GB2526184A - Diamond based electrochemical sensor heads - Google Patents

Abstract

A diamond electrochemical sensor head comprises a planar sensing surface, a rear surface through which electrical connections are provided, one or more boron doped diamond (BDD) electrodes disposed within trenches in an electrically insulating diamond support matrix at the sensing surface, the electrodes extending partially through the support matrix from the planar sensing surface towards the rear surface of the support matrix, one or more vias extending from the rear surface of the support matrix to a rear surface of the BDD electrodes, one or more ohmic contacts on the rear surface of the BDD electrodes within the vias in the support matrix, and one or more electrical connectors extending through the vias to the ohmic contacts; wherein the sensor head has a thickness in a range 50 micrometres to 1.5mm, the BDD electrodes extend through support matrix from the planar sensing surface towards the rear surface of the support matrix with a depth in a range 20 micrometres to 500 micrometres, and wherein the ohmic contacts each have a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than 10 mV.

Description

DIAMOND BASED ELECTROCHEMICAL SENSOR HEADS

Field of Invention

Certain embodiments of the present invention relate to diamond based electrochemical sensor heads and methods of fabricating the same.

Background of invention

It has already been proposed in the prior art to provide a diamond based sensor for measuring the electrochemical properties of a solution, Diamond can be doped with boron to form semi-conductive or frilly metallic conductive material for use as an electrode. Diamond is also hard, inert, and has a very wide potential window making it a very desirable material for use as a sensing electrode for an electrochemical cell, particularly in harsh chemical, physical, and/or thermal environments which would degrade standard metal based electrochemical sensors. In addition, it is known that the surface of a boron doped diamond electrode may be frmnctionalized to sense certain species in a solution adjacent the electrode.

One problem with using diamond in such applications is that diamond material is inherently difficult to manufacture and form into suitable geometries for sophisticated electrochemical analysis, To date, diamond electrodes utilized as sensing electrodes in an electrochemical cell have tended to be reasonably simple in construction and mostly comprise the use of a single piece of boron doped diamond configured to sense one physical parameter or chemical species at any one time, More complex arrangements have involved introducing one or more channels into a piece of boron doped diamond through which a solution can flow for performing electrochemical analysis. However, to date the present applicant is unaware of sophisticated diamond based electrochemical sensors which can perform multiple sensing functions at the same time, particularly configured for use in harsh environments. Due to the inherent difficulties involved in manufacturing and forming diamond into multi-structural components, even apparently relatively simple target structures can represent a significant technical challenge.

In terms of prior art arrangements, the present applicant already has a number of patent publications relating to diamond based electrochemical sensor structures and methodologies including WO200S/012894 and W02012/1 56203. :i.

W02005/012894 describes a microelectrode comprising a diamond layer formed from electrically non-conducting diamond and containing one or more pin-like projections of electrically conducting diamond extending at least partially through the layer of non-conducting diamond and presenting areas of electrically conducting diamond at a front sensing surface. A number of different fabrication routes are disclosed for obtaining such structures. One route is described as follows. An insulating diamond layer (natural or synthetic) may be polished flat and parallel. This layer may then have features etched into one face, using for example plasma or chemical etching through a mask, chemical etching of implant damaged features, or laser etching, to generate etch features which when filled will become the electrodes. A boron doped diamond layer may then be grown by CVD methods onto this patterned surface, such that it fills the recesses processed into the surface, Polishing from the front removes the continuous boron doped layer and reveals a planar sensing surface with isolated regions of boron doped diamond material surrounded by an electrically insulating diamond support matrix. Further processing from the rear face can provide either a continuous contact layer or individual connections. For example, one approach involves laser drilling to a rear surface of the boron doped regions followed by filling with metallisation and individual wire bonding, W02005/O 12894 also discloses that an interconnection layer or layers on the back of the layer presenting the analysis surface may also be boron doped diamond or as an alternative may be graphite, which may be grown or implanted, or generated in situ by laser or implantation damage, possibly modified by subsequent annealing. Other alternatives include metals which may be deposited and patterned using any standard technique such as vapour deposition, sputter deposition, electroplating, laser ablation etc. WO2012/126802 describes a similar diamond based electrochemical band sensor comprising a plurality of boron doped diamond band electrodes disposed within a diamond body. Each boron doped diamond electrode has a length / width ratio of at least 10 at a front sensing surface of the sensor. It is described that it is advantageous from a fUnctional perspective for certain sensing applications to provide band electrodes which have a high aspect ratio at the sensing surface such that a length of a band electrode across the sensing surface is very much larger than a width of the band electrode.

W02013/135783 disclosed optimized boron doped diamond materials for electrochemical sensing applications which have a high boron content and a low sp2 carbon content which can be used to form boron doped diamond electrodes which has a broad solvent window, a high degree of reversibility, and a low capacitance.

One problem the present inventors have tbund with the aforementioned approach is that ibllowing the technique of patterning the surface of an electrically insulating diamond substrate to form trenches which are then ifiled with boron doped diamond material using a CYD overgrowth technique, the geometry of the trenches and the growth parameters used in the CVD overgrowth technique must be carefully controlled to obtain tilling of the trenches with boron doped diamond without the formation of voids and also to ensure that such filling can be achieved with high quality boron doped diamond material optimized for electrochemical sensing applications.

One way to mitigate the aforementioned problem is to form shallow trenches in the electrically insulating diamond substrate which are easier to fill with boron doped diamond material using a CVD overgrowth technique. However, this approach is itself problematic for a number of reasons. For example, after overgrowth of boron doped diamond material the deposited boron doped diamond material must be processed back to reveal the boron doped diamond electrode structure. Diamond surface processing techniques are not as accurate as, for example, silicon processing techniques, and if the boron doped diamond electrode structure is very thin then it can be difficult to process back the deposited boron doped diamond material to reveal the electrode structure without processing off the electrode structure or alternatively not consistendy revealing the electrode structure without leaving some boron doped diamond material over other portions of the sensing surface.

Furthermore, if the electrode structure is made very thin then the electrodes have a higher electrical resistance in the plane of the sensing surface leading to a potential drop over each electrode sensing surface and a decrease in the reversibility of the electrodes in electrochemical sensing applications.

Further still, if the electrode structure is made very thin then it can be difficult to cut through the electrically insulating diamond support matrix to a rear surface of a thin boron doped diamond electrode disposed therein to form vias for rear electrical connections without either falling short of the rear surface of the thin boron doped diamond electrode or otherwise over-shooting the rear surface and drilling through to the front sensing surface of the boron doped diamond electrode.

Yet another problem the present inventors have found with the aforementioned approach is that it can be difficult to form a good ohmic contact on a rear surface of the boron doped diamond electrode within a via though an electrically insulating diamond support matrix and particularly one which forms a good ohmic contact with a low resistance and/or a low capacitance and/or a contact which is formed with a consistent geometry and performance due to variations in the internal surface of the via formed in the diamond material, If the electrode structure is very thin to aid formation via a CVD overgrowth technique then it can also be problematic to form a good ohmic contact in a rear surface of a boron doped diamond electrode without impinging on the sensing surface.

It is an aim of certain embodiments of the present invention to solve one or more of the aforementioned problems. In particular, it is an aim of certain embodiments of the present invention to provide optimized electrochemical sensor structures fabricated from diamond materials.

Summary of Tnvention

According to one aspect of the present invention there is provided a diamond electrochemical sensor head comprising: a planar sensing surface; a rear surface through which electrical connections are provided; one or more boron doped diamond electrodes which are disposed within trenches in an electrically insulating diamond support matrix at the planar sensing surface, the one or more boron doped diamond electrodes extending partially through the electrically insulating diamond support matrix from the planar sensing surface towards the rear surface of the electrically insulating diamond support matrix; one or more vias extending from the rear surface of the electrically insulating diamond support matrix to a rear surface of the one or more boron doped diamond electrodes within the electrically insulating diamond support matrix; one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix; and one or more electrical connectors extending through the one or more vias to the one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix, wherein the diamond electrochemical sensor head has a thickness from the planar sensing surface in which the one or more boron doped diamond electrodes are disposed to the rear surface of the electrically insulating diamond support matrix in a range 50 micrometres to 1,5 millimetres, wherein the one or more boron doped diamond electrodes extend through the electrically insulating diamond support matrix from the planar sensing surface towards the rear surface of the electrically insulating diamond support matrix with a depth in a range 20 micrometres to 500 micrometres, and wherein the one or more ohmic contacts within the one or more vias on the rear surface of the one or more boron doped diamond electrodes each have a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than tO mY, where the ohmic drop is defined by Ix R with I being current and R being total resistance.

According to another aspect of the present invention there is provided a method of fabricating a diamond electrochemical sensor head as previously defined, the method comprising: starting with an electrically insulating diamond substrate having planar front and rear surfaces and a thickness between said planar front and rear surfaces in a range 50 micrometres to 1.5 millimetres; cutting one or more trenches in the planar front surface of the electrically insulating diamond substrate, wherein the one or more trenches have a depth in a range 20 micrometres to 500 micrometres; growing boron doped diamond material over the front surface of the electrically insulating diamond substrate and into the one or more trenches; processing back the boron doped diamond material over the planar front surface of the electrically insulating diamond substrate to form a planar sensing surface comprising one or more boron doped diamond electrodes surrounded by an electrically insulating diamond support matrix, the planar sensing surface having a surface roughness Ra less than 100 nm afler processing, the one or more boron doped diamond electrodes extending through the electrically insulating diamond support matrix from the planar sensing surface towards a rear surface of the electrically insulating diamond support matrix with a depth in a range 20 micrometres to 500 micrometres, and wherein a distance between the planar sensing surface and the rear surface of the electrically insulating diamond support matrix is in a range 50 micrometres to 1.5 millimetres; forming one or more vias extending from the rear surface of the electrically insulating diamond support matrix to the rear surface of the one or more boron doped diamond electrodes within the electrically insulating diamond support matrix; forming one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix, wherein the one or more ohmic contacts within the one or more vias on the rear surface of the one or more boron doped diamond electrodes each have a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than 10 mY, where the ohmic drop is defined by Ix R with I being current and R being total resistance; and forming one or more electrical connectors extending through the one or more vias to the one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix.

