Production and distillative recovery of valuable lignin-derived products from biocrude

Abstract

Value-added chemicals present in liquid products from direct liquefaction of biomass can be recovered as bioproducts in an integrated biorefinery to enhance the economic viability of biomass-to-liquid fuel technologies and afford a higher return on investment. In recent efforts, RTI International is exploring and developing processes that could be used to produce bioproducts in addition to liquid fuel from biocrude intermediates. The present study focused on the production and recovery of high-value lignin-derived compounds from biocrude. Catalytic pyrolysis with a nonzeolite, alumina-based catalyst was used to produce the biocrude from loblolly pine in RTI's 1-tonne per day pilot-scale plant. Distillation was evaluated as a potential separation method given that the biocrude is relatively thermally stable and distillable compared to raw pyrolysis oils. The distillation studies were performed with a laboratory distillation unit (PETRODIST 300 CC) and the biocrude was fractionated into four distillate cuts: cut 1 (IBP-110 °C), cut 2 (110–200 °C), cut 3 (200–300 °C), and cut 4 (300–400 °C). In a case study, cut 3 was distilled further to concentrate the lignin-derived chemicals boiling between 200 and 270 °C. The results showed that a biocrude containing about 10 wt% guaiacols could be concentrated to about 35 wt% of guaiacols (representing 75% recovery) in the first distillation step. The second distillation step increased the concentration of guaiacols to about 53 wt% (representing 80% recovery). These findings suggest that it is possible to isolate and concentrate the lignin-derived products using a distillation column with a number of plates/trays. The yields, physicochemical properties, and chemical composition of the fractions, as well as the overall distillation separation efficiency, are reported.

---

Introduction

The production of biochemicals from biomass has gained renewed interest and has experienced a market “pull” by major companies in recent years¹ partly due to consumer demand for more sustainable and environmentally benign products. In 2015, biochemicals were expected to constitute 10% of the chemical market.² Moreover, the industry may have the potential to produce two-thirds of the total volume of chemicals from biobased materials, representing an annual market of approximately $1 trillion.¹ According to a recent U.S. Department of Agriculture (USDA) report, the U.S bio-based products industry contributed $369 billion to the U.S. economy with 4 million American jobs. Additionally, each job in the biobased products industry was responsible for generating 1.64 jobs in other sectors of the economy.²

Biomass pyrolysis is one of the promising technology options available for producing biobased fuels and chemicals. Uniquely, the liquid product (bio-oil) from biomass fast pyrolysis contains numerous oxygenated compounds^3,4 that are not easily synthesized from petroleum feedstocks, such as, aldehydes, ketones, alcohols, esters, anhydrosugars, furans, phenols, guaiacols, syringols, and large molecular weight oligomers.^5,6 Importantly, some of these compounds could be recovered as specialty chemicals or as building blocks for synthesis of other bio-based products. A number of the chemicals have ready markets and could be used to improve the environmental performance of similar products obtained from petroleum-based sources. Strategically, these chemicals could be produced alongside biofuels to enhance the economic viability of biomass-to-liquid fuel technologies and afford a higher return on investment⁷ as it is being practiced in the petroleum industry. Therefore, it is vital to develop efficient separation technologies for recovery of high-value chemicals from the liquid products of direct biomass liquefaction processes.

The recovery of chemicals from bio-oils has been extensively studied over the past several years. Many previous studies have explored options of extracting high-value chemicals from bio-oils.^8,9 For example, the British Columbia Research Corporation developed a method for extraction and recovery of levoglucosan at high purity.¹⁰ The recovery of acetol, phenols, hydroxyacetaldehyde, and glyoxal have also been studied by other research groups.^8,9 In fact, several attempts have been made to recover phenolic compounds from bio-oil.^11–19 Fractionation of bio-oil into different streams of similar functionalities has indeed enhanced chemical recovery.^12,20–23 Nevertheless, many of these efforts have not yet resulted in the production of a marketable chemical product. The technical challenge is fundamentally due to the chemical complexity of the mixture, and because the concentration of any single compound is too low to warrant efficient recovery.

Selective processes other than conventional pyrolysis have the potential of converting biomass into a liquid with a narrower chemical composition, thereby enhancing the ability to effectively recover chemical components of interest by simple separation methods. It is well known that catalytic biomass pyrolysis is relatively more selective than conventional pyrolysis, because the product slate can be controlled by applying catalysts and optimizing operating conditions. In many instances, the product slate is narrower and the chemicals of interest are in relatively high yields, making it much more feasible for their recovery. For example, the use of zeolite catalysts such as ZSM-5 results in the production of aromatic hydrocarbons including benzene, toluene, xylenes (BTX), and naphthalenes.²⁴ Likewise, the use of metal oxides like CeO₂ and TiO₂ lead to the formation of ketones, furans, and phenols.²⁵ Other catalysts have also been used to increase production of levoglucosan, levoglucosenone, furfural, hydroxyacetaldehyde, acetic acid, gluconic acid,²⁰ and phenolic compounds.^6,26 Additionally, selective processes using bimetallic catalysts and hydrogen donor solvents have been explored for lignin conversion to phenolics and other aromatic chemicals.^27–30

