The Effect of Acid Catalysis on Hydroxycinnamate Recovery from Corn Stover Using Hydrothermal and Organosolv Treatments

Abstract

Corn stover (CS) is a highly abundant type of agricultural biowaste, largely composed of lignocellulosic material. CS may be a particularly rich pool of hydroxycinnamates, represented primarily by p-coumaric acid and ferulic acid; yet, these compounds are bound onto the lignocellulosic matrix, and their release requires an appropriate acid and/or alkaline catalysis. This being the case, this study herein aimed to develop an effective process to boost hydroxycinnamate recovery by employing acid-catalyzed hydrothermal and organosolv treatments. To this end, oxalic acid was tested as a benign, natural acid catalyst, along with the well-examined sulfuric acid. A kinetic assay showed that both the acid catalyst and the use of an organic solvent (ethanol) may greatly impact the rate and level of polyphenol recovery. Under optimized conditions, determined by implementing response surface methodology, it was demonstrated that the organosolv treatment was far more effective than the hydrothermal one, with regard to total polyphenol recovery, while the oxalic acid catalysis was equally efficient as the sulfuric acid one. This treatment afforded 17.8 ± 2.3 mg gallic acid equivalents per g of dry CS mass. However, a thorough insight into the polyphenolic composition of the extracts produced revealed that hydrothermal treatment may enable, apart from p-coumaric and ferulic acid release, the formation of a compound tentatively identified as an ester of p-coumaric acid with a pentose. Furthermore, it was shown that sulfuric acid-catalyzed organosolv treatment provided almost 25 and 34% higher yields for p-coumaric and ferulic acid, respectively, but it strongly inhibited p-coumaric acid-pentose ester formation. These compositional differences appeared to impact the antioxidant activity of the corresponding extracts. It was concluded that the oxalic acid-catalyzed ethanol organosolv treatment of CS may have important potential in a biorefinery context, but improvements are required to further enhance treatment performance. This would lead to replacing corrosive catalysts, such as sulfuric acid, with benign ones, thereby establishing a fully sustainable process for the recovery of bioactive phytochemicals.

1. Introduction

The expansion of the world’s population and the concomitant intensification of agriculture and food production have led to overwhelming pressure on the environment. Both the increase in crop production and food manufacturing are major sources of biowaste generation, which is mainly rejected or harnessed for low-value activities, thus contributing to ecosystem degradation and the depletion of natural resources. Such a critical standoff results from insisting on a linear economy model, whereas the ongoing implementation of circular economy strategies provides viable solutions by establishing sustainable routes for bioresourse exploitation and agri-food production [1,2]. In this context, processing residues arising from the agricultural and food industries can be prime raw materials for the generation of novel commodities and chemicals, thus fueling, in a green way, the food, pharmaceutical and cosmetic sectors. By deploying biorefinery technologies, processing wastes rich in precious biomolecules can be effectively used for producing food additives, pharmaceutical formulations and active cosmetic ingredients, thus upcycling side streams and providing high-value-added bioactive compounds and platform chemicals [3,4].

Cereals constitute a large percentage of the global food industry, and cereal processing results in the generation of an enormous by-product volume. Corn (Zea mays) is a major cereal crop, widely used to produce human food but also animal feed and industrial products (e.g., cornstarch, alcohol, etc.). The global corn production was around 1.03 billion tons in 2017 [5], while in Greece, the average annual corn production for the period of 2002–2020 was estimated to reach about 2042 thousand tons [6]. Corn harvesting and processing inevitably generates very high amounts of by-products, with one of the primary by-products being corn stover (CS). This residue comprises stalks, husks and leaves that remain in the field after harvesting, and it is produced at a rate of 1 dry kg per dry kg of corn grain [7]. Thus, global CS production mounts up to almost 1 billion tons.

One of the principal biopolymers associated with CS is lignin, which occurs in CS at around 12% w/w, while cellulose and hemicellulose may occur at corresponding levels of about 35 and 20 w/w. Thus, taking into consideration its abundance as reported above, CS is regarded as one of the key lignocellulosic feedstocks for high-value-added chemicals, such as chemical building blocks and lignin-derived polyphenolic antioxidants [5]. In fact, lignin-associated hydroxycinnamates in CS include extensive amounts of pendant units, attached onto lignin mainly through both ester and ether linkages (~18% p-coumaric acid on a lignin basis) [8], and recent studies have shown that the major antioxidant hydroxycinnamates in various parts of CS are p-coumaric acid and ferulic acid [9]. These compounds may act as cross-linking agents between polysaccharide chains, and also between proteins and lignin, but they can also cross-couple with monolignols and be incorporated as co-polymers into lignins [10].

