# US 9,518,941 B1 · Weight-percent analysis for prompt gamma neutron activation substance analyzers > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 9,518,941 B1 - Document ID: US9518941B1 - Status: Granted - Kind: B1 - Application number: 15/224,189 - Filed: 2016-07-29 - Granted: 2016-12-13 - Earliest claimed: 2016-07-29 - Assignees: Sabia Inc. - Applicants: Sabia Inc. - Inventors: Clinton LaMar Lingren; James Francis Miller; Stephen James Foster - Web page: https://patentgrep.com/patent/US9518941B1 ## CPC - G01N 23/222: using neutron activation analysis [NAA] - G01N 2223/0745: neutron-gamma activation analysis - G01N 2223/643: object on conveyor - G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q) ## Abstract Methods and systems are provided to improve PGNAA substance analyzers. In one aspect, an analyzer includes: a source of neutrons; an opening to receive a substance; a gamma ray detector; and computational device(s) configured to receive spectral data of detected gamma rays, perform a regression algorithm on the spectral data using spectral responses for known atomic elements to determine coefficients of spectrum, sum effective weight values (corresponding to the coefficients of spectrum) together to form a total effective weight value for the substance being analyzed, divide effective weight values by the total effective weight value for the substance to generate effective weight-percent values corresponding to two or more respective ones of the atomic elements detected in the spectral data, and generate final weight-percent values based on a correlation of previous effective weight-percent values obtained for known samples with elemental or molecular weight-percent values obtained for the known samples. ## Claims ### Claim 1 A PGNAA (Prompt-Gamma, Neutron-Activation Analysis) substance analyzer comprising: a source of neutrons; an opening arranged with respect to the source of neutrons to receive a substance; a gamma ray detector arranged with respect to the opening to detect gamma rays emanating at least from the substance in response to absorption of neutrons by the substance; and one or more computational devices configured to receive spectral data of the detected gamma rays, perform a regression algorithm on the spectral data using a data repository of spectral responses for known atomic elements to determine coefficients of spectrum for atomic elements detected in the spectral data, sum effective weight values, corresponding to the coefficients of spectrum, together to form a total effective weight value for the substance being analyzed, divide each of two or more of the effective weight values by the total effective weight value for the substance being analyzed to generate effective weight-percent values corresponding to two or more respective ones of the atomic elements detected in the spectral data, and generate final weight-percent values from the effective weight-percent values, for the substance being analyzed, based on a correlation of (a) previous effective weight-percent values obtained for known samples with (b) elemental or molecular weight-percent values obtained for the known samples. ### Claim 2 (depends on 1) The PGNAA substance analyzer of claim 1, wherein the effective weight values are the coefficients of spectrum determined by the regression algorithm. ### Claim 3 (depends on 2) The PGNAA substance analyzer of claim 2, wherein the final weight-percent values are molecular weight-percent values for the substance being analyzed. ### Claim 4 (depends on 2) The PGNAA substance analyzer of claim 2, wherein the regression algorithm comprises MLR (multiple linear regression). ### Claim 5 (depends on 4) The PGNAA substance analyzer of claim 4, wherein the spectral responses for the known atomic elements are provided in the data repository in units of counts per minute per gram. ### Claim 6 (depends on 1) The PGNAA substance analyzer of claim 1, wherein the one or more computational devices are configured to generate the effective weight values in accordance with a determined relationship between magnitudes of elemental coefficients and weight values of elements or molecules. ### Claim 7 (depends on 6) The PGNAA substance analyzer of claim 6, wherein the one or more computational devices are configured to generate the effective weight values by multiplying the coefficients of spectrum by respective ratios of molecular weight to atomic weight for detected elements. ### Claim 8 (depends on 1) The PGNAA substance analyzer of claim 1, wherein the source of neutrons is an isotopic radioactive source, and the PGNAA substance analyzer comprises: a neutron moderating material arranged with respect to the isotopic radioactive source to slow velocities of neutrons from the isotopic radioactive source; and a radiation shielding material arranged with respect to the opening and the gamma ray detector to reduce an amount of radiation escaping the substance analyzer. ### Claim 9 (depends on 1) The PGNAA substance analyzer of claim 1, wherein the source of neutrons is an electrically powered neutron generator. ### Claim 10 (depends on 1) The PGNAA substance analyzer of claim 1, wherein the one or more computational devices are