Adsorption membranes are one of the best ways to obtain semiconductor composites from aqueous solutions. The established pore size less than 2 nm indicates that the PA 6 film is microporous. Therefore, it can adsorb Se micro particles and Cd2+ and Ag+ cations followed by their diffusion into the polymer membrane. In most cases, adsorption membranes are charged. There is an attraction of charges between the ions and the surface of the adsorption membrane.
3.1. XRD Analysis
To determine the structural properties of the PA 6 and the obtained composites, XRD patterns were analysed. The XRD pattern of PA 6 sample heated in distilled water is shown in
Figure 2and 4, while XDR pattern of the obtained composites is shown in Figure 4.
PA 6 crystallizes into stable α- and γ-, as well as unstable β-forms [
41]. XRD studies of the PA 6 sample revealed the presence of a main peaks at 2
θ 9.52°, 20.26°, 23.56° and 28.64°. The small diffraction peak at 2
θ = 9.52°, which was attributed to the (020) reflection of the γ crystalline form [
40]. The other two peaks at 2
θ = 20.26° and 23.56° were assigned to the (200) reflection and (002)/(202) reflections of the α crystalline form of PA 6 [
41]. The peak at 28.64° appears after polymer treatment in hot water and cannot be assigned for neither forms of the crystalline PA 6. When the polymer is exposed directly to water, water initiates degradation/hydrolysis reactions in the polymer.
Mixing H
2SeO
3 and Na
2SO
3 solutions leads to the release of Se according to the following reaction [
42]:
A red film was observed on PA 6 (
Figure 3). The precipitate, characterised by XRD analysis (
Figure 3 and
Figure 4), was identified as red amorphous selenium (a-Se).
Because intensities of the diffraction peaks of PA 6 are very high and overlap with main peaks of the materials inserted/deposited on the polymer, the XRD patterns of the obtained composites were given in the 2
θ assortment of 30–65°. As can be seen from
Figure 4, in the a- Se/PA6 composite, Se formed at the end of reaction (9) is in the amorphous form, i.e., not detected by XRD analysis.
The obtained positions of the diffraction peaks at 2
θ 42.74°, 43.46°, 48.68°, 51.09° in the Cd-Se/PA6 composite (
Figure 4) were compared with standard values and mainly indexed as CdSe (PDF 77-2307). The peaks at 2
θ 32.66°, 33.49°, 37.64°, 38.46°, 45.16°, 52.66°, 53.77°, 54.83° and 59.73° are indexed as Se
8 (PDF no. 71-528) with hexagonal and monoclinic unit cells, respectively (
Figure 4, Cd-Se/PA6 composite). It was found that CdSe has a polycrystalline nature with predominant (110) orientation plane. These experimental results agree well with the literature data. It is well known [
43] that in the case of hexagonal lattices the strongest chemical bonds are formed by atoms located in a semi crystalline configuration on crystallographic planes (110) or (101).
Based on the analysis of the XRD results, it can be concluded that the immersion of the a-Se/PA6 nanocomposite in a Cd(NO
3)
2 solution at a temperature of 80±1
oC contributes to the gradual crystallisation of red a-Se into grey cr-Se. It is very reactive and reacts with many metal cations [
44,
45]. The solubility of a-Se (2.2∙10
‒8 mol/dm
3 [
46]) was approximately an order of magnitude higher than the solubility of cr-Se (1.5∙10
‒9 mol/dm
3 [
46]). The formation of CdSe can be explained by the following reaction [
47,
48,
49]:
The formation of the Cd-Se/PA 6 composite leads to a colour change of from red to brown (
Figure 1).
