Bei posters tcbiomass2013 pdfs by Bioeconomy Institute

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

September 2013, Chicago

Xianglan Bai*, Robert C. Brown, Jie Fu, Brent H. Shanks, Matthew Kieffer

The Role of Acid Pretreatment in Solvolysis of Switchgrass 14

Using water as co-solvent improved the yield of xylose compared to pure 1, 4dioxane system.

1 0

100

150 Time (s)

200

250

Xylose

100

150 Time (s)

200

300 Deg. C

350 Deg. C

0 0

100 Time (s)

150

200

250

100

150 Time (s)

200

250

 Thermal stability of the sugars decreases with increasing temperature. The yield of total sugar monomers is lower at higher temperature after stabilization due to decomposition.

acid infused

1. 1-methoxy-2-Propanone 3. 2-vinylethyl acetate 5. 1,2-Ethanediol diformate 7. 2,3-dihydroxy-propanal 9. Glycerin 11. Levoglucosan

acid infused

2 0

100

200

300

400

100

Time (s) 8

300

400

16 14

acid washed acid infused

Yield (wt%)

untreated

acid infused

8 4 2 0

100

200 Time (s)

300

400

100

200

300

400

Time (s)

Both glucose and levoglucosan are thermally stable in the mixture solvents compared to water as the single solvent.

Levoglucosan

15 10

1, 4-dioxane water water/1, 4-dioxane

Glucose

6 1, 4-dioxane water water/1, 4-dioxane

4 2

4 1. 1-methoxy-2-Propanone 2. 1,2-Ethanediol monoformate 3. Tetrahydro-3-Furanol 4. Levoglucosan

100

150 Time (s)

200

250

Xylose

100

150 Time (s)

200

Glucose

1, 4-dioxane water water/dioxane

The presence of AAEM did not significantly affect the deconstruction of polysaccharides during solvolysis. The presence of solvent appears to mitigate the catalytic effect of AAEM. The yields of sugars increased significantly for acid infused switchgrass with acid acting as a strong catalyst for the depolymerization and hydrolysis of polysaccharides. Acid also catalyzes dehydration reactions that lead to the formation of furfural, hydroxymethylfurfural, levoglucosenone, etc.

 Levoglucosan is the major decomposition product of glucose.  Cellulose and levoglucosan can both hydrolyze in the presence of water to glucose, while glucose can dehydrate to levoglucosan when 1, 4-dioxane is the solvent.  Conversion between glucose and levoglucosan, although not completely reversible, could greatly improve stability of levoglucosan and glucose in 1, 4-dioxane.

 The choice of solvent impacts not Levoglucosan only primary products but also secondary products of solvolysis.

Conclusions

 Although AAEM is known to catalyze ring scission of pyranose and furanose in cellulose and hemicellulose during fast pyrolysis, the presence of AAEM did not significantly affect the deconstruction of polysaccharides during solvolysis.

4 3 2

 A small amount of acid infused into biomass prior to solvolysis strongly catalyzes the depolymerization of polysaccharides.

1 0

100

150 Time (s)

Financial support from National Advanced Biofuels Consortium is gratefully acknowledged. The authors like to thank Ngoc Phan for solvolysis experiments and Kwang Ho Kim for assistant in GC/MS analysis

200

250

 Levoglucosan is relatively stable to thermal decomposition.

Time (min) 9

250

Levoglucosan+Glucose+Xylose

1.1-methoxy-2-Propanone 2. Furfural 3. 2,3-dihydroxy- Propanone 4. 1,3-dihydroxy- 2-Propanone 5. Glycerin

Partial hydrolysis of levoglucosan to glucose is observed when water is co-solvent.

Glucose formed is extremely unstable in water medium.

acid washed

Fresh switchgrass contains up to 60% of water. While levoglucosan is the major sugar when 1, 4-dioxane is the solvent, glucose becomes the major sugar when water is the solvent.

Levoglucosan + Glucose + Xylose

Xylose

200 Time (s)

2. Acetic anhydride 4. Furfural 6. Tetrahydro-3-furanol 8. 1,3-dihydroxy-2-Propanone 10. 5-hydroxymethyl furfural

9 7 345 2 6

200

GC/MS chromatogram of decomposition products

acid washed

Pressurized water is known to be an excellent solvent for solvolysis reactions.

100 150 Time (s)

Degradation of Monomeric Sugars in 1, 4-dioxane

Yield (wt%)

acid washed

Effect of Water as Co-solvent on Sugar Production from Acid-infused Switchgrass

350 Deg. C

Higher reaction temperature leads to higher maximum yields of levoglucosan, glucose and total sugar monomers, but lower maximum yield of xylose.

Glucose

300 Deg. C

 Higher maximum yield of total sugars is obtained from the solvent mixture.

1, 4-dioxane water water/1, 4-dioxane

Thermal stability of the sugars improved greatly in the solvent mixtures compared to water alone.

Levoglucosan + Glucose + Xylose

Yield (%)

250

20 Yield (wt%)

Yield (wt%)

untreated

Yield (wt%)

 Higher maximum yield of xylose is produced when water is the solvent. However, xylose rapidly degrades.

untreated

Yield (wt%)

350 Deg. C

Sugar Production Under Different Acid Pretreatments Using 1, 4-dioxane as Solvent

350 Deg. C

Reactors: assembled with two, 3/8 inch Swagelok cap and connector Heating mechanism: Molten tin bath

300 Deg. C

 Apparatus

300 Deg. C

Xylose

washed/water rinsed (AWWR) switchgrass, acid-infused (AI) switchgrass (with 2wt% of sulfuric acid), levoglucosan, xylose, glucose Solvents: 1, 4-dioxane, water, and the mixture of 1, 4-dioxane and water (9: 1 ratio)

Yield (wt%)

 Materials Feedstocks: Untreated switchgrass, acid-

Glucose

Levoglucosan

(a)

Experimental Methods

12 Yield (wt%)

Thermochemical conversion technologies like fast pyrolysis and solvolysis are able to rapidly depolymerize whole lignocellulosic feedstocks into liquid products which include sugars, phenols. Both are attractive as deconstruction processes for the production of fuels. While pyrolysis of untreated biomass usually yields very little sugars, both acid washing/water rinse and acid infusion have been observed to dramatically increase the production of anhydrosugars during fast pyrolysis of lignocellulosic biomass, a phenomenon attributed to passivation of alkali and alkaline earth metals (AAEM) that otherwise catalyze scission of pyranose and furanose rings in plant polysaccharides. However, problems like char agglomeration during fast pyrolysis of acid infused biomass, and the tendency of levoglucosan to polymerize and subsequently dehydrate to char may limit the potential for maximizing sugar production. We hypothesize that solvolysis of biomass may be able to migrate these problems and therefore improve sugar production.

Effect of Water as Co-solvent (continued)

Effect of Temperature on Solvolysis of Acid Infused Switchgrass in 1, 4-dioxane

Yield (wt%)

Introduction

The sugars produced were more stable in a mixture of 1, 4-dioxane and water compared to pure water, resulting in a maximum yield of total sugars of 19.8wt%.

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies Karl M. Broer (kbroer@iastate.edu), Patrick Johnston, Patrick Woolcock, Robert C. Brown

Biomass fuel bound nitrogen (FBN) conversion in two fluidized bed gasifiers

Background

Results for tests in the 20 kg/h gasifier

0.35

0.40

0.45

Ammonia Cyanide

0.25

0.30

0.35

0.40

0.45

Figure 5. Percent yield and concentration data for NH3 and HCN in syngas produced by the 20 kg/h gasifier in response to changes in equivalence ratio.

Least Desirable

Char Bound Nitrogen Tar Bound Nitrogen

Figure 1. The five major nitrogen bearing products of gasification.

The concentrations of nitrogen in feedstocks can vary by nearly two orders of magnitude [1], leading to NH3 and HCN concentrations in syngas which also vary widely [2-7]. Equilibrium modeling using software programs such as STANJAN demonstrate that NH3 and HCN should be minor products of gasification, yet for typical gasification conditions, well over half of the FBN is converted to NH3 or HCN.

Materials and Methods

0.30

Two fluidized bed gasifiers were used in these tests: A pilot-scale 20 kg/h gasifier and a laboratory-scale 0.1 kg/h gasifier. The large gasifier was designed to approximate adiabatic operation and used a steam-oxygen mixture (36% oxygen) as fluidizing agent. The equivalence ratio (ER) was varied to explore its affect on NH3 and HCN yields. Gas samples were extracted from the large gasifier via an iso-kinetic sampling line (Figure 2). Char was removed via a hot ceramic filter. Tars were removed for gravimetric analysis via a dry condensing method. Liquid samples of nitrogen compounds were collected using glass impingers. 5% HCl was used to collect NH3. 100 mM NaOH was used to collect HCN.

Figure 2. Sample line setup for measurement of char, heavy tar, water, major permanent gas, and nitrogen compound composition of the syngas from the 20 kg/h gasifier. Gas sampling on the small reactor was conducted in a similar manner, but using an electrostatic precipitator instead of a tube furnace and dry condenser.

The small gasifier was operated allothermally (ER of zero). The short residence time of this reactor (1.2 s) allowed us to investigate the primary products of FBN devolatilization. Tars were removed via an electrostatic precipitator instead of a pressure cooker. Due to the reactor’s small size, all of the syngas produced by the reactor was bubbled through the impingers, rather than using a slip stream. A drumtype gas meter was used for both small and large scale tests to measure the total flowrate through the sampling lines. Depending on the anticipated amount of NH3 and HCN, two to four impingers were used in series. Gas was allowed to flow through the impingers at 1.0 SLPM for 30 min per sampling session (Figure 3). After sample collection, the resulting NH4+ solution was analyzed via distillation and titrimetric analysis. The CNsolution was analyzed via Ion Chromatograph (Figure 4).

Results for tests in the 0.1 kg/h gasifier

Allothermal gasification in the small gasifier revealed that it is possible for HCN concentrations to exceed NH3 concentrations (Figure 6). Additionally, the total amount of NOX precursors is lower by a factor of about five compared to the large gasifier. Because the small gasifier tests were conducted without addition of limestone, it is not yet certain whether this reversal is caused by the anoxic (ER=0) operating condition or the absence of limestone, which might play a catalytic role. 14 12 10 8 6 4 2 0

Concentration (ppm, dry basis)

Biomass FBN

0.25

12000 10000 8000 6000 4000 2000 0 0.20

Equivalence Ratio

Ammonia Hydrogen Cyanide

70 60 50 40 30 20 10 0 0.20

Concentration (ppm, dry basis)

Diatomic Nitrogen (N2)

Yield (% of Fuel Bound Nitrogen)

Diatomic Nitrogen (N2) in Oxidizing Agents

Ammonia was higher in concentration than HCN for all tests in the large gasifier; however, HCN was still present at much higher concentrations than presented by other researchers [2-7] (Figure 5). Yield (% of Fuel Bound Nitrogen)

When biomass is gasified, its fuel bound nitrogen (FBN) converts into five major forms (N2 contained in the gasifying agent(s) is not reactive at fluidized bed gasification temperatures) (Figure 1). The ammonia (NH3) and hydrogen cyanide (HCN) in the syngas are of particular interest because they are NOX precursors if the syngas is combusted. They are catalyst poisons if the syngas is to be utilized for chemical or syn-fuel products. It is of interest to determine gasifier operating conditions that minimize conversion of FBN to NH3 and CHN.