Preferably, during the step of growing boron doped diamond material over the front surface of the electrically insulating diamond substrate and into the one or more trenches, growth conditions are controlled to ensure that the one or more trenches are filled and to ensure that the boron doped diamond material filling the one or more trenches has the following characteristics: a solvent window meeting one or both of the following criteria as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode in a solution containing only deionised water and 0,1 M KNO3 as a supporting electrolyte at pH 6: the solvent window extends over a potential range of at least 4.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 38 mA cm'2; and the solvent window extends over a potential range of at least 3.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 0.4 mA cm'2; a peak-to-peak separation AE (for a macroelectrode) or a quartile potential AE314,114 (for a microelectrode) of no more than 70 mY as measured by sweeping a potential of the boron doped synthetic diamond material at a rate of too mY s' with respect to a saturated calomel reference electrode in a solution containing only deionised water, 01 M KNO3 supporting electrolyte, and t mM of FcTh4A or Ru(NH2)63 at pH 6; and a capacitance of no more than 10 tF cm'2 as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode between 70 mV and -70 mV in a solution containing only deionised water and 0.1 M KNO3 supporting electrolyte at pH 6, measuring resultant current, subtracting a current value at 0 V when sweeping towards negative potentials from a current value at 0 V when sweeping towards positive potentials, dividing the subtracted current value by 2, and then dividing the result by an area (cm2) of the boron doped synthetic diamond material and by a rate at which the potential is swept (Vs1) to give a value for capacitance in F cm'2, Preferably, growth conditions are controlled such that an interface region between the one or more boron doped diamond electrodes and the electrically insulating diamond support matrix comprises no voids having a largest lateral dimension greater than O micrometres, 5 micrometres, 1 micrometres, 500 nanometres, 300 nanometres, or 100 nanometres, In one approach the one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix are formed of non-diamond carbon, e.g. graphite. One problem with the aforementioned methodology is that it is difficult to form a metal ohmic contact on the rear surface of the boron doped diamond electrodes through narrow laser-cut blind holes or trenches forming vias in the rear of the electrically insulating diamond support matrix, It has been found that the rear surface of the boron doped diamond electrodes can be graphitized using a laser and such a graphitized or non-diamond-carbon surface within the through-hole provides an ohmic contact which is suitable for electrochemical sensor application. Some of the non-diamond-carbon rear contact can be removed by acid cleaning after formation. In this regard, it is advantageous to provide a relatively thin layer of non-diamond carbon for the ohmic contact to present a low resistance contact, As such, advantageously the ohmic contact is formed to have a low resistance per unit area of, for example, no more than 10 0 cm'2.

Laser pulsing and cross-hatching can be used to ensure an even cut during laser cutting to form the vias which can ensure that there is no "drill-through" to the front sensing surface and can also aid in providing a relatively uniform, low resistance ohmic contact,

Brief Description of the Drawings

For a befter understanding of the present invention and to show how the same may be carried into effect, embodiments of the present invention will now be described by way of example only with reference to the accompanying drawings, in which: Figure 1 shows a schematic flow diagram illustrating a fabrication method for manufacturing a diamond electrochemical sensor head.

Figure 2(a) is a further schematic illustrating the step-by-step fabrication of an all-diamond electrode.

Figure 2(b) shows 3D schematics of different all-diamond pBDD electrodes fabricated using the process illustrated in Figure 2(a) including: (i) dual band; (ii) triple band; (iii) multiple bands; (iv) disk; (v, vi) ring-disk.

Figures 2(c)(i) to 2(c)(iii) show FE-SEM images of all-diamond devices: (i) triple band (in-lens); (ii) macrodisk (secondary electron; SE); and (iii) ring-disk (SE),

S

Figure 2(c)(iv) shows an optical photograph of a top contacted ring-disk (disk is 3 mm diameter).

Figure 3 is a further schematic illustrating a step-by-step fabrication for the production of co-planar, individually addressable boron doped diamond (BDD) electrodes insulated in diamond, top 2md back contacted, at both the micro-and macroelectrode scale.

Figure 4(a) illustrates back contacting of all-diamond electrodes via an in-lens FE-SEM image of a cross-sectioned macroelectrode disk structure. The pBDD is clearly evident as a region of lighter contrast at the top of the image. The internal surface of the electrode is contacted by laser micromachining a hole through the insulating diamond to the pBDD layer.

Inset shows a photograph looking through the transparent top surface of a multiple band electrode (at least seven black, boron doped, electrodes are visible). The lasered back contacts are visible as black cones. The bands are each 180 jim wide.

Figure 4(b) illustrates back contacting of all-diamond electrodes, White light interferometry cross-sectional depth profiles are shown which were taken at various stages during the laser micromachining process, demonstrating a high level of depth control. The y axis has been normalized with respect to the thickness of the diamond wafer (1 mm). The dotted line indicates the front electrode face of the diamond sensor head.

Figure 5(a) shows, for all-diamond, individually addressable, back contacted band electrodes, typical Raman spectra recorded from higher (lower trace) and lower (upper trace) doped grains on an electrode with width of 200 jim. The inset shows a photograph of a ten band electrode device, Band lengths were all 10mm and of widths in the range SO jim to 1 mm, Figure 5(b) shows experimental (solid lines) and simulated (dashed lines) cyclic voltamograms for 1 miv1 Ru(NH3)63 reduction/oxidation recorded at 100 my sfl', in 0.1 M KNO3, on different width pBDD band electrodes, 1 mm, 500 rim, 200 jim, 00 jim, and 50 jim. The inset shows solvent windows recorded in 0.1 M KNO3 at 100 mV for the same five electrodes.

Figure 6 shows a schematic diagram of a 3D band system. The BDD electrode is labelled surface I. The insulating diamond surface is labelled 2. The external surfaces of the solution that define the volume are labelled 3. For clarity, a 2D section has been taken through the 3D model. The identity of the faces is shown on the projected face on the right hand side. The radial model for the ring-disk is also shown, where the ring and disc are labelled Ia and Ib, respectively. The inert diamond surface is boundary 2, the closed volume edge boundaries are labelled 3. The ring-disk system has an additional boundary of axial symmetry that is labelled 4.

Figure 7 shows experimental (solid lines) and simulated (dashed lines) CVs for 1 mM Ru(NH3)63 &ectrolysis in 0.1 NI KNO3 at a BDD ring electrode of dimensions 3.1 mm and 3.2 mm inner and outer diameters respectively at a scan rate of 10, 20, 30, 50, 100, 300, and 500 my 5* Figure 8 shows in-lens FE-SEM images (i) and corresponding AFM images (ii) of the boundary between insulating arid pBDD regions across two different pBDD band electrodes (a) and (b). Regions of the FE-SEM images imaged by AFM are highlighted with black boxes. Height profiles across the insulating-pBDD and grain boundaries in the pBDD region are shown.

Figure 9 shows FE-SENT images of all-diamond electrodes including: (a) an in-lens SENT image of a dual band structure with the inset shows a cross section of the same device; and (b) a secondary electron FE-SENT image of a disk electrode where the laterally growing grains are highlighted with arrows, Figure 10 shows an example of a defected ring-disk electrode system. Figure 10(a) shows an FE-SEM image of a ring-disk system with the defect clearly visible as a dark line along the inner perimeter of the ring electrode. The defect region is expanded in the inset. Both scale bars are 100 jim. Figure 10(b) shows a corresponding solvent window for the ring electrode in 0,] Ni KNO3 at 100 mY s'.

Figure 11 shows characterisation data for a 1.02 mm diameter all-diamond macro electrode: (a) CV with 1 mM Ru(NH3)(,3E; and cb) solvent window and (inset) Raman data showing no evidence of non-diamond-carbon (NDC).

Detailed description of Certain Embodiments

Boron doped diamond (BDD) is an interesting electrode material due to its ultra-wide solvent window in aqueous media, low capacitive currents, and high resistance to fouling and corrosion (Swain, G. lvi.; Ramesham, R., Anal. Chem. 1993, 65, 345-351; Panizza, M., Cerisola, G., Electrochim. Acta 2005, 51, 191-199; and Martin, H. B., Argoitia, A., Landau, U., Anderson, A. B., Angus, J. C., J. Electrochem. Soc. t996, 143, L133-L136).

BDD for electrochemical applications is typically either as-grown in nanocrvstalline (nc) thin film form on a conducting or insulating substrate, or polycrystalline (p) and thick enough to be removed from the growth substrate in a freestanding form, Non-diamond-carbon (NDC) content can be high in ncBDD and unless growth is very carefUlly controlled is also prevalent in pBDD materials (Hutton, L. A., lacobini, J G., Bitziou, E., Channon, R. B., Newton, M. E., Macpherson, J. V., Anal, Chem. 20t3, 85, 7230-7240).

For electrochemical studies a cell of defined geometry is typically placed, or clamped, over the electrode to create a defined geometric area (Granger et al., AnaL (hern. 2000, 72, 3 793- 3804; Yano et al., .1. Liectrochern. Soc. 1998, 145, 1870-1876). Electrical contact is either made to the base (if the substrate is conducting) or through a top contact (if the substrate is insulating). For freestanding pBDD the material can also be machined into a defined structure e.g. a cylinder, cuboid cc, electrically back-contacted and then sealed in glass (Hutton et al., Anal. Chem. 2008, 81, 1023-t032) or an insulating polymer, such as PTFE (Rao et al., Anal.

