Corn Stalks-Derived Hemicellulosic Polysaccharides: Extraction and Purification

Abstract

Nowadays, agricultural biomass is one the most valuable sources of natural polysaccharides. In addition to primary agricultural goods, agricultural waste is abundant, diverse, and renewable and can also be utilized as raw material for the production of polysaccharides and their derivatives. The extraction and purification of agri-waste-derived polysaccharides involves multiple processes that can vary depending on the type of raw material and the specific polysaccharides targeted. This study proposes a particular pathway from corn waste to hemicellulosic polysaccharides, which involves alkaline treatment and several physicochemical separation/purification phases using precipitation and ion exchange resins (Purolite A400, Purolite A100+, Purolite C100H). The ion exchange separation stage was optimized to retain most of the acid-soluble lignin derivatives from the extraction liquors. The process parameters considered for optimization included the solid (resin) liquid (black liquor pH 4.5) ratio, contact time, and temperature. These ranged from 0.05 to 0.15 g·mL⁻¹, 30 to 180 min, and 20 to 50 °C, respectively. The chemical composition of the separated hemicelluloses varied from 44.43 to 75.28% for xylan, 2.43 to 3.93% for glucan, 1.86 to 2.44% for galactan and 8.93 to 12.68% for arabinan. The total carbohydrate content increased from 57.65 to 96.3%.

1. Introduction

In 2011, the European Polysaccharide Network of Excellence (EPNOE) made some predictions regarding the main challenges and openings in polysaccharides research and technology considering both research and education. The targeted areas, for which the application possibilities of polysaccharides and polysaccharide derivatives were anticipated, were bio-based materials, food, and medicine/pharmaceuticals [1]. As anticipated by this EPNOE report, the primary research directions over the last ten years have been: (i) finding new and suitable sources of polysaccharides [2,3,4]; (ii) developing effective separation/fractionation processes [5,6]; (iii) improving characterization techniques; and (iv) discovering novel applications in a variety of fields. In an attempt to cover the future development of polysaccharides as renewable bio-resources, EPNOE recently published a much larger report [7] that emphasizes the value of professional education in the field of biopolymers and polysaccharides, in addition to some very important technological aspects like separation, purification, characterization, and applications.

It is commonly known that processing lignocellulosic biomass presents a number of difficulties, partly because of the chemical composition’s variability but also because of the relationships between cellulose, hemicellulose, and the lignin–carbohydrate complex, or LCC [8]. Given the nature of the chemical bonds in LCC, breaking these bonds is a crucial step in the effective separation of biomass components [9].

Our previous studies fall within the trends anticipated by EPNOE reports. The possibilities of using widely available, highly renewable, and mass-produced agricultural residues such as wheat straw and/or corn husks as sources of hemicelluloses, papermaking fibers, and lignin were studied [10,11]. Hot alkaline extraction (HAE) and soda pulping (SP) were identified as the most appropriate extraction methods for xylan-based hemicelluloses and the procedure was further optimized [10]. Special attention was paid to improving pretreatment development and optimization [11], sequential fractionation to obtain multiple products [12], and to improve the technological aspects of mass balance [13]. In our most recent study [13] we demonstrated that with the use of HAE/SP, corn stalks (CS) can yield up to 90% of lignin and 60% of hemicelluloses (HC), but only half of the lignin can be recovered. Similarly, the yield for hemicelluloses recovery can only reach 15% to 20%, of which up to 75% represents polysaccharides-based products (PBP)—mainly xylan. Therefore, it is critical to identify efficient methods for raising the yield of HC and lignin extraction and purification from spent liquors. This may improve the economic efficiency of using CS as a raw material for PBP while also partially addressing an environmental issue, as the residual liquid produced by chemical pulping of lignocellulosic biomass, known as black liquors (BLs), is a serious environmental threat [14,15].

BLs are produced in massive amounts in the pulp and paper industry [16,17] and usually present a complex chemical composition (sodium salts of organic acids, dissolved lignin, HC, and others) that is heavily influenced by the nature of the raw materials and the pulping process parameters [18,19,20]. Lately, BL has been acknowledged as a raw material for the production of lignin, hemicelluloses, and value-added chemicals [17,21,22]. However, due to its high complexity in chemical composition, both processing and chemical characterization pose significant challenges [16]. Generally, two precipitation stages are needed to separate the desired components from the BLs. The first one involves the precipitation of lignin through acidification, followed by its separation through centrifugation. The remaining supernatant (as such or concentrated) goes through a second stage of precipitation where non-solvents are used. The precipitated polysaccharides (HC) can also be separated by centrifugation. Since a part of the lignin or lignin-derived compounds are soluble in acidic media [16,23], the pH value is a critical parameter that influences both the yield and purity of recovered lignin and PBP [24,25,26].

