Influence of the ligand nature on the in situ laser-induced synthesis of the electrocatalytically active copper microstructures

1 Introduction

The method of laser-induced metal deposition from solution (LCLD) is one of the effective techniques for metallization of the surface of dielectrics and semiconductors of various types (Kochemirovsky et al., 2015; Manshina et al., 2007; Shafeev, 1993). This method is based on the metal reduction reaction, which proceeds in a local volume of a solution within the focal point of the laser beam and results in deposition of metal nano- and microstructures on the surface of a dielectric or a semiconductor. The advantage of LCLD lies in the fact that even the surface of the dielectric with large band gap (>3 eV) can be metalized without preliminary preparation in contrast with other similar techniques. In addition, LCLD is simple and cheap, i.e. there is no need to utilize complex and expensive equipment. Moreover, it provides a much higher rate of metal deposition compared to other similar methods, and, even more importantly, it allows to in situ synthesize the conductive small-sized metal structures of any given length with highly developed surface area, which can be further applied as sensory platforms for manufacturing the enzymeless electrochemical microsensors (Panov et al., 2016a, 2016b, 2017). At present, the synthesis of a series of metals including copper was performed using LCLD approach (Kochemirovsky et al., 2015; Shafeev, 1993). Here, copper attracts a special attention since it is cheap and highly conductive material, which is widely applied in microelectronics and in fabrication of the electrochemical sensors (Kordas et al., 2001; Shafeev, 1993).

As a rule, the results of Red-Ox reactions that occurred during the laser-induced metal deposition essentially differs from the results of the processes accompanying the conventional chemical metallization (Kochemirovsky et al., 2015). The reasons of this difference have not been studied in detail so far. Typically, in regard of LCLD, most of the researchers pay attention only to the effects of power and wavelength of laser radiation, a scanning speed of a laser beam, and the number of scans. However, basically nobody takes into consideration the influence of the solution components assuming that the sequence of the laser-induced autocatalytic reactions is similar with those corresponding to other methods.

It is known that Rochelle salt is one of the main components of aqueous solutions used in analytical chemistry, electrochemistry, metal-organic chemistry, and a large number of industrial technologies (Fu et al. 2011a, 2011b, 2013a, 2013b; Horner and Klufers, 2016; Rashidipour et al., 2017). The reason of such a widespread application is its unique ability to form strong soluble complexes with the most transition metals that allows to proceed the metal salt reactions in an alkaline environment preventing the formation of hydroxo complexes and hydroxides. Furthermore, tartrate as a ligand is widely applied in LCLD (Kochemirovsky et al., 2015), in which it plays a role of the component of solutions used for deposition upon laser irradiation. As it was shown earlier (Lozhkina et al., 2015; Tumkin et al., 2015), the physical and electrical properties of the copper deposits synthesized by LCLD as well as the rate of a chemical copper plating (Panov et al., 2016a, 2016b) are in the relationship with the structure of the copper complexes in solution.

Many works have been devoted to the study of the structure of the copper tartrate complexes (Blomqvist and Still, 1984; Horner and Klufers, 2016; Johansson, 1980; Prout et al., 1971; Schoenberg, 1971). It is considered to be that mono- and dinuclear species such as CuL, CuLH and Cu₂L₂ (where, L = C₄H₄O₆²⁻) are formed in an acidic medium, whereas oligonuclear species such as Cu₈L₆H₋₁₀ or Cu₆L₄H₋₇ are formed in a neutral medium (Horner and Klufers, 2016). Nevertheless, the structure of the copper tartrate complexes at high pH (greater than 12) is either not well-known or poorly studied even despite the fact that the systems with such a pH are frequently implemented in practice (e.g. LCLD, chemical copper plating, and Fehling's solution). Recently, the crystalline sodium, potassium, and cesium cuprates were obtained from tartrate solutions at pH equals to 13 (Horner and Klufers, 2016). The obtained forms of copper complexes were studied using combined potentiometric titration and UV–Vis spectroscopy in a wide range of pH; however, it was difficult to investigate the strongly alkaline region using the proposed approach due to a marked increase in the so-called alkali ion error. Thus, it is clearly seen that even despite the abundance of works, there is no reliable information on the structure of tartrate copper complexes formed in the strongly alkaline medium.

