Deliverable Report

Grant Agreement number: 227464 Project acronym: CANEBIOFUEL Project title: Conversion of Sugar Cane Biomass into Ethanol Deliverable title: D1.2 Quantification and qualitative analysis of selected cane biomass fraction

Page 1 of 8

Deliverable 1.2

CaneBioFuel (227464)

Work progress and achievements during the period

Both selected cane biomass fractions were characterized in relation to their chemical compositions. Inorganic materials (ashes and soil contaminants), low molecular weight organic compounds (extractives), water-soluble extractives, cell wall polysaccharides (cellulose and hemicelluloses) and lignin, present in both samples, were analysed using chromatographic and spectrometric methods. Results are summarized below. Inorganic components Ash content was determined in cane biomass fractions by the TAPPI T211 om-93 method. Calcination was carried out at 700 C for 2 h. Ash was determined gravimetrically and expressed in relation to the dry weight of the original samples specimen. Cane bagasse was shown to have a total ash content of 6.6 wt%, whereas cane straw contained 6.2 wt% of this same inorganic component. However, when the calcination temperature was raised to 900 C, these values decreased to 5.2 wt% for both cane biomass fractions. This decline in ash content was attributed to the loss of carbonate during calcination of cane biomass at higher temperature ranges. On the other hand, cane straw was expected to have much higher ash contents than cane bagasse. However, cane bagasse was analysed as-received while cane straw had to be water-washed to remove soil dust from its surface. Hence, water washing was partially responsible for the lower ash content of cane straw because the silica found on the surface of cane leaves were either solubilized or partly removed by shearing. The chemical composition of cane biomass ashes was determined by X-ray fluorescence in a Philips Analytical XRF apparatus, model PW 2400/00, provided with a Sample Changer model 2510. Results of these XRF analyses are shown in Tables 1 and 2. The total amount of oxides, which were present above the detection limit of the XRF method, corresponded to 94.3 and 99.7 wt% of the total ash content of cane bagasse and cane straw, respectively (Table 1). The lower total accountable oxides in cane bagasse were attributed to the presence of sulfur compounds as determined qualitatively by XRF. Further determinations will have to be carried out to validate this observation and quantify the presence of sulfur in these cane biomass fractions. Several other oxides were also present in both fractions but their quantification was not possible because they were present in concentrations lower than 1000 ppm (XRF detection limit). The quantification of these oxides (SrO, CuO, Nb 2 O 5 , ZnO, ZrO 2 e Rb 2 O) is under way by Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES). Unfortunately, these results were not ready at the time this report was written but their availability is expected for no longer than mid September. SiO 2 was the predominant oxide in both cane biomass samples, corresponding to 33.9 and 51.2 wt% in cane bagasse and cane straw ashes, respectively (Table 2). On the other hand, the CaO content in cane straw was seven times higher than that of cane bagasse, with MgO and K 2 O also displaying larger contributions to the overall oxide content but not to the same extent as observed for CaO. These data indicate differences in the buffering capacity of each of these cane biomass fractions, probably having consequences in their susceptibility to high pressure steaming (steam explosion) as part of the acid catalyst (being exogenous or not) will be neutralized at the pre-impregnation stage.

Page 2 of 8

Deliverable 1.2

CaneBioFuel (227464)

Table 1 - Oxides present in the ashes of cane biomass fractions (bagasse and straw) as determined by X-ray fluorescence (XRF). Values were expressed in relation to the total amount of ash. Oxide (%) CaO MgO SiO 2 Al 2 O 3 Fe 2 O 3 Na 2 O K2O SrO TiO 2 MnO P2O5 loss
1

wt% in bagasse ashes 2.09 0.00 1.50 0.01 33.88 0.08 18.52 0.01 29.13 0.26 <0.01 0.00 3.14 0.01 not detected 8.60 0.00 0.27 0.05 2.10 0.05 0.47 0.00 94.29 0.41

wt% in straw ashes 14.68 0.05 4.55 0.04 51.17 0.25 6.62 0.05 5.76 0.02 0.18 0.01 6.22 0.01 0.10 0.00 1.97 0.01 0.37 0.00 2.02 0.01 0.64 0.00 99.71 0.10

Total

Loss inherent to heating (flame loss)

