Investigation of the effects of pretreatment on the elemental composition of ash derived from selected Nigerian lignocellulosic biomass

Lignocellulosic biomass is an important source of renewable energy and a potential replacement for fossil fuels. In this work, the X-ray fluorescence (XRF) method was used to analyze the elemental composition of raw and pretreated lignocellulosic biomass of cassava peels, corn cobs, rice husks, sugarcane bagasse, yam peels, and mixtures of cassava peels and yam peels, corn cobs and rice husks and all five biomass samples combined. The influence of particle size on elemental properties was investigated by screening the selected biomass into two size fractions, of an average of 300 and 435 µm, respectively. The total concentration of Mg, Al, Si, P, S, Cl, Ca, Ti, Cr, Mn, Fe, Co, Cu, Zn, Sn, Ni, Br, Mo, Ba, Hg, and Pb were determined for each of the biomass samples before and after the different pretreatments adopted in this study. From the results of the analysis, there was a significant reduction in the concentration of calcium in all the analyzed biomass after the alkaline pretreatment with rice husks biomass having the lowest concentration of 66 ppm after the alkaline pretreatment. The sulfur content of the acid pretreated biomass increased considerably which is likely due to the sulfuric acid used for the acid pretreatment. The fact that a mixture of biomass feedstock affects the properties of the biomass after pretreatment was validated in the mixed biomass of cassava peels and yam peels biomass as an example. The concentration of Mg in the mixed biomass was 1441 ppm but was 200 ppm and 353 ppm in individual cassava peels and yam peels respectively. The results of this study demonstrated that pretreated mixtures of biomass have varied elemental compositions, which could be an important factor affecting downstream processes, especially if a hybrid feedstock is used in a large-scale application.


Experimental procedures/methods
Biomass preparation. The corncobs were obtained from Ogume, a village in Ndokwa West Local government area of Delta State, Nigeria. The rice husk used for this work was obtained from a rice mill in Ekperi in Etsako Central Local Government area of Edo State, Nigeria. The other biomass, yam peels, cassava peels, and sugarcane bagasse were locally sourced within Effurun in the Uvwie Local Government Area of Delta State, Nigeria. All biomass were sundried for about 7 days and then taken to the mill where they were ground into a powder. After grinding, the ground biomass were sieved into particle sizes of 300 μm and 425 μm. Pretreatment of biomass was carried out using analytical grade chemical reagents such as sodium hydroxide pellets, hydrogen peroxide, and tetraoxosulphate (VI) acid. 1500 g of each biomass was collected into different containers except for sugarcane bagasse for which 1000 g was collected. To verify the effect of biomass combination on the elemental composition, 300 g each of 300 μm sized particles for all five biomass samples were collected into another container and 750 g of 300 μm sized particles of cassava peels and yam peels combined and corn cobs and rice husks combined were measured into two Table 1. Effect of acid, alkaline, and hydrothermal pretreatment on the raw biomass samples. R = Raw. A = Acid pretreated. B = Alkaline Pretreated. C = Hydrothermal Pretreated.  R  55  175  630  117  287  47  2014 284  9  19  1074 0  0  18  0   A  354  1111 4014 239  14,044 18  1805 454  9  15  654  0  0  9  523   B  0  436  821  58  223  40  599  100  0  0  383  0  0  14  0   C  393  664  1104 219  333  57  1136 176  0  11  628  0  0  9  0   Mean  201  597  1642 158  3722  40  1389 253  4  11  684  0  0  12 A  287  1235 3574 353  7641  17  217  1085 0  33  1170 0  0  7  0   B  92  265  534  374  385  307 168  132  0  12  201  0  0  9  1364   C  1032 780  1356 1154 434  120 302  463  0  68  1411 11  0  15  0   Mean  353  653  1524 554  2181  130 240  567  0  44  999  3  0 10 720 Ash content. The procedure depicted in ASTM D2017 (1998) was used in determining the ash content of the biomass samples in this study. 1 g of sample was placed in a pre-weighed crucible and incinerated in a muffle furnace at 760 °C until complete ashing was achieved. The crucible was then transferred into a desiccator for cooling. Three replicates were made. The cooled samples were then weighed. The ash content was calculated using the equation below.