The Ag-Se-6/PA 6 nanocomposite shows diffraction peaks corresponding to the orthorhombic phase of Ag
2Se naumannite (PDF No. 01-071-2410). As can be seen from
Figure 4, the dominant peak (121) of the orthorhombic system represents the preferred orientation along this plane. The identified peak positions are in good agreement with those indicated in the literature for Ag
2Se nanowires [
50] and nanoparticles [
51]. The diffraction pattern of the Ag-Se/PA 6 nanocomposite shows two sharp lines of approximately the same intensity along the (112) and (121) planes respectively. The reaction involved in this process can be summarised as the following [
47,
48,
49]:
The formation of Ag
2Se leads to a colour change from red to black. Together with the Ag
2Se phase, a minor amount of monoclinic Se
8 (PDF No. 71-528) at 2
θ 32.66° Se
8 may remain unreacted in the deposited film phase. Ag-Se/PA6 composite (
Figure 4) XRD analysis showed that not only Ag
2Se but also the metallic Ag phase (PDF No. 04-003-1472) was identified in the diffractogram. Typically, the metallic structure of Ag is depicted by a sharp XRD peak at 2
θ 38.12° corresponding to the preferred (111) texture. Metallic Ag is the most likely impurity in chemically deposited Ag
2Se layers [
50] as shown in the reaction as follows:
As discussed in [
52], excess Ag can be introduced in several ways: in the form of adsorbed metal chains, in the form of point defects, or in the form of three-dimensional nano- or micro-inhomogeneities.
The XRD pattern of the Ag-Cd-Se/PA 6 composite shows the multiphase crystalline composition of the film. Compared to the XRD pattern of the Cd-Se/PA 6 composite, Se
8 and CdSe phases remain. The XRD peaks corresponding to CdSe are not shifted, which indicates the same structure, regardless of the cation exchange with Ag
+ ions and the appearance of new phases associated with orthorhombic Ag
2Se and metallic Ag. On
Figure 4 also shows that the intensity of the peaks of the Ag-Cd-Se/PA 6 composite increases with the appearance of new phase, which additionally shows an increase in crystallinity. Broadening of the diffraction peaks is associated primarily with the finite sizes of the crystallites. The formation of Ag
2Se in Cd-Se/PA6 composite leads to a colour change from brown to black.
Since the other phases were not found in Cd-Se/PA6, Ag-Se/PA6 and Ag-Cd-Se/PA6 nanocomposites, it could be assumed that the SeO32–, SO32– and SO42– ions diffused from the a-Se/PA 6, reacted with the Cd2+ and Ag+ ions to form these compounds in solution near the sample area. In another case, by-products of Equations (10) and (11) due to the sufficiently high solubility (Table 2, CdSeO3 (KSP = 5.0·10‒9 (mol/dm3)2) and Ag2SeO3 (KSP = 1.35·10‒16 (mol/dm3)3) were removed from the surface of the samples by rinsing the as-synthesized composites with excess hot water.
The structural parameters of PA 6 and the prepared composites calculated from XRD data are listed in
Table 1.
A theoretical framework provides a general background to support our investigation. The sequence of phase formation in multicomponent systems depends on the structure of the starting material (SM). It is common knowledge that the exchange of cations can be achieved through simple mutual diffusion with other cations [
53]. In addition, considering phase formation, the driving force of the solid-state interaction is also determined by the change in the Gibbs free energy, since the system always tends to the lowest possible free energy state [
54]. The feasibilities of possible CE reactions were studied by thermodynamic deduction based on the thermodynamic law. CE reactions based on major difference in solubility product (
KSP) between the starting material (SM) and the final product (FP). Gibbs free energy (Δ
G) for insoluble materials expressed according to Equation (15):
where
R—universal gas constant (8.319 J/mol·K),
T—temperature (K),
KSP—the solubility product of the SM (in our case CdSe),
ICP—the product of the concentration of FP ions (in our case Ag
2Se).
ICP can be calculated according to Equation 16:
where
and
—the concentration of ions (mol/dm
3). In this way, the reaction of transition from a material with higher
KSP (CdSe) to a lower one (Ag
2Se) can proceed spontaneously. However, it is not easy to judge whether the CE reaction can proceed or not since this process is very complex and is associated with many factors: barrier to the activation energy of the phase transition, change in surface free energy, temperature, the active concentrations of ions in the solution, etc. [
55]. Therefore, a more correct criterion should be the difference in solubility (
S) of the SM and the FP (see
Table 2).
