Customized Atmospheric Catalytic Hydropyrolysis of Biomass to High-Quality Bio-Oil Suitable for Coprocessing in Refining Units

: This study aimed to investigate the critical elements of the biomass ex situ catalytic hydropyrolysis (CHP) concept to improve the quality of fast pyrolysis bio-oil (FPBO) for further coprocessing in a fluid catalytic cracking (FCC) refining unit. Generally, the high oxygen and low hydrogen contents of biomass result in a bio-oil with low quality, necessitating its upgrading, which can be performed as integrated in the pyrolysis process via in situ or ex situ configuration. In this work, the quality of stem wood-derived pyrolyzates (520 ° C) was improved via ex situ CHP (400 ° C) using a continuous bench-scale drop tube pyrolyzer (60 g h − 1 ), and then the produced FPBO was coprocessed with vacuum gas oil (VGO) fossil oil using a lab-scale FCC unit (525 ° C). CHP of stem wood was carried out using different metal-acid catalysts such as Ni/HZSM-5, Ni/HBeta, Mo/TiO 2 , and Pt/TiO 2 at atmospheric pressure. FCC runs were performed using an equilibrium FCC catalyst and conventional fossil FCC feedstock cofed with 20 wt % stem wood-derived bio-oil in a fluidized bed reactor. Cofeeding the nonupgraded FPBO with fossil oil into the FCC unit decreased the generation of hydrocarbons in the range of gasoline and naphtha, indicating that bio-oil needs to be upgraded for further coprocessing in the FCC unit. Experimental results showed that different catalysts significantly affected the product composition and yield; Ni-based catalysts were strongly active tending to generate a high yield of gas, while Mo-and


INTRODUCTION
The 2050 global target of reaching net zero CO 2 emissions necessitates a gradual transition to renewable fuels in the transport sector.In Sweden, one potential way to reach this target is to utilize forest residual products (biomass) and convert them into transport fuels.So far, the thermochemical technique of fast pyrolysis has been vastly and efficiently used for biomass conversion into the energy-valued product of pyrolysis oil.−4 Therefore, upgrading of the pyrolysis oil is needed before it reaches the quality required for a fuel in the current transportation infrastructure either via blending or coprocessing strategies.An alternative to postpyrolysis upgrading processes is that the quality of pyrolysis oil is improved by integrating the pyrolysis process with a catalytic upgrading system in one operation.The catalyst can be used in both in situ and ex situ configurations, naming the whole process as in situ catalytic pyrolysis and ex situ catalytic pyrolysis, respectively.The ex situ configuration has the potential of optimizing the catalytic operation section separate from the pyrolysis part, preventing the inorganics from reaching the catalyst chamber, as well as regenerating/reusing the catalyst, which is economically defensible.Besides the application of the catalyst in the biomass pyrolysis process, introducing hydrogen into the system (hydropyrolysis process) enhances the hydrogen content of the hydrocarbon pool of the pyrolysis intermediates, causing the selective deoxygenation of O-containing pyrolyzates to valuable chemicals in the range of transport fuels.
Coprocessing of the FPBO in industrially available refinery processes, such as fluid catalytic cracking (FCC), is considered beneficial to produce high-value chemicals of gasoline, light olefins, and middle distillates, while part of their carbon is biogenic.The coprocessing of the FPBO in the FCC unit has been widely investigated considering different upgrading methods, 5−8 such as catalytic upgrading 9 and hydroprocessing, 10,11 FCC catalyst advancement, 12,13 as well as catalytic cracking condition optimization.The addition of pyrolysis biooil (up to 10 wt % feed share) was reported to provide successful processing with minor effect on the produced fuel quality. 6,11,14One of the main reasons for upgrading the FPBO via a hydrotreatment process is to be able to introduce this valuable renewable product into the current refinery units and replace the fossil-based carbon resources with renewable ones.In fact, the focus of FPBO upgrading could be to convert the dimers and trimers of this oil into light monomers, making them ready for FCC-type reaction processing.The transition from the petroleum era into the renewable era should not necessarily require major modifications to the refineries, high costs, and large amounts of time.By replacing fossil-based carbon resources in the existing refinery units with renewable ones, refiners can produce sustainable fuels and chemicals while meeting the final product standards and other regulatory requirements.Preem, the Swedish petroleum and biofuel company, started up with FCC demonstration trials of Energy & Fuels coprocessing FPBO in 2021 in the Lysekil refinery in Sweden.From January 2022, the FPBO of sawdust produced by Pyrocell in Sweden is being continuously coprocessed in Preem's plant in Lysekil.The demonstration trials were supported by the Swedish Energy Agency, and it was reported that no significant overall influence on the FCC unit was observed.However, the long-duration demonstration trials showed that certain oxygenates affect the downstream processes within the refinery.Improved properties of the FPBO could therefore be beneficial to be able to coprocess it in larger volumes in the FCC.
Hence, this study aimed to investigate the critical elements of the biomass ex situ catalytic hydropyrolysis concept to improve the quality of FPBO for further coprocessing in an FCC refining unit.−19 Therefore, in this work, the quality of stem wood-derived pyrolyzates was improved via ex situ catalytic hydropyrolysis using a continuous bench-scale drop tube pyrolyzer under atmospheric hydrogen flow.Catalytic hydropyrolysis of stem wood was carried out using different metal-acid catalysts such as Ni/ HZSM-5, Ni/HBeta, Mo/TiO 2 , and Pt/TiO 2 .These metalacid bifunctional catalysts, containing strong hydrogenation and hydrogenolysis functionalities, were selected to enhance the production of O-free compounds during the hydropyrolysis process under an atmospheric pressure of hydrogen.The produced FPBO was coprocessed with vacuum gas oil (VGO) fossil oil using a lab-scale FCC unit with the ratio of 20/80 w/ w.

