Sustainable Valorization of Wood Residue for the Production of Biofuel Materials Via Continuous Flow Hydrothermal Liquefaction

When forests are extensively cleared for infrastructure and agricultural purposes, over half of the wood generated is regarded as waste. To minimize the negative consequences and successfully use renewable energy sources, further applications must be investigated. Biomass has the potential to be converted into biofuels using a process known as hydrothermal liquefaction (HTL). The conversion of wood residue under continuous flow, subcritical HTL, under both untreated and alkaline pretreated conditions was examined in this study. According to the results, pretreatment with 4% NaOH sped up recovery times compared to using raw wood and increased glucose production by 1.8 times (equivalent to 90 g/L) in comparison to using raw wood. Additionally, the fluid's pH was raised from 6.1 to 3.5 due to the alkaline pretreatment, which also switched the hydrolysis of arabinose and cellobiose to rhamnose at 0.2 MPa. The average Net Energy Ratio (NER) for glucose during liquefaction reached values as high as 63%, while the energy output from glucose during the pretreatment reached 246 kJ. The move to green energy transition toward net zero will be aided by these findings, which will usher in new waste conversion techniques to produce sustainable biofuels using continuous flow HTL.


Introduction
Despite the implementation of preventive measures by local governments in the Borneo region such as No Deforestation zones, the island continues to experience substantial deforestation activities. For instance, approximately 6 million hectares of forest were cleared between 2000 and 2017 to make room for palm oil plantations and other agricultural and infrastructure projects as well as logging of tropical trees, specifically the prevalent Dipterocarpaceae [1][2][3][4].
The low grading process of transforming this type of wood into timber deems most of it as waste [5]. Although the impact of harmful practices is clear, there is a chance to mitigate the overall effects on climate and local society by valorizing the lignocellulosic material into high value products such as biofuel materials, thus giving rise to a sustainable, circular economy system [6,7].
Nevertheless, the conversion of lignocellulosic feedstocks into biofuel materials can be achieved using various techniques such as torrefaction, liquefaction, gasification and pyrolysis [3]. Hydrothermal liquefaction (HTL), a method that works well with wet biomass and may produce liquid and solid fuel with low oxygen concentration, has attracted the attention of the scientific community, policy makers, and energy providers [8][9][10][11][12]. HTL can be used in many reactor configurations but continuous flow HTLs are considered as the next stage in the industrialization of the production of renewable energy. Continuous flow HTL has the capacity to create vast quantities of aqueous streams which are appealing due to the high content of fermentable sugars. Through further fermentation the aqueous phase can be converted into valuable products. However, lignocellulosic biomass exhibits a resistance to decomposition under HTL conditions. For instance, under normal operating circumstances increased temperatures and lengthy retention durations are used to produce the large sugar yields required for the manufacturing of biofuels [13,14]. The increased temperature drawbacks are two fold, firstly, higher temperatures enhance cellulose and hemicellulose degradation as well as the hydrolysis of biomass into inhibitors that adversely influence enzyme fermentation in downstream processing and secondly, the energy consumption associated with high input to convert the biomass in HTL is not sustainable [15,16]. Lignocellulosic biomasses contain carbon, hydrogen, oxygen, and small amounts of metal ions in their chemical composition. Lignin, hemicellulose, and cellulose, which have a tightly packed three-dimensional ultrastructure, are produced when these constituents are combined. The cellulose and hemicellulose polymers are joined to the lignin by covalent connections, like ɑ-ether bonds [17]. Plant cell walls gain strength and rigidity from lignin, a structural polymer [18]. Prior disruption of lignocellulosic biomass (through delignification) can increase the yields of fermentable sugar by exposing the cellulose and hemicellulose to water at the critical temperature, which in turn can reduce inhibitors by shortening the exposure time. Additionally, it can also reduce energy consumption by lowering the conversion temperature.
