MgO Catalyst - American Chemical Society

Subscriber access provided by UOW Library

Article

Liquid Phase Hydrogenolysis of Glycerol over Highly Active 50%Cu-Zn (8:2)/MgO Catalyst: Reaction Parameter Optimization by Using Response Surface Methodology (RSM) Al Ameen A, Smita Mondal, Satyanarayana Murty Pudi, Nitin Naresh Pandhare, and Prakash Biswas Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b00766 • Publication Date (Web): 10 Jul 2017 Downloaded from http://pubs.acs.org on July 16, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Liquid Phase Hydrogenolysis of Glycerol over Highly Active 50%Cu-Zn (8:2)/MgO Catalyst: Reaction Parameter Optimization by Using Response Surface Methodology (RSM)

Al Ameen A, Smita Mondal, Satyanarayana Murty Pudi, Nitin Naresh Pandhare and Prakash Biswas*

Department of Chemical Engineering, Indian Institute of Technology Roorkee, Roorkee247667, Uttarakhand, India.

*Corresponding Author: Tel.: (+91)-1332-28-5820; Fax: (+91)-1332-27-6535 E-mail address: [email protected]; [email protected]

1 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 40

Graphical abstract

OH 212oC, 45 MPa, 20 wt% glycerol, 7 wt% catalyst, 12 h

HO 50%Cu-Zn(8:2)/MgO

OH HO 1,2 -propanediol (1,2-PDO)

OH Glycerol Conversion (~ 99%)

RSM


Experimental X glycerol : 97% S 1,2-PDO : 94%

Page 3 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Research highlights •

Highly active and stable 50% Cu-Zn(8:2)/MgO catalyst was developed for selective conversion of glycerol to 1,2-PDO

•

The effect of different interaction parameter on glycerol conversion and 1,2-PDO selectivity was studied to optimize the reaction condition.

•

Response surface methodology (RSM) based on Central Composite Design (CCD) was used to develop an empirical process model.

•

The analysis of variance showed that the empirical models fitted very well the experimental data with very high values of determination coefficient (R2>0.95)

•

Maximum glycerol conversion of 96.9% with 94.2 % selectivity to 1,2 PDO was obtained at the optimum reaction condition obtained by numerical optimization by using RSM.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 40

Abstract In this study, a highly promising bimetallic 50%Cu-Zn(8:2)/MgO

catalyst was

developed for selective hydrogenolysis of glycerol to 1,2-propanediol (1,2-PDO). The catalytic activity was evaluated in a high pressure autoclave reactor. Results demonstrated that the incorporation of Zn into Cu/MgO catalysts enhanced the glycerol conversion and selectivity to 1,2-PDO due to the hydrogen spill over effect. Experimentally maximum glycerol conversion of 98.7% with 94.6% selectivity to 1,2-PDO was achieved at mild reaction conditions. Response surface Methodology (RSM) was employed to study the interaction of the various reaction parameters and develop an empirical process model followed by the optimization of the reaction parameters. The high values for determination coefficients (>0.95) obtained by analysis of variance indicated that the quadratic regression models developed were highly satisfactory. The results of numerical optimization demonstrated that 97.8% glycerol conversion with 93.5% selectivity to 1,2-PDO can be achieved at 212 oC and at 4.5 MPa pressure.

Keywords: hydrogenolysis of glycerol, response surface methodology, central composite design, empirical process model, optimization


Page 5 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Nomenclature α – Coded dimensionless value representing final level of independent variable a0– Constant term in empirical model ai – Coefficient of linear term in empirical model. aij – Coefficient of interaction term in empirical model. aii – Coefficient of square term in empirical model A – Temperature B – Glycerol concentration c – Replicate number for centre point C – Catalyst loading D – Reaction time k – Number of independent variables (factors) N – Total number of experiments performed as per the experimental matrix R2– Determination coefficient xi – Value of independent variable ∆x – Step change between any two levels Xi– Coded dimensionless value of independent variable. Y1 – Glycerol conversion (response 1) Y2– Selectivity to 1,2-PDO (response 2)


