丙烯在氧氣及氫氣共存下合成環氧丙烷之活性及穩定性研究

Abstract

本研究以固體觸媒Au/TS-1催化丙烯在H2及O2共存下反應生成環氧丙烷(PO)，該研究銜接國際尖端研究團隊美國Delgass的成果繼續研發。Au活性基會先催化H2及O2反應生成H2O2，然後TS-1上的Ti活性基催化H2O2氧化丙烯形成PO。TS-1沸石擔體是由水熱法合成，並以沉澱沉積法擔載Au。本研究首先進行GC分析系統改善，以提升實驗數據值的準確性。之後期望藉由增加Au活性基，增加H2O2產量，並由添加純TS-1共觸媒增加PO產率。另外也期望藉由增加H2濃度或由添加H2O，以增加H2O2，進而提升PO產率。本研究發現Au/TS-1催化反應的初始活性很高，但會隨時間衰退，所以研究下列幾種方式嘗試提升觸媒穩定性：1. 減少進料的反應氣體(丙烯、H2或O2)濃度或降低反應溫度；2. 以有機矽化合物修飾觸媒表面，去除酸性或親水基對反應之影響；3. 使用未煅燒之TS-1(孔洞尚塞滿合成用之模板)擔載Au化合物以製備觸媒，嘗試使丙烯催化反應僅在擔體顆粒外表面進行，而不在孔道內進行。 第一部分研究GC分析系統改善方面，經改用CP-Wax 52 CB (capillary column)分離PO已能獲得更正確及精確分析數據，Au/TS-1觸媒催化丙烯環氧化實驗已能再現文獻數據。 第二部分研究各Au/TS-1觸媒，基於200℃反應溫度，GHSV =14000mLgcat-1h-1，各進料濃度70% He、10% Propylene、10% H2及10%O2，反應初始活性可獲得PO最高產率11.3%(丙烯轉化率13.3%，PO選擇率84.6%)，而反應二小時後，雖然活性明顯衰退至PO產率9.0%(丙烯轉化率11.0%，PO選擇率82.1%)，但之後活性衰退會較慢，反應活性較穩定，因此之後反應活性皆以第二小時數據做比較。已發現金擔載量越高，越容易產生副反應，不利於PO產率。當金擔載量低於0.86 wt%(金顆粒小於5 nm)，金粒子的催化活性基類似，所以在相同的丙烯轉化率會有相近PO選擇率。但是當丙烯轉化率增高時PO選擇率會降低，因此使產率在反應第二小時後無法高於9.3%。 在第三部分研究發現，Au/TS-1觸媒中添加TS-1共觸媒無法提升整體觸媒活性，事實上TS-1共觸媒的存在反而會產生較多副反應而降低PO產率。因此，藉由酸液中添加碳酸銫處理TS-1，結果發現共觸媒活性及PO產率能恢復Au/TS-1觸媒水準但仍無法增加，顯示TS-1催化活性不是反應瓶頸。 在第四部分，增加H2進料濃度能些微增加PO產率，但過度增加時，PO產率反會明顯下降，而且H2會和O2較易反應生成H2O，顯示H2O2及H2O的生成是平行競爭反應。因此藉由進料中添加H2O，使H2和O2反應時不易往H2O的方向進行，結果發現環氧丙烷產率確實有提升。另外發現，雖然增加H2濃度及添加H2O都能提升PO產率，但也會使觸媒活性衰退更嚴重。 在第五部分，減少進料的反應氣體(丙烯、H2或O2)濃度或降低反應溫度，發現反應活性雖下降，但無法減少催化活性隨時間衰退速度。而將觸媒表面種植有機矽化合物經過修飾後，發現能明顯改善衰退問題，但會降低催化活性(GHSV = 14000mLgcat-1h-1, 丙烯轉化率 = 3.9%)。雖然藉由增加觸媒使用量(GHSV = 5250mLgcat-1h-1)可稍增加轉化率至5.0%，但會降低PO選擇率。而以未煅燒TS-1擔載體Au，結果發現雖能明顯減緩活性衰退速率，但活性仍會衰退，且反應活性不高(GHSV = 14000mLgcat-1h-1, 丙烯轉化率 = 3.0%)，增加觸媒用量(降低GHSV = 5250mLgcat-1h-1, 丙烯轉化率 = 4.6%)仍無法明顯改善觸媒床活性及相關反應結果。

Following the results of last year, more study of Au/TS-1 has been carried out for catalyzing partial oxidation of propylene (in presence of hydrogen and oxygen) to produce propylene oxide (PO) in gas phase in a fixed bed reactor at 200℃. At least two kinds of catalytic active sites on Au/TS-1. Those of gold catalyze the reaction between hydrogen and oxygen to form hydrogen peroxide. Those of TS-1 catalyze the reaction between propylene and hydrogen peroxide to form PO. TS-1 is synthesized by hydrothermal method, followed by deposition-precipitation method for loading gold nano-particles to form Au/TS-1. In this research, a high initial activity has been found, which can produce PO with a yield of 11.3% (propylene conversion of 13.3%，PO selectivity of 84.6%) at GHSV = 14000 mLgcat-1h-1 with reactor inlet compositions of 70% He, 10% Propylene, 10% H2 and 10% O2 at 1 atm total pressure. There is reaction decaying problem of Au/TS-1 system. PO yield decays to 9.0% (propylene conversion 11.0%、PO selectivity 82.1%) after 2 hours reaction, after which the decay rate slows down. Several measures were adopted to try to increase more activity of Au/TS-1, including increasing the loading of gold on Au/TS-1 for H2O2 production, and mixing Au/TS-1 with TS-1 to form a co-catalyst for propylene expoxidation. It has been found that loading more gold and no more than 0.86 wt% is effective to enhance the activity for PO production. Nevertheless, TS-1 in the co-catalyst provide no positive effect. It suggests that epoxidation reaction on active site of TS-1 support of Au/TS-1 is not the reaction rate limiting step for producing PO. On the other hand, only slight increase of hydrogen concentration can slightly increase propylene conversion, PO selectivity, and PO yield. The presence of steam in the feed of the reactor can also increase PO yield. Because both are for more production of H2O2, it indicate that H2O2 generation is the rate limiting reaction for producing PO in the whole reaction system. For solving the problem of Au/TS-1 activity decay during the reaction, several attempts have been carried out. It has been found that the decrease of reactants concentrations and the reaction temperature can lower the reaction activity for producing PO; nevertheless, the decaying rate of the reaction activity can not be improved. A much more steady activity of Au/TS-1 can be achieved either by coating the catalyst with silane, or by using uncalcined TS-1 for loading gold to form Au/TS-1. However, both methods would cause the catalysts with lower activities and resulting in lower propylene conversions and PO yields.