The ash intended for nickel decontamination comes from an industrial sludge incinerator (Figure 1A). The ash was analyzed on the basis of a requirement for landfill waste recovery (Table 1). The sample was taken as a so-called mixed sample, i.e. from ten partial point samples from a temporary storage site (a landfill), namely so that all properties of the waste are representative. The pre-treatment consisted of quartation, homogenization and subsequent fractionation through a fractional sieve (mesh size 0.63 mm).

The basis of the method is the measuring of the atomic emission by optical spectroscopy. After mineralization (decomposition of solid samples into an aqueous matrix), the samples are sucked using a peristaltic pump through a nebulizer into a misting chamber, where they are nebulized. The aerosol created is then carried into a plasma, where excitation of the atoms occurs. Characteristic atomic line spectra are then generated by high frequency inductively coupled plasma (ICP). The radiation emitted is decomposed using a spectrometer grid and the intensities of the lines are monitored by detectors.

The determination of metals for pre-treatment of the ash took place by extraction under a reflux condenser. The original homogenized sample was weighed in the amount of 0.3 g into a reaction vessel and 21 ml of HCl and 7 ml of HNO₃ were added.

The reaction vessels were left open until the initial reaction (foaming) took place. The mixture was covered with a watch glass and left to stand at room temperature for 16 h. The vessels were subsequently placed on a heater and the reflux condenser was installed. The temperature of the reaction mixture was slowly raised until reflux conditions were reached. This state was maintained for 2 h to ensure that the cooling zone is lower than one-third the height of the reflux condenser. The reaction mixture was then allowed to cool. The condenser was rinsed with 10 ml of 1 mol L^–1 nitric acid in the reaction vessel. The mineralizate was then poured quantitatively through a filter paper into a 100 ml measuring cup, filled up to the mark and then poured into a 100 ml plastic storage container.

The genus Chlorella sp. as a representative of algae ranks among the most numerous groups of green algae. In all the experiments an algal biomass of Chlorella sp. obtained from Institute of Microbiology of the Czech Academy of Sciences (Opatovický mlýn, Tr̆ebon̆, Czechia), was used (Figures 1B–D). The dry biomass of algae was prepared by washing the biomass in deionized water and drying in an oven at 70°C for 24 h.

A cosmopolitan strain of A. niger (Figure 2A) was isolated from Dystric Cambisol (contaminated and eroded) without vegetation. The chemical characteristics of the substrate showed an ultra-acidic soil reaction (pH 3.12), a very low amount of organic matter (%C_ox 0.49) and exceeded the value of aluminum (Al 727 mg kg^–1) (Šimonovičová et al., 2021).

Aspergillus niger fungal pellets were prepared in a 45 ml of SDB (Sabouraud Dextrose Broth Liquid Medium, Himedia, Mumbai, India) enriched with a 5 ml suspension of conidia from a pure culture of A. niger. Cultivation took place over 5 days at 25°C in 250 mL Erlenmeyer flasks, with stirring at 170 rpm (Unimax 2010 shaker, Heidolph, Germany) under laboratory condition at 25°C (Figure 1F). The A. niger fungal pellets formed very quickly, and after 5 days they were removed by filtration, washed with a large amount of distilled water and used in the experiments (Figure 2B).

The figures of the dry green algae biomass (Figure 1C), the filtered A. niger pellets (Figures 2C–E, 3B–E), were made with a Canon IXUS 16.1 megapixel camera (Japan).

The figures of the Chlorella sp. surface structure (Figure 1D) were observed under a JEOL JXA 840A Scan Electron Microscopy (SEM) (Japan).

The figures of the A. niger strain (Figure 2A), the A. niger pellets (Figures 2C–E), and the A. niger with Chlorella sp. pellets (Figures 3B–E) were observed under an Axio Scope A 1 Carl Zeiss Jena light microscope in a drop of lactic acid enriched with a cotton blue stain (0.01%).

Microscopic measurements [Microscope: AXIO–IMAGER A, CARL ZEISS (Germany), Filter set 02–excitation G 365, Beam splitter FT 396, Emission LP 420], chlorophyll in a metabolically inactive biomass (Figure 1E).

