BACKGROUND OF THE INVENTION

The present invention relates generally to refractometry and, more particularly, concerns a method and apparatus for improving the accuracy and reliability of critical-angle refractometry measurements.

The purpose of a refractometer is to measure the refractive index of a sample material M, for example, a liquid. A prior art critical-angle refractometer is shown schematically in figure 1. The liquid sample M is typically contained in a dish in which one surface of an optical prism P, having a high index of refraction, forms the floor of the dish. Rays of light R from a Light Source S are focused by a Lens LI into prism P. It is characteristic of the interface between two disparate optical media that a light beam entering the interface will be, in part, refracted or bent upon traversing the interface, and it will be, in part, reflected from the interface. However, for any interface, there is a "critical-angle" such that a light beam incident on the interface with that or a greater angle to the normal will be totally reflected. The critical angle is defined by Equation (1). sin (critical angle) = (refractive index of sample)/(refractive index of prism) (1)

By design of the prism/sample interface I in figure 1, rays R incident on interface I are at least partially reflected therefrom. The reflected rays exit the prism and are directed by a lens L2 onto a linear optical detector O. The detector O is a linear array of light sensitive pixels, each of which produces an electrical signal proportional to the level of light incident on it. Each pixel receives a bundle of light rays that reflect from interface I at a particular angle of incidence. If this angle of incidence is greater than the critical angle, then 100% of the rays in that bundle are reflected from the interface I. This is called total internal reflection (TIR). If this angle of incidence is less than the critical angle, then the rays are only partially reflected by the prism/sample interface, because a portion of the light is refracted into the liquid sample and, ideally, does not reach the detector. A pixel receiving partially reflected light will sense a lower intensity of light than a pixel receiving totally reflected light. The light intensity pattern on the array of pixels therefore provides a map of the amount of reflection at the prism/sample interface of the bundle of light beams corresponding to each pixel. From the light intensity detected at each pixel (and the associated angle of incidence of the light beams it receives), the reflectance at the prism/sample interface I can be calculated and the result represented as a graph of reflectance versus angle of incidence (or pixel), i.e. a reflectance graph. An exemplary reflectance graph is illustrated in figure 2. As can be seen, the graph exhibits 100% reflection (TIR) over a range of pixels corresponding to low angles of incidence relative to interface I (high angles of incidence relative to the normal) and then transitions sharply to a relatively low percentage value of reflectance. The determination of the refractive index of a sample is performed by locating the pixel where the transition from total to partial reflection occurs in the graph of figure 2 and using the associated angle of incidence, together with the critical angle equation (1) above, to solve for refractive index. A convenient way to locate the transition is to search for the region of pixels where the reflectance has the greatest slope or rate of change. Because of the limited pixel resolution of existing detectors, interpolation between the pixels is usually also required. The range of refractive index values that can be measured is a function of the range of angles of incidence of the light rays that reach the detector, and a function of the refractive index of the prism material. A typical range that can be measured is refractive indices from 1.3 to 1.7.

Those skilled in the art will appreciate that, although the reflectance graph is shown visually in figure 2 for illustrative purposes, it need not exist in that form. For example, it may merely be a collection of data in a computer which can be processed in the manner described herein.

With existing refractometers, a number of operating conditions may lead to a false result being reported. For example, the user may not have cleaned the prism adequately prior to loading the liquid sample. This can result in a mixture of materials at the prism/sample interface. Sometimes, the user simply forgets to load a new sample and, instead, accidentally performs a duplicate measurement on the previously loaded sample. Another difficulty arises when an insufficient quantity of liquid sample is placed in the prism dish. If the quantity of liquid in the dish is too small, the upper free surface (or meniscus) of the liquid sample may be close enough to interface I to reflect some of the light back into the system that ideally should not reach the detector. This will result in incorrect reflectances being calculated for the pixels receiving this additional light. Users may also attempt to measure samples that cannot be well characterized by a single refractive index, for example inhomogeneous materials like colloidal suspensions. Finally, users may attempt to measure samples that have refractive indices that are outside the range of refractive index that the instrument can measure.

