WO2016090407A1 - Cuvette for optical spectroscopy - Google Patents

Abstract

A cuvette for optical spectroscopy comprising a substrate and a wicking structure in a spectral region on the substrate, the wicking structure comprising an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height in the spectral region, the wicking structure configured such that interpillar spacings between adjacent pillars in the spectral region are filled with an analyte solution when the solution is placed in contact with part of the wicking structure to thereby form a spectral sample of the analyte solution in the interpillar spacings that is suitable for optical spectroscopy.

Description

CUVETTE FOR OPTICAL SPECTROSCOPY

PRIORITY DOCUMENT

[0001 ] The present application claims priority from Australian Provisional Patent Application No.

2014905004 titled "CUVETTE FOR OPTICAL SPECTROSCOPY" and filed on 10 December 2014, the content of which is hereby incorporated by reference in its entirety.

TECHNICAL FIELD

[0002] The present disclosure relates to cuvettes for use in spectroscopy, such as UV or visible spectroscopy.

BACKGROUND

[0003 ] Optical spectroscopy is used to measure the absorption, fluorescence, phosphorescence, scattering, emission or chemiluminescence of liquid samples in various research and analytical fields. For example, in absorption spectroscopy the optical absorption spectra of liquid substances and mixtures are measured. Optical spectroscopy offers a fast and reliable analytical technique and can now be achieved using compact and relatively inexpensive instruments in remote locations.

[0004] In a simple spectrophotometer, the sample substance which is to be studied, usually a liquid, is placed into a transparent container, generally known as a cuvette or sample cell. Collimated light of a known wavelength (e.g. ultraviolet, infrared, visible, etc.) and intensity I₀, is incident on one side of the cuvette. A detector, which measures the intensity of the exiting light, I, is placed on the opposite side of the cuvette (in the case of transmission spectroscopy) or on the same side of the cuvette (in the case of reflection spectroscopy). Independent of the cuvette design, the measured absorbance, A, increases linearly with analyte concentration, [X], over a specific concentration range according to the Beer- Lambert relationship:

A = -\og ( ) = s - h - [X] ( 1 ) where 7₀ and / are the intensities of incident and transmitted light (7- is the transmittance 7), ε is the molar

'0

absorptivity, and h is the path length of the sample. Path lengths (h) in cuvettes are normally between 1 and 100 mm, typically 10mm or 2 mm. At high concentration, the relationship A <x [X] departs from linearity until the absorbance reaches a plateau. This limit is influenced by the magnitude of ε, which varies greatly for different analytes. In some cases ε is very large, such as for example for bloodj l ], chlorophyll[2,3], inks [4] and sensitizers[5]. Wherever the magnitude of ε and/or [X] is large, it is necessary to reduce h and/or dilute the sample before measurement. The latter is highly undesirable, as it is laborious, time-consuming, and a possible source of error, particularly where sample preparation is remotely carried out by unskilled personnel.

[0005 J Furthermore, these spectrometric techniques are limited for valuable samples (e.g. biological samples) because the cuvettes typically used generally require a minimum of 1 mL sample, which is typically discarded after measurement. Large sample volume and loss is problematic for valuable biological samples which may be present in limited quantities.

[0006] For analysis of fluids of limited volume (e.g. less than 10 μΐ_^) or fluids of high value (e.g.

biological samples which may be present in limited quantities), it is desirable to keep the sample volume required for spectroscopic analysis low. As sample volume decreases, it becomes more difficult to apply a sample into a measurement area of a cuvette without generating air bubbles, or sectioned liquid spots, leading to errors. Micro-cuvettes can be used for microlitre samples, but they are not easy to use because they can be difficult to fill and rinse. Known cuvettes that use capillarity for filling also suffer from difficulties with rinsing. Other cuvettes use injection of liquid, which allows rinsing, but requires filling of a syringe and displacement of connecting tubing with dead volume. So-called "cuvetteless sampling" is an alternative technique that has been used with very small sample volumes. In cuvetteless sampling, a narrow beam of light is directed to a sample stage that consists of a 1 to 2 xL liquid droplet suspended between two multi-mode optical fibres, one source-side fibre which provides light from a light source to the droplet and a detection-side fibre that guides light from the droplet to appropriate detection optics. The close proximity between the source-side and detection-side fibres allows enough of the light cone emanating from the source-side fibre to be collected by the detection-side fibre after passing through a liquid sample. These cuvetteless instruments typically require a clamping surface that can be wetted with sample to avoid an air-bubble interface. However, adding a small amount of sample (typically 5

to the centre of the clamping surface is a complicated lab technique and carry-over contamination resulting from failure to completely remove previous samples is a frequent source for error.

10007] There is a need for reliable techniques for the analysis of very small volume samples using various methods of optical spectroscopy. This is particularly true in the case of biochemical analysis, pharmaceutical screening, forensics and medical diagnostics.

[0008] Furthermore, the demand for fast, straight-forward, and inexpensive analytical methods for unskilled users is growing rapidly, driven by the convergence of innovation in portable electronics, data analysis, and microfluidics/lab-on-a-chip technology. Taking the laboratory to the sample in a compact and inexpensive device has major advantages, including short analysis-to-action turn-around times. Applications in security[6], health monitoring[7], environmental detection[8], and process control in industry [9] are a few examples where rapid on-site analysis is highly desirable. In these examples, sample preparation must not be laborious or time-consuming and conventional laboratory tasks, such as dilution, are a source of error and often impractical.

[0009 ] There is a need for cuvettes that are suitable for use in dilution-free spectroscopic analysis of small volume samples, such as high ε or [X] samples or valuable samples.

SUMMARY

[0010 ] We have developed a cuvette that permits dilution-free spectroscopic analysis of samples with high molar absorptivity, such as pigments and high concentration metal solutions, such as

hexachloroplatinate (PtCl₆ ^2~), potassium permanganate, methyl orange, and food dyes.

[001 1 ] According to a first aspect, there is provided a cuvette for optical spectroscopy comprising a substrate and a wicking structure in a spectral region on the substrate, the wicking structure comprising an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height in the spectral region, the wicking structure configured such that interpillar spacings between adjacent pillars in the spectral region are filled with an analyte solution when the solution is placed in contact with part of the wicking structure to thereby form a spectral sample of the analyte solution in the interpillar spacings that is suitable for optical spectroscopy.

[0012 ] Specifically, we have developed a pillar based cuvette that is spontaneously and precisely filled by capillarity to yield a reproducibl e optical path length of tens of micrometres. Only one droplet (~ 2 μί) of analyte solution is required. The cuvette is also open which allows for quick and easy rinsing between samples. Advantageously, the array of pillars also acts as a filter to remove particulates from the analyte solution as it fills the interpillar spacings.

[00131 In certain embodiments of the first aspect, the pillars and/or the substrate are optically transparent in the spectral region. In these embodiments, the cuvette is particularly suitable for transmission spectroscopy.

[0014 ] According to a second aspect, there is provided a cuvette when used in an optical spectrometer, the cuvette comprising a substrate and a wicking structure in a spectral region on the substrate, the wicking structure comprising an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height in the spectral region, the wicking structure configured such that interpillar spacings between adjacent pillars in the spectral region are filled with an analyte solution when the solution is placed in contact with part of the wicking structure to thereby form a spectral sample of the analyte solution in the interpillar spacings that is suitable for optical specti oscopy. [0015] In certain embodiments of the first and second aspects, the substrate comprises an analyte solution loading region on the substrate, the analyte solution loading region configured to receive a droplet of the analyte solution such that the droplet is in fluid contact with the wicking structure such that the analyte solution is dispersed within the wicking structure.

