DAQ-2025-001243

Particulate Chloride
in the Urban Environment
Dr. Kevin Perry
Department of Atmospheric Sciences
University of Utah
Presentation for the Utah Division of Air Quality
January 9, 2025
Outline
•Background/Motivation
•Project Goals
•Research Results
•Conclusions
What are Halogens?
•F and Cl are gases at room temperature/Br is liquid/I, At, and Ts are solids
•Halogens are very electronegative, meaning they strongly attract electrons
from other elements in chemical reactions.
•Due to their reactivity, Halogens are important in a wide variety of chemical
transformation mechanisms.
•Results from NOAA’s Utah Winter Fine
Particulate Study (UWFPS)
•Modeled wintertime PM formation in northern
Utah
•Confirmed the importance of halogens in PM
formation
•Demonstrated that U.S. Magnesium halogen
emissions were responsible for 10 - 25% of
regional PM2.5 during winter pollution episodes
•Noted that significant coarse-mode particulate
chloride, of unknown origin, might exist in the
region
•Coarse-mode particulate chloride should be
included in future modeling efforts if confirmed
Why is Coarse-mode Chloride
Important?
•Chloride is a key player in halogen chemistry
•Coarse-mode chloride can contribute to the release of
reactive chlorine species, which can break down ozone
molecules
•Chloride-containing particles can serve as nuclei for the
formation of secondary aerosol particles
•Acid-base chemistry can be altered by the presence of
coarse-mode chloride affecting aerosol pH
•Non-sea-salt sources of coarse-mode chloride are not
generally included in atmospheric chemical transport models
(Dr. Jennifer Haskins – exception)
Why Coarse-mode Chloride Uncertainty?
PM10 Filter Sites PM2.5 PM10
Bountiful XRF ICP-MS (but not Cl due to
polyatomic interferences)
Hawthorne XRF Collected but not routinely
analyzed
Lindon XRF Collected but not routinely
analyzed
Utah Technical Center Collected but not routinely
analyzed
Collected but not routinely
analyzed
Roosevelt Collected but not routinely
analyzed
Collected but not routinely
analyzed
Prison Site Collected but not routinely
analyzed
Collected but not routinely
analyzed
Lake Park Collected but not routinely
analyzed
Collected but not routinely
analyzed
Brigham City Collected but not routinely
analyzed
Collected but not routinely
analyzed
No PM10 Chloride measurements exist prior to this project!
Project Goals
1)Analyze 225 PM10 filters from the Hawthorne and Lindon
PM10 monitoring sites to determine the prevalence and
temporal variations of coarse-mode chloride
2)Determine the chemical composition of road salt used by
the Utah Department of Transportation
3)Use existing size-resolved elemental composition
measurements to determine the temporal variation of
particulate chloride particle size distributions
4)Perform source apportionment modeling of
coarse-particulate chloride
Research Results
•Coarse-mode Particulate Chloride
•Road Salt Composition
•Particulate Chloride Size Distributions
•Sources of Coarse-mode Particulate Chloride
Analytical Technique Options
X-Ray Fluorescence (XRF)Inductively-Coupled Plasma Mass
Spectrometer (ICP-MS)
Proton-Induced X-ray Emission (PIXE)
Pe
a
k