In this work, catalytic pyrolysis with a nonzeolitic, alumina-based catalyst was used to produce a phenolic rich biocrude from loblolly pine, which was then distilled to recover valuable lignin-derived chemicals. The partially deoxygenated biocrude (oxygen contents between 20–25 wt%) is thermally stable, distillable, and can be upgraded further into biofuels using conventional technologies such as hydrotreating. Unlike the biocrude from zeolitic catalytic pyrolysis, which consists mainly of aromatic hydrocarbons, the partially deoxygenated biocrude obtained by the reported process contains useful phenolic chemicals that are difficult to make from petroleum feedstocks. Therefore, in this work, distillation was evaluated as a potential separation method to recover guaiacols from loblolly pine biocrude as chemical intermediates; the findings from this investigation are reported herein.

Experimental

Production of biocrude by catalytic pyrolysis

The biocrudes used in this study were produced from loblolly pine (particle size of 2 mm top size, average moisture content of 15 wt%) in RTI's 1-tonne per day (1 TPD) catalytic biomass pyrolysis unit, which consists of a biomass screw feeder, riser reactor, regenerator, cyclone separators, and a condensation train (a quench, gas/liquid separator, condenser, and coalescing filters). The design and operation of the 1 TPD catalytic biomass pyrolysis unit has been reported elsewhere.³¹ The catalyst employed was a commercially available spray-dried, nonzeolite, alumina-based catalyst with a BET surface area of 114.6 m² g⁻¹ and a mean particle size of approximately 70 μm. The catalytic pyrolysis temperatures were 425 °C (biocrude-A) and 465 °C (biocrude-B). In each experiment, a biomass feed rate of 38 kg h⁻¹ was used, and operation was continuous for a minimum of 12 hours. In the catalytic pyrolysis process, biomass particles are continuously contacted with hot regenerated catalyst in the mixing zone of the riser reactor to promote partial deoxygenation of the primary pyrolysis vapors. After reaction has been established for a total residence time of ∼0.5 s in the mixing zone, the entrained char and catalyst are separated from the product vapors and gases in a cyclone. The separated solids (catalyst, char, and ash) are then transferred to the regenerator through a loop seal, where the char and coke on the catalyst are combusted with air. The pyrolysis vapors that exit the cyclone are subsequently condensed into an organic-rich fraction (biocrude) and water-rich fraction (aqueous phase). Finally, the permanent gases are directed into an electrically heated catalytic thermal oxidizer. The pyrolysis products were organic liquid phase, aqueous phase, char/catalyst coke, and gases. These products were analyzed to evaluate the process.

Distillation

Fig. 1 shows a fully automated, processor-controlled 200 mL bench-scale laboratory distillation unit (PETRODIST 300 CC) with one theoretical stage column used to fractionate the biocrude and recover a mid-cut fraction that is rich in phenolics. The unit consists of a glass distillation apparatus, flask heater/stirrer, vacuum pump, cooling system, and a control console. The unit is designed to distil up to a maximum 400 °C flask temperature.

Fig. 1 Simplified schematic diagram of PETRODIST 300.

For each experiment, 200 mL of biocrude was charged into a 500 mL round-bottom flask and the distillation was performed under vacuum (20 kPa) according to ASTM D-1160 method. The boiling point ranges for the distillation cuts were: cut 1 (IBP-110 °C), cut 2 (110–200 °C), cut 3 (200–300 °C), and cut 4 (300–400 °C). The four cuts were continuously collected using automatic fraction receivers. A laser level follower system measured the volume of each distillate cut during the distillation. In some cases, the distillation automatically stopped if vapor cracking was detected or if the maximum flask temperature value was reached. After each run, the weight distributions of the cut fractions were determined gravimetrically.

Analytical methods

The biocrude and the distillate fractions were analyzed by several analytical methods. Moisture content was measured with a Karl Fischer titrator (V20, Mettler Toledo). The test method followed the ASTM E203 standard test method for water using Hydranal-composite 5 K reagent. Organic elemental composition (CHONS) was determined with an elemental analyzer (FLASH2000, Thermo Scientific). Oxygen content was by difference. Density measurements were made with an Anton Paar DMA 35 portable density meter at 23 °C according to ASTM D4052. The values are accurate to 3 decimal places in g cm⁻³. The kinematic viscosity of the biocrude was measured with a 350 Cannon-Fenske upflow viscometer according to ASTM D-445-88. Thermogravimetric analysis was conducted on the biocrude using TA Instruments 2050 TGA. About 10 mg of sample was placed in an alumina crucible, and temperature was ramped to 600 °C at a rate of 10 °C min⁻¹ with 60 mL min⁻¹ of helium. The temperature was held at 600 °C, and the carrier gas was switched to 10% O₂ in argon. The experiment was stopped when no change in the weight was observed.

