Diego P.
Rocha
ab,
Christopher W.
Foster
a,
Rodrigo A. A.
Munoz
b,
Gary A.
Buller
a,
Edmund M.
Keefe
a and
Craig E.
Banks
*a
aFaculty of Science and Engineering, Manchester Metropolitan University, Chester Street, Manchester, M1 5GD, UK. E-mail: C.Banks@mmu.ac.uk; Tel: +44 (0)161 247 119
bInstitute of Chemistry, Federal University of Uberlandia, 38400-902, Uberlandia, Minas Gerais, Brazil
First published on 10th March 2020
Additive manufacturing is a promising technology for the rapid and economical fabrication of portable electroanalytical devices. In this paper we seek to determine how our bespoke additive manufacturing feedstocks act as the basis of an electrochemical sensing platform towards the sensing of manganese(II) via differential pulse cathodic stripping voltammetry (DPCSV), despite the electrode comprising only 25 wt% nanographite and 75 wt% plastic (polylactic acid). The Additive Manufactured electrodes (AM-electrodes) are also critically compared to graphite screen-printed macroelectrodes (SPEs) and both are explored in model and real tap-water samples. Using optimized DPCSV conditions at pH 6.0, the analytical outputs using the AM-electrodes are as follows: limit of detection, 1.6 × 10−9 mol L−1 (0.09 μg L−1); analytical sensitivity, 3.4 μA V μmol−1 L; linear range, 9.1 × 10−9 mol L−1 to 2.7 × 10−6 mol L−1 (R2 = 0.998); and RSD 4.9% (N = 10 for 1 μmol L−1). These results are compared to screen-printed macroelectrodes (SPEs) giving comparable results providing confidence that AM-electrodes can provide the basis for useful electrochemical sensing platforms. The proposed electroanalytical method (both AM-electrodes and SPEs) is shown to be successfully applied for the determination of manganese(II) in tap water samples and in the analysis of a certified material (drinking water). The proposed method is feasible to be applied for in-loco analyses due to the portability of sensing; in addition, the use of AM-printed electrodes is attractive due to their low cost.
Several techniques can be used for the determination of manganese such as atomic absorption spectrometry (AAS), inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS).11 However, although these techniques present high sensitivity for manganese determination, they present drawbacks such as high cost, being highly specialized, need for preconcentration and/or separation and the non-feasibility of carrying out in-loco analyses.3,4,12,13 Alternatively, electroanalytical methodologies, such as stripping voltammetry, can be applied for the determination of trace levels of manganese due to the selective, sensitive, rapid and portable nature of the techniques.14 Anodic stripping voltammetry (ASV) can be applied for the determination of manganese(II) using mercury and bismuth-modified electrodes;6,15 this is based upon manganese deposition at the working electrode surface and subsequent anodic redissolution of the deposited manganese (Mn0). However, this technique presents a disadvantage of requiring a large reduction potential (Edep = −1.7 V vs. SCE) to electrochemically reduce Mn2+ to Mn0; this potential may lead to intermetallic compound formation16 limiting its analytical utility. To overcome this disadvantage, cathodic stripping voltammetry (CSV) with different working electrodes can be used.14,17 The technique is based on first applying a pre-concentration step, where the electrode potential is held sufficiently positive to form insoluble manganese(IV) dioxide on the electrode surface. After a chosen pre-concentration time, the potential is then swept negative, producing a characteristic voltammetric stripping peak arising from the reduction of manganese(IV) dioxide back to manganese(II).18 Due to the underlying electrochemical mechanism and the limited range of metal ions that can undergo such a reversible transformation, the approach of cathodic stripping voltammetry is very selective and suffers from limited interferents.