Owing to their multiple biological activities, hydroxycinnamates are considered as natural compounds of particular nutritional and pharmaceutical significance, and as abundant dietary antioxidants. The antioxidant activity of major hydroxycinnamates has been well documented, while several studies have been carried out to substantiate structure–activity relationships [11,12,13]. However, the biological importance of hydroxycinnamates is not limited to their antioxidant properties, since their beneficial influence on human health has been evidenced by a number of pertinent examinations [14,15,16,17].

The recovery of bound polyphenols from cereal tissues requires specific treatments that involve acid and/or alkaline catalysis, and the polyphenols released from such treatments depend on the type of catalyst (acid/alkaline) used. Thus, various hydroxycinnamate derivatives have been identified in maize extracts, obtained by applying acid or alkaline catalysis and variable treatment conditions [18]. Recently reported data demonstrated that sulfuric acid-catalyzed organosolv treatments of wheat bran may yield extracts with significantly differentiated compositions than those obtained with alkaline catalysis [19], affecting the antioxidant properties of the extracts generated. Furthermore, it was also shown that citric acid, a natural organic weak acid, may act as an effective acid catalyst in releasing hydroxycinnamate-derived compounds from wheat bran, using hydrothermal treatments [20].

On such a basis, this work was undertaken with the scope of testing both hydrothermal and ethanol organosolv treatments for their efficiency in releasing bound hydroxycinnamates from CS. In both cases, strong (sulfuric acid) and mild (oxalic acid) acid catalysis was used to examine their effect on both the treatment yield and the polyphenolic composition. To appraise treatment performance, the extracts obtained were put under scrutiny regarding their polyphenolic profile and antioxidant activity. To the best of the authors’ knowledge, such an approach to examine CS as a bioresource of polyphenolic antioxidants is reported for the first time.

2. Materials and Methods

2.1. Chemicals

The procurement of 2,2-diphenyl-1-picrylhydrazyl (DPPH) was from Alfa Aesar (Karlsruhe, Germany) and 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ) from Fluka (Steinheim, Germany). Sodium carbonate anhydrous was purchased from Penta (Prague, Czechia). L-Ascorbic acid was obtained from Carlo Erba (Milano, Italy). Iron chloride hexahydrate (FeCl₃), oxalic acid and citric acid anhydrous were from Merck (Darmstadt, Germany). Tartaric acid, Folin–Ciocalteu regent and absolute ethanol were from Panreac (Barcelona, Spain). Ferulic acid and p-coumaric acid were from Sigma-Aldrich (Steinheim, Germany). All solvents used for chromatographic determinations were of HPLC grade.

2.2. Collection and Handling of Corn Stover (CS)

Several units of CS, composed of dried residues of canes and leaves, were collected from a maize plantation located outside the town of Karditsa (Central Greece), in October 2023. This material was manually broken down to smaller pieces, freeze-dried overnight, ground in a laboratory mill, and sieved to give a powder with a mean particle size < 500 μm. This feed was maintained in air-tight containers, at 4 °C, and used in all treatments tested.

2.3. Hydrothermal and Organosolv Treatments

Dried and comminuted CS (2.5 g) was placed in a 100 mL screw-cap glass vial, along with 50 mL of either water (hydrothermal treatment) or 60% v/v hydroethanolic mixture (organosolv treatment), to provide a constant liquid-to-solid ratio of 20 mL g⁻¹. Both hydrothermal and organosolv treatments were conducted with variable acid catalyst concentrations. In the case of organosolv treatment, the ethanol proportion was also a subject of examination. For preliminary experiments, all hydrothermal treatments were performed for 180 min, at 90 °C, whereas in the case of organosolv treatments the temperature was set at 80 °C to avoid excessive vapor pressure due to the presence of ethanol. For the response surface optimization, residence time and temperature were dictated by the experimental design. Both heating and stirring (at 400 rpm) were accomplished by a temperature-controlled hotplate (Witeg, Wertheim, Germany). Upon the completion of each treatment, a suitable volume of extract was centrifuged at 10,000× g to separate cell debris, and the clear supernatant was used for all subsequent analyses.