The chemical composition of the biocrude and the distillate fractions were determined using gas chromatography (GC) with mass spectrometric detection (Agilent 6890GC and 5975C MS). An HP-5MS column (30 m × 0.25 mm, 0.25 μm film thickness with 5% phenyl-methyl-polysiloxane as the stationary phase) was used for the separation of the components. The GC column followed a temperature program of 35 °C for 2 min, 5 °C min⁻¹ to 145 °C, 15 °C min⁻¹ to 290 °C held for 15 min. The helium carrier gas flow rate was controlled to maintain a constant linear velocity of 1 mL min⁻¹. The ion source and the interface of the mass spectrometer (MS) detector were held at 230 °C and 250 °C, respectively. The National Institute of Standards and Technology (NIST) mass spectral library was used for the initial identification of the most abundant compounds. The GC-MS was then calibrated with eugenol, isoeugenol, guaiacol, 4-methylguaiacol, 4-ethylguaiacol, and 4-propylguaiacol to specifically quantify their concentration in the distillate fractions. These pure compounds were purchased from Sigma-Aldrich. The analytical standards for calibration were prepared in accordance with ASTM method D4307. For each compound, a linear calibration curve was established for concentrations of 1000, 2000, and 3000 μg mL⁻¹. The calibration curves had R² value greater than 0.97.

Results

Physicochemical properties of biocrude

The exploratory work focused on separating valuable phenolics—specifically, guaiacol and its derivatives—from loblolly pine biocrude using fractional distillation. Table 1 shows the physicochemical properties of the biocrudes that were employed in the distillation experiment. Biocrude-A, produced at 425 °C, had carbon and oxygen contents of 63.1 wt% and 30.10 wt%, respectively. Biocrude-B produced at 465 °C, had carbon and oxygen contents of 70.2 wt% and 22.4 wt%, respectively. Biocrude-B also had a KF moisture content of 8.66 wt% and was relatively more viscous (169 cSt) and denser (1.218 g m⁻³). Conversely, biocrude-A was less viscous (132 cSt) and less dense (1.116 g m⁻³), probably due to its higher moisture content (20.20 wt%).

Table 1 Physical properties of biocrudes

Property	Analysis method	Biocrude-A (425 °C)	Biocrude-B (465 °C)
a By difference.
Moisture content (wt%)	Karl-Fischer titration	20.20	8.66
Density@23 °C, g m^−3	ASTM D4052	1.116	1.218
Kinematic viscosity, at 40 °C (cSt)	Cannon-Fenske routine viscometer (ASTM D445)	132	169
Elemental composition, wt% (dry basis)	ASTM D3176
C		63.15	70.15
H		6.28	7.21
N		0.48	0.32
O^a		30.10	22.37

---

Based on the chemical composition analysis by GC-MS in Table 2, lignin-derived products accounted for about 50% of the total ion chromatogram. The main phenolics were isoeugenol, eugenol, guaiacol, 4-methylguaiacol, 4-ethylguaiacol, and 4-propylguaiacol. Other multifunctional phenolics included 4-hydroxy-3-methoxybenzaldehyde, 2-(4-hydroxy-3-methoxy-phenyl) acetic acid, 4-hydroxy-3-methoxybenzoic acid methyl ester, 4-hydroxy-2-methoxycinnamaldehyde, and 1-(4-hydroxy-3-methoxyphenyl) ethanone. Anhydrosugars such as levoglucosan constituted about 11–16% of the total ion chromatogram. Other chemical species identified in the biocrudes include polyaromatic hydrocarbons, acetic acid, hydroxyacetone, furfural, furans, and other hydroxyl-carbonyls. Monoaromatic hydrocarbons, a characteristic of catalytic pyrolysis of biomass with zeolite, were not obvious in these biocrudes produced with the nonzeolite alumina-based catalyst at the report conditions.

Table 2 GC-MS analysis of biocrudes (peak area%)

List of major compounds in biocrude	GC-MS analysis (relative area%)
	Biocrude-A	Biocrude-B
Lignin-derived products
2-Methoxy-4-(1-propenyl)-(E)phenol (isoeugenol)	14.64	14.60
3-Allyl-6-methoxyphenol (eugenol)	5.24	4.83
2-Methoxy-4-methylphenol (4-methylguaiacol)	1.97	6.50
4-Ethyl-2-methoxyphenol (4-ethylguaiacol)	3.96	3.62
2-Methoxy-4-propylphenol (4-propylguaiacol)	8.38	1.84
2-Methoxyphenol (guaiacol)	1.97	2.34
4-Hydroxy-3-methoxybenzaldehyde	2.40	2.66
1,2-Benzendiol (catechol)	2.16	3.21
Alkylphenols	2.30	3.86
4-Hydroxy-3-methoxy-benzeneacetic acid (vanillin)	3.53	2.97
4-Hydroxy-3-methoxy-methyl ester-benzoic acid (methyl vanillate)	0.69	2.13
1-(3-Hydroxy-4-methoxyphenyl)-ethanone (apocynin)	1.01	2.95
4-Hydroxy-2-methoxycinnamaldehyde	0.99	2.19
Other chemicals	49.23	53.70
1-Methyl-7-(1-methylethyl)-phenanthrene	3.14	3.42
5-Methyl-2-furancarboxaldehyde	2.60	1.91
Acetic acid	4.07	1.16
1-Hydroxy-2-propanone	2.67	1.87
Furfural	1.70	1.61
2(5H)-Furanone	1.13	0.66
2-Hydroxy-3-methyl-2-cyclopenten-1-one	2.18	1.63
1,6-Anhydro-beta-D-glucopyranose (levoglucosan)	11.56	15.80

