Section: Appendix 6.9.1
Date: 4 March 88
AUTOMATED ANALYSIS OF AMBIENT AIR FOR SELECTED VOLATILE ORGANIC
COMPOUNDS BY A PORTABLE GAS CHROMATOGRAPH
Signed Dennis Wesolowski 3/4/88
QUALITY CONTROL COORDINATOR
Signed David A. Payne 3/4/88
TABLE OF CONTENTS
I SCOPE AND APPLICATION
II SAFETY AND WASTE HANDLING
III SUMMARY OF METHOD
IV SAMPLE HANDLING AND PRESERVATION
IX QUANTITATION CALCULATIONS
X QUALITY ASSURANCE
XI INSTRUMENT MAINTENANCE AND TROUBLESHOOTING
TABLES AND ATTACHMENTS
TABLE I ....................................... MDL Study Results
TABLE II....................................... Additional Detectable Compounds
TABLE III .................................... Troubleshooting Items
ATTACHMENT I........................... Calculation of Air Standard Concentrations
ATTACHMENT II.......................... Calculation of Events 3 (Serial Mode) Timing
STANDARD OPERATING PROCEDURE
Analysis of Ambient Air For Selected Volatile
Organic Compound By a Portable Gas Chromatograph
1. SCOPE AND APPLICATION
This method is used for automated analysis of ambient air for selected organic compounds. It is dependent on the target compounds being detect able by a photoionization detector (PID) incorporated in the Photovac gas chromatograph (GC) 10S55 instrument. THIS METHOD PRESUPPOSES THE USER HAS PRIOR OR INDEPENDENT KNOWLEDGE (ORTHAT A HIGH PROBABILITY EXISTS) OF THE IDENTITY OF SPECIFIC VOC's TO BE DETERMINED AT A SITE, SO THAT THE INSTRUMENT CAN BE CALIBRATED FOR THESE VOC's PRIOR TO USE. Unambiguous identification of presumptive identified VOC's at a site requires reference to an additional analytical system, because this Photovac instrument uses only one chromatography column and non-specific detector at a time. Possible tentative identification of other unknown VOC's with this Photovac system requires the user to have independent knowledge of the retention time order of the other VOC's, relative to the VOC's used for instrument, calibration, for the specific chromatograph column being used.
The ionization potential of a molecule must generally be less than the ionizing energy produced by the PID in order for it to be detected. Aromatic and unsaturated molecules are usually detectable by the PID. However, other types of molecules may also demonstrate detectability. Although other compounds may be measurable to this method, the following compounds are known to be detectable in air and are provided as illustrative examples. They are: vinyl chloride, methylene chloride, trans 1,2-dichloroethylene, benzene, trichloroethylene, and toluene. The sensitivity of these compounds vary but the method covers a range of 0.05 ppm (V/V) to 10 ppm (V/V) easily. The detection limits using this instrument in the automated mode have not been determine by the Central Regional Laboratory. However, D.L.'s determined manually may be used as a working model until others are determined (Table 1). Any program that requires rapid identification and quantitation of a known group of detectable analytes could use this procedure. This extends to automated determination of time weighted average values. However, modification in instrument parameters and analytical columns would need to be made to address the need at hand. A list of some compounds that are detectable by this instrument is attached in Table II.
Only personnel trained in the operation and calibration of the Photovac 10S55 GC as well as the interpretation of the raw data should use this method. Training in the use of this method is the responsibility of the Central Regional Laboratory in Region 5.
II. SAFETY AND WASTE HANDLING
Good laboratory practices should be the guide in field use of the Photovac GC, standard materials, syringes, compressed gases, and samples. The exit port of the detector will need to be vented via tubing away from the analytical area so as not to allow standard compounds or contaminants to come into the breathing zone of the analyst and coworkers. Disposal of the remains of air samples from the Tedlar bags could be done by drawing them through a large charcoal trap or other sorbant material known to retain the compounds of interest. The traps can then be disposed of in a proper fashion. Another safety consideration is handling the compressed gases. The calibration gas is a small aerosol type can which is placed in the cradle provided in the instrument and fastened with velcro strips. A special adapter is needed to connect it to the "cal in" port. The can must have a maximum internal pressure of 40 psi. If it is greater, the internal pressure sensor will interpret the calibrant as a sample.