Chem. 999, 71, 2506-H; Salimi et al., The Analyst 2004, 129, 9), epoxy (Prado et al., Analyst 2002, 127, 329-332) or PEEK (Svorc et aL, Bioelectrochem. 2012, 88, 36-41) to produce electrode formats more akin to commercially available, conventional electrode materials such as Pt, Au, glassy carbon, Disk electrodes with diameters in the range tens of jim (Wakerley et al., Chem. Comm. 2013, 49, 5657-9) to several mm (Hutton et al., Anal.

Chem. 2009, 81, t023-1032) have been fabricated this way. pBDD microelectrodes have also been produced by growing thin films of pBDD onto sharpened W wires and sealing with epoxy and glass (Sarada et a!., Electrochem. Soc. 1999, 146, 1469-1471). However, in all cases the mechanical and chemical stability of the material used to insulate is always inferior to that of the pBDD, limiting potential applications.

Moving to an all-diamond format offers several advantages, including: (i) formation of a robust seal between the pBDD electrode and insulating diamond; (ii) resistance to chemical and thermal stress and abrasive wear; and (iii) longevity in extreme environments. There have so far been limited attempts to produce all-diamond electrodes. Using freestanding pBDD, laser ablation was used to define pillar stmctures in pBDD. Subsequent overgrowth with insulating diamond and polishing revealed an array of co-planar pBDD microdisk electrodes, -50 jim in diameter. However, this approach meant that the array was not individually addressable and insulating diamond growth needed to be carefully controlled to ensure no defects, pinholes or cracks in the overgrowth layer (Colley et al., Anal. Chem. 2006, 78, 2539-2548; Pagels et al., Anal. Chem. 2005, 77, 3705-3708). A three layer diamond growth process was also described, where a microlayer of pBDD was sandwiched between two electrically insulating diamond layers. A macro hole cut through the structure resulted in the formation of an all-diamond ring electrode, suitable for use in flow studies (Hutton et al., Anal. Chem. 20t 1, 83, 5804-5808). Finally, thin film ncBDD, patterned to isolate insulating diamond overgrowth from defined regions resulted in a non-addressable recessed ncBDD UI1VIE electrode array (Hees et al., Chem. 2013, 19, 11287-92).

An alternative fabrication route is described herein resulting in the production of co-planar all-diamond electrodes, of any geometry, both at the macro-and micro-scale, A micro-electrode is one with at least one lateral dimension less than 1 mm. Single or multiple individually addressed electrodes can be produced on one device, The methodology is exemplified by demonstrating the fabrication and characterisation of individually addressable disks, bands, and ring-disk geometries, suitable for a wide range of electrochemical applications. All-diamond devices are fully characterized using Raman spectroscopy, electron microscopy, conductivity measurements, and electrochemistry.

Figure 1 shows a schematic flow diagram illustrating a fabrication method for manufacturing a diamond electrochemical sensor head as described below, The fabrication method starts (Figure 1(a)) with an electrically insulating diamond substrate 2 having planar front and rear surfaces 4, 6 and a thickness tin a range 50 micrometres to L5 mm. At least the planar front surface may be processed to have a surface roughness less than mn although this is not considered essential for the starting substrate. The substrate should have a suitable thickness to receive the overgrown electrodes and providing sufficient mechanical robustness while not being overly thick so that it is suitable for forming reliable back contacts to the electrodes through a rear surface of the substrate in a cost effective manner. For example, the thickness from the planar sensing surface in which the one or more boron doped diamond electrodes are disposed to the rear surface of the electrically insulating diamond support matrix may be no less than 75 micrometres, 100 micrometres, 150 micrometres, or 200 micrometres, and/or no more than 1 millimetre, 750 micrometres, 500 micrometres, 300 micrometres, or 200 micrometres, or any combination of these upper and lower values.

The planar front surface 4 of the electrically insulating diamond substrate 2 is then patterned by cutting one or more trenches 8 therein (Figure 1(b)). The one or more trenches 8 have a depth d in a range 20 micrometres to 500 micrometres and/or a width-to-depth aspect ratio w:i of at least I. The minimum depth of the trenches is defined such that boron doped diamond electrodes formed therein can be reliably surface processed at the front sensing surface, reliably back-contacted without forming through-holes to the front sensing surface, and so that the resultant electrodes are not too thin that in-plane resistance is too high and reversibility of the electrodes in use is impaired. Conversely, the maximum depth of the trenches is defined such that the trenches can be reliably filled with boron doped diamond matcrial via ovcrgrowth without forming voids within thc trcnchcs and in particular so that such trench filling can be achieved using suitable CVD diamond growth conditions to simultaneously provide high quality boron doped diamond material having excellent electrochemical characteristics, For example, the depth of the one or more trenches may be no more than 400 micrometres, 300 micrometres, 200 micrometres, or tOO micrometres, and/or no less than 30 micrometres, 40 micrometres, or 50 micrometres, or any combination of these upper and lower values.

In addition to the above, the cutting of the one or more trenches 8 should be controlled such that the trench walls and base have a root mean squared surface roughness Rq no more than 5 micrometres, 4 micrometres, 3 micrometres, 2 micrometres, or I micrometre as this aids overgrowth and filling of the one or more trenches without the formation of voids, Further still, while the trench walls in Figure 1(b) are illustrated as being vertically oriented, according to certain embodiments the trench walls may be angled inwards to aid in trench filling during overgrowth (indicated by dotted lines in Figure t(b)) without the formation of voids. In this regard, trench walls which are angled outwards to form an overhanging structure should be avoided as this can lead to the formation of voids under such an overhanging structure Boron doped diamond material 10 is then grown over the front surface of the electrically insulating diamond substrate 2 and into the one or more trenches 8 as illustrated in Figure 1(c). CVD diamond growth parameters can be controlled in such a way as to achieve filling of the one or more trenches 8 without the formation of voids. Furthermore, it has been found that CVD diamond growth parameters can simultaneously be controlled such that the boron doped diamond material 10 has a high concentration of boron dopant with a low sp2 carbon content to ensure that the boron doped diamond material filling the one or more trenches has the following characteristics: a solvent window meeting one or both of the following criteria as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode in a solution containing only deionised water and 0.1 lvi KNO3 as a supporting electrolyte at pH 6: the solvent window extends over a potential range of at least 4.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 38 mA cm2; and the solvent window extends over a potential range of at least 3.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 0.4 mA cm2; a peak-to-peak separation AE (for a macroelectrode) or a quartile potential AE1l/1 (for a microelectrode) of no more than 70 my as measured by sweeping a potential of the boron doped synthetic diamond material at a rate of 100 mV W1 with respect to a saturated calomel reference electrode in a solution containing only deionised water, 01 M KNO3 supporting electrolyte, and t mlvi of FcTh4A or Ru(NH2)63 at pH 6; and a capacitance of no more than 10 RE cm2 as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode between 70 mV and -70 mY in a solution containing only deionised water and 0.

Ni KNO3 supporting electrolyte at pH 6, measuring resultant current, subtracting a current value at 0 V when sweeping towards negative potentials from a current value at 0 V when sweeping towards positive potentials, dividing the subtracted current value by 2, and then dividing the result by an area (cm2) of the boron doped synthetic diamond material and by a rate at which the potential is swept (Vs') to give a value for capacitance in F cm2, In light of the above, it is important to tune the CVD diamond growth parameters such that both void-free filling of the trenches and high quality "sensor grade" boron doped diamond material is achieved simultaneously.

Materials as described above have been fabricated using a microwave plasma activated chemical vapour deposition (CVD) synthesis process. A microwave plasma activated CVD diamond synthesis system typically comprises a plasma reactor vessel coupled both to a supply of source gases and to a microwave power source. The plasma reactor vessel is configured to form a resonance cavity supporting a standing microwave, typical frequencies used for this heating application include 2,45 0Hz and approximately 900 MHz depending on the RF spectrum allocation of each country. In this work the example conditions are given for a system equipped with a 2.45 0Hz microwave source. Source gases including a carbon source and molecular hydrogen are fed into the plasma reactor vessel and can be activated by the standing microwave to form a plasma in high field regions. If a suitable substrate is provided in close proximity to the plasma, reactive carbon containing radicals can diffuse from the plasma to the substrate and be deposited thereon. Atomic hydrogen can also diffuse from the plasma to the substrate and selectively etch off non-diamond carbon from the substrate such that diamond growth can occur, If a source of boron such as diborane gas is introduced into the synthesis atmosphere then boron doped synthetic diamond material can be grown. Single crystal synthetic diamond materials are typically fabricated via homoepitaxial growth on single crystal diamond substrates. In contrast, polycrystalline synthetic diamond material can be grown on silicon substrates, refractory metal substrates, or on polycrystalline diamond substrates.

important growth parameters include the microwave power density introduced into the plasma chamber (typically ranging from less than or equal to I kW to 5 kV or more for a substrate area < 20 cm2), the pressure within the plasma chamber (typically ranging from less than or equal to 50 Torr (i.e. 6.67 kPa) to 350 Torr (i.e. 46.66 kPa) or more), the gas flow velocity flowing through the plasma chamber (typically ranging from a few lOs of sccm (standard cm3 per minute) up to hundreds or even thousands of sccm), the temperature of the substrate (typically ranging from 600 to 1200°C) , arid the composition of the synthesis atmosphere (typically comprising 0.5 to 20% by volume of carbon containing gas (usually methane) with the remainder of the synthesis atmosphere been made up of hydrogen). For boron doping the synthesis atmosphere will typically comprise a boron containing gas such as diborane at a concentration from equal to or less than 0.01% up to several % by volume.