s
e
n
s
i
t
i
v
i
t
y
Pe
a
k

s
e
n
s
i
t
i
v
i
t
y
Proton-Induced X-ray Emission
(PIXE)
•Non-destructive technique
•Highest sensitivity is for S and Cl
•Can also quantify Na which is not possible with XRF
•PIXE was performed at Crocker Nuclear Laboratory
(UC-Davis)
•Same lab where PM2.5 XRF measurements are made
for the Hawthorne and Lindon sites
•QA performed by UC-Davis by comparing PM10 and
PM2.5 measurements
PIXE Detection Rates
Cl detection rate = 93%
PM2.5 vs PM10 Chloride
Avg. PM10 Cl = 371.4 ng/m3
Avg. PM2.5 Cl = 161.7 ng/m3
Δ = 209.7 ng/m3
30% more coarse-mode Cl
than PM2.5 Cl
County Comparison
PM2.5 vs PM10 Chloride
(2019)
Site Average
PM10 Cl
Average
PM2.5 Cl
Average
Coarse PM Cl
Hawthorne
(Salt Lake County)
548.0 ng/m3 216.5 ng/m3 331.5 ng/m3
Lindon
(Utah County)
151.7 ng/m3 118.8 ng/m3 32.9 ng/m3
There was an order of magnitude more coarse PM
chloride in Salt Lake County vs Utah County in 2019
Seasonality of Coarse-mode Chloride
(Salt Lake County)
48 filters 53 filters 30 filters 29 filters
January 2019 through July 2020
Research Results
•Coarse-mode Particulate Chloride
•Road Salt Composition
•Particulate Chloride Size Distributions
•Sources of Coarse-mode Particulate Chloride
UDOT Road Salt Vendors
Road salt samples from each company were obtained directly from UDOT and
subsequently analyzed for major and trace elements using ICP-MS
Road Salt Major Constituents
Broken Arrow Compass Morton
NaCl 96.0%97.4%97.7%
SO4 1.4%0.9%1.2%
MgCl2 1.2%0.9%0.1%
CaCl2 0.5%0.4%0.9%
KCl 0.9%0.4%0.1%
All others <0.01%<0.02%<0.01%
ICP-MS analysis performed at the University of Utah Metals Lab
(Department of Geology & Geophysics – Dr. Diego Fernandez)
Road Salt Trace Species
(Average)
Soil
Evaporites Mostly Rare EarthsMostly Industrial
Research Results
•Coarse-mode Particulate Chloride
•Road Salt Composition
•Particulate Chloride Size Distributions
•Sources of Coarse-mode Particulate Chloride
Atmospheric Mercury Monitoring Site
(UT96)
Size- and Time-Resolved
PM Composition Measurements
8-stage rotating drum impactor
10 to 5 μm
0.75 to 0.56 μm
5.0 to 2.5 μm
2.5 to 1.1 μm
0.34 to 0.26 μm
0.26 to 0.09 μm
1.1 to 0.75 μm
0.56 to 0.34 μm
4 weeks
8-Stage Rotating Drum Impactor
•Greased mylar substrates to minimize particle bounce
•Analyzed using synchrotron X-Ray Fluorescence (S-XRF) at
Lawrence Berkeley National Laboratory
•Provides size-resolved elemental concentration
measurements with 3-hr resolution
•Deployed at UT96 site from September 2009 through
December 2010
September 2009
89% of the Chloride
mass is contained in
particles with
diameters > 1 μm
Unimodal
February 2010
62% of the Chloride
mass is contained in
particles with
diameters < 1 μm
Bi-modal
What Have We Learned About
Particulate Chloride?
•Significant coarse-mode chloride is present in
northern UT
•Coarse-mode chloride peaks in winter (DJF)
•Winter-time chloride is bimodal, soil chloride is
unimodal
•There is significantly more coarse-mode chloride in
Salt Lake County vs Utah County
•UDOT road salt is primarily NaCl with smaller
amounts of CaCl2, MgCl2, and KCl
Research Results
•Coarse-mode Particulate Chloride
•Road Salt Composition
•Particulate Chloride Size Distributions
•Sources of Coarse-mode Particulate Chloride
Potential PM Chloride Sources
•Sea-salt (GSL)
•Playa dust (GSL)
•Road salt (GSL – sourced)
•Coal combustion
•Biomass burning
•Vehicle emissions (fuel additives)
•Municipal incineration
•Agricultural fertilizers
•Industrial processes
•Petroleum refining
•Metal smelting
•Cement production
•Magnesium production (electrolysis)
•Volcanic emissions
PM10 Pearson Correlation Matrix (R)
=R so R2 = 0.52
Slightly more than half of the PM10
Cl variance is explained by PM10 Na
PM10 Na vs PM10 Cl Comparison
NaCl stoichiometry
Excess Cl
indicates
sources
beyond NaCl
PM10 Soil vs PM10 Cl Comparison
Overall no correlation with soilSoil Chloride
Source?
Non-soil Chloride
Sources?
Analysis of Top 10% of PM10 Cl
•Average temperatures
•Days since last precipitation
•Average windspeed
Evaluate the meteorological
conditions on the highest PM10
Cl days to determine if road
salt is a viable source
Date PM10 Cl
(ng/m3)
1/2/2019 2321
1/3/2019 1349
1/4/2019 1418
1/5/2019 1975
1/25/2019 1721
2/8/2019 4868
2/9/2019 2757
2/10/2019 1500
2/11/2019 1921
2/12/2019 4448
2/27/2019 1982
12/2/2019 2413
12/5/2019 3672
12/17/2019 2460
12/20/2019 3572
12/22/2019 1603
2/3 of the highest PM10
Cl days had average
temperatures below 32F
Average Temperatures on the Highest PM10 Cl Days
Precipitation on the Highest PM10 Cl Days
87% of the days had no precipitation
or only a trace
0 1 2 3 4 5 6
Meteorological Summary
•87% of high PM10 Chloride days (i.e., top 10%) had
average daily temperatures of 36F or colder.
•75% of high PM10 Chloride days had no
precipitation within the last 24 hours.
•69% of high PM10 Chloride days had precipitation
within the last 2 days.
•All high PM10 Chloride days had precipitation
within the last week.
2019 and 2020 were Low Cl Years
Conclusions
•52% of the PM10 chloride variance is explained by the
aerosolization of road salt, which is most likely to occur during
the first few dry days following a snowstorm
•PM10 chloride is not correlated with soil
•Positive Matrix Factorization (PMF) failed to complete a source
apportionment because PM10 Cl was only correlate with a single
other element
•The results presented here likely underestimate typical coarse
particulate chloride concentrations
•Coarse-mode chloride sources need to be investigated further
and included in future atmospheric chemical transport models
Questions?