Electrochemical methods for manganese detection have been reported previously.2,3,6,12–14,18–26 For instance, Rusinek and colleagues2 developed a new sensor based on polymer-coated indium tin oxide for manganese detection in natural water samples. Kang et al.3 utilized a copper-based electrochemical sensor with a palladium electrode to determine manganese in water samples. Banks and collaborators6 compared ASV and CSV for manganese detection in marine sediments and they concluded that CSV coupled with a boron-doped diamond electrode (BDDE) provided better results. Saterlay and colleagues20 developed a sono-cathodic stripping voltammetry method using as a working electrode a BDDE for manganese determination in instant tea reaching a very low detection limit of 10−11 mol L−1 (2 min deposition). However, several procedures using modified electrodes3,18,22 and/or high-cost electrodes, such as palladium,3 platinum,22 and BDDE,6,20,25 present some drawbacks.
Screen-printed electrodes (SPEs) may currently be the most suitable electrochemical sensors for field analysis due to their low cost, high reproducibility and ability to mimic the electroanalytical performance of conventional solid electrodes, being feasible for field analysis in the biomedical, environmental and industrial areas.27,28 Another type of electrode that has gained prominence in recent years is additively manufactured (AM) electrodes. AM-electrodes can be fabricated quickly at very low costs, requiring only a 3D design (designed in specialized software), a 3D printer and polymer material cartridges, being possible to print modified conductive electrodes with other materials (nanotubes, graphene and metal particles) for different analyses.29–34
In this work, we report for the first time, the use of AM-electrodes as the basis of electrochemical sensing platforms for the trace determination of manganese(II) in water samples via differential pulse cathodic stripping voltammetry (DPCSV) and critically compare these to SPEs. Under optimized conditions, the analytical outputs obtained using the SPE are compared with those acquired at AM-printed electrodes, where the latter provided improvements over the former, despite the electrode being composed mostly of plastic (25 wt% nanographite and 75 wt% plastic PLA). Real samples were analysed and the manganese(II) concentrations found were below the detection limit; thus recovery tests were performed and acceptable values between 96 and 105% were achieved. The electrochemical sensors were also verified to determine manganese(II) in a certified water sample. Both electrochemical sensors are shown to be able to determine manganese at trace levels (nmol L−1 region), with a wide linear range, with the additional benefit of low cost.
The AM-electrodes were prepared from the fabricated bespoke feedstocks (filaments) comprising nanographite (NG) and polylactic acid (PLA). This was achieved via the pre-mixing of NG (Sigma Aldrich, UK) and PLA. For this, 3.75 g of mesoporous NG were dispersed within excess xylene by sonication for 10 minutes. Thereafter, the resulting mixture was heated (under reflux) at 160 °C for 3 hours in a silicon oil bath, 11.25 g of PLA were then added to the mixture and maintained for another 3 hours. After this, the resulting mixture was then recrystallized using excess methanol, vacuum filtered and dried overnight to evaporate the remaining xylene. In order to obtain the filament, the resulting NG/PLA material was placed within a MiniCTW twin-screw extruder (Thermo Scientific) and heated at 200 °C with a screw speed of 30 rpm. The resulting filament was AM (3D printed) using a ZMorph® printer (Wroclaw – Poland) with a direct drive extruder at 190 °C. The AM-electrodes were cut using scissors with the following dimensions: 34 mm × 12 mm. Prior to electrochemical measurements, the AM-electrodes were polished with wet sandpaper. When the AM-electrode was used, counter and reference electrodes were, respectively, a platinum wire and Ag/AgCl (3 mol L−1 KCl). Measurements were carried out using a 3D-printed cell with an internal volume of 5 mL (Fig. 1). The design of this cell was based on previously developed templates described in the literature;40 an O-ring defines the electrochemical working area (geometric area: 0.22 cm2). Note that this is the limit (25% conductive material) that can be successfully fabricated and can be reliability printed (additively manufactured).
Next the response of the AM-electrodes towards the sensing of manganese(II) was explored. Using DPCSV, aliquots of manganese(II) were made into a pH 6 BR buffer with the analytical signal monitored as a function of concentration as shown in Fig. 2. A wider linear range between 9.10 × 10−9 mol L−1 and 2.70 × 10−6 mol L−1 using a 350 seconds deposition time was found to be achievable, with an excellent coefficient of determination (R2) of 0.998. A linear dependence of the stripping signal (peak area (Y, μA V)) versus manganese concentration ([Mn2+], μmol L−1) can be described by the equation Y = 0.0014 ± 0.0001 + 3.42 ± 0.07 [Mn2+]. The detection limit (99.7% confidence level) and the RSD (N = 10 for 1 μmol L−1) (Fig. ESI-4†) were found to be 1.6 × 10−9 mol L−1 and 4.9%, respectively. The proposed AM-electrode electroanalytical sensing platform shows excellent analytical features, a wide linear range and a low detection limit with potential for “in the field” manganese(II) determination in real samples at trace levels.