The problem to be solved is what growth parameters to select in order to fabricate synthetic boron doped diamond materials with optimized electrochemical sensing properties. Suitable growth parameters for such boron doped diamond materials have been described in W02013/135783. However, it has been found that CVD growth parameters should advantageously be changed when growing boron doped diamond materials into trenches in order to achieve void-free filling of the trenches while simultaneously achieving boron doped diamond materials with optimized electrochemical sensing properties. The problem is how to achieve high levels of boron doping while avoiding incorporation of sp2 carbon during growth and also simultaneously ensuring void-free filling of the trenches. This has been achieved by controlling substrate temperature in a range 650 to 850°C, using a synthesis atmosphere which has a relatively low concentration of carbon containing gas (e.g. in a range 0,5% to 3% of total gas flow), a high power density (e.g. 2.8 to 3.8 kW over a 50 mm diameter substrate) in combination with a relatively high reactor pressure (e.g. in the range to 160 Torr (i.e. 16.00 kPa to 21.33 kPa)) using a high gas flow configuration and total flow rate as previously described for single crystal growth. When compared to the growth conditions disclosed in W02013/]35783, the substrate temperature has been reduced, the carbon containing gas concentration has been reduced, the power density has been reduced, and the pressure has been reduced, As such, a new parameter space has been found which is capable of achieving void-free boron doped diamond growth into trenches cut in the growth surface of a diamond substrate while also achieving boron doped diamond material which has a high level of boron doping while avoiding incorporation of sp2 carbon during growth.

Advantageously, the trenches are filled with boron doped diamond material such that an interface region between the one or more boron doped diamond electrodes and the electrically insulating diamond support matrix comprises no voids having a largest lateral dimension greater than 10 micrometres, 5 micrometres, 1 micrometres, 500 nanometres, 300 nanometres, or 100 nanometres, Such voids can be detected and measured using a microscopic imaging technique such as scanning electron microscopy.

After growth of the boron doped diamond material 10, the boron doped diamond material over the planar front surface 4 of the electrically insulating diamond substrate 2 is processed back to form a planar sensing surface 12 comprising one or more boron doped diamond electrodes t4 surrounded by an electrically insulating diamond support matnx 16 as illustrated in Figure 1(d).

The planar sensing surface 12 advantageously has a surface roughness Ra of no more than nm, 75 nm, 50 nm, 20 nm, or 10 nm after processing. It has been found that the surface finish affects fhnctional performance of the one or more boron doped diamond electrodes 14 in electrochemical sensing applications.

In addition to the above, the one or more boron doped diamond electrodes 14 extending through the electrically insulating diamond support matrix 16 from the planar sensing surface 12 towards a rear surface 18 of the electrically insulating diamond support matrix with a depth din a range 20 micrometres to 500 micrometres. For example, the one or more boron doped diamond electrodes 14 may have a depth a' of no more than 400 micrometres, 300 micrometres, 200 micrometres, or 100 micrometres, and/or no less than 30 micrometres, 40 micrometres, or 50 micrometres, or any combination of these upper and lower values.

A distanceD between the planar sensing surface U and the rear surface 18 of the electrically insulating diamond support matrix is in a range 50 micrometres to 1.5 mm. The geometry of the structure illustrated in Figure 1(d) is adapted to account for several factors. First, as previously indicated, the depth a' is defined such that it is not too shallow so as to reliably process and back contact the electrodes and not too thick that it is difficult to achieve filling via overgrowth without forming voids within the one or more trenches. Secondly, the distance D should be sufficiently small that vias can reliably be cut through the electrically insulating diamond support matrix 16 to a rear surface of the one or more electrodes 14 in a cost effective manner while being sufficiently large as to provide mechanically robustness to the sensor head structure during fabrication and in use. For example, the diamond sensor head may have a thickness no less than 50 micrometres, 75 micrometres, 100 micrometres, micrometres, or 200 micrometres, and/or no more than 1.5 millimetres, 1 millimetre, 750 micrometres, 500 micrometres, 300 micrometres, or 200 micrometres, or any combination of these upper and lower values.

Next, one or more vias 20 are formed extending from the rear surface 18 of the electrically insulating diamond support matrix 16 to the rear surface of the one or more boron doped diamond electrodes 14 within the electrically insulating diamond support matrix as illustrated in Figure 1(e). The vias 20 may be in the form of holes or in the form of elongate trenches, in either case, this step must be carefully controlled to ensure that the vias extend to a rear surface of the one or more boron doped diamond electrodes 14 without extending through the one or more boron doped diamond electrodes 14 to the front sensing surface 12. In one approach, laser cross-hatching can be used to ensure an even cut during formation of the vias which can ensure that there is no drill-through to the front sensing surface 12. In addition, providing a smooth surface within the vias 20 can also aid in providing a relatively uniform ohmic contact at a rear surface of the one or more boron doped diamond electrodes 14 which has a low resistance and/or capacitance. In one example, the one or more vias 20 are formed as elongate trenches in one, or a small number, of cutting passes as this provides a cost effective route to fabrication of the vias. In this case, the distance between the rear surface of thc clcctrodcs and thc rcar surfacc of thc clcctrically insulating diamond support matrix should be sufficiently small that the vias can be formed in this manner.

In relation to the above, it has been noted that the aspect ratio of the via during cutting provides an indicator as to the quality of the cutting procedure and successful achievement of the ultimate goal to cut a via with a smooth internal surface and which reaches a doped diamond electrodes without punching through to the front sensing surface, In this regard, the via tends to have a tapered profile as indicated in Figure 1(e) with side walls which are sloped at an angle in a range 30° to 60° relative to a vertical direction perpendicular to the plane of the sensing surface, That said, while such sloped internal walls are typical of a laser cutting technique, if another technique such as an etching technique is utilized to form the vias then the side walls can be substantially vertical, One or more ohmic contacts 22 are formed on the rear surface of the one or more boron doped diamond electrodes 14 within the vias 20 in the electrically insulating diamond support matrix 16. The one or more ohmic contacts 22 within the vias on the rear surface of the one or more boron doped diamond electrodes are formed such that each contact has a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than 10 my, 8 my, 6 my, 4 my, 2 my, or i my where the ohmic drop is defined by I x R with I being current and R being total resistance and/or has a resistance per unit area of no more than 10 0 cm2, 8 0 cm2, 5 0 cm2, 3 0 cm2, or 1 0 cm2. These values are selected to define an ohmic contact which does not cause undue interference with the electrochemical sensing measurements. It is more difficult to fabricate such a contact within a via 20 as compared to an external surface of a sensor head. The formation of such a contact within a via 20 is at least partially enabled by ensuring that the cutting technique used to form the via is sufficiently well controlled to achieve a relatively low roughness surface within the via, e.g. an ElviS surface roughness Rq of no more than 5 micrometres, 4 micrometres, 3 micrometres, 2 micrometres, or I micrometre. Laser pulsing and cross-hatching can be used to achieve this goal. In general, the following laser parameters may be utilized: 1) Laser at a power above the ablation threshold for diamond (i GW cm2), 2) Hatch pitch and pulse pitch is equal to the radius of the laser spot.

3) Use interferometry or other profiling method to determine depth of cut per pass.

Repeat passes as appropriate to obtain desired depth.

4) Step the focus of the beam towards the sample to keep the same focal point at the cutting surface.

5) Ensure that the acceleration and deceleration of the stage does not affect the shape of the geometry produced.

6) Perform repeat passes to obtain the required depth.

Calibration is generally required to determine the depth of each laser pass. Hatch direction is decided by geometry; squares or circles can be hatched in multiple directions, rotating the hatch direction with each pass. Long thin shapes can be hatched in a single direction, parallel to the long axis of the shape. Insulating diamond surfaces can be coloured, e.g. with black pen (glass pen). Focussing can then be performed by eye (microscope camera) to the closest tm from the surface. The following laser parameters were utilized in the examples described herein: 1) Power(attenuator): initial pass set in a range 50 -75% (sufficient to cut but not cause fracturing/blow out) and subsequent passes controlled in a range 20 -80% (much more flexibility after initial pass).

2) Frequency: 100-333 Hz.

3) Hatch pitch: 3 jim (radius of spot size).

4) Pulse pitch: 3 jim (radius of spot size).

5) F speed: Fl -F0.3 (mm/s) (essentially defined by frequency and pulse pitch).

The maximum F-speed is limited by acceleration and deceleration of the stage.

Generally shapes of size -100 jim will have significant variability in depth at F-speeds > 0.3 mm/s.

6) Calibration of z-step: on a test piece of the same material a small shape is cut for calibration, A three-pass laser calibration is performed without any z-step.

Interferometry is used to measure the depth of each laser pass and calculate the z-step.

If the bottom is rough then reduce the F speed or do a further three-pass laser calibration using the z-step determined previously.

The one or more ohmic contacts 22 disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support mathx may be formed of non-diamond carbon, e.g. graphite. It has been found that the rear surface of the boron doped diamond electrodes can be graphitized using a laser as outlined above and such a non-diamond-carbon surface within the through-hole provides an ohmic contact which is suitable for electrochemical sensor application. Some of the non-diamond-carbon rear contact can be removed by acid cleaning after formation to reduce the resistance of the contact, In this regard, it is advantageous to provide a relatively thin layer of non-diamond carbon for the ohmic contact to present a low resistance contact as described previously, As previously indicated, laser pulsing and optionally cross-hatching can be used to ensure an even cut during laser cutting to form the vias which can ensure that there is no cut-through to the sensing surface and can also aid in providing a relatively uniform and reproducible ohmic contact, Given that the critical parameters for the trenches and vias have no%v been deduced, it is also envisaged that other techniques may be used, in a controlled manner, to form suitable trenches 8 and/or vias 20 including, for example, electron beam cutting, ion beam cutting, hot metal dissolution, or plasma etching techniques. Furthermore, the contacts may be formed by a brazing process or using a thermal metal dissolution technique.

