Ανάπτυξη καταλυτών υδρογονοεπεξεργασίας στηριγμένων σε τιτάνια

Abstract

The main subject of the present dissertation is the development of hydro processing catalysts supported on industrial anatase with relatively high specific surface area (120m2.g-1), the application of these catalysts in the hydrodesulfurization and hydro oxygenation reactions and the investigation of the active h.d.s. sites developed on the surface. The dissertation involves five parts. The main findings of each part constitute the inputs for the next one. An extended critical review concerning the hydrodesulfurization catalysts based on titania as well as the recent evolutions relevant to the structure of the molybdenum species developed on the titania surface at each step of the preparation procedure (deposition, drying, calcination and activation) is the subject of the first part. Two main conclusions may be drawn from this literature survey. The first conclusion is that the intrinsic h.d.s. activity is higher in the titania than alumina based catalysts. This mainly concerns the un-promoted molybdena catalysts and to a lesser extent, if any, the cobalt or nickel promoted ones. However, as the most titanias used in the past exhibited relatively low specific surface area (50 m2.g-1), the specific activity of the titania based catalysts is lower than the γ-alumina based ones. This prompted efforts for preparing and use titanias of high specific surface area or using mixed supports of titania-alumina, titania –silica or titania-zirconia. The second conclusion is that titania stabilizes related small clusters of MoS2 with an edge length smaller than 4.5 nm in a perpendicular or tilted configuration bonded with the surface by Ti-O-Mo or Ti-S-Mo bridges whereas on γ-alumina the perpendicular orientation is no longer thermodynamically stable for clusters with 4 Mo atoms per edge. Moreover, the creation of edge S-vacancies on large clusters is easier on anatase than on γ-alumina surface. In the second part of the dissertation we have synthesized (using the equilibrium deposition filtration technique, EDF) and characterized (using nitrogen adsorption, X-ray powder diffraction, diffuse reflectance spectroscopy, Temperature programmed reduction, X-ray photoelectron spectroscopy and Raman spectroscopy) nine non-promoted molybdena catalysts with molybdenum surface concentration in the range 1.7-5.5 nm2 g-1 (monolayer catalysts).The h.d.s. of thiophene was used as model reaction. The catalytic tests were performed using a differential fixed bed reactor working at atmospheric pressure. The following were the main purposes of this work. (i) The optimization of the thermal treatment of the anatase before usage. It was found that heating at 4000C for 12h with a heating rate corresponding to 20min for increasing temperature for room temperature to 4000C did not affected considerably the pore volume distribution and thus the specific surface area of the anatase used. (ii) The optimization of the deposition parameters of the molybdenum species on the anatase surface [equilibration time (1-48h) and pH (7.0-2.0) of the impregnation suspension]. The increase in the equilibration time and the decrease in pH result to an increase in the surface concentration of the supported molybdenum while they decrease the intrinsic activity. The conditions for obtaining the optimum surface Mo concentration relevant to the maximum specific activity are 3 hours of equilibration time and pH = 4.5. (iii) The optimization of the surface concentration of the supported molybdenum. It was found a volcano curve in the plot “specific activity vs. surface Mo concentration” with a maximum predicted at 3.8atom Mo /nm2. The most active sample prepared in this work corresponded to 3.2 atom Mo /nm2. (iv) Investigation of the structure of the active site and the promoting action of the anatase surface. It was concluded that isolated monomer molybdenum species with a monopodal configuration prevail before calcination in the samples with relatively low Mo surface concentration. These are transformed into monomer molybdenum species with a tripodal configuration upon calcination. In these species the transfer of electronic density from the titanium to molybdenum atom is facilitated through the Ti-O-Mo bridges the number of which per molybdenum atom is maximized (three).The aforementioned transformation decreases the capacity of the titania surface relevant to the formation of a monolayer and thus causes the transformation of the monomer into the oligomeric molybdenum species in which the number of the Ti-O-Mo bridges is lower than three This structure does not facilitate so much the transfer of electronic density from the titanium to molybdenum atom. The monomer molybdenum species with a tripodal configuration