Heat Treatment of Spent Liquors to Recover Chemically Bound Xylose

liquors are subjected to a constant temperature heat treatment to hydrolyze the alkyl xylosides to allow full recovery of xylose and alcohol. Complete...
0 downloads 0 Views 1MB Size
Download PDF
Article Cite This: Ind. Eng. Chem. Res. XXXX, XXX, XXX-XXX

pubs.acs.org/IECR

Heat Treatment of Spent Liquors to Recover Chemically Bound Xylose and Alcohol Asif M. Sharazi,*,† Adriaan R. P. van Heiningen,† and Ivan Sumerskii‡ †

Department of Chemical and Biological Engineering, University of Maine, 5737 Jenness Hall, Orono, Maine 04469-5737, United States ‡ Division of Chemistry of Renewable Resources, Department of Chemistry, University of Natural Resources and Life Sciences, Vienna, Konrad-Lorenz-Str. 24, A-3430 Tulln, Austria S Supporting Information *

ABSTRACT: SO2−ethanol−water (SEW) and SO2−isopropanol− water (SPW) spent liquors are obtained by fractionating sugarcane straw. The SEW and SPW liquors contain significant amounts of chemically bound xylose as ethyl xylosides (EX) and isopropyl xylosides (PX) respectively. The liquors are subjected to a constant temperature heat treatment to hydrolyze the alkyl xylosides to allow full recovery of xylose and alcohol. Complete hydrolysis of EX and PX is achieved at 121 °C in 70 and 30 min, respectively. The first-order kinetics of EX and PX hydrolysis are determined at temperatures from 100 to 121 °C. At full hydrolysis of the alkyl xylosides, the quantity of alcohol produced is greater than stoichiometric. Other sources of covalently bound alcohols in the spent liquors are identified to explain the excess alcohol produced. also confirmed by Lancefield et al.12 in butanol pretreatment (mixed with 5% H2O and 0.2 molar HCl at reflux) of different lignocellulosic feedstocks. These alkyl xylosides decrease the monosaccharides yield in the spent liquor stream and chemically bind a significant amount of solvent. Alkyl pyranosides formation can be controlled by employing higher alcohols as proposed by Drouet et al.13 The longer chain alcohols are less likely to form alkyl pyranosides.13,14 However, in organosolv processes solvent recovery becomes more expensive when using long-chain alcohols because of their decreased volatility. In contrast, the reduced safety requirements and improved lignin solubility15 in isopropanol (IPA) as compared to those in ethanol (EtOH) make IPA also of interest for application in SO2−alcohol−water (SAW) fractionation besides EtOH. It has been shown for SAW fractionation of SCS that the delignification potential with IPA is better than that with EtOH, while the xylan removal rate is nearly the same for both solvent systems.15 The maximum concentration of monomeric xylose in SEW spent liquor is obtained by cooking SCS at 155 °C for 60 min. The maximum solvent xylosides concentration also occurs under the same reaction conditions, suggesting that EX and xylose are in equilibrium during fractionation.10 In the present study, the SCS liquors obtained after solvent distillation of the SO2−ethanol−water (SEW) and

1. INTRODUCTION Concern about the environmental impact of fossil fuels burning continues to stimulate research to produce fuels and chemicals from renewable and carbon-neutral lignocellulosic feedstocks. Sugarcane straw (SCS) is a preferred raw material because it is abundantly available and has a low cost.1−4 However, economical production of biochemicals/fuels from lignocellulosic biomass by biochemical conversion requires efficient pretreatment/fractionation of the lignocellulosic structure to make the sugars readily available for fermentation. For fractionation into the three main components, cellulose, lignin and hemicellulose, the process must produce a cellulosic residue easily hydrolyzed by enzymes, and a spent treatment liquor containing the hemicelluloses mostly as monomeric sugars, and lignin which can easily be separated, for example by precipitation.5 The SO2−ethanol−water (SEW) process, which has been intensively studied at Aalto University in Finland,4−9 potentially meets these requirements. Although hemicelluloses are extensively dissolved in SEW fractionation, only 45−50% of the dissolved hemicelluloses are reported to be present as monosaccharides, with the remaining fraction identified as oligomers.5,9 However, recently it has been quantified that a significant fraction of the oligomers are actually alkyl pyranosides.10 The formation of alkyl xylosides has recently also been reported in a number of other organosolv pretreatments studies. Zhang et al.11 propose that ethyl xylosides (EX) formation is most likely the cause for poor xylan mass balances in organosolv pretreatment of plant biomass. The presence of alkyl xylosides in the solubilized fraction of hemicelluloses are © XXXX American Chemical Society

Received: Revised: Accepted: Published: A

August 31, 2017 October 31, 2017 November 2, 2017 November 2, 2017 DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research SO2−isopropanol−water (SPW) spent liquors are subjected to a simple heat treatment to hydrolyze the alkyl xylosides for full alcohol and xylose recovery. The kinetics of alkyl xylosides hydrolysis of the EtOH and IPA liquors are determined, and the excess solvent generated is interpreted in terms of additional hydrolysis reactions involving lignin and non-xylose sugars.

