supplement of evaluation of the performance of a …...cc attribution 3.0 license. supplement of...
TRANSCRIPT
Supplement of Atmos. Meas. Tech., 7, 2121–2135, 2014http://www.atmos-meas-tech.net/7/2121/2014/doi:10.5194/amt-7-2121-2014-supplement© Author(s) 2014. CC Attribution 3.0 License.
Supplement of
Evaluation of the performance of a particle concentrator for onlineinstrumentation
S. Saarikoski et al.
Correspondence to:S. Saarikoski ([email protected])
Supplements
Table S1. The shift of particles size for monodisperse AS and DOS particles in the m-VACES.
Particle size (nm) AS % (nm) DOS % (nm)
50 20 (10) 30 (5.9) 70 7.7 (5.4) 7.7 (5.4) 100 4.0 (4.0) 4.0 (4.0) 200 3.5 (7.0) 3.5 (7.0) 300 0.0 (0.0) 0.0 (0.0)
Fig. S1. Time series for the size distributions of particle number (a-b), mass (c-d) and enrichment factor. (a) and (c) are for ambient and (b) and (d) for concentrated aerosol. Particle mass was calculated by using the density of 1.48 g cm-3.
0.4
0.3
0.2
0.1
0.0
5 6 7 8 9100
2 3 4 5 6 7 8 91000
Particle diameter (nm)
10
8
6
4
2
0
35
30
25
20
15
10
5
0
Co
nce
ntra
ted
(µg
m-3)
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Am
bie
nt
(µg
m-3
)
Acidic particles
Neutral particles
ambient concentrated
ambient concentrated
Fig. S2. Size distribution of ammonium during the period when the particles were acidic (April 12, 2010) and nearly neutral (April 13, 2010; 6 am to midnight). Ambient size distributions were smoothed by one point.
Fig. S3. Scatter plots of the mass fragments in ambient and concentrated OA separated into different compound classes. Carbon clusters (Cx) were measured only with the laser on and the concentration of C1 (m/z 12) according to the fragmentation Table in Onasch et al. (2012). Concentrated aerosol is divided by the average EF for organics (27.08; a, c-e) and r-BC (37.61; b) (Table 2). One to one ratio is shown by dash lines.
0.001
2
4
6
0.01
2
4
6
0.1
2
4
conc
entr
ated
aer
osol
(µg
m-3
)
0.0012 4 6 8
0.012 4 6 8
0.12 4
ambient aerosol (µg m-3
)
2829
30
3142
43
45
46
5354
55
56
57
5859
66
67
68
69
7071
7279
8182
83
8485
86
93 94
95
9697
98
99101
103
104105
106
107
108
109
110111
112113
Organic compoundswith one oxygenatom (CxHyO)
0.001
2
4
6
0.01
2
4
6
0.1
2
4
conc
entr
ated
aer
osol
(µg
m-3
)
0.0012 4 6 8
0.012 4 6 8
0.12 4
ambient aerosol (µg m-3
)
44
45
4647
56
5859
60
61
68
69
70
71
72
73
74
7576
8081
82
8384
85
86
87
969798
99
100167
Organic compoundswith more than oneoxygen atom(CxHyOz (z>1))
0.001
2
4
6
0.01
2
4
6
0.1
2
4co
ncen
trat
ed a
eros
ol (
µg
m-3
)
0.0012 4 6 8
0.012 4 6 8
0.12 4
ambient aerosol (µg m-3
)
30
42
5254
5672
84
Organic compounds withone nitrogen atom (CxHyN)
0.001
2
4
68
0.01
2
4
68
0.1
2
4
conc
entr
ated
aer
osol
(µg
m-3
)
0.0012 4 6 8
0.012 4 6 8
0.12 4
ambient aerosol (µg m-3
)
13
15
25
26
2729
3037
38
39
40
41
42
43
44
49
5051
52
53
54
55
56
57
5961
62
6365
66
67
68
69
7071
72
737475
76
77
78
79
80
81
82
83
84
85
86
8788
89
90
91
92
93
94
95
96
97
98
99100
101
102103
104
105
106
107
108
109
110
111
112113
114
115
Hydrocarbonfragments (CxHy)
0.001
2
4
6
0.01
2
4
6
0.1
2
4
conc
entr
ated
aer
osol
(µ
g m
-3)
0.0012 4 6 8
0.012 4 6 8
0.12 4
ambient aerosol (µg m-3
)
1224
36
Carbon clusters (Cx)
4
3
2
1
0
CH
4N,
C2H
4N,
C3H
6N (
µg
m-3
)
9-Apr-10 10-Apr-10 11-Apr-10 12-Apr-10 13-Apr-10 14-Apr-10 15-Apr-10Date
80
60
40
20
0
EF
for
NH
4
20
15
10
5
0
NH
4 (µg
m-3)
NH4
CH4N C2H4N C3H6N
a)
b)
Fig. S4. Time series of ammonium and amines for the concentrated aerosol (a) and the enrichment factor for ammonium (b).
