Multivariate Analysis in Selective Nitroacetophenone Conversion by

Subscriber access provided by University of Newcastle, Australia

Full Paper

Multivariate analysis in selective nitroacetophenone conversion by hydrogen sulfide under phase transfer catalysis Ujjal Mondal, and Sujit Sen Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00287 • Publication Date (Web): 21 Dec 2016 Downloaded from http://pubs.acs.org on December 23, 2016

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.

Organic Process Research & Development 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 16

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

Organic Process Research & Development

Multivariate Analysis in Selective Nitroacetophenone Conversion by Hydrogen Sulfide under Phase Transfer Catalysis Ujjal Mondal and Sujit Sen* Catalysis Research Laboratory, Department of Chemical Engineering, National Institute of Technology Rourkela, Odisha – 769008, India

Corresponding author: Tel: +91-9938246590; fax: +91-6612462022 Email addresses: [email protected], [email protected] (S. Sen) *

ACS Paragon Plus Environment


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

ABSTRACT: Response Surface Methodology (RSM) was employed to model and optimize an improved Zinin reduction of a nitroarene, p-Nitroacetophenone (p-NAP), by hydrogen sulfide (H2S) under bi-liquid phase transfer catalysis. A novel Zinin reagent, H2S laden aqueous N-methyldiethanolamine (MDEA) solution was prepared and used for this purpose. A quadratic regression model was tested with a multivariate experimental design based on the relationship between p-NAP conversion (Response) and four independent variables - temperature, catalyst concentration, p-NAP: sulfide mole ratio and MDEA concentration. The optimum values of the independent variables were found as temperature 339.45 K, catalyst concentration of 0.082 kmol/m3, p-NAP/sulfide mole ratio of 0.452, MDEA concentration of 2.20 kmol/m3 and maximum p-NAP conversion of 96.31% has been attained. The analysis of variance (ANOVA) has been used to evaluate the goodness of the fit of the model, and the desirability function has been used to find the value of the optimized parameters to maximize the pNAP conversion. Keywords: Response Surface Methodology; p-Nitroacetophenone; hydrogen sulfide; Zinin reduction; phase transfer catalysis.


Page 2 of 16

Page 3 of 16

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


1. INTRODUCTION Hydrogen sulfide (H2S) is a very poisonous and hazardous gas. Minimal exposure of H2S may lead to different fatal conditions in any living organism and environment if not properly handled.1–3 Presence of H2S in pipelines also affects many industrial operations. 3,4 A number of techniques have been adopted by industrialists for capture and utilization of H2S gas. Among them, Claus process is the mostly accepted industrially practised process of treating H2S gas. In the convensional process of treating H2S, chemisorption of H2S gas is performed in aqueous alkanolamines, followed by partial oxidation of the gas via conventional Claus process to produce elemental sulfur.5-12 Above mentioned processes have some shortcomings and thus the development of a viable alternative process for the conversion of H2S gas to valuable chemicals is highly desirable.7,13,14 A number of absorbents have been reported in the literature for removal of H2S gas 14–21 and among all the absorbents, alkanolamine based separation processes have gained maximum popularity.9,11 In our current reaction, H2S absorbed in N-methyldiethanolamine (MDEA) 18 has been used for the synthesis of commercially valuable chemicals like p-aminoacetophenone (p-AAP). High-value chemical intermediates can be produced from aromatic amines, which are widely used as raw materials for many utility chemicals like, fibre, explosives, polymers, cosmetics, pesticides, dyes, etc.22–27. Anilines are industrially used as an intermediate for the production of dyes, artificial pigments 28,29.Some patents are available on the reduction of p-NAP to corresponding amines or sometimes to alcohols by hydrogenation under high temperature and pressure in the presence of palladium or Raney nickel as catalysts.30–34 Reduction of p-NAP to the aniline-ketone, anilinemethylene, aniline-alcohol by hydrogenation reaction with different palladium, platinum, and rhodium catalyst have been studied by Hawkins et al.35 A few researchers have produced p-AAP from p-NAP by using homogeneous or heterogeneous catalysts such as bimetallic Rh3Ni1 nanoparticles, ionic liquids, 36 rhodium/silica catalyst. 37 There are some constraints related to those processes like the selectivity of product, waste minimization, environmental safety, overall operational costs. The current reaction also follows Zinin reduction pathway, but the approach of utilization of industrially available H2S-laden MDEA is a new one. Different aqueous solution alkanolamines which are used for the absorption of H2S gas are diethanolamine (DEA), monoethanolamine (MEA) 20,21, methyldiethanolamine (MDEA) 18. MDEA is advantageous over other alkanolamines as it possesses some qualities like minimum corrosion effect, reduced solvent loss due to less vapour pressure, chemically stable and economically beneficial. The reduction reaction was carried out in a batch reactor in biphasic (liquid-liquid) mode. To augment the reaction rate, tetrabutylphosphonium bromide (TBPB) was used as a phase transfer catalyst. Development of kinetic equations is difficult for the system under study as Zinin reduction under L-L PTC mode is a complex reaction system and partitioning of catalyst in both phases is unknown and not available in the literature. The transport properties and concentration of active catalyst species (𝑄𝑆𝐻, 𝑄𝑆𝑄, 𝑄𝑆2 𝑄, 𝑄𝐻𝑆𝑂3 ) in both the phases cannot be obtained experimentally as they could not be isolated from solution. In order to scale up of current reaction from laboratory scale to industry scale, an optimization study based on the mathematical model is therefore highly awaited. Optimization of a chemical process is all needed to be done before its industrial implementation and practice, which includes understanding and finding variables responsible for a good outcome (yield, conversion). The most common approach to optimizing a chemical reaction is by changing one variable at a time (OVAT).38 The interaction between the operational variables of the process can be understood from this univariate approach. During recent times, inefficient OVAT approach is often replaced by effective chemometric methods, such as response surface methodology (RSM), based on statistical approach such as the design of experiments (DoE).39 This statistical method can be applied to study the relationship between independent variables and response for multivariable systems.40,41