MATERIALS AND METHODS
2.1.Materials.Stem wood of pine and spruce, with a particle size of 125−250 μm, was used as the pyrolysis feedstock.Before each pyrolysis test, the feedstock was dried at 105 °C to reach a moisture content of below 5 wt %.The ultimate and proximate properties as well as the inorganic content of the stem wood were analyzed by ALS Scandinavia AB, and the results are presented in Supporting Information, Table S1.5Ni/HZSM-5 (5 wt % Ni), 5Ni/HBeta (5 wt % Ni), 7Mo/TiO 2 (7 wt % Mo), 10Mo/TiO 2 (10 wt % Mo), and 7Pt/TiO 2 (7 wt % Pt), in pelletized form, were used as catalysts for ex situ upgrading of the stem wood-derived pyrolyzates.These metal-acid bifunctional catalysts, containing strong hydrogenation and hydrogenolysis functionalities, were selected to investigate the efficiency of the pyrolysis process in the production of O-free compounds under an atmospheric pressure of hydrogen.Titanium(IV) oxide (TiO 2 , pelletized, Alfa Aesar), ZSM-5 (SiO  O, Sigma-Aldrich] were used as metal catalyst precursors.H 2 and N 2 were used as reactants and carrier gases, respectively.All catalysts, Pt/TiO 2 , Mo/TiO 2 , Ni/ HBeta, and Ni/HZSM-5, were prepared by the incipient wetness impregnation method.Prior to impregnation, the pellets of TiO 2 and ZSM-5 and the powder Beta were calcined in air at 550 °C for 9 h.Then, the Ni, Mo, and Pt catalysts were loaded on the supports with the corresponding metal content (5−10 wt %).A proper amount of metal precursor was dissolved in deionized water, and then the solution was added dropwise to the support materials.The metalloaded TiO 2 , HZSM-5, and HBeta catalysts were first dried at room temperature for 24 h and at 110 °C for overnight, and then calcined at 550 °C for 9 h.The powder Ni/HBeta catalyst was pelletized using a manual press machine under 100 bar of pressure.Before each experiment, the catalysts were reduced at 450 °C for 2 h under a mixed flow (9 l min −1 ) of H 2 /N 2 (8.5/0.5 v/v).For the reusability test, the catalyst regeneration was carried out at 550 °C for 9 h in an air atmosphere.
Physicochemical properties of the metal oxide-supported catalysts were determined by N 2 isothermal adsorption−desorption (Micromeritics ASAP 2020 Plus), X-ray diffraction (XRD, BRUKER D2 PHASER), and scanning electron microscopy (SEM, thermoscientific, Apreo 2S) techniques.The amounts of metals loaded on the catalyst supports of TiO 2 , HZSM-5, and HBeta were determined by the inductively coupled plasma-mass spectrometry (ICP-MS) analysis technique at ALS Scandinavia AB.
2.2.Pyrolysis Experiments and Product Analysis.CHP experiments were conducted at atmospheric pressure in a drop tube pyrolyzer coupled to a catalytic fixed-bed reactor (Figure 1).Finely grounded and sieved stem wood (125−250 μm) was fed through a vibrating water-cooled feeder tube into the pyrolyzer (60 g h −1 ) for pyrolysis at 520 °C.The pyrolyzates of stem wood passed through the fixed-bed catalytic reactor for further upgrading.The feedstock was introduced into the furnace with 10 l min −1 carrier gas flow of N 2 /H 2 1.5/8.5 v/v.Thermal pyrolysis of stem wood, as the reference test, was carried out under 10 l min −1 of N 2 flow.The char product of pyrolysis process was collected in a cylindrical vessel and a hot-gas cylindrical filter placed successively after the pyrolyzer.The solid collector sections were kept heated at 500 °C.The pressure was measured in the pyrolyzer before and after the hot-gas filter to monitor pressure behavior inside the system.The pyrolysis gas products were passed through the catalyst bed for ex situ catalytic upgrading at 350−400 °C.A certain amount of pelletized bifunctional catalyst (60 g) was used for upgrading purpose.Before each experiment, the catalyst was reduced at 450 °C for 2 h under a mixed flow of H 2 (8.5 l min −1 ) and N 2 (0.5 l min −1 ).Different timeon-streams (90, 180, and 280 min) were investigated to understand the efficiency of the Mo/TiO 2 catalyst.To efficiently cool the downstream gas flow, it was mixed with a nitrogen flow of 5 l min −1 , indirectly cooled by liquid N 2 , and added through a sintered stainless pipe.The downstream gas flow was then passed through four glass bottles, placed in a cooling bath of glycol (−15 °C), to collect liquid products including aqueous and organic phases.Two of the glass bottles were equipped with microfilter candles (16−40 μm) to collect aerosols.Temperature of the gas leaving the cooling system was about 10 °C.After passing through a glass wool filter, components of the gas flow were continuously analyzed by μGC and FTIR.In addition, in a scheduled time, the gas products were collected in gas sampling bags for offline analysis using a GC-TCD.Data of online and offline gas analyzers were used to measure the mass of pyrolysis gas products.Yields of solid residue (char and catalytic coke) and liquid and gas products were measured to determine the mass efficiency of the pyrolysis process.All added material and products were weighed before and after the experiment to determine the yields of solid residue (char and catalytic coke) and pyrolysis liquid.
The liquid product obtained from CHP experiments consisted of the two phases of aqueous and organic.Yields of the aqueous and organic phases were measured by weighing the collected samples.Semiquantitative evaluation of the liquid samples was carried out using a Shimadzu GCMS-QP2010 Ultra chromatograph with two detectors, a mass spectrometer (MS) and a flame ionization detector (FID).The gas chromatograph column was a Restek RTX-1701.Prior the GC-MS/FID analysis, the liquids were diluted in acetone (GC ultratrace analysis grade, Scharlau).The MS peak identification was based on the NIST library search, and the corresponding FID peaks were used for quantitative assessment (% of total FID area) of each component selectivity.Karl Fischer titration, according to ASTM E203-08, was used to determine the water content of the liquid samples.Carbonyl titration, according to ASTM E3146-18a, was performed to quantitatively investigate the stability of the organic liquids.A sample of the organic liquid was dissolved in dimethyl sulfoxide (DMSO), and thereafter the solutions of hydroxylamine Energy & Fuels hydrochloride (NH 2 OH•HCl) and triethanolamine (TEA) were added.The mixture was then sealed, stirred, and heated to 80 °C for 2 h.Due to the reaction of carbonyl compounds (aldehydes and ketones) with NH 2 OH•HCl, the corresponding oxime was produced, and HCl was liberated.HCl liberated was consumed by TEA.After the reaction, unconsumed TEA was titrated with a standard HCl titrant to determine the molar concentration of carbonyls in the sample.C NMR and P NMR analyses were carried out to investigate the structural H−C framework and hydroxyl functional groups of organic liquids.The NMR were performed on a 500 MHz Bruker AV1 spectrometer equipped with a 5 mm QNP probe head with Zgradients operating at 25 °C.In the case of 13C-NMR, the FPBO (300 μL) was mixed with DMSO-d6 (deuterated dimethyl sulfoxide, 300 μL), and the spectra were referenced to DMSO-d6 at 39.5 ppm.In the case of 31 P NMR, the FPBO was derivatized with 2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane (TMDP) in a solution of pyridine:CDCl 3 1.6:1.0 in the presence of cyclohexanol as an internal standard and Cr(III)(OAcAc) 3 as a relaxation agent.In addition, elemental analysis (CHNS-O) was carried out to determine the heating value of the oil products.
In order to investigate the reusability of Mo/TiO 2 , the spent catalyst was collected after pyrolysis test and regenerated at 550 °C for 9 h under an air environment.The reusability test was carried out two times, and the regenerated catalysts are named as Mo/TiO 2 � Reg.1 and Mo/TiO 2 �Reg.2.

