Agilent Solutions for the Petrochemical and Oleochemical Industries

e-seminar Series - Agilent Solutions for the Petrochemical and Oleochemical Industries Date: 12 June 2009 Time : 10am Singapore Time For audio, please dial one of the phone numbers listed: International Telephone Number for Singapore: +65-6622 1044 or Toll Free 8008 523 396 International Telephone Number for Hong Kong: +852 3006 8101 Toll Free International Numbers: Australia 1800 999 130 New Zealand 0800 450 755 China 800 876 5011 Malaysia 1800 807 180 Indonesia 0018 038 526 350 Taiwan 0080 185 5735 India* 0008 008 521 136 South Korea 0079 885 214 717 India (Backup)* 0008 001 006 542 Philippines 1800 185 50065 Thailand* 0018 008 526 361 Vietnam 120 650 065 * Phone line needs to be IDD enabled. Not accessible from Mobile Phones

Measurement of low levels of FAME in Aviation Fuel by GC-MS Development of IP Method PM-DY/09 Tom Lynch and Alex Ttofi BP Technology Centre Pangbourne Presenter: Jim McCurry Agilent Technologies

Trace FAME in Jet Increasing quantities of biodiesel and jet are being co-transported in multi-product pipelines (MPP). In MPP transportation trace amounts of FAME can be found in jet parcels following biodiesel parcels due to FAME trail back. Following pipeline trials to establish the amount and profile of FAME trail back into jet and JIG PQ committee work on the effect of various FAMEs (up to 400 ppm) on the specification properties of jet fuel the main engine OEMs gave their verbal and written approval of up to 5ppm FAMEs in jet fuel. The DFG now consider it is necessary to formalize this acceptance of 5ppm FAMEs in jet fuel in the Def Stan 91-91, Issue 6 specification and have issued a draft specification amendment for consideration by the AFC, ExCo and OEMs (engine and airframe).

Trace FAME in Jet Following from this proposed amendment there is an urgent need to develop a referee test method with an acceptable precision at the 5 ppm FAME level. Trace FAME in Jet The method will need to be robust Applicable to all FAME types / unknown blends Suffer no interference from common diesel additives Be subjected to a ruggedness trial

Methods Considered Primary/Short Term methods requirements Sensitivity ideally less than 1ppm per FAME species Selectivity Secondary method requirements Confirmation measured species are FAME Fast, equipment readily available, minimal sample preparation 2D GC Comprehensive not widely available, too specialised, data analysis can be complex Multi heart-cut potential but worries about FID sensitivity FTIR & NMR Good selectivity but poor sensitivity would require a pre-concentration step. Different FAMES have different absorption coefficients in FTIR. HPLC (modified IP 436) Looks to have potential but modern stationary phases do not give good results potential for phase development Potentially good sensitivity and selectivity but as yet unproven. GC-MS Even most basic quadrupole systems offer high sensitivity and selectivity Mass spectrum can confirm presence of FAME

Fatty acids composition of bio diesel feedstocks (vegetable derived) Acid (% wt) Castor Coconut Corn Linseed Olive Palm Peanut Rape Soya Sunflower Jatropha Saturated acids Caproic C6:0 0.8 Caprylic C8:0 5.4 Capric C10:0 8.4 Lauric C12:0 45.4 Myristic C14:0 18.0 1.0 1..5 0.06 Palmitic C16:0 10.5 7..5 9.0 48 5.0 4.2 6.5 3.5 14.6 Stearic C18:0 0.3 2..3 3.5 3.5 2..3 4.0 5.0 1.5 4.2 2.9 7.15 Arachidic C20:0 0.4 0.5 0.2 4.0 0.7 0.6 0.2 Behenic C22:0 0..5 0.1 Lignoceric C24:0 0.2 3.0 2.4 0.4 Mono-unsaturated acids Palmitoleic C16:1 0.4 0.85 Oleic C18:1 8.0 7..5 46.3 5.0 82.5 38 60.0 60.0 33.6 33.4 46.3 Ricinoleic C18:1 87.8 0.2 Erucic C22:1 Bi-unsaturated acids Linoleic C18:2 3.6 42.0 61.5 6.0 9 21.0 21.0 52.6 57.5 30.8 Poly-unsaturated acids Linolenic C18:3 25.0 10 2.3

Initial method for first pipeline trial Developed using an RME (rape seed) fuel as supplied. Principle:- Use a polar GC column to get maximum retention of polar FAME and Selected Ion Monitoring (SIM) for selectivity and sensitivity In summary the specifics of the method are Column HP Innowax (Agilent 19091N-213) 30m x 0.32um id x 0.5um film thickness Initial oven temp 170ºC hold 5mins,ramp 6ºC until 248ºC & hold for 5 mins. Total of run 23 mins Inlet at 325ºC, constant flow of 1.2ml/min Splitless injection of 0.2µl neat sample SIM and Scan data collected