[0016] In certain embodiments of the first and second aspects, the pillars in the array in the spectral region of the substrate extend substantially orthogonally from the surface of the substrate.

[0017] The pillars can be any shape in cross section, such as square, round, elliptical, hexagonal, etc. In certain embodiments, the pillars are circular in cross section. In other certain embodiments, the pillars are square in cross section.

[0018 ] The array of pillars may be an ordered array in which the dimensions of the pillars and hence the interpillar spacings are substantially uniform in the array or, alternatively, the array may be a disordered array in which the dimensions other than the height of the pillars and hence the interpillar spacings are non-uniform in the array.

[0019 ] In certain embodiments of the first and second aspects, the height of each pillar is from about 0. 1 micrometres to about 500 micrometres, such as from about 0.1 micrometres to about 100 micrometres. In specific embodiments, the height of each pillar is about 10 micrometres. In other specific embodiments, the height of each pillar is about 20 micrometres.

[0020 ] In certain embodiments of the first and second aspects in which the pillars are circular in cross section, the diameter of each pillar is from about 0.1 micrometres to about 500 micrometres. In specific embodiments, the diameter of each pillar is about 10 micrometres.

[0021 ] In certain embodiments of the first and second aspects, the distance between adjacent pillars is from about 0.1 micrometres to about 250 micrometres.

[0022 ] In certain embodiments of the first and second aspects, the surface area fraction of the pillars is from about 0.01 to about 0.90, such as about 0.05, about 0.10, about 0.20 or about 0.30.

[0023 ] Optionally, a surface of the substrate and/or pillars that comes into contact with the analyte solution further comprises a functional coating that is able to react or bind with an analyte or release a species to generate or enhance an optical signal.

[0024] Advantageously, the dimensions of the interpillar spacings may be suitable for filtering particulates from the analyte sample. [0025] In certain embodiments of the first and second aspects, the cuvette is suitable for use in transmission spectroscopy and the pillars and/or the substrate is/are optically transparent in the spectral region so that light from a light source on one side of the cuvette is transmitted through the cuvette (and the sample) and the transmitted light is detected by a detector on another side of the cuvette. In these embodiments, the cuvette may for example be used in an optical spectrometer with incident light directed parallel to the pillars. In these embodiments, the optical path length is defined by the height of the pillars in the spectral region. However, the cuvette is not limited to that specific use and can also be used in reflection spectroscopy in which the light source and the detector of the spectrometer are each on the same side of the cuvette. In either case, the angle of incident and transmitted light may be from 0 degrees to 90 degrees.

[0026] For use in optical transmission spectroscopy, the pillars do not need to be optically transparent and they can be semi-transparent or opaque but the substrate will usually be transparent.

10027] The incident light could also be directed orthogonal to the pillars.

[0028] In certain embodiments, the dimensions of the interpillar spacings are suitable for filtering particulates from the analyte sample. In this way, a drop of an analyte sample containing particulate material can be placed on the analyte solution loading region with the droplet in fluid contact with the wicking structure and the analyte solution is allowed to disperse into the wicking structure but at least some of the particulate material is unable to pass through the interpillar spacings and is, therefore, filtered from the sample in the spectral region of the cuvette.

[0029] According to a third aspect, there is provided an optical spectroscopic method comprising: providing a cuvette according to the first aspect; contacting a droplet of an analyte solution to be analysed by optical spectroscopy with part of the wicking structure of the cuvette under conditions to fill the interpillar spacings in the spectral region with the analyte solution to provide a filled cuvette; exposing a first side of the filled cuvette to a light source such that an incident light beam is projected through the spectral region of the cuvette to a detector located at a second side of the filled cuvette that is opposite the first side, or exposing a first side of the filled cuvette to a light source such that an incident light beam passes through the analyte sample in the filled cuvette and is reflected from the substrate to a detector located at the first side of the cuvette; and measuring at least one parameter of the light at the detector.

10030] According to a fourth aspect, there is provided a process for producing a cuvette for optical spectroscopy, the process comprising: providing a substrate; forming an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height; and forming analyte solution loading region on the substrate, the analyte solution loading region configured to receive a droplet of an analyte solution such that the droplet is in fluid contact with the array of pillars such that the analyte solution is dispersed within the array of pillars.

[0031 ] In certain embodiments of the fourth aspect, the array of pillars is formed by forming a template on the surface of the substrate to give a templated substrate; etching the templated substrate under conditions to form an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height.

[0032] In these embodiments, the templated substrate can be formed by coating a surface of the substrate with a photoresist layer, exposing the photoresist layer to UV radiation through a patterned photomask to form a patterned template on the surface of the substrate.

[0033] Pillar arrays could also be directly stamped or moulded into a substrate, or a photoresist could be directly used as the pillars.

BRIEF DESCRIPTION OF THE FIGURES

[0034] Embodiments of the present disclosure will be discussed with reference to the accompanying figures wherein:

[0035] Figure 1 shows: (a) a plot of critical contact angle predicted by Equation 1 against projected area fraction pillars with heights from 5 to 50 μτη; (b-c) experimental wicking results for (b) h_n= 10 μιη and (c) h_n= 20 μηι. The crosses and circles represent wicking and non-wicking events. The solid line is calculated using Equation 2, see frame (a) also; [0036] Figure 2 shows: cross-section of the pillar cuvette loaded with the reference solution and the sample shown separate for clarity;

10037] Figure 3 shows scanning electron microscopy images of 20 μιη high pillars for φ = 0.05 (i), 0. 10 (ii), 0.20 (iii), and 0.30 (iv) array. Etching defects appear between the pillars as the lattice spacing decreases;

10038] Figure 4 shows (a) UV-Vis measurement method using the pillar cuvette; and (b) an image showing a droplet of aqueous sample in contact with a pillar array, in which a thin film (h_n = 10 μηι) has spontaneously formed. To the left is an empty pillar array for comparison;

[0039] Figure 5 shows the results of optical profilometry of the liquid (aqueous sample)-vapour interface at the tops of the pillars for h_n = 10 mm: (a) φ = 0.05 and (b) φ = 0.20. The curvature of the meniscus between the pil l ars is negligible in both cases; (c) Profiles of pillar arrays before and after filling with the liquid sample. The scale bar is 20 mm;

[0040] Figure 6 shows absorbance spectra for (a) pillar cuvettes at different φ (area fraction) with and without 0.5 M HCl(aq), and (b) for solvents with different refractive indices;

[0041 ] Figure 7 shows absorbance versus time for PtCl₆ ^2~ _(a(]) concentrations from 0.5 to 10 g/L measured in a pillar cuvette (φ = 0.05; h_n = 13.3 mm);

[0042] Figure 8 shows plots of absorbance against PtCl₆ ² _(aq) concentration measured in cuvettes with opaque pillars (see Table 1): (a) absorbance, A, and (b) absorbance normalized by the path length, A*;

[0043] Figure 9 shows plots of (a) absorbance, A, and (b) absorbance normalized to the path length, A*, at 259 nm for PtCl₆ ² _(aq) standard solutions in pillar cuvettes with semitransparent pillars; and

[0044] Figure 10 shows (a) UV-Vis spectra of a 30 ppm PtCl₆ ^2" solution collected using a conventional 2 mm cuvette, with (solid line) and without (dashed line) 6 mm polystyrene spheres present. The particles distort the spectrum greatly, (b) UV-Vis spectra of a 3 g/L PtCl₆ ^2" solution collected using a pillar cuvette (h = 5 μηι; φ = 0.40) with (solid line) and without (dashed line) 6 mm polystyrene particles present. The very good agreement shows that particles do not interfere with the analysis because the particles are filtered in-situ by the pillar array during cuvette filling, see frame (c). (c) Microscopy showing the 6 μιη particles (green spheres to the left hand side) being filtered during filling of the pillar cuvette. The particle-free liquid film used for spectroscopy is located on the right hand side. The scale bar is 28 mm. DESCRIPTION OF EMBODIMENTS

[0045 ] Provided herein is a cuvette for optical spectroscopy. As used herein, the term "cuvette" means a device designed to hold samples for spectroscopic experiments. Cuvettes are also commonly referred to as "cells" or "sample cells" and the term cuvette used herein is intended to include within its scope cells and sample cells. Prior art cuvettes are typically in the form of small tubes that are circular or square in cross section and sealed at one end to contain a fluid sample to be analysed. These cuvettes are nonnally optically transparent on opposite sides so that a single beam of light is able to pass therethrough when used for transmission spectroscopy.

