Authors Accepted Manuscript

Optimization of Transesterification of Palm-based

Methyl Palmitate and Triethanolamine Towards
Maximum Di-esteramine Content

Haliza Abdul Aziz, Mohamed Kheireddine Aroua,

Rozita Yusoff, Noor Azeerah Abas, Zainab Idris
www.elsevier.com/locate/bab

PII: S1878-8181(17)30104-4
DOI: http://dx.doi.org/10.1016/j.bcab.2017.04.014
Reference: BCAB550
To appear in: Biocatalysis and Agricultural Biotechnology
Received date: 16 February 2017
Revised date: 21 April 2017
Accepted date: 25 April 2017
Cite this article as: Haliza Abdul Aziz, Mohamed Kheireddine Aroua, Rozita
Yusoff, Noor Azeerah Abas and Zainab Idris, Optimization of Transesterification
of Palm-based Methyl Palmitate and Triethanolamine Towards Maximum Di-
esteramine Content, Biocatalysis and Agricultural Biotechnology,
http://dx.doi.org/10.1016/j.bcab.2017.04.014
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
Optimization of Transesterification of Palm-based Methyl Palmitate and

Triethanolamine Towards Maximum Di-esteramine Content

*
Haliza Abdul Aziza , Mohamed Kheireddine Arouab, Rozita Yusoffb, Noor Azeerah Abasa,

Zainab Idrisa
a
Advanced Oleochemical Technology Division, Malaysian Palm Oil Board,43000,Kajang,

Selangor, Malaysia
b
Department of Chemical Engineering, Faculty of Engineering, University of

Malaya,50603,Kuala Lumpur, Malaysia

*Corresponding author. Address: Advanced Oleochemical Technology Division,

Malaysian Palm Oil Board, 43000, Kajang, Selangor, Malaysia. Tel./Fax: +603 87694340,

E-mail address: liza_azi@mpob.gov.my

Abstract

This study involved optimization of transesterification of palm-based methyl palmitate

with triethanolamine towards maximum di-esteramine content for an ideal performance of

esterquats. This reaction was aided by CaAl, a hydrotalcite-like compound (HTLC) as the

heterogeneous catalyst and the reaction was conducted in a 500 ml reaction unit. Response

surface methodology (RSM) in combination with face-centered design was used to

optimize the operating parameters. Results show that temperature was the most significant

variable affecting the di-esteramine composition, followed by Ca-Al dosage, quadratic

effect of Ca-Al dosage, vacuum and interaction of vacuum and Ca-Al dosage. The

optimum reaction conditions obtained were at temperature of 169 oC, Ca-Al dosage of 0.74

% and vacuum of 278.6 mbar. The result from ICP-OES exhibited insignificant leaching of

1
Ca and Al active species into fatty esteramine mixtures and the reusability of Ca-Al

catalyst was successfully sustained for three subsequent cycles.

Keywords: esteramine, transesterification, methyl palmitate, heterogeneous catalyst,

1. Introduction

Nowadays, esterquats cationic surfactant is commonly derived from biological resources

such as vegetable oils and animal fats. Excellent environmental profile and the good

softening effect of the esterquats have made this surfactant a better replacement for the

conventional distearyl (tallow) dimethyl ammonium compounds as an active ingredient in

fabric softeners. Esterquats is an interesting class of quaternaries, which are characterized

by the presence of hydrophobic alkyl chains that link to the positively charged nitrogen

molecule via ester bonds. The presence of this ester link will provide easy hydrolysis into

fatty acids and short chain quats (Tyagi et al, 2006; Puchta et al, 1993). Generally,

esterquats are produced via a two-stage process in which, triethanolamine is first esterified

with fatty acid in the presence of homogeneous acid catalyst or transesterified with ester

using a base catalysts to produce the desired esteramine. Esteramine, which consists of

mixtures of mono, di and tri-esteramine is then quaternized with methylating agent to

introduce a positive charge onto the esterquats molecule. Optimization of the esterification

or transesterification process for maximum conversion into esteramine is the key step in

the preparation of esterquats. However, homogeneous catalysts employed in the process

are difficult to be separated from the product (Jiang et al, 2012). Besides, homogeneous

acid catalysts such as hypophosphorous acid, para-toluenesulfonic acid and sulfuric acid

are toxic and corrosive, while sodium methoxide used in transesterification process can be

a serious health hazard to humans upon contact. These homogeneous catalysts are also not

2
reusable. Due to these limitations, heterogeneous catalysts have been proposed as

alternative to the homogeneous catalysts for the production of esteramine.

