Homogenizer-intensified  room temperature biodiesel production using heterogeneous palm bunch ash catalyst

Waste bio-based  materials  generated  as by-products  of palm  oil  mills  have  been successfully  used as a hetero- geneous  catalyst  in  homogenizer-intensified biodiesel  production. Palm  bunch  ash  (PBA)  contains  potassium oxide as a major component that could catalyze palm oil to biodiesel at room temperature.  The influential transesterification  conditions  such  as ratio  molar  palm  oil  to methanol, rotational  speed,  catalyst  weight  and reaction  time  were  investigated   on  biodiesel  conversion.   The  highest  conversion   of  98.9%  was  obtained  in presence of a ratio molar of 1:15,  rotational  speed of 4000 rpm, catalyst  weight  of 18 wt.% and 10 min reaction times.  The catalytic  stability  test revealed  that both catalyst  amount  and K2O  concentration  decreased after one reaction cycle.  However,  no calcination is required and the availability of PBA is abundantly  in palm oil producer countries  increasing  its competitiveness  to use as a heterogeneous  catalyst.  In addition, this process could  save 6–98% of electricity  consumption  and 67–87%  of reaction  time compared  to other methods.

1.  Introduction

 Commercial biodiesel production commonly uses homogeneous base substances i.e.  sodium  or potassium hydroxide  as a catalyst  which  is performed  in  a  methanol  boiling  point  (Mahlia  et  al.,   2020).  Even though  a  high  biodiesel  yield  could  achieve  in  a  short reaction  time using those catalysts,  some drawbacks particularly  related to product purification  and highly  base waste disposal could increase the produc- tion cost (Atadashi,  2015; Sitepu et al.,  2020). Heterogeneous catalyst offers a simple separation from the product and the possibility to reuse (Chua  et al.,  2020; Mohod et al.,  2020). However,  the reaction rate is low  due  to  mass transfer limitation  as  the  reactants  are  in  different phases (Mansir et al., 2017; Tarigan et al., 2021). Therefore a high concentration of catalyst operating at a high temperature is required to increase the mass transfer (Abdullah  et al.,  2017).  Recently  heteroge- neous catalysts  particularly  those containing  potassium ions reported having  a high catalytic  activity  which is similar to homogeneous cata- lysts (Alagumalai  et  al.,  2021).  Vadery  et al.  (2014) demonstrated a biodiesel conversion of 97% achieved in 30 min at a reaction tempera- ture of 45◦C  using calcined  coconut  husk.  Generally,  those potassium catalysts are derived from renewable waste biomass such as mango peel, Tucuma peel, Brassica nigra and more (Laskar et al., 2020b; Mendonça et al., 2019; Nath et al., 2019). Calcination with temperature varied in the range of 300 - 900◦C is required to activate the catalyst (Miladinovi´c et al., 2020) which therefore could raise the biodiesel production cost in total.

Palm  bunch ash is a by-product produced from the combustion of palm  empty  fruit  bunch,   has  been  identified  contains  a  significant amount of potassium ion (Shan et al., 2018; Yaakob et al., 2012) and the stock is abundant particularly in Indonesia as palm oil largest producer in the world (Indonesia, 2021). Recently some researchers have utilized PBA  as  a  catalyst  in  biodiesel  production,  adsorbent for  wastewater treatment and air purifier and concrete industry (Boey et al.,  2011; Ho et al.,  2014; Tangchirapat  et al.,  2007; Yaakob  et al.,  2012; Zainudin et al.,  2005). Boey et al. (2011) used PBA which have been calcined in three different temperatures as a catalyst in transesterification of palm olein and obtained 90% biodiesel conversion after 30 min reaction time. Therefore it concluded that calcination  is solely helpful just for catalyst reuse. The non-calcined PBA at a concentration of 3 wt% produces 90% of  biodiesel  conversion  (Boey  et  al.,   2011).  However,   the  reaction remains to perform in reflux conditions.

