UNDERSTANDING CHAR AND TERRA PRETA SOIL CHEMISTRY FROM PYROLYSIS MASS

SPECTROMETRIC ANALYSIS

K. Magrini, S. Czernik, R. Evans

October 6, 2008

OUTLINE

Rapid SOC AnalysisPy-MBMS Instrument and MethodsMultivariate Analysis

ResultsCharacterized and Managed SoilsTerra Preta Soils

Conclusions

MOLECULAR BEAM MASS SPECTROMETRY

• Combustion mode provides

C, N, and S concentrations

• Pyrolysis mode provides

SOC chemistry

• Handles large number of samples (100-300 samples/day)

• Analysis time is short (2-5 minutes)

Transportable MBMS

150-mg Soil Samples

0

50

100

150

200

250

300

0 2 4 6 8 10 12 14Time, Min.

Tota

l Ion

Inte

nsity

, Arb

itrar

y U

nits

#1719855-15 cm1.22% C #24

198515-30 cm1.59% C

#2319855-15 cm1.65% C

#2219850-5 cm4.16% C

#21198515-30 cm0.79% C

#2019855-15 cm1.52% C

#1919850-5 cm5.46% C

#18198515-30 cm0.94% C

0

50

100

150

200

250

300

0 2 4 6 8 10 12 14Time, Min.

Tota

l Ion

Inte

nsity

, Arb

itrar

y U

nits

#1719855-15 cm1.22% C #24

198515-30 cm1.59% C

#2319855-15 cm1.65% C

#2219850-5 cm4.16% C

#21198515-30 cm0.79% C

#2019855-15 cm1.52% C

#1919850-5 cm5.46% C

#18198515-30 cm0.94% C

py-Reactor

Pyrolysis MS Traces

Heated Quartz Reactore-

Argon Collision Gas

Collisions(P+) (D+)

DetectorQ1 Q2 Q3Soil

Samples

Heated Quartz Reactore-

Argon Collision Gas

Collisions(P+) (D+)

DetectorQ1 Q2 Q3Soil

Samples

CID provides a mass spectrum of a specific fragment ion for unique identificationi.e. Mass 396 is fragmented to identify ergosterol as the parent species

• Correlate with other spectroscopic techniques MIR, NIR, DRIFTS• Correlate with other characterization data• Build comprehensive SOC database for research use

0

50

100

150

200

250

300

0 2 4 6 8 10 12 14Time, Min.

Tota

l Ion

Inte

nsity

, Arb

itrar

y U

nits

#1719855-15 cm1.22% C #24

198515-30 cm1.59% C

#2319855-15 cm1.65% C

#2219850-5 cm4.16% C

#21198515-30 cm0.79% C

#2019855-15 cm1.52% C

#1919850-5 cm5.46% C

#18198515-30 cm0.94% C

0

50

100

150

200

250

300

0 2 4 6 8 10 12 14Time, Min.

Tota

l Ion

Inte

nsity

, Arb

itrar

y U

nits

#1719855-15 cm1.22% C #24

198515-30 cm1.59% C

#2319855-15 cm1.65% C

#2219850-5 cm4.16% C

#21198515-30 cm0.79% C

#2019855-15 cm1.52% C

#1919850-5 cm5.46% C

#18198515-30 cm0.94% C

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

20 70 120 170 220 270 320 370 420m/z

Ion

Inte

nsity

43,44

2728

5557 94 109

256135

8068

PRINCIPAL COMPONENT ANALYSIS AND PLS MODELING

A projection method that helps to visualize information

• PCA helps to determine which samples are different and which variables are contributing

• to the difference

• Helps to determine groupings

• Helps quantify the amount of useful information

• PLS builds predictive models with quantitative information

Py-MBMS SOC ANALYSIS

• Analytical pyrolysis coupled with molecular beam mass spectrometry (py-MBMS) and multivariate statistics for rapid soil organic carbon (SOC) chemistry analysis of varied, characterized soils to establish breadth of application

RESULTS• Forest soils: distinguish disturbance,depth and location• Agro forest: SOC accumulation with poplar rotation (litter

and root inputs with depth)• Grassland CRP soils: distinguish management impacts• Native prairie soils: quantitating SMBC, SOC, POM C, Cmin,

fractions• Agricultural management for SOC accumulation (switchgrass)• Terra preta chemistry

Soil characterization data

PARAMETER RANGE

Depth (cm) 0-225

SOC (wt %) 0.02-9.8

SMBC (µg/g soil) 3.6-4933

POM C (g/g soil) 0.01-7.73

Cmin C (g/g soil) 0.61-1.16

Calendar Age (YBP) 0-16157 yrs

Native prairie soils: quantitating SMBC, SOC, POM C, C min

SOC PLS 1 modelfor all samples

• Lacustrine soils different

• Complex SOC spectra containing lignin, fatty acidsergosterol, and carbohydrate species

-1

0

1

2

3

4

5

6

7

8

9

0 2 4 6 8 10 12

Measured Y

Pred

icte

d Y

Measured YPredicted Y

Lacustrinesoils

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

25 125 225 325 425

m/z

regr

essi

on C

oeffi

cien

ts

368340

374

420

438

392396

254

270 312296

115117

105

919663

585960

7379 137

150164168 211

130144

7780

170172

125 Fatty acids

Lignin andcarbohydrates

SOC model regression coefficients

ESTIMATING SOC

Magrini et al., 2007.

