Carbon nanotubes as a suitable material for electrochemical sensor used in voltammetric determinations of titanium

We report the use of carbon nanotubes as a material for the preparation of an electrochemical sensor that acts as a substrate for film metal electrodes used in stripping voltammetry. The sensor is based on a mixture of multiwall carbon nanotubes, glassy carbon spherical powder, and epoxy resin. The properly selected composition of the sensor made it possible to obtain a new substrate, competitive in relation to glassy carbon, for creating film metal electrodes. In this work, the proposed new sensor was used to determine the trace amounts of Ti(IV) on the lead film electrode. Measurements were carried out with in situ mode in an acetate buffer by adsorption stripping voltammetry with the use of cupferron as a complexing agent. Linear response to Ti(IV) ions in the concentration range of 7 × 10–10–7 × 10–8 mol L−1 and the limit of detection (LODs) 2.4 × 10–10 were obtained, respectively. These promising results revealed that a mixture of carbon nanotubes, epoxy resin, and spherical glassy carbon powder used for the determination of titanium ions on PbFE might represent an important addition to existing electrochemical sensor technologies. The proposed procedure was successfully used as a new and powerful analytical tool for determination of Ti(IV) in horsetail extracts.


Introduction
Stripping voltammetry (SV) is a powerful technique for determination of low levels of metal ions.The key element in voltammetric procedures is the selection of the working electrode.The first procedures were developed using hanging mercury drop electrodes; however, their major drawback was high toxicity (Vyskočil and Barek 2019;Ciglenecki et al. 2018).With the growing awareness of the toxicity of mercury, there is a search for new electrode materials that could replace mercury electrodes with more environmentally friendly and lab-friendly materials.In recent years, the literature has described many different working electrodes which use the formation of metallic films, among others lead film electrodes (PbFE).The formation of these electrodes can take place both ex situ and in situ (Bobrowski et al. 2008;Bobrowski et al. 2018;Deniz et al. 2017;Gęca et al. 2016;Grabarczyk and Adamczyk 2019;Korolczuk et al. 2014;Tyszczuk-Rotko et al. 2014).In the case of in situ, we add the salt of the lead from which the film is to be formed to the sample.After the appropriate potential is applied to the electrode acting as a substrate, a film forms on the electrode as a result of the reduction of metal ions present in the solution and the substance to be determined is concentrated there on.In this case, the film-forming conditions must be similar to the optimal assay conditions, which is not always possible.In the ex situ method, a film metal electrode is prepared from a solution other than the sample solution.This solution contains electrolyte and metal ions from which the metal film will be generated, that is Pb(II).After washing, the electrode is transferred to the sample solution from which the actual voltammetric measurement is carried out.Obviously, the method of creating an in situ film electrode is preferred due to the faster measurements and the saving of reagents.Regardless of the method of film metal electrode creation, a very important element is the electrode acting as the substrate on which the film is created.Such an electrode in the case of PbFE is most often a glassy carbon electrode (Deniz et al. 2017;Gęca et al. 2016;Grabarczyk and Adamczyk 2019;Tyszczuk-Rotko et al. 2014) and less frequently carbon paste electrode (Korolczuk et al. 2014).The aim of our research was to find another material competitive to glassy carbon, acting as a substrate for lead film electrode, allowing to improve the parameters of procedures for the determination of trace amounts of analytes.In our research, we used nanotubes to modify the electrode thanks to which the sensitivity of the determinations was increased compared to the procedure based on PbFE formed on classic glassy carbon electrodes and additionally the range of linearity has been increased.Carbon nanotubes, due to their extraordinary properties (mechanical, strength, electrical), have a wide area of application, and they also turned out to be a very valuable electrode material (Das et al. 2022;Gupta et al. 2019;Maheswaran and Shanmugavel 2022).Also in the case of voltammetric sensors, you can find the use of multiwall carbon nanotubes, such as carbon-nanotube-modified glassy carbon electrode (Baranowska and Bijak 2013;Kumar and Vicente-Beckett 2012), carbon-nanotube-modified carbon paste electrode (Deng et al. 2008) as well as bismuth-modified carbon nanotube electrode (Hwang et al. 2008;Stočes et al. 2012) or copper film with carbon-nanotubes-modified electrode (Wasąg and Grabarczyk 2021).To the best of our knowledge, the modification of carbon nanotube electrode for lead film formation proposed by us is a new and innovative solution, so far not described in the literature.