In addition, a thermal sensor can be positioned in front of the planar sensing surface to determine the depth of the cut during formation of the via by taking thermal measurements.

This can aid in accurately and reproducibly cutting to a rear surface of the electrodes when forming the vias, Further still, it should be noted that while the via 20 shown in Figure 1(e) is oriented in a substantially vertical direction, the via can be formed at other angles relative to the planar front sensing surface.

Finally, as illustrated in Figure 1(f), one or more electrical connectors 24 are provided extending through the one or more vias to the one or more ohmic contacts 22 disposed on the rear surface of the one or more boron doped diamond electrodes 14 within the vias in the electrically insulating diamond support matrix 16. The one or more electrical connectors 24 may be provided by bonding wires to the contacts 22 or by deposition of an electrically conductive material into the vias 20.

In relation to the above, it will be noted that while the generic technique of forming diamond electrochemical sensor heads via patterning and CVD diamond overgrowth is known in the art (see, for example, W02005/0 12894), the present inventors have encountered a number of unforeseen technical problems when fabricating such structures and have performed a significant body of research in order to solve these technical problems and arive at a commercially viable fabrication route and sensor head structure which is optimized for electrochemical sensing applications. While embodiments of the invention have been described above in general terms with reference to Figure 1, further more specific worked examples are described in more detail below.

Reagents and Solutions: All solutions were prepared with Milli-Q water (Millipore Corp., UK) with a resistivity of 18.2 [2 at 25 °C, To characterize the pBDD electrodes, solutions of 0.] M KNO3 (Sigma Aldrich, UK) were used for solvent windows and capacitance experiments. For cyclic voltammetry (CV) measurements solutions contained mM hexaamineruthenium(III) chloride (Ru(NII-I3)6: Strem Chemicals, Newbury Port, USA) in 0.1 MKNO5.

Ma ferials: Microwave chemical vapor deposition (MW-CVD) polycrystalline insulating synthetic diamond was grown by Element 6 (1-larwell, Oxford, UK) in wafer form (25 mm diameter). Typically the substrates were 1 mm in thickness, mechanically lapped (polished) to a surface roughness of < 2 nm, and then subj ect to laser micromachining using a high power laser micromachiner (E-3551-1-3-ATHI-O, Oxford Lasers, UK), The resulting structures were acid cleaned in hot (--200°C) concentrated sulfuric acid saturated with KNO3, to remove any NDC produced during lasering (Wilson et al., J. Phys. Chem. B 2006, t to, 5639-5646). pBDD overgrowths were carried out using MW-CVD on the processed substrates, The approximate growth conditions used to prepare the appropriately doped "metal-like" pBDD films have been described previously (Balmer et al., J. Phys. Cond. Matt.

2009, 21, 364221). Electrical top contacts to pBDD contact pads on the all-diamond electrodes were made by sputtering a Ti/Au (10 nm / 300 nm) layer (Moorfields sputter system, UK) through a Kapton'TM tape (DuPont, UK) mask and annealing at 400°C for 5 hrs to produce an ohmic contact (Das et al., Thin Solid Films 1992, 212, 19-24). NDC (graphitized) contacts were made using the laser micromachiner (as described above).

Electrical conductivity measurements were carried out on x I x 0 mm pBDD bars (DIAFILM Electroanalytical grade -of a similar quality to the overgrowth pBDD -Element Six) on which four collinear contacts had been placed, either by graphitization and AgDAG (Silver conductive paint, RS, UK) or by Au/Ti sputtering. Material characterization of the wafers was carried out using optical microscopy (BH-2, Olympus, UK), white light interferometry (Contour GT-IC, Bruker, UK), field emission scanning electron microscopy (FE-SEM: Supra 55 VP, Zeiss), atomic force microscopy (AFN'l: Nano Enviroscope with Nanoscope IV controller, Bruker, UK) and Micro-Raman spectroscopy (Renishaw inVia Raman, Af laser at 514.5 nm excitation) using a xSO objective lens and a spot size of'-'-S pm.

Electrochemical measurernentc: Electrochemical characterization of the all diamond electrodes was performed in a standard three-electrode configuration using a Pt counter electrode and a reference electrode (saturated calomel electrode, SCE) for solvent window, capacitance and redox electrochemical analysis. CV was carried out using a potentiostat (CHI74Oa, CII Instruments Inc., USA). The back contacted BDD electrodes were typically mounted into the lid of a polypropylene Falcon tube (Fisher Scientific, UK) using silicone sealant 786 (Dow Corning, UK) or cast in epoxy (RX771CINC, Robnor, UK) such that the electrode bearing face was exposed to solution. For the top-contacted electrodes the accessible electrode area was defined by epoxy glue or KaptonTM tape. Current-voltage (resistance) curves were measured in air using a Keithley current source (Model 6220, Keithley).

All-Diamond Electrode Fabrication: The fabrication process employed to produce the all- diamond electrodes is shown in Figures 2 and 3. Figure 2(a) and Figure 3 show the step-by-step fabrication of an all-diamond electrode, The starting substrate, insulating polycrystalline diamond, was laser micromachined using a high power laser micromachiner to produce recessed structures "trenches" where the base geometry reflects the resulting electrode geometry. pBDD was overgrown into the lasered trenches and a co-planar structure revealed through multidirectional polishing of the overgrown surface using a resin-bonded scaife embedded with diamond micro-particles, 2 m -20 m in size, Using this process all-diamond macro-and microelectrode structures such as disks (cylindrical machined holes), bands (trenches) and ring electrodes (circular trenches) were produced, Typical trench depths, i.e. initial electrode depth, varied from ca. 30 jim to 00 jim, although after overgrowth and polishing the resulting pBDD electrode depth was reduced. Figure 2(b) shows 3D schematics of different all-diamond pBDD electrodes fabricated using the process illustrated in Figure 2(a) including: (0 dual band; (ii) triple band; (Hi) multiple bands; (iv) disk; (v, vi) ring-disk, Figures 2(b)(i, ii and vi) illustrate configurations containing contact pads for top electrical contacts, The minimum feature size is theoretically defined by the minimum laser spot diameter, which for this system is 6 jim (Smirnov et a],, Anal, Chem. 2011, 83, 7438-43; Kiran et al,, Sensors 2012, 12, 7669-81). However, the entry hole for the laser spot is 20 -30 jim in diameter as laser fluence (uniformity of energy density across beam diameter) and the ablation threshold of the material are critical. The heterogeneous nature of polycrvstalline diamond also means the laser ablation efficiency will vary between grains, It was found that the minimum repeatable feature size on polycrystalline diamond was limited to Ca. 50 rim jim with the laser system employed. Furthermore, as CVD diamond growth is dominated by surface kinetics, the bottom of narrow and deep channels, where transport of growth species is lower, are unlikely to be completely filled before lateral growth from the channel opening closes off the top of the channel. Hence to avoid sub-surface voids, aspect ratios (width/depth) of channels were typically limited to »= L Figures 2(c)(i-iii) show typical in-lens and secondary electron FE-SEM images of all-diamond electrode structures produced using the above procedure, including: (i) a triple individually addressable pBDD band electrode (band widths 90 tm, 64 jim and 460 jim); (ii) 1.02 mm diameter macrodisk electrode and; (iii) a ring-disk pBDD electrode (t.02 mm diameter disk, ring 1.10 mm inner diameter and 1.33 mm outer diameter). Also displayed in Figure 2(c)(iv) is an optical image of a top contacted ring disk electrode where the black electrode structures are clearly visible, A typical tapping mode AFM image recorded at the boundary between insulating and conducting diamond is shown in Figure 8 (and is discussed in more detail later) for two different pBDD band electrodes (of dimension 0.1 mm 10 mm and 1 mm 10 mm). A height line profile taken across the boundary reveals that the pBDD electrode was recessed by approximately 9 nm relative to the insulating region. A very slight recess is expected as the hardness of the pBDD lattice is partially compromised by the presence of boron and so higher doped regions polish slightly faster, This recess is minimal in size compared to those produced using alternative etch-based fabrication processes, which can be several hundreds of nm in size (Smirnov et al., Anal. Chem. 2011, 83, 7438-43; Kiran et al., Sensors 2012, 12, 7669-8 1). AFM topography scans recorded in either the pBDD or insulating only regions of the surface revealed similar surface roughness values (<2 nm) for both.

The polycrystalline nature of the pBDD is clearly visible in the FE-SEM images in Figure 2(c), with grains varying in size typically in the range 1 -50 jim. The contrast differences in the FE-SEM, most evident in the conducting regions of the surface, are associated with the non-uniform uptake of boron into different crystal facets in the polycrvstalline material (Patten et al., Angew. Chem. tnt, Ed. 2012, 51, 7002-7006). Interestingly, at the interface between the pBDD and insulating diamond, the grains appear to be generally aligned in a direction perpendicular to the interface (Figure 2(c)). For clarity, grain orientation is further highlighted in the FE-SEM images shown in Figure 9 and described in more detail later.

