AU2232700A - Reflectance measurement probe - Google Patents

Description

AUSTRALIA

Patents Act 1990

ORIGINAL

COMPLETE SPECIFICATION STANDARD PATENT Invention Title: REFLECTANCE MEASUREMENT PROBE 4 The following statement is a full description of this invention, including the best method of performing it known to me:

I

W:\DELlLAH\DCNODELEkDAV!DW -pr..do.

-2- REFLECTANCE MEASUREMENT PROBE TECHNICAL FIELD This invention relates to a probe for measuring reflectance of light from a sample surface.

BACKGROUND

10 A common requirement in ultraviolet-visible spectrophotometry is to measure the oooe Stotal reflectivity of a sample surface. Such measurements are complicated by the fact that the light reflected from the sample may be reflected in a number of ways.

It could be scattered equally in all directions (a surface that reflects light in this Sway is called a Lambertian reflective surface), it could be reflected as if from a mirror (specular reflection) or it could be reflected in any manner between these S.two extremes.

o To avoid this complication, it is common to make total reflectance measurements l with the aid of an integrating sphere. An integrating sphere is a hollow cavity 20 (normally, but not necessarily, spherical in shape) that is lined with a material having a surface that is as close as possible to a 100% Lambertian reflective surface. The purpose of this surface is to scatter any light entering the cavity so as to make the density of light energy uniform at all points on the wall of the cavity. At least three openings (ports) are made into this cavity. The first (the entrance port) is for entry of light into the cavity. The second (the sample port) is for presentation of the sample. The sample port is so positioned that light from the entrance port is incident on a sample placed over the sample port. The third port (the detector port) is for placement of an optical detector, the function of which is to sample the light incident on that portion of the cavity wall. In operation, a beam of light enters the cavity through the entrance port and is incident on the sample presented at the sample port. Any light reflected from the sample, irrespective of angle, is collected by the cavity and its intensity is averaged by -3multiple reflections from the cavity walls. The effect of this averaging is that the resultant intensity of light energy at any point of the cavity wall is linearly dependant on the total light reflected from the sample and independent of the angle at which that light was originally reflected. In these circumstances, the light incident on the detector can be taken as a representative measure of the total light reflected from the sample. The averaging effect of the cavity improves as the ratio of the cavity wall area to total port area increases. Consequently it is common to specify this ratio as one measure of performance of an integrating sphere, with 2% to 3% total port area being a common target. By ratioing the light intensity signal obtained when a reference material that is essentially 100% reflective is placed over the sample port to the signal obtained when the sample .I is placed over the sample port, a relative estimate of the reflectance of the sample can be obtained.

In practice it is common to use powdered polytetrafluoroethylene for the inside coating of the cavity. This material is an almost perfect Lambertian reflector, having a reflectivity of around 0.994 in the visible region and about 0.96 in the ultraviolet and near-infrared regions of the spectrum. ("Reflection properties of pressed polytetrafluoroethylene powder", V. Weindner and J. Hsia, Journal of the Optical Society of America, Volume 71, page 856, July 1981). With this wall coating, after 20 reflections, the light loss by absorption in the walls over the visible portion of the spectrum is 1-(0.994)2o 0.11 11%) and at the extremes of the ultraviolet and near-infrared regions up to 1-(0.96)2o 0.56 56%).

From this it can be seen that most of the light is lost from the cavity by exit through the ports rather than by absorption at the cavity wall. Therefore the light loss can be analysed (at least to a first approximation) by considering the relative port sizes and ignoring wall losses.

Consider a simple model of a sphere with three ports inlet, sample and detector.

The area of each port, expressed as a fraction of the total combined port area is as follows: inlet port area: I sample port area: S -4detector port area: D.

The reflectivity of the sample is R (where 0 R 1) and the incident light energy is taken as 1 per unit.

The light incident on the sample is then 1 per unit. The reflected light will be R per unit. The reflected light will repeatedly reflect from the coated walls of the cavity until it contacts one of the ports. Essentially all the reflected light will thus eventually reach the ports and the fraction incident on each port will be directly proportional to the relative area of each port, thus I*R will reach the inlet port and be lost, D*R will reach the detector and be converted to an electrical output, S*R will reach the sample port and be re-reflected.