There are four basic ways to start the delignification process: physical, chemical, biological, and mechanical. In an effort to boost yields and lower inhibitors in batch reactors, an amalgamation of the HTL and delignification procedures has been tried. Using coconut pulp as an example, Mariano et al. [23] combined acid hydrolysis (121 °C, 0.1 MPa, and 20 min) with low severity hydrothermal treatment (100 °C for 60 min in a batch reactor). From 2.67 wt% HTL alone to 12.7 wt% for the combination treatment, the study found a fourfold increase in sugar monomers. The production of xylooligosaccharides, xylose, and glucose from maize stover was studied by Zhang et al. [22] using a combination of ball milling, ultrasonication, and hydrothermal processing. In comparison to the control sample, the yields for the samples that were ball milled for five hours and subjected to an hour of hydrothermal treatment at 230 °C were increased by 37.5% for xylooligosaccharides and 2.1% for xylose. In another study, after being exposed to an alkaline solution, acacia wood was hydrothermally processed in a stirred batch reactor at 200 °C for 10 to 30 min by Lee and Yu [19]. They claimed that 20% more glucose is produced following enzymatic treatment when alkaline pretreatment is used. In a related investigation, Kaur et al. [20] found that 90 °C treatment of a castor residue and NaOH mixture in an incubator for up to three hours enhanced the output of bio-oil. HTL conditions, such as 60 min time and various temperatures were then applied in a stirred batch reactor. To generate usable bio-oil, additional purification and separation techniques were needed because the batch reactor produced mixed chemicals. The primary constraints of these studies are the high energy consumption brought on by the biomass's resistance to decomposition and the alkaline slurry's further treatment. The return of fermentable sugars combined with degraded materials in the aqueous solution is the second factor, which is associated with the use of batch reactors.
Despite significant advancements in HTL and delignification pretreatments, the majority of studies focused on employing alkaline treatments after the hydrothermal processes. Others treated the biomass with an alkaline solution while it was still in the reactor and then moved the final feedstock to batch processing. However, hydrothermal liquefaction of destabilized biomass at low temperatures is required to offer a sustainable conversion of the waste wood. Furthermore, having the option to harvest aqueous phase at desired retention period is beneficial to avoid inhibitors or to collect high yield fluid only. In order to close the gap, the wood waste in this research is pretreated in an alkaline solution before being exposed to subcritical water conditions in a continuous flow HTL. The intention is to look at how chemical processing affects the lignocellulosic biomass's ability to liquefy and to observe the glucose yields required for further conversion into biofuel.
The output of this work will benefit local community through the sustainable integration of wood waste from development projects with a continuous flow HTL conversion technology by receiving valuable products in the form of biofuel materials. In a similar manner, wood waste remnants from various geographical places could also be converted into biofuel components. We also use this wood with the intention of contributing to the scientific literature providing a benchmark for valorization of other lignocellulosic biomasses in a continuous flow HTL. Additionally, research projects requiring continuous process monitoring and control might benefit from data generated by this system configuration as a source of technological breakthroughs.