GC-MS method for RME in Jet Fuel Abundance 750000 Main problem is resolving the RME FAME components from the hydrocarbons and aromatics in the kerosene. Cannot do this with conventional boiling point GC column. Use of a polar Innowax column increases the retention of the RME FAME species relative to the kerosene components However, even with the conventional scanning mode GC- MS did not give sufficient sensitivity (top trace) GC-MS used with multiple single ion monitoring (SIM) to achieve a detection limit of circa 0.1ppm (lower trace) Time--> 700000 650000 600000 550000 500000 450000 400000 350000 300000 250000 200000 150000 100000 50000 Abundance Time--> 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 5ppm, 3ppm, 1ppm, 0.5ppm and 0.1ppm FAME in Kerosene using conventional GC-MS full scan mode 15.50 16.00 16.50 17.00 17.50 18.00 18.50 5ppm, 3ppm, 1ppm, 0.5ppm and 0.1ppm FAME in Kerosene using GC-MS multi SIM mode 15.50 16.00 16.50 17.00 17.50 18.00 18.50

GC-MS multi-sim for RME in Jet Fuel 60000 Calibration 0.1 to 1ppm FAME in kero by GC-MS 50000 y = 42548x + 5160.9 R 2 = 1 40000 peak area 30000 area Linear (area) 20000 10000 0 0 0.2 0.4 0.6 0.8 1 ppm FAME

Pipeline Study Fame content of Gas/Kero line during Jet import. FAME content (ppm) 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 0.00 50.00 100.00 150.00 Time (mins) Fame Content Limit of detection

Limitations of initial method Set up for one trial with known RME Used only one component (C18:1) No scope to include Other FAME components as required by customers Required the fuel and FAME blanks to reach maximum sensitivity Require greater resolution of FAME species from aromatics This lead to development of our current method

Fatty acids composition of bio diesel feedstocks (vegetable derived) Acid (% wt) Castor Coconut Corn Linseed Olive Palm Peanut Rape Soya Sunflower Jatropha Saturated acids Caproic C6:0 0.8 Caprylic C8:0 5.4 Capric C10:0 8.4 Lauric C12:0 45.4 Myristic C14:0 18.0 1.0 1..5 0.06 Palmitic C16:0 10.5 7..5 9.0 48 5.0 4.2 6.5 3.5 14.6 Stearic C18:0 0.3 2..3 3.5 3.5 2..3 4.0 5.0 1.5 4.2 2.9 7.15 Arachidic C20:0 0.4 0.5 0.2 4.0 0.7 0.6 0.2 Behenic C22:0 0..5 0.1 Lignoceric C24:0 0.2 3.0 2.4 0.4 Mono-unsaturated acids Palmitoleic C16:1 0.4 0.85 Oleic C18:1 8.0 7..5 46.3 5.0 82.5 38 60.0 60.0 33.6 33.4 46.3 Ricinoleic C18:1 87.8 0.2 Erucic C22:1 Bi-unsaturated acids Linoleic C18:2 3.6 42.0 61.5 6.0 9 21.0 21.0 52.6 57.5 30.8 Poly-unsaturated acids Linolenic C18:3 25.0 10 2.3

Major components of bio diesel feeds Fatty Acids (% wt) Castor Corn Linseed Olive Palm Peanut Rape Soya Sunflower Jatropha Saturated Palmitic C16:0 0.9 13 6 14 48 14 5 6.5 4 14.6 Stearic C18:0 0.3 4 4 2 4 4 2 4.2 3 7.2 Mono-unsaturated Oleic C18:1 3.3 29 22 66 38 48 61 33.6 34 46.3 Ricinoleic C18:1 89 Bi-unsaturated Linoleic C18:2 3.7 54 16 16 9 30 19 52.6 58 30.8 Poly-unsaturated Linolenic C18:3 52 10 2.3 Total C18 96 87 94 84 51 82 92 93 95 84 Total C18 + C16:0 97 100 100 98 99 96 97 99 99 99 Data taken from Bailey s Industrial Oil and Fat Products Vol 1, 4 th Ed, Wiley ISBN 0-471-839574 Total C18 recovers between 80 and 95% of most feeds listed except Palm Total C18 +C16:0 recovers >95% of all feeds listed.

