US Department of Energy
US Department of Energy Publications
University of Nebraska - LincolnYear
Characterization of Aerosols Containing
Zn, Pb, and Cl from an Industrial Region
of Mexico City
Ryan C. Moffet, University of California
Yury Desyaterik, Pacific Northwest National Laboratory
Rebecca J. Hopkins, Lawrence Berkeley National Laboratory
Alexei V. Tivanski, Lawrence Berkeley National Laboratory
Mary K. Gilles, Lawrence Berkeley National Laboratory
Y. Wang, University of California
V. Shuthanandan, Pacific Northwest National Laboratory
Luisa T. Molina, Molina Center for Energy and the Environment
Rodrigo Gonzalez Abraham, Molina Center for Energy and the En-
Kirsten S. Johnson, Massachusetts Institute of Technology
Violeta Mugica, Universidad Auto´noma Metropolitana-Azcapotzalco
Mario J. Molina, University of California
Alexander Laskin, Pacific Northwest National Laboratory
Kimberly A. Prather, University of California
This paper is posted at DigitalCommons@University of Nebraska - Lincoln.
Characterization of Aerosols
Containing Zn, Pb, and Cl from an
Industrial Region of Mexico City
R Y A N C . M O F F E T ,† , §Y U R Y D E S Y A T E R I K ,‡
R E B E C C A J . H O P K I N S ,§
A L E X E I V . T I V A N S K I ,§M A R Y K . G I L L E S ,§
Y . W A N G ,†V . S H U T T H A N A N D A N ,‡
L U I S A T . M O L I N A ,| , ⊥
R O D R I G O G O N Z A L E Z A B R A H A M ,|
K I R S T E N S . J O H N S O N ,⊥
V I O L E T A M U G I C A ,#M A R I O J . M O L I N A ,†
A L E X A N D E R L A S K I N , *, ‡A N D
K I M B E R L Y A . P R A T H E R *, †
Department of Chemistry and Biochemistry, University of
California, San Diego, California 92093-0314, W.R. Wiley
Environmental Molecular Sciences Laboratory, Pacific
Northwest National Laboratory, Richland, Washington 99352,
Chemical Sciences Division, Lawrence Berkeley National
Laboratory, Berkeley, California 94720-8226, Molina Center
for Energy and the Environment (MCE2), La Jolla, CA,
Department of Chemistry, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139-4307, and
Applied Chemistry, Universidad Autónoma
Metropolitana-Azcapotzalco, Av. San Pablo 180,
México D.F. 02200
Received December 6, 2007. Revised manuscript received
June 16, 2008. Accepted July 21, 2008.
Recent ice core measurements show lead concentrations
increasing since 1970, suggesting new nonautomobile-related
sources of Pb are becoming important worldwide (1).
Developing a full understanding of the major sources of Pb
internally mixed with Zn, Pb, Cl, and P. Pb concentrations
were as high as 1.14 µg/m3in PM10and 0.76 µg/m3in PM2.5.
Real time measurements were used to select time periods of
interest to perform offline analysis to obtain detailed aerosol
were found to be composed of ZnO and/or Zn(NO3)2·6H2O.
The internally mixed Pb-Zn-Cl particles represented as much
as 73% of the fine mode particles (by number) in the morning
hours between 2-5 am. The Pb-Zn-Cl particles were
primarily in the submicrometer size range and typically mixed
with elemental carbon suggesting a combustion source.
The unique single particle chemical associations measured in
this study closely match signatures indicative of waste
incineration. Our findings also show these industrial emissions
play an important role in heterogeneous processing of NOy
emitted by the same source underwent heterogeneous
transformations into nitrate particles as soon as photochemical
production of nitric acid began each day at ∼7 am.
(2). Although relatively little is known about the specific
chemical constituents responsible for the adverse health
of studies (3–5). The solubility of metal ions present in
oxidation state greatly affects their toxicity (5). Other factors
respiratory tract a particle may travel. Smaller particles with
they are more likely to be retained by the body (6–8).
Anthropogenic particles created by high temperature pro-
cesses (e.g., combustion and ore processing) possess many
of the properties responsible for adverse health effects.
In urban areas, anthropogenic sources of submicron
metal-containing particles are plentiful. For example, the
burning of fossil fuel leads to the association of Ni and V
within particles (9). Prior to 2000, tetra-ethyl-lead was used
related emissions of submicron lead particles (10). In
in heavy metals (11). Combustion of municipal waste pro-
duces submicron particles composed of Zn, Pb, and Cl as
well as numerous other metals (12). Zn and Pb are often
geographic locations (10). Dust particles are predominantly
supermicrometer sizes and when stirred up through an-
compositions which can ideally be identified using the
appropriate combination of analytical techniques.