Although the use of anion exchange resins in the purification of liquid streams resulting during the processing of lignocellulosic biomass is fairly well documented, the majority of the papers focus on the purification of acidic hydrolysate. Chen et al. (2016) discuss the purification of hot water extracts with pH values ranging from 3.5 to 5.5 obtained from corn stalks at different process times and temperatures using Amberlite IRA900, a weak base anion exchange resin. According to their study, the efficiency of lignin removal under the working conditions reached 85% [27]. Another Amberlite variety, IRA 400, was used by Dalli and co-workers to remove inhibitors such as furfural from acidic hydrolysate of poplar biomass [28]. By employing sequential cationic-anionic exchange resin treatments, Kumar et al. were able to remove phenols, 5-hydroxymethylfurfural, and nitrate salts with relatively high efficiencies (~70%). Improvements in the fermentation process are then reported by the study [29]. In 2022, Han et al. reported even higher efficiencies (>90%) when they removed lignin, acetic acid, and furfural from corncob hydrolysate (pH-1.5) using ion exchange resins [29].

Enzymatic processing of lignocellulosic biomass may be a great alternative for the extraction and recovery of polysaccharides because enzymes can handle the recalcitrance of lignocellulosic biomass [30] and help isolate hemicelluloses. Enzyme application has been shown to enhance the separation of lignin from BL [31]; however, to the best of the author’s knowledge, its effect on the separation of HC from raw BL has not been documented.

In line with the most recent EPNOE predictions, this work aims to improve the separation, purification, and characterization of PBP derived from CS, by proposing an innovative sequence of technical operations involving HAE/SP, precipitation-centrifugation, and ion exchange/adsorption accompanied by specific characterization methods (HPLC, FTIR, UV-Vis and others). The objectives of the study are: (i) chemical analysis of liquid streams and solid materials involved in PBP production by HAE/pulping using CS as raw material, (ii) separation, purification, and chemical characterization of the targeted components—lignin, HC, PBP, (iii) testing and optimization of anion exchange resins and activated carbon for advanced component separation from BL, and (iv) performing a comprehensive mass balance to highlight the advantages of using CS as a raw material in biorefinery when an advanced recovery of lignin and HC is attained.

To the best of the authors’ knowledge, no study has employed ion exchange resins in this manner to purify the liquid streams resulting from the alkaline treatment of lignocellulosic biomass.

2. Materials and Methods

2.1. Materials

Corn stalks were collected free of charge from Romanian farmers in the autumn of 2023. The CS (free of leaves and any other components) was examined to observe any rotting parts (which were removed) and further cut into smaller pieces about 20–30 mm long. The initial conditioning also included natural drying from a moisture content of around 15% to an equilibrium moisture of about 9%. To be used for chemical analysis, stalks were supplementarily ground and sieved through a 1 mm sieve.

Two types of anion exchange resins, Purolite A400 (R1) [32] and Purolite A100+ (R2) [33], and one type of cation exchange resin, Purolite C100H (R3) [34], were supplied by Ecolab (Victoria/Brașov, Romania)Purolite S.R.L Romania as free-of-charge samples for evaluation purposes.

The adsorption experiments were performed using irregular shape particles of activated carbon (AC) supplied by Buzău Romcarbon Company (Buzău, Romania).

Purified xylan extracted from beechwood was purchased from Sigma Aldrich (X4252 10G, St. Louis, MO, USA) and was used as reference material. Solutions of 99% purity of cellobiose, glucose, xylose, galactose, and arabinose, provided by Flucka, were used to obtain the calibration curves in the concentrations range of 0.1–1 g·L⁻¹.

ChemPUR Gmbh (Karlsruhe, Germany) provided acetic acid with a purity of 98% and ethanol with a 95% analytical purity.

2.2. Soda Pulping—SP

The SP trials were performed in a stainless steel, pressurized, rotating (15 r.p.m.), and temperature-controlled laboratory autoclave. The alkali charge during SP experiments was 20% NaOH, and the pulping temperature was maintained at 150 °C for 40 min. At the end of the pulping, performed at a solid to liquid ratio of 1:10, the reactor valve was opened to release the accumulated pressure (0.4 MPa) and collect the produced soda pulp black liquor (SPBL). The obtained SPBL was first filtered through a G2 crucible to remove any remaining solid biomass material. The subsequent calculations were performed with the assumption that all of the black liquor had been recovered. The produced pulp was washed and saved for further chemical composition analysis.

2.3. Separation by Precipitation

2.3.1. Lignin Separation

Acetic acid was used to treat the SPBL to lower its initial pH from 12.9 to 4.5, which allowed for the initial lignin separation. In this way, the dissolved lignin is precipitated and further removed by a centrifugation stage of 10 min at 3000 rpm. After centrifugation, the soda-pulping black liquor supernatant) with a pH of 4.5 was collected (SPBL4.5. The resulting raw lignin precipitate was stored for later chemical analysis to determine its composition, including its lignin, ash, and carbohydrate content.

