Dilute Sulfuric Acid Pretreatment of Corn Stover in a Pilot-Scale Reactor:

Investigation of Yields, Kinetics, and Solids

Enzymatic Digestibilities

Daniel J. Schell, Jody Farmer, Millie Newman, and James D. McMillan

National Bioenergy Center

National Renewable Energy Laboratory

Golden, CO

Introduction

Dilute sulfuric acid pretreatment of corn stover was performed in a continuous pilot-scale reactor

Data was collected on xylose hydrolysis yields and enzymatic digestibilities of the pretreated cellulosic solids

The xylose yield data was used to calculate kinetic rate parameters for a biphasic xylan hydrolysis model incorporating formation of intermediate xylo-oligomers

Materials and Methods Corn stover

Used approximately 9 month old stover that was tub-ground and washed and dried to 10% moisture

Analytical Performed solids compositional analysis of raw and pretreated

solids Liquor

Measured concentration of monomeric and total sugars, acetic acid, furfural, and HMF by HPLC; measured pH

Measured insoluble solids in the pretreated slurry

Enzymatic digestibility SSF was performed on the washed pretreated solids at 6%

cellulose loading, 15 FPU/g cellulose enzyme loading, 32C Calculated cellulose conversion from ethanol yield results

Pretreatment

41 pretreatments runs at a variety of conditions were conducted in a continuous 1 dry ton/d pilot-scale reactor (see flow diagram)

Pretreatment conditions 165-183C 3-12 min 0.5-1.4% (w/w) nominal acid concentration 20% (w/w) total reactor solids concentration

Pretreatment Experimental Protocol

Steady state operating conditions were established and then samples were taken; data acquisition system collected sensor data every 30s

Samples taken Raw feedstock (for moisture) Pretreated slurry Vapor vent streams

Calculations Xylan hydrolysis yields (monomeric and oligomeric xylose,

furfural, unconverted xylan) were calculated from stream composition and flow rate data

Results were also analyzed using a combined severity factor (CSF)

pHTtLogCSF

=

75.14100exp10

Dilute Acid Kinetics

2

43

1

k

kk

k

xylanreactingSlowproductionDecompositXyloseoligomersXylose

xylanreactingFast

+= RTEkkk ipHHi

oii exp)}10({

An Excel add-in (Evolver) using a genetic algorithm was used to estimate the parameter values (fast hydrolyzing xylan fraction, kio, kiH, and Ei) by minimizing an error term composed of the sum of the square of the errors between measured and predicted values for monomeric and oligomer xylose and unconverted xylan.

Pretreatment System

Figure 1. Monomeric and Total Xylose Yields

0

10

20

30

40

50

60

70

80

0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1Combined Severity Factor

Xylo

se Y

ield

(%)

MonomericTotal

Figure 2. Xylan Fractionation and Mass Balance Closure

0

20

40

60

80

100

12015 14 13 7 4 12 22 10 26 23 34 21 17 24 29 18 19 33 37 1 2 3 20 27 6 5 38 11 36 35 8 16 9 28 25 30 31 32

Run #

Yiel

d (%

theo

retic

al)

Monomeric Xylose Oligomeric Xylose Xylan Furfural

Figure 3. Enzymatic Digestibilities

55

60

65

70

75

80

85

90

95

1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1

Combined Severity Factor

Cel

lulo

se C

onve

rsio

n (%

)

Highlight Results - Experimental

Monomeric and total xylose yields reach a peak of 70-71%, but at different pretreatment severities (Fig. 1)

With increasing severity, furfural yields increase and unconverted xylan and oligomeric xylose yields decrease (Fig. 2)

Mass balance closure ranges from 85-105%, but are significantly lower at higher severities (Fig. 2)

Cellulose conversion increases with increasing severity reaching a value of 87% (Fig. 3)

Figure 4. Predicted versus Experimental Yields

0

10

20

30

40

50

60

70

80

0 10 20 30 40 50 60 70 80

Experimental Yield (%)

Pred

icte

d Yi

eld

(%)

Monomeric Xylose

Oligomeric Xylose

Unconverted Xylan

Figure 5. Kinetic Modeling Results:Based on Maximizing Monomeric Xylose

35

40

45

50

55

60

65

70

75

80

85

90

160 165 170 175 180 185 190

Temperature (C)

Max

imum

Xyl

ose

Yiel

d (%

) pH 1.0

Total xyloseMonomeric xylose

pH 1.5

pH 2.0

Experimental data range

Figure 6. Kinetic Modeling Results Based on Maximizing Monomeric or Total Xylose

35

45

55

65

75

85

95

160 165 170 175 180 185 190

Temperature (C)

Max

imum

Xyl

ose

Yiel

d (%

)

Maximizing total xylose yield

Total xyloseMonomeric xylose

Maximizing monomeric xylose yield

Total xyloseMonomeric xylose

pH 1.0

pH 2.0

Lines for monomeric and total xylose yield at pH 1.0 lie on top of each other regardless of maximization criteria

Highlight Results - Kinetic Model The kinetic model predicts xylan hydrolysis

performance across the data set fairly well (Fig. 4) Low pH and high temperatures are required to

achieve the highest xylose yields (Fig. 5) Total xylose yields are greater than monomeric

xylose yields, although the differences are small at low pH (Fig. 5)

At low pH, xylose yields are the same regardless of which criteria is used to maximize yields (Fig. 6)

At high pH, higher total xylose yields can be achieved by maximizing on total xylose (Fig. 6)

Acknowledgements

The Biochemical Conversion Element of the Office of Fuels Development of the US Department of Energy funded this work. We gratefully acknowledge the technical assistance of Bob Lyons and Bryan Rohrbackwith pilot plant operation, Raymond Ruiz, David Templeton, Amie Sluiter, and Bonnie Hames with chemical analysis of process samples, and David Templeton with procurement of the corn stover.