---

Distillation yields

The biocrudes (A & B) were fractionated into four different cuts as follows: cut 1 (IBP-110 °C), cut 2 (110–200 °C), cut 3 (200–300 °C), and cut 4 (300–400 °C). These cut points were selected to concentrate similar chemical families or functionalities in a single fraction. For example, the empirical boiling range for most of the phenolics is from 200 °C to 300 °C to coincide with cut 3. Nonetheless, other phenolics, like 1-(3-hydroxy-4-methoxyphenyl) ethanone and others that boil above 300 °C, will be concentrated in cut 4. Similarly, most of the carbonyl functionalities like furfural, cyclic ketones, and acetic acids will be concentrated in cut 2.

Table 3 shows the yields of each cut fraction for 12 distillation experiments; the first 6 experiments were with biocrude-A and the last 6 were conducted with biocrude-B. The weight distribution of each distillate fraction and the residue formed were determined gravimetrically. The volumetric yields were automatically measured with the aid of a laser level follower system. The distillation curves for liquids collected from all the experiments are also shown in Fig. 2; the dashed lines are for biocrude-A and the solid lines are for biocrude-B. Clearly, the distillation characteristics of the two biocrudes were different. The total yield of the distillates collected from biocrude-A varied between 60 wt% and 80 wt%, and the amount of residue formed ranged from 14 wt% to 29 wt%. In the case of biocrude-B, the total yield of the distillates was between 65 wt% and 75 wt%; the residue was between 16 wt% and 23 wt%. Also, about 4 wt% to 9 wt% of distillate escaped condensation and was collected in the cold trap. The mass closures were between 93% and 101%. It is worth pointing out that the variation in the overall distillation yields is partly due to differences in the distillation end points. In a number of the experiments, the distillation was terminated prematurely due to automatic detection of cracking during distillation; hence, high amounts of bottoms/residue were recorded. The final boiling point (FBP) set-point (400 °C) was only reached in a few of the experiments.

Table 3 Distillation yields (wt%), (experiments A1–A6 used biocrude-A and experiments B1–B6 used biocrude-B)

Experiment #	Cut 1	Cut 2	Cut 3	Cut 4	Cold trap	Residue	Mass balance
A1	12.9	5.9	21.4	25.4	7.5	22.3	95.4
A2	17.8	8.1	28.8	19.6	7.2	16.7	98.3
A3	12.8	6.5	21.8	27.3	8.5	22.3	99.1
A4	11.9	7.1	24.0	26.3	6.9	20.1	96.2
A5	12.5	7.8	22.4	27.3	7.7	21.7	99.4
A6	13.0	7.5	20.7	28.2	4.6	19.4	93.3
B1	7.7	4.5	18.9	35.1	6.3	28.4	101.0
B2	7.8	1.5	13.3	47.8	5.8	22.5	98.7
B3	6.0	4.5	18.2	44.7	6.8	20.2	100.5
B4	8.9	4.3	19.1	28.3	6.3	28.0	94.9
B5	4.1	3.2	17.3	55.1	4.1	14.9	98.6
B6	5.6	3.5	17.3	47.0	6.0	19.9	99.1

---

Fig. 2 Distillation curves.

The highest distillate fraction collected from the biocrudes was cut 4 in most cases; the yields were between 19 wt% and 55 wt%, depending on the run and the biocrude used. As shown by the distillation curves, a large fraction of the biocrude boiled above 300 °C. Specifically, about 35% and 50% of the volume of biocrude-A and biocrude-B, respectively, distilled below 300 °C. The cut 3 fraction had the second highest yield; the yields varied between 13 wt% and 29 wt%. The cut 2 fraction had the lowest yield, suggesting that chemical components like acetic acid, furfural, hydroxyacetone, and other cyclic ketones with boiling points between 110 °C and 200 °C were of lower concentration in the biocrudes. As expected, the cut 1 yields (components boiling below 110 °C) were more reflective of the initial moisture contents of the biocrudes. For instance, biocrude-A had a moisture content of 20.20 wt% and its cut 1 yield ranged from 12 wt% to 18 wt%. Similarly, biocrude-B had a moisture content of 8.66 and the cut 1 distillate yield averaged about 6.7 wt%.

Overall, the distillation yields reported herein for the biocrudes produced by catalytic biomass pyrolysis are higher than the usual yields observed from the distillation of raw bio-oils. This is simply because of the improved stability of the biocrudes used. Typically, distillation of fast pyrolysis oils results in excessive formation of solid residue (over 50 wt%) even at lower temperatures (<200 °C);³² this observation is attributed to the thermochemical instability of raw bio-oils. For instance, a study by Zhang et al.¹³ on mass production of chemicals from bio-oil reported a distillation yield of about 51.86 wt%. Zheng and Wei³³ also reported 61 wt% yield for distillation of bio-oil at reduced pressure (2.0 kPa). Nevertheless, it has been shown by other studies that the use of approaches such as molecular distillation at very low pressures (<0.1 kPa) improves the distillation of bio-oils^23,34,35 and prevents excessive formation of residue due to polymerization reactions.