The stoichiometric reactions at an interface between the Cd-Se/PA 6 and Ag
+ salt aqueous solution can be considered as follows:
The negative values of ΔG°
reac,298 indicate that thermodynamic conditions are provided for the proceeding of this reaction. However, it should be emphasised that in the case of solid-phase interaction in multicomponent systems, phase formation at the precursor/solution interface is a dynamic non-equilibrium process that requires careful consideration of the mechanistic pathways along which they proceed [
57]. Apparently, the determining criterion for the possibility of a thermodynamic reaction in hybrid materials is the solubility of CdSe in the starting Cd-Se/PA6 material. The higher solubility of CdSe (
Table 2) favours the transformation into Ag
2Se. When Cd-Se/PA6 is immersed in AgNO
3 solution, Ag
+ reacts with Se
2‒ ions formed from the ionisation of dissolved CdSe leading to the formation of Ag
2Se particles, which results in the reduction of Cd
2+ ions in the solution. Before explaining the formation of the Ag
2Se phase, it is necessary to discuss the active species that might be involved in the reaction path. Based on the discussion presented above, we speculated that the formation of Ag
2Se phases in the Cd-Se/PA6 composite could be explained through complex mechanism reactions: Ag
+ ions were attracted to the surfaces of metal chalcogenides particles due to the large specific area, then adsorbed Ag
+ replaced Cd
2+ ions from the CdSe lattice and formed Ag
2Se. In parallel, some Ag
+ ions featuring relatively high ion diffusivity [
58,
59] propagate into the interior of the polymer and react with the cr-Se to form Ag
2Se according to reaction (13). The obtained composites (denoted in the text as a-Se/PA6, Cd-Se/PA6, Ag-Se/PA6 and Ag- Cd-Se/PA 6) were homogeneous with good adhesion.
3.2. AFM Analysis
The morphological changes of the PA 6 surface before and after subsequent chemical incorporation of a-Se, and Cd2+ and Ag+ cations were characterized by AFM.
One can see that the surface morphology of all nanocomposites varies depending on the incorporated element. AFM was used to elucidate the formation mechanism and homogeneity of obtained composites. On a selected small scan area of 5 µm × 5 µm, 2D height data (
Figure 5A) and 3D topographic (
Figure 5B) AFM images of the respective PA 6 membrane, a-Se/PA 6, Cd-Se/PA 6, Ag-Se/PA 6 and Ag-Cd-Se/PA 6 nanocomposites with the corresponding measured topographic parameters are shown in
Figure 5. The morphological pattern of PA 6 membrane revealed a wavy surface. a- Se/PA 6, Cd-Se/PA 6, Ag-Se/PA 6 and Ag-Cd-Se/PA 6 composites had a rough surface. The 3D topographic images (
Figure 5B) show an irregular pattern resembling a valley of nodules. Since the composites did not have a pronounced texture but were characterized by a complex relief morphology, the roughness parameters were calculated from profilograms obtained for different surface areas (
Figure 5C). The length of the scan areas was approximately 4.5–6.5 μm. Surface roughness is an important indicator of the quantitative characteristics of the sample surface. The average height, Z
mean (156.9 nm), and the average surface roughness, R
a (27.61 nm), of the a-Se/PA 6 composite are clearly different from those of the unreacted PA 6 membrane (Z
mean 60.4 nm, R
a 10.78 nm) (
Figure 5A,B). As shown in
Figure 5 local areas of the surface of the PA 6 membrane and the a-Se/PA 6 composite have different morphological structures. Simultaneously, isolated solid particles of amorphous Se (
Figure 5) functioned as seeds to start the formation of CdSe and Ag
2Se. When a-Se/PA 6 composite was exposed to an aqueous Cd(NO
3)
2 and AgNO
3 solutions, a-Se reacts with Cd
2+ and Ag
+ cations to form insoluble CdSe and Ag
2Se nanoparticles that deposit in situ on the a-Se/PA 6 and produce the Cd-Se/PA 6 and Ag-Se/PA 6 composites (according to Equations (12) and (13)). The average height Z
mean (53.57 nm) and the average surface roughness R
