Introduction
Low molecular weight (LMW) monocarboxylic acids such as formic (HCOOH) and
acetic (CH3COOH) acids are present in the atmosphere as major gaseous
and particulate organic components (e.g., Kawamura and Kaplan, 1984; Kawamura
et al., 2000; Paulot et al., 2011). Gaseous and particulate formic and acetic
acids have been reported in urban (Kawamura et al., 2000), forest (Andreae et
al., 1988), high mountain (Preunkert et al., 2007), marine (Miyazaki et al.,
2014), and Arctic samples (Legrand et al., 2004). Salts of organic acids in
aerosols are water-soluble and thus influence the radiation budget of the
earth's atmosphere by acting as cloud condensation nuclei (CCN) (Kanakidou et
al., 2005). In addition, LMW monocarboxylic acids have been detected in wet
deposition such as rain, cloud and fog water, and snow samples (Maupetit and
Delmas, 1994; Keene et al., 1995; Kawamura et al., 1996, 2012). Thus, organic
acids are scavenged by wet deposition from the upper troposphere. It is also
important to note that organic acids largely contribute to total acidity of
rainwaters (Kawamura et al., 1996; Keene et al., 1983).
Formic and acetic acids have a variety of sources such as primary emission
from motor exhausts (Kawamura et al., 2000) and vegetation (Kesselmeier et
al., 1998), and secondary formation via the oxidation of anthropogenic and
biogenic precursors such as toluene (laboratory data in Kawamura) and
isoprene (Paulot et al., 2011). Kawamura et al. (2000) reported that
photochemical oxidations of various organic precursors are more important as
a source of monocarboxylic acids in the troposphere. On the other hand,
microorganisms are known to produce branched chain (iC4 and iC5)
monocarboxylic acids (Allison, 1978).
The Japanese islands are located in the western North Pacific Rim, which is
influenced by the Asian outflow of dusts and air pollutants. Asian dust
(Kosa) events in the desert areas of North China promote the delivery of air
pollutants with dust particles to the western North Pacific by westerly winds
(e.g., Iwasaka et al., 1983). LMW monocarboxylic acids have
been detected in alpine snow samples collected near the summit of Mt.
Tateyama (Kawamura et al., 2012). They reported higher concentrations of
monocarboxylic acids in snow pit samples with dust layers, suggesting that
monocarboxylic acids may be associated with Asian dust during long-range
atmospheric transport.
During the winter monsoon season, the Japanese high mountains facing the Sea
of Japan are known to have heavy snowfall, which is associated with a
significant evaporation of water vapors from the warm Tsushima Current in the
Sea of Japan under a strong westerly wind condition. Alpine mountain snow
sequences would provide useful information on the chemical states of Asian
dust deposited over the snowfield, in which atmospheric organic acids are
well preserved in snow layers (Osada et al., 2004).
In the present study, we collected snowpack samples from a pit sequence in
the Murodo-Daira snowfield (ca. 6 m in depth) near the summit of Mt.
Tateyama, central Japan, in April of 2009 and 2011. To better understand the
sources of monocarboxylic acid and their long-range transport by Asian dust
over the Japanese islands, 16 snowpack samples were analyzed for
monocarboxylic acids, inorganic ions, and dissolved organic carbon (DOC) as
well as reference dust materials of Chinese loess deposit samples collected
from the Tengger and Gobi deserts. We discuss the contributions of LMW
monocarboxylic acids to DOC as well as the association of monocarboxylic
acids with alkaline dust particles during long-range atmospheric transport.
Relations between monocarboxylic acids and pH values of the snowmelt water
will also be discussed in terms of atmospheric titration of alkaline dust
particles by acidic species, including organic acids during atmospheric
transport.
Location of the snowpack sampling site (Murodo-Daira) near Mt.
Tateyama, central Japan. Sites are also shown for the loess deposit reference
samples, which were collected from the Tengger and Gobi deserts in China and
Mongolia (Nishikawa et al., 2000, 2013).
Material and methods
Sample collection
The details on the snow collection and sample storage methods were described
in Kawamura et al. (2012) and Mochizuki et al. (2016). Snowpack samples were
collected at the Murodo-Daira site (36.58∘ N, 137.36∘ E;
elevation 2450 m) near Mt. Tateyama (elevation 3015 m), central Japan
(Fig. 1). A snow pit hole (depth, ca. 6 m) was dug down to the ground.
Table 1 provides descriptions of snow samples collected from the snow pit
sequence, in which several brown-colored dirty layers were recognized by
visual observation. Five snowpack samples including three dirty layers were
collected from the pit sequence on 18 April 2009. Eleven snowpack samples
including four dirty layers were collected from the pit sequence on 17 April
2011. In order to evaluate the consistent distributions of snow samples
within the same snow horizon with dirty layers, another snowpack sample (no.
4′) was collected at ca. 1 m away from the location of sample no. 4.
Because the thickness of dirty layers in the snow pit sequence is ca. 10 cm
or more, brown-colored particles are deposited together with snowflakes
during snow precipitation rather than dry deposition.
Descriptions of snowpack samples collected from a pit at
Murodo-Daira near Mt. Tateyama, Japan, in 2009 and 2011. Snowpack sample no.
4′ was collected from different snow pit sequences parallel to sample no.
4′.
Bold letters represent dust layers.
