Nitrogen Migration and Conversion in Chars from Co-Pyrolysis of Lignocellulose Derived Pyrolysis Model Compounds and Urea-Formaldehyde Resin Adhesive

: In thermal conversion utilization, nitrogen-rich biomass such as waste wood-based panels will release a large amount of NO x into the atmosphere, causing serious harm to the surroundings. By means of co-pyrolysis, N in waste wood-based panels can be ﬁxed in chars instead of discharging into the atmosphere in the form of volatile matter, which can reduce NO x emission and lay a foundation for the preparation of nitrogen-rich carbon materials with high added value. As the most commonly used adhesive in the production of wood-based panels, urea-formaldehyde resin adhesive (UF) is the main nitrogen source in waste wood-based panels. Therefore, the purpose of this paper is to explore the effects of glucose, ethyl maltol and 2,5-dimethyl-4-hydroxy-3(2H)-furanone (DMHF) on nitrogen migration and conversion during UF pyrolysis by adjusting the different proportions of model compounds and UF. Thermogravimetric analysis showed that ethyl maltol and DMHF had lower thermal stability and the pyrolysis process was concentrated in the range of 90–168 ◦ C, which does not coincide with the mass loss temperature of UF. UF can promote the pyrolysis of these three model compounds at the initial stage to some extent. The elemental analysis showed that the N retention in co-pyrolysis chars increased in varying degrees with the increase of the addition of model compounds; the nitrogen retention in chars of glucose-UF, ethyl maltol-UF and DMHF-UF increased by 28.47%, 3.48% and 16.45% with the increase of the model compound content from 50% to 90%, respectively. The XPS results showed that the relative content of N-6 in chars increased with the increase of ethyl maltol content, and the relative content of N-5 in chars increased with the increased addition of DMHF. Glucose had little effect on the distribution of N-functional groups in chars.


Introduction
After the global oil crisis in the 1970s, clean energy represented by biomass energy has attracted worldwide attention.By 2020, the total installed capacity of global biomass energy has reached 145 GW [1].At present, the annual output of biomass resources such as wood processing residue, straw, algae, livestock manure and domestic waste is about 3.494 billion tons in China [2].If these biomass wastes can be used effectively, a lot of conventional fossil energy will be saved every year, and it will make energy utilization more sustainable.However, some biological resources are rich in nitrogen, which may cause air pollution; these include: crop straw, which is rich in protein, chlorophyll and free amino acids, making its nitrogen content close to that of coal [3]; algae can be used to prepare high calorific value biofuel, but its N content is as up to 10 wt% [4]; livestock manure is also rich in nitrogen and releases a large amount of ammonia gas during composting [5]; domestic waste (such as plastic waste) and wood processing waste contain a large number of synthetic nitrogen-containing compounds such as polyamide [6,7].These nitrogen-rich bioresources, like traditional fossil fuels, have high emissions of NO x and its precursors during thermochemical utilization [8], causing serious environmental problems, such as the greenhouse effect, acid rain and photochemical smog, all of which endanger human health.Therefore, the efficient use of biomass energy needs to be accompanied by a concern for NO x pollution.