The carbon-nanotubes-modified electrode used in this work was obtained by mixing multiwall carbon nanotubes (CNTs) and epoxy resin to a homogeneous mass and then, after centrifuging, mixing with spherical glassy carbon (SGC) powder.The resulting paste was put under pressure into a 2 mm diameter hole drilled in epoxy resin.Such composition turns out to be a very good basis for metal film formation.The use of this multiwall carbon nanotubes/spherical glassy carbon electrode (CNTs/SGCE) has the additional advantage: the deoxygenating of the sample before measurement is unnecessary.It also simplifies the apparatus necessary to perform metal ions determination in field conditions.To confirm the applicability of nanotubes to modify the electrode acting as the substrate for PbFE, the stripping voltammetric determination of Ti(IV) ions in nanoconcentrations in natural water samples was performed.The choice of Ti(IV) ions for research was dictated primarily by the fact that the trace analysis of Ti(IV) ions in the environment is a topical issue and efforts are still being made to lower the detection limit of these ions.Titanium is no longer considered as an expensive or hard-to-reach material.The range of applications of this metal and its alloys is very wide, and with the development of technology, it constantly expands.Titanium is used, inter alia, in shipbuilding, space technology, armaments, chemical, aviation, jewelry, biomedical engineering, in the production of products used in extreme sports, including mountaineering equipment or sailing equipment.The most common compound of this element is titanium oxide.It is otherwise referred to as titanium white and is a very popular dye used, among others, in the production of paints, pigments, glass fibers, varnishes, etc.With the help of titanium oxide, an intensely white color is given to high-quality paper products.Titanium white is also added to food products and color cosmetics.It is a common dye in powders, foundations, eye shadows, blushes, and nail polishes.As it inhibits the penetration of solar radiation, it is also used as a filter in sunbathing cosmetics (Liu et al. 2004;Saurabh et al. 2022;Zhang et al. 2011).Due to such a wide use of metallic titanium and its alloys and compounds, it is necessary to constantly monitor the environment in terms of contamination with this element.The literature describes many titanium determination procedures based on various instrumental methods, including voltammetric methods using various working electrodes of which the vast majority use a toxic mercury-based electrodes (Croot 2011;Gawryś and Golimowski 2000;Gawryś and Golimowski 2001;Gawryś and Golimowski 2003;Gawryś et. al. 2000;Grabarczyk and Wasąg 2015;Pereira and Pereira 1999;Romanus et al. 1991;Vukomanovic and Loon 1994;Zhao et al. 1995).The mercury-free electrodes used for Ti(IV) determination are modified paste electrode, bismuth and lead film electrodes generated on the glassy carbon created by ex situ and in situ methods, respectively (Bobrowski et al. 2009;Cui et al. 2013;Stadlober et al. 1996;Wasąg and Grabarczyk 2016).As proved in this study, the use of carbon nanotubes as a base material for lead film electrode allowed for further improvement and increased sensitivity of Ti(IV) determinations with the use of a mercury-free eco-friendly electrode.

Apparatus
All voltammetric experiments were carried out with an Autolab PGSTAT 10 analyzer (Utrecht, The Netherlands) employing multiwall carbon nanotubes/spherical glassy carbon electrode (CNTs/SGCE) as a working electrode, a Pt auxiliary electrode and an Ag/AgCl reference electrode.Mineralization of horsetail samples was performed using a MARS 5 CEM microwave mineralizer.The pH measurements were made on a CI-316 pH meter (Elmetron).The morphology of the CNTs/SGCE was characterized using an inverted metallurgical microscope Nikon ECLIPSE MA200 (Japan).

