Introduction

Lignin is a highly underutilized commodity available from lignocellulosic agricultural biomass waste. Over 5 billion metric tons of lignin waste are produced from pre-consumer biomass annually [1]. Paper production also contributes approximately 600,000 tons per year of additional lignin waste [2]. In terms of potential commodities for carbon sequestration, lignin notably sequesters ~ 30% of organic carbon on the planet [1]. A barrier to lignin utilization is its complex structure deriving from its biosynthesis that links primarily 4-p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol, giving rise to the p-hydroxyphenyl, guaiacyl, and syringyl subunits, respectively, in lignins (Fig. 1) [3]. Not only is lignin a mixture of these diverse subunits, but their relative contribution to a particular lignin also varies with the part of the plant from which it is derived and the plant species. Lignin from grasses, for example, contains significant quantities of all three structural subunits, while hardwood lignin features significant quantities of guaiacyl and syringyl subunits but few p-hydroxyphenyl subunits, and softwood lignins feature primarily guaiacyl subunits [1].

Fig. 1
figure 1

Lignin biosynthetic precursors (A) and structure showing features of some lignins (B)

Full size image

Efficient valorization of lignin involves both “upstream” and “downstream” processes. Upstream processes includes the separation and isolation of lignin, lignin bioengineering, and the catalytic conversion of lignin, while downstream processes includes both lignin depolymerization as well as lignin upgrading [4]. Among the most promising efforts to improve lignin utilization is through depolymerization, giving bio-oils or lignin oils [5]. In regards to sample preparation, it is important to find green methods when considering solvent consumption. A few of these “green” solvents used for extraction procedures include, supercritical fluids, bio-based solvents, and liquified gases. Supercritical fluid extraction (SFE) involves shorter extraction time, improved efficiency, better selectivity, and the extracting solvent is easy to remove. Liquified gases such as butane and n-propane are advantageous because they require gentle pressure (< 1 MPa) to stay a liquid and is easily evaporated at low temperatures. Thus, liquified gas extraction can be performed at room temperature with little energy consumption and the extracts containing a negligible residual amount of solvent. Bio-based solvents can be obtained from a wide variety of biomass sources. They can be obtained through methods such as extracting vegetable oils, carbohydrate fermentation, and the steam distillation of wood [6]. Ethanol is a solvent of particular interest due to its availability via the fermentation of renewable resources such as starches, sugars, and lignocellulosics. Ethanol is also fairly low cost and readily available [7].

A recent report [8] on mild thermolytic solvolysis of lignin to yield a solubilized lignin oil was particularly attractive for the current work because the process could be accomplished at low temperatures (100–350 °C), using preferred green solvents [9] like ethanol. With ethanol at only 100 °C, this process conveniently accomplished conversion of 64 wt. % of lignin to lignin oil.

Elemental sulfur is another abundant and underemployed by-product of human industry, and is produced at a rate of 70 million metric tons per year from fossil fuel refining [10, 11]. In recent years, we have reported several efforts to prepare composites combining lignin and sulfur via the inverse vulcanization method [12,13,14]. It has proven difficult to obtain inverse vulcanized polymers that possess concurrent high strength, toughness, and elongation [15]. Additional mechanisms for S–C bond-formation will not only expand the organic substrate scope for preparing high sulfur-content materials (HSMs) but may also achieve more favorable mechanical properties.

Because classic inverse vulcanization requires olefins, prior lignin-sulfur composite synthesis required modification of lignin with olefins before reaction with sulfur [13, 14]. Allyl-derivatized lignin composites LSx (x = wt. % sulfur in monomer feed, varied from 80 to 99) and ELSx@T (where x = wt. % sulfur in the monomer feed, varied from 80 to 90, and T is the reaction temperature in °C) achieved the goal of producing high strength lignin-sulfur composites, but additional synthetic / separations / use of non-green solvents detract from the greenness, atom economy, and affordability of those processes.

