Analysis of the Pyrolytic Behaviour of Birch, Maple, and Rowan Leaves

energies

Article Analysis of the Pyrolytic Behaviour of Birch, Maple, and Rowan Leaves

Valentina Zubkova * , Andrzej Strojwas and Marcin Bielecki

The Institute of Chemistry, Jan Kochanowski University, Uniwersytecka Str. 7, 25-369 Kielce, Poland; [email protected] (A.S.); [email protected] (M.B.) * Correspondence: [email protected] or [email protected]

Abstract: A research study was conducted on the thermal behaviour of leaves of urban greenery (birch, maple, and rowan) and the products of their pyrolysis and extraction as assisted by mi- crowaves. The obtained products of pyrolysis and extraction were investigated with the use of FT-IR and UV spectroscopies and XRD techniques. A contractive analysis of samples of chars, condensates, after-extraction residue, and extracts showed that the changes in structural-chemical parameters of leaves of different types of trees during pyrolysis and extraction take place in distinct ways. About 22% of material was removed from birch leaves during extraction, and more than 17% of material was extracted from maple and rowan leaves. It was determined that, during pyrolysis of after-extraction residue of leaves, many fewer PAH compounds with carbonyl groups along with alcohols and phenols are emitted than during pyrolysis of non-extracted leaves. Taking into account that pyrolysis is the first stage of combustion, a decrease in the amount of dangerous compounds in the volatile products of pyrolysis leads to a lower contribution of such compounds in combustion products. This indicates that leaves of urban greenery can be subjected to combustion after extraction, and the obtained extracts can be used as a source of phytochemicals and chemical reagents. Keywords: leaves; pyrolysis; extraction; TG/FT-IR; XRD; UV-spectroscopy Citation: Zubkova, V.; Strojwas, A.; Bielecki, M. Analysis of the Pyrolytic Behaviour of Birch, Maple, and Rowan Leaves. Energies 2021, 14, 2091. 1. Introduction https://doi.org/10.3390/en14082091 Tree leaves falling in autumn are considered urban green waste. The green waste is not classified as hazardous waste. Nonetheless, their accumulation in the place of origin is Academic Editor: Biagio Morrone harmful to the environment and requires special disposal. Tree leaves have diverse and complex composition. Phenol, flavonoid and flavan-3-ol, 2,2-diphenyl-1-picrylhydrazyl, Received: 25 February 2021 and 2,2-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) have been identified in their Accepted: 6 April 2021 Published: 9 April 2021 composition [1]. Plant leaves can be used to obtain phytochemicals (polyphenols and triterpenoids) [2] and bioactive antioxidant compounds such as oleuropein and luteolin.

Publisher’s Note: MDPI stays neutral Eucalyptol, a chemical compound present in eucalyptus leaves, was found in bio-oils from with regard to jurisdictional claims in leaves in high concentration [3]. The leaves of Polyalthia longifolia [4] and olive trees [3] can published maps and institutional afﬁl- be successfully applied to produce bio-oils. iations. Other applications of leaves are widely known. Mango leaves can be used for production of quantum dots [5] and biochars from orchid tree leaves can be utilized as “fuel of a direct carbon solid oxide fuel cell” [6]. Attempts have also been made to include tree leaves in the composition of forage for mature male sheep [7]. Moreover, biochars from banana tree leaves can be used for the production of construction insulating material [8]. Celtuce Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. leaves are used for the production of porous coal with “outstanding supercapacitance and This article is an open access article CO2 capture performance” [9], and activated ginkgo-leaf-derived carbons were analysed distributed under the terms and as a source of porous carbons made of cheap biomass that can be applied as supercapacitor conditions of the Creative Commons electrode materials in the acid electrolyte [10]. Carbon materials from bamboo leaves in Attribution (CC BY) license (https:// combination with copper oxide/cuprous oxide nanoparticles can serve as a supercapaci- creativecommons.org/licenses/by/ tor [11], and batat leaves can be used to obtain three-dimensionally interconnected carbon 4.0/). nanorings [12].

Energies 2021, 14, 2091. https://doi.org/10.3390/en14082091 https://www.mdpi.com/journal/energies Energies 2021, 14, 2091 2 of 18

Composting is the most widespread and known way of utilizing the leaves of garden trees [13,14]. However, the leaves of urban trees contain large amounts of heavy metals because of their close proximity to roads [15–18] and therefore cannot be used for soil fertilization. Limitation of the use of fossil fuels makes us pay attention to tree leaves as a lignocellulosic biomass that can be applied for energy purposes [19]. According to Makkawi et al., the leaves of Phoenix dactylifera along with other wastes arising from its cultivation can be converted into energy without any harm to the environment or costs for disposal of these wastes [20]. Nevertheless, it is believed that forest waste has a much lower parameter of high heating value (HHV) compared to agricultural waste [21]. Despite this, tree leaves can be used as components of pellets applied in commercial devices [22]. According to He et al. [3], eucalyptus leaves can be used for energy purposes. It is characteristic of leaves that they behave differently from the wood of the same species: “leaves tend to give a higher yield of char and a lower yield of organics in bio-oil than the wood” [3]. A serious problem arising during combustion is the formation of ash and the emission of suspended dust along with volatile organic compounds, including terpenoids [23–25]. An analysis of the composition of suspended dust proved that coal atoms prevail in its elemental composition. According to transmission electron microscope (TEM) investiga- tions, the composition of coal atoms corresponds to the presence of fringes formed due to pyrolysis and the combustion of compounds containing aromatic hydrocarbons [26,27]. This gives reasons to suggest that hydrocarbons in the composition of volatile pyrolysis products (as the ﬁrst stage of combustion) along with inorganic particles lead to the formation of suspended dust. In order to verify the aforementioned statements and to propose a more economical way to dispose of leaves from selected tree types of urban greenery, the following aims were set: (i) to investigate the course of the pyrolysis process of birch, maple, and rowan tree leaves using a pyrolytic oven manufactured by the Czylok company and a thermogravimetric Fourier transform-infrared (TG/FT-IR) measuring unit; (ii) to compare structural-chemical parameters of the pyrolysis products of birch, maple, and rowan samples; (iii) to determine the extraction yield of birch, maple, and rowan samples and to investigate the obtained products of extraction using the techniques of Fourier transform- infrared (FT-IR) and ultraviolet (UV) spectroscopies and X-ray diffraction (XRD).

2. Materials and Methods The objects of the research were birch leaves (BL), maple leaves (ML), and rowan leaves (RL). These tree species are among the most common ones in Polish urban greenery. The washed leaves were dried at a temperature of 105 ◦C and ground to a diameter of particles <3 mm. The elemental analysis of samples was carried out in a CHNS Elementar Vario Micro Cube analyzer. The results of elemental analysis and the values of HHV parameters of the studied samples are presented in Table1.

Table 1. The main characteristics of leaves.

Main Characteristics BL ML RL C [%] 49.39 ± 0.21 41.99 ± 1.02 48.55 ± 0.04 H [%] 6.97 ± 0.05 5.71 ± 0.03 6.69 ± 0.05 N [%] 0.78 ± 0.71 1.46 ± 0.03 1.33 ± 0.07 S [%] 0.06 ± 0.05 a 0.07 ± 0.01 O a [%] 34.42 ± 0.22 34.49 ± 0.28 33.05 ± 0.04 A [%] 8.38 ± 0.06 16.35 ± 0.02 10.32 ± 0.02 HHV b [MJ/Kg] 21.72 ± 0.22 17.46 ± 0.28 21.19 ± 0.04 (a) Calculated by difference, O [%] = 100 − C − H − N − S − Ash; (b) Calculated by HHV [MJ kg−1] = 0.3491 * C + 1.1783 * H + 0.1005 * S − 0.0151 * N − 0.1034 * O − 0.0211*Ash; a—absent; BL—birch leaves; ML—maple leaves; RL—rowan leaves. Energies 2021, 14, 2091 3 of 18

Table2 presents the results of the determination of the content of inorganic elements in the samples. This determination was made using elemental analysis with the energy dispersive X-ray ﬂuorescence (ED-XRF) technique, using a Niton XL3T Goldd+ analyzer manufactured by Thermo Fisher Scientiﬁc Inc (Waltham, MA, USA).

Table 2. Content of inorganic elements.