Table 2 Oxides and elements present in the ashes of cane biomass fractions (bagasse and straw) as determined by X-ray fluorescence (XRF). Values were expressed in relation to the dry weight of the original cane biomass. Oxide type CaO MgO SiO 2 Al 2 O 3 Fe 2 O 3 Na 2 O K2O SrO TiO 2 MnO P2O5 loss1 Total
1

Corresponding element wt% in straw 0.77 0.01 0.24 0.01 2.68 0.01 0.35 0.01 0.30 0.02 0.01 0.01 0.33 0.01 0.01 0.01 0.10 0.01 0.02 0.01 0.11 0.01 0.03 0.01 4.93 0.02 type Ca Mg Si Al Fe Na K Sr Ti Mn P g/Kg in bagasse 0.79 0.01 0.48 0.01 8.35 0.02 5.17 0.01 10.74 0.09 0.04 0.01 1.37 0.01 nd 2.72 0.01 0.11 0.02 0.48 0.01 g/Kg in straw 5.49 0.01 1.44 0.01 12.51 0.06 1.83 0.01 2.11 0.02 0.07 0.02 2.70 0.01 0.04 0.01 0.62 0.01 0.15 0.01 0.46 0.01

wt% in bagasse 0.11 0.01 0.08 0.01 1.79 0.01 0.98 0.01 1.54 0.01 0.01 0.01 0.17 0.01 nd 0.45 0.01 0.08 0.01 0.11 0.01 0.02 0.01 5.25 0.01

Loss inherent to heating (flame loss)

Page 3 of 8

Deliverable 1.2

CaneBioFuel (227464)

The SiO 2 content found in both ashes was much lower than what has been reported in the specialized literature (Table 2). Cane bagasse ashes are reported to contain nearly 85 wt% in silica and not as much aluminum, iron and titanium oxides as observed in this study. The Al 2 O 3 content in cane biomass, which was three times higher in cane bagasse as compared to cane straw, was probably due to soil contamination. On the other hand, higher Fe 2 O 3 and TiO 2 contents in cane bagasse seemed to be due to oxidation and wearing of blades and other cane mill components, which represented an important source of contamination of the resulting material. The former was six while the latter was four times higher in cane bagasse as compared to cane straw, whose Fe 2 O 3 and TiO 2 contents agreed with data available in the literature. Macromolecular components The selected cane biomass fraction, reduced to two samples primarily composed of industrial cane bagasse fibers and freshly harvested cane straw from the same cane stock, were characterized in relation to their chemical composition after extraction with solvent of increasing polarity in a Soxhlet apparatus (see below for details on quantitative and qualitative analysis). Klason lignin analysis was carried out to determine both the acid-insoluble (TAPPI Standard Method T222 os-74) and the acid-soluble (TAPPI Useful Method 250) lignin components of cane bagasse and cane straw. Acid hydrolysates derived from Klason lignin analyses were also characterized by high performance liquid chromatography (HPLC) to reveal carbohydrate components such as glucans (cellulose) and polyoses (hemicelluloses), as well as acetic acid present in hemicelluloses as acetyl groups. Table 3 shows the chemical composition of both cane bagasse and cane straw in relation to the dry weight of the starting material, whereas Figure 1 shows the HPLC profile of both cane biomass acid hydrolysates. HPLC analyses were carried out by ion exchange chromatography in an Aminex HPX-87H column at 65 C, having H 2 SO 4 as mobile phase at 1 mL/min. Table 3 HPLC analysis of cane biomass acid hydrolysates. Component Total extractives in organic solvents Water-soluble extractives Anhydroglucose Anhydroxylose Acetyl group
2 2 1

Bagasse (wt%) 2.88 0.03 1.23 0.05 38.39 0.13 19.11 0.06 3.38 0.02 3.53 0.02 0.23 0.04 20.81 0.26 6.60 0.2 96.16

Straw (wt%) 3.50 0.06 3.52 0.01 37.41 0.01 20.40 0.13 4.29 0.03 2.32 0.01 0.77 0.03 17.70 0.23 6.20 0.2 96.11

Anhydroarabinose2 Acid-soluble lignin Acid-insoluble lignin Ash Total

1 2

Present in cane bagasse and cane straw as (1-4)-D-glucans (cellulose) Present in cane bagasse and cane straw as heteroxylans (hemicelluloses)

Page 4 of 8

Deliverable 1.2

CaneBioFuel (227464)

