UNIVERSITY OF CINCINNATI

September 26, 2006

Date:___________________

Fuyan Liang
I, _________________________________________________________,
hereby submit this work as part of the requirements for the degree of:
Doctor of Philosophy
in:
Environmental Engineering
It is entitled:
Composition and Formation Mechanism of Diesel Particulate Matter
Associated with Various Factors from A Non-road Diesel Generator

This work and its defense approved by:

Dr. Mingming Lu
Chair: _______________________________
Dr. Tim C. Keener
_______________________________
Dr. M. Eileen Birch
_______________________________
Dr. George Sorial
_______________________________
_______________________________
Composition and Formation Mechanism of Diesel Particulate Matter
Associated with Various Factors from A Non-road Diesel Generator

A dissertation submitted to the

Division of Research and Advanced Studies

of University of Cincinnati

In partial fulfillment of the

requirements for the degree of

DOCTORATE OF PHILOSOPHY (Ph.D.)

in the Department of Environmental Engineering

of the College of Engineering

2006

by

Fuyan Liang

B.S. (Environ. Eng.), Tsinghua University, P.R. China, 2000

M.S. (Environ. Eng.), Tsinghua University, P.R. China, 2002

Committee Chair: Mingming Lu, Ph.D.

 ABSTRACT

Diesel engines emit large quantities of fine particles to the atmosphere, on which numerous organic
compounds are absorbed, such as PAHs, nitro-PAHs, and sulfur-containing PAHs. It is well established
that exposure to PAHs and their derivatives may represent a high risk for human health. Considering its
organic fractions and inhalable properties, diesel particulate matter (DPM) is considered a probable
human carcinogen. The concerns of DPM health effects have simulated studies on DPM chemical
composition and its formation mechanisms.

This dissertation presents the results of an experimental investigation into the composition and formation
mechanism of DPM at various engine operating conditions and fuel sulfur contents. The influence of
sampling methods also was examined. High volume dilution sampling and ultrasonic assisted extraction
methods were developed for collection and extraction of DPM. Gas chromatography/mass spectrometry
(GC/MS) with selective ion chromatogram (SIC), tandem mass spectrometry (MS/MS) with selected ion
monitoring (SIM), and gas chromatography with atomic emission detection (GC/AED) were applied for
determination of numerous hydrocarbons and organosulfur species in diesel emissions. The results of the
comparisons between DPM extracts, diesel fuel, and engine oil indicate that diesel fuel, engine oil, and
combustion process were the three major sources of alkanes, organic acids, and PAHs and alkylated
PAHs in DPM, respectively. As expected, the distribution of organic compounds between gas and particle
phase diesel emissions was directly correlated with their vapor pressures. Adsorption of gas phase
organics onto diesel particles was the predominant mechanism controlling the formation of diesel aerosols.
This research revealed strong effects from engine operation conditions and sampling methods on the
organic composition and formation mechanism of DPM, as well as the compound distribution. Higher
engine loads caused the increase in DPM emission rate, its elemental carbon fraction, and the heavier
components, which were pyrogenic species, formed during combustion processes, and may present
greater health risks. The comparison between the results for dilution method and stack sampling method
demonstrated that the dilution process significantly increased the organic fraction of DPM through the
condensation of gas-phase organics. The research outcomes provide important knowledge for
understanding DPM nature and formation mechanisms.
 ACKNOWLEDGEMENTS

First and foremost, I would like to express my deepest gratitude to my advisor Dr. Mingming Lu for her
guidance, support, and friendship throughout my Ph.D. study. I am in eternal debt to her for the various
ways in which she supported me. I am also very grateful to Dr. Tim C. Keener, Dr. M. Eileen Birch, and
Dr. George Sorial for serving on my committee and for providing me with very valuable comments and
feedback. Although they did not serve as members of my committee, I would like to thank Dr. Soon-Jai
Khang and Dr. Sumana Keener for their advice and support.

I would like to acknowledge the National Institute for Occupational Safety and Health for the use of their
instrument and facilities at the laboratory in Cincinnati, Ohio. In particular, I want to thank Dr. M. Eileen
Birch for the loaning of the instrumentation used for my research, for her help, friendship, great advice,
and technical support throughout the method development and sample measurements.

This work is dedicated to my parents and my brother in China. They have always been so supportive and
loving. Thank you very much. Because having you in mind, I could finish this mission. I hope one day I
can reward you for all the time I was far from you.

My acknowledgement is extended to my friends at UC: Zhangli Cai, Kai zhang, Qiang Zhang, Peng Jin,
Qiuli Lu, Kessinee, and Phirun. I thoroughly enjoyed their friendship. To Zifei, for the help with sampling
in cool and hot weathers, and for driving me wherever needed. To Jun, for his kindness and help with the
GC/MS questions, problems, and skills.

But most especially this work is dedicated to my dearest Hao, who always showed me the power of
unconditional love. With him I have walked most of my path here in Cincinnati. I cannot thank him
enough for his love, his friendship, his support in the worst moments, his help with my work and for
feeding me with the most delicious food for so many years. Without you I would have never
accomplished this task.

Also I would like to thank the National Institute for Occupational Safety and Health for their financial
support.
 TABLE OF CONTENTS

Chapter 1 INTRODUCTION.................................................................................................... 1
1.1 BACKGROUND & MOTIVATION .......................................................................... 1
1.2 LITERATURE REVIEW ......................................................................................... 4
1.2.1 Diesel Fuel Composition....................................................................... 4
1.2.2 Engine Oil Composition ....................................................................... 8
1.2.3 Overview of DPM Composition ........................................................... 9
1.2.4 DPM Formation .................................................................................. 11
1.2.5 DPM Measurement Methods .............................................................. 12
1.2.6 Effects of Testing Conditions on DPM............................................... 14
1.3 RESEARCH OBJECTIVES ................................................................................... 15
1.4 REFERENCES .................................................................................................... 17

Chapter 2 EXPERIMENTAL DESIGN AND METHOD.................................................... 24

2.1 EXPERIMENTAL DESIGN ................................................................................... 24
2.2 SAMPLING METHOD ......................................................................................... 26
2.2.1 EPA Method 5..................................................................................... 26
2.2.2 Natural Dilution .................................................................................. 28
2.2.3 High Volume Dilution Sampling ........................................................ 30
2.3 EXTRACTION METHOD ..................................................................................... 35
2.4 ANALYTICAL METHOD .................................................................................... 37
2.4.1 OC/EC Analysis.................................................................................. 37
2.4.2 Hydrocarbon Analysis with GC/MS................................................... 40
2.4.2.1 Instrument and Operating Parameters.................................. 40
2.4.2.2 Compound Identification and Quantification ...................... 41
2.4.2.3 Standard Compounds ........................................................... 43
2.4.2.4 Quality Assurance and Quality Control............................... 46
2.5 EXPERIMENT ON SOURCE CONTRIBUTION TO DPM COMPOSITION .................... 47
2.6 REFERENCES .................................................................................................... 50

i
Chapter 3 THE ORGANIC COMPOSITION OF DIESEL PARTICULATE MATTER,
DIESEL FUEL AND ENGINE OIL OF A NON-ROAD DIESEL
GENERATOR........................................................................................................ 56
3.1 INTRODUCTION .................................................................................................. 56
3.2 EXPERIMENTAL METHOD ................................................................................... 57
3.3 RESULTS AND DISCUSSION ................................................................................. 57
3.3.1 Chemical Composition........................................................................ 58
3.3.2 Alkanes ............................................................................................... 65
3.3.3 PAHs and Alkylated PAHs................................................................. 67
3.3.4 Alkylbenzenes..................................................................................... 69
3.3.5 Organic Acids ..................................................................................... 70
3.3.6 Isomer Distribution ............................................................................. 71
3.4 CONCLUSIONS .................................................................................................... 73
3.5 REFERENCES ...................................................................................................... 73

Chapter 4 PHASE DISTRIBUTION OF SEMI-VOLATILE ORGANIC COMPOUNDS

IN DIESEL COMBUSTION AEROSOLS.......................................................... 75
4.1 INTRODUCTION ................................................................................................ 75
4.1.1 Gas/Particle Partitioning ..................................................................... 75
4.1.2 Partitioning Constant .......................................................................... 76
4.1.3 Partitioning Mechanisms and Predictive Models ............................... 77
4.1.4 Impact Factors on Kp Values .............................................................. 79
4.2 EXPERIMENTAL METHOD ................................................................................. 80
4.3 RESULTS AND DISCUSSION............................................................................... 81
4.3.1 Phase Distribution............................................................................... 81
4.3.2 Log Kp vs. log p L0 ................................................................................ 84
4.4 CONCLUSIONS .................................................................................................. 91
4.5 REFERENCES .................................................................................................... 92

Chapter 5 FACTORS AFFECTING HYDROCARBON DISTRIBUTION IN DIESEL

EMISSIONS ........................................................................................................... 96

ii
 5.1 INTRODUCTION ................................................................................................ 96
5.2 EXPERIMENTAL METHODS ............................................................................... 99
5.3 RESULTS AND DISCUSSION............................................................................... 99
5.3.1 Effect of Engine Load......................................................................... 99
5.3.1.1 DPM versus Load ................................................................ 99
5.3.1.2 Organic Composition versus Load..................................... 107
5.3.2 Effect of Sampling Method............................................................... 121
5.3.2.1 Dilution Effect on DPM, OC and EC ................................ 121
5.3.2.2 Dilution Effect on the Organic Composition of DPM ....... 124
5.3.3 Effect of Fuel Sulfur Content............................................................ 133
5.3.3.1 Effect of Fuel Sulfur Content on DPM, OC, and EC......... 133
5.3.3.2 Effect of Fuel Sulfur Content on the Organic Composition135
5.4 CONCLUSIONS ................................................................................................ 135
5.5 REFERENCES .................................................................................................. 137

Chapter 6 DETERMINATION OF POLYCYCLIC AROMATIC SULFUR

HETEROCYCLES IN DIESEL FUEL AND DIESEL PARTICULATE
MATTER BY GAS CHROMATOGRAPHY WITH ATOMIC EMISSION
DETECTION........................................................................................................ 142
6.1 INTRODUCTION .............................................................................................. 142
6.2 EXPERIMENTAL METHOD ............................................................................... 144
6.2.1 Sampling and Extraction Method ..................................................... 144
6.2.2 Analytical Method ............................................................................ 144
6.2.3 Analytical Standards ......................................................................... 145
6.3 RESULTS AND DISCUSSION ............................................................................. 147
6.3.1 AED Response Factor....................................................................... 147
6.3.2 PASH Speciation in Diesel Fuel ....................................................... 149
6.3.3 PASHs in Diesel Emissions .............................................................. 154
6.3.3.1 PASH Distribution between Gas and Particle Phase ......... 158
6.3.3.2 Total Organic Sulfur in DPM ............................................ 159
6.3.3.3 Effect of Fuel Sulfur on DPM PASH Distribution ............ 162

iii
 6.3.3.4 Effect of Engine Load on DPM PASH Distribution.......... 163
6.4 SUMMARY AND CONCLUSIONS....................................................................... 166
6.5 REFERENCES .................................................................................................. 167

Chapter 7 CONCLUSIONS AND RECOMMENDATIONS............................................. 170

7.1 CONCLUSIONS ................................................................................................ 170
7.2 RECOMMENDATIONS ...................................................................................... 174
7.3 REFERENCES .................................................................................................. 178

Appendix A CHROMATOGRAMS FOR SAMPLES ............................................................. 179

Appendix B DATA FOR DIESEL FUEL COMPONENT....................................................... 186

Appendix C DATA FOR DIESEL EMISSIONS ..................................................................... 188

Appendix D PASH DATA FOR DIESEL FUEL AND DIESEL EMISSIONS ...................... 204

Appendix E DATA FOR PHENANTHRENE-ENRICHED FUEL......................................... 207

iv
 LIST OF TABLES

Table 1-1 Hydrocarbon specifications of some commercial diesel fuel oils [32]...................... 5

Table 2-1 Performance characteristics of high volume dilution sampler................................. 34

Table 2-2 The selected quantitative fragment ions for alkylated PAHs and alkylbenzenes .... 42

Table 2-3 GC/MS/MS segment program for 16PAHs ............................................................. 43

Table 2-4 List of standard compounds and their chemical properties. .................................... 44

Table 3-1 Organic compounds present in low sulfur diesel fuel, engine oil and diesel
particulate matter. .................................................................................................... 60

Table 3-2 Ratio of 1-MN/2-MN and 1-MPh/2-MPh in diesel fuel and DPM.......................... 72

Table 4-1 Vapor pressures for some alkanes, PAHs and alkanoic acids at 25 ºC (torr). ......... 82

0
Table 4-2 Values of mr and br in the equation of log Kp = mr log p L + br at 25 ºC for alkanes
and PAHs based on the experimental data and model estimations.......................... 86

0
Table 4-3 Literature values of mr and br in the equation of log Kp = mr log p L + br for
gas/particle partitioning at urban areas. ................................................................... 87

Table 4-4 Activity coefficients for n-alkanes, PAHs and alkanoic acids from literature. ....... 88

Table 5-1 DPM emission rate from the non-road diesel generator at 0-75 kW (LSDF, dilution
sampling). .............................................................................................................. 100

Table 5-2 OC and EC emission rate (in mg/m3 and mg/hr) from the non-road diesel generator
at 0-75 kW (LSDF, dilution sampling).................................................................. 102

Table 5-3 Comparison of OC and EC percentage in total DPM mass from this study with the
results from other studies....................................................................................... 104

Table 5-4 Percentage of insoluble, soluble, unidentified and identified fraction in total DPM
mass at 0-75 kW (LSDF, dilution sampling)......................................................... 105

v
Table 5-5 The concentration of compounds in DPM collected with EPA Method 5, the natural
dilution method, and the dilution method (DR=3.4) (LSDF)................................ 125

Table 5-6 Fractions of organic compounds, EC, and non-carbonaceous materials in DPM at
low and high sulfur diesel fuel (dilution sampling)............................................... 134

Table 6-1 GC/AED operating parameters. ............................................................................. 145

Table 6-2 List of thiophene standards and related information.............................................. 146

Table 6-3 Sulfur response factors for three sulfur compounds. Results are based on five
injections................................................................................................................ 148

Table 6-4 PASHs in low and high sulfur diesel fuels (LSDF and HSDF)............................. 151

Table 6-5 PASHs in gas and particle phase diesel emissions (0 kW through 75 kW) when
burning low and high sulfur diesel fuels................................................................ 156

Table 7-1 Results for phenanthrene survivability experiment. .............................................. 176

Table 7-2 Literature results for contributions from unburned fuel and pyrosysnthesis to PAHs
in diesel emissions. ................................................................................................ 176

vi
 LIST OF FIGURES

Figure 1-1 Typical structure of diesel particulate matter (Kittelson, 1998 [59])....................... 10

Figure 2-1 Diesel emission source (diesel generator and load simulator)................................. 25

Figure 2-2 Schematic of an EPA Method 5 sampling train....................................................... 27

Figure 2-3 Schematic of the natural dilution sampler for DPM collection. .............................. 29

Figure 2-4 Dilution stack sampler schematic diagram designed by Hildemann et al. [15]....... 32

Figure 2-5 Schematic of high volume dilution sampler for particulate and gas phase diesel
emissions.................................................................................................................. 33

Figure 2-6 Thermogram for filter sample containing organic carbon (OC), carbonate (CC), and
elemental carbon (EC). PC is pyrolytically generated carbon or ‘char.’ Final peak is
methane calibration peak. Carbon sources: pulverized beet pulp, rock dust
(carbonate), and diesel particulate (from reference [3]). ......................................... 39

Figure 3-1 Total ion chromatogram of diesel fuel and DPM (time in minutes)........................ 59

Figure 3-2 Concentration and percentage of each category of compounds in low sulfur diesel
fuel and DPM........................................................................................................... 63

Figure 3-3 Relative percentage of alkanes in total identified compounds for low sulfur diesel
fuel and DPM........................................................................................................... 66

Figure 3-4 Chemical structures of isoprenoids present in diesel fuel and diesel emissions...... 67

Figure 3-5 Relative percentage of PAHs in total identified compounds in low sulfur diesel fuel
and DPM. ................................................................................................................. 68

Figure 3-6 Relative percentage of alkylated PAHs in total identified compounds in low sulfur
diesel fuel and DPM. ............................................................................................... 68

Figure 3-7 H/C ratio vs. PAH molecular mass found by GC/MS in low sulfur diesel fuel and
DPM......................................................................................................................... 69

Figure 3-8 Relative percentage of alkylbenzenes in total identified compounds in low sulfur
diesel fuel and DPM. ............................................................................................... 70

vii
Figure 3-9 Relative percentage of n-alkanoic acids from C6H12O2 to C18H36O2 in DPM and
engine oil.................................................................................................................. 71

Figure 3-10 Relative percentage of isomers in low sulfur diesel fuel and DPM......................... 72

Figure 4-1 The distribution of n-alkanes in gas and particle phase diesel emissions (400 ppmw
S, 0 kW, DR = 3.4). ................................................................................................. 81

Figure 4-2 The distribution of PAHs in gas and particle phase diesel emissions (400 ppmw S, 0
kW, DR = 3.4). ........................................................................................................ 83

Figure 4-3 The distribution of alkanoic acids in gas and particle phase diesel emissions (400
ppmw S, 0 kW, DR = 3.4). ...................................................................................... 84

0
Figure 4-4 Log Kp (m3/µg) vs. log p L (torr, @ 25 ºC) based on experimental measurements and
estimations for n-alkanes in diesel emissions at 0-75 kW. ...................................... 85

0
Figure 4-5 Log Kp (m3/µg) vs. log p L (torr, @ 25 ºC) based on experimental measurements and
estimations for PAHs in diesel emissions at 0-75 kW............................................. 85

Figure 4-6 Sources contributing to the variability in mr and br values. .................................... 90

Figure 5-1 Correlation between DPM mass concentration and diesel engine load (LSDF,
dilution sampling). ................................................................................................. 101

Figure 5-2 The concentration of organic and elemental carbon versus diesel engine loads
(LSDF, dilution sampling)..................................................................................... 102

Figure 5-3 Percentage of organic compounds, elemental carbon, and non-carbonaceous

materials in DPM at 0-75 kW (LSDF, dilution sampling). ................................... 103

Figure 5-4 Percentage of each identified category in total identified particle phase compounds
at 0-75 kW (LSDF, dilution sampling).................................................................. 106

Figure 5-5 The concentration of n-alkanes from n-decane (C10) to n-pentacosane (C25) in gas
and particulate phase diesel emissions at 0-75 kW (LSDF, dilution sampling). ... 108

Figure 5-6 The concentration of branched alkanes in gas and particulate phase diesel emissions
at 0-75 kW (LSDF, dilution sampling).................................................................. 110

Figure 5-7 The concentration of cycloalkanes from C13 to C25 in gas and particulate phase
diesel emissions at 0-75 kW (LSDF, dilution sampling)....................................... 112

viii
Figure 5-8 The concentration of EPA 16 priority PAHs in gas and particulate phase diesel
emissions at 0-75 kW (LSDF, dilution sampling). ................................................ 114

Figure 5-9 The concentration of alkylated PAHs in gas and particulate phase diesel emission at
0-75 kW (LSDF, dilution sampling)...................................................................... 117

Figure 5-10 The concentration of alkylbenzenes in gas and particulate phase diesel emission at
0-75 kW (LSDF, dilution sampling)...................................................................... 117

Figure 5-11 The concentration of alkanoic acids from C6H12O2 to C18H36O2 in diesel emissions
at 0-75 kW (LSDF, dilution sampling).................................................................. 118

Figure 5-12 Normalized emission rates for identified compounds expressed in terms of mass of
compound in the exhaust/mass of compound from the fuel input (LSDF, dilution
sampling). .............................................................................................................. 120

Figure 5-13 Comparison of DPM mass concentration between EPA Method 5 and the dilution
method with a dilution ratio of ~3.4 (LSDF)......................................................... 122

Figure 5-14 OC and EC concentration in DPM collected with EPA Method 5 and the dilution
method (LSDF). ..................................................................................................... 123

Figure 5-15 Fractions of organic compounds, EC, and non-carbonaceous materials in DPM
collected with EPA Method 5 and the dilution method (LSDF). .......................... 124

Figure 5-16 The percentage of individual alkanes in total identified (a) n-alkanes, (b) branched
alkanes, and (c) cycloalkanes in DPM collected with EPA method 5, the natural
dilution method, and the dilution method with a dilution ratio of ~3.4 (LSDF, 75
kW). ....................................................................................................................... 130

Figure 5-17 The percentage of individual aromatics in total identified (a) PAHs, (b) alkylated
PAHs, and (c) alkylbenzenes in DPM collected with method 5, the natural dilution,
and the dilution method with a dilution ratio of ~3.4 (LSDF, 75 kW).................. 131

Figure 5-18 The percentage of individual alkanoic acids in total alkanoic acids in DPM
collected with method 5, the natural dilution method, and the dilution method with
a dilution ratio of ~3.4 (LSDF, 75 kW). ................................................................ 133

Figure 5-19 Ratios of the concentration of DPM, OC, EC, and non-carbonaceous materials at
high sulfur diesel fuel over the concentration at low sulfur diesel fuel (dilution
sampling). .............................................................................................................. 134

ix
Figure 6-1 Structures of PASH standard compounds.............................................................. 147

Figure 6-2 Carbon (179 nm) and sulfur (181 nm) AED chromatograms for low sulfur (433
ppmw) diesel fuel................................................................................................... 150

Figure 6-3 Carbon (179 nm) and sulfur (181 nm) AED chromatograms for high sulfur (2284
ppmw) diesel fuel................................................................................................... 150

Figure 6-4 Possible alkylated PASH groups in low and high sulfur diesel fuels (LSDF and
HSDF, respectively)............................................................................................... 152

Figure 6-5 PASHs in low and high sulfur diesel fuels (LSDF and HSDF, respectively)
expressed in (a) µg S/g DF, and (b) mg S/g TS in DF........................................... 153

Figure 6-6 Sulfur (181 nm) AED chromatograms of gas and particulate phase diesel emissions
generated at 0 kW with low sulfur (433 ppmw) diesel fuel................................... 155

Figure 6-7 Organosulfur compounds in gas phase and particulate phase diesel emissions at 0
kW and low sulfur diesel fuel. ............................................................................... 159

Figure 6-8 Total organic sulfur in particulate phase and gas + particulate phase diesel
emissions for LSDF and HSDF at various load conditions. .................................. 160

Figure 6-9 Conversion rate of fuel sulfur to the emitted organic sulfur for HSDF and LSDF at
different engine loads............................................................................................. 161

Figure 6-10 Organosulfur distribution in DPM under 0 kW and 75 kW for LSDF (L) and HSDF
(H).......................................................................................................................... 163

Figure 6-11 Sulfur (181 nm) AED chromatograms of gas and particulate phase diesel emissions
generated at 75 kW with low sulfur (433 ppmw) diesel fuel................................. 164

Figure 6-12 PASH concentration, expressed in terms of µg S/g DPM and µg S/kg DF, in DPM
generated with HSDF vs. engine loads.................................................................. 165

x
 GLOSSARY OF ACRONYMS AND SYMBOLS

A Gas phase concentration, ng/m3

ASTM American Standard Testing Method
ATSP Specific surface area of suspended particles, cm2/µg
BC Black Carbon
br Intercept in a regression of log Kp vs. log p L0
BSTFA N,O-bis[trimethylsilyl]trifluoroacetamide
C1 Concentration of phenanthrene in normal fuel, mg/L
C2 Concentration of phenanthrene in phenanthrene-enriched fuel, mg/L
CC Carbonate Carbon
CDF Compound concentration in diesel fuel, g/kg DF
CF Carbonaceous Fraction
CH4 Methane
CIC Compound Independent Calibration
CO2 Carbon dioxide
CPh Concentration of phenanthrene in a fuel used, mg/L
CPM Condensable Particulate Matter
CS Compound concentration in diesel emissions (gas + particulate phase), µg/m3
DCM Dichloromethane
DHHS Department of Health and Human Services
DPM Diesel Particulate Matter
DR Dilution Ratio
EC Elemental Carbon
EPA Method 5 Determination of Particulate Emissions from Stationary Sources
ER1 Mg compound emission/kg of fuel combusted, mg/kg DF
ER2 Mg compound emission/g compound in diesel fuel, mg/g
F Particle phase concentration, ng/m3
F1 Concentration of phenanthrene in DPM for normal fuel, ng/m3
F2 Concentration of phenanthrene in DPM for phenanthrene-enriched fuel, ng/m3
FID Flame Ionization Detection
fom Weight fraction of organic matter on the TSP
FPh Concentration of phenanthrene in DPM, ng/m3
G/P Gas/Particle
GC/AED Gas Chromatography with Atomic Emission Detection
GC/MS Gas Chromatography with Mass Spectrometry
HEPA High efficiency particulate air filter

xi
HPLC High Performance Liquid Chromatography
HSDF High Sulfur Diesel Fuel
IARC International Agency for Research on Cancer
Kp Gas-particle partitioning coefficient, m3/µg
LSDF Low Sulfur Diesel Fuel
mr Slope in a regression of log Kp vs. log p L0
MS/MS Tandem Mass Spectrometry
MWom Mean molecular weight of the obsorbing organic matter, g/mol
ND Not Detected
NIOSH National Institute for Occupational Safety and Health
NIST National Institute of Standards and Technology
NOx Nitrogen oxides (NOx = NO + NO2)
Ns Surface concentration of sorption sites for adsorbing surface, mol/cm2
OC Organic Carbon
PAHs Polycyclic Aromatic Hydrocarbons
PASHs Polycyclic Aromatic Sulfur Heterocycles
PC Pyrolytically generated carbon
p L0 Vapor pressure of the pure compound at the temperature of interest, torr
PLE Pressurized Liquid Extraction
PM Particulate Matter
ppmv Parts per million, by volume
ppmw Parts per million, by weight
PUBDF The percentage of phenanthrene contributed by unburned diesel fuel in the total
particulate phenanthrene emission, %
PUF Polyurethane foam
Q1 Enthalpy of desorption from adsorbing surface, kcal/mol
Q2 Enthalpy of volatilization of the pure liquid, kcal/mol
QS Stack gas flow rate at standard conditions, m3/hr
R Gas constant, (= 8.314×10-3 kJ/mol·K = 0.00199 kcal/mol·K = 8.2×10-5
m3·atm/mol·K)
RCD Respirable Combustible Dust
RDF Fuel consumption rate, L/hr
RPD Relative Percent Difference
RSD Relative Standard Deviation
SFE Supercritical Fluid Extraction
SIC Selective Ion Chromatogram
SIM Selected Ion Monitoring
SO2 Sulfur dioxide

xii
SO3 Sulfur trioxide
SO42- Sulfate
SOF Soluble Organic Fraction
SRM Standard Reference Material
SS Size Selective
SVOCs Semi-Volatile Organic Compounds
T Temperature, K
TBDS tert-butyl disulfide
TC Total Carbon
TIC Total Ion Chromatogram
TOS Total Organic Sulfur
TSP Total Suspended Particulate Matter, µg/m3
UHC Unburned hydrocarbons
US EPA United States Environmental Protection Agency
WHO World Health Organization
γ Activity coefficient of a compound in the organic matter on the mole fraction scale
µg Micrograms, 1 ×10-6 g
ρDF Density of diesel fuel, 800 g/L

MN Methylnaphthalene
DMN Dimethylnaphthalene
TMN Trimethylnaphthalene
MPh Methylphenanthrene
DMPh Dimethylphenanthrene
MT Methylthiophene
DMT Dimethylthiophene
ET Ethylthiophene
PT Propylthiophene
BT Benzothiophene
MBT Methylbenzothiophene
DMBT Dimethylbenzothiophene
TMBT Trimethylbenzothiophene
TTMBT Tetramethylbenzothiophene
DBT Dibenzothiophene
MDBT Methyldibenzothiophene
DMDBT Dimethyldibenzothiophene
TMDBT Trimethyldibenzothiophene
Ph45T Phenanthro[4,5-bcd]thiophene
BN12T Benzo[b]naphtho[1,2-d]thiophene

xiii
BN21T Benzo[b]naphtho[2,1-d]thiophene
BN23T Benzo[b]naptho[2,3-d]thiophene
Ph34T Phenanthro[3,4-b]thiophene
BPh9,10T Benzo[b]phenanthro[9,10-d]thiophene
DiAT Diacenaphthothiophene

xiv
 Chapter 1

INTRODUCTION

1.1 Background & Motivation

Non-road diesel engines are extensively used in various applications, such as construction,

power generation, underground mines, agriculture, and so on. It is reported that in the United

States, 103,490 units of generator sets rated between 56 and 130 kW were sold from the years

1996 to 2000, which represented 11% of the total sales volume (905,000) of all non-road diesel

equipment sales of that size during that period [1]. However, the emissions of diesel particulate

matter (DPM) from non-road sources are significant [2-4]. According to the United States

Environmental Protection Agency (US EPA), 44% of total DPM and 12% of total NOx

emissions from mobile sources are emitted from non-road diesel vehicles [5].

Roadway studies have shown higher DPM contributions than ambient measurements

acquired at compliance monitoring sites. At a Manhattan bus stop in New York City, with idling

buses and heavy bus traffic, DPM contributions were in the range of 13-47 µg/m3 [6]. In-vehicle

emissions of DPM are responsible for a large portion of human exposure. Fruin et al. [7]

measured average black carbon (BC) concentrations in Los Angeles, finding ~5 µg/m3 when no

vehicles in front, and ~130 µg/m3 when following an urban transit bus making frequent stops.

Occupational DPM exposures are also high. DPM exposures for mineworkers range from

10 to 1280 µg/m3 [8-10]. DPM exposures for those working near diesel-powered forklifts

averaged 31 µg/m3 BC as reported by Zaebst et al. [11]. In other job categories, average personal

1
BC exposures were 27 µg/m3 for mechanics and 5.2 µg/m3 for truckers [11]. These levels greatly

exceeded the residential background BC concentrations of 1.1 µg/m3.

Exposures to diesel exhaust have resulted in adverse human health effects due to the

hazardous composition of diesel emissions. Studies have indicated that occupational exposure to

diesel exhaust may be related to lung cancer for bus garage workers, miners, and forklift workers,

and others [3, 12]. DPM consists of a large number of organic compounds [12-14], among which

some polycyclic aromatic hydrocarbons (PAHs), such as chrysene and benzo[a]pyrene, nitro-

PAHs and oxygenated PAHs are known or suspected human carcinogens [15-20]. It is well

established that exposure to PAHs and their derivatives may result in a higher health risk [21].

This is due to the small size and toxic composition of DPM, as approximately 90% of the DPM

mass is within the inhalable range (<1 µm) and the organic compounds tend to be adsorbed into

those fine particles that penetrate deep into the lungs.

Although further characterization of organic speciation is on-going, the hazardous effects

of DPM emissions have been confirmed and regulated. In 1988, the National Institute for

Occupational Safety and Health (NIOSH) first recommended that diesel exhaust be regarded as a

potential occupational carcinogen [22]. In 1989, the International Agency for Research on

Cancer (IARC) declared that DPM was possibly carcinogenic to humans [23]. The U.S.

Environmental Protection Agency (EPA) has classified diesel exhaust as a probable human

carcinogen [24]. The World Health Organization (WHO) [25], the California Environmental

Protection Agency [26], and the U.S. Department of Health and Human Services (DHHS) [27]

have also identified DPM as a probable human carcinogen.

Considering the adverse health effects of DPM emissions, significant efforts have been

made on studying DPM emissions from on-road diesel vehicles. However, there have been few

2
studies on the chemical composition of DPM emissions from non-road diesel engines, even

though the use of non-road diesel engines has increased rapidly and the resultant DPM emissions

are of significant quantity. Unlike diesel vehicles, diesel generators run at a fixed rpm, and the

amperage and voltage it produces can vary with the load [28]. The difference in the operating

mode can potentially result in different emission characteristics. Therefore, the aim of this study

is to characterize the organic composition of DPM emissions, especially carcinogenic polycyclic

aromatic hydrocarbons (PAHs) and polycyclic aromatic sulfur heterocycles (PASHs), from a

non-road diesel generator, and to investigate the source of contributions from fuel, combustion,

engine oil, and other possible sources.

In addition, it is essential to consider engine testing conditions in determining the

chemical composition of DPM emissions. It is generally accepted that diesel emissions are

influenced by the engine testing conditions: types of operation (steady-state or transient), fuel

quality (high or low sulfur), and after-treatment devices. Application of new technologies for

reducing particle mass emissions can also change the nature of the particles, such as increasing

the volatile fraction [29]. The above-mentioned considerations promote the understanding on the

impact of fuel composition, engine operation conditions, and emission control technologies on

the chemical composition of DPM for developing effective emission control strategies. Therefore,

this research also aims to study DPM composition and formation mechanisms as a function of

fuel properties and engine operation conditions.

Research on emissions from diesel engines has shown that sampling conditions can alter

measured DPM size distributions [30]. Many studies have shown that DPM remains in a state of

flux for some time after it is emitted to the atmosphere. During this state, coagulation, adsorption,

and condensation of organics and inorganics which are usually present in diesel exhaust with

3
significant quantities are continuously occurring. The fate of these condensable compounds is

significantly affected by dilution of the exhaust stream. Dilution sampling is a technique that has

been developed to examine the influence of rapid cooling and dilution on PM emissions from

combustion systems. Although a dilution sampler cannot simulate the actual atmospheric mixing,

it allows examination of the effects of dilution on PM emissions in order to better understand the

PM transformations that occur in the atmosphere. Another objective of this study is to investigate

the effects of dilution on DPM chemical composition and compound distribution by comparing

the composition of DPM collected with dilution and stack sampling methods.

1.2 Literature Review

1.2.1 Diesel Fuel Composition

Diesel fuel consists mainly of saturated and aromatic hydrocarbons [31, 32]. Paraffins,

cycloparaffins, monoaromatics, diaromatics, polynuclear aromatics and sulfur compounds are the

major compound types typically found in diesel fuel [33]. Their relative distribution depends on

the feedstock and fuel processing schemes. Table 1-1 listed the composition of different types of

diesel fuels. Sjogren et al. [13] studied the chemical composition and physical characteristics of

ten diesel fuels and obtained similar results. Generally, diesel fuel consists of 65-85% saturates,

5-30% aromatics, and 0-5% olefins. But the percentage of each category will be different for

different types of diesel fuel.

Alkanes: Normal, branched and cyclic alkanes are the most abundant components of

diesel fuel, which account for about 65-85% or even higher (Table 1-1). In general, diesel fuels

contain alkanes from C5 to C30. Song stated that long-chain alkanes with carbon numbers in the

range of 10-20 are the major paraffinic components [31]. There can be some lighter (C9-) and

4
heavier (C20+) alkanes, but they exist in small quantities in most high-way diesel fuels. However,

for different types of diesel fuels, the range of carbon numbers is different. According to

statistics of the World Health Organization, C10-C28 alkanes are the dominant, saturated

hydrocarbons for general diesel fuel (i.e., diesel fuel No. 2), C9-C16 for diesel fuel No. 1, and C10-

C30 for diesel fuel No. 4 [32]. Some other studies have reported similar results [34].

Table 1-1 Hydrocarbon specifications of some commercial diesel fuel oils [32].
Specification Diesel fuel a Kerosene b Distillated marine diesel c Diesel fuel d
Saturates (volume%) 65-95 78-96 60-90 59.4-76.6
Aromatic (volume%) 5-30 4-25 5-40 23.4-39.6
Olefins (volume%) 0-10 0-5 0-5 0-1.0
a
From CONCAWE (1985); fuel oil similar to diesel fuel (general). b From CONCAWE (1985, 1995);
fuel oil similar to diesel fuel No. 1. c From CONCAWE (1985); fuel oil similar to diesel fuel No. 4. d
From German Scientific Association for Petroleum, Natural Gas, and Coal (1991); three samples of diesel
fuel (general).

Branched alkanes are important in diesel fuels due to their large percentage and the

characteristics of some specific species as tracers, which are referred to as isoprenoids. The

regular isoprenoids, which have a head-to-tail structure, are the most abundant of this class of

compounds. These isoprenoids are naturally present in crude oil [35] and therefore would be

expected to be found in diesel fuel. The isoprenoids can be used as tracers for diesel engine

exhaust in conjunction with elemental carbon, hopanes, and steranes. The latter two compounds

classed are common markers of engine oil. The ratios of pristine to heptadecane and of phytane

to octadecane will help to identify the source of a fuel spill; furthermore, they can be used to

estimate the age of an environmental contamination and the degree of elimination as these ratios

increase during biological degradation [32].

5
 Alkylated cycloalkanes are also present in diesel fuels. The percentage of cycloalkanes is

relatively low compared with normal and branched alkanes. Alkylated cyclopentane,

cyclohexane, and cycloheptane are common components [36, 37].

Aromatics: Aromatic compounds constitute 5-30% of automotive diesel fuel, 5-40% of

marine diesel fuel, and 10-30% of diesel fuel No. 2 [38]. Alkyl benzenes are common

components of diesel fuel. Polycyclic aromatic hydrocarbons (PAHs), e.g. naphthalene,

phenanthrene, acenaphthene, acenaphthylene, fluorene and fluoranthene, are also present, as well

as alkyl- and cycloalkyl- substituted homologues of these substances. Generally naphthalene and

its methyl-substituted derivatives are predominant species. Trace amounts of heavy PAHs (3+-

ring aromatic compounds) such as chrysene, pyrene, benzanthracene and perylene can also be

present. The PAH content of diesel fuels varies widely.

Some species of PAHs are know or potential carcinogens. For instance, benzo(a)pyrene is

proven to be related to lung cancer [39]. As a result, seven PAHs have been listed in the 1999

EPA National Toxics Inventory. Those PAHs include benz(a)anthracene, Benzo(a)pyrene,

Benzo(b)fluoranthene, Benzo(k)fluoranthene, Chrysene, Dibenz(a,h)anthracene, and

Indeno(1,2,3-cd)pyrene.

Sulfur Content: The sulfur content of diesel fuel depends on the source of crude oil. For

most diesel fuels, sulfur content is 0.1-0.5% by weight. Currently, the U.S. EPA is proposing

stringent regulations for sulfur content in diesel fuels. For highway diesel fuel, fuel sulfur content

was reduced from 500 ppmv to 15 ppmv in June 1, 2006 [40]. For the non-raod applications, the

current EPA regulation is 3400 ppmv, but it will be reduced to 500 ppmv in 2007 and further to

15 ppmv in 2010 [41].

6
 A variety of sulfur compounds are present in diesel fuels, which include mercaptans,

sulfides, disulfides, cyclic sulfides, alkyl sulfates, sulfonic acids, sulfoxides, sulfones and

thiophenes [34, 42]. Among these compounds, benzothiophene, dibenzothiophene (DBT) and

their alkylated homologue are the most abundant species in diesel fuel [43, 44]. Sulfur in diesel

fuels is known to contribute to DPM emissions [45]. During combustion processes, fuel sulfur is

oxidized to SOx, while the majority of which is SO2. In the exhaust tail pipe, when exhaust gases

are cooled, sulfates, sulfuric acid, bound water and hydrocarbons with relatively low vapor

pressures will condense onto soot and increase DPM growth. It is reported that the amount of

sulfates adsorbed will affect the mass of hydrocarbons in the particles [46], including PAHs, due

to the formation of heavy hydrocarbons in the condensed phase from the reaction of adsorbed

sulfuric acid with organic compounds in the exhaust [47]. Neeft et al. [48] reported that 7-12%

particulate reduction can be obtained with a reduction of fuel sulfur content from 0.2% to 0.05%.

Baranescu et al. [45] evaluated the effect of fuel sulfur on the brake specific particulate

emissions on medium and heavy-duty trucks using three diesel fuels with 0.05%, 0.19%, and

0.29% sulfur. The results indicated that for an increase of 0.1% in fuel sulfur, particulates

increased by about 0.025 g/bhp-h.

Fuel additives: Additives are chemicals introduced in very small proportions to the fuel

to improve performance, enhance its desirable characteristics and to reduce the undesirable ones.

Some additives function by physical interactions with fuel components, while others are

involved in chemical reactions. There are various additives designed to improve certain

properties of diesel fuels, such as cold flow improvers (ethylene vinyl acetate, polyolefin ester,

polyamide), antioxidants (2,6-di-tert-butyl-4-methyl-phenol), metal deactivators, corrosion

7
inhibitors (alkyl phosphate), dispersants (polyamides, amines), detergents (amines, amides and

imidazones), lubricity agents, and demulsifiers [49-52].

1.2.2 Engine Oil Composition

Engine oil is one of the most important liquid lubricants. Engine oils are made from base

oils and a set of additives. It is recognized that the performance of engine oils is sensitive to the

properties of base oils and additives used. The known lubricating oil additives include oxidation

inhibitors, detergents, dispersants, corrosion inhibitors, rust inhibitors, viscosity index improvers,

depressants, and foam inhibitors. As mentioned above, some organic compounds in DPM may be

emitted from the evaporation of engine oils. However, there have been limited studies on the

composition of engine oils since the dominant components of engine oils are heavy compounds

and complex additives.

Kohler and Heeb [53] characterized ageing products of ester-based synthetic lubricants

by liquid chromatography with electrospray ionization mass spectrometry and by electrospray

ionization (tandem) mass spectrometry. They found liquid chromatography-electrospray

ionization mass spectrometry (LC-ESI-MS) to be a powerful technique to study the chemical

composition and the molecular distribution of homologues. In their results, the mixed esters of

C5, C7 and C9 carboxylic acids and pentaerythritol were observed in new lubricants. In used jet

engine oils, pentaerythritol tetraesters and dipentaerythritol hexaesters were formed. Netten and

Leung [54] studied the constituents of two jet engine lubricating oils and their volatile pyrolytic

degradation products using GC/MS. They identified some esters of carboxylic acids and

phosphoric acid. However, in both studies the identification of some compounds was done using

the NIST library rather than standard compounds. Also the compounds were not quantified.

8
1.2.3 Overview of DPM Composition

Diesel particulate matter consists of highly agglomerated solid carbonaceous material and

ash, as well as organic and sulfur compounds. Carbon in the fuel is mostly oxidized during

combustion with the residue exhausted in the form of solid carbon. A small fraction of the fuel

and evaporated lubricant oil escape oxidation and appear as soluble organic fraction (SOF). A

major proportion of SOF in DPM comprises high-molecular-weight semi-volatile organic

compounds which cannot be resolved by gas chromatography. This portion is composed

predominantly of branched and cyclic hydrocarbons. The resolvable portion of SOF contains

alkanes, PAHs, organic acids, and can also contain some hetero-atoms such as oxygen, nitrogen,

and sulfur. The SOF varies with engine design and operating condition, ranging from less than

10% to more than 90% by mass [55]. Most of the sulfur in the fuel is oxidized to SO2, but a

small fraction is oxidized to SO3 that leads to sulfuric acid and sulfate aerosol. Metal compounds

in the fuel and lubricant oil lead to a small amount of inorganic ash. Figure 1-1 illustrates the

structure of DPM.

Many studies have been conducted on the measurements of diesel emissions from on-

road diesel vehicles. Rogge et al. [56] characterized the organic particulate matter in gasoline-

and diesel-powered vehicle exhaust by GC/MS. In their study, more than 100 organic

compounds were quantified, including n-alkanes, n-alkanoic acids, benzoic acids, benzaldehydes,

PAHs, oxy-PAHs, steranes, pentacyclic triterpanes, nitrogen-containing compounds, and others.

Among these compounds, n-alkanes composed the main portion (60.8%) of the identified

resolved organic mass, and n-alkanoic acids constituted the second largest fraction. Schauer et al.

[57] measured both gas and particle phase emissions from medium duty diesel trucks using a

9
dilution source sampling system. Their results indicated that C1-C13 carbonyls accounted for 60%

of the gas phase organic compound mass emissions. In the particulate diesel emissions, PAHs

and alkylated PAHs were the major proportion (44%).

Figure 1-1 Typical structure of diesel particulate matter (Kittelson, 1998 [58]).

Among all the organic compounds identified in diesel emissions by various studies, ,

PAHs are important due to their carcinogenic and mutagenic properties and many efforts have

been put on the concentration, formation and ambient behavior of PAHs in diesel particles. In the

studies by Rogge et al. [56] and Schauer et al. [57], alkylated 2- to 3-ring PAHs were the most

abundant PAHs in particle phase diesel emissions. Heavy PAHs with more than 4 rings, such as

chrysene, benz[a]anthracene, and benzo[a]pyrene, were also present in particulate diesel

emissions. Zielinska et al. [59] studied phase distribution of PAHs in diesel vehicle emissions,

and found that semivolatile PAHs and nonvolatile 4- to 6-ring PAHs were present predominantly

10
on particles. They also found that the gas/particle phase distribution of semivolatile PAHs

depended on the engine loads.

In addition to PAHs, nitro-PAHs (N-PAHs) and oxygenated PAHs (O-PAHs) have also

been linked with increases in mutagenicity. Oxygenated PAHs have been identified in particles

emitted from diesel and gasoline engines [60, 61]. Nitro-PAH are of particular interest due to

their higher mutagenic (2 × 105 times) and carcinogenic (10 times) properties of certain

compounds compared to PAHs [62]. Many nitro-PAHs have been found in both air particulate

materials and diesel particulate matter [63, 64]. Typically nitro-PAHs are found at low µg/g even

ng/g concentration levels in diesel particulate matter [63, 64].

1.2.4 DPM Formation

The knowledge of the nature and relative influence of processes affecting diesel exhaust

components is critical to understand the important differences between diesel aerosol measured

in the atmosphere and in the laboratory. Many processes affect the formation, behavior, and

measurement of diesel aerosols, including condensation/adsorption, evaporation/desorption,

coagulation, and chemical reactions. In addition to these mechanisms, dilution ratio and

residence time can also affect particulate behavior.

Particle-to-Particle Processes [55]: Important particle-to-particle processes leading to

losses in an exhaust and sampling system include the following processes. (1) Thermophoresis,

which is the motion of a particle from asymmetrical forces that arise from a temperature gradient,

is important for systems with long sampling lines and cooled exhaust that promote

thermophoretic deposition. (2) Diffusion of particles may occur from Brownian motion or the

movement of particles down a particle concentration gradient, and is important for particles

11
smaller than 0.05 µm in diameter. (3) Coagulation is the process of agglomeration from particle-

to-particle collisions. Coagulation changes the size distribution of particles, but does not change

the particle mass and compound concentration.

Gas-to-Particle Conversions [55]: The amount and fate of volatile organic compounds

are important in characterizing diesel aerosols, since volatile organics are associated with the

smallest diesel particles (nano-particles). Volatile organics may remain as gaseous organics or

may be converted to particle phase and form the soluble organic fraction (SOF). Gas-to-particle

conversion of volatile organics may occur through two paths: adsorption on existing particles or

nucleation to form new particles. Adsorption of volatile organics onto diesel particles can occur

at low vapor pressures of adsorbed compounds. The amount of organics adsorbed on diesel

particles depends on the saturation ratios for the organic compounds, the surface area of diesel

particles available for adsorption, adsorption energy, and the residence time. Nucleation of

gaseous organic compounds occurs when saturation ratios are sufficiently high. Nucleation

causes the formation of the nuclei-mode particles in diesel aerosols, which primarily consist of

volatile materials. Nuclei mode is often observed when diesel exhaust is rapidly diluted.

Condensation is an extension of nucleation and adsorption, which requires supersaturation and

existing particles.

1.2.5 DPM Measurement Methods

Three methods are routinely used in mines to determine the concentration of diesel

particulate matter (DPM): the respirable combustible dust method (RCD), the size selective

method (SS), and the elemental carbon method (EC) [65]. In the RCD method, the RCD is the

amount of materials burned off from the silver membrane, which is used to collect respirable

12
dust, by controlled combustion. RCD is composed of all combustible materials collected on a

filter, including EC and the soluble organic fraction of DPM. In the SS method, it is assumed that

particles with diameters below 0.8 µm originate from diesel emissions. In the SS sampler, a

personal sampling pump draws air through a cyclone followed by an inertial impactor with a 0.8

µm cut point. DPM, which is mostly smaller than 0.8 µm, is collected on a polyvinyl chloride

filter (or glass fiber filter), and is determined gravimetrically. In the EC method, samples are

collected with or without an inertial pre-selector to remove particles larger than 0.8 µm. The

thermal-optical method (NIOSH 5040) is used to determine carbonaceous aerosols, speciation of

OC/EC under controlled temperature and atmosphere by continuously monitoring filter

transmittance. EC is considered as a specific marker of occupational exposure to DPM since EC

is a product of combustion and is composed of inert graphitic carbon.

