Upgrade to a Faster D2887 Analysis with a GC Accelerator Kit
Reduce Analysis Time to 9 Minutes in a 120 V GC
Abstract
The results of a simulated distillation analysis are vitally important to the operation of refineries globally. For the analysis of petroleum samples within the boiling point range from 55.5 ˚C to 538 ˚C, ASTM D2887 is a standard method that has been accepted and is used industry-wide. This application note will demonstrate how a creative use of the Restek GC Accelerator oven insert kit can permit analysts using Agilent 6890/7890 GCs with 100/120 V ovens to successfully migrate from D2887’s slower Procedure A conditions to the accelerated Procedure B conditions without new instrumentation or software, resulting in a 9-minute analysis time that meets all method requirements.
Introduction
The distribution of boiling point temperatures of petroleum products or liquid fuels is one of the properties that govern the quality of these products. That distribution is also used for the optimization of refinery processes and the characterization and prediction of other physical parameters of the fuels. Traditionally, the boiling point distribution range is determined using classical physical distillation, which results in volumetrically measurable fractions of known boiling point ranges. A major disadvantage of physical distillation is that it can be a labor-intensive, time-consuming process that requires large sample volumes.
However, simulated distillation (SimDist) can be a much more efficient alternative. SimDist analysis is a gas chromatographic technique designed to provide the same information as a physical distillation. The SimDist approach is simpler and faster, and it can be automated. It also requires significantly less sample.
ASTM D2887 [1] is a SimDist method, published by ASTM International (formerly the American Society for Testing and Materials), for analyzing samples with a boiling point range of 55.5 °C to 538 °C (C5-C44). The method contains two procedures: Procedure A and Procedure B, which each include several options that offer a variety of column choices and analytical conditions. The boiling point range for ASTM D2887 includes light crudes, naphtha, and kerosene, and Procedure B allows for a variety of biodiesel mixtures (B5, B10, and B20).
Procedure B was developed to significantly reduce the analysis time of the original method by taking advantage of faster oven heating rates, high flows, and open tubular capillary columns. A 0.53 mm ID column facilitates the use of high flow rates, and a thick film provides the sample loading capacity to allow neat samples to be injected without sacrificing peak shape due to overloading. The oven heating rate is 35 °C/min from 40 °C to 350 °C. An Agilent GC equipped with a 100/120 V oven cannot meet these aggressive oven heating rates, especially at temperatures greater than 250 °C, so typically analysts are restricted to using the longer Procedure A methods, which often take more than 20 minutes. This study investigated a novel use of the Restek GC Accelerator oven insert kit to increase the oven ramp rate of an Agilent GC with a 100/120 V oven in order to meet the requirements for Procedure B oven ramp rates.
Experimental
Sample
The instrument was set up with a GC Accelerator (plate only) installed and evaluated against the method requirements using an ASTM D2887-12 calibration standard (cat.# 31674). The configuration was further evaluated using a reference gas oil (RGO) standard. RGO oil is an oil sample with a certified boiling range distribution. It is run to evaluate the system prior to analyzing samples. The RGO values must be in agreement with the published values for the lot of RGO that is used (Separation Systems, reference gas oil, lot #1, batch #2, cat.# SD-016-05).
Analysis Conditions
As shown in Table I, the D2887 analysis conditions followed ASTM Method D2887-16a [1], Procedure B, Column Option 2. The sample amount was reduced from 0.1 µL to 0.05 µL.
For this study, a 10 m, 0.53 mm ID, 2.65 µm film thickness MXT-1HT SimDist column was cut to the 5 m length called for by the Procedure B, Column 2 method. A custom 5 m column is available (contact your Restek representative and request cat.# 573912).
The PTV purge flow was turned on near the end of the analysis (at 8 minutes) to mitigate the potential for sample carryover, while allowing enough time to achieve reproducible and representative sample transfer from the inlet to the column.