2.4. Kinetics of Polyphenol Recovery

Polyphenol release kinetics was determined using a first-order model, as previously implemented [21], as follows:

Y_TP(t) = Y_TP(s)(1 − e^−kt)

(1)

The term Y_TP(t) represents the total polyphenol yield at any time t, Y_TP(s) is the total polyphenol yield at saturation (equilibrium), and k is first-order extraction rate constant (min⁻¹).

2.5. Determination of Severity Factor

Treatment severity was evaluated by determining the alternative combined severity factor, termed as CSF’, as previously proposed [22], as follows:

CSF’ = logR_o + |pH − 7|

(2)

where R_o is the severity, defined as [23]

$$R_{\mathrm{o}}=t \times e^{({\displaystyle \frac{T-100}{14.75}})}$$

(3)

The value 100 is the reference temperature (100 °C) and 14.75 is an empirical parameter associated with the treatment temperature and activation energy.

2.6. Response Surface-Based Treatment Optimization

Optimization was undertaken to identify the ideal combination of two pivotal variables, residence time, t, and temperature, T, on the treatment performance, as appraised by the response, which was the yield in total polyphenols (Y_TP). On this basis, these two crucial independent variables (t, T) were selected to set up the experimental design for the response surface methodology. The central composite design deployed encompassed 11 design points, containing 3 central points. Variable codification to three levels, −1, 0 and 1, was accomplished as previously described [24]. Variable levels in both codified and actual form are analytically displayed in

Table 1. The variables used in the response surface methodology, and their codified and actual levels.

Process Variables	Codes	Coded Variable Level
		−1	0	1
t (min)	X1	60	180	300
T (°C) (hydrothermal treatment)	X2	50	70	90
T (°C) (organosolv treatment)		40	60	80

.

The actual ranges of both variables were chosen based on a kinetic assay, but also considering previous results [25]. Mathematical model significance (R², p) and the significance of each of the models’ coefficients were evaluated by statistical analyses (lack-of-fit and ANOVA tests), taking into account 95% as the minimum significance level.

2.7. Treatment Yield Determination and Antioxidant Assays

The total polyphenol concentration in the extracts obtained was determined using a previously established Folin–Ciocalteu assay [26]. Gallic acid was used as the calibrating standard and results were expressed as gallic acid equivalents (GAE). Yield in total polyphenols (Y_TP) was given as mg GAE per g of dried CS. Both the antiradical activity (A_AR) and the ferric-reducing power (P_R) of the extracts were assessed using the protocols published elsewhere [27]. The results for the A_AR assay were given as μmol DPPH per g of dried CS, while the results for the P_R assay were given as μmol ascorbic acid equivalents (AAE) per g of dried CS.

2.8. Chromatographic Analyses

For the liquid chromatography–diode array–tandem mass spectrometry (LC-DAD-MS/MS) analyses, the equipment and the elution program employed have been previously described in full detail [28]. For the tentative peak identification, all mass spectra were acquired in negative ionization mode. For quantification, calibration curves of p-coumaric acid (R² = 0.9982) and ferulic acid (R² = 0.9980) were constructed using commercially available standards (0–50 μg mL⁻¹). Standard solutions were prepared in HPLC-grade methanol immediately prior to determinations.

2.9. Data Processing and Statistics

JMP™ Pro 16 software (SAS, Cary, NC, USA) was used to build the design of the experiment, to deploy response surface methodology and to compute the relevant statistics (ANOVA, lack-of-fit). Likewise, SigmaPlot™ 15.0 (Systat Software Inc., San Jose, CA, USA) was used to perform non-linear and linear regressions, at least at a 95% significance level. Since the data obtained were not normally distributed, as shown by the Shapiro–Wilk test, statistically significant differences were revealed by the Kruskal–Wallis test, using IBM SPSS Statistics™ 29 (SPSS Inc., Chicago, IL, USA). All treatments were carried out at least twice, and the chemical analyses (chromatographic, spectrophotometric) in triplicate. The values reported represent means ± standard deviation (SD).