FCC Coprocessing and Product Evaluation.
A prospect of further processing of produced bio-oil was investigated by coprocessing in a FCC unit.In this work, a lab-scale FCC test unit was used to perform FCC refining in a semibatch fluidized reactor.A more detailed description of the FCC test unit can be found elsewhere. 20A fossil feed and the pyrolysis oil product were simultaneously fed with the cofeeding ratio of 20 wt % of FPBO into the reactor through two separate feed lines due to the feedstocks' immiscibility.Both the fossil and pyrolysis oils were preheated to 70 and 30 °C, respectively, to improve the viscosity characteristics without reaching thermal decomposition conditions.Temperature of the reactor preloaded with commercial FCC catalyst was 525 °C.Total mass ratio of the catalyst to feed was 5.After the FCC reaction, the vapor products were stripped from the product collection system to separate into liquid and gas products.After the product stripping step, a catalyst regeneration process followed with air at 650 °C to incinerate the remaining material (coke) deposited on the catalyst and the reactor surfaces, which was quantitively characterized based on the CO 2 formed.Properties of the fossil FCC feed and the commercial FCC catalyst can be found in a study held by Johansson et al. 20 The FCC-upgraded liquid products were evaluated qualitatively.The boiling point distributions of the liquid fractions were analyzed with GC-Simdist using the Standard Test Method for Boiling Range Distribution of Petroleum Fractions by Gas Chromatography ASTM D2887.The chemical composition of the upgraded liquid products was analyzed by the same gas chromatograph as used for the pyrolysis liquids, Shimadzu GCMS-QP2010 Ultra, but with another column, DB-Petro.The identified components for the upgraded products were categorized by paraffins, isoparaffins, olefins, naphthenes, aromatics, and oxygenated hydrocarbons.

