Optimization in synthesize of organic macrocyclic compounds in presence of nano copper chromite catalyst

Abstract

Nowadays, different structures of organic macrocyclic compounds are considered because of their attractive applications. One of the main problems in the synthesis of these materials is their long reaction time but low reaction yield. The use of catalysts can be effective in solving this problem. Among the catalysts, nano-copper chromite can be a good choice due to its good performance in the synthesis of organic compounds. In addition, the Response Surface Methodology was used to investigate the effective parameters in the synthesis more precisely. Based on the previous results of the synthesis and experiments, the catalyst content from 0% to 5% to raw material and reaction time between 24 and 96 h was chosen for the design of the experiment. After determining the reaction yield results, a suitable model was selected and its accuracy was evaluated. Results showed for yields above 95% with minimum catalyst (2.29%) the reaction time of 88 h and for minimum time (65 h), 3.85% of the catalyst is required. This yield with copper chromite nanocatalysts approximately compared to conventional methods for the synthesis of calix[4]resorcinarene was doubled.

Keywords

Synthesize optimization response surface methodology copper chromite nanocatalyst calix[4]resorcinarene

1 Introduction

Ring compounds with several aromatic rings, which are synthesized by the ring-closure mechanism (macrocylisation), are called macrocyclics or cyclic oligomers. The well-known members of this family are calixarenes, cyclodextrins, porphyrins and crown ethers [1 –3]. One of the most famous calixarenes is calix[4]resorcinarenes, which are synthesized by the condensation reaction of resorcinol with various types of aldehydes in the presence of acids [4, 5]. The notable structural feature of calix[4]resorcinarenes, i.e. bowl-shaped cavities with aromatic walls, has led to widespread application of these compounds in supramolecular chemistry. Due to the upper hydrophilic edge (hydroxyl groups) and the lower hydrophobic edge (alkyl groups), calix[4]resorcinarenes are easily functionalized through electrophilic substitution reactions of aromatic rings as well as acylation and alkylation of hydroxyl groups [6, 7]. calix[4]resorcinarenes have antibacterial properties and have been used in the medical and food packaging industries. In addition, these compounds can be used as host–guest complex templates [8], phase transfer catalysts, liquid crystals, as extractants [9], modifying agents [10] and sensors [11, 12]. Due to the importance of these materials, the main problem in their synthesis is long reaction time and low product yield that should be solved [13 –16].

Nanoparticles have been recently used as a suitable framework for the synthesis of catalysts. Nano catalysts have a higher surface to volume ratio than bulk catalysts. The higher surface area of grains affects their physical properties. There has been a growing interest in recent decades in transition metal oxide nano crystals and mixed-metal oxides due to their catalytic, optical and electrical properties and also applications in hydrogen production and many other engineering and scientific areas. Among transition metals, chrome and copper compounds are of great importance in the copper chromite (CuCr₂O₄.CuO) catalysts due to multiple applications. These catalysts are used in oxidation, hydrogenation, dehydrogenation, etc. [17, 18]. Copper chromite nano catalyst is prepared by hydrothermal and co-precipitation methods and the like. In the meantime, chemical co-precipitation is extensively used because of simple reaction conditions, high purity and ease of separation and lower cost [19]. This catalyst in nano scale has been widely used in the synthesis of organic compounds and acceptable effects in chemical reactions are reported [20 –22].

In this research, due to high yield, simple preparation, facile separation and purification of the catalyst, the use of nano-copper chromite catalyst is introduced for the synthesis of calix[4]resorcinarenes. This method can also be used for the synthesis of other calix compounds. Since the fictionalization of calix[4]resorcinarene increases its potential for the preparation of multifunctional compounds with higher yield, novel methods have received great attention for the synthesis of calix[4]resorcinarene. Also for a closer look at the effects of the parameters on the reaction yield and optimizing the reaction conditions, Response Surface Methodology (RSM) is used.

2 Experimental

2.1 Reagents

Resorcinol, acetaldehyde, ethanol (99%) and hydrochloric acid (37%) were purchased from Merck Company (Germany). Argon gas was purchased from an Iranian company. Copper chromite (CuCr₂O₄.CuO) nano catalyst with an average particle size of 50 nm was synthesized according to our previous work [23].