[0046] In contrast to prior art cuvettes and with reference to Figures 2 and 3, the cuvette 10 described herein comprises a substrate 12 and a wi eking structure 14 forming a spectral region 16 on the substrate 12. The wicking structure 14 comprises an array of pillars 18 extending from a planar surface of the substrate with each pillar 18 in the array being of substantially equal height in the spectral region 16. The wicking structure 14 is configured such that the interpillar spacings 20 are filled with an analyte solution 22 by capillary action when the solution is placed in contact with part of the wicking structure to thereby fonn a spectral sample of the analyte solution in the interpillar spacings suitable for optical spectroscopy.

10047] The cuvettes described herein can be used for any form of optical spectroscopy, including "reflectivity", "absorbance", "transmission", and "fluorescence" spectroscopy. For ease of description, further reference will be made to the use of the cuvettes in UV and visible (UV-Vis) spectroscopy.

However, it will be appreciated that the cuvettes are not limited to that particular spectroscopic use and they can equally be used in other spectroscopic techniques.

[0048 ] The substrate and pillars of the cuvette can be formed from any suitable material. For use in optical transmission spectroscopy, the pillars can be optically transparent, semi-transparent or opaque and the substrate will usually be transparent. Transparent pillars and/or substrates can be formed from an optically transparent material and suitable materials for this purpose include plastics such as

Thermanox™ vinyl and cellulose acetate or glasses such as borosilicate glass, fused silica, quartz, soda lime, and silicate glass. Alternatively, the substrate could be chosen to reflect incident light back to a detector that is located on the same side of the cuvette as the light source. In these cases, the substrate may be any reflective material, such as a silicon wafer, metal or metal-coated polymer or glass, or other material with a reflective coating.

[0049] For opaque pillars, an optically transparent substrate including plastics such as Thermanox™ vinyl and cellulose acetate or glasses such as borosilicate glass, fused silica, quartz, soda lime, and silicate glass can be first coated with a thick layer of an opaque material and the pillars then formed by etching. Suitable opaque materials include any opaque metal, such as Cr, Pt, Au, Ti, Ti₂0, Ag and/or Cu that can be sputter coated onto the substrate. An opaque polymer that can be sputtered could be used.

Alternatively, still, the pillars can be painted or stamped with any opaque paint or ink.

[0050] Optionally, a surface of the substrate and/or pillars that is in contact with the analyte solution may have a functional coating that is able to react, bind or release a species to generate or enhance an optical signal. For example, a coating comprising an antibody, protein, peptide, small molecule, metal ion, metal complex or the like may be formed on the surface of the pillars and substrate and may bind a molecule or analyte of interest in the analyte solution, which binding is then qualitatively or quantitatively detected using optical spectroscopy. For example, the cuvette may be used to measure antibody-antigen binding wherein the analyte solution is suspected of containing either the antibody or the antigen. The binding can then be measured or detected by absorption, transmittance, reflection or fluorescence spectroscopy. The analyte of interest can be any atom, chemical, molecule, compound, composition or aggregate of interest for detection and/or identification, such as an amino acid, peptide, polypeptide, protein, glycoprotein, lipoprotein, nucleoside, nucleotide, oligonucleotide, nucleic acid, sugar, carbohydrate, oligosaccharide, polysaccharide, fatty acid, lipid, hormone, metabolite, cytokine, chemokine, receptor, neurotransmitter, antigen, allergen, antibody, substrate, metabolite, cofactor, inhibitor, drug, pharmaceutical, nutrient, prion, toxin, poison, explosive, pesticide, chemical warfare agent, biohazardous agent, radioisotope, vitamin, heterocyclic aromatic compound, carcinogen, mutagen, narcotic, amphetamine, barbiturate, hallucinogen, waste product and/or contaminant. The coating can be formed on the substrate and/or pillars by any suitable method, such as physical adsorption, self-assembly, chemical reaction on the surface from solution or vapour, plasma deposition, atomic layer deposition, electrostatic adsorption, etc. For example, a quartz substrate can be chemically modified by silanization using an alkylsilane in which the tail of the silane contains the appropriate chemical/biological functionality.

[0051 ] The cuvette described herein takes advantage of 'wicking', which is the spontaneous imbibition of liquid into a small porous structure via capillarity. [1 1 ] Taking into account the surface energy and geometry of the solid-liquid, solid-vapour, and liquid-vapour interfaces involved and ignoring gravity (assuming that the parent droplet is smaller than the capillary length), wicking can be thermodynamicaHy predicted for regular arrays of pillars, such as those found in the cuvettes described herein:

[0052] In Equation 2, Θ is the contact angle measured on a flat substrate of the same material, or 'material contact angle', r is the Wenzel[12] roughness factor (e.g., r = 1 +— W 40 for cylindrical pillars), and φ is the projected surface area fraction of the pillars (e.g., φ = for cylindrical pillars). [1 1 ] In practice, Equation 2 is only able to inform when wicking is not possible, cosO < ^ , as pinning effects are caused by energy barriers that oppose the wicking effect, as described elsewhere[ 13 |. For pillars with square and circular cross-sections, these pinning effects shift the wicking criterion to larger ^ and and, unexpectedly, wicking is prevented again for square cross-section pillars at very high area fraction <p.[ 13] For these reasons, the pillars in the ordered array are preferably cylindrical pillars of uniform dimensions.

[0053] Clean quartz is completely wetted by water and most other solvents in air, such that the material contact angle is intrinsically small. However, high-energy surfaces are readily contaminated in air, which raises the material contact angle and could prevent wicking altogether in practical applications. Figure 1 (a) plots the critical contact angle predicted by Equation 2 against the area fraction of the pillars for different pillar heights. Note that wicking is encouraged by increasing the height of the pillars.

Experimental results for water on quartz pillar cuvettes exposed to ambient laboratory conditions or, in selected cases, hydrophobization are shown in Figure l (b-c). The results are in qualitative agreement with previous observations by Semprebon et al.[ 13] who showed that pinning plays a major role in determining the critical contact angle for wicking. To compensate for pinning effects, lower contact angles than those predicted by Equation 2 are required. Figure 1 (b-c) indicates that filling of cuvettes with φ = 0.05 and h_n = 10 μηι would be less reliable than the other arrays studied because the material contact angle must be less than ~ 20° for wicking to occur. Thus, both the pillar height, lattice spacing, and material contact angle of the pillar array is important for reliable operation of the pillar cuvette.