3. Results and Discussion

3.1. Effect of Acid and Ethanol Concentration

The first stage in developing an effective medium for polyphenol recovery from CS was the appraisal of the effect of the acid catalyst. Based on a previous study, which demonstrated the efficacy of citric acid in liberating hydroxycinnamates (ferulic acid) from wheat bran [20], three organic acids were tested, namely citric acid (CiAc), tartaric acid (TarAc) and oxalic acid (OxAc). The selection of these acids was based on their natural occurrence and their strength, since all three possess a relatively low pK_a. Sulfuric acid (SuAc) was also considered, since it is a strong acid catalyst. Various concentrations of aqueous solutions were assayed, and the results were assessed by comparing the yield in total polyphenols (Y_TP). The outcome depicted in Figure 1 indicated that 5% OxAc outperformed all other media tested, and it was more effective even in comparison with SuAc (p < 0.05). Thus, 5% OxAc and 1% SuAc were chosen to carry out hydrothermal treatments.

The second stage was to evaluate the ethanol proportion in the treatment medium in order to select the most appropriate solvent for the organosolv treatment. Preliminary experimentation suggested 60% hydroethanolic solution to be the highest-performing mixture, and thus this water/ethanol proportion was selected to investigate the combined effect of ethanol and acid. Figure 2 illustrates the outcome of this assay and, as can be seen, a significantly higher Y_TP was attained with 60% ethanol/10% OxAc and 60% ethanol/1.5% SuAc (p < 0.05). Therefore, these two systems were used for deploying the organosolv treatment.

3.2. Polyphenol Recovery Kinetics

To better depict the influence of the type of acid catalysis and the role of ethanol in the recovery of CS polyphenols, a range of temperatures from 50 to 90 °C was used to trace kinetics (Figure 3). Using a first-order model, the extraction rate constant, k, and the yield in total polyphenols at saturation, Y_TP(s), could be determined (

Table 2. Extraction rate constant (k) values and yield in total polyphenols at equilibrium (Y_TP(s)) determined for the various hydrothermal and organosolv treatments carried out.

Treatment	T (°C)	R2 *	k (min−1) × 10−3	YTP(s) (mg GAE g−1 DM)
Hydrothermal treatment
1% SuAc	50	0.949	56.4	2.2 ± 0.1 a
	70	0.984	23.3	5.0 ± 0.2 b
	90	0.977	12.6	9.4 ± 0.5 c
5% OxAc	50	0.987	36.5	3.4 ± 0.1 d
	70	0.979	51.8	5.2 ± 0.2 b
	90	0.991	27.7	10.4 ± 0.7 c
Organosolv treatment
60% ethanol/1.5% SuAc	40	0.982	153.4	7.9 ± 0.3 e,g
	60	0.949	112.8	10.0 ± 0.6 c
	80	0.974	36.4	18.5 ± 1.2 f
60% ethanol/10% OxAc	40	0.999	205.2	7.5 ± 0.2 e
	60	0.957	107.0	8.5 ± 0.3 g
	80	0.976	24.7	16.5 ± 0.9 f

* Indicates correlation of the kinetic model fitting to the experimental data. Values denoted with different letters (a–g) have statistically significant difference (p < 0.05).

). For the SuAc-catalyzed hydrothermal treatment, the maximum extraction rate k (56.4 min⁻¹) was achieved at 50 °C, whereas a gradual decline in k was observed by switching the temperature from 50 to 90 °C. However, the maximum Y_TP(s) (9.4 mg GAE g⁻¹ DM) was attained at 90 °C. For the OxAc-catalyzed hydrothermal treatment, the maximum k (51.8 min⁻¹) was recorded at 70 °C, but it declined to 27.7 min⁻¹ at 90 °C. In this case too, the maximum Y_TP(s) (10.4 mg GAE g⁻¹ DM) was obtained at 90 °C and showed no significant difference (p > 0.05) with the one obtained with the SuAc-catalyzed treatment.

On the other hand, the SuAc-catalyzed organosolv treatment exhibited much higher k values compared with the corresponding hydrothermal one, at any temperature tested. Moreover, the maximum Y_TP(s) was 18.5 mg GAE g⁻¹ DM, which was almost double that achieved with the hydrothermal treatment. Similarly, the OxAc-catalyzed organosolv treatment displayed a higher k at both 50 and 70 °C compared with the hydrothermal one, but at 90 °C the k recorded was close to that seen in the hydrothermal treatment. The maximum Y_TP(s) in this case was 16.5 mg GAE g⁻¹ DM, which is almost 37% higher than that obtained with hydrothermal treatment. On the grounds of these results, it was clear that polyphenol extraction was greatly facilitated in 60% ethanol, showing higher release rates and higher recovery yields.