Wood Samples
Wood residue from the Dipterocarpaceae family, Dipterocarpus caudatus (also known as Keruing bulat) was received from the Brunei Timber Sawmill (Tutong, Brunei Darussalam). The wood was air dried, milled using a Wiley mill and sieved through 30-100 mesh size to separate particles between 150 and 500 mm. Although not a process requirement, the samples were air-dried in the oven at 70 °C for 48 h prior to chemical treatment, hydrothermal liquefaction, and analysis. This step was necessary to ensure measurement consistency by using the oven-dry weight. In practice, air drying the biomass in the oven can be disregarded by accounting for the moisture content (MC) through mathematical means (Eq. 1) [21].

Chemical Pretreatment
Big efforts and associated costs has seen low efficiency returns during physical, physicochemical, and biological approaches [22,23]. The least expensive and most energyefficient option, however, is the chemical treatment [24]. Chemical agents that can create the alkaline conditions include calcium hydroxide, sodium sulphite and sodium hydroxide (NaOH). The former two are rarely mentioned in scientific literature whereas the latter is commonly used and it is the predominant method in the biorefinery industries [25,26] . In the presence of NaOH, the delignification of the biomass happens via the solvation and saponification reactions. At ambient temperature, lignin has a mild polarity in water. Consequently, the deprotonation of phenolic OH-groups causes the solubility to be higher in an alkaline solution [27]. The disadvantage of alkaline pretreatment is that it solubilizes some hemicellulose and alters the crystallinity of the cellulose, which could lead to some loss of useful compounds. By processing the lignocellulosic biomass at low alkali concentrations, the losses are reduced [28].
Therefore, the lignin linkages, phenyl glycosides, benzyl ethers, and gamma esters from the cellulose and hemicellulose were broken down by alkaline pretreatment. A weight of 5.0 g of Dipterocarpus caudatus were submerged into a 250 mL Erlenmeyer flask loaded with 167 mL of 4% NaOH (diluted from sodium hydroxide 32% in aqueous solution) (VWR, Singapore) for 3 h at room temperature. After the pretreatment time, the solution was filtered through a 0.3 μm filter paper and the biomass was neutralized to pH 7.0 by flushing with distilled water (Whatman, Cytiva). Approximately 1.5 L of distilled water was used to neutralize the biomass from pH 14.

Hydrothermal Liquefaction
The hydrothermal process engaged in this study was a continuous flow reactor equipped with a high-pressure pump, a pre-heater, and a heated reactor cell, both controlled by PID controllers. The reactor cell consisted of a 316 stainless steel cylindrical tube (18.0 mm ID) with a 7 mL volume, end capped with 100 microns stainless steel sintered mesh filters. Flow path was from bottom up, where the cylinder had a 4.0 mm (OD) aperture for inlet and the 18.0 mm opening for outlet. As the medium flowed through the pressurized system, the pre-heater matched its temperature to the reactor cell. One advantage of this setup over other reactor configurations, such as batch autoclave or stirred reactors is the reduced heating up time. Under swift heating rates, the biomass stability is maintained whereas in slow ramp up heating rates, the biomass experiences repolymerization and cracking. In the continuous flow reactor setup, we collect the aqueous fluid at desired retention times, therefore facilitating in-line separation of extracted solvents. An image of the hydrothermal liquefaction system used in this study is shown in Fig. 1. Four experiments were conducted where 1.8 g of sample was loaded in the reactor cell for each test. The sample consisted of raw wood residue (control sample) and alkaline pretreated wood residue. The 1.8 g weight filled up the reactor cell. Deionized water was used as a reactant at subcritical conditions, at a temperature of 210 °C and individual pressures which we will be referred to hereafter as low for 0.2 MPa and high, for 6.0 MPa, respectively. The fluid flow through the system was set to 10 mL/min. The system rampup time to reach 210 °C from 23 °C was 18 min and 35 s. Once the temperature reached the setpoint, the collection of the aqueous fluid was carried out. Six samples were collected using a systematic random sampling method for 60 s each.

Fermentable Sugars Compound Analysis
Aqueous samples collected from the hydrothermal experiments were pre-processed by centrifugation through 0.2 mm membrane (Nanosep, Pall Corp.). The concentration of fermentable sugars derived from the combined alkaline pretreatment and HTL was detected using a high-performance liquid chromatography (HPLC) system (Nexera, Shimadzu Corp., Kyoto, Japan) equipped with a refractive index detector (RID-20A). The column oven was fitted with a silica column (Luna 5 µm Silica (2) 100 Å, Phenomenex, USA, length: 250 mm, I.D.: 4.6 mm). This column is recommended for separating polar compounds. The mobile phase used was type 1 quality ultra-pure water freshly dispensed from a water system (Puris, Turkey). The flow rate was set to 0.6 mL/min, and the oven temperature was maintained at 40 °C. The injection volume was 10.0 mL. The sugars and other by-products cleared off the column after less than 8 min. However, under the current procedure, a total run time of 20 min was applied to ensure the complete identification and removal of all components from the column.
Reference samples of fermentable sugar solutions of 1 mg/mL of ultra-pure water were prepared for arabinose, cellobiose, galactose, glucose, mannose, rhamnose, sucrose, xylose (Nacalai Tesque, Kyoto, Japan), and fructose (Univar, Ajax Finechem). These were scanned to determine the retention times and the concentration slope through the silica column. All the reference sugars had unique retention times, except for xylose and mannose, which shared the exact value of 5.711 min.