Production of esteramine via fatty acid route using solid catalysts has been reported in

several literatures. Lipase-catalyzed esterification of oleic acid and triethanolamine for

esteramine production was reported, whereby response surface methodology (RSM) was

used to optimize the enzymatic process conditions. The optimum conditions were

established at enzyme amount of 4.77 %, reaction time of 24 h, temperature of 61.9 oC,

mole ratio of 1:1 (oleic acid to triethanolamine) and agitation speed of 480 rpm with the

highest conversion of 63.57 % obtained (Masoumi et al, 2013). However, the study did not

reported on the optimization of the operating parameters towards maximum di-estermine

content. Esterification between stearic acid and triethanolamine to produce esteramine

catalysed by 20% loaded zirconium sulfate supported on SBA-15 and aluminium on SBA-

15 mesoporous molecular sieves have been reported (Geng et al, 2010; Geng et al, 2011;

Jiang et al, 2012). The optimum operating conditions employing these catalysts were

reported at temperature of 190 oC, catalyst dosage of 0.25 %, mole ratio of 1.8:1 of stearic

acid to triethanolamine and 6 h reaction time. The solid acid catalyst of 20% loaded

zirconium sulfate supported on SBA-15 (with pore diameter of 6 nm) was reported to give

higher composition of mono and di-esteramine of 16.61% and 67.45% respectively as

opposed to the homogeneous catalyst H3PO3, with tri-esteramine composition of 15.94%.

Meanwhile, aluminium supported on SBA-15 catalyst also appeared to be selectively

produced higher di-esteramine content of 68.49% as compared to H3PO3, with mono and

tri-esteramine content of 16.30% and 15.21% respectively. Although several development

of selective catalysts to produce higher di-esteramine contents were reported, the

production of esteramine via direct esterification of fatty acids and triethanolamine in the

3
presence of solid acid catalyst will requires expensive corrosion-resistant equipment.

Furthermore, the preparation of the fatty acid substrates produces many impurities, which

lead to formation of dark-coloured product. As such, further downstream processes such as

bleaching and distillation are needed to improve the purity and colour of the products that

will increase the production cost [Narula, 1995].

Due to drawbacks of the direct esterification of fatty acids, alternative route of producing

esteramine from methyl ester has been explored. Production of palm-based fatty

esteramine from a novel heterogeneous transesterification process from palm-based methyl

palmitate has been revealed in which, Ca-Al catalyst was employed as the heterogeneous

catalyst (Haliza et al, 2016). The previous research work reported on the optimum

transesterification operating conditions towards achieving maximum conversion of methyl

palmitate of approximately 98%. Among the various HTLC investigated, Ca-Al catalyst

showed the best catalytic activity as it gave the highest conversion of methyl palmitate and

yielded mono, di and tri-esteramine compositions of 23.50%, 47% and 23.90%

respectively. Further effort to obtain the optimum operating conditions with maximum di-

esteramine content will be highlighted in this present study, which has not been previously

reported based on the scope of our knowledge. Esteramine derived from triethanolamine

will consists of a mixture of mono, di and tri-esteramine. Higher mono-esteramine content

would give better dispersibility and formulation stability but exhibits low softening effects,

whereas higher tri-esteramine content results in better softening but give poor formulation

stability (Mishra et al, 2007). Ideally, higher di-esteramine content is desirable as it

exhibits both excellent softening and good stability of fabric softeners formulation. As

such, optimization of the transesterification process towards obtaining maximum di-

esteramine content will be investigated in this present study.

4
Interaction among the process parameters as well as optimization of the operating

conditions towards maximum di-esteramine content using multivariate statistic technique

will be highlighted in this present study. Face-centered cube design was used to design the

experiment and RSM was used as the optimization tools. A mathematical model was

generated to demonstrate the relationship between the operating parameters and di-

esteramine content. Stability of Ca-Al catalyst in terms of its reusability as well leaching of

the calcium and aluminium active species into the fatty esteramine mixtures shall also be

discussed in this paper.

2. Materials and Methods

2.1 Materials

Palm-based methyl palmitate (purity, 98 %) was purchased from Carotino Sdn Bhd and

triethanolamine (purity, 99 %) was purchased from Fisher Scientific (M) Sdn Bhd. Ca-Al

solid catalyst was prepared in-house by the Malaysian Palm Oil Board.

2.2 Catalyst preparation and characterization

The Ca-Al catalyst system was prepared by co-precipitation technique at high

supersaturation in a single container of two diluted solutions A and B, in which solution A

contained Ca2+ (67 g) and Al3+ (150 g) dissolved in 1 L deionized water. Solution B acting

as precipitating reagents was prepared by dissolving Na2CO3 (42.4 g) and NaOH (16 g) in

0.5 L deionized water. The prepared solutions A and B were then mixed with high

supersaturation technique for 18 h at 80 oC with vigorous stirring. Finally, the precipitate

was filtered, washed with deoinized water up to pH 7 and dried at 120 oC for overnight.

The morphology of Ca-Al was examined by scanning electron microscopy (SEM) using

5
FEI Quanta 200 modal. Leaching of the calcium and aluminum metals of the catalyst into

the reaction medium was determined using the Inductively Coupled Plasma-Optical

Emission Spectrometry (ICP-OES) conducted using Perkin Elmer Emission Spectrometer

Model Plasma 1000.