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 40

1. INTRODUCTION Recently, conversion of biomass and biomass derived intermediates to valuable chemicals has gained significant interest due to uncertain supply of fossil fuels and global warming issues.1,2,3 In this context, biodiesel has recently emerged as a potential alternative and regarded as a renewable liquid transportation fuel derived from biomass. The world production of biodiesel is estimated to be 180 million tonnes by 2016 at an average growth rate of 42% per year.3,4 However, during the production of biodiesel via transesterification of vegetable oils and animal fats, surplus amounts (~ 10 wt.%) of glycerol is obtained as a byproduct.5 Thus, the conversion of crude glycerol to valuable products is essential for economic viability of biodiesel industry. Among the various glycerol conversion process proposed, hydrogenolysis of glycerol to 1,2-propanediol (1,2-PDO) has received a significant commercial importance. 1,2-PDO is currently produced either by chlorohydrin process or hydroperoxide process, which involves the hydration of propylene oxide derived from fossil resources.6 Moreover, production of 1,2-PDO from biomass derived glycerol is an innovative and green process. 1,2-PDO is a major commodity chemical with the overall growth rate of 4% and the global demand for 1,2-PDO is estimated to rise ~ 1.5 million metric tons per annum (MMTPA).7 1,2-PDO is widely used as a raw material for the production of unsaturated polyester resins, food additives, paints, cosmetics, liquid detergents, printing ink, plasticizers, functional fluids such as antifreeze, de-icing and heat transfer fluid etc.8.9 Development of a suitable catalyst is essential which will favour the selective cleavage of C-O bond of glycerol to obtain higher selectivity to 1,2-PDO.10 Various noble metals (Ru, Rh,

Pt, Pt) Al2O3,

11-14

and non-noble metals (Cu, Co, Ni)

ZrO2,

MgO,

3,10,15-21

catalysts supported on SiO2, ZnO,

Cr2O3, ZnO–Al2O3, and zeolites have been developed and their

performance have been evaluated in recent past. However, the disadvantages of noble metal catalysts are high cost and poor selectivity to 1,2-PDO due to the formation of degradation


Page 7 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

products by the cleavage of the C-C bonds of glycerol. The performance of noble metal catalysts was improved by the external addition of solid acids or bases.13,21 Among the transition metal catalysts, Cu based catalyst showed significant performance for hydrogenolysis of glycerol.22,17 Cu-boehmite,22 acid-base bi-functional Cu-MgO on USY,2 Cu0.4/Zn0.6 Mg 5.0 Al2O8.6,23 Cu-MgO,9,17,18 catalysts have been reported as active and selective to 1,2-PDO in the pressure range of 2-4 MPa pressure and at 200oC. Wang et al.15 reported 90% selectivity to 1,2-PDO over Cu-ZnO catalyst at very low glycerol conversion (< 5%). Balaraju et al.18 obtained 50% conversion of glycerol with 93% selectivity towards 1,2-PDO over 20wt% Cu/MgO catalyst at 40 bar pressure and at 200°C in presence of 6 wt% catalyst. Yuan et al.17 reported 72% conversion of glycerol and 97.6% selectivity to 1,2-PDO at 180°C and 3MPa pressure by using Cu(15)/MgO catalyst. Jiang et al.24 investigated the effect of addition of ZnO on Cu-Ni/Al2O3 catalyst in absence of added H2 and they found that ZnO increases the catalytic activity significantly. Optimization is a process which is used to obtain the best possible solution for a system. Conventionally, optimization is done by analyzing the effect of one parameter at a time on the experimental responses. Since this method considers only one variable at a time, it is not possible to study the interaction effects of multiple process parameters. Also, the conventional technique requires more number of experiments which increase the time as well as the expense of the process. Therefore, it is always preferable to carry out the optimization by using multivariate statistical technique. Response surface methodology (RSM) is an effective method for the optimization of multiple process parameters.25 RSM make use of statistical, regression modelling and optimization techniques to find an appropriate relation between multiple process variables and one or more responses and subsequently locate an optimum within the design space. 26,27 The method generates immense information within the design space and enable to study the interaction effects of individual variable on responses.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RSM optimizes multiple factors simultaneously, the relative easiness in handling and efficiency made it extensively useful in solving analytical problems. RSM methodology is summarized in the flowchart shown in Scheme 1. The RSM study for hydrogenolysis of glycerol is very limited in the previous literature. 28, 29 RSM generates large amounts of data in the domain of study which needs to explain the behaviour of the system with a minimum number of experiments. The contour and surface plots shows the variation and influence of interaction effects of two variables at a time on the responses while the statistical analysis helps to determine the significance of these parameters. This clarifies the binary effects of the critical independent parameters on responses thereby increase the accuracy of the prediction and directing a displacement towards the optimal region. Wolosiak-Hnat et al.28 reported the single-response optimization of technological parameters of glycerol hydrogenolysis to 1,2-PDO over Cu/Al2O3 catalyst. Glycerol conversion of 88.7% with 94.3% selectivity to 1,2-PDO was reported at 205oC, in presence of 75wt.% glycerol, 5.5wt.% catalyst after 23 h of reaction. Dam et al.29 evaluated the performance of Pt/Al2O3 catalyst with silicotungstic acid as an additive for the formation of 1,3-propanediol (1,3-PDO). The reaction parameters were optimized by experimental design and obtained 49% conversion of glycerol with 28% selectivity to 1,3-PDO. The main focus of this study was to develop a promising catalyst and the catalytic process for selective conversion of glycerol to 1,2-PDO. Previous literature information demonstrated that, the catalyst morphology, structure, acid-base bi-functional property are solely responsible for high activity and selectivity towards 1,2-PDO. It is also observed from the previous study that hydrogenolysis of glycerol to 1,2-PDO followed two steps reaction process i,e. dehydration followed by hydrogenation. Acidic and/or basic sites of the catalyst propagated dehydration step and metallic sites favoured hydrogenation step. Therefore, the metal catalyst having bi-functional acidic and/or basic sites may be a good choice for