Catalyst Screening Tests.
Figure S1 shows the representative SEM images and the corresponding energydispersive X-ray spectroscopy (EDS) analysis results of the TiO 2 , Pt/TiO 2 , Mo/TiO 2 , HZSM-5, Ni/HZSM-5, HBeta, and Ni/HBeta catalysts.The metal-oxide sites are observed to be well-dispersed on the TiO 2 surface.The elemental compositions are acquired by means of EDS.Both platinum and molybdenum were detected by EDS.A smoother surface of the catalyst was observed after Ni impregnation on HZSM-5, while no surface smoothness change was observed by impregnating Ni on HBeta.The EDS data clearly show that the nickel phase is present in both Ni/HZSM-5 and Ni/HBeta catalysts.
XRD of TiO 2 , Pt/TiO 2 , Mo/TiO 2 , HZSM-5, Ni/HZSM-5, HBeta, and Ni/HBeta catalysts was carried out to detect the oxide species of the catalysts (Figure S2).A comparison between the XRD patterns of TiO 2 and Pt/TiO 2 reveals the platinum oxide phases at 2Theta values of 39°, 48°, 67°, and 81°.However, no molybdenum-oxide phase was detected by XRD, showing that there were no crystalline phases of molybdenum oxide in the catalyst structure.The XRD results of HZSM-5 and Ni/HZSM-5 (Ni oxide) samples showed the reflections of NiO at 2Theta values of 37.4°and 43.6°.However, no nickel oxide phase was detected on Ni/HBeta, showing a good dispersion of Ni oxide sites on the HBeta support.The difference between the XRD results of Ni/ HZSM-5 and Ni/HBeta can be due to the catalyst preparation procedure.In contrast to Ni/HZSM-5 which was prepared by impregnating pellets of HZSM-5, Ni/HBeta was prepared by impregnation of HBeta powders by Ni at first and then pelletizing the powder of the Ni/HBeta catalyst.In fact, loading the metal precursors over the powder form of the support causes better penetration of the precursor solution and consequently a better distribution of the metal sites than the pellet support.
Textural properties of the catalysts were analyzed by using the nitrogen isothermal adsorption−desorption technique.Table 1 shows the BET surface area, t-plot micropore surface area, and pore volume of the catalysts.The total pore volume of the supported metal catalysts is almost similar.Although the impregnation of the nickel catalyst on the zeolite supports e Determined by the ICP-MS method.f Negligible.
reduced the surface area, the impregnation of platinum on titania increased the surface area.For TiO 2 and HZSM-5, which were in the pelletized form, the loading of metal oxides slightly increased the micropore surface area, whereas for HBeta, which was in the powder form, loading of nickel oxides reduced the micropore surface area.The amounts of platinum and nickel metals loaded on the catalyst supports of TiO 2 , HZSM-5, and HBeta were determined by the ICP-MS) analysis technique at ALS Scandinavia AB.As shown in Table 1, the contents of Pt and Ni on Pt/TiO 2 , Ni/HZSM-5, and Ni/HBeta were 7.5, 4.3, and 4.6 wt %, respectively.The addition of catalyst and hydrogen to the pyrolysis process reduced the yield of organic FPBO due to the oxygen removal via CO, CO 2 , and water formation resulting in coke deposition on the catalyst and an enhanced yield of hydrocarbon gas.The type of catalyst significantly affected the pyrolysis product composition and yield (Figure 2 and Table 2).A suitable catalyst for biomass hydropyrolysis needs to have high potential to add hydrogen atoms into the oxygenated pyrolyzates and eliminate the oxygen-containing groups.Except for Ni/HZSM-5, rest of the catalysts formed a separate aqueous fraction as a result of water production via hydrodeoxygenation reaction mechanisms.As shown in Figure 2, the Ni catalyst supported on HZSM-5 zeolite showed completely different behavior in upgrading mechanism compared to Ni supported on HBeta zeolite; the Ni/HZSM-5 resulted in only organic FPBO, while the Ni/HBeta formed only aqueous FPBO.In comparison to Ni, the Mo-and Ptbased catalysts seemed more promising to producing FPBOs with lower concentration of reactive oxygenates and improved quality.The decarbonylation and decarboxylation mechanisms, forming CO and CO 2 , respectively, occurred selectively during the stem wood hydropyrolysis over all of the catalysts except for 5Ni/HBeta.Decarboxylation and decarbonylation reactions stabilize the reactive oxygenated intermediates, reducing their polymerization tendency. 21The HBeta-supported nickel catalyst was very selective for light hydrocarbons, mainly methane.The total carbon efficiency of the CHP process, when Ni/HBeta was used as the catalyst, was 79 wt % in which 58 wt % of carbon was recovered via formation of CH 4 with  the HHV of 55.5 kJ g −1 .In a two-stage pyrolysis-catalytic hydrogenation reaction system, used by Jaffar et al., 22 it was also shown that the Ni catalyst supported on Al 2 O 3 was actively selective to methane formation.Although the two Mo/ TiO 2 catalysts resulted in almost similar yields of liquid and gas products (Figure 2), the distribution of chemicals in the organic FPBO was significantly different over the 7Mo/TiO 2 and 10Mo/TiO 2 catalysts (Tables 2 and S2).The lower loading of molybdenum on titania resulted in a better deoxygenation degree of pyrolysis vapors due to the higher surface area and pore volume.7Pt/TiO 2 and 7Mo/TiO 2 resulted in a selective production of phenolics and phenolics together with aromatic HCs (mono-and polycyclic aromatic HCs), respectively.A higher selectivity to oxygen-free compounds was achieved by using the 7Mo/TiO 2 catalyst.Therefore, 7Mo/TiO 2 was selected as the reference catalyst for further study in this work, and from now on, this catalyst will be referred as Mo/TiO 2 .

Pyrolysis Process Optimization for Atmospheric Hydrodeoxygenation.
Ex situ CHP of stem wood using Mo/TiO 2 as the catalyst yielded pyrolysis oil (aqueous and organic fractions), noncondensable gas, char solid residue, and catalytic coke as products.Catalytic activity of Mo/TiO 2 was measured at two temperatures of 350 and 400 °C.The product yields, including those from the reference experiment of noncatalytic pyrolysis, are summarized in Figure 3. Application of catalyst and hydrogen increased the formation of gas products of CO, CO 2 , and C1−C7 hydrocarbons accompanied by a lower formation of organic liquid and the appearance of an aqueous fraction.Lower formation of char was observed while hydrogen was used as the carrier and reactant gas, showing that the hydrogen facilitated thermal decomposition  Energy & Fuels of pyrolysis intermediates.Generally, the presence of hydrogen causes complex hydrotreating reactions which occur simultaneously with the thermal decomposition of the biomass feedstock.The yields of CHP products were almost similar at the two catalysis temperatures tested.However, the distribution of compounds in the FPBO fractions (Table 3) differed considerably while changing the upgrading temperature from 350 to 400 °C, meaning that the catalysis temperature could affect the quality of chemicals produced without changing the mass efficiency of the process.Increasing the upgrading temperature improved the conversion of furans, ketones, and aldehydes as well as the formation of acids and phenols.Sugars were completely transformed in CHP at 400 °C.The hydrogen atoms activated on the catalytic sites were used by the aldehyde and ketone intermediates activated on the adjacent sites to produce their corresponding alcohols.The hydrogenation of the carbonyl group of the aldehydes and ketones occurs via adsorption of these molecules onto the metal catalyst surface and the dissociative adsorption of hydrogen.
Reduction of ketones and aldehydes can occur through either an alkoxy or a hydroxy mechanism.The hydrogen is added to the carbon end and to the oxygen of the carbonyl group to form alkoxy and hydroxyalkyl intermediates via alkoxy and hydroxy pathways, respectively. 23Further hydrogenation of the oxygen of the adsorbed alkoxy intermediate generates alcohols.Moreover, the increase in the catalysis temperature enhanced the efficiency of hydrogenation and hydrodeoxygenation mechanisms, leading to higher selectivity to aromatic hydrocarbons.As shown in Table 3, the acids, furans, and ketones mainly ended in the aqueous fraction of FPBO, which is interesting since showing that the CHP was accompanied by fractional condensation.In the fractional condensation technique, the pyrolysis vapor products are condensed as different chemical groups into separate fractions, while the quality of FPBO is improved by separating the lightest and water-soluble compounds including water, aldehydes, ketones,