[0054] In practice, the height (h_n) of each pillar in the array is from about 0.1 micrometres to about 500 micrometres, such as from about 0.1 micrometres to about 100 micrometres. As described earlier, the pillar height (h_n) and surface area fraction of the pillars (ø) need to be considered together in order to effectively achieve wicking. Thus, for example, for arrays having a surface area fraction of the pillars (ø) of at least about 0.05 and diameter 10 micrometres, the height of each pillar in the ordered array may be at least about 10 micrometres, such as about 10 micrometres or about 20 micrometres. The term

"substantially" as used herein with respect to the height of the pillars means that said dimension(s) for all pillars in the spectral region are within ±5%, ±4% , ±3% , ±2% or ±1 % of one another.

[0055] In the illustrated embodiments, the pillars in the array in the spectral region of the substrate extend substantially orthogonally from the surface of the substrate. As described in more detail later, the height of the pillars determines the path length of the sample because the height of the pillars determines the "depth" of the analyte sample in the interpillar spacings. In the case where the cuvette is used in transmission spectroscopy and the direction of the incident light is orthogonal to the substrate (or parallel to the length of the pillars) the path length is substantially the same as the height of the pillars. In the case where the cuvette is used in reflective spectroscopy and the direction of the incident light is orthogonal to the substrate (or parallel to the length of the pillars) the path length is substantially double the height of the pillars. [0056 ] Whilst the illustrated embodiments provide examples in which the direction of the incident light is orthogonal to the substrate, the angle of incident light could be anywhere from 0 degrees to 90 degrees. For example, the direction of the incident light may be 0 degrees in which case is it parallel to the surface of the substrate and the incident light passes through the array of pillars in a direction orthogonal to the height of the pillars. Alternatively, the direction of the incident light may be, for example, 45 degrees to the substrate. This may be particularly suitable for reflective spectroscopy.

[0057] In the illustrated embodiments, the pillars are circular in cross section. However, in practice the pillars can be any shape in cross section, such as square, round, elliptical, hexagonal, etc. The diameter of each pillar can be from about 0.1 micrometres to about 100 micrometres, such as about 10 micrometres.

[0058 ] The array of pillars may be an ordered array in which the dimensions of the pillars and hence the interpillar spacings are substantially uniform in the array or, alternatively, the array may be a disordered array in which the dimensions other than the height of the pillars and hence the interpillar spacings are non-uniform in the array.

[0059] In use, the interpillar spacings between adjacent pillars are filled with analyte sample in the spectral region. In the illustrated embodiments in which the array is an ordered array the interpillar spacings are substantially equidimensional which means that each interpillar spacing has substantially the same height (i.e. depth) and width as each other interpillar spacing in the ordered array. The dimensions of the interpillar spacings are determined by the height, diameter and surface area fraction of the pillars.

[0060] In certain embodiments, the surface area fraction of the pillars (ø) is from about 0.01 to about 0.90, such as 0.05, 0.06, 0.07, 0.08, 0.09, 0.10, 0.1 1 , 0. 12, 0.13, 0. 14, 0. 15, 0. 16, 0. 17, 0. 1 8, 0.19, 0.20, 0.21 , 0.22, 0.23, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.30, 0.3 1 , 0.32, 0.33, 0.34, 0.35, 0.36, 0.37, 0.38, 0.39, 0.40, 0.41 , 0.42, 0.43, 0.44, 0.45, 0.46, 0.47, 0.48, 0.49 or 0.50. In specific embodiments, the surface area fraction of the pillars is from about 0.05, about 0.10, about 0.20 or about 0.30.

[ 0061 ] In deriving the Beer-Lambert law relationship for a sample that has pillars embedded in a thin liquid film, Absorbance, A, is defined as

A = -log? (3) where the transmittance T— / and 7₀ are the intensities of the transmitted and incident light (see

also eq 1). To consider the absorbance measured against a reference sample (or 'blank'), first consider the pillar cuvette filled with the reference solution, as shown in Figure 2. The incident light, /₀, is effectively divided into two light paths: one through the pillars and one between the pillars through the solution. The portion of light in each path is related to the area fraction of pillars, so that the incident intensity for the pillars and the reference solution are given by <pl₀ and (1 — φ)Ιο, respectively. The quartz substrate (excluding the pillars) can be ignored, since it is optically homogeneous (equivalent to the optical window of a conventional cuvette). It follows that the transmitted intensities for the pillars, I_p, and the reference solution, /,-, are given by:

Ip = ΦΙ₀Τ_Ρ (4)

/,- = (1 - φ)1₀Τ_τ (5)

[0062 ] When the sample is loaded, it attenuates the light that is able to pass through the reference sample, I_r, according to the transmittance of the sample, T_s:

I_s = (1 - <P)I₀T_rT_s (6) and, therefore, the overall measured transmittance, T_m, is

[0063 ] Equation 7 is unusual in that T_p and T_r are relevant to the measurement, despite being included in the spectrum of the blank. This is due to the different transmitted intensities in the two light paths (pillars and reference solution), coupled with only one path being affected when the sample is loaded.

Furthermore, light that travels to the detector without passing through the sample is generally termed 'stray light' and is undesirable for analysis of high absorbance samples. The 'stray light' effect can be avoided, by preparing opaque pillars, i.e. T_p = 0, which recovers T_m— T_s and, thus, the traditional Beer- Lambert law:

A_s = - log T_s = e - h - [X] (8)

10064] For pillar cuvettes where T_p≠ 0, the transmittance of the pillars and reference solution are both required to calculate T_s from T_m. Given that the optical properties of the pillar material (quartz in the experiments presented here) and the reference solution are typically known (or can be easily determined), this is not a major barrier to the use of cuvettes with transparent pillars for absorbance spectroscopy.

[0065 ] The cuvettes described herein can be formed by providing a substrate. A template is then fonned on the surface of the substrate to give a templated substrate having a template pattern that reflects desired pattem of pillars in the array. For example, the template pattern may be in the form of a regular or ordered array of circular voids, with each void having a diameter approximately equivalent to the desired diameter of the pillars. The templated substrate is then etched under conditions to form the ordered array of pillars extending substantially orthogonally from a planar surface of the substrate.

[0066] The templated substrate can be formed using known photolithographic and etching procedures including soft lithography, such as near-field phase shift lithography, microtransfer molding, solvent- assisted microcontact molding, microcontact printing, and other lithographic microfabrication techniques employed in the semiconductor industry and elsewhere. Direct machining or forming techniques may also be used. Such techniques may include hot embossing, cold stamping, injection moulding, direct mechanical milling, laser etching, chemical etching, reactive ion etching, physical and chemical vapour deposition, and plasma sputtering.

[0067] By way of example, the templated substrate can be formed by coating a surface of the substrate with a blanket photoresist layer formed from a photoresist material which is susceptible to radiation induced curing, such as an SU8 layer. The photoresist layer is then exposed to UV radiation through a patterned photomask for selective irradiation of the blanket photoresist layer to form a patterned template on the surface of the substrate. The types of radiation employed on the photoresist material include, but are not limited to, deep ultraviolet (DUV) (i.e. 248 nm) radiation, near ultraviolet (NUV) (i.e. 365 nm) radiation, multiple wavelength flood ultraviolet (UV) radiation, electron radiation, and ion radiation.

[0068] The templated substrate is then etched with a plasma etching agent under conditions to form the ordered array of pillars extending substantially orthogonally from a planar surface of the substrate.

Conditions for etching quartz, glass and other substrates are known in the art and can be used in the process described herein. For example, fluorocarbon etching agents, such as C₃F₈, can be used in an anisotropic plasma etching process. It will be appreciated that the etching time is used to determine the height of the pillars, with a longer etching time resulting in higher pillars for specific plasma conditions.