configured to: receive calibration spectral data of gamma rays detected from the known samples; perform the regression algorithm on the calibration spectral data using the data repository of spectral responses for known atomic elements to determine calibration coefficients of spectrum for atomic elements detected in the calibration spectral data; sum calibration effective weight values, corresponding to the calibration coefficients of spectrum, together to form a total calibration effective weight value for each of the known samples; divide each of two or more of the calibration effective weight values by the total calibration effective weight value for each respective known sample to generate the previous effective weight-percent values for the known samples; and correlate the previous effective weight-percent values with the elemental or molecular weight-percent values obtained for the known samples to generate calibration data; wherein generating the final weight-percent values comprises applying the calibration data to the effective weight-percent values for the substance being analyzed. ### Claim 11 A PGNAA (Prompt-Gamma, Neutron-Activation Analysis) substance analyzer comprising: an isotopic radioactive source of neutrons; a neutron moderating material arranged with respect to the isotopic radioactive source to slow velocities of neutrons from the isotopic radioactive source; an opening arranged with respect to the source of neutrons to receive a substance; a gamma ray detector arranged with respect to the opening to detect gamma rays emanating at least from the substance in response to absorption of neutrons by the substance; and one or more computational devices configured to receive spectral data of the detected gamma rays, perform a regression algorithm on the spectral data using a data repository of spectral responses for known atomic elements to determine coefficients of spectrum for atomic elements detected in the spectral data, use the coefficients of spectrum directly as effective weight values of the atomic elements detected in the spectral data, sum the effective weight values together to form a total effective weight value for the substance being analyzed, divide each of two or more of the effective weight values by the total effective weight value for the substance being analyzed to generate effective weight-percent values for the atomic elements detected in the spectral data, and generate molecular weight-percent values from the effective weight-percent values, for the substance being analyzed, based on a correlation of (a) previous effective weight-percent values obtained for known samples with (b) molecular weight-percent values obtained for the known samples. ### Claim 12 (depends on 11) The PGNAA substance analyzer of claim 11, wherein the regression algorithm comprises MLR (multiple linear regression). ### Claim 13 (depends on 12) The PGNAA substance analyzer of claim 12, wherein the spectral responses for the known atomic elements are provided in the data repository in units of counts per minute per gram. ### Claim 14 (depends on 11) The PGNAA substance analyzer of claim 11, wherein the one or more computational devices are configured to: receive calibration spectral data of gamma rays detected from the known samples; perform the regression algorithm on the calibration spectral data using the data repository of spectral responses for known atomic elements to determine calibration coefficients of spectrum for atomic elements detected in the calibration spectral data; use the calibration coefficients of spectrum directly as calibration effective weight values of the atomic elements detected in the spectral data, sum the calibration effective weight values together to form a total calibration effective weight value for each of the known samples; divide each of two or more of the calibration effective weight values by the total calibration effective weight value for each respective known sample to generate the previous effective weight-percent values for the known samples; and correlate the previous effective weight-percent values with the molecular weight-percent values obtained for the known samples to generate calibration data; wherein generating the molecular weight-percent values comprises applying the calibration data to the effective weight-percent values for the substance being analyzed. ### Claim 15 A non-transitory machine readable medium encoding instructions operable to receive PGNAA (Prompt-Gamma, Neutron-Activation Analysis) spectral data and to cause a computer to perform operations comprising: running a regression algorithm on the PGNAA spectral data using a data repository of spectral responses for known atomic elements to determine coefficients of spectrum for atomic elements detected in the PGNAA spectral data; summing effective weight values, corresponding to the coefficients of spectrum, together to form a total effective weight value for the substance being analyzed; dividing each of two or more of the effective weight values by the total effective weight value for the substance being analyzed to generate effective weight-percent values corresponding to two or more respective ones of the atomic elements detected