The carrier gas (air) can be put into the internal reservoir of the Photovac GC via a special high pressure gas filling tube supplied by the manufacture. It is made specifically for the instrument and must not be altered in any way. The contents gauge should read 1600 psi when the reservoir is filled but must never exceed 1750 psi. The pressure relief valve of the filling tube should prevent this from happening. A further safety measure is a rupture disc inside the instrument which is rated at 3000 psi.
III. SUMMARY OF METHOD
Aliquots of ambient air are automatically taken via an internal pump and sampling loop controlled by timed valve events in a Photovac 10S55 portable GC. Only the eluate in the sample that contains the target compounds is allowed to pass through a precolumn which is in series with the analytical column. The components with longer retention time (RT) are backflushed from the precolumn before they can enter the analytical column and contaminate it thereby increasing analytical time per sample. This is controlled by a timed event through the system computer. The on board computer system is calibrated for each target compound in the laboratory and updated in the field. A calibrant containing one of the target compounds in air (usually the one with the longest RT) is injected by a timed cycle event to update all the RT values for compound identification and to recalibrate the response factor of all the target compounds for quantitation.
The chromatogram can be plotted and the final report printed with any values found for identified target compounds. After a specified number of cycles, a timed weighted average (TWA) will be printed out if desired. All of the timed events and cycled events depend on the target compounds being analyzed, the flow rates during the backflush made and serial mode, the type and length of the precolumn and analytical column, and the temperature of the environment under which the GC is operated. These are determined in the laboratory before field workers can begin.
IV. SAMPLE HANDLING AND PRESERVATION
Since the sample is taken through a Teflon line up to 20 feet long by an internal pump, no sample handling and preservation is necessary. This is a cyclic real time sampling. The standard involved is in a pressurized can that is directly connected to the calibration port of the instrument. Therefore no standard handling is required other than checking the fittings on the cylinder and instrument and keeping it strapped in the cradle provided.
Many compounds that volatilize in ambient air may be detectable by a PID at widely different concentration levels. Although a chromatographic column may separate compounds according to chemical and physical traits, there is the likelihood that several compounds, if present, could co-elute. If this happens to a target compound, an interference occurs. There may be only partial interference with a target compound as evidenced by an elevated baseline or peak shape. In such a case its identity and quantity may still be determined depending on the severity of the problem. The data may need to be qualified as estimated. In such a situation an experienced chromatographer should be consulted.
1. Photovac 10S55 gas chromatograph with a 4 foot of 1/8 inch Teflon analytical column with 5% SE-30 on 100/120 mesh chromosorb G-AW packing and a 6 inch Teflon precolumn with the same packing specification. There is a 1 to 8 length ratio between the columns.
2. 1 liter Tedlar bags with closures.
3. Teflon tubing 1/8 inch O.D.
4. Various sizes gas tight syringes from 0.1 ml to 2.0 ml such as those supplied by Precision Scientific Co.
5. Liquid syringes from 1.0 ul to 10.0 ul capacity.
6. Lecture size bottles of zerograde or better compressed air.
7. Flow rate measuring device or bubble meter and stop watch.
Stock standard of target compounds in methanol in 5,000 to 10,000 ug/ml range.
Neat standards of target compounds.
Cylinders of standard concentrations of target compounds in air (optional).
Cylinder of trichloroethylene or toluene in air for use as calibrant.
Methanol - reagent grade.
General operational procedures are described in the Photovac 10S50
Operating Manual. The settings and sequences of operations given in this method represent only one of many variations that could be used depending on field conditions, target compounds, and data quality objectives.
The internal reservoir is filled via the special flexible filter hose to 1600 psi. The specific instructions are in the 10S55 Operating Manual.
There are 3 flow rates to be adjusted for proper chromatography. Two of them are interactive and are adjusted in the backflush mode which is the standby mode of the instrument. They are adjusted relative to each other with the needle valve at the auxillary out port and flow controller for injector Number 1.
The flow rate measured at the detector out port and controlled by injector 1 knob should be set to about 25 milliliters per minute. The measured flow rate at the auxillary out port should be 5 to 10 ml/min greater than the detector out flow rate. If it is not, adjust it with the needle valve.