To characterize metal-containing particles in the Mexico
decade, several field studies in MCMA have employed a
variety of instrumentation to address particulate pollution
part of the city is characterized by high concentrations of
industrial emissions, in particular, Zn- and Pb-containing
particles (14, 17, 18). Notably, blood Pb levels in children in
the northeastern part of the city were 10.9% higher than
concentrations in the northern MCMA have been generally
attributed to industry, and the exact identity has not been
(MILAGRO) field study held in March 2006 (see http://
mce2.org/). As part of the MILAGRO study, a sampling site
(T0) was established in northern MCMA and was impacted
by industrial, vehicle, and residential emission sources (20).
Single particle mass spectrometry was used to identify a
unique period with a high abundance of Zn and Pb particles
on March 24, 2005. Detailed studies using complementary
speciation, size, mixing state, and morphology of metal-
* Address correspondence to either author. Phone: 858-822-
5312 (K.A.P.); 509-371-6129 (A.L.). E-mail: email@example.com (K.A.P.);
†University of California.
‡Pacific Northwest National Laboratory.
§Lawrence Berkeley National Laboratory.
|Molina Center for Energy and the Environment.
⊥Massachusetts Institute of Technology.
#Universidad Auto ´noma Metropolitana-Azcapotzalco.
Environ. Sci. Technol. 2008, 42, 7091–7097
10.1021/es7030483 CCC: $40.75
Published on Web 08/29/2008
2008 American Chemical SocietyVOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY97091
This article is a U.S. government work, and is not subject to copyright in the United States.
containing particles along with quantitative mass loadings.
Spatial, temporal, physical, and chemical information on
these particles suggests they were produced by local waste
not only the concentration and source of these industrial
emissions, but also insight into the atmospheric processing
these particles undergo after they are emitted. Microscopy
and spectro-microscopic techniques provide morphology
the potential environmental effects of these particles.
Sampling site T0 was located in the northern part of MCMA
N, 99°08′55.60W and 2200 m above sea level. The site was
surrounded by industrial and residential areas as well as
techniques were used in this study. The aerosol time-of-
flight mass spectrometer (ATOFMS) measures real-time
single particle size and composition with hourly time
resolution. Computer controlled scanning electron micros-
copy with energy dispersive X-ray analysis (CCSEM/EDX)
provides quantitative information on thousands of the
substrate collected particles. Scanning transmission X-ray
microscopy with near edge X-ray spectroscopy (STXM/
resolved bulk mass concentrations of Z > 23 elements with
a 6 h time resolution. A more detailed description of these
techniques is presented in the Supporting Information.
Results and Discussion
Historically, the high concentrations of Zn and Pb particles
in the northern MCMA have broadly been attributed to
the ATOFMS and PIXE data presented in Figure 1 show
episodes of particles containing Pb and Zn were common
PM2.5size ranges determined with high volume filters. The
highest PM10levels for Zn and Pb were 1.4 and 1.14 µg/m3,
respectively, with the majority of the Pb mass in PM2.5. The
episodes of Pb and Zn particles usually occurred between
originated from the industrial region of the city, north of the
of the high Zn and/or Pb concentration to industrial emis-
sions (14, 17, 18). The source assignment of industrial
emissions is supported by the low concentrations of these
particles during the holiday weekend of March 18-20 as
highlighted in Figure 1.
On March 24, 5:15-5:30 CST the ATOFMS measured a
peak in the percentage of particles containing Zn and Pb as
indicated in Figure 1. This paper focuses on this episode,
serving as a representative event that occurred nearly every
that significantly decreased the background aerosol con-
showed that following the heavy rain event, the number
percentage of particles containing Pb or Zn was 61% (by
into those containing metals (Me > 0) or no metals (Me )
0). The metal-containing particles were further separated
into two classes; those containing Zn and/or Pb (Me > 0,
ZnPb > 0), and those without Zn or Pb (Me > 0, Pb ) 0, Zn
FIGURE 1. Time series of unscaled ATOFMS counts of Zn- and/or Pb-containing particles compared with the PIXE bulk mass
concentrations of Zn and Pb. The major and minor tick marks occur at 12 am and 12 pm, respectively. PIXE data are shown for
0.36-2.5 µm particles.
TABLE 1. Twenty-four Hour Average PM10and PM2.5(µg/m3) for
Pb and Zn Determined between March 6 and 30
7092 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
) 0). The label “max” indicates that those particular metals
made up the maximum signal out of all of the elements
The percentages shown in black and blue in Figure 2
represent the relative number of particles classified for the
contained Pb or Zn, leaving a PbZn ) 0 class mixed with
other metals including Na, Pb, Zn, K, Cu, Cr, and Cd; 3 and
12% of the Me > 0 particles contained the health relevant
metals Cd and Cr. The two data sets were in excellent
agreement for the Na-max, Fe-max, and Mg-max particle
classes. In general, the CCSEM/EDX and ATOFMS results
agreed except for the higher percentages of the K max and
sensitivity of the ATOFMS toward these metals. In fact, 80%
of the Pb max particles were internally mixed with Zn, so it
is likely that many Zn-rich particles detected by the CCSEM
are actually contained in the ATOFMS Pb max class.