As a typical nitrogen-rich biomass, NO x pollution during the thermochemical utilization of waste wood-based panels has received much attention [9].Urea-formaldehyde resin adhesive (UF), with a nitrogen content of up to 35 wt%, is the most popular type of adhesive resin for the production of wood-based panels [10], and the main nitrogen source in the pyrolysis of waste wood-based panels [11,12].The N element in UF transfers into the solid, liquid and gaseous products to different degrees during the pyrolysis of wood-based panels, and the high N content makes nitrogen-rich biochar exhibit good adsorption and electrochemical properties [13,14].Therefore, by studying the migration and conversion of nitrogen during the pyrolysis of wood-based panels, nitrogen can be preferentially fixed in the solid products to prepare nitrogen-rich biochar [15,16]; thus, high-value biochar materials can be prepared, while NO x emissions from the thermal utilization of waste wood-based panels can be reduced.In the thermal conversion of biomass, Chen et al. [17] found that the introduction of cellulose significantly increased the nitrogen content in chars by studying the co-pyrolysis of algae and lignocellulosic biomass.Lai et al. [18] investigated the nitrogen migration in the pyrolysis chars of UF with cellulose, hemicellulose and lignin and drew similar conclusions as Chen et al. [17].To increase the nitrogen content of the chars, Falco et al. [19] pretreated glucose and microalgae by co-hydrothermal conversion, and then pyrolyzed the co-hydrothermal char, concluding that the addition of glucose increased the nitrogen content in pyrolysis chars.Xu et al. [20,21] conducted a series of studies on the co-pyrolysis of glucose and fiberboard, and the results showed that the introduction of glucose significantly increased N retention in chars.Yang et al. [22] further pointed out that when fiberboard was co-pyrolyzed with glucose, the N-rich volatiles decomposed from pyrolysis of fiberboard could be captured by O-rich chars produced from pyrolysis of glucose.In addition, the above research efforts also pointed out that the introduction of cellulose and glucose could promote the conversion of N-A in algae and wood-based panels to more stable N-5 and N-6 [17,19,21,22].At the same time, Torren et al. [23] and Kawamoto et al. [24] found that the pyrolysis of cellulose and glucose both produced various furanic and pyranic compounds.The above research studies provide a feasible basis for nitrogen fixation in pyrolysis chars of nitrogen-rich biomass such as wood-based panels, but rarely involve the directional conversion of nitrogen-functional groups in chars; additionally, the species of pyrolysis intermediates of lignocellulose can affect the conversion of N-functional groups in chars, something that no study has investigated, and the related research may explore a method to regulate nitrogen conversion in thermal conversion of biomass.
In this study, UF and three model compounds (glucose, ethyl maltol and 2,5-dimethyl-4-hydroxy-3(2H)-furanone (DMHF)), which are common intermediates or have similar chemical structures with intermediates (monosaccharide, pyrans and furans) derived from the pyrolysis of lignocellulose, were selected as raw materials for the investigation of the conversion and migration of nitrogen in co-pyrolysis chars through adjusting different ratios of model compounds and UF.This work has focused on the effect of model compounds on the relative content of nitrogen functional groups in chars in order to provide basic research results for the migration and conversion of nitrogen in pyrolysis chars of nitrogen-rich biomass and the regulation of nitrogen structure in chars.

Materials
Glucose (C 6 H 12 O 6 ), ethyl maltol (C 7 H 8 O 3 ) and DMHF (C 6 H 8 O 3 ), which have similar chemical structures, with common lignocellulose derived pyrolysis intermediates, and are rich in oxygen bonds, were selected as model compounds.Among them, glucose is the monomer of cellulose, ethyl maltol is the homologue of maltol with six-membered ring structure, and DMHF evinces a five-membered ring structure [18,[25][26][27]].These three model compounds were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd.(Shanghai, China).
The UF powder was mixed with ultra-pure water at a mass ratio of 1:1, and then the liquid mixture was cured under 100 • C for 18 h.Finally, the cured UF was sieved to particle size less than 80 mesh for use.The prepared UF powder was dried for around 12 h in an oven at 105 • C and then kept in the silica gel dryer.
The elemental analysis result of UF is shown in Table 1.UF was mixed with glucose, ethyl maltol and DMHF, respectively, according to different mass ratios to obtain glucose-UF, ethyl maltol-UF and DMHF-UF samples for this co-pyrolysis experiments.The mass of each sample was 3 g, of which the mass of UF accounted for 10%, 20%, 30%, 40% and 50% of the sample mass.