2. EXPERIMENTAL SECTION 2.1. Feedstock and Chemicals. Sugarcane straw (SCS) was obtained from the Thomaston, GA biorefinery of API (American Process Inc.), which was delivered from its Brazilian partner Granbio. SO2 was purchased from Tri-Gas Matheson (Bangor, ME). EtOH (95 v/v %) and IPA (purity > 99.9%) were purchased from Fisher Scientific (Suwanee GA, USA). A detailed description of the SCS screening procedure and final composition analysis can be found in our earlier study.15 2.2. Fractionation. SCS was subjected to fractionation at 155 °C for 1 h including 10 min heat-up. The cooking liquor for fractionation was prepared by injecting gaseous SO2 into the aqueous alcohol mixture (50% w/w). The final composition of the cooking liquor was SO2/alcohol/water = 12/44/44 (wt %). Stainless-steel cooking digesters (220 mL), filled with 22.6 g of oven-dried (o.d.) SCS and cooking liquor (liquid-to-feedstock ratio of 4 L/kg) were immersed in a temperature-controlled multi-digester oil bath (MDOB) (manufactured in-house) preheated at 155 °C. The digesters were rotated at a rate of 2 rpm. After 1 h, the digesters were removed from the MDOB and put into an ice bath to quench and stop further reaction. The spent liquor was separated from the solid residue (pulp) by manually squeezing the fractionated SCS contained in a nylon bag (75 mesh). This liquor is called SO2−alcohol−water (SAW) liquor. The obtained solid residue was washed twice with 45 mL of 50% (w/w) aqueous alcohol at 60 °C and three times with 45 mL of deionized (DI) water at room temperature. The alcohol and water washings were combined with fresh SAW spent liquor as would be done in industry. The final combined liquor is designated as mixed SAW (MSAW) liquor (see Figure 1). The SEW and SPW pulps (or solid residues) produced using the above-mentioned treatment were analyzed for yield and kappa number. The pulp kappa numbers were determined according to SCAN-C 1:00 method. 2.3. Lignin and Alcohol Removal. The addition of alcohol and water washings to the spent liquor resulted in substantial lignin precipitation. This is called precipitate I in this paper (Figure 1). The precipitated lignin was separated from the MSAW liquor, vacuum-dried (25 °C), and then gravimetrically quantified (Table 1). The amount of lignin precipitate I was corrected for covalently bound sugars, determined by subjecting the solids to double-stage hydrolysis (72% H2SO4, 30 °C, 1 h; 4% H2SO4, 121 °C, 1 h), and for ash content (at 525 °C for 24 ± 6 h). After precipitate I removal, the MSAW liquors were subjected to alcohol and SO2 evaporation in a Heidolph rotary evaporator equipped with a Maxima vacuum dryer. The water bath temperature was set at 80 °C for rotary evaporation. It took about 40−60 min to achieve ∼65−75% mass reduction of the MSAW liquors. Rotary evaporation induced more lignin precipitation shown as precipitate II in Figure 1. The resultant liquor is called EVAP in accordance with nomenclature in earlier studies.5,16 Precipitate II was separated from the EVAP liquors (EVAPE for ethanol and EVAPP for isopropanol) by centrifugation (4000 rpm, 15 min) and subjected to heat treatment (see section 2.5).

Figure 1. Schematic flow diagram showing sugarcane straw (SCS) fractionation and processing of resultant streams. L/F is liquor to feedstock ratio (L/kg) while SAW and MSAW denote SO2−alcohol− water and mixed SO2−alcohol−water liquor streams, respectively. EVAP represents liquor from rotary evaporation. SEW, SO2−ethanol− water; SPW, SO2−isopropanol−water; MSEW, mixed SO2−ethanol− water; MSPW, mixed SO2−isopropanol−water.

Table 1. Lignin Distributiona fractionation process b

lignin in SCS lignin in pulpc lignin in MSAW (SCS−pulp) precipitate I dissolved Id lignin in EVAP precipitate II dissolved IId

SEW

SPW

19.8 1.2 18.6 7.5 11.1

19.8 1.1 18.7 6.8 11.9

7.9 3.2

5.5 6.4

a

All values are based on g/100 o.d. g SCS. bAsh-free Klason lignin. Calculated based on ref 15. dCalculated as numerical difference “lignin in MSAW or EVAP − precipitate I or precipitate II”. c

2.4. Alkoxy Groups in Lignin. To confirm lignin alkoxylation, the lignin from fresh spent liquor (SEW or SPW) was precipitated and analyzed for alkoxy groups (methoxy, ethoxy, and isopropoxy). The precipitated lignin samples were prepared by removing alcohol and SO2 from spent liquor through rotary evaporation and then adding DI water. The precipitated lignin was separated from water by centrifugation (4000 rpm, 15 min). The degree of alkoxylation was quantified by analyzing the precipitated lignin samples using a novel headspace-isotope dilution (HS-ID) gas chromatography mass spectrometry (GC-MS) method based on standard hydroiodic acid cleavage of alkoxy groups.17 2.5. Heat Treatment and Analytical Calculations. EVAP liquors were subjected to heat treatment for alkyl xylosides hydrolysis at 100, 110, and 121 °C in glass vials (2.0 mL) immersed in preheated polyethylene glycol (PEG). The heat treatment duration ranged from 10 to 130 min including about 2−3 min for heat up of the vial content. After heat treatment, the samples were filtered (0.45 μm) and analyzed for monomeric (free) xylose by high performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD). The instrument was equipped with a Dionex B

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Table 2. Effect of Washings and Rotary Evaporation on Spent Liquors Contenta SEW fractionation xylose xylosides alcohol lignin pH

SPW fractionation

SEW

MSEW

EVAPE

SPW

MSPW

EVAPP

9.9 (29.1) 5.8 (18.5) 129 (429) 15.8 (52.5) 0.97

12.8 (10.0) 7.0 (5.7) 328 (272) 18.6 (15.4) 1.23

12.2 (43.6) 6.8 (24.6) 0.03 (0.1) 3.2 (16.0) 1.15

10.5 (32) 2.7 (9.3) 113 (385) 15.3 (52.1) 0.96

14.5 (11.5) 4.7 (3.9) 274 (235) 18.7 (16.4) 1.32

14.3 (38.1) 4.7 (12.5) 0.02 (0.1) 6.4 (23.6) 1.08

a

Mass concentration factors (mi/mf) 4.1 and 3.0 were used for SEW and SPW, respectively, in the calculations of EVAP liquors. All quantities except pH are in g/100 o.d. g SCS. The numbers in brackets show the concentrations in g/L.

