石油产品模拟蒸馏的冷柱头技术评价
互联网
981
石油产品模拟蒸馏的冷柱头技术评价
Evaluation of a New Cold On-Column Capillary Inlet for Simulated Distillation of Petroleum Products
Introduction
In the petroleum refining process the boiling point distribution range of the feedstock streams is very important and can be used for product specification testing. The analysis is determined by gas chromatography (GC) as specified in ASTM D2887. The method is applicable to samples with an initial boiling point (IBP) of greater than 55°C and a final boiling point (FBP) of less than 538°C. Since a conventional splitless injection has shown discrimination over a wide boiling point range, the cold on-column injection technique is preferred. The separation, sensitivity, and precision criteria for Method D2887 were evaluated using a CE Instruments GC cold on-column injector with a Flame Ionization Detector (FID) and Model AS 800 auto-sampler. All GC parameters were optimized as specified in the method. The CE Instrument automated cold on-column injector and FID provided the instrument performance required meeting the criteria for Method D2887.
Experimental
The GC was set up for on-column injection with an auto sampler. The samples and standards were diluted in carbon disulfide (CS2). Air was used as the coolant for the cold on-column injector, Figure 1. A guard column was connected to the analytical column with a press fit connector. The instrument parameters are listed below: CE Instruments GC
Instrument Parameters:
Cold On-column inlet at ambient
OC Coolant valve: On for 60 seconds with air
Oven: 55°C, 20°C/min; 360°C, 10 minutes
FID: 360°C
H2: 30 mL/min
Air: 300 mL/min
Make Up: 10 mL/min nitrogen
Range: 1
Column:Restek Rtx 2887 0.53 mm × 10 m; 2.65 micron film,
Column Flow: 5 mL/min nitrogen
Guard Column: Restek Phenylmethyl deactivated 0.53 mm×5m
Injection Volume: 1 µL
AS800 Autosampler
The Injection:
The method specifies that when using an on-column injector, the entire capillary column must be temperature programmed including the point of sample introduction. In an on-column injection, the needle must enter the column inside the oven. The injector used has a butterfly valve, which is opened as the turret of the autosampler rotates over the injector, . The sample is injected through a septum into the front section of the column in the oven with a syringe with an 80 mm long needle. With solvents that have lower boiling points than the analytes, the oven is set below the boiling point of the solvent at the time of injection and no secondary cooling of the inlet is required. The sample is introduced as a liquid plug, and as the solvent evaporates the analytes are focused. Figure 2 is a representative chromatogram of an injection below the boiling point of the solvent of a diesel range organics (DRO) standard in methylene chloride.
Figure 2. Cold On-column Injection of DRO Standard in Methylene Chloride at 50 ppm each with an Oven Programof 35°C, 2 min; 10°C/min; 290°C, 10 min
When a heavier solvent like CS2 is used, the more volatile analytes like C6H14 through C11H22 show the adverse effects of solvent flooding. To eliminate peak distortion, the oven is set at 10-15°C above the boiling point of the solvent and the secondary coolant valve for the injector is turned on during the injection to prevent flashing back of the sample inside the inlet.
Figure 3. Adverse effects of Solvent Flooding with CS2 at Different Initial Oven Temperatures
Although CS2 does not respond on the FID, the adverse effects of the solvent may be seen in starting at lower initial oven temperatures as shown in Figure 3.The literature has shown that "the spreading of the liquid sample along the inlet part of the column produces broadening, distortion, and peak splitting. This flooding effect of the solvent may be eliminated by the incorporation of a guard column and operation of the initial oven temperature in a transition region at 10-15°C above the boiling point of the solvent. By injecting the sample above the boiling point of CS2 , the solvent plug is reduced to an insignificant volume. The injection of the D2887 standard into a 5 meter guard column connected to the analytical column at an initial temperature of 55°C eliminated the adverse flooding effects of the CS2 , Figure 4.
Figure 4. D2887 Standard at 1% in with an Oven Program of 55°C, 0 min,20°C/min; 360°C, 10 min (See Table 1 for concentrations)
The Separation
In order to simulate the true physical distillation in Method D2892, the chromatography was deoptimized. Nitrogen was selected for the carrier gas, because the resolution of C16H34 and C18H38 was too good with helium. The column flow was set to 5 mL/min to retard the elution of the IBP for the D2887 Reference Gas Oil to at least 1.0 minute, Figure 5. An oven program was established to elute all of the analytes through a boiling point of 538°C before the end of the ramp, Figure 4. The software automatically performed the baseline compensation analysis on all of the files using the first run of the solvent blank. Figure 6 shows the calibration curve generated from the data in Figure 4. Table 1 shows the uniformity of the hydrocarbons over the boiling point range of the D2887 calibration standard.