For the characterization of DPM organic composition, analytical techniques such as gas

and liquid chromatography with mass spectrometry (GC/MS and LC/MS) are applied. Both

methods have adequate resolution and sensitivity to be used for the quantification of trace

quantities of PAH in complex environmental matrices [20, 66-68]. Compared with LC/MS,

GC/MS is a more general technique for studying the organic composition of complex DPM

matrix. The GC/MS is most sufficient in DPM analysis and is effective in isomer identification,

and low detection limits are obtained with the GC/MS technique.

1.2.6 Effects of Testing Conditions on DPM

Several environmental or testing parameters influencing the chemical composition of

diesel emissions have been identified: engine load condition, fuel sulfur content, and sampling

condition which is primarily represented by dilution ratio.

13
 Engine load has been reported to affect DPM mass concentration, size distribution,

organic carbon and elemental carbon (OC/EC) distribution, morphology and microstructure [69,

70]. A study by El-Shoboksh [69] has shown that the particulate emission factor increased

linearly with engine load, and the amount of submicron (0-0.4 µm) and coarse (9-10 µm)

particles increased as well. Liu et al. [70] reported that the fractions of EC in DPM increased

from 21% to 84% as engine load was increased from 0 kW to 75 kW.

Sulfur in diesel fuels is known to contribute to DPM emissions [56, 71-73]. Neeft et al.

reported that 7-12% particulate reduction can be obtained with a reduction of fuel sulfur content

from 0.2% to 0.05% [48]. A study by Baranescu et al. [45] indicated that for an increase of 0.1%

in fuel sulfur, particulates increased by about 0.025 g/bhp-h. The effect of fuel sulfur content on

OC and EC distribution has been reported by Liu et al. [70].

Many studies have shown that the fate of the condensable organics and inorganics

presented in diesel exhaust is significantly affected by atmospheric aging and dilution of the

exhaust stream [74, 75]. It has been reported that the dilution ratio has a significant effect on the

size distribution and the total number of particles emitted. Increasing the dilution ratio increases

the concentration of ultrafine particles [30, 76]. The results from England et al. have indicated

that traditional source testing methods (EPA 201A and EPA 202 in the paper) may significantly

overestimate particulate emissions, especially the ultrafine condensable particle fraction [77].

1.3 Research Objectives

The overall goal of this study is to explore DPM chemical composition and the physical

and chemical formation mechanisms, provide important knowledge for assessing the human

health impact of DPM and other ultrafine particles, and establish criteria for engine design,

14
operation, and fuel reformulation. Based on this overall aim, the following subtasks will be

conducted:

1. Characterize the chemical composition of DPM emissions from a non-road diesel

generator and investigate source contributions from diesel fuel, engine oil, and the combustion

process.

Organic components of DPM may be directly from the evaporation of diesel fuels and

lubricant oils used, or may be generated during combustion processes. The DPM from a non-

road diesel generator will be collected, and its organic composition will be characterized using

GC/MS instrument. Organic components of potential sources, including diesel fuels and

lubricant oils used, will be instrumentally identified. Studies on DPM chemical composition can

provide important information for assessing DPM health effects. The characterization of DPM

composition, combined with the identification of source compositions, will help to identify

products that are formed directly during the combustion process and to determine finger print

components for DPM. In addition, this information will be helpful in fuel reformulation and

emission mitigation technologies.

2. Investigate the effects of fuel properties, engine operation conditions, and sampling

methodology on DPM composition and formation mechanisms.

It has been recognized that fuel properties, engine operation conditions, and sampling

methodology can affect DPM mass concentration and size distribution. The chemical

composition of DPM may be influenced by these parameters as well. In this study, diesel

emissions from a non-road diesel generator will be collected with dilution and stack sampling

method under varied engine load conditions (0-75 kW) and fuel sulfur contents (400 ppmw and

2200 ppmw). The organic composition of the collected samples will be analyzed with GC/MS

15
instrument and compared under different engine loads, fuel sulfur contents, and sampling

methods. The results of DPM composition, especially organic components, as a function of

engine loads will help to improve diesel engine operating conditions. Studying the effects of fuel

sulfur on DPM composition will be helpful to promote the establishment of more stringent

regulations for diesel fuel, improve fuel reformulation technologies, and therefore reduce DPM

emissions. In addition, this process will be helpful for understanding the function of sulfur in

organic condensation and particulate formulation. Investigating the effects of sampling methods

on DPM composition will provide essential information for understanding the formation,

condensation, and transportation of organic compounds, and therefore help in evaluating DPM

contribution and exposure.

3. Study gas/particle partitioning of organic components in diesel emissions and

investigate the mechanisms by which the organic composition is formed and distributed between

gas and particle phases.

Gas and particle phase diesel emissions will be collected using dilution methods and will

be instrumentally analyzed. The gas/particle phase distribution of organic compounds in diesel

emissions will be investigated through correlating partitioning coefficients with vapor pressures

of targeted compounds. The study of gas/particle partitioning of organic compounds in diesel

emissions is helpful in understanding DPM formation mechanisms, especially the adsorption and

absorption contributions in forming secondary diesel aerosol.

4. Develop analytical methods for organosulfur species in diesel emissions, identify and

quantify polycyclic aromatic sulfur heterocycles (PASHs) in DPM, and investigate their

distribution in DPM under varied conditions.

16
 The study on polycyclic aromatic sulfur heterocycles in diesel emissions is promoted by

the recognition of the potential carcinogenic and mutagenic properties associated with some

PASHs. However, it’s challenging to identify and quantify PASH species in DPM due to the

complexity of the DPM matrix, as well as the instrumentation limitations. In this study,

analytical method using gas chromatography with the atomic emission detection (GC/AED) will

be developed and evaluated. PASHs in DPM will be identified and quantified; their

concentrations and distributions in DPM will be investigated under varied engine operation

conditions and fuel sulfur contents. This experiment provides additional information about DPM

chemical composition, and the developed analytical method for PASHs in DPM can be applied

not only to air pollution studies, but also to characterizations of other engineered materials.

1.4 References

1. United States Environmental Protection Agency, Nonroad Engines, Equipment, and

Vehicles; Reducing Nonroad Diesel Emissions; Industry Characterization., Available at
http://www.epa.gov/nonroad-diesel/2004fr/420r04007b.pdf (accessed June, 2006).

2. Leotz-Gartziandia, E.; Tatry, V.; and Carlier, P., Sampling and analysis of organic
compounds in diesel particulate matter, Environ. Monit. Assess., 65(1-2): 155-163, 2000.

3. Kuusimaki, L.; Peltonen, Y.; Mutanen, P.; et al., Urinary hydroxy-metabolites of

naphthalene, phenanthrene and pyrene as markers of exposure to diesel exhaust, Int. Arch.
Occup. Environ. Health, 77(1): 23-30, 2004.

4. McDonald, J.D.; Zielinska, B.; Sagebiel, J.C.; et al., Characterization of fine particle
material in ambient air and personal samples from an underground mine, Aerosol Sci.
Technol., 36(11): 1033-1044, 2002.

5. United States Envrionmental Protection Agency, Nonroad Engines, Equipment, and

Vehicles; Reducing Air Pollution from Nonroad Engines, http://www.epa.gov/nonroad,
Accessed 2006.

6. Kinney, P.L.; Aggarwal, M.; Northridge, M.E.; et al., Airborne concentrations of PM2.5
and diesel exhaust particles on Harlem sidewalks: A community-based pilot study,
Environ. Health Perspect., 108(3): 213-218, 2000.

17
7. Fruin, S.A.; Hui, S.P.; Jenkins, P.L.; et al., Fine particle and black carbon concentrations
inside vehicles, in 10th Annual Conference of the International Society of Exposure
Analysis. 2000: Monterey, CA.

8. Cantrell, B.K.; Rubow, K.L.; Jr. F. W. Winthrop; et al. Pollutant levels in underground
coal mines using diesel equipment. in Proceedings of the 6th US Mine Ventilation
Symposium. 1993. Littleton, CO: Society for Mining, Metallurgy, and Exploration, Inc.
(SME).

9. Rogers, A.; and Whelan, B., Exposure in Australian mines, in HEI Communication # 7--
Diesel Workshop: Building a Research Strategy to Improve Risk Assessment (Stone
Mountain, GA). Health Effects Institute: Cambridge, MA, 1999.

10. Saverin, R., German Pptash miners: cancer mortality, in HEI Communication #7--Diesel
Workshop: Building a Research Strategy to improve Risk Assessment (Stone Mountain,
GA). Health Effects Institute: Cambridge, MA, 1999.

11. Zaebst, D.D.; Clapp, D.E.; Blade, L.M.; et al., Quantitative determination of trucking
industry workers' exposures to diesel exhaust particles, Am. Ind. Hyg. Assoc. J., 52(12):
529-541, 1991.

12. Sauvain, J.J.; Duc, T.V.; and Guillemin, M., Exposure to carcinogenic polycyclic
aromatic compounds and health risk assessment for diesel-exhaust exposed workers, Int.
Arch. Occup. Environ. Health, 76(6): 443-455, 2003.

13. Sjogren, M.; Li, H.; Rannug, U.; et al., A Multivariate statistical analysis of chemical
composition and physical characteristics of 10 diesel fuels, Fuel, 74(7): 983-989, 1995.

14. Fernandes, M.B.; and Brooks, P., Characterization of carbonaceous combustion residues:
II. Nonpolar organic compounds, Chemosphere, 53(5): 447-458, 2003.

15. Shi, J.P.; Mark, D.; and Harrison, R.M., Characterization of particles from a current
technology heavy-duty diesel engine, Environ. Sci. Technol., 34(5): 748-755, 2000.

16. Reilly, P.T.A.; Gieray, R.A.; Whitten, W.B.; et al., Real-time characterization of the
organic composition and size of individual diesel engine smoke particles, Environ. Sci.
Technol., 32(18): 2672-2679, 1998.

17. Pal, R.; Juhasz, M.; and Stumpf, A., Detailed analysis of hydrocarbon groups in diesel
range petroleum fractions with on-line coupled supercritical fluid chromatography gas
chromatography mass spectrometry, J. Chromatogr. A, 819(1-2): 249-257, 1998.

18. Lee, S.L.; De Wind, M.; Desai, P.H.; et al., Aromatics reduction and cetane improvement
of diesel fuels, Fuel Reformulation, 5: 26-31, 1993.

19. Adonis, M.; Martinez, V.; Riquelme, R.; et al., Susceptibility and exposure biomarkers in
people exposed to PAHs from diesel exhaust, Toxicol. Lett., 144(1): 3-15, 2003.

18
20. Gratz, L.D.; Bagley, S.T.; Leddy, D.G.; et al., Interlaboratory comparison of HPLC-
fluorescence detection and GC/MS: analysis of PAH compounds present in diesel
exhaust, J. Hazard. Mater., 74(1-2): 37-46, 2000.

21. IPCS, Environmental Health Criteria 202. Selected nonheterocyclic Polycyclic aromatic
hydrocarbons. 1998, pp. 1-883, WHO: Geneva, ISBM. 9241572027.

22. NIOSH, Carcinogenic effects of the exposure to diesel exhaust. 1990: Current Inteligence
Bulletin 50. Dept. Health and Human Services, Pulication No. 88-116.

23. International Agency for Research on Cancer, IARC monographs on the evaluation of
carcinogenic risks to humans: diesel and gasoline engine exhausts and some nitroarenes.
Lyon, France, v. 46, p. 458, 1989.

24. EPA, Health assessment document for diesel emissions. 1990: Workshop Review EPA-
600/8-90/057A, Washington, DC.

25. World Health Organization, Diesel fuel and exhaust emissions-Environmental Health
Criteria 171, Geneva, Switzerland, 1996.

26. California Environmental Protection Agency, Proposed identification of diesel exhaust as

a toxic air contaminant: health risk assessment for diesel exhaust. Office of
Environmental Health Hazard Assessment, Sacramento, CA, 1998.

27. U.S. Department of Health and Human Services, National Toxicology Program, 9th
Report on Carcinogens, Research Triangle Park, NC, 2000.

28. Manual of the Generac diesel generator (Model SD080), Tri-state generator,
http://www.powercompany.org/PDF/SD6080.pdf, accessed July 2005.

29. Burtscher, H., Physical characterization of particulate emissions from diesel engines: a
review, J. Aerosol. Sci., 36(7): 896-932, 2005.

30. Lipsky, E.; Stanier, C.O.; Pandis, S.N.; et al., Effects of sampling conditions on the size
distribution of fine particulate matter emitted from a pilot-scale pulverized-coal
combustor, Energy Fuels, 16(2): 302-310, 2002.

31. Chemistry of diesel fuels, ed. C. Song, C. S. Hsu and I. Mochida, Taylor & Francis, New
York, 2000, p. 18-19.

32. Diesel fuel and exhaust emissions, Geneva: World Health Organization, 1996, p 11-13.

33. Hsu, C.S., Diesel Fuels Analysis, in Encyclopedia of Analytical Chemistry. Wiley: New
York. pp. 6613-6622, 2000.

34. Speight, J.G., Handbook of Petroleum Product Analysis, New York: Wiley-Interscience,
2002, p 177.

19
35. Wagner, T.; and Wyszynski, M.L., Aldehydes and ketones in engine exhaust emissions -
A review, Proc. Inst. Mech. Eng. Part D-J. Automob. Eng., 210(2): 109-122, 1996.

36. Andresen, J.M.; Strohm, J.J.; and Song, C.S., Characterization and pyrolytic performance
of hydrotreated derivatives from a light cycle oil for jet fuel applications, Abstr. Pap. Am.
Chem. Soc., 218: 15-PETR, 1999.

37. Andresen, J.M.; Strohm, J.J.; and Song, C.S., Thermal degradation of naphthenic jet fuels
in the pyrolytic regime and effects of hydrotreating, Abstr. Pap. Am. Chem. Soc., 216:
061-PETR, 1998.

38. Block, R.N.; Allworth, N.; and Bishop, M., Assessment of diesel contamination in soil, in
Hydeocarbon contaminated soils, Vol. 1. Remediation techniques, environmental fate,
risk assessment, analytical methodologies, regulatory considerations, Calabrese, E.J. and
Kostecki, P.T., Editors. Lewis Publishers: Chelsea, MI. pp. 135-148, 1991.

39. Denissenko, M.F.; Pao, A.; Tang, M.S.; et al., Preferential formation of benzo(a)pyrene
adducts at lung cancer mutational hotspots in P53, Science, 274(5286): 430-432, 1996.

40. US EPA, Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur
Control Requirements, EPA 420-F-00-057, December 2000.

41. US EPA, Office of Transportation and Air Quality, Regulatory Announcement, EPA420-
F-04-032, May 2004.

42. Breitkreitz, M.C.; Raimundo, I.M.; Rohwedder, J.J.R.; et al., Determination of total
sulfur in diesel fuel employing NIR spectroscopy and multivariate calibration, Analyst,
128(9): 1204-1207, 2003.

43. Ma, X.L.; Sakanishi, K.Y.; and Mochida, I., Hydrodesulfurization reactivities of various
sulfur compounds in diesel fuel, Ind. Eng. Chem. Res., 33(2): 218-222, 1994.

44. Shi, H.; Tayor, L.T.; Fujinari, E.M.; et al., Sulfur determination in heavy diesel fuel with
open-tubular column supercritical fluid chromatography, LC GC-Mag. Sep. Sci., 16(3):
276-+, 1998.

45. Baranescu, R.A., Influence of fuel sulfur on diesel particulate emissions, SAE Technical
Paper Series 881174, 1988.

46. Duran, A.; Carmona, M.; and Ballesteros, R., Competitive diesel engine emissions of
sulphur and nitrogen species, Chemosphere, 52(10): 1819-1823, 2003.

47. Yu, F.Q.; Turco, R.P.; and Karcher, B., The possible role of organics in the formation and
evolution of ultrafine aircraft particles, J. Geophys. Res.-Atmos., 104(D4): 4079-4087,
1999.

48. Neeft, J.P.A.; Makkee, M.; and Moulijn, J.A., Diesel particulate emission control, Fuel
Process. Technol., 47(1): 1-69, 1996.

20
49. Garrett, T.K., Automotive fuels and fuel systems, Vol. 2: Diesel. London: Pentech Press:
pp 334, 1994.

50. Owen, K.; and Coley, T., Automotive fuels reference book, Warrendale, PA, USA:
Society of Automotive Engineers, Inc., 1995, p 963.

51. Pipenger, G., Making 'premium' diesel fuel, Hydrocarb. Process., 76(2): 63, 1997.

52. Henly, T.J.; Kulinowski, A.M.; and Roos, J.W., The role of fuel additives in formulating
clean transportation fuels, chap. 4, in Designing Transportation Fuels for a Cleaner
Environment, Reynolds, J.G. and Khan, M.R., Editors. Taylor & Francis: Philadelphia.
pp. 61-78, 1999.

53. Kohler, M.; and Heeb, N.V., Characterization of ageing products of ester-based synthetic
lubricants by liquid chromatography with electrospray ionization mass spectrometry and
by electrospray ionization (tandem) mass spectrometry, J. Chromatogr. A, 2001.

54. Netten, C.V.; and Leung, V., Comparison of the constituents of two jet engine lubricating
oils and their volatile pyrolytic degradation products, Applied Occupational and
Envrionmental Hygiene, 15(3): 277-283, 2000.

55. Kittelson, D.B.; Watts, W.F.; and Arnold, M., Aerosol Dynamics, Laboratory and On-
road Studies - Supplemental Report No. 2 to EPA Grant Review of Diesel Particulate
Matter Sampling Methods. 1998.

56. Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; et al., Sources of fine organic aerosol. 2.
Noncatalyst and catalyst-equipped automobiles and heavy-duty diesel trucks, Environ.
Sci. Technol., 27(4): 636-651, 1993.

57. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; et al., Measurement of emissions from air
pollution sources. 2. C-1 through C-30 organic compounds from medium duty diesel
trucks, Environ. Sci. Technol., 33(10): 1578-1587, 1999.

58. Kittelson, D.B., Engines and nanoparticles: A review, J. Aerosol. Sci., 29(5-6): 575-588,
1998.

59. Zielinska, B.; Sagebiel, J.; Arnott, W.P.; et al., Phase and size distribution of polycyclic
aromatic hydrocarbons in diesel and gasoline vehicle emissions, Environ. Sci. Technol.,
38(9): 2557-2567, 2004.

60. Choudhury, D.R., Characterization of polycyclic ketones and quinones in diesel emission
particulates by gas-chromatography mass-spectrometry, Environ. Sci. Technol., 16(2):
102-106, 1982.

61. Konig, J.; Balfanz, E.; Funcke, W.; et al., Determination of oxygenated polycyclic
aromatic hydrocarbons in airborne particulate matter by capillary gas-chromatography
and gas-chromatography mass-spectrometry, Anal. Chem., 55(4): 599-603, 1983.

21
62. Durant, J.L.; Busby, W.F.; Lafleur, A.L.; et al., Human cell mutagenicity of oxygenated,
nitrated and unsubstituted polycyclic aromatic hydrocarbons associated with urban
aerosols, Mutat. Res.-Genet. Toxicol., 371(3-4): 123-157, 1996.

63. Bezabeh, D.Z.; Bamford, H.A.; Schantz, M.M.; et al., Determination of nitrated
polycyclic aromatic hydrocarbons in diesel particulate-related standard reference
materials by using gas chromatography/mass spectrometry with negative ion chemical
ionization, Anal. Bioanal. Chem., 375(5): 381-388, 2003.

64. Zwirner-Baier, I.; and Neumann, H.G., Polycyclic nitroarenes (nitro-PAHs) as

biomarkers of exposure to diesel exhaust, Mutat. Res. Genet. Toxicol. Environ. Mutagen.,
441(1): 135-144, 1999.

65. Watts, W.F.J.; and Ramachandran, G., Diesel Particulate Matter Sampling Methods:
Statistical comparison, 2000, Final Report Submitted to the Diesel Emission Evaluation
Program, http://www.deep.org/reports/watts_final.pdf, Accessed 2006.

66. Poster, D.L.; de Alda, M.J.L.; Schantz, M.M.; et al., Development and analysis of three
diesel particulate-related standard reference materials for the determination of chemical,
physical, and biological characteristics, Polycycl. Aromat. Compd., 23(2): 141-191, 2003.

67. Wise, S.A., Standard reference materials (SRMs) for the determination of polycyclic
aromatic compounds - Twenty years of progress, Polycycl. Aromat. Compd., 22(3-4):
197-230, 2002.

68. Wise, S.A.; Sander, L.C.; and May, W.E., Determination of polycyclic aromatic
hydrocarbons by liquid chromatography, J. Chromatogr., 642(1-2): 329-349, 1993.

69. El-Shobokshy, M.S., The effect of diesel-engine load on particulate carbon emission,
Atmos. Environ., 18(11): 2305-2311, 1984.

70. Liu, Z.F.; Lu, M.M.; Birch, M.E.; et al., Variations of the particulate carbon distribution
from a nonroad diesel generator, Environ. Sci. Technol., 39(20): 7840-7844, 2005.

71. Dobbins, R.A.; Fletcher, R.A.; Benner, B.A.; et al., Polycyclic aromatic hydrocarbons in
flames, in diesel fuels, an in diesel emissions, Combust. Flame, 144(4): 773-781, 2006.

72. Jang, M.; Kamens, R.M.; Leach, K.B.; et al., A thermodynamic approach using group
contribution methods to model the partitioning of semivolatile organic compounds on
atmospheric particulate matter, Environ. Sci. Technol., 31(10): 2805-2811, 1997.

73. Pankow, J.F., An absorption model of gas/particle partitioning of organic compounds in

the atmosphere, Atmos. Environ., 28(2): 185-188, 1994.

74. Ahlvik, P.; Ntziachristos, L.; Keskinen, J.; et al., Real time measurements of diesel
particle size distribution with an electrical low pressure impactor, SAE Technical Paper
Series 980410, 1998.

22
75. Kittelson, D.B.; and Abdul-Khalek, I.S. Formation of nanoparticles during exhaust
dilution. in Proceedings of 2nd ETH Workshop on Nanoparticle Measurement. August
1998. Zurich, Switzerlan.

76. Brown, J.E.; Clayton, M.J.; Harris, D.B.; et al., Comparison of the particle size
distribution of heavy-duty diesel exhaust using a dilution tailpipe sampler and an in-
plume sampler during on-road operation, J. Air Waste Manage. Assoc., 50(8): 1407-1416,
2000.

77. England, G.C.; Zielinska, B.; Loos, K.; et al., Characterizing PM2.5 emission profiles for
stationary sources: comparison of traditional and dilution sampling techniques, Fuel
Process. Technol., 65: 177-188, 2000.

23
 Chapter 2

EXPERIMENTAL DESIGN AND METHOD

2.1 Experimental Design

In this study, a Generac diesel generator (1992, Model SD080, model No. 92A-03040-S)

located at the UC Center Hill Research Facilities in Cincinnati, Ohio, was used as the non-road

source of diesel particulate matter (Figure 2-1). This unit is powered by a Generac diesel engine

which features direct-injection, turbocharging, and compression-ignition, and is rated at 80 kW,

60 Hz and 1800 rpm. Unlike diesel vehicles, diesel generators run at a fixed rpm and voltage and

the amperage it produces can vary with the load [1]. The difference in operating mode can

potentially result in different emission characteristics. A load simulator (Merlin 100

manufactured by SIMPLX) was used to simulate loads by applying steady-state banks of heaters

to the generator at 0 kW, 25 kW, 50 kW and 75 kW, respectively.

Samples were collected on quartz filters or sorbents using EPA method 5 (Source Test

Sampling for Particulate) and the dilution method. The tests were performed under four engine

load conditions with two different sulfur-containing diesel fuels (low sulfur diesel fuel [LSDF]

with sulfur content about 400 ppmw, and high sulfur diesel fuel [HSDF] with sulfur content

about 2200 ppmw; both are from Steve Krebs Oil Company, Inc.).

24
Figure 2-1 Diesel emission source (diesel generator and load simulator).

Two types of samples were collected: for the analysis of organic carbon and elemental

carbon (OC/EC), and for compositional analysis with GC/MS. Prior to sampling, the quartz fiber

filters were baked at 550°C for a minimum of 12 hours to reduce residual carbon levels

associated with new filters. After sampling, the filters were put in a desiccator for 24 hours,

weighed, and then stored in a refrigerator until the samples were extracted or OC/EC analyzed.

The collected samples for compositional analysis were extracted with solvent, filtered, and

evaporated by blowing nitrogen to 0.5~1mL. The small amount of extracts was split into two

separate fractions. One fraction was then derivatized with N,O-

bis[trimethylsilyl]trifluoroacetamide (BSTFA, Pierce) to convert organic acids to their methyl

ester analogues, which are able to be identified and quantified with GC/MS. The other fraction

was directly analyzed with GC/MS and other instruments. If the concentrations of selected

25
compounds in the extracts are not high enough to be identified, extracts from several samples

were combined.

For the analysis of diesel fuel and engine oil (Valvoline, SAE 10W-30), they were diluted

with HPLC grade dichloromethane (DCM, Fisher Scientific) to some certain concentrations and

analyzed. If organic acids were targeted compounds, the diluted samples were derivatized by

adding BSTFA, and then analyzed with GC/MS.

2.2 Sampling Method

Two major types of sampling methods have been used for collecting PM emissions from

combustion systems: source-level sampling and ambient-level sampling. In general, source-level

sampling is currently the accepted approach for the total suspended particulate and PM10

measurements for stationary sources, while ambient level sampling using dilution is the accepted

approach for mobile source particulate emission measurements. In this study, three sampling

methods were used, including EPA Reference Method 5, high volume dilution sampling, and

natural dilution sampling.

2.2.1 EPA Method 5

EPA Reference Method 5, Determination of Particulate Emissions from Stationary

Sources, which can be found in the US Code of Federal Regulations [2], is the most common

method used to determine whether a source is in compliance with the established particulate

emission standards. In EPA Method 5, the undiluted exhaust is sampled through a heated

sampling probe. Particulate matter is withdrawn isokinetically from the source and collected on a

glass fiber filter maintained at 120 ± 14°C. The particulate mass, which includes any material

26
that condenses at or above the collection/filtration temperature, is determined gravimetrically

after the removal of uncombined water. The Method 5 sampling train is shown in Figure 2-2.

For the compositional analysis and OC/EC measurement, the following modifications

were made to the EPA Method 5 sampling procedures: (1) particles were collected on quartz

fiber filters instead of glass fiber filters, as quartz filters are required for thermal-optical analysis

by NIOSH 5040 [3, 4]; (2) due to the maximum loading limitation of the OC/EC analysis, the

sampling duration for DPM collection ranged from about 8 to 15 minutes, which is much shorter

than a typical Method 5 sampling period of approximately 60 minutes; (3) in order to obtain

enough mass for compositional analysis, samples were collected for 1 to 2 hours without too

high of a pressure drop observed.

Figure 2-2 Schematic of an EPA Method 5 sampling train.

27
2.2.2 Natural Dilution

Natural dilution sampling is generally used for studies on on-road vehicle emissions.

Kittelson et al. [5-11] have conducted a series of studies on the sampling and analysis of particle

emissions from diesel engines and gasoline vehicles. In most of these studies, a natural dilution

sampling system, which is referred to as the University of Minnesota Mobile Emissions

Laboratory (MEL), was used. The MEL was installed on a truck, which allows the sampling of

exhaust from both mobile source and stationary source. When sampling the exhaust from mobile

source, the MEL followed the target vehicle and the sample was collected through the air intake

which was located in front of the MEL. A suite of aerosol instrumentation was used in MEL to

measure particle number concentration, size distribution, surface area, and gas concentration.

This natural dilution system was large-scale, and suitable for on-road emission measurements.

However, the big problem with this natural dilution system includes less control, poor

repeatability and seasonal variation of sampling results.

The natural dilution sampler was built and used for DPM compositional analysis at the

beginning of this study (Figure 2-3). A high volume blower (General Metal Works Model 2000)

was used to draw air through the system. Quartz filters with a diameter of 90 mm were used to

collect DPM. The sampling flow rate, which is adjustable from 250 L/min to 350 L/min, was

measured by an orifice meter which was pre-calibrated by a Pitot tube with a PVC air duct

(length: about 3 feet, inner diameter: 2 inches) [12]. The sampler was set 1 m away from the

exhaust gas outlet with the sampling inlet about 1.4 m from the ground.

28
Figure 2-3 Schematic of the natural dilution sampler for DPM collection.

The natural dilution sampler was similar to the dilution sampler mentioned in the

previous section, except that the exhaust gas and the dilution air were drawn in through the same

inlet with an unknown dilution ratio and were mixed naturally in the inlet duct. The flow rates of

dilution air and the exhaust gas were not measured separately. The dilution ratio depends on the

characteristics and the relative position of the exhaust outlet and the sampler inlet, which include

their diameters, the distance between them, the direction, the height, and the exhaust flow rate.

The natural dilution sampler was used for characterization of DPM composition in the
29
atmosphere. With the natural dilution sampler, the actual DPM composition after it was emitted

into the atmosphere can be obtained, while the developed dilution sampler can only simulate the

real situation. However, natural dilution sampling was significantly affected by ambient

conditions and cannot provide satisfactory repeatability. Therefore, when examining the effects

from various factors, the developed dilution sampling was applied.

2.2.3 High Volume Dilution Sampling

It is challenging to sample particulate matter emissions from diesel engines due to the

elevated temperatures of the exhaust gases and the complexity of DPM. Many studies have

shown that DPM remains in a state of flux for some time after it is emitted to the atmosphere.

During this state, coagulation, adsorption and condensation of organics and inorganics, which are

usually present in diesel exhaust with significant quantities, are continuously occurring. The fate

of these condensable compounds is significantly affected by dilution of the exhaust stream.

Research on emissions from diesel engines has shown that sampling conditions can alter

measured DPM size distributions [13, 14].

Dilution sampling is a technique that has been developed to examine the influence of

rapid cooling and dilution on PM emissions from combustion systems. A dilution sampler

rapidly mixes hot exhaust gases with the pretreated dilution air, cools down the exhaust gas

temperature, and allows nucleation, condensation, and coagulation to occur. Although a dilution

sampler cannot simulate the actual atmospheric mixing, it allows the examination of effects of

dilution on PM emissions in order to better understand the PM transformations that occur in the

atmosphere. In the design of a dilution sampling device, Hildemann et al. [15] suggested that the

following elements must be considered :

30
 1. The sampler should simulate atmospheric dilution as closely as practicable. To

achieve this, the emissions should be diluted and cooled to ambient temperature.

2. The sampler must be designed to minimize the contamination to the sample, which

requires the sampler materials to be inert to organics and stable to high temperature.

3. The sampler should provide enough residence time to allow condensation to occur.

The Code of Federal Regulations also has some requirements for exhaust gas sampling

and analytical systems for emissions from diesel engines [16, 17]. One important requirement is

that the sampling system should provide the dilution of the exhaust to a temperature less than

51.7 °C at the sample filter.

In addition to the dilution requirement for the sampling system to collect diesel emissions,

high volume sampling is another necessity for compositional analysis of diesel emissions. The

sampler should be able to collect enough materials in a practicable time for effective speciation

analysis of organic compounds in diesel emissions, such as PAHs, nitro-compounds, and

oxygenated compounds.

Hildemann et al. [15, 18, 19] designed a dilution stack sampler specifically intended to

collect fine organic aerosol from combustion sources while minimizing sample contamination

(Figure 2-4). This sampler drew the stack exhaust at a flow rate of 30 L/min, and mixed the stack

exhaust with preconditioned dilution air at a dilution ratio ranging from 25 to 100. In the dilution

tunnel, the exhaust was mixed with dilution air under turbulent flow (Re = 10000) and was

cooled to near-ambient conditions. The dilution tunnel can provide a residence time of

approximately 2 seconds, and samples collected under this condition can be used for chemical

analysis. Additional residence time was provided by the residence time chamber, after which

samples were collected for particle size distribution measurements and chemical analysis.

31
Figure 2-4 Dilution stack sampler schematic diagram designed by Hildemann et al. [15].

32
Figure 2-5 Schematic of high volume dilution sampler for particulate and gas phase diesel
emissions.

In this study, a high volume dilution sampler was designed for sampling both gas phase

and particulate diesel emissions from a diesel generator (Figure 2-5). This sampler is similar in

33
principle to the dilution stack sampler designed by Hildemann et al., but some control devices

are simplified and the dilution ratio is lower due to the limitation of experimental conditions.

Besides the sampling of particulate diesel emission, this sampler is designed to collect gas phase

emissions as well, which is a supplement to Hildemann’s sampler. The sampler consisted of two

stages. The first stage was a 90-mm quartz fiber filter (Millipore) for the collection of diesel

particulate matter. The second stage consisted of polyurethane foam plugs (60 mm diameter) in

combination with the 10 g of adsorbent resin XAD-4 (PUF/XAD/PUF “sandwich” cartridge) that

was placed downstream of the filter to collect the gas phase organics. Diesel exhaust is drawn

from the exhaust duct through the sample inlet probe (12.7-mm diameter, 1.22-m length), and

then mixed with particle-removed ambient air in the dilution pipe, a 38.1-mm-inner diameter,

2.44-m-long PVC duct.

A high volume blower (General Metal Works Model 2000) is used to draw air through

the system. The total flow rate (Qt) of this sampler is approximately 300 L/min and was

measured with an orifice meter. The dilution air was filtered at the dilution air inlet using a high-

efficiency particulate air filter (HEPA) that has successfully achieved a minimum particle

removal efficiency of 99.97%. The flow rate of dilution air (Qd) was measured with a flow meter

(Dwyer). The performance specifications for the sampler are listed in Table 2-1.

Table 2-1 Performance characteristics of high volume dilution sampler.

Parameter Range
Sample flow 50-60 L/min
Dilution air flow 150-240 L/min
Total flow 200-300 L/min
Dilution Ratio ~3.4
Residence time 0.5-1 second
Reynolds number of flow through mixing duct 7000-11000, turbulent flow

34
 A dilution ratio of 3.4, which was calculated from the two measured flow rates, was

achieved with the sampler. Even though the dilution ratio is low relative to other studies [20, 21],

the dilution sampler provides sufficient dilution air to maintain the exhaust stream at a

temperature ranging from 17.2 °C to 31.7 °C, which is lower than the temperature (51.7 °C)

required for sample filters when dilution sampling by the Code of Federal Regulations [16].

The DPM samples were taken at four different engine load conditions: 0 kW (idle

condition), 25 kW, 50 kW, and 75 kW. The sampling duration for one filter was from 15 min (75

kW) to 1 hr (0 kW), depending upon DPM concentration which related with the engine load. The

breakthrough tests were performed for the sampling duration under different load conditions,

which indicated that no significant breakthrough was observed for filters and sorbents with the

designated sampling duration. An amount from 15 mg (for 0 kW) to 30 mg (for 75 kW) of DPM

was collected to perform the GC/MS analysis.

2.3 Extraction method

In this study, three extraction methods that are generally used for DPM extraction [22-26],

shaking by hand, soxhlet, and ultrasonic assisted extraction, were tested and compared. Although

several new extraction methods have been developed, such as pressurized liquid extraction

(PLE), supercritical fluid extraction (SFE), and subcritical water extraction [27, 28], their

applications have been limited due to the high instrumentation and operation expense. Extraction

of shaking by hand is not sufficient, especially for heavy compounds. Previous studies have

indicated that soxhlet extraction needs long operation time (normally 24 hours) [25, 26], and thus

the recovery for light compounds is low. The extraction recovery that can be obtained ranges

from 67% to 116% [29-33]. The extraction recoveries from these three methods were also

35
measured in this study and were 68-87% (shaking by hand), 34-76% (soxhlet) and 87-98%

(ultrasonic assisted extraction) respectively. Regarding all these considerations (including

operation complexity, operation time, the amount of solvent used and extraction recovery for all

compounds with different volatility), the ultrasonic assisted extraction method was used in this

study.

Two types of solvent are generally used, which are pure solvent and mixture solvent.

DCM is the common pure solvent used for extracting diesel particulate matter, since most

organic compounds can be solved in DCM [24-26]. Mixture solvents, such as hexane/ether

(90:10) [26], Benzene/methanol (4:1) [23], DCM/benzene (70:30), DCM/hexane (20:80), and

methanol/DCM (20:80) [25] are preferred due to their ability to extract many compounds with

different structures and chemical properties. However, solvent exchange needs to be performed

after evaporation due to the requirement of most analytical instruments during which some light

compounds may be lost. Therefore, DCM was used as the extraction solvent in this study.

During the extraction, deuterated internal standards were spiked on filters to determine

the extraction recovery [26]. In some studies, several categories of deuterated internal standards

were used to measure the recoveries of compounds with corresponding chemical properties

respectively [20, 34]. For instance, deuterated polycyclic aromatic hydrocarbons (PAHs),

ranging in volatility from naphthalene-d8 to coronene-d12, were often used as the internal

standards for PAH analysis. The extraction recovery can also be approximated by using one

deuterated compound (mostly deuterated tetrocosane (n-C24D50)) as an internal standard [35]. In

this study, naphthalene-d8 (Aldrich) and phenanthrene-d10 (Aldrich) were used to measure

extraction recoveries for all identified compounds.

36
 Before extraction and prior to sampling, the quartz fiber filters were baked at 550 °C for a

minimum of 12 h to reduce residual carbon levels associated with new filters, cooled to ambient

temperature, and weighed. After sampling, the filters were desiccated at 20 °C ± 5.6 °C and

under ambient pressure for at least 24 hours and weighed. Then the filters were stored in the

refrigerator until the extraction. Before extraction, the quartz fiber filters were spiked with a

mixture of deuterated internal standards to measure the recovery. Samples were extracted in

DCM with sonication for 1 h. Extracts were filtered through a Teflon filter (Altec, pore size: 0.2

µm) with a vacuum filtration system to remove the insoluble fractions (such as soot), evaporated

under a gentle stream of nitrogen to ~1 mL, and separated into two fractions. One fraction was

first derivatized with N,O-bis[trimethylsilyl]trifluoroacetamide (BSTFA) to convert organic

acids into their trimethylsilyl ester analogues, which can be easily identified and quantified with

GC/MS. Then, both fractions were analyzed with GC/MS.

2.4 Analytical Method

2.4.1 OC/EC Analysis

In this study, organic carbon and elemental carbon (OC/EC) analysis with NIOSH

Method 5040 were applied for DPM measurement.

DPM has been known to consist mostly of carbonaceous materials, about 70% to 88%,

which is often classified as EC and OC [36]. EC and OC act in different ways due to their

different optical, physical, chemical, and toxicological properties. As the core of diesel soot

particles, EC is a byproduct of incomplete combustion consisting of carbon layers that are

structurally similar to graphite. Particulate OC consists of liquid droplets and soot-associated

organics. The OC fraction of DPM is a complex mixture of unburned diesel fuel, oil, and

37
numerous organic compounds including PAHs [37-39]. EC has been linked to dysrhythmia and

cardiovascular diseases [40], while PAHs in OC fraction are reasonably anticipated to be human

carcinogens [24, 41-45].

The contributions of EC and OC to DPM vary with a number of factors, such as the fuel

type, engine type, duty cycle, engine maintenance, individual operators, the use of emission

control devices, and the compositions of the lubricant oil. Some previous studies have reported

that the OC/EC distribution varies with load conditions for new heavy duty diesel vehicles [45],

military vehicles [46], and non-road diesel generators [36].

The thermal-optical method NIOSH 5040 is most commonly used in North America for

OC/EC analysis [3, 4]. In this method, speciation of OC and EC is accomplished through

temperature and atmosphere control and by continuous monitoring of filter transmittance. Laser

light passed through the filter allows continuous monitoring of filter transmittance. Because

temperatures in excess of 850 °C are employed during the analysis, quartz-fiber filters are

required. A punch from the sample filter is taken for analysis, and OC and EC are reported in

terms of µg/cm2 of filter area. The total OC and EC on the filter are calculated by multiplying the

reported values by the deposit area. In this approach, a homogeneous sample deposit is assumed.

Thermal-optical analysis proceeds essentially in two stages. In the first, organic and

carbonate (if present) carbon are evolved in a helium atmosphere as the temperature is stepped

up to about 850 °C. The evolved carbon is catalytically oxidized to CO2 in a bed of granular

MnO2, then reduced to CH4 in a Ni/firebrick methanator. CH4 is quantified by an FID. In the

second stage, the sample oven temperature is reduced, an oxygen-helium mix is introduced, and

the temperature is stepped (to about 940 °C). As oxygen enters the oven, pyrolytically generated

carbon (PC) is oxidized and a concurrent increase in filter transmittance occurs (see Figure 2-6).

38
The point at which the filter transmittance reaches its initial value is defined as the "split"

between OC and EC. Carbon evolved prior to the split is considered as OC (including carbonate),

and carbon volatilized after the split is considered as EC. TC is the sum of OC and EC. Blank

filters were analyzed for a background check. The mass concentration of organic compounds can

be estimated from OC by a multiplicative factor of 1.2-1.4 to account for other elements, (e.g.,

hydrogen, oxygen, nitrogen, sulfur) in addition to carbon [47].

Figure 2-6 Thermogram for filter sample containing organic carbon (OC), carbonate (CC),
and elemental carbon (EC). PC is pyrolytically generated carbon or ‘char.’ Final
peak is methane calibration peak. Carbon sources: pulverized beet pulp, rock dust
(carbonate), and diesel particulate (from reference [3]).

39
2.4.2 Hydrocarbon Analysis with GC/MS

2.4.2.1 Instrument and Operating Parameters

A number of analytical methods have been developed and applied to the quantification of

organic compounds in diesel emissions. Among these, gas chromatography / mass spectrometry

(GC/MS), and high performance liquid chromatography (HPLC) coupled with fluorescence

detection, are the most typical analytical methods. Both of these methods have adequate

resolution and sensitivity to be used for the quantification of trace quantities of PAH in complex

environmental matrices [24, 48-50]. However, co-elution of compounds where separation and

especially quantitation of species within a peak cannot be overcome with HPLC, since for this

technique compound identification still relies on retention indices. Thus, errors may be involved

in compound recognition. GC/MS is a better choice for such complex matrices. The GC/MS

method is more efficient and provides results for more isomers than the LC-fluorescence method

[51], and low detection limits are obtained with the GC/MS technique.

In this study, the Varian GC/MS system (Varian, CP-3800 GC, Saturn 2200 ion trap MS)

equipped with a CP-8400 automatic sampler was used. The column used for the analysis was

CP-Sil 8 CB low Bleed/MS (30 m × 0.25 mm × 0.25 mm, equivalent of DB5-ms).

The chromatographic procedure for GC/MS can be described as follows: The injection

(0.5 µL) was in splitless mode at a temperature of 280 °C. The column temperature was first set

at 40°C for 2 min, then raised to 200 °C at 10 °C/min, to 270 °C at 5 °C/min, to 300 °C at 10

°C/min, and held at 330 °C for 10 min. Helium was used as the carrier gas with a flow rate of 1.2

mL/min. The total analysis time was 45 min.

The transfer line between the chromatograph and the mass analyzer was maintained at

280 °C, and the conditions for the mass analyzer follow. The mass analyzer manifold and the ion

40
trap were set to 45 ºC and 170 ºC, respectively. The initial ionization mode for the ion trap was

electron impact, and the filament emission current for that purpose was set to 10 µA. The

filament and electron multiplier were delayed for 3.5 min to eliminate the solvent peak. For mass

data acquisition, the threshold was 1 count.

2.4.2.2 Compound Identification and Quantification

Compound identification was performed by comparing the retention times and mass

spectra of samples with standard reference compounds under the help of National Institute of

Standards and Technology (NIST) standard library. The method of selective ion search was used

to enhance the identification of the PAH compounds.

Compound quantification was based on external standards. The response factors were

determined by injections of standard compounds with multilevel concentrations using the same

analytical method. Although GC/MS is used to analyze compounds in diesel exhaust, it is

difficult to quantify all components with conventional total ion chromatogram (TIC), since some

components exist in trace amount, such as PAHs, and they are mixed with numerous components

that comprise the very convoluted exhaust. Therefore, in this study three techniques were used to

quantify different categories of compounds, TIC, selective ion chromatogram (SIC), and tandem

mass spectrometry (MS/MS).

TIC: TIC was used to quantify these compounds which have high concentrations in diesel

emissions and do not have prominent fragment ions. Normal alkanes, branched alkanes,

saturated cycloalkanes, and organic acids were quantified with TIC.

SIC: SIC is a post-analysis technique that operates on mass signals that the computer

previously saved as a function of time. The analyst can search a previous chromatogram

41
selectively for a particular molecular or fragment ion and simultaneously reject the interferences

to generate a SIC with signals that correspond only to the mass to charge (m/z) of interest [52].

This dual function enhances the detection and sensitivity. An important advantage is that the

established procedures for producing the traditional gas chromatography/mass spectrometry

(GC/MS) spectrum remain unchanged.

In this study, SIC was used for the quantification of alkylated PAHs and alkylbenzenes,

which have prominent ions and medium concentrations in diesel emissions. The selected

fragment ions for each compound were listed in Table 2-2.

Table 2-2 The selected quantitative fragment ions for alkylated PAHs and alkylbenzenes.
Compound Ions Compound Ions
MN 141, 142, 115 C1-B 91, 92
DMN 156, 141, 115 C2-B 106, 91
TMN / MEN 170, 155, 128 C3-B 105, 120,91
MPh / MAn 192, 191, 165 C4-B 119, 134, 91
DMPh / DMAn 206, 191, 189 C5-B 133, 105, 148
Hopane 191 C6-B 147, 119, 162

MS/MS: MS/MS is an approach that reduces the background caused by the complex

matrix. All ions have been excluded, except for the selected ions (generally one to three ions for

each compound), which can then be fragmented to produce a unique product ion mass spectrum.

Because of the ability of MS/MS to preferentially isolate the ions of the target compounds, the

background noise is lower and higher sensitivities are possible due to the absence of interfering

ions [53-55].

The MS/MS process is a pre-programmed mass data acquisition event. Selected ion

monitoring (SIM) was used for the analysis of all sixteen EPA priority PAHs. An analysis

16PAHs with an ion trap can be further segmented, with each segment optimized for particular
42
PAHs in that window. The time segment when to assign the m/z search for a particular

compound was based on the chromatographic elution time for the standard compound. The time

segment and monitored ions for all 16PAHs are listed in Table 2-3.

Table 2-3 GC/MS/MS segment program for 16PAHs.

Compound Name Abbr. MW # of Time Ions monitored(m/z)
rings segment
(min)
Naphthalene Nap 128 2 4.0-12.0 101-103, 127-129
Acenaphthylene Acy 152 3 12.0-14.8 75-77, 125-127, 151-153
Acenaphthene Ace 154 3 14.8-15.5 75-77, 125-127, 152-154
Fluorene Flu 166 3 15.5-16.5 82-84, 138-140, 165-167
Phenanthrene Phe 178 3 16.5-20 74-76, 151-153, 177-179
Anthracene Ant 178 3 16.5-20 74-76, 151-153, 177-179
Fluoranthene Flt 202 4 20-25 88-89,100-102, 201-203
Pyrene Pyr 202 4 20-25 88-89,100-102, 201-203
Benzo(a)anthracene Baa 228 4 25-30 99-101, 112-114, 227-229
Chrysene Chy 228 4 25-30 99-101, 112-114, 227-229
Benzo(b)fluoranthene Bbf 252 5 30-35 111-113, 124-126, 251-253
Benzo(k)fluoranthene Bkf 252 5 30-35 111-113, 124-126, 251-253
Benzo(a)pyrene Bap 252 5 30-35 111-113, 124-126, 251-253
Indeno[1,2,3-cd]pyrene Ind 276 6 35-45 123-126, 136-139, 275-279
Dibenz(a,h)anthracene Dba 278 5 35-45 123-126, 136-139, 275-279
Benzo(ghi)perylene Bgp 276 6 35-45 123-126, 136-139, 275-279

2.4.2.3 Standard Compounds

Quantification was based on the response factors of standard compounds and the

extraction recoveries of deuterated compounds. The standard compounds listed in Table 2-4

were injected into GC/MS system. All these standard compounds were dissolved in HPLC grade

dichloromethane.