600

650

700

750

800

850

900

600 500 400 300 200 100 0

Ammonia Cyanide

600

650

700

750

800

850

900

Operating Temperature, °C in syngas produced by Figure 6. Percent yield and concentration data for NH3 and HCN the 100 g/h gasifier in response to changes in operating temperature.

Conclusions and Future Plans

Figure 3. Collecting liquid samples of nitrogen compounds using glass impingers downstream of tar condensing pressure cooker. Works Cited

1. Brown, R.C. Biorenewable resources: Engineering new products from agriculture. 2003. p.67. 2. Jong, W., et al. Thermochemical conversion of brown coal and biomass in a pressurised fluidised bed gasifier with hot gas filtration using ceramic channel filters: measurements and gasifier modeling. Applied Energy 74 2003. p 425-437. 3. de Jong, W. Nitrogen compounds in pressurised fluidised bed gasfication of biomass and fossil fuels. PhD Dissertation, Technische Universiteit Delft, 2005.

Acknowledgements: Alex Haag, Jacob Broer, Tannon Daugaard, and Nate Hamlett

Figure 4. Preparing CN- samples for Ion Chromatograph analysis.

• Our research confirms that for typical operating conditions for industrial gasifiers, NH3 is the most important syngas nitrogen species [2-7]. However, we also found that HCN is a significant product, especially for herbaceous energy crops and waste feedstocks, which have far more FBN than woody feedstocks. • Certain operating conditions, like those carried out in the small gasifier, can lead to yields of HCN that are greater than for NH3. This contrasts with reports in the literature that indicate NH3 is always the predominant nitrogen compound in syngas. • Further work is needed to determine the mechanisms responsible for variations in the relative amounts of NH3 and HCN in syngas.

4. Goldschmidt et al. Ammonia formation and NOX emissions with various biomass and waste fuels at the Varnamo 18 MWth IGCC plant. 2001. 5. Kurkela, E., Laatikainen-Luntama, J., Stahlberg, P., Moilanen, A. 1996. Pressurised fluidised-bed gasification experiments with biomass, peat and coal at VTT in 1991-1994 Part 3. Gasification of Danish wheat straw and coal. p. 28 6. Zhou, J., Masutani, S.M., Ishimura, D.M., Turn, S.Q., and Kinoshita, C.M. Release of fuel-bound nitrogen during biomass gasification. Ind. Eng. Chem. Res 2000., p. 630 7. Yu, Q-Z., Brage, C., Chen, G-X., Sjostrom, K. The fate of fuel-nitrogen during gasification of biomass in a pressurised fluidised bed gasfier. Fuel. 2007., p. 611-618.

Bioeconomy Institute, Iowa State University

tcbiomass2013

Online Access

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013/

Nicholas Creager, Lysle Whitmer, Song-Charng Kong, and Robert Brown

High Pressure, Oxygen Blown, Entrained Flow Gasification of Bio-oil Project Overview:

Gasification at pressure has advantages for downstream power production and fuel synthesis. However, conveying solid biomass into a pressurized reaction vessel adds expense and reduces system reliability. We are investigating conversion of biomass into bio-oil in an atmospheric pressure pyrolyzer followed by injection of the bio-oil into a pressurized gasifier, which overcomes some of the problems of directly gasifying raw biomass. The bio-oil also has reduced sulfur, nitrogen, and ash content, which reduces subsequent cleaning of the syngas product.

Gasification System

1 7

The gasification vessel incorporates a novel design to allow near adiabatic gasification at small scale. A silicon carbide tube is wrapped with a high temperature cable heater, which provides reliable heat for start up. The heater and silicon carbide are wrapped with high density insulation to protect the pressure vessel from the extreme gasification conditions. This heated section is contained within a pressure vessel. This section is followed by a quench vessel to reduce gas temperature before the pressure is let down. The Watlow controllers inside the control panel provide continuous monitoring of multiple sensors and functional components required to operate the reactor

Experiments were preformed using methanol as a control fuel for the liquids gasifier to prove reliability and consistency of the equipment and micro gas-chromatograph (Micro GC) for analyzing the produced syngas. All systems were operational and high quality results for methanol gasification were recorded as shown below.

Temperature Fluctuations:

1) 2) 3) 4) 5) 6) 7)

Bio-oil Mixing Duel Syringe Pumps Oxygen Supply De-ionized water supply Steam Generator Gasifier Water Quench

A sharp temperature gradient is observed along the axis of the gasification vessel. Very high temperatures (>1000 C) are observed near the oxygen-assisted atomization nozzle (see TC 2 and 3) where locally high equivalence ratios drive combustion reactions. As the reactions consume the available oxygen, temperature drops as endothermic gasification reactions brings the syngas toward equilibrium composition.

8 Liquid Flow Gas Flow Low Pressure Line High Pressure Line

8) 9) 10) 11) 12) 13)

Liquid Throttling Valve Water Drain Gas Throttling Valve Micro GC Flare/Vent Watlow Control Panel

Operational Goals:

The goal of this research is to compare syngas from conventional biomass gasification with syngas from bio-oil gasification. The bio-oil gasifier is designed to operate at pressures approaching 50 atmospheres and heater temperatures up to 1000oC as well as with the ability to inject compressed air, pure oxygen or any mixture of the two. The combination of these parameters along with internal heating capability, give the reactor immense flexibility in operation, including a wide range of equivalence ratios and high-pressure/low temperature gasification reactions needed to understand the fundamentals of pressure effects on syngas production.

Acknowledgements: CSET colleagues

Baseline testing revealed variability in syngas composition between bio-oil gasification trials. The table demonstrates the importance of proper fuel atomization and its effect on gas composition during development of the nozzle configuration. Nozzles one and two developed agglomerates at the end of the nozzle due to poor atomization. Nozzle three mixes the oil with the oxygen before it is introduced to the high temperature environment, reducing the chance for clogging and increasing the oxygen's interaction with the bio-oil.

Nearly 12% of the dry bio-oil mass was being converted to insoluble tars indicating further optimization is needed. Bio-oil homogeneity is also expected to play an important role in both syngas composition and conversion efficiency. These tests were run with whole BTG pine bio-oil due to its relatively clean and consistent properties.

Bio-oil Gasification Micro GC Data in Mole % Equivalence ratio (ø) = 25% Heater Temperature 850oC Atmospheric Pressure Nozzle Nozzle Nozzle Config 1 Config 2 Config 3 H2 23.9% 22.9% 18.6% CO 39.6% 34.1% 41.1% CH4 6.1% 7.7% 7.6% CO2 26.9% 31.4% 28.2% C2H6 0.3% 0.2% 0.4% C2H4 1.7% 2.1% 2.6%

Operational Verification with Model Fuel:

2 5

Bio-oil Gasification Data and Results:

Reactor Design:

Atmospheric Bio-oil Gasification Mass Balance Reactants Products Bio-oil Oxygen Syngas Water Tar Mass (g) 22 9 22.6 3.7 3.6 % Mass 72.0% 28.0% 75.7% 12.4% 11.9% Recovered Mass 97% Total Reactants Mass (g) 30.8

Hypothesis: At elevated pressures H2 formation will decrease in favor of CH4 Bio-oil Gasification Micro GC Data in Mole % Heater Temperature ø = 25% 850oC 1 atmg 6.8 atmg H2 23.4% 31.3% CO 44.6% 40.6% CH4 8.0% 9.2% CO2 21.1% 18.2% C2H6 0.4% 0.2% C2H4 2.0% 0.3%

Bio-oil Gasification Parameters:

All runs on the high pressure reactor were performed with the heater set at 850oC and at a gasification equivalence ratio of 25% Oxygen. The liquid feedstock was injected at a rate of 10 mL/min. Compressed nitrogen gas was used as a tracer at 1.0 SLPM in order to calculate the volume of syngas produced. Syngas was sampled for gas analysis before being tested for flammability with an online burner (see photo).

Two exploratory tests were preformed at elevated pressure to check the validity of the above hypothesis. Despite the variability in syngas composition observed previously, it was observed that hydrogen production increased with pressure while no changes in CH4 concentration were apparent as depicted in the table to the left. Further testing is needed to validate these results and further explore pressure effects. Additional tests will be conducted at pressures up to 50 bar to identify any optimum operating pressures for the production of H2 from bio-oil and the following hypothesis has been proposed:

Revised Hypothesis: H2 conversion from bio-oil will increase with relatively low pressure increases (at least < 6.8atmg) before favoring CH4 production at high pressures.

Next Steps: • •

Duplicate gasification runs with bio-oil under pressure Complete a response surface methodology study exploring reaction temperatures and pressures with respect to syngas composition

References: Lin, S.-Y., Brown, R.C, Catalytic Production of Ethanol from Biomass-Derived Synthesis Gas, Overcoming Recalcitrance of Cellulosic Biomass, 2008

Funded by: Iowa Energy Center

Bioeconomy Institute, Iowa State University

tcbiomass2013

Download it!

Center for Sustainable Environmental Technologies Bernardo G. del Campo, Robert C. Brown & Max J. Morris

OPTIMIZATION OF THE PRODUCTION OF ACTIVATED CARBON FROM FAST PYROLYSIS CHAR Reactor design and operation

Introduction Fast pyrolysis of biomass yields approximately 10-20% biochar by mass, which currently is considered a low-value co-product. Through further activation techniques, biochar can be converted into activated carbon (AC). The yield and quality of AC is a result of several different parameters such as: temperature and pressure, gas type and flow rates, feedstock type and pretreatment, heating rates and holding times, etc.

Operating parameters: - Temperature: 400-800°C - Residence time: 5-60min - Steam flow: 1 mL/g BC/min - Pressure: 0-3 PSI - Sample size 1-2 g - Extra port for adding gases - 2 Temperature controllers

In order to efficiently investigate activation methods with a large amount of operating conditions other more sophisticated statistical methods such as Response Surface Methodology (RSM) were explored.

Figure: Predictor profiler and desirability functions for surface area and burn off

Figure: Batch reactor used for steam activation

• Produce Activated carbon from red oak fast pyrolysis biochar. • Use Response Surface Methodology to optimize production, performance and economics of the process.