As shown in Figure 1, the quartz tube reactor (1000 mm × 80 mm) was put in the center of the horizontal tube furnace.Then, the mixed samples were laid onto the alumina crucible and the crucibles were pushed to the center of the quartz tube reactor.During the co-pyrolysis experiments, the temperature in the tube furnace was heated at the rate of 5 • C/min from room temperature to 320 • C through the temperature control system, with the holding time of 60 min, in the heating process; the difference between sample temperature and furnace temperature was less than 2 • C. The whole co-pyrolysis process was controlled under argon (Ar) atmosphere, with gas flow rate of 300 mL/min.The pyrolysis gas was first washed by a flask containing water and then discharged into the atmosphere.

Thermogravimetric Analysis (TGA)
Thermogravimetric analysis was employed to investigate the co-pyrolysis of UF and model compounds.UF was mixed with glucose, ethyl maltol and DMHF (1:1 by mass), respectively.Glucose-UF, ethyl maltol-UF and DMHF-UF samples are denoted as samples 1, 2 and 3, respectively.The thermogravimetric analysis was performed with the TG 209 F3 Tarsus Simultaneous Thermal Analyzer (STA) (NETZSCH Scientific Instruments, Germany).During the thermogravimetric analysis, the samples were heated at the rate of 5

Thermogravimetric Analysis (TGA)
Thermogravimetric analysis was employed to investigate the co-pyrolysis of UF and model compounds.UF was mixed with glucose, ethyl maltol and DMHF (1:1 by mass), respectively.Glucose-UF, ethyl maltol-UF and DMHF-UF samples are denoted as samples 1, 2 and 3, respectively.The thermogravimetric analysis was performed with the TG 209 F3 Tarsus Simultaneous Thermal Analyzer (STA) (NETZSCH Scientific Instruments, Germany).During the thermogravimetric analysis, the samples were heated at the rate of 5 • C/min from 30 • C (initial temperature) to 500 • C (final temperature).The initial sample masses were in range of 10-15 mg, and the purging gas and protective gas were ultra-pure argon.

Characterization of Nitrogen Content and Nitrogen Retention in Chars
The nitrogen content in chars from co-pyrolysis of the model compounds and UF were detected by the Element Analyzer (Vario EL III Elementar, Langenselbold, Germany).The measuring range of N element by this type of element analyzer is from 0.03 mg to 2 mg.Each sample was measured twice, and the nitrogen content of chars were obtained by taking the average value of two measurements.
The nitrogen retention in chars was calculated by the following formula: where N r is the nitrogen retention in char, %; N C is the nitrogen content in char, %; N f is the nitrogen content in raw material, %; M C is the mass of char, g; and M f is the mass of raw material, g.

Characterization of N Functional Groups in Chars
The structure of nitrogen functional groups on the surface of the chars was explored by using a Thermo Scientific K-Alpha X-ray photoelectron spectroscopy analyzer (Thermo Fisher Scientific Technology, Waltham, MA, USA).This XPS analyzer has an X-ray spot size of 50-400 µm and a maximum specimen thickness of 20 mm.The binding energy of the extraneous carbon (284.8 eV) and the peak binding energy of the C1s XPS spectra were used as a reference for the electric charge correction of the N1s XPS spectra, and then the corrected data were imported into XPSPEAK for peak resolving and fitting.During the treatment, the Gauss-Lorentz parameter was fixed at 20%, and the binding energies of N functional groups ranged from 398.5 ± 0.3 eV (N-6), 399.8 ± 0.3 eV (N-A), and 400.5 ± 0.3 eV (N-5) [28].
The relative content of the N functional groups in chars were calculated by the percentage of a single peak area in the total peak areas: where RC N* is the relative content of N functional group, %; A N* is the single peak area of N functional group; ΣA N* is the total peak area.