2.2 calix[4]resorcinarene synthesis

calix[4]resorcinarene was synthesized according to the method reported previously [1]. 75 mL distilled water and 37.5 mL HCl 37% were added to a 0.15 M (16.5 g) resorcinol solution, 75 mL ethanol and different percentage of copper chromite (CuCr₂O₄.CuO) nano catalyst in a 250 mL two-necked flask under argon atmosphere. Then 0.15 M (6.6 g) acetaldehyde was added dropwise to the reaction mixture for 30 min at 15°C. Milky sediment was obtained after rising the reaction temperature to 50°C and stirring for another 1 h at the same temperature. The reaction mixture was then cooled down to the ambient temperature. The reaction mixture was stirred under an argon atmosphere for completing the reaction. The reaction mixture was then filtered and the precipitate was re-crystallized in water and ethanol for further purification and reaction yield was calculated.

Fourier-transform infrared (FTIR) spectrum was recorded to confirm the structure of synthesized calix[4]resorcinarene. The results of completely characterization of the synthesized calix[4]resorcinarene have been fully reported in our previous research [24].

2.3 Design of experiments

RSM was used to study statistical and mathematical methods to optimize the factors influencing the reaction [25, 26]. The data obtained from the appropriate experiments are used by the central composite design (CCD) of RSM to estimate the optimized condition and regression model equation. The optimization of the two variables of reaction time and catalyst amount in the synthesis of calix[4]resorcinarenes was done through CCD of RSM. The ranges and levels of variables (catalyst weight and reaction time) investigated in the research are given in Table 1. The levels of the parameters varied from –1 to +1 and zero levels of the variables indicate the centre values.

Table 1
The levels of variables in RSM

Variable Symbol Low (–1) Level middle (0) High (+1)

Catalyst (% wt.) A 0 2.5 5

Time (h) B 24 60 96

Variable	Symbol	Low (–1)	Level middle (0)	High (+1)
Catalyst (% wt.)	A	0	2.5	5
Time (h)	B	24	60	96

The results were analyzed by applying the response plots and analysis of variance (ANOVA) in Design Expert Version 7.0 software [27, 28]. Design Expert software suggested 16 experimental runs according to Table 2. All the 16 synthesizes were done and yields of reaction were calculated.

Table 2

Synthesis conditions design and response of RSM

Std	Run	Factor 1 Catalyst (%)	Factor 2 Time (h)	Responce Yield (%)
4	1	2.14	64.29	89.23
5	2	0.00	24.00	8.71
10	3	0.00	60.00	27.85
3	4	3.38	24.00	17.96
12	5	0.00	96.00	36.11
2	6	5.00	44.01	53.81
11	7	4.48	24.00	19.58
16	8	2.50	96.00	88.91
13	9	5.00	96.00	92.08
1	10	5.00	96.00	92.89
6	11	0.00	96.00	46.09
14	12	0.00	24.00	8.65
15	13	4.84	24.00	23.54
8	14	1.73	33.00	29.71
9	15	4.38	69.49	92.51
7	16	2.50	96.00	91.35

2.4 Instrumental methods

Fourier-transform infrared spectroscopy (FTIR, BOMEMMB-Series 1998) with a spectral resolution of 4 cm^–1 and a scanning rate of 21 cm^–1, is used to study the chemical structure of products. Precipitates were filtered by a Buchner funnel and a vacuum pump (VE135, VALUE, China) operated with a power of 0.33 hp. Design Expert software version 7.0 used to help optimize the process.

3 Results and discussion

3.1 FTIR spectroscopy of calix[4]resorcinarene

The product was dried in a vacuum oven (6 h, 50°C, 0.04 mm Hg) white solid, mp 360°C. To ensure the correct synthesis of all 16 samples, four synthesized samples at different times were selected and FTIR spectroscopy was done. The FTIR spectra of samples 12, 6, 15 and 8, synthesized at 24, 44, 69 and 96 h, respectively, have been compared (Fig. 1) [29]. IR spectrum, KBr, ν, cm^–1: 3367 (O–H), 2972, 2929, 2874 (C–H Aliphatic), 1518 (C = C Aromatic), 1118, 1096 (C–O), 1424, 1372 (CH₃). All the spectra were the same and indicating the correct synthesis of calix[4]resorcinarene.

Fig. 1

Synthesis method and FTIR spectrum of calix[4]resorcinarene synthesized at different reaction times.