Particle Mixing State and Source Profiles. The chemical
mixing state of a particle describes which chemical species
from ATOFMS and CCSEM are presented in the Supporting
ATOFMS and CCSEM/EDX showed that ZnPb > 0 particles
are internally mixed with Cl, Na, K, P, S, and a multitude of
other elements. This mixing state can be directly compared
to source mixing states measured in previous source char-
and municipal waste incinerators and showed that these
exception of P and Cl which were more abundant in
incineration (21). Figure 3 compares the Me > 0 particles
incineration and copper smelting from Tan et al. (21). The
Me > 0 and nonferrous smelter particles did have sulfur at
m/z ) -32 and -64 in common, a marker not reported for
incineration by Tan et al. However, the dot product of the
percent contribution vectors shown in Figure 3 is 0.88
between incineration and Me > 0, while it is 0.53 between
Me > 0 and nonferrous smelting. The best match with the
Me > 0 particles from Mexico City is with incineration, with
the key elemental tracers P and Cl both abundantly present
in each sample.
The mixing state analysis is further supported by PIXE
to extract an industrial factor showing strong contributions
Table S2, the strongest associations within the industrial
factor were, from weakest to strongest, between Cl, Na, Zn,
Pb, Cr, and P. This was similar to the mixing state identified
by the ATOFMS and CCSEM/EDX. Furthermore, many of
from smelting in the work of Tan et al. (21). However, the
nitrogen and carbon. Thus, as described below, other
also be used to obtain more detailed chemical composition.
ATOFMS shows that the Zn and Pb particles were found to
be mixed with carbonaceous species such as elemental
carbon (EC), aromatics, other metals, and organic species
such as the oxalate ion. The ATOFMS spectra showed that
at least 40% of the Pb and Zn particles were mixed with
soot/EC supporting the assertion that they were generated
from combustion. In Supporting Information Figure S2, it is
shown that the nonmetal (Me ) 0) class present in this
industrial plume had a distinct aromatic hydrocarbon
signature with 75% of these particles containing aromatic
13-15% of the Pb- and Zn-containing particles were found
other species (including Mn and Cu) is further quantified in
the Supporting Information (Figures S1-S4).
Particle Morphology. Morphology is an important mi-
possible health effects (2), speciation (3), phase, formation
mechanism, and source identification. The SEM images
shown in Figure 4 illustrate the morphology of metal rich
and Pb-containing particles (panels A, B, and C), as well as
spherical Zn particles containing PbFe and Mn (not shown)
were detected. Figure 4a shows Zn-containing particles
results indicating that at least 40% of the Zn-containing
particles were mixed with elemental carbon. SEM/EDX
analysis confirmed that the needle like structures shown in
amounts of other elements suggesting a similarity to previ-
ously reported tetrahedral ZnO particles (22). ATOFMS op-
tical measurements also indicated that the particles contain-
ing Zn and Pb were nonspherical with effective densities
that exceeded 2.4 g/cm3(23). A high abundance of cubic
FIGURE 2. Rule based classification from ATOFMS and CCSEM/
EDX data. Percent of particles in a class out of the total number of
analyzed particles is indicated in black (ATOFMS) and blue
(CCSEM/EDX). Pb or Zn in particles is indicated by ZnPb > 0. The
size range analyzed for both instruments was approximately 0.2-
FIGURE 3. Comparison of MCMA aerosol on Mar 24 from 5:15-
5:30 am with two industrial sources. The color is proportional to
the percentage of particles in the sample given on the vertical
axis containing a particular marker indicated by the horizontal
axis; this percentage is also indicated inside of each box. Mexico
City metal aerosol more closely matches waste incineration than
smelting. Key elemental tracers include Pb, Zn, Cl, and P. The
ATOFMS data used for this figure is the “Me > 0” particle set
defined in Supporting Information Table S1. (* Source data ob-
tained from Tan, 2002 (21).)
VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7093
NaCl particles were also observed to be internally and
externally mixed in the same samples as the metal rich
particles, indicating they were likely produced by the same
Using STXM/NEXAFS, two major particle classes were
distinguished from one another based on morphology and
NEXAFS spectra. Representative STXM images are shown in
Figure 5. Particles shown in Figure 5a (type A) were found
to be internal mixtures of Zn, EC, and K, whereas particles
discussed previously (22). The presence of EC and K in type
A was confirmed with the corresponding carbon K-edge
To elucidate the chemical speciation of Zn-containing
particles, the Zn L-edge NEXAFS spectra of ZnO, ZnS,
Zn(NO3)2·6H2O, ZnCl2, ZnSO4standard reference materials
were measured for comparison with particles of types A and
are shown in Figure 5. The spectra in Figure 5a and b
particles located in different regions of the substrate. ZnS,
A and B, provide chemical identification of these two types.
A and Zn(NO3)2·6H2O are observed at ∼1025 eV and ∼1050
eV. These deviations may arise due to the presence of minor
amounts of other Zn compounds and/or unidentified com-
ponents in the ambient particles.