1.70
1.60
1.50
24201612840Diurnal hour
0.45
0.40
0.35
0.30
0.25
1.8
1.6
1.4
1.2
1.0
Ele
me
nta
l ra
tio
20
10
0
x10
-3
OM:OC
O:C
H:C
N:C
concentrated ambient
concentrated ambient
concentrated ambient
concentrated ambient
Fig. S5. Diurnal trends for OM:OC, O:C, H:C and N:C for concentrated and ambient OA.
Fig. S6. Correlation of mass spectra for PMF factors for ambient and concentrated OA.
0.001
2
4
6
80.01
2
4
6
80.1
Fra
ctio
n in
co
nce
ntr
ate
d O
A
0.0012 3 4 5 6
0.012 3 4 5 6
0.1
Fraction in ambient OA
C
CH
CH3
HO
H2O
C2H
C2H2
C2H3
CO
CHO
C2H5
CH2O
CH4N
CH3O
C3H
C3H2
C3H3
C3H4
C3H5
C2H2O
j13CC2H5
C3H6
C2H3O
CO2
j13CCH3O
C3H8
CHO2
C2H5O
CH3O2
C4H2
C4H3
C4H4C3HO
C4H5
C3H2O
C4H6
C3H3OC4H7
C3H4OC4H8
C3H5O
C4H9
C2H2O2
C3H6O
C2H3O2
C3H7O
C2H4O2
C2H5O2
C5H2
C5H3
C5H5
C5H6
C4H3O
C5H7
C4H4OC5H8C3HO2
C4H5O
C5H9
C3H2O2
C4H6O
C5H10
C3H3O2C4H7O
C5H11
C3H4O2
C3H5O2
C3H9SiC3H6O2
C6H3
C6H4
C6H5
C6H6
C6H7
C4O2
C6H8
C4HO2
C5H5O
C6H9
C5H6O
C6H10C4H3O2
C5H7O
C6H11
C4H4O2C5H8O
C4H5O2
C5H9O
C4H6O2C4H7O2C7H5
C7H7
C7H9
C6H6O
C6H7OC7H11
C6H8OC5H5O2
C6H9OC7H13C8H2
C6H10O
C5H7O2
C8H5
C8H7
C8H9
C7H7O
C8H11C7H9O
C8H13C9H3C7H11OC9H5
C9H7
OOA-2
0.001
2
4
6
80.01
2
4
6
80.1
Fra
ctio
n in
co
nce
ntr
ate
d O
A
0.0012 3 4 5 6
0.012 3 4 5 6
0.1
Fraction in ambient OA
C
CH
CH3
HO
H2O
C2H
C2H2
C2H3
COCHO
C2H5
CH2OCH4N
CH3O
C3H
C3H2
C3H3
C3H4
C3H5C2H2O
C2H4N
C3H6
C2H3O
C3H7
CO2
j13CCH3O
C3H8
CHO2
CH3NO
C2H5O
CH3O2C4H2
C4H3
C4H4C3HO
C4H5
C3H2O
C3H4N
C4H6
C3H3O
C4H7C3H4O
C3H6N
C4H8
C3H5O
C4H9
C2H2O2
C3H8NC2H3O2
C3H7O
C2H4O2
C2H5O2
C5H2
C5H3C5H5
C5H6C4H3O
C5H7C4H4O
C5H8
C3HO2
C4H5O
C5H9C3H2O2
C4H6O
C3H3O2
C4H7O
C3H4O2
C3H5O2
C6H2
C3H6O2
C6H3C2H4O3C6H4
C6H5
C6H6
C6H7
C6H8
C5H5O
C6H9
C4H2O2
C5H6OC4H3O2
C5H7OC4H4O2C5H8O
C4H5O2
C5H9OC4H6O2C3H3O3
C4H7O2C7H5
C7H7
C6H5O
C7H9C6H6O
C6H7O
C6H8O
C5H5O2C6H9OC4H3O3
C5H7O2C8H5
C8H7
C7H5OC8H9
C7H7O
C7H8O
C7H9OC9H3C7H11O
C9H4
C9H5
C9H7
OOA-1
0.001
2
4
6
80.01
2
4
6
80.1
Fra
ctio
n in
co
nce
ntr
ate
d O
A
0.0012 3 4 5 6
0.012 3 4 5 6
0.1
Fraction in ambient OA
C
CH
CH3
HO
H2O
C2H
C2H2
C2H3COC2H5
CH2O
CH4N
CH3O
C3H
C3H2
C3H3
C3H4
C3H5
C2H2O
j13CC2H5
C3H6
C2H3O
C3H7
CO2
j13CCH3O
C3H8CHO2
C2H5O
CH3O2
C4H2C4H3
C4H4
C3HO
C4H5
C3H2O
C4H6
C3H3O
C4H7
C3H4Oj13CC3H7
C4H8
C3H5O
C4H9
C3H6O
j13CC3H9
C3H7OC2H4O2
C5H2
C5H3
C5H5
C5H6
C5H7
C4H4O
C5H8
C3HO2
C4H5O
C5H9
C4H6O
j13CC4H9
C5H10
C3H3O2
C4H7O
C5H11
C3H5O2
C6H4
C6H5
C6H6
C6H7
C4O2C6H8
C4HO2
C6H9
C5H6O
C6H10
C5H7O
C6H11
C5H8O
C6H12C4H5O2C5H9O
C6H13
C4H7O2C7H5
C7H7
C7H8
C7H9
C7H10
C6H7O
C7H11
C7H12
C7H13
C5H7O2C8H7
C8H9C8H11
C7H9O
C8H13
C7H11OC8H15
C9H7
HOA
Fig. S7. The variation of Q/Qexpected with the value of fPEAK for ambient (a) and concentrated (b) PMF solution.