3.2 Results of the experiment design

According to the results of yield from the synthesized experiments (Table 2) and quadratic model, the ANOVA was conducted for yield results and presented in Table 3. Results of ANOVA in Table 3 showed that the ‘lack of fit F-value’ is not significant relative to the pure error and the p-values lower than 0.0001 indicated that the model and model terms were statistically significant. These indicated that the quadratic model was valid for this research. Also high value and closeness of the regression coefficient (R² = 0.984) and adjusted regression coefficient ( $R_{adj}^{2}$ = 0.976) corroborated the suitability of the model [27].

Table 3
Analysis of variance (ANOVA) for the response of the yield percentage of calix[4]resorcinarenes

Source Sum of squares df Mean square F value p-Value Prob > F

Model 1.95 5 0.39 122.98 <0.0001 Significant

A-Catalyst 0.37 1 0.37 117.12 <0.0001

B-Time 1.29 1 1.29 406.05 <0.0001

AB 1.254E-004 1 1.254E-004 0.040 0.8463

A² 0.092 1 0.092 29.11 0.0003

B² 0.17 1 0.17 53.21 <0.0001

Residual 0.032 10 3.168E-003

Lack of Fit 0.023 5 4.557E-003 2.56 0.1626 Not significant

Pure Error 8.896E-003 5 1.779E-003

Cor Total 1.98 15

Std. Dev 0.056 R-Squared 0.9840

Mean 1.59 Adj R-Squared 0.9760

C.V. % 3.54 Pred R-Squared 0.9585

PRESS 0.082 Adeq Precision 31.035

Source	Sum of squares	df	Mean square	F value	p-Value Prob > F
Model	1.95	5	0.39	122.98	<0.0001	Significant
A-Catalyst	0.37	1	0.37	117.12	<0.0001
B-Time	1.29	1	1.29	406.05	<0.0001
AB	1.254E-004	1	1.254E-004	0.040	0.8463
A²	0.092	1	0.092	29.11	0.0003
B²	0.17	1	0.17	53.21	<0.0001
Residual	0.032	10	3.168E-003
Lack of Fit	0.023	5	4.557E-003	2.56	0.1626	Not significant
Pure Error	8.896E-003	5	1.779E-003
Cor Total	1.98	15
Std. Dev	0.056	R-Squared	0.9840
Mean	1.59	Adj R-Squared	0.9760
C.V. %	3.54	Pred R-Squared	0.9585
PRESS	0.082	Adeq Precision	31.035

3.3 Theoretical and experimental data comparison

Relationship between the response (yield of reaction) and the independent variables (reaction time and amount of catalyst) in terms of coded factors and actual factors are expressed by the quadratic models 1 and 2, respectively. $\log_{10} Yield = + 1.89 + 0.19 \times A + 0.33 \times B - 3.925 E - 003 \times A \times B - 0.18 \times A^{2} - 0.26 \times B^{2}$ (1) $\begin{matrix} \log_{10} Yield & = + 0.245 + 0.221 \times A + 0.0334 \times B - 4.361 E - 005 \times A \times B - 0.029 \\ \times A^{2} - 1.989 E - 004 \times B^{2} \end{matrix}$ (2) where, A and B are amount of catalyst (% wt.) and reaction time (h), respectively.

To ensure the accuracy of the selected model, the normal probability diagram was plotted as shown in Fig. 2; the normal probability diagram shows the fit and distribution of the data obtained by the software. If the graph is linear and the data are scattered on or near the line, and the smaller amount of data are far from the straight line, the data correspond to the normal probability distribution, and the correct model is chosen unless this model is not suitable. As can be seen in Fig. 2, data are near the straight line and there are no data out of standard and a proper linear regression model is seen [30, 31].

Fig. 2

The normal probability diagram.

The graph of the predicted values and the actual values obtained from the experimental results is shown in Fig. 3 This diagram actually shows the accuracy of the chosen model versus the results. Since the results are as close as possible to the line, the accuracy of the model is indicated. As can be seen in Fig. 3, the results are near to the predicted line, especially at high yields where three data points are found on the trend line. Therefore, the accuracy of our model is verified [31, 32].

Fig. 3

The diagram of predicted verses actual.