In the case of each sample, approximately 1 g of bulk material was pulverized into a fine powder in an agate mortar. After homogenization, the samples were dried at 60°C for 24 h. Before each measurement 2 mg (± <0.05 mg) of homogenized sample was thoroughly mixed with 200 mg of KBr and pressed into a small pellet. FTIR analysis was performed in transmission mode and the spectral data were expressed as absorbance values. Each spectrum was recorded in the mid-infrared region (from 4000 to 400 cm^–1) by averaging 128 scans, with a spectral resolution of 4 cm^–1. The analysis was performed using a NICOLET 6700 FTIR spectrometer and the OMNIC 8 software (Thermo Scientific). The FTIR spectra obtained were not corrected by any procedures, filters, or algorithms, such as smoothing, absorbance normalization, baseline corrections, etc.

Mass concentrations of C, H, N, and S were determined by combustion of the sample (app. 10 mg) in the atmosphere with elevated oxygen content and subsequent detection of liberated gases by the gas chromatography principle. The O content was measured similarly, but the sample was heated in an atmosphere of pure He, without the addition of a catalyst. The measurement was carried out by the elemental analyzer FLASH 2000 equipped with the Eager Xperience software (Thermo Scientific).

All the experiments were organized according to Table 2. In all samples 3 g of ash were used. Dry biomass of Chlorella sp. in an amount of 1 g was used only in samples 5–15. A. niger fungal pellets were also used in all samples. Distilled water in a volume of 30 ml was added to samples 1, 3, 5, 6, 9, 10, and a volume of 50 ml was added only to sample 15. Samples 1, 3, 5, 6, 13, and 15 were performed on a shaker, while all other samples were run under static conditions.

We chose the variability of individual samples within the experiments based on the different decontamination conditions both under natural conditions (exterior) and under laboratory conditions at a temperature of 25°C. The samples for analysis were taken at a defined time interval of 48 h. At the end of the experiments in the case with the addition of distilled water, the suspension was separated from the solution by membrane filtration (Millipore, pore size 0.45 μm). The metal analysis of all samples was carried out on an emission spectrometer with inductance coupled plasma (ICP Profile Plus, Teledyne Leeman Labs, United States). All experiments were performed in duplicate.

The input values of the ash from the sludge incinerator are documented in Table 1. The removal of Ni from the material ranged as follows: 13.61% (sample 10) < 17.25% (sample 3) < 18.58% (sample 9) < 19.68% (sample 4) < 22.49% (sample 12) < 28.37% (sample 11) < 30.95% (sample 8) < 34.56% (sample 7) < 37.90% (sample 6) < 44.01% (sample 15) < 46.28% (sample 5), as shown in Figure 4.

From the various experimental conditions, the combination of A. niger pellets and dry algal Chlorella sp. biomass, with the ash substrate moistened with 30 ml to 50 ml of distilled water, proved to be the most effective. The experiment ran on a shaker under laboratory conditions. During shaking, in addition to the even distribution of pellets with the ash substrate, the coagulation of pellets with the algal biomass also occurred (Figures 3A–E). The coagulates formed not only absorbed Ni on the surface but also bioleached in the presence of A. niger (samples 5, 6, and 15).

Under laboratory conditions, when only A. niger pellets were applied to the ash, we recorded a slightly higher efficiency under static conditions and without the addition of distilled water (sample 4) compared with the addition of 30 ml of distilled water on a shaker (sample 3). In the exteriors, experiments were conducted under static conditions, and in this case, too, we recorded lower values of Ni content reduction in the aqueous medium (samples 9 and 10) compared with the anhydrous environment (samples 11 and 12). The species A. niger is a known producer of many organic acids and secondary metabolites (Šimonovičová et al., 2020, 2021) which are used in bioleaching processes. It can be assumed that in the case of adding distilled water, the organic acids produced were diluted, which caused a slowing down/reduction of the bioleaching process from the solid substrate.

