Identification of cement in atmospheric particulate matter using the hybrid method of laser diffraction analysis and Raman spectroscopy.

The production of cement is associated with the emissions of dust and particulate matter, nitrogen oxides (NOx), sulfur dioxide (SO2), carbon monoxide (CO), heavy metals, and volatile organic compounds into the environment. People living near cement production facilities are potentially exposed to these pollutants, including carcinogens, although at lower doses than the factory workers. In this study we focused on the distribution of fine particulate matter, the composition, size patterns, and spatial distribution of the emissions from Spassk cement plant in Primorsky Krai, Russian Federation. The particulate matter was studied in wash-out from vegetation (conifer needles) using a hybrid method of laser diffraction analysis and Raman spectroscopy. The results showed that fine particulate matter (PM10 fraction) extended to the entire town and its neighbourhood. The percentage of PM10 in different areas of the town and over the course of two seasons ranged from 34.8% to 65% relative to other size fractions of particulate matter. It was further shown that up to 80% of the atmospheric PM content at some sampling points was composed of cement-containing particles. This links the cement production in Spassk-Dalny with overall morbidity of the town population and pollution of the environment.

Introduction

Many research studies show that cement plants are major sources of particulates polluting the environment (Mehraj et al., 2013; Mishra and Siddiqui, 2014; Fell and Nordby, 2017). The production of cement, cement materials, and asbestos-cement products is associated with the emissions of dust and particulate matter, nitrogen oxides (NOx), sulfur dioxide (SO2), carbon monoxide (CO), heavy metals, and volatile organic compounds into the environment. The aerosol pollution also includes industrial dumping of overburden rock formed during the extraction of raw materials. Pollutants are also formed and emitted during storage, crushing, and transporting of raw materials as well as during packing and loading of the cement.

Harmful effects of cement dust and related substances include irritation and diseases of the respiratory tract (asthma, chronic bronchitis, and silicosis), the skin (e.g. allergic dermatitis), nasopharyngeal mucosa and the oral cavity, and digestive organs. In the medical literature, a specific cement-caused lung disease (a form of pneumoconiosis) has been described (Elsewefy and Metwalli, 1970; Bazas, 1980; Aminian et al., 2014; Moghadam et al., 2017), which is considered to be the most significant occupational hazard for cement industry workers. In addition, epidemiological studies have reported a high risk of developing cancer in cement workers, since the raw materials for cement production can contain free silica, hexavalent chromium, cadmium, and other known carcinogens (Bazas, 1980; Donato et al., 2016).

People living near cement production facilities are potentially exposed to these pollutants, including carcinogens, although at lower doses than the factory workers. Epidemiological studies have reported that such populations suffer from higher levels of respiratory and skin diseases and eye irritation than control populations not exposed to such pollutants (Bertoldi et al., 2012; Oguntoke et al., 2012; Nkhama et al., 2017).

PM10 particles are a complex heterogeneous mixture of solids and liquid droplets floating in the air in the size range of 2.5–10 μm in diameter and, depending on the composition, they may pose a threat to human health (WHO Regional Office for Europe, 2016). In developing countries, the concentrations of PM10 and PM2.5 in the vicinity of cement plants are often above the exposure limit, which leads to the increase in the number and severity of above mentioned diseases (Oguntoke et al., 2012; Nkhama et al., 2017). According to literature, an increased incidence of illness in children living near cement plants is noted even in developed countries, where the content of PM does not exceed the exposure limit (Marcon et al., 2014). There is evidence of an increased level of cancer and an increase in mortality among residents near cement plants (Giordano et al., 2012; Eom et al., 2017).

In this study we focus on the distribution of PM10 particles, the composition, size patterns, and spatial distribution of the emissions from Spassk cement plant in Primorsky Krai, Russia. This work is a continuation of a series of works by the authors on the monitoring of pollution particles in the environment arising from anthropogenic pressure, including emissions of buses (Chernyshev et al., 2019) and motorcycles (Chernyshev et al., 2018), as well as particles from welding (Kirichenko et al., 2018).

