Molecular‐Level Characterization of Native Lignin Oligomers From Flow‐Through Solvolysis of Biomass by High‐Performance Liquid Chromatography–Tandem Mass Spectrometry

ABSTRACT

Rationale

Comprehensive characterization of lignin‐derived oligomers remains a major analytical challenge because of their high structural diversity and the lack of methods capable of resolving native interunit linkages in complex biomass‐derived mixtures. Molecular‐level understanding of these oligomers is essential for elucidating lignin depolymerization mechanisms and guiding biomass valorization strategies.

Methods

Lignin oligomers generated from methanol‐based flow‐through solvolysis of birch sawdust were analyzed by high‐performance liquid chromatography coupled with ultraviolet detection and tandem mass spectrometry (HPLC‐UVD/MSⁿ). Chromatographic separation was performed on tandem reversed‐phase C18 columns, and structural characterization was achieved using electrospray ionization Orbitrap mass spectrometry with higher‐energy collisional dissociation (HCD).

Results

Distinct chromatographic behaviors were observed for lignin‐derived products with different degrees of polymerization, enabling differentiation of monomeric and oligomeric species. Tandem MS analysis of representative oligomers revealed characteristic fragmentation pathways that enabled structural elucidation of oligomers. β‐O‐4 linkages were identified as the predominant interunit structures, and native β‐β motifs were also characterized.

Conclusions

Untargeted molecular‐level characterization of lignin‐derived products generated from methanol‐based flow‐through solvolysis of birch biomass provides direct insights into the composition and interunit linkages of native‐like lignin oligomers. These findings reveal the structural diversity of biomass‐derived oligomers and bridge the gap between model‐compound studies and real biomass‐derived lignin systems.

1. Introduction

The efficient conversion of lignocellulosic biomass into fuels and value‐added chemicals has attracted increasing attention as a sustainable strategy to reduce dependence on fossil resources and mitigate carbon emissions [1, 2, 3]. Among the three major biomass components, lignin is particularly significant because it represents the largest renewable source of aromatic compounds in nature. Unlike cellulose and hemicellulose, lignin possesses a highly heterogeneous polyphenolic macromolecular structure formed through radical‐mediated polymerization of three primary monolignols: p‐coumaryl alcohol (H‐unit), coniferyl alcohol (G‐unit), and sinapyl alcohol (S‐unit). These monomers are interconnected through a wide variety of ether and C‐C linkages, such as β‐O‐4, β‐β, β‐5, and 5‐5 bonds, resulting in a nonlinear architecture with substantial structural variability across different biomass species [4, 5]. This structural complexity not only contributes to the rigidity and recalcitrance of plant cell walls but also poses major challenges for selective lignin depolymerization and product control [6].

To better understand lignin conversion pathways, extensive efforts have been devoted to the study of lignin model compounds. Monomeric and dimeric molecules containing representative functional groups or interunit linkages have been widely employed to investigate bond cleavage mechanisms, hydrogenolysis pathways, and recondensation reactions [7, 8, 9, 10, 11, 12]. These studies have provided valuable mechanistic insights into lignin depolymerization chemistry. However, such simplified model compounds preserve only limited structural features of native lignin and often neglect critical factors including steric hindrance, solubility, and hydrogen bonding [13, 14, 15, 16]. Consequently, even closely related phenolic and non‐phenolic β‐O‐4 model compounds can exhibit markedly different reactivities, highlighting the limitations of extrapolating model‐compound results to native lignin systems [13]. To address these limitations, more structurally sophisticated trimeric or polymeric lignin models have been recently developed to better mimic native lignin [14, 17, 18]. Nevertheless, a significant gap remains between these synthetic model compounds and the native lignin fragments generated in biomass depolymerization, which typically consist of multi‐unit oligomers with average molecular weights (MW) around 1500 Da [19], motivating increasing interest in the direct investigation of native lignin systems and heavy oligomeric intermediates.