2.3.2. Hemicelluloses Separation

The ethanol precipitation method was used to separate the HC. Before the HC separation, the SPBL4.5 solution was concentrated using vacuum distillation in a rotary evaporator system. The solution volume was reduced to one-third of its initial value. The SPBL4.5, which resulted after concentration, was mixed with two volumes of ethanol (p.a.) and kept at −18 °C for 60 min. The obtained mixture was then centrifuged for 10 min at 3000 rpm to separate the HC. The obtained HC samples were washed twice with fresh ethanol and further dried in a laboratory-ventilated oven at 40 °C. A similar procedure for HC separation was performed using the SPBL4.5 after the ion exchange/adsorption treatments.

2.4. Ion Exchange/Adsorption Treatments

2.4.1. Anion Exchange

The anion exchange treatments aiming to reduce the amount of acid soluble lignin (ASL) in SPBL4.5 were optimized using the Response Surface Methodology (RSM). The trials were carried out under static, batch mode conditions. The process parameters (independent variables) considered for the RSM experimental design were: (i) the resin charge (solid/liquid ratio) ranging from 0.05 to 0.15 mg·mL⁻¹, (ii) temperature, ranging from 20 to 50 °C, and (iii) treatment time, ranging from 30 to 180 min. The removal efficiency (system response) was calculated using Equation (1):

$$E(\%)={\displaystyle \frac{{C_{iASL}-C}_{fASL}}{C_{iASL}}}·100,$$

(1)

where C_iASL represents the concentration of ASL in untreated SPBL4.5, and C_fASL represents the concentration of ASL after the anion exchange treatment. The ASL concentrations was determined using Lambert–Beer law, monitoring the UV absorbance at 320 nm and an ε value of 30 L·g⁻¹·cm⁻¹ [35].

2.4.2. Activated Carbon (AC) Adsorption

AC adsorption was performed for comparison purposes, to evaluate the anion exchange purification efficiency. The AC treatment was performed at 20 °C at a solid-to-liquid ratio of 0.1 g·mL⁻¹ for a total time of 90 min.

2.4.3. Cation Exchange

The cation exchange treatment was performed, with the aim of removing the minerals from the HC samples. In brief, the treatment procedure consisted of dissolving the previously obtained HC samples in ultrapure water up to a concentration of 15 g·L⁻¹ dry substance. The resulting solutions of HC were treated by using the R3 for 90 min at room temperature (25 °C) and a liquid-to-solid ratio of 0.1g·mL⁻¹.

2.5. Chemical Analysis of Liquid Streams and Solid Materials

The polysaccharides and lignin content from solid samples of CS and CS pulp were determined by acid hydrolysis as mentioned in Ref. [36]. Based on the methodology described by Sluiter and co-workers, the only neutral sugars that can be identified using the available equipment are those that are shown in this investigation. It was necessary to lower the pH of the hydrolysate in order to preserve and prevent damage to the HPLC column. Since CaCO₃ and BaCO₃ facilitate the removal of organic acids as salts, attempts to use them as neutralization agents failed. The inability to fully eliminate the sulfate anions, which are also harmful to the column and result in strong co-elutions, restricting the scope of additional research. Using diluted sulfuric acid as an eluent and specialized columns made to function in acidic environments (HPX-87H), other authors [37] were able to identify additional sugars and uronic acids. Due to similar equipment limitations, they were better at separating derived acids from sugars but less effective at separating neutral sugars.

Following the complete hydrolysis of the samples, the acid insoluble lignin (AIL) was quantified as a solid remaining on the G3 crucible. At the same time, the ASL was determined after measuring the 320 nm absorbance of the filtered hydrolysate. Samples of the same hydrolysate were neutralized to pH 5.5–6 and after 0.22 µm filtration, HPLC was employed to determine the monosaccharides content. Further calculations based on the concentrations of mono-sugars, sugar recovery standards, and monomer-to-polymer conversion coefficients were made to determine these samples’ glucan, xylan, galactan, and arabinan content. Separation of simple sugars was achieved on a Shodex (SP0810 sulfo Pb²⁺) column heated at 65 °C while the RI detector was set at 40 °C. The HPLC analytical system (Shimadzu Nexera) was operated at a flow rate of ultrapure water of 0.6 mL/min. The solid samples of lignin and separated HC were treated similarly, accepting excepting the initial hydrolysis step at 30 °C, which was replaced with a simple dissolution/dispersion in distilled water. Additionally, a solution of 72% sulfuric acid was added, and the full hydrolysis of carbohydrates was carried out at 121 °C.