The absorption of Ni from the ash substrate occurs by means of several processes. When A. niger pellets are used alone, the process of bioleaching takes place. When a consortium of the microorganisms Chlorella sp. and A. niger were used, aside from the bioleaching, the efficiency is increased by involving the cell wall of green algae biomass. The cell wall of algae is made up of cellulose microfibrils, pectins, hemicelluloses, and proteins. These polymers form a complex network of pores and channels that influence the movement of substances in this space. But the surface negative charge of polymers can limit the movement of minerals as well as heavy metals by binding them. However, in general, the cell wall is permeable. The cytoplasmic membrane, as well as the vacuole membrane, represent a highly selective semipermeable barrier between the internal and external cell environment, and its role is to control the transfer of substances into and out of the cell to the surrounding space. It is only permeable for certain substances, for example, the molecules of some gases (O₂, CO₂, N₂) and for water, but less so for glycerol or ethanol. For substances that cannot pass freely into the cell through the cytoplasmic membrane, transmission is provided via pumps, transporters, and channels (which are membrane proteins), which are more or less specific for transport of the given substance.

For us, the so-called ABC carriers, which is a general term used for a large and diverse group of carriers, are of interest. ABC transporters on a tonoplast provide transport, for example, of heavy metals or herbicides into the vacuole as well as their sequestration. In an environment where different ions are present, competition often occurs between, for example, Ca²⁺ and other cations at the interaction site. Thus, instead of a “good cation,” an unwanted ion, e.g., Cd or ion of another heavy metal, enters the cell. However, with damaged cells and a disrupted cell membrane system a loss of membrane semipermeability and uncontrolled movement of substances occurs. Microorganisms have the ability to accumulate many heavy metals and toxic elements (Pb, Ag, Pt, Pd, Au, Hg, Ga, Cd, Cu, Ni) from the external environment (Bulgariu and Bulgariu, 2020). Living and dead cells, the products of cell metabolism, extracellular polysaccharides and cell wall components all have the ability to receive and accumulate metals (Timková et al., 2018). The primary carrier surface of biosorption, however, is the cell wall of the biosorbent. The composition of the cell wall consists particularly of polymeric substances that are rich sources of various functional groups, by which the contaminants bind. The following groups can be included here: carboxyl (–COOH), hydroxyl (–OH), sulfhydryl (–SH), phosphate (–PO₄^3–), amino (–NH₄⁺), and others (Gupta and Diwan, 2017).

The metabolically inactive, i.e., the dead culture of an algal biomass, can isolate metal ions and metal complexes from solution due to its unique chemical composition. When selecting metal-sorbent biomaterials, it is necessary to start from the factor of origin and thus resolve the issues of availability, quantity and potential demand (de Rome and Gadd, 1991; Tiemann et al., 1999). The advantage of applying dead biomass compared to a living one, i.e., an active biomass, is that the living biomass cells require the addition of a fermentation medium (which are costly), and the simultaneous addition of medium increases biological oxygen demand (BOD) and chemical oxygen demand (COD). Furthermore, dead biomass is not affected by the toxicity of metal ions and can be subjected to various chemical and physical pre-treatments to increase sorption. As a result, adsorbed metals can be easily obtained from the biomass by chemical or physical methods, which leads to repeated use of the biomass, thus reducing the costs of the process. Ash classified as hazardous waste is disposed of by several methods, such as heat treatment (e.g., sintering, melting and vitrification) (Park et al., 2005), the setting of cement (Bie et al., 2016), chemical stabilization (Nzihou and Sharrock, 2002; Wang F. H. et al., 2015), or extraction (Tang and Steenari, 2015; Wang W. et al., 2018). Other alternative methods for reducing hazardous substances in ash, which run with a lower energy load or better performance, include hydrothermal treatment (Jin et al., 2013; Qiu et al., 2017, 2018) and mechanical-chemical treatment (Chen et al., 2019).

There were several differences between the spectra of Chlorella sp. and A. niger, respectively, which are related to the different chemical composition of the two microorganisms. According to Johnson (1965), the main components building the mycelium of Aspergillus are carbohydrates (73–83%), with smaller portions of lipids (2–7%) and proteins (0.5–2.5%), while in the case of Chlorella sp., approximately 50% of the biomass is made up of proteins (Wild et al., 2019). Apart from proteins, Chlorella sp. Also contains a higher amount of lipids (∼18%) and a significantly lower concentration of carbohydrates (∼23%) in comparison to A. niger (Prabakaran et al., 2018). The results of FTIR spectroscopy were in general accordance with these differences. IR absorption in the Chlorella sp. spectrum may be associated mainly with amides and aliphatic structures, whereas the spectrum of Aspergillus showed more peaks characteristic for carbohydrates (polysaccharides). It should be noted that amide groups are not only a part of proteins but are also present in certain polysaccharides, such as chitin. In some cases, the amides bound in different structural units can be distinguished, as the vibrations of the respective functional groups manifest at different frequencies. The nitrogen content detected in the biomass was lower in the case of A. niger in comparison to Chlorella sp., and the same accounts for the N/C ratio (Table 3). This is associated with total amount of proteins in the biomass, which was probably lower in the case of Aspergillus.

