1 Introduction

Atmospheric particulate matter (PM) plays a vital role in the global climate system [1]. It directly affects the climate through the processes of scattering, transmission, and absorption of solar radiation. PM acts as cloud condensation nuclei (CCN), thereby indirectly affecting the climate system [2]. The chemical nature of the PM impacts the CCN processes. PM also adversely impacts human health [3]. Due to its effects on human health and the environment, PM has been listed as a criteria air pollutant by countries worldwide to regulate it for its control [4]. Inhalation of organic species and heavy metals associated with the fine particles (PM2.5: PM ≤ 2.5 µm) causes serious health effects [5]. Bioaerosols also adversely affect humans, animals, and plants [6].

Morphology and composition have a bearing on their physical, chemical, and optical properties [7]. These properties undergo change with respect to space and time and differentially affect human health, visibility, and energy balance on a regional scale [8 and references therein]. Detailed characterization of individual atmospheric particles is indicative of their sources of origin, transportation, and removal [9]. Morphology and chemical composition of the particles control their settling velocity by applying a drag force and hence affect the transportation process of the particles [10, 11]. Morphological characteristics like shape, surface roughness, and edge sharpness affect the scattering property of a particle [12]. The chemistry of PM is also the driver of the climatic effects of PM [13].

The carbonaceous aerosols play a greater role in absorbing and scattering the incoming solar radiations and also indirectly affecting the hydrological cycle [14]. The morphology of carbonaceous aerosols is also a key attribute considered while calculating the global warming contribution of carbonaceous particles [15].

The mineral dust particles are an important constituent of atmospheric aerosols and are contributors to total atmospheric aerosol loading. Mineral dust makeup to about 40% of global aerosol emissions from natural sources [16]. They help in the removal, deposition, and transport of atmospheric pollutants [17] including the reduction in the levels of ambient ozone to an extent of 5.5% [18]. Mineral aerosol particles carried by dust storms can be transported to a long distance, which may have impacts of regional and global scales [19].

Characterization of dust aerosols reveals important information on their source of origin [20, 21]. Mineralogical, chemical, and physical results of mineral dust could provide useful information for applications in climate modeling, visibility, remote sensing, medical geology, and other studies [22], as the heterogeneous composition of mineral aerosols increases the uncertainty in atmospheric climate models [23].

Dust events add lots of crustal aerosol to the atmosphere, which alters radiation balance when present in higher concentrations [24]. It leads to a decrease in the number of accumulation particles, an increased number of coarse particles, and a favorable condition for new particle formation [25]. Among the sources of dust particles in the atmosphere, rivers are one of the important sources of atmospheric dust particles [26]. The Brahmaputra River of Assam is a trans-boundary river flowing through Tibet, India, and Bangladesh. Earlier studies reported the Aerosol Optical Depth (AOD) over Brahmaputra Valley attributed to dust aerosols [27, 28] and also reported crustal dust and soil resuspension as important sources of PM10 (PM ≤ 10 µm) over the mid-Brahmaputra Valley [29, 30]. Detailed studies on the characterization of particulates during dust events of this region was lacking. Therefore, the present study was taken up and so designed to investigate the physicochemical attributes of atmospheric particulates during a dust event.

2 Experimental methods

2.1 Study site

The study was carried out at Tezpur University, a rural residential institutional area (26°37′ N and 92°50′ E) (Fig. 1). Tezpur is situated on the north bank of the Brahmaputra River and represents the regions around the middle stretch of the river in India. The region is mostly rural with agricultural lands, forests, tea plantations, riverine, and marshy areas. Many small-scale industries that run on coal and wood like tea processing units, brick kilns, auto work stations, and minor agriculture-based industries are growing in the region with time. Moreover, two major national highways NH 15 and NH 715 meet at a point and pass through Tezpur.