3.3. Treatment Optimization

The kinetic trials provided a valuable matrix of data, which formed the basis for treatment optimization by implementing RSM. The two independent variables considered were the two critical treatment parameters, residence time, t, and temperature, T, and Y_TP was chosen as the response. The experimental design used was a Box–Behnken with three central points, aimed at evaluating the effect of the treatment variables and identifying any synergistic functions between them. The assessment of model fitting and validity was accomplished by performing an analysis of variance (ANOVA) and lack-of-fit tests (Figures S1–S4), considering the closeness of predicted and measured response values (Tables S1–S4). The equations (mathematical models) derived from the RSM contained only the statistically significant terms (p < 0.05), as shown in

Table 3. Response surface-derived models describing the effect of independent variables (X₁, X₂) on treatment and the yield in total polyphenols (Y_TP).

Treatment	Equation (Model)	R2	p
Hydrothermal (5% OxAc)	YTP = 5.6 + 3.8X1 + 1.4X12	0.99	<0.0001
Hydrothermal (1% SuAc)	YTP = 4.6 + 3X1 + 0.6X2 + 0.7X1X2	0.98	0.0002
Organosolv (60% EtOH/10% OxAc)	YTP = 9.5 + 3.8X1 + 1.4X2 + 1.6X1X2 + 2.4X12	0.96	0.0016
Organosolv (60% EtOH/1.5% SuAc)	YTP = 10.4 + 4.8X1 + 1.6X2 + 1.7X1X2 + 2X12	0.98	0.0003

.

The square correlation coefficients determined (R²) were credible indicators of the total variability around the mean provided by the mathematical models (equations). Considering a confidence interval of 95% and taking into account the R² of the p value for lack-of-fit, it was concluded that the adjustment of the models to the experimental data was very satisfactory. Visual model representation was provided by 3D diagrams, which portrayed the concurrent effect of both treatment variables on the response (Y_TP) (Figure 4).

For the OxAc-catalyzed hydrothermal treatment, the time within the limits tested had a non-significant impact (p > 0.05), whereas temperature was the only determinant for controlling Y_TP, obeying a quadratic function (Table 3). On the contrary, for the SuAc-catalyzed treatment, t was also a significant factor, as was its cross effect with T (X₁X₂). For both organosolv treatments, it was found that cross effects, as well as the quadratic effects of T (X₁²), had a positive and significant effect on Y_TP.

By exploiting the desirability function (Figures S1–S4), the optimum values for both treatment variables (T, t), as well as the maximum response value (Y_TP), could be calculated (

Table 4. Optimized conditions, maximum predicted response (Y_TP) values and severity for the treatments tested.

Treatment	Optimal T (°C)	Optimal t (min)	Maximum Predicted YTP (mg GAE g−1 DM)	Severity (CSF’)
Hydrothermal (5% OxAc)	90	120	11.4 ± 0.9 a	7.03
Hydrothermal (1% SuAc)	90	240	9.5 ± 1.3 a	8.23
Organosolv (60% EtOH/10% OxAc)	80	240	17.8 ± 2.3 b	8.63
Organosolv (60% EtOH/1.5% SuAc)	80	240	21.0 ± 2.0 b	8.73

Letters a and b denote statistically significant differences (p < 0.05).

). The OxAc-catalyzed hydrothermal treatment afforded a maximum Y_TP of 11.4 ± 0.9 mg GAE g⁻¹ DM, which was virtually equal to the Y_TP achieved with the SuAc-catalyzed treatment. On the other hand, the OxAc-catalyzed organosolv treatment gave a Y_TP of 17.8 ± 2.3 mg GAE g⁻¹ DM, which was almost 1.6-fold higher than the Y_TP obtained with the hydrothermal treatment. Moreover, the difference in Y_TP between the OxAc- and SuAc-catalyzed organosolv treatments was non-significant.