Nearly 100% of the Pb and Zn particles contained Cl
containing chloride can undergo atmospheric processing
where the chloride is displaced by nitrate:
where M represents Zn, Pb, Na, or K. The CCSEM/EDX and
a strong temporal anticorrelation between nitrate and
chloride for Pb-, Zn-, and Na-containing particles. Each
morning, the initial spike of chloride containing particles
ed in nitrate and depleted in chloride. The reaction started
at approximately 7 am each morning and peaked at ∼12-2
pm each day. The ZnCl2 and NaCl particles began to be
HNO3 began to form in the mornings (27). Thus, the
displacement reaction shown in R1 explains the Zn(NO3)2
particles observed with the STXM/NEXAFS.
The extent of chemical processing may be different
because particles contain different amounts of reactive
chloride. CCSEM/EDX and ATOFMS show that Cl is present
in over 70% of Zn- and Pb-containing particles (Supporting
Information Figure S2) whereas “Pb max” particles contain
larger amount of Cl compared to the “Zn max” particles
(Supporting Information Figure S4). Also, all of the PbZn >
0 particles contained Na, therefore the chloride was also
industrial particles represented as much as 73% of the total
particles during morning hours, their presence can strongly
influence the chemistry of aerosols in the MCMA as
demonstrated in Figure 6.
Sources of Metal-Containing Particles in Mexico City.
Numerous industrial emission sources exist in northern
(within 5 km) obtained by the Secretary of the Environment
of the Government of the Federal District (SMA/GDF)
indicates that metallurgical sources release a substantial
fraction of the PM2.5 compared to other industries. While
here, previous studies do not report large amounts of Cl
have reported such associations with chloride in a recent
also noted high SO2concentrations within smelter plumes.
weak correlations were observed among SO2, Zn, and Pb.
Zn- and Pb-containing particles identified by ATOFMS in
this study also came from the northeast (20), and were not
FIGURE 4. Images illustrating the morphology of metal containing
particles observed in Northern Mexico City. (A) Zn containing
crystal attached to soot; (B) Needle-like particles containing Pb,
Zn, and Cl; (C) Tetrahedral Zn containing particles with compact
Pb rich particles.
7094 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
in our study did contain primary sulfur markers at m/z -32
(S-)and -64 (S2-) (Figure 3). Nevertheless, since the time
trends and indicated source direction of the Cl particles
reported by Salcedo et al. (27) and Johnson et al. (18) are
similar to those observed for the Zn/Pb/Cl-containing
particles studied here, it is highly likely that the chloride
particles observed in these previous studies came from the
Municipal and hazardous waste incineration have been
shown to emit Pb, Zn, and Cl particles (21, 28–31). For this
source, the Cl is primarily from the burning of plastics such
as polyvinyl chloride and paper, whereas the Pb and Zn can
be produced from a variety of waste materials. Electronic
waste represents a good example of a waste stream that
contains an abundance of these three elements. The PIXE,
carbonaceous material suggest these particles result from
combustion. ATOFMS identified organic particles that con-
tained an abundance of aromatic markers that may indicate
the presence of diphenyl or other similar aromatic com-
pounds typical for incineration emissions. Incineration is
however, multiple waste incinerators exist in the northern
part of Mexico City (20).
state, morphology, oxidation state and heterogeneous chem-
indicates metal-rich particles peaked at the sampling site
during the early morning hours. This is consistent with
previous studies showing high Zn and Pb concentrations
correlated with early morning air masses from the north-
eastern MCMA. Evidence of industrial incineration was
supported by the time series data showing the lowest metal
particle concentrations during a holiday weekend. Direct
comparison of our data with single particle source samples
from waste incineration supports incineration as the most
FIGURE 5. Upper panel: single energy (1024 eV) STXM images of two types of Zn containing particles observed in Northern Mexico
City. (A) Zn-containing crystals internally mixed with elemental carbon; (B) - Needle-like Zn-containing particles. Lower panel:
representative Zn L-edge normalized NEXAFS spectra for (C) particle type A and Zn(NO3)26H2O standard and (D) particle type B and
FIGURE 6. Average peak area for Cl and NO3 as a function of
time and size for Pb- and Zn-containing particles mixed with
Cl. This demonstrates the displacement of chloride by nitrate
within individual Pb and Zn particles.
VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7095
techniques confirmed that many of the ZnCl-rich particles
as the source of these particles.
Complementary ATOFMS, CCSEM/EDX, and PIXE data
show that particles containing Zn and/or Pb are mixed with
Cl and nitrate. ATOFMS measured the conversion to metal
nitrate particles immediately following the metal-chloride
particle spikes. The time lag of nitrate behind chloride
suggests a heterogeneous displacement reaction of metal
chlorides with nitric acid. This assertion is supported by
STXM/NEXAFS results showing Zn rich particles composed
of Zn(NO3)2·6H2O. Because these particles sometimes domi-
nated number concentrations in the 0.2-1 µm range, this
heterogeneous chemistry may be an important source of
submicron nitrate mass while impacting the gas phase NOy
The frequent observation of these metal-rich particles in
an urban area with a high population density also has
of the Pb-containing particles is less than 2.5 µm, meaning
that these particles may be efficiently inhaled. Also, there
may be important health ramifications if salts such as
Pb(NO3)2 are formed because lead nitrate is soluble, and
therefore more mobile within the human body.