Year
Sample ID
Snow depth (cm)
Description
2008–2009
No. 1
325–335
Weak dust layer Asian dust events were observed on 2 February 2009 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 2
410–420
Clean snow layer
No. 3
425–435
Dusty snow layer Asian dust events were observed on 1 January 2009 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 4
520–530
Dusty snow layer Asian dust events were observed on 10 December 2008 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 4'
520–530
Dusty snow layer
No. 5
530–540
Granular snow
2010–2011
No. 6
115–125
Granular snow with ice plate
No. 7
169–178
Dusty and granular snow Asian dust events were observed on 22–24 February 2011 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 8
290–300
Compacted snow layer
No. 9
390–400
Compacted snow layer
No. 10
400–410
Dusty and compacted snow Asian dust events were observed on 31 December 2010 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 11
430–440
Compacted snow layer
No. 12
460–466
Dusty and compacted snow Asian dust events were observed on 25–26 December 2010 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 13
507–527
Compacted snow with ice plate
No. 14
542–548
Dusty and compacted snow Asian dust events were observed on 6 December 2010 by a lidar over Toyama. Air masses are derived from the Taklamakan and Gobi deserts.
No. 15
590–605
Granular and compacted snow
No. 16
630–635
Granular snow
The snow samples were placed in a pre-cleaned glass jar (8 L) using a clean
stainless steel scoop. To avoid microbial degradation of organic compounds,
mercuric chloride (HgCl2) was added to the glass jar prior to collecting
the snow sample. The sample jars were sealed with a Teflon-lined screw cap
and transported to the laboratory in Sapporo within 4 days by a commercial
refrigerated transport service, which kept the samples in darkness at ca.
5 ∘C and constant humidity. The samples were stored in a dark
refrigerator room at 4 ∘C prior to analysis.
We also analyzed the reference dust materials (Kosa) including Chinese loess
deposits from the Tengger (CJ-1, < 250 µm and CJ-2,
< 100 µm) and Gobi deserts (Gobi, < 10 µm). The
reference materials were purchased from the National Institute for
Environmental Studies, for the measurements of LMW monocarboxylic acids,
inorganic ions, and DOC. Reference dust samples (0.1 g) were extracted with
ultra-pure water by the methods as described below. The detailed information
of reference samples is reported elsewhere (Nishikawa et al., 2000, 2013).
Chemical analysis
Monocarboxylic acids were determined as p-bromophenacyl esters using the
capillary gas chromatography (GC) and GC-mass spectrometry (GC-MS) methods
(Kawamura and Kaplan, 1984); 150 mL of melted snow samples were transferred
to a pear-shaped glass flask (300 mL). To avoid the evaporative loss of
volatile monocarboxylic acids from samples during analytical procedure, pH
was adjusted to 8.5–9.0 by adding several drops of 0.05 M KOH solution to
form organic acid salts (e.g., CH3COO-K+). The sample was
concentrated down to 10 mL using a rotary evaporator under vacuum
(20 mm Hg) at 50 ∘C. The concentrates were filtered through quartz
wool packed in a Pasteur pipette. The filtrates were concentrated down to
0.5 mL. To convert all organic acids to RCOO-K+ form, the
concentrates were passed through a glass column (Pasteur pipette) packed with
cation exchange resin (DOWEX 50W-X4, 100–200 meshes, K+ form). Organic
acids were eluted with pure water and transferred in a 25 mL pear-shaped
flask. The pH of the sample was checked to be 8.5–9.0 and then dried using a
rotary evaporator under vacuum (20 mm Hg), followed by blow-down with pure
nitrogen gas for 30 s. The former process generally requires 15–20 min.
Acetonitrile (4 mL) was added to the dried sample, and RCOO-K+
salts were reacted with α, p-dibromoacetophenone (0.1 M,
50 µL) as a derivatization reagent and dicyclohexyl-18-crown-6
(0.01 M, 50 µL) as a catalyst to derive p-bromophenacyl esters at
80 ∘C for 2 h. The reaction mixture was dried using a rotary
evaporator under vacuum at 30 ∘C. The derived esters were dissolved
in 0.5 mL of n-hexane/dichloromethane (2 : 1) mixture and then purified on
a silica gel column (Pasteur pipette). Excess reagent was eluted with
n-hexane/dichloromethane (2 : 1) mixture (7 mL) and then p-bromophenacyl
esters were eluted with dichloromethane/methanol (95 : 5) mixture (2 mL)
into a glass vial (2 mL). The esters were dried by blow-down using pure
nitrogen gas and then dissolved in n-hexane (100 µL). In addition,
the esters of hydroxyacids (lactic and glycolic acids) were reacted with
N,O-bis-(trimethylsilyl) trifluoroacetamide (BSTFA) with 1 %
trimethylsilyl chloride and 10 µL of pyridine to derive
trimethylsilyl (TMS) ethers for the hydroxyl (OH) group at 70 ∘C for
3 h.
p-Bromophenacyl esters and their TMS ethers were determined using a capillary
gas chromatograph (HP GC6890, Hewlett-Packard, USA) equipped with a flame
ionization detector. The esters were separated using a fused silica capillary
column (HP-5, 30 m × 0.2 mm i.d., film thickness
0.5 µm). The derivatives were also analyzed by GC-MS (Agilent
GC7890A and 5975C MSD, Agilent, USA). The compounds were identified by
comparing GC retention time and mass spectra of authentic standards. Details
of analytical procedure were described previously (Kawamura et al., 2012),
except for the pH adjustment with KOH solution. We tested the recoveries of
authentic monocarboxylic acid standards (C1–C10,
iC4–iC6, benzoic, toluic, lactic, and glycolic acids) that were
spiked into ultra-pure water. The results showed that the recoveries of
organic acids were better than 80 %. Analytical errors in the GC/FID
analysis using authentic standards were within 2 %. Total relative
standard deviations based on triplicate analysis of real samples were within
12 %. Detection limits of organic acids were estimated to be
0.001–0.004 ng g-1.
