**4. Results and discussion**

### **4.1. Reference material spectra**

Figure 1 (a-c) shows the normalized spectral results obtained from a typical reference cellulose (100%), hemicellulose (100%), and lignin (100%), respectively, after smoothening and baseline corrections. Generally, FTIR-PAS techniques permit to obtain spectra which present specific characteristic key bands of individual components. In principle, such band position allows the discrimination of different species and provides important information about the chemical compositions of the material.

**Figure 1.** FTIR-PAS spectra of a) pure lignin (hydrolytic), b) cellulose (microcrystalline powder), and c) hemicellulose (xylan from birch wood) powders showing prominent peaks/band positions at the various characteristics wavenum‐ bers.

#### **4.2. Characteristic peak positions**

The various prominent (characteristic) peak (band) positions and the corresponding peak assignments for cellulose, hemicellulose, and lignin are presented in Table 3. These are distinct peaks at different wavenumbers associated with these three biomass components. Lignin spectrum has six prominent peaks at wavenumbers of 1599 (X1), 1511 (X2), 1467 (X3), 1429 (X4), 1157 (X5), and 1054 cm-1 (X6) (Figure 1a). Cellulose spectrum has five distinct peaks at wave‐ numbers of 1431 (X1), 1373 (X2), 1338 (X3), 1319 (X4), and 1203 cm-1 (X5) (Figure 1b). Hemicel‐ lulose spectrum has six characteristic peaks at wavenumbers of 1606 (X1), 1461 (X2), 1251 (X3), 1213 (X4), 1166 (X5), and 1050 cm-1 (X6) (Figure 1c).


(a) Lignin (b) Cellulose

(c) Hemicellulose

**Figure 1.** FTIR-PAS spectra of a) pure lignin (hydrolytic), b) cellulose (microcrystalline powder), and c) hemicellulose (xylan from birch wood) powders showing prominent peaks/band positions at the various characteristics wavenum‐

The various prominent (characteristic) peak (band) positions and the corresponding peak assignments for cellulose, hemicellulose, and lignin are presented in Table 3. These are distinct peaks at different wavenumbers associated with these three biomass components. Lignin spectrum has six prominent peaks at wavenumbers of 1599 (X1), 1511 (X2), 1467 (X3), 1429 (X4), 1157 (X5), and 1054 cm-1 (X6) (Figure 1a). Cellulose spectrum has five distinct peaks at wave‐ numbers of 1431 (X1), 1373 (X2), 1338 (X3), 1319 (X4), and 1203 cm-1 (X5) (Figure 1b). Hemicel‐ lulose spectrum has six characteristic peaks at wavenumbers of 1606 (X1), 1461 (X2), 1251 (X3),

bers.

312 Biofuels - Status and Perspective

**4.2. Characteristic peak positions**

1213 (X4), 1166 (X5), and 1050 cm-1 (X6) (Figure 1c).

**Table 3.** Characteristics peaks/bands position and assignment of pure cellulose, hemicellulose, and lignin (adapted from [21])

### **4.3. Radio frequency-alkaline treatment on lignin**

It was observed that NaOH concentration is a major factor in the pretreatment process. The swelling initiated by NaOH creates pores in the biomass matrix [46-48], which helps enhance the reactivity of the biomass matrix to any externally added material such as enzyme [8]. A similar finding was reported by [49]. The lignin values obtained from the traditional wet chemistry chemical composition analysis and through FTIR-PAS as shown in Table 4 depicts that in most cases, there is a decrease in the lignin in the pretreated samples as compared to the non-treated samples. This may be due to the unloosened lignified matrix in the non-treated sample which is tightly bound to the other constituents, unlike the RF alkaline pretreated samples with broken bonds which has been structurally separated, and disrupted. This disaggregation may be attributed to the interaction between the biomass and NaOH solution in the presence of the RF heating which is believed to be responsible for this solid loss [8]. It also seems that the disruption and deconstruction of the lignified matrix is associated with the dipole interaction, flip flop rotation, and friction generated between the electromagnetic charges from the RF radiation and the ions and molecules from the NaOH solution and the biomass [8]. Lignin removal is an important part of the pretreatment process, because lignin can effectively inhibit/prevent the cellulase enzymes from hydrolysing the cellulose. This lignin reduction resulting from the alkaline pretreatment had also been reported by [50-51].