Energy & Fuels
and acids from the heavier organic compounds. 24,25However, the aqueous fraction produced via CHP consisted mainly of water with a very low content of carbonous compounds due to the deoxygenation of organic compounds, making them more hydrophobic.In the fractional condensation, the organic fractions (mainly lignin-derived components) had an increased carbon content and a decreased oxygen content compared to the original FPBO, however not to the same extent as in the organic fractions derived in the CHP.In other words, CHP results in a significantly better oil quality than the separation technique of fractional condensation since deoxygenation is needed to chemically reduce the oxygen content and improve the quality of the oil.Water content of the organic oil produced via thermal pyrolysis, CHP at 350 °C, and CHP at 400 °C was 12.3, 1.2, and 0.8 g, respectively.

Catalyst Reusability.
To study the consistency of catalyst activity, time-on-stream (TOS) behavior of Mo/TiO 2 was investigated at 400 °C.Exposing the Mo/TiO 2 catalyst to the pyrolytic vapors for a longer TOS of 180 min (biomass/ catalysts mass ratio of 3) and 270 min (biomass/catalysts mass ratio of 4.5) resulted in constant mass balance (Figure 4), reliable product selectivity (Table 4), and stable gas product streams (Figures S3−S6).As depicted in Figure 4 and Table 4, no significant activity change was observed while prolonging the reaction time from 90 min (biomass/catalyst mass ratio of 1.5) to 270 min.According to the chemical distribution in the FPBOs (Table 4), at longer TOS, less conversion of aldehydes and ketones happened.This is also in accordance with the carbonyl content measurement of the FPBOs listed in Table 5.As the selectivity of gas products during the experiment showed (Figure S3), a reduction in the formation of CO was observed because of the reduced decarbonylation mechanism, while the generation of CO 2 via decarboxylation was almost unchanged.Conversion of guaiacols was also reduced by raising the TOS, showing that the reaction pathways of guaiacol transformation were somewhat hindered.Mechanistically, guaiacol is reacted via different reactions of demethylation, demethoxylation, hydrogenolysis, and alkylation to produce their analogues intermediate/products of catechol, phenol, anisole, and methyl catechol, respectively. 26s shown in the online gas data (Figure S4), methane formation was also decreased by increasing the TOS (and/or biomass to catalyst mass ratio).Generally, among the different transformation mechanisms of guaiacols, a higher activation energy is required for demethylation of guaiacols to their hydrogenated counterparts, such as methane and catechol.Therefore, prolonging the reaction time could cause reduction of catalyst tendency for demethylation, shifting the reaction pathways to produce other intermediates/products like phenols.Catalytic coke generation decreased by increasing the TOS due to the reduced tendency of the catalyst for conversion of problematic intermediates, such as guaiacols,