The orientation suggests that these pBDD grains grew laterally from the diamond grains of the side walls of the insulating recess, rather than vertically from grains at the bottom. For the smaller width band and ring structures e.g. Figures 2(c)(i) and 2(c)(iii), the grain structure indicates that the entire band is dominated by lateral side wall growth as opposed to the larger structures, such as the pBDD disk in Figure 2(c)(ii), which is composed of both lateral and vertical growth grains. Importantly, for structures grown in the regions where vertically growing pBDD meets lateral growth pBDD, in all bar one of the eighteen structures examined in this study, optical microscopy and FE-SEM revealed the absence of surface voids in the region of the overgrown pBDD. Furthermore, from lasered cross sections of all-diamond electrodes, e.g. as shown in Figure 4(a) and in Figure 9, no subsurface voids were observed, indicating that the grains intergrow.

Electrical contact to the all-diamond electrodes was achieved in two ways. The first was a top contacting Ti!Au method (see Experimental) and required the additional machining of recessed structures to form both BDD contact pads (Figures 2(b)(i) and 2(b)(ii)) and BDD electrical contact wires (Figure 2(b)(vi)). This approach was found suitable for electrode structures such as bands, where contact pads could be incorporated without compromising the geometry of the electrode, when employed in an electrochemical cell, However, in the case of other electrodes such as the individually addressable ring-disk electrode schematically presented in Figure 2(b)(vi), the top contact approach results in the ring structure being partially compromised (unable to form a complete ring). Furthermore, the electrical contact wires may also contribute to the electrochemical current.

Four point probe electrical measurements made using Ti/Au contacts (ii = 3) over the current range +1 mA at T = 25 °C revealed a linear current-voltage relationship (R1 > 0.99) and an overall resistance, R0j of 1.46 0 for the Ti/Au contacts. As: = where R80 is the electrically insulating resistance of the pBDD (resistivity = 0.45 mO m (Diafilm EA: Enabling new electroanalytical applications, www. e6.com!sensors, accessed 3 1/01/14), hence for the bar length and area dimensions employed R111 = =0.88 0) and R01101 is the contact resistance, Rco,caa was determined as 1.7 ± 0.12 mQ cm2 for the Ti/Au contacts.

A more attractive approach to top contacting, as illustrated in Figure 2(b)(v), is electrically back contacting which also means that the entire front face of the all-diamond electrode can be exposed to solution. It has previously been shown that laser micromachining insulating diamond produces a conductive material with electrical properties similar to graphite (Alemanno et al,, Diam,Rel, Mat, 20H, 38, 32-35). Back contacting was thus achieved by laser machining a blind hole just through to the rear side of the pBDD electrode, as shown pictorially in the FE-SEM image, Figure 4(a).

During lasering, a plasma is formed at the lasered face; rapid expansion during heating causes most of the diamond to be ablated from the hole in the gaseous oxidized form e.g. CO, CO2.

A small proportion is however left behind in the form of a black, conductive soot-like deposit, which forms the INIDC-conducting diamond contact. Note that most of this material can be removed by an aggressive acid clean if required e.g. when preparing the structure for a pBDD overgrowth. Blind holes were machined by removal of successive -30 pm thick layers of insulating diamond using a cross hatching approach, where the direction of the hatch was rotated by 60° with each layer. Rotational cross hatching was employed to ensure an even cut during lasering. White light interferometry was periodically employed to monitor the depth of the machined hole and to ensure that the cut extended to the bottomed of the pBDD-filled recess but did not penetrate through to the top-side of the electrode, as shown in Figure 4(b).

The use of optical grade diamond made location of the laser in the black pBDD regions on the transparent diamond face a facile process, as shown by the optical image in the inset to Figure 4(a). Note that opaque mechanical-grade insulating diamond could be still be used, however alignment marks are required to accurately position the laser. To determine the resistance of the lasered surface contact, an experiment equivalent to that described above for the Ti/Au contact was performed. R0111,1 (n = 3) was determined to be 2.5 + 1.3 mQ cm2. The low magnitude of Rc(mtact for both the Ti/Au and lasered NDC contacts implies negligible ohmic drop due to R01110.1, To form the final contact, the lasered back contact holes were filled with conducting Ag DAG, Cu multi-core wires inserted and a hard epoxy used to fix the wires in place. When back contacting structures containing multiple closely spaced electrodes, it is important to consider the size of the entrance hole and the separation of the electrodes. This is to ensure that overlapping entrance holes do not electrically connect two adjacent electrodes.

FE-SEM images of cross sectioned back contact holes revealed a tapered structure. For example, drilling 1 mm deep into insulating diamond with a 500 pm diameter entry hole resulted in a base diameter of300 jim. This is most likely due to sidewalls receiving a lower energy density due to the perpendicular orientation with respect to the beam direction and therefore machining less efficiently. The tapering of the back contact hole is clearly evident in the inset picture to Figure 4(a), showing multiple band structures individually addressed. It is evident from this image that that electrical shorting has not occurred during the contacting procedure.

BDD Characterization: The all-diamond electrodes where characterized by FE-SEM (as shown in Figure 2(c)), Raman spectroscopy, and electrochemistry i.e. solvent windows, capacitance and CV, the latter for the fast outer sphere one electron transfer redox species, Ru(INH3)o32t In total ii = 18 different electrodes were assessed, including ten band electrodes with width dimensions in the range 50 pm -1000 pm; the length was fixed at 10 mm; one dual band electrode (width dimensions 60 jim and 460 pm); one triple band electrode (width dimensions 64 jim, 90 jim and 460 jim); one disk electrode of diameter t.02 mm and two ring-disk electrodes of the dimensions given in Figure 2(c)(iii), Of the eighteen electrodes fabricated, only one was deemed unsuitable for further investigation due to an electrical contact failure; the NDC back contact had partially penetrated through to the front electrode face resulting in a significant increase in the measured capacitance and damage to the electrode surface.

Representative data is shown in Figure 5, for five individually addressable, NDC back-contacted pBDD band electrodes of widths 50 pm, 100 jim, 200 jim, 500 pm and 1000 jim (micro to macroscale), shown optically in the inset to Figure 5(a). Further characterization data is provided later for the ring and disk electrodes, respectively.

Micro-Raman spectroscopy (514.5 nm) is useful for ascertaining NDC and qualitatively assessing boron concentration (Fujishima et al., Diamond Electrochemistry, Elsevier: 2005), however the results must be treated with caution as micro-Raman will only sample a small area of the surface at any one time, spot size depends on optical magnification but typically for /53, spot size is 5 tm. Hence either micro-Raman NBC mapping must be undertaken, although this is a very lengthy procedure especially if the electrode is macrosized, or the data must be used in combination with electrochemical analysis.

Figure 5(a) shows typical Raman spectra, recorded on both high (black line) and lower doped (red line) facets of the 200 rim wide band electrode and is typical of those recorded on other bands and in other regions of the same electrode. The sp3 peak, which is observed at 1332 cm1 in undoped polycrystalline diamond, is clearly seen, although has shifted slightly in peak position and decreased in peak intensity, due to the high boron concentration. The asymmetry of the peak (Fano resonance) also reflects the high levels of boron in the lattice, where a Fano resonance is typically observed for boron dopant levels > 1020 B atoms cm3, which decreases in symmetry the higher the average boron content, Peaks present at 500 cm" and 1230 cm'1 are also indicative of the high doping levels. NBC is observed by the presence of broader peaks between 1400 cm" and 1600 cm" (Prawer et al, Phil. Trans. Royal Soc. A 2004, 362, 253 7-2565) and would be more likely to occur in the higher doped regions. We see no evidence of NDC in the Raman spectra recorded.

The most effective way of ascertaining NDC is to record CV solvent windows in background electrolyte. Electrochemical analysis samples the entire electrode area and in background electrolyte both capacitive and surface oxidation/reduction processes contribute to the observed CV response. In particular, non-faradaic oxidation of NDC results in an appreciable current flow just before water oxidation. Furthermore, although dissolved oxygen cannot be electrocatalytically reduced on a BDD surface, oxygen is sluggishly reduced in the presence of NDC, resulting in an appreciable cathodic current in the negative window. CVs were recorded at 100 mV s' in 0.1 lvi KNO3 and are shown for the five band electrodes in the inset to Figure 5(b). Importantly for all five electrodes, the window is featureless until water electrolysis takes place indicative of negligible NDC, For these electrodes solvent windows in the range 3,62 -3.78 V were recorded (where the anodic and cathodic potential limits are defined as the potential at which a current of 0.4 mA cm'2 is passed for water electrolysis) comparing well with that recorded previously for pBDD grown under similar conditions. The one electrode which contained a surface void, showed a notable difference in solvent window characteristics, compared to the other electrodes, as shown in Figure 0. Here a solvent window of 2.45 V was recorded and dear features were present in the CV, which correlate with the presence of NDC.

Capacitance measurements (double layer) were also used to assess (i) the quality of the pBDD synthesized and (ii) the seal between the pBDD and insulating diamond (poor seals lead to increased capacitance). Note an anomalously high capacitance reading could also indicate a poorly contacted electrode. For the five band electrodes double layer capacitances, Cdl, were determined by cycling at 100 mV s' over the potential range 0.08 V. Since; = (nv Cdl can be determined by measuring the charge, Q, at 0 V at a fixed scan rate, v. Cd! values 6.61 pF cm2 (50 pm band); 7.28 pF cm2 (100 pm band); 6.73 pF cm2 (200 pm band); 7.70 pF cm2 (500 pm band) and 5.26 pF cm2 (1 mm band) were obtained in line with that expected for the boron dopant levels employed during growth, Finally, to assess the electrochemical characteristics of the all-diamond electrodes, CV was typically performed at too mV s' in a solution containing 1 mM Ru(NIH3)53 and 0.1 M KNO3; more detailed studies involved varying the scan rate, typically in the range 10 mV s -500 mV s (as shown in ESI, section 1, for an all-diamond ring electrode). Figure 5(b) shows the CV characteristics, recorded at 100 mV s for the five band electrodes (solid lines) of different widths. COMSOL, as described in ESI, section 1, was employed to simulate the expected CV response for each electrode (dotted line), assuming diffusion is rate limiting.

