Versatile DIY Route for Incorporation of a Wide Range of Electrode Materials into Rotating Ring Disk Electrodes

Rotating ring disk electrodes (RRDEs) are a powerful and versatile tool for mechanistically investigating electrochemical reactions at electrode surfaces, particularly in the area of electroanalysis and catalysis. Despite their importance, only limited electrode materials (typically glassy carbon, platinum, and gold) and combinations thereof are available commercially. In this work, we present a method employing three-dimensional (3D) printing in conjunction with machined brass components to produce housing, which can accommodate any electrode material in, e.g., pressed powdered pellet, wafer, rod, foil, or vapor deposited onto a conductive substrate form. In this way, the range and usability of RRDEs is extended. This custom do-it-yourself (DIY) approach to fabricating RRDEs also enables RRDEs to be produced at a significant fraction of the cost of commercial RRDEs. To illustrate the versatility of our approach, coplanar boron-doped diamond (BDD) RRDEs are fabricated for the first time using the approach described. Experimental collection efficiencies for the redox couple FcTMA+/FcTMA2+ are found to be very close to those predicted theoretically. BDD electrodes serve as an ideal electrocatalyst support due to their low background currents, wide solvent potential window in aqueous solution, and chemical and electrochemical stability in acid and alkali solutions. The BDD RRDE configuration is employed to investigate the importance of surface-incorporated nondiamond carbon in BDD on hydrogen peroxide generation via the oxygen reduction reaction in acid solutions.


Supporting Information
S-3 size a male thread, but it can also be done with a female. If the thread dimensions are not known, follow the below procedure.
Metric: Place the thread up against a ruler and measure the distance from one thread to the next, this is the thread pitch i.e. how far up or down the thread moves with each full rotation.
Next, measure the diameter of the thread and put an M in front of it. From here you can create a thread classification that should be of the format diameter followed by thread pitch (i.e. M4 × 0.5). If this thread corresponds to a standard combination the easiest way to check is to screw the opposite gender of thread into it. A mechanical workshop will likely have pitch gauges which will allow them to do this more accurately, although this method works well in the absence of more specialized equipment. A schematic representation of this can be found in  Again, if this thread corresponds to a standard combination the easiest way to check is to screw the opposite gender of thread into it. A schematic representation of this can be found in Figure   S1.2.

Supporting Information
S-4 Figure S1.2: Schematic showing how to size imperial threads without a pitch gauge.

Assembly:
This section describes the assembly process for a RRDE electrode consisting of two BDD electrodes which are 360 µm thick and polished on the top surface to ~ nm roughness. Advice on how to modify the assembly method for other materials is detailed later in this section.

1.
Collect all the prepared parts for the RRDE including (i) the insulating outer case (which prevents the brass being exposed to the solution and lies co-planar with the ring and the disc), (ii) the brass outer core (which serves as both the ring contact and main body), (iii) the brass inner core, (the disk contact), (iv) optionally, a stand to assist with assembly, (v) the insulating spacer (which prevents the two brass parts from touching), (vi) the insulating tube (which separates and spaces the brass inner and outer), (vii) the ring and (iix) disk electrodes and (ix) the insulating resin, rapid epoxy, and conductive epoxy. This step can be seen in Figure S1.3a.

2.
If using 3D printed parts remove from the supports and remove any defects from printing with a knife or file. This step can be seen in Figure S1.3b.

Supporting Information
S-5 3. Dry fit all the parts together to ensure they fit before gluing. Ensure that the brass inner fits flush with the brass outer. If it doesn't, sand the back of the insulating inner until they do. This step can be seen in Figure S1.3c. ii) brass outer core, iii) brass inner core, iv) RRDE assembly stand, v) insulating spacer, vi) insulating tube, vii) BDD Ring Electrode, iix) BDD Disk Electrode. b) Removing 3D-printed parts from support. c) Checking alignment between inner and outer brass parts with a ruler.

4.
Glue all the body parts together with a rapid epoxy, starting with the inner brass core and working outwards. Leave the RRDE face up for 30 minutes to allow the rapid epoxy to set. A spare insulating outer can be used as a stand for the RRDE. This step can be seen in Figure S1.4a.

5.
Once set dry fit the ring and disk electrodes into the recesses, which should be sized to match the diameters of the electrodes and ideally be ca. 300 µm deeper than the thickness of the electrodes used. This step can be seen in Figure S1.4b.

6.
Mix the conductive epoxy (if using) according to the manufacturer's instructions and apply a small amount to the brass ring and disk contacts. This step can be seen in Figure   S1.4c.