Table 1. RF
Compound Conc (%) Area RF
nC6 0.06 3964599 66076650
nC7 0.06 3379799 56329983
nC8 0.08 4450280 55628500
nC9 0.08 4496167 56202088
nC10 0.12 6798554 56654617
nC1 0.12 6751962 56266350
nC12 0.12 6779067 56492225
nC14 0.12 6862533 57187775
nC16 0.10 5615236 56152360
nC18 0.05 2820382 56407640
nC20 0.02 1129601 56480050
nC24 0.02 1108412 55420600
nC28 0.01 550048 55004800
nC32 0.01 544953 54495300
nC36 0.01 538015 53801500
nC40 0.01 528177 52817700
nC44 0.01 524163 52416300
Mean 56107908
SD 2911128
% RSD 5.18
Results
Eight replicates of the D2887 standard were analyzed to check the precision of the instrument. The retention time had a standard deviation of 0 for most of the compounds with a value of +/0.005 minutes for C44H99 . The relative percent standard deviation for response in area counts was less than 1.88%. The data is shown for the retention time and area counts as listed in Tables 2 and 3. The FID was set at a less sensitive range value from 0 to 1. No skewing of the peaks was seen. A customer crude oil sample was run on the system and shown in an overlay view with the D2887 Reference Gas Oil and calibration standard in .Figure 7
Table 2. Retention Time Precisioundefined
Compound Mean SD % RSD
nC6 0.868 0.0028 0.326
nC7 1.233 0.0000 0.000
nC8 1.751 0.0028 0.162
nC9 2.408 0.0000 0.000
nC1 0 3.129 0.0043 0.137
nC11 3.858 0.0000 0.000
nC12 4.567 0.0000 0.000
nC14 5.877 0.0037 0.063
nC16 7.057 0.0028 0.040
nC18 8.108 0.0000 0.000
nC20 9.067 0.0000 0.000
nC24 10.750 0.0000 0.000
nC28 12.183 0.0000 0.000
nC32 13.433 0.0000 0.000
nC36 14.548 0.0037 0.025
nC40 15.598 0.0037 0.024
nC44 16.913 0.0048 0.028
Conclusion
The CE Instruments automated cold on-column injection technique met the criteria for ASTM Method D2887 and showed uniform response over the boiling point distribution range of the method. The separation and reproducibility were well within the guidelines of the method. The sensitivity of the FID was more than required for the method and had to be run at a less sensitive amplification. The unique design of the autosampler and inlet provided the critical alignment of the needle on injection, so that the same hole in the septum is penetrated each time. The septum did not leak during the entire time of the study and was never replaced. The instrument was very simple to use with no maintenance required.
Table 3. Area Counts Precisioundefined
Compound Conc % Mean SD % RSD
nC6 0.06 3319066 35317 1.06
nC7 0.06 3377209 44156 1.31
nC8 0.08 4520913 54669 1.21
nC9 0.08 4600788 71279 1.55
nC1 0 0.12 6960088 111591 1.60
nC11 0.12 6908294 110542 1.60
nC1 2 0.12 6936057 114187 1.65
nC1 4 0.12 7018981 118351 1.69
nC1 6 0.10 5745943 98803 1.72
nC1 8 0.05 2889801 50764 1.76
nC2 0 0.02 1157439 19957 1.72
nC2 4 0.02 1134537 19205 1.69
nC2 8 0.01 563181 9833 1.75
nC3 2 0.01 559222 9540 1.71
nC3 6 0.01 551708 10166 1.84
nC4 0 0.01 539767 9346 1.73
nC4 4 0.01 535884 10082 1.88
1. D2887-93 "Standard Test Method for Boiling Range Distribution of Petroleum Fractions by Gas Chromatography",, Vol. 05.02 Annual Book of ASTM Standards
2. C.A. Saravalle, F. Munari, S. Trestianu, "Influence of Sample Solvent and Stationary Phase Polarity on Peak Broadening, Distortion and Splitting due to the Flooding Effects", 279 (1983) 241-242.Book of ASTM Standards Journal of Chromatography
Evaluation of a New Cold On-Column Capillary Inlet for Simulated Distillation of Petroleum Products
Introduction
In the petroleum refining process the boiling point distribution range of the feedstock streams is very important and can be used for product specification testing. The analysis is determined by gas chromatography (GC) as specified in ASTM D2887. The method is applicable to samples with an initial boiling point (IBP) of greater than 55°C and a final boiling point (FBP) of less than 538°C. Since a conventional splitless injection has shown discrimination over a wide boiling point range, the cold on-column injection technique is preferred. The separation, sensitivity, and precision criteria for Method D2887 were evaluated using a CE Instruments GC cold on-column injector with a Flame Ionization Detector (FID) and Model AS 800 auto-sampler. All GC parameters were optimized as specified in the method. The CE Instrument automated cold on-column injector and FID provided the instrument performance required meeting the criteria for Method D2887.