The liquid samples (SPBL, SPBL4.5) were analyzed for lignin and carbohydrates according to method described in Ref. [35]. At the same time, the ash or inorganic materials were determined by combustion of CS, CS pulp, lignin, and HC samples at 550–600 °C following the specification in Ref. [38].

The FTIR spectra of HC-separated samples were recorded using potassium bromide disks containing finely ground samples of HC on an Agilent Cary 630 FTIR instrument Agilent Cary 630 (Santa Clara, California, USA) (64 scans at a 4 cm⁻¹ resolution and 4000–400 cm⁻¹).

HC color value (CV) was determined after HC samples dissolution in a 0.5 M NaOH solution up to a concentration of 1.5 g·L⁻¹. The solutions were then filtered (0.45 μm syringe PTFE Roth filter) followed by absorbance measurements at 420 nm against a 0.5 M NaOH solution [39]. Absorbance measurements were performed using a Jasco V530 UV-VIS spectrometer (Tokyo, Japan). To calculate the CV, Equation (2) was used:

$$\mathrm{C}\mathrm{V}={\displaystyle \frac{{\mathrm{A}}_{420}}{\mathrm{b}·\mathrm{C}}},$$

(2)

where A₄₂₀ denotes the absorbance at 420 nm, b is the optical path length, and C is the sugar concentration in g·L⁻¹.

All the experimental results are displayed as a means of at least three determinations unless other specifications were provided by standard analytical procedures. A maximum accepted RSD value of 5% was accepted.

3. Results and Discussion

3.1. Chemical Composition of CS, CS Pulp, and Solid Product Separated from SPBL4.5

Table 1. Chemical composition of raw materials and obtained pulp and SPBL4.5 precipitated lignin.

Sample Type	Glucan (%)	Xylan (%)	Galactan (%)	Arabinan (%)	CH Tot (%)	AIL (%)	ASL (%)	Ash (%)
Raw CS	35.65	18.42	0.85	2.69	57.61	25.36	1.14	3.81
CS pulp	54.94	19.27	0.77	2.09	77.08	5.83	0.71	0.15
BLPP	1.31	12.42	1.42	4.27	19.42	50.93	2.56	15.68

CH tot = total carbohydrates, AIL = acid insoluble lignin, ASL = acid soluble lignin.

shows the chemical compositions of unprocessed CS, CS pulp obtained after SP, and black liquor precipitate (BLPP) separated by centrifugation after BL pH reduction. (oven dried—o.d.).

The cellulose content of the CS, reported as glucan, is comparable with literature data stated by various authors, ranging from 32% to 45%: 38.7% [40]; 36.89% [41]; 39.0% [42]; and 44% [43]. As for our previous studies, we reported CS cellulose contents of 37.6% [13]; and 39.7%], respectively, 43.7% [10]. Hajkova and co-workers reported a very high amount of cellulose (47.6%) in corn plant residues, although their study does not provide sufficient information regarding which parts of the plant were used [44]. The HC of corn stalks are of a xylan-type structure with branches of arabinose and galactose anhydrous residues. The raw CS had an amount of about 22% HC, as revealed by the sum of xylan, galactan, and arabinan contents. The value falls in the range of data reported in the literature, ranging from 20% to 30%: 29.65% [45]; 21–25% [46]; and 19.64–27% [10]. The lignin content of the unprocessed CS (Table 1) is higher than the values reported by the referred studies: 15.5% [45]; 4.9–8.7%], respectively, 20–23% [10]. The ash content is lower in comparison with our previous studies [12,13] but falls within the range and is close to those reported by other authors [47,48]. As pointed out by Lizzote and co-workers, the chemical composition of CS is linked to the variety of the cultivated corn, growing conditions (soil, climate), and time of harvest [47].

The SP process generates a series of changes in the chemical composition of the solid material (Table 1). The total pulp yield initially observed of 36.7% led to a sorted pulp yield of 35.6%. This is significantly higher than the value reported by Hajkova and co-workers [44], although these authors reported a very high cooking time at a temperature of 160 °C. When comparing the percentages of HC and, in particular, xylan, in raw CS and CS pulp (Table 1), the variation appears insignificant; however, when the calculations are applied to the initial amount of xylan in CS, the variation is 53%. The other HC polysaccharides removal rate also recorded values in the range of 59–65%. The carbohydrate content increases in CS pulp (77.08%), compared with the raw CS (57.61%), the most significant variation being registered for cellulose content 54.94% vs. 35.65%. When referring to the initial amount in raw CS, the CH tot suffered a negative variation of 40%. When compared to the initial amount in raw CS, the total carbohydrate content experienced a 40% decrease. The AIL percentage dropped from 25.36% in CS to 5.83% in CS pulp, which corresponds to a lignin removal rate of nearly 90%, while the ASL removal rate touched 72%. As for the ash, more than 98% of the initial amount was removed during SP treatment.