We thank the individuals who helped out at the T0 site. The
UCSD, the MIT, and the MCE2groups acknowledge support
provided by the U.S. National Science Foundation (ATM-
0528227), the Atmospheric Science Program of the office of
Biological and Environmental Research (ASP/OBER) of the
U.S. Department of Energy (DOE), (grant DE-FG02-
05ER63980). The PNNL and the LBNL research groups
acknowledge support provided by ASP/OBER DOE. The
AUMA research group acknowledges support provided by
Mexican Comision Ambiental Metropolitana, The CCSEM/
EDX and PIXE particle analyses were performed in the
Environmental Molecular Sciences Laboratory, a national
scientific user facility sponsored by the Department of
at the ALS were partially supported by the Director, Office
of Science, Office of Basic Energy Sciences, Division of
Department of Energy at Lawrence Berkeley National Labo-
ratory under Contract No. DE-AC02-05CH11231.
Supporting Information Available
Details regarding experimental procedures, rules based
particle classification, and single particle mixing state. This
material is available free of charge via the Internet at http://
(1) Osterberg, E.; Mayewski, P.; Kreutz, K.; Fisher, D.; Handley, M.;
Sneed, S.; Zdanowicz, C.; Zheng, J.; Demuth, M.; Waskiewicz,
M.; Bourgeois, J. Ice core record of rising lead pollution in the
north pacific atmosphere. J. Geophys. Lett. 2008, 35.
(2) Pope, C. A.; Dockery, D. W. Health effects of fine particulate air
pollution: Lines that connect. J. Air Waste Manage. 2006, 56,
J. K.; Ghio, A. J.; Costa, D. L. Soluble transition metals mediate
Health 1997, 50, 285–305.
effects of metals and metal compounds. Biometals 2006, 19,
(5) Hodgson, M. J.; Bracker, A.; Yang, C.; Storey, E.; Jarvis, B. J.;
2001, 39, 616–628.
(6) Schwartz, J.; Dockery, D. W.; Neas, L. M. Is daily mortality
associated specifically with fine particles. J. Air Waste Manage.
1996, 46, 927–939.
(7) Londahl, J.; Massling, A.; Pagels, J.; Swietlicki, E.; Vaclavik, E.;
Loft, S. Size-resolved respiratory-tract deposition of fine and
rest and exercise. Inhalation Toxicol. 2007, 19, 109–116.
(8) Broday, D. M.; Georgopoulos, P. G. Growth and deposition of
Technol. 2001, 34, 144–159.
(9) Suarez, A. E.; Ondov, J. M. Ambient aerosol concentrations of
cytokine-active metals from coal- and oil-fired power plants.
Energ Fuel 2002, 16, 562–568.
(10) Murphy, D. M.; Hudson, P. K.; Cziczo, D. J.; Gallavardin, S.;
Thomson, D. S.; Thornberry, T.; Wexler, A. S. Distribution of
lead in single atmospheric particles. Atmos. Chem. Phys. 2007,
(11) Pina, A. A.; Villasenor, G. T.; Jacinto, P. S.; Fernandez, M. M.
Scanning and transmission electron microscope of suspended
lead-rich particles in the air of San Luis Potosi, Mexico. Atmos.
Environ. 2002, 36, 5235–5243.
(12) Linak, W. P.; Wendt, J. O. L. Toxic metal emissions from
incineration - mechanisms and control. Prog. Energy Combust.
Sci. 1993, 19, 145–185.
J. G. Chemical composition of fugitive dust emitters in Mexico
City. Atmos. Environ. 2001, 35, 4033–4039.
R. PIXE analysis of airborne particulate matter from Xalostoc,
B 1999, 150, 445–449.
(15) Molina, L. T.; Kolb, C. E.; de Foy, B.; Lamb, B. K.; Brune, W. H.;
Jimenez, J. L.; Molina, M. J. Air quality in North America’s most
populous city - overview of MCMA-2003 Campaign. Atmos.
Chem. Phys. 2007, 7, 2447–2473.
(16) Molina, M. J.; Molina, L. T. Air Quality in the Mexico Megacity:
An Integrated Assessment; Kluwer Academic: Norwell, MA2002.
and spatial variations of metal content in TSP and PM10 in
Mexico City during 1996-1998, J. Aerosol Sci. 2002, 33, 91-102.
Xie, Y.; Laskin, A.; Shutthanandan, V. Aerosol composition and
source apportionment in the Mexico City Metropolitan Area
Phys. 2006, 6, 4591–4600.
(19) Schnaas, L.; Rothenberg, S. J.; Flores, M. F.; Martinez, S.;
in a Cohort of Children in Mexico City (1987-2002). Environ.
Health Perspect. 2004, 112, 1110–1115.