The light reaching the sample port will be re-reflected with a resultant intensity of S*R*R. On leaving the sample port it will again be dispersed by the cavity with the result S*I*R*R will reach the inlet port and be lost, S*D*R*R will reach the detector and be converted to an electrical output, S*S*R*R will reach the sample port and be re-reflected.

It is readily apparent that this argument can be continued indefinitely and the total resultant light reaching the detector is the summation of the series D*R*(1 (S*R) 2

(S*R)

3 4 Reflectivity measurements are ratios of the sample signal to that from a standard material having a reflectivity of essentially 100% The measured reflectivity is therefore D*R*(1 (S*R) 2

(S*R)

3 3 D*(1 S S 2

S

3

S

4 and this is not equal to R. Indeed, the error can be quite significant. For example, if S is 20% of the total port area, a sample of reflectivity 0.5 will be measured as 0.44 an error of 12%. For sample port area of 30% of the total port area, the error is 17.5%.

This error is a major problem for single beam measurement systems. For this reason it is sometimes called the single-beam sample presentation error. There are many situations where a single beam measurement approach (in contrast to the use of a double beam system in which the error can be eliminated) has advantages especially where the measurement is to be made at a location remote from the rest of the optical system.

There is also a second problem with existing integrating sphere technology. As derived above the total light received by the detector is D*R*(1 (S*R) 2

(S*R)

3 4 S• The limitations of geometric optics applied to conventional UVNIS light sources requires that the inlet and sample ports are of substantial size to accommodate the optical beam. By contrast, the detectors used are relatively small so that D as a fraction of the total port area is small (typically about 0.1 or less). This means that the detector at best only receives a small portion of the light incident onto the sample. For example, if we take D as 0.1 and both S and I at 0.45, the detector receives at best only 18% of the incident light even for a 100% reflective sample, and less if the cavity coating is less than 100% reflective. As a result, the signal is lower than it would otherwise be and the signal-to-noise ratio is commensurately worse.

A third problem with integrating spheres is the fragility of the reflective coating applied to the inside of the cavity. This coating is easily damaged through mechanical contact and easily contaminated via airborne pollutants. Such damage reduces the magnitude and uniformity of the coating reflectivity, deteriorating measurement accuracy. The coating is necessarily exposed to the environment via the sample port and thus very vulnerable. Consequently such systems require careful handling and a clean environment, and there is some limitation on the types of samples that can be measured. For example, measurements of open containers of liquid or powders would be extremely difficult and would involve a substantial risk of contamination of the coating of the -6integrating sphere. Even with careful handling, such systems tend to deteriorate in performance with time, and require periodic re-coating.

The present invention, at least as realised in specific embodiments, seeks to reduce one or more of these three problems.

There have been numerous previous attempts to overcome some or all of the problems described above. One prior art technique is to use a highly reflective specular surface of known and carefully controlled shape to collect and image light from a diffuse reflection. The ideal surface shape required is ellipsoidal which re-focuses such light after a single reflection. Harrick Scientific of Ossining New York, an optical accessory manufacturer for UVNIS, has marketed a reflectance accessory called the "praying mantis" accessory (a name derived from the overall appearance of the device) based on an ellipsoidal collector. Such an approach has however severe limitations. An ellipsoidal surface only correctly focuses diffusely reflected light when such light emanates from a point source.

As the size of the source increases, aberrations rapidly increase. As a consequence the fraction of reflected light collected drops and, even more importantly, the light collected becomes unrepresentative of the total amount of light reflected from the sample. There is no known geometric surface which can in theory image 100% of the light emanating from an extended diffuse source.

Another problem is that it is difficult to position such a mirrored surface so as to intercept all the reflected light rays and still a further problem is the cost of producing the surface. Producing ellipsoidal surfaces of the required quality is difficult and thus expensive. Also the mirrored surface is again very vulnerable and subject to deterioration.

Other prior art approaches have been based on the use of optical fibres in various configurations. Some of these prior art devices are designed for use in the infrared or near-infrared regions of the spectrum, where the measurement problems are substantially different and they therefore have diminished relevance. Still other such devices are specifically for use with the attenuated total reflectance -7technique, which is quite different from the reflectance measurement problem being considered here.