The Mass Loss Behavior of Model Compounds and UF Studied by Thermogravimetric Analysis
The TG and DTG curves of glucose, ethyl maltol, DMHF and UF are shown in Figure 2.  The decomposition process of UF can be divided into three stages: drying preheating, rapid decomposition and high temperature decomposition, which is consistent with previous results of decomposition characteristics of UF obtained by Lai et al. [18] and Chen et al. [29].In the first stage, the temperature increased from the room temperature to 186 °C, which was when the drying of water and the release of free formaldehyde mainly occurred [30].The mass loss of UF was 5.51%, which was slightly different from the study of Feng et al. [30].This may be due to the fact that the moisture absorption capacity of UF was strong and moisture was absorbed in the preservation period.The decomposition of the second stage was the most violent, with a maximum mass loss temperature of 269 ℃.In this temperature range, the C-N bond at the end of the UF broke, producing N2, NH3 and other gases [31].At the same time, it can be observed that the DTG curve of UF had an obvious shoulder peak at about 246 °C, which was consistent with the results of Chen The decomposition process of UF can be divided into three stages: drying preheating, rapid decomposition and high temperature decomposition, which is consistent with previous results of decomposition characteristics of UF obtained by Lai et al. [18] and Chen et al. [29].In the first stage, the temperature increased from the room temperature to 186 • C, which was when the drying of water and the release of free formaldehyde mainly occurred [30].The mass loss of UF was 5.51%, which was slightly different from the study of Feng et al. [30].This may be due to the fact that the moisture absorption capacity of UF was strong and moisture was absorbed in the preservation period.The decomposition of the second stage was the most violent, with a maximum mass loss temperature of 269 • C. In this temperature range, the C-N bond at the end of the UF broke, producing N 2 , NH 3 and other gases [31].At the same time, it can be observed that the DTG curve of UF had an obvious shoulder peak at about 246 • C, which was consistent with the results of Chen et al. [29].As for the third stage, since most of the chemical bonds in UF were broken before this stage, the structure of the remaining substances was relatively stable.
The TG curve of glucose indicates that glucose was basically not decomposed before 191 • C, and the mass loss was only 0.80%.The first stage of glucose decomposition occurred at 191-270 • C, and glucose began to decompose dramatically from 191 • C, which was consistent with the results of Xu et al. [20].Up to 270 • C, the decomposition of glucose in the first stage was basically completed, with a mass reduction of about 22%.This was because the initial decomposition of glucose was accompanied by the formation of a molten liquid membrane, which hindered the release of volatile matter and then led to a low mass loss [32].Then, the second stage of glucose decomposition occurred, with the maximum mass loss temperature occurring at 299 • C, and most of the decomposition was completed before 399 • C. At this stage, the release of volatiles causing the mass loss of glucose was as high as 57% [32].In the third stage, the pyrolysis rate of glucose became slow, and the TG curve tended to be smooth.
It can also be observed that the TG and DTG curves of ethyl maltol and DMHF almost overlap, and their mass loss properties were very similar.Before 90 • C, the mass loss of ethyl maltol and DMHF was about 1%, and presumably, the removal of water stabilizer occurred in this range.With the temperature higher than 90 • C, ethyl maltol and DMHF underwent dramatic mass loss until 168 • C; the mass loss of both reached about 98%, and the maximum mass loss temperature was 164 • C and 161 • C, respectively.The melting points and boiling points of ethyl maltol and DMHF are 85-95 • C, 290.3 ± 40.0 • C at 760 mm Hg and 73-77 • C, 215.5 ± 40.0 • C at 760 mm Hg, respectively.TG results showed that the mass loss process of ethyl maltol and furanone was concentrated in the temperature range of 90-168 • C which is below their boiling points; it is believed the sharp and complete weight losses were due to the vaporization of these two compounds.Further, pyrolysis experiments