43
Table 2-4 List of standard compounds and their chemical properties.
Compound name Formula Abbreviation MW
n-Alkanes (Restek)
n-Decane C10H22 C10 142
n-Undecane C11H24 C11 156
n-Dodecane C12H26 C12 170
n-Tridecane C13H28 C13 184
n-Tetradecane C14H30 C14 198
n-Pentadecane C15H32 C15 212
n-Hexadecane C16H34 C16 226
n-Heptadecane C17H36 C17 240
n-Octadecane C18H38 C18 254
n-Nonadecane C19H40 C19 268
n-Eicosane C20H42 C20 282
n-Heneicosane C21H44 C21 296
n-Docosane C22H46 C22 310
n-Tricosane C23H48 C23 324
n-Tetracosane C24H50 C24 338
n-Pentacosane C25H52 C25 352

Branched alkanes (Chiron)

Norfarnesane C14H30 C14 198
Farnesane C15H32 C15 212
Norpristane C18H38 C18 254
Pristine C19H40 C19 268
Phytane C20H42 C20 282

Cycloalkanes (Chiron)
n-Heptylcyclohexane C13H26 C13 184
n-Octylcyclohexane C14H28 C14 198
n-Nonylcyclohexane C15H30 C15 212
n-Decylcyclohexane C16H32 C16 226
n-Undecylcyclohexane C17H34 C17 240
n-Dodecylcyclohexane C18H36 C18 254
n-Tridecylcyclohexane C19H38 C19 268
n-Tetradecylcyclohexane C20H40 C20 282
n-Pentadecylcyclohexane C21H42 C21 296
n-Hexadecylcyclohexane C22H44 C22 310

44
Table 2-4 (Continued)
Compound name Formula Abbreviation MW
n-Heptadecylcyclohexane C23H46 C23 324
n-Octadecylcyclohexane C24H48 C24 338
n-Nonadecylcyclohexane C25H50 C25 352

PAHs (Aldrich)
Naphthalene C10H8 Nap 128
Acenaphthylene C12H8 Acy 152
Acenaphthene C12H10 Ace 154
Fluorene C13H10 Flu 166
Phenanthrene C14H10 Phe 178
Anthracene C14H10 Ant 178
Fluoranthene C16H10 Flt 202
Pyrene C16H10 Pyr 202
Benzo(a)anthracene C18H12 Baa 228
Chrysene C18H12 Chy 228
Benzo(b)fluoranthene C20H12 Bbf 252
Benzo(k)fluoranthene C20H12 Bkf 252
Benzo(a)pyrene C20H12 Bap 252
Indeno[1,2,3-cd]pyrene C22H12 Ind 276
Dibenz(a,h)anthracene C22H14 Dba 278
Benzo(ghi)perylene C22H12 Bgp 276

Alkylated PAHs (Aldrich)

1-Methylnaphthalene C11H10 1-MN 142
2-Methylnaphthalene C11H10 2-MN 142
1,2-Dimethylnaphthalene C12H12 1,2-DMN 156
1,3-Dimethylnaphthalene C12H12 1,3-DMN 156
1,4-Dimethylnaphthalene C12H12 1,4-DMN 156
1,5-Dimethylnaphthalene C12H12 1,5-DMN 156
1,6-Dimethylnaphthalene C12H12 1,6-DMN 156
1,7-Dimethylnaphthalene C12H12 1,7-DMN 156
1,8-Dimethylnaphthalene C12H12 1,8-DMN 156
2,3-Dimethylnaphthalene C12H12 2,3-DMN 156
2,6-Dimethylnaphthalene C12H12 2,6-DMN 156
2,7-Dimethylnaphthalene C12H12 2,7-DMN 156
1,3,5-Trimethylnaphthalene C13H14 1,3,5-TMN 170

45
Table 2-4 (Continued)
Compound name Formula Abbreviation MW
1-Methylphenanthrene (Ultra Scientific) C15H12 1-MPh 192
2-Methylphenanthrene (Ultra Scientific) C15H12 2-MPh 192
3,6-Dimethylphenanthrene (Ultra Scientific) C16H14 3,6-DMPh 206

Alkylbenzenes (Fisher)
Toluene C7H8 C1-B 92
C2-Benzenes C8H10 C2-B 106
C3-Benzenes C9H12 C3-B 120
C4-Benzenes C10H14 C4-B 134
C5-Benzenes C11H16 C5-B 148
C6-Benzenes C12H18 C6-B 162

n-Alkanoic acids (Fisher)

Hexanoic acid C6H12O2 C6 116
Heptanoic acid C7H14O2 C7 130
Octanoic acid C8H16O2 C8 144
Nonanoic acid C9H18O2 C9 158
Decanoic acid C10H20O2 C10 172
Undecanoic acid C11H22O2 C11 186
Dodecanoic acid C12H24O2 C12 200
Tridecanoic acid C13H26O2 C13 214
Tetradecanoic acid C14H28O2 C14 228
Pentadecanoic acid C15H30O2 C15 242
Hexadecanoic acid C16H32O2 C16 256
Heptadecanoic acid C17H34O2 C17 270
Octadecanoic acid C18H36O2 C18 284

Aromatic acids (Fisher)

Benzoic acid C7H6O2 122

2.4.2.4 Quality Assurance and Quality Control

In analytical procedure, QA-QC was assured by the following steps:

(1) Before processing any samples, a reagent solvent blank was analyzed to demonstrate

that interferences from the analytical system, glassware, and reagents are under control. Each

46
time a set of samples is extracted or there is a change in reagents, a reagent solvent blank should

be processed as a safeguard against laboratory contamination. The blank samples should be

carried through all stages of the sample preparation and measurement steps.

(2) For each analytical batch (up to 20 samples), a reagent blank, matrix spike, and

deuterated/surrogate samples were analyzed. The blank and spiked samples were carried through

all stages of the sample preparation and measurement steps.

(3) Calibration curves were obtained from the standard compounds with the R2 value

greater than 0.999. During the analysis of samples, one standard solution was injected to check

the response stability of instruments. If the responses of standard compounds changed more than

5%, calibration curves were re-made and new response factors were used;

(4) All sample and standard solutions were sealed in vials and stored in refrigerators to

prevent sample quality decay. After 3-6 months, several fresh standard solutions were prepared

to check the quality of the stored solutions. If the concentrations of stored solutions changed, the

stored solutions were discarded and new solutions were prepared;

(5) Sample containers, sampling equipments, and laboratory glassware were cleaned and

stored in a clean environment to ensure no contamination from external sources.

2.5 Experiment on Source Contribution to DPM Composition

For the investigation of the quantitative contribution from a source, it requires data on the

amount of a compound in DPM emissions and knowledge on either of the following two

parameters: the amount of a compound formed during combustion processes, and the amount of

a compound from a direct source, such as unburned diesel fuel and evaporated engine oil. The

parameter expressed in the amount of a compound in diesel emissions divided by the amount of

47
this compound in the diesel fuel is referred to as survivability. While it is difficult to quantify the

amount of a compound from combustion formation, due to the complexity and variability of the

actual combustion processes, survivability measurement of compounds in diesel fuels is

relatively simple and applicable.

Several studies applied a radiotracer technique to investigate the contributions of

unburned diesel fuels and other sources, which are mainly pyrosynthetic sources, to compounds

in diesel emissions [56-60]. The survivabilities for fluorene, 2-methylnathphalene, naphthalene,

pyrene, and benzo(a)pyrene were reported in these studies. Based on the survivability of a

compound and its emission rate in the diesel exhaust, the contribution from unburned diesel fuel

and the combustion source can be determined. Tancell et al. [60] reported that the majority of

benzo(a)pyrene (>80%) in the exhaust was derived from unburned fuel. A study by Rhead and

Pemberton [58] indicated that naphthalene in diesel emission was mainly from unburned diesel

fuel (76.2%), while other sources only contributed 23.8%. In the study by Rhead and Hardy [57],

the majority of fluoranthene in diesel emissions was from unburned diesel fuel. The

contributions to pyrene and fluorene in diesel emissions from unburned diesel fuel and

pyrosynthesis depended on engine speed and engine load.

A simpler method for estimating the contribution to a compound in DPM emission from

the diesel fuel was developed by Hori et al. [61]. In their study, the compound of interest was

added into the fuel, which is referred to as the compound-enriched fuel. By the comparison

between the compound emission rate from the combustion of the enriched fuel and the emission

rate from the normal fuel, the survivability of the compound could be calculated and thus the

contribution to the compound in DPM from the fuel can be obtained.

48
 In this study, similar method was used to study the contribution from unburned diesel

fuel to the compound in particulate diesel emissions. From preliminary study and literature

results, it was observed that phenanthrene was abundant in both diesel fuel and diesel emissions.

Therefore, phenanthrene was added into low sulfur diesel fuel, which is referred to as

phenanthrene-enriched (Ph-enriched) fuel. The amount of phenanthrene added should be

controlled so that this additive would not alter the combustion process in diesel engine.

Approximately 1% (by weight) of phenanthrene was added into diesel fuel [62]. Phenanthrene

content in normal fuel and phenanthrene-enriched fuel was measured. DPM samples were

collected for both fuels under the same engine operation conditions. Then the amount of

phenanthrene in the two types of samples was measured by GC/MS. The following equations

were used to calculate survivability of phenanthrene and the percentage of phenanthrene

contributed by unburned diesel fuel in the total particulate phenanthrene emission:

( F2 − F1 ) × QS
Survivability = ÷ 10 6 (2-1)
(C 2 − C1 ) × R DF

C Ph × RDF × Survivability
PUBDF = × 10 6 (2-2)
FPh × QS

where F1 and F2 are the concentration of phenanthrene in DPM for normal fuel and

phenanthrene-enriched fuel (ng/m3), QS is the stack gas flow rate at standard conditions (m3/hr),

C1 and C2 are the concentration of phenanthrene in normal fuel and Ph-enriched fuel (mg/L), RDF

is fuel consumption rate (L/hr), PUBDF is the percentage of phenanthrene contributed by unburned

diesel fuel in the total particulate phenanthrene emission (%), CPh is the concentration of

phenanthrene in a fuel used (mg/L), FPh is the concentration of phenanthrene in DPM (ng/m3),

and 106 is the conversion factor to convert mg to ng. The assumption for Eq. (2-1) and (2-2) is

49
that the amount of phenanthrene from sources other than unburned diesel fuel, which are mainly

pyrosynthetic formation, is not affected by the addition of phenanthrene.

Low sulfur diesel fuel with phenanthrene concentration of 198 mg/L was used as the

normal fuel. After phenanthrene was added, its concentration in Ph-enriched fuel was 10000

mg/L. The experiment was performed at 0 kW and DPM samples were collected using high

volume dilution sampler.

2.6 References

1. Manual of the Generac diesel generator (Model SD080), Tri-state generator,

http://www.powercompany.org/PDF/SD6080.pdf, accessed July 2005.

2. Federal Register, Part II, Environmental Protection Agency. Amendments for testing and
monitoring provisions. Final Rule, 40 CFR Part 60, 61, and 63; Tuesday, October 17,
2000. See http://www.epa.gov/ttn/emc/promgate.html for method details.

3. U.S. Department of Health and Human Services (DHHS), Public Health Service, Centers
for Disease Control, National Institute for Occupational Safety and Health (NIOSH). In
NIOSH Manual of Analytical Methods (NMAM); Cassinelli, M.E., O'Connor, P.F., Eds.;
Second Supplement to NMAM, 4th ed.; Publication No. 94-113. Cincinnati, OH, 1998.

4. Birch, M.E.; and Cary, R.A., Elemental carbon-based method for monitoring
occupational exposures to particulate diesel exhaust, Aerosol Sci. Technol., 25(3): 221-
241, 1996.

5. Kittelson, D.B.; Watts, W.F.; and Johnson, J.P., Particle Sampling Methodology: On-road
and laboratory measurements of nanoparticles from Diesel engines. Presentation to the
11th CRC On-road Vehicle Emissions Workshop, San Diego, CA, March 26-28, 2001.
Available at http://www.me.umn.edu/centers/mel/, accessed August 2006.

6. Kittelson, D.B.; Watts, W.F.; and Johnson, J.P., On-road nanoparticle measurements,
2003. Proc. 8th International Conference on Environmental Science and Technology,
Lemnos Island, Greece, September 8-10, 2003, Vol. A, pp. 453-460. Athens: Global
Network of Environmental Science and Technology (GLOBAL NESTGNEST) and the
Department of Environmental Studies of the University of the Aegean.

7. Kittelson, D.B.; Watts, W.F.; and Johnson, J.P., Diesel aerosol sampling methodology -
CRC E-43 technical summary and conclusions, 2002. Available at
http://www.me.umn.edu/centers/mel/, accessed August 2006.

50
8. Kittelson, D.B.; Watts, W.F.; and Johnson, J.P., Nanoparticle emissions on Minnesota
highways, Atmos. Environ., 38(1): 9-19, 2004.

9. Kittelson, D.B.; Watts, W.F.; Johnson, J.P.; et al., Fine particle (nanoparticle) emissions
on minnesota highways, 2001. Presentation to 7th Diesel Engine Emissions Reduction
(DEER) Workshop, Portsmouth, VA, August 5-9, 2001, available at
http://www.me.umn.edu/centers/mel/, accessed August 2006.

10. Kittelson, D.B.; Watts, W.F.; Johnson, J.P.; et al., On-road exposure to highway aerosols.
1. Aerosol and gas measurements, Inhal. Toxicol., 16: 31-39, 2004.

11. Kittelson, D.B.; Watts, W.F.; Johnson, J.P.; et al., Gasoline vehicle exhaust particle
sampling study, 2003. Proc. of 9th Diesel Engine Emissions Reduction Conference
(DEER 2003), August 24-28, 2003, Newport, RI. Washington, DC: Department of
Energy. Available at http://www.osti.gov/fcvt/deer2003/kittelsonpresentation.pdf,
accessed August 2006.

12. Liu, Z.F., The carbon and sulfur speciation of diesel emissions from a non-road generator,
M.S. thesis, University of Cincinnati, Cincinnati, OH, 2005.

13. Kittelson, D.B.; and Abdul-Khalek, I.S. Formation of nanoparticles during exhaust
dilution. in Proceedings of 2nd ETH Workshop on Nanoparticle Measurement. August
1998. Zurich, Switzerlan.

14. Lipsky, E.; Stanier, C.O.; Pandis, S.N.; et al., Effects of sampling conditions on the size
distribution of fine particulate matter emitted from a pilot-scale pulverized-coal
combustor, Energy Fuels, 16(2): 302-310, 2002.

15. Hildemann, L.M.; Cass, G.R.; and Markowski, G.R., A dilution stack sampler for
collection of organic aerosol emissions: design, characterization and field tests, Aerosol
Sci. Technol., 10(1): 193-204, 1989.

16. Exhaust gas sampling and analytical system for gaseous emissions from heavy-duty
diesel-fueled engines and particulate emissions from all engines. Electronic Code of
Federal Regulations (eCFR), Title 40, Part 86.1310-2007.

17. Exhaust gas sampling and analytical system; diesel engines. Electronic Code of Federal
Regulations (eCFR), Title 40, Part 86.1310-90.

18. Hildemann, L.M.; Markowski, G.R.; and Cass, G.R., Chemical composition of emissions
from urban sources of fine organic aerosol, Environ. Sci. Technol., 25(4): 744-759, 1991.

19. Hildemann, L.M.; Markowski, G.R.; Jones, M.C.; et al., Submicrometer aerosol mass
distributions of emissions from boilers, fireplaces, automobiles, diesel trucks, and meat-
cooking operations, Aerosol Sci. Technol., 14(1): 138-152, 1991.

51
20. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; et al., Measurement of emissions from air
pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks,
Environ. Sci. Technol., 33(10): 1578-1587, 1999.

21. Zielinska, B.; Sagebiel, J.; Arnott, W.P.; et al., Phase and size distribution of polycyclic
aromatic hydrocarbons in diesel and gasoline vehicle emissions, Environ. Sci. Technol.,
38(9): 2557-2567, 2004.

22. Duran, A.; de Lucas, A.; Carmona, M.; et al., Simulation of atmospheric PAH emissions
from diesel engines, Chemosphere, 44(5): 921-924, 2001.

23. Farrarkhan, J.R.; Andrews, G.E.; Ishaq, R.; et al., Quantitative diesel particulate analysis
using gas-chromatography mass-spectrometry, Proc. Inst. Mech. Eng. Part A-J. Power
Energy, 207(A2): 95-106, 1993.

24. Gratz, L.D.; Bagley, S.T.; Leddy, D.G.; et al., Interlaboratory comparison of HPLC-
fluorescence detection and GC/MS: analysis of PAH compounds present in diesel
exhaust, J. Hazard. Mater., 74(1-2): 37-46, 2000.

25. Leotz-Gartziandia, E.; Tatry, V.; and Carlier, P., Sampling and analysis of polycyclic
aromatic hydrocarbons (PAH) and oxygenated PAH in diesel exhaust and ambient air,
Polycycl. Aromat. Compd., 20(1-4): 245-258, 2000.

26. McDonald, J.D.; Zielinska, B.; Sagebiel, J.C.; et al., Characterization of fine particle
material in ambient air and personal samples from an underground mine, Aerosol Sci.
Technol., 36(11): 1033-1044, 2002.

27. Hawthorne, S.B.; Grabanski, C.B.; Martin, E.; et al., Comparisons of Soxhlet extraction,
pressurized liquid extraction, supercritical fluid extraction and subcritical water
extraction for environmental solids: recovery, selectivity and effects on sample matrix, J.
Chromatogr. A, 892(1-2): 421-433, 2000.

28. Pineiro-Iglesias, M.; Lopez-Mahia, P.; Vazquez-Blanco, E.; et al., Comparison between
Soxhlet, ultrasonic and microwave assisted extraction of polycyclic aromatic
hydrocarbons from atmospheric particulate matter, Fresenius Environ. Bull., 9(1-2): 17-
22, 2000.

29. Arditsoglou, A.; Terzi, E.; Kalaitzoglou, M.; et al., A comparative study on the recovery
of polycyclic aromatic hydrocarbons from fly ash and lignite coal, Environ. Sci. Pollut.
Res., 10(6): 354-356, 2003.

30. Ochsenkuhn-Petropoulou, M.; Staikos, K.; Matuschek, G.; et al., On-line determination
of polycyclic aromatic hydrocarbons in airborne particulate matter by using
pyrolysis/GC-MS, J. Anal. Appl. Pyrolysis, 70(1): 73-85, 2003.

31. Pandit, G.G.; Srivastava, P.K.; Sharma, S.; et al., Monitoring of persistent organic
pollutants (POPs) in aerosols using HPLC, J. Liq. Chromatogr. Relat. Technol., 25(8):
1271-1281, 2002.

52
32. Shu, Y.Y.; Tey, S.Y.; and Wu, D.K.S., Analysis of polycyclic aromatic hydrocarbons in
airborne particles using open-vessel focused microwave-assisted extraction, Anal. Chim.
Acta, 495(1-2): 99-108, 2003.

33. Swartz, E.; and Stockburger, L., Recovery of semivolatile organic compounds during
sample preparation: Implications for characterization of airborne particulate matter,
Environ. Sci. Technol., 37(3): 597-605, 2003.

34. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; et al., Measurement of emissions from air
pollution sources. 1. C1 through C29 organic compounds from meat charbroiling, Environ.
Sci. Technol., 33(10): 1566-1577, 1999.

35. Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; et al., Sources of Fine Organic
Aerosol.2. Noncatalyst and Catalyst- Equipped Automobiles and Heavy-Duty Diesel
Trucks, Environ. Sci. Technol., 27(4): 636-651, 1993.

36. Liu, Z.F.; Lu, M.M.; Birch, M.E.; et al., Variations of the particulate carbon distribution
from a nonroad diesel generator, Environ. Sci. Technol., 39(20): 7840-7844, 2005.

37. Fernandes, M.B.; and Brooks, P., Characterization of carbonaceous combustion residues:
II. Nonpolar organic compounds, Chemosphere, 53(5): 447-458, 2003.

38. Sauvain, J.J.; Duc, T.V.; and Guillemin, M., Exposure to carcinogenic polycyclic
aromatic compounds and health risk assessment for diesel-exhaust exposed workers, Int.
Arch. Occup. Environ. Health, 76(6): 443-455, 2003.

39. Sjogren, M.; Li, H.; Rannug, U.; et al., A multivariate statistical analysis of chemical
composition and physical characteristics of 10 diesel fuels, Fuel, 74(7): 983-989, 1995.

40. Tolbert, P.E.; Klein, M.; Metzger, K.B.; et al., Interim results of the study of particulates
and health in Atlanta (SOPHIA), J. Expo. Anal. Environ. Epidemiol., 10(5): 446-460,
2000.

41. Adonis, M.; Martinez, V.; Riquelme, R.; et al., Susceptibility and exposure biomarkers in
people exposed to PAHs from diesel exhaust, Toxicol. Lett., 144(1): 3-15, 2003.

42. Lee, S.L.; De Wind, M.; Desai, P.H.; et al., Aromatics reduction and cetane improvement
of diesel fuels, Fuel Reformulation, 5: 26-31, 1993.

43. Pal, R.; Juhasz, M.; and Stumpf, A., Detailed analysis of hydrocarbon groups in diesel
range petroleum fractions with on-line coupled supercritical fluid chromatography gas
chromatography mass spectrometry, J. Chromatogr. A, 819(1-2): 249-257, 1998.

44. Reilly, P.T.A.; Gieray, R.A.; Whitten, W.B.; et al., Real-time characterization of the
organic composition and size of individual diesel engine smoke particles, Environ. Sci.
Technol., 32(18): 2672-2679, 1998.

53
45. Shi, J.P.; Mark, D.; and Harrison, R.M., Characterization of particles from a current
technology heavy-duty diesel engine, Environ. Sci. Technol., 34(5): 748-755, 2000.

46. Kelly, K.E.; Wagner, D.A.; Lighty, J.S.; et al., Characterization of exhaust particles from
military vehicles fuelled with diesel, gasoline, and JP-8, J. Air Waste Manage. Assoc.,
53(3): 273-282, 2003.

47. Gray, H.A.; Cass, G.R.; Huntzicker, J.J.; et al., Characteristics of atmospheric organic
and elemental carbon particle concentrations in Los Angeles, Environ. Sci. Technol.,
20(6): 580-589, 1986.

48. Poster, D.L.; de Alda, M.J.L.; Schantz, M.M.; et al., Development and analysis of three
diesel particulate-related standard reference materials for the determination of chemical,
physical, and biological characteristics, Polycycl. Aromat. Compd., 23(2): 141-191, 2003.

49. Wise, S.A., Standard Reference Materials (SRMs) for the determination of polycyclic
aromatic compounds - Twenty years of progress, Polycycl. Aromat. Compd., 22(3-4):
197-230, 2002.

50. Wise, S.A.; Sander, L.C.; and May, W.E., Determination of polycyclic aromatic
hydrocarbons by liquid chromatography, J. Chromatogr., 642(1-2): 329-349, 1993.

51. Schubert, P.; Schantz, M.M.; Sander, L.C.; et al., Determination of polycyclic aromatic
hydrocarbons with molecular weight 300 and 302 in environmental-matrix standard
reference materials by gas chromatography/mass spectrometry, Anal. Chem., 75(2): 234-
246, 2003.

52. Hoffmann, E.D.; Charette, J.; and Stroobant, V., Mass Spectrometry-Chromatography
Coupling; Tandem Mass Spectrometry (MS/MS), in Mass Spectrometry Principles and
Applications. Wiley: Paris. pp. 99-142, 1996.

53. Nicol, S.; Dugay, J.; and Hennion, M.C., Simultaneous determination of polycyclic
aromatic dydrocarbons and their nitrated derivatives in airborne particulate matter using
gas chromatography-tandem mass spectrometry, J. Sep. Sci., 24(6): 451-458, 2001.

54. Nicol, S.; Dugay, J.; and Hennion, M.C., Determination of oxygenated polycyclic
aromatic compounds in airborne particulate organic matter using gas chromatography
tandem mass spectrometry, Chromatographia, 53: S464-S469, 2001.

55. Sandra, P.; Beltran, J.; and David, F., Enhanced selectivity in the determination of
triazines in environmental samples by benchtop CGC-MS-MS, HRC-J. High Resolut.
Chromatogr., 18(9): 545-550, 1995.

56. Buchholz, B.A.; Mueller, C.J.; Martin, G.C.; et al., Tracing fuel component carbon in the
emissions from diesel engines, Nucl. Instrum. Methods Phys. Res. Sect. B-Beam Interact.
Mater. Atoms, 223-24: 837-841, 2004.

54
57. Rhead, M.M.; and Hardy, S.A., The sources of polycyclic aromatic compounds in diesel
engine emissions, Fuel, 82(4): 385-393, 2003.

58. Rhead, M.M.; and Pemberton, R.D., Sources of naphthalene in diesel exhaust emissions,
Energy Fuels, 10(3): 837-843, 1996.

59. Tancell, P.J.; Rhead, M.M.; Pemberton, R.D.; et al., Survival of polycyclic aromatic
hydrocarbons during diesel combustion, Environ. Sci. Technol., 29(11): 2871-2876, 1995.

60. Tancell, P.J.; Rhead, M.M.; Trier, C.J.; et al., The sources of benzo[a]pyrene in diesel
exhaust emissions, Sci. Total Environ., 162(2-3): 179-186, 1995.

61. Hori, S.; Sato, T.; and Narusawa, K., Effects of diesel fuel composition on SOF and PAH
exhaust emissions, JSAE Review, 18(3): 255-261, 1997.

62. Mi, H.H.; Lee, W.J.; Chen, C.B.; et al., Effect of fuel aromatic content on PAH emission
from a heavy-duty diesel engine, Chemosphere, 41(11): 1783-1790, 2000.

55
 Chapter 3

THE ORGANIC COMPOSITION OF DIESEL PARTICULATE

MATTER, DIESEL FUEL AND ENGINE OIL OF A NON-ROAD

DIESEL GENERATOR

3.1 Introduction

Three sources may contribute to diesel emissions: combustion process, unburned diesel

fuel, and engine oil. Elemental carbon, as well as some fractions of organic carbon, was

generated during the combustion process. In addition to combustion (pyrosynthesis), the organic

fractions of DPM may be resultant from unburned diesel fuel and engine oil. The composition of

the diesel fuel has been identified and categorized into fractions of fuel input, such as 65-85%

saturates, 5-30% aromatics, and 0-5% olefins [1]. The percentage may vary with manufactures,

the mining locations, the refining processes and sulfur content. Sjogren et al. studied ten diesel

fuels from different suppliers, and obtained considerably differed percentages for each category

[2]. High sulfur content in the diesel fuels has been associated with more DPM emissions [3].

More stringent regulations have been proposed to reduce sulfur content in the near future to

reduce particulate emissions. Engine oil is also believed to be a possible source of DPM. It was

reported that some organic compounds in DPM are emitted from the evaporation of engine oil,

such as some short-chain alkanoic acids, hopanes and steranes [4, 5].

Despite the progress in DPM sampling and analysis, there are still gaps to fill, especially

in the compositional correlations among DPM, diesel fuel and engine oil. The aim of this study is

to compare the organic composition of DPM to diesel fuel and engine oil and seek the correlation

56
of DPM from petrogenic (raw materials such as diesel fuel and engine oil) and pyrogenic

(combustion) sources.

3.2 Experimental method

In this study, a Generac diesel generator (1992, Model SD080, model No. 92A-03040-S)

was used as the non-road source of diesel particulate matter. A high volume sampler with a flow

rate of approximately 300 L/min was used for sampling DPM with the sampling time 1 hour.

The DPM was taken at idle mode (0 kW) on quartz filters. The low sulfur diesel fuel (Steve

Krebs Oil Company, Inc., sulfur content: 433ppmw), engine oil from this generator, and DPM

samples were analyzed with GC/MS. The detailed experimental setup, DPM sampling, extraction

procedures, and instrumental analysis are detailed described in Chapter 2.

3.3 Results and discussion

In this study, the sample was taken at the exit of the exhaust pipe, and natural dilution

was used. The DPM mass concentration is approximately 3.7 mg/m3. Compared with other

studies, our DPM mass concentration is in the range of the reported results [3]. However, the

DPM emissions can vary significantly due to the size and type of diesel machines used, the

sampling method, and the load reported. The mass concentration of DPM from a Jeep Bobtail

(engine speed: 1700 rpm, engine load: 9-87 kW) reported by Zielinska et al. is in the range of

1687-3800 µg/m3 [6]. Chan et al. reported that the DPM mass concentration from the Ford FSD

425T turbocharged DI diesel engine (four-cylinder, direct injection, 800-4000 rpm, maximum

power output 52 kW) was from 0.2 mg/m3 to more than a hundred mg/m3 at zero load condition

57
[7]. In addition, the DPM emissions from diesel vehicles are often reported as mass per mile,

which further complicates the comparison.

3.3.1 Chemical Composition

Figure 3-1 (a) and (b) shows the total ion chromatograms (TIC) of diesel fuel and DPM

from GC/MS, respectively. The chemical composition of low sulfur diesel fuel, DPM and engine

oil is listed in Table 3-1, and the compounds labeled “ND” (not detected) can be either below the

detection limit or not present in the sample. Figure 3-2 summed up the compounds by categories,

i.e. n-alkanes, branched alkanes, saturated cycloalkanes, PAHs, alkylated PAHs, alkylbenzenes,

n-alkanoic acids and aromatic acids for the diesel fuel and DPM, and the relative fractions of

each category were calculated. It is indicated that DPM contains higher fractions of PAHs,

alkylated PAHs and n-alkanoic acids comparing with the diesel fuel, while the diesel fuel has

more fractions of branched alkanes and cycloalkanes.

58
 (a) Diesel fuel. Peaks 1–11: n-alkanes C10–C20

(b) DPM. Peaks 1–8: n-alkanes C14–C21; peak 9: Phe; peak 10: 2-MN; peak 11: 1-MN; peak 12:
DMN; peak 13: TMN; peak 14: MPh; peak 15: DMPh.

Figure 3-1 Total ion chromatogram of diesel fuel and DPM (time in minutes).

59
Table 3-1 Organic compounds present in low sulfur diesel fuel, engine oil and diesel
particulate mattera.
Compounds Diesel fuel DPM composition Engine oil
composition (µg/g) (µg/g) composition (µg/g)
n-Alkanes
n-Decane (C10) 12115 9.71 ND
n-Undecane (C11) 11271 21.6 ND
n-Dodecane (C12) 17149 31.3 ND
n-Tridecane (C13) 28834 57.1 ND
n-Tetradecane (C14) 25604 204 ND
n-Pentadecane (C15) 27660 564 ND
n-Hexadecane (C16) 23965 1164 ND
n-Heptadecane (C17) 26082 2645 ND
n-Octadecane (C18) 8727 1479 ND
n-Nonadecane (C19) 4988 1257 ND
n-Eicosane (C20) 2193 742 ND
n-Heneicosane (C21) 1092 138 ND
n-Docosane (C22) 756 10.3 ND
n-Tricosane (C23) 220 7.42 ND
n-Tetracosane (C24) 107 5.19 ND
n-Pentacosane (C25) ND 3.57 ND
Sum of n-alkanes 190763 8341

Branched alkanes
Norfarnesane (C14) 11469 12.8 ND
Farnesane (C15) 9719 28.4 ND
Norpristane (C18) 7992 135 ND
Pristine (C19) 5871 147 ND
Phytane (C20) 4775 185 ND
Other branched alkanes 328578 7084 ND
Sum of branched alkanes 368404 7592

Cycloalkanes
n-Heptylcyclohexane (C13) 13144 11.1 ND
n-Octylcyclohexane (C14) 11467 26.8 ND
n-Nonylcyclohexane (C15) 10582 28.5 ND
n-Decylcyclohexane (C16) 9135 43.3 ND
n-Undecylcyclohexane (C17) 6207 38.8 ND
n-Dodecylcyclohexane (C18) 2073 60.7 ND
n-Tridecylcyclohexane (C19) 165 12.3 ND
n-Tetradecylcyclohexane (C20) 25 9.82 ND
n-Pentadecylcyclohexane (C21) ND 6.79 ND

60
Table 3-1 (Continued)
Compounds Diesel fuel DPM composition Engine oil
composition (µg/g) (µg/g) composition (µg/g)
n-Hexadecylcyclohexane (C22) ND 4.29 ND
n-Heptadecylcyclohexane (C23) ND ND ND
n-Octadecylcyclohexane (C24) ND ND ND
n-Nonadecylcyclohexane (C25) ND ND ND
Sum of saturated cycloalkanes 52798 242

PAHs
Naphthalene (Nap) 753 4.06 ND
Acenaphthylene (Acy) 159 2.03 ND
Acenaphthene (Ace) 85 3.04 ND
Fluorene (Flu) 100 7.10 ND
Phenanthrene (Phe) 247 77.1 ND
Anthracene (Ant) 7.5 4.06 ND
Fluoranthene (Flt) ND 25.4 ND
Pyrene (Pyr) 5.0 67.0 ND
Benzo(a)anthracene (Baa) ND 5.07 ND
Chrysene (Chy) ND 5.07 ND
Benzo(b)fluoranthene (Bbf) ND 3.04 ND
Benzo(k)fluoranthene (Bkf) ND 2.03 ND
Benzo(a)pyrene (Bap) ND 4.06 ND
Indeno[1,2,3-cd]pyrene (Ind) ND 6.61 ND
Dibenz(a,h)anthracene (Dba) ND 1.65 ND
Benzo(ghi)perylene (Bgp) ND 3.04 ND
Biphenyl 437 25.9 ND
Sum of PAHs 1793 246

Alkylated PAHs
1-Methylnaphthalene (1-MN) 585 29.8 ND
2-Methylnaphthalene (2-MN) 2291 86.0 ND
1,2-Dimethylnaphthalene (1,2-DMN) 373 31.0 ND
1,3-Dimethylnaphthalene (1,3-DMN) ND 235 ND
1,4-Dimethylnaphthalene (1,4-DMN) 1540 ND ND
1,5-Dimethylnaphthalene (1,5-DMN) ND 134 ND
1,6-Dimethylnaphthalene (1,6-DMN) 1807 61.0 ND
1,7-Dimethylnaphthalene (1,7-DMN) 2548 24.8 ND
1,8-Dimethylnaphthalene (1,8-DMN) ND ND ND
2,3-Dimethylnaphthalene (2,3-DMN) ND ND ND
2,6-Dimethylnaphthalene (2,6-DMN) 1224 54.2 ND
2,7-Dimethylnaphthalene (2,7-DMN) 1837 48.7 ND

61
Table 3-1 (Continued)
Compounds Diesel fuel DPM composition Engine oil
composition (µg/g) (µg/g) composition (µg/g)
Methylethylnaphthalene (MEN) 273 ND
Trimethylnaphthalene (TMN) 12327 2641 ND
1-Methylphenanthrene (1-MPh) 242 1770 ND
2-Methylphenanthrene (2-MPh) 528 2168 ND
Dimethylphenanthrene (DMPh) ND 3420 ND
Sum of alkylated PAHs 25302 10977

Alkylbenzenes
Toluene (C1-B) 1377 10.7 ND
C2-Benzenes (C2-B) 12932 20.5 ND
C3-Benzenes (C3-B) 10003 14.9 ND
C4-Benzenes (C4-B) 9724 15.7 ND
C5-Benzenes (C5-B) 5538 6.15 ND
C6-Benzenes (C6-B) 5222 3.17 ND
Sum of alkylbenzenes 44796 71.1

n-Alkanoic acids
Hexanoic acid (C6) ND ND ND
Heptanoic acid (C7) ND 17.8 37.4
Octanoic acid (C8) ND 39.6 55.9
Nonanoic acid (C9) ND 87.8 98.7
Decanoic acid (C10) ND 289 120
Undecanoic acid (C11) ND 354 100
Dodecanoic acid (C12) ND 591 1240
Tridecanoic acid (C13) ND 444 237
Tetradecanoic acid (C14) ND 1955 3400
Pentadecanoic acid (C15) ND 1237 1429
Hexadecanoic acid (C16) ND 1991 6109
Heptadecanoic acid (C17) ND ND ND
Octadecanoic acid (C18) ND 4.08 ND
Sum of n-alkanoic acids 7012 12527

Aromatic acids
Benzoic acid ND 226 495
a
All of the compounds detected were quantified with standard compounds. ND: not detected, i.e. the
concentration is lower than detection limit or the compound is not present in the sample.

62
 100%
Alkylbenzenes, 6.6%
90% (4.5% of DF)
Unidentified
80% (31.6%) Alkylated PAHs, 3.7%
(2.5% of DF)
70%

60% PAHs, 0.26%

(0.18% of DF)
50%
Cycloalkanes, 7.7%
40% Identified (5.3% of DF)
(68.4%)
30% Branched alkanes, 53.9%
(36.8% of DF)
20%
n-Alkanes, 27.9%
10% (19.1% of DF)
0%

(a) Diesel fuel

Aromatic acids, 0.65%
100% (0.02% of DPM)
Insoluble
n-Alkanoic acids, 20.2%
16% Unidentified (0.70% of DPM)
80% 95.9%
(80.5% of DPM) Alkylbenzenes, 0.21%
(0.007% of DPM)

60% Alkylated PAHs, 31.7%

(1.1% of DPM)
Soluble
84% PAHs, 0.64%
40% (0.02% of DPM)
Cycloalkanes, 0.70%
Identified (0.02% of DPM)
20% 4.1%
(3.5% of DPM) Branched alkanes, 21.9%
(0.76% of DPM)
n-Alkanes, 24.0%
0% (0.83% of DPM)
(b) DPM

Figure 3-2 Concentration and percentage of each category of compounds in low sulfur diesel
fuel and DPM.

Approximately 70% (on mass basis) of the diesel fuel has been identified. Among the

identified fractions, n-alkanes, branched alkanes, saturated cycloalkanes, PAHs, alkylated PAHs

and alkylbenzenes account for 27.90%, 53.87%, 7.72%, 0.26%, 3.70% and 6.55% respectively.
63
This is consistent with several published studies [2, 8, 9]. For branched and cyclic alkanes,

studies have indicated that compounds ranging from C10-C20 are the most abundant [10], and the

C10-C28 fraction ranges from 65-85% [1], and 72-96% for the C8-C24 fraction [2]. From Table 3-

1, the aromatic compounds in the low sulfur diesel fuel include naphthalene, fluorene,

phenanthrene, and alkylated naphthalenes and phenanthrenes with a total percentage of about 4%.

Other studies indicated that volume fraction of PAHs in diesel fuels varies widely from 5% to

30%, even less than 5% for some high-quality fuels. Some heavy PAHs identified in DPM, such

as fluoranthene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene,

benzo(a)pyrene, indeno[1,2,3-cd]pyrene, dibenz(a,h)anthracene, and benzo(ghi)perylene, are not

identified in diesel fuel. This is consistent with some other studies where those heavy PAHs are

either not present or below detection level in diesel fuel [1, 2, 5].

1.3% of engine oil was able to be resolved by GC-MS, which are mainly organic acids. n-

Alkanoic acids from C6H12O2 to C18H36O2 are the main components and benzoic acid is the most

important aromatic acid in the 1.3% of engine oil that is amenable to identification by GC/MS

(Table 3-1).

84% of DPM can be dissolved in the solvent. This is consistent with the 82-86 % of

organic fraction obtained from thermal-optical analysis (NIOSH Method 5040, a standard

method for organic and elemental carbon (OC/EC) measurement) [11].

The consistency of dissolved fraction with the fraction of organic compounds also

indicated the high efficiency of the extraction method. In DPM, the identified organics account

for 3.47% of total DPM mass, which falls into the range of other studies [5, 12, 13].

64
 Among the identified fractions, n-alkanes account for 24.02%, branched alkanes 21.90%,

alkylated aromatic hydrocarbons 31.67% and n-alkanoic acids 20.23%. This distribution is

consistent with the DPM emissions from medium duty diesel trucks [5].

3.3.2 Alkanes

As suggested in Figure 3-2, alkanes, which include n-alkanes, branched alkanes and

saturated cycloalkanes, account for a large fraction of the diesel fuel and the identifiable portion

of DPM. The relative percentages of n-alkanes, branched alkanes and saturated cycloalkanes in

both the diesel fuel and DPM were illustrated in Figures 3-3(a-c) respectively.

A series of n-alkanes from C10 to C25 have been identified in diesel fuel and DPM. From

Table 3-1, the concentrations of n-alkanes in DPM range from several hundred to several

thousand µg/g while in diesel fuel the concentrations can reach as high as tens of thousands µg/g.

Figure 3-3(a) suggests that in the identified alkanes the most abundant species are distinct for

diesel fuel and DPM. For diesel fuel, the most abundant identified n-alkanes are C13-C17, which

is consistent with Schauer’s results [5]. While for DPM, the distribution represents a bell shape

with C17 as the most abundant. The relative fractions of C17-C21 are much higher in DPM than in

diesel fuel while the fractions of C10 to C16 are much less. The enrichment of aliphatic

hydrocarbons with higher carbon number was also reported by Rogge et al. [13, 14]. The shift to

heavier n-alkanes from diesel fuel to DPM can be the results of combustion synthesis, as shorter-

chained alkanes are more likely to undergo complete combustion than the heavier ones.

65
 Percentage of identified organics (%)
8.0
7.0 Diesel fuel
6.0 DPM
5.0
4.0
3.0
2.0
1.0
0.0
C10

C11

C12

C13

C14

C15

C16

C17

C18

C19

C20

C21

C22

C23

C24

C25
(a) n-Alkanes
Percentage of identified organics (%)
Percentage of identified organics (%)

1.4 1.6
1.4
1.2 Diesel fuel
Diesel fuel
DPM 1.2
1.0 DPM
1.0
0.8
0.8
0.6 0.6
0.4 0.4
0.2 0.2

0.0 0.0
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C14 C15 C18 C19 C20

(b) Branched alkanes (c) Saturated cycloalkanes

Figure 3-3 Relative percentage of alkanes in total identified compounds for low sulfur diesel
fuel and DPM.

For branched alkanes, the identified compounds are also known as isoprenoids. These

compounds are naturally present in crude oil and can be potentially used as tracers for diesel

engine emissions [5]. Branched alkanes phytane and pristane are produced by plankton from

isoprenoids; both are relatively resistant to breakdown and are abundant in new oil sediments.

Phytane is more persistent than pristane, so the phytane/pristane ratio is sometimes used to age

oil [15]. The ratio of pristane to heptadecane and phytane to octadecane for a given source of

66
diesel fuel is frequently distinctive enough to enable source identification in a spill investigation

[15]. These ratios also increase during biological degradation and are useful in evaluating the age

of environmental contamination [15].

The structures of selected isoprenoid alkanes are shown in Figure 3-4. Figure 3-3(b)

shows the distribution of the isoprenoids from C14 through C20 in diesel particulate matter

compared with diesel fuel used for the generator. As can be seen in the figures, DPM contains a

larger fraction of higher molecular weight compounds.

2,6,10-Trimethylundecane (Norfarnesane) 2,6,10-Trimethyldodecane (Farnesane)

2,6,10-Trimethyltridecane 2,6,10-Trimethylpentadecane (Norpristane)

2,6,10,14-Tetramethylpentadecane (Pristane) 2,6,10,14-Tetramethylhexadecane (Phytane)

Figure 3-4 Chemical structures of isoprenoids present in diesel fuel and diesel emissions.

Cycloalkanes are more toxic than alkanes or branched alkanes [15]. Figure 3-3(c) shows

the similar distributions of saturated cycloalkanes from C13 through C22 in DPM and diesel fuel.

The shift toward higher molecular weight compounds is similar to that of n-alkanes.

3.3.3 PAHs and Alkylated PAHs

Figure 3-5 and Figure 3-6 illustrate the relative percentages of dominant PAHs and

alkylated PAHs in diesel fuel and DPM. It can be seen that compared with diesel fuel, the

relative fractions of PAHs and alkylated PAHs are significantly higher in DPM.
67
 Percentage of identified organics (%)
0.25
Diesel fuel
0.20 DPM

0.15

0.10

0.05

0.00

Ind
Bbf
Bkf
Pyr
Nap

Ace
Flu
Phe

Baa

Bap

Dba
Bgp
Ant
Flt
Acy

Chy
Figure 3-5 Relative percentage of PAHs in total identified compounds in low sulfur diesel
fuel and DPM.
Percentage of identified organics (%)

12.0

10.0
Diesel fuel

8.0 DPM

6.0

4.0

2.0

0.0
MN DMN TMN MPh DMPh

Figure 3-6 Relative percentage of alkylated PAHs in total identified compounds in low sulfur
diesel fuel and DPM.

Many more PAH species are observed in DPM that are not present in diesel fuel, such as

benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene,

indeno[1,2,3-cd]pyrene, dibenz(a,h)anthracene, and benzo(ghi)perylene, with the mass range of

228-278 u (Figure 3-7). More phenanthrene and pyrene are identified in DPM than the diesel fuel.

68
This result indicates that many PAH compounds may originate from the combustion process in

the diesel engine rather than from direct evaporation of diesel fuel. These pyrogenic PAHs of

greater carcinogenic potency are precursors to soot particle formation and account for 14% of

total PAHs identified in DPM. Similar PAH distribution in DPM and diesel fuel was also

reported by Dobbins [16].

1.00
H/C Ratio of PAHs

0.80

0.60

0.40
Diesel fuel
0.20 DPM

0.00
100 150 200 250 300
Molecular Mass (u)

Figure 3-7 H/C ratio vs. PAH molecular mass found by GC/MS in low sulfur diesel fuel and
DPM.

For identified alkylated PAHs, the dominant species are trimethylnaphthalene,

methylphenanthrene and dimethylphenanthrene for DPM and dimethylnaphthalene and

trimethylnaphthalene for diesel fuel. This also suggests the enrichment of higher molecular

weight compounds in DPM. Consistent with Shauer et al. [5] and Koziel et al. [17], alkylated

PAHs are the most abundant identified species in our study.

3.3.4 Alkylbenzenes

Figure 3-8 represents percentages of alkylbenzenes from C1-benzene to C6-benzene in

low sulfur diesel fuel and DPM. Alkylbenzenes account for 6.55% of the identified fraction in

69
diesel fuel and 0.21% in DPM. It is suggested that alkylbenzenes have been significantly

consumed by the combustion processes. The dealkylation from alkylbenzenes and alkylated

PAHs and recombination of alkyl groups onto the reactants have altered the relative isomeric

distributions of these two categories [18].

Percentage of identified organics (%)

1.6
1.4
Diesel fuel
1.2
DPM
1.0
0.8
0.6
0.4
0.2
0.0
C1-B C2-B C3-B C4-B C5-B C6-B

Figure 3-8 Relative percentage of alkylbenzenes in total identified compounds in low sulfur
diesel fuel and DPM.

3.3.5 Organic Acids

Both benzoic and n-alkanoic acids ranging from C6 to C18 have been identified in DPM

and engine oil as listed in Table 3-1. The distributions of n-alkanoic acids in DPM and engine oil

are also shown in Figure 3-9.

In DPM, organic acids account for 20% of the total identified organics with

concentrations as high as several thousand µg/g. n-Dodecanoic (C12), n-tetradecanoic (C14), n-

pentadecanoic (C15), and n-hexadecanoic acid (C16) are the most abundant species identified.

From Figure 3-9, similar distribution of n-alkanoic acids is observed in engine oil, which

indicates that the organic acids from DPM may be originated from the evaporation of engine oil.

70
 60
Percentage of identified organics (%) 7
50 6
5
4
40 3
2
1
30 0
C7

C8

C9

C10

C11

C12

C13

C14

C15

C16
20
Engine oil
10 DPM

0
C6

C7

C8

C9

C10

C11

C12

C13

C14

C15

C16

C17

C18
Figure 3-9 Relative percentage of n-alkanoic acids from C6H12O2 to C18H36O2 in DPM and
engine oil.

3.3.6 Isomer Distribution

From the results, the distribution variation of isomers in diesel fuel and DPM was

observed. The relative percentages of isomers for methylnaphthalenes and methylphenanthrenes

in Ultra-low sulfur diesel fuel and DPM are shown in Figure 3-10 (a) and (b). Figure 3-10

indicates that the amount of 2-isomers is higher than 1-isomers in both diesel fuel and DPM. The

theory of heat of formation can be used to explain the higher abundance of 2-isomers. Heat of

formation of a compound refers to the heat evolved (heat of formation is negative) or absorbed

(heat of formation is positive) during the formation of one mole of a substance from its

component elements. Higher heat of formation (positive) means more heat is needed to form this

compound. That means it’s more difficult to form a compound with higher heat of formation.