[0069] Whilst the aforementioned process is particularly suitable for forming the pillar arrays and cuvettes, it will be appreciated that the pillars could also be directly stamped or moulded into a substrate, or a photoresist could be directly used as the pillars.

[0070] In addition to the wicking structure, the substrate also comprises an analyte solution loading region. The analyte solution loading region is configured to receive a droplet of the analyte solution such that the droplet is in fluid contact with the wicking structure such that the analyte solution is dispersed within the wicking structure. The volume of the droplet applied need only be large enough to fill the volume of the pillar array in the spectral region. For single droplet measurements, a sample volume of about 2 mL is sufficient. After placement on the analyte solution loading region, the droplet spontaneously spreads to meet the edge of the pillar array region, which triggers the wicking effect. The film spreads rapidly through the pillar array - typically between 1 and 4 mm/s for φ = 0.20 and h = 10 mm. The filled cuvette is then ready for use in spectroscopy.

[0071 1 As mentioned, the cuvette described herein is suitable for optical spectroscopy using an optical spectrometer. Typical optical spectrometers contain a source of radiant energy, a cuvette holder, a device that isolates a restricted region of the spectrum for measurement (e.g., ultraviolet, visible, or infrared), a radiation detector, and a signal processor and readout. Thus, provided herein is an optical spectroscopic method comprising providing a cuvette as described herein, contacting a droplet of an analyte solution to be analysed by optical spectroscopy with part of the wicking structure of the cuvette under conditions to fill the interpillar spacings in the spectral region with the analyte solution to provide a filled cuvette; exposing a first side of the filled cuvette to a light source such that an incident light beam is projected through the spectral region of the cuvette to a detector located at a second side of the filled cuvette that is opposite the first side, or exposing a first side of the filled cuvette to a light source such that an incident light beam passes through the analyte sample in the filled cuvette and is reflected from the substrate to a detector located at the first side of the cuvette; and measuring at least one parameter of the light at the detector. The light source can include monochromatic, multiple wavelengths, single bandpass, broadband, or polychromatic light. The light can be circularly polarised, linearly polarised, elliptically polarised, nonpolarised, or selectively polarisable, wherein the polarisation angle and or phase of linearly polarised light may be adjusted. The light source can also include light emitting diodes, organic light emitting diodes, quantum dot light emitting diodes, carbon nanotube LEDs, lasers, tunable lasers, vertical cavity surface emitting lasers, filament lamps, discharge lamps, sunlight, or super luminescent diodes. Other light can include UV, visible, IR, coherent, semi-coherent, incoherent, light selectively filtered by bandpass, multipass, longpass, or shortpass filters, or multiple filters in combination.

[0072 ] The optical spectroscopic method may also include a step of obtaining a reference spectrum. A reference spectrum can be obtained by adding a background solvent to the cuvette, collecting the spectrum, and using it as a baseline for the subsequent measurements. Unlike conventional cuvettes, when the cuvettes described herein are used in optical spectroscopy the incident beam encounters two materials (liquid and substrate material, such as quartz) with different refractive indices and a periodicity that causes diffraction and interference patterns of the incident light. However, we found that the diffraction and interference patterns in the collected reference spectra was reproducible and did not affect the accuracy of subsequent UV-Vis absorbance measurements. Opaque pillars are convenient because the area fraction and transmittance of the pillars and reference samples is irrelevant. However, transparent and semitransparent pillars can also be used, provided that a stray light correction is applied.

[0073 ] To obtain a spectrum, a droplet of analyte solution is placed in the analyte solution loading region. After placement of the droplet of analyte solution, liquid wicks into the wicking structure and, hence, the spectral region and a sharp spike in absorbance may be observed due to the curvature of the leading meniscus. Thereafter, the absorbance may decrease again and begin to rise linearly after that due to steady evaporation of the solvent from the cuvette. The evaporated solvent is immediately replaced by fresh analyte solution from the droplet to maintain the liquid film in the pillar array. The evaporation rate is constant and, on the timescales of the measurement, diffusion of analyte from the pillar array to the droplet is negligible so that the change in concentration is linear and can be used for measurement of unknown samples.

[0074 ] In the embodiment shown in Figure 3, the pillars and the substrate itself are optically transparent in the spectral region. In these embodiments, the cuvette is used in an optical spectrometer with incident light directed parallel to the direction in which the pillars extend from the substrate. When the ordered array of pillars is filled with liquid the curvature of the meniscus profile of the liquid at the top of each interpillar spacing is small. This means that the path length through the sample is substantially the same as the height of the pil l ars. In use, the liquid neither engulfs the tops of the pillars nor fills part way up the pillars, ensuring that capillary filling of the cuvette is reproducible between measurements.

[0075] Thus, using the cuvettes described herein provides a reliable, dilution-free method for UV-Vis analysis of high concentration or high ε samples in both organic and aqueous phases.

EXAMPLES

[0076 ] Solution Preparation

[00771 A platinum solution was prepared by dissolving sodium hexacholoroplatinate (IV) from Johnson Matthey in 0.5 M HC1. The concentration of the prepared stock solution was 24 g/L Pt⁴ . This solution was used to prepare the different dilutions.

[00781 Cuvette Fabrication

[0079] A pillar cuvette with transparent pillars was prepared using UV-photolithography and plasma etching on optical grade quartz. The quartz substrate was polished 4" diameter, 700 μηι thick wafer (Shin- Etsu). The wafer was dehydrated by heating to 200°C, then cooled to room temperature to spin-coat (Suss Microtec, Delta 80RC) SU8-10 photoresist. The sample was then pre-baked, exposed to UV at 365 nm (EVGroup, EVG 620) through a chrome-on-glass photomask (Optofab, Australian National Fabrication Facility). The photoresist was developed and hard baked at 200°C. The backside of the substrate was sputter coated (HHV/Edwards TF500) with a chromium layer to enable adhesion to the electrostatic chuck in the plasma etching tool (ULVAC 570NLD). Anisotropic plasma (C₃F₈) etching was used to replicate the SU8 structure in the quartz wafer. The wafer was diced and the photoresist removed by ashing in a furnace at 550°C. Some residue remained at the centre of the top of the pillars but this had no adverse effect on wicking or the performance of the cuvette.

[0080] For opaque pillars, the wafer was first sputter-coated with an approximately 100 ran thick layer of Cr and an approximately 50 nm layer of Au in a vacuum (2 χ 10-5 mbar). The etching procedure described in the preceding paragraph was then followed. Finally, the photoresist was removed by etching the Au layer using 4: 1 of I/I₂ in water, then washed using Piranha solution (7:3 ratio of 98% H₂S0₄ and 30% H₂0₂) at room temperature.

[0081 ] The pillars were cylindrical because pinning effects are reduced compared to other geometries with vertical edges. [13] Four projected surface area fractions of the pillars, φ, were used, where w (w = 10 mm) and d are the pillar diameter and lattice spacing of the array, respectively. Figure 2 shows the arrangement of the pillars (square lattice) with relevant dimensions labelled and scanning electron microscopy (SEM) images for three different area fractions. For low coverages, i.e., φ = 0.05 and 0.10, the pillars are resolved well; however, as the lattice spacing of the pillars reduces, so do interstitial defects. These imperfect pillar arrays were also used to determine how sensitive the analysis is to the quality of the surface structure.

[0082 ] Two sets of samples were prepared with nominal pillar heights, h, at 10 and 20 mm. However, in practice, the pillar height varies (for fixed etch time and plasma conditions) depending on the aspect ratio of the etched region, such that the rate of etching is slower for pillar arrays with smaller d. Thus, the different pillar arrays had slightly different heights (Table 1) but the fabrication of the pillar cuvettes was reproducible for a fixed φ, w, and d. Height measurements on 10 different locations on a single pillar array (φ = 0.20; h = 10 mm) gives a representative uncertainty in the pillar height of +0.33 mm, or 3 %.