-500

0

500

1000

1500

2000

2500

3000

0 500 1000 1500 2000 2500 3000

Measured Y

Pred

icte

d Y

MeasuredPredicted

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

25 125 225 325 425

X Variables

Reg

ress

ion

Coe

ffici

ents

394

282

96

111

81

60

107

256137138

121

84

6853

9173

213227

185

199

157

Palmitic acid: 74, 87, 101, 115, 129, 143, 157, 171, 185, 199, 213, 227, 256SMBC biomarker

SMBC PLS 1 model for all samples

SMBC model regression coefficients

Magrini et al., 2007.

ESTIMATING SMBC

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.00 0.10 0.20 0.30 0.40 0.50

Measured Y

Pred

icte

d Y

MeasuredPredicted

PLS1 model regressing SMBC vs. mass 256

Single mass has potential to be used as an SMBC biomarker

Magrini et al., 2007.

ESTIMATING SMBC FROM MASS 256

SUMMARY OF PLS MODELS CONSRUCTEDWITH SAMPLE CHARACTERIZATION DATA

Variable # of Samples Correlation # Principal Components

Slope RMSEP

% Carbon 90 0.96 6 0.92 0.37POM C 27 0.96 3 0.92 0.61C min 28 0.92 1 0.83 0.53SMBC 48 0.96 6 0.97 173

14C Calendar Age 20 0.91 2 0.83 2210

K. A. Magrini, R. F. Follett, J. Kimble, M. F. Davis, and E. Pruessner.2007. Soil Science.

Depth 0-225 cm

SOC 0.02-9.8 wt%

SMBC 3.6-4900 µg/g

POM C 0.01-7.7 µg/g

C min 0.6-1.2 g/g

Calendar Age 0-16000 yrs

-3

-2

-1

0

1

2

3

4

-20 -10 0 10 20

2

22

2

22

2

23

33

3

3

3

331

1

1

1

1 1

11

1

44

4

4

4

4

4

4 4PC1

PC2 Scores

Native

NoTill

18-yrCRP

ContinuousTill

Increasing total soil carbon <1.5 wt%3.0 wt%

0-5 cm

IncreasingSOM

-3

-2

-1

0

1

2

3

4

-20 -10 0 10 20

2

22

2

22

2

23

33

3

3

3

331

1

1

1

1 1

11

1

44

4

4

4

4

4

4 4PC1

PC2 Scores

-3

-2

-1

0

1

2

3

4

-20 -10 0 10 20

2

22

2

22

2

23

33

3

3

3

331

1

1

1

1 1

11

1

44

4

4

4

4

4

4 4PC1

PC2 Scores

Native

NoTill

18-yrCRP

ContinuousTill

Increasing total soil carbon <1.5 wt%3.0 wt% Increasing total soil carbon <1.5 wt%3.0 wt%

0-5 cm

IncreasingSOM

-4

-3

-2

-1

0

1

2

3

4

-15 -10 -5 0 5 10 15

2

2

2

2

2

2

22

3

3

3

3

3

31

1

1

1

11

4

44

4

4

PC1

PC2 Scores

Native

CRP

CT

NT

5-10 cm

-4

-3

-2

-1

0

1

2

3

4

-15 -10 -5 0 5 10 15

2

2

2

2

2

2

22

3

3

3

3

3

31

1

1

1

11

4

44

4

4

PC1

PC2 Scores

Native

CRP

CT

NT

5-10 cm

-4

-3

-2

-1

0

1

2

3

-15 -10 -5 0 5 10 15

2

2

2

2

2

22

2

3

3

3

3

31

1

1

1

1

1

4

4

4

4

4

4

PC1

PC2 Scores10-30 cm

CRP

Native

CT

NT

Samples from Idaho farm in CRP 18 years.

PCA shows land mayneed to be in CRP longer to recoverto native standard

SOC levels.

Can we determine sequestration

potentials of soil types and

management impacts?

