lignin properties contributing to monomer yildields did ...lignin properties contributing to monomer...

38
Lignin properties contributing to monomer i ld d i i l bi yields during simultaneous biomass delignification and lignin depolymerization Thanaphong Phongpreecha, Nicholas Hool, Kendall Christy, David Hodge Chemical Engineering & Materials Science Chemical Engineering & Materials Science Michigan State University 22 J 2016 22 June, 2016 www.glbrc.org

Upload: others

Post on 26-Mar-2020

6 views

Category:

Documents


1 download

TRANSCRIPT

Lignin properties contributing to monomer i ld d i i l biyields during simultaneous biomass 

delignification and lignin depolymerization

Thanaphong Phongpreecha, Nicholas Hool, Kendall Christy, David Hodge

Chemical Engineering & Materials ScienceChemical Engineering & Materials ScienceMichigan State University

22 J 201622 June, 2016

www.glbrc.org

Cellulosic DuPont DuPont –– NevadaNevada,,Iowa, USA Iowa, USA –– 2525M gal/yrM gal/yr

Ethanol AbengoaAbengoa –– Hugoton, Hugoton, Kansas, USAKansas, USA25M gal/yr25M gal/yr

RaízenRaízen ‐‐ Piracicaba, São Paulo, Piracicaba, São Paulo, BrazilBrazil12M gal/yr12M gal/yr

25M gal/yr25M gal/yr

Beta Renewables, Beta Renewables, CrescentinoCrescentino, Italy , Italy ‐‐ 25M gal/yr25M gal/yr

GranBioGranBio –– São São Miguel dos Miguel dos Campos, Campos, AlagoasAlagoas, Brazil , Brazil –– 1212M gal/yrM gal/yrPoetPoet‐‐DSM DSM –– Emmetsburg, Emmetsburg, Iowa, USA Iowa, USA ‐‐ 25M gal/yr25M gal/yr

Cellulosic DuPont DuPont –– NevadaNevada,,Iowa, USA Iowa, USA –– 2525M gal/yrM gal/yr

Ethanol AbengoaAbengoa –– Hugoton, Hugoton, Kansas, USAKansas, USA2525M gal/yrM gal/yr

RaízenRaízen ‐‐ Piracicaba, São Paulo, Piracicaba, São Paulo, BrazilBrazil12M gal/yr12M gal/yr

2525M gal/yrM gal/yr

Challenges1. Logistics – Year‐round supply, storage, transportation2. Low atom efficiency 

1000 k bi 300 k th l bi f lBeta Renewables, Beta Renewables, CrescentinoCrescentino, Italy , Italy ‐‐ 25M gal/yr25M gal/yr

e.g. 1000 kg biomass  < 300 kg ethanol biofuel3. High CAPEX relative to sugar‐ and starch‐derived ethanol4 Processing challenges – Solids handling biomass4. Processing challenges  Solids handling, biomass 

heterogeneity5. Need to diversify co‐product portfolio beyond ethanol 

GranBioGranBio –– São São Miguel dos Miguel dos Campos, Campos, AlagoasAlagoas, Brazil , Brazil –– 12M gal/yr12M gal/yrPoetPoet‐‐DSM DSM –– Emmetsburg, Emmetsburg, Iowa, USA Iowa, USA ‐‐ 25M gal/yr25M gal/yr

y p p y‐ Opportunity for co‐products from lignin if conversion and 

recovery challenges can be overcome

p‐xylene Terephthalicacid

Isobutanol

RenewablePolymersPolyethylene, 

FDCA C2H2

PLAPHAsPET, …

EthyleneAcrylic Acid

Lactic Acid

Renewable

C2H5OHEthanol

h dRenewableSugars (Glucose)

3‐hydroxy‐propionic acid

ic

Succinic Acid 1,4 butanediol(BDO) 

al 

al/Catalyti

Propylene, Isobutene, Butadiene 

Biologic

Chem

ica

Lignin Applicationsh ll

Combustion

Gasification

Challenges1. Structural complexity and variability 

of ligninsSynthesis of fuels and chemicals

Polymeric material

of lignins2. Process lignins often significantly 

modified, reduced reactivity/utility 3 Need to isolate & purify the processPolymeric material

Dispersant, emulsifier, stabiliser or sequestrant

3. Need to isolate & purify the process lignins

4. Lignin conversion processes tend to Materials: 

• Adhesives, thermosetresins

require severe conditions  5. Complex mixtures of products

• Low yields of single compounds, • Polymer modifiers

Carbon fiber

Aromatic monomers

purification required6.  Costs vs. petrochemical alternatives

Aromatic monomers

Scope of PresentationScope of Presentation

1. Linking Fractionated, Alkali Hardwood Lignin Properties to Aromatic Monomer YieldsProperties to Aromatic Monomer Yields

