Molecular Products and Radicals from Pyrolysis of Lignin

Nov 6, 2012 - Lignin is a highly cross-linked polyphenolic polymer without any ..... Radical intermediates from lignin pyrolysis at 450 °C were colle...
5 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Molecular Products and Radicals from Pyrolysis of Lignin J. Kibet, L. Khachatryan, and B. Dellinger* Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803, United States ABSTRACT: Thermal degradation of lignin under two reaction regimes (pyrolysis in N2 and oxidative pyrolysis in 4% O2 in N2) has been investigated in a tubular, isothermal, flow-reactor over the temperature range 200−900 °C at a residence time of 0.2 s. Two experimental protocols were adopted: (1) Partial pyrolysis in which the same lignin sample was continuously pyrolyzed at each temperature and (2) conventional pyrolysis, in which new lignin samples were pyrolyzed at each pyrolysis temperature. The results identified common relationships between the two modes of experiments, as well as some differences. The majority of products from partial pyrolysis peaked between 300 and 500 °C, whereas for conventional pyrolysis reaction products peaked between 400 and 500 °C. The principal products were syringol (2,6-dimethoxy phenol), guaiacol (2-methoxy phenol), phenol, and catechol. Of the classes of compounds analyzed, the phenolic compounds were the most abundant, contributing over 40% of the total compounds detected. Benzene, styrene, and p-xylene were formed in significant amounts throughout the entire temperature range. Interestingly, six ringed polycyclic aromatic hydrocarbons were formed during partial pyrolysis. Oxidative pyrolysis did not result in large differences from pyrolysis; the main products still were syringol, guaiacol, phenol, the only significant difference being the product distribution peaked between 200 and 400 °C. For the first time, low temperature matrix isolation electron paramagnetic resonance was successfully interfaced with the pyrolysis reactor to elucidate the structures of the labile reaction intermediates. The EPR results suggested the presence of methoxyl, phenoxy, and substituted phenoxy radicals as precursors for formation of major products; syringol, guaiacol, phenols, and substituted phenols.

■

INTRODUCTION Biomass pyrolysis remains a critical chemical process in the utilization of renewable energy and feed stocks, generation of aromatic feed stocks, and destruction of forests by fires.1,2 The intricate nature and uncertain processes occurring in the pyrolysis of biomass have generated interest in the study of thermal degradation of lignin. Lignin is a highly cross-linked polyphenolic polymer without any ordered repeating units and is perhaps one of the most complex organic aromatic polymers in nature.3−6 The lignin fraction of biomass is an important source of benzene, phenol, and dihydroxybenzenes during burning.7 Among the major components of biomass, lignin presents the greatest difficulty in understanding the relationship between structure and the devolatilization mechanisms occurring during typical thermochemical conversion processes.8 This has been attributed to the complexity of its structure and the difficulty of isolating lignin without altering its structure.8 Structural Units of Lignin. The composition of the cell wall changes with the type of tree or plant, but in general 40− 45% of wood is cellulose, 25−35% hemicellulose, 15−30% lignin, and up to 10% other compounds.9,10 Linkages between the different components consist of hydrogen bonding and covalent ether, ester, and glycoside bonds. The structure is based on three different cinnamyl alcohols as precursors: pcoumaryl alcohol, coniferyl alcohol, and sinapyl alcohol compounds (cf. Figure 1A).9,10 The respective aromatic © 2012 American Chemical Society

constituents of these alcohols in the polymers are phydroxyphenyl (H), guaiacyl (2-methoxyphenyl), (G), and syringyl (2,6-dimethoxyphenyl), (S) units5,11 (cf. Figure 1B). The formulation of lignin and the ratio of the three units change with type of cell and plant. In view of this diversity, the exact chemical structure of any lignin cannot be resolved completely.12 Decomposition Mechanism of Lignin. Lignin decomposition occurs by several competing, bond-cleavage reactions at different temperatures depending on the bond energies.13 The most frequently studied reaction is the thermal scission of the α- and β- alkyl-aryl ether bonds (cf. Figure 1C) due to their prominent role in lignin chemistry.13 Ether-linkages in lignin are cleaved in heat treatment, leading to depolymerization of the lignin macromolecule, and formati on of many products with ether linkages.14 Lignin has a tendency to form volatile products when thermally decomposed between 200 and 500 °C.13,15 Thermogravimetric analysis of various lignin samples indicated the primary pyrolysis of lignin occurred between 200 and 400 °C,13,16,17 with the highest degradation rates occurring at ∼380 °C.18,19 This observation is attributed to the thermal Received: Revised: Accepted: Published: 12994

July 20, 2012 November 1, 2012 November 6, 2012 November 6, 2012 dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

