James Schauer, Glynis C Lough

and Martin M. Shafer

Environmental Chemistry and Technology ProgramUniversity of Wisconsin – Madison

Madison, Wisconsin

TRACE METAL EMISSIONS FROM MOTOR VEHICLES

OUTLINE

• Objectives

• Sampling and analysis

• Measurement of motor vehicle emissions

• Ambient sampling and results

MOTIVATION

• Observed association between human health effects and increased atmospheric PM concentrations– Mechanisms of health effects from PM are not understood– Trace metals may be important components

• Origins, concentrations, and impacts of trace metals in atmospheric PM are not well defined– Traditional analytical methods are not sensitive enough for the

low levels in many aerosol samples

• Improving understanding of the metals content of atmospheric PM will be important for– Source reconciliation modeling efforts– Health effects studies

• Motor vehicles may be an important source of metals to the atmosphere, especially in urban areas

OBJECTIVES

• Characterize total roadway particulate matter emissions from motor vehicles

• Construct source profiles for specific sources of roadway trace metal emissions, including:

– Tailpipe emissions - Brake wear– Tire wear - Resuspended road dust

• Quantify and apportion contributions of these specific sources to measured total roadway emissions

• Understand the impact of motor vehicle emissions on metals levels in the ambient urban atmosphere

APPROACH

• Conduct tunnel tests to obtain a total roadway particulate matter emissions profile for on-road vehicles

• Develop source profiles for specific sources of trace metal and particulate matter emissions from vehicles– Conduct source characterization tests to develop profiles directly

parallel to tunnel emission tests • Sample vehicles similar to tunnel fleet vehicles• Apply identical collection and analysis methods to all samples

• Characterize chemical composition of all PM emissions– Focus on trace metals analysis with ICP-MS

• Use chemical mass balance model to apportion total roadway emissions to specific sources

• Conduct parallel ambient sampling, and apply factor analysis to determine contribution of vehicle emissions

BACKGROUND

• PM sizes– PM10

• smaller than 10 µm • primarily mechanically generated

– PM2.5• smaller than 2.5 µm • primarily from combustion and

chemical formation

• Sampling– Sizes are separated with impactors

or cyclones– Multiple filters of PM10 / PM2.5

collected simultaneously– Collected on filters selected and

prepared for individual analyses

MOUDI IMPACTOR

PM10 / PM2.5 SAMPLER

ANALYSIS

• Mass • Elemental and Organic carbon

– NIOSH 5040: thermal evolution and combustion

• Inorganic ions: Cl-, NO3-, SO4

2-, NH4+

– Ion Chromatography

• Metals– Inductively-Coupled Plasma Mass Spectrometry (ICP-MS)

• Organic compounds– Gas Chromatography Mass Spectrometry (GC-MS)

Trace Metals Analysis by ICPMS

• Trace metals have historically been used to track crustal materials (XRF, PIXE)

• New opportunities using ICPMS techniques– Excellent sensitivity and accuracy for heavier elements– Can explore isotopes as tracers – Can begin speciation of metals

• Many sources can only be tracked by trace metal signatures– Brake wear– Some industrial source

Critical Issues for ICPMS

• Critical points in development of ICP-MS methods for analysis of particulate matter:– Effective sample solubilization– Minimization of contamination– Removal of instrumental polyatomic interferences

• Solubilization - Microwave-assisted acid digestion in Teflon bombs with high purity acids is the most robust method for solubilization of trace metals in environmental samples

• Contamination minimization– Pre-cleaning of Teflon filters– Use of rigorous cleaning methods and clean

handling techniques – All work done in a dedicated Class 50 clean lab,

analyses in a Class 100 lab

Digestion

• Complete digestion of all elements requires an acid mixture of HF, HCl and HNO3– HF is required to digest clays– HCl is needed for digestion of platinum group metals– Closed bomb is needed to prevent vaporization of silicon

fluoride compounds– Microwave assisted digestion using an acid mixture of HF,

HCl, and HNO3 has been validated with several Standard Reference Materials (SRM)