2.3 Transesterification Reaction

The process was performed in a 500 ml reaction flask equipped with magnetic stirrer,

vacuum line and a heating jacket. The reaction was carried out under the optimum mole

ratio of 1.8:1 (methyl palmitate:triethanolamine) and duration of 2 h as reported in

previous study (Haliza et al, 2016). The reactants were mixed and heated to the desired

reaction temperature and vacuum. When the reaction was completed, Ca-Al solid catalyst

was separated via vacuum filtration and the product was then analyzed using gas

chromatography. For reusability study of Ca-Al, several reactions were performed under

the optimum operating conditions by recycling Ca-Al from one cycle to the subsequent

cycle without washing or drying the catalyst cake after its separation from the reaction

mixture. Leaching of Ca-Al catalyst active species into fatty esteramine was tested,

whereby the Ca and Al metals content in the product were measured after first, second,

third and forth cycles in the recycling process. For samples preparation, 1 g of fatty

esteramine was added into 10 ml nitric acid (50 % concentration) and the sample mixture

was heated for 10 minutes at 95 oC. 5 ml of concentrated nitric acid (HNO3) was then

added to the mixture and heated for another 30 minutes at 95 oC. After the mixture was

cooled down, 2 mL of deionized water and 3 mL of hydrogen peroxide (30 %

o
concentration) were added into the mixture and heated again for 2 h at 95 C.

Subsequently, 7 mL of hydrogen peroxide (30 % concentration) and 10 mL of concentrated

hydrochloric acid were added into the mixture and heated again for 15 minutes at 95 oC.

6
Finally, the sample mixture was cooled down, filtered and marked up to 100 mL with

deionized water.

2.4 Analytical method

Compositions of mono, di and tri-esteramine were determined via gas chromatography

analysis, performed using the capillary column ZB-5HT inferno (12 m x 530 m x 15 m).

The oven temperature was set initially at 100 C, held for 3 min, and then increased at

6C/min to 350 C. Helium was used as the carrier gas at a flow rate of 20 mL/min. Injector

temperature was set at 350 C with split ratio set at 10: 1. Flame Ionization Detector (FID)

was used and temperature was set at 360 C.

2.5 Design of experiments and optimization method

Design-Expert software (version 8) was employed in design the experimental works and

optimizing the operating conditions. The experimental design was a three-level face-

centered cube design involving 30 experiments. Face-centered design is usually used in

order to avoid the experimental regions over the process limit (Norhazimah et al, 2012).

Reaction temperature, vacuum, agitator speed and Ca-Al dosage were chosen as the

independent variables while the di-esteramine content was chosen as the response for the

optimization study. These independent variables were studied at three levels with 6

replications of central point and 8 axial points. The value of alpha () was set at 1 for the

face-centered design configuration.The actual and coded levels of each variables are shown

in Table 1. The selection of each variable was chosen based on various literatures reported

on the application of heterogeneous catalysis for transesterification reaction. The

maximum temperature studied was 170 oC as the boiling point of methyl palmitate

7
reported is at 180 oC under maximum vacuum of 10 mbar (Bonhorst et al, 1948). The

lower level of temperature was 130 oC as lower than that, the reaction rate was very slow.

Maximum catalyst dosage of 1% was chosen based on the results reported on the use of

solid acid catalyst to produce the esteramine (Jiang et al, 2012).

Table 1 Actual and coded levels of independent variables

Variation levels

Independent variable Unit (-1) (0) (+1)

Temperature (A) (oC) 130 150 170

Vacuum (B) (mbar) 500 255 10

Catalyst Dosage ( C ) (%) 0.1 0.55 1.0

Agitator Speed (D) (rpm) 150 200 250

3. Results and discussion

3.1 Catalyst characterization

A SEM technique was used to investigate the morphology of Ca-Al solid catalyst

employed in the reaction. The image shown in Fig. 1 clearly indicates that Ca-Al particles

exhibit hexagonal shape. However, needle shape forms are also observed. The morphology

of Ca-Al obtained from the SEM technique is similar to the solid catalyst reported in the

literatures (Savita et al, 2008). The specific surface area, pore size diameter and pore

volume of the Ca-Al were reported at 256 m2/g, 10 nm and 0.27 cm3/g respectively (Haliza

et al, 2016).

8
 Fig. 1. SEM image of Ca-Al

3.2 Data Analysis

Face centered design matrix represents in terms of coded value and the experimental

results of di-esteramine content are tabulated in Table 2.