Lignin estimated by FTIR-PAS was higher than those obtained with NREL procedures. The laborious and time consuming traditional method that uses 72% H2SO4 seems to create a more stringent condition that may lead to altering and further degrading the native cell wall compositions (such as lignin and complex carbohydrates), structure, and possibly generating artifacts. FTIR-PAS is a rapid, direct, non-invasive, and non-destructive chemical analytical technique. This rapid method can detect molecular chemical characteristics of biological materials at high spatial resolutions without altering the inherent biomass structure such as the tissue [40, 44].

The difference in the lignin values from the traditional approach and FTIR-PAS may also be attributed to the spectrum manipulations and the assumption that the reference lignin sample is 100% pure. The FTIR-PAS qualitative and quantitative analytical chemical information can be connected to the structural information within cellular dimension [42].

Equation 1 shows the predictive model for lignin generated from the training and verification analysis of the combined values from RF and SE using the methods described in section 6.3.8.

### **Lignin predictive model:**

$$\text{V\%L} = -33.92X\_1 + 52.61X\_2 + 32.16X\_3 + 208.14X\_4 + 98.46X\_5 - 56.59X\_6 + 17.33 + \varepsilon \tag{1}$$

%L=%lignin (dry matter basis), Xi =regressors/explanatory variables (normalized data based on the respective characteristic peaks, as shown in Figure 1a), with the regressors representing the respective wavelengths as 1599 (X1), 1511 (X2), 1467 (X3), 1429 (X4), 1157 (X5), and 1054


**4.3. Radio frequency-alkaline treatment on lignin**

314 Biofuels - Status and Perspective

the tissue [40, 44].

**Lignin predictive model:**

%L=%lignin (dry matter basis), Xi

It was observed that NaOH concentration is a major factor in the pretreatment process. The swelling initiated by NaOH creates pores in the biomass matrix [46-48], which helps enhance the reactivity of the biomass matrix to any externally added material such as enzyme [8]. A similar finding was reported by [49]. The lignin values obtained from the traditional wet chemistry chemical composition analysis and through FTIR-PAS as shown in Table 4 depicts that in most cases, there is a decrease in the lignin in the pretreated samples as compared to the non-treated samples. This may be due to the unloosened lignified matrix in the non-treated sample which is tightly bound to the other constituents, unlike the RF alkaline pretreated samples with broken bonds which has been structurally separated, and disrupted. This disaggregation may be attributed to the interaction between the biomass and NaOH solution in the presence of the RF heating which is believed to be responsible for this solid loss [8]. It also seems that the disruption and deconstruction of the lignified matrix is associated with the dipole interaction, flip flop rotation, and friction generated between the electromagnetic charges from the RF radiation and the ions and molecules from the NaOH solution and the biomass [8]. Lignin removal is an important part of the pretreatment process, because lignin can effectively inhibit/prevent the cellulase enzymes from hydrolysing the cellulose. This lignin reduction resulting from the alkaline pretreatment had also been reported by [50-51].