• Protocols have been developed to leach aerosols samples with surrogate fluids that can be used for ICP analysis and bioassays:– Artificial lake water – Sheesley et al. (2003)– Surrogate lung fluid – Schauer et al. (2003)

Trace Metal Clean Room

• ICPMS has excellent sensitivity– Detection limits for ICPMS analysis are determined by

field and lab blanks and not instrument detection limits– Key issue for quantification is addressing polyatomic

interferences• Broad range of ICPMS Instruments

– Traditional ICPMS– Traditional ICPMS with collision cell and plasma shield

technology• Allows analysis of light elements by removing

polyatomic interferences– High resolution ICPMS

• Allows analysis of virtually all elements increased mass accuracy

– Multi-sector Multi-collector ICPMS• Allows analysis of isotopic distribution

TRACE METAL ANALYSIS WITH ICP-MS

Recoveries of Extraction and Analysis of NIST SRMsUW-Madison ICP-MS

(Average and Standard Deviation of Multiple Extractions and Analyses)So

dium

(23)

Mag

nesi

um (2

4)

Alu

min

um (2

7)

Pota

ssiu

m (3

9)

Cal

cium

(40)

Van

adiu

m (5

1)

Chr

omiu

m (5

2)

Iron

(54)

Man

gane

se (5

5)

Cob

alt (

59)

Nic

kel (

60)

Cop

per (

63)

Zinc

(68)

Ars

enic

(75)

Rub

idiu

m (8

5)

Stro

ntiu

m (8

8)

Mol

ybde

num

(95)

Rho

dium

(103

)

Palla

dium

(108

)

Silv

er (1

09)

Cad

miu

m (1

11)

Tin

(120

)

Ant

imon

y (1

21)

Ces

ium

(133

)

Bar

ium

(137

)

Lant

hanu

m (1

39)

Cer

ium

(140

)

Haf

nium

(178

)

Tung

sten

(182

)

Plat

inum

(195

)

Gol

d (1

97)

Tha

llium

(205

)

Lead

(208

)

Ura

nium

(238

)

Rec

over

ies

(%)

0

20

40

60

80

100

120

140

NIST SRMs: San Joaquin Soil (NIST 2706), Used Auto Catalyst (NIST 2556) , and Urban Dust (NIST 1649a)

Metal Speciation

• Metal speciation has predominately been pursued to better understand toxicological effects

• Metal speciation is likely to aid source attribution efforts– Most metals present in minerals tend to have low

water solubilities– Many processed metals are more oxidized and have

higher solubilities• Metal speciation includes

– Valence state– Leachability– Isotopic analysis

• Need to consider potential impact of atmospheric processing

METAL SPECITAION

MOTOR VEHICLE EMISSIONS

TOTAL ROADWAY EMISSIONS

TUNNEL TESTS• 2 Tunnels in Milwaukee, WI• 12 Summer tunnel tests (total ~ 35,000 vehicles)

– 3 distinct sampling conditions • Courthouse Tunnel - Afternoon Rush Hour• Airport Tunnel - Weekday Rush Hours• Airport Tunnel - Weekend Traffic

• 6 Winter tunnel tests (total ~ 45,000 vehicles)– Consistent sampling conditions

• All Airport Tunnel• Weekdays, including afternoon rush hour

• Vehicle fleets characterized 3 ways:– DOT loops for basic vehicle counts– Videotaping for basic vehicle classification– Videotaping for license plate ID and detailed vehicle data

TUNNELS

AIRPORT• Howell Ave• 3 lanes in southbound direction• Similar to Van Nuys Tunnel (CA)

– Completely separate opposing bores• 770 feet long - No curvature• Constant speeds - very limited braking• ~8% truck traffic on weekdays• Not cleaned - noticeable road dust

COURTHOUSE• I-43 entrance ramp• 2 Lanes merge into one lane• Forced ventilation, exit at center• 1270 feet long - ~ 45 degree curve

– Inlet section: 715 feet - sample collection• Moderate braking and acceleration• ~2% truck traffic on weekdays • Minimal road dust

TUNNEL SAMPLING

• Samplers placed at the roadside at vehicle entrance and exit of tunnel– Upwind samples - 15 feet inside entrance– Inside Tunnel