Table I: D2887 Analysis Conditions
ASTM D2887 Procedure B, Column 2 Option | |
---|---|
Column | 5 m x 0.53 mm x 2.65 µm MXT-1HT SimDist |
Inlet | PTV 100 °C to 350 °C at 35 °C/min, purge flow 100 mL/min at 8 min |
Flow Conditions | Helium, 35 mL/min, constant flow |
Oven Program | 40 °C to 350 °C at 35 °C/min with GC Accelerator plate installed |
Detector | FID @ 360 °C Air 400 mL/min, hydrogen 40 mL/min, make-up gas (nitrogen) at 5 mL/min |
Sample | ASTM D2887-12 calibration standard (Restek, cat.# 31674); Reference gas oil (Separation Systems, lot# 1, batch #2, cat.# SD-016-05) |
Sample Size | 0.05 µL injection, autosampler, D2887 standard sample was heated 5 min at 40 °C prior to injection |
Liner | Topaz 2.0 mm ID single baffle CIS4/TDU inlet liner (cat.# 23283) |
Injection | Splitless, 0.05 µL injection, autosampler, slow sample drawing speed, fast injection |
Instrument Setup and GC Accelerator Installation
An Agilent 7890B equipped with a 120 V oven was used for this study. This particular GC is designed for two column analyses; therefore, it has two inlets and two detectors. For this experiment, the column was installed at the back PTV inlet and the back FID detector. This was to facilitate the novel use of the GC Accelerator kit.
The GC Accelerator oven insert kit (cat.# 23849) comes with one plate insert, one large block insert, and one small block insert and was originally intended to be installed in an Agilent GC to accommodate a column installed in the front or back inlet positions, the front or back detector positions, or principally, for use with a mass spectrometer. Installed in this way, however, a 120 V oven is unable to achieve the 35 °C/min ramp rate called for by the method at the upper oven temperature range (300-350 °C).
In this study, only the plate insert was used, and it was installed behind the front inlet and detectors in the GC oven (Figure 1). This position further reduced the volume of the oven and allowed the oven to meet the accelerated oven ramp rate required for this method (Figure 2). With the plate installed in this position, there is no room for the large or small block inserts, so they were not used.
Data Analysis Software
Since the required data analysis for actual samples can be a mathematically labor-intensive process, the results are routinely evaluated using one of the commercially available SimDist software packages. In this study, Dragon SimDist software developed by Envantage was used. The software comes pre-built with all the data calculations, performance measurement tools, and reporting requirements for all normal and high-temperature simulated distillation analyses.
Results and Discussion
D2887 Analysis of Retention Time Calibration Standard
The setup described here was initially tested using a calibration standard and evaluated against the method requirements for skewness (asymmetry at 10%) and sample discrimination. The chromatogram is shown below in Figure 3. Table II demonstrates how this novel setup permits an analyst with an Agilent 100/120 V oven to meet Method D2887 skewness and discrimination requirements. Measuring the ratio of a given peak’s area to the area of decane (C10) provides a measure of discrimination, or unintended loss of signal for a given compound, where a value of 1 is ideal. The skewness, or asymmetry, of the peak is a measure of potential column overloading where again, a value of 1 represents a perfectly symmetrical peak.
Figure 3: D2887 analysis of a calibration standard following method D2887-16a, Procedure B, Column 2 conditions (Table I) using the GC Accelerator plate in a 120 V GC oven.
Column | See notes |
---|---|
Standard/Sample | ASTM D2887-12 calibration standard (cat.# 31674) |
Diluent: | Carbon disulfide |
Conc.: | 1% w/w |
Injection | |
Inj. Vol.: | 0.05 µL PTV splitless |
Liner: | Topaz 2.0 mm ID single baffle CIS4/TDU inlet liner (cat.# 23283) |
Inlet Temp. Program: | 100 °C to 350 °C at 35 °C/min |
Oven | |
Oven Temp.: | 40 °C to 350 °C at 35 °C/min |
Carrier Gas | He, constant flow |
Flow Rate: | 35 mL/min |
Detector | FID @ 360 °C |
---|---|
Instrument | Agilent 7890B GC |
Notes | The analytical column used here is a 5 m section cut from a 10 m, 0.53 mm ID, 2.65 µm MXT-1HT SimDist column. A 5 m length can be ordered by contacting your Restek representative and requesting custom part number 573912. The analysis was run in a 120 V oven using the plate insert from a GC Accelerator kit (cat.# 23849) in order to meet the aggressive oven ramp rate. |
Table II: Evaluation of the chromatogram in Figure 3 for skewness (asymmetry at 10%) and sample discrimination (Cx/C10 area ratio). All the method requirements were met.