of ethyl maltol and furanone were performed in tube furnace (5 • C/min from room temperature to 320 • C and 1 h holding time) respectively, and the results indicated that tiny char were formed; meanwhile, some white crystals condensed at the outlet of the quartz tube reactor.The results of elemental analyses of the white crystals are shown in Table 2, which shows that the contents of C, H and O in the white crystals are almost the same with that of calculated from the original ethyl maltol and furanone.This indicated that ethyl maltol and furanone undergo vaporization at lower temperatures in the heating process.In order to explore the interaction between model compounds and UF, the TG and DTG curves of samples were drawn in Figure 3, and recorded as TG-experiment (TG-E) and DTGexperiment (DTG-E).At the same time, in order to facilitate the comparison, the average values of TG and DTG curves of individual pyrolysis of model compounds and UF were calculated, which were recorded as TG-calculation (TG-C) and DTG-calculation (DTG-C).As shown in Figure 3a, TG-E and TG-C of glucose-UF were obviously different.Before 310 ℃, the mass loss of glucose-UF co-pyrolysis was significantly higher than the average value of glucose and UF alone, and Xu et al. [20] also pointed out that the copyrolysis of fiberboard containing UF with glucose significantly accelerated the mass loss, indicating that co-pyrolysis of mixed samples containing with UF needed lower activation energy than pyrolysis of fiberboard and glucose [33], which made the sample easier to lose weight by pyrolysis.After 310 °C, the mass loss of glucose-UF co-pyrolysis was lower than the calculated value, which may increase the yield of co-pyrolysis chars.Comparing these two DTG curves, only two obvious peaks can be observed in DTG-E, while four peaks can be observed in DTG-C.The first peak of glucose-UF co-pyrolysis mainly came from the pyrolysis of glucose in the sample.Relatively speaking, the first peak of co-pyrolysis was about 25 °C earlier than the calculated peak, and the mass loss of the sample at this stage was also higher than the calculated value, suggesting that UF can significantly promote the pyrolysis of glucose due to the fact that it can catalyze the dehydration reaction and the formation reaction of hydroxyl and carbonyl groups [34].
Considering the co-pyrolysis of Ethyl maltol-UF and DMHF-UF, the co-pyrolysis mass loss at the initial stage of both Ethyl maltol-UF and DMHF-UF were obviously ahead of the calculated value derived from pyrolysis of ethyl maltol, DMHF and UF.DTG results showed that the mass loss peaks at the initial stage of DTG-E of Ethyl maltol-UF and DMHF-UF moved forward by 30 °C and 23 °C more than that of DTG-C, respectively; this indicated interaction affect occurred in co-pyrolysis and the addition of UF might further promote the decomposition of ethyl maltol and DMHF.With elevating pyrolysis temperature, it seemed the interaction in co-pyrolysis was mild for the TG and DTG since the curves of the experiment changed with same trend as those from calculation; meanwhile, As shown in Figure 3a, TG-E and TG-C of glucose-UF were obviously different.Before 310 • C, the mass loss of glucose-UF co-pyrolysis was significantly higher than the average value of glucose and UF alone, and Xu et al. [20] also pointed out that the co-pyrolysis of fiberboard containing UF with glucose significantly accelerated the mass loss, indicating that co-pyrolysis of mixed samples containing with UF needed lower activation energy than pyrolysis of fiberboard and glucose [33], which made the sample easier to lose weight by pyrolysis.After 310 • C, the mass loss of glucose-UF co-pyrolysis was lower than the calculated value, which may increase the yield of co-pyrolysis chars.Comparing these two DTG curves, only two obvious peaks can be observed in DTG-E, while four peaks can be observed in DTG-C.The first peak of glucose-UF co-pyrolysis mainly came from the pyrolysis of glucose in the sample.Relatively speaking, the first peak of co-pyrolysis was about 25 • C earlier than the calculated peak, and the mass loss of the sample at this stage was also higher than the calculated value, suggesting that UF can significantly promote the pyrolysis of glucose due to the fact that it can catalyze the dehydration reaction and the formation reaction of hydroxyl and carbonyl groups [34].