In recent years, advances in extraction and analytical techniques have significantly improved the ability to preserve and characterize native‐like lignin structures. Particularly, flow‐through solvolysis has emerged as a promising strategy for the extraction of highly reactive native‐like lignin from lignocellulosic biomass [20, 21, 22, 23, 24, 25]. By minimizing the residence time of lignin fragments within the reaction zone, this approach effectively suppresses condensation reactions and preserves chemically labile β‐O‐4 linkages. The resulting lignin products maintain high chemical reactivity and offer considerable flexibility for downstream catalytic upgrading. Despite these advantages, the comprehensive characterization of flow‐through solvolysis products remains highly challenging because the dominant products are high‐molecular‐weight lignin oligomers that cannot be adequately analyzed using conventional GC‐based techniques [26, 27, 28]. Although GPC and NMR provide valuable information regarding average molecular weight distributions and bulk structural motifs [29, 30, 31], they lack sufficient molecular‐level resolution to fully resolve the complexity of lignin oligomer products.

High‐resolution mass spectrometry (HRMS), particularly when coupled with chromatographic separation, provides powerful capabilities for the molecular‐level characterization of complex lignin‐derived oligomers [32, 33]. Among these techniques, high‐performance liquid chromatography coupled with tandem mass spectrometry (HPLC‐MSⁿ) offers unique advantages in lignin conversion studies, as it enables both efficient separation of complex oligomeric mixtures and structure elucidation based on multistage fragmentation information [34, 35, 36, 37]. This technique is highly effective for revealing the molecular composition, interunit linkages, and functional group distributions of lignin oligomers, thereby providing critical insights into lignin depolymerization and condensation behaviors. Despite these advances, most previous studies have focused on technically isolated lignins, such as organosolv lignin, Kraft lignin, and soluble lignin extracts, which have often undergone substantial structural modifications during extraction and processing. Consequently, the molecular composition and interunit linkages of native‐like lignin oligomers directly generated from biomass remain largely unexplored, leaving a significant gap between model‐compound investigations and real biomass‐derived lignin systems. This lack of molecular‐level information limits our understanding of the formation, preservation, and transformation of native lignin fragments during biomass depolymerization.

In this work, the methanol‐based flow‐through solvolysis of birch sawdust was systematically investigated. The combined use of HPLC, ultraviolet detection (UVD), and MSⁿ enabled in‐depth molecular‐level characterization of the native lignin oligomers from flow‐through degradation, allowing detailed analysis of their compositional and structural features. The findings not only provide important insights into the formation and evolution mechanisms of oligomeric species during flow‐through lignin solvolysis, but also help bridge the gap between simplified model compound studies and the conversion behavior of native lignin systems, thereby contributing to the rational valorization and targeted conversion of lignin.

2. Materials and Methods

2.1. Flow‐Through Solvolysis

Thermal solvolysis was carried out using a home‐made flow‐through system in a single‐bed reactor configuration (I.D. = 4 mm, length = 100 mm) [28]. Local birch wood was milled and sieved to a particle size below 250 μm. Prior to use, the sawdust was subjected to extractive removal by immersion in a boiling toluene/ethanol mixture (2:1, v/v) for 15 min, followed by Soxhlet extraction for 3 h to remove waxes, fats, resins, and terpenoids/steroids. The extracted solids were then washed thoroughly with ethanol and dried overnight at 80 °C [38, 39]. The resulting extractive‐free birch feedstock was loaded into the center of the reactor for subsequent flow‐through solvolysis experiments. For each run, 50 mg of the feedstock (∼10 mm in length with an estimated void fraction of 0.7) was used. The remaining reactor volume was filled with glass beads and quartz wool to maintain flow stability and prevent feedstock displacement. After sealing with VCR fittings, the reactor was purged with nitrogen five times and pressurized to 6 MPa to ensure that methanol remained in the liquid phase under the reaction conditions. Reactions were performed for a total of 6 h under a temperature‐programmed protocol. The external temperature was held at 180°C for 2 h, then increased to 200°C and held for another 2 h (including ~10 min of ramping), and finally raised to 220°C for an additional 2 h. The corresponding internal reactor temperatures approximately 15 °C lower than the external temperatures. Once the target temperature was reached, methanol was introduced at a constant flow rate of 0.2 mL/min, giving an estimated solvent‐feedstock contact time of ~25 s. Following each run, the liquid product was collected for HPLC‐UVD/MSⁿ and gel permeation chromatography (GPC) analysis without further pre‐treatment.