It was previously demonstrated that the optimal salt precursor for laser-induced copper deposition is copper(II) chloride (Manshina et al., 2007). Implementation of chloride rather than, for example, sulfate (Ramasubramanian et al., 1998) allows to use ATR-FTIR spectroscopy for the studies of high-alkaline solutions, where the chloride ion has no its own absorption bands. The crystals of Rochelle salt (Shyju et al., 2012), tartaric acid and its numerous derivatives were investigated using IR-spectroscopy (Bhattacharjee et al., 1989; Hasmuddin et al., 2015; Martin Britto Dhas and Natarajan, 2007). However, the assignment of the absorption bands was carried out using the standard frequency correlations rather than quantum chemical calculations, which are quite effective in the interpretation of the experimental results as well as in the modelling of both photochemical and thermal processes (El-Khoury et al., 2009; Sumita et al., 2009). Here, it should be also emphasized that influence of the ligand nature and the regime of copper coordination on the laser-induced metal deposition process was not studied in detail to date. This circumstance disables the fully establishment of the most important parameters for the in situ laser-induced synthesis of the conductive metal deposits. Moreover, the relationship between structure of copper complexes forming in solutions used for the laser deposition and electrochemical characteristics of the synthesized copper deposits was also not determined so far. In turn, the investigation of influence of the solution composition, concentration of the components and other chemical factors on the LCLD process will provide deeper understanding the reaction mechanisms involved in the laser-induced formation of the conductive metal structures with high electrocatalytic activity. Thus, in the current paper, based on DFT calculations and ATR-FTIR, UV–Vis, and Raman spectroscopic measurements, we illustrate how the structure of a ligand and a reducing agent affects the in situ laser-induced synthesis of the conductive copper microstructures exhibiting good sensory properties and suitable for further application in fabrication of new non-enzymatic electrochemical sensors and biosensors.

2 Experimental

All chemicals used in this work were high-purity reagents and were purchased from Sigma Aldrich. The experimental setup applied for the in situ laser-induced synthesis of copper microstructures has been described in detail elsewhere (Kochemirovsky et al., 2015). In general, the output from a continuous wave 532 nm diode-pumped solid-state Nd:YAG laser is split into two sets. The first one is focused on a special 0.5-mm-thick cell, more precisely, at the boundary region between solution used for the copper deposition and dielectric substrate, which are both located in this cell. In turn, this cell is placed on the computer controlled motorized stage and moved horizontally and vertically so that the focused laser beam is literally drawing copper lines on the surface of a dielectric substrate. On the other hand, the second portion of the laser output is sent to web-camera utilized for the in situ monitoring the LCLD reaction. As a result, the laser-induced synthesis of the 1-cm-long copper lines at the scanning speed of 2.5 μm s⁻¹ was carried out upon laser irradiation varied from 400 to 1000 mW. The laser-induced copper deposition was performed on glass-ceramics (Sitall ST-50-1). Sitall ST-50-1 is a crystalline glass-ceramics material composed of SiO₂ (60.5%), Al₂O₃ (13.5%), CaO (8.5%), MgO (7.5%), and TiO₂ (10.0%) (Kochemirovsky et al., 2015).

Energy dispersion of X-ray spectroscopy (EDX) and scanning electron microscopy (SEM) were used to study the atomic composition and topology of the synthesized copper deposits, respectively. The EDX set-up coupled with a Zeiss Supra 40 VP scanning electron microscope was equipped with X-ray attachment (Oxford Instruments INCA X-act).

The electrical resistance of these lines was measured using impedance meter Z-2000 (Elins Co.) operated within the frequency range between 20 Hz and 2 MHz at the signal amplitude of 125 mV.

ATR-FTIR spectra of aqueous solutions containing copper(II) chloride and Rochelle salt were recorded on a FTIR spectrometer Nicolet 8700 (Thermo Scientific) in a frequency range between 650 and 4000 cm⁻¹ with a 0.25 cm⁻¹ resolution. The solution of interest was placed on the surface of the diamond element. The measurements were conducted at the light angle of 42 degrees, and at the number of reflections and scans equals to 1 and 128, respectively. These experiments have shown negligible influence of the optical effects on ATR-FTIR spectra even at high concentrations of the components in aqueous solution. In order to obtain solid-state FTIR spectra the crystalline compounds were pressed into tablets (1.2 mg of the sample per 400 mg of potassium bromide).

Raman spectra of solid samples and aqueous solutions were recorded using Raman spectrometer Senterra (Bruker) coupled with Olympus confocal microscope attachment. The measurements were conducted upon 488-nm excitation at the laser power of 20 mW in a frequency range between 80 and 4500 cm⁻¹ with number of scans equals to 100.

Electronic absorption spectra were obtained using two-beam scanning spectrophotometer Lambda 1050 equipped with double monochromator. The measurements were performed between 250 and 850 nm with spectral resolution of 2 nm.