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

Success in the proposed work can well be thought of as a viable replacement to Claus process as for the proposed work, the operational cost is lower, the process would be more environment-friendly and more valuable product would be obtained in comparison with the Claus process where elemental sulfur is the sole product. Hence, the current work was taken up to propose the optimum conditions for p-NAP conversion in the NAP-H2S-laden MDEA-TBPB-Toluene system using DoE. RSM with central composite design (CCD) has been employed to evaluate the influence of different responsible factors on p-NAP conversion. Consequently, to develop an efficient process for achieving maximum p-NAP conversion, the desirability function for optimization of the overall process has been employed, and an optimized response model was proposed. 2. MATERIALS AND METHODS 2.1. Materials. N-methyldiethanolamine (MDEA) (≥99%), toluene (≥ 99%), p-nitroacetophenone (NAP), sodium thiosulfate pentahydrate (Na2 S2 O3 . 5H2 O) and sodium hydroxide pellets (NaOH) of analytical grade were procured from Merck (India), Ltd., tetra-n-butylphosphonium bromide was obtained from Sigma-Aldrich (China) Ltd. potassium Iodide (KI) and potassium iodate (KIO3) of analytical grade were purchased from Rankem (India) Ltd., New Delhi, India. Starch soluble, ferrous sulfide sticks (FeS), sulfuric acid (≥ 98%) were purchased from Thermo Fisher Scientific India Pvt., Ltd., Mumbai, India. 2.2. Preliminary Study. In order to select the most influencing parameters affecting p-NAP conversion, the parameters were screened by carrying out some preliminary experiments and constructing main effect plot with the help of Minitab (version 16.1.1). The slope of the line of main effect plot is the measure of the effectiveness, higher the slope of the line, higher would be the effect on the response. Horizontal line to the x-axis indicates the absence of main effect and the response mean is independent of the levels of the factors.

2.3. Multivariate Experimental Design. The most influencing parameters can be found after the screening of the parameters. This problem involves a standard RSM with five level four-factor CCD. Selection of the levels was accomplished on the basis of results obtained in the preliminary study, considering limits for the experimental set-up, working conditions for each chemical substances and the previous experiences in dealing with variables to get desired results. Multiple regression analysis of experimental data is studied with the statistical software package Design-Expert for fitting developed polynomial equations.42 Two variables were changed simultaneously while other two variables were kept at a fixed value corresponding to the stationary point during regression analysis of experimental data. Based on the fitted polynomial equations contour plots and 3D surface plots were prepared. Analysis of variance (ANOVA) and the coefficient of determination (adjusted R2 and R2) are the main factors for determining the quality of fit of the model. The statistical parameters are estimated by ANOVA. Some insignificant coefficients were reviewed and manually removed to refine the model. 2.4. Experimental method. For the preparation H2S-laden MDEA, H2S gas was bubbled through aqueous MDEA solution kept in an ice bucket and the process continued till required sulfide concentration was achieved. Sulfide concentration of the solution was measured by iodometric titration.43 A 150 ml glass full baffled batch reactor of 6.5 cm internal diameter was used to carry out all the experiments. The agitation effect for proper mixing was provided with a six-bladed glass turbine stirrer of 2 cm diameter, and the stirring speed was regulated with a digital speed regulation system. The reaction temperature was kept uniform by placing the reactor in an isothermal water bath and temperature was controlled by a PID controller (±1°C). Every experimental run was consisting of