• Courthouse: 50 feet upwind of exhaust outlet• Airport: 50 feet upwind of tunnel outlet

– Incremental concentration increase is vehicle emissions in the tunnel

• Dilution measurements with SF6: – SUMMA cans filled at inlet and inside tunnel in selected tests

– SF6 released at inlet, downwind of samplers

– Windspeed measured in all tests

– SF6 and windspeed used to calculate volumetric flow through tunnel for all tests

EMISSION RATES

• Increase in species concentration in the tunnel converted to emission rates

• Measured concentration increase: (mass species) / (m3)

• Multiplied by SF6-derived dilution rate: m3 air through tunnel

• Divided by tunnel distance and number of vehicles passing through the tunnel during the test:

total km driven

• To yield emission rates:(mass species) / (km driven)

TUNNEL EMISSIONS PROFILETUNNEL EMISSIONS PROFILE

Summer tunnel test data: average emission rates from on-road vehicles at Milwaukee, Wisconsin for PM10.

Error bars indicate standard errors.

Ele

men

tal C

Org

anic

CTot

al c

arbon

Lith

ium

Sod

ium

Po

tass

ium

Rubid

ium

Ces

ium

M

agnes

ium

Cal

cium

Str

ontium

Bar

ium

Alu

min

um

Sca

ndiu

m

Lanth

anum

Cer

ium

Titan

ium

Zirco

niu

m

Haf

niu

m

Cob

alt

Nic

kel

Cop

per

Zin

c Cad

miu

m

Lead

Silv

er

Thal

lium

Chro

miu

m

Man

gan

ese

Iron

Ruth

eniu

m

Rhod

ium

Pa

lladiu

m

Irid

ium

Pl

atin

um

Van

adiu

m

Ars

enic

M

olyb

den

um

Tungst

en

Ura

niu

m

Sili

con

Ger

man

ium

Sel

eniu

m

Tin

Antim

ony

Tel

lurium

Ph

osphor

os

Sulfur

Chlo

rine

Bro

min

e Io

din

e

Em

issi

on

Rate

(u

g/

km

)

10-1

100

101

102

103

104

105

106

Weekdays, Courthouse (Test #1 - #6, 2% Trucks) Weekdays, Airport (Test #8, #11 & #12, 6-10% Trucks)Weekends, Airport (Test #9, #10, 2% Trucks)

OC

EC

TUNNEL CONCLUSIONS

• Emissions in different seasons are similar, with exception of salt-impacted test– Several important elements emitted

• Pb, expected from tailpipe emissions of older vehicles• Ba, high in brake wear emissions • Ce, expected in tailpipe emissions• Noble platinum group metals from catalytic converters not

seen as high as expected

• Tunnel emissions are dominated by PM10– Low levels of PM2.5 in these tests

ROAD DUST

• Roadway dust is resuspended by passing vehicles

• Dirt was collected from the surface of the tunnel roadways– Ensured that samples were parallel to

dust resuspended by tunnel fleet– Vacuumed from road surface in

Summer and Winter tests– Dust was resuspended in a dilution

chamber and sampled • PM 2.5 and PM 10: Bulk chemistry and

trace metals

ROAD DUST

0

20

40

60

80

100

120

140

160

180

Summer Courthouse TunnelWinter Courthouse TunnelSummer Airport TunnelWinter Airport Tunnel

(mg

spec

ies)

/ (g

mas

s)

OC EC ChlorideNitrate SulfateAmmonium Na Mg Al K Ca Fe Ti0

20

40

60

80

100

120

140

160

180

PM10

PM2.5

(mg

spec

ies)

/ (g

mas

s)

Cl - NO3- SO4

2- NH4+

ROAD DUST

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Summer Courthouse TunnelWinter Courthouse TunnelSummer Airport TunnelWinter Airport Tunnel

(mg

spec

ies)

/ (g

mas

s)

V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

PM10

PM2.5

(mg

spec

ies)

/ (g

mas

s)

SALT-IMPACTED ROAD DUST

• High levels of road salt components were emitted from the roadway during one Winter Airport tunnel test