Reagents
Multiwall carbon nanotubes, O.D. × I.D. × L 10 nm ± 1 nm × 4.5 nm ± 0.5 nm × 3~6 μm (Sigma-Aldrich).Spherical glassy carbon powder size 0.4-12 µm (HTW Hochtemperatur-Werkstoffe GmbH).Acetate buffers were prepared from acetic acid and sodium hydroxide Suprapure Merck.A standard solution of 1 g L −1 of Pb(II), cupferron, solution of 1 g L −1 of Ti(IV) were purchased from Fluka.Horsetail products were purchased from a supermarket of such Polish supplier as "Flos," and "Herbapol" (plant materials were stored in paper bags in a dry and shady place at ambient temperature until use).All solutions were made using triply distilled water.

Preparation of CNTs/SGC electrode
The preparation of the working electrode involved first mixing a multiwall carbon nanotubes with epoxy resin (in a ratio of 1:25) to a homogeneous mass.Then to remove air bubbles, the resulting mass was heated to 115 °C and hot centrifuged.The prepared mass was then mixed (in a 2:1 ratio) with spherical glassy carbon powder (size 0.4-12 µm).The resulting paste was put under pressure into a 2 mm diameter hole drilled in epoxy resin.Copper wire was utilized to secure an electric contact.The electrode prepared in this way was polished on sandpaper with coarse grit (P120) and then with fine grit (P2000).After polishing, the electrode was rinsed with large amounts of triply distilled water and kept in an ultrasonic bath (Sonic-3, Polsonic, Poland) for 30 s to remove any residual polishing material.The electrode prepared in this way was ready for measurements and ensured obtaining stable and reproducible results over a long period of time.The research was carried out over the period of 1 year and during this time, no changes in the obtained signals were observed.On each measurement day, the electrode was polished using 0.3 µm alumina slurry on a Buehler polishing pad and immersed for 30 s in an ultrasonic bath.

Measurement procedures
Initial measurements have been carried out by AdSV mode using ex situ and in situ procedures as described below.
Ex situ procedure It is as follows: first film formation and then determining the metal ion.
Film formation The electrochemical deposition of lead at the CNTs/SGC electrode was performed in an electrochemical cell with 10 mL of 0.2 mol L −1 acetate buffer (pH = 3.5), 1 × 10 -4 mol L −1 Pb(II) applying potential − 1.7 V for 50 s, stirring solution.After this pre-concentration step, the system was aborted.
Metal determination The measurement was performed in an electrochemical cell with 10 mL of 0.2 mol L −1 acetate buffer (pH = 6.0), 4 × 10 -4 mol L −1 cupferron with electrodeposition at − 0.8 V of 50 s and stirring solution.Following the deposition step, the stirring was stopped and after 5 s, the voltammogram was recorded by applying a negative-going potential scan from − 0.6 V to − 0.9 V.
In situ procedure It is as follows: film formation and determination of the metal were performed simultaneous.The measurement was taken in an electrochemical cell with 10 mL of 0.2 mol L −1 acetate buffer (pH = 6.0), 1 × 10 -4 mol L −1 Pb(II), 4 × 10 -4 mol L −1 cupferron with electrodeposition at successive potentials − 1.7 V for 50 s and − 0.8 V of 50 s with mixing the solution.Following the deposition step, the stirring was stopped and after 5 s, the voltammogram was recorded by applying a negative-going potential scan from − 0.6 V to − 0.9 V.
For repetitive measurements, the electrode was cleaned both in method 1 and 2 in the same way i.e., by applying potential 0.2 V for 20 s under stirring conditions to remove the residual metals.All experiments were done at room temperature.
As the obtained results of the determination of Ti(IV) were comparable, the in situ method was selected as a standard procedure as faster and more convenient due to the fact that the entire measurement was carried out in one measuring cell.