Page 4 of 16

Page 5 of 16

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


30 ml each of an aqueous phase and an organic phase. First, the aqueous phase was poured into the glass batch reactor, and then agitation of the aqueous phase was provided by glass stirrer till a steady state temperature was reached. After that organic phase was added to the reactor which is containing organic substrate (p-NAP) and phase transfer catalyst (TBPB) dissolved in organic solvent (toluene). Then the reaction was started by switching on the stirrer. The stirrer was turned off after eight hours of reaction, and a slight amount (0.2 ml) of the sample was withdrawn from the organic phase for analysis in Gas Chromatography (GC).

Figure 1. Schematic diagram of the experimental setup for H2S absorption followed by reaction in a batch reactor and finally sample collection and analysis in GC-MS. 2.5. Analysis of phases. The organic samples were analyzed by gas chromatograph (Agilent GC 7820A model). The column used was a 2 m x 3 mm capillary column (DB-5MS) and nitrogen was used as a carrier gas with a flow rate of 1.6 cm3/min. The injection port was heated at a fixed temperature of 250 °C, and a FID detector was used at a temperature of 300 °C. The synthesized product was confirmed by GC-MS (Agilent 5977A model). The conversion of p-NAP was calculated with respect to the concentration of limiting reactant (p-NAP). 3. RESULT AND DISCUSSION 3.1. Overall reaction. The overall reduction of p-Nitroacetophenone (1) by H2S-laden MDEA in the presence of TBPB as phase transfer catalysis to produce p-Aminoacetophenone (2) as sole product is shown in Scheme 1 below. Scheme 1. Overall reaction of p-NAP reduction by H2S

3.2. Proposed mechanism of the reaction. The mechanism of reduction reaction of aromatic nitroaromatic compounds by Zinin reduction under L-L PTC mode of reaction has long been established.44–46 Current L-L PTC system consists of an aqueous phase (H2S absorbed in aqueous MDEA solution) and an organic phase (substrate p-NAP dissolved in toluene) and a quaternary phosphonium salt (TBPB). TBPB is a phase transfer catalyst in the reaction, is soluble in both the phases. Thus, the reaction mechanism can be explained by Stark’s extraction mechanism. Valency of sulphur can vary between -2 to +6. Therefore during the course of the reaction sulphur can be in − 46 multiple anionic forms ((HS − , HSO− , HSO− 2 , HSO3 ).



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

In the aqueous phase, H2S gas reacts with MDEA and forms two anions (Sulphide (S 2− ) and hydrosulphide (HS − ) ions).47–49 These two anions form an ionic equilibrium in the aqueous phase, which is same as observed in aqueous ammonium sulfide solution,50 as shown in Scheme 2. Quaternary cations (Q+) present in the aqueous phase readily binds with hydrosulphide anions (HS − ) and transfer to the organic phase. Then Q+ HS − ion pair reacts with p-NAP to form p-AAP through the formation of multiple intermediates. The reduction reaction of the aromatic nitro compound to aromatic amines is an electron transfer reaction which requires 6 electron transfer through the formation of intermediates 3 and 4 (nitroso benzene and hydroxyamines).51, 52 The existence of these intermediates was undetected during GC-MS analysis because the formation and disappearance of these molecules are very fast. At the end of the organic phase reaction (7), the quaternary cation 2− became inactive Q+ HSO− 3 . Then it transfers to aqueous phase and reacts with sulfide (S ) to + − regenerate again (Q HS ) as shown in Equation (6). This completes the full catalytic cycle.

Scheme 2. Proposed mechanism for selective reduction of p-NAP by H2S-laden MDEA solution in Zinin reduction.