Aqueous Fluid pH
The pH value is vital for subsequent biofuel production processes. Feeding medium within specific pH ranges reduces the need for adjustments, increasing process simplicity, and cost-effectiveness. The pH of the aqueous solution collected for 60 s interval was measured at room temperature using a benchtop pH meter (Fisherbrand accumet AB150 pH).

Chemical Properties by FTIR -ATR Spectroscopy
The spectra of the solids were collected at room temperature using an Agilent Cary 630 Fourier transform infrared (FTIR) spectrometer equipped with attenuated total reflectance (ATR) within the 4000-650 cm −1 range, at a resolution of 4 cm −1 . Solid samples were pressed against a diamond crystal with a torque knob for uniform pressure distribution. Peak identification was carried out offline using OriginPro 2021b. The applied procedure was baseline mode maximum constant, no baseline treatment, peak filtering by height, and the auto threshold was set to 20%.

Degree of Crystallinity by XRD
X-ray diffraction (XRD) was used to determine changes in the crystallinity of the raw, chemically treated, and solid Fig. 1 A view of the continuous flow hydrothermal liquefaction system. The HTL is built around a stainless-steel skid with the liquefaction part at the top and the system pressurizing part at the bottom, highlighted in red. The latter is only used at the beginning of the experiment to pressurize the system 1 3 residue samples. XRD experiments were performed using an XRD-7000 under normal atmospheric conditions, equipped with a copper X-ray tube, with a wavelength l (Ka1) = 1.5406 nm radiated at 40 kV voltage and 30 mA. Diffraction intensities were measured between 5° to 50° (2θ) using a step size of 2° per minute and a sampling pitch of 0.02°. Crystallinity area was measured using OriginPro 2021b. The degree of crystallinity can be derived from various equations. The Segal crystallinity, for example, considers only the cellulose intensity represented by the highest spectrum peak [29]. The Rietveld powder diffraction method, considers the entire biomass crystallinity. Since the biomass has an amorphous XRD spectrum, Rietveld powder diffraction is more suited for our application (Eq. 2) [30,31].

Energy Efficiency Analysis
Performance analysis of the aqueous product resulted from the hydrothermal liquefaction from pretreated and untreated lignocellulosic biomass was carried. The potential of the energy output from the compounds identified was calculated using Eq. 3.
The benefit of the energy output from the hydrothermal liquefaction process of a particular sugar compound was articulated as a function of Net Energy Ratio (NER). We discarded the energy input during heat up and system pressurizing, as with prolonged retention times their contribution is trivial. As such we focused on the energy input and the yield during the liquefaction stage only using Eq. 4.

Aqueous Analysis
It has been acknowledged that multiple complex reactions occur during hydrothermal liquefaction of lignocellulosic biomass, leading to different by-products. Lignocellulosic biomass decomposition pathways are dependent on operating conditions such as retention time, temperature, and pressure. Carefully manipulating them can control the (2) % Crystallinity = Area under the crystalline peaks Area under all peaks × 100 release of various compounds of interest. In this study, we have focused solely on the assessment of fermentable sugars that are highly important for biofuel production. The solid assessment was done with the scope of understanding changes in the biomass structure resulting from the delignification and hydrothermal treatment.

Compounds Identification Based on Retention Times
Applying low-pressure hydrothermal conditions decomposed the raw wood into arabinose and cellobiose as fermentable sugars. The values for arabinose ranged between 0.02 to 0.04 wt%, peaking at t 1 , whereas for cellobiose, 0.01 wt% at t 4 Fig. 2a.
The lignin limited the water's access to the hemicellulose and cellulose polymer chains. On the other hand, increased pressure led to decomposition into fermentable sugars such as glucose, xylose, or mannose. The latter was released within the first two measured retention times, whereas the former was recorded at t 2 until the end of the experiment Fig. 2b. This indicates that pressure is an important parameter in destabilizing and controlling the release and conversion of valuable products. Process conditions for biomass fractionation are commonly summarized in terms of severity factors. While this approach might be helpful in batch reactors, we showed above that pressure plays an important role in the conversion process. For example, the equations for severity factors in [32] only account for the time and temperature and do not consider the pressure parameter. The application of low-pressure hydrothermal conditions to the alkaline pretreated samples generated rhamnose with a value of 0.02 wt% Fig. 2c. In this instance, there is a different decomposition path compared to the sample in its raw state. Alkaline pretreatment changes the conversion pathway of lignocellulosic biomass; however, the low alkali concentration was not sufficient to return high amounts of fermentable sugars. The absence of arabinose and cellobiose seen in the control sample could be due to the partial removal of hemicellulose. On the other hand, increased pressure led to hydrolysis of the alkaline pretreated lignocellulosic biomass straight into glucose, at t 0 , 0.16 wt%, increasing to 0.85 wt% at t 1 and steadily decreasing afterwards Fig. 2d. Compared to the raw wood experiment, this method did not produce xylose or mannose and it generated glucose from t 0 onwards. This could be a result of further destabilization of the biomass structure by increased pressure and the complete exposure of the cellulose to the process conditions. The total estimated values of sugar released were calculated by finding the slope between two points interval (y = mx + b) and estimating for the time intervals (x) in between the time series based on the graphs above (y-y 1 = m(x-x 1 )) ( Supplementary  Information, Table S 1). The standard deviation was derived from all total estimated values Fig. 3.