Table 2 Face-centered design matrix

Run Type of Coded Value Response

Order Factor A (oC) B (mbar) C (%) D (rpm) (Di-esteramine Content, %)

5 Center 0 0 0 0 32.82

6 Center 0 0 0 0 38.52

11 Center 0 0 0 0 38.45

13 Center 0 0 0 0 32.52

20 Center 0 0 0 0 39.79

21 Center 0 0 0 0 42.31

7 Axial 0 0 1 0 30.40

9
 14 Axial 0 -1 0 0 27.40

16 Axial 1 0 0 0 43.26

17 Axial -1 0 0 0 4.78

22 Axial 0 0 -1 0 7.16

25 Axial 0 0 0 -1 35.55

26 Axial 0 0 0 -1 37.69

28 Axial 0 1 0 0 43.30

1 Factorial 1 -1 -1 1 14.85

2 Factorial 1 -1 1 -1 33.16

3 Factorial -1 -1 1 -1 4.67

4 Factorial -1 -1 -1 -1 3.00

8 Factorial -1 1 1 1 41.25

9 Factorial 1 -1 1 1 40.12

10 Factorial -1 1 -1 1 1.85

12 Factorial 1 -1 -1 -1 22.96

15 Factorial 1 1 -1 1 27.13

18 Factorial 1 1 -1 -1 36.68

19 Factorial -1 1 -1 -1 1.00

23 Factorial 1 1 1 -1 45.00

24 Factorial 1 1 1 1 45.62

27 Factorial -1 -1 1 1 5.55

29 Factorial -1 1 1 -1 40.10

30 Factorial -1 -1 -1 1 1.4

Note: A, temperature; B, vacuum; C, catalyst dosage; D, agitator speed

A quadratic polynomial model was obtained based on the experimental data and this model

was used to describe the mathematical relationship between the independent variables and

10
dependent response. The following equation (1) in the terms of coded factors was

generated to predict the di-esteramine content:

Di-esteramine content (%) = 35.34 + 12.57 A + 7.75 B + 9.33 C + 4.47 BC

13.92 C2 (1)

where A: temperature, B: vacuum, C: catalyst dosage, BC: interaction between vacuum and catalyst

dosage and C2 : quadratic effect of catalyst dosage.

Analysis of variance (ANOVA) was performed to investigate the significance and fitness

of the model as well as the effects of the individual independent variables and their

interaction on the response. From the ANOVA shown in Table 3, it can be observed that

the quadratic model is significant as the p-value (probability of error value) is less than

0.0001 to predict the di-esteramine content. The lack of fit p-value of 0.0551 (lack of fit p-

value > 0.05 is not significant) implies that the model is fitted to all data (not significant

lack of fit is good). In this study, temperature (A), vacuum (B) and Ca-Al dosage (C),

interaction of vacuum catalyst dosage (BC) and quadratic effect of catalyst dosage (C2)

are significant model terms with p-value less than 0.05 (p-value < 0.05 indicates that the

model terms are significant for 95 % confidence interval).

Statistically, the larger the magnitude of the F-value from the ANOVA indicated that the

corresponding model term is more significant. By referring to the individual model term

from ANOVA shown in Table 3, it appeared that temperature has the highest F-value of

45.84, implying that this process parameter has the most significance influence towards

maximum formation of di-esteramine. Ca-Al catalyst dosage observed to be the next most

significance process parameter with F-value of 25.26, which can promote the formation of

di-esteramine. Vacuum level of the system has the least influence towards promoting the

11
maximum formation of di-esteramine with F-value of 17.41. Agitator speed was found not

to have any significance impact on the formation di-esteramine and thus this process

parameter was excluded in order to improve the model and optimization results. The

ranking of the overall significance terms from the ANOVA are as follow; temperature >

Ca-Al catalyst dosage > quadratic effect of catalyst dosage > vacuum > interaction of

vacuum and catalyst dosage. The value of R-squared of 0.8287 indicates that the model fits

the experimental data, because according to reported works, value of R2 > 0.75 indicates

fitness of the model (Reddy et al, 2008; Haaland, 1989). Predicted R-squared of 0.7218 is

in agreement with the adjusted R-squared of 0.7931 as the adjusted and predicted R-

squared values are within 0.2 of each other. This confirms the precision of the

experimental data and model (Zabeti et al, 2009). Adequate precision measures the signal

to noise ratio and a value greater than 4 is desirable. Result shows that the adequate

precision value was 16.832, indicating that the model can be used to navigate the design

space.

Table 3 ANOVA results for the model and model terms

Source F-value p-value, Prob>F

Model 23.23 <0.0001 Significant

A-temperature 45.84 <0.0001

B-vacuum 17.41 0.0003

C-catalyst dosage 25.26 <0.0001

12
BC 5.16 0.0324

C2 22.48 <0.0001

Lack of fit 4.35 0.0551 not significant

R-squared = 0.8287, Adjusted R-squared = 0.7931, Predicted R-squared= 0.7218, Adequate

Precision = 16.832

Fig. 2 shows the plot of predicted versus actual di-esteramine content (%), which verified

that, the predicted response (di-esteramine content, %) is in agreement with the observed

data (actual) in the range of the operating conditions. The value R2 = 0.8287 indicated that

the model fits the experimental data.