For comparative purposes, manganese(II) determination was also performed using SPEs. Using the same experimental conditions reported above for the AM-electrodes, the analytical performance towards the sensing of manganese(II) was explored with the analytical signal monitored as a function of concentration. The resulting DPCSV curves can be observed in Fig. ESI-5† where a linear concentration range of peak area versus concentration from 9.10 × 10−9 mol L−1 to 1.82 × 10−6 mol L−1 is achievable with an excellent coefficient of determination (R2) of 0.996 observed. The relationship between the peak area (Y, μA V) and manganese concentration ([Mn2+], μmol L−1) can be described by the equation Y = 0.032 ± 0.002 + 2.51 ± 0.09 [Mn2+]. The relative standard deviation (RSD) value was estimated to be 6.74% for ten consecutive measurements using a low concentration of manganese (1 μmol L−1) (Fig. ESI-6†). The detection limit of 2.4 × 10−9 mol L−1 (99.7% confidence level) was calculated as follows: 3sB/S, in which sB and S were the standard deviation for ten consecutive measurements of baseline noise and the slope of the analytical curve, respectively.
The (electro)analytical performances of the AM-electrodes and SPEs are summarised within Table 1. The use of AM-electrodes shows an improvement in the analytical characteristics, with 36% increase in sensitivity obtained. A decrease of 33% in the detection limit is also achieved. In addition, a larger linear range and a smaller RSD were obtained compared to the SPE results. However, if one normalises the sensitivity with the geometric electrode area, the sensitivities are 15.5 and 35.7 μA V μmol−1 L cm−2 for the AM-electrodes and SPEs, respectively. That said, both are completely different electrochemical sensing platforms with the AM-electrodes comprised of 75% plastic! Note that this is the limit that can be successfully fabricated and can be reliability printed (additively manufactured); thus the limit in this case (PLA) is 25% conductive material.
Analytical feature | AM-electrodes | SPEs |
---|---|---|
Linear range (mol L−1) | 9.10 × 10−9 to 2.70 × 10−6 | 9.10 × 10−9 to 1.82 × 10−6 |
R 2 | 0.998 | 0.996 |
Sensitivity (μA V μmol−1 L) | 3.4 | 2.5 |
RSD (N = 10 for 1 μmol L−1) | 4.9% | 6.7% |
Detection limit (mol L−1) | 1.6 × 10−9 | 2.4 × 10−9 |
We next turn to exploring the AM-electrodes and SPEs for the sensing of manganese(II) in three samples of tap water. The water samples were 2-fold diluted in the supporting electrolyte. Using the proposed protocol described above and optimized conditions, similar results were obtained for both sensors. Since manganese(II) signals were not achieved in all samples as the concentration found was below the detection limit, the samples were fortified with a very low manganese(II) concentration (3.6 × 10−7 mol L−1, final concentration in the cell after the sample is 2-fold diluted). Satisfactory recovery values between 96% and 105% were acquired, showing acceptable accuracy of the developed method considering the low manganese concentration adopted for the experiment. Standard addition curves showed good linearity (R2 > 0.99) and the results are shown in Fig. 3. In addition, the accuracy of the proposed protocol was evaluated by the analysis of ALPHA APS 1075 (Trace Metals in Drinking Water) certified material again measurements made via the standard addition methodology. The type and concentration of the trace metals (including manganese) present in this certified material are shown in ESI Table 2.† A statistically significant difference was not observed between the AM-electrodes, 7.10 (±0.25) × 10−4 mol L−1 and SPEs 7.20 (±0.25) × 10−4 mol L−1 comparing well with the certified material value of (7.28 × 10−4 mol L−1) at a confidence level of 95%. As can be seen in ESI Table 2† many other metal ions are present in the ALPHA APS 1075 in very high concentration and despite that, no interference problems were observed in the analysis. Additionally, we must emphasize that due to the high deposition potential of MnO2 and subsequent voltammetric scanning of positive to negative potentials, the interference caused by several metal ions is largely eliminated3 and the use of cathodic stripping voltammetry is highly selective for the manganese determination. Thus, we have demonstrated that the AM-electrodes and SPEs can be successfully applied for the trace determination of manganese(II) in tap water samples.
