NMR-Based Test Methods - How to Change the Paradigm of f NMR - - PowerPoint PPT Presentation
The Practical Non-Existance of f Official NMR-Based Test Methods - How to Change the Paradigm of f NMR Exclusion The Perspective of a Small NMR Service Company John Edwards, Process NMR Associates Validation Workshop PANIC NMR Conference
Validation Workshop – PANIC NMR Conference – February 19, 2016 – Houston Texas
ASTM NMR Test Methods E2977: Standard Practice for Measuring and Reporting Performance of Fourier Transform Nuclear
Magnetic Resonance (FT-NMR) Spectrometers for Liquid Samples
D7171-05: Standard Test Method for Hydrogen Content of Middle Distillate Petroleum Products by Low-
Resolution Pulsed Nuclear Magnetic Resonance Spectroscopy
D4273-05: Standard Test Methods for Polyurethane Raw Materials Determination of Primary Hydroxyl
Content of Polyether Polyols
D4875-05: Standard Test Methods of Polyurethane Raw Materials: Determination of the Polymerized
Ethylene Oxide Content of Polyether Polyols
D5017-96: Standard Test Method for Determination of Linear Low Density Polyethylene (LLDPE)
Composition by Carbon-13 Nuclear Magnetic Resonance
D5292-99: Standard Test Method for Aromatic Carbon Contents of Hydrocarbon Oils by High Resolution
Nuclear Magnetic Resonance Spectroscopy
F2259-03: Standard Test Method for Determining the Chemical Composition and Sequence in Alginate by
Proton Nuclear Magnetic Resonance (1H NMR) Spectroscopy
F2260-03: Standard Test Method for Determining Degree of Deacetylation in Chitosan Salts by Proton
Nuclear Magnetic Resonance (1H NMR) Spectroscopy
USP <2251> Adulteration of Dietary Supplements with Drugs and Drug Analogs.
Poloxamer Testing
USP <761>: General Chapter – Nuclear Magnetic Resonance Spectroscopy
Krill Oil Monograph – 1H and 31P NMR, Beta Glucan Monograph – 1H NMR, Heparin Sodium Monograph – 1H NMR
European Pharmacopoeia Salmon Oil, Farmed 01/2005:1910 13C NMR for distribution of fatty acids.
ISO 10565:1998 Oil and Moisture in Seeds by Low Field NMR Spectroscopy 8292 Animal and vegetable fats and oils -- Determination of solid fat content -- Pulsed nuclear magnetic resonance method 21461:2006 Rubber - Determination of the aromaticity of oil in vulcanized rubber compounds 21561:2005 Styrene-butadiene rubber (SBR) - Determination of the microstructure of solution- polymerized SBR 11543:2000 Modified starch -- Determination of hydroxypropyl content -- Method using proton nuclear magnetic resonance (NMR) spectrometry 10632:2000 Oilseed residues -- Simultaneous determination of oil and water contents -- Method using pulsed nuclear magnetic resonance spectroscopy AOCS Cd 16b-93 : (revised in 2000): Direct Method - Determination of Solid Fat Content -- Pulsed Nuclear Magnetic Resonance Method Cd 16-81 : (revised in 2000): Indirect Method - Determination of Solid Fat Content -- Pulsed Nuclear Magnetic Resonance Method Ak5-01 (01) : Oil Content of Oilseed Residues by NMR Ak3-94 (00) : Oil Content in Rapeseed and other Oilseeds by NMR Ak4-95 (99):Simultaneous Determination of Oil and Moisture Contents of Oilseeds using Pulsed Nuclear Magnetic Resonance Spectroscopy American Herbal Pharmacopoeia Aloe Vera - 1H NMR Method for Quality Inspection and Identification Canadian Crude Quality Technical Association Maxxam V1.04 11/05 : Determination of Olefin Content of Crude Oils, Condensates and Diluents by 1H NMR
Guide to NMR Method Development and Validation – Part I: Identification and Quantification
1H qNMR is very well
referenced in the academic literature but is still not an official standard method validated by a standards organization