To measure inorganic ions, samples were passed through a membrane disk filter
(0.22 µm, Millipore Millex-GV, Merck, USA) and the filtrates were
injected into an ion chromatograph (Model 761 compact IC, Metrohm,
Switzerland) equipped with an AS-09 autosampler (Kawamura et al., 2012).
Anion analysis was conducted using a Shodex SI-90 4E column and a 1.8 mM
Na2CO3+ 1.7 mM NaHCO3 solution as eluent. Cation analysis
was conducted using a C2-150 column and a 4.0 mM tartaric acid + 1.0 mM
dipicolinic acid solution as eluent. The total analytical precision is
4 % (Miyazaki et al., 2010).
After removing the particles in the samples on a disk filter
(0.22 µm, Millipore Millex-GV, Merck, USA), DOC was determined
using a total organic carbon (TOC) analyzer (Model TOC-Vcsh, Shimadzu)
(Miyazaki et al., 2011).
Non-sea-salt ions
Concentrations of non-sea-salt ionic species X(Mnss-x) were
estimated by the following equation:
Mnss-x=Mx-(X/Na)swMNa,
where Mx and MNa are the concentrations of X and of Na,
respectively. (X/Na)sw means the mass ratio of species X to Na
in seawater (Duce et al., 1983). The ratios are 0.25 (SO42-), 0.037
(K+), 0.038 (Ca2+), and 0.12 (Mg2+) (Berg and Winchester,
1978). The ratio of F- is 0.000146 (Yang et al., 2009).
Example of lidar measurements of dusts obtained at Imizu, Toyama
(ca. 40 km northwest of Mt. Tateyama) during 1–31 December 2008. The color
scale indicates the extinction coefficient of dust particles based on lidar
measurements. Black line represents clouds and gray shade above the black
lines represents no data.
Seven-day airmass back trajectories at a level of 3000 m a.s.l.
over the Murodo-Daira site in (a) 2008–2009 and
(b) 2010–2011. Color lines show the trajectories associated with
dust layers as observed by a lidar.
Lidar observation and back trajectory analysis
We detected Asian dust events by the lidar observation (data are provided by
the National Institute for Environmental Studies) over Imizu
(36.70∘ N, 137.10∘ E), ca. 40 km northwest of Mt.
Tateyama, Toyama Prefecture, Japan, during December to March in each year.
The observation wavelength of the laser is 532 nm. Details of the extinction
coefficient of dust particles were given in Shimizu et al. (2004). One
example of a lidar image is presented in Fig. 2. Dense dust layers were
recorded at the upper layers (3–4 km) over Imizu on 10 December 2008, whose
dust event should be recorded in the snow pit sequences collected in 2009
(possibly corresponds to no. 4; see Table 1). This dust event was also
recognized by the lidar observations at Niigata, Sendai, and Tsukuba in
Japan. We estimated that Asian dust events observed on 10 December, 1
January, and 2 February during 2008–2009 and 6, 25–26 and 31 December, and
22–24 February during 2010–2011 correspond to sample ID nos. 4, 3, 1, no.
14, 12, 10, and 7, respectively (Table 1).
To investigate the source of air masses during the snow season (November to
April), 7-day backward air mass trajectories were calculated at a level of
3000 m a.s.l. using an online program, Meteorological Data Explorer
(METEX), which was developed by the National Institute for Environmental
Studies (NIES), Japan. Meteorological data were obtained from the National
Centers for Environmental Prediction (NCEP) Reanalysis data. Figure 3 shows
the back air mass trajectories corresponding to selected dust layers
(Table 1). The heights of air masses over the Asian continent and the Sea of
Japan ranged from 2500 to 6000 m.
Concentrations (ng g-1) of monocarboxylic acids in snowpack
samples collected from a snow pit sequence at Murodo-Daira near Mt. Tateyama,
Japan, in 2009 and 2011 and reference dust materials. Bold letters represent
dust layers.
Acid species
Snow sample ID (2009)
Snow sample ID (2011)
Reference dust materials
No.1
No.2
No.3
No.4
No.4′
No.5
No.6
No.7
No.8
No.9
No.10
No.11
No.12
No.13
No.14
No.15
No.16
CJ-1
CJ-2
Gobi
Aliphatic acids
Formic, C1
476
137
344
99.4
112
41.8
2.21
21.3
8.05
6.38
34.4
16.1
62.0
5.41
55.4
15.1
19.7
2420
3940
4402
Acetic, C2
708
273
456
121
140
51.2
9.01
52.6
31.0
25.1
61.5
36.4
50.8
21.9
40.1
22.2
25.6
1435
18540
11170
Propionic, C3
66.9
14.2
37.1
5.48
6.57
2.66
1.64
8.57
3.21
2.28
8.52
6.04
7.30
3.71
4.78
3.95
1.77
95
770
98
Isobuthyric, iC4
5.09
2.37
3.17
0.90
1.08
0.51
0.36
1.03
0.11
0.10
1.35
0.80
1.15
0.56
0.79
0.69
0.35
n.d.
n.d.