Improved method details Instrument Agilent 5973 MS & 6890 GC (diffusion pump) Column- HP INNOWAX (Agilent 19091N-205) 50 m x 200 um id x 0.4 um film. Constant flow 0.6ml/min Splitless Injector temp - 325ºC, MS source: 230ºC & MS Quad 150ºC Neat sample Injection 1 µl, Oven: Initial 150ºC hold 5 mins, Ramp 12ºC / min until 210ºC hold 17min Ramp 3ºC / min until 252ºC hold 2 min Total run time 43 mins Data collection delay 16 mins MS SIM & SCAN data collected simultaneously, SIM ions - see table on next slide Sum of all 6 SIM ion peaks measured for results

SIM Ions for FAMEs in Jet Fuel FAME Species C16:0 C17:0 C18:0 C18:1 C18:2 C18:3 Approximate RT (min.) 24.3 29.5 34.2 35.0 35.5 (2 summed isomer peaks) 37.2 39.8 SIM Ions to be used for quantification (m/z) 227, 239, 270, 271 241, 253, 284 255, 267, 298 264, 265, 296 262, 263, 264, 294, 295 236, 263, 292, 293

Standards Standards of C16:0, C17:0 and C18 Fames 1.000 mg/kg FAME stock standard Individual FAME species dissolved in 99% n-dodecane Eight GC/MS calibration standards prepared in n-dodecane from stock standard 0.5, 1, 2, 3, 5, 10, 20 and 50 mg/kg + n-dodecane blank All standards stored at 0 o C when not in use

Mixed FAME STD in Jet: 200ppm Total ion count Total 6 SIM C14:0 C15:0 C16:0 C18:1 C18:2 C18:3 C17:0 C18:0

Calibration overlay for C18 components Abundance 10000 9500 9000 8500 8000 C18:1 C18:2 5ppm C18:3 7500 7000 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 C18:0 2ppm 1ppm 0.5ppm 0.1ppm Time--> 32.50 33.00 33.50 34.00 34.50 35.00 35.50 36.00 36.50 37.00 37.50 38.00 38.50 39.00 39.50

Detection limit C18:2 Abundance 11000 10000 9000 5ppm 8000 7000 6000 5000 4000 3000 2000 1000 2ppm 1ppm 0.5ppm 0.1ppm Time--> 0 35.10 35.20 35.30 35.40 35.50 35.60 35.70 35.80 35.90 36.00 36.10 36.20 36.30 36.40 36.50 Reference fuel

Calibrations: Total C18 C18 methyl esters peak areas summed C18 methyl esters are C18:0, C18:1, C18:2 & C18:3 3000000 2500000 Total Area C18 Calibration R 2 = 0.9995 2000000 Total Area 1500000 1000000 500000 0 0 5 10 15 20 25 30 35 40 45 Conc FAME ppm

FAME STD - C18 individual components FAME in Avtur Low Range Calibration Peak Area 3000000 2500000 2000000 1500000 1000000 500000 C16:0 C17:0 C18:0 C18:1 C18:2 C18:3 Linear (C17:0) Linear (C16:0) Linear (C18:2) Linear (C18:1) Linear (C18:3) Linear (C18:0) C17:0 y = 338280x R 2 = 0.9998 C16:0 y = 549880x R 2 = 0.9998 C18:0 y = 338472x R 2 = 0.9998 C18:1 y = 189088x R 2 = 0.9996 C18:2 y = 136534x R 2 = 0.9998 0 C18:3 y = 53002x R 2 = 0.9994 0 1 2 3 4 5 6 mg.kg FAME

Limit of detection C16:0 Abundance 38000 36000 34000 32000 30000 28000 26000 24000 22000 20000 18000 16000 14000 12000 10000 Reference fuel 5ppm 2ppm 1ppm 0.5ppm 0.1ppm 8000 6000 4000 2000 22.90 23.00 23.10 23.20 23.30 23.40 23.50 23.60 23.70 Time-->

PME calibration Calibration also developed using Palm oil methyl ester based bio fuel (PME) in Jet Fuel PME in Jet Fuel calibration based on Total C16:0 C18 by GC-MS 3000000 2500000 R 2 = 0.9997 Total C16:0 C18 peak area 2000000 1500000 1000000 500000 0 0 5 10 15 20 25 30 35 40 45 50 Conc PME (ppm)

RME calibration 1300000 RME calibration Total C16:0 + C18 Peak Area 1100000 900000 700000 500000 300000 100000 R 2 = 0.9999-100000 0 2 4 6 8 10 12 14 16 18 20 Conc RME (ppm)

Repeatability Total Individual FAME FAME C18:3 C18:2 C18:1 C18:0 5ppm standard Standard deviation 0.09 0.08 0.08 0.15 0.07 95% limits 0.18 0.16 0.16 0.30 0.14 Mean 4.69 4.47 4.73 4.69 4.95 Variance 0.01 0.01 0.01 0.02 0.01 Number 10 10 10 10 10 50ppm standard Standard deviation 0.88 0.65 0.90 1.05 1.41 95% limit 1.80 1.30 1.80 2.10 2.82 Variance 0.82 0.43 0.83 1.13 2.21 Mean 48.95 49.23 48.67 49.87 47.71 number 10 10 10 10 10