Fig. 1
figure 1

a The study area (Tezpur) in the Brahmaputra Valley and the neighboring regions of Asia, b Tezpur locality and sampling location, Tezpur University

Full size image

2.2 Collection of PM

The collection of PM was done by a fabricated grab sampler with an airflow rate of 1.2 L/min. Sampling was carried out at the rooftop of the Department of Environmental Science building, Tezpur University, at a height of 20 m from the ground. The collections were made on 47 mm diameter Whatman Glass Fiber filters, Grade GF/A for morphology study and on polytetrafluoroethylene (PTFE) membrane filters for XRD and FTIR analyses [31]. The pore size of Whatman glass fiber filters and PTFE membrane filter is 1.6 µm and 1.0 µm, respectively. Sampling was carried out during the four-day-long dust event of April 2013. Two sets of samples were collected—(i) 24-h samples and (ii) 3-h samples. The shorter duration samples were used for the single-particle analysis. The wind speed during the sampling period ranged from 2 ms−1 to 9.10 ms−1. Heavy accumulation of coarse and fine particles was found on the filter paper. A 24-h sample was also collected during the non-dusty period (3 days after a rain event during June 2013) to understand the background characteristics of PM. The details of the sampling are provided in Supplement 1.

Meteorological condition is the main driving factor controlling the distribution of air pollution in a region. In the present study, meteorological data were downloaded from wunderground.com (Supplement 2). Temperature, humidity, wind speed, and rainfall data during the sampling days are provided in Supplement 2. No rainfall was recorded in April 2013.

2.3 Analyses

A square of the sample filters (1 mm × 1 mm) was cut and coated with platinum for the morphological analysis under the scanning electron microscope (SEM) (JEOL JSM 6390 LV). Energy-dispersive X-ray spectrometer (EDS) INCAx Sight microanalysis system (Oxford Instruments, Model 7582), hyphenated with the SEM, was used for elemental analysis. The silicon (lithium-drifted) crystal detector of the EDS had an acquisition rate of 50,000 cps. The Super Atmosphere Supporting Thin Window (SATW) of the detector confirms good resolution at the low energy end of the spectrum with a resolution of 137 eV at 5.9 keV. The detector had a minimum quantification limit of 0.01 wt. %. The PM deposited on the PTFE filter was used for single-particle analysis under SEM.

For the XRD analysis, samples collected on a PTFE filter were used because it is very stable and absorbs negligible water or gases [31]. The particles on the PTFE filter were scooped out using a teaser and placed in the quartz sample holder for X-ray diffractometer (XRD, Rigaku, Miniflex) analysis. The X-ray diffraction data were collected at Bragg angle 2θ ranging from 10 to 800 with a scanning speed of 0.050 min−1. The X-ray source was a Cu Kα line with a wavelength of 1.54 Å. The Braggs' equation nλ = 2dhklSinθ was used to calculate the interplanar spacing (dhkl). In this equation, n is the order of reflection (n = 1) and λ is the wavelength of X-ray used (λ = 1.54 Å). The peak positions (2θ) and interplanar spacing (dhkl) were compared with the Joint Committee on Powder Diffraction Standards (JCPDS), version 2.4 database for mineral identification [32].

Samples collected on PTFE filters were also used for Fourier transform infrared (FTIR) study. FTIR spectra from 400 to 4000 cm−1 were recorded on the FTIR spectrometer (Perkin Elmer; Frontier MIR-FIR). 1 mg of dust sample was dispersed with KBr (spectroscopic grade) in the ratio of 1:20. The sample was then pelletized at a pressure of about 1 MPa before measuring the FTIR spectra. The FTIR analysis was carried out at a spectral resolution of 4 cm−1. The reported spectrum is the average of four scans.

To trace the pathway of airmasses during the dust event, 72-h airmass back trajectories at 100 m AGL were computed using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model. The reanalysis data were downloaded from the Gridded Meteorological Data Archives (ready.noaa.gov/archives.php). Geographic Information System (GIS)-based software TrajStat (http://www.meteothink.org/products/trajstat.html) was used to project the HYSPLIT back trajectory clusters using the 72-h trajectories reaching the site hourly on 1st and 3rd April 2013.