There was a peak around 3280 cm^–1 in the A. niger spectrum (Figure 5A) which was much less visible in the case of Chlorella sp. (Figure 5B). This absorption may be assigned to amines, which were detected in various fungal species and strains (Stüttgen et al., 1978; Karimi et al., 2019). Johnson (1965) reported that the cell wall of A. niger contains approximately 9–13% hexosamines. Amine and carboxylate groups were identified as major functional groups for metal biosorption in the A. niger biomass (Cai et al., 2016). This is in accordance with results presented here, as one of the peaks of the correlation spectrum reached a maximum at 3284 cm^–1, which lies within the amine frequency interval. Specifically, IR absorption around 3280 cm^–1 is caused by the N–H stretching of secondary amines (Smith, 2019). However, this peak may also be associated with the amides bound in proteins (Forfang et al., 2017) or chitin (Rinaudo, 2006), as stretching vibrations of the N–H group and associated absorption around 3280 cm^–1 are characteristic for both of these components. The peak with a maximum at 2143 cm^–1 in the A. niger spectrum can be associated with isocyanides, which have been identified in various microorganisms, including Aspergillus species (Massarotti et al., 2021). According to Bittante and Cecchinato (2013), isocyanides show characteristic absorption between 2110 and 2165 cm^–1.

Another difference observed in the two spectra is the peak at 1602 cm^–1, which was absent in case of Chlorella sp. The absorption at frequencies around 1600 cm^–1 is associated with vibrations of the aromatic ring and the presence of melanin pigments in the fungal biomass (Meenu and Xu, 2019). Information on the synthesis of various pigments by filamentous fungi, including Aspergillus species, were reviewed by Kalra et al. (2020). Significant differences between the spectra of A. niger and Chlorella sp. were detected in the regions of amide vibrations, particularly in the amide I and II spectral bands. This is related mainly to the differing overall protein content in the biomass of the two compared microorganisms, which was probably significantly higher in the case of Chlorella sp. Amide I and II bands are typically represented by two distinct peaks with maxima at around 1656 and 1547 cm^–1, respectively (Lorenz-Fonfria, 2020). These two peaks are very apparent in the Chlorella sp. spectrum, whereas they are less prominent in the case of A. niger. The absorption in each of the two amide bands is associated with the combination of vibrations of several functional groups. The amide I peak results from the stretching of the C=O and C–N groups, respectively, as well as N–H bending. Amide I absorption is primarily caused by the C=O stretching vibration of the amide group, with weaker contributions from the amide C–N stretching and the N–H bending. In the case of the amide II band, the absorption is caused mainly by N–H bending and C–N stretching (van Hoogmoed et al., 2003; Lorenz-Fonfria, 2020). However, due to the complex and variable structure of proteins, slightly differing frequencies are reported for the amide I and amide II bands when data from various sources are compared.

The spectra of A. niger and Chlorella sp. also exhibited different absorption patterns between 1470 and 1430 cm^–1. Whereas the spectrum of Chlorella sp. showed only one peak (1454 cm^–1) in the considered interval, the spectrum of A. niger showed a triplet of peaks, with maxima at 1469, 1450, and 1431 cm^–1, respectively. There are two adjacent bands in which the vibrations of aliphatic structures are detected: asymmetrical CH bending in the methyl group at 1470 cm^–1 and methylene scissoring at 1465 cm^–1 (Stuart, 2004). Other works report slightly different frequencies for these two bands. For example, Hollas (2004) associates the 1444 cm^–1 frequency with asymmetric deformations of CH₃, while Haneef et al. (2017), who used fungal mycelia as a source of various fibrous materials, linked symmetric bending of CH₂ with the absorption around 1450 cm^–1. Although various authors who analyzed a fungal biomass associated the peaks between 1465 and 1450 cm^–1 with the vibrations of the CH groups in lipids (Bhat, 2013; Lecellier et al., 2014; Haneef et al., 2017; Kosa et al., 2018), it is worth noting that the interval between 1500 and 1300 cm^–1 is often termed as the “mixed” region. Besides lipids, vibrations of functional groups bound in polysaccharides, proteins and triterpene compounds may be detected in the considered interval (Meenu and Xu, 2019; Saif et al., 2021). The majority of absorption at 1430 cm^–1 was probably caused by C–O–H bending in polysaccharides (Song et al., 2015; Meenu and Xu, 2019), but some contributions from the COO^– group in carboxylic acids cannot be ruled out (Max and Chapados, 2004; Stuart, 2004).

