid ratio (0.1-0.2 %wt), liquid flow rate (0.01-0.1 ml min-1 ), and nozzle tip and capillary tube diameter. We had four categories of spray performance: 'a', 'b', 'c' and 'd' in decreasing order of performance. We used a mixed design of experiments including two factors: four temperatures (300 °C, 350 °C, 400 °C, 450 °C) and two oxygen partial pressures (3 %vol, 10 %vol) with VPP catalyst. We used type 'a' atomization to verify the effect of other parameters and type 'b' to verify the effect of atomization performance on yield. We also tested sequentially feeding of xylose-oxygen followed by air at a frequency of 5 min-1 . The experimental plan included 2 h and 4 h runs to test catalyst stability.;Residence time inside the catalytic bed was 0.2 s and experiments lasted 4 h. The fluidization gas contained 3 %vol and 10 %vol oxygen in nitrogen and entered the reactor through a fritted glass distributor. The inlet fluidizing gas stream varied between 80-150 ml/min to meet 3 % vol and 10 %vol oxygen. We carried out most experiments with 1 g of calcined VPP catalyst. We metered the 3 %wt xylose solution at 0.04 ml min-1 with a syringe pump. We atomized the liquid into small drops through a 0.25 mm capillary tube constricted at the end nozzle. We fed nitrogen and the liquid solution with the gas-to-liquid ratio of 0.18 %wt to produce an effervescent spray. The droplets vapourized rapidly and the xylose reacted to form maleic anhydride, acrylic acid and acrolein.;We absorbed the liquid products (acids) from the reactor effluent in distilled water in a series of quenches. We sampled and analyzed the accumulated acids offline with high performance liquid chromatography (HPLC). We validated the HPLC analysis with gas chromatography (GC) and tried to identify other possible products.;Operating conditions have a considerable effect on product distribution and production rates. Acrylic acid was the most desirable and maleic acid the most abundant. Vanadyl pyrophosphate is both active and selective for this process and in the best case, at 300 °C and 10 %vol oxygen, maleic acid, acrylic acid and acrolein yields were 25 %, 17 % and 11 %, respectively. We also detected gaseous carbon dioxide with GCMS during the reaction. Thermogravimetric analysis for the VPP samples we withdrew at the end of the reaction confirmed that no coke formed on the catalyst. Powder agglomeration and caramelization were only problematic when the reactor operated outside the range established during the scouting experiments.