3.3. Screening of parameters. The main effect plot is shown in Fig. 2. The effect of stirring speed and reaction time on p-NAP conversion is insignificant as shown in the figure. Based on the preliminary studies, it can be concluded that increasing stirring speed more than 1000 rev/min does not help to enhance total conversion of the substrate.43 In order to eliminate mass transfer effect entirely, the stirring speed was maintained at 1500 rev/min. Reaction rate became slower after high initial conversion and to achieve maximum conversion all experiments were carried out for the duration of 8 hours. So, the most influencing parameters are temperature, catalyst conc., p-NAP/ sulfide concentration ration and MDEA concentration. Catalyst concentration and MDEA concentration have a strong effect on p-NAP conversion, and the temperature is having a weak effect on the conversion of p-NAP. p-NAP/ sulfide concentration ratio has a negative impact on conversion.


Page 6 of 16

Page 7 of 16

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


As p-Aminoacetophenone (p-AAP) is the only product identified by GC-MS, p-NAP conversion is the only response chosen for our design.

Figure 2. The main effect plot of control factors The levels and range of all the independent variables taken are listed in Table 1. The levels of each independent variable are set accordingly the knowledge gained from preliminary studies to achieve desired results. Table 1. Coded levels and range of independent variables for experimental design Level

lowest low mid high highest

Coded level XI -2 -1 0 1 2

A: Temperature (K) 303 313 323 333 343

Uncoded level B: Catalyst Conc. C: p-NAP/ Sulfide (kmol/m3) mole ratio 0.006 0.1211 0.025 0.2422 0.044 0.3633 0.063 0.4844 0.082 0.6055

D:MDEA Conc. (kmol/m3) 1.00 2.00 3.00 4.00 5.00

3.4. Development of Regression Model Equation. Table 2 is the design matrix generated based on the range and levels of Table 1. The total number of experimental runs conducted were 30 as proposed by the central composite design and the response (p-NAP conversion) was calculated to fit a second order polynomial model. Actual levels of independent variables, temperature (A), Catalyst concentration (B), p-NAP: sulfide mole ratio (C) and MDEA concentration (D) with ranges are shown in Table 2.



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 8 of 16

Table 2. Experimental design matrix and results - A 24 full factorial CCD with six replicates of the Centre point Run

Actual level of variables Response Temperature Catalyst Conc. p-NAP/Sulfide MDEA p-NAP (K) (kmol/m3) mole ratio Conc. (kmol/m3) Conversion (%) 1 323 0.044 0.6055 3.00 44.20 b 2 323 0.044 0.3633 3.00 47.92 3 313 0.025 0.2422 2.00 50.16 4 333 0.063 0.4844 2.00 61.60 5 313 0.025 0.2422 4.00 66.50 6 323 0.006 0.3633 3.00 44.49 7 313 0.063 0.4844 2.00 46.10 8 333 0.025 0.4844 4.00 49.33 9 333 0.025 0.2422 2.00 60.91 10 333 0.025 0.2422 4.00 72.00 11 313 0.025 0.4844 4.00 45.10 12 323 0.082 0.3633 3.00 82.40 13b 323 0.044 0.3633 3.00 47.00 14 333 0.063 0.4844 4.00 73.11 15 323 0.044 0.3633 1.00 43.63 16 313 0.063 0.4844 4.00 58.46 17 323 0.044 0.1211 3.00 88.50 18 343 0.044 0.3633 3.00 70.00 19b 323 0.044 0.3633 3.00 48.00 20 313 0.063 0.2422 2.00 66.73 21 333 0.063 0.2422 2.00 82.00 22 333 0.063 0.2422 4.00 95.21 b 23 323 0.044 0.3633 3.00 47.50 24 323 0.044 0.3633 5.00 69.00 25b 323 0.044 0.3633 3.00 47.40 26 313 0.025 0.4844 2.00 32.83 27 303 0.044 0.3633 3.00 45.69 28 313 0.063 0.2422 4.00 81.22 29 333 0.025 0.4844 2.00 41.87 30b 323 0.044 0.3633 3.00 46.40 b centre points