In the case of the Chlorella sp. spectrum there is a relatively broad peak at 1397 cm^–1, which can be associated with asymmetric stretching of COO^– (van Hoogmoed et al., 2003) and at the same time with the symmetric bending of CH in the CH₂ and CH₃ groups, respectively (D’Souza et al., 2008). A smaller peak at 1377 cm^–1 can be assigned to symmetric bending of the aliphatic CH₃ (Kosa et al., 2018; Meenu and Xu, 2019). Haneef et al. (2017) associated the absorption around 1375 cm^–1 with the CH bending of chitin molecules in the fungal mycelium. In our case, the peak was present only in the A. niger spectrum, which supports the assumption that it was caused by chitin. The relatively sharp absorption at 1312 cm^–1 was probably caused by the C–O stretching of carboxylic acids (Meenu and Xu, 2019) and/or the asymmetric C–N–C stretching of aromatic amines (Stuart, 2004; Silverstein et al., 2005).

The peaks at 1250 (A. niger) and 1242 cm^–1 (Chlorella sp.), respectively, can in both cases be linked to the amide III band, where the absorption is associated with CN stretching, NH bending and CO in-plane bending (López-Lorente and Mizaikoff, 2016). It is probable that the absorption in this region was also partially caused by asymmetric P=O stretching in nucleic acids (Haneef et al., 2017) and phospholipid substances (Kosa et al., 2018). From a comparison of the spectra it follows that the peak around 1250 cm^–1 was more intense in the case of Chlorella sp. than in A. niger. This suggests that the absorption was caused mainly by proteins and also partially by phospholipids, as their portion in the biomass is higher in comparison to nucleic acids (especially in Chlorella sp.).

The majority of the absorptions detected between 1200 and 900 cm^–1 originate from the vibrations of carbohydrates. Specific functional groups and related frequencies include stretching vibrations of the COC (1150 cm^–1), CO (1075 cm^–1), and CC (1052 or 1034 cm^–1) groups (Bhat, 2013; Baeva et al., 2019). In addition to carbohydrates, the C–O stretching of alcohols and sulfoxides occur at 1053 and 1034 cm^–1, respectively (Rahman et al., 2014). However, the concentration of these substances in the biomass of both organisms is probably significantly lower in comparison to carbohydrates. Although the spectra of the A. niger and Chlorella sp. biomass were interpreted separately in this section, the adsorption experiments were carried out using the “mixture” of these two microorganisms. The spectrum of this this sample (without adsorbed Ni) is shown in Figure 5C.

The interaction between the microbial biomass and the Ni substrate was assessed on the basis of a correlation spectrum. In the spectrum each value expresses the collinearity between the absorbance detected at certain wavelength and the amount of Ni retained by the sample. This approach enables better identification of the sorption sites in comparison to simple visual observation of the spectra. This follows from Figure 6, where the spectra are sorted according to the concentration of adsorbed Ni. It is hard to identify the change in absorbance or the shift in peak positions in the plotted spectra which would be related to concentration of Ni in the sample. Among the few visible changes in the spectra is the increase in absorbance between 1600 and 1400 cm^–1, which was at the same time accompanied by a higher amount of adsorbed Ni. In this regard, the correlation spectrum is more informative, as it indicates which functional groups are involved in the sorption of Ni.