a (9.82 nm) of the Cd-Se/PA 6 composite significantly decreased compared to the a-Se/PA 6 composite. It implies that CdSe formation proceeded mainly through deposited a-Se layers within the PA 6 surface. The average height Z
mean (35.10 nm) and the average surface roughness R
a (8.25 nm) of the Ag-Se/PA 6 composite significantly decreased compared to the a-Se/PA 6 composite. After exposure of the Cd-Se/PA 6 composite in an aqueous solution of AgNO
3, submicron pyramidal particles, presumably belonging to Ag
2Se (according to Equation 13), are found on the surface of the resulting Ag-Cd-Se/PA 6 composite (
Figure 5). Pyramid-like structures with a lateral size of about 150 nm are also visible. Some of these particles aggregate on the composite surface. Small dark spots at the boundaries of grains and agglomerates may be due to depressions in the layers. From the roughness values presented in
Figure 5, we can conclude that the quantitative parameters of the surface of the Ag-Se/PA 6 and Ag-Cd-Se/PA 6 composited decreased compared to the Cd-Se/PA 6 composite. For Ag-Cd-Se/PA 6 composite R
a decreased from 9.82 nm to 8.36 nm, and R
q decreased from 12.17 nm to 11.55 nm. In addition, the peak-to-valley roughness R
t of the surface of the Cd-Se/PA 6 and Ag-Cd-Se/PA 6 composites remain virtually unchanged. That confirmed our assumption about the formation of compact films. RMS is defined as the standard deviation of the surface height profile from the average height. The profilograms indicate a slightly asymmetric surface with more peaks than valleys (
Figure 5C). It is important to note that the RMS of the deposited films decreased during the formation of CdSe, Ag
2Se and mixed CdSe-Ag
2Se nanoparticles compared to the a-Se films deposited on the PA 6 surface.
3.3. ATR-FTIR Analysis
The FTIR study allows to discover the conformational structural changes of polymer chain as well as and the complexation and interaction between the polymeric matrices and inorganic nanoparticles. The ATR-FTIR spectra of the PA 6 and obtained nanocomposites within a wave number range from 650 to 1650 cm
−1 are displayed in
Figure 6. PA 6 consists of (CH
2)
5 segments separated by a parallel or antiparallel arrangement of secondary amide groups [
60]. The bands related to NH–CO fragments are mainly affected by the mutual arrangement of H–bonds, while the vibrations from CH
2 groups reflect the effect of the chain conformation [
61]. In PA 6 spectrum the absorption bands at 834 cm
‒1, 929 cm
‒1, 960 cm
‒ 1, 1031 cm
‒ 1 and 1200 cm
‒ 1 (
Table 3), confirm predominantly the α-crystalline phase. The spectral features at 1476 cm
−1 and 1417 cm
‒1 represent CH
2 scissoring vibrations next to −NH and >C=O groups, respectively. The band at 1463 cm
−1 corresponds to CH
2 scissoring vibrations. Two typical absorption bands at 1633 and 1535 cm
‒1 represents the amide I and II group, respectively.
The FTIR vibration bands of Se
8 rings are reported at 487.9 cm
−1 and 737.7 cm
−1 [
62]. These peaks are absent in the FTIR spectrum of a-Se/PA 6, which confirms the XRD data that the inserted selenium is amorphous. Incorporation of solid inorganic nanoparticles into flexible polyamide network causes chain stretching. This should cause a conformational change in the chains of the polymer matrix [
61,
63]. It can be seen that most of the characteristic vibration bands of PA 6 in the wavenumber region from 650 to 1650 cm
−1 are retained in the FTIR spectrum of a-Se/PA 6; however
the intensity of the peaks changes: some of them become more intense and some
of them expand.
It can be concluded that as a-Se nanoparticles are introduced between the polymer chains and stretched, the position and intensity of the peaks change. Thus, the peak of neat PA 6 at 1124 cm
−1 in a-Se/PA 6 spectrum shifts to 1112 cm
−1, and two new small peaks are recorded at 974 cm
−1 and 1390 cm
−1. The shoulder at 974 cm
−1 shows the appearance of a small amount of the γ crystalline form of PA 6, and C–C symmetry vibration causes IR absorption at the
1390 cm–1 [
64].