Energy & Fuels
which are considered as one of the main reasons for coke formation over the catalytic sites during biomass pyrolysis. 27he high contact between catalytic sites and pyrolysis intermediates/reactants causes coke formation on both the surface and pores of the catalyst structure.The coke deposition on the catalyst active sites during the operation resulted in low accessibility of catalytic sites for further reactants.Since the activity of Mo/TiO 2 involves the contacts between catalyst active sites and pyrolyzates, any restructuring caused by carbon deposition will affect the catalysis interfaces.
Representative TEM images of the fresh and spent Mo/ TiO 2 revealed a changed structural appearance of the spent catalyst (Figure S7).The fresh samples had clean surfaces on the TiO 2 grains, while Mo/TiO 2 after 270 min operation showed a layer of amorphous carbon deposited on the surfaces of catalyst grains.The variation of the thickness of the amorphous layer could be related to the dispersion of molybdenum active sites controlling the main reaction mechanisms, leading to coke formation on the surface and in the pores of the catalyst.In addition, a comparison between the STEM images of TiO 2 and Mo/TiO 2 (Figure S8) indicated some addition structure of the Mo/TiO 2 particles, and some of the projected particle surfaces appeared brighter indicating a layer of high-atomic number materials (Mo).
Catalyst lifetime and stability are considered as key factors in the sustainable CHP process of biomass.Hence in this study, the Mo/TiO 2 catalyst used for the CHP of stem wood at 400 °C for 180 min was regenerated and reused in the process for two more times.Textural properties of the regenerated catalysts listed in Table 1 showed that the BET surface area of Mo/TiO 2 did not change after three times regeneration.The XRD patterns of the fresh and regenerated catalysts demonstrated that the crystallinity of the catalyst structure did not change even after three times regeneration (Figure 5).However, an increase in the particle size was observed after reapplication of the catalyst in the CHP process, as shown by the SEM images (Figure S9).The growth of catalyst particles after the CHP reaction could be attributed to slight thermal sintering, which lowers the dispersion degree of metal catalysts, increasing the particle size.In addition, due to the exothermic nature of coke ignition, the heat generated during the regeneration process could create hot spots between catalytic active sites and coke precursors causing particle agglomeration.Figure 6 shows the yields of products obtained from the CHP of stem wood with regenerated catalysts.Reapplication of regenerated Mo/TiO 2 two times did not change the product yield.Water content of the organic oil produced via CHP using the fresh, Reg.1, and Reg.2 catalysts was almost constant, changing in the range of 2.4−2.6 g.The regenerated Mo/TiO 2 retained hydrodeoxygenation activity during CHP, although the guaiacol transformation was slightly lower than that observed with fresh Mo/TiO 2 (Figure 7).Compared to the  fresh catalyst, the selectivity to alcohols in the organic fraction of FPBO was decreased using regenerated catalysts.Selectivity to phenols and monoaromatic HCs over regenerated catalysts was almost similar to that of the fresh one.Formation of polycyclic aromatic hydrocarbons (PAHs) reduced after the first catalyst recycling and then remained constant.
Water, an undesirable product of biomass pyrolysis, is the most abundant component in FPBO.For coprocessing of FPBO in a refinery unit, almost all moisture and most of the organic oxygens must be removed.In fact, less water is favorable for the stability, energy density, and acidity of pyrolysis oil, as well as the transportation costs. 28Water content of the FPBO was reduced almost three times (from 18.81 to 6.14 wt %) via CHP of stem wood at 350 °C.Increasing the catalysis temperature to 400 °C further reduced the water content to 4.07 wt %.Prolonging the reaction time reduced the conversion to water and led to the amount that was close to those of CHP using regenerated catalysts.Overall, the moisture contents of the upgraded oils were significantly low.Compared to the water content (35−45 wt %) of the FPBO obtained from catalytic hydropyrolysis of loblolly pine using NiMo in a high-pressure fluidized bed reactor, 29 very much less water was produced under the reaction conditions used in this study, which can be due to the atmospheric pressure of H 2 that reduced the selectivity to deoxygenation via the dehydration mechanism.Meanwhile, the oxygen content in the organic fraction of FPBO obtained from the CHP at 400 °C was between 13.8 and 19.2 wt %, which was reduced considerably compared to the reference thermal FPBO (43.72 wt %).Compared to the thermal FPBO (HHV of 17.6 kJ kg −1 ), the upgraded oils were enriched in carbon and hydrogen contents, leading to their higher HHV values (28.4−32.5 kJ kg −1 ).The FPBO obtained from CHP at 400 °C for 90 min contained the highest HHV of 32.54 kJ kg −1 , while increasing the TOS to 180 min reduced the HHV for 2.9 kJ kg −1 and further increasing the TOS to 270 min maintained the energy value.Besides, the regenerated Mo/TiO 2 catalysts also kept their high activity for removing organically bonded oxygens.Carbonyl compounds, such as aldehydes and ketones, highly react during FPBO storage resulting in an increase in the average molecular weight and viscosity of the oil. 30Aldehydes and ketones will involve in a number of reactions during FPBO storage: (i) reaction with water to form hydrates, (ii) reaction of aldehydes with alcohols to produce hemiacetals or acetals and water, (iii) reaction of aldehydes with phenolics to generate water and resins, and (iv) reaction of proteins with aldehydes to form oligomers. 30,31 The proposed mild upgrading conditions in this work were sufficient to convert the very reactive compounds of carbonyls to more stable compounds like phenols and monoaromatic hydrocarbons.The content of carbonyls in the organic fraction of FPBO reduced nearly for 52−71% via CHP with the minimum amount of 1.6 mol kg −1 while using fresh Mo/TiO 2 at 400 °C for 90 min.As the results listed in Table 5 show, increasing the TOS and regenerating/recycling the catalyst led to almost similar total contents of carbonyls in the upgraded FPBO.
The closure of the carbon balance showed that the application of the catalyst and hydrogen proceeded the reaction in a way that a considerable part of the carbon in the feedstock ended in the gas products (Table 5).Hydrogenation reactions facilitated the formation of noncondensable light hydrocarbons, reducing the carbon recovery by FPBOs.As inferred from the results listed in Table 5, the higher activity of the catalyst resulted in lower carbon recovery in the FPBOs.The total carbon efficiency of the proposed system was about 70−86 wt %.Reduction of the carbon recovery using the catalyst could be due to the formation of gas products not detected by gas analyzers.A better carbon recovery via gas products caused a better total carbon efficiency, indicating that the carbon loss could be mainly due to the gas products that cannot be detected by the analysis instruments.The regenerated Mo/TiO 2 catalysts produced FPBOs with almost similar carbon efficiency as the fresh Mo/ TiO 2 .The highest carbon recovery in FPBO (38.86 wt %) was observed when stem wood was hydropyrolyzed over the fresh Mo/TiO 2 at 400 °C for 180 min.Our data in carbon recovery in condensable products (29−39 wt %) are in accordance with those (29.2−32.2wt %) achieved by Stummann et al. 32 in the catalytic hydropyrolysis of beech wood using a sulfided Mo catalyst in a fluidized bed pyrolyzer.In a study held by Yung et al. 33 38 wt % of the carbon in biomass feedstock was recovered in the oil produced from CHP of the mixture of pine and forest residue (50/50 w/w) using the Pt catalyst supported on TiO 2 .Compared to the expensive noble Pt catalyst, which is generally a very efficient hydrogenation catalyst, the equal catalysis performance of the transition metal of Mo supported on TiO 2 in carbon recovery could be a valuable achievement.
Analysis of functional groups in the FPBOs via NMR methods provides detailed information about the FPBO's structural properties.The main advantage of NMR analysis to GC−MS is that the information about the whole functional groups in the whole FPBO will be obtained by NMR independent of the volatility of the components, which is a limiting function for GC−MS.Hence, 31 P NMR analyses of thermal and upgraded FPBOs were carried out to quantitatively determine the selectivity to hydroxyl functional groups (Table 5).The total hydroxyl content in thermal FPBO (7.2 mmol g −1 ) was higher than the total hydroxyl contents in the FPBOs obtained via CHP.Catalytic hydropyrolysis of stem wood significantly converted the aliphatic alcohols (from 3.8 mmol g −1 in thermal FPBO to 0.5−0.7 mmol g −1 in upgraded FPBOs).On the contrary, the selectivity to phenolic hydroxyls enhanced from 2.7 mmol g −1 in the thermal FPBO to 3.0−3.4mmol g −1 in the upgraded FPBOs via CHP.These data are in accordance with those determined by the GC−MS/FID method.Moreover, 13 C NMR analysis was performed to investigate the effect of the catalyst and hydrogen on the structural H−C framework of FPBO (Figure 8).The C NMR integration results revealed that the chemicals belonging to the aldehydes, ketones, carboxylic acids, esters, aliphatic alcohols, ethers, carbohydrates, and methoxy functional groups converted significantly to their hydrogenated counterparts via CHP process.This significant change was accompanied by an increase in the selectivity to the components of aromatics,  Energy & Fuels alkenes, and aliphatics, resulting in a lower oxygen and a higher carbon content in the upgraded FPBOs.
The Van Krevelen plot (H/C atomic ratio vs O/C atomic ratio; Figure 9) was provided to investigate the efficiency of different reaction conditions on the quality of FPBO obtained via CHP of stem wood.The H/C ratio displays the compounds consistent with their saturation degree, while the O/C ratio represents the oxygen content.A higher H/C ratio accompanied by a lower O/C ratio is desired for FPBO as fuel chemicals.In comparison to the thermal FPBO, the upgraded FPBO had lower H/C and O/C ratios which could be due to the significant structure change of pyrolysis oil, as well as dehydration, decarboxylation, and decarbonylation reactions happened during the CHP.The higher catalytic activity resulted in lower H/C and O/C ratios showing that one of the main deoxygenation pathways was dehydration. 34The decrease in the H/C and O/C ratios could be also due to the aromatization tendency of the Mo/TiO 2 catalyst proceeded via dehydrogenation and deoxygenation mechanisms (C NMR data in Figure 8).Mechanistically, a catalyst with a lower aromatization activity results in organic compounds with higher hydrogen atoms.Even though the thermal FPBO had a high value of H/C ratio, its high O/C ratio causes a low HHV as compared to the upgraded FPBO.The hydrogen-to-carbon ratio of the upgraded FPBOs is lower than those of the conventional fossil fuels, like diesel with H/C molar ratio of 1.8, indicating that the upgraded oils need to be mixed with fossil fuels before being processed in the refinery units.
3.4.FCC Upgrading.The results from the FCC coprocessing were characterized by the qualitative assessment of the produced liquid product in terms of using the FPBO as a feedstock for existing biorefineries.Both thermal and upgraded FPBOs were coprocessed with the fossil commercial FCC  feedstock at equivalent conditions to assess the effect of FPBO pretreatment on FCC-upgraded liquid product quality.To get valuable results, a rather high cofeeding ratio of 20 wt % for FPBO was chosen in these experiments.Compared to the thermal FPBO, the upgraded oil was very easily fed into the FCC system due to its better quality, considering the lower viscosity and lower content of oxygenated components.In fact, one of the main issues in coprocessing a bio-oil into the current refinery units is the difficulty in feeding the biobased oil due to its high viscosity and the high content of oxygenated compounds, which lead to system clogging.Therefore, coprocessing of FPBO in the FCC unit is technically feasible.The simulated distillation of the liquid products from the FCC process was carried out to determine the weight fractions for different distillation ranges, as presented in Figure 10.Our results demonstrated that coprocessing of the FPBO had a significant effect on the conversion to light product fractions with an observed increase of unconverted bottom fraction in the upgraded product.Moreover, the cracking of both thermal and upgraded FPBOs together with fossil feed produced more paraffins, olefins, and napthenes.While thermal FPBO was more selective toward napthenes, the upgraded FBPO was significantly selective toward paraffins and olefins (Figure 11).At the same time, the formation of aromatics, as the main chemical fraction in the FCC feed product, was reduced by the copresence of FPBO, which could be due to the side reactions caused by the presence of FPBO hindering the aromatization efficiency of the catalyst.The distribution of simulated mass fractions of FCC fossil feedstock and FCC-upgraded liquid products indicated an increase in the bottom fraction (boiling point higher that 360 °C) for 2 and 2.6 times, from 12 wt % to 24 and 31 wt %, while coprocessing thermal and upgraded FPBO, respectively.In a study held by Pinho et al., 35 the bottom fraction yield from FCC of VGO was in the range of 12.5−16.6wt % for the conversion rate of 63.5−70.5 wt %, while the lower conversion corresponded to the higher bottom fraction.As shown by Pinho et al. 35 and Wang et al., 9 coprocessing of 5−10 wt % of FPBO from pine woodchips and beech wood, respectively, with VGO did not change the yields of the bottom fraction, while our results revealed that the coprocessing of high concentration of thermal and upgraded FPBOs (20 wt %) from stem wood increased the distribution of the bottom fractions.Considering the selectivity of compounds in the oil products of the FCC process (Figure 11), this high molecular weight fraction (bottom fraction and LCO fraction) consisted mainly of long chain paraffins and olefins.Distribution of the components in the oil products of the FCC process revealed that the presence of FPBO somehow facilitated the cracking performance, resulting in a lower formation of heavy molecular weight chemicals that cannot be detected by GC-MS.This can be attributed to the improved reaction tendency of the intermediates of the FCC feed and FPBO.
The comparison between pure fossil feed product yields and the bio-oil cofeed products differ from the ones obtained in previously mentioned scientific papers; 6,9 in this work, the opposite effect of coprocessing FPBOs was observed in terms of decrease in lighter product fraction conversion.However, a similar component selectivity trend regarding olefin and aromatics is recognized in a study held by Gueudréet al. 10 The results from comparing the different bio-oils cofeeding in terms of the grade of pretreatment showed that the liquid products are similar in both yield and quality, while the component of liquid product obtained from a higher pretreatment grade resulted in a richer component selectivity of primary petrochemical products, such as olefins.