To better illustrate the effect of acid catalysis on the polyphenol recovery from CS, treatments with neat water and 60% ethanol were also carried out, under optimal T and t settings (90 °C/180 min and 80 °C/300 min, respectively). The corresponding Y_TP values were 3.1 ± 0.1 and 8.3 ± 0.7 mg GAE g⁻¹ DM, a finding that pointed emphatically to the crucial role of the acid catalysts in boosting polyphenol recovery. The above findings highlighted the potency of OxAc in assisting both hydrothermal and organosolv treatments, since its presence was pivotal in acquiring maximum Y_TP. Furthermore, a comparison between hydrothermal and organosolv treatments revealed that ethanol also had a critical role, fostering significantly higher recoveries.

To further appraise the combined effect of T and t, but also to take into account the acidity of the treatment media tested, the alternative combined severity factor (CSF’) was also determined (Table 4). The OxAc-catalyzed organosolv treatment had a CSF’ value of 8.63, which was very close to 8.37, determined for the maximum Y_TP from wheat bran, using 10% aqueous citric acid [20], but higher than the 7.61 required in the OxAc-catalyzed ethanol organosolv treatment of coffee silver skin for effective polyphenol recovery [29]. On the other hand, the SuAc-catalyzed organosolv treatment had a CSF’ of 8.73, which was also higher than the 7.93 and 7.71 reported for the SuAc-catalyzed ethanol organosolv treatment of wheat bran [19] and coffee silver skin [29] to achieve the maximum Y_TP. Even lower CSF’ values of 7.44 were required for the maximum Y_TP from red grape pomace, employing a citric acid-catalyzed organosolv treatment [30].

The CSF’ is widely used to evaluate and optimize ethanol organosolv treatments [31], yet it should be considered as merely indicative, since differences might arise from the recalcitrance of the treated material, but they may also be related to the polyphenolic composition. For example, bound phenolics may require stronger acid and/or alkaline conditions, longer residence times and higher temperatures for their liberation and recovery, as opposed to non-bound compounds, which may be easily entrained in the liquid medium (solvent). This would inevitably increase process severity. In any case, when attempting to design a suitable treatment for effective polyphenol recovery, it is of paramount importance to consider pH, temperature and residence time [22]. Based on recent investigations that have demonstrated sound correlations between CSF’ and Y_TP [19,29,30], severity might be regarded as an additional tool in implementing and assessing pertinent processes.

3.4. Treatment Effect on the Polyphenolic Composition

The extracts produced from the hydrothermal and organosolv treatments, but also the control obtained from treatments carried out with neat water and 60% ethanol, were subjected to HPLC analysis, to examine the effect of both the acid catalyst and the solvent on the polyphenolic profile. In Figure 5, it can be seen that the treatment with water provided an extract composed essentially of CouA and FA. However, in the extracts obtained with either catalyst (OxAc or SuAc), another compound (peak 1) appeared. This finding evidenced that the presence of acid affected the polyphenolic composition. To obtain information regarding the nature of peak 1, the extract was examined by performing LC-DAD-MS/MS analyses. The UV–vis spectrum of this peak was identical to that of a p-coumaric acid standard, while the mass spectrum acquired in negative ionization mode showed a molecular ion at m/z = 295 and a diagnostic fragment at m/z = 163. The latter fragment corresponded to the p-coumarate moiety and based on this outcome, peak 1 was tentatively identified as a p-coumaric acid-arabinofuranose ester, in line with previous findings [32,33]. Peak 1 was then designated as CouE.

When treatments were performed with 60% ethanol, peak 1 occurred at significantly low levels, yet several other peaks of lower polarity than FA appeared in the chromatogram (Figure 6). These compounds were not detected in the extract generated with any hydrothermal treatment, and thus they were presumably ethanol-extractable substances. Furthermore, the use of either acid catalyst did not result in the generation of CouE, a finding that suggests that ethanol might have inhibited its formation.

To gain deeper insight into the effect of both acid catalyst and solvent (ethanol), the major phytochemicals detected (CouE, CouA and FA) were quantified in all extracts produced (

Table 5. Yields of the major hydroxycinnamates in the extracts produced by employing hydrothermal and organosolv treatments. Note: CouA and FA correspond to p-coumaric acid and ferulic acid.