(20) Moffet, R. C.; de Foy, B.; Molina, L. T.; Molina, M. J.; Prather,
K. A. Measurement of ambient aerosols in northern Mexico
City by single particle mass spectrometry. Atmos. Chem. Phys.
2008, 8, 4499–4516.
(21) Tan, P. V.; Fila, M. S.; Evans, G. J.; Jervis, R. E. Aerosol laser
ablation mass spectrometry of suspended powders from PM
sources and its implications to receptor modeling. J. Air Waste
Manage. 2002, 52, 27–40.
and evolution of fugitive particles from a copper smelter.
Environ. Sci. Technol. 1981, 15, 1208–1212.
(23) Moffet, R. C.; Qin, X.; Rebotier, T.; Furutani, H.; Prather, K. A.
Chemically Segregated Optical and Microphysical Properties
of Ambient Aerosols Measured in a Single Particle Mass
Spectrometer J. Geophys. Res., [Atmos.] 2008, in press, doi:
(24) Wu, C. Y.; Biswas, P. An equilibrium-analysis to determine the
speciation of metals in an incinerator. Combust. Flame 1993,
(25) Eighmy, T. T.; Eusden, J. D.; Krzanowski, J. E.; Domingo, D. S.;
Stampfli, D.; Martin, J. R.; Erickson, P. M. Comprehensive
approach toward understanding element speciation and leach-
precipitator ash. Environ. Sci. Technol. 1995, 29, 629–646.
7096 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
(26) Dall’Osto, M.; Booth, M. J.; Smith, W.; Fisher, R.; Harrison, R.
M. A study of the size distributions and the chemical characteri-
sation of airborne particles in the vicinity of a large integrated
steelworks. Aerosol Sci. Technol. 2008, in press.
(27) Salcedo, D.; Onasch, T. B.; Dzepina, K.; Canagaratna, M. R.;
Zhang, Q.; Huffman, J. A.; DeCarlo, P. F.; Jayne, J. T.; Mortimer,
P.; Worsnop, D. R.; Kolb, C. E.; Johnson, K. S.; Zuberi, B.; Marr,
L. C.; Volkamer, R.; Molina, L. T.; Molina, M. J.; Cardenas, B.;
A.; Shutthanandan, V.; Xie, Y.; Brune, W.; Lesher, R.; Shirley, T.;
Jimenez, J. L. Characterization of ambient aerosols in Mexico
City during the MCMA-2003 campaign with aerosol mass
Phys. 2006, 6, 925–946.
(28) Walsh, D. C.; Chillrud, S. N.; Simpson, H. J.; Bopp, R. F. Refuse
YorkCityduringthe20thcentury.Environ. Sci. Technol. 2001, 35,
An improved paradigm for the structure and dynamics of the
urban mid-Atlantic aerosol. Environ. Sci. Technol. 1998, 32,
(30) Hu, C. W.; Chao, M. R.; Wu, K. Y.; Chang-Chien, G. P.; Lee, W. J.;
Chang, L. W.; Lee, W. S. Characterization of multiple airborne
particulate metals in the surroundings of a municipal waste
incinerator in Taiwan. Atmos. Environ. 2003, 37, 2845–2852.
(31) Chang, M. B.; Huang, C. K.; Wu, H. T.; Lin, J. J.; Chang, S. H.
Characteristics of heavy metals on particles with different sizes
VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7097
Supporting Information to:
Characterization of Aerosols Containing Zn, Pb, and Cl from an
Industrial Region of Mexico City
R.C. Moffet1, Y. Desyaterik2, R.J. Hopkins3, A.V. Tivanski3, M.K. Gilles3, Y. Wang1, V.
Shutthanandan2, L.T. Molina4,5, R. Gonzalez Abraham4, K.S. Johnson5, V. Mugica6, M.J.
Molina1, A. Laskin2,* and K.A. Prather1*
1Department of Chemistry and Biochemistry, University of California, San Diego, CA 92093-0314
2W.R. Willey Environmental Molecular Sciences Laboratory,
Pacific Northwest National Laboratory, Richland, WA 99352
3Chemical Sciences Division, Lawrence Berkeley National Laboratory,
Berkeley, CA 94720-8226
4Molina Center for Energy and the Environment (MCE2), LaJolla, CA 92037
5Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA 02139-4307
6Applied Chemistry, Universidad Autónoma Metropolitana-Azcapotzalco (UAM-A). Av. San Pablo 180,
This supporting information contains: 6 pages of text, 2 Tables and 5 Figures
This article is a U.S. government work, and is not subject to copyright in the United States.