Experimental
The GC was set up for on-column injection with an auto sampler. The samples and standards were diluted in carbon disulfide (CS2). Air was used as the coolant for the cold on-column injector, Figure 1. A guard column was connected to the analytical column with a press fit connector. The instrument parameters are listed below: CE Instruments GC
Instrument Parameters:
Cold On-column inlet at ambient
OC Coolant valve: On for 60 seconds with air
Oven: 55°C, 20°C/min; 360°C, 10 minutes
FID: 360°C
H2: 30 mL/min
Air: 300 mL/min
Make Up: 10 mL/min nitrogen
Range: 1
Column:Restek Rtx 2887 0.53 mm × 10 m; 2.65 micron film,
Column Flow: 5 mL/min nitrogen
Guard Column: Restek Phenylmethyl deactivated 0.53 mm×5m
Injection Volume: 1 µL
AS800 Autosampler
The Injection:
The method specifies that when using an on-column injector, the entire capillary column must be temperature programmed including the point of sample introduction. In an on-column injection, the needle must enter the column inside the oven. The injector used has a butterfly valve, which is opened as the turret of the autosampler rotates over the injector, . The sample is injected through a septum into the front section of the column in the oven with a syringe with an 80 mm long needle. With solvents that have lower boiling points than the analytes, the oven is set below the boiling point of the solvent at the time of injection and no secondary cooling of the inlet is required. The sample is introduced as a liquid plug, and as the solvent evaporates the analytes are focused. Figure 2 is a representative chromatogram of an injection below the boiling point of the solvent of a diesel range organics (DRO) standard in methylene chloride.
Figure 2. Cold On-column Injection of DRO Standard in Methylene Chloride at 50 ppm each with an Oven Programof 35°C, 2 min; 10°C/min; 290°C, 10 min
When a heavier solvent like CS2 is used, the more volatile analytes like C6H14 through C11H22 show the adverse effects of solvent flooding. To eliminate peak distortion, the oven is set at 10-15°C above the boiling point of the solvent and the secondary coolant valve for the injector is turned on during the injection to prevent flashing back of the sample inside the inlet.
Figure 3. Adverse effects of Solvent Flooding with CS2 at Different Initial Oven Temperatures
Although CS2 does not respond on the FID, the adverse effects of the solvent may be seen in starting at lower initial oven temperatures as shown in Figure 3.The literature has shown that "the spreading of the liquid sample along the inlet part of the column produces broadening, distortion, and peak splitting. This flooding effect of the solvent may be eliminated by the incorporation of a guard column and operation of the initial oven temperature in a transition region at 10-15°C above the boiling point of the solvent. By injecting the sample above the boiling point of CS2 , the solvent plug is reduced to an insignificant volume. The injection of the D2887 standard into a 5 meter guard column connected to the analytical column at an initial temperature of 55°C eliminated the adverse flooding effects of the CS2 , Figure 4.
Figure 4. D2887 Standard at 1% in with an Oven Program of 55°C, 0 min,20°C/min; 360°C, 10 min (See Table 1 for concentrations)
The Separation
In order to simulate the true physical distillation in Method D2892, the chromatography was deoptimized. Nitrogen was selected for the carrier gas, because the resolution of C16H34 and C18H38 was too good with helium. The column flow was set to 5 mL/min to retard the elution of the IBP for the D2887 Reference Gas Oil to at least 1.0 minute, Figure 5. An oven program was established to elute all of the analytes through a boiling point of 538°C before the end of the ramp, Figure 4. The software automatically performed the baseline compensation analysis on all of the files using the first run of the solvent blank. Figure 6 shows the calibration curve generated from the data in Figure 4. Table 1 shows the uniformity of the hydrocarbons over the boiling point range of the D2887 calibration standard.