A more efficient one-pot synthesis of lignin-sulfur composite CLS90 was developed by reacting sulfur (90 wt. %) with chlorolignin (10 wt. %), a waste product of paper/pulping. Preparation of CLS90 occurred via radical-induced aryl halide-sulfur polymerization (RASP, Scheme 1B) rather than inverse vulcanization [16]. RASP involves S–Caryl bond formation via thermal generation of aryl radicals via C–Halide σ-bond scission and subsequent reaction of aryl halides with S–S bonds. Although this process proved valuable as a simple one-pot synthesis of durable composites from two industrial waste products, it produces 0.5 equivalents of toxic S2Cl2 for each S–C bond formed, making RASP generally less ecologically desirable and less atom economical than inverse vulcanization.

Scheme 1
scheme 1

Major S–C bond-forming reactions in the formation of GS80 from sulfur and guaiacol

Full size image

Given results from the aforementioned lignin-sulfur composites (LSx, ELSx@T, and CLS90), it was of interest to investigate other possible S–C bond-forming reactions involving lignin derivatives. Lignin itself does not contain alkenes or halogens, but features a high abundance of anisole derivative subunits. We recently unveiled new S–C bond-forming mechanisms that occur when elemental sulfur is heated above 220 °C with anisole derivatives like O,O′-dimethylbisphenol A, [17] guaiacol, [14] and others [18]. These studies revealed a range S–C bond-forming processes. The S–C bond-forming reaction of guaiacol (80 wt. %) and sulfur (20 wt. %) at 230 °C, for example, gave composite GS80, comprising sulfur catenate crosslinkers attached to aryl units (Scheme 1) [14]. This is especially pertinent to the current study because guaiacol is a common constituent of many lignin oils. GS80 exhibited improved tensile strength and elongation at break compared to most HSMs derived from renewable organics that had been reported at the time of its synthesis.

Given the range of S–C bond-forming reactions unveiled in previous studies and the ability of such bond-forming reactions to stabilize composites of guaiacol and sulfur, we hypothesized that a durable composite would be attainable by the reaction of elemental sulfur with lignin oil as an alternative to whole lignin that can be used without prior derivatization. Herein, composite LOS90 was synthesized by heating lignin oil (10 wt. %) with elemental sulfur (90 wt. %) at 230 °C for 24 h. A series of model compounds (Fig. 2) featuring functional groups present in lignin subunits or thermal degradation products previously observed (Scheme 2) were reacted with elemental sulfur under identical conditions to further inform on S–C bond-forming reactions operable in LOS90. Properties of composite LOS90 were characterized using differential scanning calorimetry (DSC, mechanical test stand analysis, thermogravimetric analysis (TGA), infrared spectroscopy, flexural strength analysis, and scanning electron microscopy with elemental mapping by energy-dispersive X-ray analysis (SEM–EDX). These studies reveal that LOS90 is a microscopically homogeneous composite with good thermal stability and compressive strength higher than is required of ordinary Portland cement for residential building foundations.

Fig. 2
figure 2

Lignin derivatives/model compounds reacted with elemental sulfur in the current study to inform on the range of chemistries operational during synthesis of LOS90 from lignin oil

Full size image
Scheme 2
scheme 2

Reaction of model compounds 2–6 with elemental sulfur provides insight into reactivity between elemental sulfur and variably-functionalized aryl units

Full size image

Materials and Methods

Materials

Sulfur powder (99.5%) was purchased from Alfa Aesar. Compounds 6-methylguaiacol (> 98%) and 1,2-dimethoxybenzene (99%) were purchased from TCI (Tokyo Chemical Industry). Compounds o-cresol (> 99%) and syringol (99%) were purchased from Sigma Aldrich. Catechol (99%) was purchased from Acros Organics. These chemicals were used without further purification. The lignin oil used in this study was produced by a reported method using thermal solvolysis of kraft lignin in ethanol [8].

General Considerations and Instrumentation

All 1H NMR spectra were recorded on a Bruker Avance spectrometer operating at 300 MHz.