Elements BL ML RL Ca [%] 2.39 ± 0.02 3.06 ± 0.02 3.92 ± 0.03 K [%] 3.30 ± 0.03 1.64 ± 0.02 1.84 ± 0.02 Fe [mg/kg] 125 ± 18 231 ± 21 579 ± 28 P [mg/kg] 603 ± 115 2999 ± 118 1152 ± 133 S [mg/kg] 2475 ± 94 1707 ± 71 3563 ± 106 Si [mg/kg] 2129 ± 133 2224 ± 114 11,577 ± 227 Sr [mg/kg] 38 ± 1 78 ± 1 99 ± 2 Zn [mg/kg] 98 ± 6 12 ± 3 27 ± 4

It follows from Table2 that the greatest amount of determined inorganic elements (apart from K atoms) was observed in the RL sample. The biomass samples were pyrolyzed in a PRC 70 × 708/110 M pipe oven manufactured by the Czylok company under the atmosphere of nitrogen of 99.999% purity. The heating rate in the oven was 10 ◦C/min. The samples were heated to a temperature of 450 ◦C. The volatile pyrolysis products were removed to a retort with methanol cooled by water with ice to obtain a condensate, i.e., the volatile products of pyrolysis absorbed by methanol. Dried samples weighing 2 g were extracted with a mixture of chloroform and methanol (25:25 mL) in an Anton Paar multiwave 3000 microwave extractor at a temperature of 150 ◦C under the pressure of 18 bars for 20 min. After the solvent evaporated, the samples of extracts and after-extraction residues (samples RBL, RML, and RRL) were subjected to further analysis. The samples of initial biomass (BL, ML, RL) and after-extraction residues (RBL, RML, RRL) were pyrolyzed in a TG/FT-IR unit under the atmosphere of nitrogen of high purity up to a temperature of 780 ◦C in order to determine the composition of volatile products of pyrolysis being formed. The volatile products were directed via interface to a Nicolet iS10 spectrometer, and their FT-IR spectra were registered in the range of a wavenumber of 4000–600 1/cm. The FT-IR spectra were elaborated and the deconvolution of differential thermogravimetric (DTG) curves was performed using OMNIC9 software. During deconvolution, a fragment of the DTG curve was analysed in the range of 20–600 ◦C. The samples of biomass, chars, condensates, and extracts were studied by the attenuated total reﬂectance (ATR) technique using a Thermo Fischer Scientiﬁc Nicolet iS10 spectrometer. For this purpose, a Smart MIRacle module with a ZnSe monocrystal was used. The automatic correction of baseline was performed by OMNIC9 software, taking into account the local optical minima near 2400, 2000 1/cm. The UV spectra of the samples of condensates and extracts dissolved in acetonitrile were registered by a Jasco UV-Vis V-630 spectrophotometer in the range of wavelength of 195–350 nm. For this purpose, 0.0001 g was taken from the samples and dissolved in 50 mL of acetonitrile. The dissolved samples were investigated in quartz cuvettes. The obtained spectra were normalized and deconvoluted using Jasco Spectra Manager software. The biochars were heated in a pyrolytic oven at a temperature of 450 ◦C and investigated with the use of X-ray diffraction. Using Cukα radiation, the powder diffractograms were recorded. The X-ray lamp mode was U = 40 kV, I = 25 mA, and the exposure time at every spot was 20 s. Energies 2021, 14, x FOR PEER REVIEW 4 of 18

The biochars were heated in a pyrolytic oven at a temperature of 450 °C and investi- Energies 2021, 14, 2091gated with the use of X-ray diffraction. Using Cukα radiation, the powder diffractograms 4 of 18 were recorded. The X-ray lamp mode was U = 40 kV, I = 25 mA, and the exposure time at every spot was 20 s.

3. Results and3. Discussion Results and Discussion 3.1. Analysis of Pyrolysis Products 3.1. Analysis of Pyrolysis Products Table3 presents the yields of char and condensate obtained during pyrolysis of the Table 3 presents the yields of char and condensate obtained during pyrolysis of the samples of leaves. samples of leaves.

Table 3. Table 3. Yield of chars andYield condensates of chars and obtained condensates during obtained pyrolysis during of samples. pyrolysis of samples.

Sample Sample Yield of Char [%] Yield of Char Yield [%] of Condensate Yield of [%] Condensate [%] BL BL33.8 33.84.1 4.1 ML ML41.8 41.8 1.9 1.9 RL 35.1 3.1 RL 35.1 3.1

◦ It follows fromIt Table follows 3 that from at Table a temperat3 that ature a temperatureof 450 °C, the of ML 450 sampleC, the had ML the sample great- had the greatest est char yield, charand the yield, BL andsample the BLhad sample the lowest had theone. lowest one. Figure 1 presentsFigure the1 ATRpresents spectra the ATR of the spectra dried of samples the dried of samples leaves. ofThe leaves. spectra The coin- spectra coincided cided in such ain way such that a waythe height that the of the height band of near the band1600 1/cm near was 1600 the 1/cm same. was This the height same. This height corresponds tocorresponds the presence to of the bonds presence of the of C=C bonds type of in the the C=C material type in of the samples. material of samples.

C-O stretching in cellulose H-bonds and lignin

Cal-H

C=O C=C

Car-H BL ML RL

3500 3000 2500 2000 1500 1000 Wavenumber (1/cm) Figure 1. ATR spectra of examinedFigure 1.leaves.ATR spectra of examined leaves.

This enables aThis visual enables estimation a visual of estimationthe contribution of the contributionof particular ofbonds particular and groups bonds and groups of of atoms with atomsrespect with to the respect content to the of C=C content bonds of C=C in the bonds studied in the samples. studied The samples. shape Theof shape of the the curves in Figurecurves 1 in implies Figure 1that implies the sample that thes varied samples from varied each from other each by othera contribution by a contribution ratio ratio of hydrogenof hydrogen bonds, bonds bonds, of bonds the Cal of-H the type, Cal-H carbonyl type, carbonyl groups, groups,and bonds and of bonds the C ofar- the Car-H type. H type. However,However, the most the mostnoticeable noticeable difference difference is theis contribution the contribution of C-O of stretching C-O stretching in in cellulose cellulose and ligninand lignin in the in samples. the samples. For comparison,For Figure comparison, 2 presents Figure the2 ATR presents spectra the of ATR initial spectra samples of initial of leaves samples as well of leaves as well as their condensatesas their and condensates chars. The and ATR chars. spectra The of ATR the samples spectra of (except the samples the BL (exceptcondensate the BL condensate sample) were normalizedsample) were with normalized respect to with the height respect of to the the band height near of the1600 band 1/cm. near 1600 1/cm. In the ATR spectraIn the of ATRall samples, spectra there of all we samples,re some wide there bands were somein the range wide bandsof wave- in the range of numbers of 3700–2400wavenumbers 1/cm. These of 3700–2400 bands point 1/cm. to Thesethe presence bands of point −OH to groups the presence that form of −OH groups a range of hydrogenthat form bonds a range(OH π of, self-associated hydrogen bonds OH, (OH OH OR,··· π, self-associatedtightly bound cyclic OH, OH- OH··· OR, tightly ... tetramers, OH N,bound carboxylic cyclic OH-tetramers, acids dimmers, OH SH···...N,N) carboxylic [28,29] in the acids studied dimmers, samples SH N) and [28 to,29 ] in the stud- the bonds of theied C samplesal-H type and (range to the of bonds wavenumbers of the Cal -Hof 2950–2850 type (range 1/cm). of wavenumbers The ATR spectra of 2950–2850 1/cm). suggest that >C=OThe ATRgroups spectra (range suggest of wavenumbers that >C=O groupsof 1800–1650 (range 1/cm) of wavenumbers are present in of the1800–1650 1/cm) initial samplesare and present condensates, in the initialand C=C samples groups and (near condensates, 1600 1/cm) and are present C=C groups in the (near initial1600 1/cm) are samples, condensatespresent (except in the initial the BL samples, sample), condensates and chars. In (except the dactyloscopic the BL sample), range, and in chars. the In the dactyloscopic range, in the ATR spectra there are visible one to eight bands that are characteristic of groups occurring in cellulose, hemicellulose, and lignin [30–32]. The assignment of the bands is explained in the caption for Figure2. Near the wavenumber of 1030 1/cm, there

is an intensive band of β-1,4-glycoside bond (marked as 9 in the ﬁgure). The presence of Energies 2021, 14, x FOR PEER REVIEW 5 of 18

Energies 2021, 14, 2091 5 of 18

ATR spectra there are visible one to eight bands that are characteristic of groups occurring in cellulose, hemicellulose, and lignin [30–32]. The assignment of the bands is explained in the caption aforementioned for Figure 2. Near bands the shows wavenumber that the of 1030 studied 1/cm, samplesthere is an ofintensive leaves band should be rated as of lignocelluloseβ-1,4-glycoside bond biomass (marked [31– as33 9]. in The the spectrafigure). The in Figure presence2 suggest of the aforementioned that the height of band nine, bandscorresponding shows that the to studied the presence samples of of leaves C-O should stretching be rated (in as cellulose lignocellulose and biomass lignin) in the samples, [31–33].decreased The spectra during in Figure the pyrolysis 2 suggest process.that the height This of implies band nine, that corresponding this β-1,4-glycoside to the bond under- presence of C-O stretching (in cellulose and lignin) in the samples, decreased during the went degradation: the height of band nine decreased in char of the BL sample by ﬁve times, pyrolysis process. This implies that this β-1,4-glycoside bond underwent degradation: the heightin char of band of the nine ML decreased sample in bychar three of the times, BL sample and inby charfive times, of the in RL char sample of the ML by 2.5 times. As a sampleresult by ofthree pyrolysis, times, and the in char contribution of the RL sample ratio ofby the2.5 times. bonds As of a result the C ofal-H pyrolysis, type decreased in all thechars. contribution Moreover, ratio of the the bands bonds correspondingof the Cal-H type todecreased the presence in all chars. of carbonyl Moreover, groups the disappeared. bandsThe corresponding contribution to ratio the presence of hydrogen of carbonyl bonds groups decreased disappeared. in the The chars contribution of ML and RL. In all ratiochars of hydrogen of the studied bonds decreased samples in in the the chars range of ML of 1450–1350and RL. In all 1/cm, chars thereof the appearedstudied a wide band samplesthree, in which the range can of be 1450–1350 attributed 1/cm, to there aromatic appeared skeletal a wide vibration band three, combined which can withbe CH in plane attributed to aromatic skeletal vibration combined with CH in plane deformation for lig- nindeformation and cellulose along for lignin with CH and deformation cellulose alongin fragments with CHof cellulose deformation and hemicellulose in fragments of cellulose [32].and hemicellulose [32].