Lignin estimated by FTIR-PAS was higher than those obtained with NREL procedures. The laborious and time consuming traditional method that uses 72% H2SO4 seems to create a more stringent condition that may lead to altering and further degrading the native cell wall compositions (such as lignin and complex carbohydrates), structure, and possibly generating artifacts. FTIR-PAS is a rapid, direct, non-invasive, and non-destructive chemical analytical technique. This rapid method can detect molecular chemical characteristics of biological materials at high spatial resolutions without altering the inherent biomass structure such as

The difference in the lignin values from the traditional approach and FTIR-PAS may also be attributed to the spectrum manipulations and the assumption that the reference lignin sample is 100% pure. The FTIR-PAS qualitative and quantitative analytical chemical information can

Equation 1 shows the predictive model for lignin generated from the training and verification analysis of the combined values from RF and SE using the methods described in section 6.3.8.

on the respective characteristic peaks, as shown in Figure 1a), with the regressors representing the respective wavelengths as 1599 (X1), 1511 (X2), 1467 (X3), 1429 (X4), 1157 (X5), and 1054

=regressors/explanatory variables (normalized data based

e(1)

%L 33.92 52.61 32.16 208.14 98.46 56.59 17.33 *XX X XXX* 12 3 4 56 =- + + + + - + +

be connected to the structural information within cellular dimension [42].


Standard error=standard deviation between the wet chemistry and FTIR-PAS values divided by the square root of 2; W=washed after pretreatment; DW=pretreated with distilled water; TW=pretreated with tap water.

**Table 4.** Lignin composition of RF-alkaline pretreated and non-treated biomass grind obtained using the traditional wet chemistry and FTIR-PAS methods

cm-1 (X6). 17.33=intercept, and ε=error term/stochastic variable which describes the noise (errors that could emanate from the equipment, environment, or the experimenter).

Table 5 shows the R2 and mean square error values of the various biomass components from the regression analysis. However, the R2 values are too low; this might be due to the associated stochastic variables and the spectrum manipulations.


**Table 5.** R2 and mean square error values from the regression analysis

Washing the pretreated samples reduces the lignin. This may be due to loss of solid lignin during the washing process [8]. This investigation shows that the concentration of NaOH solution and the ratio of biomass to the NaOH solution are the dominant contributing factors, while temperature plays a lesser role. The heat provided by the RF is needed to assist the alkaline solution in the deconstruction and disaggregation of lignocellulosic biomass matrix. It was also observed from this investigation that biomass can be alkaline pretreated even at room temperature if the required ratio of biomass and NaOH solution is applied.

### **4.4. Radio frequency-alkaline treatment on cellulose and hemicellulose**

The difference in the cellulose and hemicellulose values from the traditional approach and FTIR-PAS may be attributed to the spectrum manipulations and the assumption that the reference cellulose and hemicellulose samples are 100% pure. Equations 2-3 show the cellulose and hemicellulose predictive models generated from the training and verification analysis of the combined values from RF and SE using the methods described in section 3.


Photoacoustic Spectroscopy in the Assessment of the Quantitative Composition of the Biomass — Barley Straw http://dx.doi.org/10.5772/59225 317


Standard error=standard deviation between the wet chemistry and FTIR-PAS values divided by the square root of 2; W=washed after pretreatment; DW=pretreated with distilled water; TW=pretreated with tap water.

**Table 6.** Cellulose and hemicellulose compositions of RF-alkaline pretreated and non-treated biomass grind obtained using the traditional wet chemistry and FTIR-PAS methods

#### **Cellulose predictive model:**

Biomass composite R2

316 Biofuels - Status and Perspective

**Table 5.** R2

**Temperature**

**Biomass: NaOH solution**

**Wet chemistry FTIR-PAS Standard**

**ratio**

**( oC)**

% Lignin 0.68 9.10 % Cellulose 0.34 13.86 % Hemicellulose 0.31 15.86

and mean square error values from the regression analysis

value Mean square error

Washing the pretreated samples reduces the lignin. This may be due to loss of solid lignin during the washing process [8]. This investigation shows that the concentration of NaOH solution and the ratio of biomass to the NaOH solution are the dominant contributing factors, while temperature plays a lesser role. The heat provided by the RF is needed to assist the alkaline solution in the deconstruction and disaggregation of lignocellulosic biomass matrix. It was also observed from this investigation that biomass can be alkaline pretreated even at