GC–MS studies were carried out on a Shimadzu QP2010SE system with an auto-injector (AOC-20i) equipped with the mass selective detector, having an interface temperature of 250 °C, a solvent cut time of 3.00 min, threshold of 70 eV and mass range of 45 to 900 m/z. Compounds were separated using a SH-Rxi-5 MS capillary column (Restek Company, Bellefonte, USA: crossbond 5% diphenyl/ 95% dimethyl polysiloxane) having dimensions 30 m (length) × 0.25 mm (diameter) × 0.25 μm (film thickness). The temperature of the injector was initialized to 250 °C. The temperature was programmed to 50 °C for 1 min, then from 50 °C to 320 °C at a rate of 20 °C/min, then held at 320 °C for 4 min.

Fourier transform infrared spectra were obtained using an IR instrument (Shimadzu IRAffinity-1S) with an ATR attachment. Scans were collected over the range 400–4000 cm−1 at ambient temperature with a resolution of 8 cm−1.

SEM was acquired on a Schottky Field Emission Scanning Electron Microscope SU5000 operating in variable pressure mode with an accelerating voltage of 15 keV.

Compressional strength measurements were undertaken with cylinders using a Mark-10 ES30 (USA) Manual Test Stand equipped with a Mark 10 M3-200 Force Gauge (USA) by a modified ASTM C39 standard. The LOS90 composites were melted, formed in molds, and then cooled to room temperature. The cylinders were aged for 4 d before they underwent compressional strength testing. The stability for these materials longer-term is unknown.

Flexural strength analyses were carried out using a Mettler Toledo DMA 1 STARe System in single cantilever mode. The samples were made using silicon resin moulds. The samples dimensions were 1.5 × 10.7 × 5.0 mm. The temperature was 25 °C and the clamping force was 1 cN m. The samples were tested in triplicates and the reported flexural strength and moduli is an average of the three trials.

TGA data were recorded (Mettler Toledo TGA 2 STARe System) over the range 20–800 °C with a heating rate of 10 °C·min−1 under a flow of N2 (100 mL·min−1).

DSC data were recorded using a Mettler Toledo DSC 3 STARe System with a temperature range of –60 °C to 140 °C with a heating rate of 10 °C·min–1 under a flow of N2 (200 mL·min–1). Each DSC measurement was performed over three heating and cooling cycles. The data reported herein are from the third cycle. The first cycle removes any solvent impurities. The melting and crystallization transitions and the glass-transition temperature were seen in during the third cycle. The change of percentage crystallinity between the sulfur and the composites was calculated from the DSC data utilizing the following equation:

$$\Delta {\chi }_{c}=1-\left\{\frac{\Delta {H}_{m}-\Delta {H}_{cc }}{\Delta {H}_{m(S)}-\Delta {H}_{cc(S)}}\right\} \times 100\%$$

where ΔXc is change of percentage crystallinity with respect to sulfur, ΔHm is the melting enthalpy of the composite, ΔHcc is the cold crystallization enthalpy of the composite, ΔHm(S) is the melting enthalpy of sulfur, and ΔHcc(S) is the cold crystallization enthalpy of sulfur.

Synthetic Procedure Safety Note

CAUTION: Heating elemental sulfur with organics can result in the formation of H2S or other gases. Such gases can be toxic, foul-smelling, and corrosive. Temperature must be carefully controlled to prevent thermal spikes, contributing to the potential for H2S or other gas evolution. Rapid stirring shortened heating times, and very slow addition of reagents can help avoid unforeseen temperature spikes.

Model Compound Reactivity Studies

For each of the compounds shown in Fig. 2, 1.600 g elemental sulfur and 0.400 g of the organic compound were added directly into a glass pressure tube under an atmosphere of dry nitrogen gas in a VTI glovebox. Each pressure tube was equipped with a magnetic stir bar and sealed under N2 with a Teflon screwcap having a Viton O-ring. Each tube was placed into an oil bath heated to 230 °C. Heating was continued for 24 h with continuous stirring by a magnetic stir bar. After 24 h of heating the pressure tubes were cooled to room temperature before opening. No mass loss was observed from any of the tubes. For 1H NMR analysis, 50 mg of each product mixture was placed in 1.0 mL of CDCl3 and shaken to facilitate dissolution. The mixture was then filtered through a 0.2 µm PTFE syringe filter and transferred to the NMR tube. GC–MS samples, 20 mg of each reaction mixture was extracted into CH2Cl2 followed by filtration through a 0.2 µm PTFE syringe filter.