Figure 2. ATR spectra of examined samples and their pyrolysis products—chars and condensates: (1) aromatic skeletal vibrations of guaiacyl rings; (2) pyran ring symmetric scissoring C-H deformation, asymmetric in-plane for lignin and hemicellulose; (3) aromatic skeletal vibration combined with C-H in-plane deformation of cellulose; (4) C-H deformation in hemicellulose; (5) C-O vibration in syringil and guaiacyl rings; (6) guaiacyl ring breathing, C-O stretch in lignin, C-O linkage in guaiacyl aromatic methoxyl groups; (7) C-O-C stretching vibration ring skeletal pyranose; (8) C-O stretching in cellulose; (9) C-O stretching in cellulose; (10) C-H out-of-plane glucose ring in cellulose and hemicellulose or 1-2-disubstituted aromatic rings [30–33]. Energies 2021, 14, x FOR PEER REVIEW 6 of 18

Figure 2. ATR spectra of examined samples and their pyrolysis products—chars and condensates: (1) aromatic skeletal vibrations of guaiacyl rings; (2) pyran ring symmetric scissoring C-H deformation, asymmetric in-plane for lignin and hemicellulose; (3) aromatic skeletal vibration combined with C-H in-plane deformation of cellulose; (4) C-H deformation in hemicellulose; (5) C-O vibration in syringil and guaiacyl rings; (6) guaiacyl ring breathing, C-O stretch in lignin, C-O linkage in guaiacyl aromatic methoxyl groups; (7) C-O-C stretching vibration ring skeletal pyranose; Energies 2021, 14, 2091 (8) C-O stretching in cellulose; (9) C-O stretching in cellulose; (10) C-H out-of-plane glucose ring6 of in 18 cellulose and hemicellulose or 1-2-disubstituted aromatic rings [30–33].

The bands present in the spectra of condensates correspond to the disappearance of the aforementionedThe bands present bands in in the the spectra spectra of ofcondensates chars. Figure correspond 2 shows that to there the disappearance is a high contributionof the aforementioned ratio of the bands bands corresponding in the spectra to ofthe chars. presence Figure of bonds2 shows of thatthe Cal-H there istype, a high the bandscontribution corresponding ratio of the to aromatic bands corresponding C-H out-of-p tolane the bend presence in the of range bonds of of 900–600 the Cal-H 1/cm, type, and the bands corresponding to aromatic C-H out-of-plane bend in the range of 900–600 1/cm, and other bands one to nine, characteristic of dactyloscopic range of biomass in the composi- other bands one to nine, characteristic of dactyloscopic range of biomass in the composition tion of condensates (for explanation, see figure caption). Special attention should be given of condensates (for explanation, see ﬁgure caption). Special attention should be given to to band 10 in the condensates of the BL and RL samples, which correspond to the presence band 10 in the condensates of the BL and RL samples, which correspond to the presence of of the bond of C-H bending in 1,2-disubstituted aromatics in these samples. It is important the bond of C-H bending in 1,2-disubstituted aromatics in these samples. It is important to mention the fact that the contribution ratio of polar compounds able to form hydrogen to mention the fact that the contribution ratio of polar compounds able to form hydrogen bonds in the condensates of the ML and RL samples increased in comparison to their bi- bonds in the condensates of the ML and RL samples increased in comparison to their ochars. In contrast to the BL and RL samples, no substantial increase in the contribution biochars. In contrast to the BL and RL samples, no substantial increase in the contribution ratio of carbonyl groups was observed in the condensate of the ML sample. ratio of carbonyl groups was observed in the condensate of the ML sample. The spectra of the initial samples of leaves and the products of their pyrolysis pre- The spectra of the initial samples of leaves and the products of their pyrolysis presented in Figure 2 show that the differences in contribution ratio of particular groups of sented in Figure2 show that the differences in contribution ratio of particular groups of atomsatoms in in initial initial samples samples of of different different leaves leaves also also cause cause differences differences in in the the contribution contribution ratio ratio ofof these these groups groups in in condensates condensates and and biochars. biochars. In the In condensates the condensates of samples, of samples, there are there some are compoundssome compounds able to ableform to hydrogen form hydrogen bonds and bonds compounds and compounds with polar with groups. polar groups.This implies This thatimplies there that can there be intermolecular can be intermolecular interactions, interactions, leading leading to self-aggregation to self-aggregation of these of com- these poundscompounds in volatile in volatile products products and to and the format to theion formation of aerosol of aerosolparticles. particles. Taking into Taking account into thataccount pyrolysis that pyrolysis is the first is thestage ﬁrst of stagethe combus of the combustiontion process, process, it can be it canpresumed be presumed that these that aerosolthese aerosol particles particles will form will particles form particles of suspended of suspended dust and dust additionally and additionally poison poison the envi- the ronment.environment. To describe To describe the observed the observed differences differences in composition in composition of the condensates of the condensates in Figure in 2,Figure UV spectroscopy2, UV spectroscopy was included was included in the analysis. in the analysis. FigureFigure 3 presentspresents the the UV UV spectra spectra of of the the condensate condensate material material of ofthe the BL, BL, RL, RL,and andML samples.ML samples.

1.1 BL 1 ML RL

Abs0.5

0 195250 300 350 Wavelen gth [n m] FigureFigure 3. 3. UVUV spectra spectra of of obtained obtained condensates. condensates.

SimilarSimilar to to the the ATR ATR spectra, spectra, the the UV UV spectra spectra of of condensates condensates have have different different shapes, shapes, whichwhich implies implies differences differences in in their their composition. composition. These These differences differences concern concern the the presence presence of of compoundscompounds containing containing chromophore chromophore groups. groups. Th Thee condensate condensate of of the the ML ML sample sample has has the the greatestgreatest value value of of absorbance absorbance in in the the entire entire discussed discussed range range of of wavelengths wavelengths (195–350 (195–350 nm). nm). ThisThis suggests suggests that that there there is is a a greater greater concen concentrationtration of of the the mixture mixture of of various compounds withwith chromophore chromophore groups groups in in this this condensate. condensate. The Thecondensate condensate of the of BL the sample BL sample is a more is a more homogeneous material with respect to the occurrence of compounds with particular chromophore groups. In the spectrum of this condensate near 235 nm, there is a distinct maximum, which implies the presence of a prevailing amount of compounds of the same

type. In the ATR spectrum of the condensate of the BL sample, there is no band near 1600 1/cm; however, in the ATR spectra of the RL and ML samples, such a band is visible (see Figure2). This implies a greater homogeneity of the material of the BL condensate. On the basis of the ATR spectra of this condensate (Figure2), it can be suggested that the compounds with carbonyl chromophore group in the acyclic aliphatic compounds prevail in its composition [34]. Energies 2021, 14, 2091 7 of 18

3.2. Analysis of Extraction Products The studied samples of leaves were extracted in order to explain other details about their composition. Table4 presents the results of the extraction of dried samples of leaves. The data in this table imply that the birch tree has the largest amount of the material soluble in a mixture of chloroform and methanol under extraction conditions.

Table 4. Yield of material soluble in mixture of chloroform and methanol and yield of residual material.