Acidity Measurements of Aqueous Fluid
The pH value affects bacterial contamination, determines criteria selection for yeast fermentation, and influences hydrolysis pathways, fermentation rate, and by-product formation. Generally, medium pH values between pH 3.0 and pH 5.0 provide the optimum conditions for biofuel production [33][34][35]. Deionized water used in the decomposition of the samples had an initial pH value of 6.0 at room temperature. It was found that the majority of the data points fluctuate between pH 3.5 and pH 4.5. One obvious outlier is the hydrothermal liquefaction of raw wood at low pressure, which fluctuates from 8.7 at t 0 to as low as 4.3 at t 3 , having an average pH of 6.1 Fig. 4. The higher pH values occur due to lignin decomposition. It is a complex decomposition process, and it is unclear which compounds contribute the most [36].
In contrast, the reduced pH values result from hemicellulose decomposition and partial hydrolyzation of cellulose. The lower pH values could be attributed to the pentoses presence such as xylose or mannose and rhamnose. When only glucose (hexose) was found in the aqueous fluid, the pH value was higher compared to the pentosehexose mix. Longer retention times tend to lower pH value even more. This could be due to further hydrolyzation of glucose into furfural, 5-hydroxymethylfurfural or through the formation of organic acids such as acetic acid, via cleavage of the acetyl groups [37,38]. It is worth mentioning that at increased pH values, the aqueous fluid had shades of brown. As the pH valued dropped, the appearance of the aqueous fluid turned translucent.

Visual Analysis
The effect of each treatment and process visibly changes the color shades of the solid biomass and residues. The raw wood started with a light brown color Fig. 5a. After the chemical treatment, it changed its color to dark brown Fig. 5d. Exposure of both to the hydrothermal liquefaction at 210 °C and 0.2 MPa showed a similar shade of brown Fig. 5b and e. A further increase in pressure resulted in a transformation from brown to black, with a darker color for raw wood Fig. 5c and f.

Chemical Structure
The alkaline treated solid of the Dipterocarpus caudatus overall resembles the spectrum shape of its untreated version of each other. The small changes in functional groups are primarily due to the stability of the phenolic skeleton of lignin [39]. Offline peak identification (Table 1) revealed that the most significant effect the alkaline treatment has on raw wood is around the triple bond region, where the dissolution of the vibrational stretching at ~ 1718 cm −1 occurred. This bond is responsible for the ketone/aldehyde, C=O stretch in the hemicellulose [40]. The bands at ~ 667, ~ 1103, and ~ 3150 cm −1 have also dissipated. In contrast to the dissolution of some bands, the chemical-treated wood showed the formation of new functional groups, especially at the stretching vibrations of ~ 1264 cm −1 , which is thought to be responsible for the syringyl ring breathing and C-O stretching in lignin and xylan [41]. A wavenumber backsliding from raw wood at ~ 1029 cm −1 , down to ~ 1018 cm-1 for chemically treated wood is also observed. These vibrations are responsible for the aromatic C-H in-plane deformation bonds. This change is due to the increase in the reduced mass of molecules and lower force constant [42].
After hydrothermal liquefaction of raw wood Fig. 6b, the solid residue from low-pressure conditions (0.2 MPa) showed only minor changes in the transmittance percentage (%T), while the sample processed at higher pressure sample at ~ 1029 cm −1 %T = 74, down to 72% at 0.2 MPa and 66%, at 6.0 MPa). In addition, the vibrational stretch at ~ 1718 cm −1 did not appear, indicating that the C=O double bond was broken.
The hydrothermal conditions applied to the chemically treated wood Fig. 6c influenced the fingerprint region (650-1500 cm −1 ), especially in the lower vibrational frequencies Table 1. Similar to the processing of raw wood, a significant effect is a reduction in transmittance percentage at ~ 1029 cm −1 . Additionally, there are minor vibrational stretches of the C-H in the single bond region (2500-4000 cm −1 ).