2. Integrated Catalytic Deconstruction of Woody Biomass: Simultaneous Delignification and Lignin Conversion

1. Linking Fractionated, Alkali Hardwood Lignin Properties to Aromatic Monomer YieldsProperties to Aromatic Monomer Yields

Fractionated Hybrid Catalytic Depolymeri ation

Aromatic Poplar Lignins Catalytic Depolymerization Monomers

Correlating Properties to Yields

1. Linking Fractionated, Alkali Hardwood Lignin Properties to Aromatic Monomer Yieldsp

• Lignins– Mild soda pulping of hybrid poplar

Goal: Relate lignin properties to• Fractionation 

– SLRP process– Unpressurized CO2; ambient T

properties to product yields

Unpressurized CO2; ambient T• Conversion approaches– Catalytic oxidationThi id l i– Thioacidolysis

• Lignin characterization– 1H NMR ArOH; Aliphatic OH; p– 13C NMR β‐O‐4– HSQC NMR β‐β, β‐5, β‐O‐4, S/G GPC Molecular weight– GPC  Molecular weight

– GC/MS  Lignin products

Impact of Delignification Conditions on Poplar Enzymatic Hydrolysis Yields

Enzymatic Hydrolysis

Alkaline Delignification

p y y y

St kl t l (2015)

HydrolysisDelignification

LigninsPopulus nigra x )

80

90

100

Stoklosa et al. (2015). Bioenerg Res., 8:1224‐1234.

maximowiczii cv. NM6)

40

50

60

70

80

ose Yield (%

)

Pulping Conditions:Pulping Conditions:Pulping Conditions:

0

10

20

30

Gluco Pulping Conditions: 

170°C, 1 hour54% biomass/pulp yield85% of lignin removed

Pulping Conditions: 150°C, 3 hour67% biomass/pulp yield62% of lignin removed

Pulping Conditions: 150°C, 30 minutes78% biomass/pulp yield51% of lignin removed

0 24 48 72 96 120

Time (hours)

H‐Factor = 579 H‐Factor = 1430H‐Factor = 166

Lignocellulose Deconstruction and ConversionLignocellulose Deconstruction

Enzymatic Hydrolysis

Alkaline Delignification HydrolysisDelignification

LigninsPopulus nigra x ) Ligninmaximowiczii cv. NM6)

Deconstruction

Lignin (Fractionation? +) Depolymerization

IntermediatePhenolic Monomers Lignin Oligomers

IntermediateProducts

Additional Processing

Additional Processing

Additional Processing

Conversion of Intermediates

“Other” Lignins

Final Products(Targeting Diverse 

k S )

Processing Processing Processing

High‐value flavor and fragrance 

d

Bulk Aromatics

Novel bio‐based foams, adhesives 

d ti

Intermediates

Lignin‐derived solid fuels for 

h tMarket Sectors)

Increasing Product Value

compounds and coatings

Product 1      Product 2       Product 3           Product 4

process heat

Lignin Fractionation by CO2 Acidification

1 SLRP Process Recovered Lignin Stoklosa et al., 2013. Green Chem. 15:2904‐29121.  SLRP Process

• CO2 acidification of alkaline black liquor at elevated temperature pH 11 10 5

gFractions

liquor at elevated temperature and pressure– Generation of hydrated lignin phase

d l d

pH 11, 10.5

– Liquid‐liquid separation

2.     Atmospheric acidificationpH 2, 9, 10, 10.6, 11

• 50 mg/g catalyst (CuSO )

Lignin Depolymerization: Oxidationh d d d/50 mg/g catalyst (CuSO4) 

loading; 2 M NaOH; T = 180°C; 0.5 MPa O2

• Maximum yields achievable

• Further degradation and/or repolymerization of monomers

• Maximum yields achievable at 5‐7 minutes

14

16

Lignin 

8

10

12

d on

 Klason 

ass Ba

sis)

2

4

6

onom

er Yield

(% M

a

0

2

0 3 5 7 10 14

Mo

Time (min)

Lignin Depolymerization: Hydrogenolysis• Lignin depolymerization in ethanol with Ni/CLignin depolymerization in ethanol with Ni/C catalyst at 220˚C

• No addition of H• No addition of H2

– Simultaneous alcohol reforming and hydrogenolysis

ld h• Lower yields than                                               thioacidolysis or                                               

doxidation – Lower solubility?