Characterization of electrode surface
Real images of the solid CNTs/SGC electrode were taken using an inverted metallurgical microscope Nikon.Figure 1 is the surface image of the electrode, where white particles with irregular shape and size are spherical glassy carbon powders with 0.4-12 µm particle size and the gray surface is nanotubes mixed with epoxy resin.

Repeatability and reproducibility of the CNTs/SGC electrode
Series of voltammetric measurements were also performed to establish the reliability of the CNTs/SGC electrode.The reliability was evaluated by examining the repeatability and reproducibility of the CNTs/SGC electrode by performing voltammetric measurement for a 1 × 10 -8 mol L −1 Ti(IV) under the optimized parameters described above, that is of 0.2 mol L −1 acetate buffer (pH = 6.0), 1 × 10 -4 mol L −1 Pb(II), 4 × 10 -4 mol L −1 cupferron with electrodeposition at successive potentials − 1.7 V for 50 s and − 0.8 V of 50 s, respectively.Thus, to examine the repeatability of sensor, six voltammetric measurements one after the other were carried out; in so doing, the relative standard deviation (% RSD) was found to be 3.3% The reproducibility was evaluated from the measurements performed in six subsequent days as RSD and was 3.7%.The obtained results clearly justify the good repeatability and reproducibility of the produced electrode.

Electrolyte
As is known, for the voltammetric measurement, the sample to be analyzed must be supported by the presence of a suitable electrolyte, most often they are buffer, acids or bases.In the case of the determination of Ti(IV) ions by the AdSV method with the use of cupferron as a complexing agent, as reported in the literature, the most appropriate medium is the acetate buffer (Baoxian and Shuxun 1994;Croot 2011; Wasąg and Grabarczyk 2016).When examining the effect of the pH of the acetate buffer on the height of the titanium peak, it was found that, also in our procedure, the experiments showed that the highest peak current was obtained in acetate buffer pH = 6.0 (Fig. 2).The optimum concentration was in the range of 0.05-0.2mol L −1 , so 0.2 mol L −1 was selected for further experiments.
Another necessary addition to the analyzed sample in the AdSV method is an appropriate complexing agent.In the case of adsorptive accumulation of complexes with Ti(IV) ions, for example cupferron, chloranilic acid or oxalate can be used as a complexing agent (Baoxian and Shuxun 1994;Cui et al. 2013;Croot 2011;Grabarczyk and Wasąg 2015;Stadlober et al. 1996).However, in our procedure, cupferron proved to be the most suitable, which is one of the most frequently used complexing agents in voltammetric procedures based on the AdSV method (Adamczyk and Grabarczyk 2020;Baoxian and Shuxun 1994;Cui et al. 2013;Croot 2011).Conducting experiments aimed at selecting the appropriate concentration of cupferron, it was observed that with increasing concentration of cupferron to 4 × 10 -4 mol L −1 , the analytical signal of titanium increases and then the signal remains constant (Fig. 3).Cupferron concentration equal to 4 × 10 -4 mol L −1 was used as the most optimal value in all measurements.

Lead concentration
The concentration of Pb(II) used for formation of the lead film is known to influence the height of the stripping peak of  3 The influence of cupferron concentration on current response of 7 × 10 -9 mol L −1 Ti(IV) the target metal.In the absence of lead ions, the CNTs/SGC electrode exhibited no titanium signal.So the concentration of Pb(II) in the solution was changed in the concentration range from 1 × 10 -6 mol L −1 to 2 × 10 -4 mol L −1 in the environment of 0.2 mol L −1 of the acetate buffer at pH = 6.0.As lead ions content was increased up to 1 × 10 -4 mol L −1 , the stripping titanium responses increased due to the increase of active sites resulting from the metals film formation on the CNTs/SGC electrode.The experiments were carried out for the concentration of Ti(IV) equal to 1 × 10 -8 mol L −1 .The heights of the titanium peaks depending on the lead concentration recorded from the respective solutions are presented graphically in Fig. 4.