The difference in the cellulose and hemicellulose values from the traditional approach and FTIR-PAS may be attributed to the spectrum manipulations and the assumption that the reference cellulose and hemicellulose samples are 100% pure. Equations 2-3 show the cellulose and hemicellulose predictive models generated from the training and verification analysis of

> **Wet chemistry**

Non-treated - 42.51 50.37 3.93 29.98 23.82 3.08 1:4 22.25 28.65 3.20 23 18.10 2.45 1:4 22.37 27.69 2.66 22.14 28.47 3.16 1:4 26.93 26.09 0.42 26.24 20.73 2.75 1:5 24.21 4.98 9.61 21.63 30.37 4.37 1:5 21.07 29.42 4.18 21.38 24.56 1.59 80W 1:5 27.69 26.55 0.57 21.6 15.81 2.90 1:5 24.65 37.03 6.19 21.05 8.44 6.31 90W 1:5 33.44 34.50 0.53 26.08 22.58 1.75 1:6 30.93 14.95 7.99 29.12 21.00 4.06 1:6 28.25 29.02 0.39 23.18 22.88 0.15 1:6 30.37 20.85 4.76 22.36 39.31 8.47

**Cellulose (%) Hemicellulose (%)**

**FTIR-PAS Standard error**

room temperature if the required ratio of biomass and NaOH solution is applied.

the combined values from RF and SE using the methods described in section 3.

**error**

**4.4. Radio frequency-alkaline treatment on cellulose and hemicellulose**

$$2\% \text{C} = 28.63 \text{X}\_1 + 48.60 \text{X}\_2 + 35.83 \text{X}\_3 - 51.71 \text{X}\_4 - 29.24 \text{X}\_5 + 37.16 + \varepsilon \tag{2}$$

Where %C=%cellulose value, Xi =regressors/explanatory variables (normalized data based on the respective characteristic peaks, as shown in Figure 1b), with the regressors representing the respective wavelengths as 1431 (X1), 1373 (X2), 1338 (X3), 1319 (X4), and 1203 cm-1 (X5), 37.16=intercept, and ε=error term/stochastic variable.

#### **Hemicellulose predictive model:**

$$\text{V\#H} = -14.25\text{X}\_1 - 90.42\text{X}\_2 + 34.14\text{X}\_3 - 39.18\text{X}\_4 + 71.17\text{X}\_5 + 122.90\text{X}\_6 + 30.42 + \varepsilon \tag{3}$$

%H=%hemicellulose wet chemistry value, Xi =regressors/explanatory variables (normalized data based on the respective characteristic peaks, as shown in Figure 1c), with the regressors representing the respective wavelengths as 1606 (X1), 1461 (X2), 1251 (X3), 1213 (X4), 1166 (X5), and 1050 cm-1 (X6), 30.42=intercept, and ε=error term/stochastic variable.

Ramesh and Singh (1993) [2] reported that barley straw theoretically contains about 40% cellulose, 20% hemicellulose, and 15% lignin. Marsden and Gray (1985) [52] also reported that barley straw theoretically contains about 44% cellulose, 27% hemicellulose, and 7% lignin. These values are comparable with the values obtained from the non-treated sample in this investigation. It should be noted that the variance in the chemical composition between the reported theoretical values and the values obtained from this investigation may be attributed to differences in locations where the crop was grown, weather conditions, the barley variety grown, and different methods of analysis.

### **4.5. Steam explosion treatment on lignin**

The efficacy of FTIR-PAS techniques for studying changes in plant cell wall composition following steam explosion pretreatment has been evaluated. Table 7 shows that the SE pretreated samples have higher lignin content as compared to the non-treated. This may be due to the carbonization of the sample resulting from the direct contact of biomass with the walls of the reactor during the steam explosion pretreatment.


Standard error=standard deviation between the wet chemistry and FTIR-PAS values divided by the square root of 2; \*= % mass fraction of water.