SUMMARY OF THE INVENTION The present invention provides a probe for measuring reflectance of light from a sample surface. The probe includes a means for directing a light beam along an axis towards a sample surface, the probe being positionable relative to the sample surface such that the axis extends substantially normally to that surface, with the sample surface spaced from the light beam directing means. A light detector of the probe is located, in use, opposite the sample surface and preferably adjacent to and surrounding the light beam directing means, and a surrounding reflecting surface is located between the light detector and the sample surface. Light from the light beam which is reflected by the sample 15 surface either impinges directly on the detector or is reflected by the surrounding reflecting surface to the detector. The surrounding reflecting surface is shaped the detector is relatively sized such that reflected light is prevented from impinging on the sample surface to be re-reflected. Preferably reflecting surface and detector are also designed to ensure all reflected light reaches the detector after a minimal number of reflections from the surrounding reflecting surface. The '****surrounding reflecting surface effectively provides a specular reflecting collector S. surface.

Prior art attempts to use specular reflecting collector surfaces were based on conventional optical imaging concepts. This is needlessly restrictive in at least two regards. Firstly, there is no need to create an image at the detector location, all that is necessary is that substantially all the light from the sample reaches the detector surface. Secondly, it is not necessary for all rays of light to undergo the same number of reflections from the collecting surface. All that is necessary is that all the light reflected from the sample whether diffuse or specular reaches the detector after a finite and small number of reflections from a highly reflective surface and without re-impinging on the sample. The invention meets these criteria. In the invention the number of reflections is kept small to ensure the light -8lost in these reflections does not become significant. This is not just to reduce overall light loss, it also affects the inherent accuracy of the probe. As an example, consider a design which directed specularly reflected rays, emanating from the sample, to the detector after very few reflections but directed diffusely reflected rays, emanating from the sample, to the detector only after many reflections. The diffuse rays would suffer greater attenuation. Such a probe would falsely record the reflectance of a highly specular sample as greater than a highly diffuse sample of identical total reflectance.

The maximum number of reflections from the surrounding reflecting surface, that is, from the collector surface, which is allowed for any ray of light is chosen 0'.i depending on the overall accuracy required from the probe and generally this will be fifteen or less, although as stated above the lower the number of reflections S" the better. In an embodiment described below the maximum number of reflections is one with a significant percentage of all rays travelling directly from the sample surface to the detector without contacting the reflecting collector surface at all. Preferably the member of reflections allowed from the surrounding reflecting surface is between one and five.

Preferably the means for directing a light beam along an axis is an optical fibre.

Thus an optical fibre may pass through a small hole at the centre of the detector and terminate adjacent the detecting surface of the detector such that a conical beam of light from the optic fibre end is directed onto the sample surface.

Preferably the probe comprises a body or cell which is made from a suitable solid material which is transparent to the wavelength of the light being used. Thus the detector and light beam directing means may be mounted on one face of this cell of material which provides an opposite face at which the sample surface is relatively locatable. An outer surface of the cell extending between this one face and the opposite face is shaped and possibly coated to provide the surrounding reflecting surface, as described in more detail below.

-9- Embodiments of the invention will now be described, by way of non-limiting example only, with reference to the accompanying drawings.

BRIEF DESCRIPTION OF DRAWINGS Fig. 1 schematically illustrates a first embodiment.

Fig. 2 illustrates some design parameters for the invention.

Fig. 3 schematically illustrates another embodiment and Fig. 4 illustrates a cell formation for a probe according to another embodiment.

DETAILED DESCRIPTION OF EMBODIMENTS A general form of embodiment for the invention is explained by reference to figure 1 "1 which illustrates a probe having cylindrical symmetry. Thus a mirrored surface 3 of a body (ie. a surrounding reflecting surface) in this figure is conical in form and a detector 2 is circular with a small hole 7 in the middle through which an optic fibre 1 (ie the means for directing a light beam along an axis) presents.

In this configuration, light is directed onto the sample 4 through optic fibre 1 in a conical beam 5 along an axis 8. The angle of divergence of beam 5 is determined by the numerical aperture of the fibre 1 and is thus easily calculated according to the formula sin(half angle of cone of light) numerical aperture of fibre I refractive index of medium (1) Some of the diffusely reflected light will travel directly from sample 4 to detector 2 as shown by reflected ray 9. Other rays 6 will strike the mirrored surface 3 and be reflected onto the detector 2. The detector 2 may be of any appropriate type but for the UVNIS region of the spectrum a silicon detector is the preferred detector.