A Davis Rotating drum Universal size-cut Monitoring (DRUM) impactor (1) and
Time Resolved Aerosol Collector (TRAC) (2,3) were used to collect particulate matter
samples for laboratory analysis. The DRUM impactor collected bulk particle samples for
PIXE analysis in size ranges of 1.15– 2.5µm (Stage A), 0.34– 1.15µm (Stage B), and 0.07–
0.34µm (Stage C) onto three Teflon strips at a fixed air flow of 10 slpm with a rotation
rate of 2mm per 12 h. The TRAC sampled particles onto a rotating impaction plate
containing pre-arranged microscopy substrates. In this study, each substrate was exposed
for 15 minutes of sample collection while the collection plate was continually advanced
to prevent particle overlap and provide time resolution. The effective aerodynamic cut-off
size D50 of the TRAC is ~0.36 µm. Two different types of substrates were used in this
study: (a) Copper 400 mesh TEM grids coated with Carbon Type-B films (Ted Pella,
Inc.) for the computer controlled scanning electron microscopy with energy dispersed
analysis of X-rays (CCSEM/EDX), and (b) 100-nm thick silicon nitride (Si3N4)
membranes (Silson Ltd, Inc.) for the Scanning Transmission X-ray Microscopy with
Near Edge X-ray Absorption Fine Structure spectroscopy (STXM/NEXAFS).
Aerosol Time-of-Flight Mass Spectrometry (ATOFMS). The ATOFMS was a
single particle mass spectrometer used for the MILAGRO study and is described
elsewhere (4). The ATOFMS analyzes sizes between 200 – 3000 nm. The single particle
mass spectra, size, scattering intensity, and temporal information were imported into the
MATLAB YAADA database (http://www.yaada.org/) (5).
Proton Induced X-Ray Emission (PIXE). PIXE analysis was done shortly after
the MILAGRO campaign at the Environmental Molecular Sciences Laboratory (EMSL)
at Pacific Northwest National Laboratory (PNNL). Experimental procedures have been
described in detail elsewhere (6). Briefly, Teflon substrates mounted on a special sample
holder were placed inside a vacuum chamber evacuated to 2×10−7Torr. A 3.5 MeV
proton beam was used for PIXE analysis and the spectra were evaluated by the GUPIX
program (7). Concentrations of elements were determined by calibration to known
standards (MicroMatter, Deer Harbor, WA) with approximately 5% uncertainty. The
mass concentrations were used to perform a principle components analysis that is
described in detail in the supplementary information section.
Computer Controlled Scanning Electron Microscopy/Energy Dispersive X-Ray
Analysis (CCSEM/EDX). CCSEM/EDX particle analysis was performed at
EMSL/PNNL following the campaign. A FEI XL30 digital field emission gun
Environmental Scanning Electron Microscope equipped with an EDX microanalysis
spectrometer (EDAX, Inc.) was used in this study. Specific details on the CCSEM/EDX
analysis of particles collected on the filmed grids are published elsewhere (2). The
elements detected in particles for this work were C, N, O, Na, Mg, Al, Si, P, S, Cl, K, Ca,
Ti, Mn, Fe, Ni, Cr, Zn, and Pb. The atomic percentage data for light elements (C, N, O)
are considered semiquantitative (2) and therefore were omitted for particle classification
purposes. For all other elements, particle composition is reported either as normalized
atomic percentage or atomic ratios.
Scanning Transmission X-Ray Microscopy/Near Edge X-Ray Fine Structure
Spectroscopy (STXM/NEXAFS). Zinc L-edge NEXAFS spectra and images were
acquired using the STXM instrument on beamline 11.0.2 of the Advanced Light Source
(Berkeley, CA) in a ~0.5 atm He-filled chamber. The apparatus and experimental method
are described in detail elsewhere (8-12). The spatial resolution for these experiments was
~35 nm. Dwell times used to acquire a single energy image were typically 1 ms per pixel.
The X-ray energy calibration (±0.1 eV) was afforded by addition of CO2gas (~6 Torr) to
the STXM chamber through comparison of the position of CO2Rydberg transitions at
292.74 and 294.96 eV (13).
Standard ZnO, ZnS, Zn(NO3)2• 6H2O, ZnCl2powders and a 0.05 M solution of
ZnSO4 in water were purchased from Sigma-Aldrich (St. Louis, MO). The purity of
standard reference materials was ≥ 95%. Thin coatings of the standards were deposited
directly onto Si3N4windows by applying gentle contact between the crushed fine powder
sample and the Si3N4window and subsequently removing any loose sample. The ZnSO4
solution sample was prepared by placing a small drop on a Si3N4 window and then
allowing it to evaporate leaving a crystalline residue.
High-Volume Filter Sampling. 24-h integrated samples of PM2.5 and PM10 were
collected at T0 with high volume Tisch samplers (Cleaves, Ohio) equipped with quartz
microfiber filters. Filters were stabilized in a controlled temperature and humidity room
before and after sampling for gravimetric analysis. Extraction of metals from one half of
the filter was carried out with an OI Analytical (College Station, TX) microwave oven
using 2 ml of HNO3, 1 ml of HCl and 2 ml of HF. After digestion the solutions were
neutralized with 2% boric acid and gauged. Metals analysis was performed using an ICP-
OES (Thermo Jarrel Ash, Waltham, MA). A standard reference material (High Purity,
QC-TMFM-A) was used to validate and verify the digestion method accuracy and
efficiency, as well as to calculate metal recoveries, percent recovery and traceability
2. Single Particle Mixing State and Factor Analysis of Bulk Measurements In the
following analysis, we compare single particle mixing states obtained from ATOFMS
and CCSEM/EDX data (Figures S1-S4) using the rule based classification scheme
detailed in Table S1. As the methodologies behind these techniques are fundamentally
different, instrumental characteristics must be considered when comparing the results. In
the ATOFMS technique, atomic and molecular ionization efficiencies may differ, and
matrix effects may be important for particles of different composition. In turn,
CCSEM/EDX provides relative abundance of various elements calculated from EDX
spectra as atomic percentages (2).