that is, the same sample is continually heated at each pyrolysis temperature and (2) conventional pyrolysis, in which new lignin samples were pyrolyzed at each pyrolysis temperature, that is, a new sample is heated at every pyrolysis temperature. Partial pyrolysis is defined as a selective in situ conversion of biopolymers to desired products.24 The partial pyrolysis technique offers some advantages in comparison with conventional pyrolysis. First, only one loading of biomass material is used and can be heated multiple times. Second, it provides partial accumulation of any fraction and analysis for products in the gas phase as well as in the charred material. Third, the intermediate neutral, but unstable, products may be collected before they disappear in the secondary processes. Experimental Protocol for Pyrolysis. The thermal degradation of lignin was investigated in a straight-tube, isothermal flow-reactor over the temperature range 200−900 °C at atmospheric pressure, typically in 100 °C increments under two reaction regimes (pyrolysis in N2 and oxidative pyrolysis in 4% O2 in N2) using the System for thermal diagnostic studies, STDS.25 The gas flow rate was designed to maintain a constant residence time of 0.2 s. Thirty ±0.2 mg of lignin were loaded into the straight-tube isothermal quartz reactor (0.30 cm i.d. × 17.7 cm, volume 1.25 mL) and held in place by quartz wool to avoid being swept by carrier gas flowing through the reactor. The reactor containing the sample was then placed inside an electrically heated furnace at heating rate of 10 °C/sec for 3 min, before turning off the heater and cooling the sample with flowing N2 while exposing the reactor to a cooling fan. This method of thermolysis of lignin closely resembles the case where a boat is used to hold the sample in the reactor. The benefits of this technique are 2-fold: (1) the sample is held intact in the reactor, and (2) the carrier gas flows uniformly through the sample during the entire analysis, resulting in highly reproducible analyses. Due to high gas flow rates, the contact time with charred material is short enough (0.2 s) to minimize secondary reactions. Considering the challenging task involved in biomass pyrolysis, transport studies (test runs) were carried out at three selected temperatures 200, 400, and 900 °C in order to determine the reproducibility of major reaction products of lignin pyrolysis. All experimental points presented in Figures 36 are average values of two runs. GC-MS Analysis of Molecular Products. The mass of solid sample remaining after every pyrolysis temperature was determined by weight difference. The pyrolysate leaving the reactor in a transfer line (1 m × 0.53 mm) heated at 275 °C and entered a GC injection port where it was collected and condensed at the head of the column at −60 °C. The gas chromatography−mass spectroscopy (GC/MS) analysis of the pyrolysate was conducted with an Agilent 6890N gas chromatography equipped with a 5973N mass selective detector (MSD) with an ion source of electron impact (EI) at 24 ev to minimize extensive fragmentation.3 Two GC columns, a Gas-pro column (60 m × 0.32 mm i.d × 0.25 μm) for analysis of low molecular weight products, and a DB5-MS column (30 m ×0.25 mm × 0.25 μm) for the determination of high molecular weight products, were used. The temperature programming was typically: −60 °C initial temperature; holding for 3 min → heating rate of 15 °C/min →130 °C intermediate temperature; holding for 1 min → heating rate of 25 °C/min →300 °C for the DB5-MS column and 260 °C for the Gas-Pro column, with a final hold time of 5 min. The injector, FID detector, and MSD detector temperatures were

Figure 1. A. The three monolignols and B. H, G, and S derivatives (B), C. The main linkages in lignin polymer (β-O-4 and α-O-4) and substituted phenoxyl radical from monolignols.

scission of the α- and β- alkyl-aryl ether bonds, C−C and the C−O bonds that have lower bond dissociation energies (∼346 and 358 Kjmol−1, respectively) than the C−OCH3 bond (410 Kjmol−1), (cf. Figure 1C).13 All these processes involve appearance of free radicals, elimination of water, formation of carbonyl, carboxyl, and hydroperoxide groups (especially in air), evolution of CO and CO2, and eventually production of a charred residue.8,20 Consequently, these findings point to the importance of interaction of various functional groups and their influence on the thermal decomposition of lignin.13 Nevertheless, lignin is believed to thermally decompose via a free radical mechanism.8,21,22 We here report on the pyrolytic decomposition of lignin using the System for Thermal Diagnostic Studies (SDTS) to analyze for molecular products and Low Temperature Matrix Isolation EPR (LTMI-EPR) to identify free radical intermediates. These data are discussed in relation to the mechanism of lignin decomposition and the toxicity of its decomposition byproducts.

■

EXPERIMENTAL SECTION Materials. The lignin (hydrolytic lignin) used for this study was obtained from Sigma Aldrich Inc. (USA) and was used without further treatment. This is the type of lignin, prepared when dilute sulphuric acid is used to hydrolyze most of the polysaccharides to produce fermentable sugars, leaving lignin as a solid byproduct.23 Partial and Conventional Pyrolysis. Two experimental protocols were adopted: (1) Partial pyrolysis in which the same lignin sample was continuously pyrolyzed at each temperature, 12995

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

250, 275, and 280 °C, respectively. Ultra high purity (UHP, 99.99%) helium was used as the carrier gas at constant flow of 3.3 mL/min. The MS was operated in the total ion current mode (TIC) with a mass scan range of 15−600 amu. The products were identified using the NIST Library. Only compounds with percent yield contributions greater than 1% are reported. EPR Analysis of Radicals. To elucidate the existence of gas-phase radicals in the thermal degradation of lignin, pyrolysis was investigated in an isothermal flow reactor in conjunction with a coldfinger - EPR (electron paramagnetic resonance) assembly depicted in Figure 2.26 The term (EPR) refers to the