• Salt-impacted test was conducted on a cold, dry day, when the surface of the road had dried– Roadside snow had melted on several previous warmer days– Salt and slush dragged into tunnel by traffic– On the cold, dry day, no additional snow melt, and roadway dried

• Impact of road salt was determined by comparing elemental emission rates with similar winter tests

• Mole balance corroborates high emissions of road salt components NaCl, MgCl2, KCl, and CaCl2

• Profile is not pure salt, also shows components of sampled road dust from Airport Tunnel

SALT-IMPACTED ROAD DUST

OC EC chloride nitrate sulfateammonium Na Mg Al K Ca Fe Ti0

20

40

60

80

100

120

140

Average Winter TestSalt-Impacted Winter Test

(mg

spec

ies)

/ (k

m d

riven

)

OC EC ChlorideNitrate SulfateAmmonium Na Mg Al K Ca Fe Ti0

50

100

150

200

250

300

350

(mg

spec

ies)

/ (g

mas

s)

ROAD SALT PM10, calculated (mg species / g mass)

Cl - NO3- SO4

2- NH4+

Cl - NO3- SO4

2- NH4+

WINTER TUNNEL TESTS PM10 emission rates (mg species / km)

SALT-IMPACTED ROAD DUST

V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Average Winter TestSalt-Impacted Winter Test

(mg

spec

ies)

/ (k

m d

riven

)

V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

(mg

spec

ies)

/ (g

mas

s)

ROAD SALT PM10, calculated (mg species / g mass)

WINTER TUNNEL TESTS PM10 emission rates (mg species / km)

ROAD DUST DISCUSSION

• Dust collected from the tunnels is enriched with elemental emissions from other sources– Ba from brake wear– Pb from older vehicles’ tailpipe emissions, leaded gasoline

residues– Cr, Mn, Cu, Zn from brake wear, engine wear

• Salt applied to roads is an important factor in cold areas

• More road dust mass is PM10 than PM2.5

• Brake emissions are difficult to measure– Tunnel and dynamometer tests measure a mixture of emissions– Dynamometer tests are expensive, and access to dynos is limited– Specialized brake dynamometers are also expensive/limited

• Need method for estimating emissions composition– Relating emissions to brake pad composition would allow easy,

inexpensive characterization of a wide range of brake types

• Set of tests conducted to characterize brake wear emissions and compare: – Actual brake wear emissions from driving cycles – Chemical composition of the bulk brake pads– Chemical composition of brake housing dust

BRAKE WEAR

BRAKE WEAR EMISSIONS

• Actual brake wear emissions– Specialized Dynamometer: California Air Resources Board Facility

– RL-SHED: Running Loss - Sealed Housing for Evaporative Determinations

– Sealed environment, stainless steel, constant volume and temperature, engine inlet air piped in, exhaust piped out

– Emissions in the shed are:– Tire wear - Evaporative emissions– Brake wear - Resuspended dust

– VEHICLE tests• 2 vehicles with original brakes• 3 types of driving cycle, plus background tests• Vehicle test results corrected with blank/background tests

DYNAMOMETER

RUNNING LOSS SHED

SAMPLING SETUP

AIR INLET AND

EXHAUST

EMISSION RATES – selected abundant species, by testem

issi

on r

ate

(ug/

km)

0.01

0.1

1

10

100

1000

10000

100000

PM10

Ti V Cr Mn Fe Cu Zn Sr Mo Sb Ba

emis

sion

rat

e (u

g/km

)

0.01

0.1

1

10

100

1000

10000

100000

PM2.5

VEHICLE A VEHICLE B

FTP A1UC ASS AFTP A2

FTP B1UC B (NO PM10 DATA)SS BFTP B2 (NO PM10 DATA)

• Relationships between composition of brake wear emissions, brake pad composition, and brake housing dust– Brake pads removed from dyno test vehicles and pulverized– Brake housing dust also collected from dyno test vehicles– Both dusts resuspended and sampled for PM10 and PM2.5