2.4 Quality control

Care was taken to prepare contamination-free samples. Filters were desiccated before and after the collection of samples to avoid traces of moisture. Acetone and methanol were used to clean the tweezers, scissors, and sampling stubs before each use. The sample holder of the XRD was cleaned properly before each use.

3 Results and discussion

3.1 Morphological characteristics

SEM micrographs of PM are shown in Fig. 2. Particles of various shapes and sizes—spherical, irregular, cluster, flaky, rod-like, angular, agglomerates chain, and triangular particles—are seen (Fig. 2a and b). Previous workers reported extremely irregular shapes of particles [23, 33]. Aggregation could be an important formation process of irregular shape particles [34]. SEM micrographs of aerosols collected during the non-dusty period showed greater accumulation of carbonaceous agglomerates, explicit chain-like structures, and spherical soot particles (Fig. 2c) [24, 33]. Soot particles are generally emitted from incomplete combustion of biomass and fossil fuels [7]. Carbonaceous agglomerate and chain-like structures may explain the emissions from fossil fuel combustion in vehicles and industries [24, 33]. Spherical shape particles are generally indicative of combustion processes or high-temperature processes [24], which were vividly seen in the micrographs. Thus, based on morphological expression, we have categorized the particles vis-a-vis their origins, viz. biogenic, geogenic, and anthropogenic. This is clearer from high-resolution SEM images (above 10,000×).

Fig. 2
figure 2

SEM micrographs at different magnification showing different morphologies: a, b bulk sample collected at Tezpur during dusty periods, c carbonaceous agglomerate and soot aerosol during non-dusty periods, after a rainfall d clay-like crustal matter along with spherical fly ash particles, e crustal dust with sharp-edged and sheet-like structures, and f image showing the presence of biogenic aerosols like diatom

Full size image

Clay-like structure (sheets) and fine carbonaceous spherical particles were seen (Fig. 2d). The presence of clay-like structures may explain the geogenic origin, which later resuspended from the crustal surface by the wind gust. Figure 2e shows the crustal dust with sharp-edged particles, emitted from anthropogenic construction activity and movement of vehicles on road. We also found diatoms in the bulk sample analysis (Fig. 2f). Figure 3 shows the particles of biogenic origin which include plant debris, pollen, and brochosomes secreted by leafhoppers of the family Cicadellidae—an abundant species of bioaerosols in ambient air [35]. Brochosomes are spherical, honeycomb-like nanoparticles made of proteins and lipids [36]. It acts as a water-repellent protective surface coating [35, 36]. Pollen can be transported up to 100–1000 km [37] and can also act as CCN [38]. Figure 4 shows fine particles of different origins with size measurements. Fine particles smaller than 1.5 µm were emitted from fly ash generated from combustion [24] and fine dust from construction and vehicular movement. Biogenic particles of minimum size ~ 275 nm (Fig. 3d) were also present in the samples. Spherical soot particles and irregular shape particles are seen in Fig. 4a. The spherical soot particulates could be of combustion origin like coal and biomass burning [39]. There are several coal-fed brick kiln industries nearby areas of the sampling station. The irregular-shaped particles could be from crustal and construction activities. Also, soot agglomerates from vehicular emission are seen in Fig. 4b.

Fig. 3
figure 3

Biogenic aerosols: a plant parts, b, c pollen and d brochosomes secreted from insects

Full size image
Fig. 4
figure 4

Fine particles originated from anthropogenic activities: a industrial and construction activities b soot agglomerates from vehicular emission