a

– Contaminating                                                             metals?

ssing  Data

Mi

Atmospheric Fractionation SLRPpH    2         9         10      10.6    11      10.5    11

Monomer Yields as a Function of Lignin Recovery Conditions 7

8

9

10

12

14

16

18

drogen

olysis

Oxidation

of  Lignin Recovery Conditions• Diverse range of yields

• Comparable trends between3

4

5

6

4

6

8

10

mer Yield by Hyd

nomer Yield by O

• Comparable trends between methods 0

1

2

0

2

4

0 5 10 15 20

Mon

om

Mon

Monomer Yield by Thioacidolysis

16

18

20

Lignin

Acetosyringone AcetovanilloneSyringic acid Vanillic acidSyringaldehyde Vanillin4‐Hydroxybenzaldehyde 16

18

20

Lignin 

ass)

p‐Hydroxyphenyl Monomers

Guaiacyl Monomers

Syringyl Monomers

8

10

12

14

eld on

 Klason L

Mass Ba

sis)

10

12

14

elds on Klason

 olyis (%

 by Ma

2

4

6

8

Mon

omer Yie

(% M

4

6

8Mon

omer  Y

ieby

 Thioacid

0

2

2.00 9.00 10.00 10.60 11.00 10.5‐10 11‐10.5 12.0‐11

pH

0

2

2.00 9.00 10.00 10.60 11.00 10.5‐1011‐10.512.0‐11

pHAtmospheric Fractionation SLRPpH    2        9      10    10.6   11    10.5   11

Atmospheric Fractionation SLRPpH    2       9    10  10.6  11  10.5 11

Lignin Aromatic and Aliphatic Hydroxyl• 1H NMR to quantify MeO ArOH Aliphatic OH• 1H‐NMR to quantify MeO, ArOH, Aliphatic OH• Normalized to a per monomer basis using MeO and S/G ratio• Strong correlation between aromatic hydroxyl content and g y yalkaline solubility– pKa of ArOH ≈ 10.5

• Correlation with other properties?• Correlation with other properties?

1 60

1.801.00 % molPhOH % mol AlOH

r mer

1.20

1.40

1.60

0.70

0.80

0.90

r mon

omer

per mon

om

Ph lhβ l0.60

0.80

1.00

0.50

0.60

0.70ArO

Hpe

r

phatic OH p

‐(β‐O‐4)

Phenylcoumaran(β‐5 + α‐O‐4)

etherβ aryl 0.400.40 A

lip

Atmospheric Fractionation SLRPpH   2      9    10   10.6 11  10.5 11

Alkyl‐Aryl Ether Content • β‐O‐4 contents estimated by integrating Cγ peak of β‐O‐4 per MeO with S/G ratios taken into account

• Clear trend of decreasing β‐O‐4 content with decreasing pH of separation

QuantitativeQuantitative 13C NMR β-O-4 per 100 Ar Mn

7000

8000

30

35

Mnβ‐O‐4

4000

5000

6000

15

20

25β O 4

1000

2000

3000

5

10

δ (ppm)

002 9 10 10.6 11 10s 11spH     2      9      10    10.6  11    10.5   11Atmospheric Fractionation SLRP

2‐D 1H‐13C HSQC NMR: Lignin Structures

β‐O‐4 content βas recovery pH 

) m)

13C δ(ppm

)

13C δ(ppm

1

1H δ (ppm)1H δ (ppm)

Comparison of HSQC NMR Results

(ppm

)

δ(ppm

)

13C δ

13C δ

1H δ (ppm)1H δ (ppm)

Correlating Properties

R² = 0.98920.60

0.70

0.80

0.90

per A

r)

0.20

0.30

0.40

0.50

ArO

H Con

tent (

0.00

0.10

0 10 20 30 40

A

β‐O‐4 Content by 13C NMR (per 100 Ar)

5000

6000

7000

8000

age MW

R² = 0.9857

1000

2000

3000

4000

Num

ber A

ver

0

1000

0 10 20 30 40

β‐O‐4 Content by 13C NMR (per 100 Ar)‐1        ‐0.5        0          0.5       1

Correlating Properties

R² = 0.98920.60

0.70

0.80

0.90

per A

r)

0.20

0.30

0.40

0.50

ArO

H Con

tent (

0.00

0.10

0 10 20 30 40

A

β‐O‐4 Content by 13C NMR (per 100 Ar)

5000

6000

7000

8000

age MW

R² = 0.9857

1000

2000

3000

4000

Num

ber A

ver

‐0.99

0.99

0

1000

0 10 20 30 40

β‐O‐4 Content by 13C NMR (per 100 Ar)‐1        ‐0.5        0          0.5       1