2.2. HPLC‐UVD/MSⁿ analysis

HPLC‐UVD/MSⁿ analysis was carried out using a Thermo U3000 system coupled to an Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific). After chromatographic separation, the eluents first passed through the UV detector cell and were subsequently introduced into the negative mode electrospray ionization (‐ESI) source of the mass spectrometer through a capillary, resulting in an approximate delay of 15 s between the UV and MS signals. Product separation was achieved using two tandem reversed‐phase columns (Syncronis C18, 100 mm × 2.1 mm × 1.7 μm). Injection volumes of 2.5 μL and 10 μL were used for UVD and MS analysis, respectively. The column temperature was maintained at 40 °C. The UV detector was operated at 275 nm, a wavelength particularly suitable for detecting aromatic conjugated systems, especially phenolic compounds. Model compounds dissolved in methanol were used to confirm their retention times and UV responses for calibration and semi‐quantitative analysis. Details of the gradient elution program are summarized in Table 1.

TABLE 1.

Solvent gradient used for HPLC separation.

Time/min	Flow/(mL·min−1)	%Acetonitrile (B)
0	0.2	30
6	0.2	30
27	0.2	80
30	0.2	80
33	0.2	30
35	0.2	30

MS data were acquired and processed using Xcalibur 4.0 software. Elemental composition assignment was performed under the following constraints: ¹²C between 3 and 80, ¹H between 4 and 120, ¹⁶O between 1 and 40, with a mass error less than 3 mDa. The double‐bond equivalent (DBE), calculated according to DBE = 1 + x − y/2 for C_xH_yO_z species, was restricted to values between 0 and 30. In addition, the nitrogen rule for even‐electron ions was applied during formula assignment. The parameters for the Orbitrap mass spectrometer are summarized in Table 2.

TABLE 2.

Key parameters for the Orbitrap mass spectrometer.

Experimental parameters	Value
ESI voltage (V)	−2700
Mass range (Da)	120~2000
Resolution (at 200 Da)	500 000
Sheath gas (L/min)	3.6
Aux gas (L/min)	12.1
Sweep gas (L/min)	5.4
Ion transfer tube temperature (°C)	320
Vaporization temperature (°C)	150

The MSⁿ experiments were conducted using higher‐energy collisional dissociation (HCD). Precursor ions were isolated using a quadrupole mass analyzer with an isolation width of 0.1 Da and subsequently accelerated to induce collisions with inert nitrogen gas. During this process, kinetic energy was converted into internal vibrational energy, leading to bond cleavage and the formation of characteristic fragment ions once the dissociation threshold was exceeded. HCD collision energies were varied from 0% to 100% in 10% increments to generate different fragmentation patterns, enabling structural elucidation of lignin oligomers and related products.

2.3. GPC Analysis

The products dissolved in tetrahydrofuran were analyzed by GPC using an EcoSEC HLC‐8320 system (Tosoh Corporation) equipped with a refractive index detector (RID). MW distributions ranging from approximately 100 to 10 000 Da were obtained with a total acquisition time of 20 min.

3. Results and Discussion

3.1. Molecular Weight Distributions of Lignin Fragments

Lignin fragments collected from the flow‐through solvolysis of birch wood were first analyzed by GPC to evaluate their MW distribution. As shown in Figure 1, the extracted lignin fragments exhibited relatively high molecular weights, with a number‐average molecular weight (Mn) of 507 Da and a weight‐average molecular weight (Mw) of 1559 Da. In addition to monomers and dimers, a substantial fraction of oligomeric species extending beyond 1000 Da was observed, with the dominant distribution ranging from 600 to 3000 Da and peaking near 2000 Da. Since the response of the RID is approximately proportional to mass concentration, the relatively high Mw (1559 Da) suggests that oligomers corresponding roughly to heptamers to decamers account for a significant mass proportion of the lignin fragments, consistent with previous reports on native‐like lignin fragments. In contrast, the lower Mn value indicates that these high‐MW oligomers are comparatively less abundant in number, with trimers representing the dominant species on a molecular‐count basis.

FIGURE 1.

GPC profiles of lignin fragments obtained from flow‐through solvolysis of birch feedstocks (180~220 °C, 6 h).