Methods

Spassk-Dalny is a town in the Primorsky Krai, Russia, located in an open area in the Kuleshovka and Spasovka river basin, 243 km to the north of Vladivostok. The climate is moderately monsoon; the prevailing wind direction is southwest and the average annual wind speed is 2.5–3 mps. The population in 2017 was more than 41,000. The main enterprise of Spassk-Dalny is a cement plant. The first plant was built in 1907 to provide cement for the construction of the Trans-Siberian Railway (which runs through Spassk-Dalny) and the Vladivostok Fortress. In 1976, the construction of a new Spassk cement plant (NSCP) was completed on the southern outskirts of the town. Raw materials are mined in the limestone quarry of the Dlinnogorsk deposit, located 6 km south of the Spassk-Dalny town center and 2 km east of the new Spassk cement plant. Raw material from the quarry is transported by conveyor belt to the mill at the factory. The aggregate capacity of the JSC Spassk cement enterprise is over 3 million tons of cement per year. At present, most of the production is at the new Spassk cement plant.

To study the atmospheric pollution in the area we collected needles (leaves) from conifers and used the ultrasound-enhanced washing method for sample preparation following the method described by Kodintsev et al. (2017).

The needle samples were collected in two series or seasons of the year. The first series of samples were collected during the summer of 2017 at 10 sampling points. Sampling points were located in Spassk-Dalny, a neighboring Spasskoye settlement and near the New-Spassk cement plant and the original Spassk cement plant in dry, sunny weather at a temperature of +30 to +33 °C during a weak south wind. The second series of samples were collected in the winter season of 2017–2018. A control group of samples was collected in local gardener's community, remote from town center. The schematic map with the sampling points is shown in Figure 1.

Figure 1

Conifer needle sampling points in Spassk-Dalny. Point No. 1 is next to the entrance to the New-Spassk cement plant; point No. 4 is near the gate to the Spassk cement plant; point No. 6 is in the town center of Spassk-Dalny.

Needles were collected from two coniferous tree species i.e. Pinus sylvestris and Abies nephrolepis at a height of approximately 1–1.5 m. The needles were placed in sterile plastic containers prewashed with double-distilled water and transported to the laboratory for analysis.

In the laboratory, the containers with sample needles were filled with double-distilled water and treated with ultrasound using a Sonopulse 3100 HD ultrasonic homogenizer (Bandelin electronic GmbH & Co. KG, Germany) at 22 kHz, 100 W and a 5 min exposure to remove the pollution particles from the needles.

Following the ultrasonic treatment, a 40 ml portion of the resulting washout liquid was taken from each sample and analyzed for particulate matter size distribution using the Analysette 22 NanoTec plus laser particle sizer (Fritsch GmbH, Germany). The measurements were run at the settings of quartz/water at 20о C. The detection range was 0.008–2000 μm. Laser diffraction analysis of a control sample (Portland cement produced by JSC Spasskcement) was performed as well.

The remaining liquid in the containers was filtered through a 0.22 μm filter and dried, and then the residue was further analyzed. Dried samples, the surface of control needles (not ultrasonic treated) and the reference Portland cement sample were analyzed with the Morphologi G3-ID equipment (Malvern Instruments Ltd., UK), which combines the dispergation system and automated static imaging features with chemical identification of individual particles using Morphologically-Directed Raman spectroscopy. Using the Malvern's patented chemical correlations analysis software, the proportion of cement in the particulates from all samples was determined. Raman spectra of PM samples were also automatically compared with the library of spectra of minerals widespread in the area: calcite, datolite, danburite, calcium-iron garnet, hedenbergite, and other minerals of common sedimentary rocks.