The black liquor precipitate (BLPP), resulting after pH adjustment of the SPBL to 4.5 with acetic acid, was separated by centrifugation, dried to a relative moisture content of 10%, and further used for chemical analysis. The AIL represents slightly over half of the BLPP (50.93%), while the ASL is 2.56%. The amount of carbohydrates in the BLPP was 19.42%, of which xylan constituted over half (~64%). The remaining main components of the BLPP are arabinan (~22%), galactan (7.3%), and glucan (6.8%). The high ash content suggests that some components of the SPBL are precipitated as sodium salts.

3.2. Chemical Composition of SPBL and SPBL4.5

Following the pulping procedure, the generated SPBL was extracted, examined, and used once more to recover carbohydrates. When the SPBL is acidified to pH 4.5, lignin and, to a lesser extent, other constituents like polymeric carbohydrates, precipitate together.

Table 2. Chemical characteristics: SPBL vs. SPBL4.5.

Sample Type	pH	C (mS·cm−1)	DM (g·L−1)	IM (g·L−1)	OM (g·L−1)	Glucan (g·L−1)	Xylan (g·L−1)	Galactan (g·L−1)	Arabinan (g·L−1)	CH (g·L−1)	L (g·L−1)
SPBL	12.9	44.1	78.74	30.45	48.29	1.15	9.90	1.15	0.31	12.60	20.28
SPBL4.5	4.5	25.4	69.83	23.43	46.41	0.55	7.21	1.07	0.24	9.07	8.37

C = conductivity, DM = dry matter, IM = inorganic matter, OM = organic matter, CH = carbohydrates, L = lignin.

displays some of the SPBL’s physio-chemical properties as well as how they altered after the pH adjustment and centrifugation.

High pH and high conductivity values, which are caused by the presence of sodium ions (residual sodium hydroxide and/or different organic acid salts formed during delignification), are typical characteristics of black liquors. When the pH is adjusted from the initial 12.9 to the required 4.5, the conductivity drops by roughly 42%. Moreover, the pH modification induces a reduction in dry matter content (11%), inorganic matter (23%), and organic matter (3.9%).

The amount of dissolved lignin decreases by 59% compared to a 27% decrease in CH concentration. The amounts of glucan, xylan, arabinan, and galactan in the acidified black liquor decrease by 36.3%, 27.7%, 22.6%, and 6.9% when compared to the initial quantity in SPBL.

3.3. Anion Exchange Treatment Optimization

As stated in the materials and methods section, this treatment aimed to eliminate the non-sugar-derived components that were present in the SPBL4.5. The treatment efficiency was determined using Equation (1) based on the evolution of the global parameter ASL. Resin charge [49,50], temperature [51,52], and contact time [53,54] are typically the three main factors that affect the efficiency of ion exchange treatments and constant initial concentrations of ions to be removed. The goal of the optimization study was to ascertain the collective and individual effects of these process variables on the treatment’s effectiveness. As shown in

Table 3. Experimental design and results for anion exchange resin treatment optimization.

Exp. Run	X1-Resin Charge (g·mL−1)	X2-Temperature (°C)	X3-Contact Time (min)	E R1 (%)	E R2 (%)
1	0.10	35	105	73.9	80.3
2	0.05	50	30	43.7	46.4
3	0.15	50	30	56.1	65.4
4	0.15	35	105	88.0	86.5
5	0.10	35	105	73.9	79.1
6	0.05	50	180	72.1	76.1
7	0.10	35	30	43.1	61.8
8	0.15	20	30	65.6	76.0
9	0.10	20	105	71.6	82.4
10	0.05	20	30	46.2	40.6
11	0.10	35	105	73.9	80.7
12	0.10	35	105	72.7	79.6
13	0.05	20	180	51.4	72.8
14	0.10	35	105	76.6	79.1
15	0.10	35	180	77.8	87.1
16	0.15	50	180	92.8	95.2
17	0.10	50	105	80.7	87.4
18	0.10	35	105	73.9	79.8
19	0.15	20	180	89.6	93.7
20	0.05	35	105	62.7	62.6
21	0.05	50	105	66.9	70.1
22	0.05	20	105	58.5	62.1
23	0.15	20	30	79.0	89.4
24	0.15	50	105	87.5	91.3

, a set of 24 experiments (including 5 replications at the center point) were statistically programmed for each anionic resin using the design of experiments approach (DOE).

Regarding efficacy, the two resins show relatively comparable outcomes, with R1 showing an efficiency range of 43.1% to 92.8% and R2 showing an efficiency range of 40.6% to 95.2%. However, the efficacy evolution for the R1 and R3 resins exhibits some slight variations when individual parameters are considered, as illustrated in Figure 1, Figure 2 and Figure 3.