Treatment	Y (mg g−1 DM)
	CouE	CouA	FA	CouA/FA Ratio	Total
Water	<0.01 a	0.57 ± 0.03 a	0.23 ± 0.02 a	2.5	0.80
Hydrothermal (5% OxAc)	0.48 ± 0.03 b	1.18 ± 0.06 b	0.33 ± 0.01 b	3.6	1.99
Hydrothermal (1% SuAc)	0.45 ± 0.02 b	0.67 ± 0.04 c	0.19 ± 0.02 a,c	3.5	1.31
60% EtOH	<0.01 a	0.98 ± 0.03 d	0.18 ± 0.01 c	5.4	1.16
Organosolv (60% EtOH/10% OxAc)	<0.01 a	2.21 ± 0.09 e	0.46 ± 0.02 d	4.8	2.67
Organosolv (60% EtOH/1.5% SuAc)	<0.01 a	2.96 ± 0.07 f	0.70 ± 0.02 e	4.2	3.66

Different letters (a, b, c, d, e and f) within columns denote statistically significant differences (p < 0.05).

). The water extract was the poorest, as it afforded only 0.80 mg g⁻¹ DM of total hydroxycinnamates. SuAc-catalyzed hydrothermal treatment boosted CouE formation, and it also resulted in a CouA yield increase from 0.57 to 0.67 mg g⁻¹ DM; overall, the increase in total hydroxycinnamates was almost 39%. To the contrary, the OxAc-catalyzed hydrothermal treatment gave extracts significantly more enriched in all hydroxycinnamates considered, providing in total 1.99 mg g⁻¹ DM.

In the case of organosolv treatment, the role of both catalysts tested was crucial in attaining a higher polyphenol yield, but the SuAc-catalyzed treatment was proven more efficient compared with the OxAc-catalyzed one, as it gave approximately 25% higher CouA and 34% higher FA yields. In total, the SuAc-catalyzed organosolv treatment provided 3.66 mg g⁻¹ DM, and outperformed all other treatments employed. This outcome highlighted that (i) OxAc and SuAc may have different potencies in an aqueous and a hydroethanolic environment, (ii) the ethanol organosolv treatment was superior to the hydrothermal treatment in obtaining extracts enriched in both CouA and FA, but not in CouE. In any case, acid catalysis was the key factor in boosting polyphenol recovery.

In CS, apart from the lignin fraction, which is a polymeric network of covalently connected monolignols, the polyphenolic composition is dominated by CouA and FA [8,18], which was also clearly demonstrated in this study (Figure 5 and Figure 6). These compounds occur in CS as constituents of the lignocellulosic matrix, attached onto lignin and/or hemicellulose chains through various bonds. With reference to CouA, it has long been shown that almost 90% is attached to syringyl moieties of lignin via ester bonds [34,35]. On the other hand, FA is mainly esterified with arabinosyl residues of arabinoxylan chains. However, FA may also occur as ether- or ester-linked onto lignin structures, forming cross-links between these two polymeric matrices [18,36].

The high-performance retrieval of both CouA and FA using acid-catalyzed organosolv treatment may thus be associated with phenomena embracing lignin/lignocellulose deconstruction and solubilization, but most importantly hydroxycinnamate detachment from the lignocellulosic complex through hydrolytic reactions, through the cleavage of ester and ether bonds [8]. In general, the reactions involved in ethanol organosolv treatment include the hydrolysis of the internal bonds in lignin, but also the bonds associated with the lignin–hemicellulose complex, liberating lignin fragments, and the hydrolysis of glycosidic bonds in cellulose and hemicellulose [31].

Depending on the conditions deployed, lignin may undergo depolymerization, generating low-molecular weight fragments, and ether linkage cleavage contributes to such fragmentation. In ethanol organosolv treatment, lignin breakdown may largely depend on the cleavage of α-aryl and β-aryl ether linkages, which may take place in the presence of an acid. Ether-linked hydroxycinnamates may be cleaved only upon acid catalysis and, under acidic conditions and in the presence of a solvent (ethanol), depolymerization may be greatly promoted, increasing phenolic monomer yield [37]. At this point the involvement of ethanol may be critical, assisting lignin solubilization [31], but also permitting the facile accessibility of ester/ether bonds by the catalyst, which is a rate-limiting step in the hydrolysis process [38].