Fig. S8. The variation of time series and mass spectra for the PMF factors with the fPEAK values from -0.4 to 1.6. Ambient (a) and concentrated (b) OA.
Detection of trace elements
HR-AMS can detect several trace elements with lower melting and boiling points in near
real-time (Salcedo et al., 2012), and furthermore, the SP-AMS can detect some additional
metals and trace elements due to the laser vaporizer (Onasch et al., 2012). However, the
concentrations of trace metals are typically very low in ambient air, especially in Finland,
and therefore their real-time detection is challenging. Thus the use of a concentrator may
allow the real-time detection of elements that otherwise could not be observed.
Five trace elements were detected in ambient air without the m-VACES; aluminum,
vanadium, iron, zinc and rubidium. Except zinc, all of the elements were most likely
associated with the laser vaporization as there was a steep drop in the recorded
concentrations when the laser vaporizer was switched off. Rubidium can undergo surface
ionization with a tungsten vaporizer making its quantification difficult (Drewnick et al.,
2006), similar to sodium and potassium. In this study surface ionization in tungsten
vaporizer was not likely to be a confounding issue, as rubidium seemed to be totally
vaporized by the laser and not hit the tungsten vaporizer, however, surface ionization can
also take place on the hot surface of the r-BC particles discussed more in Carbone et al.
(2014). Slow evaporation is another issue related to the detection of semi-refractory
elements with the AMS, requiring special data analysis procedures (Salcedo et al., 2012).
Zinc signals were consistent with slow evaporation at tungsten vaporizer the ratio of
closed to open being around 0.70 and 0.25 for ambient and concentrated aerosol,
respectively (Fig. S9). However, a part of the difference could be due to the uncertainty
in the detection of low signal for Zn.
Trace elements gave rather similar EFs to the other species measured, however, the
standard deviations for EFs were large probably due to the high uncertainty associated
with the determination of their low ambient concentrations (Table 2). Of all the elements,
rubidium had the smallest enrichment factor and aluminum the highest. EFs were
calculated for the most abundant isotopes (m/z 26.982 for Al, m/z 50.944 for V, m/z
55.935 for Fe, m/z 84.912 for Rb), except for zinc that was detected at the isotope 68Zn
(m/z 67.925). All the other isotopes of V, Fe and Zn could not be resolved from the
neighboring peaks or were below their detection limits. The isotopic ratio of 85Rb and 87Rb is shown in Fig. S10 for open and closed modes. The isotopic ratio of 87Rb to 85Rb
was near the natural isotopic ratio of 0.386 for both the modes. Unfortunately, it was not
possible to investigate size-dependent EFs for trace elements as they composed only a
minor fraction of the total signal at each unit mass. Additionally, due to the limited data
coverage from the laser vaporizer, diurnal patters for the EF could not be calculated.
Aluminum, iron and rubidium correlated with r-BC. This suggests that they have similar
combustion-related origins in Helsinki, mostly in the form of traffic sources, either from
exhaust or non-exhaust (e.g. road dust) emissions. Zinc has been shown to be associated
with regional or long-range transported air masses in Helsinki whereas vanadium has
more local sources like heavy oil combustion (Pakkanen et al., 2001).
Besides those five trace elements that were detected in ambient air without the
concentrator, three additional elements were detected only with the m-VACES. These
elements were strontium, zirconium and cadmium (Fig. S11) and they were identified
based on their exact m/z. Strontium was detected at m/z 87.906 and zirconium at m/z
89.905 that were their most abundant isotopes. Cadmium was observed at m/z 111.903.