Page 8 of 40

Page 9 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

selective conversion of glycerol to 1,2-PDO. Cu based catalysts are highly selective to cleave C-O bond and suppress the cleavage of C-C bond. It is also found that incorporation of Zn increase the acidity/basicity of the catalyst which enhanced the activity and also the selectivity to 1,2-PDO.15,24,30,31 Several previous reports also suggested that, ZnO can act as a reservoir for atomic hydrogen which enhance hydrogen spillover, thereby increasing the hydrogenation capability of the catalyst.3,23 In this study, highly active and selective Cu-Zn bimetallic catalyst supported on MgO was developed and the catalytic performance for the liquid phase hydrogenolysis of glycerol to 1,2-PDO was evaluated. An empirical process model was developed using RSM to study the effect of various reaction parameters such as temperature, glycerol concentration, catalyst loading and reaction time on glycerol conversion and 1,2-PDO selectivity and finally optimize the reaction parameters. The interaction effects among the various reaction parameters are discussed, and the empirical models were validated with the experimental data and applied to predict the responses. The results suggested that 50%Cu-Zn(8:2)/MgO catalyst was highly active and selective to 1,2PDO at mild reaction condition as compared to the results reported earlier. Optimization results also demonstrated that the experimental value of glycerol conversion and 1,2-PDO selectivity is very close to the optimum value obtained by using Design expert. Design expert is a statistical software package capable of performing design of experiments (DOE). It allows screening and optimization up to fifty number of experimental factors. The graphical tools and statistical analysis in the software determines the effects and interaction between the factors. The numerical optimizer gives the optimum values according to the constrains set by the user. The results obtained in this study established that, 50%Cu-Zn(8:2)/MgO catalyst may be promising for commercial application.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2. EXPERIMENTAL SECTION 50%Cu-Zn(8:2)/MgO catalyst was prepared by precipitation-deposition method. In a typical synthesis, aqueous solution (1.0 mol/L) of NaHCO3 (>99.8%, Thomas Baker, India) was added drop-wise to the aqueous solution of copper (II) nitrate trihydrate (>99%, >99%, Thomas Baker, India) and zinc (II) nitrate hexahydrate (>99%, Thomas Baker, India) with constant stirring until the pH was 7-8. After precipitation, the required amount of MgO light (>99.8%, Thomas Baker, India) was added to the solution under vigorous stirring. The slurry obtained was further aged at room temperature for 12 h followed by filtration. The obtained solid was dried in an oven at 110 oC overnight followed by calcination in air at 550 oC for 4 h. The catalyst was characterized by various methods, such as Brunauer-Emmett-Teller (BET) surface area measurement, X-ray diffraction (XRD), NH3-temperature programmed desorption (NH3-TPD), temperature programmed reduction (TPR) and scanning electron microscopy (SEM) with energy dispersive X-ray spectra (EDX). Detail experimental procedure and product analysis were discussed in our previous study.9,32

3. RESULTS AND DISCUSSION 3.1 Catalyst characterization The BET surface area (SBET) and pore volume (VP) of 50%Cu-Zn(8:2)/MgO catalyst was determined from N2 adsorption–desorption isotherm. The BET surface area and total pore volume of pure MgO support was 92 m2/g and 0.18 cm3/g, respectively. The surface area (24 m2/g) of the catalyst was decreased significantly after metal impregnation due to the formation of multi-layer clusters of metal oxide and blockage of pores of support by copper and zinc particles.18,33 The characteristics X-ray diffraction patterns of reduced 50%Cu-Zn(8:2)/MgO and MgO catalysts were compared in order to determine the changes in structure of the catalyst if