The findings of the simplified ANOVA analysis, which are shown in

Table 4. Synthetic ANOVA analysis results presentation.

	Model F Value	Adequate Precision	R2	R2 Adjusted	Predicted R2
Model 1 (R1)	39.6	23.8	0.94	0.92	0.83
Model 2 (R2)	71.5	31.2	0.96	0.95	0.89

, show that the models that were produced are adequate for navigating the design space. The two models appear to be accurate based on their F-values. The “Adequate Precision” value, which is greater than 4, as well as the higher-than-0.9 values for R² and R² adjusted provide additional support for the models’ precision. For both models, there was a correlation observed between the adjusted R² and the predicted R² values.

Table 5. Coefficients of the model equations.

	β0	X1	X2	X3	X1X2	X2X3	X12	X32
Model 1 (R1)	6.3	231.0	0.0918	0.409	−5.16	0.00512	972	−0.002
Model 2 (R2)	−8.03	648.0	0.224	0.380	−2.67	0.00108	1580	−0.0013

displays the coefficients for the model equations that show how the chosen process parameters affect the resin treatment’s efficiency in removing ASL. The general form of the model is given by Equation (3). Several factors were removed to simplify the model with the condition to preserve its hierarchy.

$$Y={\beta }_0+{\displaystyle \sum _{i=1}^k}{\beta }_i{X}_i+{\displaystyle \sum _{i=1}^k}{\beta }_{ii}{X}_i^2+{\displaystyle \sum _{i<j}}{\beta }_{ij}{X}_i{X}_j+\epsilon$$

(3)

Examining how the investigated variables affected the final E (%) reveals that, in both scenarios, the resin charge has the biggest influence on the effectiveness of the ASL removal. Thus, the elimination of ASL is positively impacted by an increase in resin charge. Nonetheless, it appears that the impact of resin charge is even greater in the case of R2 (second order vs. first order for R1). The temperature has an unexpectedly low, but positive, effect on the E%. It is important to note that the temperature range was constrained to maintain the resin’s thermal stability. The impact of contact time is depicted in Figure 1, Figure 2 and Figure 3. The linear and relatively rapid progress of the ASL removal efficiency for nearly 100 min is followed by a plateau region where the E% growth is considerably slower and non-linear.

The values of the process parameters that, according to the mathematical models, produce the highest E% values, are shown in

Table 6. Predicted and experimentally validated optimal conditions for ion exchange removal of ASL from SPBL4.5.

Resin Type	X1—Resin Charge (g·mL−1)	X2—Temperature (°C)	X3—Contact Time (min)	Predicted E (%)	Experimental E (%)
R1	0.1	50	167	73.3	70.2
R2	0.1	50	180	80.2	79.6

, which also presents the results of the experimental validation. There is a reasonable error between the predicted and actual values, as the experimental validation shows.

3.4. Hemicelluloses Characterization

3.4.1. UV Analysis and Color Value of Isolated Hemicelluloses

In Figure 4, the UV spectra of the obtained HC samples are compared with those of the pure xylan XSA. All samples, including XSA, displayed shoulders of absorbance in the range 270–290 nm (1.12 at 270 nm and 1.86 at 280 nm, respectively).

The absorbance values in the mentioned range result from the presence of lignin and derivatives [55,56]. In the case of all CS HC samples, shoulders may be observed at 300–330 nm and as well at 360–380 nm, which are also attributed to the presence of lignin. Both of these are a consequence of the presence and alkaline ionization of various forms of phenolic hydroxyl groups of hemicelluloses containing lignin [57]. The main observation on these shoulders is the reduction in their intensity as a result of the ion exchange treatment.

The color value (CV) is an important feature of HC materials, and determining it provides a rapid way to assess the presence of contaminants.

Figure 5 shows that the color values of each hemicellulose differ from one another. The HC SPBL4.5 sample had the highest CV, almost 80% higher than the commercial xylan XSA. Aside from ASL, several other chemical components resulting from the Maillard reaction may contribute to high absorptivity in the 400–450 nm region and, consequently, high CV values [39,58,59]. The ion exchange/adsorption treatments on SPBL4.5 reduced the CV of all separated HC. R1 treatment resulted in an 80% decrease in the CV for the corresponding HC, whereas R2 treatment resulted in a 60.6% reduction in the CV. The CV was further reduced by a supplementary treatment with the cationic resin R3 (Figure 5).