CarboPac PA1 (4 × 250 mm) column (30 °C), CarboPac PA1 (4 × 50 mm) guard column, IonPac NG1 (4 × 35 mm) guard column, GP50 Pump, ED40 Detector (gold electrode), and an AS50 Autosampler. The eluent flow rate was 1.0 mL/min (degassed H2O at 0.7 mL/min + 300 mmol NaOH from the post column at 0.3 mL/min). The HPAEC injection samples were prepared by adding fucose (1 g/L, 1 mL) to the EVAP liquors (0.1 mL) and were then diluted to 10 mL with NaOH (0.1 molar). The amount of alcohol (EtOH or IPA) in all liquor samples was determined by high performance liquid chromatography (HPLC, Shimadzu) equipped with RI detector and BIO-RAD Aminex HPX-87K column (45 °C). The mobile phase was aqueous H2SO4 (5 mmol, 0.6 mL/min). The amount of ethyl xylosides (EX) was determined by treating the 0.2 mL liquor with H2O2 (6.4 μL, 35 wt %) and diluting it with water (dilution factor = 4−5) as described earlier to convert SO2 into sulfate so as to avoid peak interference between EX (retention time; 11.3 min for beta anomer and 12.7 for alpha anomer) and SO2 (retention time 12.7 min).10 A much higher amount of hydrogen peroxide (32 μL, 35 wt %) was used for EX determination in SEW and MSEW liquors because they contained much higher SO2 concentrations.10 Isopropyl xylosides (PX) in EVAPP and corresponding heat-treated liquors were estimated using calibration curves developed in similar fashion as for EX10 (see also the Supporting Information). However, the IPA samples were analyzed without H2O2 addition as SO2 does not elute at the same location as the PX anomers during HPLC analysis (retention time; beta anomer 11.7 min and alpha anomer 13.7 min).

much lower (3−7 wt %). This may be attributed to a higher carry over of sugars with precipitate I in the separation procedure. The dissolved lignin concentration in the EVAP liquors is 66−84% lower than that in the MSAW liquors. This decrease in lignin content due to removal of precipitated lignin (I and II) is comparable to the 75% (w/w) reduction in lignin content reported by Sklavounos et al.5 who used the same treatment for spruce MSEW spent liquor. Our results also agree with those of You et al.16 who found only 2−3% (based on feedstock) dissolved lignin in EVAPE liquor after rotary evaporation of MSEW liquors obtained from SCS fractionation (155 °C, 58 min, SEW liquor composition 12:44:44 w/w %). It indicates that lignin removal from spent liquors by precipitation depends on alcohol dilution/depletion level irrespective of lignocellulosic biomass. 3.2. Composition of Spent Liquor. Table 2 shows the comparison of fresh (SAW), mixed (MSAW) and rotovap (EVAP) spent liquors obtained after SEW and SPW fractionation of SCS (see Table S2 for molar concentrations). A 3−4% increase in monomeric xylose content is seen after adding alcohol and DI water washings to fresh SEW and SPW spent liquors to produce MSEW and MSPW liquors, respectively (Figure 1). Similarly, washing removes an additional 2.5−3.5% lignin from fractionated SCS. Some alkyl xylosides are also removed from the solid residue (or pulp) in the washing process. These findings agree with those of Iakovlev et al.18 and Sklavounos et al.5 that alcohol/water washing removes a significant amount of lignin and carbohydrates from fractionated biomass. Very small changes in monomeric xylose and in xylosides content (SCS basis) occur when MSAW liquors are subjected to rotary evaporation at 80 °C indicating that xylosides and xylo-oligomers hydrolysis are not significant at this temperature. About 4.3% xylooligomers, 4% EX, and 9.5% monomeric xylose (all based on SCS) are reported in SEW fresh liquor produced under similar fractionation conditions.10 The concentrations of EtOH and IPA in EVAPE and EVAPP liquors are very low at 0.1 g/L (∼0.02−0.03 g/100 o.d. g SCS). Thus, more than 99% of the alcohol present in the MSAW liquors is removed in the rotovap step. However, alcohol chemically bound to carbohydrates or lignin remains in the EVAP liquor. The fresh spent liquors have a high acidity (pH ∼1.0) due to the presence of lignosulfonic acids. As expected, the pH increases to 1.2−1.3 due to dilution with washing liquid which is free of lignosulfonic acid and SO2. The pH again slightly goes down when SO2 and alcohol are removed from the liquors by rotary evaporation because the increase in acidity due to concentration is only partly balanced by the removal of SO2. 3.3. Ethyl Xylosides Hydrolysis. EVAPE spent liquor contains 6.8 g of EX per 100 o.d. g SCS (Table 2). Figure 2