Physicochemical properties of the distillation fractions

The distillate cuts from selected experiments (B3, B4, & B5) were analyzed for moisture, density, and organic CHNO as shown in Table 4. The cut 1 fraction had KF moisture content between 75 wt% and 89 wt%; as such, the density of cut 1 (0.998–1.015 g cm⁻³) was close to that of water. Consequently, the carbon and hydrogen contents on wet basis averaged 12.30 wt% and 10.21 wt%, respectively. In the case of cut 2, the KF moisture content varied between 23 wt% and 45 wt%, and the density was slightly higher than 1 g cm⁻³ with an average value of 1.072 g cm⁻³. Also, the average carbon and hydrogen contents were 45.6 wt% and 8.5 wt%. The cut 3 fractions had lower moisture content (7 wt% to 9 wt%) and the carbon and hydrogen contents averaged at 67.2 wt% and 8.1 wt%, respectively. As expected, the cut 4 fraction collected at temperatures higher than 300 °C had higher average carbon (72.8 wt%) and hydrogen (8.6 wt%) contents. Additionally, the density of cut 4 fraction was 1.093–1.210 g m⁻³ and similar to the original biocrude. It is worth noting that the oxygen content of cut 3 fractions is comparable to the oxygen content of phenolics such as isoeugenol (19%) and 4-methylguaiacol (23%).

Table 4 Physicochemical properties of distillate fractions

Sample	Moisture (wt%)	Density (g cm^−3)	Carbon (wt%)	Hydrogen (wt%)	Oxygen (wt%)
Cut 1 (IBP-110 °C)
Experiment B3	81.20	1.013	13.31	10.01	76.67
Experiment B4	88.70	1.015	5.64	10.32	84.04
Experiment B5	76.70	0.998	17.93	10.30	71.76
			12.30	10.21	77.49
Cut 2 (110–200 °C)
Experiment B3	27.65	1.079	52.16	8.17	39.67
Experiment B4	45.30	1.081	35.35	8.72	55.93
Experiment B5	23.05	1.055	49.24	8.57	42.18
		1.072	45.584	8.487	45.930
Cut 3 (200–300 °C)
Experiment B3	8.85	1.085	67.20	8.10	24.70
Experiment B4	7.05	1.140	65.18	8.23	26.59
Experiment B5	6.69	1.068	69.18	8.06	22.75
			67.19	8.13	24.68
Cut 4 (300–450 °C)
Experiment B3	0.90	1.093	73.15	8.70	18.15
Experiment B4	1.10	1.208	67.13	7.97	24.90
Experiment B5	1.13	1.210	78.13	9.26	12.60

---

Analysis of distillation residue

A sample of the residues formed during distillation of the biocrude oils was also analyzed. Organic elemental analysis showed that the residue had an average composition of 75.5% carbon, 5.1% hydrogen, and 19.0% oxygen. The CHNO analysis indicated that the polymerized material formed during distillation was not free of oxygenated compounds. In Fig. 3, a Fourier transform infrared (FTIR) spectrum of the residue suggests that compounds with carbonyl and phenolic functionalities were present. Specifically, the C=O band around 1705 cm⁻¹ indicates ketones and aldehydes as well as acids. Also, the O–H stretching vibration around 3350 cm⁻¹ and the broad band between 1300 cm⁻¹ and 1000 cm⁻¹ as well as the aromatic C–H stretch at 1600 cm⁻¹ are probably due to heavy phenolic compounds. Typically, phenolic C–O–H deformation stretches are seen around 1268 cm⁻¹ and aromatic ethers of C–O–C stretches between 1200 and 1240 cm⁻¹. Also, the absorption bands between 1030 and 1150 cm⁻¹ due to C–O stretching vibrations may be originating from alicyclic ring alcoholic compounds such as sugars. Furthermore, a thermogravimetric analysis (Fig. 4) showed that about 40 wt% of the residue degraded when it was heated at a rate of 10 °C min⁻¹ to 600 °C under helium. Most of the decomposition occurred between 300 °C and 500 °C. The remaining fraction at 600 °C degraded completely when the atmosphere was changed from helium to zero air. In general, the analyses suggest that the residue contain nonvolatiles like lignin oligomers and other higher molecular weight compounds.

Fig. 3 FTIR spectra of biocrude oil and residue formed during distillation.

Fig. 4 TGA of residue formed during biocrude distillation.