The heat of formation is 32.10 kcal/mol for 1-methylnapthalene, 31.23 kcal/mol for 2-

methylnapthalene, 51.23 kcal/mol for 1-methylphenanthrene, and 47.95 kcal/mol for 2-

71
methylphenanthrene (Heat of Formation calculated by WinMopac7.21 (PM3)). Therefore, 2-

isomers are more stable and abundant in environment.

90 80
80 1-MN 70 1-MPh
70
Percentage (%)

Percentage (%)
2-MN 60 2-MPh
60
50
50
40
40
30 30
20 20
10 10
0 0
Diesel Fuel DPM Diesel Fuel DPM

(a) methylnaphthalenes (b) methylphenanthrene

Figure 3-10 Relative percentage of isomers in low sulfur diesel fuel and DPM.

Table 3-2 listed the ratio of 1-/2-isomers in diesel fuel and DPM. For the

methylnaphthalenes, the isomer ratio in DPM is relatively consistent with that of the diesel fuel.

This can be an indication that the two isomers are consumed or formed at similar rates, and this

may be caused by the close resemblance of the heats of formation between the two. For the

methylphenanthrenes, however, higher concentrations are found in the DPM, which suggests that

the two isomers may be pyrogenic. Higher 1-MPh/2-MPh ratio in DPM than in diesel fuel

indicates 1-MPh is formed more than 2-MPh. The phenanthrene groups are more affected by

combustion, which promotes the interest in studying phenanthrene enriched fuel.

Table 3-2 Ratio of 1-MN/2-MN and 1-MPh/2-MPh in diesel fuel and DPM.
1-MN/2-MN 1-MPh/2-MPh
Diesel fuel 0.26 0.46
DPM 0.35 0.82

72
3.4 Conclusions

The chemical composition of low sulfur diesel fuel, engine oil and diesel particulate

matter emitted from a non-road diesel generator has been investigated. A total of approximately

90 organic compounds are identified and quantified, including a series of n-alkanes, branched

alkanes, saturated cycloalkanes, PAHs, alkylated PAHs, alkylbenzenes, n-alkanoic acids and

aromatic acids. It is shown that the organic composition of DPM is mainly resulted from three

sources: n-alkanes in DPM are likely from unburned diesel fuel, the main source of organic acids

is from engine oil evaporation, and most PAHs and alkylated PAHs are likely formed during

combustion. Compared with diesel fuel, DPM contains fewer fractions of alkanes and more PAH

compounds. The enrichment of higher molecular weight alkanes, including n-alkanes, branched

alkanes and cycloalkanes, is likely resultant from fuel combustion. The pyrogenic process also

resulted in the loss of alkylbenzenes and the alteration of isomeric distribution of PAHs and

alkylated PAHs.

3.5 References

1. Diesel fuel and exhaust emissions. 1996, World Health Organization, Geneva. p. 11-13.

2. Sjogren, M.; Li, H.; Rannug, U.; et al., A multivariate statistical analysis of chemical
composition and physical characteristics of 10 diesel fuels, Fuel, 74(7): 983-989, 1995.

3. Saiyasitpanich, P.; Lu, M.M.; Keener, T.C.; et al., The effect of diesel fuel sulfur content
on particulate matter emissions for a nonroad diesel generator, J. Air Waste Manage.
Assoc., 55(7): 993-998, 2005.

4. Kawamura, K.; Ng, L.L.; and Kaplan, I.R., Determination of organic acids (C1-C10) in the
atmosphere, motor exhausts, and engine oils, Environ. Sci. Technol., 19(11): 1082-1086,
1985.

5. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; et al., Measurement of emissions from air
pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks,
Environ. Sci. Technol., 33(10): 1578-1587, 1999.

73
6. Zielinska, B.; Sagebiel, J.; Arnott, W.P.; et al., Phase and size distribution of polycyclic
aromatic hydrocarbons in diesel and gasoline vehicle emissions, Environ. Sci. Technol.,
38(9): 2557-2567, 2004.

7. Chan, S.H.; and He, Y.S., Measurements of particulate mass concentration using a
tapered-element oscillating microbalance and a flame ionization detector, Meas. Sci.
Technol., 10(4): 323-332, 1999.

8. Pal, R.; Juhasz, M.; and Stumpf, A., Detailed analysis of hydrocarbon groups in diesel
range petroleum fractions with on-line coupled supercritical fluid chromatography gas
chromatography mass spectrometry, J. Chromatogr. A, 819(1-2): 249-257, 1998.

9. Hsu, C.S., Diesel Fuels Analysis, in Encyclopedia of Analytical Chemistry. Wiley: New
York. pp. 6613-6622, 2000.

10. Chemistry of diesel fuels. ed. Song, C.;Hsu, C.S.; and Mochida, I. Taylor & Francis: New
York. p. 18-19, 2000.

11. Liu, Z.F.; Lu, M.M.; Birch, M.E.; et al., Variations of the particulate carbon distribution
from a nonroad diesel generator, Environ. Sci. Technol., 39(20): 7840-7844, 2005.

12. McDonald, J.D.; Zielinska, B.; Sagebiel, J.C.; et al., Characterization of fine particle
material in ambient air and personal samples from an underground mine, Aerosol Sci.
Technol., 36(11): 1033-1044, 2002.

13. Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; et al., Sources of fine organic aerosol. 2.
Noncatalyst and catalyst equipped automobiles and heavy duty diesel trucks, Environ. Sci.
Technol., 27(4): 636-651, 1993.

14. Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; et al., Sources of fine organic aerosol.8.
Boilers burning No. 2 distillate fuel oil, Environ. Sci. Technol., 31(10): 2731-2737, 1997.

15. Alkanes Entry, in Environmental Contaminants Encyclopedia, Irwin, R.J., Ed. National
Park Service: Fort Collins,CO, 1997.

16. Dobbins, R.A.; Fletcher, R.A.; B. A. Benner; et al. PAHs in flames, diesel fuels and
diesel emissions. in 30th International Symposium on Combustion, Work-In-Progress
Poster Session. July 25-30, 2004. Chicago, IL.

17. Koziel, J.A.; Odziemkowski, M.; and Pawliszyn, J., Sampling and analysis of airborne
particulate matter and aerosols using in-needle trap and SPME fiber devices, Anal. Chem.,
73(1): 47-54, 2001.

18. Yang, J.; and Lu, M., Thermal growth and decomposition of methylnaphthalenes,
Environ. Sci. Technol., 39(9): 3077-3082, 2005.

74
 Chapter 4

PHASE DISTRIBUTION OF SEMI-VOLATILE ORGANIC

COMPOUNDS IN DIESEL COMBUSTION AEROSOLS

4.1 Introduction

4.1.1 Gas/Particle Partitioning

Diesel emissions consist of a large number of organic compounds [1-3], among which

some polycyclic aromatic hydrocarbons (PAHs), such as chrysene and benzo[a]pyrene, are

known or suspected human carcinogens [4-9]. Studies have indicated that occupational exposure

to diesel exhaust may be related to lung cancer for bus garage workers, miners, and forklift

workers etc. [2, 10]. It is well established that exposure to PAHs and their derivatives may result

in a higher health risk [11]. Therefore, there is a necessity to understand the physical-chemical

behavior of the carcinogenic and toxic organic compounds present in diesel exhaust. One of the

important physical-chemical processes determining the behavior of organic compounds is the

partitioning between the gas and particle phases. Such gas/particle (G/P) partitioning determines

the distribution of a given compound between the two phases and is strongly related to human

exposure. With pure compound liquid vapor pressures ( p L0 ) in the range of 10-6-100.0 Pa (~10-8-

10-2 Torr), semi-volatile organic compounds (SVOCs), which include many of the PAHs and

substituted PAHs, are present significantly in both gas and particle phases due to their

intermediate volatilities. Therefore, SVOCs are of special interest in the G/P partitioning studies.

Volatile organic compounds (e.g., benzene) are mostly present in the gas phase, and low-

volatility organic compounds are almost exclusively in the particulate phase.

75
 The significance of understanding G/P partitioning for these organic compounds lies in

the fact that their health effects are strongly related to the physical forms in which the

compounds are present when exposure occurs. In addition, G/P partitioning is important in

predicting the behavior of diesel emission tracers and the effectiveness of control technologies in

removing diesel aerosols.

4.1.2 Partitioning Constant

There have been some theoretical and experimental studies in understanding the

distribution of chemical species between the gas and particulate phases. Junge [12] proposed a

linear Langmuir isotherm equation to describe gas/particle partitioning in the atmosphere. The

partitioning of organic compounds between gas and particulate phases can be parameterized as

follows [13, 14]:

F / TSP
Kp = (4-1)
A

where Kp is a compound- and temperature-dependent partitioning coefficient (m3/µg), F is the

particle phase concentration of the compound of interest (ng/m3), A is the gas phase

concentration (ng/m3), and TSP is the amount of total suspended particulate matter (µg/m3).

Generally, F represents the mass of a compound measured on a filter, and A represents the mass

of a compound measured on a downstream adsorbent.

The constant Kp represents the sorbed/gaseous concentration ratio. Increasing Kp implies

increasing partitioning to the solid phase. Multiplying Kp by TSP gives a ratio of the particle to

gas phase concentrations:

Kp × TSP = F/A (4-2)

76
When Kp × TSP is greater than 1, the compound partitions primarily into the particle phase; Kp ×

TSP values less than 1 indicate partitioning primarily to the gas phase.

4.1.3 Partitioning Mechanisms and Predictive Models

Seinfeld and Pandis [15] summarized theories describing gas/particle partitioning of

organic compounds. Dissolution and adsorption are the two most important mechanisms, which

are similar to the absorption and adsorption mechanisms suggested by Pankow et al. [14, 16-21].

When organic species are already present in particles, organic vapors tend to dissolve in the

particle phase organics. This is referred to as dissolution. The gas/particle partitioning of organic

vapor species via dissolution is determined by the molecular properties of the species and the

organic phase and the amount of the particulate organic phase available for the vapor phase

compound to dissolve into. When an organic particulate phase is absent, adsorption is the

mechanism to initiate gas/particle partitioning. The adsorption process involves two steps: the

particle surface is partially covered by vapor molecules, and a monolayer is formed.

Occasionally additional layers can be formed following the monolayer, which can be regarded as

the third step. The adsorption behavior of a surface is generally characterized by an adsorption

isotherm. Langmuir isotherm is frequently used to describe the equilibrium adsorption for the

formation of a monolayer. One of the most widely used isotherms is based on the BET theory,

which is an extension of the Langmuir isotherm to include the adsorption of two or more

molecular layers.

Pankow and co-workers have developed the modeling framework for the sorption

processes in a series of studies [14, 16-21]. Two mechanisms describing the gas/particle

partitioning of SVOCs have been recognized: adsorption onto the aerosol surface and absorption

77
into the aerosol organic matter. When the partitioning is dominated by adsorption, the partition

coefficient is given by [14]:

N s ATSP Te ( Q1 −Q2 ) / RT
Kp = (4-3)
16 p L0

where Ns is the number of adsorption sites (mol/cm2), ATSP is the surface area of the TSP

(cm2/µg), T is the temperature (K), Q1 and Q2 are the enthalpies of desorption and volatilization,

respectively (kcal/mol), R is the universal gas constant (= 8.314×10-3 kJ/mol·K = 0.00199

kcal/mol·K = 8.2×10-5 m3·atm/mol·K), and p L0 is the sub-cooled liquid vapor pressure of the pure

compound (torr). When partitioning is dominated by absorption into organic matter of aerosol,

Kp is given by [19]:

760 f om RT
Kp = (4-4)
MWom γp L0 10 6

where fom is the fraction of organic matter on the TSP, MWom is the mean molecular weight of

the organic matter (g/mol), and γ is the activity coefficient of the absorbate in the organic matter.

Usually, it is difficult to experimentally determine the predominant mechanism in the

environment. In any given situation, both adsorptive and absorptive partitioning may exist

simultaneously. The observed value of Kp will then contain contributions from both mechanisms

so that for a given compound [19]:

1 ⎡ N s ATSP Te (Q1 −Q2 ) / RT 760 f om RT ⎤

Kp = ⎢ + ⎥ (4-5)
p L0 ⎣ 1600 MWom γ 10 6 ⎦

78
The actual relative importance of adsorptive and absorptive partitioning in a given situation will

depend on the values of the various parameters in Eq. (4-5); increasing ATSP and e (Q1 −Q2 ) / RT will

favor adsorption; increasing fom, decreasing MWom, and decreasing γ will favor absorption.

Regardless of the relative importance of adsorptive vs. absorptive partitioning, the

partition coefficient is inversely proportional to the sub-cooled liquid vapor pressure. Taking the

logarithm of Eq. (4-5) gives the linear relationship shown in Eq. (4-6):

log K p = mr log p L0 + br (4-6)

where mr and br (expressed as Eq. (4-7)) are the slope and intercept of the linear regression,

respectively.

N s ATSP Te (Q1 −Q2 ) / RT 760 f om RT

br = + (4-7)
1600 MWom γ 10 6

A plot of log Kp vs. log p L0 for a series of compounds may give a slope mr = -1 if br

remains constant for the compounds of interest, which occurs only when γ is equal for each

compound in the absorbed phase and particle size effects are negligible. However, γ may vary

depending on the absorbing aerosol [21]. Therefore, values calculated from Eq. (4-5) are often

verified with field measurements. The deviations from mr = -1 are possibly attributed to

sampling artifacts, non-exchangeable material within particles, or non-equilibrium partitioning.

4.1.4 Impact Factors on Kp Values

In general, gas/particle partitioning of organic compounds is a function of a number of

physical and chemical factors. First of all, the state of the compound in the emission (e.g. vapor,

adsorbed on the surface of particles or absorbed into particles) implies the possible partitioning

79
mechanisms which directly determine the Kp values to be estimated from Eq. (4-3) or Eq. (4-4).

The second important factor is the physical and chemical properties of the compound which may

include the vapor pressure of the compound ( p L0 ), the reactivity and stability of the compound

represented by its activity coefficient (γ), and the affinity of the compound for the particle’s

organic matrix which is parameterized by Q1 and Q2 in Eq. (4-5). The amount of particles, which

is expressed in terms of the number of adsorption sites (Ns) and the surface area of the TSP

(ATSP), and the composition of particles, which is described by the fraction of organic matter (fom)

[22], are essential for determining Kp when the dominant partitioning mechanism is adsorption

and absorption, respectively. Pankow [16] has reported that low TSP level increases the

deviation of calculated log[(F/TSP)/A] from the estimated Kp from Eq. (4-3). In addition, the

collection methodology and sampling conditions, i.e. sampling temperature (T) [16], relative

humidity (RH) [23], and filters (quartz or Teflon) [24], and sampling duration [25] may also

influence the partitioning.

4.2 Experimental Method

The experiments were performed on a Generac diesel generator with burning low sulfur

containing diesel fuel (400 ppmw S). The samples were collected using the developed dilution

method (DR = ~3.4), with particle phase emissions collected on quartz fiber filters and gas phase

emissions collected by the PUF/XAD/PUF cartridge. A detailed description of the experimental

setup and procedures, sampling methodology, and extraction and analytical methods has been

described in Chapter 2.

80
4.3 Results and Discussion

4.3.1 Phase Distribution

The gas phase and particle phase n-alkanes emitted from the diesel generator at 0 kW are

shown in Figure 4-1. Overall, the amount of n-alkanes in gas phase is much higher than the

amount in particle phase. Volckens and Leith [26] have reported that when the value of F/A is

greater than 1, the compound is primarily present in the particle phase; the value less than 1

indicates partitioning primarily to the gas phase. The experimental results in this study are

consistent with the theory: lighter alkanes (carbon number between 10 and 19) with F/A values

in the range of 0.0-0.41 are predominantly in the gas phase, while heavier alkanes (carbon

number higher than 20), which have the F/A values between 1.48 and 31.2, are primarily present

in the particle phase. Table 4-1 listed the vapor pressure for some alkanes and PAHs at 25 ºC,

which is close to the sampling temperature. From Figure 4-1 and the vapor pressure listed in

Table 4-1, alkanes occur primarily in gas phase when the vapor pressures are higher than 10-4.36

torr and mainly in particulate phase when the vapor pressures are lower than 10-4.84 torr.

250000
Gas phase
Concentration (ng/m3)

200000
Particle phase

150000

100000

50000

0
C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25

Figure 4-1 The distribution of n-alkanes in gas and particle phase diesel emissions (400
ppmw S, 0 kW, DR = 3.4).

81
Table 4-1 Vapor pressures for some alkanes, PAHs and alkanoic acids at 25 ºC (torr).
0 0
Compounds Log p L Ref. Compounds Log p L Ref.
n-Hexadecane (C16) -3.16 [27] Fluoranthene (Flt) -4.07 [27]
n-Heptadecane (C17) -3.27 [27] Pyrene (Pyr) -4.24 [27]
n-Octadecane (C18) -3.72 [27] Benzo(a)anthracene (Baa) -5.71d [21, 23, 28]
n-Nonadecane (C19) -4.36 [27] Chrysene (Chy) -5.42 [27]
n-Eicosane (C20) -4.84 [27] Benzo(b)fluoranthene (Bbf) -7.12a [15]
n-Heneicosane (C21) -5.18 [27] Benzo(k)fluoranthene (Bkf) -7.13a [15]
n-Docosane (C22) -5.70 [27] Benzo(a)pyrene (Bap) -7.33a [15]
n-Tricosane (C23) -6.43a [23] Dodecanoic acid (C12) -5.48e [28]
n-Tetracosane (C24) -6.58b [29] Tetradecanoic acid (C14) -5.84 [30]
n-Pentacosane (C25) -7.08c [29] Pentadecanoic acid (C15) -6.12 [30]
Fluorene (Flu) -2.24 [27] Hexadecanoic acid (C16) -7.11 [30]
Phenanthrene (Phe) -2.98 [27] Heptadecanoic acid (C17) -7.08 [30]
Anthracene (Ant) -3.01 [27] Octadecanoic acid (C18) -8.39 [30]
a
@ 20 ºC. b,c @ 31 ºC. d The vapor pressure was estimated from the vapor pressures at other temperatures
listed in references [21, 23, 28]. e @ 15 ºC.

Figure 4-2 represents the PAH distribution between gas phase and particle phase diesel

emissions at 0 kW. As Figure 4-2 shows, 2- and 3-ring PAHs ( p L0 > 10-3.01 torr) are

predominantly present in gas phase emission, 4- to 6-ring PAHs ( p L0 < 10-4.07 torr) are primarily

present in particulate phase with 5- and 6-ring PAHs ( p L0 < 10-5.71 torr) occurring exclusively in

particulate phase. The correlation between the gas/particle distribution of PAHs and their vapor

pressures indicates that the relative abundance in the particulate phase increases as the vapor

pressure decreases. Comparing G/P distributions for PAHs and alkanes, it has shown that the

boundary vapor pressure between gas and particle phase differs for PAHs and alkanes, which

suggests that the G/P partitioning depends not only on the vapor pressure, but also on some other

physical-chemical properties of the compound. These physical-chemical properties include the

reactivity and stability of the compound (γ), and the affinity of the compound for the particle’s

organic matrix (Q1 and Q2). These parameters, especially the activity coefficient γ, determine the
82
differed br values according to Eq. (4-7). Higher activity coefficients for PAHs than those for

alkanes were reported by Jang et al. [21], which may be a confirmation for the prediction in this

study.

10000
Concentration (ng/m3)

Gas 200
8000 150
Particle 100
6000 50
0
4000

Chy

Bbf

Bkf
Baa

Bap

Ind

Dba
Bgp
2000

Ind
Bbf

Bkf
Pyr
Nap

Ace

Flu
Phe

Baa

Bap

Dba

Bgp
Ant

Flt
Acy

Figure 4-2 Chy

The distribution of PAHs in gas and particle phase diesel emissions (400 ppmw S,
0 kW, DR = 3.4).

Figure 4-3 shows the concentrations of alkanoic acids in both gas and particle phase

diesel emissions at 0 kW. Alkanoic acids with a carbon number between 6 and 12 ( p L0 > 10-5.48

torr) are primarily in gas phase, and particulate phase contains more fractions of high-molecular

weight acids with the vapor pressures lower than 10-5.48 torr. It has been expected that the

gas/particle partitioning of alkanoic acids differs from that of alkanes due to the differences

between their physical-chemical properties, such as polarity and activity. The experimental

results have indicated that alkanoic acid partitions less into particle phase than the alkane and

PAH with similar vapor pressure. For instance, tetradecanoic acid (C14) has similar vapor

pressure as n-docosane (C22) and benzo(a)anthracene (Baa) (Table 4-1), the ratio of particle

phase concentration/gas phase concentration (F/A) for the former is lower than that for the latter

two: the F/A value is 1.60 for tetradecanoic acid, while 5.91, and 5.51 for n-docosane and

benzo(a)anthracene, respectively. The reactivity and stability of the compound and the affinity of

83
the compound for the particle’s organic matrix are the two important factors affecting G/P

partitioning. It has been found that alkanoic acids have similar activities coefficients as n-alkanes

(~1.0). Therefore, the affinity, which may be related to the polarity of the compound, is the

possible reason leading to the difference in their partitioning behaviors of alkanoic acids and n-

alkanes.

12000
Gas phase
Concentration (ng/m3)

10000
Particle phase
8000

6000

4000

2000

0
C6

C7

C8

C9

C10

C11

C12

C13

C14

C15

C16

C17

C18
Figure 4-3 The distribution of alkanoic acids in gas and particle phase diesel emissions (400
ppmw S, 0 kW, DR = 3.4).

Though the above discussions are based on the data at 0 kW, similar results and

conclusions can be obtained for other load conditions.

4.3.2 Log Kp vs. log p L0

The partitioning coefficient (Kp, m3/µg) was calculated from the measured gas and

particle phase concentrations of a compound (A and F, ng/m3) and the TSP concentration (µg/m3)

in diesel emissions according to Eq. (4-1). Log Kp values were plotted vs. log p L0 (torr, @ 25 ºC)

for n-alkanes and PAHs (Figure 4-4 and Figure 4-5, respectively) and the linear regressions were

illustrated in the two figures as well. The slope (mr), intercept (br), and regression coefficient (R2)

84
values for the regressions were given in Table 4-2. Table 4-3 summarized the slopes and

intercepts reported by other studies [17, 26, 31-33].

-1
Measured Kp at 0 kW
-2 Measured Kp at 25 kW

Measured Kp at 50 kW
Log Kp (m3/µg)

-3
Measured Kp at 75 kW

Expected Kp
-4
(Adsorption+Absorption)
Expected Kp
(Adsorption)
-5 Expected Kp
(Absorption)

-6
-6 -5 -4 -3 -2
Log pL0, torr

Figure 4-4 Log Kp (m3/µg) vs. log p L0 (torr, @ 25 ºC) based on experimental measurements
and estimations for n-alkanes in diesel emissions at 0-75 kW.

-1

Measured Kp at 0 kW
-2
Measured Kp at 25 kW
Log Kp (m3/µg)

-3 Measured Kp at 50 kW

Measured Kp at 75 kW

-4 Expected Kp
(Adsorption+Absorption)
Expected Kp
(Adsorption)
-5 Expected Kp
(Absorption)

-6
-6 -5 -4 -3 -2
Log pL0, torr

Figure 4-5 Log Kp (m3/µg) vs. log p L0 (torr, @ 25 ºC) based on experimental measurements
and estimations for PAHs in diesel emissions at 0-75 kW.

85
Table 4-2 Values of mr and br in the equation of log Kp = mr log p L0 + br at 25 ºC for alkanes
and PAHs based on the experimental data and model estimations.
n-Alkanes PAHs
mr br R2 mr br R2
Experimental data
0 kW -0.83 -7.46 0.99 -0.64 -6.14 0.92
25 kW -0.82 -7.69 0.97 -0.62 -6.58 0.95
50 kW -0.68 -7.81 0.97 -0.61 -6.81 0.94
75 kW -0.66 -8.03 0.95 -0.63 -7.12 0.94
Model estimations
Adsorption -1.00 -7.53 NAa -1.00 -7.50 NA
Absorption -1.00 -7.59 NA -1.00 -7.62 to -8.19 b NA
Adsorption + Absorption -1.00 -7.26 NA -1.00 -7.27 to -7.45 b NA
a
Not available since mr and br values were obtained from Eq. (4-3), Eq. (4-4), and Eq. (4-5). b The value
of br differed for different compounds due to the varied values of activity coefficient (γ).

These figures and tables have suggested that the correlation of log Kp vs. log p L0 was

statistically significant, which was also reflected by the high R2 values (R2 > 0.92). The slopes of

log Kp vs. log p L0 plots range from -0.66 to -0.83 for n-alkanes, and from -0.61 to -0.64 for PAHs,

which are close to the expected value of -1. The corresponding intercepts br range from -7.46 to -

8.03 and from -6.14 to -7.12 for n-alkanes and PAHs, respectively. The slopes and intercepts

determined in this study are in the range reported by other studies (Table 4-3) although some

differences are observed. The dissimilarities between the results in this study and in other studies

may be caused by different aerosol type and sampling method. In this study, diesel aerosols were

collected from a direct source, while in most other studies, ambient aerosols were collected.

Sampling method (high-vol or not), filter type (glass fiber, Teflon, or quartz fiber filter), and

sampling condition (temperature and humidity) may also result in the dissimilarities between

studies.

86
Table 4-3 Literature values of mr and br in the equation of log Kp = mr log p L0 + br for gas/particle partitioning at urban areas.
Compound type Temperature (ºC) Locationa mr br Reference
Alkanes NA Portland -0.70 to -1.15 -6.05 to -8.51 Hart (1989) cited in [17]
Alkanes 5 Denver -0.86 -7.29 Foreman and Bidleman [34]b
Alkanes 37 Los Angeles -1.03 -8.68 Liang et al. [29]c
Alkanes 25-32 Athens -0.28 to -0.57 -4.27 to -5.87 Mandalakis et al. [31]d
PAHs 5 Portland -0.69 to -1.06 -5.74 to -7.35 Hart (1989) cited in [17]
PAHs 7 Portland -1.09 -7.42 Ligocki and Pankow [35]e
PAHs 37 Denver -0.76 -6.71 Foreman and Bidleman [34]b
PAHs 6-8 Baltimore Harbor Tunnel -0.57 to -0.76 -6.05 to -6.61 Benner et al. [36]f
PAHs 6-10 Osaka, Japan -0.86 to -1.17 -6.92 to -9.05 Yamasaki et al. [13]g
PAHs -10 to 3.7 Chicago -0.34 to -1.04 -3.48 to -8.29 Cotham and Bidleman [37]h
PAHs NA Chicago -0.64 -3.47 Simcik et al. [38]i
PAHs NA Lake Michigan -0.67 -3.71 Simcik et al. [38]i
PAHs NA Columbia -1.18 -9.06 Foreman and Bidleman [39]j
PAHs 25-32 Athens -0.47 to -0.70 -4.52 to -5.97 Mandalakis et al. [31]d
PAHs in diesel exhaust aerosols at 5 ºC -0.67 to -1.06 -4.96 to -10.4 Volckens and Leith [26]k
a
In most of these studies, samples were ambient aerosols. b Sampling method: high-vol sampling (0.35-0.5 m3/min) with glass fiber filter
(GFF)/PUF/PUF. c Aerosols were generated in a Teflon smog chamber under controlled conditions. Samples were collected on GFF/PUF/PUF at a
flow rate of ~70 L/min. d Sampling method: high-volume air sampler, GFF/PUF; sampling duration: 12 or 24 h; sampling volume: 900 or 450 m3.
e
Sampling method: high-volume sampling with GFF/GFF/PUF/PUF or Teflon membrane filter (TMF)/TMF/PUF/PUF; flow rate: 100-190 L/min;
face velocity: 30-60 cm/s. f Samples were collected in exhaust rooms of the Baltimore Harbor Tunnel. Sampling method: high-volume sampling
with TMF/PUF or GFF/PUF; flow rate: ~0.8 m3/min. g Sampling method: high-volume air sampler with GFF/PUF/PUF; flow rate: 0.75-0.8
m3/min; linear velocity on the GFF: 30.6-32.6 cm/s. h Sampling method: high-volume sampling with GFF/GFF/PUF/PUF; flow rate: 0.5 m3/min;
sampling volume: 260-390 m3. i Sampling method: high volume air sampler equipped with GFF/PUF. j Sampling method: high volume air sampler
(1.0 m3/min) with GFF/PUF/PUF/PUF. k Four types of sampler were used to collect diesel exhaust aerosols at a flow rate of 4.2 L/min for each
sampler: (1) XAD denuder/Fiberfilm PTFE coated GFF/XAD denuder, (2) Fiberfilm PTFE coated GFF/Fiberfilm PTFE coated GFF/XAD
denuder, (3) electrostatic sampler/XAD denuder, and (4) Zeflour PTFE filter/XAD denuder/XAD denuder.

87
 Kp values can also be estimated from Eq. (4-3), (4-4), and (4-5). In the estimation of Kp

values, the values for some parameters in the equations were selected based on some literature.

Pankow [14] presented that for simple physical sorption, Ns is not very compound-dependent,

and is of the order of 4 × 10-10 mol/cm2. Thus 4.0 × 10-10 mol/cm2 was selected for Ns in this

study. Regarding the ATSP for urban aerosol, Corn et al. [40] have measured by the BET method

and obtained the values in the range of 0.019-0.031 cm2/µg for particulate matter collected in

Pittsburgh, Pennsylvania. Calculations based on particle-size distribution data reported by

Whitby et al. [41] for Pasadena, California give an ATSP of about 0.05 cm2/µg. Since the data of

Corn et al. [40] was obtained by direct surface area measurements, a value of 0.025 cm2/µg was

selected for ATSP by Pankow [14] and was used in this study as well. The value of (Q1-Q2) has

been reported to be in the range of +2 to +4 kcal/mol for PAHs by Yamasaki et al. [42] and was

selected as +3 kcal/mol in this study. The values of fom and MWom depend on the type of

particulate matter and fom = 0.5 and MWom = 300 g/mol were assumed in this study based on the

literature data [26]. When studying the partitioning of compounds with different polarities and

functional groups, activity coefficient (γ) should be considered since γ differs from compound to

compound [21]. Table 4-4 listed activity coefficients for n-alkanes, some PAHs, and alkanoic

acids summarized from the literature.

Table 4-4 Activity coefficients for n-alkanes, PAHs and alkanoic acids from literature.
Compounds γ Reference Compounds γ Reference
n-Alkanes 1.2 Jang et al. [21] Pyrene 4.3 Jang et al. [21]
Fluorene 1.3a Chandramouli et al. [43] Benzo(a)anthracene 4.8 Jang et al. [21]
Phenanthrene 2.8 Jang et al. [21] Chrysene 4.8 Assumedb
Anthracene 2.8 Jang et al. [21] Alkanoic acids 1.1 Jang et al. [21]
Fluoranthene 4.3 Jang et al. [21]
a
The activity coefficient of fluorene was assumed the same as that of fluorene-d10 considering their
similar physical-chemical properties. b The activity coefficient of chrysene was not available and was
assumed the same as benzo(a)anthracene in this study.

88
 The estimated log Kp values were also plotted vs. log p L0 in Figure 4-4 and Figure 4-5 and

the calculated mr and br values were listed in Table 4-2. The estimated Kp values for alkanes are

similar for adsorption, absorption, and combination mechanisms, while for PAHs the difference

among the three mechanisms is evident. As described above, the activity coefficient is important

in gas/particle partitioning of the compound, and increasing the activity coefficient can cause the

partitioning to favor absorption. As Table 4-4 shows, the activity coefficient is near 1.2 for n-

alkanes while it increases from 1.3 for 3-ring PAHs to 4.8 for 4-ring PAHs. As a result of the

increased activity coefficient for heavy PAHs, adsorption may become the dominant mechanism

for low-vapor pressure compounds, which is also indicated by Figure 4-5: the measured Kp

values approach the Kp values estimated from the absorption equation when γ is low.

Table 4-2 shows that the measured mr and br values deviate from the theoretical values,

which might be caused by the sampling bias and the variability of compound properties [17, 44].

The effects from various sources on the measured values of mr and br were discussed by Pankow

and Bidleman [17] and were summarized as follows. (1) The variability in the compound-to-

compound differences in (Q1-Q2) could affect both mr and br values. (2) The event-to-event

variability in ATSP and T will not affect mr while it can cause variability in br of about ±0.5 and

±0.15, respectively. (3) When some fractions of sorbing compounds are bound and cannot

exchange with the gas phase, the measured log Kp values will be higher, the mr value is

shallower, and br value is greater than the expected values at full equilibrium. (4) The particle-to-

gas desorption of compounds will result in a steep mr and a low br. (5) The gas-to-particle

adsorption of compounds will result in a shallow mr and a high br. (6) Within a given sampling

event, if ambient concentrations (e.g. A, F, and TSP) increase and/or ambient temperature (T)

decreases, the measured value of mr will become more shallow and br will be increased;

89
inversely, if ambient concentrations decrease and/or T increases, the measured value of mr will

steepen and br will be lowered. The six types of effects have been illustrated in Figures 4-6(a-f).

Figure 4-6 Sources contributing to the variability in mr and br values.

(a) Variability in the value of mq (the slope of a linear regression of Q1 against Q2
for a group of compounds sorbing on a specific type of particulate matter); (b)
variability in ATSP and T; (c) non-exchangeable effects; (d) desorption kinetics; (e)
adsorption kinetics; (f) changing concentrations or changing T within a given
event. (Pankow and Bidelman [17])

Figure 4-5 and Table 4-2 have presented shallow mr and high br values for PAHs under

the measurement conditions compared with predicted values. Based on the study by Pankow and

Bidleman [17], gas-to-particle adsorption of PAHs and non-exchangeable effects may occur

during the sampling events. However, regarding the partitioning of n-alkanes, the shallow mr and

low br values match none of above-mentioned six situations. Another study by Volckens and

Leith [44] has indicated that when particle phase mass is measured erroneously as gas phase, a

shallow mr and low br would be observed for the measurement, which matches the partitioning

90
results for n-alkanes in this study. This result suggests that the filter breakthrough for n-alkanes

may occur during the sampling events due to the high load of n-alkanes and/or long sampling

time. It is possible that gas-to-particle adsorption for PAHs and filter breakthrough for n-alkanes

occur simultaneously, since the concentration of n-alkanes in diesel exhaust is much higher than

that of PAHs.

From Figure 4-4 and Figure 4-5, it also has been observed that for both PAHs and n-

alkanes, the regression lines are almost parallel for four load conditions and move downward as

the load increases. According to the discussion by Pankow and Bidleman, this result means that

the temperature T increases and the surface area of the TSP (ATSP, cm2/µg) decreases as the load

increases, which is true since the combustion temperature is higher and the particle size is larger

at higher loads.

4.4 Conclusions

The gas and particulate phase diesel exhausts from a non-road diesel generator were

collected with the developed dilution method under different load conditions, the gas/particle

partitioning of compounds was investigated, and the measured partitioning coefficients were

compared with the estimated values from adsorption, absorption, and combination mechanisms.

The results have indicated that the amount of compounds in gas phase is much higher than the

amount in particle phase. Light n-alkanes with the vapor pressures higher than 10-4.36 torr (C10-

C19) occur primarily in gas phase emissions, and heavy n-alkanes with vapor pressures lower

than 10-4.84 torr are mainly in particulate phase. 2- and 3-ring PAHs ( p L0 > 10-3.01 torr) are

primarily present in gas phase emission, while 4- to 6-ring PAHs ( p L0 < 10-4.07 torr) occur

primarily or exclusively in particulate phase.

91
 The G/P partitioning coefficient Kp of a compound could be related to its vapor pressure

p L0 by the equation: log Kp = mr log p L0 + br. The linear regression results have indicated that the

correlation of log Kp vs. log p L0 is statistically significant. For n-alkanes, the slopes (mr) of log Kp

vs. log p L0 plots range from -0.66 to -0.83 and the intercepts (br) range from -7.46 to -8.03. The

shallow mr and low br values compared with estimated values suggest that the filter breakthrough

for n-alkanes may occur during the sampling events. For PAHs, shallow mr (-0.61 to -0.64) and

high br (-6.14 to -7.12) values have been observed, which indicates gas-to-particle adsorption of

PAHs and non-exchangeable effects may occur during the sampling events. Engine loads affect

G/P partitioning of compounds; the regression lines are almost parallel for four load conditions

and move downward as the load increases. This result may relate with the increased temperature

and the decreased surface area of the TSP with loads.

4.5 References

1. Fernandes, M.B.; and Brooks, P., Characterization of carbonaceous combustion residues:

II. Nonpolar organic compounds, Chemosphere, 53(5): 447-458, 2003.

2. Sauvain, J.J.; Duc, T.V.; and Guillemin, M., Exposure to carcinogenic polycyclic
aromatic compounds and health risk assessment for diesel-exhaust exposed workers, Int.
Arch. Occup. Environ. Health, 76(6): 443-455, 2003.

3. Sjogren, M.; Li, H.; Rannug, U.; et al., A Multivariate statistical analysis of chemical
composition and physical characteristics of 10 diesel fuels, Fuel, 74(7): 983-989, 1995.

4. Adonis, M.; Martinez, V.; Riquelme, R.; et al., Susceptibility and exposure biomarkers in
people exposed to PAHs from diesel exhaust, Toxicol. Lett., 144(1): 3-15, 2003.

5. Gratz, L.D.; Bagley, S.T.; Leddy, D.G.; et al., Interlaboratory comparison of HPLC-
fluorescence detection and GC/MS: analysis of PAH compounds present in diesel
exhaust, J. Hazard. Mater., 74(1-2): 37-46, 2000.

6. Lee, S.L.; De Wind, M.; Desai, P.H.; et al., Aromatics reduction and cetane improvement
of diesel fuels, Fuel Reformulation, 5: 26-31, 1993.

92
7. Pal, R.; Juhasz, M.; and Stumpf, A., Detailed analysis of hydrocarbon groups in diesel
range petroleum fractions with on-line coupled supercritical fluid chromatography gas
chromatography mass spectrometry, J. Chromatogr. A, 819(1-2): 249-257, 1998.

8. Reilly, P.T.A.; Gieray, R.A.; Whitten, W.B.; et al., Real-time characterization of the
organic composition and size of individual diesel engine smoke particles, Environ. Sci.
Technol., 32(18): 2672-2679, 1998.

9. Shi, J.P.; Mark, D.; and Harrison, R.M., Characterization of particles from a current
technology heavy-duty diesel engine, Environ. Sci. Technol., 34(5): 748-755, 2000.

10. Kuusimaki, L.; Peltonen, Y.; Mutanen, P.; et al., Urinary hydroxy-metabolites of
naphthalene, phenanthrene and pyrene as markers of exposure to diesel exhaust, Int. Arch.
Occup. Environ. Health, 77(1): 23-30, 2004.

11. IPCS, Environmental Health Criteria 202. Selected nonheterocyclic Polycyclic aromatic
hydrocarbons. 1998, pp. 1-883, WHO: Geneva, ISBM. 9241572027.

12. Junge, C.E., Basic considerations about trace constituents in the atmosphere as related to
the fate of global pollutants, in Fate of pollutants in the Air and Water Environments.
Part I., Suffet, I.H., Ed. Wiley: New York. pp. 7-26, 1977.

13. Yamasaki, H.; Kuwata, K.; and Miyamoto, H., Effects of ambient temperature on aspects
of airborne polycyclic aromatic hydrocarbons, Environ. Sci. Technol., 16(4): 189-194,
1982.

14. Pankow, J.F., Review and comparative analysis of the theories on partitioning between
the gas and aerosol particulate phases in the atmosphere, Atmos. Environ., 21(11): 2275-
2283, 1987.

15. Seinfeld, J.H.; and Pandis, S.N., Organic atmospheric aerosols, in Atmospheric chemistry
and physics: from air pollution to climate change. Wiley: New York. pp. 700-765, 1998.

16. Pankow, J.F.; and Bidleman, T.F., Effects of temperature, TSP and percent non-
exchangeable material in determining the gas-particle partitioning of organic compounds,
Atmos. Environ., 25A(10): 2241-2249, 1991.

17. Pankow, J.F.; and Bidleman, T.F., Interdependence of the slopes and intercepts from log-
log correlations of measured gas-particle partitioning and vapor pressure. 1. Theory and
analysis of available data, Atmos. Environ., 26A(6): 1071-1080, 1992.

18. Pankow, J.F.; Storey, J.M.E.; and Yamasaki, H., Effects of relative humidity on
gas/particle partitioning of semi-volatile organic compounds to urban particulate matter,
Environ. Sci. Technol., 27(10): 2220-2226, 1993.

19. Pankow, J.F., An absorption model of gas/particle partitioning of organic compounds in

the atmosphere, Atmos. Environ., 28(2): 185-188, 1994.

93
20. Pankow, J.F., An absorption model of the gas/aerosol partitioning involved in the
formation of secondary organic aerosol, Atmos. Environ., 28(2): 189-193, 1994.

21. Jang, M.; Kamens, R.M.; Leach, K.B.; et al., A thermodynamic approach using group
contribution methods to model the partitioning of semi-volatile organic compounds on
atmospheric particulate matter, Environ. Sci. Technol., 31(10): 2805-2811, 1997.

22. Mader, B.T.; and Pankow, J.F., Study of the effects of particle-phase carbon on the
gas/particle partitioning of semivolatile organic compounds in the atmosphere using
controlled field experiments, Environ. Sci. Technol., 36(23): 5218-5228, 2002.

23. Storey, J.M.E.; Luo, W.; Isabelle, L.M.; et al., Gas/solid partitioning of semivolatile
organic compounds to model atmospheric solid surfaces as a function of relative
humidity. 1. Clean quartz, Environ. Sci. Technol., 29(9): 2420-2428, 1995.

24. Mader, B.T.; and Pankow, J.F., Gas/solid partitioning of semivolatile organic compounds
(SOCs) to air filters. 3. An analysis of gas adsorption artifacts in measurements of
atmospheric SOCs and organic carbon (OC) when using Teflon membrane filters and
quartz fiber filters, Environ. Sci. Technol., 35(17): 3422-3432, 2001.

25. Baek, S.O.; Goldstone, M.E.; Kirk, P.W.W.; et al., Phase distribution and particle-size
dependency of polycyclic aromatic hydrocarbons in the urban atmosphere, Chemosphere,
22(5-6): 503-520, 1991.

26. Volckens, J.; and Leith, D., Comparison of methods for measuring gas-particle
partitioning of semivolatile compounds, Atmos. Environ., 37(23): 3177-3188, 2003.

27. Pankow, J.F.; Isabelle, L.M.; Buchholz, D.A.; et al., Gas/particle partitioning of
polycyclic aromatic hydrocarbons and alkanes to environmental tobacco smoke, Environ.
Sci. Technol., 28(2): 363-365, 1994.

28. Jang, M.; and Kamens, R.M., A predictive model for adsorptive gas partitioning of SOCs
on fine atmospheric inorganic dust particles, Environ. Sci. Technol., 33(11): 1825-1831,
1999.

29. Liang, C.K.; Pankow, J.F.; Odum, J.R.; et al., Gas/particle partitioning of semivolatile
organic compounds to model inorganic, organic, and ambient smog aerosols, Environ. Sci.
Technol., 31(11): 3086-3092, 1997.

30. Tao, Y.; and McMurry, P.H., Vapor pressures and surface free energies of C14-C18
monocarboxylic acids and C5-dicarboxylic and C6-dicarboxylic acids, Environ. Sci.
Technol., 23(12): 1519-1523, 1989.

31. Mandalakis, M.; Tsapakis, M.; Tsoga, A.; et al., Gas-particle concentrations and
distribution of aliphatic hydrocarbons, PAHs, PCBs and PCDD/Fs in the atmosphere of
Athens (Greece), Atmos. Environ., 36(25): 4023-4035, 2002.

94
32. Goss, K.U.; and Schwarzenbach, R.P., Gas/solid and gas/liquid partitioning of organic
compounds: Critical evaluation of the interpretation of equilibrium constants, Environ.
Sci. Technol., 32(14): 2025-2032, 1998.

33. Naumova, Y.Y.; Offenberg, J.H.; Eisenreich, S.J.; et al., Gas/particle distribution of
polycyclic aromatic hydrocarbons in coupled outdoor/indoor atmospheres, Atmos.
Environ., 37(5): 703-719, 2003.

34. Foreman, W.T.; and Bidleman, T.F., Semivolatile organic compounds in the ambient air
of Denver, Colorado, Atmos. Environ., 24A(9): 2405-2416, 1990.

35. Ligocki, M.P.; and Pankow, J.F., Measurements of the gas/particle distributions of
atmospheric organic compounds, Environ. Sci. Technol., 23(1): 75-83, 1989.

36. Benner, B.A.; Gordon, G.E.; and Wise, S.A., Mobile sources of atmospheric polycyclic
aromatic hydrocarbons: A roadway tunnel study, Environ. Sci. Technol., 23(10): 1269-
1278, 1989.

37. Cotham, W.E.; and Bidleman, T.F., Polycyclic aromatic hydrocarbons and
polychlorinated biphenyls in air at an urban and a rural site near Lake Michigan, Environ.
Sci. Technol., 29(11): 2782-2789, 1995.

38. Simcik, M.F.; Franz, T.P.; Zhang, H.X.; et al., Gas-particle partitioning of PCBs and
PAHs in the Chicago urban and adjacent coastal atmosphere: States of equilibrium,
Environ. Sci. Technol., 32(2): 251-257, 1998.

39. Foreman, W.T.; and Bidleman, T.F., An experimental system for investigating vapor
particle partitioning of trace organic pollutants, Environ. Sci. Technol., 21(9): 869-875,
1987.

40. Corn, M.; Montgomery, T.L.; and Esmen, N.A., Suspended particulate matter: seasonal
variation in specific surface areas and densities, Environ. Sci. Technol., 5(2): 155-158,
1971.

41. Whitby, K.T.; Husar, R.B.; and Liu, B.Y.H., The aerosol size distribution of Los Angeles
smog, J. Colloid Interface Sci., 39(1): 177-204, 1972.

42. Yamasaki, H.; Kuwata, K.; and Kuge, Y., Determination of vapor pressure of polycyclic
aromatic hydrocarbons in the supercooled liquid phase and their adsorption on airborne
particulate matter, Nippon Kagaku Kaishi, (8): 1324-1329, 1984.

43. Chandramouli, B.; Jang, M.; and Kamens, R.M., Gas-particle partitioning of semi-
volatile organics on organic aerosols using a predictive activity coefficient model:
analysis of the effects of parameter choices on model performance, Atmos. Environ.,
37(6): 853-864, 2003.

44. Volckens, J.; and Leith, D., Effects of sampling bias on gas-particle partitioning of semi-
volatile compounds, Atmos. Environ., 37(24): 3385-3393, 2003.

95
 Chapter 5

FACTORS AFFECTING HYDROCARBON DISTRIBUTION IN

DIESEL EMISSIONS

5.1 Introduction

Emissions from diesel engines represent an important source of particles in ambient air.

Non-road diesel engines contribute to approximately 44% of total diesel particulate matter (DPM)

emissions and 12% of NOx emissions from mobile sources nationwide [1]. DPM contains

numerous organic components, among which polycyclic aromatic hydrocarbons (PAHs) and

their oxygenated and nitrated derivatives are highly hazardous [2-7]. Diesel particulate matter

has been regarded as a potential occupational carcinogen by several government agencies, such

as the U.S. Environmental Protection Agency (EPA) [8] and the National Institute for

Occupational Safety and Health (NIOSH) [9].

It is generally accepted that diesel particulate emissions are influenced by the engine

testing conditions: types of operation (steady-state or transient) [10-14], fuel quality (high or low

sulfur) [12, 15, 16], sampling conditions [17], and after-treatment devices [18]. Therefore, it is

important to understand the impact of different fuels, engines, and emission controls on the

chemical composition of particles for developing effective emission control strategies.

Effect of Engine Load: Engine load is an important factor that affects the combustion

process, diesel exhaust temperature, and the composition of diesel particulate matter. It has been

reported that engine load has effects on DPM mass concentration, size distribution, organic

carbon and elemental carbon (OC/EC) distribution, morphology, and microstructure. El-

96
Shoboksh [10] studied the exhaust particle size distribution and particulate emissions for a single

cylinder test diesel engine running at 0.25, 0.50, 0.75, and full load and at a constant speed of

600 rpm. The particulate emission factor (g particles/kg fuel) has been shown to increase linearly

with engine load. For particle size distribution, submicron (0-0.4 µm) and coarse (9-10 µm)

particles were found to increase, while the intermediate size showed insignificant variation with

engine load. Liu et al. [19] reported that the relative contributions of OC and EC vary

significantly with engine load. The fractions of EC over DPM increase with increasing load from

21% at 0 kW to 84% at 75 kW.