Now remeasure the flow rate at the detector and bring it back to about 25 ml/min if it has changed significantly. Remeasure at the needle valve and see if the 5 to 10 ml/min difference is maintained. This is obviously an iterative process that may take some time to adjust. It should first be adjusted in the laboratory before field work is done. All rates should be recorded.
Depress the power on button and wait for the lamp to light. Now the third flow rate can be adjusted. The third rate is measured after the first two rates are established. It is measured during the foreflush mode or serial mode in which the precolumn and analytical columns have the carrier gas directed through them in series. This rate should be the same as the rate through the analytical column during the backflush mode. Therefore, it should be about 25 ml/min and is measured at the detector out port.
In order to measure this flow rate and adjust it, the serial mode of operation must be entered. To do this for adjustment purposes put Event 3 on at 10 seconds and off at 0 seconds. This allows serial flow continuously.
When the proper flow rate has been reached using the injector number 2 knob, change Event 3 to some number greater than 10 seconds for its "off" status. This will put the instrument in the backflush mode again.
Use library number 1 and update the internal calendar and clock.
Now enter the Event settings as follows:
The instrument may now be calibrated.
An appropriate amount (1 to 5 microliters) of methonolic solution of a mixture of the six target compounds is added to a one liter Tedlar bag of clear air. Allowing about five minutes for proper diffusion, it is attached via a swagelock fitting to the probe in port on the instrument.
The concentration of each component in the air mixture is calculated using the equation in the Attachment I.
At concentrations of approximately 0.05 ppm (V/V) for vinyl chloride and 3 ppm (V/V) for methylene chloride a gain of 50 is used.
The analysis time is now set via the cycle key. The first question is the plotter delay which starts the plotter. It should be set to 10 seconds.
The analysis time is then set in seconds. Since the Events and flow rates are set for 65 oC to 70 oC temperature range, the run time should be set at about 1000 seconds to be sure to get the toluene peak. It can later be changed to fit conditions.
Now start the analysis and let it run to completion.
All of the six target compound peaks should be shown on the chromatogram. If the last peak, which is somewhat broad, is not present, the analysis run time may have been too short or the Event 3 valving was turned off at too short a time.
Setting the off time for Event 3 is described in detail in the Operator's
Manual. An example is given in Attachment II.
At the end of the analysis the report will give the peak number and RT in seconds as well as its (millivolt second) reading. The later is a measure of the integrated peak area.
All of the compounds in the standard are first stored in the library using the store key and answering the prompts displayed on the LED. The analyst must supply the name of the compounds and their concentrations in ppm. If more peaks appear than are in the calibration mixture, only name those which are known to be target compounds. Their peak numbers may be out of sequence.
After all are stored, list the library with the appropriate key. The ID numbers shown are important in the calibration sequence.
Calibrating the system is done next using the Cal key and answering the prompts for each compound. All of the compounds will then be calibrated in the computer. The computer associates the area of each peak with its concentration and with the RT for peak identification.
The library storage, listing, and subsequent calibration is fully described in the Operator's Manual.
If all peaks were seen and integrated before the end of the run, the system is ready for calibrations.
After calibration the instrument is set up to automatically sample the ambient air and continually calibrate itself through a timed cycle sequence. This is accessed through the Cycle key. Answer all the prompts including the TWA prompt if desired. All prompts and use of the key is described in the Operator's Manual.
It should be noted that whichever calibrant is used for the automatic calibration update, both the RT and the responses for all of the compounds in the original calibration will be updated based on this one compound.
If the RT is 5% longer then the original all target compound RT's will be increased by 5%. If the response is 5% less, then all responses will be reduced by 5%. The concentration of their calibrant should be the same as that used in the original calibration.
Other parmeters necessary for the identification and quantitation of the peaks are set with the Set-Up keys on the instrument. These values may be used:
on with setup speed0.5cin/min
Sens - -
upslope -15 downstope -15
+/- 10 seconds
All are explained in the Operator's Manual.
The amount of chart paper, battery power, carrier gas supply, chart pens, and calibrant should be checked. There should be enough for the period of time the sampling is to continue.