Compound | Structural Name | Boiling Point (°C) | Time | Area | Cx/C10 Area Ratio | Cx/C10 Spec | Asymmetry (10%) | Asymmetry Spec |
---|---|---|---|---|---|---|---|---|
n-Pentane | C5 | 36 | — | — | — | — | — | — |
n-Hexane | C6 | 69 | 0.15 | 8665 | 0.93 | 0.9-1.1 | 1.06 | 0.8-1.3 |
n-Heptane | C7 | 98 | 0.249 | 9239.3 | 0.99 | 0.9-1.1 | 1.04 | 0.8-1.3 |
n-Octane | C8 | 126 | 0.438 | 9267.2 | 0.99 | 0.9-1.1 | 1.01 | 0.8-1.3 |
n-Nonane | C9 | 151 | 0.726 | 9495.9 | 1.01 | 0.9-1.1 | 1.00 | 0.8-1.3 |
n-Decane | C10 | 174 | 1.085 | 9362.1 | 1.00 | 0.9-1.1 | 0.98 | 0.8-1.3 |
n-Undecane | C11 | 196 | 1.469 | 9314.8 | 0.99 | 0.9-1.1 | 0.98 | 0.8-1.3 |
n-Dodecane | C12 | 216 | 1.852 | 9318.9 | 1.00 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Tetradecane | C14 | 254 | 2.573 | 9379.6 | 1.00 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Pentadecane | C15 | 271 | 2.907 | 9318.7 | 1.00 | 0.9-1.1 | 0.96 | 0.8-1.3 |
n-Hexadecane | C16 | 287 | 3.224 | 9363.1 | 1.00 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Heptadecane | C17 | 302 | 3.524 | 9344.4 | 1.00 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Octadecane | C18 | 316 | 3.811 | 9363.9 | 1.00 | 0.9-1.1 | 0.96 | 0.8-1.3 |
n-Eicosane | C20 | 344 | 4.344 | 9355.8 | 1.00 | 0.9-1.1 | 0.96 | 0.8-1.3 |
n-Tetracosane | C24 | 391 | 5.28 | 9520.4 | 1.02 | 0.9-1.1 | 0.98 | 0.8-1.3 |
n-Octacosane | C28 | 431 | 6.081 | 9374.2 | 1.00 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Dotriacontane | C32 | 466 | 6.778 | 9259.2 | 0.99 | 0.9-1.1 | 0.98 | 0.8-1.3 |
n-Hexatriacontane | C36 | 496 | 7.394 | 9253.6 | 0.99 | 0.9-1.1 | 0.98 | 0.8-1.3 |
n-Tetracontane | C40 | 522 | 7.946 | 9252.5 | 0.99 | 0.9-1.1 | 0.97 | 0.8-1.3 |
n-Tetratetracontane | C44 | 545 | 8.445 | 9277.6 | 0.99 | 0.9-1.1 | 0.97 | 0.8-1.3 |
When plotting retention times against boiling points, it is typical to observe a curve that is nonlinear for the first four paraffins (C5-C8), but the plot must be linear from C9-C44. Figure 4 illustrates the performance of the method under the conditions and the setup used in this study. The results show the expected performance for D2887 analysis.
If the GC oven had not been able to achieve or maintain the necessary temperature program, the variation would result in a nonlinear curve, especially for the later eluting compounds where the temperature lag would be more pronounced. The agreement shown is further evidence that the novel use of the GC Accelerator plate insert upgrades performance to meet the accelerated method conditions, allowing labs to speed up their D2887 analyses using the thoroughly vetted and approved method conditions found in Procedure B with their existing 100/120 V GC oven.
Analysis of Reference Gas Oil (RGO) Sample
Using the certified RGO sample and the Dragon SimDist software package, the setup was evaluated to determine if it would produce results that agreed with the published boiling points, which is required by the method to evaluate a system prior to the analysis of test samples. An initial instrument blank was also collected to demonstrate system cleanliness.
Figure 5 illustrates the chromatogram from D2887 analysis of the RGO sample using the GC Accelerator plate with data analysis performed by the software package overlaid. Table III shows the passing results from the analysis, demonstrating that the GC Accelerator plate insert allowed the 100/120 V GC oven to meet the method requirements.
Table III: Dragon SimDist report of D2887 analysis of RGO sample. All the percent mass increments are in good agreement and within the permitted deviation range from the certified RGO values.