3.5. Model Selection and Fitting. The model is selected on the basis of different statistical parameters which were accurately analyzed by DOE. Fitting and significance of developed model can be understood by “p-value” and “F-value” of ANOVA analysis. Further refinement of the model can be done based on the results obtained by ANOVA analysis. The software suggested that the quadratic model is the most suitable for the provided data based on the best lack of fit test, favorable F value, “prob>F” value, standard deviation, and R2 value. It is shown in the form of the quadratic polynomial equation: 4

4

4

4

Y   0    i X i    ii X     ij X i X j i 1

i 1

2 i

i 1 j i 1


(8)

Page 9 of 16

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


Where Y is (the response of dependent variable) the conversion of p-NAP conversion (%) and βis are the model coefficients calculated from experimental data. Table 3 shows the ANOVA result summary of the quadratic response surface model fitting. With 0.0001 probability and F-value of the model of 602.37 at the 0.01% level suggests a high degree of precision of the model. The "Lack of Fit F-value" is low as 3.28 signifies that there is a 10.12% chance of occurring of the large "Lack of Fit F-value" due to noise. The coefficient of determination (R2) of 99.82% implies that the model fitting was accurate and only 0.18% of the total variability was unable to explain by the model. The "Pred R-Squared" value of 0.9908 in comparison with the "Adj R-Squared" of 0.9966; which signifies that the model prepared can give a very well estimation of the response of the system. The low value of the coefficient of variation (CV = 1.62%) of the model also suggests a high accuracy and better reliability of the experimental data. Adequate precision was also found to be more than 4. Figure 3A shows the comparison between predicted values and actual values for p-NAP conversion. From Figure 3A, it is clear that the experimental values obtained are close to the predicted ones. Thus, we may imply that the developed model can successfully predict the conversion of p-NAP. The normal plot of residues for p-NAP (Figure 3B) was found significant as didn’t show any deviation of the variance. All the linear, quadratic and interaction model terms are found to be significant as P-value for all terms are below 0.01. AC, BD, BC are the interaction parameter that is having P-value more than 0.05 and considered as insignificant. From the coded values of the main effect, quadratic and interaction parameters, it could be seen that A, B, C, D, AB, AD, CD, A2, B2, C2 and D2 are major influencing parameters. AC, BC and BD are the insignificant terms and are not necessary for explaining the conversion of p-NAP. Therefore, the insignificant terms are ignored for better fitting of the model. The significance of these quadratic and interaction effects between the variables had indeed been lost in earlier studies53 where experiments were carried out following OVAT approach. Final model equation for p-NAP conversion in terms of actual factors (subjected to range shown in Table 1) has been obtained as follows: % NAP Conversion = +2811.40715 - 17.04116 × A - 3688.96024× B - 308.71952 × C + 19.13035× D + 9.83454× AB - 0.076206 × AD - 5.95634× CD +0.026951× A2 +11343.68075× B2 + 328.75893 × C2 + 2.31257× D2

(9)

Table 3. ANOVA for response surface quadratic model for p-NAP conversiona Source

Sum of Squares

Degree of Freedom

Mean

F-Value

Square

Model

7612.22

14

543.73

Residual

13.54

15

0.90

Lack of Fit

11.75

10

1.17

Pure Error

1.79

5

0.36

Cor Total

7625.76

29

a

p-value Prob > F

602.37

< 0.0001

3.28

0.1012

Coefficient of determination (R2) =0.9982; Adj. R2 =0.9966; Pred. R2 =0.9908; C.V. % = 1.62; Adeq. Precision = 94.703; Std. Dev.= 0.95.



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

Figure 3. Plot of (A) predicted values versus actual values and (B) Normal % probability plot of internally studentized residuals for p-NAP conversion.

3.6. Response surface analysis. 3D surface and contour plots are the graphical representation of the regression equations implemented to achieve or predict the optimum influence of variables and to gain a better understanding of combined interaction variables the range provided.54, 55 3D surface plots show not only the combined but also the sole impact of the variables on the response. Figure 4A & 4B shows the interaction effect of catalyst concentration and temperature on the conversion of p-NAP. Conversion increased with the increase of catalyst concentration and temperature, either individually or combined. The catalyst concentration is having pronounced effect on conversion. Highest conversion of 71% achieved with 0.063 kmol/m3 catalyst concentration and at 333K temperature. As the catalyst loading increased the number of active catalyst increase in the reaction mixture and so the conversion. The transition state theory explains that the rates of most organic reactions can be enhanced by the increase of reaction temperature, and the collision theory suggests that at higher temperature the number of effective collision between molecules increases and hence the rate of reaction also increases. Figure 4C & 4D shows the interaction between MDEA concentration and temperature on p-NAP conversion. While the effect of MDEA concentration has a positive influence on p-NAP conversion, the effect of temperature is higher. MDEA does not react with p-NAP but it is affecting equilibrium among MDEA-H2S-H2O and the solution is dominated by sulfide and hydrosulfide anions. With more MDEA concentration the equilibrium between active anions shift more towards concentration of sulfide anions, and the conversion p-NAP is enhanced (equation 1). Highest conversion of 63.5% was achieved with MDEA concentration of 4.0 and at 333K. Figure 5A and 5B shows the interactive effect of p-NAP: sulfide concentration ratio and MDEA concentration on the conversion of p-NAP. According to the figure, the conversion of p-NAP is