CRP and CT soils likely influencedby specific crop roots at 5-10 cm

• NT and CT: corn (shorter roots)• CRP: switchgrass (longer roots)

AT 10-30 cm, CT and CRP soilshave (CT-short roots) and switchgrass (CRP-long roots) crops. Possibly seeing SOC inputs from each root type

CRP Impacts on SOC Kimble 2003

Model for SOC ESTIMATION IN CRP SOILS

-3.00E-04

-2.00E-04

-1.00E-04

0.00E+00

1.00E-04

2.00E-04

3.00E-04

4.00E-04

5.00E-04

50 100 150 200 250 300 350 400 450

X-Variables

Reg

ress

ion

Coe

ffici

ents 60

68

78

96

110

114115

128132

142144

156

104106

53

396382

368340278280284

256161163 175 189 225

151137121123125

11111397

99

838569

71

5557

168

Fatty Acid Fragments (-CH 2-) 14 amu units55-69-83-97-111-125

57-71-85-99-113

Fatty AcidsLinoleicLinolenicStearic

308326 354

73 126

82

Carbohydrates60, 73, 82, 114

94

92

Monoaromatics78, 92, 106

Phenolics94, 110

Plant-derived

Microbial Contributions?

-20

-15

-10

-5

0

5

10

15

-30 -20 -10 0 10 20 30 40

X-expl: 56%,9%

HA

NH

WSWS HA

NHWS

HA

NH WSWS

HA

NH

WSWSHA

NH

WS

HA

NH

WS

WS

NHWS

HA

NH

WS

WS PC1

PC2Scores

Cropped

AKRON, CO SOIL FRACTIONS

Native

WS: whole soilHA: humic acidNH: non hydrolyzable

0

2

4

6

8

10

12

14

16

18

25 75 125 175 225 275 325 375 425 475

m/z

Ion

Inte

nsity

Akron, CO Whole Soil Complex spectrum Lignin and carbohydrates present

44 amuCO2

9194959698

555760

818284

110

124138

150164

184

210

264

110, 124, 138, 150 amu "Lignin"

Aromatics

Mass spectra of Akron, COwhole soil and non hydrolysablefraction

0

2

4

6

8

10

12

14

16

18

25 75 125 175 225 275 325 375 425 475

m/z

Ion

Inte

nsity

284

9798 Aromatics91-95

44

57

83

138

194196

110

256

124150 Fatty

acids

Masses at 71, 83, 91, 111, 122, 252, 262, 284: may indicate fatty acids likely derived from microbialbiomass during decomposition

Non-hydrolyzable soil fractionLess complex spectrumEnriched in aromatics

m/z 94

OH

CH3

OH

OMe

NH O

M/z 67

M/z 164

Py-MBMS SHOWS:

• Fractions have different chemistry

• Humic acid similar spectrum to whole soil; enriched mass 91

• Lignin fragments and microbial products present in whole soil, humic acid, and humin fractions

• Non acid hydrolyzable fraction contains fatty acids. Less complex spectra than the whole soil and other

fractions – similar to terra preta

Principal Component 1Scores Terra Preta

C. Steiner – TP samples

Principal Component 1 Loadings – Terra Preta

Principal Component 2Scores Terra Preta

Principal Component 2 Loadings – Terra Preta

Palmitic Acid

Microbial BiomassMarkers (n-acetyl glucosamine)

Terra Preta and Forest Soils

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

20 70 120 170 220 270 320 370 420

m/z

Ion

Inte

nsity

Forest Soil

TerraPreta

44 amu

0

2

4

6

8

10

12

14

16

18

25 75 125 175 225 275 325 375 425 475

m/z

Ion

Inte

nsity

284

9798 Aromatics91-95

44

57

83

138

194196

110

256

124150 Fatty

acids

0

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 50 100 150 200 250 300 350 400 450 500

110

91, 94-6

137

Terra preta

Non hydrolyzablefraction AK, CO

NH and TPspectra aresimilar to each other and to highlycharredbiomass

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

20 70 120 170 220 270 320 370 420

m/z

Ion

Inte

nsity

43,44

2728

5557 94 109

256135

8068

TERRA PRETA MASS SPECTRUM

Proteins

Fatty Acids

CONCLUSIONS• Py-MBS rapidly characterizes SOC species

• Data can be used to estimate SOC and SMB contents

• Terra Preta soils have less SOM complexity and aresimilar to non hydrolyzablesoil fractions

• Working on characterizing“black carbon”

PS + NPK

PS + 5% Char

Potting Soil (PS)

Figure 1: Corn plant roots 35 days post germination in potting soil (PS), PS and NPK addition, PS and 5% peanut char.

ACKNOWLEDGEMENTS

NREL LDRD Program

USDA NRCS and ARS

USDA FOREST SERVICE

BOISE-CASCADE

CHRIS STEINER