3.4 Dependence of reaction parameters

Figure 4 shows the dependence of the catalyst content (Fig. 4-a) and reaction time (Fig. 4-b) on the reaction yield. As can be seen, by increasing the catalyst content and reaction time the reaction yield increased and then started decreasing. This increase is about 3.75% for the catalyst value and 84 h for the time.

Fig. 4

Dependence of (a) amount of catalyst and (b) reaction time with error bars to yield.

The results in Fig. 4 show only the effect of one parameter on a constant value of the other parameter and cannot give a precise idea of the influence of the parameters that are working in parallel.

For better understanding the interactions of the parameters affecting the reaction yield, from the test results, a contour plot was prepared and presented in Fig. 5. This graph illustrates favorably the interactions of the synthetic parameters that simultaneously influence the synthesis process. The red areas in this graph indicate high yields and as moving from red to yellow, green and blue areas, the reaction yields decrease [33]. This graph can indicate well the timing of the synthesis of calix[4]resorcinarene with the specified amount of catalyst to achieve the highest yield, which is economically important for the design of experiment.

Fig. 5

Contour plots of optimal design for calix[4]resorcinarene synthesis.

The curves in Fig. 5 indicate points with the same reaction yield and the midpoints of the two curves are the ranges of variations. The point in this diagram plays the most pronounced role of reaction time on yield, and increasing the amount of catalyst from a certain amount does not have a significant effect on increasing yield (optimum value) and the lines go almost parallel to the catalyst axis. Figure 5 has the advantage of achieving a predetermined yield with different amounts of catalysts and reaction times, which can be selected according to the time and material constraints [34]. Therefore, according to the specific value of yield, the synthesis conditions can be obtained by drawing horizontal and vertical lines. The method helps to estimate the value of the yield reaction based on the desired values of the variables.

Figure 6 illustrates the response surface and contour plot for the influence of catalyst and reaction time on the yield of calix[4]resorcinarene synthesis. This figure shows more effective role of reaction time on synthesis yield.

Fig. 6

Response surface of three-dimensional plot indicating the effect of interaction between catalyst and reaction time on yield.

To obtain the optimum synthesis conditions, for example, with minimum amount of catalyst and reaction time, the contour plot was plotted in the range of 24–95 h and 0% to 4.45% catalysts (Fig. 7). By drawing the tangent line to the maximum yield range curve and the perpendicular line to this line at the lowest point, the optimal point is obtained.

Fig. 7

Contour plots for determine optimum condition.

As can be seen, it takes 83 h to achieve a yield above 95% with a minimum catalyst value (2.29%) and for the shortest time (about 65 h), approximately 3.85% of the catalyst is required. The amount of catalyst can help to progress the reaction but in a limited way and more amount of catalyst is without a positive result. Therefore, too much catalyst will not be more effective.

4 Conclusion

In this research, for increasing the yield of calix[4]resorcinarene synthesis, copper chromite nano catalyst was used. In order to determine the effect of synthesis parameters of calix[4]resorcinarene, the experimental design by Expert Design software with the variables of catalyst amount and reaction time was chosen. After confirming the selected model, the interactions of the parameters affecting the reaction yield in the contour plots were determined. Results showed the time has a greater effect on the yield and to achieve a yield around 95%, about 65h and 3.85% catalyst are required. Also, using the contour plots, the dependence of yield on variables on synthesis can be obtained.

Footnotes

Acknowledgments

We are sincerely grateful to the Ahvaz Branch, Islamic Azad University, for the financial support of this project.

Competing interests

The authors declare that they have no competing interests.

References

Martí-Centelles

, Pandey

M.D.

, Burguete

M.I.

and Luis

S.V.

, Macrocyclization Reactions: The Importance of Conformational, Configurational, and Template-Induced Preorganization, Chem Rev 115 (2015), 8736–8834.

Conzatti

, Brunengo

, Utzeri

, Castellano

, Hodg

and Stagnaro

, Macrocyclic Oligomers as Compatibilizing Agent for Hemp Fibres/Biodegradable Polyester Eco-Composites, Polymer 146 (2018), 396–406.

Anjali

and Chawla

H.M.

, Synthesis and Applications of Chromogenic Calix[4]Resorcinarene Derivatives, Res J Chem Environ 18 (2014), 45–49.