BRAKE WEAR

Elemental emissions compared to dust and pad compositionsPM10: SHED tests, resuspended brake dust and pads from test vehicles

Fra

ctio

n m

easu

red

elem

enta

l mas

s

0.0

0.2

0.4

0.6

0.8

1.0Fe Na Mg Al K Ca Ti V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U

FTP UC SS FTP

HO

US

ING

D

US

T

CR

US

HE

D

PA

D

FTP SS

HO

US

ING

DU

ST

CR

US

HE

D

PA

D

VEHICLE 1 VEHICLE 2

BRAKE WEAR PROFILE

• Trace metal emissions are more similar to housing dust than to bulk pad composition– May reflect simultaneous grinding of the rotor, rivets, etc

• could explain increased contribution of Fe, Cu in housing dust

– Some species in the pad may be emitted in larger particle sizes (ie, larger than PM10)

• may explain decreased contribution of Ba in housing dust• may explain disappearance of Ca in vehicle B dyno emissions

• Developing emissions profile for brake wear– Brake housing dust and brake pads sampled from random

vehicles at local Wisconsin garages• Samples are parallel with WI tunnel emissions

– Dusts resuspended for size segregation and chemical analysis

BRAKE WEAR PROFILE

• Dust and pads from random vehicles show same composition trends as dyno test samples

• increased Fe and Cu, Zn in housing dust

• decreased Ba contribution on housing dust

• Housing dust samples averaged to create two profiles, as observed:

• HIGH levels of Ba, Cu, and other ‘trace’ metals in dust• LOW levels of ‘trace’ metals in housing dust

BRAKE DUST PROFILE

0

50

100

150

200

HIGH TRACE ELEMENT AVERAGELOW TRACE ELEMENT AVERAGE

(mg

spec

ies)

/ (g

mas

s)

OC EC chloridenitrate sulfateammoniumNa Mg Al K Ca Fe Ti0

200

400

600

800

(mg

spec

ies)

/ (g

mas

s)

Cl - NO3- SO4

2- NH4+

PM10

PM2.5

BRAKE DUST PROFILE

0

5

10

15

20

HIGH TRACE ELEMENT AVERAGELOW TRACE ELEMENT AVERAGE

(mg

spec

ies)

/ (g

mas

s)

V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U0

20

40

60

80

(mg

spec

ies)

/ (g

mas

s)

PM10

PM2.5

BRAKE WEAR DISCUSSION

• Housing dust is more similar to actual brake wear emissions

– Dust and pads from random vehicles show same composition trends as dyno test samples

– Dust from a range of vehicles can be sampled to estimate emissions

• Emissions from brake wear are seen in the absence of braking

– Tunnel tests and RL-SHED dynamometer test both showed brake wear from vehicles driven at 30mph with no braking

– PM2.5 emissions are seen during braking, but little PM2.5 is seen during non-braking cycles

– PM10 likely builds up in brake housing, and is emitted continuously during cycles with no braking

TIRE WEAR

• RL-SHED dynamometer test emissions also include tire wear

• Tires from dyno test vehicles sampled– Removed from vehicles– Abraded to obtain wear samples– Bulk chemistry and trace metals analysis

• Organic carbon dominates tire composition– Trace metal levels extremely low– Tires contribute negligible levels of

metals to brake and tire wear dynamometer test emissions

TIRE WEAR(m

g sp

ecie

s)/(

g m

ass)

0

200

400

600

800

Tire 1Tire 2Tire 3

PM10

OCECNaMg Al K Ca Fe Ti V Cr MnCu Zn As Rb Sr MoRu Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U

(mg

spec

ies)

/(g

mas

s)

0

200

400

600

800

PM2.5

TIRE WEAR DISCUSSION

• Tire wear contribution to metals levels is very low

– Emissions from tires are 70% organic carbon

– Zn is the only significant measured element (~0.1% by mass)

– Brake wear dominates metals emissions in the SHED

TAILPIPE EMISSIONS

• Tailpipe emissions include– Unburned fuel - Partially combusted fuel– Lube oil - Engine wear