Unfortunately, any other isotopes of these elements could not be detected, and therefore
their existence could not be verified by the isotopic ratios. Of those three, only strontium
seemed to be associated with r-BC and vaporized with the laser. However, because
zirconium in its elemental form requires about 4400 ºC to be vaporized, one possibility is
that this trace element was associated with other elements forming for example salts or
oxides that could be vaporized with lower temperatures. Neither zirconium nor cadmium
showed evidence of slow evaporation in tungsten vaporizer indicated by a signal in the
closed mode.
Strontium and zirconium did not have any clear correlation with the other species
measured whereas cadmium correlated with organics and r-BC. That was unexpected as
cadmium did not seem to be in the same particles with r-BC (detected also without laser).
Previously cadmium has been related to several sources in Helsinki and it was found to
be mostly regional or long-range transported (Pakkanen et al., 2001). Strontium, on the
other hand, has been associated with traffic-related components from vehicle motor oil
but it can also originate from crustal material. Zirconium has not been reported in
Helsinki urban air before but in US it has been reported e.g. from the emission of the
coal-fired boilers and in samples of geological material (Watson et al., 1995).
7
6
5
4
3
2
1
0
Am
bie
nt (
Hz)
10-Apr-2010 11-Apr-2010
Date
80
60
40
20
0
Con
cen
tra
ted
(H
z)
Open Closed
Open Closed
Fig. S9. Slow evaporation of zinc. Open and closed signal for ambient and concentrated aerosol.
70
60
50
40
30
20
10
0
87R
b (n
g m
-3,
RIE
=1)
16014012010080604020085
Rb (ng m-3
, RIE=1)
14
12
10
8
6
4
2
0
87R
b (
ng
m-3
, RIE
=1
)
40302010085
Rb (ng m-3
, RIE=1)
Open Closed
y = 0.391*x 2
r2= 0.931
y = 0.378*x 2
r2= 0.930
Fig. S10. Scatterplot of signal from 87Rb and 85Rb for open and closed modes. Expected natural isotopic ratio is shown by a dash line.
100x10-3
80
60
40
20
0
Diff
Hz/
ns
88.288.188.087.987.8m/z
2.0
1.5
1.0
0.5
0.0
Diff
Hz/
ns
88.288.188.087.987.8m/z
50x10-3
40
30
20
10
0
Diff
Hz/
ns
90.290.190.089.989.8
m/z
1.0
0.8
0.6
0.4
0.2
0.0
Diff
Hz/
ns
90.290.190.089.989.8m/z
100x10-3
80
60
40
20
0
Diff
Hz/
ns
112.3112.2112.1112.0111.9111.8
m/z
2.0
1.5
1.0
0.5
0.0
Diff
Hz/
ns
112.3112.2112.1112.0111.9111.8
m/z
StrontiumAmbient
StrontiumConcentrated
ZirconiumAmbient
ZirconiumConcentrated
CadmiumAmbient
CadmiumConcentrated
Raw fitted
Raw fitted
Raw fitted
Raw fitted
Raw fitted
Raw fitted
87.9056
89.9047
111.9028
Fig. S11. Strontium, zirconium and cadmium peaks for ambient and concentrated aerosol. Additional isotopes of these ions could not be detected.
References
Carbone, S., Onasch, T., Saarikoski, S., Timonen, H., Sueper, D., Worsnop, D. R., and Hillamo, R.: Trace metals characterization with the SP-AMS: detection and quantification, in preparation, 2014. Drewnick, F., Hings, S. S., Curtius, J., Eerdekens, G., Williams, J.: Measurement of fine particulate and gas-phase species during the New Year’s fireworks 2005 in Mainz, Germany, Atmos. Environ. 40, 4316–4327, 2006. Onasch, T. B., Trimborn, A., Fortner, E. C., Jayne, J. T., Kok, G. L., Williams, L. R., Davidovits, P., and Worsnop, D. R.: Soot Particle Aerosol Mass Spectrometer: Development, Validation, and Initial Application, Aerosol Sci. Technol., 46, 804-817, 2012. Pakkanen, T. A., Loukkola, K., Korhonen, C.H., Aurela, M., Mäkelä, T., Hillamo, R., Aarnio, P., Koskentalo, T., Kousa, A., and Maenhaut, W.: Sources and chemical composition of atmospheric fine and coarse particles in the Helsinki area, Atmos. Environ. 35, 5381–5391, 2001. Salcedo, D., Laskin, A., Shutthanandan, V., and J.-L. Jimenez, J.-L.: Feasibility of the Detection of Trace Elements in Particulate Matter Using Online High-Resolution Aerosol Mass Spectrometry, Aerosol Sci. Technol., 46, 1187-1200, 2012. Watson, J. G., Chow, J. C., and Houch, J. E.: PM2.5 chemical source profiles for vehicle exhaust, vegetative burning, geological material, and coal burning in Northwestern Colorado during 1995, Chemosphere, 43, 1141-1151, 1996.