3. RESULTS AND DISCUSSION 3.1. Lignin Distribution. SAW fractionation removes most of the lignin from SCS as shown in Table 1. The resultant SEW and SPW pulps have yields of 38.9 ± 0.2 and 37.3 ± 0.5 wt % respectively and a kappa number of 23 indicating that about 94% lignin is removed from SCS in SEW and SPW fractionation. This lignin becomes part of the spent liquor stream containing dissolved sugars and unreacted alcohol and SO2. The addition of water washings in the spent liquor dilute its alcohol concentration (Figure 1). Initially, spent liquors have 44 wt % alcohol, and the washings addition bring this concentration down to 25−28 wt % in mixed liquor stream (MSAW, Figure 1). As a result, lignin starts to precipitate because of its hydrophobic nature. This precipitated lignin fraction is shown as precipitate I in Figure 1 and Table 1. Precipitate I contains 17−21% sugars and 4−7% ash by weight. Thus, precipitate I represents 34−38 wt % of total lignin in the original SCS. The alcohol depletion in MSAW liquor by rotary evaporation forms precipitate II. The precipitate II amounts are comparable to precipitate I except that their content of sugars is C

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

An activation energy for EX hydrolysis of 66 kJ/mol is determined from the rate constants based on the Arrhenius equation. The rate constant data show that hydrolysis is two times faster at 110 °C than at 100 °C and another 1.5 times faster at 121 °C. The PX hydrolysis results will be discussed later in the paper. Herein, 1 mol of EX hydrolysis produces 1 mol of xylose and EtOH (see reaction 1): C7H14O5 + H 2O ⇌ C2H5OH + C5H10O5

(1)

The EVAPE liquor initially has 6.8 g of EX (per 100 o.d. g SCS) corresponding to 38 mmol (per 100 o.d. g SCS) of EX. It also contains 12.2 g of xylose (or 81.2 mmol) which is subtracted from the total monomeric xylose content measured after EX hydrolysis. The development of the molar concentration of the newly formed xylose at different hydrolysis conditions is plotted in Figure 4. The rate of xylose release increases dramatically as Figure 2. Ethyl xylosides (EX) hydrolysis: ●, 100 °C; 121 °C.

▲,

110 °C; ■,

shows the hydrolysis of EX (sum of α- and β-ethyl xyloside) during heat treatment at different temperatures. At 100 °C, a gradual disappearance in EX is seen over a period of about 2 h. The rate of disappearance increases with temperature and negligible EX remains after 70 min at 121 °C. Plotting the natural logarithm of the EX molar concentration vs reaction time (Figure 3) produces straight lines. The data

Figure 4. Xylose release from ethyl xylosides (EX) hydrolysis: ●, 100 °C; ▲, 110 °C; ■, 121 °C.

Figure 3. Kinetics of ethyl xylosides (EX) hydrolysis: ●, 100 °C; 110 °C; ■, 121 °C; ◇, initial EX concentration.

the temperature is increased from 100 to 121 °C. Figure 4 suggests that complete EX hydrolysis requires more than 2 h at 100 °C, while the same can be achieved in 70 min at 121 °C. A similar heat treatment method is patented by Retsina and Pylkkanen19 to recover lignin from SEW spent liquors. In the patented process, it is claimed that heat treatment of SEW spent liquors produces water insoluble lignin, while all “oligomeric sugars” are converted to monomers. However, based on our recent study,10 it is now known that the described “hydrolysis of oligomeric sugars” in the patent involves mostly the hydrolysis of EX instead of xylo-oligomers. In the current heat treatment of EVAPE liquors, no additional acid catalyst, such as H2SO4, is needed for EX hydrolysis because the SAW process produces its own strong soluble acid catalyst, lignosulfonic acid. Although the addition of H2SO4 converts all EX and xylo-oligomers into monomeric xylose,10 it also promotes sugar dehydration reactions leading to reduced sugar yield.14 Grisel et al.14 encounter high sugar losses in organosolv (94 wt % EtOH or 95 wt % MeOH) treatment of wheat straw with mineral acid (H2SO4) as catalyst. The reported losses of more than 25 wt % are attributed to humins formation. In contrast, during hydrolysis in the present study the sugar dehydration is very small as evidenced by insignificant amounts of furfural, and by a good molar mass balance of EX and (newly

▲,

points after 70 min at 121 °C are not included in the kinetic analysis as it produces inaccurate results because of very low EX concentrations. Thus, hydrolysis follows first-order kinetics in EX. The first-order rate constants were determined from the slopes of the straight lines in Figure 3 and are listed in Table 3. Table 3. Kinetic Parameters of Hydrolysis khyd × 102 (1/min) EX PX

100 °C

110 °C

121 °C

activation energy (kJ/mol)

1.68 2.13

3.49 5.97

5.23 20.06

66 130.7 D

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

walnut shell, 84% Douglas fir) indicating that lignin ethoxylation occurs at relatively low temperatures. Hydrolysis of these α-ethoxylated linkages are reported to occur under relatively mild acidic conditions (0.1 molar HCl, 2:1 dioxane/ water, 100 °C, 4 h) to produce native-like lignin.12 Thus, heat treatment of SAW liquor leads to not only alkyl xylosides hydrolysis but also lignin de-ethoxylation which is a very important finding for commercial implementation of organosolv processes since full EtOH recovery is needed for a viable biochemical conversion technology. The advantage of the SAW technology is that there is no need to add any additional mineral/organic acids since the naturally produced lignosulfonic acid provides the required acidity (pH of about 1.0) for the hydrolysis reactions. However, since xylo-oligomers appear to survive the present heat treatment conditions, further research is needed to establish at which temperature above 121 °C the xylo-oligomers will hydrolyze fully to monomers. Alcohol released from dealkoxylation of lignin during heat treatment of EVAP liquors is estimated from the alkoxy groups. The alkoxy groups chemically bound to lignin are shown in Figure 6 in the form alkoxy/C9 (C9 represents phenyl-

formed) monomeric xylose over the entire hydrolysis period at all 3 temperatures studied (see Table S3). This also suggests that xylo-oligomers survived the present heat treatment conditions since the maximum amount of xylose formed at 121 °C is essentially the same as that at the two lower temperatures. Heat treatment of EVAPE liquor also releases EtOH as is shown in Figure 5. This is expected based on reaction 1. After