are transformed, upon activation, into relatively small size clusters of molybdenum sulfide oligomeric molybdenum species bonded with the surface by Ti-O- Mo or Ti- S- Mo bridges through edges. The number of these bridges per Mo atom takes its maximum value in these clusters. These isolated structures being in perpendicular or titled configuration facilitate the transfer of electronic density from the titanium to molybdenum through the aforementioned bridges which in turn decreases the intensity of the Mo-S bonds and favor the removal of S2- ions and thus the stabilization of a relatively large number of coordinatively unsaturated sites (CUS) considered as active sites. Therefore, these isolated structures are very active. On the other hand the oligomeric molybdenum species in the oxidic precursor state are transformed into considerably larger clusters, eventually in perpendicular or titled configuration, with more molybdenum planes and smaller number Ti-O- Mo or Ti- S- Mo per Mo and thus CUS. These clusters are relatively less active than the previous ones. As the surface concentration of Mo increases, the contribution of the larger and less active clusters increases justifying the observed linear decrease in the intrinsic activity with the Mo surface density. The observed volcano curve for the specific activity [“specific activity vs. surface Mo concentration”] may be attributed to two opposite effects: to the increase of active sites and the aforementioned decrease in the intrinsic activity. It is important to note that the maximum in the volcano curve is obtained at 3.8 Mo atoms per nm which correspond to the monolayer of monomer tripodal molybdenum species. In the third part of the dissertation we continued the optimization of preparative parameters related to the synthesis of non-promoted molybdenum catalysts taking into account the main findings of the second part. Specifically, we have synthesized nine additional non -promoted molybdena catalysts, most of which with the optimum molybdenum surface concentration (3.2 atoms Mo/nm2). The catalysts were characterized using nitrogen adsorption, X-ray powder diffraction, diffuse reflectance spectroscopy, temperature programmed reduction and X-ray photoelectron spectroscopy. The h.d.s. of thiophene was used as a model reaction. The catalytic tests were performed using a differential fixed bed reactor working at atmospheric pressure. The optimization procedure concerned mainly the method of deposition of the molybdenum species (equilibrium deposition filtration, typical wet impregnation, dry impregnation and modified wet impregnation, namely wet impregnation under controlled immersion time and pH). For the samples prepared under similar impregnation conditions, the application of equilibrium deposition filtration technique provided the most active catalyst. It follows the catalyst prepared by typical wet impregnation and then the catalysts prepared by dry (or modified wet impregnation). Large impregnation time, activation after drying, without calcination, and no calcination of the support before the deposition of the molybdenum species provided catalysts with relatively low activity. In order to examine the influence of the support we have synthesized a catalyst with aforementioned surface Mo concentration supported on γ-alumina following dry impregnation. It was found that the specific activity of this catalyst is smaller than those obtained on all catalysts supported on anatase. The influence of the impregnation technique and the aforementioned preparation parameters was interpreted taking into account the size of the supported molybdenum clusters obtained after activation which is related to the number of the Ti-O-Mo (Ti-S-Mo) bridges per Mo atom. Suitable choice of the impregnation technique and the aforementioned preparation parameters leads to the increasing contribution of supported molybdenum clusters with small size and thus with high number of the Ti-O-Mo (Ti-S-Mo) bridges per Mo atom. The relatively low activity exhibited by the catalyst supported on γ-alumina was attributed to the different structure of the supported molybdenum species obtained after activation and will be explained in the next part of work. The fourth part of the dissertation deals with promoted catalysts. On the base of the three most active non-promoted molybdena catalysts supported on anatase, which were developed in the third part of the dissertation, we synthesized cobalt and nickel promoted molybdena catalysts supported on this support. We studied the influence of the kind of the promoter as well as the method of deposition of the molybdenum and the promoter species (equilibrium deposition filtration for the molybdenum species/ dry impregnation for the promoter, wet impregnation for the molybdenum