**Table 7.** Lignin composition of SE pretreated and non-treated biomass grind obtained using the traditional wet chemistry and FTIR-PAS methods

Hemicellulose degrades easily and some volatile organic compounds vaporize as volatile components, while cellulose behaves as a fixed carbon (solid combustible residue). This may account for the increase in the lignin content. Lam et al. (2011) [18] investigated the steam explosion of Douglas fir (*Pseudotsuga menziesii*) at a reaction temperature of 200-220o C and a retention time of 5-10 min. These researchers reported that there was increase in lignin content from 30 to 43% attributed to the thermal degradation of hemicellulose during the steam explosion treatment. Chen and Kuo (2011) [53] reported that cellulose and lignin are both locked in biomass from the mild carbonization process. This indicates that the degraded cellulose may appear as residue resulting to the increase in the lignin content.

The interaction between moisture content and temperature and also between retention time and moisture content had a statistically significant effect (P<0.01) on the lignin. The interaction among the three variables (moisture content, temperature, and retention time) also had a significant effect on the lignin. The difference in the lignin values obtained using the FTIR-PAS and the traditional approach may be attributed to the spectrum manipulations and the assumption that the reference lignin sample is 100% pure.

### **4.6. Steam explosion treatment on cellulose and hemicellulose**

These values are comparable with the values obtained from the non-treated sample in this investigation. It should be noted that the variance in the chemical composition between the reported theoretical values and the values obtained from this investigation may be attributed to differences in locations where the crop was grown, weather conditions, the barley variety

The efficacy of FTIR-PAS techniques for studying changes in plant cell wall composition following steam explosion pretreatment has been evaluated. Table 7 shows that the SE pretreated samples have higher lignin content as compared to the non-treated. This may be due to the carbonization of the sample resulting from the direct contact of biomass with the

**Lignin (%)**

Standard error=standard deviation between the wet chemistry and FTIR-PAS values divided by the square root of 2; \*=

**Table 7.** Lignin composition of SE pretreated and non-treated biomass grind obtained using the traditional wet

**Wet chemistry FTIR-PAS Standard error**

grown, and different methods of analysis.

318 Biofuels - Status and Perspective

**4.5. Steam explosion treatment on lignin**

**Moisture content (%)\***

**Temperature**

% mass fraction of water.

chemistry and FTIR-PAS methods

**( oC)**

walls of the reactor during the steam explosion pretreatment.

**Retention time**

Non-Treated - - 20.12 22.62 1.25 8 5 23.79 26.25 1.23 8 5 22.72 31.46 4.37 8 5 40.58 21.41 9.58 30 5 22.05 28.23 3.09 30 5 21.69 24.53 1.42 30 5 33.01 25.99 3.51 50 5 21.18 25.41 2.11 50 5 23.31 33.21 4.95 50 5 25.04 32.60 3.78 8 10 21.56 31.63 5.04 8 10 21.66 25.13 1.73 8 10 32.75 33.45 0.35 30 10 21.25 29.37 4.06 30 10 20.9 27.47 3.28 30 10 37.31 29.21 4.05 50 10 21 23.54 1.27 50 10 31.82 31.16 0.33