3.2. Molecular Weight Distributions by HPLC‐UVD/MS

To further elucidate the molecular composition of the extracted lignin fragments, HPLC‐UVD/MS analysis was subsequently performed. In methanol‐based biomass conversion systems, the earliest chromatographic peak (before ~3 min, Figure 2a) corresponds to the methanol solvent front. Because methanol is present in large excess, trace amounts of products with limited retention on the chromatographic column may partially co‐elute with the solvent and are not completely resolved. In contrast, most of the reaction products are sufficiently retained by the column and are separated from methanol, eluting at longer retention times. Compounds subsequently elute according to their polarity and hydrophobicity. Carbohydrate‐derived products, primarily monosaccharides and their derivatives released from hemicellulose, together with a smaller contribution from cellulose‐derived carbohydrates, generally elute before 3.5 min because of their high polarity arising from abundant hydroxyl and oxygen‐containing functional groups. Consequently, they are well separated from lignin‐derived phenolic compounds, which elute at longer retention times. Such separation behavior helps reduce matrix interference and improves the ESI efficiency of the later‐eluting phenolic species.

FIGURE 2.

HPLC‐UVD/MS analysis of birch solvolysis products. (a) UV chromatogram. (b–e) Base peak chromatograms via HPLC‐MS with m/z range in lignin monomer (120~230 Da, b), dimer (270~500 Da, c), trimer (460~720 Da, d), and tetramer (700~1000 Da, e) regions, respectively. Assignment of the numbered peaks are summarized in Table 3. Monomeric species in (a) and (b) are denoted by S, G, and H according to the lignin unit type (syringyl, guaiacyl, and p‐hydroxyphenyl, respectively), with the numerical suffix indicating the assigned peak in Table 3. Dimeric, trimeric, and tetrameric species in (c–e) are denoted by D, T, and Te, respectively, followed by the assigned peak number in Table 3.

For low‐degree oligomers (monomers to tetramers), retention behavior is highly sensitive to polarity differences arising from the type, number, and position of functional groups, resulting in substantial variation in retention times among different structures. Low‐MW phenolics, particularly monomers, dimers, and trimers, therefore exhibit a broad elution window extending beyond approximately 4 min. Based on this characteristic, tracking the temporal evolution of base peak intensities within selected mass ranges provides more meaningful compositional information, enabling identification of the dominant products at different degrees of polymerization (DP). Figure 2b–e presents the base peak chromatograms obtained by HPLC‐MS in the lignin monomer (120–230 Da), dimer (270–500 Da), trimer (460–720 Da), and tetramer (700–1000 Da) mass regions, respectively, and the assignments of the numbered peaks are summarized in Table 3. The selected mass windows serve only as approximate boundaries for differentiating species with different degrees of polymerization.

TABLE 3.

List of major monomers to tetramers reflected from the base peak monitoring of HPLC‐MS.