17
Butyric, C4
10.0
4.21
7.20
1.31
1.64
0.85
0.60
1.12
0.39
0.32
2.60
1.52
1.76
1.20
1.46
1.31
0.35
39
319
44
Isopentanoic, iC5
40.3
37.4
114
44.9
53.4
30.6
0.55
2.02
1.34
1.37
2.56
1.22
3.66
1.75
2.03
0.68
1.33
3
581
271
Pentanoic, C5
2.55
1.51
2.41
0.92
1.12
0.48
0.33
0.71
0.09
0.08
1.04
0.44
0.58
0.47
0.57
0.52
0.17
15
204
25
Isohexanoic, iC6
n.d.
n.d.
n.d.
n.d.
n.d.
0.12
0.03
0.04
0.01
n.d.
0.08
0.01
0.03
0.02
0.06
0.09
n.d.
n.d.
n.d.
n.d.
Hexanoic, C6
4.03
3.70
4.38
4.23
5.19
1.46
0.76
0.93
0.09
0.04
1.59
0.50
0.83
0.60
1.37
0.74
0.58
5
227
39
Heptanoic, C7
2.01
1.03
2.09
0.61
0.73
0.74
0.03
0.41
0.11
0.08
1.39
0.11
0.49
0.25
0.29
0.37
0.13
3
82
0
Octanoic, C8
1.04
0.25
1.38
1.03
1.26
0.55
0.16
0.15
0.02
0.01
0.46
0.11
0.09
0.15
0.23
0.19
0.07
4
125
14
Nonanoic, C9
6.47
7.23
5.50
3.66
4.78
3.61
1.12
1.38
0.14
0.04
1.62
0.73
0.88
1.08
1.09
0.79
0.66
47
1400
3255
Decanoic, C10
3.57
0.38
2.68
2.40
2.88
1.40
0.14
0.36
0.05
0.38
0.69
0.32
0.32
0.31
0.45
0.42
0.25
n.d.
n.d.
n.d.
Sub-total
1330
481
981
286
331
136
16.9
90.6
44.7
36.1
117.8
64.3
129.9
37.4
108.6
50.4
51.0
4066
26 190
19 340
Aromatic acids
Benzoic, Benz
6.89
3.75
8.74
2.02
2.28
1.29
0.25
1.14
0.12
0.08
3.47
0.61
1.00
0.93
1.98
1.12
0.21
26
62
68
o-toluic
n.d.
n.d.
0.04
0.07
0.06
n.d.
n.d.
0.01
n.d.
n.d.
0.02
n.d.
n.d.
0.01
0.01
0.01
n.d.
n.d.
n.d.
n.d.
m-toluic
0.44
0.71
0.50
0.33
0.37
0.30
0.03
0.08
n.d.
n.d.
0.05
n.d.
0.04
0.02
0.02
0.02
n.d.
n.d.
n.d.
n.d.
p-toluic
0.09
0.06
0.11
n.d.
0.03
0.03
0.01
0.02
0.00
n.d.
0.07
0.01
0.03
0.02
0.03
0.02
0.01
n.d.
n.d.
n.d.
Sub-total
7.42
4.53
9.39
2.42
2.74
1.62
0.29
1.25
0.12
0.08
3.61
0.62
1.06
0.97
2.04
1.16
0.22
26
62
68
Hydroxyacids
Lactic, Lac
1.46
1.11
5.06
1.73
1.89
1.26
0.16
0.01
0.15
0.28
0.07
0.15
0.21
0.21
0.27
0.38
0.14
192
2124
1215
Glycolic, Glyco
0.08
0.19
0.70
0.28
0.33
0.16
0.04
0.01
0.15
0.19
0.05
0.12
0.21
0.31
0.30
0.32
0.20
112
1020
385
Sub-total
1.55
1.30
5.76
2.02
2.22
1.42
0.20
0.02
0.30
0.47
0.12
0.26
0.42
0.52
0.56
0.70
0.34
304
3144
1600
DOC
1360
508
2380
865
936
469
507
904
544
381
1580
723
427
743
704
2110
576
73 000
403 000
267 000
Total MA-C/DOC (%)
35.6
36.7
15.8
13.1
14.1
12.1
1.5
3.9
3.2
3.7
2.9
3.5
10.8
2.1
5.4
1.0
3.2
2.0
2.9
3.3
Results
Tateyama snow pit samples
Homologous series of low molecular weight normal aliphatic
(C1–C10), branched chain (iC4–iC6), hydroxy (lactic and
glycolic), and aromatic (benzoic acid and o-, m-, and p-toluic acid isomers)
monocarboxylic acids were detected in the snow pit samples (Table 2). We
found that differences in the concentrations of each monocarboxylic acid
between sample nos. 4 and 4′ are comparable to the total relative standard
deviations based on triplicate analysis of real samples. Thus, we consider
that each horizontal layer in the snow pit site is homogenous and that each
snow sample is representative of the snowfall events over the Murodo site.
Concentrations of selected low molecular weight monocarboxylic acids
in Mt. Tateyama snow samples.
Concentrations (ng g-1) of major ions and pH in snowpack
samples collected from a snow pit sequence at Murodo-Daira near Mt. Tateyama,
Japan, in 2009 and 2011 and reference dust materials. Bold letters represent
dust layers.