The total accountable matter in Table 3 reached 96.16 wt% for cane bagasse and 96.11 wt% for cane straw. These numbers are consistent with the fact that 4-Omethyl-glucuronic acid is hardly detectable using ion exchange chromatography under the conditions used in this study. As this hemicelluloses component may correspond to as much as 4 wt% of the cane biomass dry weight, the totals shown in Table 3 are consistent with an almost quantitative mass balance in the determination of both cane biomass chemical composition. Measurement of 4-Omethyl-glucuronic acid using colorimetric methods will be immediately carried out to estimate the contribution of this residue in the primary structure of cane biomass hemicelluloses. The Klason lignin content of cane bagasse was higher than that of cane straw. These data indicate that cane straw fibers are less lignified than cane bagasse, particularly if one considers that the protein content of these fractions has not been determined as yet. Proteins do not hydrolyse under mild hydrolysis conditions and easily precipitate in low pH. Therefore, if present, they are likely to be accounted for as acid-insoluble lignin. Protein content in cane straw is probably higher than that of cane bagasse, leading to an over-estimation of its acid-insoluble lignin component. Measurement of nitrogen content by Kjeldahl in both fresh cane biomass samples and acid-insoluble lignins will easily settle this question. By contrast, the acid-soluble lignin in cane straw was higher than that of cane bagasse and this was attributed to the presence of secondary metabolites in the former, which is more metabolically active than the latter.

Page 5 of 8

Deliverable 1.2

CaneBioFuel (227464)

Figure 1 HPLC analysis of Klason lignin acid hydrolysates derived from (A) cane bagasse and (B) cane straw. Glc, glucose; Xyl, xylose; Ara, arabinose; RID, reffractive index; UV, ultraviolet spectroscopy. Another useful observation from Table 3 is that the hemicelluloses content of cane straw was slightly higher than that of cane bagasse while the cellulose content, expressed as glucans, was practically the same (Figure 1). The higher arabinose content (Figure 1a) in the former was also noticeable and this is one of the reasons why cane straw Klason lignin hydrolysates contained a higher concentration of furfural (Figure 1b). As mentioned above, these dehydration by-products arise from pentoses and arabinose, being easily hydrolysed as a pentosan side group and less stable in acid media due to its furanosidic conformation, is the primary origin for the accumulation of furfural in acid hydrolysates. Altogether, these data indicate that high pressure steaming hemicellulosic hydrolysates coming from cane straw will contain a greater amount of potential hydrolysis and/or fermentation inhibitors due to its greater hemicellulose content and higher arabinose-to-xylose ratio. On the other hand, the cane straw hemicelluloses seem to be less acetylated than the corresponding cane bagasse hemicelluloses but the effect of this property on steam treatment cannot be easily anticipated as lower acetyl content generates less acetic acid in autohydrolysis experiments but steam-treated substrates with lower acetyl content are bound to have greater accessibility to cellulase enzymes which are not rich in acetyl xylan esterases. Following the same context discussed above, HPLC analyses were also useful to quantify furfural and hydroxymethylfurfural in Klason lignin acid hydrolysates. These dehydration by-products were converted to pentoses and hexoses by applying a suitable conversion factor and the resulting amounts were added to the quantification based on the individual standards to express the total amount of anhydroglucose (cellulose) and anhydroarabinose (hemicelluloses) of both cane biomass fractions. In this case, the entirety of the detectable furfural was converted to anhydroarabinose because this residue (arabinose) is by far more susceptible than xylose to acid hydrolysis and dehydration. Interestingly, cane straw produced Klason lignin hydrolysates with much higher concentration of dehydration by-products than cane bagasse (3.41 0.17 vs. 1.22 0.47 wt% for furfural and 1.39 0.01 vs. 0.22 0.01 wt% for hydroxymethylfurfural). This is indicative of a more accessible hemicelluloses component in cane straw which might translate into a different accessibility for pretreatment and hydrolysis as well. However, this is yet to be demonstrated because the steam treatment of cane straw is yet to be realized. The furfural-tohydroxymethylfurfural ratio in these hydrolysates was also different and this was consistent with the higher water-soluble extractives content of cane straw (see below for details). Being primarily composed of starch, these extractives contribute more to hydroxymethylfurfural formation than cellulose, reason for what the ratio was higher for cane straw as compared to cane bagasse. Extractives The selected cane biomass fraction, reduced to two samples primarily composed of industrial cane bagasse fibers and freshly harvested cane straw from the same cane stock, were characterized in relation to their chemical composition. Both samples were initially extracted with five different solvents with increasing polarity and the solvent extracts were quantified gravimetrically (Table 4) and analysed by Fouriertransformed infrared spectroscopy (FTIR) as a preliminary characterization step towards the elucidation of their chemical composition (Figures 2 and 3).