There will be a minimum angle for the conical surface 3 to ensure that even the most extreme ray will reach the detector after a single reflection. The most extreme ray is one emanating nearly parallel to the sample 4 surface and figure 2 shows the means by which the critical angle c may be computed. In practice, the detector 2 surface and fibre I numerical aperture may represent the starting point for the design with the length L and sample port size S (see Fig. 2) being chosen to meet the single reflection criterion. It should be appreciated that a conical reflecting surface 3 is used here as a simple example of the form of the probe. In practice other profiles are possible and indeed preferred as described below.

If mirrored surface 3 is produced in the conventional way the surface would be vulnerable to damage and deterioration. This eventuality is avoided in embodiments of the invention by making the region or cell 10 through which the incident and reflected light rays pass from some suitable solid but transparent material. Many suitable materials exist. For example, for operation in the visible region, glass or clear plastic such as acrylic may be used. For operation in the UVNIS regions quartz or Sapphire may be used. Possible materials are however not limited to these and the final choice may be determined by other factors as well. For example, a hard material such as Sapphire may be preferred to minimise scratching of the sample face. Refractive index is also an important consideration both because it affects the divergence of the incident beam 5 as indicated in the formula above and because it may affect the operation of the reflective surface 3 as described below.

The mirrored surface 3 in Figure 1 may be provided by making the outside surface of this cell 10 conform to the profile required of the mirrored surface and then coating that surface with a highly reflective material. The faces serving as detector port and sample port would of course be left uncoated. Some possible coatings are an evaporated layer of Aluminium (for UVNIS) or Gold (for NIR regions). Alternatively the coating may be formed from a number of dielectric layers so as to form an interference filter. It is also possible to use a coating with a refractive index significantly lower that that of the cell 10 material and rely on total internal reflection at the cell surface. A special case of that approach is to leave the surface uncoated but in contact with air. It will be appreciated that many other reflective coatings could be devised without departing from the spirit of the above disclosure.

-11- Since the reflective surface of importance is that between the cell 10 and the coating 3 the exposed surface of the coating may be protected by a variety of means. For example additional cladding could be applied. Alternatively the entire cell 10 may be glued or otherwise fastened into a housing 12 which protects the coating, cell, detector and optical fibre as shown in figure 3. In this way the entire measurement cell may be sealed and rendered largely immune from contamination or degradation in use. It is also possible with a configuration as illustrated in Fig. 3 to make reflectance measurements of liquids such as paint by immersing the end of the probe in the liquid, something that would instantly destroy an integrating sphere of the prior art.

Typical diameters for the detector 2 and fibre 1 are about 10mm and 0.6mm with the hole 7 in the detector being about 1 mm. Using these diameters, reflected Iin S. S light loss through the central hole 7 of the detector 2 is only about In practice, if the detector 2 is a silicon detector, there will be an intrinsic region around the hole 7 which is not sensitive to light. This increases the total light loss to about but this does not significantly affect either accuracy or signal to noise levels.

S

S

In practice, there are three other factors which should be considered in the detailed design of the cell 10. Firstly, some samples allow penetration of incident light to some depth below the surface of the sample. This light interacts with the sample below the surface and is reflected back to emerge potentially at some distance from the point at which it entered. Thus the patch of sample emitting reflected light is larger than the size of the incident light patch. If the size of the sample port is only made just big enough to accommodate the incident light patch from beam 5 then some of the reflected light will be excluded from the measurement giving an erroneous result. In practice it has been found necessary to make the sample port at least 2 mm larger than the incident light patch.

A second factor which should be considered is that to just bring the light back to the detector 2 surface is inadequate. The angle of incidence at this surface should be considered because the detector sensitivity varies with angle of -12incidence, especially for angles approaching grazing incidence. Indeed, if the angle is too far from normal to the surface and there is an air space between the cell and the detector 2 the rays may be totally internally reflected within the cell and not reach the detector at all. This problem can be avoided by using a refractive index matching material between the cell 10 and detector 2 (for example silicone grease). Nonetheless, ideally the light should reach the detector surface as close to normal as possible.