Table 1. RF
Compound Conc (%) Area RF
nC6 0.06 3964599 66076650
nC7 0.06 3379799 56329983
nC8 0.08 4450280 55628500
nC9 0.08 4496167 56202088
nC10 0.12 6798554 56654617
nC1 0.12 6751962 56266350
nC12 0.12 6779067 56492225
nC14 0.12 6862533 57187775
nC16 0.10 5615236 56152360
nC18 0.05 2820382 56407640
nC20 0.02 1129601 56480050
nC24 0.02 1108412 55420600
nC28 0.01 550048 55004800
nC32 0.01 544953 54495300
nC36 0.01 538015 53801500
nC40 0.01 528177 52817700
nC44 0.01 524163 52416300
Mean 56107908
SD 2911128
% RSD 5.18
Results
Eight replicates of the D2887 standard were analyzed to check the precision of the instrument. The retention time had a standard deviation of 0 for most of the compounds with a value of +/0.005 minutes for C44H99 . The relative percent standard deviation for response in area counts was less than 1.88%. The data is shown for the retention time and area counts as listed in Tables 2 and 3. The FID was set at a less sensitive range value from 0 to 1. No skewing of the peaks was seen. A customer crude oil sample was run on the system and shown in an overlay view with the D2887 Reference Gas Oil and calibration standard in .Figure 7
Table 2. Retention Time Precisioundefined
Compound Mean SD % RSD
nC6 0.868 0.0028 0.326
nC7 1.233 0.0000 0.000
nC8 1.751 0.0028 0.162
nC9 2.408 0.0000 0.000
nC1 0 3.129 0.0043 0.137
nC11 3.858 0.0000 0.000
nC12 4.567 0.0000 0.000
nC14 5.877 0.0037 0.063
nC16 7.057 0.0028 0.040
nC18 8.108 0.0000 0.000
nC20 9.067 0.0000 0.000
nC24 10.750 0.0000 0.000
nC28 12.183 0.0000 0.000
nC32 13.433 0.0000 0.000
nC36 14.548 0.0037 0.025
nC40 15.598 0.0037 0.024
nC44 16.913 0.0048 0.028
Conclusion
The CE Instruments automated cold on-column injection technique met the criteria for ASTM Method D2887 and showed uniform response over the boiling point distribution range of the method. The separation and reproducibility were well within the guidelines of the method. The sensitivity of the FID was more than required for the method and had to be run at a less sensitive amplification. The unique design of the autosampler and inlet provided the critical alignment of the needle on injection, so that the same hole in the septum is penetrated each time. The septum did not leak during the entire time of the study and was never replaced. The instrument was very simple to use with no maintenance required.
Table 3. Area Counts Precisioundefined
Compound Conc % Mean SD % RSD
nC6 0.06 3319066 35317 1.06
nC7 0.06 3377209 44156 1.31
nC8 0.08 4520913 54669 1.21
nC9 0.08 4600788 71279 1.55
nC1 0 0.12 6960088 111591 1.60
nC11 0.12 6908294 110542 1.60
nC1 2 0.12 6936057 114187 1.65
nC1 4 0.12 7018981 118351 1.69
nC1 6 0.10 5745943 98803 1.72
nC1 8 0.05 2889801 50764 1.76
nC2 0 0.02 1157439 19957 1.72
nC2 4 0.02 1134537 19205 1.69
nC2 8 0.01 563181 9833 1.75
nC3 2 0.01 559222 9540 1.71
nC3 6 0.01 551708 10166 1.84
nC4 0 0.01 539767 9346 1.73
nC4 4 0.01 535884 10082 1.88
1. D2887-93 "Standard Test Method for Boiling Range Distribution of Petroleum Fractions by Gas Chromatography",, Vol. 05.02 Annual Book of ASTM Standards
2. C.A. Saravalle, F. Munari, S. Trestianu, "Influence of Sample Solvent and Stationary Phase Polarity on Peak Broadening, Distortion and Splitting due to the Flooding Effects", 279 (1983) 241-242.Book of ASTM Standards Journal of Chromatography