Page 6 of 8

Deliverable 1.2

CaneBioFuel (227464)

Table 4 Extractives found in cane biomass (bagasse and straw). Sample Bagasse Straw EE (%) 0.79 0.01 1.31 0.07 DCM (%) 0.23 0.02 0.27 0.04 EtOH:Tol (%) 1.42 0.05 1.45 0.06 EtOH 95% (%) 0.44 0.04 0.47 0.05 Water (%) 1.23 0.05 3.52 0.01

EE, ethyl ether; DCM, dichloromethane; EtOH:Tol, ethanol, toluene; EtOH, ethanol

Quantitatively speaking, cane bagasse and cane straw differed considerably in the amount of detectable EE and water-soluble extracts. While the nature of the EE extracts will be soon characterized by GC-MS, the larger amount of water-soluble material in the cane straw is probably due to the presence of secondary metabolites and energy storage compounds (starch) that are critical for photosynthesis and plant growth. The preliminary characterization of cane biomass extractives was carried out by Fourier transformed infrared spectroscopy. The FTIR spectra of cane bagasse extractives in ethyl ether (EE), dichloromethane (DCM), ethanol:toluene 2:1 vol/vol (EtOH:Tol), ethanol (EtOH) and water are shown in Figure 2, while Figure 3 contains the corresponding FTIR spectra derived from cane straw. In general, the extractives obtained by treating cane bagasse with EE in a Soxhlet apparatus had a FTIR spectrum typical of aliphatic organic compounds of low polarity (Figure 2). This conclusion arose from the assignment of the axial deformations in 2920 and 2850 cm-1 to the C-H linkage in CH 3 e CH 2 groups, respectively, as well as to the angular deformation of C-H in CH 2 at 1463 cm-1. Further evidence in the FTIR spectra, including the absence of the stretching vibration of associated O-H groups, suggest that these extractives are primarily composed by lipids, fatty acids, terpenes and waxes. The breathing vibration of the aromatic ring was initially observed at 1540 cm-1 in the FTIR spectrum of the DCM extract and became relatively predominant in the FTIR spectrum of the EtOH:Tol extract, together with other bands that are characteristic of phenolic acids, sterols, lignans and other oxygenated aromatic compounds. The 2400-3300 cm-1 region was attributed to the stretching vibration of associated O-H, while the band at 1168 cm-1 corresponded to the stretching vibration of O-C-O in acetals, most likely belonging to the glycosidic linkages of cane polysaccharides. These bands were predominant in the FTIR spectra of the latter two cane biomass extractive families, particularly in the water extract, whose hydrophilicity and FTIR profile was typical of water soluble carbohydrates such as starch.

Page 7 of 8

Deliverable 1.2

CaneBioFuel (227464)

3500

3000

2000 2500 Wavenumber (cm-1)

1500

1000

500

Figure 2 - FTIR analysis of extractives found in cane bagasse using ethyl ether, dichloromethane, ethanol/toluene, ethanol and water in a typical Soxhlet extraction apparatus. The FTIR spectra of cane straw EE, DCM, EtOH/Tol, EtOH and water soluble extracts (Figure 2) were quite different to those previously obtained from cane bagasse (Figure 3). However, the assignments were typically the same, with differences only applying to the occurrence and to the relative intensity of each of the preassigned absorption bands. In this case, the similarity among the FTIR spectra of cane straw EE, DCM and EtOH/Tol extracts was even more striking, suggesting that a single extraction procedure with one of these solvent (probably DCM, for technical reasons) would suffice for this specific cane biomass fraction. Also, the FTIR spectra of these three fractions revealed that oxygenated compounds are present even when the solvent of lowest polarity was used. Further information about the chemical composition of these families of organic compounds will be generated by GC-MS.

Page 8 of 8

Deliverable 1.2

CaneBioFuel (227464)

3500

3000

2500 2000 Wavenumber (cm-1)

1500

1000

500

Figure 3 - FTIR analysis of extractives found in cane straw using ethyl ether (EE), dichloromethane (DCM), ethanol/toluene 2:1 (EtOH/Tol), ethanol (EtOH) and water in a typical Soxhlet extraction apparatus.