Effect of Fuel Sulfur Content: Sulfur in diesel fuels is known to contribute to DPM

emissions [16, 20-23]. It is reported that the amount of adsorbed sulfates affects the mass of

hydrocarbons in the particles [21], including PAHs, due to the formation of heavy hydrocarbons

in the condensed phase from the reaction of adsorbed sulfuric acid with organic compounds in

the exhaust [23]. Neeft et al. reported that 7-12% particulate reduction can be obtained with a

reduction of fuel sulfur content from 0.2% to 0.05% [22]. Baranescu et al. [20] evaluated the

effect of fuel sulfur on the brake specific particulate emissions on medium and heavy-duty trucks

using three diesel fuels with 0.05%, 0.19%, and 0.29% sulfur. The results indicated that for an

increase of 0.1% in fuel sulfur, particulates increased by about 0.025 g/bhp-h. The effect of fuel

sulfur content on OC and EC distribution has been reported by Liu et al. [19].

Due to the above mentioned effect of fuel sulfur content on DPM emissions, the U.S.

EPA has proposed more stringent regulations for sulfur content in diesel fuels. For highway

diesel fuel, fuel sulfur content was reduced from 500 ppmv to 15 ppmv in June 1, 2006 [24]. For

non-road applications, the current EPA regulation is 3400 ppmv, but it will be reduced to 500

ppmv in 2007 and further to 15 ppmv in 2010 [24].

97
 Effects of Sampling Conditions: Two major types of sampling methods have been used

for collecting PM emissions from combustion systems: source-level sampling and ambient-level

sampling. In general, source-level sampling is currently the accepted approach for the total

suspended particulate and PM10 measurements for stationary sources, while ambient level

sampling using dilution is the accepted approach for mobile source particulate emission

measurements. The particulate mass, size distribution, and chemical composition differ for the

two methods due to the aerosol processes such as coagulation, condensation, and nucleation

occurring during dilution. Many studies have shown that the fate of the condensable organics and

inorganics presented in diesel exhaust is significantly affected by atmospheric aging and dilution

of the exhaust stream [25, 26]. It has been reported that the dilution ratio has a significant effect

on the size distribution and the total number of particles emitted. Increasing the dilution ratio

increases the concentration of ultrafine particles [27, 28]. The results from England et al. have

indicated that traditional source testing methods (EPA 201A and EPA 202 in the paper) may

significantly overestimate particulate emissions, especially the ultrafine condensable particle

fraction [29].

Most of the above-mentioned studies focused on the impacts of each factor on PM mass

concentration and size distribution. However, the effects of engine load conditions, fuel sulfur

contents, and sampling methods on DPM chemical composition, especially individual organic

compound, have not been well studied. The purpose of this study is to investigate the

concentration and distribution of many types of organic compounds variated with engine load

conditions, fuel sulfur contents, and sampling conditions.

98
5.2 Experimental Methods

The experiments were performed under four engine load conditions (0 kW, 25 kW, 50

kW and 75 kW) to investigate load effect; two different sulfur-containing diesel fuels, which

were low sulfur diesel fuel (LSDF 400 ppmw S) and high sulfur diesel fuel (HSDF, 2200 ppmw

S), were used for studying fuel sulfur impact; and three sampling methods (EPA method 5,

natural dilution, and developed dilution method with a dilution ratio of 3.4) were applied for the

investigation of the influence of sampling conditions. The detailed experimental setup and

procedures have been described in Chapter 2.

5.3 Results and Discussion

5.3.1 Effect of Engine Load

5.3.1.1 DPM versus Load

DPM was collected using dilution sampling method (dilution ratio (DR) = 3.4) under four

engine load conditions (0 kW, 25 kW, 50 kW and 75 kW) when burning low sulfur diesel fuel

(400 ppmw S). The total DPM mass was determined by the gravimetric method, and the DPM

mass concentration was simply obtained by dividing the total mass with the total sample volume

that was corrected to standard conditions (1 atm, 293 °K). The DPM emission rate was

calculated by multiplying DPM mass concentration by stack gas flow rate.

The DPM mass concentration and emission rate for four load conditions were listed in

Table 5-1. When the load is increased from 0 kW to 75 kW, the DPM mass concentration

increases from 5.01 mg/m3 to 16.88 mg/m3, and the emission rate increases from 1.32 g/hr to

6.45 g/hr. The similar results have been reported by Burtscher et al., which indicated that

emission factors for diesel engine increased with increasing load [30]. This has to be expected

99
because higher load for diesel engines means lower air/fuel ratio and less complete combustion.

It is expected that higher DPM mass concentration should be obtained by dilution method than

EPA Method 5 since dilution method allows some organics to condense on particles. However,

compared with a study by Saiyasitpanich et al. [16], which used EPA Method 5 as the sampling

method, the DPM mass concentration measured in this study is close to their results. This study

also applied EPA Method 5 to measure DPM mass concentration and obtained lower

concentration than Saiyasitpanich’s results. The lower DPM mass concentration in this study

may be attributed to DPM mass from probe wash. In this study, due to the consideration of

correlating OC, EC and compositional results to DPM mass, DPM from probe wash was not

included for both dilution method and EPA Method 5. If this fraction is included, the DPM mass

concentration should be higher than the reported results. The DPM mass concentration as a

function of the applied engine load can be plotted as a straight-line (Figure 5-1), which is

consistent with Saiyasitpanich’s study.

Table 5-1 DPM emission rate from the non-road diesel generator at 0-75 kW (LSDF,
dilution sampling).
Engine Fuel Stack gas Air/fuel ratio DPM DPM
load consumption flow rate (mass/mass concentration emission
(kW) rate (L/hr) (m3/hr) basis) (mg/m3) rate (g/hr)
0 5.0 263 85 5.01 ± 0.58 1.32 ± 0.15
25 10.0 286 46 7.35 ± 1.13 2.10 ± 0.32
50 15.0 341 37 10.79 ± 1.57 3.68 ± 0.53
75 19.5 382 32 16.88 ± 1.59 6.45 ± 0.61

100
 20
18 y = 0.1562x + 4.1526

Mass concentration (mg/m3)

16 R2 = 0.9546
14
12
10
8
6
4
2
0
0 25 50 75
Load (kW)

Figure 5-1 Correlation between DPM mass concentration and diesel engine load (LSDF,
dilution sampling).

The OC and EC on quartz filters were measured using NIOSH method 5040 [31] and the

emission rates for four load conditions were summarized in Table 5-2. The results indicate that

both OC and EC emission rates increase with load. When engine load is increased from 0 kW to

75 kW, OC emission rate increases by twice, while EC emission rate increases by a factor of

about 13. The regression results for OC and EC emission rates also indicate that EC emission

rate increases more significantly with load than OC emission rate (Figure 5-2). Figure 5-2

suggests that the OC emission rate increases linearly with engine load, while EC emission rate

increases exponentially. The greater increase in EC emission rate is the result of the higher fuel

usage, lower air/fuel ratio, and higher temperature at higher loads, as EC is a product of

incomplete combustion.

101
Table 5-2 OC and EC emission rate (in mg/m3 and mg/hr) from the non-road diesel
generator at 0-75 kW (LSDF, dilution sampling).
Engine load OC emission rate EC emission rate OC emission rate EC emission rate
(kW) (mg/m3) (mg/m3) (g/hr) (g/hr)
0 3.64 ± 0.49 0.63 ± 0.09 0.96 ± 0.13 0.17 ± 0.02
25 4.43 ± 0.65 1.55 ± 0.26 1.27 ± 0.18 0.44 ± 0.07
50 5.38 ± 0.78 3.77 ± 0.54 1.84 ± 0.27 1.29 ± 0.18
75 6.42 ± 0.91 8.42 ± 0.88 2.45 ± 0.35 3.22 ± 0.34

10
9 OC
8
Concentraiton (mg/m3)

EC
7
y = 0.0373x + 3.5694
6
R2 = 0.9963
5
4
3
2 y = 0.6436e0.0347x
R2 = 0.9993
1
0
0 25 50 75
Load (kW)

Figure 5-2 The concentration of organic and elemental carbon versus diesel engine loads
(LSDF, dilution sampling).

Figure 5-3 represents the fractions of EC, organic compounds, and non-carbonaceous

materials in DPM. The amount of organic compounds was calculated from the amount of OC by

multiplying a factor of 1.2 (according to the study by Pierson and Brachaczek [32]) to account

for the hydrogen, oxygen and other non-carbon elements that are associated with OC mass but

were not measured by the thermal/optical reflectance technique directly. The carbonaceous

fraction (CF) of total DPM was estimated by

CF = (EC + 1.2 × OC)/DPM (5-1)

102
where the EC, OC, and DPM refer to the mass of EC, OC, and DPM collected, respectively.

100%

80% Non-carbonaceous
materials
Percentage in DPM

EC
60%

Organic compounds
40% (1.2*OC)

20%

0%
0 25 50 75
Load (kW)

Figure 5-3 Percentage of organic compounds, elemental carbon, and non-carbonaceous

materials in DPM at 0-75 kW (LSDF, dilution sampling).

As Figure 5-3 shows, the emissions were almost entirely carbonaceous materials, with CF

estimated to be 1.00, 0.93, 0.95, and 0.96 for 0 kW, 25 kW, 50 kW, and 75 kW, respectively.

The remaining non-carbonaceous materials may include sulfates, nitrates, metals, and other trace

elements and ions, although further analysis is required to confirm this hypothesis. A closer

examination suggests that the relative fractions of EC and OC change dramatically with change

in engine load. OC is the major species for low load conditions and its fraction decreases with

load, while EC is the major species for high load conditions and its fraction increases with load.

Therefore, it is necessary to account for varying DPM composition while estimating DPM health

impacts and emission source profile. The EC and OC emissions as a percent of total DPM during

this study are consistent with those reported by others as shown in Table 5-3.

103
Table 5-3 Comparison of OC and EC percentage in total DPM mass from this study with the
results from other studies.
Source Cycle Fuel EC OC
Shah, [33] Cold start/idle, HHDDTa ULSD (< 15 ppm S)b 17.3 ± 12.3 72.7 ± 14.8
CARBc creep 36.8 ± 13.6 60.1 ± 14.1
CARBc transient 67.8 ± 7.2 28.8 ± 9.1
CARBc cruise 61.0 ± 7.3 33.0 ± 15.0
Kelly, [34] 1500 rpm and idle Diesel (220 ppm S) 13.4 75.5
1500 rpm and 27% load 2.58 73.6
1700 rpm and 23%, 9% loadd 17.9 52.8
1700 rpm and 23% load 29.7 56.2
1700 rpm, 93%, 23% load 35.3 51.6
Rogers, [35] Running at fixed rpm for each Diesel (380 ppm S) 20 ± 8 60 ± 30
measurement
Shi, [2] 1600 rpm and 25% load Disel (427 ppm S) 24.58 ± 3.47 57.66 ± 8.28
1600 rpm and 50% load 35.88 ± 3.20 44.67 ± 8.40
1600 rpm and 100% load 33.87 ± 3.85 24.53 ± 11.87
2600 rpm and 25% load 35.16 ± 1.48 49.89 ± 8.85
2600 rpm and 50% load 47.30 ± 2.62 28.87 ± 2.80
2600 rpm and 100% load 51.52 ± 1.58 24.46 ± 10.16
Schauer, [36] MDDT, Hot start FTPe CA RFDf 30.8 ± 3.6 19.7 ± 1.6
Lowenthal, [37] Central Business District Diesel No. 2 43.3 ± 20.1 35.4 ± 17.8
This study 1800 rpm and idle Diesel No. 2 12.60 ± 1.80 72.55 ± 9.78
1800 rpm and 25% load (400 ppm S) 21.05 ± 3.54 60.20 ± 8.84
1800 rpm and 50% load 34.92 ± 5.00 49.86 ± 7.23
1800 rpm and 75% load 49.91 ± 5.21 38.05 ± 5.39
a
HHDDT: Heavy Heavy-Duty Diesel Truck (gross vehicle weight > 33,000 lb). b ULSD (< 15 ppm S):
ultra-low sulfur diesel with sulfur content less than 15 ppm. c CARB: California Air Resources Board. d
Decelerations during the middle of the text. e MDDT: Medium Duty Diesel Trucks; FTP: Federal Test
Procedure. f CA RFD: California reformulated diesel.

Material balances that describe the chemical composition of the organic mass detected by

GC/MS are shown in Table 5-4. The total DPM mass can be divided as soluble organic fraction

(SOF) and unsolvable fraction by extraction using DCM as solvent. The unsolvable fraction is

mainly elemental carbon (EC), often resulting as a byproduct of incomplete combustion, and

non-carbonaceous materials such as metals. The soluble organic fraction can be subdivided into

identified and unidentified organics by GC/MS, the latter of which mainly consists of the

104
unresolved complex mixture and those compounds that could not be separated as discrete peaks.

From Table 5-4, the SOF is reduced from 84.3% to 52.5% as load is increased from 0 kW to 75

kW, which is consistent with the variation in organic fractions from OC/EC results. The decrease

in identifiable fraction was observed as load increasing. Small fractions of identifiable organics

in DPM (1.63-8.39%) were also observed by other studies [36, 38, 39].

Table 5-4 Percentage of insoluble, soluble, unidentified and identified fraction in total DPM
mass at 0-75 kW (LSDF, dilution sampling).
0 kW 25 kW 50 kW 75 kW
Insoluble fraction 15.7 22.4 34.0 47.5
Soluble organic fraction 84.3 77.6 66.0 52.5

Unidentified fraction 75.9 72.2 64.1 50.8

Identified fraction 8.39 5.41 1.95 1.63

n-Alkanes 2.36 1.75 0.60 0.42

Branched alkanes 2.68 1.78 0.56 0.46
Saturated cycloalkanes 0.11 0.09 0.07 0.04
PAHs 0.07 0.04 0.03 0.02
Alkylated PAHs 2.21 1.17 0.20 0.15
Alkylbenzenes 0.02 0.03 0.23 0.27
n-Alkanoic acids 0.90 0.51 0.21 0.21
Aromatic acids 0.03 0.04 0.05 0.07

The identified compounds were classified into eight categories (Table 5-4), and the

relative percentage for each category in the identifiable mass for four load conditions was shown

in Figure 5-4. The normal alkanes, branched alkanes, and alkylated PAHs are the major species

in the identified compounds, and the remaining small percentage is attributed to saturated

cycloalkanes, PAHs, alkylbenzenes, and organic acids. As Table 5-4 shows, the percentage in

total DPM mass for each category (except alkylbenzenes and aromatic acids) decreases with

105
increasing load, which results from the greater increase in EC proportion. However, the variation

of relative percentage with load is different for each category. The relative percentage for

alkanes (56.60-66.54%) and n-alkanoic acids (9.55-13.55%) does not significantly vary with

load. In contrast, a great dependence on engine load was observed for aromatics, including PAHs,

alkylated PAHs, alkylbenzenes, and aromatic acid. As engine load is increased from 0 kW to 75

kW, the percent increases from 0.29% to 16.97% and from 0.37% to 4.29% for alkylbenzenes

and aromatic acids, respectively, while it decreases from 26.36% to 9.53% for alkylated PAHs.

The load dependence or independence for different species may be related to the sources of these

compounds. According to the study by Dobbins et al. [40], the load independent compounds are

most probably from petrogenic sources (raw materials such as diesel fuel and engine oil) and

load dependent compounds are from pyrogenic sources (from combustion process).

100%
Percentage in total identified compounds

90%
Aromatic acids
80%
n-Alkanoic acids
70%
Alkylbenzenes
60%
Alkylated PAHs
50%
PAHs
40%
Cycloalkanes
30%
iso-Alkanes
20%
10% n-Alkanes

0%
0 25 50 75
Load (kW)

Figure 5-4 Percentage of each identified category in total identified particle phase
compounds at 0-75 kW (LSDF, dilution sampling).

106
5.3.1.2 Organic Composition versus Load

Alkanes were identified and classified as normal alkanes, branched alkanes, and saturated

cycloalkanes. Alkanes account for a large percentage of the identified compounds and are mainly

sourced from diesel fuels. Normal alkanes with the carbon number ranging from C10 through C25

are the major components of diesel fuel, and the selected branched alkanes are known as the

isoprenoids, which are naturally present in crude oil and have the potential to be used as tracers

for diesel engine exhaust.

Normal Alkanes: Even though the percentage of n-alkanes is almost stable under

different load conditions, the absolute concentrations of n-alkanes (ng/m3 flue gas) in diesel

emissions vary significantly with load as shown in Figures 5-5(a) and (b).

For the particulate phase (Figure 5-5(a)), a bell-shaped distribution for n-alkanes from

C10 to C25 was observed with C17-C20 as the most abundant n-alkanes for all load conditions. This

bell-shaped distribution is related with the similar distribution of n-alkanes in diesel fuel [41],

since diesel fuel is the major source for n-alkanes in DPM. From Figure 5-5(a), it was also

observed that the concentrations of n-alkanes at 0 kW and 25 kW are much higher than at 50 kW

and 75 kW, even though the fuel consumption rate linearly increases with load. The flue gas

temperature at different load (113 ºC, 178 ºC, 259 ºC, and 310 ºC for 0 kW, 25 kW, 50 kW, and

75 kW, respectively) may be the answer. Flue gas temperature affects the gas/particle phase

partitioning of compounds, which is dominated by the adsorption or absorption [42-44]. At

higher loads, the higher flue gas temperature resulted in a higher sampling temperature due to the

limitation of dilution ratio and residence time in the dilution tunnel, and therefore large

percentage of unburned compounds cannot condense on particles and remain in the gas phase.

Similar results were reported by Zielinska et al. [45].

107
 18000
16000 0kW
14000 25kW
Concentration (ng/m3)

50kW
12000
75lW
10000
8000
6000
4000
2000
0
C10

C11

C12

C13

C14

C15

C16

C17

C18

C19

C20

C21

C22

C23

C24

C25
(a) Particle phase
350000
0 kW
300000
25 kW
Concentration (ng/m3)

250000 50 kW
75 kW
200000

150000

100000

50000

0
C10

C11

C12

C13

C14

C15

C16

C17

C18

C19

C20

C21

C22

C23

C24

C25

(b) Gas phase + particle phase

Figure 5-5 The concentration of n-alkanes from n-decane (C10) to n-pentacosane (C25) in gas
and particulate phase diesel emissions at 0-75 kW (LSDF, dilution sampling).

Figure 5-5(b) shows that the total n-alkane emissions (gas phase + particle phase) are

much higher than particle phase emissions, which suggests that most emissions are from the gas

phase. Compared with the particle phase, the most abundant n-alkanes in the total emissions have

shifted to lighter n-alkanes (the n-alkanes with smaller carbon number). This is to be expected,

since during the dilution process, the flue gas is cooled down and the lighter compounds are

108
condensed onto particle phase from gas phase. However, the change in flue gas temperature does

not significantly affect the total emissions for heavier compounds since they exist mainly in

particle phase.

From Figure 5-5(b), two types of trends can be observed regarding the concentrations of

n-alkanes versus load. For n-alkanes with a carbon number from 10 to 14, an increase followed

by a decrease is indicated, which represents a bell-shaped distribution. For n-alkanes with a

carbon number higher than 14, the concentration increases with load. The different effects of

load on n-alkanes with a different carbon number are probably due to the coexistent influence

from varied temperature and fuel consumption rate. As the load is increased from 0 kW to 75 kW,

the combustion temperature and fuel consumption rate increase. Higher combustion temperature

can result in an increase of the conversion fraction of n-alkanes [46, 47] according to the kinetics

reaction mechanism of combustion, which was built based on thermal decomposition, reaction

with O2, propagation, radical decomposition, and isomerization [46]. Therefore, as the load

increases, the unburned fuel will be reduced due to the higher conversion fraction and the

increased volatilization at higher temperatures, which leads to the reduction of the compound

concentration in DPM. On the other hand, the compound input was increased since more fuel

was introduced into the engine at high load conditions. This will cause an increase in the

compound concentration in DPM. The combination of temperature and fuel input effects led to

the two types of trends for the concentrations of n-alkanes versus load. When fuel input is the

dominant impact factor, the concentration increases with load. The bell-shaped distribution may

be caused by the shift of the dominant impact factor: from 0 kW to 25 kW, the increase in fuel

input is dominant and results in the increase in compound concentration, while from 25 kW to 75

kW, the temperature predominates over fuel input and a concentration decrease is the result.

109
 Branched Alkanes: Highly branched isoprenoid alkanes normally occur in crude oil [48].

It is known that branched alkanes are more resistant to biological degradation than straight-chain

alkanes, which makes them valuable for fingerprinting oil sources. The distributions of branched

isoprenoid alkanes from C14 to C20 in diesel particulate matter and total diesel emissions (particle

phase and gas phase emissions) are presented in Figures 5-6(a) and (b), respectively.

12000
0 kW
Concentration (ng/m3)

10000
25 kW
8000 50 kW
6000 75 kW

4000

2000

0
C14 C15 C18 C19 C20

(a) Particle phase

600000
0 kW
500000 25 kW
Concentration (ng/m3)

400000 50 kW
75 kW
300000

200000

100000

0
C14 C15 C18 C19 C20

(b) Total (gas phase + particle phase)

Figure 5-6 The concentration of branched alkanes in gas and particulate phase diesel
emissions at 0-75 kW (LSDF, dilution sampling).

110
 In diesel particulate matter, the selected isoprenoids C18-C20 are the major species which

have much higher concentrations than C14 and C15 (Figure 5-6(a)). The result indicates that the

most abundant species for branched alkanes has shifted to the alkane with a larger carbon

number (C20) compared with normal alkanes (C18). The possible reason is that branched alkanes

are more active due to their side chains compared with n-alkanes having the same carbon number.

This can also be observed from the lower boiling points and shorter GC/MS retention times of

branched alkanes compared with n-alkanes having the same carbon number.

Figure 5-6(b) represents the total branched alkane emissions (gas phase and particle

phase emissions) for engine loads from 0 kW to 75 kW. The concentration of branched alkanes

in total emissions significantly increases compared with particle phase emissions, especially for

C14 and C15 (increasing by 150-240 times). The most abundant branched alkane in total

emissions has shifted to lighter alkanes compared with the particle phase, which is similar to the

result observed for n-alkanes.

As the engine load is increased, the concentrations of branched alkanes C18-C20 increase

for both particle phase emissions (Figure 5-6(a)) and total emissions (Figure 5-6(b)). For

branched alkanes C14 and C15, the load-resulting variation in their concentrations shows less

dependence on the engine load. The difference between the engine load effect on branched

alkanes C18-C20 and the effect on C14 and C15 is possibly due to the relatively higher stabilities of

C18-C20 compared with the stabilities of C14 and C15. The load-resulting variation in the

concentration for less stable branched alkanes C14 and C15 depends on many factors, such as fuel

input, combustion consumption, and combustion generation, while for branched alkanes C18-C20,

the variation dependents primarily on fuel input.

111
 1200
0kW
Concentration (ng/m3) 1000 25kW
50kW
800
75kW
600

400

200

0
C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25

(a) Particle phase

30000
0 kW
25000
25 kW
Concentration (ng/m3)

20000 50 kW
75 kW
15000

10000

5000

0
C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25

(b) Total (gas phase + particle phase)

Figure 5-7 The concentration of cycloalkanes from C13 to C25 in gas and particulate phase
diesel emissions at 0-75 kW (LSDF, dilution sampling).

Saturated Cycloalkanes: Alkanes are the major components of the diesel fuel. In

general, alkanes in the diesel fuel are mostly straight-chain and branched alkanes, but mono-

cycloalkanes are also present as an important fraction [49]. Figures 5-7(a) and (b) represent the

distributions of saturated cycloalkanes from C13 to C25 in particle phase emissions and total

emissions. Figures 5-7(a) and (b) show that the distribution of saturated cycloalkanes is similar

112
to that of normal alkanes, even though the concentration of saturated cycloalkanes is much lower

than the concentration of n-alkanes. The most abundant cycloalkanes in total emissions are C13-

C15. This distribution is similar to the distribution of n-alkanes in total emissions, and is related

with the distribution of cycloalkanes in the diesel fuel.

PAHs and Alkylated PAHs: Many studies have been performed on the formation and

ambient behavior of PAHs and their substituted homologues due to their carcinogenic and

mutagenic properties [11, 40, 45, 50-54]. The research efforts have indicated that fuel

aromaticity, PAH accumulation in lubricant oil, lubricant oil combustion, and cold start behavior

influence PAH emissions [2, 11, 17, 30, 39, 45, 51, 53, 55]. Engine load is an important factor

which can influence combustion temperature, air/fuel ratio, and consequentially influences PAH

formation and behavior [11, 45]. Most of these studies are focused on on-road vehicle emissions.

However, limited studies have been conducted for emissions from non-road diesel generators,

which have different operating modes and emission characteristics from on-road vehicles.

In this study, more than 30 PAHs and alkylated PAHs have been identified and quantified

in particle phase and gas phase emissions from a non-road diesel generator under different load

conditions. The distributions of EPA 16 priority PAHs in particle phase emissions and total

emissions and their variation with load are illustrated in Figures 5-8(a) and (b).

113
 1000
900 0kW
800 25kW
Concentration (ng/m3)

700 50kW
600 75kW
500
400
300
200
100
0
Nap Acy Ace Flu Phe Ant Flt Pyr Baa Chy Bbf Bkf Bap Ind Dba Bgp

(a) Particle phase

30000
0 kW 300
Concentration (ng/m3)

25000
25 kW
Concentration (ng/m3)

20000 50 kW 200
75 kW
15000 100

10000
0
5000 Baa Chy Bbf Bkf Bap Ind Dba Bgp

0
Nap Acy Ace Flu Phe Ant Flt Pyr Baa Chy Bbf Bkf Bap Ind Dba Bgp

(b) Total (gas phase + particle phase)

Figure 5-8 The concentration of EPA 16 priority PAHs in gas and particulate phase diesel
emissions at 0-75 kW (LSDF, dilution sampling).

Figure 5-8(a) shows that phenanthrene, fluoranthene, and pyrene are the major PAHs in

diesel particulate matter. As the engine load is increased, the variation of PAH concentration is

not closely related to load conditions, especially for 2- and 3-ring PAHs. Some other studies

have reported that the relation between PAH concentration and the engine load differs for

different PAH species [11, 17, 45, 53]. Jones et al. studied the effects of air/fuel ratio (up to 3.94)

114
on the PAH content in diesel particles and found that higher air/fuel ratio, which is the case for

low load conditions, reduced high molecular weight PAHs [11]. However, since the air/fuel ratio

of the combustion in diesel engines is much higher than the studied range by Jones et al. and the

combustion condition is more complex, the results from this study are not consistent with Jones’

results. Waldenmaier et al. has shown that with load increasing the concentration decreased for

benz[a]anthracene and chrysene while it increased for fluoranthene [17]. A study by Williams et

al. [55] suggested that when the air/fuel ratio was in the range of 20-60, PAH emissions in DPM

were high at low load, decreased at mid-load (having the air/fuel ration of about 35), and

increased at higher load. The examined air/fuel ratio in this study is in the range of 32-85, and

the result is consistent with the work of Williams et al.

As Figure 5-8(b) shows, the major PAHs in total emissions are some lighter PAHs (2-

and 3-ring PAHs) instead of 3- and 4-ring PAHs. As the engine load is increased, the

concentration increases for all 16 PAHs. However, the mechanisms leading to the increase may

differ between 2- to 4-ring PAHs and 5- and 6-ring PAHs. Two main mechanisms were

recognized in explaining the occurrence of PAHs in diesel emissions [40, 50, 53]. PAHs in diesel

emissions can result directly from fuel PAH survivals through the combustion process, which is

referred to as survival pathway or petrogenic origin. On the other hand, PAHs in the exhaust

emissions can be formed during combustion processes by decomposition, dealkylation, and

recombination, which are called pyrosynthetic pathway or pyrogenic origin. Both mechanisms

are dependent on compounds as well as the combustion conditions, including engine load,

temperature, and air/fuel ratio. Most studies have indicated that pyrosynthetic pathway

represented the majority of PAH formations in the DPM emission [53, 56, 57]. However, Tancell

et al. [58] found that benzo[a]pyrene in the fuel, which had survived combustion, was the major

115
source of benzo[a]pyrene in the exhaust. Williams et al. [55] concluded that for engine

conditions used in their work, 2- to 4-ring PAHs in the exhaust particulate were primarily

unburned fuel components.

According to the study by Williams et al. [55], 2- to 4-ring PAHs are expected to be from

fuel PAH survivals. The increasing concentration with loads for 2- to 4-ring PAHs in this study

may have resulted from the increased amount of unburned fuel at high loads. This conclusion is

based on two factors: 2-to 4-ring PAHs are present in the diesel fuel, and at high loads, the low

air/fuel ratio causes the amount of unburned fuel increased, resulting in the increase of unburned

PAHs. However, 5- and 6-ring PAHs are believed to be pyrogenic species, which are of greater

carcinogenic potency [59] and contribute directly to soot particle formation [60], since they are

not present in diesel fuels. As the load is increased, the pyrosynthetic process is promoted and

consequently the concentration of pyrogenic PAHs increases. This conclusion has to be

examined by further studies.

Figures 5-9(a) and (b) present distributions of alkylated PAHs in particle phase emissions

and total emissions, and variations with loads in their concentrations. Emissions of alkylated

PAHs are relatively high compared with emissions of unsubstituted PAHs, which is consistent

with other studies [39]. For particle phase emissions, alkylated PAHs are reduced by 76% as the

load is increased from 0 kW to 75 kW. Similar results have been reported by Jensen and Hites

[61]. When gas phase emissions are included, major alkylated PAHs are shifted from heavier

species (trimethylnaphthalenes, methyl- and dimethylphenanthrenes) to some lighter species

(dimethylnaphthalenes).

116
 30000 300000

Concentration (ng/m3)
Concentration (ng/m3)

0kW 0 kW
25000 250000
25kW 25 kW
20000 200000
50kW 50 kW
15000 75kW 150000
75 kW
10000 100000
5000 50000
0 0

TMN
TMN

MN

DMN
MN

DMN

MPh

DMPh

MPh

DMPh
(a) Particle phase (b) Total (gas phase + particle phase)

Figure 5-9 The concentration of alkylated PAHs in gas and particulate phase diesel emission
at 0-75 kW (LSDF, dilution sampling).

Alkylbenzenes: Alkylbenzenes are common components in diesel fuels, as well as in

diesel emissions. As the engine load is increased, the fuel input is increased, which results in the

increase of alkylbenzene emissions (Figures 5-10(a) and (b)). A second source for alkylbenzene

emissions may be the generation from the decomposition, dealkylation, and recombination from

other compounds.

12000 350000
Concentration (ng/m3)
Concentration (ng/m3)

0kW 300000 0 kW
10000
25kW 25 kW
250000
8000 50 kW
50kW 200000 75 kW
6000 75kW 150000
4000 100000
2000 50000
0 0
C1-B

C2-B

C3-B

C4-B

C5-B

C6-B

C1-B

C2-B

C3-B

C4-B

C5-B

C6-B

(a) Particle phase (b) Total (gas phase + particle phase)

Figure 5-10 The concentration of alkylbenzenes in gas and particulate phase diesel emission at
0-75 kW (LSDF, dilution sampling).

117
 Alkanoic Acids: As known from Chapter 3, alkanoic acids in diesel emissions are from

engine oil used. Their distribution and variation with load are different from the fuel sourced

compounds (Figures 5-11(a) and (b)).

10000
9000 0kW
Concentration (ng/m3)

8000
25kW
7000
50kW
6000
5000 75kW
4000
3000
2000
1000
0
C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18

(a) Particle phase

100000
90000 0 kW
Concentration (ng/m3)

80000
25 kW
70000
50 kW
60000
50000 75 kW
40000
30000
20000
10000
0
C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18

(b) Total (gas phase + particle phase)

Figure 5-11 The concentration of alkanoic acids from C6H12O2 to C18H36O2 in diesel
emissions at 0-75 kW (LSDF, dilution sampling).

For alkanoic acids from C6H12O2 to C10H20O2, the concentrations increase with load in

both particle phase and total emissions, while decrease with load for some heavier acids

(C14H28O2 to C18H36O2). This is related to the combustion temperature, which can influence the

118
evaporation of engine oil components. At higher loads, the higher combustion temperature

greatly increases the evaporation of lighter compounds which have lower boiling points, while it

has less influence on the evaporation of heavy compounds. The reduction of heavy acid

emissions at higher loads is caused by the increase in the flow rate of stack gas at higher loads.

Compound Emission Rate Normalized to Fuel Usage: The emission rate of identified

compounds in total diesel emissions (gas phase + particle phase) at 0-75 kW was calculated

based on the data of compound concentrations in emissions, the stack gas flow rate, the fuel

usage rate, and compound concentrations in the diesel fuel according to Eq. (5-2) and (5-3). The

results were expressed as mg compound emission/kg of fuel burned and as mg compound

emission/g compound in diesel fuel using the method of Williams et al. [55].

C S × QS
ER1 = (5-2)
RDF × ρ DF

ER1
ER2 = (5-3)
C DF

Where:

ER1 = mg compound emission/kg of fuel combusted, mg/kg DF

CS = compound concentration in diesel emissions (gas + particulate phase), µg/m3

QS = stack gas flow rate, m3/hr

RDF = fuel usage rate, L/hr

ρDF = diesel fuel density, 800 g/L

ER2 = mg compound emission/g compound in diesel fuel, mg/g

CDF = compound concentration in diesel fuel, g/kg DF

119
 Figure 5-12 shows the emission rate for several categories of compounds, expressed in

terms of mass of compound in the exhaust/mass of compound in the fuel; also shown are the

total unburned hydrocarbons (UHC).

1.40 0.30
PAHs
Mg compound in emission
/ g compound from fuel

1.20

G total UHC emission

0.25

/ kg fuel combusted
Alkylated
1.00
PAHs 0.20
0.80 n-Alkanes
0.15
0.60 Branched
alkanes 0.10
0.40 Cycloalkanes
0.05
0.20 Alkylbenzenes
0.00 0.00
0 25 50 75 0 25 50 75

Engine load (kW) Engine load (kW)

(a) Each category of compounds (b) Total unburned hydrocarbons (UHC)

Figure 5-12 Normalized emission rates for identified compounds expressed in terms of mass
of compound in the exhaust/mass of compound from the fuel input (LSDF,
dilution sampling).

As Figure 5-12 suggests, total unburned hydrocarbons are 0.015-0.024 wt% of the fuel

input, and the UHC emission rate varies with engine loads, following the “U” type trend which

shows high emissions at low load, decreasing at 50 kW, and increasing at 75 kW. Similar

survivabilities and trends are observed for n-alkanes, branched alkanes, cycloalkanes, and

alkylated PAHs. However, unsubstituted PAHs and alkylbenzenes behave differently. A higher

proportion of 2- to 4-ring PAHs (0.02-1.0 wt%) survive the combustion process, and 5- and 6-

ring PAHs are formed during the combustion process. Total PAH emissions are 0.12-0.13 wt%

of PAHs from the diesel fuel. The fraction of PAHs in the emission to PAHs in the fuel increases

120
with loads. Compared with a study by Williams et al. [55], the normalized emission rate of

PAHs and UHC is relatively low. This may be caused by different sampling methods. In the

study by Williams et al., the exhaust gases were cooled by means of a water cooled condenser

and glass fibre filters were used. Figure 5-12 also shows that the emission rate for alkylbenzenes

increases with loads. Higher normalized emission rates for PAHs and alkylbenzenes suggest the

production of PAHs and alkylbenzenes during the combustion process, especially at high load

conditions.

5.3.2 Effect of Sampling Method

5.3.2.1 Dilution Effect on DPM, OC and EC

In this study, three sampling methods described in Chapter 2 have been used, i.e. EPA

Method 5, natural dilution, and dilution method with a dilution ratio of about 3.4. However, the

DPM mass, OC, and EC concentration expressed in terms of mg/m3 are not available for natural

dilution method, since the actual exhaust gas volume collected and the dilution ratio cannot be

measured for natural dilution. Therefore, the DPM mass, OC, and EC concentration are

compared between EPA Method 5 and dilution method (DR=3.4) only.

Figure 5-13 indicates that DPM mass concentration linearly increases with loads for both

methods, which is consistent with other studies [16, 19]. Compared with Method 5, higher DPM

mass concentration is obtained with the dilution method. This is a result of the continuous

adsorption, nucleation, coagulation, and condensation of the large quantities of condensable

organics and inorganics during dilution processes. England et al. [29] compared PM2.5 mass

measured by the dilution method and the in-stack method (EPA 201A / 202) for the gas-fired

boiler and the gas-fired process heater. They found that the amount of filterable particulate

121
collected using EPA method 201A was lower than the particulate collected with the dilution

method, which is similar to the result in this study. However, large amounts of condensable

particulate matter (CPM), which was believed to be mostly sulfates and chlorides [29], were

present in the exhaust and could be measured by EPA method 202. If this fraction was included,

the total filterable matter plus CPM mass was much higher than the total PM2.5 mass obtained

with the dilution method. Since particle condensation mechanisms are dependent on both vapor

concentration of compounds in the emission and the exhaust temperature, the dilution method

simulates conditions that more closely represent true atmospheric condensation conditions

compared with traditional source testing methods and is recommended as the standard reference

method for the measurement of automotive and stationary emissions.

20
DPM mass concentration (mg/m3)

Dilution (DR=3.4)
16
EPA Method 5

12

0
0 25 50 75
Load (kW)

Figure 5-13 Comparison of DPM mass concentration between EPA Method 5 and the dilution
method with a dilution ratio of ~3.4 (LSDF).

From Figure 5-13, it is also observed that the difference of DPM mass concentration

between EPA Method 5 and the dilution method is highest at 75 kW. At higher loads, the

exhaust gas temperature, as well as the CPM mass, is higher. Therefore, the DPM mass
122
concentration is significantly increased at high loads with the dilution method due to the

condensation of the large quantities of CPM during dilution processes.

Figure 5-14 illustrates the OC and EC concentration in diesel emissions collected using

EPA Method 5 and the dilution method under different load conditions. Multiple measurements

were made at each load condition and sufficient reproducibility is evidenced by the figure.

Figure 5-14 indicates that the EC concentrations are similar for EPA Method 5 and the dilution

method under each load condition. This has to be expected as EC is the carbonaceous carbon that

is produced by the incomplete combustion and is not affected by the sampling condition, i.e. the

sampling temperature, which is the major difference between the two methods.

12
OC-EPA Method 5
10
Concentration (mg/m3)

OC-Dilution (DR=3.4)
8 EC-EPA Method 5
EC-Dilution (DR=3.4)
6

0
0 25 50 75
Load (kW)

Figure 5-14 OC and EC concentration in DPM collected with EPA Method 5 and the dilution
method (LSDF).

The organic carbon collected with the dilution method is much higher than that collected

with EPA Method 5, especially at high loads (Figure 5-14). The fraction of organic compounds

in DPM, which is calculated from 1.2×OC/DPM, is from 87% (0 kW) to 46% (75 kW) for the

dilution method, while it is from 62% (0 kW) to 9% (75 kW) for EPA Method 5 (Figures 5-15 (a)

123
and (b)). The higher organic fraction in DPM collected with dilution method is caused by the

condensation of the condensable organic particulate matter and is consistent with the higher

DPM mass concentration. Liu et al. [19] reported that at the collection temperature of 25 °C ± 3

°C, the organic compounds account for 27-75% of the DPM mass, as opposed to 9-62% at the

collection temperature of 120 °C ± 14 °C. The result from this study is similar to the study by

Liu et al. [19].

100% 100%

80% 80% Non-carbonaceous

Fraction in DPM

Fraction in DPM

materials
60% 60%
EC
40% 40%

20% 20% Organic compounds

(1.2*OC)
0% 0%
0 25 50 75 0 25 50 75
Load (kW) Load (kW)

(a) EPA Method 5 (b) Dilution method (DR = 3.4)

Figure 5-15 Fractions of organic compounds, EC, and non-carbonaceous materials in DPM
collected with EPA Method 5 and the dilution method (LSDF).

5.3.2.2 Dilution Effect on the Organic Composition of DPM

As mentioned above, the actual exhaust gas volume and the dilution ratio for natural

dilution method are not measurable; therefore, the concentration of individual compounds in

particulate diesel emissions is expressed in terms of µg/g DPM for the three sampling methods

(Table 5-5). The concentration of identified compounds in DPM is compared among three

sampling methods at low (0 kW) and high (75 kW) load conditions.

124
Table 5-5 The concentration of compounds in DPM collected with EPA Method 5, the
natural dilution method, and the dilution method (DR=3.4) (LSDF).
Compoundsa Concentration (µg/g DPM)
0 kW 75 kW
Method 5 Natural Dilution Method 5 Natural Dilution
dilution method dilution method
n-Alkanes
n-C10 5.3 9.7 116 1.1 3.8 24
n-C11 16 22 164 1.5 5.6 47
n-C12 21 31 222 1.6 7.4 55
n-C13 26 57 337 3.2 12 79
n-C14 48 204 500 5.2 14 116
n-C15 79 565 1085 7.2 18 203
n-C16 159 1164 2206 11 30 308
n-C17 219 2645 3189 14 46 441
n-C18 375 1479 3851 20 67 516
n-C19 427 1257 3415 23 92 560
n-C20 384 742 2769 28 81 528
n-C21 276 138 1921 25 69 404
n-C22 162 10 1041 15 40 225
n-C23 107 7.4 574 9.9 33 146
n-C24 75 5.2 421 6.5 17 113
n-C25 64 3.6 324 5.7 14 128
Total n-alkanes 2444 8341 22137 176 552 3893

Branched alkanes
C14 7.6 13 154 3.0 6.5 61
C15 16 28 269 6.0 6.8 91
C18 104 135 1065 68 89 635
C19 123 147 1204 86 107 697
C20 152 185 1454 95 116 760
Other branched alkanes 923 7084 22894 326 288 2295
Total branched alkanes 1325 7593 27040 584 613 4538

Cycloalkanes
C13 2.8 11 21 0.57 0.90 8.7
C14 4.4 27 46 0.80 1.5 23
C15 9.1 29 101 1.6 2.6 50
C16 12 43 168 3.0 5.0 70
C17 21 39 209 4.0 7.5 86
C18 32 61 150 5.8 13 82
C19 24 12 138 7.9 15 63

125
Table 5-5 (Continued)
Compounds Concentration (µg/g DPM)
0 kW 75 kW
Method 5 Natural Dilution Method 5 Natural Dilution
dilution method dilution method
C20 17 9.8 78 7.0 9.6 42
C21 13 6.8 54 4.0 5.5 28
C22 11 4.3 45 3.0 4.1 17
C23 8.5 0.00 31 2.6 2.8 14
C24 7.2 0.00 26 1.9 2.0 10
C25 5.5 0.00 13 1.6 1.9 7.1
Total saturated cycloalkanes 168 242 1079 44 71 500

PAHs
Nap 1.5 4.1 35 0.60 3.2 7.4
Acy 1.2 2.0 17 0.20 1.6 4.3
Ace 1.3 3.0 33 0.20 3.7 11
Flu 4.4 7.1 32 0.40 5.3 11
Phe 23 77 224 4.9 30 70
Ant 1.8 4.1 17 0.40 3.3 5.8
Flt 16 25 78 5.0 20 42
Pyr 28 67 146 9.7 35 56
Baa 7.8 5.1 21 3.8 8.3 12
Chy 7.8 5.1 29 4.3 11 16
Bbf 3.0 3.0 5.9 1.4 3.0 7.5
Bkf 2.7 2.0 5.0 1.2 1.5 6.2
Bap 4.9 4.1 11.1 2.1 5.3 10
Ind 3.8 6.6 8.2 1.7 3.7 7.4
Dba 1.8 1.7 2.9 0.94 1.8 2.5
Bgp 2.8 3.0 6.4 1.8 2.9 6.1
Total PAHs 112 220 673 39 139 275

Alkylated PAHs
MN 33 116 340 6.0 19 51
DMN 125 589 2076 25 99 394
TMN 757 2914 5625 34 123 422
MPh 1167 3938 7135 27 76 274
DMPh 1032 3420 6268 30 95 362
Total alkylated PAHs 3114 10977 21445 122 412 1503

Alkylbenzenes
C1-B 6.8 11 41 4.3 27 198

126
Table 5-5 (Continued)
Compounds Concentration (µg/g DPM)
0 kW 75 kW
Method 5 Natural Dilution Method 5 Natural Dilution
dilution method dilution method
C2-B 13 21 72 8.9 50 358
C3-B 11 15 43 16 79 464
C4-B 12 16 47 35 148 826
C5-B 5.6 6.2 23 26 105 554
C6-B 3.0 3.2 11 13 55 278
Total alkylbenzenes 51 71 236 104 464 2678

n-Alkanoic acids
C6 0 0 10 1.5 23 45
C7 3.6 18 27 6.8 62 134
C8 8.1 40 57 10 94 168
C9 18 88 125 20 140 256
C10 45 289 398 26 198 318
C11 55 354 478 19 133 184
C12 89 591 786 28 183 217
C13 93 444 582 11 54 69
C14 425 1955 2503 22 125 170
C15 281 1238 1559 9.7 50 62
C16 474 1991 1899 10 62 84
C17 25 0.0 5.2 4.2 2.5 3.1
C18 11 4.1 10 1.7 5.2 5.3
Total of n-alkanoic acids 1528 7012 8441 170 1133 1716

Aromatic acids
Benzoic acid 58 226 305 35 430 677

Total identified compounds 8799 34682 81354 1274 3814 15780

a
The abbreviation in the table represents the corresponding compounds in Table 5-5. For n-alkanoic acids,
C6-C18 abbreviations represent the alkanoic acids from C6H12O2 to C18H36O2.

Table 5-5 indicates that the concentration of most compounds is highest for the dilution

method while lowest for EPA Method 5 under both low and high load conditions. As the exhaust

gas temperature is 113 ºC, 178 ºC, 259 ºC, and 310 ºC for 0 kW, 25 kW, 50 kW, and 75 kW,

respectively, it is expected that the dilution process has the greatest effect on compound
127
condensation at 75 kW. Therefore, the compound concentration and distribution are compared

for the three methods at 75 kW in the following discussion. At 75 kW, the concentration ratio of

individual compounds collected with the dilution method over those collected with EPA Method

5 is in the range of 2.6-53.3. While the concentration ratio of individual compounds collected

with the natural dilution method over those collected with EPA Method 5 is mostly less than

10.0 and is between 10.0 and 18.7 for several compounds. These results suggest that the

developed dilution method with a dilution ratio of 3.4 is more effective in converting organics

from the gas phase to the particle phase than the natural dilution method with a sampling

distance of about 1.0 m. This means the dilution ratio for the natural dilution with a sampling

distance of about 1.0 m is less than 3.4. The lower sampling temperature for the developed

dilution method than the natural dilution also confirms this prediction. Brown et al. [27] studied

the relationship between the dilution ratio and the sampling distance and found that the dilution

ratio increased with the sampling distance. At the sampling distance of 1.0 m, the dilution ratio

was about 10, which is different from the result in this study. This may be explained from the

fact that the dilution ratio not only depends on the sampling distance, but also depends on

diameters and directions of the exhaust outlet and the sampling inlet.

Kittelson et al. [62] has reported that as the engine exhaust is diluted and cooled, the

soluble organic fraction (SOF) may move from the gas phase to the particle phase by two paths:

the adsorption on existing particles or the nucleation to form new particles. The condensation can

be viewed as an extension of the nucleation and adsorption. Both adsorption and nucleation

processes are influenced by the saturation ratio for the various organic species, which is defined

as the partial pressure of the gaseous organic species divided by the vapor pressure of the same

organic species. The partial pressure of the volatile components in the hot exhaust decreases with

128
dilution. The vapor pressure of those components is a function of temperature, which also

decreases with dilution. Since the vapor pressure-versus-temperature relationship is nonlinear,

the saturation ratio reaches a maximum at some dilution level. Kittelson et al. [62] has reported

that the saturation ratio has the highest value for dilution ratios of about 5 to 50. Thus, the

strongest force driving gas to particle conversion occurs in the critical dilution ratio range (5 -

50:1). The dilution ratio for the developed dilution method is about 3.4 and falls in the range

where the saturation ratio increases with the dilution ratio. Therefore, the concentration of

compounds is the highest for the developed dilution method, lower for the natural dilution, and

lowest for EPA Method 5.