resonant absorption of the electromagnetic radiation by electronic systems which possess permanent magnetic moments due to the orbital as well as spin angular momentum of electrons which are therefore paramagnetic.27 Accordingly, EPR is a spectroscopic technique used to detect species having one or more unpaired electrons (e.g., radicals). A straight tube isothermal quartz flow reactor (10 mm × 50 mm) was used for pyrolysis of lignin at a fixed temperature 450 °C. 10−15 mg of lignin was loaded into the inlet of the reactor at ∼200 °C and held in place by quartz wool. Elimination of low molecular products of lignin pyrolysis initiated between 50 and 150 °C.20,28 The flow of N2/CO2 gas at less than 0.3 Torr pressure swept the evaporated volatile components into the reactor. The pyrolyzed products exiting the reactor were pumped directly onto a coldfinger. The CO2 carrier easily freezes at liquid nitrogen temperature, creating an ideal matrix for condensation of radicals.26 To avoid product condensation on the walls, all transfer lines from the reactor to the EPR cavity were maintained at 100 °C regardless of the pyrolysis reactor temperature. The Dewar was also equipped with a special PTFE pressure − vacuum valve (PV-ANV, Wilmad) which allowed the Dewar (maintained at liquid N2 temperature) to be separated from the reactor and evacuated to 10−4 Torr for EPR analysis. To generate reference phenoxy-type radicals, the frozen aquatic solutions of different phenols in 4 mm EPR tubes were subjected to UV photolysis in a Dewar with liquid nitrogen at 253.7 nm. The 253.7 nm light was generated using a conventional, mercury vapor, ozone-free pencil lamp from Jelight, Inc. This double bore lamp, with a 9 mm OD, produced a 4 in. light at a power of ∼9 mW/cm2 at 254 nm measured at a distance of 15 mm from the lamp. The phenoxy, ohydroxyphenoxy, and p-hydroxyphenoxy radicals were also

Figure 2. Cold finger assembly for LTMI-EPR.

Figure 3. Yields (based on GC area counts) of major oxygenated products (A−C) and hydrocarbons (D) from partial pyrolysis of lignin in N2. 12996

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

Figure 4. Yields (based on GC area counts) of major oxygenated products (A and B) from conventional pyrolysis of lignin in N2.

Figure 5. Yields (based on GC area counts) of major oxygenated products (A) and hydrocarbons (B) from partial oxidative pyrolysis of lignin in 4% oxygen in N2.

marker compounds of lignin (syringyl, guaiacyl, and hydroxyphenyl units) should be the major products. Of the classes of compounds analyzed, phenols (phenol, p-cresol, and catechol), syringol, 4-propenyl syringol, and guaiacols (guaiacol, eugenol, 4-ethylguaiacol, and 5-methylguaiacol etc.) were the most abundant products contributing over 40% of the total compounds analyzed. Acetic acid and furfuryl alcohol achieved a maxima at ∼330 °C while methanol, furan, 2-methyl furan, and 2,5-dimethylfuran maxima were at ∼430 °C (cf. Figure 3C). The low molecular weight, oxygenated products peaked between 250 and 400 °C, while the majority of the phenolic compounds exhibited maxima between 350 and 500 °C. The aromatic hydrocarbons (benzene, toluene, and styrene) exhibited maxima between 500 and 700 °C (cf. Figure 3D). The benzene concentration peaked at ∼650 °C, whereas that of toluene peaked at 520 °C. Conventional Pyrolysis. Product distributions for pyrolysis of fresh lignin samples at every temperature were very similar to results from partial pyrolysis of lignin; however, the concentration maxima were >400 °C (cf. Figure 4). Syringol and 4-vinylguaiacol were the primary products (cf. Figure 4A) while catechol and phenol were the main products from the simple phenol family (cf. Figure 4B). Some compounds, such as 3-methoxycatechol and 3,4-dimethyl phenol did not increase significantly with increased temperature,(cf. Figure 4B). Partial Oxidative Pyrolysis. The maximum product distributions were between 200 and 400 °C. (cf. Figure 5). The major products were syringol, guaiacol and phenol, (cf. Figure 5A. The syringol maximum was at ∼350 °C, whereas

produced from gas-phase photolysis of phenol, catechol and hydroquinone, respectively, at room temperature and very low pressure (≤0.1 Torr). All EPR spectra were recorded on a Bruker EMX-20/2.7 EPR spectrometer (X-band) with dual cavities, modulation and microwave frequencies of 100 kHz and 9.516 GHz, respectively. The typical parameters were: sweep width of 200 G, EPR microwave power of 1−20 mW, and modulation amplitude of ≤4 G. Time constant and sweep time were varied. Values of gfactors were calculated using Bruker’s WINEPR program, which is a comprehensive line of software, allowing control of the Bruker EPR spectrometer, data-acquisition, automation routines, tuning, and calibration programs on a Windows-based PC.29 The exact g-values for key spectra were determined by comparison with a 2,2-diphenyl-1-picrylhydrazyl (DPPH) standard. In some experiments, gradual warming of the Dewar was employed to allow annealing of the matrix and annihilation of mobile or very reactive radicals. This resulted in production of cleaner, sharper spectra of single radicals under environmentally isolated conditions.