GC-MS quantification of guaiacols in the distillation fractions

The amount of guaiacols in the cut fractions 2, 3, and 4 obtained from seven selected distillation experiments (A1, A2, and B2–B6) were determined by GC-MS, as shown in Fig. 5. Calibration curves developed with pure eugenol, isoeugenol, guaiacol, 4-methylguaiacol, 4-ethylguaiacol, and 4-propylguaiacol were used to quantify the signals. Notably, other phenolics, such as 1-(4-hydroxy-3-methoxyphenyl) ethanone, 4-hydroxy-3-methoxybenzaldehyde, 4-hydroxy-3-methoxy-benzeneacetic acid, and 4-hydroxy-2-methoxycinnamaldehyde were not quantified. In cut 2, the total amount of eugenols and alkylguaiacols was between 6.3 wt% and 17.8 wt%. On average, the amount of guaiacols, eugenol, and isoeugenol were 10.6 wt%, 2.0 wt%, and 1.9 wt%, respectively. For cut 3 distillate fractions, the total amount of eugenols and guaiacols varied between 21 wt% and 38 wt%. Isoeugenol was the compound with the highest concentration (7 to 15 wt%), followed by eugenol (4.1 wt% to 7.7 wt%). The total amount of guaiacols was between 9.0 wt% and 15 wt%. The cut 4 distillates contained about 3–6.5 wt% of eugenols and guaiacols. The average amounts of eugenols and guaiacols were 2.38 wt% and 2.53 wt%, respectively.

Fig. 5 Concentration of eugenols and guaiacols in cut 2, cut 3, and cut 4 fractions from seven selected distillation experiments (A1, A2, and B2–B6).

Case study: evaluation of a two-stage distillation

A biocrude-B sample containing about 9.4 wt% of eugenols and guaiacols was distilled in two stages with the goal of examining further the efficiency of distillation as a separation method. Specifically, the biocrude contained about 0.6 wt% guaiacol, 1.5 wt% 4-methylguaiacol, 0.7 wt% 4-ethylguaiacol, 0.7 wt% 4-propylguaiacol, 1.3 wt% eugenol, and 4.6 wt% isoeugenol. In the first stage of distillation, the yields for cut 1, cut 2, cut 3, and cut 4 were 17.8 wt%, 8.1 wt%, 28.8 wt%, and 19.6 wt%, respectively. About 16.7 wt% residue was formed and 7.2 wt% light organics were collected in the cold trap. When the cut 3 fraction was distilled again, the yields for fractions that boiled between the IBP and 200 °C was 12.6 wt%, and the yield for the new cut 3 fraction collected between 200 °C and 280 °C was 53 wt%. The last fraction collected between 280 °C and 350 °C was 26.7 wt%. Fig. 6 shows the distillation curve of the biocrude in the first stage and the redistilled cut 3 in the second stage; Fig. 7 shows the GC-MS chromatograms. It can clearly be seen from the cut 3 distillation curve that more distilled between 200 °C and 300 °C, suggesting that the new cut 3 fraction would contain a high amount of phenolics. Compositional analysis showed that the first cut 3 fraction contained about 5.6 wt% of eugenol, 15 wt% of isoeugenol, 12.3 wt% of methyl- and ethyl-guaiacols, and 2.6 wt% of 4-propylguaiacol. As expected, the concentration of these phenolics increased in the second cut 3 fraction collected in the second stage of distillation: 9.0 wt% eugenol, 20.0 wt% isoeugenol, 19.7 wt% methyl- and ethyl-guaiacols, and 4.2 wt% 4-propylguaiacol.

Fig. 6 Distillation curves of the biocrude in the first stage and cut 3 in the second stage.

Fig. 7 GC-MS of biocrude: cut 3 in the first stage and cut 3 in the second stage.

Discussion

Fractionation of bio-oils enables the concentration of targeted chemicals for ease of recovery and purification. Over the years, several solvent fractionation procedures have been developed for the recovery of a phenolic fraction²¹ from bio-oil. In particular, the systematic column chromatographic fractionation of bio-oil with different solvents, as well as liquid- liquid extraction, have been extensively studied by different research groups.²¹ For example, a combination of a water-addition step and a base extraction followed by solvent separation can result in a stream of phenolics and neutrals.²¹ Recently, Fu et al. studied the extraction of phenols from lignin-derived bio-oil using a switchable hydrophilicity solvent.¹¹ Fele Žilnik et al.¹⁴ recovered a phenolic fraction from pyrolysis oil by aqueous extraction and simultaneous use of a hydrophobic-polar solvent and antisolvent. Patel et al.¹⁵ investigated extraction of cardanol and phenol from bio-oil using supercritical fluid extraction method. Additionally, Naik et al.¹⁶ employed supercritical CO₂ to fractionate furanoids, pyronoids, and benezenoids from bio-oil. Li et al.¹⁷ explored alkaline extraction, followed by dichloromethane extraction to recover phenol derivatives from bio-oil.