species/ typical wet impregnation for the promoter).The aforementioned optimization study was extended to nickel promoted tungsten catalysts supported on anatase. In order to study the influence of the support on the specific activity we synthesized and studied a number of the aforementioned promoted catalysts supported on γ-alumina. The catalysts were characterized using nitrogen adsorption, X-ray powder diffraction, diffuse reflectance spectroscopy, temperature programmed reduction and X-ray photoelectron spectroscopy. The h.d.s. of thiophene was used as model reaction. The catalytic tests were performed using a differential fixed bed reactor working at atmospheric pressure. Fifteen catalysts were tested in this part of the dissertation. The promoted catalysts supported on anatase are (at least) one order of magnitude more active than the un-promoted catalysts. This indicates that the cobalt or nickel promote furthermore the activity with respect to promotion exhibited by the anatase surface through the Ti-O-Mo or Ti-S-Mo bridges. The deposition procedure followed does not influence so much the catalytic behaviour. The specific activity of the nickel promoted tungstate catalysts supported on anatase is placed between that exhibited by the corresponding cobalt and nickel promoted molybdate catalysts. All the promoted catalysts supported on anatase are proved to be more active than the corresponding ones supported on γ-alumina. The active species in the promoted catalysts supported on anatase are molybdenum clusters of relatively small size (somewhat larger in the tungstate catalysts) in perpendicular or titled configuration bonded with the support through Ti–O–Mo(W) and Ti–S–Mo(W) bridges. These clusters involve also Co- S - Mo(W) or Ni- S- Mo(W) bridges in a certain distance from the anatase surface. The combined promoting action of the anatase and the promoter is due to the transfer of electronic density from the titanium or promoter to the molybdenum or tungsten through the aforementioned bonds. The first transfer weakens the Mo(W)-S facilitating CUS formation. The second transfer becomes more strong the (Co or Ni) –S bonds whereas it becomes less strong the S–Mo (W) bonds. Thus, a sulfur atom located between molybdenum and a promoter atom is related to the medium intensity of the “metal-sulfur” bonds. This should be related to the manifestation of high activity according to the Sabatier principle. This means that the promoter optimizes the intensity of the “metal-sulfur” bonds allowing the CUS formation and in the same time allowing the adsorption of the reactant in the CUS through its sulfur atom leading to the formation of weak surface bonds of the Mo with the reactant. The second mechanism may be also takes place through the Ti–S–Mo(W) sites. The relatively high activity of the catalysts supported on anatase with respect to the corresponding ones supported on γ-alumina may be attributed to the following reasons. In the first case this due to the combined action of all promoter atoms and the anatase surface described above. In the second case this due only to the promoter atom remained in the surface and decorating theMoS2 slabs. In fact, it is not manifested promoted action through the bridges Al-Ο-Mo of the Co-Mo-S or Ni-Mo-S phases of type I. The conformation of this type is somewhat similar to that of the drastic sites described above for the anatase. Moreover, the catalytic activity of the slabs of the very active Co-Mo-S or Ni-Mo-S phases of type II (hexagonal slabs weakly bonded with the γ-alumina surface where the CUS are formed at the corners and edges) does not influenced form the support surface. This simply facilitates the dispersion of these slabs. In the fifth part of the dissertation we have chosen certain very active catalysts which we had synthesized mainly in the third part. These have been tested in the hydrodeoxygenation of phenol considered as a probe reaction for deoxygenating of bio-oil fractions drawn from the cellulose- lignin biomass (bio fuels of second generation). The catalysts were tested after reduction using a fixed bed reactor working at high pressure. It was found that the most active catalyst was a NiMo catalyst supported on aloumina. Moreover, non promoted and Co promoted molybdena catalysts supported on anatase were proved to be also very active. Three of the most active catalysts were tested in the sulfided state for the simultaneous h.d.s. and h.d.o. of dibenzothiophene and phenol. These combined reactions are used as model ones for the co-processing of bio- and petrol-oil fractions. Concerning h.d.o. it was fount that the kind of the support does not affect the catalytic behaviour. On the other hand the influence of the kind of the support on the h.d.s. is rather complicated.