3.4.2. FTIR Analysis of Isolated Hemicelluloses

The FTIR analysis of the obtained CS hemicelluloses samples yielded the spectra shown in Figure 6. All the samples analyzed displayed specific absorption bands at ~898 and ~1162 cm⁻¹ that result from the stretching of C-O-C β-(1-4) inter-anhydro-xylose units glycosidic bond [60]. The bands occurring at ~1043 cm⁻¹ and ~1060 cm⁻¹ were assigned to various C–O bond stretching [61]. Several other peaks that occur in the range 1100–1400 cm⁻¹ may be assigned to both C-H vibrations and O-H vibrations [62]. The band at ~1412 might be a consequence of the symmetric stretching of carboxylates [63]. Carboxylates also cause broad absorption in the range 1630–1600 cm⁻¹ as a result of antisymmetric stretching. The presence of absorbed water that generates absorption bands by H-O-H bending vibration at ~1640 cm⁻¹ might be overlapped by other functionalities [64]. FTIR spectra of samples obtained after both anion and cation exchange treatment displayed bands at ~1720–1730 cm⁻¹ that suggest the presence of C=O functionalities [65]. FTIR spectra also displayed large absorption bands at ~3400–3500 cm⁻¹ corresponding to O–H bonds stretching vibration and also in the range 2950–2850 cm⁻¹ corresponding to C-H stretching vibrations.

3.4.3. Chemical Composition of the Isolated Hemicelluloses Samples

The chemical composition of the isolated HC, presented in

Table 7. Chemical composition of the obtained hemicelluloses.

Sample Type	Glucan (%)	Xylan (%)	Galactan (%)	Arabinan (%)	HC Tot (%)	AIL (%)	ASL (%)	Ash (%)
HC_SPBL4.5	2.43	44.43	1.86	8.93	57.65	0.35	2.54	11.96
HC_CS AC	2.97	56.71	2.78	12.12	74.58	0.55	1.01	11.15
HC_CS R1	3.75	60.89	2.55	10.98	78.17	0.25	0.75	10.14
HC_CS R2	3.53	64.46	2.19	11.00	81.18	0.22	0.55	10.03
HC_CS R1 R3	4.96	74.36	3.03	13.95	96.31	0.29	0.78	0.15
HC_CS R2 R3	3.93	75.28	2.44	12.68	94.34	0.27	0.82	0.13

CH tot = total carbohydrates, AIL-acid insoluble lignin, ASL-acid soluble lignin.

, was established based on the HPLC results. Some chromatogram examples are presented in the Supplementary Materials in Figures S1–S5. The HC separated from the SPBL4.5 designated HC_SPBL4.5 (Table 7) was used as a reference for the other samples. The HC obtained after SPBL4.5 treatment with activated carbon, Purolite A400, Purolite A100+, Purolite C100H were denoted as HC_CS AC, HC_CS R1, HC_CS R2, and HC_CS R3. The samples with the labels HC_CS R1 R3 and HC_CS R2 R3 were subjected to sequential anionic and cationic treatments.

The predominant type of polysaccharide in all the studied HC samples is xylan. The polymeric chain of xylan also includes various branches consisting mostly of arabinose anhydro-units but also of glucose and galactose. Under the study’s conditions, the chemical composition of about 28% of the reference sample was not determined. Nevertheless, certain hypotheses concerning these 28% could be developed in light of the literature reviews. Based on the research of Sun and co-workers [66], Huan and co-workers [67], and Dafchahi and Acharya [68], it is likely that 4-O-methyl-D-glucuronic acid, acetic acid and methyl alcohol residues are the main components that have not yet been reckoned.

In the reference sample (HC_SPBL4.5 in Table 4), the xylan represents 44.43%. Both adsorption and ion exchange treatments increase xylan content: 27.6% for AC, 37% for R1, and 45% for R2. The addition of the R3 treatment (cationic exchange) to the R1 and R2 treatments (anionic exchange) resulted in an increase in xylan content of 67% and 69.4%, respectively. In comparison to the reference, treatments containing AC, R1, R2, and R1R3, and R2R3 sequences increased the purity (CH tot) of the HC samples by 29.4%, 35.6%, 40.8%, 67.1%, and 63.6%, respectively.

3.5. Mass Balance

Figure 7 depicts the corn stalks soda pulping process flowsheet, including component recovery from BL, as well as some quantitative data on potential products generated in a CS soda pulping biorefinery approach. The mass balance calculations were carried out with an initial quantity of 100 g of CS. The soda pulping of 100g CS leads to an amount of 36.7 g of CS pulp and 26.5 g of raw lignin (see Figure 7). Papermaking is one of pulp’s possible applications, as we have already mentioned in earlier work [12]. In the meantime, lignin can be utilized for a number of purposes, such as energy valuation, chemical synthesis, or even as an antioxidant.

Several calculations were performed using the chemical composition of the initial CS and the resulting pulp (presented in Table 1).