CONCLUSIONS
In order to improve the quality of stem wood pyrolysis oil for coprocessing in the FCC refinery unit, one-step upgrading of stem wood pyrolysis vapors was carried out in a continuous drop tube pyrolyzer coupled with a fixed-bed catalytic reactor (ex situ configuration) under atmospheric hydrogen.The performance and efficiency of different catalysts were investigated in the CHP of stem wood.Experimental results showed that the application of different catalysts significantly affected the product composition and yields.Ni-based catalysts were strongly active, tending to generate a high yield of gas (mainly light hydrocarbons), while Mo-and Pt-based catalysts seemed better for production of liquid with improved quality.The carrier material of the metal catalysts also affected the results obtained from the hydropyrolysis of stem wood; HBeta zeolite resulted in higher yields of gas products than HZSM-5 zeolite and TiO 2 .The quality of FPBO significantly improved with atmospheric catalytic hydropyrolysis using Mo/TiO 2 while reducing the formation of reactive oxygenates such as aldehydes and ketones.The composition of oil from CHP also showed that the selectivity to phenols and aromatic hydrocarbons was relatively high, indicating that the proposed system had the potential to extract chemicals with higher heating value than the conventional composition of FPBO.The 29−39 wt % carbon recovery in the condensables was considerably high and reasonable.The Mo/TiO 2 catalyst kept high hydrodeoxygenation activity by increasing the TOS (90− 270 min), and almost no change in the activity of Mo/TiO 2 was observed after its regeneration and reapplication in the CHP process for two times.The stability is one of the key quality standards for FPBO in FCC coprocessing.CHP leads to an upgraded FPBO with considerably lower contents of water and carbonyls compared to the thermal pyrolysis, increasing the chance of coprocessing in the FCC unit.Therefore, the upgrading technique of ex situ catalytic hydropyrolysis could be a benefit to refiners, improving the chance of coprocessing a biobased oil with fossil oil in the FCC unit and offering them options to meet the renewable fuel standards and other regulatory requirements.Cofeeding the upgraded FPBO from stem wood with VGO fossil oil (20 wt % of FPBO) into the FCC unit significantly increased the formation of paraffins and olefins, while it had a negative trend in the formation of isoparaffins and aromatic hydrocarbons.
Proximate and ultimate properties of pyrolysis feedstock; chemicals detected by GC-FID in the organic oil fractions; SEM-EDS analysis of catalysts; XRD analysis of catalysts; consistency of gas products in catalyst stability tests; TEM and STEM images of catalysts; and SEM-STEM images of catalysts (PDF) Energy & Fuels