Materials and Methods

Feedstock: Red Oak (RO) fast pyrolysis biochar was produced in a fluidized bed reactor (~500°C and ~20s residence time). Material pretreatment: methanol was used to wash RO at rate of 8:1 v/v (methanol to biochar) to remove remaining bio-oil in the sample. Acid wash with 0.1 M of sulfuric acid followed, at a rate of 10:1 v/v (acid solution:biochar) and dried at 105°C for 48 hrs. Experimental design: A complete factorial design with 3 temperature levels by three levels of activation times with 3 replications was performed. Biochar Activation: Roughly 1 gram of biochar was activated in every run. Steam continuously flushed the sample at 1mL/min/gr for the designed temperature and residence time. Physi-sorption analysis: 0.1g of sample was degassed for 4 hours at 300°C with a vacuum reaching at least 100 Pa. Afterwards, samples were measured for BET surface area. Promising biochar-activated carbons were analyzed for the volume of micropores based on v-t plot, and pore size distribution based on Quenched Solid Functional Theory for disordered carbonaceous materials. Activated and non activated biochars present a type II isotherm as described by Brunauer, Deming and Teller. Non-activated fast pyrolysis chars present very low surface area typically less than 10m2/g and mostly representing external surface area.

Energy Center

Figure: RSM for burn off rate with different temperatures and residence time

Results and Discussion Surface Area m2/g

Objectives

Funding and support provided by the Iowa

R2=82.5%

The maximum desirability function for surface area. and burn off resulted in 365m2/g and 18%, which was predicted for 600°C and 60 min of residence time.

15 10

R2=86.6%

5 0

Corn Stover Loblolly Pine Red Oak

Switchgrass

Sand

Soil

Biomass R.O.

Figure: BET surface area for non activated Fast Pyrolysis Biochar

Note that commercial activated carbon typically range from 500 to 1500 m2/g R2=84.2%

Table: Surface area model; estimated parameters and level of significance Term Estimate Std Error Intercept 370.5 38.1 Temp 176.4 21.7 Res. Time -6.3 22.9 Temp*R.Time -69.6 26.3 Temp2 -84.5 38.6 Res. Time2 1.4 36.9

t Ratio Prob>|t| 9.72 <.0001 8.14 <.0001 -0.27 0.788 -2.64 0.018 -2.19 0.044 0.04 0.971

Figure: RSM for biochar surface area for different temperatures and holding times. The predicted model closely fits the experimental data (R2=84%), with significant quadratic term for temperature and interaction between residence time and temperature. The overall model is significant p< 0.0001 and level of significant for the lack of fit test is p=0.26. The quadratic model is appropriate and predicts very well the experimental data.

Figure: Predictor profiler and desirability functions for gross income $/ton of biochar Figure: Gross income $/ton of biochar for the different treatment combinations Based on hypothetical value, of $4 per m2/g for commercial activated carbon, an optimum region can be found for maximizing income. Highest gross income for activating biochar was reached around 800°C and 5 min of residence time. At this condition, approximately $1540 can be gained for every ton of activated biochar.

Conclusion • Low cost adsorbents can be produced by steam activation of fast pyrolysis char. • Red oak chars were converted into activated carbon (>500 m2/g) • The use of RSM helped to optimize activation parameters and experimental work (by decreasing runs and replication) and optimizing process economics.

Future Work • Chemisorption and functional attributes should be studied to identify different uses and applications of activated biochars.

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013/

Preston Gable, Patrick Johnston, Lysle Whitmer and Robert Brown

Time-Resolved Sampling of Pyrolysis Vapors and Aerosols from a Free-fall Reactor Char Collection

Pyrolysis of biomass yields a large variety of volatile and nonvolatile chemical compounds that likely change as a result of secondary reactions. The goal of this research is to (1) determine the effect of operating conditions on pyrolysis product yields; and (2) evaluate the temporal profile of both volatile and non-volatile compounds in the pyrolysis stream.

Hot solid residues fall directly into a catch-pot at the bottom of the freefall reactor, with the balance entrained in the sweep gas and recovered by the gas cyclones. It is generally thought that rapid heating of the biomass followed by rapid quenching of products is important to high bio-oil yields. The effect of residence time and temperature of hot solids in the catch-pot on bio-oil yield was investigated by either cooling the catch-pot or augering the solids out at different rates. Reactor Conditions Reactor temperature: 550°C; sweep gas flow rate: 5 SLPM; Heated length of reactor: 2.13 m (1.8 s heating time)

Reactor Design • 316 1¼” schedule 40 stainless steel pipe • GC/MS sampling ports positioned every 15.2 cm • 3.05 m reactor length • Temperature range: 300-1000°C • Pressure range: 0-100 kPa • Feed rate range: 0.25-2 kg/h • Maximum heating time of 2.5 seconds • Ability to operate continuously • Ability to manipulate particle velocities • Ability to manipulate reaction duration

Results from Different Char Collection Methods The residence time and the temperature of the char in the catch-pot was found to dramatically influence bio-oil yield. As shown in the figure below, moderately long residence times in the catch-pot at elevated temperature favored high bio-oil yield.

70.00%

Reactor Advantages % Yield of biomass

60.00%

Time resolved sampling of pyrolysis products Simple manipulation of important pyrolysis variables Identification and semi-quantification of both volatile land nonvolatile pyrolysis products • Reactor can be used with a variety of analytical methods • Mass Spectroscopy (MS) • Gas Chromatography/Mass Spectroscopy (GC/MS) • Mass Spectroscopy/Mass Spectroscopy (MS/MS) • Solid Phase Micro Extraction (SPME) • Fast turn-around of experiments • • •

Time-Resolved Measurements of Py Products

Comparison of Liquid and Solid Yields Manipulating Hot Solids Residence Time Solids Liquids NCGs

50.00% 40.00% 30.00% 20.00% 10.00% 0.00% Char Quench 300C

Char Auger Char Auger Char Auger 0.50 1.92RPS 0.625RPS RPS Char Collection Method

Char Catch

Increasing Hot Solids Residence Time

MW Distribution from TOF Mass Spectrometer

If the free-fall reactor is operated in steady state, then vertical positions along the axis of the reactor can be related to residence time of falling particles in the heated zone. Gas sampled from the reactor also contain condensable organic vapors and aerosols, which Beginning constitute bio-oil when recovered. A Time-of-Flight Mass of Heated Spectrometer (TOF MS) is used to analyze this gas stream because it Zone is able to detect a wide range of both volatile and non-volatile compounds in virtually real-time. The TOF MS is located on an elevator that brings the instrument very close to the desired sampling port, which are located at 15.2 cm intervals along the height of the reactor, providing equivalent temporal resolution of about one-tenth second in pyrolysis reactions. 0.150 Seconds

2000 0

100

200

300

400

500

600

700

800

900

1000

100

200

300

400

500

600

700

800

900

1000

100

200

300

400

500

600

700

800

900

1000

100

200

300

400

500

600

700

800

900

1000

Mass to Charge Ratio (m/z)

10000 8000 6000 4000

8000

Observations From Time-Resolved Tests

Fixed parameters for DOE • Feedrate: 1 kilogram per hour • Reactor headspace purge: 3 standard liters per minute • Solids collection configuration: Char Auger at 0.5 RPS • Liquid collection configuration: Cold Quench Version 2.1

4000

10000

0.332 Seconds

• Upon initial heating pyrolysis vapors containing relatively low molecular weight (MW) compounds are released very quickly; • Many of these initial low MW (100-250 Da) compounds reach their maximum concentrations within 0.5 seconds of initial heating; • As time further increases, these low MW compounds disappear. Since these are thought to be phenolic monomers and dimers, they are likely polymerizing to higher MW oligomers; • The number of high MW compounds (250-1000 Da) increases dramatically after 0.5 s. Since the data has not been quantified, it is difficult to conclude whether their mass accounts for the loss of mass from low MW compounds. Residence time tests with the char catch-pot suggests that some of this increase in high MW compounds is due to “slow devolatilization” of the hot solids; • After 0.7 s the high MW compounds reach their maximum concentration after which they decline or shift to MW > 1000 Da (detector limited to 100-1000 Da for the present test)

6000

6000 4000 2000 0 10000 8000

0.514 Seconds

6000 4000 2000 0 10000 8000

0.696 Seconds

6000 4000 2000 0

100

200

300

400

500

600

700

800

900

1000

100

200

300

400

500

600

700

800

900

1000

100

200

300

400

500

600

700

800

900

1000

10000 8000

0.878 Seconds

A Design of Experiments (DOE) is being performed to create an Response Surface Methodology (RSM) to better understand the relationship between reactor conditions and the pyrolysis products. The dependent variables include yields of liquid, solid, and noncondensable gases. The independent variables of reactor temperature, particle heating time, and sweep gas dilution.

Acknowledgements: CSET colleagues

8000

2000

The first time-series data of molecular weight distribution of pyrolysis products are displayed to the right. The reactor had a heated length of 1.8 m with heated wall temperature of 600°C. Sweep gas was 18 SLPM. Average particle heating time was1.06 s.

Other Experiments in Progress

Free-fall Reactor System

10000

Counts

Introduction

6000 4000 2000 0 10000 8000 6000

1.06 Seconds

4000 2000 0

Schematic of Free-fall Reactor With On-line Sampling System

Funded by: Phillips 66 Company

Bioeconomy Institute, Iowa State University

tcbiomass2013

Evaluation of Bio-oil Corrosion Characteristics Center for Sustainable Environmental Technologies

Patrick A. Johnstona patrickj@iastate.edu, Robert C. Browna, b aCenter for Sustainable Environmental Technologies (CSET), Iowa State University, Ames, IA 50011 bDepartment of Mechanical Engineering, Iowa State University, Ames, IA 50011

Results

Experimental

Introduction

Petroleum crude oils contain naphthenic acids that are one of the main causes of corrosion in the petroleum industry during storage, transport, and refining. Naphthenic Acid Corrosion (NAC) has been related to three parameters: Total Acid number (TAN), temperature, and flow rate [1, 2, 3, 4]. Bio-oils do not contain naphthenic acids but do however contain other organic acids such as acetic, formic, glycolic and propanoic acid that may have similar consequences to metal surfaces. A comprehensive corrosion study needs to be completed to determine the actual effects of the organic acids in bio-oil on select metal surfaces. As a result of this study, acid number and corrosion potential of pyrolysis bio-oils can be correlated and ranked to produce a fraction that is compatible with petroleum refining equipment. The proposed project will evaluate and compare corrosion and acid number (AN) on fast pyrolysis bio-oil produced at Iowa State University (ISU). The corrosion characteristics are evaluated using a vacuum distillation apparatus that simulates a distillation tower in a petroleum refinery. Selected metal specimens are subjected to both the liquid and vapor phase of the bio-oil. Method D664 “Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration” was used to determine acidity of the bio-oil samples. Ion Chromatography (IC) is used to quantify organic acid concentration. The corrosiveness is ranked gravimetrically according to mass loss during the study. Selected metal specimens are examined by Scanning Electron Microscopy (SEM) equipped with energy-dispersive spectroscopy (EDS) to determine morphological and composition changes on the metal surfaces.