Page 10 of 40

Page 11 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

any after metal impregnation (Figure 1(A)). As shown in Figure 1(A), pure MgO shows the major diffraction peaks at 2θ = 36.9o, 42.9o, 62.2o, 78.6o, respectively, corresponding to the (111), (200), (220) and (222) crystal planes of cubic MgO (JCPDS: 78-0430).9,32 For 50%CuZn(7:3)/MgO catalyst, the peaks corresponding to metallic Cu and ZnO phase were observed indicating well dispersion of Cu and ZnO on MgO support. Metallic copper peak was detected at the 2θ values of 43.3°, 50.4°, 74.1° corresponds to (111), (200) and (220) crystal plane, respectively (JCPDS: 85-1326).9 However, the diffraction peaks detected at the 2θ angle of 31.74°, 34.5°, 36.2°, and 56.5° for 50%Cu-Zn(8:2)/MgO catalyst, ascribed the (101), (100), (002), (101), (110) crystal planes of hexagonal ZnO phase (JCPDS: 80-0075). The average crystallite size of metals in the reduced catalysts was calculated by using Scherrer formula from the line width of their respective XRD peaks. The copper crystallite size was estimated from the line width of the peaks corresponding to (111), (200) and (220) crystal planes and it was ~ 37.1 nm. Reduction behaviour of 50%Cu-Zn(8:2)/MgO and MgO are shown in Figure 1(B). Pure MgO, did not show any reduction peak up to 800oC. For 50%Cu-Zn(8:2)/MgO catalyst, a broad reduction peak was observed at 290°C. In our previous study 9 it was observed that at higher copper loading (>25wt %), CuO/MgO catalyst followed single step reduction at around 263-293°C. At this temperature, CuO directly reduced to Cuo without formation of Cu2O or other intermediate oxides. It is also found that complete reduction of ZnO to Zn was quite difficult.34 Xia et al.23 reported ZnO increased the reducibility of CuO species by hydrogen spill over and the reduction of Cu0.4Zn

0.3Mg 5.3Al2O9

catalyst was obtained at

~300°C. In this study, 50%Cu-Zn(8:2)/MgO catalyst showed a reduction peak at ~290°C implied synergistic effect of CuO and ZnO as well as hydrogen spill over of ZnO. The acidic strength of the catalyst was determined by NH3-TPD analysis and the obtained NH3-TPD pattern of the catalyst is shown in Figure 1(C). The result showed that the


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

acidic sites were distributed in the temperature region of 250-500oC and >700oC. These regions were attributed to the desorption of ammonia from the medium and strong acidic sites, respectively.35,36 Total acidic strength of the catalyst was estimated based on the desorbed amount of ammonia and it was 2.13 mmol NH3 gcat-1. The basic strength of the catalyst was determined by CO2-TPD and the TPD profile is shown in Figure 1(D). Very high temperature (>600oC) desorption of CO2 represented the presence of strong basic sites on the surface of the catalyst. Total basic strength of the catalyst was 1.8 mmol CO2 gcat-1. SEM analysis was performed to determine the morphology of 50%Cu-Zn(8:2)/MgO catalyst and rod like morphology was obtained as shown in (Figure 1(E)) which is in a good agreement with the previous literature.37

3.2 Experimental design in Response Surface Methodology (RSM) and empirical process model Central composite design (CCD) in RSM was used to develop an empirical process model to optimize and study the interaction of reaction parameters. The optimization process involved mainly three steps which included experimental design by using a statistical approach, evaluation of coefficients in the developed model and prediction of responses followed by adequate check of the model.26, 38 For optimization and parametric study, four factors and five levels CCD consisting 30 numbers of experiments of different level combinations were designed and performed in a liquid phase autoclave reactor. The factors were the independent experimental variables and the levels were the value of these variable at which experiments were conducted. The variables that have major effects on the responses were identified and two level factorial experiments were conducted to identify the levels of the selected variables. The levels were selected in accordance with previous literature


Page 12 of 40

Page 13 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

keeping the objective to maximize glycerol conversion and selectivity od 1,2-PDO. The total number of experiments required for CCD is calculated according to equation (1), given by

N = k 2 + 2k + c

(1)

Where N is the total number of experiments, k denotes factor number and c is the replicate number (c=6) for the centre point. The reaction factors and the experimental domain chosen for analysis were reaction temperature (190-220 oC), glycerol concentration (10-50 wt.%), catalyst loading (5-9 wt%) and reaction time (6-14h), respectively. The design of experiments (DOE) was classified into three sets, factorial points consisting of 16 (2k) number of experiments, 6 centre points and 8 (2k) star points (axial points). For factorial points, each factor was varied into two levels; all factors were kept at midrange for centre points while axial points were identical to centre points except one factor was varied in each experimental run. The five levels were represented as [–α, -1, 0, 1, α]. The value of α depend on the k 4 number of factors and is given by α = 2 . Generally, the values are 1.48, 1.68 and 2.00 for

two, three and four factors. The factorial points corresponded to all combinations of -1 and +1. The responses studied were glycerol conversion (Y1) and selectivity of 1,2-PDO (Y2). The empirical model developed was tested with the analysis of variance (ANOVA) for responses and reaction parameters studied.