[0083 ] Table 1 - Actual plasma etch depths (pillar heights) for pillar arrays at different projected surface area fractions, φ, and different nominal etch depths, h_n.

Actual pillar height h (μιη)

Φ h_n= 10 μηι h_n= 20 μηι

0.05 13.3 26.1 0.10 12.5 24.1

0.20 1 1.1 20.6

0.30 10.2 20.0

[0084 ] Spectroscopy

[0085 ] For all spectroscopic measurements an Ocean Optics DT-Mini-2 light source and Ocean Optics QE65000 detector were used. The light source and detector are compact and easily coupled with lab-on-a- chip platfonns for application in remote locations.

[0086] Results and Discussion

[0087] A conventional 2 mm path length quartz cuvette was used for comparison with the pillar-cuvette measurements. The molar absorptivity, ε, at 259 nm for PtCl₆ ^2" complex in the aqueous phase was found to be very high (2.055 x 10^"3 m²/mol) so that the measurable concentration range is limited to values less than ~ 50 ppm in this cuvette. Industrial processing and refining of precious metals, including platinum, is carried out at much higher concentrations. For example, platinum concentrations in excess of 20 g/L are common in refinery solutions. Analysis using UV-vis would therefore require pre-dilution factors of up to 1000. These dilutions are not only time-consuming. Large amounts of solvent (here, a 0.5 M HCl aqueous solution) are required even for microvolume aliquots of sample; 100 xL sample requires 100 ml of solvent for a single measurement. The additional sample handling may also lead to greater analytical uncertainty.

[0088] Capillary Filling

[0089] The UV-Vis spectroscopy analysis using the pillar-cuvette proceeds as follows. The pillar cuvette is cleaned with a wash bottle with water and or ethanol and dried with compressed air before a single droplet of neat sample was placed immediately adjacent to the pillar array. The applied volume is unimportant for the operation of the cuvette and UV-Vis measurements. In fact, partially dipping a pillar- cuvette into a bulk liquid will give the same results. However, for the single droplet measurements carried out, the sample volume is ~ 2 mL. This is 50 times less than that required for accurate dilution, no additional solvent is required, no special dispensing equipment is required (e.g., a micro-pipette), and loading and rinsing the pillar cuvette takes only a few seconds. After placement, the droplet spontaneously spreads to meet the edge of the pillar array region, which triggers the wicking effect. The film spreads rapidly through the pillar array - typically between 1 and 4 ram/s for φ = 0.20 and h = 10 mm, based on high-speed microscopy - and is dependent on the viscosity of the liquid and the distance travelled as predicted by Washburn for capillary filling of porous media. [ 10]

[0090] Once the pillar array was filled with the sample, we used optical profilometry (VEECO, WY O NT9100) to characterise the morphology of the liquid-vapour interface. Figure 4 shows the meniscus profiles for φ = 0.05 and 0.20, with the line scans for the empty pillar arrays shown for comparison. The curvature of the meniscus profile is small, i.e., ~ 300 nm (within the uncertainty of the etch depth), leading to a negligible change in the actual path length. It is important to note that the liquid neither engulfs the tops of the pillars nor fills part way up the pillars, ensuring that capillary filling of the cuvette is reproducible between measurements. Thus, the meniscus curvature is consistent between

measurements, so that the precision of the measurements is unaffected. The meniscus morphology was found to be stable for long periods of time, provided the volume of the droplet was in excess of the interstitial volume of the pillar array. The excess liquid in the placed droplet acts to continuously replenish the film of liquid in the pillar cuvette as evaporation occurs. For a typical droplet volume, the meniscus morphology is maintained for tens of seconds.

[0091 ] Spectroscopy

[0092 ] Accurate UV-Vis absorbance measurement requires a reliable reference spectrum. The reference spectrum is obtained by adding the background solvent to the cuvette, collecting the spectrum, and using it as a baseline for the subsequent measurements. As discussed, filling the pillar cuvette is reproducible in terms of film thickness and meniscus morphology; however, unlike conventional cuvettes, the incident beam encounters two materials (liquid and quartz) with different refractive indices and a periodicity that causes diffraction of the incident light. Solving for the diffraction pattern is complex; however, the nature of the diffraction pattern in the collected reference spectra was reproducible and did not affect the accuracy of the UV-Vis absorbance measurements. Figure 5 shows the effect of diffraction on the reference spectra with and without solvent (0.5 M HCl_(aq)) present. The number of peaks increases with increasing difference between the refractive index of the quartz and the liquid or vapour, while the peak intensity increases with increasing φ. The results are off-set with respect to the absorbance axis for clarity.

[0093] Using the reference spectrum of 0.5 M HCl_(aq), we proceeded to test the analytical performance of the pillar cuvette using PtCl₆ ^2~ _(aq) solutions. The absorbance at 259 nm was collected from before droplet placement until sufficient data was collected for analysis. After placement of the droplet of sample, the measured absorbance remained at zero until the wicking liquid reached the detection area of the pillar array and a sharp spike in absorbance was observed, Figure 6. This spike may be due to the curvature of the leading meniscus, which refracts the incident light away from the detector, or a transient concave curvature of the liquid-vapour interface between the pillars. Within seconds, the absorbance decreases again and begins to rise linearly over tens of seconds. The linear increase in absorbance is due to steady evaporation of the solvent from the pillar array. The evaporated solvent (which selectively leaves behind non-volatile PtCl₆ ^2~) is immediately replaced by fresh sample from the droplet to maintain the liquid film in the pillar array. The evaporation rate is constant, due to the very well defined geometry of the film, and therefore the change in concentration is linear. Here, we assume that the volume of the excess droplet is sufficiently large so that evaporation from that droplet has a negligible effect on concentration (volume of the film roughly 0.5 μL· for φ = 0.10 and h„ = 10 μηι) and that diffusion of analyte from the pillar array to the droplet is negligible. Fitting the data within the region denoted by the dotted lines, see Figure 6, then calculating the absorbance at 6 s (first dotted line) based on the best fit (immediately after the pillar cuvette is filled; first dotted line) gives values for standard solutions that are in excellent agreement with Beer's Law and can be used for measurement of unknown samples. This approach ensures that evaporation rates have no influence on the measurements. The method could also be automated using appropriate software so that the user was only required to place the droplet.

[0094 ] Opaque Pillars

[0095 ] Equation 7 relates the measured transmittance, 7_m, to the transmittance through the sample, T_s, for semitransparent and opaque pillars. For opaque pillars, we recover T_m = _s and, therefore, the

Beer-Lambert law (eq 1 ). To confirm this experimentally, we prepared pillar arrays with varied φ with the top surface of the pillars capped by a 100 nm thick Cr layer. The latter made the pillars opaque, that is, Γ_ρ = 0, while allowing light to pass through the liquid that wicks between the pillars. Table 2 gives the measured dimensions of the pillar cuvettes with opaque pillars, which were used for the spectroscopy measurements presented in this section (see also Figure 4).