Lab Scale Distillation Tower for Evaluating Bio-oil Corrosion Characteristics

Scanning Electron Microscopy – 1500X 304SS Control 304SS Vapor Phase (top) 304SS Liquid Phase

316SS Control

316SS Vapor Phase (top)

316SS Liquid Phase

17-4 Control

17-4 Vapor Phase (top)

17-4 Liquid Phase

Objectives

• Determine if existing petroleum refinery infrastructure can handle biooil as a “drop in” crude oil alternative. • Determine if the acidic components in bio-oil cause corrosion issues with specific types of metals found in the petroleum industry. • Evaluate metal specimens and acid composition of the bio-oil produced at ISU using a lab scale distillation tower apparatus [5]. • Determine whether acid number is a good predictor of corrosion potential.

Modified ASTM Titration for Modified Acid Number

Conclusions

• Bio-oil presents potential corrosion issue with both 316 and 304 stainless steel metals. • Precipitate-hardened metals such as 17-4 are little affected. • MAN appears to be a good indicator of corrosion potential for bio-oils. • These results do not take into account the overall type and corrosiveness of acids present.

Materials and Methods

https://www.cset.iastate.edu/tcbiomass2013/

• Longer duration tests. • Different types of metals. • Energy-dispersive x-ray spectroscopy (EDS). • Determine morphological and compositional changes on the metal surfaces.

F o rm 2

100

ce A -

ro p io n at e

ly c

ol at e

ta te

References

4 3 2 1 0

-20 0.0

Funding provided by: John Pappajohn & Iowa Farm Bureau

Future Work

Ion Chromatography for Organic Acid Composition at e

1 2 3 4 120 µS

Conductivity [µS]

I. Modified ASTM D664—“Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration”—to determine modified acid number (MAN) of bio-oil. II. Ion Chromatography (IC)—used to determine organic acid composition and concentration of bio-oil. III. Tested metals 304, 316, and 17-4 in lab scale distillation tower (750mL of bio-oil at 110°C). IV.Gravimetrically analyzed metal specimens. V. Scanning Electron Microscopy (SEM)—used to determine if the acidic composition of bio-oils is a good representation of potential corrosion.

min 2.0

4.0

6.0

8.0

10.0

12.0

16.0 14.0 Retention Time [min]

18.0

20.0

22.0

24.0

26.0

28.0

30.0

1. A. Jayaraman, H. Singh, and Y. Lefebvre, Rev. Inst. Fr.Pet.41, 265 (1986). 2. C. M. Cooper, Hydrocarbon Process.51, 75 (1972). 3. E. Babaian-Kibala, H. L. Craig, G. L. Rusk, et al., Mater.Perform. 32, 50 (1993). 4. E.B. Zeinalov, V.M Abbasov, and L.I. Alieva, Petro. Chem.49, 185 (2009). 5. H. D. Dettman, N. Li, J. Lou, Refinery Corrosion. CanmetEnergy-Nat. Res. Canada (2009).

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies Matthew Kieffer, Xianglan Bai, Robert C. Brown

Hydrothermal processing of microalgae using inorganic catalysts Introduction

Results and Discussion

• Algae is composed of lipids, proteins, and carbohydrates that can be used for biofuel production with higher photosynthetic efficiency, faster growth rates, and higher area specific yields than terrestrial biomass.

• HTP avoids the energy intensive drying process necessary in pyrolysis to convert biomass with high moisture content (80-90%) to fuels.

Aqueous

Organics Bio-oil

Solids

• Solvent: Water (20 mL) • Catalysts: Sodium carbonate, sodium formate, sodium hydroxide (1.5 g)

Equipment: • Stainless steel batch reactors

(volume of 32 mL)

• Heating mechanism: Techne Industrial Fluidized Bed 51

Initial Algae Composition

Averages St. Dev.

Solids

St. Dev.

No Catalyst

43.6%

1.2%

8.7%*

0.0%

0.4%*

0.3%

Sodium Carbonate

38.0%

0.2%

53.1%*

0.7%

7.3%*

0.8%

Sodium Formate

31.5%

0.6%

46.2%

4.3%

3.2%

0.6%

Sodium Hydroxide

27.7%

0.4%

54.0%

1.0%

8.4%

2.7%

No Catalyst Sodium Carbonate Sodium Formate Sodium Hydroxide

N(%) 5.92 4.10 4.84 3.77

C(%) 75.76 78.08 76.53 79.06

•

The yield of bio-oil reached maximum of 44 wt.% with no catalyst.

•

Bio-oil yields from catalytic HTP were overall lower than that from noncatalytic HTP.

• While catalysts are assumed to remain in the solids after the reaction, high yields of the aqueous fraction suggest that catalysts are possibly distributed between solid and aqueous phases after reaction. • All three catalysts deoxygenated the biooil, resulting in bio-oil with higher heating values than the oil produced with no catalyst.

Reaction temperature: 350°C

• Reaction time: 60 minutes

H(%)

O (% by HHV(MJ/Kg) difference) 10.39 34.47 9.38 36.36 9.98 36.10 8.54 37.20

S(%)

7.47 8.14 8.40 8.39

0.45 0.31 0.26 0.25

Data is on a dry, ash free wt.% *Requires re-run

Product Recovery Process:

• Solution was then centrifuged to separate water soluble, organic, and solid fractions • Extraction solvent is evaporated from the organic phase at 50ºC for 12 hours • Aqueous phase was dried at 70ºC and solids at 105ºC for 24 hours

Moisture (%)

Volatiles (%)

Fixed C (%)

Ash (%)

8.9% 0.6%

61.4% 3.6%

4.4% 1.2%

25.3% 3.4%

Financial support from Iowa Energy Center is gratefully acknowledged.

Sodium Formate

6.00E+07

4.00E+07

2.00E+07

0.00E+00

• It was found that bio-oil dissolved in dichloromethane provided higher definition of product compounds than when dissolved in acetone (not shown) for GC/MS analysis.

40.0% 35.0% 30.0%

• All three catalysts reduced the nitrogen containing compounds in the oil.

25.0%

• Sodium formate produced the highest yields of light oxygenates and cyclic pentanones. • All three catalysts produced a higher yield of light oxygenates, suggesting lower bio-oil yields from catalytic HTP is possibly due to the loss of light compounds during the solvent evaporation process.

20.0% 15.0% 10.0% 5.0%

Conclusions

0.0%

No Catalyst

Sodium Carbonate

Sodium Formate

Sodium Hydroxide

• While all the tested catalysts produced better quality of bio-oil compared to the case with nocatalyst, there were significant differences among the products from each catalyst. • Sodium carbonate and sodium hydroxide had the strongest effect in removing nitrogen and oxygen in the bio-oil among the four scenarios. • The overall lower bio-oil yield while using sodium carbonate in this study compared to literature may be a result of the high ash content in the algae and its catalytic effect. Lower bio-oil yields could possibly be due to the loss of light compounds during the solvent evaporation process.

600 550 500 450 400

Sodium Carbonate

No Catalyst Sodium Carbonate Sodium Formate Sodium Hydroxide

45.0%

650

No Catalyst

8.00E+07

50.0%

Average Molecular Weights of Bio-Oil

• Product was extracted with 16 mL of dichloromethane

Sodium Carbonate

Bio-Oil Yields

Average Bio-Oil Molecular Weight

• Maximum reaction pressure: 24.8 MPa

Average Molecular Weight (Da)

• Sample: Chlorella Vulgaris (3.5 g)

•

St. Dev.

• Sodium hydroxide resulted in bio-oil with the lowest N and highest heating value.

Conditions:

Microalgae HTP in Batch Reactor:

Aqueous

Percent of DAF Algae (%)

Organic solvent addition and phase separation

St. Dev.

Catalyst

Hydrothermal Processing

Gases

Oil

Bio-Oil Composition

• Sodium carbonate is a catalyst commonly used in HTP of microalgae to increase the yield and decrease N yields of the bio-oil. Sodium formate and sodium hydroxide are suggested as intermediate compounds resulting from the participation of sodium carbonate with the watergas shift [2] . The present study compared the catalytic effects of the three related sodium salts during HPT of microalgae with high ash content.

Products

Catalyst

No Catalyst

Sodium Hydroxide

• Hot compressed subcritical water has fewer and weaker hydrogen bonds, promoting a high availability of H+ and OH- for acid and base catalyzed reactions such as hydrolysis [1].

Microalgae Slurry

GC/MS Chromatograms and Compound Yields

Product Yields

• Hydrothermal processing (HTP) is the decomposition of biomass in the presence of water at elevated temperature and pressure.

Materials and Methods

Results and Discussion

Sodium Formate

Sodium Hydroxide

•

The average molecular weight ranged from 474 Da to 579 Da

•

The bio-oils produced from catalytic HTP have increased molecular weights compared to the bio-oil obtained from non-catalytic HTP. This is possibly because light oxygenates have been removed during the evaporation of the solvent from the bio-oil fraction.

• Challenges arise in the separation of the product fractions when working with HTP. There are many variations in equipment, techniques, feedstocks and solvents used that create lots of variance among different studies. Future studies should consider these factors when performing experiments as well as comparing and analyzing data.

References 1.

López Barreiro, D., et al. (2013). "Hydrothermal liquefaction (HTL) of microalgae for biofuel production: State of the art review and future prospects." Biomass and Bioenergy 53(0): 113-127.00

Elliott, D. C. and L. J. Sealock (1983). "Aqueous catalyst systems for the water-gas shift reaction. 1. Comparative catalyst studies." Industrial & Engineering Chemistry Product Research and Development 22(3): 426-431.