Electrochemical properties of the synthesized copper deposits were studied using a cyclic voltammetry and amperometry (potentiostat, Elins P30I). All measurements were carried out in a standard three-electrode cell using a platinum wire counter electrode, a Ag/AgCl reference electrode and an indicator electrode based on the deposited copper structures. The range of frequencies varied from 100 kHz to 1 MHz, the sweep speed of the potential was set at 50 mV s⁻¹, and the potential range was 8 V. The current sensitivity of the etalon pure bulk copper and the synthesized copper electrode was 10 and 1000 μA, respectively. The 0.1 M solution of Na₂SO₄ (pH = 9.8) was used as a background solution, and hydrogen peroxide of different concentrations was added to a background solution for the sensory activity studies. The amperometric response of the synthesized copper electrode to the consecutive additions of hydrogen peroxide was obtained at −240 mV.

For structures optimization and frequencies calculations we used a Minnesota density functional M11 and basis set 6-31g(d) for C, H, O and Lanl2DZ pseudopotentials and basis set were used for Cu, K and Na. To take into account electrostatic solvent-solute interaction polarizable continuum model (PCM) with dielectric constant epsilon = 80.4 (water, T = 293 K) were used. All calculations were performed using Gaussian 03C (Frisch et al., 2004).

3 Results and discussion

The typical copper deposits synthesized by LCLD are micro-sized nanostructured lines with width of 20–150 µm deposited on the surface of the inert material. Optical micrograph, SEM image, and results of EDX analysis of the synthesized copper structure are demonstrated in Fig. 1. Depending on the composition of the solutions used for the laser-induced deposition, topology of the synthesized copper deposits can be different, in turn, their electrical conductivity can vary from a high values, which are close to the conductivity of pure copper, to the complete lack of it. It should be noted that the electrical conductivity of copper structures produced by LCLD is a parameter that indirectly characterizes the rate of copper deposition, the amount of the deposited copper, dispersion, porosity of the copper deposit, and packing density of micro- and nanoparticles of which it consists, which is directly associated with topology of these copper materials.

Close

Fig. 1

Typical optical micrograph (a), SEM image (b) and EDX spectrum (c) of the copper structure (line) deposited upon laser irradiation from solution containing CuCl₂, КNaC₄H₄O₆·4H₂O, and NaOH at pH 12.5.

Export to PPT

In order to evaluate the relationship between the composition of solution and topology, the dependence of the electrical conductivity of the copper deposits on concentration of copper(II) chloride in the plating copper solution was obtained. This experiment was conducted at pH 12.5 using concentrations of CuCl₂ varied from 0.001 to 0.05 M and the ratio between CuCl₂ and КNaC₄H₄O₆·4H₂O concentrations equals 1–3. As a result, the conductivity of copper deposits changes nonlinearly and exhibits its maximum value at 0.01 M solution of copper(II) chloride (Fig. 2a). As can be seen from Fig. 2b, there is a region of component concentrations within the indicated pH range, where it is possible to produce a conductive copper deposits. Moreover, the study of a broader pH range (from 3.5 to 13.5) using solution containing 0.01 M CuCl₂ showed that in an acidic media (pH < 6) copper is produced as clusters and deposition of a conductive deposit is not possible (Fig. 2b).

Close

Fig. 2

The dependencies of the conductivity of the synthesized copper deposits on concentration of copper(II) chloride (a) and pH values (b).

Export to PPT

It is clear that pH has impact on the structure of the copper complexes in solution; therefore, apparently, topology of copper deposits depends not only on the solution composition but also the form of these complexes present in solution. On the other hand, the distribution of species in the copper(II) and L-(+)-tartaric system in acidic and neutral media have been already well-studied (Horner and Klufers, 2016) and the maximum electrical conductivity was observed at pH greater than 12. Therefore, all our further discussions will be focused on this range of pH. In order to understand the mechanisms involved in transformation of tartrate molecule from the form of crystalline tartrate salt to the form of copper complex the detailed spectroscopic study has been applied. Table 1 presents the results of DFT calculations, which consistent with ATR-FTIR experimental data on the maxima frequencies of the corresponding vibrations of sodium potassium tartrate tetrahydrate and the tartrate copper complex.

The ATR-FTIR spectra of solutions containing Rochelle salt, 0.1 M sodium hydroxide and water, in which concentration of КNaC₄H₄O₆·4H₂O was varied from 0.01 to 0.06 M, were recorded (Fig. 3a). For the sake of investigation of the complexation reaction between tartrate anions and copper(II) cations in alkaline medium at the concentrations optimal for formation of the conductive deposits, a series of solutions corresponding to system CuCl₂–КNaC₄H₄O₆·4H₂O–NaOH–H₂O was studied using ATR-FTIR spectroscopy (Fig. 3b).