No.	m/z	Formula	DBE	R.T.	No.	m/z	Formula	DBE	R.T.
S1	225.08	[C11H13O5]−	5	3.85	T1	649.21	[C31H37O15]−	13	5.20
G1	195.07	[C10H11O4]−	5	3.85	T2	529.21	[C28H33O10]−	12	5.50
H2	137.02	[C7H5O3]−	5	4.05	T3	619.24	[C31H39O13]−	12	6.55
S2	197.05	[C9H9O5]−	5	4.15	T4	633.22	[C31H37O14]−	13	8.00
G2	167.04	[C8H7O4]−	5	4.15	T5	659.23	[C33H39O14]−	14	9.40
S3	209.08	[C11H13O4]−	5	4.95	T6	691.26	[C34H43O15]−	13	10.80
G3	179.07	[C10H11O3]−	5	5.25	T7	663.23	[C32H39O15]−	13	12.20
S4	181.05	[C9H9O4]−	5	6.20	T8	643.24	[C33H39O13]−	14	13.10
G4	151.04	[C8H7O3]−	5	6.20	T9	643.24	[C33H39O13]−	14	13.55
S5	207.07	[C11H11O4]−	6	8.20	T10	613.23	[C32H37O12]−	14	14.00
G5	177.06	[C10H9O3]−	6	8.55	T11	643.24	[C33H39O13]−	14	14.55
S6	223.10	[C12H15O4]−	5	13.65	T12	613.23	[C32H37O12]−	14	15.00
G6	193.09	[C11H13O3]−	5	14.40	T13	535.20	[C30H31O9]−	15	15.45
D1	451.16	[C22H27O10]−	9	4.10	T14	641.22	[C33H37O13]−	15	15.60
D2	421.15	[C21H25O9]−	9	4.30	T15	677.25	[C33H41O15]−	13	16.00
D3	423.13	[C20H23O10]−	9	4.50	T16	595.22	[C32H35O11]−	15	16.55
D4	393.12	[C19H21O9]−	9	4.65	T17	657.26	[C34H41O13]−	14	17.10
D5	435.17	[C22H27O9]−	9	5.80	T18	549.21	[C31H33O9]−	15	17.70
D6	405.16	[C21H25O8]−	9	6.20	T19	639.21	[C33H35O13]−	16	18.35
D7	433.15	[C22H25O9]−	10	7.25	T20	611.21	[C32H35O12]−	15	18.55
D8	403.14	[C21H23O8]−	10	7.50	T21	547.20	[C31H31O9]−	16	19.90
D9	437.15	[C21H25O10]−	9	8.60	T22	549.21	[C31H33O9]−	15	20.50
D10	415.14	[C22H23O8]−	11	9.30	T23	623.21	[C33H35O12]−	16	21.00
D11	329.10	[C18H17O6]−	10	10.05	T24	563.23	[C32H35O9]−	15	22.10
D12	433.15	[C22H25O9]−	10	10.50	Te1	845.32	[C42H53O18]−	16	8.35
D13	355.12	[C20H19O6]−	11	12.65	Te2	885.32	[C44H53O19]−	18	13.35
D14	367.12	[C21H19O6]−	12	13.10	Te3	869.32	[C44H53O18]−	18	14.20
D15	359.11	[C19H19O7]−	10	14.45	Te4	839.31	[C43H51O17]−	18	14.50
D16	327.09	[C18H15O6]−	11	14.85	Te5	869.32	[C44H53O18]−	18	15.00
D17	403.10	[C20H19O9]−	11	15.85	Te6	867.31	[C44H51O18]−	19	15.80
D18	371.15	[C21H23O6]−	10	16.40	Te7	883.34	[C45H55O18]−	18	17.10
D19	343.12	[C19H19O6]−	10	18.40	Te8	853.33	[C44H53O17]−	18	17.40
D20	369.13	[C21H21O6]−	11	19.55	Te9	881.32	[C45H53O18]−	19	18.40
D21	373.13	[C20H21O7]−	10	20.75	Te10	897.35	[C46H57O18]−	18	19.55
D22	385.17	[C22H25O6]−	10	22.40	Te11	713.26	[C40H41O12]−	20	23.65

Note: Monomeric species are designated as S, G, and H according to the lignin unit type (syringyl, guaiacyl, and p‐hydroxyphenyl, respectively), as labeled in Figure 2a,b, whereas dimeric, trimeric, and tetrameric species shown in Figure 2c–e are denoted by D, T, and Te, respectively. The numerical suffix identifies the corresponding peak in Figure 2 and the associated entry in this table. The formulas listed correspond to the deprotonated ions, [M‐H]⁻, observed in negative ESI mode.

Notably, several monomeric compounds, such as C₁₁H₁₄O₄ (S3 in Table 3) and C₁₂H₁₆O₄ (S6 in Table 3), exhibited weak or negligible response under ESI conditions and were therefore identified based on their UV signals by comparison with authentic standards (Figure 2a,b). The monomeric products directly obtained from flow‐through solvolysis exhibited relatively limited structural diversity and strong UV responses, reflecting both the high molar absorptivity of the conjugated aromatic‐alkenyl chromophores (DBE mostly 5~6) conjugated with aromatic rings. Compared with higher oligomers, monomers also exhibited better chromatographic resolution, whereas dimers and oligomers were distributed over numerous structural isomers with broader chromatographic peaks. Together with their inherently lower molar concentrations resulting from higher molecular weights and, in some cases, reduced conjugation, these factors collectively led to substantially weaker UV peak intensities for oligomers. These side chains largely remained non‐deoxygenated, resulting in relatively high polarity and consequently early elution within 15 min. According to the lignin unit type, monomeric products were classified as syringyl (S), guaiacyl (G), and p‐hydroxyphenyl (H) species, as summarized in Table 3. S‐type and G‐type monomers, differing by one CH₂O unit, typically appeared as paired species, with S‐type compounds eluting slightly earlier or nearly simultaneously with their G‐type analogues. The dominant monomeric products were identified as sinapyl alcohol (C₁₁H₁₄O₄, S3 in Table 3) and coniferyl alcohol (C₁₀H₁₂O₃, G3 in Table 3).