Synthesis of LOS90

Elemental sulfur (45.0 g) was added to a large Erlenmeyer flask and then placed in an oil bath. The sulfur was then heated to 180 °C and allowed to melt until a reddish liquid was formed. The lignin oil (5.00 g) was then added to the flask and covered with aluminum foil. Rapid mechanical stirring was then applied. The temperature was then increased to 230 °C. The mixture was heated for 2 h under continuous rapid mechanical stirring. Within the 2 h, the reaction mixture turned dark brown in color and appeared homogeneous. After the 2 h, the stirring was stopped, and the reaction was removed from the heat. The resulting product was allowed to cool to room temperature, forming a dark brown solid. To form the cylinders for compressive strength tests and rectangular prisms for flexural strength tests, the product was remelted at 180 °C then poured into molds and cooled to room temperature. Elemental analysis theoretical: found: C 3.08, H 0.16, S 94.17.

Dark Sulfur Quantification

This procedure follows a modified literature procedure from Hasell’s group [19] employing toluene to extract and quantify dark sulfur. Extraction of dark- sulfur was performed by suspending 100 mg of finely ground LOS90 in 20 mL of toluene in a centrifuge tube. After an hour, the tube centrifuged for approximately 30 s, allowing the solid to collect at the bottom of the tube. The supernatant was then transferred by pipette into a round bottom flask. The toluene was evaporated from each fraction under reduced pressure and both solids weighed to determine the amount of dark sulfur and insoluble material. Elemental analysis confirmed that the soluble material was > 98.5% sulfur.

Results and Discussion

Synthesis and Chemical Characterization of Composite

The lignin oil used in this study was produced by a reported method using thermal solvolysis of kraft lignin in ethanol (1:5 w/v ratio) at 100 °C. This route to lignin oil was selected because of the relatively low temperature and use of a preferred green solvent [8]. Analyses of the lignin oil by 2D HSQC NMR spectrometry (Fig. S1, Supplementary Material) and MALDI (Fig. S2, Supplementary Material) support the same composition as was previously reported for lignin oils prepared by this method (illustrative structural subunits are provided in Fig. S3 in the Supplementary Material). The fragmentation of the whole lignin into lower molecular weight monomers was further confirmed by MALDI-TOF spectrometry, which verified the presence of significant quantities of whole lignin as well as of higher oligomers consistent with reported composition of the lignin oils prepared by this method. (Fig. S2 in the Supplementary Material).

Preparation of LOS90 involved the reaction of lignin oil (5.0 g, 10 wt. %) and sulfur (45 g, 90 wt. %). Sulfur was first melted in an Erlenmeyer flask and heated to 180 °C, followed by the addition of lignin oil with rapid mechanical stirring. The temperature was increased to 230 °C under continuous stirring and the mixture was stirred for 2 h while the top of the flask was covered with aluminum foil. The mixture was visually homogenous throughout the heating period, allowing a much shorter reaction time than the 24 h of heating required for synthesis of lignin-sulfur composite LSx. LOS90 solidified into a dark brown, remeltable solid after cooling to room temperature, in an isolated yield of 98.5% of LOS90. Heating sulfur with organics that have readily abstractable H atoms can produce toxic hydrogen sulfide gas, with attendant mass loss of up to 34% [20]. The high isolated yield of LOS90 and the lack of discoloration of the aluminum foil covering both indicate negligible if any hydrogen sulfide generation in the current case. Upon melting, the material was poured into molds and allowed to cool to room temperature to form samples for both compressive and flexural strength tests.