in the PGNAA spectral data; and generating final weight-percent values from the effective weight-percent values, for the substance being analyzed, in accordance with calibration data generated by correlating effective weight-percent values obtained for known samples with elemental or molecular weight-percent values obtained previously for the known samples. ### Claim 16 (depends on 15) The non-transitory machine readable medium of claim 15, wherein running the regression algorithm on the PGNAA spectral data comprises performing MLR (multiple linear regression) processing of the PGNAA spectral data. ### Claim 17 (depends on 15) The non-transitory machine readable medium of claim 15, wherein the effective weight values are the coefficients of spectrum determined by the regression algorithm. ### Claim 18 (depends on 17) The non-transitory machine readable medium of claim 17, wherein the spectral responses for the known atomic elements are provided in the data repository in units of counts per minute per gram. ### Claim 19 (depends on 15) The non-transitory machine readable medium of claim 15, wherein the operations comprise generating the effective weight values in accordance with a determined relationship between magnitudes of elemental coefficients and weight values of elements or molecules. ### Claim 20 (depends on 19) The non-transitory machine readable medium of claim 19, wherein generating the effective weight values comprises multiplying the coefficients of spectrum by respective ratios of molecular weight to atomic weight for detected elements. ### Claim 21 (depends on 19) The non-transitory machine readable medium of claim 19, wherein the operations comprise: receiving calibration spectral data of gamma rays detected from the known samples; running the regression algorithm on the calibration spectral data using the data repository of spectral responses for known atomic elements to determine calibration coefficients of spectrum for atomic elements detected in the calibration spectral data; summing calibration effective weight values, corresponding to the calibration coefficients of spectrum, together to form a total calibration effective weight value for each of the known samples; dividing each of two or more of the calibration effective weight values by the total calibration effective weight value for each respective known sample to generate the effective weight-percent values obtained for the known samples; and correlating the effective weight-percent values obtained for the known samples with the elemental or molecular weight-percent values obtained for the known samples to generate the calibration data. ### Claim 22 A non-transitory machine readable medium encoding instructions operable to receive PGNAA (Prompt-Gamma, Neutron-Activation Analysis) spectral data and to cause a computer to perform operations comprising: running a regression algorithm on the PGNAA spectral data using a data repository of spectral responses for known atomic elements to determine coefficients of spectrum for atomic elements detected in the PGNAA spectral data; summing the coefficients of spectrum together to form a total effective weight value for the substance being analyzed; dividing each of two or more of the coefficients of spectrum by the total effective weight value for the substance being analyzed to generate effective weight-percent values for the atomic elements detected in the PGNAA spectral data; and generating molecular weight-percent values from the effective weight-percent values, for the substance being analyzed, in accordance with calibration data generated by correlating effective weight-percent values obtained for known samples with molecular weight-percent values obtained previously for the known samples. ### Claim 23 (depends on 22) The non-transitory machine readable medium of claim 22, wherein running the regression algorithm on the PGNAA spectral data comprises performing MLR (multiple linear regression) processing of the PGNAA spectral data. ### Claim 24 (depends on 22) The non-transitory machine readable medium of claim 22, wherein the spectral responses for the known atomic elements are provided in the data repository in units of counts per minute per gram. ### Claim 25 (depends on 22) The non-transitory machine readable medium of claim 22, comprising: receiving calibration spectral data of gamma rays detected from the known samples; running the regression algorithm on the calibration spectral data using the data repository of spectral responses for known atomic elements to determine calibration coefficients of spectrum for atomic elements detected in the calibration spectral data; summing the calibration coefficients of spectrum together to form a total calibration effective weight value for each of the known samples; dividing each of two or more of the calibration coefficients of spectrum by the total calibration effective weight value for each respective known sample to generate the effective weight-percent values obtained for the known samples; and correlating the effective weight-percent values obtained for the known samples with the molecular weight-percent values obtained for the known samples to generate the calibration data.