Homogenizer  is controlled stirring equipment that has a stator and rotor. A gap that varies from 100 to 3000 μm between them can cause large turbulence (Vikash et al., 2021). The stirring strength produced by this homogeneous dispersing device is known to be 3 times greater than that of an ordinary mixer at the same speed (Vikash et al.,  2021). Thus, very high intermolecular interactions occur in this homogeneous dispersing device causing reactions to occur quickly (Hsiao et al.,  2018; Vikash  et al.,  2021).  The biodiesel conversion of 99.1% was obtained using a homogenizer  in just 1 min using NaOH  as a catalyst  at room temperature  conditions  (Sa´nchez-Cantú  et  al.,   2017).  Similar  results were obtained by other researchers using soybean oil with a biodiesel conversion of 97.1% was obtained  within  40 s using a NaOH  catalyst (Sanchez-Cantú et al., 2019). To date, no literature found regarding homogenizer intensified biodiesel production using waste bio-based heterogeneous catalysts.

This study reports the outcomes of homogenizer-intensified the transesterification of palm oil using PBA as catalyst operates in batch mode at room temperature.  The sensitivity of biodiesel conversion to processing  parameters such  as  ratio  molar  of  palm  oil  to  methanol, rotational speed (rpm), catalyst weight (wt.%), and reaction time (min) was systematically determined in triplicate to define the maximum conversion.  The leaching  of PBA is conducted to investigate the possi- bility of recyclability of the catalyst. Furthermore, the Fourier Transform Infrared (FT-IR), X-ray diffraction  (XRD),  X-ray fluorescence and scan- ning electron microscope – energy dispersive spectroscopy (SEM-EDS) are used to characterize the PBA catalyst.

Fig. 1.  The schematic  image  of the experimental  set-up.

layer was collected  and stored in a desiccator for GC  analysis.  All  the experiments were repeated three times and the statistical error bars are included in the corresponding figures.

2.4.   Leaching and reusability test of PBA catalyst

A  leaching  test was performed to explain  the decreasing  catalytic activity due to losing active species during the reaction. The amount of potassium as the  main  elements contained  in  the  catalyst  was deter- mined using the EDS analysis before and after high mixing intensity was conducted.  The protocol for this study is only involving  methanol and PBA.  A 4 gram of PBA and 80 ml methanol were fed to the cavitation reactor and was homogenized at 4000 rpm for 30 min. The catalyst was filtered and dried in an oven at 40◦C overnight until constant weight and stored in a desiccator  for elemental  analysis.  The divergence  catalyst weight before and after dispersing is calculated as leaching percentage. The reusability study was conducted using reaction conditions of ratio molar palm oil to methanol of 1:15, catalyst weight of 7 wt.%, a reaction time of 30 min and rotational speed of 4000 rpm. After the reaction was completed, the catalyst was separated from the solution phase and directly used with fresh palm oil and methanol for the subsequent cycle while the biodiesel layer was collected for GC  analysis.

2.5.   Biodiesel conversion analysis

After  homogenizer-intensified  transesterification of  palm  oil  using PBA as a catalyst,  1μL of biodiesel was injected into GC  for conversion determination  with methyl  heptadecanoate  as internal standard following  previous protocol (Ma’arof et al.,  2021). The gas chromato- graph spectrometer (Shimadzu type 2010) is equipped with a capillary column  (length 15 m x 0.25  mm ID) and a flame  ionization  detector. Helium was used as carrier gas with a constant flow of 1 ml/min while the column oven was set at 90◦C initially  and rising with a rate of 7◦C min. The injection port and flame ionization  detector temperature was set to 260◦C.

2.6.   Statistical analysis

The statistical differences of parameters were validated  by analysis of variance (ANOVA)  followed by Turkey Post-hoc pairwise comparison at P < 0.05 using Statistica v13.