Full size image

3.2 Composition using energy-dispersive X-ray (EDX) spectroscopy

Elemental analysis of bulk sample (Fig. 5a, spectrum 1) showed the following sequence of abundance: O > C > Si > Al > Fe > Ca > Na > Ti > Mg > K > S (Table 1). The carbon (C) in the PM is a product of combustion coming from biomass and/or fossil fuel burning [40, 41]. The detection of sulfur (S) and potassium (K) together with the C indicated the presence of carbonaceous soot particles which could have originated from agricultural burning or wood burning in household activities and coal burning in factories and brick kiln industries of nearby areas [30]. The presence of oxygen (O), silicon (Si), iron (Fe), magnesium (Mg), sodium (Na), and calcium (Ca) showed a high accumulation of crustal clay minerals coming from resuspended dust (Fig. 5a). The EDX analysis of bulk samples may interfere with the compositional characteristic of particles originating from different sources. So EDX spectrum of single-particle with different morphology in the bulk sample was tried. Figure 5b shows the EDX pattern of two particles; one large particle (Fig. 5b, spectrum 1) of irregular shape with uniform surface and another small spherical shape particle (Fig. 5b, spectrum 2). Both particles contained similar elements (Al, Ca, C, Fe, Mg, Mn, Ni, O, K, Si, Na, Ti, V, and Zn) irrespective of their shape and size (Table 1). Detection of these element showed the presence of alumina silicate, mica, quartz-like minerals and carbonaceous material. Comparison of the present study with other reported works from India showed accumulation of geogenic mineral/clay particles generated by resuspension of dust from soil/road and other anthropogenic particles from the burning of fossil fuel and biomass [31, 39, 42, 43].

Fig. 5
figure 5

a SEM–EDX analysis of bulk sample b SEM–EDX of a large (spectrum 1) and a small (spectrum 2) particle at the bulk sample (the Pt peak was due to the coating used)

Full size image
Table 1 Elemental composition (%) of PM
Full size table

The EDX analyses of single particles collected for a short duration are shown in Fig. 6, and the elemental concentrations are presented in Table 2. Mineral coarse particles with irregular shapes and carbonaceous particles were observed. Along with mineral dust particles, soot of sulfates was observed (Sample-1 spectrum-4 and Sample-2 spectrum-1). The presence of sulfur confirms their origin from the combustion process. These particles probably originated from soil dust, dust resuspension from road and earth crust, and other anthropogenic activities like construction and vehicular movement on-road, combustion activities, and agricultural fields. The irregular shaped particles are formed by two processes; (i) they may form by chemical reactions among pre-existing solid, liquid or gaseous phases, or (ii) by aggregation of pre-existing particles [34]. The elemental composition of these particles mainly consisted of Al, Si, O, C, Na, K, Fe Mg, and Ca indicating the presence of CaCO3 and aluminosilicates which were most likely from geological sources (Table 2) [31, 42]. A high content of Si and Al with varying Mg, K, Fe, and Ca is characteristic of aluminosilicates [44]. The compositions of the particles with smooth spherical shape (Fig. 7a) and biogenic particle (Fig. 7b) were also analyzed, and their elemental concentrations are presented in Table 2. Figure 7a shows the presence of soot/fly ash particles emitted from various kinds of burning processes and coal combustion. Tezpur site is located in a rural area with low or no vehicular traffic, and during the pre-monsoon season (March to May) lots of coal-fed brick kilns industries remain active nearby. The fly ash particles are mainly composed of aluminosilicates and iron and/or calcium-rich particles [45]. The size of fly ash particles ranges between 2 μm and 10 μm and generally spherical [45].

Fig. 6
figure 6

EDX analysis of single particles of two samples along with 4 different spectra of point scanning (samples collected were for a short period; 3-h sampling)

Full size image
Table 2 Elemental composition (%) of individual particles of different origin
Full size table
Fig.7
figure 7

a SEM–EDX of anthropogenic soot particles with spherical morphology b SEM–EDX of biogenic (Diatom) particle

Full size image

Figure 7b shows a biogenic particle—a diatom. Biogenic particles such as pollen with (C+O) > 80% and a minor amount of Na, Si, Al < 10% were observed (Fig. 3). These particles also have various shapes and sizes. Single-particle analysis using SEM–EDX indicated the probable biogenic (plant parts, pollens, diatoms), geogenic (road dust, resuspended soil dust), and anthropogenic (carbonaceous particle, fly ash) origin of particles.