Supporting Information
S-6 throughout. This step can be seen in Figure S1.5a.

8.
Once the conductive epoxy has set, the RRDE can be removed from the stand and placed face up. The electrode face is then flooded with the insulating epoxy chosen for sealing.
It is advantageous to heat the RRDE and resin in a lab oven to an appropriate temperature, to lower the viscosity of the resin and ensure a good seal around the electrodes. Try not to fill the surface with resin past flush as this will make it harder to polish the surface coplanar in later steps. This step can be seen in Figure S1.5b.

9.
Cure the insulating resin according to manufacturer's instructions. This step can be seen in Figure S1.5c. Figure S1.5: Steps 7-9 of RRDE fabrication including, a) curing the conductive epoxy on a stand to ensure the electrodes are co-planar. b) Flooding the electrodes surfaces with insulating resin. c) UV curing the insulating resin.

S-7
10. Once the insulating resin is cured it can be polished away to reveal the surface of the electrodes, increasingly fine grades of abrasive paper can be used until the electrode surfaces are completely exposed. Due to the large size of the electrodes, it is easy to polish them flat with some care, pause frequently during polishing and hold the RRDE at right angles to the abrasive paper throughout. This step can be seen in Figure S1.6a and b.

11.
After the electrode surfaces are exposed, and typically before each use the surfaces should be alumina polished to ensure the electrodes are clean. This step can be seen in Figure   ESI  The electrodes are exposed. c) Alumina polish to clean the electrodes.

Rough Electrodes:
In the case of a ring or disk electrode with significant surface texture it may be desirable to protect the surface prior to Step 8 where the face is flooded with epoxy, as it may otherwise be impossible to remove all the epoxy from the surface. This could be achieved in two ways; the first method would be to place a piece of tape such as Kapton tape, over the electrode before Step 6. Kapton mask is easy to remove by polishing and should leave no residue. The second method is to flood the surface with PVA Glue before Step 8. The PVA glue is easy to remove by polishing but can also be electrochemically removed by cycling in acid e.g. 0.1 M H2SO4.

Foil and Rod Materials:
While the method of fabrication used was designed for freestanding sheets (wafers) of material or powder compacts (> 100 µm thick) the authors believe it could be easily adapted for other electrode formats. If foil electrode material (< 100 µm thick) was

Supporting Information
S-8 to be used, we suggest attaching it to both the brass inner and outer between Steps 3 and 4 and then omitting steps 5-7. As metal foils are available at relatively low cost for almost all metals this would be a cost-effective way of making a metal electrode. Using rod material is possible, electrodes could be turned to size on a lathe and the RRDE assembled as normal. If the electrodes are thicker than 360 µm, for which the RRDE parts are designed herein, it would be essential to increase the height of both the insulating inner and outer to be at least 100 µm thicker than the electrode (but ideally 300 µm).

Supporting Information
S-9 Table S1 gives a bill of materials for a single RRDE electrode, excluding the cost of the electrodes themselves. If the body is SLA printed that UV resin should be used for sealing as this will ensure adhesion. If PEEK is used a resin with good adherence to PEEK should be used.

Supporting Information
S-10 Figure S2 shows CVs in the redox mediator FcTMA + for the ring and disk electrodes of both RRDEs.

Supporting Information
S-11

ESI.4 Uncompensated Resistance Measurements
Uncompensated resistance, Ru, measurements were made by collecting the i-t data from a series of potential pulses in the non-faradaic region (± 0.1 V) of all electrodes in 0.1 M KNO3 (n=5).
To collect sufficient data points over the timescale of the i-t decay curve a potentiostat fitted with a fast scan module was used (PGSTAT128N with ADC10M.S, Metrohm Autolab, Swizerland). The i-t response was fitted to Equation S1, 1 where ΔE represents the height of the potential pulse and Cdl is the electrochemical double layer capacitance, enabling values of both Cdl and Ru to be obtained.

Equation S1
: The Ru values for the five pulses ( Figures S3 and 4) per electrode were then averaged to obtain the value for Ru presented in Table S2.  Figure S3: i-t experimental non-faradaic decay data (black points) with fits (purple line, equation S1) from which Ru values were extracted for the BDD RRDE electrode.