The close agreement of the experimental and simulated data indicates that the pBDD electrodes are behaving as diffusion-limited "metal-like" electrodes, for their defined geometries. Characterization data for the all-diamond disk electrode is also shown in Figure 11.

Using a laser micromachining approach, where recessed structures can be machined into insulating diamond, then over-grown with pBDD and finally polished (lapped) flat to reveal a co-planar structure, it is possible to fabricate individually addressable all-diamond electrodes of any geometry. This approach is demonstrated herein with the fabrication of individually addressable all-diamond disks, ring-disk and macro-and micro-band dectrodes. We find that the smallest controllable electrode size is limited to 50 pm >< 50 pm, taking into account the technical limit of the laser micromachiner, the aspect ratio of the recessed structure produced and the ease at which pBDD can be grown into the recess, avoiding defects and NDC growth.

The resolution of this fabrication methodology could be increased further by employing improved lithographic and processing techniques.

Low resistance electrical contacting is possible using either Ti/Au top contacts or NDC back contacts, the latter increasing the range of electrode geometries which can be processed and the usability of the resulting electrode, especially in harsh and aggressive environments for long periods of time. The all-diamond electrodes are shown to exhibit the same characteristics of highly doped, negligible NDC content, pBDD grown in bulk form, thus demonstrating that the quality and performance capabilities of the pBDD have not been compromised via growth into the recessed structures. Note, the use of optical grade insulating diamond also paves the wave for combined electrochemical spectroscopic measurements using the same device.

Theoretical uzodeling: A COMSOL (COMSOL multiphysics, COMSOL, SWE) model was employed to simulate the electrochemical CV behavior of the various electrode geometries fabricated as described below.

Simulations were performed using the commercial finite element modelling package COMSOL 4.3b (COMSOL AB, Sweden) on a Desktop i7 2700 with 8 GB RAM. Briefly, the structure was defined in the appropriate geometry (disk, ring-disk or band) based upon optical and FE-SEM microscopy measurements of the different electrodes. Due to the different electrode geometries, two approaches were used in the simulations. A full 3D approach was used to simulate the band geometries to accurately represent the behaviour of the ends of the bands, Ring-disk electrodes, which have rotational symmetry, were simulated in a radially symmetric pseudo 3D space. A 2D plane through the radius of the disk was considered, where one boundary was defined as the axis of symmetry of the system. The simulation domains of the band and ring-disk electrodes are illustrated in Figure 6 which shows a schematic diagram of a 3D band system. The BDD electrode is labelled surface 1. The insulating diamond surface is labelled 2. The external surfaces of the solution that define the volume are labelled 3. For clarity, a 2D section has been taken through the 3D model, The identity of the faces is shown on the projected face on the right hand side. The radial model for the ring-disk is also shown, where the ring and disc are labelled a and Ib, respectively.

The inert diamond surface is boundary 2, the closed volume edge boundaries are labelled 3.

The ring-disk system has an additional boundary of axial symmetry that is labelled 4.

Typically 35O,UUU triangular mesh elements were used in the 3D band simulation and 75,OOO were used in the 2D axisymmetric ring-disk simulation with the greatest mesh resolution at the pBDD-insulating diamond boundary where there is -0.5 mesh elements per jim of electrode, Increasing the number of mesh elements was not found to significmtly change the results obtained.

The one electron reduction of Ru(NH3)63 species in solution was considered as follows

A

+ e RthH3)t where ki and k1, are the rate constants of electron transfer for the reduction and oxidation reactions respectively. Fick' s second law of diffusion was solved to describe diffusion firstly for the 2D axisymmetric system of the ring-disk electrodes iBc; - -= LJ.L_A __ _ r r and secondly in the 3D system of the band electrodes at av fli where c1 (mol cm'3) and D1 represent the concentration and the diffusion coefficient of species I (Ru(NH3)632), (x, y, z) and (r, z) represent the cardinal dimensions of the 3D and 2D axisymmetric systems, respectively, and / is the time. D1 was assumed to be identical for both oxidation states of the redox species (D = 8.8 x 106 cm2 1), The model was used to simulate the current at the pBDD electrode by solving the diffusion equation for the appropriate geometry subject to the boundary conditions summarized in the table below which includes the set of equations governing flux of species into and out of the domain: Surface Descnption Equation = - 1 Electrode = +

RT

= k.e RY 2 Inert diamond -ii N1 = 0 3 External faces c, = 4 Axis of symmetry = a _________ ______________________ Or In the above table, /c° represents the standard rate of reaction, ii is the number of electrons involved in the reaction (here n = 1), a is the charge transfer coefficient (assumed to be 0,5), 17 is the overpotential, which is defined as the applied potential minus the half-wave potential (-0.18 V), T is the temperature (298 K), F and R are the Faraday and Gas constants, respectively. For these simulations electron transfer was assumed to he sufficiently fast (k0 = 0.1 cm s') such that diffusion was rate limiting, in accordance with previous MET measurements of k° on similar dopant density BDD. For each electrode geometry the current, i, was simulated as a function of E for a given potential scan rate. The current is proportional to the flux through the electrode boundary for the 2D axi symmetric ring-disk system by . =1 n-dr and for the 3D band system by eee)ec = jJ -dxdy where and I are the current and the flux in the z dimension through the boundary, respectively, and Ye/cc and Ze/ec are the coordinates that define the electrode area.

Figure Sb and Figure 7 show typical experimental CV data matched to simulation. In particular, Figure 7 shows experimental (solid lines) and simulated (dashed lines) CVs for an all-diamond BDD ring electrode (as shown in Figure 2c)ivfl of 3.10 and 3.20 mm inner and outer diameters, respectively at potential scan rates of 10, 20, 30, 50, 00, 300, and 500 my s A FM analysis qf all-diamond electrodes Figures 8(ai) and 8(bi) show in-lens FE-SEM images of the boundary between the insulating diamond and pBDD for two different band electrodes of width 100 pm (a) and width 1000 p (b). Figures 8(aii) and 8(bii) show tapping mode AFM images recorded at 20 pm the conesponding area is also highlighted on the FE-SEM images, which were taken at an accelerating vohage of 2 kV and a working distance of 5 mm, Height profiles of the surface topography across the insulating diamond-pBDD boundary, as shown in Figures 8(aii) and 8(bii), reveal a height difference of 9 nm, Height profiles of the surface topography across grain boundaries in the pBDD region shown in Figures 8(aii) and 8(bii) indicate a height difference of only t-2 nm, fE-SEA] images of all-diamond electrodes A representative selection of FE-SEM images of all-diamond electrodes is shown in Figures 9(a) and 9(b). In Figure 9(a) an individually addressable dual band electrode of width 60 pm and 500 pm, separated by a 40 pm thick layer of electrically insulating diamond is shown, This structure is of use, for example, in generation-collection type electrochemical experiments. In Figure 9(b) the boundary region of the BDD 1.02 mm diameter macro electrode is shown, with the grains that grow laterally from the side wall of the insulating diamond trench highlighted. Lateral grain growth can also be seen on each of the bands shown in Figure 9(a), where the smaller band is entirely composed of laterally grown grains whereas the larger band is composed of both laterally and vertically growing grains.

fE-SEA'] and solvent window characterization of a stir/ace void Of the 18 electrodes assessed in this study, only one was found to contain a defect, as shown in Figure 10. This defect, at the electrode-insulator boundary, is likely to have occurred during overgrowth due to the grains being unable to properly inter-grow at this location, Lapping of the overgrown pBDD to reveal the co-planar electrode also revealed the defect, -300 pm'< 50 pm in size, positioned along the inner perimeter of the ring electrode. In addition to a defect itself, there is evidence of aberrant grain structure, typified by many small grains, in the region around the defect (inset of Figure 10(a)). The presence of the smaller grains, in addition to the larger defect, supports the conclusion that aberrant grain growth has occurred in this region. Furthermore, a relatively narrow solvent window, 2.45 V, was observed for this electrode, in addition to peaks in the CV indicative of NDC presence, as well as NDC catalysed oxygen reduction. These data indicate that the quality of the electrode was compromised by defect presence in the surface.

Electrochernical Characterization qf an A li-Diamond Disk Electrode As illustrated in Figure ii, a disk electrode of 1.02 mm diameter was electrochemically 3-4-.

charactensed by: (a) CV with 1 mM Ru(NH3)( in 0.] M KNO3; (b) CV solvent windows in 0.1 lvi KNO3 at 100 mV S1 and by Raman spectroscopy as shown in the inset to Figure 11(b). The theoretical i, for a 1.02 mm diameter electrode, in accordance with Randles Sevcik, is 0,252 mA cm2 (assuming.1) = 8.8 >< 106 cm2 1) which is close to the experimentally measured /,, of 0.276 mA cm* The solvent window is 3.7 V and shows no evidence of 02 reduction or NDC oxidation peaks. Finally the Raman, which is representative of six spectra recorded on this electrode, shows no evidence of NDC in either light or dark grains.

In summary, the present specification described how to fabricate optimized electrochemical sensor structures fabricated from diamond materials via techniques involving insulating diamond substrate patterning, boron doped diamond overgrowth, and back contacting techniques. Using the methodology as described herein the following advantageous features can be achieved: (i) suitable geometries and CVD diamond growth parameters to obtain filling of the trenches with boron doped diamond without the formation of voids and also to ensure that such filling can be achieved with high quality boron doped diamond material optimized for electrochemi cal sensing applications; (ii) suitable geometries and diamond processing techniques to achieve back-contacting of boron doped diamond electrodes through an insulating diamond support matrix without either falling short of the rear surface of the thin boron doped diamond electrodes or otherwise over-shooting the rear surface and drilling through to a front sensing surface of the diamond sensor head; and (iii)fabrication of good ohmic contacts on a rear surface of the boron doped diamond electrodes within a via though an insulating diamond support matrix and particularly ohmic contacts which have a low resistance, a low capacitance, and a uniform and reproducible geometry and functional performance.