Inorganic
Snow sample ID (2009)
Snow sample ID (2011)
Reference dust materials
species
No.1
No.2
No.3
No.4
No.4'
No.5
No.6
No.7
No.8
No.9
No.10
No.11
No.12
No.13
No.14
No.15
No.16
CJ-1
CJ-2
Gobi
Anion
F-
96
15
115
3
42
20
19
27
16
15
94
20
10
17
27
21
10
374 000
148 000
43 600
MSA-
1080
62
1250
204
172
83
83
62
64
75
128
93
51
129
117
66
74
665 000
224 000
215 000
NO3-
534
130
458
316
224
150
791
1340
208
114
2020
88
104
428
1120
843
174
126 000
1 376 000
138 000
SO42-
1250
430
1460
728
536
364
845
1360
439
315
3010
310
260
549
1330
1070
282
2 493 000
24 038 000
1 853 000
Total
2960
637
3280
1250
974
617
1740
2790
727
520
5250
511
424
1120
2600
2000
540
3 659 000
25 806 000
2 251 000
nss–F-
96
15
115
3
42
19
19
27
16
15
94
20
9
17
27
21
10
374 000
148 000
43 400
nss–SO42-
434
112
155
325
120
98
756
903
293
220
2440
215
n.d.
453
1040
967
192
2 008 000
23 010 000
1 592 000
Cation
Na+
3240
1270
5210
1610
1660
1060
356
1840
586
380
2310
380
1420
385
1160
417
362
1 942 000
4 111 000
1 047 000
NH4+
111
41
200
268
243
78
235
291
54
36
842
47
34
56
517
190
41
336 000
1 460 000
18 700
K+
215
16
292
148
105
96
86
111
n.d.
50
302
n.d.
40
n.d.
119
43
n.d.
943 000
4 614 000
2 148 000
Ca2+
3120
485
3390
1600
1890
505
184
639
148
140
1060
113
515
n.d.
574
220
200
10 798 000
18 877 000
8 864 000
Mg2+
190
6
195
334
152
24
35
127
n.d.
13
78
n.d.
19
n.d.
35
33
n.d.
1 869 000
1 045 000
754 000
Total
6880
1817
9290
3960
4050
1760
896
3010
787
619
4590
540
2020
442
2410
903
602
15 908 000
30 107 000
12 831 000
nss–K+
95
n.d.
99
88
44
57
72
43
n.d.
36
217
n.d.
n.d.
n.d.
76
28
n.d.
871 000
4 462 000
2 110 000
nss–Ca2+
3000
436
3190
1540
1820
464
170
569
125
125
976
99
462
n.d.
530
204
186
10 725 000
18 721 000
8 824 000
nss–Mg2+
n.d.
n.d.
n.d.
140
152
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
1 636 000
552 000
628 000
pH
6.9
6.1
6.7
6.7
6.3
6.0
4.7
6.0
5.2
5.0
6.2
5.1
6.2
4.9
5.9
4.4
5.4
–
–
–
Acetic acid (C2) was found to be the dominant species (2009:
51.2–708 ng g-1; 2011: 9.01–61.5 ng g-1), followed by formic
acid (C1) (2009: 41.8–476 ng g-1; 2011:
2.21–62.0 ng g-1). Concentrations of C3–C10 acids were
1–2 orders of magnitude lower than C2. In contrast, iC5 acid
(2009: 30.6–114 ng g-1; 2011: 0.55–3.66 ng g-1) was detected
as the most abundant branched chain acid. Lactic and glycolic acids were also
detected as hydroxyacids in the snow pit samples. Concentrations of lactic
and glycolic acids are 1 and 2 orders of magnitude lower than those of major
monocarboxylic acids (C1 and C2), respectively. The concentration
of benzoic acid ranged from 0.08 to 8.74 ng g-1. Total concentrations
of toluic acid isomers were found to be significantly lower (average
0.07 ng g-1) than that of benzoic acid (2.11 ng g-1). Average
concentrations of total monocarboxylic acids in the dust layers (2009:
739 ng g-1; 2011: 114 ng g-1) were greater than those without
dust layers (2009: 313 ng g-1; 2011: 43 ng g-1) (Fig. 4).
Concentrations of DOC ranged from 469 to 2380 ng g-1 in 2009 and 381
to 2110 ng g-1 in 2011 (Table 2). The highest concentration of DOC
(2380 ng g-1) was found in sample no. 3, in which a dust layer was
observed.
We detected cations (Ca2+, Na+, Mg2+, K+, and
NH4+) and anions (F-, NO3-, SO42-, and
MSA-) in snow pit samples collected in both 2009 and 2011 from the
Murodo-Daira site near Mt. Tateyama (Table 3). Concentrations of
nss–Ca2+, nss–Mg2+, nss–K+, nss–F-, and
nss–SO42- were calculated as shown in Table 3. NO3- and
nss–SO42- are two major anions. The highest concentrations of
NO3- (2020 ng g-1) and nss–SO42- (2440 ng g-1)
were obtained in sample no. 10, in which a dust layer was observed. On the
other hand, Na+ and nss–Ca2+ are two major cations. Higher
concentrations of Na+ and nss–Ca2+ were found in sample nos. 1
(Na+: 3240 ng g-1; nss–Ca2+: 3000 ng g-1) and 3
(Na+: 5210 ng g-1; nss–Ca2+: 3190 ng g-1), both of
which showed the presence of a dust layer. The pH of melt snow samples ranged
from 4.4 to 6.9 (Table 3). Higher pH was found in sample nos. 1, 3, and 4
(pH = 6.7–6.9), in which dust layers were observed.