CdSe and Ag
2Se have no specific vibrations in the spectral region between 650 and 1650 cm
– 1, but exhibit absorption in the region of 200–400 cm
–1 [
65,
66,
67]. Thus, it is possible to avoid the problem of overlapping of metal selenides and PA 6 characteristic peaks. The IR spectra of the Cd-Se/PA 6, Ag-Se/PA 6 and Ag-Cd-Se/PA 6 composites are very similar to the a-Se/PA 6 spectrum, but show an even greater shift of the peak at 1112 cm
−1 to lower frequencies (
Table 3) and the appearance of another new peak at 877 cm
−1. By analogy with data reported for CdSe [
61], those weak vibrations can be attributed to the Cd–Se or Ag–Se bond. Another dimension of the functionality and complexity of inorganic-organic materials is introduced if the selenides are combined with each other. For example, the negative energies of chemical band formation of CdSe and Ag
2Se promote the creation of Ag–Cd–Se– interfacial bonds, which leads to non-selective nucleation in Ag
2Se/CdSe hetero-nanostructures [
68]. Thus, the peak at 1108 cm
−1 in the IR spectrum of the Ag-Se/PA 6 composite shifts to 1094 cm
−1 in the IR spectrum of the Ag-Cd-Se/PA 6 composite.
The shift and broadening of the C–C stretching vibration peaks, accompanied by the appearance of new small peaks of low intensity, allows us to conclude that we are observing the formation of intermediate structures inside the polyamide chains. Thus, we can be conclude, that the spectral changes after insertion of a-Se and ions of Cd
2+ and Ag
+ ions with respect to PA 6 sample belong to the bands related to the C–C conformation of the main chain. The peak at 1124 cm
−1, arising from the all-trans C–C conformation of the main chain, loses its spectral position. In addition, the peak is broadened, and its intensity noticeably increases in all spectra of the obtained composites (
Figure 6).
3.4. UV-Vis Analysis
The experimental absorption spectra are presented in
Figure 7.
The absorption spectrum of PA6, (
Figure 7.) displays absorption peak at 225 nm. The color of the obtained a-Se/PA6 nanocomposite becomes red, suggesting the change of optical properties. In the UV-Vis absorption spectrum of the a-Se/PA6 nanocomposite, a number of peaks appear at 320 nm, 370 nm, and 475 nm, which characterize the formation of a band structure: red a-Se nanoparticles are introduced into the system of PA 6 micropores at the first stage, and then they combine into aggregates and agglomerates. According to the literature data [
69,
70] bulk a-Se absorbs 475 nm light.
In the UV-Vis spectrum of the Cd-Se/PA6 nanocomposite, peaks appear at the same 320 nm and 370 nm, which are related to Se nanoparticles and agglomerates, and a new peak at 535 nm. According to the literature data, CdSe quantum dots have a wide absorption band in the spectral range of 410–545 nm [
71], and the absorption maximum of CdSe nanocrystals, depending on the particle size, is in the range from 492 to 578 nm [
72]. Therefore, the peak at 535 nm can be attributed to CdSe.
In Ag-Se/PA6 UV-Vis spectrum the peaks appear at 320 nm, 365 nm and 475 nm. Ag
2Se nanoparticles absorb in a wide spectral region from 300 to 600 nm [
73], while two sharp peaks are observed in the spectrum of quantum dots at 440 and 607 nm, respectively [
74]. Monodisperse silver nanoparticles, depending on the size, shape, and distribution in nanostructures, have two absorption bands; one broad peak in the range 420-430 nm and a shoulder at 580-590 nm [
75]. Therefore, a broad peak from 380 to 590 nm with a maximum at 475 nm can be attributed to the Ag
2Se phase with a small amount of Ag nanoparticles.