Among oligomeric species, dimers and trimers exhibited the most abundant base peaks, with more than 20 molecular formulas assigned for each category. In contrast, tetramers displayed less prominent chromatographic features, likely because the polarity differences among higher oligomers became increasingly small, limiting chromatographic separation efficiency. While the actual number of molecular species is expected to be much higher, the identified compounds likely represent the predominant products. Many adjacent chromatographic peaks either shared identical molecular formulas, suggesting the presence of structural isomers, or differed by a CH₂O unit, indicating substitution between S‐ and G‐type monolignol units within otherwise identical or similar structures.

For higher‐molecular‐weight oligomers (DP > 4), the influence of polarity differences on retention gradually diminished, whereas increasing molecular size and hydrophobic surface area became the dominant factors governing chromatographic behavior. As a result, retention time generally increased with DP, giving rise to the characteristic broad hump observed in the later portion of the chromatogram (Figure 2a). Extracted ion chromatogram analysis showed that oligomers from DP 4 to DP 8 were mainly distributed between 8 and 26 min and exhibited approximately Gaussian retention profiles. Adjacent DP classes differed by one lignin monomer unit, corresponding to a mass increment of ~200 Da and a typical elemental composition of C_10~11H_12~14O_3~5 per repeating unit (Figure S1).

3.3. Structural Identification by Tandem Mass Spectrometry

Tandem MS analysis was performed on representative flow‐through solvolysis products from birch wood. Many β‐O‐4 linkages were identified in both dimeric and higher oligomeric species. Representative tandem MS spectra of dimers, trimers, and tetramers are shown in Figure 3, along with the corresponding structural assignments derived from fragmentation behavior and known lignin structural characteristics.

FIGURE 3.

Tandem MS analysis of birch‐derived dimers, trimers, and tetramers, with proposed structural assignments. MS² spectra of (a) [C₂₁H₂₅O₉] ⁻ (D2 in Table 3) with a collision energy of 20%, (b) [C₃₃H₃₉O₁₄] ⁻ (T5 in Table 3), and (c) [C₄₄H₅₃O₁₈] ⁻ (Te3 in Table 3), both with a collision energy of 15%. Asterisks (★) denote the precursor ions.

At relatively low collision energy (HCD = 15~20%), the precursor ions predominantly underwent inter‐unit bond cleavage, producing a series of mono‐, di‐, and trimeric fragments, which is consistent with the relatively low bond dissociation energy of β‐O‐4 linkages. For example, the MS² spectrum of C₂₁H₂₆O₉ (D2 in Table 3) acquired at HCD = 20% (Figure 3a) showed dominant cleavage at the C_β‐O bond, yielding G‐ and S‐type fragments. These monomeric fragments further exhibited typical CH₂O and H₂O neutral losses, closely resembling the fragmentation patterns observed for lignin model compounds (Supplementary Figures S2–S12) [12], thereby enabling plausible structural reconstruction of the dimer.

Similarly, as shown in Figure 3b, the trimer C₃₃H₄₀O₁₄ (T5 in Table 3) produced a prominent monomeric fragment ion at [C₁₁H₁₁O₄]⁻ under HCD = 15%. This fragment likely originated from two pathways: direct formation from a sinapaldehyde unit (highlighted in dark gray in the proposed structure) and dehydration of the fragment ion [C₁₁H₁₃O₅]⁻, whose fragmentation behavior is analogous to that observed for C₂₁H₂₆O₉ (D2 in Table 3) shown in Figure 3a. Combined with the characteristic structural features of hardwood lignin, the observed dimeric fragment [C₂₂H₂₇O₁₀]⁻ strongly supports a unique trimer structure assignment.