Thirdly, it should be noted that when a highly specular sample 4 is being measured, very little of the light will reflect off the mirrored surface 3. Indeed, if the cell 10 is made very short (ie L is small) the image generated by a specular *sample on the detector 2 may be significantly smaller than the detector port aperture, which is not ideal. Ideally the image from a highly specular sample S"should come as close as possible to filling the area of the detector, otherwise the l• proportion lost in the blind centre of the detector, while small, will be different for diffuse and specular samples leading to small but non zero errors.

o..o When these three factors are taken into account it becomes significantly more difficult to meet all the goals with a simple conical reflecting surface 3. Other

C

profiles, while more expensive to produce, are capable of achieving better overall performance. In particular a constant radius profile shown in figure 4 is preferred over the conical profile shown in Fig. 1 as a profile capable of meeting the performance requirements while still being relatively easy to produce. Many other profiles are possible; for example, a parabolic profile may be used, or a profile made up of conical sections of different angles, or a profile resulting from a hemisphere with a flat sample face ground parallel to the detector face etc. Many possible profiles can be devised to give acceptable results without departing from the core concept of this invention which is to prevent reflected light from impinging on the sample surface.

Comparative tests on a probe according to the invention and a Labsphere DRA (a prior art integrating sphere type reflectometer which is claimed to be a high performance sphere) gave at least comparable results. However the invention, -13whilst providing comparable results to the prior art, provides a much more robust and portable probe.

The invention described herein is susceptible to variations, modifications and/or additions other than those specifically described and it is to be understood that the invention includes all such variations, modifications and/or additions which fall within the spirit and scope of the following claims.

SC X:\DELILAH\COMPLETEFIBRE.DOC

S

*g0g *0* o 4oo

Claims (14)

1. A probe for measuring reflectance of a sample surface including a body positionable against the sample surface, means for directing a beam of light across the body along an axis which extends substantially normally to the sample surface for the light to be reflected from that surface, a light detector carried by the body to be located opposite the sample surface, the body providing a surrounding reflecting surface which, in use, is located between the sample surface and the detector, the surrounding reflecting surface being shaped and the detector being relatively sized for the reflected light to be detected without again being reflected from the sample surface. •1 *0 4 S 15

2. A probe as claimed in claim 1 wherein the surrounding reflecting surface is shaped and the detector is relatively sized such that the reflected light reaches 4 the detector either directly or after a minimal number of reflections being fifteen or less. 4o

3. A probe as claimed in claim 2 wherein the minimal number of reflections is between one and five.

4. A probe as claimed in any one of claims 1 to 3 wherein the means for directing a beam of light is an optical fibre having an end an which, in use, is spaced from the sample surface for a conical beam of light from the optical fibre to be directed onto the sample surface.

A probe as claimed in any one of claims 1 to 4 wherein the body is solid and formed from a material which is transparent to the wavelength of the light of the beam.

6. A probe as claimed in claim 5 wherein the body includes opposite faces joined by a continuous surface, one of said faces having mounted thereon the means for directing a beam of light and the detector, the opposite face being positionable against the sample surface, the continuous surface providing the surrounding reflecting surface.

7. A probe as claimed in claim 6 wherein the continuous surface is coated with a reflective material.

8. A probe as claimed in claim 6 wherein the continuous surface is coated with a number of layers of dielective material to form an interference filter.

9. A probe as claimed in claim 6 wherein the continuous surface is coated .i :with a material having a refractive index lower than that of the material of the body such that reflected light is totally internally reflected at the continuous surface of the body.

S A probe as claimed in claim 6 wherein the continuous surface is uncoated and the body, in use, is surrounded by a medium having a refractive index lower than that of the material of the body such that reflected light is totally internally reflected at the continuous surface of the body.

11. A probe as claimed in any one of claims 1 to 10 wherein the body is contained in a protective housing or coated by a protective material.

12. A probe as claimed in any one of claims 1 to 11 wherein the body is frusto- conical in shape.

13. A probe as claimed in any one of claims 1 to 11 wherein the surrounding reflecting surface has a convexly curved constant radius profile.

14. A probe as claimed in any one of claims 1 to 11 wherein the surrounding reflecting surface has a convexly curved parabolic profile. -16- A probe for measuring reflectance of a sample surface substantially as hereinbefore described with reference to the accompanying figures. DATED: 17 March, 2000 PHILLIPS ORMONDE FITZPATRICK Attorneys for: VARIAN AUSTRALIA PTY LTD 9 0 9 0 9996 6 09