Close

Fig. 3

(a) ATR-FTIR spectra of aqueous solutions containing КNaC₄H₄O₆·4H₂O of different concentrations and 0.1 M NaOH. (b) ATR-FTIR spectra of aqueous solutions containing КNaC₄H₄O₆·4H₂O of different concentrations, 0.01 M CuCl₂, and 0.1 M NaOH. In legend, the concentrations of КNaC₄H₄O₆·4H₂O are indicated in mol L⁻¹.

Export to PPT

Fig. 3b shows that with addition of copper a new absorption band centered at 1100 cm⁻¹ appears, moreover, there is no a frequency shift of the absorption band maxima except those corresponding to the absorption band centered at 1585 cm⁻¹. The comparison of the dependencies of the characteristic peak intensities on concentration of tartrate ions between КNaC₄H₄O₆·4H₂O–NaOH–H₂O and CuCl₂–КNaC₄H₄O₆·4H₂O–NaOH–H₂O systems results in the following features (Fig. 4). Fig. 4a shows the dependence of intensity of asymmetric vibration of carboxylate group at 1585 cm⁻¹ on concentration of sodium potassium tartrate tetrahydrate. The similar dependencies were obtained for other frequencies illustrated in Table 1 (Figs. 4b-f). As can be seen in Fig. 4a, Lambert-Beer law is well obeyed within concentration of tartrate range between 0.01 and 0.06 M, this fact allows to plot the calibration curves for “free” tartrate ligand and use them in order to obtain information about the nature of interaction of tartrate ion in systems containing copper salts. The intensities of vibrations at 1120 and 1068 cm⁻¹ corresponding to hydroxyl groups in solutions containing Rochelle salt are sensitive to the presence of copper cation in solution (Figs. 4b and c). The intensity of the absorption band centered at 1362 cm⁻¹ behaves analogically (Fig. 4d). According to calculations, the frequency at 1362 cm⁻¹ corresponds to a mixed vibration of the angles C—OH and C—C—H. Thus, the presence of copper in solution indirectly affects the value of optical density at the absorption band maximum. The intensity and position of the absorption bands centered at 1585 cm⁻¹, which correspond to asymmetric stretching vibration of carboxylate group (—CO₂—), are not changed with addition of Cu²⁺ into the solution and remain identical with ones observed for solutions without copper(II) chloride within entire concentration range (Fig. 4e). Furthermore, the analogical scenario was displayed by these two systems (with and without copper) with respect to the frequency corresponding to symmetric stretching vibration of carboxylate group (—CO₂—) (Fig. 4f). This suggests that carboxylate groups remain unbound in the solutions containing different proportions of the studied components. Raman spectroscopy also confirms this observation (Fig. 5a).

Close

Fig. 4

(a) The dependence of intensity of asymmetric vibration of carboxylate group at 1585 cm⁻¹ on concentration of sodium potassium tartrate tetrahydrate. The dependencies of intensity of the absorption bands centered at 1120 (b), 1068 (c), 1362 (d), 1585 (e) and 1400 (f) cm⁻¹ on concentration of КNaC₄H₄O₆·4H₂O. In legends, the concentrations of CuCl₂ are indicated in mol L⁻¹.

Export to PPT

Close

Fig. 5

(a) Raman spectrum of aqueous solution containing 0.32 M CuCl₂, 0.96 M КNaC₄H₄O₆·4H₂O and 1.28 M NaOH. (b) ATR-FTIR spectra of aqueous solutions containing 0.02 M CuCl₂, 0.06 M КNaC₄H₄O₆·4H₂O, and NaOH of different concentration. In legend, the concentrations of sodium hydroxide are indicated in mol L⁻¹. (c) UV–Vis absorption spectra of aqueous solutions containing of various concentrations of CuCl₂, КNaC₄H₄O₆·4H₂O, and NaOH. The concentration ratio between these three components was kept to be 1:3:4. In legends, the concentrations of sodium potassium tartrate tetrahydrate and copper(II) chloride are indicated in mol L⁻¹.

Export to PPT

The peak centered at 1400 cm⁻¹ is one of the most intense in Raman spectra and is more intense than the near located band centered at 1585 cm⁻¹. Raman spectra with such intensity ratios can be exhibited by vibrations of fully symmetric types. Indeed, if the 1585 cm⁻¹ band is the most intense in FTIR spectrum (Figs. 3a and c) and is referred to asymmetric stretching vibration of carboxylate group (—CO₂—), then the symmetric vibration of the same group should be most intense in Raman spectrum (Nikolskiy, 1982).