Component Dry Matter Composition Wt% Product Composition Wt% Product Composition (mg/L) Acemannan Wt% 13.96 0.158 1549 Glucose Wt% 16.00 0.181 1776 Malic Acid Wt% 27.89 0.315 3096 Lactic Acid 0.00 0.000 Citric Acid 22.07 0.249 2449 Pyruvic Acid 0.16 0.002 18 Sorbate 0.00 0.000 Benzoate 0.00 0.000 Acetoin 0.00 0.000 2,3-butandiol 0.00 0.000 Whole Leaf Marker (WLM) Not Observed Not Observed Trace Observed Isocitrate (WLM-1)
0.00
0.000 Isocitrate Lactone (WLM-2)
0.00
0.000 Acetic Acid 0.00 0.002 17 Succinic Acid 0.00 0.000 Formic Acid 0.00 0.000 Fumaric Acid 0.00 0.000 Ethanol 0.00 0.000 Glycine 0.00 0.000 Propylene Glycol 0.00 0.000 Sucrose 0.00 0.000 Dry Matter Weight (%) Organic Aloe Vera Juice Hand Fillet with Pulp Batch OAVWP0116 1.13
Alcohol and Tobacco Tax and Trade Bureau (TTB)
300 MHz NMR – PDE5i Standards
Component Wt% = 100 x (Component Molar Integral / Std Molar Integral) x (Wt Std / Wt Sample) x (Component Mw / Std Mw)
Component Wt% = 100 x (Component Molar Integral / Std Molar Integral) x (Wt Std / Wt Sample) x (Component Mw / Std Mw)
Experimental Varian Mercury - 300 MHz NMR Supercon Spectrometer 4 pulse on pure sample in 5 mm tube, Run Unlocked In MNova 8.1.2 SPC Files Imported, Stacked, Binned at 3 Hz interval, Area Normalized to 100, Saved as Transposed Ascii Matrix PLS Regression Performed Thermo Grams IQ and Eigenvector Solo
DHA EPA-a
Ole Olefin ins =C =C-CH CH2-C= C= In chain CH2 EtO tOOC-CH CH2-R
DHA EPA
CH3 CH3-CH CH2-O-OC OC-CH CH2-R CH3 Ethyl CH3 R-CH2-C= CH3 w6/ w9 w3
EPA-b
www.process-nmr.com www.nmrblog.com Tel: 203-744-5905
www.process-nmr.com www.nmrblog.com Tel: 203-744-5905
0.5 Hz binning interval 3 Hz binning interval
www.process-nmr.com www.nmrblog.com Tel: 203-744-5905
www.process-nmr.com www.nmrblog.com Tel: 203-744-5905
Spinsolve 1H NMR data at 42.5 MHz of a series of fish oils and esterified omega-3 fatty acids
10 20 30 40 1 2 3 4 5 6 Hotelling T^2 (99.91%) Q Residuals (0.09%) 0.2 0.4 0.6 0.8
1 2 3 Leverage Y Stdnt Residual 1 20 40 60 80
20 40 60 80 Y Measured 1 Y CV Predicted 1
50
5 10 15 Scores on LV 1 (95.78%) Scores on LV 2 (2.48%)
Model Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 27-Jul-2014 15:03:05.984 Author: John@JCEPNA X-block: Final Fish Oil Spectra - 43MHz.csv 55 by 830 (John@JCEPNA@20140727T140753.60670200 m:20140727140753.619) Included: [ 1-55 ] [ 1-830 ] Preprocessing: Mean Center Y-block: EPA-GC Data.xlsx 55 by 1 (John@JCEPNA@20140727T141758.52236711 m:20140727141758.529) Included: [ 1-55 ] [ 1 ] Preprocessing: Mean Center
Cross validation: venetian blinds w/ 7 splits RMSEC: 1.40566 RMSECV: 2.20172 Bias: 0 CV Bias: -0.00054578 R^2 Cal: 0.995028 R^2 CV: 0.987838 SSQ Table Percent Variance Captured by Regression Model
Comp This Total This Total
1 95.78 95.78 84.51 84.51 2 2.48 98.26 3.47 87.98 3 0.29 98.55 9.21 97.19 4 0.76 99.31 0.59 97.78 5 0.43 99.73 0.60 98.38 6 0.07 99.80 0.77 99.15 7 0.06 99.86 0.17 99.33 8 0.04 99.91 0.18 99.50
Omega-3 EPA Content (wt%) in Fish Oil and Pharmaceutical Grade Supplements Correlation between 40 second 1H NMR at 42.5 MHz and primary analysis method – 50 minute GC analysis
10 20 30 40 1 2 3 4 Hotelling T^2 (99.91%) Q Residuals (0.09%) 0.2 0.4 0.6 0.8
2 4 Leverage Y Stdnt Residual 1 10 20 30 40 50 60
20 40 60 80 Y Measured 1 Y CV Predicted 1
50
10 20 Scores on LV 1 (95.78%) Scores on LV 2 (2.23%) R^2 = 0.989 8 Latent Variables RMSEC = 0.75363 RMSECV = 1.2463 Calibration Bias = 0 CV Bias = 0.098796