The significance of the correlation (P < 0.05) was exceeded at frequencies characteristic for various organic groups (Figure 7). The first two peaks at the margin of the displayed spectral range, around 3284 and 3062 cm^–1, may be assigned to stretching vibrations of the NH group in amides (Parker, 1971; van Hoogmoed et al., 2003; Forfang et al., 2017). However, the origin of the 3284 cm^–1 peak may also be associated with OH stretch due to the formation aquo-metal complexes (Zhang et al., 2021). The four adjacent peaks, within the 2960 and 2850 cm^–1 interval, represent the different stretching vibrations of aliphatic CH groups, specifically: methyl symmetric CH stretching at 2960 cm^–1, methylene asymmetric CH stretching at 2930 cm^–1, methyl asymmetric CH stretching at 2870 cm^–1, and methylene symmetric C–H stretching at 2850 cm^–1 (Stuart, 2004). A significant correlation between absorbance and the content of adsorbed Ni was also observed at 1695 and 1632 cm^–1, respectively. The more apparent peak at 1695 cm^–1 is due to the stretching vibrations of the C=O group, which is a part of various organic structures, including carboxylic acids (COOH), amides (CONH) or esters (COOR). On the other hand, the smaller peak with a maximum around 1630 cm^–1 probably represents the C=O stretching vibration of a dissociated COO^– group (Antisari et al., 2011; Kang et al., 2011).

The correlation between absorbance and the content of adsorbed Ni showed higher values in the amide I and II bands, which correspond with the intervals of 1710–1600 and 1580–1520 cm^–1, respectively (Lorenz-Fonfria, 2020). As was mentioned in the previous section, absorbance in amide I band is associated with the symmetric stretching of C=O and C–N together with N–H bending, whereas in case of the amide II band, the absorbance is indicative mainly of C–N stretching and N–H bending vibrations (Lecellier et al., 2014).

Although the correlation spectrum did not show peaks with maxima at 1650 or 1600 cm^–1, the r values at these wave numbers clearly exceeded the significance level (P < 0.01). The mentioned frequencies are associated with vibrations of aromatic carbon (Rahman et al., 2014; Meenu and Xu, 2019). Some amino acids (and their metabolites), which are synthesized by algae and other microorganisms, contain an aromatic ring in their structure (Żyszka-Haberecht et al., 2019). Aromatic structural units are also bound in various pigments, some of which (melanins) are present in fungal mycelium (Rahman et al., 2014). It cannot be ruled out that the mentioned substances contributed to the sorption of Ni by the microbial biomass.

A significant correlation between adsorbed Ni and absorbance was observed at 1558 and 1508 cm^–1, respectively. Both peaks are within the amide II interval, where IR absorption is caused by the combination of N–H bending and C–N stretching vibrations. In addition, the band around 1558 cm^–1 was also associated with asymmetric stretching of the COO^– group of amino acids (Barth, 2000, 2007), carboxylic acids and their salts (Max and Chapados, 2004). This suggests that carboxyl groups can effectively contribute to Ni adsorption. The coefficient of correlation reached the highest value (0.9) at 1508 cm^–1. In a study of a similar nature, where the removal of Cd from water by cyanobacterial biomass was assessed, Bon et al. (2021) assigned the IR adsorption at 1510 cm^–1 to N–H bending. Besides the mentioned C–N stretching and N–H bending, vibrations of aromatic carbon may be detected at frequencies around 1508 cm^–1, which means that various compounds containing an aromatic ring cause the absorption at the given wave number. In a number of works dealing with the composition of soil organic matter, the peak around 1510 cm^–1 was associated with the stretching vibrations of aromatic carbon (Haberhauer and Gerzabek, 1999; Pietikäinen et al., 2000; Jiménez-González et al., 2019). Similar findings are reported in studies on the analysis of plant materials, where absorption at 1515 cm^–1 indicates valence vibrations of aromatic C in phenolic compounds (Heredia-Guerrero et al., 2014). It is worth noting that Barth (2007) associated the band around 1510 cm^–1 with the detection of amino acids containing an aromatic ring in their structure.