This statement is confirmed by quantum-chemical calculations (Table 1). As it was shown before, the formation of the copper(II)–tartrate ion complex is accompanied by appearance of the 1100 cm⁻¹ band, which is absent in solutions free from a salt of copper (Fig. 3). Preliminarily, this band can be interpreted as a valence vibration of the hydroxyl group of tartrate anion attached to the central atom of the complex (1120 cm⁻¹). The red shift of the free ligand band is apparently associated with proton elimination caused by insertion of OH⁻ group to Cu²⁺.

Fig. 5b demonstrates ATR-FTIR spectra of solutions with different concentration of sodium hydroxide. It is shown here that the 1100 cm⁻¹ band is absent when the ratio of [Cu²⁺]/[L]/[OH⁻] equals to 1:3:2, whereas this band appears while [Cu²⁺]/[L]/[OH⁻] is 1:3:4 and further remains unchanged at 1:3:6 and 1:3:8. Thus, sodium hydroxide concentration has a fundamental importance in formation of the tartrate copper complex in an alkaline medium. This observation is supported by the complex structure previously proposed and described as Cu(C₄H₄O₆)₂(OH)₂⁴⁻ (Horner and Klufers, 2016), i.e. two equivalents of alkali is necessary for formation of two Cu—OH bonds and two equivalents of alkali are required to neutralize two protons eliminated by the displacement of hydrogen ions from the hydroxyl groups of the tartrate, which interact with copper. Summarizing the results of the spectral studies, it can be argued that in the system CuCl₂–КNaC₄H₄O₆·4H₂O–NaOH–H₂O the complexation of copper with tartrate anion occurs through a hydroxyl group. The form of such tartrate copper complex with an excess of sodium hydroxide is shown in Fig. 6.

Close

Fig. 6

The structure of the tartrate copper complex in alkaline environment optimized using Minnesota density functional M11 and basis set 6-31g(d) for C, H, O along with Lanl2DZ pseudopotentials and basis set for Cu, K and Na.

Export to PPT

The charge transfer occurred during the LCLD reaction can be described using thermodynamic criteria of copper cations interaction with functional groups of ligands (OH⁻, H₂O, L). The affinity of copper(II) cation to electron is among the most important once. This parameter is characterized by the following values: the second ionization potential of copper atom is 20.29 eV, the first ionization potential is 7.72 eV (Nikolskiy, 1982). Fig. 7 demonstrates the diagram representing the ionization potentials of different ligands in comparison with electron affinity of copper(II) cation as the approximate characteristic of the electron donor ability of these ligands and electron acceptor ability of copper(II) cation. Here, the higher the ionization potential of the ligand the more difficult the reduction of the copper atom proceeds. As it is shown in Fig. 7, in order to reduce copper ion to metallic copper in solution the ionization potential must be lower than 7.72 eV. Thus, free OH⁻ group is the most favorable in terms of energy for formation of copper hydroxide.

Close

Fig. 7

The ionization potentials of different ligands in comparison with electron affinity of copper(II) cation as the approximate characteristic of the electron donor ability of these ligands and electron acceptor ability of copper(II) cation (Champion, 2003; Nikolskiy, 1982).

Export to PPT

Moreover, the bond formation via one hydroxyl group could not provide the reduction of copper to metallic form, however, if this process is chelate in nature (i.e. two OH⁻ groups involved) then the energy level is enough for copper ion reduction.

On the other hand, in an aqueous alkaline solution the carboxylate ion is hydrated, as a result, its complexation with copper cation is unfavorable. In turn, in acidic medium its ionization potential is also insufficient for the copper reduction. Amino groups demonstrate similar behavior. Thus, the comparison of the ionization potentials shows that the necessary requirement for the copper reduction in LCLD solutions is formation of chelate complexes with ligand via alcohol groups followed by insertion of free OH⁻ groups, which provide energetically favorable process of electron transfer from oxidizing to reducing agents due to the smallest ionization potential in the system under consideration. All the mentioned above explains the specific properties and role of the components in the process of complex formation between tartrate anion and copper cation occurred in solution used for the in situ laser-induced synthesis of the conductive copper microstructures.