Yield of Extract Calculated Yield of Extract Calculated from Sample from the Mass of the the Mass of the Extract [%] After-Extraction Residue [%] BL 22.12 ± 0.20 24.13 ± 1.74 ML 17.20 ± 0.03 17.93 ± 1.77 RL 17.78 ± 1.30 19.26 ± 2.12

Figure4 presents the ATR spectra of extracts and after-extraction residues of leaves normalized with regard to the C=C band near 1600 1/cm. For better comparison, the spectra of initial samples are presented by a dotted line. The assignment of bands one to ﬁve in the ATR spectrum is explained in the ﬁgure caption. Attention should be drawn to the fact that the β-1,4-glycoside bond, giving band six in the spectra near 1030 cm−1, underwent degradation as a result of extraction. The relative height of this band decreased in the after-extraction residue but increased in the extract. This implies that the part of the material of leaves that contained this bond was extracted. The analysis of the spectra in Figure4 shows that, in the extracts, after-extraction residue, and initial samples there occur some bands near the same wavenumbers. Hence, it can be suggested that the same groups of atoms occur in these samples. However, the contribution ratios of these groups of atoms and bonds in the formation of the ATR spectra of the discussed samples are different. The contribution ratio of hydrogen bonds in the RBL sample is lower, whereas in the RRL sample it is higher than in initial samples. In the RML sample, the contribution ratio of hydrogen bonds in the after-extraction residue and initial sample is the same. Attention should be drawn to the fact that, similar to the condensates, the samples of BL, ML, and RL have a high contribution ratio of carbonyl groups. However, the compounds, the presence of which in the spectra causes the occurrence of bands in the range of somehow greater values of wavenumbers, prevail in the extracts of the BL and ML samples. These can be esters (1750–1735 1/cm), carboxyl acids (1760 1/cm), aldehydes (1740–1720 1/cm), and aliphatic ketones (1725–1705 1/cm). In the extract of the RL sample (the bands in the area of somehow higher values of wavenumbers), the presence of dimmers of conjugated acid (1710–1680 1/cm), conjugated aldehyde (1710–1685 1/cm), and conjugated ketone (1685–1666 1/cm) should be expected [35]. The discussed results show that there is some material moved out as a result of the microwave extraction of leaves. The compounds able to form hydrogen bonds and the compounds with carbonyl groups are removed together with this material. The contribution ratio of almost all functional groups decreases in the after-residue extract. This implies that extraction before pyrolysis makes it possible to decrease the content of hydrocarbons hazardous to the environment in the composition of volatile products of pyrolysis. Ipso facto, the contribution of hazardous compounds in the composition of combustion products emitted to the environment will be decreased. However, the obtained material of extracts after fractionation can be used as a source of platform chemicals. Energies 2021, 14Energies, x FOR 2021PEER, 14 REVIEW, 2091 8 of 18 8 of 18

Figure 4. ATRFigure spectra 4. ofATR examined spectra samplesof examined and theirsamples extraction and their products—extracted extraction products—extracted and residual materials: and residual (1) aromatic skeletal vibrationsmaterials: of guaiacyl (1) aromatic rings; skeletal (2) pyran vibrations ring symmetric of guai scissoringacyl rings; C-H (2) pyran deformation, ring symmetric asymmetric scissoring in plane C- for lignin and hemicellulose;H deformation, (3) C-H deformation asymmetric in in hemicellulose; plane for lignin (4) an guaiacyld hemicellulose; ring breathing, (3) C-H C-O deformation stretch in lignin, in hemicel- C-O linkage in guaiacyl aromaticlulose; methoxyl (4) guaiacyl groups; ring (5)breathing, C-O-C stretchingC-O stretch vibration in lignin, ring C-O skeletal linkage pyranose; in guaiacyl (6) aromatic C-O stretching methoxyl in cellulose groups; (5) C-O-C stretching vibration ring skeletal pyranose; (6) C-O stretching in cellulose and C- and C-O deformation [31–33,36]. O deformation [31–33,36]. Figure5 shows the UV spectra of the material extracted from the initial samples The discussedof leaves. results show that there is some material moved out as a result of the microwave extraction of leaves. The compounds able to form hydrogen bonds and the compounds with carbonyl groups are removed together with this material. The contribution ratio of almost all functional groups decreases in the after-residue extract. This implies that extraction before pyrolysis makes it possible to decrease the content of hydrocarbons hazardous to the environment in the composition of volatile products of pyroly-

Energies 2021, 14, x FOR PEER REVIEW 9 of 18

sis. Ipso facto, the contribution of hazardous compounds in the composition of combustion products emitted to the environment will be decreased. However, the obtained material of extracts after fractionation can be used as a source of platform chemicals. Energies 2021, 14, 2091 9 of 18 Figure 5 shows the UV spectra of the material extracted from the initial samples of leaves.

1.1 BL 1 ML RL

Abs0.5

0 195250 300 350 Energies 2021, 14, x FOR PEER REVIEW 10 of 18 Wavelength [nm] FigureFigure 5. 5. UVUV spectra spectra of of extracted extracted material. material.

from ThetheThe coursepresence presence of ofpyrolysis of distinct distinct maximaof maxima wood inof inthe the the UV same UV spectra spectraspecies. of extracts ofIn extractsthe (FigureDTG (Figure curves 5) implies5 )near implies thatthe temperaturestherethat thereare certain are of certain 470 classes and classes 675of chemical °C, of there chemical compounds appear compounds some with peaks. similar with The similar chromophorepresence chromophore of these groups peaks groupsoccur- can beringoccurring caused in the by in composition interactions the composition ofbetween extracts of extracts cellulose of the of thestudied and studied lignin samples samples[39,40] but or but ligninnot not a amixtureand mixture hemicellulose of of various various [41,42].chemicalchemical The compounds course of interaction (what cancan beofbe the suggestedsuggested inorganic on on thecomponents the basis basis of of the onthe shape the shape surface of of spectra spectra of char of theof with the RL theRLand volatileand ML ML samples products samples as seenof as the seen in pyrolysis Figure in Figure3). of This leav 3). fact esThis cannot will fact make bewill excluded the make separation the [43]. separation These of extract interactions of material extract couldmaterialinto particularhave into caused particular fractions the formation fractions easier. easier.of the new products, the decomposition of which contributed to the increase in mass loss rate at the temperatures of 470 and 675 °C and the 3.3. Analysis of Pyrolysis Process in a TG/FT-IR Unit formation3.3. Analysis of maximaof Pyrolysis visi Processble in thein a DTG TG/FT-IR curve. Unit FigureFigure 7a6a presents presents the the thermogravimetric TGA and DTG curves curves of and BL initialthe results samples of deconvolution and their after- of Figure 6a presents the TGA and DTG curves of BL initial samples and their after- theextraction DTG curve residues. of ML and RML samples (Figure 7b). extraction residues.

The results of deconvolution of the DTG spec3.7 tra are seen in Figure 6b. The curves of 8.5 mass loss in Figure 6a imply that the RBL sample10.2 in the temperatureNew pseudo- range of 200–600 °C component has an inclination towards a yield in volatile products of13.8 pyrolysis lower than the BL sam- Pseudo-component ple. However, in the temperature range of 270–320 °C and higherLignin than 430 °C, it shows a 9.0 28.6 Pseudo-component greater mass loss rate. In case of the RBL sample, the mass loss rateCellulose decreased II after extraction in the ranges of 120–260 °C and 320–420 °C, but at the temperaturesPseudo-component of 260–320 °C and higher than 430 °C, it increased. This implies that duringCellulose extraction I not only were 34.3 Pseudo-component 18.4 some components that underwent decomposition at lower temperaturesHemicellulose removed from the BL sample, but the structure of material of the entire samplePseudo-component loosened. On the basis of 6.6 Extractives thermogravimetric studies presented in works [37,38], the temperaturePseudo-components ranges were es- Moisture and LVS tablished in which basic components of biomass (hemicellulose,21.5 cellulose, and lignin) 26.8 should undergo decomposition during pyrolysis. The deconvolution of the DTG curves 2.9 of BL and RBL samples was carried out taking into account these temperature ranges. It 10.1 follows from the comparison of shape of the DTG5.7 curves of samples that, during extrac- BL RBL tion of BL, the extractives in the composition of volatiles were removed as the first from the sample. The contribution of these extractives during pyrolysis of RBL decreased by almost( a9.5) times (Figure 6b). The contribution of pseudo-components(b) of moisture and light volatile substances (LVS) increased during pyrolysis. Attention should be drawn to FigureFigure 6. 6. TGATGA and and DTG DTG curves curves (a ()a )along along with with the the results results of of deconvolution deconvolution of of the the DTG DTG curve curve (b (b) )of of BL BL and and RBL RBL samples. samples. an increase in contribution of the pseudo-components of hemicellulose and cellulose I and a decreaseThe results in contribution of deconvolution of cellulose of the II DTGin the spectra formation3.5 are seenof the in DTG Figure curve.6b. The This curves implies of ◦ that, under the influence of microwave extraction14.3 and anNew increased pseudo- pressure, there occur mass loss in Figure6a imply that the RBL sample in the temperaturecomponent range of 200–600 C destabilization and loosening of the structure of hemicellulose, cellulose, and lignin of the has an inclination towards a yield in volatile35.4 products of pyrolysisPseudo-component lower than the BL sample. BLHowever, sample, in which the temperature caused a decrease range of in 270–320 their thermal◦C14.0 and higherstability.Lignin than The 430 contribution◦C, it shows of a greatera new Pseudo-component pseudo-componentmass loss rate. In case in the of theformation RBL sample, of the theDTG mass curve loss Celluloseof ratethe II decreasedRBL sample after increased. extraction In ◦ ◦ ◦ workin the [3], ranges the authors of 120–260 statedC andthat 320–420the courseC, of but pyrolysis at thePseudo-component temperatures of leaves of forest of 260–320 trees differsC and 23.5 Cellulose I ◦ 20.3 higher than 430 C, it increased. This implies that duringPseudo-component extraction not only were some components that underwent decomposition at lower temperaturesHemicellulose removed from the Pseudo-component BL sample, but the structure of material of the entireExtractives sample loosened. On the basis

of thermogravimetric studies presented in works33.1 [37,38Pseudo-components], the temperature ranges were 28.5 Moisture and LVS established in which basic components of biomass (hemicellulose, cellulose, and lignin)

9.1 9.6 ML RML (a) (b) Figure 7. TGA and DTG curves (a) along with the results of deconvolution of the DTG curve (b) of ML and RML samples.