Στόχος της εργασίας αυτής είναι η ανάπτυξη καταλυτών υδρογονοεπεξεργασίας που είναι στηριγμένοι σε βιομηχανικό ανατάση με σχετικά μεγάλη ειδική επιφάνεια, καθώς και η εφαρμογή τους σε αντιδράσεις υδρογονοαποθείωσης και υδρογονοαποξυγόνωσης. Η εργασία διαιρείται σε πέντε ενότητες, που αφορούν: α) Βιβλιογραφική ανασκόπηση, β) Μη ενισχυμένους μολυβδαινικούς καταλύτες που έχουν παρασκευαστεί με την τεχνική της «ισορροπίας εναπόθεσης – διήθησης», γ) Μη ενισχυμένους μολυβδαινικούς καταλύτες που έχουν παρασκευαστεί με διαφορετικές τεχνικές, δ) Ενισχυμένους με κοβάλτιο ή νικέλιο μολυβδαινικούς ή βολφραμικούς καταλύτες στηριγμένους σε ανατάση ή γ-αλούμινα, και ε) Αξιολόγηση δραστικών καταλυτών που είχαν παρασκευαστεί στις προηγούμενες ενότητες και είχαν αξιολογηθεί για αντιδράσεις υδρογονοαποθείωσης σε ατμοσφαιρική πίεση, σε για αντιδράσεις υδρογονοαποξυγόνωσης και ταυτόχρονης υδρογονοαποξυγόνωσης – υδρογονοαποθείωσης σε υψηλή πίεση.Βασικές επιδιώξεις αυτής της διατριβής είναι η αριστοποίηση των παρασκευαστικών παραμέτρων και η διερεύνηση των δραστικών θέσεων. Για την επίτευξη των στόχων της εργασίας συνθέσαμε έναν πολύ μεγάλο αριθμό καταλυτών χρησιμοποιώντας διάφορες τεχνικές εμποτισμού και διάφορες παρασκευαστικές συνθήκες. Τους καταλύτες αυτούς τους αξιολογήσαμε για την διεργασία της υδρογονοαποθείωσης χρησιμοποιώντας ως αντιδράσεις μοντέλα την υδρογονοαποθείωση του θειοφαινείου και του διβενζοθειοφαινείου και για την διεργασία της υδρογονοαποξυγόνωσης την αντίδραση της υδρογονοαποξυγόνωσης της φαινόλης. Για την εδραίωση «σχέσεως δομής – δραστικότητας» και την διερεύνηση των δραστικών θέσεων οι καταλύτες χαρακτηρίστηκαν με διαφορετικές τεχνικές όπως: φασματοσκοπία διάχυτης ανάκλασης ορατού–υπεριώδους (DRS), θερμοπρογραμματισμένη αναγωγή (TPR), περίθλαση ακτίνων–Χ (XRD), μέτρηση του πορώδους και της ειδικής επιφάνειας (BET), φασματοσκοπία Ramman και φασματοσκοπία ηλεκτρονίων ακτίνων-Χ.Η προσπάθεια αριστοποίησης τεχνικών εμποτισμού και παρασκευαστικών παραμέτρων έδειξε ότι η για τους μη ενισχυμένους καταλύτες η εφαρμογή της τεχνικής της «ισορροπίας εναπόθεσης – διήθησης», κάτω από συνθήκες που να οδηγεί σε στηριγμένους καταλύτες με επιφανειακή συγκέντρωση μολυβδαινίου 3.2 άτομα Μο/nm2 (που αντιστοιχεί σε μονόστρωμα μονομερών τροδοντικών μολυβδαινικών ειδών), οδηγεί στην σύνθεση πολύ δραστικού καταλύτη