As it was figured out before, with the increase of the component concentrations in the system CuCl₂–КNaC₄H₄O₆·4H₂O–NaOH–H₂O, while maintaining the ratio of [Cu²⁺]/[L] equals to 1–3 and physical parameters of the deposition, there is no possibility to obtain the conductive copper deposits at concentrations of Cu²⁺ more than 0.06 M. In order to explain this phenomenon we prepared a series of solutions with the ratio of [Cu²⁺]/[L]/[OH⁻] equals to 1:3:4, where concentrations of CuCl₂, КNaC₄H₄O₆·4H₂O and NaOH were varied from 0.04 to 0.4 M, from 0.12 to 1.2 M, and from 0.16 to 1.6 M, respectively. For comparison, the ATR-FTIR spectra of aqueous solutions of КNaC₄H₄O₆·4H₂O (0.12–1.2 M) and NaOH were used. In ATR-FTIR spectra recorded in solutions of 0.01–0.4 M CuCl₂ a shift towards low frequency region of the absorption band corresponding to asymmetric stretching vibration of carboxylate group (—CO₂—) was observed (Fig. 8). In the saturated solution this shift reaches the size of 23 cm⁻¹, whereas in solution containing free ligand the shift is 5 cm⁻¹. This allows to suggest the possibility of formation of the coordination bonds of tartrate anion with copper ion via hydroxyl groups. The proposed assumption is confirmed by the ATR-FTIR spectra presented as dependencies of optical density at the maximum of the absorption band corresponding to asymmetric stretching vibration (1585 cm⁻¹) of carboxylate group (—CO₂—) vs. concentration of solutions with and without Cu²⁺ (Fig. 9).

Close

Fig. 8

(a) ATR-FTIR spectra of aqueous solutions containing КNaC₄H₄O₆·4H₂O, CuCl₂, and NaOH of different concentrations. The concentration ratio between these three components was kept to be 1:3:4. (b) ATR-FTIR spectra of aqueous solutions containing КNaC₄H₄O₆·4H₂O and NaOH of different concentrations together with 0.04 M CuCl₂. In legends, all concentrations are indicated in mol L⁻¹.

Export to PPT

Close

Fig. 9

The dependencies of intensity of the absorption band centered at 1585 cm⁻¹ on concentration of alkaline solutions containing КNaC₄H₄O₆·4H₂O with (a) and without (b) CuCl₂. The concentration ratio between copper(II) chloride, sodium potassium tartrate tetrahydrate, and sodium hydrate was kept to be 1:3:4. (c) The dependence of optical density at the maximum of the absorption band centered at 1585 cm⁻¹ on concentration of the aforementioned components. The concentration range, in which the conductive copper micro-sized deposits are produced upon laser irradiation, is highlighted in red.

Export to PPT

The Figs. 9a and b show that these dependencies do obey Lambert-Beer law for both types of solutions, however, in solutions containing copper(II) cations the linear regime is smaller (Fig. 9a) and coincides with those corresponding to the range of concentrations suitable for the LCLD experiments.

Furthermore, the increase of concentrations more than 0.2 M leads to deviation from Lambert-Beer law and inability to form the conductive copper deposits. Hence, this deviation at concentrations between 0.025 and 1.2 M is connected with enhancement of intermolecular interactions in the copper complex. The increase in viscosity of solution containing Cu²⁺ results in its transformation into a gel at high concentrations; as a result, this fact may explain the polymerization process, which occurs in the tartaric acid solutions. In the solutions close to neutral pH values, it is possible to isolate and characterize these structures of hydrated polymers using X-ray spectroscopy (Prout et al., 1971); however, in alkaline solutions similar studies are difficult. Isothermal evaporation method was used to obtain the crystalline forms of the copper tartrate complex in alkaline solutions. The obtained crystals were identified as sodium chloride and unidentifiable oily mass indicating the presence of the polymerization process. Moreover, the 1585 cm⁻¹ band observed in diluted solutions is shifted to 1565 cm⁻¹ in concentrated solutions. Thus, such a low-frequency shift indicates on polymerization via the carbonyl group of the copper complex.

In addition, the results of IR and Raman studies were supported by UV–Vis absorption spectroscopy. Fig. 5c illustrates the electronic absorption spectra of aqueous solutions of the copper tartrate complex (CuCl₂–КNaC₄H₄O₆·4H₂O–NaOH–H₂O). The maximum absorption band is centered at 676 nm (ε = 50 mol⁻¹ L cm⁻¹) in diluted solutions, which is consistent with literature (Liu et al., 2015). This band is attributed to d-d transition (²E_q → ²T_2q) of the copper tartrate complex. The increase of Cu²⁺ concentration leads to the shift of the maximum of absorption band from 676 to 682 nm and the decrease in extinction coefficient from 50 to 41.7 for solutions containing Cu²⁺ of concentrations varied from 0.04 to 0.4 M, respectively. This shift and decrease in extinction coefficient is due to weakening of the bonds in the coordination polyhedron of copper, while maintaining its symmetry, and due to the distorted octahedral geometry. Here, it should be noted that the bathochromic shift caused by the increase of concentration of Cu²⁺ accompanied by the hypochromic effect does not contradict the results of ATR-FTIR and Raman spectroscopy. Thus, the results of UV–Vis measurements also support the idea that the copper coordination through the carboxylate groups with increase of the component concentrations leads to the polymerization of the tartrate complex.