It follows from comparison of the curves of mass loss of ML and RML samples in Figure 7a that the extraction did not substantially influence the course of the pyrolysis process of the RML sample. The rate of removal of volatile products from the RML sample decreased in the temperature ranges of 170–275 °C and 360–610 °C and increased in the temperature range of 270–360 °C. The results of deconvolution in Figure 7b proved that

Energies 2021, 14, x FOR PEER REVIEW 10 of 18

from the course of pyrolysis of wood of the same species. In the DTG curves near the temperatures of 470 and 675 °C, there appear some peaks. The presence of these peaks can Energies 2021, 14, 2091 10 of 18 be caused by interactions between cellulose and lignin [39,40] or lignin and hemicellulose [41,42]. The course of interaction of the inorganic components on the surface of char with the volatile products of the pyrolysis of leaves cannot be excluded [43]. These interactions shouldcould undergo have caused decomposition the formation during of the pyrolysis. new products, The deconvolution the decomposition of the DTG of which curves con- oftributed BL andRBL to the samples increase was in carriedmass loss out rate taking at the into temperatures account these of temperature 470 and 675 ranges. °C and It the followsformation from of the maxima comparison visible of shapein the ofDTG the curve. DTG curves of samples that, during extraction of BL, theFigure extractives 7a presents in the the composition thermogravimetric of volatiles curves were and removed the results as theof deconvolution ﬁrst from the of sample.the DTG The curve contribution of ML and of these RML extractives samples (Figure during 7b). pyrolysis of RBL decreased by almost 9.5 times (Figure6b). The contribution of pseudo-components of moisture and light volatile 3.7 substances (LVS) increased during pyrolysis. Attention should8.5 be drawn to an increase in 10.2 New pseudo- contribution of the pseudo-components of hemicellulose and cellulosecomponent I and a decrease in 13.8 contribution of cellulose II in the formation of the DTG curve. ThisPseudo-component implies that, under the Lignin inﬂuence of microwave extraction and an increased pressure,9.0 there occur destabilization 28.6 Pseudo-component and loosening of the structure of hemicellulose, cellulose, and ligninCellulose of II the BL sample, Pseudo-component which caused a decrease in their thermal stability. The contributionCellulose of I a new pseudo- component in the formation of the DTG curve of the RBL sample34.3 increased.Pseudo-component In work [3], the 18.4 authors stated that the course of pyrolysis of leaves of forest trees differsHemicellulose from the course of Pseudo-component pyrolysis of wood of the same species. In the DTG curves6.6 near the temperaturesExtractives of 470 and ◦ Pseudo-components 675 C, there appear some peaks. The presence of these peaks can beMoisture caused and byLVS interactions 21.5 between cellulose and lignin [39,40] or lignin and26.8 hemicellulose [41,42]. The course of

interaction of the inorganic components on the surface of char2.9 with the volatile products of the pyrolysis of leaves cannot be excluded [43]. These interactions10.1 could have caused the 5.7 formation of the new products, the decomposition of which contributed to the increase in BL RBL mass loss rate at the temperatures of 470 and 675 ◦C and the formation of maxima visible

in the DTG curve. Figure(a7) a presents the thermogravimetric curves and the(b) results of deconvolution of Figure 6. TGA and DTG thecurves DTG (a) curvealong with of ML the and results RML of samplesdeconvolution (Figure of 7theb). DTG curve (b) of BL and RBL samples.

3.5

14.3 New pseudo- component

35.4 Pseudo-component 14.0 Lignin Pseudo-component Cellulose II Pseudo-component 23.5 Cellulose I 20.3 Pseudo-component Hemicellulose Pseudo-component Extractives

33.1 Pseudo-components 28.5 Moisture and LVS

9.1 9.6 ML RML (a) (b)

FigureFigure 7. TGA 7. TGA and and DTG DTG curves curves (a) ( alonga) along with with the the results results of deconvolutionof deconvolution of theof the DTG DTG curve curve (b) ( ofb) MLof ML and and RML RML samples. samples.

It followsIt follows from from comparison comparison of of the the curves curves of of mass mass loss loss of of ML ML and and RML RML samples samples in in FigureFigure7a that7a that the the extraction extraction did did not not substantially substantially inﬂuence influence the the course course of theof the pyrolysis pyrolysis processprocess of theof the RML RML sample. sample. The The rate rate of of removal remova ofl of volatile volatile products products from from the the RML RML sample sample ◦ ◦ decreaseddecreased in in the the temperature temperature ranges ranges of of 170–275 170–275C °C and and 360–610 360–610C °C and and increased increased in thein the ◦ temperaturetemperature range range of 270–360of 270–360C. The°C. The results results of deconvolution of deconvolution in Figure in Figure7b proved 7b proved that the that extraction did not cause any signiﬁcant changes in contribution of extractives, moisture, and LVS. However, it increased the removal of hemicellulose from the RML sample by 4.6%, the removal of cellulose I by 3%, and the removal of cellulose II by almost 12%. As a result of the extraction, the removal of volatile products corresponding to lignin pseudo-components decreased by almost 20%. Energies 2021, 14, x FOR PEER REVIEW 11 of 18

the extraction did not cause any significant changes in contribution of extractives, moisture, and LVS. However, it increased the removal of hemicellulose from the RML sample Energies 2021, 14, 2091 11 of 18 by 4.6%, the removal of cellulose I by 3%, and the removal of cellulose II by almost 12%. As a result of the extraction, the removal of volatile products corresponding to lignin pseudo-components decreased by almost 20%. FigureFigure8 presents 8 presents the TGAthe TGA and DTGand DTG curves curves of RL of and RL RRL and samplesRRL samples and the and results the results of deconvolutionof deconvolution of the of DTG the curvesDTG curves of these of these samples. samples.

0.4 1.2

20.6 New pseudo- 32.4 component Pseudo-component Lignin Pseudo-component Cellulose II 36.1 17.8 Pseudo-component Cellulose I Pseudo-component Hemicellulose 5.0 Pseudo-component 27.8 Extractives 22.2 Pseudo-components Moisture and LVS 6.6 6.9 3.4 10.8 8.8

RL RRL (a) (b)

FigureFigure 8. TGA 8. TGA and and DTG DTG curves curves (a) along(a) along with with the the results results of deconvolution of deconvolution of the of the DTG DTG curve curve (b) of(b) RL of andRL and RRL RRL samples. samples.