Average concentrations of C1 (202±170 ng g-1), C2
(292±249 ng g-1), and iC5 (53.5±30.8 ng g-1)
in 2009 are 1 order of magnitude higher than those in 2011 (C1:
22.4± 20.1 ng g-1; C2: 34.2± 15.8 ng g-1;
iC5: 1.69± 0.88 ng g-1). Similar tends were found for the
average concentrations of minor monocarboxylic acids (C3–C10,
iC4, and iC6) in the snow pit samples in 2009 and 2011. The average
concentration of DOC in 2009 (1090± 712 ng g-1) is slightly
higher than that in 2011 (836±534 ng g-1). The contribution of
total monocarboxylic acids to DOC (total MCA-C / DOC) in 2009 (21.2±11.6 %) is 6 times higher than that in 2011 (3.75±2.62 %).
Average concentrations of NO3- (657±633 ng g-1) and
nss–SO42- (748±682 ng g-1) in 2011 are 2–3 times
higher than those in 2009 (NO3-: 302±166 ng g-1;
nss–SO42-: 207±139 ng g-1). In contrast, average
concentrations of nss–Ca2+ in 2009 (1740±1190 ng g-1) are
5 times higher than those in 2011 (345±285 ng g-1).
Reference dust materials
We detected LMW monocarboxylic acids, inorganic ions, and DOC in the water
extracts from three reference dust materials (CJ-1, CJ-2, and Gobi) (Tables 2
and 3). Concentrations of total LMW monocarboxylic acids in the reference
dusts were 4370 ng g-1 (CJ-1), 29 390 ng g-1 (CJ-2), and
21 010 ng g-1 (Gobi). The dominant LMW monocarboxylic acids were
formic and acetic acids. Concentrations of DOC were 73 000 ng g-1
(CJ-1), 403 000 ng g-1 (CJ-2), and 267 000 ng g-1 (Gobi).
Total MCA-C/DOC ratios in reference dust materials were 2.0 % (CJ-1),
2.9 % (CJ-2), and 3.3 % (Gobi). Concentrations of nss–Ca2+ in
the reference dust materials were 10 700 µg g-1 (CJ-1),
18 700 µg g-1 (CJ-2), and 8820 µg g-1
(Gobi).
Scatter plot of concentrations of formic plus acetic acids vs.
nss–Ca2+ in Mt. Tateyama snow samples. The dotted line represents the
Deming linear regression.
Discussion
Influence of Asian dust
High concentrations of nss–Ca2+ were obtained in the dust layers of
both 2009 and 2011. Ca2+ is known as a major metal ion to be transported
from arid regions in North Asia with Asian dust (Mori et al., 2002; Tsai and
Chen, 2006). In this study, contributions of nss–Ca2+ to Ca2+ in
2009 and 2011 are 95 and 91 %, respectively. In addition, the mass
concentration ratios of Mg / Ca at the Murodo-Daira site in 2009 and 2011
are 0.08 and 0.12, respectively. These values are comparable to those in
reference dust materials such as CJ-1 (0.17), CJ-2 (0.06), and Gobi (0.09).
Therefore, nss–Ca2+ can be used as an indicator of mineral dust. High
abundances of nss–Ca2+ in snowpack samples indicate that a strong
outflow of dust particles from the Asian continent was involved with a heavy
snow precipitation.
To investigate the effect of Asian dust on LMW monocarboxylic acids, we
plotted major LMW monocarboxylic acids (i.e., formic plus acetic acids)
against nss–Ca2+ using all the data points (Fig. 5). Concentrations of
formic plus acetic acids were found to increase linearly with that of
nss–Ca2+ (r= 0.88). The air mass trajectories have passed over the
Asian continent including North China and Mongolia (Fig. 3). Asian dust
particles may be a carrier of formic and acetic acids via acid–base
interaction, forming carboxylate salts, when the Asian dust activity
maximizes in North China. The pathways of long-range transport and sources of
formic and acetic acids will be discussed in the following Sects. 4.2 and
4.3.
Average concentrations of formic and acetic acids and nss–Ca2+ in 2009
are higher than those in 2011. This may be related to a strong influence of
the Asian dust events, although the detailed records of the Asian dust events
in North China are not available at this moment.
Scatter plots of the natural logarithm of formic plus acetic acids
and pH, and natural the logarithm of nss–Ca2+ and pH. The solid and
dotted lines represent the Deming linear regression.
Long-range transport of formic and acetic acids and aerosol
acidity/alkalinity
Figure 5 presents the relationship between formic plus acetic acids and the
pH of melt snow. Concentrations of formic plus acetic acids were found to
increase exponentially with pH (r= 0.87). Interestingly, concentrations
of nss–Ca2+ were also found to increase exponentially with pH (r= 0.89) (Fig. 6). Because LMW monocarboxylic acids have high vapor pressure
(Saxena and Hildeman, 1996), they should be largely present as gases in the
atmosphere (e.g., Kawamura et al., 1985; Liu et al., 2012). During long-range
atmospheric transport, alkaline dust particles may be subjected to
atmospheric titration by gaseous monocarboxylic acids.
We calculated ion balance in the snow pit at the Murodo-Daira site near Mt.