■

RESULTS AND DISCUSSION Partial Pyrolysis. The primary compounds detected and their relative distributions are presented in Figure 3. Syringol, 4propenyl syringol, guaiacol (and its derivatives) were the most abundant products of lignin pyrolysis (cf. Figure 3A). The second most abundant products were catechol, phenol, and their derivatives (cf. Figure 3B). These data are consistent with work performed by other researchers, indicating the three 12997

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

Figure 6. Yields (based on GC area counts) of major oxygenated products (A and B) from conventional oxidative pyrolysis of lignin in 4% oxygen in N2.

guaiacol and phenol maxima were at ∼330 °C. While formaldehyde and acetone were formed in significant quantities under oxidative pyrolysis, they were only formed in trace quantities under pyrolysis. Formaldehyde achieved a maximum concentration at ∼250 °C while acetone achieved a maximum concentration at ∼350 °C (cf. Figure 5A). Most of the phenol compounds, that is, catechol, and 3-methoxyphenol were formed in low yields, compared to pyrolysis. PAHs were not observed, probably due to oxidation of precursors.30 The already partially oxidized lignin components, that is. syringol, guaiacol and phenol did not exhibit a significant decrease in yield. Conventional Oxidative Pyrolysis. The maximum yields for most compounds were at slightly lower temperature, 400 − 450 °C, with syringol, guaiacol, catechol, and phenol being the dominant products above 300 °C. Only a few hydrocarbon products were detected, including benzene, toluene, and pxylene As a general rule the product yields in conventional pyrolysis (or conventional oxidative pyrolysis) should be higher than that in partial oxidative (or conventional partial oxidative) pyrolysis due to specificity of these two processes as described in the Experimental Section. Particularly, the same sample is continually heated under partial pyrolysis, and this leads to exhaustion of products from the sample due to continuously decreasing amounts of initial lignin concentration at each pyrolysis temperature. However, some product yields may show unique behavior, for instance in the case of catechol and phenol, Figures 5A and 6B. It is well established in general, catechol is a product formed due to further secondary reactions of guaiacol (one of the major products).31,32 As a result, catechol yields mimic the yields of guaiacol although in much lower yields (Figures 5 and 6). A significant difference between guaiacol and catechol has been observed under oxidative partial pyrolysis (Figure 5A). From this observation, it would appear catechol oxidizes much faster under oxidative partial pyrolysis in comparison to phenol (a similar product as catechol). Interestingly, a highly oxidative environment has been observed in char formation process during partial oxidative pyrolysis, Figure 7. This may imply that at each pyrolysis temperature, some highly active intermediate species (for instance hydroperoxides) may adsorb on char surfaces during cool-down processes and initiate the process of lignin pyrolysis at the next pyrolysis temperature. Based on polarization data which shows the dipole moment of catechol is significantly higher (2.21D)

Figure 7. % char yields for pyrolysis of lignin 4% oxygen in N2 and oxidative pyrolysis of lignin in in N2.

than that of phenol (1.54 D), catechol being highly polar is better adsorbed on char surfaces than phenol.33,34 This means the amounts of catechol adsorbed on the surface is much higher than that of phenol. Therefore, less amounts of catechol are released into the gas phase due to decomposition of catechol by adsorbed intermediates, Figure 5A. Char Yield. The thermal degradation profile of lignin under a wide range of pyrolysis conditions is presented in Figure 7. At 200 °C, the weight loss of lignin under pyrolytic conditions (partial and conventional pyrolysis) was small, however; a rapid weight loss of ∼20% was recorded between 300 and 400 °C. For partial oxidative pyrolysis, the weight loss was more rapid over the same temperature range, viz. 40%. A percent weight loss of ∼30% was observed for conventional oxidative pyrolysis. Two fundamental temperature zones were observed in the decomposition profile of lignin. The first zone, with high weight loss (200−500 °C), yielded the majority of the volatile components (cf. Figures 5−6). The second stage of weight loss (500−900 °C), the decomposition of lignin was nearly constant for pyrolysis experiments, and the lignin char was largely aromatic. This resulted in the formation of hydrocarbon products such as, propene, propane, benzene, toluene, and styrene, etc., (cf. Figure 3D). Radicals from Conventional Pyrolysis. Radical intermediates from lignin pyrolysis at 450 °C were collected and analyzed using the LTMI-EPR technique. A representative spectrum of trapped radicals at 77 K is depicted in Figure 8, spectrum 1. The spectrum is an unstructured singlet (with some anisotropy) with g = 2.0072 and ΔHp-p = 14.0G. The 12998

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

was subtracted from the spectrum of EPR radicals from lignin pyrolysis (cf. Figure 9, spectrum 1) a residue spectrum was observed with a high g-value of 2.0064 and ΔHp-p = 18G (cf. Figure 9, spectrum 3). This difference in spectrum closely resembles that of a phenoxyl or substituted phenoxy, such as a hydroxyphenoxyl (neutral semiquinone radical).44 Indeed, the radicals from phenol and hydroquinone/catechol pyrolysis (and photolysis), produced as molecular products from lignin decomposition, have previously been identified as phenoxy and semiquinone radicals, respectively.36,42,44−47 These EPR spectra were structureless singlet lines detected by the LTMI-EPR technique at 77 K. The phenoxy radical spectrum exhibited a broader (ΔHp-p = 16G) than semiquinone radical (ΔHp-p = 12G).43 Because phenoxy linkages are key structural units while semiquinones are secondary linkages, phenoxy-type radicals may be higher in concentration than semiquinone radicals from lignin pyrolysis.48 Accordingly, it can be concluded that intermediate radicals are mostly derived from phenolic linkages in lignin and are probable precursors for formation of phenolic compounds, that is, 2,6 - dimethoxy phenoxy (syringyl groups), 2-methoxy phenoxy (gualacyl groups), and phenols for (phenoxy goups) etc. For this to be true, these intermediate radicals should be present in residue EPR spectrum. Additionally, this argument is supported by results from GC-MS analyses which indicate that phenolic compounds are the major reaction products of lignin pyrolysis. The yields of the principal phenol-type products drop significantly in the order: syringol > guaiacol > phenol > cresols∼catechol (cf. Figure 3). A key issue is the broad character of the EPR spectra detected from lignin pyrolysis. Comparing the broadening effect of substituent groups on EPR spectra of phenoxy radical is useful to understand this. For instance, the position and number of Cl atoms on the aromatic ring, as a typical electronegative (electron-withdrawing) substituent, slightly affects the total spectral width.49 The g-value slowly increases from g = 2.0062 for mono-, to g = 2.0065 for di-, and g = 2.0076 for trichlorophenoxy radicals (the g-value for pure phenoxy is g = 2.0053).49 In contrast to chlorine substituents, methyl group are electron-donating and broaden the EPR spectra of phenoxy.50 Methoxy substituted phenoxy radicals, which form in lignin pyrolysis, may have dual impacts on total EPR linewidth, because of their ability to be either electron-donating or electron-withdrawing, depending on the position of substitution.50,51 The spectral width of EPR spectra presented in the residue spectrum is broader (ΔHp-p = 18G) than the phenoxyl radical EPR spectrum (ΔHp-p = 16G) detected from phenol pyrolysis using the same LTMI-EPR technique.36 To determine if the observed spectra were of substituted phenoxyls, additional experiments were initiated. Radicals were generated by UV photolysis of hydroquinone (HQ), catechol