It is known that the solution components used for LCLD exhibit partially overlapping functions (Kochemirovsky et al., 2012, 2014, 2015). For example, sodium potassium tartrate tetrahydrate demonstrates reduction activity in the method of laser-induced metal deposition. At the same time, many reducing agents (Kochemirovsky et al., 2014), in particular polyols, are able to coordinate copper atoms in solution (Shafeev, 1993). As a result, possible mechanisms of formation of the conductive copper deposits with taking into consideration both coordinating and reduction abilities cannot be uniquely interpreted based on only literature data. Therefore, in order to study the reduction ability we chose several ligands, which have different coordinating functional groups (Table 2).

As one can see in Table 2, the implementation of oxygen-containing ligands leads to formation of the continuous, monolithic and, in some cases, conductive copper deposits. Laser-induced deposition of copper from solutions containing ligands coordinating via nitrogen proceeds poorly since LCLD reaction periodically and spontaneous is interrupted, and the synthesized deposits display non-monolithic structure. The results of atomic absorption spectrometric studies showed that the amount of copper in the deposits, which were obtained from solutions of oxygen-containing copper complexes, is significantly larger than that of copper deposited from solutions containing ligands coordinating via nitrogen; moreover, these deposits exhibit no electrical conductivity. In contrast, it is possible to distinguish the group of ligands among oxygen-containing ones, which results in deposition of the most monolithic copper structures with good electrical conductivity properties. This group includes sorbitol, erythritol, sodium potassium tartrate and glycerol. It should be also noted that these ligands are coordinated with copper via hydroxyl group in alkaline medium (Kochemirovsky et al., 2012, 2014, 2015). In turn, remaining ligands coordinate using other functional groups such as carboxylate and nitrogenous. It was noticed that last-step stability constant plays a major role rather than total instability constant when copper reduction from organic complexes occurs (Kochemirovsky et al., 2015). Indeed, for example, in the case of salicylic acid (K_instab = 7.94 × 10¹⁶), which has low stability constant for the last step (k = 5.0 × 10⁻⁷), the formation of wide metal structures is observed (line width of 200–400 nm, depending on laser power output). Apparently, the oxidative decarboxylation reaction occurs during complexation of copper with ligand, which is bound with copper via oxygen. Similar processes can be observed at high temperature. Perhaps, carboxylic acids can be also formed during oxidation of polyols upon intense laser irradiation (Kochemirovsky et al., 2015). These acids may further undergo decarboxylation upon influence of temperature within the adjacent temperature zones outside of the laser beam focus (Shafeev, 1993). This leads to formation of wide copper structures (2–4 times wider than the focal spot size). It should be noted that these processes cannot be easily studied by chromatographic and spectroscopic methods since the amount of the generated organic products is extremely low in comparison with high concentrations of the starting materials. Therefore, in this work we use indirect approaches in order to describe the mechanisms of the reactions involved in the LCLD process. Nevertheless, since the coordination of copper via hydroxyl group favors the processes of dehydration of alcohols on metal catalysts (Kochemirovsky et al., 2015), it is possible to assume that the side products of this process such as hydrogen and fragments of organic molecules can be involved in the autocatalytic process of the reduction of copper oxides to its metallic state.

Summarizing the discussed above results the following important observations can be made. The comparison of the dependencies illustrated in Fig. 4 and taking into account the amphoteric character of copper hydroxide allows to suggest that at pH 7–13 the form of copper complex – Cu(C₄H₄O₆)₂(OH)₂⁴⁻ is dominant. In turn, probably, only this form of the copper complex is responsible for the laser-induced synthesis of metallic copper (Kochemirovsky et al., 2015) explaining the necessity of binding of copper atom with tartrate via hydroxyl groups in order to produce the conductive copper deposits. Indirect confirmation of this conclusion is present by a series of works (Kochemirovsky et al., 2015; Kordas et al., 2001), in which it is shown that better topology of copper deposits is provided by ligands and reducing agents coordinating with copper via OH⁻ groups. This consistent with the results obtained in the current work according to which we found that the stability of the tartrate copper complex in an alkaline medium is referred to hydroxyl groups that provide the formation of stable chelate cycles at a certain ratio of Cu²⁺ and OH⁻. In opposite, the copper coordination through carboxyl groups at lower pH values ends up in weakening of the complex or formation of the low-soluble compounds. On the other hand, the negatively charged tartrate complex in alkaline solutions is able to polymerize via carboxylate groups at concentrations of copper(II) chloride more than 0.06 M leading to collapse the LCLD reaction.