Broadly, it is an object of the present invention to avoid shortcomings of existing refractometry and methods and systems. It is specifically contemplated that the invention should prevent users from making false measurements in all of the situations discussed above.

In accordance with one aspect of the present invention, a critical-angle refractometer which utilizes an in image of light reflected from an optical interface with a vessel containing a sample under test to determine an optical property of the sample, sample properties are evaluated to prevent improper testing of the sample. This evaluation includes establishing reflectance information associating the amount of reflection with locations in the image; and utilizing a plurality of properties of the reflectance information to determine if the vessel contains a proper sample under test.

Preferably, the properties include the point of transition from complete to partial reflection and at least one of the following: the average rate of change of reflectance as a function of position in the image in a predefined range of interest (ROI) about the point of transition; the maximum reflectance in the ROI; the minimum reflectance in the ROI; the range of reflectance in the ROI; and the average reflectance in the ROI.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing brief description and further objects, features, and advantages of the present invention will be understood more completely from the following detailed description of a presently preferred, but nonetheless illustrative, embodiment in accordance with the present invention, with reference being hand to the accompanying drawings, in which:

Figure 1 is a schematic representation of a prior art critical-angle refractometer;

Figure 2 is an exemplary reflectance graph obtained with a prior art prior art critical- angle refractometer;

Figure 3 illustrates a window or region of interest (ROI) established around the transition in a reflectance graph;

Figure 4 illustrates the reflectance graph corresponding to a "Dish Clean" condition;

Figure 5 illustrates the reflectance graph corresponding to a "Sample Index Out of Range" condition; Figures 6 and 7 illustrate reflectance graphs corresponding to a "Sample Present" condition, with figure 6 illustrating a poor quality sample and figure 7 illustrating a normal quality sample.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

In addition to determining the refractive index by locating the pixel in the reflectance graph (e.g. figure 2) where the transition from total to partial reflection occurs, embodiments in accordance with the present invention also characterize the transition with three additional parameters determined from the pixels and their associated reflectances. After the transition pixel is determined, a window or region of interest (ROI) is established around the transition (see figure 3). The three additional parameters are: the average steepness or slope of the graph across the ROI (scaled to 100% for a perfect sample), the maximum reflectance in the ROI, and the minimum reflectance in the ROI. Equivalently, the range and average of reflectance across ROI may be used instead of the maximum and minimum.

In a preferred embodiment, these parameters are used to discriminate among four possible states at the prism/sample interface:

1. Sample Present

2. Dish Clean

3. Sample Index Out of Range

4. Other

When the dish floor (interface I) is clean, all of the ray bundles undergo TIR at the prism/sample interface, and for each pixel across the detector a reflectance close to 100% is calculated. Such a graph is shown in figure 4. The transition region of interest, which is located by searching for the region of pixels where the reflectance has the greatest slope or rate of change, will be found at a random pixel determined by the noise in the signal. Since for this example all regions are the same, the location is not important to the analysis. The three parameters may be compared with stored preset values or thresholds as per the following example: If the reflectance range across the ROI is less than 5%, and the minimum reflectance is greater than 95%, and the average scaled slope is less than 5%, then the prism/sample interface state is "Dish Clean". In other words, this condition is characterized by a relatively constant reflectance range, a minimum reflectance above a high threshold and an average slope below a low threshold.