Start the analysis and the system will continue to cycle for the required time. It may be left unattended.
IX. QUANTITATION CALCULATIONS
Although the on board computer may be updated for RT and response, it should be remembered that only one calibration concentration is entered for each compound. So long as levels of contaminants are close to that level, the quantitative results should be good.
No study has yet been done by the CRL to determine how close the sample component concentration needs to be for accurate quantitation. Therefore, an arbitrary limit of +/- 50% of each standard could presently be used.
For example, if one of the target compound standards was 1 ppm (V/V) at calibration, the level for good quantitation can be expected in the range of 0.5 ppm to 1.5 ppm. If values outside of this range are computed and reported by the instrument they should be considered "estimated values."
The computer forms a ratio of the peak area of a component to its listed concentration in ppm. If the sample is in the proper RT window, it calculates its concentration by solving a simple proportional equationsuch as:
peak area standard = peak area of Sample Component
concentration of Standard (ppm) X (concentration of Sample Component)
The amount of the injection is invariant throughout the analytical cycle sequence so this is not a factor. If this value were to be changed, by altering Event 5 "on" time, a proportional change in the quantitation would need to be made manually.
A 2 second injection time is equivalent to a 330 ul injection of air.
Therefore, 4 second injections as is presented in this method is equivalent to a 660 ul injection, the DL's in Table I were determined for a 1 ml manual injection. So to use them as a working guide, they should all be increased by a factor of 1.5.
X. QUALITY ASSURANCE
Outside of initial calibration and continued calibration updates, nothing can be done during an automated run to monitor the system short of stopping the sequence.
However, a mixture of target compound should be injected through the probe in port or with the standards at the end of the sequence to determine if the system has operated properly. The concentration of these target compounds should be different from the standards and made from different solutions.
Also, all chromatograms are automatically numbered in sequence by the plotter. If any problems were noticed from the plots, for instance, a calibration update problem, a written record of the problem should be made and all analyses associated with it should be labelled as suspect.
It is very important to document all procedures done with the system and date and initial them in a field notebook. Any maintenance procedures must be listed and the reason they were needed.
XI. INSTRUMENT MAINTENANCE AND TROUBLESHOOTING
Obviously all expendable items should be checked before any automated sequence to insure that the cycling will go to completion. This includes the battery charge if the system cannot be connected to an AC source. See Operator's Manual.
Carrier gas flow should constantly go through the Teflon column since contaminants can permeate it. In this way they will be continuously swept from the column and not be allowed to become concentrated on the packing.
The GC injection port septum should remain unpierced so there should be no need to change it during the course of the analysis.
If a leak is suspected, check the column fittings inside the instrument. They should be finger tight.
The instrument will operate at its optimum if it is located in a stable temperature environment.
Table III has a list of common troubleshooting items and their probable cause and solution.
1. Photovac 10S50 Operator's Manual, Rev. A
2. "Ambient Monitoring for Specific Volatile Organics Using a SensitivePortable PID GC," Spittler, T.M.
MDL* (ppb) (V/V) in Air
Trans - 1, 2 dichloroethylene
MDL's were determined using microliter quantities of methanolic solutions of the compounds injected into a 0.5 liter gas sampling bulb. A 1 milliliter injection was made at a gain of 50 and a flow rate of 25 ml/min. They are the statistical estimates from seven replicate injections.
Additional Detectable Compounds
RELATIVE RETENTION TIMES (Relative to Benzene)
ON SE-30 COLUMN
RELATIVE RETENTION TIME
SHAPE OF PEAK
2,4 & 2,5-dimethylhexane
A - sharp symmetric peak
B - broad peak shape
C - broad and skew peak shape
LCD indicates LAMP NOT READ PLEASE WAIT, more than 3 mins after switch-on.
Source needs tuning
Tune source (see CP1, end of this section)
Instrument shuts itself down a few seconds after switch-on. Indicates low battery.
Battery needs charging.
Connect to mains, switch on, disable CYCLE if necessary and leave for 10 hours to charge.
Upon starting analysis, printer does not function at all.
Momentarily depress FEED key, releases printer.
Upon starting, plotter starts then stops immediately
Analysis time isA0"
Set proper analysis time using CYCLE key
Upon starting, no peaks appear at all.
Improper valve settings
Obtain setting list by pressing TEST + ENTER. Check settings listed.
Leaky gas fittings
Tighten all fittings but use fingers only. BE VERY CAREFUL NOT TO OVER-TIGHTEN FITTINGS ON SOLENOID VALVES.
Septum in injection port needs changing or tightening
Early peaks do appear but later peak(s) missing.
Too early backflush
Increase EVENT 3 OFF time
Unwanted late peak appears, maybe even during next analysis.
Too late backflush
Decrease EVENT 3 OFF time
Valve remains ON (as indicated by LCD).
Improper timing, ON time greater than OFF time.
Obtain LIST or check EVENT status through LCD.
Sensitivity seems too low.
Calibrant is wrong.
Leak in system.
Tighten (finger tight) all fittings. Suspect column attachment fittings and injection port. DO NOT OVER- TIGHTEN FITTINGS ON SOLENOID VALVES.
Valve timing wrong.
Check EVENTs, it is possible the backflush is too fast-see above. Also possible INJECT time too short.
Lamp is failing.
Replace with spare and try again.
Peak appears but is not recognized.
Peak not calibrated
Perform qualitative cal. as in p 37.
Increase subsequent cal. frequency.
Check that peak is listed in library and that it was assigned proper ID and plotter numbers.
Peaks appear too slowly.
Flow rate(s) too low.
Check all flow-rates.
Peaks appear too quickly.
Flow rate(s) too high.
Check all flow-rates.
Battery life is too short.
Lamp power too high.
Press TEST + ENTER to obtain status report, if SOURCE POWER is greater than 40, see CP 1.
Printer/valve cycle is very frequent.
If your cycle time is 2 mins and if you are using full print-out, the battery life will be reduced. Use external battery pack Cat# 201 or 202. Also, minimize printer format.
Instrument uses air carrier too fast.
You have a high flow rate(s).
There is a leak.
Tighten all fittings, especially on column DON=T OVERTIGHTEN SOLENOID VALVE FITTINGS!
ATTACHMENT IThe following formula is used to calculate the Concentration of each component in the air mixture:
C = m
4.1 X 10-8 x M X V
C is the concentration of compound in Air in ppm (V/V).
m is the mass of compound added to bag in grams.
M is the molecular weight of the compound in a.m.u.
V is the volume of the sampling bag in liters.
4.1 x 10-8 is the number of moles derived from the ideal gas equation, contained in a 1 ul volume.
For example, if 1 ul of a solution containing benzene at 500 ug/ml is added to 1 liter of air, what will be its concentration? 500 ug/ml is 0.5 ug/ul
C = 5 x 10-6g
4.1 x 10-8 X 78 X 11
C = 0.156 ppm or 156 ppb (V/V)
Assuming a 6 inch precolumn and a 4 foot analytical column, the time the carrier flow should be in series is determined by a simple proportion involving the elution time of the last eluant of interest. In this case, it is toluene. For example, if the RT Of toluene in 900 seconds:
6 inches = 1/8
1/8 x 1000 = 125 seconds
This is the time it takes for most of the toluene to leave the pre-column and elute onto the analytical columns. A factor of 20% is added to this time to allow for the tail of toluene to elute and some room for drift due to environmental conditions. Therefore, 150 seconds should be used as the event 3 off time.
Selected Halogenated and Aromatic VOC's
Automated Analysis of Ambient Air For Selected VOC's
0.1 ppm (V/V) to 100 ppm (V/V).
METHOD DETECTION LIMIT:
To Be Determined
Container - N/A Preservation - N/A Direct Air Sampling into Chromatographic Holding Time - N/A System.
To Be Determined
In addition to Field Controls, the following will be analyzed within each analytical run: Laboratory Calibration of Instrument (Photovac 10S55) For target compounds, a calibrant gas will go into the field with the instrument and will be used to adjust retention times for identification purposes on every alternate cycle thus bracketing the analytical runs. This will also serve as a rough check on calibration of all the target compounds.
1. Photovac 10S50 Operating Manual Rev. A
2. Various Remedial Investigation QAPPS.