3.  Results and discussion

3.1.   Catalyst characterization

The  FT-IR  spectra of  the PBA  catalyst  exposed several absorption peaks as shown in Fig. 2A. A band at 2981.9 cm  1 can be attributed to stretching vibration of the C–H band from the organic compound which presumably resides in the PBA catalyst.  The low transmittance band at 2117.1 cm  1 denoted the presence of M – O – K bands (M = Si, Mg, etc.) (Yaakob  et  al.,  2012).  Further,  a  minor  and  sharp band  occurred  at 1654.9 and 1379.1 cm  1 representing the carbonates due to adsorption of CO2  onto mineral oxides (Laskar et al., 2020b). The peaks observed at 969.1  and  864.7  cm-1  indicate  the  presence of  the  Si  – O  – Si  band (Yaakob  et al.,  2012).  The elemental  composition of the PBA catalyst was determined using XRF analysis (Fig. 2B) which showed the presence of Si,  P, S,  K, Ti,  Mn,  Fe, Cu,  Zn,  Br, Rb, Mo,  Ba, Eu and Re elements. Potassium, K, was the major element observed in PBA which constituted 81.1% followed  by Si detected by 7.28% and Mo,  Fe, Rb and P which consists of 3.3,  2.69,  2.02 and 1.6%, respectively. Other elements were found to be less than 1%. Further XRD analysis of PBA catalyst (Fig. 2C) exhibits peaks at 2θ of 28.62◦;  30.00◦;  31.07◦,  31.83◦,  32.48◦;  40.74◦; 43.73◦; 45.25◦  and  50.40◦  which  represent hexagonal  crystal  of  K2O (JCPDS file no.  00–026–1327).  The peaks observed at 32.45◦;  37.55◦; 59.1◦   and  67.8◦   correspond to  the  presence of  CaO  (JCPDS file  no. 00–001–1160) while  MgO  was detected appeared in peaks of 37.07◦; 43.74◦  and 73.85◦  (JCPDS file no. 01–087–0653). The quantity of K2O was examined  using  EDS  analysis  which  revealed  a  concentration  of 35.9%. This result is higher than the K2O content of 19.8% reported by the previous researcher (Boey et al., 2011, 2012). The SEM image of PBA (Fig. 2D) was found to be agglomerated due to high temperature during combustion  of  palm  fiber  and  shell  (Attari  et  al.,   2022)  rendering reducing its catalytic activity (Changmai  et al.,  2020).

3.2.   Effect of ratio molar oil to methanol

Theoretically,  3 mol of methanol and 1 mole of triglycerides are reacted to produce 3 mol of fatty acid methyl ester. However, an addi- tional  mole  of  methanol  is  required  to  drive  the  reaction  towards products (Laskar et al.,  2020b; Sitepu  et al.,  2020).  Mostly  a ratio of molar  oil  to  methanol  of  1:6  is  used  to  ensure a  complete  reaction (Sitepu et al.,  2020). However, based on the published results, an opti- mum  biodiesel  conversion  was achieved  using a ratio molar  of  more than 1:6 (Kasirajan,  2021; Tan et al.,  2019).  Therefore this parameter was optimized by conducting the homogenizer-intensified trans- esterification reaction with various oil to methanol ratios of 1:9; 1:12; 1:15; 1:18 and 1:21 under reaction condition of catalyst weight of 7 wt. %, the reaction time of 30 min and rotational  speed of 4000 rpm. As shown  in  Fig.   3A,   the  biodiesel   conversion   was  increased  in  the increasing ratio molar to reach the maximum conversion of 92.6 ± 0.6% at a ratio of 1:15. Further, increasing methanol volume deteriorates biodiesel conversion. This presumably due to higher methanol content in the reaction solution could  trigger the reversible transesterification reaction  to  produce  monoglycerides  and  diglycerides  (Gohain  et  al.,2017; Pathak et al.,  2018).  In addition,  the mass transfer and surface interaction between the reactants were decreased in diluted solution due to an excess volume of methanol (Tan et al., 2019). This result is similar to the previously published result of transesterification palm olein using walnut  shell ash as a catalyst  which  obtained  biodiesel conversion of 98% at a ratio of 1:15 (Miladinovi´c et al., 2020). Mendonça et al. (2019) reported identical  results (conversion  of  97.3%)  in  the  utilization  of waste tucum˜a peels as heterogeneous catalyst in transesterification of soybean oil.  In contrast, a biodiesel conversion of 97.5% was obtained using a ratio of molar oil to methanol of 1:6 from transesterification of soybean oil (Sa´nchez-Cantú et al., 2019). However, this result was achieved using a homogeneous catalyst that distributed equally throughout  the  reaction  mixture  rendering  high  activity  (Changmai et al.,  2020). 

A one-way analysis of variance revealed that the biodiesel conver- sions were significantly  different  for all  the  different  ratios  molar  of palm oil to methanol.  Further, the Tukey test showed that the biodiesel conversion  was significantly  different when the transesterification re- action was conducted using a ratio molar of 1:21.

