Determining percent solids in suspension using Raman spectroscopy
Methods and apparatus are provided for determining weight percent of solids in a suspension using Raman spectroscopy. The methods can be utilized to acquire Raman spectral data from the suspension and to determine weight percent of solids in a process being carried out, for example, in a vessel, without the need to remove samples for analysis. The weight percent of the solids can be determined with a desired accuracy in a relatively short time, typically 10 minutes or less. The acquired Raman spectral data may be processed by chemometric software using, for example, a partial least squares algorithm and data pretreatment to provide a predicted value of weight percent solids. In some embodiments, the invention is used to determine the weight percent of microparticles of a diketopiperazine in an aqueous solution.
1. A method for determining weight percent of solids in a suspension, comprising:
mixing the suspension;
acquiring Raman spectral data of the solids in the suspension; and
processing the Raman spectral data to determine weight percent of the solids in the suspension.
2. A method as defined in claim 1 , wherein the weight percent of the solids in the suspension is in a range of 5% to 20%.
3. A method as defined in claim 1 , wherein the solids are microparticles of a diketopiperazine.
4. A method as defined in claim 3 , wherein the microparticles have a size in a range of 0.1 to 34 micrometers.
5. A method as defined in claim 3 , wherein acquiring Raman spectral data includes acquiring data in a range of 1180 to 1800 cm −1 and a range of 2782 to 3136 cm −1 .
6. A method as defined in claim 1 , wherein processing the Raman spectral data includes executing a partial least squares algorithm with Savitzky-Golay first derivative data pretreatment.
7. A method as defined in claim 1 , wherein processing the Raman spectral data includes executing a Partial Least Squares algorithm using two principal components.
8. A method as defined in claim 1 , wherein the Raman spectral data is processed by a computing device executing a partial least squares algorithm with data pretreatment.
9. A method as defined in claim 1 , wherein processing the Raman spectral data includes executing a regression algorithm comprising a partial least squares, multiple linear regression, or principal component regression, with or without data pretreatment.
10. A method as defined in claim 9 , wherein the data pretreatment is selected from multiplicative and extended multiplicative scatter correction, standard normal variate, various derivatives, and normalizations.
11. Apparatus for determining weight percent of solids in a suspension, comprising:
a Raman spectrometer to acquire Raman spectral data of the solids in the suspension; and
a computing device including a processor and a computer-readable storage medium, the computer-readable storage medium containing computer instructions that, when executed by the processor, perform a method of processing the Raman spectral data to determine weight percent of the solids in the suspension.
12. Apparatus as defined in claim 11 , wherein the Raman spectrometer includes a short focus optical device adapted for immersion in the suspension.
13. Apparatus as defined in claim 12 , wherein the Raman spectrometer and the method of processing the Raman spectral data are configured to determine weight percent of microparticles of a diketopiperazine.
14. Apparatus as defined in claim 13 , wherein the Raman spectrometer is configured to acquire Raman spectral data in a range of 1180 to 1800 cm −1 and a range of 2782 to 3136 cm −1 .
15. Apparatus as defined in claim 11 , wherein the Raman spectrometer is configured to acquire Raman spectral data in less than 10 minutes.
16. Apparatus as defined in claim 11 , wherein the Raman spectrometer is configured to acquire Raman spectral data for an exposure time of 45 seconds and for 4 accumulations, and to perform cosmic ray filtration.
17. Apparatus as defined in claim 11 , wherein the processing of the Raman spectral data includes executing a Partial Least Squares algorithm with Savitzky-Golay first derivative data pretreatment.
18. Apparatus as defined in claim 11 , wherein the Raman spectral data is processed by the computing device executing a Partial Least Squares algorithm with data pretreatment.
19. Apparatus as defined in claim 11 , wherein processing the Raman spectral data includes executing a regression algorithm comprising a partial least squares, multiple linear regression, or principal component regression, with or without data pretreatment.
20. Apparatus as defined in claim 19 , wherein the data pretreatment is selected from multiplicative and extended multiplicative scatter correction, standard normal variate, various derivatives, and normalizations.