Another rather important aspect, which is not discussed in detail here, deserves a particular attention. In our previous studies it was observed that the copper micro-sized deposits with highly developed surface area and good electrical conductivity properties produced by LCLD also exhibit rather good sensory activity towards hydrogen peroxide, which is known to be one of the most important disease markers in human blood (Panov et al., 2016a, 2016b). In this regard, the electrochemical studies were carried out, in which the synthesized copper deposits (lines) were used as indicator electrodes in a standard three-electrode electrochemical cell (Panov et al., 2016a, 2016b). Fig. 10 illustrates the cyclic voltammograms (CVs) of the copper microstructures deposited from the plating solution, in which the components should exist in the form of the proposed above complex (Cu(C₄H₄O₆)₂(OH)₂⁴⁻), in the presence and the absence of the additional polyol reducing agents. The area of CV characterizes the charge flowing through the electrode at given concentration of the studied analyte. In turn, this parameter is responsible for sensitivity of the electrode towards the analyzed substrate. As can be seen from Fig. 10, the in situ laser-induced deposition of copper from solution containing sodium potassium tartrate tetrahydrate with and without sorbitol results in fabrication of the electrodes with much higher (according to the CV area) analytical response in comparison to a pure bulk copper with similar geometric parameters (Panov et al., 2016a, 2016b, 2017), and the electrodes synthesized from solutions containing the reducing agents with less number of hydroxyl groups (glycerol and erythritol).

Close

Fig. 10

The cyclic voltammograms (CVs) of the copper microstructures deposited from the plating copper solution containing sodium potassium tartrate tetrahydrate and the following reducing agents: sorbitol (1), no reducing agent (2), glycerol (3), erythritol (4), and the cyclic voltammogram of a pure bulk copper (5). All CVs were recorded in the background solution of 0.1 M Na₂SO₄. Inset shows linear dependencies of the measured Faraday current of 1–5 on different concentrations of hydrogen peroxide.

Export to PPT

In addition, the inset in Fig. 10 presents the linear dependencies of the measured Faraday current of 1–5 on different concentrations of hydrogen peroxide. These curves allow to obtain another important characteristic of any electrochemical sensor – limit of detection (LOD), which can be calculated using the following equation: LOD = 3S/b, where S is the standard deviation of the signal and b is the slope of the calibration curve. As a result, it is obvious that the lower limits of detection towards hydrogen peroxide are also displayed by the copper electrodes produced from solutions containing polyols with higher number of hydroxyl groups.

4 Conclusion

In this paper we demonstrate how the ligand nature and it's coordination mode with copper ions present in the plating solution affect the in situ laser-induced synthesis of the sensor-active micro-sized copper structures. It was observed that the deposition of copper with good electrical conductivity and electrochemical properties occurs only from solutions containing OH-coordinating ligands and reducing agents (polyols); in turn, the higher the number of the functional hydroxyl groups in their structures, the higher the developed surface area and better sensory characteristics of the resulting copper deposits. In addition, ATR-FTIR, UV–Vis, and Raman spectroscopic studies revealed that, in order to be successful, the LCLD synthesis of the sensor-active copper materials should be performed from the plating solutions at pH 7–13, in which the concentration of copper(II) chloride lies within 0.01–0.05 M and the ratio of the components ([Cu²⁺]/[L]/[OH⁻]) is kept as 1:3:4. At such conditions the components of the plating solution form the tartrate copper complex, which exists in the form of Cu(C₄H₄O₆)₂(OH)₂⁴⁻, where copper ion is coordinated by four hydroxyl groups of the ligand and two hydroxyl groups of the environment. The proposed structure of the tartrate copper complex was also predicted by quantum chemical calculations performed at DFT level of theory. Thus, in the current work we obtained rather important and useful information for deeper understanding the LCLD reaction mechanisms, which has a great interest not only for theoretical purposes but also for practical applications dealing with fabrication of the small-sized sensory platforms suitable for development of new enzymeless sensors and biosensors with better characteristics.