In HSMs the sulfur present as crosslinking –Sx– chains is generally accompanied by small quantities of physically entrapped sulfur species not covalently linked to the organic comonomers. The physically entrapped sulfur is known as dark sulfur and its contribution to an HSM is generally determined by its extraction and quantification either by UV–vis spectroscopy or by evaporating the extraction solvent and quantifying the mass directly [19, 21]. In the current case, dark sulfur was extracted with toluene, followed by evaporation of the toluene and weighing the extractable material. Elemental analysis was also performed on the soluble fraction, confirming it was > 98% sulfur. The composite LOS90 comprised 76 wt. % insoluble material and 23 wt. % dark sulfur, demonstrating high incorporation of sulfur into the polymer network. LOS90 exhibits a similar dark sulfur content to the 16 wt. % dark sulfur content of LS90 (a composite prepared from allylated lignin with 90 wt. % sulfur).

Evidence of C–S bond formation between the lignin oil constituents and the polymeric sulfur was observed by infrared spectroscopy (Fig. 3). A peak was observed at 660 cm–1 indicating the C–S stretch [22,23,24]. The IR spectra (Fig. S4, Supplementary Material) also confirmed that LOS90 retained characteristic peaks for the lignin oil such as O–H stretching (3200–3500 cm–1), C–H stretching (2930 cm–1), and C–O stretches (1030–1200 cm–1).

Fig. 3
figure 3

Portion of the FT-IR spectrum of LOS90 (blue trace) compared to that of lignin oil (orange trace). Full spectra are provided in the Supplementary Material (Fig. S4) (Color figure online)

Full size image

Scanning electron microscopy (SEM, Fig. S5, Supplementary Material) imaging with elemental mapping by energy dispersive X-ray analysis (EDX) was performed to assess the distribution of components in the LOS90 composite (Fig. 4). SEM/EDX data showed homogeneous distribution of S, C, and O. Small dots of silica-rich material observed are residue from the SEM sample preparation.

Fig. 4
figure 4

Elemental mapping of carbon (green, upper left), oxygen (blue, upper right), and sulfur (lower right) in LOS90 by energy dispersive X-ray analysis (EDX) (Color figure online)

Full size image

Model Compound Reactivity Studies

Compounds 2–6 (Fig. 1) were selected as models of lignin oil components that may engage in S–C bond-forming reactions like those we previously reported for guaiacol (1, Scheme 1C). Each compound 2–6 was reacted with elementals sulfur under conditions identical to those used to prepare LOS90. Following reaction, the soluble compounds were extracted under conditions that will give thiols at each position where a crosslinking –Sx– chain is expected to attach to the organic scaffold in such units in LOS90. This allowed for identification of small molecular species present in the reaction mixtures by GC–MS analysis (Figs. S6–S65 in the Supplementary Material). The observed chemistry is summarized in Scheme 2.

Compounds 2b, 3b, 4a, and 5a each result from forming one S–Caryl bond. There does not appear to be a strong regioselectivity trend for the formation of these S–Caryl bonds, as all three of the possible regioisomers of 2b, one of the three possible regioisomers of 3b, both regioisomers of 4a, one regioisomer of 5a, and two of the four regioisomers of 5c were observed. Two other products featuring one new S–Caryl bond each, 4e and 5c, were also observed, but S–C bond formation was accompanied by additional backbone modifications described below.

Several species featuring two S–Caryl bonds were also observed, modelling modes of crosslinking in LOS90. A sulfur atom joins two aryl rings in compounds 2c (three regioisomers observed) and 4b. For 4b, only one regioisomer forms as a result of the steric encumbrance at other potential attachment sites. Similar diaryl thioether derivatives 1b were reported from reaction of 1 with sulfur (Scheme 1C).

In one instance the formation of two S–Caryl bonds to the same aryl ring was observed to give product 6a (two regioisomers observed). One compound, 4f, was observed in which two aryl rings were joined by a new Caryl–Caryl bond, and each aryl ring featured a thiol substituent as a result of two newly-formed S–Caryl bonds.