Xrf analysis. X-ray fluorescence (XRF) spectroscopy is a rapid method used to determine the biomass ash composition. The XRF provides simple analytical solutions to a wide range of quality and process control requirements when compared to other analytical techniques. It can provide detailed analysis on non-destructive analysis, minimal preparation samples, simultaneous multi-element quantitative and qualitative analysis and these results are displayed in seconds. In this study, X-ray fluorescence (x-supreme 8000), incorporated with an application-optimized Oxford instrument's high-reliability X-ray tube and high-performance silicon drift (SDD) was used to analyze the elemental composition of the ashes obtained from the selected lignocellulosic (1) Ash Content (%) = weight of ash original weight of sample × 100  www.nature.com/scientificreports/ biomass being studied. To minimize the error, multiple experiments (triplicate) were carried out for each sample, and the average value was selected for the chemical composition of the samples. XRF analysis was carried out on the ash obtained from the different biomass selected for this study. The elemental composition in 1 kg of the different biomass and biomass mixtures was given in ppm.

Results
General observations. From the result of the experiment, it was seen that the chemical composition of the mineral matter of ashes depends largely on the biomass type, origin, and combustion conditions as corroborated by previous researches on ash components and the varying composition of ash in comparison to the combustion temperature. The lignocellulosic biomass used in this study has varying ash chemical composition for cassava peel, corn cobs, rice husk, sugarcane bagasse, and yam peels as shown in Table 1. The main inorganic mineral constituents of the biomass are Ca, Fe, S, Si, Al, P, and Sn. The five major inorganic constituents of the biomass are shown in decreasing order of abundance as shown in Table 2. The concentration of the elements in the raw biomass samples follows the same trend as that obtained by Cavalaglio et al. 19 in their work centered on the  Effects of particle size. Particle size affected the elemental composition of the raw biomass being studied.
This could be as a result of the presence of mineral matters of technogenic origin that are present in the biomass particles that were not broken to the particle size of ≤ 300 μm sizes or the presence of these minerals in 300 μm sized particles and absent in this same proportion in 425 μm sized particles. For example, from Table 3, Sn is present in the 425 micron-sized cassava particles but absent in the 300 m sized cassava particles. Also, Cr, a trace element, was present in the 300 micron-sized cassava peel biomass particles but absent in the 425 micron-sized particles. Sn, a major constituent of unpretreated or raw 300 micron-sized corn cobs, was absent in the 425 m size. This explains why varying particle sizes may also vary the inorganic mineral constituents or composition (ash content) of the biomass as corroborated by Lori et al. 20 in their work on proximate and ultimate analyses of bagasse, sorghum and millet. Cassava peel biomass is typically rich in Ca with a 2014 ppm and 2098 ppm mass for 300 and 425 μm sized particles respectively. This kind of variation in the concentration of the elemental composition is observed all through the biomass considered in this research. The result also shows that magnesium has a higher concentration in the 425 microns sized biomass for all the biomass being studied except for corn cobs biomass where its concentration is higher in the 300 microns sized biomass. The concentration of zinc in all the biomass in this study is low and the effect of particle size on it is not too pronounced.
Effects of pretreatment. The chemical pretreatment employed in this study was a determinant factor in the elemental composition of the ash content in the biomass being studied. The analysis of the effects of pretreatment on the elemental composition was focused on the 300 μm sized particles. The alkaline pretreatment was seen to reduce the Ca concentration in all the biomass tremendously. The acid pretreatment was seen to increase the sulfur content of the biomass. This is due to the presence of sulfur in the acid (H 2 SO 4 ) used for the pretreatment. High concentrations of Si and Ca form low-melting-point eutectics, which can cause slagging. Salts of these elements do form surface deposits on heating equipment 21 . The five-biomass mixture has a relative concentration of chlorine with maximum concentration in corn cob and sugarcane bagasse with values of 487 ppm and 250 ppm respectively. Chlorine is a major parameter in ash deposits. Its presence reduces the melting point of ash and therefore allows for an easier deposition of ash. Al compounds also play a key role in reducing the melting point of ash. Sulfur oxides form sulphates and condense on the surface of heating equipment. They also form fly ash particles. Generally, fuels with high Ca content will have