Tateyama. In this study, we could not use the data of a chloride ion
(Cl-) because of the addition of HgCl2 into snow samples as
bactericide. To calculate ion balance, we used equivalent ratios of Cl-
to Na+ (1.26) obtained in the same snow pit in 2011 (Watanabe et al.,
2012). Figure 7 shows total cations (Na+, NH4+, K+,
Ca2+, and Mg2+) against total anions (F-, MSA-,
NO3-, SO42-, and organic anions including normal
(C1–C10), branched chain (iC4–iC6), aromatic (benzoic
and toluic acid isomers), and hydroxyl (lactic and glycolic) monocarboxylic
acids) (r= 0.95). The slope (1.26) of more than unity indicates that
excess cations exist in the snow pit at the Murodo-Daira site near Mt.
Tateyama, although CO3- and HCO3-, and unidentified organic
anions were not taken into consideration.
Linear regression plots between total cation equivalents (neq) and
total anion equivalents (neq) in melt snow samples at the Murodo-Daira site
near Mt. Tateyama.
We calculated the ratios of formic plus acetic acids / nss–Ca2+ for
the Murodo-Daira snow pit samples and compared the ratios of formic plus
acetic acids / nss–Ca2+ in the reference materials such as CJ-1,
CJ-2, and Gobi. We found that formic plus acetic acids / nss–Ca2+
ratios for the Murodo-Daira snow pit samples (ave. 0.27) are significantly
higher than those from CJ-1 (0.00036), CJ-2 (0.0012), and Gobi (0.0018)
reference samples collected from the arid areas of North China. These results
indicate that alkaline dust particles can adsorb gaseous MCAs in the
atmosphere and largely control the long-range transport of LMW monocarboxylic
acids from the Asian continent to the western North Pacific Rim. Based on a
good correlation between monocarboxylic acids and nss–Ca2+, it is very
likely that organic acids in aerosols exist in the form of salts such as
Ca(HCOO)2, Ca(HCOO)(CH3COO), and/or Ca(CH3COO)2.
Scatter plots of (a) concentrations of benzoic acid vs.
nss–Ca2+, (b) formic plus acetic acids vs. benzoic acid,
(c) formic plus acetic acids vs. nss–K, and (d) formic
plus acetic acids vs. nss–F- in Mt. Tateyama snow samples. The dotted
line represents the Deming linear regression.
Prince et al. (2008) reported that
gas-phase acetic acid is adsorbed on the surface of calcite (CaCO3), a
major mineral of dust particles. Acetic acid can form calcium acetate in the
atmosphere (Alexander et al., 2015). Vapor pressures of those organic anions
are significantly lower than those of free monocarboxylic acids. In addition,
the lifetimes of formic and acetic acids with OH radicals are estimated to be
25 and 10 days, respectively, at -13 ∘C assuming the OH
concentration of 1.0×106 molecules cm-3 (Paulot et al.,
2011). This timescale is much longer than that of the atmospheric transport
time of air mass from the Asian continent to Mt. Tateyama. Therefore, the
acidity/alkalinity of an aerosol surface is an important factor in
controlling the uptake of gaseous organic acids, and thus organic acid salts
can be long-range transported as particles in the atmosphere from the Asian
continent to the Japanese islands. Zhang et al. (2012) reported that pH of
wet deposition for the last 2 decades showed a slight increase in the
southeastern Tibetan Plateau, China, due to the presence of Ca2+ that is
derived from Asian dust. We suggest that long-range atmospheric transport of
LMW monocarboxylic acids associated with Asian dust over the Japanese islands
would be changed in the future due to the changes in the emission of Asian
dusts from the Asian continent that are associated with global warming and
changes in land use (Zhang et al., 2003; Song et al., 2016).
Scatter plot of concentrations of branched chain
(iC4–iC6) monocarboxylic acids vs. lactic acid in Mt. Tateyama
snow samples. The dotted line represents the Deming linear regression.
Scatter plots of (a) concentrations of branched chain
(iC4–iC6) monocarboxylic acids vs. nss–Ca2+ and
(b) lactic acid vs. nss–Ca2+ in Mt. Tateyama snow samples. The
dotted line represents the Deming linear regression.
Major contributions of anthropogenic monocarboxylic acids
Benzoic acid is directly emitted from fossil fuel combustion (Kawamura et
al., 1985) and also produced in the atmosphere by photo-oxidation of aromatic
hydrocarbons such as toluene (Forstner et al., 1997), which are derived from
human activities. Benzoic acid positively correlated with nss–Ca2+ (r= 0.90) (Fig. 8a). In addition, the average benzoic
acid / nss–Ca2+ ratio obtained for the Murodo-Daira snow pit
samples (0.0029) is 3–4 orders of magnitude higher than those obtained from
the Kosa reference materials such as CJ-1 (0.0000024), CJ-2 (0.0000033), and
Gobi (0.0000078). Benzoic acid may also be adsorbed on the pre-existing
particles via atmospheric titration of alkaline dust particles derived from
the Asian continent. The air mass trajectories arriving at the Murodo-Daira
site have passed over North China, where many industrial regions and
mega-cities (e.g., Beijing) are located (Fig. 3).