Figure 5. Ethanol (EtOH) release during heat treatment of EVAPE liquor: ●, 100 °C; ▲, 110 °C; ■, 121 °C; dashed green line, theoretical maximum EtOH.

rotary evaporation of MSEW the resultant liquor (EVAPE) still has ∼0.1 g/L EtOH (see Table 2). This corresponds to 0.7 mmol of EtOH per 100 o.d g SCS. This amount is subtracted from the EtOH concentration in the heat-treated liquor, and the corrected EtOH concentrations are plotted in Figure 5. On the basis of the stoichiometry of reaction 1 and the results in Figure 4, it can be calculated that the maximum EtOH concentration after complete EX hydrolysis would be 38 mmol. However, the measured maximum EtOH concentration in the heat-treated liquors reaches a value of about 78 mmol which is about twice as high (Figure 5). Other possible sources of EtOH are ethylated lignin and/or ethyl pyranosides of non-xylose sugars. We did not measure ethyl pyranosides of non-xylose sugars (arabinose, galactose, glucose, mannose) in MSEW liquors, but their presence is likely, albeit at much lower concentrations in comparison to that of EX. Thus, the most likely source for the additional EtOH is ethylated lignin. Bauer et al.20 show that acidic organosolv treatment of miscanthus with EtOH (95% EtOH/5% 0.2 molar HCl (v/v), 8 h reflux) results in modification of the lignin structure. Under these conditions, EtOH behaves as a nucleophile and ethoxylates the α-hydroxyl group of the lignin propane side units containing a ß-O-4 linkage. The α-ethoxylation is the only major lignin modification seen in this ethanosolv pretreatment process. Similar findings have been reported by Lancefield et al.12 for organosolv treatment (10 mL/g, 95% alcohol/5% 4 molar aqueous HCl (v/v), 6 h reflux) of different lignocellulosic biomass (beech, walnut shell, and Douglas fir). Interestingly, nearly 96% of the propane side units containing a ß-O-4 linkage are α-ethylated for all these species at 110 °C, while a lower degree of ethoxylation is observed under reflux conditions at the azeotropic boiling point of 78.15 °C (65% beech, 71%

Figure 6. Moles of methoxy, ethoxy, and isopropoxy functional groups per mole of phenylpropanoid lignin unit (C9). Molar weight of C9 is 204 g/mol. SEW, SO2−ethanol−water; SPW, SO2−isopropanol− water

propanoid unit) molar ratio. The lignin precipitate is corrected for sugars retained in the samples (∼5 wt %), and the molar weight (MW) of lignin phenylpropanoid unit is taken as 204 g/ mol. This MW is used for SCS lignin based on its weight distribution of guaiacyl (G), syringyl (S), and p-hydroxyphenyl (H) units as described earlier15 for calculation of the SCS lignin Hildebrand solubility parameter. As expected, both SEW and SPW derived lignin samples contain a significant amount of methoxy groups since SCS lignin is composed mostly of G and S units. However, the SEW and SPW lignins also contain significant amounts of ethoxy and isopropoxy groups, respectively. On the basis of Figure 6, it can be calculated that nearly 0.18 mol of ethoxy groups are present per mol of SEW lignin. Since there are 18.6% lignin in SEW liquor (Table 1), it follows that there are 16.4 mmol of ethoxy groups (0.18 × 1000 × 18.6/ 204) chemically bound to dissolved SEW lignin, so complete hydrolysis of EX and SEW lignin produces 54.4 mmol EtOH/ E

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Table 4. Non-Xylose Sugars in Fresh and Heat-Treated Spent Liquorsa EVAPE fresh arabinose galactose glucose mannose total a

2.4 0.5 3.5 0.4 6.8

(16.0) (2.8) (19.4) (2.0) (40.2)

EVAPP

heat-treated 3.2 0.7 4.7 0.7 9.3

(21.3) (4.1) (26.3) (3.7) (55.4)

pyr 0.8 0.2 1.2 0.3 2.5

fresh

(5.3) (1.3) (6.9) (1.7) (15.2)

2.5 0.6 4.4 0.5 8.0

(16.6) (3.1) (24.4) (2.6) (46.7)

heat-treated 2.8 0.6 4.7 0.7 8.8

(18.8) (3.6) (25.9) (3.8) (52.1)

pyr 0.3 0.0 0.3 0.2 0.8

(2.2) (0.5) (1.5) (1.2) (5.4)

Quantities are reported as g/100 o.d. g SCS. The numbers in parentheses show concentrations as mmol/100 o.d. g SCS. Pyr, pyranosides.