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013/

Kwang Ho Kim, Xianglan Bai, Erica Dalluge, Carolyn Hutchinson, Dustin Dalluge, Young-Jin Lee and Robert C. Brown

Thermal Depolymerization of Lignin Introduction

Results & Discussions

 Bio-oil produced from biomass pyrolysis contains 25-30 wt% of lignin derived

 GPC of fresh SR bio-oil, aged solvent-evaporated SR bio-oil and pyrolytic lignin

phenolic oligomers with molecular weight (MW) up to 2,500 Da

 GPC of SR pyrolysis oil of several phenolic monomers at 500°C

• Fresh SR bio-oil • Aged solvent-evaporated bio-oil

 In contrast to phenolic monomers which can be utilized as value-added

- SR bio-oil was stored two weeks after

chemicals, phenolic oligomers are detrimental to the process due to the instability

evaporation of solvent • Pyrolytic lignin

and the deactivation of catalysts

- Extracted from bio-oil produced from

 The mechanism of phenolic oligomer formation and the mode of transport of

fast pyrolysis of red oak • Lignin

these non-volatile compounds are still under debate Monomers

- corn stover organosolv lignin

Monomers

Monomers Lignin

Oligomers

Lignin

Monomers

Oligomers

Oligomers Char

Hypothesis 1: Phenolic oligomers are thermally ejected from biomass

Char

•

The fresh SR bio-oil showed two sharp, prominent peaks around 172 and 372 Da. The average MW of the bio-oil was 286 Da, which is significantly lower compared to original lignin or pyrolytic

Hypothesis 2: Phenolic oligomers are formed by reoligomerization of monomers

 To understand the origin of phenolic oligomers found in bio-oil, fast pyrolysis of lignin was conducted and the pyrolysis products were investigated

lignin recovered from conventional fast pyrolysis •

In the aged solvent-evaporated bio-oil, the molecular weight shifted to higher MW region as it aged

 High-resolution mass spectra (HRMS) of SR bio-oil from lignin pyrolysis at 500°C *Peak intensity of FT-ICR is relative to m/z 340

Experimental

and peak intensity of orbitrap is related to m/z 150. Low mass compounds (m/z<300)

 Material: Cornstover organosolv lignin, 4-ethylphenol, guaiacol, syringol and 2-

underrepresented by FT-ICRMS are detected by

methoxy-4-vinylphenol

oritrap-MS.

 Micropyrolysis: Frontier Lab Micropyrolyzer (PY-2020id)  Recovery of pyrolysis products in cold solvent: injection needle from

*The peaks above 504 Da and below 690 Da

micropyrolyzer was directed into cold solvent (methanol or tetrahydrofuran); the

were unable to be assigned using the elements

product is referred to as solvent recovered (SR) bio-oil

C, H, N, and O within an acceptable mass error range.

 Analytical methods - Gel permeation chromatography (GPC): molecular weight distribution of SR bio-oil

• The MW distributions of HR-MS match the GPC chromatogram very well

- Atmospheric pressure photoionization (APPI) high resolution mass spectrometry:

•

A total of 569 phenolic compounds were identified by the two APPI-HRMS and Py-GC/MS

SR bio-oil of lignin was analyzed with Fourier transform ion cyclotron resonance

•

FT-ICRMS shows that the MWs of phenolic compounds were <504 Da while pyrolytic lignin had

(FT-ICRMS) and orbitrap-MS - The compounds with MW <300Da were better detected by

MWs >1000 Da. •

trimers (DP=3) and tetramers (DP=4)

were better detected by FI-ICRMS

were assigned based on Double Bond Equivalence (DBE) value.

< Frontier Micropyrolyzer and SR pyrolysis oil collection system >

1.manual injection sample holder; 2. micropyrolyzer oven; 3. Vial filled with solvent, injection needle from part 2 is immersed in the solvent; 4. helium gas line; 5. supporting plate

SR bio-oil is dominated by monomers and dimers and it is followed by trimers. In comparison, the abundance of tretramers is very low.

Double bond equivalent (DBE) value vs. molar mass (M/Z) for compounds detected in SR bio-oil

This work was supported by National Advanced Biofuels Consortium (NABC)

Methoxyl group containing compounds produced higher MW products than the reactants and larger MW compounds were produced with increasing pyrolysis temperature

•

Pyrolysis of 2-methoxy-4-vinylphenol produced larger MW compounds and the average MW of SR bio-oil increased from 181 Da to 274 Da with increasing pyrolysis temperature

•

The results indicated that phenolic monomers with reactive functional groups could rapidly initiate reoligomerization during pyrolysis via side chain addition and rearrangement.

Conclusions  A total of 569 phenolic compounds with molecular weight less than 504 Da were found from SR bio-oil, which are significantly smaller than conventional pyrolytic lignin  Phenolic monomers and dimers are most abundant products during lignin pyrolysis and it is followed by trimers  Phenolic oligomers in bio-oil is formed by reoligomerization of small phenolic molecules originating from the primary pyrolysis reactions

Monomers (DP=1), dimers (DP=2),

orbitrap-MS and the compounds with MW >300Da

•

Monomers

Dimers

Oligomers

Lignin

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013/

Kwang Ho Kim, Xianglan Bai and Robert C. Brown

Hydrogen Donor Solvent Assisted Thermal Decomposition of Lignin to Alkylphenols Introduction

Results & Discussions

 Lignin is the second most abundant natural polymer in the biosphere and produced in large amounts

 Solvolysis with tetralin

from the pulping process  In US, lignin is extracted over 50 million tons from pulping industry per year and yet only 2% of lignin is used commercially because of its structural recalcitrance  The interest in lignin as renewable feedstock for biofuel and aromatic chemicals has been increasing

5 min 5.52 1.90 2.83 34.4

Total phenols (wt.%) Alkylphenols (wt.%) Vinylphenols (wt.%) Selectivity (%)*

Alkylphenols R1= CH3, CH2CH3 or CH2CH2CH3 R2 and R3 = H or OCH3

R3 OH

300 °C 10 min 15 min 5.76 5.90 2.32 2.79 2.62 2.25 40.3 47.4

350 °C 10 min 15 min 9.85 10.18 6.58 7.45 2.09 1.48 66.6 73.1

5 min 9.62 6.22 2.16 64.5

4-Ethylphenol

products could undergo condensation through radical coupling reactions that lead to the new carbon-

1.5

carbon linkages and ultimately char  Solvolysis of lignin with hydrogen donor solvents is known to improve the quality of bio-oil. The

decreased

300°C

1.0

350°C 400°C

0.5

4.92 59.64 34.48 0.97

350°C

1.0

400°C

0.5

• Tetralin as the solvent: the molecular weight of bio-oil (tetrahydrofuran solubles, THF solubles) decreased from 356 to 258 Da with increasing reaction time from 5 min to 15 min

10 Reaction time (min)

w/ naphthalene

• Isopropanol as the solvent: the molecular weight of THF solubles decreased from 335 to 315 as the reaction time increased from 5 min to 15 min

Total phenols (wt.%) Alkylphenols (wt.%) Vinylphenols (wt.%) Selectivity (%)*

5 min 3.86 1.59 1.71 41.2

300 °C 10 min 15 min 4.05 4.28 1.86 2.42 1.62 1.21 45.9 56.6

5 min 7.35 5.50 0.97 75.0

350 °C 10 min 15 min 7.93 7.89 6.87 6.98 0.38 0.26 86.7 88.5

5 min 9.30 8.42 0.23 90.6

400 °C 10 min 15 min 9.62 10.60 8.89 10.06 0.23 0.20 92.5 94.9

 Reaction mechanism

*Selectivity of alkylphenols

62.80 5.75 29.81 1.64

2.0 300°C

1.5

350°C

1.0

400°C

0.5

GPC for molecular weight distributions

0.0 5

10 Reaction time (min)

350°C

400°C

10 Reaction time (min)

• The yield of total monomeric phenols was lower, but the selectivity of alkylphenols were higher compared to using tetralin as the solvent

 Solvolysis with naphthalene 1.6

• Naphthalene is a non-hydrogen donor solvent.

Yield (wt.%)

1.0 0.8 0.6

5min

0.4

10min

0.2 0.0

< Online Microsolvolysis System >

This work was supported by National Advanced Biofuels Consortium (NABC)

15min

isopropanol

naphthalene

+H· Δ

300°C

• The yields of total monomeric phenols and alkylphenols decreased with increasing reaction times.

H·

acetone

R·

+ etc.

+H·

R1 and R2 = H or OCH3

• Hydrogen atoms donated by tetralin and isopropanol stabilize highly reactive primary phenols by saturating side chains. For example, vinyl group is converted to ethyl group.

Conclusions  Lignin solvolysis with hydrogen donor solvents improves the quality of bio-oil mainly by saturating phenolic compounds with reactive functional groups to stable alkylphenols  Without hydrogen donor solvent, the primary products could undergo secondary condensation reactions and convert to large molecular weight compounds H3CO

Lignin

OCH3 O

O OCH3

• The primary products likely repolymerize to larger molecular weight compounds or char.

• Similarly, increasing temperatures and reaction times improved the selectivity of alkylphenols whereas reduced the yield of vinylphenols during lignin solvolysis with isopropanol

1.2

4-Vinylphenol

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 5

1.4

+ 4H·

tetralin

4-Ethylphenol

2.5

• Naphthalene as the solvent: the molecular weight of bio-oil increased with increasing reaction times

increased

 Solvolysis with isopropanol

Yield (wt.%)

Proximate analysis (wt%) Moisture Volatiles Feedstock: Cornstover organosolv lignin Fixed carbon Solvent: Tetralin, isopropanol (Hydrogen donor solvent) Ashes Ultimate analysis (wt%) Naphthalene (non-Hydrogen donor solvent) C Reaction temperature: 300, 350 and 400°C H O1) Reaction time: 5, 10 and 15 min N Analyses: Online GC-FID/MS system for identification and quantification

300°C

1.5

0.0

Frontier Lab) attached to a micro-furnace pyrolyzer (PY-3030D, Frontier Lab) with online product analysis

2.0

Yield (wt.%)

 Solvolysis of lignin was conducted in sealed glass capsule and heated in a micro reactor (PY1-1050,

decreased

• The yield of total monomeric phenols increased with increasing reaction temperatures and reaction times in the range tested • While higher solvolysis temperature and longer reaction time improved the yield of total phenols, the selectivity of alkylphenols is also enhanced • 4-ethylphenol, a major alkylphenol, significantly increased at the expense of 4-vinyl phenol, the primary product of lignin, with increasing temperatures and reaction times

Experimental



w/ isopropanol

2.5

solvents are investigated in the present study



w/ tetralin

3.0

0.0

primary reaction pathways of lignin decomposition during solvolysis conversion with hydrogen donor



5 min 10.94 7.73 2.18 70.6

400 °C 10 min 15 min 11.31 11.36 8.33 8.68 1.99 1.77 73.7 76.4

4-Vinylphenol

Yield (wt.%)

2.0

Yield (wt.%)

 Thermal decomposition of lignin could produce a various type of monomeric phenols. However, those



 Molecular weight distributions

*Selectivity of alkylphenols

over the last decade



Vinylphenols R1= CHCH2 R2 and R3 = H or OCH3

OH OH Lignin

H3CO HO

Lignin

OCH3

Solvolysis

OCH3 OH

‘Stable’ monomeric phenols

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013/

Longwen Ou, Yihua Li, Tristan R. Brown, Guiping Hu, Robert C. Brown

Techno-Economic Analysis of the Production of Hydrocarbons from Pyrolytic Sugars Introduction • It is possible to produce monosaccharides from biomass fast pyrolysis by recovering bio-oil in several stages according to boiling point. • The resulting monosaccharides are readily upgraded to liquid alkanes by aqueous-phase processing. • Alkanes produced by direct aqueous-phase dehydration/hydrogenation of sugars has low value due to high volatility. C-C bonds formation is required to produce gasoline and diesel range fuels from sugars. • This analysis is performed to understand the economics of fast pyrolysis with monosaccharide extraction and upgrading to liquid fuels as well as to identify possible ways to improve the process economics.