[0096] Table 2 - Actual Values of Area Fraction (φ), Pillar Width (w), and Pillar Height (h) for the Cuvettes with Opaque Pillars

[0097] Note that the path length decreases later accounted for by reporting the absorbancc normalized by the measured path length (A*). Accurate UV-vis absorbance measurement requires a reliable reference spectrum. The reference spectrum is obtained by adding the background solvent to the cuvette, collecting the spectrum, and using it as a baseline for the subsequent measurements. As shown above, filling the pillar cuvette is reproducible in terms of film thickness and meniscus morphology; however, unlike conventional cuvettes, the incident beam is split into two paths. Since no light can pass through the opaque pillars, an interference pattern results exclusively from light that is transmitted through the liquid. We found that the interference pattern observed for the empty cuvette is reproducible for a given cuvette and, for the filled cuvettes and small φ, the wavelength dependence of the absorbance is not very significant. Regardless of the choice of cuvette, the reference spectra were reproducible and did not influence the accuracy of the UV-vis absorbance measurements. A plot of the transmittance of the cuvette itself (no liquid) against φ was linear, as expected for opaque pillars. Using the reference spectrum for 0.5 M HCl(aq), we tested the analytical performance of the opaque pillar cuvette using PtCl₆ ² (aq) solutions. The absorbance at 259 nm was collected from before droplet placement until sufficient data was collected for analysis. After placement of the droplet of sample, the measured absorbance remained at zero until the wicking liquid reached the detection area of the pillar array and a sharp spike in absorbance was observed. This spike may be due to the curvature of the leading meniscus, which refracts the incident light away from the detector. Within seconds, the absorbance decreased again and began to rise linearly over a relatively long time-scale (tens of seconds). The linear increase in absorbance is due to steady evaporation of the solvent from the pillar array. The evaporated solvent (which selectively leaves behind nonvolatile PtCl₅ ² ) is immediately replaced by a fresh sample from the droplet to maintain the liquid film in the pillar array. The evaporation rate is constant, due to the very well-defined geometry of the film, and therefore, the change in concentration is linear. Here, we assume that the volume of the excess droplet is sufficiently large so that evaporation from that droplet itself has a negligible effect on concentration (the volume of the film is ~0.5

for φ = 0.10 and h = 10 μιη) and that diffusion of the analyte from the pillar array to the droplet is negligible over the time scale of the measurement. Fitting the data within the linear region then calculating the absorbance at the intercept of the linear fit with the initial spike (the moment the pillar array is filled) gives values for the standard solutions that are in excellent agreement with the Beer-Lambert Law and can be used for measurement of unknown samples. This approach ensures that evaporation rates have no influence on the measurements. The method could also be automated using appropriate software that detects the initial spike and fits the data, so that the user is only required to place the sample on the pillar-cuvette. Figure 8a shows experimental results for six different pillar cuvettes with varied φ and each with slightly different path lengths (pillar heights) due to limitations in our etching methodology, see Table 2. All six cuvettes report a linear dependence of absorbancc with concentration in agreement with the Beer-Lambert Law, see Figure 8a. The slopes of the best linear fits to the data decrease with increasing φ due to the decreasing path lengths. When the measured absorbance is normalized (A*) to account for the actual path length (pillar height), all data collapses onto a single line. Based on the collapsed data and eq 1 , the molar absorptivity, ε, can be determined (241 16 L mol '-cm '). This value is in good agreement with ε determined in this study using 2 mm (ε = 23098 L- mol ¹ -cm ¹) and 10 mm (ε = 23459 L- mol ¹ cm path length conventional cuvettes and reported values. Thus, the optical path lengths appear to be reasonably well defined by the pillar etch depths; however, for alternative geometries, including tapered pillars, an "effective path length" may be more relevant.

[0098 ] Semi-transparent Pillars

[0099 ] The preparation of quartz pillar cuvettes with opaque pillars requires the additional step of metal deposition and etching compared with bare quartz pillars. We sought to eliminate these steps by studying cuvettes with transparent quartz pillars. This was achieved by etching the Cr layer from the tops of the opaque pillars used previously, followed by hydrofluoric acid etching to remove residues and clean the surface. The etching reduced the pillar width slightly (w = 5.8 μπι). Thus, the actual φ ranged from 0.02 to 0.17, see Figure 9, with no change in the lattice constant or pillar height (refer to Table 2). The pillars are semitransparent and the measured transmittance, if_m, through the cuvette containing a liquid sample is theoretically predicted by eq 7. The input parameters are φ, 7_P, 7_r, and 7'_s, of which φ is known and T_T ~ \ for water at the relevant wavelength (259 nm). For quartz, T_v is expected to be relatively high; however, the transmittance of the pillars is likely to be reduced due to increased scattering, for example, due to roughness, finite wall angle (typically 87-90°), or imperfections in the fabrication. Thus, we are left with two unknowns in eq 7: 7_P and T_s. T_v and φ are critically important due to the effect of stray light on absorbance spectroscopy. Where 7_P and φ are large, stray light can dominate the transmission causing departure from linearity. The effect of φ is clearly revealed in the experiments (Figure 9a). Absorbance is plotted for various pillar cuvettes (φ = 0.02 to 0.17). For the smallest area fraction, φ = 0.02, the departure from linearity is barely noticeable; however, for φ = 0.17, the measured absorbance is much lower over the full range of concentrations. While this is undesirable, it differs from a conventional "stray light" effect because it can be quantitatively accounted for using eq 7. Conveniently, we have independently determined T_s using the cuvettes with opaque pillars, for which φ, T_v, and T_r are unimportant. Using the experimentally derived T_s (opaque pillars) and assuming 7_P = 0.60 and T_r = 1 , we arrive at the predicted curves for pillar cuvettes with semitransparent pillars shown in Figure 9a. The prediction is in good agreement with the measurements, especially at concentrations up to ~4 g/L. The only "fitting" parameter in this comparison is T_p; however, as is shown below, this value can be independently verified from experiment. The transmittance of the liquid sample alone, Ί , can be obtained by rewriting eq 7:

[00100] ^Ts = ^Tm + T^ T^T _r - - 1) (9) [00101 ] For a given path length, reference liquid and cuvette material, (T_P/T_T) is constant, and therefore, the prefactor {(φΙ( \ - φ))- (T_V/T_T)) in eq 9 varies with φ alone. Furthermore, for measurements of a single sample using any two cuvettes with distinct values of φ, the constant ( Τ^Τ_τ) can be determined from T_m, because T_s is unchanged for a given path length. Based on the results presented in Figure 9a and Tr = 1 , we find that T_v = 0.60. Since A = -log T_s increases linearly with concentration and T_s is independent of cuvette design, a plot of A* (A normalized by the path length) against platinum concentration provides a validation of eqs 7 and 9. Figure 9b is a strong validation that pillar cuvettes with semitransparent pillars can be successfully used for absorbance spectroscopy of high ε samples (up to 3-4 g/L of platinum in this study), despite the "stray light", which travels through the pillars to the detector.

[00102 ] In-sit Filtration of Particles

[00103] We also considered the case where microscopic particles are present in the sample and adversely affect the UV-Vis measurement. Particles obstruct the incident light, shifting the baseline and, in some cases, may absorb or emit light at the same wavelength at the target analyte. Filtration or sedimentation requires additional materials and time. However, the pillar cuvette provides the opportunity to filter microscale particles in-situ, without any additional sample handling. Aqueous PtCl₆ ^2" (3 g/L Pt(IV) for the pillar array and 30 ppm Pt(IV) for the 2 mm cuvette) solutions were spiked with 6 mm polystyrene spheres (Polysciences Inc.) and collected the UV-Vis spectra in both types of cuvette. Figure 9(c) shows an image of the pillar array close to the edge where the solution began wicking into the pillars. The polystyrene particles clearly are arrested by the pillar array during the capillary flow, leaving the filtered solution free to travel into the detection zone of the array. The spectra for the two types of cuvette with and without particles added clearly show that the simple one-step analysis is highly effective, Figure 9(a-b).