Relating Monomer Yields to Lignin Properties

‐1        ‐0.5        0          0.5       1

Relating Monomer Yields to Lignin Properties

• Clear correlations betweenClear correlations between many properties

i ld l l d d• Yields clearly dependent on initial β‐O‐4 content

‐1        ‐0.5        0          0.5       1

Prediction of Maximum Possible Monomer Yields from LigninMonomer Yields from Lignin

triadn‐3

m3d1 m4

d2 d3m1 m2 mn 1dn‐3 mdn‐2 dn‐1mn 2mn 3

triad1

mi

triadn‐2

m3 m4m1 m2 mn‐1 mnmn‐2mn‐3

triad2

mi

n‐8n = number of monomers in polymer m is a monomerp yn‐1 = number of dyads (adjacent monomer pairs)

that could be linked by an aryl ethern 2 = number of triads (adjacent dyads)

d is a dyadFlory and Fujiwara. (1969). Macromol. 2(4):315‐327

• Simplified model of lignin assuming: 

n‐2 = number of triads (adjacent dyads) ( )

– Population of linear polymers of length n with randomly distributed β‐aryl ethers (no cross‐linking/branching) 

– Depolymerization is only C–O cleaving

Prediction of Maximum Possible Monomer Yields from LigninMonomer Yields from Lignin

triadn‐3

m3d1 m4

d2 d3m1 m2 mn 1dn‐3 mdn‐2 dn‐1mn 2mn 3

triad1

mi

triadn‐2

m3 m4m1 m2 mn‐1 mnmn‐2mn‐3

triad2

mi

n‐8n = number of monomers in polymer m is a monomerp yn‐1 = number of dyads (adjacent monomer pairs)

that could be linked by an aryl ethern 2 = number of triads (adjacent dyads)

d is a dyadFlory and Fujiwara. (1969). Macromol. 2(4):315‐327n‐2 = number of triads (adjacent dyads) ( )

• Probability that any triad contains two adjacent β‐aryl ethers is the β‐aryl ether content squaredethers is the β aryl ether content squared

• P((A ∩ B)  C) = P(A) P(B) + P(C)• Expected value for yield is:

[ ] [ ])Content4-O-(2)Content4-O-()2n(n1YieldE 2 β×+β×−=

Expected value for yield is: 

Prediction of Maximum Possible Monomer Yields from LigninMonomer Yields from Lignin

• Probability that any triad contains two adjacent β‐aryl ethers is the β‐aryl ether content squaredethers is the β aryl ether content squared

• P((A ∩ B)  C) = P(A) P(B) + P(C)• Expected value for yield is:

[ ] [ ])Content4-O-(2)Content4-O-()2n(n1YieldE 2 β×+β×−=

Expected value for yield is: 

Prediction of Maximum Possible Monomer Yields from LigninMonomer Yields from Lignin

• Expected value for monomer yield is: 

[ ] [ ])Content4-O-(2)Content4-O-()2n(n1YieldE 2 β×+β×−=

p y

Prediction of Maximum Possible Monomer Yields from Lignin

• Thioacidolysis yields match model prediction

• Higher yields for oxidation due to cleavage of C C bonds?

Monomer Yields from Lignin

• Higher yields for oxidation due to cleavage of C–C bonds? • Lower yields for hydrogenolysis

16%

12%

14%

16%

 Predicted

 l/mol)

ThioacidolysisOxidationHydrogenolysis

6%

8%

10%

Mean Max. 

er Yield (m

ol n = 10n = 15n = 25

2%

4%

6%

Measured or 

Mon

ome n  100

0%0% 5% 10% 15% 20% 25% 30% 35%

M

β‐O‐4 Content

Prediction of Maximum Possible Monomer Yields from Lignin

• Thioacidolysis yields match model prediction for: – Fractionated alkali lignins from hybrid poplar (NM6)

Monomer Yields from Lignin

– In situ lignins from diverse Populus trichocarpa genotypes

– Extracted lignins from silver birch (Betula pendula)

ed Yield n

Birch CEL Lignin

r Qua

ntifie

edicted or

Pre

Scope of PresentationScope of Presentation

1. Linking Hardwood Lignin Properties to Aromatic Monomer YieldsMonomer Yields

2. Integrated Catalytic Deconstruction of Woody Biomass: Simultaneous Delignification and Lignin Conversion

2. Integrated Catalytic Deconstruction of Woody Biomass ‐ Backgroundy g

• Background ‐ Hydrogenolysis– Challenge of low yields from technical lignins due to modification during lignin removal/processing

– High temperature and high H2 pressure                                              required at the expense of selectivity

Meier et al. 1994. Biomass Bioenerg. 40(2):171-177required at the expense of selectivity

– Ni/C capable of H‐transfer from alcohols to lignin with potential for integration with organosolv pretreatment to achieve high 

Bioenerg. 40(2):171 177

monomer yields

• Scope of workWang and Rinaldi, 2012. Energ Environ Sci. 5: 8244-8260; Song, et al. 2013. Energ Environ Sci. 6, 994-1007.