Monosaccharide Extraction and Upgrading Process

Capital and Operating Costs $9

Installed Equipment Cost (Million $) Reforming and Hydroprocessing Monosaccharides Extraction Pyrolysis & Oil Recovery Storage Glucose Dehydration Combustion Pretreatment Partial Separation HMF Upgrading Utilities

$71 $114

$3 $26 $7 $3

$43

Sugar Upgrading Reaction Pathway

$35

$53

Annual Operating Costs (Million $) $250

OH OH O

HO HO

OH 3H2O

O OH

isomerization

OH Fructose

Glucose

HMF

4H2O

C9-alkane

5H2

3H2

hydrogenation

Simplified Process Flow Diagram Biomass

Pretreatment

Solids Removal

Pyrolysis

Char

Natural Gas Process Heat

Bio-oil Recovery

Heat Generation

Noncondensable Gases

Noncondensable Gases Hydrogen

Heavy Ends

Sugar Extraction

Phenolic Oligomers Light Ends

Fuel Gas Excess Hydrogen Gasoline and Diesel

HMF Upgrading

HMF

Partial Separation

Gasoline and Diesel

• Bio-oil is recovered in two stages according to boiling point. 1.“Heavy ends” consisting mostly of water-soluble sugars and waterinsoluble phenolic oligomers. 2.“Light ends” consisting of water, carboxylic acids and aldehydes. • Heavy ends separated into sugars and phenolic oligomers, the latter of which is hydroprocessed to gasoline and diesel. • Light ends are steam reformed to produce hydrogen for the process.

Co-product Credits

$100

Utilities

Fixed Costs Waste Disposal Electricity Other raw materials and catalysts Feedstock

Monosaccharide Dehydration Levulinic Acid

Assumptions Cost year Stream factor Plant Life Equity Loan term

$US 2011 90% 30 years 40% 10 years

Equity

40%

Internal rate of return (after tax)

10%

Income tax rate

39%

Results Total Purchased Equipment Cost (TPEC) Working Capital Total Fixed Capital Investment Total Investment (with land)

$86

100%

Sensitivity Analysis Fixed Capital cost ( $742;$571;$400MM) Feedstock cost ( $97.5;$75;$52.5/dry ton)

MM $ % of TPEC $122

• HMF upgrading accounts for 30% of the total capital cost because several large volume, high pressure reactors are used. • Reforming and hydroprocessing is the second largest contributor to the capital cost, accounting for 19% of the total capital cost. • Feedstock cost contributes to 27% of the operating cost. • Utilities cost is high due to the existence of several distillation columns.

Pessimistic: Base case: Optimistic

Economic Assumptions and Main Results

Monosaccharide

Bio-oil Upgrading

$150

Capital Depreciation

($50)

dehydration/ HO hydrogenation

Average Income Tax

$200

$50

aldol condensation OH

Average Return on Investment

Annual fuel yield

Yield per 15% of FCI feedstock

$571

469%

Total project Investment

664

546%

MFSP

37.2

MM gal/yr

45.9 gal/dry ton 17.8

$/annual gal

5.15

$/gal

The authors would like to acknowledge the financial support of the U.S. Department of Energy through the Pacific Northwest National Laboratory.

$4.44

$5.87 $4.71

$5.60

Working Capital (22.5%;15%;7.5%)

$4.98

$5.33

Income tax rate (49%;39%;29%)

$5.01

$5.33

Natural gas price ($209;$169;$129/MT) Hydrogen price ($2.44;$3.49;$4.54/kg)

$5.06 $5.10

$5.25

Optimistic Pessimistic

$5.21

$4.1 $4.4 $4.7 $5.0 $5.3 $5.6 $5.9 $6.2 MFSP($/gal)

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies Marjorie Rover, Patrick Johnston, Ryan Smith, and Robert C. Brown

Sugar and Phenolic Oligomer Recovery from the Heavy-Ends of Fractionated Bio-Oil

Introduction

Original Bio-Oil Heated to 80°C

Materials and Methods

Results

Sugars and Phenols Recovery from the Heavy Ends of Fractionated Bio-Oil SF 1 Sugar Solution

SF 2 Sugar Solution

SF 1

15 10 5 0

25 20 15 10 5 0

2 3 4 Water-to-Heavy Ends Ratio (n:1)

SF 2

2 3 4 Water-to-Heavy Ends Ratio (n:1)

Moisture in Phenolic Raffinate After Removal of Sugars Water-to-Heavy Phenolic Oligomer-Rich Ends Ratio Raffinate SF1 Moisture (%)

Phenolic Oligomer-Rich Raffinate SF2 Moisture (%)

0.5:1

27.02±1.74

22.57±0.70

1:1

18.30±0.17

17.94±0.49

2:1

22.57±0.42

22.02±0.59

5:1

21.89±0.62

19.70±0.26

20 SF1

SF2

10 5 0

2 3 Number of Washes

30 25

SF 1

SF 2

15 10

40 60 80 100 Extraction Temperature (°C)

120

Raffinate Relative Molecular Weight Distribution

3.E+04 Area (mAU*min/gm)

3. The minimum amount of water required for the phase separation of bio-oil water-soluble constituents from the water-insoluble constituents was determined by the addition of deionized water drop-wise into SFs 1 and 2 while stirring thoroughly by hand after each addition. The phenolic oligomerrich raffinate was separated from the water-soluble components using a known amount of oil mixed at different ratios by weight with deionized water.

Sugars (wt% db)

2. Bio-oil constituents were evaluated and quantified using GC with a flame ionization detector (GC/FID), acid content was determined by ion-exchange chromatography, sugar content was determined using ultraviolet-visible range spectroscopy [3], and gel permeation chromatography (GPC) was used to determine the molecular weight distributions [2]. Water-insolubles were determined by an in-house method and moisture was determined by Karl Fischer [1]. Ultimate analyses were performed with oxygen determined by difference [2].

Extraction of Sugars as a Function of Water-to-Heavy Ends Ratio

Sugars (wt% db)

1. Bio-oil was produced in a fast pyrolysis unit consisting of a fluidized bed operated at 450-500°C and a bio-oil recovery system that recovers bio-oil in distinct multiple stage fractions (SF) [1]. A sugar-rich aqueous phase and a phenolic-rich raffinate were separated from SF 1 and SF 2 [2].

Clean Phenolic Oligomer Raffinate at 21°C After Washing

Total Sugar (wt% db)

Total Sugars (wt% db)

This study explores separate recovery of sugars and phenolic oligomers produced during fast pyrolysis. Bio-oil fractionation is accomplished with a fivestage system, recovering bio-oil according to condensation temperatures of chemical constituents. The first two stages capture “heavy ends”; mostly water soluble sugars and water insoluble phenolic oligomers. Exploiting differences in water solubility allows for recovery of a sugar-rich aqueous phase and a phenolicrich raffinate. Over 93 wt% of the sugars in stage fractions (SF) 1 and 2 are removed in two water washes. Pyrolytic sugars from SF1 and SF2 are suitable for upgrading to biofuels catalytically or by fermentation. The phenolic raffinate, representing 44-47 wt% dry basis (db) of both SF1 and SF2, is less sticky and viscous than the unwashed SFs. It shows potential for production of fuels, aromatic chemicals, polymers, resins, asphalt, etc.

Phenolic Raffinate

One Wash at 80-100°C is Sufficient

Viscosity Difference

3.E+04

Trimers

Phenolic

Dimers

2.E+04

SF 1 Heavy-Ends Raffinate SF 2 Heavy-Ends Raffinate

2.E+04 Monomers

1.E+04 5.E+03 0.E+00

Conclusions

100 Mw (Da)

1000

10000

1. We successfully demonstrated the ability to separate sugars and lignin-derived phenolic oligomers from the heavy fractions of biomass. 2. The sugars were extracted effectively at over 93 wt% with two water washes. 3. Approximately 3-7 wt% of other water-soluble or partially soluble constituents were removed with the water-soluble sugars. 4. This research has shown that sugars and phenolic oligomers can be separated, providing two separate streams for fermentation, catalytic upgrading, or other kinds of conversions to value-added products.

References

1. A.S. Pollard, M.R. Rover and R.C. Brown, Journal of Analytical and Applied 3. M.R. Rover, P.A. Johnston, Pyrolysis, (2012). B.P. Lamsal, R.C. Brown, accepted Journal of Analytical and Applied 2. M.R. Rover, P.A. Johnston, R.G. Smith, R.C. Brown, under review, Pyrolysis, (2013). Bioresource Technology, (2013). https://www.cset.iastate.edu/tcbiomass2013/

Funding provided by: Phillips 66 Company

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies Marjorie Rover, Patrick Johnston, Ryan Smith, and Robert C. Brown

Comprehensive Thermal Degradation Study of Bio-Oil Phenolic Oligomers

Materials and Methods

1. Bio-oil was produced in a fast pyrolysis unit consisting of a fluidized bed operated at 450-500°C and a bio-oil recovery system that recovers bio-oil in distinct multiple SFs [1]. SF 1 and SF 2 were included in this study.

Temperature Range (°C)

SF 2 Residue Remaining (wt%)

58 - 108

80.3

31 - 56

91.7

108 - 155

70.4

57 - 108

83.5

153 - 201

59.7

108 - 152

77.3

201 - 251

51.8

152 - 201

67.8

251 - 301

44.8

201 - 253

58.1

302 - 352

39.5

252 - 302

46.3

352 - 402

35.7

304 - 353

39.2

355 - 403

35.5

2.0E+05 1.5E+05

3. GPC was used to determine the molecular weight distributions [2] of the residue remaining after each temperature ramp and hold time. The polystyrene standards used had a molecular weight range of 162 – 38,640 g mol-1. Each initial sample size was approximately 60 mg. After each temperature ramp the residue remaining in the cup was put into tetrahydrofuran (THF) to dissolve the THF solubles. A new sample cup was placed on the TGA and the temperature ramp was followed until reaching the next highest hold temperature of the preceding sample until reaching 300°C.