The empirical model developed to correlate the responses as function of reaction factors is presented in the form of a quadratic equation (2) as follows:

k

k

k

Yu = a0 + ∑ai xi + ∑∑aij xij + ∑aiixi2 i=1

i< j

i=1

(2)

where Yu is the predicted response, ao, ai, aii and aij are the constant, linear, squared and interaction terms, respectively. xi denotes the real value of independent variable, i = 1,


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 40

2……………n and u = 1, 2………n. For the regression analysis, the independent variables are transformed into coded dimensionless value according to the following equation (3):

Xi =

xi − xo ∆x

(3)

where Xi is the coded dimensionless value of the independent variable, xi and xo are the corresponding real values of the independent variable at ith level and centre point, respectively, and ∆x is the step change. The optimization of glycerol conversion and 1,2-PDO selectivity was performed by using Design Expert 7.0 software. The levels studied for each factor was transformed into a coded dimensionless value. The actual and coded values of each factor as per experimental design and their range are shown in Table 1. The complete set of experiments and the real values of the responses are shown in Table 2. The empirical model equations developed by regression analysis to determine the influence of reaction factors on responses in terms of dimensionless coded values are given in Table 3. The p and F values are estimated by the software via regression analysis. The numerical values of p and F are used to improve the goodness fit of the model. For each response, the significant terms in the model were identified and the goodness fit of the model was examined by analysis of variance (ANOVA). The terms with high probability value (pvalue > 0.2) indicate that the interaction of these reaction factors is not significant to the model. Hence, terms BC,CD in the quadratic polynomial model developed for glycerol conversion and terms AD,CD,B2 from the model for selectivity to 1,2-PDO were neglected. This improved the fitness of the model and the modified empirical model for glycerol conversion and selectivity to 1,2-PDO are given in Table 4 and Table 5, respectively. Finally, the F-value (65.93 and 32.23) and p-value (0.95) and adjusted R2 (>0.92) for both models (Table 6) indicated that the model fitted the expermental data very well. Since the value of R2 increased with number of terms in the model equation irresepective of the significance of the terms, a more unbiased estimation of model goodness was adjusted R2. Moreover, the small difference between adjusted R2 and predicted R2 (95% at a temperature of >205oC and at the glycerol concentrations of 95.5%. The maximum glycerol conversion of 99.2% was achieved at 215-220oC and after 12-14 h of reaction time. In the temperature range of 205-212oC and after 10-12h of reaction glycerol conversion was 94-98%. At initial stage of reaction i,e. after 6-8 h, even at high reaction temperatures (>205oC), glycerol conversion was low (75-83%). This result suggested that the reaction temperature of 210-220oC and the reaction time of 12-14h were beneficial to achieve higher glycerol conversion. 50%Cu-Zn(8:2)/MgO catalyst provided excellent selectivity to 1,2-PDO (95.3-96.6%) at 10 h) reaction time and at 7.5wt.%), a gradual increase in glycerol conversion was observed. Maximum glycerol conversion of 99.2% was achieved at 215-220oC in presence of 8 wt% catalyst loading. At low catalyst loading (5-7 wt%), glycerol conversion was found to be 75-86 % at low reaction temperatures (195-205oC) and 89-99.3% at 205oC and higher. The increase in glycerol conversion with increasing the catalyst loading was due to the accessibility of more active centres in the catalyst. 9,12,18 As observed earlier, the selectivity to 1,2-PDO decreased with reaction temperature and reached the minimum value of 91% at


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

220oC. The 1,2-PDO selectivity decreased at higher temperature due to over hydrogenolysis of 1,2-PDO. However, the effect of catalyst loading on the selectivity to 1,2-PDO was insignificant. With increasing the catalyst loading, the ratio of the products formed in the reaction mixture were not varied significantly. As a result the ratio of carbon present in a product to the total carbon present in all products were almost constant which gives almost constant selectivity. This results demonstrated highly selective nature of the catalyst. Maximum of 96.2% selectivity to 1,2-PDO was observed at 190-195oC and it was higher than 95% at

MgO Catalyst - American Chemical Society

Recommend Documents