Distribution of Alkanes: The dilution process affects not only the absolute concentration

of compounds but also the compound distribution, since the dilution effect on the adsorption and

condensation differs for light and heavy compounds. Therefore, the percentage of individual n-

alkanes in the total emitted normal alkanes is calculated and the result is compared between the

three sampling methods (Figure 5-16(a)). Figures 5-16(b) and (c) present the distribution of

identified branched alkanes and cycloalkanes in DPM collected at 75 kW with the three methods.

As Figure 5-16(a) shows, the n-alkane distribution is similar for the three methods, which

indicates that the percentage of individual n-alkanes is highest for mid-molecular weight n-

alkanes, while lower for low- and high-molecular weights. However, compared with EPA

Method 5, the most abundant n-alkane has shifted to the left side of the figure (from C20 to C19),

which represents lower molecular weight compounds. The percentage of n-alkanes with lower

molecular weights (the carbon number less than 18) is highest for the dilution method, lowest for

EPA Method 5, and falls in between for the natural dilution method. The percentage of n-alkanes

with higher molecular weights (the carbon number greater than 20) is in reverse order for the

129
three methods. These results indicate that the formation of lower molecular weight compounds in

DPM is more significantly increased by the dilution process than the formation of higher

molecular weights.
Percentage of individual n-alkanes (%)

18
16 EPA Method 5
14 Natural dilution
12 Dilution (DR=3.4)
10
8
6
4
2
0
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
(a) n-Alkanes
Percentage of individual branched

Percentage of cycloalkanes (%)

40 25
35 EPA
20 Method 5
30
Natural
alkanes (%)

25 15 dilution
EPA
20 Dilution
Method 5 10
15 Natural (DR=3.4)
10 dilution
Dilution 5
5
(DR=3.4)
0 0
C14

C15

C18

C19

C20

C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25

(b) Identified branched alkanes (c) Cycloalkanes

Figure 5-16 The percentage of individual alkanes in total identified (a) n-alkanes, (b) branched
alkanes, and (c) cycloalkanes in DPM collected with EPA method 5, the natural
dilution method, and the dilution method with a dilution ratio of ~3.4 (LSDF, 75
kW).

130
 The result from the comparison of cycloalkane distribution among the three methods is

similar to the result for normal alkanes, while for the selected branched alkanes no significant

difference is observed among the three sampling methods.

Distribution of PAHs, Alkylated PAHs, and Alkylbenzenes: The distributions of

PAHs, alkylated PAHs, and alkylbenzenes for EPA Method 5, the natural dilution and the

developed dilution method are illustrated in Figures 5-17(a-c).

Percentage of individual PAHs (%)

30
(a)
25

20

15

10

0
Ind
Bbf

Bkf
Pyr
Nap

Ace

Flu

Phe

Baa

Bap

Dba

Bgp
Ant

Flt
Acy

Chy

35 40
(b) (c)
35
Percentage of individual
Percentage of individual

30
alkylated PAHs (%)

alkylbenzenes (%)

30
25
25
20
20
15
15
10 10
5 5
0 0
MN DMN TMN MPh DMPh C1-B C2-B C3-B C4-B C5-B C6-B

EPA Method 5 Natural dilution Dilution (DR=3.4)

Figure 5-17 The percentage of individual aromatics in total identified (a) PAHs, (b) alkylated
PAHs, and (c) alkylbenzenes in DPM collected with method 5, the natural
dilution, and the dilution method with a dilution ratio of ~3.4 (LSDF, 75 kW).

131
 Although Table 5-5 has indicated that the concentration of all these aromatics is highest

for the dilution method and lowest for EPA Method 5, the distribution of these aromatics is

similar and the most abundant species are the same for the three sampling methods. A detailed

examination of the relative abundance for each compound has suggested that among the

compounds which have similar structures and chemical properties, larger fractions of lower

molecular weights are collected with the dilution method while it is opposite for EPA Method 5.

For instance, among 16 EPA priority PAHs, 2- and 3-ring PAHs account for a higher percentage

of total PAHs for the dilution method (39.9%) than for EPA Method 5 (17.5%), while 5- and 6-

ring PAHs account for a lower percentage for the dilution method (14.4%) than for EPA Method

5 (23.3%).

Distribution of Alkanoic Acids: Figure 5-18 represents the distribution of alkanoic acids

in DPM at 75 kW for method 5, the natural dilution method, and the dilution method with a

dilution ratio of ~3.4. Similar to the above-mentioned result, which suggests that the abundance

shifts to the lower molecular weights for the dilution method compared with EPA Method 5, the

percentage of alkanoic acids with carbon numbers from 6 to 10 in total particulate alkanoic acids

is 37.57% for EPA Method 5, increases to 45.64% for the natural dilution, and further increases

to 53.67% for the developed dilution method. On the contrary, the proportion of alkanoic acids

with higher molecular weights (C12H24O2 to C18H36O2) increases for dilution method compared

with EPA Method 5.

132
 Percentage of individual alkanoic
20
EPA Method 5
16 Natural dilution
Dilution (DR=3.4)
acids (%)

12

0
C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18

Figure 5-18 The percentage of individual alkanoic acids in total alkanoic acids in DPM
collected with method 5, the natural dilution method, and the dilution method
with a dilution ratio of ~3.4 (LSDF, 75 kW).

5.3.3 Effect of Fuel Sulfur Content

5.3.3.1 Effect of Fuel Sulfur Content on DPM, OC, and EC

As illustrated in Figure 5-19, the concentration of DPM collected at high sulfur diesel

fuel (2200 ppmw S) is approximately 1.2 times the concentration of DPM collected at low sulfur

diesel fuel (400 ppmw S). Higher DPM concentration at higher fuel sulfur content have been

reported by other studies [16, 19]. Figure 5-19 also indicates that in the DPM at high sulfur

diesel fuel, the organic compounds are about 1.3 times as high, and non-carbonaceous materials

are 1.7-2.8 times as high, as those at low sulfur diesel fuel. The higher emission of organic

compounds at higher fuel sulfur content is consistent with the result reported by Wall and

Hoekman [63], and is due to the increased nucleation of organic compounds by the sulfuric acid

[64]. The higher non-carbonaceous material emission may result from the increase of sulfate

fractions at higher fuel sulfur content.

133
 3.0

2.5 Non-carbonaceous
Concentration ratio
materials
(HSDF / LSDF)
2.0 OC
1.5
DPM
1.0
EC
0.5

0.0
0 25 50 75
Load (kW)

Figure 5-19 Ratios of the concentration of DPM, OC, EC, and non-carbonaceous materials at
high sulfur diesel fuel over the concentration at low sulfur diesel fuel (dilution
sampling).

Table 5-6 lists the fractions of organic compounds, EC, and non-carbonaceous materials

in DPM at low and high sulfur diesel fuel. The contribution of organic compounds and non-

carbonaceous materials to DPM is higher for the high sulfur diesel fuel than for the low sulfur

diesel fuel due to the above-mentioned occurrence of larger amounts of sulfuric acid and sulfate.

Table 5-6 Fractions of organic compounds, EC, and non-carbonaceous materials in DPM at
low and high sulfur diesel fuel (dilution sampling).
Load Fraction of organic Fraction of EC Fraction of non-
(kW) compounds (1.2×OC) carbonaceous materials
LSDF HSDF LSDF HSDF LSDF HSDF
0 87.1% 90.5% 12.6% 9.0% 0.3% 0.5%
25 72.2% 76.2% 21.0% 15.1% 6.7% 8.8%
50 59.8% 63.1% 34.9% 27.8% 5.2% 9.1%
75 45.7% 49.4% 49.9% 40.1% 4.4% 10.5%

134
5.3.3.2 Effect of Fuel Sulfur Content on the Organic Composition

Compared with low sulfur diesel fuel, the concentration of organic compounds in DPM

collected at high sulfur diesel fuel is 1.2-3.0 times as high, which may result from the increased

nucleation. However, regarding the fraction of each category and the distribution of the

identified compounds, no significant difference is observed between low and high sulfur diesel

emissions. Consistent with this result, the distribution of these compounds in high sulfur diesel

fuel is similar to their distribution in low sulfur diesel fuel. Similar enhanced nucleation for these

organics by sulfuric acid is another hypothesis and should be further investigated.

5.4 Conclusions

The engine load significantly affects the concentration of DPM, OC, EC, and individual

compounds, and also affects the relative distribution of these compounds. As the load is

increased, the concentration of DPM, OC, and EC increases, as well as the concentration of

individual compounds in total diesel emissions (gas phase plus particle phase). However, the

fraction of OC in DPM decreases with loads. In the identified particle phase compounds, alkanes

account for more than 50% and this fraction does not significantly vary with loads; PAH and

alkylbenzene fractions increase with loads; while the fraction of alkylated PAHs decreases.

When normalized to the diesel fuel usage, the emission rate increases with loads for most

pyrogenic species (PAHs and alkylbenzenes), while it decreases with loads for most petrogenic

species (alkanes, alkylated PAHs).

The sampling method is an important factor that affects the concentration and distribution

of organics in DPM. The developed dilution method with a dilution ratio of about 3.4 is more

effective in converting organics from the gas phase to the particle phase than the natural dilution

135
with a sampling distance of about 1.0 m, and both methods are more effective than the source

sampling method (EPA Method 5). With the developed dilution method, DPM concentration

increases 40-80% compared with EPA Method 5, OC concentration is 2.7-11.6 times as high as

that for EPA Method 5, and EC concentration is similar for the two methods. The dilution

process not only affects the concentration of organic compounds, but also affects their

distribution. Compared with the source sampling method, the most abundant species has shifted

to those compounds with lower molecular weights for the dilution method, which suggests that

the dilution process favors the formation of light compounds more than the formation of heavy

compounds.

Higher fuel sulfur content results in higher emissions of DPM, OC, and non-

carbonaceous materials. As the fuel sulfur is increased from 400 ppmw to 2200 ppmw, the

emission of DPM, OC, and non-carbonaceous materials increases to 1.2, 1.3, and 1.7-2.8 times

respectively, while the EC emission is not significantly affected. Although the emission of

individual organic compounds in diesel exhaust at high sulfur diesel fuel increases to 1.2-3.0

times as that at low sulfur diesel fuel, their relative distribution does not vary with fuel sulfur

contents since the distribution of these compounds in low and high sulfur diesel fuels is similar.

136
5.5 References

1. United States Envrionmental Protection Agency, Nonroad Engines, Equipment, and

Vehicles; Reducing Air Pollution from Nonroad Engines, http://www.epa.gov/nonroad,
Accessed 2006.

2. Shi, J.P.; Mark, D.; and Harrison, R.M., Characterization of particles from a current
technology heavy-duty diesel engine, Environ. Sci. Technol., 34(5): 748-755, 2000.

3. Reilly, P.T.A.; Gieray, R.A.; Whitten, W.B.; et al., Real-time characterization of the
organic composition and size of individual diesel engine smoke particles, Environ. Sci.
Technol., 32(18): 2672-2679, 1998.

4. Pal, R.; Juhasz, M.; and Stumpf, A., Detailed analysis of hydrocarbon groups in diesel
range petroleum fractions with on-line coupled supercritical fluid chromatography gas
chromatography mass spectrometry, J. Chromatogr. A, 819(1-2): 249-257, 1998.

5. Lee, S.L.; De Wind, M.; Desai, P.H.; et al., Aromatics reduction and cetane improvement
of diesel fuels, Fuel Reformulation, 5: 26-31, 1993.

6. Adonis, M.; Martinez, V.; Riquelme, R.; et al., Susceptibility and exposure biomarkers in
people exposed to PAHs from diesel exhaust, Toxicol. Lett., 144(1): 3-15, 2003.

7. Gratz, L.D.; Bagley, S.T.; Leddy, D.G.; et al., Interlaboratory comparison of HPLC-
fluorescence detection and GC/MS: analysis of PAH compounds present in diesel
exhaust, J. Hazard. Mater., 74(1-2): 37-46, 2000.

8. EPA, Health Assessment Document for Diesel Emissions, Workshop Review EPA-
600/8-90/057A, EPA, Washington, DC, 1990.

9. US Department of Health and Human Services, Public Health Service, Centers for
Disease Control, National Institute for Occupational Safety and Health, DHHS (NIOSH),
Current Intelligence Bulletin No 50 - Carcinogenic Effects of Exposure to Diesel Exhaust,
Publication No. 88-116, Cincinnati, OH, 1988.

10. El-Shobokshy, M.S., The effect of diesel-engine load on particulate carbon emission,
Atmos. Environ., 18(11): 2305-2311, 1984.

11. Jones, C.C.; Chughtai, A.R.; Murugaverl, B.; et al., Effects of air/fuel combustion ratio
on the polycyclic aromatic hydrocarbon content of carbonaceous soots from selected
fuels, Carbon, 42(12-13): 2471-2484, 2004.

12. Liang, F.; Lu, M.; Birch, M.E.; et al. Effects of engine load and fuel sulfur on the
organosulfur content of diesel particulate matter. in proceedings of the 98th A&WMA
Annual Conference and Exhibition. June 21-24, 2005. Minneapolis, MN.

13. Lindgren, M.; and Hansson, P.A., Effects of transient conditions on exhaust emissions
from two non-road diesel engines, Biosyst. Eng., 87(1): 57-66, 2004.

137
14. Zhu, J.Y.; Lee, K.O.; Yozgatligil, A.; et al., Effects of engine operating conditions on
morphology, microstructure, and fractal geometry of light-duty diesel engine particulates,
Proc. Combust. Inst., 30: 2781-2789, 2005.

15. Corro, G., Sulfur impact on diesel emission control - A review, React. Kinet. Catal. Lett.,
75(1): 89-106, 2002.

16. Saiyasitpanich, P.; Lu, M.M.; Keener, T.C.; et al., The effect of diesel fuel sulfur content
on particulate matter emissions for a nonroad diesel generator, J. Air Waste Manage.
Assoc., 55(7): 993-998, 2005.

17. Waldenmaier, D.A.; Gratz, L.D.; Bagley, S.T.; et al., The influence of sampling
conditions on the repeatability of diesel particulate and vapor phase hydrocarbon and
PAH measurements, SAE Technical Paper Series 900642, 1990.

18. Burtscher, H., Physical characterization of particulate emissions from diesel engines: a
review, J. Aerosol. Sci., 36(7): 896-932, 2005.

19. Liu, Z.F.; Lu, M.M.; Birch, M.E.; et al., Variations of the particulate carbon distribution
from a nonroad diesel generator, Environ. Sci. Technol., 39(20): 7840-7844, 2005.

20. Baranescu, R.A., Influence of fuel sulfur on diesel particulate emissions, SAE Technical
Paper Series 881174, 1988.

21. Duran, A.; Carmona, M.; and Ballesteros, R., Competitive diesel engine emissions of
sulphur and nitrogen species, Chemosphere, 52(10): 1819-1823, 2003.

22. Neeft, J.P.A.; Makkee, M.; and Moulijn, J.A., Diesel particulate emission control, Fuel
Process. Technol., 47(1): 1-69, 1996.

23. Yu, F.Q.; Turco, R.P.; and Karcher, B., The possible role of organics in the formation and
evolution of ultrafine aircraft particles, J. Geophys. Res.-Atmos., 104(D4): 4079-4087,
1999.

24. US EPA, Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur
Control Requirements, EPA 420-F-00-057, December 2000.

25. Ahlvik, P.; Ntziachristos, L.; Keskinen, J.; et al., Real time measurements of diesel
particle size distribution with an electrical low pressure impactor, SAE Technical Paper
Series 980410, 1998.

26. Kittelson, D.B.; and Abdul-Khalek, I.S. Formation of nanoparticles during exhaust
dilution. in Proceedings of 2nd ETH Workshop on Nanoparticle Measurement. August
1998. Zurich, Switzerlan.

27. Brown, J.E.; Clayton, M.J.; Harris, D.B.; et al., Comparison of the particle size
distribution of heavy-duty diesel exhaust using a dilution tailpipe sampler and an in-

138
 plume sampler during on-road operation, J. Air Waste Manage. Assoc., 50(8): 1407-1416,
2000.

28. Lipsky, E.; Stanier, C.O.; Pandis, S.N.; et al., Effects of sampling conditions on the size
distribution of fine particulate matter emitted from a pilot-scale pulverized-coal
combustor, Energy Fuels, 16(2): 302-310, 2002.

29. England, G.C.; Zielinska, B.; Loos, K.; et al., Characterizing PM2.5 emission profiles for
stationary sources: comparison of traditional and dilution sampling techniques, Fuel
Process. Technol., 65: 177-188, 2000.

30. Burtscher, H.; Kunzel, S.; and Huglin, C., Characterization of particles in combustion
engine exhaust, J. Aerosol. Sci., 29(4): 389-396, 1998.

31. Birch, M.E.; and Cary, R.A., Elemental carbon-based method for monitoring
occupational exposures to particulate diesel exhaust, Aerosol Sci. Technol., 25(3): 221-
241, 1996.

32. Pierson, W.R.; and Brachaczek, W.W., Particulate matter associated with vehicles on the
road. 2, Aerosol Sci. Technol., 2(1): 1-40, 1983.

33. Shah, S.D.; Cocker, D.R.; Miller, J.W.; et al., Emission rates of particulate matter and
elemental and organic carbon from in-use diesel engines, Environ. Sci. Technol., 38(9):
2544-2550, 2004.

34. Kelly, K.E.; Wagner, D.A.; Lighty, J.S.; et al., Characterization of exhaust particles from
military vehicles fuelled with diesel, gasoline, and JP-8, J. Air Waste Manage. Assoc.,
53(3): 273-282, 2003.

35. Rogers, C.F.; Sagebiel, J.C.; Zielinska, B.; et al., Characterization of submicron exhaust
particles from engines operating without load on diesel and JP-8 fuels, Aerosol Sci.
Technol., 37(4): 355-368, 2003.

36. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; et al., Measurement of emissions from air
pollution sources. 2. C-1 through C-30 organic compounds from medium duty diesel
trucks, Environ. Sci. Technol., 33(10): 1578-1587, 1999.

37. Lowenthal, D.H.; Zielinska, B.; Chow, J.C.; et al., Characterization Of Heavy-Duty
Diesel Vehicle Emissions, Atmos. Environ., 28(4): 731-743, 1994.

38. McDonald, J.D.; Zielinska, B.; Sagebiel, J.C.; et al., Characterization of fine particle
material in ambient air and personal samples from an underground mine, Aerosol Sci.
Technol., 36(11): 1033-1044, 2002.

39. Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; et al., Sources of fine organic aerosol. 2.
Noncatalyst and catalyst-equipped automobiles and heavy-duty diesel trucks, Environ.
Sci. Technol., 27(4): 636-651, 1993.

139
40. Dobbins, R.A.; Fletcher, R.A.; Benner, B.A.; et al., Polycyclic aromatic hydrocarbons in
flames, in diesel fuels, an in diesel emissions, Combustion And Flame, 144(4): 773-781,
2006.

41. Liang, F.Y.; Lu, M.M.; Keener, T.C.; et al., The organic composition of diesel particulate
matter, diesel fuel and engine oil of a non-road diesel generator, J. Environ. Monit., 7(10):
983-988, 2005.

42. Jang, M.; Kamens, R.M.; Leach, K.B.; et al., A thermodynamic approach using group
contribution methods to model the partitioning of semivolatile organic compounds on
atmospheric particulate matter, Environ. Sci. Technol., 31(10): 2805-2811, 1997.

43. Pankow, J.F., An absorption model of gas/particle partitioning of organic compounds in

the atmosphere, Atmos. Environ., 28(2): 185-188, 1994.

44. Pankow, J.F., An absorption model of the gas/aerosol partitioning involved in the
formation of secondary organic aerosol, Atmos. Environ., 28(2): 189-193, 1994.

45. Zielinska, B.; Sagebiel, J.; Arnott, W.P.; et al., Phase and size distribution of polycyclic
aromatic hydrocarbons in diesel and gasoline vehicle emissions, Environ. Sci. Technol.,
38(9): 2557-2567, 2004.

46. Dagaut, P., On the kinetics of hydrocarbons oxidation from natural gas to kerosene and
diesel fuel, Phys. Chem. Chem. Phys., 4(11): 2079-2094, 2002.

47. Ristori, A.; Dagaut, P.; and Cathonnet, M., The oxidation of n-hexadecane: Experimental
and detailed kinetic modeling, Combust. Flame, 125(3): 1128-1137, 2001.

48. Atlas, R.M., Microbial degradation of petroleum hydrocarbons - An environmental

perspective, Microbiol. Rev., 45(1): 180-209, 1981.

49. Granata, S.; Faravelli, T.; and Ranzi, E., A wide range kinetic modeling study of the
pyrolysis and combustion of naphthenes, Combust. Flame, 132(3): 533-544, 2003.

50. Dobbins, R.A.; Fletcher, R.A.; B. A. Benner; et al. PAHs in flames, diesel fuels and
diesel emissions. in 30th International Symposium on Combustion, Work-In-Progress
Poster Session. July 25-30, 2004. Chicago, IL.

51. Collier, A.R.; Rhead, M.M.; Trier, C.J.; et al., Polycyclic aromatic compound profiles
from a light-duty direct- injection diesel-engine, Fuel, 74(3): 362-367, 1995.

52. Courtois, Y.; Molinier, B.; Pasquereau, M.; et al., Influence of the operating-conditions of
a diesel-engine on the mutagenicity of its emissions, Sci. Total Environ., 134(1-3): 61-70,
1993.

53. Rhead, M.M.; and Hardy, S.A., The sources of polycyclic aromatic compounds in diesel
engine emissions, Fuel, 82(4): 385-393, 2003.

140
54. Zielinska, B.; Sagebiel, J.; McDonald, J.D.; et al., Emission rates and comparative
chemical composition from selected in-use diesel and gasoline-fueled vehicles, J. Air
Waste Manage. Assoc., 54(9): 1138-1150, 2004.

55. Williams, P.T.; Bartle, K.D.; and Andrews, G.E., The relation between polycyclic
aromatic compounds in diesel fuels and exhaust particulates, Fuel, 65(8): 1150-1158,
1986.

56. Marr, L.C.; Kirchstetter, T.W.; Harley, R.A.; et al., Characterization of polycyclic
aromatic hydrocarbons in motor vehicle fuels and exhaust emissions, Environ. Sci.
Technol., 33(18): 3091-3099, 1999.

57. Rhead, M.M.; and Pemberton, R.D., Sources of naphthalene in diesel exhaust emissions,
Energy Fuels, 10(3): 837-843, 1996.

58. Tancell, P.J.; Rhead, M.M.; Trier, C.J.; et al., The sources of benzo[a] pyrene in diesel
exhaust emissions, Sci. Total Environ., 162(2-3): 179-186, 1995.

59. Larsen, J.C.; and Larsen, P.B., Chemical carcinogens, in Air Pollution and Health, Hester,
R.E. and Harrison, R.M., Editors. The Royal Society of Chemistry. pp. 33-56, 1998.

60. Dobbins, R.A.; Fletcher, R.A.; and Chang, H.C., The evolution of soot precursor particles
in a diffusion flame, Combust. Flame, 115(3): 285-298, 1998.

61. Jensen, T.E.; and Hites, R.A., Aromatic diesel emissions as a function of engine
conditions, Anal. Chem., 55(4): 594-599, 1983.

62. Kittelson, D.B.; Watts, W.F.; and Arnold, M., Aerosol Dynamics, Laboratory and On-
road Studies - Supplemental Report No. 2 to EPA Grant Review of Diesel Particulate
Matter Sampling Methods. 1998.

63. Wall, J.C.; and Hoekman, S.K., Fuel Composition Effects on Heavy-Duty Diesel
Particulate Emissions, SAE Technical Paper Series 841364, 1984.

64. Shi, J.P.; and Harrison, R.M., Investigation of ultrafine particle formation during diesel
exhaust dilution, Environ. Sci. Technol., 33(21): 3730-3736, 1999.

141
 Chapter 6

DETERMINATION OF POLYCYCLIC AROMATIC SULFUR

HETEROCYCLES IN DIESEL FUEL AND DIESEL

PARTICULATE MATTER BY GAS CHROMATOGRAPHY WITH

ATOMIC EMISSION DETECTION

6.1 Introduction

Organic sulfur compounds are the most abundant non-hydrocarbon constituents in

petroleum. The sulfur content of petroleum fuels contributes to the formation of sulfur dioxide

(SO2), which causes both acid deposition and poisoning of the catalytic converters in vehicles.

An increase in fuel sulfur also results in increased diesel particulate emissions. Fuel sulfur

reduction has been mandated by US Environmental Protection Agency (EPA) in the near future.

A large fraction of the organic sulfur in diesel fuels occurs in aromatic structures,

especially as alkylated homologues of polycyclic aromatic sulfur heterocycles (PASHs). It was

reported that benzothiophene (BT), dibenzothiophene (DBT) and their alkylated homologues are

the most abundant organosulfur compounds in diesel fuels [1]. Recently, increasing interest has

been focused on PASHs for several reasons. Specifically, some PASHs have been reported for

their potential mutagenic and carcinogenic properties [2, 3]; some PASHs, especially alkylated

DBTs, are difficult to remove in the desulfurization process for production of low sulfur fuels [4,

5]; and some PASHs can be potential indicators of the origin and maturity of crude oil [6].

142
 Given the above-mentioned roles of sulfur compounds, efforts have been placed on the

chemical characterization of PASHs in crude oil, diesel fuel, and other petroleum products [1, 7,

8]. However, there have been few studies on the organosulfur content of diesel particulate matter

(DPM), which is regarded as a carcinogen by the EPA [9] and the National Institute for

Occupational Safety and Health (NIOSH) [10]. PASHs in DPM can originate directly from the

diesel fuel or be generated by the combustion process. They are adsorbed on the DPM, which

includes a high number of ultrafine particles, and therefore penetrate into the deep lung. Thus,

the characterization of PASHs and other particle-borne organic compounds is important and

necessary for evaluating and controlling any adverse health effects associated with DPM

exposure. Studies on the PASH content of DPM are difficult because their concentrations are

low, there is a great variety of compounds present [11], and quantitative determination of

individual PASH isomers in the complex mixture is difficult.

The identification and quantification of individual PASH require selective and sensitive

methods of detection. Gas chromatography with the atomic emission detection (GC/AED) is a

powerful technique that offers high-resolution separation of components in a complex matrix and

highly selective spectrometric detection. Its application to the analysis of complex matrices,

including petroleum products, has been demonstrated [12-15]. The AED is an element-selective

and universal detection that provides relatively constant elemental response factors for different

compounds [12-14], which makes a compound independent calibration (CIC) possible. CIC is

highly useful because it minimizes the number of analytical standards required and permits

quantification of compounds for which no standards exist. CIC is particularly attractive when

dealing with highly toxic chemicals because relatively nontoxic surrogates can be used for

instrument calibration.

143
 In this paper, PASHs in two different sulfur-containing diesel fuels (low sulfur diesel fuel

[LSDF] and high sulfur diesel fuel [HSDF]) and the resultant DPM were identified and

quantified by GC with sulfur-selective atomic emission detection. The distribution of PASHs in

DPM was investigated under different fuel sulfur and engine load conditions. For the

convenience of description, low-molecular weight or lighter PASHs were defined as the PASHs

with one or two rings, and high-molecular weight or heavier PASHs were defined as three-, four-

or five-ring PASHs. The precision of a CIC was evaluated with a calibration solution containing

several PASHs having different structures and molecular weights.

6.2 Experimental method

6.2.1 Sampling and Extraction Method

Sampling and extraction methods were described in Chapter 2.

6.2.2 Analytical Method

A 6890 GC equipped with a G2350A AED (Agilent Technologies, Palo Alto, CA) was

used for quantification of the sulfur components examined in this study. Table 6-1 lists the

GC/AED operating conditions. The HP-5 MS, column used in this study, and equivalent columns

such as DB5-MS have been proven to be effective for PASH separation in several studies [2, 6,

13]. Both carbon (179 nm) and sulfur (181 nm) selective modes were monitored for all samples.

Sulfur determination was based on external and internal standards. For the external standard

calibration, the average response factor of a standard solution containing three sulfur compounds

at different concentrations was used. As an internal standard, a solution of t-butyl disulfide

144
(TBDS) in DCM was spiked into tested fuels and DPM extracts, which were determined to be

free of TBDS. Tentative compound identities are based on retention time.

Table 6-1 GC/AED operating parameters.

GC Conditions
Injection port temperature 280 °C
AED transfer line 310 °C
GC capillary column 30 m x 0.25-mm id x 0.25-µm film HP-5 MS
Oven temperature program 40 °C to 300 °C at 10 °C /min, hold 10 min
Column flow 1.3 mL/min
Carrier gas Helium
Injection 20:1 split for fuels; splitless for DPM extracts
Injection volume 1 µL
AED Parameters
Reagent gases Oxygen, 50 psi; hydrogen, 45 psi
Makeup flow 66 mL/min
Cavity temperature 300 °C
Carbon emission line 179 nm
Sulfur emission line 181 nm

6.2.3 Analytical Standards

Standard compounds listed in Table 6-2 were prepared in dichloromethane. Thiophene

and alkylated thiophenes were from Fisher Scientific, and other PASHs were obtained from a

laboratory in Germany (Dr. Jan T. Andersson, Institut für Anorganische und Analytische Chemie,

Correns-Str. 30, D-48149 Münster, Germany). Figure 6-1 shows the structures for the PASH

standards. As a check of method performance, the following National Institute of Standards and

Technology (NIST) standard reference material (SRM) also was used: SRM 2724b, sulfur in

diesel fuel oil (426.5 ppmw total sulfur).

145
Table 6-2 List of thiophene standards and related information.
No. Compound Abbreviation Formula MWa Bp or Mpb
1 Thiophene T C4H4S 84 84.4 °C (Bp)
2 2-Methylthiophene 2-MT C5H6S 98 113 °C (Bp)
3 3-Methylthiophene 3-MT C5H6S 98 114 °C (Bp738c)
4 2,3-Dimethylthiophene 2,3-DMT C6H8S 112 142-144 °C (Bp)
5 2,5-Dimethylthiophene 2,5-DMT C6H8S 112 134 °C (Bp740)
6 2-Ethylthiophene 2-ET C6H8S 112 132-134 °C (Bp)
7 2-Propylthiophene 2-PT C7H10S 126 157.5-159.5 °C (Bp)
8 Benzothiophene BT C8H6S 134 221-222 °C (Bp)
9 2-Methylbenzothiophene 2-MBT C9H8S 148 51-52 °C (Mp)
10 3-Methylbenzothiophene 3-MBT C9H8S 148 125-127 °C (Bp25)
11 5-Methylbenzothiophene 5-MBT C9H8S 148 66-67 °C (Bp0.6)
12 3,5-Dimethylbenzothiophene 3,5-DMBT C10H10S 162 125-126 °C (Bp14); 118-
118.5 °C (Bp9)
13 2,3,5-Trimethylbenzothiophene 2,3,5-TMBT C11H12S 162 145-146 °C (Bp15)
14 2,3,7-Trimethylbenzothiophene 2,3,7-TMBT C11H12S 162 143-144 °C (Bp15)
15 2,3,4,7- 2,3,4,7- C12H14S 176 167.5-168 °C (Bp14)
Tetramethylbenzothiophene TTMBT
16 Dibenzothiophene DBT C12H8S 184 332-333 °C (Bp)
17 4-Methyldibenzothiophene 4-MDBT C13H10S 198 >334 °C (Bp)
18 4,6-Dimethyldibenzothiophene 4,6-DMDBT C14H12S 212 ~340 °C (Bp)
19 2,4,6-Trimethyldibenzothiophene 2,4,6-TMDBT C15H14S 226 NAd
20 Phenanthro[4,5-bcd]thiophene Ph45T C14H8S 208 139-140 °C (Mp)
21 Benzo[b]naphtho[1,2-d]thiophene BN12T C16H10S 234 103-104 °C (Mp)
22 Benzo[b]naphtho[2,1-d]thiophene BN21T C16H10S 234 232-234 °C (Mp)
23 Benzo[b]naptho[2,3-d]thiophene BN23T C16H10S 234 158-159 °C (Mp)
24 Phenanthro[3,4-b]thiophene Ph34T C16H10S 234 82.5-83.5 °C (Mp)
25 Benzo[b]phenanthro[9,10- BPh9,10T C20H12S 284 129-130 °C (Mp)
d]thiophene
26 Diacenaphthothiophene DiAT C24H12S 332 285-286 °C (Mp)
a b c
Molecular weight. Bp: boiling point; Mp: melting point. Bp738 is the boiling point at the pressure of
738 mm Hg. d Not available.

146
Figure 6-1 Structures of PASH standard compounds.

6.3 Results and discussion

6.3.1 AED Response Factor

Atomic emission detection should permit a compound independent calibration with

surrogate reference compounds that contain the element of interest. Results with errors of a few

147
percent have been reported when the compounds to be quantified had elemental compositions,

retention times, and concentrations similar to the reference compounds [12-15].

In this study, three organosulfur compounds (Table 6-3) that are most abundant in diesel

fuel were chosen as external calibration standards. The sulfur response factor was calculated as

the average of the three compounds. The sulfur response factor was not influenced significantly

by differences in molecular weight or retention time; the deviation from the average value is

about 5%.

Table 6-3 Sulfur response factors for three sulfur compounds. Results are based on five
injections.
Compound MWa Formula Response factorb RSDc (%) (n-5)
DBT 184.255 C12H8S 54.06 1.8
4-MDBT 198.282 C13H10S 52.73 0.7
4,6-DMDBT 212.309 C14H12S 52.54 1.1
a
Molecular weight. bArea count per nanogram sulfur. cRelative standard deviation.

The total sulfur in diesel fuel was estimated by integrating the sulfur response from 7 min

(the beginning of the sulfur emission) to 25 min for the fuel and 32 min for DPM. The

integration endpoint was the point where the sulfur emission fell to baseline and remained stable.

With the average response factor obtained for the three sulfur compounds, the calculated sulfur

content of NIST SRM 2724b is 406 ppmw, which is a 4.9% deviation from the certified value of

426.5 ppmw. Based on the AED results, the sulfur contents of the diesel fuels (LSDF and HSDF)

used in this study are 433 ppmw and 2284 ppmw respectively. These results are in good

agreement with results (400 ppmw S for LSDF and 2200 ppmw S for HSDF) reported by another

laboratory (OKI Analytical, Cincinnati, Ohio, 45212, USA) using ASTM D1552-03 [16] for the

analysis. The relative percent difference (RPD) is 3.9% for LSDF and 1.9% for HSDF. Therefore,

148
the accuracy of this approach for total sulfur in the diesel fuels was verified through analysis of a

NIST standard and comparison with results obtained for the two fuels by an ASTM method.

6.3.2 PASH Speciation in Diesel Fuel

As was done in other studies, both carbon and sulfur selective modes were used to

identify the compounds. Carbon (179 nm) and sulfur (181 nm) AED chromatograms for LSDF

and HSDF are shown in Figure 6-2 and Figure 6-3, respectively. The total sulfur content and

concentration of individual sulfur compounds are listed in Table 6-4. Samples were screened for

twenty six different compounds, with 13 (T, 2-MT, 3-MT, 2,3-DMT, 2,5-DMT, 2-ET, 2-PT,

BN12T, BN21T, BN23T, Ph34T, BPh9,10T, and DiAT) not detected in diesel fuels. In Table 6-4,

3-MBT and 5-MBT were identified based on chemical standards. However, coelution may exist

for 3- and 4-MBT, and for 5- and 6-MBT, which was reported by Depauw et al. [17].

The chromatograms for both LSDF and HSDF indicate a large number of alkylated

PASH isomers exist in the two fuels. Although we don’t have standards for each of the peaks, an

isomer grouping method can be used to categorize these compounds into several groups

according to their retention times and other properties [6]. The following possible groups can be

assigned based on this comparison: C1-BTs (12.0-13.0 min), C2-BTs (13.5-14.5 min), C3-BTs

(14.6-15.8 min), C4-MBTs (16.8-17.2 min), C1-DBTs (18.8-19.6 min), C2-DBTs (19.6-20.9 min),

and C3-DBTs (21.0-22.0 min). These groups account for more than 50% of the total sulfur in

diesel fuel, with C2-DBTs as the largest fraction for LSDF and C3-BTs for HSDF (Figure 6-4).

Compared to previously reported chromatograms [6] for crude oil, the grouping pattern of diesel

fuel resembles the middle fraction of crude oil.

149
 Retention time

Figure 6-2 Carbon (179 nm) and sulfur (181 nm) AED chromatograms for low sulfur (433
ppmw) diesel fuel.

Retention time

Figure 6-3 Carbon (179 nm) and sulfur (181 nm) AED chromatograms for high sulfur (2284
ppmw) diesel fuel.

150
Table 6-4 PASHs in low and high sulfur diesel fuels (LSDF and HSDF).
Compound Formula Sulfur contenta (µg S/g DF) ± std (δ; n = 3)
LSDF HSDF
Total organic sulfurb S 433±9 2284±12
T to 2-PT (7 compounds)c ND d
ND
BT C8H6S 1.34±0.05 8.21±0.17
2-MBT C9H8S 0.11±0.02 7.68±0.14
3-MBTe C9H8S 3.21±0.09 8.39±0.12
5-MBTf C9H8S 2.86±0.13 15.29±0.45
3,5-DMBT C10H10S 3.85±0.16 18.71±0.42
2,3,5-TMBT C11H12S 8.28±0.24 63.56±2.46
2,3,7-TMBT C11H12S 10.72±0.19 75.91±2.78
2,3,4,7-TTMBT C12H14S 4.32±0.11 24.36±0.56
DBT C12H8S 15.23±0.29 83.99±1.92
4-MDBT C13H10S 21.22±0.52 66.78±1.64
4,6-DMDBT C14H12S 10.95±0.31 20.81±0.35
2,4,6-TMDBT C15H14S 4.34±0.15 6.63±0.24
Ph45Tg C14H8S 2.08±0.13 3.03±0.12
BN12T to DiAT (6 compounds)h C16H10S ND ND
a
For both fuels (LSDF and HSDF), three sets of samples were injected. b Total sulfur was calculated by
integrating the entire sulfur response as one peak. Integration began where the sulfur emission line
increased from baseline or at the beginning of the first peak (7 min for diesel fuel). The final integration
point was designated after the sulfur emission returned to baseline and remained stable (25 min for diesel
fuel). The standard temperature program indicated in Table 6-1 was used. c The 7 compounds include
compound No. 1 to No. 7 listed in Table 6-2. d ND (not detected) indicates result below the minimum
detectable level (MDL). MDLs are based on results for a test mixture (Agilent part 8500-5067). The
specification for sulfur 181 is 2 pg/sec. The MDL for our laboratory was typically about 0.5 pg/sec, or 1
pg S for a peak having a 2-sec width (at half height). e The coelution between 3- and 4-MBT may occur. f
The coelution between 5- and 6-MBT may occur. g This peak is assigned as Ph45T based on the retention
time of a standard compound. It is possible that this peak also contains a C3-DBT because C3-DBTs can
also elute in this region. Currently, we do not have a C3-DBT standard to examine this possibility. h The 6
compounds include compound No. 21 to No. 26 listed in Table 6-2.

151
 60

Percentage of total sulfur (%)

50
C3 -DBTs

40 C2 -DBTs
C1 -DBTs
30
C4 -BTs
20 C3 -BTs

C2 -BTs
10
C1 -BTs
0
LSDF HSDF

Figure 6-4 Possible alkylated PASH groups in low and high sulfur diesel fuels (LSDF and
HSDF, respectively).

For most diesel fuels, the sulfur content is 0.01-0.5% by weight. Currently, the EPA is

proposing stringent regulations for the sulfur content of diesel fuels. For highway diesel fuel,

sulfur content will be reduced from 500 ppmv to 15 ppmv on June 1, 2006 [18]. For non-

road applications, the current EPA regulation is 3400 ppmv, but it will be reduced to 500

ppmv in 2007 and further to 15 ppmv in 2010 [19]. As the total sulfur content of diesel fuel is

reduced from high to low, the distribution of organosulfur compounds also changes, as illustrated

in Figure 6-5(a) and (b). In Figure 6-5(b), the concentration was normalized by total sulfur

content in diesel fuel, which is different from Table 6-4.

152
 90
80
Sulfur (µg S/g DF)

70 (a) LSDF (TS=433 ppm)

60 HSDF (TS=2284 ppm)
50
40
30
20
10
0

60
Sulfur (mg S/g TS in DF)

50 (b) LSDF (TS=433 ppm)

40 HSDF (TS=2284 ppm)

30

20

10

0
2,3,4,7-TTMBT
2,3,5-TMBT

2,3,7-TMBT

2,4,6-TMDBT
BT

2-MBT

3-MBT
5-MBT

3,5-DMBT

DBT

4-MDBT
4,6-DMDBT

Ph45T

BN12T

BN21T

BN23T
Ph34T

BPh9,10T

DiAT
Figure 6-5 PASHs in low and high sulfur diesel fuels (LSDF and HSDF, respectively)
expressed in (a) µg S/g DF, and (b) mg S/g TS in DF.

Figure 6-5(a) shows that the concentrations of all the sulfur compounds are higher in

HSDF than in LSDF. The most abundant PASH species in both the low and high sulfur diesel

fuels are DBT, 4-MDBT, 2,3,5-TMBT, and 2,3,7-TMBT. In LSDF, 4,6-DMDBT is also one of

the major PASHs, whereas its relative abundance in HSDF is low. Figure 6-5(b), in which the

concentrations of compounds are expressed in terms of mg S/g total sulfur (TS) in DF, shows

that the relative abundance of the compounds is different for the two fuels. In the HSDF, the

relative abundance of the lighter compounds is higher, while in LSDF, the abundance of heavier

compounds is higher. Based on the chromatograms (Figure 6-2 and Figure 6-3), it also can be

153
seen that there are some lighter sulfur compounds (presuming compounds with shorter retention

time are relatively lighter compounds) that are major components of HSDF, which is not the case

for LSDF. This is consistent with the results of desulfurization in that lighter compounds are

more easily removed from the fuel [5, 20]. It is also reasonable to expect that with the future fuel

sulfur reduction, the organosulfur species remaining in the fuel will be more inclined toward the

heavier (more than 3 rings) side.

6.3.3 PASHs in Diesel Emissions

In diesel emissions, sulfur compounds exist in both the gaseous and particulate fractions.

Sulfur dioxide (SO2) is a priority air pollutant regulated under the National Ambient Air Quality

Standards. Sulfate (SO42-) is thought to play an important role in fine particle formation and the

nucleation of organic compounds, including polycyclic aromatic hydrocarbons (PAHs),

reportedly through formation of heavy hydrocarbons from reaction between adsorbed sulfuric

acid and organic compounds in the exhaust [21-23]. Gas-phase sulfur compounds (SO2, SO3) and

particulate sulfate (SO42-) in diesel exhaust have been investigated, but there have been limited

studies on the organosulfur species in DPM and the influence of fuel sulfur on this fraction.

Although this fraction is of small percentage of sulfur emissions from diesel engines, there is a

necessity to study these organosulfur species, their concentrations and distribution, since these

organosulfur species are hazardous, some of which are carcinogenic, sorbed on DPM and

transferred into the lung with DPM.

Both gas and particulate diesel emissions were collected and the extracts were injected

into the GC/AED system. Samples were screened for carbon 179 and sulfur 181 and the sulfur

181 AED chromatograms for the extracts are shown in Figures 6-6(a-c). A total of twenty three

154
sulfur compounds were identified and quantified in either gas phase or particulate phase extract.

The concentrations of identified sulfur compounds are listed in Table 6-5, the total organic sulfur

content (TOS) for each sample was calculated and summarized in Table 6-5 as well.

Retention time

Figure 6-6 Sulfur (181 nm) AED chromatograms of gas and particulate phase diesel
emissions generated at 0 kW with low sulfur (433 ppmw) diesel fuel.
(a) Collected with PUF, (b) collected with XAD, and (c) collected on filter.

155
Table 6-5 PASHs in gas and particle phase diesel emissions (0 kW through 75 kW) when
burning low and high sulfur diesel fuels.
Compound Formula Sulfur content (µg S/g DPM) ± std (δ; n = 3)a
Gas phase emission at LSDF L0-Gb L25-G L50-G L75-G
Total organic sulfurc S 775±35 786±27 450±26 400±30
d
T C4H4S ND ND ND ND
2-MT C5H6S 5.56±0.62 10.15±0.89 3.42±0.33 5.41±0.39
3-MT C5H6S 9.90±0.78 19.01±1.25 6.10±0.52 10.37±0.89
2,3-DMT C6H8S 3.86±0.45 5.32±0.63 2.09±0.29 4.03±0.44
2,5-DMT C6H8S 4.17±0.54 6.68±0.56 2.45±0.28 4.59±0.38
2-ET C6H8S 4.03±0.49 6.82±0.61 2.72±0.32 4.13±0.26
2-PT C7H10S 1.73±0.28 4.35±0.52 1.54±0.27 2.76±0.37
BT C8H6S 9.56±0.82 7.76±0.67 3.96±0.36 4.18±0.43
2-MBT C9H8S 9.52±0.71 5.11±0.43 3.67±0.41 4.41±0.51
3-MBTe C9H8S 12.20±0.88 9.22±0.79 5.88±0.51 5.87±0.55
f
5-MBT C9H8S 4.90±0.52 2.24±0.33 2.06±0.28 2.14±0.30
3,5-DMBT C10H10S 9.11±0.63 9.34±0.86 6.49±0.56 5.25±0.49
2,3,5-TMBT C11H12S 23.31±1.57 17.87±1.49 12.59±0.97 10.92±0.93
2,3,7-TMBT C11H12S 31.36±2.05 23.86±1.84 16.96±1.27 14.92±1.22
2,3,4,7-TTMBT C12H14S 2.28±0.33 1.85±0.26 1.38±0.18 0.85±0.17
DBT C12H8S 22.82±1.55 13.00±1.06 10.76±0.93 7.68±0.68
4-MDBT C13H10S 9.86±0.81 5.29±0.55 5.91±0.56 5.31±0.47
4,6-DMDBT C14H12S 3.59±0.42 ND ND ND
Compound No. 19 to 26 listed in Table 6-2 ND ND ND ND

Particle phase emission at LSDF L0-P L25-P L50-P L75-P

Total organic sulfur S 679±32 734±27 819±25 631±35
Compound No. 1 to No. 14 listed in Table 6-2 ND ND ND ND
2,3,4,7-TTMBT C12H14S 1.48±0.15 1.96±0.28 1.53±0.25 1.63±0.22
DBT C12H8S 2.84±0.23 2.68±0.32 1.77±1.89 1.51±1.75
4-MDBT C13H10S 9.07±0.68 7.33±0.88 4.71±0.55 3.55±0.41
4,6-DMDBT C14H12S 10.30±1.08 12.40±1.03 11.57±1.21 11.34±1.09
2,4,6-TMDBT C15H14S 13.37±1.22 34.50±2.94 27.69±2.49 22.38±2.23
Ph45Tg C14H8S 6.42±0.75 22.39±2.19 14.86±1.78 11.68±0.94
BN12T C16H10S 0.99±0.37 2.37±0.36 4.09±0.83 5.58±0.63
BN21T C16H10S 1.13±0.25 2.87±0.43 3.83±0.75 5.11±0.82
BN23T C16H10S 0.95±0.22 1.74±0.31 2.23±0.41 4.69±0.76
Ph34T C16H10S ND ND ND 2.25±0.50
BPh9,10T C20H12S ND ND ND ND
DiAT C24H12S ND ND ND ND

156
Table 6-5 (Continued)
Compound Formula Sulfur content (µg S/g DPM) ± std (δ; n = 3)
Particle phase emission at HSDF H0-P H25-P H50-P H75-P
Total organic sulfur S 1585 1599 1349 923
Compound No. 1 to No. 12 listed in Table 6-2 ND ND ND ND
2,3,5-TMBT C11H12S 2.81 1.93 0.98 1.85
2,3,7-TMBT C11H12S 3.44 3.40 2.00 1.25
2,3,4,7-TTMBT C12H14S 14.28 12.67 8.52 2.49
DBT C12H8S 30.29 25.18 11.21 4.41
4-MDBT C13H10S 32.39 26.86 19.81 11.39
4,6-DMDBT C14H12S 22.91 29.28 15.31 8.25
2,4,6-TMDBT C15H14S 28.08 42.40 27.20 18.56
Ph45T C14H8S 19.05 34.78 17.80 9.55
BN12T C16H10S 1.89 3.22 5.54 7.68
BN21T C16H10S 1.69 3.06 4.61 6.86
BN23T C16H10S 1.50 2.98 4.33 6.49
Ph34T C16H10S ND 0.47 2.48 0.48
BPh9,10T C20H12S ND ND ND 0.81
DiAT C24H12S ND ND ND ND

Total organic sulfur in gas phase at HSDF 1768 1654 1282 1077
a
For DPM at LSDF, three sets of samples were collected and injected, but for DPM at HSDF only one set
of samples was collected. b L or H indicates low or high sulfur diesel fuel was used; 0 through 75 is the
engine load (kW); G or P indicates gas phase or particle phase. c Total sulfur was calculated by
integrating the entire sulfur response as one peak. Integration began where the sulfur emission line
increased from baseline or at the beginning of the first peak (7 min for DPM). The final integration point
was designated after the sulfur emission returned to baseline and remained stable (32 min for DPM). The
standard temperature program indicated in Table 6-1 was used. Notes d-g are the same as those for Table
6-4.