Supporting Information
S-13

Supporting Information
S-14

ESI.5 White Light Interferometry (WLI) of Disk Electrode Surfaces
WLI images of both the BDD ( Figure S5a) and NDC-BDD ( Figure S5b) disk electrodes were collected using a Bruker ContourGT (Bruker Nano Inc, USA) using a 5× objective. Data was processed in Gwydion 2.55. The plane was levelled and then a 4 th order polynomial background subtracted before roughness measurements were taken. Roughness values are taken as RMS from the entire image. Figure ESI.4b shows the raster of the laser used to create the sp 2 bonded carbon layer on the surface.

Supporting Information
S-15

ESI.6 Quinone Surface Coverage (QSC) Measurements
QSC measurements were performed to quantify the difference in NDC content between the BDD and NDC-BDD disk electrodes following the protocol described in reference. 2 It has been previously shown that QSC measurements show a high degree of correlation with surface NDC both as-grown and from incorporation by laser micromachining. 2 The results of the QSC measurements can be seen in Figure S6. These measurements were collected using digital staircase voltammetry. The area under the quinone oxidation peak was converted to a quinone surface coverage using Equation S2. iA is the integrated area under the peak, n is the number of electrons transferred, A is the area of the electrode measured from WLI, and v is the scan rate.

Γ =
This gave a QSC of 2.0 × 10 -12 mol cm -2 for the BDD disk electrode and a QSC of 2.6 × 10 -10 mol cm -2 for the NDC-BDD disk electrode.

Supporting Information
S-16

ESI.7 Generation Collection Response of the NDC-BDD disk and BDD ring electrode (RRDE)
The collection efficiencies of the RRDE for the redox couple FcTMA + / FcTMA 2+ were measured for comparison with the theoretical efficiency. The potential of the NDC-BDD disk electrode was scanned from a value where no electron transfer occurred (0.00 V vs SCE) to one where the oxidation of FcTMA + to FcTMA 2+ was mass transport limited. The potential of the BDD ring electrode was held at a value where reduction of FcTMA 2+ was also mass transport limited (0.05 V vs SCE). The experiments were performed as a function of rotation rate from 1000 RPM to 2000 RPM in 500 RPM steps, which can be seen in Figure S7.

Supporting Information
S-17

ESI.8 Ring and Disk Currents at 0.60 V vs SCE
The theoretical collection efficiency for the RRDE design employed were calculated using the spreadsheet provided by Pine Research (https://pineresearch.com/shop/wpcontent/uploads/sites/2/2017/02/RRDE-Collection-Efficiency-Calculator-Worksheet-REV003.xlsx). Calculations are based on the those detailed in section 9.4 of Bard and Faulkner. 3 The disk electrode has a diameter of 5.00 mm, the ring has an inner and outer diameter of 7.00 and 9.00 mm respectively. This gives rise to a theoretical collection efficiency of 35%. The empirical collection efficiencies for the BDD RRDE and NDC-BDD RRDE can be seen in tables S3 and S4 respectively.  at rotation rates from 1,000 to 2,500 RPM. Note the Y intercept is fixed at 0.

S-19
Levich analysis of the BDD disk gives an R 2 of 0.999, demonstrating that this electrode performs in good agreement with theory. Extracting a D value from the gradient gives 5.1 × 10 -6 cm 2 s -1 , which is in agreement with that measured using the UME in the same solution.
The Levich plot for the NDC-BDD disk can be seen in Figure S9. Levich analysis of the NDC-BDD disk demonstrates a similar adherence to theory with an R 2 of 0.999. Extracting a FcTMA + diffusion coefficient from this data gives a value of 4.9 × 10 -6 cm 2 s -1 .

Supporting Information
S-20

ESI.10 SEM of Pt-NP Modified Ring Electrode
Field emission scanning electron microscopy (FE-SEM) images of the Pt-NP modified BDD ring electrode was taken using the SE2 secondary electron detector of a Zeiss Supra 55VP FE-SEM (Zeiss, Germany) operating at 10 kV ( Figure S10). The FE-SEM image revealed a highdensity of Pt NPs, 10's of nm in size, with preferential deposition observed on the more highly doped grains of the BDD. Calibration plot of the average current over the last 5 s vs H2O2 concentration after generationcollection experiments. The ring was calibrated before and after generation-collection experiments to ensure that the before calibration was still valid.

Supporting Information
S-23

ESI.13 Faradic Efficiency
Faradaic efficiency of H2O2 generation on both disk electrodes was calculated at -0.6, -0.7, and -0.8 V vs SCE, according to Equation S3. 3 Where iring is the current measured on the ring, iDisk is the current measured on the disk, and NEmpirical is the empirical (measured) collection efficiency for the electrode used.