For the tetramer C₄₄H₅₄O₁₈ (Te3 in Table 3), as shown in Figure 3c, analysis of the major fragments across various DP suggests a structure composed of two C₁₁H₁₄O₅ units, corresponding to fragments such as [C₁₁H₁₃O₅]⁻ and [C₂₂H₂₇O₁₀]⁻, together with a typical resinol‐type substructure represented by [C₂₂H₂₅O₈]⁻. The latter corresponds to a typical β‐β type dimer, likely formed by two S‐type lignin units. This assignment is consistent with the previously reported structural characteristics of birch lignin and with β–β‐linked structures previously identified by NMR in birch RCF products [38, 39].

Additional tandem MS analyses of several lignin‐derived oligomers are summarized in Table 4, listing the major fragment ions along with the proposed building blocks, corresponding inter‐unit linkage types and proposed structures. The MS/MS spectra and proposed structures for each entry are provided in Figures S13–S18. It should be noted that the structures proposed herein represent plausible assignments derived from the observed MSⁿ fragmentation behavior rather than definitive assignments. Because multiple isomeric structures may share the same molecular formula and generate similar fragment ions, the present method cannot unambiguously distinguish among all possible isomers.

TABLE 4.

Structural analysis of typical oligomers from flow‐through solvolysis of birch wood via tandem MS.

Entry (m/z)	Formula (DBE)	R.T. /min	HCD energy	Primary fragments	Units	Inter‐unit linkages	Proposed structure
1 (619.24) T3	[C31H39O13]− (12)	6.55	20%	571.22 (−CH4O2)	S‐C3‐2OH G‐C3‐2OH G‐C3‐2OH	β‐O‐4 × 2
				393.16 (−C11H14O5)
				★ 331.12 [C18H19O6]−
				★ 225.08 [C11H13O5]−
				195.07 [C10H11O4]−
2 (663.23) T7	[C32H39O15]− (13)	12.20	20%	451.16 (−C10H12O5)	S‐C3‐2OH S‐C3‐2OH S‐COOMe	β‐O‐4 × 2
				225.08 [C11H13O5]−
				★ 211.06 [C10H11O5]−
				196.04 [C9H8O5]−
3 (643.24) T9	[C33H39O13]− (14)	13.55	15%	★ 595.22 (−CH4O2)	S‐C3‐2OH 2S‐resinol	β‐O‐4 β‐β
				★ 417.15 (−C11H14O5)
				★ 225.08 [C11H13O5]−
				195.07 [C10H11O4]−
4 (657.26) 17	[C34H41O13]− (14)	17.10	20%	★ 595.22 (−C2H6O2)	S‐C3‐OH‐OMe 2S‐resinol	β‐O‐4 β‐β
				★ 417.15 (−C12H16O5)
			30%	220.07 [C12H12O4]‐·
				181.05 [C9H9O4]−
5 (839.31) Te4	[C43H51O17]− (18)	14.50	10%	791.29 (−CH4O2)	S‐C3‐2OH G‐C3‐2OH 2S‐resinol	β‐O‐4 × 2 β‐β
				643.24 (−C10H12O4)
				★ 421.15 [C21H25O9]−
				★ 417.16 [C22H25O8]−
				225.08 [C11H13O5]−
				★ 195.07 [C10H11O4]−
6 (853.33) Te8	[C44H53O17]− (18)	17.40	20%	791.29 (−C2H6O2)	S‐C3‐2OH G‐C3‐OH‐OMe 2S‐resinol	β‐O‐4 × 2 β‐β
				595.22 (−C12H18O6)
				★ 417.16 [C22H25O8]−
				★ 225.08 [C11H13O5]−
				★ 195.07 [C10H11O4]−

Note: T and Te denote trimeric and tetrameric species, respectively, corresponding to Table 3 entries. Primary fragments are listed as neutral or radical losses from precursor ions to indicate characteristic functional groups (e.g., −CH₄O₂ represents concerted elimination of –OH and –CH₂OH). Monomeric fragments are given as molecular formulas for direct structural assignment. ★ indicate dominant fragment ions. S and G denote syringyl and guaiacyl units. C_n indicates the number of carbon atoms in the C4 alkyl side chain (e.g., C₃ = propyl), and terminal labels (e.g., −2OH) denote side‐chain functional groups.