Considering the co-pyrolysis of Ethyl maltol-UF and DMHF-UF, the co-pyrolysis mass loss at the initial stage of both Ethyl maltol-UF and DMHF-UF were obviously ahead of the calculated value derived from pyrolysis of ethyl maltol, DMHF and UF.DTG results showed that the mass loss peaks at the initial stage of DTG-E of Ethyl maltol-UF and DMHF-UF moved forward by 30 • C and 23 • C more than that of DTG-C, respectively; this indicated interaction affect occurred in co-pyrolysis and the addition of UF might further promote the decomposition of ethyl maltol and DMHF.With elevating pyrolysis temperature, it seemed the interaction in co-pyrolysis was mild for the TG and DTG since the curves of the experiment changed with same trend as those from calculation; meanwhile, the residual mass of experiment value and calculation value showed little difference, considering ethyl maltol and DMHF decomposed completely at low temperature as shown in Figure 2a; it assumed the residual chars were mainly formed from the pyrolysis of UF in co-pyrolysis of Ethyl maltol-UF and DMHF-UF.

Effect of Model Compounds: UF Ratio on Nitrogen Migration in Co-Pyrolysis Chars
The nitrogen retention in chars was calculated from the char yields and the nitrogen content in chars.The curves of nitrogen content, char yields and nitrogen retention with UF content were plotted in Figure 4.Meanwhile, the nitrogen content, char yield and nitrogen retention for individual pyrolysis of UF (3 g) were marked as horizontal lines in Figure 4.When 50% UF was added into the mixture, the char yield of glucose-UF co-pyrolysis was significantly higher than that of UF alone, and the char yield increased by about 2.7%, with the UF content decreased from 50% to 10%.It can be inferred that the interaction of glucose and UF facilitated the formation of chars, which confirmed the discovery in 3.1.2,and that the addition of glucose can promote the conversion of N from volatile-N to char-N due to higher char yield [21].To the contrary, during the co-pyrolysis of ethyl maltol and DMHF with UF, the char yields decreased almost linearly with the decrease of UF content; however, the char yields improved significantly compared with co-pyrolysis in the thermal analyzer as shown in Figure 3b,c, which may due to much more raw materials used in tube furnace pyrolysis (3 g) than in thermogravimetry experiment (10-15 mg) and the dramatic increase of raw materials may suppress the release of volatiles and thus benefit for the formation of chars [35].The improvements of char yields indicate that a part of ethyl maltol and DMHF take part in the formation of chars in co-pyrolysis, which is helpful for fixing more nitrogen in chars.
The nitrogen content in glucose-UF, ethyl maltol-UF and DMHF-UF co-pyrolysis chars decreased linearly with the decrease of UF content in the mixtures, as shown in Figure 4a, which due to nitrogen source reduced with decreasing of UF addition.Additionally, glucose-UF derived chars showed the lowest nitrogen content, which due to the addition of glucose, increased the yield of char, while ethyl maltol-UF chars showed higher nitrogen content due to lower char yields.When considering the nitrogen retention in chars, as shown in Figure 4c, the nitrogen retention of co-pyrolysis chars of these three model compounds with UF were all increased to different degrees compared to UF pyrolysis alone.In co-pyrolysis, a part of nitrogen was fixed by char that derived from the py- When 50% UF was added into the mixture, the char yield of glucose-UF co-pyrolysis was significantly higher than that of UF alone, and the char yield increased by about 2.7%, with the UF content decreased from 50% to 10%.It can be inferred that the interaction of glucose and UF facilitated the formation of chars, which confirmed the discovery in Section 3.1.2,and that the addition of glucose can promote the conversion of N from volatile-N to char-N due to higher char yield [21].To the contrary, during the co-pyrolysis of ethyl maltol and DMHF with UF, the char yields decreased almost linearly with the decrease of UF content; however, the char yields improved significantly compared with co-pyrolysis in the thermal analyzer as shown in Figure 3b,c, which may due to much more raw materials used in tube furnace pyrolysis (3 g) than in thermogravimetry experiment (10-15 mg) and the dramatic increase of raw materials may suppress the release of volatiles and thus benefit for the formation of chars [35].The improvements of char yields indicate that a part of ethyl maltol and DMHF take part in the formation of chars in co-pyrolysis, which is helpful for fixing more nitrogen in chars.