3.3.   Effect of rotational speed

The different polarity of oil and methanol has caused low interaction and mass transfer (Laskar et al.,  2020a). Hence mixing  intensity could increase miscibility leading to high biodiesel conversion (Boocock et al., 1998; Sitepu et al.,  2020).  Thus the homogenizer-intensified  palm oil transesterification using PBA as catalyst was performed through testing various rotational speeds (3000 – 7000 rpm with increment 1000 rpm). In  all  experiments,  the  ratio  molar  of  palm  oil  to  methanol,  catalyst weight and reaction time was fixed as in the legend of Fig.  3B. A biodiesel  conversion  of  >90%  routinely  occurred  for  all  the  rotational speeds tested with the highest conversion of 95 ± 1.1% achieved using a rotational speed of 5000 rpm. This phenomenon presumably due to the transesterification reaction  was controlled  by  the  diffusion  of  the  re- actants (Tarigan  et al.,  2022).  The conversion was constant when the reaction reached equilibrium (Noureddini and Zhu, 1997). This result is in line with the result reported by a previous researcher that obtained a plateau conversion from the rotational speed of 2000 to 4000 rpm (Sanchez-Cantú et al., 2019). Furthermore, Feng et al. (2020) obtained a similar graphical trend in homogenizer-facilitated  lipid extraction from black soldier fly larvae study. The constant lipid extraction efficiency of 70% occurred  in  homogenization  intensity  of  12,000  to  16,000  rpm (Feng et al.,  2020). The ANOVA  analysis showed that there was a sig- nificant effect of rotational speed on biodiesel conversion. However, the Tukey post hoc test did not detect the significance.

3.4.   Effect of catalyst weight

The low catalytic activity of heterogeneous catalysts due to the different  phases with  the  reactant  is one  of  its drawbacks (Abdullah et al.,  2017; Tarigan et al.,  2021).  Therefore in the present study,  the catalyst  weight  in  the range of  5 to 21 wt.%  was investigated  under reaction conditions of ratio molar oil  to methanol  of 1:15,  a reaction time of 30 min and rotational speed of 4000 rpm. The further rise in the catalyst  weight  was attended with an increment in the biodiesel conversion  (Fig.  3C).   The  biodiesel  conversion  achieved  >90%  for  all investigated  catalyst  weight  except  for  5  wt.%.  Lower conversion  at catalyst weight of 5 wt.% presumably due to low availability  of catalytic site (Gohain  et al.,  2017).  The highest biodiesel conversion of 98.9  ± 0.1% was observed at a catalyst weight of 18 wt.%. Similar results were observed by other authors that reported increasing biodiesel conversion in increasing catalyst concentration (from 87% using 0.35 wt% to 97% using 0.5 wt% of catalyst) (Sa´nchez-Cantú et al.,  2019). This finding is also in agreement with Gohain  et al.  (2017) who observed a constant biodiesel yield after reaching a 100% conversion. One-way ANOVA showed a significant  interaction between catalyst weight and biodiesel conversion which was driven by low conversion at catalyst weight of 5 wt.% as detected by Tukey test post hoc.

3.5.   Effect of reaction time

Reaction time has a considerable impact on biodiesel conversion as if not properly adjusted there are possibilities for an incomplete or back- ward reaction which will decrease the yield/conversion of the product (Tan et al.,  2019). In this study, this parameter was determined at six different reaction times from 10 min to 60 min. All the experiments were set to operate using a ratio molar oil to methanol of 1:15, catalyst weight of  7 wt.%  and  rotational  speed of  4000 rpm.  Regularly  the biodiesel conversion  obtained  more  than  90% with  the  lowest  (92.6  ± 0.6%) achieved at a reaction time of 30 min and the highest conversion of 98.7 ± 0.2 procured at 60 min (Fig. 3D). Interestingly, a reaction time of 10 min is sufficient to produce biodiesel with a conversion of 97.4 ± 1.5%. This proves that mass transfer limitation was diminished due to the large turbulence produced by the homogenization device. In addition, a small droplet as a result of high-speed stirring has increased contact intense between reactants increasing reaction rate (Hsiao et al.,  2018; Vikashet al.,  2021). This result agrees with Sanchez-Cantú  et al.  (2019) who reported  conversion  of  97.1%  in  a  dispersion  time  of  only  40  s  in homogenizer-intensified  transesterification of  soybean  oil  using a homogenous catalyst. In contrast, a longer reaction time is required when using heterogeneous CaO  as a catalyst.  Joshi  et al.  (2017) obtained  a biodiesel yield of 84% after 30 min reaction time. However, this could be explained  as PBA contains potassium ions in the forms of oxide or carbonate which  have  catalytic  activity  higher  than CaO  (Alagumalai et al.,  2021). The analysis of variance detected that reaction time has a significant effect on biodiesel conversion which is driven by a low conversion of 92.6 ± 0.6% at a reaction time of 30 min.