Confirmation of FAME present The mass spectrometer adds the possibility to confirm the peaks as FAMES From TIC we can get a spectrum and library match down to circa 20ppm. From pattern of 6 SIM ions we can also get some confirmation of FAME Abundance Abundance Abundance 24000 22000 20000 18000 16000 14000 12000 10000 8000 6000 4000 2000 0 m/z--> Abundance 18000 17000 16000 15000 14000 13000 12000 11000 10000 Scan 697 (24.220 min): K03D006.D\datasim.ms (-667) (-) 227.0 270.0 22000 SIM 16:0 20000 18000 16000 14000 239.0 12000 10000 8000 6000 4000 2000 0 220 225 230 235 240 245 250 255 260 265 270 275 m/z--> Abundance Scan 1961 (35.492 min): K03D006.D\datasim.ms (-1988) (-) 9500 264.0 9000 8500 SIM 18:1 8000 7500 7000 6500 6000 5500 5000 Scan 1263 (29.465 min): K03D006.D\datasim.ms (-1233) (-) 241.0 253.0 SIM 17:0 284.0 26000 24000 22000 20000 18000 16000 14000 12000 10000 8000 6000 4000 2000 Scan 1812 (34.186 min): K03D006.D\datasim.ms (-1779) (-) 255.0 298.0 SIM 18:0 0 235 240 245 250 255 260 265 270 275 280 285 290 245 250 255 260 265 270 275 280 285 290 295 300 305 Abundance m/z--> Scan 2134 (37.194 min): K03D006.D\datasim.ms (-2115) (-) 294.0 263.0 SIM 18:2 4000 3500 3000 2500 267.0 Scan 2363 (39.762 min): K03D006.D\datasim.ms (-2347) (-) 236.0 292.0 SIM 18:3 9000 8000 4500 4000 2000 263.0 7000 6000 3500 3000 1500 5000 4000 2500 2000 1000 3000 2000 296.0 1500 1000 500 1000 0 255 260 265 270 275 280 285 290 295 300 m/z--> m/z--> 500 0 255 260 265 270 275 280 285 290 295 300 m/z--> 0 230 235 240 245 250 255 260 265 270 275 280 285 290 295 300

5 ppm FAMEs in MEROX Jet Fuel Abundance 12500 TIC: K03D023.D\datasim.ms TIC: K03D010.D\datasim.ms (*) 12000 11500 11000 10500 10000 9500 9000 8500 8000 7500 7000 Blue trace = 5ppm Rape Seed Methyl Esters in MEROX Black trace = MEROX 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 Time--> Abundance 12500 TIC: K03D023.D\datasim.ms TIC: K03D018.D\datasim.ms (*) 12000 11500 11000 10500 10000 9500 9000 8500 8000 Blue trace = 5ppm Mixed FAME in Merox Black trace = Merox 7500 7000 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 Time-->

5 ppm FAMEs in HDT Jet Fuel Abundance 13000 TIC: K03D020.D\datasim.ms TIC: K03D007.D\datasim.ms (*) 12500 12000 11500 11000 10500 10000 9500 9000 8500 8000 7500 Blue trace = 5ppm Rape Seed Methyl Esters in HDT Black trace = HDT 7000 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 Abundance Time--> 0 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 14000 13000 TIC: K03D020.D\datasim.ms TIC: K03D013.D\datasim.ms (*) 12000 11000 10000 9000 Blue trace = 5ppm Mixed FAME in HDT Black trace = HDT 8000 7000 6000 5000 4000 3000 2000 1000 Time--> 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00

Benefits of method Meets 5 ppm detection limit for most first generation bio-diesel feeds. Uses readily available off the shelf equipment No sample preparation samples injected neat Under 1 hour per sample Can confirm peaks are FAMES Good linearity of FAME over wide concentration range Can match FAME contamination to possible source Equipment also very useful for other bio-fuel related analysis

Limitations and Options for Improvement Limitations Detection Limits can be influenced by base jet fuel Ideally you need the uncontaminated jet fuel to provide a zero baseline Higher detection limits for lower Carbon number FAME s (C12 for coconut not possible) Mineral diesel fuel contamination interferes. Development Options Better stationary phase with better selectivity for FAME over aromatics Heart cut FAMES using microfluidics to achieve better resolution and sensitivity Chemical Ionisation (CI) may offer greater selectivity and sensitivity, will look at in near future.

??????? Thank you for listening Any further questions or comments

Next e-seminar Series Tuesday, 16 th June, 2009 10:00 AM (Singapore Time) Simplifying sample preparation and speeding up the analysis for the characterisation of simulated distillation samples Speaker: Roger Firor, Agilent Technologies