• Series of dynamometer tests done to investigate tailpipe emissions– More than 100 vehicles, both gasoline and diesel– Several different driving cycles tested on chassis dyno– Emissions collected with dilution source sampler

• Allows exhaust to cool and condense before collecting particles

• Fuel and lube oil samples collected from test vehicles– Determine impact of lube oil and engine wear on tailpipe

emissions of trace metals

TAILPIPE DISCUSSION

• Elemental content of emissions varies between vehicles

• Zn, Cu, and Fe generally have highest emissions– Expected to be from engine wear, seen in used motor oil

• Some vehicles emit high Ca– Also emit high Zn, and significant Fe and Cu, possibly indicating

combustion of lube oil that contains engine wear products– Ca is consistently high in all used lubricating oil samples– Agrees with other studies

• Ce is detected in tailpipe emissions – significant primarily in gasoline-powered vehicles

VEHICLE EMISSIONS APPORTIONMENT

• PM2.5 is relatively low in the tunnels, PM10 is high– Bulk of road dust in the tunnels is PM10– Brake wear emitted during driving without braking is mainly PM10

• Basic, visual comparison of tunnel profiles with source profiles shows:– Roadway metal emissions from motor vehicles are dominated by

brake wear and resuspended road dust– Resuspension of road salt is a very significant source of PM from

roadways at some times

IMPACTS ON AMBIENT ATMOSPHERE

• Ambient sampling network - Source apportionment– Determine the contribution of motor vehicles to total ambient PM in

the urban atmosphere– Powerful molecular marker techniques have been developed with

organic compounds as source tracers– The broad spectrum of metals that can be analyzed with ICP-MS

provides many potential new tracers– Combination of metals data with organics data will allow very

robust and accurate source reconciliation

• 3 sampling sites:• Milwaukee, WI - urban• Waukesha, WI - industrial• Denver, CO – industrial/residential

• EPA 6th day sampling schedule • February 2001 – January 2002

• PM 2.5 and PM 10• Same sampler design as source tests at each location• Chemical analyses parallel to source tests