Page 10 of 16

Page 11 of 16

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


decreased with the increased ratio of p-NAP/sulfide concentration ratio. MDEA concentration has the weak effect on p-NAP conversion in comparison to p-NAP/sulfide concentration ratio. Reducing the p-NAP/sulfide concentration ratio from 0.48 to 0.24, have increased the conversion from 49% to 72%. As the ratio of p-NAP/sulfide concentration decreases, the number of moles of sulfide per moles of pNAP increases and that results in an increase in the conversion.

Figure 4. Contour and 3D surface plot of the effect of different parameters on the conversion of pNAP (A) contour plot of the interaction of Temperature and Catalyst concentration. (B) 3D surface plot of the interaction of Temperature and Catalyst concentration. (C) Contour plot of the interaction of Temperature and MDEA concentration. (D) 3D surface plot of the interaction of Temperature and MDEA concentration.



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 12 of 16

Figure 5. Contour and 3D surface plot of the effect of different parameters on the conversion of pNAP (A) Contour plot of the interaction of p-NAP: sulfide concentration ratio and MDEA concentration. (B) 3D surface plot of the interaction of p-NAP: sulfide concentration ratio and MDEA concentration.

3.7. Optimization of influencing factors. Optimal conditions for conversion of p-NAP was obtained with the response surface methodology on the basis of the desirability function. The weight or importance of the goal is set accordingly to achieve the desired goal. The goal is set for each constraint in range among four other possibilities (none, minimum, maximum, and target). The criterion for all parameters that are adopted to achieve maximum p-NAP conversion are shown in Table 4. Based on the range of the parameters provided, the optimum criteria for a maximum p-NAP conversion of 96.31% were found to be temperature of 339.45 K, catalyst concentration of 0.082 kmol/m3, p-NAP/sulfide concentration ratio of 0.452 and MDEA concentration of 2.20.

Table 4. Optimized reaction conditions based on selected criteriaa

Criteria

Temperature (K)

Cat. Conc. (kmol/m3)

pNAP:Sulfide mole

MDEA conc.

Conversion Desirability

ratio Maximize p-NAP conversion and all parameters are in range

339

0.082

0.452

2.20

96.31

1.000

Maximize p-NAP conversion and minimum catalyst concentration

341

0.006

0.218

4.84

96.04

1.000

Maximize p-NAP conversion and

303

0.008

0.148

3.95

95.72

1.000


Page 13 of 16

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


minimum temperature Maximize p-NAP conversion and minimum MDEA concentration

342

0.012

0.127

1.0

95.40

1.000

Maximize p-NAP conversion and minimum catalyst concentration and temperature

303

0.006

0.127

3.46

95.97

1.000

a

Limit of % p-NAP conversion = 32.83 – 95.21% ;

4. Model verification and confirmation. In order to confirm the established model, a number of verification experiments were carried out based on the optimum conditions for achieving maximum pNAP conversion. After three replicate experiments, an average of 96.07% conversion was achieved, as shown in Table 5. So, it can be concluded that there is a good agreement between the model's predicted value and experimental value and thus the validity of the model simulated for p-NAP is confirmed. Table 5. Results of validation experimentsa. Run No. Conversion% Selectivity% 1 96.0 100.0 2 96.12 100.0 3 96.09 100.0 Average 96.07 100.0 Predicted by Eq. (5) 96.07 a Experiments conducted at conditions: Temperature = 339.5K; Catalyst conc. =0.082 kmol/m3; pNAP: Sulfide mole ratio=0.452; MDEA Conc. = 2.20 kmol/m3.