Mouradzadegun

, Abadast

and Elahi

, Tetraphenolate c-Methylcalix[4]Resorcinarene as a New Heterogeneous Phase Transfer Catalyst for Ring-Opening of Triaryl Substituted Pyrylium Salts in Aqueous Biphasic Medium, Monatsh Chem 145 (2014), 1663–1667.

Gutsche

C.D.

and Lin

L.G.

, Calixarenes 12: The Synthesis of Functionalized Calixarenes, Tetrahedron 42 (1986), 1633–1640.

Cram

D.J.

and Cram

J.M.

, Container Molecules and Their Guests, Monograph in Supramolecular Chemistry, Editor J. F. Stoddart, The Royal Society of Chemistry, Cambridge 4 (1994), 2771–2772.

Kamboj

and Singh

D.P.

, Macrocyclic Complexes: Potent Molecules against Drug Resistant Microbes, Int J Curr Res Chem Pharm Sci 3 (2016), 15–22.

Gawhale

S.T.

, Rathod

N.V.

, Kalyani

V.S.

, Singh

P.K.

, Sapkal

R.S.

, Sapkal

V.S.

, Chaudhari

G.N.

and Malkhede

D.D.

, Investigation of Host-guest Complexation of the Ternary Model: p-Sulfonatothiacalix[4]arene-fluorescein-M^2 +, Polyhedron 177 (2020), 114268.

Deska

, Dondela

and Sliwa

, Selected Applications of Calixarene Derivatives, Arkivoc 6 (2015), 393–416.

10.

Velásquez-Silva

B.A.

, Castillo-Aguirre

, Rivera-Monroy

Z.J.

and Maldonado

, Aminomethylated calix[4]resorcinarenes as Modifying Agents for Glycidyl Methacrylate (GMA) Rigid Copolymers Surface, Polymers 11 (2019), 1147–1160.

11.

Eddaif

, Shaban

, Telegdi

and Szendro

, A Piezogravimetric Sensor Platform for Sensitive Detection of Lead (II) Ions in Water Based on calix[4]resorcinarene Macrocycles: Synthesis, Characterization and Detection, Arab J Chem 13 (2020), 4448–4461.

12.

Sarkar

and Srinives

, Single-Walled Carbon Nanotubes-Calixarene Hybrid for Sub-ppm Detection of NO₂, Microelectron Eng 197 (2018), 28–32.

13.

Talotta

, Gaeta

and Neri

, Stereoprogrammed Direct Synthesis of Calixarene-Based [3]Rotaxanes, Org Lett 14 (2012), 3104–3107.

14.

García-Rio

and Basílio

, Supramolecular Surfactants Derived from Calixarenes, Curr Opin Colloid Interface Sci 44 (2019), 225–237.

15.

Wang

, Ding

and Guo

, Assembly Behaviors of Calixarene-Based Amphiphile and Supra-Amphiphile and the Applications in Drug Delivery and Protein Recognition, Adv Colloid Interface Sci 269 (2019), 187–202.

16.

Fuller

R.O.

, Koutsantonis

G.A.

and Ogden

M.I.

, Magnetic Properties of Calixarene-Supported Metal Coordination Clusters, Coord Chem Rev 402 (2020), 213066.

17.

Maniecki

T.P.

, Mierczynski

and Jozwiak

W.K.

, Copper-Supported Catalysts in Methanol Synthesis and Water Gas Shift Reaction, Kinet Catal 51 (2010), 843–848.

18.

Perumal

, Nagendra

and Ramakrishna

P.A.

, Effect of Ageing on Combustion of Ammonium Perchlorate with Copper Chromite Additive, Combust Flame 212 (2020), 352–354.

19.

Santacesaria

, Carotenuto

, Tesser

and Di Serio

, Ethanol Dehydrogenation to Ethyl Acetate by Using Copper and Copper Chromite Catalysts, Chem Engin J 179 (2012), 209–220.

20.

, Xiao

, Van Bokhoven

J.A.

and Liang

, A Non-Alkoxide Sol–Gel Route to Highly Active and Selective Cu-Cr Catalysts for Glycerol Conversion, J Mater Chem 20 (2010), 755–760.

21.

Prasad

and Singh

, A Review on CO Oxidation over Copper Chromite Catalyst, Catal Rev Sci Eng 54 (2012), 224–279.

22.

Pouresmaeily

S.M.