Degree of Crystallinity
Similar with the characterization by FTIR of the chemical structure, we characterized the solid material before and after the chemical treatment and the hydrothermal liquefaction step. We sought to observe the changes in the crystallinity index of the overall biomass and correlate them with the hydrolysis efficiency especially. The X-ray diffraction showed that the wood material had an amorphous structure with a broad peak between 10° and 30° (Fig. 7), attributed to the amorphous carbon [46]. The samples crystallinity changed based on the treatment and process conditions applied. Table 2 highlights the degree of crystallinity values for each sample. The Dipterocarpus caudatus sample was highly amorphous in its raw state, with approximately 5% crystallinity. Exposing the samples to 4% NaOH for three hours increased the crystallinity percentage by over 4%. Hydrothermal liquefaction of raw wood under low pressure did not influence the crystallinity to a great extent as the solid residue maintained its crystallinity below 6%. However, an increase in the pressure conditions doubled the crystallinity percentage. Hydrothermal liquefaction had a more significant impact on alkaline treated wood than it did on raw wood. The degree of crystallinity for the alkaline pretreated biomass went as high as 25% at 0.2 MPa and 15% at 6.0 MPa.

Relationship Between Sugar Yields and Biomass Residue
Based on this data we can derive that the alkaline pretreatment has a significant impact when combined with HTL at high pressure conditions (6.0 MPa). Alkaline pretreatment almost doubles the amount of glucose hydrolyzed from lignocellulosic biomass in the interval tested, from 5 wt% for control sample to 9 wt% for the alkaline pretreated specimen.
Additionally, the alkaline pretreatment also speeds up the decomposition rate by bypassing the hydrolyzation of xylose or mannose. This is highly important for ethanol production in downstream steps since most microbial fermentation enzymes prefer hexoses, such as glucose, instead of pentoses, such as xylose and arabinose. Microbial hosts find it challenging to convert pentoses into ethanol; instead, genetically engineered yeast is used. The yields from pentoses, however, have not shown promising results [47].
The control sample released 1.07 wt% glucose at t 2 , equivalent of 7 min. The alkaline pretreated sample yielded 4.23 wt% glucose in the same time interval, overtaking the control sample yield after 3 min of retention time which showed a fourfold increased recovery rate. At t 3 , glucose production for control sample was 2.60 wt%, and the alkaline treated was 6.16 wt%, equivalent of 2.4-fold increase in recovery rate. The rate reduces to twofold and 1.7-fold at t 4 and t 5 , respectively.
The results from the analysis of the functional groups of solid residues strongly imply that glucose formation is dependent on breaking the C=O bonds of ketone/aldehyde at ~ 1718 cm −1 , the C-H bending vibrations occurring at ~ 700 cm −1 and ~ 840 cm −1 , the formation of C-O bonds at ~ 1055 cm −1 and reducing the intensity of the aromatic C-H bonds in the fingerprint regions (~ 1000-1100 cm −1 ). The solubilization of raw wood into arabinose and cellobiose is supported by the small changes in transmittance percentage and the changes in the out-of-plane bending vibrations of aromatic rings between ~ 600-900 cm −1 . On the other hand, the solubilization of the alkaline treated wood into rhamnose indicates that small changes in the C-H outof-plane vibrations with additional breaking effects on the single bonds of C-H stretching occurred. Solid residues from the control sample that resulted in xylose or mannose does not provide any extra structural information to determine which functional group is responsible for their release. In this case, the analysis of the solid residue after the short residence time is necessary.
The raw wood under high-pressure HTL returns more mixed sugar compounds with an acidic character. One interpretation of this finding is that the alkaline treatment Fig. 6 The transmittance spectra of a the raw wood and the sample after chemical treatment, b raw wood and its solid residues after hydrothermal liquefaction, and c the chemically treated wood and its solid residues after the hydrothermal liquefaction Table 1 shows the significant peaks and changes occurred as a result of the alkaline pretreatment and hydrothermal processing of raw wood. Comprehensive details are listed in Table S2 ----------1230 87  --1264 89  ----1264 87 1267 86 Ketone/aldehyde, C=O stretch [40] 1718 93 - [45] 3150 93 might have removed the hemicellulose simultaneously while aiming to destabilize the lignin. This effect is visible in the crystallinity measurements, where raw wood at low pressure did not significantly change its degree of crystallinity. In contrast, the alkaline treatment of wood returned the highest crystallization value of all tests. Because cellulose is not a branched polymer like hemicellulose or lignin, the exposure to high temperature increased the crystallization percentage [11,39].