The nitrogen content in glucose-UF, ethyl maltol-UF and DMHF-UF co-pyrolysis chars decreased linearly with the decrease of UF content in the mixtures, as shown in Figure 4a, which due to nitrogen source reduced with decreasing of UF addition.Additionally, glucose-UF derived chars showed the lowest nitrogen content, which due to the addition of glucose, increased the yield of char, while ethyl maltol-UF chars showed higher nitrogen content due to lower char yields.When considering the nitrogen retention in chars, as shown in Figure 4c, the nitrogen retention of co-pyrolysis chars of these three model compounds with UF were all increased to different degrees compared to UF pyrolysis alone.In co-pyrolysis, a part of nitrogen was fixed by char that derived from the pyrolysis of UF, while the other part was fixed by chars derived from glucose, ethyl maltol and DMHF in co-pyrolysis.The second part of nitrogen was fixed in the char mainly due to the interaction of nitrogen-containing volatiles decomposed from UF and oxygen-rich char matrix derived from glucose, ethyl maltol, DMHF [15,22,36].The three model compounds and their decomposition products were rich in carboxyl and carbonyl groups, which increased the nitrogen retention in chars through Maillard reaction with N-containing compounds [15,19,21].The nitrogen fixation ability of ethyl maltol was relatively weak, and the nitrogen fixation rate did not increase significantly with the increase of ethyl maltol, which was due to low char yield in co-pyrolysis of ethyl maltol and UF.The increase of glucose and DMHF improved the nitrogen retention dramatically, as the content of the glucose and the DMHF increased from 50% to 90%, as the nitrogen retentions of glucose-UF and DMHF-UF co-pyrolysis chars increased by 28.5% and 16.5%, respectively.In particular, the nitrogen retention of glucose-UF co-pyrolysis chars tended to increase more significantly with the decrease of UF content from 50% to 10%, which may be due to the fact that glucose was easily converted to the molten state in pyrolysis; the molten glucose can clad UF matrix and thus hinder the release of volatile-N [20,21].

Effect of Model Compounds: UF Ratio on the Conversion of Nitrogen Functional Groups in Co-Pyrolysis Chars
XPS was used to characterize the N-functional groups in chars to investigate the effect of the addition of ethyl maltol, DMHF and glucose upon the nitrogen conversion of co-pyrolysis chars (N1s spectra of chars are shown in Figures S1-S4).Three types of N functional groups can be detected on the surface of the co-pyrolysis chars, namely amide-N (N-A), pyridine-N (N-6) and pyrrole-N (N-5) [6,9,16], while N-6 and N-5 were converted from N-A during co-pyrolysis [30].From the N1s XPS spectra of co-pyrolysis chars, it was clear that N-A always occupied a large proportion; a previous study indicated the complete decomposition of N-A needing a pyrolysis temperature beyond 400 • C [20].
As shown in Figure 5a, the relative content of N-6 in the co-pyrolysis chars increased by 11.86%, with the content of ethyl maltol in the sample increased from 50% to 90%, showing an obviously increasing trend, while the relative content of N-A decreased from 45.68% to 35.81%, showing a stable decreasing trend, and the relative content of N-5 in chars basically remained stable.Similarly, as shown in Figure 5b, the relative content of N-5 in co-pyrolysis chars increased significantly with the increase of DMHF content in the mixtures.With the increase of DMHF content from 50% to 90%, the relative content of N-5 in chars increased by 15.05%, while the relative content of N-A decreased by 10.31%.The relative content of N-6 in glucose-UF co-pyrolysis chars was slightly higher than that of N-5 when the content of glucose was 50%.With the increase of glucose content, the relative content of N-6 tended to decrease, while the relative content of N-5 increased slowly.The relative content of N-6 was higher than that of N-6 until the glucose content was 90%.Generally speaking, the relative content of nitrogen-functional groups in glucose-UF co-pyrolysis chars remained relatively stable.