Area (mAU*min/g)

0.0E+00

Relative Mw=108 Da

2.5E+04

Relative Mw=87 Da

2.0E+04 1.5E+04

Volatiles 250-300°C Volatiles 300-350°C

0.0E+00

Trimers [4]

1.8E+04

Dimers [4]

1.6E+04

100 Mw (Da)

Raffinate 30-50°C Raffinate 30-105°C Raffinate 30-150°C Raffinate 30-200°C Raffinate 30-250°C Raffinate 30-300°C

2.0E+05 1.5E+05 1.0E+05

1000

10000

1.4E+04

SF 2

Trimers [4] Hexamers [4] Dimers [4]

Octamers [4]

1.0E+04

Volatiles 150-200°C

8.0E+03

Volatiles 200-250°C

6.0E+03

Volatiles 250-300°C

Relative Mw=364 Da

Volatiles 300-350°C 1

100

1000

10000

Conclusions

1. A.S. Pollard, M.R. Rover and R.C. Brown, Journal of Analytical and Applied Pyrolysis, (2012).

100 Mw (Da)

1000

Mw (Da)

References

5.0E+04 10

SF 2

1. SF 1 and SF 2 contain 44-47 wt% phenolics (22-24 wt% g/g whole bio-oil). 2. SF 2 phenol oligomeric raffinate appears to be more stable than SF 1  Dramatic increase in molecular weight does not occur until 200°C, whereas SF 1 occurs at 150°C.  The 300°C SF 1 raffinate sample was not soluble in THF while SF 2 raffinate was. 3. SF 2 raffinate is comprised of larger molecular weight species versus SF 1.

Monomers [4]

10000

Relative Mw=87 Da

1.2E+04

0.0E+00

1000

Relative Mw=278 Da

2.0E+03 10

100 Mw (Da)

Relative Mw=185 Da

4.0E+03 1

Volatiles 150-200°C Volatiles 200-250°C

2.0E+04

Monomers [4]

2.5E+05

0.0E+00

3.0E+04

5.0E+03

SF 1

1.0E+05

3.0E+05

Phillips 66 Company

Raffinate 30-50°C Raffinate 30-105°C Raffinate 30-150°C Raffinate 30-200°C Raffinate 30-250°C Raffinate 30-300°C

Hexamers [4]

Relative Mw=185 Da

1.0E+04

SF 1 and SF 2 Raffinate Relative Molecular Weight Distribution 2.5E+05

SF 1

3.5E+04 Area (mAU*min/g)

SF 1 Residue Remaining (wt%) 90.5

5.0E+04

Funding provided by:

4.0E+04

Temperature Range (°C) 33 - 57

2. The weight loss analyses were done using TGA. The bio-oil sample size was approximately 45 mg. Approximate temperature ramps included 30-50°C, 50105°C, 105-150°C, 150-200°C, 250-300°C, 300-350°C, 350-400°C at a rate of 5°C min-1. Each endpoint in the individual ramp was held 30 minutes prior to proceeding to the next ramp. Nitrogen purged the system at 100 mL min-1 for the entire experiment.

4. The relative molecular weight distribution of the volatiles was also performed using the same described methodology. Each approximate initial sample size was 180 mg. The first sample was not acquired until after the 150°C hold time. The next sample was acquired after the 200°C hold time and so forth until after the 350°C hold time. The volatiles were captured in a 25 mL impinger utilizing ice as the coolant and THF as the solvent. These samples were analyzed by GPC, as previously described.

Relative Molecular Weight of Raffinate Volatiles SF 1 and SF 2

SF 1 and SF 2 Phenol Oligomeric Raffinate Weight Loss (30- 400°C)

Area (mAU*min/g)

Bio-oil from fast pyrolysis is a promising economical approach to advanced biofuels production and is considered a potential candidate to replace or reduce the use of petroleum fuels. Bio-oil is unstable under long-term storage or upon heating. Understanding the thermal stability of bio-oil is important to upgrading it to transportation fuels, especially if it is to be integrated into petroleum refining operations. The release of vapors from bio-oil samples and their correlation to molecular structure will provide insight for both emissions, as well as sample degradation characteristics that occur during heating processes for specific upgrading technologies. This study explores the thermal behavior of biooil phenolic oligomers for its potential to be successfully upgraded to liquid fuels. Bio-oil fractionation was accomplished utilizing Iowa State University’s five-stage system that recovers bio-oil as stage fractions (SF) according to the constituents condensation points. The first two SFs collect “heavy ends” comprised of both water-soluble sugars derived from polysaccharides and waterinsoluble phenolic oligomers derived from lignin. The phenolic oligomers were heated by a thermogravimetric analyzer (TGA) to determine weight loss characteristics from 30-350°C. Vapors were also collected during the experiments and gel permeation chromatography (GPC) analyses were utilized to determine relative molecular mass. The phenolic oligomer volatiles from SF 1 showed molecular weight (Mw) ranges consistent with monomers-pentamers while SF 2 showed molecular weight ranges consistent with monomershexamers. The GPC studies indicated SF 2 phenolic oligomers were more stable versus SF 1.

Results

Area (mAU*min/g)

Introduction

10000

2. M.R. Rover, P.A. Johnston, R.G. Smith, R.C. Brown, under review, Bioresource Technology, (2013).

3. Mullen CA, Boateng AA. Characterization of water insoluble solids isolated from various biomass fast pyrolysis oils. Journal of Analytical and Applied Pyrolysis (2011)..

https://www.cset.iastate.edu/tcbiomass2013/

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies Marjorie Rover, Patrick Johnston, Tao Jin, Laura Jarboe, and Robert C. Brown

Funding provided by: Phillips 66 Company National Science Foundation (NSF) # CBET-1133319

Formic Acid

2,6-methoxyphenol 0.04±0.02

Valeric Acid

Furfuryl Alcohol

0.20±0.13

0.09±0.04

Butyric Acid

2(5H)-furanone

0.09±0.01

0.29±0.02

Acetic Acid

Guaiacol

0.10±0.04

0.05±0.01

Vanillin

0.06±0.01

Phenol

0.06±0.01

5-HMF

0.32±0.07

0.33±0.02

0.89±0.01

0.56±0.02

0.57±0.002

0.34±0.03

0.84±0.01

0.43±0.04

0.08±0.003

0.07±0.01

Furfural Guaiacol Acetol

Compound Concentrations Acetic acid Below Inhibitory Wt% • Acetol: (5.0, IC100) [9] Formic acid • 5 HMF: (0.45, IC100) [10] Glycolic acid • Guaiacol: (0.30, IC100) [11] Propionic acid

0.08±0.01

Untreated

Liquid-Liquid Extraction

Ionic Liquid

NaOH untreated

Ion-Exchange Resin

5-HMF

Phenol

Vanillin

Guaiacol

2(5H)-Furanone

2,6Dimethoxyphenol

Furfural

NaOH treated

5.0 4.0 3.0 2.0 1.0 0.0

NaOH Untreated @ 48 h NaOH Treated @ 48 h Glucose Control @ 48 h

0.0

Conclusions

Comparison of Liquid-Liquid Extraction, Ionic Liquid, and Ionic Resin

0.5

1.0

1.5 2.0 2.5 3.0 Concentration of sugar (wt%)

3.5

4.0

1. NaOH Overliming  It was the most successful detoxification method with no loss of sugars.  Utilization of 1 and 2 wt% pyrolytic sugars was improved relative to the untreated sugars, but 3 wt% sugars was inhibitory. 2. Other Overliming Treatments  Ca(OH)2 removed 7 wt% sugars and continued precipitating for several days.  NH4OH did not perform as well as the NaOH and it showed no loss of sugars. 3. Other Detoxification Methods A. Liquid-Liquid Extraction  It removed phenolics, acids, and reduced the furans without loss of sugar. B. Ionic Liquid  Phenols were not removed as effectively however 5-HMF and acids were successfully removed with no loss of sugars. C. Ion-Exchange Resin  Acids were successfully removed although phenolic compounds and furans were not, additionally 8 wt% sugars were lost.

References

Acetic Acid

Formic Acid

1. A.S. Pollard, M.R. Rover and R.C. Brown, Journal of Analytical and Applied 7. B. Alriksson, A. Sjöde, N.-O. Nilvebrant and L. Jönsson, Applied Pyrolysis, (2012). Biochemistry and Biotechnology, 130, (2006) 599.

Glycolic Acid

1.2 1.0 0.8 0.6 0.4 0.2 0.0

0.6 0.5 0.4 0.3 0.2 0.1 0

Acetol

Chemical (wt% wb)

0.25±0.06

NaOH is the Optimum Overliming Candidate

Utilization of Pyrolytic Sugars by Ethanologenic E. coli

OD 550

Furfural

5-HMF

7. Three treatments by overliming were evaluated for detoxification of the sugar solutions. The solutions were brought to pH 7 after each treatment using H2SO4.  Ca(OH)2: 30°C for 3 h at pH 11 [6]  NaOH: 80°C for 3 h at pH 9 [7]  NH4OH: 55°C for 3 h at pH 9 [7]

Acetol

Phenol

6. An ion-exchange resin: Dowex 66, was mixed in a ratio of 1 part to 5 parts watersoluble sugars, vortex for 30 min and centrifuge 20 min at 2635 g.

Compound 5-Hydroxymethyl furfural (5-HMF)

SF1 (wt% wb) SF2 (wt% wb) 0.55±0.07 0.31±0.01

Vanillin

5. An ionic liquid:,1-methyl-3-octylimidazolium tetrafluoroborate [5], was mixed in a ratio of 1 part ionic liquid to 5 parts water-soluble sugars, vortex for 30 min and centrifuge for 20 min at 2635 g.

Compound

Guaiacol

4. The sugars and fermentation media were inoculated with ethanol-producing Escherichia coli (KO11 strain). The growth was presented by optical density (OD) at 550 nm. KO11 was cultured in 10 mL medium of Luria Broth and pyrolytic sugars in 50 mL centrifuge tubes at 37°C.

SF 2 Sugar Solution

Compounds Remaining in Sugars

3-Methyl-1,2cyclopentanedione

3. Liquid-liquid extraction detoxification method was evaluated. A solution of: 25% tri-n-octylamine in 1-octanol [4] was mixed in a 1 to 1 ratio by weight of the watersoluble sugars and placed on a shaker table for 2 h and centrifuged for 10 min at 2635 g.

Known Escherichia coli Inhibitors [8]

2(5H)-Furanone

2. Bio-oil constituents were evaluated and quantified using GC with a flame ionization detector (GC/FID), acid content was determined by ion-exchange chromatography, and the sugar content was determined using ultraviolet-visible range spectroscopy [3].

SF 1 Sugar Solution

2,6-Dimethoxyphenol

Original Bio-Oil

Furfural

Materials and Methods

Are These Sugars Dirty?