100 o.d. g SCS (38 from EX + 16.4 from lignin). This is still smaller than 70 mmol of EtOH/100 o.d. g SCS at complete EX hydrolysis at 121 °C for 70 min (see Figure 5). Thus, it was investigated whether the difference of 15.6 (70−54.4) mmol EtOH/100 o.d. g SCS could be produced from hydrolysis of ethyl pyranosides of non-xylose sugars in the spent liquor. Table 4 shows the concentration of non-xylose monosaccharides in EVAPE and EVAPP spent liquors before and after heat treatment. Non-xylose monomeric sugars in fresh EVAPE are measured and compared with their corresponding quantities in the heat-treated liquor (121 °C, 70 min). The same treatment and analysis is done for fresh EVAPP and heattreated liquor (121 °C, 30 min) as will be discussed later in this paper. The increase in non-xylose sugar content after heat treatment indicates the contribution from hydrolysis of corresponding pyranosides and oligomers in fresh spent liquors. When neglecting the contribution from non-xylose oligomers (only 0.98 wt % glucomannan in SCS), it can be calculated that the maximum increase in monosugars caused by hydrolysis of non-xylose ethyl pyranosides could be 15.2 mmol of EtOH/100 o.d. g SCS. Thus, the theoretical maximum amount of ethanol is 15.2 + 38 + 16.4 = 69.6 mmol. This value is added as a horizontal line in Figure 5. The data points above the horizontal line are thought to be due to experimental errors. This shows that now there is a reasonable ethanol mass balance coming from three sources: EX, ethoxylated lignin, and nonxylose ethyl pyranosides. However, for unequivocal verification that the additional ethanol (15.2 mmol) is coming from the non-xylose pyranosides, quantification is needed of the pyranosides of all non-xylose sugars. At all three temperatures, the molar sum of EX and EtOH increases to about 70 mmol/ 100 o.d. g SCS over the duration of the hydrolysis, but the rate to reach this level is of course much higher at 121 °C than at 100 °C (see Table S3). 3.4. Isopropyl Xylosides Hydrolysis. EVAPP contains 4.7 g of PX (per 100 o.d. g SCS) as shown earlier in Table 2. This represents 3.7 g of the equivalent xylose. Comparing this with 2.7 g PX/100 o.d. g SCS in fresh SPW spent liquor (Table 2) indicates that addition of the wash liquors increases the amount of PX confirming that PX is washed out of the cooked SCS. Table 2 also shows that rotovap treatment of MSPW liquor at 80 °C to produce EVAPP from MSPW is unable to hydrolyze PX and xylo-oligomers at this temperature as was also found for EX. Subsequently the EVAPP liquors, which are essentially free of SO2 and IPA, are subjected to heat treatment at 100, 110, and 121 °C. The effect of heat treatment temperature on PX hydrolysis is shown in Figure 7. The PX hydrolysis rate increases with increasing temperature and complete hydrolysis is achieved in about 30 min at 121 °C. It takes about 100 min at 110 °C for complete hydrolysis but EVAPP still has about 0.3 g or 6.4% PX remaining after 2 h heat treatment at 100 °C.

Figure 7. Isopropyl xylosides (PX) hydrolysis: ●, 100 °C; ▲, 110 °C; ■, 121 °C.

The initial PX content of EVAPP of 4.7 g/100 o.d. g SCS is equivalent to 25 mmol/100 o.d. g SCS. As for EX, a natural logarithm plot of the molar concentration of PX in the heattreated liquors versus reaction time at different temperatures is shown in Figure 8. The straight lines again indicate a first-order hydrolysis reaction for PX as is found for EX. The first-order reaction rate constants calculated from the slopes of straight lines (Figure 8) are listed in Table 3. It shows that the PX hydrolysis rate constants are higher than those of EX at 100,

Figure 8. Kinetics of isopropyl xylosides (PX) hydrolysis: ●, 100 °C; ▲, 110 °C; ■, 121 °C;◇, PX initial concentration. F

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research 110, and 121 °C and that the PX hydrolysis is 10 times faster at 121 °C than at 100 °C. This is reflected by the high activation energy of PX hydrolysis (130.7 kJ/mol) which is about two times higher than that of EX. The higher rate constants for PX hydrolysis can be related to reduced steric strain imposed on the xylose half-chair transition state during hydrolysis as was found by Timel21 who investigated the effect of differently sized substituents on glycosides hydrolysis. The reason for difference in activation energy for PX and EX hydrolysis is unclear. However, it is evident that heat treatment of EVAPP at high temperatures releases IPA and xylose much quicker than EtOH and xylose during heat treatment of EVAPE. The maximum amount of xylose released during heat treatment of EVAPP at different temperatures is 22 mmol as shown in Figure 9. This is close to the initial amount of 25

Figure 10. Isopropanol (IPA) release during heat treatment of EVAPP liquor: ●, 100 °C; ▲, 110 °C; ■, 121 °C; dashed green line, theoretical maximum IPA.

per mol of lignin monomer. This isopropoxy/C9 ratio is similar to the ethoxy/C9 ratio suggesting that lignin alkylation is determined by the reactivity of lignin and not the solvent. This data also implies that complete deisopropoxylation of dissolved lignin from SPW fractionation will produce 20 mmol of IPA (0.22 × 1000 × 18.7/204). In addition, isopropanol could be produced by hydrolysis of non-xylose isopropyl pyranosides. This is quantified in Table 4 as 5.4 mmol of IPA. Thus, complete hydrolysis of PX, isopropoxylated lignin, and nonxylose isopropyl pyranosides produces 50.4 mmol of IPA. This theoretical amount is shown in Figure 10 as a horizontal line. Clearly this calculated maximum amount is more than the 32 mmol of IPA found in EVAPP heat-treated liquor (121 °C, 30 min). This suggests a comparatively slower dealkoxylation of isoproxylated lignin than ethoxylated lignin in the presence of lignosulfonic acids. This argument is supported by Lancefield et al.12 who reports better de-etherification of ethanol than butanol lignins in acidic media (0.1 molar HCl in 2:1 dioxane/ water at 100 °C). Further support of the slow deisopropoxylation of dissolved lignin is seen in Table S4 which shows that the molar sum of PX and IPA increases significantly faster and more at 121 °C than at 100 °C. This behavior and continuing increase suggest that lignin deisopropoxylation is not complete even after 130 min heat treatment at 121 °C.