, Shirkavand Hadavand

, Shahidzadeh

and Ebrahimi

, N-Alkylation of Aniline by Copper-Chromite Catalyzer by Auto-Transfer Hydrogen Methodology, Chem Express 10 (2017), 1–7.

23.

Pouresmaeily

S.M.

, Shahidzadeh

, Shirkavand Hadavand

and Ebrahimi

, Synthesis of Cu-Chromite Catalyzer by Citrate Sol-Gel, Open J Synth Theory Appl 6 (2017), 13–22.

24.

Barekat

, Shirkavand Hadavand

, Rayatzadeh

and Badri

, Silane-Functionalized Organic Macrocyclic Compounds-Modified UVCurable Polyurethane Coatings: Synthesisand Properties, J Appl Chem Res 15 (2021), 56–66.

25.

Shirkavand Hadavand

, Saeb

M.R.

, Najafi

and Malekian

, Response Surface Analysis for Understanding the Effects of Synthesis Parameters on the Microstructure of Hyperbranched Polyesters, J Macrom Sci Part A Pure Appl Chem 53 (2016), 741–749.

26.

Shohani

, Pourmahdian

and Shirkavand

, Hadavand Response Surface Methodology for Design of Porous Hollow Sphere Thermal Insulator, IOP Conf Ser Mater Sci Eng 269 (2017), 012073.

27.

Mohamadpour

, Pirdashti

, Rostami

A.A.

and Shahrokhi

, Response Surface Methodology for the Evaluation of Lysozyme Partitioning in Poly (Vinyl Pyrrolidone) and Potassium Phosphate Aqueous Two-Phase System, Iran J Chem Chem Eng 38 (2019), 197–208.

28.

Mohammadi Doust

, Rahimi

and Feyzi

, An Optimization Study by Response Surface Methodology (RSM) on Viscosity Reduction of Residue Fuel Oil Exposed Ultrasonic Waves and Solvent Injection, Iran J Chem Chem Eng 13 (2016), 3–19.

29.

Eddaif

, Shaban

, Telegdi

and Szendro

, A Piezogravimetric Sensor Platform for Sensitive Detection of Lead (II) Ions in Water Based on calix[4]Resorcinarene Macrocycles: Synthesis, Characterization and Detection, Arabian J Chem 13 (2020), 4448–4461.

30.

Galedari

, Mehdipour Ghazi

and Mirmasoomi

S.R.

, Photocatalytic Process for the Tetracycline Removal Under Visible Light: Presenting a Degradation Model and Optimization Using Response Surface Methodology (RSM), Chem Eng Res Des 145 (2019), 323–333.

31.

Singh

, Belova

, Singh

and Chandra Sharma

, Biodiesel Production Using a Novel Heterogeneous Catalyst, Magnesium Zirconate (Mg₂Zr₅O₁₂): Process Optimization Through Response Surface Methodology (RSM), Energy Convers Manage 174 (2018), 198–207.

32.

Behera

S.K.

, Meena

, Chakraborty

and Meikap

B.C.

, Application of Response Surface Methodology (RSM) for Optimization of Leaching Parameters for Ash Reduction from Low-Grade Coal, Int J Min Sci Technol 28 (2018), 621–629.

33.

Ghosh

and Das

, Optimization of Copper Adsorption by Soil of Polluted Wasteland using Response Surface Methodology, Indian Chem Eng 56 (2014), 29–42.

34.

Krishna

and Padma Sree

, Artificial Neural Network and Response Surface Methodology Approach for Modeling and Optimization of Chromium (VI) Adsorption from Waste Water using Ragi Husk Powder, Indian Chem Eng 55 (2013), 200–222.

Optimization in synthesize of organic macrocyclic compounds in presence of nano copper chromite catalyst

Abstract

Keywords

1 Introduction

2 Experimental

2.1 Reagents

2.2 calix[4]resorcinarene synthesis

2.3 Design of experiments

Table 1 The levels of variables in RSM Variable Symbol Low (–1) Level middle (0) High (+1) Catalyst (% wt.) A 0 2.5 5 Time (h) B 24 60 96

3 Results and discussion

3.1 FTIR spectroscopy of calix[4]resorcinarene

Footnotes

Acknowledgments

Competing interests

References

Table 1
The levels of variables in RSM

Variable Symbol Low (–1) Level middle (0) High (+1)

Catalyst (% wt.) A 0 2.5 5

Time (h) B 24 60 96