5 in chars increased by 15.05%, while the relative content of N-A decreased by 10.31%.The relative content of N-6 in glucose-UF co-pyrolysis chars was slightly higher than that of N-5 when the content of glucose was 50%.With the increase of glucose content, the relative content of N-6 tended to decrease, while the relative content of N-5 increased slowly.The relative content of N-6 was higher than that of N-6 until the glucose content was 90%.Generally speaking, the relative content of nitrogen-functional groups in glucose-UF co-pyrolysis chars remained relatively stable.Based on the results from the nitrogen functional groups' conversion in co-pyrolysis chars, it can be concluded that abundant oxygen-containing radicals from glucose, ethyl maltol and DMHF can react with the nitrogen-rich species from UF through a Maillard reaction and prompt the conversion from N-A to heterocyclic N during co-pyrolysis.Additionally, the addition of ethyl maltol (six-membered structure) and DMHF (five-membered structure) led to more N-6 and N-5 being converted, respectively, which indicated that the structure of oxygen-containing heterocyclic compounds formed during pyrolysis of lignocellulose may affect the conversion of nitrogen in chars when nitrogen is involved in the pyrolysis.In conclusion, the co-pyrolysis between the model compounds of Based on the results from the nitrogen functional groups' conversion in co-pyrolysis chars, it can be concluded that abundant oxygen-containing radicals from glucose, ethyl maltol and DMHF can react with the nitrogen-rich species from UF through a Maillard reaction and prompt the conversion from N-A to heterocyclic N during co-pyrolysis.Additionally, the addition of ethyl maltol (six-membered structure) and DMHF (five-membered structure) led to more N-6 and N-5 being converted, respectively, which indicated that the structure of oxygen-containing heterocyclic compounds formed during pyrolysis of lignocellulose may affect the conversion of nitrogen in chars when nitrogen is involved in the pyrolysis.In conclusion, the co-pyrolysis between the model compounds of lignocellulose thermal conversion products and UF not only improved the nitrogen retention in chars, but also changed the distribution of nitrogen-functional groups.Furans and pyrans may be effective additives to regulate the distribution of nitrogen-functional groups in chars.

Conclusions
This study has investigated N retention and migration in chars prepared from copyrolysis of UF with three model compounds (ethyl maltol, DMHF and glucose).Based on the discussion above, the following conclusions can be made.
(1) Thermogravimetric analysis indicated the mass loss was accelerated at the initial stage during co-pyrolysis of all three model compounds with UF.
(2) The addition of the three model compounds improved the nitrogen retention rate of co-pyrolysis chars, which due to abundant oxygen-containing groups sourced from model compounds, enhanced the Maillard reactions; meanwhile, co-pyrolysis could improve char yield, especially for the addition of glucose which also would be of benefit in fixing more nitrogen in char.
(3) In co-pyrolysis, the addition of ethyl maltol (six-membered structure) and DMHF (five-membered structure) led to more N-6 and N-5, respectively, being converted, indicating that the structure of oxygen-containing heterocycle affected the formation of the nitrogen-functional group in chars during nitrogen-involved lignocellulose pyrolysis.
Analysis of Decomposition Characteristics 3.1.1.The Mass Loss Behavior of Model Compounds and UF Studied by Thermogravimetric Analysis The TG and DTG curves of glucose, ethyl maltol, DMHF and UF are shown in Figure 2.

Table 1 .
Elemental analysis of UF.

Table 2 .
The content of C, H, O in ethyl maltol, DMHF (calculation) and white crystals (measuring).