The surface of conifer needle samples was also analyzed for the content of heavy metals. A non-destructive method of analyzing the test material was required for these samples, so the analysis was performed by X-ray fluorescence (XRF) using the Innov-X SyStemS portable spectrometer (USA). The XRF method is based on the collection and subsequent analysis of the spectrum arising from the irradiation of the material with X-ray radiation.

Particle morphology was studied on electron probe micro analyzer JXA 8100 (JEOL, Japan) with an Oxford Instruments INKA-sight energy dispersive analytical spectrometer (UK). The particles were studied in the dried wash-out from the conifer needles.

Results

According to our results, the percentage of PM10 particles was high in all areas of the town and in the surrounding neighborhood. Particulate matter with a diameter under 10 μm can penetrate the human respiratory system, even the deep lungs (WHO Regional Office for Europe, 2016). It is noteworthy that high concentrations of PM10 particles were found at a considerable distance from the proposed main source of the atmospheric pollution – NSCP. For example, 58.8% of the PM10 particles were 10.2 km away from the NSCP (point No. 7), and 49.6% at 12 km away from the NSCP (point No. 8).

The particle size distribution and physical parameters of the atmospheric particulates in the wash-out from the conifers needles in the first, summer series of samples taken in Spassk-Dalny town are presented in Table 1.

Table 1

Particle size distribution and physical parameters of particulate matter in the washout from the conifer needles in the summer series of samples.

Fraction, μm	Sampling points
	1	2	3	4	5	6	7	8	9	10
	Percent of particles
Under 1	7.9	5	3.4	6.5	3.5	5.2	6.5	5.7	4.6	7.7
1–10	50.6	35.5	25.2	53.9	34.5	44.9	58	49.7	41	59.5
10–50	27.7	27.5	12.4	18.4	45.6	45.5	34.1	38.2	38.5	32.2
50–100	0.6	0.1	1	0	5.2	1.3	0.5	0.8	7.7	0.6
100–400	12	18.2	16.8	0.6	1.4	1.2	0.9	0.1	2.7	0
400–700	1.2	12.3	31.8	3	0.2	0	0	0.1	0	0
Over 700	0	1.4	9.4	17.6	9.6	1.9	0	5.4	5.5	0
Mean diameter, μm	16.4	57	185.2	155.7	226.8	112.7	9.6	180.3	20.5	9.2
Mode, μm	15.7	14.7	542.1	10.6	972.7	15.7	7.4	1004.8	14.2	5.2

Ultrasound treatment of conifer needles was an effective method to obtain samples with increased content of particulate matter for analysis. Photomicrographs of the surface of needle samples from sampling points No. 2 (distance to NSCP is 0.8 km) and No. 8 (distance to NSCP is 12 km) before and after ultrasound treatment, obtained with an Axio Imager Z2 microscope (Zeiss AG, Germany) are presented in Figure 2.

Figure 2

Photomicrographs of the surface of needles. Magnification x100. a) Pinus sylvestris needle from sampling point No.2 before ultrasound treatment; b) Pinus sylvestris needle from sampling point No.2 after ultrasound treatment.

Figure 3 shows particle scattering diagrams by diameter in needle samples collected during the winter season 2017–2018 at sampling points No.1, 6, and 9. As shown in the figure, a particle size of approximately 2 μm (PM2.5) prevailed in samples No. 1 and No. 9; while particles smaller than 1 μm followed by 5 μm prevailed in the sample taken in the center of Spassk-Dalny. In total, 100,000 particles in each sample were analyzed using the Morphologi G3-ID (Malvern Instruments Ltd., UK), which combines the automated static imaging features with Raman spectroscopy.

Figure 3

Particle scattering diagrams by diameter in needle samples collected during the second stage of the study (winter season 2017–2018). a) sampling point No. 1 (entrance to the NSCP); b) sampling point No. 6 (town center); c) sampling point No. 9 (Local forestry).