R2= 0.8287

Fig. 2. Scatter plot of predicted di-esteramine content (%) against actual di-esteramine

content (%)

13
3.3 Correlation between empirical equation and the response surface

The interaction between vacuum and catalyst dosage and their effect on di-esteramine

content is illustrated in the response surface 3D plot as shown in Fig. 3(c), generated based

on equation (1) and by having the temperature and agitator speed at their center points. The

response surface indicates that at low vacuum, the di-esteramine content increase with an

increase in Ca-Al catalyst dosage. However, higher catalyst dosage leads to a slight

reduction in the di-esteramine content, due to the fact that the quadratic term of the catalyst

dosage is more significant with a negative effect (equation 1). Considerable increment in

di-esteramine content was observed with increasing Ca-Al catalyst dosage at higher

vacuum. This is due to the fact that the interaction term of vacuum catalyst dosage (BC)

is more significant with a positive effect on the response (equation 1).

3.4 Interaction effects of process parameters on di-esteramine content

In order to determine the optimum levels of each process variables for maximum di-

esteramine content, response surface 3D and contour plots were constructed by plotting the

response (di-esteramine composition) against any two independent variables, while

maintaining the third variable at their center points. Response surface 3D and contour plots

of the predicted di-esteramine content as a function of two independent variables; (a)

temperature and vacuum, (b) temperature and catalyst dosage (c) vacuum and catalyst

dosage are demonstrated in Fig. 3(a) (c).

3D surface and contour plots of di-esteramine content versus temperature and vacuum

shown in Fig. 3(a) at fixed catalyst dosage of 0.55 % and agitator speed of 200 rpm reveals

14
that generally, increase in both variables resulted in an increase in di-esteramine content.

However, the composition of the di-esteramine was more affected by temperature

compared to vacuum. Temperature gave a profound impact on transesterification, in which

the rate of reaction towards the formation of di-esteramine increases as the temperature is

increased. Furthermore, it appeared that the miscibility between palm-based methyl

palmitate and triethanolamine improved at higher temperature thus enhanced the contact

between these two reactants and Ca-Al solid catalyst. Higher di-esteramine content was

obtained at higher temperature from 150 to 170 oC and higher vacuum from 150 to 10

mbar.

Fig. 3(b) demonstrates the response surface 3D and contour plots of di-esteramine

composition versus temperature and Ca-Al dosage with vacuum fixed at 255 mbar and

agitator speed of 200 rpm. It was observed from the figure that minimal di-esteramine

content was obtained at low temperature and low amount of Ca-Al dosage. An increase in

temperature resulted in an increase in di-esteramine content and higher content of di-

esteramine content was observed at temperature in the range of 162 to 169 oC. It was also

appeared that the di-esteramine content increased with increase in Ca-Al catalyst dosage

from 0.5 to 0.8 %. When the Ca-Al dosage was increased beyond 0.8 %, reduction in the

composition of di-esteramine was experienced, presumely due to increase in viscosity of

the reaction mixture when higher amount of Ca-Al catalyst was incorporated. This might

reduce the external mass transfer rates between bulk phases of the reactants towards the

surface of the solid catalyst. This finding is in agreement with several published works

reporting similar trend on the effect of catalyst dosage (Wenlei et al, 2015). It was also

reported in the biodiesel production from Styrax confuses Hemsl seed oil, in which the

yield started to reduce when the magnetic solid acid catalyst concentration was more than 5

15
g/l. The higher amount of catalyst could results in agglomeration and settled out from

reaction mixture and thus, reducing the catalyst surface area (Yigang et al, 2015). Other

previous work described that the percentage of conversion for the enzymatic synthesis of

esteramine using lipase was increased by increasing the catalyst dosage from 3.0 to 5.8 %

and decreased with further increase to 7 %. Unlike heterogeneous transesterification,

higher reaction temperatures were reported to induce the inactivation of the enzyme used

in the study (Masoumi et al. 2013). However, this reported work did not mention on the

interaction effects of the process parameters towards di-esteramine formation.

The interaction of vacuum and Ca-Al catalyst dosage towards di-esteramine content is

illustrated in the response surface 3D and contour plot at fixed temperature 150 oC and

agitator speed of 200 rpm shown in Fig. 3(c). It was found that the composition of di-

esteramine was more affected by the Ca-Al dosage than vacuum. Increase in Ca-Al dosage

resulted in an increase in the composition until it reached 0.8 %, whereas, an increase in

vacuum did not significantly increase the content of di-esteramine. High vacuum of 10

mbar gave maximum conversion when the catalyst dosage was at 0.78 %, suggesting that

an efficient methanol removal system favored the formation of di-esteramine. Lower

formation of di-esteramine was observed at vacuum lower than 200 mbar when 0.78 % Ca-

Al dosage was incorporated, presumely due to insufficient energy to pull the methanol

away from the system. The excessive accumulation of methanol could lead to higher

decomposition of di-esteramine into mono-esteramine, hence resulting in lower percentage

of di-esteramine in the product mixture.

16
 (a)

(b)

(c)

17
Fig. 3. (a) Response surface 3D and contour plots of di-esteramine content versus
temperature and vacuum at catalyst dosage of 0.55 % and agitator speed of 200 rpm. (b)
Response surface 3D and contour plots of di-esteramine content versus temperature and
Ca-Al dosage at vacuum of 255 mbar and agitator speed of 200 rpm. (c) Response surface
3D and contour plots of di-esteramine content versus vacuum and Ca-Al dosage at
temperature of 150 oC and agitator speed of 200 rpm.