5. CONCLUSION Based on experimental design methodology, a statistical model of single response surface optimization was carried out on the liquid-liquid phase transfer catalyzed reduction of p-NAP with H2S-ladden MDEA. A quadratic model was developed based on the relationship between p-NAP conversion and four independent variables (temperature, catalyst conc., p-NAP/sulfide mole ratio and MDEA conc.). The contour plots, 3D surface plots and the value of different regression coefficients have clearly explained the single parameter effect and interaction effects of dual parameters on the pNAP conversion. The model fits well and is verified with optimum conditions for highest p-NAP conversion. R2 values of the regression analysis confirm a good agreement between the experimental data and predicted response.

ASSOCIATED CONTENT Supporting Information



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

GC spectra of reactant and products; MS spectra of products Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS UM is thankful for doctoral fellowship from Ministry of Human Resource and Development (MHRD), India during the tenure of the work.

REFERENCES (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)

Polychronopoulou, K.; Fierro, J. L. G.; Efstathiou, a. M. Appl. Catal. B Environ. 2005, 57 (2), 125–137. Sabereh Rezaei; Tavana, A.; Sawada, J. a.; Wu, L.; Junaid, A. S. M.; Kuznicki, S. M. Ind. Eng. Chem. Res. 2012, 51 (38), 12430–12434. Dhage, P.; Samokhvalov, A.; Repala, D.; Duin, E. C.; Tatarchuk, B. J. Phys. Chem. Chem. Phys. 2011, 13 (6), 2179–2187. Li, L.; King, D. L. Catal. Today 2006, 116 (4), 537–541. FARAJI, F. Int. J. Hydrogen Energy 1998, 23 (6), 451–456. S. Cervera-March, L. Borrell, J. Giménez, R. Simarro, J. M. A. Int. J. Hydrogen Energy 1992, 17 (9), 683–688. Huang, H.; Yu, Y.; Chung, K. H. Energy & Fuels 2009, 23 (9), 4420–4425. Petrov, K.; Srinivasan, S. Int. J. Hydrogen Energy 1996, 21 (3), 163–169. Z. Mao, A. Anani, R. E. White, S. S. and A. J. A. J. Electrochem. Soc. 1991, 138 (5), 1299– 1303. Noringt, J. O. N. E.; Fletchers, E. A. Energy 1982, 7 (8), 651–666. Groisil, M.; Ibrahim, S.; Gupta, A. K.; AlShoaibi, A. Energy Procedia 2015, 75, 3066–3070. Mark A Plummer. Process for recovering sulfur and hydrogen from hydrogen sulfide. US5334363 A, 1994. Yadav, G. D.; Jadhav, Y. B.; Sengupta, S. J. Mol. Catal. A Chem. 2003, 200 (1–2), 117–129. Belmabkhout, Y.; De Weireld, G.; Sayari, A. Langmuir 2009, 25 (23), 13275–13278. Prabu, M.; Ramalingam, K. RSC Adv. 2015, 5 (24), 18554–18564. Wang, X.; Wang, X.; Song, C.; Song, C. Transform 2009, No. 13, 5777–5783. Pipatmanomai, S.; Kaewluan, S.; Vitidsant, T. Appl. Energy 2009, 86 (5), 669–674. Mandal, B.; Bandyopadhyay, S. S. Environ. Sci. Technol. 2006, 40 (19), 6076–6084. Kim, Y. J.; You, J. K.; Hong, W. H.; Yi, K. B.; Ko, C. H.; Kim, J. Sep. Sci. Technol. 2008, 43 (4), 766–777. Al-Baghli, N. a.; Pruess, S. a.; Yesavage, V. F.; Selim, M. S. Fluid Phase Equilib. 2001, 185 (1–2), 31–43. Kaewsichan, L.; Al-Bofersen, O.; Yesavage, V. F.; Selim, M. S. Fluid Phase Equilib. 2001, 183–184, 159–171. Catino SC, F. E. Concise Encyclopaedia of Chemical Technology; John Wiley & Sons: New York, 1985. Sakaue, S.; Tsubakino, T.; Nishiyama, Y.; Ishii, Y. J. Org. Chem. 1993, 58, 3633–3638. Shabbir H. Gheewala, A. P. A. Water Sci. Technol. 1997, 36 (10), 53–63. Boon, N.; De Gelder, L.; Lievens, H.; Siciliano, S. D.; Top, E. M.; Verstraete, W. Environ. Sci. Technol. 2002, 36 (21), 4698–4704. Abdessamad Grirrane, Avelino Corma, H. G. Science 2008, 322 (December), 1661–1664. Dell’Anna, M. M.; Intini, S.; Romanazzi, G.; Rizzuti, A.; Leonelli, C.; Piccinni, F.; Mastrorilli, P. J. Mol. Catal. A Chem. 2014, 395, 307–314. Chen, Jin-Fang; Jia, Tao; Huang, X.-L. Yingyong Huaxue 2000, 17 (6), 672–674. Manieh, A. A.; Sayed, A. Z. Al-Azhar Bull. Sci. 1995, 6 (1), 35–48. W. Michalowicz, L. Haven. Preparation of m-amino-.alpha.-methylbenzyl alcohol. U.S Patent


