Automated mineral classification

Patent metadata

Field	Value
Publication number	US-9778215-B2
Application number	US-201213662072-A
Country	US
Kind code	B2
Filing date	Oct 26, 2012
Priority date	Oct 26, 2012
Publication date	Oct 3, 2017
Grant date	Oct 3, 2017

Abstract

The present invention discloses a combination of two existing approaches for mineral analysis and makes use of the Similarity Metric Invention, that allows mineral definitions to be described in theoretical compositional terms, meaning that users are not required to find examples of each mineral, or adjust rules. This system allows untrained operators to use it, as opposed to previous systems, which required extensive training and expertise.

First claim

We claim as follows:

1. A method for determining the mineral content of a sample, said method comprising:
placing a sample in a scanning electron microscope;
directing an electron beam to a point on the sample to obtain an x-ray spectrum of the sample and a backscatter electron intensity value, the x-ray spectrum and the backscatter electron intensity value corresponding to a data point, the data point having an associated mineral composition;
sequentially applying rules from a first set of rules to the x-ray spectrum of the data point, wherein each rule identifies a mineral composition and comprises one or more criteria corresponding to the mineral composition that the rule identifies, and applying the rule comprises comparing the x-ray spectrum of the data point to the one or more criteria of the rule;
identifying the mineral composition of the data point as the mineral composition of the first rule of the first set of rules applied to the x-ray spectrum of the data point in which the one or more criteria are satisfied;
if and only if the x-ray spectrum of the data point fails to satisfy the one or more criteria of any of the rules of the first set of rules, comparing the x-ray spectrum of the data point to x-ray spectra of different mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point; and
if the value of a metric of similarity between the x-ray spectrum of the data point and the x-ray spectrum of the mineral definition for the best matching mineral in the library meets a predetermined threshold, identifying the mineral composition of the data point as the mineral composition corresponding to the best matching mineral defined in the library.

2. The method of claim 1 , further comprising if the value of the metric of similarity between the x-ray spectrum of the data point and the x-ray spectrum of the best matching mineral defined in the library fails to meet the predetermined threshold:
sequentially applying rules from a second set of rules to the data point, wherein each rule of the second set of rules identifies a mineral composition and comprises one or more criteria corresponding to the mineral composition that the rule identifies, and applying the rule comprises comparing the data point to the one or more criteria of the rule; and
identifying the mineral composition of the data point as the mineral composition of the first rule of the second set of rules applied to the data point in which the one or more criteria are satisfied.

3. The method of claim 1 , in which comparing the x-ray spectrum of the data point to x-ray spectra of different mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point further comprises comparing the backscatter electron intensity value of the data point to backscatter electron intensity values corresponding to minerals defined in the library.

4. The method of claim 1 , in which sequentially applying rules from a first set of rules to the x-ray spectrum of the data point identifies bad data points or identifies parts of the sample that do not need detailed analysis.

5. The method of claim 1 , in which comparing the x-ray spectrum of the data point to x-ray spectra of different mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point comprises calculating similarity metrics between the data point and each of the mineral definitions in the library of mineral definitions.

6. The method of claim 5 in which calculating similarity metrics includes determining probabilities that the library mineral will produce the values observed for the data point.

7. The method of claim 5 in which calculating similarity metrics includes calculating a chi squared value.

8. The method of claim 1 , in which sequentially applying rules from a first set of rules to the x-ray spectrum of the data point includes comparing the data point to a criterion that includes the height of an x-ray peak at a specific energy range.

9. The method of claim 8 , in which one of the criteria is a backscatter electron intensity value.

10. A scanning electron microscope system for determining the mineral content of a mineral sample, comprising:
an electron beam column for directing an electron beam toward a mineral sample;
a detector for measuring the energy or wavelength of x-rays emitted from the mineral sample in response to an impingement on the mineral sample of electrons in the electron beam to create an x-ray spectrum of a data point;
a processor for executing computer instructions to determine the minerals present in the mineral sample; and
a computer memory containing computer instructions configured for:
sequentially comparing the x-ray spectrum of the data point to criteria within a first set of rules, each criterion corresponding to a mineral composition, and identifying the mineral composition of the data point as the mineral composition corresponding to the first satisfied criterion;
if and only if the x-ray spectrum of the data point fails to satisfy any of the criteria in the first set of rules, comparing the x-ray spectrum of the data point to x-ray spectra of mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point; and
if the match between the x-ray spectrum of the data point and the x-ray spectrum of the mineral definition for the best matching mineral in the library meets a predetermined threshold, identifying the mineral composition of the data point as the mineral composition corresponding to the best matching mineral defined in the library.

11. The scanning electron microscope system of claim 10 , in which the computer instructions further comprise instructions for, if the match between the x-ray spectrum of the data point and the best matching mineral defined in the library fails to meet the predetermined threshold, sequentially comparing the data point to criteria within a second set of rules, each criterion corresponding to a mineral composition, and identifying the mineral composition of the data point as the mineral composition corresponding to the first satisfied criterion within the second set of rules.

12. The scanning electron microscope system of claim 10 , in which the computer instructions for comparing the x-ray spectrum of the data point to x-ray spectra of different mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point further comprise computer instructions for comparing a backscatter electron intensity value of the data point to backscatter electron intensity values corresponding to minerals defined in the library.

13. The scanning electron microscope system of claim 10 , in which the computer instructions for sequentially comparing the x-ray spectrum of the data point to criteria within a first set of rules comprise computer instructions for identifying bad data points or identifying parts of the mineral sample that do not need detailed analysis.

14. The scanning electron microscope system of claim 10 , in which the computer instructions for comparing the x-ray spectrum of the data point to x-ray spectra of different mineral definitions in a library of mineral definitions to determine which of the minerals defined in the library is a best match for the data point comprise computer instructions for calculating similarity metrics between the data po

Assignees

• Fei Co

Inventors

• Owen Michael James
• Buhot Michael

Classifications

• H01J2237/2442

  Energy-dispersive (Si-Li type) spectrometer · CPC title

• H01J2237/24425

  Wavelength-dispersive spectrometer · CPC title

• H01J2237/2807

  X-rays · CPC title

• G01N23/2252 Primary

  Measuring emitted X-rays, e.g. electron probe microanalysis [EPMA] · CPC title

• H01J2237/2804

  Scattered primary beam · CPC title