Acetol

This study focuses on the effective removal of contaminants from pyrolytic sugar to produce a suitable fermentation substrate. Iowa State University utilizes a bio-oil recovery system from fast pyrolysis of lignocellulosic biomass as stage fractions (SF). The first two SFs collect “heavy ends” comprised of both sugars and phenolic oligomers. Exploiting differences in water solubility, we are able to recover a sugar-rich aqueous phase and a phenolic-rich raffinate. The sugar-rich aqueous phase contains small percentages of other water-soluble constituents such as low molecular weight acids, furans, and phenols that are possibly inhibitory to successful fermentation. Analyses of the sugar-rich aqueous phase by gas chromatography/flame ionization detector (GC/FID) indicated several compounds, acetol, guaiacol, and 5hydroxymethylfurfural (5-HMF) known to be inhibitory to microbes/bacteria, were below the inhibitory wt% without additional detoxification. However, other compounds such as acetic acid, formic acid, and furfural require removal before fermentation. Current methods of detoxification were evaluated. These included overliming, liquid-liquid extraction, ionic liquid and ionic resin for removal of contaminants. Our research has shown the optimal candidate for detoxification of the pyrolytic sugars was sodium hydroxide overliming which showed maximum growth measurements utilizing ethanol-producing Escherichia coli (E. coli). We successfully removed the following percentages of compounds present in the initial sample utilizing sodium hydroxide overliming: 80% acetol, 80% furfural, 56% 2,6dimethoxyphenol, 47% guaiacol, 74% vanillin, 91% phenol, and 82% 5-HMF with no degradation or loss of pyrolytic sugars.

Results

Chemcals (wt% wb)

Introduction

Clean Pyrolytic Sugars Solution

2. M.R. Rover, P.A. Johnston, R.G. Smith, R.C. Brown, under review, Journal of Analytical and Applied Pyrolysis, (2013).

8. L. Jarboe, Z. Wen, D. Choi and R. Brown, Applied Microbiology and Biotechnology, 91, (2011) 1519.

3. M.R. Rover, P.A. Johnston, B.P. Lamsal, R.C. Brown, under review, Journal of 9. L. Jarboe laboratory, unpublished. Analytical and Applied Pyrolysis, (2013). 10. J. Zaldivar, A. Martinez and L.O. Ingram, Biotechnology and 4. J.K.S. Chan and S.J.B. Duff, Bioresource Technology, 101, (2010) 3755 Bioengineering, 65, (1999) 24. 5. J. Fan, Y. Fan, Y. Pei, K. Wu, J. Wang and M. Fan, Separation and Purification 11. J. Zaldivar, A. Martinez and L.O. Ingram, Technology, 61, (2008) 324. Biotechnology and Bioengineering, 68, (2000). 6. R. Millati, C. Niklasson and M.J. Taherzadeh, Process Biochemistry, 38, (2002) 515.

https://www.cset.iastate.edu/tcbiomass2013/

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

https://www.cset.iastate.edu/tcbiomass2013

Kaige Wang and Robert C. Brown

Compare in-situ and ex-situ catalytic pyrolysis of lignocellulosic biomass using H-ZSM5 zeolite

5.0E+08 4.0E+08

ex-situ

Temperature of pyrolysis reactor: 500oC Temperature of catalytic bed: 500oC

1.5E+08

3.0E+08 2.0E+08

0.0E+00

Methods

Materials: Hybrid poplar purchased from Wood Residual Solutions (USA) and ZSM5 catalyst (CBV2314) from Zeolyst (USA) was used in this study.

2.0E+08

1.0E+08

For ex-situ catalytic pyrolysis, investigate the effects of catalyst loading and reaction temperature on yields

Pyrolysis experiment: Experiments were conducted with a Frontier 3050 Tandem Micro-Reactor System (see figure below), which consists of two furnaces connected in series. For ex-situ pyrolysis, biomass alone was pyrolyzed in the first furnace and the resulting pyrolysis vapors pass through a fixed bed of catalyst in the second furnace. For in-situ catalytic pyrolysis, a mixture of biomass and catalyst (1:20) was pyrolyzed in the first furnace while the second furnace was empty.

In-situ

Benzene Toluene Xylene Indenes

C10+

• Same reaction temperature and catalyst to biomass ratio (20:1) was used for the two methods (indenes include indene as well as alkyl-indenes; C10+ includes napthalenes and higher polyaromatics). • In-situ catalytic pyrolysis of hybrid poplar produced more aromatic hydrocarbons compared to ex-situ (twice as much for some aromatics) • Ex-situ catalytic pyrolysis produced higher olefin yield than In-situ • It is hypothesized that primary reactions of fast pyrolysis produce small enough molecules to diffuse into and react with the in-situ catalysts mixed with the pyrolyzing biomass. • It is hypothesized that secondary condensation reactions downstream of the pyrolysis reactor produce aerosols that foul and coke the ex-situ catalysts, reducing aromatic yields compared to in-situ catalysts.

Ex-situ catalytic pyrolysis: Effect of catalyst loading 2.0E+08

Peak Area

1.5E+08

Catalyst load:

10mg

20mg

40mg

Temperature of pyrolysis reactor: 500oC Temperature of catalytic bed: 500oC

500°C

600°C

700°C

Temperature of pyrolysis reactor: 500oC

5.0E+07 0.0E+00

Benzene Toluene Xylene Indenes

C10+

• Higher temperatures favor the formation of small aromatics such as benzene and toluene. • 500oC is an optimal temperature for catalytic conversion of pyrolysis vapor in terms of yield of valuable aromatics (BTX).

Ex-situ catalytic pyrolysis: Effect of pyrolysis temperature 2.0E+08 1.5E+08

400°C

500°C

600°C

700°C

Temperature of catalytic bed: 500oC

1.0E+08 5.0E+07

Benzene Toluene Xylene Indenes

C10+

• Yield of aromatics increased with increasing pyrolysis temperature.

Future work:

5.0E+07

The authors are thankful for the finical support from Iowa Energy Center

400°C

1.0E+08

0.0E+00

1.0E+08

0.0E+00

Ex-situ catalytic pyrolysis: Effect of catalytic bed temperature

Peak Area

Objectives

Ex-situ catalytic pyrolysis produced less aromatics compared with in-situ

Peak Area

Catalytic pyrolysis is a promising technology to produce olefins and aromatics including benzene, toluene, and xylene when using HZSM-5 catalyst. Generally, there are two approaches to catalytic pyrolysis: insitu and ex-situ. For in-situ catalytic pyrolysis, catalyst is introduced in the pyrolysis zone. For ex-situ catalytic pyrolysis, biomass is separately pyrolyzed and the pyrolysis vapors passed through a catalysis reactor downstream of the pyrolysis reactor.

Results

Peak Area

Introduction

Comparison of In-Situ and Ex-Situ Catalytic Pyrolysis

• Compare in-situ and ex-situ catalytic pyrolysis of carbohydrates and lignin. • Evaluate catalyst life.

Benzene Toluene Xylene

Indenes

C10+

Bioeconomy Institute, Iowa State University

tcbiomass2013

Center for Sustainable Environmental Technologies

Yanan Zhang, Guiping Hu and Robert C Brown

Life Cycle assessment of the Production of Hydrogen and Transportation Fuels from Corn Stover via Fast Pyrolysis Abstract

Comparison

Process

The goal of this project is to perform a life cycle assessment (LCA) and quantify the environmental impacts of the production of hydrogen and transportation fuels from the fast pyrolysis of corn stover and upgrading of the resulting bio-oil.

Methodology

0.037

0.25

0.117

1.7

0.0422

0.4

0.0975

1.22

0.115

1.5

Case C: Ethanol via gasification

0.15

1.2

Case D: 2005 petroleum-based gasoline

0.3

4.5

Sensitivity Analysis

8 Interpretation

• Indirect land use change is not included • Electricity from fuel mix in the Midwest region of the U.S • Corns stover is with 25% moisture and dried to 7% via pretreatment • Hydrogen is set as the avoided produce with credits • IPCC 2007 GWP 100a, TRACI 2 method • Cumulative Energy Demand (CED)

(MJ/km)

Results Fossil energy input (MJ/km)

Impact assessment

( kg CO2eq/km)

Pyrolysis-based gasoline for the current study (co-production of hydrogen and gasoline) Case A: Pyrolysis-based gasoline from forest residue (external hydrogen) Case B1: Pyrolysis-based gasoline from corn stover (hydrogen from bio-oil reforming) Case B2: Pyrolysis-based gasoline from corn stover (hydrogen from natural gas steam reforming) Case B3: Pyrolysis-based gasoline from forest residue (hydrogen from natural gas steam reforming)

This study finds that Global Warming potential for this scenario is 88% and 94% lower than for petroleum-based gasoline and diesel fuel (2005 basis), respectively.

Inventory

Fossil energy

Scenario

Input data for this analysis come from Aspen Plus modeling, a GREET model database and a U.S Life Cycle Inventory Database. SimaPro 7.3 software with Eco-invent 2.2 data base is employed to estimate the environmental impacts.

Goal and scope

GHG emissions

Gasoline fuel economy (125%; 100%; 75%)

Biochar credit

Hydrogen credit Products transportation

2 0 -2 -4

Bio-oil yield (125%; 100%; 75%)

Bio-oil upgrading

Electricity usage for pyrolysis (125%; 100%; 75%)

Bio-oil production

Electricity usage for upgrading (125%; 100%; 75%)

Biomass preprocessing

Preprocessing electricity (75%; 100%; 125%)

Biomass transportation

Nitrogen fertilizer(75%; 100%; 125%)

Biomass production

-6

Gasoline yield (125%; 100%; 75%)

Biomass transportation distance(75%; 100%; 125%)

-0.1 -0.05 0 0.05 0.1 0.15 0.2 Global Warming Potential (kg CO2eq/km on gasoline basis)

-8 Gasoline basis

Diesel fuel basis

GWP (kg CO2eq/km)

0.8 0.6

Biomass CO2 aborption

0.4

Vehicle operations Products transportation

0.2

Bio-oil upgrading

Bio-oil production

-0.2

Biomass preprocessing

-0.4

Biomass transportation

-0.6

Biomass production

-0.8 Gasoline basis

Conclusions • • • • •

Net fossil energy input is 0.25 MJ and 0.23 MJ per km traveled for a light duty vehicle fueled by gasoline and diesel fuel, respectively. In the overall system, bio-oil production has the largest fossil energy input. The Global Warming Potential (GWP) is 0.037 kg CO2eq and 0.015 kg CO2eq per km traveled for a vehicle fueled by gasoline and diesel fuel, respectively. Vehicle operations contribute up to 33% of the total positive GWP. The GWPs in this study are 88% and 94% lower than for petroleum based gasoline and diesel fuel (2005), respectively.

Diesel fuel basis

We appreciate the support of the Bioeconomy Institute and the Biobased Industry Center of Iowa State University.

https://www.cset.iastate.edu/tcbiomass2013