Figure 9. Xylose release from isopropyl xylosides (PX) hydrolysis: ●, 100 °C; ▲, 110 °C; ■, 121 °C.

mmol PX/100 o.d. g SCS. It also suggests that xylo-oligomers present in EVAPP liquor (not quantified) do not hydrolyze to monomeric xylose and may require more severe heat treatment, i.e., temperatures higher than 121 °C. The molar balance of PX and (newly formed) monomeric xylose over the entire hydrolysis period shows a small decrease (4−6 mmol/100 o.d. g SCS), with higher loss occurring at 121 °C (see Table S4). The decrease suggests dehydration/ condensation of xylose to furfural and humins, respectively. This is supported by the presence of about 4 mmol furfural in heat-treated EVAPP liquor after 130 min at 121 °C. As for EVAPE heat treatment, the maximum molar amount of IPA formed during of EVAPP heat treatment is significantly larger than the stoichiometric amount of 25 mmol/100 o.d. g SCS expected from the PX hydrolysis reaction: C8H16O5 + H 2O ⇌ C3H 7OH + C5H10O5

4. CONCLUSIONS Spent liquors from SEW and SPW fractionation of SCS containing significant amounts of ethyl xylosides (EX) and isopropyl xylosides (PX), respectively, are subjected to heat treatment to recover alcohol and chemically bound xylose. The kinetics of alkyl xylosides hydrolysis shows that EX and PX are completely hydrolyzed to xylose and their respective alcohols. The hydrolysis does not require addition of any mineral acids since lignosulfonic acids, present in the spent liquors, provide the required acidity (pH of about 1.0) for the alkyl xylosides hydrolysis reaction. However, heat treatment at 121 °C or lower without acid addition is not sufficient for hydrolysis of xylo-oligomers to monomers. The rate of hydrolysis is higher for PX than EX at temperatures of 100 °C or higher, while the hydrolysis does not occur or only slightly occurs at 80 °C. At 121 °C, the hydrolysis of PX and EX is completed after 30 or

(2)

The release of IPA during heat treatment of EVAPP at 100, 110, and 121 °C is shown in Figure 10. As with EX hydrolysis, concentrations of IPA higher than 25 mmol suggests that the solvent is also covalently bound to lignin and non-xylose sugars. EVAPP liquor treated at 121 °C for 30 min contains 32 mmol of IPA, i.e., 7 mmol (or 28%) more than expected from PX hydrolysis stoichiometry. Figure 6 shows that SPW lignin contains 0.22 mol of isopropoxy units G

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research 70 min, respectively. Longer times at this temperature promotes xylose dehydration reactions. Higher concentrations of alcohols (EtOH or IPA) are obtained than the stoichiometric amounts expected from hydrolysis of the respective alkyl xylosides. It is argued that the additional alcohol is mostly coming from hydrolysis of ethoxylated (or isopropoxylated) lignin and of alkyl pyranosides of non-xylose sugars, i.e., arabinose, galactose, glucose and mannose. SEW and SPW lignins contain 0.18 and 0.22 ethoxy or isopropoxy groups per C9 lignin monomer unit, respectively. Their close molar ratio suggests that lignin alkylation is determined by the reactivity of lignin and not the solvent. The isopropoxylated lignin is harder to hydrolyze than ethoxylated lignin, and 130 min heat treatment at 121 °C is insufficient to fully release all IPA while these conditions appear to be adequate for lignin deethoxylation of EVAPE liquor.