Performance Analysis
The energy output potential from the glucose increased by more than 50% in energy yield, from 112.92 kJ for the control sample to 246.62 kJ for the alkaline pretreated sample, under the same HTL conditions. This in turn results in greater NER performance for glucose production. The energy requirements for the conversion of lignocellulosic biomass are detailed in Table 3.
Due to the nature of the continuous water flow of the system, it is possible to analyse the NER performance across time. From Fig. 8 it can be seen that alkaline pretreatment exhibits a NER value for glucose production at t 1 above 60%, which is reducing with retention time. In such cases, it would be ideal to maintain high NER performance collection of aqueous fluid between 2 to 7 min of retention time. During the HTL of the untreated biomass, the glucose conversion begins at t 1 onwards. Because to this delay in Fig. 7 XRD spectra of a the wood before and after the chemical treatment, b the raw wood and the solid residues resulted from the hydrothermal liquefaction and c the chemically treated solid and the solid residues from the hydrothermal liquefaction conversion, a mean NER below 20% is achieved. Our NER equation considers only the input (e.g., 571 kJ) and output during the liquefaction process. The NER presented takes into consideration the energy required to maintain the system at set temperature. This is an on/off process operated by the proportional-integral-derivative (PID) controller, which resets within ± 1 °C of the set point. Realistically, the NER is higher by 1/3 or more. The energy to heat up the system, a requirement that will become imperceptible with scale, was not considered an influential factor, although in the current set up it was only 663 kJ. Similarly, the energy required to pressurize the system, which happens as a one-off and is negligible (e.g., 10 kJ). No energy is required to pressurize to 0.2 MPa, and one-off energy input is only due to the temperature increase.