Figure 8. The EPR spectra of radicals accumulated on coldfinger from lignin pyrolysis at 450 °C (spectrum 1, g = 2.0071, ΔHp-p = 13.5G) and from Burley tobacco pyrolysis at 450 °C (spectrum 2, g = 2.0056, ΔHp-p = 13G).

small peaks on both sides of the main spectrum (marked with an asterisk in Figure 8) indicate the presence of trace quantities of oxygen as E-lines (K = 1, J = 2, M = 1→2).35 These are readily removed by annealing.36 Because the pyrolysis of tobacco has much in common with the pyrolysis of lignin,37,38 an EPR spectrum from Burley tobacco pyrolysis at 450 °C in the presence of less than 1 Torr of air was overlaid with the spectrum of lignin (cf. Figure 8, spectrum 2). The tobacco spectral parameters were g = 2.0056 and ΔHp-p = 13G. Both spectra were similar and exhibited similar anisotropy, which is believed to be due RO•2 easily formed in the pyrolysis of tobacco, catechol, hydroquinone, and other organics in presence of small quantities of oxygen.36,39−43 When the expected spectrum of RO•2 (cf. Figure 9, spectrum 2)

Figure 9. The EPR spectra of radicals accumulated on coldfinger from lignin pyrolysis at 450 °C and 0.1 Torr air (black line, g = 2.0073, ΔHp-p = 15.0 G) and overlaid red reference EPR spectrum of RO•2 (g = 2.0089) produced from heating of tobacco to 450 °C in vacuum. The blue spectrum (g = 2.0064, ΔHp-p = 18G) is the subtraction spectrum of the lignin and RO•2 .

Table 1. EPR Parameters of Radicals Generated by UV Photolysis of Hydroquinone (HQ), Catechol (CT), Phenol (PhOH) and Some Substituted Phenols in Frozen Aquatic Solution, pH 7.0a HQ ΔH p-p, G g-value molarity, M

12.5 2.0049 8.0 × 10−2

HQ 11.0 2.0049 annealinga

HQ 9.5 2.0050 annealinga

HQ 11.5 2.0042 na

CT 15.5 2.0058 1 x10−1

CT 12.7 2.0049 na

PhOH c

16.0−21.0 2.0051 6.0 × 10−2

PhOH

tyrosineb

4-Cl−PhOH

21.0 2.0050 na

21.0 2.0048 5.0 × 10−3

19.0 2.0063 1 × 10−1

a na, radicals were generated from very low pressure, gas-phase photolysis of precursors and accumulated on the cold finger at 77 K. aGradual annealing of the frozen solution of HQ at 8.0 × 10−2 M after UV irradiation. bTyrosine: (OH)C6H4CH2CH(NH2)CO2H. cDepending on irradiation time