**(min)**

From Table 8, in comparison with the non-treated biomass, it is evident that retention time, moisture content, and temperature had a significant effect on the cellulose and hemicellulose content. The decrease in the sugars content increased at higher retention time and temperature. Wang et al. (2009) [54] also reported that the retention time and temperature are the process parameters required for the optimization of steam explosion process. In this present study, less degradation of the simple sugars was observed at higher moisture content. High feedstock moisture content acts as acid catalyst to hydrolyze biomass during steam explosion. However, the direct contact of biomass with the walls of the reactor will limit and affect the extent of the hydrolysis. Therefore, a combination of carbonization and acid catalyzed hydrolysis occurred which ultimately led to the degradation of the simple sugars and increase in the lignin. The obtained results from the wet chemistry demonstrated that the hemicellulose contained in the biomass was highly degraded (79% to 89%) compared to cellulose (58% to 77%) as reported in chapter 4. The high degradation of hemicellulose was due to its amorphous nature, which degrades easily and evaporates as volatile components during the carbonization process. Presumably, the crystallinity of cellulose was responsible for the less degradation of this component. These degradations can be explained by considering the monomers of hemicel‐ lulose and cellulose which consist primarily of sugars. Degradation of cellulose and hemicel‐ lulose during steam explosion/thermal pretreatment of biomass has been reported by [16, 19-20, 55-57]. It was reported that hemicellulose is very reactive and was nearly completely removed at 200o C, while both cellulose and lignin can be dissolved partially at higher tem‐ peratures. Shaw (2008) [57] performed steam explosion on poplar and wheat straw at 200-205o C, steam pressure of 1.66-1.73 MPa for 4-4.5 min. This author reported a decrease in cellulose and hemicellulose content with an increase in the lignin content after the steam explosion treatment of both biomass samples. Chen and Kuo (2011) [53, Yang et al. (2007) [59], and Khezami et al. (2007) [60] studied the effects of thermal process on biomass. These researchers showed that thermal pretreatment removes moisture and light volatiles from biomass. Bergman et al. (2005) [61] and Lipinsky et al. (2002) [62] reported that during the torrefaction process, biomass was partly decomposed giving off various condensable and noncondensable gases, with a carbon-rich solid as a final product. Mohammad and Karimi (2008) [7] reported a corresponding decrease in total sugar recovery with increasing temperature during steam explosion.

Furthermore, the difference between the cellulose and hemicellulose values estimated using the FTIR-PAS and the measured values using the traditional approach might also be associated with the spectrum manipulations and the assumption that the reference cellulose and hemi‐ cellulose samples are 100% pure. The main advantage of this correlation is that based on collection of FTIR-PAS spectra, it provides a rapid, easy, economical, non-destructive, and nonlaborious estimation of the chemical composition of biomass. This may be of particular interest in the contexts where more sophisticated and expensive equipments for experimental meas‐ urement of biomass chemical compositions are not always available. However, care must be taken in selecting the steam explosion pretreatment conditions in order to prevent excessive degradation of the chemical properties of the complex carbohydrates, since the yields of hemicelluloses and cellulose were dependent on the pretreatment conditions of the steam explosion.



Standard error=standard deviation between the wet chemistry and FTIR-PAS values divided by the square root of 2; Tempt.=temperature; M.C.=moisture content; R.T.=retention time.

**Table 8.** Cellulose and hemicellulose compositions of SE pretreated and non-treated biomass grind obtained using the traditional wet chemistry and FTIR-PAS methods

### **5. Conclusion**

researchers showed that thermal pretreatment removes moisture and light volatiles from biomass. Bergman et al. (2005) [61] and Lipinsky et al. (2002) [62] reported that during the torrefaction process, biomass was partly decomposed giving off various condensable and noncondensable gases, with a carbon-rich solid as a final product. Mohammad and Karimi (2008) [7] reported a corresponding decrease in total sugar recovery with increasing temperature

Furthermore, the difference between the cellulose and hemicellulose values estimated using the FTIR-PAS and the measured values using the traditional approach might also be associated with the spectrum manipulations and the assumption that the reference cellulose and hemi‐ cellulose samples are 100% pure. The main advantage of this correlation is that based on collection of FTIR-PAS spectra, it provides a rapid, easy, economical, non-destructive, and nonlaborious estimation of the chemical composition of biomass. This may be of particular interest in the contexts where more sophisticated and expensive equipments for experimental meas‐ urement of biomass chemical compositions are not always available. However, care must be taken in selecting the steam explosion pretreatment conditions in order to prevent excessive degradation of the chemical properties of the complex carbohydrates, since the yields of hemicelluloses and cellulose were dependent on the pretreatment conditions of the steam