Figure 1 .
Figure 1.Schematic diagram of the continuous bench-scale drop tube pyrolyzer coupled with a fixed-bed catalytic reactor.

Figure 2 .
Figure 2. Effect of catalyst type on the yields of gas and liquid pyrolysis products.

Figure 5 .
Figure 5. XRD patterns of the fresh and regenerated Mo/TiO 2 catalyst.

Figure 9 .
Figure 9. Van Krevelen diagram plotting the atomic H/C ratio against atomic O/C ratio of the organic FPBOs.

Figure 11 .
Figure 11.Selectivity of components of FCC-upgraded liquid products obtained from FCC fossil feedstock, coprocessing of thermal FPBO, and coprocessing of upgraded FPBO.

Table 1 .
Textural Properties of Catalysts Analyzed by N 2 Adsorption−Desorption and ICP-MS Calculated in the range of relative pressure (P/P 0 ) of 0.01−0.20.b Evaluated by the t-plot method.c Total volume of pores less than 109 nm diameter at P/P 0 a

Table 2 .
Effect of Catalyst Type on the Selectivity of Pyrolysis Products in Organic Oil Fractions

Table 3 .
Selectivity of Components of FPBO Obtained from Thermal Pyrolysis and CHP of Stem Wood a T pyrolysis = 520 °C, T catalysis = 350 and 400 °C, WHSV in CHP experiments = 1 h −1 , and reaction time = 90 min.
a Reaction conditions:

Table 4 .
Effect of TOS on the Selectivity of Components of FPBO Obtained from Thermal Pyrolysis and CHP of Stem Wood a a Reaction conditions: T pyrolysis = 520 °C, T catalysis = 400 °C, WHSV in CHP experiments = 1 h −1 , and TOS = 90−270 min.