– Assess select metal catalysts on an inert support as potential bi‐functional catalysts for simultaneous delignification, alcohol reforming, and hydrogenolysis of lignin directly from birch woodg, y g y g y

– Characterize the effect of lignin removal on enzymatic hydrolysis 

– Identify descriptors for hydrogenolysis of lignin model

2. Integrated Catalytic Deconstruction of Woody Biomassof Woody Biomass

Additional Valorization by Catalytic or Biochemical Conversion4. 4. 4.

Sorbitol2. Hydrogenolysis;Hydrogenation

4.

H2Xylitol

4‐propylsyringol4‐propylguaiacol

2

1. Fractionation  +Catalytic H‐transfer

LigninXylan

“Other” 3. Hydrolysis

D‐XyloseXylan

CelluloseD‐GlucoseH‐donating Solvent

Enzyme‐or Acid‐Catalyzed

Cellulose

Si l d li ifi i l h l f i d• Simultaneous delignification, alcohol reforming, and hydrogenolysis

Yields and Selectivities at 200°CMonomer Selectivities

(%)

Yi ld b d l i i i l li i• Yields based on total initial lignin

• High selectivities for small range of monomer products

• Cleavage of C–C bonds in side chain region for Pd

Yields and Selectivities at 220°C

(%)

Monomer Selectivities

• Exceptionally high yields for several systemsExceptionally high yields for several systems• No evidence of aromatic ring saturation• Increasing saturation of side chain region for Ru, Co, Ni,      and Pd– Reaction products provide evidence for simple reaction pathway

Enzymatic Hydrolysis Yields and DelignificationVisible macroscopic differences and100 Lignin removal Visible macroscopic differences and differences in enzymatic hydrolysis yields due to differences in surface lignin adsorption

60

70

80

90

olysis Yield

gcorrelated to hydrolysis yields 

20

30

40

50

Glucose Hydro

220 C

200 C

90

100

(%)

220 C200 C

0

10

0 20 40 60 80 100

G

Lignin Removal

200 C

100d

°°

50

60

70

80

olysis Yields 

60

70

80

90

100

olysis Yield

Co

Ni

RuPd

ed

10

20

30

40

ucose Hydro

o Data

20

30

40

50

lucose Hydro Highest monomer yields comparable to highest sugar yields U

ntreate

0

10

None Fe Co Ni Cu Zn Ru Pd Ag

Glu No

0

10

0 20 40 60

G

Aromatic Monomer Yield

Ongoing Work – DFT and MD Studies

Identifying potential descriptors for lignindescriptors for lignin model depolymerization

Adsorption mechanism conformation forAdsorption mechanism, conformation for hypothetical lignins on Ni (1 1 1) surface

Summary

• Lignins can be enriched in select properties by fractionationfractionation

• In fractionated alkali lignins many lignin properties can be correlated to each other and to monomercan be correlated to each other and to monomer yields

• β‐O‐4 content is an important predictor of monomer• β‐O‐4 content is an important predictor of monomer yields from fractionated and native lignins

• Unmodified in situ lignins capable of highest yields• Unmodified in situ lignins capable of highest yields

• Potential for integrated biomass deconstruction processprocess

Acknowledgements

CollaboratorsCollaborators• Cliff Foster, GLBRC• Yue Qi, MSUM k Thi Cl

Funding:• U.S. Department of Agriculture: 

• Mark Thies, ClemsonNortheast Sungrant Initiative

Research Group:Dr Ryan Stoklosa Jacob Crowe Thanaphong “Joe” PhongpreechaDr. Ryan Stoklosa, Jacob Crowe, Thanaphong Joe  Phongpreecha, Lisaura Maldonado‐Pereira, Georginelly Ferreira Inácio, Angel Santiago‐Colón, Nicholas Hool, Ben Gardner, Nick Ferringa, Kendall Christy

Thank you!

Questions? David HodgeDavid Hodge

Chemical Engineering Mi hi St t U i itMichigan State University Email: [email protected]

Web: http://www.chems.msu.edu/groups/hodge/