3.5 Optimization of operating conditions and validation of model

For optimization, the transesterification reaction was performed under the recommended

conditions and the experimental results were compared with predicted values. The RSM

software suggested that the optimized condition for obtaining maximum di-esteramine

content were temperature of 169 oC, vacuum of 278.6 mbar, Ca-Al catalyst dosage of 0.74

% and agitator speed of 200 rpm (Table 4). The di-esteramine composition under the

optimum reaction conditions was predicted to be 47.9 %. After a triple run of the

experiments, it was found that the actual di-esteramine content obtained under the

optimum operating conditions suggested was 45.5 %, which was 95 % in agreement with

the predicted value. Therefore, this implies that the generated quadratic model derived

from RSM experimental design has sufficient accuracy to predict the di-esteramine

content.

Table 4 Optimum operating conditions and di-esteramine content using RSM

Temperature Vacuum Catalyst Agitator Di-esteramine Content (%)

(oC) (mbar) Dosage (%) Speed
(rpm) Experimental Predicted

169 278.6 0.74 200 45.5* 47.9

* Experimental data are mean values from three replicates measurement (SD 0.3)

18
Previous study reported the optimum conditions established for maximum conversion of

methyl palmitate of approximately 98 % was at 170 oC, vacuum of 10 mbar, Ca-Al catalyst

dosage of 0.5 % and agitator speed of 150 rpm (Haliza et al, 2016). It appeared that there is

a similarity between the optimum operating conditions towards achieving maximum

conversion of methyl palmitate and maximum di-esteramine yield, implying that a good

correlation between these two factors was observed. Significant difference was only

noticeable on the vacuum level, whereby maximum di-esteramine content can be obtained

at lower vacuum of 278.6 mbar.

In this study, the reaction performed under the optimum condition yielded 20.9 % mono-

esteramine and 27 % tri-esteramine. Although the tri-esteramine content obtained from

RSM optimization was 27 %, which was slightly higher than the reported composition of

25 % for a better fabric softening formulation stability (Franklin et al, 2002). This might be

tackled by performing the reaction at mole ratio lower than 1.8:1 (methyl palmitate :

triethanolamine). The fatty esteramine compositions obtained in this present study

demonstrated similar fatty esteramine compositions with the conventional approach using

sodium methoxide of 19.6, 49 and 28.5 % of mono, di and tri-esteramine respectively

(Haliza et al, 2016).

3.6 Reusability of catalyst

In conventional process, sodium methoxide was commonly used as the catalyst for

transesterification of methyl ester and triethanolamine. However, the catalyst was difficult

to be separated from the product and could not be reused. Reusability is an important

property for heterogeneous catalyst as it can reduce the cost of the catalyst and the

19
operation of regenerating the catalysts. Therefore, heterogeneous catalyst that has a

potential for industrial application should have a good stability property. Reusability

experiments were carried in this present study to test the stability of the Ca-Al catalyst.

Based on Fig. 4, the reusability of CaAl was successfully tested for three subsequent

cycles, where the mono, di and tri- esteramine compositions obtained for the second and

third cycles were close to the first cycle using fresh Ca-Al catalyst. However, Ca-Al

catalyst suffered a dramatic change of fatty esteramine compositions in the fourth cycle,

whereby only 22 % di-esteramine was obtained compared to approximately 46 % formed

in the first, second and third cycles of the recycling process. It was noticeable that higher

composition of mono-esteramine of approximately 33 % obtained in the fourth cycle,

indicating that the catalytic activity of Ca-Al in converting the mono-esteramine into di-

esteramine was reduced.

50

40
Fatty Esteramine Composition, %

30

20

10

0
1 2 3 4

Number of cycle

Mono-esteramine Di-esteramine Tri-esteramine

Fig. 4. Reusability of CaAl. (Optimum reaction conditions: 169 oC, 278.6

mbar, 200 rpm and 0.74 % Ca-Al dosage)

20
The reduction of the catalytic activity probably caused by the mechanical stress of the

catalyst particles that collide against each other and also against the walls of the reaction

unit caused by continuous agitation from one cycle to another in the reusability study.

Moreover, the mass loss of Ca-Al catalyst in the recycling process might also lead to

decrease in fatty acid composition in the forth cycle. The reusability of Ca-Al catalyst

performed under the optimum operating conditions established in this study were

compared with reported reusability study of other solid catalysts in transesterification

processes. Ca-Al catalyst was used in methanolysis and ethanolysis of soybean oil for

biodiesel production, however, the reusability of this catalyst for the process was not

reported (Simonetti et al, 2014). Similarly, research work on the employing Ca-Al mixed

oxide for biodiesel production from transesterification of rapeseed oil did not mention on

the stability in terms of reusability of the catalyst (Kock et al, 2015). Reusability of other

catalyst such as CaO supported on Al2O3 employed in transesterification of palm oil for

biodiesel production was reported whereby, the catalyst was reused for only two

subsequent cycles at optimum condition of 65 C, catalyst content of 6 wt.% and

alcohol/oil molar ratio of 12:1 (Zabeti et al, 2010). Meanwhile, study on the reusability of

CaO catalyst employed in the production of glycerol carbonate via transesterification of

glycerol and dimethyl carbonate was reported, whereby this catalyst showed a weak

catalytic activity in the second cycle of the reaction conducted at 95 oC, catalyst/glycerol

molar ratio = 0.06 and dimethyl carbonate/glycerol molar ratio of 3.5 (Jose et al, 2009).