Reference Gas Oil QC Detail - Type: RGO, Lot# 2 | |||||
Percent Off | Certified BP Temp (°C) | Temp Window (°C) | Actual BP Temp (°C) | Variance | Result |
IBP | 115.556 | 7.611 | 113.961 | -1.594 | -PASSED- |
10 | 175.556 | 4.111 | 174.216 | -1.340 | -PASSED- |
20 | 223.889 | 4.833 | 222.207 | -1.682 | -PASSED- |
30 | 259.444 | 4.667 | 258.292 | -1.152 | -PASSED- |
40 | 288.889 | 4.278 | 288.846 | -0.043 | -PASSED- |
50 | 312.222 | 4.278 | 311.961 | -0.262 | -PASSED- |
60 | 331.667 | 4.278 | 331.19 | -0.477 | -PASSED- |
70 | 353.333 | 4.278 | 353.626 | 0.293 | -PASSED- |
80 | 377.778 | 4.278 | 378.145 | 0.367 | -PASSED- |
90 | 406.667 | 4.278 | 407.524 | 0.857 | -PASSED- |
95 | 428.333 | 4.278 | 429.713 | 1.380 | -PASSED- |
FBP | 475.556 | 11.778 | 476.986 | 1.431 | -PASSED- |
Method Precision (Repeatability)
The method’s repeatability was measured from nine replicate injections of RGO oil and a coefficient of variation was calculated from these values. A coefficient of variation, or relative standard deviation, is expressed as the standard deviation divided by the average of data set in percent. The data set, shown in Table IV, resulted in a spread below 0.5%, which demonstrates acceptable method precision.
Table IV: Dragon SimDist values for nine replicate injections of RGO with calculated standard deviations and relative standard deviations.
Percent Off | Certified BP Temp (°C) | Temp Window (°C) | Run 1 | Run 2 | Run 3 | Run 4 | Run 5 | Run 6 | Run 7 | Run 8 | Run 9 | Std. Dev. | % RSD |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
0.5 | 115.6 | 7.6 | 114.0 | 114.9 | 114.7 | 115.2 | 114.2 | 115.0 | 114.0 | 115.3 | 115.5 | 0.6 | 0.5 |
10 | 175.6 | 4.1 | 174.2 | 175.7 | 175.1 | 176.0 | 175.1 | 175.9 | 174.9 | 175.6 | 176.0 | 0.6 | 0.3 |
20 | 223.9 | 4.8 | 222.2 | 225.1 | 223.9 | 225.5 | 224.6 | 225.5 | 224.1 | 224.5 | 225.3 | 1.1 | 0.5 |
30 | 259.4 | 4.7 | 258.3 | 260.5 | 259.6 | 260.9 | 260.2 | 260.8 | 259.8 | 260.0 | 260.7 | 0.8 | 0.3 |
40 | 288.9 | 4.3 | 288.9 | 290.0 | 289.5 | 290.3 | 289.9 | 290.2 | 289.7 | 289.8 | 290.2 | 0.5 | 0.2 |
50 | 312.2 | 4.3 | 312.0 | 312.7 | 312.4 | 312.9 | 312.7 | 312.9 | 312.6 | 312.6 | 312.9 | 0.3 | 0.1 |
60 | 331.7 | 4.3 | 331.2 | 331.5 | 331.5 | 331.7 | 331.6 | 331.6 | 331.5 | 331.5 | 331.6 | 0.1 | 0.0 |
70 | 353.3 | 4.3 | 353.6 | 354.0 | 353.9 | 354.2 | 354.1 | 354.1 | 354.0 | 354.0 | 354.1 | 0.2 | 0.0 |
80 | 377.8 | 4.3 | 378.2 | 378.4 | 378.4 | 378.5 | 378.5 | 378.5 | 378.4 | 378.4 | 378.5 | 0.1 | 0.0 |
90 | 406.7 | 4.3 | 407.5 | 407.6 | 407.6 | 407.8 | 407.8 | 407.7 | 407.7 | 407.7 | 407.7 | 0.1 | 0.0 |
95 | 428.3 | 4.3 | 429.7 | 429.8 | 429.8 | 429.9 | 429.9 | 429.9 | 429.9 | 429.9 | 429.9 | 0.1 | 0.0 |
99.5 | 475.6 | 11.8 | 477.0 | 477.1 | 477.3 | 477.1 | 477.2 | 477.1 | 477.2 | 477.2 | 477.2 | 0.1 | 0.0 |
Conclusion
A novel use of the GC Accelerator oven insert kit in an Agilent instrument with a 120 V oven upgraded the instrument’s performance to achieve the accelerated D2887-16 Procedure B method conditions where the oven ramp rate is 35 °C/min from 40 °C to 350 °C. All the method requirements were met in for the D2887 analyses performed, making this a viable option for labs wanting to take advantage of a faster 9-minute analysis time, while still using the industry-accepted method conditions established in ASTM method D2887-16.
References
- ASTM D2887-16a, Standard Test Method for Boiling Range Distribution of Petroleum Fractions by Gas Chromatography, ASTM International, West Conshohocken, PA, 2016. https://www.astm.org/Standards/D2887.htm