3.3 Mineralogy

Figure 8 shows the XRD pattern of PM. The broad hump of the spectra appears due to the amorphous nature of the sample and the irregular arrangement of the materials. The distinct small XRD peaks position (2θ values) along with calculated interplanar spacing (dhkl) is provided in Table 3. The mineral peaks of quartz [46], feldspar [47], kaolinite [48], illite [49], augite [50], and calcium aluminum silicate [31] are distinctly observed.

Fig. 8
figure 8

XRD pattern for the aerosol sample collected at Tezpur. Quartz (Q), feldspar (F), kaolinite (K), illite (I), augite (A) and calcium aluminum silicate (CAS)

Full size image
Table 3 Minerals, corresponding XRD peak positions (2θ) and calculated interplanar spacing (dhkl)
Full size table

To further confirm the presence of different minerals in the samples, FTIR analysis was carried out (Fig. 9a and b). For proper visualization of the absorption peaks, the spectrum collected was shown in two separate wave ranges: 400–1999 cm−1 (Fig. 9a) and 2000–4000 cm−1 (Fig. 9b) [20]. The FTIR spectrum indicated the presence of feldspar, quartz, augite, cerussite, calcite, organic carbon along with clay mineral kaolinite, illite, montmorillonite, and imogolite (Table 4). The Brahmaputra River has high abundances of feldspar, kaolinite, chlorite, illite, and quartz-like minerals [56, 57]. Kaolinite, formed from feldspar under acidic conditions, is the dominant species of the Brahmaputra River [56]. The presence of the same types of minerals in aerosol samples might indicate that dried beds of the Brahmaputra River are one of the major sources of atmospheric aerosol particles during the windy events of pre-monsoon season. This is also supported by the presence of diatoms in the samples. The diatoms are a group of microalga present in freshwater and marine environments, and the Brahmaputra River is rich in diatom species [58].

Fig. 9
figure 9

FTIR spectrum of dust sample for the frequency range 400–4000 cm−1. The spectrum is shown in two separate wave ranges, 400–1999 cm−1 in (a) and 2000–4000 cm−1 in (b)

Full size image
Table 4 The FTIR spectra peak observed with corresponding to the literature FTIR spectra peak for different minerals
Full size table

Researchers had applied trajectory analysis to trace the pathway of airmasses into a region [40, 41]. The airmass backward trajectory clusters reaching the site during the dust event are shown in Fig. 10. It is explicit that local wind gusts prevailed over the region during the dust event. Saikia et al. [59] found that during the dry season the sand bar of the Brahmaputra River covered an area of 154 km2 in 2014 within the length of 40 km near the Tezpur area. More so, during the dry season there is no vegetation on the open beds, which makes it easy for the aeolian removal of matter from the surface. A minimum threshold wind speed (U) of 6 ms−1 is sufficient to remove a particle from the dry surfaces with < 5% vegetation cover [60]. The maximum wind speed was 9.1 ms−1 with an average wind speed of 5 ms−1 during the study period. This indicated the lifting of the aerosol particles from dried riverbeds of the Brahmaputra River during the windy period which could be transported to a long distance.

Fig. 10
figure 10

Airmass back trajectory clusters reaching Tezpur (the sampling station) of mid-Brahmaputra Valley

Full size image

4 Conclusion

Morphology study showed that particles originated from geogenic and anthropogenic sources have mainly spherical and irregular shapes and particles from biogenic origin have a distinct spherical structure. Both fine and coarse particles were present in the atmospheric aerosol samples. Mineralogy study revealed dried beds of Brahmaputra River as a source of atmospheric aerosol particles during the windy events of pre-monsoon season. This is also supported by the presence of diatom in the samples. More detailed morphological study and chemical characterization in different seasons and different episodic periods over Brahmaputra Valley would reveal more information on atmospheric particles and associated sources.