The findings from this study also revealed that the rapid decrease in both conversion and

yield presumably due to deactivation of CaO caused by the exposure with air between the

catalytic runs and reduction of its surface area as the catalyst was lump together when it

was reused in the subsequent cycles.

21
3.7 Leaching of Ca-Al catalyst actives species into reaction system

Leaching of active sites into the reaction medium is one of the causes for deactivation of

the heterogeneous catalyst in chemical processes. The deactivation mechanism affects the

industrial application as extensive leaching may threaten the reusability of the

heterogeneous catalyst. The degree of leaching directly influences the number of cycles the

catalyst can be reused. Insignificant amount of Ca metal of 32 mg/L was detected in the

fatty esteramine sample from the first cycle as illustrated in Fig.5. It is noticeable that only

minor amount of Ca active species were leached into the product from second, third and

forth cycles of approximately 5, 6 and 6.2 mg/L respectively and thus they were found to

be negligible. Similarly, the leaching of Al active species was also found to be minimal.

The trace amount Al metal of 5 mg/L detected in the fatty esteramine samples from the

first, second, third and forth cycles are depicted in Fig.5. Overall, it was observed that

leaching of Ca and Al active species into fatty esteramine samples were insignificant and

these findings were in agreement with the reusability trend of this Ca-Al catalyst, which

capable of maintaining its activity for three subsequent cycles. However, it appeared that

leaching of Ca and Al metals was not the main reason for the deactivation of the catalyst

causing the decrease in fatty esteramine compositions from the forth cycles as insignificant

amount of the actives species were detected in fatty esteramine sample at that cycle. This

findings strongly supports that the weak catalytic activity of Ca-Al in the forth cycle

presumably due to mechanical stress of the catalyst and weight loss upon recycling

processes.

22
 Fig. 5. Leaching test profile for Ca and Al metals in fatty esteramine

The leaching profile of Ca-Al catalyst obtained in this present study was compared with

findings reported in heterogeneous transesterification processes. Leaching of Ca active

species from the Ca-Al mixed oxide used in transesterification of rapeseed oil for biodiesel

production was reported (Kock et al, 2015), whereby the leaching was started to increase

rapidly when the conversion was high. High conversion of 92 % was achieved within 90

minutes reaction time and almost no Ca metal (~ 5 mg/L) was detected in the reaction

mixtures within this period. When the transesterification was prolonged to more than 90

minutes, the leaching of the Ca metal increased rapidly up to 75 mg/L as more glycerol

was formed. Investigation on the leaching of Ca metal from Ca-Al mixed oxides catalyst in

each reaction components of biodiesel such as methanol, rapeseed oil, methyl ester and

glycerol also was reported in the study. They found that the maximum content of Ca active

species of approximately 70.8 mg/L was detected in glycerol. The study also mentioned

23
that calcination conditions (e.g. nitrogen or air flow, calcination temperature) of the Ca-Al

mixed oxide catalyst also affected the degree of leaching of the Ca actives species. Another

study has evaluated the stability of CaO supported on alumina catalyst in transesterification

of palm oil for biodiesel production and reported that the leaching of Ca active species into

biodiesel phase was insignificant (Zabeti et al, 2010). The amount of Ca metal in the

biodiesel samples after the first and second cycles were measured at 31 mg/L and 12 mg/L

respectively. Overall, the amount of Ca active species leached into fatty esteramine

detected in this present study were observed to be lower than that of the reported literatures

and the leaching of Al active species into fatty esteramine was also found to be minimal.

4. Conclusion

The optimum heterogeneous transesterification operating conditions for achieving

maximum di-esteramine content of approximately 46 % established via RSM optimization

were at 169 oC, 278.6 mbar, 0.74 % Ca-Al dosage and agitator speed of 200 rpm. The

statistical analysis from ANOVA revealed that temperature has the most profound

influence in promoting the formation of maximum di-esteramine content followed by Ca-

Al catalyst composition, whereas vacuum has the least impact on the di-esteramine

formation. The experimental di-esteramine composition was obtained at 45.5%, which was

in good agreement with the predicted composition of 47.9%. Ca-Al catalyst can be re-used

and the fatty esteramine compositions were sustained for three subsequent cycles.

Leaching test on Ca-Al catalyst exhibited insignificant amount of calcium and aluminium

active species detected in the fatty esteramine samples throughout the recycling process.

24
Acknowledgement

The author acknowledged the financial support by the Malaysian Palm Oil Board. This

work was carried in collaboration with the Chemical Engineering Department of

University of Malaya.