[00104] Conclusion

[00105] We used micro pillar structures as a convenient and reliable way to form 10 and 20 mm thick liquid films for UV-vis spectroscopy analysis. The film thickness is precisely controlled by the height of pillars and filling of the 'pillar cuvettes' is spontaneous, driven by capillarity. We chose hexachloroplatinate (IV) as the analyte, due to the very strong absorbance peak at 259 nm. The microscale path length of the pillar cuvette allowed us to directly determine the concentration of undiluted samples at concentrations as high as 10 g/L without loss of precision (compared to measuring diluted samples in a conventional 2 mm cuvette). The open structure allows rapid cleaning of the cuvette and reloading. Furthennore, the pillars act to retain microscopic particles, permitting spectroscopic analysis of the liquid phase can be carried out without prior filtration. [00106] Throughout the specification and the claims that follow, unless the context requires otherwise, the words "comprise" and "include" and variations such as "comprising" and "including" will be understood to imply the inclusion of a stated integer or group of integers, but not the exclusion of any other integer or group of integers.

[00107] It will be appreciated by those skilled in the art that the invention is not restricted in its use to the particular application described. Neither is the present invention restricted in its preferred embodiment with regard to the particular elements and/or features described or depicted herein. It will be appreciated that the invention is not limited to the embodiment or embodiments disclosed, but is capable of numerous rearrangements, modifications and substitutions without departing from the scope of the invention as set forth and defined by the following claims.

REFERENCES

[00108] The reference to any prior art in this specification is not, and should not be taken as, an acknowledgement of any form of suggestion that such prior art forms part of the common general knowledge.

[00109] ( 1) von Schenck, H.; Falkensson, M; Lundberg, B. Clinical Chemistry 1986, 32, 526-

529.

[001 10] (2) Zheng, W.; Shan, N.; Yu, L.; Wang, X. Dyes and Pigments 2008, 77, 153-157.

[001 1 1 ] (3) Lichtenthaler, H. FC.; Buschmann, C. In Current Protocols in Food Analytical

Chemistry; John Wiley & Sons, Inc., 2001.

[001 12] (4) Adam, C. D.; Sherratt, S. L.; Zholobenko, V. L. Forensic Science International 2008,

174, 16-25.

[001 13] (5) Yu, Q.; Liu, S.; Zhang, M.; Cai, N.; Wang, Y.; Wang, P. The Journal of Physical

Chemistry C 2009, 113, 14559-14566.

[001 14] (6) Borowsky, J.; Collins, G. E. Analyst 2007, 132, 958-962.

1001 15] (7) Chin, C. D.; Under, V.; Sia, S. K. Lab on a Chip 2012, 12, 21 18-2134.

[001 16] (8) Waggoner, P. S.; Craighead, H. G. Lab on a Chip 2007, 7, 1238- 1255.

[001 17] (9) Chow, A. W. AIChE Journal 2002, 48, 1590-1595.

[001 18] (10) Washburn, E. W. Physical Review 1921, 17, 273-283.

[001 19] ( 1 1) Bico, J.; Thiele, U.; Quere, D. Colloids and Surfaces A 2002, 206, 41 -46.

[00120] ( 12) Wenzel, R. N. Ind. Eng. Chem. 1936, 28, 988.

[00121 ] (13) Semprebon, C; Forsberg, P.; Priest, C; Brinkmann, M. Soft Matter 2014, 10, 5739- 5748.

Claims

1. A cuvette for optical spectroscopy comprising a substrate and a wicking structure in a spectral region on the substrate, the wicking structure comprising an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height in the spectral region, the wicking structure configured such that interpillar spacings between adjacent pillars in the spectral region are filled with an analyte solution when the solution is placed in contact with part of the wicking structure to thereby form a spectral sample of the analyte solution in the interpillar spacings that is suitable for optical spectroscopy.

2. A cuvette when used in an optical spectrometer, the cuvette comprising a substrate and a wicking structure in a spectral region on the substrate, the wicking structure comprising an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height in the spectral region, the wicking structure configured such that interpillar spacings between adjacent pillars in the spectral region are filled with an analyte solution when the solution is placed in contact with part of the wicking structure to thereby form a spectral sample of the analyte solution in the interpillar spacings that is suitable for optical spectroscopy.

3. The cuvette according to either claim 1 or claim 2, wherein the substrate further comprises an analyte solution loading region on the substrate, the analyte solution loading region configured to receive a droplet of the analyte solution such that the droplet is in fluid contact with the wicking structure such that the analyte solution is dispersed within the wicking structure.

4. The cuvette according to any one of claims 1 to 3, wherein the pillars are transparent.

5. The cuvette according to any one of claims 1 to 3, wherein the pillars are semi-transparent.

6. The cuvette according to any one of claims 1 to 3, wherein the pillars are opaque.

7. The cuvette according to any one of claims 1 to 6, wherein the array is an ordered array.

8. The cuvette according to any one of claims 1 to 7, wherein the pillars are circular in cross section.

9. The cuvette according to any one of claims 1 to 7, wherein the pillars are square in cross section.

10. The cuvette according to any one of claims 1 to 9, wherein the height of each pillar is from about 0.1 micrometres to about 500 micrometres.

1 1. The cuvette according to claim 10, wherein the height of each pillar is about 10 micrometres.

12. The cuvette according to claim 10, wherein the height of each pillar is about 20 micrometres.

13. The cuvette according to any one of claims 1 to 12, wherein the surface area fraction of the pillars is from about 0.01 to about 0.90.

14. The cuvette according to claim 13, wherein the surface area fraction of the pillars is about 0.05.

15. The cuvette according to claim 13, wherein the surface area fraction of the pillars is about 0.10.

16. The cuvette according to claim 13, wherein the surface area fraction of the pillars is about 0.20.

17. The cuvette according to claim 13, wherein the surface area fraction of the pillars is about 0.30.

18. The cuvette according to any one of claims 1 to 17, wherein the substrate is optically transparent in the spectral region.

19. The cuvette according to any one of claims 1 to 18, wherein the substrate is reflective in the spectral region.

20. The cuvette according to any one of claims 1 to 19, wherein a surface of the substrate and/or pillars that comes into contact with the analyte solution further comprises a functional coating that is able to react or bind an analyte or release a species to generate or enhance an optical signal.

21. The cuvette according to any one of claims 1 to 20, wherein the dimensions of the interpillar spacings are suitable for filtering particulates from the analyte sample.

An optical spectroscopic method comprising providing the cuvette according to any one of claims 1 to 21 ; contacting a droplet of an analyte solution to be analysed by optical spectroscopy with part of the wicking structure of the cuvette under conditions to fill the interpillar spacings in the spectral region with the analyte solution to provide a filled cuvette; exposing a first side of the filled cuvette to a light source such that an incident light beam is projected through the spectral region of the cuvette to a detector located at a second side of the filled cuvette that is opposite the first side, or exposing a first side of the filled cuvette to a light source such that an incident light beam passes through the analyte sample in the filled cuvette and is reflected from the substrate to a detector located at the first side of the cuvette; and measuring at least one parameter of the light at the detector. 23. A process for producing a cuvette for optical spectroscopy, the process comprising: providing a substrate; forming an array of pillars extending from a planar surface of the substrate with each pillar in the array being of substantially equal height; and forming analyte solution loading region on the substrate, the analyte solution loading region configured to receive a droplet of an analyte solution such that the droplet is in fluid contact with the array of pillars such that the analyte solution is dispersed within the array of pillars.