Most of the identified oligomers were found to contain at least one β‐O‐4 linkage, confirming that a significant fraction of native ether linkages was preserved during flow‐through solvolysis. In addition, β‐β linkages were identified in several oligomers, including C₃₃H₄₀O₁₃, C₃₄H₄₂O₁₃, C₄₃H₅₂O₁₇, and C₄₄H₅₄O₁₇. Compared with β‐O‐4 linkages, β‐β linkages exhibit substantially higher stability and generally require higher collision energies to undergo dissociation. These β‐β linkages originate from resinol‐type structures in lignin, which are formed through β‐β radical coupling of two monolignols followed by intramolecular cyclization and therefore contain two internal ether bonds. During reductive lignin depolymerization, these internal ether bonds can be cleaved to generate ring‐opened resinol structures while preserving the central C‐C linkage.

4. Conclusions

Lignin‐derived products generated from methanol‐based flow‐through solvolysis of birch biomass were investigated through untargeted molecular‐level characterization based on HPLC‐UVD/MSⁿ. Distinct chromatographic behaviors associated with molecular size and degree of polymerization enabled differentiation of monomeric and oligomeric species and provided insights into their structural characteristics. Accurate‐mass measurements and tandem MS analysis revealed the compositions and interunit linkages of representative dimers, trimers, and tetramers. Characteristic fragmentation pathways provided direct evidence for the preservation of native β‐O‐4 linkages and β‐β resinol motifs in the flow‐through solvolysis products. These findings provide molecular‐level insights into the composition, structural diversity, and preservation of native‐like lignin oligomers directly generated from biomass, thereby bridging the gap between model‐compound investigations and real biomass‐derived lignin systems and improving our understanding of lignin depolymerization and valorization pathways.

Author Contributions

Zhongyue Zhou: conceptualization, funding acquisition, writing – original draft, writing – review and editing. Linyu Zhu: investigation, data curation, writing – original draft, writing – review and editing. Xintong Xiao: visualization, writing – review and editing. Cunhao Cui: methodology, investigation. Fei Qi: supervision, writing – review and editing.

Funding

This work was supported by the National Natural Science Foundation of China (22574105).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: (a) HPLC‐MS spectrum for the birch sample (average over RT from 8 to 26 min), revealing well‐resolved lignin oligomers from tetramers to octamers. (b) EIC of lignin‐derived oligomers from solvolysis of birch wood with DP from 4 to 8 obtained by HPLC‐MS, showing an overall increase in retention time with molecular size.

Figure S2: Tandem MS spectra of dihydroconiferyl alcohol.

Figure S3: Tandem MS spectra of homovanillyl alcohol.

Figure S4: Tandem MS spectra of vanillin.

Figure S5: Tandem MS spectra of p‐hydroxybenzoate.

Figure S6: Tandem MS spectra of apocynin.

Figure S7: Tandem MS spectra of coniferyl alcohol.

Figure S8: Tandem MS spectra of coniferyl aldehyde.

Figure S9: Tandem MS spectra of homovanillic acid methyl ester.

Figure S10: Tandem MS spectra of dihydrosinapyl alcohol.

Figure S11: Tandem MS spectra of syringic acid.

Figure S12: Tandem MS analysis of GGGE. (a) MS2 spectra. (b) and (c) MS3 spectra.

Figure S13: Tandem MS spectrum of [C31H39O13]− (Entry 1 in Table 4).

Figure S14: Tandem MS spectrum of [C32H39O15]− (Entry 2 in Table 4).

Figure S15: Tandem MS spectrum of [C33H39O13]− (Entry 3 in Table 4).

Figure S16: Tandem MS spectrum of [C34H41O13]− (Entry 4 in Table 4).

Figure S17: Tandem MS spectrum of [C43H51O17]− (Entry 5 in Table 4).

Figure S18: Tandem MS spectrum of [C44H53O17]− (Entry 6 in Table 4).

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.