higher sulfur fixation in the ash 20,22 . Ca and Mg in a biomass fuel increase the ash melting point temperature of the fuel, thus making it more suitable for power plant fuel as against the high concentration of potassium which will, in turn, lead to slagging and formation of hard deposit in the furnace and reboiler. The high phosphorus content of rice husks in hydrothermal pretreatment will influence the burning properties as well as cause the formation of low melting temperature ash 23 . Elements often associated with environmental toxicity are present in the biomass in very minute concentrations. Heavy elements such as Co and Cu were present only in sugarcane bagasse and yam peels and absent for all other samples. It was also noticed that Co and Cu which were present in raw sugarcane bagasse and yam peels were absent after both samples underwent alkaline pretreatment. The impact of alkaline pretreatment was evident in Mg and most of the biomass samples, reducing to 0.000 ppm or very low values. Alkaline pretreatment could be said to be relevant in removing some toxic minerals present in the biomass. Other toxic substances in the biomass include Al, Mn, Cr, and Zn but in trace concentration. It is important to also note that the presence of these substances in the given samples of sugarcane bagasse and yam peels does not validate their presence in all  Acid pretreatment. Table 4 and Fig. 1 show the elemental composition of the different inorganic matter present in the raw and pretreated samples of 300 μm sized particles of cassava peels, corn cobs, rice husks, sugarcane bagasse, and yam peels. From the table of results, it was seen clearly that acid pretreatment varied the elemental composition of the biomass by changing their concentration and not necessarily the elemental constituents. Nevertheless, few samples showed the presence or absence of an element before or after pretreatment. Generally, Mg, Al, Si, P, S, Ti, and Mn showed an increase in concentration across all biomass samples with exception of corn cobs for Mg, Al, Si, and P, and sugarcane bagasse for P alone. The raw sample of corn cobs contained 712 ppm of Mg whereas the acid pretreated sample contained 0.000 ppm of Mg. Also, the raw sample of yam peel biomass contained 0.000 ppm of Mg while the acid pretreated sample contained 287 ppm of Mg. The increase in the sulfur content of the acid pretreated sample against the raw sample was very high, this could be as a result of the sulfuric acid used in the acid pretreatment. There was a decrease in the concentration of Cl and Ca for all samples after acid pretreatment. Acid pretreatment could be useful in reducing the number of certain minerals present in the biomass.
Alkaline pretreatment. Table 5 and Fig. 2 show the results of alkaline pretreatment on the elemental composition of the different biomass. Alkaline pretreatment of the biomass samples brought about a different variation in the elemental composition of ashes from the different biomass samples. The concentration of Si and Cl in all five biomass samples decreased after alkaline pretreatment. Except for yam peels biomass, Mg, P, S, and Cl also show a decrease in concentration after pretreatment in an alkaline medium as well as Al, Cr, Mn, and Fe. Zn and Sn were seen to show a general increase in their concentration after alkaline pretreatment. The two elements were absent (showing 0.000 ppm for both) in the raw/unpretreated sample of rice husk but present after the samples were pretreated in an alkaline medium showing 18 ppm and 430 ppm respectively. The calcium concentration decreased drastically in all the biomass samples most probably because of the sodium hydroxide used for the alkaline pretreatment. Calcium is higher up the electrochemical series and for positive ions, the ones higher up the series displaces the ones lower in the series which could be the reason for the reduced calcium after the alkaline pretreatment. Another factor that could have led to the reduction of the calcium after the alkaline pretreatment could be leaching due to the high pH of the NaOH used for the pretreatment as corroborated by Osman et al. in their work 18 .
Hydrothermal pretreatment. Like alkaline pretreatment, hydrothermal pretreatment brought about a different variation in the elemental composition in the ash content of the five biomass samples. Except for Co and Cu, all the other samples had no clear increase or decrease in the concentration of the elements under consideration. Al, S, Cl, Ti, and Sn had a decrease in their concentration in the different samples but with exceptions in cassava peels, yam peels, for Al, S, and Cl; and corn cobs and rice husks for Ti and rice husks for Sn. Ca, Mg, Si, P, Mn, Fe, Co, Cu, and Zn all showed a general increase in their concentration in the ash content of the biomass samples after pretreatment though with exceptions. Co was 0.000 ppm in raw yam peels but 11 ppm in the hydrothermally pretreated sample of the same biomass. Table 6 and Fig. 3 show the effect of hydrothermal pretreatment.