AMBIENT SAMPLING

AMBIENT PM10 COMPOSITION

DENVER PM10ug

/m3

0

5

10

15

20

25

1.4*OC EC CHLORIDE NITRATE SULFATE AMMONIUM SUM OXIDE MASS

MILWAUKEE PM10

ug/m

3

0

5

10

15

20

25

WAUKESHA PM10

ug/m

3

0

5

10

15

20

25

FebruaryMarch2001

AprilMayJune2001

JulyAugust

September2001

OctoberNovemberDecember

2001

January 2002

AMBIENT PM2.5 COMPOSITION

DENVER PM2.5

ug/m

3

02468

10121416

1.4*OC EC CHLORIDE NITRATE SULFATE AMMONIUM

MILWAUKEE PM2.5

ug/m

3

0

5

10

15

20

25

WAUKESHA PM2.5

ug/m

3

0

5

10

15

20

25

FebruaryMarch2001

AprilMayJune2001

JulyAugust

September2001

OctoberNovemberDecember

2001

January 2002

BULK CHEMISTRYPM10: 3 Ambient sites 2001

ug/m

3

0

2

4

6

8

10

12FEBRUARY-MARCHAPRIL-MAY-JUNEJULY-AUGUST-SEPTEMBEROCTOBER-NOVEMBER-DECEMBERJANUARY

ug/m

3

0

1

2

3

4

5

6

7

OC EC CHLORIDE NITRATE SULFATE AMMONIUM

ug/m

3

0

1

2

3

4

5

6

7

DENVER PM10

MILWAUKEE PM10

WAUKESHA PM10

CRUSTAL ELEMENTSPM10: 3 ambient sites 2001

ug/m

3

0.0

0.2

0.4

0.6

0.8

1.0ug

/m3

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7FEB-MARAPR-MAY-JUNJUL-AUG-SEPOCT-NOV-DECJAN

Na Mg Al K Ca Fe

ug/m

3

0.0

0.1

0.2

0.3

0.4

0.5

0.6

DENVER PM10

MILWAUKEE PM10

WAUKESHA PM10

TRACE ELEMENTSPM10: 3 Ambient sites 2001

ug/m

3

0.00

0.02

0.04

0.06

0.08

0.10

ug/m

3

0.000

0.005

0.010

0.015

0.020

0.025FEB-MARAPR-MAY-JUNJUL-AUG-SEPOCT-NOV-DECJAN

Ti V Cr Mn Cu Zn As Rb Sr Mo Ru Rh Pd Ag Cd Sb Cs Ba Ce W Pt Tl Pb U

ug/m

3

0.00

0.01

0.02

0.03

0.04

0.05

DENVER PM10

MILWAUKEE PM10

WAUKESHA PM10

Ba v CuPM10 at 3 ambient sites

Copper (ng/m3)

0 20 40 60 80

Ba

(ng/

m3)

0

10

20

30

40

50

60

70

Denver BaMilwaukee BaWaukesha Ba

Sr v. CuPM10 at 3 ambient sites

Copper (ng/m3)

0 10 20 30 40 50 60 70 80

Sr

(ng/

m3)

0

5

10

15

20

Denver SrMilwaukee SrWaukesha Sr

Ti v FePM10 at 3 ambient sites

Iron (ng/m3)

0 500 1000 1500 2000 2500 3000 3500

Ti (

ng/m

3)

0

50

100

150

200

250

Denver TiMilwaukee TiWaukesha Ti

Waukesha v Milwaukee2001 Ambient PM10

Milwaukee (ug/m3)0 10 20 30 40 50 60 70

Wau

kesh

a (u

g/m

3)

0

10

20

30

40

50

60

70

Milwaukee (ug/m3)0 2 4 6 8 10 12

Wau

kesh

a (u

g/m

3)

0

2

4

6

8

10

12

Milwaukee (ug/m3)0.0 0.5 1.0 1.5

Wau

kesh

a (u

g/m

3)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

MASS (Measured) ORGANIC CARBON

ELEMENTAL CARBON

Milwaukee (ug/m3)0 5 10 15 20

Wau

kesh

a (u

g/m

3)

0

5

10

15

20

SULFATE

Waukesha v MilwaukeePM10 Copper and Lead

Milwaukee (ng/m3)0 20 40 60 80 100

Wau

kesh

a (n

g/m

3)

0

20

40

60

80

100

Copper

Milwaukee (ng/m3)0 5 10 15 20 25

Wau

kesh

a (n

g/m

3)

0

5

10

15

20

25

Lead

AMBIENT DISCUSSION

• Metals emitted from motor vehicles are detected in ambient atmosphere

• Differences in regional sources at three ambient sites are apparent– Crustal materials, sources, altitude, vehicle operation

• Wisconsin sites – Data agree well for regionally-influenced species– Contribution of local sources is apparent– Chloride contribution to PM10 obvious in Winter months

Metal Speciation

• Dissolution of particles deposited in the lungs is a likely pathway to bioavailability

• Ultimate impact of PM on organisms may be more directly tied to the leachable fraction of metals than to the total metals concentration

• Leachability of trace metals in a biologically relevant fluid is sought

• Investigating leachability of trace metals in a biologically relevant fluid requires analytical methods that can handle the complex inorganic and organic matrix of a synthetic lung fluid

SYNTHETIC LUNG FLUID

• Natural extracellular lung fluid – coats the alveoli and walls of the lungs

– is a buffered solution of salts with surfactants

• Two synthetic matrices used for extractions– Gamble Solution: buffered salt solution

– Surfactant-Like Solution: Gamble solution plus an organic surfactant to more closely mimic the environment in the lungs

• Duplicate metals samples analyzed– analysis of complete digestion compared to the leachable fraction

of the metals in synthetic lung fluid

CHEMICAL COMPOSITION

• Sodium 3350 mg/liter• Potassium 160 mg/l• Calcium 100 mg/l• Magnesium 25 mg/l

• Chloride 4050 mg/l• Bicarbonate 1900 mg/l• Phosphate 100 mg/l• Sulfate 50 mg/l

• Albumin 25 mg/l• DPPC 730 mg/l• Cetyl Alcohol 75 mg/l• Tyloxapol 50 mg/l

• The DPPC, Cetyl Alcohol, and the Tyloxaphol are modeled after Exosurf Neonatal, a commercial lung surfactant for newborn babies