12999

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

Notes

(CT), phenol (PhOH) and some substituted phenols from their frozen aquatic solutions. The UV photoexcitation of phenol resulted in partial photo dissociation to phenoxyl radical and a hydrogen atom,52,53 and the photodecomposition of HQ/CT should occur similarly.54 All EPR spectra generated were simple unstructured singlet lines (cf. summary in Table 1). The common feature for all spectra was the high g-values characteristic for oxygen centered radicals 55 and broad singlet lines. The ΔHp-p for radicals produced from phenol, tyrosine and 4-chlorophenol were much broader (19−21 G) than for radicals from HQ or CT (10−15G) (cf. Table 1). The effect of concentration broadening on the EPR spectra of radicals (hydroxyphenoxyl or neutral semiquinone radical) produced by UV photolysis of frozen aquatic solutions of hydroquinone is clear from the data in Table 1.47 For instance, the ΔHp-p = 12.5 G for semiquinone radicals derived from stock solution of HQ (normalized intensity, I = 1.5, arbitrary units) dropped slowly by annealing procedure to ΔH p-p = 11.0 G (I = 0.14) and ΔH p-p = 9.5.0 G (I = 0.07) at almost the same g-value (Table 1). The broad signals derived from phenol, tyrosine and 4-chlorophenol (ΔH p-p = 19−21 G) most resembled the signal produced from lignin pyrolysis (ΔH p-p = 18.0 G), with a high g-value of 2.0064. Due to their high g-value and broad line-width, the EPR data strongly suggest the EPR spectra from lignin gas−phase pyrolysis are phenoxy and substituted phenoxy radicals. To the best of our knowledge, these EPR data supported by molecular product analysis are new and successfully identify the intermediate character of radicals in the gas-phase pyrolysis of lignin. Toxicological Considerations of Phenoxy and Substituted Phenoxy radicals. The toxicology of intermediate radicals and molecular products from pyrolysis of lignin is a subject of environmental significance. It is known that phenoxy and semiquinone radicals produced from biomass and tobacco burning are resonance stabilized environmentally persistent free radicals (EPFRs) with long lifetimes and may cause extensive cellular damage, DNA damage, oxidative stress, tumors, and cancer.55−58 The phenolic compounds, including substituted phenols, guaiacol, syringol, catechol, 4-vinyl guaiacol, and vanillin, found in biomass burning and cigarette smoke, as well as in this work, are also considered lethal.56,59 Phenol by itself affects liver enzymes, lungs, kidneys, cardiovascular system, and may attack the nervous system.60 Following H atom abstraction from the phenol hydroxyl group, the resultant phenoxy radical exhibits some electron-deficient character 61 which would be stabilized by electron-donating substituents such as amino, methoxy, and methyl groups. Consequently, phenols substituted in this way possess longer half-lives and may lead to health issues.55−58 Such radicals with longer lifetimes are considered persistent free radicals (PFRs) and are thus very toxic. Additionally, phenoxy radicals are precursors for formation of unchlorinated dibenzo-p-dioxin/dibenzofuran, which are easily chlorinated by a little chlorine and a redoxactive transition metal to form polychlorinated dibenzo-pdioxin/dibenzofurans (PCDD/Fs).62−68 The experimental data presented here establish a critical base for further elucidation and modeling of the gas-phase pyrolysis of lignin as well as environmental implications.

■

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge partial support of this work through Reynolds’ Tobacco Company and the Patrick F. Taylor Chair held by Dr. Barry Dellinger.

■

REFERENCES

(1) Shin, E. J.; Nimlos, M. R.; Evans, R. J. A study of the mechanisms of vanillin pyrolysis by mass spectrometry and multivariate analysis. Fuel 2001, 80 (12), 1689−1696. (2) Jiang, G. Z.; Nowakowski, D. J.; Bridgwater, A. V. Effect of the temperature on the composition of lignin pyrolysis products. Energy Fuels 2010, 24, 4470−4475. (3) Sharma, R. K.; et al. Characterization of chars from pyrolysis of lignin. Fuel 2004, 83 (11−12), 1469−1482. (4) Wan, J. K. S.; Depew, M. C. Applications of ESR and CIDEP to mechanistic studies of lignin chemistry. Res. Chem. Intermed. 1998, 24 (8), 831−847. (5) Lewis, N. G.; Yamamoto, E. Lignin - Occurrence, Biogenesis and Biodegradation. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1990, 41, 455−496. (6) Kuzina, S. I.; et al. Free radicals in the photolysis and radiolysis of polymers: IV. Radicals in gamma- and UV-irradiated wood and lignin. High Energy Chem. 2004, 38 (5), 298−305. (7) Scheijen, M. A.; et al. Evaluation of a tobacco fractionation procedure using pyrolysis mass-spectrometry combined with multivariate-analysis. Beitr. Tabakforsch. Int. 1989, 14 (5), 261−282. (8) Evans, R. J.; Milne, T. A.; Soltys, M. N. Direct mass-spectrometric studies of the pyrolysis of carbonaceous fuels 0.3. Primary pyrolysis of lignin. J. Anal. Appl. Pyrolysis 1986, 9 (3), 207−236. (9) Dorrestijn, E.; et al. The occurrence and reactivity of phenoxyl linkages in lignin and low rank coal. J. Anal. Appl. Pyrolysis 2000, 54 (1−2), 153−192. (10) Baliga, V.; et al. Physical characterization of pyrolyzed tobacco and tobacco components. J. Anal. Appl. Pyrolysis 2003, 66 (1−2), 191−215. (11) Freudenb.K, LIGNINIts constitution and formation from phydroxycinnamyl alcohols. Science, 1965, 148(3670), 595 (12) Yuan, T. Q.; et al. Structural characterization of lignin from triploid of populus tomentosa carr. J. Agric. Food Chem. 2011, 59 (12), 6605−6615. (13) Jakab, E.; Faix, O.; Till, F. Thermal decomposition of milled wood lignins studied by thermogravimetry mass spectrometry. J. Anal. Appl. Pyrolysis 1997, 40 (1), 171−186. (14) Brezny, R.; Mihalov, V.; Kovacik, V. Low-temperature thermolysis of lignins 0.1. Reactions of beta-o-4 model compounds. Holzforschung 1983, 37 (4), 199−204. (15) Ramiah, M. V., Thermogravimetric and differential thermal analysis of cellulose, hemicellulose, and lignin. J. Appl. Polym. Sci., 1970. 14 (5), 1323 (16) Kawamoto, H.; Ryoritani, M.; Saka, S. Different pyrolytic cleavage mechanisms of beta-ether bond depending on the side-chain structure of lignin dimers. J. Anal. Appl. Pyrolysis 2008, 81 (1), 88−94. (17) Nakamura, T.; Kawamoto, H.; Saka, S. Pyrolysis behavior of Japanese cedar wood lignin studied with various model dimers. J. Anal. Appl. Pyrolysis 2008, 81 (2), 173−182. (18) Jiang, Z. H.; Argyropoulos, D. S. Coupling P-31 NMR with the Mannich reaction for the quantitative analysis of lignin. Can. J. Chem. 1998, 76 (5), 612−622. (19) Ben, H. X.; Ragauskas, A. J. NMR characterization of pyrolysis oils from Kraft lignin. Energy Fuels 2011, 25 (5), 2322−2332. (20) Shafizadeh, F. Introduction to pyrolysis of biomass. Journal of Analytical and Applied Pyrolysis 1982, 3 (4), 283−305. (21) Cai, X. M.; Dass, C. Conformational analysis of proteins and peptides. Curr. Org. Chem. 2003, 7 (18), 1841−1854.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. 13000