Accumulation potential and time
When analyzing the conditions of lead film formation in many procedures described in the literature, it was observed that the potential applied to the electrode being the substrate for the film assumed a different value from − 0.65 V in the negative direction up to − 1.7 V (Bobrowski et al. 2008;Bobrowski et al. 2018;Deniz et al. 2017;Gęca et al. 2016;Grabarczyk and Adamczyk 2019;Korolczuk et al. 2014;Tyszczuk-Rotko et al. 2014).As for the adsorption of Ti (IV) complexes on the surface of the working electrode in various procedures described in the literature, the potential also changed in a wide range depending on the working electrode and the complex (Bobrowski et al. 2009;Cui et al. 2013;Croot 2011;Gawryś and Golimowski 2000;Gawryś and Golimowski 2001;Gawryś and Golimowski 2003;Gawryś et al. 2000;Grabarczyk and Wasąg 2015;Pereira and Pereira 1999;Romanus et al.1991;Stadlober et al. 1996;Vukomanovic and Loon 1994;Wasąg and Grabarczyk 2016;Zhao et al. 1995).Therefore, we proposed to conduct an in situ measurement in which it would be carried out at two successive potentials, taking into account both the optimal conditions for film formation and titanium accumulation.The first potential was varied from − 2.1 V to − 1.1 V for a 50 s accumulation time, while the second potential was varied from − 1.1 V to − 0.7 V for a 50 s accumulation time.The obtained results are presented in Fig. 5 in graphical form as the dependence of the titanium peak height on the changing values of the accumulation potential.Analyzing the obtained results and selecting the most optimal potentials of lead film formation on the CNTs/SGC electrode and titanium accumulation, the following values were selected: − 1.7 V for 50 s and − 0.8 V for 50 s.The accumulation times were also selected, changing it in the range from 10 to 80 s with a frequency of 10 s and the obtained results are shown in Fig. 6.As can be seen in the case of first accumulation potential − 1.7 V, the titanium peak increased with the accumulation time to 50 s and then did not increase significantly.In the case of second accumulation potential − 0.8 V, the titanium peak increased with accumulation time to 50 s and then slowly decreases.

Calibration and detection limit
To evaluate the analytical parameters, we conducted measurements with successive additions of titanium ions to standard solutions (0.2 mol L −1 acetate buffer (pH = 6.0), 1 × 10 -4 mol L −1 Pb(II), 4 × 10 -4 mol L −1 cupferron) with increments of its concentration in connection with − 1.7 V for 50 s and − 0.8 V of 50 s accumulation step.Lead film CNTs/SGC electrode revealed a linear response for titanium in examined concentration ranges from 7 × 10 -10 to 7 × 10 -8 mol L −1 .The calibration plot obeyed the equation y = 0.228x + 0.084, where y is the peak current (µA) and x is the Ti(VI) concentration (nmol L −1 ).The linear correlation coefficient was r = 0.994.For the developed procedure, the detection limit was found to be 2.4 × 10 -10 mol L −1 estimated from three times the standard deviation for the lowest studied Ti(VI) concentration.Comparing the obtained parameters to the parameters previously published for Ti(VI) determination using a lead film electrode generated on a classical glassy carbon electrode, a significant improvement in efficiency was obtained as the detection limit is lower by almost an order of magnitude and the linearity range is wider (Wasąg and Grabarczyk 2016).