In the previous work reported above, most of the fractionation studies to isolate phenolic compounds from bio-oil have used liquid–liquid extraction separation techniques. Distillation, on the other hand, is not as common, even though it is a well-known separation technique. Distillation has long been the dominant separation process in industrial applications with respect to the overall number of units implemented and in terms of total capital investment. Additionally, the relative simplicity, efficiency, and maturity of distillation make it the obvious first choice in the chemical processing industry. Nevertheless, its application for bio-oil separation has been hindered by excessive formation of solid residue (over 50 wt%) even at lower temperatures (<200 °C).³² This is attributed to the thermochemical instability of bio-oils produced by conventional fast pyrolysis. The work by Murwanashyaka et al.¹⁸ and Amen-Chen et al.¹⁹ are among the few studies that have employed distillation as a separation technique to recover phenolics. Specifically, Murwanashyaka et al.¹⁸ utilized steam distillation to concentrate phenolic compounds, which were further purified by solvent extraction to yield syringol with a purity of 92.3%. Amen-Chen et al.¹⁹ used distillation in combination with a five cross-current stage liquid–liquid extraction to recover valuable phenolics such as cresols, guaiacols, catechols, and syringol from wood tar.

The results from this study suggest that distillation could be used exclusively to recover a phenolic fraction from bio-oil, given that it is relatively stable and has narrow chemical distribution. As shown in the GC-MS data, the amount of undesirable oxygenates like hydroxy-carbonyls, acids, and anhydrosugars which are predominant in non-catalytic bio-oils were rather lower in the biocrude used. Also, the biocrude contained largely aromatic and phenolic compounds. The difference in the chemical composition of the biocrude used and what is typically expected from non-catalytic pyrolysis oils can be attributed to the catalytic effect of the alumina-based catalyst. The alumina catalyst is a solid acid catalyst, and as such, it is able to promote cracking reactions to break down higher molecular weight compounds like oligomers and also facilitate dehydration and decarbonylation reactions of many of the hydroxy-carbonyls and anhydrosugars. Consequently, the biocrude from catalytic pyrolysis tend to be relatively thermally stable, less acidic, and less oxygenated; and as such can be distilled without the usual excessive formation of residue. As shown in the present work, the loblolly pine bio crude produced from RTI's 1 TPD catalytic biomass pyrolysis unit was up to 80% distillable for temperatures up to 400 °C. Moreover, the study showed that phenolics such as eugenols and guaiacols in biocrude were concentrated from 10 wt% to 35 wt% in the first stage of distillation and to 53 wt% in a second stage of distillation. The data also showed that the recovery of the eugenols and guaiacols was about 75% in the first stage and 80% in the second stage. The findings suggest that a number of stages would be required to obtain higher purities.

Clearly, catalytic pyrolysis plays a critical role in the production of chemical products of interest by producing a bio-oil or biocrude with desirable thermochemical characteristics that enable traditional separation techniques to be employed. For instance, the eugenols and guaiacols targeted in this study are not selectively produced during conventional biomass pyrolysis; as such, their concentration in raw bio-oils is low (up to 4 wt%)¹⁴ compared to the biocrude from catalytic pyrolysis, which has high concentration of eugenols (up to 10 wt%) and guaiacols (up to 6.5 wt%).³¹ The reason for the difference in concentration is that, in conventional pyrolysis, most of the lignin biopolymer breaks down mainly into polyphenols, also called pyrolytic lignin/insoluble lignin. However, in catalytic pyrolysis, the pyrolytic lignin is further cracked/depolymerized, which consequentially enhances the formation of monomeric phenolics. As a result, the selection of a particular type of separation technique will depend on the thermochemical characteristics and thermodynamic properties of the chemical product of interest.³⁶

Conclusions

This study investigated distillation as a separation method for recovering value-added phenolics, specifically guaiacols, from biocrude produced from catalytic pyrolysis of loblolly pine. The distillation process was tailored to fractionate the biocrudes to concentrate compounds with key functional groups, with the goal of recovering the phenolics of interest. Two sets of experiments were conducted. In the first experiment, single-stage distillation was performed on two different biocrude samples, and in the second experiment, a biocrude sample was subjected to a two-stage distillation process with the goal of increasing the concentration of guaiacols. The results of this investigation show that up to 80 wt% of distillates for temperatures up to 400 °C is achievable with the biocrudes. The cut 3 (200–300 °C) fraction, which is of interest in this work, had distillate yields between 13 wt% and 29 wt%, and the total concentration of eugenols and guaiacols varied between 21 wt% and 38 wt%. The results from the second experiment show that a biocrude containing about 10 wt% guaiacols can be distilled to concentrate guaiacols up to 35 wt% (representing 75% recovery) in the first step and up to about 53 wt% (representing 80% recovery) in the second distillation step. This work suggests that it is possible to isolate and concentrate the lignin-derived products with a distillation column that includes a number of plates/trays. Nevertheless, a comprehensive separation strategy consisting of various separation methods may be the ultimate approach to enhancing the recovery of value-added products from direct liquefaction of biomass processes.

Acknowledgements

We acknowledge RTI's Internal Research and Development funding in support of the distillation work. The U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Bioenergy Technologies Office, is also acknowledged for financial support under contract EE-0005358 (Catalytic Upgrading of Thermochemical Intermediates to Hydrocarbons). The authors would also like to acknowledge the analytical support of Kelly Amato and Alexandra Zapata.