In Ag-Cd-Se/PA6 nanocomposite UV-Vis spectrum the peaks appear at 315 nm and 555 nm. According to the literature data [
76,
77,
78], the introduction of Ag
+ ions into CdSe nanocrystals via the CE reaction does not lead to noticeable changes in the absorption spectra of CdSe in the UV-visible region. In contrast, other authors suggested that the formation of the CdSe-Ag
2Se core-shell leads to the complete disappearance of the CdSe absorption peak in the UV-Vis spectra [
79]. Thus, it can be stated that the absorption spectrum of the Ag-Cd-Se/PA6 composite is a superposition of the individual absorption spectra of the CdSe, Ag
2Se, and Ag phases, its absorption edge is shifted to the longer wavelength region, and their absorption increases in the range of 315–555 nm compared to with the region (320–535 nm) of Cd-Se/PA6 and Ag-Se/PA6 composites. This red shift of the
λg may be related to the common complex alignment of the valence and conduction bands, as well as to the presence of multiple defects and vacancies. The determination of the phase composition of the obtained composites according to the data of optical studies is in good agreement with the XRD data.
The experimental UV-Vis spectra better correspond to the rectilinear section on the graphs 𝐴
2·𝜆
–2 = 𝑓(𝜆
–1) (
Figure 8), which confirms the direct transitions of the obtained nanocomposites.
All the calculated optical parameters are given in
Table 4 for comparison.
The decrease in
Eg of each subsequent nanocomposite compared to PA 6 is due to a change in the they optical properties.
Eg of nanocomposites can be affected many factors, such as composition, particle size and distribution, shape, agglomeration/aggregation state, solubility, surface morphology/topography, structure crystallinity, and many different structural defects or deformations [
80,
81]. The wavelength at which the signal gets saturated is interpreted as being the value of the absorption edge or the
Eg of the semiconductor. For a-Se/PA6 composite
Eg is 2.23± 0.02 eV.
Eg measured for bulk a-Se is 1.99 ± 0.02 eV [
82]. Introduction of Cd
2+ ions proseed at 80
oC, and at heat treatment amorphous Se crystallizes.
Eg measured for cr-Se is 1.85 ± 0.02 eV [
82]. An increase in the crystallinity of the obtained composites (
Figure 4) and the carrier concentration is accompanied by a decrease in
Eg from 2.05±0.02 eV for the Cd-Se/PA 6 composite to 1.64±0.02 eV for the Ag-Cd-Se/PA 6 composite (
Table 4). The obtained values are in good agreement with the literature data: CdSe and Ag
2Se have a band gap of 1.9 eV [
20] and 1.8 eV [
21], respectively.
One of the sources of unfavorable recombination is the subband absorption tail, which is characterized by the Urbach energy. Absorption tails are defined as densities of states that extend from the bands into the band gap of the photoactive layer. Tail states inside the band gap act as traps and recombination centers and trap a charge carrier. The
EU is influenced by thermal and structural disturbances in semiconductor materials [
83,
84,
85]. In this work, we analyze structure-dependent
EU. The Urbach energy plots were constructed and shown in
Figure 9, and the calculated values are given in
Table 4.
EU reflects the density of states in the band tails and, hence, the local microstructural disorder [
86]. As can be seen from the values presented in
Table 4, the
EU of PA 6 is 1.31 eV, which is higher as compared with obtained nanocomposites. The relatively high value of
EU indicates a greater propensity for the transformation of weak structural bonds into defects [
87]. The a-Se/PA 6 nanocomposite (amorphous phase) is more structurally disordered. With the stepwise incorporation of Cd
2+ and Ag
+ ions,
EU decreases. This may be due to an increase in the structural ordering of nanocomposites, which is confirmed by an increase in crystallinity detected by XRD analysis. In addition, it was previously reported [
88] that thicker selenide films have less structural randomness, since with an increase in the thickness and density of the layers, structural defects are minimized, which minimizes the
EU. The formation of compact these films was confirmed by AFM analysis.
The σ characterizes the broadening of the optical absorption edge arising from the electron-phonon interactions [
40,
89]. The steepness parameter of PA 6 is lower than that of the obtained nanocomposites at the same temperature (
Table 4). The lower differences in electronegativity decrease the orbital overlap, and decrease the band gap.