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001

Environmental Science & Technology

Article

(22) Britt, P. F.; et al. Pyrolysis mechanisms of lignin - surfaceimmobilized model-compound investigation of acid-catalyzed and freeradical reaction pathways. J. Anal. Appl. Pyrolysis 1995, 33, 1−19. (23) Jiang, G. Z.; Nowakowski, D. J.; Bridgwater, A. V. A systematic study of the kinetics of lignin pyrolysis. Thermochim. Acta 2010, 498 (1−2), 61−66. (24) Agblevor, F. A.; et al. Fractional catalytic pyrolysis of hybrid poplar wood. Ind. Eng. Chem. Res. 2010, 49 (8), 3533−3538. (25) Rubey, W. A.; Grant, R. A. Design aspects of a modular instrumentation system for thermal diagnostic studies. Rev. Sci. Instrum. 1988, 59 (2), 265−269. (26) Khachatryan, L.; et al. Formation of cyclopentadienyl radical from the gas-phase pyrolysis of hydroquinone, catechol, and phenol. Environ. Sci. Technol. 2006, 40 (16), 5071−5076. (27) Lancaster, G. Electron paramagnetic resonance (a review). J. Mater. Sci. 1967, 2 (5), 489−495. (28) Serio, M. A.; et al. Measurement and modeling of lignin pyrolysis. Biomass Bioenergy 1994, 7 (1−6), 107−124. (29) Eaton, G. R., et al., Quantitative EPR; Springer Wien: New York , 2010; p 185. (30) Sharma, R. K.; Hajaligol, M. R. Effect of pyrolysis conditions on the formation of polycyclic aromatic hydrocarbons (PAHs) from polyphenolic compounds. J. Anal. Appl. Pyrolysis 2003, 66 (1−2), 123−144. (31) Asmadi, M.; Kawamoto, H.; Saka, S. Thermal reactions of guaiacol and syringol as lignin model aromatic nuclei. J. Anal. Appl. Pyrolysis 2011, 92 (1), 88−98. (32) Dorrestijn, E.; Mulder, P. The radical-induced decomposition of 2-methoxyphenol. J. Chem. Soc., Perkin Trans. 2 1999, No. 4, 777−780. (33) Lander, J. J.; Svirbely, W. J. The Dipole Moments of Catechol, Resorcinol and Hydroquinone. J. Am. Chem. Soc. 1945, 67 (2), 322− 324. (34) Goode, E. V.; Ibbitson, D. A. Dipole Moments of a Series of Substituted Phenols in Benzene. J. Chem. Soc. 1960, No. NOV, 4265− 4270. (35) Carlier, M.; Pauwels, J. P.; Sochet, L.-R. Application of ESR techniques to the study of gas-phase oxidation and combustion phenomena. Oxid. Commun. 1984, 6 (1−4), 141−156. (36) Khachatryan, L.; Adounkpe, J.; Dellinger, B. Radicals from the gas-phase pyrolysis of hydroquinone: 2. Identification of alkyl peroxy radicals. Energy Fuels 2008, 22 (6), 3810−3813. (37) Sharma, R. K.; et al. Characterization of char from the pyrolysis of tobacco. J. Agric. Food Chem. 2002, 50 (4), 771−783. (38) Zhou, S.; et al. Pyrolysis behavior of pectin under the conditions that simulate cigarette smoking. J. Anal. Appl. Pyrolysis 2011, 91 (1), 232−240. (39) Pryor, W. A.; Prier, D. G.; Church, D. F. ESR Study of mainstream and sidestream cigarette smoke: Nature of free radicals in gas-phase smoke and in cigarette tar. Environ. Health Perspect. 1983, 47, 345−355. (40) Flicker, T. M.; Green, S. A. Comparison of gas-phase free radical populations in tobacco smoke and model systems by HPLC. Environ. Health Perspect. 2001, 109, 765−771. (41) Valavanidis, A.; Haralambous, E. A comparative study by EPR of free radical species in the mainstream and sidestrem smoke of sigarettes with conventional acetate filters and ’bio-filters″. Redox.Report 2001, 6 (3), 161−171. (42) Adounkpe, J.; et al. Radicals from the atmospheric pressure pyrolysis and oxidative pyrolysis of hydroquinone, catechol and phenol. Energy Fuels 2009, 23, 1551−1554. (43) Khachatryan, L.; Adounkpe, J.; Dellinger, B. Radicals from the gas-phase pyrolysis of hydroquinone 2. Identification of alkyl peroxy radicals. Energy Fuel 2008, 22, 3810−3813. (44) Khachatryan, L.; Adounkpe, J.; Dellinger, B. Formation of phenoxy and cyclopentadienyl radicals from the gas-phase pyrolysis of phenol. J. Phys.Chem., A 2008, 112, 481−487. (45) Khachatryan, L.; et al. Radicals from the gas-phase pyrolysis of catechol. 2. Comparision of the pyrolysis of catechol and hydroquinone. J. Phys. Chem., A 2010, 114, 10110−10116.