PIXE data compliments the single particle data by supplying quantitative bulk
mass concentrations of individual elements (Z > 11). Table S2 presents factor loadings
obtained using a principal components analysis. Six components of the variation in the
elemental concentrations were identified on the basis of the magnitude of eigenvalues
(greater than one), and these six factors accounted for 82% of total variance in the entire
dataset. The first factor was highly loaded in Al, Si, Ca, Ti, Mg, and Fe. It explained 27%
of the total system variance. This factor appears to represent the crustal sources,
including air-borne road dust, construction dust, and fugitive dust. The second factor had
strong factor loadings for Cl, Na, Zn, and moderately loaded with Pb, Cr, P. It accounted
for 22% of the total variance. The fifth factor grouped Cu and K together but with only a
slight correlation with Zn, Pb and other metals.
Mixing State of the Me>0, ZnPb>0 Class. To better understand the sources,
formation routes and variability of metal rich particles, it is useful to determine the
chemical associations (mixing state) in the individual particles. Both the CCSEM/EDX
and ATOFMS can determine the chemical mixing state of single particles. Figure S1
shows stacked bar charts for approximately 4400 (CCSEM/EDX) and 800 (ATOFMS)
individual metal containing (Me>0) particles detected shortly after the rain event on
March 24th. Stacked bars are plotted for individual particles and the colored area indicates
the percent of each metal normalized to sum of all constituents. Potassium was omitted
from Figure S1b because it often dominated the ATOFMS signal and hindered our ability
to graphically compare mixing states of particles detected with ATOFMS and
CCSEM/EDX. Figure S1 clearly shows that Zn, Pb, and Fe are the most abundant metals
in the ZnPb>0 class and are internally-mixed. The relative proportions of the Pb max and
Zn max particles differ between the two techniques due to the enhanced sensitivity of
ATOFMS for Zn and Pb. Additional metals in the ZnPb>0 class include Mg, Al, Ca, Mn,
K and Si. Here, Si, Al, and Mg show only minor presence although they are more
abundant in the Fe max and “ other” unidentified (or organic) classes.
Another way to examine mixing state involves comparing the percentages of
particles in a certain class that are mixed with a specific chemical species. In Figure S2,
the mixing states of the major classes within the ZnPb > 0 group can be seen for both the
ATOFMS (S2a) and CCSEM/EDX (S2b) data sets. Particle classes are indicated on the
y-axis and the color scale represents the fraction of the particles in these classes that
contain the chemical markers on the x-axis. Both CCSEM/EDX and ATOFMS confirm
that Zn and Pb containing particles are strongly mixed with Na, K, and Cl. Within the Zn
max class, both CCSEM/EDX and ATOFMS data show a strong association with sulfate
(detected as S in the EDX spectra). The ATOFMS data indicate that the secondary
nitrogen markers 46NO2-and 62NO3-are strongly coupled with all particles containing Zn
and Pb although the amount of NO3is a strong function of the time of day due to the
nitrate/chloride heterogeneous chemistry described in the main manuscript. Phosphorous
95PO4), represented in Figure S2a as PO, are found to be mainly
associated with the ZnPb > 0 set and weakly associated with Me>0, Pb & Zn = 0 particles.
The CCSEM/EDX measurements reveal that P is mostly present in the ZnPb>0 class as
well, but in fewer particles. This result suggests that P is a trace element in ZnPb>0
As seen in Fig S2a, ~50% of the ZnPb>0 particles were also mixed with carbon.
Compared to Me>0, Pb & Zn = 0 classes, the ZnPb>0 classes contain a higher fraction of
particles with a 36C3+peak. This marker is an important identifier for elemental carbon
(EC), and indeed EC was most strongly mixed within the ZnPb>0, Zn max type. This
was further corroborated by the STXM/NEXAFS identification of soot in Zn-containing
A cluster accounting for 6.3% of the total Pb containing particles contained Cu,
EC, oxalate, Zn, and Pb. Here, this is reflected in the fact that Zn max particles also
contain a higher proportion of Cu and oxalate ion (C2O4-2). Approximately 65% of the Zn
max particles contain the oxalate marker (Fig S2a). Although the exact reason for this is
not known, it is common knowledge that oxalate binds strongly to metal ions within fog
water (14). The formation of metal-oxalate complexes from the constituent ions has an
equilibrium constant Keq > 106M-1 and increases with the metal oxidation state. If the
metals are present along with hot water vapor in industrial emissions, oxalate complexes
may form in the aqueous particles.