Compared with the chromatograms of diesel fuels, the chromatograms of diesel emission

extracts (Figure 6-6(a-c)) show more sulfur peaks which include the peaks with the retention

times shorter than 12 min in PUF and XAD extracts and the peaks with retention times longer

than 22 min in DPM extracts. The large hump in Fgiure 6-6(c) represents a complex sulfur

compound matrix in DPM. From Table 6-5, C3-BT, DBT, C1-DBT and C2-DBT, which are

abundant in diesel fuels, are still predominant in diesel emissions. However, some lighter

157
compounds in gas phase, such as C1-T, BT, and C1- and C2-BT, and some heavier compounds in

particulate phase, such as C3-DBT and PH45T, are also important and account for large

percentages in diesel emissions. The occurrence of light compounds in diesel emissions indicates

the possibility of decomposition of the sulfur compounds in diesel fuels during the combustion

process. On the other hand, some heavier PASHs which do not exist in diesel fuels, such as

BN12T, BN21T, BN23T, Ph34T and BPh9,10T, are present in DPM with high fractions,

especially at high load conditions. These PAHs appear to be generated through the pyrogenic

pathway during the combustion process. These larger PASHs may increase the toxicity of DPM,

as some three-, and especially four- and five-ring species, are known to be mutagenic [24, 25].

6.3.3.1 PASH Distribution between Gas and Particle Phase

Figure 6-7 presented the organosulfur compounds in both gas and particulate phase diesel

emissions at 0 kW when burning low sulfur diesel fuel. Although the organosulfur species are

different in gas and particulate phase diesel emissions, their concentrations are comparable

which are in the range of 0-35 µg S/g DPM, the total organic sulfur contents in gas and

particulate phases are comparable as well (400-820 µg S/g DPM). Thiophene, benzothiophene,

and their alkylated homologue are primarily present in gas phase, while dibenzothiophene, its

alkylated homologue, and other high-molecular-weight sulfur compounds (4- and 5-ring PASHs)

are present predominantly in DPM. 7-ring PASH diacenaphthothiophene was not detected in

diesel emissions, however, heavier PASHs, which might include 6- or 7-ring PASHs, could exist

in DPM as shown by Figure 6-6(c). Due to the lack of standard reference compounds, those

heavier PASHs were not measured in this study. Similar gas/particle distributions of

organosulfur compounds were obtained for other engine load conditions.

158
 35

30
L0-Gas
Sulfur (µg S/g DPM)

25 L0-Particle
20
15

10
5
0
2-MT
3-MT

2,3-DMT

2,5-DMT

2-MBT
3-MBT

5-MBT
3,5-DMBT

2,3,5-TMBT
2,3,7-TMBT

2,3,4,7-TTMBT

4-MDBT
4,6-DMDBT
2,4,6-TMDBT
2-ET

2-PT
BT

DBT

Ph45T
BN12T
BN21T
BN23T
Figure 6-7 Organosulfur compounds in gas phase and particulate phase diesel emissions at 0
kW and low sulfur diesel fuel.

The results have shown that the organosulfur compounds occurring in gas phase diesel

emission are mostly light ones which are less toxic compared with those heavier one present in

DPM, therefore, when discussing the concentrations and distributions of PASHs in the following

context, only the particulate phase PASHs will be referred to.

6.3.3.2 Total Organic Sulfur in DPM

Results for the total organic sulfur (TOS) in particulate phase and gas + particulate phase

diesel emissions for LSDF and HSDF under different engine loads are shown in Figure 6-8. As

evidenced by these figures, fuel sulfur content has a significant effect on the concentrations of

total organic sulfur in diesel emissions. When fuel sulfur is increased from 433 ppmw to 2284

ppmw, the total organic sulfur in DPM is increased by a factor of 1.34 at 0 kW. At high load

conditions, the total organic sulfur content in DPM is less dramatically affected by fuel sulfur,

increased by a factor of 0.46 at 75 kW. As reported, when fuel sulfur was reduced from 0.2% to

159
0.05%, a 7-12% reduction in particulate mass can be obtained [26]. Relative to DPM mass, the

amount of total organic sulfur content is much more affected by the change in fuel sulfur content.

Figure 6-8 also indicates that engine loads have some effects on total organic sulfur content in

both gas and particulate phase diesel emissions, especially when burning HSDF. As the engine

load increasing, the total organic sulfur content in gas + particulate phase diesel emissions

decreased.

3600
3300
3000 at HSDF (G+P)
Sulfur (µg S/g DPM)

2700
2400
2100 at LSDF (G+P)
1800
1500
1200 at HSDF (P)
900
600
300 at LSDF (P)
0
0 25 50 75
Engine Loads (kW)

Figure 6-8 Total organic sulfur in particulate phase and gas + particulate phase diesel
emissions for LSDF and HSDF at various load conditions.

When examining the effects of fuel sulfur and engine load, it is convenient and

reasonable to compare the conversion rate of sulfur. In this study, the conversion rate is defined

as the mass (µg) of organic sulfur in diesel emissions (or DPM) per gram of sulfur input from the

fuel. Figure 6-9 represents the conversion rate of fuel sulfur to organic sulfur in diesel emissions

for LSDF and HSDF at different engine loads.

160
 900
800
(µg S in DPM / g S input)
700 at LSDF (G+P)
Conversion rate

600
500 at HSDF (G+P)
400
at LSDF (P)
300
200
at HSDF (P)
100
0
0 25 50 75
Engine loads (kW)

Figure 6-9 Conversion rate of fuel sulfur to the emitted organic sulfur for HSDF and LSDF at
different engine loads.

As shown in Figure 6-9, for LSDF, only 0.069-0.079% of fuel sulfur is converted to (or

remains as) organic sulfur in diesel emissions, among which 47-65% is in particulate phase. This

means more than 99.9% of fuel sulfur is converted to inorganic sulfur, including gaseous sulfur

dioxide (SO2), sulfur trioxide (SO3), and sulfuric acid (H2SO4) and particulate sulfate (SO42-).

Liu et al. [27] has reported that the conversion of fuel sulfur to total particulate sulfur was in the

range of 0.12-0.50%, among which 50-100% occurs in the form of particulate sulfate. This

result means 0-50% of particulate sulfur, which is 0-0.25% of fuel sulfur input, may be in the

form of organic sulfur, which is consistent with our results. For high sulfur diesel fuel, the

conversion of fuel sulfur to organic sulfur in diesel emissions is even lower, ranging from

0.027% to 0.043%. A possible reason for the lower recovery is that higher sulfur diesel fuel

contains a higher proportion of lighter sulfur components, and these lighter sulfur compounds are

mostly combusted and converted to SO2 and SO42-.

Figure 6-9 also indicates that the conversion of fuel sulfur to particulate organic sulfur is

not well correlated with engine load and the trend is different for LSDF and HSDF. For LSDF,

161
the conversion of fuel sulfur to particulate organic sulfur is higher at 50 kW and 75 kW than at 0

and 25 kW, while for HSDF, the situation is opposite. Some studies have reported that the sulfur

conversion rate to particulate sulfur is not strongly correlated with engine load, which is

consistent with our results [27]. The conversion of fuel sulfur to total organic sulfur in diesel

emissions, including both gas phase and particulate phase, is basically higher at low loads than at

high loads. The possible reason is that at high loads, the combustion temperature is high, which

causes the fuel sulfur being more completely converted to SO2 and SO42-.

6.3.3.3 Effect of Fuel Sulfur on DPM PASH Distribution

Fuel sulfur content not only affects the concentrations of organosulfur compounds in

diesel emission, but also affects their distribution. Table 6-5 has shown that as fuel sulfur is

increased from 433 ppmw to 2284 ppmw, the concentration of individual compounds in DPM

increases by 2-9 times. The effect of fuel sulfur on the distribution of organosulfur in DPM is

similar for each load condition. Figure 6-10 shows the PASH distributions in DPM for LSDF and

HSDF with 0 kW and 75 kW as examples. A significant influence from fuel sulfur content on the

distribution of sulfur compounds in DPM has been observed. With low sulfur diesel fuel, the

distribution curves are sharp, with 2,4,6-TMDBT as the most abundant PASH; while for DPM

from HSDF, the curves are broader (more uniformed distribution), especially at 0 kW. In

addition to 2,4,6-TMDBT, DBT, 4-MDBT, and 4,6-DMDBT are all major PASH products. This

distribution suggests that DPM generated from HSDF contains more PASHs with lower

molecular weights relative to LSDF. The presence (Figure 6-10) of heavier PASHs (i.e.,

BPh9,10T) was observed at maximum load in DPM from HSDF, but not from LSDF. This may

162
indicate a greater possibility of forming higher molecular weight PASHs when the sulfur

concentration is higher.

35
30
Sulfur (µg S/g DPM)

25 L0
20 H0
15 L75
10 H75
5
0
2,3,4,7-TTMBT
2,3,5-TMBT

2,3,7-TMBT

2,4,6-TMDBT
DBT

4-MDBT

4,6-DMDBT

Ph45T

BN12T

BN21T

BN23T

Ph34T

BPh9,10T

DiAT
Figure 6-10 Organosulfur distribution in DPM under 0 kW and 75 kW for LSDF (L) and
HSDF (H).

6.3.3.4 Effect of Engine Load on DPM PASH Distribution

With both fuels, engine load affects the distribution of organosulfur compounds due to

the variations in fuel usage rate, excess oxygen and combustion temperature, etc. Figures 6-11(a-

c) illustrate the sulfur 181 AED chromatograms for the diesel emission extracts from PUF, XAD,

and filter fractions. Comparing the chromatograms at 0 kW and 75 kW, engine load does not

significantly affect organosulfur compounds in gas phase diesel emissions, while alters the

organosulfur species and their distribution in particulate phase. In particulate phase diesel

emissions, an obvious shift to heavier sulfur compounds is observed at high load, with most of

the lighter compounds being absent.

163
 Retention time

Figure 6-11 Sulfur (181 nm) AED chromatograms of gas and particulate phase diesel
emissions generated at 75 kW with low sulfur (433 ppmw) diesel fuel.
(a) Collected with PUF, (b) collected with XAD, and (c) collected on filter.

As the engine load was increased from 0 kW to 75 kW, three different types of PASH

distributions in DPM at LSDF and HSDF were observed: increasing with load, decreasing with

load, and bell-shaped distribution. In Figures 6-12(a-c), DPM at HSDF was used as an example,

and the concentrations were expressed in terms of µg S/g DPM and µg S/kg DF (normalized to

diesel fuel usage).

164
 35 10
Sulfur (µg S/g DPM)
2,3,5-TMBT

Sulfur (µg S/kg DF)

30 8
25 2,3,7-TMBT
20 6
2,3,4,7-TTMBT
15 4
10 DBT
5 2
4-MDBT
0 0
0 25 50 75 0 25 50 75
Engine load (kW) Engine load (kW)

(a) Decreasing with load

50 14
Sulfur (µg S/g DPM)

Sulfur (µg S/kg DF)

12 4,6-DMDBT
40
10
30 8
2,4,6-TMDBT
20 6
4
10 2 Ph45T
0 0
0 25 50 75 0 25 50 75
Engine load (kW) Engine load (kW)

(b) Bell-shaped distribution

10 2.5
Sulfur (µg S/g DPM)

Sulfur (µg S/kg DF)

BN12T
8 2.0
6 1.5 BN21T
4 1.0
BN23T
2 0.5
0 0.0
0 25 50 75 0 25 50 75
Engine load (kW) Engine load (kW)

(c) Increasing with load

Figure 6-12 PASH concentration, expressed in terms of µg S/g DPM and µg S/kg DF, in DPM
generated with HSDF vs. engine loads.

For PASHs with lower molecular weights (MW ≤ 198), such as 2,3,5-TMBT, 2,3,7-

TMBT, 2,3,4,7-TTMBT, DBT, and 4-MDBT (2,3,5-TMBT and 2,3,7-TMBT were not found in

DPM from LSDF), the concentration was reduced as load increased (Figure 6-12(a)). These

165
PASHs also are present in diesel fuels. As the load was increased, the fuel input increased, which

should result in a concentration increase for these lighter PASHs. However, as load increased,

the combustion temperature also increased and a greater fraction of the lighter sulfur compounds

was completely burned and converted to SO2 and SO3. As a result, a lower amount of these

PASHs remained in the particulate fraction.

For higher molecular weight PASHs (MW ≥ 234), which include BN12T, BN21T,

BN23T, Ph34T and BPh9,10T, the concentration increased with load (Figure 6-12(c)). These

heavier PASHs were not present in diesel fuels and are possibly formed during combustion. As

the load increased, so did the combustion temperature and fuel input, which may promote

formation of higher molecular weight PASHs.

For intermediate molecular weight PASHs (198 < MW < 234), such as 4,6-DMDBT,

2,4,6-TMDBT and Ph45T, a bell-shaped distribution was observed (Figure 6-12(b)). As load was

increased from 0 kW to 75 kW, the concentration first increased and then decreased, with the

highest concentration observed at approximately 25 kW. These compounds are present in diesel

fuels, and this distribution may be the result of competition between the increase in fuel input

and combustion decomposition. As load increased, more fuel input led to a larger amount of

PASHs in the DPM, while the increased combustion decomposition lowered the PASH

concentration. The combination of both effects resulted in the distribution shown in Figure 6-

12(b).

6.4 Summary and Conclusions

PASHs in LSDF and HSDF and the corresponding DPM emissions were determined by

GC/AED. Results for calibration standards, NIST SRM 2724b, and fuels indicated the sulfur

166
response factor in this application is relatively independent of analyte structure. As expected, the

PASH concentration and distribution in diesel fuels depends on the sulfur content. The most

abundant sulfur species are DBT, C1-DBTs, and C3-BTs in both diesel fuels. In low sulfur diesel

fuel, heavier PASHs such as DMDBT are also higher in concentration, which indicates the LSDF

contains a larger fraction of heavier components than the HSDF.

In additions to C3-BT, DBT, C1-DBT, and C2-DBT which are abundant in diesel fuel, the

most abundant PASHs in diesel emissions also include some lighter organosulfur compounds in

gas phase, such as C1-T, BT, and C1- and C2-BT, and some heavier PASHs in particulate phase,

such as C3-DBT and PH45T. Some heavier PASHs (e.g., BNTs) not found in diesel fuels, which

are more hazardous, are present in DPM and may be generated through the pyrogenic pathway

during the combustion process. 0.027-0.079% of fuel sulfur has been converted to emission

organic sulfur. The organic sulfur content and PASH concentration and distribution in DPM are

affected by both fuel sulfur and engine load. Higher fuel sulfur led to the following two

consequences: higher total organic sulfur in diesel emissions and more PASHs with lower

molecular weights in DPM. Engine load is an important factor affecting PASH distribution. With

engine load increases, the dominant PASH species in DPM shift to higher molecular weight.

Three types of distributions were observed as the engine load was increased from 0 kW to 75 kW.

The concentrations decreased for lighter PASHs and increased for heavier PASHs as load was

increased from 0 kW to 75 kW. For intermediate molecular weight PASHs, a bell-shaped

distribution was observed, with the highest concentrations at 25 kW.

6.5 References

1. Ma, X.L.; Sakanishi, K.Y.; and Mochida, I., Hydrodesulfurization reactivities of various
sulfur compounds in diesel fuel, Ind. Eng. Chem. Res., 33(2): 218-222, 1994.

167
2. Mössner, S.G.; and Wise, S.A., Determination of polycyclic aromatic sulfur heterocycles
in fossil fuel-related samples, Anal. Chem., 71(1): 58-69, 1999.

3. Jacob, J., Sulfur Analogues of Polycyclic Aromatic Hydrocarbons (Thiaarenes),

Cambridge University Press, Cambridge, UK, 1990, p 70.

4. Shafi, R.; and Hutchings, G.J., Hydrodesulfurization of hindered dibenzothiophenes: an

overview, Catalysis Today, 59(3-4): 423-442, 2000.

5. Song, C.S., An overview of new approaches to deep desulfurization for ultra-clean

gasoline, diesel fuel and jet fuel, Catalysis Today, 86(1-4): 211-263, 2003.

6. Hegazi, A.H.; Andersson, J.T.; and El-Gayar, M.S., Application of gas chromatography
with atomic emission detection to the geochemical. investigation of polycyclic aromatic
sulfur heterocycles in Egyptian crude oils, Fuel Process. Technol., 85(1): 1-19, 2003.

7. Breitkreitz, M.C.; Raimundo, I.M.; Rohwedder, J.J.R.; et al., Determination of total

sulfur in diesel fuel employing NIR spectroscopy and multivariate calibration, Analyst,
128(9): 1204-1207, 2003.

8. Speight, J.G., Handbook of Petroleum Product Analysis, New York: Wiley-Interscience,

2002, p 43.

9. US Environmental Protection Agency (EPA) Health Assessment Document for Diesel

Engine Exhaust. Prepared by the National Center for Environmental Assessment,
Washington, DC, for the Office of Transportation and Air Quality; EPA/600/8-90/057F,
2002. Available from: National Technical Information Service, Springfield, VA;
PB2002-107661.

10. NIOSH, Current Intelligence Bulletin 50 - Carcinogenic Effects of Exposure to Diesel

Exhaust, US Department of Health and Human Services, Public Health Service, Centers
for Disease Control, National Institute for Occupational Safety and Health, DHHS
(NIOSH), Cincinnati, OH, NIOSH Publication no. 88-116, 1988.

11. Liang, F.Y.; Lu, M.M.; Keener, T.C.; et al., The organic composition of diesel particulate
matter, diesel fuel and engine oil of a non-road diesel generator, J. Environ. Monit., 7(10):
983-988, 2005.

12. Andersson, J.T., Some unique properties of gas chromatography coupled with atomic-
emission detection, Anal. Bioanal. Chem., 373(6): 344-355, 2002.

13. Becker, G.; and Colmsjö, A., Gas chromatography atomic emission detection for
quantification of polycyclic aromatic sulfur heterocycles, Anal. Chim. Acta, 376(3): 265-
272, 1998.

14. Andersson, J.T.; and Schmid, B., The atomic-emission detector in gas-chromatographic
trace analysis - some studies on the performance and applications, Fresenius J. Anal.
Chem., 346(4): 403-409, 1993.

168
15. Pedersen-Bjergaard, S.; Asp, T.N.; and Greibrokk, T., Determination of sulfur-containing
and chlorine-containing compounds using capillary gas-chromatography and atomic
emission detection, Anal. Chim. Acta, 265(1): 87-92, 1992.

16. American Society for Testing and Materials (ASTM), Annual Book of ASTM Standards,
vol. 05.01, ASTM, Conshohocken, PA, 2005.

17. Depauw, G.A.; and Froment, G.F., Molecular analysis of the sulphur components in a
light cycle oil of a catalytic cracking unit by gas chromatography with mass
spectrometric and atomic emission detection, J. Chromatogr. A, 761(1-2): 231-247, 1997.

18. US EPA, Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur
Control Requirements, EPA 420-F-00-057, December 2000.

19. US EPA, Office of Transportation and Air Quality, Regulatory Announcement, EPA420-
F-04-032, May 2004.

20. Schulz, H.; Böhringer, W.; Ousmanov, F.; et al., Refractory sulfur compounds in gas oils,
Fuel Process. Technol., 61(1-2): 5-41, 1999.

21. Shi, J.P.; and Harrison, R.M., Investigation of ultrafine particle formation during diesel
exhaust dilution, Environ. Sci. Technol., 33(21): 3730-3736, 1999.

22. Duran, A.; Carmona, M.; and Ballesteros, R., Competitive diesel engine emissions of
sulphur and nitrogen species, Chemosphere, 52(10): 1819-1823, 2003.

23. Yu, F.Q.; Turco, R.P.; and Karcher, B., The possible role of organics in the formation and
evolution of ultrafine aircraft particles, J. Geophys. Res.-Atmos., 104(D4): 4079-4087,
1999.

24. Pelroy, R.A.; Stewart, D.L.; Tominaga, Y.; et al., Microbial mutagenicity of 3-ring and 4-
ring polycyclic aromatic sulfur heterocycles, Mutat. Res., 117(1-2): 31-40, 1983.

25. McFall, T.; Booth, G.M.; Lee, M.L.; et al., Mutagenic activity of methyl-substituted
tricyclic and tetracyclic aromatic sulfur heterocycles, Mutat. Res., 135(2): 97-103, 1984.

26. Neeft, J.P.A.; Makkee, M.; and Moulijn, J.A., Diesel particulate emission control, Fuel
Process. Technol., 47(1): 1-69, 1996.

27. Liu, Z.; Lu, M.; Keener, T.C.; et al. The sulfur speciation of diesel emissions from a non-
road generator. in Proceedings of the 98th A&WMA Annual Conference & Exhibition.
June 21-24, 2005. Minneapolis, MN.

169
 Chapter 7

CONCLUSIONS AND RECOMMENDATIONS

7.1 Conclusions

Diesel particulate matter (DPM) has been regarded as a potential occupational carcinogen

and the hazardous health effect of DPM emissions has been confirmed and regulated.

Considering the adverse health effect of DPM emissions, significant efforts have been put on

studying DPM emissions from on-road diesel vehicles. However, there have been few studies on

the chemical composition and formation mechanisms of DPM from non-road diesel engines,

especially as a function of fuel properties and testing conditions, even though the use of non-road

diesel engines has increased rapidly and the resultant DPM emissions are of significant quantity.

In this study, GC/MS with selective ion chromatogram (SIC) and tandem mass

spectrometry (MS/MS) with selected ion monitoring (SIM) were developed and applied for the

identification and quantification of PAHs. The successful measurement of PAHs in the complex

DPM matrix indicates that SIC and MS/MS techniques are effective and provide satisfactory

selectivity and low detection limit for measuring trace components in complex matrices.

Diesel particulate matter from a non-road diesel generator burning low- and high-sulfur

diesel fuel was collected using the dilution method and the stack sampling method at variated

engine load conditions. The DPM chemical composition was instrumentally determined and the

results were compared under variated fuel sulfur contents, engine loads, and sampling methods.

The chemical composition of possible sources contributed to DPM emissions was also

characterized and compared with DPM composition. These experiments were conducted to

170
provide important information for understanding DPM formation mechanisms and assessing its

health effects.

The research findings suggest that DPM contains 52-84% soluble organic fraction,

among which only a small percentage can be resolved by GC/MS. In the GC resolvable fraction,

alkanes, PAHs and alkylated PAHs, and organic acids, which are attributed to the unburned

diesel fuel, the combustion process, and the evaporated engine oil, respectively, are predominant

species. This result suggests that DPM control strategies should emphasize the control of fuel

properties, and engine designs, as well as the engine oil quality. The comparison between the

composition of the diesel fuel and DPM has indicated that DPM contains more fractions of

higher molecular weight compounds, which are likely resultant from pyrogenic processes.

The experimental findings have indicated that the gas/particle distribution of organic

compounds in diesel emissions is significantly related with their vapor pressures. Light

compounds with vapor pressures higher than 10-4.07 torr, such as n-alkanes C10-C18 and 2- and 3-

ring PAHs, are predominantly present in the gas phase emission, while heavy compounds with

vapor pressures lower than 10-4.84 torr, such as n-alkanes C20-C25 and 5- and 6-ring PAHs, occur

primarily or exclusively in the particulate phase. The G/P partitioning coefficient Kp of a

compound could be related to its vapor pressure p L0 by the following equation: log Kp = mr

log p L0 + br. The linear regression results have indicated that the correlation of log Kp vs. log p L0

is statistically significant. The shallow slopes of the linear regression results compared with

estimated values suggest that the gas-to-particle adsorption of compounds is the major

mechanism. The filter breakthrough for n-alkanes and non-exchangeable effects for PAHs may

occur during the sampling events. The understanding of G/P partitioning for organic compounds

in diesel emissions helps to evaluate the health effects of diesel emissions, which are strongly

171
related to the physical forms in which the compounds are present when the exposure occurs. In

addition, studies on G/P partitioning of compounds in diesel emissions are helpful in

understanding DPM formation mechanisms, especially the adsorption and absorption

contributions in forming secondary diesel aerosol.

The applied research examined and confirmed the effects of fuel properties, engine

operation conditions, and sampling methods on DPM emission rate, its chemical composition,

and/or DPM formation mechanism. Increased engine loads result in higher emission rate of DPM,

organic carbon (OC), and elemental carbon (EC). However, the fraction of OC in DPM

decreases with loads. DPM formation mechanisms are also affected by engine loads. Higher

engine load increases the pyrogenic formation of compounds while it weakens the role of

petrogenic pathway. The sampling method is an important factor, affecting the concentration and

distribution of organics in DPM. Compared with the stack sampling method, the dilution process

results in an increase of 40-80% in DPM emission, and significantly increases the amount of

individual organic components of DPM due to the continuous adsorption, nucleation,

coagulation, and/or condensation of the large quantities of condensable organics and inorganics

during dilution processes. The dilution process also alters the distribution of those organic

compounds, causing the enrichment of low molecular weight components. This result suggests

that the dilution process favors the formation of light compounds more than the formation of

heavy compounds. The study of fuel sulfur effect on diesel emissions has indicated that when the

fuel sulfur content is increased from 400 ppmw to 2200 ppmw, DPM emissions are increased by

20%, and an increase between 20-200% is obtained for individual compounds presented in diesel

particulate matter. This result suggests that sulfur in diesel exhausts may act as a nucleation core

for all components presented in diesel exhausts. However, the relative distribution of these

172
organic components does not significantly vary with fuel sulfur contents, which may relate to the

similar composition of low and high sulfur diesel fuel. These findings will help to improve diesel

engine designs, promote the establishment of more stringent regulations for diesel fuels, improve

fuel reformulation technologies, and therefore reduce DPM emissions. In addition, investigating

the effects of sampling methods on DPM composition will provide essential information for

understanding the formation, condensation and transportation of organic compounds, and

therefore helps to evaluate DPM contribution and exposure.

Polycyclic aromatic sulfur heterocycles (PASHs) are of interest due to their potential

mutagenic and carcinogenic properties. However, studies on PASH content in DPM are difficult

since the concentration of PASHs is low in DPM, there is a great variety of compounds present,

and the quantitative determination of individual PASH isomers in the complex mixture is

difficult. This study successfully applied gas chromatography with the atomic emission detection

(GC/AED) and compound independent calibration (CIC) techniques on the measurement of

PASHs in DPM, using optimized temperature program and specific reagent gases. The research

also provides a potential, highly selective, and effective analytical technology for measuring

organosulfur species in other complex sample matrices.

PASHs in low and high sulfur diesel fuel and the corresponding DPM emissions were

determined by GC/AED. Results from fuel analysis have indicated that the PASH concentration

and distribution in the diesel fuel vary with fuel sulfur contents. The amount of PASHs in high

sulfur diesel fuel is higher than in low sulfur diesel fuel. However, low sulfur diesel fuel contains

relatively more fractions of heavy PASHs, which are possibly more hazardous. This result

suggests that more stringent regulations on fuel quality and further fuel sulfur reduction are

required.

173
 Experimental findings on PASHs in DPM indicate that 0.027-0.079% of the fuel sulfur

has been converted to the emitted organic sulfur. DPM contains more complex PASH species

than diesel fuels, with benzothiophene, dibenzothiophenes, and their alkylated homologue as the

major species. Some heavier PASHs (e.g., BNTs) not found in diesel fuels are present in DPM.

They are more hazardous and may be generated through the pyrogenic pathway during the

combustion process. Both fuel sulfur contents and engine loads affect the formation of PASHs in

DPM. Higher fuel sulfur content leads to higher PASH emissions. The effect of engine loads on

the PASH formation in DPM is more complex, and three types of distributions are observed as

the engine load is increased from 0 kW to 75 kW: decreasing with loads for lighter PASHs,

increasing with loads for heavier PASHs, and bell-shaped distributions for intermediate

molecular weight PASHs.

7.2 Recommendations

Although the research findings have provided important information for DPM health

effect assessments, its formation mechanisms, and control strategies, more research is needed on

DPM chemical composition, especially as a function of DPM size distribution, quantitative

contribution to DPM components from various sources, and sampling methodology and

characterization of nano-sized DPM. The following studies are recommended:

1. The investigation on the quantitative correlation of diesel emissions from petrogenic

and pyrogenic sources is recommended.

This research has identified the three sources of DPM emissions and their contributions

to DPM chemical components; however, the quantitative contribution to a compound from each

source has not been well studied. This research has preliminarily studied the survivability of

174
phenanthrene using procedures described in Chapter 2. The selection of phenanthrene as the

additive is due to its abundance in both diesel fuel and diesel emissions. The amount of

phenanthrene added (~1% by weight) is determined from the consideration that this additive

would not alter the combustion process in diesel engine. The survivability value for

phenanthrene calculated from Eq. (2-1) using the data listed in Table 7-1 is 0.0115%. The

survivability value is considered to differ with the type of compounds, type of engines, operating

condition, collection condition, and other factors [1, 2]. From Eq. (2-2), the unburned diesel fuel

contributes to the major proportion (61.2%) of phenanthrene recovered in DPM, while other

sources of phenanthrene, presumably pyrosynthetic formation, is 38.8% of the recovered

phenanthrene in DPM. Table 7-2 summarized literature results for survivabilities and sources of

PAHs in diesel emissions. The relative importance of pyrosynthesis and unburned fuel greatly

depends on PAH species and engine operating conditions.

Both radiotracer method and the method used in this study can be applied for

investigations on source apportionment of PAHs in diesel emissions. Especially, the radiotracer

method can also be applied for studying engine oil contributions.

From the experiment for Ph-enriched fuel, it is also found that the addition of

phenanthrene affects aromatic emissions from diesel engines. For Ph-enriched diesel fuel, the

amount of aromatics, especially alkylated naphthalenes and phenanthrenes and alkylbenzenes, in

particulate diesel emissions has been increased to 1.3-3 times as high as for normal fuel.

However, the emission of alkanes and alkanoic acids was not affected by the addition of

phenanthrene in fuel. Further investigations on combustion processes and aromatic formation

mechanisms when burning Ph-enriched fuel are recommended.

175
Table 7-1 Results for phenanthrene survivability experiment.
Fuel type Phenanthrene Phenanthrene Fuel Stack gas
concentration in concentration in consumption flow rate
diesel fuel (mg/L) DPM (ng/m3) rate (L/hr) (m3/hr)
Normal fuel 198 709 5 263
Ph-enriched fuel 10000 22213 5 263

Table 7-2 Literature results for contributions from unburned fuel and pyrosysnthesis to
PAHs in diesel emissions.
Compounds Survivability Contribution from unburned fuel or pyrosynthesis
Phenanthrene 0.0115% 61% from unburned fuel, 0 kW
(this study)
Most PAHs 0.05-0.2% [3]
Naphthalene 0.48% [4]; 24% from unburned fuel [4]; Pyrosynthesis is the major
0.35-1%a [2] source [2].
Benzo(a)pyrene 0.04% [5] 80% from unburned fuel [5]
Fluorene 0.87% [6]; Unburned fuel is the major source except for low-speed and
0.2-1.3%a [2] low-load condition [2].
Pyrene 0.17 [6]; At low load, pyrosynthesis is the major source; at high load,
0.2-0.4%a [2] pyrosynthesis and unburned fuel contribute almost equally[2].
Fluoranthene 0-0.1%a [2] Pyrosynthesis is the major source [2].
a
The data was estimated from literature results.

2. Measurements of nitro-PAHs and oxygenated PAHs, which are highly carcinogenic

and mutagenic, carbonyl compounds, and metals are recommended for the integrality of DPM

chemical composition and the accuracy of DPM health effect assessments.

3. In addition, characterization of DPM from the combustion of a biodiesel blend, which

is the mixture of biodeisel blended with petroleum diesel at some level, and the comparison of

DPM composition between biodiesel and normal petroleum diesel are also recommended.

Biodiesel is a domestic, biodegradable, and renewable fuel for diesel engines, derived from

natural oils like vegetable oil or animal fat. Due to the increasing interest in the use of biodiesel,

176
the U.S. Environmental Protection Agency has conducted a comprehensive analysis of the

emission impact of biodiesel using publicly available data. This investigation shows that the use

of biodiesel in diesel engines results in substantial reduction of unburned hydrocarbons, carbon

monoxide, and particulate matter [7]. Detailed compositional analysis of DPM from biodiesel

blend will help to assess its health effects and to instruct the proper application of biodiesel.

4. The investigation on sampling methods for ultrafine and nano-sized diesel particles,

and the investigation on DPM size-dependent composition are also recommended.

Measurements of ultrafine and nano-sized diesel particles are challenging, since much longer

sampling time is required in order to get enough materials. High-throughput sampling method for

ultrafine and nano-sized diesel particles is a necessity. The composition of diesel particles varies

with their size, which leads to the size-dependent toxicity. Studying the size-dependent

composition of diesel particles will be essential for a better understanding of their health effects

and the development of effective emission control strategies.

5. The implementation of efforts using mathematical modeling to predict DPM emission

and its chemical composition at specific conditions is recommended. This research has correlated

DPM composition with source contribution, engine operation condition, and sampling condition.

Combining the information from this research through some mathematical methods should

deduce a predicting model, which will significantly contribute to estimating DPM exposure,

evaluating DPM health effects, and developing effective control strategies.

It is hoped that the present applied research will contribute to the assessment of DPM

health effect, the establishment of regulations on fuel quality and DPM exposure, the evolution

of engine design and operations, and the development of DPM control technologies.

177
7.3 References

1. Williams, P.T.; Abbass, M.K.; Andrews, G.E.; et al., Diesel particulate emissions: The
pole of unburned fuel, Combust. Flame, 75(1): 1-24, 1989.

2. Rhead, M.M.; and Hardy, S.A., The sources of polycyclic aromatic compounds in diesel
engine emissions, Fuel, 82(4): 385-393, 2003.

3. Abbass, M.K.; Andrews, G.E.; and Williams, P.T., The influrence of diesel fuel
composition on particulate PAH emissions, SAE Technical Paper Series 892079, 1989.

4. Rhead, M.M.; and Pemberton, R.D., Sources of naphthalene in diesel exhaust emissions,
Energy Fuels, 10(3): 837-843, 1996.

5. Tancell, P.J.; Rhead, M.M.; Trier, C.J.; et al., The sources of benzo[a]pyrene in diesel
exhaust emissions, Sci. Total Environ., 162(2-3): 179-186, 1995.

6. Tancell, P.J.; Rhead, M.M.; Pemberton, R.D.; et al., Survival of polycyclic aromatic
hydrocarbons during diesel combustion, Environ. Sci. Technol., 29(11): 2871-2876, 1995.

7. United States Environmental Protection Agency, Draft Technical Report, A

Comprehensive Analysis of Biodiesel Impacts on Exhaust Emissions, EPA420-P-02-001,
October 2002.

178
 Appendix A

CHROMATOGRAMS FOR SAMPLES

A.1 GC/MS Total Ion Chromatogram (TIC) for Diesel Fuel and Diesel
Emissions

kCounts RIC all No.2DF-1000--4-29-2004.SMS

600 Diesel fuel
500

400

300

200

100

0
MCounts RIC all L0-Fillter.SMS
1.25 DPM
1.00

0.75

0.50

0.25

0.00
MCounts RIC all L0-PUF-Dilute.SMS
Gas phase diesel
0.75 emission: PUF fraction

0.50

0.25

0.00
MCounts RIC all L0-XAD-Dilute.SMS
Gas phase diesel
1.25 emission: XAD fraction
1.00

0.75

0.50

0.25

0.00

Figure A.1 Total ion chromatogram of diesel fuel and DPM (time in minutes).

179
A.2 GC/MS Selected Ion Chromatogram (SIC) for Diesel Fuel and Diesel
Emissions
kCounts Ion: 128 all No.2DF-1000--4-29-2004.SMS
7
6 Naphthalene Ion: 128
5
4
3
2
1
0
kCounts Ion: 152 all No.2DF-1000--4-29-2004.SMS
1.50 Acenaphthylene Ion: 152
1.25
1.00
0.75
0.50
0.25
0.00
kCounts Ion: 154 all No.2DF-1000--4-29-2004.SMS
1.00 Ion: 154
0.75 Acenaphthene
0.50

0.25

0.00
kCounts Ion: 166 all No.2DF-1000--4-29-2004.SMS

1.5 Ion: 166

Fluorene
1.0

0.5

0.0
11 12 13 14 15 16
minutes
0
kCounts Ion: 178 all No.2DF-1000--4-29-2004.SMS
Phenanthrene
1.25 Ion: 178
1.00 Anthracene
0.75

0.50

0.25

0.00
Counts Ion: 202 all No.2DF-1000--4-29-2004.SMS

200 Ion: 202

150
Pyrene
100

50

0
18 19 20 21 22 23
minutes

Figure A.2 Selected ion chromatogram (SIC) for PAHs in diesel fuel (time in minutes).

180
 0 00
kCounts Ions: 141+142+115 all No.2DF-1000--4-29-2004
2-MN
Ions: 141+142+115
40 1-MN
30

20

10

0
kCounts Ions: 156+141+115 all No.2DF-1000--4-29-2004
C2-N 4 Ions: 156+141+115

2, 3 5
30

20
6, 7
1
10 8
9
0
kCounts Ions: 170+155+128 all No.2DF-1000--4-29-2004
C3-N Ions: 170+155+128
15

10

0
12 13 14 15 16 17 18 19
minutes

kCounts Ions: 192+191+165 all No.2DF-1000--4-29-2004

2.5
Ions: 192+191+165
2.0

1.5 C1-178

1.0

0.5

kCounts Ions: 206+191+189 all No.2DF-1000--4-29-2004

Ions: 206+191+189
2.0 C2-178
1.5

1.0

0.5

19.5 20.0 20.5 21.0 21.5 22.0

minutes

Figure A.3 Selected ion chromatogram (SIC) for alkylated PAHs in diesel fuel (time in
minutes).

181
 0
kCounts Ions: 91+92 all No.2DF-1000--4-29-2004.SMS
Toluene Ions: 91+92
12.5
10.0

7.5

5.0
2.5

0.0
kCounts Ions: 106+91 all No.2DF-1000--4-29-2004.SMS
20 2 Ions: 106+91
3 C2-B
15

10 1
5

0
kCounts Ions: 120+105+91 all No.2DF-1000--4-29-2004.S
50
7 Ions: 120+105+91
40 3
30 C3-B 8
4,5,6
20
2
10 1
0
kCounts Ions: 134+119+91 all No.2DF-1000--4-29-2004.S
Ions: 134+119+91 8-10 12
25 6 11
20 C4-B
15 2 5
10 3 4 7
5 1
0
4 5 6 7 8 9
minutes
0
kCounts Ions: 148+133+105 all No.2DF-1000--4-29-2004
15.0 Ions: 148+133+105
12.5
C5-B
10.0
7.5
5.0
2.5
0.0
kCounts Ions: 162+147+119 all No.2DF-1000--4-29-2004
15.0
Ions: 162+147+119
12.5 C6-B
10.0
7.5
5.0
2.5
0.0
9.5 10.0 10.5 11.0 11.5 12.0 12.5
minutes

Figure A.4 Selected ion chromatogram (SIC) for alkylbenzenes in diesel fuel (time in
minutes).

182
 Counts Ion: 128 all 0kW-74.18mg4-4-2004.SMS
250
Ion: 128
200 Naphthalene
150

100

50

0
9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5
minutes
0
kCounts Ion: 152 all 0kW-74.18mg4-4-2004.SMS
Acenaphthylene Ion: 152
2.0

1.5

1.0

0.5

0.0
kCounts Ion: 154 all 0kW-74.18mg4-4-2004.SMS
Acenaphthene
Ion: 154
1.5

1.0

0.5

0.0
kCounts Ion: 166 all 0kW-74.18mg4-4-2004.SMS
Ion: 166
3 Fluorene
2

0
14.5 15.0 15.5 16.0 16.5 17.0
minutes

kCounts Ion: 178 all 0kW-74.18mg4-4-2004.SMS

30 Phenanthrene
25
Ion: 178
Anthracene
20
15
10
5
0
18.0 18.5 19.0 19.5 20.0 20.5
minutes

kCounts Ion: 202 all dpm_zifei.sms

6 Pyrene Ion: 202
5
4 Fluoranthene
3
2
1
0
21.5 22.0 22.5 23.0 23.5 24.0
minutes

Figure A.5 Selected ion chromatogram (SIC) for PAHs in DPM (time in minutes).

183
 0
kCounts Ions: 141+142+115 all 0kW-74.18mg4-4-2004.S
2-MN
12.5 Ions: 141+142+115
10.0 1-MN
7.5

5.0

2.5

0.0
kCounts Ions: 156+141+115 all 0kW-74.18mg4-4-2004.S

C2-N 4 Ions: 156+141+115

20 5
15 2,3 6,7
10 1 8
9
5

0
kCounts Ions: 170+155+128 all 0kW-74.18mg4-4-2004.S
25
Ions: 170+155+128
20

15 C3-N
10

0
13 14 15 16 17
minutes
0
kCounts Ions: 192+191+165 all 0kW-74.18mg4-4-2004.S

12.5 Ions: 192+191+165

10.0
C1-178
7.5

5.0

2.5

0.0
kCounts Ions: 206+191+189 all 0kW-74.18mg4-4-2004.S

Ions: 206+191+189
7.5 C2-178

5.0

2.5

0.0
20 21 22 23 24
minutes

Figure A.6 Selected ion chromatogram (SIC) for alkylated PAHs in DPM (time in minutes).

184
 0
kCounts Ions: 91+92 all 75kW-50.66mg4-5-2004.SMS

3 Toluene Ions: 91+92

0
kCounts Ions: 106+91 all 75kW-50.66mg4-5-2004.SMS

6
C2-B 2 Ions: 106+91
5 3
4
3
1
2
1
0
3.5 4.0 4.5 5.0 5.5 6.0 6.5
minutes
0
kCounts Ions: 120+105+91 all 75kW-50.66mg4-5-2004.S
C3-B
60 7 Ions: 120+105+91
50
40 8
30 1 2 3,4,5 6
20
10
0
6.5 7.0 7.5 8.0 8.5 9.0 9.5
minutes
0
kCounts Ions: 134+119+91 all 75kW-50.66mg4-5-2004.S
C4-B 8,9,10 13
40 6 11 Ions: 134+119+91
30
12
20 2 3
45 7
10
1
0
8.0 8.5 9.0 9.5 10.0 10.5 11.0
minutes
0
kCounts Ions: 148+133+105 all 75kW-50.66mg4-5-2004.
20
Ions: 148+133+105
C5-B
15

10

0
kCounts Ions: 162+147+119 all 75kW-50.66mg4-5-2004.
Ions: 162+147+119
20
C6-B
15

10

0
9.5 10.0 10.5 11.0 11.5 12.0 12.5
minutes

Figure A.6 Selected ion chromatogram (SIC) for alkylbenzenes in DPM (time in minutes).

185
 Appendix B

DATA FOR DIESEL FUEL COMPONENT

Table B.1 Organic compounds present in low and high sulfur diesel fuela, in µg/g DF.
Compound LSDFb HSDFc Compound LSDF HSDF
n-Alkanes
n-Decane (C10) 12115 4388 n-Octadecane (C18) 8727 11288
n-Undecane (C11) 11271 13163 n-Nonadecane (C19) 4988 6038
n-Dodecane (C12) 17149 25688 n-Eicosane (C20) 2193 1425
n-Tridecane (C13) 28834 32888 n-Heneicosane (C21) 1092 1088
n-Tetradecane (C14) 25604 36375 n-Docosane (C22) 756 450
n-Pentadecane (C15) 27660 36413 n-Tricosane (C23) 220 338
n-Hexadecane (C16) 23965 37913 n-Tetracosane (C24) 107 225
n-Heptadecane (C17) 26082 27863 n-Pentacosane (C25) NDd ND
Sum of n-alkanes 190763 235650

Branched alkanes
Norfarnesane (C14) 11469 9936 Pristine (C19) 5871 4045
Farnesane (C15) 9719 7639 Phytane (C20) 4775 3203
Norpristane (C18) 7992 6939 Other branched alkanes 328578 287624
Sum of branched alkanes 368404 319386

Cycloalkanes
n-Heptylcyclohexane (C13) 13144 12837 n-Tetradecylcyclohexane (C20) 25 225
n-Octylcyclohexane (C14) 11467 10025 n-Pentadecylcyclohexane (C21) ND ND
n-Nonylcyclohexane (C15) 10582 8885 n-Hexadecylcyclohexane (C22) ND ND
n-Decylcyclohexane (C16) 9135 7450 n-Heptadecylcyclohexane (C23) ND ND
n-Undecylcyclohexane (C17) 6207 5480 n-Octadecylcyclohexane (C24) ND ND
n-Dodecylcyclohexane (C18) 2073 3471 n-Nonadecylcyclohexane (C25) ND ND
n-Tridecylcyclohexane (C19) 165 337
Sum of saturated cycloalkanes 52798 48710

PAHs
Naphthalene (Nap) 753 1425 Benzo(a)anthracene (Baa) ND ND
Acenaphthylene (Acy) 159 250 Chrysene (Chy) ND ND
Acenaphthene (Ace) 85 287 Benzo(b)fluoranthene (Bbf) ND ND
Fluorene (Flu) 100 525 Benzo(k)fluoranthene (Bkf) ND ND

186
Compound LSDFb HSDFc Compound LSDF HSDF
Phenanthrene (Phe) 247 425 Benzo(a)pyrene (Bap) ND ND
Anthracene (Ant) 7.5 50 Indeno[1,2,3-cd]pyrene (Ind) ND ND
Fluoranthene (Flt) ND 25 Dibenz(a,h)anthracene (Dba) ND ND
Pyrene (Pyr) 5.0 50 Benzo(ghi)perylene (Bgp) ND ND
Biphenyl 437 1079
Sum of PAHs 1793 4116

Alkylated PAHs
1-MN 585 3475 1,8-DMN ND 0
2-MN 2291 5650 2,3-DMN ND 1225
1,2-DMN 373 650 2,6-DMN 1224 4425
1,3-DMN ND 3425 2,7-DMN 1837 4725
1,4-DMN 1540 925 TMN 12327 13025
1,5-DMN ND 125 1-MPh 242 893
1,6-DMN 1807 2250 2-MPh 528 3280
1,7-DMN 2548 2725 DMPh ND 2675
Sum of alkylated PAHs 25302 49473

Alkylbenzenes
Toluene (C1-B) 1377 2059 C4-Benzenes (C4-B) 9724 10021
C2-Benzenes (C2-B) 12932 13258 C5-Benzenes (C5-B) 5538 7645
C3-Benzenes (C3-B) 10003 10689 C6-Benzenes (C6-B) 5222 5790
Sum of alkylbenzenes 44796 49462
a
All of the compounds detected were quantified with standard compounds. b Low sulfur diesel fuel with
sulfur content 400 ppmv. c High sulfur diesel fuel with sulfur content 2200 ppmv. d ND: not detected, i.e.
the concentration is lower than detection limit or the compound is not present in the sample.

187
 Appendix C

DATA FOR DIESEL EMISSIONS

C.1 Data for Diesel Emissions

Table C.1 Parameters for the non-road diesel generator.
Engine load Fuel consumption rate Stack gas flow rate Air/fuel ratio (mass/mass
(kW) (L/hr) (m3/hr) basis)
0 5.0 263 85
25 10.0 286 46
50 15.0 341 37
75 19.5 382 32

Table C.2 DPM mass, OC fraction, and EC fraction in diesel emissions.