Efficiency Comparison of the Acid Pretreatment and Continuous Flow HTL and Other Recent Published Methods
It is important to evaluate the effectiveness of the current approach and acquire perspective on those of others that have been published. Since there is little research on continuous flow HTL, conventional batch HTL equipment is generally used in conversion procedures. The authors employ a variety of data interpretation methodologies concurrently, which renders an efficient comparison relatively neutral. Details of each interpretation is available Table S2 (Supplementary Information). In our investigation, we compared the energy ratios and yields of the ones provided by us with those found in recent literature. For instance, Zhuang et al. [48] looked into batch HTL of lignocellulosic biomass using microwaves. It revealed that the energy ratio of biocrude increased by 0.4 value points, from 1.69 to 1.73, when compared to standard HTL during 60 min retention time and 240 °C. Under typical HTL circumstances and with microwave assistance, the comparable ratio for a retention time of 16 min would be 0.45 and 0.46, respectively. About 10% of the biocrude in both treatments contained hydrocarbons; as a result, it may be estimated that the energy ratio from carbohydrates like glucose was less than 0.05%.
A low-lipid marine (Tetraselmis sp.) was examined by Aljabri et al. [49] using a typical batch HTL method with a 60-min retention time and 350 °C. Carbohydrates recovered 10% of their initial energy after 10 min, 15% after 20 min, and 29% after 30 min. At 10 min, the NER in our approach is higher than those reported, being 47.15 percent for the pretreated alkaline biomass and 30.4 percent for untreated biomass (Fig. 8). Taghipour et al. [50] investigated the impact of recycling the aqueous phase in a catalytic HTL using this Spirulina as a feedstock. The energy consumption recovery (ECR), which ranged from 38 to 80%, was recorded under various temperature settings and catalytic use. We did see, however, that when reporting the ECR, the authors did not take the heating-up/cooling cycle into account. Hong et al. [51] observed an energy recovery rate of 49%, increasing with the cycle numbers up to 65% using a similar method and evaluating penicillin residue. The energy recovery rate increased with the addition of the catalyst, indicating 63% without any cycle but dropping to 55% after three cycles. Similar to Taghipour et al. [50], their equation calculation does not account for the re-heating of the reactor brought on by aqueous collection cycles. Using several catalysts in conjunction with HTL, Ding et al. [52] studied a variety of biomass feedstocks. The energy recovery values ranged from 24% for starch to 22% for cellulose, tripling with the inclusion of Co-Mo catalyst. The authors combined highly pure biomass with specific catalysts. Because low processed biomass feedstocks have compositional variation, this method needs a more practical proof of concept study, which is what our work offers. Sequential HTL treatments, where the sugar molecules were recovered after the initial operational phase, is one of the other methods for biomass pretreatment. For instance, Gu et al. [53] looked at two different temperature scenarios for the conversion of poplar wood chips, 140 °C and 180 °C. Yields of 0.5% and 3% glucose were reported. At these low temperatures, it is therefore doubtful that recovering fermentable sugars will result in high yields during the first stage. In a related investigation, Gu et al. [54] explored the pretreatment of poplar wood by fungal treatment. This method raised the glucose yields to 5% and, by adding fungal manganese feed, reached glucose yields of 55%, which was higher than what was attained with our method. However, there is no mention of the energy requirements related to fungal inoculation. We may infer that the evaluation of a biomass conversion system is somewhat dependent on the feedstock employed, in addition to the system setup. According to Lee and Yu [19], who used acacia wood as feedstock, alkaline post-treatment increases glucose yields following enzymatic fermentation by 52% to 73%. While Kim & Um [55] showed that alkaline pretreatment of Korean native kenaf and miscanthus did not result in any yield increase, we reported an increase in yields from Dipterocarpus caudatus. As opposed to that, it decreased the yield by 2 and 3 percentage points. According to Mariano et al. [23], coconut pulp can yield over 60% hexose sugars. When the biomass structure is light, more ecologically friendly processes like ball milling and ultrasonication [22] may be used to disturb the cellulose covering layers.

Conclusion
In conclusion, our research shows that pretreating lignocellulosic biomass in alkaline solution before hydrothermal liquefaction in continuous medium flow promotes compound valorization (e.g., glucose) at an early stage, shortens processing durations, and streamlines the energy-intensive processes of HTL. The neutralization of the biomass following the alkaline pretreatment is one area that could benefit of improvements. We flushed it with distilled water, but it might still produce a lot of waste water. In the future, water recirculation systems or titration studies may be used to enhance waste management.
Additionally, we showed that pressure has a significant impact on the conversion process. In this way, separate compounds can be exploited, such as rhamnose if low pressures (0.2 MPa) are utilized. With no prior biomass treatment but high pressure (6.0 MPa) it is possible to hydrolyze xylose and mannose within the first 6 min of retention period, switching to glucose thereafter. Conversely, low pressure conditions can yield sugars such as arabinose and cellobiose. From a functional perspective, this system outperforms the traditional batch processing, allowing for early separation of compounds and a decent NER, with possibility to increase the glucose energy return by enzyme fermentation thereafter. Furthermore, stronger delignification through longer soaking times or biomass homogenization could improve the NER values, generating biofuel materials, such as fermentable sugar more sustainably.
The local community will profit from the work's outcomes by acquiring valuable products in the form of biofuel materials through the sustainable integration of wood waste from development projects with a continuous flow HTL conversion technology.