(46) khachatryan, l.; et al. Radicals from the gas-phase pyrolysis of catechol: 1. O-semiquinone and ipso-catechol radicals. J. Phys. Chem., A 2010, 114 (6), 2306−2312. (47) Adounkpe, J.; Khachatryan, L.; Dellinger, B. Radicals from the gas-phase pyrolysis of hydroquinone 1. Temperature dependence of the total yields of radicals. Energy Fuels 2008, 22, 2986−2990. (48) Leary, G., Chemistry of reactive lignin intermediates 0.1. Transients in coniferyl alcohol photolysis. J. Chem. Soc., Perkin Trans. 2, 1972 (5), 640. (49) Graf, F.; Loth, K.; Gunthard, H.-H. Chlorine hyperfine splittings and spin density distribution of peroxy radicals. An ESR and Quantum chemical study. Helv. Chim. Acta 1977, 60 (76), 710−721. (50) Stone, T. J.; Waters, W. A. Aryloxy-radicals. Part1. ESR spectra of radicals from some substituted monohydric phenols. J. Chem. Soc. 1964, 213, 213−218. (51) Hansch, C.; Leo, A.; Taft, R. W. A survey of hammett substituent constants and resonance and field parameters. Chem. Rev. 1991, 91 (2), 165−195. (52) Bussandri, A.; Willigen, H. van FT-EPR study of the wavelength dependence of the photochemistry of phenols. J. Chem. Phys. 2002, 106, 1524−1532. (53) Jeevarajan, A. S.; Fessenden, R. W. Unusual chemically induced dynamic electron polarization of electrons by photoionization. J. Chem. Phys. 1992, 96, 1520−1523. (54) Calvert, J. G., Pitts, J. N. Jr, Photochemistry; John Wiley & Sons, Inc., 1966; 499. (55) Dellinger, B.; et al. Formation and stabilization of persistent free radicals. Proc. Combust. Inst. 2007, 31, 521−528. (56) Smith, C. J.; Hansch, C. The relative toxicity of compounds in mainstream cigarette smoke condensate. Food Chem. Toxicol. 2000, 38 (7), 637−646. (57) Dellinger, B.; et al. Role of free radicals in the toxicity of airborne fine particulate matter. Chem. Res. Toxicol. 2001, 14 (10), 1371−1377. (58) Dellinger, B.; et al. The role of combustion-generated radicals in the toxicity of PM2.5. Proc. Combust. Inst. 2000, 28, 2675−2681. (59) Selassie, C., D.; et al. On the toxicity of phenols to fast growing cells. A QSAR model for a radical-based toxicity. Journal of the Chemical Society. Perkin Transactions 2 1999, No. 12, 2729−2733. (60) Talhout, R.; et al. Hazardous compounds in tobacco smoke. Int. J. Environ. Res. Public Health 2011, 8 (2), 613−628. (61) Dellinger, B.; et al. Formation and stabilization of persistent free radicals. Proc. Combust. Inst. 2007, 31, 521−528. (62) Evans, C. S.; Dellinger, B. Mechanisms of dioxin formation from the high-temperature pyrolysis of 2-bromophenol. Environ. Sci. Technol. 2003, 37 (24), 5574−5580. (63) Evans, C. S.; Dellinger, B. Mechanisms of dioxin formation from the high-temperature pyrolysis of 2-chlorophenol. Environ. Sci. Technol. 2003, 37 (7), 1325−1330. (64) Louw, R.; Ahonkhai, S. I. Radical/radical vs radical/molecule reactions in the formation of PCDD/Fs from (chloro)phenols in incinerators. Chemosphere 2002, 46 (9−10), 1273−1278. (65) Khachatryan, L.; Asatryan, R.; Dellinger, B. Development of expanded and core kinetic models for the gas phase formation of dioxins from chlorinated phenols. Chemosphere 2003, 52 (4), 695− 708. (66) Nganai, S.; Lomnicki, S.; Dellinger, B. Ferric oxide mediated formation of PCDD/Fs from 2-monochlorophenol. Environ. Sci. Technol. 2009, 43 (2), 368−373. (67) Nganai, S.; Lomnicki, S.; Dellinger, B. Formation of PCDD/Fs from oxidation of 2-monochlorophenol over an Fe2O3/silica surface. Chemosphere 2012, 88 (3), 371−376. (68) Nganai, S.; Lomnicki, S. M.; Dellinger, B. Formation of PCDD/ Fs from the copper oxide-mediated pyrolysis and oxidation of 1,2dichlorobenzene. Environ. Sci. Technol. 2011, 45 (3), 1034−1040.

13001

dx.doi.org/10.1021/es302942c | Environ. Sci. Technol. 2012, 46, 12994−13001