■

■

REFERENCES

(1) Goncalves, A. R.; Ruzene, D. S.; Moriya, R. Y.; Oliveria, L. R. Pulping of Sugarcane Bagasse and Straw and the Biobleaching of the Pulps: Conditions, Parameters and Recycling of Enzymes. In Proceedings, 59th Appita Conference: Auckland, New Zealand, May 16−19, 2005; Australian Pulp and Paper Industry Technical Association: Carlton, Victoria, Australia, 2005. (2) Saska, M.; Gray, M. Pre-treatment of Sugarcane Leaves and Bagasse Pith with Lime Impregnation and Steam Explosion for Enzymatic Conversion to Fermentable Sugars. In Proceedings, 28th Symposium on Biotechnology for Fuels and Chemicals, Nashville, TN, April 30−May 2, 2006; Humana Press: Totowa, NJ, 2006. (3) Moutta, R. O.; Chandel, A. K.; Rodrigues, R.; Silva, M. B.; Rocha, G.; Silva, S. S. Statistical Optimization of Sugarcane Leaves Hydrolysis into Simple Sugars by Dilute Acid Catalyzed Process. Sugar Tech 2012, 14, 53−60. (4) You, X.; van Heiningen, A.; Sixta, H.; Iakovlev, M. Kinetics of SO2-Ethanol-Water (AVAP®) Fractionation of Sugarcane Straw. Bioresour. Technol. 2016, 212, 111−119. (5) Sklavounos, E.; Iakovlev, M.; van Heiningen, A. Study on SO2Ethanol-Water Spent Liquor from Spruce Chips/Softwood Biomass for ABE Fermentation. Ind. Eng. Chem. Res. 2013, 52, 4351−4359. (6) Iakovlev, M. SO2-Ethanol-Water (SEW) Fractionation of Lignocellulosics. School of Chemical Technology, Helsinki, 2011. (7) Yamamoto, M. SO2-Ethanol-Water Fractionation and Enzymatic Hydrolysis of Forest Biomass. School of Chemical Technology, Helsinki, 2014. (8) Sklavounos, E. Conditioning of SO2-Ethanol-Water (SEW) Spent Liquor from Lignocellulosics for ABE Fermentation to Biofuels and Chemicals. School of Chemical Technology, Helsinki, 2014. (9) Iakovlev, M.; van Heiningen, A. Efficient Fractionation of Spruce by SO2-Ethanol-Water: Closed Mass Balances for Carbohydrates and Sulfur. ChemSusChem 2012, 5, 1625−1637. (10) Sharazi, A. M.; van Heiningen, A. Ethyl Xylosides Formation in SEW (AVAP) Fractionation of Sugarcane Straw; Implications for Ethanol and Xylose Recovery. Holzforschung 2017. DOI: 10.1515/hf2017-0041. (11) Zhang, Z.; Harrison, M. D.; Rackemann, D. W.; Doherty, W. O.; O’Hara, I. M. Organosolv Pretreatment of Plant Biomass for Enhanced Enzymatic Saccharification. Green Chem. 2016, 18, 360−381. (12) Lancefield, C. S.; Panovic, I.; Deuss, P. J.; Barta, K.; Westwood, N. J. Pretreatment of Lignocellulosic Feedstocks using Biorenewable Alcohols; Towards Complete Biomass Valorization. Green Chem. 2017, 19, 202−214. (13) Drouet, P.; Zhang, M.; Legoy, M. D. Enzymatic Synthesis of Alkyl ß-D-Xylosides by Transxylosylation and Reverse Hydrolysis. Biotechnol. Bioeng. 1994, 43, 1075−1080. (14) Grisel, R.; van der Waal, J.; de Jong, E.; Huijgen, W. Acid Catalysed Alcoholysis of Wheat Straw: Towards Second Generation Furan Derivatives. Catal. Today 2014, 223, 3−10. (15) Sharazi, A. M.; van Heiningen, A. Kinetics of Sulfur dioxideAlcohol-Water (SAW) Pulping of Sugarcane Straw (SCS). Tappi J. 2017, 16, 313−328. (16) You, X.; van Heiningen, A.; Sixta, H.; Iakovlev, M. Lignin and Ash Balances of Sulfur dioxide-Ethanol-Water Fractionation of Sugarcane Straw. Bioresour. Technol. 2017, 244, 1111−1120. (17) Sumerskii, I.; Zweckmair, T.; Hettegger, H.; Zinovyev, G.; Bacher, M.; Rosenau, T.; Potthast, A. A Fast Track for Accurate Determination of Methoxy and Ethoxy Groups in Lignin. RSC Adv. 2017, 7, 22974−22982. (18) Iakovlev, M.; Päak̈ könen, T.; van Heiningen, A. Kinetics of SO2Ethanol-Water Pulping of Spruce. Holzforschung 2009, 63, 779−784.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting of Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.7b03613. Synthesis of isopropyl xylosides, calibrations, effect of washings and rotary evaporation on liquor molar content, molar balance for EX and PX hydrolysis (PDF)

■

C9, phenylpropanoid unit khyd, first-order hydrolysis rate constant (1/min) MDOB, multi digester oil bath PEG, polyethylene glycol

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: +1-207-581-2278. Fax: +1-207-581-2323. ORCID

Asif M. Sharazi: 0000-0002-2134-7017 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Financial support from the Fulbright Association and the J. Larcom Ober Chair, University of Maine Orono, ME, USA is highly appreciated. We are also thankful to Thomaston, GA, biorefinery of API for providing feedstock material for this research, and to Dr. Mikhail Iakovlev of API for his insightful and helpful suggestions for our research.

■

ABBREVIATIONS SCS, sugarcane straw SAW, SO2−alcohol−water MSAW, mixed SO2−alcohol−water liquor SEW, SO2−ethanol−water SPW, SO2−isopropanol−water MSEW, mixed SO2−ethanol−water liquor MSPW, mixed SO2−isopropanol−water liquor EVAP, liquor from rotary evaporation EVAPE, liquor from rotary evaporation of MSEW EVAPP, liquor from rotary evaporation of MSPW mi, initial mass (g) mf, final mass (g) mi/mf, mass concentration factor EX, ethyl xylosides PX, isopropyl xylosides Pyr, pyranosides IPA, isopropanol EtOH, ethanol H

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research (19) Retsina, T.; Pylkkanen, V. Separation of Lignin from Hydrolyzate. U.S. Patent 8585863B2, 2013. (20) Bauer, S.; Sorek, H.; Mitchell, V. D.; Ibáñez, A. B.; Wemmer, D. E. Characterization of Micanthus giganteus Lignin Isolated by Ethanol Organosolv Process under Reflux Conditions. J. Agric. Food Chem. 2012, 60, 8203−8212. (21) Timell, T. E. The Acid Hydrolysis of Glycosides. I. General Conditions and the Effect of the Nature of the Aglycone. Can. J. Chem. 1964, 42, 1456−1472.

I

DOI: 10.1021/acs.iecr.7b03613 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX