This study was carried out from December 2015 to April 2016 in the supply and housing areas of the La Gloria mill, Veracruz. These areas are geographically located between 19º 29’-19º 23’ N, and 96º 21’-96º 31’ W in the central coastal plain region of the state of Veracruz, Mexico, where the average wind speed is 43 km/h. In the studied area there are some rural communities close to La Gloria mill, along with the rural community called La Gloria, which is practically adjacent to it.

The predominant climate is Aw₁, being warm subhumid with a mean annual temperature greater than 22 ºC and a coldest month temperature greater than 18 ºC. The precipitation of the driest month is less than 60 mm; the summer rainfall occurs with precipitation/temperature index between 43.2 and 55.3; the annual precipitation is around 1400 mm according to the Köppen climate classification (García 2004). There are three clearly distinguished seasons throughout the year: drought, rainy and with strong northerly winds.

This activity was carried out during and after the sugarcane harvesting, because in this time the sugarcane mill was operating day and night. This gave rise to the particulate matter (ashes) emission from La Gloria mill. Sometime after this process was completely established, the sample collection started in each selected site.

Two studied areas, zone A and zone B, were selected for the collection of the particulate matter samples, with a total amount, for both zones, of 51 sampling sites, being zone A the closest to the La Gloria mill (Fig. 1). During the operation of La Gloria mill, 50 samples were collected for each site, each one consisting of 20 g/d. After each sampling, all particulate matter samples were taken to the laboratory, until the 50 samples for every site on zones A and B were collected. Subsequently, all particulate matter accumulated during the sampling was carefully mixed up, from which a composed sample of 500 g was obtained and then sieved with 250 μm (No. 60) mesh for the removal of undesirable material. The resulting samples were taken to the laboratory for further analyses. The collection of particulate matter samples in all the 51 sampling sites of zones A and B was done in relatively small cylindrical containers with the following dimensions: 40 cm of base diameter and 15 cm of height. These sampling containers were placed 30 cm above the soil surface to reduce, as much as possible, the entrance of any particles that could contaminate the particulate matter samples gradually entering into the referred containers. From each sieved sample, 50 g were taken and packaged in a previously labeled 16.5 × 14.9 cm Ziploc bag. These bags were stored in closed desiccators to avoid moisture absorption from the environment and remained there until their analysis in the laboratory (Fig. 2). These analyses were performed using the atomic absorption spectrophotometry technique.

Fig. 1

Fig. 2

Each sieved and homogenized 50-g sample was taken and deposited in a 1-L Imhoff cone containing 500 mL of distilled water. Then it was stirred up using a glass rod and allowed to rest for 1 min. The fine particles that were not ashes precipitated to the bottom of the cone, while ashes floated and were decanted immediately on a long neck glass funnel, upon which a filter paper was placed to retain them. Subsequently, the ashes were placed in a porcelain capsule to be dried up in an oven at 100 ºC until a constant weight was reached. The dry ash sample was prepared to carry out its digestion in the CEM trademark microwave oven.

To carry out the laboratory analyses, a Thermo Scientific iCE 3500 AAS System atomic absorption spectrophotometer was used, which is installed and operated in the laboratory of the Technological Institute of Boca del Río, Veracruz. The following analysis technique was applied to the ash samples:

1. Preparation of laboratory material for analysis of the particulate matter (ash) samples. The preparation of the material used in the laboratory was carried out under the specifications established by NOM-117-SSA1-1994 and NOM-242-SSA1-2009 for the analysis of heavy metals, in which the various methods for determining Cd, Pb, Cu, and Zn in food, drinking water, purified water and specific substrates by atomic absorption spectrometry are described.

2. Cleaning of Teflon and glass materials for sample analysis. The cleaning of the Teflon materials for digestion and glassware was carried out with phosphate-free mild (neutral pH) soap at 10 % per liter of distilled water. This soap was used in order to avoid ionic interferences in the spectrophotometer reading. After soaping the material and rinsing it with tap water, it was immersed in a distilled water solution with 20 % nitric acid (HNO₃) (per liter).

The Teflon beakers remained for 1 to 2 h in this solution, while the glassware remained for 24 h. Once the process of removing any metal residue in the HNO₃ solution was completed, the materials were immersed in type II or double-distilled water. The glassware and Teflon beakers remained there for up to 24 h to ensure a complete removal of the acid.

Finally, the materials were removed from the type II water and drained onto the absorbent material. The clean and dry material was suitably placed in duly labeled, airtight (Ziploc) bags or containers with a lid.

3. Digestion of particulate matter samples. First, 0.5 g of a particulate matter sample were weighed. Each sample was then placed in an HP-500 Teflon beaker, to which 10 mL of 70 % (suprapur) reactive grade HNO₃ were added. Previously, a safety membrane was placed inside each valve or blue plug. The valve was placed in the beaker cup and gently adjusted to the bottom. The beakers were then placed in a CEM MARS X microwave oven.

All samples were analyzed with a blank sample containing 0.5 mL of double-distilled water and 10 mL of HNO₃ of the same characteristics, as well as a reference control. Programming of the digestion process was made by the Soil-3051 HP500 method.

The power was adjusted based on the number of beakers used: 1-3 (300 W), 4-6 (600 W) and 6-12 (1200 w) (Table I).

TABLE I

Stage	Power		Ramp	PSI	ºC
	Max	%
1	1200 W	100	2:00 min	300	165
2	1200 W	100	10:00 min	300	175

Note: The power was adjusted based on the number of beakers used: 1-3 (300 W), 4-6 (600 W) and 6-12 (1200 W)

Once the digestion was completed, the Teflon beakers were removed from the oven and placed in the fume hood for toxic vapors, using the required safety measures such as the use of glasses, mask and gloves, in order to facilitate the handling of the samples. The blue valve on the beakers was carefully opened to gradually lower the pressure. Subsequently, the beakers were carefully opened to prevent the change of pressure derived from ejecting the sample outside of them.

The samples obtained were filtered using a Nalgene bottle with a 0.45 μm Millipore filter and a vacuum pump. The filtrate was poured into a 25 mL volumetric flask and brought to volumetric level with type II or double-distilled water. The filled flasks were stirred to homogenize the sample, which was then poured into a previously labeled glass or plastic amber container. Finally, parafilm was placed in the mouth of the amber containers, which were then placed in airtight (Ziploc) bags for storage at about 4 ºC until read on the atomic absorption spectrophotometer.

Prior to the sample reading, a calibration curve had to be prepared for each metal analyzed, for which certified standards of trademark High Purity Standards were used. Thus, the sample readings and standards used for each metal are described as follows:

Cd: a wavelength of 228.8 nm with a 50 % current lamp, a flame of air-C₂H₂, with a lighter height of 4.1 mm and a combustion flow of 1.0 L/min were used.

In addition, the optimization of instrumental specifications such as airflow, acetylene gas, lamp wavelength and lighter height for each metal was done. After this, the equipment became stable and the corresponding standards (with known concentrations and a lower to higher concentration range of analyte) were read for each studied metal. To obtain a calibration curve with a correlation coefficient greater than 0.99, the reading of the digested samples absorbance was made. Hence, the obtained results remained within the concentrations of the calibration curve.

A heavy metals index was constructed through a principal component analysis (PCA) for the concentrations of particulate matter collected at the sampling sites under study. By means of this analysis, the eigenvalues for each component were identified. Subsequently, component one (C₁) was chosen because of its greater explained variance (Table II). According to the PCA, this indicates that the first component (factor) has the highest eigenvalue of all factors, differentiating the highest and lowest value of the components.