Some products observed from reactions of 4 and 5 involve scission of the methoxy C–O σ-bond, also observed in reactions of guaiacol with sulfur (Scheme 1C). Sulfur is not required to facilitate this pathway, however, as heating pure 4 without sulfur leads to similar C–O σ-bond scission [25]. Loss of one or both methyl groups from 4 led to products 4c and 4d, while loss of one of the methyl groups from 5 led to 5b. In addition to the loss of methyl groups, 4 also underwent O–Caryl σ-bond scission with a net loss of one hydroxy/methoxy unit from the structure, accompanied by S–Caryl bond formation to give 4e. Loss of methoxy groups via O–Caryl σ-bond scission is also a known thermal decomposition route for 4 in the absence of sulfur [25].

Thermal and Morphological Properties of LOS90

The thermal stability of lignin oil and LOS90 were assessed using thermogravimetric analysis (TGA, Fig. 5 and Table 1).

Fig. 5
figure 5

Thermogravimetric analysis (TGA) traces for LOS90 (blue trace) and lignin oil (orange trace) (Color figure online)

Full size image
Table 1 Thermal and morphological properties of the lignin oil-sulfur composites LOS90 with comparison to sulfur and other composites
Full size table

The lignin oil decomposed over a broad temperature range and showed three decomposition events at 93 °C, 159 °C, and 309 °C. The first decomposition temperature is attributable to loss of entrapped ethanol (observed by NMR analysis and not removable under vacuum) and water. The decomposition feature with onset at 159 °C is due to the degradation of the propanoic side chains in whole lignin oligomers, while the decomposition feature with onset of at 309 °C is attributed to the cleavage of β–β and C–C linkages in lignin oil oligomers [26]. The TGA trace for LOS90 showed a single decomposition event revealing a decomposition temperature (Td) of 229 °C, typical for HSMs and for elemental sulfur (Td of 229 °C) and attributable to sublimation of entrapped cyclo-S8 in LOS90. The smaller decomposition event at > 300 °C in LOS90 is attributable to decomposition of the lignin oil-derived organics.

Differential scanning calorimetry (DSC, Fig. 6) of the LOS90 composites revealed thermal transitions associated with cyclo-S8 and polymeric sulfur. LOS90 exhibited a melting peak expected for cyclo-S8 at 114 °C, consistent with the effective removal of entrapped cyclo-S8 by toluene extraction. The shoulder to the low-temperature side of the prominent melt transition is attributable to sulfur’s α-to-β phase transition. A glass transition temperature (Tg) was also observed in the third heating cycle at –36 °C, diagnostic for polymeric sulfur. The melting and cold crystallization enthalpies was used to calculate the percent crystallinity of LOS90 compared to orthorhombic α-sulfur. Based on these data, the crystallinity of LOS90 was 57% that of α-sulfur, quite similar to that of LS90 (67%) that, like LOS90, comprises 90 wt. % sulfur.

Fig. 6
figure 6

DSC thermogram (top) for LOS90 (first heating cycle (blue line) and first cooling cycle (orange line)). The bottom image is an inset that more clearly reveals the glass transition temperature (Tg) observed in the third heating cycle. Endothermic transitions are downward in these thermograms (Color figure online)