While this invention has been particu'arly shown and described with reference to a number of embodiments, it will be understood to those skilled in the art that various changes in form and detail may be made without departing from the scope of the invention as defined by the appending claims,

Claims (10)

1. Claims A diamond electrochemical sensor head comprising: a planar sensing surface; a rear surface through which electrical connections are provided; one or more boron doped diamond electrodes which are disposed within trenches in an electrically insulating diamond support matrix at the planar sensing surface, the one or more boron doped diamond electrodes extending partially through the electrically insulating diamond support matrix from the planar sensing surface towards the rear surface of the electrically insulating diamond support matrix; one or more vias extending from the rear surface of the electrically insulating diamond support matrix to a rear surface of the one or more boron doped diamond electrodes within the electrically insulating diamond support matrix; one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix; and one or more electrical connectors extending through the one or more vias to the one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix, wherein the diamond electrochemical sensor head has a thickness from the planar sensing surface in which the one or more boron doped diamond electrodes are disposed to the rear surface of the electrically insulating diamond support matrix in a range 50 micrometres to I.5 millimetres, wherein the one or more boron doped diamond electrodes extend through the electrically insulating diamond support matrix from the planar sensing surface towards the rear surface of the electrically insulating diamond support matrix with a depth in a range 20 micrometres to 500 micrometres, and wherein the one or more ohmic contacts within the one or more vias on the rear surface of the one or more boron doped diamond electrodes each have a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than 10 mV, where the ohmic drop is defined by I x R with I being current and R being total resistance.
2. 2. A diamond electrochemical sensor head according to claim, wherein the thickness from the planar sensing surface in which the one or more boron doped diamond electrodes are disposed to the rear surface of the electrically insulating diamond support matrix is no less than 75 micrometres, 100 micrometres, 150 micrometres, or 200 micrometres.
3. 3, A diamond electrochemical sensor head according to claim 1 or 2, wherein the thickness from the planar sensing surface in which the one or more boron doped diamond electrodes are disposed to the rear surface of the electrically insulating diamond support matrix is no more than 1 millimetre, 750 micrometres, 500 micrometres, 300 micrometres, or 200 micrometres.
4. 4. A diamond &ectrochemical sensor head according to any preceding claim, wherein the depth of the one or more boron doped diamond electrodes is no more than 400 micrometres, 300 micrometres, 200 micrometres, or 100 micrometres.
5. 5. A diamond &ectrochemical sensor head according to any preceding claim, wherein the depth of the one or more boron doped diamond electrodes is no less than micrometres, 40 micrometres, or 50 micrometres.
6. 6. A diamond electrochemical sensor head according to any preceding claim, wherein the one or more boron doped diamond electrodes each have a front sensing surface roughness Ra ofno more than 100 nm, 75 nm, 50 nm, 20 nm, or 10 nm,
7. 7. A diamond electrochemical sensor head according to any preceding claim, wherein an interface region between the one or more boron doped diamond electrodes and the electrically insulating diamond support matrix comprises no voids having a largest lateral dimension greater than 10 micrometres, 5 micrometres, I micrometres, 500 nanometres, 300 nanometres, or 100 nanometres,
8. 8. A diamond electrochemical sensor head according to any preceding claim, wherein the one or more boron doped diamond electrodes are formed of boron doped diamond material having the following characteristics: a solvent window meeting one or both of the following criteria as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode in a solution containing only deionised water and 0.1 M KNO3 as a supporting electrolyte at pH 6: the solvent window extends over a potential range of at least 4.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 38 mA cm'2; and the solvent window extends over a potential range of at least 3.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 0.4 mA cm'2; a peak-to-peak separation AE (for a macroelectrode) or a quartile potential AE314,114 (for a microelectrode) of no more than 70 mV as measured by sweeping a potential of the boron doped synthetic diamond material at a rate of 100 mV with respect to a saturated calomel reference electrode in a solution containing only deionised water, 0.1 M KNO3 supporting electrolyte, arid 1 mM of FcTh4A or Ru(NH2)63 at pH 6; and a capacitance of no more than 10 RE cm'2 as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode between 70 my and -70 mY in a solution containing only deionised water and 0.1 M KNO3 supporting electrolyte at pH 6, measuring resultant current, subtracting a current value at 0 V when sweeping towards negative potentials from a current value at 0 V when sweeping towards positive potentials, dividing the subtracted current value by 2, and then dividing the result by an area (cm2) of the boron doped synthetic diamond material and by a rate at which the potential is swept (Vs'1) to give a value for capacitance in F cm'2,
9. 9. A diamond electrochemical sensor head according to any preceding claim, wherein the resistance of each of the one or more ohmic contacts within the vias on the rear surface of the one or more boron doped diamond electrodes is sufficiently low that the ohmic drop in the faradaic electrochemical experiment is no greater than 8 my, 6 my, 4 my, 2 my, or] my,
10. 10. A diamond electrochemical sensor head according to any preceding claim, wherein the one or more ohmic contacts within the vias on the rear surface of the one or more boron doped diamond electrodes each have a resistmce per unit area of no more than 0 cm'2, 5 0 cm'2, or 1 0 cm'2, 1], A diamond electrochemical sensor head according to any preceding claim, wherein the one or more ohmic contacts are formed of non-diamond-carbon.12, A method of fabricating a diamond electrochemical sensor head as previously claimed, the method comprising: starting with an electrically insulating diamond substrate having planar front and rear surfaces and a thickness between said planar front and rear surfaces in a range 50 micrometres to 1.5 millimetres; cutting one or more trenches in the planar front surface of the electrically insulating diamond substrate, wherein the one or more trenches have a depth in a range 20 micrometres to 500 micrometres; growing boron doped diamond material over the front surface of the electrically insulating diamond substrate and into the one or more trenches; processing back the boron doped diamond material over the planar front surface of the electrically insulating diamond substrate to form a planar sensing surface comprising one or more boron doped diamond electrodes surrounded by an electrically insulating diamond support matrix, the planar sensing surface having a surface roughness Ra less than 100 nm after processing, the one or more boron doped diamond electrodes extending through the electrically insulating diamond support matrix from the planar sensing surface towards a rear surface of the electrically insulating diamond support matrix with a depth in a range 20 micrometres to 500 micrometres, and wherein a distance between the planar sensing surface and the rear surface of the electrically insulating diamond support matrix is in a range 50 micrometres to 1.5 millimetres; forming one or more vias extending from the rear surface of the electrically insulating diamond support matrix to the rear surface of the one or more boron doped diamond electrodes within the electrically insulating diamond support matrix; forming one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix, wherein the one or more ohmic contacts within the one or more vias on the rear surface of the one or more boron doped diamond electrodes each have a resistance which is sufficiently low that an ohmic drop in a faradaic electrochemical experiment is no greater than tO mV, where the ohmic drop is defined by Ix R with I being current and R being total resistance; and forming one or more electrical connectors extending through the one or more vias to the one or more ohmic contacts disposed on the rear surface of the one or more boron doped diamond electrodes within the vias in the electrically insulating diamond support matrix.13. A method according to claim 12, wherein, during the step of growing boron doped diamond material over the front surface of the electrically insulating diamond substrate and into the one or more trenches, growth conditions are controlled to ensure that the one or more trenches are filled and to ensure that the boron doped diamond material filling the one or more trenches has the following characteristics: a solvent window meeting one or both of the following criteria as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode in a solution containing only deionised water and 0.1 M KNO3 as a supporting electrolyte at pH 6: the solvent window extends over a potential range of at least 4.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 38 mA cm2; and the solvent window extends over a potential range of at least 3.0 V wherein end points of the potential range for the solvent window are defined when anodic and cathodic current density measured at the boron doped synthetic diamond material reaches 0.4 mA cm2; a peak-to-peak separation AE (for a macroelectrode) or a quartile potential AE314114 (for a microelectrode) of no more than 70 mY as measured by sweeping a potential of the boron doped synthetic diamond material at a rate of 100 mY s with respect to a saturated calomel reference electrode in a solution containing only deionised water, 01 M KNO3 + 3+ supporting electrolyte, arid 1 mM of FcTMA or Ru(NH3)6 at pH 6; and a capacitance of no more than 10 RE cm2 as measured by sweeping a potential of the boron doped synthetic diamond material with respect to a saturated calomel reference electrode between 70 mV and -70 mY in a solution containing only deionised water and 0. I M KNO3 supporting electrolyte at pH 6, measuring resultant current, subtracting a current value at 0 V when sweeping towards negative potentials from a current value at 0 V when sweeping towards positive potentials, dividing the subtracted current value by 2, and then dividing the result by an area (cm2) of the boron doped synthetic diamond material arid by a rate at which the potential is swept (Vs1) to give a value for capacitance in F cm2.14. A method according to claim 12 or 13, wherein the one or more vias are formed using a laser cutting technique.15. A method according to claim N, wherein the laser cutting technique uses a pulsed laser beam.16. A method according to any one of claims 12 to 15, wherein the one or more ohmic contacts comprise non-diamond-carbon formed by a cutting technique used to form the one or more vias.17. A method according to claim 16, wherein a proportion of the non-diamond-carbon formed by the cutting technique is removed from the one or more vias prior to forming one or more electrical connectors.