Electrodea | Deposition time (s) | Linear range (mol L−1) | Detection limit (mol L−1) | Ref. |
---|---|---|---|---|
a TMGE – thick-film graphite-containing electrode modified with formazan; CBSP – copper-based electrochemical sensor with a palladium electrode; SSEBS-ITO – polystyrene-block-poly(ethylene-ran-butylene)-block-polystyrene-sulfonate coated indium tin oxide; G-SA – graphite/styrene-acrylonitrile copolymer composite electrodes composite electrodes; CFE – carbon film electrodes fabricated from carbon resistors of 2 Ω; GE – graphite electrode; RGCE – rotating glassy carbon disk electrode; PTE – micro fabricated platinum thin–film electrode; EPPGE – edge plane pyrolytic graphite electrodes; SpCI – stencil-printed carbon ink electrodes; BDDE – boron-doped diamond electrode; Mn(II)-IIP/MWCNT/Chit/IL/GCE – glassy carbon electrode modified with a multi-walled carbon nanotube-chitosan-ionic liquid nanocomposite decorated with an ion imprinted polymer; GCDE – glassy carbon disk electrode; CPE-carbon paste electrode; SPE – graphite screen-printed electrode; NG/PLA AM-electrode – nano graphite (NG)/PLA. b NM – Not mentioned. | ||||
SSEBS-ITO | 180 | 1.8 × 10−8 to 8 × 10−7 | 1 × 10−9 | 2 |
CBSP | 600 | 4.5 × 10−7 to 10.9 × 10−6 | 3.3 × 10−7 | 3 |
CPE | 120 | 1 × 10−8 to 6 × 10−8 | 3.5 × 10−9 | 6 |
BDDE | 60 | 1.25 × 10−6 to 2.5 × 10−5 | 7.4 × 10−7 | 6 |
TMGE | 60 | 1.8 × 10−9 to 5.5 × 10−7 | 7 × 10−10 | 12 |
G-SA | 180 | 3.6 × 10−9 to 1.8 × 10−7 | 3.6 × 10−9 | 13 |
CFE | 180 | NMb | 4 v−9 | 14 |
RGCE | 900 | NMb | 6 × 10−9 | 18 |
EPPGE | 120 | 2.5 × 10−8 to 2.5 × 10−7 | 1.4 × 10−8 | 19 |
BDDE | 120 | 1 × 10−11 to 3 × 10−7 | 1 × 10−11 | 20 |
GE | 600 | NMb | 5.5 × 10−8 | 21 |
PTE | 900 | 9.1 × 10−8 to 9.1 × 10−7 | 1.6 × 10−8 | 22 |
GCDE | 120 | 1 × 10−9 to 4 × 10−8 | 2.5 × 10−10 | 23 |
Mn(II)-IIP/MWCNT/Chit/IL/GCE | 120 | 2 × 10−6 to 9 × 10−6 | 1.5 × 10−7 | 24 |
BDDE | 120 | 4 × 10−8 to 1.7 × 10−7 | 8.9 × 10−9 | 25 |
SpCI | 180 | 5 × 10−7 to 2.5 × 10−5 | 5 × 10−7 | 26 |
SPE | 350 | 9.1 × 10−9 to 1.8 × 10−6 | 2.4 × 10−9 | This work |
AM-electrodes | 350 | 9.1 × 10−9 to 2.7 × 10−6 | 1.6 × 10−9 | This work |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0an00018c |
This journal is © The Royal Society of Chemistry 2020 |