TABLE II

Variable	Factor 1	Factor 2	Factor 3	Factor 4
Lead	0.825	-0.174	0.169	-0.511
Cadmium	0.032	-0.910	0.320	0.261
Copper	-0.006	-0.395	-0.912	-0.108
Zinc	0.815	0.209	-0.190	0.506
Exploratory variable	1.345	1.058	0.999	0.597
Total proportion	0.336	0.265	0.249	0.149

The PCA also allowed the acquisition of the factorial loads (index) for each variable, in which the sites with the highest and lowest concentration of heavy metals were detected (Fig. 7).

Fig. 7

In this sense, it is convenient to emphasize that the source of energy to operate La Gloria mill is the combustion of fuel oil combined with bagasse of cane; this results in the emission of large volumes of particulate matter containing heavy metals (i.e., Pb and Zn). In the opinion of the authors, this causes the adherence of Pb and Zn during the process.

In addition, it has been reported that Zn concentrations in the environment are increasing due to anthropogenic activities. A considerable amount of Zn is added as a result of industrial and mining activities as well as carbon combustion (Pacyna 1996, Shallari et al. 1998, He et al. 2005, Arruti et al. 2010). Therefore, when one can see the results derived from the principal component analysis, it is observed that Pb and Zn show a good correlation with factor 1. Therefore, it is reasonable to asseverate that the source of major impact in the Pb and Zn concentration is fuel oil and bagasse combustion, which is enhanced by sugarcane burning and the permanent traffic of lorries transporting the harvested sugarcane and particular vehicles from local inhabitants.

Furthermore, the PCA was a useful statistical tool to group the concentrations of the analyzed heavy metals into the following categories: high, medium and low. It also made possible to build up a map that included isolines for Cd, Cu, Pb and Zn.

Table III shows that factor one has a high correlation with Pb and Zn, which makes them the most representative metals.

TABLE III

Factor	Eigenvalue	Total variance (%)	Total cumulative value	Cumulative (%)
1	1.345	33.627	1.345	33.627
2	1.058	26.462	2.403	60.089
3	0.999	24.979	3.402	85.069
4	0.597	14.930	4.000	100.000

Particulate matter is emitted through the industrial chimney stack. The higher the chimney is, the greater the probability that pollutants will be dispersed and diluted before affecting neighboring populations. The chimney’s visible emission is called plume, and its height is determined by the speed and thrust of the gases coming out of the chimney (Lou et al. 2009). According to the distance at which the highest concentrations of particulate matter occur these are not dependent on wind speed (Solís y López 2003). The highest concentration is maintained at a distance of 500 m from the emitting source. Usually wind favors the dispersion of pollutants and humidity plays a negative role in the dispersion, due to the concentration of smoke and particulate matter.

The distribution of particulate matter in the surrounding area of La Gloria sugar mill and the sugarcane burning area presents a high level of concentration in places closest to the emission source, with distant sites presenting medium and low concentrations. This is attributable to the dispersion, i.e., the transport of particulate matter in the air, which depends on the prevailing weather conditions such as wind speed and direction that were observed while collecting the particulate matter samples. This information was not available for the studied zone and it was not possible to retrieve it from institutions like the Water National Commission (CONAGUA, Spanish acronym), hence it was not included in the article. The different types of volatile particle dispersion phenomena caused accumulation in areas close to the source of emission and their transport to closer areas. Figure 8 shows the location of the sampling sites; the high, medium and low concentration levels are indicated by using the following geometric figures: circles, squares, and triangles.

Fig. 8

It is necessary to clarify that the scope of this study was constrained to the determination of concentrations and spatial distribution of the heavy metals Cd, Cu, Pb and Zn present in particulate matter emitted by the La Gloria sugar mill and sugarcane burning in regional agroecosystems, as a function of the distance at which such material was deposited during the emission process. The reason to stress this issue is the lack of financial support, which hindered the extension of this research work to other aspects.

By means of an analysis of variance of the high, medium and low concentration types, it was found that there were significant statistical differences (p = 0.0001) between types (Fig. 9).

Fig. 9