The identification of cement particles in the samples was carried out by comparing the Raman spectra of cement, the particles in the dried residue, and aggregates of the particles on the needle samples themselves using the Morphologi G3-ID equipment. The use of micro-Raman spectrometry makes it possible to identify various mineral phases in ordinary Portland cement without special sample preparation (Potgieter-Vermaak et al., 2006). This method is unique as it allows determination of the size and the basic chemical composition of each particle in the sample (Kholodov and Golokhvast, 2016).

The main mineral components of cement (alite - C3S, belite - C2S, tricalcium aluminate - C3A) have been previously studied using Raman spectroscopy and characteristic Raman spectra are found in the literature (Liu and Sun, 2016; Girotto et al., 2020). These mineral components are the products of high-temperature processing of mineral raw materials. They are labile compounds reacting with water to form other substances (Bullard et al., 2011), and therefore they can't be related to the natural mineral background and are of anthropogenic origin. Other mineral particles (calcite, gypsum etc.) possibly related to cement production were found as well, but we could not reliably distinguish their origin.

Figure 4 shows a comparison of spectra of particulates in the washed and dried residue from the needle sample taken at the sampling point No. 7 with the spectrum of Portland cement manufactured by JSC Spasskcement.

Figure 4

Comparison of spectra of particulates in the dried residue of the washout from the needle sample (sampling point No. 7) with Portland cement spectra. a) Sample spectra; b) Portland cement spectra. The correlation of spectra in this sample is from 86.3% to 91.5%.

The labeled peaks on the red and blue lines in Figure 4 correspond to alite with a high content of amorphous carbon (1270, 1495 cm−1) and quartz (426 cm−1). Alite is a cement-specific component that correlated well with the spectra of particles from the needle samples reported in this study. The origin of quartz in the samples is questionable, as it may be both of the natural mineral background and of cement origin.

The software chemical correlation based on Raman spectra showed the presence of cement in all the needle samples collected in summer 2017 in Spassk-Dalny in quantities up to 90%, both in the ultrasound-treated and dried washed samples, and directly on the surface of the needles. The spectra of 108 of 120 particles from the needle surface samples from sampling point No. 2 (distance to the New-Spasskcement plant entrance is 0.8 km) showed over an 80% chemical correlation with the Portland cement spectrum. The spectra of 96 of 120 particles from sampling point No. 8 (distance to the NSCP is 12 km) showed over a 60% chemical correlation with the Portland cement spectrum, which was 80% of all particles in the sample.

The results of the software chemical correlation of the particle spectra in samples collected during the second stage of the study (winter 2017–2018) with the spectrum of the Portland cement are presented in Table 2. As shown in Table 2, 64% and 82 % of all analyzed PM10 particles from sampling points 1 and 9, respectively, were similar in chemical composition to cement particles. It should be noted that in the needle sample No. 6 collected from the town center during the winter season, particles with a spectrum similar to the spectrum of Portland cement were found only in the larger particle size fraction and only in the amount of 24%.

Table 2

Software chemical correlation based on Raman spectra of needle washed samples (winter 2017–218 samples) with a sample of Portland cement.

Sample No. and sampling point	The percentage of particles in the sample with 50% or higher correlation with the cement spectrum
	PM10 particles	Particles diameter = 15–25 μm
Pont No. 1 (Entrance to NSCP)	64%	82%
Pont No. 2 (Distance to NSCP = 0.8 km)	90%	Not measured
Pont No. 6 (Town center)	Not detected	24%
Pont No. 8 (Distance to NSCP = 12 km)	80%	Not measured
Pont No. 9 (Local forestry)	82%	93%

The results of the analysis of the conifer needles for the content of heavy metals are shown in Table 3. The increased content of titanium, copper, and zinc observed in the samples can be explained by the anthropogenic impact on the town's environment.

Table 3

Analysis of samples collected in the winter season 2017–2018 for the content of metals. Bold sigifies the increased content of titanium, copper and zinc.