GAMBLE SOLUTIONINORGANIC SALTS

SURFACTANT ORGANIC COMPONENTS

Gamble solution and surfactant are combined to closely mimic the composition of fluid in the lungs

METHOD: LEACHING

DUPLICATE SAMPLES

COLLECTED

DIGESTION AND ICP-MS ANALYSIS

FOR TOTAL METALS

DUPLICATE SAMPLE OR SRM

GAMBLE SOLUTION PREPARED

CLEANED WITH

CHELEX COLUMN

BLANK TESTED ON

ICP-MS

COMBINED AND SHAKEN

IN 37oC WATER BATH

LEACHATE FILTERED

FILTRATE ANALYZED

UNDIGESTED BY ICP-MS

LEACHED FRACTION

COMPARED TO TOTAL METALS

LEACHING SAMPLES

• NIST Standard Reference Materials (SRM’s)– Used Auto Catalyst– Urban Dust

• Total Roadway Emissions from Motor Vehicles– PM10 samples from vehicle tunnel sampling

• Tire and Brake Wear emissions– PM10 samples from tire and brake wear dynamometer

testing

RESULTS

Solubility of Trace Metals of Two Standard Reference Materials Urban Dust (NIST1649a) and Used Auto Catalyst (NIST2556)

in Simulated Lung FluidA

lum

inum

Tita

nium Iron

Man

gane

seC

oppe

rZ

inc

Rub

idiu

mZ

irco

nium

Neo

bium

Rho

dium

Silv

erC

adm

ium

Tin

Ant

imon

yT

ellu

rium

Ces

ium

Bar

ium

Lan

than

umC

eriu

mH

afni

umT

ungs

ten

Lea

d

Lea

chab

le F

ract

ion

(%)

0

20

40

60

80

100

Urban Dust Test #1 Urban Dust Test #2 Auto Catalyst Test #1 Auto Catalyst Test #2

RESULTS

S o lubil i ty of PM 10 Trace M etals in S imulated Lung Flu idM oto V e h icle E m issio n Tests:

B r a k e a n d T ire W ear and R o a d w a y T u n n e l Em issionA

lum

inum

Tita

nium Iron

Man

gane

se

Cop

per

Zin

c

Ant

imon

y

Bar

ium

Cer

ium

Lea

d

Lea

chab

le F

ract

ion

(%)

0

20

40

60

80

100

B rake and Tire Wear Tes t #1B rake and Tire Wear Tes t #2R o a d w a y T u n n e l Emiss ions Tes t #1R o a d w a y T u n n e l Emiss ions Tes t #2R o a d w a y T u n n e l Emiss ions Tes t #3R o a d w a y T u n n e l Emiss ions Tes t #4

CURRENT WORK

• CHEMICAL MASS BALANCE MODEL– Apportion specific sources of roadway emissions to total

emissions from motor vehicles• Roadway profile and specific sources have been fully characterized• Use of parallel sampling and analysis techniques allows all profiles

to be directly compared• CMB relates source profiles to overall profile to determine

contributions of each

• FACTOR ANALYSIS OF AMBIENT DATA– Exploratory factor analysis

• All sources of ambient PM are not fully defined

• Parallel sampling and analyses available only for motor vehicle emissions

• EFA can apportion ambient concentrations to various factors

• Factors can be related to known sources

ACKNOWLEDGEMENTS

• Funded by Health Effects Institute (HEI#99-13)• California Air Resources Board Haagen-Smit Laboratory staff

• Tailpipe emissions tests funded by NREL• US EPA and Eastern Research Group

WI State Laboratory of HygieneJeff DeMinter, John Strauss, Al Spallato, Mike Arndt, Chris Worley, Dustan Helmer

Dr. June Soo Park, Min-Suk Bae, Rebecca Sheesley, Charles Christensen, James Bucholz, M Danijarsa

Dr. Mike Hannigan, Sara Bruening, Clifton Hackbarth