Concentration (mg/m3) Fraction in DPM
Engine
Organic compounds Non-carbonaceous
load DPM OC EC EC
(1.2×OC) materials
High volume dilution method, low sulfur diesel fuel
0 kW 5.01±0.58 3.64±0.49 0.63±0.09 87.1% 12.6% 0.3%
25 kW 7.35±1.13 4.43±0.65 1.55±0.26 72.2% 21.0% 6.7%
50 kW 10.79±1.57 5.38±0.78 3.77±0.54 59.8% 34.9% 5.2%
75 kW 16.88±1.59 6.42±0.91 8.42±0.88 45.7% 49.9% 4.4%

High volume dilution method, high sulfur diesel fuel

0 kW 6.05 4.56 0.54 90.5% 9.0% 0.5%
25 kW 9.35 5.94 1.41 76.2% 15.1% 8.8%
50 kW 13.29 6.99 3.69 63.1% 27.8% 9.1%
75 kW 20.08 8.27 8.04 49.4% 40.1% 10.5%

EPA Method 5, low sulfur diesel fuel

0 kW 2.74 1.41 0.58 61.7% 21.2% 17.2%
25 kW 5.08 2.19 1.73 51.8% 34.1% 14.2%
50 kW 7.48 2.45 3.90 39.3% 52.1% 8.6%
75 kW 9.38 0.71 7.89 9.2% 84.1% 6.7%

188
Table C.3 Composition of DPM collected with EPA Method 5, in µg/g DPM.
Compound 0 kW 75 kW Compound 0 kW 75 kW
n-Alkanes
n-Decane (C10) 5.3 1.1 n-Octadecane (C18) 375 20
n-Undecane (C11) 16 1.5 n-Nonadecane (C19) 427 23
n-Dodecane (C12) 21 1.6 n-Eicosane (C20) 384 28
n-Tridecane (C13) 26 3.2 n-Heneicosane (C21) 276 25
n-Tetradecane (C14) 48 5.2 n-Docosane (C22) 162 15
n-Pentadecane (C15) 79 7.2 n-Tricosane (C23) 107 9.9
n-Hexadecane (C16) 159 11 n-Tetracosane (C24) 75 6.5
n-Heptadecane (C17) 219 14 n-Pentacosane (C25) 64 5.7
Sum of n-alkanes 2444 176

Branched alkanes
Norfarnesane (C14) 7.6 3.0 Pristine (C19) 123 86
Farnesane (C15) 16 6.0 Phytane (C20) 152 95
Norpristane (C18) 104 68 Other branched alkanes 923 326
Sum of branched alkanes 1325 584

Cycloalkanes
n-Heptylcyclohexane (C13) 2.8 0.57 n-Tetradecylcyclohexane (C20) 17 7.0
n-Octylcyclohexane (C14) 4.4 0.80 n-Pentadecylcyclohexane (C21) 13 4.0
n-Nonylcyclohexane (C15) 9.1 1.6 n-Hexadecylcyclohexane (C22) 11 3.0
n-Decylcyclohexane (C16) 12 3.0 n-Heptadecylcyclohexane (C23) 8.5 2.6
n-Undecylcyclohexane (C17) 21 4.0 n-Octadecylcyclohexane (C24) 7.2 1.9
n-Dodecylcyclohexane (C18) 32 5.8 n-Nonadecylcyclohexane (C25) 5.5 1.6
n-Tridecylcyclohexane (C19) 24 7.9
Sum of saturated cycloalkanes 168 44

PAHs
Naphthalene (Nap) 1.5 0.60 Benzo(a)anthracene (Baa) 7.8 3.8
Acenaphthylene (Acy) 1.2 0.20 Chrysene (Chy) 7.8 4.3
Acenaphthene (Ace) 1.3 0.20 Benzo(b)fluoranthene (Bbf) 3.0 1.4
Fluorene (Flu) 4.4 0.40 Benzo(k)fluoranthene (Bkf) 2.7 1.2
Phenanthrene (Phe) 23 4.9 Benzo(a)pyrene (Bap) 4.9 2.1
Anthracene (Ant) 1.8 0.40 Indeno[1,2,3-cd]pyrene (Ind) 3.8 1.7
Fluoranthene (Flt) 16 5.0 Dibenz(a,h)anthracene (Dba) 1.8 0.94
Pyrene (Pyr) 28 9.7 Benzo(ghi)perylene (Bgp) 2.8 1.8
Sum of PAHs 112 39

Alkylated PAHs
1-MN 8.2 1.9 1,8-DMN NDa ND

189
Compound 0 kW 75 kW Compound 0 kW 75 kW
2-MN 25 4.1 2,3-DMN 1.3 ND
1,2-DMN 7.3 0.74 2,6-DMN 11 2.6
1,3-DMN 46 12 2,7-DMN 7.9 ND
1,4-DMN 5.3 ND TMN 757 34
1,5-DMN 30 6.3 1-MPh 481 11
1,6-DMN 12 3.5 2-MPh 686 16
1,7-DMN 5.3 ND DMPh 1032 30
Sum of alkylated PAHs 3114 122

Alkylbenzenes
Toluene (C1-B) 6.8 4.3 C4-Benzenes (C4-B) 12 35
C2-Benzenes (C2-B) 13 8.9 C5-Benzenes (C5-B) 5.6 26
C3-Benzenes (C3-B) 11 16 C6-Benzenes (C6-B) 3.0 13
Sum of alkylbenzenes 51 104

n-Alkanoic acids
Hexanoic acid (C6) 0 1.5 Tridecanoic acid (C13) 93 11
Heptanoic acid (C7) 3.6 6.8 Tetradecanoic acid (C14) 425 22
Octanoic acid (C8) 8.1 10 Pentadecanoic acid (C15) 281 9.7
Nonanoic acid (C9) 18 20 Hexadecanoic acid (C16) 474 10
Decanoic acid (C10) 45 26 Heptadecanoic acid (C17) 25 4.2
Undecanoic acid (C11) 55 19 Octadecanoic acid (C18) 11 1.7
Dodecanoic acid (C12) 89 28
Sum of n-alkanoic acids 1528 170

Aromatic acids
Benzoic acid 58 35
a
ND: not detected, i.e. the concentration is lower than detection limit or the compound is not present in
the sample.

190
Table C.4 Composition of DPM collected with natural dilution method and high volume
dilution method, in µg/g DPM.
Compound Natural dilution method High volume dilution method
0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
n-Alkanes
n-Decane (C10) 9.7 14 11 3.8 116 86 42 24
n-Undecane (C11) 22 26 11 5.6 164 123 77 47
n-Dodecane (C12) 31 31 16 7.4 222 179 99 55
n-Tridecane (C13) 57 47 21 12 337 273 123 79
n-Tetradecane (C14) 204 159 33 14 500 364 167 116
n-Pentadecane (C15) 565 229 51 18 1085 783 261 203
n-Hexadecane (C16) 1164 440 73 30 2206 1414 459 308
n-Heptadecane (C17) 2645 981 101 46 3189 2287 655 441
n-Octadecane (C18) 1479 865 129 67 3851 2625 896 516
n-Nonadecane (C19) 1257 724 179 92 3415 2509 948 560
n-Eicosane (C20) 742 506 133 81 2769 2293 836 528
n-Heneicosane (C21) 138 291 126 69 1921 1578 637 404
n-Docosane (C22) 10 124 76 40 1041 949 354 225
n-Tricosane (C23) 7.4 49 44 33 574 420 252 146
n-Tetracosane (C24) 5.2 27 21 17 421 320 140 113
n-Pentacosane (C25) 3.6 25 18 14 324 213 165 128
Sum of n-alkanes 8341 4537 1045 552 22137 16417 6112 3893

Branched alkanes
Norfarnesane (C14) 13 12 10 6.5 154 136 81 61
Farnesane (C15) 28 19 18 6.8 269 293 164 91
Norpristane (C18) 135 119 102 89 1065 1088 876 635
Pristine (C19) 147 134 124 107 1204 1197 940 697
Phytane (C20) 185 148 139 116 1454 1317 1067 760
Other branched alkanes 7084 2835 418 288 22894 13958 2823 2295
Sum of branched alkanes 7593 3267 812 614 27040 17988 5950 4538

Cycloalkanes
n-Heptylcyclohexane (C13) 11 4.5 2.0 0.90 21 14 14 8.7
n-Octylcyclohexane (C14) 27 7.8 3.5 1.5 46 36 32 23
n-Nonylcyclohexane (C15) 29 15 7.2 2.6 101 82 72 50
n-Decylcyclohexane (C16) 43 23 14 5.0 168 122 102 70
n-Undecylcyclohexane (C17) 39 38 22 7.5 209 163 127 86
n-Dodecylcyclohexane (C18) 61 44 30 13 150 129 117 82
n-Tridecylcyclohexane (C19) 12 29 27 15 138 116 98 63
n-Tetradecylcyclohexane (C20) 9.8 18 15 9.6 78 79 64 42
n-Pentadecylcyclohexane (C21) 6.8 14 11 5.5 54 45 38 28

191
Compound Natural dilution method High volume dilution method
0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
n-Hexadecylcyclohexane (C22) 4.3 9.3 8.2 4.1 45 40 30 17
n-Heptadecylcyclohexane (C23) NDa 7.9 6.6 2.8 31 22 22 14
n-Octadecylcyclohexane (C24) ND 5.6 4.2 2.0 26 18 14 10
n-Nonadecylcyclohexane (C25) ND 5.3 3.5 1.9 13 8.1 5.7 7.1
Sum of saturated cycloalkanes 242 221 154 71 1079 876 735 500

PAHs
Naphthalene (Nap) 4.1 1.6 1.3 3.2 35 22 12 7.4
Acenaphthylene (Acy) 2.0 0.8 0.7 1.6 17 11 7 4.3
Acenaphthene (Ace) 3.0 1.6 0.7 3.7 33 25 11 11
Fluorene (Flu) 7.1 3.2 2.0 5.3 32 21 16 11
Phenanthrene (Phe) 77 42 28.8 30 224 143 91 70
Anthracene (Ant) 4.1 2.4 1.3 3.3 17 12 9 5.8
Fluoranthene (Flt) 25 29 33.3 20 78 51 52 42
Pyrene (Pyr) 67 62 66.0 35 146 88 77 56
Benzo(a)anthracene (Baa) 5.1 11 13.1 8.3 21 15 16 12
Chrysene (Chy) 5.1 13 15.0 11 29 19 17 16
Benzo(b)fluoranthene (Bbf) 3.0 11 10.5 3.0 5.9 7.20 7.35 7.5
Benzo(k)fluoranthene (Bkf) 2.0 4.8 3.3 1.5 5.0 6.64 6.74 6.2
Benzo(a)pyrene (Bap) 4.1 10 8.5 5.3 11.1 12.14 11.71 10
Indeno[1,2,3-cd]pyrene (Ind) 6.6 6.3 3.9 3.7 8.2 9.64 7.63 7.4
Dibenz(a,h)anthracene (Dba) 1.7 0.0 0.0 1.8 2.9 3.01 2.45 2.5
Benzo(ghi)perylene (Bgp) 3.0 3.2 1.3 2.9 6.4 7.53 8.24 6.1
Sum of PAHs 220 202 190 139 673 454 353 275

Alkylated PAHs
1-MN 30 4.5 5.5 4.8 112 51 26 16
2-MN 86 18 18 14 228 118 54 35
1,2-DMN 31 14 5.4 2.2 145 78 21 15
1,3-DMN 235 128 69 39 562 434 210 113
1,4-DMN ND ND 53 19 274 152 101 68
1,5-DMN 134 203 12 18 326 248 84 68
1,6-DMN 61 41 21 12 327 161 83 45
1,7-DMN 25 21 ND ND 101 75 33 27
1,8-DMN ND ND ND ND 0 0 0 0
2,3-DMN ND ND ND ND 24 16 12 5
2,6-DMN 54 27 13 9 171 102 48 44
2,7-DMN 49 24 ND ND 147 81 27 11
TMN 2914 1591 256 123 5625 3230 543 422
1-MPh 1770 747 50 41 3157 1528 164 118

192
Compound Natural dilution method High volume dilution method
0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
2-MPh 2168 949 161 35 3978 2019 289 156
DMPh 3420 1821 236 95 6268 3199 577 362
Sum of alkylated PAHs 10977 5590 899 412 21445 11490 2271 1503

Alkylbenzenes
Toluene (C1-B) 11 14 24 27 41 46 175 198
C2-Benzenes (C2-B) 21 22 37 50 72 69 322 358
C3-Benzenes (C3-B) 15 17 48 79 43 54 346 464
C4-Benzenes (C4-B) 16 15 106 148 47 43 644 826
C5-Benzenes (C5-B) 6.2 22 88 105 23 65 550 554
C6-Benzenes (C6-B) 3.2 13 42 55 11 35 227 278
Sum of alkylbenzenes 71 103 346 464 236 312 2264 2678

n-Alkanoic acids
Hexanoic acid (C6) ND ND 9.0 23 10 15 23 45
Heptanoic acid (C7) 18 23 56 62 27 52 100 134
Octanoic acid (C8) 40 60 78 94 57 127 174 168
Nonanoic acid (C9) 88 104 123 140 125 208 271 256
Decanoic acid (C10) 289 256 208 198 398 487 443 318
Undecanoic acid (C11) 354 259 179 133 478 467 294 184
Dodecanoic acid (C12) 591 406 231 183 786 691 343 217
Tridecanoic acid (C13) 444 256 130 54 582 410 172 69
Tetradecanoic acid (C14) 1955 856 361 125 2503 1285 481 170
Pentadecanoic acid (C15) 1238 421 190 50 1559 590 236 62
Hexadecanoic acid (C16) 1991 547 216 62 1899 712 295 84
Heptadecanoic acid (C17) ND 1.3 0.0 2.5 5.2 8.4 4.8 3.1
Octadecanoic acid (C18) 4.1 8.2 5.3 5.2 10 11 8.2 5.3
Sum of n-alkanoic acids 7012 3201 1785 1133 8441 5063 2845 1716

Aromatic acids
Benzoic acid 226 289 376 430 305 426 469 677
a
ND: not detected, i.e. the concentration is lower than detection limit or the compound is not present in
the sample.

193
Table C.5 Compound concentration (ng/m3) in particulate and gas phase diesel emissions
collected with high volume dilution method (LSDF).
Compound Particulate phase Gas phase
abbreviationa 0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
n-Alkanes
C10 414 515 352 309 136976 174113 122843 112078
C11 584 732 644 589 155926 203587 180209 166199
C12 791 1066 823 701 182801 259092 220656 194255
C13 1201 1629 1022 996 199363 294799 263693 263940
C14 1780 2173 1388 1473 195823 293366 315076 312276
C15 3863 4666 2171 2567 206106 251963 280072 310565
C16 7854 8428 3825 3891 183518 203853 231871 276279
C17 11354 13638 5455 5572 153544 183653 217305 250734
C18 13711 15651 7466 6526 86420 109513 140640 151958
C19 12161 14960 7893 7075 29933 51851 76285 101382
C20 9859 13674 6965 6678 6651 17459 33575 50164
C21 6840 9406 5308 5103 2741 3083 16703 21240
C22 3708 5661 2952 2846 627 1004 2068 2268
C23 2045 2506 2097 1852 146 846 1028 1394
C24 1498 1911 1169 1431 68 412 592 669
C25 1155 1270 1375 1617 37 272 322 372
Sum of n-alkanes 78817 97887 50906 49227 1503570 2172268 2251197 2458172

Branched alkanes
C14 549 812 672 777 108678 171370 160596 180255
C15 958 1745 1368 1145 140850 277479 250412 254205
C18 3791 6485 7295 8026 132686 237981 375362 488322
C19 4287 7136 7828 8813 63275 147633 254892 367380
C20 5177 7852 8884 9606 29501 89386 174902 243950
Other branched 81514 83223 23514 29021 1059676 1414790 752453 1102803
alkanes
Sum of branched 96276 107253 49562 57389 1534665 2338639 1968617 2636915
alkanes

Cycloalkanes
C13 76 86 113 110 8263 11722 17242 17427
C14 163 216 265 291 9760 14688 21494 24177
C15 358 490 600 634 9287 15271 21184 21684
C16 599 729 847 883 7776 11260 15390 18178
C17 742 973 1058 1085 6252 9812 13635 15449
C18 533 770 974 1034 2327 4898 7440 8666
C19 491 693 814 803 1013 2235 4682 5982

194
Compound Particulate phase Gas phase
abbreviationa 0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
C20 276 473 531 528 295 909 2134 2733
C21 192 269 314 359 108 232 458 689
C22 159 238 252 212 17 157 233 324
C23 111 132 184 173 7.5 67 137 185
C24 94 106 120 121 3.2 49 86 117
C25 46 49 47 90 NDb 16 35 60
Sum of saturated 3841 5224 6120 6323 45109 71316 104150 115671
cycloalkanes

PAHs
Nap 126 132 101 94 9600 19138 23086 26520
Acy 59 68 62 55 2100 3812 5160 5745
Ace 119 147 92 134 3320 6412 9955 12590
Flu 114 128 134 143 2440 4184 6386 7593
Phe 797 854 760 887 3388 7756 11798 12901
Ant 62 69 72 74 355 738 841 1234
Flt 279 307 434 532 215 382 571 816
Pyr 519 524 644 706 139 540 718 884
Baa 76 92 129 152 14 34 57 79
Chy 103 115 145 198 11 23 65 67
Bbf 21 43 61 95 ND ND 1.6 2.6
Bkf 18 40 56 78 ND ND ND ND
Bap 40 72 98 127 ND ND 2.0 2.0
Ind 29 57 64 93 ND ND 2.0 1.0
Dba 10 18 20 31 ND ND 0.73 1.7
Bgp 23 45 69 77 ND ND ND 1.7
Sum of PAHs 2395 2709 2940 3475 21741 43157 58825 68677

Alkylated PAHs
1-MN 399 303 217 205 25520 27543 32574 34988
2-MN 812 704 450 438 52831 63142 65441 75169
1,2-DMN 518 465 175 190 12144 15611 18189 18485
1,3-DMN 2003 2585 1749 1428 30632 37206 43556 57594
1,4-DMN 975 903 837 854 15723 32760 39958 52836
1,5-DMN 1159 1479 700 855 25664 41285 46329 57014
1,6-DMN 1163 957 687 571 16177 25902 33338 35498
1,7-DMN 359 447 275 341 9126 13485 17665 21550
1,8-DMN ND ND ND ND ND ND ND ND
2,3-DMN 86 95 99 58 1995 2692 3470 3660
2,6-DMN 608 608 400 550 10175 13082 15849 21420

195
Compound Particulate phase Gas phase
abbreviationa 0 kW 25 kW 50 kW 75 kW 0 kW 25 kW 0 kW 75 kW
2,7-DMN 523 483 229 141 11484 7212 4656 3864
TMN 20027 19259 4523 5336 27801 39899 60421 72386
1-MPh 11242 9111 1366 1492 940 1307 2176 2848
2-MPh 14162 12038 2407 1973 2444 2513 2894 3517
DMPh 22318 19073 4804 4578 1053 1526 1682 1978
Sum of alkylated 76353 68508 18917 19010 243709 325165 388199 462805
PAHs

Alkylbenzenes
C1-B 144 274 1459 2510 8333 15719 69533 95941
C2-B 255 411 2680 4533 7659 13285 120220 219847
C3-B 152 321 2885 5866 3709 8494 112611 259674
C4-B 167 257 5365 10439 2175 4978 132269 293494
C5-B 82 387 4578 6999 770 4897 62162 149636
C6-B 39 208 1894 3520 331 2316 27113 61804
Sum of 840 1858 18862 33867 22978 49689 523908 1080395
alkylbenzenes

n-Alkanoic acids
C6 36 89 191 573 2189 4465 9572 30768
C7 95 308 836 1690 2803 7824 29095 49411
C8 204 757 1447 2126 4699 18172 37616 63959
C9 444 1243 2255 3239 6156 20300 46378 84557
C10 1419 2906 3694 4016 11326 29057 55405 68273
C11 1703 2785 2446 2332 8733 17002 19640 20254
C12 2800 4118 2854 2741 5916 13274 12230 12961
C13 2073 2447 1429 876 2707 6320 5554 3997
C14 8911 7660 4009 2152 5585 8037 8650 5384
C15 5552 3515 1970 788 1336 1707 2208 1148
C16 6763 4242 2456 1056 384 424 1409 743
C17 18 50 40 39 1.6 6.9 7.5 8.7
C18 36 65 68 66 1.4 3.9 7.6 8.6
Sum of n-alkanoic 30054 30185 23695 21696 51838 126592 227770 341472
acids

Aromatic acids
Benzoic acid 1085 2540 3907 8561
a
The abbreviation used in this table refers to the corresponding compound listed in Table C.4. b ND: not
detected, i.e. the concentration is lower than detection limit or the compound is not present in the sample.

196
Table C.6 Normalized emission rates of compounds in gas and particulate phase diesel
emissions collected with high volume dilution method (LSDF).
Compounds mg emission/kg fuel used mg emission/g compound in fuel
0kW 25kW 50kW 75kW 0kW 25kW 50kW 75kW
n-Alkanes
n-Decane (C10) 9.03 6.24 3.50 2.75 0.75 0.52 0.29 0.23
n-Undecane (C11) 10.29 7.30 5.14 4.08 0.91 0.65 0.46 0.36
n-Dodecane (C12) 12.07 9.30 6.29 4.77 0.70 0.54 0.37 0.28
n-Tridecane (C13) 13.19 10.60 7.52 6.49 0.46 0.37 0.26 0.22
n-Tetradecane (C14) 12.99 10.57 8.99 7.68 0.51 0.41 0.35 0.30
n-Pentadecane (C15) 13.81 9.17 8.02 7.67 0.50 0.33 0.29 0.28
n-Hexadecane (C16) 12.58 7.59 6.70 6.86 0.53 0.32 0.28 0.29
n-Heptadecane (C17) 10.84 7.05 6.33 6.28 0.42 0.27 0.24 0.24
n-Octadecane (C18) 6.58 4.47 4.21 3.88 0.75 0.51 0.48 0.44
n-Nonadecane (C19) 2.77 2.39 2.39 2.66 0.55 0.48 0.48 0.53
n-Eicosane (C20) 1.09 1.11 1.15 1.39 0.49 0.51 0.53 0.63
n-Heneicosane (C21) 0.63 0.54 0.63 0.65 0.58 0.41 0.57 0.59
n-Docosane (C22) 0.29 0.24 0.14 0.13 0.38 0.32 0.19 0.17
n-Tricosane (C23) 0.14 0.12 0.09 0.08 0.65 0.54 0.40 0.36
n-Tetracosane (C24) 0.10 0.08 0.05 0.05 0.96 0.78 0.47 0.48
n-Pentacosane (C25) 0.08 0.06 0.05 0.05 NAa NA NA NA
Total n-alkanes 104 81 65 61 0.55 0.43 0.34 0.32

Branched alkanes
Norfarnesane (C14) 7.18 6.16 4.58 4.43 0.63 0.54 0.40 0.39
Farnesane (C15) 9.32 9.98 7.15 6.25 0.96 1.03 0.74 0.64
Norpristane (C18) 8.97 8.74 10.87 12.15 1.12 1.09 1.36 1.52
Pristine (C19) 4.44 5.53 7.47 9.21 0.76 0.94 1.27 1.57
Phytane (C20) 2.28 3.48 5.22 6.21 0.48 0.73 1.09 1.30
Other branched alkanes 75 54 22 28 0.23 0.16 0.07 0.08
Total branched alkanes 107 87 57 66 0.29 0.24 0.16 0.18

Cycloalkanes
n-Heptylcyclohexane (C13) 0.55 0.42 0.49 0.43 0.04 0.03 0.04 0.03
n-Octylcyclohexane (C14) 0.65 0.53 0.62 0.60 0.06 0.05 0.05 0.05
n-Nonylcyclohexane (C15) 0.63 0.56 0.62 0.55 0.06 0.05 0.06 0.05
n-Decylcyclohexane (C16) 0.55 0.43 0.46 0.47 0.06 0.05 0.05 0.05
n-Undecylcyclohexane (C17) 0.46 0.39 0.42 0.40 0.07 0.06 0.07 0.07
n-Dodecylcyclohexane (C18) 0.19 0.20 0.24 0.24 0.09 0.10 0.12 0.11
n-Tridecylcyclohexane (C19) 0.10 0.10 0.16 0.17 0.60 0.63 0.95 1.01
n-Tetradecylcyclohexane (C20) 0.04 0.05 0.08 0.08 1.50 1.98 3.03 3.19
n-Pentadecylcyclohexane (C21) 0.02 0.02 0.02 0.03 NA NA NA NA

197
Compounds mg emission/kg fuel used mg emission/g compound in fuel
0kW 25kW 50kW 75kW 0kW 25kW 50kW 75kW
n-Hexadecylcyclohexane (C22) 0.01 0.01 0.01 0.01 NA NA NA NA
n-Heptadecylcyclohexane (C23) 0.008 0.007 0.009 0.009 NA NA NA NA
n-Octadecylcyclohexane (C24) 0.006 0.006 0.006 0.006 NA NA NA NA
n-Nonadecylcyclohexane (C25) 0.003 0.002 0.002 0.004 NA NA NA NA
Total saturated cycloalkanes 3.22 2.74 3.13 2.99 0.06 0.05 0.06 0.06

PAHs
Naphthalene (Nap) 0.64 0.69 0.66 0.65 0.85 0.91 0.88 0.87
Acenaphthylene (Acy) 0.14 0.14 0.15 0.14 0.89 0.87 0.93 0.89
Acenaphthene (Ace) 0.23 0.23 0.29 0.31 2.66 2.76 3.36 3.67
Fluorene (Flu) 0.17 0.15 0.19 0.19 1.68 1.54 1.85 1.89
Phenanthrene (Phe) 0.28 0.31 0.36 0.34 1.11 1.25 1.44 1.37
Anthracene (Ant) 0.03 0.03 0.03 0.03 3.65 3.84 3.46 4.27
Fluoranthene (Flt) 0.03 0.02 0.03 0.03 NA NA NA NA
Pyrene (Pyr) 0.04 0.04 0.04 0.04 8.66 7.60 7.74 7.79
Benzo(a)anthracene (Baa) 0.006 0.005 0.005 0.006 NA NA NA NA
Chrysene (Chy) 0.008 0.005 0.006 0.006 NA NA NA NA
Benzo(b)fluoranthene (Bbf) 0.001 0.002 0.002 0.002 NA NA NA NA
Benzo(k)fluoranthene (Bkf) 0.001 0.001 0.002 0.002 NA NA NA NA
Benzo(a)pyrene (Bap) 0.003 0.003 0.003 0.003 NA NA NA NA
Indeno[1,2,3-cd]pyrene (Ind) 0.002 0.002 0.002 0.002 NA NA NA NA
Dibenz(a,h)anthracene (Dba) 0.001 0.001 0.001 0.001 NA NA NA NA
Benzo(ghi)perylene (Bgp) 0.001 0.002 0.002 0.002 NA NA NA NA
Total PAHs 1.59 1.64 1.76 1.77 1.17 1.21 1.29 1.30

Alkylated PAHs
Methylnaphthalenes (MN)b 5.23 3.28 2.80 2.71 1.82 1.14 0.98 0.94
Dimethylnaphthalenes (DMN)c 9.24 7.05 6.48 6.78 0.99 0.76 0.69 0.73
Trimethylnaphthalenes (TMN)d 3.14 2.11 1.85 1.90 0.26 0.17 0.15 0.15
Methylphenanthrenes (MPh)e 1.89 0.89 0.25 0.24 2.46 1.16 0.33 0.31
Dimethylphenanthrenes (DMPh)f 1.54 0.74 0.18 0.16 NA NA NA NA
Total alkylated PAHs 21 14 12 12 0.83 0.56 0.46 0.47

Alkylbenzenes
Toluene (C1-B) 0.56 0.57 2.02 2.41 0.40 0.42 1.47 1.75
C2-Benzenes (C2-B) 0.52 0.49 3.49 5.49 0.04 0.04 0.27 0.42
C3-Benzenes (C3-B) 0.25 0.32 3.28 6.50 0.03 0.03 0.33 0.65
C4-Benzenes (C4-B) 0.15 0.19 3.91 7.44 0.02 0.02 0.40 0.77
C5-Benzenes (C5-B) 0.06 0.19 1.90 3.84 0.01 0.03 0.34 0.69
C6-Benzenes (C6-B) 0.02 0.09 0.82 1.60 0.00 0.02 0.16 0.31

198
Compounds mg emission/kg fuel used mg emission/g compound in fuel
0kW 25kW 50kW 75kW 0kW 25kW 50kW 75kW
Total alkylbenzenes 1.57 1.84 15 27 0.03 0.04 0.34 0.61

Total unburned hydrocarbons 239 189 155 171

a
Not available since these compounds are not present in diesel fuels. b Methylnaphthalenes include 1- and
2-methylnaphthalene. c Dimethylnaphthalenes include all isomer that have been identified as dimethyl- or
ethyl- naphthalene based on NIST library search or standard compound. d Trimethylnaphthalenes include
all isomer that have been identified as trimethyl- or ethylmethyl- naphthalene based on NIST library
search or standard compound. e Methylphenanthrenes include all isomer that have been identified as
methylphenanthrene or methylanthracene based on NIST library search or standard compound. f
Dimethylphenanthrenes include all isomer that have been identified as dimethyl- or ethyl- phenanthrene
or anthracene based on NIST library search or standard compound.

199
Table C.7 Compound concentration (µg/g DPM) in DPM collected with high volume
dilution method (HSDF).
Compound 0kW 25kW 50kW 75kW Compound 0kW 25kW 50kW 75kW
abbreviation abbreviation
n-Alkanes
C10 6.9 7.7 6.1 12 C18 4635 4619 2460 1511
C11 28 17 13 12 C19 4229 4953 3423 1681
C12 54 18 10 8 C20 4557 5457 3014 1483
C13 234 157 54 29 C21 3516 4239 3028 1411
C14 633 442 215 126 C22 1660 2311 1213 882
C15 1194 1089 624 429 C23 1014 1476 1002 565
C16 3193 2160 2006 1268 C24 725 997 666 351
C17 4173 3233 1415 892 C25 492 1305 568 390
Sum of n-alkanes 30345 32481 19718 11048

Branched alkanes
C14 75 131 76 119 C19 2321 2232 1402 1049
C15 133 172 61 38 C20 2733 3091 1587 1073
C18 1720 1993 1134 863 Other branched 37576 20506 11950 4410
alkanes
Sum of branched 44558 28125 16209 7553
alkanes

Cycloalkanes
C13 71 33 26 16 C20 783 1357 1326 988
C14 109 99 68 48 C21 879 1010 336 207
C15 237 289 133 112 C22 445 453 210 148
C16 630 694 323 161 C23 466 514 201 127
C17 965 1191 587 329 C24 323 432 135 76
C18 1396 1589 949 593 C25 272 319 104 68
C19 1392 1303 572 421
Sum of saturated 7967 9283 4969 3294
cycloalkanes

PAHs
Nap 35 33 24 28 Baa 184 86 27 21
Acy 39 11 6.4 4.3 Chy 130 87 30 19
Ace 82 26 10 5.4 Bbf 60 30 27 11
Flu 90 79 25 4.0 Bkf 43 26 10 8.2
Phe 553 549 154 129 Bap 24 18 7.4 2.6
Ant 46 38 22 5.2 Ind 28 26 18 10
Flt 225 178 71 20 Dba 10 14 3.6 3.4

200
Compound 0kW 25kW 50kW 75kW Compound 0kW 25kW 50kW 75kW
abbreviation abbreviation
Pyr 428 308 87 59 Bgp 58 29 8.2 8.2
Sum of PAHs 2034 1538 532 339

Alkylated PAHs
1-MN 89 29 16 19 1,8-DMN 0 0 0 0
2-MN 255 100 51 48 2,3-DMN 21 20 11 4
1,2-DMN 151 97 19 12 2,6-DMN 155 104 38 36
1,3-DMN 548 579 185 141 2,7-DMN 160 84 22 11
1,4-DMN 146 178 86 50 TMN 8385 5557 695 585
1,5-DMN 347 713 78 99 1-MPh 4481 2369 202 133
1,6-DMN 165 212 70 67 2-MPh 5789 2999 360 202
1,7-DMN 111 95 30 24 DMPh 11606 4738 666 423
Sum of alkylated 32410 17875 2528 1854
PAHs

Alkylbenzenes
C1-B 80 107 82 70 C4-B 83 94 321 342
C2-B 137 159 116 166 C5-B 90 119 255 284
C3-B 72 117 131 196 C6-B 47 74 107 103
Sum of 509 671 1011 1161
alkylbenzenes

n-Alkanoic acids
C6 12 20 29 60 C13 857 633 293 120
C7 36 66 136 216 C14 3754 1970 826 301
C8 77 167 255 306 C15 2527 866 377 111
C9 188 276 439 388 C16 2969 1225 433 149
C10 494 620 643 471 C17 7.7 15 9.2 6.0
C11 692 647 482 320 C18 15 19 14 7.8
C12 1322 1098 645 382
Sum of n-alkanoic 12950 7623 4583 2838
acids

Aromatic acids
Benzoic acid 448 592 680 896
a
The abbreviation used in this table refers to the corresponding compound listed in Table C.4. b ND: not
detected, i.e. the concentration is lower than detection limit or the compound is not present in the sample.

201
C.2 Conversion of Compound Emissions between Different Units

The concentration of compounds in DPM is expressed in terms of µg/g DPM and ng/m3.

The conversion between the two units is as follows:

Compound concentration (ng/m3) = compound concentration (µg/g DPM) × mass of

DPM collected (mg) / volume of stack gas collected (m3) (C.1)

The following data are used to convert the concentration of compounds in DPM between

the two units.

Table C.7 Sampling record for DPM collected with high volume dilution method (LSDF).
Engine load (kW) Mass of DPM collected (mg) Volume of stack gas sampled (m3)
0 53.82 15.12
25 72.80 12.21
50 103.84 12.47
75 124.53 9.85

For the emission rate of compounds in total diesel emissions (including gas and

particulate phase), two expressions are used: mg compound emission/kg of fuel combusted and

mg compound emission/g compound in diesel fuel. The following two equations are used to

calculate the two normalized emission rates.

C S × QS
ER1 = (C.2)
RDF × ρ DF

ER1
ER2 = (C.3)
C DF

Where:

ER1 = mg compound emission/kg of fuel combusted, mg/kg DF

CS = compound concentration in diesel emissions (gas + particulate phase), µg/m3

QS = stack gas flow rate, m3/hr, listed in Table C.1

202
RDF = fuel usage rate, L/hr, listed in Table C.1

ρDF = diesel fuel density, 800 g/L

ER2 = mg compound emission/g compound in diesel fuel, mg/g

CDF = compound concentration in diesel fuel, g/kg DF

203
 Appendix D

PASH DATA FOR DIESEL FUEL AND DIESEL EMISSIONS

Table D.1 PASHs in low and high sulfur diesel fuels (LSDF and HSDF).
Compounda Formula Sulfur contentb (µg S/g DF) ± std (δ; n = 3)
LSDF HSDF
Total organic sulfurc S 433±9 2284±12
T to 2-PT (7 compounds)d ND e
ND
BT C8H6S 1.34±0.05 8.21±0.17
2-MBT C9H8S 0.11±0.02 7.68±0.14
3-MBTf C9H8S 3.21±0.09 8.39±0.12
5-MBTg C9H8S 2.86±0.13 15.29±0.45
3,5-DMBT C10H10S 3.85±0.16 18.71±0.42
2,3,5-TMBT C11H12S 8.28±0.24 63.56±2.46
2,3,7-TMBT C11H12S 10.72±0.19 75.91±2.78
2,3,4,7-TTMBT C12H14S 4.32±0.11 24.36±0.56
DBT C12H8S 15.23±0.29 83.99±1.92
4-MDBT C13H10S 21.22±0.52 66.78±1.64
4,6-DMDBT C14H12S 10.95±0.31 20.81±0.35
2,4,6-TMDBT C15H14S 4.34±0.15 6.63±0.24
Ph45Th C14H8S 2.08±0.13 3.03±0.12
BN12T to DiAT (6 compounds)i C16H10S ND ND
a
The compound names for the abbreviations are as listed in Glossary of acronyms and symbols. b For
both fuels (LSDF and HSDF), three sets of samples were injected. c Total sulfur was calculated by
integrating the entire sulfur response as one peak. Integration began where the sulfur emission line
increased from baseline or at the beginning of the first peak (7 min for diesel fuel). The final integration
point was designated after the sulfur emission returned to baseline and remained stable (25 min for diesel
fuel). The standard temperature program indicated in Table 6-1 was used. d The 7 compounds include T,
2-MT, 3-MT, 2,3-DMT, 2,5-DMT, 2-ET, and 2-PT. e ND (not detected) indicates result below the
minimum detectable level (MDL). MDLs are based on results for a test mixture (Agilent part 8500-5067).
The specification for sulfur 181 is 2 pg/sec. The MDL for our laboratory was typically about 0.5 pg/sec,
or 1 pg S for a peak having a 2-sec width (at half height). f The coelution between 3- and 4-MBT may
occur. g The coelution between 5- and 6-MBT may occur. h This peak is assigned as Ph45T based on the
retention time of a standard compound. It is possible that this peak also contains a C3-DBT because C3-
DBTs can also elute in this region. Currently, we do not have a C3-DBT standard to examine this
possibility. i The 6 compounds include BN12T, BN21T, BN23T, Ph34T, BPh9,10T, and DiAT.

204
Table D.2 PASHs in gas and particle phase diesel emissions (0 kW through 75 kW) when
burning low and high sulfur diesel fuels.
Compounda Formula Sulfur content (µg S/g DPM) ± std (δ; n = 3)b
Gas phase emission at LSDF L0-Gc L25-G L50-G L75-G
Total organic sulfurd S 775±35 786±27 450±26 400±30
e
T C4H4S ND ND ND ND
2-MT C5H6S 5.56±0.62 10.15±0.89 3.42±0.33 5.41±0.39
3-MT C5H6S 9.90±0.78 19.01±1.25 6.10±0.52 10.37±0.89
2,3-DMT C6H8S 3.86±0.45 5.32±0.63 2.09±0.29 4.03±0.44
2,5-DMT C6H8S 4.17±0.54 6.68±0.56 2.45±0.28 4.59±0.38
2-ET C6H8S 4.03±0.49 6.82±0.61 2.72±0.32 4.13±0.26
2-PT C7H10S 1.73±0.28 4.35±0.52 1.54±0.27 2.76±0.37
BT C8H6S 9.56±0.82 7.76±0.67 3.96±0.36 4.18±0.43
2-MBT C9H8S 9.52±0.71 5.11±0.43 3.67±0.41 4.41±0.51
3-MBTf C9H8S 12.20±0.88 9.22±0.79 5.88±0.51 5.87±0.55
5-MBTg C9H8S 4.90±0.52 2.24±0.33 2.06±0.28 2.14±0.30
3,5-DMBT C10H10S 9.11±0.63 9.34±0.86 6.49±0.56 5.25±0.49
2,3,5-TMBT C11H12S 23.31±1.57 17.87±1.49 12.59±0.97 10.92±0.93
2,3,7-TMBT C11H12S 31.36±2.05 23.86±1.84 16.96±1.27 14.92±1.22
2,3,4,7-TTMBT C12H14S 2.28±0.33 1.85±0.26 1.38±0.18 0.85±0.17
DBT C12H8S 22.82±1.55 13.00±1.06 10.76±0.93 7.68±0.68
4-MDBT C13H10S 9.86±0.81 5.29±0.55 5.91±0.56 5.31±0.47
4,6-DMDBT C14H12S 3.59±0.42 ND ND ND
8 compoundsi ND ND ND ND

Particle phase emission at LSDF L0-P L25-P L50-P L75-P

Total organic sulfur S 679±32 734±27 819±25 631±35
14 compoundsj ND ND ND ND
2,3,4,7-TTMBT C12H14S 1.48±0.15 1.96±0.28 1.53±0.25 1.63±0.22
DBT C12H8S 2.84±0.23 2.68±0.32 1.77±1.89 1.51±1.75
4-MDBT C13H10S 9.07±0.68 7.33±0.88 4.71±0.55 3.55±0.41
4,6-DMDBT C14H12S 10.30±1.08 12.40±1.03 11.57±1.21 11.34±1.09
2,4,6-TMDBT C15H14S 13.37±1.22 34.50±2.94 27.69±2.49 22.38±2.23
Ph45Th C14H8S 6.42±0.75 22.39±2.19 14.86±1.78 11.68±0.94
BN12T C16H10S 0.99±0.37 2.37±0.36 4.09±0.83 5.58±0.63
BN21T C16H10S 1.13±0.25 2.87±0.43 3.83±0.75 5.11±0.82
BN23T C16H10S 0.95±0.22 1.74±0.31 2.23±0.41 4.69±0.76
Ph34T C16H10S ND ND ND 2.25±0.50
BPh9,10T C20H12S ND ND ND ND
DiAT C24H12S ND ND ND ND

205
Table D.2 (Continued)
Compound Formula Sulfur content (µg S/g DPM) ± std (δ; n = 3)
Particle phase emission at HSDF H0-P H25-P H50-P H75-P
Total organic sulfur S 1585 1599 1349 923
12 compoundsk ND ND ND ND
2,3,5-TMBT C11H12S 2.81 1.93 0.98 1.85
2,3,7-TMBT C11H12S 3.44 3.40 2.00 1.25
2,3,4,7-TTMBT C12H14S 14.28 12.67 8.52 2.49
DBT C12H8S 30.29 25.18 11.21 4.41
4-MDBT C13H10S 32.39 26.86 19.81 11.39
4,6-DMDBT C14H12S 22.91 29.28 15.31 8.25
2,4,6-TMDBT C15H14S 28.08 42.40 27.20 18.56
Ph45T C14H8S 19.05 34.78 17.80 9.55
BN12T C16H10S 1.89 3.22 5.54 7.68
BN21T C16H10S 1.69 3.06 4.61 6.86
BN23T C16H10S 1.50 2.98 4.33 6.49
Ph34T C16H10S ND 0.47 2.48 0.48
BPh9,10T C20H12S ND ND ND 0.81
DiAT C24H12S ND ND ND ND

Total organic sulfur in gas phase at HSDF 1768 1654 1282 1077
a
The compound names for the abbreviations are as listed in Glossary of acronyms and symbols. b For
DPM at LSDF, three sets of samples were collected and injected, but for DPM at HSDF only one set of
samples was collected. c L or H indicates low or high sulfur diesel fuel was used; 0 through 75 is the
engine load (kW); G or P indicates gas phase or particle phase. d Total sulfur was calculated by
integrating the entire sulfur response as one peak. Integration began where the sulfur emission line
increased from baseline or at the beginning of the first peak (7 min for DPM). The final integration point
was designated after the sulfur emission returned to baseline and remained stable (32 min for DPM).
Notes e-h are the same as those for Table D.1. i The 8 compounds include 2,4,6-TMDBT, Ph45T, BN12T,
BN21T, BN23T, Ph34T, BPh9,10T, and DiAT. j The 14 compounds include T, 2-MT, 3-MT, 2,3-DMT,
2,5-DMT, 2-ET, 2-PT, BT, 2-MBT, 3-MBT, 5-MBT, 3,5-DMBT, 2,3,5-TMBT, and 2,3,7-TMBT. k The
12 compounds include T, 2-MT, 3-MT, 2,3-DMT, 2,5-DMT, 2-ET, 2-PT, BT, 2-MBT, 3-MBT, 5-MBT,
and 3,5-DMBT.

206
 Appendix E

DATA FOR PHENANTHRENE-ENRICHED FUEL

Table E.1 Composition of DPM collected when burning phenanthrene-enriched diesel fuela
(high volume dilution sampling, 0 kW).
Compound µg/g ng/m3 Compound µg/g ng/m3
DPM DPM
n-Alkanes
n-Decane (C10) 112 331 n-Octadecane (C18) 4740 13988
n-Undecane (C11) 195 574 n-Nonadecane (C19) 4344 12819
n-Dodecane (C12) 226 667 n-Eicosane (C20) 4293 12670
n-Tridecane (C13) 423 1248 n-Heneicosane (C21) 2826 8339
n-Tetradecane (C14) 554 1635 n-Docosane (C22) 1289 3803
n-Pentadecane (C15) 1006 2969 n-Tricosane (C23) 668 1971
n-Hexadecane (C16) 2680 7909 n-Tetracosane (C24) 432 1274
n-Heptadecane (C17) 3639 10739 n-Pentacosane (C25) 363 1070
Sum of n-alkanes 27788 82005

Branched alkanes
Norfarnesane (C14) 172 507 Pristine (C19) 1394 4114
Farnesane (C15) 301 889 Phytane (C20) 1771 5228
Norpristane (C18) 1195 3527 Other branched alkanes 28487 84070
Sum of branched alkanes 33321 98334

Cycloalkanes
n-Heptylcyclohexane (C13) 21 61 n-Tetradecylcyclohexane (C20) 100 294
n-Octylcyclohexane (C14) 50 149 n-Pentadecylcyclohexane (C21) 69 203
n-Nonylcyclohexane (C15) 133 392 n-Hexadecylcyclohexane (C22) 49 145
n-Decylcyclohexane (C16) 236 696 n-Heptadecylcyclohexane (C23) 36 105
n-Undecylcyclohexane (C17) 214 633 n-Octadecylcyclohexane (C24) 27 79
n-Dodecylcyclohexane (C18) 217 642 n-Nonadecylcyclohexane (C25) 16 49
n-Tridecylcyclohexane (C19) 193 569
Sum of saturated cycloalkanes 1361 4016

PAHs
Naphthalene (Nap) 88 260 Benzo(a)anthracene (Baa) 37 110
Acenaphthylene (Acy) 47 140 Chrysene (Chy) 53 157

207
Compound µg/g ng/m3 Compound µg/g ng/m3
DPM DPM
Acenaphthene (Ace) 93 274 Benzo(b)fluoranthene (Bbf) 11 31
Fluorene (Flu) 84 248 Benzo(k)fluoranthene (Bkf) 9 26
Phenanthrene (Phe) 7527 22213 Benzo(a)pyrene (Bap) 17 50
Anthracene (Ant) 44 131 Indeno[1,2,3-cd]pyrene (Ind) 14 43
Fluoranthene (Flt) 155 458 Dibenz(a,h)anthracene (Dba) 4.8 14
Pyrene (Pyr) 294 868 Benzo(ghi)perylene (Bgp) 10 28
Sum of PAHs 8488 25049

Alkylated PAHs
1-MN 315 929 1,8-DMN ND 0
2-MN 665 1963 2,3-DMN 52 155
1,2-DMN 397 1171 2,6-DMN 415 1226
1,3-DMN 1339 3951 2,7-DMN 323 953
1,4-DMN 662 1955 TMN 10759 31751
1,5-DMN 814 2402 1-MPh 12058 35584
1,6-DMN 821 2424 2-MPh 15694 46315
1,7-DMN 245 724 DMPh 23249 68612
Sum of alkylated PAHs 67810 200114

Alkylbenzenes
Toluene (C1-B) 90 265 C4-Benzenes (C4-B) 97 285
C2-Benzenes (C2-B) 166 491 C5-Benzenes (C5-B) 58 171
C3-Benzenes (C3-B) 88 260 C6-Benzenes (C6-B) 26 76
Sum of alkylbenzenes 524 1547

n-Alkanoic acids
Hexanoic acid (C6) 36 106 Tridecanoic acid (C13) 715 2111
Heptanoic acid (C7) 73 217 Tetradecanoic acid (C14) 2837 8373
Octanoic acid (C8) 157 464 Pentadecanoic acid (C15) 1959 5782
Nonanoic acid (C9) 508 1500 Hexadecanoic acid (C16) 2262 6676
Decanoic acid (C10) 533 1574 Heptadecanoic acid (C17) 7.1 21
Undecanoic acid (C11) 968 2857 Octadecanoic acid (C18) 13 37
Dodecanoic acid (C12) 36 106
Sum of n-alkanoic acids 10070 29718

Aromatic acids
Benzoic acid 363 1071
a
Low sulfur diesel fuel with 198 mg/L phenanthrene was used as the normal fuel. After phenanthrene
was added, its concentration in Ph-enriched fuel was 10000 mg/L. b ND: not detected, i.e. the
concentration is lower than detection limit or the compound is not present in the sample.

208