υδρογονοαποθείωσης. Η προσπάθεια αυτή έδειξε επίσης ότι γι’ αυτή την επιφανειακή συγκέντρωση και ο τυπικός υγρός εμποτισμός καταλήγει σε πολύ δραστικό καταλύτη. Η εισαγωγή του ενισχυτή (κοβαλτίου, νικελίου) μετά την εισαγωγή και σταθεροποίηση της μολυβδαινικής φάσης με ξηρό ή υγρό εμποτισμό, οδήγησε στην παραγωγή ενισχυμένων καταλυτών με δραστικότητα μια τάξη μεγέθους μεγαλύτερη από εκείνη των μη στηριγμένων καταλυτών. Τα ίδια ακριβώς ισχύουν και για στηριγμένους βολφραμικούς καταλύτες ενισχυμένους με νικέλιο. Οι καταλύτες που παρασκευάστηκαν σε ανατάση είναι γενικά πιο δραστικοί από τους αντίστοιχους καταλύτες που παρασκευάστηκαν σε γ-αλούμινα.Οι δραστικές θέσεις στους μη ενισχυμένους καταλύτες που είναι στηριγμένοι σε ανατάση είναι σχετικά μικρού μεγέθους πλειάδες θειούχου μολυβδαινίου (κάπως μεγαλύτερες βολφραμίου) σε κάθετη ή πλάγια διαμόρφωση προς την επιφάνεια του ανατάση που συνδέονται με γέφυρες Ti–O–Mo(W) και Ti–S–Mo(W). Η δράση του ανατάση εκδηλώνεται μέσω των γεφυρών αυτών με μετακίνηση ηλεκτρονικής πυκνότητας από το τιτάνιο προς το μολυβδαίνιο (βολφράμιο), κάτι που μειώνει την ισχύ του δεσμού μολυβδαινίου (βολφραμίου) – θείου, αυξάνει τον αριθμό των συντακτικά ακόρεστων θέσεων και επομένως την δραστικότητα, καθώς στα κενά αυτά προσροφώνται μέσω ατόμων θείου οι αντιδρώσες θειούχες ενώσεις.Στη δραστική θέση των ενισχυμένων καταλυτών και σε κάποια απόσταση από την επιφάνεια σχηματίζονται επιπλέον και γέφυρες Co-S-Mo(W) ή Ni-S-Mo(W). Η συνδυασμένη δράση του ενισχυτή εκδηλώνεται μέσω των γεφυρών αυτών με μετακίνηση ηλεκτρονικής πυκνότητας από τον ενισχυτή προς το μολυβδαίνιο (βολφράμιο), κάτι που ενώ ισχυροποιεί το δεσμό ενισχυτή - οξυγόνου (θείου) μειώνει την ισχύ του δεσμού μολυβδαινίου (βολφραμίου) - θείου. Έτσι ένα άτομο θείου ευρισκόμενο ανάμεσα σε ένα άτομο μολυβδαινίου και σε ένα άτομο ενισχυτή παρουσιάζει ενδιάμεση ισχύ δεσμού «μετάλλων – θείου», που σύμφωνα με την αρχή του Sabatier σχετίζεται με την εκδήλωση της μέγιστης δραστικότητας. Αυτό με άλλα λόγια σημαίνει ότι η παρουσία του ενισχυτή αριστοποιεί την ισχύ του δεσμού «μετάλλων – θείου» έτσι που να δημιουργείται μεν ανιονικό κενό θείου, αλλά να επιτρέπει στην θειούχο αντιδρώσα ουσία να προσροφηθεί στο ανιονικό αυτό κενό μέσω του ατόμου του θείου κάνοντας ενδιάμεσους επιφανειακούς δεσμούς με το μολυβδαίνιο ή /και τον ενισχυτή.