In contrastIn contrast to the to RBLthe RBL and RMLand RML samples, samples, the RRL the sampleRRL sample has a muchhas a lowermuch masslower loss mass thanloss the than RL samplethe RL sample in the temperature in the temperature range above range 225 above◦C (Figure225 °C 8(Figurea). In the 8a). temperature In the temper- ◦ rangeature of range 155–450 of 155–450C, the °C, RRL the sample RRL sample emits muchemits much fewer fewer volatile volatile products products of pyrolysis of pyroly- comparedsis compared to the RLto the sample. RL sample. It follows It follows from Figure from8 Figureb that the8b extractionthat the extraction did notincrease did not in- thecrease yield inthe extractives yield in extractives but decreased but decreased the yield in the moisture yield in and moisture LVS along and withLVS along the thermal with the stabilitythermal of hemicellulosestability of hemicellulose pseudo-components pseudo-components by 23.2%. The by thermal23.2%. The stability thermal of pseudo-stability of componentspseudo-components of cellulose of I andcellulose lignin I increasedand lignin substantially increased substantially after the extraction. after the extraction. TheThe results results presented presented in Figuresin Figures6–8 6–8imply imply that th someat some part part of organic of organic components components that that areare present present in the in compositionthe composition of volatile of volatile products products during during pyrolysis pyrolysis is removed is removed from BL, from RL, BL, andRL, ML and samples ML samples during during extraction. extraction. Although Although the studied the studied leaves belongleaves belong to lignocellulosic to lignocellu- biomass,losic biomass, the extraction the extraction affects theaffects loosening the loosen of structureing of structure and the and thermal the thermal stability stability of its of basicits componentsbasic components in different in different ways ways and thereforeand therefore the interactionsthe interactions taking taking place place between between thesethese components. components. In In turn, turn, this this means means that that during during pyrolysis pyrolysis of initialof initial samples samples of leavesof leaves andand their their after-extraction after-extraction residues, residues, the the structural structur changesal changes in thesein these materials materials will will take take place place in differentin different ways. ways. Figure9 presents the diffractograms of chars of leaves and their after-extraction Figure 9 presents the diffractograms of◦ chars of leaves and their after-extraction res- residuesidues pyrolyzed pyrolyzed at at a atemperature temperature of of450 450 °C. InC. the In thediffractograms, diffractograms, there there are visible are visible reflexes reflexes from the biochars included in biochars of inorganic components and from the reflex from the biochars included in biochars of inorganic components and from the reflex of of (002) NaF. In this case NaF is used as an internal standard. With regard to NaF reflex, (002) NaF. In this case NaF is used as an internal standard. With regard to NaF reflex, the the differences in integral intensity of the reflex from the biochar of samples can be visually differences in integral intensity of the reflex from the biochar of samples can be visually evaluated. Such differences point to the degree of ordering of pyrolyzed material. Based evaluated. Such differences point to the degree of ordering of pyrolyzed material. Based on the data in Table2 and the results published in works [ 31,44], it can be suggested that on the data in Table 2 and the results published in works [31,44], it can be suggested that the peaks in the diffractograms of biochars point to the presence of inorganic components the peaks in the diffractograms of biochars point to the presence of inorganic components in the form of quartz, calcite, K-feldspar, plagioclase feldspar, magnetite, etc. The shape in the form of quartz, calcite, K-feldspar, plagioclase feldspar, magnetite, etc. The shape of the reflex from biochar in the range of 2 theta 14–34 implies that, in biochar material, of the reflex from biochar in the range of 2 theta 14–34 implies that, in biochar material, except in the amorphous phase, the occurrence of two types of ordering, γ-fraction and except in the amorphous phase, the occurrence of two types of ordering, γ-fraction and crystalline phase, can be expected. The presence of γ-fraction is caused by local molecular crystalline phase, can be expected. The presence of γ-fraction is caused by local molecular ordering of aliphatic or alicyclic chains whereas crystalline (graphite-like) phase is formed from lamellas that have a structure similar to graphene layers [45,46].

Energies 2021, 14, x FOR PEER REVIEW 12 of 18

ordering of aliphatic or alicyclic chains whereas crystalline (graphite-like) phase is formed from lamellas that have a structure similar to graphene layers [45,46]. The comparison of the diffractograms implies that the char of RML sample has the best degree of ordering. Compared to the diffractogram of ML sample, it has a smaller angle of deviation of the background line on the side of smaller 2 theta angles. The char of RRL sample is less ordered, its diffractogram has the greatest angle of deviation of the background line on the side of smaller 2 theta angles. Taking into account that the loosening of the structure of material facilitates its ordering, it can be suggested that a better ordering of char from the RML sample results from a decrease in thermal stability of hemicellulose, cellulose I, and cellulose II. The results of deconvolution of the DTG curve presented in Figure 7b point to this. A worse ordering of the char of the RRL sample could have been caused only by a lower thermal stability of hemicellulose. The data in Figure 8b imply that the pseudo-components of cellulose and lignin of the RRL sample increased their thermal stability. The ordering of the char of the RBL sample took the medium position because the pseudo-components of hemicellulose, cellulose I, and lignin reduced their thermal stability (see the results of deconvolution in Figure 6b). The FT-IR spectra of samples were registered at different temperatures in order to evaluate the influence of extraction on the composition of volatile products of pyrolysis. Energies 2021, 14, 2091 Figure 10 presents the FT-IR spectra of volatile products of pyrolysis of BL and RBL12 ofsam- 18 ples. To make a contrastive analysis easier, the spectra were presented in such a way that the height of the CO2 band in them was the same.

Figure 9. Diffractograms of pyrolyzed leaves and their pyrolyzed after-extraction residues: q—quartz, k—K-feldspar (e.g., microline), p—plagioclase feldspar (e.g., albite, anorthite), c—calcite, m—magnetite (magnetite and maghemite overlap).

The comparison of the diffractograms implies that the char of RML sample has the best degree of ordering. Compared to the diffractogram of ML sample, it has a smaller angle of deviation of the background line on the side of smaller 2 theta angles. The char of RRL sample is less ordered, its diffractogram has the greatest angle of deviation of the background line on the side of smaller 2 theta angles. Taking into account that the loosening of the structure of material facilitates its ordering, it can be suggested that a better ordering of char from the RML sample results from a decrease in thermal stability of hemicellulose, cellulose I, and cellulose II. The results of deconvolution of the DTG curve presented in Figure7b point to this. A worse ordering of the char of the RRL sample could have been caused only by a lower thermal stability of hemicellulose. The data in Figure8b imply that the pseudo-components of cellulose and lignin of the RRL sample increased their thermal stability. The ordering of the char of the RBL sample took the medium position because the pseudo-components of hemicellulose, cellulose I, and lignin reduced their thermal stability (see the results of deconvolution in Figure6b). The FT-IR spectra of samples were registered at different temperatures in order to evaluate the inﬂuence of extraction on the composition of volatile products of pyrolysis. Figure 10 presents the FT-IR spectra of volatile products of pyrolysis of BL and RBL samples. To make a contrastive analysis easier, the spectra were presented in such a way that the height of the CO2 band in them was the same. Energies 2021, 14, x FOR PEER REVIEW 13 of 18

Figure 9. Diffractograms of pyrolyzed leaves and their pyrolyzed after-extraction residues: q— quartz, k—K-feldspar (e.g., microline), p—plagioclase feldspar (e.g., albite, anorthite), c—calcite, m—magnetite (magnetite and maghemite overlap).

It follows from the analysis of the shape of the FT-IR spectra of volatile products in Figure 10 that during pyrolysis of the after-extraction residues of birch leaves there are much lower contribution ratios of a mixture of saturated and unsaturated hydrocarbons, a mixture of compounds containing carbonyl groups, and a mixture of alcohols, phenols, and esters formed as a result of defragmentation of the β-1,4-glicoside bond observed in the composition of volatile products. Such a shape of the spectra proves that during extraction a significant part of the aforementioned compounds (over 20%) was extracted from birch leaves. Somehow, less material (about 17%) was extracted from the ML and RL samples. It was expected that the transformation of the shape of the spectra of volatiles from these samples, which was caused by extraction, would take place in similar way. However, it follows from the shape of the spectra in Figure 11 that, in the case of the ML and RML samples, the differences in shape are visible in the FT-IR spectra of volatile products that Energies 2021, 14, 2091 were emitted in the temperature range of 300–400 °C. In the composition of volatile 13prod- of 18 ucts of the RML sample (like RBL sample), there is an increased contribution ratio of water vapour and a decreased contribution ratio of saturated and unsaturated hydrocarbons.

BL Water vapour RBL CO2

CO2 Mixture of Mixture of compoundswith phenol, CO groups alcohol, and esters Water vapour Mixture of o hydrocarbons 250 C

300oC

350 oC

400 oC

o CH4 450 C

3500 3000 2500 2000 1500 1000 Wavenumber (1/cm) Figure 10. FT-IR spectra of volatile products emitted during pyrolysis of BL and RBL samples.samples.

It follows from the analysis of the shape of the FT-IR spectra of volatile products in Figure 10 that during pyrolysis of the after-extraction residues of birch leaves there are much lower contribution ratios of a mixture of saturated and unsaturated hydrocarbons, a mixture of compounds containing carbonyl groups, and a mixture of alcohols, phenols, and esters formed as a result of defragmentation of the β-1,4-glicoside bond observed in the composition of volatile products. Such a shape of the spectra proves that during extraction a signiﬁcant part of the aforementioned compounds (over 20%) was extracted from birch leaves. Somehow, less material (about 17%) was extracted from the ML and RL samples. It was expected that the transformation of the shape of the spectra of volatiles from these samples, which was caused by extraction, would take place in similar way. However, it follows from the shape of the spectra in Figure 11 that, in the case of the ML and RML samples, the differences in shape are visible in the FT-IR spectra of volatile products that were emitted in the temperature range of 300–400 ◦C. In the composition of volatile products of the RML sample (like RBL sample), there is an increased contribution ratio of water vapour and a decreased contribution ratio of saturated and unsaturated hydrocarbons. Energies 2021, 14, x FOR PEER REVIEW 14 of 18

No increase in the contribution ratio of water vapour was observed in the volatile products of pyrolysis of the RRL sample. There is a visible decrease in the contribution ratio of saturated and unsaturated hydrocarbons in the temperature range of 300–450 °C and a tendency toward a decrease in the contribution ratio of the mixture of compounds with carbonyl groups in the temperature range of 250–400 °C. Attention should be drawn to the shape of fragments of the spectrum in the range of wavenumbers corresponding to the occurrence of bands of water vapour at the temperatures of 350–450 °C (Figure 12). This shape may have been caused by the interactions of water vapour with the surface of formed char, and a decrease in its contribution ratio was observed as a result. It cannot be Energies 2021, 14, 2091 excluded that these interactions of water vapour with the surface of char could have14 been of 18 catalysed by inorganic components present on the char surface. The peaks on the diffractograms of chars point to the presence of these inorganic components [47].