It is noteworthy that approximately 63% of the original hemicelluloses are removed during the pulping process (13.84 g as an absolute value). During the pulping progression, nearly 10% of the extracted carbohydrates are degraded within the mass of the reaction (12.5 g remained in 1 L of BL). The precipitation of lignin from BL after pH reduction to 4.5 results in an additional drop of ~34.7% in carbohydrate concentration. A further decrease of about 34.7% in the concentration of carbohydrates is caused by the precipitation of lignin from BL after the pH is lowered to 4.5. Consequently, the amount of polysaccharides that can be recovered by ethanol precipitation is reduced to 9.07 g per 1 L of SPBL4.5. That represents nearly 65% of the HC removed during pulping and 41.3% of the original HC content of CS. Some mass balance results and recovery yields are presented in

Table 8. Values of the recovery yield of HC obtained during the study.

Sample Type	Raw HC (g/100 g CS)	Pure HC (g/100 g CS)	RY CS (%)	RY BL (%)	RY SPBL4.5 (%)
HC_SPBL4.5	12.3	7.09	32.29	56.68	78.18
HC_CS R1	10.9	8.52	38.80	68.11	93.94
HC_CS R2	10.2	8.28	37.71	66.19	91.29
HC_CS R1 R3	8.9	8.57	39.03	68.52	94.50
HC_CS R2 R3	8.6	8.11	36.95	64.85	89.45

, where RY stands for recovery yield. RY CS is reported to the initial amount of HC in raw oven-dry CS (Equation (4)), RY BL (Equation (5)) is reported to the amount of HC in the BL, and RY SPBL4.5 is reported to the amount of HC in SPBL4.5 (Equation (6)).

Table 8 displays some mass balance findings along with the corresponding recovery yields; RY stands for recovery yield. The calculations for RY CS refer to the initial amount of HC in raw CS, for RY BL to the amount of HC in the BL, and RY SPBL4.5 to the amount of HC in SPBL4.5.

$$RYCS\left(\%\right)={\displaystyle \frac{m_{pHC}}{m_{CS}·{HC}_{tot}(\%)}}·100$$

(4)

$$RYBL\left(\%\right)={\displaystyle \frac{m_{pHC}\left(BL\right)}{V_{BL}·C_{CHtot}}}·100$$

(5)

$$RYSPBL4.5\left(\%\right)={\displaystyle \frac{m_{pHC}\left(SPBL4.5\right)}{V_{SPBL4.5}·C_{CHtot}}}·100$$

(6)

where m_pHC is the mass of pure HC (g) recovered from the oven-dried amount of CS (m_CS), which produced a certain amount of HC (HC_tot); m_pHC (BL) is the mass of pure HC (g) obtained by processing a certain volume of BL (V_BL, in L), having a concentration of carbohydrates C_CHtot (g·L⁻¹); and m_pHC (SPBL4.5) is the amount of pure HC (g) obtained by processing a volume of SPBL4.5 liquid V_SPBL4.5 (L) having a concentration of carbohydrates C_CHtot (g·L⁻¹).

4. Conclusions

The black liquor resulting as a byproduct of corn stalk lignocellulosic biomass soda-pulping can be regarded as an important source of hemicellulosic polysaccharides. The first step in producing high-purity HC samples was to remove lignin from BL using acidic precipitation. The pH adjustment to 4.5 induces lignin precipitation, but it also causes a loss of carbohydrate content, as nearly 20% of the CHtot is lost by mixing with the lignin precipitate. A significant amount of the original SPBL lignin (the acid-soluble lignin—ASL) was still found in the supernatant SPBL4.5. A second precipitation step with ethanol resulted in hemicelluloses samples containing 57.65% CHtot. To lower the AIL from the SPBL4.5, anion exchange resin treatment was also employed. In this regard, the RSM was used to study and optimize the Purolite A400 and Purolite A100+ (ECOLAB, Victoria/Brașov, Romania) treatment of SPBL4.5. The obtained quadratic dependency models revealed the optimal conditions in terms of resin charge, temperature, and contact time. The predicted values were confirmed through experimentation, yielding HC samples with a CHtot content of 78% and 81%, respectively. The analysis of these HC samples showed that they contained between 10% and 12% minerals. As a result, the supernatants from the anion exchange treatment received an additional treatment using cation exchange resin (Purolite C100H). The HC products that were subsequently precipitated with ethanol had a content of 94–96% CHtot and significantly less ash. Color value, FTIR, and UV analysis were performed to highlight the benefits of using ion exchange resins for HC recovery.

The best results in terms of HC recovery yield and purity degree were obtained by following this sequence: soda pulping, acid precipitation at pH 4.5, anionic exchange (Purolite A100+), first ethanol precipitation, cationic exchange (Purolite C100H), and second ethanol precipitation. The mass balance analysis provides significant quantitative data and indicates that resin treatment raises the HC recovery yields. Certain types of ion exchange resins that are targeted and specific to individual HC components may be designed and used to improve recovery efficiency.