**Cellulose (%) Hemicellulose (%)**

**Wet chemistry FTIR-PASStandard**

**error**

**error**

**Wet chemistry FTIR-PASStandard**

Non-treated - - 42.51 50.37 3.93 29.98 23.82 3.08 8 5 11.19 20.53 4.67 3.42 12.39 4.48 8 5 11.17 17.93 3.38 3.29 13.10 4.90 8 5 9.58 18.54 4.48 3.74 6.60 1.43 30 5 14.82 19.87 2.53 3.8 17.79 7.00 30 5 14.63 17.53 1.45 4.13 12.54 4.21 30 5 12.07 17.30 2.62 3.88 11.44 3.78 50 5 15.81 22.06 3.12 6.58 15.95 4.69 50 5 14.51 19.74 2.62 5.11 15.18 5.04 50 5 15.77 20.24 2.23 5.23 14.02 4.40 8 10 10.02 16.91 3.45 3.9 16.01 6.05 8 10 9.11 19.49 5.19 4.47 9.44 2.49 8 10 8.74 16.08 3.67 3.23 12.03 4.40 30 10 15.38 15.66 0.14 5.03 14.91 4.94 30 10 13.57 16.95 1.69 6.59 6.50 0.05 30 10 12.7 16.24 1.77 4.54 12.47 3.96

during steam explosion.

320 Biofuels - Status and Perspective

explosion.

**M.C. (%)**

**R.T. (min.)**

**Tempt. ( oC)**

Lignocellulosic biomass has been identified as a potential feedstock for the biofuel industry. Quantitation of lignocellulosic biomass components (lignin, cellulose, and hemicellulose) is often performed using the traditional acid hydrolysis followed by gravimetric determination. This approach is complicated and time consuming. FTIR-PAS was used in light of the need for rapid analysis of biomass materials and wood-based materials at large. The samples were initially pretreated using RF-alkaline and steam explosion techniques and analyzed gravimet‐ rically using the traditional approach to elucidate compositional information. Thereafter, the effect of the pretreatment conditions on barley straw grind was further analyzed based on their FTIR-PAS spectra. In order to develop a predictive model that will be swiftly used for the quantitative prediction of the chemical composition of the biomass, reference materials: pure cellulose (microcrystalline powder), hemicellulose (xylan from birch wood), and lignin (hydrolytic) powders were mixed in different proportions with known concentrations. The reference materials were used to generate standard spectra to determine the relationship between the respective quantity of components in the mixture and the FTIR spectra of representative biomass sample. The FTIR wavenumber-dependent instrumental effects were corrected by using carbon black reference spectrum. Multiple linear regression models for cellulose, hemicellulose, and lignin were developed based on the generated regression parameters, with coefficient of determination 0.68, 0.34, and 0.31, respectively, and mean square error of 9.10, 13.86, and 15.86, respectively. This study reflects that pretreatment can also lead to the degradation of the energy potentials. However, non-treated biomass resulted in the least conversion yield during the enzymatic hydrolysis [63]. The FTIR-PAS technique has advantage because it is a quick, easier, and non-destructive method. The structure of the biomass is maintained when spectra are measured directly from the bulk of grind biomass surface. Consequently, this study has led to the following conclusions: i) Lignin matrix structurally disrupted and released during pretreatment process; ii) Pretreatment enhances the accessibility and digestibility of the cellulose and hemicellulose; iii) This increased the conversion rate and assisted in reducing the costs and amount of enzymes required for the next stage of process (enzymatic hydrolysis) by 64% and 33% for RF and SE pretreatment, respectively. This implies that the PA infrared spectra can be used for biofuel feedstock identification and analysis of the chemical composition of biomass before it is processed. This innovative approach could be easily adopted by the biofuel industry and extended to any form of lignocellulosic biomass feedstock.