As an acid catalyst, OxAc has been studied in various biomass treatments, such as corncob processing, where it displayed a higher performance than SuAc in producing xylose [39], while in pineapple treatment OxAc was of equal efficacy to SuAc in the production of cellulose nanofibers [40]. OxAc-catalyzed treatments were also demonstrated to have a high efficacy in the generation of feruloylated oligosaccharides from corn bran [41] and lignin-containing cellulose nanocrystals [42]. Most importantly, aqueous OxAc systems were shown to be promising lignin cleavage media, promoting the breaking down of β-O-4 lignin model compounds under oxidative conditions. On such grounds, OxAc has been proposed as a bio-based, benign and recycled reagent for the production of high-value-added aromatic chemicals, such as vanillin, without the need for metal catalysts or organic solvents [43]. In concurrence with these findings, the potency of OxAc to catalyze hydrolytic reactions has also been illustrated by recent studies pertaining to quercetin glycosides from onion solid waste using a glycerol/oxalic acid deep eutectic solvent, where optimized conditions of time and temperature were used to produce extracts highly enriched in quercetin [44].

3.5. Antioxidant Properties

The extracts generated by applying optimized treatment conditions, along with water and hydroethanolic extracts (no acid catalyst addition) were appraised for antiradical activity (A_AR) and ferric-reducing power (P_R) to obtain evidence regarding their antioxidant potency. The extract produced with 60% ethanol/1.5% SuAc was shown to display considerably a higher A_AR compared with all other extracts (p < 0.05), followed by the extract obtained with 60% ethanol/10% OxAc (Figure 7A). This outcome was in line with the polyphenolic richness of the extracts (Table 5). However, P_R did not exhibit the same trend, and the most active extract was the one produced with 1% aqueous SuAc, followed by that obtained with 5% OxAc (Figure 7B). Similar discrepancies between A_AR and P_R have been previously reported for ferulic acid-containing extracts from wheat bran [19,20], and putatively ascribed to the differences in the polyphenolic composition, but also the manifestation of possible synergistic/antagonistic effects amongst the various polyphenolic constituents of the extracts.

Taking into account that the major difference between the hydrothermal- and organosolv-treated extracts was the presence of a p-coumaric acid-arabinofuranose ester (Figure 5), then it could be argued that its presence in the extracts obtained with the hydrothermal treatments might have contributed to the increased P_R found. However, this hypothesis remains to be elucidated. On the other hand, it is well established that both CouA and FA have radical-scavenging capacity [13] and, being the principal extract polyphenols, could define the overall antiradical activity. In fact, in wheat bran extracts, ferulic acid was claimed to play a prominent role in the antioxidant activity as a major antioxidant contributor [45,46]. This argument was further supported by wheat bran subjected to hydrolysis, where ferulic acid-enriched extracts exhibited more powerful antiradical activity [47]. The results presented herein were in accordance with such a phenomenon, since the extracts obtained with acid catalysis (either with hydrothermal or organosolv treatment) were far more enriched with both CouA and FA compared with those obtained without catalysis and displayed stronger antioxidant effects. This finding highlighted the importance of hydroxycinnamate liberation in the expression of antioxidant activity.

4. Conclusions

The examination presented herein revealed for the first time the potential of using a natural organic acid (oxalic acid) as an effective catalyst to boost the release of antioxidant hydroxycinnamates from corn stover, a highly abundant agricultural biowaste. Oxalic acid was a more efficient catalyst in hydrothermal treatment, providing higher recoveries for both p-coumaric and ferulic acids. However, the sulfuric acid-catalyzed ethanol organosolv treatment was shown to be the highest-performing system in this regard. Under optimized time and temperature conditions, which lend the treatment a rather moderate severity, the methodology developed afforded a recovery of total hydroxycinnamates of about 3.7 mg per g of dry mass. The recovery of these compounds may be an important stage in an integrated biorefinery process, with the prospect of producing antioxidant substances that could be of high value to the food, cosmetics and pharmaceutical industries, owing to their multiple biological properties. The green character of the process could be significantly enhanced by replacing corrosive sulfuric acid with benign oxalic acid, but there may be a compromise regarding recovery yields. The issue of effectiveness, along with the higher oxalic acid cost and the cost associated with its recycling might be an important shortcoming of a pertinent methodology; however, further optimization could increase the overall treatment performance to deliver a highly effective and fully sustainable process. Currently, work is in progress towards this objective, which is anticipated to further bring out the potential of corn stover as a unique bioresource of bioactive phytochemicals.