Effects of mixed biomass. Cassava peels + Yam peels mixed biomass. Table 7 and Fig. 4 show the results of cassava + yam peels biomass. The pretreated samples for cassava + yam peels mixed biomass have varying www.nature.com/scientificreports/  www.nature.com/scientificreports/ concentrations of the different elements. The average Mg, P, S, and Zn concentration in the cassava + yam peels mixed biomass was seen to have increased more than they were for the individual biomass. For example, the concentration of Mg in the mixed biomass was 1441 ppm but was 200 ppm and 353 ppm in cassava and yam respectively. A decrease in concentration was however noticed in the cassava + yam peels mixed biomass for Al, Si, and Ti, when compared to the average composition of the elements in the individual cassava and yam, peels biomass. Ti in the cassava + yam peels mixed biomass was 181 ppm but was 253 ppm and 567 ppm in individual cassava and yam peels biomass respectively. The concentration of Cl, Ca, Cr, Mn, Fe, and Sn in the cassava + yam peels mixed biomass fell between the concentration of these elements in the individual cassava and yam peels biomass. As stated by Smith et al. 25 the composition of biomass changes when two or more biomasses are combined. The choice of biomass for a given use depends on the requirements of the operation or use as each of these elements has its environmental and energy impact on the surroundings and the system. The results for magnesium, calcium and chlorine are in tandem with those obtained by Sadawi et al. 26 in their work on commodity fuels from biomass through pretreatment and torrefaction: effects of mineral content on torrefied fuel characteristics and quality.
Corn cobs + Rice Husks mixed biomass. The pretreated corn cobs + rice husks mixed biomass has a varying concentration of elemental composition as seen in Table 8 and Fig. 5 respectively. The average Mg, S and Al concentration in the corn cobs + rice husks mixed biomass increased more than they are for the individual biomass. For example; the concentration of Mg in the mixed biomass was 800 ppm but 224 ppm and 670 ppm in individual corn cobs and rice husks respectively. Sulfur (S) in the mixed biomass was 9128 ppm but 276 ppm and 3398 ppm in individual biomass samples of corn cobs and rice husks respectively. A decrease in concentration was observed in the corn cobs + rice husks mixed biomass for Ca, and Sn when compared to the average composition of the elements in the individual biomass with 71 ppm for the mixed biomass and 399 ppm and 92 ppm for corn cobs and rice husks respectively. The other elements present in the biomass mixture took a position between the concentration of the individual biomass of corn cobs and rice husks with exception of Zn and Ti which had values equal to the least in the range. The result of the analysis shows there is a considerable change in the elemental composition of the mixture of the corn cobs and rice husks after pretreatment as compared to the singular biomass buttressing the fact that the combination of biomass feedstocks has a significant effect on  www.nature.com/scientificreports/   www.nature.com/scientificreports/ their properties and subsequent output as corroborated in the previous work by the authors 14 . The high value of Phosphorus in the ash of individual and combined biomass of corn cobs and rice husks could be due to the use of fertilizers for crop cultivation 27 .
Mixture of all five biomasses. In analyzing the impact of the mixture of all five biomass samples on the overall elemental composition of the pretreated samples, the average values of the concentration of each element in the mixture of all the five biomass are as shown in Tables 9 and 10 and also depicted in Fig. 6. From Table 10, it was seen that the concentration of Al, Si and Ti in the different biomass considered was the peak in the mixture of all the five biomass samples, with values of 909 ppm for the five biomass mixture and 565 ppm, 523 ppm, and 387 ppm for S, C + Y, and Cc + R respectively for Al. It was also evident from Table 9 that Mg (with exception of sugarcane bagasse where it was 0 ppm), Mn, and Zn had the least values for the five biomass mixtures with 28 ppm in all biomass mixture and 40 ppm, 38 ppm, and 29 ppm for S, C + Y, and Cc + R respectively. Although the difference in the concentration of Mn and Zn is insignificant as compared to the difference in the concentration of Mg.

Conclusion
The results obtained from the XRF analysis of the five lignocellulosic biomass as well as their mixed forms show that pretreatment of samples plays a significant role in modifying the elemental composition of the biomass and this varies with pretreatment type as well. Thus, the effect of acid pretreatment on a biomass sample differs from that of alkaline pretreatment on that same sample. The alkaline pretreatment on the biomass samples show better results, especially on the hybridized (mixed) feedstocks with sulfur having lower concentration than that of acid pretreatment. In the hybridized biomass, the concentration of Al, Si, and Ti in the different biomass considered was the highest in the overall mixture of all the five biomass. For Aluminum, the reported values were 909 ppm for the five-biomass mixture and 565 ppm, 523 ppm, and 387 ppm for S, C + Y, and Cc + R respectively. Information on the elemental and ash composition of biomass is vital for a bioethanol processing plant as the elemental and ash components of the biomass should be given adequate consideration during process design