Sample preparation
Horsetail products before high-pressure mineralization were ground using a pestle and mortar.For acid digestion, horsetails (0.25 g) were weighed into 100 mL Teflon HP-500 Plus vessels.Then 10 mL of concentrated nitric acid (60%) was added to each vessel, the vessel was covered and then mineralized at 130 °C for 30 min.The obtained extracts were cooled to ambient temperature before being transferred quantitatively into clean 100 mL volumetric flasks.The samples were then diluted to volume by the addition of ultrapure water.Three replicate digestions were made for each horsetail products.

Determination of Ti(IV) in horsetail extracts using lead film CNTs/SGC electrode
In the horsetail extracts, titanium was determined by proposed AdSV procedure using lead film CNTs/SGC electrode.
The extracts of samples were analyzed using the standard addition method.Because the prepared extracts contained nitric acid, an appropriate quantity of sodium hydroxide was added directly to the solution in the voltammetric cell to obtain pH = 6.0.The results for each sample are the mean of the four replicates and are quoted on a dry weight basis and are presented in Table 1.The relative standard deviation (RSD) was calculated in the range from 1.7% to 3.3% for the Herbapol and Flos horsetail extracts, confirming the acceptable reproducibility of the results of Ti(IV) determination  6 The influence of accumulation time on 7 × 10 -9 mol L −1 Ti(IV) peak current: (a) first accumulation time was changed at -1.7 V, second accumulation time was constant at 50 s at − 0.8 V; (b) first accumulation time was constant at 50 s at − 1.7 V, second accumulation time was changed at -0.8 V Table 1 The Ti(IV) content in extracts of horsetail from Herbapol and Flos analyzed using the standard addition method The results for each sample are the mean of the four replicates and are quoted on a dry weight of the horsetail

Number of sample
The Ti(IV) content in the dry matter of horsetail [µg g −1 ] The average Ti(IV) content in the dry matter of horsetail [µg g using the lead film CNTs/SGC electrode.The obtained results confirmed literature reports that horsetail is rich in titanium (Emsley 2001) and the titanium content in studied samples was between 40.4 ± 2.4 and 53.8 ± 2.2 μg g −1 in dry horsetail.Figure 7 presents exemplary voltamperograms recorded during the determination of titanium in the horsetail extract of the Flos.

Conclusion
This paper describes for the first time a new eco-friendly electrode based on a mixture of multiwall carbon nanotubes, epoxy resin, and glassy carbon spherical powder, which is a substrate competitive to glassy carbon electrode for the formation of metallic film electrodes.To confirm this, the electrode was successfully used as a substrate for the formation of a lead film to increase the sensitivity of Ti(IV) in the AdSV determination.As it has been proven, the detection limit has been reduced by almost an order of magnitude, and at the same time, the range of quantification has been increased in relation to the procedure for determining Ti(IV) on a film lead electrode generated on glass carbon.The novel lead film CNTs/SGC electrode method developed here enabled the simple and rapid detection of Ti(IV) with good stability, RSDs for repeatability and reproducibility were all < 4%.Lead film CNTs/SGC electrode was used as a new and powerful analytical tool for determination of Ti(IV) in horsetail extracts with RSD range 1.7% and 3.3%.Additional analysis of extracts confirmed that horsetail are distinguished by a high concentration of titanium compared to most plants which contain about 1 ppm of titanium, while in the case of horsetail, the concentration of this element is several dozen times higher.The obtained results confirm that titanium tends to accumulate in plants containing high amounts of silica, which include horsetail.

Fig. 1
Fig. 1 A real image of the solid CNTs/SGC electrode taken using an inverted metallurgical microscope Nikon

Fig. 4 Fig. 5
Fig.4The influence of Pb(II) concentration on current response of 7 × 10 -9 mol L −1 Ti(IV) Fig.6The influence of accumulation time on 7 × 10 -9 mol L −1 Ti(IV) peak current: (a) first accumulation time was changed at -1.7 V, second accumulation time was constant at 50 s at − 0.8 V; (b) first accumulation time was constant at 50 s at − 1.7 V, second accumulation time was changed at -0.8 V