The correlation between absorbance and the content of adsorbed Ni also showed a maximum at 1456 cm^–1. This peak probably represents the bending of CH₂ groups in lipids, aliphatic structures (Mayers et al., 2013; Lecellier et al., 2014; Kosa et al., 2018), and polysaccharides (Meenu and Xu, 2019). The adjacent peak at 1410 cm^–1 may be attributed to the asymmetric stretching of the COO^– group (van Hoogmoed et al., 2003). Even though the correlation culminating around 1265 cm^–1 reached lower values in comparison to the IR bands mentioned above, the P < 0.05 significance threshold was exceeded. This band may be associated mainly with amide III vibrations, which consists of CN stretching, NH bending and CO in-plane bending (López-Lorente and Mizaikoff, 2016). Besides amides, IR absorption in this region is caused by asymmetric stretching of the P=O group in nucleic acids and phospholipids (Mayers et al., 2013; Lecellier et al., 2014; Kosa et al., 2018).

Absorption sites for Ni were found to provided mainly by amide, esters and lipidic structural units, whereas carbohydrates probably played a less important role in the sorption of Ni. This follows from the correlation between absorbance and the content of adsorbed Ni. In addition to the mentioned functional groups and structural moieties, it is also possible that other components of microbial biomass contributed to Ni adsorption, for example, amines or compounds containing an aromatic ring. On the other hand, the r values were mostly insignificant in the frequency interval characteristic for the vibrations of carbohydrates (between 1200 and 900 cm^–1). It should be noted that carbohydrates also cause absorption in a relatively broad region above 3000 cm^–1, where OH groups are detected. However, interpretation of the peaks between 3800 and 3000 cm^–1 is often ambiguous, as the absorbance in this interval is strongly affected by the water content in the sample. Although we dried all samples before FTIR measurements, it is difficult to prevent the absorption of water from the atmosphere by sample material, which proceeds quite rapidly.

The following section compares our findings with the results of other works assessing biosorption. Based on the peak position shift, Kang et al. (2011) and Guarín-Romero et al. (2019) both suggested that mainly OH and HN groups are involved in the metal uptake by the algal biomass. This is in general accordance with the results presented here, as we observed that the amount of adsorbed Ni and absorbance in bands of OH and NH vibrations show significant correlation. Rahman et al. (2014), who studied the adsorption of metal ions on the surface of Trichoderma presented similar findings. They found that the adsorption of metal caused changes in the FTIR spectra mainly in the 3800–3000 and 1600–1700 cm^–1 intervals, where the vibrations of hydroxyl, carbonyl, carboxyl, and amide groups are detected, respectively. Besides the mentioned bands and functional groups, we also observed that other structural units may be involved in biosorption, which follows from the significant correlation of the 1600–1400 cm^–1 interval. A substantial part of this region corresponds with the amide II band, but at lower wave numbers the vibrations of the other groups manifest, including CH and COO^– groups or an aromatic ring. Our observations are in relative accordance with the findings of Pradhan et al. (2007), who reported an increase of absorbance at 1550, 1538, and 1513 cm^–1, respectively, due to a change of the C=O stretching after Ni sorption on Microcystis capsules. We observed a significant correlation between absorbance detected at mentioned frequencies and the content of adsorbed Ni. The calculated r values were in this case above 0.87 (P < 0.001), and the correlation reached a maximum at 1508 cm^–1.

Although the value of the correlation coefficient and the peak positions indicate which functional groups are involved in the sorption of Ni, some aspects of this approach should be considered. A significant positive correlation between the content of adsorbed Ni and the absorbance detected at specific frequency does not prove that the group itself is directly involved in the sorption of Ni. Even if we exclude the overlapping of the respective spectral bands (which is a factor that is hard to eliminate), it is still possible that the considered functional group is part of a molecule that may be involved in the sorption process, but the group itself is not. In this study we observed a significant positive correlation in the case of IR bands associated with several functional groups, including OH, CH, COO^–, C=O, and NH groups, respectively. However, the presented FTIR results did not prove that all the mentioned groups were actively participating in the adsorption of Ni. It is possible that the structural units (or molecules) of the microbial biomass which were involved the adsorption of Ni contain the mentioned functional groups but that some of them (CH groups) did not participate in the adsorption. On the other hand, there are different adsorption mechanisms. Besides chemical bonding, which takes place between Ni cation and negatively charged functional groups (such as COO^–), weaker electrostatic attraction plays a role in biosorption, as well (Ren et al., 2021). In such a case, less reactive structural units, including non-polar CH groups, also take part in the retention of Ni. These factors probably contributed to the significant positive correlation that was observed between the content of adsorbed Ni and absorbance in the bands of CH vibrations (and possibly also vibrations of an aromatic ring).