Model Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 27-Jul-2014 14:50:021.58 Author: John@JCEPNA X-block: Final Fish Oil Spectra - 43MHz.csv 55 by 830 (John@JCEPNA@20140727T140753.60670200 m:20140727140753.619) Included: [ 1-55 ] [ 1-830 ] Preprocessing: Mean Center Y-block: DHA-GC Data.xlsx 55 by 1 (John@JCEPNA@20140727T140259.39467736 m:20140727140259.401) Included: [ 1-55 ] [ 1 ] Preprocessing: Mean Center
Cross validation: venetian blinds w/ 7 splits RMSEC: 0.753632 RMSECV: 1.2463 Bias: 0 CV Bias: 0.0987959 R^2 Cal: 0.99605 R^2 CV: 0.989345 SSQ Table Percent Variance Captured by Regression Model
Comp This Total This Total
1 95.78 95.78 73.08 73.08 2 2.23 98.01 6.53 79.61 3 0.55 98.56 15.91 95.53 4 0.50 99.06 1.63 97.16 5 0.67 99.73 0.78 97.93 6 0.07 99.80 1.42 99.35 7 0.07 99.87 0.15 99.50 8 0.04 99.91 0.10 99.61
Omega-3 DHA Content (wt%) in Fish Oil and Pharmaceutical Grade Supplements Correlation between 40 second 1H NMR at 42.5 MHz and primary analysis method – 50 minute GC analysis
Picospin-80 NMR Data – 82.3 MHz
10 20 30 40 1 2 3 4 5 6 Hotelling T^2 (99.91%) Q Residuals (0.09%) 0.2 0.4 0.6 0.8
1 2 3 Leverage Y Stdnt Residual 1 20 40 60 80
20 40 60 80 Y Measured 1 Y CV Predicted 1
50
5 10 15 Scores on LV 1 (95.78%) Scores on LV 2 (2.48%)
Model Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 27-Jul-2014 15:03:05.984 Author: John@JCEPNA X-block: Final Fish Oil Spectra - 43MHz.csv 55 by 830 (John@JCEPNA@20140727T140753.60670200 m:20140727140753.619) Included: [ 1-55 ] [ 1-830 ] Preprocessing: Mean Center Y-block: EPA-GC Data.xlsx 55 by 1 (John@JCEPNA@20140727T141758.52236711 m:20140727141758.529) Included: [ 1-55 ] [ 1 ] Preprocessing: Mean Center
Cross validation: venetian blinds w/ 7 splits RMSEC: 1.40566 RMSECV: 2.20172 Bias: 0 CV Bias: -0.00054578 R^2 Cal: 0.995028 R^2 CV: 0.987838 SSQ Table Percent Variance Captured by Regression Model
Comp This Total This Total
1 95.78 95.78 84.51 84.51 2 2.48 98.26 3.47 87.98 3 0.29 98.55 9.21 97.19 4 0.76 99.31 0.59 97.78 5 0.43 99.73 0.60 98.38 6 0.07 99.80 0.77 99.15 7 0.06 99.86 0.17 99.33 8 0.04 99.91 0.18 99.50
Omega-3 EPA Content (wt%) in Fish Oil and Pharmaceutical Grade Supplements Correlation between 40 second 1H NMR at 42.5 MHz and primary analysis method – 50 minute GC analysis
Model Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 15-Nov-2014 23:10:039.19 Author: John@JCEPNA X-block: Fish Oils - 40 Sample - X Matrix.xlsx 37 by 57 (John@JCEPNA@20141115T230315.72489196 m:20141115231022.946) Included: [ 1-11 13-27 29-32 34-40 ] [ 1-2 71-125 ] Preprocessing: Mean Center Y-block: DHA.xlsx 37 by 1 (John@JCEPNA@20141115T230330.44825991 m:20141115231022.953) Included: [ 1-11 13-27 29-32 34-40 ] [ 1 ] Preprocessing: Mean Center
Cross validation: venetian blinds w/ 6 splits RMSEC: 1.09711 RMSECV: 1.20325 Bias: 3.55271e-015 CV Bias: 0.0624194 R^2 Cal: 0.992002 R^2 CV: 0.99041 SSQ Table Percent Variance Captured by Regression Model
Comp This Total This Total
1 79.58 79.58 77.43 77.43 2 11.40 90.97 20.16 97.59 3 7.91 98.88 1.61 99.20
5 10 15 20 25 30 0.5 1 1.5 2 2.5 3 Hotelling T^2 (98.88%) Q Residuals (1.12%) 0.2 0.4 0.6 0.8
1 2 3 4 Leverage Y Stdnt Residual 1 10 20 30 40 50 60
20 40 60 Y Measured 1 Y CV Predicted 1
10 20
5 10 15 20 Scores on LV 1 (79.58%) Scores on LV 2 (11.40%)
DHA model – 0.02 ppm bin Picospin 80.3 MHz 1H NMR Data
www.process-nmr.com www.nmrblog.com Tel: 203-744-5905
standards organization. i) Internal Standard Approach ii) Electronic Standard Approach iii) Spectral Deconvolution Approach
NMR systems
AOAC, AOCS, ISO, ASTM, US Pharmacopoia, European Pharmacopoeia ? Personal Objective: