GC2 PFT Analysis: User Guide
Rev. 376 | June 2018 – V.Percuoco
Draft 6/7/2011; Rev. 1/2/1014; Rev. 371T | July 2017 – D.Houpt
The major concern in shipboard microbiological study is whether microbes from the drilling fluid are introduced into the recovered core material during coring. Therefore, it is critical to verify whether recovered cores are contaminated. Perfluorocarbon tracer (PFT) can be used to quantify the amount of contamination due to drilling fluid. It is strongly recommended that this test be routinely conducted when coring for microbiological studies.
PFTs are chemically inert and can be detected with high sensitivity. The JRSO has two chemicals it uses as chemical tracers to monitor potential contamination of sediment and rock samples on the JOIDES Resolution.
Table 1: Physical and chemical properties of perfluorocarbon tracers
Property | PFMCH | PFMD |
CAS Number |
|
|
Molecular Formula | C7F14 | C11F20 |
Molecular Weight (g/mol) | 350.05 | 512.09 |
Boiling Point °C | 76 | 160 |
Density (g/mL) | 1.788 | 1.972 |
Solubility in Water (mg/L) | ~2 | ~10 |
Solubility in Methanol (mg/L) | 104 |
|
Solubility in Hexane (mg/L) |
| 470,000 |
Vapor Pressure @ 25C (kPa) | 14.11 | 0.29 |
Both perfluoromethyldecalin (PFMD) and perfluoromethylcyclohexane (PFMCH) are miscible in each other. The vapor pressure of PFMCH is fairly high so it evaporates readily and quickly at standard room pressure and temperature. The evaporation of PFMD is less significant. The low solubility in water for either compound facilitates gas-phase partitioning and quantitative headspace analysis.
The purity of PFMD purchased from Oakwood Chemical is approximately 85-90%. The predominant contaminants tend to be perfluorodecalin and perfluoro-tert-cyclohexane. The purity of a batch may be found by navigating to Oakwood Chemical's website and entering the LOT # located on the bottle label.
At 70°C, 100% of the PFMCH can be expected to be in the volatile phase, whereas the Antoine-derived curve for PFMD predicts that the partial pressure of PFMD in the 20 mL vial would be approximately 3% of atmosphere (0.03 bar).
It is impossible to heat the vial to the boiling point of the PFMD, however, because it will boil the water in the samples and exceed the pressure capacity of the vial. Even temperatures close to 100°C will liberate significant water vapor, which can be problematic for the gas chromatography conditions.
It is therefore recommended that the samples be heated at 85°C, at which temperature, the PFMD can be expected to have a partial pressure of approximately 6% of atmosphere (0.06 bar). Although the majority of the PFMD is not volatilized in the vial, it is consistently volatilized at a steady temperature in both the standard vials and the sample vials, so a consistent concentration of PFMD will be measured by the GC-µECD.
PFT is continuously fed into the stream of drilling fluid using an Alltech 301 high-performance liquid chromatography (HPLC) pump located within a black cabinet in the Mud Pump room (Figure 1). The tracer is delivered into the drilling fluid stream through a valve on the low-pressure side of the mud charge pump (Figure 2). The rate of the tracer injection is tied to the pumping rate of the drill fluid in order to maintain a final concentration of ~1 mg/L in the drilling fluid through the entire drill string. Operation of the HPLC pumps is usually controlled autonomously by the drill shack. In order to manually operate either pump:
Ensure the lines exiting the cap of the tracer reservoir are sealed with putty to prevent tracer from evaporating away.
|
|
|
|
Figure 5: Different syringe types for preparing PFT standards and for injecting samples. Listed from top to bottom: 1 µL glass analytical syringe from Scientific Glass Engineering (SGE), 0.5 mL Teflon-fitted Pressure-Lok glass syringe from Precision Sampling Corp, 10 µL glass analytical syringe from SGE, 0.10 mL (100 µL) Microliter #710 glass syringe from Hamilton Co.
The insolubility of PFMCH creates challenges in creating the dilution curve. In order to deal with this problem, a volatile gas dilution scheme was created as follows:
Stock solution is pure PFMCH. Follow the dilution schema in Table 1.
Calibration Level | Reagent added to 20 mL crimp top headspace vial |
Concentration (ng/mL [ppb]) |
7 | 10 µL PFMCH | 900,000 |
6 | 0.5 mL Level 7 (900,000 ng/L) | 22,500 |
5 | 0.1 mL Level 7 (900,000 ng/L) | 4,500 |
4 | 0.5 mL Level 6 (22,500 ng/L) | 562.5 |
3 | 0.5 mL Level 5 (4,500 ng/L) | 112.5 |
2 | 0.5 mL Level 4 (562.5 ng/L) | 14.1 |
1 | 0.5 mL Level 3 (112.5 ng/L) | 2.81 |
Table 1. Dilution scheme for PFMCH.
The crimp top headspace septa are good for only a few injections; remake standards after five or six injections.
Perfluoromethyldecalin (PFMD) Standard CurveTable 1. Dilution scheme for PFMCH.
The crimp top headspace septa are good for only a few injections; remake standards after five or six injections.
PFMD's ready solubility in hexane (47% w/v) makes serial dilutions of this tracer much more straightforward, but does require some caution on the user's part because the hexane will evaporate at the oven temperature in the incubation oven.
Warning! The user should be careful how much of the hexane-dissolved standard is injected into a headspace vial. At the incubation oven temperature, nearly 100% of the hexane will move into the gas phase. Adding 1 mL of hexane to a 20 mL headspace vial will create nearly 11 atmospheres of pressure at 70°C, which will likely shatter the vial!
Assumptions: All PFMD is volatilized in the vial when heated to 70°C.
Use the following equation to calculate the concentration of the primary standard (PFMD in hexane solvent).
[STD1°] = Concentration of PFMD in standard (g/mL hexane)
pPFMD= Density of PFMD (1.972 g/mL)
Purity = Purity of the PFMD solution (%). Enter the LOT # from the bottle on Oakwood’s website to find the purity of the bottle
VPFMD = Volume of PFMD pipetted into the hexane solvent (mL)
Vhexane = Volume of hexane solvent (mL)
Use the following equation to calculate the concentrations for secondary standards prepared from the primary standard:
[STD2°] = Concentration of PFMD in secondary standards (g/mL hexane)
V1° = Volume primary standard pipetted into the headspace vial (mL)
Vvial = Volume of the headspace vial (20 mL)
The calibration curve consists of measurements of the secondary standards. While extracting, the additional volume of the connected syringe adds to the total volume of the vial and thus slightly dilutes the PFMD concentrations. This factor is taken into account in the following equation using the ratios of the syringe volume to the vial and syringe volumes. Injected mass would otherwise have an error of 12.5% (for a 2.5 mL extraction of a 20 mL vial).
To determine the mass of PFMD injected in the GC from an extraction of the secondary standard:
M = mass of PFMD injected on column (g)
Vsyringe = Volume of the secondary standard extracted via the autosampler or manual syringe.
Using a cemented needle syringe, add 5.1 µL of neat PFMD reagent into 25 mL of Optima Grade hexane (mixed hexanes will serve equally well).
Prepare serial dilutions of stock level A into separate headspace vials. Use a cemented needle syringe from SGE Corp. (not a plastic-tip pipettor) to add the specified levels (Table 1) of stock solution A to 20 mL crimp-top headspace vials by injecting through the septum. Use the high precision 1-µL analytical syringe to accurately pipette small volumes. The crimp top headspace septa are good for only a few injections; remake standards from the stock solutions after five or six injections.
Table 2: Calibration standards dilution scheme
Batch | Calibration Level | Reagent STD A added to 20 mL crimp top headspace vial | Concentration (ng/mL headspace) |
Low Level | 4 | 1 µL | 20 |
3 | 0.75 µL | 15 | |
2 | 0.50 µL | 10 | |
1 | 0.25 µL | 5 | |
Blank | 0 µL | 0 | |
High Level | 5 | 62 µL | 1,240 |
4 | 31 µL | 620 | |
3 | 6.2 µL | 124 | |
2 | 0.78 µL | 15.5 | |
1 | 0.25 µL | 5 | |
| Blank | 0 µL | 0 |
Table 2: Serial dilution scheme for PFMD.
The GC2 system comprises an HP 6890 gas chromatograph (GC) with a micro-electron capture detector (µECD).
The GC inlet is operated in splitless mode. PFT gas samples obtained using the headspace extraction method may be injected manually after incubation for 30 minutes at 85 deg. C, or can be injected by the Gerstel autosampler (whose incubator oven should be set to 85°C). The injection port liner assembly is connected to a megabore column (Rt-Alumina BOND/KCl, 50 m, 0.53 mm ID, 10 µm thickness), and then to a µECD detector, which requires both carrier and makeup gases (nitrogen).
Ensure the syringe installed in the autosampler has the Teflon-tipped plunger (Figure 6).
Figure 6: Different syringes used by the Gerstel Autosampler. The rubber plunger of the syringe shown on top causes significant sample carryover, likely due to tracer penetrating pore spaces within the rubber. It is best to use the syringe with the teflon-tipped plunger shown on bottom.
Nitrogen gas is used in all three flow lines (column carrier, detector carrier, and makeup gases). Nitrogen suffices as the detector makeup gas for this procedure because chromatographic efficiency is not an issue and it is readily available aboard ship because of the nitrogen generator.
The µECD is designed to operate best with a flow rate of at least 20 mL/min. Carrier flow of capillary columns, typically 10 mL/min, requires make-up gas to ensure the optimum total flow rate for the detector.
Nitrogen supply settings are:
The µECD cell contains 63Ni, a radioactive isotope emitting high-energy electrons (β-particles) with a nominal radioactivity of 10 mCi. These undergo repeated collisions with carrier gas molecules, producing ~100 secondary electrons for each initial β-particle.
Further collisions reduce the energy of these electrons into thermal range. These low-energy electrons are then captured by suitable sample molecules, which reduces the total electron population within the cell. Therefore, with higher sample concentration the conductivity of an existing gas will drop noticeably, which is recorded by the µECD outcoming signal detector.
(Note that the raw signal represents a drop in electron current signal, flipped over to positive peaks through the GC electronics and software.)
PFT is pumped into the drilling fluid during coring. When core is delivered to the deck, small core samples are placed in headspace vials, sealed, and heated before headspace analysis on the GC2. The presence of a PFT peak from a sample from the interior of a core indicates core contamination from drill fluid, which may contain contaminating microbes.
Sediment samples are collected from the edge and center of the core on the catwalk immediately after cores are retrieved. The sample from the outer edge is used to confirm successful delivery of the tracer to the core, whereas the interior sample is used to estimate the quantity of intrusion of drill water into the core. Because the exterior of the core liner is coated with drilling fluid, contact with the liner should be avoided while collecting core samples for PFT analysis.
Sample analysis includes the following steps:
Important! The nitrogen gas supply to the Agilent 6890 GC-µECD must be of sufficient purity to protect the 63Ni source, so before the detector is brought to operating temperature, be sure that no significant nitrogen demands are being made throughout the laboratory. For example, if the microbiologists are using the "Berkley bucket" technique to flush nitrogen through a container, do not proceed.
If time-critical measurements must be made without waiting for other usage to go down, talk to the Laboratory Officer about hooking up a UHP nitrogen tank from the reserve tanks in the hold.
The area of the PFT peak is integrated and converted to the amount of PFT using values from the standard curve. The amount of sample is determined by weighing each vial and subtracting the weight of an empty vial. The total headspace volume is calculated by subtracting the volume of sample from the total volume of the vial. Total tracer concentration in the sample is corrected to account for the fraction of the headspace that is injected. The amount of drilling fluid in the sample is calculated assuming that the tracer was present at 1 mg/L.
Use the following equations to determine the amount of drill-water intrusion in a sample:
(Drill water, L)/Core material, g)/[(PS – PB)/(CDW x a x W x FI)]
where
PS = integrated peak area of PFT in sample (in arbitrary units),
PB = integrated peak area of PFT in blank (in arbitrary units),
a = slope derived from the calibration curve (in arbitrary units per gram),
CDW = concentration of PFT in drilling fluid (in grams per liter),
W = weight of sample (in grams), and
FI = fraction of the total headspace gas injected:
Vinj/[Vvial – (W/rbulk)]
where
Vinj = volume of sample injected (in liters),
Vvial = volume of vial (in liters),
rbulk = sample density (in grams per liter), and
W = weight of sample (in grams).
Due to its cis-trans isomerism, perfluoromethylcyclohexane elutes as a peak doublet around 31 minutes. Since the isomers cannot be resolved by the column and temperature regime, it is best to integrate the area under both peaks. Other features of a typical chromatogram include a peak around 30 sec due to the co-elution of the permanent gases (O2, CO2, CO, etc); a small peak before 11 minutes due to the hexane solvent; and peaks at 22 (doublet), 24 (single), 29 (single), 33.5 (single), and 35.5 minutes (single) from other perfluorocarbons (perfluoromethyldecalin, and perfluoro-tert-cyclohexane) occupying nearly 10% by volume of the original PFMD reagent.
The ECD is extremely sensitive to changing concentrations of PFMD. A vial of 10 ppm in the headspace is enough to saturate the detector. In most cases, the concentrations analyzed in samples will be sub ppm level. Injecting high levels of tracer will cause significant carryover between samples, especially if the A/S is used.
Figure 7: Typical chromatogram for PFMD
Figure 8: Chromatogram in which PFMD has saturated the detector
The analytical batch is a group of samples run together with a single set of QC parameters, such as calibration/calibration verification, blank, and other QA/QC samples.
Blanks are analyzed to determine the instrumental and procedural backgrounds. These blanks consist of 0.25 mL injections of air collected in the gas-tight syringe from outside the laboratory or headspace gas from empty vials prepared at the same time and location the samples are taken.
Calibrating the instrument produces instrument response factors to absolute component concentrations. To prepare a calibration for quantitation of unknown samples, the retention time(s) for the peak(s) of interest and the amount of component injected must be known.
The graphical representation of the amount and response (peak area) for PFT from the calibration samples defines the calibration curve. Because the ECD is not linear across its range of detection, multiple calibration standards are run to calibrate for PFT. Various curve-fit calculations are available to determine optimum regression coefficient including linear, log, power, exponential, quadratic, and cubic.
The correlation coefficient is the square root of the regression coefficient and gives a measure of the fit of the calibration curve to the data points. The value of the correlation coefficient ranges from 0.000 (no fit) to 1.000 (perfect fit). The calibration coefficient for PFT must be >0.995 to be considered an acceptable calibration.
A multilevel calibration is valid over the range of concentrations used in the calibration samples. Extrapolation of a calibration curve, especially if it is not linear, gives at best an approximation result.
Primary safety issues are centered around the electron capture detector, the tracer handling, and GC oven operation
The electron capture detector measures the current flow caused by ¿-particle emission from the 63Ni foil source. This source is completely contained and is considered to be safe for humans if the system is properly maintained.
If the ECD detector is heated above ~150°C without supply gas flow or is overheated above ~400°C, the 63Ni foil internal to the detector may oxidize and be damaged. Such damage may result in a release of 63Ni into the laboratory atmosphere.
Perfluoromethylcyclohexane and perfluoromethyldecalin are used as the perfluorocarbon tracer compounds. These compounds (one of which, PFMCH, is highly volatile), is chemically inert and of extremely low toxicity. Although it is relatively harmless, PFMCH, especially, can permeate widely if not used under properly ventilated conditions and cross-contamination of environment-to-sample can occur.
Safe handling guidelines for the PFT chemical compound consist of the following:
Both PFMCH and PFMD are considered to be non-hazardous by US and EU classifications under normal conditions.
For procedure or GC operation problems, call a chemistry technician for help.
The rate at which PFMD is pumped is controlled by the Sci-ops engineer or by the drillshack. To change the flowrate the Sci-ops will need to manually delve into the RIGWATCH formula and to add a correction factor.
This is most likely due to a blockage in the lines or the pump not correctly primed. Prime the pump. Verify there are no visible leaks in the tubing and tubing connections. Verify that the valves after the pumps are correctly configured (Figure 4), if the pumps are switched, the values will need to be switched accordingly. If everything is configured correctly and no leaks are present, close the valve connected to the drilling fluid pipeline (Figure 2), remove the Swagelok fitting, press "Prime" on the operating HPLC pump and watch for tracer to begin flowing from the open line. If tracer does flow then the issue is likely due to a miscommunication with the Ethernet and RIGWATCH (see below). If the tracer is not flowing, work backwards along the flowpath to find a blockage.
When the PFT pump is turn on/off via RIGWATCH a command is sent to the National Instruments Ethernet box, and the NI box pings back RIGWATCH. The control display of RIGWATCH will indicate one of the two channels for the PFT pumps as "OFF" if the systems are miscommunicating. The miscommunication may be due to an incorrectly set BAUD rate within the NI box or RIGWATCH COM Ports. Verify that the pump and Ethernet box are on, and then ask the Marine Computer Specialist to verify/adjust the BAUD rates.
The manual autosampler control is overridden by ChemStation once the program is started. To enable manual control of the autosampler (i.e. controlling it by the physical A/S control panel) navigate within ChemStation to Instrument>Change PAL Configuration and select "Release Terminal". To give control back to ChemStation press "Lock Terminal" on the same menu. Use manual control to help in replacing bent needles.
If the A/S keeps bending needles, verify the current needle is straight then change out the A/S tension cord. The cord ensures the magnetic tip of the A/S stays in position as it injects. Otherwise, it collides with the cap of the sample inlet and raises slightly, missing the septum and bending the needle on the surrounding metal.
While the A/S syringe heats the heater block temperature readings will occasionally fluctuate between the actual temperature and -999.99C. If this occurs, enable manual control of the A/S, remove the syringe needle heater block and inspect the circuit contacts—four brass colored circles. Clean them off with a Kim Wipe then reinstall the heater block and give control of the A/S back to ChemStation. Try running a sample, if the problem persists, replace the heater block.
Issues with the chromatograms tend to be due to a poor injection, injecting water or particles, or the degradation of front inlet septa. Particles from the injection or the degradation of the septa cause periodic noisy baselines as the particulates elute the column. Baking out the column or cutting off a small portion of column at the inlet may help in clearing up contamination. Consult Agilent's user guides for walkthroughs. Important: Always note the temperature of the oven to ensure no gases above the ECD's temperature rating are contacting it.
An injection did not occur properly if the peak for the permanent gases is not present within the first minutes of the analysis. Try a second injection. If the second injection fails, change the septum. If this doesn't fix the problem, consult Agilent's user guide about cleaning out the front inlet.
To prevent permanent damage, never heat the column without adequate carrier gas flow through the column. In most flow failure cases, the system will give a warning beep followed by emergency shutdown procedures. Lack of carrier gas flow may be caused by:
A critical issue is column degradation due to oxygen penetration, especially at high temperatures
Before installing the column, inspect the quality of the cut end with a magnifying glass. If the cut is not square and smooth, recut the column with a column cutter until quality is satisfactory. Ensure the correct length of the inner column end part for each connection.
For ECD, the column should be conditioned with the detector end disconnected. To condition the column, follow the instructions contained in the column's box.
If column contamination is suspected, rinse the column with a small injection of pure methanol. If the baseline does not improve significantly try the following:
Warning: Never rinse or inject the capillary column with inorganic acids or bases!
Agilent Technologies, Inc., 2007. Agilent Chemstation for GC, LC, LC/MSD, CD, and A/D Systems Revision B.03.01. Hewlett Packard.
Agilent Technologies, Inc., 2008. Understanding Your Agilent Chemstation, Manual G2070-91125. Hewlett Packard.
F2 Chemicals, Ltd. MSDS for perfluoromethylcyclohexane and perfluoromethyldecalin:
PFMCH: http://www.f2chemicals.com/pdf/sds/Perfluoromethylcyclohexane - SDS20122 - ENG.pdf
PFMD: http://www.f2chemicals.com/pdf/sds/Perfluoromethyldecalin - SDS20132 - ENG.pdf
Harvey, R.W., George, L.H., Smith, R.L., and LeBlanc, D.R., 1989. Transport of microspheres and indigenous bacteria through a sandy aquifer: results of natural- and forced-gradient tracer experiments. Environ. Sci. Technol., 23:51.
McKinley, J.P., and Colwell, F.S., 1996. Application of perfluorocarbon tracers to microbial sampling in subsurface environments using mud-rotary and air-rotary drilling techniques. J. Microbiol. Meth., 26:1-9.
Plank, T., Ludden, J.N., Escutia, C., et al., 2000. Proc. ODP, Init. Repts., 185. doi:10.2973/odp.proc.ir.185.2000
Senum, G.I., and Dietz, R.N., 1991. Perfluorocarbon tracer tagging of drilling muds for the assessment of sample contamination. In Fliermans, C.B., and Hazen,T.C. (Eds.), Proc. First Int. Symp. Microbiology of Deep Subsurface. Westinghouse Savannah River Co. Information Service Section Publications Group, 7-145.
Smith, D.C., Spivack, A.J., Fisk, M.R., Haveman, S.A., Staudigel, H., and the Leg 185 Shipboard Scientific Party, 2000. Methods for quantifying potential microbial contamination during deep ocean coring. ODP Tech. Note, 28. doi:10.2973/odp.tn.28.2000
Use the GC 6890 parameters listed below to recreate the method for measuring PFMD.
Oven
| FRONT INLET (SPLIT/SPLITLESS) | COLUMN 1 |
FRONT DETECTOR (uECD) | SIGNAL 1 | POST RUN |
PAL SAMPLER AND METHOD | PARAMETERS OF PAL CYCLE |
|