Notes and references

1. L. Zhang, Biofuels & Biochemicals Then and Now: Innovation Trends from Feedstocks to End Products, Cleantech Group, 2014.
2. J. S. Golden and R. Handfield, Why Biobased? Opportunities in the Emerging Bioeconomy, http://www.biopreferred.gov/BPResources/files/WhyBiobased.pdf, accessed October 2015.
3. D. C. Elliott, Analysis and comparison of biomass pyrolysis/gasification condensates: Final report, Report PNL-5943; other: ON: DE87003597, 1986, pp. 1–101.
4. S. Czernik and A. V. Bridgwater, Energy Fuels, 2004, 18, 590–598.
5. J. Piskorz, D. S. Scott and D. Radlein, in Pyrolysis Oils from Biomass, American Chemical Society, 1988, vol. 376, ch. 16, pp. 167–178.
6. X.-f. Zhu and Q. Lu, in Advanced Biofuels and Bioproducts, ed. J. W. Lee, Springer New York, 2013, ch. 10, pp. 129–146,  DOI:10.1007/978-1-4614-3348-4_10.
7. R. A. Sheldon, Green Chem., 2014, 16, 950–963.
8. A. V. Bridgwater, Appl. Catal., A, 1994, 116, 5–47.
9. A. V. Bridgwater, Catal. Today, 1996, 29, 285–295.
10. C. J. Longley, J. Howard and D. P. C. Fung, in Advances in Thermochemical Biomass Conversion, ed. A. V. Bridgwater, Springer Netherlands, Dordrecht, 1993, pp. 1441–1451,  DOI:10.1007/978-94-011-1336-6_114.
11. D. Fu, S. Farag, J. Chaouki and P. G. Jessop, Bioresour. Technol., 2014, 154, 101–108.
12. Y. Elkasabi, C. A. Mullen and A. A. Boateng, ACS Sustainable Chem. Eng., 2014, 2, 2042–2052.
13. X.-S. Zhang, G.-X. Yang, H. Jiang, W.-J. Liu and H.-S. Ding, Sci. Rep., 2013, 3, 1120.
14. L. Fele Žilnik and A. Jazbinšek, Sep. Purif. Technol., 2012, 86, 157–170.
15. R. N. Patel, S. Bandyopadhyay and A. Ganesh, Energy, 2011, 36, 1535–1542.
16. S. Naik, V. V. Goud, P. K. Rout and A. K. Dalai, Bioresour. Technol., 2010, 101, 7605–7613.
17. J. Li, C. Wang and Z. Yang, J. Anal. Appl. Pyrolysis, 2010, 89, 218–224.
18. J. N. Murwanashyaka, H. Pakdel and C. Roy, Sep. Purif. Technol., 2001, 24, 155–165.
19. C. Amen-Chen, H. Pakdel and C. Roy, Biomass Bioenergy, 1997, 13, 25–37.
20. J.-S. Kim, Bioresour. Technol., 2015, 178, 90–98.
21. D. Mohan, C. U. Pittman and P. H. Steele, Energy Fuels, 2006, 20, 848–889.
22. J. A. Capunitan and S. C. Capareda, Fuel, 2013, 112, 60–73.
23. S. Wang, Y. Gu, Q. Liu, Y. Yao, Z. Guo, Z. Luo and K. Cen, Fuel Process. Technol., 2009, 90, 738–745.
24. T. Carlson, G. Tompsett, W. Conner and G. Huber, Top. Catal., 2009, 52, 241–252.
25. O. D. Mante, J. A. Rodriguez, S. D. Senanayake and S. P. Babu, Green Chem., 2015, 17, 2362–2368.
26. D. Santhanaraj, M. R. Rover, D. E. Resasco, R. C. Brown and S. Crossley, ChemSusChem, 2014, 7, 3132–3137.
27. A. Rahimi, A. Ulbrich, J. J. Coon and S. S. Stahl, Nature, 2014, 515, 249–252.
28. J. Zhang, J. Teo, X. Chen, H. Asakura, T. Tanaka, K. Teramura and N. Yan, ACS Catal., 2014, 4, 1574–1583.
29. J. Zhang, H. Asakura, J. van Rijn, J. Yang, P. Duchesne, B. Zhang, X. Chen, P. Zhang, M. Saeys and N. Yan, Green Chem., 2014, 16, 2432–2437.
30. J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552–3599.
31. D. C. Dayton, J. R. Carpenter, A. Kataria, J. E. Peters, D. Barbee, O. D. Mante and R. Gupta, Green Chem., 2015, 17, 4680–4689.
32. A. V. Bridgwater, Therm. Sci., 2004, 8, 49.
33. J.-L. Zheng and Q. Wei, Biomass Bioenergy, 2011, 35, 1804–1810.
34. X. Guo, S. Wang, Z. Guo, Q. Liu, Z. Luo and K. Cen, Appl. Energy, 2010, 87, 2892–2898.
35. Z. Guo, S. Wang, Y. Gu, G. Xu, X. Li and Z. Luo, Sep. Purif. Technol., 2010, 76, 52–57.
36. B. Singh, A. Kumari and S. Datta, in Commercializing Biobased Products: Opportunities, Challenges, Benefits, and Risks, The Royal Society of Chemistry, 2016, pp. 92–131,  10.1039/9781782622444-00092.

---