Page 14 of 16

Page 15 of 16

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


(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)

4021487, 1977. W. Emerson, R. A. H. Preparation of 2-amino-alpha-methylbenzyl alcohol, 1968. Tinsley Samuel W. Process for the hydrogenation of meta-nitroacetophenone, 1976. Rylander Paul N. Hydrogenation of nitrophenyl alkyl ketones, 1968. Lower, A.; Carbinols, A. Preparation of monocyclic aminoaryl lower alkyl carbinols, 1954. Hawkins, J. M.; Makowski, T. W. Org. Process Res. Dev. 2001, 5 (3), 328–330. Elavarasan, P.; Kondamudi, K.; Upadhyayula, S. Chem. Eng. J. 2009, 155, 355–360. Currall, K.; David Jackson, S. Top. Catal. 2014, 57 (17–20), 1519–1525. Schwindt, M. A.; Lejon, T.; Hegedus, L. S. Organometallics 1990, 9 (13), 2814–2819. Tanyildizi, M. S.; Özer, D.; Elibol, M. Process Biochem. 2005, 40 (7), 2291–2296. D.C. Montgomery. Design and Analysis of Experiments, fifth.; John Wiley and Sons: New York, 2001. A.I. Khuri, J. A. C. Response Surfaces: Design and Analyses; Dekker: New York, 1987. The Design-Expert software was utilized for generate the design matrix, regression analysis and the graphical representation of the results and models; see: (a) Using Design-Expert version 8.0.2, Stat-Ease, Inc., Minneapolis, USA 2016. Singh, G.; Nakade, P. G.; Mishra, P.; Jha, P.; Sen, S.; Mondal, U. J. Mol. Catal. A Chem. 2016, 411, 78–86. Yadav, G. D.; Lande, S. V. Adv. Synth. Catal. 2005, 347 (9), 1235–1241. Yadav, G. D.; Lande, S. V. Ind. Eng. Chem. Res. 2007, 46 (10), 2951–2961. Yadav, G. D.; Jadhav, Y. B.; Sengupta, S. Chem. Eng. Sci. 2003, 58 (12), 2681–2689. Xu, H.; Zhang, C.; Zheng, Z. Society 2002, 6175–6180. Xu, H.; Zhang, C.; Zheng, Z. Simulation 2002, 2953–2956. Maity, S. K.; Pradhan, N. C.; Patwardhan, A. V. Ind. Eng. Chem. Res. 2006, 46, 7767–7774. Maity, S. K.; Pradhan, N. C.; Patwardhan, A. V. Applied Catalysis A: General. 2006, pp 251– 258. Klausen, J.; Ranke, J.; Schwarzenbach, R. P. Chemosphere 2001, 44 (4), 511–517. Zhou, B.; Song, J.; Zhou, H.; Wu, T.; Han, B. Chem. Sci. 2016, 7 (1), 463–468. Maity, S. K.; Pradhan, N. C.; Patwardhan, A. V. Appl. Catal. A Gen. 2006, 301 (2), 251–258. Liu, H. L.; Chiou, Y. R. Chem. Eng. J. 2005, 112 (1–3), 173–179. Vildozo, D.; Ferronato, C.; Sleiman, M.; Chovelon, J.-M. Appl. Catal. B Environ. 2010, 94 (3– 4), 303–310.



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

Table of Contents Graphic 99x32mm (300 x 300 DPI)


Page 16 of 16

Multivariate Analysis in Selective Nitroacetophenone Conversion by

Recommend Documents