References

Bonhorst CW, Althouse PM, Triebold HO. 1948. Esters of naturally occurring fatty acids.

Ind Eng Chem. 40: 2379 84.

Cavani F. Trifirb, Vaccari A. 1991. Hydrotalcite-type anionic clays preparation, properties

and applications. Catal Today 11: 173-301.

Franklin R, Mendello R, Albert P, Steichen D, Trinh P-N. 2002. High di (alkyl fatty ester)

amines and quaternary ammonium compounds derived therefrom. Patent No. US

200200 25915A1

Geng T, Li Q, Jiang Y, Wang W. 2010. Synthesis of stearic acid triethanolamine ester over

solid acid catalysts. Chinese Chem Lett 21:1020-1024.

Geng T, Li Q, Jiang Y, Wang W. 2011. Esterification of stearic acid with triethanolamine

over zirconium sulfate supported on SBA-15 mesoporous molecular sieve. J. Surfact

Deterg 14:1522

Haliza AA , Aroua MK, Yusoff R, Abas NA, Idris Z, Hazimah AH. 2016. Production of

Palm-Based Esteramine Through Heterogeneous Catalysis. J Surfact Deterg 19:11

18.

Haaland PD. 1989. Experimental design in biotechnology. New York: Marcel Dekker Inc.

Jiang Y, Geng T, Li Q. 2012. Synthesis of stearic acid triethanolamine ester over Al-SBA-

15,. J Porous Mater 19:369 -374.

25
Jose RO, Olga G., Belen M., Amaia P, Camilo R, Leire L, Jesus T, Mara CV. 2009.

Synthesis of glycerol carbonate from glycerol and dimethyl carbonate by

transesterification: Catalyst screening and reaction optimization. Appl Cataly A Gen

366: 315324

Kock, J., Hjek M., & Troppov I. 2015. The factors influencing stability of CaAl mixed

oxides as a possible catalyst for biodiesel production. Fuel Processing Technology

134: 297302

Lijing G, Guangyuan T, Guomin X, Ruiping W. 2010. Biodiesel from palm oil via loading

KF/Ca-Al hydrotalcite catalyst. Biomass Bioenerg 34:1283-1288

Masoumi HRF, Kassim A., Basri M, Abdullah DK, Abdollahi Y, Ghani SS, Rezaee M.

2013. Statistical optimization of process parameters for lipase-catalyzed synthesis of

triethanolamine-based esterquats using response surface methodology in 2-liter

bioreactor. The Sci World J.2013:1-9

Mishra S, Tyagi VK. 2007. Esterquats: The Novel Class of Cationic Fabric Softeners.

J Oleo Sci 56:269276

Narula O.P (1995) Treatise on Fats, Fatty Acids & Oleochemicals; Industrial Consultants:
New Delhi, India 1.

Norhazimah AH, Faizal CKM. 2012 Optimization of fermentation parameters for the

production of bio-ethanol from oil palm trunks sap using face-centered (CCD).

Australian and New Zealand Chemical Engineers Conference Proceeding

(Chemeca).

Puchta R, Krings P, Sandkuehler P. 1993. A new generation of softeners. Tenside, Surfact

Deterg 30:186-191.

26
Reddy LVA., Wee YJ, Yun JS, Ryu HW. 2008. Optimization of alkaline protease

production by batch culture of Bacilus sp. RKY3 through Plackett-Burman and

response surface methodological approaches. Bioresour. Technol. 99: 2242-2249.

Savita GD, Agarwal D, Susanta B. 2008. Synthesis and characterization of hydrotalcites:

Potential thermal stabilizer for PVC. Indian J Chem 47A:1004-1008.

Simonetti EAN, Thim GP, Cortez GG. 2014. Ca-Al hydrotalcites as catalysts for

methanolysis and ethanolysis of soybean oil. Mod Res Catal 3: 117 127.

Tyagi R, Tyagi VK, Khanna RK. 2006. Synthesis, characterization and performance of

tallow fatty acids and triethanolamine based esterquats. J Oleo Sci; 55: 337- 345.

Wenlei X, Xinli Y, Xuezhen Z. 2015. Interesterification of soybean oil methyl stearate

catalyzed by guanidine-functionalized SBA 15 silica. J Am Oil Chem Soc 92: 915 -

925.

Yigang W, Xiaoan N, Zhenxing L. 2015. Biodiesel Synthesis from Styrax confuses Hemsl

Catalyzed by S2O82- / ZrO2 TiO2 Fe3O4. J Am Oil Chem Soc 92: 813 820.

Zabeti M, Daud W.M.A.W, Aroua M.K. 2009. Optimization of the activity of CaO/Al2O3

catalyst for biodiesel production using response surface methodology. Appl Cataly A

Gen 366:154159

Zabeti M, Daud W.M.A.W, Aroua M.K. 2010. Biodiesel production using alumina-

supported calcium oxide: An optimization study. Fuel Process Tecnol; 91:243248

27