3.6.   Leaching test of PBA

The  main  advantage  of  heterogeneous catalysts  is their  ability  to recycle  and  reuse (Chua  et al.,  2020).  However,  its catalytic  activity reduces after several attempts which is due to the leaching  issue (Ala- gumalai  et al.,  2021). Therefore in this study, the leaching  test is per- formed using PBA and methanol with a rotational speed of 4000 rpm for 30 min. The concentration of K2O as the main compound contained in PBA was determined using EDS analysis and compared after 2 cycles of leaching  test. As shown in Table  1, both catalyst  weight  and concen- tration of K2O  decreased after consecutive  use. The significant  weight loss of the catalyst after cycle 1 presumably due to organic compound dissolved in methanol.  The loss was causing less catalyst in the system generating  a low conversion.  In consequence of the significant  weight loss  of   the   PBA   catalyst,   the   biodiesel   conversion   was  markedly decreased from 97.4% to 70.2%.  Boey et al.  (2011) reported a similar phenomenon  and  concluded  that  methanol  hydrolyzes  K2O  to  form KOH. This result is in contrast with previous published catalytic stability of  calcined   wood  ash  which  reported  non-leaching   of  the  catalyst (Sharma   et   al.,    2012).   Mendonça   et   al.   (2019)   also   detected   a non-leaching activity of the catalyst. However, those bio-based catalysts have been calcined at high temperatures before being used generating better  stability   and   diminishing   the   organic   compound   contained (Sharma et al.,  2012).

3.7.   Electricity and time consumption

One important factor to consider for biodiesel commercialization  is electricity  consumption  and reaction time (Tarigan  et al.,  2022).  The electricity consumption of homogenizer intensified biodiesel production using PBA as heterogeneous catalyst was calculated  roughly  based on lab-scale.  Table  2 presented the comparison  of  this study with  tradi- tional reflux, microwave- and ultrasound-assisted methods in electricity consumption and reaction time.  The regular magnetic  stirrer hotplate which consumes 600 W per hour was used in the traditional reflux method (Boey et al., 2011) while microwave and ultrasound require 300 and 299.66 W per hour, respectively to conduct the transesterification reaction  (Attari  et al.,  2022;  Hsiao  et al.,  2020).  The electricity  con- sumption of the homogenizer device to facilitate  the transesterification reaction producing 1 kg of biodiesel was 31.5 kW h which is equivalent to 12.96  MJ  kg  1. The electricity  consumption for PBA catalysing  the transesterification of palm oil using homogenizer device was lower than reflux  and  microwave  methods  and  slightly  higher  than  ultrasound method. The homogenizer device showed an energy saving of 97.8% and 5.75% for reflux and microwave methods, respectively. However, in comparison  with ultrasound device,  the homogenizer  required 30.6% more energy. Furthermore, the homogenizer device is time-wise as this method saved time at 67%, 87% and 75% compared to reflux,  micro- wave and ultrasound methods, respectively.

4.  Conclusion

Transesterification reaction of palm oil was performed using waste PBA as catalyst facilitated by homogenizer device. Large turbulence produced from the high rotation speed of the rotor-stator was increase the reaction rate. The biodiesel conversion in the range of 73% to 98.9% was obtained  with the highest achieved  in reaction condition  of ratio molar palm  oil  to methanol  of 1:15,  a rotational  speed of 5000 rpm, catalyst weight of 18 wt.% and reaction time of 10 min. However, some catalyst amounts eroded due to the leaching  process during the trans- esterification reaction. The concentration of K2O also decreased gradually after several cycles decreasing its catalytic  activity.  The electricity consumption of the homogenizer device was lower than reflux and mi- crowave methods except for the ultrasound method. However, in terms of   time-saving,   this  method   could   save  67–87%  of   reaction   time compared to other transesterification methods.

Declaration of competing interest

The authors declare that they have no known competing  financial interests or personal relationships that could have appeared to influence the work reported in this paper.