Formic plus acetic acids showed a strong positive correlation with benzoic
acid (r= 0.90) (Fig. 8b), indicating that they are derived from
anthropogenic sources in the Asian continent. In contrast, nss–K+, a
tracer of biomass burning (Zhu et al., 2015), did not show a positive
correlation with formic plus acetic acids (r= 0.18) (Fig. 8c).
nss–F-, a tracer of coal burning (Wang et al., 2005), shows a positive
correlation with formic plus acetic acids (r= 0.72) (Fig. 8d); however,
they were rather scattered. Biomass and coal burning is not a major source of
monocarboxylic acids in the snow pit samples collected from the Murodo-Daira
site near Mt. Tateyama. We consider that formic and acetic acids are both
derived from anthropogenic and photochemical processes in the atmosphere of
North China. They are adsorbed on the pre-existing alkaline Kosa particles
via the atmospheric titration during a long-range atmospheric transport over
the Japanese islands.
The mean concentrations of formic and acetic acids in our samples in 2009 are
higher than those reported in mountain snow samples from southern California
(Kawamura et al., 1996), Tateyama (Kawamura et al., 2012) and the southern
French Alps (Maupetit and Delmas, 1994), and ice core samples from Antarctica
(de Angelis et al., 2012). The total MCA-C / DOC ratio (av. 21 %) in
2009 is significantly higher than those reported in rainwater samples from
Los Angeles (4.4 %) (Kawamura et al., 2001), Shenzen, China (2.3 %)
(Huang et al., 2010), and reference dust materials (CJ-1: 2.0 %; CJ-2:
2.9 %; and Gobi: 3.3 %). These results indicate that water-soluble
LMW monocarboxylic acids in the snow pit samples near Mt. Tateyama constitute
a significant fraction of water-soluble organic carbon, suggesting that
entrainment of organic acids in alkaline dusts and snowflakes is significant
during the atmospheric transport from China to Japan.
Minor contributions of biogenic monocarboxylic acids
Branched chain (iC4–iC6) monocarboxylic acids are produced by
bacterial activity of Bacteroides ruminicola, Megasphaera elsdenii, and Streptomyces avermitilis (e.g., Allison, 1978; Hafner
et al., 1991). It is of interest to note that iC5 has not been reported
in motor exhaust (Kawamura et al., 2000) and urban rainwater (Kawamura et
al., 1996). Bacteria (lactobacillus) and plant tissues are known to
produce lactic acid (Cabredo et al., 2009; Baker and El Saifi, 1953).
Lactobacillus mainly exists in soil (Huysman and Verstraete, 1993).
We found a strong positive correlation between branched chain
(iC4–iC6) acids and lactic acid (r= 0.98) (Fig. 9). This
strong correlation suggests that these organic acids are closely linked in
the biosynthetic processes associated with bacterial activity in soils.
Branched chain (iC4–iC6) acids (r= 0.85) (Fig. 10a) and lactic
acid (r= 0.81) (Fig. 10b) showed a positive correlation with
nss–Ca2+. Maki et al. (2011, 2014) reported that bacterial communities
are present in the layers of snow pit sequences at Murodo-Daira near the
summit of Mt. Tateyama and are considered to be associated with Asian dust
events. Bacterial species responsible for branched monocarboxylic and lactic
acids have not been reported in the Tateyama snow samples at this time.
However, our results suggest that branched chain monocarboxylic acids may be
produced by bacterial process in soils of the Asian continent and transported
over the Japanese islands with Asian dust. The contribution of biogenic
monocarboxylic acids is much lower than anthropogenic monocarboxylic acids.
Summary and conclusions
Low molecular weight normal (C1–C10), branched chain
(iC4–iC6), hydroxyl (lactic and glycolic), and aromatic (benzoic
and toluic isomers) monocarboxylic acids were detected in the snow pit
samples collected from Murodo-Daira snowfield near the summit of Mt.
Tateyama, central Japan. Acetic acid was detected as the dominant species
(125 ng g-1), followed by formic acid (85.7 ng g-1) and
isopentanoic acid (20.0 ng g-1). Enhanced concentrations of
monocarboxylic acids and nss–Ca2+ were obtained in the snow pit samples
with dust layers. We found that abundances of formic and acetic acids largely
depend on non-sea-salt Ca2+ (r= 0.88). These acids positively
correlated with benzoic acid (r= 0.90) that is primarily produced by
fossil fuel combustion and secondary photochemical oxidation of anthropogenic
toluene and other aromatic hydrocarbons, indicating that monocarboxylic acids
were mainly of anthropogenic and photochemical origin. Formic plus acetic
acids exponentially correlated with pH (r= 0.87) (pH = 4.7–6.9).
Alkaline dust particles may be subjected to atmospheric titration by gaseous
monocarboxylic acids.
In addition, we analyzed reference dust materials including Chinese loess
samples from the Tengger and Gobi deserts for the measurements of LMW
monocarboxylic acids and inorganic ions. The ratio of total monocarboxylic
acid / nss–Ca2+ at the Murodo-Daira snow pit samples (0.27) was
found to be significantly (2 to 3 orders of magnitude) higher than those of
Chinese loess reference samples (0.00036–0.0018). These comparisons suggest
that gas-phase monocarboxylic acids are easily adsorbed on the surface of
pre-existing dust particles derived from the Asian continent to result in
organic acid salts. Our study demonstrates that Asian dust is a key factor in
promoting a long-range atmospheric transport of LMW monocarboxylic acids
emitted and produced over North China to the western North Pacific Rim under
a strong influence of the East Asian winter Monsoon. By forming the organic
acid salts, LMW monocarboxylic acids can be more stabilized against the
photochemical decomposition during long-range atmospheric transport.