Full size image

Mechanical Properties and Environmental Impact Considerations

Elucidating the mechanical properties of LOS90 (Table 2) was undertaken to assess its viability as an alternative to less sustainable materials. Previous work [27,28,29] has shown that HSMs often have compressive strengths competitive with legacy mineral products used in structural applications. For example, ordinary Portland cement (OPC) generally has a flexural strength greater than 3 MPa and is required to have a compressive strength greater than 17 MPa to meet minimum specifications for use in residential building foundations according to American Concrete Institute code ACI 332.1R-06. Published data on the compressive strengths of sulfur-lignin derivative composites in particular are limited. The compressive strength of LOS90 (22.1 MPa, Fig. 7 and Table 2, stress–strain plots are provided in Fig. S67) exceeded the 17 MPa requirement. The compressive strength of LOS90 is also similar to those of composites comprising 10 wt. % whole lignin derivative and 90 wt. % sulfur (Table 2), for example ELS90@180 (20.1 MPa) and ELS90@230 (26.8 MPa), while being considerably higher than that of ELS80@180 (10.9 MPa) [13]. The preparation of ELSx@T precursor oleic-esterified lignin is a much more energy and time-intensive process than that used to prepare the lignin oil required for LOS90 preparation, so LOS90 is a more viable construct for attaining similar compressive strength characteristics. It also valuable to compare the cost of bulk sulfur to the cost of ordinary Portland cement. When purchased at scale (as of February 26, 2024), bulk sulfur can be purchased for $80 per metric ton and lignin for $100 per metric ton (all costs are provided in USD). Taking the 64% conversion of whole lignin to lignin oil, making one ton of LOS90 requires 0.16 tons of whole lignin to give the 0.1 tons of lignin oil to combine with the 0.9 tons of sulfur, for a material cost of $88 per metric ton, somewhat lower than the average price of OPC ($130 per metric ton) [30]. These material cost estimates do not include the ethanol for use in lignin oil production (ethanol would be recovered for reuse on an industrial scale) or the water used for OPC curing. The comparable material costs are promising especially in light of ongoing rising costs of unsustainable resources.

Table 2 Mechanical properties of the lignin-sulfur composites LOS90 with comparison to other composites
Full size table
Fig. 7
figure 7

Comparison of flexural strength (blue bars) and compressive strength (orange bars) of LOS90 with lignin-sulfur composites (LSx), lignin/cellulose containing sulfur composites (ELSx@T, CLS80, mAPS95), normalized to values of ordinary Portland Cement (17 MPa for compressive strength and 3.5 MPa for flexural strength). Bars not displayed indicate that the attendant value was not previously reported for the material listed (Color figure online)

Full size image

The flexural strength and moduli of LOS90 was tested and evaluated, revealing high flexural strength (5.7 MPa) and moduli (186 MPa) at room temperatures compared to previously reported lignin sulfur composites (Table 2, stress–strain curves are provided in Fig. S68). The aforementioned ELSx@T composites, for example, had flexural strengths of 2.7–3.9 MPa. Composites LSx (x = wt. % sulfur), prepared by inverse vulcanization of allylated lignin, showed flexural strengths (1.5–2.1 MPa) of less than 50% that of LOS90 and moduli of 57–87 MPa. A chlorolignin-sulfur composite (CLS80) comprising 80 wt. % sulfur and 20% chlorolignin likewise exhibits somewhat lower flexural strength (3.6 MPa) than that of LOS90. Whereas the flexural strength of LOS90 exceeds those of other sulfur-lignin HSMs comprising high (80–90 wt. %) sulfur, a similar flexural strength of 5.6 MPa has been reported for mAPS95 (prepared by inverse vulcanization of 2 wt. % allyl lignin, 3 wt. % allyl cellulose, and 95 wt. % sulfur). The flexural strength and modulus has also been reported for other high sulfur polymers with organic crosslinkers. Dicyclopentadiene (DCPD), a cheap industrial by-product, was reacted with elemental sulfur to form S-DCPD (50 wt. % sulfur and 50 wt. % dicyclopentadiene) [31]. The flexural strength of LOS90 (5.7 MPa) compares fairly to S-DCPD which exhibits a flexural strength of 6.0 MPa. However, when organic crosslinkers are incorporated with S-DCPD, the flexural strength decreases. When linseed oil is reacted with sulfur and DCPD to form S-DCPD-linseed (50 wt. % S, 25  wt. % dicyclopentadiene, 25 wt. % linseed oil), a slightly lower flexural strength of 4.7 MPa is observed. However, when limonene, a low-cost biomass material, was reacted with sulfur and DCPD to form S-DCPD-limonene (50 wt. % sulfur, 25 wt. % dicyclopentadiene, 25 wt. % limonene), a significantly lower flexural strength of 1.9 MPa was observed. However, the flexural moduli of S-DCPD (3700 MPa), S-DCPD-linseed (1250 MPa), and S-DCPD-limonene (1750 MPa) [31] greatly exceed that of LOS90 (186 MPa). It is also important to note that LOS90 exceeds the flexural strength of OPC (3.7 MPa), making it competitive with commercial building materials in this regard as well.