When the user loads a sample that has a refractive index that is above the range of the instrument, none of the ray bundles undergoes TIR at the prism/sample interface, and for each pixel across the detector a low reflectance is calculated. Such a graph is shown in figure 5. The transition region of interest, which is located by searching for the region of pixels where the reflectance has the greatest slope or rate of change, will again be found at a random pixel determined by the noise in the signal. Since for this example as well all regions are the same, the location is not important. The three parameters may be compared with stored preset values or thresholds as per the following example: If the reflectance range across the ROI is less than 5%, and the maximum reflectance is less than 5%, and the average scaled slope is less than 5%, then the prism/sample interface state is "Sample Index Out of Range". In other words, this condition is characterized by a relatively constant reflectance range, a maximum reflectance below a low threshold and an average slope below a low threshold

When the user loads a sample that has a refractive index that is within the range of the instrument, some of the ray bundles undergo TIR at the prism/sample interface. Exemplary graphs are shown in figures 6 and 7 for poor quality and normal quality samples, respectively. The transition region of interest is located by searching for the region of pixels where the reflectance has the greatest slope or rate of change. The three parameters may be compared with stored preset values or thresholds as per the following example: If the reflectance range across the ROI is greater than 20%, and the maximum reflectance is greater than 95%, and the average scaled slope is greater than 50%, then the prism/sample interface state is "Sample Present". In other words, this condition is characterized by a relatively large reflectance range (above a high threshold), a maximum reflectance above a high threshold and an average slope above a high threshold.

When the user loads a sample and a comparison of the three parameters with stored preset values does not satisfy the conditions for "Sample Present", "Dish Clean", or "Sample Index Out of Range"; then the prism/sample interface state is "Other". This may occur, for example, because scaled slope across the ROI has been lowered by lack of homogeneity in the sample or contamination at the prism/sample interface (see figure 7). It also may occur because additional light reflected from the sample meniscus has reduced the reflectance range across the ROI. Alternatively, the user may have only partially covered the interface with sample, so that more than one distinct refractive index is present at the boundary. This also will reduce the scaled slope over the ROI.

To prevent errors that are caused by the user forgetting to unload, clean, or reload the sample, the present invention also incorporates a computerized process (state machine) driven by the transitions in the prism/sample interface state. This "Message State" process provides instructions to the user in the form of indicators or messages displayed on a screen of the refractometer. The three Message States are:

1. Unloading

2. Loading

3. Measuring

"Unloading" is the initial Message State. An example Message State process is described by the following pseudo-code:

Start

Message State = "Unloading"

If the Message State is "Unloading"

If the prism/sample interface state is "Other"

Display message: "Sample quality is too low to measure.

Please clean dish. "

If the prism/sample interface state is "Sample Present"

Display message: "Please remove sample and clean dish. "

If the prism/sample interface state is "Sample Index Out of

Range"

Display message: "Sample index too high to measure.

Please clean dish. "

If the prism/sample interface state is "Dish Clean"

Transition to Message State "Loading"

If the Message State is "Loading"

Display message "Please load sample. "

If the prism/sample interface state is "Sample Present"

Transition to Message State "Measuring"

If the prism/sample interface state is "Other" or "Sample Index

Out of Range"

Transition to Message State "Unloading"

If the Message State is "Measuring"

Display message "Measuring, please wait. "

If measurements are complete for this sample

Transition to Message State "Unloading"

Go to Start For proper operation of this algorithm it is understood that the prism/sample interface state transitions must be suitably debounced to give the operator time to clean the dish or load samples.

The software of the instrument can be configured to only allow measurement while the Message State is in the state "Measuring". This prevents the user from performing a measurement without unloading prior samples and cleaning the prism/sample interface. The user is also reminded to unload the sample after measurements are complete. A further benefit is that the measurement can be initiated automatically when the Message State transitions to "Measuring" so that the user is not required to press a button to start the measurement.

For differing types of samples, it is useful to adjust the stored preset values (thresholds) that are compared to the three parameters used to discriminate among the prism/sample interface states. For example, for a given sample it may be known a priori that the scaled slope over the ROI should never be below 80%, whereas for other samples it may be determined empirically that a lower value of scaled slope is acceptable. Therefore the present refractometer includes a database of named samples that includes user configurable fields to change the stored preset values used to discriminate among the prism/sample interface states.

Although a preferred embodiment of the invention has been disclosed for illustrative purpose, those skilled in the art will appreciate that many additions, modifications and substitutions are possible without departing from the scope and spirit of the invention, as defined by the accompanying claims.