Water vapour CO2

ML CO2 RML Mixture of compoundswith Mixture CO groups of phenol, alcohol, and esters Water vapour Mixture of hydrocarbons 250oC CO

300oC

350oC

400oC

450oC CH4

3500 3000 2500 2000 1500 1000 Wavenumber (1/cm) FigureFigure 11. 11. FT-IRFT-IR spectra spectra of of volatile volatile products products emitted emitted during pyrolysis of ML and RML samples.

No increase in the contribution ratio of water vapour was observed in the volatile products of pyrolysis of the RRL sample. There is a visible decrease in the contribution ratio of saturated and unsaturated hydrocarbons in the temperature range of 300–450 ◦C and a tendency toward a decrease in the contribution ratio of the mixture of compounds with carbonyl groups in the temperature range of 250–400 ◦C. Attention should be drawn to the shape of fragments of the spectrum in the range of wavenumbers corresponding to the occurrence of bands of water vapour at the temperatures of 350–450 ◦C (Figure 12). This shape may have been caused by the interactions of water vapour with the surface of formed char, and a decrease in its contribution ratio was observed as a result. It cannot be excluded that these interactions of water vapour with the surface of char could have been catalysed by inorganic components present on the char surface. The peaks on the diffractograms of chars point to the presence of these inorganic components [47]. Energies 2021, 14, x FOR PEER REVIEW 15 of 18

Energies 2021, 14, 2091 15 of 18 The FT-IR spectra of volatile products of RL and RRL samples show some differences starting at a temperature of 300 °C (Figure 12).

Water vapour CO2

RL CO2 RRL Mixture of compoundswith Mixture CO groups of phenol , alcohol, and esters Water vapour Mixture of hydrocarbons 250oC CO

300oC

350oC

400oC

CH4 450oC

3500 3000 2500 2000 1500 1000 Wavenumber (1/cm) FigureFigure 12.12. FT-IRFT-IR spectraspectra ofof volatilevolatile productsproducts emittedemitted duringduring pyrolysispyrolysis ofof RLRL andand RRLRRL samples.samples.

4. ConclusionsThe FT-IR spectra of volatile products of RL and RRL samples show some differences ◦ startingThe at leaves a temperature of urban oftrees 300 fallingC (Figure in autumn 12). should be utilized. The research on the course of the pyrolysis process of leaves (as the first step of the combustion process) has 4. Conclusions proven that leaves can be subjected to combustion after the extraction operation. The ad- vantagesThe leavesof such of an urban approach trees are falling as follows: in autumn should be utilized. The research on the course of the pyrolysis process of leaves (as the ﬁrst step of the combustion process) has proven that leaves can be subjected to combustion after the extraction operation. The advantages of such an approach are as follows:

Energies 2021, 14, 2091 16 of 18

(i) The yield of extraction of birch, maple, and rowan leaves is 22.12 ± 0.20%, 17.20 ± 0.03%, and 17.78 ± 1.30%, respectively. (ii) The removal of such an amount of material reduces the burden on the environment caused by dangerous chemical compounds present in the composition of volatile products of pyrolysis (aromatic hydrocarbons, compounds with carbonyl groups, alcohols, phenols, water vapour). (iii) The composition of extracted material has greater homogeneity in groups of compounds with carbonyl chromophore groups compared to the composition of the condensate that is formed during pyrolysis. (iv) The separation of extracted material into fractions will enrich the raw material platform with materials that can be used to obtain phytochemicals and other chemical reagents. (v) The removal of dangerous compounds from the material of leaves will increase the amount of energy resources that can be solely subjected to combustion or added to other species of combusted biomass.

Author Contributions: Conceptualization, V.Z.; methodology, V.Z.; investigation, V.Z., A.S. and M.B.; data curation, V.Z., A.S. and M.B., resources, A.S.; validation, A.S.; writing—original draft preparation, V.Z.; writing—review and editing, V.Z.; project administration, V.Z. and A.S.; funding acquisition, A.S. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Rector of Jan Kochanowski University in Kielce, grant number SIGR.RN.20.078.611. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conﬂicts of Interest: The authors declare no conﬂict of interest.

Abbreviations

ATR attenuated total reﬂectance BL birch leaves DTG differential thermogravimetric curve ED-XRF energy dispersive X-ray ﬂuorescence FT-IR Fourier-transform infrared spectroscopy HHV higher heating value ML maple leaves RBL residual birch leaves RML residual maple leaves RL rowan leaves RRL residual rowan leaves TEM transmission electron microscope TG/FT-IR thermogravimetric Fourier-transform infrared measured unit UV ultraviolet spectroscopy XRD X-ray diffraction

References 1. Tálos-Nebehaj, E.; Hofmann, T.; Albert, L. Seasonal changes of natural antioxidant content in the leaves of Hungarian forest trees. Ind. Crop. Prod. 2017, 98, 53–59. [CrossRef] 2. Taamalli, A.; Sánchez, J.L.; Jebabli, H.; Trabelsi, N.; Abaza, L.; Carretero, A.S.; Cho, J.Y.; Román, D.A. Monitoring the Bioactive Compounds Status in Olea europaea According to Collecting Period and Drying Conditions. Energies 2019, 12, 947. [CrossRef] 3. He, M.; Mourant, D.; Gunawan, R.; Lievens, C.; Wang, X.S.; Ling, K.; Bartle, J.; Li, C.-Z. Yield and properties of bio-oil from the pyrolysis of mallee leaves in a ﬂuidised-bed reactor. Fuel 2012, 102, 506–513. [CrossRef] 4. Mythili, R.; Venkatachalam, P.; Subramanian, P.; Uma, D. Characterization of bioresidues for biooil production through pyrolysis. Bioresour. Technol. 2013, 138, 71–78. [CrossRef] Energies 2021, 14, 2091 17 of 18