Element	Pont No. 1		Pont No. 6		Pont No. 9		Control sample
	Mg/kg	%	Mg/kg	%	Mg/kg	%	Mg/kg	%
K	15 669.6	32.9	24 818.9	30.2	29 796	30.9	14989.9	48.8
Ti	2 471.6	5.2	5 535	6.7	6 837.5	7.1	934.4	3
Cr	65.1	0.1	97.7	0.1	100	0.1	16.5	0.1
Mn	614.3	1.3	803.3	0.9	944.2	0.9	511.8	1.6
Fe	27 464.8	57.7	48 355.9	58.8	54 973.4	57	13560.4	44.1
Cu	72.1	0.1	333.3	0.4	793.7	0.8	51.0	0.1
Zn	326.4	0.7	801.4	0.9	987.5	1	118.4	0.3
Pb	93.9	0.2	202.8	0.2	287.4	0.3	74.3	0.2
Sr	418.5	0.8	570.5	0.7	835	0.9	102.9	0.3
Mo	3.3	0	8.1	0	10.6	0	1.4	0
Cd	4.5	0	6.3	0	10.8	0	2.6	0
Sn	19.6	0	20.4	0	26.9	0	6.2	0
Sb	15.1	0	41.8	0.1	65.9	0.1	10.5	0
Ba	239.8	0.5	451.9	0.5	396.3	0.4	224.6	0.7
Co	13.2	0	28.3	0	68.1	0.1	12.6	0
Ni	57.5	0.1	92.1	0.1	161.1	0.2	69.6	0.2
As	17.7	0	43.9	0.1	108.3	0.1	11.2	0

Additionally, according to particle morphology analysis, one of the most common types of particles in the samples was torn metal particles. Among these, particles of iron oxide, steel, and pyrite were predominant. Small impurities contained particles with the presence of Pb, Ni, Mg, Al, Sr, Zn, Ba, and Mn. Another common type of particle was spherical type, which was mostly composed of calcium oxide (CaO), silicon dioxide (SiO2), aluminum oxides (Al2O3), and iron (Fe2O3).

Discussion

The cement manufacture process at the new-Spassk plant is a “dry” process, which is more energy-efficient than the “wet” process. Among the both processes, the former method allows a dramatic increase in product output. It is known that the filtration equipment was installed to reduce the dust emission into the atmosphere. However, environmental studies conducted in the study area showed that emissions from the cement plant affected all areas of the town.

According to our results, high concentrations of PM10 particles were observed in all areas of the town. At all sampling points, the percentage of PM10 ranged from 34.8% to 65% relative to other size fractions of particulate matter. The current data corresponded to the results of previous studies, indicating a high level of air pollution with particulate matter in Spassk-Dalny (Kolomiytseva and Khristofrova, 2012; Kholodov et al., 2018).

Earlier we studied the particle size distribution in the atmospheric precipitation of snow water in Spassk-Dalny using laser diffraction analysis. It was shown that at 8 of the 12 sampling points the content of PM10 particles in the atmosphere was high, ranging from 16.8% to 31.5% (Kholodov et al., 2018). Some deviation in the percentage of the PM10 fraction in atmospheric air between our previous study of particle size distribution in snow water samples and the new data can be explained by the fact that the coniferous trees contributed to greater retention of microparticles and their constant concentration for a longer time. These needle sample analyses can be considered more representative of the atmospheric air than the analysis of the snow water samples (Kodintsev et al., 2017). In addition, the small increase of the PM10 percentage in these latter samples can be explained by a more efficient ultrasonic-enhanced sample preparation. We found no correlation between the season of conifer needle sampling and particle size distribution or other properties of the analyzed PM.