The viscoelastic properties of LOS90 were assessed using dynamic mechanical analysis (DMA). DMA data (Table 3 and Fig. S69 in the Supporting Information) were collected over the temperature range of –60 to + 80 °C in order to analyze the temperature dependence of storage modulus (E′), loss modulus (E"), and the damping factor (tan δ). The Tg values of –38, –38, and –40 °C observed from tan δ (local maximum), loss modulus (local maximum) and storage modulus (inflection point) agree well with the Tg of –36 °C observed from DSC data (Fig. 6 and Table 1). The transitions observed above 20 °C in both the storage and loss moduli curves are attributable to chain freedom of relative movement in less-ordered parts of the polymer film, [32] a feature observable for LS90 [12] and several other HSMs [33, 34] comprising polymeric sulfur chains as crosslinkers similar to those in LOS90.

Table 3 Data from dynamic Mechanical Analysis of storage modulus (E′), loss modulus (E″) and damping factor (tan δ) from –60 to 80 °C
Full size table

The E factor [35] is a useful metric for estimating of the relative greenness a chemical process, where E factor is equal to mass of waste produced divided by mass of useful product produced, such that a lower E factor is more desirable. Considering the process as beginning with whole lignin and elemental sulfur, an E factor of 0.057 was calculated for LOS90, with the waste generated being lignin char left behind from the whole lignin not converted to the oil. The E factor was calculated according to the following equation:

$$E\mathrm{ factor}= \frac{{\text{waste}}}{\mathrm{useful product}}= \frac{\mathrm{lignin char}}{({{\text{LOS}}}_{90})}=\frac{2.81\mathrm{ g}}{(49.25\mathrm{ g})}=0.057$$

The ethanol used is recovered by distillation for reuse, so is not considered as a waste product in this calculation. Other sulfur-lignin derivative composites have E factors that are 1–2 orders of magnitude higher (Table 4) and produce notably more environmentally detrimental waste products than the insoluble lignin char that is the only waste material resulting from LOS90 synthesis. For comparison, the E factors for bulk chemicals range from < 1 to over 50 [36]. All of the E factors for sulfur-lignin derivative composites are notably lower than that of OPC, however. The E factor for LOS90 is also over an order of magnitude lower than that of OPC (1.4) [36]. Moreover, the primarily waste product of Portland cement production is CO2, with about 1 kg of CO2 released for every kilogram of cement produced, such that Portland cement production is responsible for ~ 8% of all anthropogenic CO2 production [37]. These data emphasize the potential of waste-derived cements or geopolymer cements [38, 39] to replace legacy building materials in a future green economy.

Table 4 E factors for reported sulfur-lignin derivative composites
Full size table

Conclusion

Lignin oil made at low temperature (100 °C) in ethanol, a preferred green solvent, undergoes reaction with fossil fuel waste sulfur at 230 °C for two hours. The resulting composite, LOS90, has compressive and flexural strengths exceeding those required for use in load-bearing components of the built environment. The method described herein for preparing LOS90 satisfies numerous facets of the Principles of Green Chemistry, such as achieving a 98.5% atom economy from lignin oil and sulfur to the product, thereby minimizing waste and facilitating the complete integration of reactants into the final product. Moreover, this method uses ethanol, a green solvent, utilizes renewable plant-derived lignin, and provided a valuable avenue for sulfur utilization. Given that OPC has high blue and grey water footprints and its manufacture accounts for 8% of anthropogenic CO2, accomplishing this greener synthesis of a material having compressive strength appropriate for replacing OPC in some structural applications also holds promise for greener construction practices in the future. In addition to the fundamental chemical and mechanical tests reported herein, however, biological and environmental studies on toxicity and the potential of the materials for soil acidification should to be undertaken by those seeking to use the new composites in specific contexts to validate the safety of materials in those proposed applications.