5. Singh, J.; Kaur, S.; Lee, J.; Mehta, A.; Kumar, S.; Kim, K.-H.; Basu, S.; Rawat, M. Highly fluorescent carbon dots derived from Mangifera indica leaves for selective detection of metal ions. Sci. Total. Environ. 2020, 720, 137604. [CrossRef][PubMed] 6. Cai, W.; Zhou, Q.; Xie, Y.; Liu, J.; Long, G.; Cheng, S.; Liu, M. A direct carbon solid oxide fuel cell operated on a plant derived biofuel with natural catalyst. Appl. Energy 2016, 179, 1232–1241. [CrossRef] 7. Raghuvansi, S.K.S.; Prasad, R.; Mishra, A.S.; Chaturvedi, O.H.; Tripathi, M.K.; Misra, A.K.; Jakhmola, R.C.; Saraswat, B.L.; Jakhmola, R.C. Effect of inclusion of tree leaves in feed on nutrient utilization and rumen fermentation in sheep. Bioresour. Technol. 2007, 98, 511–517. [CrossRef] 8. Bilba, K.; Arsene, M.; Ouensanga, A. Study of banana and coconut fibers: Botanical composition, thermal degradation and textural observations. Bioresour. Technol. 2007, 98, 58–68. [CrossRef] 9. Wang, R.; Wang, P.; Yan, X.; Lang, J.; Peng, C.; Xue, Q. Promising Porous Carbon Derived from Celtuce Leaves with Outstanding Supercapacitance and CO2 Capture Performance. ACS Appl. Mater. Interfaces 2012, 4, 5800–5806. [CrossRef] 10. Zhu, X.; Yu, S.; Xu, K.; Zhang, Y.; Zhang, L.; Lou, G.; Wu, Y.; Zhu, E.; Chen, H.; Shen, Z.; et al. Sustainable activated carbons from dead ginkgo leaves for supercapacitor electrode active materials. Chem. Eng. Sci. 2018, 181, 36–45. [CrossRef] 11. Wang, Q.; Zhang, Y.; Xiao, J.; Jiang, H.; Hu, T.; Meng, C. Copper oxide/cuprous oxide/hierarchical porous biomass-derived carbon hybrid composites for high-performance supercapacitor electrode. J. Alloys Compd. 2019, 782, 1103–1113. [CrossRef] 12. Wei, X.; Li, Y.; Gao, S. Biomass-derived interconnected carbon nanoring electrochemical capacitors with high performance in both strongly acidic and alkaline electrolytes. J. Mater. Chem. A 2017, 5, 181–188. [CrossRef] 13. Kalamdhad, A.S.; Singh, Y.K.; Ali, M.; Khwairakpam, M.; Kazmi, A.A. Rotary drum composting of vegetable waste and tree leaves. Bioresour. Technol. 2009, 100, 6442–6450. [CrossRef] 14. Moharana, P.C.; Biswas, D.R. Assessment of maturity indices of rock phosphate enriched composts using variable crop residues. Bioresour. Technol. 2016, 222, 1–13. [CrossRef][PubMed] 15. Shahid, M.; Dumat, C.; Khalid, S.; Schreck, E.; Xiong, T.; Niazi, N.K. Foliar heavy metal uptake, toxicity and detoxification in plants: A comparison of foliar and root metal uptake. J. Hazard. Mater. 2017, 325, 36–58. [CrossRef] 16. Yalaltdinova, A.; Kim, J.; Baranovskaya, N.; Rikhvanov, L. Populus nigra L. as a bioindicator of atmospheric trace element pollution and potential toxic impacts on human and ecosystem. Ecol. Indic. 2018, 95, 974–983. [CrossRef] 17. Pourkhabbaz, A.; Rastin, N.; Olbrich, A.; Langenfeld-Heyser, R.; Polle, A. Influence of Environmental Pollution on Leaf Properties of Urban Plane Trees, Platanus orientalis L. Bull. Environ. Contam. Toxicol. 2010, 85, 251–255. [CrossRef] 18. Alaimo, M.G.; Varrica, D. Recognition of Trace Element Contamination Using Ficus macrophylla Leaves in Urban Environment. Int. J. Environ. Res. Public Health 2020, 17, 881. [CrossRef] 19. Oo, H.M.; Karin, P.; Chollacoop, N.; Hanamura, K. Physicochemical characterization of forest and sugarcane leaf combustion’s particulate matters using electron microscopy, EDS, XRD and TGA. J. Environ. Sci. 2021, 99, 296–310. [CrossRef][PubMed] 20. Makkawi, Y.; El Sayed, Y.; Salih, M.; Nancarrow, P.; Banks, S.; Bridgwater, T. Fast pyrolysis of date palm (Phoenix dactylifera) waste in a bubbling fluidized bed reactor. Renew. Energy 2019, 143, 719. [CrossRef] 21. Enes, T.; Aranha, J.; Fonseca, T.; Lopes, D.; Alves, A.; Lousada, J. Thermal Properties of Residual Agroforestry Biomass of Northern Portugal. Energies 2019, 12, 1418. [CrossRef] 22. Regueiro, A.; Jezerská, L.; Pérez-Orozco, R.; Patiño, D.; Zegzulka, J.; Neˇcas,J. Viability Evaluation of Three Grass Biofuels: Experimental Study in a Small-Scale Combustor. Energies 2019, 12, 1352. [CrossRef] 23. Zhao, F.; Shu, L.; Wang, Q.; Wang, M.; Tian, X. Emissions of volatile organic compounds from heated needles and twigs of Pinus pumila. J. For. Res. 2011, 22, 243–248. [CrossRef] 24. Samae, H.; Tekasakul, S.; Tekasakul, P.; Furuuchi, M. Emission factors of ultrafine particulate matter (PM < 0.1 µm) and particle- bound polycyclic aromatic hydrocarbons from biomass combustion for source apportionment. Chemosphere 2021, 262, 127846. [CrossRef] 25. Li, W.; Shao, L. Transmission electron microscopy study of aerosol particles from the brown hazes in northern China. J. Geophys. Res. Space Phys. 2009, 114, 1–10. [CrossRef] 26. Davis, J.; Tiwari, K.; Novosselov, I. Soot morphology and nanostructure in complex flame flow patterns via secondary particle surface growth. Fuel 2019, 245, 447–457. [CrossRef] 27. Chang, Q.; Gao, R.; Gao, M.; Yu, G.; Wang, F. The structural evolution and fragmentation of coal-derived soot and carbon black during high-temperature air oxidation. Combust. Flame 2020, 216, 111–125. [CrossRef] 28. Li, D.; Li, W.; Chen, H.; Li, B. The adjustment of hydrogen bonds and its effect on pyrolysis property of coal. Fuel Process. Technol. 2004, 85, 815–825. [CrossRef] 29. Fang-Jing, L.; Wei, X.Y.; Zong, Z.-M.; Fan, M. Characterization of the oxygenated chemicals produced from supercritical methanolysis of modified lignites. Energy Fuels 2016, 30, 2636–2646. [CrossRef] 30. Lazdovica, K.; Kampars, V.; Liepina, L.; Vilka, M. Comparative study on thermal pyrolysis of buckwheat and wheat straws by using TGA-FTIR and Py-GC/MS methods. J. Anal. Appl. Pyrolysis 2017, 124, 1–15. [CrossRef] 31. Zubkova, V.; Strojwas, A.; Bielecki, M.; Kieush, L.; Koverya, A. Comparative study of pyrolytic behavior of the biomass wastes originating in the Ukraine and potential application of such biomass. Part 1. Analysis of the course of pyrolysis process and the composition of formed products. Fuel 2019, 254, 115688. [CrossRef] 32. Traoré, M.; Kaal, J.; Matínez Cortizas, A. Application of FTIR spectroscopy to the characterization of archeological wood. Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 2016, 153, 63–70. [CrossRef] Energies 2021, 14, 2091 18 of 18

33. Zhao, X.-B. Peracetic acid pretreatment of sugarcane bagasse for enzymatic hydrolysis: A continued work. J. Chem. Technol. Biotechnol. 2008, 6, 950–956. [CrossRef] 34. Rao, C.N.R. Ultra-Violet and Visible Spectroscopy. Chemical Applications; Butterworth and Co (Publishers) Ltd.: London, UK, 1975; pp. 58–62. 35. Ramos-Carmona, S.; Martínez, J.D.; Pérez, J.F. Torrefaction of patula pine under air conditions: A chemical and structural characterization. Ind. Crop. Prod. 2018, 118, 302–310. [CrossRef] 36. Coates, J. Interpretation of Infrared Spectra, A Practical Approach. In Encyclopedia of Analytical Chemistry; Meyers, R.A., Ed.; John Wiley and Sons Ltd.: Chichester, UK, 2000; pp. 10815–10837. 37. Yang, H.; Yan, R.; Chen, H.; Lee, D.H.; Zheng, C. Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel 2007, 86, 1781–1788. [CrossRef] 38. Yang, H.; Yan, R.; Chen, H.; Zheng, C.; Lee, D.H.; Liang, D.T. In-Depth Investigation of Biomass Pyrolysis Based on Three Major Components: Hemicellulose, Cellulose and Lignin. Energy Fuels 2006, 20, 388–393. [CrossRef] 39. Wu, S.; Shen, D.; Hu, J.; Zhang, H.; Xiao, R. Cellulose-lignin interactions during fast pyrolysis with different temperatures and mixing methods. Biomass Bioenergy 2016, 90, 209–217. [CrossRef] 40. Hosoya, T.; Kawamoto, H.; Saka, S. Cellulose–hemicellulose and cellulose–lignin interactions in wood pyrolysis at gasiﬁcation temperature. J. Anal. Appl. Pyrolysis 2007, 80, 118–125. [CrossRef] 41. Liu, Q.; Zhong, Z.; Wang, S.; Luo, Z. Interactions of biomass components during pyrolysis: A TG-FTIR study. J. Anal. Appl. Pyrolysis 2011, 90, 213–218. [CrossRef] 42. Hu, J.; Jiang, B.; Liu, J.; Sun, Y.; Jiang, X. Inﬂuence of interactions between biomass components on physicochemical characteristics of char. J. Anal. Appl. Pyrolysis 2019, 144, 104704. [CrossRef] 43. Wang, M.; Tian, J.; Roberts, D.G.; Chang, L.; Xie, K. Interactions between corncob and lignite during temperature-programmed co-pyrolysis. Fuel 2015, 142, 102–108. [CrossRef] 44. Lei, S.; Shi, Y.; Qiu, Y.; Che, L.; Xue, C. Performance and mechanisms of emerging animal-derived biochars for immobilization of heavy metals. Sci. Total. Environ. 2019, 646, 1281–1289. [CrossRef][PubMed] 45. Siddiqui, M.N.; Ali, M.F.; Shirokoff, J. Use of X-ray diffraction in assessing the aging pattern of asphalt fractions. Fuel 2002, 81, 51–58. [CrossRef] 46. AlHumaidan, F.S.; Hauser, A.; Rana, M.S.; Lababidi, H.M.S.; Behbehani, M. Changes in asphaltene structure during thermal cracking of residual oils: XRD study. Fuel 2015, 150, 558–564. [CrossRef] 47. Couhert, C.; Commandre, J.-M.; Salvador, S. Is it possible to predict gas yields of any biomass after rapid pyrolysis at high temperature from its composition in cellulose, hemicellulose and lignin? Fuel 2009, 88, 408–417. [CrossRef]