According to Raman spectra analysis, up to 80% of the atmospheric PM content at some sampling points was composed of cement particles. The chemical correlation based on the Raman spectra showed the presence of cement particles in all of the needle samples. Despite the use of filtration equipment to reduce the dust emission of the cement plant into the atmosphere, cement particles were found in the atmospheric particulate matter far beyond the sanitary protection zone of the cement plant, up to 12 km from it. The correlation of Raman spectra of the atmospheric particulates at this distance from the plant with the spectrum of the Portland cement produced by JSC Spassk cement was very high for all samples and all particles. However, dust emissions from the limestone quarry of the Dlinnogorskoye deposit and other stationary and mobile pollution sources should also be taken into consideration in evaluating the overall pollution sources affecting this region.

According to a previous research (Kolomiytseva and Khristofrova, 2012), the concentrations of poly-element compounds of heavy metals – copper, manganese, chromium, and lead in the air of Spassk-Dalny exceeded the maximum permissible concentrations (exposure limit) by a factor of 1.5–4 with a maximum distance of 1–2 km from the cement plants. A similar pattern of distribution of these heavy metals was observed at all sampling points in the current study in Spassk-Dalny. The effects of metals in particulate matter on health are associated with reduced lung function, respiratory diseases, increased mortality rates, myocardial infarction incidence, and other deceases (Geiger and Cooper, 2010; Fortoul et al., 2015).

The morphological analysis of particles in the samples showed that one of the dominant forms of particles were spheres. Some of the spherical particles are products of fuel combustion, possibly fly ash or other combustion products (Gieré et al., 2003). It is highly likely that the source of at least a portion of such spherules is the cement factory, where the raw materials are heat treated (Sánchez-Soberón et al., 2015; Rovira et al., 2018). Additionally, the presence in the composition of a number of particles of silicates, aluminoferrite, and magnesium oxide suggests that some of these spheres may be cement products, as these are typical components of ordinary cement.

A strong relation was reported between the emission of particulate matter and the general morbidity of the population in the area (Kiku et al., 2017). The contribution of air pollution to the frequency and severity of the most common diseases of the respiratory system, digestion, skin, allergic reactions, etc. is reported to be up to 30% of the total amount of factors affecting health of the population (Kiku et al., 2017). The cement plant is one of the major pollutants in Spassk area, and its operation is linked to overall morbidity of the town population.

Conclusions

The study of atmospheric particulate matter using a combined method of laser diffraction analysis, Raman spectroscopy, and scanning electron microscopy in a township affected by a local cement plant (Spassk-Dalny, Primorsky Krai, Russian Federation) showed that cement-containing PM were present in all samples taken from the town. Despite the use of filtration equipment at the cement plant, cement-containing particles were found beyond the sanitary-protection zone of the enterprise.

Cement-containing particles in the atmosphere are known to cause irritation and diseases of the respiratory tract (asthma, chronic bronchitis, silicosis), skin (for example, allergic dermatitis), mucous membranes of the nasopharynx and oral cavity, digestive organs, etc. The cement production in Spassk-Dalny, among other sources of pollution, is linked to overall morbidity of the town population and pollution of the environment.

According to the results of our study, cement production once again proves to be a manufacture hazardous for human health. We have shown that more than half of the particles of the PM10 fraction have anthropogenic origin associated with the production of cement. These results may improve our understanding of risks that cement production poses to human health, and distances that pollution from cement manufacture may cover, further strengthening the necessity for cement pollution control and mitigation.

Declarations

Author contribution statement

Aleksei Kholodov: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Alexander Zakharenko, Alexander Karabtsov: Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Vladimir Drozd, Valery Chernyshev, Konstantin Kirichenko, Ivan Seryodkin, Igor Vakhnyuk: Performed the experiments.

Svetlana Olesik, Ekaterina Khvost: Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Vladimir Chaika: Contributed reagents, materials, analysis tools or data.

Antonios Stratidakis: Contributed reagents, materials, analysis tools or data; Wrote the paper.

Marco Vinceti: Analyzed and interpreted the data; Wrote the paper.

Dimosthenis A. Sarigiannis, A. Wallace Hayes, Aristidis Tsatsakis: Analyzed and interpreted the data.

Kirill Golokhvast: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.